159078182 447 Chevron Instrumentation[1]

INSTRUMENTATION AND CONTROL MANUAL Volume 1: Part 1: Engineering Guidelines and Appendices

CHEVRON RESEARCH AND TECHNOLOGY COMPANY RICHMOND, CA

July 1999

Manual sponsor:

For information or help regarding this manual, contact M.S. (Mike) Shigemura, (510) 242-4551

Printing History Instrumentation and Control Manual First Edition First Revision Second Revision Third Revision Second Edition

June 1989 May 1992 June 1993 July 1996 July 1999

Restricted Material Technical Memorandum This material is transmitted subject to the Export Control Laws of the United States Department of Commerce for technical data. Furthermore, you hereby assure us that the material transmitted herewith shall not be exported or re-exported by you in violation of these export controls.

The information in this Manual has been jointly developed by Chevron Corporation and its Operating Companies. The Manual has been written to assist Chevron personnel in their work; as such, it may be interpreted and used as seen fit by operating management. Copyright  1989, 1992, 1993, 1996, 1999 CHEVRON CORPORATION. All rights reserved. This document contains proprietary information for use by Chevron Corporation, its subsidiaries, and affiliates. All other uses require written permission.

Chevron Corporation

July 1999

Maintaining This Manual Instrumentation and Control Manual If you have moved or you want to change the distribution of this manual, use the form below. Once you have completed the information, fold, staple, and send by company mail. You can also FAX your change to (510) 242-2157. ❑ Change addressee as shown below. ❑ Replace manual owner with name below. ❑ Remove the name shown below. Previous Owner:

Title: Last

First

M.I.

Current Owner:

Title: Last

First

M.I.

Company:

Dept/Div:

Street:

P.O. Box:

City:

State:

Requesting Signature

Zip:

Date

Send this completed form to: Document Control, Room 50-4328 Chevron Research and Technology Company 100 Chevron Way (P.O. Box 1627) Richmond, CA 94802

CRTC Consultants Card The Chevron Research and Technology Company (CRTC) is a full-service, in-house engineering organization. CRTC periodically publishes a Consultants Card listing primary contacts in the CRTC specialty divisions. To order a Consultants Card, contact Ken Wasilchin of the CRTC Technical Standards Team at (510) 242-7241, or email him at “KWAS.”

Chevron Corporation

July 1999

Reader Response Form Instrumentation and Control Manual We are very interested in comments and suggestions for improving this manual and keeping it up to date. Please use this form to suggest changes; notify us of errors or inaccuracies; provide information that reflects changing technology; or submit material (drawings, specifications, procedures, etc.) that should be considered for inclusion. Feel free to include photocopies of page(s) you have comments about. All suggestions will be reviewed as part of the update cycle for the next revision of this manual. Send your comments to:

Page or Section Number

Document Control, Room 50-4328 Chevron Research and Technology Company 100 Chevron Way (P.O.Box 1627) Richmond, CA 94802 Comments

Name Address

Phone

Chevron Corporation

July 1999

100 System Design Abstract This section discusses the major phases of the design of instrumentation and control systems. It references other sections of the manual for detailed information on each aspect of the design process. It presents the overall picture of how the many components of an instrumentation design develop, from job scope to turnover to Operations.

Chevron Corporation

Contents

Page

110

Introduction

100-2

120

Preliminary Design Considerations

100-2

121

Getting Off on the Right Foot

122

Designing the Better Control System

123

Choosing a Control System

124

Evaluating Viable Alternatives

125

Life Cycle Costs

130

Instrumentation Design Engineering

131

Detailed Design Development

132

Design Specifications

133

Specification of Instrumentation

134

Documentation

135

Instrumentation Database

140

Construction and Startup

141

Documenting Field Changes

142

Commissioning

143

System Startup

144

Closing Documentation

100-1

100-7

100-10

July 1999

100 System Design

Instrumentation and Control Manual

110 Introduction The Instrumentation and Control Manual is intended to help engineers and designers design, construct, start up, and maintain typical Company instrumentation systems. It is intended to be used as a guide, with the understanding that no guide can replace sound engineering judgement. This section introduces the many aspects and procedures involved in designing an instrumentation system. Whether designing a small field job or a large facility, the elements of system design are similar. Protecting People and the Environment is a cornerstone of how Chevron does business, and must become an integral part of the design of any Chevron facility. With this commitment firmly in mind, a structured approach to defining, designing, and implementing a control system must be used to ensure success.

120 Preliminary Design Considerations 121 Getting Off on the Right Foot For the Control Systems Engineer, this first step is defining the objectives of the system he/she intends to install. This is a function embedded in the Chevron Project Development and Execution Process (CPDEP), and the analyses described below form an integral part of Chevron’s Policy 530. The Control Objectives Analysis (COA) is a facilitating process for defining what a control system does. The process consists of plant operators, process engineers, and control engineers reviewing plant process flow diagrams and defining the objective of each regulating device (control valve, damper, variable speed drive, etc.) on the drawing. The format of the objective is a single-sentence statement, defining what the regulating device always does to a process variable. (Example: CV-1002 maintains overhead pressure between limits.) In the case of a new process on which plant engineers and operators do not have hands-on experience, the Control Objectives Analysis should be done with the assistance of engineers and operators from other facilities presently operating the process. Finally, experience operating similar processes should serve as a basis for the COA. Similar facilitating processes define the objectives of safety shutdown systems (Shutdown Objectives Analysis, or SOA) and alarm systems (Alarm Objectives Analysis, or AOA). Only after these objectives have been defined and agreed to by Operations and Engineering, can the design of the control, safety, and alarm systems begin.

July 1999

100-2

Chevron Corporation


100 System Design

122 Designing the Better Control System With Objectives firmly in hand, the Control Systems Engineer needs to define the HOW of the control system. Items which are defined in the Control Design Analysis process include:

Chevron Corporation

•

System Geography - is control hardware in the field, in remote instrument enclosures or in a rack room in the control center? Will there be multiple sites where the operator can access the control system? Will multiple sites be peer or hierarchical?

•

System Architecture - what will be the defined and potential data transfer links to other control systems? To control and/or monitoring computers? To a Management Information system?

•

Control Architecture - how much of the control will be done by the system’s front end? Will there be advanced control such as DMC? Will there be a separate computer for advanced control?

•

Environmental - what is the Area Classification for the plant and for field sites where controls will be located? For the area where operator interfaces will be located? What are the measurable airborne contaminants for these locations?

•

Operator Interface - does the operator “see” the process via a CRT, an array of controller faceplates, or field indicating controllers? Or via a combination of two or three of these methods?

•

Operability - can the process be manually controlled in the field using a manual bypass around the regulating valve? Will there be field operators to perform this function when required?

•

Reliability - what is the minimum acceptable operating factor for the control system? What is the economic incentive for increasing reliability by a defined percentage?

•

Failure Modes - what will be the status of the control system if individual instruments fail? Do all failures result in the control system going to (or tending to) the defined Fail-Safe condition of control valves and drives?

•

Expandability - if the control system intended for a mature, well-defined process with little potential for expansion, or is this a pioneering process or first step in a multiphase project?

•

Cutover Plans - for reinstrumentation projects, the plan for cutting over from existing to new instrumentation should be a part of the control design process.

•

Hot cutovers are typically more labor intensive than conversion en masse during a planned shutdown; however, most plants opt for the hot cutover, since it allows a more gradual conversion, and results in one less “unknown” during a plant startup.

•

Urban Renewal - the amount of “re-engineering” of existing facilities (reverification of the suitability of reused field instrumentation such as orifice plates

100-3

July 1999

100 System Design


and control valves) needs to be determined at the initial design phase. This process is labor intensive if done properly - process conditions need to be field verified.

123 Choosing a Control System Digital technology now dominates the control hardware market. Electronic analog controls have all but vanished from the scene, and pneumatic controls are viable only in specialized applications where reliable electric power is not available, or where the Area Classification prevents installation of electronic controls without using elaborate cabinet or control room purging. Note current environmental regulations in California virtually prohibit the existence of Class 1 Division I areas.

Control & Operator Interface Digital electronic control should be considered as the default selection for all control systems installed in strategic facilities. With the exception of projects adding to existing control systems, solid justification must be given for deviating to electronic analog or pneumatic controls. Digital electronic control is available on a broad spectrum of platforms, from Single Loop Digital Controllers (SLDCs) through Programmable Logic Controllers (PLC) to multi-plant Distributed Control Systems (DCS). Note most SLDCs currently offered are in fact multi-loop digital controllers, with the capacity to control up to four valves from a 3-inch x 6-inch panel mounted faceplate. Operator interfaces range from panel mounted faceplates (which emulate traditional panel-mounted controllers) to color CRTs using interactive graphics for display and control of operating parameters. The industry trend is toward the use of generic color CRTs running system-specific display and control software.

Transmitters ‘Smart’ process variable transmitters should also be considered as the default standard. These instruments offer higher accuracy and reliability than their electronic or pneumatic analog counterparts, and add the bonus of remote diagnostic data acquisition and calibration checking.

Control Valves ‘Smart’ control valves are an emerging technology which offers extensive valve and process diagnostics, using the valve positioner - actuator as a sensor, or using pressure and temperature sensors embedded in the control valve body, or a combination of both. This technology should be considered on installations where maintenance access to control valves and drives is restricted.

July 1999

100-4

Chevron Corporation


100 System Design

Field Communications Communications between ‘Smart’ field instrumentation (transmitters and/or valves) and the control room are typically done over the same twisted pair of wires carrying the transmitter output or valve positioner input signal. Communications protocols range from vendor proprietary systems such as Honeywell DE- 6 Byte to multivendor open systems such as HART™. The Fieldbus communications protocol, which is being developed by an international consortium of instrument manufacturers, will offer the ability to link field instrumentation (transmitters, controllers, field indicators, valve positioners, and auxiliaries) on a multidrop power - communications wire pair. Control functions (algorithm execution) will be downloaded to the lowest possible tier of the system architecture, freeing up higher level computation capacity for running advanced control strategies.

Intrinsic Safety Intrinsically Safe (I. S.) construction is intended to prevent sources of ignition (electric sparks) in Hazardous Areas by limiting the transmission of power from nonHazardous areas and by limiting the storage of energy in field devices. Use of I.S. construction permits opening field enclosures (including transmitter and valve positioner housings) without first powering down circuits or sniffing the area to verify the absence of flammable mixtures. There is no necessary correlation between I. S. construction and Hazardous Area Classification ratings nor between I. S. construction and Explosion-proof housing construction. Because Intrinsically Safe construction severely limits the voltage and current which can be transmitted into Hazardous Areas, special attention must be given to limiting the number and type of field devices which cause voltage drops, and to the quality of field terminations. (Corrosion on field terminals can cause indeterminate voltage drops on current loops.) Final determination of whether this level of protection is appropriate for an installation should be made only after an extensive review of local Electrical and Safety Codes. The use of ‘Smart’ field instrumentation, which permits communications from a non-Hazardous area, has diminished the use of Intrinsically Safe instrumentation systems in domestic petrochemical installations.

124 Evaluating Viable Alternatives Once the scope and function of the control system is defined, the control systems engineer can focus on selecting hardware. In all likelihood, more than one commercially available system will meet the requirements of the project. Consider the following factors in evaluating viable, competing systems:

Chevron Corporation

100-5

July 1999

100 System Design


•

System Integration - determine how well a system integrates: –

–

horizontal integration, or the breadth of control hardware (regulatory and discrete control algorithms; continuous, batch, or state control operation) from a single manufacturer. vertical integration, or the depth of control hardware (transmitters and valve positioners, controllers, network interfaces, operator consoles, advanced control computers, MIS links) and software to tie the pieces together.

Avoid the entanglements of multiple sources of interface software. •

Uniformity - avoid putting “one of everything” in a control center. A strategic objective should be to have a single type of operator interface for all controls in a center; an absolute requirement should be a single operator interface for each group of plants under the control of a single board operator.

•

System Maturity - reject “sunset technology” unless you’re fitting in the last piece of a multi-phase control replacement project. Recognize that even though a manufacturer is legally bound to provide spare parts support for a limited period of time following obsolescence of a product, he has limited control over keeping competent engineers in a support function on an obsolete system.

•

Product Stability - avoid the control system which appears to be in a constant state of evolution. (These are typically a maintenance nightmare.)

•

Technical Support - investigate the communications paths available for connecting plant support personnel to technical resources at the Factory. Consider also the level of local support you can expect, especially during the first years of system operation.

•

Configurability - evaluate the magnitude of the configuration task. A system requiring special-skills programmers for initial set-up will require these same specialists, a significant expense to the plant, for every future modification. By contrast, systems which configure with a higher level Operating System can be set up and modified by plant control or process engineers, maintenance technicians, or selected operators.

•

Track Record - past performance is a valid indicator of future actions. Be wary of “born again” control systems companies with a trail of dissatisfied clients but a promise that all is changed. Check out recent references on a potential vendor’s “Happy Camper” list; be prepared for candid dialog.

125 Life Cycle Costs The quoted cost of a control system is the tip of a financial iceberg. Determine the Life Cycle cost of a system by reviewing the other cost components: •

July 1999

Training - engineers, maintenance technicians, trainers, and operators will all need training on a new system. Significant cost sub-components are tuition, time, travel, and frequency of refresher training courses.

100-6

Chevron Corporation


100 System Design

•

System Engineering - include an estimate of the cost of documentation, configuration, and commissioning services.

•

Acceptance Testing - this procedure persists in a quality-conscious world. Determine where the system will be staged, how completely application software can be loaded, and how many User representatives will be needed to give the system a thorough Factory Acceptance Test.

•

Spare Parts - request a realistic list of recommended spare parts/components from the system vendor; advise him that the cost of these spares will be factored into the cost evaluation of the overall system. (The magnitude of the Recommended Spare Parts List increases once the basic system order is placed.)

•

Redocumentation - how readily does the system adapt to self-documentation for changes in field instrumentation, control strategy definition, or configuration? Does the system use a “fill in the blanks” configuration format, or does it require proficiency in a high-level computer language?

•

Maintenance - how much maintenance effort is required to keep the system in reliable operation? Can reliability be increased (and the cost of ownership decreased?) through minor adjustments to system architecture?

130 Instrumentation Design Engineering 131 Detailed Design Development Detailed design fleshes out the control system ‘skeleton’ defined in the Preliminary Design phase of the Project. Successful installations—and thus successful projects—are rooted in the patient attention to an almost limitless number of details.

Designs Engineering Contractors The detailed design of a system is labor- and document-intensive. For this reason, detailed design work is frequently done by engineering contractors. They offer the advantage of being able to supply skilled technical personnel at short notice, and only for the duration of the project. The downside is that any technical expertise paid for by the Client and acquired by the contractor vanishes at the completion of the Project. Engineering contractors must be provided with current Chevron or plant Standards, Specifications, Drawings, and Forms, if the goals of Uniformity and Quality are to be realized.

Systems Integrators In a similar fashion, systems integrators and packaged systems suppliers must be provided with Chevron or plant specifications stating minimum requirements for controls which they provide, integration with other systems, documentation and all other information normally supplied by vendors of non-packaged instrumentation.

Chevron Corporation

100-7

July 1999

100 System Design


132 Design Specifications Design specifications are used to guide system designers. The application and the type of contract are important factors in determining the extent of design specification needed. A “typical” design specification (Model Design and Construction Specification, Section J, “Instrumentation and Controls”) is available from the Projects and Engineering Technology Group (P&ET) of CRTC. This specification is modeled to allow for a number of options and is adaptable to fit specific jobs. The design phase of a job produces the construction specification, which usually comes in two parts: a written specification and a construction drawing package. These two parts fully define how an instrumentation system is supposed to be built. Changes in specifications after a bid has been awarded can be very costly. It is therefore important to form an accurate bid package (specifications and drawings). Because an instrumentation system has many inter-related components, a thorough end-of-job review is recommended.

133 Specification of Instrumentation The specification of individual instrumentation is usually done on ISA (Instrumentation Society of America) specification forms. These forms are widely used throughout the industry, and most contractors and vendors are familiar with them. These forms are found in ISA S20 which is included in Volume 2 of this manual. The ISA form is used during design and construction and startup and by the maintenance group after startup. The ISA instrumentation specification forms include brief instructions for filling in the form. For additional guidance, this manual includes data sheet guides. Various sections of this manual also discuss instrumentation selection and specification. Consult with the Monitoring and Controls Unit of CRTC for the latest information on computer generated ISA data sheets.

134 Documentation A system designed in-house by Chevron, or designed by an engineering contractor, or designed and built by a system integrator/packaged systems supplier generally includes complete documentation for design, construction, operation and maintenance. These documents will usually satisfy the Federal and/or local safety and health legal compliance requirements for critical instrumentation. The following should be considered as minimum documentation requirements for control systems installations:

July 1999

•

Area Electrical Classification maps

•

Plot Plans & Elevations, showing location of and access to major equipment and critical instrumentation.

100-8

Chevron Corporation


100 System Design

•

Piping & Instrumentation Diagrams (P&IDs) - refer to Section 200.

•

Process Flow Diagrams, showing flow rates and conditions of pressure, temperature and chemical composition for all streams in the plant.

•

Process Control Diagrams, showing the configuration of front-end control strategies, as determined by the Control Objectives Analysis (COA), and verified by the Control Designs Analysis (CDA).

•

Advanced Control Strategy documents, including Control Narratives, describing Strategies for optimizing the Process. These would include complete DMC documentation.)

•

Logic Diagrams, defining the functionality of Safety Interlock Systems, as defined by the Safety Objectives Analysis.

•

Vessel Drawings, showing the elevation and orientation of nozzles and the maximum, normal, and minimum levels of product and/or interfaces within the vessel. (These are required for designing instrumentation bridles and ordering level instruments.) The vessel drawing shall also tabulate the following data for each level instrument connected to the vessel. Type of instrument Alarm setpoints Specific gravity of process fluid(s) Specific gravity of seal or capillary fluid Instrument span with calculations Zero suppression or elevation

Chevron Corporation

•

Instrument Data Sheets, describing the physical construction of the instrument hardware items procured for the project. (These Data Sheets must be complete enough to permit ordering instruments without additional descriptive documentation.)

•

Orifice Data Sheets, detailing flowing conditions for all orifice flowmeters. These data should be supplied by Process Engineers familiar with the plant of similar processes. Inaccurate process data will come back to haunt ALL engineering disciplines.

•

Loop Diagrams, showing the interconnections among all hardware specific components in a control loop.

•

Junction Box Wiring Diagrams, showing the layout of termination strips and their connection to Main or Branch cables or to field wiring.

•

Cable Schedules, listing the cables and pairs (or conductors) used for interconnection of instrumentation components. (In some cases, these may be combined with Junction Box drawings described above.)

100-9

July 1999

100 System Design


•

Installation Details, showing the relative position of process connections, instruments, and Utilities (power, heat tracing, vents, drains, etc.), and listing the materials used for installation.

•

Configuration Forms, describing the software or firmware (or both) used for creating Control Strategies, Operator Displays, and Reports.

•

Critical Alarm, Instrumentation, & Emergency Shutdown testing programs (procedures and frequencies for testing).

•

Indexes, for cross-referencing all of the above.

Documentation may be developed and maintained using paper or electronic media, or a combination of both. In all instances, current documentation must be available to plant Operations, Maintenance, and Technical organizations. Use of electronic documentation systems with relational data bases increases the speed and accuracy of the documentation effort, since a single data entry event generates (or edits) parameters on multiple, related data files.

Design Reviews Periodic reviews of project design documentation ensures that costly rework or reordering of material is eliminated. The frequency of these design reviews is best determined by the Project Management team, to which the Control Engineer reports.

135 Instrumentation Database The efficient handling of the vast array of instrumentation information for a project is a key issue in any instrumentation design. It is desirable to create a “Master” instrumentation database. Data need only to be entered once and changes are automatically updated for all sub-databases. Software tools are available to control instrumentation information and generate reports and schedules. An instrumentation schedule can be used to document most of the instrumentation information.

140 Construction and Startup 141 Documenting Field Changes A well-planned and designed control project minimizes the number and nature of field engineering changes. These field changes should be documented on a master markup set of Drawings maintained in the Project Engineering Office, and should include all necessary supporting documentation. Field changes must be signed off by the same level of authority as original drawings or formal revisions thereto. Prior to issuing Field Change Orders, all appropriate Management of Change (MOC) requirements must be satisfied.

July 1999

100-10

Chevron Corporation


100 System Design

142 Commissioning The commissioning process consists of verifying the proper installation, connection, and calibration of all instrumentation items on the Project. Specification ICM-MS-1586, “Instrument Commissioning,” is a guide for preparing newly installed instrumentation prior to plant startup. It describes the contractor’s responsibility for inspecting, checking, adjusting, and calibrating the instrumentation and documenting all of the work for approval by the Company. Recent trends in instrumentation have eased the burden of the Commissioning process: •

Most ‘Smart’ process transmitters can be interrogated from the control console or from termination panels in the rack room, to verify that the ‘right’ transmitter is connected to the right terminations. (Forcing the transmitter to identify itself by Tag Number is a technique for electronically ringing out a transmitter installation.)

•

The accuracy of digital electronic transmitters far exceeds that of field test equipment, and digital transmitters show no tendency to drift. Therefore, shop or field calibration of transmitters becomes superfluous. Instruments can move directly from the Tally Room to the installation site.

•

The increasing use of ‘Smart’ valve actuators or positioners permits calibration checks and recalibration of control valves from the rack room or marshaling panel. This minimizes the requirement for cycling control valves through the valve or instrument shop prior to installation.

All instrument installations should be signed off by the installer, the instrument inspector, and an Operator. OSHA regulations (29 CFR 1910) require that “Critical instrumentation” be installed and inspected by qualified workers.

143 System Startup At the system startup phase, operation of final control elements is switched over to the new control system. A major effort is required to tune control loops, especially on a grass-roots project, or loops on a reinstrumentation project which did not previously exist. Prior to attempting to tune control loops, verify that any control configurations which inhibit controller response on changes in set point have been disabled. (These features will need to be re-enabled following controller tuning.) The use of a high speed data recorder (with chart speed selectable up to 6 in. / min.) will aid in the capture of process response to changes in set point or changes in controller tuning constants. Use of high speed trending on a CRT display is acceptable, especially if the set point, process variable, and controller output can be trended on the same display.

Chevron Corporation

100-11

July 1999

100 System Design


A moderate amount of process variable damping is required for flow loops on digital controllers, to prevent the control algorithm from chasing noise on the PV signal. (The magnitude of this noise is not apparent on analog instrumentation, due to the inherent damping of inputs from volumetric capacity (pneumatic controls) or input R-C filters (electronic analog controls). When tuning Cascade Controllers, tune the ‘slave’ controller first, then the ‘master’ controller. At the conclusion of controller tuning, note tuning constants in a secure logbook, which can be used for future reference to determine is controller tuning constants have been altered.

144 Closing Documentation All field changes must be transferred to permanent documentation following turnover of the control system to the Proprietor of the project. Final, “As-Built” drawing revisions must be provided to plant Operations, Maintenance, and Engineering offices as part of the project’s documentation. In addition to the documentation described on Section 134, this final documentation project must include operating instructions, maintenance manuals, and spare parts lists for all equipment installed.

July 1999

100-12

Chevron Corporation

200 P&ID Development Abstract This section is an introduction and comprehensive guide to the planning, layout, preparation, and review of piping and instrumentation diagrams (P&IDs). It follows the P&ID development process from start to finish and is applicable to drawings of any scale and complexity. Piping, equipment, and instrumentation aspects of the P&ID are given equal weight, and considerable attention is given to the inclusion of specific elements on the drawing. Particular emphasis is given to the P&ID as a major factor in determining the efficiency, operability, maintainability, and safety of a facility. Note The foldout P&ID drawings referred to in this section are located at the end of this section.

Chevron Corporation

Contents

Page

210

The P&ID and Its Uses

200-3

220

Planning the P&ID

200-3

221

Developmental Stages

222

Layout Styles

223

Types of P&IDs

230

P&ID Symbol Standards

231

Symbol Standards—Piping and Equipment

232

Symbol Standards—Instrumentation and Controls

240

P&ID Content

241

Instrumentation

242

Piping and Equipment

250

Numbering Systems

200-16

260

Additional Information

200-19

270

P&ID Review

200-21

280

P&ID Drawings and Engineering Forms

200-24

281

P&ID Drawings

200-8

200-11

200-1

July 1999

200 P&ID Development

July 1999


282

Engineering Forms

290

References

200-25

200-2

Chevron Corporation



210 The P&ID and Its Uses Piping and instrumentation diagrams (P&IDs) are the graphic and symbolic summation of the processing aspects of a facility. Although the piping, instrument and equipment information collected on a P&ID can be found elsewhere in a facility’s design records, only the P&ID displays them in comprehensive, coherent relationship to one another. Activities in which P&IDs have a key role include the following: •

Design and design review. Defines piping, instrumentation and control systems

•

Design and construction progress. Provides a graphic framework in which to monitor design and construction

•

Construction completion check. At plant completion, the construction agency (either a contractor or Company) is responsible for delivering completed asbuilt P&IDs. This permits a piece-by-piece review of compliance with the design

•

Startup. Provides critical information during startup of a new facility

•

Operation. Provides the primary source of operating information and training aid for a plant or facility

•

New engineer training. Provides a sound example from which to design similar facilities

•

Maintenance planning and safety. Provides a framework for planning and monitoring cleanup and isolation, inspection, and similar work prior to startup, as well as future maintenance

•

Governmental communications. Provides a vehicle for communication with regulatory and governing agencies

•

Additions and modifications. Up-to-date P&IDs provide a basis for estimating, design and implementation of future additions and modifications

220 Planning the P&ID All major P&ID decisions and approval should be secured early in the design process to avoid costly changes. Because this isn’t always practical, the P&ID must usually accommodate some additions. Planning consists of determining the number of P&IDs and their arrangement, content, and style. (For more on style see Section 222.)

Safe Design Practices Safe design practices promote operating continuity, prevent upsets and alarm failures, and reduce unnecessary shutdowns. They are the foundation for employee and community safety.

Chevron Corporation

200-3

July 1999



Space Allocation To prevent overcrowding and a confusing process flow, 25% to 50% of the space on drawings should be allowed for future equipment. Disorderly P&IDs may impede the design process, and can be a liability during plant upsets—when quick comprehension is important. D-size (22-inch by 34-inch) drawings are commonly used for P&IDs because they are a manageable desktop size; however, some systems would be seriously overcrowded on a single D-size drawing. For a “grouped” P&ID (see Section 222) a longer R-size (28-inch-by-unlimited) drawing ensures that all closely related processing equipment is included on the same P&ID. The longer drawing may be avoided by separating stand-alone process, utility, or package systems and placing them on their own major equipment or auxiliary P&IDs.

Arrangement of Elements The initial arrangement of each process P&ID is submitted for owner/operator approval. These P&IDs include equipment, piping and instrument manifolds, instrument symbols (or reserved areas for them), piping runs (or reserved areas—horizontal and vertical), auxiliary systems and subsystems. Arrangement of equipment and piping should follow a sequence that flows logically across the sheet from left to right; for example, feed comes in on the left, products go out on the right. The main flow lines should be heavier than secondary process lines and utility lines, and should not double back. Lines should be spaced evenly, with a minimum of lines crossing. In general, the P&ID should be kept readable.

221 Developmental Stages To avoid the need for extensive rework and decrease the chance of error, P&IDs are usually revised and reissued several times during their development. This staged approach also makes the job more manageable and allows critical path items to proceed before all aspects of the P&ID are firm. The stages might proceed as follows:

Preliminary stage (permits P&ID layout to proceed) Stage 1 (permits facility layout and P&ID development to proceed) This revision affects the following critical path elements: site preparation, foundations, underground features, structures and pipeways, piping, platforms, ladders and walkways, power and utility supply and distribution systems, etc. This revision is typically issued for design. P&ID elements necessary for this revision include the following: •

July 1999

Major process and utility systems equipment, including driver and numbering system selection. Though not entirely a P&ID function, the estimated size and location of major equipment such as the air cooler, furnace, reactors, etc., is also required

200-4

Chevron Corporation



•

Estimated sizing and most major valving (with sizes) for major process lines and utility and relief headers (with major supply and return branches)

•

Major piping and control valve manifolds

•

Instrumentation, including the majority of field sensors, transmitters, recorders and controllers

Stage 2 (permits major instrumentation purchase and equipment fabrication) The second P&ID issue follows closely upon the first. This issue permits major instrumentation purchase and equipment fabrication to proceed, and finalizes the plot plan. In addition, work starts on detailed piping design, and relief and utility areas. P&ID elements necessary for this revision include the following: •

Selection of instruments and numbering system, and approval of all equipment and instrument connections

•

Platform layout and specification of platform attachment clips so that vessel suppliers can begin fabrication

•

All revisions to previously approved elements

•

Columns, vessels, tanks, drums, and heat exchangers

•

Connection sizes and types (flanged or stub welded), location, flange facing and ratings

•

Relief valve settings

•

Control valve failure mode: fail-OPEN (FO) or fail-CLOSE (FC)

•

Setpoints of critical shutdown instruments

•

All instrumentation. Control valve manifolds have been sized, all major instruments numbered. Shared-display mounting, board mounting, or field mounting has been specified

•

Piping. Final valving and sizing of all process lines (and major utility connections) their numbers, insulation requirements and heat tracing. All small piping and utility connections shown

Stage 3 (permits finished piping layout to be completed) P&ID elements necessary for this issue include the following:

Chevron Corporation

•

All revisions to previously approved elements

•

Equipment, piping and instrumentation. All necessary additional detail. Sizing and specifying of all relief, utility, and sample connections

•

All small piping sizes, connections, and fittings, including startup, shutdown, pumpout, steamout, washdown, etc.

•

Plot limit block valves, fully detailed or on separate drawings, as warranted

200-5

July 1999



Stage 4 (receives approval for construction) P&ID elements necessary for this issue include the following: • • • •

All revisions to previously approved elements All operating elements All maintenance elements All safety elements

222 Layout Styles Note

Figures 200-3 through 200-10 are 11x17 foldouts at the end of this section.

The three primary P&ID layouts used by the Company are the grouped equipment layout, serial equipment layout, and geographical layout.

Grouped Equipment Layout This layout style emphasizes processing interrelationships between closely associated, often interactive equipment. It is used for plants where several feed/product streams are processed concurrently, such as on-plot process facilities (the major manufacturing areas of plants, as opposed to off-plot, or support, areas), utility generation facilities, water and waste treating facilities, etc. (see Figure 200-3). To keep drawing lengths manageable, the facility is divided into essentially independent functioning elements. For a large processing plant these elements might include furnaces, reactors, distillation columns (towers), compressors, etc., that can be conveniently grouped on separate drawings. On the separate drawings, lines handling lighter products are drawn along the top, lines handling the heavier products along the bottom.

Serial Equipment Layout This is a convenient layout for plants with a single major process stream that is acted upon sequentially at essentially independent stations, for instance, a packaging plant or production facility (see Figure 200-4). The P&IDs for plants laid out in this style can be many feet long when on a roll or multifold paper. When properly laid out, these may be broken down into individual drawings to more easily fit desktops or for inclusion in record books. Each segment holds usually one, sometimes two processing elements. Serial-style P&IDs often have equipment information blocks along the top, process gas, relief, vent and flare headers just below, the equipment in the middle, interconnection lines just below the equipment, and pumps and compressors along the lower edge.

Geographical Layout This layout is used for collections of independent processing elements that are not linked by process relationships, such as tankfields, utility distribution systems, plot limit manifolds, and interconnection diagrams. A roughly geographical layout is often the most logical way to present them. (See Figure 200-5).

July 1999

200-6

Chevron Corporation



223 Types of P&IDs Main P&IDs The main P&IDs show process flow, mechanical equipment, and instruments and controls. For small plants the main P&ID is all that is required.

Major Equipment P&IDs Major processing equipment such as compressors, reactors, furnaces, treaters, and refrigeration systems are often placed on separate P&IDs (See Figure 200-6). This accomplishes the following: •

Provides the space to show the interrelationships of complex mechanical elements with their instrumentation and supporting supply systems

•

Shows precise location details, particularly for critical temperature points

•

Unclutters the main P&IDs

Auxiliary P&IDs Equipment not directly in the main processing stream is often referred to as auxiliary equipment. Examples are seal, flush, and purge systems; lube oil, hot oil, and oil mist systems; and glycol heating systems (see Figure 200-7). These may be placed on separate P&IDs to reduce crowding on the main P&ID or when they serve equipment on different P&IDs. When small, they may be combined on a single drawing with other auxiliary systems. When auxiliary equipment is supplied assembled in a “package unit” from a vendor, it should be depicted within a dashed-line box, with attention given to the following Company/vendor interface areas: •

Equipment supplied at the boundaries. Otherwise, pickled pipe may arrive without mating flanges, the pipe material may be wrong, or both Company and vendor may supply block valves

•

Instruments. Otherwise, both parties may supply duplicate sets, or Companysupplied instruments may not fit vendor-provided connections

Plot Limit Block Valve Manifold P&IDs This is a type of geographical layout (see Section 222). In major petroleum and petrochemical processing facilities individual plants or groupings of plants are set up as isolable entities. A single major assemblage of block valves at the end of a central pipeway, the plot limit block valve manifold (plot limit manifold), ties the individual plant headers into an interconnecting pipeway system serving other facilities (see Figure 200-8). With a few exceptions (primarily underground lines) all lines in the plant pass through the plot limit manifold. This facilitates supervisory review of plant isolation prior to a major planned shutdown. For small plants the plot limit block valves may be shown on the process P&IDs themselves. For larger plants a plot limit manifold drawing is prepared.

Chevron Corporation

200-7

July 1999



Interconnection Diagrams These specialty drawings show, on one drawing, the relationship between control systems located in different plants.

Utility Distribution System P&IDs These are usually laid out geographically to preserve the sequencing and relative locations of all elements (see Figure 200-5). In mid-sized plants, several utility systems (steam and condensate, all gases, water, etc.) may either be layered on a single drawing in separate well-defined strips or superimposed. To reduce clutter, only the tie-in portions of utility systems should be shown on the main P&IDs. These should include all valving and instrumentation associated with the control or isolation of the processing equipment—checks, block valves, flow indicators, etc.; the utility P&IDs themselves show little valving. The tie-ins should be labeled with the utility P&ID line and drawing numbers, and, if desired, their service. Small, in-plant utility facilities are usually shown on their associated utility P&IDs—instrument air dryers, fuel gas knock-out drums (separators that remove entrained water from the gas), condensate dryers, etc. Larger utility supply and processing systems are usually shown on separate process P&IDs—water treatment plants, boiler plants, etc.

Relief System P&IDs Figure 200-9 shows a typical geographical layout for a relief system. Relief valves and bypasses are not shown here, but are included on the main process P&IDs—the usual practice for process operations information. All calculated relief loads should be recorded on this drawing, since they are not always found in the plant design records. Relief system P&IDs are very helpful in determining relief system modifications when adding major equipment in the future.

230 P&ID Symbol Standards The following drawings show the P&ID symbols commonly used in newer plants. These symbols are derived from the nationwide Instrument Society of America (ISA) standards (see Standard Drawings and Forms): • • • •

ICM-EF-824A, Standard Piping and Equipment Symbols ICM-EF-824B, Standard Instrument Symbols ICM-EF-824C, Standard Logic and Instrument Symbols ICM-EF-824D, Guidelines for P&ID Presentation of Level Instrumentation

EF drawings may be adapted and condensed to a single sheet for a major facility.

July 1999

200-8

Chevron Corporation



231 Symbol Standards—Piping and Equipment When making plant additions or modifications, it is sometimes best to continue using the existing symbology familiar to plant personnel. However, new facilities should use modern ISA symbology to ensure clear communications with installation personnel. An ISA symbol exists or can be adapted for almost any instrument. Equipment symbols are another matter. It is often necessary to elaborate on ICM-EF-824A for complex machinery such as compressors, multistage pumps, and materials handling equipment. All project-specific symbols and other unusual symbology should be clearly recorded either on the project P&ID symbol drawings (if incorporated as project drawings) or in the notes column of the P&IDs themselves. Such symbols must also be used consistently throughout the project. This is vitally important because P&IDs may be the only drawings available to those unfamiliar with a particular project or facility, such as engineers involved in facility additions and modifications. Often what is regarded as a universal “standard” symbol by one organization is found to be different elsewhere.

232 Symbol Standards—Instrumentation and Controls The following should be agreed upon before much work is done on the P&IDs for a project: • • • •

A standard for continuous modulating controls A standard for process safety and sequencing logic How to document P&ID special symbols The degree of details to be shown on the P&ID

Continuous Modulating Controls Modulating controls indicate and control variables that can change continuously over a range of values. ISA Standard S5.1 (see Appendices) is the preferred standard.

Process Safety and Sequencing Logic Variables for process safety and sequencing logic can normally assume only two states; a pump is either on or off; a temperature either is or is not too high; a burner either is or is not lit; a filter is or is not ready to be backwashed. The logic symbol standard used most often is ISA Standard S5.2, (see Volume 2, Industry Codes and Practices). These symbols are most suitable for representing binary process logic, thus for documenting most safety systems. They do not easily represent sequencers such as drum programmers which have many output states. In most cases the sequencing logic will be complex enough to require separate functional logic diagrams. The S5.1 symbols connect the individual instrument symbols to a box labeled with the name of the logic system. For example, a boiler P&ID may

Chevron Corporation

200-9

July 1999



show a box labeled BURNER MANAGEMENT SYSTEM. There may be more than one logic box shown on the P&ID to represent different logic systems (see ISA S5.2, Appendix A, Figure 1). The box should direct the reader to a logic document that is recoverable by future users of the P&ID. This logic document might be a drawing, such as that shown in S5.2, Appendix A, Figure 2. Note that this drawing is tied to Figure 1 by the instrument balloons on the interlock system in Figure 1 and adjacent to the logic in Figure 2. Word descriptions can supplement or replace logic drawings such as that provided in S5.2, Appendix A, Section 3.1. Figure 200-1 is a type of word description called a control philosophy. This is a very effective way to communicate complex or simple process control schemes. Fig. 200-1

Control Philosophy Dirty Water Tank

EQUIPMENT:

T-4

REFERENCES:

P&ID F-40001

PROCESS DESCRIPTION:

Tank T-4 is the dirty water surge tank for the produced water plant. It is 38 feet diameter and 24 feet high. It receives produced water from the FWKO vessels, coalescers, and other locations. This water may contain some oil that needs to be skimmed. From this tank the liquid goes to the flotation units which further separate the oil from the water.

PROCESS CONTROL Level Control—

Level control is very important in this tank. It is controlled by LIC-T4 at 18 plus or minus 2 feet. There needs to be a constant head in the tank so that relatively constant flow can be supplied out of the tank to the flotation units. (See the control philosophy of the flotation units.)

PROCESS UPSETS:

July 1999

High Level—

High level could occur if there is a block in the outlet line, the controller or control valve fails, or more water is coming into the system than can be handled. LAH-T4 will alarm at 22 feet in the control room if this happens. (The operator may then decide to manually control the flow out of the tank with the bypass valve to lower the level.) It will also close FV-T5 to keep from transferring to T-4.

Low Level—

The level may fall below the control range if there is a controller or control valve failure or there is a leak. In this case LAL-T4 will alarm at 7 feet in the control room.

Emergency S/D—

LV-T4 closes during an ESD #1.

200-10

Chevron Corporation



Control philosophies (when used) are an integral part of the P&ID. They can be placed in an expanded note section of a P&ID or on separate P&IDs. Simpler safety and sequencing logic can be shown entirely on the P&ID using the symbols of ISA Standard S5.1. For example, a low-level shutoff for a tank valve actuated by a level switch may be depicted without the need for a separate logic diagram.

Special Symbols Any special symbols should follow the rules in ISA S5.1 and S5.2, and be defined on each drawing.

Degree of Detail ISA S5.1 identifies three levels of detail, depending on user requirements, as follows: •

Simplified loop. See ISA S5.1, Section 6.12, Figure 1. Simplified symbolism and abbreviated identification identify the principal measurement and control functions. Process control diagrams often use simplified loops

•

Conceptual loop. See ISA S5.1, Section 6.12, Figure 2. Functionally oriented symbolism and abbreviated identification show the control function but not the implementing hardware. Advanced process control diagrams and P&IDs intended primarily for the process operator normally use conceptual loops. Detailed loops are frequently shown on additional drawings

•

Detailed loop. See ISA S5.1, Section 6.12, Figure 3. Detailed symbolism and more complete identification show the type of hardware and kinds of signals. Detailed loops are often needed by the plant control engineer and the design, control engineering and maintenance staffs. For operator training, the conceptual loops must frequently be shown on additional drawings

240 P&ID Content 241 Instrumentation The development of process and equipment control schemes, and the placement of minor instruments are discussed in this section and depicted in Figure 200-2.

Process Control Schemes Process flow and control diagrams generally do not show equipment controls, safety controls, and miscellaneous minor instruments. When incorporating process controls on the P&ID, the plant designer makes hardware and software choices that were unavailable to the process control designer. These choices can affect the function of the process controls and should be reviewed with the process control expert.

Chevron Corporation

200-11

July 1999



Fig. 200-2

Development of the Instrumentation Portion of P&IDs

Equipment Control Schemes If they are very simple, major equipment controls should be shown on the process P&ID. Otherwise, they should be shown in detail on a separate equipment P&ID and referenced on the process P&ID. Equipment control schemes should be developed in coordination with equipment vendors, and Company and design agency equipment control specialists. The resulting P&IDs should be reviewed with an equipment control expert. Review and approve equipment controls that are completely determined by the Vendor in packaged systems. With packaged systems, a Company instrumentation expert should be consulted before it is too late to make changes.

Minor Instrumentation The P&ID designer is responsible for putting all minor instrumentation on the P&ID, including the following: • • • • • • • • •

July 1999

Locally mounted pressure gages Remote temperature indicators Local temperature indicators Local level indication Remote flow indicators Alarm and shutdown systems Pressure sensors for automatic pump starters Toxic and combustible gas monitors Transmitter output indicators

200-12

Chevron Corporation



The following paragraphs give guidance on the proper application of each of the minor instruments listed.

Locally Mounted Pressure Gages Pressure gages should be installed on the following process equipment and piping locations to monitor operation and performance: •

Discharges of pumps and compressors

•

Vessels and the bottom vapor space of columns

•

Near the process connection for nonindicating pressure transmitting instruments

•

Furnace fuel oil, fuel gas and atomizing steam branch headers

•

Furnace draft. A single draft gage should be manifolded to the inlet of the convection section and to a position below the stack damper on each furnace

Pressure test points consisting of a process connection with a plugged valve are located in process equipment and piping, as follows: • • • •

Near the inlet and outlet of all packed vessels and columns At all indicating pressure transmitter instruments Inlets and outlets (both shell and tube side) of each heat exchanger and reboiler Inlet and outlet of each air cooler

Remote Temperature Indicators (Thermocouples and Resistance Temperature Devices [RTDs]) Remote temperature indication should be provided, as follows, on most process equipment and piping:

Chevron Corporation

•

Columns. All inlet and outlet lines

•

Vessels. All inlet and outlet lines expected to have different temperatures

•

Fired heaters. Inlet line and outlets from each pass, header pass points from the convection to the radiant section, on the tube wall as recommended by the furnace supplier (a minimum of three per pass), and on the stack just ahead of the damper

•

Process stream junctions. Downstream of the junction point of all important process streams

•

Coolers. All liquid product inlets and outlets

•

Temperature controllers and transmitters. These instruments should have an additional thermocouple and thermowell separate from the controller or transmitter. Instruments on high pressure piping and reactors may use a common thermowell

•

Orifice flow meters. For heavy hydrocarbons. Used to estimate viscosity and make flow corrections from fluid temperature changes

200-13

July 1999



•

Parallel piping lines. Temperature transmitter thermowells should be installed in both lines, and the sensing bulb for the transmitter in one line. The installation should permit transfer of the sensing bulb to the other well

•

Process compressors or blowers. Inlet and outlet lines. One point is required on the combined inlet and one on the combined outlet of compressors or blowers in parallel on the same service

Local Temperature Indicators (Dial Thermometers) Local temperature indication should be provided for process equipment and piping where required for manual field control. Such temperature indicators should measure outlet water temperature from all condensers or coolers, discharge of all blowers, discharge of each compressor cylinder, and lube oil and water for pumps, turbines, compressors and similar mechanical equipment. •

Heat exchanger thermowells. Thermowells should be located at the inlets and outlets of heat exchangers (shell side and tube side) that don’t have remote temperature indicators. If a thermocouple point or dial thermometer is present temperature test points are not needed

•

Compressor temperature alarms and shutdowns. High discharge temperature alarms are necessary on each cylinder of a main reciprocating compressor and, frequently, on other compressors as well. Thermocouples and thermistors may be used for this service; filled thermal systems should not be. Because high temperatures must be detected at very low or zero flow, the sensing point should be either in the compressor nozzle or immediately downstream of it

Local Level Indication Local level indication should be provided for all columns, vessels and drums to determine total and interface (if any) level. •

Gage glasses. Gage glasses are preferred for local level indication, with the following exceptions: – – –

July 1999

At pressures above 900 psig, except for steam or water service Where they are unsuitable for the process fluid (dirty stocks that will coat the glass, etc.) Determine if a gage glass for an elevated vessel will be readable from grade and, if not, include an additional indicator at grade

•

Displacer-type level transmitters. When level glasses cannot be used, include a displacer-type level transmitter with a local receiver gage. Usually, any level alarm should be taken from this transmitted signal. This transmitter should be separate from the level controller loop

•

Differential pressure level transmitters. Use with a flange-mounted diaphragm capsule when neither a gage glass nor a torque-tube displacement type instrument is suitable

200-14

Chevron Corporation



•

Pyrometer-type level sensors (ram’s horns). Use for heavy oil columns (e.g., atmospheric and vacuum columns) if approved by the operations representative. No fewer than five should be used

•

Automatic tank gages. Tanks used for inventory control should have automatic tank gages readable from the ground, and level transmitters that display in the central control house (if there is one). Heated tanks and tanks storing product at above-ambient temperature should have remote readout of spot tank temperatures. If there is an existing tank gaging system, a project decision should determine whether automatic tank gages should read out on it

Remote Flow Indicators All feed, product and utility lines should have remote flow indication.

Alarms and Shutdown Systems •

Shutdowns and interlocks. Automatic shutdown and interlock systems (see Section 1200) prevent the startup of equipment or portions of the plant when operation would be a serious hazard. Alarms may be anticipatory or activated at the time of shutdown, and are displayed on the central control room alarm system

•

Alarm and safety setpoints. Setpoints for alarms and safety trips should be recorded on the P&ID if they require setting in the field

•

Low flow shutdowns in high energy systems. When a centrifugal pump is injecting liquid into a high energy system, shutdown of the pump can cause disastrous reverse rotation if the check valve fails to hold. In such cases, a low flow reading on the feed meter should close a control valve to prevent the backflow

•

Computer communication. The process computer (if used) should monitor alarm and shutdown status

Pressure Sensors for Automatic Pump Starters In services where continuous flow is critical, drivers for spare pumps should automatically start on loss of flow from the prime unit.

Toxic and Combustible Gas Monitors Toxic and combustible materials require special attention. Facilities handling hydrogen sulfide (H2S) require an ambient H2S monitoring system. Monitoring stations should be judiciously located around equipment handling high H2S concentrations.

Transmitter Output Indication Blind transmitters (except level) should have at least one indicating gage on the transmitted signal. If a control valve is associated with several transmitted variables (directly or indirectly as with flow and level indicators on the same stream) the gages should be readable from the manual bypass valve. Gages for split range instruments should be readable from each valve. A gage is not required for level

Chevron Corporation

200-15

July 1999



transmitters if the gage glass can be read from the control valve. The same applies to any indicating transmitter that can be read from the control valve.

242 Piping and Equipment Physical Size and Mechanical Design Information The physical size and mechanical design information almost universally found on P&IDs for all major processing and production facilities is as follows: •

Piping elements. Nominal pipe diameters and sizes of valves, flanges, reducers, connections and miscellaneous and special fittings

•

Columns, vessels, tanks. Internal diameter(s) (“ID”), seam-to-seam height(s) or length(s) and equivalent boot and dome dimensions

•

Relief valve setpoints

Additional Mechanical and Process Information Major processing plants usually control operating conditions by varying feed stream composition, throughput, heat input, etc. As a result, relief valve setpoints are usually the only mechanical design information shown on the P&IDs. By contrast, additional mechanical design and (sometimes) process information (such as design temperatures and pressures, duties, horsepowers, speeds, capacities, and throughput) are shown on production facility P&IDs, because actual conditions can be quite different than anticipated. Some operators of major processing facilities now request expanded equipment information on their P&IDs. This information can be of considerable help in operations, and in estimating and designing additions and modifications.

250 Numbering Systems Plant, Equipment, and Piping Numbering Systems There are so many systems in use throughout the Company that it would serve little purpose to discuss more than a few general principles here. API RP 14C, Table 2.2, “Component Identification” is another system of line numbering and equipment identification.

Facility Names and Plant Numbering Systems Major facilities (both upstream and downstream) are generally named for their location. Small facilities such as small producing gas plants and small stand-alone asphalt plants are not further subdivided. Larger processing facilities are broken up into distinct plants. These plants are generally named for their function (crude unit, gas dehydration, boiler plant, effluent treating plant, etc.). They are usually also assigned a number. During construction and later, during maintenance, this provides a rough-cut way of segregating, by construction area, the hundreds (and sometimes

July 1999

200-16

Chevron Corporation



thousands) of items of delivered equipment. Local management should be consulted in determining plant numbers for a new project. Operations approval should be obtained for numbering systems (plant, piping, equipment, and instruments) at the beginning of a new project. Once drawings and specifications are issued for quotation, the cost of changing numbering systems is surprisingly high. A confirming letter or memorandum emphasizing the importance of this decision is helpful.

Equipment Numbering Systems Where plant numbers are used, they should be incorporated into equipment numbers, and, often, into the instrument numbers. Where plant numbers are not used (such as for offshore platforms) many prefer that the instrument numbers relate to uniquely numbered equipment. This method is used on platforms built to API RP 14C to associate safety devices with the equipment they protect. Some facilities incorporate a plant number into the equipment number and associate instruments with equipment. Most facilities also use alphanumeric systems with a letter prefix that indicates the type of equipment involved. For instance, the prefix MAF designates a 7-tray glycol contactor for an offshore platform (see API RP 14C). The same equipment in a downstream major processing plant would be denoted by a C for column. The prefix system allows the number series to be restarted for each type of equipment, so that numbers can usually be limited to two digits. Equipment is numbered serially or in decade steps for major equipment. Thus, for plant 20, three sequential columns might be numbered 2010, 2020, 2030. P-2021 might be the reflux pumps for column C-2020. Skipped numbers are acceptable in a system such as this.

Instrument Numbering Systems Assignment of instrument numbers must be coordinated with all design agencies that are developing P&IDs or subsections of P&IDs. A unique identification for each instrument is assigned in accordance with ISA Standard S5.1 (see Appendices), as follows: 70-FIC-101 Reading left to right, the first element of the code is the plant number. This is normally omitted on the P&IDs, but it is included in the instrument number for other purposes such as ordering, I. D. tags, etc. The letters that follow represent the instrument type according to ISA Standard S5.1. The final element is the loop identification common to all instruments and components in a loop. If possible, loop identifiers should be in order of their positions upstream in the process flow; that is, feed loops have the lowest numbers, and product and final effluent loops the highest.

Chevron Corporation

200-17

July 1999



Loop Identification Systems For additions, the existing instrument numbering system is usually extended to include new instruments. For new construction, the following methods are available: •

Functional orientation. This system was developed for large distributed control systems. Its restricted identifier size accommodates electronic databases. It is organized as follows: –

– – •

Hardware orientation. This system was developed when individual controllers were mounted in control panels. Seven blocks of numbers are used: – – – – – – –

•

100 to 899. Except for safety relief devices, major instrument loops take their identifiers from this block. An alphabetical suffix is added where more than one of the same component is present in a loop, as with split range control valves. Temperature points that are part of the control display system use this block 900-999. Safety relief devices, relief valves and bursting disks have 3-digit groups from this block Four-digit groups. Minor instruments have 4-digit identifiers

100 to 399. Loops monitored or controlled from the control center 400 to 499. Field controlling, recording and indicating loops, including dial thermometers 500 to 599. Field contacts for alarms 600 to 699. Pressure gages 700 to 799. Level gages 800 to 899. Board temperature points, including test wells 900 to 999. Relief valves and bursting disks

Major equipment orientation. This system provides much information to the plant operator. However, it is unwieldy and is not recommended except where already used

Line Labeling Systems Lines generally require four to seven identifying elements. Symbology depends on the facility and organization involved. These elements are as follows:

July 1999

•

Plant number

•

Service. Also called a “line identification letter” (i.e., process, instrument air, caustic)

•

Line number. Often (and best) a separate unbroken series restarting with each service designation (critical for large jobs to keep number length reasonable)

•

Nominal line size

•

Piping classification. Service classification, service, etc. Specifies pipe, valves and fittings

200-18

Chevron Corporation


•

Insulation

•

Heat tracing


See Standard Drawing ICM-EF-824A. For major processing plants in which many additions and changes are anticipated during the design phase and later, sequential line numbering that follows process and utility flow is recommended.

Processing Facilities To minimize line numbers and better indicate process relationships, line numbers at processing facilities usually run unchanged from one piece of equipment to the next, including branches to multiple or similar pieces of equipment such as a pump and its spare. Also, the number is not changed for a change in pressure or materials.

Producing Facilities COPI and some producing organizations change line numbers when the pressure classification (piping classification) changes, on branches to multiple or similar equipment, and when a materials change is required. Their requirements differ from major processing plants. They have many fewer piping classifications and much larger pressure changes that need to be clearly indicated. COPI assigns a different series of sequential numbers for each service, and has a very organized numbering procedure.

Line Schedules In larger plants (particularly those constructed by large contractors) it is necessary to keep track of assigned line numbers using line schedules. Otherwise, accidental reuse of the same number would surely occur. In addition, line schedules are required by some governmental agencies for permitting. A line schedule often becomes a valuable control document summarizing all piping design criteria, including the following: •

Design/process information critical to line design and specification (service flow, pressures, temperatures, viscosity, density, pour point, etc.)

•

Resulting design information (pipe size, piping classification, insulation, heat tracing specifications, etc.)

•

Line connection points (to/from)

260 Additional Information Miscellaneous Elements A variety of components, details, and descriptions are shown on typical P&IDs, including the following: •

Chevron Corporation

P&ID revisions. Changes should be clearly identified by sequentially numbered symbols, such as “diamonds,” to pinpoint the location of each revision on the drawing. These changes are listed in the revision block or on a

200-19

July 1999



separate sheet issued and filed with the P&ID. Avoid the use of vague terms such as “general revision” •

Line classification changes. Line classification changes may occur where different streams join, for instance: – –

Where utility piping ties into process lines with a higher corrosion rate, temperature or pressure Lines entering and leaving a vessel where processing changes occur that require additional valves of a higher class than dictated by conditions within the line

P&IDs should be carefully reviewed to ensure that all line classification change symbols are shown. This is particularly important when the plant is modeled, because there are no piping layout drawings (plans and elevations). The model, piping isometrics (spool drawings) and P&IDs are then the only records of line classifications. •

Level, alarm, and shutdown setpoints and operating ranges. Although often shown on other drawings, such as vessel drawings (see Section 134) and the level instrument piping drawings, these should be shown on the process P&IDs when they are critical to the safe or proper operation of the process

•

Flanges. Most often, all 2-inch and larger equipment connections and valves are flanged, but there are some which are not, such as welded stub nozzles (a welded line-to-equipment connection often used on high, hard-to-reach vessel connections and between stacked exchanger pairs) and weld-in valves used in higher pressure services To distinguish welded from flanged connections, either show all the flanges or stipulate that all 2-inch and larger connections are flanged except where a symbol (sometimes “WE” for weld-end) is placed adjacent to the connection. The P&IDs may then be used as a blinding control drawing during shutdowns, and to indicate flanged connections to the design draftsman. When different from the piping classification, flange sizes and rating are also shown on the P&IDs.

July 1999

•

Entering and leaving line designations. Careful labeling of lines as they enter and leave the P&ID allows good continuity from one P&ID to another. Labeling includes the reference P&ID drawing number, the line identification (noted along the line or enclosed in a rectangular “tag” or “balloon”), the to/from equipment number, and a service description

•

Equipment internals. P&IDs should include a graphic representation of vessel internals whose function (or lack of it) may impact the operation of the facility. Examples include column and vessel internals, gas and liquid distribution and segregation mechanisms, internal level floats, heat exchanger overflow weirs and tubes, furnace tubes and dampers, and many more. For complicated equipment such as reactors, a separate major equipment P&ID is often prepared to document critical bed temperature points, process gas flow path, quench feed points, etc.

200-20

Chevron Corporation



•

Detail reduction. Level, pressure, and flow instrument piping details on process P&IDs (e.g., testing and maintenance isolation valving and connections) can be greatly reduced by using auxiliary instrumentation drawings (see Figure 200-10). Repetitive (and sometimes complicated) vent, drain, and sample system details may also be included on a separate schedule

•

Reference Drawings. A reference drawing block is often incorporated along the lower edge of the P&ID. It lists major associated drawings—plot plans, piping layout, electrical, etc., with type of drawing, item or area covered, and drawing number. This can help locate associated drawings that may be scattered among hundreds of project drawings. When modifying a facility it is equally important to add new reference drawings to the drawing reference blocks

•

Drawing Titles. Many styles are used throughout the Company for drawing titles. Titles may contain helpful information on the type of drawing (P&ID, instrument, piping, etc.) the item or area covered, the project title or plant name, and the name of the facility or division. Depending on the organization, title blocks may be sequenced differently or omit some items

Information Not Shown on P&IDs The following information is usually not shown on P&IDs: •

Pipefitting details are not shown, except for reducers. These details include hydrotest high- and low-point vents and drains, elbows, tee’s, other joints, and (sometimes) unions

•

Pipe supports are not shown. These supports include hangers, anchors, guides and pipe expansion loops

•

Structural information is usually not shown. This information includes most support structures, platforming, ladders, etc.

•

Electrical information is not shown, except for special controls such as threeway switches

•

Instrument piping/tubing is not shown, except for level instrumentation piping. In major processing facilities, level piping valves and details are placed on auxiliary P&IDs or piping drawings, leaving only a skeletal piping outline on the main process P&IDs. By contrast, producing organizations usually retain valving and level piping details on the main P&IDs

270 P&ID Review Various P&ID review techniques may be used to ensure full consideration of facility design, including safety, operability, maintainability, reliability, etc. These techniques include the following: •

Chevron Corporation

Parallel element review. A comparative analysis of all occurrences of a single detail to verify design consistency. The types of review should be discussed and agreed upon

200-21

July 1999



•

Line-by-line review. A stepwise review of small related groupings of design elements (such as all the elements associated with a single line). It is used primarily to obtain owner/operator approval of the design

•

Design practices review. Focuses on all features ensuring operating continuity (the foundation for employee and community safety). Covers piping, equipment, and instrumentation and control systems that protect against upsets and failures

•

Hazard Assessment, Mitigation, and Hazard Abatement. An evolving set of lengthy, formal techniques (both qualitative and quantitative) being adopted on a national scale. Primarily intended for the analysis of new or untried processes or facilities (or elements thereof) where there is a potentially significant hazard to employees or the community.

Parallel Elements Review This technique involves the comparative analysis of all occurrences of a single design element on all P&IDs. These elements are easily located even on complex P&IDs by the design engineers or an individual familiar with the type of facility. Parallel element review is fast, thorough, and very revealing of errors and omissions. It works well with groups and for individual review. Optimally, only one element at a time is selected for analysis, because several aspects of each element may need examination simultaneously, such as use, need, aptness, and engineering rationale. One might take several passes through the P&IDs to review the following equipment connections: vessel drain size for each vessel, flanged versus stub weld connections, all flanged thermocouple locations, piping classification changes for lines tying into vessels in corrosive service, etc.

Line-by-Line Review (Operational and Maintenance Review) Conducted by the design/operations team, line-by-line review is the stepwise review of related groupings of design elements. Groupings may comprise an individual piece of equipment or a line with its valving, drains, sample connections, piping classification, insulation, heat tracing, etc. The review follows the general path of process flow. This review is often used to obtain acceptance or approval by operations. It is logical and sequential (often a yellow marking pencil shows progress, a red pencil additions or changes). However, since the elements in each grouping usually have different functions, the immediate impact of a parallel element review is lost. Further, the process can be tedious, particularly when involving many lines with repeating features. Suggestions for conducting a line-by-line review:

July 1999

•

Break up sessions with parallel element review to dispose of the most repetitive elements

•

Use interactive role playing, in which the operator walks through all steps needed to start up, run and shut down the plant. Design engineers, process representatives, etc., point out erroneous or missing piping, equipment and instrument elements

200-22

Chevron Corporation



Design Practices Review This is employed, particularly in larger, more complex facilities, to confirm that good design practices have been used. It is best conducted using finished P&IDs. Its objectives are as follows: •

Identify elements whose failure could, through malfunction or human error, endanger operating continuity, employees, or equipment

•

Confirm that good design practices have been incorporated, including safety systems, mitigating systems, alarms and shutdowns, etc., to eliminate or reduce the consequences of failure to acceptable levels

•

Estimate the potential for alternate failure modes

•

Determine whether the consequences of failures constitute an acceptable risk

•

Modify (or provide additional) design features or safeguards to reduce consequences to acceptable levels

Design practices review focuses on the active elements of the facility—instruments and controls, pumps and drivers, furnaces, compressors, utility supply systems, etc. These elements are examined in “brainstorming” sessions that consider both historic and unusual equipment and control system failure modes. Techniques of inquiry include the “what if” method and the related, more powerful Hazard and Operability Study Method (see the American Institute of Chemical Engineering (AIChE) “Hazard Evaluation Procedures” and, in this manual subsection, Hazard Assessment). Usually an abbreviated, verbal run-through of these techniques resolves concerns or identifies problem areas for later evaluation.

Hazard Assessment When used in the early design phase, hazard assessment review techniques uncover necessary changes that can be made at minimum cost. Later, errors may be extremely costly to correct. Almost without exception a representative of the owner/operator/client must be present at every review meeting. Other interested organizations include process, designs, maintenance, operations, safety, reservoir engineering, plant or drilling foremen, area superintendent, other management, etc. Hazard assessment may also include mitigation and abatement techniques such as the Hazard and Operability Study, Failure Mode and Effects Study, Fault Tree Analysis, SAFE charts, Blast Effect Analysis, Atmospheric Dispersion Study, Radiant Heat Study, etc. These can be quite costly, particularly when full documentation is required. They are used when mandated by federal, state or local laws and regulations or as judged appropriate by the responsible manager. California and New Jersey have passed legislation requiring the application of these techniques to plant processes and equipment for stipulated toxic or flammable materials whose catastrophic release could impact the general population (see API RP 14C, Analysis, Design, Installation, and Testing of Basic Surface Safety Systems for Offshore Production Platforms). AIChE has published “Vapor Cloud Dispersion,”

Chevron Corporation

200-23

July 1999



“Vapor Release Mitigation,” “Guideline for Safe Storage and Handling of High Toxic Hazard Materials,” and “Hazard Evaluation Procedures.” The Hazard and Operability Study covered in “Hazard Evaluation Procedures” has been accepted by the EPA as definitive. Future publications planned are: “Quantitative Risk Assessment” and “Guidelines for Process Control.” Because of the many safe design practices built into Company standards and procedures, these AIChE Procedures—and their reporting requirements—may be found to be unnecessarily formal and lengthy. Modification and shortening should be considered if the full procedure is not legally mandated. The AIChE guidelines do not define what level of risk is acceptable, and this is a complex subject affected by conditions peculiar to a facility such as closeness to a population, amounts and types of materials involved, and regulatory emission limits. A guideline recommending “applicable projects, techniques, sources of help, references, suitable consultants, waiver procedures, and other appropriate guidance to operating company personnel” is under consideration by the Hazard Assessment Steering Committee. Contact HE&LP for an update.

280 P&ID Drawings and Engineering Forms 281 P&ID Drawings The end of this section contains the following figures referred to in the text of Section 200: Figure 200-3

P&ID—Grouped Equipment Layout

Figure 200-4

P&ID—Serial Equipment Layout

Figure 200-5

P&ID—Geographical Layout

Figure 200-6

Major Equipment P&ID—Furnace

Figure 200-7

Auxiliary P&ID—Tempered Oil

Figure 200-8

P&ID—Plot Limit Manifold

Figure 200-9

P&ID—Relief System

Figure 200-10

Auxiliary Instrumentation Drawing

282 Engineering Forms

July 1999

ICM-EF-824A

Standard Piping and Equipment Symbols

ICM-EF-824B

Standard Instrument Symbols

ICM-EF-824C

Standard Logic and Instrument Symbols

ICM-EF-824D

Guidelines for P&ID Presentation of Level Instrumentation

200-24

Chevron Corporation



290 References Instrument Society of America (ISA) • •

ISA Standard S5.1, Instrument Symbols and Identification ISA Standard S5.2, Binary Logic Diagrams for Process Operations

American Petroleum Institute (API) •

API RP 14C, Analysis, Design, Installation, and Testing of Basic Surface Safety Systems for Offshore Production Platforms, Table 2.2, “Component Identification”

American Institute of Chemical Engineering (AIChE) • • • •

Chevron Corporation

“Guideline for Safe Storage and Handling of High Toxic Hazard Materials” “Hazard Evaluation Procedures” “Vapor Cloud Dispersion” “Vapor Release Mitigation”

200-25

July 1999


Fig. 200-3


P&ID—Grouped Equipment Layout

Chevron Corporation

200-27

July 1999


Fig. 200-4


P&ID—Serial Equipment Layout

Chevron Corporation

200-28

July 1999

Instrumentation and Control Manual Fig. 200-5


P&ID—Geographical Layout

Chevron Corporation

200-29

July 1999


Fig. 200-6


Major Equipment P&ID—Furnace

Chevron Corporation

200-30

July 1999


Fig. 200-7


Auxiliary P&ID—Tempered Oil

Chevron Corporation

200-31

July 1999



P&ID—Plot Limit Manifold

Chevron Corporation

200-32

July 1999



P&ID—Relief System

Chevron Corporation

200-33

July 1999



Fig. 200-10 Auxiliary Instrumentation Drawing

Chevron Corporation

200-34

July 1999



Fig. 200-3 P&ID—Grouped Equipment Layout

Chevron Corporation

200-27

July 1999



Fig. 200-4 P&ID—Serial Equipment Layout

Chevron Corporation

200-29

July 1999



Fig. 200-5 P&ID—Geographical Layout

Chevron Corporation

200-31

July 1999



Fig. 200-6 Major Equipment P&ID—Furnace

Chevron Corporation

200-33

July 1999



Fig. 200-7 Auxiliary P&ID—Tempered Oil

Chevron Corporation

200-35

July 1999



Fig. 200-8 P&ID—Plot Limit Manifold

Chevron Corporation

200-37

July 1999



Fig. 200-9 P&ID—Relief System

Chevron Corporation

200-39

July 1999



Fig. 200-10Auxiliary Instrumentation Drawing

Chevron Corporation

200-41

July 1999

300 Process Control Abstract This section is an introductory reference to process control. It discusses control theory, control modes and problems and includes guidelines for typical process control situations. This section also discusses controller tuning and control mode selection.

Chevron Corporation

Contents

Page

310

Introduction

300-2

320

Control Loops

300-2

321

Open Loop Control

322

Closed Loop Control

330

Control Modes

331

Proportional Control

332

Integral Control

333

Proportional-Plus-Integral Control

334

Derivative Control

340

Advanced Control

341

Cascade Control

342

Feed-forward Control

350

Controller Tuning

351

Quarter Decay Method

352

Ultimate Sensitivity Method

353

Process Reaction Curve Method

360

References

300-5

300-14

300-18

300-23

300-1

June 1989

300 Process Control


310 Introduction Process control is fundamental to most industrial processes. Although control technology has evolved greatly in arriving at today’s microprocessor and digital implementations, all control methods rely on the same basic structure, called a “control loop.” Control loops have six basic constituents, as follows: •

Controlled variable. The condition that is being controlled

•

Setpoint. The value at which a controlled variable must be maintained

•

Manipulated variable. A condition (variable) that can be changed to cause the controlled variable to change

•

Controller. A device that keeps the controlled variable at the setpoint

•

Final control element. The device adjusted by the controller(s) to change the manipulated variable

•

Disturbances. Process conditions that tend to change the value of the controlled variable

320 Control Loops Control loops can be either manual or automatic. A manual control loop requires a human being to observe the value of the controlled variable. If this variable is not at the setpoint, the human observer adjusts a manipulated variable (see Figure 300-1). An automatic control loop employs a controller to keep the controlled variable at the setpoint. In Figure 300-2, the controller receives a signal from a transmitter (the circled X) representing the condition of the controlled variable, and sends an output signal to a valve regulating the manipulated variable. Fig. 300-1

Manual Control

Fig. 300-2

Automatic Control

In a refinery furnace, a controller monitors the outlet temperature (controlled variable). If the outlet temperature is not at the desired value (setpoint), the controller

June 1989

300-2

Chevron Corporation


300 Process Control

changes the fuel flow (manipulated variable) by changing the position of the fuel valve (final control element). Automatic control may be open loop (feed forward) or closed loop (feedback).

321 Open Loop Control In open loop control, the controller adjusts the final control element without measuring the process. An example of open loop control is a cycle timer that operates a drain valve, as in the simple gas-liquid separation process shown in Figure 300-3. At predetermined intervals, the timer causes the drain valve to open even if there is nothing to drain. Fig. 300-3

Open Loop Control

A more common example of open loop control would be an automatic lawn sprinkler system. Here a clock timer opens a water valve for several minutes each day. It would not check to see if the lawn needed water and would even turn on the sprinklers in the rain. Open loop control like these examples is not widely used. Open loop control operating in a feed forward mode is frequently used along with closed loop control. Feed forward control is discussed in Section 342.

322 Closed Loop Control Closed loop control, also known as feedback control, is the most widely used type of automatic control. If feedback control were used in Figure 300-3, the controller would open the drain valve only when the liquid level rose above the controller setpoint and would continue to adjust the valve as needed to keep the liquid drained from the vessel. The gas separation process in Figure 300-4 has a feedback (closed loop) level control system in which the controller LC receives a signal from the level transmitter LT. The controller compares this measurement with the setpoint and adjusts the outlet valve as necessary. The difference between the controller measurement and controller setpoint is the error signal. When the error is not zero, the level controller opens or closes the outlet valve to return the level to the setpoint.

Chevron Corporation

300-3

June 1989

300 Process Control


Fig. 300-4

Closed Loop Control

On/Off Control On/off control is the simplest mode of automatic control. It has only two outputs— on (100%) or off (0%)—and only responds to the sign of the error—positive or negative; i.e., whether it is above or below the setpoint. Because of an effect known as constant cycling, on/off control is not generally suitable for continuous automatic feedback control. If the control valve in Figure 300-4 were to remain completely open when the level is above setpoint, and completely closed when the level drops below setpoint, a constant cycling of valve position and level would result (see Figure 300-5). As with open loop control, the varying level resulting from constant cycling may be acceptable in some noncritical level applications. Fig. 300-5

On/Off Control

Differential Gap Control Differential gap control is a refinement of on/off control. Instead of changing output from on (100%) to off (0%) at a single setpoint, differential gap action changes output at high and low limits called boundaries. As long as the measurement remains between the boundaries, the controller holds the last output. This

June 1989

300-4

Chevron Corporation


300 Process Control

extends the period and limits the amplitude of the controlled variable oscillations (see Figure 300-6). On many controllers the size and position of the differential gap is adjustable, permitting fine-tuning. Fig. 300-6

Differential Gap Control

Differential gap control is suitable for some continuous automatic feedback control loops. It slows the rapid cycling of on/off control, reducing wear on the final control element while maintaining much of the simplicity of on/off control. A typical application of differential gap control is the operation of a dump valve or pump to keep a vessel level within an acceptable range.

330 Control Modes Controllers can be adjusted to function correctly in many different applications. Each controller usually has three adjustment modes: •

Proportional. Controller output changes by an amount related to the size of the error

•

Integral. Controller output changes by an amount related to the size and duration of the error

•

Derivative. Controller output changes by an amount related to the rate of measurement change

With pneumatic controllers and early electronic controllers, each mode added to a controller made it more expensive. Most electronic controllers available today are equipped with all three modes at no additional cost. The unneeded modes can be turned off. Most control applications use proportional-plus-integral control. Proportional-plusintegral-plus-derivative is sometimes used for temperature control with delays (deadtime) of several minutes. Proportional-only control is sometimes used in noncritical services such as draining vessels.

Chevron Corporation

300-5

June 1989

300 Process Control


Note that the proportional and integral actions depend on the error (defined as setpoint measurement), but the derivative action only depends on the measurement. Controllers are constructed this way so there will be no large change in controller output when the operator enters a new setpoint for the controller.

331 Proportional Control (Controller output can go directly to a valve or to the setpoint of another controller. In the following discussions, it is assumed that controllers send their output directly to a valve.) Figure 300-7 shows the relationship between valve position and error that is characteristic of proportional control: The valve position changes in exact proportion to the amount of error, not to its rate or duration. The response is almost instantaneous, and the valve returns to its initial value when the error returns to zero. Fig. 300-7

Proportional Control Response

Control Algorithm The linear relationship between the setpoint deviation (error) and the valve position (controller output) for proportional action can be expressed as follows: O = Kc E (Eq. 300-1) where: O = Controller output Kc = Controller Gain = ∆Output / ∆Error E = Error = (Setpoint - Measurement) This equation is called the control algorithm. The gain, Kc, is also called the controller sensitivity. It represents the proportionality constant between the control valve position and controller error.

June 1989

300-6

Chevron Corporation


300 Process Control

Proportional Band Another way of characterizing a proportional controller is to describe its proportional band. The proportional band is the percent change in value of the controlled variable necessary to cause full travel of the final control element. The proportional band, PB, is related to its gain as follows: Kc= 100/PB (Eq. 300-2) Both proportional band and gain are expressions of proportionality. Manufacturers may call their adjustments gain, sensitivity, or proportional band. Figure 300-8 shows the relationship between valve opening and proportional bands of different percentages. High percentage proportional bands (wide bands) have a less sensitive response than low percentage proportional bands (narrow bands). Fig. 300-8

Effect of Proportional Band

Bias Bias is the amount of output from a proportional controller when the error is zero. Equation 300-1 implies that when the error is zero, controller output is zero. The valve is either fully open or fully closed and provides no throttling action. Adding a bias provides this throttling action. Equation 300-1 then becomes: O = Kc E + B (Eq. 300-3) where: B = Bias (percent of full output) Typically, manufacturers set the bias at 50%. To prevent a process bump, the operator is sometimes allowed to adjust the bias before putting the controller in automatic. Figure 300-9 shows controller output versus error at different proportional bands with a 50% bias. At zero error, the controller output is 50% of full range for any proportional band.

Chevron Corporation

300-7

June 1989

300 Process Control


Offset A controller’s error is the difference between its setpoint and measurement. In a proportional-only controller, a change in setpoint or load introduces a permanent error called offset (see Figure 300-10). It is impossible for a proportional-only controller to return the measurement exactly to its setpoint, because proportional output only changes in response to a change in the error, not to the error’s duration. Fig. 300-9

Effects of Proportional Band with 50% Bias

Fig. 300-10 Proportional Control Response to a Load Change

Assume that a proportional-only controller controls the outlet temperature of a furnace and that the temperature is at the setpoint. If the feed rate to the furnace increases, more fuel will be needed. This disturbance represents a load change to the furnace. To get more fuel, the fuel valve must be opened more. As is suggested by Equation 300-3, the only way that the valve can be at some value other than its starting point is for an error to exist. Thus, the proportional controller alone cannot return the outlet temperature to its setpoint. As mentioned, some controllers allow the operator to adjust the bias until the value of E (the error, or offset) is zero. Offset is determined by the proportional band value for the controller and the change in valve position that occurs when a disturbance takes place: ∆E = PB (∆O) / 100 (Eq. 300-4) where: ∆E = Change in error PB = Proportional band ∆O = Change in valve position The proportional-only controller is the easiest continuous controller to tune. It provides rapid response and is relatively stable. If offset can be tolerated (loose control), proportional-only control can be used.

June 1989

300-8

Chevron Corporation


300 Process Control

332 Integral Control Integral (reset) action is the result of an integration of controller error with time. With integral action, controller output is proportional to both the size and duration of the error. As long as a deviation from setpoint exists, the controller continues to drive its output in the direction that reduces the deviation. The rate of change of controller output is proportional to the magnitude of the error. Figure 300-11 illustrates the open loop response of integral action. Fig. 300-11 Integral Controller Response (Open Loop)

Integral action is normally used in conjunction with proportional action; it is rarely used by itself. Integral action is quantified as the time (the reset time) required to change controller output by an amount equal to the change caused by proportional action. In other words, it is the time required to repeat the contribution of the proportional action. On some controllers, integral settings are in repeats, meaning repeats per minute; on others, settings are in minutes, meaning minutes per repeat. One setting is the reciprocal of the other; decreasing the integral time increases the amount of integral action.

Chevron Corporation

300-9

June 1989

300 Process Control


333 Proportional-Plus-Integral Control Proportional-plus-integral control is the recommended control action for most applications. Often called PI control, it combines proportional action and integral action in one controller. The resulting control action has the fast response and stability of proportional action, but no offset. In eliminating offset, integral action serves as an automatic bias adjustment. The output from a proportional-plus-integral controller may be expressed as follows: n

  1 E∆T m = K c E + ------  TR   o

∑

(Eq. 300-5) where: O = controller output Kc = controller gain E = error TR = reset time, minutes per repeat Σ = summation from time 0 to time n ∆T = interval between summations Figure 300-12 shows the open loop response of proportional-plus-integral control. Proportional control immediately acts to reverse the error. Integral action then continues to change controller output until the error equals zero. Fig. 300-12 Proportional-Plus-Integral Control Response (Open Loop)

June 1989

300-10

Chevron Corporation


300 Process Control

Figure 300-13 depicts proportional-plus-integral control for a closed loop. In response to a step change in load (top graph), the controlled variable (middle graph) falls below the setpoint. The integral action adjusts the bias from 50% initially to about 75% after the load change and shifts the position of proportional band (shaded area) on the scale. Notice that the percentage value of the proportional band is not changed. The lower graph shows the output of the controller. Fig. 300-13 Proportional-Plus-Integral Control Response (Closed Loop)

Wind-up A basic problem with integral controllers is that integral action continues as long as an error exists. Assume a proportional-plus-integral controller is used to maintain the level in the gas-liquid separator vessel in Figure 300-4. If a valve is closed upstream of the vessel, the level drops below the setpoint. The controller then closes the control valve in the outlet line to maintain the level setpoint. With no inlet flow, the control valve closes completely and the vessel level is still less than the setpoint. A pneumatic control valve will typically be fully closed at a controller output of 15 psig. Since the measured vessel level is less than the setpoint, the integral action of the controller continues to increase the controller output to the air supply pressure (typically 20-30 psig). The action of the integral controller trying to exceed the normal range of the controller output is called wind-up. If the upstream valve is opened and flow is restored, the vessel level will rise above the setpoint. The response of the controller to this high level will be delayed by the wind-up. When the controller does respond, the output goes to the opposite limit. In this case, the control valve will fully open and the vessel level will drop sharply. The controller may oscillate through several cycles, stroking the control valve from stop to stop on each cycle, before the oscillations cease and control is restored.

Chevron Corporation

300-11

June 1989

300 Process Control


Such oscillations overwork the control valve and, depending on the fluid and pressures involved, can cause mechanical damage and seriously disrupt the process downstream on the valve. An anti-wind-up feature may be included on controllers that are frequently subject to this type of disturbance. This limits the controller output range and thus prevents wind-up. When the process returns to normal, the controller lag is eliminated and the oscillations are no worse than those in a proportional controller.

Integral Time Integral time should be proportional to the time it takes for the process to respond to control action. When the process responds quickly, the integral time can be shorter. If the integral time is too short, the control valve reaches its limit before the measurement has time to respond. When the measurement does respond, it will overshoot the setpoint, causing the integral to drive the valve to its opposite limit. The time lag built into the gradual response of integral action lengthens the period of oscillation of a loop. For a loop with proportional-plus-integral control, the period of oscillation after a load change is longer than for proportional alone. For loops where the exact value of the controlled variable is not critical, the shorter period of the proportional-only controller can be an advantage. For example, a vessel may operate within a wide range of liquid level without adversely affecting pressure or gas quality. Therefore, the system level does not have to be accurately controlled, and proportional control is often sufficient.

334 Derivative Control With derivative action (also called rate action), the controller output is proportional to the rate of change of the error. This means the faster the change in level, the faster the change in controller output and control valve settings. By the same token, if the level remains constant, even with a large error, the controller output would be zero. This makes the use of derivative action by itself impractical. Derivative action is normally combined with proportional action or proportionalplus-integral action. Derivative action, being proportional to the rate of change of the measured variable, introduces a “lead” (anticipation) element into the controller. This increases the speed of response of the controller and compensates for the lags introduced by proportional and integral actions. Figure 300-14 illustrates derivative action. The output from a proportional-plus-derivative controller may be expressed as follows: M n – Mn – 1 O = K c E n + T D  -----------------------------    S (Eq. 300-6) where: O = controller output

June 1989

300-12

Chevron Corporation


300 Process Control

Kc = controller gain En = error at time n TD = derivative time, minutes Mn = measurement at time n Mn-1 = measurement at previous sampling time S = Time between measurements (sampling time) The derivative action is greatest when integral and proportional action are just beginning to respond. Derivative action also responds to the change in sign of the measured variable. This opposes the tendency of integral and proportional action to overshoot the setpoint and enables the controlled variable to settle out faster than with either proportional or proportional-plus-integral action. In Figure 300-14, area A represents the proportional component of controller output. Note that the proportional response is a function of the difference between the setpoint and the measured variable. Areas B and C represent the component added or subtracted by derivative action. As the measured variable stops decreasing and starts increasing, the sign of the derivative function changes. The integral action (area D) eliminates offset by not returning to zero when the proportional and derivative actions return to zero output. Areas E and F represent the corrections that result from all three actions taken together. Fig. 300-14 Proportional-Plus-Integral-Plus-Derivative-Control Action (Closed Loop)

Chevron Corporation

300-13

June 1989

300 Process Control


Derivative action, being sensitive to the rate of change of the measured variable, cannot be used in processes that require fast response, or that have rapid fluctuations or high noise levels. These conditions cause instability through large increases in the derivative gain, and rapidly change direction (sign). Although derivative action is difficult to tune because of its extreme sensitivity to measurement noise and other high frequency disturbances, it does have some applications. Most importantly, it is used with proportional and integral action in temperature processes that have large time lags. Derivative action can be very helpful in controlling processes that have significant deadtime, but using it can be difficult. Sometimes adding derivative action can make the control loop appear slow and inactive with some types of process disturbances. This sluggishness might lead one to increase the amount of derivative and perhaps also increase the controller gain. However, these new tunings might make the controller unstable when a different disturbance occurs in the plant.

340 Advanced Control Because this section of the Instrumentation and Control Manual is meant to be introductory in nature, we will define the term “advanced control” to be anything more sophisticated than simple, single-loop feedback control. Advanced control would therefore include cascade control, feed forward control, signal selector control, adaptive gain control, self-tuning controllers, multivariable control, matrix control, and many other techniques too numerous to mention. We will only deal here with cascade. The reader is encouraged to consult the references listed in Section 360 for additional information. The Monitoring and Control Systems Division in the Engineering Technology Department is also available for consultation.

341 Cascade Control Cascade control should also be considered when the primary control variable is slow to react to disturbances. Like any feedback control loop, a cascade control loop has a controlled variable, a setpoint and a controller. However, instead of having a valve as its final control element, a cascade controller sends its output to the setpoint of another controller, adjusting this setpoint to correct an error in the controlled variable. This other controller is called the secondary or slave controller. The cascade controller is called the primary or master controller. If disturbances in the process can be recognized and quickly corrected, the primary control loop will not be affected. This suggests that the secondary control loop must operate faster than the primary loop. In fact, general guidelines suggest that the secondary loop should respond at least five times faster than the primary loop. Looking again at the example of the furnace, let us assume that the fuel system provides fuel to several other furnaces as well. Over the course of hours, the pressure in the system might well vary as the fuel demand in all of the furnaces changes. A change in the fuel header pressure changes heat transfer in the furnace.

June 1989

300-14

Chevron Corporation


300 Process Control

Because heat transfer is a slow process, the outlet temperature controller cannot be tuned well enough to eliminate the effect of changing fuel flow (see Figure 300-15). (For details on controller tuning, see Section 350.) Fig. 300-15 Feedback Control Performance

On the other hand, if the fuel flow remains steady while the pressure is changing, the furnace temperature will be more constant. Fuel flow changes almost immediately when the control valve is moved. Therefore, the flow controller can be tuned to eliminate most of the disturbances in fuel flow. Such circumstances lend themselves to the use of cascade control: a fast process (fuel flow), a slow process (furnace heat transfer), and a disturbance (fuel pressure) that affects the fast loop. Figure 300-16 shows the cascade control system for the furnace. Fig. 300-16 Cascade Control

Chevron Corporation

300-15

June 1989

300 Process Control


Compare Figure 300-17 to 300-15. With cascade control the outlet temperature is much more steady. The fuel gas controller (secondary controller) has eliminated almost all fuel pressure disturbance from the furnace. Fig. 300-17 Cascade Control Performance

342 Feed-forward Control Feed-forward control measures a disturbance before it can affect the controlled variable, and changes the manipulated variable to compensate for the disturbance. Of course, for feed-forward control to work properly, the magnitude and timing of the effect on the controlled variable must be known. The process might be worse off if the manipulated variable is changed too much or too quickly. In Figure 300-18, a gas-fired furnace process is equipped with a temperature controller (TC), a feed-forward controller (FFC), and a summer, which adds the two controller outputs together. The feed-forward controller, also called a flow fraction controller, operates like a simple multiplier: The output of the FFC consists of its input (from the flow transmitter FT) multiplied by a ratio entered by the operator. Fig. 300-18 Feed-forward Control

June 1989

300-16

Chevron Corporation


300 Process Control

Figure 300-19 shows what might happen in a real furnace as the feed rate is changed. In the top graph, the feed rate to the furnace is raised at time 1. By time 2, the furnace outlet temperature begins to drop below setpoint. The fuel valve then begins to open and raises the outlet temperature back to the setpoint by time 3. In the bottom graph, the fuel valve has begun to open by time B, and by time C the furnace temperature is back to the original setpoint. With feed-forward and feedback control, the process has recovered from the feed rate disturbance much faster than with feedback control alone. Note that the temperature’s period of oscillation is the same in both cases. This period is a dynamic characteristic of the furnace and cannot be changed by the control system. However, the feed-forward controller has been able to reduce the size of the temperature disturbance and has speeded up the recovery. Fig. 300-19 Feedback/Feed-forward Control Performance

Feed-forward control should not be used by itself, but always with feedback control, because the rate and magnitude of the reaction of a process to a disturbance is rarely consistent.

Chevron Corporation

300-17

June 1989

300 Process Control


350 Controller Tuning Several methods are available to tune a controller to function in a specific loop. The following discussion considers some of the methods commonly used. Several of the references in Section 360, particularly Reference 5, should be useful when difficult situations are encountered.

351 Quarter Decay Method The quarter decay method is a closed loop controller tuning method. This means that the controller remains in automatic while tuning adjustments are made. The quarter decay method defines the ultimate limit for tight controller tuning. Often, the tuning constants it produces are too tight (too sensitive) in processes that have sticky valves and noisy measurements. To prevent controllers from going unstable unexpectedly, tuning constants should be set to values one-half as sensitive as those obtained with the quarter decay method. After these less sensitive tunings are exposed to actual upsets and irregularities, and the operators gain confidence in the controller tuning, it may be appropriate to make the tunings more sensitive. The general tuning sequence is as follows: 1.

With the controller in automatic, adjust all tuning constants to their least sensitive (least effective) setting. Proportional band should be at its highest value (proportional gain should be at its lowest value). Integral time should be at its highest value (most minutes per repeat or least repeats per minute). Derivative time should be at its highest value.

2.

Make a small step change in controller setpoint and record the controller measurement until it settles out.

3.

Change the setpoint back to its original value. Record the measurement as before.

4.

Increase the proportional gain (reduce the proportional band) in small steps and repeat steps 1-3 until the recording of the output resembles Figure 300-20, curve B; that is, until the amplitude of the first positive excursion of curve B is approximately four times that of the second (thus the name, quarter decay method).

5.

Measure the period of oscillation. Set the reset and derivative: TR = P/1.5 minutes (Eq. 300-7) TD = P/6 minutes (Eq. 300-8)

June 1989

300-18

Chevron Corporation


300 Process Control

Fig. 300-20 Quarter Decay Method Tuning

6.

With TR and TD set at above values, reestablish controller gain for quarter decay.

Figures 300-21, 300-22, and 300-23 show how the three tuning parameters affect the response of a controller. With proportional-only control, settling time is fairly long and there is a permanent offset from the setpoint. Adding integral control reduces settling time and eliminates offset. Adding derivative control to proportional control reduces settling time but not offset. Only integral control eliminates the offset. Fig. 300-21 Proportional-only Controller Response

Chevron Corporation

Fig. 300-22 Proportional-Plus-Integral Controller Response

300-19

June 1989

300 Process Control


Fig. 300-23 Proportional-Plus-Derivative Controller Response

352 Ultimate Sensitivity Method The ultimate sensitivity method ( Figure 300-24) is also a closed loop test. Adjust the integral time and/or the derivative time to their minimum values. Then narrow the proportional band (increase gain) in small steps, each time changing the setpoint as described in Section 351, until the controller measurement just begins to cycle continuously. This proportional band setting is called the ultimate proportional band, denoted “PBu.” The period of oscillation at the ultimate proportional band is called the ultimate period, measured in minutes and denoted “Pu.” The amplitude of the oscillations in Figure 300-24 has been exaggerated for clarity. The ultimate proportional band, PBu, and the ultimate period, Pu, are then used to calculate tuning constants as shown in Figure 300-25. These constants give the quarter damping response already discussed. Note that Figures 300-25 and 300-26 show two sets of equations for a proportionalplus-integral-plus-derivative controller. The set identified as “Commercial” should be used for controllers encountered in industry. The set identified as “Ideal” is based on an ideal control algorithm equation commonly used in universities. They are included here for completeness.

353 Process Reaction Curve Method This is an open loop tuning method. The controller remains in manual while response tests are made. The tuning method measures two parameters to describe the response characteristic of the process: process deadtime and process time constant. The deadtime is the delay between a change in valve position and the resulting change in the controlled variable. The process time constant is the time required for the controlled variable to reach approximately 60% of its final value.

June 1989

300-20

Chevron Corporation


300 Process Control

Fig. 300-24 Ultimate Sensitivity Method

Fig. 300-25 Ultimate Sensitivity Method Tuning Constants Proportional Band (%)

Reset Time (minutes)

Derivative Time (minutes)

Proportional Controller

0.5 PBu

—

—

Proportional + Integral Controller

0.45 PBu

Pu / 1.2

—

Proportional + Integral + Derivative Controller Ideal

0.6 PBu

Pu / 2.0

Pu / 8.0

Proportional + Integral + Derivative Controller Commercial

0.3 PBu

Pu / 4.0

Pu / 4.0

Notes

PBu = Ultimate Proportional Band, % Pu = Ultimate Period, minutes

To perform this test, change the controller valve position by a small amount and record the controlled variable. The deadtime, TD, and time constant, TC, are measured and their values used to calculate the controller tuning constants. Figure 300-26 shows how the measurements are made and used.

Chevron Corporation

300-21

June 1989

300 Process Control


Fig. 300-26 Open Loop Reaction Curve Method Tuning Constants

Note that the process reaction curve method cannot be used to integrate processes such as level control; when a valve controlling a level is changed the level continues to change until the vessel overflows or empties. Level controllers can be tuned using the ultimate sensitivity method or more advanced methods discussed in Reference 5. Figure 300-27 gives typical ranges of controller tuning constants for various processes. Use these values with caution; your process might not be “typical.” The

June 1989

300-22

Chevron Corporation


300 Process Control

exact values must be determined by one of the above methods. For future reference, always record the control loop ID number (e.g., FRC-123), the date, and the tuning constant when you have finished tuning a control loop. Fig. 300-27 Tuning Constants for Typical Process PROPORTIONAL BAND %

RESET TIME (MINUTES)

DERIVATIVE TIME (MINUTES)

Flow

100 - 500

0.02 - 0.1

none

Liquid Pressure

100 - 500

0.02 - 0.1

none

Gas Pressure

1- 50

0.1 - 0.5

none

Level

1- 50

0.05 - 0.25

none

10 - 100

1 - 10

0.5 - 20

LOOP TYPE

Temperature

360 References

Chevron Corporation

1.

Fundamentals of Process Control Theory. Instrument Society of America, 1981.

2.

Process Control Systems. McGraw-Hill, 1979.

3.

Process Instruments and Controls Handbook. McGraw-Hill, 1974.

4.

Controllers & Control Theory. Production Facility Bookware Series, International Human Resources Development Corp., 1987.

5.

Tuning and Control Loop Performance. Instrument Society of America, 1983.

300-23

June 1989

400 Pressure Measurement Abstract This section is a practical guide to the selection, specification, and installation of instruments for indicating, recording, and controlling pressure. Section 410 discusses general concepts of pressure measurement—particularly-as they bear on selecting a pressure instrument—describes and discusses specific devices and provides guidance in their application and specification. Section 420 gives general and specific guidance for the installation of pressure instruments. Section 440 lists reference material for further reading.

Chevron Corporation

Contents

Page

410

Application and Specification of Pressure Instruments

400-2

411

General Information

412

Pressure Elements

413

Pressure Gages

414

Field Pressure Recorders

415

Field Pneumatic Pressure Controllers

416

Pressure Transmitters

417

Pressure Switches

418

Draft Gages

419

Diaphragm Seals

420

Installation of Pressure Instruments

421

General Requirements—Field Pressure Instruments

422

Specific Requirements—Pressure Instruments

430

Model Specifications, Standard Drawings, and Engineering Forms 400-18

431

Standard Drawings

440

References

400-16

400-19

400-1

July 2000

400 Pressure Measurement


410 Application and Specification of Pressure Instruments 411 General Information Pressure instruments should be suitable for the process pressure, the process fluid, and the environment for which they are installed.

Range Field indicators, recorders, and transmitters should have ranges approximately double the expected operating pressure. The field pressure controller range should cover the minimum and maximum operating pressures.

Units of Calibration Pressure instruments should read in the following units: •

Above atmospheric: Pounds per square inch gage (psig), inches of water (inches H2O), or inches of mercury (inches Hg)

•

Below atmospheric: Pounds per square inch absolute (psia), ounces per square inch, inches H2O vacuum, or inches of mercury vacuum (inches Hg vac). Absolute pressure instruments should be ordered with compensation for barometric pressure changes

Force-Balance vs. Motion-Balance Pressure instruments with Bourdon tube, bellows, or diaphragm sensing elements may use either force-balance or motion-balance mechanisms to convert the element movement into an output signal. In a motion-balance transmitter (Figure 400-1), the moving tip of the element is connected to an indicator, the flapper of a pneumatic transmitter, or the current-producing section of an electronic transmitter. In a forcebalance transmitter (Figure 400-2), the force that tends to move the element is opposed by an equal force generated by the electronic or pneumatic output signal. A force-balance transmitter has no moving parts and therefore is not subject to hysteresis or dead band effects.

Overrange Protection Pressure instruments should withstand the maximum operating pressures encountered. Under unusual conditions, such as thermal expansion, the instrument may exceed its range. Most instruments can withstand overpressures up to 1.4 times their maximum range. Instruments exposed to a vacuum should be selected to withstand full vacuum. Certain pressure elements can withstand high overrange. Diaphragm elements with capsules backed up by a metal housing have a high overrange capacity, and many modern electronic pressure transmitters can withstand extreme overpressure. Pressure limiting valves, or “gage savers,” block the inlet pressure at a preset limit, but are rarely used.

July 2000

400-2

Chevron Corporation


Fig. 400-1


Motion-Balance Pressure Sensor

Fig. 400-2

Force-Balance Pressure Sensor

Pressure Element Materials Pressure elements should be designed to minimize corrosion. Unless the process fluid requires better corrosion resistance, pressure elements should be 316 stainless steel (316 SS). For salt water or services where chloride stress corrosion is possible, pressure elements should be Monel. Pressure elements for instrument air or sweet water should be bronze. Most other process fluids require better corrosion resistance, for which 416 SS is usually acceptable. Consult a materials engineer regarding highly corrosive services. See API RP 551, Section 4.2.10.

Connection Size The process connection for all pressure instruments should be ½-inch male or female National Pipe Thread (NPT). Receiver gages should have a ¼-inch NPT process connection.

412 Pressure Elements Pressure instruments may use mechanical sensing elements such as Bourdon tubes, bellows, or diaphragms. These elements are mechanically connected to an indicator, recorder, controller, or transmitter. Electronic pressure transmitters, on the other hand, have sensing elements such as resonant wires, strain gages, capacitors, and piezoelectric crystals that convert pressure into an electronic output signal. Specific techniques are often proprietary. The Company has not established a preference. Field pressure switches (see Section 417) may use Bourdon tube, bellows, diaphragm, spring piston, or spring disk sensors. Spring disk pressure switches are preferred for most applications. The following descriptions and comparisons for different sensing methods are for information only. In practice, the design engineer may have little or no say over the type of element that the manufacturer has utilized for the selected instrument. However, it may sometimes be necessary for a measurement to be performed that

Chevron Corporation

400-3

July 2000



requires a specific type of sensing element. In this case it is important to know the differences between various instruments.

Manometers A manometer is a U-shaped tube filled with a liquid, usually mercury or water. A differential pressure will cause the column on the high-pressure side to fall and the column on the low-pressure side to rise. The value of the differential pressure equals the difference in hydrostatic head between the two columns of liquid. Manometers are used to indicate low pressures, such as furnace drafts, and to calibrate low-pressure instruments.

Bourdon Tubes A Bourdon tube (Figure 400-3) is a curved metal tube closed at one end and fixed to a pressure source at the other. Application of pressure at the fixed end results in movement at the free end as the tube cross section deforms. Bourdon tubes come in a C shape, spiral, or helix, depending on the pressure to be measured. The Bourdon tube is the most commonly used element for pressure gages.

Bellows A bellows pressure element expands when pressure is applied to the inside, actuating an indicator, transmitter or controller. Bellows elements are generally used in pressure ranges from 0 to 10 inches H2O and from 0 to 10 psig, and for vacuum ranges from 10 to 20 inches H2O. In pneumatic instruments, bellows usually operate at approximately 3 to 15 psig.

Diaphragms The diaphragm sensor is a thin flexible metal disk. Pressure applied to one side of the disk causes a deflection that actuates the indicator, transmitter, or controller. Diaphragm elements are used to measure very low pressures, and vacuums from ¼ inch to 5 inches H2O. They are commonly used to measure furnace draft pressures.

Resonant Wire Pressure Sensors The resonant wire electronic pressure transmitter operates on the principle that a taut wire vibrates at a natural frequency that varies with wire tension. Resonant wire transmitters are accurate and stable. Ranges are 0 to 5 inches H2O and 0 to 6000 psig (Figure 400-4).

Strain Gage Pressure Sensors Strain gage electronic pressure transmitters operate on the principle that metallic conductors subject to strain exhibit a change in electrical resistance. The application of pressure bends the measuring element, and the resulting change in resistance is converted to an output signal. Ranges are 0 to 30 inches H2O and 0 to 18,000 psig. Strain gages require regulated power supplies. They are very compact and have a high speed of response.

July 2000

400-4

Chevron Corporation


Fig. 400-3


Three Types of Bourdon Tubes

(a) C SHAPE

(b) SPIRAL

(c) HELICAL

Chevron Corporation

400-5

July 2000


Fig. 400-4


Resonant Wire Pressure Transmitter (Courtesy of The Foxboro Company)

Capacitance Pressure Sensors Capacitance pressure sensors operate on the principle that the change in capacitance of an elastic element is proportional to the applied pressure. Ranges are 0 to 3 inches H2O and 0 to 5000 psig. Capacitance pressure transmitters have good linearity and frequency response.

Spring-piston Pressure Sensors Spring-piston elements consist of a piston opposed by a spring. The application of pressure moves the piston. The position of the piston opens or blocks instrument air ports in the cylinder wall, changing the pneumatic output signal. These sensors are used in pneumatic pressure switches, which are often called pressure pilots or stick pilots. See Figure 400-5.

Spring-disk Pressure Sensors The spring-disk pressure sensor (Figure 400-6) is used in pneumatic and electronic pressure switches (see Section 417). The sensor is a convex metal spring disk (Belleville spring). Pressure applied against the convex side gradually deflects the disk until it snaps to a concave shape. When the pressure is released, the disk snaps back to its original shape. The snap action is connected to an electrical or pneumatic switch. The switch usually includes a helical spring behind the disk for adjustment of the trip point. The switch contacts can be ordered to close or open at the set pressure on either an increasing or decreasing pressure. Pressure switches may be either field-adjustable or preset at the factory. This type of pressure switch is dependable and has good repeatability.

July 2000

400-6

Chevron Corporation



Fig. 400-5

Pressure Pilot (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.)

Fig. 400-6

Spring-disk Pressure Switch (Courtesy of Custom Control Sensors, Inc.)

413 Pressure Gages It is recommended that locally mounted pressure gages be installed at the following locations to monitor the performance of equipment and pressure instruments:

Chevron Corporation

•

Pump and compressor discharges

•

Pump and compressor interstages

400-7

July 2000



•

Many pump and compressor suctions

•

Vessels and the vapor spaces of columns

•

Nonindicating pressure transmitters

•

The sensing side of each direct-operated pressure regulator and field pressure controller

•

Each field pressure switch

•

Each field pressure recorder

•

Across filters

•

Furnace fuel oil, fuel gas, and atomizing steam branch headers

•

Each furnace, using a single draft gage manifolded to the convection section inlet, bridgewall, and below the stack damper

•

Any application where local pressure indication is desirable, such as inlets of production manifolds

Pressure Gage Specification Pressure gages are available in a wide range of qualities and prices. Standard quality pressure gages with blowout backs are acceptable. The Richmond Refinery has standardized on stainless steel liquid-filled pressure gages, some of the other locations use a combination of stainless steel liquid filled and unfilled gages.

Rating For most process applications, pressure gages with an accuracy grade 2A in accordance with ANSI B40.1, Gauges—Pressure Indicating, Dial Type, Elastic Element, are acceptable. Grade 2A means the pressure gage is accurate to 0.5% of span.

Gage Construction Pressure gages should have 4½-inch dials, ½-inch gage connections, and stainless steel movements (gages with 6-inch dials are also acceptable in most locations). The case can be cast iron or cast aluminum with a blowout plug or plastic (fiberglassreinforced polypropylene or phenolic) material with a blowout back. The gage connection should have wrench flats. The lens should be shatterproof glass. White dials with black figures and letters are standard. Other colors and graphic designs are available. Adjustment of the zero should be possible without removing the pointer from its shaft. All pressure gages should display the tube, tip, and socket material on the front dial. The information should be imprinted by the manufacturer and should clearly describe the materials.

Range Pressure gage range should be selected so that the gage operates in the middle third of the scale. Overrange and underrange travel stops should be provided.

July 2000

400-8

Chevron Corporation



When the maximum scale reading is 1000 psig or higher, metal case gages should have solid fronts separating the dial from the movement (this is standard with plastic cases). Also, the Bourdon tube should be bored instead of drawn, and the connection between the Bourdon tube and the socket and tip should be threaded and backwelded.

Liquid-filled Gages The cases of liquid-filled gages are filled with a viscous lubricating fluid. Such gages maintain their accuracy much longer than conventional gages. They should be used in pulsating or vibrating services, such as the discharges of reciprocating pumps and compressors, and should be considered for other severe services. The fill liquid should be glycerine for ambient temperatures above 0°F and silicone oil for ambient temperatures below 0°F, or if the process fluid and glycerine are incompatible. Mineral oil (sometimes specified under the brand name “Kaydol”) is acceptable for services above 0°F. Liquid-filled gages should be filled until only a trace of a bubble is left in the face, to allow for thermal expansion and to show that the gage case is filled. The maximum temperature for liquid-filled gages is 150°F.

Receiver Gages Receiver gages should have 4½-inch diameter dials, bronze Bourdon tubes and stems, and ¼-inch NPT process connections. Local instrument air pressure gages on control valves, pneumatic transmitters, and controllers should have 1½-inch or 2½-inch diameter dials. Receiver gages and instrument air pressure gages should be rated Grade B in accordance with ANSI B40.1. Grade B pressure gages are accurate to 2%.

Instrument Air Pressure Gages Instrument air pressure gages are usually installed in the following locations: •

Each instrument air header (manifold)

•

Each instrument air supply regulator outlet (if the regulator does not have a pressure gage)

•

Control valve diaphragm actuators (where there are no positioners)

Pressure Test Points Pressure test points consisting of a process connection with a plugged processquality root valve should be provided on process equipment and piping at the following locations:

Chevron Corporation

•

Inlet and outlet of packed vessels and columns

•

Indicating pressure transmitters, controllers, and recorders. The pressure test point is the same as the calibration connection on the instrument manifold

•

Inlets and outlets (both shell and tube side) of all heat exchangers and reboilers

•

Inlet and outlet of each air cooler

400-9

July 2000



414 Field Pressure Recorders Field pressure recorders should have weatherproof cases of fiberglass-reinforced plastic or aluminum. Colored epoxy or baked vinyl is used to provide a textured finish. If the case contains electrical components, it should meet the electrical classification requirements for the area. The pressure element should be suitable for the process pressure and fluid. The range should be selected so that the normal process pressure is in the middle third of the chart. If suppressed-range pressure recorders are specified on critical applications, such as compressor suction or discharge, a full-range pressure recorder should also be provided. A dual-pen, dual-range recorder may be used instead of a full-range recorder. The process connection should be ½-inch female NPT. The chart should be 12 inches in diameter. The case should include a socket or yoke for mounting on a 2-inch pipe. The typical chart drive has a 7-day rotation and a clockwork drive with an 8-day wind. Electric and pneumatic chart drives are also available.

415 Field Pneumatic Pressure Controllers The field pneumatic pressure controller (pressure controller) compares the measured process pressure to a setpoint and sends an output signal to a final control element (control valve) which acts to hold the process pressure at the setpoint. Field pressure controllers are used when local control of pressure is required and the following conditions hold: • • • •

It is not necessary to change the pressure setpoint from a central control house It is not necessary to record the pressure from a central control house It is not necessary to change the controller tuning from a central control house The controlled pressure is not part of a cascade loop

Field pressure controllers may be either indicating controllers or recording controllers. Blind controllers are rarely used.

Specification of Field Pressure Controllers Field pneumatic pressure controllers should have weatherproof cases of fiberglassreinforced plastic, or aluminum. Colored epoxy or baked vinyl is used to provide a textured finish. If the case contains electrical components, it should meet the electrical classification requirements for the area. The pressure element should be suitable for the process pressure and fluid. The controller should include an indicator for the process pressure and setpoint. The setpoint should be easily adjusted using a knob either inside or outside the case, depending on the need to make adjustments. The controller should operate on an instrument air pressure of 18 to 22 psig. The output should be 3 to 15 psig. (Although some locations have safely used other supply mediums such as natural gas, nitrogen, and carbon dioxide, caution is advised since some instrument components may be incompatible with the medium selected. When available, instrument quality air is preferable.)

July 2000

400-10

Chevron Corporation



The controller should include a two-position, bumpless, auto/manual switch that is internally mounted to avoid accidental switching. The controller action should be specified as either direct or reverse acting, and it should be possible to change the action in the field. The controller should include supply and output pressure gages. The process connection should be ½-inch female NPT. The case should include a socket or yoke for mounting on a 2-inch pipe.

Control Mode The control mode is normally selected from the following options: •

On/off control. For alarms, protective devices, and startup and shutdown of equipment

•

Proportional only control. For simple pressure control where small variations from the setpoint are unimportant. The proportional band should be adjustable at least from 0% to 200%

•

Proportional plus reset control. This is the usual control mode for field pressure controllers. The proportional band should be adjustable at least between 0% and 200%. For applications where the controller occasionally operates above or below setpoint or in intermittent service, it should be specified to include anti-reset windup

416 Pressure Transmitters Process fluids should not be piped directly to a control building. Pressure transmitters permit the control, recording and indicating of pressure at a central location.

Electronic Pressure Transmitters The electronic pressure transmitter is used to transmit a signal to a remote electronic receiver. The receiver may be a controller in a remote building, a remote marshalling cabinet for numerous signals, or a remote terminal unit (RTU). Transmitters should be compatible with the remote control house instrumentation. Indicating transmitters should not be specified; they generally use Bourdon tube pressure elements and motion-balance mechanisms, and are not as reliable as forcebalance transmitters. An electronic pressure transmitter may be specified with a suppressed range to allow for range and span adjustments without changing electronic or mechanical components. Transmitters should include adjustable damping to smooth out the noise from high frequency pressure transients. The accuracy of electronic transmitters has drastically improved since the introduction of “smart” transmitters. Accuracy should be within 0.25% of calibrated span for standard electronic transmitters and 0.1% of calibrated span for “smart” transmitters. It is advisable to consult individual manufacturer’s specifications; for example, Rosemount gage-pressure transmitters have a span accuracy of 0.05% of span. The transmitter signal may be digital or 4 to 20 milliamps DC. The transmitter should be loop powered, which means that the operating voltage is carried on the

Chevron Corporation

400-11

July 2000



same two wires that carry the output signal. “Smart” transmitters that operate on a 24-volt DC power supply are preferred. Some advantages of “smart” transmitters include lower cost of ownership, improved diagnostic capabilities, ease of calibration and commissioning. The electronics enclosures for newer transmitters minimize the effects of radio frequency interference (RFI). In addition, much of the RFI in the signal wiring can be eliminated by metal conduit. Electronic pressure transmitters must be specified to meet the appropriate electrical classification. Intrinsically safe transmitters must be specified as such and so labeled by the manufacturer. Transmitters should be supplied with linear output meters and with accessories for mounting on a 2-inch pipe.

Pneumatic Pressure Transmitters Pneumatic pressure transmitters transmit pressure signals to remote pneumatic receivers. Both blind and indicating transmitters are available. Blind transmitters are preferred because they use force-balance mechanisms (See Section 411). Accuracy should be within 0.5% of the calibrated span. Pneumatic pressure transmitters should have weatherproof cases of fiberglass or aluminum. The pressure element should be compatible with the process pressure and fluid. The transmitter should be designed to operate on an instrument air supply pressure of 18 to 22 psig. The output pressure should be 3 to 15 psig. Output is typically specified as direct acting, but reverse acting is an available option. The transmitter should have supply and output pressure gages. The process connection should be ½-inch female NPT. The case should include a socket or yoke for mounting on a 2-inch pipe.

Receiver Pressure Gages A receiver pressure gage operates on the 3 to 15 psig pneumatic signal from a pneumatic transmitter. It provides a local indication of the transmitter’s performance. The dial is specified to match the range of the transmitter. All blind pneumatic transmitters except (in most cases) level transmitters should have at least one receiver gage on the transmitted signal. The receiver gage should be mounted at a location convenient for the operator. For instance, if a control valve is associated with the transmitted variable, the gage should be readily visible from the manual bypass valve for field manual control. For split-range instruments, a gage should be visible from each bypass valve. A receiver gage is not required on level transmitters if the gage glass is readily visible from the bypass valve. Indicating pressure transmitters should have receiver gages if the transmitter cannot be easily read from the bypass valve.

417 Pressure Switches Field pressure switches protect equipment and machinery from overpressure and underpressure without reliance on a remote control house. Hermetically sealed

July 2000

400-12

Chevron Corporation



switch contacts are used in many producing applications. They are usually electrical but can be pneumatic.

Electrical Pressure Switches Electrical pressure switches provide on/off contact closures for use with equipment or an alarm or shutdown system. They should have one of three standard case designs, depending on the electrical classification where they are installed: • • •

Explosionproof (NEMA 7) Weather-resistant (NEMA 3 & 4) General purpose (NEMA 1)

Electrical pressure switches should have snap-acting, dual, single pole double throw (SPDT) or double pole double throw (DPDT) contacts. The contacts should be rated to supply the operated device with a minimum of 10 amperes at 115 volts AC and 5 amperes at 28 volts DC. For potentially corrosive environments and for intrinsically safe systems, hermetically sealed switch contacts should be specified. The engineer should study the switch specifications to verify that its ratings are compatible with the application. Switch contacts should be open in the alarmed condition. Terminal blocks or terminal strips should be provided. Dead front or shrouded terminal blocks are acceptable. The electrical conduit connection should have a minimum diameter of ½ inch. If the switch contacts are handling milliamp signals, the contacts may be specified with a gold or other conformal coating to minimize oxidation. This usually adds only about 20 dollars to the cost of the switch. The pressure element may be a Bourdon tube, a bellows diaphragm, or a spring disk. Spring disks are preferred if the application permits a relatively large dead band (about 7% to 8% of the differential pressure range). Otherwise, specify Bourdon tubes for set pressures above 100 psig and bellows diaphragms for set pressures less than 100 psig. The pressure switch setpoint should be in the middle third of the range. Proof pressure should be higher than the maximum process pressure. Electrical pressure switches are available with either a fixed or an adjustable dead band between the setpoint and the reactivation point (see Figure 400-7). Closedifferential switches are generally factory set at 0.5% to 1% of the span. On double adjustment switches, both the set and reactivation points can be adjusted. The minimum differential varies from 2% to 8% of the span. The type of switch depends on the application. Electrical pressure switches should have internally adjustable setpoints with calibrated scales. Dual control electrical pressure switches are available with two independent switches in the same housing. Selection of the adjustable range for a specific installation should consider both the setpoint actuation accuracy and the life factor. For greatest accuracy, the setpoint should fall in the upper half of the range. For the most favorable component life, it

Chevron Corporation

400-13

July 2000


Fig. 400-7


Pressure Switch Dead Band

should be in the lower half. The usual compromise is to specify a setpoint in the middle third of the range.

Pneumatic Pressure Switches Pneumatic pressure switches are on/off pneumatic controls used to operate equipment, alarms, and shutdown systems. Pneumatic pressure switches are also called pressure pilots and are of two basic types: stick pilots and Bourdon tube pilots. Stick pilots are spring piston sensors with a three-way spool valve block-and-bleed device. Spring and piston size is determined by the pressure range. They are used in wellheads and flow lines in production fields because they are resistant to plugging by solids. Stick pilots should be constructed of 316 stainless steel with Viton O-rings. The trip point should have a repeatability of 3% of the set pressure or 5 psi, whichever is least. The connection for the transmitted air signal should be ¼-inch NPT. Bourdon tube pressure pilots are pressure switches with a 0% to 2% proportional band to give on/off control. The range is determined by the Bourdon tube. They are available in high switch point only, low only, and high/low configurations. Bourdon tube pressure pilots are used in producing applications that are not subject to plugging by solids.

418 Draft Gages Draft gages are low-pressure indicators or transmitters that are installed on process furnaces to measure draft. To improve accuracy, a single draft gage is normally piped to various locations on the furnace.

419 Diaphragm Seals Diaphragm seals, also called chemical seals or gage protectors, use a thin flexible diaphragm to isolate the pressure element from the process fluid (see Figure 400-8). The space between the diaphragm and the sensing element is filled with a suitable

July 2000

400-14

Chevron Corporation



noncompressible liquid. Diaphragm seals can be an integral part of the pressure instrument or be connected by capillary tubing. Capillary tubes up to 25 feet long Fig. 400-8

Typical Diaphragm Seal (Used with permission. Ashcroft is a registered trademark of Dresser Industries Instrument Division.)

are available. In vibrating service, diaphragm seals should be remotely mounted and have armored stainless steel capillary tubing. Diaphragm seals have the following applications: •

Steam

•

Water, if the leads are subject to freezing

•

Process streams that are corrosive to the pressure element

•

Dirty process streams containing solids that can plug the pressure element

•

Viscous process streams that can solidify in the pressure system

•

High-temperature process streams that exceed the maximum temperature rating of the instrument

Diaphragm seals can be used at temperatures ranging from -40°F to 1500°F. They can have either a ¾-inch or 1-inch NPT screwed or flanged process connection, depending on the piping classification. The bottom housing and the diaphragm material should be 316SS or better and should be compatible with the process fluid. Filling fluid identification or maximum temperature limit should be stamped on the body or nameplate. The diaphragm should be welded or attached to the body so that fluid will not escape when the diaphragm is disassembled.

Chevron Corporation

400-15

July 2000



420 Installation of Pressure Instruments 421 General Requirements—Field Pressure Instruments Field pressure instruments may be affected by ambient conditions such as humidity, dust, and airborne corrosive vapors and mists. Temperature extremes will also affect performance. For example, a pneumatic pressure transmitter may show as much as a 3% zero shift in span for a 75°F rise or fall in ambient temperature.

Accessibility and Visibility Pressure instruments should be located so that they are easy to observe, calibrate, and repair. Figure 400-9 gives the access recommendations for specific kinds of pressure instruments. Fig. 400-9

Access Requirements for Pressure Instruments Acceptable Means of Access Platform or Grade

Stepladder or Rolling Platform(1)

Permanent Ladder

Pressure Transmitter

Yes

Yes

No

Field Pressure Controller

Yes

No

No

Field Pressure Recorder

Yes

No

No

Field Pressure Switch

Yes

No

No

Field Pressure Gage

Yes

Yes

Yes

Instrument Type

(1) Provided the instrument is less than 10 feet above grade and the site is safe for a ladder or platform.

Pulsation and Vibration Pressure instruments for lines or equipment such as reciprocating compressors, high-pressure pumps, and high-pressure drop control valves whose vibration can impair instrument performance or cause connection failure should be remotely mounted, with tubing or a diaphragm seal and capillary between the instrument and the root valve. See API RP 551, Sections 4.2.5 and 4.2.6.

Heat Tracing, Purges, and Seals Pressure leads handling fluids that may solidify or become viscous enough to impair measurement should be heat traced. Alternatively, the process fluid can be isolated from the pressure instrument by diaphragm seals, seal pots, flowing seals, or other means. Diaphragm seals and heat tracing up to the seal are preferred. If heat tracing is provided, the viscosity of the process fluid in the leads to pressure instruments should be maintained at 4000 centistokes or less at the plant minimum recorded ambient temperature. The process fluid should not be traced at a temperature higher than the maximum working temperature of the pressure instrument.

July 2000

400-16

Chevron Corporation



For process leads, pretraced, preinsulated tubing bundles, heated either by steam or electricity, are preferred. In liquid service where plugging is likely (i.e., heavy fuel oil) pressure taps should be located in the top half of the pipe.

Siphons Siphons, or pigtails, should be provided for vapor services above 150°F and for steam service to prevent the condensing vapor from overheating the instrument. In liquid service, if the pressure gage is mounted above the pressure tap, a siphon should be provided when the process fluid temperature exceeds 300°F. See Standard Drawings GB-J1143, GB-J1146, and GB-J1147. See also API RP 551, Section 4.3.4.

Tubing AP RP 551 recommends ½-inch or 3/8-inch O.D. tubing to connect remote-mounted pressure instruments to process connection block valves. For manufacturing applications, ½-inch O.D. stainless steel tubing should be used. For production applications, many facilities have standardized on 3/8-inch O.D. stainless steel tubing.

Pressure Instrument Piping Root valves should use ¾-inch process connections and ¾-inch process valves to the process tap. Normally, gate valves are used, but the valve should match the piping classification. Do not install valves that can trap pressure without bleedoff when the pressure gage or switch has been removed.

Restriction Fittings Restriction fittings are installed at the root valve to minimize the release of process fluid should the instrument or instrument tubing fail. They are subject to plugging and are generally used only in toxic (H 2S), corrosive (NH3), hazardous (LPG) and clean refinery stock services. Remotely mounted pressure instruments should include a restriction fitting screwed into the outboard end of the root valve. The fitting should be a special blind tubing fitting with a small-diameter drilled hole. (See Standard Drawing GB-J1223). Pressure instruments mounted directly on the root valve should include a restriction adapter screwed into the outboard end of the root valve. The adapter should be made of steel barstock with NPT threads at both ends.

422 Specific Requirements—Pressure Instruments Pressure Gages Direct-mounted pressure gages may be installed with a ¾-inch root valve that meets the piping standards or with proprietary pressure gage block and bleed valves or approved equivalent. See Standard Drawings GB-J1143 and GB-J1144. Remote-mounted pressure gages should be installed with a ¾-inch root valve that meets the piping standards. See Standard Drawing GB-J1146.

Chevron Corporation

400-17

July 2000



Field Pressure Recorders, Controllers, and Transmitters Field pressure recorders, controllers, and transmitters should be piped in parallel with a process pressure gage and should include a three-valve manifold to facilitate field calibration. See Standard Drawings GB-J1145, GB-J1146, and GB-J1147.

Pressure Switches Facilities for testing alarms and shutdowns should include but not be limited to: •

Field High Pressure Switch. A pressure gage on the switch process lead and a valved connection should be provided to permit testing with a portable pressure source. See Standard Drawings GB-J1146 and GB-J1147

•

Field Low Pressure Switch. A pressure gage on the switch process lead and an atmospheric bleed connection should be provided to bleed off pressure and permit testing against the pressure gage reading. In LPG, high H2S, and other hazardous services, the bleed should be piped to the relief system or another safe place. See Standard Drawings GB-J1146 and GB-J1147

Draft Gages See Standard Drawing GB-J1148 for typical installation.

Automatic Pump Starters The process connection for a pressure-operated automatic pump starter (APS) should be made between the prime pump unit discharge and its check valve (first valve). The APS pressure pilot should be located at the process connection if the steam APS control valve is visible from that location. If the steam valve is not visible from the process connection, the pressure pilot should be located where the steam valve can be observed. Steam-driven APS pumps are used in refineries where the process must keep running through power failures. Most other plants use a pressure switch on the common discharge line and select between the main pump and a standby motordriven pump with HOA (Hand-Off-Auto) switches.

430 Model Specifications, Standard Drawings, and Engineering Forms 431 Standard Drawings

July 2000

GB-J1143

Instrument Installation Details— Pressure Gage Installation—¾" Root Valve

GB-J1144

Instrument Installation Details— Pressure Gage with Diaph. Seal—¾" Root Valve

GB-J1146

Instrument Installation Details— Remote Mounted Pressure Instrument—With Pressure Gage

400-18

Chevron Corporation



GB-J1147

Instrument Installation Details— Two Remote Mounted Pressure Instruments with Pressure Gage

GB-J1148

Instrument Installation Details— Draft Gage

GB-J1223

Details for Gage Adapters and Restriction Fittings

440 References The Company employs the following industry codes, standards, and recommended practices for the design, specification, and installation of pressure instrumentation:

Chevron Corporation

1.

API Recommended Practice 551, Process Measurement Instrumentation, American Petroleum Institute.

2.

ISA S20, Specification Forms for Process Measurement and Control Instruments, Primary Elements and Control Valves, Instrument Society of America, 1981.

3.

Pressure Instrumentation, Production Facility Bookware Series, Paragon Engineering Services, Inc., 1987.

4.

ANSI B40.1, Gauges—Pressure Indicating, Dial Type, Elastic Element.

5.

PIP PCCPR001, Pressure Measurement Criteria.

400-19

July 2000

500 Flow Measurement Abstract The term “flow measurement” covers volume and mass measurement for liquid, gas, and vapor (steam). Many new flow measurement technologies have been developed in the past 25 years, and the accuracy of traditional flow measurement devices has improved. This section is primarily concerned with plant, or process, flow measurement. Custody transfer flow measurement is briefly discussed, with reference to the appropriate chapters of the API Manual of Petroleum Measurement Standards (MPMS) and to the standards of the American Gas Association (AGA). The section gives technical information for the selection, application, and installation of flow measurement devices, or flow meters, that are commonly used in the petroleum and petrochemical industries. Operating principles, performance characteristics, and applications and limitations are discussed.

Chevron Corporation

Contents

Page

510

Introduction

500-3

520

Selection of Flow Measurement Devices

500-3

521

Differential Pressure (Head-type) Flow Meters

522

Positive Displacement (PD) Meters

523

Turbine Flow Meters

524

Magnetic Flow Meters

525

Ultrasonic Flow Meters

526

Variable Area Meters (Rotameters)

527

Vortex Shedding Flow Meters and Swirl Flow Meters

528

Mass Flow Meters

530

Other Flow Measurement Devices

531

Open Channel and Partially Filled Pipe Flow Measurement

532

Flare Flow Meters

533

Two-Phase and Multi-Phase Flow Metering

534

Other Flow Meters

500-1

500-92

July 1999

500 Flow Measurement

July 1999


540

Meter Provers

500-102

541

Proving Liquid Flow Meters

542

Proving Gas Flow Meters

543

Proving Mass Flow Meters

550

Flow Switches

560

Model Specifications, Standard Drawings, and Engineering Forms 500-104

561

Model Specifications

562

Standard Drawings

570

References

571

Included Materials

572

Other References

500-103

500-105

500-2

Chevron Corporation



510 Introduction In this section, all of the flow meters covered in API Recommended Practice 551, Section 2, “Flow” (API RP 551) are discussed, with additional information supplemental to the RP. As stated in its scope, API RP 551 is primarily concerned with refinery process flow measurement. It is, however, also applicable to gas and chemical plants and to other production facilities. For entry level engineers and those with limited experience with flow meters, API RP 551 is a good document to start with. It provides a basic, unbiased view on commonly used flow meters. A copy of API RP 551 is included in Volume 2 of this manual.

520 Selection of Flow Measurement Devices Once the need for flow measurement and the location for a flow meter have been determined, the first question usually is what type of device should be used. The first consideration in selecting a suitable flow measurement device is the service, or the type of fluid for which the flow meter will be used. Figure 500-1 is a chart for preliminary selection of flow meters based on their intended service. Keep in mind that there are often exceptions. Figure 500-2 provides a quick comparison of the main types of flow meters on the basis of typical performance criteria, design conditions, and cost of equipment. Each of the flow meters listed in Figures 500-1 and 500-2 will be discussed in more detail below. Once a flow meter is selected on the basis of the service, other considerations are as follows: • • • • • •

Type of fluid to be handled Process conditions under which the flow meter will be operating Performance requirements Electrical area classification and safety Installation conditions and maintenance requirements Economics

Type of Fluid To Be Handled Characterize the fluid: Is it liquid, gas or steam, or two-phase? Clean, dirty, or a slurry? Corrosive or noncorrosive? For steam flow metering, Appendix D describes, in a flowchart, the process of selecting the right type of flow meter.

Process Conditions Each of the following criteria should be considered. Note that the maximum value for a particular criterion or the extreme process conditions are often used as the design conditions.

Chevron Corporation

500-3

July 1999


Guide to Flow Measurement

July 1999

Fig. 500-1

500-4


Chevron Corporation


Flow Meter Selection Table

Chevron Corporation

Fig. 500-2

500-5


July 1999



•

Normal and extreme operating conditions

•

Minimum and maximum flow rates

•

Physical properties of the fluid, such as viscosity and entrained solid particles or air

•

Pressure and temperature

•

Dynamic properties of the fluid, characterized by the Reynold’s number (Re)

The accuracies of many flow meters can be affected by the velocity profile of the fluid, which is measured by a nondimensional number called “Reynold’s number” (Re). Reynold’s number is the ratio of the inertia forces to viscous forces, and is defined by the following equation: ρVD Re = -----------µ (Eq. 500-1)

where: ρ = density of the fluid V = average velocity of the fluid D = inside diameter of the pipe µ = absolute viscosity of the fluid. Up to approximately Reynold’s number 2,000, the flow is called laminar, viscous, or streamline flow. The flow profile in this regime is not affected by the wall roughness of the pipe. Above 10,000, the flow is called fully developed turbulent. The region between 2,000 and 10,000, where the flow is shifting from laminar to turbulent, is not clearly defined and is called transitional. Within the transition regime, the flow profile exhibits a flat parabolic geometry and is called “plug flow.” Generally speaking, the velocity profile in the transition between laminar and turbulent flow regimes can be unstable and difficult to predict, as the flow may exhibit properties of both laminar and turbulent regions or oscillate between them.

Performance Requirements The performance of a flow meter is often described by its accuracy, linearity, repeatability, flow range (rangeability), and resolution. These are defined below. Accuracy. Accuracy is a measure of how close to true or actual flow the instrument indication may be. It is often expressed as a percent of full scale (FS) and/or of reading (“rate”). Accuracy at a particular flow rate may be an order of magnitude better than rated flow range accuracy. Accuracy is specified for a given turn-down, or flow range. Generally speaking, the accuracy of a flow meter decreases with increased turn-down.

July 1999

500-6

Chevron Corporation



Unless a flow meter can be proved by another device, such as a “master meter” or a “prover” traceable to the National Bureau of Standards (NBS), now called National Institute of Standards Technology (NIST), accuracy is often difficult to prove. Furthermore, improper installation may adversely affect the accuracy of a flow meter. The term “accuracy” may not always appear on a vendor’s performance specification sheet but may be expressed as “repeatability” and “linearity.” These two terms are discussed below. Repeatability. Repeatability is the ability of a flow meter to indicate the same reading each time the same flow conditions exist. It is usually expressed as percent of full scale. Often, repeatability in a vendor’s specifications is based on shop tests conducted by the manufacturer of the flow meter within its recommended range of flow, temperature, pressure, and viscosity for a given type of fluid (e.g., air, water, or other solvent). Linearity. Linearity measures how linear the output signal is expected to be in proportion to flow rate. The output signal from a flow meter may or may not be linear depending upon the operating principle of the flow meter used. For example, the output signal of the primary element of a differential pressure flow meter is not linear to flow rate. Rather, the square root of the differential pressure produced by the meter is linearly proportional to the flow rate. Therefore, the raw signal produced by a differential pressure flow meter has to be conditioned (“scaled”) by another device to provide a linear signal. Flow Range (Rangeability). The maximum and minimum flow rates must be determined. The ratio of these two rates, often called “turndown,” “flow range,” or “rangeability” of a flow meter, depends upon the operating principle of the flow meter’s primary element. The flow range of a particular flow meter may make it unsuitable for some applications. Resolution. Resolution is a measure of the smallest increment of total flow that can be individually recognized.

Installation Conditions and Maintenance Requirements Consideration must be given to how the flow meter and accessories are to be installed and what is required for maintenance and calibration. Availability of parts and vendor support are also factors. Sometimes, even if a flow meter meets all the process conditions and performance requirements, it may not be suitable for a particular job because of difficulty of installation (e.g., requiring plant shutdown or too much space) or maintenance. Finally, the flow meter and its installation must be acceptable to operators and must be maintainable, preferably by local personnel.

Electrical Area Classification and Safety The meter and accessories must be checked for proper electrical area classification in accordance with the National Electrical Code (NEC), API Recommended Practice (RP) 500, and Section 300 of the Electrical Manual, as appropriate.

Chevron Corporation

500-7

July 1999



In addition, personnel safety should be considered for proper installation, especially in high temperature and high pressure applications.

Economics As shown in Figure 500-2, the cost of a flow meter ranges from a few hundred to tens of thousands of dollars. Sometimes, several types of flow meters could do the job—at significantly different costs. Installation is often a major portion of the total installed cost. Installation costs can vary significantly depending upon the flow meter selected, its location, installation method, auxiliary components needed, and utility requirements. Operating cost is another consideration. The energy cost resulting from permanent pressure loss is often significant. For example, pumping costs could be significant in larger sizes of differential pressure flow meters with higher permanent head loss. Where large meters are needed, therefore, the selection of a more expensive flow meter that has a lower permanent pressure loss coefficient or that is obstructionless might be justified. It is difficult to identify all the factors that an engineer must consider when selecting a flow meter for a particular application. In some situations, factors which are usually of minor significance may become relatively important. For custody transfer, the contract often specifies the type of flow meter, the minimum accuracy specifications, piping configuration, auxiliary equipment, and meter proving. Consult the API Manual of Petroleum Measurement Standards, Chevron Petroleum Measurement Manual, Part C, and Company specialists for this special application.

521 Differential Pressure (Head-type) Flow Meters Many types of differential pressure (d/p) flowmeters are available. Some of the commonly used types (including orifice plates, flow nozzles, venturi tubes, flow tubes, pitot tubes and pitot venturi tubes) are described in API RP 551, included in Volume 2 of this manual. Outside the U.S.A., ISO 5167 Standard is widely used for orifice meters. It also covers other types of d/p flow meters. This section gives information about these flow meters and describes three additional d/p flow meters: annubar, elbow, and wedge. When selecting d/p flow meters, note that the rangeability of the differential pressure transmitter is limited to approximately 10:1, while the flow turndown is usually limited to 3.5:1 due to the nature of a squared output. Differential pressure transmitters with different calibrated ranges can be “stacked” or installed in parallel across the orifice plate to achieve flow turndowns of 10:1 or better, when the orifice plate is operated within its operating constraints. However, care must be taken in measuring fractional inches of water column in developed differential.

July 1999

500-8

Chevron Corporation



Piping requirements (meter runs) are very important for the accuracy and repeatability of differential pressure flow meters.

Orifice Meters Orifice meters are perhaps the most commonly used d/p flow meters in refineries, production facilities, and chemical plants. The accuracy of an orifice meter depends on many factors, such as fluids, upstream piping configuration, Beta ratio (d/D, or β, the ratio of orifice bore to pipe size), Reynold’s number limits, and correction factors used. The overall accuracy ranges from ±0.5 to ±5%. The overall accuracy of orifice metering can be affected by: •

improper meter run design (i.e. insufficient flow conditioning or straight run of pipe)

•

incorrect installation of the plate (e.g., backward) and sensing line (e.g., trapped vapor in liquid service)

•

rough meter tube surface

•

problems of the plate (as described later)

•

high or low beta ratio

•

pulsating flow

•

liquid entrained in gas flow

•

transmitter problems (e.g., drift)

•

inadequate secondary element (e.g., should it be a chart recorder or an electronic flow computer)

•

gas sampling and analysis for specific gravity measurement used in the calculation

•

the inherent accuracy limitation (“uncertainty”) of the orifice discharge coefficient in the calculation. The orifice coefficients used in the industry standards were developed based on limited lab and field test data.

Improperly designed and installed orifice meter systems tend to under-measure flow because of the inherent characteristics and imperfections of the current industry standards. Orifice meters tend to over-measure flow in a pulsating-flow situation. Orifice Bore. The standard thin plate orifice comes in three basic throat configurations: concentric, eccentric, and segmental (Figure 500-3). The concentric throat design is the most common. Used for clean fluids in one fluid state, it is ideally suited for gas, steam, water, air, and clean liquid hydrocarbon and chemicals. The eccentric throat design is used primarily for liquids that contain gas. The eccentrically located orifice bore allows free passage of gas.

Chevron Corporation

500-9

July 1999


Fig. 500-3


Orifice Bores

Segmental throat design is used when liquids containing solids are measured. The opening allows for free passage of the solids because the bottom of the segment is tangential to the lower circumference of the pipe. The accuracy of concentric orifice meters is generally better than that of the eccentric and segmental meters. Edge. Two orifice edge configurations are in general use in the United States. The square-edge type is commonly used with gas, steam, and low viscosity liquids. The edge is sharp and can be straight or bevelled. The quadrant-edge type is a better choice for lower Reynold’s numbers (i.e., below 10,000). It is recommended if the viscosity of the liquid is above 10 centipoise (Cp) but does not exceed 50 Cp. The conical-edge type is similar to the quadrant-edge type but is not commonly used in the United States. As a “rule-of-thumb,” Reynold’s (Re) number constraints are as follows: Re Constraint Concentric (under 2 in.)

1000+

(2 in. and over)

5000 d+

Conical

250β < RD <200,000β

Eccentric

10,000 — 1,000,000

Integral

1000 d/D

Quadrant

250 — 3200 < RD <60,000 — 280,000

Segmental

10,000 — 1,000,000

The document entitled “Orifice Design Calculations by Mainframe Computer,” included as Appendix A, in Volume 1, Part I, of this manual, provides more information on applications and limitations of both designs. Orifice Plate Specifications. Orifice plate specifications are covered in API MPMS, Chapter 14.3/AGA Report No. 3 (1992) and include for example, the following:

July 1999

500-10

Chevron Corporation



•

Plate thickness: The minimum and maximum thickness depend on plate size, but 1/8 inch is the minimum.

•

Plate flatness tolerance: The maximum departure from flatness varies with orifice diameter (d) and size (D). It may be calculated as 0.005(D-d).

•

Roughness: The orifice plate surface roughness should not exceed 50 microinches.

•

Upstream edge: The square-edge (sharp-edge) type should not show a beam of light when checked with an orifice edge gage. Alternately, the square-edge type should not reflect a beam of light when reviewed without magnification. The orifice should not have a burred or feathered edge. Nicks on the edge can be expected to increase the measurement uncertainty.

•

Centering: The concentric orifice should have the bore centered within 3% of the inside diameter of both the upstream and downstream sections of the orifice meter run.

•

Beta ratio: A recent study by Chevron Petroleum Technology Company (CPTC) suggests that the optimal beta ratio is 0.4 to 0.55 for natural gas. Beta ratios departing from this range increase measurement uncertainty.

Bent orifice plate: According to a CPTC study, the measurement error of a bent orifice plate in gas service was up to 4.5% lower than the true flow rate. A bent orifice plate may be caused by plant upset, sudden valve opening or closure, or other reasons. Materials. Orifice plates are made of “zero-corrosion” materials, usually stainless steels (304SS, 316SS), monel, or other alloys, depending on the service. Pressure Tap Locations. The choice of pressure tap locations is dependent on many factors, such as flow conditions, pipe size, cost, and accuracy required. Orifice taps may be located as flange taps, corner taps, radius taps, vena contracta taps, and pipe taps. Their applications are discussed in Appendix A. Piping Specifications. API Standard 2530/AGA Report No. 3, ASME Standard MFC-3M-1984, and ISO 5167 specify meter run requirements for orifice meters used in custody transfers. Plant-type orifice meters that do not require ±0.5 to ±1.5% accuracy may require less stringent flow conditioning. The minimum length of upstream and downstream meter runs varies depending on piping configurations and beta ratio. Use of flow straightening vanes reduces the meter run length requirement. Consult the industry standards (API/AGA and ISO/ASME) for details. Secondary Elements. Secondary elements are required to perform one or more of the following functions:

Chevron Corporation

•

Convert the differential pressure measurement to electrical or digital output signal

•

Transmit the signal to a remote readout

500-11

July 1999


•


Record the signal

For higher accuracy, flow computers may be justified, especially when dealing with large volumes. Specifying and Sizing. Copies of the ISA S20 Specification Forms with instructions for specifying orifice meters and secondary elements are provided in Volume 2 of this manual. Sizing can be accomplished by any of the following three methods: 1.

By manual calculation—Methods of manual sizing are given in Appendix B, Volume I, Part I, of this manual. Understanding the manual sizing methods will give the engineer a better sense of what the computer programs are doing.

2.

By consulting vendors—Some vendors may size an orifice meter for you; some will charge a fee while others will not. If sizing is done by the user, ask vendors to confirm the calculation.

3.

By computer—A mainframe computer program called “ORIFICE” is available. The user’s manual, “Orifice Design Calculations by Mainframe Computer,” is included as Appendix A in Volume 1, Part I, of this manual. Users can run the calculations in two ways: –

Through VM mainframe computer program “ORIFICE.” The logon procedure is described in Appendix A. – Through Plant Equipment Information System (PEIS), which is used by some Refining locations in Chevron1. A number of PC-based (personal computer) orifice design calculations - both inside and outside Chevron - have been developed and some are being used. For example, –

– – –

The Orifice program by Pascagoula refinery as part of the plant meter database. The program was written based on the ISO/Miller equations as used in the mainframe program, and it also includes other calculations TechData (“Kyle”) ISA’s FLOWEL by Kenonics Controls Ltd. Gulf Publishing’s INSTRUCALC

These programs are in general suitable for plant orifice meter design calculations. Since it is not practical to keep track the accuracy of various PC-based orifice programs on a continuous basis, the following criteria may be helpful to evaluate the accuracy of a “new” PC-based orifice calculation. If a “new” PC-based orifice program is “based on ISO 5167/ASME-MFC 3M,” it can be tested to determine if the calculations and implementation are accurate and consistent with the ISO/ASME equations by comparing the results from the mainframe ORIFICE program. 1.

PEIS will be replaced by Meridium system in 1996 as part of the implementation of the Chevron AFIS.

July 1999

500-12

Chevron Corporation



In general, the following criteria can be used to evaluate the calculations by a new PC-based orifice program to determine if it is adequately accurate for plant operations: a.

Multiple test cases should be included covering normal and worst situations (e.g., low Re, small pipe size—near the limits of the equations in the industry standard).

b.

If the PC-based program is supposed to be based on ISO-ASME/Miller equations, the difference between the orifice discharge coefficients calculated by the PC-based program and by the mainframe ORIFICE program should be within 0.1%, and the difference on flow rates within 0.15%.

c.

If the PC-based program is supposed to be based on the API MPMS, chapter 14.3/AGA-3 (1992) equations, the orifice discharge coefficients calculated by the PC-based program and by the GAZ program (for gas flow) should be within 0.1% and the difference on flow rates within 0.15%.

Note For custody transfer gas flow calculations, the GAZ program developed by CPTC follows the equations in the current API Manual of Petroleum Measurement Standards (MPMS), Chapter 14.3 (and AGA 3). It can be used for custody transfer orifice calculations. The GAZ calculations can also be used to verify calculations by a gas supplier or buyer, or the calculations by commercially available electronic flow computers. Appendix A in Volume I, Part I of this manual provides test cases with results calculated by the mainframe VM ORIFICE program, for users who want to test their new PC-based programs. Integral Orifice. An orifice meter comprises an orifice installed integrally with a differential pressure transmitter (Figure 500-4). Although it provides for a compact installation, the overall accuracy is lower: ±2% to ±5%. Fig. 500-4

Chevron Corporation

Integral Orifice Assembly (Courtesy of the Foxboro Company)

500-13

July 1999



Integral orifices are used for 1½ inch, 1 inch and ½ inch pipe and will work on the same fluids as standard orifice assemblies. Care must be taken when using orifices under 0.125 inch bore; 0.125 inch and smaller bored orifices are vulnerable to plugging or dirty fluids. At present, no API, ASME, or ISO standards for integral orifice are available.

V-Cone Meters Like an orifice meter, a V-Cone meter is a differential type meter based on the principle of correlating the observed pressure drop due to an obstruction in the line to the volumetric flow rate. As the name implies, the obstruction is a V-shaped cone hanging in the center of the pipe (Figure 500-5). The size of the central V-Cone determines the beta ratio of the meter which is so defined as to have the same opening area as a same-size orifice meter at the same beta ratio. Fig. 500-5

Cutaway Drawing of V-Cone Differential Pressure Flow Meter (Courtesy of McCrometer)

Discharge Coefficient. While the discharge coefficient of an orifice meter (which is approximately 0.6) is calculated by an elaborate equation based on various parameters about the meter and the fluid properties, the discharge coefficient of a V-Cone meter (which is approximately 0.85) is usually supplied by the manufacturer based on factory calibration. Performance. One manufacturer claims a ±0.5% accuracy of reading for the primary element. Depending on secondary instrumentation, possible V-Cone meter system accuracy ranges from ±1% to ±2%. The V-Cone primary element exhibits repeatability to ±0.1% or better. The turndown of the V-Cone meter is claimed to be 15 to 1—exceeding the traditional differential head type meters. The V-Cone meter is not significantly affected by non-ideal flow conditions. This is in contrast to orifice meters which are known to be sensitive to non-ideal flow conditions that are usually caused by upstream elbows, valves, and other fittings. Tests conducted at CPTC in 1995 have showed that the V-Cone flow rate measure

July 1999

500-14

Chevron Corporation



ments are within 0.5% of the no-swirl baseline measurement even in highly swirling flow (e.g., swirl angles up to 40 degrees). Applications. Since their accuracy is not compromised by short pipe lengths between the disturbances (elbows, valves, etc.), V-cone meters may find process measurement applications at locations where long meter runs are not available due to space limitation, for example, on platforms and congested process plants. On an offshore platform, with shorter pipes, the space and weight requirement of a metering system can be reduced. This is a very important consideration for offshore platform construction and maintenance. The cramped platform deck area precludes long straight pipes to condition the gas flow before being measured by the traditional orifice meters.

Flow Nozzles The ASME Flow Nozzles are the most commonly used. The standard published by the American Society of Mechanical Engineers (ASME), ASME MFC-3M-1984, “Measurement of Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi,” covers the following three types of ASME Flow Nozzles. • • •

ASME high beta ratio nozzle: for beta between 0.50 and 0.80 ASME low beta ratio nozzle: for beta between 0.20 and 0.50 ASME throat tap flow nozzle: for beta between 0.25 and 0.50

These flow nozzles, shown in Figure 500-6, are elliptical-inlet, long-radius, wall-tap style nozzles. Another standard flow nozzle is the ISA 1932 type, which is widely used in Europe but uncommon in the U.S. This type of flow nozzle is described in the ISO Standard 5167. It is not covered in the ASME standard or in this manual. Applications and Limitations. Flow nozzles are generally used in steam (vapor) at high pipeline velocities (over 100 feet per second) as well as in water and light slurry. They are selected for these applications because their rigidity makes them more stable at higher temperatures and velocities than orifices. A special type of flow nozzle, the critical flow nozzle, is used to operate at critical (“choked” or “sonic”) flow for flow limiting or as a secondary flow standard, for example, in proving natural gas flow meters. When a flow nozzle and a square-edge orifice are sized to create the same differential at the same flow rate, the pressure loss of both is approximately the same. Performance Characteristics. Properly installed and calibrated flow nozzles are nearly equal in accuracy to sharp-edge orifices. As with orifice meters, a major factor affecting accuracy of flow nozzles is the uncertainty of the discharge coefficient. Typically, this uncertainty is ±2.0% for the high beta ratio and low beta ratio nozzles with wall taps. For ASME throat tap flow nozzles, consult ASME Performance Test Codes (PTC) 6 and 19.5 for more information. Sizing. The procedure for sizing a flow nozzle is similar to that for sizing an orifice meter. The ASME flow nozzle discharge coefficient for wall taps is given in the

Chevron Corporation

500-15

July 1999

ASME Nozzles (From ASME: Std. MFC-3M-1984, “Measurement of Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi.” Courtesy of ASME.


July 1999

Fig. 500-6

500-16 Instrumentation and Control Manual

Chevron Corporation



standard. The discharge coefficient of an ASME flow nozzle with throat taps is given in the ASME Performance Test Codes. Usually, major suppliers can either size the flow nozzle or verify your calculations. Installation. A flow nozzle can be mounted between two piping flanges, or welded in the pipe, or installed with the curved nozzle inlet facing upstream in a section of internally bored pipe. Figure 500-7 shows a typical type of mounting. Tap requirements are similar to those for orifice meters. Fig. 500-7

A Typical Flow Section with Flange-Type Flow Nozzle (Courtesy of Badger Meters, Inc.)

Piping Requirements. Like an orifice meter, a flow nozzle requires adequate straight pipes (“meter runs”) to achieve the best accuracy. Figure 500-8, from ASME Fluid Meters (1971), presents the piping requirements for orifice, flow nozzles, and venturi tubes (discussed below).

Venturi Tubes The most commonly used venturi tubes are of the ASME classical or Herschel style (Figure 500-9). Venturi tubes are used primarily in low static line pressure applications where high pressure recovery is important. They may be used in air, steam, water, gas, chemical, and light slurry. The long-term permanent head loss is typically between 10 to 14% of the measured differential. Figure 500-10 compares the overall pressure loss through several primary elements. Piping Requirements. Generally, a venturi tube requires about one-half of the upstream and downstream runs of an orifice meter (Figure 500-8). Figure 500-11 presents information on meter run requirements, reprinted from the ASME Fluid Meters publication (1971). The accuracy of a venturi tube is nearly equal to that of a thin plate orifice meter. For pipe Reynold’s numbers greater than 100,000, discharge coefficients for the

Chevron Corporation

500-17

July 1999


Fig. 500-8

July 1999


Piping Requirements for Orifices, Flow Nozzles, and Venturi Tubes (From ASME Fluid Meters, 1971. Courtesy of ASME.) (1 of 2)

500-18

Chevron Corporation


Fig. 500-8


Piping Requirements for Orifices, Flow Nozzles, and Venturi Tubes (From ASME Fluid Meters, 1971. Courtesy of ASME.) (2 of 2)

Chevron Corporation

500-19

July 1999

ASME Venturi Tube (From ASME Fluid Meters, 1971, “Piping Requirements for Venturi Tubes.” Courtesy of ASME.)


July 1999

Fig. 500-9


Chevron Corporation


Fig. 500-10 Pressure Loss vs. Beta Ratio (From ASME Fluid Meters, 1971, Courtesy of ASME.)


Fig. 500-11 Piping Requirements for Venturi Tubes (From ASME Fluid Meters, 1971, Courtesy of ASME.)

venturi tubes are constant and predictable to within ±0.5% to ±2%, depending on design. Venturi tubes can be sized using the ASME Standard MFC-3M-1984, or ISO Standard 5167. The discharge coefficient can be calculated if the following limitations are not exceeded: •

Size: 4 to 48 inches. Smaller sizes (down to 2 inches) or larger sizes (up to 84 inches) are also commercially available.

•

Beta ratio (d/D): between 0.40 and 0.75.

•

Reynold’s number (RD, based on pipe diameter): between 200,000 and 6,000,000.

Usually, major suppliers can size venturi tubes or verify calculations. For more information on sizing, construction, and installation, consult ASME Fluid Meters (1971) and ASME Standard MFC-3M-1984.

Pitot Tubes and Annubars These devices are used for larger pipe sizes when the fluid (gas, steam or liquid) is clean. For a pitot tube, the difference between the total (stagnation) pressure at the tip of the pitot tube and the static pressure outside the pitot tube follows the square-root relationship. Flow rates are thus measured.

Chevron Corporation

500-21

July 1999



An annubar is a multiple-ported pitot tube that attempts to provide the average velocity. A typical pitot tube is shown in Figure 500-12. Figure 500-13 is a schematic of an annubar. Fig. 500-12 Pitot Tube

Applications. The main reasons for choosing pitot tubes or annubars are as follows: • • •

Very low pressure loss Ease of installation (i.e., they may be inserted into existing piping or duct) Relatively low cost

Pitot tubes and annubars can be inserted into process lines through either flange or threaded connections. Various connections and mounting techniques are usually available in vendor technical brochures and manuals. The accuracy of a pitot tube is often lower than that of orifices and venturis. Typical accuracy is ±5% to ±10% on volume measurement. (The velocity measurement made by the pitot tubes may be accurate to within ± 1%, but this measurement is not often of interest.) The accuracy of an annubar may be better than that of the pitot tube. Some vendors claim ± 1% accuracy for the primary element over a turndown of 10 to 1. With the transmitter, the combined accuracy may be between ±1.5% and ±2%, provided a uniform velocity profile is available. Limitations. The minimum Reynold’s number based on pipe diameter is 10,000. Because annubars have pressure sensing ports facing flow, clean fluids are preferred. Because pitot tubes are single-point measuring devices, it is important to have a uniform velocity profile when using them. Figure 500-13 shows typical upstream and downstream piping requirements.

July 1999

500-22

Chevron Corporation



Fig. 500-13 Annubar Piping Requirements (Courtesy of Dieterich Standard)

Chevron Corporation

500-23

July 1999



Specifying and Sizing. Vendor technical data are usually used to select a pitot tube or to size and select an annubar. When specifying an annubar, its structural compatibility (i.e., the maximum allowable differential pressure) must be known. This information, which varies from one design to another, can be found in vendors’ product brochures. All annubar sensors are subject to resonance vibration. A vibration force perpendicular to the flow stream occurs due to the shedding of alternate vortices. It is important to check the frequency of the shedding against the natural resonance frequency of the sensor. A resonance calculation may be done either by the vendor or by the user, provided the equations and sensor design data are available. In any case, if the calculated differential pressure at maximum flow rate for an application is below 25% of the total maximum differential pressure allowed for the selected sensor, no resonance calculation is required.

Elbow Taps A typical elbow, or elbow tap, is shown in Figure 500-14. The pressure taps are located along a radius that is at a 45-degree angle to the face of the flange. Generally speaking, elbows that measure 4 inches and larger are of the “short radius” type, in which the radius of the elbow equals the inside diameter of the elbow. Elbows that measure less than 4 inches are usually the “long radius” type, in which the elbow radius equals one and one-half (1.5) times the inside diameter of the elbow. Fig. 500-14 Elbow

Because the elbow depends on centrifugal force to develop a differential pressure, measuring the flow of gas, with its inherent low density, is not practical. Liquids, with their much higher densities, develop reasonable differentials. The radius of curvature has a considerable effect on the differential. Thus, a short radius elbow will develop more differential than a long radius elbow. Elbows are not adversely affected by changes in viscosity because both pressure taps are located along the same diameter.

July 1999

500-24

Chevron Corporation



Uncalibrated elbows typically measure flow within ± 5% of upper range value. Better accuracy can be obtained for calibrated elbows. The minimum Reynold’s number (based on pipe diameter) for elbows is typically 10,000. Although elbows are often sized by vendors, a simple formula for measuring flow through an elbow is provided below. Q = 5.663 K D2 (h/gf) (Eq. 500-2)

where: Q = flow rate in gallons per minute (gpm) at flow temperature K = flow coefficient for given elbow D = pipe inside diameter of elbow in inches h = differential pressure in inches of water gf = specific gravity at flow conditions Figure 500-15 gives the flow coefficient K for various elbow sizes and radii of curvature. These are nominal values, subject to a tolerance of up to ±5%. Fig. 500-15 Elbow Flow Coefficient K Pipe Size (Inch)

Short Radius (R = D)

Long Radius (R = 1.5D)

1

0.850

0.975

2

0.800

0.920

3

0.780

0.895

4

0.760

0.875

6

0.730

0.840

8

0.710

0.815

10

0.695

0.800

12

0.685

0.785

24

0.660

0.760

Wedge Flow Meters A wedge element is nothing more than a “V” shaped restriction welded to the top of the pipe with a segmental opening at the bottom of the pipe to allow for free passage of any solids present in the process fluids. Wedge flow meters come in various designs. They are used in liquid or gas service where orifice meters are not suitable because of entrained solid or gas; for example, they are used in production field gathering systems. The chemical tee type (Figure 500-16) with its associated remote seal transmitter is used for liquids with

Chevron Corporation

500-25

July 1999



solids, dirty viscous fluids, and hot fluids which tend to solidify when cooled. The pipe tap type is used for clean viscous fluids which can be dead-ended in conventional lead lines. Fig. 500-16 Tee Type Wedge Flow Element Configurations (Courtesy of ABB Kent-Taylor)

The smaller integral elements, typically limited to ½ to 1½ inch size, are also used for clean viscous fluids. They are usually mounted directly to conventional d/p transmitters. Wedge flow meters are of the d/p type, in which the square root of the differential is linearly proportional to volume flow rate within a certain range (or Reynold’s number range). Wedge flow meters maintain this square-root relationship over much

July 1999

500-26

Chevron Corporation



lower values of Reynold’s numbers than many other differential producing elements. The flow coefficient for wedge flow meters remains relatively constant in the lower Reynold’s number ranges. For example, a 1½-inch segmental wedge flow coefficient is constant from Reynold’s numbers of 40,000 down to 2,000, as compared to the change on flow coefficient (K) for a 1½-inch orifice plate with flange taps (Figure 500-17). Fig. 500-17 Flow Coefficient K Compared to Reynold’s Number (Courtesy of ABB-Kent Taylor)

The accuracy of uncalibrated wedge flow meters is typically ±3% of upper range value. Calibrated meters may achieve ±0.5% accuracy. CPTC tested a 4-inch Combustion Engineering/Taylor Wedge meter in 1988 with a high gravity natural gas. The results revealed the following: •

Taylor Wedge meters should be properly calibrated against a suitable reference flow device before use. The flow coefficients supplied by the manufacturer cannot be relied upon to provide accurate (±0.5%) flow rates. Rather, they were found to undermeasure flow by 2 to 8%.

•

The meter showed acceptable repeatability and its flow coefficients have a weak dependency on flow rate, making flow rate calculation relatively simple.

•

The wafer-type wedge meter should be avoided if frequent change of the wedge element is required, because this is time-consuming.

Included in the following figures are a few typical sizing charts and equations, based on product brochures from vendors. For help in sizing and specifying a wedge flow meter, consult the manufacturer.

Chevron Corporation

•

Figure 500-18: Determination of Integral Wedge Element H/D Ratio and Differential Pressure (in three parts)

•

Figure 500-19: Integral Wedge Capacity Table

500-27

July 1999


•


Figure 500-20: Wedge Element Capacity Tables (½ inch to 12 inches)

Lo-loss Meters The Lo-Loss flow meter (Figure 500-21) is a proprietary design of Badger Meter Inc. (Tulsa, OK). Figure 500-22 shows a plan view of a PMT Lo-Loss flow meter. The Lo-Loss flow meter is a modification of the venturi design. The major advantages of the Lo-Loss are as follows. •

It produces high differentials with lower permanent pressure loss. This high pressure recovery feature can greatly enhance operating economy.

•

It is lower in weight.

•

It requires shorter overall laying lengths.

Several designs are available. Construction materials vary from exotic alloys to cast iron and plastic. Selection of material is based on operating temperature and fluid type. The accuracy of uncalibrated Lo-Loss flow meters is claimed to be ±0.75% of rate. Calibrated meters may achieve ±0.25%—according to Badger Meter. The manufacturer should be consulted for help in sizing and specifying a Lo-Loss flow meter. Outside the U.S.A., a similar meter called “Dall Tube” is also available. Installation Of Secondary Devices To optimize the accuracy of differential pressure flow meters, the secondary devices (field flow controllers, transmitters, recorders, and indicators) should be installed such that the inaccuracies of the flow measurement will be minimized. The installation must also provide reasonable access for routine operator functions and maintenance procedures. Secondary devices must be installed to keep the piping (tubing) between them and the primary element clear of trapped material. If the measured stream is hot, the measuring element of the secondary devices must be protected from the direct contact of the hot fluid. The secondary devices and the tubing connecting it to the process must also be protected against freezing (solidifying) of the measured or sealing fluid. Supports for secondary devices, piping and tubing methods, as well as sealing, purging and winterizing are covered in detail in Chapters 1500 and 1600 of this manual. In the following paragraphs, the word “flow meter” will be used to refer to the secondary devices. Flow Meter Location and Accessibility Location. Use the following guidelines to determine the location of the flow meter:

July 1999

500-28

Chevron Corporation



Fig. 500-18 Determination of Integral Wedge Element H/D Ratio and Differential Pressure (1 of 3) (Courtesy of ABBKent Taylor) H/D Ratio To determine the H/D ratio for a particular application, use the following procedure: 1. Multiply a given maximum flow rate by appropriate correction factor below: Liquid (U.S. gpm): Correction Factor = 1.0 (Eq. 500-3) 0.216 × T Gas (scfh): Correction Factor = --------------------P (Eq. 500-4) Steam (lb/hr):

Correction Factor = 0.3634 × V (Eq. 500-5)

where: P = process pressure in psia (psig + 14.7) T = process temperature in °R (°F + 460) V = specific volume of steam (cu ft/lb) 2. Match this corrected flow rate to values shown in the Capacity Table (Figures 500-19 or 500-20) to determine H/D ratio required to produce a desired differential pressure. Example: 8000 scfh of CO2 gas at 50 psig and 90°F in a 1-inch line. We want to choose an H/D ratio to produce a desired differential pressure of approximately 100 inches H2O. 0.216 × ( 90 + 460 ) 8000 × ------------------------------------------- = 14 ,689 scfh ( 50 + 14.7 )

(Eq. 500-6)

From the Capacity Table, a 1-inch integral Wedge element having an H/D of 0.50 will produce 100 inches H2O of differential pressure for 14,926 scfh. We would choose a 0.50 H/D ratio. Wedge element ratio (H/D) is ratio of Wedge element height to pipe inside diameter. Differential Pressure To determine the differential pressure for a particular application, use the following procedure: 1. To correct a value for differential pressure (h), read directly from Capacity Table for actual flowing conditions; use the following: Liquid: h = h (from chart) *gf (Eq. 500-7) Gas:

G × T × 0.216 h = h ( from chart ) × ----------------------------P (Eq. 500-8)

Steam: h = ( from chart ) × V × 0.3634 (Eq. 500-9)

Chevron Corporation

500-29

July 1999



Fig. 500-18 Determination of Integral Wedge Element H/D Ratio and Differential Pressure (2 of 3) (Courtesy of ABBKent Taylor) 2. To calculate exact differential pressure produced at known flow rate, use one of the following equations: q ( in U.S. gpm ) 2 Liquid: h = g f × ----------------------------------------5.663 × F a × Kd 2

(Eq. 500-10)

2 G×T Q ( in scfh ) Gas: h = ------------- -------------------------------------------------P 7727 × F × Y × Kd 2 a

(Eq. 500-11)

2 W ( in lb/hr ) Steam: h = V × ----------------------------------------------359 × F × Y × Kd 2 a

(Eq. 500-12)

where: Kd2 =

integral Wedge element coefficient

Fa = expansion factor G = specific gravity of gas h = differential pressure in inches H2O gf = specific gravity of liquid at flow conditions P = process pressure in psia (psig + 14.7) Q = flow rate of gas in scfh q = flow rate of liquid in U.S. gpm T = process temperature in °R (°F + 460) V = specific volume of steam in ft3/lb W = flow rate of steam in lb/hr Y = correction factor Example: Continuing with the preceding example, having selected an H/D ratio of 0.500 (Kd2 = 0.440), we find, from the charts below, that: Fa = 1.0002 Y = 0.9917 2 8000 1.517 × 550 Therefore: h = -------------------------- -----------------------------------------------------------------------64.7 7727 × 1.0002 × 0.9917 × 0.440 h =

July 1999

(Eq. 500-13)

72.57 inches H2O

500-30

Chevron Corporation



Fig. 500-18 Determination of Integral Wedge Element H/D Ratio and Differential Pressure (3 of 3) (Courtesy of ABBKent Taylor)

•

Field indicators, recorders and controller should be located adjacent to the process equipment of interest to the flow measurement and should be installed for easy operator viewing of the indicated or recorded flow value.

•

Flow transmitters should be specified with integral output meters and located so that the output meter is visible from the related control valve or drive. Where such orientation is not practical, a remote output meter should be installed adjacent to the related control valve or drive.

•

In an integrated operating facility, plot limit field indicators and recorders should be located adjacent to major process and utility plot-limit manifolds, away from areas of high fire or corrosive potential. One of the vital uses of the flow information displayed on these meters is to verify the function of emergency block valves and utility stream supply to plants during major plant upsets or mishaps.

Accessibility. To optimize flow measurement accuracy, flow meters should be installed as close to the process connection as possible. Process piping and tubing leads should be kept to a minimum.

Chevron Corporation

•

Smart flow transmitters, which require little routine maintenance, should be mounted adjacent to the flow elements in pipeways, provided they can be accessed using ladders, portable staging, or mechanical lifts.

•

Analog flow transmitters, indicators, recorders, and controllers require more frequent access. Consideration should be given to locating these instruments at grade or adjacent to a stairway-accessible platform. When these instruments are used for process balance or utilities “hard charge” service, frequent maintenance calibration checks are required, and accessibility is an absolute requirement in planning the installation.

500-31

July 1999



Fig. 500-19 Capacity Table for a Typical Integral Wedge (Courtesy of ABB Instrumentation)

July 1999

500-32

Chevron Corporation



Fig. 500-20 Capacity Table for Typical Wedge, ½" to 12" (In Gallons per Minute of Water) (Courtesy of ABB Instrumentation)

Chevron Corporation

500-33

July 1999



Fig. 500-21 Lo-Loss Flow Meter (Courtesy of Badger Meters, Inc.)

Fig. 500-22 Plan View of PMT Lo-Loss Meter (Courtesy of Badger Meters, Inc.)

July 1999

500-34

Chevron Corporation



Orientation of Flow Meters Use the following guidelines to determine the location of the flow meter relative to the primary measuring element. Gas Flow Meters. This section applies to non-condensing gases. (Refer to Standard Drawing No. GB-J1177 for details of installation of field mounted flow controllers, indicators, and recorders. Refer to Standard drawing No. GB-J1182 for details of installation for flow transmitters.) The objective of this orientation is to ensure that all condensed liquids will drain back into the process line. •

Locate the flow meter above the line. For vertical lines, locate the flow meter above the orifice plate.

•

For horizontal meter runs, use the flange taps located above the center line of the pipe.

•

Mount the flow meter and install the three- (or five-) valve manifold so that the flow meter and manifold are self-draining.

•

Route the tubing leads between the flow meter and the root valves so that they slope continuously downward toward the root valves and are, therefore, selfdraining. Ensure that there are no pockets where condensing vapors can get trapped.

(Refer to Gas Flow Meters Installed Below the Process Line, below, for installations where the layout of plant piping, structures, and equipment prevent use of the flow meter orientation described above.) Liquid Flow Meters (Low Pour Point). This section applies to liquids having a pour point significantly below the lowest ambient temperature encountered at the field location of the process line and the flow meter. (Refer to Standard Drawing No. GB-J1178 for details of installation of field mounted flow controllers, indicators, and recorders. Refer to Standard Drawing No. GB-J1183 for details of installation of for flow transmitters.) The objective of this orientation is to ensure that all non-condensable gases entrained or dissolved in the process liquid will vent back into the process line.

Chevron Corporation

•

Locate the flow meter below the line. For vertical lines, locate the flow meter below the orifice plate.

•

For horizontal meter runs, use the flange taps located below the center line of the pipe.

•

Mount the flow meter and install the three- (or five) valve manifold so that the flow meter and manifold are self-venting.

•

Route the tubing leads between the flow meter and the root valves so that they slope continuously upward toward the root valves and are, therefore, selfventing. Ensure that there are no pockets where gases can get trapped.

500-35

July 1999



Vapor Flow Meters. This section applies to steam and to gases which will condense at the highest expected ambient temperature. (Refer to Standard Drawing No. GB-J1181 for details of installation of field mounted flow controllers, indicators, and recorders. Refer to Standard Drawing No. GB-J1186 for details of installation for flow transmitters.) The objective of this orientation is to maintain a constant liquid static head (from the condensed steam or gas) on both impulse leads while ensuring that all entrained non-condensable gases will vent back to the process line. •

Locate the flow Meter below the line. For vertical lines, locate the flow meter below the orifice plate.

•


•

Install condensation chambers above the orifice plate root valve. Locate the condensate chambers so that they are self-draining back to the process line.

•

Mount the flow meter and install the three- (or five) valve manifold so that the flow meter and manifold are self-venting.

Route the tubing leads between the root valves and condensate chambers so that they slope continuously downward toward the root valves. Slope the tubing between the condensate chambers and the flow meter downward toward the meter. Ensure that the leads are self-venting with no pockets where vapor can be trapped.

Application Note Use of Condensation Chambers in Steam Service. Condensation chambers are required for all high-displacement meter bodies (bellows-type controllers, indicators, and recorders) and are recommended for all other steam flow meters. Condensation chambers perform the following functions in steam metering: •

Their large volume provides a reservoir for supplying condensate to the impulse leads on high displacement flow meter bodies.

•

Their large surface area permits the rapid generation of cool condensate for filling the impulse leads and permits the meter to be returned to service faster.

•

Excess condensate backflushes the lines between the chambers and the root valves, preventing the concentration and crystallization of soluble salts in impulse lines.

Flow meters installed with condensation chambers may require winterizing protection to prevent freezing of impulse leads and meter bodies. Refer to Section 1500, “Instrument Seals, Purges, and Winterizing,” for additional details. Gas Flow Meters Installed Below the Process Line. This section applies to noncondensing gas flow meters where the physical layout of plant piping, structures, or equipment requires that the flow meter be mounted below the line. In these installations, condensate accumulators must be installed.

July 1999

500-36

Chevron Corporation



(Refer to Stadard Drawing No. GB-J1180 for details of installation of field mounted flow controllers, indicators, and recorders. Refer to Standard Drawing No. GBJ1185 for details of installation for flow transmitters.) The objective of this orientation is to ensure that all condensed liquid will drain out of the impulse leads and into the condensate accumulators. The condensate will have to be periodically manually drained or pressured out of the accumulators. •


•

Locate condensate accumulators below the flow meter, in an area accessible for periodic draining of the condensate.

•

Mount the flow meter and install the three- (or five-) valve manifold so that the flow meter and manifold are self-draining to the condensate accumulators.

•

Route the tubing leads between the flow meter and the root valves so that they slope continuously downward toward the condensate accumulators and are, therefore, self-draining. Ensure that there are no pockets where condensing vapors can get trapped.

•

Route the accumulator lines to an environmentally acceptable drain system or sump. (Frequently accumulators are under pressure and the drain system is not. When such condensate is emptied into a low pressure system it can vaporize. The vapor must be vented into a closed system. This is especially a concern if the condensate contains dissolved toxic gases, e.g., H2S.)

Special Case Flow Meter Installations. Flow meter installations not described above require special installation designs to ensure that the accuracy of measurement and reliability are maintained. If the measured fluid has a low pour-point, contains solids, is viscous, etc., refer to Section 1500, “Instrument Seals, Purges, and Winterizing” to determine the type of special treatment that must be applied.

522 Positive Displacement (PD) Meters General Positive displacement (PD) meters measure volumetric flow directly by continuously separating (isolating) a flow stream into discrete volumetric segments and counting them. Therefore, PD meters are not inference-type meters, as are turbine meters and d/p meters. PD meters are often used in refineries, chemical plants, pipeline pump stations, marine loading terminals, and marketing truck loading racks. They are commonly used in liquid service, and in the gas industry, gas PD meters are used as domestic gas meters by utility companies. This section is primarily concerned with liquid PD meters for use by the petroleum and petrochemical industries.

Design and Construction of PD Meters The three basic subassemblies in a bare PD meter are the external housing, internal measuring element, and counter drive train.

Chevron Corporation

500-37

July 1999



External Housing. PD meters are usually built with inlet and outlet connections from ¼ inch to 16 inches in diameter and are designed to handle pressures to 1440 psi (600 pounds ANSI) and flow rates to 12,500 BPH. Typical housing materials are carbon steel, cast iron, ductile iron, aluminum, bronze, or stainless steel. Consult the manufacturer for help in selecting the proper material. PD meters may be of single- or double-case construction. The advantages of doublecase construction are that (1) piping stress is not transmitted to the measurement element; (2) the measuring element can be easily removed for service or inline flushing on startup; and (3) the differential pressure across the measuring element walls is minimal, thus eliminating the possibility of dimensional changes in the measuring element due to system pressure variations. Small meters of materials other than carbon steel are normally single-case. Meters over 6 inches almost always use carbon steel double-case construction. Internal Measuring Element. Six types of internal measuring elements (Figure 500-23) are in general use. These designs are discussed below in “Applications and Limitations.” All PD meters have some clearances between moving and stationary surfaces, with differential pressure across the clearances. Thus there will always be some fluid that bypasses the measuring chamber by “slipping” through these clearances. Counter Drive Train. A typical counter drive train is shown in Figure 500-24. The counter drive train consists of a gear train, a rotary shaft seal or magnetic coupling, and a calibrator. The gear ratio of the gear train is chosen to convert the fixed volume per revolution of the measuring element to some nominally convenient volume per revolution of the counter input shaft. The rotary shaft seal is required where the counter drive train penetrates the meter internal housing or measuring chamber. It is normally designed as a module or gland for easy access. Alternately, a magnetic drive coupling can be used instead of a packing gland to eliminate the need for frequent servicing of a shaft seal. A meter calibrator is a device containing gears for adjusting in fine (e.g., 0.05%) increments the output speed (RPM) from a meter counter drive train over a relatively narrow (e.g., 1.0%) total adjustment range. Use of a calibrator may produce error; therefore, it should be used only when necessary. A calibrator is necessary whenever the mechanical counter on the meter must register actual volume throughput. If a meter factor (ratio of actual/registered volume) is to be applied to the registered volume (a common practice with pipeline and bulk marine custody transfer meters), nominal 100% meter gearing and no calibrator (or a “dummy” calibrator) would typically be used. One or more pulse transmitters can be connected to the output shaft. The pulse transmitters are used to generate high and low resolution pulses to a remote receiver or computer that performs calculations and operates the metering system.

July 1999

500-38

Chevron Corporation



Fig. 500-23 Six Types of Measuring Elements (Figures 1 and 6 are Courtesy of Smith Meter, Inc.; Figures 3 and 4 are Courtesy of Brooks Instrument Division, Emerson Electric)

Chevron Corporation

500-39

July 1999



Fig. 500-24 Typical Counter Drive Train (Courtesy Smith Meter, Inc.)

Performance Characteristics of PD Meters PD meters can achieve better than ± 0.05% repeatability accuracy. Linearities of ± 0.15% to ± 0.5% are typical over flow ranges of 8:1 to 12:1, with a normal range of 10:1. Figure 500-25 shows the accuracy of a commonly used PD meter. Fig. 500-25 Typical Accuracy Curve (Courtesy Smith Meter, Inc.)

The dominant factor affecting meter accuracy is the amount of slippage (“bypass”) through the meter clearances. Ideally, if the amount of slippage remains constant, no error would occur and the meter would be perfectly repeatable. In reality, this is often not the case because of the variation of fluid dynamic properties (e.g., viscosity) and operating conditions (e.g., flow rate, pressure and temperature). To meet the specified linearity accuracy, a minimum flow rate is often required. Because of the viscosity effect, the lower the fluid viscosity, the higher the minimum flow rate.

July 1999

500-40

Chevron Corporation



Applications and Limitations of PD Meters Applications. Applications for various types of PD meters are as follows. Rotary vane. The rotary vane PD meter is the most common design. It is widely used for crude oil, mid-weight and heavy distillates, lube oil and additives, and chemicals of similar viscosity. The rotary vane meter is also used for asphalt (iron trim required). Many custody transfer measurements use rotary vane PD meters. Bi-rotary lobed. The applications for rotary lobed (also called bi-rotor) meters are similar to those for rotary vane meters. They are both considered to be the most accurate of all PD meter designs (see the discussion of “Performance Characteristics” above). Piston. Oscillating piston and rotary piston PD meters are used for chemicals, LPG, and batching processes with ±0.5% to ±1% accuracy. They are available in small sizes (1-inch to 3-inch sizes are typical) and are used as dispensing meters at service stations. Oval gear. Oval gear PD meters can be used for heavy oils with higher viscosity and for corrosive liquids but not for asphalt service. Typical designs allow maximum viscosity of 200 to 300 CP (1,000 to 1,599 SSU), at maximum flow rate. Higher viscosity, up to 1,000 to 1,500 CP (4,800 to 7,00 SSU), may be possible but the maximum flow rate has to decrease proportionally and special clearance must be used. Nutating disc. The nutating disc PD meter is primarily used in water meters and sometimes for chemical batchings. Accuracy is approximately ±1%. Rotating paddle. Rotating paddle PD meters often come in small sizes (1 to 2 inches). They are used in oil fields for water flooding. Accuracy is usually ± 2% or less. In summary, liquid PD meters can be used for measuring the flow of a wide range of fluids from LPG to asphalt. They are used instead of inferential type (e.g., orifice, turbine) meters where one or more of the following circumstances obtains. •

Requirement for higher accuracy

•

Requirement for wider rangeability (10:1)

•

Fluid viscosity too high for turbine meters

•

Lack of space for long meter runs as required by differential pressure meters and turbine meters (PD meters do not need flow straightening devices)

•

Specification by parties involved in custody transfer

Rotary vane and bi-rotor PD meters generally provide better accuracy than inferential flow meters, which include differential pressure (e.g., orifice) and velocity (e.g., turbine) meters.

Chevron Corporation

500-41

July 1999



Minimum accuracy requirements depend upon applications or contractual agreements. Typical minimum accuracy requirements for PD meters are shown in Figure 500-26. Fig. 500-26 Typical Minimum Accuracy Requirements for a PD Meter Applications

Accuracy Specifications

Custody Transfer

Non-Custody Transfer

Linearity as percent of flow rate

±0.15%

±1.0%

Repeatability as percent of flow rate

±0.05%

±0.05%

5:1 to 8:1

8:1 to 12:0

Rangeability

With the exception of the nutating disc meter, PD meters are ideal for viscous liquids not suitable for turbine meters. Generally speaking, PD meters are ideal for fluid viscosities greater than about 4 CP (typical for No. 2 fuel oil or 40° API gravity crude oil) because at these viscosities PD meters are relatively insensitive to flow rate and viscosity variations that may occur during operation. In other words, the accuracy and rangeability of a PD meter increase at higher viscosities (opposite to that of a turbine meter) because the meter factor shift due to bypass decreases with higher viscosities. Figure 500-27 shows graphically the effects of viscosity on PD meter accuracy. Fig. 500-27 Effects of Viscosity on PD Meter Accuracy

Heavier high viscosity oils will tend to cling to the internals of meters. Oils which are not effectively wiped off the blades and other moving parts of a PD meter will form a coating on the meter’s internal surfaces and reduce the volume of the measuring chamber.

July 1999

500-42

Chevron Corporation



The maximum viscosity for a typical 4-inch rotary vane PD meter is 400 CP (2,000 SSU), considered standard by the manufacturers. Higher viscosity is obtainable by using wider blade clearance. Until now, many pipeline meters have been designed and operated at 30% to 80% of the manufacturer’s specified maximum. This figure is too conservative, even for custody transfer applications. PD meters usually stay near or within the linearity specification at 70% to 100% of the maximum flow rate. At low flow rates (i.e., less than 20% of maximum), however, the linearity of a PD meter becomes unpredictable. Therefore, a more practical view is to allow the meters to operate between 20% to 90% of their maximum flow rates. A meter should operate at the flow rate at which it was calibrated (“proved”). It should be recalibrated whenever the flow rate changes more than ±10% of the maximum flow rate (full scale). In some places, the practice of recalibration is ±10% of the operating flow rate—a conservative and sometimes impractical practice. Lubricity has a significant effect on wear and meter life. Low lubricity usually causes increased friction between moving parts and accelerates wear. PD meters should run at lower rates with low lubricity fluids (note: turbine meters may perform better at higher rates with low lubricity fluids because the rotor tends to float better at higher rates). LPG, condensates, and water generally exhibit low lubricity. Most crude oils and diesel oils usually exhibit good lubricity. Gasolines and jet fuels are intermediate. Crude oils containing corrosive water and/or sand should flow through PD meters at fairly high rates and as continuously as possible to keep those materials from settling out in the bottom bearing. Cyclic operation with short run periods and long dormant periods tends to distort the meter factor and permit abrasive materials to build up in the bottom of the meter. If an automatic temperature compensator (ATC) is used, it may not be able to adequately compensate in short cycle operations because of lack of response time. Limitations. A PD meter is often more expensive than a turbine meter for the same rated capacity. This is generally true for sizes greater than 4 inches. For the same size meters, a turbine meter has a much higher rated capacity than a PD meter. For example, a typical 6-inch inline turbine meter has a maximum flow rate of about 3,000 GPM as compared to 1,000 GPM for a 6-inch rotating vane PD meter. The maximum pressure rating is usually lower for PD meters, as compared to turbine meters. PD meters require more maintenance than turbine meters and meters with no moving parts.

Specifying and Sizing A copy of the Instrument Society of America (ISA) “Standards and Practices for Instrumentation,” Specification Form S20.25, and instructions for completing it are

Chevron Corporation

500-43

July 1999



included in Volume 2 of this manual. When specifying a PD meter, consider the following criteria:

Meter Design •

Service—The type of liquid(s) and the nature and approximate amount of abrasives and corrosives must be specified. The application (process, custody transfer, pipeline, or load rack) and service (continuous or intermittent) must be known. Use the space at the bottom of the form as needed.

•

Temperature—Specify normal and maximum temperatures (in degrees F or C). These temperatures have direct impact on the range of viscosities at which the meter will be operating. Temperature and fluid type also affect selection of materials for the meter.

•

Size—The capacity of a PD meter may vary by manufacturer and type. Obtain information from the manufacturer’s technical specification. Consider multiplemeter configuration for cost effectiveness and operating flexibility. It is usually better to over-range than to under-range a PD meter if a choice is necessary.

•

Temperature and Pressure Rating—Specify normal and maximum temperatures and pressures so that proper material for and construction of the meter can be evaluated. Consider pressure drops across the strainer, deareator, meter, and other piping components to determine back pressure control.

•

Enclosure Class—In most cases, housings for electrical outputs are required to be explosionproof. As a minimum, the housings should be suitable for NEC Class I, Group D, Division 2 areas and should meet the specifications of NEC Article 500, Hazardous (Classified) Locations.

•

Materials of Construction—Consult the manufacturer concerning metallurgy. Smaller meters (e.g., under 4 inches) usually have iron (ductile or cast) or carbon steel housing. The larger sizes (6 inches and up) are usually carbon steel. A wide variety of trim materials is available for different types of internals. Trims are typically categorized as “standard trim,” “all iron trim,” or “LPG trim.” They are made of aluminum, bronze, iron, steel, stainless steel, or alloys. Seals are usually Viton or Teflon. Consult the manufacturer for recommended materials.

Counter When specifying counter components, do not put excessive weight on the meter and piping. Stress due to excessive load on the meter and piping may degrade measurement accuracy and cause fatigue. •

July 1999

Register Type—Specifying a register on the meter that indicates gross volume is a common practice. A second register on the meter stacking indicating “net” or “standard” volume may be necessary in remote custody transfer locations. This second register puts additional weight on the meter (and piping) and therefore should be avoided unless necessary.

500-44

Chevron Corporation



•

Totalizer Type—A local flow totalizer may not be required if the pulses only need to be totalized remotely, for example, in the control room by a panelmounted totalizer or a flow computer.

•

Reset, Capacity, Set-Stop—These items are particularly related to marketing terminal meters used for truck or tank car loading.

Fluid Data •

Minimum and Maximum Flow Rate—Specification of the minimum flow rate is a requirement for meeting the specified linearity. Maximum flow rate is based on continuous and intermittent ratings. Commonly used engineering units are gallons per minute (GPM) and barrels per hour (BPH).

•

Viscosity—Specify normal and maximum viscosity. These parameters are used to determine clearances, type of external housing (single or double) and type of rotary shaft seal. The engineering units and conversion factor are as follows. Cs = (Cp)/(SG) SSU × 0.21589 = Cs (approximate) (Eq. 500-14)

where: Cs = Kinematic viscosity in centistokes Cp = Absolute viscosity in centipoise SSU = Saybolt Universal Unit SG = Fluid Specific Gravity (water = 1 at 60°F)

Options

Chevron Corporation

•

Temperature Compensator—Temperature compensation (to provide standard volume) can be achieved by using a temperature sensor (e.g., RTD), a temperature transmitter, and a computer. This arrangement is the preferable method for large facilities. For remote, stand-alone LACT units, the traditional filled temperature bulb with an “automatic temperature compensator” (ATC) may be a better choice. The disadvantages of the bulb are that it is less accurate and it adds the weight of “stacking” on the meter.

•

Transmitter Type—Use a high resolution pulse transmitter when high resolution of the output signal is required. Sometimes, a high resolution pulse transmitter and a low resolution pulse transmitter are used—one to pace an automatic in-line sampler.

•

Transmitter Output—The format of the output signal must be compatible with the receiving unit that will scale it to volumetric units.

•

Air Eliminator—An air eliminator (de-areator) is needed where the piping configuration or operation may allow gas or air to be trapped in the metering system.

500-45

July 1999


•


Strainer—A strainer upstream of the meter must be provided to protect the meter from abrasives and foreign objects. Determine the pressure drop across the strainer. A differential pressure gage (or two pressure gages) often can be very useful in determining the pressure drop across the strainer. (Sometimes, a d/p transmitter is used to remotely monitor the pressure drop across the strainer.) Consult the meter manufacturer for type, size, basket design, and mesh (typically 4, 10, 20, 40, or 80 mesh are standard sizes).

Note Mounting Configuration—PD meters are usually installed horizontally. In some truck loading racks, PD meters are installed vertically because of space requirements.

Installation The API Manual of Petroleum Measurement Standards, Chapter 5, Section 2, provides a standard for a liquid PD meter system for custody transfer measurement. Some of the design criteria described in the standard are applicable in non-custody transfer measurements. Figure 500-28 is a typical schematic arrangement of a meter station with three positive displacement meters. Figure 500-29 shows a PD meter with peripheral equipment for LPG at a gas plant or marketing terminal. Fig. 500-28 Typical Schematic Arrangement of Meter Station with Three Displacement Meters

Meter Provers and Proving A liquid PD meter can be proved by a pipe prover, a master meter, or a tank prover. The frequency of proving depends upon the application of the meter. For example, a custody transfer meter should be proved more frequently than a process stream meter.

July 1999

500-46

Chevron Corporation



Fig. 500-29 PD Meter with Auxiliary Equipment for LPG at Gas Plant or Marketing Terminal

Refer to Chevron Petroleum Measurement Manual, Part C for more information on meter proving and meter provers. “Guidelines for Meter Prover Design,” included as Appendix C of this manual, and Specification ICM-MS-2498, also included in Volume 1, provide more details.

523 Turbine Flow Meters Unlike positive displacement (PD) meters, turbine meters are inferential-type meters that infer volumetric flow rate from the measurement of rotational movement (angular velocity) of a bladed rotor or impeller suspended in the flow stream. Two basic assumptions are necessary to obtain volumetric flow rate from a turbine meter: 1.

Volumetric flow rate is proportional to the average stream velocity.

2.

Average stream velocity is proportional to the rotor (blade) angular velocity.

As fluid passes smoothly through the meter, it causes the rotor to revolve with an angular velocity proportional to flow. The rotor blades, passing through the magnetic field of the pick-up, generate a pulsing voltage in the coil of the pick-up assembly. Each voltage represents a discrete volume. The frequency of the voltage generated is proportional to the rotor speed or the rate of flow. The number of pulses per unit of flow is termed the K factor. The actual K factor for each meter is determined by factory calibration and is provided with each meter. The total number of pulses, integrated over a period of time, represents the total volume metered. Turbine meters can be either in-line or insertion type. In-line turbine meters are commonly used in both liquid and gas service. Insertion turbine meters are used in gas, steam, air, water, or oil. In liquid service, they may be used to pace automatic line samplers. A bi-directional flow option is often available on each of these meters.

Chevron Corporation

500-47

July 1999



Turbine Meter vs. PD Meter To determine whether to use a turbine meter or a PD meter for a particular application, consider the following: 1.

Maximum Viscosity: If the maximum viscosity exceeds 10% of the Reference Viscosity for the sizes of turbine meters to be considered (Figure 500-30), PD meters would normally be more accurate (which is important in custody transfer).

Fig. 500-30 Effect of Viscosity on Linear Range of Meter (Used with permission from Smith Meter, Inc.)

If only low-viscosity refined products such as propane, gasoline, kerosene, or diesel are being metered, turbine meters normally would be selected because of their longer service life for continuous duty operation and because their accuracy is normally equal to or better than that of PD meters for these types of products. Field tests indicate that dual-bladed helical turbine meters (see Figure 500-32) are able to maintain good repeatability and linearity for crude oils with a viscosity range of 4 to 32 Cs. Figure 500-31 is a selection guide based on flow rate and viscosity (in centipoise) for PD meters and conventional turbine meters.

July 1999

2.

Maximum Flow Rate: If the total flow rate required at the metering station exceeds about 100,000 BPH, many parallel PD meter runs of the maximum size (typically 16 inches) would be required. In this case, turbine meters would normally be considered. Generally speaking, mid-size and large-size PD meters are more expensive than turbine meters for the same design flow rate.

3.

Maximum Pressure: If the meter pressure rating must be greater than 600 pounds ANSI (1440 psig), a PD meter cannot be used.

500-48

Chevron Corporation



Fig. 500-31 P/D and Conventional Turbine Meter Selection Guide (Used with permission from Smith Meter, Inc.)

4.

Back Pressure: For a turbine meter, the back pressure (i.e., meter outlet pressure) should be at least 25 psig at maximum flow rate for low vapor pressure fluids, and 1.25 times the maximum vapor pressure for high vapor pressure fluids (e.g., propane). Low back pressure is of particular concern if the turbine meter is to be located close to a receiving tank. With PD meters, the back pressure must only exceed the vapor pressure by a small amount.

5.

High Paraffin Content: Turbine meters should not be used with liquids containing paraffin or other similar substances that can precipitate out on the surfaces of the meter, changing its cross-sectional flow area. Waxy crudes, for example are not suitable for conventional turbine meters.

In-line Liquid Turbine Meters Design and Construction. The three basic sub-assemblies in an in-line liquid turbine meter are the housing, internal parts, and detector. (See Figure 500-32). The housing is normally constructed of a flanged pipe spool in sizes from ¼ inch to 24 inches with pressure ratings from 150 pounds to 2400 pounds ANSI (275 to 6000 psi working pressure). Internal parts include the bladed rotor suspended or supported on a bearing and shaft and the upstream and downstream stators. The detector normally consists of a magnetic or RF pick-up coil, explosionproof junction box, and optional pre-amplifier (if the signal cable exceeds a certain length). Two types of bearing design are common: journal bearing and ball bearing. Journal bearings are usually made of cemented tungsten carbide. Ball bearings are made of stainless steel or special plastic material.

Chevron Corporation

500-49

July 1999



Fig. 500-32 Turbine Meters

Conventional In-Line Liquid Turbine Meter (Courtesy of Daniel Industries, Inc.)

Dual-Bladed Helical Turbine Meter (Courtesy of Faure Herman Meter, Inc.)

July 1999

500-50

Chevron Corporation



Journal-bearing meters are often used in process plants, pipelines, and marine loading terminals for more continuous services. Ball-bearing meters have been mainly used for intermittent services (e.g., at truck loading racks). Performance Characteristics. Linearity accuracy of ±0.15% over the reduced flow range of 7:1, ±0.25% over the normal flow range of 10:1, and ±0.5% over the extended flow range (12:1) is typical. Repeatability of ±0.05% to ±0.1% is commonly seen in vendor’s specifications. Rangeability may be increased to above 12:1, but overall accuracy will decrease. Minimum accuracy requirements depend upon the applications and contractual agreements. Typically, the minimum accuracy specifications are as shown in Figure 500-33. Fig. 500-33 Typical Minimum Accuracy Specifications for an In-Line Liquid Turbine Meter Applications Custody Transfer

<2½-in. Sizes and Non-Custody Transfer

Linearity as percent over normal flow range

±0.15%

±0.5%

Repeatability as percent over normal flow range

±0.05%

±0.05%

4:1 to 7:1

7:1 to 12:1

Accuracy Specifications

Rangeability

Figure 500-34 illustrates the effect of fluid dynamic properties, as implied by the Reynold’s number (Re): ( Velocity ) ( Diameter ) ( Density ) Re = ---------------------------------------------------------------------------( Viscosity ) (Eq. 500-15)

Design temperature of turbine meters varies from −40°F to +500°F depending upon the material of construction. Pressure ranges are usually available for all ANSI and API flange ratings. Fig. 500-34 Universal Turbine Meter Performance Curve

Chevron Corporation

500-51

July 1999



Applications. In-line liquid turbine meters are used primarily because of their accuracy and rangeability. They are used in light, thin (not viscous) products, chemicals, and crude oil custody transfer measurement. For example, in-line liquid turbine meters are used for in-line product blending, additives, NGL, LPG, ammonia, and high API gravity crude oil. Many industrial users consider minimum allowable flow rates to be in the 30% to 40% zone, and maximum flow rates to be up to 100% of manufacturer’s specified maximum. Some meter manufacturers consider 50% to be a reliable minimum, with 40% an absolute minimum. Manufacturers also claim that for many applications turbine meters can run at 120% (the “extended range”). At this time, the data are not sufficient to make a quantitative evaluation to run at 120%; therefore, use as a maximum 40% to 100% of the manufacturer’s specified maximum, with the qualification that the maximum may be extended to 120% in clean service. Limitations. The limitations of liquid turbine meters include the following: • • • • • •

Higher cost (compared to orifice meters) Susceptibility to wear or damage due to dirty or nonlubricating streams Unsuitability for viscous liquids Susceptibility to rotor damage from over-speeding Need for flow straightening devices and air-eliminators Need for more maintenance than other nonmoving-type meters

Specifying and Sizing. A copy of the ISA Specification Form (S20.24) with instructions is included in Volume 2 of this manual. When specifying a turbine meter, consider the following criteria: • • • • • • • • •

Minimum back pressure Materials for internal parts (bearing, blade, etc.) and body Strainer and air-eliminator requirements Pressure drop of meter run Maximum and minimum flow rates Proper meter size Pulse security (i.e., single- or dual-pulse transmitters) Pre-amp and resolution of pulse transmitter(s) Uni- or bi-directional flow

Minimum Back Pressure: The meter must be protected from air trapped and entrained and from vapor developed in the fluid as a result of vaporization or cavitation. Therefore, the metering system must, under all conditions, have a pressure sufficiently high due to flow separation. The API Standard recommends that the minimum back pressure (BP) be as follows: BP = 2 (dP) + 1.25 (Vp) (Eq. 500-16)

July 1999

500-52

Chevron Corporation



where: Bp = Minimum back pressure at a distance of 5 pipe diameters downstream of the meter. dP = Pressure drop through the meter at the maximum rate of flow. Vp = Absolute vapor pressure of the liquid at the highest operating temperature. Figure 500-35 shows how the meter K factor could be affected if the back pressure is too low. Fig. 500-35 TM Performance with Cavitation

Maximum Viscosity: The maximum viscosity for liquid turbine meters varies by make, model, and options. Consult the manufacturer for the limits. Figure 500-31 can be used as a general guide for preliminary selection of a meter. Turbine meter manufacturers often quote higher viscosity limits. Field tests conducted by a number of oil companies in early to mid 1990s indicated that the dual-bladed helical turbine meters were able to maintain good repeatability (±0.05%), and linearity (±0.15%) over a 4 to 1 turndown for crude oils with viscosity ranges of 4 Cs to 32 Cs. One of the meters was even tested with crude oil of high viscosity of 115 Cs. The helical meter outperformed the conventional turbine meters, especially in moderate to high viscosity applications, and was found less susceptible to filamentary debris. Specific Gravity: Changes in specific gravity, or density, do not affect the average meter K factor value; however, such changes do affect the overall linear range of the meter. The maximum flow rate rating increases with decreased specific gravity. The minimum flow rate rating is lower with higher specific gravity. Strainer: Turbine meters should be protected by use of an adequate strainer upstream of the meter run. The strainer mesh depends on the meter size and the cleanliness and viscosity of the liquid.

Chevron Corporation

500-53

July 1999


☞


Caution The strainer is used to protect the meter from damaging objects in the liquid stream; it is not intended for filtering the liquid. To use a strainer as a liquid cleaning filter requires an extremely fine mesh screen that will cause a high pressure drop. Consult the manufacturer for the proper strainer size, basket mesh, and pressure drop. Air Eliminators (Deareators): The presence of gas or vapor, either bulk or entrained, degrades meter performance and can cause damage. An air eliminator upstream of the meter run is required where air or vapors are present. Proper Meter Size: Usually, turbine meter manufacturers provide capacity tables in their technical catalogs. Find out the pressure drop through the meter to calculate the back pressure requirement. Figure 500-36 shows typical pressure drop at different flow rates. Usually, the pressure drop at maximum flow ranges from 4 to 6 psi.

Fig. 500-36 Pressure Drop through Meter vs. Flow Rate for Typical Liquid Turbine Meters (Courtesy of Daniel Industries, Inc.)

Flow Conditioner: The accuracy of a turbine meter can be adversely affected by swirls and irregular flow patterns caused by valves, elbows, and other pipe fittings. Sufficient length of straight pipe spools and/or straightening vanes upstream and downstream of the meter should be provided (see “Installation” below).

July 1999

500-54

Chevron Corporation



Installation. The API Manual of Petroleum Measurement Standards (MPMS), Chapter 5, Section 3, “Turbine Meters,” should be used as the standard for proper installation. Turbine meters are usually furnished with journal bearings. They are usually installed horizontally. Whenever vertical installation is needed, consult the manufacturer. Generally, only ball-bearing types can be installed vertically. Turbine meters require proper upstream and downstream flow conditioning assemblies to reduce or eliminate the swirl or irregular flow patterns. API MPMS recommends the use of a pipe spool of 10 pipe diameters containing a straightening vane section immediately upstream of the meter and a pipe spool of 5 pipe diameters downstream of the meter (Figure 500-37). Fig. 500-37 Recommended Flow Conditioning Assemblies with Straightening Elements (From Manual of Petroleum Measurement Standards Ch.5, Sec. 3, Courtesy of American Petroleum Institute)

In some process plant applications where high accuracy is not required and the available space makes the meter runs described above impractical, shorter meter runs may be used—at the cost of accuracy. As stated earlier, a strainer is required to protect the meter from damage from foreign objects in the fluid. A deareator is needed if air may be trapped in the meter. A back pressure control valve should be provided to maintain proper back pressure to avoid cavitation. Figure 500-38 shows a typical piping schematic with connections for proving. Proving. Liquid turbine meters can be proved by a pipe prover, a master meter, or a meter prover.

Chevron Corporation

500-55

July 1999



Fig. 500-38 Turbine Meter System Schematic Diagram (Used with permission from Smith Meter, Inc.)

The frequency of proving depends upon the application of the meter. For example, custody transfer meters are proved more often than process stream meters. Refer to Chevron Petroleum Measurement Manual, Part C for more information or meter proving and meter provers. The API Manual of Petroleum Measurement Standards, Chapter 4, “Proving Systems,” provides detailed requirements and procedures for meter proving. ICM-MS-2498, “Stationary Meter Prover—Displacement Type Conventional Pipe Prover,” and Appendix C of this manual, “Guidelines for Meter Prover Design,” provide more details on pipe provers. These guidelines cover the selection, sizing, and proving procedures for both conventional pipe provers. The specification covers requirements for design, fabrication, inspection, and calibration of conventional uniand bi-directional pipe provers.

In-line Gas Turbine Meters Design and Construction. Axial-flow in-line gas turbine meters are commonly used in the industry. The American Gas Association standard designated as “Report 7” (AGA 7) should be used to specify the design and construction of this type of meter. In an axial-flow gas turbine meter (Figure 500-39), gas entering the meter increases in velocity through the annular passage formed by the nose cone and the interior wall of the body. The movement of gas over the angled rotor blades imparts a force to the rotor, causing it to revolve. The rotational speed is directly proportional to the flow rate. The actual rotational speed is a function of the passageway size and shape and the rotor design. It is also dependent upon the load imposed due to internal mechanical friction, external loading, and dynamic properties (e.g., gas density and fluid drag) of the fluid.

July 1999

500-56

Chevron Corporation



Fig. 500-39 Axial Flow Gas Turbine Meters (Courtesy of Equimeter, Inc.)

Rotor bearings should be stainless steel with factory-sealed or continuous lubrication. Rotor hubs and blades should be constructed in one piece and from the same material. Gas turbine meters are usually available in sizes ranging from ¼ inch to 24 inches and of various materials (aluminum, carbon steel, stainless steel, or alloy body with metal or non-metal rotors). They are constructed to meet ANSI ratings from 150 to 2500 pounds. Performance Characteristics. Repeatability accuracy of ± 0.1% to ± 0.15% over the normal flow range is usually required. Rangeability of most gas turbine meters is 10:1 or higher, with decreased accuracy. The overall accuracy should be better than ± 1.0% of the flow rate for a calibrated meter. Applications. In-line gas turbine meters are used in gas pipelines, gas plants, largevolume production facilities, and refineries. Generally speaking, a properly installed gas turbine meter offers better accuracy than an orifice meter. In addition, turbine meters offer better rangeability (10:1 or higher) than differential-pressure meters (4:1). Limitations. Like liquid turbine meters, gas turbine meters have the following limitations: •

Susceptibility to wear or damage due to dirty streams

•

Higher cost (compared to orifice meters; this may not be true if a senior orifice fitting is needed)

•

Need for flow straightening devices (meter runs and straightening vanes which are required for orifice and other d/p type meters)

•

Susceptibility to rotor damage from over-speeding

•

Need for more maintenance than other non-moving-type meters

Specifying and Sizing. The considerations for specifying a gas turbine meter are similar to those for liquid turbine meters. Gas turbine meters are generally designed for a maximum flow rate not to exceed a certain rotor speed in rpm. This maximum flow rate of the meter remains the same

Chevron Corporation

500-57

July 1999



for all pressures within the stated maximum meter operating pressure, i.e., the maximum rotor speed remains the same regardless of the pressure. Manufacturers’ specifications should be consulted when sizing a gas turbine meter. Installation. AGA 7, Section 3, describes recommended installation for in-line gas turbine meters. Figure 500-40, which illustrates various piping configurations, is based on AGA 7. Fig. 500-40 Piping Configurations for In-Line Gas Turbine Meters (1 of 2)

July 1999

500-58

Chevron Corporation



Fig. 500-40 Piping Configurations for In-Line Gas Turbine Meters (2 of 2)

Proving. As described in AGA 7, a gas turbine meter can be calibrated and proved by a bell prover, a master meter, a sonic-nozzle prover, or a critical-flow-orifice prover. Generally speaking, both the critical-flow-orifice and sonic-nozzle are capable of calibration at operating conditions to an accuracy of ±0.25% of actual flow rate.

Chevron Corporation

500-59

July 1999



The method and frequency of proving depends upon the availability of proving device(s), installation configurations, and accuracy requirements.

Insertion Turbine Meters Construction. An insertion turbine meter (Figure 500-41) typically consists of two major components: (1) a rotor head and (2) a retractor. The rotor head is inserted in the process line. The retractor provides the basic means of support for the rotor head, mounting fitting, pressure seal, pick-up coil, and pre-amp. Fig. 500-41 Typical Insertion Turbine Meters (Courtesy of the FMC Corporation)

Principle. The underlying principle of operation for insertion turbine meters assumes the average velocity of flow can be inferred from a measurement of the local velocity. In other words, it is assumed that the angular velocity of the rotor is proportional to the average flow velocity in an ideal fluid. In reality, the flow rate, Q, is related to the local velocity by the following equation: Q = (A) (Fp) (Fo) (Vy) (Eq. 500-17)

where: Q = flow rate, in ft3/min A = Insertion area, in ft2

July 1999

500-60

Chevron Corporation



Fp = Profile Factor, which relates to the fluid dynamic properties (Reynold’s number) Fo = Obstruction factor due to the effect of the rotor Vy = Local velocity, in ft/min The average velocity can be measured at a position at which it equals the local velocity (i.e., at “critical position”). The average velocity can also be measured at the centerline position because it is proportional to the centerline velocity. Installation. The rotor can be located in either of two positions: •

Center Position: This position is usually chosen for large pipe sizes. The basis for selecting center positioning varies among manufacturers. As a general rule, center positioning is used when the diameter of the velocity detector (rotor) becomes significant compared to the cross-sectional area of the pipe. A major advantage of center positioning is the comparatively low sensitivity to positioning error, which has one-to-one impact on linearity accuracy.

•

Critical Position: This position is usually chosen for small pipe sizes. By definition, the critical position is that position in the pipe at which the local velocity is equal to the average velocity. Critically positioned meters are nearly independent of flow. For a wide range of fluids and flow rates, critically positioned meters usually exhibit comparable or better linearity than in-line meters.

Obviously, center position is easy to locate. Two ways to locate the critical position are to calculate the location based on the Reynold’s number (equations are not shown in this manual) or to consult with the manufacturer. Usually, major suppliers are able to provide users with this information if they have the flow conditions and pipe size. Both “fixed” and “retractable” insertion turbine meters are available. Fixed type meters should only be installed or removed from service when the meters are isolated from the fluid and de-pressured (e.g., for a new plant, during plant shutdown, blind-off, or block-and-bleed). Retractable type meters are designed to allow installation or removal under normal operating conditions. Sometimes, the meters can be installed by hot-trapping. Performance Characteristics. Typically, insertion turbine meters can achieve ±1% of full scale. Rangeability is 10:1 to 50:1. Some turbine rotors can operate at temperatures as high as 750 degrees Fahrenheit. Applications and Limitations. Insertion turbine meters are used to measure steam flow, relief gas in a header, and other gas and air flow parameters. Liquid insertion turbine meters have been used to pace automatic line samplers in custody transfer applications. They are acceptable only if the liquid is clean. The limitations of insertion turbine meters are similar to those for in-line meters.

Chevron Corporation

500-61

July 1999



Specifying and Sizing. Many of the criteria described for in-line turbine meters are applicable for insertion turbine meters. Consult the manufacturer for proper selection and size.

524 Magnetic Flow Meters An electromagnetic flow meter measures volumetric liquid flow rate. Its fundamental operating principle is based on Faraday’s law of electromagnetic induction. That is, the voltage induced across any conductor, as it moves through a magnetic field at right angles to the lines of flux, is proportional to the velocity of that conductor. In an in-line magnetic flow meter, a straight tube is located in a magnetic field created by the coils on the tube. Electrodes, mounted in a plane at right angles to the magnetic field, contact the liquid and act like brushes in a generator. They provide a means by which the voltage induced in the moving liquid is brought out for external measurement. This induced voltage is the flow signal. It is linearly proportional to the velocity of the liquid. For a full pipe, the measured voltage is a direct indication of the volume flow rate through the pipe. This relation can be expressed by the formula: E =k(v)(B)(D) (Eq. 500-18)

where: E = signal generated by conductive liquid flow through the pipe k = a calibration constant v = velocity of fluid at the sensing point B = density of magnetic flux generated by means of field coils D = pipe diameter Two types of designs are commercially available today: 1.

In-line magnetic flow meter—Usually a spool flanged on the process pipe.

2.

Insertion magnetic flow meter—Flow sensor mounted on a probe (Figure 500-42) to measure local velocity. The local velocity is considered representative.

Applications and Limitations Applications. Magnetic flow meters are typically used to measure slurries, acid streams, and viscous fluids (conductive). In the petroleum and chemical industries, magnetic flow meters are used in water, waste water (effluent) and sludge, acids, alkalies, and other chemicals.

July 1999

500-62

Chevron Corporation



Fig. 500-42 Insertion Magnetic Flow Meter (Courtesy of Monitek Technologies)

One of the advantages of magnetic flow meters is that they are obstructionless; they have no head loss except due to pipe friction. In addition, they may be used in slurries and in dirty and corrosive liquids. Magnetic flow meters have a wide flow range and are independent of the density, viscosity, and static pressure of the fluid. Some probe-type magnetic flow meters are designed for open channel (or partially filled pipe) flow measurement, as discussed in Section 531. Limitations. In-line magnetic flow meters are relatively expensive, especially if line size is large. The accuracy of magnetic flow meters may be affected by change of fluids, and the process fluid typically must have a minimum conductivity of 5 micromhos per centimeter. Some models may have a lower minimum conductivity (2 micromhos per centimeter). Most aqueous liquids can easily meet the minimum conductivity requirement. Pure liquid organic chemicals may or may not be sufficiently conductive. Crude oil and refined petroleum products are electrically nonconductive and therefore not suitable for use with magnetic flow meters. Magnetic flow meters cannot be calibrated in place.

Performance Characteristics Accuracies of ±0.5% to ±1% of full scale are typical. Linearities of ±0.5% of full scale and repeatabilities of ±0.2% to ±0.5% are also common. Rangeability may be as high as 30:1.

Chevron Corporation

500-63

July 1999



The results from a test conducted in 1985 by the International Instrument User’s Association (SIREP — WIB) on an in-line magmeter suggest that: 1.

The repeatability may achieve ±0.03% of rate above 10% of full scale.

2.

The accuracy may achieve ±0.03% of full scale, ±0.35% of rate at 10% of full scale, and ±0.65% of rate at 10% of full scale.

3.

The accuracy can be affected by the following factors: – – – –

Shift of frequency of power supply Distorted flow profile due to upstream valves and piping elements Misalignment of the flow meter Magnetic interference from external magnetic field

The installed accuracy may, therefore, be affected by these factors. At present, commercially available insertion magmeters are less accurate than inline meters. In a test on an insertion probe magmeter by SIREP, the accuracy of the flow meter was approximately ±5% at 1 foot per second or less. Insertion magmeters can be used in open channels or partially filled pipes (Section 531) where high accuracy is often not a concern.

Specifying and Sizing Consult the manufacturer when specifying and sizing magnetic flow meters. A copy of the ISA Specification Form with instructions is included in Volume 2 of this manual.

Installation Installation methods may vary according to the type and make of the magnetic flow meter. The manufacturer should be consulted to ensure proper installation.

525 Ultrasonic Flow Meters The two most common principles of operations used by ultrasonic flow meter manufacturers are “transit time” and “Doppler effect.” A schematic of a typical flow meter is shown in Figure 500-43.

Transit Time Ultrasonic Flow Meters Principle of Operation. A “transit time” ultrasonic flow meter uses two transducers installed on the process line at an angle (usually 45 degrees) to the flow (Figure 500-44). The transducers function as pulse transmitters and receivers. Sonic pulses are alternately sent and received between the transducers in opposite directions. Because the pulses travelling upstream (i.e., against flow) take more time to reach the receiver than those traveling downstream (i.e., with flow), the alternate pulses yield a frequency difference. This difference is proportional to the average flow velocity, independent of the specific fluid being measured.

July 1999

500-64

Chevron Corporation



Fig. 500-43 Ultrasonic Flow Meter Schematic

Fig. 500-44 Transit Time Ultrasonic Flow Meters

Because the measurement is based on frequency shift or time delay, the transit time type of flow meter is also called “time-of-flight” or “contrapropagating.” Applications. Transit time ultrasonic flow meters are used in gas and liquid service. In gas service, they are used in natural gas (not for custody transfer measurement) and air. One application is for flare flow measurement, discussed in Section 532. In liquid service, transit time flow meters are limited to relatively clean liquids with less than 0.2% (by volume) of entrained gas bubbles and solid particles combined. The flow meters rely upon ultrasonic signals traversing the pipe; therefore, the liquid must be relatively free of solids and air bubbles. Bubbles in the flow stream seem to cause more attenuation of the acoustic signals than solids do. Consult the manufacturer to determine the type of flow meter to use.

Chevron Corporation

500-65

July 1999



Limitations. A uniform velocity profile is required for both gas and liquid flow meters. Typically, they perform better at turbulent flow with a Reynold’s number greater than 100,000 and fluid velocity greater than 1 foot per second. Like turbine meters, both gas and liquid flow meters require a minimum straight pipe (meter run) from valves, tees, elbows, pumps, compressors, etc., to meet performance specifications. Typically, 10 to 20 diameters upstream and 5 diameters downstream are required. Performance Characteristics. For liquid flow meters, accuracy is typically ±1.0 in turbulent flow (a Reynold’s number greater than 100,000 and a fluid velocity greater than 1 foot per second). The accuracy would be reduced to ±2.5% of rate over a 10:1 flow range if these conditions are not met. Repeatability is typically better than ±0.5% of rate, depending upon velocity range and manufacturer. The accuracy of gas flow meters is typically ±2% to 5% of rate over 50:1 flow range. Repeatability is ±0.5% of rate. The accuracy of ultrasonic flow meters may be affected by changes in viscosity and temperature. A test conducted by an independent testing agency in 1983 revealed that higher liquid viscosities caused positive shifts in the output. In general, transit time ultrasonic flow meters are more accurate than Doppler meters. From 1973 to 1974, the Company and another major oil company conducted separate tests on ultrasonic flow meters. The Company tested a flow meter against two 16-inch positive displacement meters for custody transfer at a liquid pipeline terminal. The test results indicate that the ultrasonic flow meter tested “meets its factory specified ±0.5% accuracy,” but it “is not accurate enough for normal (1/20%) custody transfer measurement. It is, however, a reliable, rangeable, low maintenance meter for applications such as in-line blending or run-down accounting, where its apparent 1% accuracy is acceptable.” It was found that “the meter factor shifted with flow, viscosity and gravity.” In another test conducted in the El Segundo refinery in the mid-1970s, an ultrasonic flow meter produced erroneous output when the medium changed from crude oil to water. In 1993, the company tested a clamped-on ultrasonic flow meter on a crude oil line against a positive displacement (PD) meter. The average error, from measurement of multiple batches of transfer, was found to be 3%. With a meter factor, which is from using a pipe prover, applied to the ultrasonic flow meter, the average error was 0.3%. In Europe, multi-path transit-time ultrasonic flow measurement is now available for certain custody transfer applications.

Doppler Effect Ultrasonic Flow Meters Principle of Operation. A “Doppler effect” ultrasonic flow meter uses one or two transducers to send ultrasonic waves, typically at about 500 kHz frequency, at an angle through the pipe wall (Figure 500-45) into the fluid (usually liquid). Part of the energy is reflected by particles or bubbles in the fluid and is returned through the

July 1999

500-66

Chevron Corporation



Fig. 500-45 Doppler Effect Ultrasonic Flow Meter

pipe wall to a receiving transducer. The receiving transducer may or may not be the same one transmitting the wave. Because the reflectors are travelling at the fluid velocity, the frequency of the reflected wave is shifted according to the Doppler principle. The velocity is proportional to the difference between transmitted and received frequency. An insertion ultrasonic flow meter is a special design. Its hardware and installation are similar to those for insertion turbine meters. Given a known pipe diameter, flow rate can be measured. Applications. Doppler effect ultrasonic flow meters are used in aerated fluids or in slurry with a particle concentration of 0.2% (to 50%) by volume and approximately 100-micron-sized particles. The flow meters are used in most petroleum products and chemicals and in brine, liquid sulfur, sulfuric acid, plant effluent, and sewage. Consult the manufacturer to determine the type of flow meter to use. Limitations. Like transit time flow meters, Doppler effect flow meters require a minimum straight pipe (meter run) from valves, tees, elbows, pumps, and compressors to meet performance specifications. Typically, 10 to 20 diameters upstream and 5 diameters downstream are required. In addition, the outside pipe diameter must be at least 1 inch. Doppler effect flow meters are susceptible to the velocity profile of the fluid. Even distribution and uniform size of particles in the fluid helps to improve meter accuracy. The flow meters generally perform better in turbulent flow at higher Reynold’s numbers (i.e., over 100,000). Pipe vibration at no-flow conditions can sometimes cause an upscale flow indication due to particle or bubble motion. Some manufacturers simply turn down the sensitivity of the detection circuitry, while others have proprietary circuitry that ensures a zero indication at no-flow conditions.

Chevron Corporation

500-67

July 1999



As with the transit time and other flow meters, the pipe must be full in order for the Doppler effect flow meter to properly indicate volumetric flow. A Doppler effect flow meter will, however, indicate velocity in a partially full pipe as long as the transducers are mounted below the liquid in the pipe. Further information on open channel or partially full pipe flow measurement is provided in Section 531. The acoustic waves from the transducer of a Doppler effect flow meter do not penetrate air. This type of meter, therefore, cannot be used with porous pipe materials, such as concrete, concrete-line, cast iron, or many types of fiberglass. Generally speaking, ultrasonic flow meters with “clamp-on” transducers are less accurate than “wetted” flow meters because the errors introduced by wave reflection in pipe walls due to wave path-length change with temperature (pipe and fluid), acoustic short circuits, and limitations due to critical angle. Insertion ultrasonic flow meters to date have limited application. One application is for pacing an automatic line sampler in crude oil custody transfers. The user should consult Company specialists and the manufacturer of the automatic line sampler selected. Performance Characteristics. Typically, accuracy is ±1% to 3% of rate over a 10:1 flow range for liquid flow meters in turbulent flow. Repeatability is typically better than ±0.5% of rate, depending upon velocity range and manufacturer. Transducer Mounting. Usually, the manufacturer’s recommended mounting method should be followed. Basically, the clamp-on style is designed for convenient mounting to the user’s flow tube in the field, for temporary installation, or as a portable device. The wetted flow meter is designed to be a permanent part of a fully assembled flow tube and is usually prefabricated by the flow meter manufacturer. Wetted transducers can also be applied to pipe walls made of sonically opaque materials such as concrete, which will not accommodate clamp-on transducers. The transducers should be mounted “sideways” (i.e. not top and bottom in liquid service where debris may accumulate on the bottom of the pipe. The transducers should also be located where no (or a minimum amount of) air or vapor is entrained in the liquid.

526 Variable Area Meters (Rotameters) Operating Principle A rotameter’s operation is based on the variable area principle, where the flow raises a float in a tapered tube and consequently increases the area for passage of the flow. The greater the flow, the higher the float is raised. The height of the float is directly proportional to the flow rate. With liquids, the float is raised by a combination of buoyancy and velocity head. With gases, the buoyancy is negligible and the float responds to the velocity head alone. The float moves up and down in the tube in proportion to fluid flow rate. Its movement is also proportional to the annular area between the float and the tube

July 1999

500-68

Chevron Corporation



wall. The float reaches a stable position in the tube when the upward force exerted by the flowing fluid equals the downward gravitational force exerted by the weight of the float. A change in flow rate upsets this force balance. The float then moves up or down, changing the annular area, until it again reaches a position where the forces are in equilibrium.

Types of Rotameters The two general types of rotameters are as follows. Glass Tube Rotameters - The basic rotameter is the glass tube indicating type. The tube is made of precision-machined borosilicate glass. The float is also very accurately machined, usually from metal or plastic. Stainless steel is often used to provide corrosion resistance. Safety-shielded glass tube rotameters are in general use throughout industry for measuring both liquids and gases. Figure 500-46 shows typical glass tube rotameters. Fig. 500-46 Typical Glass-Tube Rotameters (Courtesy of Brooks Instrument Division, Emerson Electric)

Metal Tube Rotameters (Armored Meters) - For higher pressures and temperatures beyond the practical range of glass tubes, or for meeting electrical area classification requirements, metal tube rotameters are used. They are usually manufactured of stainless steel, and stainless steel floats are often standard. A version of the armored rotameter is the straight-through flow type, which is most often used for dirty or viscous liquids. Figure 500-47 shows typical metal tube rotameters.

Performance Characteristics The accuracy of a rotameter is ±2% to ±10% of full scale over a 10:1 flow range. Repeatability is typically ±1% to ±2% of indication.

Chevron Corporation

500-69

July 1999



Fig. 500-47 Typical Metal Tube Rotameters (Courtesy of Brooks Instrument Division, Emerson Electric)

Applications and Limitations Applications. Rotameters have been used in liquid and gas service in various process plants. They are used as bypass meters, or indicators in analyzer sample systems, or as purge flow indicators. They can be made with integrated transmitters, alarm switches, and controllers, and coupled with magnetic indicators. Limitations. All rotameters should be installed vertically. Glass tube rotameters often have been used in areas where explosionproofing is not required. They should not be used in liquids which attack the glass metering tube. For example, a glass tube rotameter should not be used in boiling water with high pH, which softens the glass; in a wet steam, which also softens the glass; in caustic soda, which dissolves glass; or in hydrofluoric acid, which etches glass. Rotameters have a so-called “viscosity immunity ceiling.” Below this ceiling, the meter is not influenced by the viscosity of the liquid. When the viscosity immunity ceiling is exceeded, the rotameter calibration will be influenced by the viscosity. Glass tube rotameters have temperature and pressure limits. The safe working pressures of borosilicate glass tubes recommended by ISA (ISA RP16.1,2,3) are listed in Figure 500-48. Up to the maximum temperatures borosilicate glass tubes are resistant to thermal shock, but not to hydraulic shock. The maximum working pressure ratings are for non-shock (no water hammer) conditions.

Specifying and Sizing Consult an applications engineer or a knowledgeable sales representative to determine the required size and features of the rotameter. To size a rotameter use the

July 1999

500-70

Chevron Corporation



calculation methods given in Figure 500-49 (5 sheets) as a sizing guide. The guide was prepared by a domestic rotameter supplier. A chart showing the limit of viscosity immunity for rotameters is included. Fig. 500-48 Typical Safe Working Pressures of Borosilicate Glass Tubes Nominal Tube Inlet Bore (Inches)

Max. Working Press. (psig) Up to 200°F

Press. Reduction Above 200°F (psig/°F)

Maximum Temperature °F

1/16— 1/8

550

0.75

400

1/4

450

0.75

400

3/8

350

0.75

400

1/2

300

0.75

400

3/4

240

0.60

400

1

200

0.45

400

1½

130

0.33

400

2

100

0.25

400

3

70

0.15

300

4

50

0.10

300

A copy of the ISA Rotameter Specification Form S20.22 with explanatory instructions is included in Volume 2 of this manual. Because design and rated capacity vary from one manufacturer to the other, consult the vendor for the proper size.

Industry Standards ISA has published the following four Recommended Practices (RPs) on rotameters. •

ISA RP16.1,2,3, Terminology, Dimensions and Safety Practices for Indicating Variable Area Meters (Rotameters, Glass Tube, Metal Tube, Extension Type Glass Tube)

•

ISA RP16.4, Nomenclature and Terminology for Extension Type Variable Area Meters (Rotameters)

•

ISA RP16.5, Installation, Operation, Maintenance Instructions for Glass Tube Variable Area Meters (Rotameters)

•

ISA RP16.6, Methods and Equipment for Calibration of Variable Area Meters (Rotameters)

These RPs should be used to supplement the information provided in this section.

Chevron Corporation

500-71

July 1999



Fig. 500-49 Rotameter Sizing Guide (Courtesy of Brooks Instrument Division, Emerson Electric) (1 of 5) Sizing Rotameters Rotameters are sized by converting the flow rate of the fluid to be metered to an equivalent flow rate of water or air. The manufacturers’ capacity tables give meter capacities in GPM water and SCFM air. The conversion may be computed mathematically with sizing equations, or a table showing sizing factor versus specific gravity may be used. When a rotameter is shifted to a different service, the capacity change or applicable correction factor may also be computed using the conversion equations. Sizing factors are given below, and complete sizing equations follow, based on Brooks Instrument. Sizing Equations for Rotameters with Stainless Steel Floats 1. Liquid Flow To obtain equivalent flow rate in GPM water, multiply process flow rate in FPM by the sizing factor corresponding to Specific Gravity of the process fluid. (GPM water equivalent = GPM process flow rate × sizing factor) Sp. Gr.

Sizing Factor

Sp. Gr.

Sizing Factor

0.60

0.75

1.00

1.00

0.65

0.79

1.05

1.03

0.70

0.82

1.10

1.06

0.75

0.85

1.15

1.08

0.80

0.88

1.20

1.11

0.85

0.91

1.25

1.14

0.90

0.94

1.30

1.17

0.95

0.97

1.35

1.19

The flow rate of the “V” float rotameter is independent of the viscosity of the liquid, provided the viscosity does not exceed the value shown (viscosity ceiling). Glass Tube

Metal Tube

Tube Size

½"

¾"

1"

1½"

2"

½"

1"

1½"

2"

3"

Brooks Size

8

9

10

12

13

8

10

12

13

15

Viscosity Limit, 5 centistokes

10

20

30

45

5

15

25

60

125

2. Gas Flow To obtain equivalent flow rate in SCFM air at STP, correct the process SCFM for operating temperature and pressure as shown. SCFM air equivalent = SCFM process flow rate ×

SG To 14.7 ------- × -------- × --------1.0 530 Po (Eq. 500-19)

where: SG = is specific gravity at STP (14.7 psia and 70°F).

July 1999

500-72

Chevron Corporation



Fig. 500-49 Rotameter Sizing Guide (Courtesy of Brooks Instrument Division, Emerson Electric) (2 of 5) To = Temperature at operating conditions, in °R (°F + 460) Po = Pressure at operating conditions, in psia (psig + 14.7) Viscosity Considerations When Sizing Rotameters The sharp edge metering float gives a range of viscosity immunity to the rotameter. The degree of this immunity is dependent on meter size and float weight and therefore the meter capacity. When sizing meters, compare the viscosity of the fluid being metered to the so-called “viscosity immunity ceiling” of the meter selected to be sure the viscosity of the fluid does not exceed that ceiling. When the fluid viscosity is less than the ceiling, the meter will not be influenced by the viscosity, even though viscosity might vary within the immunity range or be different from the value actually specified. In this case the capacity data from manufacturers’ bulletins may be used directly. (The table under “Sizing Equations” above gives the rotameter viscosity limits.) Where the fluid viscosity is greater than the “viscosity immunity ceiling,” the rotameter calibration will be influenced by the viscosity. In this case, each meter must be individually calibrated to determine the precise calibration at the particular fluid viscosity. It is then important that the meter be used only at the viscosity for which it has been calibrated. The guide below, based on the water equivalent flow rate, shows the “viscosity immunity ceiling” or viscosity value above which the rotameter is subject to viscosity effect. Regardless of the style of meter, this curve gives the viscosity with respect to flow rate, below which the meter may be purchased without an individual viscosity calibration.

Chevron Corporation

500-73

July 1999



Fig. 500-49 Rotameter Sizing Guide (Courtesy of Brooks Instrument Division, Emerson Electric) (3 of 5) Complete Sizing Equations for Rotameters 1. Liquids GPM water equivalent = GPM metered liq. flow ×

( 7.04 ) ( Sp. gr. liq. ) -----------------------------------------------------------( Sp. gr. float - Sp. gr. liq. ) (Eq. 500-20)

Sp. gr. liquid = Specific gravity (water equals 1.0) of metered liquid at operating conditions Sp. gr. float = Specific gravity (water equals 1.0) of rotameter float This equation converts the metered flow to an equivalent flow of water through a rotameter with 316 stainless steel float, and making allowance for the specific gravity of the metered liquid at operating conditions and the specific gravity of the float to be used. Brooks capacity tables give the following water flows for rotameters with 316 stainless steel floats. Specific gravities of float materials: Aluminum 2.80 Hastelloy B 9.24 Durimet

8.02 Hastelloy C 8.94

Monel 8.84

316 Stn. Stl. 8.04

Teflon

Nickel 8.91

Tantalum

Titanium 4.50

16.60

2.20

2. Gases SCFM air equivalent = SCFM metered gas flow ( SG ) ( To ) ( 14.7 ) ( 8.04 ) × --------------------------------------------------------------------( 1.0 ) ( 530 ) ( Po ) ( Sp. gr. float ) (Eq. 500-21) SG = Specific gravity (air equals 1.0) of metered gas (at STP) To = Temperature at operating conditions, °R (°F + 460) Po = Pressure at operating conditions, psia (psig + 14.7) This equation converts the metered flow to an equivalent flow of air in SCFM at 70°F and 14.7 psia making allowance for the operating temperature and pressure. Brooks capacity tables are in scfm at standard temperature and pressure of 14.7 psia and 70°F. For flow in gravimetric units, the basic sizing factor portions of the equations remain unchanged. The flow portion is modified to confirm to gravimetric flow units. lbs. per min. liq. GPM water equivalent = -----------------------------------------( 8.33 ) ( sp. gr. liq. ) ×

( 7.04 ) ( Sp. gr. liq. ) -----------------------------------------------------------( Sp. gr. float - Sp. gr. liq. ) (Eq. 500-22)

July 1999

500-74

Chevron Corporation



Fig. 500-49 Rotameter Sizing Guide (Courtesy of Brooks Instrument Division, Emerson Electric) (4 of 5) ( lbs. per min. gas (13.34 ) SCFM air equivalent = ---------------------------------------------------------SG ( SG ) ( To ) ( 14.7 ) ( 8.04 ) × --------------------------------------------------------------------( 1.0 ) ( 530 ) ( Po ) ( Sp. gr. float ) (Eq. 500-23) 3. Steam Steam rotameters are sized by a conversion equation showing specific volume, available directly from the steam tables. SCFM air equivalent = (lbs. per min.) (3.65) ×

Specific Volume (Eq. 500-24)

This equation may also be used for other gas or vapor where the specific volume (or its reciprocal, density in pounds per cubic foot at operating conditions) is known. Sizing Equations for Ball Float Rotameters The capacity tables for the low flow, 150 mm, indicating rotameters using spherical floats, give rotameter capacities for the various individual float materials. Because these meters are sensitive to viscosity, for both liquid and gas service sizing should be left to the manufacturer. Changing Rotameter Metering Conditions, Liquid Flow (Does not apply to small Ball Float Rotameters) The change in rotameter capacity as a result of a change in fluid being metered (a change in fluid specific gravity) may be calculated from the equation below. The equation assumes there is no significant change in viscosity. Q 2 = Q1 ×

( Sp. gr. float – Sp. gr. liq. 2 ) ( Sp. gr. liq. 1 ) -------------------------------------------------------------------------------------------------( Sp. gr. float – Sp. gr. liq. 1 ) ( Sp. gr. liq. 2 ) (Eq. 500-25)

where: Q1 = GPM flow rate of rotameter as originally calibrated Q2 = GPM flow rate for new fluid conditions Sp. gr. float = specific gravity (water = 1.0) of float Sp. gr. liquid 1 = specific gravity (water = 1.0) of liquid for which the meter was originally calibrated Sp. gr. liquid 2 =

specific gravity of liquid for new conditions

Note that the above equation is for volumetric flow rate or GPM. The equation below applies for gravimetric units.

W

( Sp. gr. float – Sp. gr. liq. 2 ) ( Sp. gr. liq. 2 )

2

= W 1 × -------------------------------------------------------------------------------------------------( Sp. gr. float – Sp. gr. liq. 1 ) ( Sp. gr. liq. 1 ) (Eq. 500-26)

Chevron Corporation

500-75

July 1999



Fig. 500-49 Rotameter Sizing Guide (Courtesy of Brooks Instrument Division, Emerson Electric) (5 of 5) Changing Rotameter Metering Conditions, Gas Flow (Does not apply to small Ball Float Rotameters) The correction factor for a gas rotameter used at a temperature, pressure, and specific gravity (Condition 2) other than the originally specified temperature, pressure, and specific gravity (Condition 1), may be determined mathematically as shown below. This correction applies for scales etched in standard volume units, such as scfh or scfm.

Correction Factor =

SG1 T1 P2 -----------------------SG2 T2 P1 (Eq. 500-27)

where: T =

absolute temperature °R (°F + 460)

P = psia (psig + 14.7) SG = specific gravity (air equals 1.0)

July 1999

500-76

Chevron Corporation



527 Vortex Shedding Flow Meters and Swirl Flow Meters Vortex Shedding Flow Meters Vortex shedding flow meters were first introduced in the late 1960s. A schematic of a vortex shedding meter is shown in Figure 500-50. Basically, a vortex shedding flow meter is a short section of pipe flanged at both ends for direct insertion in a pipe line. Some forms are not flanged but rather consist simply of a wide ring mounted between standard pipe flanges of various pressure ratings. Fig. 500-50 Schematic Vortex Shedding Meter

Principle of Operation. Figure 500-51 shows that when a fluid stream flows around a bluff body (vortex shedder), viscosity-related effects produce vortices downstream. The vortices are formed against the bluff body and are shed off its downstream faces in a regularly oscillating pattern. The fluid velocity is directly proportional to the frequency of oscillation. Fig. 500-51 Vortex Shedder and Vortices

Chevron Corporation

500-77

July 1999



Mathematically, the Von Karman vortex shedding frequency (f), flow velocity (v), and shedder width (d) relate as follows: f = (St)(v)/(d) (Eq. 500-28)

Where the dimensionless constant St is called the “Strouhal number” and is a significant parameter in vortex flow measurement. Figure 500-52 shows a typical graph of Strouhal numbers vs. Reynold’s numbers for a cylindrical vortex shedder. Within a wide range of Reynold’s numbers, vortex shedding frequency is directly proportional to fluid velocity and is unaffected by changes in density and viscosity. If the Strouhal number for a given vortex shedder is known, flow rate can be measured by means of the vortex shedding frequency. Fig. 500-52 Strouhal Number vs. Reynold’s Number

Vortex shedding frequency of oscillation can be sensed by many different sensors, including: • • • • • •

Thermistor Piezoelectric Ultrasonic High frequency pressure transducer Magnetic pick-off Differential switched capacitor

Applications. Vortex shedding flow meters are used in many liquid, gas, and steam services to replace orifice meters for better flow range (typically 10:1 as compared to 3:1 for orifice meters). The head loss to a typical vortex flow meter is usually smaller than that to an orifice plate. Limitations. Viscous fluids can not normally be metered by vortex shedding flow meters because of the minimum Reynold’s number constraint. The minimum Reynold’s number must be maintained, and it varies with the shape of the bluff body, typically ranging from 4,000 to 40,000 for common shapes (e.g., triangular,

July 1999

500-78

Chevron Corporation



rectangular, square, trapezoidal, and tee). Figure 500-53 shows a graph of typical minimum flow velocity vs. viscosity (in centistokes) for 1- to 4-inch meter sizes. Fig. 500-53 Minimum Flow Velocity vs. Viscosity

Installation. Upstream and downstream straight pipes are required. The length of the straight pipe (i.e., “meter run”) depends mainly on the piping elements upstream of the meter. Typical requirements are shown in the table below. Upstream Piping Elements

Length of Upstream Meter Run (as x Pipe Diameters)

Length of Downstream Meter Run (as x Pipe Diameters)

Reducer

15

5

Expander

35

5

Single 90° elbow

40

5

Two or more 90° elbows in the same or different plates

40

5

Globe/gate valve - full open

35

5

Control Valve

50

5

8 after the straightener

5

Flow straightener

Most vortex shedding flow meters have maximum flow velocities. Typically they are: 250 feet per second (fps) for gases and steam and 25-30 fps for liquids. These maximum flows sometimes limit the rangeability of vortex shedding flow meters to less than 10:1. It is important to select the proper type of sensor for the process. Sensors with ports can clog when used with dirty fluids. Sensors with moving parts can have maintenance problems if operated at high frequencies. Wetted sensors must be able to

Chevron Corporation

500-79

July 1999



withstand the extremes in recalibration of electronics. Sensors which require a process shutdown and draining of the line for maintenance may not be practical. Performance Characteristics. Above their minimum flow rates, most vortex shedding flow meters can achieve an accuracy of ±1% or better over a 10:1 flow range. Actual rangeability may be reduced (e.g., to 8:1) because of the minimum Reynold’s number constraint. During the period of 1985 to 1988, WIB-SIRED tested five vortex meters: Foxboro, Yokogawa (Japan), Neptune, Kent (U.K.) and Bopp and Reuther (West Germany). The test results indicate the following. 1.

The accuracies based on the meter factors provided by vendors range from ±1 to ±2% of rate. The repeatabilities of most of the meters were within ±0.2%. Only some of the meters tested met manufacturers’ accuracy specification.

2.

Meter factors provided by vendors may not always be accurate. The calibration curve of vortex flow meters can be determined most accurately by performing a wet calibration with the medium to be used under operating conditions. A different fluid can be used to accurately calibrate the meter, provided the flow rate of the calibration fluid covers the same Reynold’s number range as the actual operating fluid under operating conditions.

3.

Low viscosity fluids, like propane, yield rather high Reynold’s numbers at which the vortex flow meters perform well.

4.

Installed accuracy can be adversely affected because of misalignment of the meter, upstream valves, and piping elements. Use of Schedule 80 pipe may also degrade accuracy.

Sizing. Following is a size selection guide prepared by a vortex shedding manufacturer (Figure 500-54). The minimum and maximum flow rates may vary by make and model, but the information in this guide gives a good example of how to size a vortex shedding flow meter.

Swirl Flow Meter A “swirl flow meter” or “Swirlmeter” uses its stationary swirl blades to force the axial flow of fluid into a rotation (Figure 500-55). A vortex generated at the center of the rotation is forced by a backflow into a secondary rotation that is proportional to the flow rate and linear within a broad measuring range. This frequency is measured by a piezosensor. The frequency signal is then converted in the converter to an electrical or digital signal output. Performance. The accuracy of a Swirlmeter is similar to that for a vortex shedding flow meter - ±0.5 to ±1% of the rate. Applications. Swirlmeters have been in use in gas, oil and chemical (liquid) applications. Company experience is limited.

July 1999

500-80

Chevron Corporation



Fig. 500-54 Vortex Shedding Flow Meter Sizing Guide (1 of 4) (Courtesy of Endress + Hauser) The operating range of a vortex shedding flow meter depends on the characteristics of the fluid being monitored. The following size selection guide is designed as a general guide for properly sizing the flow meter. 1. General Criteria Maximum Flow velocity: Steam and gas, 246 fps (75 mps); liquids, 30 fps (9 mps). Minimum flow velocity is dependent upon three criteria: a. Fluid density (influences sensor sensitivity). b. Minimum Reynold’s Number of 3800. c. Minimum vortex frequency of 1 Hz, for large sizes (8-12 inches). 2. Saturated Steam Flow Minimum and maximum mass flow rates by meter size are shown in Figure 500-54A below. Example: What is the measuring range for 60 psig saturated steam in a 4-inch line? a. From Figure 500-54A, scan across the 4 inch row and read the measuring range under the 150 psig column (1100 to 28,580 lb/hr).

Figure 500-54A b. The saturation temperature is 366°F and the density is 0.363 lb/ft3 (last two lines of Figure 500-54A). 3. Superheated Steam Flow a. Read the density from Figure 500-54B based on operating temperature and pressure (interpolation may be required).

Chevron Corporation

500-81

July 1999



Fig. 500-54 Vortex Shedding Flow Meter Sizing Guide (2 of 4) (Courtesy of Endress + Hauser)

Figure 500-54B b. If flow rate is known only in lb/hr, convert to volumetric flow rate using the formula below. m q = -----------------( 60 ) ( ρ ) (Eq. 500-29) where: q = volumetric flow rate in ACFM m = mass flow rate in lb/hr ρ = density in lb/ft3 c. Refer to Figure 500-54C for minimum/maximum flow rates by meter size. Example: Determine the proper line size and measuring range for superheated steam at 500°F and 200 psig, flowing at 20,000 lb/hr. a. From Figure 500-54B, the steam density is 0.396 lb/ft3 20 ,000 b. q = --------------------------- = 842 ACFM ( 60 ) ( 0.396 )

(Eq. 500-30)

c. From Figure 500-54C, a 4-inch meter would be a good choice, offering a 17.5:1 turndown. The measuring range for a steam density of 0.396 lb/ft3 is 48-1300 ACFM, or 1150-31,000 lb/hr.

July 1999

500-82

Chevron Corporation




Figure 500-54C 1. Gas Flow a. If the flow rate is only known in SCFM, convert to ACFM:

q

( qs )( T ) = ---------------------A ( 35.4 ) ( p ) (Eq. 500-31)

where: qA = Actual volumetric flow in ACFM qs = Volumetric flow rate at standard conditions (60°F, 14.7 psia) T = Actual process temperature in °R (°F + 460) p = Actual process pressure in psia b. Determine the actual gas density. Standard densities of some common gases are shown in Figure 500-54D. Calculate actual density as:

ρ

( ρs ) ( p ) ( 35.4 ) = --------------------------------A (T ) (Eq. 500-32)

where: ρA = Actual density in lb/ft3 ρs = Density at standard conditions (60°F, 14.7 psia) in lb/ft3 p = Actual process pressure in psia T = Actual process temperature in °R (°F + 460)

Chevron Corporation

500-83

July 1999




Figure 500-54D c. If the density is unknown, it can be determined from the molecular weight or specific gravity of the gas as: M G ρ = -------- = --------381 13.1 (Eq. 500-33) where: ρ = Density in lb/ft3 M = Molecular weight of the gas G = Specific gravity of the gas d. Refer to Figure 500-54C for minimum/maximum flow rates by meter size. Example: Determine the proper line size and measuring range for air at 70°F and 100 psig, flowing at 5000 SCFM. a. Converting from SCFM to ACFM: ( 5000 ) ( 70 + 460 ) q A = ------------------------------------------ = 653 ACFM ( 35.4 ) ( 100 + 14.7 ) (Eq. 500-34) b. Actual gas density: ( 0.0761 ) ( 100 + 14.7 ) ( 35.4 ) ρ A = -------------------------------------------------------------- = 0.583lb/ft 3 ( 70 + 460 ) (Eq. 500-35) c. From Figure 500-54C, a 4-inch meter would be a good choice, offering a 16:1 turndown. The measuring range for a gas density of 0.583 lb/ft3 is 40-1300 ACFM.

July 1999

500-84

Chevron Corporation



Fig. 500-55 “Swirlmeter Operating Principle” (Courtesy of Bailey, Fischer & Porter Co.)

Installations. The upstream and downstream straight run requirements are much less stringent than vortex meters. A manufacturer recommended the following: Length of Upstream Meter Run (as x Pipe Diameters)

Lengths of Downstream Meter Run (as x Pipe Diameters)

Reducer

3

1

Expander

3

1

Single 90° elbow

3

1

Globe/gate valve - full open

5

1

Upstream Piping Elements

The thermowell, if needed, should be installed at least 2D downstream of the process connection of the meter.

528 Mass Flow Meters A mass flow meter is one that measures mass directly. Mass flow meters can be classified by operating principle as the “Coriolis effect” and “Thermal” types. Coriolis effect mass flow meters are often used in liquid and slurries. Some can be used in compressed gases. Thermal mass flow meters are used in gas and liquid. This section discusses Coriolis effect mass flow meters. Thermal mass flow meters are discussed in Section 532.

Principle of Operation Typically, a Coriolis effect mass flow meter consists of a flow sensor unit and an electronic transmitter unit. The flow sensor unit is made of one or more flow tubes. Figure 500-56 is one of many flow tube assembly designs. Each sensor utilizes a magnetic coil mechanism to vibrate the flow tubes. The tubes vibrate at their natural or harmonic frequency like a tuning fork. The peak amplitude is typically less than one-tenth of an inch.

Chevron Corporation

500-85

July 1999



Fig. 500-56 Typical Flow Tube Assembly (Courtesy Micro Motion, Inc.)

As fluid flows through the tubes, it generates a so-called “Coriolis force.” This Coriolis force always opposes entering fluid and aids departing fluid. As the fluid accelerates on the inlet side and decelerates on the outlet side, it causes the tubes to twist. The amount of twist is directly proportional to the mass flow rate of the fluid flowing through the tubes. Two position detectors, located on each side of the flow tubes, send this information (as a difference of phase shift) to the electronic transmitter unit, where it is processed and displayed. A Coriolis force mass flow meter may also be configured to indicate volumetric flow rate. In this case, the frequency of the vibrating tube or tubes is measured and used to determine the density of the fluid. The density is determined in the same manner as other types of vibrating tube densitometers, and is independent of the mass flow rate determination. By dividing the mass flow rate by the measured density, the actual volumetric flow rate is determined.

Applications and Limitations Applications. Coriolis effect mass flow meters are relatively new and are still being developed. They are used for chemicals and plastics, slurries, produced oil, liquefied petroleum gases, and and wherever mass flow measurement (i.e., by weight) is desirable or where volumetric flow measurement is difficult. For example, the Company has been using a number of mass flow meters to measure unstabilized liquid flow of produced oil. Installing a stabilizer at each wellhead, the alternative to using mass flow meters to measure this nearly two-phase flow, would be very expensive. A Coriolis effect mass flow meter can also be used to measure net oil in producing fields. From 1986 to 1987, Chevron Petroleum Technology Company used Micro Motion mass flow meters and an electronic unit to develop a “Net Oil Computer” (NOC) system. The NOC can measure water (in percentage) for production well testing and product allocation. Since then, the Company has licensed several meter manufacturers to use NOC calculations in their products.

July 1999

500-86

Chevron Corporation



Limitations. The cost of a mass flow meter rises rapidly with size; 4- to 6-inch meters are usually considered the large “standard size.” Large meters are very heavy. The flow tube assembly for a typical 3-inch flow meter (0 - 7000 lb/min flow range) weighs over 200 pounds. The weight of a typical 6-inch meter may exceed 1400 pounds. Although there is currently no API standard governing field proving of Coriolis force mass flow meters, API published a technical report (in 1995) outlining procedures for proving this type of flow meter. In general, if the Coriolis flow meter is configured to indicate volume, the same proving equipment and procedures for conventional volumetric flow meters (i.e., PD and turbine meters) can be adapted. If the Coriolis meter is configured to measure mass, then the density of the fluid would need to be measured and incorporated in the proving calculations.

Performance Characteristics Accuracy of ±0.2% to ±0.4% of rate is common for typical Coriolis effect mass flow meters. Rangeability of 20:1 and repeatability of 0.1% are common for most meters. As the process pressure increases, the stiffness of metering tubes also increases, causing a negative measurement error if not compensated. The magnitude of pressure effect ranges typically from nil to - 0.009% per psi, depending on the size, wall thickness and design of the meter. In general, meters with smaller tubes or thicker tube walls have negligible or very low pressure coefficients. Large size meter and/or meters with thin tube walls would have higher pressure coefficients. Operating temperature ranges from -400°F to 400°F for the flow tube assembly. The electronic control unit is typically limited to -40°F to 140°F.

Specifying and Sizing Flow range, fluid type, and operating temperature and pressure must be provided when specifying a mass flow meter. Wetted material ranges from carbon steel to stainless steel to alloys, depending upon its intended application. Consult the manufacturer for help in the proper selection of materials. Figure 500-57 is a specification for a typical Coriolis effect mass flow meter. Sensor sizes shown are for ¼ inch to 6 inch tube size (same size flanges are usually used for end connections). Figure 500-58 shows typical pressure drops versus flow rate. This chart is only valid for process fluids with viscosities near 1 Cp. For specific gravities other than 1, divide by the specific gravity (based on water at 70°F). Consult flow meter manufacturer for higher viscosities and gases.

Installation Manufacturer’s recommendations on installation, specifically mounting and orientation of the meter should be followed. Since vendor instructions may not be

Chevron Corporation

500-87

July 1999



Fig. 500-57 Specifications for a Typical Coriolis Mass Flow Meter (Courtesy of Micro Motion, Inc.)

July 1999

500-88

Chevron Corporation



Fig. 500-58 Pressure Drop vs. Flow Rate for a Typical Mass Flow Meter (Sensor Size (DXX) Explained in Figure 500-57) (Courtesy of Micro Motion, Inc.)

available or sufficient, the following are some general considerations on installation of Coriolis mass flow meters: 1.

Locate the meter flow sensor at a low point of the piping, or “pipe drop” to avoid air or vapor trapped in the meter.

2.

Use an air eliminator or other control devices to liquid-pack the meter.

3.

Avoid installing the meter flow sensor in a vertical line to avoid particle accumulation in the flow tubes when measuring liquid slurries. This orientation also facilitates cleaning, if the process lines are purged with gas or steam.

4.

Locate the flow sensor unit at least two feet from any large transformer or motor to reduce electromagnetic interference from other electrical equipment.

5.

Secure the flow sensor unit and avoid excessive vibration.

6.

Locate the flow meter sensor at a location with minimum pressure fluctuation, if practical. Avoid locations where large pressure changes occur due to control valves, change of pipe size, and piping elements. Avoid locations where a large pressure drop may occur which can cause cavitation.

Figure 500-59 shows the orientation guidelines recommended by Micro Motion for their “Elite” series mass flow meters larger than ¼ inch.

Chevron Corporation

500-89

July 1999



Fig. 500-59 Orientation of Flow Sensors (Courtesy of Micro Motion, Inc.) (1 of 2)

July 1999

500-90

Chevron Corporation



Fig. 500-59 Orientation of Flow Sensors (Courtesy of Micro Motion, Inc.) (2 of 2)

Chevron Corporation

500-91

July 1999



530 Other Flow Measurement Devices 531 Open Channel and Partially Filled Pipe Flow Measurement Fluid flow in an open channel can be measured by the following methods. • •

Measurement of the fluid depth (i.e., “head”) to infer volumetric flow rate Measurement of the fluid depth plus velocity to calculate flow rate

Fluid flow in a partially filled pipe (closed conduit) can also be measured by these methods. A prerequisite for open channel or partially filled closed conduit flow measurement is that a free flow condition must exist at all flow rates. If this condition is met, the Manning/Chezy Equation can be used: Q = 1.49 (R0.167)(S0.5)(A)/(n) (Eq. 500-36)

where: Q = flow rate in appropriate engineering units 1.49 = constant, dimensionless R = the hydraulic radius ( area of stream cross section ) = ----------------------------------------------------------------------( wetted perimeter ) S = slope or gradient of the channel/pipe A = channel/pipe cross sectional area n = roughness factor of the channel or pipe wall; e.g., n is 0.015 for a concrete surface and 0.014 for cast iron pipe in fair condition The equation and the constants are based on experimental data with water in turbulent flow. The flow rate measurement accuracy is typically ±5% to 10% for the first method (level-only) and ±3% to 5% for the second method (velocity plus level), provided other parameters in the equation are comparably accurate. Open channel flow meters based on the measurement of level only include those with flumes, weirs, flow tubes, static pressure sensors, and capacitance and ultrasonic level sensors as the primary element. Flow meters based on the measurement of velocity plus level typically consist of one of the primary elements for fluid depth from those listed above, plus an ultrasonic or a magnetic flow meter as the velocity sensor. This section briefly describes some of the most commonly used open channel flow meters: flumes, ultrasonic and magnetic open channel flow meters, combination of velocity and level measurement flow meters, and weirs.

July 1999

500-92

Chevron Corporation



Flumes The Parshall flume and the Palmer Bowlus Flume are two major types of flumes with a long history of use in waste water flow measurement. Flumes are not often used in Chevron applications. See Weirs and Flumes by P. Ackers for more information.

Ultrasonic Open Channel Flow Meters Figure 500-60 shows some typical open channel applications using ultrasonic flow meters for fluid depth and velocity measurements. Rangeability of this type of flow meter may be as much as 200:1. Fig. 500-60 Various Mounting Methods for Ultrasonic Open Channel Flow Meters (Courtesy of Badger Meters, Inc.)

Magnetic Open Channel Flow Meters A magnetic flow probe can be used for velocity measurement. It creates a magnetic field through which fluid (usually conductive) cuts the field’s lines of force. Thus a voltage is created proportional to velocity, in accordance with Faraday’s Law. As is typical for magnetic flow meters, the measurement is unaffected by bubbles. Like ultrasonic flow meters, magnetic open channel flow meters can detect flow direction and have wide flow range (typically, 100:1).

Chevron Corporation

500-93

July 1999



Combination of Velocity and Level Measurements A number of commercially available open channel flow meters utilize a velocity sensor (e.g., ultrasonic, magnetic) and a level measurement device (e.g., ultrasonic, static pressure sensor, flume, flow tube, or weir) to achieve better accuracy (±3% to 5%) as compared with the level-only open channel flow measurement (±5% to 10%). These flow meters are usually unaffected by solids or air bubbles or by backwater (back flow) effects. They can detect flow direction. The flow meters provide much better accuracy than the level-only type when flow velocity is not constant.

Weirs A weir is a partial obstruction in an open channel over which the fluid accelerates with a free surface. Weir plates are simple head-producing primary devices for open channel flow measurement. They are suitable for clear fluids where hydraulic conditions are stable. Operation of the weir is sensitive to the approach velocity of the liquid, often necessitating a stilling basin or pond upstream of the weir. Such a basin reduces the fluid velocity and provides a place for debris to settle out. Accumulation of foreign material and debris adjacent to the flow meter will affect the operation of the flow meter. Self-cleaning bar screens well upstream of the flow meter may be considered if debris is a persistent problem. To avoid submergence of the weir, the crest must be located higher than the maximum possible downstream elevation of the fluid surface. The head measurement should be made far enough upstream from the weir so that it will not be affected by the downward curve of the fluid surface. Weirs can achieve accuracies of 2% to 5% of rate and turndowns of as high as 25:1. However, the reduced accuracy of the level transmitter may become significant in the lower portion of the flow range. The V-notch weir has a very good turndown and its coefficient does not vary excessively over a wide range of flow. Figure 500-61 illustrates three types of weirs: rectangular, Cipolletti, and triangular or “V-notch.” Weir size may be estimated by using the graphs of the relationship between flow and the liquid head upstream of the flow meter as shown in Figure 500-62. Information on flow measurement for open channels and partially filled pipes can be obtained from the following references and standards. Ackers, P., et al., Weirs and Flumes (Wiley, New York, 1978). ISO, Standard 1438, Liquid Flow Measurement in Open Channels Using Thin Plate Weirs and Venturi Flumes (ISO 1438-1975E, 1975). ISO, Standard 1438/1, Water Flow Measurement in Open Channels Using Weirs and Venturi Flumes, Pt. 1, Thin Weirs (ISO 1438/1-1980E, 1980). National Bureau of Standards, A Guide to Methods and Standards for the Measurement of Water Flow (NBS Publication 421, 1975).

July 1999

500-94

Chevron Corporation



Fig. 500-61 Common Types of Weirs (Courtesy of Bailey, Fischer and Porter, by W. H. Nagel, P. E.)

Chevron Corporation

500-95

July 1999



Fig. 500-62 Relationship of Flow and Liquid Head (Weirs) (Courtesy of Bailey, Fischer and Porter, by W. H. Nagel, P. E.)

July 1999

500-96

Chevron Corporation



532 Flare Flow Meters A flare flow meter is needed for energy conservation and for environmental compliance. Flare flow meters provide the following essential information. •

Energy conservation—By measuring relief rates of individual process plants, operators may be able to identify and improve processing at troubled plants. Flare gas flow measurement may also be required by local regulatory authorities in Europe.

•

Control of Leaks—Some flare flow meters can calculate molecular weight of the gas in the stack. Molecular weight may be used to identify sources of leaks into the flare system. Quick identification of leak sources, such as partially unseated relief valves, often produces significant savings.

•

Environmental compliance—Mass flow rate measured by a flare flow meter may be used to control flare tip steam injection. If the mass flow in the stack is known, the correct amount of steam required at the flare tip may be accurately controlled, reducing steam usage while maintaining compliance with pollution control regulations. With a fast-acting steam control valve, the flare flow meter can be used for automatic flare control to reduce the extent of or to prevent a “smoking stack” during a process surge.

Several types of flare flow meters have been tried in the past decade. At present, the most commonly used flare flow meters are ultrasonic (transit-time) and thermal (thermistor) types.

Ultrasonic Flare Flow Meters A transit-time (“time-of-flight”) ultrasonic flare flow meter uses two ultrasonic transducers in the gas flow. Each transducer is capable of sending and receiving ultrasonic pulses. A pulse travelling in the direction of flow arrives at the opposite transducer in a shorter period than a pulse travelling against the flow. The measured time difference can be used to calculate the velocity of the flow. Ultrasonic flare flow meters operate in both laminar and turbulent flow regions. Their velocity range is typically 0.1 to 50 feet per second, bi-directional. If the pipe diameter is known, the volumetric flow rate can be calculated. With temperature measured by a thermocouple or RTD and with pressure measured by a pressure sensor, the flow meter can use average velocity to calculate molecular weight and then mass flow rate of the gas.

Thermal (Thermistor) Flare Flow Meters The operating principle of a thermal flare flow meter is based on the measurement of heat dissipated by one or more pairs of matched thermistors. A thermistor is a detector that uses a temperature-sensitive resistor as its sensing element. Infrared radiation causes the temperature of the resistor to change. The single-point flow meter uses a pair of thermistors while the multi-point uses multiple pairs. Figure 500-63 shows the mounting for a typical thermistor flare flow meter.

Chevron Corporation

500-97

July 1999



Fig. 500-63 Direct Insertion FM713 MKII Thermistor Flare Flow Meter Installation (Courtesy of Peek Measurement, Sarasota Products)

July 1999

500-98

Chevron Corporation



Single-point mass flow meters are more suitable for smaller relief lines, relief headers with sufficient straight runs, and higher flow velocities. In other words, they are suitable for even flow profiles in which the actual flow can be inferred by the average linear velocity at a point selected locally. In a fully developed flow, this point is at approximately two-tenths of the pipe diameter if measured from the inner wall of the pipe. Figure 500-64 conceptually compares the single-point against the multi-point design by a thermal mass flow meter supplier. Fig. 500-64 Typical Single-point vs. Multi-point Thermal Mass Flow Meter (Courtesy of Fluid Components, Inc.)

Chevron Corporation

500-99

July 1999



Multi-point mass flow meters are better suited for installations where an irregular flow profile is anticipated.

Other Comments In 1987, an ultrasonic and a single-point thermal mass flow meter were tested sideby-side at Chevron USA’s Richmond refinery. Following are some findings of interest. •

The realistic accuracy of the flare flow meters tested was 5% to 15% of rate. This accuracy depended upon piping configuration, range of flow rate, and how well the meters were calibrated.

•

In practice, it is almost impossible to physically calibrate and prove a flare flow meter in the field. If this were possible, the accuracy would improve.

•

The typical response time of both flow meters was 30 to 60 seconds.

•

The molecular weights “measured” by the ultrasonic flow meter (with temperature and pressure inputs) were found to be 11% to 19% lower than those analyzed by a laboratory gas chromatograph with grab samples.

•

Stratification may exist as a result of very low flow and a large relief header. If this is the case, accuracy of the flow meter will be adversely affected.

•

If the response time is acceptable, the mass flow rates measured by these flare flow meters may be used to set steam injection rates for flare control.

•

Thermal (mass flow) instrumentation is usually applied as a switch indicating flow or no-flow. See Section 550.

In 1988, Chevron USA’s Pascagoula refinery replaced the “thermal pile” on a relief line with a thermal mass flow meter and reported satisfactory performance.

Other Flare Flow Meters In addition to the ultrasonic (transit-time) and thermal flow meters, vortex and annubar flow meters have been tried for flare flow measurement. The results have been fair to relatively unsatisfactory because of their low rangeability (3:1 for annubar and 10:1 for vortex) and the minimum Reynold’s number (10,000) required by both flow meters.

533 Two-Phase and Multi-Phase Flow Metering Two-phase Flow Meters Two-phase flows are difficult to measure directly. It is difficult to accurately measure the total flow (i.e., two phases combined), and more difficult to measure each phase accurately. For two-phase flow application which does not require measuring each phase, some conventional flow meters can be used with a reasonable degree of success, for example, a wedge meter in a wet gas stream.

July 1999

500-100

Chevron Corporation



For a two-phase flow application which requires measuring each phase, commercially available flow metering systems include: •

Coriolis mass flow meters which have been successfully used in oil-water flow measurement, as net oil computer (NOC), in various upstream applications.

•

Combination of microwave technology and conventional meters (e.g., turbine meter) as an NOC for two-phase meter for oil/water measurement.

The accuracy of these two-phase flow meters ranges from 2 to 5%.

Multi-phase Flow Meters Direct measurement of multi-phase (i.e., oil, gas and water) production flows offers potentially large savings in facilities and operating costs. In the past several years, various multi-phase metering systems were tested and developed by a number of flow meter system vendors and oil companies. One of them is the Chevron Multiphase Metering Loop (CMML) by Chevron Petroleum Technology Company (CPTC). These multi-phase flow meters usually employ more than one measurement technology, and some designs require the physical separation of gas from liquid in a slip stream. The accuracy of these metering systems ranges from 1% to 10%, and the accuracy varies with each of the three phases. For more information, consult with fluid measurement specialists in CPTC.

534 Other Flow Meters Other flow meters have been used by the industry to overcome specific problems and for special purposes. These include target flow meters and unconventional flow meters. Target Flow Meters. The target flow meters (Figure 500-65) were developed to overcome the problems of dirt buildup in front of an orifice in liquid streams and of liquid buildup in a moist gas stream. These meters also eliminate freezing or plugging of lead lines. The primary element of a target flow meter consists of a sharp leading edge disk (“target”) tied to a supporting rod or bar. The differential pressure produced by the reduced annular area creates a disk drag force. This pressure is transmitted through a bar to a secondary device. Like all other d/p flow devices, the square-rooted output is linearly proportional to the volume flow rate. Target flow meters are well suited for dirty flows and for flows with low-Reynold’s number (RD>100), as well as for clean fluids. Accuracy of uncalibrated target flow meters ranges from ±1% to ±5% of upper range value, depending on line size, beta ratio, and Reynold’s number. Unconventional Flow Meters. In addition to those covered so far, a number of unconventional, “high-tech” flow measurement technologies are available. Some of them are being tested or developed. These technologies include neutron

Chevron Corporation

500-101

July 1999



Fig. 500-65 Target Insertion Flow Meter

bombardment/gamma re-radiation tagging, nuclear, and others. They are not discussed in detail here because they have not yet gained industry acceptance.

540 Meter Provers 541 Proving Liquid Flow Meters Liquid flow meters can be proved by one of the following uni-directional prover devices: a pipe prover, a tank prover, or a master meter. •

A pipe prover is a type of continuous flow volumetric prover comprising a length of pipe from which a known volume is displaced by a displacer to or from a meter being proved at normal operating conditions.

•

A tank prover is an open or closed vessel of known capacity designed for the accurate determination of the volume of liquid delivered into or out of it during a meter proving operation. The prover tank is typically an upright cylindricalshaped tank.

•

A master meter is a meter that has been previously proved by another certified prover. The master meter can then be used either to prove a flow meter or to prove a prover.

The API Manual of Petroleum Measurement Standards (MPMS), Chapter 4, “Proving Systems,” and Chapter 12, “Calculation of Liquid Petroleum Quantities Measured by Turbine or Displacement Meters,” should be read for additional information on meter proving. The Chevron Petroleum Measurement Manual, Part C also provides information of meter provers and meter proving.

July 1999

500-102

Chevron Corporation



A document entitled Guidelines for Meter Prover Design is included as Appendix C in Volume 1 of this manual to provide more details on typical applications of each of these provers. The design guide also discusses typical meter/prover systems for custody transfer, with emphasis on conventional pipe provers. In addition, a model specification, “Stationary Meter Provers—Displacement Type Conventional Pipe Provers” (ICM-MS-2498), is included in the Specification section of this manual.

542 Proving Gas Flow Meters Gas flow meters used for custody transfer can be proved by a master gas turbine meter, a sonic nozzle prover, a bell prover, or a critical-flow orifice prover. These devices are discussed briefly in Section 523. Because proving gas flow meters is a specialized topic, this subject is not covered in depth here. More information on proving gas turbine meters can be found in American Gas Association (AGA) Report No. 7.

543 Proving Mass Flow Meters To date, no industry standard on proving mass flow meters is available. No continuous field proving technique or device has yet been developed. Currently, mass flow meters are calibrated in the shop through the use of weights certified by NIST.

550 Flow Switches Thus far, we have described the most commonly used flow meters. This section briefly describes flow switches, two-phase flow measurement, and other unconventional flow measurement technologies.

Flow Switches Three basic types of flow switches have been available for many years: mechanical, magnetic, and thermistor or thermal.

Chevron Corporation

1.

Mechanical: The flow physically moves a paddle, a float, or a similar device, thereby opening or closing a contact closure in a switch at the preset flow rate. Another widely used flow switch uses a differential pressure transmitter with an orifice.

2.

Magnetic: Two designs are available. The first design uses a magnetic piston that is moved by the flow to activate the switch. The second design operates on Faraday’s Law of Magnetic Induction by measuring electrically conductive fluid that moves through a magnetic field to induce a voltage proportional to flow. This design is virtually a magnetic flow meter with a set-point.

3.

Thermistor or thermal: This switch measures rate of heat dissipation (thermal dispersion) in the surrounding area. The temperature differential between a reference sensor and the measuring sensor is the greatest in a noflow condition and decreases as the flowing fluid passes across the sensing assembly. Flow can be measured by the rate of heat dispersion.

500-103

July 1999



Of these three basic types, the thermal or thermistor flow switches have been found to be accurate, reliable, and easy to install. Flow switches can be used in liquid, gas, and air. The set-point of some flow switches is field-adjustable.

560 Model Specifications, Standard Drawings, and Engineering Forms 561 Model Specifications Specification ICM-MS-2498 Stationary Meter Prover—Displacement Type Conventional Pipe Prover.

562 Standard Drawings The following standard drawings are included Volume 1, Part 2, of this manual.

July 1999

GB-J1177

Differential Pressure Flow Instrument Gas Service—Instrument Above Taps

GB-J1178

Differential Pressure Flow Instrument Liquid Service—Instrument Below Taps

GB-J1179

Differential Pressure Flow Instrument Dry Gas—Instrument Below Taps

GB-J1180

Differential Pressure Flow Instrument Wet Gas—Instrument Below Taps

GB-J1181

Differential Pressure Flow Instrument Steam Service—Instrument Below Taps

GB-J1182

Differential Pressure Flow Transmitter Gas Service—Meter Above Taps

GB-J1183

Differential Pressure Flow Transmitter Liquid Service—Meter Below Taps

GB-J1184

Differential Pressure Flow Transmitter Steam Service—Without Seal Pots

GB-J1185

Differential Pressure Flow Transmitter Gas Service—Meter Below Taps

GB-J1186

Differential Pressure Flow Transmitter Steam Service—With Condensate Pots

GB-J1187

Differential Pressure Flow Throat Tap Connections

GB-J33475

Piping Requirements for Meter Runs

GC-J99504

Standard Orifice Plates

500-104

Chevron Corporation



570 References 571 Included Materials The following appendices are included in Volume 1 of this manual. Appendix A, Orifice Design by Mainframe Computer Appendix B, Hand Calculation Method for Orifice Design Appendix C, Guidelines for Meter Prover Design Appendix D, Flowmeter Selection Charts for Process Plant Meters The following materials are included in Volume 2 of this manual. ISA S20

Specification Forms for Process Measurement and Control Instruments, Primary Elements and Control Valves (with instructions):

S20.20

Orifice Plates and Flanges

S20.21

Differential Pressure Instruments

S20.22

Rotameters

S20.23

Magnetic Flow Meters

S20.24

Turbine Flow Meters

S20.25

Positive Displacement Meters

572 Other References Ackers, P., et al. Weirs and Flumes. New York: Wiley, 1978. AGA, Report No. 7, Measurement of Fuel Gas by Turbine Meters. API RP 500, Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Division 1 and Division 2. API RP 551, Process Measurement Instrumentation API RP 505, Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Zone 0, Zone 1, and Zone 2. API, Manual of Petroleum Measurement Standards. ASME, Standard MFC-3M-1984, Measurement of Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi. ASME Publication, Fluid Meters - Their Theory and Application. Chevron Petroleum Measurement Manual, Part C Chevron Corporation, Natural Gas Measurement Manual, 1988. ISA Standards and Practices for Instrumentation, particularly the following:

Chevron Corporation

500-105

July 1999



ISA RP16.1, 2, 3 Terminology, Dimensions and Safety Practices for Indicating Variable Area Meters (Rotameters, Glass Tube, Metal Tube, Extension Type Glass Tube) ISA RP16.4 Nomenclature and Terminology for Extension Type Variable Area Meters (Rotameters) ISA RP16.5 Installation, Operation, Maintenance Instructions for Glass Tube Variable Area Meters (Rotameters) ISA RP16.6 Methods and Equipment for Calibration of Variable Area Meters (Rotameters) ISO, Standard 1438, Liquid Flow Measurement in Open Channels Using Thin Plate Weirs and Venturi Flumes (ISO 1438-1975E, 1975). ISO, Standard 1438/1, Water Flow Measurement in Open Channels Using Weirs and Venturi Flumes, Pt. 1, Thin Weirs (ISO 1438/1-1980E, 1980). ISO, Standard 5167, Measurement of Fluid Flow by Means of Orifice Plates, Nozzles and Venturi Tubes in Circular Cross-Section Conduits Running Full. Miller, R. W. Flow Measurement Engineering Handbook. New York: McGraw-Hill. National Bureau of Standards, A Guide to Methods and Standards for the Measurement of Water Flow (NBS Publication 421, 1975).

July 1999

500-106

Chevron Corporation

600 Temperature Measurement Abstract This section is a practical guide to the application, selection, specification, and installation of instruments for measuring, indicating, recording, and controlling temperature in refineries, chemical plants, and producing facilities. Section 610 discusses general concepts of temperature measurement, particularly as they bear on selecting a temperature instrument. It also describes and discusses specific devices and provides guidance in their application and specification. Section 630 gives general and specific guidance for the installation of temperature instruments. Section 650 lists reference material for further reading.

Chevron Corporation

Contents

Page

610

Application and Specification of Temperature Instruments

600-3

611

General Information

612

Temperature Primary Elements

613

Bimetallics

614

Filled Thermal Systems

615

Thermocouples

616

Resistance Temperature Devices (RTDs)

617

Thermistors

618

Thermowells

619

Furnace Skin Points

620

Local Temperature Indicators (Dial Thermometers)

621

Remote Temperature Indicators (Thermocouples)

622

Temperature Test Points (Thermowells)

623

Compressor Temperature Alarms and Shutdowns

624

Self-contained Temperature Regulators

625

Temperature Transmitters

626

Field Temperature Recorders

627

Field Pneumatic Temperature Controllers

600-1

600-29

December 1999

600 Temperature Measurement

December 1999


628

Temperature Switches

629

Multipoint Temperature Systems

630

Installation of Temperature Instruments

631

General Requirements—Field Temperature Instruments

632

Specific Requirements—Temperature Instruments

640

Model Specifications, Standard Drawings and Engineering Forms 600-41

641

Model Specifications

642

Standard Drawings

650

References

600-39

600-42

600-2

Chevron Corporation



610 Application and Specification of Temperature Instruments 611 General Information Temperature instruments should be suitable for the process temperature, the process fluid, and the environment where they are installed. Temperature can be measured electrically or mechanically. The following paragraphs define some terms commonly used in temperature measurement.

Units of Measurement Several different units of temperature measurement have been established, but the two most commonly used in oil and gas measurement are Fahrenheit and Centigrade (or Celsius). Fahrenheit arbitrarily assigns the number 32 to the freezing point of water and 212 to the boiling point of water, dividing the interval into 180 equal parts. Centigrade sets the freezing point of water at 0 and the boiling point at 100. The conversion between Fahrenheit and Centigrade degrees is as follows: T°C = 5/9 (T°F − 32) T°F = (9/5 T°C) + 32 where: T°F = Temperature, degrees F T°C = Temperature, degrees C In the United States, temperature instruments are normally ordered to read in degrees Fahrenheit (°F). Canada and most overseas locations use degrees Centigrade (°C).

Ranges The range of a temperature instrument is defined by the minimum and maximum temperature that it can indicate, record, measure, or transmit within a specified accuracy. Primary temperature elements have an element range—the range of the thermocouple, resistance bulb, or filled thermal system. Temperature controllers and recorders also have an element range (inherent range limits of the primary element in the controller or recorder due to its construction) and a calibrated range that is adjusted to match the required measurement. Temperature instruments may be calibrated for full range or for a narrower suppressed range. The calibrated range is usually adjusted by the instrument manufacturer before the instrument is shipped. Field indicators, recorders, and transmitters should have ranges approximately one and one-half times the expected operating temperature. The upper end of the range for temperature controllers should be as close to the maximum operating temperature as practical. A typical choice of elements would be a 0°F to 200°F filled thermal system with a calibrated range of 50°F to 150°F and a scale to match.

Chevron Corporation

600-3

December 1999



Scales The calibrated marking of an indicating, recording, or controlling temperature instrument is called the scale. The scale should match the calibrated range.

Spans The span is the difference between the maximum and the minimum values of the calibrated range of a temperature instrument. The minimum span is the smallest that the manufacturer can accurately calibrate within the instrument’s element range. The span of the element should match that of the process. Narrow span instruments allow more readability and control on critical process control loops. Span selection should be based on operating conditions, but the following usually apply: •

The span should be 100°F if the normal operating temperature is 200°F or less. If a 100°F span is not available, select the narrowest span available.

•

The span should not exceed 50 percent of the operating temperature, if the operating temperature is higher than 200°F.

Accuracy The accuracy of a temperature instrument depends on the selection of the temperature element and the frequency of calibration. For most applications, resistance temperature devices (RTDs) are most accurate. Thermocouple instruments and filled thermal systems are intermediate. Dial thermometers are least accurate.

Stability Temperature instruments that use filled thermal system elements or thermocouples tend to drift and require periodic recalibration. Filled thermal system temperature instruments should include ambient temperature compensation for best accuracy and stability.

612 Temperature Primary Elements Every temperature instrument requires a primary element to convert the temperature of process fluid into a measurable form. Many different primary elements are available but the five most widely used in the petroleum/petrochemical industry are: • • • • •

Bimetallic sensors Filled thermal systems Thermocouples RTDs Thermistors

These are discussed in varying detail in the following subsections.

December 1999

600-4

Chevron Corporation



613 Bimetallics The sensing element consists of a strip of two different metals welded together. The metals have a large difference in thermal expansion coefficients. When the temperature changes, one metal expands more than the other. This causes the strip to deflect and drives a pointer or other mechanism. The bimetallic element is often shaped into a spiral or a helix for compactness. See Figure 600-1. Fig. 600-1

Bimetallic Ambient Air Thermometer

Bimetallic temperature elements are less accurate than most others, however they are very reliable. They are most often used for dial thermometers or temperature switches. They are often used for local temperature indication and control in offshore production.

614 Filled Thermal Systems Filled Thermal Systems vs Electronic Measurement For most applications, electronic temperature measurement elements are more accurate, more stable, and less expensive than filled thermal systems. Field electronic instruments are less practical because of Electrical Area Classification restrictions and because they should be protected from the environment. Electronic temperature instruments also require electric power and do not function during a power outage. The availability of reliable electric power, the location, and the Electrical Area Classification are the key considerations in the decision. Electronic control systems should use electronic temperature measurement. Pneumatic control systems should use filled thermal systems.

Chevron Corporation

600-5

December 1999



Selection of Filled Thermal Systems Filled thermal systems sense temperature by measuring the change in either the volume or the pressure of the fill fluid as the temperature changes. The basic system consists of a temperature sensor, or bulb, connected by a small bore tube, or capillary, to a bourdon, bellows, or diaphragm motion element. The system is filled with a suitable fill fluid. The temperature sensor is inserted into the process (in a thermowell) where the temperature is to be measured. When the fill fluid expands, it deforms the motion element, which is connected to an indicator, a controller, or a pneumatic or electronic transmitter. There are two basic classes of filled thermal systems. Liquid filled systems contain an incompressible fluid whose thermal expansion changes the volume of the motion element. Gas or vapor filled systems contain a gas or a volatile liquid/vapor mixture. Thermal expansion increases the pressure in the motion element, with similar results. Ambient temperature variations affect the accuracy of filled thermal systems. The size of the error depends on the type of system, the temperature range, the length of the capillary tubing, the size of the thermal bulb, the fill fluid and its pressure, and other factors. In liquid and gas-filled systems, both the ambient changes of the motion element and of the capillary can be compensated (full compensation), or just the motion element can be compensated (case compensation). Vapor systems do not need compensation because they are unaffected by ambient temperature changes. The Scientific Apparatus Makers Association (SAMA) divides filled thermal systems into four major classifications, according to the fill fluid, as follows: • • • •

SAMA Class I, Liquid Filled SAMA Class II, Vapor Pressure SAMA Class III, Gas Filled SAMA Class V, Mercury Filled

These four classes of filled thermal system are further subdivided according to their design details and their ambient temperature compensation.

SAMA Class I, Liquid Filled Systems SAMA Class I systems are completely liquid filled. The fill fluid is usually a hydrocarbon such as xylene or ethylbenzene. The fill fluid should remain in the liquid phase over the entire temperature range. The vapor pressure of the fill fluid should be below the maximum pressure of the filled thermal system at maximum temperature to prevent bubbles. The fill fluid should never freeze or solidify because this would change the calibration. See Figure 600-2A. Class I systems are subdivided as follows: SAMA Class IA systems are fully compensated for ambient temperature changes. This is accomplished with a second bourdon tube and capillary. See Figure 600-2B. SAMA Class IB systems are case compensated for ambient temperature changes at the instrument case. This is usually accomplished with a bimetallic mechanism in the case. See Figure 600-2C.

December 1999

600-6

Chevron Corporation


Fig. 600-2

Chevron Corporation


Liquid Filled Thermometers (From Process Instruments and Controls Handbook, D.M. Considine, Editor, 2nd Ed., 1974. Used with permission from McGraw Hill.)

600-7

December 1999



SAMA Class II Vapor Pressure Systems SAMA Class II systems are partially filled with a volatile fluid. The interface between liquid and vapor should occur in the bulb. The vapor pressure increases with temperature and the pressure acts on the pressure element. The fill fluid should have a vapor pressure-temperature relationship that is both large and linear. There should be a measurable vapor pressure at the lowest usable temperature. The useful temperature range of the various fill fluids is limited. SAMA Class II is divided into Class IIA, IIB, IIC, and IID categories, based on the temperature of the bulb relative to the rest of the system. SAMA Class II systems do not require ambient temperature compensation.

SAMA Class III, Gas Filled Systems SAMA Class III systems are entirely filled with gas. The gas pressure varies directly with temperature according to Charles’ Law. The deviation of the gases used from a perfect gas are slight enough to permit acceptable accuracy. Nitrogen is the favorite fill gas because it is inert, inexpensive, and has a linear temperature-pressure relationship for normally encountered temperatures. Helium is sometimes used for temperatures below 400°F or above 800°F. SAMA Class IIIA systems are fully compensated for ambient temperature changes with a second bourdon tube and capillary. SAMA Class IIIB systems are case compensated for ambient temperature changes at the instrument case with a bimetallic mechanism in the case. The additional expense of Class IIIA compensation is rarely justified because the error caused by ambient temperature changes is very small.

SAMA Class V, Mercury Filled Systems SAMA Class V systems are seldom ordered today because of environmental and safety concerns with mercury. The following information is applicable if you have an existing system. SAMA Class V systems are fully filled with liquid mercury. They are similar to Class I systems except for the unique thermal properties of mercury. The coefficient of expansion of mercury is nearly uniform, which allows the use of linear charts and scales. Class V systems respond rapidly and they have enough power to operate control elements. SAMA Class VA systems are fully compensated for ambient temperature changes with an Invar wire inside the capillary and a bimetallic mechanism in the case. SAMA Class VB systems are case compensated for ambient temperature changes at the instrument case with a bimetallic mechanism in the case. All Class V systems require case compensation to neutralize the effect of bourdon element thermal expansion.

December 1999

600-8

Chevron Corporation



Force- vs Motion-Balance Filled thermal system transmitters may use either force-balance or motion-balance mechanisms to convert the element movement into an output signal. Motion-balance transmitters are often indicating. In a motion-balance transmitter, the moving tip of the element is connected to an indicator, or to the flapper of a pneumatic transmitter, or to the current producing section of an electronic transmitter. Force-balance transmitters are always blind. A receiver gage or field ammeter is required if the transmitter output is to be read in the field. In a force-balance transmitter, the force that tends to move the element is opposed by an equal force that is generated by the electronic or pneumatic output signal. Because a force-balance transmitter has no moving parts, it has less hysteresis and is more accurate.

Specification of Filled Thermal Systems Selection of the filled thermal system depends on required range, span, capillary length, accessibility, and space limitations for the placement of the sensor. For routine applications, only SAMA Class IA (liquid filled) or SAMA Class IIIB (gas filled) should be specified. The two systems are compared in Figure 600-3. Selection depends on the ambient temperature, the height of the bulb and the accuracy required. ISA Specification Form S20.11a should be used to order these systems. A copy of the form and instructions for completing it can be found behind Tab DS-4780.

Capillaries The capillaries should be all-welded 304 or 316 stainless steel with flexible corrosion-resistant armor. The maximum capillary length should not exceed 20 feet.

615 Thermocouples If an electrical circuit is made of two wires of different metals, and the junctions where the different wires are joined are at different temperatures, an electrical voltage will be generated. This voltage or electromotive force (EMF) is caused by the Seebeck Effect. Thermocouple instruments measure temperature by measuring the EMF. See Figure 600-4.

Types of Thermocouples A thermocouple consists of two wires of different materials that are joined at the hot junction where the temperature is to be measured. The other end of the wires, the reference junction, is joined at the thermocouple measuring instrument. The instrument measures the temperature of the reference junction and the small voltages generated by the temperature difference between the two junctions. Any two dissimilar metals can be used to make a thermocouple. The Instrument Society of America (ISA) has selected a small number of combinations of precise composition and assigned letters to them. They prepared millivolt output versus temperature tables for these combinations. EMF values are tabulated with the

Chevron Corporation

600-9

December 1999


Fig. 600-3


Application Data for Filled Thermal Systems Type

Gas Pressure

Liquid Expansion

IIIB

IA

−460°F to +1400°F

−300°F to +600°F

200°F(2)

40°F

1000°F

600°F

1400°F

100% of span

100 feet

100 feet

10 × 7/8 in.

6 × 3/8 in.

6 × 5/8 in.

3 × 1/4 in.

2 to 8 seconds

6 seconds

Medium Low

High

Uniform

Uniform

SAMA Class Temperature Limits(1) Minimum Span Maximum Span Limits of Overrange

(3)

Max. Tubing Length(4) Max. Bulb Size Min. Bulb Size(5) 63% Time Constant

(6)

Relative Cost Scale

(1) The maximum bulb temperature is generally limited to 600°F. However when necessary, maximum bulb temperatures up to 1000°F may be specified. (2) The minimum span for a force-balance transmitter is 50°F. (3) The limits of overrange protection are reduced for the narrowest spans. (4) Limit capillary length to 20 feet. Longer lengths may result in larger bulbs, slow response, and poor ambient temperature compensation. (5) Minimum bulb sizes are available only with force-balance transmitters. (6) Time constant is the time for the temperature to reach 63% of a step change. Short capillary lengths and minimum bulb diameters give minimum time constants.

Fig. 600-4

Thermocouple Principle

reference junction at 32°F (0°C), also called ice-point. When both the hot junction and the reference junction are at the same temperature, no EMF is generated. The tables thus show that at 32°F all thermocouples generate zero EMF. Most measuring instruments have cold junction compensation built into their circuitry. Cold junction compensation corrects the thermocouple readout for the temperature of the reference junction at the readout electronics. The ISA Standard has been adopted by the American National Standards Institute as ANSI/MC 96.1-1982.

December 1999

600-10

Chevron Corporation



Thermocouple instruments are calibrated to match the EMF ranges. Most installations in the petroleum/ petrochemical industry use Iron-Constantan, (Type J); Chromel-Alumel, (Type K). Copper-Constantan, (Type T) is used occasionally.

Recommended Temperature Limits for Thermocouples Figure 600-5 is taken from American National Standard ANSI/MC96.1-1982. These upper temperature limits are only for thermocouples that are made up in the field. The intent of this table is to show typical temperatures at which thermocouples fabricated from the wire sizes listed may be expected to provide satisfactory service. Fig. 600-5

Thermocouple Upper Temperature Limits

Thermocouple Type

No. 8 Gage

No. 14 Gage

No. 20 Gage

Type J Iron-Constantan

1400°F (760°C)

1200°F (650°C)

895°F (480°C)

Type K Chromel-Alumel

2300°F (1260°C)

1995°F (1090°C)

1795°F (980°C)

——

700°F (370°C)

500°F (260°C)

Type T Copper-Constantan

Prefabricated, swaged, magnesia-insulated thermocouples should normally be used because they are better protected from oxidation. Type J swaged thermocouples are normally used from -300°F to 1400°F. Type K swaged thermocouples are normally used from 32°F to 2300°F. Type T thermocouples are sometimes used for temperatures below 0°F because they have better corrosion resistance than Type J. Type J thermocouples have traditionally been specified instead of Type K thermocouples at lower temperatures because Type K generates such a low EMF signal. With today’s solid-state readouts, many locations use Type K thermocouples at all temperatures from ambient up to the upper limit of Type K, thus avoiding Type J’s problems with corrosion and oxidation. There are many other types of thermocouple materials available for special services, but Types J, K, and T fill the vast majority of our needs.

Advances in Thermocouple Technology In recent years, thermocouple design has improved: •

Type N Thermocouples In the 1980s, Incotherm Pty. Ltd. of Hereford, England introduced the Type N thermocouple, consisting of a Nicrosil™ positive wire and a Nisil™ negative wire. Type N thermocouples with Inconel-600 sheaths are capable of sustained operating temperatures at 1,850°F. The main objection to using Type N is that, when retrofitted into existing installations designed for Type K thermocouples, Type N requires extensive rewiring between field junction boxes and T/C multiplexers. Most process-plant thermocouple multiplexers are now smart enough to calculate temperatures of all common types of thermocouples, including Type N.

Chevron Corporation

600-11

December 1999



Incotherm also introduced a series of Nicrobel™ metal sheaths, which have extended temperature ratings. These exhibit coefficients of linear expansion virtually identical to those of Type K and Type N thermocouple wires. They claim an upper temperature limit of 2,340°F for the combination of Type N couple and Nicrobell-B metal sheath. Note that Type NN (Nisil) wire melts at 2,280°F. •

High Performance Type K Thermocouples In the early 1990s, both Incotherm and Hoskins Manufacturing Co. of Hamburg, Michigan announced the development of high-performance ISA Type K thermocouple wires, which they claim overcome most of the objections to standard Type K couples. These are called SuperK thermocouples. Calibration stability to temperatures in excess of 2,000°F is achieved by eliminating the trace metals that can migrate between thermocouple wires and the sheath metal, while maintaining the Type K calibration curve. This practice permits reuse of existing Type K thermocouple extension wire and multiplexers on existing installations when retrofitting to the high performance thermocouple material.

Thermocouple Accuracy Thermocouples are relatively inaccurate. Thermocouples suffer from calibration drift and corrosion as they age. The millivolt output signal is quite low and subject to noise pickup. Figure 600-6 is taken from American National Standard ANSI/MC96.1-1982. Fig. 600-6

Initial Calibration Tolerances for Thermocouples

Thermocouple Type

Temperature Range

Standard Tolerance

Special Tolerance

Type J Iron-Constantan

0 to 1400°F 0 to 750°C

±4°F or ±0.75% ±2°C or ±0.75%

±2°F or ±0.4% ±1°C or ±0.4%

Type K Chromel-Alumel

0 to 2300°F 0 to 1250°C

±4°F or ±0.75% ±2°C or ±0.75%

±2°F or ±0.4% ±1°C or ±0.4%

Type T Copper-Constantan

0 to 660°F 0 to 350°C

±2°F or ±0.75% ±1°C or ±0.75%

±1°F or ±0.4% ±0.5°C or ±0.4%

Cryogenic Range

-330 to 0°F 200 to 0°C

±2°F or ±1.5% ±1°C or ±1.5%

The length, composition, and condition of the thermocouple lead wires also affects the accuracy of the measurement. However, the initial material cost and the installed cost of thermocouple measurement are lowest as compared to other systems. This is particularly true when a large number of thermocouples are multiplexed into a single receiver.

December 1999

600-12

Chevron Corporation



Extension Wire Selection The extension or lead-in wiring from the thermocouple in the field to the thermocouple measuring instrument should be of material similar to that of the thermocouple. Otherwise, each field termination would generate a false EMF. The extension wire can be either thermocouple wire, i.e., iron-constantan, or it can be a proprietary alloy extension wire that has the same properties as the thermocouple wire.

Field-made vs Prefabricated Thermocouples Thermocouples can be (and in the past, were) field fabricated by welding the wires together and insulating them with ceramic beads. This is rarely done today. Prefabricated mineral insulated, metal sheathed, swaged, thermocouples should be used.

Duplex Thermocouples Duplex thermocouples consist of two separate thermocouple elements contained in a single mineral insulated, metal sheath. The intent is to have a single thermocouplethermowell assembly provide separate control and monitoring signals, or to have a backup element for a measurement point in the event of failure of the primary element. CRTC discourages the use of duplex thermocouples for the following reason. In order to fit two thermocouple junctions and lead wires in the same interior volume as a simplex thermocouple, smaller diameter wire has to be used. The thinner wires are less robust and more prone to burning out or breaking than the heavier simplex element. It is apparent that any thermal stress which causes the failure of one of the elements would cause the other element to fail as well. Whenever multiple discrete temperature measurements are required at one location (e.g. for measurement validation, or to separate monitoring from control), multiple thermowells with simplex thermocouples located in proximity are recommended. Duplex thermocouples should only be used as a last resort, for example, when it is impossible to install another thermowell; or when it is necessary to validate the measurement of localized hot spots or severely stratified temperature patterns. CRTC instrumentation specialists should be consulted for these applications.

Thermocouple Specification ISA Form S20-12a is used to order thermocouples. A copy of the form and instructions on how to complete it can be found behind the Tab labeled DS-4780. The following information may also be helpful. Prefabricated thermocouples for general services should be a minimum 0.25-inch outside diameter, sheathed with 304 or 316 stainless steel (or better), and mineral insulated. The sheath material should terminate with suitable sealing material to keep the mineral insulation dry. The thermocouple should allow positive grounding at the thermocouple well. See Figure 600-7. Thermocouple properties, color coding, and limits of error should conform to ANSI/MC96.1, Temperature Measurement Thermocouples. The measured resistance between the leads and the sheath should be in accordance with ASTM E236,

Chevron Corporation

600-13

December 1999


Fig. 600-7


Thermocouple Grounding

Specification for Thermocouples, Sheathed, Type K, for Nuclear or Other HighReliability Applications. Thermocouples can be specified as assemblies with the head mounted on the thermowell, or the head can be mounted remotely on conduit. Assemblies are used most often and are described below.

Thermocouple Head Specifications Thermocouple heads should be weatherproof, epoxy-coated, cast aluminum. They should be mounted directly on the thermowell with a union and a spring-loaded thermocouple. The thermocouple head should provide a grounded terminal for connecting the ground wire from the thermocouple and the drain wire from the extension wire shield. Connection to the thermowell should be ½-inch NPT. The head cover should be threaded and gasketed with a stainless steel retaining chain connected to the body.

616 Resistance Temperature Devices (RTDs) As the temperature of a conductor increases, its electrical resistance also increases. Resistance temperature devices (RTDs) are calibrated resistors that are used to measure temperature. RTDs are more accurate than thermocouples and almost all other temperature elements and they maintain their accuracy for long periods. The current flow of an RTD is much higher than that of a thermocouple so they are less subject to noise pickup or errors from lead-in wires. The change in the ice-point resistance (at 0°C) of the RTD should not exceed 0.5°F in the first year of service. RTDs should be used in place of thermocouples: • • • •

December 1999

For highly accurate temperature measurement, such as custody transfer service For narrow span temperature measurement (under 100°F) For temperature difference measurement For control and for other critical applications

600-14

Chevron Corporation



Two-wire, Three-wire and Four-wire RTDs RTDs are available for connection to the readout devices with two, three or four wires. The three-wire system essentially eliminates the error caused by the ambient temperature changes in the resistance of the lead-in wires. Three-wire sensors are preferred because they maintain the high inherent accuracy of the RTD. The lead-in wiring error can be entirely eliminated by using a four-wire system. Four-wire systems are mostly used for extremely accurate laboratory type measurements. Figure 600-8 shows the connections for three- and four-wire RTDs.

Types of Resistance Temperature Sensors RTDs are commonly made of copper, nickel, or platinum. Nickel elements may be used from -40°F to +500°F. Platinum elements, enclosed in 316 stainless steel or Inconel sheaths, may be used from -320°F to +1200°F.

Specification of Resistance Temperature Sensors Although expensive, platinum RTDs are preferred because platinum has a higher maximum temperature limit, a longer service life, and the resistance versus temperature curve is extremely linear. Platinum RTDs should conform to International Standard DIN 43760, calibrated for 100 ohms at 0°C, and with a slope defined by an alpha of 0.00385 ohms/ohms/degree C. Open circuit protection should be provided. Nickel elements are calibrated with a resistance of 120 ohms and cannot be interchanged with platinum elements. Scientific Apparatus Makers’ Association (SAMA) also publishes RTD curves like those published by the International Standard Organization. Most RTD manufacturers, however, reference the International Standard curves because they are more universally accepted.

617 Thermistors Thermistors are very small ceramic resistors with a high temperature coefficient of resistance. Thermistors are usually made of a sintered mixture of metallic oxides. For a given change of temperature, the resistance of a thermistor changes approximately 10 times as much the resistance of a platinum RTD. Thermistors are sensitive to very small changes in temperature. When mounted in small thermowells, they respond very quickly because of their small thermal mass. Thermistors are usually made in the form of a tiny bead, which is encapsulated in glass. Disc thermistors are also available. Thermistors are usually not interchangeable and the temperature instrument should be calibrated to match the specific thermistor. Thermistors are subject to long term drift due to aging. Their accuracy and ambient temperature compensation are usually less than conventional temperature sensors. Thermistors are normally used in lower priced digital temperature indicators or temperature switches.

Chevron Corporation

600-15

December 1999


Fig. 600-8


Resistance Temperature Devices; Diagrams for Three- and Four- Wire Systems

December 1999

600-16

Chevron Corporation



618 Thermowells Few temperature sensors can withstand continuous exposure to process fluids. Additionally, it is desirable to be able to remove the sensor without shutting down the process unit. Therefore, temperature sensors should be installed in thermowells to protect the temperature sensor from direct exposure to the process fluid and to allow removal of the sensor. Thermowells provide mechanical protection from the static and dynamic forces of the process stream, and they provide chemical protection from corrosive process fluids. The same basic types of thermowells are used for thermocouples, RTDs, filled thermal systems, and bimetallic thermometers. Test thermowells are sometimes provided to allow the process to be checked with either a mercury-in-glass test thermometer or a portable electronic temperature indicator. Test thermowells should be provided with a stainless steel plug and chain to keep dirt out of the bore of the thermowell.

Thermowell Length Thermowells and temperature sensors are normally available in lengths ranging from 2½ inches to 12 inches and longer for special orders. The minimum length for use in piping should be 9 inches, which provides about 4 inches of immersion. Refer to Standard Drawings GB-J1196 and GB-J1198. For vessels, the minimum length should be 9 inches and the maximum length should be 12 inches, unless special designs with provision for thermowell support are used. For the maximum thermowell length (U) should be 9 inches. “U” is defined on Standard Drawings GB-J1195 and GB-J1197. These thermowell lengths provide approximately 6 inches of immersion when installed as shown in Standard Drawings GB-J1196 and GB-J1198. For vessels or tanks where the fluid is not flowing or agitated, the thermowell should be long enough to provide at least 9 inches of immersion into the process fluid, measured from the inside face of the vessel or tank wall. Longer thermowells may be required for specific applications, such as FCC or Isocracking reactors, or asphalt storage tanks. They should be specially designed for the application. The selection of a thermowell length should be based on the geometry of the installation and on a calculation of the forces acting on the thermowell under flowing conditions as described below.

Harmonic Vibration Long thermowells that extend into high velocity process fluids are susceptible to vibration induced by the Von Karman vortex frequency. If the thermowell is longer than a certain critical length, it may vibrate at its natural frequency, causing noise and possible failure due to metal fatigue.

Chevron Corporation

600-17

December 1999



The fluid velocity of a gas should be calculated as follows: 0.086Q g T U = -----------------------P ( D )2 (Eq. 600-1)

where: U = Fluid velocity, ft/sec Qg = Gas flow rate, SCFM T = Operating temperature, degrees R °R = t(°F)+459.69 P = Operating pressure, psia D = Pipe inside diameter, inches The fluid velocity of a liquid should be calculated as follows: 0.41Q L U = -----------------D2 (Eq. 600-2)

where: U = Fluid velocity, ft/sec QL = Liquid flow rate, gpm D = Pipe inside diameter, inches The Von Karman wake vortex frequency is a function of fluid velocity, the outside diameter of the thermowell, and the Reynolds Number. Calculate it as follows: A. Calculate the Reynolds Number (N R) for the flowing fluid:

N R = 3163QG b ⁄ ( dνG f )

(Liquids)

where: Q = Flow rate (gpm) Gb = Liquid specific gravity at 60°F d = Thermowell diameter (inches) Largest diameter in flow ν = Kinematic viscosity (centistokes) Gf = Liquid specific gravity at flowing temperature

December 1999

600-18

Chevron Corporation



N R = 0.48297QG/ ( dµZ b )

(Gases)

where: Q = Flow rate (SCFH) G = Gas specific gravity at 60°F and 1 atm. (air = 1.00) d = Thermowell diameter (inches) Largest diameter in flow µ = Absolute viscosity (centipoises) Zb = Compressibility factor

NR = 6.316 W/(dµ)

(Steam)

where: W = Mass flow rate (lbs/hr) d = Thermowell diameter (inches) Largest diameter in flow µ = Absolute viscosity (centipoises) B. Determine the Strouhal Number (NS): For NR < 40,000: use NS = 0.21 For 40,000 ≤ NR ≤ 400,000: use NS = 0.24 (log10 NR) - 0.894 For NR > 400,000:

use NS = 0.45

Note For NR < 40,000, this Harmonic Vibration analysis is nearly identical to the results obtained by the ASME method (reference 3, ASME Performance Test Codes, Supplement on Instruments and Apparatus, Part 3: Temperature Measurement). For NR > 40,000, this Harmonic Vibration analysis is more conservative than the ASME method, i.e., the thermowell must be shorter. This is because a larger NS gives a higher vortex frequency (f v) which means the thermowell length must be shorter in order to meet the 0.8 ratio requirement for fv/fn. Reference 8, Brock, J.E., “Stress Analysis of Thermowells,” is the source for NS when NR > 40,000. C. Calculate the vortex frequency (f v, Hertz) as follows: fv = 12 NS U/dt (Eq. 600-3)

where: U = Fluid velocity (feet/second)

Chevron Corporation

600-19

December 1999



dt = Thermowell diameter at the tip (inches) Calculate the lowest natural elementary frequency (fn) of the thermowell: fn = (Rf)(Ff)(dr/L2)

E ⁄ ( γ + γ′ ) (Eq. 600-4)

where: Rf = Frequency response factor = [1 - 0.4(dr+dt)/L] Ff = Frequency factor = [1.65 + 1.21(dr/dt)(1.0 - 0.094 dr/dt)] dr = Outside root diameter of thermowell (inches) L = Length of thermowell (root to tip), (inches) E = Young’s modulus (28.6 x 106 for stainless steel) γ = Specific weight of well material (lbs/in3) γ´ = Specific weight of fluid (lbs/in3) The vortex frequency (fv) should never exceed 80 percent of the natural frequency (fn) of a thermowell. fv ≤ 0.8 fn (Eq. 600-5)

Thermowell lengths that meet these criteria should not be subject to fatigue failure. Alternatively, Figures 600-9 and 600-10 show the maximum fluid velocity for standard screwed thermowells of various lengths, based on water at 68°F. [Reference 8] Fig. 600-9

Allowable Lengths for 1-Inch NPT Screwed Thermowells

Maximum Fluid Velocity, ft/sec

Thermowell Length, in

193

4

93

6

54

8

40

10

30

12

17

18

10

24

Fig. 600-10 Allowable Lengths For ¾-Inch NPT Screwed Thermowells

December 1999

Maximum Fluid Velocity, ft/sec


115

4

54

6

600-20

Chevron Corporation



Fig. 600-10 Allowable Lengths For ¾-Inch NPT Screwed Thermowells Maximum Fluid Velocity, ft/sec


32

8

24

10

18

12

11

18

8

24

Response Time For rapid response to changes in temperature, the thermowell should locate the active portion of the measuring element in the flow stream. Mechanical strength requires thick, short, thermowells. These factors tend to reduce the accuracy of the temperature measurement and the response time to react to temperature changes.

Thermowell Specifications ISA Form S20.12a is used to order thermowells alone, thermocouples alone, or complete thermocouple/thermowell assemblies. A copy of the form and instructions for completing it are found behind Tab DS-4780. The basic thermowell materials are usually 304 or 316 stainless steel. Other stainless alloys can be used for thermowells in high temperature applications, e.g., type 309 stainless steel for furnace convection section thermowells. Exotic alloy thermowells, e.g., Alloy 20 or Hastelloy, are frequently used for corrosive services. Tantalum sheathing is used in services corrosive even to exotic alloys (spray-on or electro-plated applications of tantalum are usually porous and are not acceptable). Ceramic wells are used where the process is at a low pressure and may be both corrosive and hot, e.g., chemical waste incinerators. Thermowell material should always be selected for nil or near nil corrosion. Flanged thermowells should be used for the following services:

Chevron Corporation

•

Alloy and alloy-clad vessels

•

Coking service lines and vessels, such as crude unit atmospheric and vacuum column bottoms, residuum stripper columns, thermal cracking vessels, FCC fractionators, etc.

•

Columns and vessels where it is specified that all connections should be flanged

•

Any location where severe vibration is expected, such as reciprocating machinery, high pressure drop control valves, etc.

•

All services, except flue gas, if the temperature is above 750°F

•

Columns, vessels and piping in LPG, H2S, Freon, or hydrogen services. Hydrogen service is defined as 50% or more volume percent hydrogen

•

Caustic at temperatures above 140°F

600-21

December 1999


•


Steam at 400 psig or higher

Flanged thermowells should be 1½ or 2 inch. Flanged connections similar to Van Stone type flanges should be used. Refer to Standard Drawing GB-J1197. The internal thread for mounting the head should be ½ inch NPT. Screwed thermowells (not to be seal-welded) should be used for all services not requiring flanged thermowells. Screwed thermowells should be a tapered design machined from 11/16-inch minimum hex 304 or 316 stainless steel bar stock with a ¾-inch NPT process connection. Refer to Standard Drawing GB-J1195. Thermowell bores should be specified as the diameter of the sensing element (Ds) plus one hundredth of an inch (Ds + 0.010 in.). The internal bore diameter for test thermowells should be 3/8 of an inch. Throwaway plastic plugs for internal thread protection should be furnished with thermowells which will contain a thermal element. Test thermowells should be furnished with a steel plug.

619 Furnace Skin Points Summary and Recommendations The key to successful Skin Point Thermocouples (referred to below as SPTCs) is selecting the right materials and installing them properly at the right places in the fired heater. •

Material At the present time, Chevron has not standardized on a particular material for SPTCs. Due to manufacturing inconsistencies with Hoskins-2300™ and resultant problems with installed SPTCs in Chevron facilities, the Company no longer recommends the use of this material. Pending performance testing on other high-temperature alloys, the Company is specifying 310 stainless steel and Inconel 600 as sheath materials for mineral insulated metal sheathed (MIMS) thermocouple wire. The Gay Engineering & Services Co. (Gayesco) Refractopad SPTC design remains the recommended standard for fired heaters at Company facilities. Extreme-service units such as CCR units and ethylene crackers require the use of retractable SPTCs, such as the Gayesco Retractopad. CRTC instrumentation specialists should be consulted for all SPTC applications. Services which form a coke layer on the tube ID may require special consideration. Contact the CRTC furnaces or instrumentation specialist.

•

Installation To achieve the longest possible service life, use the standard drawing, GDJ1201, for planning the installation of SPTCs.

December 1999

600-22

Chevron Corporation



Each installation is unique and requires fine-tuning of the basic design. Fullservice engineered equipment manufacturers such as Gayesco can provide a range of services from consultation to physical installation and testing. Thermal-scanning services can be helpful in identifying the hottest areas of a fired heater and in spotting sites for SPTC installation. They can also be used as a tool to diagnose burner and/or air-register problems.

SPTC Construction The Gayesco Refractopad SPTC used on fired heaters today consists of the following components: •

A low-mass, weld pad, approximately ¾-in. × 1/8-in. square and radiused to the O.D. of the tube. The thermocouple assembly is factory-welded to the pad, and the pad is shop- or field-welded to the tube at the point of temperature measurement.

•

A thermocouple assembly, consisting of measuring junction, connecting thermocouple-grade wires, surrounding magnesia insulation, and protective metallic sheath. The T/C assembly is factory pre-formed to fit the tube being measured and includes sufficient sheathed thermocouple wire to exit the furnace setting at a predetermined point. Where required, a coil is formed in the MIMS cable to allow for thermal expansion of the furnace tube. Alternately, sufficient MIMS cable is installed outside the furnace to account for expansion inside the furnace.

•

A radiation shield and thermal insulation to protect the measuring junction and weld pad from radiant and convective heating at the point of measurement.

•

Mechanical supports and mounting clips to provide mechanical integrity to the assembly when installed in the fired heater.

•

Fittings and transition wire to permit routing the MIMS cable outside the fired heater to the junction box for connection to monitoring instrumentation.

Failure Mechanisms on Conventional SPTCs At the elevated temperatures encountered in refining and chemicals fired heaters, several failure mechanisms combine to shorten the life of SPTCs: •

Internal Differential Expansion Ideally, sheath/thermocouple wire-pair wires have identical coefficients of thermal expansion and can withstand temperature fluctuation. When the coefficients differ, stresses are imposed on the assembly as temperature rises from ambient to operating and encounters the normal cycling from changes in firingrate and flame-patterns. Both 300-series stainless steel and the common Inconel alloys have a higher coefficient of thermal expansion than the Chromel and Alumel pair. The different coefficients cause a strain at the tip, where the thermocouple pair and sheath cap are welded into a single sensing point. With time

Chevron Corporation

600-23

December 1999



and thermal cycling, this repeated strain can cause one of the thermocouple wires to break, thus opening the thermocouple. •

External Differential Expansion SPTC cables are secured to the measured furnace tube (ideally, on the dark side) with a series of metal clips, which hold the sheath snug against the tube. This arrangement provides a heat sink to conduct away any heat absorbed from reflected radiation. Because the common sheath materials all have a coefficient of thermal expansion higher than that of tube metals, the sheath portion of the SPTC tends to pull away from the furnace tube at elevated temperatures, moving the thermocouple sheath away from its needed heat sink.

•

Large Grain Growth At elevated temperatures, commonly-available sheath metals exhibit large grain growth. Such growth opens paths through the lattice structure to permit the entry of molecular oxygen and weakens the structure of the sheath.

•

Oxidation Oxidation of Type K thermocouple wires decreases the mV/°F generated. Oxidizing the Cr in the KP (Chromel) wire decreases the potential generated for a given temperature; oxidizing the Al in the KN (Alumel) wire increases the potential for a given temperature. Since the Alumel wire is negative with respect to the Chromel wire, the net result is a loss in potential equal to the sum of the component losses. Preferential oxidation of the Cr in the KP wire is sometimes referred to as “green rot,” after the color of the chromium oxide at the point of chemical activity.

•

Corrosion Fuels with high concentrations of sulfur or vanadium can cause severe corrosion of the SPTC sheath at elevated temperatures. For most refinery and chemicals applications, environmental regulations limit the amount of these elements in fuel. However, plant incinerators or sulfur- recovery furnaces may have high sulfur concentrations. Also, some offshore power plants can burn resid with high heavy-metals concentrations.

•

Mechanical Damage Most thermocouple failures occur in the sheath because of repeated bending and movement. This usually occurs in the area where the sheath jumps from the tube to the furnace roof, wall, or floor. It is therefore important that the sheath exit the furnace as close as possible to where the tubes are anchored, to minimize sheath bending during tube expansion and contraction. Thermocouple assemblies that are not properly supported inside the furnace of the fired heater are also subject to mechanical damage during periodic inspection and maintenance.

December 1999

600-24

Chevron Corporation


•


Transition Wiring Thermal Damage At the transition piece, thermocouple wires inside the MIMS cable are connected to more pliable thermocouple extension wire that carries a significantly lower temperature rating. Connections are insulated, and the overall splices are epoxied into the transition piece for mechanical protection. If this transition becomes overheated, the epoxy will ooze out of the fitting, permitting shorting of the thermocouple wires and formation of a secondary, low-temperature measuring junction.

•

Damage During Installation Arc welding onto the sheath of the MIMS cable is a common failure mechanism on SPTCs and must be avoided.

Sources of Errors on Conventional SPTCs •

Metal Migration At elevated temperatures, metals with high vapor pressure (Mn and Al) can migrate from the KN wire through the magnesia insulation to the KP wire, altering the potential generated for a given temperature. Mn present in sheath alloys (300-series stainless steels or Inconel) can migrate to either the KP or the KN or both thermocouple wires, inducing errors not quantifiable by magnitude or direction.

•

Thermocouple Shunt effect Magnesia becomes a semi-conductor at elevated temperatures. A Chromel vs. Magnesia and a Magnesia vs. Alumel thermocouple can form at a localized hot spot. This “volunteer” thermocouple generates a potential equal to a Type K couple for a given temperature and is electrically in parallel with the “official” measuring thermocouple. In other words, the indicated temperature is somewhere between the official and volunteer actual temperatures. If the hot spot becomes large and hot enough, the sheath becomes a continuous averaging thermocouple. In this case, the indicated temperature is that of the sheath, regardless of the temperature of the measuring junction. This abnormality is referred to as “high-temperature shunt effect” in scientific journals. The severity of this interference is a function of the following factors: – –

sheath temperature: onset is above 1,500°F. length of sheath at elevated temperature: twice as long = twice the induced error. – diameter of the thermocouple wires: more loop resistance increases the severity. – dielectric constant of the magnesia insulation: higher resistance reduces error. In practice, the ¼-in. O.D. MIMS cable used for fired-heater SPTCs does not exhibit measurable shunt effect below a temperature of 1,850°F. The error

Chevron Corporation

600-25

December 1999



induced is fully reversible when temperatures return to normal, provided that the thermocouple-wire metallurgy is not permanently altered during the hightemperature transient. •

Long-Term Drift At elevated temperatures, an ordering reaction common in all Ni - Cr systems causes a positive drift with time. Positive drift means increased potential for a given temperature. Long-term drift tests of >20,000 hours at temperatures above 1,400°F show an initial downward drift (oxidation), followed by an upward drift from the “ordering reaction” of the Ni - Cr system, and finally a second drift downward (additional oxidation) prior to T/C failure. The speed or magnitude of this drift cannot be predicted, but some errors of up to 25°F have been noted.

•

Radiant Energy Interference For SPTCs used on fired heaters, improperly installed radiation shields and the high- temperature insulation they contain causes a measurement to indicate higher than actual. The magnitude of error is typically 50°F to 200°F. Use of a high-density insulation, e.g., castable refractory, instead of the recommended Kaowool low-density fiber insulation, slows down the SPTC speed of response to changes in actual tube-metal temperature.

•

Secondary Junction Measurement If MIMS cable is crimped or otherwise mechanically damaged during installation, one or both of the thermocouple conductors can come in contact with the sheath or with each other. This contact forms a secondary measuring junction, and the indicated temperature is somewhere between the temperatures of this junction and the official measuring junction. The minimum bend-radius data found in the specifications for various MIMS cables is quite conservative. There is no evidence of secondary junctions forming when these limits are adhered to during installation.

Vendor Recommendation - SPTCs Gayesco has received a patent on the Retractopad removable SPTC assembly, which permits replacement of thermocouple measuring elements and connecting MIMS cable without re-welding on the furnace tubes. Consider this design a standard for all high-chrome furnace tubes that require elaborate pre- and post-weld heat treating once the tube metal has been exposed to process conditions. They will continue to be the sole source for this design for the remaining life of their patent. (Gayesco’s patent on the Refractopad expired several years ago, giving rise to a freshet of copied designs from their competitors.)

Installation Details (See Drawing No. GD-J1201-1) Every fired heater has unique requirements for installing SPTCs. The installation details below are common to most applications. Services which form a coke layer

December 1999

600-26

Chevron Corporation



on the tube ID may require special consideration, particularly with regard to the radiation shield. Contact the CRTC furnace or instrumentation specialist. •

Welding Procedures Procedures for welding the SPTC weld pad to the tube at the point of measurement vary with the size, age, service, and material of the tube metal. Specifics for these procedures must be developed for each furnace by the plant engineering department.

•

Radiation Shields To protect the measurement from radiated-heat interference, radiation shields are installed over the weld pad/thermocouple assembly after its attachment to the tube being monitored. Shields are light-gauge, corrosion-resistant metal channels, often 300-series stainless steel, formed to the radius of the tube and filled with a high-temperature insulation, typically Kaowool rated to 3,000°F. Radiation shields extend from the point of measurement circumferentially around the tube to its dark side. Radiation shields should be installed over the weld pads as soon as possible after completion of weld-pad welding. Verify that the Kaowool insulation supplied with the radiation shield is in place prior to installing the shield. Radiation shields are typically single-pass welded on the front and both sides, leaving the back side, where the MIMS cable exits the radiation shield, unwelded. To insure measurement accuracy, the weld bead geometry must be exactly as the manufacturer specifies.

•

MIMS Cable Routing and Support The MIMS cable runs between the point of heater-tube measurement and the transition to outside the furnace. It should be routed along the coolest possible route to minimize flame-impingement damage and high-temperature shunt effect errors. Typically, this route is on the dark side of the tube, away from the radiant energy of the fuel burners. Also consider routing MIMS cables away from manways and hoisting-gear access doors where they can be damaged during inspection or periodic maintenance. Stainless-steel weld clips are installed to hold the MIMS cable snug against the heater tube on which it is routed, providing a heat sink to absorb as much radiant energy as possible. The spacing of clips is a function of differential expansion and temperature of the furnace tube and MIMS cable along the route to the transition. For high-temperature fired heaters with both sides of the tube fired, spacing of as little as 8 in. might be required. Double-fired furnace tubes should be treated as a special case. Have the installation design and the MIMS cable routing and support reviewed by the SPTC Manufacturer.

•

Chevron Corporation

Point of Entry

600-27

December 1999



All installations must provide access to the SPTC assembly at the point of furnace entry and at the point of connection to thermocouple extension wire. Install Junction Boxes at a point accessible from grade or from walkways or platforms. In general, MIMS cable on vertical-tube furnaces should exit the furnace in the area of the tubing supports—top entry for top hung furnace tubes. This exit location minimizes expansion-coil requirement; the coils must be able to handle only the differential expansion between the fired heater tubes and the MIMS cable—typically a differential of 0.5 µ-in./in./°F. For horizontal tube fired heaters, the wall tube SPTCs should exit adjacent to the point of measurement (for single-side fired heater tubes) or at the coldest end of the furnace (for double-side fired tubes). •

Transitions Wherever possible, use a Unistrut tubing clamp and strut to anchor the fixed end of the expansion coil outside the furnace area. Otherwise, you must build the transition between MIMS cable and low-temperature extension wire small enough to pass through the tubing nut and ferrule of a thermocouple fitting or bulkhead connector. For the transition through the furnace setting and casing, use a 1½-in. pipe section packed with Kaowool or similar high-temperature mineral wool insulation. Use slotted-guide washers and a drilled-pipe cap on the outside of this pipe section to prevent dropout of the mineral wool insulation.

•

Expansion Coils Use sufficient expansion coil to accommodate the expansion (or differential expansion) at the temperature limit of the tube being measured. Wherever possible, locate the expansion coil outside the furnace to minimize the amount of MIMS cable exposed to furnace heat. Doing so reduces the potential for flame impingement damage or for high temperature shunt effect errors. If expansion coils must be inside the furnace, locate them in the shadow of the tube being measured.

•

Junction Boxes and Thermocouple Extension Cable Junction Boxes for terminating MIMS cable and initiating runs of thermocouple extension wire should be used only for this purpose and should not be shared with other signal cables. Because SPTCs are grounded at the field end (at the weld pad on the furnace tube being measured), T/C extension-wire cable and pair shields should be connected to MIMS cable sheaths in the Junction Box. The conductors in Type KX thermocouple extension wire have the same alloy formula as conventional Type K thermocouple wire. Because of slight irregularities in alloying, and the use of low-temperature insulation, this wire cannot be used to form bead-end thermocouples. Thermocouple extension wires carry

December 1999

600-28

Chevron Corporation



the same conductor color-coding as the thermocouple: Yellow, Positive = Chromel (KP); Red, Negative = Alumel (KN). Note

Thermocouple and extension-wire color-coding is a national standard.

The following color codes for ISA Type K thermocouples are used by major industrial nations outside the United States: Nation

Chromel (KP)

Alumel (KN)

Germany

Red

Green

Japan

Red

White

United Kingdom

Brown

Blue

620 Local Temperature Indicators (Dial Thermometers) Locally mounted dial thermometers should be installed on all points of process equipment and piping, where such indication is required for hand control in the field. This includes at least the following locations: •

Outlet water streams from all condensers or coolers

•

Discharge of all blowers and each compressor cylinder

•

Lube oil and cooling water circulating systems for pumps, turbines, compressors and similar mechanical equipment

•

All tanks and storage vessels

•

At important points listed under Section 621 in locations where a multipoint thermocouple indicator is not provided

Dial Thermometer Specification ISA Form S20.14a should be used to order dial thermometers. A copy of the form and instructions on how to complete it can be found behind Tab DS-4780. Dial thermometers for process streams should be nonresettable, hermetically sealed, heavy duty, industrial type with helical coil elements and five- or six-inch diameter dials. Accuracy should be less than one percent of span. Smaller three-inch dials with two percent accuracy may be used on auxiliary systems, such as lube or seal oil. Dial thermometer cases should be stainless steel with an external adjustment screw. Normal operating temperatures for dial thermometers should be in the center one third of the scale. Dial thermometers are available with back connections (angle form), bottom connections (straight form), or every-angle (adjustable). Every-angle dial thermometers should be used to permit adjustment of the viewing angle in the field.

Chevron Corporation

600-29

December 1999



621 Remote Temperature Indicators (Thermocouples) If a facility has a centralized control room, remote temperature indication should be provided on process equipment and piping to check on the operation and performance of the equipment and the temperature instruments. This includes at least the following locations: •

Columns: All inlet and outlet lines.

•

Vessels: All inlet and outlet lines (where different temperatures are expected).

•

Heat Exchangers including reboilers: Inlet and outlet lines from the shell and/or tube sides. Multiunit heat exchangers should have test thermowells between units, and thermocouples only on the inlet and outlet of the exchanger train.

•

Fired Heaters: The inlet line, the outlets from each pass, header pass points from convection to radiant sections, tube wall temperatures on representative radiant section tubes as recommended by furnace supplier (minimum of three per pass), stack flue gas just ahead of the damper.

•

Process Stream Junctions: Downstream of the junction point of all important process streams.

•

Coolers: All liquid product inlets and outlets.

•

All temperature controlling and transmitting instrument locations (as a check for each instrument). A thermowell separate from the controller or transmitter thermowell is required, except in high pressure piping.

•

Each heavy hydrocarbon line having an orifice flow meter. The purpose is to approximate flow corrections with fluid temperature change.

•

If an electronic control system is used and advanced control strategies are implemented to optimize the process, thermocouples or RTDs should be provided at every orifice flow meter in the process.

•

Both lines, when parallel piping is used, such as the twin vapor lines from a large fractionating column. Temperature transmitter thermowells should also be installed in both lines, with the sensing bulb for the transmitter installed in one of the lines. The installation should permit transfer of the sensing bulb from one thermowell to the other.

•

Process Compressors or Blowers: All inlet and outlet lines. Only one point is required on the inlet and outlet of compressors or blowers in parallel on the same service.

Duplication of remote temperature indicating points resulting from a combination or series arrangement of any of the above requirements is not intended.

622 Temperature Test Points (Thermowells) Temperature test points (thermowells) should be provided on process equipment and piping to check on equipment performance and operation on an occasional basis.

December 1999

600-30

Chevron Corporation



Locations of such temperature points should generally include, but should not be limited to, the inlet and outlet of all heat exchangers, shell side and tube side, that are not provided with remote temperature indicators. Duplication of temperature test points with a thermocouple point or a dial thermometer is not intended. Temperature test points that check temperature sensing points should be located within 12 inches of the sensing point. Thermowell specifications are covered in Sub-section 618.

623 Compressor Temperature Alarms and Shutdowns Rotating machinery requires extensive monitoring to assure that operating problems do not develop unexpectedly. The Machinery and Electrical Systems Team determines the extent of monitoring that is required to ensure reliability. Where temperature instrumentation is required, the following instruments are typically used: •

Electric motors are normally furnished with RTDs embedded in motor windings.

•

High-discharge temperature alarms should be provided on each cylinder of important reciprocating compressors. They may also be necessary on less important compressors. The temperature can rise very rapidly, particularly if a discharge valve fails closed. The temperature element should provide rapid response (less than 60 seconds). Thermocouples, not filled thermal systems, should be used for this service. The sensor should be located either in the compressor nozzle or immediately downstream of it, so that it will detect the high temperature, even with blocked flow. The Company prefers a suitably designed internal sensing arrangement.

624 Self-contained Temperature Regulators Self-contained temperature regulators combine a filled thermal system with a direct operated control valve. The fluid in the thermal system operates directly on the metal diaphragm of the control valve. They are relatively inaccurate and subject to temperature droop as the load increases. They are normally specified for simple field temperature control applications such as steam coils on tanks.

625 Temperature Transmitters Temperature transmitters should be specified and installed where it is desirable to control, record, or indicate the temperature at a central location. Field-mounted RTDs or thermocouples can be wired directly to remotely mounted indicators, recorders, or controllers, but this is rarely done (except with the modern distributed control systems) because of possible noise pickup. Temperature transmitters or converters should be used to transmit the temperature signal. The transmitted signal may be either converted to pneumatic (3 to 15 psig) or electronic (4 to 20 mA). The transmitter output should be compatible with the receiver instrument. Descriptions of several types follow.

Chevron Corporation

600-31

December 1999



Pneumatic Temperature Transmitters Pneumatic temperature transmitters should be used when transmission of a temperature signal to a remote pneumatic receiver is required. Pneumatic temperature transmitters should have weatherproof cases of fiberglass, aluminum, or steel coated with epoxy paint. Both blind and indicating transmitters are available. Blind, suppressed range, filled thermal system transmitters should be used because their force-balance mechanisms are more reliable and have fewer moving parts. Thermal lag compensation (rate action) should be provided when fast response is required. The filled thermal system sensor should be connected to the transmitter with capillary tubing. Capillary tubing should be 316 stainless steel with 316 stainless steel armor. Capillary length should be kept as short as possible (maximum of 20 feet) to minimize ambient temperature errors. Accuracy should be within 0.5 percent of the calibrated span. The transmitter should be designed to operate on an instrument air supply pressure of 18 to 22 psig. The output pressure should be 3 to 15 psig. Output should be direct acting. The transmitter should have supply and output pressure gages. The case should include a socket or yoke for mounting on a 2-inch pipe.

EMF/Pneumatic Temperature Transmitters EMF/pneumatic temperature transmitters may be used when a thermocouple signal is available in a control house with pneumatic recorders and controllers. A reliable source of AC power is required. The transmitter should have an output of 3 to 15 psi for 0 to 100 percent of the calibrated range. Adjustable span and zero suppression is required. The transmitters should meet the requirements for the National Electrical Code (NEC) hazardous area where they are installed. This would normally be for general purpose use.

Electronic Temperature Transmitters Electronic temperature transmitters should be used when transmission of a signal to a remote electronic receiver is required. Filled thermal system electronic transmitters are available but are rarely used. Best accuracy and reliability is obtained with integral 100-ohm, three-wire, platinum RTDs. Less demanding services should use thermocouples. With modern distributed control systems (DCS), thermocouples and RTDs can either be multiplexed or directly wired into the DCS. They can be used as process measurement inputs without the need for a field transmitter.

Electronic Temperature Transmitter Specifications Electonic transmitters with smart electronics are preferred. A single smart transmitter may be programmed for either thermocouple or RTD input, thus minimizing spare parts inventories. Thermocouple transmitters should ground the thermocouple at the thermowell. The temperature measuring circuit for thermocouples should be a zero balance-type or

December 1999

600-32

Chevron Corporation



feedback balance-type with integral cold junction compensation for thermocouple and millivolt inputs. RTD transmitters should use resistance bridge-type temperature measuring circuits. Temperature transmitters should have NEMA 4 housings with dual compartments (or equivalent) to separate the electronics from the wiring connections, which should be in an integral or attached junction box. Temperature transmitters should be protected from radio frequency interference (RFI) and electromagnetic and electrostatic interference. They should include adjustable damping to smooth out transient signal noise. Range and span should be adjustable without changing electronic components. Inputs and outputs should be isolated. Burnout protection should be provided to cause fail safe action of the final control element. Loss of input and indication should be upscale for all input types. The transmitter should be loop powered, which means that the operating voltage is carried along the same two wires as the output signal and originates from the receiver instrument. This is commonly referred to as a “two-wire” transmitter. Transmitters that operate on 24-volt DC power are preferred. Consult the manufacturer’s literature for details concerning power requirements and electrical connections. Output should be 4 to 20 mA DC isolated. Other signals, such as proprietary digital communication, are available if required. The transmitter output should be compatible with the receiver instrument. Electronic transmitters should meet the requirements for the National Electrical Code (NEC) hazardous area where they are installed. The minimum should be explosionproof, Class I, Division 2, Group D. Intrinsically safe (IS) electronic transmitters are preferred, but they are not available from all manufacturers. When used in IS systems, transmitters should be specified as IS and should have an IS certification label.

626 Field Temperature Recorders Field temperature recorders should have weatherproof fiberglass, aluminum, or steel cases coated with epoxy paint. If the case contains electrical components, it should meet the electrical classification for the area. The range should be selected so that the normal process temperature is in the middle third of the chart. Direct-connected field recorders with filled thermal systems may be used if instrument air is not available. If instrument air is available, pneumatic receiver recorders should be used with a filled thermal system pneumatic transmitter. Selection of the filled thermal system depends on the required range, span, capillary length, accessibility, and limitations on the size and location of the sensor. The capillary should be 316 stainless steel with flexible 316 stainless steel armor.

Chevron Corporation

600-33

December 1999



The chart should be 12 inches in diameter. The case should include a socket or yoke for mounting on a 2-inch pipe. The normal chart drive has a 7-day rotation with a clockwork drive with an 8-day wind. Electric and pneumatic chart drives are also available.

627 Field Pneumatic Temperature Controllers A field pneumatic temperature controller compares the measured process temperature against a setpoint and sends a pneumatic output signal to a final control element (control valve) which acts to hold the process temperature at the setpoint. These are used when local control of temperature is required and the following conditions hold: •

It is not necessary to change the temperature setpoint from a central control house

•

It is not necessary to record or trend the temperature from a central control house

•

It is not necessary to change the controller tuning from a central control house

•

The controlled temperature is not part of a cascade loop. Temperature controllers may be either indicating or recording. Blind controllers are rarely used

For temperature control applications which require only local control and no transmission to a remote receiver, a pilot-operated pneumatic indicating controller with a filled thermal system should be used. For applications that require a remote controller mounted in a local panel or console, a pneumatic receiver indicating controller and a pneumatic temperature transmitter with a filled thermal system should be used. Refer to Sub-section 614 on filled thermal systems. Span selection should be based on the following guidelines: The span should be 100°F if the normal operating temperature is 200°F or less. If a span of 100°F is not available, select the narrowest span available. The span should not exceed 60% of the operating temperature, if the operating temperature is higher than 200°F.

Field Pneumatic Temperature Controller Specifications Temperature controllers should have weatherproof fiberglass, aluminum, or steel cases coated with epoxy paint. If the case contains electrical components, it should meet the electrical classification for the area. Temperature controllers should be equipped with proportional plus integral (reset) plus derivative (PID) control action. Proportional band should be at least 0 to 500%. Reset action should be adjustable down to 20 minutes per repeat, or longer. Derivative action should be adjustable from 0.05 to 50 minutes. Controller output should be 3 to 15 psig, or 6 to 30 psig, as required by the final control element. The controller options should include anti-reset windup to prevent the integral action from winding up when the controller is in “manual.”

December 1999

600-34

Chevron Corporation



Temperature controllers should include separate indicators for process temperature and setpoint. The setpoint should be easily adjustable from either inside or outside the case, depending on the need to make adjustments. The controller should operate on an instrument air pressure of 18 to 22 psig. The output should be 3 to 15 psig. For offshore applications, the output should be 6 to 30 psig and a higher supply pressure will be required. The controller should include a two-position, bumpless, balanceless, auto/manual transfer switch that is internally mounted to avoid accidental switching. The controller output should be either direct or reverse acting and it should be possible to change the action in the field. The controller should include supply and output pressure gages. The case should include a socket or yoke for mounting on a 2-inch pipe.

628 Temperature Switches Field temperature switches protect equipment or machinery from overtemperature or undertemperature, without reliance on instrumentation located in a remote control house. Signals from them are usually electric but can be pneumatic. Shutdown switches should be independent of other devices.

Electrical Temperature Switches Electrical temperature switches provide on/off contact closures for use either with equipment or an alarm or shutdown system. Figure 600-11 defines the terminology used in temperature switches. Fig. 600-11 Temperature Switch Technology

Temperature switch cases should be either epoxy painted or all stainless steel. They should have stainless steel exterior screws and a hermetically sealed electrical assembly. They should be explosionproof (NEMA 7), weather-resistant (NEMA 4 and 4X), or general purpose (NEMA 1), depending on the electrical classification where they are installed.

Chevron Corporation

600-35

December 1999



The available circuit arrangements are very flexible. Standard arrangements include single-pole, single-throw (SPST), single-pole, double-throw (SPDT), and doublepole, double-throw (DPDT) designs, but units are available with up to four poles. Switches should normally be specified with snap-acting, dual, single-pole, double throw (SPDT) contacts. The contacts should be hermetically sealed and rated to supply the operated device with a minimum 10 amperes at 120 volts AC, and 6 amperes at 28 volts DC. Switch contacts should open in the alarm condition. Terminal blocks or terminal strips should be provided. Dead front or shrouded terminal blocks are acceptable. Entrance bushings to the temperature switch case should be provided. The electrical conduit connection should be ½ inch minimum. The temperature sensor may be a filled thermal system or a bimetallic element mounted in a suitable thermowell. Filled thermal system switches normally have a slow response time, especially in vapor service. Temperature switches should have an internally adjustable setpoint(s) with a calibrated scale. The setpoint should fall in the middle third of the range. Temperature switches are available with either fixed or adjustable differentials between the setpoint and the reactivation point. Fixed differential temperature switches have a single adjustment for setpoint. They are factory set with differentials of 0.5% to 1% of span. On double adjustment switches, both setpoint and reactuation point can be independently adjusted. The maximum differential in such designs is the range of the switch, while the minimum varies between 2 and 8% of span. The type of switch should match the application. Dual control electrical temperature switches are available with two independent switches in the same housing.

Pneumatic Temperature Switches Pneumatic temperature switches are on/off pneumatic controls used to operate equipment, alarms, or shutdown systems. They are most frequently used in producing applications and the output is a 0 to 60 psig pneumatic signal which is part of the control or alarm and shutdown system of the facility. They are also used for local control of packaged equipment. Pneumatic temperature switches should be adjustable and they should use either bimetallic or filled thermal system sensors. The pneumatic switch should be a stainless steel block and bleed-type valve.

EMF/Alarm Relays EMF/alarm relays may be used in control houses where thermocouple signals are available. A reliable source of AC power is required. Dual units should be provided if high and low functions are required, except on shutdown service. The EMF/alarm relay should meet the requirements for the National Electrical Code (NEC) area where they are installed. This would normally be Division 2 or General Purpose.

December 1999

600-36

Chevron Corporation



629 Multipoint Temperature Systems Multipoint thermocouple temperature systems can be either: •

An integral part of the main cathode ray tube (CRT)-based distributed control system

•

A stand-alone thermocouple system that displays all thermocouple temperatures to the plant operators on demand and communicates to the process monitoring computer

•

Stand-alone, solid-state, digital temperature indicators mounted on the instrument panel

For all three systems, connected thermocouples should not use up more than 75% of the system’s installed capacity for thermocouple connections. Thermocouples linked to the temperature system should not be connected to any other instrumentation. Duplex thermocouples should be installed in a common thermowell to provide parallel functions where necessary.

CRT-Based Distributed Control Systems The current makers of distributed control systems can also provide multipoint thermocouple readout. The thermocouples should tie into the distributed control system data highway and the temperatures should read out on the operator’s CRT/keyboard work stations. The temperature information should be accessible to the plant process computer (if provided) through the common data communication system. The system should be designed so that failure of any one component or subsystem will not cause total failure of the temperature system and also so that any fault can be cleared by the plant operators in less than 5 minutes. Techniques such as dual redundancy, subsystem shedding, and self-diagnostics can be used to accomplish this.

Digital Multipoint Temperature Systems (Digital TI System) For the past twenty years, stand-alone digital TI systems have been available to display thermocouple temperatures. The digital TI system should provide one or more digital select and display stations mounted on the main control panel to provide operator access to the temperature information. Each select and display station should be designed to provide complete and independent access to all input signals. These visual displays should show both the measured temperature and the address or other identification of the point being measured. The select station should have a ten-digit keyboard to address the temperature points on the system. Each station should have access to all temperature points on the system. The system should include a directory which lists all points accessible from that station. Electromechanical devices such as stepping switches or relays should not be used with the exception of the analog signal input multiplexer. The multiplexer may use hermetically sealed, mercury-wetted relays or fully encapsulated contacts, such as magnetic reed relays, to connect the analog signals to the digital TI system input.

Chevron Corporation

600-37

December 1999



Contacts and connectors should be gold plated. Plugs and other electronic junctions should be suited for the specified environment. The system should be designed so that two input signals will not be connected in parallel. Switching of thermocouple types and ranges should be fully automatic. Automatic continuity checks to indicate thermocouple burnout or high loop resistance should be made when the thermocouple is selected. A visual alarm on the display panel should indicate malfunction of the loop and/or the digital TI system. Addition of alarms, computers, or other equipment in parallel with the input should not interfere with the continuity checks, and conversely, the continuity checks should not interfere with the function of the alarms, computers, or other equipment. The digital TI system, including multiplexers, should be field expandable to accommodate additional inputs as specified. As a minimum, the digital TI system should accommodate 10% additional inputs per process unit without additional hardware. The digital TI system should be designed so that the temperature reading will not vary by more than plus or minus 0.1% for the specified ambient temperature and humidity ranges. Remote stations and master stations should be equipped with double-pole circuit breakers for disconnecting AC power from each station. Switches should be provided to separate any remote station bus from the mainframe without affecting the rest of the station. A computer interface should be provided to make block transfers of temperature data to the process computer on demand. The system should be designed such that failure of any one component or subsystem will not cause total failure of the temperature system and also so that any fault can be cleared by the plant operators in less than 5 minutes. Techniques such as dual redundancy, subsystem shedding, and self-diagnostics can be used to accomplish this. Field-mounted remote station cabinets should be weatherproof, designed for the electrical area classification specified, and pressurized to prevent corrosive atmosphere from entering the cabinet.

Multipoint Digital Temperature Indicators (Digital TIs) Small existing plants usually have thermocouples connected to toggle switches on the control panel. Digital multipoint temperature indicators are available for temperature display. The digital TI should be solid-state digital indicators mounted on the instrument panel. Console desks should not be used. The maximum number of temperature points per indicator should be 100. Not more than 75% of the switch capacity installed for each type of thermocouple should be used. Thermocouple switches should be toggle switches (not pushbuttons), with spring-return to neutral position. Switches should be mounted on a separate switch cabinet, and they should be arranged so that additional switches up to full capacity can be added later.

December 1999

600-38

Chevron Corporation



630 Installation of Temperature Instruments 631 General Requirements—Field Temperature Instruments Field temperature instruments should be unaffected by ambient conditions such as temperature extremes, freezing, excessive moisture, high humidity, dust, solar heat, or corrosive vapors.

Accessibility and Visibility Temperature instruments should be located so that they are easy to observe and accessible for calibration and repair. Figure 600-12 gives the access requirements for specific kinds of temperature instruments. Fig. 600-12 Access Requirements for Temperature Instruments Platform or Grade

Stepladder or Rolling Platform

Permanent Ladder

Temperature Transmitter

Yes

Yes

No

Field Temperature Controller

Yes

No

No

Field Temperature Recorder

Yes

No

No

Field Temperature Switch

Yes

No

No

Field Dial Thermometer

Yes

Yes

Yes

Thermocouples and Resistance Bulbs

Yes

Yes

Yes

Thermowell Temperature Test Points

Yes

Yes

Yes

Instrument Type

Pulsation and Vibration Temperature instruments with filled thermal systems use a capillary to connect the thermal bulb to the instrument. The instrument should be separately supported when vibration of piping or equipment could impair instrument performance. Instruments using bimetallic elements are generally vibration resistant. Proper design of thermowells to resist harmonic vibration is covered in Sub-section 618.

632 Specific Requirements—Temperature Instruments Dial Thermometers Dial thermometers should be located so that they are readable from grade or platform (preferably grade). They should be readable from the control valves associated with temperature control to permit manual operation. Dial thermometers should be installed in 304 or 316 (or better) stainless steel protective thermowells, which should be specified with the dial thermometer. Dial thermometers should be identified with stainless steel tags attached with wire. The tag should list the instrument number. The indicating range should be printed on the dial of the thermometer.

Chevron Corporation

600-39

December 1999



Dial thermometers for tanks should be installed in accordance with Standard Drawing GC-D99612 (see the Tank Manual).

Thermowells and Test Wells All temperature sensing elements (filled thermal system bulbs, thermocouples, dial thermometers, etc.) should be installed in thermowells. Installation of bare elements is not recommended. Screwed thermowells should be installed in accordance with Standard Drawing GB-J1196. Flanged thermowells should be installed in accordance with Standard Drawing GB-J1198. Thermowells should be installed in self-draining positions when the process fluid is at or below 32°F. The adjustable threaded union between the thermowell and the thermowell head is optional. It is required with filled thermal system bulbs. Thermowells should be stamped with material type, connection size and type, and the nominal length. Thermowells for furnace stacks should be installed in accordance with Standard Drawing GB-J1200.

Thermocouples Thermocouples should be installed in a suitable thermowell with a union and a thermocouple head. The head is sometimes installed on conduit remote from the thermowell. The tip of the thermocouple should make good thermal contact with the thermowell. The hot junction should be positively grounded to the thermocouple sheath unless prohibited by the design of readout or transmission equipment. The thermocouple should allow positive grounding at the thermocouple head. Thermocouples should not be connected to more than one device. Duplex thermocouples installed in a common thermowell are recommended. Alternately, the thermocouple signals can be converted to 4 to 20 mA. Take particular care with Iron-Constantan extension wire in wet climates. The iron wire is subject to serious rusting unless it is protected with moisture-proof insulation. Thermocouples should be identified with stainless steel tags attached with wire. The tag should list the thermocouple type and stem length.

Resistance Temperature Devices Resistance temperature devices should be mounted in the same manner as thermocouples. The tip of the RTD should make good thermal contact with the thermowell.

Furnace Skin Points Furnace skin points should be installed in accordance with Standard Drawing GD-J1079. Failure of the lead-in wires is the main reason for premature failure of

December 1999

600-40

Chevron Corporation



furnace skin points. The lead-in wire should have adequate flexibility and should be protected from direct flame impingement.

Filled Thermal Systems for Temperature Recorders, Controllers, Transmitters, and Switches The filled thermal system capillary should be protected from mechanical damage by proper supports. The maximum distance between capillary supports should be 18 inches.

Field Pneumatic Temperature Controllers Field pneumatic temperature controllers should be located within 20 feet of the point of temperature measurement.

Electronic Temperature Transmitters Electronic temperature transmitters are normally mounted remote from their associated thermowells to protect against high ambient temperature. They may also be mounted directly on the thermowell.

640 Model Specifications, Standard Drawings and Engineering Forms 641 Model Specifications ICM-DG-4780

Instructions for Ordering Temperature Instruments

642 Standard Drawings

Chevron Corporation

GB-J1195

¾ Inch Screwed Thermowell

GB-J1196

Screwed Thermowell Installations in Piping

GB-J1197

Flanged Thermowell Details for 1½ Inch and 2 Inch Flanged Connections

GB-J1198

Flanged Thermowell Installations in Piping

GB-J1199

Installation Details for Thermocouple Extension Wire and Cable

GB-J1200

Thermowell for Furnace Stack

GB-J1201

Installation Details for Furnace Tube Skin Point Thermocouple

GB-J1202

Installation Details for Reactor Skin Point Thermocouple

GB-J1203

Fabrication Details for Reactor Skin Point Thermocouple

GC-D99612

Standard Thermometer Assembly for Oil Storage Tanks (See the Tank Manual)

600-41

December 1999



650 References Chevron employs the following industry codes, standards, and recommended practices for the design and installation of temperature instrumentation. 1.

API Recommended Practice RP 551, Process Measurement Instrumentation.

2.

Specification Forms for Process Measurement and Control Instruments, Primary Elements and Control Valves, Instrument Society of America, 1975.

3.

ASME Performance Test Codes, Supplement on Instruments and Apparatus, Part 3: Temperature Measurement, and Chapter 8, Bimetallic Thermometers.

4.

ANSI/MC96.1-1982, Temperature Measurement Thermocouples.

The following references are recommended:

December 1999

5.

Temperature Measurement in Engineering, Volume 1, by Baker, Ryder, and Baker, (J. Wiley & Sons, Inc.).

6.

Fundamentals of Temperature, Pressure, & Flow Measurements, by R.P. Benedict, (J. Wiley & Sons, Inc.).

7.

Process Instruments and Controls Handbook, by Considine, (McGraw-Hill Book Company, Inc.).

8.

Brock, J.E., Stress Analysis of Thermowells, Report No. NPS-59BC74112A (Unclassified), Naval Post Graduate School, Monterey, California, November 11, 1974.

600-42

Chevron Corporation

700 Level Measurement Abstract This section introduces common level measurement methods and equipment. It describes various detecting devices for continous and point level measurement of process vessels. It includes discussions of level gages, displacement transmitters and controllers, differential pressure transmitters, capacitance level sensors, ultrasonic, vibration, microwave, nuclear, float switches, level bridles and pyrometers (Rams’ Horns). Information on the application of automatic tank gages and level switches will be found in the Section 700,“Instrumentation/Measurement,” of the Tank Manual. The information in this section includes and adds to information provided about level measurement in API RP 551. Along with discussions of level measurement principles, are application guidelines reflecting Company experience. These guidelines will assist the engineer in the design of a liquid measurement system. This section does not include all of the special methods for unusual situations or the measurement of solid materials.

Chevron Corporation

Contents

Page

710

Gage Glasses

700-3

711

Construction and Operating Principles

712

Application Guidelines

713

Installation Guidelines

720

Float & Displacer Devices

721

Displacement Transmitters and Controllers

722

Float and Displacer Switches

730

Level Bridles

700-18

740

Pressure Transmitters

700-19

741

Principles of Operation

742

Application Guidelines

743

Installation Guidelines

750

Other Electronic Devices

700-1

700-8

700-29

July 1999

700 Level Measurement

July 1999


751

Hydrostatic Level Measurement

752

Capacitance and RF Admittance Level Sensors

753

Ultrasonic

754

Vibration

755

Nuclear Level Measurement

756

Microwave Level Measurement

757

Pyrometers

760

Model Specifications, Standard Drawings, and Engineering Forms700-43

761

Standard Drawings

770

References

700-44

700-2

Chevron Corporation



710 Gage Glasses 711 Construction and Operating Principles For columns (towers), vessels, and small process tanks the most commonly used gage glasses in the petrochemical industry are the armored type with tempered borosilicate glass. The liquid chamber and covers are usually made of steel, but a variety of alloys is available. The covers, which are used to bolt the glass to the body, are usually carbon steel, and have slots for viewing the fluid. Armored glasses are usually specified in multiple sections, which are fabricated together by the vendor to get the desired overall viewing length. Manufacturers can supply these sections in nine glass sizes. However, most facilities try to standardize to one of the larger sizes. The largest size has an overall length of 14-1/8 inches and a viewing length of 12-5/8 inches. This means that when multiple sections are used there are 1-1/2 inch blind spots between sections. The overall length of a gage glass should be limited to five sections or about 69 inches of visible length. This limitation of length is imposed to limit the weight of the unit and to control the stress imposed when gage glasses are mounted on a vessel or tank. Gage glasses are normally used to backup other level instruments when maintenance, calibration, or manual operation is necessary. Rarely are they used as the primary level device, unless installed on small storage vessels containing low usage chemicals. When used as the backup for level transmitters and/or switches, gage glasses should cover the entire operating range, including the spans of the level transmitters and high and low alarm and shutdown switches. It is often desirable to install scales on gage glasses for the following reasons: (1) to help operations keep track of level changes or product transfer rates; (2) to help maintenance calibrate instruments; (3) or when gage glasses are used on small tanks holding low usage chemicals. The two types of armored gage glasses, reflex and transparent, are discussed below.

Reflex Gage Glasses Reflex armored gage glass (see Figures 700-1 and 700-2) has a slotted cover on only one side of the chamber. Longitudinal ridges on the liquid side of the glass act as prisms to reflect light. Above the liquid level light is reflected and appears mirror-like; below the liquid level light is not reflected and appears black in color. Reflex gauges are used for gas-liquid interfaces when clean fluids are present.

Transparent Gage Glasses Transparent armored gage glass (see Figures 700-3 and 700-4) has slotted covers on opposite sides of the chamber to allow the operator to look through the unit. Transparent glasses are required to see the interfaces between two liquids, and when the process fluid is viscous and coats the surface of the glass.

Chevron Corporation

700-3

July 1999



Fig. 700-1

Reflex Glass

Fig. 700-2

Reflex Armored Gage Glass Construction

Fig. 700-3

Transparent Glasses

Fig. 700-4

Transparent Armored Gage Glass Construction

When transparent gauge glasses are used in steam service greater than 200 psig, or to monitor caustics or hydrofluric acid, mica shields are required to protect the glass from chemical attack and discoloration. Reflex gauges should not be used for these services since the mica shielding would hinder the reflection of the prisims cut into the gauge glass. Backlighting with illuminators may be required in these cases. Critical levels which must be continuously observed at night should also have backlighting. See Figure 700-5, which shows typical multiple level gages with illuminators. Normally, flashlights provide sufficient light for operators.

July 1999

700-4

Chevron Corporation


Fig. 700-5


Typical Multiple Level Gages with Illuminators

Tubular Gage Glasses Normally tubular gage glasses are not used because of their fragility. The typical tubular gage may be adequate, however, for some minor utility applications on skidded units. These cases should be carefully reviewed for safety and avoided whenever possible. “Joggler” gage glasses with thick walled borosilicate glass tubes and solid steel armored shields have occasionally been used in the Company in services below 150 psig. These gages are constructed with continuous viewing, red-lined tubes for gas-liquid interfaces, calibration strips, and corrosion-resistant process wetted surfaces. “Halar” tubular gage glasses with impact-resistant plastic tubes have also been used for low pressure and temperature service (e.g., below 500 psig and 300°F).

Magnetic Level Gages Magnetic gages have been used where the fluids being handled are toxic or flammable, and a release of these fluids caused by glass failure would be hazardous to personnel or the environment. A float with an internal magnet is mounted inside a nonmagnetic chamber. The indicator assembly mounted on the outside has magnetized “wafers” which rotate 180 degrees as the magnet passes. The sides of the wafer are distinctly colored to indicate the level. The units have flanged end connections. Magnetic gages should be used in clean services only.

Chevron Corporation

700-5

July 1999



712 Application Guidelines Refer to the project or plant piping classification to determine the temperature and pressure requirements for the gage glass, materials of construction, and the type of process connections. For steam applications be sure to check the Steam Working Pressure (SWP) rating of the gage glass. Coordinate the selection of the gage glass with the design of the piping and vessels. It is most important to determine and document the physical elevations inside the equipment that must be measured. This information will be used to set the elevation and orientation of the vessel or tank connections. If process fluids will foul or coat the glasses so badly they will not be usable, consider another device as a backup, such as a capacitance level transmitter. Gage glasses should cover the entire operating range of all the process instruments. When the process viewing requirements exceed 69 inches, two shorter gage glasses should be used. Each gage glass should have its own process block valves. Backlighting requirements need to be determined with the operating personnel. Backlighting is required on important transparent gage glasses when the liquid level is checked frequently and/or the interface is hard to see because of film or other deposits. When viewing reflex gages is important, increase the area lighting to provide a higher level of visibility at night. If a gage glass on a steam drum is not readable from grade, the ASME Boiler Code requires an additional level indicator at grade. Mirrors for reading elevated gage glasses on steam drums are not acceptable. Vendor supplied options that can be considered for the installation of a gage glass are:

July 1999

•

Weather and corrosion-resistant scales

•

Large chambers for boiling, or turbulent fluids, to minimize paraffin buildup, or for slow flowing viscous fluids

•

Flanged, socket weld or screwed connections both on the ends and sides

•

Illuminators for backlighting

•

Mica shields for steam service over 200 psig or caustic or hydrofluoric acid service

•

Welding pads for vessels and tanks

•

Frost proof extensions

•

Jacketed bodies for heating and cooling

•

Internal tubes for heating and cooling

700-6

Chevron Corporation



713 Installation Guidelines Gage glasses should be installed with dedicated root valves to isolate the gage from the process. Flanged end connections should be specified for gage glasses in services where process piping calls for seal welding or socket welding. Typically, for process applications, the valving will be furnished separately from the gage glasses as specified by the piping classification for the block, drain, and vent valves.

ASME Boiler Code The ASME Boiler Code requires that gage glasses on boilers be installed with offset gage cocks (e.g., specially designed angle valves with an integral shutoff ball check on the vessel side). This is the only application where gage cocks are required. Although many Company facilities use gage cocks, they are not recommended. Gage cocks have been promoted by contractors and vendors because they have built-in ball checks. They are supplied with a drain and vent plug instead of valving, and the offset feature allows them to be easily cleaned. Because of fouling, however, the ball checks are extremely unreliable. There are two types of gage cocks available: 1.

The union bonnet gage cock is designed with adjustable eccentric process connections to compensate for poor vessel fabrication. There have been a number of gage cock failures with union bonnet design, where the handwheel and stem screwed out of the gage cock body and the operator was injured. The stem threads are also in contact with the process fluid and could corrode. Union bonnet valves should not be considered safe.

2.

The outside stem and yoke (OS&Y) gage cock requires accurately located vessel process connections. This gage cock offers no advantages over valves selected from the piping classification, particularly when drain and vent valves are required. This is the only type of gage cock that should be specified for boiler service.

Gage glasses in services that can become coated should be capable of being cleaned with a brush. Piping and platform designers should be aware of this requirement. Gage glasses may require steam or electric tracing and insulation. Hot fluids that are viscous at ambient temperatures need tracing. Some gage glasses may only require insulation for good operation or for personnel safety. This subject is covered in detail in Section 1500, “Instrument Seals, Purges, and Winterizing.” Each gage glass, single or multiple section, shall have its own process, drain and vent valves. Do not “string out” gage glasses. See Standard Drawing GC-J1170 for details of gage glass installation. Normally the drain and vent connections for level glasses are specified plugged. Many times the drain is located over a funnel. Drains and vents are covered in detail in Section 753 below. Between sections there is a 1½ inch blind spot. When knowing the exact level at any time is critical, gage glasses should be overlapped to give continuous coverage.

Chevron Corporation

700-7

July 1999



Care must be taken in the installation of magnetic gages to eliminate steel supports, steam tracing, conduit, or heater wires that may affect the magnetic fields.

720 Float & Displacer Devices 721 Displacement Transmitters and Controllers Principle of Operation and Construction The operation of the displacer is based on Archimedes’ Principle—that a body immersed in a liquid will be buoyed up by a force equal to the weight of the liquid displaced. Displacers are cylindrical in shape (Figure 700-6) so that for each increment of submersion depth an equal increment of buoyancy change will result. The torque tube is designed to twist a specific amount for each increment of displacer buoyancy change. The displacer rod which connects the displacer to the torque tube is designed to absorb lateral forces and minimize friction by use of a knife edge bearing. The torque tube is tubular so that a small rotary force may transmit the degree of rotation accurately to the outside of the vessel. The process end of the torque tube isolates the process from the instrument. A transmitter or a controller is attached to the end of the shaft to convert the rotary motion into varying pneumatic or electric signals.

Application Guidelines Displacers are widely used to measure total liquid level or interface level. Displacers can be supplied as controllers or transmitters. Controllers can be proportional band, proportional band plus reset, or differential gap. Transmitters can have 3 to 15 psig pneumatic or 4 to 20 mA DC electric output signals. Transmitters are normally used when signals are sent back to a remote location such as a control house, or when the level device is mounted on a platform and adjustments may be needed during normal operations. Displacers in total liquid level service can be calibrated with water and then adjusted for the specific gravity of the fluid. It is important to determine and document the elevations of the minimum, midpoint, and maximum process settings in relation to the vessel or tank drawings. This information will be used in determining the displacer cage style required and the vessel connection locations. For cageless displacers it is necessary to specify the length from the head flange to the top of the displacer. Displacers can measure a variety of instrument ranges or spans from 14 inches to 120 inches, though most applications are for 14 inches or less. Span ranges of 32, 48, and 60 inches are often used. For ranges over 60 inches, other types of level sensors are normally used. Because of the weight of the cage, some locations limit displacer lengths to 48 inches.

July 1999

700-8

Chevron Corporation


Fig. 700-6


Cross-Section of a Caged Displacer (Courtesy of Fisher Controls International, Inc.)

Chevron Corporation

700-9

July 1999



Caged, or external, displacers are preferred over cageless, or internal, units. Because caged displacers are outside the vessel, they are: • •

Easier to maintain Less susceptible to turbulence inside the vessel

Cageless displacers can be mounted on the tops of tanks or vessels. They can be considered for use when turbulence is not a factor, when there are no vessel obstructions, when they can be easily removed for maintenance, and when process set point changes are rarely made. Side-mounted cageless displacers are seldom used. Cageless displacers should be considered for interface control applications that require large displacers. They are also advantageous when the liquid contains solid materials, when the liquid has a high melting point and might solidify in a cagemounted unit, or when the application requires a corrosion-proof upper chamber, like teflon-lined steel, and expensive alloy-wetted parts. Cages can be made of steel or 316 stainless steel. Internal wetted parts can be furnished in a variety of corrosion-resistant materials. The pressure and temperature ratings of displacers in some facilities exceed those of the vessel and process piping in order to be able to interchange units. For example, the minimum ANSI rating for all units in a facility may be 300#, even though many applications only require 150# rating.

Installation Guidelines Figure 700-7 shows the four types of nozzle connections available for external displacers. Each has advantages. Selection should be coordinated with the vessel and tank platform design. Displacer gages also come in a variety of process connection configurations (Figure 700-7). Fig. 700-7

July 1999

Displacer Cage Connections

700-10

Chevron Corporation



Some Company facilities standardize on style F-3 which has the least amount of piping flexibility. Displacers should not be ordered until the access to the instrument and to the bridle piping has been finalized. It is also necessary to specify the orientation of the transmitter or controller case in relation to the displacer cage. Right-hand mounting means the instrument case is to the right of the displacer cage when looking at the front of the case. Left-hand mounting indicates the instrument case is to the left of the displacer cage. When specified with mid-flanges, the heads of caged displacers may be rotated in 45-degree increments. Heads may be easily rotated in the field, but the casemounting orientation should be specified before purchasing. The caged displacers should be mounted with separate block, drain, and vent valves. The P&ID may specify test fluid connections and drain funnels. Many facilities provide a separate armored gage glass for caged displacers on the same bridle when the vessel or tank can not be emptied for calibration. If this is not done, cages should be specified with bosses so that an armored gage glass can be added later for calibration purposes. Some facilities use removable tygon tubing which is connected to the drain valve for calibration with nontoxic fluids. Caged displacers can be supplied with either flanged or screwed connections: 2 inch size connections are usually preferred over 1½ inch connections. Cageless displacers are normally supplied with a 4 inch top-mounting flange. Displacers that require steam or electric tracing with insulation should be identified. Hot fluids that are viscous at ambient temperatures need tracing. Some displacers just require insulation for good operation or personnel safety. See Section 1500, “Instrument Seals, Purges, and Winterizing.”

722 Float and Displacer Switches Operation Principles Float Switches The most commonly used level switch is the float switch. In a float switch a metal float or ball rises or falls as the level changes and moves a magnetic attraction sleeve attached to a float rod. The attraction sleeve moves up and down inside a sealing tube. Outside the tube a permanent magnet is attached to a spring loaded pivot with an electrical switch. The switch contacts open or close an electrical circuit (Figure 700-8). Float switches are made with pneumatic relays for use where electricity is not available or the shutdown system is pneumatic, as on some of the smaller platforms. Most level switch applications are specified with electrical contacts.

Displacer Switches Displacer switches work like float switches, but they offer features not found in float-operated switches. The basic sensor of a displacer switch utilizes a pair of

Chevron Corporation

700-11

July 1999


Fig. 700-8


Float Switch (Courtesy of Magnetrol)

weights or displacers, heavier than the liquid. These weights are suspended from a spring. When the liquid engulfs the upper weight, a buoyancy force is produced, which causes the effective weight of the displacer to change, in turn causing the spring to seek a new balance position. This moves the magnetic attraction sleeve into the field of the magnet. See Figure 700-9. Multiple displacers and electrical switches can be combined on a single suspension cable to provide multiple output functions, e.g., pump off, pump on, high level alarm.

Application Guidelines Float Switches Float switches are used when the sensing point requires little or no adjustment. Most float switches are mounted in their own chambers or cages outside of a vessel or tank. Therefore, to change the setting, the piping must be modified within the limits of the vessel nozzles. Float switches work best in clean process services. If applied to dirty or viscous liquids, a testing program is essential in order to ensure that they are functioning. The use of another type of switch or a transmitter signal (dedicated transmitter required for shutdown service) that can be monitored should be considered in difficult services.

July 1999

700-12

Chevron Corporation


Fig. 700-9


Displacer Switch (Courtesy of Magnetrol)

Internal float switches (see Figure 700-10) should be used only when they can be safely removed from the equipment during operation and there is some means to test them conveniently. Most float switches are specified with chambers or cages. External float switch chambers must be specified with either: • •

Flanged or screwed end connections (one inch is standard) Flanged or welded chamber design

The piping classification provides a guideline for specifying the end connections. Facilities normally prefer to standardize in order to have interchangeable units. See Figure 700-11. Welded chamber design is less expensive and should be considered for most applications. Two exceptions are: • •

Chevron Corporation

In fouling or coking services In corrosive services

700-13

July 1999



Fig. 700-10 Typical Internal Float Level Switches

In a welded chamber design, once the float is damaged the device must be thrown away rather than repaired. Contractors like to supply the welded chambers because they are cheaper. Electrical or steam heat tracing can be added to a viscous process application to improve the level switch performance. Float switch wetted parts must be suitable for the process fluid. Normally the floats are supplied in 304 and 316L stainless steels. Chambers or cages are available in carbon steel. Other alloys can be furnished.

Displacer Switches It is very important to have accurate process data on the specific gravity of the fluid at the operating conditions, especially if the specific gravity can vary. Displacer switches are useful when a level is to be measured over a large span or considerable adjustment of the sensing point may be needed during startup or normal operation. The switching point(s) can be easily adjusted by moving the displacer(s) up or down on the suspension wire. Displacers are specified with either a fixed narrow differential or span adjustable, or a wide differential or span adjustable between the displacers. The distance from the top connection to the high level point is also adjustable. These dimensions must be carefully selected, documented, and specified. See Figure 700-12. Displacers can also withstand higher pressures than ball floats, because they are usually solid. They can be furnished with corrosion-resistant parts like ceramic floats and Hasteloy wire. Displacers are normally specified for internal mounting in a tank or vessel, when the equipment can be safely opened and drained for maintenance and calibration

July 1999

700-14

Chevron Corporation



Fig. 700-11 Types of External Float Switches (Courtesy of Magnetrol)

Chevron Corporation

700-15

July 1999



Fig. 700-12 Narrow and Wide Differential Displacer Switches (Courtesy of Magnetrol) Narrow Differential 2-1/2 in. NPT

Wide Differential 4 in. 150# R.F Flange

2-1/2 in. NPT

4 in. 150# R.F. Flange

changes. When the process can not be interrupted for service, displacers should be mounted on external isolatable standpipes. Flange ratings for internal level switches must be specified so they are the same as the vessel or tank flange. Displacer switches with external chambers are often used instead of float switches for low specific gravity services like propane or for high pressure applications requiring Class 600 flanges or higher. These have a narrow differential of less than 1 inch. Displacers can be specified with dual switching stages (see Figure 700-13). The span of operation of each switch must be specified. When dual stage points are required, then another type of level transmitter with a continuous output signal should be considered, especially if the output can be sent to a digital device like a programmable controller.

Electrical Switches and Housings Several types of electrical switches are available. Normally, hermetically sealed electrical contacts are preferred. They provide corrosion resistance as well as eliminate arcing contacts. Both mercury or magnetic switches can be provided, although the electrical current ratings for magnetic switches are much lower and must be greater than required. Micro switches should be avoided. Normally, only one or two sets of single pole double throw (SPDT) contacts are specified. When float or displacer level switches are installed in Class I, Division 2, areas, either a weather tight (NEMA 4) or an explosion proof (NEMA 7) housing may be specified, if the electrical contacts are hermetically sealed. The casting of the NEMA 7 housing may be preferred to provide additional corrosion protection. Most facilities have standardized these requirements.

July 1999

700-16

Chevron Corporation



Fig. 700-13 Displacers with Dual Switching Stages (Courtesy of Magnetrol)

Normally the switch and housing are required to have either a UL or FM listing to comply with local government agencies having jurisdiction. Often a CSA listing is acceptable.

Installation Guidelines Drains, Vents, and Test Fluid Connection When several float switches are “ganged” on a bridle or standpipe, each float switch should have its own required block, drain, vent, and test fluid valves. The bridle or standpipe should also have its own block, drain, and vent valves. See Figure 700-14. Standard Drawing GB-J1170 shows how this can be done. Many environmental designs now require piping the drain valve directly to a closed drain system, instead of using a funnel. The vent valve may also be piped directly to a vapor recovery system. The design of drains and vents must be coordinated with environmental and safety personnel as well as operating personnel. Since these types of float switches will be routinely tested to verify operation, adequate access around the housings must be provided. During hydrotesting, the pressures will often exceed the pressure rating of ball floats inside the external chambers and may collapse them. These ratings must be known, so that the float switch can be blinded during the hydrotest.

Chevron Corporation

700-17

July 1999



Because both float and displacer switches have only a single barrier between the process and the conduit, a drain and seal should be installed in the conduit in Class I, Division 2, areas, to comply with Article 501-5(f)(3) of the National Electric Code. Fig. 700-14 Vents, Drains, and Test Fluid Connections for Gage Glasses and Level Instruments Requirements for Water, CO2, and Non-Toxic Hydrocarbons(1)

Instruments

Requirements for Acid, Caustic, Toxic Hydrocarbons and Acid Hydrocarbon Mixtures(2)

Gage Glass Only

Plugged drain valve and plugged vent valve (clean out)

Plugged vent and drain valves, drain valve located over drain funnel at bottom of glass (3)

Total Level Gage Glass with Level Switch(s)

Plugged vent and drain valves with test fluid connection

Vent and drain piped to drain funnel near bottom of glass. Both valves accessible at bottom of glass. Test fluid connection.(3)

Interface Gage Glass with Level Switch(s)

Plugged drain valve. Vent piped to near bottom of glass. Both valves accessible at bottom of glass. Test fluid connection.

Vent and drain piped to drain funnel near bottom of glass. Both valves accessible at bottom of glass. Test fluid connection.(3)

(1) In hydrocarbon service, where the drain valve terminates on a platform, a drain funnel shall be piped to grade. Drain funnels are not required for water or where a gage glass in hydrocarbon service is located at grade, unless such a termination would present a hazard. (For example, leaking water freezing in a cold climate.) On new installations, drains and vents from level instruments in all services except air and water, should be piped to closed systems and should not be piped to funnels. This practice not only complies with the 1990 Clean Air Act but is also a safer approach. Piping any volatile hydrocarbons (hydrocarbons at or above bubble point) or compounds to funnels endangers operating and maintenance personnel, especially if the vapors are toxic or flammable. (2) Alarms and/or gage glasses in caustic service shall have a water wash connection consisting of gate and check valve at the bottom of the glass. Water shall be permanently connected. The water wash connection shall also serve as the test fluid connection. (3) Where funnels are used, they shall be piped to an appropriate drain system at grade.

730 Level Bridles Bridle Applications When horizontal vessel connections for level instruments exceed 2½ feet in length, level bridles are used to “extend the vessel” closer to the instruments and to provide adequate support for the weight of the instruments. The bridle is usually a 3-inch pipe. The lower nozzle is usually oriented horizontal to the shell of the vessel. Normally bridles have their own block valves at the vessel and their own vent and drain connections. Other reasons for using bridles are:

July 1999

•

For minimizing vessel connections, particularly for stress-relieved vessels. The use of bridles allows for relocation of instruments without welding on stressrelieved vessels.

•

For long level ranges. Bridles reduce the number of vessel connections.

700-18

Chevron Corporation



Instrument Connections to Bridles Connect level glasses to bridles with dedicated block valves. Each level glass should have its own drain and vent valves. Do not “string-out” level glasses with closeconnected nipples. Provide dedicated block valves for each level switch. Drain, vent, and test fluid valves are also provided for each level switch, so they may be maintained without taking several instruments out of service at the same time. Provide dedicated block valves for each displacer instrument, or any other level instrument. Level glasses, level switches, and level transmitters and controllers may be installed on the same bridle. See Standard Drawings GB-J1170 for bridle connections and GB-J1167, GB-J1168, or GB-J1169 for standard vessel connections without bridles.

740 Pressure Transmitters 741 Principles of Operation Hydraulic Head Pressure exerted by the hydraulic head of a fluid can be used to measure liquid level. The relationship of pressure, specific gravity, and height of liquid level above a reference point can be shown as: P=G×H (Eq. 700-1)

where: P = static head pressure, inches of water G = specific gravity of fluid H = level height above reference, inches Since most tank and vessel dimensions are in feet or inches, the level is specified in inches of water. To sense a change in inches of water under process pressure conditions, differential pressure (d/p) transmitters are commonly used.

Differential Pressure Transmitter Application Guidelines Differential Pressure transmitters are one of the most versatile and commonly used instruments for measuring level as well as flow and pressure. They are available in a variety of materials and can be specified with optional chemical seals.

Chevron Corporation

700-19

July 1999



Differential pressure transmitters can be used to measure the liquid level in either open or closed (pressurized) tanks and vessels. The different types of calibration required, depending on the specific installation, are summarized in Figure 700-15. Fig. 700-15 Type of Calibration Required for Various Liquid Level Applications Transmitter Application Service

Initial Conditions

Type of Calibration

Open Tank or Closed Tank with dry leg

Minimum level at datum line

Zero Based

Minimum level above datum line

Suppressed Zero

Minimum level at or above datum line

Elevated Zero

Closed Tank with wet leg

The center of the d/p transmitter measuring element is often called the datum line. All liquid level measurements are referenced from this datum line, or point. When the zero point of the desired level range is above the d/p transmitter, zero suppression of the range must be made. This is normally true for both open tank installations and for closed tank installations with dry legs. See Figure 700-16. In an open tank installation (see Figure 700-17), the high pressure or HI side of the measuring element is connected near the bottom of the tank and the low pressure or LO side is vented to atmosphere. The hydraulic head pressure developed in the HI side is a direct measure of the liquid level. The effect of atmospheric pressure is canceled because this pressure is applied to both sides of the measuring element. Fig. 700-16 Zero Suppression

Fig. 700-17 Differential Pressure Transmitter Used for Liquid Level Measurement on an Open Tank Installation

In a closed tank installation (Figure 700-18), the effect of tank or vessel pressure is canceled by connecting both the HI and LO sides of the measuring element to the tank or vessel. The HI side is connected near the bottom of the tank and the LO side is connected near the top.

July 1999

700-20

Chevron Corporation



Fig. 700-18 Differential Pressure Transmitter Used for Liquid Level Measurement on a Closed Tank Installation

The line between the LO side of the measuring element and the vapor space at the top of the vessel can be either wet or dry depending on the characteristics of the process vapor. Any change in the liquid level in this leg will cause measurement error. For d/p transmitters with wet legs, the fluid in the process lead will exert a head pressure on the LO side of the d/p transmitter’s measuring element. A zero elevation of the range must be made. See Figure 700-19. Fig. 700-19 Zero Elevation

Chevron Corporation

700-21

July 1999



Wet and Dry Legs In a closed tank or vessel installation, the condition of the LO side connecting line is very important. Care must be taken to make sure that this line is either completely free from any liquid (dry leg) or is always fully filled (wet leg). Any change in liquid level in a wet leg or any liquid accumulation in a dry leg will cause a significant error in the level measurement. If a dry leg is used, care must be taken to make sure that condensate does not collect in the measuring element. If a wet leg is used, the immiscible fluid in this leg must remain at a constant level under all process conditions. When the process vapor is not condensable, a dry leg can be used. Some of the techniques used to make sure no liquids or fluids collect in the measuring element are: •

Install a liquid trap at the bottom of the process lead

•

Run the process lead vertically up for several feet after leaving the tank before reversing the direction down to the instrument

•

Heat trace the process lead

•

Purge the process lead with inert gas; use bubbler type installation with inert gas purges

•

Use matching chemical seals

A combination of these techniques will ensure a successful installation of a dry leg. When the process vapor is even slightly condensable, a wet leg should be used. Process vapor will condense even when the process temperature is close to or higher than ambient temperature. To prevent condensation and possible plugging in the instrument leads, chemical seals may be used, provided the capillary length of each leg does not exceed 25 feet. The interface level between two clean fluids can also be measured using a d/p transmitter. For interface measurement, the low-pressure lead must always be immersed in the lighter density fluid and the high-pressure lead must always be immersed in the heavier fluid. The low-pressure lead must never see a vapor phase above the low density fluid.

Span and Range Calculations Span is a function of the change in level. The range values are the end points of the span. The lower range value is the net head pressure applied to the measuring element at minimum level. The upper range value is the net head pressure applied at maximum level. The required span and range values must fall within the limits specified for the transmitter being specified. To determine span and range values for a specific application use the following equations:

July 1999

700-22

Chevron Corporation



Span = A × Gt (Eq. 700-2)

Lower Range Value = (S × Gt) - (E × Ge) (Eq. 700-3)

Upper Range Value = (A + S)Gt - (E × Ge) (Eq. 700-4)

where: A = dimensions in inches as shown in Figures 700-17 and 700-18 S = dimensions in inches as shown in Figures 700-17 and 700-18 E = dimensions in inches as shown in Figures 700-17 and 700-18 Gt = specific gravity of the liquid in the tank Ge = specific gravity of the seal fluid in the wet leg (if different from the liquid in the tank) Example 1: Open Tank (Figure 700-17) Determine span and range values for a Zero Based calibration: Given: A = 100 in.; S = 0 in.; Gt = 1.2 Span = 100 × 1.2 = 120 in. water Lower Range Value = 0 Upper Rate Range Value = 100 × 1.2 = 120 in. water Calibrated range is 0 to 120 in. water Example 2: Dry Leg (Figure 700-18) Determine span and range values for a Suppressed Zero Calibration. Given: A = 100 in.; S = 10 in.; Gt = 1.2 Span = 100 × 1.2 = 120 in. water Lower Range Value = 10 × 1.2 = 12 in. water Upper Range Value = (100 + 10) 1.2 = 132 in. water Calibrated range is 12 to 132 in. water Example 3: Closed Tank - Wet Leg (Figure 700-18) Determine span and range values for an Elevated Zero calibration.

Chevron Corporation

700-23

July 1999



Given: A = 100 in.; S = 10 in.; E = 130 in.; Gt = 0.9; Ge = 1.1 Span = 100 × 0.9 = 90 in. water Lower Range Value = (10 × 0.9) - (130 × 1.1) = -134 in. water Upper Range Value = (100 + 10)0.9 - (130 × 1.1) = -44 in. water Calibrated range is -134 to -44 in. water The minus signs indicate that positive pressures must be applied to the LO side of the measuring element when calibrating a transmitter for this range.

Liquid Interface Calculations When two clean immiscible liquids with different specific gravities are in a tank or vessel together, the d/p transmitter can be specified to measure the height of the interface between the maximum and minimum operation point. The calculation for a typical level interface, as given in Figure 700-20 would be: Span = (b) (G2-G1) (Eq. 700-5)

∆H max = (a) (G1) + (b+c+d) (G2) - (h) (G1) (Eq. 700-6)

∆H min = (a+b) (G1) + (c+d) (G2) - (h) (G1) (Eq. 700-7)

where: G1 = specific gravity of upper liquid G2 = specific gravity of lower liquid a = distance between upper process connection and maximum measured interface level b = span within which interface level is to be measured c = distance between lower process connection and minimum measured interface level d = distance d/p cell is mounted below the lower process connection h = distance between the d/p cell and the upper process connection (a + b + c + d) Example: G1 = 0.8

July 1999

700-24

Chevron Corporation



G2 = 1.1 a = 5 inches b = 50 inches c = 10 inches d = 20 inches then Span = 15 inches ∆H max = 24 inches ∆H min = 9 inches (suppression)

Fig. 700-20 Typical Level Interface

742 Application Guidelines Differential pressure transmitters should be considered whenever the operating range is 48 inches or greater. Differential pressure transmitters with diaphragm seals and capillaries should be considered for shorter ranges in chemical (corrosive) and plugging applications. First, the maximum and minimum operating levels must be determined and documented. Then the tank or vessel connections as well as the location of the d/p transmitter can be determined. One of the advantages of using a d/p transmitter is that the

Chevron Corporation

700-25

July 1999



operating levels can be changed after startup with only a recalibration of the transmitter. The materials of construction for the instrument and process tubing should be specified from the piping classification and corrosion information. These should agree with standardized requirements for the facility. If process conditions are unusual, then several alternative d/p transmitter installations should be considered. Highly corrosive liquids could damage the instrument and cause hazardous leaks. Slurries could cause plugging. Liquids that are viscous at ambient temperatures could cause plugging or false signals. Two successful ways to modify d/p transmitters to solve these problems are to use bubblers and chemical seals.

Bubblers Bubblers are the simplest level measuring devices suitable for corrosive liquids, slurries, and most viscous liquids. They have also been successfully used on materials that solidify at ambient temperature. Figure 700-21 shows a typical bubbler installation. Fig. 700-21 Bubbler Systems with d/p Transmitter

Only the bubbler tube needs to be corrosion-proof. A variety of corrosion-resistant tubing materials are available: Hasteloy C, monel, tantalum, kynar, Teflon, and Teflon-coated steel. Gas that will not react with the process flows continuously through the bubbler tube where it pushes against the hydrostatic head of the liquid. Dry instrument air, nitrogen, or clean natural gas are the three most commonly used gases. A low purge gas flow rate (usually about one scfh) is adequate for most bubbler tubes, which are usually made from ½-inch diameter pipe. The tip of the bubbler tube should be notched or sloped to provide a constant reference point where the bubbles enter the process fluid.

July 1999

700-26

Chevron Corporation



The application of bubblers is normally limited to tanks and low pressure vessels, because the supply pressure of the gas in the facility is typically 100 psig or less. Small variable area rotometers with needle valves and constant flow regulators are used to control the continuous gas flow. The d/p transmitter may be mounted above the maximum operating level for convenience and to prevent plugging of the process leads in case the purge gas fails. Bubblers should not be used in saturated solutions, because of solids buildup and plugging. Wide process temperature swings may prohibit their use in viscous applications. The level measurement can be affected when the density of the liquid is stratified.

Chemical Seals Chemical seals or diaphragms have also been successfully used with d/p transmitters to prevent plugging or where corrosive fluids are involved. The hydrostatic pressures are sensed at the diaphragms that are either connected directly to the d/p transmitter or through remote chemical seals connected by armored capillaries. In both cases the seals and transmitters are factory sealed with special inert fluids like silicone. The capillaries must be the same length or “matched,” and are limited to 25 feet. Flush-type diaphragms (see Figure 700-22a) can be mounted directly on valves. The diaphragms are made of a variety of corrosion-resistant materials, such as: 316 SS, monel, Hasteloy C, tantalum, and Teflon-coated 316 SS. These diaphragms normally have 3-inch flanges. Extended diaphragm d/p transmitters are also available (Figure 700-22b). These are mounted directly on 4-inch tank flanges with the diaphragm protruding inside the tank. The choice of corrosion-resistant wetted materials for the diaphragm extensions is limited. Differential pressure transmitters with remote diaphragm seals can eliminate the need for wet or dry legs with their associated seal liquid, condensate, or purge design considerations (Figure 700-23).

743 Installation Guidelines After the type of d/p transmitter installation has been selected, the initial calculations made, and the specifications for the transmitter determined, follow the installation guidelines for differential pressure transmitters from standard drawings (e.g., GB-J1171, GB-J1172, or GB-1173), or from other sources when bubblers or chemical seals are specified. Process leads can be 3/8 inch or ½ inch seamless Type 316 stainless steel tubing. Process leads should be sloped towards the instrument, as mentioned on the installation detail drawing. Process connections are normally ¾ inch. Installations with wet legs connected to the low side of the d/p transmitter should be filled with a seal fluid. The seal fluid must be compatible with the process and

Chevron Corporation

700-27

July 1999



Fig. 700-22 Flat Diaphragm Cell d/p Transmitter (a) and Extended Diaphragm d/p Transmitter (b)

Fig. 700-23 Differential Pressure Transmitter with Remote Diaphragm Seals (Courtesy of the Foxboro Company)

July 1999

700-28

Chevron Corporation



suitable for all environmental conditions, such as high process temperature and freezing ambient temperature. It may be necessary to insulate or heat trace this lead to keep the seal fluid from freezing. Filling tees and constant head pots should be included in wet leg installations. Provisions should be made to allow the operator to safely fill and check wet legs. Drain pots should be installed at the bottom of dry legs to catch any unforeseen condensation. Armored capillaries for remote diaphragm chemical seals should be protected by supporting them in tube or channel.

750 Other Electronic Devices 751 Hydrostatic Level Measurement Operation Principles Hydrostatic level measurement works on the same principle as differential pressure level measurement. It measures the hydrostatic head in a vessel. In atmospheric vessels, two “smart” electronic pressure transmitters (accuracy 0.05% of upper range value), are mounted at predetermined elevations and are connected to a microprocessor based system (see Figure 700-24). Since the difference in elevations is fixed and is known, the difference in readings allows the microprocessor to compute the density of the fluid. Using the computed density and the hydrostatic head reading, the microprocessor can then accurately compute the total mass or total level in a vessel. If the vessel is pressurized, a third transmitter must be used to measure and to compensate for vessel pressure. Fig. 700-24 Hydrostatic Level Measurement System Schematic

Application Guidelines Because the difference in elevations of the transmitters must be sufficiently great to allow the microprocessor to calculate the density (5-ft), hydrostatic level

Chevron Corporation

700-29

July 1999



measurement is best suited for vertical storage vessels and tanks. Hydrostatic Tank Gauging is described in detail in the Chevron Tank Manual. Level measurement by hydrostatic principle can be affected by product density stratification and by thermal stratification. Because the density measured by the two lower pressure transmitters is used to compute the level, if stratification exists, the measured density will not represent the average density and the computed level will be affected. Hydrostatic level measurement has been successfully used by several Chevron Chemical facilities and by many Chevron USA Marketing facilities.

752 Capacitance and RF Admittance Level Sensors There are two common types of capacitance level sensors: straight capacitance measurement instruments and radio frequency (RF) admittance-type sensors. RF means that a reluctance circuit has been added to the capacitance circuit making the instrument operate at a fixed radio frequency of typically 100 kHz. The Company has had limited success with capacitance sensors because they are affected by solids buildup or heavy hydrocarbon coating of the probe. RF sensors are not affected by solids buildup and have been successfully used in services where nothing else has worked. RF admittance-type capacitance sensors take into account both the dielectric constant and conductivity properties of a fluid or solid and therefore level measurement is virtually unaffected by conductivity. This is very important when measuring processes that leave conductive or insulating coatings on the sensing element. Capacitance level and radio-frequency (RF) admittance sensors are used when buoyancy and hydrostatic head instruments will not work reliably. For capacitance level sensors to work, the dielectric constants between the upper and lower fluids must be significantly different. Heavy oil/water interfaces and emulsion pads are two of the most frequent applications for capacitance level instruments. In other industries capacitance probes have been used to measure total liquid level, slurries, and granular materials.

Advantages and Disadvantages The advantages of a capacitance level sensor are:

July 1999

•

They have no moving parts.

•

They are simple and rugged.

•

The probe is designed to resist corrosion.

•

They can be designed to be removed and cleaned without shutting down the process.

•

They can accommodate a wide range of process temperatures and pressures.

700-30

Chevron Corporation


•


Installation is easier than with the typical displacer transmitter or ball float switch.

The disadvantages are: •

If application guidelines are not closely followed and vendor assistance is not used for calibration and training before startup, then problems usually occur and the reliability of capacitance level sensors is often questioned. (Suppliers of this equipment offer training and startup assistance at no extra cost.)

•

If the dielectric constant of the fluid or fluids the sensor is measuring varies or changes, then continuous transmitters or point level will be affected. This happens (a) when vessels or tanks are not dedicated to one service, (b) when the product in a reactor changes continually during the process, or (c) when the process temperature variations change the dielectric constant. For unstratified continuous applications, a second reference sensor may be needed below the lowest liquid level to measure the changing dielectric constant of the material so that the true level may be ascertained.

•

When the probe is inside the vessel, it is difficult to confirm the correlation between the output and the actual level without level glasses or sample connections.

•

For interface applications in coalescers, a continuous level sensor can either indicate the oil/emulsion interface, the emulsion/water interface, or some point in between, depending on how the instrument was tuned during startup.

•

Continuous sensing element assemblies mounted inside coalescers must not interfere with any internals and are difficult to remove and insert because of their length and installation geometry (long probes may need to be anchored on the bottom to maintain a vertical orientation).

•

The cost will be significantly higher when the length of the probe exceeds vendor’s standard length.

Measurement Principles The following discussion of the basic physics of RF level measurement ignores the conductive component of current, which is actually included in the sensor’s modern electronics. The simplified, capacitive only model is sufficient for practical application to level measurement problems. Capacitance is the property of a system of conductors and a dielectric or insulator that permits the storage of electrically separated charges when potential (voltage) differences exist between the conductors. A capacitor consists of two conducting materials separated by an insulator. Figure 700-25 schematically represents a capacitor that is formed by two flat plates. Capacitance is determined by the area of the conductors, the distance between them, and the dielectric constant of the insulator. The capacitance of a flat plate capacitor is given by the following equation:

Chevron Corporation

700-31

July 1999



Fig. 700-25 Capacitor Formed by Two Plates

C = 0.225 K A/d (Eq. 700-8)

where: C = capacitance (picofarads) K = dielectric constant A = area of plates (square inches) d = distance between plates (inches) The sensor is a variable capacitor. It typically consists of a metal rod mounted in a vessel with flanged or screwed connections. It is insulated from the metallic vessel and comprises one of the conductors, while the vessel is normally the other conductor. The liquid in the vessel covers or uncovers the sensor, resulting in a change in capacitance. A simple capacitance system for level measurement has several drawbacks. First of all, after the level has covered the probe and then falls away, there may be a coating left on the probe that can cause it to react as though the probe were still covered. Secondly, the connecting cable between the sensor and the electronics can act as a large capacitor which is not stable with temperature changes. Changes in cable capacitance appear to be the same as changes in vessel level. The RF admittance-type level instrument solves the problem of coatings on probes. It takes into account both the dielectric constant and the conductivity of the material, and can measure the level of most fluids with a conductivity of 20 micromhos/cm or more with standard designs. Below this point frequencies can be adjusted to measure an insulating fluid with a conductivity as low as 1 micromho/cm. To eliminate the problem of the cable acting as a large capacitor which is affected by temperature changes, a three-terminal sensing element is connected to the transmitter with a coaxial cable.

July 1999

700-32

Chevron Corporation



On-Off Sensing Elements To overcome the problems discussed above, three-terminal or “Cote-Shield” point sensing elements have been developed for on-off sensing elements. See Figure 700-26. Figure 700-26 shows an exaggerated view of a three terminal probe and how a coating may look in a level control on-off system with an electrical equivalent circuit of the coating left on the probe. Fig. 700-26 On-Off Three-Terminal (Cote-Shield) Sensing Element Showing Product Coating (Cote-Shield is a trade name of Drexelbrook. Similar design features are available from other manufacturers, but in different trade names.)

The center wire of the coaxial cable leading to the amplifier is connected to the center core of the sensor. The shield of the coaxial cable is connected to the middle element of the probe. The ground wire of the cable is connected to the conduit and thus to the vessel body. The electronic instrument measures only the current that travels from the probe’s center element to ground, because no current flows through the coating. During calibration the ideal place to set the calibration adjustment is midway between the empty vessel capacitance and the capacitance when the sensing element is covered. For simple interface systems this point would be the separation between the oil/water layer. For vessels like a coalescer, the interface would either be the oil/emulsion layer or the emulsion/water layer.

Continuous Sensors Sensing elements used to make analog measurements are mostly the two-terminal insulated (coated) type. The sensing element must provide a change in capacitance that is proportional to the change in level being measured. For a fluid that can coat the probe or for an insulating fluid (i.e., most petroleum products), the capacitance will not be linear if the vessel is an irregular shape or the

Chevron Corporation

700-33

July 1999



probe is mounted off center. In these cases concentric shields, cages, or ground rods are added for linearity and accuracy to make the sensing element parallel to the grounded reference (see Figure 700-27). Probe size is an important part of continuous capacitance level measurement. The differential capacitance over the range of level measurement determines the instrument input span. Normally it is desirable to have a differential capacitance of 10 pF or above for good instrument accuracy. Fig. 700-27 Continuous Sensing Element Assemblies for Coating or Insulating Fluids

Another important consideration for good accuracy and resolution in a continuous sensing element is to keep the ratio between the differential or span capacitance and the empty vessel saturation capacitance in the range of 0.25 to 4.0. Sensing elements for continuous measurement must be mounted vertically and be long enough to extend the length of the desired level range. Forces on any long sensing elements may be great and may require special heavy duty construction, a bottom anchor, or some sort of support from the vessel walls or internals.

Application Guidelines The following conditions must be satisfied when specifying either on/off (point) level switches or continuous level transmitters.

July 1999

•

The vendor must include, at no extra cost, startup field assistance from a trained factory service representative.

•

Maximum operating temperature and pressure requirements must be determined for the probe design.

•

Unusual forces acting on the probe must be determined (like agitation on a long top mounted probe that may require heavy duty construction and/or anchoring).

700-34

Chevron Corporation



•

If required, the compression gland must be of fire-safe design so the sensing element will remain in the vessel and not be an additional source of fuel for a fire.

•

The operating voltage available must be specified. Either 24 volts DC or 115 volts AC can be supplied.

•

NEMA enclosure rating for the electrical classification of the area must be determined (NEMA 7 for classified areas and NEMA 4 for outdoor areas).

•

UL or FM listing requirements must be met for the level electronics unit for the area classification as required by local agencies or plant practices.

•

The fail-safe status of the electrical switch contacts must be determined (either high or low and whether the process fluid is above or below the sensing element. See Figure 700-28).

•

Time delay in switching contacts that may be caused by agitation or waves in a vessel must be considered.

•

Corrosion-resistant materials or coatings that may be needed for the electronic enclosure(s) when appropriate.

•

The compatibility of probe materials with the process must be determined (304 SS is standard but materials like 316 SS and Hasteloy C are available).

Fig. 700-28 Typical Fail-Safe Relay Connections for High and Low Levels (Courtesy of Drexelbrook Engineering Co.)

On-Off Level Switches These conditions should be determined when specifying on-off capacitance level switches: •

Chevron Corporation

The mounting of the sensing element (horizontal or vertical). Horizontal installation is most often specified. If a vertical installation is required, then the

700-35

July 1999



distances between the switching point and the vessel or tank connection must be specified. •

An accurate description of the fluid being measured so the vendor can determine the dielectric constant(s) and verify that the application is satisfactory. The dielectric constant of the fluid being tested should be at least 3 pF (picofarads) or more when the sensing element is covered and uncovered. The fluid conductivity (micromho/cm) and viscosity (centipoise) should be included. Sometimes it is necessary to obtain samples for laboratory measurement.

•

Type of nozzle connection specified, so that the “Cote-Shield” extends through the nozzle, through any product buildup on the wall, and at least 2 inches into the vessel.

•

Requirements for probe removal during operation and/or under pressure. A full port valve, a packing gland, and a safety chain would be needed.

•

Electronics mounted remotely from the sensing element or directly on the probe or sensing head. Remote mounting is recommended for most outdoor applications in classified areas.

Continuous Level Transmitters The following factors should be considered when specifying continuous level transmitters. •

An accurate service description of the fluid or interface fluids must be specified so that the vendor can accurately size the sensing element and calculate the calibration capacitances. This should include the conductivity (micromho/cm) and viscosity (centipoise), as well as the dielectric constant(s) when known. Most materials handled in the oil and petrochemical industry tend to coat sensing elements. Sometimes it is necessary to send samples to the manufacturer to ascertain the properties and the operation of the instrument.

•

All continuous sensing elements are top or angle mounted. They can be long and may have a coating that must be protected from damage.

•

Field indication of the level is often required by the addition of a milliampindicating gage.

Installation Guidelines Arrange for the calibration and training by a factory service representative after installation to ensure success during startup. This service is included at no extra charge by major vendors. Most plants and facilities where capacitance probes are used successfully continue to use field service by factory-trained personnel. For continuous applications:

July 1999

•

Make sure the probe assembly (cage, concentric shield, or ground rod) is properly installed.

•

Be sure there is adequate overhead clearance to be able to remove the sensing element carefully and not damage the coating.

700-36

Chevron Corporation



•

Be sure this assembly does not touch any internals in the vessel or tank.

•

Make sure that it is anchored or supported properly, particularly if it is a long probe and/or if there is a possibility of forces like agitation affecting the probe.

•

Review the installation with the vendor’s field service representative before closing up the vessel or tank.

Make sure the Cote-Shield of the three-element on-off sensing probe extends through the vessel nozzle and into the vessel by at least 2 inches. In order to comply with Article 501-5(f)(3) of the National Electric Code, two barriers must be between the process and the conduit with an opening between them. When the electronics are directly mounted on the sensing element, a drain and seal must be installed. The drain exposes the electronics to the atmospheric conditions. When electronics are remote-mounted they are connected to the sensor with a coaxial cable which acts as the opening between the process and the conduit, but will not allow moisture to enter the electronics enclosure. The remote-mounted electronics design is preferred.

753 Ultrasonic Operation Principles (Ultrasonic) There are two types of ultrasonic level measurement systems. One type consists of a transceiver (ultrasonic wave generating and receiving unit) which emits a signal then waits for an echo before emitting another signal and a transducer (electronics to process the signal). In the second type of system ultrasonic excitations are induced into a metallic probe and the transducer looks for probe vibration frequency shifts as liquid comes in contact with the vibrating probe. In the first type of system the transducer can be located remotely and be connected to the transceiver by cable. In the second type of system the transducer is integrally connected to the transceiver.

Point Measurement Metallic probes are usually used for point measurement. Some probes are designed with a vapor gap, which is the ultrasonic wave path, located near their end. Transceiver components are inside the probe and are located on either side of the vapor gap. Transducer electronics are mounted in a housing at the external end of the probe (see Figure 700-29). Vibration frequency shift probes do not have an air gap and appear to be a solid cylindrical metal rod. A point probe monitors a change of state, and switches a set of contacts whenever a change of state (liquid level reaches the probe or liquid level drops below and uncovers the probe).

Continuous (Analog) Measurement In continuous or analog systems sound waves are emitted towards a liquid surface by the sending unit, the waves are reflected from the liquid level surface, and are sensed by a receiving unit which is located at or integral with the sending unit. The transit time between wave generation and reception is proportional to the distance

Chevron Corporation

700-37

July 1999



Fig. 700-29 Ultrasonic Vapor Gap Point Measurement Probe

that the liquid level surface is from the measurement system. The system can be located above or below the surface of the liquid level being measured and the sound waves may travel either through the vapor to the surface and back (above the surface mounting) or through the liquid (below the surface mounting unit). The “cone of vision” of analog ultrasonic systems is approximately 9 degrees.

Application Guidelines (Ultrasonic) Ultrasonic is an alternative technology for continuous or point level measurement. The top-mounted sensors designed for continuous level measurement can used in many differenct types of liquids. Ultrasonic sensor probe with a vapor gap (Figure 700-29) can be affected by coating, or fouling. If the gap on a point probe is closed (filled) due to fouling or excessive condensation, the probe will not work reliably and is likely to give false signals. Analog ultrasonic systems are not suitable for agitated fluid surfaces, e.g., surface agitation due to “boiling” such as in LPG orammonia storage vessels. Also, the fluid surface must be relatively uniform, i. e., there should be no foaming or froth on the surface of the fluid. Surface irregularities produce random reflections which reach the receiver at differing intervals thus “confusing” the receiver. Because most ultrasonic level systems are installed inside the vessel that they monitor, they can be physically affected by process vapors. Applications should be reviewed to ensure that the vapors will not corrode and will not condense on (coat) the transceiver. Ultrasonic level systems are affected by fluid and vapor properties and the range of ultrasonic level systems can vary from less than 40 to over 150 feet. Changes in vapor pressure, temperature, echoes, etc., can affect the accuracy of a system. Pressure and temperature compensation should be provided since both affect vapor density and wave propagation speed in the vapor.

July 1999

700-38

Chevron Corporation



Installation (Ultrasonic) Point Measurement Point probes can be installed either in a horizontal or vertical orientation (Figure 700-30). To permit withdrawal for maintenance, point probes can be installed through isolating valves/packing glands (see Figure 700-31). To restrict the initial probe withdrawal distance (to prevent the probe from being pulled out too far or"blown" out of the packing gland by process pressure during maintenance) most manufacturers offer metal lanyards (cables) as an option. The lanyards ensure that the probe tip clears the isolating valve allowing the isolating valve to be closed without damaging the probe tip. Fig. 700-30 Point Probe Orientations

Fig. 700-31 Point Probe Shown in Inserted and Withdrawn Positions

Separate power supply and signal wiring must be run to the probe electronics housing.

Analog Measurement Analog ultrasonic systems are usually installed inside the vessel, with the transceiver inserted into the vessel through an opening (flange) at the top of the vessel. The transducer would be located either near the vessel or remotely in a controlled environment. Whenever maintenance must be performed on the transceiver, the vessel must be depressured and purged.

Chevron Corporation

700-39

July 1999



754 Vibration Vibration Level Switches A variation of the ultrasonic probe which monitors shifts in resonant frequency is the vibrating paddle level switch. A vibrating paddle switch works on the principle of detecting a shift in the probe’s vibration (induced versus actual) frequency. Paddle switches usually vibrate at a sub-sonic frequency. Only one or two Chevron facilities have reported success with vibrating paddle switches. Most facilities report poor reliability and excessive maintenance.

755 Nuclear Level Measurement Operation Principle A nuclear level measurement system consists of a radiation source installed on one side of a vessel and a receiving unit (sensing element) on the opposite side of the vessel. Radiation is usually in the form of gamma rays, but the radiation source is selected to emit the lowest radiation level needed for the specific application. The radiation source is encapsulated in a lead lined enclosure and gamma rays are emitted through an aperture in the enclosure. The aperture is designed specifically for each application and determines the range of the device, i.e., whether it is a point measurement or an analog system. The sensing element must match the aperture design. Elongated sensing elements (up to several feet) are used for analog applications. Short sensing element (less than six inches) are used for point measurements. Radiation is directed at the sensing element through the walls and the contents of the vessel. When the vessel is empty, the walls absorb a portion of the radiation. As the vessel fills, the fluid in the vessel absorbs additional radiation. The change in radiation absorption is proportional to the change in level (see Figure 700-32). Fig. 700-32 Radioactive Level Measurement System

July 1999

700-40

Chevron Corporation



Nuclear Regulatory Commission (NRC) permits are required for each nuclear level system installation. An initial permit must be acquired to install a nuclear system in a facility. The process of obtaining permits is fairly rigorous and can be time consuming. Additional permits for subsequent installations are easier to obtain.

Application Guidelines Nuclear level measurement is used for difficult process level measurements where intrusive type level measurement technologies are not practical, e. g., monitoring catalyst level in a hydrocracker reactor (which operates at high pressure and temperature and vessel penetrations need to be minimized).

Installation Installation of a nuclear level measurement system requires careful design and alignment. The radiation source is installed on one side of a vessel and the sensing element is mounted on the side opposite the radiation source. Direction and alignment are critical to ensure that radiation is directed at, and only at, the sensing element. For maintenance access and for stability, a nuclear measurement system should be supported (braced) from the structure that is built around the process vessel. Because nuclear level measurement systems are frequently installed on vessels that operate at high temperature, thermal expansion of the vessel and its consequent movement relative to the structure must be considered in the installation design. Operations and maintenance must be trained in handling radiation sources and in the potential hazards of nuclear radiation. If a source is not completely isolated during maintenance, personnel can be accidentally exposed to nuclearradiation. All personnel working on or around nuclear level systems should wear radiation monitors to alert them if the system was not properly installed, commissioned, or secured and if there is any stray radiation in the area.

756 Microwave Level Measurement Microwave level measurement, also called radar tank level gauging, has been developed primarily for tank level gauging applications and is described in detail in the Tank Manual. Microwave technology provides another means of accurate level measurement, with no moving parts in a tank or vessel. This technology is considered superior to the ultrasonic and it has been sucessfully applied to tank level measurement. Both continuous and point monitors are available. Although microwave systems are not as greatly affected by light surface foam as ultrasonic level gauges, movements such as waves may affect readings. Heavy foam may result in a loss of radar echo. Vessel internals design may also affect the accuracy of a radar gauge. Any obstacles above the liquid surface which fall within a cone of approximately 12 degrees (cone of vision of a radar gauge) may influence the gauge accuracy.

Chevron Corporation

700-41

July 1999



757 Pyrometers Where Used Pyrometer level sensors in Company refineries (also known as “Rams’ Horns”) are widely used in Company refineries for bottom level detection of heavy oil columns, such as a crude unit’s atmospheric and vacuum columns. In more recent installations they are used as backup level devices to nuclear level transmitters.

Construction Figure 700-33 shows a single Rams’ Horn level sensor. The construction details are shown on Standard Drawing GB-J1174, included in Volume 1, Part 2, of this manual. A thermowell and thermocouple are inserted in the outboard end of the 4-inch diameter pipe. Insulation should not exceed the minimum required to meet personnel protection and pour point requirements. Fig. 700-33 Pyrometer Level Sensor Well

Operating Principle Process fluid in the horn loses heat through the light insulation and a horn filled with vapor will lose heat faster than a horn filled with liquid. Comparing temperatures of the various horns indicates which horns are exposed to condensing vapor and which are immersed in hot liquid. The lighter the insulation on the horn, the greater the temperature difference between the vapor and liquid temperatures. In a three horn system typical temperatures of one crude unit from top to bottom might be 440°F, 600°F, and 640°F.

July 1999

700-42

Chevron Corporation



When the vapor enters the horn, it condenses on the walls and drains back through the 2-inch diameter pipe. The temperature sensors in the condensing vapor read the vapor temperature. When the liquid enters the horn, it cools, becomes slightly more dense, and drains back through the 2-inch diameter pipe.

Application Guidelines When used as the primary level device, no fewer than five Rams’ Horns should be used together. When they are used to back up nuclear level transmitters, three is typical. Normally they are spaced 12 to 15 inches apart. The thermocouple readouts should be in a remote location. Each readout normally has its own display.

760 Model Specifications, Standard Drawings, and Engineering Forms 761 Standard Drawings GB-J1158

Level Gage with Screwed Connections

GB-J1159

Level Gage with 150# Flanged Connections

GB-J1160

Level Gage with 300# Flanged Connections

GB-J1161

Level Switch, External Float Type with Screwed Connections

GB-J1162

Level Switch, External Float Type with #150 Flanged

GB-J1163

Level Switch, External Float Type with #300 Flanged

GB-J1164

Level Instrument, Displacer Type with Screwed Connections

GB-J1165

Level Instrument, Displacer Type with 150# Flanged

GB-J1166

Level Instrument, Displacer Type with 300# Flanged

GB-J1167

Vessel Connections for Level Instruments with Screwed Connections

GB-J1168

Vessel Connections for Level Instruments with 150# Flanges

GB-J1169

Vessel Connections for Level Instruments with 150# Flanges

GB-J1170

Vessel Connections and Level Instrument Bridle Connections

GB-J1171

Diff/Press Level Transmitter, Mounted Below Lower Tap

GB-J1172

Diff/Press Level Transmitter, Mounted At Lower Tap

GB-J1173

Diff/Press Level Transmitter, Mounted Above Upper Tap

GB-J1174

Standard Well for Pyrometer Type Liquid Level Indicator

762 Documentation Requirements Commissioning and calibration of certain types of level instruments require a knowledge of process parameters neither provided on P&IDs nor easily measurable in the field by maintenance technicians. Process fluid specific gravity, specific

Chevron Corporation

700-43

July 1999



gravity differential for interface applications, seal fluid specific gravity, calculation of instrument span, zero elevation and suppression, and alarm setpoints need to be provided on a single document which can be used, and easily updated, by the maintenance technician. Two documents required for control system installations (refer to I&CM 100, “System Design,” Section 134, “Documentation”) are the vessel drawing and the instrument data sheet. In order to consolidate all level instrument data for a particular vessel in a single document, the vessel drawing shall contain the following information for each level instrument associated with that vessel. • • • • • • • •

Nozzle elevation and orientation Maximum, normal and minimum levels Type of level instrument Alarm setpoints Specific gravity of process fluid(s) Specific gravity of seal or capillary fill fluids Instrument span with calculations Zero suppression or elevation, if required

770 References

July 1999

1.

Process Measurement Instrumentation, API Recommended Practice 551, First Edition, 1993.

2.

A Practical Guide to RF Level Controls, Drexelbrook Engineering Company, 1981.

3.

Measurement and Control of Liquid Level, Instrument Society of America, 1982.

4.

Tank Manual, Section 700, “Instrumentation/Measurement,” Chevron Corporation.

5.

Process Installation for Differential Pressure Transmitters, Taylor Instrument, 1985

6.

National Electric Code, National Fire Protection Association, 1984 Edition.

700-44

Chevron Corporation

800 Analyzer Instruments Abstract Analyzers monitor a specific process, product quality, or meet environmental and safety regulatory requirements. They also provide timely process analyses for process computer control systems. Analyzers are categorized as on-line analyzers (often called process analyzers), laboratory analyzers, and ambient monitors/gas detectors. This section is directed toward engineers who have varying levels of experience with on-line analyzers or analyzer systems. (Laboratory analyzers and gas detector/ambient air monitors are excluded from this section).

Chevron Corporation

Contents

Page

810

Introduction

800-3

820

Analyzer Project Execution

800-4

821

Analyzer Project Development

822

Analyzer Project Organization

830

Sample Conditioning System Design

831

Designing Sample Systems

832

Steps in Developing a Sample Conditioning System

833

Sample Point

834

General Sample Line Considerations

835

Construction Materials

836

Sample System Components

837

Process Analyzer Sample Systems

840

Analyzer Specification

841

Required Documents

842

Analyzer System Inspection and Acceptance Procedure

843

Analyzer Shelters

850

Analyzer Installation, Commissioning, and Startup

800-1

800-17

800-70

800-75

July 1999

800 Analyzer Instruments

July 1999


851

Analyzer Installation Checkout Procedure

852

Analyzer Commissioning and Startup

860

Calibration and Validation of Analyzer Output

861

Calibration

862

Continuous Validation of Analyzer Output

870

Safety

871

General

872

Sample Line and Sample System Components

873

Leak Detection

874

Sample Disposal System

875

Electrical and Ignition Problems

880

References

800-76

800-80

800-82

800-2

Chevron Corporation



810 Introduction Successful analyzer designs require knowledge of a wide range of disciplines. Analyzers monitor and control processes throughout the refining and petrochemical industry and also help the company comply with environmental regulatory requirements. Analyzers determine the composition or properties of the sample in or near the process. They are expected to give accurate, timely results on a 24-hour-a-day basis with little human intervention or maintenance. All analyzer projects require detailed engineering, proper analyzer selection, and correct installation. The manual user will learn to specify analyzers and the accessories that are necessary for a successful installation. The annotated version of the analyzer specifications included in this manual should be helpful for completing the installation. The general requirements for the installation, startup, and commissioning of analyzers are presented here as well as extensive information about sample system design which is a particularly important and often neglected subject. Detailed information on specific analyzers is not presented in this section but can be found in handbooks and analyzer repair manuals or by consulting with analyzer specialists with access to the Company’s Refinery Analyzer Applications Reference Document (see Reference [15] at the end of Section 800). The following topics are discussed in detail in this manual: Analyzer Project Execution. This section summarizes the important factors in developing and executing a successful analyzer project. These factors include scope development and cost estimating, project organization, and assigning roles and responsibilities. Sample Conditioning System Design. Sample handling system design is difficult and should not be left entirely to the vendor. The purpose of sample-handling systems is to deliver a clean, representative stream sample to the analyzer. Particularly difficult design problems occur in mixed-phase samples and when heat tracing is required. Sample-handling systems should be built around individual, singlepurpose, discrete components that can be interchanged easily or replaced. A crucial factor in sample handling is the order in which conditioning operations are carried out. The purpose of the sample-handling system and the quality of the stream must be kept in mind. Proper Specification of Analyzer, Shelter, Installation, and Bid Evaluation. Specifications in this manual are useful as guidelines for the integration of analyzers and shelters and for the installation of analyzers. The well-documented analyzer specification sheets help attain the proper analytical results required by operations. Proper Installation, Start-up, and Commissioning. Properly designed analyzer systems may become unreliable as a result of utility failure, electronic noise, improper tracing, and lack of maintenance. Calibration and Validation for Computer Control. All analyzers require verification of results for operations’ purposes. This section offers the means of verification and some considerations for proper calibration.

Chevron Corporation

800-3

July 1999



Safety Requirements. This section outlines safety requirements that are specific to analyzer installations and provides referrals to the appropriate codes. Appendices. Appendix E includes typical analyzer drawings. References. A list of published materials on this subject is included for instrument engineers and analyzer specialists.

820 Analyzer Project Execution This section is designed to aid in the execution of analyzer projects. This assistance includes developing the cost estimate and justification, along with the project organization, roles, and responsibilities.

821 Analyzer Project Development The objective of this section is to provide a proposal, justifying the purchase of an analyzer system. Take the following steps prior to requesting approval to purchase an analyzer system: 1.

Review and define analysis requirements

2.

Locate sample point and shelter location

3.

Collect process data

4.

Evaluate and select method of analysis

5.

Prepare preliminary scope

6.

Prepare cost estimate

7.

Prepare justification for analyzer (cost vs. pay out)

8.

Obtain approval to purchase analyzer system

9.

Write analyzer system specification

Step 1.

Review and Define Analysis Requirements

Input from Operations and Maintenance is very valuable. When reviewing and defining analysis requirements, an analyzer specification sheet helps to prompt the necessary responses from Operations and Maintenance. The analyzer specialist’s knowledge of not only unit operations but also analytical and sampling requirements is key to answering questions about the limitation of the proposed analytical installation. Discourage the requesting organization from requiring measurements that are “nice to know.” Measurements of additional components in, e.g., gas chromatographs, could increase the complexity of the system and the cycle time of the analysis. There are, however, several questions to be asked when defining analysis requirements:

July 1999

800-4

Chevron Corporation



•

What does the operational unit expect to accomplish with a new analyzer?

•

Is the analysis feasible?

•

Are there any vendors who have performed this analysis acceptably in the past? If so, who are they?

•

Are there any upsets that may make the analysis difficult?

•

What level of precision and accuracy is required of the analyzer?

•

What composition or properties are required for control?

•

What is the required response time?

•

What is the required on-stream time?

•

What are the requirements for calibration and readout?

•

Are there any special maintenance requirements?

•

Are there any environmental or safety concerns specific to this process stream?

•

What is the required disposal method for the sample?

•

What is the method and frequency of lab testing on the same stream?

Once these questions are answered, the scope of the project can be developed. Step 2.

Locate the Sample Point and Shelter

When selecting the analyzer site, consider the area electrical classification, shelter accessibility, availability of utilities, distance to the control room, and response time of the analyzer. Also consider operation and maintenance requirements for other equipment in the area. •

Where is the sample point and return location?

•

Are there long sample line runs?

•

Are utilities available (steam, power, nitrogen, instrument and plant air, chemical sewers)?

•

Is there an existing house with room for a new analyzer?

•

Is the signal cable in place?

•

How long will the signal cable run be to the control room?

When locating the sample point, consider response time, varying temperature, pressure available, and cleanliness of the sample. Keep the sample transport time to a minimum. Generally, balance the required conditions of the sample against expense and required response time.

Chevron Corporation

800-5

July 1999



Step 3.

Collect Process Data.

Collecting process data helps to determine whether or not the analysis is feasible and, if it is, which method is appropriate. The maximum and minimum extremes during operation upset, start-up, and shut-down should be used. If they are not, a process analyzer specification sheet that itemizes the volume percent of each component to three or four decimal places is useless. It is important to consider any possible contaminants, solids, pressure and temperature swings. Use typical process analyzer specification sheets such as those in the specification section of this manual which have a line-by-line explanation of requirements. Important and often overlooked points include the maximum and minimum temperatures and pressures of the sample and solids content. There are simple design changes that can prevent problems arising from bubble point or dew point of the sample. While various types of probes and filters can be included in the design, the suitability of each type depends upon load requirements and sample characteristics. Consult with CRTC’s M&CS Unit specialists and local maintenance personnel about their experiences with each type. Step 4.

Evaluate and Select Method of Analysis

The analysis method must fit process conditions. Factors that influence the choice of analysis method include reliability, cost, response time, ease of operation, calibration, and maintenance. It is good practice to talk to maintenance personnel, operators, lab personnel, and analyzer specialists before selecting a method. In particular, avoid an analyzer system that is unproven for the particular application unless Operations and Maintenance are aware of this situation and support this work. If you do prefer an unproven system, request references and investigate the application further, before informing Operations and your management about your decision and giving them the reasons and possible consequences. Refer to Figure 800-1 through 800-4 for a list of typical analyzers and applications. Consult with CRTC’s M&CS Unit specialists for a complete list of vendors available for each application and recommendations. Chevron currently maintains alliance agreements with both Applied Automation, Inc. and Rosemount Analytical, Inc. Fig. 800-1

Commonly Used Analyzer System Integrators

ATI

TASC

Applied Automation

Measurementation

Pastech

Step 5.

Prepare Preliminary Scope.

The following elements must be included in the preliminary scope of the project: • • •

July 1999

Design Integration Training

800-6

Chevron Corporation


Fig. 800-2


Composition Analyzers

Analyzer Type

Applications

Typical Manufacturers

Gas Chromatograph (GC)

Column Control Feed Analysis BTU content

Applied Automation(1) Asea Brown Boveri

GC with Temp Program (Distillation analysis)

Feedstocks Column Sidecuts

Applied Automation(1) Asea Brown Boveri

Mass Spectrometer

Ammonia Plant Feed Area Monitoring Ethylene Cracking Units

Asea Brown Boveri Perkin Elmer Fisons(1)

(1) Preferred

• • • • •

Spare parts Startup Design Key project personnel Maintenance support Technical support

Make some decisions early in the process, particularly those involving design and key project personnel. (Review Section 822 for information about roles and responsibilities.) Avoiding these decisions until after the cost estimate is completed may impact both the project cost and schedule. Design. The designer may be a Company employee, an engineering contractor, or an analyzer systems integrator. The designer must be an analyzer engineer. The quality of design work varies tremendously: •

Analyzer systems are generally complex and require a protective environment and conditioned sample.

•

An installation can become maintenance intensive unless the designer is experienced with analyzers.

A wide variety of disciplines are involved in designing an analyzer system properly. Ideally, Operations and Maintenance provide considerable input for the design early in the project. Integration. It is best to select qualified “analyzer system houses” or integrators, keeping in mind that the quality of integrators’ work varies dramatically. If you are unfamiliar with the quality of an integrator’s recent work, visit the integrator shop before submitting a request for quotation. Ask others at your location about their experiences with integrators and consult with CRTC’s M&CS Unit specialists to obtain recommendations about integrators. The current (1996) list of experienced integrators are: Applied Automation, Pastech, ATI, Measurementation, and TASC. Training. Training begins long before the analyzer system is complete. If the analyzer system is new to a particular location, training is a requirement of the project, not the maintenance department. If possible, involve maintenence personnel

Chevron Corporation

800-7

July 1999


Fig. 800-3


Physical Property Analyzers

Property

Applications


Calorific Value (BTU)

Process Fuel Gas Natural Gas Custody Transfer

Fluid Data Applied Automation, Asea Brown Boveri(1) Daniel

Color

Jet Fuel Vacuum Col Sidecut

Ametek Precision Scientific

Cloud Point

Diesel Lube Oil

Precision Scientific(1)

Conductivity

Waste Water Boiler Feed Water Steam & Condensate

Rosemount(1) Foxboro Analytical TBI Leeds and Northrup

Density (Gas)

Gas Purity Fuel Gas Blending

Sarasota (Redland), Solartron (Schlumberger) UGC

Density (Liquid)

Feedstocks Column Sidecuts Products

Automation Products Sarasota (Redland) Solartron (Schlumberger)(1)

Flash Point

Jet Fuel Diesel Fuel

Precision Scientific

Freeze Point

Jet Fuel

Precision Scientific

Octane (Comparators)

Gasoline Blending

Core Labs NIR

Octane (In-line)

Reformer Product

NIR (Consult with CRTC for vendor)

Opacity

Stacks

Rosemount(1)

pH

Acid Strength Waste Water Corrosion Control

Rosemount(1) TBI Great Lakes Instruments Foxboro

Pour Point

Lube Oils

Precision Scientific(1)

Turbidity (Suspended Solids)

Waste Water

Monitek Hach(1)

Vapor Pressure

Gasoline Blending

Precision Scientific Asea Brown Boveri(1)

Viscosity

Feedstocks Crude Col Sidecuts Lube Oils

Precision Scientific(1) Brookfield

Viscosity..less critical

Fuel Oil

Automation Products

(1) Preferred

July 1999

800-8

Chevron Corporation


Fig. 800-4


Analyzers for Single or Grouped Components

Components / Range

Applications


Method

Ammonia ppm

Waste Water

Orion, Ionics

Selective-Ion Flow Injection

Aromatics Sulfur Compounds

Waste Water Stack Gases

Ametek Rosemount 9100 Bovar (Western Research 900 Series) (1)

Ultra-Violet

Hydrogen Sulfide ppm

Gas Streams

Bovar (Western Research) Houston Atlas, MDA Scientific Ametek

H2S/SO2 Ratio

Sulfur Plant Tail Gas

Ametek; Bovar (Western Research)(1)

Methane, Ethane CO,CO2,H2S

Stack Gases Combustion Control Gas Purity

Rosemount Asea Brown Boveri MSA

Nitrogen Oxides ppm

Stack Gases

Rosemount(1) TECO

Oxygen percent

Combustion Control

Rosemount Ametek (Thermox) Yokogawa

In-Stack On-Stack Extractive

Oxygen ppm

Product Gas Blanket Gas

Teledyne Delta F, Anacon

Electro-Chemical

Oxygen (Dissolved)

Waste Water

Rosemount Orbisphere

Phosphates ppm

Boiler Feedwater Steam Condensate

Hach

Silica ppm

Boiler Feedwater

Hach

Sodium ppm

Boiler Feedwater

Orion

Sulfur Dioxide ppm

Stack Gases

Bovar (Western Research) Rosemount(1)

Total OrganiCarbon

Waste Water

Astro Resources, Ionics, Rosemount

Total Sulfur ppm

Fuel Gas

Houston Atlas, Applied Automation(1)

Moisture in Gas ppm

Reformer Recycle Gas Instrument Air Natural Gas

Ametek Panametrics

Water in Liquids ppm

Feedstocks

Panametrics

Water in Ambient Air (Hygrometers)

Humidity

Panametrics

Infra-Red

Titration

Selective-Ion

(1) Preferred

Chevron Corporation

800-9

July 1999



in the physical and operational shop inspection. Schedule training to begin when the factory personnel are on site for start up. Spare Parts. Provide a higher level of reliability by ensuring that there is an adequate stock of spare parts. If necessary, prepare a list of specific, suggested spare parts. Occasionally, manufacturers offer spare parts at a slight discount when purchased with the new equipment. If the analyzer system is a new type for the location, the project should fund the purchase of the items on the manufacturers recommended spare parts list (usually a one-year supply). Startup. For a new installation, startup assistance may be necessary from the manufacturer or systems integrator; and this assistance can also include training maintenance personnel. Make arrangements for assistance and training while developing the project scope. Note: Startup assistance is mandatory if the analyzer technology is new to the location. Step 6.

Prepare Cost Estimate

Listed below are the major factors contributing to the cost of an analyzer project. This information should be available after completing steps 1-5 above. • • • •

Analyzer and sample system design, engineering and integration Analyzer house or shelter design, engineering and integration Training and startup costs Installation Costs – – – – – – – – –

• • • • •

Power Signal Sample lines and probes Steam Cooling water supply and return Sewer connections Concrete pad Control room and/or analyzer room work Craft manpower

Readout devices/Host computer interface Spare parts Engineering and inspection costs Calibration manifold Maintenance and technical support estimated costs

Estimates of cost vary with location, plant size, labor market, and the analyzer requirements. Several volatile factors that influence project cost estimation include:

July 1999

•

Base materials from recent cost quotations.

•

Labor, based on days worked, rate of pay, and efficiency.

800-10

Chevron Corporation



•

Design, typically 10-15% of total project cost and higher, if contracted. (Includes drafting and contract engineering.)

•

Freight, based on mode of transportation and distance. Dedicated transportation should be used with analyzer shelters.

•

Escalation of costs, typically 0.1% per month for labor and material during the course of project.

•

Overhead, typically 50-60% of labor costs.

•

Undeveloped project scope, 10% for miscellaneous materials and labor costs that were not firm on initial project estimate.

Step 7.

Prepare Justification For Analyzer (Cost vs. Payout)

If the analyzer is not installed for safety or environmental purposes, it must be justified on the basis of economy. Calculate payout based on the amount of fuel saved, the increase in product yield, higher operating efficiency, decrease in off-specification product or the price of noncomformance. Include preventive maintenance and technical support costs as part of the cost vs. payout calculation. Process engineers should be able to define the cost savings provided by the analyzer measurement. Step 8.

Obtain Approval to Purchase Analyzer System

This step depends on local practice. It is important to include alternatives and to explain why they were not chosen. This approach reassures management that alternatives were considered. For additional support, it may be useful to discuss your applications with someone from another location or with someone from CRTC’s M&CS Unit. Step 9.

Write Analyzer System Specification.

This manual contains standard specification sheets and an analyzer system specification. The analyzer system specification is also available electronically (MS Word). If you need more detailed information, contact CRTC’s M&CS Unit analyzer specialists. Another valid method of developing analyzer specifications is to employ the Equipment Supplier Alliance (ESA) Process. This process is outlined in the Applied Automation Chevron Alliance document and in the Rosemount Analytical – Chevron Alliance Document. In summary, an analyzer system integrator is selected through other means (not bidding) and the specification is developed in cooperation with them. This improves the quality of the specification and can eliminate the adversarial positions fostered by the bid/award process.

Chevron Corporation

800-11

July 1999



822 Analyzer Project Organization Introduction This section contains information for the engineer who is responsible for the application, selection, purchase, installation, successful start-up, and ongoing success of an analyzer system. It is also useful to contractors, integrators, and any others who need to know how to organize and execute an analyzer project successfully. Included in this section are: •

Specifications and documents necessary for communicating among the owner, builder, and contractor during the engineering phase and construction phases.

•

Requirements for coordination with other groups, planning hints, and a checklist for project roles and responsibilities.

•

Milestones for ensuring the successful completion of the project.

Roles and Responsibilities Analyzer projects involve several different engineering disciplines of the Company, the prime engineering and construction contractor, and the analyzer systems integrator personnel. Close cooperation and interaction of these personnel is necessary for a successful project. During the detailed engineering phase follow the suggested format in EF-885, Analyzer Project Roles and Responsibilities. This form serves as a guide to assign the major responsibilities of each organization and individual. Lead, approval, and review roles are defined for each activity where appropriate. The size of the project may affect the applicability of certain items but in general they have been developed to cover situations found on all analyzer projects. The engineer (COMPANY) or CONTRACTOR, if designated as a lead engineer, should take a lead role in assigning roles and responsibilities. The roles should be defined at a meeting incorporating representatives from Operations, Maintenance, Engineering, Process Control, and Laboratory (where appropriate) so that all affected parties agree to the roles. The roles are defined as follows: Lead - this person or organization is responsible to see that the task is completed on time and that all affected parties buy into the decision. Approval - Final approval authority required before implementation. Review - The person or organization must make timely comments on the items presented. The individuals who take each role can come from the COMPANY, CONTRACTORS, and INTEGRATOR. Of course it is expected that different individuals will take different roles in various aspects of the project. It is also expected that the analyzer maintenance personnel will have input and review roles. Any questions

July 1999

800-12

Chevron Corporation



concerning the implementation of these roles can be referred to CRTC’s M&CS Unit analyzer specialist. Early in the project, work with the contractor to assign roles and responsibilities to avoid project delays and to fulfill all obligations of the project. Be certain to give someone the task of revising existing equipment. Send a copy of the project roles and responsibilities to the analyzer systems integrator along with the request for quotations.

Maintenance and Technical Support (Design) Technical support and maintenance personnel with direct, on-line, analyzer experience and training provide expert knowledge that is key to the successful purchase and installation of analyzer systems. Experienced analyzer engineers often ask that analyzer maintenance personnel be present during early design and review stages of an analyzer project. Involving maintenance personnel early in the project offers several advantages: •

Having the analyzer maintenance technicians make contributions during the planning steps leverages their field knowledge and gives them a pride of ownership in a new installation.

•

Being closest to the actual daily work process, analyzer maintenance technicians are very sensitive to such design considerations as low-maintenance systems, proper maintenance clearances, and safety.

•

Obtaining advice of skilled analyzer technicians helps to reduce the number of field modifications required.

•

The analyzer team which includes maintenance will estimate when Operations should assume responsibility for the analyzer system.

If experienced analyzer technicians and analyzer engineers are not available locally to work on the project, CRTC’s M&CS Unit Analyzer group can provide such support for every phase of design, training, and technical/maintenance support program development.

Analyzer Systems Integrator To ensure a successful installation, Chevron personnel, the contractor, and the analyzer systems integrator must interact closely. The contractor transmits all official communication between the Company and the integrator. Figures 800-1 and 800-2 are lists of acceptable analyzer systems integrators, analyzers, and component manufacturers for the project. The contractor provides a scope of work to each of the analyzer systems integrators on Chevron’s list and asks them to respond with a technical proposal and bid. The contractor also monitors the performance of the analyzer systems integrator, the complete installation, and the commissioning of the analyzer system. The analyzer systems integrator provides easily maintained and correct analyzer installations and ensures that analyzer systems are installed with all required

Chevron Corporation

800-13

July 1999



ancillary equipment. Such equipment includes the means for properly conditioning and transporting sample, protecting the analyzer from the environment, and providing an accurate and reliable output for monitoring and control. The analyzer systems integrator is also responsible for producing the complete engineering design of the final analyzer system and a review of the tasks undertaken by the contractor.

Pre-Bid Meeting The contractor schedules a pre-bid meeting to be held with the integrators approximately two weeks after sending the request for quotation. (Depending on the scope of the project, Chevron may waive this meeting.) During the meeting, the contractor reviews the technical proposal and defines roles and responsibilities, answering questions or resolving technical issues to ensure that the integrators understand the scope of the work. The integrators leave the meeting with sufficient detail to determine the quality of the engineering design, sample system components, analyzers, and ancillary hardware. If necessary, the contractor may revise and re-transmit the bid request or issue an addendum to it to obtain uniform bidding from all integrators.

Bid Review The contractor reviews the quotations and prepares a bid evaluation and recommendation for review by Chevron personnel.

Alternate Bid Approval The project may elect to work with an alliance partner to supply the analyzers, analyze system integration and engineering at a pre-negotiated price. This approach eliminates the time and cost of the bid process.

Project Schedules The contractor coordinates the analyzer work to fit within the schedule for the overall project and advises Chevron personnel and the integrator early enough to determine any cost impact to the Company. In the request for quotation to integrators, the contractor issues a proposed project schedule, with milestones for the installation work. (See following sample.)

Proposed Project Schedule

July 1999

1.

Analyzer system proposal issued for bid

(__/__/__)

2.

Pre-bid meeting

(__/__/__)

3.

Bid quotations due date

(__/__/__)

4.

Contractor and Company review

(__/__/__)

5.

Contractor awards analyzer system

(__/__/__)

6.

Analyzer team kickoff meeting

(__/__/__)

7.

Order long-delivery analyzers, equipment

(__/__/__)

8.

Sample system review meeting

(__/__/__)

800-14

Chevron Corporation


9.


Receipt of preliminary drawings

(__/__/__)

10. Checkout at analyzer manufacturer

(__/__/__)

11. Receipt of certified installation drawings

(__/__/__)

12. Physical inspection at integrator

(__/__/__)

13. Operational inspection at integrator

(__/__/__)

14. Ship to field

(__/__/__)

15. Documentation of as-builts

(__/__/__)

16. Installation

(__/__/__)

17. Field checkout

(__/__/__)

18. Pre-commissioning

(__/__/__)

19. Technician training

(__/__/__)

20. Turnover of system to Operations

(__/__/__)

Bid Proposal Required Format The following is an outline of the required format for a bid proposal submitted by the integrator: •

Submit a technical proposal and bid to the contractor.

•

When proposing a custom item, include a standard commercial item for comparison bid. – –

Indicate the proposed benefits, including cost and performance of the custom item. Include design specifications for the custom item to enable comparison between custom and standard items.

Include the following in the bid proposal: •

One-page project summary.

•

List of equipment identified by manufacturer, model, etc.

•

Pricing summary, itemizing the price of each analyzer, each sample system, shelter, etc.

•

Separate statements of cost of – – – – –

Chevron Corporation

Any special equipment required but not supplied. Engineering and design for analyzer system. Three sets of loop folders for each analyzer system and analyzer shelter, with incremental cost for each additional set. Required calibration cylinders and racks. Shipment of all material.

•

Project execution summary.

•

Project schedule per attached Company specification.

800-15

July 1999



•

Technical comments, exceptions, and clarifications (analyzers, sample systems, calibration, and ancillary equipment).

•

Technical comments on analyzer shelters, buildings (construction, features, lifting methods, etc.)

•

Comments on custom instruments (compare to commercial instruments to indicate performance and cost savings).

•

Special items required for the analyzer system but not included in bid.

•

Optional and alternative equipment or services.

•

Documentation to include all “as-built” or “as-installed” drawings not only as electronic files for computerized drafting system compatible with plant software but also as hard copy.

•

Integrator participation in checkout of designated analyzers at manufacturer’s site prior to shipment.

•

Acceptance test at integrator’s site.

•

Schedule of delivery from acceptance date.

•

Startup assistance.

•

Proposed integrator project team listing.

Detailed Engineering Phase After the awarding of the contract for the analyzer system to the selected integrator, the analyzer team attends a kick-off meeting. Company, contractor, and integrator representatives discuss the project in detail and confirm the scope of the engineering work and finalize the project schedule.

Systems Integration Phase The analyzer systems integrator has the primary responsibility for implementing this phase of the project; however, other members of the analyzer team monitor all aspects of integration to ensure that the analyzer system is delivered on schedule, which allows time for analyzer pre-commissioning.

Physical and Operational Inspections At the integrator’s site, the contractor or Company personnel, or both carry out a physical inspection of the items listed in Section 842. The integrator gives Chevron and the contractor a minimum of two weeks’ notice to schedule the physical and operational inspections. Representatives of the Company and contractor are present at the integrator’s site for the operational inspection during which:

July 1999

•

All major components operate correctly and reliably.

•

Analyzers demonstrate their ability to meet their design specifications.

800-16

Chevron Corporation


•


All analyzers perform a repeatability run of at least eight hours and preferably 24 hours.

For analyzers having a non-linear output, the manufacturer or integrator must provide data defining the relationship between analyzer output and the true value. Further details about operational inspections are included in Section 842.

Installation The installation contractor arranges for all aspects of the analyzer system installation at the Company site. This includes temporary storage of analyzer systems and related equipment, scheduling of Company or contract labor, gathering all necessary work permits, and ensuring adherence to all Company and other related construction guidelines.

Checkout The checkout contractor advises the integrator when the construction phase is completed. The integrator provides, at the Companys request, technical checkout staff to perform an inspection of the installed system and to verify not only that all utilities are properly connected but also that all wiring conforms to the drawings.

Pre-commissioning The integrator tests all aspects of the analyzer systems, including all utilities, safety systems, HVAC, and analyzers. Analyzers are run on calibration standards before running them on the process stream. The contractor audits this work and approves the integrator work before they leave the site.

Turnover to Operations After reviewing the turnover criteria, the contractor turns the system over to Operations at a formal meeting with the appropriate personnel represented. Refer to ICM-DS-4362, Analyzer Enclosure Check-off Specification for details.

Post-audit The post audit is conducted by Company personnel and may be attended by representatives of the contractor, Operations, and the integrator, if appropriate. All aspects of the project are reviewed and both the positive and negative aspects noted. They record improvements for future projects and include these recommendations with site-specific documentation and standards.

830 Sample Conditioning System Design Few process streams are compatible with available analyzers without some sample modification. Hardware is required for taking a sample from the process, for transporting it to the analyzer, and for adjusting pressure, temperature, and flow rate. Hardware is also required for removing particulates, separating phases, scrubbing, and drying. In this section, the function, operation, and limitations of individual pieces of hardware are explained; and, wherever possible, specific

Chevron Corporation

800-17

July 1999



recommendations are made. An overall design guide demonstrating four types of samples is provided below. While you may not find it necessary to design a sample system, you should know what constitutes a good design so that you can approve those submitted by integrator. You may use the steps in Section 832 as a checklist. Neglecting any of these requirements may result in developing an erroneous analysis or no analysis at all. The portion of the overall installation that is left to the customer’s discretion may determine the success or failure of the analytical instrument. Other resources are included as Appendix E: typical sample system drawings for combustion control systems (extractive), pH, gas chromatographs, fuel gas specific gravity, boiler feedwater conductivity, and infrared analyzers.

831 Designing Sample Systems The sample must be representative of the process. The sampling time delay should meet process control requirements. Locate the sample system in a serviceable area, preferably on the outside wall of the shelter. Plan the local design and installation with the advice and aid of the instrument manufacturer, process engineer, and analyzer specialist. Ask an analyzer specialist to review any designs prepared by an integrator and keep the sample system as simple as possible.

832 Steps in Developing a Sample Conditioning System Obtain the required analyzer sample conditions from the vendor. These conditions dictate the degree to which the sampling system must alter the process sample. Failure to supply a sample compatible with the analyzer may affect the accuracy of analysis and may either render the instrument inoperable or cause permanent damage. Most analyzers operate on a single-phase sample that is relatively free of moisture, corrosive substances, and particulate material (rust, scale, or catalyst fines). The sample should be at, or near, ambient temperature and atmospheric pressure. Few samples meet these criteria, therefore, conditioning is necessary. Also, many process streams are well above ambient temperature, therefore, require some means of cooling. A few analyzers are housed in heated compartments and are operated at temperatures as high as 300°F. Sample pressure must be reduced in most sampling systems. Concurrent with reductions in temperature and pressure are problems with maintaining phase integrity. These problems are discussed later. The steps for developing a sample system are as follows:

July 1999

1.

Ensure that there are adequate safety provisions. Safety should be the primary consideration when developing a sample system. Check the area electrical classification and review NFPA 496, the NEC, and API standards.

2.

Determine if there are any special analyzer requirements. What are the conditions of the sample? Are there pressure and temperature limitations? Does it require a liquid or vapor phase sample? Will the output be affected by

800-18

Chevron Corporation



interfering compounds? Can the analyzer be damaged by corrosion, moisture, or particulates? 3.

It is important to follow the correct order for carrying out certain operations for the sample stream. For each sample type, a specific sequence of operations is recommended. (See Figure 800-5). There are exceptions to any operations sequence, and you must decide which optional steps to select.

4.

Review the process piping and visit the site if possible. Choose the sample point where the condition of the process sample fits the needs of the analyzer and maintenance. Should plugging occur at the sample tap, the analyzer could be damaged permanently if the operator were unable to reach a sampling point.

5.

Remove the sample from the process line with a suitable sample probe. The sample probe should extend into the process line one-third to one-half the width of the process line and should be located on the side or top of the line.

6.

Provide a means of sample transport from the probe to the analyzer. The choice of tubing material is dictated by the nature of the sample. The motive force may be a pressure drop across a process pump. Orifices or in-line metering pumps are less desirable choices. A throttle control valve should not be used at all. In some cases, aspirating probes and eductors are acceptable.

7.

Condition the sample to conform to the analyzer requirements as described above.

8.

Provide sample switching arrangements for multistream analyzers and introduction of calibration mixtures. Sample switching should be kept to an absolute minimum or avoided if possible.

9.

Investigate the availability of modular sampling systems. Some manufacturers provide sample conditioning components as options or as an integral part of the analyzer. Although these systems must be reviewed carefully for reliability, they are usually less costly than systems that are developed in-house.

10. Provide for sample disposal. Regulatory agencies are applying increasingly stringent restrictions on venting samples. Samples can often be returned to lowpressure points in the process which saves production costs overall. Sometimes the sample must be returned to a chemical sewer, recycle tank, or a vent or flare header.

Chevron Corporation

800-19

July 1999



Fig. 800-5

Sequence of Operations for Sample Types

Gas Sample with Condensibles Retained 1. Prefilter (hot) 2. Reduce pressure (hot) 3. Cool (above dewpoint, optional) 4. Knockout (optional) 5. Coalesce (optional) 6. Heat 7. See final sequence Gas Sample with Condensibles Rejected 1. Prefilter (hot), optional 2. Cool (to required dewpoint) 3. Knockout 4. Coalesce 5. Heat 6. Reduce pressure 7. See final sequence Liquid Sample to be Vaporized 1. Cool (optional) 2. Prefilter 3. Vaporize 4. Reduce pressure 5. Coalesce (optional) 6. Heat 7. See final sequence Liquid Sample to remain a Liquid 1. Prefilter 2. Cool (optional) 3. Reduce pressure (optional) 4. Phase separate (optional) 5. Coalesce (optional) 6. Warm to analyzer temperature 7. Degas(optional) 8. See final sequence Final Sequence 1. Dry (optional) 2. Filter (When possible, use a self-cleaning filter in the bypass loop and an inline filter in the slipstream to the analyzer.) 3. Regulate pressure (optional) 4. Switch streams (avoid when possible) 5. Switch calibration stream (only if auto cal is required) 6. Adjust flow 7. Temperature match 8. Analyzer or disposal 9. Sample return Note

July 1999

See Section 837 for a more detailed description of some typical analyzer sample systems that have been successfully applied at Chevron.

800-20

Chevron Corporation



833 Sample Point Avoid using an existing sample tap unless it meets the following conditions: 1.

The tap is located at the top or side of the process line.

2.

Mixing is complete and the process reaction is stable.

3.

The best point is selected to provide the analysis needed for process control and is located near the place where corrective control action is applied.

4.

The sample point is located in a live stream away from poor mixing points. Avoid possible water traps, gas pockets, and dead legs in piping.

5.

The sample tap is located where a single phase exists (if possible).

6.

In the process line, there is a point having moderate and constant pressure. Sample circulation loops can be used to minimize lag time.

7.

The sample temperature is moderate and constant. Try to obtain such a sample by taking it downstream from a condenser or heat exchanger. Letting the process do the sample conditioning not only reduces heater or cooler requirements in the sample system but also reduces costs and sample system volume.

8.

The physical location of the sample point is chosen to be readily accessible for maintenance and servicing and has the required utilities available. Sample, return, and drainage problems are minimal. Sample line runs must be direct, short, and located at a distance from hazardous work areas or hazardous equipment.

Sample Probe Typically, sample probe design and placement have been neglected. Do not re-use existing probes. A properly designed probe can be the beginning of sample cleanup. In its simplest form for clean streams, a probe can be an open-ended tube inserted into a process stream. Never insert a probe flush with the wall of the pipe. Samples collected from the walls do not represent the stream because of the creep along the pipe wall and because the flow rate along the wall is nearly zero. For gaseous streams containing mists or suspended solids or for liquid streams containing bubbles or solids, use a probe with the opening opposite the direction of the flow to help reject non-homogeneous materials. For ease of insertion or removal from a pressurized line, provide the probe with a full port gate or a ball valve with packing gland. Be sure to consult with the analyzer specialist, analyzer maintenance, and operations before selection of an insertion probe which can be removed through a packing gland. It is recommended to use a flanged probe which cannot be extracted while the line is operational, when ever possible. The simplest stack probe is a long tube with a series of openings on the opposite side of the flow direction. To remove suspended material completely, add a filter to

Chevron Corporation

800-21

July 1999



the tip of the probe. Common filter materials are sintered or woven stainless steel and porous ceramic. In very dirty streams, minimize filter plugging by adding a baffle to deflect suspended matter. Filtered probes will require occasional cleaning. They can be removed from the stack for cleaning but will commonly be cleaned by a blow-back system that introduces air or steam into the sample line automatically at preset intervals. Use steam eductor (ejector) probes, such as the Taylor steam eductor, for sampling streams at pressures that are near atmospheric or slightly negative. Steam enters the end of the probe through an aspirating nozzle that sucks the sample from the flue into the sample transport line. The sample is cleaned in a water wash separator ahead of the analyzer. To prevent line blocking, use dry steam, place a strainer in the sample line, and slope the probe to grade. Air or water educators are also acceptable and operate on the same principle. Systems are designed with ejectors, located either upstream or downstream of the analyzer. Some systems may require a back-pressure regulator. Immediately following the probe, install a coarse y-strainer to prevent large particles from entering the remainder of the sample system.

834 General Sample Line Considerations 1.

Keep sample lines short to minimize transport lag time.

2.

Consider using a fast sample bypass loop, as shown in the typical drawings in Appendix E.

3.

Select smallest diameter line, suitable for the required flow rate and for the available pressure drop. Capacities for tubing of varying diameters are shown in Figure 800-6.

4.

Provide sufficient pressure to maintain adequate velocities. Suggested flow velocities are 5 to 10 feet/second for liquids and 20 to 40 feet/second for gases.

5.

Install flow indicators and check valves to ensure that the sample flows in the proper direction.

6.

Control the temperature by steam tracing, electrical tracing. or steam or hot water jacketing to prevent the formation of a second phase. Preinsulated steam and electrically traced tube bundles are available commercially.

7.

The following techniques are helpful in minimizing trapped particulates: – – –

Avoid an excessive numbers of fittings. Use tubing bends instead of ells when possible. Ream and deburr tube endings before making up the fittings.

A further incentive for avoiding excessive fittings is to minimize leakage. Inleakage can occur even in relatively high-pressure lines. On vacuum systems, a single-run tube bundle may be appropriate.

July 1999

800-22

Chevron Corporation


Fig. 800-6


8.

Never apply pipe dope in sampling systems. Pipe dope may absorb or desorb trace elements selectively or it may bleed into the sample, which would affect the analysis.

9.

Provide adequate support to avoid root-valve failure.

Volume of Tubing and Pipe

Tube Size Inches

Wall Thickness Inches

Inside Diameter Inches

Volume Cu. Ft. Per 100 Ft.

Volume Gal. Per 100 Ft.

Volume CC Per Foot

1/4

0.028

0.194

0.0205

0.150

5.8

0.035

0.180

0.0177

0.132

5.0

0.049

0.152

0.0126

0.094

3.6

0.035

0.305

0.0057

0.378

14.4

0.049

0.277

0.0418

0.313

11.8

0.035

0.430

0.1008

0.754

28.5

0.049

0.402

0.0881

0.659

24.9

0.065

0.370

0.0746

0.558

21.1

5/8

0.049

0.527

0.1514

1.133

42.9

3/4

0.035

0.680

0.2521

1.886

71.4

0.065

0.620

0.2095

1.567

59.3

0.065

0.870

0.4126

3.087

116.8

0.095

0.810

0.3577

2.676

101.3

Pipe Size Inches

Schedule

Inside Diameter Inches

Volume Cu. Ft. Per 100 Ft.

Volume Gal. Per 100 Ft.

Volume CC Per Foot

1/2

40

0.622

0.211

1.578

60

80

0.546

0.162

1.212

46

40

0.824

0.370

2.768

105

80

0.742

0.300

2.244

85

40

1.049

0.600

4.489

170

80

0.957

0.499

3.733

141

3/8

1/2

1

3/4

1

Chevron Corporation

800-23

July 1999



Sample Line Routing Once the sample tap is located, examine the needs of the tubing for transporting the sample to the analyzer. Accessibility is the most important factor from a maintenance point of view. The locations of the lines are important to replace or tighten fittings to eliminate leaks. The routing should take advantage of any existing structures or lines to which the sample lines can be strapped or of any instrument line trays in which the sample tubing can be laid. Avoid sharp bends or kinks in the lines. If possible, slope the lines down to the analyzer location and do not build liquid traps into the system. Support tubing in trays; clamp it to the structural steel on pipes or, as a last resort, select heavier gage tubing to protect it from vibration and damage during construction. Do not clamp sample tubing to vibrating structures. Avoid low spots in preinsulated tubing bundles by supporting them in trays. These bundles are heavy and will sag from their own weight, therefore, do not support bundles with only unistrut clamps. To minimize transport time, keep the sample and return lines short, usually a maximum of 300 feet. (For additional information, see Sample Line Sizing in this manual.)

Method of Sample Transportation The following factors determine the method selected to transport the sample to the analyzer: The simplest installation, a single line to the analyzer (shown in Figure 800-7a), is acceptable for applications having the analyzer field mounted at the sample point to keep the sample lines short. Sulfur plant tail-gas analyzers are installed this way to eliminate the possibility of sulfur plugging the sample lines. The sample is returned to the process because it is hazardous. A single line with a bypass stream is the most efficient way of transporting samples to an analyzer (shown in Figure 800-7b). This system minimizes sample waste and reduces time lag in the sampling system to an acceptable value. Normally, a oneminute lag time is considered to be the maximum allowed for sample transportation. Analyzers capable l of faster analysis and used in control loops may require a more rapid sample circulation. When sufficient pressure is not available. select a single-line sample system with a steam, air, or water eductor that pulls the sample through the analyzer sample valve. This system, shown in Figure 800-7c, must be installed with care. It is under vacuum and any leakage in the sample system, up to and through the sampling valve, results in air being drawn into the system. Drawing air into the system makes the analysis worthless. As leakage is difficult to detect, use long, uninterrupted runs of preinsulated tubing to decrease the number of fittings to be checked for leaks. A back-pressure regulator is required when the eductor is in this location.

July 1999

800-24

Chevron Corporation


Fig. 800-7


(a, b, & c) Single-line Sample Systems

Chevron Corporation

800-25

July 1999



Select single-line sample systems only when response time is unimportant and the material is non-hazardous. The obvious limitations of sending the fast loop to waste are as follows: •

Pollution Avoidance - The handling and disposing of toxic sample is an issue to be resolved because waste vented to the sewer or to the atmosphere is monitored and may require clean up in another process unit.

•

Cost - Disposing of waste that consists of large quantities of hydrocarbons or other material can be expensive.

Figure 800-8 illustrates various methods of obtaining sample circulation loops.

Process Pumps Process pumps are the best means of creating differential pressure for sampling systems and is accomplished by taking the sample inlet off the pump discharge and the sample return to the pump suction. Select this procedure whenever: •

a relatively constant differential pressure is maintained.

•

the pump is not so remote from the process control point that time lag in the process piping becomes a factor.

Make certain that the sample circulating back to the pump suction does not subject the pump to overheating. Overheating is a concern if there is a restricted process line downstream of the pump. This arrangement would increase the flow through the sample loop and, in effect, open a bypass from the discharge of the pump to the pump suction. If allowed to continue over a long period, this action could cause the pump bearings to overheat.

Process Equipment The amount of differential pressure that is available for a sample circulating loop depends on the process and its auxiliary equipment. The pressure drop across a heat exchanger and an accumulator can provide sample circulation to the analyzer.

Control Valves These installations are normally avoided because control valves tend to provide variable differential pressure. The control valve operation and the sample loop interact when the control valve is operating at travel extremes. The differential pressure providing sample flow in the sample loop is insufficient when the valve is fully open; conversely, all of the process pressure differential is applied to the sample loop when the valve is fully closed. This condition causes considerable flow around the valve, possibly causing process upsets. If you choose this type of installation, calculate the circulating loop line size based on the existing differential pressure when the valve is at design flow. From this calculation, be certain that, if the control valve is fully closed, the flow in the sample loop will not exceed 10 percent of the control valve design flow. If the amount of bypass flow is too great, investigate another means of transporting sample flow to the analyzer.

July 1999

800-26

Chevron Corporation


Fig. 800-8


Circulating Sample Loops

Chevron Corporation

800-27

July 1999



Restriction Orifice Installations of this type are restricted to low, differential pressures since small restrictions in the process lines are unacceptable. Orifice runs for flow measurement cannot provide analyzer-sample circulation as the analyzer flow bypasses the flow-measuring device, thereby, causing an error in the flow measurement. Restriction orifices may also be subject to clogging and waste process energy.

Sample Pumps Sample pumps are the least desirable alternative as they require significant maintenance. Select sample pumps only if there is no source of pressure differential in the process unit. Avoid positive displacement pumps that are a potential hazard if any portion of the sample loop is blocked downstream of the pump. Install a proper relief valve to maintain sample system pressure at a safe level should any equipment malfunction. Do not install internal relief valves in the pumps as they might overheat if the internal relief valve should open. but instead install valves that relieve to a remote process line. The section entitled. “Sample System Components,” describes different types of pumps.

Eductors In certain situations, eductors may replace pumps. Eductors have no moving parts. They are shown in stack sampling in the drawings in Appendix E of this manual.

Other Considerations Observe certain precautions when circulating loops handling volatile liquids are to be sampled in a vapor state. Always choose a sample-return-point pressure of at least 30 psig higher than the vapor pressure of the sample. This precautionary measure prevents flashing of the sample in the circulation loop and the resultant non-representative sampling. By installing a liquid bypass stream, it is possible to reduce lag time for liquid to be vaporized. Minimize the amount of wasted sample by vaporizing and pressurereducing the sample adjacent to the process sample connection. By using a probe, minimize the volume of pipe, valve, and fittings upstream of the vaporizer. Remember that there is a 300:1 volume ratio of vapor hydrocarbon to liquid hydrocarbon which means that a dead liquid leg causes a long delay and renders the sample non-representative of the process. In situations where extremely long sample lines are a necessity (such as retrofitting an analyzer into an existing house), install a bypass loop (see Figure 800-9). Avoid installing this type of loop indiscriminately because it may be possible to overload the filter or other sample system components in the secondary loop. A sample pump may be required to achieve the extra flow in the longer lines.

Sample Disposal When a sample return point does not exist in the process, it is necessary to decide whether or not to discard the sample effluent to the sewer system or to provide the means for pumping the sample back into the process.

July 1999

800-28

Chevron Corporation


Fig. 800-9


Bypass Loop to Reduce Response Time

Consider sample disposal systems when: •

The sample is too valuable to discard.

•

The sample is toxic or corrosive and too hazardous to discard in a chemical sewer system.

•

Discarding of hydrocarbon samples in a sewer system or to the atmosphere must be minimized for environmental reasons.

•

Discarding of hydrocarbons is prohibited.

•

The cost of discarding the sample is very high.

Every effort should be made to return hydrocarbon samples to the process or, if that is not practical, to the relief system. There are two forms of liquid sample disposal systems: one requires an electric motor driven pump; and the other, an all pneumatic system. Both systems have surge pots that are maintained at atmospheric pressure through a vented standpipe. The vent serves to eliminate back-pressure buildup on the analyzer outlet, and the surge pot makes is possible for the pumps to operate infrequently.

Electric Sample Disposal System Figure 800-10 illustrates an electric sample disposal system that can be purchased or built with a surge pot that vents to atmosphere through an explosion proof vent. Two level switches actuate the pump: the high-level switch starts the explosion-proof, motor-driven pump which continues operating until the low-level switch shuts it off. The pump remains inoperative until the level rises to actuate the high-level switch. The pump has sufficient head to return the sample to the process.

Chevron Corporation

800-29

July 1999



Fig. 800-10 Example of Electric Sample Disposal Systems

Sizing Sample Lines Design sample systems to provide the desired sample flow according to the following requirements: •

Adequate speed of response. Fast response may require higher flows with excess bypassing.

•

The monetary value of the sample.

Keep the maximum allowable time delay for normal installations under one minute. Normally, the sample lines are ½-inch; and, if a line of this size cannot provide the required flow, add a sample pump. To maintain the proper instrument response, multistream sequencing requires high, continuous flow in each individual line from the process lines being monitored. If such flows become excessive, they overload the sample-system components (filters and coalescers) which may result in an improperly conditioned sample that eventually could damage the analyzer. If the pressure differential in the sample system varies, the flow varies. If the pressure at the sample point is sufficient to overcome the pressure drop in the sampling system and provide the required sample flow, pressure control may be the only flow control necessary. Design sample-system components with residence volumes small enough to prevent both transport delays and capacitance averaging of sample concentration transients. Sample-system components with large volumes require more sample flowing through them to be flushed adequately. Even with considerable flushing. changes in composition are not readily transported to the analyzer due to the mixing effect in the large volume sample system components. Use minimal retention volume and residence time when designing sample systems. While you

July 1999

800-30

Chevron Corporation



can reduce residence time by maintaining sample system volume and increasing the flow, this can result in overloading the sample-system components. It is better to reduce the volume of the sample system by maintaining short sample lines and to optimize velocity by selecting tubing with small cross-sectional areas.

Lag Time When designing new sample systems or checking existing systems to determine proper flow conditions, use these methods and equations to obtain liquid and vapor samples. Sample transport lag time is the amount of time required to extract a representative sample from a process line, to transport the sample to the analyzer, and to condition the sample to be compatible with the analyzer. Do not confuse lag time with the analyzer response time, the time required by the analyzer to indicate 95 percent of the initial value when a step change is introduced to the analyzer inlet. The sample transport lag time and the analyzer response time determine the total delay that exists between an actual change in the process stream and the corresponding analyzer output. Analyzer response time is fixed by the choice of the analyzer. To optimize sample transport lag time. keep the volume of the sample system to a minimum without allowing excessive pressure drop. Usually sample flow through the analyzer is small. Normally, fast-flow bypass streams are preferred to transport the sample to the analyzer. In liquid application, calculate the sample transport lag time easily by dividing the internal volume of lines and components by the sample flow. Because liquid can be considered non-compressible, the sample volume does not change significantly with temperature or pressure; therefore, the lag time can be controlled by establishing the proper flow. For gas streams, the compressibility of gases makes it more difficult to calculate the total lag time. For a fixed quantity of gas, the volume varies directly with absolute temperature and inversely with absolute pressure. For a given sample flow, therefore, the actual velocity of the gas through any portion of the system depends on the temperature and pressure of the gas. Normally, a gas analyzer is operated at atmospheric pressure for stability; therefore, the simplest method for determining sample transport lag time is to relate the total gas volume in the system to standard conditions (STP). As a result. for various portions of the system, the volume at STP can be determined by means of the following equation: ( 14.7 + P ) ( 530 ) V STP = V STAT --------------------------------------( 14.7 ) ( 460 + T ) (Eq. 800-1)

VSTP = Gas volume adjusted to standard conditions VSTAT = Static volume (internal capacity)

Chevron Corporation

800-31

July 1999



P = Pressure in PSIG T = Temperature in °F. Under standard conditions, the total volume of the system is the amount of gas between the process line and the analyzer. The flow through the system (analyzer and bypass streams) can then be established to provide the desired lag time for the sample transport. Refer to Figure 800-11 as a sample problem. Due to non-laminar flow characteristics and the shape of various sample system components. it is necessary to add more sample system response time for complete volume turnover. For both liquids and gases, a rule of thumb is to calculate the time based on the system volume multiplied by three. The ultimate method is to run tests on the installed system itself.

Liquid and Vapor Sample Pressure Drop and Velocity The above description for gas streams does not take into account pressure drop through the system. For relatively low flow and short line length, the pressure drop should not be significant. At a high flow or long line length, however, the pressure drop may become an important factor in the system design. In addition, although the specific gravity of a gas or liquid does not affect the lag time calculation, it must be considered when specifying flow control and indicating devices. Determine the pressure drop using Figures 800-12 and 800-13.

Vaporized Liquids An objectionable transport lag is introduced if a sample to be vaporized remains liquid until the point where it enters the analyzer without a bypass circulating loop. For light hydrocarbons, the ratio of vapor volume to liquid volume is greater than 200:1. If a vapor flow of 2000 cc/min is maintained downstream of a vaporizer, a liquid flow of less than 10 cc/min flows upstream from the vaporizer. Every 10 cc of liquid holdup in the sample line creates one minute of transportation lag. Separation of as little as one foot can cause transportation lags of over five minutes. Vaporize the sample either very close to the sample tap or immediately downstream of a circulation loop, as shown in the typical gas chromatograph sample system in Appendix E.

835 Construction Materials Select construction materials to minimize corrosion. Analyzers and sample system components can be harmed by the following environmental factors:

July 1999

1.

Condensation that may occur in conduit that has not been sealed.

2.

Water that can collect in instrument enclosures.

3.

Rodents who can eat wire insulation.

800-32

Chevron Corporation



Fig. 800-11 Sample Lag Time Problem (Courtesy of Pastech Corporation) Sample Lag Time Problem

SOLUTION

A gas sample handling system has been designed to meet the requirements for a specific analyzer application. The following information has been established or calculated.

Step 1: Determine adjusted volume (STP) from bypass point in sample system to the analyzer. Volume in item 5 @ 30 psig = 5 cc x 3 Atm = 15 scc Volume in item 6 @ 15 psig = 5 cc x 1 Atm = 5 scc

1. Process Conditions a. Temperature: Ambient

Total Volume = 20 scc

b. Pressure: 210 psig (15 Atm)

Step 2: Determine lag time from bypass point in sample system to the analyzer.

Note:

Since the temperature effects are not significant, they will be neglected for purposes of simplicity

2. Sample Probe & Field Station (reduce to outlet pressure of 30 psig)

20 scc Lag Time = ------------------------ = 0.1m = 6s 200 scc/m

a. Sample Probe: 1’ of ¼" tubing, .049" wall, static volume = 3.6 cc

Allowable transport lag time from process line to the bypass point becomes 24 s (30 s - 6 s)

b. Block Valve, Filter, Regulator: Static volume = 6.4 cc

Step 3: Determine adjusted volume (STP) from process line to bypass point in sample system.

3. Transport Line a. Tubing: 3001 of ¼" tubing, .035" wall, static volume = 1500 cc

Volume in item 2 @ 210 psig = 10 cc x 15 Atm = 150 scc

4. Sample System to Bypass Stream a. Tubing: 5' of ¼" tubing, .035" wall, static volume = 25 cc

Volume in item 3 @ 30 psig = 1500 cc x 3Atm = 4500 scc

b. Filter: Static volume = 20 cc

Volume in item 4 @ 30 psig = 50 cc x 3 Atm = 150 scc

c. Valves, etc.: Static volume = 5 cc

Total Volume = 4800 scc

5. Bypass Stream to Analyzer Flowmeter

Step 4: Determine bypass flow for transport lag time of 24s from process line to bypass point in sample system.

a. Tubing: ½' of ¼" tubing, .035" wall, static volume = 2.5 cc b. Valve, Flowmeter, etc.: Static volume = 2.5 cc 6. Transport Line from Sample System to Analyzer

4800 scc 200 scc Bypass Flow = ---------------------- = ------------------- = 12,000 scc/m 24s s or approx. 24 SCFH

a. Tubing: 1' of ¼" tubing, .035" wall, static volume = 5 cc PROBLEM Given an analyzer sample flow of 200 scc/m, (standard cubic centimeters/minute) determine the required bypass flow (at STP) to give a maximum Sample Transport Lag Time of 30 seconds from the process line to the analyzer inlet.

Chevron Corporation

Please note that for a Sample Transport Lag Time of 60 seconds, the Bypass Flow would be approx. 11 SCFH. The above is a means for estimating the Sample Transport Lag Time. It is important to note the items which have a serious impact on the lag time as opposed to those having a negligible effect. Should more precise determinations be required (this would be very unusual), it would be necessary to use complex equations, obtain accurate component volume data and line measurements, confirm exact process information and, as a final step, perform dynamic testing on the system as installed.

800-33

July 1999



Fig. 800-12 Pressure Drop per 100 Feet of Tubing (Gases) Darcy Pressure Drop per 100 Feet (psi) Flow (ft/m)

1/4 Tubing

3/8 Tubing

1/2 Tubing

1

0.009

---

--

2

0.035

---

--

3

0.081 < 1.00

< 1.00

4

0.144

5

0.225

---

--

10

0.898

---

--

0.13 × f × p × v 2 Darcy Pressure Drop per 100 Ft. = ---------------------------------------d f = Friction Factor = 0.014 p = Density (C1-C4 Hydrocarbons) = 0.08 v = Velocity d = I.D. Tubing 1/4 O.D. × .035 Wall Tubing has an I.D. = 0.18 in. 3/8 O.D. × .035 Wall Tubing has an I.D. = 0.305 in. 1/2 O.D. × .035 Wall Tubing has an I.D. = 0.43 in. Q × 0.1079 Velocity (ft/sec) = ------------------------d2 Q = Flow (ft/m) d = I.D. Tubing (inches) Flow (ft/m)

1/4 Tubing

3/8 Tubing

1/2 Tubing

1

3.33

1.16

0.58

2

6.66

2.32

1.17

3

9.99

3.48

1.75

4

13.32

4.64

2.33

5

16.65

5.80

2.92

10

33.3

11.6

5.83

July 1999

800-34

Chevron Corporation



Fig. 800-13 Darcy Pressure Drop per 100 Feet (psi) Flow (ft/m)

1/4 Tubing

3/8 Tubing

1/2 Tubing

1

4.6

0.33

0.06

2

8.5

1.3

0.24

3

41.7

3.0

0.54

4

-

5.3

0.95

5

-

8.3

1.5

10

-

33.1

5.95

0.13 × f × p × v 2 Darcy Pressure Drop per 100 Ft. = ---------------------------------------d f = Friction Factor = 0.014 p = Density Liquid HC = 41.3 (as n-Hexane) v = Velocity d = I.D. Tubing 1/4 O.D. × .035 Wall Tubing has an I.D. = 0.18 in. 3/8 O.D. × .035 Wall Tubing has an I.D. = 0.305 in. 1/2 O.D. × .035 Wall Tubing has an I.D. = 0.43 in. Q × 0.1079 Velocity (ft/sec) = ------------------------d2 Q = Flow (ft/m) d = I.D. Tubing (inches)

Flow (ft/m)

1/4 Tubing

3/8 Tubing

1/2 Tubing

1

3.33

1.16

0.58

2

6.66

2.32

1.17

3

9.99

3.48

1.75

4

13.32

4.64

2.33

5

16.65

5.80

2.92

10

33.3

11.6

5.83

Chevron Corporation

800-35

July 1999



4.

Mud-daubers who can build nests in sample system components prior to use.

5.

Components that may be subjected to elevated temperatures where air purges have not been provided.

6.

Corrosive material that may leak into the analyzer or the sample system.

Take the following precautions to help minimize corrosion problems: 1.

Protect sample tubing from mechanical abuse by routing it properly or running the tubing in tray.

2.

Specify hypodermic-grade stainless or other alloy tubing and clean it thoroughly to remove oil film or purchase it cleaned. When selecting tubing, keep in mind that thin-walled material is affected significantly by external corrosion.

3.

Check all sample systems thoroughly for leaks.

4.

Inspect all socket or seal-weld connections on sample systems. These connections can leave small internal cracks or pockets in the pipe wall where contaminants may accumulate.

5.

Evaluate all sample system components for material that may absorb portions of the sample or react with the sample. Do not use cork or felt filter elements in service where the sample contains low moisture or solvents. Substitute ceramic filters.

Material Selection Follow the specifications in this manual and the enclosed list of noncorrosive parts. Take care when selecting nonmetallic materials for sampling systems because the sample composition might include components that attack nonmetallic materials. Two methods of attack of nonmetallic materials are (1) direct solubility (i.e., solubility of plastic in the sample material, as in acetone on plexiglass), and (2) solubility of the plasticizers or fillers used in plastics (i.e., failure of Kel-F in hydrocarbon solvent). Dissolving the plasticizer results in initial swelling then final disintegration of the base material. Reduce not only costs but also fire-and-health hazards by selecting material for the sampling system properly.

836 Sample System Components Piston Pumps Piston or plunger pumps are often selected for metering applications. At the suction stroke, the check valve on the inlet port opens and allows the chamber to fill with fluid. At the discharge stroke, the inlet check valve closes and the moving pistons force the fluid through the outlet port. Piston pumps are capable of a wide range of flows by varying the stroke length and/or drive speed. Some units are capable of generating pressures up to 30,000 psi.

July 1999

800-36

Chevron Corporation



Some disadvantages of piston pumps are as follows: • • • •

Piston pumps produce pulsating flow. Packings are subject to leakage and require periodic maintenance. Piston lubricant may contaminate the sample. Check valves are troublesome especially in dirty streams.

Diaphragm Pumps Diaphragm pumps need either a mechanical or a hydraulic coupling to transmit the plunger movement into diaphragm displacement. To vary flow, adjust the stroke length and stroke frequency. An advantage of the diaphragm pump is that there is no leakage past the piston and no contamination of the sample from pump lubricant. The discharge pressure is limited because of the stress on the elastomeric diaphragm. In mechanical linkage, the upper limit is about 300 psi whereas some hydraulic-type pumps can withstand up to 5,000 psi.

Bellows Pumps Bellows pumps move fluid by alternately extending and compressing flexible chambers and may be fabricated from stainless steel or from a variety of polymeric materials. To adjust the pump capacity over a wide range, alter the length of the stroke and stroke frequency. The advantages of bellows pumps are that they can isolate the process fluid from the pumping mechanism and can eliminate sliding seals. Discharge pressures are limited to 50 psi.

Gear Pumps Do not choose gear pumps for samples containing abrasives because the abrasive cause progressively increasing internal leakage or slippage. Seal leakage is the most common problem with this type of pump.

Centrifugal Pumps Centrifugal pumps move large volumes of sample at relatively low pressure and operate at comparatively high speeds. Choose centrifugal pumps with a magnetic linkage to avoid seal and contamination problems. Centrifugal pumps are not suitable for high-viscosity samples and are not self-priming.

Peristaltic Pumps In peristaltic pumps, fluid is forced along by waves of contraction produced mechanically on flexible tubing. The primary advantage of peristaltic pumps are their low cost, low sample holdup, and freedom from leakage and contamination.

Eductors Eductors have the advantage of no moving parts, low maintenance, and low cost. The motive fluid under pressure, discharges at high velocity through a nozzle and

Chevron Corporation

800-37

July 1999



entrains the suction fluid. The motive fluid may be air, the process stream, steam, or another convenient fluid. Eductors may be installed in sampling systems at the end of the sampling train, downstream from the analyzer, so that the sample is sucked through the analyzer without being contaminated by the motive fluid. Alternatively, the motive fluid may be mixed with sample and separated in the sample conditioning train. Appendix E shows a sample system with eductors for extracting oxygen and carbon monoxide from stack gas.

Flow Indicators Rotameters, the most widely used flow indicators, are variable orifices consisting of a tapered tube and float which are available armored or in glass. Usually the rotameter manufacturer supplies sizing information for water and air. Calculate the approximate conversions to other fluids with the equations available in vendor literature. Sometimes, a sight-flow bubbler can provide continuous regulation of gas flow if you set the adjustment to very small flow rates and time the bubbles through the liquid. Fifty psi is the maximum pressure that can be applied to a bubbler with a pyrex bowl and 100 psi with a plastic bowl.

Heat Exchangers In some cases, the process stream temperatures may be above the temperature limit of the analyzer, and heat must be removed from the sample. Sometimes heat loss through the walls of the sample line provides adequate cooling, but often a sample cooler is required. The heat exchange media can be water, air, freon, steam, or other fluid. Mechanical refrigerators are acceptable for low temperature applications. Vortex coolers are useful for cooling small flows but consume large quantities of air. Gassample coolers may result in the formation of a second phase that requires a separator to remove the condensed material. A trap is necessary for removing large amounts; coalescers, for small amounts.

Cooling and Heating The following guidelines help determine if liquid samples require cooling: 1.

When separation is required between condensibles and noncondensibles.

2.

For analysis of steam samples.

3.

When taking samples for initial boiling point, freeze cloud, and flash point.

4.

To prevent flashing in sample valves.

Cool other samples if the sample temperature exceeds the temperature compensation range of the analyzer or if the temperature exceeds the capabilities of internal heat exchangers. The most common cooling medium is cooling water.

July 1999

800-38

Chevron Corporation



If cooling water is not available, is too expensive, or its temperature is not low enough, take cold air from a vortex tube or a mechanical refrigerator. Some samples, such as long and short residue streams, may require cooling to just above 212°F as they may solidify at lower temperatures. In such cases, the cooling medium may be hot condensate, the temperature of which is affected by controlling the pressure of the steam in the steam condenser. Heating may be required to prevent condensation of gaseous samples or to lower viscosity of liquid samples. Ensure that the conditioned temperatures are maintained up to the analyzer. Gas samples approach the temperature of their container rapidly. Heat tracing helps to avoid plugging and corrosion from condensation of water in gas samples.

Separators Removal of Entrained Liquids in Liquids. By gravity difference, coalesce and further separate liquid entrainment (i.e. water in oil or opposite) and also separate entrained gases. To prevent fouling of the coalescer packing, filter it first. Remove solids by filtration. Dispose of the separated entrainments into the return line of the circulation loop. For single-line systems, use an automatic drain or gas vent. Coalescers are effective for removing dispersed water from gas. Most process streams contain small amounts of water which may have to be removed before the sample reaches the analyzer. A coalescer contains a fine pore filter element that is wetted by hydrocarbons and not water. Finely dispersed water agglomerates into large droplets that fall out to the bottom of the coalescer. The large internal volume of commercial coalescers require either large flows or sample by-pass flows to avoid long lag times. The sample leaving the coalescer is saturated and no further cooling should occur in the sample system. Note Adhere to the flow rate recommendations provided by the coalescer manufacturer. If the flow rate is too high, the coalesced material may break up into small particles that remain in floating condition and prevent separation by gravity differences. Removal of Entrained Liquids in Gases. By selecting the sample tap location properly, you minimize the carryover of liquid or solid entrainment. Obtain further separation in the sampling system by: (1) installing cyclone filters when the sample velocity is adequate; (2) coalescing droplets for liquid entrainment and following with gravity separation; (3) filtering solids. Removal of Entrained Gases in Gas Streams. Remove polar compounds, such as NH3 and CO in stack gases by scrubbing with a water cooler or water eductor. The mixed phase enters the separator from the side or from the top. The abrupt change in direction and the reduced velocity separates the higher-density, entrained

Chevron Corporation

800-39

July 1999



material from the lower-density fluid. Heavy material is drained off periodically either manually or automatically. Centrifugal Separators. A cyclone separator consists of an inverted cone with a tangential inlet near the top. Overflow connections are located at the top and center of the cone; the drain connection, at the bottom. The efficiency of cyclones depends on the velocity of the sample and the difference in density between the particles and the fluid. The overflow is normally maintained at 50 percent of the total flow. The underflow is returned continuously to the sample line.

Filters Most process streams contain small amounts of solid or semisolid material in the form of scale, rust, and catalyst fines. Remove at least part of this material to prevent plugging lines and to avoid damaging the analyzer. Mechanical filtration is available in a variety of materials and pore sizes. To maximize analyzer performance and minimize sample system maintenance, do not overfilter. While overfiltering is not harmful, particle removal has an effect on the eventual plugging of filters and increased maintenance. It is important to know the requirements of the particular installed analyzer. For example, samples for a pH meter may require no more than a coarse wire-mesh screen filter. Gas chromatograph samples, however, must be free from abrasive material down to 0.3 microns to protect delicate sample valves. For this application, select bypass filters whenever possible as their self-cleaning capability allows for much longer periods between replacing and cleaning elements than other types of filters. Sample systems that need fast-flow circulation loops can take advantage of the bypass configuration to provide the required, small, slip stream to the analyzer. Be sure to provide sufficient flow rate in the bypass stream to keep the filter clean or install two filters in parallel as an acceptable alternative when it is too costly to shut down the system for tiller cleaning or essential maintenance. Install pressure gages at either end of the manifold to indicate when filters are plugged. Size the filter housing properly to maintain a low sample-system volume for decreased lag time.

Valves Valves can be categorized according to their intended use, such as the following: shutoff valves, throttling valves, and remotely operated valves. Shutoff Valves. Shutoff, or on/off valves can be subdivided into gate, ball, plug, diaphragm, and specialty valves, such as rotary or slider valves, common in gas chromatographs. Sometimes, small gate valves are selected for throttling; however, large sizes vibrate excessively in the partially open state.

July 1999

800-40

Chevron Corporation



The diaphragm valve provides a sample that is free of contamination from valvestem lubricant or packing. The sealed construction of this valve makes it useful for highly corrosive streams. When the valve is in the closed position, ball valves trap liquid in the ball cavity which is undesirable as the trapped liquid may contaminate the next sample. Throttling Valves. Most throttling or flow-adjust valves are variations of the basic globe valve. The globe may be a disc, ball, plug, vee, or needle configuration. For small flows, as in analyzer sample systems, select the miniature vee valve. For precision metering, choose a tapered needle valve. Do not use valves of this type for shutoff, however, because of the possible damage to the stem and potential for leakage. Remotely Operated Valves. The most common remotely operated valves in sample systems are pneumatic valves or electrically operated solenoid valves. Solenoid valves are generally limited to a maximum pressure drop of 200 psig. Special coils are available for temperatures as high as 350°F. Multiport valves are useful for multistream and calibration arrangements. Air-operated valves are often chosen for high-temperature applications as their performance for these applications can be significantly better than solenoid valves.

837 Process Analyzer Sample Systems Introduction. The process analyzer sample system designs presented in this section are intended as guides for engineers and designers who are involved in the installation of process analyzers. All but one of the designs are based on actual installations in various parts of the Company. They are not meant to be standards, as sample systems are often enhanced in the field not only by experienced maintenance and plant operation personnel but also as improved components are introduced. No sample system is trouble-free, and scheduled preventive maintenance is essential to attain a high analyzer service factor. The basic requirement of a sample system is to extract a representative sample from a process line under process conditions of temperature, pressure and particle load and to present it, in a short time and in a suitable condition to an on-line analyzer. Initially, compromise is essential to make most important decisions in the installation: the location of the sample take-off and of the analyzer shelter for short sample lines. This is due to competition from other requirements of a process plant. Therefore, it is necessary to plan the analyzer installations early in a project and to involve other engineers and designers in the planning process. Sample system design is usually left to an analyzer systems contractor. The design examples that are given in this section will be useful for evaluation of the contractor’s work.

Chevron Corporation

800-41

July 1999



Chevron Research and Technology, M&CS Unit should be used as a resource and support in all phases of the process analyzer project when there is not a site process analyzer engineer. Design Notes. Earlier 800 sections of this manual covered many aspects of sample system design. Two basic requirements are mentioned here: In almost all cases, a sample probe in the process line should be used for extracting an analyzer sample since a probe provides initial inertial filtering, and takes a more representative sample at a distance from the pipe wall. A fast sample loop is a basic feature of most sample systems. This fast flowing line brings the sample to the analyzer, thereby reducing sample lag time. Also, only the sample needed by the analyzer is fully conditioned by filters, regulators or heat exchangers, thereby saving maintenance. Analyzer Integrity. When a process analyzer is in a control loop, or it is accumulating data for regulatory purposes, it is essential to know if the analyzer output is believable. Various checks have been devised to indicate whether or not this is true. These checks vary to some extent from one type of analyzer to another. The following are some examples: • • • • •

Low sample pressure in the fast loop. Low sample flow to the analyzer. Analyzer calibration off by more than a set percentage. Analyzer signal large step change. Analyzer being calibrated or repaired.

In the sample system designs that follow, the only integrity checks shown are flow switches, but the need for other integrity checks should always be reviewed.

Stack Gas Sample System. Nitrogen Oxides and Continuous Emission Monitoring Analyzers The sample systems for continuous emissions monitors (CEMS - CO, Oxygen and NOx) and for nitrogen oxide analyzers alone, are very similar. However, the regulations which control the monitoring of NOx emissions alone, apply to all existing furnaces, some of them fairly small. Consequently, the requirement that calibration gases should be introduced at the sample probe on the stack does not apply. A bypass filter, item 6, at the sample branch to the analyzer from the fast loop may sometimes be omitted, depending on the size of the filter at the end of the probe in the stack. A membrane filter which holds back liquid water, has been added downstream of the refrigerated sample drier, as a back-up device. This filter can also be installed ahead of the diaphragm pump in case the sample line temperature drops below the stack gas dew point.

July 1999

800-42

Chevron Corporation



The following sample system components are shown in Figure 800-14. These components are suitable for areas with a general purpose electrical classification. Fig. 800-14 Stack Gas Sample System: For NOx & CEMS Analyzers

1. Sample Probe in Stack. This 1" or 1.5" stainless steel probe provides inertial filtering by changing direction of flow of the sample gas. For air quality control an external filter is recommended - the filter being accessible from the outside of the stack. (Fabricated by contractor) 2. On-Stack Sample Control Box. This box is required to be near the sample probe so that calibration gases and blowback air for the probe can flow through the complete sample system. The solenoid valves in the box are remotely controlled from a switching device such as a PLC in the analyzer house. The box is heated and the temperature is controlled above the dew point of the stack gases. (Fabricated by contractor) 3. Heated Sample Line. This electrically traced tubing bundle maintains the sample at a temperature above the dew point to prevent condensation in the line. This bundle should be of the type that uses constant wattage with a temperature controller. Condensed liquid might absorb some of the desired components and/or freeze in the line. (Dekoron, Parker Hannifin, or O’Brien)

Chevron Corporation

800-43

July 1999



4. Air Blow-back Valve. This air-operated, 3-way ball valve is remotely operated during the calibration cycle. (Whitey) 5. Fast Loop Air Eductor. Since the stack gas is below atmospheric pressure, a pump is required to extract the sample. Eductors or “jet pumps” are frequently used since they are practically maintenance free. A fast flowing sample (typically 5 ft/sec) is brought right down to the analyzer house to reduce sample lag time. A diaphragm pump may also be used. (½" SS jet pump with an air requirement 50 psig is fabricated by Penberthy, Fox.) 6. Sample By-pass Filter. A filter (10 micron nominal) may be required here for the sample branch to the analyzer, depending on the probe filter size. (PermaPure, PAI, Fluid Data) 7. Pressure Gauge. A local pressure gauge (0-30 psig nominal) in the sample line to the analyzer is a valuable indicator especially for maintenance purposes. (Ashcroft) 8. Diaphragm Sample Pump with Bypass. This is necessary to provide a positive sample pressure for the analyzer(s). Moisture drop-out ahead of the pump reduces the life of the diaphragm. The sample line inside the box should be traced and insulated right up to the pump. See component 11 below on using a membrane filter. (ADI) 9. Relief Valve. A relief valve around the pump protects the pump against a blockage downstream, whatever the cause - plugging of a filter, malfunction of a regulator, or a valve closure. (Nupro) 10. Sample Cooler/Drier. This component should be designed for minimal contact between the sample and the condensible materials to reduce the chance of gases dissolving into the liquid. There is evidence that some NO2 is lost in the coolers. (Refrigerated Sample Cooler/Drier for 5°C dew point are manufactured by M & C, Universal Analyzers, Hankison.) 11. Membrane Filter. This filter with ¼" ports will prevent liquid water or oil from passing and is recommended as a back-up device for the cooler/drier. (A+ Company) 12. Pressure Regulator and Gauge. Sample conditions must be carefully controlled to attain the sensitivity and stability of the analyzers in this application. The regulator has ¼" connections and a range of 0-10 psig. (Go, Fairchild) 13. Valving for Analyzer Zero and Span Gases. Calibration checks are carried out automatically every 24 hours using certified cylinder gases. The output signals of the analyzers are automatically adjusted by auto zero and span units. These functions are controlled by a dedicated PLC or other device in the analyzer house. The calibration gases are connected to the system through valves in a block & bleed arrangement with a bubbler leak indicator. Some regulations require the calibration gases to be introduced at the sample probe, see component 2 above. (The 3-way, multipurpose solenoid valves are manufactured by Skinner, Asco.)

July 1999

800-44

Chevron Corporation



14. Sample Flow Indicator. To maintain correct calibration, the flow rates of the sample and calibration gases must be the same, at the same analyzer outlet pressure. The flowmeter should have ¼" connections and a range of 0.2 to 1.2 scfh. (Brooks, Wallace and Tiernan) 15. Sample Flow Switch. The sample flow is one of the items monitored to check “Analyzer Integrity” described in the introduction of this sub-section. (Fluid Component Inc.) 16. Final Sample Filter. This is the final “polishing” filter before the analyzer. It is packaged as a tee-type filter with a two micron element. (Nupro) 17. Nitrogen Dioxide (NO2) to Nitrous Oxide (NO) Converter. This converter is necessary if a total NOx analysis is required, which is normal, since the Chemiluminescent. The most efficient and stable converters operate at a temperature of about 800°F. Low temperature converters (about 350°F) using a molybdenum catalyst are more likely to lose efficiency, and will convert a small amount of ammonia to NO, in spite of the manufacturers’ claims. (High Temperature converter, built into analyzer by analyzer manufacturer. Rosemount.) 18. Sample System Box. Experience gained in cogeneration units indicates that the sample diaphragm pump needs more protection from condensation. The sample system box should be controlled at a temperature of about 180°F and the sample line inside the heated box should be traced right up to the pump. This ensures that the pump has some protection against condensation when the box is opened. 19. Analyzer. The following analyzers are common for these applications: • • •

Oxygen: Paramagnetic Carbon Monoxide: Infra-Red, with CO2 and H2O compensation NOx: Chemiluminescent

Chromatograph sample system: FCC Stack Gas with Heavy Particle Loading This sample system is similar to the next chromatograph sample system - gas sample under pressure returned to header - with the following special features: •

There are two sample fast loops with two bypass filters to clean the sample. A cyclone filter is in the first loop, a swirlklean filter is in the second loop.

•

The stack gas is below atmospheric pressure so that an eductor is needed to extract the sample.

•

The sample flowing through the chromatograph sample valve can exhausted to atmosphere. However, the fast loops are returned to the stack or duct.

The following sample components are shown in Figure 800-15. 1. Sample Probe. A plain probe with heated tubing take-off from the top of a tee fitting at the end of the probe. The end connection is plugged so that it can be used for cleaning the probe.

Chevron Corporation

800-45

July 1999



Fig. 800-15 Chromatograph Sample System: FCC Stack Gas with Heavy Particle Loading

2. Sample Tubing. An electrically traced, self-limiting, ½" tubing bundle keeps the sample above its dew point, to prevent condensation and possible freezing in the line. (Parker Hannifin, Dekoron, O’Brien) 3. Cyclone Filter. This SS filter removes most particles larger than 10 microns without an element. This type of filter is described in more detail in the component notes on the Extended Natural Gas sample system described later in this section. (Mooney Analytical, PAI) 4. Flow Indicator. The flow through the filter must be maintained at the specified rate to ensure maximum particle removal. (Brooks) 5. Swirlkleen Bypass Filter. A second sample loop with a Swirlkleen bypass filter provides further filtering. This filter is also described in the component notes for the Extended Natural Gas Sample System later in this section. (Collins) 6. Flow Indicator. A second sample loop flow indicator has the same function as component 4 above since the filter requires a continuous bypass flow to sweep the particles along with the return sample to the stack. (Brooks, Wallace and Tiernan) 7. Sample Diaphragm Pump. This pump drives the sample through the chromatograph. To protect the diaphragm it is advisable to extend the electrical tracer, heating the line right up to the pump, to prevent condensation in the line when the heated enclosure is opened. (ADI, Universal Analyzers)

July 1999

800-46

Chevron Corporation



8. Pressure Gauge. The sample pressure must be sufficient to drive the sample through the chromatograph sample valve. At this point a pressure of about 20 psig is adequate. The gauge should have a range of 0-30 psig on a 4½" dial. (Ashcroft) Sample components 9 through 15 are the same as in Figure 800-16 below. Fig. 800-16 Chromatograph Sample System: Gas Sample Under Pressure Returned to Header

16. Heated Enclosure. This enclosure is necessary to keep the components in the sample system above the dewpoint. For a stack gas the temperature could be approximately 130 deg F. The column in the chromatograph should be chosen to accommodate the water vapor. (Hoffman)

Chromatograph Sample System: Gas Sample Under Pressure Returned to Header The following sample components are shown in Figure 800-16. 1. Sample Probe with Gate Valve. This full port valve has a packing gland for probe extraction. A probe is always the preferred method of sampling for analyzers, due to flow direction filtering, and more representative sample. (Integrated by Contractor with gate valve for extraction, where piping code allows.) 2. Sample Block Valve. This valve at the process line is usually covered by the piping code, and is not selected by the sample system designer. This valve is often a ¾" gate valve and is installed by field. 3. Filter. A Y-strainer is recommended upstream of the pressure regulator. (Yarway)

Chevron Corporation

800-47

July 1999



4. Pressure Regulator. A differential pressure of about 20 to 25 psig will generally be sufficient to provide gas sample flow through the system. A suitable sample return point for the fast sample loop is necessary. (Moore) 5. Sample Line. A prefabricated line is preferred since the sample must be kept safely above the dewpoint temperature of the gas. All expected ambient temperatures and process conditions should be considered - including start-up and upsets. (For light, dry gas streams this is not necessary.) (Electrically traced, self limiting tubing bundle with ¼" SS lines are made by Dekoron, Parker Hannifin, O’Brien). 6. Filter. The fast sample loop is taken to the analyzer house to reduce sample lag time. A Swirlkleen filter is required at the analyzer sample take-off from the fast loop. (Collins) 7. Fast Sample Loop. This return line must be connected to a low pressure point in the process so that adequate differential pressure is available for the sample system. The line back to the process may need to be heated to prevent condensation which could cause plugging. 8. Pressure Gauge. Local indication of sample pressure is valuable for routine checking and maintenance purposes. The pressure gauge should have a range of 0–25 psig. (Ashcroft) 9. Final Filter. An in-line filter with a maximum 2 micron element is required to safeguard the chromatograph column. (Nupro) 10. Bypass Flow Indicator. This bypass takes the excess gas not flowing through the chromatograph sample valve. It also allows the flow to continue when the sample is blocked and vented a few seconds before the sample valve injects the measured volume into the carrier gas, allowing the gas to come to atmospheric pressure. (Brooks, Wallace and Tiernan) 11. Sample Block and Vent Valves. These are air operated, ¼", SS ball valves which are controlled by the chromatograph cycle program. As stated above, they block and vent the sample flowing through the sample valve in the chromatograph, to ensure that the measured volume in the loop is at atmospheric pressure. (Whitey) 12. Flow Indicator. The flow through the sample valve is continuous except when it is blocked for a few seconds at the beginning of each measuring cycle. See component 14 below. (Brooks, Wallace and Tiernan) 13. Differential Pressure Regulator. This component is necessary due to vent header pressure variations. A differential pressure regulator can be used here since the effect of sample pressure variations is eliminated by component 11 above. (Moore) 14. Sample Valve. This valve is an integral part of the chromatograph. It directs a continuous flow of gas through a fixed volume sample loop. At the beginning of an analysis cycle, the valve operates and the carrier gas sweeps the volume of sample into the column where it separates into components.

July 1999

800-48

Chevron Corporation



15. Block and Bleed Valves. If calibration is controlled automatically, which is normally the case, the calibration gases are introduced through a block and bleed solenoid valve arrangement. See also the discussion on Figure 800-27 later in this sub-section.

Chromatograph Sample System: Liquid Sample with Liquid Injection Valve There are many cases where the sample is liquid under process conditions but it can be analyzed as a vapor in a chromatograph column. Normally, it is better to transport the sample as a liquid under pressure and inject it directly into the carrier gas, where it vaporizes at the temperature of the chromatograph column. The following sample components are shown in Figure 800-17. Fig. 800-17 Chromatograph Sample System: Liquid Sample with Liquid Injection Valve

1. Sample Probe with Gate Valve. This full port valve has a packing gland for extraction. It is used where piping code allows. (Integrated by contractor or chromatograph manufacturer). 2. Sample Block Valve. This ½" SS ball valve is used for isolating the sample system from the process line. (Whitey) 3. Filter. This Y-strainer with 100 mesh screen is recommended before the pressure regulator. (Yarway) 4. Pressure Regulator. It is always advisable to regulate the sample pressure because of the level and variations in the process pressure. Also the differential

Chevron Corporation

800-49

July 1999



pressure between take-off and return should be sufficient to drive the sample through the sample system and return the sample to a lower pressure point in the process - preferably across a pump. Range can be 20-25 psig differential. (Moore, Go) 5. Bypass Filter. The fast sample loop is taken to the analyzer house to reduce sample lag time, where the sample line to the analyzer is filtered by a Swirlkleen bypass filter. (Collins) 6. Flowmeter. An armored flowmeter is required in the fast flow loop for safety reasons for liquid hydrocarbons under pressure. (Brooks, Wallace and Tiernan) 7. Return Line Root Valve. The analyzer sample system must be protected from the downstream process pressure even though the sample take-off valve is closed. The root valve is normally a ¾" gate valve constructed to line specifications. 8. Sample Block Valve. This ¼" SS ball valve is necessary to have at the analyzer house for maintenance purposes. (Nupro) 9. Bypass Flowmeter. Sample bypass at the analyzer takes excess flow not required by the sample valve and keeps the sample flowing during calibration to reduce lag time. (Brooks) 10. Check Valves. These ¼" SS valves prevent circulation of the sample around the loop. (Nupro) 11. Final Filter. This ¼" SS filter protects the liquid injection valve and the chromatograph column. (Nupro) 12. Flowmeter. This flowmeter (0–20 ml/min) controls the flow through the liquid sample valve. (Brooks) 13. Liquid Sample Valve. This valve injects a measured volume of the sample by means of a grooved piston directly into the carrier gas where it vaporizes and flows into the chromatograph column. (ABB, MAT Valve) 14 & 15. Back Pressure Regulator and Gauge. It is necessary to keep the liquid in the system above the bubble point at the prevailing temperatures. The regulator setting also depends on the pressure required to return the sample to the process line. (Moore) 16. Standard Injection - This system is used when it is possible to calibrate with a standard liquid mixture kept under pressure during storage and when sample is injected by a liquid valve. The cylinder containing the standard has two sections connected by a piston, with nitrogen under pressure on one side, the standard on the other. A block and bleed valve arrangement can be used as shown. (The 3-way, air operated, ¼" ball valves are made by Whitey.)

Chromatograph Sample System: Liquid Sample with Local Vaporizing Regulator A liquid stream to be analyzed by a chromatograph may contain heavy components that would damage the chromatograph column or it may be necessary only to

July 1999

800-50

Chevron Corporation



analyze the lighter components in the stream. In these cases, a vaporizer regulator is located ahead of the chromatograph, and the chromatograph column is protected from any liquid that has not been vaporized by a liquid knock out pot. The following sample components are shown in Figure 800-18. Fig. 800-18 Chromatograph Sample System: Liquid Sample with Vaporizing Regulator

1 & 2. Sample Probe with Gate Valve. (See appropriate notes in previous sample systems.) 3. Sample Shut Off Valves. These ½" SS ball valves isolate the sample system for maintenance. It is important to protect the system from back pressure from the fast sample loop return point. (Nupro) 4. Heat Exchanger. This may be necessary to reduce the chance of the liquid vaporizing in the sample system ahead of the vaporizer. The sample should be transported at low temperatures, to allow the pressure to be reduced without the risk of vaporizing. The bubble points under various pressures should be estimated. (Princo 2" Dial Thermometer/Sentry Cooler) 5 & 7. Fast Sample Loop Flow Indicator and Valve. See sample system design notes at the beginning of this section. This 0–2 gpm armored flowmeter provides control of the flow in the fast loop at the analyzer house. (Brooks) 6. Primary Filter. This Y-strainer has a 10 micron SS element. (Yarway)

Chevron Corporation

800-51

July 1999



8. Secondary Sample Loop and Filter. This loop and filter bypasses excess liquid not flowing through the vaporizer regulator and keeps the sample updated close to the analyzer. (Collins) 9. Flow Indicator and Valve. This armored rotameter provides control for the secondary sample loop. (Brooks) 10. Vaporizing Regulator. The liquid sample is vaporized in a heated block at the entrance to the regulator. The vaporized sample pressure is regulated. 11. Vapor Sample Pressure Gauge. This 0–20 psig gauge indicates the pressure driving the sample through the rest of the sample system. (Ashcroft) 12. Relief Valve. This relief (set at 10 psig) to the vent header protects the system against a blockage downstream or malfunction of the regulator. (Nupro) 13. Liquid Knock-out Pot. This pot drains off heavy components in the sample that have not been vaporized. (Integrated by Contractor) 14. Flow Indicator and Valve. This flowmeter (0–1 L/min) and valve control flow of vapor to the sample valve. (Brooks, Wallace and Tiernan) 15. Sample Block and Vent Valves. These ¼", SS, air-operated ball valves bring the sample to atmospheric pressure a few seconds before it is injected into the carrier gas. The valves are operated by the chromatograph cycle program. (Whitey) 16 & 17. Block and Bleed Valves. This system is used when the calibration gas is introduced automatically by the chromatograph, which is the normal case. (Skinner, Asco) 18. Differential Pressure Regulator. Most analyzer samples must be vented to a closed vent header, not to the atmosphere. A back pressure or differential pressure regulator is needed because of the variations in pressure in the header. The pressure in the sample valve does not affect the calibration of the chromatograph, due to the operation of component 15 above. (Moore) 19. The Bypass Loop takes care of excess flow when the sample block operates or when the chromatograph is in calibration. 20. Heated Enclosure. The sample system components from the vaporizer downstream must be in a heated enclosure, to prevent liquid dropping out in the lines. The controlled temperature depends on the dew point of the vapor.

Chromatograph Sample System: Extended Analysis of Natural Gas This design includes probe and bypass filter options in the sample system for a continuous extended analysis of untreated natural gas by capillary chromatograph. The term “extended” means that all the heavier components in the stream are included in the analysis. Since some of these may be present in the line as liquids, the extraction of a representative sample is more difficult. The following sample components are shown in Figure 800-19.

July 1999

800-52

Chevron Corporation



Fig. 800-19 Chromatograph Sample System: Extended Analysis of Natural Gas Analysis

Probe Options. In each probe option, a static mixer is recommended ahead of the probe in the line to improve the chance of extracting all components. •

Option 1. Here a heated sidestream is intended to vaporize all the components and to allow the flow to become laminar, so that an isokinetic probe will be effective. The sample probe internal diameter must be large enough to minimize any selective effect. The isokinetic probe receives sample at the same flow rate as the gas flow rate in the line, avoiding composition changes which might be caused by inertial effects.

•

Option 2. A simpler installation would have a regular ASTM type sample probe or an isokinetic probe immediately following the static mixer without the heated sidestream.

•

Option 3. To compare with the above, a plain probe with square end opening could be used in the line.

ByPass Filter Options. The following bypass filter options can be used: •

Chevron Corporation

4. Cyclone Filter. In this filter the sample stream enters at the large diameter of a conical chamber. Particles move down the walls of the chamber, gaining speed as the flowrate increases, and are swept out with the discharge. The clean

800-53

July 1999



sample is taken from the center at the top of the chamber. The filter is effective for removing particles above 10 micron at 50 psi differential between the inlet and the clean sample. No element is needed, thus it is very effective as a primary element in sample conditioning. •

5. In-line Filter. This filter is a straight-through system, in which a cylindrical sintered metal element is held between O-rings. The sweepstream flows through the inside of the cylindrical element, and the filtered sample is taken off at a side connection. Elements are available down to a pore size of 0.5 microns. The whole of the element is theoretically in the filtration path, but the area close to the outlet plugs more quickly and increases the pressure drop across the filter.

•

6. Swirlkleen Filter. This filter combines both an element and a centrifugal filter effect. The sample enters one side of the circular wall of a cylindrical chamber and exits the other. This action sweeps the particles through as they are forced to the walls under centrifugal action. The filtered sample is taken from both ends of the chamber through sintered metal filter elements. Pore sizes may be down to 0.5 microns. (Collins, Rosemount)

•

7. ByPass Filter/Coalescer. This filter contains a filter element, either woven or sintered metal, which is screwed on to the top cap of the filter assembly, surrounding the filtered sample connection. The sweepstream is around the outside of the element, from top to bottom. A continuous flow is maintained from the bottom, in bypass service, or the connection can be valved or plugged for in-line service. The volume hold-up may be a problem in some applications.

8. Pressure Regulator. This ¼", SS, metal diaphragm regulator is located close to the line to reduce the pressure rating of the sample system components and to reduce the gas held up in the sample lines, both safety requirements. The pressure in the sample system must be kept at a minimum - just sufficient to keep the sample flowing, to avoid raising the dew point. The regulator range will depend on the application. (GO) 9. Sample Line. The electrically traced, Teflon tubing bundle keeps the sample above the dewpoint to prevent condensation and possible freezing in the line. (Parker Hannifin, Dekoron, O’Brien) 10. Valves. These valves are provided for a sample bomb connection if a manual sample is required. 11. Heated Enclosure. This enclosure keeps all sample components above the sample dew point.

Supercritical Chromatograph Sample System A supercritical chromatograph uses carbon dioxide above its critical point as a carrier. Thus it is capable of analyzing very heavy hydrocarbon samples which are soluble in the CO2. The most important feature of the sample handling system is that it maintains the sample at a temperature high enough for it to flow freely through the lines.

July 1999

800-54

Chevron Corporation



The sample system shown in Figure 800-20 is the section next to the analyzer, and does not include any components near the process line. Fig. 800-20 Supercritical Chromatograph Sample System

1. Valves. Local sample shut-off and bypass valves allow the flow to be maintained in the sample loop during analyzer shut downs for maintenance, etc. (Hoke) 2. Bypass Filter. The sample fast loop inside the sample box flows continuously to keep the lag time to a minimum. The connection for the sample to the chromatograph is taken through a Swirlkleen bypass filter. See component notes for Figure 800-16 for more details on this type of filter. (Collins) 3. Local Pressure Gauge. Range 0–100 psig, ¼" connections. (Ashcroft) 4. Armored Flowmeter. Bypass flow is maintained about 2.5 gpm on this indicator. Full scale range is 0–5 gpm. (Wallace & Tiernan) 5. Bypass Flowswitch. This 316 SS flow switch activates an alarm when flow goes below 1 gpm. (ChemTec) 6. Bypass Flow Control Valve. ½" SS needle valve (Hoke) 7. Double Block and Bleed Valves. Standard liquid is introduced through these for calibration of the chromatograph. (The notes for Figure 800-27 describe these valve arrangements more fully.) (Whitey)

Chevron Corporation

800-55

July 1999



8. Flow Indicator. This armored flowmeter with a range 0–6 gph indicates the sample flow through the analyzer. It is located on the sample outlet line from the analyzer. (Wallace & Tiernan) 9. Sample Flowswitch. This flow switch has a range of 0–6 gph and activates an alarm on low sample flow (set at 3 gph). (Fluid Component Inc.) 10. Sample Flow Control Valve. ¼" SS needle valve. (Hoke) 11. Block and Bleed Valve. This second valve arrangement returns the process sample to line or the standard used for calibration to the liquid standard supply. (See component 7 above). 12 & 13. Valves. These single, air-operated, 4-way ball valves are located in the standard liquid introduction system. They switch the stream from the sample to the standard. (Whitey) 14. Sample System Component Box. This insulated steel box is heated with a steam coil and temperature controlled by a thermostat for 210° F. A window in the box allows the two flowmeters to be seen from the outside. (The sample system box with a subpanel for components is made by Hoffman. The temperature indicator, Range 50°–400° F, 2" dial thermometer is made by Ashcroft).

Recycle Gas Moisture Analyzer Sample System See reference cited in RAARD for design.

Tanker Vapor Recovery Sample System This sample system uses an infrared gas analyzer which is a back-up for a control system in which natural gas is mixed with vapors recovered from the tankers during loading operations. The control system is designed to keep the vapor in the line well above the explosive limit, and is based on flow ratio. The analyzer is sensitized to methane. The following sample components are shown in Figure 800-21. These components are suitable for an area electrical classification of Class 1, Gp D, Division 1. 1. Sample Probe in Line with Gate Valve. This full port valve has a packing gland for probe extraction. (Integrated by Contractor) 2. Sample Probe Block Valve. This 3/8" SS ball valve is used for sample cut-off at the line. (Whitey) 3. Dual Head Diaphragm Pump. This sample pump is located near the line to reduce the length of the pump suction line and to provide positive pressure in the sample line. (Air Dimension) Note that all sample lines outside the sample box are electrically traced, selflimiting sample lines to ensure that there is no liquid drop-out in the system. 4. Sample Fast Loop Back Pressure Regulator. This 0–25 psig regulator is mounted in the sample component heated enclosure with pressure gauge. (GO)

July 1999

800-56

Chevron Corporation



Fig. 800-21 Tanker Vapor Recovery Sample System

5. Fast Sample Loop Flow Indicator. This flowmeter and valve has ¼" connections and a range of 0–15 scfh. (Brooks) 6 & 9. In-line Filters. These filters have a ¼" 0.5 micron element. (Nupro) 7. Block and Bleed Valves. This valve arrangement protects against leaky valves in the zero and span gas system, and also ensures that the lines are swept with the selected gas, avoiding cross-contamination of one gas with another. The valves are air-operated, 3-way, 1/8", SS ball valves. (Whitey) 7A. Flowmeter. Any leaks in the test gas valves will be shown by this flow indicator. The flow indicator range is 0–2 scfh. (Brooks) 8. Sample Flowmeter and Valve. This 0–2 scfh flowmeter and valve are used to control the sample flow to the analyzer. (Brooks) 9. Sample Filter. This in-line SS filter has 1/8" connections, a 0.5 micron element, and provides final filtering before the analyzer. (Nupro) 10. Flow Switch. This 1/8", 10 L/min, explosion-proof switch is part of the analyzer integrity check discussed in the introduction to this sub-section. (Autoflow)

Chevron Corporation

800-57

July 1999



11 & 12. Back Pressure Regulator and Gauge. The regulator is necessary to maintain calibration when a gas is being measured in this type of analyzer, where infra-red absorption is measured in a fixed volume cell. The subatmospheric pressure regulator is set at 0 psig. The pressure gauge has a range of 30" to 30 psig. (GO, WIKA) 13. Heated Enclosure. This Hoffman box houses the sample system components to prevent condensation. (Hoffman) 14. Vent Valve. This 4-way ball valve is located downstream of the analyzer. It is used to vent the line to atmosphere when calibration is being carried out during the time the vapor recovery system is down. It is necessary because the pressure in the main vapor line can then be 15 to 25 psig, making the calibration invalid under operating conditions. (Whitey)

Capillary Viscometer Sample System: Highly Viscous or High Wax Point Sample The special requirements for a sample of this kind are: •

The sample must be kept above its wax point temperature at every part of the sample system. Provision must be made to flush out the sample system with cutter stock when the system is being shut down. This flushing avoids plugging the lines with solid or viscous sample during maintenance and reduces start-up time.

•

The sample system components at the analyzer shelter are installed in a heated box. Valves are provided to flush this section also.

•

All electrical components must be suitable for Class 1, Gp D, Div. 2 electrically classified area.

The following sample components are shown in Figure 800-22. 1. Sample Take-Off and Return Root Valves. These ¾" gate valves should be heated and insulated. An extension of the tubing bundle electric tracer may be used for short lengths. (By piping to conform with Code) 2. Flush and Drain Valves. These ½",SS ball valves should also be heated and insulated. They are installed for flushing the sample lines with a cutter stock before shutting down the system for maintenance etc. (Whitey) 3. Heated Sample Line. The electrically traced, self-limiting tubing bundle is used to maintain the temperature of the line safely above the wax point of the sample. All parts of the system outside the heated box must be heat traced and insulated. (Dekoron, Parker Hannifin, O’Brien) 4. Heated Sample Box. This sample box must be brought up to temperature before the viscometer pump is switched on. The temperature indicator and pressure gauge should be visible from outside without opening the box. This insulated sample system box includes an electrical heater and temperature control unit. (Integrated by contractor.)

July 1999

800-58

Chevron Corporation



Fig. 800-22 Capillary Viscometer Sample System: Highly Viscous or High Wax Point Sample

5. Filters. Two Y-strainers in parallel with 100 mesh screens are recommended so that elements can be cleaned with minimum interruption of service. (Yarway) 6 & 7. Pressure Regulator and Gauge. The viscometer is not designed to handle process pressure variations. However, the double pump model of Precision Scientific is less sensitive than the Hallikainen type. The pressure setting depends on the return line pressure. The pressure regulator has SS components and ½" connections. 8. Temperature Indicator. This 2" dial thermometer should be visible outside the sample box. (Princo) 9. Sample Filter. This Y-strainer with 20 micron SS mesh protects the gear pump(s) in the viscometer from wear due to particles in the sample. A clean sample is the single most important factor in viscometer maintenance. (Yarway). 10 & 11. Flow Indicator and Control. The sample response time is kept to a minimum by maintaining a fast flow in the loop. The flow indicator has a typical range of 0–30 GPM range and has ½" connections. (Universal Flow Monitors) 12. Sample Box Flush and Drain Valves. See component 2 above. (Whitey)

Chevron Corporation

800-59

July 1999



13. Calibration Valves. A viscometer can be calibrated by collecting a sample at the outlet while noting the viscometer reading. The sample is analyzed in the lab, and any necessary adjustment to the viscometer signal (pneumatic or electronic) is made after the fact. Alternatively, a calibration standard may be introduced directly into the viscometer using these ¼" SS ball valves. Care is taken to circulate enough samples of the standard to flush out the process sample. (Whitey) 14. Temperature Indicator. The sample must be ten degrees F cooler than the viscometer bath for good control, since this depends on maintaining a load on the bath heater. A cooling coil is also provided in the bath for circulation of cooling water for the same reason. The temperature indicator has a 2" dial. (Princo) 15. Pressure Gauges. These gauges have diaphragm protection. Their range is fixed by the sample return pressure (Ashcroft)

Freeze Point Analyzer Sample System (Precision Scientific) The jet fuel freeze point analyzer runs through a cycle in which a measured volume of sample is cooled in the analyzer cell. The freeze point is associated with the temperature at which wax crystals disappear from a sample warming up from below its freeze point. If the sample is hot initially, the cell cooling system may not be able to bring the sample down to the freeze point temperature. So an important part of the sample system is a chiller for the plant cooling water. The following sample components are shown in Figure 800-23. 1. Sample Probe. The probe has a packing gland for extraction. 2. Root Valve. The ¾" gate valve is built to piping code. (By Field) 3. Local Sample Block Valve. This ½" SS ball valve is used to shut off sample at the shelter for maintenance. (Whitey) 4. Sample Pump. This pump is needed due to variations in the process line pressure. The pump bypass relief valve is set at 100 psig. 5. Sample Heat Exchanger is an initial sample cooler, using raw (not chilled) plant cooling water. 5A. Fast Sample Loop Flow Indicator. This armored SS flowmeter with ½" connections is used to indicate total flow. (Brooks, Wallace and Tiernan) 6. Temperature Indicator. This 2" dial thermometer with a 0-200°F. range gives the sample temperature before entering the second heat exchanger. See component 11 below. (Ashcroft) 7. Pressure Gauge. The bypass pressure (0–100 psig) is needed when setting the bypass flow using the valve described in component 19 below. (Ashcroft) 8. Bypass Filter. This filter with ½" connections and a 5 micron element handles only the sample flowing to the analyzer. The fast loop acts as a flushing stream. (Collins)

July 1999

800-60

Chevron Corporation



Fig. 800-23 Freeze Point Analyzer Sample System (Courtesy of Precision Scientific Petroleum Instruments Co.)

9. Pressure Regulator. The pressure is regulated to meet the limits of the analyzer cell. The SS regulator with ¼" connections has a range of 0–50 psig, set at 20 psig. (GO, Fairchild) 10. Pressure Gauge. Range 0–50 psig. (Ashcroft) 11. Sample Heat Exchanger. The sample is cooled by chilled water from the chiller. The analyzer inlet temperature limits are 50 to 120 deg F. If the freeze point is very low the analyzer will be more reliable if the sample is cool. 12. Calibration System. The calibration of the analyzer may be checked using a liquid standard which is introduced into the sample line from a cylinder pressurized by plant nitrogen at 20 psig. If for any reason the sample lines or the sample cell become coated with wax, a solvent can be introduced in the same way. 13. Sample Solenoid Valve. This valve is under the control of the analyzer which allows the sample to flow through the cell to flush out the previous sample and to warm the cell before the test sample is trapped. 14 & 15. Flowmeter and Valve. The sample flow rate through the analyzer is controlled and indicated by this valve and flowmeter. 16. Level Switch. This switch is installed to provide an alarm when the level of the collected sample is too high in the collection vessel, component 17 below.

Chevron Corporation

800-61

July 1999



17. Collection Vessel. The sample is collected in this vessel for return to the line under pressure. A manual drain valve can be used to lower the level in case of emergency. 18 & 18A. Sample Return Pump. The pump is set at 500 cc/min. A low sample flow is maintained into the vessel from the sample inlet line to provide a steady feed for the pump. A pump is needed since the sample return is in the same line as the take-off. 19, 20 & 21. The sample return flow is regulated and the pressure indicated by these components. 22. The remainder of the sample system is designed to circulate cooling water through the heat exchangers and the chiller. The design basis is for plant water at 50 psig and 65 to 80 deg F. The chiller can supply 2 gpm at 40 deg F.

Sulfur Plant Tail Gas Sample System: Excess H2S or H2S/SO2 Ratio (Ametek Analyzer)

The biggest problem in handling this sample is avoiding drop-out of elemental sulfur. The Ametek method mounts the analyzer above the sample take-off, requiring a special platform. A demister section is installed vertically above the take-off to allow sulfur particles to drop back into the line. The whole sample system is steam jacketed and insulated, and the sample is drawn through the lines and back into the process by a steam eductor. The following sample components are shown in Figure 800-24. 1. Sample Lines. These lines are 1" jacketed pipe with connections for 65 psig steam at 30 to 60 lbs/hr. They are kept as short as possible. 2. Sample Demister. A steam jacketed demister is designed to drop out elemental sulfur. 65 psig steam requirements are 10 to 20 lb/hr. Demister temperature is controlled at 260–265°F. 3. Demister Temperature Indicator. (Princo, Ashcroft) 4 & 6. Steam Jacketed Ball Valves. These 2" shutoff valves isolate the analyzer for calibration and maintenance. 5. Steam Eductor. An eductor is more reliable than a pump for this type of sample and keeps the sample temperature elevated. The eductor uses 90 to 110 lbs/hr of 65 psig steam regulated to 45 psig. (Part of the analyzer assembly) 7. Double Photometer Assembly. Two 400 series photometers are needed due to overlap of the absorption bands of SO 2 and H2S. One photometer measures SO2, the other measures SO2 + H2S. The H2S signal is separated in the electronics. 4620 Double Photometer Assembly. (Ametek)

Sulfur Plant Tail Gas sample system: Excess H2S, H2S/SO2 Ratio, or Air Demand (Western Research)

The following sample components are shown in Figure 800-25.

July 1999

800-62

Chevron Corporation



Fig. 800-24 Sulfur Plant Tail Gas Sample System: Excess H2S or H2S/SO2 Ratio (Courtesy of Ametek)

1. Sample Probe. This probe is a custom design in 316 SS positioned in the center of the tail gas line. 2. Sample Block Valves. These steam jacketed ball valves on both outlet and return are heated to avoid condensation of sulfur and blockage of the lines. 3. Sample Lines. These lines are custom, electrically traced, flexible Teflon. 4. Sulfur Condenser. This exchanger is designed to condense sulfur and bypass it to the return line before the sample is introduced into the analyzer cell. If sulfur were not removed it would fog the optical windows and affect calibration. 5. Sample Cell. A single cell is sufficient in this system, since the absorption of ultraviolet (UV) energy is measured at three wavelengths (see 9 below). The quartz windows that allow the passage of the UV energy are accessible through screw-off caps. Sample flow rate through the cell is 5 L/min with a velocity of 20 cm/sec. 6. Heated Sample Box. This box contains the sample cell, condenser, air aspirator, air drive heater, and sample shut off valves. The temperature of the complete system is maintained by thermistor sensors and a proportional controller. If any part of this system fails, a back purge is automatically started, to prevent plugging in a cold zone. 7. Sample Eductor. This eductor is a Teflon “jet pump” driven by heated air, drawing the sample through the cell, exhausting the sulfur from the condenser, and driving the sample in the return line.

Chevron Corporation

800-63

July 1999



Fig. 800-25 Sulfur Plant Tail Gas Sample System: Excess H2S, H2S/SO2 Ratio, or Air Demand (Courtesy of Western Research Div., Bon Valley Resource Services Ltd.)

8. Ultraviolet Lamp. This lamp is DC operated with high intensity at constant radiation level for long-term stability. 9. Three UV Radiation Filters. These filters are rotated through the beam emerging from the sample cell. 10. Photomultiplier Detector. This detector is sequentially exposed to three UV wavelengths by the filter wheel. The electronic circuits transform the output signal from the detector into an excess H2S, or “Air Demand” signal for the tail gas unit. 11. Pressure Regulator. The air supply is regulated for the eductor, zero air and condenser. 12. Air Preheater. The air preheater for the eductor ensures that the sample temperature is maintained. 13. Zero Calibration Valve. Zero adjustment is made with zero air in the sample cell after all the sample has been flushed out of the cell. Gradual deterioration of the UV source or deposits on the cell windows do not affect calibration due to the use of a reference frequency method. 14. Air Eductor. This eductor is driven by heated air so that the sample will not be cooled in the return line to the process. The eductor is needed to return the sample back into the tail gas line because there is not enough differential pressure between the take-off and return points.

July 1999

800-64

Chevron Corporation



Hydrocarbon in H2S Sample System The sample in this application is highly corrosive and toxic. Special materials are required in all parts of the system in contact with the sample. For example, ethylene propylene O-rings and seals are recommended for valves etc., and teflon seals in ball valves. Provision is made for purging the sample system with nitrogen before any break is made in the tubing for maintenance. The sample is automatically shut off if the liquid in the knock-out pot reaches a high level. The following sample components are shown in Figure 800-26. Fig. 800-26 Hydrocarbon in H2S Sample System

1. Pressure Control Valve. The sample can be taken on either side of this pressure control valve in the process line. 2. Sample Block Valves. These gate valves (usually ¾") are field installed gate valves to piping code. 3. Local Block Valves. These manual ½" 316 SS ball valves with Teflon seals are needed at the analyzer shelter to shut off the sample before purging the line for maintenance. (Whitey)

Chevron Corporation

800-65

July 1999



4. Heated Sample Line. The 1" sample line is heated to 125°F to prevent condensation in the line. The line is sloped back to the process line as a further precaution in case of upsets. (Field installed) 5. Liquid Knock-out Pot. This pot is located at the lowest point in the sample system and is drained back into the process line at least 8 feet downstream of the take-off. This 304 SS pot has an internal volume of 2250 cc. (Integrated by analyzer manufacturer) 6. Level Switch. This switch operates by sensing a high liquid level in the pot and actuating an alarm. Sample is shut off to the analyzer by way of component 7. The switch is an ultrasonic level sensor made by National Sonics. 7. Sample Shut-off Valve. This valve closes on high level in the knockout pot to protect the analyzer against liquid that would seriously affect the calibration and cause erratic readings. This ½" SS air operated ball valve with Teflon and EPR seals is normally closed and will shut off on air failure. (Whitey) 8 & 9. Pressure Gauge and Regulating Valve. A bypass flow is required to reduce sample lag time. The flow is adjusted to maintain at least 5 psig on the pressure gauge. The ¼" SS valve has teflon seals (Whitey). The pressure gauge is a 0-30 psig range and has a 4½" dial (Ashcroft). 10. Bypass Filter. This filter is considered the primary filter element, even though the knock-out pot provides inertial filtering through the change in direction of the gas flow. This filter also has liquid coalescing action. The bypass filter has a SS body, 0.6 micron element, 1" bypass connections, and ¼" sample connections. (Balston) 11. Differential Pressure Regulator and Regulating Valve. This flow control uses a differential pressure regulator controlling the pressure across a restriction valve. The 316 SS differential pressure regulator with Kynar has ¼" connections (Moore). The ¼" SS valve has Teflon seals (Whitey). 12. Flow Indicator. The flow to the analyzer must be the same for calibration gases and sample. The ¼" SS armored flowmeter has a range of 0.5–5 scfh (air) and ethylene-propylene O-rings. (Brooks) 13. Final Filter. This ¼" SS inline filter has a 0.6 micron element. (Nupro) 14. Back Pressure Regulator. Since the sample must be returned to the line, a back pressure regulator is required to maintain constant pressure on the analyzer cell, for correct calibration. The regulator has ¼" connections, SS and Teflon construction, and a range of 0–10 psig. (Go) 15. Solenoid Valves. The zero and span gases for calibration are introduced through a block and bleed valve arrangement. (See also Figure 800-27). This type of analyzer measures the absorption of infrared energy in a fixed volume gas cell. The pressure in the cell during calibration must be the same as the sample when it is

July 1999

800-66

Chevron Corporation



Fig. 800-27 Examples of Double Block and Bleed Valve Systems

being analyzed. These ¼" 3-way SS, 120 Volt, Class 1, Gp D, Div. 1 classification valves have Buna-N O-rings. (Asco) 16. Sample/Calibration Gas Select Valve. This air operated, ¼", SS, 4-way ball valve operates with the solenoids above. 17. Eductor. The ¾" 316 SS eductor draws the sample through the sample system and the analyzer and returns it to the process line. (Penberthy) 18. Enclosure. The sample system components are mounted in an electrically heated enclosure controlled at 120°F. 19. Auto Zero and Span Unit. This unit automatically adjusts the analyzer output to the correct level when the calibration gases are introduced. It can actuate an alarm when the adjustment exceeds a selected amount. 20. Check Valve. This valve safeguards the analyzer against steam back-up if there is a blockage downstream of the eductor. The ½" SS check valve has a Teflon coated spring and ethylene propylene O-rings. (Nupro)

Chevron Corporation

800-67

July 1999



Examples of Double Block and Bleed Valve Arrangements for Analyzer Sample Systems An analyzer may monitor more than one stream and stream switching will normally be automatic. Further, a single stream analyzer will very often be calibrated by a gas or liquid which is introduced automatically by a timing device. In any of these cases, if a simple block valve arrangement is installed, it is impossible to know if one stream is contaminating another due to a leaky valve. The best safeguard against cross contamination is to use a “double block and bleed” valve arrangement. The name applies to any valve system which has a low pressure bleed to drain, downstream of the first block in the non active stream and a second block in the line taking the sample to the analyzer. In a multi-stream system, a fast loop of each sample should be kept flowing continuously near the analyzer, so that a fresh sample is always available at the stream switching valve. Examples of double block and bleed valve arrangements are shown in Figure 800-27. The two systems on the left are for solenoid valves. There is one system on the right for air operated valves, using common air actuators. It is shown in two positions, one with the “sample” selected, the other with “span” selected. The selected stream sweeps the line to the analyzer in each case. Leaks in the valves are indicated by continuous flow through a bubbler or a flowmeter in the drain/vent line. If the valves are not leaking, the flow will stop shortly after stream switching has taken place.

Sample System for Gasoline Analyzers This is a section of a sample system which is common to three gasoline analyzers, having the same sample source and the same return line. The vapor pressure analyzer shown in Figure 800-30 is a continuous flow model by ABB. The minimum sample tubing size is ½", to maintain adequate sample flow through the analyzer. The following sample components are shown in Figure 800-28. 1. Shutoff Valve. The sample can be shut off locally from all of the analyzers by this 1" SS ball valve. (Red-White) 2. Pressure Gauge. Common sample source pressure is indicated locally by this 0–400 psig gauge. (WIKA) 3. Bypass Filter. The sample take-off from the fast sample loop is through the side branch of this ½" SS filter. (PermaPure) 4. Flowmeter. Fast sample loop flow is indicated on the meter. The flow can be maintained when any or all of the analyzers are shut down and to reduce sample lag time on start-up. The flowmeter has a range of 0–15 gpm and is armored. Flow rate is set at 8 gpm. (Brooks) 5. Regulating Valve. Fast sample loop flow is regulated by this ¾" SS rising plug valve. (Whitey)

July 1999

800-68

Chevron Corporation



Fig. 800-28 Sample System for Gasoline Analyzers

6. Sample Shutoff Valve. This ½" SS ball valve isolates the sample system while maintaining fast sample loop flow. (Red-White) 7. Flow Indicator. This meter indicates the cooling water flow through the sample heat exchanger. The 3/8" SS flowmeter with a range of 0–3 gpm is set at 1.5 gpm. (Brooks) 8. Cooling Water Flow Regulator. This is a ½" SS needle valve. (Whitey) 9. Sample Heat Exchanger. The sample temperature is controlled at 100°F in the vapor pressure analyzer so the sample inlet temperature should be below 100°F. (The heat exchanger is integrated by the analyzer systems manufacturer). 9A. Temperature Gauge. This ½" dial thermometer indicates the sample temperature at the outlet of the heat exchanger so that the cooling water flow rate can be adjusted. (Princo) 10. Coalescing Filters. The ½" SS filters are installed in parallel, to allow for changing of elements without disturbing the sample flow. (Balston) 11. Filter Bypass Valve. ¼" SS ball valve. (Whitey) 12. Bypass Flowmeter. The coalescing type of filter should have a continuous bypass flow, to avoid accumulating water. The flow rate is adjusted to the minimum required to keep the sample clear of water. The ¼" SS flowmeter with needle valve and a range of 0–5.5 gph is set at set at 2 gph. (Brooks)

Chevron Corporation

800-69

July 1999



13. Pressure Gauge. The pressure downstream of a filter is an indication of the need to change the filter element. The gauge has a 2½" dial and a range of 0–300 psig. (WIKA) 14. Shutoff Valves. Each of the sample lines to the analyzers can be shut off independently of the others using these ½" SS ball valves. (Red-White) 15. Check Valve. A check valve offers added protection against the sample flowing back in a loop. The check valve is ½", 316 SS, and has a 1/3 psi spring. (Nupro) 16. Air Operated Valves. These ½" SS ball valves are operated remotely for a variety of reasons to isolate the analyzer by shutting off the sample line both upstream and downstream of the analyzer. (Whitey) 17. Sample/Calibrate Selection Valve. This manually operated, ½", 3-way ball valve introduces a standard into the analyzer for calibration check. (Whitey) 18. Check Valve. A check valve protects against back flow if valve 17 is opened when there is a failure of the calibration system pump. The valve is ½" and has a 1 /3 psi spring. (Nupro) 19. Valve. This ½" SS needle valve controls the flow through the vapor pressure analyzer. (Whitey) 20. Valve. This ½" SS ball valve is operated together with valve 5 above when it is necessary to maintain the fast sample loop flow while the sample to the analyzers is blocked. (Red-White) 21 & 22. Check Valves and Block Valves. These valves are in the return lines from the other gasoline analyzers. (Red-White) 23. Pressure Gauge. Return line sample pressure must be sufficient to keep the sample flowing into the process line. The SS gauge has a 2½" dial and a range of 0–100 psig. (WIKA)

840 Analyzer Specification Analyzer specifications not only help define instrument, utility, and shelter needs but also provide a basis for comparing competing bids. Because process analyzer requirements are very much application-dependent, certain additional documentation is required for a successful installation. The responsibility for providing the documentation can be the Company’s, the integrator’s, or both. Responsibility is also defined by whether or not the job is small and to be carried out in-house, partially in-house, or turn-key. The following annotated specifications, data sheets and forms are included in this manual: • • • •

July 1999

ICM-MS-4362 Fabrication of Process Analyzer Systems ICM-DS-4362, Analyzer Enclosure Check-off Specification ICM-MS-4363 Installation of Process Analyzer Systems ICM-DG-4809 Process Analyzer Specification Sheets

800-70

Chevron Corporation


•


EF-885 Analyzer Project Roles and Responsibilities

Bid Evaluation. Reviewing proposals or quotes from an acceptable integration shop is no different from any other type of materials-management procedure. It is, however, more complicated than placing an order for a field transmitter or a dial thermometer because of the diverse aspects involved. You must write clear, rigorous specifications and know the integrators and vendors.

841 Required Documents Refer to the specifications for documents which are required from vendors, integrators, and contractors. The following sections discuss additional documents that the responsible engineer must fill out:

Analyzer Specification Sheets Analyzer specification sheets are ISA-type forms that must be prepared for each analyzer. The accuracy of each form is essential to the success of a project as an unknown contaminant or an interfering component can render a sample system or an analyzer useless. Specification sheets contain information about the process conditions that determine both the analyzer selection and the sample system design. Number specification sheets individually and sequentially and make certain that they contain references to the applicable engineering flow diagrams and the requisition numbers. A process specification that spells out the weight percentage of each component to four decimal places may be totally useless unless the minimum and maximum ranges possible during any type of operation (including upset, startup, shutdown, and turnaround conditions) are listed and known to the designers. The temperature and pressures of all streams must be obtained under all possible conditions. It is very important to know if a component is in the liquid or gaseous phase or if it is at the bubble point. To ensure that you have the information required, fill out an analyzer specification sheet.

Acceptable Vendors and Integrators List Compile a list of acceptable analyzer vendors alter considering the requirements for each application. (This includes training and spare parts requirements.) Refer to the analyzer selection section of this manual and discuss your selection with analyzer specialists. Note that analyzer installations usually have greater reason to justify single-sourcing hardware based on the expense of spare parts and local maintenance training.

Additional Documentation for In-House Jobs Instructions for Company in-house jobs require documentation and drawings for both specification and construction group walk-throughs. Typical engineering job instructions are included in Appendix E. Small jobs are often engineered in-house and require the following documentation:

Chevron Corporation

800-71

July 1999



Plot Plan. Include the location of the shelter, utilities. and sewers on an existing plot plan or a sketch. Sewers may require separate drawings. Analyzer House. Locate the analyzers, sample system components, and utility headers on the house drawing(s).

Foundation Drawings System Overview. Include the sample tap, return, fast loop, sample conditioning plate or box, sample lines, data transmission, analyzer, maintenance recorder, and operator readout devices. Locate all utilities and steam addition and traps on the sample line.

Power and Signal Drawings Cylinder Racks, Control Room or Analyzer Maintenance Room. As required, these drawings are for boilermakers. Sample Probe Sketch. Refer to the sketch in Appendix E. Typical Engineering Job Instructions. The instructions can be modeled alter the form in Appendix E.

842 Analyzer System Inspection and Acceptance Procedure The following documentation is to be given to the Company at the acceptance test: 1.

A copy of the final acceptance procedures. Given to the inspection representative by the system integrator six weeks prior to the test.

2.

A general information sheet indicating the gases or liquids used to develop the application and a copy of a calibration or liquid analysis used for calibration. The original stream data is furnished by the Company representative.

3.

A record showing 48 hours of uninterrupted operation on a calibration standard. This record should be fully documented with the time, date, chart speed. range, sample size. flow temperature. and any available information to indicate the conditions.

Performance The analyzer assembly and operations are to be inspected. In the calibration standard, the systems assembler may or may not have all of the components listed in the process stream data sheet, provided by the Company representative. The calibration standard must contain components that are designated by an asterisk. The system integrator warrants that the analyzer will perform within the required tolerances of repeatability and accuracy on a sample, corresponding to the original Chevron stream data. Performance is verified during the acceptance test. The company representative observes the analyzer operations for eight continuous hours and is responsible for conveying the information received at this test to the plant maintenance file.

July 1999

800-72

Chevron Corporation



Require that documentation of the 48-hour repeatability test be given to you in advance of your visit.

843 Analyzer Shelters The analyzer shelter can be three sided, fully enclosed, or free standing and is assembled according to the systems diagrams and materials of construction described in the specification. Normally, analyzers are mounted on wall racks, except for those requiring all-around access. The system is assembled in such a way that any one analyzer can be removed without interrupting the operation of the others. All piping and tubing connections are made through bulkheads. The objectives in providing an analyzer house are as follows: •

To create a space within an area that would be a nonhazardous area under normal operating conditions to test or calibrate analyzers and perform maintenance with opened casings and live electrical circuits. This space is achieved by pressuring the shelter to lower the electrical classification and by limiting the amount of hydrocarbon allowed to enter the shelter. This includes keeping cylinders and sample conditioning outside the walls of the shelter whenever possible.

•

To create an environment in which analyzers and the ancillary equipment are protected adequately from the weather and which provides a stable operating temperature for the analyzer.

•

To provide space for future equipment. Consolidating a number of analyzers in the same building has the advantage of all-weather and multiple servicing. Most installations allow multiple groupings. In a complex process unit, there may be a number of groups strategically located.

Choose analyzer locations for the following reasons: accessibility; proximity to the sample probe and utilities; and greatest possible distance from explosives and other hazards. Technicians must be able to access the building. The locations must also be near an unloading point for carrier gas, calibration gas cylinders, and equipment. Grouping analyzers reduces costs as they share common instrument cable and tubing trays. See Figure 800-29 Analyzer Shelter Documentation for information that the integrator provides for completing the Process Analyzer Specifications Sheet (ICM-DS-4809).

Field Documentation Complete the installation details of probes, sensors. and analyzer hardware located in the field outside the analyzer shelter. Check for items in the following list:

Chevron Corporation

1.

Size and location of the first block valve and line identification

2.

Location of sample tap hardware

3.

Equipment support

800-73

July 1999



Fig. 800-29 Analyzer Shelter Documentation Analyzer Shelter Documentation. The integrator should provide information concerning the following: •

Outside dimensions of each shelter.

•

Detail anchoring method.

•

All piping, tubing, and electrical field connection points.

•

Recommended lifting methods.

•

Utility demand and requirements.

•

Cable scheduling and tagging from the shelters to the control room.

•

Ground bus location for field connection.

Company engineer should consider the location of the following equipment: •

Sample systems mounted on plates or in enclosures.

•

Flare headers for sample returns.

•

Vent headers.

•

Steam headers and steam regulator.

•

Drain headers.

•

Block valves for sample, vent, and flare lines.

•

Cylinder racks.

•

Bulkheads.

•

H2S Detectors if appropriate.

•

O2 deficiency detector, to be considered.

•

LEL sensor, to be considered.

Analyzer building inlet and outlet connections. These include samples, headers, utilities, (water, steam, power, etc.) entering and leaving the building are to be consolidated in one area. The tubing , pipe, and headers on the building walls should terminate at the most convenient end of the building for final termination. Drain headers should be routed to the proper underground drain. The goal is to be able to run one support for building installation connections.

4.

Methods of attachment

5.

Utility requirements and connection details

6.

Signal and cable routing, schedule, tagging and connections details (The integrator and the contractor coordinate incorporating this information on the contractor’s routing drawings.)

Tag Information. To facilitate field installation, the integrator must tag all elements of the analyzer system that are shipped as individual components or modules in accordance with the sample system and the installation drawings.

July 1999

800-74

Chevron Corporation



850 Analyzer Installation, Commissioning, and Startup 851 Analyzer Installation Checkout Procedure The purpose of this procedure is to outline the methods for checking field-installed analyzers. This procedure can be followed by contract personnel and Company inspectors prior to continuity, loop, and leak checks (refer to ICM-MS-4363, Installation of Process Analyzer Systems). Visually inspect all analyzers and accessories for damage alter installation. If damage is observed, take corrective action. Each analyzer should have a sheet with calibration and check -out data for every analyzer and analyzer loop that is inspected. Compare all analyzer installations against flow and installation drawings for proper location and for compliance with any special provisions in these drawings. Check that all analyzers are free of shipping restraints. Glass should be clean. Temporary protective devices must be removed before startup. Verify that all analyzers and tubing are correctly tagged, labeled, and installed at the correct location. All sample systems and sample probes must be tagged and checked for proper installation with respect to the direction of flow, elevation, orientation, insertion, and depth. Mount phenol material, 3 inches × 6 inches, on a bracket designed to attach to the root valve piping, and post a warning to close the valve on the probe (see the sample probe drawing in Appendix E). Check analyzer wiring for circuit continuity. proper terminal connection, and labels. Check analyzer piping and tubing which is process-connected for agreement with P&ID’s and for appropriate materials, connection sizes, and rating. Verify that adequate brackets and supports have been used to maintain rigid construction. Verify the correct installation of all utilities. including proper steam addition and trapping of prefabricated steam-traced bundles. Check the manufacturer’s engineering data for these. Do not commission sample lines until the process line has been flushed clean. Isolate all analyzers from the process lines by blocking off or disconnecting them. Clean the sample tubing lines thoroughly. Ensure that all tubing connections are made according to the manufacturer’s recommendation. Replace overtightened connections and tubing that is scarred by a tube cutter or a knife used to remove the PVC jacket.

852 Analyzer Commissioning and Startup These commissioning procedures provide the details for verifying the operability of analyzer installations. Ensure that all checks, tests, and adjustments are conducted by qualified analyzer personnel and verified by Company personnel.

Chevron Corporation

800-75

July 1999



Leak Testing Instrument Utility Headers and Process Tubing. Perform a leak test on the sample system and utility header. Test clean tubing, not hydrostatically tested, with 100 psig instrument air and spray all fittings with a soapy. solution. Close block valves and vent all analyzers. Remake, with new ferrules, all tube fittings that leak alter one additional turn of the compression nut. Clean tubing for GC support gases (helium, hydrogen, and air) with a volatile solvent and purge it with nitrogen. Solvents, such as methylene chloride, have been used in the past. Solvents, such as hexane, are not sufficiently volatile for cleaning purposes. Turning Power on the Analyzer System. Check electrical terminations for continuity and location; verify the utilities; ensure that power is turned on by the party responsible for the analyzer.

Analyzer Calibration Field Procedure The process sample lines remain closed until after a calibration check of the analyzer. Documentation. A general information sheet is required that indicates the gases or liquids used to develop the application and that gives an analysis report of the calibration standard. Determine the appropriateness of the standard by using the correct stream data from the Analyzer Specification Sheets. The analyzer technician fills out the attached form for chromatograph checkout in the field. Performance. Inspect the analyzer operation. The integrator may not have all the components of the stream in the calibration standard. The calibration standard must contain the components of interest and any possible interfering components. The integrator must guarantee that the analyzer will correctly analyze a stream containing the components which are outlined on the Analyzer Specification Sheets. The prime contractor is responsible for field calibration, although he may delegate the work to others. The analyzer shall show the same repeatability as in the factory test. Refer to the Analyzer Startup Procedure in the Company’s data sheet ICM-DS-4362

860 Calibration and Validation of Analyzer Output 861 Calibration The only way to determine whether or not an analyzer is accurate when received is to check it against a series of calibration standards. After this initial calibration, the analyzer must be checked periodically against at least one suitable standard to ensure continued accuracy. Specification of calibration standards are discussed in ICM-MS-4362. This specification considers the limitations of the external standard method and the normalization method.

Weight Versus Volume Basis Almost all analyzers can provide direct results on a molar or volume basis, because a fixed volume of sample is actually analyzed. If sample density is not constant,

July 1999

800-76

Chevron Corporation



inaccuracies can occur. Normalization allows for calculations on a weight basis. If sample density is constant, calculations by both weight and volume are possible.

External Standard Method The calibration and validation of analyzer output is obtained with the external standard method and can be applied to most analyzers. Analysis can be determined on a volume basis only when the sample density is constant. For photometric analyzers, both a span and a zero standard are required.

Normalization Method In this method, normally only for chromatographs, each peak area is corrected for normal variations in detector response. The results can be determined on a weight or a volume basis. All components must be measured, however, the sample size is relatively unimportant.

862 Continuous Validation of Analyzer Output Continuous validation of analyzer output is required for closed-loop control. The validation can be internal and/or external via switches and relays that are added by the user. An example of analyzer integrity is provided in Figure 800-30. In Appendix E there is an electrical loop drawing of an analyzer system with pressure alarm. Errors are detected by failed function at the analyzer or its controller. At the DCS or host computer, such errors are classified in a programmable hierarchy so that the control system can take appropriate response. Some error functions may only be diagnostic in nature, while others (caution, warning, and error) indicate an inaccurate analysis. At present, because vendors have systems with various options and users require different configurations. it is advisable to consult with process control and process analyzer experts and with the vendors to determine where alarms should be classified in the hierarchy. Serial Communication check. In digital data transmission, there is a possibility that one or more bits of a digital number may be in error. Commonly, errors are checked by using parity. An extra bit, such as an ASCII character, added to a binary word is set on or off to achieve an even or an odd parity. If a bit in the binary word is dropped or incorrect, the total number of bits is checked. This is not a fool-proof method because more than one bit may be dropped or be incorrectly high. If the checksum method is used (where a sum is transmitted by a block of characters), the receiving device then sums the characters and compares the sum with the received checksum. If the two do not agree, the entire transmission block is rejected. Analyzer self validation check. Analyzers can be designed to flag errors in temperature, drift, and carrier pressure. These checks are outlined for chromatographs in ICM-MS-4362. Self-validation checks are also available for other analyzers. It is preferable that such signals form a common alarm to the computer (good/bad).

Chevron Corporation

800-77

July 1999



Fig. 800-30 Analyzer Integrity Example — Boiling Point Analyzer

Sample system check. Usually, sample-system check is limited to sample flow in the fast sample circulation loop and in the slipstream to the analyzer. Such checks are necessary for exceedingly dirty streams or in applications so critical that daily checks by an analyzer technician are insufficient. Sample flow shut-off is likely to affect composition in the percent level very quickly and is easily detected by manipulation of computer controller data. You must consider if this will guarantee the necessary validation. Computer controller data manipulation. The system computer that accepts the analyzer signal can check its validity. Integrity checks include minimum or maximum rate of change of signal and analyzer out of range (high/low limit). See Analyzer Output Integrity Checks in Figure 800-31..

July 1999

800-78

Chevron Corporation



Fig. 800-31 Analyzer Output Integrity Checks Process chromatographs should have the following internal integrity checks. • Detector balance greater than limit - Balance required to zero detector is greater than instruments limit.

•

Chip failure.

•

Any undefined table reference.

•

Invalid data type, analog or digital.

•

I/O errors.

•

Temperature deviation exceeded.

•

Power failure.

•

Low carrier pressure.

•

Low sample pressure.

•

Peak outside concentration limit.

•

Purge failure.

•

Oven shutdown.

Oxygen Analyzers Oxygen analyzers used for furnace control require special error warning message to ensure proper operation of furnaces. There are extractive type analyzers and close coupled as well as in situ (zirconium) analyzers used in this service.

•

Any insufficient memory problem in programmer tables.

Signals required include: •

Instrument in calibration.

•

No access to analyzer data highway.

•

Calibration failure.

•

Host computer interface does not respond.

•

Instrument in “blow-back” mode.

•

Message transmission failure on data highway.

•

•

No stream selected at cycle start.

Cell temperature out of range (zirconium analyzers).

•

Detector flame out.

•

Cell failure (zirconium).

•

Detector balance greater than limit.

•

Furnace shut down / flame out.

•

Detector zero greater than limit.

•

Power failure.

•

Peak retention drift.

•

Sample flow or pressure (extractive type).

•

Analog to digital converter failure.

•

Any programming error, i.e. gates overlap or too many gates.

Typical oxygen sample systems with integrity checks are contained in Section 830.

•

Number of gates on does not match number of gates off.

•

Tables unavailable.

•

Any failed signal i.e. failure to signal balance to detector board.

•

Other Analyzers In critical control applications the analyzers will require the following signals where applicable. •

Cell failure.

•

Low sample pressure at analyzer inlet or low flow.

•

Calibration.

Numeric value found where string expected and vice versa.

•

Loss of power.

•

Loss of enclosure purge.

•

Total area change outside limit.

•

Failure to calibrate.

•

Failure to reset cycle.

•

Any queue, peak, or arithmetic overflow.

•

Invalid address.

•

Failure of component to respond.

Chevron Corporation

800-79

July 1999



870 Safety 871 General Consider the safety of operating and maintenance personnel when designing an analyzer installation. Several safety items have already been discussed: observing proper metallurgy in choosing sample system piping, providing relief valves to prevent overpressuring, and providing drain valves for relieving system pressure prior to performing maintenance. Safety considerations discussed in this section are sample line installation. sample disposal problems, and electrical ignition problems in sample piping and analyzer areas.

872 Sample Line and Sample System Components Ensure that sample lines and all connections to sample system components are leakproof and that small piping is seal welded. Select tubing and tubing fittings (considered more reliable) wherever possible. Where pumps are required in the sampling system, provide a means for collecting and disposing of possible leakage. Where an air supply is available, install air-driven rather than electrically driven pumps to increase sample pressure or to handle sample disposal with a minimal hazard from malfunctioning. To protect personnel, leak proof and insulate or isolate hot sample lines. Locate outside analyzer shelters outside all high-pressure, high-temperature, and highvolume sample loops that reduce time lag, along with their associated pumps, filters. rotameters, and heat exchangers. Select 1/8-inch or ¼-inch stainless steel tubing for sample lines entering the analyzer shelter. More exotic alloys may be required in some applications involving highly corrosive service. Closed-couple all samplesystem components inside the analyzer enclosure to minimize sample volume and locate them in a sample system housing that can be purged to reduce the possibility of an explosion in the event of a leak. Check pressure/temperature ratings of each component and operate at 85 percent or less. Some locations require certification by a nationally recognized testing agency (UL or FM)

873 Leak Detection Due to the large number of small fittings, the risk of leaks is high. Ensure that enclosed sample systems handling hazardous material (such as H2S) have leakdetection systems that warn the operator or maintenance worker of leaks before he or she opens the enclosure. This warning can range from H2S paper at a vent opening to tieing into an existing remote, leak-detection system.

July 1999

800-80

Chevron Corporation



874 Sample Disposal System Gas and Vapor Some sample system installations do not have sufficient pressure drop available to allow sample return to the process without the addition of a pump or some other means of sample disposal. In most locations, it is not acceptable practice to discharge small amounts of light hydrocarbon vapors to the atmosphere. Venting requires the necessary piping to bring vented hydrocarbons to a safe location. For example, because most chromatographs are sensitive to backpressure in the vent lines, the vent line size should be large enough so that vented sample flow cannot result in a buildup of backpressure. GC carrier gas vents are still acceptable as long as they are piped outside the analyzer shelter. Future installations face more stringent environmental requirements governing the allowable amount of effluent vented to the atmosphere. Means of sample disposal will include relief lines. furnaces, or return to the process. If such an installation is required, take the precautions necessary to eliminate the potential backpressure of a closed system. If sample disposal is made in a furnace, provide a means to shut the vent automatically in case the furnace is shut down. This arrangement eliminates venting hydrocarbons into the furnace while maintenance personnel are working on the furnace.

Liquid Liquid drains for analyzers are two-inch (minimum) pipes sloped to an oil sewer. Vent the high point of this drain line to the atmosphere as described above for vent lines. Do not allow valves or goosenecks in drain lines as they might create liquid seals. Heavy components being sampled might solidify in the drain line at atmospheric temperature, therefore, it may be necessary to steam trace or electrically trace and insulate the drain line.

875 Electrical and Ignition Problems Sample systems and sample-system components include electrical components, such as motors, electrical heat tracing, solenoid valves. and pressure switches. It is necessary. therefore. to be familiar with the electrical classification of the installation area and to select the proper location, material, and equipment consistent with the classification. The analyzers (and some of their associated sampling system components carrying hydrocarbons) will be located in cases or enclosures that may also contain analyzer electrical equipment. The National Electric Code defines conditions for specific classifications. Refer to Section 300 of the Electrical Manual for a more detailed discussion on area classification. Place the analyzer in a nonhazardous area. If this is not possible, build the analyzer or hazardous portions of it into explosion-proof housing to allow safe operation in Class 1, Group C&D, Division 1, areas. If such an installation has been made, explain to the operating and maintenance personnel that there is a possibility of unsafe conclusions existing during maintenance and calibration. These conditions are related to opening explosionproof housings and exposing electrical and electronic elements while they are operating. Obviously, soldering components and other hot work cannot be carried out in

Chevron Corporation

800-81

July 1999



areas where hydrocarbons are present without a special hot work permit. If such conditions exist, remove the analyzer or the faulty unit. Moreover, it should be understood that precautions must be taken when removing and replacing explosionproof housing covers and conduit covers. Screwed or flanged-type, explosion-proof housings require constant attention if they are to remain safe. For example, the explosion-proof cover is safe only for as long as all the bolts of a flange-type cover are installed and tightened properly. If maintenance tightens only half the bolts, the housing can no longer contain an explosion. If the cover is dropped, corrodes, or if the smooth mating surface is damaged, it is probably no longer safe. Purging an enclosure following the requirements of NFPA 496 is another way to protect electronic equipment. Section 9 of this bulletin refers specifically to analyzers and analyzer shelters. Intrinsically safe electrical systems have become popular recently because of their convenience and the savings in the initial equipment costs and in installation costs. (See Section 1400 of this manual for a more detailed discussion on intrinsic safety.) Equipment that is intrinsically safe can be maintained conveniently as heavy covers are not required to be in place when the equipment is in operation. Calibration checks and adjustments can be made safely in the field without special precautions even though the equipment is installed in a Division 1 area. In conclusion, safety for operating and maintenance personnel in hazardous areas can be obtained by: 1.

Containing a hazardous device in an approved enclosure (explosion proof for Division 1 areas; hermetically sealed for Division 2 areas).

2.

Changing the nature of the atmosphere surrounding the source of ignition by providing a forced ventilation system.

3.

Installing of the hazardous device in an isolated, non-hazardous area.

4.

Removing the source of ignition with intrinsically safe circuit designs.

880 References For further information consult with local maintenance personnel, engineers, vendors, and analyzer specialists in CRTC’s M&CS Unit. The following list of references will also be of help:

July 1999

1.

Process Analyzer Technology. John Wiley and Sons, 1986.

2.

On-Line Process Analyzers, Wiley Interscience, 1988.

3.

Principles of Sample Handling and Sampling Systems Design for Process Analysis. ISA,1972.

4.

The Design and Application of Process Analyzer Systems. Wiley Interscience, 1984.

800-82

Chevron Corporation



5.

General Handbook of On-line Process Analyzers. Ellis-Horwood, 198 1.

6.

Gas Chromatographs as Industrial Process Analyzers. Pergamon Press, 1911,

7.

Sampling Systems for Process Analyzers. Butterworths, 1981.

8.

Quality Measuring Instruments in On-Line Process Analysis. Ellis- Horwood, 1982.

9.

pH Control. ISA, 1984.

10. API RP 555, Process Analyzers 11. ANADATA. Measurementation. 12. ISA Directory. 13. Intech, ISA. 14. Instrumentation and Control Systems. Chilton. 15. Refinery Analyzer Applications Reference Document, CRTC, 1996. 16. Analytical Instrumentation. Chilton Book Co., 1994.

Chevron Corporation

800-83

July 1999

900 Control Valves Abstract The intent of this guideline is to assist instrument engineers in the selection and specification of control valves. The function of a control valve within the process is explained, and the types of control valves available are discussed. Subsections offer information on valve sizing, trim characteristics, accessories, and selection of material for both valve bodies and trim. Other subsections review actuator types, selection, sizing, and discuss self-contained pressure regulators. In addition, this guideline suggests how to avoid common control valve problems and discusses valve installation and piping considerations. Pertinent industry standards are listed, and a brief reference list is included. Contents

Page

910

Introduction

900-3

911

Definition of a Control Valve

912

Safety Considerations

913

Control Valve Alliance

920

Hydraulics

921

Process Hydraulics

922

Valve Hydraulics

930

Control Valve Basics

931

Types of Control Valves

932

Applications

933

Trim Characteristics

934

Control Valve Sizing

935

Material Selection

940

Actuators

941

Spring-and-Diaphragm Actuators

942

Actuator Sizing

ChevronTexaco Corporation

900-5

900-9

900-31

900-1

April 2002

900 Control Valves

April 2002


950

Control Valve Accessories

900-35

960

Control Valve Installation

900-42

970

Control Valve Problems

900-44

971

Mechanical Problems

972

Design Problems

980

Pressure Regulators

900-48

990

References

900-49

900-2



910

900 Control Valves

Introduction The control valve guidelines are directed at the entry level design engineer but are structured as a reference for experienced instrument engineers as well. For example, subjects such as the Health, Environment, and Safety Group restrictions on use of “between-flange-mounting” control valves for hydrocarbon service are critical to anyone selecting a control valve. “Safety,” below, and Section 931, “Types of Control Valves,” discuss this in detail. The engineer must be aware of the control valve’s function in a hydraulic system and the relative pressure drop that must be allocated to a control valve. The engineer must also know that the control valve must be able to shut off against all flow conditions—that the worst-case upstream pressure (under upset conditions) must be specified so that the actuator can be selected and sized to actuate the valve safely. Section 942, “Actuator Sizing,” covers this area in detail. A control valve must function through all ranges of operating conditions, and all operating cases must be considered in its engineering, e.g., startup (when the system operates under reduced capacity, the piping is clean, and the friction drop is minimal); normal (when the operators try to run the system at, or slightly above, its design specifications with the piping in fouled condition); and every other possible process circumstance (including upsets) that could affect the rangeability of the control valve. Section 932, “Applications,” covers this area. For the engineer familiar with hydraulics and rangeability considerations who needs guidance in selecting the type of control valve to use for an application, valve types and Company experiences and preferences are discussed in Section 931, “Types of Control Valves.” Applications for fire-safe valves are also discussed in Section 931. Control valve sizing methods are covered in Section 934, “Control Valve Sizing, ” This section also discusses personal computer (PC) software for sizing control valves. Section 970, “Control Valve Problems,” includes information which, although it may not be applicable for day-to-day, low-pressure, constant-flow applications, has, when ignored, resulted in substantial operating and equipment losses at Company facilities. This information is provided in hopes of precluding similar occurrences. If, after reading this guideline, an engineer still has concerns about the selection or application of control valves, ERTC Instrumentation and Control specialists, listed in the ERTC Consulting Services card, will be glad to assist.

911 Definition of a Control Valve A control valve is a device, installed in a hydraulic system, that controls flow through the system by introducing and modulating pressure drop within its body. By increasing pressure drop through the system, the control valve creates a resistance to flow, thus decreasing the flow.


900-3

April 2002

900 Control Valves


Proper selection and engineering of a control valve will prevent the following types of problems: •

Installation of a valve that is unsafe for the application, e.g., a between-flangemounting valve installed in hydrocarbon service

•

Cavitation, flashing, and choked flow (in downstream piping), which can occur in liquid service

•

Noise and choked flow (in the valve body), which can occur in vapor service

•

Erosion, which can occur at high-pressure drop in all applications

•

An actuator without the power to either maintain a constant position or one that is not capable of overcoming the process pressure and to shut off against it

912 Safety Considerations Tests conducted jointly by the Company’s Health, Environment and Safety Group and by several major oil companies showed that between-flange mounting control valves (installed in piping with long exposed bolts) are not fire-safe and, with specific exceptions, must not be used in hydrocarbon service. (See Section 2082 of the Company’s Fire Protection Manual.) Examples of such valves are flangeless Fisher Vee-Ball, Masoneilan Camflex designs, and many butterfly valves, etc. In a flash fire, the exposed bolts expand faster than the valve body, open the space between the valve body and the flanges, and allow the contents of the piping to spill out and feed the fire. Metal shields around the bolts were believed to preclude this problem. Tests show that metal shields delay the failure by only a matter of minutes. Also, these shields are seldom replaced when maintenance is performed on a valve. The Company’s Health, Environment, and Safety Group must approve all exceptions to the above, e.g., Fisher Vee-ball valves are approved for use on the decks of Company tankers. The logic is that the valves are on deck where they are not filled with hydrocarbon and are not under significant pressure. A tanker fire would destroy the ship before the contents would leak through the flanges of these control valves. Most control valve manufacturers have now modified their designs, and Fisher VeeBall and Masoneilan Camflex valves are offered both in the flangeless design and with flanged end connections. Most rotary control valve manufacturers also offer “between-flange-mounting” control valves either with lugged bodies, which protect the bolts from flame impingement, or with bodies drilled and tapped to receive studs or machine bolts to secure piping flanges directly to the valve body. Actuator material selection is also a fire-safety concern. An actuator must drive a control valve to its fail-safe position and hold it there during a fire. To do this, all components associated with maintaining the fail-safe position must be metallic and must have a melting point of at least 1750° F.

April 2002

900-4



900 Control Valves

913 Control Valve Alliance In 1995, after rigorously implementing THE COMPANY’S Supplier Quality Improvement Process (CSQIP), THE COMPANY signed an international alliance with Fisher-Rosemount to cover control valves, regulators, level instruments, field mounted instruments, Posi-Seal butterfly valves and services related to engineering, installation, maintenance, and diagnostics. An Alliance Improvement Team (AIT) and a number of Local Alliance Improvement Teams (LAITs) have been formed and are actively working to improve their product application, process improvement and new technology development opportunities, and to pursue both initial cost savings and life-cycle cost reduction of control valves. You can read about Chevron/Fisher Alliance successes, review the Alliance savings, and report on your cost savings by clicking the following link to the Chevron/Fisher Control Alliance Home Page. http://cpln-pub6.sr.chevron.com/Purchase/alliance.nsf/1dfd425a51dfa7f0882563cd007172c6/47cb71ae83a5d85488256702000f82 c4?OpenDocument

920

Hydraulics 921 Process Hydraulics A centrifugal pump is the most common device used to introduce pressure into a hydraulic system. In many processes, e.g., production separator, a constant upstream pressure is readily available and a pump is not used. Because the discharge pressure of a centrifugal pump varies, we will use the pump curve to discuss process hydraulics. In a constant pressure system the “pump curve” would be a straight horizontal line.

Pump Curve The pressure that a centrifugal pump can generate is a function of the flow rate. The head pressure produced by a centrifugal pump decreases as the flow rate increases. Conversely, an increase in flow results in a decreased pump discharge pressure. (See Figure 900-1.)

System Friction Losses Pressure drop produced by friction losses through piping and equipment increases as the flow through the system increases. This pressure drop increases as the square of the increase in flow rate.

System Pressure Drop System pressure drop is the difference between the pump discharge pressure and the system outlet pressure. It is a combination of piping system friction losses and the control valve pressure drop at any given flow rate. A pressure profile through a hydraulic system is shown in Figure 900-2.


900-5

April 2002

900 Control Valves


Fig. 900-1 Pump Curve

Fig. 900-2 System Pressure Profile

April 2002

900-6



900 Control Valves

922 Valve Hydraulics Valve Pressure Drop It is important to understand that a control valve does not define the pressure drop across it. It has to consume whatever pressure drop is available to maintain a system at set point. This concept is shown in Figure 900-1 (the difference between the pump discharge pressure curve and the system friction loss curve). At a high flow rate, the control valve does not have to consume as much pressure drop as it does at low flow rates. The control valve can only regulate flow by consuming pressure drop in the system. Its ability to increase or decrease flow rapidly depends on the portion of the total system pressure drop allocated to the control valve. For a control valve to be able to control a process effectively, the rule of thumb is to allocate to the control valve, at normal operating conditions, a pressure drop equal to about one-half of the piping and equipment friction drop (one-third of the total system pressure drop). This rule of thumb applies to processing facilities and must be re-evaluated for applications such as pipelines, where providing one-half of the pipeline friction drop for a control valve would not be practical.

Valve Flow Velocities In order to visualize flow velocity through a control valve, imagine a liquid flowing in a piping system in which a restriction such as a concentric orifice plate has been placed. Upstream of the orifice plate the liquid must have a static pressure greater than its vapor pressure and a velocity head due to the flow rate. As the liquid flows through the restriction, its velocity must increase. Because the sum of the static pressure and velocity head will remain equal, an energy interchange takes place, with the static pressure losing what the velocity head gains. (See Figure 900-3.) At the orifice outlet, the stream will reach its maximum velocity and minimum static pressure. The point of maximum velocity is called the vena contracta (Pvc).

Pressure Recovery Further downstream, as the fluid stream expands into a larger cross-sectional area, velocity head decreases and static pressure increases. However, the downstream static pressure never recovers completely to the pressure that existed upstream of the restriction. In a control valve, the pressure differential (∆p = P1 - P2) that remains after the pressure recovery, is the amount of energy that was dissipated within the control valve. The amount of pressure recovery depends on the type of valve being used. Figure 900-4 shows relative vena contracta pressure drops for a globe and a rotary valve with the same inlet (P1) and outlet pressures (P2).


900-7

April 2002

900 Control Valves


Fig. 900-3 Valve Velocity-Pressure Profile

Fig. 900-4 Valve Pressure Recovery

April 2002

900-8



900 Control Valves

The pressure differential between the valve inlet and the vena contracta is of interest relative to flashing and cavitation. If pressure at the vena contracta should drop below the vapor pressure of the fluid, vapor bubbles will form in the stream. If valve outlet pressure recovers above the vapor pressure of the liquid, the bubbles will collapse and cavitation will occur. If valve outlet pressure stays below the vapor pressure of the liquid, flashing will occur. (See Figure 900-5.) Fig. 900-5 Pressure Recovery—Possible Conditions

In a rotary valve, the velocity at the vena contracta is higher, giving more capacity through a smaller valve. The problem is that the static pressure at the vena contracta (Pvc) is more likely to drop below the vapor pressure of the fluid and can put the control valve into cavitation. Section 970, “Control Valve Problems,” addresses cavitation in more detail.

930

Control Valve Basics The next few paragraphs explain the terminology and provide the definitions used throughout the rest of the guideline.


900-9

April 2002

900 Control Valves


931 Types of Control Valves Globe Valves The term “globe valve” describes a vast array of body configurations. “Globe valves” are usually a variation of the piping-type globe pattern valve, where the distance between the plug and the seat can be manually adjusted to change the flow through the valve. (See Figure 900-6.) Fig. 900-6 Globe Valve

The following valves fall into the “globe valve” category: • • • • • •

Globe valves having “cage” trim (both balanced and unbalanced plugs) Globe valves having conventional single or double port trim Globe valves having angle bodies (may be used in special cases) Globe valves having split bodies (may be used in special cases) Three-way globe valves (may be used in special cases) Multistep globe valves (may be required for noise control)

Rotary Valves Whereas the globe valve uses a linear motion to change the relative position of the plug and seat, the butterfly and ball valves use a rotary-motion to position the disk or ball into the path of fluid flow. (See Figures 900-7 and 900-8.)

Other Types Numerous types of valve designs are manufactured to meet specific applications. Some of the more common designs are flangeless constructed valves, angle valves, pinch valves, and three-way valves. The angle valve body is essentially the same as a globe body except either the inlet or outlet flow axis is in line with the valve stem. Depending upon the inner valve configuration, flow may be either in the side and out the bottom or vice versa.

April 2002

900-10



Fig. 900-7 Butterfly Valve

900 Control Valves

Fig. 900-8 Ball Valves

Fire-Safe Valves Fire-safe valves are either linear or rotary motion valves which, when engulfed in a fire, will, for a specified period of time: • •

Fail to or maintain a predetermined position Meet a specific leakage rate in the shut off position

If the fire is not controlled in that period, the fire is considered out of control, and valve integrity and performance capabilities become secondary. When fire safety is a concern, valves should be so specified to the valve supplier at the time the valve is ordered. • • • • • • • • •

A number of industry specifications cover fire-safe valves. Among them are the following: API-607 (Onshore installations) API RP-6F (Offshore installations) British Standard Institute (BS) 5146 French Standard AFNOR M98-411 Exxon BP3-14-1 U. S. Coast Guard Factory Mutual 6033 API 589, Fire Test for Evaluation of Valve Stem Packing

These specifications define the type of test that the valves must pass to receive the fire-safe rating. Flangeless Valves. Flangeless valves are valves that are mounted (sandwiched) between flanges with long exposed bolts. Such bolts, when exposed to flames, heat and expand faster than the valve. The differential growth opens up the flanges and allows the contents of the piping to leak out, and, if the material is a hydrocarbon, to feed the flame.


900-11

April 2002

900 Control Valves


Use of metal shields over exposed bolts is not acceptable because it delays valve failure by a very short time. Also, such metal shields are very seldom replaced after maintenance is performed on a valve. ChevronTexaco’s Health, Environment and Safety Group requires that all valves in hydrocarbon service shall either be flanged or shall have body lugs which completely enclose the mounting bolts. Globe Body Valves. Almost all globe body valves, except between-flangemounting valves such as Fisher Style BF, which are installed with long exposed bolts, are suitable for fire-safe applications. Rotary Valves. Rotary valves in hydrocarbon service must be of a type that have body flanges, e.g., Masoneilan Camflex II, or must have lugs that completely enclose the mounting bolts, e.g., Fisher E-Disk, which must be specified with body lugs.

932 Applications Globe Valve Applications Globe valves are available in a variety of body sizes and body ratings ranging from 1-inch to over 24-inches for flanged ANSI Classes 150 to 600 and 1-inch to 8-inches for flanged ANSI Classes 900 to 2500. When purchasing new control valves for high pressure applications, i. e., Class 900 through Class 2500 valve bodies, you should specify non-destructive evaluation (NDE). Typical NDE tests include hydrostatic testing, x-raying to ensure that there are no areas of porosity in the castings, and liquid-penetrant testing. Recommendations when to specify NDE are given in Data Sheet Guide ICM-DG-2350, Instructions for Selecting Sizing, and Specifying Control Valves. NDE requirements are spelled out in detail in ICM-MS-2350, Control Valve Fabrication Specification. Copies of both of these documents are in Part I, Volume 2 of this Manual. Company practice, as well as petroleum industry practice in general, is to standardize on ANSI Class 300 process connections for globe body control valves as a minimum. The cost of ANSI Class 300 and ANSI Class 150 globe body control valves is the same. The only cost difference is the difference in price between two Class 300 and two Class 150 piping flanges. Labor costs to install the valves are also the same. Manufacturers use the same globe valve body castings to manufacture Class 150 through Class 600 globe valves—the only difference is how much metal is removed from each casting. Furthermore, stress analyses performed by several major oil companies show that control valve piping manifolds induce a significant load on globe valve body flanges. The overall conclusion is that, although the Class 300 and Class 150 installations are economically comparable, Class 300 installations are stronger, safer, and more reliable. There are additional benefits from the standpoint of reduced spare parts inventory and interchangeability—a spare Class 300 valve can be installed in either a Class 150 or a Class 300 piping system.

April 2002

900-12



900 Control Valves

Globe valves, in general, can be used in services for nongritty liquids, gases, and steam at moderate to high temperature and pressures. However, the operating variables will determine the type of valve to be selected to handle a specified set of service conditions. Special trims are available when noise and cavitation problems have been identified. Once all the service conditions are known, the valve can be properly engineered. One of the disadvantages of using a globe valve is that as the valve increases in size beyond 4 inches, the cost becomes high compared to a rotary-type, high-recovery valve.

Rotary Valve Applications In valve sizes of 6 inches and greater, ball or butterfly valves may show an economic advantage over globe valves to handle a specified set of process conditions. However, several critical areas must be considered to ensure suitability. The first concern is that because all rotary valves are high-pressure recovery valves, they must be sized to avoid problems of cavitation, noise, and choked flow. Pipe swages also tend to affect this type of valve more by reducing the actual pressure head available for the control valve. The large Cv of a ball or butterfly valve does not necessarily compensate for head loss in pipe reducers. Because of higher velocity at the vena contracta, high-recovery valves tend to cavitate in liquid service and generate more aerodynamic noise than conventional globe valves in vapor service. Evaluation of cavitation and Sound Pressure Level (SPL) is therefore extremely important when high recovery valves are used. Throttling ball valves should be considered for the following services where they are often particularly suitable: •

Wide rangeability is required, such as for fuel oil or gasoline blending.

•

Fluids carrying suspended solids such as coking service, chemical slurries, liquids with catalyst in suspension, etc.

•

Viscous fluids such as asphalt and tars.

•

Pipelines, where you want to minimize valve losses during normal operation.

Characterized-port ball valves offer the best available rangeability. However, these valves are usually offered in the “flangeless” construction, e.g., “Vee-Ball.” Flanged bodies are required to be used for hydrocarbon service. One of the most common types of rotary valves used for control is the butterfly valve. Typical applications are for low or moderate operating pressures and low or moderate pressure drops. Butterfly valves are subject to pressure drop limitations because of the combination of shear and torsional stress on the shaft and disk. The shaft and disk must be designed to withstand the torque that can be produced either at maximum pressure drop or when the valve is only partially open. Rotation of conventional disks should be limited to about 60 degrees from the closed position in throttling service, but 90 degrees rotation is acceptable for on/off


900-13

April 2002

900 Control Valves


service. Rotation of specially designed disks such as Fisher’s “Fishtail Disk” should conform to the manufacturer’s recommendation. Tight closure with conventional metal disk and body is virtually impossible. Elastomer lining of the body will normally provide a suitable seal. However, for extreme service, piston rings or inflatable metal seals are preferred. Usually, any of the sealing methods will result in higher starting torque, and the actuator must be sized accordingly. Also, the temperature limitation of the seal must be taken into consideration. Butterfly control valves can often be used to advantage in normal process applications for valve sizes 6 inches and larger, where valve pressure drops are not excessive. Allowable pressure drop is dependent on shaft material and diameter and on operating variables. Once all the operating variables are known, the allowable torques, noise, cavitation, etc., should be checked to determine whether a butterfly is applicable. Butterfly control valves can also be used to advantage in applications, such as corrosive or erosive service, that require the use of special or exotic metals. Instead of requiring a 100 percent premium body material for these types of services, a butterfly with an inert elastomer coating or liner can be considered. The eccentric disk valves are often called high performance butterfly valves. The less restricted flow results in greater capacity (higher Cv). These valves can also control high-pressure drops with standard construction. The shutoff capabilities of these valves are excellent. They can be used where bidirectional flow is required. However, the flow capacity in the reverse direction is reduced. Torque requirements throughout the operating range are reduced because of the eccentric motion of the disk. These valves are also extremely susceptible to noise and cavitation. All possible operating conditions must be reviewed to make sure they are suitable. (See Figure 900-9.) Fig. 900-9 Eccentric Disk Valve

April 2002

900-14



900 Control Valves

Flangeless valves can be used where fire safety concerns are satisfied. The eccentric spherical plug valve is constructed with single body machining. This allows the valve to be used at ANSI ratings of up to 600 pounds. However, when this type of valve must be installed between pipe reducers, sufficient space must be allowed to install the long make-up bolts. Angle valves are designed for severe pressure reducing services, for example, pressure “letdown” service. If cavitation or flashing is inevitable, an angle valve may be piped to discharge directly into a vessel or other enlarged volume thus lessening valve and outlet piping damage. Angle valves may be used in cases where the piping layout does not allow a globe valve. However, these valves require very careful sizing because their capacity is affected by the direction of flow through the body and also by the type of trim used. For example, a Venturi insert in the valve outlet will result in a higher pressure recovery downstream. When sizing the actuator, the high unbalance forces on the plug must be considered. Angle valves are not a preferred choice, except for special applications as mentioned above. A three-way valve can be used either as a diverting or a mixing valve. Three-way valves are normally provided with linear ports. (See Figure 900-10.) Fig. 900-10Three-Way Valve Application

The use of three-way valves at temperatures above 500°F or at differential temperatures exceeding 300°F is not recommended. The valve, having been installed at ambient conditions and rigidly connected at three flanges, cannot accommodate pipeline expansion caused by process temperature, and distortion results. Similarly, in mixing service, when the temperature difference between the two ports is substantial, the resulting differential expansion can also cause distortion. Distortion in the valve can cause binding, leakage at the seats, dead band, and packing friction. It is ChevronTexaco’s preference to use two separate valves instead of three-way valves because temperature effects are not as critical and because both valves can fail closed (in three-way valves one port is always open). (See Figure 900-11.)


900-15

April 2002

900 Control Valves


Fig. 900-11Preferred Approach Over Three-Way Valves

Globe vs Rotary Valve Comparison For the same nominal size, globe body control valves are bigger, heavier, and are more “rugged” than rotary vaalves. Thus, they weigh more and cost more than rotary valves. Globe valve can withstand higher shutoff pressures and take greater pressure drops than rotary valves. Globe body control valves are less susceptible to cavitation and aerodynamic noise generation. In severe service applications, globe body control valves can be retrofitted with severe service. Masoneilan is promoting their rotary Camflex valve as a “rotary globe” control valve. In reality, Camflex valves are rotary valves. All ball and segmental ball valves have globular bodies but it does not make them globe body control valves.

933 Trim Characteristics Control Valve Characteristics The flow characteristic of a control valve is the relationship between valve opening (stroke) and flow at a constant pressure drop across the valve. Figure 900-12 shows the typical control valve characteristic curve.

Inherent Trim Characteristics Control valve trim characteristics, published by manufacturers, are called inherent characteristics. Values and plots for inherent characteristics are obtained at laboratory conditions where control valve flow capacities are determined, while constant valve inlet and outlet pressures are maintained.

Quick Opening The Quick Opening flow characteristic curve provides for maximum change in flow rate at low valve travels with a fairly linear relationship. Additional increases in valve travel give sharply reduced changes in flow rate and when the valve plug nears the wide open position, the change in flow rate approaches zero. In a control valve, the quick opening valve plug is used primarily for on-off service.

April 2002

900-16



900 Control Valves

Fig. 900-12Valve Characteristic Curves (Inherent)

Linear The Linear flow characteristic curve shows that the flow rate is directly proportional to the valve travel. For instance, at 50 percent of rated valve travel, the flow rate would be 50 percent of maximum flow; and at 80 percent of rated travel, the flow rate would be 80 percent of maximum flow. The linear valve plug is commonly specified for liquid level control and for low gain applications such as vapor pressure control.

Equal Percent In the Equal Percentage flow characteristic curve, equal increments of valve travel produce equal percentage change in flow. The change in flow rate is always proportional to the flow rate just before the change in valve plug position is made. Consequently, when the valve plug is near the seat and flow is small, the change in flow will also be small; and if the flow is large, the change in flow would also be large. Valves with equal percentage flow characteristics are used for high gain applications such as flow and liquid pressure control and for applications where high rangeability (greater than 5:1 is required).

Modified Parabolic Some valves have a characteristic that falls between linear and equal percent. This curve, because it was unique to contoured plugs, is frequently referred to as modified equal percent or modified parabolic.


900-17

April 2002

900 Control Valves


Installed Trim Characteristics When a control valve is installed into a piping system, its actual (installed) characteristics will be different from its inherent characteristics. Installed characteristics refer to the way a valve will actually perform with respect to the changing friction drops in piping and will differ for each piping configuration. To understand how the piping configuration affects the characteristics one needs to develop an accurate hydraulic evaluation of a system including an actual pump curve, best available calculation of piping friction losses, and the published trim characteristic curves or data for the control valve being evaluated (See Figures 900-1 and 900-12). At lower flow rates the pump discharge pressure is near its maximum and piping friction losses are near their minimum. At high flow rates the opposite is true. When the valve is close to its seat, the pressure drop across it is much higher than when it is wide open. Installed characteristic curves can be developed by calculating and plotting flow rates, using actual pressure drops that a valve will see at its various stages of travel. At low openings, where the pressure drops are near the maximum, flow rates through a valve will be proportionately much higher than at greater openings. When these calculated flow rates are plotted against valve travel, the installed characteristic of a valve with an inherent Linear Trim begins to approach the characteristic of a Quick-Open valve. The installed characteristic of a valve with an inherent Equal-Percent Trim begins to approach the inherent characteristic of a Linear Trim valve (See Figures 900-12 and 900-13). This linearity suggests that the gain (response of the process compared to the change in valve travel) of a control valve with inherent equal-percent trim is more likely to produce stable process control. When the gain of a control valve is linear, one set of tuning constants will work throughout all stages of valve travel. When the valve gain is non-linear, different tuning constants would be required at different stages of valve travel. This is especially significant when a facility must process several different feeds or when the process is designed for a wide range of operating cases. When installed characteristics are evaluated, equal-percent trim becomes the trim of choice in over 95% of control valve applications.

Valve Trim Valve trim consists of the internal parts of the valve that are in flowing contact with the controlled fluid and that perform the throttling function. Trim components are usually of a higher metallurgy than the valve body and are removable for replacement during maintenance. Figure 900-14 shows the basic types of trim used with globe valves.

Globe Valve Trim The plug is the moving component of the valve. It throttles flow by positioning itself within the seat orifice and shutting off flow by contacting the seat. The plug is moved against dynamic fluid flow forces by stem force transmitted from the

April 2002

900-18



900 Control Valves

Fig. 900-13Typical Installed Trim Characteristic Curves

Fig. 900-14Typical Globe Body Plugs

actuator. Throttling may be done by a contoured or V-port plug or shaped opening in a cage guide. Guiding is the means by which the plug is aligned throughout plug travel. The guiding must resist all side thrust on the plug from dynamic fluid flow conditions. The common methods of guiding in globe valves are cage guiding, top guiding, top and bottom guiding, and port or skirt guiding.


900-19

April 2002

900 Control Valves


In cage guiding, the cage guides the plug. The outside diameter of the valve plug is in close proximity to the inside wall surface of the cage throughout the travel range. Cage guiding of the plug, combined with uniform distribution of flow around the plug by the cage ports, provides two principal advantages: • •

Stable throttling at higher pressure drops. Reduction in side load and friction.

Top guiding uses a single extra-long bushing to guide the plug stem above the plug and is usually used with a contoured, single-seated plug. Top-guided valves should be avoided for high pressure drop applications (over 100 psi). Top and bottom guiding is used with contoured and V-port plugs with post extensions that are guided above and below the plug. This design allows the plug to be reversed to change the “fail” action on loss of instrument air. Port guiding is also referred to as skirt guiding. The plug is guided in the seat ring by a skirt on the plug. It is used in bodies with V-port plugs and in some three-way valve designs. Port and skirt-guided valves are notorious for failure caused by the locking pin failing and the plug spinning off the stem.

Rotary Valve Trim The trim of rotary valves differs greatly from that of globe valves. The disk, which is in contact with the process fluid in a butterfly valve, can be considered the major component of the trim. The sphere with an internal passageway controls the process fluid media in a ball valve. The sphere configuration can vary from the full ball with a straight bore to segmented balls with special shaped openings or notches. Eccentric mounting of the disk in the eccentric disk valve allows the disk to pull away from the seat, minimizing seat wear. Trim life is prolonged because the disk is not in contact with the seat during throttling.

Rangeability The term “rangeability” is defined as the ratio of maximum controllable flow to minimum controllable flow. Thus a vendor will state that a valve with a characteristic curve will have a rangeability of 50:1. It is impractical, from a manufacturing standard, to characterize flow when the valve is barely open. Likewise, at shutoff it is impossible to characterize any leakage flow that may occur. If minimum flow were used, a valve with bubble-tight shutoff would have infinite rangeability. Such reasoning may be logical, but is unrealistic. The fact that a characteristic curve for an equal percentage valve shows a maximum rangeability of 50:1 does not mean that this is the available rangeability. For instance, ChevronTexaco’s practice is to design control valves to operate between

April 2002

900-20



900 Control Valves

not less than 10 percent and not more than 90 percent of valve opening or stroke. Thus the available rangeability for an equal percent valve will be less than 20:1. Because Cv is an expression of flow capacity, a review of maximum required Cv, minimum required Cv, and the manufacturer’s rated Cv is required in selecting the rangeability needed in a control valve. If this ratio is within the available rangeability of the selected control valve, control of the flowing media is possible. The best rangeability is offered by ball valves with characterized ports, e.g., Fisher “Vee-Ball.” These valves are usually offered in “flangeless” construction. Flanged bodies are required for hydrocarbon service.

Control Valve Seat Leakage Most control valves are throttling continuously, and tight shut-off is not required. Control valve seat leakage becomes a concern in applications where control valves are designed to be normally shut off and to control the process only periodically, e.g., a liquid level control valve on a compressor discharge knockout drum. If the level control valve leaked faster than the liquid was accumulating in the knockout drum, vapor blow-through, which could overpressure the downstream piping and equipment, would result. American National Standards Institute has adopted Fluid Control Institute’s document ANSI/FCI 70-2, Quality Control Standard for Control Valve Seat Leakage, as the national standard for classifying leakage through control valves. FCI 70-2 divides allowable control valve leakage into six classes. These classes are defined as follows: Class I:

No test required.

Class II:

Up to 0.5% of rated control valve capacity at full travel. Tested with air or water at 45-60 psid or maximum operating differential, whichever is lower.

Class III:

Up to 0.1% of rated control valve capacity at full travel. Same test as for Class II.

Class IV:

Up to 0.01% of rated control valve capacity at full travel. Same test as for Class II.

Class V:

0.0005 milliliter per minute of water per inch of port diameter per psi differential. Tested with water at maximum service pressure drop across the valve plug, but not to exceed ANSI body pressure rating.

Class VI:

Bubbles per minute based on nominal port diameter as shown below and tested with air or nitrogen at maximum rated differential pressure across valve plug:


900-21

April 2002

900 Control Valves


Nominal Port Diameter (inches)

Bubbles Per Minute

1

1

1½

2

2

3

3

6

4

11

6

27

8

45

934 Control Valve Sizing Masoneilan Control Valve Sizing Equations In the early 1940s, Masoneilan coined the term “Cv” to refer to a coefficient to denote control valve capacity. Cv is equivalent to the number of gallons per minute of water that will pass through a control valve at a pressure drop of 1 psi. Basically, it is a capacity index. For example, a control valve that has a flow coefficient or Cv of 10 has an effective port area (in the fully open position) that will pass 10 gpm of water at a 1 psi pressure drop. Control valve capacity ratings are generally determined by tests. The test method consists of measuring the flow through the valve, at a constant pressure differential across the valve, and at various degrees of valve opening. The basic liquid sizing equation that Masoneilan developed was: Q Cv = ------------------------------1/2 P – 1 P 2  ---------------- Sp. Gr.  (Eq. 900-1)

Cv = Valve sizing coefficient Q = Liquid flow in gpm P1 = Upstream pressure P2 = Downstream pressure Sp. Gr. = Specific gravity of the fluid Masoneilan then extrapolated the liquid sizing equations to vapor and mass flow, and a method to size all control valves was established. These equations could not predict incipient cavitation in liquid service and were ineffective in the transition regime of vapor flow from subsonic to sonic.

April 2002

900-22



900 Control Valves

Masoneilan equations did not include sizing corrections for viscosity, supercompressibility, piping geometry, etc., built into them.

Fisher Control Valve Sizing Equations Because of the problems encountered in using Cv to predict critical flow, Fisher Controls began testing valves on air and water at their own test facility. On the basis of these test results, Fisher established a sizing coefficient, Cg, to relate critical flow to the absolute inlet pressure for gases. Cg was experimentally determined by testing each style and size of valve and showing that the predicted flow at critical pressure drop was in acceptable tolerances with the curves generated from the test data. In 1963, Fisher published their Universal Valve Sizing Equations for volumetric and mass vapor flow. The equations established vapor sizing coefficients in units of Cg and steam sizing coefficients in units of Cs. Fisher equations gave excellent results in calculating compressible fluid capacities, but, unfortunately, could be used only for specifying Fisher valves. The coefficients Cg and Cs are meaningless to all other control valve manufacturers. Another drawback of the Fisher equations is that they expect the user to know when sizing corrections, e. g., compressibility, piping geometry effects (reducers), viscosity, etc., are needed. The process for applying the corrections is cumbersome. The user must make a preliminary valve selection based on initial sizing, decide that the application requires the corrections, calculate the corrections separately from the sizing calculations, adjust the calculated sizing values, then make a final valve selection. Not applying the corrections may result in a control valve being either oversized or undersized. Fisher’s current software gives the user the option to perform calculations using either ISA equations or Fisher’s proprietary coefficients.

ISA Control Valve Sizing Equations In the mid-1970’s, the Instrument Society of America (ISA), with cooperation from Fisher, developed control valve sizing equations that incorporated the benefits of the Fisher equations, but used the universal valve sizing coefficient Cv and were applicable to sizing all control valves. The ISA equations in their basic form list all the possible corrections that must be considered in sizing a control valve, e.g., supercompressibility, viscosity, swage effect of inlet and outlet piping, etc. The corrections are included so that an engineer does not forget a possible consideration, with the result that the valve is improperly sized. ChevronTexaco experience shows that control valves can be undersized by as much as twenty five percent (25%), expecially rotary valves, when swage effects of piping reducers, viscosity, etc., are not taken into consideration. The ISA equations are recommended and endorsed by the Company.


900-23

April 2002

900 Control Valves


Effects of Piping Geometry on Control Valve Capacity Typically, control valve body sizes come out to be one size smaller than the piping in which they are installed. In globe body control valves trim can be furhter reduced to accommodate process requirements. In rotary control valves, reduced trim is not an option and a smaller valve must be specified for the application. In such cases, effects of piping reducers and process hydraulics on valve capacity must be carefully evaluated. ChevronTexaco’s experience shows that use of piping reducers can cut down the rated capacity of a rotary control valve by as much as twenty five percent (25%).

Personal Computer Software A number of personal computer (PC) programs are available for sizing control valves. These programs are offered by publishing companies, ISA, and most of the major control valve vendors. To date, most of the programs are written to run on the IBM PC’s or their compatibles. Limited software is available for the Apple Macintosh computers. The early computer programs were based on Fisher equations. They gave results in Fisher sizing coefficients, i.e., Cg and Cs, and did not include sizing corrections. Most of the newer programs are based on ISA equations and have subroutines that calculate and apply all corrections automatically. The best programs allow the user to review installed trim characteristics, and, based on the type of valve type and trim selected, to evaluate control valve gain. This feature allows the user to select the optimum valve type, valve body size, and trim size for the application.

935 Material Selection Valve Bodies Valve body materials are selected based on the process fluid and its properties, i.e., type of fluid, pressure, temperature, corrosion, and erosion properties. Valve body material is usually specified the same as process piping material, e.g., carbon steel, 316 stainless steel, etc. Pressure and temperature ratings for the pressure containment parts, i.e., body, bonnet, and bolting have been established for the more common materials by the American National Standards Institute (ANSI). The recommended materials for specified physical and chemical requirements are given in standards established by the American Society for Testing and Materials (ASTM). Pages 297 through 315 of the ISA Handbook of Control Valves, 2nd edition, list extracted tables from ANSI and ASTM standards. It is suggested that for accuracy or items not listed, the specific ANSI or ASTM standard be consulted. It is good practice, when special materials are recommended to confirm the selection with ERTC Materials specialists, and then specify the materials by the proper ASTM designation.

April 2002

900-24



900 Control Valves

Valve Trim The control valve trim material selection is generally predicated on the factors of corrosion, erosion, wear and galling, and temperature. Corrosion properties of the process fluid are important factors in material selection. Tables are available that list materials and chemicals but give only a general indication of how a limited selection of materials will react when in contact with certain fluids. These tables should be used only as a guide. For example, the table may show that it is acceptable to use carbon steel for chlorine gas, but titanium is not acceptable. While this is true for dry gas, the opposite is true if the gas is wet. Concentration, temperature, contaminants, pressure, and other conditions may alter the suitability of a particular material. Erosion is damage caused by the impingement of high velocity particles on the material surfaces. Entrained sand, slurries, catalyst fines, and wet steam liquid droplets are sometimes associated with this type of wear. The valve plug and seat ring are probably the valve components that are most susceptible to erosion. The ISA Handbook of Control Valves, 2nd edition, pages 162 through 164, discusses the “properties and applications of common trim materials.” This listing may be used as a guide in the selection of materials for corrosion or erosion services. Galling is related to temperature, material pairs, surface finish, hardness, and loading. The lubricating qualities of the flowing fluid also affect wear and galling. Dry superheated steam has a notable tendency to cause this type of failure. In applications where high mass flow rates exist, wear and galling of valve plug and bushing materials are usually encountered. High temperature can either anneal or soften metals, increasing the galling potential. The differential thermal expansion of trim components can act to reduce the working clearances designed into the valve. This can occur when hot fluid initially enters a cold valve and not all trim components come up to temperature at the same rate. Galling may be lessened or prevented by the following: •

Specifying hard material for one, or preferably both, parts

•

Using different materials for both parts

•

Specifying either a required surface finish and surface hardness or special coating

•

Selecting material pairs with low galling potential

Packing Gland Assemblies Packings, installed into a packing gland assembly, prevent leakage of process fluid past the stem or shaft. (Also see the next section for hydrocarbon fugitive emissions requirements.) The packing gland assembly consists of a packing box machined into the valve bonnet, a wiper ring, a lantern ring, packing, a packing follower, a packing flange, and stud bolts and nuts. The wiper ring sits at the bottom of the packing box


900-25

April 2002

900 Control Valves


and provides mechanical support for the lantern ring. The lantern ring is a metallic spacer that provides mechanical support for the packing and also provides a cavity to receive the lubricant when lubricant is used. The packing flange contacts the follower which has a ball-shaped surface to compensate for unintentional misalignment of the packing flange thus minimizing side loading on the stem. Adjustment to nuts on packing gland studs compress the packing against the packing box walls and the surface of the stem. A loosely adjusted packing will allow process fluid to leak past the packing. Overtightened packing will “grab” the stem and will keep it from moving. Some packing gland designs use two sets of packings separated by the lantern ring. In these designs, the lantern ring cavity can be used either to accept the lubricant, or, in a dry packing system, to monitor intermediate packing leakage through the lubricator and isolating valve connection. All manufacturers specify that a torque wrench be used to adjust the stud bolt nuts to ensure that packing is tightened properly.

Packings Packing recommendations for specific petrochemical services are tabulated in Specification ICM-EG-2350, Control Valve Fabrication. Historical Overview. For control valves, past practices have been to use asbestos composition “jam-type” packings, where a packing was installed into a packing gland, the stud nuts tightened, and the resiliency of the packing provided the sealing forces needed to keep the process from leaking. To reduce packing friction and to optimize sealing, lubricants were injected into the lantern ring area. Lubricants sealed the voids in the resilient packing as well as filled in mechanical voids between the packing and the stem. Most petrochemical facilities used Teflon packings, which consisted of Teflon formed into either solid rings, split rings, or “chevron vee rings.” Teflon Chevron rings relied on the mechanical shape of the rings to provide the needed lateral loading to maintain the seal. In the late 1980’s use of asbestos was banned by the EPA. At about the same time, Teflon was banned from hydrocarbon services because it is not fire-safe. In 1990, the Clean Air Act was passed and strict fugitive emission sealing performance requirements were mandated. Current Requirements. In today’s business climate, control valve packings must be fire-safe, mechanically reliable, chemically inert, must have high sealing capability, and must produce minimal friction on the stem. To meet these criteria, control valve packing material options have been narrowed to fluorocarbons (Teflon or Kal-Rez), graphitic packings, and combination packings (combination of fluorocarbon and graphite). Fluorocarbon Packings. Fluorocarbon packings are chemically inert in most process applications, provide outstanding sealing characteristics and have a very low coefficient of friction. However, fluorocarbon packings are not fire-safe and are also temperature limited. Teflon sublimates at about 400°F, Kal-Rez decomposes at about 500°F. They are the optimum choice for most chemical and non-hydrocarbon services. Teflon packings come in solid, split-ring, and chevron form. Kal-Rez comes in chevron form only. Chevron packings are usually installed with

April 2002

900-26



900 Control Valves

anti-extrusion back-up rings to preclude packing degradation through cold flow. Fluorocarbon packings can only be used for hydrocarbon services in the “combination” form described below. Fluorocarbon packings can be installed in “jam-type” packing glands; however, for optimum performance (long intervals between maintenance while complying with low fugitive emission limits) they should be installed in “live-loaded” packing gland assemblies. Live loading is produced by belleville spring washers which deliver a constant and uniform load on the packing. Graphitic Packings. Graphitic packings (graphite or carbon) come in pre-formed rings that slip over the valve stem or in braided ribbon form. Graphitic packings are relatively inert and can be used for most hydrocarbon services. Some die-formed graphite rings have a temperature rating of up to 1,000°F. Braided ribbon packings are easy to handle but are difficult to install correctly. Proper installation procedures require that graphitic packings be installed one layer at a time with each layer being “crushed” to eliminate all voids. When multiple layers are installed without the intermediate “crushing,” only the top layer gets compressed properly. Voids remain in the uncompressed layers and packings start to leak shortly after installation. Graphitic packings, once compressed, do not have the property to “spring back” when compressive forces are released (or when packing material volume is reduced through wear). Experience shows that properly installed graphitic packings perform best when they are live-loaded. Graphitic packings have a coefficient of friction that is many times higher (up to thirty times higher) than fluorocarbon packings. To overcome packing friction forces, control valves with graphitic packings require larger actuators than valves with fluorocarbon packings. Valves which initially had Teflon packings but were retrofitted to graphitic packings (late 1980s) typically ended up with undersized actuators. Several Company facilities tried using braided graphite packings with lubricants to both reduce friction and to “seal” the voids. Experience shows that lubricated braided graphitic packings give borderline performance and require high maintenance. The lubricant must be replenished frequently while friction is not reduced enough to provide good control valve performance. Combination Packings. Combination packings, which were introduced in early to mid-90s are most efficient, and in the long run most economical. Within their temperature limits they satisfy all the criteria that a packing must meet. Combination packings consist of fluorocarbon chevron rings supplemented with solid graphite rings. Fluorocarbon rings provide low friction and excellent sealing while graphitic rings satisfy the fire-safety requirements by restricting hydrocarbon releases if the fluorocarbon packings “burn.” Combination packings also require live-loading to work effectively.

Lubricants Modern control valve packing systems are designed for long-term, maintenance-free operation and are designed to work without lubricants. Periodic injection of packing lubricant is labor intensive, drives up maintenance costs, and should be considered only as a “last resort,” e.g., in a situation where a packing has developed a leak and the valve cannot be taken out of service (no bypass manifold) until the plant is shut down for maintenance.


900-27

April 2002

900 Control Valves


Figure 900-15 shows the installation of a lubricating and isolation valve. However, due to restrictions on fugitive emissions, control valve packing systems have improved to the point where lube and isolation valves are normally not used. Fig. 900-15Lubrication and Isolating Valve

Sealing Systems for Fugitive Emission Compliance The 1990 Clean Air Act imposes strict regulations requiring elimination of fugitive emissions. The Act requires that hydrocarbon fugitive emissions are to be regularly monitored using a pre-approved Volatile Organic Compound (VOC) analyzer that measures hydrocarbon emissions in parts per million (ppm) of methane. Excessive emissions are categorized as leaks. The leaks must be promptly reported and repaired. Fines can be imposed for either excessive or unreported leaks (leaks found during unannounced inspections conducted by either the Environmental Protection Agency (EPA) or local air quality management district representatives). For volatile emissions the Clean Air Act identifies over 180 hydrocarbon components and compounds that comprise VOCs which require monitoring. Current VOC limits are 500 ppm for chemical and petrochemical plants and 1,000 ppm for refineries. Components that are classified as toxic (hazardous air pollutants or HAPs) are limited to 500 ppm even if they are in a refinery process, e.g., benzene. Local agencies are given the option to impose regulations even stricter than those required by the Clean Air Act. In the State of California these options are being fully implemented by the two dominant air quality management districts, the South Coast Air Quality Management (SCAQM) District in the Greater Los Angeles area and the Bay Area Air Quality Management (BAAQM) District in the Greater San Francisco Bay Area. SCAQMD currently imposes a 1,000 ppm VOC limit. This limit is supposed to go down to 500 ppm in 1997. BAAQMD currently imposes a 500 ppm limit. BAAQMD’s limit is scheduled to go down to 100 ppm in January 1997.

April 2002

900-28



900 Control Valves

Valve manufacturers are working feverishly to develop valve packing systems that will comply with fugitive emission regulations. The valve manufacturers’ dilemma is that for a valve to control efficiently, packing friction must be minimized. High packing friction contributes to hysteresis, which is not acceptable in control valves. To minimize packing friction, valve packing compression forces have to be minimized. Reducing valve packing compression reduces the effectiveness of the valve packing and allows VOCs to escape. Teflon packing minimizes packing friction but by itself is not acceptable to ChevronTexaco from a fire safety standpoint. Graphite packings are satisfactory for fire safety but impose large friction forces on the control valve stems and may meet the fugitive emissions requirements. In 1993 the Fire Protection Staff approved the use of the combination of graphite and teflon-type rings. Lubricants and grease seals are being used in some ChevronTexaco locations as an interim solution. In the hydrocarbon industry, lubricators and lubricants were always furnished on control valve packing glands. “Grease seals” were once the only acceptable method to control unacceptable H2S and mercaptan leakage from control valve packings. In the long term, grease seals will be considered acceptable only for emergency repair and non-lubricated systems will be required. Valve and packing manufacturers have come up with the “dry” packing combinations identified in the section on “Packing” above, which meet the fugitive emission requirements. One constraint for maintaining effective fugitive emissions control is stem and packing box condition and finish. For fugitive emission applications, Fisher-Rosemount furnished their valves with stem finishes of 4 microinches rms. Packing box finishes can be somewhat rougher because the packing does not move in a packing box. As mentioned earlier, live-loading is a must for effective fugitive emissions control valve application. As part of the Petroleum Environmental Research Forum (PERF), the Company has been monitoring packing fugitive emissions performance at several Company facilities. Test data shows that use of lubricants and grease seals to meet VOC fugitive emission compliance is only marginally successful. Lubricated valves are barely able to meet the 500 ppm emissions limit while they require frequent packing gland adjustments and lubricant injection. Also, lubricating graphitic packings does not seem to reduce the high friction that is normally associated with dry graphitic packings. Because of high friction, control performance of valves with lubricated graphitic packings has not been entirely acceptable. By comparison, live-loaded “dry” packing systems that the Company is monitoring for fugitive emissions have been delivering better than 50 ppm sealing performance (<20 ppm average emissions) for over a year without any maintenance or adjustments. Because properly sized actuators have been furnished for these “dry” packings, control valve performance has also been outstanding.

Packing Performance Standards In 1994, Fluid Control Institute published ANSI/FCI 91-1-1994, Standard for Qualification of Control Valve Stem Seals to Meet EPA Emission Guidelines for Volatile Organic Compounds. This standard defines and classifies performance levels that


900-29

April 2002

900 Control Valves


control valve packing systems must meet and requires the manufacturer to certify that his packing system will meet the specified performance level. For example, Class A1 defines a packing system which will not exceed 100 ppm emissions over a period of 100,000 mechanical cycles and three thermal cycles (60°F to 450° to 60°F = one thermal cycle). ISA is in the process of publishing a similar standard for rotary control valve packing systems.

Hermetically Sealed Valves Control valves with bellows seals fall into the category of “hermetically” sealed control valves. They provide zero leakage as long as the bellows are intact. Bellows seals have a history of work hardening, fatiguing, and failing. Bellows seals are usually furnished with Teflon back-up packing systems. Although manufacturers have changed bellows seal design and construction to extend bellows seal life, the Company’s most recent experience was that a “new design” bellows seal failed within four months after being placed in service. When a bellows seal fails the entire bonnet assembly must be replaced as a unit. Replacement bellows seals cost about six times more than live-loaded, fire-safe fugitive emission-type packings. One control valve company advertises their valves as hermetically-sealed “zeroleakage” control valves for refinery services. The valves were developed for distilled water control on nuclear powered submarines, and, with cutbacks in military spending, the manufacturer had to look for new markets. It picked the petrochemical industry because the petrochemical industry is a major user of control valves. Four valves were installed successfully in ChevronTexaco to control clean (filtered and polished) products in a pipeline application where reliable instrument air was not available. Although the sales representatives claim that the valve is the answer to fugitive emission concerns in all petroleum services, the company’s technical manager says that the valves are susceptible to plugging in over 75% of petroleum processing applications. The valves are actuated by means of a large solenoid coil which opens or closes a pilot port inside the control valve body. To open or close the valve, the pilot diverts process pressure to the upstream or downstream side of the plug. Valve position is fed back to remote control electronics through a Linear Variable Differential Transformer (LVDT). For maintenance considerations, the electronics must be mounted remotely in an environment that is protected from rain and electrically classified as General Purpose. Failure mode of these valves is indeterminable because of many parameters that can fail, e. g., power failure, electronics failure, loose wiring, loss of process pressure, etc., each of which would make the valve fail in a different direction. This would be a safety concern because a control valve must fail in a predetermined, fail-safe direction. Because of different signal levels (power to coil and LVDT feedback), two conduits must be run from the remote electronics to each control valve. As a result, the total initial installed cost of these valves can be much higher than the cost of conventional control valves. There is no operating and maintenance cost data available to establish the total life cycle cost of these control valves.

April 2002

900-30



900 Control Valves

Specification ICM-MS-2350 This specification defines ChevronTexaco’s minimum requirements for body metallurgy, fabrication, and packing and lubricator selection of all valves used for control except gate valves and direct-operated or self-contained regulators. Air-operated diaphragm actuators, piston actuators, and control valve positioning devices are also covered by this specification. The requirements, as applicable, for electric-motor operated, hydraulically operated and electrohydraulically operated control valves are discussed. Any fabrication feature that deviates from this specification must be approved by ChevronTexaco.

940

Actuators 941 Spring-and-Diaphragm Actuators The spring-and-diaphragm actuator is the oldest of all control valve actuator designs and is still the most widely used. It is simple, inexpensive, fairly efficient; and its speed is limited only by the rate at which air can be moved into and exhausted from the diaphragm case. In a direct-acting diaphragm actuator, the pressure-tight chamber is above the diaphragm and increasing pressure within that chamber, compresses the spring, and results in a downward motion of the stem. A direct-acting diaphragm actuator is shown in Figure 900-16. Fig. 900-16Typical Direct-Acting Diaphragm Actuator

In a reverse-acting actuator, the pressure-tight chamber is below the diaphragm and increasing air pressure with that chamber results in an upward motion of the stem.


900-31

April 2002

900 Control Valves


On loss of diaphragm air pressure, the valve returns to its designed failure mode: that is, the direct-acting actuator fails up, the reverse fails down. The limitation of the spring-and-diaphragm actuator is that a large diaphragm is required to produce a large stem force, and the larger the diaphragm, the less pressure the case can withstand.

Piston Actuators Piston actuators are used when spring-and-diaphragm actuator limits are exceeded. Single-acting and double-acting piston actuators are the two most common types. In a single-acting piston, the positioner drives the piston against a spring (as in a diaphragm actuator). On loss of air supply pressure, the spring moves the piston and, in turn, the control valve plug, to the fail-safe position. In double-acting actuator, air pressure is used to drive the actuator piston in both directions. These actuators react to a pressure unbalance that is created by loading supply pressure on one side of the piston and unloading the opposite side. Figure 900-17 shows a double-acting piston schematic. Fig. 900-17Operation of Double Acting Piston Actuator with Positioner (Courtesy of Fisher Controls, International, Inc.)

April 2002

900-32



900 Control Valves

Assume that a downward piston motion is required and the instrument bellows receives a corresponding change in input signal pressure. This causes the beam to pivot, covering the nozzle on relay A and increasing its output pressure. At the same time, relay B reacts to the change in beam position, decreasing the pressure to the underside of the piston. This action results in unbalanced forces acting on the piston. Because of this unbalance, the piston moves down changing the plug position. A similar analysis can be made for an upward piston movement. A feedback arrangement prevents over correction and ensures a definite position of the piston and valve plug for a given instrument signal. This is accomplished by a range spring that feeds the position of the piston back to the beam.

Selection An actuator should be selected to overcome the static and dynamic forces associated with the control valve. It should be able to move the valve plug to a specific position and maintain that position, within a specified tolerance, despite the varying forces exerted by the flowing fluid. In other words, the actuator should have power, stiffness, and frequency response qualities suitable for the application. The static unbalance of the valve created at shutoff is a major force. On singleported, unbalanced globe valves, this force is calculated by multiplying the area of the valve orifice by the maximum shutoff pressure. On “balanced” valves, the force is calculated by multiplying the cross-sectional area of the stem by the maximum shutoff pressure. Friction forces also are present in the valve body. These forces are a result of the packing, seals, and guiding surfaces. For large valve body assemblies, the weight of the valve plug and stem forces should be also considered. The static forces of rotary-motion valves are measured in torque units (in-lb) rather than force units. Ball valves and other symmetrical valves have negligible unbalance static forces; eccentric or offset valves do have unbalance forces. Ball valves and plug valves often are designed so that the closure member is always in contact with the seat throughout the travel. These valves have extremely high frictional forces at all openings. Other rotary valves, such as some offset high-performance butterfly valves and eccentric plug valves, incur sealing friction only at small angles of opening. The torque required to cope with this friction is known as “breakaway” or “breakout” torque. Special consideration must be given when specifying an actuator for a ball valve. The standard valve has a circumferential ring seat that the upstream pressure seals against the ball. This situation can generate substantial friction and require large operator torque for smooth operation. These valves are also available without seat rings, which eases the load on the actuator but eliminates the tight shutoff feature. Between these two extremes is a large spectrum of valve types that have nonlinear seal friction characteristics. For example, the standard elastomer-lined butterfly valve has a high friction interference fit, but the friction diminishes as the valve opens.


900-33

April 2002

900 Control Valves


Static unbalance is by no means the only process influence on the valve. As the valves throttles, dynamic forces act on the plug because of the fluid momentum. These forces vary in magnitude and direction, depending on valve style. In all cases, however, flow forces are proportional to the pressure differential across the valve. The valve plug stroke is determined by the control valve body size or the valve plug size to give full controllable flow. In a pneumatic-diaphragm actuator, stem travels of 0.5 to 4 inches are typical. Stem travel for the piston type actuator can vary from ¾ of an inch to 24 inches. Diaphragm actuators normally operate with air pressure ranges of 3 to 15 psig or 6 to 30 psig. When thrust and torque requirements exceed the range of diaphragm actuators, pneumatic pistons are used. Piston actuators operate on air pressures from 50 to 150 psig. These actuators can be used in either throttling or on-off applications. For applications where fast stroking speeds, high thrusts, and long strokes are required, electrohydraulic actuators can be used. A 4 to 20 ma dc signal to an electrohydraulic transducer will operate these actuators. This type of actuator will operate an assortment of large rotary and sliding stem valves. Failure-modes of actuators are specified after considering the type of process being controlled. The process may require one valve to fail-open to relieve the pressure build-up of a system and another valve in the same system to fail-close to retain a particular level. The correct selection of the failure-modes of control valves in a process unit is essential for reasons of safety. The three basic valve actuator failure-modes are fail open, fail closed, and lock-up or fail in place (hold the last position). In the diaphragm actuator, on loss of air pressure, the valve will fail open or closed depending on the spring and diaphragm arrangement. Pneumatic lock-up systems are used with piston actuators to lock in the existing actuator loading pressure in the event air supply is lost. Normal operation resumes automatically when supply pressure is restored. A piston actuator “fail-safe” system requires a volume tank that is charged with instrument air header pressure. The volume tank provides loading pressure for the actuator piston when supply pressure fails, thus driving the piston to the desired position. For actuator fire-safety requirements and fail-safe operation during a fire see Section 912.

942 Actuator Sizing Improper diaphragm actuator sizing created many problems for the Company in the past. As previously mentioned, valve vendors often try to reduce the cost of a valve by providing the smallest size actuator that they can get away with. Vendors claim that a smaller actuator with a weak spring optimizes control valve response speed, which it does. What they don’t say is that the small actuator is incapable of overcoming transient or inadequately defined dynamic fluctuations in the

April 2002

900-34



900 Control Valves

process. With small actuators, even minor process pressure fluctuations will overpower the actuator and drive the valve away from the desired control position. In gross misapplications, process pressure has driven control valves wide open, creating severe process upsets and unsafe conditions. Until about 1978, Fisher Controls was extremely conservative in sizing control valve actuators. Since then, Fisher has lowered their standards to compete with other valve manufacturers. Presently, all vendors use essentially the same actuator sizing criteria and equations. The International Society of Measurement and Control (ISA) is in the process of publishing a standard which will establish a uniform approach to actuator sizing. Until the standard is published, a “rule of thumb” is to limit the “jack-up” or bench set of a control valve to no more than one-fourth (1/4) of the diaphragm signal range, i. e., 3 psi (6 to 15 psig or 3 to 12 psig) for a 3 to 15 psig actuator or 6 psi (12 to 30 psig or 6 to 24 psig) for a 6 to 30 psig actuator. Some facilities specify ANSI/FCI Class IV seat leakage requirement. Both of these requirements force the manufacturer to supply a bigger actuator with a stiffer spring.

950

Control Valve Accessories A number of devices may be used as accessories with control valves. Some of them are mounted integrally with the valve or actuator, some must be mounted on the actuator assembly, and others may be located adjacent to the valve.

Control Valve Bonnets Control valve bonnets are an integral part of most types of control valves and are not truly accessories, but there are options that must be considered. A bonnet is the transition component between the valve body which contains the packing gland assembly, and the actuator. Bonnets are available with either screwed (union) or bolted bonnet construction. Screwed bonnets have a history of working loose so bolted bonnets should be specified whenever a choice is available. “Extension” or “radiation fin” bonnets protect the packing and the actuator from the effects of extreme process temperature. These bonnets should be specified for process temperatures below 32°F or above 450°F.

Valve Positioners A valve positioner is a device which ensures that the control valve moves to the position which corresponds to the signal from the controller. A positioner receives the signal from the controller and compares it with the valve stem position. If the stem position does not match the control signal, the positioner will change the output to the valve actuator until the correct valve stem position is obtained. Positioner output can be pneumatic or hydraulic depending on actuator design. Positioners for electrohydraulic actuators are assembled integrally with the hydraulic system and will not be discussed here. A control valve positioner can be looked at as a cascade controller. Its primary control loop is the controller that positions the stem. The secondary control loop is


900-35

April 2002

900 Control Valves


the set point (control signal) from the process controller. The system works well as long as the primary loop responds faster than the secondary loop. If the process can change faster than the positioner can respond, the loop will become unstable and control will be impossible to achieve. There are several different types of valve positioners. A control valve with a pneumatic diaphragm actuator will usually have the positioner mounted on the actuator support yoke. Mechanical linkage connects the positioner to the stem and tubing is used to connect the positioner output to the diaphragm. Depending on the design of the pneumatic piston actuator, the positioner may be mounted either on the yoke or be mounted integrally with the piston. On an integral positioner, stem feedback is provided internally through the piston actuator. At least one pneumatic connection is made internally directly into the piston. On double acting pistons, the second connection is made with external tubing. Except for electrohydraulic actuators, the output from control valve positioners is always pneumatic, although the control signal may be pneumatic or electronic. All major control valve manufacturers offer both pneumatic and electropneumatic valve positioners. Both types are acceptable and reliable and should be selected to match the type of control system used.

Pneumatic Positioners Prior to electronic control systems, pneumatic control valve positioners were specified only on valves that were larger than 2-inches and on valves that were operating at a pressure drop greater than 200 psi (See Figure 900-18). With electronic control systems, electropneumatic positioners are specified for all control valves. Fig. 900-18Pneumatic Valve Positioner Selection Guide

April 2002

900-36



900 Control Valves

Electro-Pneumatic Converters When electronic control systems became prominent, the demand for electropneumatic positioners increased. Early electropneumatic positioners were nothing more than electropneumatic transducers installed within or attached externally to conventional pneumatic positioners. Because the components were not specifically designed to work together, the outputs were not always stable. As a result, many ChevronTexaco Operating Companies were dissatisfied with electropneumatic positioners and preferred to use pneumatic positioners only. Electronic signals were converted to pneumatic signals through separate remotely mounted electropneumatic transducers. Separation of craft labor was another reason for using electro-pneumatic converters with pneumatic positioners. Unions allowed only electricians to work on electro-pneumatic converters while only pipe fitters were allowed to work on the pneumatic positioners.

Electro-Pneumatic Positioners Present electropneumatic positioners are sufficiently fast and reliable to the point where they should be specified for all control valves driven by electronic signals. Newer smart electropneumatic positioners offer built-in features such as valve stem position feedback and control valve diagnostics. Such features were never available with pneumatic positioners.

Smart Positioners Several manufacturers offer control valves with “smart” positioners. Each manufacturer offers a different set of features. Some of the offered features include: •

Communicate with a remote calibrator or with a distributed control system either in digital mode, analog mode, or in both modes simultaneously.

•

Can be calibrated or re-configured from a remote location.

•

Store calibration and maintenance history as well as valve operating history (number of strokes or cycle reversals to determine wear on stem, packing, trim, etc.). This information can be retrieved in digital mode while the positioner is communicating with its controller in analog mode. The information can be used for “troubleshooting” (“diagnosing”) control valve condition or process related problems.

•

Perform various levels of PID (proportional, integral and derivative) control.

•

Compatible with fieldbus communications.

Company experience shows that the incremental initial cost of these smart positioners is recovered the first time that a contractor calibrates/loop-checks a control loop. Smart positioners are a rapidly evolving technology with new features being added every day. Fisher-Rosemount’s FIELDVUE positioner is the most widely used smart positioner in the Company with over a thousand units in service in Upstream, Chem-


900-37

April 2002

900 Control Valves


ical, and Refining operations. Also see “Control Valve Disgnostics” below. In addition to most of the features described above, it offers many of the features and benefits available from the Flowscanner in-line diagnostic tool. As the control industry is preparing to convert to the Fieldbus Foundation and ANSI/ISA SP50 Fieldbus protocols, more features are being built into smart positioners. Some smart positioners already offer three mode (PID) controller functions which are built into the positioner electronics. Note One manufacturer offers a valve where process sensors are built into the control valve body (smart valve). For certain applications, this valve can conceivably be used as a stand-alone process sensing element, controller, and final control element — all in one control valve. The Company has limited experience with this valve. One concern is that there is no method to isolate the process sensors from the process in case any of them fail or require maintenance.

Control Valve Diagnostics Several manufacturers, e. g., Fisher-Rosemount, Valtek, Masoneilan, Neles-Jamesbury, etc., have either introduced or announced that they are developing diagnostic tools to evaluate control valve condition or performance without taking the valve out of the line. Fisher-Rosemount introduced their Flowscanner in early 1993. Use of the Flowscanner demonstrated that more than half of “problem” control valves can be repaired or recalibrated in-place without taking the valves out of the line. In the process of using the Flowscanner, Fisher found that the traditional process of adjusting the bench set on control valves was inaccurate. The bench set procedure should be performed with the actuator stem disconnected from the valve stem. ChevronTexaco started using the Flowscanner in 1994, about one-year after it was introduced to the petrochemical industry. Metrics from the first two facilities that used the Flowscanner prior to shutdowns show that it cut control valve maintenance costs. The Flowscanner is now being used by all ChevronTexaco operating companies. The Fisher-Rosemount FIELDVUE smart positioner has recently been upgraded to provide most of Flowscanner disgnostics capabilities.

Positioner Application Guidelines Pneumatic positioners should only be used with pneumatic field controllers and with older pneumatic control systems which have not been upgraded to electronics. In a pneumatic positioner, the position control mechanism can be either force balance or motion balance. In each type, a bellows receives a signal from a controller and repositions the flapper on a pilot valve. This change in flapper position changes the output to the actuator and drives the valve stem proportionally to the change in the controller signal. In motion balance positioners, the bellows moves the flapper pivot point. The change in stem position moves the flapper in the direction opposite to the pivot point returning the pilot valve to equilibrium. In force balance positioners the

April 2002

900-38



900 Control Valves

change in bellows pressure produces a force against the flapper. The flapper is pivoted near the pilot valve and is counterbalanced by a spring which is connected to the stem. To produce equilibrium, the change in stem position exerts a force, through a spring, which drives the flapper back to its original position. Because minimal flapper movement will affect the positioner output, force balance positioners tend to be extremely unstable and can be affected by outside disturbances such as pipe vibration, etc. Force balance positioners should be avoided whenever possible (See Figure 900-19). Both pneumatic and electropneumatic positioners can be split-ranged. Split ranging refers to the configuration where an initial portion of the control signal drives one valve and the remaining portion of the control valve signal drives a different valve, e. g., one positioner operates a valve over a 3 to 9 psig portion of the controller output while another positioner operates a second valve over 9 to 15 psig. Fig. 900-19Valve Positioner Types (Courtesy of Fisher Controls International, Inc.)

With pneumatic controllers, split-ranging is the only available option when more than one control valve must be driven by the same controller. With electronic control systems, although the positioners could be split-ranged, the preferred practice is to dedicate separate controller outputs to drive valves which would have previously been split-ranged. The “split-ranging” is done as part of the control loop configuration. Better control can be achieved because the valves operate over the full range of the control signal and because unique tuning constants can be developed based on process response (gain) over the range of each “split-ranged” valve.


900-39

April 2002

900 Control Valves


The output of some positioners can be characterized to produce a response different from the controller output. This is done by changing the shape of a position feedback cam within the positioner. The linkage from the stem moves the cam and produces alters the stem position feedback. Cams have been used to change the response of a linear valve to that of an equal percent valve, to split range valves, or to provide characterized response to correspond to process characteristics, e. g., pH control response curve. Cams are not used as frequently with electropneumatic positioners because it is more practical to characterize the control signal within the electronics by modifying the control algorithm. With older electronic control systems, e. g. older board-mounted electronic controllers and Foxboro Spec 200, cams have to be used if one wants to characterize the response of a control valve. Pneumatic positioners used in a pneumatic control loop may be detrimental to the quality of control in fast processes such as liquid pressure and flow processes. Sheldon G. Lloyd, Fisher Controls, International, in his paper “Guidelines for the Use of Valve Positioners and Booster Amplifiers,” comments on the use of positioners in fast systems. He considers liquid pressure control, most gas pressure applications and most flow processes in this category. He states: “While positioners can greatly reduce the apparent actuator dead band at low frequencies, the use of a positioner results in a different type of nonlinearity at high frequencies which appears to be much more troublesome that the original dead band. Fast systems with positioner-equipped actuators are much more prone to limit cycle, and tend to severely degrade controller settings. Better positioners, in the traditional sense, increase the severity of the problem. Where the best control is required, the use of a booster amplifier in lieu of a positioner is likely to result more favorable dynamics and consequently better stability for fast systems.”

Air Volume Booster Relay An air-volume booster relay is a pneumatic device that amplifies the capacity of a control signal. This greatly improves the response of the control valve and allows more effective dynamic control. Volume booster relays are used when a control valve is located far from its pneumatic controller or when the valve has a large pneumatic diaphragm and fast valve response is required.

Solenoid Operated Valves Solenoid-operated valves are used with control valves in a variety of on-off or switching applications. An example of an on-off application is closing the gas valve to a fired heater on loss of flame signal or on a loss of power to the solenoid. An example of a switching application is the switching of an alternate pressure source, such as nitrogen, into a device upon loss of instrument air. The solenoid valve is open when energized and allows the positioner output to pass into the diaphragm case. Upon power loss, the solenoid valve closes the port to the valve positioner and exhausts pressure from the diaphragm case of the control valve. (See Figure 900-20.)

April 2002

900-40



900 Control Valves

Fig. 900-20Typical Installation of a Solenoid Valve

Position Transmitter A position transmitter is used to sense the position of a valve stem and transmit that signal to a remote readout. This readout may be only for operator information or it can be used to initiate some action. The instrument can sense the valve stem position of rotary or sliding stem valves. The pneumatic position transmitter is similar in construction and mounting to the valve positioner. By a minor modification in a force balance-type valve positioner, any motion of the valve stem can be transmitted as a 3- to 15-psig signal to an indicating device. Electronic position transmitters are slide wire potentiometers mechanically connected to the valve stem. The output of this device provides a current signal that is proportional to the valve stem position. Actuator signal indication, generally 3 to 15 psig or 6 to 30 psig, is indicated by one of the three gages that have been specified as a three-gage manifold with which a positioner is to be equipped. The other two gages indicate supply pressure and instrument output, respectively. Where a positioner is not furnished, a 2-inch 0- to 30-psi gage can be mounted on the diaphragm of the control valve.

Lifting Lugs Control valves should be specified with lifting lugs, either attached to the actuator for diaphragm actuated valves or to the valve body for rotary valves, to facilitate installation and removal of control valves into and from piping. Modern control valves are furnished with positioners which are precision instruments that must be precisely aligned with the control valve actuator. Typically, when handling control valves without lifting lugs, riggers use web straps that they wrap around the actuators/positioners. This practice has the potential to damage the positioner, its


900-41

April 2002

900 Control Valves


linkage, and its alignment as well as the tubing that connects the position to the actuator.

Nameplates The nameplate attached to the yoke of the actuator contains information about the control valve. This nameplate is furnished by the manufacturer and lists information such as serial number, valve model number, size of valve, port size, body material, diaphragm pressure range, bench set pressure range, trim, order number, and valve tag number. Through serial numbers Quality manufacturers can track valve component and trim size, style and material. Because control valves are designed for fail-safe operation and because most can also be used for emergency block valve applications, they can be considered as critical control components. API 75 for Upstream applications and OSHA 1910 for Downstream applications require Positive Material Identification (PMI) for all critical control devices. API 75 and OSHA 1910 hold the user responsible for proper selection of all materials and maintenance of each component. As such, OEM components and factory certified repair facilities should be used for valve maintenance. Third-party repair of control valves, by shops that provide no material traceability, should not be used. Some operating centers require an additional nameplate, attached to the valve, that lists the valve tag number, service description, and valve failure mode.

960

Control Valve Installation Correct sizing and selection procedures, proper installation techniques, and periodic preventive maintenance are factors that can lengthen control valve service life. Some guidelines for the installation of control valves are as follows.

Accessibility for Maintenance All control valves should be installed so that they are readily accessible for maintenance. To achieve this, the piping designer should follow the manufacturer’s outline drawing. This drawing should show the required clearances in the following areas: •

Above the valve for the removal of the top works with stem and plug as an assembly

•

Below the valve for the removal of the bottom flange (if applicable)

•

On the sides of the valve for the removal of accessories such as manual operators, solenoid valves, valve positioners, etc.

•

At the flanges for removal of the flange bolts (Special consideration must be given to this area when reducers are used to connect the valve to the piping.)

Where an oversized actuator, especially on small valves, is required, the piping designer should provide the added clearance as well as the necessary support this actuator requires on the installation drawing. Notes should also be included on the

April 2002

900-42



900 Control Valves

installation drawing stating that a special field inspection will need to be made during installation of this actuator. Some processes require the piping and control valves to be steam-traced or electrictraced and at times steam-jacketed. This type of installation will require insulation to retain heat and to protect personnel from burns. Additional clearances should be provided for this type of installation so that the insulation will not interfere with the actuator or the accessories. Control valves located above grade because of design conditions should be installed so that they are readily accessible from a permanent platform or walkway with sufficient clearances for maintenance purposes. Control valves should be installed in horizontal lines with the diaphragm actuators located directly above the valve body. Control valves used in process lines or fuel lines to fired heaters should be located such that manifolds may be drained and the valves removed without danger of flashback.

Isolation of the Control Valve Most control valves are installed with block and bypass valves. By using the block valves to isolate the control valve, the bypass valve can be used to manually control the process while the control valve is being repaired. In general, the bypass valve is specified to be the same size as the control valve and is usually a globe valve that allows throttling. However, for large sizes (above 6 inches), gate valves may be considered because of the cost. Economic line sizing results in control valves normally being at least one size smaller than the pipe in which a control valve is installed. In the past, ChevronTexaco’s control valve manifold design standards called for control valve isolation (block) valves and the by-pass valve to be the same nominal size as the control valve body. One pitfall of this design is that gate valves, which are normally used as block valves, are manufactured with reduced ports. The reduced ports can be 60 to 70% of the cross sectional area of the nominal pipe diameter. Such valves function as restrictions upstream and downstream of the control valve and can limit flow through the control valve manifold. For that reason, many control valves operate in an unsafe mode with the bypass valve partly open. The latest API control valve manifold design guidelines call for block valves to be the same size as the piping. Where the greatest flexibility is to be provided for expansion, the block valves upstream and downstream of the control valve should be line size. However, where the control valve is two or more sizes smaller than line size, the block valves may be one size smaller than line size. In several areas of concern, the design of the manifolds is dictated by the requirements of the process. Batch operation, use of exotic alloys, very low available pressure drops, and slurry service where the bypass lines are likely to plug must be given special consideration when designing the manifold.


900-43

April 2002

900 Control Valves


Supports at control valve manifolds should be located such that the assembly is selfsupporting with the control valve removed. When designing control valve manifolds, refer to API RP 551, Manual on Installation of Refinery Instruments and Control Systems, and to GF-A1250, General Arrangement Standards - Plant Facilities - Control Valve Manifolds in the Piping Manual.

970

Control Valve Problems Control valve problems can be grouped into two categories: mechanical and design. All can be avoided by proper valve engineering.

971 Mechanical Problems Mechanical problems are those in which a control valve undergoes physical damage or other undesirable effects, e.g., abrasion, erosion, cavitation, noise, plug rotation, and vibration.

Abrasion Abrasion occurs when the process fluid contains suspended solids. Abrasive fluids are difficult to handle. If you think that the fluid you are trying to control may be abrasive, call an ERTC Instrumentation and Controls Specialist for support.

Erosion “Wire drawing,” a form of erosion, occurs in high pressure drop applications, where a valve that normally operates in a shut off position, e.g., a steam control valve to a spare turbine driven pump, leaks slightly (Class I, II, or III trim tightness). The leaking fluid gets “channeled” along the path of least resistance and eventually wears out a groove that resembles a wire extrusion die. This groove grows as the valve continues to leak until the leakage rate becomes unacceptable and the valve must be removed for maintenance. To preclude wire drawing, hardened (stellite-surfaced or solid stellite) trim should be specified. Although valve vendors recommend hardened trim only when pressure drop is 200 psi or greater, ChevronTexaco’s recommendation is to consider hardened trim when the pressure drop exceeds 100 psi.

Cavitation A control valve produces a pressure drop by reducing the area between the seat and the plug (disk or ball in rotary valves). Process fluid accelerates through the reduced opening. As the fluid accelerates, pressure in the area of acceleration drops, then recovers partially as the fluid enters the valve/piping cavity and decelerates (Bernoulli’s equation). If the pressure in the area of acceleration drops below the vapor pressure of the process fluid, the process fluid will flash and form vapor bubbles. If the pressure recovers above the vapor pressure of the fluid, the vapor bubbles will collapse

April 2002

900-44



900 Control Valves

(implode) and go back into solution. The implosion of the bubbles generates so much energy that metal is physically washed away from the valve body and the downstream piping. Cavitation is accompanied by noise that can be described as “a cement truck with its rotating drum full of gravel,” only louder. Severe cavitation can generate noise in excess of 110 dBA. Cavitation can be treated by selecting a valve with a lower pressure recovery coefficient, e.g., a valve where the process fluid is impinged upon itself from opposite sides of a cage (anti-cavitation cage trim) or by taking the pressure drop in several stages. (See Figure 900-21.) Fig. 900-21Pressure Drop Across a Multistage Anticavitation Control Valve

To ensure that all manufacturers rate and certify their cavitation application control valves uniformly, ISA is writing a Recommended Practice which establishes a uniform process for manufacturers to test and certify their cavitation service control valves. All major manufacturers have participated in writing this document and have endorsed the procedures and methods.

Aerodynamic Noise Aerodynamic noise is a result of a standing shock wave that is generated when vapor flowing through a control valve reaches sonic velocity at the vena contracta.


900-45

April 2002

900 Control Valves


Noise can induce vibration into and damage piping and equipment and can lead to hearing loss in personnel. The intensity of the noise is a function of the pressure drop, control valve size and style, and downstream piping size and schedule. Because noise propagates through the downstream piping, the most effective way to handle noise is to treat it at the source. Techniques for treating noise, in declining order of effectiveness and economics, include the following. 1.

Selecting a valve style designed to reduce noise, e.g., Fisher Whisper-Trim valves

2.

Using diffusers or baffle plates that take pressure drop in stages, eliminating sonic velocity conditions

3.

Using silencers (Although silencers are extremely efficient, they are expensive to buy and even more expensive to install and support properly.)

4.

Increasing the schedule of the downstream piping to help attenuate the noise

5.

Using acoustical lagging

Control valve noise can be predicted fairly accurately. Although each valve manufacturer claims to have a noise calculation method, the method developed by Fisher Controls has proven more reliable than other methods. In the Company’s experience, the Fisher method gives more accurate values, even when applied to other brands of control valves. Consult the Noise Control Manual for more guidance in this area.

Plug Rotation Plug rotation is a problem in some sliding stem valves. Hydraulic forces, produced by pressure drop and flow geometry through the valve, cause the plug to rotate along the axis of the stem. The plug, valve stem, actuator stem, and actuator are assembled with threaded connections. The rotating force tends to either tighten the threaded connections or to make them come apart; the higher the pressure drop, the greater the force. Even though these threaded connections are pinned, the cyclical forces cause the pins to work free and the connections come apart. Valves that tend to rotate comprise all skirt guided valves, including V-port trim valves, and cage-guided valve designs where openings that take the pressure drop are in the plug and not in the cage. ChevronTexaco has had a number of significant losses when plugs separated from the actuators during operation and the valves slammed shut. Although these types of valves are not safe, they are frequently furnished by vendors as part of packaged systems because of their low first cost. It is the responsibility of the engineer to screen and reject these types of valves.

April 2002

900-46



900 Control Valves

Vibration Mechanical vibration of valve components is a result of random pressure fluctuations within the valve body and/or fluid impingement upon the movable parts. The resulting physical damage incurred by the valve plug and/or associated guide surfaces is generally of more concern than the noise emitted. Mechanical vibration may be reduced by using trim with close clearances, such as cage guiding, or selecting material that reduces wear, thus preventing the enlargement of clearances. Both single- and double-ported nonpressure balanced valves can become unstable when throttling high-pressure drops at low lifts. As the plug is moved up and down by the actuator, it is subject to tremendous forces caused by flow impingement. This results in pressure fluctuation in the flowing media. Seats, plugs, and stems can be damaged by this. Throttling instability may be reduced by using a stiffer actuator (high spring rate) and a pressure-balanced, cage-guided control valve that is not subjected to the same magnitude of unbalanced forces. Water hammer should be checked, particularly for long pipeline applications. A flowing fluid possesses momentum due to its mass and velocity; when a valve is closed suddenly, the pressure buildup caused by the inertia of the moving fluid may cause damage to the valve, piping, and other process equipment.

972 Design Problems Design problems are those that cause shutoff, choked flow, and rangeability and stability difficulties.

Rangeability Problems Rangeability problems are encountered when an engineer does not consider all possible normal and abnormal conditions under which a control valve must operate. Rangeability requirements have been discussed earlier.

Stability Problems Stability problems are characterized by the inability of the actuator to hold the plug steady at a given position. The valve stem cycles up and down, literally like a sewing machine. This is usually the result of the actuator being unable to handle the process that is acting on the valve trim. This condition can occur when actual process pressure is greater than that specified by the engineer. It has also occurred when the valve sales representative was not familiar with the effect of hydraulic forces on valve trim and with stem force reversals that occur along the length of stem travel. In the first case, incomplete maximum upstream and shutoff pressure data is given to the valve vendor and the vendor selects an actuator on the basis of this incomplete process data. Although the actuator has been selected correctly, it does not have the power to overcome the actual process forces. This problem can only be


900-47

April 2002

900 Control Valves


prevented by providing the vendor with the complete range of process conditions, from shut-off through maximum design flow. This is the engineer’s responsibility. In the other, equally frequent case, a valve sales representative who is not completely familiar with valve hydraulics and possible stem thrust reversals within a control valve body attempts to become the low bidder by bidding the smallest actuator with the weakest spring rate. The engineer’s responsibility here is to spot check actuator sizing and proposed bench sets. If the valve “jack-up” exceeds onefourth of the valve control signal (3 psi “jack-up” for a 3 to 15 psig actuator or 6 psi “jack-up” for a 6 to 30 psig actuator), the engineer should reject the proposal and demand a bigger actuator with a stiffer spring.

980

Pressure Regulators The self-contained pressure regulator is simple, dependable, rugged, and inexpensive. The pressure from the flowing media acts on the diaphragm, compressing the spring and resulting in valve plug movement. The initial spring compression sets the pressure at which the valve begins to open. For each pressure on the diaphragm, there is a corresponding seat position and a corresponding flow. Many different regulator types and designs are available. The “air supply regulator,” or “airset,” is a common type of spring-loaded pressure-reducing regulator. It is usually a 0.25-inch air pressure regulator used to reduce instrument air header supply pressure to a level compatible with pneumatic instruments. These regulators are usually provided with an integral filter; hence, the name filter regulator. Many processes require that a certain pressure be maintained in the system by relieving excess pressure into a lower pressure line or area. For this application, a back-pressure regulator can be used. It consists essentially of a self-contained pressure reducer regulator reversed so that the inner valve opens when the inlet pressure reaches a predetermined pressure. In a back-pressure regulator, the upstream pressure or the controlled pressure acts on the diaphragm to control the inner valve internally or through a static connection. Such valves are not by definition emergency safety devices; rather, they are continuously operating pieces of equipment. The regulator is a complete, self-contained, feedback control loop with a narrow, fixed, proportional band. The regulated pressure will be offset by changes in the upstream pressure and flow demand in the case of a pressure reducing valve. This offset in regulated pressure with changing flow is called “droop.” Droop is a function of both the pressures (inlet and outlet) and regulator design parameters such as spring rate, valve lift, and diaphragm area. Droop will affect the operating span (flow range) of the regulator. Information on droop versus flow is therefore essential for satisfactory regulator performance. (See the manufacturer’s catalog.)

April 2002

900-48



900 Control Valves

The diaphragm in the self-contained regulator senses the change in the flowing media pressure which moves the valve plug (by compressing the spring). However, this simple, rugged, and inexpensive regulator does have some disadvantages. •

As fluids become more difficult to handle, the choice of regulators becomes more narrow and finally nonexistent.

•

The dynamic performance is fixed. It is usually impossible to make any adjustments except set point.

•

Regulators have inherent droop. That is, the controlled variable changes to some degree with load.

•

Regulators are adversely affected by low pressure drop.

•

Most regulators fail open upon diaphragm rupture, which can be an unsafe condition.

Control valves with separate controllers should be used when the disadvantages of regulators are not acceptable. Pilot-operated regulators provide more accurate regulation for a wide range of pressures and capacities than the spring loaded regulators. These types of regulators are generally used in pressure control of natural gas, fuel gas, air, and some steam applications. Because of the small passages and ports that can become plugged, these regulators should be used only with clean fluids. Pilot-operated regulators can have external or internal pilots. The upstream pressure is applied to the top of the main diaphragm and tends to close the valve. Increase in downstream pressure above the setting of the pilot spring moves the seat away from the upper ball of the pilot plug. This bleeds the pressure below the line valve diaphragm and tends to close the line valve. A decrease in controlled pressure closes the upper bleed port and opens the lower port, which allows loading pressure to open the valve. External pilot regulators are available in a wide range of sizes. This design also offers the convenience of remote set point adjustments.

990

References 1.

Hutchison, J.W., ISA Handbook of Control Valves, 2nd ed.

2.

Lloyd, Sheldon G., Guidelines for the Use of Valve Positioners and Booster Amplifiers (Fisher Governor Co).

Industry Standards A number of industry standards that are applicable when specifying a control valve. Some of the more common ones follow.


900-49

April 2002

900 Control Valves


American National Standard Institute (ANSI) Standards. •

ANSI B16.34 determines valve pressure capability at certain temperature levels for different materials.

National Association of Corrosion Engineers (NACE) Standard MR-01-75 specifies the proper materials, heat treat conditions, and strength levels required to provide good service life in a sour gas (where H2S is present) environment. American Society for Testing Materials (ASTM) Standards sets the guidelines for forgings and castings of control valve bodies and bonnets. American National Standards Institute/Fluid Controls Institute (ANSI/FCI) Standards. •

ANSI/FCI 70-2, Quality Control Standard for Control Valve Seat Leakage

•

ANSI/FCI 91-1, Standard for Qualification of Control Valve Stem Seals to Meet EPA Emission Guidelines for Volatile Organic Compounds

American National Standards Institute/International Society for Measurement and Control Standards (ANSI/ISA).

April 2002

•

ANSI/ISA Standard S75.01, Control Valve Sizing Equations.

•

ANSI/ISA Standard S75.03, Face-to-Face Dimensions for Flanged Globe-Style Control Valve Bodies (ANSI Classes 125, 150, 250, 300, and 600)

•

ANSI/ISA Standard S75.04, Face-to-Face Dimensions for Flangeless Control Valves (ANSI Classes 150, 300, and 600)

•

ANSI/ISA Standard S75.11, Inherent Flow Characteristics and Rangeability of Control Valves

•

ANSI/ISA Standard S75.16, Face-to-Face Dimensions for Flanged Globe-Style Control Valve Bodies (ANSI Classes 900,1500, and 2500)

•

ANSI/ISA Standard S75.17, Control Valve Aerodynamic Noise Prediction

•

ANSI/ISA Standard S75.19, Hydrostatic Testing of Control Valves

•

ANSI/ISA Standard S75.22, Face-to-Centerline Dimensions for Flanged Globe-Style Angle Control Valve Bodies (ANSI Classes 150, 300, and 600)

•

API Recommended Practice RP 552, Refinery Control Valves

•

API Standard 589, Fire Test for Evaluation of Valve Stem Packing

900-50


1000 Control Systems and Process Computers Abstract This section provides an introduction to the different types of control systems as well as their advantages, disadvantages, and applications within the Company. The section also provides a discussion of computers and a variety of systems used for process control and monitoring. A brief historical overview of process control helps the reader understand the basis of current control systems. Some basic concepts and terminology of control systems are introduced. More information about control systems and process computers is available from the materials listed in Section 1070, “References,” and from personnel in the corporate Engineering and Technology Department’s (ETD) Monitoring and Control Systems Division. Contents

Page

1010 Introduction

1000-3

1011 Historical Background 1012 Control Signals 1013 Types of Control 1014 Control Device Location 1020 Systems For Continuous Control

1000-12

1021 Pneumatic Control Systems 1022 Electronic Analog Control Systems 1023 Microprocessor-Based Control Systems 1030 Systems for Discrete Control

1000-20

1031 Hardwired Relay Logic 1032 Programmable Logic Controllers (PLCs) 1033 Distributed Control Systems (DCSs) 1040 Process Computer Systems

Chevron Corporation

1000-1

1000-28

June 1989

1000 Control Systems and Process Computers


1041 Introduction 1042 Computers for Monitoring 1043 Computers for Discrete Control 1044 Computers for Continuous Control 1050 Characteristics of Successful Projects

1000-51

1051 People Make The Project 1052 Control Objectives Analysis (COA) 1053 Design of Controls 1054 Operator Involvement 1055 On-Site Support 1056 Monitoring the Results 1060 Selection Criteria for Computer-based Instrumentation and Process Computers 1000-54 1061 Hardware 1062 Software 1063 System Integrity 1064 Vendor 1065 Other System Requirements 1070 References

June 1989

1000-59

1000-2

Chevron Corporation



1010 Introduction 1011 Historical Background The two major purposes of instrumentation are measurement and control. More than 2000 years ago, measurement devices were already in use to measure water flow in order to regulate material transfer and to account for consumption of water for irrigation. In the early part of the twentieth century, simple measurement devices such as pressure and temperature gages were mounted directly on the process pipes or vessels. Control was achieved by the operator walking about the plant monitoring the plant conditions and then making manual adjustments on valves, dampers and levers to best control the process. This was a true distributed control system, with the operators and instrumentation distributed at different points throughout the process. Early control mechanisms or regulators were mechanical. The movement of a diaphragm, bourdon tube, or bimetallic element operated the stem of a control valve. These mechanisms helped to automate the work and made it possible for operators to handle larger plants. In the 1920’s and 1930’s, pneumatic transmission of measurement signals was developed, making it possible to put indicators, recorders, controllers, and other equipment in a centralized area. These instruments were installed on large panels located at the process unit or in a centralized control room. A disadvantage of these pneumatic systems was that as the distance between the process and the control device grew, a significant delay, or deadtime, which was detrimental to the quality of the control, was introduced into the control system. The development of electronic analog signal transmission in the 1960’s helped alleviate this problem. The use of electronic transmission signals was hindered in some applications because of the danger of electrical energy in explosive and flammable environments. The application of digital computers to process control began in the late 1950’s. The computers in this era were large, expensive, and not reliable. It was not until the middle 1970’s that the use of computers became widespread in the process control industry. The early computers were used only for supervisory purposes because of their unreliability. Communicating with the panel-mounted controllers by setting valve positions and setpoints, the computers allowed for more complex control strategies. Use of computers also allowed changing the control system structure while the plant was still on-line. The advent of microprocessors and large scale integrated electronic circuits has made the development of distributed control systems possible. The first distributed control system (DCS) was introduced in 1976. These control systems distribute the functions of the control system among a number of processors so that a single failure does not affect the entire system and shut down the entire process. By distributing the functions and intelligence of the system, it is also possible to locate the analog/digital interface close to the process. This greatly reduces the amount of

Chevron Corporation

1000-3

June 1989



cabling required because the signals can be multiplexed and brought back to the computer over a single communication cable.

1012 Control Signals Analog Signals An analog or continuous signal comes from a process-measuring device that is continuous in nature and represents the value of the process variable. An analog signal has infinite resolution within its range boundaries. An example of an analog signal is a level measurement from a tank. The output of a level transmitter (Figure 1000-1) represents the conditions of the process. The level is measured by a differential pressure (D/P) cell and the output of the transmitter is in milliamps (mA). The output value is the analog of the level and varies linearly and continuously with the level in the tank, and as the level rises and falls, the milliampere value follows accordingly. Fig. 1000-1 Typical Analog Measurement

An analog signal can be represented as an electronic signal, with the output given either as voltage (1 to 5 volts) or current (4 to 20 milliamps), or as a pneumatic signal with the output given as pounds per square inch (3 to 15 psi). Examples of an analog signal include pressure, level, temperature, and flow. Figure 1000-2 provides a graphic representation of an analog signal. There are two types of analog electronic signals, high level and low level signals. The signals are differentiated by their voltage levels. The high level signal is typically 4 to 20 milliamps or 1 to 5 volts. The output of a transmitter is a high level signal and is typically 4 to 20 milliamps. The 1 to 5 volt signal is the most commonly used input by analog to digital converters. A 250 ohm resistor is used to convert a 4 to 20 milliamp signal to a 1 to 5 volt signal. The resistor is typically an integral part of a microprocessor-based control system. Low level signals are of low voltage, typically in the millivolt range. Examples of low level analog signals include thermocouples and resistance temperature detectors (RTDs). This voltage must first be amplified before it can be processed internally within a microprocessor-based system. Many of the new microprocessor-

June 1989

1000-4

Chevron Corporation



Fig. 1000-2 Analog Signal

based instrumentation systems will accept low level signals directly. The instrument system will amplify the low level signal internally. The low level signals can also be amplified via transmitters. Special temperature scanning devices (often referred to as temperature multiplexers) can be used to gather a large number of low level signals. The temperature scanner may have a dedicated panel readout or be linked to a computer or instrumentation system.

Digital Signals Digital signals, often referred to as sampled signals, are a set of discrete signals which approximates an analog signal over time. Digital signals are used by microprocessor-based systems. Analog signals can not be used directly by the microprocessor-based system. Analog signals are converted to digital signals through an analog/digital (A/D) converter. Similarly, signals output from a microprocessorbased system must be converted from digital to analog (D/A) before the signal can be used by the filed devices. These converters are typically an integral part of the microprocessor-based system. In a microprocessor-based system an analog signal is sampled at a predetermined frequency with the value of the last sample held until the next sample is taken. The sample frequency is a function of the instrument, the process being measured, and the capacity of the microprocessor. A digital signal is not a continuous signal as is the analog signal. The digital signal has finite resolution which approximates the analog signal. Figure 1000-3 provides a graphic comparison of an analog and digital signal.

Discrete Signals A discrete signal is a signal that represents one of two possible states, such as on/off, open/closed, or low level/normal. An electric light switch is a discrete device; it is either on or not on. Figure 1000-4 provides a drawing of a typical discrete measurement. The binary numbering system, which consists of a combination of 1’s and 0’s, is often used to encode discrete signals. Because of this, discrete signals are some-

Chevron Corporation

1000-5

June 1989


Fig. 1000-3 Digital Signal


Fig. 1000-4 Typical Discrete Measurement

times referred to as digital signals. This should not be confused with the digital or sampled signals discussed above. Discrete signals are typically the result of a switch, a relay contact, or a push button. As the switch or relay changes state, the discrete signal changes accordingly. The signal will appear as a voltage, usually 120 VAC or 24 VDC. This signal can be used to sound an alarm or light an indicator light on a local panel, used in startup and shutdown “relay” logic, or used as an input to a microprocessorbased system. Examples of devices which provide a discrete signal include a level or pressure switch, a run contact from a motor, or a limit switch on a valve. Figure 1000-5 provides a graphic representation of a discrete signal. Fig. 1000-5 Discrete Signal

Normal State (1) Abnormal State (0)

Pulse Signals A pulse signal is a special type of digital signal. Some field measuring devices generate pulses at a certain frequency. The frequency of the pulses varies according to the measured variable. Each pulse is equivalent to a specific measured amount of the variable. The pulses are then added up over time to provide a measured value. See Figure 1000-6 for an example of a pulse signal.

June 1989

1000-6

Chevron Corporation



Fig. 1000-6 Pulse Frequency Signal

Examples of instruments that generate a pulse signal are turbine meters and kilowatt (power) meters. In the case of the flow meter, each pulse represents a certain volume of flow. The pulses are then added up over time to obtain a volumetric flow rate. The frequency of the pulses is equivalent to the flow rate, while the number of pulses is equivalent to the volume. Many control systems will accept pulse signals directly. Other systems require special converters to convert the pulse signal to an electronic analog (4 to 20 milliamp) signal.

1013 Types of Control Continuous Control Continuous control is used on continuous processes. A continuous process is one in which process material is continually flowing through the process equipment. Continuous processes are usually measured and controlled by analog signals, although digital transmitters are being introduced by many vendors. Continuous control can be as simple as maintaining a flow rate or pressure in a single pipeline or as complex as control of an entire distillation column or hydrocracker. Continuous control involves the continuous measurement of a process variable via an analog signal and the adjustment of a final control element, such as a control valve to keep the process measurement at a desired value. Process values are maintained close to their targets or setpoints despite changes in the process or process upsets. Disturbances in the process caused by changes in, for example, feed composition and rate, fuel gas composition, or pressure are kept to a minimum. Continuous control is used extensively throughout the Company’s refineries, chemical plants, and producing processing plants. Continuous control is also used to a lesser extent at the Company’s pipeline and marketing facilities.

Sequential Control Sequential control is often referred to as on/off control. It is a series of discrete control actions performed in a specific order or sequence. These actions can be the opening or closing of valves or the starting or stopping of devices. The control

Chevron Corporation

1000-7

June 1989



actions can be initiated by an operator or a process condition or as a result of the passage of a given period of time. Sequential control is much used at the Company’s producing offshore, pipeline, and marketing locations. Sequential control can also be found at the Company’s refineries, but its applications are typically non-oil related. The applications within the refineries and chemical plants tend to be associated with water treatment, utilities, effluent processing, and possibly packaging.

Batch Control Batch control is a combination of sequential and continuous control. A batch process is a process where the operation is time-dependent and repeatable. The similarity between a batch process and a continuous process is that both processes can be time-dependent; however, in a continuous process, startup theoretically occurs only once. In a batch process a sequence of events is repeated over and over again. For example, consider a reactor and a series of storage tanks. Ingredients or reactants are charged into the reactor vessel, left to react, separated, and then sent to storage. Once this sequence of events is complete, the sequence starts all over again from the beginning. The sequence can be simple or quite complex, and can entail hundreds of sequence steps. Continuous control takes place within the individual steps of a batch; for example, temperature is controlled while a reaction takes place. Sequential control is used to move between the steps of the sequence. For example, a sequence of valve and pump manipulations must take place to transfer the contents of the reactor to another vessel. Batch control is used quite extensively at the Company’s chemical plants, particularly in the agricultural division. Batch tracking, a form of batch control, is a common pipeline application. Batch control is not normally a function of the Company’s refining or producing facilities.

1014 Control Device Location Self-Contained Control Devices Self-contained control devices are devices that contain the process measurement and the final control element in a single device. These devices are located in the field, mounted directly on the process and thus require no signal transmission. They are used to maintain a specific process value (setpoint) in order to meet safety and/or equipment specifications. Self-contained control devices are mechanical devices that are often used to control line pressure via use of a mechanical back-pressure regulating valve. Tank pressures can also be controlled by self-contained control devices. Self-contained control devices should only be used in situations where the setpoint never changes and operator attention is not required.

June 1989

1000-8

Chevron Corporation



Local Control Devices Local control systems are systems in which the measurement indication and the control device are adjacent to he process. A local control system can be as simple as a small controller or as large as a panel with a number of indicators and controllers on it. Local control devices are installed today in small plants or applications where it is not economical to bring the measurement and control signals to a centralized control room. In general, local controllers should only be used where very little operator attention is required to maintain performance and where the required control functions are very simple. Both continuous and sequential control can be performed locally. Local controllers can be either pneumatic or electronic (see Section 1020, “Systems For Continuous Control”).

Centralized Control Room The function of a centralized control room is to bring together in one centralized location all the measurement devices and controllers needed by the operator to monitor and control one or more facilities. The two main design formats for centralized control rooms are panel-mounted or CRT-based operator interfaces.

Panel-Mounted Instrumentation vs. CRT-Based Instrumentation. There is much debate about whether CRT-based (Figure 1000-7) or panel-mounted instrumentation (Figure 1000-8) provides the best operator interface to the process. In control rooms that use panel-mounted instrumentation, some operators feel that they can determine the state of the entire process in a few seconds by “scanning” the board. On CRT-based instrumentation systems operators often have to call up a number of displays to see where the process is and where it is going; this takes time. However, with a properly designed display system and operator interface, the benefits of a CRT-based system can far outweigh the benefits of a panel-mounted system. The Monitoring and Control Systems Division of the Company’s Engineering Technology Development Department have developed standard display guidelines for the Honeywell TDC-3000 DCS. Many of the concepts discussed in these guidelines pertain to other CRT-based operator display systems. Some of the Company’s facilities have panel-mounted digital controllers interfaced to a CRT-based system. This arrangement not only allows the flexible CRT operator interface but also enables the operator to go back to the board and control the facility if the CRT-based system fails. CRT-based instrumentation systems typically can not be economically justified for projects with less then 100 inputs (tags). For these cases, panel-mounted instruments are recommended. Each project must be looked at individually because there may be special reasons why a small system requires a CRT interface. Section 1044, “Computers for Continuous Control,” discusses the use of personal computer systems in conjunction with panel-mounted instruments to allow the operator the benefits of the CRT-based system on smaller applications. The current trend within the Company is to centralize as many process units as possible into one control room without moving the operator too far from the process. Control rooms are being centralized to help contain the rising cost of building control room buildings and reduce manpower costs. These costs are due to

Chevron Corporation

1000-9

June 1989


June 1989

Fig. 1000-7 Distributed Instrumentation Control Room (Richmond ISOMAX)

1000-10


Chevron Corporation


Chevron Corporation

Fig. 1000-8 Panel-Mounted Instrumentation Control Room (Richmond LSFO)

1000-11


June 1989



strict building safety codes and the special environmental conditioning (temperature and humidity) required by modern instrumentation.

1020 Systems For Continuous Control Pneumatic, electronic, and microprocessor-based systems can be used for continuous control. This section defines each type of system, lists its advantages and disadvantages, and provides examples of its use within the Company.

1021 Pneumatic Control Systems Pneumatic control systems communicate and control with air pressure (Figure 1000-9). The process variable is measured and converted to an air pressure signal by a pneumatic transmitter. The transmitter then sends the air pressure signal through a tube to a receiver. This receiver can be an indicator, controller, recorder, or relay. Fig. 1000-9 Pneumatic Control

The air pressure signal typically is of the 3 to 15 psi range. This pneumatic signal is an analog of the process variable that is being measured or a controller output. For example, for an instrument which has a range of 0 to 150 GPM, a signal of 3 psi would correspond to 0 GPM and a signal of 15 psi would correspond to 150 GPM. Pneumatic relays can be used for some logic functions, such as high or low select, and for some simple arithmetic functions, such as addition, multiplication, solving for the square root, and totalization. All working parts within pneumatic instruments are mechanical.

June 1989

1000-12

Chevron Corporation



Advantages of Pneumatic Control Systems One of the main advantages of pneumatic control is that the components are intrinsically safe and can be used in explosive environments without special modifications. Pneumatic instrumentation is also more resistant to corrosive environments than its electronic counterpart. Pneumatic control can have an economic advantage over other types of control when it is installed as a single, stand-alone, local controller. In these small applications, it is often easier to run instrument air then to run electrical conduit to power the instrument. Pneumatic controllers tend to be reliable for these applications. In many producing installations, especially offshore, natural gas is used to actuate controls, eliminating need for an air compressor, power supplies, etc.

Disadvantages of Pneumatic Control Systems One of the main disadvantages of pneumatic signal transmission is the problem of transmitting the signal over significant distances. A time lag is associated with the transmission of the pressure signal through the instrument tubing. As the distance that the signal must travel increases, the speed of the response of the pneumatic transmission systems increasingly becomes a problem for “fast” loops such as flow. Alternative solutions may be necessary. Pneumatic systems installed for distances up to 350 feet have a time constant of about a second and a dead time of one-third of a second. Transmission of pneumatic signals up to 1,000 feet can be used if a time constant of about 7 seconds and a dead time of 1 second are acceptable.

Applications for Pneumatic Control Systems Follow these guidelines when deciding whether to use pneumatic control systems: •

Process can tolerate the time constant of 1 second and a dead time of one-third of a second.

•

The distance that the measurement or control signal travels is relatively short (less than 350 feet).

•

The installation has no complex control requirements (no complex control algorithms).

•

Extreme accuracy (1/4% or less) is not necessary.

•

Can be used for all environments (including explosive and corrosive).

•

Can be used for applications with inadequate electrical power sources.

Almost all control valves are controlled by pneumatic actuators. The rest of the control system may be electronic or digital-based, but the final control element is pneumatic. Pneumatic control valve actuators are used because they are intrinsically safe, inexpensive, reliable, and can generate the force needed to overcome the process dynamics, which electronic value actuators cannot do.

Chevron Corporation

1000-13

June 1989



The Company’s offshore producing facilities are the prime users of pneumatic based control systems. Platforms use pneumatics because the environment is explosive and the type of control performed requires little operator attention. The pneumatic control systems are also easier for platform personnel to maintain. Although electronic control systems have been used since 1965, over 60% of the Company’s downstream facilities still run entirely on pneumatic control. These control systems are slowly being upgraded to some form of electronic control as spare parts for the older instruments become difficult to obtain and as the need grows for better and more advanced and efficient process control. Single loop digital controllers are often used for individual loop replacement as the old pneumatic instruments fail, and distributed control systems are often used to retrofit entire pneumatic control systems.

1022 Electronic Analog Control Systems Electronic signals operate by converting the process or control variable to an analog electric signal, usually a voltage (1 to 5 volts) or a current (4 to 20 milliamps). Analog electronic instruments include controllers, indicators, recorders, and relays. Electronic relays can be used for some logic functions, such as high or low select, and for some simple arithmetic functions, such as addition, multiplication, solving for the square root, and totalization. Analog electronic instruments either accept or output electronic signals even though the control system may be a mix of both pneumatic and electronic signals. For example, pneumatic signals are often used to modulate control valves. In this case the controller outputs an electronic signal and then a transducer converts the electronic signal to a pneumatic signal before it can be used by the valve. See Figure 1000-10.

Advantages of Electronic Control Systems Electronic signal transmission has a virtually instantaneous response and hence the response time is not a problem in process control applications. Because of their speed of response, electronic control systems allow nearly unlimited distances for transmission by wire, radio linkage, or microwave signals.

Disadvantages of Electronic Control Systems Electronic control systems can quickly become complex for more than just single loop control. Use of electronic instruments for anything except simple control or logic is not recommended.

Applications for Electronic Control Systems Few if any new installations within the Company use electronic instrumentation systems. Applications which in the past would have used electronic control systems now use digitally-based or microprocessor-based control. Electronic control systems can still be found in new installations on systems designed and supplied by vendors.

June 1989

1000-14

Chevron Corporation



Fig. 1000-10 Electronic Control

Many existing Company facilities are controlled and operated via electronic instrumentation systems. A number of these instrumentation systems are also interfaced to supervisory control and monitoring systems.

1023 Microprocessor-Based Control Systems The advent of the microprocessor has revolutionized the control industry. A microprocessor is, essentially, a minicomputer on a single microchip. Microprocessorbased systems are often referred to as digital control systems. Microprocess-based control systems use digital signals. Electronic signals are converted to a digital signal via an analog/digital (A/D) converter which is typically part of the control system. All microprocess-based systems require some form of programming to define the data acquisition and control functions. This definition procedure is referred to as configuration. Configuration may be as simple as filling out “fill-in-the-blank” forms or using a high level programming language. The two categories of microprocessor-based control systems are single loop control systems and distributed control systems.

Single Loop Controllers A single loop controller (SLC) is a digital controller that is responsible for the control of a single loop or control valve. The controller contains a microprocessor that makes it extremely powerful compared to its pneumatic and electronic counterparts. Single loop controllers can be stand-alone or can be linked together on a common data highway that can provide for better control and display opportunities.

Chevron Corporation

1000-15

June 1989



SLCs are small and typically panel-mounted. All data displayed on the faceplate of the controller are shown in a digital format. Typically, the faceplate will show the value of the measurement, the setpoint, and output for a loop as well as a bar graph representation of the same data. See Figure 1000-11. Fig. 1000-11 Single Loop Controller (SLC)

SLCs typically allow for one to three analog inputs and from one to two analog outputs. The output signals can be used for manipulating valves, or driving trend recorders, or as inputs to other instruments. SLCs also have a few discrete (digital) inputs and outputs for alarming and simple logic functions. The mode of configuration for SLCs differs among manufacturers. Available options include configuring from the front panel of the controller, via a small portable plug-in unit (calculator-like) or via a link to a personal computer. Advantages of Single Loop Controllers:

June 1989

•

The main advantage of an SLC is that it preserves single-loop integrity. When an SLC fails, control of only one loop is lost, and not of entire portions of the facility.

•

SLCs can be very cost effective for small applications (1 to 100 points).

•

SLCs digital controllers have tremendous control flexibility compared to its pneumatic or electronic counterparts.

•

The cost of an SLC is equivalent to that of a pneumatic or electronic controller.

•

SLCs contain math and logic functions that can eliminate the need for additional instrumentation.

•

Operators feel secure with SLCs because they closely resemble the older boardmounted instrumentation.

1000-16

Chevron Corporation



Disadvantages of Single Loop Controllers: •

For large applications (greater than 100 points), the amount of panel space needed to mount the controllers becomes prohibitive.

•

SLCs do not have the expansion flexibility of a distributed control system.

•

SLCs can control only 1 or 2 loops.

Applications for Single Loop Controllers: • • • •

New installations for systems with a small amount of I/O (1 to 100) Replacement of individual pneumatic and electronic controllers Backup of critical loops on larger distributed control systems SLCs should not be used for any “complex” control strategies.

Distributed Control Systems A distributed control system (DCS) as shown in Figure 1000-12 distributes the functions of a control system into many different microprocessors. The microprocessors form small subsystems and are linked together via a communication or data highway. These subsystems are often referred to as “nodes.” The individual nodes in a distributed control system perform a variety of functions. These functions can be located in individual nodes, or located on the same node. Fig. 1000-12 Typical Distributed Control System Layout

Control. The control subsystems are typically responsible for scanning the inputs into the system and sending outputs to the field. All simple control functions and

Chevron Corporation

1000-17

June 1989



some complex functions are performed within the control subsystem. The control functions are accomplished by the execution of function blocks that contain specific preprogrammed algorithms. Some DCS manufacturers also provide a high-level programming language within the control subsystem to perform some of the more advanced control applications. The control subsystem is also typically responsible for alarming and for gathering data for short-term trends. Display. The display subsystem is the operator’s interface to the process. It is responsible for retrieving all information from the network that is required by a display and calling up the display definition. Typically, the display subsystem will store many of the standard operating displays in its memory to help reduce display call-up time. The display subsystem can support single and sometimes multiple CRTs. The operator consoles are 13- or 19-inch color CRTs typically with touch screen and a cursor control device such as a mouse or trackball. The operator interface is one of the features of the distributed control system. Information from many geographical areas is brought together in one place before the operator. These DCS systems give the operator color graphics with dynamic process data, faceplate displays that try to replicate functions of the panel mounted controllers, trend data, and an alarm summary or annunciator displays. At a large facility, thousands of pieces of data are often at the operator’s fingertips. It is a challenge to provide all the information the operator needs in a precise and consistent manner. Data History and Collection. The history subsystem is responsible for gathering data for long term storage. A mass storage device is always associated with this node. The historical data can be called to the operator’s display screen in graphic and sometimes tabular format and can be used for control applications, optimization routines, and reporting functions. Depending upon the DCS manufacturer, the functions of the history subsystem can be on a dedicated processor or on a processor that has a number of other functions. Interface to Other Computers. Interfaces or gateways to other computer systems are sometimes required to allow communications with an advanced control system or a monitoring system. The most common computer interface is to a DEC-VAX. Interface to Other Instruments or Instrument Systems. DCS manufacturers also provide interfaces to other instruments, such as intelligent analyzers or temperature scanning devices. Interfaces are also available to many of the programmable logic controllers, smaller instrument systems, such as single loop controllers, and first generation instrumentation systems. These interfaces are also often referred to as gateways. Report and Management Functions. Report generating functions are also available. Examples of typical operating reports that a DCS can produce include shift, daily, monthly, maintenance, and emissions reports. The reports can use data from the history package or real time data. Typically, a node on the network is not dedicated to these functions. Often the report functions are on the node that contain

June 1989

1000-18

Chevron Corporation



some of the system management functions, engineering configuration, and data collection. Engineering and Configuration Functions. The system can be configured from one of the operator consoles or from a dedicated engineer workstation. The engineer workstation can be a personal computer or a console identical to the operator’s but dedicated to engineering functions. A dedicated engineer’s workstation is recommended for large systems. From the engineer’s workstation the engineer must perform system, database, display, history, and report configuration. Standard engineering utilities provide “fill-in-the-blank” forms to configure the database. Typically, a higher level computer language is provided to perform more advanced applications. The node responsible for engineering configuration will have access to a mass storage medium. The engineering and configuration functions are typically found on the system processor node. The nodes can all be located in a central location (such as a control room) or they can be physically distributed throughout the facility. There are often economic advantages to locate the node responsible for control and input and output (I/O) in the field, because then it is not necessary to wire each individual instrument to the control room. The instruments are wired to a panel in the field and then a single cable is run from the field panel to the control room. The controllers on distributed control systems differ from SLCs in that they are typically responsible for the control of a large number of loops and have a large I/O count. For example, a typical controller on one of today’s DCS systems supports 16 analog inputs, 8 analog outputs, and a combination of 48 digital inputs and outputs. Because an individual controller is responsible for a larger portion of the facility, backup systems have been developed to prevent catastrophic events upon DCS hardware failure. Distributed control systems are highly reliable. Many of the functions within the system are redundant and have automatic backup upon failure. All control functions and most input and output (I/O) processing is capable of being made either redundant or fault tolerant. The terms “redundant” and “fault tolerant” sometimes lead to confusion. Stating that a hardware subsystem is redundant indicates that identical hardware is available to take over immediately when a failure occurs. Stating that a hardware subsystem is fault tolerant indicates that certain pieces of the hardware that are less reliable are made redundant, but not the entire subsystem. For example, a redundant subsystem would have a duplicate of every hardware component in the subsystem. A fault tolerant subsystem will have only individual redundant pieces, such as a disk or a processor board. Redundancy is not always a standard feature on these systems. It is typically an added expense. Redundant control and I/O hardware are recommended for all continuous control applications where a piece of hardware is responsible for more than one process loop. Advantages of Distributed Control Systems: •

Chevron Corporation

All operator data is in a central location.

1000-19

June 1989



•

The operator interface is flexible.

•

Expansion is easy (the instrument panel does not require modification).

•

Control logic can be easily changed while the facility is operating.

•

Distributed control systems work well for continuous, sequence, and batch control.

•

The cost per loop for large systems is low.

•

They provide an extensive set of advanced continuous control algorithms.

•

All database items are tag addressable.

Disadvantages of Distributed Control Systems: •

The initial software cost is higher.

•

Software maintenance personnel are required.

•

Distributed control is costly for systems with a small I/O count (1 to 100).

•

If I/O and control is not redundant, control of multiple loops will be lost if the hardware fails.

Applications for Distributed Control Systems. Distributed control systems are appropriate for installations that have a large number of signals (100 or more), a centralized operator interface, and requirements for the following: • • • •

Reporting Advanced control and optimization Long-term data storage Refinery-wide information systems

Distributed control technology is the mainstay within the process control industry and the Company. Almost all continuous control applications of significant size are using distributed control systems. These systems can satisfy the control needs of almost any control application. Those applications requiring special or dedicated systems can be interfaced to the DCS, allowing the operator a single window to the process.

1030 Systems for Discrete Control Hardwired relay logic, programmable logic controllers (PLCs), and distributed control systems (DCSs) can be used for discrete or sequential control. This section defines each type of system, discusses its advantages and disadvantages, and provides examples of its use within the Company.

June 1989

1000-20

Chevron Corporation



1031 Hardwired Relay Logic Hardwired relays are used to perform discrete logic functions. Electromechanical relays are wired in series or parallel with field input signals to turn equipment on and off. Examples of input devices used by relay logic include the following: • • • • • • •

Pushbuttons Selector Switches Limit Switches (LS) Level Switches Proximity Switches Timer Contacts Status Contacts

These input devices either allow current to flow through the circuit or cause a break in the current flow, thereby switching an instrument on or off. See Figure 1000-13. Fig. 1000-13 Relay Logic

Instruments controlled by outputs from relay logic switched on by discrete devices include the following: • • • • • •

Pilot Lights (PL) Solenoid Valves Horns Control Relays (CR) Timers Starter Coils

Advantages of Hardwired Relay Logic The major advantage of hardwired relay logic is that it can be cost effective for small control applications (2 to 3 relays).

Disadvantages of Hardwired Relay Logic •

Chevron Corporation

Expansion is not possible without disabling the system.

1000-21

June 1989



•

The control logic cannot be changed without adding or deleting relays or disabling the system.

•

For a large number of signals, the system cost is high.

•

It is difficult to implement and troubleshoot complex logic for systems using hardwired relays.

•

Hardwired relay logic requires a fair amount of panel space for large systems.

•

Logic testing functions become complicated.

Applications for Hardwired Relay Logic Hardwired relay logic was the mainstay for interlocks, control, and emergency shutdowns for decades. The Company has many hardwired relay logic control systems currently installed and operating, but virtually all new discrete control applications within the Company use either a PLC or a DCS. The only new installation where hardwired relay logic should be applied is for very simple systems, requiring no more than two relays, interlock or emergency shutdown logic applications. Small-scale PLCs with eight inputs and outputs are now available for approximately the same cost as two hardwired relays.

1032 Programmable Logic Controllers (PLCs) A PLC is a microprocessor-based control system that was initially designed for discrete control applications. The National Electrical Manufacturers Association (NEMA) defines a PLC as a “digital electronic apparatus with a programmable memory for storing instructions to implement specific functions such as logic, sequencing, timing, counting, and arithmetic to control machines and processes.” PLCs were originally designed to replace hardwired relay logic. All of the functions performed by individual hardwired relays can now be performed within the PLC software. This eliminates the cost of extra relays and wiring and reduces the panel size required for the system’s installation.

Hardware for PLCs The hardware for PLCs was specifically designed to be installed in industrial areas. Most PLC hardware can be placed in locations where there are substantial amounts of electrical noise, electromagnetic interference, mechanical vibration, extreme temperatures, and humidity. Some PLCs are suitable for Class I, Division 2 explosive environments. The three main hardware components of a programmable logic controller are the main processor, the memory, and the I/O system (Figure 1000-14). The main processor is responsible for execution of the operating system, I/O scanning, and applications (ladder logic) software. In addition to the main processor many PLCs also have math co-processors which are responsible for mathematical functions and more advanced logic functions.

June 1989

1000-22

Chevron Corporation



Fig. 1000-14 PLC Block Diagram

The input/output (I/O) system is the interface by which the field devices are connected to the PLC. The interface conditions signals received from or sent to external field devices. Typical digital input and output devices are as follows: Inputs

Outputs

•

Pushbutton

•

Pilot Light

•

Selector Switch

•

Solenoid Valve

•

Limit Switch

•

Horn

•

Proximity Switch

•

Control Relay

•

Timer Contact

•

Timer Switch

•

Analog Measurements

•

Motor Starter Contact

•

Circuit Breaker

The I/O system is made up of individual modules that provide terminations for the field wires and the I/O interface to the PLC system. The individual modules are specific to I/O signal types. There are unique modules for analog and digital signals and for each input and output signal type. I/O modules can be located close to or some distance away from the main processor. The maximum distance that I/O modules can be located from the main processor is dependent upon the controller but can range anywhere between 1,000 and 15,000 feet. The communication media between the main processor and I/O modules can be a direct parallel connection, a twisted pair, twinaxial cable, a coaxial cable, or a fiber optic cable, depending upon the controller and the distance. Figure 1000-15 describes how the I/O modules can be interfaced to the bus. One method is to interface all the I/O modules directly to the parallel bus. The majority of applications are installed this way. This design is very fast, but the distance

Chevron Corporation

1000-23

June 1989



between the I/O module and the main processor is limited to approximately 200 feet. Fig. 1000-15 PLC I/O—Structures

To allow greater distance between the I/O and the main processor, remote I/O modules have been developed. Instead of the main processor interfacing directly with the I/O modules, the processor communicates with a local I/O processor. The local I/O processor then serially transmits the data to a remote I/O processor located near the I/O module. The remote I/O processor in turn transmits the data to the I/O modules. The two possible configurations for the local I/O processor are (1) point-to-point and (2) multi-drop. The point-to-point configuration has a dedicated local I/O processor for every remote I/O processor. In the multi-drop configuration, the local I/O processor is interfaced with two or more remote I/O processors. A PLC system can be made up of one or many multi-drop, or point-to-point I/O configurations. Scan time of the I/O is dependent upon the type of controller and the type of data transmission. Typically, PLCs have extremely fast scan times and fast logic execution. I/O connected directly to the parallel bus can typically be scanned around every 5 milliseconds. I/O coming through remote I/O processors is typically slower, sometimes as slow as 40 to 50 milliseconds.

June 1989

1000-24

Chevron Corporation



To increase the PLCs fault tolerance, different portions of the PLC system can be made redundant. These include the main processor, the communications path, the remote I/O processor and even the I/O modules. Processor redundancy is typically expensive, awkward, and not a smooth transfer. The Company typically uses only redundant processors for applications involving emergency shutdown logic or applications requiring high reliability. PLCs are typically “off the shelf” systems with specific hardware configurations. PLC manufacturers can not custom design a particular structure that is not offered in their product line.

Software for PLCs The application software for PLCs was developed so that it could be easily used by electricians and technicians. It mimics hardwired electromechanical control circuits and is called Relay Ladder Logic. The programmed rung concept of ladder logic is a direct carryover from the hardwired relay ladder rung, in which input devices are connected in series and parallel to control various outputs. The software programming tools use conventional relay ladder logic symbols. An example of PLC ladder logic is shown in Figure 1000-16. Fig. 1000-16 Relay Ladder Logic

A rung is the contact symbology required to control a single output. A single output can be controlled by one or multiple inputs. A complete ladder logic diagram will have many rungs. Each rung on the ladder logic is a combination of input conditions connected from left to right between two vertical lines, with a symbol that represents the output on the far right of the line. The symbols are connected in series, parallel, or some combination thereof, to obtain the desired logic. The processor executes the logic from the first rung (top) all the way to the last (bottom) rung. In addition to relay ladder logic, PLCs may also have other software programming tools that allow the user to configure the application via functional blocks, Boolean

Chevron Corporation

1000-25

June 1989



logic, and high-level programming languages such as Basic. The types of programming tools available depend upon the manufacturer and the type of controller. PLCs can be programmed via a proprietary PLC configurator, or via a link to a personal computer with special software.

Computer Interfaces for PLCs PLCs can also be linked to other computers via some type of communications network. These computers can provide many of the functions of a DCS, such as a CRT-based operator interface, data collection for historization and reports, and more advanced logic or control applications. PLCs can also be linked together via a communications network. This allows individual PLCs to use data from other controllers in order to perform control functions. It also allows the operator interface and report and data collection functions to be in one centralized location for the entire facility. These PLC networks resemble a distributed control system. The communication networks may use a standard protocol such as Ethernet or one specially designed by the vendor. Supervisory Control and Data Acquisition (SCADA) systems (discussed in Section 1043, “Computers for Discrete Control”) typically use PLCs as part of their system structure. The PLCs are located in remote locations and are linked together with some form of communications (radio, telephone, or cable) to a centralized computing center. This centralization makes possible a more efficient operator interface and allows for data collection and reports. All distributed control manufacturers have interfaces to PLCs available. The list of available interfaces varies depending upon the vendor and the product line. A PLC interfaced to a DCS allows the operator interface and data collection and reports for the sequential and continuous control to be in the same location.

Advantages of PLCs

June 1989

•

PLCs are very reliable.

•

PLCs have a fast scan and logic execution time compared to distributed control systems.

•

Control logic changes on-line are possible without disturbing the process.

•

PLCs allow for data collection when they are linked to another computer.

•

Control and data acquisition expansion is easy, but may be limited due to rack space.

•

Logic can be readily duplicated for similar equipment and processes; it is not necessary to rewire an entire set of new relays.

•

PLCs have minimal space requirements.

•

PLCs provide logic troubleshooting aids.

•

PLCs allow for remote I/O.

1000-26

Chevron Corporation



•

PLCs are typically less expensive than hardwired relay logic applications and distributed control systems.

•

PLCs offer limited analog control capability when few loops are required.

Disadvantages of PLCs •

PLCs do not have the advanced continuous control functionality of a DCS.

•

PLCs do not have a global database within a PLC network.

•

Database items within a PLC are not tag addressable, which makes some programming applications more difficult than on a DCS.

•

It can be awkward and sometimes expensive to develop an operator interface for a PLC (i.e., panels, buttons, lights, switches, etc.).

•

The PLC interface to supervisory computers is slow.

•

In general, most PLC applications are limited ladder logic, but some higher level programming languages are being introduced.

•

PLCs require a special programming interface.

Applications for PLCs PLCs have wide applications and use within the Company’s upstream facilities. Upstream locations use PLCs for both discrete and continuous monitoring and control applications. Downstream locations within the Company typically only use PLCs for discrete control applications and a small number of continuous control applications. Examples of PLC usage within the Company include: • • • • • • • • •

Sequence Control Emergency Shutdown Logic Batch Processes (Chevron Chemical) Materials and Handling Water and Waste Treatment Offshore Oil Production Platforms Discrete Monitoring and Alarming Applications Analog Data Acquisition Simple Continuous Control Applications

1033 Distributed Control Systems (DCSs) Distributed Control Systems also perform discrete control functions. Scan time and logic execution time on a DCS are typically an order of magnitude or more slower than PLCs. Scan times on a DCS system typically are on the order of every 250 milliseconds to 1 second compared with the PLC 5 to 50 milliseconds. For many applications the scan time on the distributed control system is sufficient.

Chevron Corporation

1000-27

June 1989



The general structure and functions of a Distributed Control System are discussed in more detail in Section 1043, “Computers for Discrete Control.” Many DCS systems now have ladder logic software programming tools available. In addition to ladder logic, DCS systems also use some type of high-level advanced control language. These languages allow for more advanced applications that may be difficult to configure in ladder logic and also provide for more operator interactions, such as messages. If fast logic is required when using a DCS for discrete control, use a PLC for the fast logic functions and then interface the PLC to the DCS for supervisory control. The Company’s downstream locations are the most common users of distributed control systems for discrete control. A DCS system is typically required for downstream applications because of their extensive and complex continuous control requirements. Since the DCS is already present at the facility, the discrete control applications are done within the DCS without much additional hardware, training, or maintenance costs.

1040 Process Computer Systems 1041 Introduction Initially, computer systems were designed with an emphasis on hardware. The software was developed to accommodate the hardware. Today’s modern process computer and control systems are very software intensive. Now hardware is designed and chosen to meet the needs of the system; however, most systems are driven by software requirements. The three types of computers are mainframes, minicomputers, and microcomputers.

The Mainframe The mainframe is physically the largest computer and provides the most computing power of the three types of computers. The central processing unit (CPU) for these computers used to be mounted on a characteristic frame; hence the name “mainframe.” Mainframes are very powerful and very expensive. They traditionally are not used as process computers. Examples of the mainframe computers include the large IBMs such as the IBM 3090 and the IBM 3084.

The Minicomputer Minicomputers are the middle size of the three types of computers. Generally, they are cheaper and less powerful than mainframes and faster and more powerful than microcomputers. Examples of minicomputers include DEC PDP, DEC VAX, HP1000, and Taylor 3106. Minicomputers are used extensively throughout the process control and monitoring industry. They are typically interfaced to a control system to perform advanced control, optimization, data historization, and advanced alarming functions.

June 1989

1000-28

Chevron Corporation



The Microcomputer (Personal Computer) The microcomputer originally got its name because its CPU was located on a single chip called a microprocessor. Today, larger computers also use microprocessors. For the purpose of this discussion, the term “microcomputers” will be used to indicate what we know today as personal computers (PCs). Examples of microcomputers are the IBM PC, the Apple MacIntosh, and other IBM clones. Typically, microcomputers are not used for process control, but they are used for process monitoring and data acquisition. Microcomputers are also used as engineering configuration interfaces to many of the control systems. The process industry uses computers with special operating systems. The two major attributes of a process computer’s operating system are operation in real time and multi-tasking. Real Time. Real time operating systems perform functions in real, or clock, time. For example, suppose a pump needs to be started and then 2 seconds later a valve needs to be opened. The application control program would first start the pump and then 2 real time seconds later would open the valve. Real time operations are used in every aspect of process computers, for example, I/O scanning, alarming, control algorithms, data collection, and historization. Multitasking. Multitasking operating systems provide special scheduling of programs or tasks that allows for more efficient use of the processor. Only one task can be executed in the CPU at a time. The multitasking system schedules the different tasks to be executed in the CPU. Each task is assigned a certain priority. The operating system schedules the tasks according to priority and suspends tasks of lower priority to allow a task of higher priority to run. This provides for efficient utilization of the processor and results in nearly simultaneous execution of the individual tasks. Also of equal importance, multitasking allows interruptions to occur during the normal execution of many programs without loss of data and the ability to finish execution once programs of higher priorities have been completed.

1042 Computers for Monitoring This section discusses the use of computers in process monitoring. Some background information is presented along with a discussion of the systems presently being used for process monitoring. A description of the computer process monitoring system currently being developed by the Company (UNICORN) is included. The purpose of computer-based process monitoring is to gather and view plant and facility information. Process monitoring is used for data acquisition and presentation, not for process control. The monitoring system can calculate items such as reactor yields and economic variables and can display process variables such as reactor temperatures, tank levels, and process flow rates. This information can be displayed to the operator and presented in an easy-to-use form. For example, color graphics can be used to present process alarms and data to operators in a form that can be easily and quickly understood. However, monitoring systems are not limited to use by the operator. The same information can be analyzed, reformatted and

Chevron Corporation

1000-29

June 1989



presented to management. This information reflects the present condition of the facility and aids in optimum planning and scheduling. Chevron is a leader in the process monitoring field. The Engineering Technology Department (ETD) first developed and installed a computer-based monitoring system in the late 1960’s. This system evolved into CHEVMON, which was designed to meet the particular needs of the petroleum industry. A powerful new process monitoring system, UNICORN, is scheduled to be released in 1989. This system has been designed by ETD to take advantage of the powerful new computers that have recently entered the market.

Process Monitoring Applications The traditional process monitoring application is designed for use by the operator. Process alarms and the status of the plant are displayed. Graphics, tables, and plots of plant variables are used to help the operator understand large amounts of data. Various mathematical operations can also be performed to aid the operator. For example, a noisy process signal can be passed through a digital filter. Another common application is the use of a program to estimate a process variable that cannot be measured; an example of this is the on-line estimation of the octane of a gasoline stream from a reforming unit. Process monitoring systems can be thought of as information managers. Modern process monitoring systems no longer are limited to operator interfaces; they also have the ability to distribute real-time information to engineers and managers. These systems are able to interface with existing DCSs, PLCs, and other sources to accomplish this goal. Management requires up-to-date information for accurate planning and scheduling decisions. Engineering needs both current information on facility activities and historical data for engineering studies. Ideally, a monitoring system will allow all functions access to the facility database and will present the data in a useful format. Management should also be able to pass information back to operations in order to implement their goals. As technology advances, the differences between a DCS and a process monitoring system become less distinct. However, distributed control systems are primarily designed as process control systems. These systems, including their communications network, are usually fully redundant. The graphics in these systems are very powerful, but they are directed toward operators and process control. The same is true of data management with a DCS; engineering studies can be performed and plant reports can be generated, but these tasks can be more easily accomplished by process monitoring systems. Finally, DCSs are generally expensive. In contrast, process monitoring systems are directed more toward information management. These systems are not as critical to plant operations and hence are not fully redundant. They are relatively inexpensive. The latest DCSs and process monitoring systems do, however, have more and more overlapping functions. For example, the Company’s future process monitoring system, UNICORN, will include some continuous and discrete control functions. Therefore, UNICORN could be used for a small number of control loops where

June 1989

1000-30

Chevron Corporation



redundancy is not critical and cost is a factor. On the other hand, some applications will use a DCS without a process monitoring system. The decision to use a DCS or process monitoring system or both systems will depend on the application. It is anticipated that many applications will run a process monitoring system in addition to a DCS. This is especially true in refineries and larger plants, where the two systems will make the process information available to a plant-wide network.

Process Monitoring at the Company—Present and Future The Company has considerable experience with process monitoring systems. The Company’s first monitoring project was implemented in the late 1960’s using IBM 1800 computers. Since then the Company has developed three unique monitoring systems: • • •

CHEVMON CHEVSCADA MODSCAN

Each of these three systems has its own niche within the corporation. The CHEVMON application area has been the process plants within refineries and chemical plants. CHEVSCADA and MODSCAN systems are applied to platforms and pipelines, respectively. Approximately 30 CHEVMON, MODSCAN, and CHEVSCADA systems are installed in numerous locations within the corporation. The CHEVSCADA and MODSCAN systems have powerful discrete control capabilities and can perform some limited continuous control functions. They can also monitor process variables. Systems used in this way are known as Supervisory Control and Data Acquisition (SCADA) systems, and are discussed in Section 1043, “Computers for Discrete Control.” The features of the CHEVMON system are discussed in more detail below. ETD’s new monitoring system, UNICORN, will incorporate the features of CHEVMON, CHEVSCADA, and MODSCAN into a single system. The full integration of MODSCAN and CHEVSCADA into UNICORN is scheduled for the early 1990’s. UNICORN will run on a network of DEC-VAX computers. It will have control capabilities similar to the COSMIC system and superior graphics and information management. UNICORN is not designed as a replacement or substitute for distributed control systems. The software is being developed in-house. Some third-party software, such as a relational database, is being added. CHEVMON, MODSCAN, and CHEVSCADA will continue to be supported after UNICORN is released. A number of other monitoring systems are also used in the Company. All of these systems depend upon operators and laboratory personnel entering the process data into the computer; that is, they are off-line, rather than real time systems. Data are often entered once or twice a shift; by comparison, the real time systems sample the process data every few seconds. These off-line systems typically use DEC-VAX computers. A system used at a number of Company locations is the Refinery Information System (RIS). The data from off-line systems are used for engineering, planning, and management studies.

Chevron Corporation

1000-31

June 1989



The future in monitoring systems is directed toward large plant-wide and refinerywide information networks. These networks will allow all operating areas within a geographical area to share the same process data. Distributed control systems and data gathering equipment will provide process information to individual monitoring systems. These monitoring systems are then networked together, allowing all process data from an entire refinery to be accessed in one location. The networks used for these information networks will be local area networks (LANs). The networks at those locations using CHEVMON and the new UNICORN systems will be Ethernet-DECNET based. Other sites may use networks based upon the standard GM-MAP protocol. CHEVMON. This section provides an overview of CHEVMON. Additional information is available from ETD’s Monitoring and Control System Division. CHEVMON is a powerful and flexible process monitoring and information system that is intended to help Operations improve the economic operation of their facilities. Generally, benefits are generated from the intelligent use of data that are automatically collected. Operations staff are able to make changes to the process in a timely manner. The use of historical data also allows the evaluation of plant performance and potential changes to a plant or its operation. Some of the specific areas of benefit are as follows: •

Closer control of operations by providing the operator with a powerful alarm subsystem

•

Reduced utilities consumption by providing real-time information on supply and demand

•

Ability to automatically calculate key operating variables on a continuous basis

•

Improved process evaluation capability provided by flexible history, trending, and reporting subsystems

•

Automation of routine operating and management reports

•

Improved ability to analyze plant upsets and mishaps by use of the event history subsystem

The CHEVMON system runs on any of the family of DEC-PDP products. A number of important features are standard parts of CHEVMON systems, including the following.

June 1989

1.

An operator console serves as a window into the process. The console can be either a 13-inch or 19-inch CRT. The number of consoles depends upon the needs of the operating unit. The color displays include trends and custom graphics.

2.

Laboratory data may be entered into the system in a variety of ways. One option is by direct connection with a laboratory computer system.

3.

Scanning subsystems have been developed for many types of instrumentation. Some of these systems are Taylor MOD 3, Foxboro Spec 200, Modicon, tank gauging equipment, and DEC computer interfaces.

1000-32

Chevron Corporation



4.

The alarming subsystem offers considerable flexibility. Three levels of alarm priorities are supported along with different alarm modes. For example, alarms can be configured for startup, steady operation, and shutdown modes.

5.

The history subsystem saves data for all measured and calculated variables for periods of time established by the system owners. The history subsystem also includes a number of powerful and flexible methods to access the history database. One-minute trend data is the maximum resolution.

6.

Event Logs record all events related to alarms and all operator commands to the system.

7.

The Expression Package is a powerful data retrieval and calculation tool that is used to configure customized graphics and reports.

8.

The Report Generator is a tool that allows the user to quickly format and schedule reports.

9.

Test Mode is a framework in which application programs (written in FORTRAN) can quickly be tested and debugged.

10. The Data Analysis Package provides tools for engineers, such as data reduction and X-Y plots. Statistical Quality Control (SQC) is also supported. 11. Links to IBM personal computers that allow the process engineer easy access to history data are provided. Data are requested by tag ID and structured so that they can be easily incorporated into spreadsheet programs such as Lotus 1-2-3. Any number of custom applications may be added to basic CHEVMON systems. A few of the applications that have been recently implemented include the following: Tankage, Blending, and Oil Movement • • •

Inventory Reports Tank Watch List Mogas Blending

Utilities and Operations Coordination • • • •

Steam balances Cogen Turbine Performance Monitoring Refinery Stockloss (Stock Transfer Recording) Effluent Diversion “What If” Scenarios

Process Surveillance

Chevron Corporation

•

Predictors for product properties, such as API, distillation cutpoints, and freeze and flash points

•

Hydrogen balances

•

Coker furnace fouling model

•

TKN reactor catalyst deactivation model

1000-33

June 1989


•


Boiler load allocation

There are currently 20 CHEVMON installations within the Company. The installations include multiple systems at some of our larger refineries and some relatively small systems at such locations as the Salt Lake Refinery, the Gavitoa Gas Plant, and the Hawaii Refinery. A network of CHEVMON computers within a refinery greatly increases the potential value of the information that is available. Standard software is available that allows access to data on any one of the systems, allowing refinery shift coordinators to look at critical information for more than one unit at a time. The software also makes it feasible to do process engineering studies of multiple units. The CHEVMON systems communicate with each other via DECNET over an Ethernet communication pathway. Figure 1000-17 shows the CHEVMON networks at the Richmond, El Segundo, and Pascagoula refineries. Fig. 1000-17 CHEVMON Networks

Connections to COSMIC control computers are also a standard feature. These links are primarily used to download laboratory information for control purposes after the information has been verified by Operations. A third type of connection is a link to corporate computers in San Ramon. UNICORN. UNICORN is an acronym for Unified Information & Control Oriented Real-time Network. It is the name of a computer system under development by

June 1989

1000-34

Chevron Corporation



ETD’s Monitoring and Computer Control Division that will fill the sensor-based process computing needs of the Company in the 1990’s. This is a preliminary discussion of some of UNICORN’S features; they may change before its release in 1989. The primary use of the UNICORN system will be for wide-area, sensor-based information management, advanced monitoring applications (like batch tracking, leak detection, or tank monitoring), and SCADA applications (such as well or pipeline monitoring and control). A special node is being designed for standard continuous control functions as well as some advanced control functions, such as dynamic matrix control (DMC). The UNICORN system is designed to enhance the features of a DCS, not to replace it. The general function of the UNICORN system will be to continuously acquire data in real time from a variety of sources, check the data for accuracy and other limits, and then put the data into a database. The system will also provide access to the database for programs that interact with the various users of the system, facilities for creating and scheduling reports, historical storage of values entering the database, and facilities for maintaining the database and other components of the system. Each UNICORN system will typically be constructed of two or more low-cost microcomputers (any of the DEC-VAX product line) linked via local area networking hardware and software. The programs that make up the system can be distributed among the various computers in order to maintain a high level of performance and to allow the isolation of certain critical programs into the more protected environment of a specific processor. Certain programs will be configured to run on multiple computers, so that in the event that one computer fails, another can immediately assume the failed machine’s duties. Other programs will be configured to run on whatever computer is “most available” at the time. It will be possible to configure a system to run on a single computer. This will normally be done for small systems with only a few hundred variables. Such systems could later be easily expanded by the purchase of additional computer hardware and the configuration of the new variables. It will also be possible to link a number of UNICORN systems into an area-wide network. Within such a network, application programs will be able to access the database of any connected system just as easily as they can access the database of the local system. UNICORN’s architecture will allow a number of smaller, geographically separated systems to appear to the user to be a single large system.

1043 Computers for Discrete Control The term discrete control computers refers to those computers used above and beyond the standard data acquisition and control system such as the PLC. The type of computer system used for discrete control systems within the Company are Supervisory Control and Data Acquisition (SCADA) systems. SCADA systems are computer systems that allow the control and monitoring of a process in remote areas from a centralized location. From this centralized location an operator can

Chevron Corporation

1000-35

June 1989



view the entire process through graphic and trend displays and initiate commands and setpoint changes to the field. SCADA systems are mainly dedicated to monitoring and discrete control applications, but they can also perform some simple continuous control functions (Figure 1000-18). Fig. 1000-18 Major Components of a SCADA System

The distance between the centralized location and the individual process can be quite large, covering many miles. Communication is therefore a key ingredient of a SCADA system. Because SCADA systems are capable of passing data over large distances, they fulfill the application needs of the Company’s pipeline and producing platform operations.

SCADA Components Following are the four major components of a SCADA system (see Figure 1000-18): • • •

June 1989

Master Station Remote Units Communications

1000-36

Chevron Corporation


•


Man/Machine Interface

Master Station. The master station is the heart of a SCADA system. The master station is a computer that is responsible for the following: • • • • • • • • •

Scanning of field devices Operator interface Alarm annunciation and logging Engineering configuration Preparation of reports Control Special applications Calculations Data trending and archival functions

The master station consists of both hardware and software. The hardware components include a computer, printers, display terminals, and disks. The software components include the operating system, the SCADA software, and custom software designed especially for a particular application. Remote Units. The remote units are the interface to the field devices. These units can be used for simple data acquisition or for data acquisition and control. The two types of remote units are remote terminal units (RTUs) and programmable logic controllers (PLCs). Remote Terminal Units (RTUs). Remote Terminal Units act as a collection station for the field instrumentation. Field signals are connected directly to terminal blocks at the RTU. The data at the RTU is then scanned by a remote communication link that connects with the master station. With the advent of the microprocessor chip, RTUs have become “intelligent” or “smart.” In many cases “smart” RTUs now accomplish calculations that were once provided by the master station. RTUs are now able to provide discrete control conversion to engineering units, alarming, and engineering calculations. Programmable Logic Controllers (PLCs). Programmable Logic Controllers also act as a collection point for field instrumentation, but the PLC has far more capability than the RTU. PLCs provide fast digital control in the form of ladder logic and can also provide limited analog control. RTUs vs. PLCs. While these devices overlap in function, some general observations can be made regarding the choice between an RTU and a PLC.

Chevron Corporation

•

PLCs generally need a more controlled operational environment than RTUs.

•

PLC vendors tend to be more reliable than RTU vendors. PLC vendors have a much better track record than RTU vendors where longevity and upward compatibility of the product are a concern.

1000-37

June 1989



•

PLCs are more capable of performing control than RTUs. PLCs package control modules with analog and discrete capabilities. PLCs perform discrete control in millisecond time spans.

•

RTUs are typically more cost effective than PLCs for monitoring applications. This gap is quickly narrowing, but in general this is still the case especially when there is a large amount of analog I/O signals.

•

RTUs are low power consumers. Solar-powered RTUs utilizing radio communications are becoming popular in applications (pipelines) that lack power sources. This enables RTUs to be remotely located in environments that do not have a power source.

•

RTUs are usually completely packaged by the vendor. Power supplies, NEMA cabinetry, and heat sources are included. PLCs requiring packaging as remote devices usually need to be taken to a third-party vendor.

Communication. A master station will communicate with its remote stations every 10 to 60 seconds. There are a variety of protocols used to communicate between the master and remote stations. A protocol that many vendors are using as a “defacto” standard is MODBUS. The Instrument Society of America (ISA) is currently developing a standard communications protocol between field devices (smart transmitters) and intelligent masters (RTUs, PLCs, and computers). This standard is called SP-50, “Field Bus.” SP-50 standards may be introduced in early 1991. A number of methods and communication equipment can be used to communicate between the master station and its remotes: •

Direct Connect. Master stations can be directly hardwired to field devices. Typically this is not feasible for remote communications. This method can be used when there is a short distance between master station and the field device, between 5 and 2000 feet.

•

Phone Connect. The use of public telephone lines is the most common communications media. Phone connect applications require modems at each end of the phone line. Phone communications typically restrict the transmission rate to 1200 baud.

•

Microwave Connect. This application is very similar to the phone connect application. This application is used when dedicated phone lines are not available at the remote site. It is a popular method of communication used by the Company’s offshore producing platforms.

•

Radio Connect. Ultra High Frequency (UHF) can be used in relatively flat areas, in a small radius from the master station. An omnidirectional antenna and a base station transceiver are located at the master station. A radio and directional antenna are located at the remote site. An FCC license is required for this application. The license typically takes 6 months to procure.

All projects requiring remote communications should contact the Communications Technology Department (CTD) of Chevron Information Technology Company (CITC) as early in the project phase as possible. They will help determine the

June 1989

1000-38

Chevron Corporation



proper form of communications required and assist in obtaining the required licenses. Man/Machine Interface. This is the hardware and software interface that allows the operator to interact with the SCADA application. Operator interfaces take the form of keyboards, touch screens, and trackballs. Devices used for reporting and displays include CRTs (color and monochromatic), printers, LED displays, annunciators, and automatic phone dialers. Many of the operator interfaces provide color graphics and data trending and archival functions.

SCADA Architecture Numerous, different system architectures are available for SCADA systems. The type of system architecture used depends upon the size and type of application. Distributed Control Systems. The distributed control systems (DCS) discussed in Section 1020 are beginning to enter the SCADA market by offering remote communications and SCADA applications. The DCS systems have interfaces available to many different PLCs. Some DCS manufacturers can also provide some type of remote I/O analogous to the RTUs. Distributed control systems do not have one master station. Instead, the functions of the master station are distributed among different processors. The DCS systems are capable of providing digital logic and also interfacing to PLCs that provide ladder logic functionality. Some of the DCS systems recently introduced into the market have ladder logic capabilities, but they in no way match the functionality of a PLC. In general, the distributed control systems have a more advanced operator and engineer interface than some of the other types of SCADA systems. Minicomputer. Minicomputer-based SCADA systems use a minicomputer as the master station. Minicomputers used most often include the Hewlett Packard (HP) ASeries, the Data General Eclipse, the Digital Equipment Company (DEC) PDP11’s, and DEC’s more powerful family of Micro-VAXs. The functions of the master station with this type of architecture include the following: • • • • • •

Scanning Data conversion Alarming Data archival functions Displays (support of multiple operator consoles) Computer-based applications

The master station will interface directly with remote units in a master/slave relationship. The master station sends requests for data to the remote units. The remote units are allowed to talk to the master station only when requested to do so by the master station. Minicomputer-based SCADA systems can also be designed to be redundant. Redundant systems have two identical master stations in constant communication with

Chevron Corporation

1000-39

June 1989



each other. One of the master stations will be designated as the primary, the other as the backup. If at any time the primary should fail, the backup master station will automatically become the primary master station. The peripherals are interfaced to the minicomputers via a hardware switch. If the primary master station should fail, the hardware switch will automatically switch all the peripherals to the backup master station. Figure 1000-19 provides a diagram of a typical redundant nondistributed minicomputer-based SCADA system. Figure 1000-20 shows a diagram of a typical redundant distributed minicomputer system. Programmable Logic Controller Network. PLCs can be networked together to form a SCADA system. These systems typically lack the functionality of a DCS or a minicomputer-based system. They do not provide historical trending, computerbased applications, or multiple operator stations. Most PLC-based SCADA systems can not communicate with RTUs and typically can only communicate with one kind of PLC, the vendor-supplied PLC. Microcomputer. Microcomputer- or personal-computer (PC)-based SCADA systems can be used to interface to smart RTUs and PLCs. PC-based SCADA systems are typically small and hence inexpensive. A disadvantage of a PC-based system is that it is often limited to four (and sometimes two) communications interfaces. PC-based systems also tend not to be as reliable nor provide the functionality of the DCS- and minicomputer-based systems. Alarm. The alarm systems are the simplest of the SCADA systems. They are typically PC-based and their man/machine interface is a single monochromatic screen and a printer. They are inexpensive but lack the functionally of all of the above types of systems. The alarm systems are used only for data and alarm monitoring.

SCADA Applications As mentioned earlier, SCADA systems within the Company are used exclusively in upstream applications such as pipelines and producing facilities. Particular applications include the following:

June 1989

•

Leak detection (net barrel line balance, pressure pack volume balance, pressure point monitoring, or predictive pressure profiling)

•

Batch tracking (monitoring the progress of an interface between two different segregations as they progress through a pipeline)

•

Well tests

•

Injection setpoint control

•

Tank monitoring

•

Reports (inventory, DOT, production, injection, shift)

•

Meter proving

•

Emission monitoring and reports

1000-40

Chevron Corporation


Chevron Corporation

Fig. 1000-19 Redundant SCADA System with Minicomputer

1000-41


June 1989


June 1989

Fig. 1000-20 Typical Redundant Distributed Minicomputer SCADA System

1000-42


Chevron Corporation



Company SCADA Systems The Company currently has two SCADA systems developed within the corporation. These systems are CHEVSCADA and MODSCSAN. CHEVSCADA is a minicomputer-based SCADA system running on an HP-1000 computer. CHEVSCADA is used for platform applications within the Company. MODSCAN is also a minicomputer-based SCADA system. MODSCAN runs on a DEC PDP computer. MODSCAN systems are used in Company pipeline applications. The Company is currently in the process of developing a new monitoring system that will incorporate the SCADA functions of CHEVSCADA and MODSCAN. Monitoring capabilities will be available on UNICORN in 1989, and SCADA functions are scheduled for availability in early 1990. Other information on the MODSCAN, CHEVSCADA, or UNICORN systems can be obtained from the Monitoring and Control Systems Division of the Company’s Engineering Technology Department.

1044 Computers for Continuous Control Almost all new instrumentation is based upon computer technology, ranging from the small microprocessor-based single loop digital controllers to the sophisticated and powerful multi-processor distributed control systems. This section discusses those computers that are interfaced to the standard instrumentation system. These computers tend to be more powerful than the base level instrumentation and can therefore provide additional control opportunities. In the past, most computer control systems were centralized processing systems, in which only one processor performs all the tasks ( Figure 1000-21). The tasks are not shared among processors as in a distributed processing system. If the main processor fails, all functions of the system are lost. Because of this, these systems are not directly connected to the process. They are interfaced with instrumentation systems to provide a fall-back operating position. Fig. 1000-21 Centralized Processing System

Chevron Corporation

1000-43

June 1989



Today, many computer control systems use distributed architecture, which enables the creation of a more powerful system with individual nodes on a network responsible for specific applications. An interface to an instrumentation system such as a DCS is still required with the new distributed control computers because such systems lack the redundancy features of an instrumentation system. A computer control system’s main function is to provide the mathematical power for advanced control applications and optimization routines. Because these functions of the computer control system are not critical for the operation of the plant, the plant can still operate when the computer is not available. Many of the computer control systems provide a separate CRT-based operator interface. In the past this was a necessity, because the display capabilities of the older distributed control systems were somewhat limited. Additionally, these computer systems were sometimes interfaced to panel-mounted instruments. The display systems of today’s distributed control systems are quite sophisticated. The type of screens or operator interfaces that the operator has to learn and to work with should be minimized. One interface would be ideal and no more than two should be used when a distributed control system is interfaced with the control computer. Some of the distributed control systems allow displays from a remote source to be called up on the standard operator’s console. In general, the computer control system need only send setpoints, calculated variables, and operator messages over the interface to the DCS. Those systems that have panel-mounted instrumentation as the front end need to have the operator interface provided by the computer control system. Control computers provide two levels of control: • •

Direct Digital Control (DDC) Supervisory Setpoint Control (SSC)

With direct digital control the control algorithm calculation (PID) is done within the computer. The computer then sends the valve position signal directly to the control valve, through the control system. In supervisory setpoint control, the setpoint is calculated in the computer and then sent to the control system. The control system then performs the control algorithm calculation and determines the proper valve position. Figure 1000-22 provides a graphical comparison of DDC to SSC. Many of the original computer control systems used DDC because the early instrumentation was not designed for SSC. The need for SSC arose because the computer control system is typically much slower than the instrumentation system, sometimes by a factor of four or more. This caused some control problems for fast loops, such as flow. Many of the older computer systems installed within the Company now use a combination of DDC and SSC. SSC is typically used with today’s modern distributed control systems. The cycle time of a loop on a modern DCS is 0.1 to 0.25 seconds, compared to 1 to 4 seconds on the control computer.

June 1989

1000-44

Chevron Corporation



Fig. 1000-22 DDC vs. SSC

It is always recommended that DDC be connected to an intermediate controller and not directly to the valve in the field. Computer control systems are typically nonredundant and backup control is required in case the computer fails. Two types of computers are used for computer control systems: • •

Minicomputers Microcomputers

Minicomputers and Advanced Control. Minicomputers provide the mathematical power for some of the advanced control functions and optimization routines. Minicomputers used within the Company for advanced control include the DEC-VAX, Taylor 3106, and Honeywell 4500. In the late 1970’s, the Company developed its own computer control system based on the Taylor 3106. This system is called Command Oriented System for Modern Industrial Control (COSMIC). Approximately 24 COSMIC systems are in use throughout the corporation. These computer systems are used exclusively within Chevron USA and to some extent Chevron Chemical. COSMIC systems provide a well-developed operator interface as well as special tools with which the engineer can perform configuration, testing, and system management functions. One of the key features of the COSMIC system is that it has a simple and informative operator interface. The state of every control loop is displayed to the operator along with the reasons why a loop is not in its normal state. One of the most attractive features of COSMIC is that it provides the operator with as much information about the process as possible on a single display. It has been Company policy since 1986 to do no further COSMIC development. The last COSMIC system was installed in 1987. Many of the functions found on

Chevron Corporation

1000-45

June 1989



the COSMIC system are now an integral part of distributed control systems. However, although the distributed systems do provide many of the COSMIC functions, many other COSMIC functions are missing. The Company’s new monitoring system, UNICORN, will provide the advanced control platform for the corporation and in addition will provide an upward migration path for the current COSMIC systems (COSMIC II). COSMIC on a VAX will be available in mid-1989. In addition to COSMIC II, the UNICORN system will also provide utilities to perform advanced control functions such as DMC. These utilities will be purchased from outside vendors. Figure 1000-23 provides a diagram of a proposed UNICORN system showing its many functionalities, including control, planning, scheduling, training, and modeling. The UNICORN system will provide a refinery-wide information system that will allow a single interface to plant information for the operator, engineer, manager, planner, and scheduler. Most distributed control system vendors provide an interface between their instrument system and the DEC-VAX. The new UNICORN system will be DEC-VAX based. The Company has many advanced control strategies available. Figure 1000-24 provides a list of the some of the computer applications available for specific process units, although not all are applicable for a given plant. Advanced control strategies should be configured so that they include but are not limited to the following features: • • • •

Limit Checks Measurement Validity Checks and Actions Secure Cascade Configuration Bumpless/Balanceless Operation

The operator interface should always display the state of the control strategy no matter where the strategy is performed. In addition to the strategy state, the operator interface should also provide messages to the operator about the current state, such as what it is doing and why it is doing it. Appendix F provides an example of an advanced control strategy specification. Any questions concerning advanced control applications can be directed to the Monitoring and Control Systems Division of the Company’s Engineering and Technology Department. Microcomputers. Personal computers fall into the microcomputer category. Personal computers (PCs) are relatively new in the control industry. They have made great advancements over the past few years. Typically, most control software is written for IBM PCs and IBM clones. Some new control software for the Apple Macintosh has recently been introduced, but it does not approach the amount of software available for the IBM. Software packages are available that interface to an extensive list of instrument systems. These software packages provide relatively good graphic displays, control,

June 1989

1000-46

Chevron Corporation


Chevron Corporation

Fig. 1000-23 Chevron’s Future Advanced Control Platform

1000-47


June 1989



Fig. 1000-24 Computer Applications Available for Specific Process Units. (1 of 3)

June 1989

Process Unit

Advanced Control Application

1.

Rheniformers

Octane Control Recycle Gas Control Maximize Feed Rate Stabilizer RVP Control Stabilizer Pressure Minimization Minimize Hydrogen Bleed Maximize C4/C3 in Stabilizer Bottom Multifurnace Automatic Combustion Control

2.

Crude Units

Maximize Heat Recovery in Crude Preheat Train Water Wash-to-Crude Ratio Sidecut Product Quality Control Degree-of-Fractionation Control Column Pressure Minimization Stabilizer Bottoms RVP Control Vacuum Column Bottoms Quench Control Flash Point Control Jet Quality Control Mode Control Automatic Combustion Control

3.

Hydrocracker

Catalyst-Average-Bed Temperature Control Recycle Gas Control Fractionator Product Quality Control Degree-of-Fractionation Control Column Pressure Minimization Automatic Combustion Control

4.

Hydrotreating

Catalyst-Average-Bed Temperature Control Recycle Gas Control Sidecut Product Quality Control Degree-of-Fractionation Control Column Pressure Minimization Automatic Combustion Control

5.

Hydrogen Plant

Steam-to-Carbon Ratio Control Hydrogen Production Control Conversion Control

6.

Furnace

Furnace Feed Ramp Furnace Pass Balancing Automatic Combustion Control

1000-48

Chevron Corporation




Chevron Corporation

Process Unit


7.

Aromatics

Minimize Xylenes in Light Reformate Toluene in Whole Xylenes Control Liquid-to-Vapor Ratio Control % C9+ in Whole Xylene Control Minimize Xylenes in Heavy Isomerate Control Heart-Cut Xylene Excess % C9+ in Heart-Cut Xylenes Control Minimize Xylenes in Light Blending Aromatics Toluene in Heart-Cut Xylenes Control

8.

Alkylation

% iC4 Recycle-to-Olefins Ratio Control Optimize DIB Sidecut Rate DIB Recycle-to-Feed Ratio Control Depropanizer Bottoms Control Debutanizer Overhead Control

9.

Ethylene

E/P Blending on Analyzer Control Feedstock Selection Control Depropanizer Bottoms and Light Key Control

10. Solvent Extraction and Dewaxing

Treater Temperature Control Solvent Ratio Optimizer

11. Fluid Cat Cracker

Feed Component Control Riser Temperature Control Reactor Pressure Control Regenerator Temperature Control Gasoline 95% Point Control

12. NH3 Plant

H-to-N Ratio Control Methane Leakage Control Steam-to-Carbon Ratio Control Feed Maximization

13. Fertilizer Plant

Fertilizer Formula Corrector Acids-to-Ammonia Ratio Control Neutralizer pH Profile Control Ammonia Distribution Control Nitric Acid Distribution Control Phosphoric Acid Distribution Control

1000-49

June 1989




Process Unit


14. Residuum Desulfurizing

Maximize Hot VGO Feed Minimize Gas-to-Oil Ratio Recycle H2 Compressor Speed Optimizer Hot Low Pressure Separator Temperature Control Reactor Heatup/Cooldown Ramp Reactor Catalyst-Average-Temperature Control

15. Delayed CokerAutomatic

Quench Cycle Control Maximize CHDN Feed Temperature Recycle Ratio Control Lean Oil-to-Offgas Ratio Control Sponge Oil-to-Offgas Ratio Control Debutanizer Overhead Control Debutanizer RVP Control

data trending, and report capabilities. Personal computers can be used for monitoring and operator interface functions only, or they can be used for relatively simple control applications. The use of supervisory setpoint control with personal computers is strongly recommended. The PC should not be used as a standalone control system connected directly to the process or as the only operator interface to the process. PCs are not reliable enough. Personal computers should be interfaced to single loop controllers (SLCs). If a PC is used for monitoring functions only, the I/O interface can be quite simple and the SLCs are not required. Figure 1000-25 shows a personal computer interfaced to a number of single loop controllers. Fig. 1000-25 Control System Using a Microcomputer and Single Loop Controllers

June 1989

1000-50

Chevron Corporation



Even though personal computer-based systems can handle larger point counts, it is recommended that personal computers should only be used when the total I/O requirement is between 50 and 100 and the control requirements are quite simple. Anything larger or more complex should be handled with a small distributed control system.

1050 Characteristics of Successful Projects 1051 People Make The Project A successful control project must include the cooperative actions of everyone who interacts with the process, from the facility manager to the newest hire. It is not enough to have a good installation and start-up. The control system must stay in use continuously in order for the desired project benefits to be realized. The people and the process will change and someone must care enough about the control system to explain it to the new personnel and adapt it to the new conditions. A feeling of shared ownership by all parties involved is the best means of keeping a control system in operation. This attitude of shared ownership is developed through understanding of the goals of and reasons for the control, participation in developing and improving the system, and continued feedback on how well the system is working and why the system may not be working.

1052 Control Objectives Analysis (COA) Control Objectives Analysis is a procedure that produces a clear set of objectives for a control project and it also is a means of getting a variety of people involved in a control project in its early stages. The original article describing the Control Objectives Analysis (COA) procedure is included as Appendix G. Three or more people representing Operations, Process Engineering, and Control Engineering work with a moderator to develop carefully worded control objectives that equal the number of control valves in the process area being considered. The control objectives must be concise, precise, and true-all-of-the-time statements of why process valves are moved. A simplified process diagram is used to show all of the control valves on the unit. The valves in the process diagram are numbered and this defines the number of control objectives which need to be developed. This number may change as the objectives are developed. A valve which is always open or closed may be crossed off the control valve list, or a variable speed pump may be missed on the first count but added later. An initial statement of the overall objective for the process is developed. The overall objective often develops in a rambling way, so that everyone tends to get involved and works together on the COA. The overall objective is reviewed again at the end of the COA. The contributions of a knowledgeable operator to the COA have proven to be invaluable. The operator will remember a number of problems on the unit that can then receive immediate engineering attention. The operator’s views enhance the control objectives by providing some elements of basic reality. Word quickly

Chevron Corporation

1000-51

June 1989



spreads to other personnel on the shifts that operator concerns are being included in the early planning for the new control system. The control objectives may then be categorized into three types: Type I:

Hold material balance or heat balance. These objectives are requirements of the process and the valves can not be manipulated for other reasons.

Type II:

Hold operations at management-set targets. These objectives are sometimes seen as being too restrictive for operations or as not demanding enough from the unit.

Type III:

Balance one objective against another: optimize, minimize, or maximize. These objectives may be used to optimize the process benefits through improved control engineering and planning.

The Type III control objectives show where the economic benefits are for the process control. Constraints on process equipment, product qualities, product markets, resource availabilities, and uncontrolled outside disturbances determine the upper limits of the achievable benefits. The potential benefits should be defined as carefully as possible. This helps determine whether or not a project should continue and provides a standard for measuring the final success of the project.

1053 Design of Controls The final control elements and their locations are affected by the control scheme selected for the process. Often the control scheme is selected because it is used elsewhere. The Process Control Group of ETD is available to review control drawings and advise on the various options which may be used to control any process. A number of engineering tools are available for helping in the control design. The Relative Gain Analysis is a tool useful for distillation columns. A computer program that runs on an IBM PC is available from the Monitoring and Control Systems Division of ETD. A number of steady-state model runs will provide excellent information on coupling of valves with process measurements and on interactions within the process. A dynamic simulation of the process is possible when the process becomes complex and little useful experience is available. The designed controls should be oriented toward the way that operators think about the process. Starting a facility, turning on a few regulatory controls, and then switching to advanced controls that optimize the process should seem like a natural progression. When the advanced controls require a fair amount of interaction between control loops, the final system must be engineered so that the operator is protected from the system subtleties. Another ideal is to provide the controls with “graceful degradation.” This implies that the controls do the right things when some parts of the system are not working. A measurement (such as a value from an analyzer) may be unavailable or a valve may be on manual by-pass, but the rest of the control scheme will keep the process running in a stable and desirable state.

June 1989

1000-52

Chevron Corporation



One advantage of the modern DCS control equipment is that the control scheme may be easily changed as new information is obtained or new requirements are defined. This luxury should not be used as an excuse for not planning for the best engineered controls and equipment installations.

1054 Operator Involvement Operators are often concerned about a new control system. The fear of the unknown can be reduced by keeping operators informed on almost any aspect of the project, from analyzers to key-boards to schedules. Training is, of course, essential prior to any startup activity. The participation of one or more operators in the project development helps in achieving acceptance of the completed job. The COA is a good opportunity to start getting the operators involved. Many projects have benefited from encouraging the operators to develop some of the graphic displays and trend groupings so that they have a part in the project development. Operations supervisors are usually deeply involved early in a project but may forget how to do some things once the system is running well. Some refresher training can help them to feel comfortable in their control room again.

1055 On-Site Support Our goal is to make each location self-sufficient in its ability to support the control systems after they are installed. This is not easy to do, considering the complexity of the systems and personnel turnover. One of the most significant aids is the use of the same hardware system for several projects, which provides some cross-coverage for instrumentation and engineering support. People will have to be trained every year and some contract maintenance will be required for some of the more sophisticated equipment. Laboratory support must also be considered. We are running less frequent analyses in our plant laboratories even though the variability in the feeds is greater than ever. Each analysis is used as the reference for more volume of product and the analytical results are being fed directly into the control systems and used to adjust the processes. The timeliness and quality of the laboratory results must be continually emphasized and monitored.

1056 Monitoring the Results To maximize the economic benefits of a process, we usually are pushing against some constraints. These restraints can cause more work than is required in a unit that is not under tight control. As a result, some controls may be turned off and the others circumvented unless the system is routinely monitored. One of the standard monitoring tools within the Company is to have the control system monitor the percent of time each control loop is in use. A record is also needed that explains why the controls are not in use. Equipment maintenance is a common reason. At times, a certain level of quality for a product may not be required or a particular control may be disturbing the whole process. The combina-

Chevron Corporation

1000-53

June 1989



tion of the in-use time record and the explanation record allows the supervisors to take reasonable actions to keep the process in good operation. However, note that a good on-line time record does not necessarily mean that the unit is achieving the control objectives.

Monitoring Utilities and Yields Most control projects are designed to produce some savings of energy and/or produce more valuable products. Rarely can either of these benefits actually be measured on the facility flow meters. Feed qualities change, feed rates change, equipment constraints develop, and different fuels are used. The results are often ambiguous unless process modifications were made or a significant change in procedures was developed. It is also common for the operators to become more interested in the unit when they see process studies being conducted and hear of the results of a COA. The unit is then running very well before the new control system is installed, which presents a formidable challenge to the control engineer.

Monitoring Product Qualities The analytical results usually provide a clearer indication of an improvement in the process control. For a well-controlled process, the operators are quite confident of their products. The laboratory results provide a confirmation and only occasionally does an unexpected result trigger trouble-shooting activities to locate a previously unnoticed process disturbance. The product qualities may be monitored by the techniques developed in Statistical Quality Control (SQC), but the X-bar and R charts become quite boring on a well-controlled process. In a number of processes, the inherent error introduced by the sampling procedures and in the laboratory results has become an important consideration in the observed product quality variations. Laboratory personnel need to be involved in resolving these problems, which are fairly complicated. We must avoid trying to adhere to product specifications that are more stringent than we are capable of measuring, but we must push toward the limit of those specifications.

1060 Selection Criteria for Computer-based Instrumentation and Process Computers It is difficult to put together a list of selection criteria that will satisfy the needs of every project. The applications, size of the system, location, and local support staff differ from project to project. Four categories of selection criteria are discussed in a general way in this section: • • • •

Hardware Software System integrity Vendor

A relative scale of importance should be assigned to the items listed below for your particular project. A review of these items will help you achieve a successful

June 1989

1000-54

Chevron Corporation



project. Before selecting a system, consult the Monitoring and Computer Systems Division of the Engineering Department.

1061 Hardware The following hardware issues should be considered before purchasing a system. State of the Art Technology. How long has the product been operating in the field? Traditionally, the Company requires a product to have 1 to 2 years of operating experience before it is installed. Hardware Expansion. Determine the maximum configuration for the system and make sure that there is room for future expansion with your current configuration without incurring unreasonable costs. System Modularity. Many of today’s systems have interchangeable parts, which reduces repair, technician training, and spare parts inventory costs. Operator Interface. The interface is of extreme importance; the operator spends all day working with it because it is the primary source of information about the process. The interface should be consistent, quick, and easy in allowing the operator to get from display to display. A standard operating display should appear on the screen within 2 seconds and should not take any more than two button pushes to select. Special attention should be given to the ease of use and layout of the keyboard. Interface with Other Equipment. Links to other instrumentation systems, computers, analyzers, etc., must be considered. The expense of both the hardware and the software needed to accomplish the interface must be considered. Distance Limitations between the Individual Modules. Due to communication and system design limitations there is often a maximum allowable physical distance between modules on a network. Additional hardware, special communication devices, or special cabling may be required to satisfy specific job requirements. Environmental Requirements. It is necessary to pay attention to the type of environment the system is to be installed in. Special design considerations may be necessary for temperature, humidity, and hazardous conditions. Special Power Requirements. The type and amount of power required for the system must be determined. Determine if an uninterrupted power supply is required. On-Line Failure Diagnostics. Does the system have on-line hardware diagnostics or does the system have to be taken off-line to perform the diagnostics? Failure Detection Capability. The system should be able to detect all hardware and software failures and report back to the operator in a clear and precise manner. Redundancy. If individual parts of the system fail, what is lost, and is the loss acceptable? If the system is redundant, does the backup takeover automatically upon failure?

Chevron Corporation

1000-55

June 1989



On-Line Repairs. Can circuit boards be removed on-line or must the system be turned off to do repairs? Intrinsic Safety. Does the environment the equipment is to be installed in have intrinsic safety requirements and does the equipment meet them? Peripheral Equipment. What peripherals are necessary? Which come as a standard part of the system and which have to be specially purchased?

1062 Software Software determines the real power of today’s modern control systems. Software design and testing is very time consuming; hence a good software package is extremely important. The following software issues should be considered before purchasing a system. Continuous Control Functionality. Is this a requirement for your system? Check how easy it is to configure and implement continuous control functions. Are there an adequate number of predefined algorithms? Items to be especially aware of include the ease of changing the range of an instrument, setting up cascade control properly, changing tuning constants and alarm limits, and changing important configuration parameters while on-line. Scan and Alarm Capability. What is the fastest scan frequency available and what percentage of the database can be scanned at that frequency? Is it possible to take points temporarily out of service? Is there a facility in the system that informs the operator which tags are out of service? How easy is it for the operator to tell when a point is in a state of alarm and how quickly can the alarm be acknowledged? Does the alarm display provide adequate information? Subsystem Communication Capability and Compatibility. Can one subsystem initiate a task in another? Can processors in a network access other’s peripherals? Can processors share databases? Can remote processors detect and act on the failure of another processor? Does the operating system ensure that the network requests do not swamp the processor? High Level Supervisory Control Language. Is there a higher language available such as Fortran, Pascal, C, or a special vendor-developed control language to perform advanced control and optimization functions? The supervisory programs should be able to access history data and have access to some type of file structure. On-Line Program Generation and Modification. Is it possible to modify control programs while on-line? Support Software. Are utilities available that allow the engineer to quickly and easily load and save the database and supervisory programs? Are special engineering utilities provided to cross reference portions of the database, print the database in different formats, and make multiple edits of the database?

June 1989

1000-56

Chevron Corporation



Program Scheduling Capability. What is the maximum number of programs that can be scheduled at a certain frequency and what will the impact be on system loading? Historic Database Functionality. What is the trending capability of the system? What is the data capture frequency and how much data can be stored? How easy is it for the operator to call up its displays and for the engineer to configure it? Can it interface with control and user programs and reports? Can the history data be easily printed out in tabular form or can it only be displayed on the screen? Console Display Flexibility and Display Building Capability. Are there any standard operating displays? If so, do they provide useful information to the operator and how easy are they to configure? How easily can custom graphic displays be built and modified? Can displays be modified and downloaded on-line without taking a console out of service? Debug Utilities. Are there special tools for the engineer to test and debug supervisory control programs, ladder logic, and function block configuration without affecting the process? Database Maintenance. Do on-line configuration changes, tuning, and alarm limit changes get updated to the database source files automatically? Effective System Loading Monitoring Tools and System Status Displays. Does the system have adequate displays for determining overall system performance and in particular performance of the control functions? Does the system provide status displays with valuable information on the current system status for the engineer and operator? Tag Building Flexibility. How many unique tags can the system support? What is the maximum number of characters allowed per tag? At least eight is recommended and twelve is preferred. Batch Control Capability. If your application is a batch application, determine if there is a special software package for batch. Is it easy to implement the application and is it easy to modify it? How large can recipes be? Are various recipe formats available? System Restart Flexibility. Is it possible to configure how a system is to come up on return from a crash or power failure? Software Documentation. Is the system to be maintained by on-site personnel? If so, the following documentation from the vendor is recommended:

Chevron Corporation

1.

Application source code listings

2.

System software documentation and source code

3.

User manuals

4.

System hardware documentation

5.

System installation drawings, electrical drawings, and other drawings for maintenance personnel

1000-57

June 1989


6.


System build documentation and source

1063 System Integrity System integrity refers to the system’s response to detected hardware or software problems. Error recovery is just as important as error detection. The following questions should be asked concerning a perspective new system: •

What happens during a short- or long-term power failure? What is the transition time between the two, and how does the system respond in case the backup power supply is not available?

•

What happens if the bulk storage device holding the database and control programs has a failure?

•

What happens to all historic data during a computer failure?

•

How does the system detect corruption of the process database and control programs? What does it do when it finds the corruption?

•

What hardware integrity checks are built in? What happens if a check fails?

•

What software integrity checks are built in?

•

What happens if a processor fails? How is control passed to a backup or a remote processor?

•

Is the system protected against outputting erroneous values to field instrumentation?

•

What happens to outputs upon a failure, cold start, and failover?

•

What happens if the primary operator interface fails? Are there additional operator interfaces either locally or remote? Backup controls and displays at a remote site should be convenient to use.

1064 Vendor The choice of a particular vendor also has great impact on the successful completion of a project. The following information about the vendor supplying the system should also be considered. Ability to provide hardware and software assistance. Does the vendor have qualified personnel and sufficient resources to provide assistance in the areas required? Both during and after the system’s installation these resources will be required until the user has built up a level of confidence and knowledge of the system. Training Courses. How developed is the vendor’s training program? What type of classes are offered and what type of resources does the vendor put forth to support their training program? Schedule. Does the vendor have an organization capable of meeting your schedule demands?

June 1989

1000-58

Chevron Corporation



Vendor Reputation. What is the vendor’s reputation for long-term support of their product, the quality of field service staff, ability to meet schedules, product reliability, and software support and assistance?

1065 Other System Requirements Many modern control and computer systems have stringent environmental limits. Also, electrical power requirements, including standby power must be studied. Proper grounding of the control system, computer system, and electrical field instrumentation is critical to proper operation.

1070 References 1.

Moore, J.A., and S.N. Moore. Understanding Distributed Process Control (Instrument Society of America).

2.

Chevron Corporation. Distributed Control System Evaluation (Chevron Corporation Engineering Technology Department Publication, 1988).

3.

Instrument Society of America (ISA). “Intech, The International Journal of Instrumentation and Control,” Instrument Society of America (ISA Monthly Publication).

4.

Williams, T.J. Use of Digital Computers in Process Control (Instrument Society of America).

5.

Machulda, J. “Fault Tolerant Control: You Can Afford It” (Intech, Volume 32, Number 9, 1985).

6.

Sierk, H. “Methods of Comparison of Distributed Controls,” in Instrumentation in the Power Industry Volume 30 (Instrument Society of America, 1987).

7.

DeBow, B. “Retrofitting Distributed Process Control Systems” (Proceedings of the Instrument & Control Systems Conference and Exhibit, Instrument Society of America, 1987).

8.

Borer, J., Instrumentation and Control for the Process Industries (Elsevier Applied Science Publishers, 1985).

9.

Roffel, B., and P. Chin. Computer Control in the Process Industries (Lewis Publishers, 1987).

10. Mellichamp, D.A. Real-Time Computing (Van Nostrand Reinhold Company, 1983). 11. Noltingk, B.E., editor. Instrumentation Reference Book (Butterworths). 12. Chevron Corporation. Application Guide for PLCs (Chevron Corporation Engineering Technology Department Publication, 1989). 13. Jones, C.T., and L.A. Bryan. Programmable Controllers—Concepts and Applications (International Programmable Controls, 1987).

Chevron Corporation

1000-59

June 1989



14. Bryan, L.A., and E.A. Bryan. Programmable Controllers—Selected Applications (Industrial Text Company, 1987). 15. Instrument Society of America. “Programmable Controls—The User Magazine” (ISA Monthly Publication). 16. Chevron Corporation. Chevron Brochure (Chevron Corporation Engineering and Technology Department Publication, 1988). 17. Siler, W.R. “Designing LAN’s for a Chemical Company Tech Center” (Intech, Volume 34, Number 8, 1987). 18. Southard, R.K. “Putting Fiber Optics to Work in Industrial Applications” (Intech, Volume 34, Number 10, 1987). 19. Hughes, K. “Trends in LAN’s for Plant Automation” (Intech, Volume 35, Number 3, 1988). 20. Chevron Corporation. SCADA Evaluation Guide. (Chevron Corporation Engineering Technology Department). 21. Chevron Corporation. Pipeline Manual, Section 500, “SCADA Systems” (Engineering Technology Division, 1989).

June 1989

1000-60

Chevron Corporation

1100 Control Room Design Abstract This section describes the practices used for the design of control rooms for process plants and for onshore producing facilities. It explains the location of the control room in relationship to the facility processes and discusses pertinent architectural criteria. Design guidelines are prescribed for types of control panels and auxiliary control equipment. Contents

Page

1110 Introduction

1100-3

1111 Definitions of Control Room Types 1120 Background

1100-4

1121 Engineering and Coordination 1130 Design Parameters

1100-4

1131 Environmental Parameters 1132 Physical Parameters 1133 Utilities 1140 Prerequisites and Procedures

1100-5

1141 Design Documentation 1142 Local Control Room Prerequisite Data 1143 Central Control Room Prerequisite Data 1144 Design Procedure Sequence 1150 System Requirements for Control Room Design

1100-6

1151 Air-Conditioning and Filtering 1152 Pressurization and Purging Systems 1153 Utilities 1154 Gas Detection and Fire Detection and Extinguishing Systems 1155 Plant Communications Systems 1160 Design Concepts

Chevron Corporation

1100-8

1100-1

July 1999

1100 Control Room Design


1161 General Considerations 1162 Building Requirements 1163 Control Room Environment 1164 Pneumatic Tubing and Cable Design 1165 Layout Considerations for Control Panels 1166 Layout Considerations for Control Rooms Housing Distributed Control Systems 1167 Layout Considerations for Auxiliary Equipment 1168 Local Control Room Design Concepts 1169 Central Control Room Design Concepts 1170 Checklist and Final Documentation

1100-24

1171 Checklist for Control Room Design 1172 Final Documentation 1180 Sample Control Room Design Criteria

1100-30

1190 References

1100-34

1191 Model Specifications 1192 Other References

July 1999

1100-2

Chevron Corporation



1110 Introduction This section provides guidelines for the construction of new control rooms and for major modifications to existing control rooms for processing plants and onshore producing facilities. This section has limited applicability for the design of control rooms on offshore platforms, because they have requirements unique to the application. Local conditions, traditional methods of control room construction, and the specific requirements of the project and ultimate user will determine the extent to which the designer should follow this guideline. The designer is urged to utilize local expertise to supplement guideline recommendations.

1111 Definitions of Control Room Types For the purposes of this section, a control room is defined as an enclosed modular building or permanent structure at a major facility that can contain control panels, consoles, and associated control equipment. It can include a field termination (rack) room, computer equipment room, offices, utility and mechanical equipment rooms, and other facilities necessary for the efficient and safe operation of the facility. The two types of control rooms discussed in this section are local control rooms and central control rooms.

Local Control Room The local control room is used in specialized areas such as turbine-driven generator systems, laboratories, drilling controls, and furnaces or heater systems. The local control room may be manned or unmanned. The manned local control room may be occupied 24 hours per day, with occasional absences for field checking of process equipment and conditions. All or only selected critical parameters from the information gathered or generated in the local control room may be transmitted to a central control room depending on ultimate design decisions. Typically, the manned local control room is adjacent to the process it maintains. The unmanned control room contains control equipment that allows local monitoring and limited control of a unit, individual process, or small plant. Personnel will visit the unmanned local control room occasionally to perform maintenance checks or to obtain process information not available in the central control room.

Central Control Room The central control room monitors and controls process operations, fire and gas protection, and communications. The control room may incorporate one or more offices depending on project requirements. In addition, the central control room can gather process data from local control rooms distributed throughout a facility or from remotely located controls, such as analyzer sheds or remote terminal units, that are commonly found in process plants and on pipeline facilities.

Chevron Corporation

1100-3

July 1999



1120 Background 1121 Engineering and Coordination The engineering disciplines required for the design of a control room include the following: 1.

Instrument engineering, to design control systems, lay out the man-machine interface, schedule equipment components into enclosures, determine power requirements, and provide overall coordination for the other disciplines

2.

Electrical engineering to provide power distribution, lighting, and cable routing

3.

Architectural and structural engineering to provide room or building design in accordance with personnel requirements and blast resistant design if appropriate

4.

Safety engineering to provide detection and alarm systems and to assist in the selection of nontoxic and fireproof materials

5.

Mechanical engineering, to provide heating, ventilation and air-conditioning (HVAC) infrastructure, piping design for fluid utilities, and assistance with fire detection and extinguishing systems

1130 Design Parameters The following parameters should be considered in order to implement the control room design procedure.

1131 Environmental Parameters •

Climatic conditions at the site—minimum and maximum temperatures and relative humidity, maximum wind velocity, and prevailing wind direction

•

Ambient air quality—location of the control room relative to known or potential sources of airborne pollutants

•

Unusual phenomena—dust or sand storms, flash flooding, salt spray, etc.

•

Seismic zone classification

1132 Physical Parameters

July 1999

•

Weight and size restrictions for major equipment or modules to be installed at the site, and access by transport and cranes

•

Soil loading conditions (Major control house expansion might require pile setting of building and underground duct banks.)

1100-4

Chevron Corporation



•

Proximity of grade to water table, if underground cable vaults are planned

•

Drainage, and susceptibility of grade areas to flash flooding

1133 Utilities • • • •

Electrical power availability Potable water availability Instrument air supply Drainage and sewer line availability

1140 Prerequisites and Procedures 1141 Design Documentation Preliminary control room design is developed from the following documentation: 1.

Area electrical classification plan

2.

Plant or platform preliminary layout

3.

Control system conceptual layout

1142 Local Control Room Prerequisite Data 1.

Specialized functions, e.g., turbine generator control room, motor control center, orlist1ation

2.

Manned or unmanned

3.

Type of communications available

1143 Central Control Room Prerequisite Data 1.

Personnel requirements—operating, technical, and administrative – – – –

2.

Control, shutdown, and environmental monitoring systems requirements – – –

Chevron Corporation

Changing room Lavatories Lunchroom Maintenance area

Physical size, layout, and location Power consumption Interface requirements

1100-5

July 1999



1144 Design Procedure Sequence 1.

Review data obtained in Sections 1131, 1132, 1133.

2.

Obtain prerequisite data as described in Sections 1141, 1142, and 1143 as applicable.

3.

Review Section 1150 to determine systems required in the control room design.

4.

Complete the system design after reviewing the design concepts covered in Section 1160.

1150 System Requirements for Control Room Design The systems that may be included in the design of a control room are air-conditioning and filtering, pressurization and purging, fire detection and protection, gas detection, and communications (if required). These systems are discussed in the following sections.

1151 Air-Conditioning and Filtering An HVAC system is required in any control room containing personnel or temperature-sensitive electronic equipment. The HVAC system may also serve as the pressurization system. The HVAC system design should follow the recommendations in Specification ICM-MS-3651, Section 5.0. Window-mounted air-conditioners may be used in place of central HVAC systems in tropical and subtropical climates where heating is not required for personnel comfort, provided that they provide equivalent air filtering for personnel protection, are rated for the electrical area classification of the control room, and draw fresh air from an area that does not increase the hazard to personnel or equipment.

1152 Pressurization and Purging Systems Although the Company does build purged and pressurized control rooms and buildings, this is done primarily to prevent air contamination. Such rooms and buildings are normally located in unclassified areas only, an exception being offshore platforms. Offshore control rooms located in hazardous areas require positive pressure inside to keep the hazardous vapors or gases from infiltrating the control room. The design rules and code requirements for purging and area classification are extensive. The design engineer should refer to the following documents:

July 1999

•

National Fire Protection Association (NFPA) Standard No. 496, “Purged and Pressurized Enclosures for Electrical Equipment”

•

API RP 500 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Division 1 and Division 2.

•

API RP 505 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Zone 0, Zone 1, and Zone 2.

1100-6

Chevron Corporation



•

Instrument Society of America (ISA) RP 12.1, “Electrical Instruments in Hazardous Atmospheres”

•

Electrical Manual, Section 300, “Area Classification”

1153 Utilities The utilities required for most control room installations include electrical power, instrument-quality compressed air, and potable water and drain systems.

Electrical power systems Control room electrical power requirements can include any or all of the following: 1.

12 VDC, 24 VDC, or 120 VAC, all with UPS, for instrumentation and control systems

2.

12 VDC or 24 VDC for paging and alarm systems

3.

120 VAC for lighting and receptacles

4.

12 VDC, 24 VDC, or 120 VDC, all with UPS, for emergency lighting systems

5.

12 VDC, 24 VDC, or 120 VAC, all with UPS, for fire and gas detection systems

6.

208 or 480 VAC, 3-phase, 60 Hz for HVAC systems

7.

12 VDC, 24 VDC, or 120 VAC, all with UPS, for radio systems

8.

48 VDC with UPS for microwave systems

9.

120/240 VAC for computers and peripherals

For the purpose of the above list, a DC “UPS” consists of a battery charger, a storage battery bank, and sufficient auxiliary instrumentation to monitor and control their functions. The electrical supply and distribution system for the control room should follow the recommendations in Specification ICM-MS-3651, Section 4.0.

Water and Drainage system A potable water supply for the control room may be required for water fountains, laboratory sinks, restrooms, and changing rooms in process plants. Piping should be designed in accordance with the project piping specifications and applicable national, state, and local building codes. Drains must also tie into the appropriate drainage systems—process or sanitary. Tie-ins to process drains must be gas sealed (Civil and Structural Manual, Section 500).

Air supply Instrument air for the control room is used for pneumatic instrumentation and possibly for the laboratory. Piping should be designed in accordance with the project

Chevron Corporation

1100-7

July 1999



piping specifications, ISA RP 60.9, “Piping Guide for Control Centers,” and ISA S7.3 “Quality Standard for Instrument Air.”

1154 Gas Detection and Fire Detection and Extinguishing Systems Gas detection and fire detection and extinguishing systems should follow the guidelines in the Fire Protection Manual and in Specification ICM-MS-3651, Section 6.0.

1155 Plant Communications Systems In-plant communications system The in-plant communications system enables communications between field operators, local control rooms, and the central control room. Equipment selected for this application can include sound-powered phones, local telephone system extensions, intercom sets, radios, or combinations of two or more types of equipment. In-plant systems can also include the ability to announce plant emergencies such as fire, accidents, and evacuation signals. System design should limit traffic to that which is of interest to a single plant area.

Interplant communications system The interplant communications system enables communications among two or more plants within a facility. Additional equipment can include dedicated phone systems for process—or system-specific communications (e.g., dedicated to utilities users or to communications between H2S producers and recovery plants) and radios for communications with agencies outside the facility.

Coordination The design of these systems should be coordinated with the Telecommunications Department.

1160 Design Concepts 1161 General Considerations The following criteria should be considered for control room designs.

Safety considerations Safety considerations should include area classification, location in relation to classified areas, fire and gas protection, blast protection, and proximity to escape routes.

Area classification The control room is preferably located in an unclassified area. This requirement is extremely critical for both personnel and electronic equipment. If a control room

July 1999

1100-8

Chevron Corporation



must be located in a hazardous location, it should be pressurized and have an airlock or satisfy API RP 500 or RP 505.

Location The control room design should account for prevailing wind. Whenever possible, locate the control room upwind of the process area. This will minimize the potential for process gas accumulation in the control room area. All control room air intakes should be located upwind of the process area and in an unclassified area. It may be desirable to install combustible or toxic gas sensors in or near air intakes.

Fire and gas protection Fire protection systems are required for all control rooms. These systems include heat and smoke detection systems and fire extinguishing systems. The systems should be designed according to the Fire Protection Manual. Combustible and toxic gas detection systems for control rooms must provide two functions: (1) alert personnel to the presence of these gases within the monitored area, and to abnormal operation of sensors and interface electronics; and (2) take corrective action upon detection of gases (e.g., shut in sources). Gas detection systems should be designed in accordance with the Fire Protection Manual. Gas detection systems for offshore platforms must be designed in accordance with API RP 14F and ISA RP 12.13.

Proximity to escape routes The control room design should consider proximity of the control room to escape routes for emergency evacuation. Control rooms should have at least two exits that are as distant from each other as is practical, on the basis of good engineering judgment.

External interference considerations The design of the control room should address and minimize the effects of noise and vibration, electromagnetic interference (EMI), and radio frequency interference (RFI). Ambient noise and vibration should be minimized by proper control room location, installing soundproofing, and providing equipment vibration damping. EMI can be minimized by segregation of power and control wiring, selection of signal levels for transmission, shielding and grounding of cables, and equipment grounding. The recommendations of Section 3.0 of Specification ICM-MS-3651 should be followed. RFI can be minimized by selection of equipment enclosures, physical isolation between sources and RFI-susceptible instruments, and implementation of procedures to minimize RFI generation near sensitive instruments.

Chevron Corporation

1100-9

July 1999



Accessibility and communications A control room design should include a method for communicating with the process areas. It may be desirable to include a page-party system with loudspeakers and handsets throughout these areas and the control room. Unmanned control rooms should provide telephone or page-party communications for occasional visits by personnel. If the unmanned control room is entirely automated, a supervisory control and data acquisition (SCADA) system may be required for data communications to a manned control room.

Ambient nose considerations Control room design and equipment selection should recognize the need to limit the ambient noise level within work areas. Data loggers and report printers are significant noise generators. Printers other than those supplying data needed by the operator at his work station should be installed in a separate room. Printers located at the operator work station can be fitted with noise reduction shrouds to lower ambient noise levels. Air-conditioning systems (especially window-mounted units) can be an additional source of high ambient noise levels. HVAC equipment selection and layout and duct design can minimize the transmission of noise into the control room. Cooling fans installed in electronic equipment racks are another potential noise generator. Where control room layout does not permit installing this equipment in a separate rack room, soundproofing on the inside of rack doors and routing of hot air exhaust ducts away from the operator work area can minimize ambient noise.

Kitchen facilities General control rooms frequently include facilities for preparing and consuming operator meals. Cooking equipment, especially range tops and conventional ovens, generates aerosol greases which can damage electronic circuit boards. Kitchen facilities design should include a vent hood over cooking equipment which discharges outside the control room structure and HVAC return air ducts from the kitchen area should be segregated from those in the remainder of the control room.

Changing rooms for central control rooms in process plants These locker and shower facilities should be designed so that personnel are provided with a “dirty” changing area having direct access to the showers and separated from the clean dressing areas with a walk-through passageway that can be used when arriving personnel change into workclothes.

1162 Building Requirements Architectural features also influence control room design. See Section 400 of the Civil and Structural Manual for guidelines on the civil and structural aspects of small buildings, blast resistant design, and hurricane resistance.

July 1999

1100-10

Chevron Corporation



General requirements The control room may be single- or multistory. Typical arrangements for single- and two-story control rooms are shown in Figure 1100-1. A local control room should be designed around the equipment to be located in the room. This applies to the physical layout of the control room as well as to the architectural design of the room.

Windows and doors Control room windows are usually located to enable viewing as much of the process area as possible. Where windows are provided, several design requirements exist: 1.

If a window is located in a wall required to be fire-resistant, it should be of fixed-frame construction, and glazed with heat-resistant, shatterproof glass.

2.

If the control room is pressurized, window frames and windows should be sealed such that internal pressure can be maintained.

3.

If high ambient noise surrounds the control room, double-glazed windows might be required to reduce noise. (This technique typically attenuates noise by 10 to 15 decibels.)

Refer to Specification ICM-MS-3651 for additional guidelines. In a local unmanned control room, windows should be kept to a minimum. Unless the design demands it, windowless rooms are preferred. In manned local control rooms, windows may be needed to view the process being controlled. Sometimes windows are not needed for viewing (for example, in a manned control room for internally mounted switchgear) but should be provided to meet the goals of human engineering. At least two doors, preferably at opposite ends of the control room, must be provided to allow multiple escape routes for personnel. Doors should open in the direction of the escape path, and be fitted with panic bar hardware. At least one of these doors should be sized to allow for the movement of major equipment into and out of the control room. The maximum size door without mechanical helpers is a double door, 6 feet wide by 7 feet high. If equipment sizing dictates a larger access opening, install a removable transom or a removable wall panel. Doors should be metal with integral frames, and might require soundproofing to attenuate ambient noise. Unless impossible, doors should not be located in a fire-resistant wall.

Lighting The designer should incorporate two lighting systems in the control room design: a primary system for normal operation and an emergency system to be activated on failure of the primary system.

Chevron Corporation

1100-11

July 1999



Fig. 1100-1 Control Rooms (Typical)

‘

July 1999

1100-12

Chevron Corporation



The lighting systems should be designed in accordance with the Electrical Manual, Section 1700, “Lighting.” Figures 1100-2 and 1100-3 show typical layouts of normal and emergency lighting. Emergency lighting is recommended in all manned control houses, to allow orderly shutdown of the facility in the event of a power failure, and permit safe exit from the control room. Fluorescent lighting fixtures are recommended for most applications to achieve the normal lighting levels required. This type of fixture reduces glare and power consumption. Ceiling, wall, and floor reflectance can also be utilized to obtain a glare-free lighting environment.

Ceiling design The issues of fire resistance and the functionality of ceiling spaces must be addressed prior to design of a control room ceiling. The need for fire-resistant construction is determined by what is above the control room ceiling. The desirability of a suspended ceiling is determined by whether the space above it will be used, or might be used at a future date, for the routing of HVAC ducting, utilities, or control system cabling. A suspended ceiling, using T-rail construction and removable acoustic panels, is recommended for control rooms housing primarily electronic instrumentation, because this configuration allows for substantial future changes to the instrumentation and its connecting cabling. A flat suspended ceiling with acoustical tiles permanently attached to the false-work can provide space for routing ducting, utilities, and cabling, but lacks flexibility for installing instrumentation cabling additions. Both configurations of suspended ceiling allow for installation of flush-mounted fluorescent lighting fixtures and the easy reconfiguration of lighting fixture layout as required by realignment of equipment and furniture within the control room. A solid flat ceiling offers none of the functionality of either configuration of suspended ceiling, but has the lowest installed cost. A flocked finish should be considered as a cost-effective method of noise and glare reduction. Ceiling height above the floor should be determined by the requirements of the equipment planned for the control room, and should be a minimum of 9 feet; use of a greater height adds to the esthetics of larger control rooms.

Floor design The floor design of the control room is dependent on the following criteria: the type of equipment to be mounted in the room (electronic or pneumatic), fire protection systems, method and location of wiring, and coordination with remaining building design considerations for lighting, wiring, and ventilation.

Chevron Corporation

1100-13

July 1999



Fig. 1100-2 Reflected Ceiling Plan

July 1999

1100-14

Chevron Corporation



Fig. 1100-3 Control Building: Lighting and Small Power Plan

Chevron Corporation

1100-15

July 1999



A raised floor in conjunction with a suspended ceiling should be considered for electronic equipment. Both floor and ceiling then provide space for cable interconnection of equipment and act as plenums for ventilation. A raised floor 12 inches above the primary floor should provide sufficient clearance in most applications. The designer should ensure that the raised floor is capable of supporting the weight of the installed equipment as well as the expected traffic load without noticeable flexure. Also, the floor height is dependent upon the seismic zone in which the facility is located. If no electronic equipment is located in the control room, the designer may consider using a solid floor to reduce costs. For either construction, both fire protection requirements and the color and texture coordination of the room for eliminating glare and optimizing lighting levels must be satisfied. For additional guidelines, see Specification ICM-MS-3651, Installation Requirements for Digital Instrumentation and Process Computers.

1163 Control Room Environment The environment of the control room should be comfortable for human occupancy and should protect the equipment located there. If the control room is located in a hazardous area, it should be purged or designed in accordance with API RP 500 or RP 505. Through-traffic paths should be arranged so that personnel track in the minimum amount of dirt as they walk through the room. Some Company locations issue disposable plastic overshoes to personnel as they enter the control room.

1164 Pneumatic Tubing and Cable Design Design considerations for cable or pneumatic line entry into a control room are dependent on whether the room is pressurized and whether entry is through walls requiring fire-resistant rating or required to be vapor tight. If the control room is pressurized, the entries should be sealed sufficiently to maintain the room internal pressure. If the control room is not pressurized, and if the walls are not required to be fireresistant or vapor tight, the entries should be weather-tight. In either case, the entries and adjacent areas for cable, conduit, or tubing routing should provide a 50% (minimum) allowance for future expansion. For a pressurized control room cable entry, a multicable transit as shown in Figure 1100-4 or equivalent entryway should be used. Individual pneumatic tubing entries to pressurized control rooms should be routed via bulkhead fittings as shown in Figure 1100-5. The interior layout of cables and pneumatic tubing is a function of the control room layout and type of entry into the cabinets or consoles. If the equipment is bottom entry, a raised floor should be considered. Conversely, if the equipment is top entry, a false ceiling should be considered. When a combination of top and bottom entry is used, the designer should consider either a raised floor and lowered ceiling or one of

July 1999

1100-16

Chevron Corporation



Fig. 1100-4 Multicable Transit (Courtesy of Nelson Fire Stop Products) COMPRESSION BOLT When tightened, the bolt applies pressure to the compression plate sealing the grooved insert modules around the cables.

Multicable transit is based on a simple but effective design. It consists of a rectangular metal frame suitable for floor or wall installation, which is available in single or multiple units. Each frame contains an arrangement of Tecron elastomer modules grooved to fit snugly around cables, pipes, or conduit passing through the frame. The Tecron modules expand when exposed to heat, providing a continuous seal even if cable jackets disintegrate. The entire assembly within each frame is locked in position to prevent dislodgement and the spread of fire and the products of combustion.

END PACKING—STANDARD End packing assembly is bolted into place to provide a fire and watertight seal above the compression plate. The standard end packing assembly is used when both sides of the transit frame are accessible. END PACKING—SPECIAL The special end packing assembly serves the same purpose as the standard and is used when the transit frame is accessible from only one side. COMPRESSION PLATE The compression plate acts as a pressure plate above the internal assembly. STAY PLATES Stay plates are inserted between every completed row to help distribute compression forces within the frame and to keep modules from dislodging under high pressure conditions. GROOVED INSERT MODULES Grooved insert modules are available in seven module sizes to accommodate a range of cable/pipe from 5/32 in. to 3-3/4 in. O.D. They fit snugly around the cable or pipe to form an air-tight, water-tight seal when compression is applied in final assembly step.

TRANSIT FRAME The transit frame is the housing into which the other components are fitted. MCT LUBRICANT (TALLOW) Used when packing. Allows the insert modules to slide easily over each other.

SPARE INSERT MODULES Solid modules are used to fill voids or allow for future addition of cables. They are available in 3 module sizes.

RTV-106 SEALER For armored cable. Sealer should be applied in the grooves to seal the space between the armor and the cable sheath in navy cables, and the groove in the interlock of industrial cables.

Chevron Corporation

FILL-IN INSERT STRIPS Used to fill space gaps. Available in two thicknesses; 5 and 10 mm. Strips are 120 mm long and are split to allow cutting at any desired length.

1100-17

July 1999



Fig. 1100-5 Typical Bulkhead Connections

the two with appropriate routing to the equipment with opposite entry. Typical routing techniques are shown in Figure 1100-6. Instrument, alarm, thermocouple, and control wiring should be separated.

1165 Layout Considerations for Control Panels Two factors affect control room arrangement: space required and use of conventional panels or shared-display (CRT) systems. Typical arrangements for conventional and shared-display systems and associated equipment are shown in Figures 1100-7, 1100-8, and 1100-9.

General considerations for operator work areas The design of the control room layout should ensure that the operator work area is separated from the flow of traffic through the control room and from nonoperational work areas. Appropriate communications to the operator work station should be provided.

Clearances The designer should maintain minimum clearances around control equipment to provide work space for operations and maintenance and to facilitate ventilation. The following minimum clearances should be included in the control room design: •

Operations and maintenance access: 4 feet, or as required by local electrical code, whichever is greater

•

Overhead clearance for panels: 1 foot

Selection of control equipment requiring front and rear access should consider its additional space requirements for access and the size limitations of the control room. Conventional control panel design is frequently a mimic representation of the process in the area above instrumentation; this frequently increases panel size.

July 1999

1100-18

Chevron Corporation



Fig. 1100-6 Cabling and Tubing Rooms, Process Plant Central Control Room (Example)

Control panels should be built in accordance with Specification EG-1348, ShopFabricated Control Panels.

Color considerations A panel background color that compromises between maximum contrast and no contrast should be specified. The compromise should lean slightly toward less contrast.

1166 Layout Considerations for Control Rooms Housing Distributed Control Systems Distributed control systems (DCS) control rooms typically contain operating consoles, one or more video displays and printers, and magnetic data storage devices. The circuit board rack comprising the microprocessor-based controllers may be located in the console under the CRTs, in a separate cabinet in the control room, or in the “rack” room. A typical arrangement found in Company process plants is shown in Figures 1100-8 and 1100-9.

Chevron Corporation

1100-19

July 1999



Fig. 1100-7 Shared Display (CRT) and Conventional Panel Control Systems

July 1999

1100-20

Chevron Corporation



Fig. 1100-8 Offshore Control Building: Electrical Plan

1. NAV AID BATTERY RACK

21. 125 VDC 30A MAIN 3 2

41. GAS DETECTION PANEL

2. 125 VDC BATTERY RACK

22. 125 VDC MAIN BREAKER PANEL

42. GAS DETECTION PANEL


23. 480 VAC POWER DISTRIBUTION PANEL

43. TRANSFORMER 4160/480 VOLT

4. 24 VDC BATTERY (CONTROL) RACK

24. 208 VAC LIGHTING PANEL

44. WATER CHILLER UNIT #1

5. 24 VDC BATTERY (GENERATOR) RACK

25. A/C DISTRIBUTION PANEL

45. WATER CHILLER UNIT #2


26. A/C DISTRIBUTION PANEL

46. WATER FOUNTAIN

7. 24 VDC DISTRIBUTION BOARD

27. 480V SWITCHGEAR

8. 24 VDC U.V. CONTACTOR

28. 480V MCC

47. WALL MTD HAND SET/SPEAKER AMPLIFIER

9. NAV AID BATTERY CHARGER

29. 125 VDC HALON LATCHING RELAY

48. LINE BALANCE ASSEMBLY

10. 24 VDC CONTACTOR FOR SOLAR GENERATOR

30. 4160 SWITCHGEAR

49. MULTI-TONE GENERATOR

11. 24 VDC BATTERY CHARGER 12. 24 VDC BATTERY CHARGER 13. 24 VDC BATTERY CHARGER 14. 24 VDC BATTERY CHARGER 15. 125 VDC EMERGENCY LUBE PUMP STARTER 16. 125 VDC DISTRIBUTION BOARD 17. 24 VDC CONTACTOR (SOLAR LUBE) 18. 125 VDC BATTERY CHARGER 19. 125 VDC BATTERY CHARGER

31. "C" TRAIN COMPRESSOR CONTROL PANEL 50. CONE SPEAKER 51. DESK TOP HANDSETS/SPEAKER 32. CENTRAL CONTROL PANEL AMPLIFIER 33. "D" TRAIN (FUTURE) 52. AMPLIFIER 34. INSTRUMENT AIR PANEL 53. P/C TERMINAL 35. 3 INVERTERS (STACKED) 54. 45 KVA TRANSFORMER 36. GENERATOR PANEL #1 55. 24 VDC 70 AMP MAIN BREAKER 37. GENERATOR PANEL #2 56. 125 VDC 100 AMP MAIN BREAKER 38. PLC LOGIC PANEL 57. 24 VDC 70 AMP MAIN BREAKER 39. FIRE DETECTION PANEL 58. 24 VDC STARTER (SOLAR LUBE) 40. GAS DETECTION PANEL 59. U.V. DETECTION PANEL

20. 125 VDC U.V. CONTACTOR

Chevron Corporation

1100-21

July 1999


July 1999

Fig. 1100-9 Typical Process Plant Central Control Room Building Arrangement with Shared Display Control System


Chevron Corporation



Clearances The designer should maintain the same minimum clearances for operations and maintenance access to non-console equipment as described in Section 1165. Additional clearance for operator access to consoles is required to allow passage behind an operator seated at a work station. Care must be taken to account for the width of integral work surfaces mounted to consoles. Operating personnel must be able to easily access and operate the consoles while seated.

Lighting and contrast considerations Normal lighting considerations (see Section 1162) are usually acceptable. However, a major difficulty is the glare problem on video displays. Glare reduction techniques include use of spotlighting, rearrangement of lighting fixtures, installation of glare shields on video monitors, and use of light diffusers on new or existing light fixtures.

1167 Layout Considerations for Auxiliary Equipment As shown in Figures 1100-8 and 1100-9, auxiliary equipment can include field wiring termination cabinets, programmable logic controllers, specialized equipment controllers (e.g., for compressors), printers, computer interface equipment, and data storage units. Only printers used by the operator for critical messages should be located in close proximity to the operator work station. All other printers should be located away from the operator work station, ideally in a separate room, to minimize noise level. The designer should specify the color of the auxiliary equipment to accomplish a “soft” contrast with the room color. Routing of cables and tubing to or between auxiliary equipment should be planned to avoid interference or difficulty in installing and maintaining the particular type of equipment. Batteries for supplying standby power to control systems, emergency lighting (except for self-contained emergency lighting packs or exit signs), and communications systems should be housed in a separate room provided with adequate ventilation to prevent the accumulation of hydrogen gas generated by battery charging.

1168 Local Control Room Design Concepts Local control rooms may be manned or unmanned, depending on the application. Usually, the local control room will be equipped to provide local monitoring and control of large equipment (drilling rig, motor control center, etc.) or a segment of the overall process. Technically, little or no attention is required for personnel comfort in the design of an unmanned local control room. However, for a number of reasons (e.g., protection of equipment, declassification of the area to permit the use of less costly instrumentation, and comfort of operating and maintenance personnel), some unmanned local control rooms should be designed so they can be manned.

Chevron Corporation

1100-23

July 1999



The designer should consider a means to transmit the monitored information back to a remote control center.

1169 Central Control Room Design Concepts The central control room is normally equipped to monitor and control process functions and peripheral support equipment for the facility. Office space for supervisory personnel should be provided within the control room building. Additional facilities that might be included within the control room building include: •

Office space for technical specialists

•

Lavatory facilities

•

Kitchen and eating facilities

•

Storage closets for consumable stores used in the control room (e.g., charts, ink, run sheets, etc.)

Highly flammable, volatile, or toxic materials should not be permitted in the control room. If laboratory facilities are provided within the control room building, they should be provided with separate outside entry doors and be furnished with an independent HVAC system or with window-mounted air-conditioners, as discussed previously in this guideline.

1170 Checklist and Final Documentation 1171 Checklist for Control Room Design The checklist shown in Figure 1100-10 lists the data required for sizing, constructing, and providing accessories and utilities for a control room. Additional data which amplifies on these requirements may be appended to this checklist.

1172 Final Documentation The following documentation should be created: 1.

Control room location plan

2.

Control room layout drawings (showing floor plans and equipment elevations)

3.

Layout drawings for pneumatic tubing and electrical power and signal cables

4.

Area lighting analysis

5.

Lighting layout drawing

6.

Specifications for the following: –

July 1999

Building walls

1100-24

Chevron Corporation



Fig. 1100-10 Checklist for Control Room Design (1 of 5)

A. CONTROL ROOM TYPE (Check one.) 1. Local Unmanned

_______

2. Local Manned

_______

3. Central

_______

B. CONTROL HARDWARE (Check and specify all that apply.)

Type

Size (L × W × H)

Front

Access: Rear Side(s)

Signal Entry: Top Bottom

1. Free-Standing Control Panel

_______

________

______

______

______

______

______

2. Console Control Panel

_______

________

______

______

______

______

______

3. DCS Console

_______

________

______

______

______

______

______

4. Control Racks

_______

________

______

______

______

______

______

5. Wall-Mounted Control Panel

_______

________

______

______

______

______

______

6. Printer

_______

________

______

______

______

______

______

a. ____________________

________

______

______

______

______

______

b. ____________________

________

______

______

______

______

______

c. ____________________

________

______

______

______

______

______

a. Operator Work Area:

________

sq. ft.

b. Equipment Area:

________

sq. ft.

c. Maintenance Area:

________

sq. ft.

d. Storage Area:

________

sq. ft.

e. Office Area:

________

sq. ft.

f. Lavatory Area:

________

sq. ft.

g. Kitchen Area:

________

sq. ft.

h. Laboratory Area:

________

sq. ft.

i. Passageways:

________

sq. ft.

j. Other:

________

sq. ft. (Specify use): ______________________

2. Ceiling Height:

________

in.

3. Other Sizing Data:

_______________________________________________________

7. Other (Specify):

C. SIZE REQUIREMENTS 1. Floor Area:

Chevron Corporation

1100-25

July 1999




D. ENVIRONMENTAL REQUIREMENTS (Check all that apply; specify required data.) 1. Cooling

_______

Maximum Ambient Temp., °F___

2. Heating

_______

Minimum Ambient Temp., °F___

3. Dehumidification

_______

Maximum Humidity,

4. Air Purification

_______

Airborne Contaminants:______________________________

5. Pressurization

_______

6. Sanitary Facilities

_______

7. Noise Abatement

_______

Ambient Noise Level:

8. Other

_______

(Specify)_________________________________________

____

_____

R.H. at ___°F

dBA

E. FLOORS 1. Type: (Check one.)

_________

a. Solid

_________

b. Raised

_________

c. Other

_________

(Specify)____________________________

2. Covering: (Check all that apply.) a. Skid-Resistant

_________

b. Chem-Resistant

_________

c. Other

_________

(Specify)____________________________

F. WALLS 1. Construction (Check all that apply.) a. Fire-Resistant

_________

b. Insulated

_________

c. Soundproofed

_________

d. Other

_________

(Specify)____________________________

2. Penetrations (Check all that apply; specify where appropriate.) a. Doors: (1) Number

_________

(2) Size

_________

(3) Accessories

_________

b. Windows: (1) Number

_________

(2) Size

_________

(3) Accessories

_________

July 1999

1100-26

Chevron Corporation



Fig. 1100-10 Checklist for Control Room Design (3 of 5) c. Entries: (1) Removable Transoms

_________

Number:

_____

Size:

______________

(2) Cable Entryways

_________

Number:

_____

Size:

______________

Method of Sealing:

_______________________________________________________

(3) Tubing Entryways

_________

Number:

_____

Size:

______________

(4) Other Entries

_________

(Specify)______________________________

d. Other Penetrations

_________

______________________________(Specify)

a. Interior Surfaces (Type)

_________

(Color)

_______________________________

b. Exterior Surfaces (Type)

_________

(Color)

_______________________________

3. Finish

G. CEILING 1. Construction (Check all that apply.) a. Fire Resistant

_______

b. Insulated

_______

c. Suspended

_______

d. Noise resistant

_______

e. Other

_______

2. Finish

Specify type:

T-rail__

Plenum height:

___ in.

Permanent______________

(Specify)________________________________________________

(Type)_________________________

(Color)__________________________

3. Other Features: ___________________________________________________________________________

H. LIGHTING (Note: If lighting requirements vary within control room, attach additional copies of this section to Checklist and reference by area or room.) 1.

Type (Check one.) a. Fluorescent, flush mount

_________

b. Fluorescent, surface mount

_________

c. Incandescent

_________

d. Spotlighting

_________

e. Other

_________

(Specify)__________________________________

2. Lighting Intensity (Specify where appropriate.) a. Operator Work Area:

____________fc

b. Equipment Area:

____________fc

c. Maintenance Area:

____________fc

d. Storage Area:

____________fc

Chevron Corporation

1100-27

July 1999



Fig. 1100-10 Checklist for Control Room Design (4 of 5) e. Office Area:

____________fc

f. Lavatory Area:

____________fc

g. Kitchen Area:

____________fc

h. Laboratory Area:

____________fc

i. Passageways:

____________fc

j. Other:

____________fc

(Specify use)____________________

3. Emergency Lighting Requirements (Check and specify where appropriate.) a. At Exit Doors

____________

Duration:___________________ mins.

b. In Control Room

____________

Duration:___________________ mins.

c. In Equipment Area

____________

Duration:___________________ mins.

d. Other

____________

Duration:___________________ mins.

(Specify Area): _________________________________________________________________________

I. SAFETY 1. Area Classification:

Class _____________

Division ________

Group(s) _______

a. Smoke Detectors

____________

Areas Requiring Protection: _________

b. Flame Detectors

____________


c. Toxic Gas Detectors

____________


d. Combustibles Detectors

____________


e. Halon Systems

____________


f. Other Firefighting

____________


2. Safety Systems (Check all that apply.)

(Specify Type): _________________________________________________________________________ g. Other Safety Systems

____________


(Specify Type): _________________________________________________________________________

J. UTILITIES 1. Electrical Power (Check all that apply; specify data where appropriate.) a. 12 VDC with UPS

______

Load, amps: _________

Battery amp-hours: ________________

b. 24 VDC with UPS

______

Load, amps: _________


c. 48 VDC with UPS

______

Load, amps: _________


d. 120 VDC with UPS

______

Load, amps: _________


e. ___ VDC with UPS

______

Load, amps: _________


f. 120 VAC with UPS

______

Load, amps: _________


g. ___ VAC with UPS

______

Load, amps: _________


h. 12 VDC (no UPS)

______

Load, amps: _________

July 1999

1100-28

Chevron Corporation




2.

3.

4.

i.

24 VDC (no UPS)

______

Load, amps: _________

j.

48 VDC (no UPS)

______

Load, amps: _________

k. 120 VDC (no UPS)

______

Load, amps: _________

l.

___ VDC (no UPS)

______

Load, amps: _________

m. 120 VAC (no UPS)

______

Load, amps: _________

n. 480 VAC (no UPS)

______

Load, amps: _________

o. ___ VAC (no UPS)

______

Load, amps: _________

Instrument Air

______

Pressure, PSIG: ______

Volume, SCFM: ___________________

Filtration, microns:

______

Water a. Potable water

______


Volume, GPM: ____________________

b. Hot water

______


Volume, GPM: ____________________

c. Cold water

______


Volume, GPM: ____________________

a. General Purpose Sewer

___________________

Volume, GPM: ____________________

b. Laboratory Sewer

___________________

Volume, GPM: ____________________

c. Other Sewer

___________________

Volume, GPM: ____________________

Sanitary Sewers

(Specify use): __________________________________________________________________________ 5.

Other Utilities a. Type: _________________________________

Capacity: _________________________________

b. Type: _________________________________

Capacity: _________________________________

– – – – – – – – – 7.

Drawings showing the following: – –

Chevron Corporation

Ceiling Floor HVAC system Gas detection system Purging system Fire detection system Fire extinguishing system Page-party system Other communications equipment

Location of multicable transit and bulkhead penetrations Layout of gas detectors, smoke detectors, fire detectors and extinguishers, and Halon system

1100-29

July 1999



– –

HVAC system, including ducting, vents, exhausts, intakes, and central plant equipment Electrical distribution and plumbing

1180 Sample Control Room Design Criteria The scenario for this design example is as follows: A manned oil and gas production platform is to be designed for installation off the west coast of Cabinda, Angola, and a central control room is required. (System capacity is not stated in this example.) 1.

Temperature: Min 54°F Max 94°F

2.

Humidity: 81 to 89%

3.

Prevailing wind: Direction—from southwest Velocity— 1 Hour average: 10 knots 3 Second average: 15 knots

4.

Classification of area where control room should be located: unclassified

5.

Unusual weather occurrences: none

6.

Weight restrictions: 75 tons, max.

7.

Size restrictions: 57 by 34 ft., max.

8.

Power availability: 120 VAC Single-phase, 60 Hz 480 VAC Three-phase, 60 Hz 120 VDC 24 VDC

9.

Water availability: sufficient

10. Air supply: sufficient The prerequisites for the design of the central control room are also assumed and are as follows:

July 1999

1.

Number of administrative personnel: one per 12 hrs

2.

Number of technical specialists: one

3.

Computer requirements: A computer is required to support the distributed control system. An operator interface CRT, a computer, and a printer interface

1100-30

Chevron Corporation



are required. Cabinets connecting the computer to the field instrumentation are located in the control room. –

– –

Computer Console: Configuration—Six bays with an integral desk; Size—Approximately 132 in. long, 48 in. deep, and 72 in. tall; Power Consumption—Computer and CRT: 1400 watts,—Printer: 300 watts,— Disk Drive: 250 watts. Interface Cabinet Configuration: eight separate I/O cabinets with field terminations; four emergency shutdown (ESD) Cabinets. Size: I/O Cabinets—each 26 in. wide x 48 in. deep x 84 in. high; Emergency Shutdown Cabinets—26 in. wide x 36 in. deep x 72 in. high; Power Consumption: I/O cabinets—800 watts each; ESD Cabinets—250 watts each.

Once the Checklist for Control Room Design is completed, the control room design can begin. The following steps are recommended: 1.

Locate the control room on the basis of wind direction, current direction, proximity to escape routes, and proximity to the process area. For purposes of this example, the control room is located in a nonhazardous area. See Figure 1100-11.

Fig. 1100-11 Control Room Location Plan

2.

Chevron Corporation

Establish the size of the control room by creating a layout showing all equipment with minimum or greater clearances. See Figure 1100-12.

1100-31

July 1999



Fig. 1100-12 Control Room Equipment Layout

3.

Perform the architectural design, including walls, windows, ceiling, flooring, and lighting. Provide a building specification based on the above requirements.

4.

Lay out the HVAC system. See Figure 1100-13.

Fig. 1100-13 Location of HVAC and Fire/Gas Protection Systems in Control Room

July 1999

1100-32

Chevron Corporation



5.

Lay out the fire protection and gas detection systems. See Figure 1100-13.

6.

Provide room for the communications equipment (including ship-to-shore, surface-to-air, and internal communications). Locate the communications equipment required to support the operation of the control room with the platform or process plant, coordinating with the telecommunications group.

7.

Establish the electrical load requirements of the equipment as determined in Steps 2 through 6. See Figure 1100-14 for an example calculation.

Fig. 1100-14 Electrical Load Requirements for Control Room Equipment DESCRIPTION

POWER REQD

VOLTAGE

UPS/PRIMARY

1,400 Watts

120 VAC

Primary

Disk Drive

250 Watts

120 VAC

Primary

Printer

300 Watts

120 VAC

Primary

I/O Cabinets (8)

6,400 Watts

24 VDC

Primary

ESD Cabinets (4)

1,000 Watts

24 VDC

UPS

Lab Equipment

3,000 Watts

120 VAC

Primary

Lighting

10,000 Watts

120 VAC

Primary

Emergency Lighting

2,000 Watts

120 VDC

UPS

Fire & Gas Detection

5,000 Watts

24 VDC

UPS

HVAC System

25,000 Watts

480 VAC/3ph/60 Hz

Primary

Radio System

3,500 Watts

24 VDC

UPS

Computer and CRT

SUMMARY This control room has the following power requirements from the respective power supplies: 24 VDC UPS

9,500 Watts

120 VDC UPS

2,000 Watts

480 VAC Primary

25,000 Watts

120 VAC Primary

14,950 Watts

24 VDC Primary

6,400 Watts

8.

Establish cable and pneumatic line layouts including the location of bulkheads and multicable transits. See Figure 1100-15.

The above example only provides a general description of a control room. When actual overall system requirements are established, a detailed design and specification must be developed.

Chevron Corporation

1100-33

July 1999



Fig. 1100-15 Typical Bulkhead Layout

1190 References 1191 Model Specifications Specification ICM-MS-3651

Installation Requirements for Digital Instrumentation and Process Computers

Specification ICM-MS-1348

Shop-Fabricated Control Panels

1192 Other References National Fire Protection Association (NFPA) Standard No. 496, Purged and Pressurized Enclosures for Electrical Equipment ISA RP 12.1

Electrical Instruments in Hazardous Atmospheres.

ISA RP 60.9

Piping Guide for Control Centers.

ISA S7.3

Quality Standard for Instrument Air.

Chevron Corporation, Electrical Manual, Section 300, “Area Classification.” Chevron Corporation, Electrical Manual, Section 1200, “Lighting.” Chevron Corporation, Civil and Structural Manual, Section 400, “Small Buildings.”

July 1999

1100-34

Chevron Corporation

1200 Relief Systems Abstract This section includes guidelines for sizing, selecting, and installing pressure relief devices and introduces the primary industry references (API RP 520, RP 521, etc.). It also provides a basis for designing pressure relief devices in coordination with the design of new or pre-existing relief headers and disposal systems. Recommendations for both onshore and offshore installations are noted. The sections listed below treat the portions of a relief system separately for convenience of description. Keep in mind that these are components of an interrelated system. Contents

Page

1210 Introduction

1200-3

1211 Purpose and Scope of this Document 1212 The Purpose of Pressure Relief Systems 1213 Types of Equipment Protected by Pressure Relief Systems 1214 Storage/Transportation Equipment 1215 The Components of Typical Pressure Relief Systems 1216 Applicable Design Codes and Regulations 1220 Causes of Overpressure

1200-16

1221 Overview of the Relief System Analysis Process 1222 Examination for Causes of Overpressure 1230 Pressure Relief Device Selection

1200-28

1231 Conventional Pressure Relief Valves 1232 Balanced Pressure Relief Valves 1233 Pilot-Operated Pressure Relief Valves 1234 Rupture Disks 1235 Breaking Pin Devices 1236 Relief Device Accessories


1200-1

November 2001

1200 Relief Systems


1237 Materials Selection 1240 Pressure Relief Device Set Pressures Sizing and Installation

1200-37

1241 Relief Device Set Pressures 1242 Relief Device Capacity Calculation 1243 Pressure Relief Valve Installation 1244 Rupture Disk Installation 1245 Relieving Thermal Expansion of Liquids in Piping 1250 Pressure Relief Disposal Systems

1200-65

1251 Introduction 1252 Choice of Disposal Method 1253 Disposal System Design Basis 1254 Disposal System Modeling 1255 Evaluation of Hydraulic Modeling Results 1256 Disposal System Components 1260 Pressure Relief System Design — New Facility

1200-99

1261 Facility Information 1262 Equipment List 1263 Causes of Overpressure 1264 Required Relief Flow Rates 1265 Required Relief Devices 1266 Relief Device Inlet and Outlet Lines 1267 Relief Discharge Collection and Disposal Systems 1270 Pressure Relief System Design — Existing Facility

1200-104

1271 Pressure Relief System Design Reviews 1272 Relief System Design Review Report 1273 Management of Change 1280 Relief Valve Testing

1200-118

1281 Test Equipment 1282 Inspection and Test Procedure 1283 Records and Reports

November 2001

1290 References

1200-138

12100 Glossary

1200-140

1200-2



1200 Relief Systems

1210 Introduction 1211 Purpose and Scope of this Document This document is intended to provide a general overview of the varied issues relevant to pressure relieving systems. It covers the purpose of these systems, the relevant industry codes and governmental regulations, the types of equipment requiring protection, the potential sources of overpressure that necessitate protection, and the types of protective devices used. It addresses issues involved in the design, installation, inspection, and maintenance of pressure relief systems, for existing facilities, modifications or additions to existing facilities, and for newly constructed facilities. It covers the requirements for documentation of the relief system design basis, and of the mechanical integrity inspection and maintenance records relevant to relief systems. This document does not address the pressure relief requirements of and protective devices for all types of pressure-containing equipment. Rather, its scope is confined primarily to process industry and pipeline transportation systems. Explicitly excluded from detailed discussion are the pressure relief requirements of and protective devices for transportation containers: rail cars, tank trailers, barge tanks, and compressed gas cylinders. This document is not intended to provide the detailed instruction required to perform analysis and design of pressure relief systems. Rather it attempts to give a broad introduction to all of the issues relevant to these systems, and to provide references to standards, codes, recommended practices and other sources for further, more detailed, information. This document’s intended audience includes design, process, maintenance, and safety engineers, plant inspectors, reliability analysts, and operations management personnel throughout the process industries. This includes the upstream (both onand offshore), refinery, chemical, and pipeline sectors. It is assumed that readers of this document will be involved in the oversight of pressure relief systems design and analysis projects, in relief system design reviews, and/or in the management of existing systems at production and process facilities.

1212 The Purpose of Pressure Relief Systems All process systems, from domestic water heaters to petroleum production facilities to high-pressure polymerization reactors to atmospheric storage tanks involve the containment of fluids. These fluids are contained in equipment and piping fabricated from materials that have limited mechanical yield strengths. Such equipment and piping are generally fabricated following the rules of one or more industry standards, rooted in engineering science and experience, that specify maximum allowable pressures at which these components are to be operated. These maximum pressures are specified with the intention of providing a sizable “safety margin” between the maximum pressure and the pressure at which the material yield strength would be reached. Stressing these materials beyond their yield strengths could lead to undesirable consequences, varying from an explosive rupture of a high-pressure


1200-3

November 2001

1200 Relief Systems


vessel to a spill of flammable liquid to a release of toxic gases into the atmosphere. Pressure relief systems are installed to protect equipment from failure caused by overpressure. As defined by American Petroleum Institute Recommended Practice 521, a “pressure relieving system is an arrangement of a pressure-relieving device, piping, and a means of disposal intended for the safe relief, conveyance, and disposal of fluids.” A relieving system, per this definition, “may consist of only one pressure relief device, either with or without discharge pipe, on a single vessel or line.” A more complex pressure relief system “may involve many pressure-relieving devices manifolded into common headers to terminal disposal equipment.”

1213 Types of Equipment Protected by Pressure Relief Systems Pressure relief devices and systems are installed on a variety of equipment present at production and processing facilities.

Process Equipment Under the umbrella of “process” equipment (which includes oilfield production processes), the types of equipment that typically are protected by pressure relief systems include vessels, heat exchangers, pumps, compressors, piping, etc. In this context, “vessel” is used in the general sense of “pressure vessels.” Accordingly, examples of “vessels” include separators, scrubbers, columns (distillation, contacting, adsorption), reactors, accumulators, and filters, as well as other types. Heat exchangers of all types are included here: shell and tube, air-cooled, plate and frame, and fired heaters. Positive-displacement pumps and compressors nearly always are equipped with some type of pressure relief system; centrifugal pumps and compressors sometimes require the installation of relief devices. Piping elements are rated for some design pressure; relief devices are therefore often installed on piping to prevent that pressure from being exceeded.

Utility Equipment Nearly all process industry facilities have on-site utility systems of some sort. Many of these utility systems are protected against causes of overpressure by pressure relief devices and systems. Some of the typical utility systems are listed below. 1.

Steam Systems Pressure relief devices are typically present throughout utility steam systems. These locations include the boiler itself, feed water drums, steam drums, and let-down stations.

2.

Steam Turbines Steam turbines used to drive pumps, compressors, and other rotating equipment are very often fitted with pressure relief devices to protect the turbine casing and downstream components against possible overpressure.

November 2001

1200-4



3.

1200 Relief Systems

Cooling and Heating Systems Heat transfer fluid systems - both heating and cooling -typically require pressure relief devices to protect against potential overpressure caused by leaks of process fluids into the (typically low pressure) heating or cooling system, or by abnormal operation of heat transfer equipment.

4.

Lubrication Systems Lubrication systems are often equipped with relief devices, typically on surge tanks and pumps.

5.

Compressed Air Systems Relief devices are usually installed on these systems, which include both instrument air and plant air systems. The equipment protected typically includes compressors, wet air receivers, dryers, and any equipment downstream of pressure let-down stations.

6.

Fire Water Systems Fire water systems often contain pressure relief systems installed on pump discharges.

7.

Refrigeration Systems In these systems, relief devices typically are required on compressor suction scrubbers, compressor discharges, liquid refrigerant surge vessels, economizers, and process heat transfer equipment (refrigerant evaporators).

1214 Storage/Transportation Equipment Storage and transportation equipment also requires protection from overpressure. 1.

Storage Vessels Pressurized storage is used for storage of gasses or liquids with low boiling points like LPG (liquified petroleum gases). Storage vessels are subject to the same rules as pressure vessels. Examples are propane bullets, butane spheres and storage vessels for chemicals. the pressure relief devices from storage vessels for flammables may be connected to flares or to other means of disposal in case of chemicals.

2.

Storage Tanks Storage tanks are for storage of liquids at atmospheric pressure. Some atmospheric hydrocarbon storage tanks may require devices to limit vapor losses. These are “breather vents” to relieve pressure by venting vapors and also relieve vacuum by letting in air. Some tanks are protected by gas blanketing systems. The pressure settings for breather vents and gas blanketing systems are in ounces of pressure per square inch. The breather vents are generally weight loaded “pallets.”


1200-5

November 2001

1200 Relief Systems


Storage tank venting is not compatible with flare systems. The low operating pressures preclude disposal of vapors by venting to a refinery flare system. Vapor discharges from atmospheric storage tanks if environmentally unacceptable would require separate dedicated vapor disposal systems. Protection of storage tanks is governed by API Standard 650 and API Standard 2000. A more complete discussion of this subject can be found in the Tank Manual. 3.

Shipping Containers (Gas Cylinders to Rail Cars to Barges) Shipping container designs are subject to Federal Regulations. It is important to observe safe handling of containers and not to disable the pressure protection devices associated with them.

4.

Transmission Pipelines Transmission pipelines are designed to Pipe line Code B31.4 for liquids or Code B31.8 for gases. Operation of the pipelines is subject to Federal Regulations. Protection from over pressure is an integral part of design and operating procedures.

1215 The Components of Typical Pressure Relief Systems A pressure relief system typically consists of pressure relief devices (pressure relief valves, rupture discs, and others), piping, and disposal systems. The following provides a brief description of the each of these subsystems; more detailed descriptions will be given in subsequent sections.

Inlet Piping “Inlet” piping serves to provide a physical connection from the equipment requiring pressure relief to the pressure relief device. This piping may be as short as a flanged or threaded connection, or may include piping segments, fittings, and valves. The significant design criterion is that this piping should not excessively restrict the flow from the protected equipment to the relief device.

Pressure Relieving Devices The role of the relieving devices is to allow, under the correct conditions, fluid flow out of or into the process equipment to relieve the build-up of excessive internal or external pressure. The following sections introduce the most common types of relief devices. A more detailed discussion of these devices, including their intended applications, and the advantages and disadvantages of each, will be given in Section 1230. Pressure relief devices are most broadly categorized according to whether or not they are designed to reclose after opening. 1.

Nonreclosing Relief Devices These relief devices are designed to undergo an irreversible mechanical failure at a specified pressure. Their failure allows fluid to flow out of the equipment

November 2001

1200-6



1200 Relief Systems

needing pressure protection, relieving the build-up of excessive pressure. These devices include rupture disks and breaking pin devices. Rupture Disk Devices A rupture disk is essentially a membrane that is engineered to contain pressures up to some critical value. When exposed to pressures above this value, the disk fails, allowing fluid to flow through the area formerly blocked by the disk. Breaking Pin Devices A breaking pin device contains a load-bearing pin that supports a pressurecontaining member. The pin is designed to break when the force exerted on it by the pressure-containing member exceeds some critical value. When the pin breaks, the pressure-containing member moves, allowing fluid to flow through the area formerly blocked by the member. 2.

Reclosing Relief Devices This category includes pressure relief valves of several different designs and descriptions. These include spring-loaded, weight-loaded, and pilot-operated pressure relief valves designed to reclose after opening. Spring-Loaded Pressure Relief Valves In this design, the valve is held closed against the system pressure by a spring having a specified force. When the system pressure exceeds the spring force, the valve opens, and flow out of the process occurs. Pilot-Operated Pressure Relief Valves This design employs two separate valves. The “main” valve, through which relief flow occurs, contains a piston with a larger surface area at its top than at its bottom. Under normal operating conditions, the system pressure is in contact with both sides of this piston. The difference in top and bottom surface areas produces a net force downward, holding the piston closed against the main valve inlet nozzle. The upper side of the piston is also in contact with the inlet of a small, spring-loaded “pilot” valve. When the system pressure exceeds the force of the pilot valve spring, the pilot valve opens, venting the system pressure from the upper side of the main valve piston. The system pressure on the inlet side of the piston then forces the piston to lift, allowing flow through the main valve. When the system pressure drops, the pilot valve recloses, returning system pressure to the top of the piston, and closing the main valve.

Effluent Handling Systems 1.

Atmospheric Discharge In many installations, the relief device effluent is routed directly to the atmosphere. The appropriateness of this design depends on the nature of the effluent, the proximity of personnel and equipment to the discharge location, the operating company’s policies, the facility’s environmental permits, and construction costs. When direct atmospheric discharge is chosen, the “effluent handling


1200-7

November 2001

1200 Relief Systems


system” consists of as little as the relief device’s outlet fitting, or as much as several pipe segments, fittings and a block valve. 2.

“Closed” Systems In many other situations, the discharge side of a relief device is routed to a “closed” effluent handling system. In this context, the word “closed” designates any type of discharge arrangement other than the one in which the relief device discharges directly to the atmosphere. Such systems take a variety of forms, but they nearly always contain one or more of the following components. Collection Piping Closed discharge systems always involve some degree of discharge piping. In most cases, the discharge of many relief devices is tied together into a common relief header. This “header,” whether coming from one or many relief devices, then leads to the next general component of the effluent handling system. In some cases, this is simply a different point in the process - the effluent is merely recycled. In other cases, the discharge is routed to some type of phase separation equipment. Separation Vessels Nearly all closed relief systems involve the installation of a knockout drum - a liquid-vapor separation vessel - to minimize the presence of liquids in the fluid reaching the disposal system, which is usually not equipped to handle liquids. Occasionally, the relief effluent contains one or more substances that are particularly hazardous or valuable. In these cases, extra separation equipment may be installed to capture these components in the relief stream. This equipment may include sparge vessels, in which the relief stream is bubbled through a liquid, which serves to capture the relevant component. Catch tanks of other design are also used as appropriate. The less hazardous components in the relief stream are then allowed to pass to some type of disposal system. Disposal Systems Disposal systems usually involve some sort of combustion apparatus, whether it be an elevated flare tip, a ground flare, or a pit flare. These systems are typically equipped with a variety of sub-components designed to ensure clean and reliable combustion, and to prevent entry of air into the effluent handling system.

1216 Applicable Design Codes and Regulations In part because pressure relief systems are installed on a variety of process, storage, and shipping systems, a comparable variety of industry codes, standards, recommended practices, and government regulations apply to pressure relief systems. The following tables provide a listing of most of these documents that bear on the purpose, design, fabrication, installation, and maintenance of pressure relief

November 2001

1200-8



1200 Relief Systems

systems. Some have a very broad application; others apply to a very narrow range of equipment or situations. None have universal utility. These tables are organized by the entity issuing the code, standard, or regulation. Documents issued by private industrial organizations are listed first, followed by regulations issued by the United States government. These regulations are very often rooted in and cite the industry standards. Many foreign governments issue regulations that bear on relief systems, in ways similar to those issued by the USA. Following the tables of US regulations is a table that lists, for each of a variety of process, storage, and transportation equipment types, the industry and government codes that apply to the pressure relief requirements of that type of equipment. This table is intended to serve as a resource for anyone interested in learning more about the requirements of various codes, standards, and regulations pertaining to the pressure relief needs of a particular equipment type.

Industry Design Codes Tables 1-1 through 1-11 provide a listing of design codes and regulations that have some impact on the design, specification, installation, and maintenance of pressure relief systems. The documents in the first eight of these tables are issued by either professional or industry associations. Many of these documents have been adopted as American National Standards by the American National Standards Institute (ANSI). Tables 1-10 and 1-11 list sections of the Code of Federal Regulations (CFR), issued by U.S. federal government agencies, that have relevance to pressure relief systems. Finally, for each of a variety of equipment types (process vessels, storage tanks, piping, etc.), Table 1-12 provides a list of industry standards and practices and government regulations that pertain to pressure relief systems for the protection of each equipment type. Fig. 1200-1

Codes Issued by The American Society of Mechanical Engineers (ASME) (1 of 2)

Code Designation

Title

BPVC

Boiler and Pressure Vessel Code

Description

Section I

Rules for Construction of Power Boilers

Section IV

Rules for Construction of Heating Boilers

Section VI

Recommended Rules for Care and Operation of Heating Boilers

Section VII

Recommended Rules for Care and Operation of Power Boilers


1200-9

Scope includes equipment in evaporation service. Fluid being evaporated is typically water, but scope includes other fluids as well. This includes hydrocarbons.

November 2001

1200 Relief Systems

Fig. 1200-1


Codes Issued by The American Society of Mechanical Engineers (ASME) (2 of 2)

Code Designation Section VIII

Title

Description

Rules for Construction of Pressure Vessels

Paragraphs UG-125 through UG-136 contain specifications for the design, construction marking, certification, testing, setting, and installation of pressure relief devices installed on equipment within the scope of this Section of the BPVC.

Division 1 Division 2 — Alternative Rules

B31

ASME Code for Pressure Piping

B31.1

Power Piping

B31.3

Process Piping

B31.4

Liquid Transmission Systems Piping

B31.8

Gas Transmission and Distribution Piping Systems

Fig. 1200-2

Codes Issued by The American Petroleum Institute (ASME)

Code Designation

Title

Description

RP520 Part I

Sizing, Selection, and Installation of Pressure-Relieving Devices in Refineries Part I - Sizing and Selection

RP 520 Part II

Sizing, Selection, and Installation of Pressure-Relieving Devices in Refineries Part II - Installation

RP 521

Guide for Pressure-Relieving and Depressuring Systems

Std 526

Flanged Steel Pressure Relief Valves

Std 527

Seat Tightness of Pressure Relief Valves

RP 576

Inspection of Pressure-Relieving Devices

API Std 2000

Venting Atmospheric and Low Pressure Storage Tanks. Refrigerated and Nonrefrigerated

API Std 2510

Design and Construction of LPG Installations

API Std 2510A

Fire-Protection Considerations for the Design and Construction of Liquefied Petroleum Gas (LPG) Storage Facilities

November 2001

1200-10



Fig. 1200-3

1200 Relief Systems

Codes Issued by The National Fire Protection Agency (NFPA)

Code Designation

Title

Description

NFPA 30

Flammable and Combustible Liquids Code

NFPA 68

Venting of Deflagrations

Fig. 1200-4

Codes Issued by The American Water Works Association (AWWA)

Code Designation AWWA D100

Fig. 1200-5

Description

AWWA Standard for Welded Steel Tanks for Water Storage

Specifies venting requirements for tanks built to this standard.

Codes Issued by The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE)

Code Designation ASHRAE 15

Fig. 1200-6

Title

Title

Description

Safety Code for Mechanical Refrigeration

Section 9 details the requirements for pressure relieving systems installed in accordance with this standard.

Codes Issued by The Compressed Gas Association (CGA)

Code Designation

Title

S-1.1

Pressure Relief Device Standards

Description

Part 1 - Cylinders For Compressed Gases S-1.2

Pressure Relief Device Standards Part 2 - Cargo And Portable Tanks For Compressed Gases

S-1.3

Pressure Relief Device Standards Part 3 - Stationary Storage Containers For Compressed Gases

S-7

Method For Selecting Pressure Relief Devices For Compressed Gas Mixtures In Cylinders


1200-11

November 2001

1200 Relief Systems

Fig. 1200-7


Codes Issued by The National Board of Boiler and Pressure Vessel Inspectors (NBBI)

Code Designation

Title

Description

NB-18

Pressure Relief Device Certifications

Lists the flow-capacity data collected for pressure relief devices in certification tests specified by ASME BPVC Sections I, III, IV, and VIII

NB-23

National Board Inspection Code

Details the procedures for inspection and repair of pressure relief valves that must be followed before the “VR” valve repair stamp may be applied to a repaired PRV.

Fig. 1200-8

Codes Issued by The National Electrical Manufacturer’s Association (NEMA)

Code Designation

Title

NEMA Standard SM-23

Steam Turbines for Mechanical Drive Service

NEMA Standard SM-24

Land Based Steam Turbine Generator Sets 0 to 33,000 kW

Description

Includes specification of the pressure relief requirements of turbines driven by steam, air, and other gases.

Governmental Regulations Fig. 1200-9

Regulations Issued by The United States Environmental Protection Agency (EPA)

Code Designation

Title

Description

40 CFR 60

Standards Of Performance For New Stationary Sources

Section 60.18 specifies maximum flare tip exit velocity for certain facilities.

40 CFR 63

National Emission Standards For Hazardous Air Pollutants For Source Categories (“NESHAP”)

Section 63.11 specifies maximum flare tip exit velocity for certain facilities.

40 CFR 68

Risk Management Program (“RMP”)

Requires regulated facilities to compile and maintain the relief system design and design basis as part of facility’s the Process Safety Information.

Note that the requirements of RMP concerning relief systems are identical to those of the OSHA Process Safety Management Standard (“PSM”) described below in item number 1. 1.

About the OSHA Process Safety Management (“PSM”) Standard Both the OSHA Process Safety Management standard (29 CFR 1910.119) and the EPA Risk Management Program rule (40 CFR 68) explicitly mention pressure relief systems in two contexts. The first requires that the relief system

November 2001

1200-12



1200 Relief Systems

Fig. 1200-10 Regulations Issued by The US Occupational Safety and Health Administration (OSHA) Code Designation

Title

Description

OSHA 1910.101

Compressed Gases

Includes specifications for pressure relief requirements of equipment containing compressed gases.

OSHA 1910.103

Hydrogen

Includes specifications for pressure relief requirements of equipment handling hydrogen.

OSHA 1910.104

Oxygen

Includes specifications for pressure relief requirements of equipment handling oxygen.

OSHA 1910.106

Flammable and Combustible Liquids

Includes specifications for pressure relief requirements of equipment containing flammable and combustible liquids.

OSHA 1910.110

Storage and Handling of Liquefied Petroleum Gas

Includes specifications for pressure relief requirements of equipment for the storage and handling of LPGs.

OSHA 1910.111

Storage and Handling of Anhydrous Ammonia

Includes specifications for pressure relief requirements of equipment for the storage and handling of anhydrous ammonia.

29 CFR 1910.119

Process Safety Management (“PSM”)

Requires regulated facilities to compile and maintain the relief system design and design basis as part of the Process Safety Information of the facility.

design and design basis be included in the process safety information pertaining to the equipment in the process. The second requires that relief and vent systems and devices be included in a facility’s mechanical integrity assurance program. Process Safety Information (29 CFR 1910.119 (d) (3) (D) and 40 CFR 68.65 (d) (iv) The intention of the process safety information requirement is to ensure that data is on hand to facilitate the other requirements of PSM and RMP: the process hazards analysis, the operating procedures, mechanical integrity, and the management of change. For pressure relief systems, this information includes the following: a.

Documentation of the pressure relief requirements of each pressurecontaining equipment item in the facility. This includes all ASME pressure vessels, storage tanks, pumps, and compressors.

b.

This documentation includes a listing of each contingency that could lead to overpressure in the equipment item, and for each of these contingencies, a calculation of the relief flow rate required to alleviate the overpressure.


1200-13

November 2001

1200 Relief Systems


Also for each contingency, the documentation must identify the relief device(s) present to provide the required flow rate. c.

Documentation of each pressure relief device in the facility. This includes data such as the manufacturer, model number, size, and materials of construction of the device. It also includes documentation of the adequacy of the device’s flow capacity for each of the causes of overpressure listed in requirement b.

d.

Documentation of any relief and vent collection systems present. This includes flare headers and associated quenching, separation, and disposal (e.g., flares) equipment. The adequacy of the flow capacity of the headers should be documented for all common contingencies, along with the acceptability of the back pressures built-up at each relief device, of the separation capacity of knockout drums, and of the thermal radiation generated by the flare.

e.

Documentation of the basis used in the design of the relief system: PIDs, PFDs, mass and energy balance data, and collection header isometric drawing.

f.

Documentation of the industry codes, standards, and recommended practices employed in the design of the relief systems.

With this information in hand, design review and PHA teams need not generate concerns asking, for example, if a certain causes of overpressure was considered when a relief device was specified, and plant engineers can more confidently recognize the impact of a change in operating conditions or the addition of a new pump on the pressure relief requirements of the process. Mechanical Integrity 29 CFR 1910.119 (j) (1) (iii) and 40 CFR 68.73 (a) (3) The purpose of a mechanical integrity program is to ensure the containment of hazardous substances used in a process. To this end, both the PSM and the RMP regulations require that pressure relief systems be covered by a program that includes:

November 2001

a.

Written procedures for inspection, testing and maintenance

b.

Training of personnel responsible for these activities

c.

Periodic inspection and testing of the devices and associated piping and equipment

d.

The inspection frequencies should be set based on manufacturer’s recommendations and operating experience.

e.

The results of these inspections and tests must be fully documented.

f.

A quality assurance program for the material selection, construction, installation, and repair of relief devices.

1200-14



1200 Relief Systems

Deficiencies revealed by testing and inspection must be remedied before continued use of the item or in a safe and timely manner provided steps are taken to assure continued safe operation. As most relief device deficiencies are usually revealed during testing and inspection in a repair shop, these deficiencies are typically addressed immediately upon discovery. Fig. 1200-11 Codes Applicable to Various Types of Equipment Items Type of Equipment

Applicable Codes

Process Pressure Vessels,

ASME BPVC Section VIII; API RPs 520, 521

Heat Exchangers, etc.

Storage (General)

API RP 521, NFPA 30, OSHA 1910.106

Storage (Atmospheric)

API Std 2000, RP 521 Publications 2028, 2210; NFPA 30; OSHA 1910.106; UL Std. 142

Storage (Pressurized)

ASME BPVC Section VIII; API RPs 520, 521; NFPA 30; OSHA 1910.106

Storage (Water Tanks)

AWWA D100

Steam Power Boilers

ASME BPVC Sections I, VII

Heating Boilers

ASME BPVC Section IV, VI

Mechanical Refrigeration Systems

ASHRAE Std. 15

Steam Powered Turbines

API Std. 611, 612

Process Facility Piping

ASME B31.3; API RPs 520, 521

Boiler Piping

ASME B31.1

Liquid Transportation Piping

ASME B31.4;

Gas Transportation Piping

ASME B31.8;

Compressed Gas Cylinders

CGA S-1.1, S-1.2, S-1.3, S-7


1200-15

November 2001

1200 Relief Systems


1220 Causes of Overpressure 1221 Overview of the Relief System Analysis Process The general steps in a pressure relief system analysis are the same, whether the analysis is of a new facility for the purpose of designing a pressure relief system, or of an existing facility for the purpose of validating and documenting an existing pressure relief system. This process can be broken down into the following five steps: 1.

Identification of all potential causes of overpressure for every piece of equipment within the scope of the analysis. This requires that each piece of equipment be checked for potential causes of overpressure. The applicable causes are typically classified at this point in the analysis according to whether they apply only to the equipment item under consideration, or potentially to other items as well.

2.

Calculation of the required relief flow rates associated with each of these potential causes of overpressure.

3.

Evaluation of existing (or design of new) relief devices to ensure that a relief device exists to provide each required flow rate determined in step 2.

4.

Evaluation of existing (or design of new) inlet and outlet piping of each device to ensure that it can provide the actual (“rated”) flow capacity of the relief device without interfering with the device’s performance.

5.

Evaluation (or design) of discharge collection and disposal systems to ensure they are capable of handling all of the potential required relief flow rates that are discharging simultaneously (“simultaneous release”).

The remainder of this section is dedicated to a discussion of the first two steps in this process, with emphasis on the causes of overpressure themselves. While some indication may be given concerning the approaches used to calculate the required relief flow rates, the details involved in step 2 are generally beyond the scope of this document.

1222 Examination for Causes of Overpressure API Recommended Practice 520 provides a list of potential causes of overpressure (also referred to as “operating contingencies”) that should always be considered when analyzing an equipment item for its pressure protection requirements. While this list covers the contingencies that should always be considered, careful examination should also be made of the specific process being analyzed to ensure that all potential causes of overpressure are covered. Of the sixteen contingencies listed in RP 520, several apply very often to equipment types that are installed in nearly all process facilities. Other contingencies on the list may apply only to certain equipment types, or perhaps to several types of equipment, but only in relatively rare circumstances. We will first discuss the contingencies that most often apply to typical equipment types, and then direct our

November 2001

1200-16



1200 Relief Systems

attention to contingencies that are less frequently relevant. The following list of contingency names is taken from API Recommended Practice 520. The evaluation of relative frequency of occurrence is not from Recommended Practice 520, but reflects industry experience. Cause of Overpressure

Relative Frequency of Occurrence

1. Closed Outlets on Vessel

Applies often, to several equipment types

2. Cooling Failure to Condenser

Applies to a limited number of equipment types

3. Top-tower Reflux Failure


4. Sidestream Reflux Failure


5. Lean Oil Failure to Absorber


6. Accumulation of Noncondensables

Not usual applicable

7. Entrance of Highly Volatile Material

Not usual applicable

8. Overfilling of Vessel


9. Failure of Automatic Controls


10. Abnormal Heat or Vapor Input

Not usually applicable

11. Split Exchanger Tube


12. Internal Explosions


13. Chemical Reaction


14. Hydraulic Expansion


15. Exterior Fire


16. Power Failure


17. Other

Prior to discussing causes of overpressure, it is important to have an understanding of the terms accumulation and overpressure. It has long been common industry practice to set a pressure relief device at the design pressure of the equipment item(s) it protects. However, it is recognized that most pressure relief devices do not open fully as soon as they reach their set pressure. For example, a conventional spring loaded pressure relief device set at 100 psig may not open fully until the pressure at its inlet has reached 110 psig, or 10% above the set pressure. The amount of pressure above the relief device set pressure is referred to as the accumulation. Equipment design codes take this pressure relief device behavior into account by making it permissible for the internal pressure to rise above the design pressure of the equipment item during a relieving episode. For example, Section VIII of the


1200-17

November 2001

1200 Relief Systems


ASME Boiler and Pressure Vessel Code states (at paragraph UG-125) that all pressure vessels “shall be protected by a pressure relief device that shall prevent the pressure from rising more than 10% or 3 psig, whichever is greater, above the maximum allowable working pressure except as permitted” in two subsequent sections. The amount of pressure rise permitted by the relevant design code is the “allowable overpressure”. Allowable overpressures vary by code, equipment type, and cause of overpressure. The most common allowable overpressure is 10% of maximum allowable working pressure (MAWP). The phrase “design pressure” will be used below to refer to the maximum set pressure permissible by the relevant design code for a single relief device on the equipment item in question. For the ASME Boiler and Pressure Vessel Code, this is the MAWP; for the ASME Piping Code, this is the design pressure; for a pump or compressor, this is normally the pressure rating of the casing.

Frequently Applicable Causes of Overpressure The following six causes of overpressure frequently apply to a variety of common equipment types. In documenting the causes of overpressure relevant to most equipment items, a rationale must be provided whenever these contingencies do not apply to an item. 1.

Closed Outlets on Vessel Here “vessel” is intended to mean any equipment item. Closed Outlets is a potential causes of overpressure when the following two criteria are met: a.

A mechanism, such as a block valve, exists to block the item’s outlet. If a system of vessels in open communication can be blocked by the action of a single outlet block valve (i.e., on the outlet of the last vessel), the potential for a blocked outlet should be considered for each item in the system.

b.

A pressure source (pump, compressor, high-pressure reservoir, heat etc.) in excess of the design pressure is present upstream. The maximum pressure of sources can generally be determined as follows: * The physical limitations of equipment such as the dead head pressure for centrifugal pumps and compressors. That is, the upstream pump or compressor must be able to exceed the design pressure in order to be a potential source of overpressure. * The maximum temperature achievable due to heat input. For example, if the heat source can not exceed 300 °F, then the internal pressure cannot exceed the vapor pressure of the fluid at 300 °F. As with the pumps and compressors, the heat source must be capable of elevating the vessel pressure above the design pressure to qualify as a potential source of overpressure.

November 2001

1200-18



1200 Relief Systems

The closed outlet causes of overpressure often applies to the following types of equipment items: Vessels

Compressors

Heat Exchangers

Pumps

Absorbers

Distillation Systems

The relief flow rate required to prevent overpressurization of the item is equal to the flow rate into the item, with the item at its design pressure plus any allowable accumulated pressure. The throughput of most compressors and of centrifugal pumps decreases with increasing discharge pressure, so the required relief flow rate often can be less than the normal operating flow rate. Similarly, if any outlets on the equipment remain open, credit may be taken for the flow out through them. 2.

Power Failure Processing facilities use a variety of power sources: electric motors drive pumps, compressors, aerial cooler fans, reactor stirrers, and remotely actuated valves. Steam turbines drive many of the same equipment items. A loss of these power sources can sometimes cause an overpressure. For example, loss of electricity to a motor-driven cooling water pump can cause loss of cooling to a condenser; loss of power to a reflux pump can cause reflux failure. The Power Failure contingency is different from the contingencies (such as Loss of Reflux) it may initiate in that a single power failure may affect many power users simultaneously. Depending on the power distribution system, a single failure might cause loss of cooling to a partial condensing system and the shut-down of a compressor that takes suction from the uncondensed column overhead stream. This contingency may thereby generate a larger required relief load than either the loss of cooling or the blocked overhead outlet would alone. The required relief flow rate in the power failure contingency is dependent on the particular effects of the potential failure. Therefore, no generally applicable method can be described.

3.

Overfilling of Vessel Overfilling of a “vessel” can generally occur in one of two ways: a.

A liquid pressure source capable of exceeding the design pressure of the vessel is present, and a mechanism exists to block the liquid outlet.

b.

A liquid pressure source capable of exceeding the design pressure of the item is present, and the liquid inlet flow rate can exceed the maximum outlet flow rate. In this case, a mechanism for blocking the liquid outlet is not required to make overfilling a credible causes of overpressure.

Common examples of the overfilling contingency are accidental overfilling of a storage vessel (either a pressure vessel or a low-pressure tank) due to operator error or level-control failure, and overfilling of certain process vessels (separa-


1200-19

November 2001

1200 Relief Systems


tors, surge tanks, and reflux accumulators) due to level control failure, outlet pump failure, or an increase in inlet flow rate. Consideration should be given to the probability of the vessel being overfilled. For example, if the liquid inlet flow rate is small relative to the vessel size and a high level alarm or shutdown is present, the potential for the vessel to be overfilled is significantly reduced. API 521 recommends an assumption of a typical operator response time of 20 minutes. Therefore if a mechanism exists to alert the operator (LAH, etc.), and the inlet flow rate is small enough that vessel will not overfill within 20 minutes of the alert, then overfilling may not be considered a credible causes of overpressure. However, shutdown systems alone that are designed to prevent overfilling are generally not considered adequate to prevent the contingency. In any case, the rationale for excluding overfilling as a potential contingency should be documented. 4.

Failure of Automatic Controls Most control valves are designed either to go to a fully open or fully closed state, or to retain their position upon the loss of their actuating signal or driving medium (often instrument air). Regardless of this “failure position” design, a failure of the primary control element can always cause the controlled valve to go fully open or fully closed. Such a control failure that causes a valve at an equipment outlet to close should be evaluated as either a potential Closed Outlet or Overfilling contingency, as described above. A control failure that causes an inlet valve to open can create a cause of overpressure for downstream equipment if the pressure upstream of the control valve can exceed the equipment’s design pressure. The two criteria that must be met for inlet control valve failure to qualify as a potential causes of overpressure are the following. In evaluating an equipment item for the applicability of control failure, it is necessary that all inlets be identified and checked against these criteria. a.

A control valve or regulator must be present on an inlet line to the equipment; and

b.

The normal operating pressure upstream of the control valve or regulator must be greater than the design pressure of the equipment being evaluated. In the event that the normal upstream operating pressure is unknown, 90% of the upstream relief device set pressure may be used.

A special case arises when the downstream side of the control valve connects to a header system with parallel outlets. In these cases, failure open of the control valve may not result in overpressure of equipment since the upstream pressure can be relieved through several outlets. Another case of special concern is that involving failure of a level control valve (LCV) that dumps high pressure liquids into a low pressure system. While the increase flow of liquid will place an increased burden of vapors flashing from the liquids as they are let down into the low pressure equipment, the more significant causes of overpressure occurs after all liquid has drained from the upstream vessel. The high pressure

November 2001

1200-20



1200 Relief Systems

gases that then flow through the failed LCV can rapidly create serious overpressure in the downstream equipment. In general, the required relief flow rate for an inlet control valve contingency is evaluated as the increase in the flow into the equipment relative to the normal operating flow rate. The full-open flow rate should be evaluated with the downstream side of the control valve at relief conditions. The relief device should be sized using full size trim in the control valve, even if the actual control valve has reduced trim. It should be noted that these conditions might involve a change in the fluid phase from normal operating conditions. Credit may be taken for the increase in flow through the equipment’s outlets due to the increase in the equipment’s pressure. API Recommended Practice 521 recommends that in taking credit for this increased flow, any valves on the equipment’s outlets are assumed to be in their normal operating positions. In all cases, consideration should be given to the possibility that the control failure under consideration may affect more than one control valve. In some cases, the failure could cause multiple inlets to a vessel to go open. On the other hand, an inlet may go open while an outlet goes closed. In such a case, of course, it would be inappropriate to take credit for the normal flow through the closed outlet. 5.

Split Exchanger Tube Internal failure of a heat exchanger is a potential causes of overpressure when the low pressure side of a shell and tube exchanger could be overpressured in the event of a tube failure that would allow the fluid from the high pressure side to enter the low pressure side. The secondary effects of tube rupture upon equipment connected to the low pressure side should also be evaluated, since the high pressure fluid could ultimately arrive at other equipment with relatively low design pressures. The following criteria are used to determine the applicability of the tube rupture causes of overpressure. The tube rupture contingency applies whenever either of the two following criteria hold: a.

The MAWP of the high pressure side of the exchanger is greater than 1.5 times the MAWP of the low pressure side.

b.

The high side design pressure exceeds 1,000 psi.

If the MAWP of the high pressure side is less than 1.5 times that of the low pressure side, then the low pressure side generally will have been hydrostatically tested at a pressure at least that of the high pressure side fluid. It is therefore assumed that the high pressure side fluid will not be able to produce catastrophic damage when it enters the low pressure side. If the high pressure side is pressure-relieved below its MAWP, this relief device set pressure is often used in place of the high-side MAWP when determining the high to low side pressure ratio in criterion 1 above. API Recommended Practice 521 notes that for double-pipe heat exchangers in which both the shell and the “tube” are constructed of standard schedule pipe, internal failure generally need not be considered a credible causes of overpres-


1200-21

November 2001

1200 Relief Systems


sure. This is because the inner pipe is no more likely to rupture than any other pipe in the system. For “double-pipe” heat exchangers in which the inner conduit is actually fabricated of gauge tubing, API recommends the experienced application of engineering judgment in evaluating the credibility of internal failure as a cause of overpressure. In order to evaluate the required relief flow rate, the assumption is normally made that the exchanger’s internal failure is a complete break across a single tube. The flow rate is then calculated for the flow from each side of the ruptured tube. This will involve the calculation of a liquid, gas, or flashing liquid discharging from the “high side” operating pressure to the “low side” relief pressure. Furthermore, consideration should be given to the possibility of vaporization of either fluid brought about by the improved heat transfer when the fluids come into direct contact. Even if liquid on the low pressure side does not evaporate, it is likely to be carried along with high-pressure vapor flowing through the rupture, creating a two-phase relief contingency. In this case, the flow of gas or vapor through the rupture would be single phase, but that the flow through the relief device would two-phase. In certain instances where a very high pressure fluid enters the low pressure side of an exchanger, the rate at which pressure builds up in the exchanger may be faster than a pressure relief valve can open. In these cases, the installation of a rupture disk, which responds more rapidly to a pressure stimulus than does a typical pressure relief valve, is often preferred. 6.

Hydraulic Expansion The Hydraulic (or Thermal) Expansion causes of overpressure is applicable to an equipment item whenever all of the following three conditions are met: a.

The item can be blocked-in

b.

The item is full of liquid.

c.

A mechanism exists to heat the blocked-in liquid. The source of heat may be any of the following: · Process heat (e.g., from a fired heater or heat exchanger) · Heat tracing by steam, hot water, electricity, or other means. · Solar Radiation · Ambient conditions (when the “cold” liquid is at or below ambient temperature).

When all three of these conditions prevail, the temperature increase leads to thermal expansion of the liquid. Because the equipment is liquid-full, this expansion produces a dramatic increase in the equipment’s internal pressure, unless a relief path is available to the expanding liquid.

November 2001

1200-22



1200 Relief Systems

Equipment types to which this contingency typically applies are: a.

Pressure vessels that operate liquid-full and at or below ambient temperature (e. g., filters and adsorbers).

b.

Heat exchangers (including heat-traced piping).

c.

Pipe segments that operate liquid-full and at or below ambient temperature.

d.

Some block valves that trap small amounts of liquid when in the fullyopened or fully-closed position.

The cold sides of heat exchangers always have a source of heat available to produce a temperature increase. If the cold side should become blocked-in liquid-full while flow continued on the hot side, the Hydraulic Expansion contingency would require pressure relief. Similarly, liquids normally at or below ambient temperature have a potential heat source in both the ambient air and in solar radiation. Note that this can even apply to the hot side of a heat exchanger if both sides operate below ambient temperature. At some facilities, operation and maintenance procedures and training are established to ensure that the hot side of a heat exchanger is blocked-in before the cold side, or that the cold side is drained before being blocked-in. In this way, the cost of purchase, installation, maintenance, and documentation of a relief device can sometimes be avoided. At other locations, reliance on such procedures is not consistent with the local safety-management philosophy. It is important to note that, when documenting the pressure-relief contingencies for the exchanger in question, Hydraulic Expansion should still be identified as an applicable contingency, even if the above-mentioned practices are used. In the documentation, these practices should be noted as in use, removing the need for a relief device. The required relief flow rate for any Hydraulic Expansion contingency can be calculated from the trapped liquid’s heat capacity and thermal expansion coefficient, together with the duty of the heat source. The first two variables are straightforward properties of the liquid involved. For some types of heat sources, the duty is usually established with little difficulty. This is the case for the cold sides of heat exchangers, for heat-traced piping, and for any equipment for which solar radiation provides the heat source (it is conservative to assume a solar heat flux of 350 Btu/hr/ft2). However, when the atmosphere is the source of heat to a sub-ambient trapped liquid, the heating duty is not easily established. Fortunately, this duty is small enough that its determination is typically not required. API Recommended Practice 521 points out that it is common practice to use a ¾" x 1" NPS relief valve in most installations requiring thermal expansion relief. Experience has shown that this size device typically provides adequate relief capacity. Possible exceptions noted by Recommended Practice 521 include long, uninsulated pipelines and large vessels. In these cases, it is the usual practice to use the solar flux of 350 Btu/hr/ft2 and the exposed surface area of the equipment item to estimate the heat duty to the item.


1200-23

November 2001

1200 Relief Systems


Finally, because the relief flow rate required for this contingency is often a nominal value, the fact that most block valves leak becomes relevant. In some cases, the leak rate through some “closed” valves may be large enough that a thermal relief valve becomes unnecessary. Leak-through flow rates are available from many valve manufacturers, so a quantitative analysis of the situation may be possible. 7.

Exterior Fire Exterior fire is a potential cause of overpressure for any equipment item that has a liquid inventory and that may be exposed to an external fire. When heated by the flames, the liquid boils, generating vapor that must be relieved. In equipment without liquid, expansion of the contained gas or vapor due to fire-heating does generate a small relief requirement. However, API 520 notes that a pressure relief device may not significantly reduce the potential for rupture of a vapor-filled vessel exposed to fire. In the absence of evaporative cooling afforded by a liquid inventory, the vessel wall temperature quickly reaches failure levels. Therefore, relief devices are not usually installed on vapor-filled equipment solely to address the Exterior Fire contingency. In some instances, however, depressuring systems may be installed to vent vapor filled vessels, minimizing the stress of internal pressure on a fire-heated liquid-less vessel. In general, the relief flow rate required in a Fire contingency is equal to the vapor generation rate. This can be calculated from a heat flux, an area through which the heat flows, and the liquid’s enthalpy of vaporization at the relieving conditions. The particulars are a function of the type of equipment involved and of the particular standard to which adherence is adopted. Type of Equipment

Industry Standard or Regulation

Pressure Vessels

API RP 520 or RP 521

Low Pressure Storage Tanks

API Standard 2000 NFPA 30 OSHA 1910.106

Refrigeration Equipment

ASHRAE Standard 15

All of these methods include the effects of insulation, drainage, and firefighting equipment in the calculation of the heat flux. Thermal insulation limits the heat absorption from fire exposure as long as it remains intact. It is important to provide effective weather protection for the insulation so that it will not be removed by high-velocity fire hose streams or adjust the calculation to consider the quality of the installation. The differences among the methods involve precisely how these effects are included, and how the heat transfer area is calculated.

November 2001

1200-24



1200 Relief Systems

Overpressure Contingencies Applicable to Certain Equipment Types The following five causes of overpressure usually apply only to a small group of equipment types. For these types, however, these contingencies may nearly always apply. The first three contingencies in this group apply only to distillation systems, the heart of which is a distillation column. Therefore, these three will be discussed together. 1.

Distillation Systems In a typical distillation system, fluid components are separated by maintaining a temperature profile across a pressure vessel filled with contacting equipment such as trays or packing. A typical distillation system employs a reboiler and a condenser to maintain a hot to cold temperature profile from the bottom to the top of the column. In addition, liquids condensed in the condenser are usually added back to the top of the column as liquid reflux. Loss of heat removal from a distillation column can result in an additional vapor flow rate out the top of the distillation column, which may require pressure relief. Loss of heat removal can result from any of the following: a.

A loss of coolant flow to the overhead condenser, which in turn can produce a loss of reflux.

b.

A loss of reflux due to failure of a reflux pump or inadvertent valve closure.

c.

A loss of a cold feed stream to the top of the column.

d.

A loss of pump-around flow.

Because of the complexity of distillation systems, it is difficult to predict whether loss of condensing or reflux will require pressure relief without performing the detailed required relief rate analysis. For this reason, if the possibility exists for any of the above four contingencies to occur, then the loss of cooling/loss of reflux contingencies should be identified as applicable. The relief flow rate required in these contingencies is sometimes evaluated by simulating the loss of cooling or loss of reflux flow using a process simulator. For cases in which this approach is not feasible, API Recommended Practice 521 provides descriptions of simplified approaches to this calculation. For complex distillation systems in which cooling is provided in several places (condenser, multiple pump-arounds, etc.), consideration should be given to the possibility of a simultaneous loss of multiple cooling sources. A review of the coolant distribution and electrical distribution systems is often required to formulate the worst case. The topic of distillation systems would not be complete without a brief discussion of the impact that a cause of overpressure in one equipment item can have on other, connected items. This point is not unique to fractionation systems, but


1200-25

November 2001

1200 Relief Systems


they do provide a very clear example of the principle. Note that a distillation system usually consists of a column, reboiler, condenser, and accumulator, which often cannot be isolated from each other by block valves. In these cases, all causes of overpressure that affect the column also affect the other pieces of equipment in the distillation system and vice-versa. For example, a ruptured tube in the reboiler could overpressure the distillation column and a fire around the reboiler, accumulator and distillation column could overpressure the associated condenser. The causes of overpressure documentation should reflect this fact, but the required relief rates need only be calculated once. 2.

Absorbent Flow Failure When absorption equipment (typically a contacting tower) is operating normally, a portion of the inlet stream passes out of the equipment with the absorbent - the process effluent flow rate is smaller than the feed flow rate by an amount equal to the absorption rate. Upon loss of absorbent flow, the effluent rate must increase to the feed rate, or an additional effluent path (i.e., a relief device) must become available. If the feed pressure can exceed the absorber’s MAWP, the evaluation of the applicability of absorbent flow failure as a cause of overpressure becomes a question of the capacity of the effluent piping and downstream equipment relative to the feed flow rate. This effluent capacity, as well as the feed flow rate, should be evaluated at the absorber’s relief pressure; the increase in absorber pressure from operating to relief conditions can lead to both an increase in effluent capacity and a decrease in feed rate. API Recommended Practice 521 notes that while loss of lean oil flow to a hydrocarbon absorber generally does not cause an overpressure, both acid-gas removal units removing large fractions of the inlet stream, and absorbers removing carbon-dioxide from syn-gas streams upstream of methanators often do require overpressure protection against loss of absorbent flow. In any case, each individual absorber should be evaluated for the relevance of this causes of overpressure.

Causes of Overpressure That Are Infrequently Applicable The following causes of overpressure are relatively infrequently applicable to any type of equipment. Nonetheless, they should always be considered, and should be documented when they are applicable. 1.

Accumulation of Noncondensables The overhead piping configurations of some distillation systems may allow noncondensable gases to accumulate to such an extent that the overhead condenser is essentially blocked. For example, if the system is designed for total condensing of the overhead flow, any noncondensable materials that may inadvertently enter the system may collect in the reflux accumulator and “back up” into the condenser.

November 2001

1200-26



1200 Relief Systems

The effect of this contingency is the same as loss of cooling to the condenser, with eventual loss of reflux. The required relief flow rate can be calculated in the same way described above for that contingency. 2.

Entrance of Highly Volatile Material In some cases, the possibility exists for a volatile material (e.g., light hydrocarbon, or water) to enter into direct contact with a relatively hot fluid (e.g. heat transfer oil, or process fluid). This direct contact results in rapid vaporization of the volatile stream, which, in turn, may create substantial overpressure. Some of the equipment configurations in which this contingency is possible include heat exchangers (in which tube rupture may bring the hot and volatile fluids into direct contact), and water or steam connections to process vessels (when a single valve separates the process from the utility fluid). If the flow rate of volatile material and the amount of heat present in the hot fluid are known, the required relief flow rate may be calculated assuming vaporization to be instantaneous upon contact of the two fluids. However, these input data for the calculation are often unknown. Furthermore, the generation of vapor volume may be so rapid that the response time of a relief device is too long to supply effective relief flow. Therefore, it is essential that proper design and maintenance procedures be used to eliminate the possibility of mixing such streams.

3.

Abnormal Heat or Vapor Input It is sometimes possible for the duty of a heat exchanger to exceed its specified value. For example, a reboiler’s duty may be specified under the assumption of the presence of some deposits on the heat exchange surfaces; when the exchanger is freshly cleaned, these deposits will be absent, and the duty can be higher than specified. In such situations, the abnormally high heating can produce greater than design vapor flow rates, which may require pressure relief. The required relief flow rate is simply calculated as the vaporization rate in excess of the normal (design) rate. In other situations, a loss of heating can lead to a cause of overpressure. For example, in series distillation systems, the bottom product of one column provides the feed to the next column. If the reboiler for the upstream column loses heating, the bottom product of that tower will be unusually rich in volatile material. When this altered feed enters the normally-operating downstream column, it will yield an above-average vapor flow rate, which the downstream column may not be able to handle. Abnormal vapor input may result, for example, from an abnormally high feed rate to a distillation tower. In this case, the required relief flow rate is taken as equal to the excess vapor flow rate.

4.

Internal Explosions If sufficient air, oxygen, or other compound capable of supporting combustion can infiltrate an equipment item containing flammable gases, vapors, or dusts, an explosive mixture may result inside the equipment. Ignition of such a


1200-27

November 2001

1200 Relief Systems


mixture can result in either a deflagration or a detonation, which differ in the speed of propagation of the combustion flame front. In the event of a deflagration, venting the overpressure generated is possible, although the methods for designing such a vent are somewhat dependent on the precise configuration of the equipment involved. These methods are discussed in NFPA 68, Guide for Venting Deflagrations. NFPA 69, Explosion Prevention Systems, provides a description of methods for controlling deflagrations other than venting – for example, total containment, suppression, and oxidant concentration reduction. Venting of a detonation, in which the propagation speed of the flame front exceeds the local speed of sound, is generally not possible. For cases that hold the possibility of a detonation, prevention of the creation of the flammable mixture and/or of a source of ignition becomes of great importance. 5.

Chemical Reaction In some cases, uncontrolled chemical reaction can present a potential causes of overpressure. Intended reactions can become uncontrolled and unintended reactions can occur without control due to any of several root causes: a.

Loss of temperature control, due to an external fire, loss of cooling, or loss of mixing.

b.

Loss of reaction mixture control due to loss of mixing, loss of a feed stream, incorrect feed of the correct compounds, or feed of incorrect compounds.

Estimation of the relief flow rate required in a chemical reaction contingency is a complex process, usually requiring bench-scale experiments simulating the expected reaction upset conditions. This is necessarily highly dependent on the specific substances involved, the specific process being evaluated, and the specific conditions expected to produce the reaction upset. Publications from the Design Institute for Emergency Relief Systems (DIERS) and the Center for Chemical Process Safety (CCPS) of the American Institute of Chemical Engineers (AIChE) provide further detail and additional sources of information on the subject of the chemical reaction causes of overpressure.

1230 Pressure Relief Device Selection The purpose of this section is to identify the common types of pressure relief devices and provide guidance that will aid in the selection of the appropriate device for a given application. A basic description of the operation, benefits, and limitations is presented for each type of device. Refer to API Recommended Practice 520.

1231 Conventional Pressure Relief Valves A conventional pressure relief valve is designed to open when the differential pressure across the valve reaches the net set pressure. The valve will reseat when the differential pressure drops below the net set pressure minus the blowdown.

November 2001

1200-28



1200 Relief Systems

Figure 1200-12 shows a schematic diagram of a typical conventional pressure relief valve. Because the fluid on the discharge side of the valve is in contact with the full area of the upper surface of the valve’s disk, the set pressure of this valve is the sum of the net set pressure (the resistance to opening supplied by the spring) and any superimposed back pressure at the valve’s outlet; see Figure 1200-13. When specifying a conventional valve for an application with superimposed back pressure, it is important to clearly specify the set pressure (i.e., the desired opening gauge pressure), the back pressure in service, and the net set pressure. Fig. 1200-12 Typical Conventional Pressure Relief Valve (Courtesy of the American Petroleum Institute)

Most conventional pressure relief valves have internals (trim) designed specifically for either liquid or vapor service. Consideration should be given to all potential phases of flow through the relief valve when selecting the trim. The conventional pressure relief valve is the most commonly used pressure relief device in the process industries. While it has some disadvantages relative to the rupture disk (see below), its ability to reclose represents an overwhelming advantage in most applications. By opening and then reclosing in response to the appearance and removal of a cause of overpressure, a pressure relief valve relieves the excess pressure without venting an entire vessel or a continuous process. Due to the simplicity of its design, a conventional valve is preferred to the other types of pressure relief valves in terms of reliability, cost, and maintenance. This also provides for future expansion and additional relief loads. Other types should be selected only when service conditions (as discussed below) or regulations require them.


1200-29

November 2001

1200 Relief Systems


Fig. 1200-13 Effect of Superimposed Back Pressure on the Set Pressure of a Conventional Pressure Relief Valve

Equipment MAWP = Desired Set Pressure

Equipment MAWP = Desired Set Pressure

Net Set Pressure Set Pressure Net Set Pressure

Set Pressure

Superimposed Back Pressure Superimposed Back Pressure

(b)

(a)

Limitations on Use Some of the circumstances in which a conventional pressure relief valve should not be used are described in the following sections. 1.

Back Pressure Because conventional pressure relief valves act on differential pressure, they should generally not be used whenever the back pressure at the valve’s outlet would interfere with the valve’s reliable operation. Normally not to exceed 10% of the set pressure of the lowest set PSV connected to the relief system. This interference can occur when: a.

November 2001

The superimposed back pressure is sufficiently variable that a net set pressure can not be chosen to keep the valve’s set pressure both less than the protected equipment’s MAWP and greater than the equipment’s operating pressure by an amount large enough to avoid leakage through the valve

1200-30



1200 Relief Systems

(see Figure 1200-13 and the discussion below concerning high operating pressure).

2.

b.

The built-up back pressure is calculated to be sufficiently high to raise concern about causing the valve to reclose before the equipment’s pressure has fallen below its MAWP. Such reclosing can lead to the valve “chattering” – see API Recommended Practice 520, Part I for a discussion of this behavior. The “concern level” of built-up back pressure for a conventional pressure relief valve is typically taken to be equal to the vessel’s allowable overpressure. See “Outlet Piping” on page 1200-47 for a discussion of the calculation of and acceptable values for built-up back pressure.

c.

The built-up back pressure from relief valves opening first is not considered a superimposed back pressure on the other relief valves opening later in response to the same contingency.

High Operating Pressure Conventional pressure relief valves should generally not be selected for applications in which the normal operating pressure can exceed 90% of the valve’s set pressure. This is because leakage of process fluid through the closed pressure relief valve, while typically zero at operating pressures below 90% of the set pressure, can increase significantly as the operating pressure increases beyond the 90% level.

1232 Balanced Pressure Relief Valves A balanced pressure relief valve is designed to be much less dependent than a conventional valve on the back pressure present at its outlet. To achieve this goal, balanced valves most often contain a bellows that serves to isolate a portion of the valve disk from the back pressure. The inside of the bellows is in pressure communication with the valve bonnet, which must be vented to a location of constant low pressure, typically the atmosphere. Other balanced valve designs exist (see API Recommended Practice 520 Part I), but the bellows valve is by far the most prevalent and has been shown to be the most reliable in most installations. With the exception of the incorporation of the vented bellows for pressure balancing, the construction and characteristics of a bellows valve are identical to those of a conventional relief valve. In fact most conventional valves can be converted to balanced performance merely by installation of the bellows and its gaskets and removal of a plug in the bonnet vent. Figure 1200-14 shows a schematic diagram of a typical balanced bellows pressure relief valve. Balanced PSVs are most often used to tie a new low-pressure relief load into an existing heavily loaded relief header or to protect the PSV topworks from corrosive gases in the relief header. A balanced pressure relief valve is generally the preferred relief device whenever back pressure conditions preclude the use of a conventional pressure relief valve. Because the bellows isolates many of the internal parts from the fluid on the discharge side of the valve, bellows-type balanced valves also find application when the discharge system contains corrosive or fouling fluids. These advantageous prop-


1200-31

November 2001

1200 Relief Systems


erties of a balanced valve come at the expense of a higher cost and increased maintenance requirements relative to a comparable conventional valve. Fig. 1200-14 Typical Bellows Type Balanced Pressure Relief Valve (Courtesy of the American Petroleum Institute)

Limitations on Use 1.

Back Pressure While designed to minimize the effects of back pressure on their performance, balanced pressure relief valves are not immune to these effects. In general, the flow capacity of balanced valves is reduced at total back pressures greater than roughly 17% of set pressure for liquid flow and 30% of set pressure for vapor flow. An individual balanced valve’s manufacturer should be consulted regarding the onset and rate of capacity reduction as a function of back pressure, and for possible onset of unstable performance (i.e., possible chattering) at high levels of built-up back pressure. Bellows-type PSVs can be used with header back pressures as high as 40% of the set pressure. Because the bellows itself is fabricated from thin, corrugated metal available in a limited number of materials, it may have a pressure rating that is below the rating of the valve body and outlet flange. Thus, the selection of a bellows type valve may impose a limit on the absolute value of the back pressure (rather than its fraction of set pressure) that would not be present if a conventional valve were selected.

November 2001

1200-32



2.

1200 Relief Systems

Potential Atmospheric Discharge The bonnet of a balanced pressure relief valve must be vented to a location of constant low pressure; the atmosphere is the typical choice of venting location. If the bellows should fail, fluid from the discharge side of the valve may be discharged from the bonnet vent. Therefore, careful consideration should be given to the nature (particularly the toxicity) of the discharge fluids and the safety of the bonnet vent’s ultimate discharge location.

3.

High Operating Pressure Balanced pressure relief valves exhibit the same behavior as conventional valves with respect to seat leakage as the operating pressure exceeds 90% of set pressure. Therefore, they should generally not be selected for applications with high ratios of maximum operating pressure to set pressure.

4.

Periodic Inspection When bellows-type PSVs are installed, it is necessary to periodically check that the bellows is intact. A leaking bellows does not provide back pressure compensation and it allows the relief header to leak to atmosphere.

1233 Pilot-Operated Pressure Relief Valves Pilot-operated pressure relief valves are characterized by the presence of a pilot valve in addition to the main valve, which encloses a floating piston. The floating piston is designed with larger effective area on its top than on its bottom. A sensing line is used to allow the process pressure to be sensed by the pilot and to act on both the top and bottom surfaces of the floating piston. At the set pressure, the pilot opens and vents the pressure from the top of the piston, allowing it to open to relieve overpressure in the protected equipment. Many pilot valve designs are available, resulting in different operating characteristics. The pilot design may result in pop action or modulating action of the piston. In addition, the process fluid may or may not actually flow through the pilot (flowing or non-flowing). Due to the variety of available designs, the manufacturer should be consulted for particular applications. Figure 1200-15 shows a schematic diagram of a typical pilot-operated pressure relief valve. Pilot-operated valves have a few primary advantages that may warrant selective application. The required margin between operating and set pressure is smaller than for either conventional or balanced valves, because the process pressure actually holds the floating piston tightly closed until the pilot vents. Designs are available that will tolerate higher levels of back pressure than conventional or balanced pressure relief valves. The effect of inlet line flowing pressure drop (see “Inlet Piping” on page 1200-43) can be avoided by connecting the pressure sensing line directly to the protected equipment (remote sensing).


1200-33

November 2001

1200 Relief Systems


Fig. 1200-15 Typical Pilot-Operated Pressure Relief Valve (Courtesy of the American Petroleum Institute)

Limitations on Use 1.

Corrosive or Fouling Fluids Flow paths through the pressure-sensing tubing and the pilot valve of a pilotoperated pressure relief valve are generally quite narrow, and therefore subject to plugging. Consequently, pilot-operated pressure relief valves should generally not be selected for installations in which service fluids are corrosive, dirty, viscous, or polymerizable. In addition, one distinct disadvantage to pilot-operated PSVs is that, if the sensing tubing comes loose due to vibration or thermal cycling, the result can be an undesirable full flow.

1234 Rupture Disks Rupture disks are nonreclosing pressure relief devices designed to burst at a specified differential pressure, enabling flow through the area formerly blocked by the burst material. Rupture disks are manufactured in a variety of designs, for both liquid and vapor services, from a variety of materials. Consequently, they can have a variety of performance characteristics with respect to corrosion resistance, operating pressure margins, burst pressure ranges, and service lifetimes. See API Recommended Practice 520, Part I, the CCPS monograph Guidelines for Pressure Relief and Effluent Handling Systems, and manufacturer’s literature for further discussions of the various designs of rupture disk devices and their performance

November 2001

1200-34



1200 Relief Systems

characteristics. Figure 1200-16 shows a diagram of a typical rupture disk and its holder. Fig. 1200-16 Typical Rupture Disk Installation (Courtesy of the American Petroleum Institute)

1.

2.

Advantages of Rupture Disks a.

They open and allow flow more rapidly than valves.

b.

They have a larger flow capacity than a pressure relief valve of identical inlet nominal pipe size.

c.

They are less prone to leakage at pressures below their stated operating pressure limits.

d.

They are simpler devices, and therefore generally cheaper to purchase. They require little or no maintenance, but, rather, are replaced periodically.

Disadvantages of Rupture Disks a.

Complete de-pressure with potentially large loss of inventory.

b.

Need to replace before process can be re-started.

c.

Need a larger gap between operating and set pressure, consequently the equipment has to be designed for higher pressure.

d.

Subject to fatigue failure.

e.

The recommended interval for replacement is generally one year.


1200-35

November 2001

1200 Relief Systems


Because of the short life, rupture disks are generally not recommended for refinery equipment to operate continuously for a number of years between shutdowns. In light of their characteristics, rupture disk devices are selected for applications requiring rapid opening (such as for protection of the low pressure side of a heat exchanger against a tube rupture contingency) and/or high flow capacity. In these applications, rupture disk devices are used as stand-alone relief devices. In an alternative application, their positive closure and low cost lead rupture disks to be installed at the inlet or outlet of pressure relief valves to isolate the valve from fluids that may be toxic, corrosive, or that might cause fouling of the valve. In these applications it is important to note that the installation of a rupture disk in series with a pressure relief valve causes a small reduction in the valve’s flow capacity (see “Rupture Disks Upstream of Pressure Relief Valves” on page 1200-43 for further discussion of the combination capacity factor). Note Retrofitting of an existing refinery relief valve with a rupture disk is not a viable option for two reasons: 1) it requires yearly replacement, and 2) it would generally require reducing the set pressure and could require increasing the relief valve size. Rupture discs should be mounted in the holder provided by the maker to avoid premature rupture due to pipe stress. If back pressure or vacuum can cause pressure reversal, a rupture disc with adequate vacuum strength is required. When ordering, specification data must include expected temperature at rupture conditions (not normal operating temperature). Disc rupture pressure is greatly affected by temperature; too many discs have ruptured prematurely because they were not specified correctly.

Limitations on Use The principle limitation on the selection of rupture disks as pressure relief devices is that once activated, they do not reclose. Therefore they are inappropriate for any application in which venting of an entire vessel or group of equipment items is not acceptable. Note also that when a rupture disk is used alone, the venting of process material due to a possible premature failure of the disk must be tolerable.

1235 Breaking Pin Devices Also known as buckling pin devices or rupture pin devices, these are nonreclosing pressure relief devices in which a pressure-containing member is held closed against the system pressure by a load-bearing pin that has been engineered to fail when the system pressure reaches some specified pressure. When the pin fails, the pressurecontaining member is moved by the system pressure, allowing relief flow through the now-open device. Breaking pin devices have many of the same advantages as rupture disks - rapid-opening performance, high capacity per unit nominal pipe size, simplicity of design, and ease of maintenance. These devices also have the additional advantage that the load-bearing component is external to the process fluid. This means that the load-bearing part is not subject to corrosion, fouling, etc. due to

November 2001

1200-36



1200 Relief Systems

the process fluid, and that the equipment need not be opened to atmosphere to reset the device after it has operated. Like rupture disks, breaking pin devices have the distinct disadvantage of nonreclosing operation. They also have the unique disadvantage of being easily rendered inoperable by the insertion of a rigid object in the space normally occupied by the breaking pin. Breaking pin devices have not yet achieved significantly wide application in the process industries.

Limitations on Use Section VIII of the ASME Boiler and Pressure Vessel Code currently restricts the use of breaking pin devices to the inlet of a pressure relief valve. In such installations, these devices are subject to the same requirements as are rupture disks.

1236 Relief Device Accessories The various types of relief devices can be equipped with accessories that may be required when the device is used in certain applications. Some of the most frequently encountered accessories are: •

Lifting lever. Pressure relief valves in contact with certain service fluids are required by Section VIII of the ASME Boiler and Pressure Vessel Code to be equipped with a lifting lever to facilitate periodic in-place testing of the valve’s freedom of movement. These fluids are water at temperatures above 140 °F, air, and steam.

•

Open bonnet. Pressure relief valves in steam service (particularly within the scope of ASME Boiler and Pressure Vessel Code Section I) are generally equipped with open bonnets (i.e., the spring and spindle are exposed).

1237 Materials Selection As for any process instrument or piping component, consideration should be given to selecting the materials of construction of a pressure relief device. All wetted parts should be compatible with all expected process fluids. The individual facility’s materials specifications and/or the facility materials engineer should be consulted to ensure selection of materials appropriate to the application.

1240 Pressure Relief Device Set Pressures Sizing and Installation The purpose of this section is to present guidelines for the proper setting, sizing, and installation of pressure relief devices.


1200-37

November 2001

1200 Relief Systems


1241 Relief Device Set Pressures The maximum allowable set pressure for a relief device is determined by the pressure rating of the equipment being protected and by the number of relief devices installed. The maximum allowable relieving pressure is equal to the maximum allowed accumulated pressure, which is a function of the MAWP, the causes of overpressure, and the number of relief devices installed. Figure 1200-17 below summarizes these maximum limits as dictated by ASME Boiler and Pressure Vessel Code Sections I and VIII and by ASME Piping Code B31.3. Figure 2 of API RP 521 shows the relationship between the various pressures involved in pressure relief, and it defines the terminology. Figure 2 shows a 10% margin between the maximum operating pressure (Po) and MAWP. To avoid PSV leakage and excessive maintenance, provide the following margins between Po and MAWP: 25 psi for Po 15% of Po for Po 45 psi for Po 10% of Po for Po

= = = =

0 - 170 psi 170 - 300 psi 300 - 450 psi above 450 psi

The above recommendations are for conventional and balanced PSVs. Pilot-operated PSVs allow the vessel to be operated closer to the MAWP. The minimum set pressure for conventional PSVs is 5 to 10 psig. The minimum set pressure for bellows-type PSVs is approximately 25 psig. Both of these minimums depend on the size of the PSV. Consult the manufacturers’ catalogs. Fig. 1200-17 Maximum Relief Device Set Pressures and Equipment Accumulated Pressures for Installations of Single and Multiple Pressure Relief Devices Allowed by ASME Codes (where the protected equipment pressure rating MAWP is 100) (1 of 2) ASME Section VII Unfired Pressure Vessels Contingency

Single Valve Installations Maximum Set Pressure

Maximum Accumulated Pressure

Multiple Valve Installations Maximum Set Pressure


Non-Fire First Valve Additional Valves Fire Only

First Valve Additional Valves Supplemental

Valves

November 2001

1200-38



1200 Relief Systems

Fig. 1200-17 Maximum Relief Device Set Pressures and Equipment Accumulated Pressures for Installations of Single and Multiple Pressure Relief Devices Allowed by ASME Codes (where the protected equipment pressure rating MAWP is 100) (2 of 2) ASME Section I Boilers

Single Valve Installations

Contingency

Non-Fire

Maximum Set Pressure




First Valve Additional Valves

Fire Only


ASME B31.3 Process Piping Contingency

Non-Fire

Single Valve Installations Maximum Set Pressure





Fire Only


For low pressure equipment, the specific design code (e.g., API Standard 620, 650, etc.) should be consulted to determine the maximum relief device set pressure and overpressure.

1242 Relief Device Capacity Calculation Pressure Relief Valve Sizing Pressure relief valve sizing is based on theoretical nozzle flow equations adjusted using various empirical factors to account for the non-ideality of the flow. The typical sizing methods are summarized below with particular emphasis on the various correction factors. With the exceptions noted, these methods are applicable to conventional, balanced, and pilot-operated pressure relief valves. 1.

Non-Flashing Liquid API Recommended Practice 520, Part I provides (in Section 4.5.1 of the 6th [1993] edition) the standard sizing equation for liquid releases that are not expected to flash across the relief valve. The flow is a function of the relief valve nozzle area, the liquid density, and the pressure drop across the valve. This equation includes correction factors for the liquid’s viscosity (Kv), the back pressure (Kw) (in the case of a balanced valve), and flow non-ideality


1200-39

November 2001

1200 Relief Systems


(Kd). The viscosity correction factor decreases from a value of one only as the flow becomes highly viscous. Values of Kv as a function of the Reynolds number are given in RP 520. Values of Kw for balanced relief valves should normally be obtained from the valve manufacturer; if the value for the specific valve under consideration is unavailable, the generic curve given in API RP 520 may be used. Values of the discharge coefficient (Kd) are determined experimentally in the process of the ASME-required certification of an individual pressure relief valve model’s flow capacity. Kd is simply the valve’s measured flow capacity divided by the valve’s theoretical capacity under the same conditions of relief pressure, temperature, and fluid. Values of Kd are given in manufacturer’s catalogs, and are published in NB-18, Pressure Relief Device Certifications. Note that in calculating the theoretical flow capacity, the valve nozzle’s actual discharge area is used. Therefore, this measured value of Kd must be used only with the actual discharge area when calculating the valve’s flow capacity. The measured value of Kd and the actual discharge area are a “matched set,” and should always be used as such. This point is raised because API Standard 526 sets forth “standard orifice areas” for flanged pressure relief devices. In the absence of specific manufacturer data for Kd, API recommends a value of 0.62 for liquid flow calculations. This value should be used only with the standard orifice areas (a matched set different from the model-specific set). Mixing a measured Kd with a standard orifice area or an actual area with API’s default Kd will generally lead to inaccurate calculation of a pressure relief valve’s flow capacity. The ASME Pressure Vessel Code formerly permitted 25% overpressure for liquid PSVs to reach full capacity; therefore, the Company sized liquid PSVs using 25% overpressure. The ASME Pressure Vessel Code was revised in 1985 to require that liquid PSVs pass their full rated capacity with 10% overpressure. The latest designs of liquid PSVs can meet this requirement. When reusing old liquid PSVs designed for 25% overpressure, it is necessary to recalculate the PSV capacity using 10% overpressure. Routine maintenance of old PSVs does not require recalculation with 10% overpressure unless the relieving case has changed. 2.

Vapor For all-vapor releases across the relief valve, the API Recommended Practice 520 provides the standard sizing equation (in Section 4.3.2.1 of the 6th edition). The flow is a function of the nozzle area, the inlet absolute pressure and temperature, and several fluid properties. In addition to the discharge coefficient (Kd), discussed above, the equation also includes a correction factor (Kb) for the effects of back pressure on the flow through a balanced pressure relief valve. The value for Kb should normally be obtained from the manufacturer. If Kb is not available for a specific valve, the generic curve given in API RP 520 may be used. The measured values of Kd are a function of the phase of the relief fluid. Therefore, the values discussed above for liquid sizing are not to be used when sizing

November 2001

1200-40



1200 Relief Systems

for gas or vapor flow. Valve manufacturers and the National Board publish separate values of Kd for vapor flow. The same caveats discussed above apply when sizing for vapor flow; the API recommended default value of Kd for vapor flow is 0.975. The vapor-sizing equation discussed above was derived for the case of critical flow – that is, flow that is independent of the downstream pressure. In the unusual occurrence that the downstream pressure is on the order of 50% of the relief pressure, the vapor flow is in the subcritical regime, and a different equation describes the flow. See API Recommended Practice 520 Part I for further description of critical and subcritical flow regimes. For convenience, API has generated a set of values of Kb as a function of pressure downstream of the pressure relief valve that enable the use of the critical flow equation to calculate the subcritical flow capacity of conventional and pilot-operated valve. Thus, with the inclusion of an additional generic curve for Kb, one equation can suffice for both flow regimes. 3.

Steam The steam sizing equation presented in API RP 520, Part I (Section 4.4.1 of the 6th edition) should be used for sizing pressure relief valves for saturated or superheated steam flow. This equation is very similar to the vapor-sizing equation, with the physical properties of saturated steam built-in. Accordingly, the steam sizing equation yields results very similar to the vapor sizing equation, but requires less physical property data. Property differences between saturated and superheated steam are accounted for via a superheat correction factor (Ksh); RP 520 provides a table of Ksh values as a function of steam pressure and temperature. Deviations from ideal gas behavior at high pressures (> 1500 psi) enter the calculation via the Napier equation factor, KN, which is a simple function of the absolute pressure. Finally, the steam sizing equation also contains the discharge coefficient, Kd, in the same manner as the liquid and vapor sizing equations. When valve-specific data are not available for steam flow, API RP 520 recommends a value of 0.975, consistent with the behavior of other vapors.

4.

Two-Phase Fluids Releases that result in both liquid and vapor flow simultaneously through the pressure relief valve are classified as two-phase. Two-phase releases typically result from either (1) a liquid release that flashes (i.e., reaches saturation) at a pressure above the valve’s constant back pressure or (2) inadequate free-board volume in the overpressured equipment to allow vapor-liquid disengagement upstream of the relief valve – the two-phase flow in the equipment then continues into the valve. In order to determine the potential for two-phase flow from an initially all-liquid release, a process simulation package should be used to flash the relief stream adiabatically from relieving conditions to the constant back pressure. In the event that the stream remains all-liquid, the methods summarized above (item 1, “Non-Flashing Liquid” on page 1200-39) should be used. If the liquid begins to vaporize, two-phase flow calculation methods are required. To determine the potential for two-phase relief flow due to a lack of vapor-liquid disengagement, consideration must be given to the fluid proper-


1200-41

November 2001

1200 Relief Systems


ties, available disengagement space, and vessel geometry. Consult a two-phase relief expert whenever adequate disengagement is an issue. Given the variation in the physical properties across the relief valve, two-phase flow capacity evaluations are not amenable to simple hand calculation. In addition, different situations may require application of specific methods to achieve an accurate result. Refer to Guidelines for Pressure Relief and Effluent Handling Systems for further information on vapor disengagement and twophase relief sizing.

Rupture Disk Sizing Either of two approaches may be used for sizing rupture disk devices: (1) nozzle flow and (2) pipe flow. Each method is described both in Section VIII of the ASME Boiler and Pressure Vessel Code (BPVC) and in API Recommended Practice 520 Part I. 1.

Rupture Disk Sizing Using Nozzle Flow Equations Rupture disks can be sized using the equations presented in “Pressure Relief Valve Sizing” on page 1200-39 for pressure relief valves. In this method, the equation applicable to the specific fluid phase is used, but the coefficient of discharge (Kd) is set to a value of 0.62 regardless of the phase. Paragraph UG127 of BPVC Section VIII expressly restricts the use of this sizing approach to rupture disks discharging directly to atmosphere, and with inlet and outlet lines at least as large as the rupture disk and no longer than eight and five pipe diameters, respectively.

2.

Rupture Disk Sizing Using Pipe Flow Equations (KR Method) For rupture disk device installations in which the discharge is routed to a closed system or which include a significant inlet piping system, the disk is sized as one component in the entire inlet line – rupture disk device – outlet line system, using standard pipe flow equations. In this approach, each fitting contributes to the overall flow resistance factor of the piping system. The resistance factor for the burst rupture disk (KR) is that determined for each disk model as specified in paragraph UG-131 of Section VIII. Certified values of KR are available from individual disk manufacturers and are published by the National Board of Boiler and Pressure Vessel Inspectors in NB-18, Pressure Relief Device Certifications. In the absence of measured values of KR for a specific rupture disk device, a value of KR = 2.4 is required by Section VIII. Once the rupture disk/piping system total resistance is defined, the system’s flow capacity is calculated by setting the inlet pressure equal to the relieving pressure and the outlet pressure equal to the constant back pressure. Alternatively, the required flow rate and the constant back pressure can be specified to determine the pressure in the vessel required to provide adequate relief. In the first approach, the calculated rupture disk/piping system capacity is compared to the calculated required relief rate, while in the second approach, the calculated pressure in the vessel is compared to the MAWP plus allowable accumulation to determine the adequacy of the flow capacity.

November 2001

1200-42



1200 Relief Systems

Rupture Disks Upstream of Pressure Relief Valves When a rupture disk device is installed upstream of a pressure relief valve, no capacity calculations are performed for the rupture disk. Rather, the effect of the rupture disk on the flow capacity of the device combination is represented by an additional capacity correction factor in sizing the valve. This “combination capacity factor” (Kc) may be determined experimentally by rupture disk manufacturers for each model of rupture disk in combination with individual pressure relief valve models, as specified in the ASME Boiler and Pressure Vessel Code, Section VIII, paragraph UG-132. These measured values of Kc are available from the disk manufacturers and are published in the National Board publication NB-18, Pressure Relief Device Certifications, for each combination of rupture disk model and pressure relief valve model tested. For disk/valve combinations for which no value of Kc has been measured, Section VIII requires the use of a value of 0.9. ASME UG-127 outlines special mandatory requirements for installing rupture discs under PSVs. These requirements include: •

A means (i.e., a tell-tale) to check that the rupture disc is intact and that there is no liquid or pressurized gas in the cavity between the rupture disc and the PSV

•

Derating relief valve capacity to 80% of valve design, unless the specific combination of rupture disc, disc holder, and PSV has been flow tested. This topic is covered in ASME UG-132. Note that it is necessary to include the rupture disc when calculating relief inlet pressure drop

•

Reduction in operating pressure to match the larger bursting tolerance of the rupture disc

The rupture disc vendors make special holders which mount under the PSV and which include a tell-tale.

1243 Pressure Relief Valve Installation To help ensure the proper operation of pressure relief valves - and thereby ensure the protection of process equipment against overpressure - certain guidelines should be followed in the installation of pressure relief valves. The following sections briefly describe these installation guidelines.

Inlet Piping 1.

Hydraulic Considerations The principal requirements on the hydraulic performance of a pressure relief valve inlet line are specified in ASME Boiler and Pressure Vessel Code, Section VIII. At paragraph UG-135 (“Installation”), Section VIII states, The opening through all pipe, fittings, and nonreclosing pressure relief devices (if installed) between a pressure vessel and its pressure relief valve shall have at least the area of the pressure relief valve inlet. The characteristics of this upstream system shall be such that the pressure drop will not reduce the relieving capacity below that required or adversely affect the proper operation of the pressure relief valve.


1200-43

November 2001

1200 Relief Systems


More specific guidance concerning acceptable inlet pressure drop is provided in Non-mandatory Appendix M of Section VIII, which specifies that for valves in compressible fluid service, “the cumulative total of all nonrecoverable inlet losses shall not exceed 3% of the valve set pressure. The inlet pressure losses will be based on the valve nameplate capacity corrected for the characteristics of the flowing fluid.” This guidance of a maximum of 3% of the set pressure for the frictional inlet flow losses is reiterated in API Recommended Practice 520 Part II, and is the generally accepted practice. The adverse effect that this is intended to avoid is the repeated opening and closing of the valve known as chatter. Chattering typically causes a reduction in relief capacity and can result in premature valve failure. See Guidelines for Pressure Relief and Effluent Handling Systems, Section 2.4.2.2.1, for additional discussion of chatter. Recall (Section 1233) that a pilot-operated pressure relief valve can be installed with its pressure sensing line connected directly to the protected equipment, rendering the opening and closing of the valve independent of the flowing inlet pressure drop. The evaluation of the non-recoverable (i.e., frictional only) inlet line pressure loss is performed using standard equations for fluid flow through pipe and fittings. Each pipe segment or fitting is represented in the calculation by a flow resistance coefficient or an equivalent length of pipe. As specified in the excerpt from Section VIII Appendix M quoted above, the flow rate is assumed to be the full nameplate (i.e., rated) capacity of the valve. This inlet pressure loss is evaluated for all causes of overpressure in which the valve is expected to relieve. “Nonrecoverable losses” refers only to the frictional losses, neglecting the effects of potential and kinetic energy. 2.

Inlet Line Stresses The discharge flow from a pressure relief valve generates forces that can result in substantial stresses on the inlet nozzle and its connection to the protected vessel. The magnitude of these forces depends on the size of the valve, the configuration of the outlet piping, and the temperature change due to the discharge. API Recommended Practice 520 Part II presents an equation for quantifying the magnitude of this force for open discharge systems. In addition to this reaction force stress, temperature gradients across a pressure relief valve discharging an auto-refrigerating fluid can induce additional stress on the inlet line due to contraction of the piping. An experienced piping designer should generally be consulted to specify support members for pressure relief valves discharging directly to atmosphere. For large closed systems, a pipe stress analysis is usually performed prior to installation.

3.

Isolation Valves Installation of an isolation (i.e., block) valve in the inlet line to a pressure relief valve enables the valve to be removed for repair or for routine inspection and maintenance without isolating or venting the protected equipment. Most applicable codes allow the installation of such a valve, provided it meets certain minimum requirements designed to ensure that the relief device is not inadvert-

November 2001

1200-44



1200 Relief Systems

ently isolated from the equipment it is intended to protect. Furthermore, as the installation of an isolation valve implies the intention to remove the pressure relief valve while continuing to operate the process equipment, an alternative protection scheme should be planned. This alternative can be a spare relief device that would be installed immediately upon removal of the “primary” pressure relief valve, a normally isolated relief device that would be connected to the equipment when the primary valve is isolated, specialized operating procedures used while the primary valve is removed, or a combination of these approaches. Note Section I of the ASME Boiler and Pressure Vessel Code explicitly forbids the installation of an isolation valve in the inlet line of a safety valve within its scope. Some specific guidelines concerning installation of isolation valves in a pressure relief valve’s inlet line are: a.

Such valves are to be full bore.

b.

They are to be capable of being car-sealed or locked open.

c.

If a gate valve is used as such an isolation valve, it is to be installed with its stem oriented horizontally so that an internal failure will not cause the gate to fall closed.

If two full-capacity pressure relief valves are installed via a tee on the same vessel nozzle, with one valve intended as a spare for the other, it is important that only one inlet isolation valve be open at a time. The excess flow capacity afforded by both pressure relief valves could produce excessive inlet pressure drop. Consideration should be given to the use of a three-way isolation valve capable of creating an open pathway to only one of the two pressure relief valves at a time. 4.

Rupture Disk on Inlet Line Several pressure relief valve installation guidelines are peculiar to the case of a rupture disk in the inlet line. Each of these is required by Section VIII of the ASME Boiler and Pressure Vessel Code. a.

The rupture disk should be of a nonfragmenting design. Fragments may interfere with the operation of the pressure relief valve.

b.

The rupture disk, when burst, should have a net flow area equal to or greater than the inlet connection area of the pressure relief valve.

c.

The space between the rupture disk and the pressure relief valve must be equipped with a free vent, a bleed valve, an excess-flow valve, a pressure indicator, or some other means of detecting rupture disk leakage. Build-up of pressure in this space would increase the burst pressure of the rupture disk, which operates on the pressure differential across it.

d.

When evaluating the nonrecoverable pressure drop across the inlet line, the rupture disk device should be represented in the list of fittings with its


1200-45

November 2001

1200 Relief Systems


certified flow resistance factor (KR). If no certified value is available, a value of 2.4 should be used. Figure 1200-18 shows a typical installation of a rupture disk in the inlet line of a pressure relief valve. Fig. 1200-18 Rupture Disk/Relief Valve Installation (Courtesy of the American Petroleum Institute)

5.

General Installation Details To reduce inlet line pressure loss, the inlet line length should be as short as practical, and the exit from the vessel to the pressure relief valve should be well-rounded.

November 2001

1200-46



1200 Relief Systems

The inlet line should be self-draining back to the vessel, and should be designed to facilitate access to the pressure relief valve for inspection and maintenance. Pressure relief valves should always be installed in the upright position, with their stems oriented vertically. Other orientations can lead to seat damage or failure to reseat in a leak-free manner.

Outlet Piping 1.

Hydraulic Considerations As described above for inlet lines, the principal hydraulic performance requirements of a pressure relief valve outlet line are also specified in the ASME Boiler and Pressure Vessel Code, Section VIII, paragraph UG-135: The size of the discharge lines (from pressure relief devices) shall be such that any pressure that may exist or develop will not reduce the relieving capacity of the pressure relief devices below that required to properly protect the vessel, or adversely affect the proper operation of the pressure relief devices.

Once again, more specific guidance concerning acceptable outlet pressure drop is provided in Non-mandatory Appendix M of Section VIII, The flow characteristics of the discharge system of high lift, top guided safety, safety relief, and pilot-operated pressure relief valves in compressible fluid service shall be such that the static pressure developed at the discharge flange of a conventional direct spring loaded valve will not exceed 10% of the set pressure when flowing at stamp capacity. Other valve types exhibit various degrees of tolerance to back pressure and the manufacturer’s recommendation should be followed.

API Recommended Practice 520 reiterates this guidance that the built-up back pressure on a conventional valve should be limited to 10% of set pressure when flowing at the rated capacity – i.e., at 10% overpressure. However, acknowledging the connection between the valve’s allowable overpressure and the back pressure at which the valve may begin to reclose, RP 520 Part I extends this guidance, Conventional pressure relief valves should typically not be used when the built-up back pressure is greater than 10 percent of the set pressure at 10 percent overpressure. A higher maximum allowable built-up back pressure may be used for overpressure greater than 10 percent.

These guidelines are specific only for conventional, spring-loaded pressure relief valves. For other types of valves, the manufacturer should always be consulted for values of the built-up back pressure at which a specific valve suffers loss in flow capacity and/or in stable performance. The generic-use data for the back pressure capacity correction factor for balanced pressure relief valves presented in API RP 520 Part I indicate that such valves suffer a loss in flow capacity at built-up back pressures exceeding 17%, and 30% of the set pressure for liquid, and vapor services, respectively. In general, pilot-operated pressure relief valves are not affected by built-up back pressure until the total


1200-47

November 2001

1200 Relief Systems


back pressure causes a transition from critical to sub-critical flow conditions, which occurs when the total back pressure exceeds roughly 50% of the relief pressure. The sizing of relief device outlet lines is generally subjected to two separate types of back pressure development evaluations. First, the built-up back pressure is calculated for each pressure relief valve, for each contingency in which that valve alone is flowing. In this calculation, the valve is assumed to be flowing at its full rated capacity of the fluid flowing in each contingency (i.e., the stamped capacity, adjusted to the properties of the flowing fluid). The back pressure developed must not exceed the acceptable value for the type of valve to be installed; if the calculation predicts an unacceptably high back pressure, the discharge line should be modified or a more tolerant valve should be selected. For valves discharging directly to atmosphere, this is the only outlet-line hydraulic performance calculation required. For valves discharging into a common collection system, however, an additional back pressure development evaluation is carried out. As discussed in more detail in Section 1250, the back pressure developed at each flowing pressure relief valve is calculated for each relief contingency. In this calculation, per the recommendation of API Recommended Practice 521, each relieving device is assumed to be flowing at the capacity required to prevent more than the allowable overpressure in the equipment it protects in the contingency under consideration. That is, the required relief flow rates, as discussed in Section 1222, are used in calculating the back pressures developed during simultaneous release from relief valves opening under the same contingency. The acceptance criteria for these back pressures are the same as those for the analysis in which one valve at a time is assumed to be flowing. As in the evaluation of inlet pressure drop, standard piping flow equations are used to calculate the pressure drop through the discharge line, with each piping segment and each fitting accounted for by a flow resistance factor or an equivalent length of pipe. Unlike the inlet line loss calculation, however, the outlet pressure drop calculation should include all contributions to the pressure drop. Sonic limitations and kinetic energy effects can significantly increase the calculated pressure drops. Potential energy effects can be particularly significant for liquid and two phase releases. 2.

Outlet Stresses The discharge of a pressure relief valve can generate forces resulting in substantial stresses on the outlet piping. The mechanical and thermal sources of stress should be evaluated in a manner consistent with the guidelines presented in item 2, “Inlet Line Stresses” on page 1200-44.

3.

Outlet Piping Materials Consideration should be given to the potential for auto-refrigeration of the discharged fluid due to the pressure drop across a pressure relief valve. The

November 2001

1200-48



1200 Relief Systems

flashing of high pressure vapors or light hydrocarbons may require special materials for the low temperatures that may be generated. 4.

Isolation Valves Isolation valves are sometimes installed in a relief valve discharge line for the same reasons they may be installed in its inlet line. Any isolation valve in a discharge line, sub-header, or main relief header should meet the same requirements specified for inlet isolation valves in item 3, “Isolation Valves” on page 1200-44.

5.

General Installation Details The outlet lines from pressure relief valves discharging to closed systems should be free draining to a knockout drum. If this is not practical, accessories such as drain valves or steam tracing should be installed in the discharge line to keep it free of liquid. Accumulation of liquids on the discharge side of a pressure relief valve can cause increased superimposed back pressure and twophase slug flow upon relief. Cases of relief headers being dislodged from the pipe rack due to slug flow have been documented. Similarly, outlet lines from pressure relief valves discharging to atmosphere should also be equipped with a means for keeping the outlet line free of liquids. Typically, a weep hole is drilled at the low point in the discharge line to allow drainage. This weep hole should be of a size adequate to prevent plugging. Also to facilitate proper drainage, connections of pressure relief valve discharge lines into collection headers or sub-headers should be made at the top of the header.

1244 Rupture Disk Installation Inlet and Outlet Piping The hydraulic performance requirements of pressure relief valve inlet and outlet lines are rooted primarily in the opening and closing behavior of the various types of valves. Since rupture disks operate fundamentally differently from pressure relief valves, their inlet and outlet lines do not have to meet the same hydraulic performance requirements. Rupture disk connecting lines need only to be only be large enough to allow the entire rupture disk/piping system to have a flow capacity at least as large as the required relief load. Accordingly, rupture disk inlet and outlet lines are sized in the same process as sizing the disk itself, as described in Item 2, “Rupture Disk Sizing Using Pipe Flow Equations (KR Method)” on page 1200-42.

Inlet and Outlet Stresses Rupture disk inlet and outlet piping is subject to the same mechanical and thermal stresses identified for pressure relief valves. Again, the complexity of the system generally will indicate whether a pipe stress analysis is required.


1200-49

November 2001

1200 Relief Systems


Isolation Valves Rupture disk isolation valves are subject to the same considerations and requirements as those stated in Item 3, “Isolation Valves” on page 1200-44.

Location of Rupture Disk Particularly when a rupture disk is selected because of its quick-opening characteristics, the disk should be installed as close to the protected equipment as practical, with no intervening elbows, etc. While an intervening isolation valve may be installed, the disk should be in a “line of sight” of the equipment’s contents. Conversely, in applications in which routine pressure surges occur, the rupture disk should be placed in a location that minimizes the severity of the pressure surges. The pressure cycling accompanying such surges can contribute to premature fatigue failure of metal rupture disks.

General Installation Details The burst pressure of most rupture disk designs is dependent on the orientation of the disk - there is definitely an intended inlet and outlet side of the disk. For this reason, it is imperative that rupture disks be installed in the proper orientation. Some types of rupture disk can burst at a pressure several times the stamped burst pressure when installed “upside down.” Disk manufacturers have devised various methods to try to make the orientation of rupture disk installation more fail-safe.

1245 Relieving Thermal Expansion of Liquids in Piping This section helps you determine when it is necessary to provide relief of pressure due to thermal expansion of liquids in piping as a result of temperature rises from external sources of heat not including fire. Some recommendations for means of providing such relief are included. A detailed analysis of some thermal relief applications, including charts and sample calculations is also given.

General Consideration of thermal relief is necessary in all sections of liquid piping, regardless of length, when it is reasonable to expect that the liquid will be blocked-in while the line is subject to temperature rises from solar radiation, warm ambient air, steam tracing, or other external sources of heat. A temperature increase will cause both the liquid inside a pipe and the pipe itself to expand in volume. Liquids have high thermal coefficients of expansion compared to metals. For example, oil will expand approximately 25 times as much as the pipe. Therefore, high pressures will build up when liquids are heated in a line sealed by block valves or blinds. Neither thermal expansion of the pipe, expansion of the pipe from internal pressure, nor compressibility of the liquid may be sufficient to relieve the liquid thermal expansion before pressures exceeding the maximum safe pressure of piping components are reached. Tests by the Manufacturing Department at El Segundo Refinery and by Crane Company, and verified by calculations, show that the pressure from thermal expansion of liquid hydrocarbons will increase about 70 to 100 psi for each °F temperature increase.

November 2001

1200-50



1200 Relief Systems

The length of line has no effect on the pressure that will result from thermal expansion of liquid in a blocked line. However, the volume of fluid that must be released to prevent excess pressure build-up will be directly proportional to the line length. Calculations show that temperatures of 150°F can be reached in small lines (10 inches and less) containing liquid hydrocarbons before heat lost by convection equals heat gained from solar radiation. Therefore, if oil is initially at 50°F, a temperature rise of 100°F is possible during extreme exposure to the sun. Tests performed at El Segundo verified that even on a normal day the temperature of oil in a line can increase by 50°F. The possible temperature increase from solar radiation is sufficient to raise the pressure in lines containing liquids by as much as 3,500 to 10,000 psi. See Figures 1200-19 through 1200-22. Such pressures may be considerably above the maximum allowable working pressure of valves and pipe, particularly if the pressure of the liquid, at time of blocking, could be at or near the maximum working pressure of the system. The principal reason more ruptures have not occurred in lines without relief valves has been that sufficient relief is usually afforded by inherent leakage of common valves. With the increasing use of positive shutoff valves, such as plug cocks and ball valves, double-block valves, and some kinds of line blinds, there is a greater likelihood of rupture. Fig. 1200-19 Pressure vs. Temperature Increase for Confined Liquids


Fig. 1200-20 Increase in Temperature of Pipe Containing Hydrocarbon Liquid by Solar Radiation

1200-51

November 2001

1200 Relief Systems


Fig. 1200-21 Coefficient of Thermal Expansion vs. API° for Hydrocarbons

Fig. 1200-22 Approximate Maximum Rate of Temperature Rise Due to Solar Heat for Pipe Containing Hydrocarbon Liquid (°F/hour)

Code Requirements The ANSI/ASME B31.1 Power Piping Code states: “Fluid Expansion Effects. Where the expansion of a fluid may increase the pressure, the piping system shall be designed to withstand the increased pressure or provision shall be made to relieve the excess pressure.”

The ANSI/ASME B31.4 Liquid Transportation Systems Piping Code and the ANSI/ASME B31.3 Chemical Plant and Petroleum Refinery Piping Code state: “Fluid Expansion Effects. Provision shall be made in the design either to withstand or to relieve increased pressure caused by the heating of static fluid in a piping component.”

No further details are given; it is left to the piping designer to ensure that thermal expansion effects are accommodated. The California Unfired Pressure Vessel Safety Orders require that every LPG or NH3 pipe line or hose that can be isolated by two or more stop valves shall have a safety relief valve installed in the pipe line or hose to prevent excessive pressure buildup. The safety relief valve required by this standard shall start to discharge at not less that 312 psi for LPG (300 psi for NH3), nor more than 400 psi, unless the system is designed for higher pressures and provided with safety devices that will

November 2001

1200-52



1200 Relief Systems

adequately protect the system. See Article 5 L.P. Gas System, Order 480 (h) and Article 6, Anhydrous NH3, Order 506 (g).

Methods of Preventing Pressure Build-up Relief Valves. Relief valves are the generally preferred method of preventing pressure buildup. Relief valves are discussed below in “When Relief Valves Are Needed.” Insulation of Lines. Several inches of insulation are needed to sufficiently reduce the heat flow to prevent significant temperature rise. The cost of insulation is the disadvantage of this method. Gas or Vapor Pressure Chambers. The disadvantages of gas or vapor pressure chambers are: (1) higher cost than relief valves, (2) possible contamination if fluid in line is changed before the pressure chamber is flushed out, and (3) gum formation in some gasolines upon contact with air. However, where maintenance of relief valves is a problem, as in some acid or chemical lines, properly designed chambers may offer advantages. Partial Draining of Lines Before Blocking. This method is subject to human error. Drilled Holes in Valves. This method is not acceptable because of uncontrolled flow through the hole at times other than when relief is necessary. Slight “Cracking” of Valves. This method is subject to human error and allows contamination if fluids separated by the valve are different.

When Relief Valves Are Needed Overpressure can occur with only a small rise in temperature. Thus all sections of lines that can be blocked off theoretically require provision for thermal relief. However, experience has shown that the probability of failure does not always justify installing a relief valve. As pressure is built up in a line, some types of valves may leak. Or conditions may be such that there is no possibility that the temperature of liquid in a blocked-off line can rise. Accordingly, there are services where provisions for relief never need be made, such as for cold water lines inside buildings. Because of environmental and safety concerns, it is no longer acceptable to assume that flanges will leak to relieve thermal expansion pressures. Any relief design must assume that relief effluent is contained within the piping.

Conditions Where Relief Is Never Needed 1.

Relief is not required for on-plot lines or for lines to plant rundown tankage where line length does not exceed 400 feet. Relief valves are provided on plant equipment, and it is assumed that all on-plot and rundown lines will be open to equipment that is relieved.

2.

Relief is not required in off-plot manifolds confined by valves of a type which can be expected to leak, such as typical solid-wedge gate valves, if there is less


1200-53

November 2001

1200 Relief Systems


than 20 feet of pipe per valve. (If the amount of pipe per valve is between 20 and 200 feet, use the “Doubtful Case” approach given below). 3.

Relief is not required in areas where no source of external heat energy is available, including most underground lines.

4.

Relief would not be needed if a line is always hot and if it can be blocked only when hot. The chances are slight that it can be heated further by solar radiation. For this reason, relief valves need not be provided in liquid hydrocarbon or water lines which can only be blocked at temperatures above 150°F. A line which is hot when blocked and then cooled will be subjected to an internal vacuum. Since standard wall pipes up to NPS 30 can withstand vacuum of 14.7 psi, and since the possibility of pulling this much vacuum is slight, there is no need for vacuum relief on standard wall pipe lines in the sun. (The assumption that a line can be blocked only when hot is not completely accurate.)

5.

Relief is not needed on lines insulated for heat retention where the source of the hot liquid has thermal relief provision.

Conditions Where Relief Is Always Needed 1.

Relief should always be provided if there is more than 200 feet of pipe per valve even though the valves are of a type that may be expected to leak, such as typical solid-wedge gate valves. If the amount of pipe per valve is between 20 and 200 feet, use the “Doubtful Case” approach.

2.

Relief should always be provided for lines confined by plug cocks, ball valves, double-block valves, or other valves which are not expected to leak, or by some kinds of line-blinds.

Note In blinding a line with a spectacle blind or with some styles of line-blinds, the line may be at least partially drained during the blinding operation. In such cases, thermal expansion of the liquid within the pipe will not cause any substantial pressure rise and no other provision for relief need be made.

November 2001

3.

Relief should always be provided for lines which are steam or electrically traced, except those that fit in the category described in item 4 above.

4.

Relief should always be provided for cold side piping of heat exchange equipment (and for warm side if it normally operates below ambient temperatures) unless pressure relief has otherwise been provided to satisfy requirements of the Pressure Vessel Code. These relief valves must be located between the heat exchanger and the block valve, preferably at the outlet.

5.

Relief should be provided for double seated valves of a type in which pressures exceeding the safe pressure of the valve can build-up in the cavity and blow the packing or damage the valve. (Note: Some styles of double block valves have spring loaded seats or other built-in provision for relieving such excess pressures.)

1200-54



1200 Relief Systems

Considerations for Determining Need in Doubtful Cases Between the above extremes, there are many services where the need for pressure relief is not certain. In such cases, the problem should be approached by calculating the amount of liquid that must be relieved and then deciding if that much leakage can be accommodated. The effect of some variables can be calculated; others cannot. The approximate quantity of liquid to be relieved can be calculated using the calculable variables derived at the end of this section. Generally one should begin with the simple equations that do not consider expansion of the pipe nor the compressibility of the liquid, thereby giving a conservative answer. More refined calculations can be carried out later. Having determined the quantity of liquid that must be relieved, consideration must then be given to the number and type of valves, and the amount of leakage through each (the incalculable variables), and other sources of possible relief. Personal judgment must also be used when determining whether or not a relief valve is needed. It is difficult and perhaps dangerous to estimate and rely on leakage through valves. More and more valves in use today are high quality, tight shutoff valves that will not allow enough leakage to relieve thermal pressures (such as ball valves, Orbit valves, General twin-seal type valves and expanding-gate valves).

Solar Heat—Effects and Considerations The amount of solar energy a pipe line may absorb and the effect on the pressure build-up from expansion will be essentially the same at all geographic locations between 60°N and 60°S latitudes. Directional orientation between latitudes 60°N and 60°S does have an effect on the total amount of heat that can be absorbed by a pipe line. A North-South line will absorb more heat than an East-West line, but the maximum rate of heat absorption (at noon) is the same for both. Wind affects the heat lost by a pipe from convection currents. Wind of 10 mph will double the convection heat losses from bare pipe in still air. Since still air is a possibility at all locations, the theoretical maximum temperatures attainable from solar heating have been calculated assuming natural-convection heat losses without any wind. (See Section 900 of the Fluid Flow Manual for wind heat transfer coefficients.) Paint and pipe coating will affect the amount of heat a pipe line can absorb from solar radiations. Most light colored paints will reflect 20 to 30% of the solar energy, but only on freshly painted surfaces. Since a line could be painted any color after its initial installation, the theoretical maximum temperatures attainable have been calculated assuming flat-black painted lines with an absorptivity of 1.0.

Disposal of Effluent from Relief Valves You must anticipate that all relief valves may leak and consider this possibility when choosing a method for disposing of effluent. For example, it is highly undesirable to connect discharges from relief valves on finished product piping to a


1200-55

November 2001

1200 Relief Systems


common closed header connected to a tank or other location which could result in a back pressure on the system. Under such circumstances, discharge from, or leakage past, a relief valve on one line could leak backward through a relief valve on another line at lower pressure than the intended disposal location and contaminate product in the other line. The quantity of effluent expected during thermal relief, as well as the possibility of continual leakage, must also be considered when choosing a disposal method. Methods that have been used for disposal include the following. Around Block Valve. The relief valve releases pressure around the block valve. This is feasible only if the fluid is the same on each side of the block valve so no contamination results, or if contamination is permissible. This is a common disposal practice for pressure relief at tank block valves. In other piping, valve back-pressure problems will probably rule out this method. An internally vented relief valve set for high back-pressure (low spring setting) could open on a low back-pressure condition, tending to nullify the intent of the block valve. To a Sump or Sewer. This is a method to use when discharge around the block valve is impractical and where discharge to ground would create a hazard or be uneconomical. Collection in a sump with a pump to return accumulated effluent to the plant will be preferable to disposal into a sewer if lines are long and large quantities of effluent can be expected. If a common line is used to carry effluent from several relief valves to a sump or sewer, the system should be designed to prevent back-flow and possible contamination of the contents of different lines. Use of an open funnel at the point of discharge from each relief valve into an open gravity drain line has been found to be advantageous. To Atmosphere or to Ground. This method should only be used when the quantity of effluent or leakage is small and when the release does not create a hazard or an undesirable condition. Relieved liquid should be directed away from piping or other equipment. In Cascade. Cascaded relief valves are not recommended. If used, cascaded valves require careful consideration both from a safety and operating viewpoint, making due allowance for back-pressure effects.

Types of Relief Valves The construction of relief valves must meet local area requirements. Valves are commonly selected on the following basis:

November 2001

•

Use steel, bolted bonnet, thermal relief valves (Crosby Series 900, Teledyne Farris 2740 or 2741, or equal) with flanged or seal welded screwed inlet and outlet connections for all LPG liquid services. It should be noted that seal welding of screwed inlet and outlet connections presents a problem when it is necessary to take the relief valves into the shop for testing and setting. For this reason, flanged ends are preferred (see Figure 1200-23).

•

Use the less expensive steel, screwed bonnet, screwed end, thermal relief valves (Crosby Series 900 or equal) for all other liquid process services except in corrosive liquid services or in piping subject to vibration. In these services, use the bolted bonnet style above. The bolted bonnet style may also be prefer-

1200-56



1200 Relief Systems

Fig. 1200-23 Flanged Relief Valve (Series 900, Model 9511019A) (Courtesy of the Crosby Valve & Gage Company)

able in severely corrosive atmospheres where external corrosion could make disassembly of the screwed bonnet difficult •

Use screwed, bronze relief valves (or equal) for water services

Size of Relief Valves The smallest practical size relief valves available have adequate capacity for the majority of piping applications; however, valves having a ¾-inch inlet are considered the minimum for adequate mechanical strength. When relieving capacities are in doubt, relief valve size can be most conveniently selected by using slide rules or sizing charts published by valve manufacturers. Use an accumulation of 10% for determining the capacity of these relief valves when fully open. The ¾-inch inlet, 0.74 in2 orifice Crosby relief valves referred to above will provide liquid relief for several miles of large pipe. They are the smallest, reasonably inexpensive relief valves generally acceptable for Company use. They will relieve 29 GPM when the relief pressure is 175 psig, and 38 GPM when the relief pressure is 300 psig (assumed 10% accumulation and specific gravity of stock as 0.80). When vapor can be generated in substantial quantities as would be the case if a high vapor pressure liquid, such as LPG or NH3, were to be blocked in on the cold side of an exchanger with steam on the hot side, a much larger safety relief valve will be required.

Setting of Relief Valves Relief valves should be set at the maximum allowable working pressure of the weakest component of the blocked-off line with allowance for static head.


1200-57

November 2001

1200 Relief Systems


Calculation of Solar Heat Gains The solar radiation coefficient is dependent on the projected area of the pipe, which varies with latitude and orientation. See Section 200 of the Evaporation Prevention Manual for solar heat tables. We assume an absorptivity of 1.0 Qs is calculated over the projected area. Q Btu ------s = 300 -----------A ft 2 hr (Eq. 1200-1)

The convection heat loss coefficient is given in Perry’s Chemical Engineers’ Handbook (page 474). Qconv is calculated over the total outside area. Q conv 1- 0.25 ( ∆T′ ) 1.25 ----------Btu-------------- = 0.5  -----  A Do ft 2 hr (Eq. 1200-2)

where: Qs = radiation heat gain, Btu/hr Qconv = convection heat loss, Btu/hr ∆Τ′ =

temperature above ambient, °F

Do = outside diameter, in. D = outside diameter, ft A = πDL L = length, ft The radiant heat loss coefficient is given on page 486. Qrad is calculated over onehalf the outside area. Q rad ---------T 1  4  T 2  4 Btu A = 0.1724  -------- – --------- -------------- 100  100 ft 2 hr 2 (Eq. 1200-3)

where: T1 = final pipe temperature, °R T2 = initial pipe temperature, °R

November 2001

1200-58



1200 Relief Systems

Calculation of Maximum Attainable Temperature of Pipe and Liquid The maximum temperature the pipe and liquid can be raised to by incident sunlight is set by the equilibrium of solar heat gain with convection and radiant heat loss. Qin = Qout Qsolar = Qconv + Qrad (Eq. 1200-4)

For 1 foot (L) of NPS 6 inch pipe at an ambient temperature of 100°F, the solar heat gain given by Equation 1200-1 is: (1 ⁄ 2) Q in = 300 -------------L = 300 ( 0.5 ) = 150 But/ft-hr Equation 1200-4 must be solved iteratively to find the maximum attainable temperature. For simplicity, we will make a first guess that is actually the answer: Try ∆T′ = 60°F Q out = Q conv A + Q rad A (Eq. 1200-5)

1 0.25 = 0.5  --- ( 60 ) 1.25 ( 0.5πL )  6 + 0.1724 ( 6.2 4 – 5.6 4 )  0.5πL ---------------  2  = 84 + 66 = 150 Btu/ft-hr Since Qin = Qout = 150 Btu/ft-hr, the maximum ∆T′ for the NPS 6 inch line is 60°F at the 100°F ambient temperature.

Calculation of Temperature Rise per Hour We can calculate the temperature rise per hour by repeated use of Equation 1200-5. Note that we estimate the average ∆T′ for each iteration. Assume the following: Pipe and liquid are at same temperature (this is conservative). Maximum, ambient temperature = 100°F


1200-59

November 2001

1200 Relief Systems


Pipe: NPS 6 inch, Schedule 40 To = 50°F Weight of pipe = 18.97 lb/ft Liquid Gasoline, Sp. Gr. = 0.86 Weight of liquid = 10.8 lb/ft Specific Heat Steel = 0.12 Btu/lb °F Specific Heat Liquid = 0.50 Btu/lb °F Total specific heat = 0.12(18.97) + 0.50(10.8) = 7.68 Btu/ft°F QinSolar = 150 Btu/ft-hr a.

During the first 3 hours most of the incident solar heat is absorbed by the pipe: 3 ( 150 ) ∆T′ = ----------------- = 58.5°F 7.68 (Eq. 1200-6)

T 3 = T o + 58.5 = 108.5°F b.

During the fourth hour the ∆Τ′ becomes great enough to make the Qout significant. Assuming an average temperature of 118°F: Q out = 0.501 ( 18 ) 1.25 + 0.1353 ( 5.78 4 – 5.6 4 ) = 18.5 + 19.1 = 27.6 Btu/ft-hr 150 – 27.6 ∆T′ = ------------------------- = 16°F 7.68 (Eq. 1200-7)

T 4 = T 3 + ∆T′ T 4 = 108.5 + 16 = 124.5°F c.

During the fifth hour: Assume average temperature is 130°F:

November 2001

1200-60



1200 Relief Systems

Q out = 0.501 ( 30 ) 1.25 + 0.1353 ( 5.9 4 – 5.6 4 ) = 35 + 31 = 67 Btu/ft-hr 150 – 67 ∆T′ = --------------------- = 10.9°F 7.68 (Eq. 1200-8)

T 5 = T 4 + ∆T′ T 5 = 135.4°F d.

Similarly, we can calculate the pipe and liquid temperature for each hour as it approaches the maximum of 168°F. See Figure 1200-24.

Fig. 1200-24 Sample Calculation: Pipe and Liquid Temperature for Each Hour, Approaching Maximum of 168°F ∆ ′ °

out

°

Sixth hour

90.1

7.8

143.2 (T6)

Seventh hour

109.4

5.3

148.5 (T7)

Eighth hour

122.5

3.6

152.1 (T8)

Ninth hour

129.5

2.7

154.8 (T9)

Calculation of Pressure and Volume We can calculate the pressure that will be attained in the pipe and the amount of liquid volume necessary to be released. We define: σc = circumferential stress in pipe, psi σl = longitudinal stress in pipe, psi σl/σc = use 0.5 D = mean diameter, in. L = pipe length, in. V = volume inside pipe, in3 α = bulk thermal coefficient of expansion of liquid, 1/°F (see Figure 1200-21) β = linear coefficient of expansion of steel, 1/°F (use 6.5 × 10-6) K = bulk modulus of elasticity of liquid, psi (use 2.1 × 105 for crudes and 1.5 × 105 for refined products) E = modulus of elasticity of steel, psi (use 30 × 106)


1200-61

November 2001

1200 Relief Systems


∆T = temperature rise of pipe and liquid, °F t = pipe wall thickness, in. υ = Poisson’s ratio for steel (use 0.3) ∆P = pressure rise of liquid, psi Expansion of steel pipe due to pressure: ∆P D σ c = ------------2t (Eq. 1200-9)

∆P D σ 1 = ------------4t (Eq. 1200-10)

σ c D υσ 1 D ∆D = --------- – -------------E E σ  σc D  -  1 – υ -----1- = --------E  σ c (Eq. 1200-11)

πD 2 A = ---------4 (Eq. 1200-12)

π πD ∆A =  --- 2D∆D = -------∆D  4 2 (Eq. 1200-13)

σ1 L σcL ∆L = --------- – υ --------E E σcL  σ1  -  ------ – υ = --------E  σc  (Eq. 1200-14)

V = AL (Eq. 1200-15)

∆V = A∆L + L∆A (Eq. 1200-16)

November 2001

1200-62



1200 Relief Systems

2 D LA ∆L- + 2L πD ----------  ∆ -------- = ----------------4  D L

∆L ∆D = V ------- + 2 -------L D (Eq. 1200-17)

σ  σcL  σ1  2σ c D  ∆V -  ------ – υ + -------------  1 – υ -----1- -------- = --------EL  σ c ED  σ c V  σ = -----c- [ ( 0.5 – 0.3 ) + 2 ( 1 – 0.3 × 0.5 ) ] E σ = 1.90 -----cE 1.90 ∆PD = ---------------------2tE (Eq. 1200-18)

Thermal expansion of steel pipe: ∆V -------- = 3β∆T V (Eq. 1200-19)

Total expansion of steel pipe: ∆V 1.90∆PD -------- = ---------------------- + 3β∆T V 2t E (Eq. 1200-20)

Thermal expansion of liquid: ∆V -------- = α∆T V (Eq. 1200-21)

Compression of liquid due to pressure: ∆ V- = – -----∆P------V K (Eq. 1200-22)

Total expansion of liquid: ∆V ∆P -------- = α∆T – ------V K (Eq. 1200-23)


1200-63

November 2001

1200 Relief Systems


Total liquid pressure increase: ∆V Equate -------- for liquid and pipe V 1.90∆PD ∆P ---------------------- + 3β∆T = α∆T – ------2t E K (Eq. 1200-24)

1.90 ( D ) 1 ∆P  -------------------- + ---- = ( α – 3β )∆T  2t E K (Eq. 1200-25)

( α – 3β )K∆ T ∆P = ---------------------------------------1.9DK ⁄ 2t E + 1 (Eq. 1200-26)

Relief volume: At a set pressure of PR the volume to be released must be ∆P R ∆V -------- =  α∆T – --------- K  V 1.90 ∆P R D - + 3β∆T –  -------------------------  2t E 1.90 DK ∆P R = ( α – 3β )∆T –  1 + --------------------- --------- 2t E  K (Eq. 1200-27)

If we wish to be conservative and neglect the expansion of the pipe (small) and compression of the liquid (moderate) then: ∆P = αK∆T (Eq. 1200-28)

∆V = αV∆T (Eq. 1200-29)

November 2001

1200-64



1200 Relief Systems

1250 Pressure Relief Disposal Systems 1251 Introduction In the previous several sections, we have covered the identification and quantification of pressure relief requirements and the selection, sizing, and installation of pressure relief devices to provide relief flow. This section discusses the issues involved in handling the fluids discharged from these relief devices and from emergency depressuring (“blowdown”) valves. The selection and design of piping and equipment to handle these fluids needs to satisfy a variety of criteria involving both the impact of the fluids on personnel and the environment, and the impact of the disposal system on the performance of the relief devices that feed into it. The relief device discharge fluids should be disposed of safely, economically, with a minimum environmental impact, and without limiting the relief of overpressure within the process equipment.

1252 Choice of Disposal Method The selection of a method for the ultimate disposition of the fluid discharged from a relief device is influenced primarily by the nature of the fluid. Innocuous materials such as water or air can be discharged directly to the air or the ground. While flammable gases and vapors may usually be discharged to the atmosphere safely, the environmental impact of such discharge may be unacceptable. Flammable liquids and all toxic substances usually require contained collection, separation, and disposal systems. The following sections briefly describe the options available for the intermediate and final disposition of relief discharges.

Return to the Process In some cases, the discharge of a relief device may be directed to another point in the same process. For example, relief valves on the discharge of reciprocating pumps often have their outlets routed back to the suction of the pump. Relief valves installed to provide pressure relief for piping segments in the event of thermal expansion of a blocked-in liquid sometimes have their discharges routed to a storage vessel or another piping segment. The advantages of this method of relief device discharge handling are that the process fluid is not transferred to a waste stream, and that the discharge handling “system” is minimal. However, there are some significant limitations on the application of this approach to discharge handling. The relieved material should be similar in composition to that at the location to which it is routed so that product purity is maintained. The normal pressure at the discharge location should be considered when establishing the differential or net set pressure of the device (see Sections 1231 and 1232). The net (or spring) set pressure of a conventional relief valve, or the burst pressure of a rupture disk, plus the normal pressure at the discharge should equal the desired set pressure (see Figure 1200-13). Therefore, a desired discharge location would not be appropriate if the pressure at that location is not known or is


1200-65

November 2001

1200 Relief Systems


subject to significant fluctuations. Use of a balanced (e.g., bellows) relief valve can often alleviate these difficulties, but care should be taken to consider the effect of back pressure on the valve’s flow capacity and that the back pressure rating of the valve (particularly of the bellows) is not exceeded.

Separation of Phases or Components Relief system disposal “processes” usually include some type of separation process to ensure the safe operation of the ultimate disposal methods. To avoid the occurrence of “burning rain”, flare systems always contain a vapor-liquid separator (the “flare knockout drum”) upstream of the flare itself. In other cases, relief disposal systems may contain a counter-current absorption contactor to remove one or more hazardous or valuable components from the relief stream. For example, refinery alkylation units route relief and vent streams acid (HF or sulfuric) through a caustic scrubber to neutralize the acid; the scrubbed vent stream is then sent through the remainder of the collection system to a flare. Flare Gas Recovery. In some cases, environmental or economic considerations may justify installation of equipment to recover noxious or valuable components from a relief stream that might otherwise simply be flared. These are typically sized to recover such components only from relatively routine venting activity, and should therefore be designed to direct the relief stream directly to the flare in emergency situations. See API Recommended Practice 521 for further discussion of these systems.

Atmospheric Discharge Discharge of a relief stream more or less directly to the atmosphere is usually the simplest and least costly disposal option available. However, care is required to ensure that atmospheric discharge is compatible with personnel safety, environmental safety, and good public relations. The criteria for evaluating the safety issues concerning relief device discharge to atmosphere are well established. 1.

Safety Considerations Non-hazardous Materials Non-hazardous fluids may be safely discharged either to the atmosphere or to grade. Included in this class of materials are air, steam, and water, provided that they are discharged in a location and direction that will prevent contact of the relief stream with personnel. Flammable and Combustible Materials Flammable and combustible liquids generally cannot safely be discharged to the atmosphere. Their density virtually guarantees that they will fall to grade level, where they will present a hazard to equipment and personnel. They should be routed to separation and collection equipment for recycle or other ultimate disposition.

November 2001

1200-66



1200 Relief Systems

Flammable vapors are discharged from relief devices at very high velocities. The jet of material exiting the tailpipe therefore rapidly entrains large amounts of air, quickly reducing the concentration of flammable components in the jet below the levels at which combustion can be supported. In the absence of wind, the portion of the discharged “cloud” whose fuel concentration is between the lower and upper flammable limits forms a well-defined, narrow cylinder whose length and width depend on the exit diameter, exit velocity, and composition of the relief stream. In the highest wind speeds, this jet may be blown entirely horizontal. However, in the absence of downdrafts, even for vapors of high specific gravity (e.g., molecular weight two or three times that of air), the flammable portion of the vapor jet never extends below the level of discharge. As long as no ignition sources exist above the discharge elevation, therefore, there is very little chance of igniting flammable vapors discharged from a relief device to atmosphere. This analysis is based on the discussion and methods outlined in API Recommended Practice 521, and backed up by vapor cloud dispersion calculations and by decades of operating experience. The analysis is also predicated on the absence of liquids in the discharged stream. When selecting atmospheric discharge for a relief device handling flammable substances, care should be used to ensure that liquid will not be present in the discharge. Possible mechanisms for liquid leaving a relief device that normally has only gases or vapors at its inlet include: overfilling of the equipment item with liquid, liquid carryover due to two-phase flow of viscous or foamy substances, and liquefaction of auto-refrigerating relief streams. Toxic Materials The considerations for relief streams that may contain toxic substances are similar to those for flammable streams. Streams with potentially toxic liquids should be contained in a closed disposal system for separation, possible treatment, and disposal. The discussion of the jet nature of a relief device discharge and the resultant dilution of the discharge with air holds equally well for gasphase streams containing toxic substances. However, the concentrations below which exposure to toxic materials can be tolerated are typically a great deal lower than those below which flammable substances will not support combustion. The lower flammable limit of most hydrocarbons, for example, is in the range of one to three percent by volume. The lower limits for acute (i.e., shortterm exposure) toxic effects vary considerably from one compound to another, but in some cases can be in the range of ten parts per million (ppm) by volume. The threshold concentrations at which longer exposures may produce undesirable effects can be even smaller, but because relief discharges are normally of short duration, these lower threshold concentrations are not usually of concern. Any decision to discharge potentially toxic gases or vapors to atmosphere should be based on modeling of the atmospheric dispersion of the discharge using carefully chosen, generally conservative modeling techniques and assumptions. The modeling results would be expected to indicate that at locations to which personnel access is possible, the predicted concentrations of toxic substances would not reach unacceptable levels.


1200-67

November 2001

1200 Relief Systems

2.


Environmental Considerations In this context, “environmental considerations” refers to the long-term, far-field effects of discharging relief streams to the atmosphere. This is in contrast to the “safety” concerns discussed above, which are restricted to the immediate impact of such releases, typically for worker safety and property protection. Environmental considerations tend to be more restrictive than safety concerns. For example, while atmospheric discharge of relief valves in gas plants and compressor stations may present no short-term hazards to worker or public safety, the cumulative effect of occasional discharges and leaks may not meet regulatory targets for VOC emissions. Engineers and managers should be informed about local environmental protection regulations and corporate and local management policies and targets concerning discharge of hydrocarbons and other compounds to the environment

3.

Nuisance Considerations Relief devices discharging to atmosphere typically generate a good deal more audible noise than do those discharging into a collection system. In addition, many compounds can be detected by their odor at concentrations significantly below those at which they pose any hazard to life or health. Consideration generally should be given to these facts and their impact on an operating facility’s neighbors when choosing between discharge to the atmosphere and routing to a collection and disposal system.

Flaring After a relief stream has undergone phase separation and, possibly, capture of valuable or hazardous components in an absorption or recovery system, the most commonly used alternative to discharge to the atmosphere is combustion in some sort of flare. While a flare gas recovery system may be appropriate for collecting and reusing portions of routine venting and small periodic relief device discharges, for disposal of large flows of flammable gases and vapors in emergency situations, there is little alternative to a flare system. 1.

Flare Combustion Noise and Light Combustion noise is approximately proportional to the square of the mixing velocity. Steam-air injection produces high frequency noise and also raises the intensity of the low frequency combustion roar. Most current flare tips use multiport steam nozzles to reduce flare noise to some extent. While it is operationally convenient to keep excess steam on the flare, it is undesirable for noise production and energy conservation. Without steam-air injection, buoyant forces aspirate air into the combustion zone and the laminar flame tends to be noiseless, billowy, and smoky. With steam-air injection, aspiration is thrust controlled, turbulence is greater, mixing is better, and the flame is smaller, stiffer, less radiant, and smokeless. Unfortunately, these conditions produce the combustion roar that is typical of thrust controlled flames.

November 2001

1200-68



1200 Relief Systems

The expected noise levels from modern flares with Coanda or multiport nozzles are shown in Figure 1200-25. Combustion roar dominates, particularly at higher relief rates. The expected sound level can be scaled by adding or subtracting 6 dB if the distance is halved or doubled. Fig. 1200-25 Expected Noise Levels for Elevated Flares with Multiport or Coanda Nozzle Steam Injection (Courtesy of Flaregas Corp.)

At high relief rates, high frequency noise is not what really disturbs the neighbors. Rather, they are bothered by the inherent rumble, vibration, and illumination. There is no technological remedy. In an emergency, an elevated flare may disturb the surrounding neighborhood. It is a poor idea to use an elevated flare to burn off small or frequently vented streams. Alternate means of controlling the pressure of the process should be considered. Any significant use of an elevated flare, particularly at night, may be objectionable to nearby residents. 2.

Ground Flares Small relief streams can be burned unobtrusively in a ground flare. The Company now has ground flares at most of its refineries. The ground flare is basically a refractory-lined, multi-burner process furnace without tubes. The stack may be as high as 100 feet. This discussion assumes a low pressure ground flare, although high pressure ground flares also exist. The capacity of the ground flare is limited: 25,000 and 50,000 Btu/hour are common sizes. Normally, the ground flare is sized for 5 to 10% of the total capacity of the relief system. The ground flare is included to take care of the great majority of the relief incidents, about 95%, that are relatively small. Ground flares are used to reduce the noise and light of elevated flares. Enclosing the turbulent burning zone eliminates the light, and employing multiple burners greatly reduces the noise. A typical ground flare is at least 15 decibels quieter than an elevated flare.


1200-69

November 2001

1200 Relief Systems


The elevated flare is reserved for emergencies. The ground flare is not an alternative to an elevated flare. It is an expensive added feature for good public relations. The diversion water seal preferentially routes relief gas to the ground flare up to its maximum capacity, where it is burned quietly, smokelessly, and without visible flame. When the flow exceeds the capacity of the ground flare, the diversion seal automatically diverts the excess relief gas to the elevated flare. See Figure 1200-39. The relief system is designed so that the ground flare can be shut down for maintenance. All relief streams are bypassed to the elevated flare, and there is no interruption in plant operation. The ground flare is sized to minimize operation of the elevated flare. This requires a review of the controllable venting activities of the plants and their flow rate and frequency. This review is used to select the capacity of the ground flare from the maker’s standard capacities. Generally, the Company selects either 25,000 or 50,000 Btu/hour capacities. Most of the Company’s experience has been with two main kinds of ground flares: cylindrical (John Zink), and rectangular (Flaregas). Figure 1200-26 shows a cylindrical ground flare. Figure 1200-27 shows a rectangular ground flare. Fig. 1200-26 Cylindrical Ground Flare

Fig. 1200-27 Rectangular Ground Flare

Ground flares are not incinerators. They lack the residence time and mixing thoroughness to handle sour gas streams. Sour gas streams must be segregated and not burned in ground flares or there will be odor problems.

November 2001

1200-70



3.

1200 Relief Systems

Energy Conservation The best way to “recover” relief gas is to not send it to the flare in the first place. Run the plant to minimize venting and PSV operation. Source control is the first and most important step. Leaking PSVs, a common source of normal venting, can be detected by ultrasonic probes (such as the UE Systems Inc., “Trouble Shooter”) or infrared imaging devices (such as the “Probe Eye”). Many refineries have installed flare gas recovery systems that have given good paybacks. The systems are quite simple. A compressor takes suction on the relief header and returns the compressed gas to the process or to the fuel system. A flow history of the relief header is needed to properly size the compressors. The relief gas recovery system includes a pressure controlled purge gas bleed line working against the water seal to maintain a minimum pressure in the relief header at all times to prevent oxygen entry. This avoids the need for upstream gas purges.

4.

Purge Gas Conservation API RP 521 recommends a fuel gas purge stream at the upstream end of the relief header and at each major branch. This purge gas maintains a hydrocarbon rich atmosphere in the relief header, the knockout drums, and the flare stack. In a large relief system with a ground flare, purge gas usage of 50 to 100 EFO/day is not unusual (EFO stands for equivalent fuel oil units measured in barrels). The purge injection station usually includes a rotameter with a globe valve bypass, block valves, and a blind. The injection rate is usually controlled by a fixed orifice to insure a constant supply, even with instrument malfunction. Often the purge gas is simply injected in addition to a “normal” venting flow that is sufficient to protect the relief system. Considering the value of the purge gas, it makes economic sense to put purge injection on pressure control. Holding the relief header on pressure control against the water seal permits shutting off the upstream restriction-orifice injection points. This pressure controlled purge does not replace the purge gas injection above the water seal that protects the flare stack. A gas seal just below the combustion tip reduces the purge requirement from 20,000 scfh to 2,000 scfh. Virtually all the Company stacks have gas seals.

5.

Steam Conservation Steam is injected to the combustion tip to suppress smoke. Once the flame is smokeless, excess steam makes more noise, wastes energy, and may force the combustion back down the tip. The amount of steam required depends on relief gas discharge rate, gas composition, and wind speed; therefore excess steam is normal. The most effective method of reducing steam consumption is to control


1200-71

November 2001

1200 Relief Systems


the rate based on flame quality or on relief gas flow and composition. Flares in producing areas often use natural gas instead of steam for smoke suppression.

Flares at Producing Sites Producing site flares for offshore and onshore are also designed in accordance with API RP 521. Sections 1253 through 1256 discuss design concepts for both locations. Additional factors must be considered in the design of offshore flares. 1.

Offshore Flare Location The offshore flare often must be located adjacent to a manned production platform. API RP 521 recommends that the flare be located so that maximum radiation level does not exceed 500 Btu/hr/ft2 (excluding solar radiation) for areas where personnel are working continuously. This recommendation is consistent with experience.

2.

Smokeless Burning Most offshore platforms do not have steam for flare operation. Therefore, they use flares that do not require additional turbulence generation or that use an air blower or a water spray from a pumping unit. Air blower flares have a greater turndown than water spray flares.

3.

Continuous Flaring It may be necessary to continuously flare tail gas or excess produced gas which cannot be economically utilized. It may also be necessary to flare gases vented during process upsets until corrective action can be taken.

4.

Offshore Flare Designs Offshore platforms typically use the following kinds of flares: –

–

November 2001

Nonsmokeless utility flares, which are used for continuous flaring of natural gas with a molecular weight of 20 or less, or for short-term emergency flaring of paraffinic hydrocarbon vapors with a molecular weight greater than 20. These flares incorporate a proprietary flame retention device to provide a more stable flame which helps to prevent liftoff (lifting of the flame from the top of the flare) or blowoff (complete extinguishment). They also include a flare windshield which prevents the flame from licking down the outside of the nozzle. See Figure 1200-28. Smokeless air blower flares, which are used for continuously flaring paraffinic hydrocarbon vapors with molecular weights greater than 20. These also incorporate a flame retention device and a windshield, but the configuration is quite different from the nonsmokeless flare. Very little forced air is required for smokeless burning. Secondary air is entrained by the flame itself. The flare has a flow sensor to start the air blower when needed. See Figure 1200-29.

1200-72



1200 Relief Systems

Fig. 1200-28 Nonsmokeless Utility Flare (Courtesy of NAO, Inc.)

Fig. 1200-29 Smokeless Air Blower Flare (Courtesy of Kaldair, Inc.)


1200-73

November 2001

1200 Relief Systems


–

Multipoint flares, which are used where smokeless flaring, low radiation, and a large turndown are required. The tip requires no seals and no assist gas or blown air. These flares require high relief gas pressures—5 to 75 psig.

–

Coanda-profile flares, which are used where smokeless flaring and low radiation are required. These flares entrain high volumes of air and create a stiff flame which resists wind effects. They require high relief gas pressures—5 to 75 psig. See Figure 1200-30.

Fig. 1200-30 Coanda-Profile Flare (Courtesy of Kaldair, Inc.)

Offshore Flare Support Structures Selection of the flare structure depends on the distance between the flare and the production platform, which in turn depends on the relief gas quantity, heating value, toxicity, and whether the flaring is intermittent or continuous. The main kinds of support structures are:

November 2001

•

Flare booms, which extend from the edge of the platform at an angle of 15 to 45 degrees. They are usually 100 to 200 feet long. Sometimes two booms oriented 180 degrees from each other are used to take advantage of prevailing winds. See Figure 1200-31.

•

Derrick-supported flares, which are located on a derrick above the production platform. They are used when space is limited and relief quantities are moderate. The disadvantages are: a possible crude oil spill onto the platform,

1200-74



1200 Relief Systems

Fig. 1200-31 Flare Boom

interference with helicopter landing, and higher radiation intensities. See Figure 1200-32. Fig. 1200-32 Derrick-Supported Flares

•

Bridge-supported flares, which are on a separate platform connected to the main platform by a bridge as much as 600 feet long. Bridge supports are usually spaced about every 350 feet. See Figure 1200-33.


1200-75

November 2001

1200 Relief Systems


Fig. 1200-33 Bridge-Supported Flare

•

Remote flares are on a separate platform connected to the main platform by a subsea relief line. The main disadvantage to these flares is that any liquid carryover or subsea condensation will collect in pockets in the connecting line, so that the line acts like a liquid trap. See Figure 1200-34.

Fig. 1200-34 Remote Flare with Subsea Relief Line

November 2001

1200-76



1200 Relief Systems

•

Ground flares, which are often used on tanker based production platforms to reduce noise and to reduce the radiation levels on deck.

•

Floating flares, which are supported on a barge. These are mainly used in emergency or short-term situations.

1253 Disposal System Design Basis Whenever multiple relief devices protecting multiple equipment items discharge into a common collection system, the impact of the collection system on the generation of built-up back pressure at the individual relief devices must be analyzed. Note that the back pressure evaluation described in “Outlet Piping” on page 1200-47 involves only the flow from one device, and only from the device outlet to the discharge line’s entrance into a large common header. The additional sources of flow introduced by the common collection system create the possibility of pressures higher than atmospheric where an individual discharge line empties into a collection header. Therefore, the analysis of the impact of a relief discharge collection system on the proper performance of the various relief devices connected to it consists of a detailed evaluation of the cause of overpressure. A “common contingency” or “simultaneous release” is one in which a single event can result in discharge to the collection system from multiple sources. This analysis should be performed whether the ultimate discharge location is the process, the atmosphere, or a flare. In addition, common contingencies that result in large flows from single devices may require consideration on a case-by-case basis.

Simultaneous Releases Caused by a Common Contingency Individual relief valves are sized for the highest relieving rate required by any of the applicable causes of overpressure for the equipment being protected. The relief system branch and main headers and the flare tip are sized for the worst case simultaneous relief rate caused by a single relieving contingency. The most common overpressure contingencies causing simultaneous releases from a number of relief valves are: fire, power failure, cooling water failure and utility failure. In addition to releases from relief valves there may be reliefs from pressure control valves and automatic de-pressuring systems. Sizing of the total relief system has to take into consideration these additional loads as well. 1.

External Fire A large external fire can cause multiple discharges to the disposal system due to vaporization of liquids within the equipment heated by the fire. Determination of Credibility As the source for an external fire of significant extent is generally a flammable liquid, processing units that contain only gaseous materials or only non-flammable liquids may not require consideration of the external fire contingency.


1200-77

November 2001

1200 Relief Systems


Determination of Impact on the Collection and Disposal System Consideration should be given to the source and location of flammable substances, the provision for liquid drainage, and the effects of natural and man-made barriers. In the absence of special drainage or barrier considerations, API Recommended Practice 521 recommends assuming that a fire will cover a surface area of between 2,500 and 4,000 ft2, typically in a circular shape. All relief devices protecting equipment requiring external fire relief, and located within the assumed fire zone, are assumed to discharge simultaneously into the disposal system. 2.

Total Electric Power Failure A total electrical power failure can affect both process equipment and utilities, resulting in multiple discharges to the relief header. Determination of Credibility A detailed evaluation of all electrical power supplies to the facility should be performed to determine the possibility of a total power failure. Large process facilities typically have multiple electrical power supplies to increase reliability. However, one should consider the possibility that a single event could affect multiple electrical power supplies. Experience indicates that most operating facilities have suffered a total loss of electrical power at some time. Determination of Impact on the Collection and Disposal System All electrically-driven machinery such as pumps, compressors, and air coolers will shutdown. Consequently, other utility systems such as cooling water, instrument air, and shutdown systems may be affected. In many installations, electrically driven machinery has an installed spare unit using a different energy source. Note that API Recommended Practice 521 states that such spares are not considered 100% reliable; therefore, credit should not normally be taken for their operation. However, where two or more units are in parallel, have different energy sources, and normally operate simultaneously, credit may be taken for a partial failure (e.g., the electric pump fails, but the steam driven pump continues to operate). Care is required when taking credit for the continued operation of machinery driven by alternate energy sources, because the power failure may affect auxiliary equipment (e.g., lube oil pumps) that would cause the machinery to shut down. In addition, equipment driven by an alternate energy source may shut down due to consequences of the power failure. For example, the power failure may cause a high level in a compressor suction drum, causing the compressor to shut down. Finally, careful consideration should be given to collateral failures that may mitigate the effects of a total power failure. Typical examples include loss of feed, loss of heat input via forced circulation reboilers, and loss of heat medium circulation.

November 2001

1200-78



3.

1200 Relief Systems

Partial Electric Power Failure A failure of an individual substation, bus, or circuit within the electrical distribution system can affect process equipment and utilities, resulting in multiple releases to the disposal system. In some cases a partial power failure may result in a larger overall load to the disposal system or (more commonly) to a single collection line within the system than a total failure. Determination of Credibility A fault, inadvertent operation of circuit breakers, or a mechanical failure can lead to a localized loss of electrical power within the distribution system. Credit for automatic or manual switching of power supplies is generally not taken when determining the credibility of a failure. Facility electrical personnel should be consulted when evaluating potential modes of failure. Determination of Impact on the Collection and Disposal System The impact on the disposal system is determined in the same manner as described for total power failure. Additional consideration should be given to equipment that continues to be supplied power by unaffected portions of the electrical distribution system.

4.

Cooling Water Failure A loss of cooling water can result in a loss of process cooling and shutdown of rotating equipment, causing multiple discharges to the disposal system. Determination of Credibility A cooling water failure is considered a realistic contingency if a single event could result in the loss of cooling water. In addition to power failure contingencies discussed above, a loss of cooling water may result from a rupture or blockage of a single cooling water supply line. Typically, consideration is given to the failure of only one cooling water circuit at a time. Determination of Impact on the Collection and Disposal System A cooling water failure will result in a loss of cooling to all water-cooled exchangers, such as process coolers, process condensers, and utility coolers. Loss of condensation typically requires relief of the non-condensed vapors. Loss of overhead cooling for distillation columns often requires relief of the column overhead vapors. Loss of condensation in refrigeration loops can result in shutdown of the refrigerant compressors and subsequent loss of refrigerant cooling to the process. The loss of utility cooling can result in shutdown of rotating equipment due to high temperature. Operating experience should be consulted to define the potential for shutdown of rotating equipment due to loss of utility cooling.


1200-79

November 2001

1200 Relief Systems

5.


Instrument Air Failure An instrument air failure may cause the failure of all air-actuated instrumentation. Discharges to the disposal system can result from overpressure of process equipment or depressuring directly to the disposal system. Determination of Credibility Instrument air may be lost due to the shutdown of the compressors used to maintain system pressure, plugging of filters or driers, or due to a line rupture within the system. The presence of multiple air compressors, air receivers, or make-up from other plant air supplies decreases the likelihood of a total loss of instrument air due to compressor shutdown. However, due to the possibility of a line rupture within the system and the difficulty in predicting the impact of such a rupture, consideration is generally given to a total loss of instrument air. Determination of Impact on the Collection and Disposal System The immediate result of an instrument air failure will be failure of air-actuated control and shutdown valves, which will fail closed, in place, or open. Particular consideration should be given to control valves or depressuring valves that fail open and discharge directly to the collection and disposal system. Note that a depressuring or shutdown valve may have an independent air supply (e.g., an air bottle) to prevent failure-mode operation of the valve if air were otherwise lost.

6.

Steam Failure A loss of one or more steam systems within the facility can affect steam driven equipment, resulting in multiple discharges to the disposal system. However, a loss of steam can also result in loss of heat input to the process, which may reduce the impact of equipment failures. Determination of Credibility Steam distribution systems are often fed by multiple sources such as boilers and waste heat steam generators. In addition, multiple steam pressures are generally present, with equipment enabling make-up of steam from higher to lower pressure systems. The potential for a loss of one or more steam pressures should be considered on a case-by-case basis. Determination of Impact on the Collection and Disposal System A steam failure can result in the loss of steam-driven equipment such as large compressors and pumps. The loss of large compressors (e.g., a wet gas compressor in an FCC Unit) can result in significant discharges to the disposal system. The impact of a steam failure is often mitigated by the loss of heat input to the process.

7.

Boiler Feedwater Failure A loss of boiler feedwater can result in a loss of process cooling in boiler feedwater preheaters and waste heat steam generators, causing multiple discharges

November 2001

1200-80



1200 Relief Systems

to the disposal system. A prolonged failure can also result in the loss of steam supply. Determination of Credibility The potential for a loss of boiler feedwater can be determined in a manner similar to loss of cooling water. In addition to a loss of boiler feedwater supply pumps, a blockage or rupture of a single supply line could also result in the loss of boiler feedwater. Determination of Impact on the Collection and Disposal System Boiler feedwater is often preheated by cooling or condensing process fluids. Therefore, a loss of boiler feedwater may result in overpressure via a loss of process cooling. In addition, an extended loss of boiler feedwater can result in a loss of inventory in steam drums. Subsequent loss of boiler feedwater circulation to waste heat generators can also result in overpressure due to loss of process cooling. In addition, the steam supply may eventually be lost. The impact of a loss of steam is described above. 8.

Large Liquid Release Regardless of its initiating contingency, the largest sustainable liquid release to the disposal system should be identified to evaluate the adequacy of liquid separation and collection equipment (e.g., knockout drums) within the disposal system. Determination of Credibility Blockage of any liquid stream within the process can result in a liquid discharge to the relief header. The most common contingency resulting in a large liquid discharge to the relief header is overfilling of unit feed surge drums. See Item 3, “Overfilling of Vessel” on page 1200-19 for a discussion of the applicability of the overfilling contingency for individual pieces of equipment. Determination of Impact on the Collection and Disposal System A large liquid release rarely leads to a hydraulic (i.e., back pressure) concern. However, such releases should be considered when evaluating or specifying the capacity of any liquid collection equipment within the disposal system. In cases in which the source of the liquid may be limited, consideration should be given to the volume of the liquid source compared to the volume of the liquid collection equipment.

Disposal System Feeds Each discharge to the disposal system is defined in terms of flow rate, pressure, temperature, and composition to enable accurate modeling of the collection and disposal system performance. In many cases, the discharge conditions will correspond to a specific causes of overpressure that has already been evaluated during the determination of the pressure relief requirements of the refinery’s equipment (see


1200-81

November 2001

1200 Relief Systems


Section 1220). In addition to these cases, guidelines are presented here for evaluating discharges from other types of devices. Where a system is protected by more than one of the devices listed below, the total relief requirement from the process should be allocated to the various devices to achieve a conservative result. For example, conventional pressure relief valves are affected by much lower back pressures than control valves are. Therefore, when both a conventional pressure relief valve and a control valve to the disposal system are installed on an equipment item, a conservative evaluation of the disposal system performance will be obtained by assuming the relief flow is entirely through the pressure relief valve. 1.

Pressure Relief Device Discharges The relief rate required for the causes of overpressure under consideration, rather than the actual or “rated” device capacity, should be used as the input flow rate to the disposal system. The required relief rate and conditions should generally already have been quantified as part of the individual equipment required relief analysis and documentation (see Section 1220 for further guidance). Where multiple relief devices provide the total relief requirement, the relief flow requirement should be allocated among them based on the relative sizes and opening pressures of the devices.

2.

Control Valve Discharges Control valves may add to the overall discharge into the disposal system. Unless the control valve is expected to fail open (e.g., due to instrument air failure), the required relief rate should be used as the input to the disposal system. Where both a control valve and a relief device are present, and both discharge to the same disposal system, the required relief rate is conservatively assigned to the relief device. Situations in which a control valve fails open to the disposal system should be analyzed consistent with the guidance presented below for depressuring valves.

3.

Depressuring Valve Discharges Depressuring valves are generally designed to de-inventory a system within a specified amount of time in the event of a process excursion or other emergency. Depressuring may occur as a direct result of any of the common contingencies described in “Simultaneous Releases Caused by a Common Contingency” on page 1200-77 and can have a significant impact on the operation of the disposal system. Depressuring valve discharges are time dependent. In the absence of a detailed analysis of the sequence of events, the initial flow, based on the maximum expected pressure upstream of the depressuring valve, should be used as the input to the disposal system. The impact of the built-up back pressure on the depressuring valve capacity should be evaluated upon completion of the hydraulic modeling of the disposal

November 2001

1200-82



1200 Relief Systems

system. An iterative calculation may be necessary to determine the expected initial flow from the depressuring valve.

1254 Disposal System Modeling To design a new (or to evaluate an existing) relief disposal system, it is necessary to model the response of the disposal system to the relief stream that would flow into it in each of the common contingency cases identified in the analysis outlined above in Section 1253. Hydraulic modeling software packages are available from several vendors for performing this analysis. The analysis method may be broken into two parts. The first is the description (and input to the software) of the physical lay-out of the existing or proposed relief stream collection system. Typically this process starts with an isometric sketch of the collection piping, fittings, and equipment that includes all relief devices, control valves, and depressuring valves that discharge into the collection system. For each piping segment in the collection system, an entry is made in the software model giving the pipe size, length, elevation change, insulation, pipe roughness, and the resistance factor for all fittings (valve, elbows, tees, diameter changes, etc.) in the segment. Typically, an identification number is given to each pipe segment thus entered, as well as to the “nodes,” or intersections of two or more segments. Any separation equipment (such as knockout drums) is also entered, and flow resistance factors are also included for flare tips and seals that may be present. Sources of data for flow resistance data for various fittings and other equipment in a relief header system are listed in Figure 1200-35. Fig. 1200-35 Sources of Flow Resistance Data for Typical Components of a Relief Disposal System Component Type

Sources of Flow Resistance Factor Data API Recommended Practice 521

Pipe Fittings & Block Valves

Guidelines for Pressure Relief and Effluent Handling Systems Flow of Fluids Through Valves, Fittings, and Pipe (Crane TP-410)

Knockout Drums & Relief Scrubbers

In the absence of internal components (packing, trays, or demister pads), only entrance and exit losses, per “fittings” above

Flare Tips

Flare tip manufacturer

Purge Reduction Seals

Seal manufacturer

Liquid Seals

Assumed liquid-free during an emergency relief contingency. Therefore, as for knockout drums.

When the physical layout of the collection system has been identified and entered into the modeling software, the second part of the hydraulic analysis can proceed.


1200-83

November 2001

1200 Relief Systems


For each contingency identified in “Disposal System Design Basis” on page 1200-77, flow data is entered for each relief device, control valve, and depressuring valve identified to be flowing in that contingency. The input flow data includes the flow rate, pressure, temperature, and composition of the stream arising at each “source.” The model is then run; the software uses the known pressure at the ultimate discharge location and the flow and resistance data to calculate the pressure developed, the fluid temperature, and the fluid phase at each node in the system, working backward from the discharge point to each system inlet.

1255 Evaluation of Hydraulic Modeling Results For each common contingency, the hydraulic modeling output data consists of the fluid temperature, pressure, composition, and phase at each node defined in the physical description of the relief collection system being evaluated. Also included are the stream velocity and mass flow rate within each defined piping segment (i.e., between nodes). These modeling results are evaluated against the design criteria outlined below. These criteria can be summarized as follows. The flow out of the relief devices and depressuring valves into the collection system should not be compromised by the pressure built-up at each device’s outlet, and the design conditions of all piping, fittings, and equipment in the collection and disposal system should be appropriate for all conditions predicted by the hydraulic modeling for all common contingencies. 1.

Back Pressure on Pressure Relief Valves The impact of the back pressure on a pressure relief valve is dependent on the type (conventional, balanced, or pilot-operated) of the valve. Section 1230 describes the effects of back pressure on the performance of each type of pressure relief valve. See Item 1, “Pressure Relief Device Discharges” on page 1200-82 for a discussion of the acceptable values of built-up back pressure for the various types of pressure relief valves. Consideration needs to be given to the mechanical back pressure limit for each balanced-bellows type relief valve connected to the disposal system. The maximum back pressure rating for the bellows can be significantly less than the outlet flange pressure rating.

2.

Back Pressure on Control Valves The impact of the back pressure on control valves should be evaluated using the valve manufacturer’s capacity calculation methods. The total input to the disposal system may be reduced by the developed back pressure. The impact of back pressure on fully open control valves is determined using the guidance presented below for depressuring valves.

3.

Back Pressure on Depressuring Valves Back pressure may cause the capacity of a depressuring valve to be reduced due to sub-critical flow. The reduction in capacity should be included when designing systems to be depressurized in a specified amount of time. The back

November 2001

1200-84



1200 Relief Systems

pressure may serve to reduce the overall initial discharge to the disposal system from the depressuring valve. 4.

Mechanical Design of Collection System Components The pressures and temperatures predicted by the hydraulic model to be developed within the collection and disposal system should be considered when specifying (or evaluating) the design limits of all system components. Particular attention should be given to the materials of construction of components at which temperatures below -20 °F are predicted.

1256 Disposal System Components Materials Specifications The materials of construction of collection and disposal system components should be specified for the extreme fluid pressures and temperatures predicted by the hydraulic modeling of all common contingencies. Careful consideration should be given to the potential for low temperatures due to auto-refrigerating fluids, and for high temperatures that may accompany the process upset causing the relieving contingency. The potential for corrosive fluids to be present in all parts of a collection and disposal system should also be considered when specifying materials of construction for the system components.

Piping and Piping Components To prevent the collection of liquids at low-point “pockets,” which might cause blockage of the disposal system, relief piping should always drain freely to a liquid separation vessel (e.g., a flare knockout drum). Similarly, piping connections to a collection header are made at the top of the header. If a collection system contains multiple separation vessels, each piping segment is installed to be free-draining to the downstream vessel. The final section of piping should drain freely in the upstream direction (i.e., back to the final knockout drum). When collection piping contains valves (e.g., for isolation of individual relief devices or of header sections), such valves should be of the full-port variety, and either car-sealed or locked in the full-open position. Such valves should be designed and installed so that, in case of a mechanical failure, the valve fails in the full-open position. Relief disposal piping should meet the requirements of ASME Piping Code B31.3 for piping support and flexibility. Particular attention should be given to loads induced by potential thermal expansion or contraction of the piping.

Separation Equipment Fluids entering collection systems may require processing by separation equipment for one or more of the following reasons. •

The fluid may contain compounds whose economic value justifies the expense of the separation equipment.

•

The fluid may contain compounds incompatible with downstream components in the collection system.


1200-85

November 2001

1200 Relief Systems

•


The fluid may contain hazardous compounds whose discharge to the atmosphere, whether before or after combustion, would be unacceptable on health, safety, or regulatory grounds.

All pressure relief and depressuring streams should be evaluated for the potential need for separation equipment prior to atmospheric discharge or flaring. 1.

Separation Equipment Upstream of Knockout Drums As described above, a variety of justifications may exist for separation equipment in a collection and disposal system. One example seen in refineries is scrubbing of acid relief streams from an HF alkylation unit with potassium hydroxide upstream of the usual knockout drum and flare arrangement. Refer to API Recommended Practice 521, the CCPS book Guidelines for Pressure Relief and Effluent Handling Systems, and the references in each for further information on the need for and design of separation equipment in relief and vent collection systems.

2.

Knockout Drums A knockout drum is a vapor-liquid separator installed in a collection system to limit the quantity of liquids carried to downstream sections of the system. A knockout drum is essential in this location because a flare stack is not capable of handling liquids, and because liquid droplets can pose a serious hazard to equipment and personnel when ignited as they exit a flare tip. In some cases, multiple knockout drums are installed in series in a given relief header. This may be necessary if an individual vessel or unit has the potential to discharge a large volume of liquid to the collection and disposal system. The Company prefers to locate knockout drums near the plot limit of each process plant where the plant relief headers enter the main interconnecting pipeway. Occasionally it is economical to provide a single knockout drum for a group of plants. Knockout drums should be designed or evaluated for each global relief contingency identified, using the methods described in API Recommended Practice 521, to meet the following criteria. Liquid Droplet Separation The final knockout drum in a collection system that discharges either to a flare or to atmosphere should separate out entrained liquid droplets larger than some specified size. To avoid “burning rain” from a flare, API Recommended Practice 521 recommends this size be in the range of 300 to 600 micrometers diameter; flares can normally handle droplets smaller than 300 micrometers. Methods for evaluating the separation performance of a knockout drum are described in RP 521. Liquid Hold-up Capacity The hold-up volume of a knockout drum is the minimum volume available for temporary containment of liquids without interfering with the drum’s ability to

November 2001

1200-86



1200 Relief Systems

achieve the specified vapor/liquid separation. The combined hold-up volume of the knockout drums in a pressure relief collection system should be sufficient to contain the greatest amount of liquid predicted to be discharged to the collection system for some specified length of time by any single contingency. Knockout drums should be sized so that a 10-minute blow at maximum rate will fill the drum half-full. this conforms with API RP 521 recommendations. When evaluating the sufficiency of a knockout drum’s hold-up volume, credit is often taken for the flow capacity of knockout drum pumps. Maximum liquid volume also depends on the operator’s reaction time to the relief event. Operating management should concur on the selection of the knockout drum size. Figure 1200-36 shows a typical knockout drum installation and illustrates the relationship between liquid hold-up volume and droplet separation. Fig. 1200-36 Knockout Drum Hold-up Volume and Droplet Separation. (Courtesy of the American Petroleum Institute)

Producing facilities often operate knockout drums at 20 to 30 psig, and liquids are forced by that pressure to temporary storage tanks or sumps from which they can be pumped whenever it is convenient. Operating the scrubber under pressure allows both the knockout drum and the associated pumping equipment to be smaller, since the knockout drum should reasonably never reach one-half capacity.


1200-87

November 2001

1200 Relief Systems


Level Control Requirements Instrumentation should normally be installed on a knockout drum for automatic control of the liquid level in the drum. A high level alarm signal should be delivered to the appropriate control room to warn of possible liquid carryover to the flare or atmospheric discharge. Due to the low pressure normally in the drum, a pump is usually required to remove liquids from the knockout. Careful analysis needs to be used in specifying the power source for this pump. Because of the low pressure normally seen by knockout drum, designers are often tempted to specify a very low design pressure for this vessel. Operating experience has proven this a false and short-sighted economy. Most operating companies currently require a minimum MAWP of at least 50 psig for flare knockout drums.

Back-flow Prevention Systems Disposal systems should include equipment to prevent the infiltration of hazardous levels of air. This objective is normally achieved by using a flow of purge gas through the entire disposal system. The required flow rate of purge gas can be significantly decreased by the installation of a purge reduction seal. The following sections provide a brief summary of purge gas flow requirements and of purge reduction and flare seals. See API Recommended Practice 521 for further discussion of these topics. 1.

Purge Gas Typically, flare headers are continuously purged with a flow of non-oxidizing gas compatible with the fluids expected to be in the header and with the materials of construction of the header and all component equipment. The purge gas should normally be introduced at the upstream end of each branch of the header system. The purge gas system is normally designed to enable additional flow for the following occasions. – –

–

Start-up of the flare system after being opened to atmosphere (e.g., for maintenance). Contraction of the vapors in the flare system as they cool following a hightemperature venting contingency. The maximum required purge flow rate will normally be dictated by this contingency, if it can occur. Release of low molecular weight fluids (e.g., H2) to the flare system.

For flare stacks without purge reduction seals, the minimum purge gas flow through the stack is usually calculated using the methods described in H. W. Husa, “How to Compute Safe Purge Rates,” (Hydrocarbon Processing and Petroleum Refiner, vol. 43, no. 5, pp. 179-182 (1964)). For flare stacks equipped with purge reduction seals, the appropriate normal purge flow should be determined in consultation with the manufacturers of the seal and of the flare tip. The volume of gas is typically enough to maintain a velocity of about 0.1

November 2001

1200-88



1200 Relief Systems

foot per second in the relief header with no other gas flow into the relief header. This injected purge gas is adequate to maintain the gas seal. In older relief systems, the amount of purge gas is flow controlled. Today, to conserve energy, the rate of purge gas injected is often controlled by using the relief header operating pressure as the measured variable. When there is enough PSV leakage or process venting to maintain the desired back pressure, no purge gas is injected. 2.

Seals Purge Reduction Seals A flare tip is normally specified with a purge reduction seal located immediately below the tip. This seal may be of either the molecular or the velocity design. A gas seal (molecular seal) can replace the liquid seal drum on an elevated flare, but not on a ground flare. Figure 1200-37 and Figure 1200-38 show diagrams of typical molecular and velocity seal designs. Refer to API Recommended Practice 521 for further discussion of purge reduction seals. Note that the pressure drop induced by any purge reduction seal should be included when performing a flare system hydraulic analysis.

The gas seal is located just below the flare tip and serves to prevent air entry into the stack. It is supplied by the flare supplier. The gas seal is often called a molecular seal because it depends on the density difference between air and hydrocarbon gas. A continuous stream of purge gas is required for proper functioning of the gas seal, but the amount of purge gas is much less than would be required without the seal. Gas seals have major advantages over liquid seals: they do not slosh and they produce much less oily water. However, the gas seal must be drained and the drain loop has to be sealed. Figure 1200-37 shows a gas seal. Perhaps several examples will serve to underscore the importance of keeping the gas seal drains clear and functioning. In an incident in a cold climate, a boiler plant failure interrupted the steam supply to several key turbine drivers. As the hydrocracker makeup compressor slowed down, excess hydrogen was vented to the flare. But no increased flame was observed and the crude unit pressure “… went off the chart!” Quick shutdown of the entire complex prevented potentially catastrophic failures. The gas seal was plugged with ice. Witnesses saw debris flying out of the flare. Ice boulders weighing 50-60 pounds were found at the base of the flare. Also found were the 3-inch steam center pipe and nozzle. In another incident in a milder climate, when the drain-plugged gas seal began to fill with water (presumably steam condensate and rain) it became a good stripper of the H2S in the relief gas bubbling through. The concentration built up and, eventually, a relatively large gas release expelled sour water onto personnel below. Such incidents serve to emphasize the importance of keeping the seal drain line free and clear.


1200-89

November 2001

1200 Relief Systems


In another incident in a milder climate, the drain line plugged and the seal filled with water. The water came mostly from the steam system which was used to assist in smokeless flaring. The liquid filled seal was calculated to have about 10 tons of water in it and as this water was off-center and sloshed back-and-forth during small gas releases, it caused the 120 foot flare stack to shake violently and. The shaking actually resulted it the guy-wire turnbuckles to loosen; one of the turnbuckles had loosened up over 1 foot when it was noticed and only had another few inches of working length before it would have let go. The plant handled sour fluids and flare system was the only relief system, thus in order to clear the flare it was necessary to simply blow out the water with a high release rate. This meant that the plant to take the chance that the fluid in the molecular seal was indeed water and not LPGs which commonly accumulated in the flare knock-out drum. Luckily they were right as liquid water rained down on the plant and not flaming liquid hydrocarbon. Since a gas seal is required with an elevated flare to keep air out of the flare stack, the liquid seal is usually omitted from an elevated-only flare system. If a vapor recovery compressor is used, a liquid seal is used to provide a minimum header back pressure. The fluidic seal is an alternative to the gas seal. It uses an open wall-less venturi that permits flow out of the flare in one direction with very little resistance but strongly resists counterflow of air back into the stack. This venturi consists of a series of baffles, like open-ended cones in appearance, mounted with the flare tip. The fluidic seal is smaller and less expensive than a molecular seal and, since it weighs less, there is less structural load on the flare stack. However, fluidic seals require more purge gas. The Company has used fluidic seals in offshore applications, but not in refineries. Figure 1200-38 shows a fluidic seal. Liquid Seals Liquid seals are often installed immediately upstream of a flare stack to maintain a positive pressure in the collection system and to provide protection against flashback into the flare header. Figure 1200-39 is a diagram of a typical flare seal drum. Refer to API Recommended Practice 521 for additional description of liquid seals and methods for sizing them. In a ground flare system, a double liquid seal provides a positive means of diverting excess relief gas into the elevated flare. The maximum volume of relief gas that can go to the ground flare is determined by the water seal depth in the upper seal. Once the relief header pressure exceeds that head, the excess relief gas goes to the elevated flare. Figure 1200-39 shows a double liquid seal. Some facilities use a dual header design with dedicated high- and low-pressure flares. The high-pressure ground flare is usually self-aspirating and requires no water or steam injection.

November 2001

1200-90



1200 Relief Systems

Fig. 1200-37 Schematic Diagram of a Gas (Molecular) Seal

Fig. 1200-38 Schematic Diagram of a Fluidic Seal (NAO) (Courtesy of NAO, Inc.)

Fig. 1200-39 Seal Drum with Double Seal for a Ground Flare and an Elevated Flare

Depth of the Water Seal The depth of liquid in the seal drum determines the relief header back pressure. This depth is set by the flare supplier, but it can usually be altered somewhat (with the supplier’s concurrence) to suit plant conditions. Typical seal depths are 2 feet for elevated flares and 6 inches for ground flares.


1200-91

November 2001

1200 Relief Systems


Pulsations API RP 521 states that surging in seal drums can be minimized by using Vnotches on the end of the dip leg. This does not conform to Company experience. If the water sloshes in the seal drum, it will cause pulsations in the gas flow to the flare, resulting in noise and light disturbances. The Company prefers either a displacement seal as shown in Figure 1200-40 or a perforated anti-slosh baffle as shown in Figure 1200-41. Fig. 1200-40 Seal Drum with Displacement Seal

Fig. 1200-41 Seal Drum with a Perforated AntiSlosh Baffle

Combustion Tips Combustion tips are available from a number of makers. They are designed to provide smokeless flaring up to some fraction of the maximum relief capacity of the relief system. Steam injection, when available, is used to suppress smoke with required steam pressures from 75 to 150 psig. Lower pressure steam can be used if the steam piping and injectors are designed for it. High pressure (15 psig) relief gas or air can be substituted for steam in some special designs. See “Flares at Producing Sites” on page 1200-72 for a discussion of flares at producing sites where steam is rarely available. Combustion tips with smokeless capacities up to one million pounds per hour are available. Few process plants, however, have sufficient excess steam to supply the flare during a major release; and furthermore, during a major upset, extra steam is often required by the process units. Combustion tips are made of stainless steel in diameters up to 6 feet. The tips and the piping are flanged so that they can be removed from the flare stack. The design depends upon the supplier selected. The Company has had good experience with two designs: Flaregas and John Zink.

November 2001

1200-92



1200 Relief Systems

The Flaregas design uses a large number of special Coanda nozzles which promote thorough mixing of air and steam with the relief gas prior to burning. Steam is distributed to the nozzles by a steam chest. As the steam flows through the nozzles into the mixing chamber, it draws air along with it. Figure 1200-42 shows a flare tip with Coanda nozzles. Figure 1200-43 shows the Coanda nozzle. Fig. 1200-42 Flare Tip Using Coanda Nozzles (Courtesy of Flaregas Corp.)

Fig. 1200-43 Schematic of a Coanda Nozzle (Courtesy of Flaregas Corp.)

The other design, John Zink, uses a number of internal pipes to distribute the steam-air mixture across the combustion zone. This design also utilizes a separate central steam injector and an outer steam ring to give three stages of smokeless combustion capacity. Each of the three injection systems has its own controls. The outer ring is very noisy and is used only in a major release where additional smokeless capacity is required with the sacrifice of noise control. Figure 1200-44 shows a flare tip with internal steam piping. With the exception of the outer ring, both designs are equally noisy.

Steam Injection Smoke is unburned carbon from incomplete hydrocarbon combustion. The oxygen in the air first combines with hydrogen atoms in the hydrocarbon molecules, freeing carbon atoms for secondary reactions. If the hydrocarbon molecules crack at the high combustion temperature, then more free carbon is produced. Unsaturated hydrocarbon molecules produce more free carbon. This unburned carbon appears as smoke. The free carbon will be converted to carbon dioxide (and not produce smoke) as long as an excess of oxygen is present within the combustion zone. This requires good air/gas mixing. Thick smoke will result if a hydrocarbon gas is flared without prior mixing with air.


1200-93

November 2001

1200 Relief Systems


Fig. 1200-44 Flare Tip with Internal Steam Piping (Courtesy of the John Zink Company)

Injection of steam helps control the carbon-hydrogen ratio. Steam catalyzes the combustion reactions and lowers the temperature in the combustion zone, thus suppressing smoke. However the main reason for injecting a high velocity steam-air mixture into the combustion zone is to increase turbulence, which improves combustion and reduces the size of the flame. Unfortunately, steam injection increases the combustion roar. The usual injection rate is about 0.5 pounds of steam per pound of hydrocarbon. Actual steam requirements for smokeless combustion are hard to predict because they depend on the molecular weight of the relief gas, the percentage of unsaturates, and the air/gas mixing ratio. Butadiene requires 0.8 to 1 pound of steam per pound of gas; methane requires 0.1 to 0.2 pounds per pound, while hydrogen burns smokelessly without steam.

Control of Steam Injection Automatic control of steam injection is difficult because the volume of relief gas is not easily measured because of the wide rangeability (more than 100 to 1). Even if the volume of relief gas is known, molecular weight and composition have a major effect on the steam requirement. Depending on the type of combustion tip, several separate steam injection valves may be needed, requiring split-range control signals.

November 2001

1200-94



1200 Relief Systems

Currently the Company uses three methods to automatically control steam injection: 1.

Flaregas Flarescan—Measuring the flame radiation with thermocouples set around the flare tip. Steam is increased until smoke is suppressed, resulting in lower infrared radiation.

2.

John Zink Zoom—Measuring the infrared radiation with a television camera. Steam is increased to reduce the amount of infrared radiation.

3.

Panametrics—Measuring the volume and the density of the relief gas with an ultrasonic flowmeter. The required amount of steam is calculated by a microcomputer.

All three methods work for low-to-medium relief rates. Smoke is produced with large rapid changes in the relief rate. Adjusting steam rate with manual controls is not acceptable.

Flares 1.

Types of Flares Refer to API Recommended Practice 521 and the CCPS Guidelines for Pressure Relief and Effluent Handling Systems for a description of the characteristics of elevated and ground flares. Flares installed at Chevron facilities will generally be the elevated variety. Figure 1200-45 contains a schematic diagram of a typical flare installation.

2.

Smokeless Operation Requirements For environmental and public relations reasons, flares should generally be designed to operate in a smokeless regime. However, as smokeless operation at the highest potential flow rates can be difficult to achieve, a design maximum flow rate for which smokeless operation is required is usually specified. Smokeless operation is typically accomplished by steam injection at the flare tip. However, other approaches to achieving smokeless flaring exist, such as air injection, and the use of high (≥~5 psig) flare discharge pressures. In some cases, these methods may be more practical and economical than steam injection.

3.

Sizing Discharge Velocity Considerations Pipe tips are have often been designed to limit the exit velocity to a maximum Mach number in the range of 0.5. This is not necessary because high velocity lifted flames are perfectly stable. Higher velocities should be considered on an economic basis. Note that EPA regulations given at 40 CFR 60.18 and 40 CFR 63.11 specify maximum normal flare exit velocities for certain types of production and process facilities. Radiant Heat Considerations The location and height of a flare stack are usually chosen so that for each flaring contingency established per “Disposal System Design Basis” on


1200-95

November 2001

1200 Relief Systems

November 2001

Fig. 1200-45 Typical Flare Installation. (Courtesy of the American Petroleum Institute)

1200-96




1200 Relief Systems

page 1200-77, the thermal radiation emitted by the flare flame does not exceed some acceptable value at locations to which facility personnel have unrestricted access. See API Recommended Practice 521 for data concerning the effects of exposure to thermal radiation. Flare Height Height determination for an elevated flare stack is not a precise science. The stack height is selected to reduce the radiant heat intensity at ground level to desired safe limits for personnel and equipment. The Company uses the method outlined in API RP 521, Appendix A. Selecting an acceptable radiant intensity is the key design decision. The radiant intensity is calculated by assuming uniform spherical radiation from the flame center. Calculation is inexact because of the following conditions. 1.

Emissivity must be approximated. Emissivity is that fraction of the thermal energy of the flared gas that is actually radiated as heat. Test emissivities range from 60% to 15% and even lower when steam is introduced into the combustion zone. While steam is used for smoke suppression in most modern flares, steam may not be available during the major emergency that causes the maximum flare release. Therefore, for purposes of calculation assume an emissivity of 40%.

2.

The flame center is unknown but is assumed to be located halfway up the flame. This varies significantly for a vertical flame in windless conditions to a nearly horizontal flame in a severe windstorm. Flames from major releases are typically several hundred feet long. Therefore, for purposes of calculation assume a flame length of 100 tip diameters.

A ground level radiation intensity of 1500 Btu/hr/ft2 is the maximum safe level (OSHA Standard) for prolonged exposure. This level is approximately the threshold of pain for bare unprotected skin. Below this level, exposure can be tolerated for virtually unlimited periods. Above this level, skin will blister. This intensity, however, is too low for a realistic flare height design onshore because personnel in the vicinity of the flare will promptly escape to a safe distance as soon as a major release begins. For offshore locations see “Flares at Producing Sites” in Section 1252. The API RP 521 sample calculations assume two different maximum allowable radiation intensities at grade—2000 and 3000 Btu/hr/ft2. Using maximum allowable radiation intensity of 2600 Btu/hr/ft2 at grade in the immediate vicinity of the flare has worked well for the design of recent installations. By using these design criteria, most Company flares are 150 to 175 feet high. Noise Considerations Local and federal regulations concerning workplace audible noise levels should be considered when designing vent stacks, flare stacks, and flare tips.


1200-97

November 2001

1200 Relief Systems


Distance Between Flares and Tankage Sometimes flares must be constructed adjacent to tankage. Normally a distance of 200 feet from the flare to the nearest tank is enough to protect the tank from fire hazard. Elevated flares sometimes spit burning liquid during a major release, and the setback distance should accommodate this hazard. Potential Hazardous Plume Considerations When the potential exists for flare gas to contain toxic compounds (e.g., H2S, NH3), the possibility of loss of ignition of the flare stream requires consideration. In most such cases, an atmospheric dispersion analysis is performed for the potentially unburned relief stream flowing from the stack. Such an analysis should be performed for each relief contingency for which the relief gas is expected to contain a toxic compound. CRTC’s Health, Environment and Safety Group is normally consulted if dispersion modeling is required. The following guidelines apply to the atmospheric concentrations predicted by the modeling. Toxic compound concentrations predicted by the modeling should not exceed the “Immediately Dangerous to Life and Health” (IDLH) level at any location accessible to personnel. The IDLH levels are specified by the National Institute for Occupational Safety and Health (NIOSH). Toxic compound concentrations predicted by the modeling should not exceed levels set by local or federal regulatory agencies at any location to which the public has unrestricted access (typically outside the facility fence line). In some cases, the potential may exist for the flare gas combustion products to contain toxic compounds (e.g., SO2). In these cases, dispersion analysis of the combustion product plume may be required to ensure that unacceptable concentrations of the hazardous products do not reach facility personnel or the public. 4.

Auxiliary Equipment Pilots A flare tip is normally specified with at least one constantly burning pilot flame. Currently available pilots are equipped with a flame monitoring device that sends an alarm signal to the appropriate control room in case of a loss of the pilot flame. The pilot flame monitoring device may also automatically activate an electronic pilot igniter upon detecting a loss of the flame. Flare Monitoring The flare is normally equipped with a method for alerting control room personnel of flare activity. This may take the form of video monitoring of the flare tip, a pressure transducer in the flare stack, or any other reliable system. Aircraft Warning Lights When required by local or federal regulations, elevated flare stacks should be equipped with appropriate aircraft warning lights.

November 2001

1200-98



1200 Relief Systems

1260 Pressure Relief System Design — New Facility The purpose of this section is to provide a brief, step-by-step summary of the process of designing a pressure relief system for a new process facility. Note that, as with most design processes, many of these steps will be performed iteratively as the design of the entire facility is revised.

1261 Facility Information Compile information about the process itself and about the vessels and other equipment that the process unit will comprise: •

Block Flow Diagrams Process Flow Diagrams (PFDs) w/ heat & material balance Piping and Instrumentation Diagrams (P&IDs) Plot plans Preliminary relief header layout

•

Process simulations

•

Pump & compressor specification sheets and design flow curves Tank specifications Piping and fitting pressure ratings

•

Vessel and heat exchanger fabrication drawings, as needed

1262 Equipment List Compile a list of all of the equipment items in the facility. This list is typically generated from the P&IDs. The list will typically exclude items that may be calledout on a P&ID, but that can be classified as piping components (e.g., steam traps, desuperheaters, etc.) •

All ASME vessels (both Section I and Section VIII of the BPVC)

•

All low-pressure equipment (tanks, etc.)

•

All Pumps, Compressors

1263 Causes of Overpressure For each equipment item in the facility, identify all potential causes of overpressure. Document the basis for the identification of causes of overpressure as applicable or inapplicable to each equipment item. Guidance for this contingency identification step is summarized in Section 1220 and in API Recommended Practice 521. Example documentation for this step is shown in Figure 1200-47. Note that when a cause of overpressure is “frequently applicable” to the class of equipment


1200-99

November 2001

1200 Relief Systems


being analyzed, the comments section of the contingency documentation contains an explanation of contingency’s identification as either applicable or inapplicable to the specific equipment item being analyzed. This level of detail in the documentation is important so that questions raised in future design reviews, PHAs, and MOC processes can be efficiently addressed by reference to the contingency identification documentation. More than any other, this step in the relief systems design process requires the services of an engineer experienced with the analysis of the pressure relief requirements of process facilities.

1264 Required Relief Flow Rates For each causes of overpressure identified as applicable in Section 1263, calculate the relief flow rate required to prevent the equipment item in question from exceeding its maximum allowable pressure (e.g., MAWP plus allowable accumulation). Fluid properties required for these calculations are taken from the process simulation, adjusted to relieving conditions. The methods for performing some of these calculations are briefly discussed in Section 1220; further discussion is given in API Recommended Practice 521. Example documentation for such a calculation is shown in Figure 1200-48.

1265 Required Relief Devices At this point in the design process, the pressure relief requirements of the process equipment have been identified and quantified. The next steps involve sizing and selection of the relief devices and sizing of associated piping to handle the required relief flow rates.

Relief Device Selection For each equipment item or group of items that requires pressure relief, select a relief device type to provide the required relief, using the selection guidance provided in “Pressure Relief Device Selection” on page 1200-28. Note that this is a preliminary choice; the type of device selected may change as the design of the inlet and outlet piping and of the collection header is finalized.

Required Device Size Size each relief device so that it will have adequate flow capacity for all causes of overpressure applicable to the equipment item or items it is intended to protect. Often the causes of overpressure that requires the largest relief valve orifice or rupture disk diameter can be determined by direct examination of the required flow rates. However, when contingencies differ with respect to the relief fluid’s composition, phase, or other property, the largest required flow rate may not correspond to the largest required relief device size. Therefore, after sizing the relief device for what appears to be the contingency requiring the largest device, the device should be evaluated and documented for all causes of overpressure identified, to ensure adequate capacity. Example relief device documentation is shown in Figure 6-3.

November 2001

1200-100



1200 Relief Systems

Again, by adequately documenting the initial design calculations, questions that arise later in the design review and PHA processes can be quickly addressed.

Discharge Fluid Disposal Method Select an appropriate method for handling the fluids to be discharged from this relief device. See Section 1252, “Choice of Disposal Method” on page 1200-65 for a discussion of typical discharge fluid handling and disposal methods. Possible choices for disposal methods include atmosphere, an oily-water sewer, and a flare system. The cost of intermediate (i.e., upstream of a flare) processing of the discharge is typically justified only by the presence of a highly valuable, hazardous, or incompatible component in the discharge stream.

1266 Relief Device Inlet and Outlet Lines Having selected and sized the required relief devices and chosen the ultimate disposition locations for their discharge streams, the designer now turns to the piping systems for connecting the relief devices to the equipment they must protect and to the disposal systems chosen.

1267 Relief Discharge Collection and Disposal Systems Collection System Requirements In “Discharge Fluid Disposal Method” on page 1200-101, an appropriate disposal method was identified for each relief device: to atmosphere, to another process location, or to a collection and disposal system. Now address whether one collection system will suffice. The devices identified for discharge to a collection system and all of the fluids they may discharge are considered on a global basis. Considerations that may indicate the need for more than one relief discharge collection system include the following. 1.

Fluid Compatibility Chemical reactions within relief headers are generally to be avoided. When the potential exists for a reaction between fluids discharged from different relief devices, consideration should be given to installing separate headers for collecting the incompatible discharges.

2.

Fluid Handling Requirements If the discharge of a particular relief device or set of devices requires some particular handling or processing - such as neutralization, or temperature quenching - a separate collection system may be appropriate. In some instances, this separate collection system may discharge into the “primary” relief collection system following the specialized “processing.” In some instances, it may be possible to route the “processed” relief stream into the main relief collection system for its ultimate disposal.


1200-101

November 2001

1200 Relief Systems

3.


Materials Requirements The properties of the fluids discharge from some relief devices may necessitate special metallurgy in the downstream relief header. In such cases it may be more economical to design and install a separate header and disposal system for those devices with the special requirements rather than to upgrade an entire single collection system.

4.

Set Pressure Range As discussed in Section 1230 and Section 1240, the operation and flow capacity of relief devices depends on the pressure at the discharge of the device as a percentage of the device’s set pressure. When the devices discharging into a single collection header system have widely disparate set pressures, header pressures that are acceptable for devices with high set pressures may render devices with low set pressures ineffective or inoperable. In such cases, parallel “high pressure” and “low pressure” relief headers may be appropriate.

Simultaneous Discharge From Multiple Sources For each relief collection system serving multiple equipment items, identify all “global” causes of overpressure. These are contingencies, initiated by a single event, that will cause discharge into the collection system simultaneously from multiple sources. See “Simultaneous Releases Caused by a Common Contingency” on page 1200-77 for further discussion of the common contingency identification process. Prepare a brief description of each contingency, including the initiating event, the process equipment effected by the event, and the devices expected to discharge into the collection system. For each contingency, compile the expected discharge data for each individual flow source (relief device, control valve, or depressuring valve). These data include flow rate and physical properties (composition, pressure, temperature, etc.) of each stream that will enter the collection system in a given contingency, and are typically available from the calculations documented in “Required Relief Flow Rates” on page 1200-100. Note that, per API Recommended Practice 521, the flow rate used in this collection system hydraulic evaluation is the rate required to prevent excessive equipment overpressure, not the actual capacity of each relief device. Calculate stream data for discharge sources not previously evaluated.

Collection System Model Prepare an isometric sketch of each planned relief collection system, from the devices discharging into it, through each relief header and any “processing” equipment (e.g., scrubbers, quench drums, knockout drums), to the final discharge point (typically a flare tip). Modeling of the flow behavior of relief discharges through collection and disposal systems is usually performed using specialized computer programs. Conceptually, however, the task is straightforward: a flow model is “built” of the collection system. Each piping segment and fitting is represented by its internal diameter and roughness, and either its length, its “equivalent length,” or its flow-resistance coefficient. The typical sources of flow resistance data are Crane Technical Paper

November 2001

1200-102



1200 Relief Systems

No. 410, API RP 521, and Perry’s handbook. Any equipment items in the collection system, from scrubbers and knockout drums to seals and flare tips, are also included in the hydraulic model, typically using data provided by the respective manufacturer.

Relief Contingency Hydraulic Modeling Once the computer model of the collection and disposal system has been assembled, it is used to evaluate the performance of the system itself in the event of each of the global relief contingencies identified in “Simultaneous Discharge From Multiple Sources” on page 1200-102. For each common contingency, stream data for each source (usually a relief device, but sometimes a control or depressuring valve) that will be flowing in the contingencie is entered into the computer program. The program then uses equations for energy, mass, and momentum balance to solve for the pressure drop developed across each segment of pipe and fittings in the collection system as well as the fluid properties at the junction between each segment. These results are then used to evaluate the design of the collection system and its impact on the performance of relief and depressuring devices. Questions typically asked at this stage in the design process include the following. Is the built-up back pressure developed at each inlet to the collection system acceptable for the proper performance of the relief device or depressuring valve at that inlet? If not, at what component(s) in the collection system is the excessive back pressure developed? Should those components be redesigned (typically enlarged) to reduce the built-up pressure, or should the relief device type be changed to accommodate the calculated back pressure? Are the temperature and pressure ratings of all pipe, piping components, fittings, and vessels in the collection and disposal system appropriate for the pressure, temperature, and flow velocity calculated at that component?

Relief Separation Equipment The hydraulic modeling done above should have included the effects of any separation equipment present in the collection and disposal system. This equipment includes absorbers, quench drums, knockout drums, etc. The model will generally have assumed a separation efficiency of 100% (some modeling programs may allow the assumption of lower efficiencies). In this step of the system design process, the equipment parameters necessary to meet the separation criteria are developed. See Section 1250 for additional information on sizing relief separation equipment and knockout drum pumps.

Relief Disposal Equipment The final step in the relief system design process is the design of equipment for the ultimate disposal of the relief discharge streams. This equipment is most frequently an elevated flare, but as described in Section 1250, other types of disposal equipment are sometimes employed. These include ground flares, flare or burn pits, and thermal oxidizers. In some cases, a collection system may simply discharge to atmosphere.


1200-103

November 2001

1200 Relief Systems


This step includes selecting a location and height for the flare stack, and designing the flare’s auxiliary equipment - purge systems, seals, ignition systems, etc. More information on the methods and issues in this design process are given in Section 1250.

1270 Pressure Relief System Design — Existing Facility Once a process unit - along with its pressure relief system - has been built and is operating, the considerations of its design often fall into the background as considerations of its operation and maintenance come to the foreground. However, there are at least two circumstances in which pressure relief system design considerations are of particular relevance for an existing process facility. The first is during a design review of the existing pressure relief system. The need for these reviews has been recognized recently for two reasons. Personnel at many process facilities have discovered that the available documentation of their pressure relief systems - typically only a set of relief device specification sheets, but in some cases less than that - is insufficient to meet the process safety information requirements of both the OSHA Process Safety Management standard (29 CFR 1910.119) and the EPA Risk Management Program rule (40 CFR 68). If it is carefully documented, a relief system design review will generate the required relief system design information. The second reason these design reviews have been recognized as necessary is that changes to a process unit over a few decades of its operation have typically resulted in some inadequacies in the unit’s pressure relief system. Occasionally no relief device is present where one is needed, or its inlet or outlet line is improperly designed or installed, or many new devices are tied into an existing collection system without proper re-analysis of the system’s available capacity. Design reviews performed at a large number of various existing process facilities have found that on the order of 40 percent of the equipment items have some deficiency in their pressure relief systems (see Berwanger, Kreder, and Lee). Such reviews currently being performed at many process facilities require a familiarity with relief system design considerations on the part of operations and maintenance personnel and management. The second situation in which relief system design principles are particularly important to an existing facility is in the development and application of procedures for the management of changes to these facilities. Such management of change (MOC) procedures have been developed during roughly the past ten years and are now required by the two US federal regulations mentioned above. Key aspects required of such management of change procedures are the thorough consideration of proposed process changes and the updating of the facility’s process safety information to reflect implemented changes. Because relief systems are vital safety systems and because their requirements and performance are sensitive to changes in process conditions, thorough evaluation of proposed changes for their impacts on relief systems must be a part of any MOC process. Because the relief system design basis is an element of the facility’s process safety information, all proposed changes must be reviewed for potential impacts on the relief system design basis in order to keep the design basis information current.

November 2001

1200-104



1200 Relief Systems

1271 Pressure Relief System Design Reviews The process for the design of a pressure relief system for a new facility was summarized in Section 1260. In general, the process for a design review of an existing pressure relief system is very similar to this process for the initial design of a new pressure relief system. There are, however, some essential differences. 1.

Because the design review is of a process and a pressure relief system that is nearly unchanging, there is generally no need for the iterative revision of the analysis that is required when designing relief systems for process equipment that has not yet been built. While the design review team should be aware of (and be kept informed of) possible changes to the process while they are conducting their review, they can generally take the existing equipment and process as fixed.

2.

While the design reviewer may not need to revise his analysis to accommodate weekly or daily revisions to a process design, it is essential that he be working with up-to-date information concerning the process equipment and the process conditions. The goal of the review, after all, is to ensure that the pressure relief system is adequate for the facility as it is currently operating, which is nearly always different from its original design. The design reviewer must be aware that equipment files may contain information for equipment that has long since been replaced. Reasonable efforts should be expended to ensure that all input data for the design review (P&IDs, equipment files, relief device data sheets, process simulations, etc.) are current. In some cases, field verification of file data may be necessary. In other cases, existing data may need to be scaled to current operating or relieving conditions.

3.

While the goal of a new relief system design is typically to specify the most economical relief system that can provide the required relief of potential causes of overpressure, the goal of a design review of an existing pressure relief system is slightly more modest. A review aims merely to verify that the existing system can supply at least the required relief flow rates for all identified contingencies. Because the relief system being analyzed has already been installed, determining the best, most efficient design is not necessary. This distinction affects the assumptions that are made when there are uncertainties in input data or methodologies for causes of overpressure evaluations or relief flow rate calculations. Because the design review’s goal is verification of adequacy, whenever the reviewer is faced with uncertainty, he is justified in making the assumption that will yield the most conservative result. Only if this assumption indicates that the existing relief system may not be able to relieve overpressure adequately is further effort to refine the evaluation justified. By making use of such simplifying – yet conservative – assumptions, the design review team can make the most efficient use of its typically limited resources.

4.

One of the key goals of any pressure relief system design review is to generate documentation of the existing relief system’s design basis that is sufficiently thorough to meet the process safety information requirements of the OSHA


1200-105

November 2001

1200 Relief Systems


Process Safety Management standard and the EPA Risk Management Program rule. This documentation can provide data that is vital to facility personnel involved in the management of changes to the process or in the design and implementation of operating and maintenance procedures for the process. 5.

Because this is a design review and not a design project, the desired product of the review (in addition to design basis documentation) is a listing of any concerns about the existing pressure relief system. This product is typically referred to as a Relief System Concern List. Upon completion of the design review, this list is reviewed with the unit’s operating, safety, and management personnel. Items on the Concern List are issues in the design and installation of the pressure relief system that, in the judgment of the review team, should be brought to the attention of the interested parties. These items may represent deficiencies in the overpressure protection of the equipment analyzed, but they typically also include cases in which data sufficient for a full evaluation of the existing system were unavailable. The design review team issues the Concern List as a set of issues that it feels should receive further study and that may require action of some sort by unit personnel. The Concern List may also contain preliminary recommendations of follow-up actions that should be considered for addressing the concerns on the list.

Keeping in mind these distinctions between the initial design of a new and the design review of an existing pressure relief system, the following sections summarize the typical workflow for a relief system design review.

Design Review Scope In general, the scope of any relief system design review should be the revalidation and documentation of the adequacy of all pressure relief systems on all equipment items in the facility. Such a review would involve evaluation of the pressure relief requirements of all equipment items in the facility. Such a thorough, equipmentbased analysis is the only way to ensure and to document that the overpressure protection needs of the entire facility are met by the existing pressure relief system. For existing facilities with no more relief system documentation than a set of relief device specification sheets, the documentation provided by such a complete review is expected by OSHA inspectors. There are a number of reasons for a relief system design review to have a more limited scope. Budgetary constraints may require that a facility’s needs for a pressure relief system review be prioritized in some way. If adequate design basis documentation is on hand, such prioritization is typically guided by the goal of overall risk-reduction. The risk-evaluation may be done in a variety of ways: according to relative age, or relative amount of alteration since initial construction or design review. For example, there is little benefit in reviewing a process unit that has not been altered since it was designed and built, regardless of how long it has been operating (in practice, however, such unaltered units are rarely encountered). Operating experience with the existing pressure relief system may be useful in prioritizing the allocation of design review resources. Another approach to prioritization may center on concern about the capacity of a relief collection system (e.g., a flare header). During the life of a process facility, relief devices protecting newly

November 2001

1200-106



1200 Relief Systems

installed equipment are usually tied into a flare system, without evaluation of the impact on built-up back pressures in global relief contingencies. For this reason, relief system design reviews are sometimes confined in their scope to relief devices and depressuring valves that discharge into a collection and disposal system. This limitation in turn limits the design review’s scope to equipment items that are protected by these relief devices. It is important to note that such limitations on the set of equipment items included in the design review’s scope result in design basis documentation that does not meet the requirements for process safety information set forth by the PSM and RMP regulations. Regardless of whether or how the scope of the design review is limited, it is essential that all interested parties agree on what the scope of the work is to be.

Data Collection The data required to perform a review of an existing pressure relief system is the same as that required to design a relief system for a new facility: •

Block Flow Diagrams Process Flow Diagrams (PFDs) with energy & material balance data Piping and Instrumentation Diagrams (P&IDs) Plot plans

•

Isometric drawings of relief headers, knock-out drums, and discharge or flare stacks showing all connecting piping back to relief devices, or other automatic inlet valves

•

Process simulations

•

Relief device data sheets Pump & compressor specification sheets and design flow curves; Tank specifications; Piping and fitting pressure ratings; Control valve data sheets, as needed

•

Vessel and heat exchanger fabrication drawings, as needed.

Most of these data elements come under the designation of process safety information according to the OSHA Process Safety Management standard and the EPA Risk Management Program rule. Accordingly, they should be accurate and up-to-date. For the purposes of the relief system design review, they really must be current. There is no benefit accrued from analyzing the pressure relief requirements of the process as it existed when originally designed or following the most recent debottlenecking project. Up-to-date PFDs, P&IDs, and plot plans should generally be readily available. Flare header isometrics should at least be checked against the P&IDs to ensure that all connections to it are shown; field verification and updating may be more efficient


1200-107

November 2001

1200 Relief Systems


and reliable. In some cases, the header drawing may need to be generated from scratch. It is a vital part of accurate modeling of the hydraulic performance of the relief collection system in common contingencies, and it provides concise documentation of a large part of the relief system. The process simulation - and hence the mass and energy balance data (flow stream properties) - should be checked against operating data to ensure it is current. If it is not current, it will generally need to be updated, as it is the source of all fluid properties for the relief system design review. If the process simulation is not current, the design review cannot be. Equipment specification sheets for an existing facility should be used with a small amount of caution. The concern, again, is whether the data is current. This concern reduces to a question of the diligence with which the facility’s process safety information has been maintained. If a pump’s impeller was replaced (not in-kind) at some point in the unit’s history, a new spec sheet and head vs. flow curve should have been inserted in the pump’s equipment and maintenance file. The same is true if a pressure vessel was re-rated, or if some other equipment change was made. Comprehensive field-verification of this data is generally both unnecessary and unwarranted; however, the person gathering the data should always be sensitive to potential inconsistencies in equipment files that may indicate information that is not current. Another concern with the use of specification sheets for existing equipment is that even if the equipment has not been changed, the data on the specification sheet may not be current. For example, if a pump’s suction pressure or the specific gravity of its working fluid has been changed, the pump will still develop the same head if its discharge is blocked, but in either case the dead-head gauge pressure will be different from that listed on the spec sheet or shown on the pump curve. Similarly, the flowing pressure drop through a flare tip shown on the tip’s spec sheet will typically be for the relief system’s original design flow rate. From that information, the relief system reviewer should be able to calculate a flow resistance factor (K) for the tip, which in turn can be used to calculate the pressure drop across the tip for the current design (or other) flaring contingency. The general point being made here is that in some cases the information on an equipment item’s specification sheet may not be invalid, but may nevertheless need to be scaled from the original to the current operating conditions. If there are concerns that the existing process safety information may be seriously inaccurate or insufficient, the need for field verification and/or retrieval of data from off-site sources (typically from equipment vendors) should be discussed and a proper course of action be chosen during the scope-definition step in the design review process. The design review documentation should contain a description of the sources of all input data for the design review, with a discussion of any verification procedures employed and any scaling to current process conditions performed.

November 2001

1200-108



1200 Relief Systems

Equipment List The generation of a list of equipment protected by the pressure relief system being reviewed should be a straightforward exercise. This list is generally created from the P&IDs, with consideration of any limitations in the scope of the design review.

Causes of Overpressure Just as for the design of a new pressure relief system, all potential causes of overpressure are identified and documented for all equipment items in the scope of the design review. See “Causes of Overpressure” on page 1200-99 for further discussion. In general, the analysis of the pressure relief requirements of the existing process equipment should proceed without regard for the presence of an existing pressure relief system - the equipment requires relief whether or not a relief system is present. The exception to this rule is that for some contingencies the presence of a relief device may specify the maximum upstream pressure assumed by the reviewer when evaluating the applicability of some causes of overpressure to downstream equipment. For example, when evaluating whether a control valve failure could overpressure a downstream vessel, the maximum pressure upstream of the control valve is compared to the vessel’s MAWP. In most cases if a relief device is present upstream of the control valve, the relief device’s set pressure is taken as the maximum upstream pressure. Similarly for the Heat Exchanger Tube Rupture contingency, the set pressure of a relief device on the high-pressure side of the exchanger may be used as the limiting pressure if it is less than the high-pressure side’s MAWP.

Pressure Relief Devices This step has no direct counterpart in the process of designing a pressure relief system for a new facility or unit. In conjunction with the identification and documentation of all causes of overpressure for specific equipment item, the reviewer should identify and note the relief device or devices that should provide pressure relief in each causes of overpressure. This is generally a straightforward process, as most equipment items have a relief device directly on them. However, there are cases in which this is not true: reboilers and condensers in distillation systems often are relieved via pressure relief valves located on the associated distillation column, and compressor discharges are often relieved downstream of a discharge cooler, perhaps even on a suction scrubber in a subsequent stage of compression. Any equipment for which no protective relief device can be identified should be documented clearly on the preliminary Concern List.

Required Relief Flow Rates For each causes of overpressure identified as applicable to an equipment item, the relief flow rate required to prevent the equipment from exceeding its maximum allowable pressure (e.g., MAWP plus allowable accumulation) is calculated. Fluid properties required for these calculations are taken from the process simulation, adjusted to relieving conditions. The methods for performing some of these calcula-


1200-109

November 2001

1200 Relief Systems


tions are briefly discussed in Section 1220; further discussion is given in API Recommended Practice 521. This step is nearly identical to its analog in the new relief system design process (Section 1262). However, subject to agreements made when specifying the scope of the design review, required relief rates are typically not calculated for equipment items for which no relief device could be identified (see Pressure Relief Devices above).

Relief Device Sizing and Rated Capacity Calculation When the required relief rates have been determined, the required size for each existing relief device is then calculated for each relief contingency with which the device is associated (see Pressure Relief Devices, above). This step in the design review process is analogous to “Required Device Size” on page 1200-100 in the design process for new facilities. Refer to that section for further discussion. At the same time, the rated capacity of each device is calculated for each relief contingency. Both the required size and the rated capacity of each size are documented for all relief contingencies. All relief devices calculated to have a rated capacity smaller than the required flow rate for any contingency for which it must provide relief should be noted on the Concern List, along with the relevant information concerning the relief contingency, the equipment protected, the required and rated capacities, the actual and required orifice sizes, etc.

Inlet and Outlet Pressure Drop Calculation After evaluating the capacity of each pressure relief valve at the overpressure appropriate to each relief contingency, the pressure drops across the inlet and outlet lines are calculated and documented. For these calculations, the pressure relief valve is assumed to be flowing at its rated capacity at the relevant overpressure, and the overpressure used should be the one appropriate to the relief contingency being analyzed. Note that this means the valve’s stamped capacity will not necessarily be used in this calculation (e.g., in a fire contingency). All relief valves calculated to have pressure drops in excess of the Code-recommended values should be noted on the Concern List, along with the relevant information concerning the relief contingency, the equipment protected, the line (inlet or outlet) involved, the calculated pressure drop, etc. Note that the design review of relief devices discharging directly to atmosphere is complete at this point in the design review process.

Simultaneous Reliefs For each relief discharge collection system, regardless of its disposal method or location, identify all contingencies that could result in simultaneous discharges from multiple equipment items into the collection system. One exception to this rule involves the fire relief contingency. For this eventuality, it is unnecessary to identify and document every possible combination of two or more equipment items that

November 2001

1200-110



1200 Relief Systems

can be relieved into a collection system in the event of an external fire. Rather, the effort should be directed toward using the agreed-upon default fire area to identify fire contingencies that will involve large numbers of vessels, or that will cause large volumes to discharge into the collection system, or that will cause discharges into sections of the collection system with large flow resistance (e.g., narrow piping, large numbers of fittings, etc.). This step in the relief system design review process is similar to the analogous step in the design process for a new pressure relief system; see “Simultaneous Discharge From Multiple Sources” on page 1200-102 for further discussion. An important difference in a review is that years of operating experience are available as an additional resource in the process of evaluating potential global relief contingencies. The knowledge of experienced unit operators and electricians should always be called upon for this task.

Collection System Model The first task in this step is to field-verify, to update, or to generate from scratch an isometric drawing of each relief discharge collection system within the scope of the design review. This drawing should depict all automatically-operated inlets into the collection system (pressure relief devices, control valves, and emergency depressuring [“blowdown”] valves), as well as all piping and fittings connecting these inlets to the collection headers, and all piping, fittings, and equipment in the headers themselves. The drawing should also show pipe diameters and lengths. The preparation and/or verification of this drawing is generally begun early in the design review process, in parallel with the steps discussed above. The current collection system(s) isometric drawing is then used in preparing a current computer hydraulic model of the collection system(s). This step is analogous to that described in “Collection System Model” on page 1200-102 for the design of a new pressure relief system. Each piping segment is represented in the model with its diameter, elevation change, length, and the flow resistance of the valves and fittings it contains. Flow resistance information is typically obtained from Crane Technical Paper No. 410, API Recommended Practice 521, and Perry’s Chemical Engineers' Handbook. Any equipment items in the collection system, from scrubbers and knockout drums to seals and flare tips, are also included in the hydraulic model, typically using manufacturers' data. Note that the manufacturers' data will generally be available only for the design flow case of the unit or facility as originally constructed. As discussed above in “Data Collection” on page 1200-107, this original flow data (e.g., pressure drop across the flare tip) will need to be scaled to current operating and common contingency conditions. Once “constructed” this model of the relief collection system can be used to evaluate the hydraulic performance of the existing collection system in each of the common contingencies identified in Simultaneous Reliefs, above.

Common Contingency Hydraulic Modeling The hydraulic performance of all relief discharge collection and disposal systems is now evaluated in the event of each global relief contingency identified in Simultaneous Reliefs, using the computer model assembled in Collection System Model.


1200-111

November 2001

1200 Relief Systems


For an individual common contingency, data is entered for each inlet to the collection system that will be flowing in that contingency: flow rate, composition, pressure, temperature, fluid phase, etc. The hydraulic modeling program then uses equations for energy, mass, and momentum balance to solve for the pressure drop developed across each segment of pipe and fittings in the collection system as well as the fluid properties at the junctions between each segment. These results are then used to evaluate the performance of the collection and disposal system. The back pressures calculated to be developed at each collection system inlet are examined to evaluate their potential effect on the operation and flow of the device at that inlet, using the criteria discussed in Section 1255. The pressures, temperatures, and flow velocities calculated in each segment of the collection system are compared to the specifications for the pipe and fittings in those segments to ensure that conditions outside of specifications are not expected at any point in the collection system for any relief contingency. Finally the performance of individual equipment items in the collection and disposal system is evaluated: for knockout drums, the liquid-vapor separation performance and the liquid hold-up requirements are analyzed for contingencies involving large liquid releases. The thermal radiation generated by flares is calculated at locations to which personnel have access, and at the location of the closest equipment or structural item. These values are compared to standard values of acceptable radiation levels published in API Recommended Practice 521 and elsewhere. Relief and depressurizing valves predicted to have developed back pressures that exceed maximum recommended values in any common contingency are noted in the Concern Identification List, with the relevant identifying information, the contingency involved, the predicted and recommended maximum back pressures, etc. Any piping segments predicted to have conditions outside of the design specification limits should similarly be noted in the Concern List, along with concerns about the performance of knock-out drums or other separation equipment or about predicted flare radiation levels.

1272 Relief System Design Review Report The design review report should satisfy the two-fold objective of the review: documentation of the current design basis of the existing pressure relief system and identification of any potential discrepancies between the performance of the system and the performance recommended or required by established industry standards and recommended practices. Accordingly, the design review report should generally contain the following elements: 1.

Statement of Scope

2.

Summary of data sources –

November 2001

Including version numbers and revision dates for PFDs, P&IDs, and process simulations

1200-112



3.

4.

5.

1200 Relief Systems

Summary of engineering calculation methods a.

Industry and Chevron Corporate Standards and Design Codes used

b.

Engineering assumptions used (e.g., maximum height above grade heated by an external fire)

c.

Equations and methods used (e.g. API RP 520 equations for relief device sizing for liquids and for vapors)

For each Equipment Item: a.

Contingency Identification Form

–

Describing all applicable causes of overpressure and explaining the reasons that any common contingencies may have been determined to be inapplicable in a given case

b.

Equipment Information Form

–

Describing the equipment item and referring to the data sources for the item (e.g., specification sheets, design drawings, ASME U-1 reports, pump curves, etc.)

c.

Contingency Detail Form (Required Rate Calculation)

–

Indicating standard calculation methods used. If some other method is used, provide a complete description of that approach.

For each Relief Device: a.

Relief Device Descriptive Data

–

Including materials of construction, and referring to data sources for the device (original specification sheet, recent inspection report, etc.)

b.

Sizing Data

–

Including relief stream composition, temperature, pressure, etc.

c.

Inlet and Outlet Line descriptions and pressure drop calculation

–

Listing all pipe sizes and lengths and all fitting types and sizes

6.

Isometric Drawings of each Relief Collection System

7.

List and description of each potential global relief contingency

8.

For each global relief contingency: a.

List of all devices expected to be flowing, with a list of each flow stream’s properties (composition, pressure, temperature, quality, flow rate, etc.)

b.

Back pressure predicted at each flowing device

c.

List of extreme values of temperature and pressure predicted in the collection system


1200-113

November 2001

1200 Relief Systems

9.


d.

Summary of the performance of each flare knock-out drum, including the minimum drop size separated and the maximum contingency duration that can be accommodated by the drum’s available hold-up volume

e.

Summary of the flare thermal radiation values predicted at accessible locations near the flare

Copies of all electronic data files and software used in the design review –

To enable local facility personnel to maintain the relief system design basis information as changes are made to the system after delivery of the design review report

10. List of concerns identified by the design review as requiring discussion and possible further action. –

Including preliminary recommendations for corrective actions for these concerns

1273 Management of Change Follow-up investigations of process industry incidents have traced a significant fraction of these incidents to changes in the process: unexpected consequences, out-ofdate operating procedures, unauthorized changes, etc. For this reason, many facilities began to develop and implement procedures for the formal review, authorization, and documentation of changes to the facility’s processes. In separate regulations directed toward minimizing accidental releases of highly hazardous substances, both OSHA and EPA chose to include a requirement for management systems to ensure such procedures were instituted at all facilities handling these substances. The requirements for such “management of change” (MOC) procedures are formally set forth in OSHA’s Process Safety Management standard (29 CFR 1910.119) and in EPA’s Risk Management Program rule (40 CFR 68). While most process industry facilities have already implemented systems for MOC, many are still striving to improve the efficiency and effectiveness of their MOC procedures (see, for example, West, et al., and Eastman and Sawyers). The several goals of a management of change procedure are specified in paragraph (L) of the OSHA PSM standard: the procedure should ensure that the proposed change has been considered from the viewpoint of its technical basis, its implications for safety and health of facility personnel, its implications for operating procedures, the time required to implement the changes, and the authorization requirements for the change. Furthermore, the MOC procedures must ensure that the facility’s process safety information (PSI) and operating procedures are updated to reflect any authorized changes, and that all facility personnel whose work activities are affected by the change are informed of and trained in the changes before the changes are implemented. The requirement for MOC procedures does not apply to changes that can be classified as “replacements-in-kind” – the replacement of one equipment item with a different one meeting the same design specifications. Rather, the requirement applies to actual changes in the operation of the process. Examples of such changes

November 2001

1200-114



1200 Relief Systems

include the addition, removal, or alteration of equipment, a change in a process control scheme, or a change in operating or maintenance procedures. The need for procedures for management of change is particularly relevant for pressure relief systems, because a change’s impact on the relief system is often overlooked. This oversight can be traced to several causes. First, a relief system is typically thought of – with some justification – as an auxiliary system. While a change in one part of the process may have a direct impact on other parts of the process, the change typically has only an indirect affect on the relief system. Unless one deliberately looks for possible relief system repercussions, one is unlikely to see them. In the general sense, of course, this is why MOC procedures are required by the US federal regulations – to look for all possible repercussions of the change. The second reason that relief system impacts are often overlooked is that many of the plant personnel performing reviews are not very experienced with identifying all potential causes of overpressure. Most process engineers are familiar with the need for pressure relief in case of external fire, closed outlets, and thermal expansion. But relatively few have the expertise with evaluating an equipment item’s overpressure relief requirements that comes from designing new or reviewing existing pressure relief systems. Without a resource such as API Recommended Practice 521 or the summary presented in Section 1220 of this document at hand, many causes of overpressure are easily overlooked. Finally, the impacts of a process change are not only indirect and somewhat subtle, they can also be pervasive: the repercussions do not stop with the new or altered causes of overpressure or required relief loads. In fact, the impact goes beyond the relief devices and their inlet and outlet lines, to the performance of collection headers and flares. Unless the relief collection and disposal system is re-evaluated to ensure adequate performance in the wake of a process change, the full potential for the change’s impact on the pressure relief system cannot be known. The following sections present a brief, step-by-step procedure for considering the potential impact on a facility’s pressure relief system of any change that requires the MOC process.

Identify the Scope of the Change The purpose of this step is to generate a list of the equipment items involved in the change. Such involvement can occur in two ways. The first is through the addition, removal, or alteration of the equipment itself. The second is through the alteration of the process conditions within an equipment item, with or without a change to the item itself. The first mechanism of change is slightly easier to discern because it involves a direct affect: we directly change one or more equipment items. Changes in the process conditions are more indirect, and may occur far downstream of the location at which the direct change (e.g., in an instrument set point) was made. How does one identify all the equipment items affected by a change? First, start at the location of any direct change: an equipment or instrument addition or modification, a change in operating procedures, etc. Process equipment items that are to be added, removed, or altered will clearly be affected by the proposed change. However, other vessels, etc., are likely to be affected as well. To determine which these are, ask the question “In which equipment items will there be a change in


1200-115

November 2001

1200 Relief Systems


process conditions (such as pressure, temperature, flow rate, level, composition, etc.)?” On a P&ID, begin at the location of the change, and proceed both upstream and downstream, evaluating the process conditions assuming the proposed change has been implemented. When an item is reached in which the new process conditions will be identical to the old, the limits of the change’s impact have probably been reached. Then ask “Can the change have any implications further (up- or) downstream? It is important that all equipment inlets and outlets be followed in this search for the limits of the change’s impacts. A process simulation, updated to include the proposed change, can be very helpful in this task.

Conduct a Design Review of the Pressure Relief System within the Scope of the Change The product of the task just described is a list of the equipment items that would be affected – both directly and indirectly – by the proposed change. The next task is to conduct a pressure relief system design review of these equipment items, assuming the proposed change has been implemented. This design review should be carried out in the manner described in Section 1271. The scope of this “mini-review” is the list of equipment generated as described above (Identify the Scope of the Change). The input data for the mini-review should assume the proposed change has occurred. The following discussion includes occasions during the MOC relief system minireview in which the design basis documentation is required to be updated to reflect the proposed change. Because this design review is conducted before the change has been authorized or implemented, the documentation modified during the review should be a copy of the facility’s relief system design basis documentation. Only after the changes have been authorized and implemented should the original design basis documentation be updated permanently. 1.

Causes of Overpressure For each equipment item in the scope, identify all causes of overpressure. If the proposed change has generated new potential causes of overpressure, identify an existing relief device to provide protection. If none exists, make a note that the proposed change will require the design and installation of a new one. If existing causes of overpressure are no longer applicable due to the proposed change, update the documentation accordingly.

2.

Required Relief Rates For each equipment item in the scope, update the calculation of the required relief flow rate, as appropriate for the particular change. For example, if the change results in an increase in the operating pressure in a reciprocating compressor’s suction scrubber, the compressor relief load for the blocked outlet contingency will need to be recalculated, because the flow rate is a function of the suction pressure. In this example, however, the relief load for the scrubber’s fire contingency may not need recalculation, because the required flow rate is a function of only the relief pressure and the contained liquid’s composition and level.

November 2001

1200-116



3.

1200 Relief Systems

Relief Devices For each existing or new relief device in the scope of the change, calculate and document the device’s required size and actual capacity. For existing devices, compare these values with the device’s actual size and required capacity. Note any devices that are calculated to be undersized for the proposed new operating conditions. For new devices proposed as part of the change, note the required size. Select a discharge disposal method, as discussed in “Choice of Disposal Method” on page 1200-65 and “Discharge Fluid Disposal Method” on page 1200-101. Develop a preliminary design for the device’s inlet and outlet lines.

4.

Device Inlet and Outlet Lines Calculate and document the pressure drops through the inlet and outlet lines of pressure relief valves when flowing at the valve’s rated capacity in each contingency. Note any existing lines whose calculated pressure drops exceed recommended values, per Section 1243, “Pressure Relief Valve Installation” and Section 1244, “Rupture Disk Installation.” Adjust the preliminary design and sizing of new lines to meet recommended pressure drop values.

5.

Discharge Collection Systems Assuming the proposed change is implemented, will the relief discharge fluids be compatible with the existing collection systems? If not, consider the need for a new or modified collection system. Will the change require installation or removal of relief devices that connect to a collection header? If so, update the collection system drawings and computer models to reflect these changes.

6.

Global Relief Contingency Modeling New common contingencies are likely to be created only by a very significant change: one that involves addition of many vessels - creating a new global fire contingency - or one that requires the addition of a new electric service buss. However, existing global relief contingencies are likely to be affected by most typical process unit changes: addition of a new relief device that ties to the collection system, or any change in the relief flow rate required to flow through any existing relief device. The existing common contingencies should be reviewed to identify which will be affected by the proposed change. The data describing the streams that flow into the collection system in these contingencies should be updated, and the hydraulic performance of the collection and disposal system should be reanalyzed. The built-up back pressures at all flowing devices and the discharge stream properties in collection system should be evaluated relative to the design values. Flare thermal radiation values should be calculated and compared to acceptable values.


1200-117

November 2001

1200 Relief Systems


MOC Relief System Design Review Report In the case of significant process changes that may have a sizable impact on the unit’s pressure relief system, a brief report of the results of the mini-review will be desirable. If so, it should have the same structure and contain the same data elements as a full design review report, as described in Section 1272.

1280 Relief Valve Testing This section provides recommendations for relief valve inspection, test frequencies, procedures, and documentation, and supplements the American Petroleum Institute’s Guide for Inspection of Refinery Equipment, Chapter XVI, “Pressure Relieving Devices.” In this section, the word “should” denotes a recommendation and the word “shall” is mandatory. This section is to be used for background information for testing relief valves. Actual valve disassembly, shop inspection, repairs, settings, and reassembly shall be performed by trained and qualified personnel. Before disassembling or testing a valve, review manufacturers’ warranty requirements and state and local codes. Some states require that certain valves be tested and repaired by qualified personnel of professional organizations, such as ASME.

Purpose for Testing Routine inspection and testing ensures that valves will operate when pressure or vacuum exceeds the design rating of equipment and/or piping. Properly maintained and operating valves: 1.

Protect personnel.

2.

Prevent damage to capital investments.

3.

Prevent plant downtime due to overpressure and/or excessive vacuum accidents.

4.

Conserve material losses from a leaking valve.

5.

Reduce fugitive emissions from a leaking valve.

Definitions Definitions used in this section are from the American National Standard Institute B95.1, Terminology for Pressure Relief Devices. Back Pressure. Static pressure existing at the outlet of a pressure relief device due to pressure within the discharge system. Blowdown. Difference between actual popping pressure of a pressure relief valve and actual reseating pressure. This value is expressed as a percentage of set pressure or in pressure units. Blowdown Pressure. Value of decreasing inlet static pressure at which point no further discharge is detected at the outlet of a resilient disk-type safety relief valve.

November 2001

1200-118



1200 Relief Systems

This pressure is measured after the valve has been subjected to a pressure equal to or above the popping pressure. Chatter. Abnormal, rapid reciprocating motion of the movable parts of a pressure relief valve in which the disk contacts the seat. Closing Pressure or Reseat Pressure. Value of decreasing inlet static pressure at which the valve disk re-establishes contact with the seat or at which lift becomes zero. Cold Differential Test Pressure. Static pressure at which a pressure relief valve is adjusted to open on the test stand. This test pressure includes corrections for service conditions of back pressure and/or temperature. Leak Test Pressure. Specified inlet static pressure at which a quantitative seat leakage test is performed in accordance with standard procedure. (This manual will use API Standard 527 as the procedure for identifying leaks.) Lift. The travel of the disk away from closed position when a valve is relieving. Opening Pressure. Value of increasing inlet static pressure of a relief valve at which there is a measurable lift or, at which the discharge becomes continuous. This state can be determined by seeing, feeling, or hearing. Popping Pressure. Value of increasing inlet static pressure at which point disk moves in the opening direction at a faster rate compared with a corresponding movement at higher or lower pressures. This value applies to the safety relief valves in compressible fluid service. Set Pressure. Value of increasing inlet static pressure at which a pressure relief valve displays one of the operational characteristics. These are defined under “opening pressure,” “popping pressure,” or “start-to-leak pressure.” Start-to-Leak Pressure. Value of increasing inlet static pressure at which the first bubble occurs when a resilient disk-type safety relief valve is tested by means of air under a specified water seal on the outlet.

1281 Test Equipment Relief Valve Testing The equipment used for relief valve testing discussed in the subsection “Shop Testing” of Section 1282 must be designed to meet each facility’s specific needs. One typical design is shown in Figure 38 of the API Guide for Inspection of Refinery Equipment, Chapter XVI, “Pressure Relieving Devices.” This design, however, does not meet all needs. To make the best choice for a particular facility one needs to answer several questions, including: 1.

Are you going to test for full valve lift as well as blowdown pressure? (These tests will increase the volume of compressed air required to do the testing.)


1200-119

November 2001

1200 Relief Systems


2.

What valve orifices and set pressures will you want to test? (These factors will also affect the test bench capacity requirement and the range of pressure measurements required.)

3.

How many valves do you have to test? (If there are many valves to be tested, the bench design may require more than one station, or it may be cost effective to purchase a hydraulic clamping bench.)

4.

Do you have pilot operated valves to test?

5.

Does the test stand need to be portable?

6.

Will you need to test liquid relief valves? (Liquid relief valves must use liquid as the test medium.)

These and other site-specific questions will help determine the requirements for the test bench. Based on the answers to these questions, one can decide between commercial or custom test benches.

Commercially Available Test Benches Relief valve test benches with a variety of features are available through manufacturers and specialty test stand manufacturers. Some of the main features offered are described below. Valve Mounting: The most frequent way to mount valves on the test stand is by using adapter flanges to resolve differences between the valve and the test stand’s flange size, type, and rating. Each time a new valve is tested, the appropriate adapter flange must be used and the valve must be bolted to the flange. An alternative offered by vendors is to use a hydraulic clamp to secure the valve to the test bench. The only adapters required on these systems are for screwed valves. Bolt-up time is eliminated. Systems using hydraulic clamping are usually more expensive than those using adapter flanges. Hydraulic clamps save testing time, however, and may be justified when there are many valves to test. System Capacity: Most commercial systems (but not all) are designed to test only the relief valve’s set pressure. They do not have sufficient capacity to achieve full valve lift and to determine the blowdown pressure. Be sure that the bench you purchase has sufficient capacity for the parameters to be tested. Instrumentation: Test bench instrumentation ranges from standard test pressure gages to digital gages with peak-hold features, as well as data logging computers. Your needs will determine what is appropriate. For systems designed to test only set pressure, standard test gages are adequate. For those systems designed to also measure blowdown pressure, a digital peak-hold (max/min) gage is desirable. Data logging systems provide a hard copy of the test measurements and can also be used to archive test data. One must also decide whether a hoist is needed; what size compressor is appropriate; if air drying equipment is needed; and what work space is required (i.e., space available vs. size of the accumulator pressure vessels). The cost range can be

November 2001

1200-120



1200 Relief Systems

from several thousand dollars to $50,000 or more, depending on the sophistication of the test bench design.

Test Bench Designs The purpose of this section is to outline the principal features of relief valve testing equipment. Drawings of relief valve test equipment from various locations are available from the M&CS Unit Library of CRTC under technical file number 720-J-0323. (It should be noted that these existing designs are used only for checking the relief valve set pressure and thus do not have the capacity to check blowdown.) Relief valve testing discussed in the remainder of this section consists of determining the popping and reseat pressures. The testing equipment consists of an air compressor, accumulator vessel, test vessel, and a suitable test manifold. (See Figure 1200-46.) Each test bench should be designed for a particular application, taking into account the questions asked above. Fig. 1200-46 Schematic of a Typical Relief Valve Test Stand

The maximum valve orifice and set pressure to be tested are very important. If there is a small number of “large capacity” valves, the test bench may be designed to test only the set pressure for these valves. Once the maximum valve capacity to be tested is determined, the size, capacity, and pressure rating of the various vessels and the compressor can also be determined. The compressor capacity is based on the time to pressure the accumulator up to its operating pressure (probably from 5 to 15 cfm). If the installation requires a threestage compressor, or if the atmospheric air has high humidity, an aftercooler should be used and an automatic drain trap installed on the accumulator. At compressor


1200-121

November 2001

1200 Relief Systems


discharge pressures below 1000 psi and with relatively dry air, this auxiliary equipment is not justified. Any moisture in the accumulator can be blown down manually. The accumulator must be of sufficient size and pressure rating to pressure up the test vessel to the valve set pressure. It must also have sufficient capacity to assist the test vessel in achieving full valve lift on larger orifice valves. One or two 10 cubic feet vessels are typical. However, the vessel volumes are a function of the pressure rating. The higher the pressure of the accumulator, the smaller the volume required. For example, one valve testing company uses Scuba Bottles filled to several thousand pounds as their accumulator. As a design starting point, a pressure 33% higher than the maximum relief valve set pressure is sufficient. The test vessel must be rated at least 25% over the maximum relief valve set pressure and must not restrict flow to the relief valve. The vessel volume required is related to the volume of the accumulator and its operating pressure. When used with two 10 cubic feet accumulators operating at 33% over the maximum valve set pressure, a test vessel of about 6 cubic feet is a good design starting point. The test manifold is discussed in the next section. To facilitate the handling of large valves, a davit or jib crane should be considered.

Pressure Gages Precision (½% of full scale graduations, ¼ of 1% accuracy) test gages should be used on all test benches. When analog test gages are used, the set pressure being tested should be in the middle one-third of the range. These gages should be tested for accuracy once every 2 weeks using a dead weight tester. Digital gages must also meet the precision requirements above. For digital gages (similar to those manufactured by Heise), the calibration frequency may be relaxed significantly compared with analog test gages. There are not many mechanical linkages to affect long term accuracy. One valve manufacturer requires the testing of digital gages once every 90 days, although experience will determine the required frequency of testing. In addition, digital gages also offer peak hold features that make them cost effective.

1282 Inspection and Test Procedure Frequency The frequency of relief valve testing and inspection is determined by operating Company procedures as well as by national, state, or local codes. Many operating companies have regional procedures that take into account national and local codes as well as their experience in setting inspection and testing frequencies. The regional engineering office or plant engineer should be consulted for questions about frequency of testing. If guidelines are not set, the conservative inspection and test frequency in this manual can be used. As a minimum, federal, state, and local codes shall determine the frequency of relief valve inspection and testing. It is not the intent of this manual to list all the codes

November 2001

1200-122



1200 Relief Systems

relating to relief valves. Common codes that apply to most of the Company operating facilities are listed below. Pipelines and Pressure Limiting and Regulating Stations: Code of Federal Regulations, Title 49-Transportation, Articles 192.729, 192.739, 192.745, and 192.749. Compressor Stations: Inspect and test at least once each calendar year. At intervals not to exceed 15 months. Pressure Limiting and Regulating Stations: Test in place at least once each calendar year at intervals not to exceed 15 months. Installations in Federal Waters: Department of Interior, Federal Register, Vol. 53, No. 63, Appendix 1, Register No. 250.124, Production Safety Systems Records, requires that relief valves shall be tested every 12 months. Boilers: National Board Inspection Code, Boiler and Pressure Vessel Testing, requires the following: •

Low pressure boilers (steam pressure or relief valve set less than 15 psig; water temperature less than 250°F, or boiler feed water pressure less than 165 psig): relief valves are to be inspected once a month by operating the lifting lever. Each valve shall be tested once a year.

•

Power boilers less than 400 psig: inspect relief valves each month by operating the valve lifting lever. The valve shall be tested once a year.

•

Power boilers greater than 400 psig: test based on operating experience.

Other considerations for determining inspection and test frequencies are: •

Valves with plugging problems or with a high failure rate on the “as-received” relieving pressure test should have decreased periods between tests. Conversely, valves without failures and/or in clean service may have the length between tests increased.

•

After an extended plant shutdown, all valves should be tested before plant startup.

•

Testing should be coordinated with scheduled plant shutdowns when possible.

•

Testing should conform to the information or recommendations of the relief valve manufacturer.

•

When constructing new facilities, the relief valves should be tested and installed before the plants are commissioned. Relief valves are precision instruments. They should be stored in a clean location and carefully installed at the last possible moment to avoid exposure to construction activities. To prevent construction delays, it is recommended that spool pieces be specified and installed in the piping during the construction and hydrotesting periods.

Recommended inspection and test frequencies should be developed and adjusted by refineries according to historical test records, regional requirements, and shutdown schedules.


1200-123

November 2001

1200 Relief Systems


Visual Field Inspection—Before Valve Removal The following are recommended items to check at the annual visual inspection: 1.

The correct valve is installed and valve number agrees with the P&IDs.

2.

Seals on all the valve external adjustments are in place. This is ASME Code requirement.

3.

All valve vents are clear.

4.

All block and bypass valves are in the correct position, either fully open or fully closed. Block valves and bypass valves are car sealed or locked open.

5.

Gags and blinds are removed or flagged with operator and maintenance tags. Before untagged blinds are removed, the operations and maintenance groups should be contacted.

6.

The external surfaces are free of cracks, corrosion, and mechanical damage.

7.

Drain holes in the discharge piping elbows are clear, to prevent liquids from settling.

8.

Valve discharge piping that vents to the atmosphere is clear of obstructions. Remove all bird and insect nests.

9.

There are no flange or relief valve leaks. A tight seal can be checked by applying a soap solution on the valve flange and screwed connections. For valves that relieve to the atmosphere, the discharge piping opening should be monitored with an infrared gun. In hydrocarbon service (i.e., a pipeline), a hydrocarbon detector can be placed at the discharge piping outlet to monitor a leaking valve. Checking for hydrocarbon emissions is a requirement of Federal Regulation 40 CFR, Part 60. The frequency of inspection is set by local agencies.

Field Safety Before removing a valve from the process, the following safety precautions shall be observed:

November 2001

1.

Operators shall be notified when the relief valve needs to be removed or isolated with its block (stop) valve or blinds. Operations will want to ensure that the process is stable or shut down before the valve is removed.

2.

When a relief valve is protecting an ASME Code stamped vessel and the relief valve’s stop valve needs to be closed, the code (ASME Boiler and Pressure Vessel Code, Section VIII, Division 1, Appendix M, M-5) requires that the inlet and outlet stop valve be closed only by an authorized person. This person should remain stationed at the valve until the stop valve is opened, sealed or locked.

3.

If the relief valve discharges into a plant header, the valve outlet should be isolated from the header with the stop valve and a blind.

1200-124



1200 Relief Systems

4.

Space between the relief valve and its stop valve or blind should be vented to a safe location. The purpose of this arrangement is to release the trapped gases or liquids and determine if the stop valve is providing a good seal.

5.

If work is to be performed on a relief valve in place, blinds should be inserted between the stop valve and relief valve.

6.

Protective clothing and equipment shall be worn when removing valves from hazardous service. Personnel shall be trained and certified for the protective equipment (i.e., Scott air packs) before use. Plant and regional safety rules shall be observed.

7.

Valves shall be chemically neutralized before transporting. Valves with pyrophoric iron should be kept moist. For relief valves protecting vessels, wait until vessel is steam cleaned before removing valve.

Visual Field Inspection After Valve Is Removed Inlet and outlet relief valve piping and valve nozzles should be inspected for thinning, corrosion, and deposits that could cause plugging and/or seat damage. All deposits should be removed from the piping before the valve is returned to service.

Transporting Relief Valves Relief valves, despite their size, are fragile. Careful handling when transporting is important. Rough handling can change the pressure setting and/or deform valve parts and thereby prevent proper setting of the valve set pressure. Damage can also occur to the seats, preventing a good seal. Operating company experience has revealed that a valve recently serviced, set, and tested can leak or not operate properly because of rough handling from the shop to the field. Relief valves are used for protecting personnel and equipment; it is important that they function properly. The following are guidelines for transporting valves. Secure and transport valves in the upright position to prevent abrasion between the disk and the inlet nozzle. Protect the inlet and outlet openings with covers to keep valve internals clean. Protective covers prevent the valve from being installed without removal of the covers. Cover threaded parts to prevent damage. (See Figure 1200-47.)

Shop Inspection Shop inspection includes visual inspection to determine the “as-received” relieving set pressure, and valve cleaning and inspecting requirements.

Visual Inspection The following items should be inspected and recorded before the valve is disassembled: •

Flanges should be free of pitts, smooth, flat, and the gasketed surface area should not be reduced. Screwed ends should have clean unobstructed threads.

•

All bonnet vents should be clear of obstructions.

•

The blowdown ring setting should be noted and recorded.


1200-125

November 2001

1200 Relief Systems


Fig. 1200-47 Recommended Method of Transporting Relief Valves

•

Flame arrestors on breather valves should be cleared of all fouling and plugging material.

Determining the “As-Received” Relieving Pressure After the initial shop inspection, the “as-received” relieving pressure should be determined. This information is used to build a historical record for the valve. From this data test frequencies can be determined and information gained on how the process is affecting the valve. The “as-received” relieving pressure is determined on the test block. Slowly increase the test pressure (and vacuum for tank breather valves) at the relief valve inlet while monitoring the valve inlet pressure gage and noting the pressure at which the valve opens or relieves. •

If the valve opens at the set pressure, record the opening pressure.

•

If the valve opens at a pressure (vacuum) higher than the set pressure (vacuum), repeat the test. It may be that the valve was stuck. After the opening pressure (vacuum) is determined, record the information.

•

If the valve opens below the set pressure, the valve spring may be weak (the weight may need adjusting), or the setting was changed. Record the opening pressure.

After the opening or the relieving pressure (vacuum) is determined, remove the test pressure at the valve inlet. Put 35 psig at the valve outlet and check the bonnet and seats for leaks with a soap solution. A lower pressure (for example, 5 psig) should be used on breather valves.

November 2001

1200-126



1200 Relief Systems

If there are leaks around seats, report this to the operating group. This information is useful in identifying continuous emissions to common relief headers or sources of product loss. Cleaning and Checking Parts. After determining the “as-received” test pressure, proceed to dismantle the valve. Valves should be cleaned, degreased, and chemically cleaned to remove all rust, grease, and deposits. Use the following as an inspection guide. Seats and Disks. Seats should be checked for flatness and compared against the valve manufacturers’ tolerances. If lapping of the seats is required, the lapping should be performed with precision machinery (such as a Lapmaster) in a temperature controlled room. The seating surfaces should be lapped to an optical flatness of 10-15 millionths of an inch (measured with a monochromatic light) and surfaces finished to at least 8 micro inches. The seating surfaces should then be polished to a high mirror finish. The fit between the guide and the disk or disk holder should be inspected for scoring. The seats, disks, and nozzles should also be inspected for roughness and damage. Spring. Inspect the spring for corrosion, cracking, or deformation. Spring cracks can be revealed with a dye check. If the valve has opened below the set pressure in the “as-received” test, a spring test should be performed. The permanent set of the spring (defined as the difference between the free height before compression and that height measured 10 minutes after the spring has been compressed solid three times, after presetting at room temperature) shall not exceed 0.5% of the uncompressed free height. This standard is defined by ASME Section I, PG-73, 73.1.2 and ASME Section VIII-Division 1, UG-136, (a)(2). If there is a need to change the valve set pressure, the valve manufacturer should be consulted to determine if the new setting is within the range of the existing spring. In general, for set pressures less than or equal to 250 psig, the existing spring shall not be used for any pressure greater than 10% above or 10% below that for which the valve is marked. For set pressures greater than 250 psig, the existing spring shall not be used for any pressure 5% above or 5% below that for which the valve is marked (see ASME Section I, 72.3). Check the relief valve test records or valve data sheet to verify that the correct spring is installed. If there is a discrepancy between the number on the spring and the number in the relief valve data sheet, the plant/area engineer should be notified for a resolution. Bellows. Check for cracks with dye. Examine bellows for thin spots. Body. Measure and record body thickness. The objective is to track the valve wear. Check for body cracks by magnetic particle, magna-glow or magna-flux. Examine all internal gasketed surfaces to determine if they are smooth.


1200-127

November 2001

1200 Relief Systems


Stem. Check the stem to confirm that it is straight per manufacturer tolerances. Visually inspect the stem guides to verify that they are free of cracks, pits, corrosion, and galling. Threads. Inspect all threaded parts and repair as needed to provide full thread engagement. Check Valves. Inspect check valves to verify that they will allow flow in only one direction. Check valves are commonly found in the pilot tubing on pilot-operated relief valves. Replacement parts should be purchased from valve manufacturers. There have been reports of high failure rates for valve replacement parts. This has prompted the manufacturers’ assemblers to test all incoming parts. With the testing and new quality control programs developed by valve manufacturers and their suppliers, there has been a significant improvement in reducing new part failures. If your current suppliers are providing good parts, continue using their services. New operating company test/repair facilities should replace the worn valve parts with original factory parts from the valve manufacturer. Improper Valve Performance. The following are common valve problems: Leaking Valve Causes: 1.

Dirt (scale, weld slag, corrosive deposits, coke) trapped under seats. This is often caused by testing a relief valve in place.

2.

Corrosion due to improper material compatibility with the process.

3.

Valve chatter. Chatter is caused by undersized inlet piping to the valve, an oversized valve, or set pressure too close to the operating pressure.

4.

Excessive pipe strain on the valve.

5.

Improper alignment of the spindle.

6.

Improper fitting of the spring to spring washer, or improper bearing between spring washers and their respective bearing contacts between the spindle and disk or disk holder.

Solutions:

November 2001

1.

Install a rupture disk between the inlet piping and the valve. (This can be expensive because piping alterations are often required.)

2.

Check the valve seat and disk materials to see if they are compatible with the process fluid.

3.

Check the valve set pressure against the process operating pressure. Perhaps the valve is set too close to the normal operating pressure of the process. Note that the set pressure shall not exceed the maximum allowable working pressure of the equipment or piping that the relief valve is protecting.

1200-128



1200 Relief Systems

4.

Check the valve size calculations and data sheet to verify that the valve is not oversized.

5.

Check the piping to verify that the piping loads are handled by the pipe supports and not the relief valve.

6.

Inform maintenance personnel of the importance of proper relief valve handling.

Cracked spindles and springs are often a result of general corrosion due to improper compatibility with the process as well as stress corrosion. Possible solutions for these problems are: (1) use material that will resist the corrosive action of the process; (2) isolate the spindle and spring from the process by using a bellows, an O-ring seal on the spindle guiding surfaces, or install a rupture disk at the relief valve inlet connection; (3) repair leaking valves. Cracked bellows are a result of general corrosion due to improper material compatibility with the process or the back pressure exceeds the bellows design. Solutions for cracked bellows are: (1) select a bellows material that is compatible with the process; (2) when bonnets are vented to a closed system, examine the types of gases in the closed system; (3) specify and install bellows that are designed to withstand the normal and operating back pressures; and (4) compare the superimposed and the developed back pressure against the design pressure of the bellows.

Shop Testing After the valve is cleaned, inspected, and reassembled, it is placed on a shop test block and tested to verify the proper opening pressure, blowdown, seat and bonnet tightness. If the relief valve will be installed in an orientation other than vertical, it must be tested in the same position in which it will operate.

Test Media The test media shall be as follows: Media

Service

Steam or air with temperature correction (ref. ASME Section 1, PG-73.4.2)

ASME power boilers

Water (ref. ASME Section VIII-Div. 1, UG-136, (d)(4))

Liquid service valves on ASME pressure vessels

Air or nitrogen

All other valves

When using air to test steam service valves operating at temperatures greater than 150°F, make temperature corrections to the set pressure to achieve the cold differential test pressure. The cold differential test pressure for a conventional relief valve is the set pressure plus the temperature correction minus the superimposed back pressure. The cold differential test pressure for a balanced relief valve is the set pressure plus the temperature correction minus the pressure in the relief valve bonnet.


1200-129

November 2001

1200 Relief Systems


The following are the temperature corrections for using air to test steam valves: Operating Temperature Temperature Correction Increase % of the difference between the set pressure and the back pressure (or bonnet pressure for a balanced valve) 0-150°F

none

151-600°F

1%

601-800°F

2%

801-1000°F

3%

For example: a temperature correction of approximately 2 psig is required for a conventional valve operating at 250°F, with a set pressure of 200 psig and a superimposed back pressure of 5 psig [(200-5)×0.01=1.95]. Therefore, the cold differential test pressure is 197 psig [200+2-5].

Blowdown Ring Adjustment for the Valve Test When possible, the valve should be tested by relieving the rated capacity to obtain the full relief lift on the valve stem. This procedure tests the spring and ensures that the stem is not binding. However, there are instances where the capacity of the test stand is considerably less than the valve being tested. When this condition exists, the valve may pop and open; but since the capacity of the test stand accumulator tank is below the valve relieving capacity, the valve will quickly slam shut. This slamming action will damage the valve seat. To allow the valve to stay open after popping, the blowdown ring can be set at a high blowdown percentage or a stop can be applied to limit the movement of the valve disk. To allow the valve to stay open for a period after popping, the blowdown ring is adjusted to a high blowdown percentage value. (Before making adjustments to the blowdown ring, mark its position and the location of the adjustment notches.) Consult the valve manufacturer for the test blowdown setting. Typically, the blowdown ring is brought up to the disk holder, then lowered by two notches on the blowdown ring. The blowdown ring locking screw shall be inserted so that it is positioned between the adjustment notches on the ring. Another method of preventing the valve from closing too quickly is to limit the travel of the disk. This can be done by limiting the upward movement of the valve stem. Often, the valve cap is removed and replaced with a cap that has a set-screw positioned directly above the stem. When the valve is in the closed position, the setscrew is lowered to touch the top of the stem and then raised by approximately ¼ inch. This will restrict the valve lift and prevent the accumulation tank from emptying too rapidly.

Opening Pressure Valves in hot service should be preheated to allow the metal to expand to its operating size. After the valve has reached its service temperature, slowly increase the test pressure until the valve starts to simmer and pop (preferably fully open). The

November 2001

1200-130



1200 Relief Systems

relief valve should start to open when the test pressure reaches the cold differential test pressure. Note pressure on the test gage (relief valves are fully open when the test pressure equals the cold differential test pressure plus the accumulation). After the safety or relief valve fully opens, the test pressure gage should show decreasing pressure. Note pressure at which the valve closes. See Figure 1200-48 for tolerances for opening pressure. Fig. 1200-48 Tolerances for Opening Pressure Equipment

Set Pressure (psig)

Tolerance (± psi or %)

Power Boiler (ref. ASME Section I, PG 72.2)

p ≤ 70 70 < p ≤ 300 300 < p ≤ 1000 p > 1000

2 psi 3% 10 psi 1%

ASME Pressure Vessel (ref. ASME Section VIIIDiv. 1 UG-134 (d) (1))

p ≤ 70 p > 70

2 psi 3%

After the valve is adjusted so that the opening pressure is within the above tolerances, pop the valve three additional times to ensure that opening pressure is repeatable and within tolerance. When the operating pressure is set, replace the cap and car-seal it to the valve bonnet.

Reseat Pressure Adjustments When the reseating pressure is not fixed (determined by the style of valve), it is adjusted by varying the position of the blowdown ring. Adjust the reseating pressure as follows. 1.

Review the reseating pressures recorded on the three final valve tests used to determine the valve opening pressure. The reseating pressures should be compared against the desired blowdown.

2.

Move the blowdown ring the required amount of notches to obtain the desired setting. Consult the valve manufacturer for the ring notch position and blowdown percentage correlation.

If the blowdown ring was adjusted to accommodate testing a large capacity valve, return the blowdown ring to its original position. Also check the position of the ring against the valve manufacturer’s recommended settings. If there is a discrepancy between the two settings, consult the valve manufacturer to discuss the proper setting required for the desired blowdown. Record the blowdown ring setting in the valve maintenance record for future reference. After the blowdown ring is adjusted, insert the ring locking screw. Visually verify that the screw is positioned between the blowdown ring notches before tightening the screw. Tighten and car-seal the locking screw.


1200-131

November 2001

1200 Relief Systems


Reseating Pressures Reseating pressures should not be set for more than 10% below the valve opening pressure. Valves set with a blowdown greater than 10% may not reseat. This can happen when the reseating pressure is set below the normal operating pressure of the process. Typically the valve manufacturers set the blowdown between 7 and 10%. Following are the resetting pressure requirements for power boilers: Equipment

Reset Pressure

Power Boiler

0.96 × set pressure

Boilers < 100 psig

(set pressure - 4 psi)

Boiler (200-300 psig)

0.99 × set pressure

All

(set pressure - 2 psi)

Seat Leakage Test Leaking valves pose a hazard to personnel and equipment. They also lead to fouled and/or inoperable valves, and loss of product. To help identify seat leakage, the American Petroleum Institute (API) has developed, Standard 527, Commercial Seat Tightness of Safety Relief Valves With Metal to Metal Seats. Use API Standard 527 as a guide for testing seat leakage (a copy of this standard is included in this manual). This test should be performed after the valve’s pop and reseating pressures are set. The test apparatus for measuring seat tightness is shown in Figure 1 of API Standard 527. After all openings in the valve’s secondary pressure zone (caps, drain holes, vents, and outlets) are closed, the test pressure at the valve inlet shall be held at 90% of the set pressure immediately after popping the valve. For valves set at or below 50 psig, the test pressure shall be set at 5 psig below the set pressure immediately after opening the valve. Before the bubble count is started, the test pressure shall be held for the following periods. For nonsteam service valves with metal-to-metal seats, the following leakage rate (in bubbles per minute) shall not be exceeded:

November 2001

Valve Inlet Size

Pre-Bubble Count Holding Period

2 inches and smaller

1 minute

2½ inches, 3 inches

2 minutes

4 inches, 6 inches, 8 inches

5 minutes

1.

For set pressures less than or equal to 1000 psig, the leakage rate shall not exceed the values listed in API Standard 527, Figure 1.

2.

For set pressures greater than 1000 psig, the leakage shall not exceed the values from API Standard 527, Figure 2. For valves with resilient (elastomeric O-ring) seat seals, there should be no leakage (0 bubbles per minute).

1200-132



1200 Relief Systems

For valves in steam service, the test media should be steam using the test pressures outlined in API Standard 527 (the test cover for the valve outlet flange may be omitted). There should be no audible or visible leaks. Visible leaks can be detected with the assistance of a mirror. Place the mirror by the side (not in front) of the outlet flange. Steam leaks will fog the mirror. For valve equipment in toxic, hazardous, corrosive, or cryogenic service, the leakage should be 0 bubble per minute.

Bonnet Leak Test After the seat tightness test is complete and the valve is depressurized, the bonnet shall be tested for leaks. This test checks for leaks in the secondary pressure zone of the valve. The test is conducted as follows: 1.

Plug all openings on the bonnet.

2.

Place a blind test flange on the valve outlet flange. The test flange has a bulkhead fitting to allow it to connect to an air source. Or, adapt the test flange (in Figure 1 of API Standard 527) that is used for testing the seat leakage.

3.

Apply approximately 35 psig air to the test flange. Caution should be used in pressurizing valves equipped with bellows. Excessive pressure can damage (collapse) the bellows. Consult the valve manufacturer for the pressure ratings of the bellows. The bonnet leak test pressure shall not exceed the pressure rating of the bellows.

4.

Apply a soap solution to all the bonnet connections and check for bubbles at the bonnet-to-base joint, adjustment pin seal, cap-to-bonnet joint, and all plugged openings. Tighten the necessary connection to eliminate bubbles. Adjustments shall be made so as not to change any of the valve settings.

Pilot Operated Valves Pilot operated valves shall be tested in the same manner as described above for spring-loaded relief valves. The exception is that the blowdown is adjusted in the pilot valve instead of the main valve. The pilot operated valve shall be tested on the shop test bench as a complete assembly (pilot connected to the main valve). Testing and setting the pilot will not assure that the valve assembly will open and reseat at the desired pressures. It is not uncommon to have different pressure settings for the pilot and the main valve. For example, the reseat pressure of the full valve assembly and pilot may differ by as much as 3%. For some styles of valves, the main valve dome pressure may be at 70% of the supply pressure while the pilot will be at 100% of the supply pressure. The valve manufacturer should be consulted for the proper pressure differential setting (if any) between the pilot and the main valve.


1200-133

November 2001

1200 Relief Systems


On-Line Testing Testing relief valves that are connected to the process plant is not recommended unless required by code. Testing the valve in-place can lead to seat damage or leakage by allowing dirt, scale, or other solids to deposit between the valve disk and the seating surface. Seat damage can also occur if the valve is opened suddenly and then slammed shut. If on-line testing cannot be avoided, it should be limited to valves operating in clean, low temperature and pressure service, or if experience indicates that the test is safe and practical. However, one must recognize that on-line testing will not always verify blowdown, seat leakage, bonnet leaks, or the physical condition of the valve internal components. Steam and air valves on receivers shall have their opening levers lifted once a year. This is an ASME code requirement. Before lifting the lever the following procedures shall be followed: 1.

Obtain the proper plant or regional operating permits.

2.

Notify the operators and confirm that the process is stable and that the steam drum or receiver has been recently blowndown (to minimize the amount of dirt in the valve seating area).

Boiler safety valve levers are normally operated when there is at least 75% of the valve set pressure in the boiler. This is to reduce the amount of force required to lift the lever. Section I of the ASME Boiler and Pressure Vessel Code does not allow block valves under boiler safety valves, thus testing on the equipment is done by raising the steam pressure and observing when the valve opens and reseats.

1283 Records and Reports Records This manual recommends that specification records, historical records, and test records be maintained for all relief valves. A record folder should be maintained for each valve to track the valve maintenance history and to determine if the inspection and testing interval is adequate for the valve. Other documents that can be placed in the records folder are: the valve data sheet and the valve drawing with the parts identified. If there is a need to change valve settings and/or parts, the changes shall be noted on all documents in the records folder. Proposed revisions in the valve materials, spring size, set pressure, and blowdown shall be reviewed with the plant and/or area engineer before these modifications are attempted.

November 2001

1200-134



1200 Relief Systems

Specification Record A specification record should be developed for each relief valve. This record provides the basic data to test and inspect the valve. A copy of this record should accompany the test record whenever a valve is tested or inspected. A sample specification record is shown in Figure 1200-49. This document also provides: •

Design information for ordering new parts (the valve data sheet should be used to order a new valve)

•

Part identifications to allow assembly of an identical valve from spare parts

Historical Record A historical record should be developed for each relief valve. This record tracks when the valve was tested or inspected. API Guide for Inspection of Refinery Equipment, Chapter XVI, Appendix B, gives an example of a historical record. The API sample form should be amended to include information on whether the valve was inspected, tested in-line, and/or shop tested. A test/inspection history is valuable for: •

Monitoring the performance and condition of the valve at the various test/inspection intervals. These data can be used to adjust the test frequency of the valve (the test/inspection frequency shall not exceed the periods set by the codes).

•

Developing a test history for valves that are required to be tested at intervals specified by codes.

•

Evaluating the valve performance data to see if the valve is suitable for the service intended. These data will indicate problems in design, materials, or a poor application for the valve.

Test Record A test record contains the results of a test and/or inspection. A sample test record is shown in Figure 1200-50. A copy of each test record should be retained in the valve report folder.

Report Monthly reports are recommended to identify which valves need testing and which valves are past the scheduled test dates. The Chevron Plant Equipment Information System (PEIS) computer program has a relief valve inspection/test reporting function. Contact Chevron Information Technology Company, Computer Application Department, Manufacturing, Chemical, and Engineering Division, for information on setting up the test/inspection reporting program.


1200-135

November 2001

1200 Relief Systems


Fig. 1200-49 Specification Record Specification Record Valve Number:

Serial Number:

Plant:

Protecting:

Service: Protecting a Code Vessel:

yes

no

Valve Inlet Size:

Rating:

Connection Type:

Valve Outlet Size:

Rating:

Connection Type:

Manufacturer:

Model Number:

Type:

Conventional

Balanced

Pilot

Breather

Orifice Size: Body and Bonnet Material:

NACE:

yes

no

Nozzle and Disk Material:

NACE:

yes

no

Trim Material:

NACE:

yes

no

NACE:

yes

no

Spring Material:

Number:

Bellows Material: Gasket Material: Cold Differential Test Pressure:

Set Pressure:

Superimposed Backpressure: Blowdown %:

Blowdown Ring Setting:

Accumulation: Normal Operating Temperature: Test Interval: Remarks: Rev.

November 2001

Date

Description

Approval

1200-136



1200 Relief Systems

Fig. 1200-50 Test Record Test Record Test Date:

Date Last Tested:

Valve Number:

Valve Protecting a Code Vessel: yes no

Cold Differential Test Pressure:

Normal Operating Temperature:

Initial “As-Received” Test Blowdown Ring Setting:

Opening Pressure:

Valve Inlet Leakage at: (pressure or vacuum value) Valve Outlet Leaks:

yes

no

Bonnet Leaks:

yes

Disk & Seats

Disk Holder

Inspection Inlet

Outlet

Nozzle/Seat

Good

Good

Good

Good

Good

Dirty

Dirty

Dirty

Dirty

Dirty

Corroded

Corroded

Eroded

Eroded

Corroded

Plugged

Plugged

Machined

Corroded

Machined

Flange dam.

Flange dam.

Lapped

Cut

Galled

Threads dam.

Threads dam.

Replaced w/

Machined

Replaced w/

#

Lapped

#

Replaced w/ # Spring

Guide

Stem

Bellows

Body

Good

Good

Good

Good

Good

Dirty

Dirty

Dirty

Corroded

Corroded

Corroded

Corroded

Corroded

Broke

Eroded

Crack/broke

Machined

Bent

Cracked/leaks

Cracks

Replaced w/

Galled

Replaced w/

Replaced w/

#

Replaced w/

#

#

Vents

Threads

Seals

Check Valves

Open

Good

Good

Tight

Plugged

Fouled

Replaced

Leaking

#

Machined

Replaced

Valve Body Minimum Thickness: Shop Test Test Media:

air

water

Opening Pressure:

steam

other__________ __

Cold Differential Test Pressure:

Acceptable Opening Pressure Tolerance: Reseat Pressure: Blowdown = [1-(Reseat Press.÷Cold Diff. Test Press.)]: Seat Leakage:


bubbles/minute

1200-137

November 2001

1200 Relief Systems


1290 References 1.

ASME Boiler and Pressure Vessel Code, Section I – Rules for Construction of Power Boilers

2.

ASME Boiler and Pressure Vessel Code, Section VIII Division 1 – Rules for Construction of Pressure Vessels

3.

ASME Code for Pressure Piping: B31.1 - Power Piping

4.

ASME Code for Pressure Piping: B31.3 - Chemical Plant and Refinery Piping

5.

ASME Code for Pressure Piping: B31.4 - Liquid Transmission Systems Piping

6.

ASME Code for Pressure Piping: B31.8 - Gas Transmission and Distribution Piping Systems

7.

National Board of Boiler and Pressure Vessel Inspectors NB-18 – Pressure Relief Device Certifications NB-18

8.

API Recommended Practice 520 - Sizing, Selection and Installation of Pressure-Relieving Devices in Refineries, Part I – Sizing and Selection

9.

API Recommended Practice 520 - Sizing, Selection and Installation of Pressure-Relieving Devices in Refineries, Part II – Installation

10. API Recommended Practice 521 - Guide for Pressure-Relieving and Depressuring Systems 11. API Standard 526 - Flanged Steel Pressure Relief Valves 12. API Standard 527 - Commercial Seat Tightness of Safety Relief Valves With Metal to Metal Seats 13. API Recommended Practice 572 - Inspection of Pressure Vessels 14. API Recommended Practice 576 - Inspection of Pressure Relieving Devices 15. NFPA 30 – Flammable and Combustible Liquids Code 16. Guidelines for Pressure Relief and Effluent Handling Systems, Center for Chemical Process Safety, American Institute of Chemical Engineers, 1998. 17. Relief Systems Handbook, Cyril F. Parry, Gulf Publishing, 1992. 18. Flare Gas Systems Pocket Handbook, K. Banerjee, N. P. Cheremisinoff, and P. N. Cheremisinoff, Gulf Publishing, 1985. 19. Flares: Design and Operation, John F. Straitz, National Airoil Burner Company, 1990. 20. Flow of Fluids Through Valves, Fittings, and Pipe, Crane TP-410, Crane Co., 1981.

November 2001

1200-138



1200 Relief Systems

21. Perry’s Chemical Engineer’s Handbook, Robert H. Perry, Don W. Green, and James O. Maloney, 6th Edition, McGraw-Hill, 1984. 22. NIOSH Pocket Guide to Chemical Hazards, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services (Publication No. 97-177-604), 1997. 23. P. C. Berwanger, R. A. Kreder, and W. R. Lee, “Non-conformance of Existing Pressure Relief Systems with Recommended Practices: A Statistical Analysis,” paper LPS-2c in LPS 1999 – Proceedings of the 33rd Annual Loss Prevention Symposium, American Institute of Chemical Engineers, 1999. 24. H. W. Husa, “How to Compute Safe Purge Rates,” Hydrocarbon Processing and Petroleum Refiner, 43 (5), 179-182 (May, 1964). 25. W. Y. Wong, “PRV Sizing for Exchanger Tube Rupture,” Hydrocarbon Processing, 71 (2), 59 – 64 (February 1992). 26. L. L. Simpson, “Estimate Two-Phase Flow in Safety Devices,” Chemical Engineering, 98 (8), 98 – 102 (August 1991). 27. H. H. West, M. S. Mannan, R. Danna, and E. M. Stafford, “Make Plants Safer with a Proper Management of Change Program,” Chemical Engineering Progress, 94 (6), 25 (June 1998). 28. M. M. R. Eastman and J. R. Sawyers, “Learning to Document Management of Change,” Chemical Processing, 61 (11), 15-22 (November 1998). 29. American National Standards Institute, ANSI B95.1, Terminology for Pressure Relief Devices. 30. API Guide for Inspection of Refinery Equipment, Chapter XVI, Pressure Relieving Devices. 31. API RP 14C. Analysis, Design, Installation, and Testing of Basic Surface Safety Systems on Offshore Production Platforms. 32. API Std 527. Commercial Seat Tightness of Safety Relief Valve with Metal-toMetal Seat. 33. API Std 620. Recommended Rules for Design and Construction of Large, Welded, Low-Pressure Storage Tanks. 34. API Std 650. Welded Steel Tanks for Oil Storage. 35. Code of Federal Regulations, Part 192, Title 49, Transportation of Natural and Other Gas By Pipelines: Minimum Federal Safety Standards Code of Federal Regulations, Department of Transportation, Federal Gas Pipeline Safety Standards. 36. Department of Interior, Federal Register, Vol. 53, No. 63, Appendix 1, Register No. 250.124, Production Safety Systems Records.


1200-139

November 2001

1200 Relief Systems


37. ISA Std S20. Specification Forms for Process Measurement and Control Instruments, Primary Elements and Control Valves. 38. ISA Std S20.53, Instructions for Ordering Pressure Relief Valves. 39. National Board Inspection Code, Boilers and Pressure Vessels. 40. NB-18, Pressure Relief Device Certifications (National Board of Boiler and Pressure Vessel Inspectors), “National Board Authorization to Repair ASME and NB Safety and Safety Relief Valves for VR Certification.” 41. Perry’s Chemical Engineers’ Handbook. [Robert H. Perry]. 6th edition. New York: McGraw-Hill, 1984.

12100 Glossary Accumulation. The pressure increase over the maximum allowable working pressure of a vessel during discharge through the pressure relief device, expressed in pressure units or as a percent. Maximum allowable accumulations are established by applicable codes for operating and fire contingencies. Note The ASME Boiler and Pressure Vessel Code does not use the word “accumulation,” but does allow for a vessel’s pressure to rise above the MAWP by an amount that varies with the nature of the overpressure event and the number of relief devices protecting the vessel. Actual discharge area. The measured minimum net area that determines the flow through a valve. Applicable (or credible) causes of overpressure. A contingency or event in which a vessel or other equipment item can be exposed to pressures in excess of its design pressure and that has sufficient likelihood of occurrence that is should be included in the design basis of a pressure relief system. Atmospheric discharge. The release of vapors and gases from pressure-relieving and depressuring devices to the atmosphere. This may be via either an open or a closed disposal system. Back pressure. The pressure that exists at the outlet of a pressure relief device as a result of the pressure in the discharge system. Back pressure can be either constant or variable. Back pressure is the sum of the superimposed and built-up back pressures. Balanced pressure relief valve. A spring-loaded pressure relief valve that incorporates a means for minimizing the effect of back pressure on its performance characteristics. Blowdown. The difference between the set pressure and the closing pressure of a pressure relief valve, expressed as a percentage of the set pressure or in pressure units.

November 2001

1200-140



1200 Relief Systems

Blowoff. If the velocity of fuel exiting a burner or flare tip exceeds the flame propagation velocity, the flame will be lifted above the burner. This condition is called blowoff. The lifting of the flame continues until a new stable position is reached as a result of turbulent mixing and dilution of the fuel with air. Breaking pin device. A nonreclosing pressure relief device actuated by inlet static pressure and designed to function by the breakage or buckling of a load-carrying section of a pin that supports a pressure containing member. Also known as a buckling pin device or a rupture pin device. Built-up back pressure. The increase in pressure at the discharge of a relief device that develops as a result of flow after the pressure relief device or devices open. Note The term “built-up back pressure” is used in some documents (e.g., Section 2.2.4.1 of API RP-520 Part I, 6th ed.) as the criterion for defining back pressure limitations on typical pressure relief valves. However, this term is not used in the ASME Boiler and Pressure Vessel Code discussion of back pressure limitations (see BPVC Section VIII, Appendix M-8, 1998 ed.). Burst pressure. The inlet static pressure at which a rupture disk device functions. Chatter. Abnormal rapid reciprocating motion of the movable parts of a pressure relief valve in which the disk contacts the seat. Choked (or Critical) Flow. The flow of a compressible fluid through a pressure relief device, piping or other equipment whose rate does not respond to a decrease in the downstream or back pressure on the device. The mass flow rate then depends only on upstream conditions. This usually occurs when the flow velocity equals the velocity of sound in the fluid at the flowing conditions. Choked flow can occur with gas/vapor or with two-phase fluid systems. Also referred to as Sonic Flow. Closed disposal system. A disposal system capable of containing pressures that are different from atmospheric pressure. Closed-bonnet pressure relief valve. A pressure relief valve whose spring is totally encased in a metal housing. This housing protects the spring from corrosive agents in the environment and is a means of collecting leakage around the stem or disk guide. The bonnet may or may not be sealed against pressure leakage from the bonnet to the surrounding atmosphere, depending on the type of cap or lifting-lever assembly employed or the specific handling of bonnet venting. Closing pressure. The value of decreasing inlet static pressure at which the valve disk reestablishes contact with the seat or at which lift becomes zero. Coefficient of discharge. The ratio of the measured relieving capacity to the theoretical relieving capacity. Cold differential test pressure. The inlet static pressure at which a pressure relief valve is adjusted to open on the test stand. This test pressure includes corrections for service conditions of temperature and/or back pressure. Collection system. A system of piping, fittings, and equipment designed to gather fluids from pressure relief devices, process vent valves, and emergency depres-


1200-141

November 2001

1200 Relief Systems


suring valves, and to prepare these fluids for their safe disposal. The collection system includes separation equipment such as hot blowdown drums, acid gas neutralizing scrubbers, and flare knockout drums. Combination capacity factor. The ratio of average flow capacity determined by tests of a pressure relief valve in combination with a rupture disk device to the flow capacity of the pressure relief valve alone. Conventional pressure relief valve. A spring-loaded pressure relief valve whose performance characteristics are directly affected by changes in the back pressure on the valve. Design basis. The basis for sizing equipment and associated piping used for process or pressure relief services. For a pressure relief system, the design basis includes the identification of all applicable causes of overpressure, the calculation of all required relief flow rates for all equipment items in the process, the calculation of the required sizes of all pressure relief devices and their connecting piping and fittings, and the design of systems and equipment for collecting and disposing fluids discharged from these devices. Design Institute for Emergency Relief Systems (DIERS). Institute under the auspices of the American Institute of Chemical Engineers, founded to investigate design requirements for relief devices and vent lines for two-phase relief of runaway chemical reactions. This activity has been carried forward by the DIERS Users Group, and the scope of activity has been broadened to include a variety of important technical topics related to pressure relief system design. Design pressure. The design pressure of a vessel is at least the most severe condition of coincident temperature and gauge pressure expected during operation. The design pressure is the pressure used in the design of a vessel to determine the minimum permissible thickness or other physical characteristics of the different parts of the vessel (see also maximum allowable working pressure). The design pressure may be used in place of the maximum allowable working pressure (MAWP) in all cases in which the MAWP has not been established. Note For low (just above atmospheric) pressure tanks, the design pressure may be expressed as a “pressure not to be exceeded.” Differential set pressure. The difference between the desired set pressure of a pressure relief valve and the superimposed back pressure on the valve at the time the valve is called upon to operate. Disk (valve). A disk is the pressure containing movable element of a pressure relief valve which effects closure. Disposal System. A system of piping and/or equipment for the safe ultimate disposal of vapors delivered from a collection system. The disposal system usually takes the form of a vent stack, a flare gas recovery unit, an incinerator, or a flare. Effective discharge area, or equivalent flow area. A nominal or computed area of flow through a pressure relief valve for use in recognized flow formulas to deter-

November 2001

1200-142



1200 Relief Systems

mine the capacity of a pressure relief valve. It is generally less than the actual discharge area. Also termed the “orifice” area. Emergency shutdown (ESD) system. The safety control system that, either automatically or manually, overrides the action of the basic control system when predetermined conditions are exceeded. Emergency venting. The venting required when an abnormal condition, such as ruptured internal heating coils or an external fire, exists either inside or outside of a low-pressure tank. Contrast with normal venting. Equivalent ideal nozzle area. The product of the actual discharge area and the coefficient of discharge. Flammability limits. The range of gas or vapor amounts in air that will burn or explode if a flame or other ignition source is present. Generally, the wider the range the greater the fire potential. See also Lower Explosive Limit/Lower Flammable Limit and Upper Explosive Limit/Upper Flammable Limit. Flare. A means for safely disposing of waste gases through the use of combustion. With an elevated flare, the combustion is carried out at the top of a pipe or stack where the burner and igniter are located. A ground flare is similarly equipped except that combustion is carried out at or near ground level. A burn pit differs from a flare in that it is primarily designed to handle liquids. Flash point. The minimum temperature of a liquid at which sufficient vapor is given off to form an ignitable mixture with air, near the surface of the liquid or within the vessel used, as determined by the appropriate test procedure and apparatus. See NFPA 30 for test procedures appropriate for various liquids. Importance The lower the flash point temperature of a liquid, the greater the chance of a fire hazard. Flashback. A situation in which the trailing edge of a flame fed by flowing gases begins to propagate backward (toward the source of the gases). It can occur if the flame propagation velocity exceeds the linear velocity of the source gas. Freeboard. The clear, vertical space above a liquid or a foam in a vessel. Global relief contingency. A contingency in which a single event results in the opening of multiple pressure relief devices or emergency depressuring valves. Hazard and Operability Study (HAZOP). A systematic qualitative technique to identify process hazards and potential operating problems using a series of guide words to study process deviations. A HAZOP is used to discover what deviations from the intention of the design can occur and what their causes and consequences may be. This is a systematic detailed review technique that can be applied to new or existing processes to identify hazards. Hazardous material. In a broad sense, any substance or mixture of substances having properties capable of producing adverse effects on the health or safety of human beings. This includes material presenting dangers beyond the fire problems


1200-143

November 2001

1200 Relief Systems


relating to flash point and boiling point. These dangers may arise from, but are not limited to, toxicity, reactivity, instability, or corrosivity. Hold-up volume. The volume of a knockout drum between the maximum level normally maintained (the level at which a drain valve is opened or a drain pump is started) and the maximum level at which the acceptable liquid droplet size can be separated. Homogeneous equilibrium flow. The flow of liquid and vapor phases under conditions in which the two phases are in thermodynamic equilibrium and are flowing at the same velocity. IDLH Concentration. A measure of the potential for harmful effects due to acute exposure to a toxic substance in air. The “Immediately Dangerous to Life and Health (IDLH) concentration” is defined such that exposure to concentrations of a toxic substance in air above this level for more than thirty minutes may result in impairment of an individual’s ability to escape to a safe location. In the analysis of potential consequences of acute toxic-substance releases, the IDLH level is normally used as the concentration not to be exceeded at locations to which personnel have access. Inapplicable (or noncredible) causes of overpressure. A contingency or event that has very low and unreasonable likelihood of occurrence or that cannot result in the exposure of an equipment item to pressures in excess of its design pressure. Such a contingency need not be included in the design basis for an emergency relief system. However, the reasons for identifying a specific contingency as inapplicable or noncredible should be documented as part of the relief system’s design basis. Inlet Size. The nominal pipe size (NPS) of the inlet of a pressure relief valve, unless otherwise designated. Isothermal. A system condition in which the temperature remains constant. Isothermal behavior implies efficient heat exchange between the system and the surroundings. Knock-out drum. A vessel installed in a relief discharge collection system in order to separate the liquid and vapor components of the discharge fluid for separate subsequent handling. Leak-test pressure. The specified inlet static pressure at which a seat leak test is performed on a pressure relief valve. Lift. The actual travel of the disk away from its closed position when a pressure relief valve is relieving. Liquid seal. An equipment item, often incorporated into a flare stack, designed to prevent both flashback and infiltration of air from the flare stack into the flare header. Gases from the flare header are bubbled through a non-flammable liquid (typically water). The liquid head maintains a slight positive pressure in the flare header.

November 2001

1200-144



1200 Relief Systems

Liquid trim valve. A safety relief valve designed for incompressible fluid service, typically with adjustable blowdown. Such a valve can be certified for use with compressible fluids. Low pressure equipment. Any equipment (typically a storage tank) with a design pressure less than or equal to 15 psig. Lower Explosive Limit (LEL) or Lower Flammable Limit (LFL). The lowest concentration of a vapor or gas (the lowest percentage of the substance in air) that will produce a flash of fire when an ignition source (heat, arc, or flame) is present. See also Upper Explosive Limit or Upper Flammable Limit. Importance burn.

At a concentration lower than the LEL, the mixture is too “lean” to

Management of Change (MOC). Procedures for the formal review, authorization, and documentation of changes to a process industry facility. Management of change is one of the required elements of two US federal regulations intended to prevent or minimize the consequences of accidental releases of highly hazardous compounds. These regulations are the OSHA Process Safety Management standard and the EPA Risk Management Program rule. Maximum allowable venting (or relieving) pressure. The pressure corresponding to the Code-allowable pressure rise (accumulation) above the maximum allowable working pressure (MAWP) under relieving conditions. Also referred to as the maximum allowable accumulated pressure (MAAP). Maximum allowable working pressure (MAWP). The maximum gauge pressure permissible at the top of a completed vessel in its operating position for a designated temperature. The pressure is based on calculations for each element in a vessel using nominal thicknesses, exclusive of additional metal thicknesses allowed for corrosion and for loadings other than pressure. The MAWP is the basis for the pressure setting of the pressure relief devices that protect the vessel. Maximum operating pressure. The maximum pressure expected during normal system operation. Molecular seal. A type of purge reduction seal in which the difference between the molecular weights of air and the flared gas is exploited to reduce the rate of infiltration of air into a flare stack. Usually located in the flare stack just below the flare tip, this is sometimes called a diffusion seal. Net flow area. The area which determines the flow after a nonreclosing pressure relief device has operated. The (minimum) net flow area of a rupture disk is the calculated net area after a complete burst of the disk, with appropriate allowance for any structural members that may reduce the net flow area through the rupture disk device. Nonreclosing pressure relief device. A pressure relief device designed to remain open after operation. A manual resetting means may be provided.


1200-145

November 2001

1200 Relief Systems


Normal venting. The venting of an equipment item required because of operational requirements or atmospheric changes. Contrast with “emergency venting.” Nozzle. The pressure-containing element of a pressure relief valve that constitutes the smooth inlet flow passage and includes the fixed portion of the seat closure. The downstream end of the nozzle is often termed the valve ''orifice'' or nozzle “throat.” Note

“Nozzle” is also used to refer to the inlet or outlet connection on a vessel.

Open disposal system. A disposal system that discharges directly from the relieving device to the atmosphere with no containment other than a short tail pipe. Open-bonnet pressure relief valve. A pressure relief valve whose spring is directly exposed to the atmosphere through the bonnet or yoke. Depending on the design, the spring may be protected from contact with vapors or gases discharged by the valve and will be cooled by the free passage of ambient air through and around the spring. Opening pressure. The value of increasing inlet static pressure of a pressure relief valve at which there is a measurable lift, or at which the discharge becomes continuous as determined by seeing, feeling, or hearing. Operating margin. The difference between the highest pressure attained in normal operation and the set or burst pressure, expressed either in pressure units or as a percentage of the set pressure. Note that the ASME Boiler and Pressure Vessel Code term is “pressure differential” [ASME VIII, Appendix M-7]. Operating pressure. The pressure to which the vessel is usually subjected in service. A pressure vessel is normally designed for a maximum allowable working pressure that will provide a suitable margin above the operating pressure to prevent any undesirable operation of the relief device. Outlet size. The nominal pipe size (NPS) of the outlet of a pressure relief valve, unless otherwise designated. Overpressure. The pressure increase over the set pressure of the relief device, expressed in pressure units or as a percent of set pressure. It is the same as accumulation when the relief device is set at the maximum allowable working pressure of the vessel, assuming no inlet pipe loss to the relieving device. Note When the set pressure of the relief device is less than the vessel’s maximum allowable working pressure, the overpressure will generally be greater than the vessel’s accumulation. Pilot. A device for maintaining a small, constant flame that serves to ignite combustible gases and vapors as they exit a flare. Pilot-operated pressure relief valve. A pressure relief valve in which the main relieving device is combined with and controlled by a self-actuating auxiliary pressure relief valve. Pressure relief device. A device actuated by inlet static pressure and designed to open during an emergency or abnormal condition to prevent a rise of internal fluid

November 2001

1200-146



1200 Relief Systems

pressure in excess of a specified value. The device also may be designed to prevent excessive internal vacuum. The device may be a pressure relief valve, a nonreclosing pressure relief device, or a vacuum relief valve. Pressure relief valve. A pressure relief device designed to reclose and prevent the further discharge of fluid after normal conditions have been restored. A generic term applied to relief valves, safety valves, and safety relief valves. Pressure/Vacuum (PV) valve. Either a pilot-operated valve or a direct-acting vent (a weight-loaded or spring-loaded valve), a PV valve is used to relieve excess pressure or vacuum that has developed in a tank or other low-pressure equipment. Pressure-relieving system. An arrangement of a pressure-relieving device, piping, and a means of disposal intended for the safe relief, conveyance, and disposal of fluids in a vapor, liquid, or gaseous phase. A relieving system may consist of only one pressure relief valve or rupture disk, either with or without discharge pipe, on a single vessel or line. A more complex system may involve many pressure-relieving devices manifolded into common headers to terminal disposal equipment. Process equipment item. Used in this document to indicate any pressurecontaining element in a process industry facility. Thus, as used here, all process equipment items require analysis of their pressure relief requirements. Examples of process equipment items include pressure vessels, piping, low-pressure storage tanks, etc. Process fluid. Used in this document to indicate the fluid in any pressure-containing element in a process industry facility. Thus, in this document “process” fluids include those that are often considered utility fluids: compressed air, cooling water, steam, refrigerants, heat transfer oils, fuel gas, fuel oil, etc. Process Hazard Analysis. An organized effort to identify and evaluate hazards associated with chemical processes and operations to enable their control. This review normally involves the use of qualitative techniques to identify and assess the significance of hazards. Conclusions and appropriate recommendations are generally developed. Occasionally, quantitative methods are used to help prioritize risk reduction. Purge gas. A gas that is continuously or intermittently added to a system to render the atmosphere nonignitable. The purge gas may be inert or combustible. Quench vessel or drum. A vessel containing liquid and a sparger for quenching an effluent stream. Commonly used for cooling or condensing vapor or vapor-liquid mixtures, or for reacting effluent with a neutralizing agent, or for absorbing hazardous components from an effluent. Quenching. Rapid cooling of a fluid from an elevated temperature by mixing it with another fluid of a lower temperature. For example, severe cooling of a reacting system in a short time prevents further reaction. Rated relieving capacity. That portion of the measured relieving capacity permitted by the applicable code or regulation to be used as a basis for the application of a


1200-147

November 2001

1200 Relief Systems


pressure relief device. Also known as the stamped capacity, certified capacity, or nameplate capacity. Relief discharge stream. The fluids discharged from pressure relief devices, emergency depressuring valves, and normal process vents. Also called effluent stream. Relief valve. A pressure relief valve having a gradual lift that is generally proportional to the increase in inlet static pressure in excess of the opening pressure. A relief valve is used primarily with incompressible fluids. Relieving capacity. The flow rate through a pressure relieving system calculated for a designated temperature, set pressure and overpressure, using a flow area and coefficient of discharge (or resistance factors for rupture disk device systems) determined in certification tests or otherwise specified in ASME BPVC. For pressure relief devices covered under the ASME BPVC but normally not capacity-certified, and for low pressure relief devices not covered by the ASME BPVC, the relieving capacity is based on the manufacturer’s calibration data. This value does not include the Code-mandated factor of 0.9. Also known as the best estimate flow rate. Relieving conditions. The inlet pressure and temperature of a pressure relief device at a specific overpressure. The relieving pressure is equal to the valve set pressure (or rupture disk burst pressure) plus the overpressure. The temperature of the flowing fluid at relieving conditions may be higher or lower than the operating temperature. Remote-sensing pressure relief valve. A pilot-operated pressure relief valve installed so that the opening pressure signal is obtained at a location remote from the valve body, typically at the vessel being protected. Required relief capacity. The relief flow rate required (to be supplied by a relief device) to prevent the pressure or vacuum in an equipment item from exceeding that allowed by applicable codes in the event of a specific causes of overpressure. Runaway reaction. A thermally unstable chemical reaction system that shows an accelerating increase of temperature and reaction rate. The runaway reaction can finally result in vessel overpressure. Rupture disk device. A nonreclosing differential pressure relief device actuated by inlet static pressure and designed to function by bursting the pressure-containing rupture disk. A rupture disk device includes a rupture disk and a rupture disk holder. Safety relief valve. A pressure relief valve characterized by a rapid opening pop action or by opening generally proportional to the increase in inlet pressure in excess of the opening pressure. It may be used for either compressible or incompressible fluids, depending on design, adjustments or application. Safety valve. A pressure relief valve actuated by inlet static pressure and characterized by rapid opening or pop action. A safety valve is normally used to relieve compressible fluids.

November 2001

1200-148



1200 Relief Systems

Note A distinction is often made between “low lift” safety valves (in which the discharge area is determined by the disk position) and “full lift” safety valves (in which the discharge area is not determined by the position of the disk). Sealed block valve. A valve that may be sealed fully open or fully closed during normal operation. Sealing may be achieved by applying car seals or positive locks (see the ASME Boiler and Pressure Vessel Code Section VIII, Paragraph UG-135, and Appendices M-5 and M-6). Seat. The pressure containing contact between the fixed and moving portions of the pressure containing elements of a pressure relief valve. Set pressure. The inlet gauge pressure at which a pressure relief valve or vent is set to open under service conditions. Specified burst pressure. The value of increasing inlet static pressure, at a specified temperature, at which a rupture disk device is designed to function. Contrast with stamped (or marked) burst pressure. Stamped burst pressure. The value of the pressure differential across the rupture disk at a coincident temperature at which rupture disks of identical manufacture have been determined to burst. It is derived from destructive tests performed on each rupture disk lot at the time of manufacture. See ASME Boiler and Pressure Vessel Code Section VIII, paragraph UG-137 for methods for determining the stamped burst pressure. The stamped burst pressure is marked on the rupture disk. Rupture disks that are manufactured with zero manufacturing range will typically be stamped at the specified burst pressure. Standard trim valve. A safety relief valve designed for compressible fluid service. With adjustable blowdown, the valve can be adjusted for use with incompressible fluids. Superimposed back pressure. The static pressure that exists at the outlet of a pressure relief device at the time the device is required to operate. It is the result of pressure in the discharge system coming from other sources, and it may be either constant or variable. Theoretical relieving capacity. The computed capacity expressed in gravimetric or volumetric units of a theoretically perfect nozzle having a minimum cross sectional flow area equal to the actual discharge area of a pressure relief valve or relief area of a nonreclosing pressure relief device. The flow path in a perfect nozzle is conventionally taken as isentropic. Thermal inbreathing. The movement of air or blanketing gas into a tank when vapors in the tank contract or condense as a result of a decrease in atmospheric temperature. Thermal outbreathing. The movement of vapors out of a tank when vapors in the tank expand and liquid in the tank vaporizes as a result of an increase in atmospheric temperature.


1200-149

November 2001

1200 Relief Systems


Turndown. An expression of the range of minimum and maximum flow rates expected under commonly encountered operating conditions. Commonly expressed as a ratio of minimum-to-maximum flow rate. Upper Explosive Limit (UEL) or Upper Flammable Limit (UFL). The highest concentration of a vapor or gas (the highest percentage of the substance in air) that will produce a flash of fire when an ignition source (heat, arc, or flame) is present. See also Lower Explosive Limit or Lower Flammable Limit. Importance burn.

At a concentration higher then the UEL, the mixture is too “rich” to

Vacuum relief valve. A pressure relief device designed to admit fluid to prevent an excessive internal vacuum; it is designed to reclose and prevent further flow of fluid after normal conditions have been restored. It may be either a direct-acting vent or a pilot-operated valve. Valve bonnet. The housing around the spring of an enclosed-spring pressure relief valve. Vapor depressuring system. A protective arrangement of valves and piping intended to provide for rapid reduction of pressure in equipment by releasing vapors. The actuation of the system may be automatic or manual. Velocity seal. A type of purge reduction seal designed to reduce the gradient in purge gas velocity across the flare stack diameter in order to reduce the rate of infiltration of air into a flare stack. Usually located in the flare stack just below the flare tip, this is sometimes called a venturi seal. Venting. Flow of vessel contents out the vessel. The pressure is reduced by adequate venting, thus avoiding excessive pressurization of the vessel. The emergency flow can be one-phase or multiphase. Each flow regime results in different flow and pressure characteristics. Vessel neck or nozzle. A piping connection on a vessel, commonly constructed of a short section of pipe welded to the vessel, with either a flanged or threaded end for connecting piping or instrumentation. VOC. Volatile organic compound.

November 2001

1200-150


1300 Process Alarm and Shutdown Systems Abstract This section provides guidance for the design, operation, and maintenance of process alarm and shutdown systems. These control systems are designed to minimize the time personnel or equipment are exposed to potentially hazardous conditions. The principles and guidelines included here apply primarily to new plant equipment, but the safety concepts may be useful in upgrading existing alarm and shutdown systems. This section does not include a discussion of wellhead shutdown systems, which can be found in Section 1600 of this manual. The user of this engineering guideline must recognize the complexity of each individual situation. Successful application depends on the judgement of the designer in providing a practical, functional system and on the operator to control the system appropriately. Formal project or plant safety reviews such as a Hazardous Operations (Haz-Ops) Review are important to the success of a shutdown system. Detailed guidelines on the application of shutdown systems are provided in Appendix H. These guidelines were produced for Chevron Products Company and Chevron Chemical Company by ChevronTexaco Energy Research and Technology Company. Contents

Page

1310 Predesign Considerations

1300-3

1320 Basic Principles

1300-3

1330 Protective System Design Considerations

1300-4

1331 What Should Be Shut Down? 1332 What Constitutes a Shutdown? 1340 Failure Modes

1300-6

1341 De-Energize to Trip 1342 Energize to Trip 1350 Sensing, Logic, and Processing Components

1300-9

1351 Sensing Devices


1300-1

January 2002

1300 Process Alarm and Shutdown Systems


1352 Alarms and Supporting Instrumentation 1353 Logic Systems 1354 Matching the Logic System with the Potential Hazard 1360 Emergency Block Valves

1300-20

1361 Testing EBVs 1362 Valve Types 1363 Actuator Sizing 1364 Actuator Types 1365 Fail-safe Design 1366 Control Components 1367 Electrical Power and Pneumatic Supplies 1368 Fireproofing 1369 Data Sheets 1370 Manual Resets and Bypasses

1300-40

1371 Manual Resets 1372 Bypasses 1380 Testing, Maintenance, and Documentation

1300-41

1390 References

1300-42

1391 Industry Associations 1392 Government Agencies 1393 Company

January 2002

1300-2




1310 Predesign Considerations Process alarm or shutdown systems provide interlocks, alarms, safety sequences, and trips to either prevent an abnormal, unsafe event from occurring or to minimize the consequences of that event. In most cases, the need for such protective systems depends upon a realistic assessment of the process being handled, the location, the probability of equipment failure or operator inattention resulting in a possibly dangerous situation, and the severity of the consequences. No single alarm or shutdown system design is suitable for every application. In general, a protective system should be considered if: •

Equipment malfunction or incorrect operation can result in risk of injury, damage to equipment, or lost production

•

Facility is unattended

•

Safe plant shutdown involves a critical sequencing of individual unit and equipment shutdowns

•

The operator cannot respond quickly enough to avoid a hazardous condition

•

There is a possibility of personnel being in or close to the plant during a potentially hazardous upset

•

There is a risk of release of flammable or toxic materials in the event of equipment malfunction

•

An installation represents a high capital investment

As the severity of the event or the likelihood of the event increases, the integrity of the shutdown system should increase. Redundancy, diversity, reviews, acceptance test, periodic tests, and keeping the design simple all increase the integrity of the shutdown system. Shutdown systems perform functions that are different from those performed by alarm-only systems. A shutdown system, upon receiving input that a process condition is possibly dangerous, automatically closes valves, opens valves, or activates other equipment to bring the plant to a shutdown condition. An alarm system merely informs the operator that a possibly dangerous condition has occurred, leaving the response to that condition up to the operator. Though functionally different, many design features of these two systems are identical.

1320 Basic Principles The specific recommendations for alarm and shutdown systems follow from these basic principles: •

Protective systems are designed to prevent or mitigate specific events.

•

Protective systems must protect equipment while providing acceptable unit availability.


1300-3

January 2002



•

Protective systems should be as simple as possible. They should be segregated from process functions to minimize the possibility of routine maintenance and modification work inducing common mode faults that defeat the protective function and the process function or spurious trips of the equipment being protected. Special care is needed so that power supplies and distribution permit non safety-related equipment to be powered down for maintenance without impairing shutdown system operation.

•

Protective systems must be designed so that in-service maintenance and checking by trained operators or maintenance personnel can be accomplished without actually shutting down the process. Maintenance and checking capabilities must also be available on a plant that is shut down. Whenever possible, planned shutdowns of equipment should utilize protective systems to test shutdown equipment and to train personnel.

•

Protective systems must be operated and maintained by properly trained personnel.

1330 Protective System Design Considerations Two of the most difficult questions to answer in designing a protective system are “What should be shut down?” and “What constitutes a shutdown condition?”

1331 What Should Be Shut Down? The functions of a protective system can vary significantly depending upon the process, the location, and the risk involved. It is useful to consider several “levels” (lowest to highest) of protection, as follows: •

Level 1.

Alarm only

•

Level 2.

Individual equipment shutdown

•

Level 3.

Process train shutdown

•

Level 4.

Complete process system shutdown

•

Level 5.

Emergency shutdown

Alarm Only (Level 1) The system that provides the lowest level of automatic physical protection is usually the alarm-only system. At this level, process conditions are monitored and impending problems are brought to the attention of the operator. The alarm-only level is used where the operator has time to take corrective action and remedy the problem. It is also used to indicate minor failures or abnormal operation of equipment or monitoring devices.

January 2002

1300-4




Individual Equipment Shutdown (Level 2) The second lowest level system shuts down individual units of equipment such as a pump or a compressor, or, on a production platform, a test separator. An example from a refinery is the automatic shutdown of a compressor due to mechanical failure.

Process Train Shutdown (Level 3) The next level of protection is a process train shutdown. At this level all of the process systems affected by a malfunction are shut down. An example is a production separator, one of several feeding gas to a compressor, which feeds a glycol contactor. Liquid from the separator is fed to a dedicated heater-treater and then to pumps. A malfunction in the separator shuts down the inlet to the separator, the heater-treater, and the pumps. It does not shut down the compressor and glycol systems. In a refinery, parallel trains in a sulfur removal facility is an example in which one process train could be shut down while the other continues to operate.

Complete Process System Shutdown (Level 4) The second highest level of shutdown is a complete process system shutdown, commonly in large gas processing plants. Upon detection of a possibly dangerous operating condition, such as high pressure in the inlet separator, the plant is isolated from all incoming and outgoing streams, and process equipment is shut down.

Emergency Shutdown (Level 5) The highest level of shutdown is usually referred to as an emergency shutdown (ESD). At this level, all process systems are shut down, as in a complete process shutdown. Additionally, the entire plant is usually depressured to the extent practical. This shutdown level usually results from either the automatic detection of catastrophic conditions such as fire or major gas leaks, or from operator decision and action to initiate the ESD. Not every protective system includes all levels of shutdown. For example, very few refinery facilities have complete emergency shutdown systems as defined above. Refineries are staffed 24 hours a day, and the operators are trained to respond to emergencies. On the other hand, manned offshore platforms are required by law to have complete emergency shutdown systems, even though they are staffed 24 hours a day by trained operators. In the event of a major release or fire, it is essential that personnel be in a position to quickly respond to the emergency. The highest level of protective system, emergency shutdown, enables the operators to concentrate on emergency response and life safety while the facility automatically shuts down. Generally, refinery protective systems go up to Level 2, with a few process train shutdown systems (sulfur plants, hydrotreaters). Most onshore gas processing plants have Level 4 shutdowns, and many have complete emergency Level 5 shutdown systems. Some of these locations operate unattended.


1300-5

January 2002



1332 What Constitutes a Shutdown? A shutdown of an individual piece of equipment, such as a compressor, may progress through up to three stages, as follows: •

Stage 1. Shutting down the driver and permitting the machinery to stop, while it is still pressurized with process fluid.

•

Stage 2. Automatically blocking in the equipment with shutdown valves, in addition to (1) above.

•

Stage 3. (1) and (2) above with an automatic bleed valve to depressurize the equipment.

The first stage might be implemented in the event of a process shutdown elsewhere in the facility or in the event of an equipment failure that does not pose a hazard. The second stage could be implemented upon detection of a potentially unsafe condition that does not present an immediate hazard and for which a complete depressuring would cause an extended shutdown of the facility. The third stage, which taken in conjunction with the first two is often referred to as “block and bleed,” would be resorted to in the event of a fire, major release, or any other event of similar magnitude. Blocking and bleeding protects the equipment by isolating it from hydrocarbon sources and by relieving hydrocarbons contained within that system, thus significantly reducing the amount of fuel that can be released or that can burn. Higher levels of shutdown utilize the same principles. A process train or process system shutdown would block in that facility, or portion of a facility, shutting down all equipment which does not need to be kept running. Typically, cooling water systems, air coolers, and other equipment which act to reduce the threat of possibly dangerous plant operation should be kept running. Similarly, any equipment critical to the safe shutdown of the facility, such as the fuel system feeding the flare, should be kept operating. The plant may or may not be depressured, depending upon the amount of potential fuel within the block valves and the threat of fire. In an automatic emergency shutdown, the facility would be blocked in, all equipment would be shut down except that necessary for a safe shutdown (usually a flare). In most cases the entire plant or predetermined sections of the plant would be depressured in a controlled fashion. (As a minimum the flare system must be designed for the maximum emergency depressuring rate.)

1340 Failure Modes Once it has been determined that a facility should be protected from equipment failure or operator error by an alarm or shutdown system, further evaluation can help to anticipate the result of a failure within the protective system itself. Should the plant, or unit, or an affected portion of a plant continue to operate, or should it shut down “fail safe?” This disposition following failure is referred to as the failure modes and is an essential design consideration.

January 2002

1300-6




The desired behavior of the plant upon failure of the protective system is often achieved through the use of normally energized or normally de-energized components. These components can be designed to either trip when de-energized or trip when energized. For example, a de-energize-to-trip component might be a valve, which is held open during normal operation by air supply and electrical power, but which would close upon failure of either air or power. Failures can be classified into two types, safe-failures and dangerous-failures. Safefailures are those failures that cause the shutdown system to put the plant in a safe state, typically shutting the plant down. Dangerous-failures are those failures that keep the shutdown system from properly executing its protective function. Dangerous-failures do not become a problem unless they occur coincident with a hazardous event. Dangerous-failures can be exposed by periodically testing the shutdown system. As an example, let’s look at a furnace fuel gas line with a Chopper valve controlled by a solenoid valve. The safe state is to de-energize the solenoid valve, close the Chopper, and stop the flow of fuel gas. A safe-failure would be loss of power forcing the solenoid valve de-energized and shutting down the furnace. A dangerous-failure would be the solenoid spring breaking. Now, the solenoid valve is stuck open and the safety system cannot control it. The safety system is in a dangerous mode. The safety system cannot chop fuel gas, even if a hazardous situation exists. Redundancy is a means to reduce safe-failures and dangerous-failures. A single, one-out-of-one (1oo1) configuration, has a certain reliability for safe-failures and dangerous-failures. Dual systems are either 1oo2 configuration (i.e., the system requires one of the two sensors to indicate trip in order to trip the plant) or 2oo2 configuration (i.e., the system requires two out of two sensors to indicate trip in order to trip the plant). A 1oo2 configuration is best for reducing dangerous-failures and the worst for safe-failures. A 2oo2 configuration is best for reducing safefailures and is the worst for dangerous-failures. A 2oo3 configuration is a compromise. 2oo3 provides a good reduction in both safe and dangerous failures. For a better understanding of these tradeoffs, refer to ISA S84.01 and ISA draft TR84.02; or contact the ERTC M&CS Safety Interlock Specialist. Fig. 1300-1

Redundancy Schemes Ranked Best to Worst, by Failure Type Failure Type

Failure Mode

Safe-Failures

Dangerous-Failures

Best

2oo2

1oo2

2oo3

2003

1oo1

1oo1

1oo2

2oo2

Worst


1300-7

January 2002



1341 De-Energize to Trip A protective system designed with de-energize-to-trip components is one in which a single component failure will place the equipment or unit in a safe state, usually shutdown. Typical single component failures are power or air failures, blown fuses, and broken input or output signals. If improperly designed, single circuit de-energize-to-trip designs may be prone to nuisance trips and alarms because one single component (out of perhaps hundreds) causes a shutdown. Obviously the nuisance trip and alarm are immediately detectable and can have serious impact on the productivity of the plant and operator confidence in the system. Nuisance shutdown frequency is a function of the complexity and number of components in the system. If excessive, the frequency of nuisance shutdowns can be reduced by the selection and installation of more reliable or redundant components. The design of a reliable de-energize-to-trip protective system must ensure that the system is placed in a shutdown state upon failure of any single system component. The following guidelines should be observed: •

Pneumatically or hydraulically operated actuators for shutdown valves should move to the safe position on loss of actuator power supply.

•

Solenoid valves should be energized during plant operation, and moved to the safe position when power is removed or lost (de-energize-to-trip).

•

Sensing-device contacts should be closed during normal or safe operations and should open when the shutdown condition is reached.

•

Failure of the air or electric supply to a measuring instrument used in a shutdown system should cause the output to move toward the trip condition. In some cases this will require the use of reverse-acting transmitters. If this cannot be achieved, consideration should be given to installing instruments to alarm on measuring instrument power failure.

•

The sample flow of an analyzer used in a trip system should be monitored. Loss of flow or significant reduction in flow should initiate an alarm.

•

Shutdown systems using thermocouples should be provided with a thermocouple burnout protection device. On thermocouple burnout, the converter output will be driven away from the shutdown trip setting. A separate alarm should be installed to detect the burnout.

Most shutdown systems, and all alarm systems, should be designed with de-energize-to-trip components in order to immediately bring protective system component failures to the attention of the operators. The biggest drawback to this type of design, the possibility of nuisance trips, is less of a concern now with the use of more reliable and redundant electronic equipment. Use of a de-energize-to-trip system does not guarantee that a component failure will immediately cause the plant to shut down. In fact the operators may not know that the protective system is not in working order. For example, a process lead to a field switch could become plugged, making it unable to sense a possibly dangerous operating condition, or a valve could become stuck in its normal operating position,

January 2002

1300-8




unable to move to the fail-safe position. Such equipment failures can be detected only through a regular, documented testing program.

1342 Energize to Trip In an energize-to-trip design, a single component failure will not shut down the equipment but will allow continued operation with the protective system partially or completely disabled. Facilities which cannot tolerate nuisance shutdowns, and for which a highly reliable de-energize-to-trip shutdown system is not available, may be designed with energize-to-trip protective systems. Documented periodic testing of such a system is essential to assure reliability.

1350 Sensing, Logic, and Processing Components Protective systems comprise process sensors, equipment condition or position sensors, logic performing systems, shutdown valves with actuators, interconnected wiring, tubing, air supply, hydraulic supply, power supplies, bypass switches, indicator lights, and horns. See Figure 1300-2 for a detailed drawing of an equipment shutdown system. There is a wide variety of equipment available that can perform each of these functions. The overall reliability of the system depends upon what type of equipment is selected, how it is installed, how well it is maintained and tested, and whether any redundancy has been provided. Protective systems can be powered pneumatically, electrically, hydraulically, or by a combination of all of these. Pneumatic systems have the advantages of being simple, reliable, and easy to maintain, and of requiring no electrical power. For these reasons they have been used extensively in the past on offshore platforms as well as in refinery process plants. As plants get bigger and shutdown systems become more complex, pneumatic components can become cumbersome. Electronic components, including programmable controllers, have proven their durability under severe operating conditions. Redundant and triple-redundant systems continue to improve. Most of the recently installed large shutdown systems have been electronic, regardless of location, and use programmable controllers to perform the logic.

1351 Sensing Devices The following points concerning sensing devices are important to consider when designing alarm or shutdown systems: •

Shutdown sensors (pressure, level, flow switches, etc.) should be located such that they are readily accessible for routine testing and maintenance.

•

A system shutdown device should be dedicated exclusively for shutdown purposes and should not serve in a dual capacity as a control device. A device for miscellaneous and pre-shutdown functions may serve in a dual capacity for control purposes.


1300-9

January 2002


Fig. 1300-2


Equipment Shutdown System—Detailed Drawing

January 2002

•

Sensors are usually connected directly to the process and therefore are often in contact with hot, corrosive, contaminating substances that can seriously affect sensor reliability. If point sensors of adequate design are not available, shutdowns may be connected to a signal from a dedicated transmitter. Then the problems of process interface can be handled by the transmitter, which can be monitored continuously. When using programmable controllers, transmitters can be directly wired to an analog card and the shutdown switch programmed in the software. Sometimes specially designed transmitters, such as flow, analysis, or interface level, result in a more reliable system.

•

To reduce confusion and accidental shutdowns, consideration should be given to equipping sensors with only one contact dedicated to the shutdown system.

•

Multiple microswitch contacts in sensors that are not snap acting will not actuate simultaneously. They can either shut down equipment without actuating the associated shutdown alarm, or activate that shutdown alarm without shutting down equipment. To reduce confusion and the chance of a shutdown occurring without any annunciation, interposing relays with multiple contacts driven by a single contact in the sensor are recommended. This assures simultaneous alarm and shutdown activation. Programmable controllers are ideal for providing software relays to do this.

1300-10



•


Dual setpoint devices should not be used for shutdown services except for vibration, analyses, or bearing-temperature instruments.

1352 Alarms and Supporting Instrumentation Figure 1300-2 shows, in addition to a shutdown system, a typical alarm-only system, consisting of a sensing device, an annunciator, and supporting instrumentation, including wiring, conduit, and valves for isolating the sensor from the process for testing. This schematic represents the minimum design sufficient to achieve a testable alarm system. The purpose of the annunciator is to alert the operator to an abnormal condition through the use of lights or horns. Many recently installed protective systems utilize electronic annunciators with solid-state circuitry and backlighted windows. Pneumatic annunciators are available for applications where there is no electric power or where electronic equipment may be unsuitable. Most facilities contain multiple, interconnected processes, with an upset to one often causing an upset to another. It is frequently difficult to determine which upset occurred first. Annunciators are available with first out indication, which allows the operator to distinguish which alarm occurred first, and to take appropriate action. Many new annunciation systems for larger facilities are using cathode ray tube (CRT) monitor screens and printers to communicate process information, alarms, and shutdowns to the operator. First out alarm indication may be specified. When sensing devices are connected to a shutdown system, the following should be considered: •

Alarms should be provided for each sensor.

•

All components of a protective system (alarms, trips, interlocks, etc.) that are associated with a sensor should actuate at the same time as the sensor unless a time delay sequence is required.

•

Pre-shutdown alarms can be provided at manned facilities to warn that a trip is impending. Pre-shutdown alarms enable the operator to take corrective action before the protective system is tripped. For some processes, the time taken to reach the trip condition after the operation of the pre-trip alarm might be too short for successful manual intervention. In these cases the pre-shutdown alarms may be omitted.

•

Bypass switches are commonly used to isolate sensors for testing and repair. Bypass switches should be installed for individual sensors so that the entire protective system is not deactivated during testing and repair work. Switches in bypass should be indicated locally or alarmed remotely so that operators do not accidentally operate without the full protection afforded by the system. Where practical, bypass timers or limit check timers are provided which allow testing of the protective system without shutting the process down. After a preset time the protective system will automatically go back in service. Programmable controllers allow the addition of limit check timers easily at no additional cost.


1300-11

January 2002



1353 Logic Systems Most protective systems are installed in critical applications and therefore must be highly reliable and not subject to extensive down time. This system availability depends on component selection and design and testing frequency. Protective systems may use microprocessors (e.g., programmable controllers) or be electrical, pneumatic, hydraulic or mechanical (rare). They are designed using several types of logic. The selection of the logic type should be determined by evaluating the availability and reliability of the utilities (power, instrument air, etc.) which will actuate the protective system, and by evaluating the complexity of the system. A conceptual drawing of a simple shutdown system is shown in Figure 1300-3. Figure 1300-4 shows the safety analysis function evaluation (SAFE) chart prepared for this system. The SAFE chart should be completed for offshore production facilities in accordance with API RP 14C. Fig. 1300-3

Equipment Shutdown System—Conceptual Drawing

Figures 1300-5, 1300-6, and 1300-7 indicate how logic can be performed pneumatically for this same shutdown system. Logic systems range in size from one or two components to hundreds of logic functions and many discrete components. Generally, as complexity increases, the tendency is toward electronic, microprocessor based logic systems, e.g., programmable controllers.

January 2002

1300-12



Fig. 1300-4


Safety Analysis Function Evaluation Chart (SAFE)


1300-13

January 2002



Fig. 1300-5

Simple Pneumatic Shutdown Logic

Fig. 1300-6

Pneumatic Shutdown System with Bypass

January 2002

1300-14



Fig. 1300-7


Pneumatic Shutdown System with Lock-out and Manual Reset

False trips resulting in nuisance shutdowns can result in complete abandonment of the protective system. High false trip rates are frequently cited as the reason operators bypass or disarm protective systems. Reducing the false trip frequency to an acceptable level means careful selection of components, thorough testing, and prudent use of redundancy. Documented periodic testing of protective systems is one means of ensuring a higher degree of availability. Because of the need to provide bypassing or deactivating features of some type while testing is in progress, all or part of the plant may not be protected by the shutdown system during the testing period. Through the use of additional or redundant measuring devices, and, in some cases, redundant final control elements, a significant improvement in the reliability of the system may be achieved. Protective systems using multiple sensors and dependable, electronic voting logic (see Triple Circuit System, following) to initiate action are also being used. The initial cost will be greater, but a significant payback can result from reliability, reduction of nuisance trips, reduction of failures that prevent proper operation, and the elimination, reduction, or simplification of testing procedures.


1300-15

January 2002



There are a number of factors which must be considered in order to determine whether or not redundancy is warranted. The potential hazard or loss caused by a failure, the relative likelihood of a failure, and the potential for such adverse consequences as increased nuisance shutdowns must all be evaluated. Whether a logic system is pneumatic, electrical or uses microprocessor based electronics, the following are some typical designs:

Single Circuit System. A single circuit system is one in which trip components are placed in series. In a deenergize-to-trip design, a single component failure in this circuit causes the system to trip and shut down the equipment. Another single circuit system is one in which energize-to-trip components are placed in parallel. If a single component fails in this circuit and that component becomes the trip initiator, nothing will happen. No action will occur; nor will that particular part of the shutdown system perform its protective function if called upon to do so by an abnormal operating condition. The facility could be unprotected.

Dual Circuit System (Electronic Only) A dual circuit system may be simply parallel component series circuits, but variations can be endless. The most reliable system would comprise two completely redundant parallel single circuits, from sensors through and including redundant final trip valves. The advantage of a dual circuit system is that if a single component fails the other component will continue to function. In fact, one system can be tested independently of the other. The disadvantage is the increased number of components and the consequent increased maintenance and testing.

Triple Circuit System (Electronic Only) Figure 1300-8 shows a schematic for a triple circuit logic system. A triple circuit system can receive one, two, or three independent inputs for a given abnormal process condition (i.e., three high pressure shutdown switches set to trip at the same pressure) and can evaluate each input through independent microprocessors. The system then compares the results, and if any two out of the three circuits indicate that a shutdown should occur (voting logic) then the protective system activates a shutdown. No single component failure can cause a false shutdown. A triple circuit system with majority voting has distinct operational advantages over single and dual circuit systems, as follows:

January 2002

•

The individual circuits can be completely tested on stream without shutdown.

•

At least two of the three parallel components must function to initiate a trip. The system is said to be fault tolerant.

•

Online repair is permissible.

•

Any component may be removed and replaced without risk of a spurious trip or of defeating the protective system. This assumes all components are triplicated.

1300-16



•

Fig. 1300-8


System failures and false alarms can be detected and annunciated without causing a shutdown and without losing the protection afforded by the system. Therefore, routine manual checking of the system becomes less critical.

Triple Circuit Logic System (Courtesy of Triconex)

Triple circuit systems are becoming more common in Company installations. For more information about operating company experience with this technology, contact the Monitoring and Control System Division (M&CS) of ChevronTexaco Energy Research and Technology Company (ERTC).

Programmable Controllers A number of programmable controllers have proven reliable in shutdown systems. For these programmable controllers, the manufacturers are dedicated to building a quality product, have well-defined software life cycles, have extensive quality assurance programs, and are committed to the customer. As maintenance and engineering personnel have become familiar with them, they have been found easier to troubleshoot, more reliable, less expensive to install and expand, and invaluable during initial startup due to the ease of making changes. Program design should take advantage of the built-in diagnostics available. It is highly desirable for the designer to use a fully annotated program to describe the logic to help people troubleshoot the process logic, make changes safely and keep the documentation updated. However, because of the nature of solid-state electronics, programmable controllers can suffer failures with symptoms which are quite different from the more traditional relay logic. A single component failure within a programmable controller may cause process input scanning to stop or outputs to remain static. If the programmable controller is of conservative design and


1300-17

January 2002



incorporates internal failure detection with fail-off action—including a proven reliable watchdog timer—then the programmable controller can imitate the failure mode of a relay system, reducing troubleshooting difficulty and false trips. Examples of faults that should be detected and annunciated by the programmable controller’s built-in diagnostic functions are: • • • • •

Processor failure Memory failure Remote I/O rack communication failure CPU power supply failure Battery backup failure

If a CRT based annunciator system is used, many other faults and potential problems can be detected and annunciated, such as: • • • •

Individual card failure Remote I/O failure location Program write protection disabled Forced inputs or outputs

If programmable controllers are used in a triple circuit system with voting, the importance of internal fault detection is reduced. The triple circuit becomes its own fault detector.

Distributed Control Systems Distributed control systems are being used in the industry and in many Company facilities to implement alarm and shutdown systems. Integrating alarms and shutdowns into control systems can improve safety and reduce cost if done properly. These systems can prioritize and sort incoming alarms to improve operator response to abnormal conditions that may possibly become dangerous. Where regulatory agencies permit, distributed control system equipment can be used to implement shutdown system logic, because it satisfies the following requirements: • •

Online testing of the sensors, shutdown valves, and logic. Segregation, documentation and verification of alarm and shutdown logic.

Dedicated hardware should be used both in the distributed control system and the sensing devices, and to implement protective logic in the most critical applications (e.g., personnel and environmental protection).

1354 Matching the Logic System with the Potential Hazard It is useful to determine whether or not there is an independent back up device which will offer protection in the event the primary system fails. Such backup devices could include a relief valve, a pond for spill control, a flare system including relief headers, temperature sensors to back up flame scanners for fire control, or any number of other safety features and design considerations.

January 2002

1300-18




Protective systems can also be prioritized by the potential for a possibly dangerous situation to personnel, for a significant exposure to the environment, for a significant loss of production, or finally for damage to major equipment. The safety of people and the prevention of environmental hazards must be considered first for the application of redundant or voting logic, multiple sensors and final elements. Low risk protective systems are used when the possibility of danger to people and exposure to the environment is very low, or the possible equipment damage minor and production losses insignificant. Figure 1300-9 summarizes the minimum recommendations for hardware and software suitability, based on whether an independent backup device exists. Single logic systems are generally adequate in most cases in the Company. Sensors and shutdown valves are the major source of false or nuisance shutdowns and not the logic systems. So the primary emphasis should be placed with selection, installation, and the ability to test the field devices. When a critical process variable is hard to measure, then dual sensors or a dedicated transmitter whose output can be continuously monitored should be considered. Redundant logic systems are not meant to be a solution to a problem of an insufficient number of trained personnel to test and maintain single safety systems. Also, large systems which combine shutdown logic with control logic can be a source of nuisance or false shutdowns and are more difficult to test. Smaller dedicated redundant programmable controllers, as well as two out of three voting microprocessors, should be considered. Fig. 1300-9

Recommended Logic Systems—Significant Risk to Personnel, Environment, Equipment, or Production

Independent Backup Device Exists?

Minimum Recommendation

YES

Single or dual shutdown system (new technology products acceptable)(1) (2)

NO

Dual shutdown system (proven products)(2)

Note

Any logic system is suitable for low risk applications

(1) Single systems include: Single sensors and single final control elements, and: a) Single safety relay b) Single microprocessor c) Redundant microprocessors with real-time updating & automatic switchover (2) Dual systems include: Dual sensors & dual final control elements, and: a) Single microprocessor and safety relay b) Two independent safety relays c) Redundant microprocessors with real-time updating & automatic switchover d) Two out of three voting microprocessors


1300-19

January 2002



1360 Emergency Block Valves Emergency Block (or shutdown) Valves (EBV), either isolate or divert flammable or toxic material, or they depressurize a piece of equipment. They are installed in situations that are potentially hazardous to humans, the environment, or major process equipment. You should not use a modulating process control valve as an EBV. An actuated quarter-turn or rotary process valve is the best choice. The EBV control components and logic should be separated from process control logic. Although this section applies specifically to refineries, chemical plants, and pipeline facilities, other ChevronTexaco facilities may also apply these guidelines. Valves for blocking in production oil and gas wellheads, called Surface Safety Valves or SSV’s, are discussed in Section 1700, Wellhead Control Systems. Of the five levels of protection defined in Section 1331, only Level 5 — Emergency Shutdown — requires automated valves to be EBVs. For Levels 2-4, the guidelines for EBV’s may be followed in whole or part. Decide if there is a need for EBV’s by conducting a process hazards analysis (PHA) during the initial design period and by reviewing the need during subsequent PHA’s as defined in API RP-750. A hazard and operability study (HAZOPS) is an example of a PHA. Be sure to establish a means of obtaining and maintaining accurate drawings, specifications, and other documentation for each EBV and its associated actuation system. Basically, EBV’s perform three functions: • • •

Emergency isolation Emergency depressurization Emergency diversion

Emergency Isolation To stop the release of a flammable or toxic material, an EBV is installed in the line upstream, downstream, or on both sides of a piece of equipment. This EBV is normally set in the open position and closes during an emergency. Consider installing emergency isolation block valves in areas, such as:

January 2002

•

Suction line of a pump that handles fire hazardous material but does not have a manual isolation valve that can be operated safely in a fire

•

Remote location, isolated from unmanned pumping or compressor stations

•

Fuel gas line(s) to a furnace, boiler, or fired heater

•

Line(s) to and from a tank or pressure vessel (i.e., LPG sphere) that contains flammable or toxic material (stops the flow if there is a fire, release, or line rupture)

•

Shore side of a wharf loading/unloading line (stops the flow if the line ruptures)

1300-20




Emergency Depressurization An EBV installed in the line downstream of a pressure vessel usually reduces the pressure by venting the vapor-space gases to a flare or by dumping the contents. Normally, this EBV is opened slowly after other EBV’s isolate the vessel and before vessel’s pressure relief valve opens. (e.g., Installed on the outlet of high-pressure reactors, the valve depressurizes the reactor in the event of unsafe conditions.)

Emergency Diversion An EBV re-routes flows from a plant area or empties the storage contents from one large vessel to a safe location in the facility. (e.g., Install this EBV on the inlet feed line(s) to protect a plant that contains flammable or toxic material or to empty the contents of a storage vessel of toxic or flammable material, such as hydrofluoric acid.)

Pump Suction EBVs During a fire, difficulties in isolating pumps safely from associated storage vessels and tanks can fuel the fire. For these applications, consider EBVs to protect: •

Pumps taking suction from vessels containing LPG

•

Hot pumps (operating above 600°F), connected to large reservoirs of hydrocarbons

•

Critical, long procurement lead-time pumps connected to large reservoirs of fuel

In these services, several facilities have begun using thermally-activated actuators on fire-safe butterfly valves. (A report about this application is listed under References.) Do not install the valve and actuator over the pump seal because the seal is the most likely point at which a fire may start. Locate the valve and actuator a minimum of six feet from the pump seal and outside any other fire zone.

1361 Testing EBVs To ensure that it will function on demand, an EBV must be tested routinely. Safety shutdown systems are normally dormant, and many faults in the system are not apparent. Testing the whole system uncovers covert faults; 30 percent of which may involve valves, according to one study. As part of procurement, require that knowledgeable representatives of the operating facility witness a Factory Acceptance Test (FAT) on the assembled EBV. During shutdowns and turnovers, treat EBVs as safety relief valves. Have shop personnel inspect them thoroughly, and run a leak test. During the plant operations the testing should include the: • • •

Actuated valve Sensor or initiating inputs (transmitter, switch, push button... ) Control system components and logic


1300-21

January 2002


• •


Pneumatic or electrical power sources or both Control room alarms and displays

Conduct a full test of each EBV and its associated actuation system on a regular schedule (at least every three months). If the EBV is meant to close, it is common practice to provide a means by which to bypass, block-in, and pressure test the EBV to verify that it closes tightly. Provide customized test procedures for each individual EBV function. Document in the procedures the set points at which the EBV is to activate. In test procedure documentation, detail a step-by-step process, leaving space for testing personnel to initial and date each step and to record results. In this way, the procedure becomes a comprehensive record of the integrity of the emergency system. Store completed test procedures in a secure location to be available for possible review. Guidelines are given in Section 8.4 Maintenance Systems, API 750. The EBV testing process also provides a means by which to log changes made to the EBV system. As changes are made, they must be verified by testing and incorporated formally into the testing process. During testing, the EBV should be fully stroked and actuated from all possible locations; such as, local pushbuttons, remote pushbuttons, and the automatic safety shutdown system. For an EBV on spared equipment, such as a pump, this requirement is easily fulfilled: one pump is shut down and its valve tested while the other pump is running. An EBV on unspared equipment can be tested easily if the equipment can be shut down for the short test period. If the equipment cannot be shut down during normal operation, then a duplicate EBV can be installed (see Figure 1300-10), allowing one valve to be tested while the other valve is in service. Although duplicate valves may seem costly, they allow the routine testing necessary to ensure that the EBV functions on demand. Around the fuel EBV on fired equipment, install a full-flow, hand-operated, bypass valve so that the EBV can be tested fully. Provide a means to verify full, tight closure with a downstream pressure gage (see Figure 1300-11). EBV testing by partial stroking is not recommended for fired equipment. This bypass valve should be car-sealed in the closed position, equipped with position switches, and connected to alarms and lights to warn the operator when the manual bypass valve is not fully closed. Occasionally, process requirements make partial stroking preferable to full stroking and the addition of a bypass valve. For HF Mitigation, some of the EBVs cannot be fully stroked without interrupting critical process streams. Partial stroking is preferable because: • • • •

January 2002

There is a strong potential for plugging the bypass line in this service. The past performance of the monel plug valves is excellent. The bypass valve increases the risk of external leakage. Some through-valve leakage can be tolerated.

1300-22




Fig. 1300-10 Testable Shutdown Valve Arrangement


1300-23

January 2002



Fig. 1300-11 Recommended Emergency Block Valve Arrangement

The cold flow properties of the TFE sleeve mean that monthly testing is essential to unseat the plug for all of the EBVs in HF service.

1362 Valve Types Normally, quarter-turn or rotary valves are used as EBVs. Ball and high-performance butterfly valves are frequently selected for typical process applications. For special applications, globe, plug, and gate valves are used as EBVs. Plants and projects have a piping specification which define piping classes for each service. Each piping class lists standard block valves and provides a direct reference to a purchase order description that includes a make and model number of a recommended valve. Follow these recommendations unless there is a good reason to upgrade the EBV. In the Corporation Piping Classification, GB-135169, only a few of the piping classifications list quarter-turn valves that can be adapted to standard actuators. Every classification includes a gate valve, and many include plug valves. To obtain a complete description of a ball or butterfly valve, either check the ERTC Piping Item Description Catalog or contact the Piping Systems engineer of ERTC’s Materials and Equipment Engineering Unit. Piping item numbers for butterfly and ball valves found in piping classifications are listed in Figure 1300-12 for the most common process classes: E1, H1, L1, and N1. Confirm these item numbers and their corresponding descriptions with the plant’s mechanical design group. ERTC’s Quality Assurance Group keeps an up-to-date list of acceptable valve manufacturers.

January 2002

1300-24




High-performance butterfly valves require smaller actuators than other types of quarter-turn valves. Their disadvantages are that they: • • •

Block part of the fluid path. May cause a pressure drop in the line. May not have a bubble-tight shutoff unless adjusted and maintained properly.

Fig. 1300-12 Emergency Block Valve Piping Item Numbers EBV Piping Item Numbers

Class

Service

Design Limits

Gate

Butterfly

Ball

Orbit

Plug

E1

Gen Service to 450°F, Stm to 185150#RF

185 & 450°F

L20BA3C A

L26BAXW 5

L25BA3 CB

L25BB3F B

L24BE3 GB

H1

LPG & Hyd.Carb-150#RF, CA 1/16 & SW

185 & 450°F

L20BA3C A

L26BAXW 5

L25BA3 CB

L25BB3F B

L24BA3 GB

L1


615 & 450°F

L20FA3DA

L26FAXW 5

L25FA3C B

L25FB3F B

L24FA3G B


505 & 750°F

L20FA3DA

L26FAXW 5

L25FA3C B

L25FB3F B

L24FA3G B

N1 Note

Item numbers in bold are found in GB-135169, other item numbers should be confirmed with mechanical design group.

Flangeless valves (wafer style) are not acceptable, even when metal shields are installed around the exposed bolts. Ball valves must be flanged, and butterfly valves must be lugged. Orbit ball valves are listed as block valves in many of the classifications in the Corporation Piping Specification. These ball valves are not standard quarter-turn valves, but they can be actuated with special thrust actuators made by Orbit. Gate valves do not accommodate on-off service readily for several reasons: •

Actuators in thrust service become extremely large and heavy and are very expensive.

•

Required actuator thrust is highly dependent on the shutoff pressure drop. Increased pressure drop may prevent the valve from closing in an emergency.

•

In many process services, gate valves do not maintain adequate shut-off capabilities after being in service.

•

Thermal expansion may impose moment loads on the valve body and cause the valve to stick when closing.

•

In the future, it may be more difficult for gate valves to pass the new, fugitiveemission standards.

Electric and air motors may be used to operate gate and globe valves, but they do not have fail-safe capabilities and must be energized to work. Consequently, both types of motors must be protected against loss of utilities and fire. Consider redun-


1300-25

January 2002



dant electrical power, backup nitrogen sources, reserve tanks, fireproofed electrical cables, and fireproofing on the motor actuator. Refer to Section 1365. Specifications Valves, at a minimum, should conform to their relevant industrial or national standard for through or seat leakage, e.g., API 598 or ANSI B16.34 (Petrochemical), API 6D (Pipeline), and API 6A (Wellheads). A seat leakage test is recommended for the fully assembled EBV at the FAT. For projects ordering many automated valves, test selected valves for seat leakage. API 598 Valve Inspection and Testing defines many requirements for valves in the refining process. This specification is preferred for defining the seat-leakage requirements for quarter-turn valves with piston actuators or for gate and globe valves with electric or pneumatic motor actuators. An overview of this specification shows: •

60–100 psi minimum test pressure for these valves: ball, gate, plug, and metalseated butterfly.

•

110% of the design differential pressure for these valves: globe and resilientseated butterfly.

•

No allowable leakage for resilient-seated valves.

•

Leakage rates for metal-seated valves in drops per minute (16 drops per ml) for a liquid test and bubbles per minute for a gas test. These rates apply to valves grouped in four size ranges: 2-inch and under, 3- to 6-inches, 8- to 12-inches and over 14-inches.

ANSI/FCI 70-2-1991 Control Valve Leakage defines only the seat leakage for control valves. (ANSI B16.104-1976 Control Valve Seat Leakage, which is essentially the same as 70-2, is discontinued.) This specification is preferred for defining the seat-leakage requirements of quarter-turn type control valves. Sliding-stem control valves should not be considered for EBV service. An overview of this specification shows: •

Six classes of leakage are defined, but only Class V (tight shut-off, liquids) and Class VI (tight shut-off, gases for resilient-seats) should be installed as EBVs.

•

Class V seat leakage is 0.0005 ml per minute per inch of orifice diameter per psi of the maximum service drop across the valve plug.

•

Class VI seat leakage is defined by the port diameter from one inch through eight inches in either ml or bubbles per minute. Test pressure is 50 psi.

A comparison of these specifications shows the allowable leakage for a four-inch, metal-seated valve at 100 psi differential would be 0.7 ml per minute, based on API 598; and 0.2 ml per minute, based on ANSI-70-2.

January 2002

1300-26




For applications tolerating no leakage (such as fuel gas line(s) to a furnace or boiler or a blowdown line from a reactor), install two valves in series (double block). When two valves are set in series, it is possible to locate an automated bleed valve between these two valves for fuel gas services, as shown in NFPA 85A. For other services, close the lead valve first and provide a pressure switch and alarm for leakage detection. Another solution for situations which do not tolerate leakage is to provide an EBV with a minimum leakage class rating of Class VI or bubble-tight as defined by ANSI/FCI 70-2-1991, or a resilient-seated valve as defined by API-598. For leakage checking, the installation must include a downstream block valve and pressure gage. When in hydrocarbon service, quarter-turn valves must be certified to the latest edition of API 607, Fire Test for Soft-Seated Quarter-Turn Valves, or to the Company-approved equivalent. Recently, molded covers made of K-MASS fireproofing were fire tested successfully by following API- 607 for a quarter-turn valve with an elastomer liner. The results showed that, in fire areas, non-metal seated valves are successful as EBVs in hydrocarbon service. Refer to Section 1368. Some special applications require different types of valves, such as: •

When pipeline-style, full-conduit, expanding-wedge-design, gate valves are involved in process applications above 350°F, the expanding wedge may bind in both the open and the closed positions.

•

Y-pattern globe valves and Orbit valves in high-pressure refinery applications, such as a hydro-treater recycle compressor service, have proven reliable. For high-pressure and high-temperature applications requiring tight shutoff, select the Y-pattern valve.

•

Maxon gate valves and actuators have been successful on low-pressure, cleanfuel services to fired units.

1363 Actuator Sizing Actuators must be sized properly for all emergency conditions. Ensure that the sizing and assembly is done only by valve manufacturers and by automation specialists who have demonstrated their capability to the Company. Ball and plug valve actuator sizing is usually independent of the differential pressure. Butterfly valve sizing is somewhat affected by the differential pressure, while globe and gate valve sizing is greatly affected by differential pressure. Butterfly valves must also be checked for the operating hydrodynamic torque. Maximum torque for a butterfly valve is normally the opening torque to break the seat. Maximum torque for ball and plug valves is the opening or closing torque. Plug valves that have TFE liners or are fully lined are used in corrosive services. Unless cycled at least once a month, the cold flow properties of teflon may prevent


1300-27

January 2002



the plug from moving. Maximum thrust for gate valves occurs when closing or seating the wedge. At a minimum, the actuator must be sized to handle the maximum differential pressure across the valve during an upset condition. To calculate the maximum differential pressure, assume a downstream pressure of zero (simulating a line break or complete depressurizing). The actuator can be sized for the maximum possible operating pressure, usually according to the setting of a safety relief valve. Whenever possible, anticipate any future operating requirements. The maximum allowable working pressure (MAWP) of the valve (refer to ANSI B16.34) can be used to size the actuator. The advantage of sizing the actuator with the MAWP is that the valve will work if the process conditions change during plant life. The disadvantage of this method is that the actuator may be much larger than needed for any anticipated operating conditions and therefore more costly to purchase and install. Valve damage may occur at these very high actuator torques. The valve manufacturers should verify that their valves can withstand these hightorque loads. The vendor should size the actuator, but the engineer must verify the calculations and make sure that the vendor has added the required additional torque or safety factor specified on the data sheets. This safety factor is needed after the valve has been in service. Corrosion or a buildup of deposits make the valve more difficult to seat or unseat. The safety factors vary depending on the type of valve and service conditions. For piston-actuated, quarter-turn valves, the minimum factor should be 25 percent. For dirty or fouling services, a 50 percent safety factor is often added. Consider these factors when making the final decision: •

The minimum torque available for various actuator sizes.

•

Whether the normal operating or a minimum design pressure will accommodate a dip in the air or gas pressure during an emergency.

•

Whether reserve tanks are used.

For motor-actuated valves, a safety factor as high as 100 percent is suggested, because the additional cost of a slightly larger motor is not significant. The horsepower required is directly related to the speed of closing. It is often desirable to decrease the closing time after an EBV is put into service. Unless requested otherwise, vendors typically use closing times of one minute per diameter foot of the valve. If the MAWP is used for sizing the valves that are affected by differential pressure, then no other safety factors may be needed.

January 2002

1300-28




1364 Actuator Types Pneumatic, air-operated, piston-type actuators are most often selected for EBVs. Other actuators may be chosen, such as: diaphragm, thermally-activated, electric motor, pneumatic motor, or hydraulic. To ensure the design is consistent in the plant, review the type of actuator being considered with operating personnel. Piston Actuator The piston-operated actuator has few moving parts, so maintenance is minimal. Both rack-and-pinion and scotch-yoke design piston actuators have been successful. Rack-and-pinion actuators are preferable because of their compact size, but they are typically limited to valve sizes of six inches and smaller. Though both the spring-return and double-acting piston actuators can be made failsafe, the spring-return actuator is preferred. The spring moves the valve into the fail-safe position when the control signal fails. To make the double-acting piston fail-safe, add a pressure switch and a reserve tank with sufficient capacity to move the valve to the fail-safe position. Piston-operated actuators can actuate ball and butterfly valves, ranging from NPS 2 to 48, and ANSI Class 150 to 2500 lb. One disadvantage of a scotch-yoke piston actuator is it’s size; e.g., a spring-return actuator is 72 inches long, and a double-acting actuator is 60 inches long for a 14-inch ANSI Class 300 ball valve, sized using the MAWP. This size could be a problem when trying to install an EBV in an existing plant with space limitations. Another disadvantage of any piston-operated actuator is its tendency to corrode if wet instrument air is used. Make sure that the internals have a corrosion-resistant surface, such as nickel-plating. The most common brands of piston actuators for quarter-turn process valves are: • • • • •

Automax Bettis Limitorque Rotork Worcester

Diaphragm Actuator Another simple actuator is the diaphragm with spring-return which is fail-safe because the spring moves the valve to the fail-safe position when the control signal fails. The disadvantage of this actuator is that it is limited to small valves in low-pressure service. It can actuate up to a three-inch ANSI Class 600, four-inch ANSI Class 300, or six-inch ANSI Class 150 ball valve, and up to a six-inch ANSI Class 600, an eight-inch ANSI Class 300, or a ten-inch ANSI Class 150 high-performance butterfly valve.


1300-29

January 2002



Do not install diaphragm actuators without fireproofing in fire hazardous areas. The yokes are often cast iron which, in a fire, may break when hit by a stream of cold water, thus causing the valve to fail open. Often, both the diaphragm actuator and valve are supplied by the same manufacturer. Coil-Spring Actuator Coil-spring-return actuators, made by Posi-Seal, have recently been selected for suction lines of hydrocarbon pumps. They can be thermally-activated with fusible links which melt in case of fire. They are listed in the FM Approved Equipment Handbook. Once tripped by either the links or by a remote electric or pneumatic signal, the valve and actuator must be opened manually with the handwheel. Melted links must of course be replaced. A report entitled, Thermally-Activated Fire-Safe Emergency Block Valve, is listed in the References. Electric and Pneumatic Motor Actuators Electric and pneumatic motor operated actuators often operate very large valves. They also actuate high-pressure globe valves, which require very high thrust loads to shut in compressors. Normally, electric motors also operate gate valves because of their high thrust loads. Their disadvantages are that they cannot be made fail-safe, that they have more moving parts, and that they require more maintenance. When used as EBVs, the electric motor’s power supply should be redundant or be backed up by an emergency power generator. In a fire zone, the motor, actuator and control housing, and the conduit must be fireproofed. Refer to Section 1368. Both electric-and pneumatic-motor actuators work with an adjustable torque switch to control the seating of the valve. Note A fail-safe quarter turn electric actuator is available from Biffi (a Tyco Flow Control company). The Biffi actuator is Type EHS described in Biffi document DT 1295. See www.biffi.it and www.tycoflow.com. To avoid corrosion problems, air or gas supply for pneumatic motors must be clean and dry. As EBVs, pneumatic motors must have a very secure source of gas, such as instrument air backed up by nitrogen. Pneumatic and electric motor actuators can operate gate, globe, ball, Orbit, and butterfly valves, ranging from NPS 1 to 36 and ANSI Class 150 to 2500. See Figure 1300-13 for typical electrical motor installation. The most common brands of electric motor-driven actuators are: • •

January 2002

Limitorque Rotork

1300-30




Fig. 1300-13 Typical Electric Motor Operated Actuator and Emergency Block Valve Installation

Limitorque also makes pneumatic, motor-driven actuators.

Hydraulic Actuator The hydraulic actuator requires excessive maintenance and therefore should be selected to operate an EBV only when no other actuator will work. A hydraulically powered actuator has its own hydraulic fluid tank. Although this tank is dedicated to the actuator and valve for an emergency, should the hydraulic system develop a leak, fluid is lost. In an emergency the actuator may not develop the pressure required to move the valve.

Manual Overrides for Actuators Manual override may be specified if it will disengage automatically when the remote signal for the actuator is activated to trip the valve to the safe position. Unfortunately, only electric and pneumatic motor actuators and thermally-activated, coil spring actuators have this capability. Piston and diaphragm actuated EBVs should not have manual overrides, because the valve cannot be tripped after it has been opened manually by an override. If an EBV is designed properly, tested at the factory, and tested regularly at least every three months, it should respond whenever a demand is placed on a system. Overrides, therefore, should not normally be required nor used for quarter-turn or diaphragm valves.


1300-31

January 2002



There are three basic types of manual overrides: • • •

Gear box Hydraulic Jackscrew

Gearbox manual overrides are the most dependable, maintenance free, and physically compact. Unfortunately, they often cost as much as the actuator. Hydraulic overrides depend on the operation of a hydraulic pump and eventually need maintenance. They are less costly than a gear box override. Jackscrew overrides are inexpensive but should have a handwheel attached. Unfortunately, the exposed threads rust over time. Also, the operator expends significant amounts of time and effort to move the valve, which is impractical in an emergency. Pneumatic wrenches can facilitate testing jackscrew overrides. To operate the manual override, the force required at the rim of the handwheel should not exceed 80 lbs. Check the size of the handwheel and the number of turns required to open or close the valve fully to make sure the handwheel is neither too large nor the number of turns required too many. The handwheel must also be accessible. If the actuator is not at grade but up a pipeway; then provide safe, unobstructed access for the operator to reach and operate the handwheel. For an electric or pneumatic motor actuator used on an EBV, it should be impossible to lock the motor in the manual or hand position.

1365 Fail-safe Design A fail-safe design of an EBV installation means that the valve will move to the safe position if either the electrical or the pneumatic power supply to the actuator fails. As mentioned previously, the results of a PHA are essential during the initial design period. The PHA should consider a fail-safe design and determine if other control or alarm signals are necessary for operating safely, for avoiding spurious trips, and for shutting down the plant and its equipment. The design of the interlock logic should be based on the outcome of the PHA.

De-energized-To-Trip Design Fail-safe or de-energize-to-trip is preferred for the design of an EBV. In general, design the EBV system so that loss of motive air or electrical power produces a safe response from the EBV. Production (onshore and offshore), the chemical facilities, pipeline, and Warren Petroleum all follow this design. The advantage of fail-safe design is that it initiates without a control system or operator action. The disadvantage is that, unless backup power and pneumatic systems

January 2002

1300-32




are incorporated in the design, it may go into the failed position if the system is accidentally de-energized. There are two fail-safe actuators: spring-return-piston and diaphragm actuators. When the control signal to the actuator fails, the springs close or open the valve. The control signal can, however, fail if there is an electrical or pneumatic power loss or if the control system itself fails. The spring-return actuator in a de-energize-to-trip service can be supplied with reserve tank to eliminate spurious tripping caused by the loss of pneumatic supply pressure. When the installation of the EBV is designed to be fail-safe, the engineer must consider all the equipment affected by the opening or closing of this valve. For example, if the actuator and valve are on the suction of a pump, the pump must be designed to shut down when the EBV closes. Figure 1300-14 shows the Typical Chopper Valve Arrangement for High Reliability, De-Energize-To-Trip (Fail-Safe) System for a process furnace using a triple modular redundant logic system. Fig. 1300-14 Dual solenoid Valve Installation in a De-Energize-to-Trip (Fail-Safe) System for a Critical Fuel Gas EBV

Energized Design Historically, to avoid spurious trips, refineries have chosen an EBV that has an operator-initiated energize-to-trip design rather than a fail-safe design.


1300-33

January 2002



There are three major problems with an energize-to-trip design: 1.

The first problem is that, if the fire or toxic spill is between the operator and the pushbuttons, the operator may be unable to reach the pushbuttons to activate the actuator and valve. To reduce but not eliminate this problem, relocate the pushbuttons to a safe area (as discussed in Section 1366).

2.

The second problem is that, unless tested frequently, the system reliability is in question.

3.

The third problem is that the emergency shutdown system may be damaged by a fire or explosion. Within a fire hazardous area, fireproofing provides limited protection (approximately 30 minutes) for the system during a fire. An explosion or another type of emergency may damage the system immediately.

To resolve these problems, provide the double-acting actuator in energize-to-trip service with a reserve tank of sufficient capacity to allow three complete cycles back and forth to its failure position. The reserve supply should: •

Be connected to the pressure source with double check valves to prevent back flow.

•

Have a pressure switch to move the valve to the failure position automatically when the pressure in the reserve tank drops to a critical level.

•

Have an alarm to warn the operator of a drop in the supply pressure.

Electric and Pneumatic Motors Pneumatic and electric motors are not inherently fail-safe. They must be energized to operate and move the valve to the safe position. Motors usually operate EBVs with gate and globe valves because they require a high thrust to work. To make the systems reliable, consider these factors: testing, fireproofing, and accessibility:

January 2002

•

Establish a frequent testing of the motors, preferably once a month.

•

Fireproof motors, any controls, and electrical conduit inside the fire hazard area.

•

Back up the electrical power to the starter motor with an emergency power generator or make sure the motor control center is double-ended and fed from two independent power sources.

•

For pneumatic motors, install reserve tanks with a low-pressure alarm and double-check valves on the supply line or back up the supply pressure with a secure form of gas pressure, such as a plant-wide nitrogen system.

•

Review the time needed to operate these emergency valves fully because the horsepower required is in proportion to the speed of operation.

•

Make sure spare motors are available when problems are highlighted during testing.

1300-34




•

Train maintenance people on these specialized drivers.

•

Provide safe access during an emergency to the manual override gearbox.

1366 Control Components The control components which may be required for the installation of an EBV are as follows: • • • • • • • •

Auxiliary limit switches Solenoid or pneumatic pilot valves Local and remote pushbuttons Fusible plugs Position indication Pneumatic reserve tanks Speed control Interlock logic and alarms

The control system for remote actuation is the final part of an EBV system. Because many details and variables must be considered, this is the most time-consuming phase of the design.

Auxiliary Limit Switches Auxiliary limit switches have two purposes: •

To shut down or operate associated equipment when the EBV either starts to move or has completed its movement.

•

To send the valve position indication signals and alarms to remote locations.

The application hardware for installing limit switches on piston actuators for quarter-turn valves has improved significantly. Historically, proximity switches have proven to be highly reliable and are now packaged inside NEMA 4 enclosures which can be stacked on top of the EBV actuator or on the stem of the bypass valve. SPDT contacts are standard. If the proximity switches are U.L. Listed for Class I, Div 2 areas, no arcing contacts are present; and, therefore, they do not require explosion-proof enclosures and conduit seals. On P&IDs and control drawings, these limit switches should be labeled ZSO and ZSC for position switch open and position switch closed respectively. For electric and pneumatic motors, auxiliary limit switches normally are microswitches with open or arcing contacts. Typically, four or more adjustable switches are available for open, closed, and intermediate positions.

Solenoid Valve and Pneumatic Pilot Valve A solenoid valve (a combination control solenoid and three-way air valve) is necessary with a remote, electrical, hand switch or shutdown system.


1300-35

January 2002



A de-energize-to-trip or fail-safe solenoid valve is recommended. To prevent spurious trips, back up the power source with an uninterruptible power supply (UPS) (see Section 1300 of the Electrical Manual for more information on UPS) and fireproof the control conduit; or locate the solenoid valve outside the fire zone. Figure 1300-14 shows how to install dual solenoid valves to reduce even further the possibility of a spurious trip in a de-energize-to-trip system for a critical fuel gas EBV. Both solenoid valves must trip to close the fuel gas valve. An energize-to-trip solenoid valve, however, must be connected to a power source backed up by a UPS. The conduit must be fireproofed inside the fire zone. Increase the reliability of an energize-to-trip system by installing dual solenoid valves and routing the conduits separately. Energizing either solenoid valve would actuate the EBV to its safe position. With a pneumatic pilot valve, take care to ensure that it is not in the fire zone and is suited to the environment.

Remote Pushbuttons Remote actuation capability is important because the initial release of hazardous material can occur without a fire. If hazardous material is released, the operator may find it difficult to enter the area safely to isolate and shut down the equipment. Conversely, with remote actuation, the operator can isolate the equipment from a safe location. Locate pushbuttons a minimum of 50 feet from the actuator and valve and outside the fire-hazardous area zone (see the Fire Protection Manual, Section 1700). Placing a protective shroud over the stop pushbutton helps to prevent accidental activation. In the field, make pushbuttons easily visible and label them clearly and permanently (white letters on a red background) with a sign or nameplate. In panels and consoles, arrange all pushbuttons or switches carefully to avoid operator confusion. Operators, under the stress of an emergency situation, may shut down the wrong equipment if shutdown switches, for different process units or equipment, look the same and are grouped together. Physically separate the shutdown switches and label their functions boldly. For maintenance safety, make sure pushbuttons can be locked in the de-energized position. If pushbuttons are part of an automatic shutdown system, the action which they initiate should latch the system so that the EBV does not return to its original position when the pushbutton is released. The capability to actuate electric motors may be required both on the electric motor actuators and from at least one set of remote pushbuttons. The remote pushbuttons must be able to override the local pushbuttons on the motor. The location of the remote open, close, and stop pushbuttons will vary with design requirements. These pushbuttons may be required in the plant, in the control house, or at both locations.

January 2002

1300-36




The open and close pushbuttons for electric motors should require two actions to activate, such as pushing in and turning to the maintained position. Activating the stop pushbutton should require only the push action. Consider installing a protective shroud over the stop pushbutton to prevent accidental activation.

Fusible Plugs A fusible plug, installed in the tubing run above a potential fire source, can send either a spring-return or double-acting piston actuator to the failure position during a fire.

Position Indicator For quarter-turn EBVs, a highly visible, impact-resistant, beacon-type, valveposition indicator can be stacked on top of the limit-switch enclosure (or on top of the actuator if limit switches are not required). Electric motors are provided with full open and closed position indication. Continuous position indication can also be provided both at the motor and to a remote location.

Position Indicating Lights Position indicating lights may be required for the EBV and for the manual bypass valve (if there is one). Locate these lights in the field, in the control house, or in both places. If position lights are required, there should be at least two lights: one for the open position and one for the closed position. Locate the lights next to their switches. For position lights located in the field, be sure to shade the lights from the sun so that the operator can see if the lights are on or off. Push-to-test lights are useful, allowing the operator to check the bulbs. There are two conventions for specifying which of the lights is red and which is green: 1.

The green light indicates the normal position, and the red light indicates the abnormal position (the position the valve will be in during the emergency), or

2.

The green light indicates that the valve is open, and the red light indicates that the valve is closed.

The engineer should ask the operating representative which convention is followed by the particular facility.

Pneumatic Reserve Tank To ensure a source of power during an emergency, double-acting, energize-to-trip piston actuator requires a pneumatic reserve tank. To minimize the probability of a spurious trip, add a reserve tank to the springreturn actuator.


1300-37

January 2002



The reserve tank should be (a) sized to cycle the valve three times, (b) connected to the pressure source with double check valves and also have bleed valves for checking. A pressure switch on the pneumatic supply line to the tank warns the operator of a problem with the source of pressure. Pressure gages can also be placed on the reserve tank and at the pressure switch. To provide a fail-safe, double-acting piston actuator, add another pressure switch which will move the valve automatically to the failure position when the pressure in the reserve tank drops below a critical level.

Speed Control When a valve is closed quickly in a line containing flowing liquid, the inertia of the flowing liquid increases the pressure at the valve. This effect is called surge, and the increase in pressure is called surge pressure. Surge pressure can cause such extremely rapid changes in pressure that it results in the metallic percussions commonly called water hammer. The surge pressure wave then propagates back up the line and can cause mechanical damage. High fluid velocities and long runs of pipe increase the possibility of water hammer. In ChevronTexaco’s Fluid Flow Manual, Section 800, there is information about surge pressure and methods for calculating it. To minimize surge pressure peaks, specify devices to control the closing rate of the valve and test them during the FAT. Normally, for quarter-turn valves, the speed is controlled by an adjustable needle valve.

Interlock Logic and Alarms The basis for the interlock logic is a well-executed PHA (conducted during the initial design period) which becomes the basis for hardware selection and software design. This subject is covered in previous parts of Section 1300.

1367 Electrical Power and Pneumatic Supplies The power supply and control signals for the actuator of the EBV are electric, pneumatic (gas or air), or hydraulic fluid. In each of these cases, the power and control system driving the actuator must be available during an emergency condition. To avoid spurious trips during normal operation, power sources should be very dependable, especially for fail-safe EBVs. An electrically controlled actuator must be connected to an emergency power source so that it will function during any emergency, including a power outage. This power source should be backed up by a UPS (see Section 1300 of the Electrical Manual for more UPS information). If the actuator cannot be connected to an emergency power source, then electricity should not be the power supply. Power to the electrical motor’s starter should be backed up by an emergency power generator or the motor control center should be double-ended and fed from two independent power sources.

January 2002

1300-38




A pneumatically powered or controlled actuator (or both) must be connected to a highly reliable source of clean and dry air or gas, which will not be lost during an emergency. To ensure that the EBVs have power and control signals if the plant loses air, install reserve tanks or an air system backed up by a plant-wide nitrogen system for emergency use. Dedicated high-pressure air or nitrogen cylinders also provide a reliable power source. Make sure the air or gas is clean and dry; or the actuator may fill up with water, corrode, and malfunction during an emergency.

1368 Fireproofing If possible, do not install an EBV and its actuator in any fire zone (as defined in the Fire Protection Manual, Section 1700). If a non-fail-safe EBV is installed in a fire-hazardous area, then fireproof all components within the fire zone. These components include the actuator, the wiring, the control devices and the valve (if it has an elastomer seat or liner). It is normal practice to mount control devices, such as limit switches, solenoid valves and air filters, in a fireproofed enclosure. Any non-fireproofed component of a non-fail-safe EBV is a weak link during a fire and will cause the installation to malfunction. The Company-approved methods for fireproofing the actuator, wiring, and control devices are presented in Section 1700 of the Fire Protection Manual. Consider fireproofing fail-safe EBVs in a fire-hazardous area if the EBV is to be operator initiated, rather than fire initiated. Historically, fire blankets and rigid enclosures providing protection have required conscientious, ongoing maintenance. A maintenance-free, epoxy-type, intumescent, spray coating, called K-MASS, has been developed and tested successfully. The ½-inch thick coating is sprayed and bonded to actuators and over control enclosures. During a fire, it expands to provide protection for the equipment for over 30 minutes. Recently, molded covers made of K-MASS were fire tested successfully on a quarter-turn valve with an elastomer liner. The results, detailed in the Southwest Research Institute Report Number 6-302, showed that soft or non-metal seated valves can be successfully used as EBVs in hydrocarbon services within fire areas. The covers are installed in the field after the flange bolts are tightened. For inspection or adjustment of the valve, the covers can be removed easily with special tools. For HF Mitigation, the operating plants chose a monel-plug valve with a TFE liner for their EBV’s. This valve would not pass the through-valve leakage requirements in Edition 3 of API Standard 607, Fire Test for Soft-Seated Quarter-Turn Valves. The historical, industry-wide performance of these plug valves in this service, however, was excellent compared to the specially designed gate valves. For any EBVs located in fire-hazardous areas, consider adding special, molded, fireproofing covers made form K-MASS over the valve body.


1300-39

January 2002



For any questions on fireproofing methods, refer to the Fire Protection Manual, Section 1700, or consult with the Fire Protection Staff in the Health, Environmental, and Loss Prevention Division or the Materials Unit.

1369 Data Sheets Use the following data sheets and their associated data sheet guides when ordering an emergency block valve: • • • •

For a quarter-turn valve with a piston actuator - ICM-DS-4892(1) For a valve with a diaphragm actuators - ICM-DS-4892(2) For a valve with an electric motor actuator - ICM-DS-4892(3) For a valve with a pneumatic motor actuator - ICM-DS-4892(4)

Use also these data sheets and data sheet guides to specify normal on-off process control valves.

1370 Manual Resets and Bypasses 1371 Manual Resets It is recommended that all protective systems be provided with a manual reset. Once a shutdown system has been activated, the plant or process remains shut down until an operator clears the protective system, preferably through the use of manual reset levers located at each ESD valve and a reset button at the protective system control panel. This encourages the operator to investigate the cause of the shutdown, as well as to physically assess the effects of the shutdown on the plant before starting it up again.

1372 Bypasses Alarm systems should not be bypassed. Shutdown systems should not be designed to run permanently in a bypassed state. Such a design would suggest a re-evaluation of the need for a protective system in the first place. Nevertheless, there are occasions when a shutdown system must be bypassed. These include testing, maintenance and certain startup conditions. Shutdown systems should be designed to be tested while the plant is operating. This is accomplished through the use of bypass switches. An example of a startup bypass is a low flow switch, which activates a shutdown. During startup, low flow conditions exist which would normally trip the switch. The preferred design for this example would include a timing circuit, which would automatically re-activate the shutdown system after a predetermined period of time. If a sensor or circuit is manually bypassed, it is recommended that a control room alarm and a field light or some other equivalent means be provided to alert operators to the bypassed condition of the protective device. These devices are shown in Figure 1300-2. Programmable controllers with CRT alarm systems can have dedi-

January 2002

1300-40




cated bypass summary displays and reports that can be printed showing the time and date of all outstanding bypasses. It is recommended that those times when a protective system is bypassed be clearly documented.

1380 Testing, Maintenance, and Documentation Periodic testing of the protective system is essential for several reasons, as follows: •

Testing provides a basis for recording the system performance. A good performance record creates operator confidence that the system will work when necessary. A poor performance record indicates that improvements are necessary.

•

Testing exercises the trip valves. Trip valves that are opened and closed on a periodic basis will get the preventive maintenance they need.

•

Testing will teach personnel how the system works. Trained personnel should be able to quickly determine the cause of the shutdown, correct the problem and start the unit back up.

•

Test frequencies and test results should be documented (a legal requirement for offshore facilities).

•

For each protective system a complete and accurate record of all alarm and shutdown trip points should be maintained along with proper documentation of corrective actions or shutdown functions performed at each trip point. For offshore installations, this documentation is required by a safety analysis functional evaluation (SAFE) chart as shown in Figure 1300-4.

During the early stages of a project, a description of the process logic for the shutdown system should be developed. It could be in the form of a word description, sometimes called a Control Philosophy (see Figure 200-8 of Section 200), or as an ISA-type logic diagram (see ISA 5.2). A SAFE chart may be adequate for platforms. This description will be used by operations for review and training, during any formal safety reviews (e.g., a haz-ops review), and to select the system hardware and develop the programming logic. Programmable controller-based shutdown systems should have fully annotated program documentation. Not only should these show input and output descriptions and instrument tag numbers, but they should also have a complete text description of the shutdown logic and the hazards involved. This information will be invaluable for operating, testing, troubleshooting, and documenting changes over the life of the system.


1300-41

January 2002



1390 References 1391 Industry Associations American Petroleum Institute (API) 10 1.

Specification 6D Specification for Steel Gates, Plug, Ball and Check Valves for Pipeline Service.

2.

RP 14C Recommended Practice for Analysis, Design, Installation and Testing of Basic Surface Safety Systems on Offshore Production Platforms.

3.

RP 14D Specification for Wellhead Surface Safety Valves for Offshore Service.

4.

RP 14F Design and Installation of Electrical Systems for Offshore Production Platforms.

5.

RP 14G Fire Prevention and Control on Open Type Offshore Production Platforms.

6.

RP 500 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Division 1 and Division 2.

7.

RP 505 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, ZOne 0, Zone 1, Zone 2

8.

STD-607 Fire Test for Soft-Seated Quarter-Turn Valves.

9.

STD-598 Valve Inspection and Testing.

American National Standards Institute (ANSI) 10. B2.1 Pipe Threads (Except Dryseal) Specifications, Dimensions, and Gaging for Taper and Straight Pipe Threads Including Certain Special Applications. 11. B16.5 Pipe Flanges and Flanged Fittings. 12. B16.34 Valves—Flanged and Buttwelding End Steel, Nickel Alloy, and Other Special Alloys. 13. FCI 70-2-1991 Control Valve Seat Leakage. 14. B31.3 Petroleum Refinery Piping.

National Association of Corrosion Engineers (NACE) 15. MR-01-75 fide Stress Cracking Resistant Metallic Material for Oil Field Equipment.

National Fire Protection Association (NFPA) 16. 70 National Electrical Code. 17. 85 A, B, D Prevention of Furnace Explosions. 18. 496 Purged and Pressurized Enclosures for Electrical Equipment.

January 2002

1300-42




Oil Companies Manufacturing Association (OCMA) 19. FSV.1 Testing for Firesafe Valves.

American Society of Mechanical Engineers (ASME) 20. Section VIII Pressure Vessels.

Manufacturers Standardization Society of the Valve and Fittings Industry (MSS)

21. SP-61 Pressure Testing of Steel Valves.

Instrument Society of America (ISA) 22. S5.2 Binary Logic Diagrams for Process Operations. 23. S18.1 Annunciator Sequences and Specifications.

Southwest Research Institute 24. SwRI Test Report Number 6-302 Fire Test for K-Mass Insulation for SoftSeated Quarter-Turn Valves, dated October 17, 1991.

1392 Government Agencies United States Department of the Interior—Minerals Management Service (MMS) 25. Outer Continental Shelf (OCS) Order No. 2 Drilling Procedures. 26. Outer Continental Shelf (OCS) Order No. 3 Plugging and Abandonment of Wells. 27. Outer Continental Shelf (OCS) Order No. 5 Subsurface Safety Devices. 28. Outer Continental Shelf (OCS) Order No. 7 Pollution and Waste Disposal. 29. Outer Continental Shelf (OCS) Order No. 8 Platforms, Structures, and Associated Equipment. 30. Outer Continental Shelf (OCS) Order No. 9 Oil and Gas Pipelines. 31. Outer Continental Shelf (OCS) Order No. 11 Oil and Gas Production Rates, Prevention or Waste, and Production of Correlative Rights. 32. Outer Continental Shelf (OCS) Order No. 13 Production Measurement and Commingling.


1300-43

January 2002



Note The referenced OCS orders contain legal requirements for offshore facilities. For example, OCS Order No. 5 specifically requires safety system designs to comply with both API RP 14C and API RP 500B. The OCS orders also delineate the type and extent of design documentation (flow diagrams, piping diagrams, electrical schematics, etc.) which must be provided for Department of the Interior (DOI) approval, as well as testing requirements for the system, including testing frequencies.

1393 Company 33. Thermally-Activated Fire-Safe Emergency Block Valve Report, dated July 1991, ERTC.

January 2002

1300-44


1400 Intrinsic Safety Abstract This section provides guidelines for the selection of intrinsically safe equipment that can be used to design intrinsically safe instrumentation systems. A number of applications are presented as well as sample calculations for specifying intrinsic safety barriers. Contents

Page

1410 Introduction

1400-2

1420 Intrinsic Safety

1400-2

1421 Reasons To Use Intrinsically Safe Designs 1430 Cost Considerations

1400-4

1440 Intrinsic Safety Standards

1400-4

1441 General 1442 Types of Equipment Certification 1443 Definition of Simple Apparatus 1444 Certification Requirements 1450 Fundamental Requirements of Intrinsic Safety

1400-9

1451 Design and Installation Considerations 1452 Protective Components 1453 Grounding and Shielding 1454 Maintenance of Intrinsically Safe Systems 1455 Inspection and Commissioning 1456 Documentation 1457 Typical Applications 1458 Sample Calculations for Cable and Barriers 1460 References

Chevron Corporation

1400-22

1400-1

July 1999



1410 Introduction Intrinsic Safety is a design and construction method that can be applied to electrical instruments and their interconnecting wiring for safe use in a hazardous (classified) location. The intrinsically safe method is preferred in some Company facilities. This section includes: •

The advantages and disadvantages of intrinsically safe, explosionproof, nonincendive, and purging methods of design and construction

•

Selected applications for intrinsically safe designs

•

Industry standards and requirements

•

Sample calculations for specifying intrinsic safety barriers

1420 Intrinsic Safety Intrinsic safety is the result of the design, manufacture, installation, and maintenance of electrical equipment that limits the energy in an instrument circuit. A hazardous or classified area or location, as defined by the National Electrical Code (NEC), is an area where fire or explosion hazards may exist due to flammable gases or vapors, flammable liquids, combustible dust, or ignitible fibers or flyings. Intrinsically safe wiring and equipment is incapable of releasing enough thermal or electrical energy to cause ignition of a specific hazardous mixture in its most easily ignited concentration. This definition applies to service in both normal operation and in abnormal fault conditions. The term specific hazardous mixture is the most hazardous material composition possible at a specific temperature and pressure. Barrier intrinsic safety uses electrical barriers as the primary means of limiting energy. The barrier is also an isolation device which electrically separates equipment in the nonhazardous location from the equipment in the hazardous area. Current and voltage-limiting power supplies also provide this energy limiting/isolation function. This section explains the technique of using electrical barriers. The requirements for equipment located on the nonhazardous side of the barrier are: 1.

Equipment must not use or generate a voltage exceeding 250 volts.

2.

Proper wiring and grounding practices must be followed.

The requirements for equipment on the hazardous side of the barrier are:

July 1999

1.

Each field circuit must have its own barrier.

2.

Field devices must comply with either system or entity approval standards for intrinsically safe installations.

3.

Proper wiring and grounding practices must be followed.

4.

Total capacitance and inductance (including the cable) must be below certain levels.

1400-2

Chevron Corporation


5.


Manufacturer’s documentation and user calculations must be obtained and maintained.

1421 Reasons To Use Intrinsically Safe Designs Circuits and systems designed to be intrinsically safe reduce, but do not eliminate, the possibility of igniting hazardous substances through equipment operation or malfunction. The margin of safety for properly designed and maintained intrinsically safe systems is always greater than that of alternative methods. The reason for this added margin of safety is that the components have been tested in a wide range of normal and fault conditions to insure that they do not release the energy needed to contribute to an ignition. Other techniques, such as air purging, provide protection only as long as the integrity of the equipment enclosure is maintained. A loss of supply air or the inadvertent opening of a purged enclosure without the power being disconnected first, for example, would render the purge useless. Intrinsically safe designs improve the maintainability and testability of circuits and systems. Intrinsically safe equipment in a hazardous area may be worked on live (i.e., circuits may be tested with all equipment energized and operating). This saves time, labor, and the extra equipment needed to insure that the environment in which the circuit is installed is free from flammable mixtures. Accessibility is also increased since intrinsically safe construction may not require explosionproof construction, which is necessary for other design methods. Thus, cost savings may be realized by reducing both the amount of explosionproof equipment required and installation and labor costs. However, some operating locations prefer to install intrinsically safe equipment and wiring in explosionproof or purged housings and conduit for physical and environmental protection. The cost of such explosionproof construction is substantial. The cost of intrinsically safe instrumentation is usually higher than that of equivalent non-intrinsically safe products. It is unlikely, however, that this premium by itself would be of sufficient magnitude to be a determining factor. The same conclusion is true for incremental design costs. Costs are likely to be less when completely new electronic instrumentation systems are being installed than when small additions are made to existing older, nonintrinsically safe systems. The benefits of intrinsic safety are (1) the incremental cost of intrinsically-safecertified equipment is more than offset by the savings in installation material and labor costs (laying armored cables in cable trays versus pulling cables in overhead or buried conduits, using weatherproof versus explosionproof terminal boxes, etc.); and (2) maintainability is improved in the operating environment. At facilities where the full advantages of intrinsic safety are not utilized, and cables are still installed in conduits, intrinsic safety is justified on the basis of eliminating explosionproof enclosures and improving maintainability. Before adopting intrinsically safe design techniques, it is essential to understand hazardous area classification and the types of equipment that are suitable for these areas. See Section 300 of the Electrical Manual.

Chevron Corporation

1400-3

July 1999



1430 Cost Considerations Two factors currently control Company policy with respect to the cost of intrinsically safe installations compared to other alternatives. 1.

Facilities that have not used intrinsically safe designs continue not to because of the perceived added costs.

2.

All facilities that originally used intrinsically safe designs typically continue the practice. Maintenance training costs are absorbed as normal operating costs. These facilities balance training and capital costs with savings that result from easier equipment maintenance. In addition, construction costs for intrinsically safe systems are significantly lower.

A $150 to $500 per loop hardware cost differential between general purpose and intrinsically safe construction and equipment can be expected. This cost difference includes barriers mounted in I/O cabinets and changes in enclosures to accommodate the spacing and partitioning requirements. The greatest potential for cost reduction is in the installed cost of explosionproof enclosures and rigid conduit compared to the cost of general purpose enclosures and armored cable in cable tray. Company experience shows that installed cost reductions of 25 to 33% are realistic. There are many variables to be considered, including the size and number of enclosures needed, the number of multi-conductor cables required, and the complexity of the tray or conduit runs.

1440 Intrinsic Safety Standards 1441 General The following is a brief overview of the certification authorities for the United States, Canada, and Europe. The publications referenced in this section are cited fully in Section 1460.

United States The National Fire Protection Association (NFPA) publishes a standard which is applicable to intrinsic safety. The National Electrical Code (NFPA 70) provides guidance on the installation of electrical equipment in hazardous locations, and is recognized as a standard by the American National Standards Institute (ANSI). NFPA does not test and/or certify equipment. Underwriters’ Laboratories Inc. (UL) is an organization that prepares its own standards and provides manufacturers with testing and listing services that signify adherence to UL standards. Document UL 913 provides construction and testing requirements for intrinsically safe apparatus and associated apparatus. Local approval authorities at some locations, such as Los Angeles, require UL certification. Certification by other organizations is not accepted.

July 1999

1400-4

Chevron Corporation



Factory Mutual Research Corporation (FM) is another U.S. organization that prepares its own standards and provides manufacturers with testing and listing services that signify adherence to FM standards. Document FM 3610 also provides construction and testing requirements for intrinsically safe apparatus and associated apparatus. The technical contents of UL 913 and FM 3610 are the same. The Instrument Society of America (ISA) publishes an ANSI-adopted standard, ANSI/ISA-RP 12.6. This document is a guideline for safe design and installation practices of intrinsically safe wiring.

Canada The Canadian Standards Association (CSA) maintains standards and certified electrical apparatus for hazardous locations in Canada. CSA approval is acceptable to local approval authorities at some Company locations in the United States.

Europe and International The International Electrotechnical Commission (IEC) is the international body for electronic standardization. It is affiliated with the International Organization for Standardization (ISO), but has technical and financial autonomy. The IEC’s standards form the basis for the standards adopted by the European nations (including the United Kingdom) who are members of the electrotechnical committee called the European Committee for Electrotechnical Standardization (CENELEC). Under the CENELEC system, each member nation has its own certifying authority. In the United Kingdom, for example, the certifying and testing agency is the British Approvals Service for the Electrical Equipment in Flammable Atmospheres (BASEEFA). The United States is not a CENELEC member nation, but the IEC does recognize ANSI standards. Figure 1400-1 shows the current intrinsic safety standards and the corresponding appropriate testing or certifying authorities.

1442 Types of Equipment Certification There are two methods that manufacturers use to evaluate their equipment in the United States: (1) The system approval method, which involves testing a specific barrier for use in combination with specific field apparatus, and (2) the entity concept approval, a more recent approach.

System Approval Method Figure 1400-2 is an example of a system approval certificate. It shows that a Rosemount transmitter, Model 1151, has been given approval for use with a Westinghouse, 75SB02 model barrier in a Class I, Division 2, Group C classified area. The system approval process is involved and expensive because it requires testing unique equipment combinations. The certifying agency requires extensive data and

Chevron Corporation

1400-5

July 1999



Fig. 1400-1 Current Intrinsic Safety Standards (Courtesy of R.Stahl, Inc.) Area

Source

Standard or Recommended Practice

International

IEC

79-11

USA

FM

Class #3610

FM

UL

UL 913

UL

ANSI/ISA

RP 12.6 (Installation)

CSA

CSA C22.2

CANADA

Testing Institution

CSA

No. 157 EUROPE

CENELEC

EN 50014 and

PTB (D)

EN 50020

BASEEFA (GB) CESI (I) INEX (B) LCIE (F)

documentation to be submitted by the equipment manufacturer in order to apply for certification.

Entity Concept Approval The entity concept allows the interconnection of approved intrinsically safe equipment to approved barriers and other associated apparatus that have not been specifically tested together under system approval. This interconnection is allowed providing certain conditions are met relative to voltage, current, capacitance, and inductance. Note that Figure 1400-2 shows the parameters and values required for entity approval. The exact requirements and calculations can be found in FM 3610 and UL 913. These are discussed below.

1443 Definition of Simple Apparatus Intrinsically safe design limits energy transmitted to a hazardous area. The storage of energy inside a classified area must also be considered. This means accounting for the presence of inductive or capacitive elements. Apparatus that do not store energy, that is, that have no capacitance or inductance, or do not contain batteries, are classed as simple apparatus. Apparatus which do not meet these requirements are called non-simple. Thermocouples, resistance temperature devices (RTDs), and switch contacts are examples of simple apparatus. The circuits they are part of are classed as simple circuits. Only capacitance and inductance of the cable need to be considered. An electronic transmitter, on the other hand, is an energy-storing device and thus is a non-simple apparatus.

July 1999

1400-6

Chevron Corporation



Fig. 1400-2 Example of Intrinsic Safety Certifications for Rosemount Model 1151 Electronic Transmitter (Courtesy of Rosemount, Inc.) (1 of 2) FM Intrinsic Safety Certifications for Rosemount Transmitter Model 1151 Intrinsic Safety, FM • Factory Mutual (FM) Intrinsic Safety approval available when used with approved barriers shown below. •

Stainless steel certification tag provided.

•

Not available with Output Codes B and G, or Option Codes V2 and V3.

Systems Approvals FM Approved for Class I,II,III, Division 1&2 Groups Barrier Manufacturer

Barrier Model

A

B

C

D

E&G

Foxboro

2AI-I2V-FGB

•

•

•

•

•

2AI-I3V-FGB

•

•

•

•

•

2AS-I3I-FGB

•

•

•

•

•

3A2-I2D-CS-E/FGB-A

•

•

•

•

•

3A2-I3D-CS-E/FGB-A

•

•

•

•

•

1130FF21000

NA

NA

•

•

•

1130FF22000

NA

NA

•

•

•

1135FF21000

NA

NA

•

•

•

1135FF22000

NA

NA

•

•

•

5850FL81100

•

•

•

•

•

5851FL81100

•

•

•

•

•

5850FL81200

NA

NA

•

•

•

5851FL81200

NA

NA

•

•

•

Taylor

Westinghouse

75SB02

•

•

•

•

•

Leeds & Northrup

316569, 316747

•

•

•

•

•

Honeywell

38545-0000-0110-113-F5B5 38545-0000-0110-111/112-F5B5

•

•

•

•

•

NA

NA

•

•

•

Measurement

115,122,128+,128-,129

Technology

188,188R,322,2441

•

•

•

•

•

Stahl

8901/30-280/070/70

•

•

•

•

•

8901/30-199/130/70

•

•

•

•

•

8901/30-199/100/70

•

•

•

•

•

8901/31-280/100/70

•

•

•

•

•

8901/31-280/070/70

•

•

•

•

•

8901/31-199/130/70

•

•

•

•

•

8901/31-199/100/70

•

•

•

•

•

8901/30-280/165/80

NA

NA

•

•

•

8901/31-280/165/80

NA

NA

•

•

•

•

•

•

•

•

•

•

•

•

•

NA

NA

•

•

•

8903/51-200/050/7 Supply} 8901/31-086/150/7 Return} 8903/31-315/050/7 Supply} 8901/31-086/150/7 Return} Fisher Controls

Chevron Corporation

AC302

1400-7

July 1999



Fig. 1400-2 Example of Intrinsic Safety Certifications for Rosemount Model 1151 Electronic Transmitter (Courtesy of Rosemount, Inc.) (2 of 2) FM Intrinsic Safety Certifications for Rosemount Transmitter Model 1151 Intrinsic Safety, FM • Factory Mutual (FM) Intrinsic Safety approval available when used with approved barriers shown below. •

Stainless steel certification tag provided.

•

Not available with Output Codes B and G, or Option Codes V2 and V3.

Entity Approvals FM Approved for Class I,II,III, Division 1&2 Groups 1151 Parameters

Associated Equipment Parameters

VMAX = 40V

Voc ≤ 40V

IMAX = 165mA

Isc ≤ 165mA

C1 = 0

CA > 0

L1 = 0

LA > 0

VMAX = 40V

Voc ≤ 40V

IMAX = 225mA

Isc ≤ 165mA

C1 = 0

CA > 0

L1 = 0

LA > 0

A

B

C

D

•

•

•

•

NA

NA

•

•

1444 Certification Requirements If the field device is labelled as simple apparatus, system certification is not required. These devices can be regarded as part of an intrinsically safe system if: 1.

They are properly connected to a barrier with intrinsic safety certification.

2.

Cable capacitance and inductance do not exceed the permissible capacitance and inductance connected to the barrier. These cable parameters must be checked, as cabling provides a source of energy storage.

If the field apparatus is non-simple, either system approval or entity approval of the field apparatus and barrier is required. With system approval, design calculations by the user are not necessary. Application of entity-approved equipment requires simple calculations by the design engineer to verify the compatibility of electrical characteristics. Barrier parameters identified by the testing authorities as part of their entity approval are: maximum open-circuit voltage, maximum short-circuit current, and maximum capacitance and inductance permitted to be connected to the barrier. The same parameters are also identified by the testing agency as part of their entity approval for non-simple field apparatus.

July 1999

1400-8

Chevron Corporation



To ensure intrinsic safety when the barrier and non-simple field apparatus have entity approvals, the following four conditions must all be met. (To apply conditions 1 through 4, refer to Examples 1 and 2 in Section 1458). 1.

The barrier open-circuit voltage must be less than or equal to the maximum voltage the field apparatus can receive: Voc ≤ Vmax (Eq. 1400-1)

2.

The barrier short-circuit current must be less than or equal to the maximum current that the field apparatus can be subjected to: Isc ≤ Imax (Eq. 1400-2)

3.

The capacitance allowed to be connected to the barrier must be greater than or equal to the sum of the maximum unprotected capacitance of the field apparatus plus the cable capacitance: Ca ≥ Ci + Cc (Eq. 1400-3)

4.

The inductance allowed to be connected to the barrier must be greater than or equal to the sum of the maximum unprotected inductance of the field apparatus plus the cable inductance: La ≥ Li + L c (Eq. 1400-4)

An accepted alternate condition to Equation 1400-4 is that the inductance-to-resistance ratio (L/R) of the cable be less than the L/R ratio of the barrier: L/R cable < L/R barrier (Eq. 1400-5)

1450 Fundamental Requirements of Intrinsic Safety This section discusses the fundamental requirements of intrinsic safety and the following: design and installation of cables, terminals, wires, and enclosures (Section 1451); protective components (Section 1452); grounding and shielding (Section 1453); maintenance and inspection and commissioning (Sections 1454 and 1455); documentation (Section 1456); and applications (Section 1457). Two examples of sample calculations for cable and barriers are also included (Section 1458). All intrinsically safe apparatus and circuits are tested and certified by approval agencies to meet two basic requirements: 1.

Chevron Corporation

The energy available at the hazardous location must not be great enough to cause ignition by arcing or high temperature during normal operation. Normal

1400-9

July 1999



operation includes use at the maximum supply voltage and with any adjustments made at the most unfavorable settings. The field wiring may be opened, shorted, or grounded. In addition, a safety factor of 1.5 is applied to energy. 2.

The energy available under fault conditions must not be great enough to cause ignition by arcing or high temperature after a single fault, with a multiplier of 1.5 applied to arc energy; and after two faults with no additional safety factor applied to the energy released. A fault is any defect or electrical breakdown which can adversely affect the electrical or thermal characteristics of the intrinsically safe circuit.

The safe release of stored energy for resistive, capacitive, and inductive circuits has been determined experimentally. From these data, curves have been published showing the relationship between voltage and current at ignition levels. These ignition curves are given in UL 913. Equipment manufacturers have tested their equipment in these specific atmospheres during the certification of their equipment. The protective device ensures that the two basic requirements above are satisfied even if specific faults occur. Wiring practices minimize the occurrence of wiring faults. If a fault does occur, the circuit is protected. Assume Figure 1400-3(a) represents a 4 to 20 milliamp signal operating at 24 VDC. If a high AC voltage enters the circuit as shown in Figure 1400-3(b) and the circuit is not designed to limit such a fault, that voltage will be present in the hazardous area. A protective interface must be included in the circuit to prevent such conditions. Various protective devices are discussed in Section 1452. The most common of these is the Zener safety barrier. Its location in the circuit is shown in Figure 1400-3(c).

1451 Design and Installation Considerations This section discusses the wiring design and installation considerations recommended by FM 3610, UL 913, NEC Article 500, and ANSI/ISA RP 12.6.

Cables and Cable Routing Interconnecting cables in hazardous areas for non-intrinsically safe circuits should be installed according to NEC for the Class and Division involved, as well as for power-limited circuits. Similar signal types (such as 4 to 20 mA) should be grouped together and separated from dissimilar types, such as thermocouple, RTD, and alarm contact signals. Install intrinsically safe wiring in conduit or raceway separate from non-intrinsically safe wiring. See ANSI/ISA RP 12.6 for certain exceptions. Intrinsically safe circuits must be visually identifiable as specified in ANSI/ISA RP12.6. Cables must be electrically compatible with the intrinsically safe components being used. The minimum specifications for cables allowed in intrinsically safe circuits are given in UL 913. Figure 1400-4 shows the tabulation for two cables commonly used in many Company facilities.

July 1999

1400-10

Chevron Corporation



Fig. 1400-3 Nonhazardous (Safe) and Hazardous Areas With and Without Circuit Fault Protection

Chevron Corporation

1400-11

July 1999



Fig. 1400-4 Table Showing Typical Parameters of Instrument Cables Typical Single-Pair Signal Cable 16 Gage

Typical Multi-Pair Signal Cable 20 Gage

UL-913 Cable Specification for Intrinsically Safe Circuits Minimum Strand Diameter 0.008 inch

0.019 inch

0.012 inch

Minimum Insulation Thickness 0.010 inch

0.015 inch

0.015 inch

Minimum Insulation Rating 500 V

1000 V

1000 V

Maximum Capacitance 60 pF/foot (picofarads/foot)

58 pF/foot

31 pF/foot

Maximum Inductance 0.020 mH/foot (microhenries/foot)

0.016 mH/foot

0.18 mH/foot

Sample calculations for cable capacitance and inductance are presented in Section 1458.

Terminals, Wiring, and Enclosures The creepage and clearance distances between uninsulated live parts as listed in UL 913 must be observed. Terminals for intrinsically safe wiring must be separated from terminals for nonintrinsically safe circuits. The preferred method is to locate each circuit type in a separate enclosure. A common enclosure is allowed if the intrinsically safe circuits are separated from the non-intrinsically safe circuits by either of two methods. Separation can be accomplished with an insulating or grounded metal partition or by a minimum distance of 2 inches between the adjacent terminals. The following practice, not specifically required by the standards, is considered good procedure and is recommended. •

Strip only enough insulation from conductors to provide a secure termination. Wiring should be secured to maintain separation and no more than 6 inches of slack should be permitted at the terminals. These precautions and attention to terminal layout will minimize the chance of a ground or short between circuit types.

•

Test points should be located to minimize the possibility of degrading the intrinsic safety of circuits. Acceptable types of test equipment should be identified by permanently attached tags located near the test points.

1452 Protective Components Intrinsically safe circuits depend on the use of components that limit current and voltage. Some components are internal to intrinsically safe devices that are sized

July 1999

1400-12

Chevron Corporation



and specified by the manufacturer as part of the certification process. Examples of these include: • • • • •

Transformers Damping windings Shunt diodes Resistors Capacitors

More detailed information on these items is provided in UL 913 and FM 3610.

Zener Barriers Zener barriers are the most common type of component in intrinsically safe instrumentation and are sized by the design engineer. Zener barriers are capable of limiting voltage and current to a hazardous area by means of the protected circuit. A circuit schematic of a typical barrier is shown in Figure 1400-5. The resistors must meet the requirements described in UL 913. The zener diodes must be able to withstand 1.5 times the maximum available voltage and current without failure. In order for smaller diodes to be used, most barriers incorporate non-replaceable fuses to limit the maximum fault current through the zener diodes. These fuses must be capable of interrupting large fault currents in one-tenth of the diode’s rated maximum conduction time for the current. This current, typically, is many times larger than anticipated fault currents in the instrument loop. Fig. 1400-5 Safety Barrier Circuit Diagram (Courtesy of R.Stahl, Inc.)

Apparatus connected to the hazardous side of the barrier must be approved for use with the barrier by either system or entity procedures. In three-wire, bridge circuit, or specially grounded loops, more than one barrier may be required to isolate all legs of the loop.

Active Isolators The devices described above are called passive isolators. Active isolators are also available. They too are sized by the design engineer. Active isolators are usually in

Chevron Corporation

1400-13

July 1999



the form of solid state isolating relays and thermocouple or strain-gage amplifiers. Active isolators should be tested and approved for their intended application.

1453 Grounding and Shielding Intrinsically safe systems require reliable and secure grounds, and design of the grounding network therefore requires care and attention. Grounding requirements are found in ANSI/ISA RP 12.6. The major requirements are: •

Barriers and cable shields must be grounded together

•

Dedicated, redundant intrinsic safety ground cables that terminate at the instrument system grounding electrode are recommended

•

The maximum allowable resistance from the ground cables to the instrument system grounding electrode is 1 ohm

1454 Maintenance of Intrinsically Safe Systems Intrinsically safe systems are not capable of releasing enough electrical or thermal energy to ignite a specific flammable mixture despite open circuits, shorts, grounds, and two faults in the system. Therefore, the system will almost always fail to perform its function long before a safety hazard develops. However, environmental conditions, such as corrosion, moisture, shock and vibration, may slowly degrade the integrity of a system. Periodic inspections are needed, especially to confirm the integrity of the grounding system. A more likely way for an intrinsically safe installation to be violated is by improper modifications. If a new device is wired into a loop, it is possible to do so without having verified whether the limiting inductance or capacitance is exceeded. Devices must be approved for use in existing loops. Another way to violate an intrinsically safe installation is to add a device to panel wiring that violates the separation requirements between intrinsically safe and non-intrinsically safe circuits. Periodic inspections are recommended to ensure that intrinsically safe systems have not been violated. All changes to intrinsically safe wiring should be documented, reviewed, and approved by a qualified engineer.

1455 Inspection and Commissioning All intrinsically safe instrumentation systems must be tested prior to startup, just as with other types of instrument installations. The engineer is usually responsible for identifying and documenting those components and circuits which must be tested during commissioning and inspected periodically thereafter. The tests and inspections should be witnessed or performed by a Company representative qualified to work with intrinsically safe systems. The operating management of the installation is responsible for ensuring periodic inspections.

July 1999

1400-14

Chevron Corporation



Items to be Inspected The items listed below require inspection or testing during commissioning and periodically thereafter. This list should be carefully reviewed by the design engineer. Intrinsically safe instrumentation systems should be tested or inspected to verify that: 1.

All components are labelled with instrument and certification tags.

2.

All junction boxes, cable trays, conduits, cables, cabinets, and instrument housings are labelled as containing intrinsically safe circuits or equipment.

3.

Terminations are neat and made in a workmanlike manner with proper partitions, cable ties, and insulation.

4.

All conductors are of the proper type, insulation, gage, and color, and are properly routed.

5.

All barriers are installed and operational.

6.

All power supplies are of the proper rating and usage.

7.

All grounds are properly installed and tested.

8.

All fuses are of the proper type and rating.

9.

No jumpers or modifications violate the integrity of the system’s intrinsic safety.

10. All documentation is accurate, complete, and available at the site. See Section 1456. 11. All inspections and testing are completed without fault or omission prior to energizing the intrinsically safe circuits. 12. A procedure and a schedule for periodic maintenance and inspection are provided to operations.

1456 Documentation All documentation for an intrinsically safe instrumentation system should be complete, assembled, and available on site for testing and commissioning (as with other types of instrument installations). All drawings, manuals, and other documentation for intrinsically safe apparatus must indicate that the components, circuits, or apparatus are part of an intrinsically safe system. Documentation for intrinsically safe apparatus should be approved as part of the design, and should include:

Chevron Corporation

1.

Equipment manuals.

2.

System or entity approval documents from equipment manufacturers including adherence to Equations 1400-1 through 1400-4.

1400-15

July 1999



3.

Replacement parts list.

4.

Grounding interconnection and wiring diagrams.

5.

Instrument wiring diagrams.

6.

Power supply interconnection and wiring diagrams.

7.

Design specifications.

8.

Procedure and schedule for periodic maintenance and inspections.

9.

Instructions for future modifications and additions.

1457 Typical Applications This section contains recommended design techniques for typical applications, including thermocouples, alarms, electrical equipment, interlocks and shutdowns, milliamp loops, fiber optics, and intelligent transmitters.

Thermocouples The voltages produced by thermocouples are usually suitable for intrinsic safety design. The main objective is to prevent dangerous voltages from being connected to thermocouple wiring from electronics in the unclassified areas through faults or maintenance errors. Barriers or active isolators may be used for individually grounded or ungrounded thermocouples. Individually field-mounted thermocouple transmitters may be used where individual thermocouples need to be grounded; these could be certified intrinsically safe or mounted in explosionproof housings with field barriers. For large thermocouple scanning systems, the alternatives are intrinsically safe multi-plexers, barriers with floating (ungrounded) thermocouples, or explosionproof construction.

Alarms Hermetically sealed alarm contacts are recommended for reliability in all alarm systems, whether or not they are intrinsically safe. They are not acceptable in Division 1 areas without explosionproof enclosures unless the system is intrinsically safe. Intrinsically safe alarm systems are readily available and are recommended. The major design effort is to assure the integrity of the intrinsically safe alarm wiring. Isolation from non-intrinsically safe wiring and circuits is important. This may be achieved using safety barriers, approved solid state relays, and approved wiring techniques. Alarm contacts in switchgear and motor starters usually are not approved for connection directly to intrinsically safe systems since a hazardous foreign voltage could be placed on the hazardous or field side of the barrier. Approved interposing relays installed between the field contacts and barriers provide the required

July 1999

1400-16

Chevron Corporation



protection. In addition, the signal wiring for these contacts should be installed in cables/conduits separate from other intrinsically safe circuits.

Electrical Equipment Associated electrical equipment (motor starters, circuit breakers, transformers, and power system instrumentation) are either located in an unclassified area or are mounted in explosionproof housing. The greatest risk with associated apparatus is the possibility of imposing unsafe voltages or currents on intrinsically safe wiring through faults or incorrect wiring. Electrical equipment associated with intrinsically safe instrumentation should be purchased with the necessary terminal spacings, ground isolation, and grounded wireway dividers or separate wire routing. It is best to have all associated circuits brought out to a separate cubicle, junction box, or terminal strip, as appropriate. Intrinsically safe alarm circuits may be wired to contacts on relays containing nonintrinsically safe circuits, provided that the non-intrinsically safe circuits comply with standard creepage and clearance distances and do not exceed the current and voltage limitations specified in UL 913. Certification of acceptable usage should be obtained for each relay. Approved barriers, solid state relays, and current repeaters may be used to provide isolation. These are preferred over electro-mechanical devices and should be conspicuously located in a separate cubicle or junction box. The design engineer should review and approve pertinent drawings and shopinspect intrinsically safe wiring in electrical apparatus.

Interlocks and Shutdowns Sensing switches, electrical equipment interlocks, solenoid valves, and sequencing relays connected to typical shutdown and interlock circuits may be located in hazardous areas. Some sensing devices are approved as intrinsically safe or are nonincendive. Most other components in this application are typically of generalpurpose construction and are located in nonhazardous areas or are mounted in explosionproof enclosures. Approved barriers and solid state relays may be used to ensure intrinsic safety. Relays shared with non-intrinsically safe circuits are permitted as stated above. Sensors with multiple hermetically sealed or independently switched contacts may also be used to separate intrinsically safe circuits from non-intrinsically safe circuits. To minimize the number of interconnections between intrinsically safe and nonintrinsically safe portions of a shutdown or interlock system, the designer should consider partitioning the system logic into intrinsically safe and non-intrinsically safe parts. Apparatus located in hazardous areas should be wired in an intrinsically safe manner to compatible logic circuits. A minimum number of contacts should then be connected from the non-intrinsically safe part of the system through appropriate barriers or isolators. Special attention should be given to the design of interconnection diagrams for interlock and shutdown systems because of the large amount of non-intrinsically safe apparatus that could be involved.

Chevron Corporation

1400-17

July 1999



Milliamp Loops Intrinsically safe milliamp control loops should be designed for a 4 to 20 milliamp range. Interconnections between multiple power supplies should be analyzed to eliminate the possibility of unsafe overvoltages, overcurrents, and overpowered circuits. Approved barriers are available for these circuits. Proper grounding is critical, and includes proper termination of cable shields.

Field-powered Milliamp Loops Analyzers and other self-contained and self-powered instrument systems are often associated with centrally located controls. These systems are most often located in nonhazardous areas or are mounted in explosionproof enclosures. Approved barriers and isolating current repeaters can be used to connect these self-powered milliamp loops to intrinsically safe controls. These protective devices must be located at the field device. Multiple barriers in series may result in such loops limiting the maximum allowable cable resistance. The field device must be referenced to the intrinsic safety ground. The field power supply should be included in the multiple power supply analysis recommended above.

Fiber Optics Not all fiber optic equipment is intrinsically safe. Some fiber-powered transmitters and high-speed communications equipment transmit and receive enough energy to ignite hazardous mixtures. All fiber optic equipment in hazardous areas must be certified as intrinsically safe. The current certification stage of instruments considered for purchase should be checked carefully. Not all manufacturers’ products have yet been certified as intrinsically safe.

Intelligent Transmitters Some makes of intelligent transmitters cannot be addressed through certain makes and models of zener barriers. Check with the vendor before purchasing equipment to verify that transmitters and barriers are compatible.

1458 Sample Calculations for Cable and Barriers Sample calculations for cable and barriers presented in examples 1 and 2 below are based on Equations 1400-1 through 1400-4, as follows: Voc ≤ Vmax (Eq. 1400-1)

Isc ≤ Imax (Eq. 1400-2)

Ca ≥ Ci + Cc (Eq. 1400-3)

July 1999

1400-18

Chevron Corporation



La ≥Li + Lc (Eq. 1400-4)

Example 1—Entity Concept Calculations for a Non-simple Device The basic steps are: 1.

Choose the non-simple device and barrier. (The requirements of Equations 1400-1 and 1400-2 must be satisfied.)

2.

Choose the cable, determine the required length, and calculate Cc, Lc.

3.

The requirements of Equations 1400-3 and 1400-4 must be satisfied. The L/R ratios may be checked as an alternate to Equation 1400-4.

4.

Calculate the loop resistance. It must be less than or equal to the maximum allowable resistance of the barrier.

5.

Ensure that the rated current of the barrier fuse is equal to or greater than the maximum operating current of the loop.

Note

These steps may require several iterations to meet all conditions.

Figure 1400-6 shows a Stahl barrier, Model Number 9001/51-280-091-14 connected to a Rosemount transmitter (a non-simple device) Model Number 3051C, FM approved, with transient protection (option code T1) and 4-20 mA output with digital signal based on HART protocol (output code A) in a Class I, Division 1, Group B classified area. The cable is a single twisted and shielded pair (16 AWG). Vendor data are: Barrier • • • • • • •

Voc = 28 V Isc = 88.2 mA Ca = 140,000 pF La = 4500 µH Maximum Resistance = 317 ohms Fuse Rating = 160 mA L/R = 14.2 µH/ohm

Transmitter • • • •

Vmax = 40 V Imax = 160 mA Ci = 10,000 pF Li = 1050 µH

Cable • • •

Chevron Corporation

Capacitance = 58 pF/foot Inductance = 0.16 µH/foot Resistance = 0.009 ohms/foot per conductor

1400-19

July 1999



Fig. 1400-6 Non-Simple Device: Example 1: Sketch (Courtesy of R. Stahl, Inc.)

•

Length = 1,500 feet

By inspection, the conditions of Equations 1400-1 and 1400-2 are met. Cc and Lc can be obtained as follows: 1500 ft x 58 pF/ft = 87,000 pF Lc = 1500 ft x 0.16 µH/ft = 240 µH Equation 1400-3 requires that: Ca ≥ Ci + C c 140,000 ≥ 10,000 + 87,000 = 97,000 pF Equation 1400-4 requires that: La ≥ L i + Lc 4500 ≥ 1050 + 240 = 1290 µH Equations 1400-3 and 1400-4 are satisfied. A check of the L/R ratios shows that: 0.16 µH/ft L/R cable = ------------------------------- = 8.9µH/ohm 0.018 ohm/ft L/R barrier = 14.2µH/ohm L/R cable < L/R barrier (Eq. 1400-6)

July 1999

1400-20

Chevron Corporation



Next, check the end-to-end resistance of the barrier with respect to the loop. The loop has a 24 volt power supply and a 4 to 20 mA signal (as shown in Figure 1400-7). Resistance is 24 volts / 0.02 amps = 1200 ohms at 20 mA. Fig. 1400-7 Basic Loop Diagram with Intrinsic Safety Barrier (Courtesy of R. Stahl, Inc.)

The minimum voltage at which the transmitter will operate is 12 volts. At 20 mA, the interior resistance of the transmitter is 600 ohms. The input impedance of the receiving equipment is 250 ohms. The total resistance so far is 850 ohms. Last, determine the signal wire resistance. Since the twisted pair has a combined resistance of 0.018 ohm/foot, the cable will have a total resistance of 1500 × 0.018 ohm/foot = 27 ohms for the loop. Thus, the total of these resistances is 877 ohms, leaving 323 ohms for the barrier. Thus, the end-to-end resistance of the specified barrier cannot exceed 323 ohms. A check of vendor data shows that the maximum allowable barrier is 317 ohms. Finally, ensure that the rated current of the barrier fuse (in this case 160 mA) is equal to, or greater than, the maximum operating current of the loop (20 mA).

Example 2—Entity Concept Calculations for a Simple Device The basic steps are: 1.

Choose a barrier. Determine the barrier cable parameters from vendor data.

2.

Choose the cable. Determine the allowable length from cable parameters.

3.

Calculate the loop resistance. It must be less than, or equal to, the maximum allowable resistance of the barrier.

Note that these steps may require several iterations to meet all conditions. In the case of a non-energy-storing (simple) device connected to a barrier, all of the capacitance and inductance permitted for the device can be allocated to the connecting cabling.

Chevron Corporation

1400-21

July 1999



An MTL Model 187 barrier is chosen for the switch contact in the field as shown in Figure 1400-8. From published vendor data: • • •

Maximum cable capacitance allowed = 130,000 pF Maximum cable inductance allowed = 80 µH Maximum resistance allowed = 340 ohms

Fig. 1400-8 Example 2: Model 187 Simple Device Sketch (Courtesy of MTL)

If a 16 AWG twisted pair cable has parameters of 58 pF/ft capacitance and 0.16 µH/ft inductance, then the allowable length cannot exceed the lesser of: 130, 000 80 --------------------- = 2, 241 feet or ---------- = 500feet 58 0.16 (Eq. 1400-7)

If the cable parameters are not known, UL 913 allows the following maximum values to be used: • •

Capacitance = 60 pF/ft Inductance = 0.2 µH/ft

A check of the total cable resistance shows that: 500 ft x 0.018 ohms/ft = 9 ohms < 340 ohms

1460 References Chevron Electrical Manual, Section 300, “Area Classification”

July 1999

1400-22

Chevron Corporation



Specification ICM-MS-3651 “Installation Requirements for Digital Instrumentation and Process Computers” API Recommended Practice RP 500 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Division 1 and Division 2 API RP 505 Classification of Locations for Electrical Installations at Petroleum Facilities Classified as Class I, Zone 0, Zone 1, and Zone 2 API Recommended Practice RP 14F, Design and Installation of Electrical Systems for Offshore Production Platforms FM 3610 Intrinsically Safe Apparatus and Associated Apparatus for Use in Class I, II, and III, Division 1 Hazardous Locations, October 1988 ISA dS12.1 (1988) Definitions and Information Pertaining to Electrical Instruments in Hazardous (Classified) Locations (Draft) ANSI/ISA RP 12.6 (1987) Installation of Intrinsically Safe Systems for Hazardous (Classified) Locations NFPA 70 National Electrical Code, 1987 Underwriter’s Laboratories (UL) 913 Intrinsically Safe Apparatus and Associated Apparatus for Use in Class I, II, and III, Division 1, Hazardous (Classified) Locations, Fourth Edition July 29, 1988

Chevron Corporation

1400-23

July 1999

1500 Instrument Seals, Purges, and Winterizing Abstract Field instruments must be protected against plugging and damage caused by low ambient temperatures and process fluids that cause fouling and chemical attack. This section covers seals, purges. and winterizing for this protection. Section 1510 shows the need for protection and describes protective techniques. Section 1520 discusses process and operating conditions that lead to problems and recommended solutions. Sections 1530, 1540, and 1550 discuss seals, purges, and winterizing in more detail. Section 1560 lists relevant model specifications, standard drawings, and engineering forms, and Section 1570 lists references. Heat tracing for process piping is covered in the Utility Manual, Section 700. Contents

Page

1510 Overview of Instrument Seals, Purges, and Winterizing

1500-3

1511 General Information 1512 Quick Guide to Design Decisions 1513 Form ICM-EF-409, Application Spreadsheet 1520 Process Fluid Considerations

1500-14

1521 Water 1522 Hydrates 1523 High Pour Point Hydrocarbons 1524 High Viscosity or Low Pour Point Hydrocarbons 1525 Fouling Fluids and Coking Hydrocarbons 1526 Salts and Other Solids 1527 Condensation 1528 Chemical Attack and Corrosion 1529 High Temperature Fluids 1530 Seals

Chevron Corporation

1500-27

1500-1

July 1999

1500 Instrument Seals, Purges, and Winterizing


1531 General Information 1532 Diaphragm Seals 1533 Liquid Seals 1534 Need for Additional Winterizing 1540 Purges

1500-35

1541 General Information 1542 Disadvantages of Purges 1543 Purging Methods 1544 Purging Fluids 1545 Need for Additional Winterizing 1550 Winterizing—Heating and Insulating

1500-49

1551 General Information 1552 Ambient Temperature Considerations 1553 Fluid Considerations 1554 General Design Principles 1555 Instrument Housings 1556 Heat Tracing 1557 Steam Tracing 1558 Electric Tracing 1559 Traced Tubing Bundles 1560 Model Specifications, Standard Drawings, and Engineering Forms 1500-70 1561 Standard Drawings 1562 Engineering Forms 1570 References

July 1999

1500-71

1500-2

Chevron Corporation



1510 Overview of Instrument Seals, Purges, and Winterizing 1511 General Information This section of the Instrumentation and Control Manual discusses the use of seals, purges, and winterizing to protect instruments and their process-connecting leads from the negative effects of process fluids and low ambient temperatures. It covers most of the conditions typically encountered in Company petroleum and petrochemical facilities. It does not cover the rarer situations found in other chemical plants and in truly severe climates, such as the north slope in Alaska or the high plateau country in Wyoming. This section is intended to be supplementary to, and should always be used in conjunction with, Company form ICM-EF-409 and API Recommended Practice RP 551, Process Measurement Instrumentation. API RP 551 is included in Volume 2 of this manual. ICM-EF-409 is included in Volume 1. Both API RP 551 and ICM-EF-409 are briefly discussed in this section.

Problems Some process fluids under the effects of ambient temperature can disable field instruments. They can plug leads, condense in leads, distort internal parts, and cause corrosion and other forms of chemical attack. Plugging is usually the result of ambient temperature conditions that cause oils to congeal or become too viscous, water to freeze, and hydrates to form. Solids deposited by process fluids also cause plugging by coke formation, sedimentation of suspended solids, and salt formation. Freezing water is usually responsible for distortion damage. Finally, some of the high temperature or chemically aggressive process fluids will cause rapid corrosion, cracking (stress corrosion cracking), or thermal breakdown of internal fill fluids or will otherwise damage sensitive exposed internal parts of instruments. Typically these parts are thin and delicate. They cannot be designed with any corrosion allowance. The best solution is to use instruments that are immune to these effects, ones that are inherently safe and do not require the installation or maintenance of any special protective techniques. For instance, use construction materials that are immune to high temperatures or chemical attack, and use in-line measurement methods (such as turbine meters or vortex shedding meters) that do not need external process connection leads. Typically, however, such immune equipment is not available, not acceptable for the application, or costs too much. Therefore, some other form of protection is required. Seals, purges, and winterizing can provide this protection.

Solutions Sealing, purging, and winterizing are usually used to protect instruments that are installed with some form of either flowing or nonflowing (pressure transmitting)

Chevron Corporation

1500-3

July 1999



process leads. Typically these are instruments used to measure flow, liquid level, and pressure, but process analyzer leads may also be included. Seals, particularly diaphragm seals, are the most reliable and require the least maintenance of the three protective techniques. They are the preferred method and should be considered first when planning a protected installation. Winterizing should be considered next. It is more costly, requires more maintenance, and is less reliable since heat tracing is an active method which can fail and cause loss of protection. Nevertheless, it is better than purging. Flowing purges are the least reliable. They are high cost and high maintenance systems. They are difficult to design, install, and maintain so that they will work consistently, reliably, and cause a minimum of problems. Consider them only as a last resort where seals or winterizing will not work. Seals. Seals are used only on nonflowing leads. They are passive techniques which exclude the process fluid from the instrument and connecting leads. Two different methods are available. Figure 1500-1 illustrates both. One is the diaphragm seal which uses a flexible diaphragm (usually made of a nil corrosion metal) with process fluid on one side and an appropriate fill fluid on the instrument side. The other method uses a seal pot with a liquid seal on the instrument side that is immiscible with the process fluid. Seals are discussed in detail in Section 1530. Purges. Purges can be used on both flowing and nonflowing leads to exclude the process fluid from the instrument. They do this by injecting a continuous flow of fluid into the process lead. This continuously sweeps the process fluid out of the leads, thus protecting the instrument and preventing the buildup of process fluid that would plug the leads. This idea is illustrated in Figure 1500-2. Purges are discussed in detail in Section 1540. Winterizing. Winterizing is generally some form of heating, usually heat tracing. This keeps the process fluid warm enough to prevent impaired operation from the effects of low ambient temperatures, such as congealing heavy oils, freezing water, or condensation. See Figure 1500-3 for a typical heat traced instrument. Winterizing is discussed in detail in Section 1550. Interrelationships. Sealing, purging, and winterizing are interrelated techniques. Sometimes one can be substituted for the other or two may be used together. Frequently installations will use winterizing in conjunction with seals or purges. A common solution is to heat trace the process side of a sealed installation. Sometimes winterizing (heat tracing or just insulation) must be used to protect stagnant seal fluid or even flowing purge fluid from low ambient temperatures. In some instances it may even be necessary to heat trace diaphragm seal fluid (such as in capillary type external seals) to maintain low enough viscosity to transmit pressure in low ambient temperature conditions. A common combination is shown in Figure 1500-4.

July 1999

1500-4

Chevron Corporation



Fig. 1500-1 Typical Seals

Chevron Corporation

1500-5

July 1999



Fig. 1500-2 Typical Purge LOCATION OPTIONAL -

DRAIN BACK PREVENTION

NEAR INSTRUMENT

OFFSET FOR PURGE OIL LIGHTER THAN PROCESS

OR NEAR PROCESS CONNECTION

FLUID

PROCESS

TO INSTRUMENT

CONNECTION

BY-PASS FOR HIGH VOLUME FLUSH

CHECK VALVE

PURGE PURGE ROTAMETER

SUPPLY

NEEDLE VALVE

1512 Quick Guide to Design Decisions Figure 1500-5 (sheets 1–5) is a logical flowchart which can be used to quickly determine if protection is needed and what protection is appropriate. It covers most of the conditions typically encountered in Company petroleum and petrochemical facilities. It does not cover the rarer situations found in other chemical plants and in truly severe climates, such as the north slope in Alaska or the high plateau country in Wyoming This flowchart list is intended to be used in conjunction with the appropriate portions of this section, with Company form ICM-EF-409, Seals, Purges, and Winterizing of Instruments, and with API RP 551.

How the Quick Guide Works The Quick Guide is an ordered flowchart of statements and conclusions. The question is whether the statement applies to the application under review. The answer is either yes (y) or no (n). These y’s and n’s establish a path through the flowchart that leads to a conclusion. Either protection is required or it is not. If it is, the flowchart indicates what type of protection is in general use in Company facilities and where to go for guidance in applying protection.

July 1999

1500-6

Chevron Corporation



Fig. 1500-3 Typical Heat Tracing

Pressure Transmitter Steam Supply Insul. Det. 1 Heat-Bolt and Heat-Pak Assembly Insul. Det.2

Glass Tape Wrap for Light Tracing Only

To Trap Weather Coat

Lead or Tracer

Lead Pipe Insulation

Tracer

DETAIL - 1

DETAIL - 2

Single Lead or Tracer Insulation and Weatherproofing

Single Lead with Tracer Insulation and Weatherproofing

Chevron Corporation

1500-7

July 1999



Fig. 1500-4 Combined Seal and Heat Tracing (From Standard Drawing GD-J99742-5)

July 1999

1500-8

Chevron Corporation



Fig. 1500-5 Quick Guide for Winterizing Design Decisions (1 of 5)

Chevron Corporation

1500-9

July 1999




July 1999

1500-10

Chevron Corporation




Chevron Corporation

1500-11

July 1999




July 1999

1500-12

Chevron Corporation




Chevron Corporation

1500-13

July 1999



1513 Form ICM-EF-409, Application Spreadsheet Company form ICM-EF-409, titled “Seals, Purging, and Winterizing of Instruments,” is a spreadsheet that presents Company practice in the application of these protective methods to field instrument installations. It takes into account the type of field instrument, the type of process fluid, its maximum expected viscosity, and the effects of ambient temperature.

How ICM-EF-409 Works Refer to form ICM-EF-409, included in Volume 1 of this manual. The numbered rows represent the different field instruments. The lettered columns represent the different process fluids at temperatures within the expected range of ambient temperatures as defined by the corresponding notes. The fields at the intersections of the rows and columns give the recommended protective methods. The extensive notes at the bottom of the form define terms, explain the use of the form, and assist in applying the recommended techniques. Be sure to read and apply all of the applicable notes. They are very important. The ones referred to in the heading for each lettered column (5, 6, 7) are an integral part of the definition of the operating conditions. The ones referred to in the internal fields (10, 12, 13, 16, 19) are integral parts of the definition of the recommended protective measures. Finally, several notes (1, 2, 3, 4, 8, 9, 11, 15, 17, 18) are of a more general nature and are essential to successful installations..

Example The use of ICM-EF-409 can be illustrated by an example. For a pressure transmitter on a wet vapor that condenses water in a freezing climate, the form works as follows: •

Enter row 3 where pressure transmitters appear.

•

Enter column D for wet vapors where freeze protection is required as called for in Note 7.

•

Use the combination of a diaphragm seal-mounted self draining with heat tracing as called for in the field at the intersection of the row and column; take into account the notes referred to in that field.

•

Note 12 is not applicable; capillary lengths are short with the transmitter line mounted.

•

Note 13 is not applicable; this is not a pressure gage.

1520 Process Fluid Considerations Process fluid is the first factor to take into account when determining the need for seals, purges, and winterizing. Ambient temperature is the second.

July 1999

1500-14

Chevron Corporation



Many fluids will disable instruments as the ambient temperature drops by forming solids that plug leads or by increasing viscosity to the point where the fluid will not flow or will not transmit pressure. With other fluids, ambient temperature is not a significant factor in the problems they cause. This section discusses specific process fluids and the problems they cause and indicates which methods to use for protection. Only those fluids most common to Company facilities are discussed. Protective methods are covered in general terms. For more complete guidance on specific protective techniques, see the following: Section 1530 for seals, Section 1540 for purges, and Section 1550 for winterizing. Company Standard Form ICM-EF-409, the application spreadsheet for seals, purges, and winterizing, shows Company practice in the use of these methods for various combinations of process fluids and temperatures.

1521 Water Water Problems Water is the most common problem fluid in petroleum processing facilities. When water freezes it can plug process connecting leads and also the supply and condensate return tubing for steam tracing. It expands when it freezes and can therefore rupture tubing and ruin the internal wetted parts of instruments. Free liquid water streams and steam services are only the most obvious sources of water problems. Many hydrocarbon streams are either water-contaminated or are wet by design. They contain free water almost all of the time. Most other hydrocarbon streams must be considered to have water in them at least some of the time. Water will accumulate at low points in instrument leads and can freeze. The water in these streams can come from many sources: steam stripping, water washing processes, steaming out lines, water flushing lines, process upsets, and water contaminated storage. Gas and vapor streams frequently contain water vapor in concentrations high enough to condense out as free liquid water when the ambient temperature drops. Sources of water are similar to those for liquid hydrocarbon streams.

How To Solve Water Problems The first and most effective step is to use designs that keep the water out of the instrument and its process-connecting leads. This is practical for nonflowing leads, such as those on differential pressure (d/p) flow transmitters and on pressure transmitters, but not for flowing leads such as those on level glasses and on analyzer sample tubing. Obviously, it will not work for liquid water streams. Self-draining Connections. Use line-mount transmitters with close-coupled leads that are fully self-draining from the instrument to the process connection. This practice will solve the problem for most liquid hydrocarbons and for water condensing out of gases and vapors once the water is drained and the leads and instrument body are filled with water-free process fluid. In cases where continuous condensation of

Chevron Corporation

1500-15

July 1999



water in the leads is expected, it will be necessary to add winterizing to prevent freezing. Frequently, insulation alone is sufficient. Diaphragm Seals. Use diaphragm seals. Mount them so that they are close-coupled to the process line and self-draining into the process. This practice keeps the water out of the seal body. If the process line is winterized, the process connection and the process side of the seal should be winterized in the same way as the line using the line tracer, if any. Figure 1500-4 illustrates these preventive measures. This method will work for liquid water applications and for steam up to the temperature limits of the seal and its fill fluid. Liquid Seals. Use liquid seals arranged so that they are self-draining to the process from the surface of the seal liquid. Use a seal liquid that is heavier than, and immiscible in, both the process fluid and water. This is the method commonly used in steam service. When the water is continuously condensing from vapor or gas, insulate that portion of the system where condensation takes place, including the “top half” of the seal pot, if any. Heat from the circulating gas or vapor usually prevents freezing, but heat tracing may be required in unusually severe situations. Winterizing. The above methods cannot be used in many applications, such as for some liquid water streams and the flowing leads on level glasses, level switches, displacement-type level transmitters and controllers, sample leads on analyzers, etc. Use winterizing in these cases. Sometimes insulation alone is sufficient, for example, for flowing leads in mildly freezing climates. When heat tracing is used, be sure to avoid temperatures that will boil the water. Boiling will cause erratic and erroneous instrument readings. For an occasional application involving potential condensation of water from gas or vapor, the best way to prevent condensation is to keep the temperature above the dew point of the water. Heat tracing is used to do this. For nonflowing instruments in liquid water streams and for steam services, use close-coupled and self-draining seals to minimize the winterizing required and to keep water out of the instrument bodies.

1522 Hydrates Hydrate Problems Hydrate problems are closely related to water problems, but are not nearly as common. Hydrates are solids that cause plugging. They do not, however, cause damage from expansion. The hydrate formers most commonly encountered are hydrogen sulfide and propane and lighter hydrocarbons. Liquid water must be present for hydrate formation. The formation temperature is a function of both pressure and the concentrations of the hydrate formers in the

July 1999

1500-16

Chevron Corporation



mixture. This means that hydrates will form well above the freezing point of the liquid water. Figure 1500-6 shows this general relationship. Fig. 1500-6 Hydrate Formation with Light Hydrocarbons; Approximate Relationships—Do Not Use for Design

How To Solve Hydrate Problems Whenever hydrate formation is possible, check process information to see where and when protection will be needed. Because liquid water must be present and its source may be condensation from vapor mixtures, hydrates can be treated much like wet hydrocarbons. Use the same principles of self-draining, sealing, insulating, winterizing, and heat tracing to maintain temperature above the dew point of water. The main difference is that protection against hydrate formation will be needed at higher ambient temperatures and probably at different temperatures for different applications. The temperature at which hydrates form is a function of pressure and composition both of which can vary in any specific process application. This means that the ambient temperature at which protection is needed will not be known with the same precision as for water freeze protection. Be much more conservative in deciding when and where protection is needed.

Chevron Corporation

1500-17

July 1999



1523 High Pour Point Hydrocarbons High Pour Point Hydrocarbon Problems As their temperature decreases, many hydrocarbon liquids reach a point called their “pour point,” at which they no longer flow freely. They will not readily pour out of a test container. At this temperature, the hydrocarbon fluids stop moving in flowing leads and will no longer transmit pressure in nonflowing leads. The instruments become useless. A high pour point hydrocarbon is one with a pour point above the minimum expected ambient temperature, frequently well above it. Pour points may be below 32°F, but pour points in the normal ambient temperature range are common. Those above 100°F are not uncommon, and some of the heavier oils and residuums can have pour points well above 200°F. Therefore, pour point protection is not just a cold weather need.

How To Solve High Pour Point Hydrocarbon Problems The main defense is heat tracing, which keeps the liquid above its pour point. This is the only solution for flowing leads, and is the preferred solution for nonflowing leads. Heat tracing should maintain the fluid a minimum of 20°F above the pour point at the lowest expected ambient temperature and the severest weather conditions of wind and precipitation. Overheating at maximum ambient temperature is possible but unlikely since most high pour point fluids have high boiling points. Purging. Although heat tracing is the only solution for flowing leads, purging with a fluid that is compatible with the process can sometimes be used on nonflowing leads. This keeps the high pour point fluid out of the instrument and its leads and reduces the need for heat tracing. Even with purges, however, the process lead should be heat traced between the process tap and the purge connection. This procedure provides protection against purge failure and any migration of the high pour point material into the lead. This protection is especially needed for applications where ambient temperatures will be at or below the pour point all or most of the time. Because high pour point process streams are frequently quite hot, be careful in selecting purge fluids to avoid any vaporization when the purge enters the traced portion or the process stream. The vaporization can come from light ends or moisture in the purge fluid. Diaphragm Seals. Diaphragm seals can be used to ensure that the high pour point process fluid is excluded from the instrument body and to minimize the amount of heat tracing required. Use of seals is limited by the maximum temperature rating of the seal and its fill fluid. Seal fill fluids that are suitable for high process temperatures are likely to be too viscous at minimum ambient temperatures. In such cases, the capillary leads are winterized to keep the fill fluid warm enough to assure adequate fluidity.

July 1999

1500-18

Chevron Corporation



Minimize Tracing. Both purged and sealed installations should be designed to minimize the amount of heat tracing needed. Put purge connections close to the process root valve and mount the diaphragm housings close-coupled directly on the root valve. When this is done, the main process line tracer can generally be used to trace from the process tap to the purge connection, or to and around the diaphragm seal body. This eliminates the need for separate instrument tracers, yet allows the equipment to be removed for servicing or replacement.

1524 High Viscosity or Low Pour Point Hydrocarbons High Viscosity or Low Pour Point Hydrocarbon Problems High viscosity low pour point hydrocarbons generally do not have specific pour points. With decreasing temperature, their viscosity increases to the point where instrument functions are impaired well before a pour point or freeze point is reached. Instruments need to be protected. The disabling effect on instruments is progressive, increasing with increasing viscosity. Unlike freezing and pour point phenomena, there is no clear cut point where instruments are disabled and protection is needed. The decision on when to protect instruments is arbitrary, based on judgement and successful experience. The Company uses two rules of thumb in making these decisions, one for nonflowing lead applications and the other for flowing leads. Nonflowing Applications (Low Displacement Instruments). The maximum viscosity for nonflowing leads on low or nil displacement instruments (force balance or equal pressure and flow transmitters) is 4000 centistokes. Good fluidity is not required, only the ability to transmit pressure. At 4000 centistokes, fluids will still adequately transmit pressure. Protection is required when expected ambient temperatures will be low enough for viscosities to go above this level. Flowing Applications (High Displacement Instruments). The maximum viscosity for flowing leads and for nonflowing leads on high displacement instruments, such as motion balance bellows-type differential pressure flow meters, is 1000 centistokes. Fluidity is required in these applications where the liquid must flow freely through process connections. Examples are level glasses and analyzer sample leads or instruments in which significant liquid movement is required when measurement changes cause volume changes within the instrument body. At 1000 centistokes liquids are still fluid enough for this to happen. Protection is required when expected ambient temperatures will be low enough for viscosities to be above this level.

How To Solve High Viscosity or Low Pour Point Hydrocarbon Problems Use the same protective methods as for high pour point hydrocarbons: heat tracing to maintain high enough fluid temperatures; sealing; and purging. In heat tracing, use the same method as for high pour point hydrocarbons to maintain a safety margin of 20°F.

Chevron Corporation

1500-19

July 1999



1525 Fouling Fluids and Coking Hydrocarbons Fouling Fluid and Coking Hydrocarbon Problems Some process streams carry solid particles. Coke or catalyst particles in oils and the solids in slurries are typical. These particles will settle out and plug instrument leads. This is true of the flowing leads on level instruments as well as on nonflowing leads. The velocity in flowing instrument leads is too low to keep the solids in suspension. Heavy hydrocarbons at or above the coking temperature cause similar problems. As long as the hydrocarbons flow fast enough in the process lines, residence times are too short and velocities are too high for plugging problems to occur. However, where these liquids slow down or are stagnant at the entrance to and just inside nonflowing instrument leads, residence times are long and the temperature remains high. Coke forms, accumulates, and plugs the leads. This will happen in flowing instrument leads also because the liquid velocities are never high enough to prevent it.

How To Solve Fouling Fluid and Coking Hydrocarbon Problems The only solution is to use purges that keep the process fluid out of the instrument leads. As discussed below, these installations must be very carefully designed. See API RP 551 and Section 1540, “Purges,” for more information. Installation Design. Here are some of the factors to keep in mind when designing these purged installations. Most of them apply to any purged installation and are covered in the Section 1540. They are mentioned here because design details are very critical in these fouling and coking services.

July 1999

•

Purge rates and pressures: Must be high enough to reliably exclude the process fluid under all normal and abnormal operating conditions without introducing unacceptable errors in the measurement.

•

Purge location: Should normally be close to the process tap to minimize the length of purged lead and the effect on the instrument reading. When it is necessary to inject the purge at or near the instrument, be sure to design for an acceptably small effect on the instrument reading.

•

Lead filling: Provisions need to be made for getting purge fluid into the instrument body and the leads between the instrument and the purge connection. Don’t forget plant startup.

•

Purge fluid selection: This is critical to the design. Consider compatibility with the process, reliability of the source of supply, availability at plant startup, and density of the purge compared with the process fluid. For hot process fluids, consider whether the purge will form vapor when it contacts the process fluid. Try to avoid purge fluids that require winterizing.

•

Rod out: provisions should be made at the root valve so the process connection can readily and safely be reamed out.

1500-20

Chevron Corporation



Diaphragm Seals. Sometimes these purges are combined with diaphragm seals to reduce the length of purged lead and to ensure that the process fluid will not enter the instrument body. The purge should be introduced at the diaphragm face so that it will “wash” the diaphragm and keep it free of solids. Be sure to consider the maximum temperature rating of the diaphragm seal and its fill fluid. The purge flow alone cannot be relied on to keep the seal below the process fluid temperature, particularly on close-coupled installations. Reasonably reliable protection against high temperature can be obtained with a “dead leg” that will remain full of purge liquid when purge flow stops. The “dead leg” must be long enough so that the seal face will be kept “cool” by the heat loss from the stagnant liquid.

1526 Salts and Other Solids Salts and Other Solids Problems Instrument leads and process taps are sometimes plugged by the deposition of salts and other solids from nonhydrocarbon streams. These streams can be either liquids, wet gases, or wet vapors. The solids are usually inorganic salts such as sulfates, sulfides, ammonium compounds, chlorides, etc. Often they appear to be complex mixtures with no clear identity. They are sometimes quite hard and difficult to remove, and often do not readily dissolve in water. Liquid Sources. The liquid streams which may cause problems are the water streams and the water solutions of various chemical reagents. Although all such streams are suspect and should be checked, the typical offenders include sour waters, waste waters, wash waters, caustics, and the specialized chemical reagents sometimes used in large quantities in process plants. Gas and Vapor Sources. The vapor and gas streams are harder to pin down. They usually are streams containing water vapor and some water soluble compounds such as ammonia, hydrogen sulfide, and sulfur dioxide. Some typical high risk applications are in sulfur plant feed, the associated tail gas plant streams, and the vapor phase from sour water and wash water separators. Solid Formation. For most applications it is difficult to predict whether solids will be deposited, where in the system they will form, and whether they will be a problem. In general, the mechanism of solid formation is not well understood. It undoubtedly varies greatly between applications. A particular process stream may be a problem in one plant while the same stream may not be a problem in a similar plant at a different geographic location. In any event, actual operating experience is the only reliable guide to which process streams will cause problems, which applications will have problems, what the problems will be, and under what operating conditions the problems will occur. The designer should seek out and use this experience to determine if a problem will exist for specific applications and to define the problem if one exists. Information from such experience is usually available, although it may be hard to find.

Chevron Corporation

1500-21

July 1999



How To Solve Problems with Salts and Other Solids Because the problems associated with the formation of salts and other solids are so varied, there are no general solutions that can be relied on. Plugging disables the instruments, but the conditions for plugging vary greatly. Different applications will therefore have somewhat different problems and may require different protective methods. Purging. Purges are most commonly used, both for liquid and gas/vapor applications. The idea is to exclude the problem fluid from the leads, sweep it out if it does get in, and wash away any solids that do form. Water purge of some sour water, waste water, and wash water applications and nitrogen purge of some sulfur plant feed streams are some examples. Successful purge fluids vary greatly. They are definitely application-specific. Sealing. In other applications, properly designed and installed seals may be the answer. Sometimes heat tracing might provide effective protection by keeping leads and connections hot enough to prevent condensation or solid formation. Use Experience. Here again, information from actual operating experience with both successful and unsuccessful protective methods is generally available, although sometimes hard to find. This is a difficult design area. Solutions that look good on paper have failed in practice. Take full advantage of available experience. Use methods that have proven successful, or at least avoid ones that have failed.

1527 Condensation Condensation Problems The problem with condensation is different from the problem with solids. The leads do not plug. They either fill or partially fill with liquid. Sometimes they fill with liquid which then boils off. This fill-and-boil cycle can be repeated as conditions change. The leads can also alternately fill and drain in self-draining arrangements where the condensation rate exceeds the capacity of the lead to carry it away. Each of these conditions results in erroneous or erratic instrument readings. In analyzer sample leads these conditions result in stalled flow or nonrepresentative samples delivered to the analyzer. Typically, the condensate is either hydrocarbon liquids from hydrocarbon vapors, water from wet vapors or gases, water from steam, or liquid ammonia from anhydrous ammonia vapor. Because liquid can condense whenever the dew point of the process fluid is above the temperature of the instrument lead, condensation can be a problem at normal and high ambient temperatures as well as at low temperatures. Most vapor or gas process streams are at their dew point under process conditions. Because process temperatures are usually well above ambient temperatures (usual process temperatures are above 100°F), there are many opportunities for trouble. Steam, fractionating column overheads, vapors from separators, and gases from knock out drums are a few examples.

July 1999

1500-22

Chevron Corporation



Certain applications are especially prone to condensation problems. These are characterized by long leads that are exposed to ambient weather conditions yet must remain dry (or at least open) to perform properly. Examples include long sample leads for analyzers, differential pressure level installations, and differential pressure measurement across reactors and fractionating columns. In these differential pressure measurements the two taps are at different elevations. Typically, the elevation difference is great enough so that the hydrostatic head of condensate in the leads exceeds the differential pressure being measured; and the differential pressure being measured usually exceeds the zero adjustment available on the instrument. That is why these leads must remain open. Another group of applications in the same general class are those where diurnal changes in ambient temperature cross over the dew point. Ammonia and LPG storage facilities are common examples. At night, when it is cool and heat loss from instrument leads is high, the ammonia and LPG will condense and fill the leads. During the heat of the day when it is warm and sunny the leads heat up and the liquid boils away. The result is wide swings in instrument readings with no real change in the storage vessel level.

How To Solve Condensation Problems A variety of protective techniques are used to solve condensation problems. The choices are application-specific. Fortunately, the basic problem is always the same (ambient temperature causes condensation) and the applications can easily be identified as falling under one of several protective methods. Self-Drain. By far the most common solution to condensation problems is to design so that the condensate will freely self-drain back to the process. The idea is to let condensation take place but handle it so it will not be a problem. Flowing lead level instruments (level glasses, switches, transmitters, etc.) use equalizing lines. Nonflowing leads use close-coupled, self-draining installations. Such solutions work for most line-mounted flow and pressure transmitters and can be made to work on some installations with longer leads. The key to success in designing these self-draining installations is to make sure that the liquid does not condense faster than the lead can take it away during operation at the minimum expected ambient temperature and the most severe weather conditions of wind and precipitation. This is usually not much of a problem on closecoupled, line-mounted installations except in severe climates; however, it can be a serious problem with longer leads in almost any climate. Sometimes insulating the leads to reduce the condensation rate will be sufficient. Increasing the diameter of the lead so that it will not overload with liquid may be the solution, particularly for some unavoidably long leads (reactor and column differential pressure). Seals can be used to minimize the length of lead that needs to self-drain. Self-Seal. This is probably the second most common method of solving condensation problems. The idea is to allow the condensation to occur but use it in a beneficial way. Designing a high point in the lead between the root valve and the instrument and close to the root valve causes the lead between the high point and the instrument to fill with condensate and stay filled. Excess condensate drains back to

Chevron Corporation

1500-23

July 1999



the process from the high point. This technique is called “self-sealing.” The condensate-filled lead is referred to as the “wet leg.” Usually a seal pot or equivalent is located at the high point. See Figure 1500-1. Self-sealing is widely used in measuring both differential pressure level and the flow and pressure of steam. In addition to solving the condensation problem, the wet leg provides a reliably constant reference pressure. It is seldom used for other process flow and pressure measurement. With self-sealing, the instrument is calibrated with an offset to compensate for the known hydrostatic effect of the wet leg. For flow measurement, the two high points are put at the same elevation above the instrument body so that no offset is needed. Self-sealing cannot be used in differential pressure measurement applications where the difference in elevation between the two taps is so great that the required offset is greater than can be specified with the instrument. Reactor and column differential pressures are likely to fall into this category. Self-sealing will not be successful if the condensate in the leg(s) will boil under some operating conditions. Boiling will upset the instrument reading. The condensate will boil if the expected maximum ambient temperature is above the boiling point of the condensate at the process pressure. Similar troubles occur if swings in process pressure are fast enough and wide enough to cause significant boiling and loss of the condensate. Self-sealing obviously will not work if the wet legs plug when the ambient temperature drops below the freezing point or pour point of the condensate. Winterizing is added to prevent such plugging. Liquid Seal. This solution is similar to self-sealing except the wet legs are filled with a different fluid, generally one that is compatible with the process but not miscible with the condensate. This design is intended to eliminate boiling and freezing in the wet leg. It has the obvious disadvantage that the level must be periodically checked and manually maintained. The applications and design considerations are much the same as for self-sealing. Diaphragm Seals. Diaphragm seals that are mounted at the root valve and freely self-draining into the process can solve the condensate problem. This eliminates wet legs and shortens the self-draining portion to the point where drain capacity problems are eliminated. Cold weather problems are also easier to handle. However, use of diaphragm seals is limited by the temperature limitations of the seal and its fill fluid. This technique is used mostly in freezing climates for pressure and flow measurement in steam and other services where water is the condensate. It eliminates concern over freezing of the wet legs and the need to maintain wet legs filled with anti-freeze fluid. It can also be used in differential pressure level applications to eliminate wet leg problems. The portion from the process tap to, and including, the seal body is treated the same as the process line: it is left bare, insulated only, or steam-traced using the process line tracer. Purges. A noncondensing gas or vapor that is compatible with the process can be used to purge the instrument leads to keep out the condensing fluid. The purge fluid should be introduced at the instrument so that it will keep the complete lead purged.

July 1999

1500-24

Chevron Corporation



Experience has shown that purging just the root connection will not give reliable results. The purge rate must be high enough to prevent any significant upstream migration of the condensibles while remaining low enough not to induce excessive error in the instrument reading. Purging is not widely used. Its major application is for differential pressure measurements that require very long leads. Purging is particularly useful at those installations where the taps are too far apart to use either diaphragm seals or liquid seals because (a) available capillary leads are too short to reach between the taps, or (b) liquid seals would require a zero offset greater than can be specified with the instrument. Purging with gas is usually successful in such situations if it is coupled with adequately sized self-draining leads. Typical applications are on reactor and fractionating column differential pressure measurements. Heat Tracing. Heat tracing can be combined with the other techniques for handling condensing services to prevent plugging from ambient temperature effects. Heat tracing can also be used by itself to maintain the lead temperature above the dew point. This is an expensive method and one that can be easily compromised by tracer failure. It is not widely used. In most condensing applications, other methods provide better protection. Where heat tracing is used to maintain a dry lead, it is advisable to provide a high point that will self-drain back to the process. This practice provides a degree of protection if the tracer fails or is shut off for maintenance. Analyzer Sample Leads. Analyzer sample leads are perhaps the most common application where heat tracing is used by itself. It is used to prevent the condensation that will result in nonrepresentative samples. The sample system should be designed to tolerate liquid without damage to the analyzer in case of tracer failure.

1528 Chemical Attack and Corrosion Chemical Attack and Corrosion Problems Some of the process fluids—chemical reagents (caustic, acids, solvents) and especially waste streams—are quite active chemically and can attack the materials used in instrument tubing and fittings, valves, and the wetted parts of instruments. In addition to simple corrosion, the hazards include chloride cracking, stress corrosion cracking, intergranular corrosion, hydrogen embrittlement, hydrogen blistering, and other forms of metallurgical damage. Temperature plays a part in the damage caused by some fluids, such as caustics and acids. The nonmetallic parts also are subject to attack. The solution to the problems of chemical attack and corrosion are applicationspecific. The need for protection and the methods used for protection are best determined from actual plant experience. Only some general considerations are discussed here to provide a guide to the factors that must be considered when doing instrument design work in this area.

Chevron Corporation

1500-25

July 1999



The designer is strongly advised to seek out and employ materials and techniques that have proven successful in actual plant experience.

How To Solve Problems of Chemical Attack and Corrosion The main defense against chemical attack and corrosion is to use materials of construction that can tolerate the process fluid environment, and to avoid the hazardous temperatures that cause temperature sensitive situations. These cautions must be observed for all of the process wetted material, not just the instrument parts. Resistant Materials. Most instruments, control valves, diaphragm seals, instrument tubing and fittings, instrument valves, etc., are available in a wide range of corrosion-resistant materials. In addition to stainless steel, hastelloy, monel, alloy 20, tantalum, titanium, other ferrous and nonferrous alloys are also readily available. Many items are available with inert nonmetallic coatings. Some are available made wholly of such nonmetallic materials. For wet hydrogen sulfide services, most manufacturers can supply equipment made to the National Association of Corrosion Engineers (NACE) recommendations for that service. Seals and Purges. Seals, purges, or a combination of the two can be used to provide protection in this area by keeping the aggressive fluid away from parts that are easily damaged. Again, specific successful plant experience is the best guide. Winterizing. Rather than a solution, winterizing is an additional concern with corrosion because improper winterizing can inadvertently magnify corrosion problems. With temperature-sensitive fluids (e.g., some caustics and acids), heat tracing must be designed so that the temperature will not rise enough to cause damage under the maximum expected ambient temperature. This concern should be addressed whenever heat tracing, especially steam tracing, is applied to any of these aggressive fluids. Consider self-limiting electric tracing for these applications.

1529 High Temperature Fluids Problems with High Temperature Fluids The problem with high temperature process fluids is very straight-forward: they can damage the instruments they come in contact with by overheating these instruments. The electronics, internal fill fluids, and materials of construction all have high temperature limits. Process fluids in petroleum processing plants are frequently above the temperature limits for instruments. Oils at 700°F and gasses at 1000°F are not uncommon. These will quickly and permanently disable most instruments.

How To Solve Problems with High Temperature Fluids The general answer to the problem is equally straight-forward: do not expose the instrument to the high process temperature. In a very few applications there may be instruments or measurement methods that either tolerate the high temperature or somehow side-step the problem. Radioactive

July 1999

1500-26

Chevron Corporation



level measurement is one example. These instruments or methods are very special cases and their use should be based strictly on successful experience. Several possible solutions using conventional instruments and methods are available. Point of Measurement. Move the point of measurement to a location where the process stream has been cooled down. The same process control or monitoring results may be obtainable at a location downstream of heat exchangers or boilers, or upstream of a furnace. Relocating control valves and the point of measurement for flow and pressure is often possible. Dead Legs. A relatively short run of tubing or piping with stagnant liquid in it is quite effective at reducing the temperature. Most nonflowing instrument connections are long enough to cool down even the hottest services to a temperature where instruments are safe. However, this is a very risky method to use. Careful design is mandatory to minimize the hazard of exposing the instrument to damaging temperatures during startup, shutdown, upsets, and maintenance. Check to see if the cooling will introduce other problems such as plugging or water accumulation. Seals. Use diaphragm seals that can tolerate the temperature. Be especially careful about the fill fluids. They are usually very viscous. In cold climates they may need heat tracing to give reliable and acceptable performance. For very high temperature applications, a sodium-potassium material is available for use as the fill in diaphragm seals and associated capillaries. Because this is a very hazardous material, its use is strongly discouraged. Purges. Purges to keep temperatures down by displacing the hot fluid are not highly recommended. However, sometimes they are the only solution. Introduce the purge at the instrument or flush through the instrument. Keep leads long enough so the dead leg will protect the instrument when the purge fails. Select the purge fluid density and arrange the installation so the hot process fluid will not drain into the instrument (or the cool purge fluid drain out) when the purge stops. In general, be very careful to minimize the hazard of exposing the instrument to damaging temperatures during startup, shutdown, upsets, and maintenance. Check to see if the cooling will introduce other problems such as plugging.

1530 Seals 1531 General Information Seals protect instruments by isolating them from the process fluid. The seal lies between the process fluid on the process side and a fill fluid on the instrument side. The fill fluid is compatible with the instrument and is suitable for the application. API RP 551 describes seals and their installation. Section 1520, “Process Fluid Considerations,” identifies specific services where seals are used and discusses design considerations that apply to those applications.

Chevron Corporation

1500-27

July 1999



Nonflowing Leads Seals are static arrangements that can only be used on instruments with nonflowing (pressure transmitting) process connections.

Diaphragm Seals and Liquid Seals Two different sealing methods are used: diaphragm seals and liquid seals. The difference is in how they separate the process fluid from the seal fill fluid. Diaphragm seals use a flexible membrane. Liquid seals use a free liquid surface. See Figure 1500-1.

Reliability Seals, especially the diaphragm-type, are reliable and require relatively little maintenance. They use static and passive techniques which are not disabled or compromised by the loss of a purge fluid, steam, or electric power.

Short Process Connections The seals can be remote from the instrument. This makes it easy to install short leads between the seals and their process connections. It also makes it easier to make these leads freely self-draining. Other characteristics of these two different types of seals are covered in Section 1532, “Diaphragm Seals,” and Section 1533, “Liquid Seals.”

Design Considerations Many of the considerations that are important to the successful application of both types of seals have been mentioned in Section 1520, “Process Fluid Considerations,” which cover each of the various types of process fluids (see the heading “Seals” for each fluid). Many of them will be repeated here so that this discussion will provide a reasonably complete guide.

1532 Diaphragm Seals Diaphragm seals are also called “gage protectors” or “chemical seals.” They use a thin flexible diaphragm to hold a seal fluid in the instrument and to separate this from the process fluid. See Figure 1500-7 and the top part of Figure 1500-1. The diaphragm is typically made of nil corrosion metal. Nonmetallic diaphragms, plastic coated diaphragms, and metallic diaphragms with a plastic membrane on the process side are all available and may be acceptable where a suitable metallic diaphragm cannot be found. The fill fluid is a suitable noncompressible liquid. It is compatible with the instrument and can usually tolerate a wide range of ambient temperatures.

Integral and Capillary Types Diaphragm seals are available close-coupled or integral with the instrument. They are mostly used on pressure gages and pressure switches and are connected to the instrument with capillary tubing of varying lengths. Integral diaphragm seals are

July 1999

1500-28

Chevron Corporation



Fig. 1500-7 Typical Diaphragm Seal (Used with permission. Ashcroft is a registered trademark of Dresser Industries Instrument Division.) GAGE CONNECTION

FILLING SCREW

TOP HOUSING

OIL FILLED DIAPHRAGM CAPSULE

BOTTOM HOUSING

FLUSHING CONNECTION

PROCESS CONNECTION

available on pressure and differential pressure instruments as well as pressure gages and switches. See Figure 1500-8.

Temperature Limits Both the diaphragm materials and the fill fluids have temperature limits that must be considered when applying diaphragm seals. At excessively high temperatures, fill fluids break down chemically or start to vaporize. As discussed in Section 1529, “High Temperature Fluids,” seals can sometimes be protected from high process temperatures by stagnant legs or cool purges. At excessively low temperatures, fill fluids may become too viscous to quickly and reliably transmit pressure. This can be handled by heat tracing.

Protection from Process Fluids Diaphragm seals can provide protection from most of the troublesome process fluids. For guidance on specific applications, refer to the sections on process fluids.

Chevron Corporation

1500-29

July 1999



Fig. 1500-8 Diaphragm-and-Capillary Systems (Adapted from API RP 550, 1, 8, Figure 8-2, 1985) Courtesy of American Petroleum Institute SCREWED OR SOCKET WELD

TRANSMITTER

TYPE DIAPHRAGM CAPSULE

CAPILLARY OF LENGTH REQUIRED

A. PRESSURE TRANSMITTER

BLIND FLANGE

P TRANSMITTER

WAFER TYPE DIAPHRAGM CAPSULE

CAPILLARIES NOTE: RUN CLOSE TOGETHER TO MINIMIZE TEMPERATURE EFFECT.

B. LIQUID LEVEL OR DIFFERENTIAL PRESSURE TRANSMITTER

Short Process Connections Diaphragm seals are generally installed with short process leads. These connections can easily be made reliably self-draining when the applications call for it. When they must be fully self-draining, mount them with the diaphragm horizontal so no pocket is formed on the process side. Such pockets will degrade performance if they fill with solids caused by freezing, sedimentation, coke formation, etc.

Winterizing Simplification Using diaphragm seals can simplify winterizing. They eliminate the need for troublesome winterizing of instrument bodies. Coupled with short process connections, they can eliminate the need for separate winterizing of long instrument leads. Only

July 1999

1500-30

Chevron Corporation



the short space from the process tap through the seal body needs to be winterized. Usually the same winterizing as the main process line is appropriate and can be done using the main line tracer. Insulation alone should be considered in applications where enough heat will be conducted from the process connection or when continuous condensation (steam) will keep the connection warm enough.

Differential Pressure Level Where diaphragm seals are used in differential pressure level measurement, it is possible to expose the seals and the instrument to a differential pressure equal to the full process pressure when the assembly is being depressured for removal (there is no “bypass valve” in the capillaries at the instrument). In high pressure applications, this may adversely affect the instrument or the seal system. To eliminate the risk, a valved balance line can be installed between the taps. This bypass can be opened to “balance” the pressure across the instrument after the root valves have been shut.

Instrument Performance Effects Diaphragm seals usually have a small adverse effect on the instruments to which they are connected. They add hysteresis and decrease accuracy, usually an acceptably small effect that should, however, be taken into account when considering diaphragm seals for narrow span instruments. This is especially true for relatively large displacement instruments such as motion balance field pressure controllers. Because of these effects, diaphragm seals are not recommended for applications where process pressures are lower than 15 Psig.

1533 Liquid Seals Liquid seals use a free liquid interface to isolate the instrument from the process fluid. They use gravity to hold the seal liquid in the instrument and its portion of the instrument lead while holding the process fluid in its portion of the lead. The installation is arranged so that the surface of the seal liquid will remain at a constant location in the lead between process tap and instrument. See Figures 1500-9 and 1500-10. The process side of the leads are arranged to prevent loss of fill liquid both during operation and shutdown (self-drain process side for fill liquids heavier than the process fluid, goose-neck offset in process side for those lighter than the process fluid).

Benefits Liquid seals provide two major benefits: they protect both the instrument and its lead from any damaging effects of the process fluid. The constant location of the liquid surface provides a constant hydrostatic head of liquid above the instrument. This eliminates the inconsistencies that could be caused by varying amounts of process liquid collecting in the leads. This constant head provides a constant reference for calibrating the instrument, so that it will always read the pressure at the fixed liquid surface. This feature is commonly referred to as the “wet leg.” It is widely used for the upper connections in differential pressure level and in steam

Chevron Corporation

1500-31

July 1999



Fig. 1500-9 Liquid Seals on Low Displacement Flow Transmitter (Adapted from API RP 550, 1, 8, Figure 8-3, 1985) Courtesy of American Petroleum Institute

ZEROING BYPASS ABOVE SEAL

LINES SELF-VENTING SEAL TO TEE

TO FLANGE TAPS

BOTH SIDES (SEE NOTE)

1/2" NPS

T

R T

A

N

O

S

IT M

T

E

R

NOTE -

SEAL POTS MAY BE USED ON ANY APPLICATION AND ARE REQUIRED FOR

SEAL LIQUID HEAVIER THAN

HIGH DISPLACEMENT INSTRUMENTS.

LINE FLUID

flow applications where the leads stay full and any excess condensate drains back into the process.

Self-Sealing A liquid seal installation is “self-sealed” if the sealing fluid is the process fluid. This technique is widely used. It is entirely satisfactory if the process fluid is both compatible with the instrument wetted parts and suitable for the full range of ambient temperatures. Self-sealing is used primarily to obtain the constant reference level of the wet leg and to protect the instrument from high process temperatures. Steam flow and pressure measurement and the upper connections on differential pressure level measurement are common examples. Self-sealed wet legs can be winterized if necessary, but this is not general Company practice. An anti-freeze liquid is generally used for the seal fill liquid in applications where the self-sealed wet leg would not be suitable for the lower ambient temperatures.

Seal Pots Seal pots, also called “seal chambers,” and “condensate pots” are very similar and both perform the same function. They form a “wide spot in the line” that provides a large pool of liquid where the seal liquid surface is located. This means that the

July 1999

1500-32

Chevron Corporation



Fig. 1500-10 Liquid Seals on Differential Pressure Level Transmitter (Adapted from API RP 550, 1, 8, Figure 8-3, 1985) Courtesy of American Petroleum Institute 6($/ 72 +(5(

%<3$66

Vessel or Bridle

6($/ 72 +(5(

72 75$160,77(5

elevation of the surface will be less affected by changes in the volume of liquid in the lead. Seal pots are mandatory for high displacement instruments so that the volume of fluid displaced by movement of the instrument internals will not significantly affect the elevation of the liquid in the wet leg. They are not mandatory, but usually are used on low and nil displacement instruments because they provide added assurance that the level will be maintained. They are most often not used on steam flow and pressure measurement because the condensation rate is so high (any loss is quickly recovered).

Seal Liquids Seal liquids should be immiscible with the process fluid and suitable for the full range of ambient temperature. Fluids both heavier and lighter than the process fluid are available. Generally, a heavier seal fluid is preferred because it is easier to fill leads and there is less risk of losing the fill if the process line is drained or runs empty. If the seal fluid is miscible with the process fluid (e.g., ethylene glycol anti-freeze in a service with condensing water), then the seal fluid must be recharged at regular intervals.

Chevron Corporation

1500-33

July 1999



Maintenance Concerns Liquid seals with or without seal pots require more maintenance diaphragm seals. Each installation must be checked periodically to be sure that the fill liquid is still in good condition (not contaminated), that the fill levels are correct, and that the seals are still functioning properly. The ethylene glycol and water mixture that is most commonly used absorbs water. It may absorb enough water from the process stream to cause it to freeze if it is not periodically replaced.

Identification of Liquid Seals To aid checking and refilling, provide each installation with permanent nameplates at the seal pots and at the sealed instrument. Engrave the nameplates with the instrument tag number and the seal fluid that must be used, like this: 5-FT-1234 SEAL WITH DC-200 These nameplates need to be durable and easy to read when they get old and dirty. Use deeply engraved laminated plastic with black letters on a white background or deeply engraved stainless steel.

Bypass Valves and Seal Liquid Loss Another problem is loss of seal liquid because of incorrect operation of the bypass valve in the manifold at the instrument or of the drain valves. The seal identification nameplate is a help here. Sometimes a seal pot is provided to give enough volume to provide some protection and to serve as a warning “flag” even though it is technically not needed. To minimize the seal fill loss problem, Figures 1500-9 and 1500-10 show pressure equalization bypasses on the instrument impulse lines, between the root valves and the seal pots. It should be noted here that the bypass shown in Figure 1500-10 is not for zeroing the transmitter. In the absence of a bypass valve at the transmitter it provides a means to “balance” the pressure on the two sides of the transmitter when it is being taken out of service. This prevents the transmitter from being exposed to a differential pressure equal to the full process pressure.

1534 Need for Additional Winterizing Winterizing is frequently combined with the use of seals. Using diaphragm seals can greatly simplify the winterizing required.

On the Process Side The process-containing part of the instrument lead, the section between the process tap and the sealing diaphragm or liquid surface, often must be winterized. Steam and some hydrate formers are typical examples. In general, if the process line or vessel is winterized, the process side of the seal will need the same winterizing. With close-coupled diaphragm seal installations, the process line tracer can frequently be used to trace the process side of the seal. Figures 1500-4 and 1500-11 illustrate these ideas.

July 1999

1500-34

Chevron Corporation



Fig. 1500-11 Winterizing Added to Diaphragm Seals (From Drawing GD-J99742-5) Pressure Instrument

Capillary Tube Gage Protector Pressure Gage

Steam Supply

Manifold Valve

To Trap

When the process side of seals is winterized, be sure to carry the winterizing far enough onto the seal pot or diaphragm seal body. Experience has shown that heat loss from the seal fill side can cause problems on the process side when this has not been done. Usually, carrying only the insulation over the remainder of the seal pot or diaphragm seal body will provide adequate protection without negative effects on the seal fill.

On the Seal Side Some applications may require that the fill fluid portion be winterized also. This will be necessary with liquid seals in severe climates where a suitable anti-freeze fill liquid cannot be found and with capillary diaphragm seals where the seal fill will be too viscous at the minimum expected ambient temperature.

1540 Purges 1541 General Information API RP 551 describes the types of applications for which purges are commonly used and describes methods for purging. Section 1520, “Process Fluid Consider-

Chevron Corporation

1500-35

July 1999



ations,” also identifies the specific services where purges are used and discusses design considerations that apply to those applications. Purges are used in liquid, gas, and vapor services. They keep the process fluid out of the instrument, its process leads, and the process taps. The purge fluid is intended to sweep the leads and taps clear of any process fluid that might enter them. To do this, the purge fluid is continuously injected at a controlled rate into the instrument leads where it flows into the process. Figures 1500-2, 1500-12, and 1500-13 illustrate these principles. Fig. 1500-12 Purge Installation for Plugging Service (Adapted from API RP 550, 1, 8, Figure 8-7, 1985, Courtesy of American Petroleum Institute)

A

CHECK VALVE

FROM AIR OR GAS PURGE UNIT TURN LINES AT A IF INSTRUMENT IS BELOW TAP

SEE FIGURES 13 AND 14

A VESSEL OR LINE

FROM LIQUID PURGE UNIT VALVES ARRANGED TO PERMIT REAMING OF TAP

The purge fluid must be clean and compatible with the instrument, its leads, and the process. It is usually suitable for the full range of ambient temperatures, although sometimes this is not possible and winterizing has to be added.

Flowing and Nonflowing Leads Purges can be used on both flowing and nonflowing instrument leads. While the design considerations are essentially the same for both types, back-flow is much stronger in flowing leads since fluids must flow in and out on level changes.

July 1999

1500-36

Chevron Corporation



Fig. 1500-13 Orifice Tap Purges (Adapted from API RP 550, 1, 8, Figure 8-8, 1985, Courtesy of American Petroleum Institute) FROM PURGE UNITS SEE FIGURES 13 & 14

TURN LINES UP IF METER IS ABOVE CHECK VALVE (TYP.) TO METER

PURGE LIGHTER THAN FLOW FROM PURGE UNITS SEE FIGURES 13 & 14

TURN LINES UP IF METER IS ABOVE CHECK VALVE (TYP.)

TO METER

PURGE HEAVIER THAN FLOW

Design Considerations Many of the considerations that are important to the successful application of purges have been mentioned in Section 1520, “Process Fluid Considerations,” which covers each of the various types of process fluids (see the heading “Purges” for each fluid). Many are also discussed here to provide a reasonably complete guide.

1542 Disadvantages of Purges Flowing purges are the least reliable of the three protective methods (seals, purges, and winterizing). They are high cost and high maintenance systems. They are difficult to design, install, and maintain so that they work consistently and reliably with few problems. Flowing purges should be used as a last resort where “in-line” methods of measurement are not available and where seals or winterizing will not work. Even so, there still are a large number of purged installations in Company facilities, such as crude unit vacuum sections, heavy oil sections of other plants, and plants handling particularly dirty fluids.

Reliability Considerations Purges are complex active systems. They are inherently less reliable than either the static seal systems or the simpler active heat traced systems, which present fewer

Chevron Corporation

1500-37

July 1999



opportunities for problems. Purges need a consistent source of purge fluid and they rely on an active system of pumps, lines, and controls to deliver the fluid at adequate pressure and flow rate with the flow to each lead properly controlled. All of these items are subject to failure and upset at any time. Clearly, great care must be taken in the design of these purge systems. Consider all of the factors that affect reliability. These include the following. •

Quality and character of available purge fluids

•

Sources of purge fluids

•

Reliability of the supply

•

Delivery system supplying the fluid

•

Delivery and distribution system supplying fluid to the leads

•

Metering system controlling the injection rate into the leads

•

Possibility and hazard of back flow of process fluid into the leads and into the purge system

•

Availability and quality of purge fluid during pre-startup, startup, upsets, and shutdown

•

Behavior of the design under both normal and abnormal operating conditions

•

Failure of the purge fluid (stalled flow, loss of pressure, etc.)

In addition, the design should consider any other features, such as adding winterizing or diaphragm seals, that may materially improve the reliability of the connected instrument.

Instrument Performance Effects The flow of the purge fluid in the instrument leads causes a pressure drop which produces an error in the instrument reading. The purge system and the flow rate must be designed so that the error is acceptably small. At the same time, the flow rate must be large enough to keep the process fluid out of the leads. For information on typical purge flow rates, see Section 1543, “Purging Methods,” under the heading “Control the Flow.” For flow instruments, design to keep the pressure drop equal in both leads. For pressure instruments, the pressure drop can usually be kept low enough that the error is within the normal percentage error of the instrument. Low pressure and vacuum measurements are the most difficult. In flowing lead installations, an additional error is introduced by the difference in density between the process liquid and the purge liquid because the purge liquid will fill the external cage or chamber of the level instrument. Sometimes this error is small enough to be ignored. When it is not, it can usually be compensated for by instrument calibration (a displacement type transmitter), by adjusting the elevation

July 1999

1500-38

Chevron Corporation



of the cage (trip point of level switch), or by external markings to show the difference between actual and indicated readings (gage glass).

1543 Purging Methods Various systems are used for supplying purge fluid and delivering it to the individual instrument leads.

Gear Pump System One system uses a small electric motor-driven positive displacement gear pump to deliver purge to several instrument leads at once. While this system has not been widely used in Company facilities, it does have several advantages and deserves serious consideration. The pump supply comes from a large reservoir of purge liquid, called a day tank, which is sized to require refilling at reasonably long intervals. The purge control configuration at each individual instrument lead includes an isolation block valve, a rotameter with a needle valve for flow control, a check valve, and another isolation block valve at the point of injection into the lead. The gear pump is sized to supply a constant flow rate greater than the sum of the connected purges. To handle this, a discharge pressure regulator returns the excess flow to the reservoir. This allows for independent adjustment of the flow rate of each purge. This system supplies a smooth nonpulsating flow at nearly constant pressure, which gives consistent and reliable control of each purge delivery rate. However, the failure of one pump will stop purge delivery to several instruments.

Lubricator Pump System A pumping system that has frequently been used on Company facilities in the past employs a multipump lubricator of the type used on large reciprocating compressors. The system has several disadvantages and has fallen out of favor. Better systems are available. The lubricator pump is discussed in some detail here because of its former popularity and wide use in existing Company facilities. The lubricators have several plunger pumps on a common shaft driven by one small electric motor. A separate plunger pump is used for each individual instrument lead. Failure of one plunger pump will stop the flow to only one instrument lead. The capacity of each pump can be adjusted to meet the needs of its instrument lead. Because no other flow control is needed, the purge configuration at each instrument lead connection is simplified to just a check valve and an isolating block valve. This system delivers a pulsating purge flow to the instrument leads. Care must be taken to minimize the effects of the pulsations on differential pressure type instruments in flow and level service and on very low pressure measurements. Injecting the purge close to the process taps and eliminating the restriction fitting at the root connection will help.

Chevron Corporation

1500-39

July 1999



In addition to the problem of pulsating flow, these systems have several other disadvantages. Failure of one driving motor will stop the flow to several instruments. Pump capacity is hard to adjust and it is very difficult to tell with any certainty what the individual flow rates are. It is impossible to provide really high flow rates for flushing out lead blockage. In fact, the maximum capacity of the pumps is so limited that they can be used only on nonflowing leads and sometimes cannot provide enough flow for adequate protection even there. The required plumbing and controls for the supply and delivery systems are complex and bulky. These lubricators and their built-in reservoirs were originally designed to supply oil to reciprocating compressor rods and bearings. The flow rates for that service are much lower than for instrument purge, so the built-in reservoir is quickly exhausted. A day tank is required along with the piping and automatic refilling controls needed to keep the built-in reservoirs from running dry. The pumps have no built-in relief valve so the discharge of each pump needs a bypass valve back to the reservoir as well as a block valve to allow taking the instrument out of service. The bypass must be opened before the block can be closed. This depressures the purge line so the system must rely on the check valve at the injection point to prevent back flow. All too often the check valve leaks and causes trouble.

Chemical Injection Pump System Another system, recently installed in Company facilities, uses a separate reciprocating positive displacement chemical injection pump for each instrument lead. It has several advantages and should be given serious consideration where the process pressure will permit its use. Each chemical injection pump is driven by its own piston or diaphragm using pressure-regulated instrument air for the motive force. At the connection to its instrument lead, each purge control configuration includes a block valve, a rotameter with a needle valve for flow control, a check valve, and another isolation block valve. This system is highly reliable and eliminates most of the disadvantages of other systems. Each pump takes suction directly from the day tank so there is no interposing reservoir to worry about. Instrument air is a highly reliable power source, usually more reliable than electric power. The pumps and drivers are simple and slow moving so they should be long lasting. Failure of one pump will stop only one purge flow. Pressure is adequately controlled by the air pressure regulation. Flow is reasonably smooth and can be adequately controlled by the rotameter and needle valve. No bypass valve is needed because the pump will merely stall without damage when an isolating block valve is closed. This greatly reduces the risk of back flow and simplifies the plumbing.

Plant Flush System This type of system takes advantage of a flush fluid system intended primarily for other use. Some plants have flush supply systems that are piped throughout the plant. These are installed to protect other equipment such as pumps and valves and to provide flushing during shutdowns and upsets.

July 1999

1500-40

Chevron Corporation



These plant-wide systems can be used to supply and deliver instrument purge where they meet the needs of that service. Perhaps the most important of these needs is the ability to provide purge fluid prior to and during start up. This is necessary because the instrument leads must be filled and purge flow established before process fluids start to flow. Using these systems for purge supply and delivery greatly simplifies the overall instrument purge system. Separate pumps and day tanks are not needed. Only isolation, flow regulation, and back flow prevention are required for each purge. Plant flush systems do have the disadvantage that all instrument purges stop when the plant-wide system fails. Since these flush systems are usually driven by centrifugal pumps, the purge lines will depressure on failure, introducing the risk of back flow from check valve leakage. Of course, if the main system fails, instruments may be the least of the problems.

Utility System Many applications can use plant utility supplies as sources for purge fluid. Typical possibilities are steam, nitrogen, hydrogen, fuel gas, natural gas, utility air, and plant water. These are treated in a manner similar to the plant-wide flush system supplies except that there is much greater concern over possible back flow into the utility systems. Because of the possibility of contamination from back flow, potable water and instrument air should not be used for purging against process fluids.

Back Flow Concerns Back flow of the process fluid into the purge delivery and supply system is always a concern. However, some applications are particularly hazardous. Back flow of certain process fluids, especially into a utility or plant flush system, can have particularly disastrous results. Acids, caustics, and other highly aggressive or highly toxic chemicals fall into this category. In these situations serious consideration must be given to independent purge supplies or at least to some means to maximize the isolation between the process and the supply in the event of serious back flow. Break tanks and adequately sized and instrumented surge vessels have been used to provide isolation.

Purge Connection Configurations The configuration of the purge system at the point of injection into the instrument lead is largely determined by the type of supply and delivery system used. Figures 1500-2, 1500-12, and 1500-13 illustrate various features. These configurations fall into two groups—those that require flow control and those that do not (their flow is controlled elsewhere). Of the supply and delivery systems discussed above, only the lubricator pump does not need flow control at the point of injection. There the purge rate is controlled by adjusting the capacity of the plunger pump. Figure 1500-2 shows a typical configuration that includes flow control. Where flow control is not required, the configuration consists of a check valve and an isolation

Chevron Corporation

1500-41

July 1999



block valve. Figures 1500-12 and 1500-13 illustrate this if the purge comes from the lubricator assembly, rather than from a purge unit as shown on the figures. These three figures illustrate other features that are important to successful purge connection configurations. An isolation block valve is located at the point of injection. This arrangement permits maintenance work on either the instrument or the purge system. A check valve is located just outside the isolation valve to eliminate, or at least minimize, back flow into the purge system on loss of purge pressure. The instrument lead between the injection point and the root valve is offset (or the instrument leads are sloped) either “up” or “down.” This arrangement prevents process fluid from draining into the instrument lead or purge fluid from draining into the process when purge flow stops. The direction of the offset depends on the relative densities of the purge and the process fluids. On close-coupled, linemounted installations the same results are obtained by mounting the instrument above or below the process taps. Of course the offset will not help if the check valve leaks and the purge pressure drops below process pressure. As Figure 1500-2 shows, a second block valve is located just upstream of the flow control assembly. With the other isolation block valve, this permits maintenance of the items in the purge connection configuration.

Purge Control System Configurations Purge rates and pressures must be high enough to reliably exclude the process fluid (or at least maintain a path for the transmission of pressure) under all normal and abnormal operating conditions without introducing unacceptable errors in the measurement. That is the purpose of the control system located at the point of injection. Figures 1500-14 and 1500-15 illustrate two of the most commonly used configurations. Many other workable configurations, including variations of these, are used, but these two illustrate the basic principles. Control the Flow. The system represented in Figure 1500-14 controls flow very simply with a properly sized restriction orifice (RO). The system works best where the process pressure is reasonably constant. The pressure regulator upstream of the RO assures a reasonably constant pressure drop and so a constant flow. The pressure regulator can be omitted if the purge supply pressure is constant and gives a reasonable pressure drop across the RO. If both the supply and process pressure vary, the pressure regulator can be replaced by a differential pressure regulator that controls the pressure drop across the RO. Restriction orifice systems are well suited to gas purge applications, especially where the supply pressure is high enough to give critical flow through the orifice. Restriction orifice arrangements have some disadvantages. The flow cannot be readily adjusted. Changing the pressure regulator setting is effective only if the supply pressure is very high relative to the process pressure. There is no direct indication of the flow rate. It must be inferred from the pressure gage readings and the size of the RO bore.

July 1999

1500-42

Chevron Corporation



Fig. 1500-14 Restriction Purge Flow Regulation—Typical Configuration

NOTE 3 NOTE 1

P1

PURGE

R0

P1

SUPPLY

NOTE 2 AT INSTRUMENT LEAD

1. REGULATOR NEEDED IF SUPPLY PRESSURE VARIES. FOR GASES, SET TO MAINTAIN CRITICAL FLOW IF POSSIBLE. FOR LIQUIDS, SET TO GIVE ADEQUATE PSID ACROSS ORIFICE. 2. BY-PASS FLUSH VALVE IF NEEDED TO CLEAR TAPS OF PLUGS. 3. RESTRICTION ORIFICE CAN BE REPLACED WITH A DRILLED GATE VALVE AND BY-PASS FLUSH VALVE DELETED IF REGULATOR WILL GIVE ADEQUATE FLOW AND PRESSURE FOR FLUSHING WITH GATE VALVE OPEN.

The system represented in Figure 1500-15 is more widely used to control flow. This system overcomes the main disadvantages of the RO systems, controlling flow by using a regulator to maintain a fixed pressure drop across a needle valve. It indicates actual flow with a rotameter. The flow can be varied while in operation by adjusting the needle valve while reading the rotameter. For liquid purges, this system maintains a constant flow even with variations in both supply and process pressures, and works adequately for gases also unless the variations in supply pressure are too great, in which case an upstream pressure regulator is needed. An acceptable purge rate has to be determined for each application. It must be high enough to keep the leads clear and low enough to keep the effects on the instrument within acceptable limits. There are no rules of thumb to make this easy. However, some idea of the typical range of flow rates may be helpful. Typical ranges for gas purge flows are 0.05 to 1 scfm for vacuum services, 0.1 to 1.5 scfm for intermediate pressure services (say up to 300 or 500 psig), and 0.5 to 5 scfm for high pressure services. Typical liquid flow may range from 0.1 to 8 GPM, with typical lubricator system flows being much lower. The lower ranges are for small nonflowing leads; the higher ranges are for large flowing leads. Confirm the Flow. The system must provide a way to confirm that it is working and maintaining purge flow. The system in Figure 1500-15 does this with a rotameter. For the system in Figure 1500-15, flow can be confirmed by manipulating the block valves while observing the readings on the two pressure gages.

Chevron Corporation

1500-43

July 1999



Fig. 1500-15 Constant Flow Regulated Purge—Typical Configuration 127(

127(

127(

3

385*( 6833/<

127(

$7 ,167580(17 /($'

1. Regulator and pressure gage needed for gases if supply pressure is not constant. Normally not required for liquids. 5(*8/$725 $1' 35(6685( *$*( 1(('(' )25 *$6(6 ,) 6833/< 35(6685( 127 &2167$17 1250$//< 127 5(48,5(' )25 /,48,'6

$5025(' 385*( 527$0(7(5

2. Armored purge rotameter. 1(('/( 9$/9( $1' ',))(5(17,$/ 35(6685( 5(*8/$725 9$/9( &$1 %( 5(3/$&(' %< 5(675,&7,21 25,),&( $7 6$&5,),&( 2) $%,/,7< 72 9$5< )/2: $1' &/($5 %/2&.$*( %< 23(1,1* 9$/9(

3. Needle valve and differential pressure regulator. Valve can be replaced by restriction orifice at sacrifice of ability to vary flow and clear blockage by opening valve. %<3$66 )/86+ 9$/9( ,) 1(('(' 72 &/($5 7$36 2) 3/8*6

4. By-pass flush valve if needed to clear taps of plugs.

Keep Out Dirt. Dirt must be kept out of the small openings in regulators, needle valves, rotameters, and ROs. Even with clean purge sources, dust and dirt can be picked up in the distribution piping. Both systems in Figure 1500-14 and Figure 1500-15 use strainers to trap dirt at the inlet to the configuration (see Section 1550, “Winterizing—Heating and Insulating”). Bypass Flush. Some applications will have a problem with slow buildup of plugging material in spite of the purge. When purge is lost or temporarily interrupted, as for servicing, some process fluid will back up into the taps and leads. In some applications this can cause varying degrees of plugging. In these cases, a means must be available to manually flush at high flow rates to clear the leads and the process taps. The bypass valve shown in both Figure 1500-14 and Figure 1500-15 does this. As a rule of thumb, the bypass flow should be 10 to 30 times the purge rate. Of course, this high rate will likely upset the instrument reading, so operators must take appropriate steps any time the bypass is used. High Pressure Services. Both pressure regulators and rotameters have upper pressure limits. When no regulators or rotameters are available to meet the pressure requirements, use variations of the configurations discussed. For instance, change

July 1999

1500-44

Chevron Corporation



the Figure 1500-14 configuration to eliminate the regulator and replace the RO with a calibrated needle valve that will allow on-line adjustment of the purge rate. Lead Filling. Any purged installation must include provisions for completely filling the instrument body and the leads between the instrument and the purge injection point. The means to do this depends upon the details of the configuration of the leads and the instrument body. In designing systems, consider the need for high point vents, low point drains, instrument body vents and drains, and filling connections at low points or at the instrument body. Back flow of the purge fluid can usually be used for filling. Remember, filling must be possible as part of commissioning and startup. Purge Injection Location. The purge fluid generally is injected into the instrument lead at either of two locations—close to the process tap or close to (or through) the instrument body. At the Process Tap. A point for injection that is close to the process tap is probably the most widely used location for nonflowing leads. The injection point is immediately outboard of the root valve. The purge fluid sweeps a minimum length of the lead and causes a minimum effect on the instrument reading. This location also keeps winterizing to a minimum where the purged section of the lead must be winterized, as with high pour point and high viscosity process fluids. Usually the purged section of the lead is short enough that the main line or vessel tracer can be used for winterizing. Most of the instrument lead remains stagnant, containing the original fill of purge fluid so that it is not affected by changes in purge fluid composition or temperature. It will not fill with process fluid if there is a flow reversal, so any resulting plugging is restricted to the process connection area and is easier to flush clear. On the other hand, if process fluid does enter the leads outboard of the purge connection, it will not be swept clear by the normal purge flow so the leads may plug and have to be manually cleared by maintenance personnel. Thus, this is a poor location for any application where flow reversal in the leads can be expected or is probable. This is the case on all flowing lead installations because level changes are likely to cause flow reversal on external chamber level instruments. Flow reversal is also probable on nonflowing leads in gas or vapor service where pressure surges or pulsations are expected. The compressibility of the gas or vapor may allow process fluid to enter the leads. At the Instrument. These applications benefit from injecting the purge immediately inboard of the instrument or its manifold valve. Sometimes it may even be advisable to purge through the instrument body. With these arrangements, the purge fluid will keep the maximum amount of the instrument lead system continuously swept clear. This method is commonly used on flowing lead instruments and on nonflowing leads in compressible fluid installations where back flow or migration of the process gas or vapor is likely. Differential pressure measurement across reactor beds and fractionating columns are common examples. While using this location helps to overcome the problems associated with back flow, it has all of the disadvantages associated with long lengths of purged leads. Greater

Chevron Corporation

1500-45

July 1999



instrument errors are introduced by the purge flow. Where it is necessary to winterize the purged leads, more winterizing will be needed and it will be necessary to use an independent tracer rather than the main line or vessel tracer. Rod Out Connection. In some purged applications, solids may slowly form or accumulate at the process tap and build up to form a hard plug. Coking hydrocarbons are particularly prone to this problem. Some salt formers and fouling fluids will also do this. In these applications a connection must be provided at the root valve so that the process connection can readily and safely be reamed out. This is called a “rod out tee,” usually a tee with the instrument lead connected to the branch, the run of the tee in line with the root valve and the process tap, and the outboard end of the tee plugged. Sometimes the outboard end of the tee is fitted with a plugged gate valve. The vessel or line connections on Figure 1500-12 illustrate this. Usual Company practice is to use a plugged tee. The outboard valve is used where fouling is expected to be so bad and process pressures are so high that the root valve cannot be relied on to shut off tight enough to safely install the rod out equipment.

1544 Purging Fluids The selection of purging fluids and their source of supply is critical to successful purging systems. The most important factors to consider are reviewed here.

Fluid Considerations Here is a summary of the major characteristics of successful purge fluids.

July 1999

•

Compatibility: The fluid must be compatible with the process, the instrument, and the instrument leads. It must not react with the process fluid to form vapor or solids. Its presence in the process fluid must not adversely affect downstream equipment or process operation. It must not be a contaminant that will degrade downstream product quality.

•

Cleanliness: The fluid must be dirt free and the distribution system designed and installed to keep it that way.

•

Density: Purge liquids both heavier and lighter than the process fluid are used. Generally a heavier purge liquid is preferred because it is easier to fill leads and there is less risk of losing the liquid from the leads when the process line or vessel is drained or runs dry.

•

Volatility: The purge liquid must not vaporize when it contacts the process liquid and it must not vaporize in hot weather. Either situation will cause erratic instrument performance.

•

Ambient temperature: Ideally the fluid should be unaffected by either low or high ambient temperatures: no vaporization on the hottest days and no problems on the coldest days with viscosity, pour point, freezing, or condensation. Exceptions are those cases where water must be used or where only wet hydro-

1500-46

Chevron Corporation



carbon liquids or condensible vapors are available. For these, winterizing will be required in freezing climates. •

Dry: Ideally purge fluids should be dry, containing neither liquid water nor water vapor that will condense. Dryness eliminates ambient temperature and dew point problems.

•

Dew point: When the purge is a gas or vapor it must stay above the dew point at all times. Condensation causes erratic operation. Condensed water can collect, freeze, and stop the purge flow.

•

Personnel hazards: Since the purge fluid has to be handled by maintenance personnel, it should not be toxic or otherwise hazardous.

•

Variability: The quality of the purge fluid should remain constant. However, since some purge fluids come from intermediate streams manufactured on site, they may be subject to variations in composition, density, pour point, etc. This variation is acceptable so long as it does not cause critical properties to fall outside acceptable limits (e.g., pour point or dew point cannot be allowed to go too high).

Source of Supply The supply of purge fluid must be reliable and the fluid must be available when needed. The source of fluid must not be cut off, exhausted, or otherwise interrupted while the plant is in operation. Instrument purging begins early in startup before the flow of troublesome process fluid starts, must continue during upsets in plant operation, and continues during shutdown until the troublesome process fluids have been removed from the lines and equipment. Purge fluid must be supplied continuously. These requirements are easily met where the purge is a readily available liquid supplied from day tanks dedicated to the purge system. The requirements are rarely a problem where the purge comes from highly reliable plant-wide utility systems such as utility air, plant water, condensate, steam, natural gas, hydrogen, and nitrogen. Plant-wide flushing systems are more likely to be a problem. When considering them as a supply for purge fluid, be sure to check their reliability, availability, and variability, especially during startup, plant upset, and shutdown. The real potential for problems comes with the use of streams that are manufactured within the plant using the purge. Be very careful about using such sources. They frequently will not be available or the quality will be unacceptable during startup, upset, or shutdown.

1545 Need for Additional Winterizing As stated in API RP 551, purge systems do not always eliminate the need for some winterizing. The reason for adding winterizing to purged installations is that ambient temperatures may fall low enough to cause trouble with the process fluid

Chevron Corporation

1500-47

July 1999



when the purge is interrupted or with the purge fluid itself. The troubles come from freezing water, high pour points, excessive viscosity, condensation, and hydrates.

Protect for Purge Failure Winterizing is added to protect against interruption of purge fluid flow where the ambient temperature drops below the process fluid’s freezing point, pour point, point of maximum allowable viscosity, or dew point. For nonflowing lead applications, winterize the part of the lead between the process tap and the point of purge injection. Include the isolation block valve. For flowing lead applications, winterize both the lead and the external cage or chamber of the instrument.

Protect Purge Fluid The character or composition of the purge fluid itself may cause problems at low ambient temperatures if winterizing is not added.

Water The low viscosity, low pour point fluids commonly used for purge may contain some water. This water can settle out and freeze in the instrument, its leads, or the purge supply and delivery system. The amount of purging being done will seldom justify the expensive drying system that would be needed to eliminate the water. In this case, winterize the purge supply system for water freeze protection. If the instrument leads cannot be made fully self-draining to the process connection, they also must be winterized for water freeze protection.

Pour Point and High Viscosity Occasionally, particularly in more severe climates, minimum ambient temperature will be below the pour point or the point of maximum acceptable viscosity of the purge liquid. In such cases, winterize all of the purge supply and delivery system and all of the instrument sensing leads. The stagnant part of the lead must be included since it will fill with purge liquid.

Condensation In applications using gas or vapor purge where the ambient temperature will fall below the dew point, use winterizing to keep the purge above the dew point. Also winterize all of the instrument lead that will not freely self-drain into the process. If hydrates will form, even the self-draining portion must be winterized.

Type of Winterizing See Section 1550 on winterizing for a discussion of various winterizing techniques. Heat tracing will be required in most cases where winterizing is added to purged installations. However, when winterizing is being added to flowing lead applications or to the purged part of nonflowing lead applications, consideration should be given to using insulation alone and without heat tracing for the following: (1) where the climate is mild; (2) when the ambient temperature does not drop far below the temperature at which trouble will occur and stays there for only a short time (say only overnight); or (3) where occasional short-term interruption of the instrument function will not cause any severe problems. Insulation alone may also be consid-

July 1999

1500-48

Chevron Corporation



ered for winterizing the purge supply and delivery system when these same criteria are met and the purge liquid flow rates are high enough to keep the line warm.

1550 Winterizing—Heating and Insulating 1551 General Information Winterizing is performed to prevent the disabling of instrument systems by the formation of solids, by increasing the viscosity of liquids, and by condensation of vapors; all of these occur at low ambient temperatures. In many cases these temperatures will be well above the freezing point of water. The formation of hydrates and the congealing of high pour point oils are typical examples of this condition. The purpose of winterizing is to keep the temperature of the fluid high enough to avoid these unwanted changes. Such winterizing normally includes some form of heating, usually heat tracing. However, with very mild temperatures, insulation alone may be sufficient. The temperature behavior of the process fluid and the characteristics of the local climate are what determine the need for winterizing and the method to use. Frequently, winterizing is combined with seals or purges to give the most effective protection against the disabling effects of low ambient temperatures.

1552 Ambient Temperature Considerations Winterize instrument systems when the temperature can be expected to be cold enough for a long enough time to cause disabling changes in the fluid in instruments or their process connections. This expectation should be based on the coldest recorded temperature, not some nominal figure such as a “normal” or “average” minimum temperature. This criterion is applied to all nonflowing leads and to most nominally flowing leads, such as on level glasses, which can be relatively stagnant for long periods of time. The need for and the extent of winterizing depend on the duration and temperature of cold spells, wind speed, precipitation, the type of fluid involved, initial temperatures, geometry, and insulation material. Few of these factors will be well defined. Calculations need to be supplemented by experience and engineering judgement. Although winterizing generally requires the use of heat tracing, sometimes insulation alone can be enough. Consider insulation alone in very mild climates where the instrument you want to protect meets one of the following criteria: •

Chevron Corporation

The instrument is a flowing lead, such as on a level glass, an analyzer sample line, or a purge fluid supply line; the lowest expected temperature is no more than 10°F below the “change” temperature; and that it will not stay at the lowest expected temperature below the “change” temperature more than 48 hours.

1500-49

July 1999


•


The instrument is a nonflowing lead; the lowest expected temperature is no more than 10°F below the “change” temperature; the instrument will stay below the lowest expected temperature only overnight; and the instrument will warm up well above this temperature (say 10°F) during the day.

Keep in mind that these are only very general rules of thumb and must be used with good engineering judgement. Sometimes winterizing, in the form of heat tracing, is required at rather high ambient temperatures. High pour point heavy oils may require temperatures to be maintained well above 100°F, hydrates can form well above freezing at higher pressures, and condensation can occur at higher temperatures.

1553 Fluid Considerations Water The most obvious need is to protect against freezing water. The water can come from condensing steam and wet hydrocarbons as well as from free water streams.

Wet Hydrocarbon Streams With water-contaminated fluids, water will settle out in any low points and plug the line when it freezes. For such lines the need for winterizing depends on the pour point, the point of maximum tolerable viscosity, or the freezing point of water, whichever is higher. Where hydrate formers such as light hydrocarbons or hydrogen sulfide are present, the formation of hydrates must be considered.

High Pour and High Viscosity Hydrocarbons Heat tracing and winterizing are almost always required when process leads contain high pour point or high viscosity oils. These are the fluids that require the highest temperatures to maintain adequate fluidity. Heat tracing is likely to be required even in the warmest climates.

Inorganic Chemicals Inorganic reagents, such as caustic solutions and acids, commonly need winterizing because many locations have minimum temperatures below the point at which these solutions solidify. These also are the process fluids most likely to cause damage from overheating. Sodium hydroxide (caustic) and phosphoric acid are typical examples.

Condensing Vapors Heat tracing can be used to prevent vapors from condensing in instrument leads where they could be cooled below the dew point by ambient temperature conditions. Furnace stack gas sample leads to an analyzer are a common example. This technique can also be used to prevent “refluxing” in dry self-draining instrument leads. In cold weather, the condensing rate may be great enough to fill all or a

July 1999

1500-50

Chevron Corporation



portion of the lead and give false or erratic readings due to the hydrostatic effect of the condensate.

Overheating Along with the benefit of maintaining adequate temperatures in cold weather, heat tracing carries the risk of producing temperatures that are too high when ambient temperatures are at their maximum. This overheating can cause a variety of problems. Boiling, the most common problem, will degrade instrument performance. Water is the most common problem fluid, with hydrocarbons in the “gasoline and lighter” range running a close second. The high temperatures can also cause chemical or metallurgical attack on the instrument leads and wetted parts. Stress cracking from hot caustic and accelerated corrosion from some acids are typical. Some sensitive organics may be subject to thermal decomposition or polymerization, which can shift instrument readings by changing the density of the fluid in the leads or can plug the leads with the decomposition products or the polymer. Finally, the materials used in the tracing hardware itself may be damaged by the high temperatures. The plastics used in self-limiting tracer cable are the most common example.

Temperature Controls Temperature controls can be used to keep the process leads from overheating or to turn on the heat tracing only when the ambient temperature falls low enough to cause trouble (e.g., 35°F for water, 75°F for 50% caustic, etc). This subject is covered in more detail in Section 1558, “Electric Tracing.”

1554 General Design Principles Eliminate or Minimize Winterizing Winterizing is undesirable. It is costly, difficult to install properly, and expensive to maintain. Heat tracing is prone to failures which can disable instruments and cause serious process upset. The ideal solution is to use measurement methods that do not need winterizing. Failing that, design to minimize the amount of winterizing that is needed. Use self-draining leads. Line-mount transmitters with short leads to the process taps. Use seals to keep the process fluid out of the instrument leads. Design purges to minimize the amount of heat tracing needed.

Self-drain the Water Eliminate low points where water can collect. Make the leads self-draining from the instrument to the process connection. Close-coupled line-mounted installation of transmitters makes this practical. Where water-contaminated hydrocarbons are involved, this is frequently all that is needed.

Chevron Corporation

1500-51

July 1999



Most liquid hydrocarbon services will have water in them at some time. It may accumulate if the system is not properly designed. The water can come from many sources. If the stream is not normally contaminated with water from storage, water washing, or steam stripping, water is likely to be introduced by or left over from water washing lines, steamout, or upset operation.

Use Seals Use seals mounted self-draining on short process connections to keep the potentially troublesome process liquids out of the instrument systems. Diaphragm seals are preferred for this but liquid seals can be used. Often this use of seals will eliminate the need for any heat tracing. Where it does not, at least it will reduce tracing to just the portion between the seal and the process tap.

Purge Design In purged installations, it is often advisable to use heat tracing between the purge connection and the process tap. Locate the purge connection close to the process tap. This will minimize the tracing needed and reduce the hazards from tracer failure. This method is commonly used when purging heavy, often dirty, viscous or high pour point stocks.

Winterize Use heat tracing and insulation to keep leads and instruments warm only when other protective methods are insufficient, impractical, unacceptable, or otherwise cannot be used. In designing tracing for leads and instruments, be alert to the hazards of overheating, which can cause damage to the leads, the instrument, and the tracer material. Trace transmitter topworks only when the coldest ambient temperature is below the manufacturers’ recommended minimum. See Figure 1500-16 for an example of an exposed topworks.

1555 Instrument Housings A variety of instrument housings are described and discussed in API RP 551. This section is supplementary to that material.

Prefabricated Housings Prefabricated insulating housings of rigid molded plastic or flexible fabric covered insulation designed and fabricated specifically for transmitters and other field instruments are available from a variety of manufacturers. They come with various types of steam or electric heating systems for freeze protection and for higher temperatures. The prefabricated finned heaters are usually preferred. Prefabricated enclosures and heating systems take up less space, require less design time, cost less, require less labor to install, are more reliable, and are cheaper and easier to maintain than other constructed housings. They are to be used in preference to sheet metal housings or field-applied insulation except where prefabricated housings are either not available or not practical.

July 1999

1500-52

Chevron Corporation



Fig. 1500-16 Transmitter Housing with Topworks Outside (Type P, Type P/L) (From Drawing DG-J99744-5) Transmitter Topworks

Heat Pack Housing with Heat Bolt in Transmitter

Insulate

Steam Supply Glass Tape Wrap for Light Tracing Only

To Trap

In selecting and using these prefabricated housings, always follow the manufacturer’s recommendations, particularly with regard to the amount of heat required to maintain the required inside temperature for the minimum expected ambient temperature.

Transmitter Enclosures Housings for transmitters are available that heat only the transmitter body and not the topworks. See Figure 1500-16. This protects the electronic (or pneumatic) components from overheating. Use this type except in very severe climates where the expected minimum temperature is too low for the topworks.

Transmitter Valve Manifolds Another option for transmitter housings is the three-valve or five-valve manifolds typically used on differential pressure flow transmitters. Either the manifold can be included inside the housing (as shown in Figure 1500-16) or the housing can be purchased without room for the manifold so that it must be insulated separately. The preferred arrangement is to bolt the manifold directly to the transmitter and include it inside the housing. This makes heating the manifold very simple.

Chevron Corporation

1500-53

July 1999



However, to avoid opening the housing to use the valves, extensions must be added to the valve stems to make the handles accessible outside the housing. This introduces a design problem. Be very careful in the selection of these extensions. Some are poorly designed and will easily slip on the stem so that the valve will not move or cannot be shut off tightly. A simple set screw is not adequate. Some of the flexible housings solve this problem by using a hatch or flap which can be opened to get to the valve handles. The hazard here is that the hatch may not be closed properly after use.

Whole Instrument Enclosures Large, rigid molded plastic enclosures with hinged doors are also available. See Figure 1500-17. Similar flexible housings are made. In place of the doors, these use large flaps which are held closed with Velcro fasteners and straps. Either electric or steam finned heaters can be used for heating the air inside the enclosure. These heaters are available from the housing manufacturers. Since these large housings enclose the whole instrument, care must be taken to avoid overheating. They should be used when the expected minimum ambient temperature is lower than the minimum temperature of the transmitter topworks or when the whole instrument must be enclosed, as is true for a field controller that has the process wetted measuring element in the same box as the controller. Fig. 1500-17 Typical Large Hinged Enclosures BACK ACCESS

WINDOW

Heating Transmitter Bodies Several methods are in use for heating transmitter bodies in prefabricated housings.

July 1999

1500-54

Chevron Corporation



Heat Bolts. Heat bolts are widely used in steam-traced applications. One or more of the transmitter body bolts or studs is replaced by a hollow stud which carries the tracing steam and condensate. This conducts the heat directly to the transmitter body. When considering this method, keep in mind that drilling the bolt weakens it and may reduce the transmitter body pressure rating below the maximum working pressure of the application. Consult the instrument manufacturer for the effect of using heat bolts. Heat bolts can be obtained from the housing manufacturer. Be sure to follow the housing manufacturer’s recommendations when using heat bolts. Heat bolts are effective and compact and they do simplify the installation. However, they complicate maintenance because they must be disconnected whenever the instrument is removed for servicing. Heat Blocks. Heat blocks or heating saddles are another option. These are held next to the instrument body inside the housing. They heat primarily by radiation and secondarily by heating the air in the housing. These also are available from the housing manufacturer, whose recommendations should be followed when using them. Heat blocks are available in both steam and electrically heated versions. Although they are bulky and somewhat more difficult to install, heat blocks do simplify servicing because they are not connected to the instrument and do not have to be disconnected from the tracer when the instrument is removed. Wraparound Heating. Finally, the tracer can simply be wrapped around the body of the instrument or other device. The wrapping must be done in a pattern that will not trap the device such that it cannot be removed without breaking the tracer. This is especially important in electric tracing. The wraparound technique is illustrated for a three-valve manifold in Figure 1500-18 and in some of the details in the electric tracing Standard Drawing GD-J1221. Housing and tracer manufacturers can give guidance on the amount of tracing required for specific applications. Fig. 1500-18 Typical Wraparound Tracing

Chevron Corporation

1500-55

July 1999



1556 Heat Tracing API RP 551 covers the subject of heat tracing very well, discussing both steam and electric tracing. The following discussion supplements that material.

Electric Tracing Both electric and steam tracing methods can be used on instruments. However, electric tracing is generally the method of choice. Electric tracing should be used where temperatures must be controlled with some accuracy or must be limited to a lower value than can be achieved with steam tracing to prevent decomposition of process fluid or chemical attack of instrument leads. Electric tracing is especially attractive where conditions call for light tracing with steam (described later). Electric tracing generally costs less than steam tracing for the short lengths used for instruments (less than 100 feet). For short runs, most of the cost is in the branch circuit for electric tracing and in the steam supply and condensate piping and trap manifold for steam tracing. Within a process unit, the branch circuits are normally less expensive than the steam and condensate piping. Electric tracing is generally cheaper and more practical than steam for water freeze protection and for prevention of hydrate formation. Ambient temperature thermostats can be used to turn on and off groups of tracers all having the same protection temperature. Energy costs for electric tracers tend to be lower than for steam for lower temperatures such as required for freeze and hydrate formation protection. When ambient temperature approaches or exceeds the required protection temperature, steam consumption continues and greatly overheats the system, while electrical tracing either shuts off (ambient thermostat) or greatly cuts back on consumption (selflimiting cable).

Off-Plot and Remote Locations Electric tracing is generally used for off-plot and other remote locations because electricity is usually more readily available than steam.

Insulation Heat tracing systems will not work properly if the wrong insulation is installed or if the insulation is installed poorly. A heat tracing system is designed by using the heat losses for a particular insulation system, including the housing. Care must be taken to be sure these requirements are made clear to the installers and that the correct materials are installed. Substitutions may cause failure because heat losses may exceed the tracer’s ability to add heat. Less likely is overheating from too much insulation. Proper installation is essential. Drawings, specifications and field inspection must be clear on this point. The entire tracing system must be carefully insulated and waterproofed. Particular attention must be given to the junctions of the instrument lead with the instrument (entry to the enclosure) and of the instrument lead with the point of measurement (at the root valve) because these are places where the insula-

July 1999

1500-56

Chevron Corporation



tion is frequently missed or removed and not replaced. The insulation must have no gaps (for example, at elbows or bends). Instrument leads must have no exposed drain valves. A durable weathercoat sheath, preferably made of metal or PVC, should be installed for protection. The system should be designed and installed to minimize the risk of damage to the insulation during routine service or maintenance. For steam tracing in freezing climates, insulation must be continuous from the steam source, through the instrument tracing, through the condensate return lead, and back to the steam trap manifold. At the steam trap(s), the traps, strainers, and valves must be insulated in a manner that permits easy removal and replacement because they require frequent checking and maintenance. Condensate return headers must be insulated. Removable blanket insulation should be used with caution. It has major shortcomings. Once removed, the blanket may not be reinstalled properly. Blankets are frequently hard to re-fit and difficult to keep firmly in place with no gaps at the edges. The result is unplanned-for heat loss and consequent failure of the measurement system in severe weather. The blankets must remain weatherproof. Some types of blanket covering materials are easily torn or have unsealed stitching penetrations so that they absorb and hold water. The water both reduces the insulation effectiveness and acts as a “sink” for the heat intended for the process leads. Again the result is loss of effective protection and failure of the measurement system.

Temperature Limitations When designing heat tracing systems, take care not to exceed the maximum temperature limits of the process fluid, the tracing material (of particular concern in electric tracing), and the instrument internals (especially important with electronic instruments). These concerns are covered in more detail below, in other parts of this section, and in API RP 551.

Light Tracing vs. Heavy Tracing As discussed in API RP 551, light tracing is used in steam tracing applications to avoid overheating sensitive process fluids. Electric tracing using self-limiting cable is the preferred way of handling heat sensitive fluids. These concerns are discussed more thoroughly in Section 1557, “Steam Tracing.” The idea behind light tracing is to avoid direct contact between the steam tracer and the instrument lead. Such contact causes overheated hot spots. The technique is to wrap the lead with insulating tape to hold off the tracer. Figure 1500-19 illustrates this. For Company practice in implementing light tracing, see the design and construction notes D-2 and C-8 on Standard Drawings GC-J1217 and GC-J1218. When steam tracing high pour point and high viscosity stocks where high temperatures must be maintained, good heat transfer is mandatory; therefore, direct contact between tracer and line is desirable. The insulating tape is omitted. This is referred to as heavy tracing. It also is illustrated in Figure 1500-19. The same direct-contact method is used with electric tracing where steam is not available and heavy tracing is indicated. The required temperature typically will be above the capabilities of self-limiting cable. In these cases constant heat density

Chevron Corporation

1500-57

July 1999



Fig. 1500-19 Light and Heavy Steam Tracing Concepts

cable is used for electric tracing. Constant heat density cable brings with it the risk of overheating during periods of higher ambient temperatures. This hazard can be minimized by controlling the electric power to the cable. Use thermostats installed to sense the temperature of the traced process leads. Successful installations require very careful design and installation. Consult electric tracer equipment manufacturers for guidance. Constant density cable and the use of thermostats are discussed in Section 1558, “Electric Tracing.”

Calculations for Heat Tracing It is not feasible to provide tables for all combinations of minimum ambient temperature, holding temperature, insulation materials, and insulation thickness and piping and tubing diameter for all steam pressures and electric heating cables. Therefore, it is generally necessary to calculate heat tracing requirements. Help is available. Manufacturers of winterizing equipment have published helpful tables, charts and formulas. They have computer programs and will either supply these programs to users or will do the calculations for them.

July 1999

1500-58

Chevron Corporation



As part of their service when supplying equipment, most electric tracing suppliers will provide complete designs, including calculations, material selection, material lists, installation designs, and documentation. To do this they must be given all required information, such as ambient temperatures and weather conditions, process and operating requirements, insulation requirements, and electrical information.

1557 Steam Tracing This section supplements and should be used in conjunction with the discussion of steam tracing presented in API RP 551. Figures 1500-3, 1500-4, 1500-11, and 1500-16 all illustrate steam traced instruments.

Temperature Considerations—Light vs. Heavy Tracing The maximum temperature limitations of the process fluid are especially important with steam tracing because the tracing steam is usually quite hot. The problems center around vaporization and chemical attack. The preferred solution is to use self-limiting electric tracing. However, where steam tracing is necessary, light tracing is used instead. Light tracing is described above in Section 1556, “Heat Tracing,” and in API RP 551. The principle is illustrated in Figure 1500-19.

Avoid Boiling When the boiling temperature of the process fluid is lower than the temperature of the steam, care must be used to prevent boiling in the leads. Such boiling will cause wide swings in the reading of the instrument and may cause a permanent error in the reading if all of the liquid is boiled off and none recondenses to replace it. Light tracing is used to deal with this circumstance.

Avoid Chemical Damage Light tracing is also called for when the steam temperature is hot enough to cause chemical or physical attack by the process fluid, especially at the hot spots where the tracer contacts the instrument lead. Phosphoric acid and caustic solutions are two cases in point. There are also other conditions where steam temperature will be high enough to cause thermal breakdown of stagnant process fluids or instrument fill fluids. Check for these problems and if they are indicated, light tracing may be a solution.

Limitations of Light Tracing Light tracing will not prevent overheating problems if the air inside the insulation that is covering both the lead and the tracer is heated above the boiling or other “damage” point of the fluid. As a rule of thumb, the steam temperature must not be more than 50°F above the boiling or “damage” point of the fluid except in specially engineered situations where careful calculations have clearly shown that boiling or damage will not result. This 50°F rule of thumb is for a 3/8 inch O.D. maximum

Chevron Corporation

1500-59

July 1999



size tracer. A pressure reducing valve may be used to lower the steam temperature. Care must be taken to assure that the steam will be down to its saturation temperature before entering the traced portion, and that the required pressure is still high enough to get the steam through the steam trap and into the condensate collection system.

Self-draining Tracers Ideally, each instrument tracer should be entirely self-draining, as discussed in API RP 551 and illustrated in Figure 1500-20. The steam source should be at the highest elevation, the steam trap at the lowest, and the steam tracer continuously sloping downward between the two. This arrangement, however, is not always possible. When pockets are unavoidable, follow the recommendations of API RP 551 as shown at the top of Figure 1500-20. In freezing climates, if the tracer is not self-draining, pockets of condensate can freeze in the tracer when the steam is shut off during maintenance of the instrument. In this situation, do one of the following: •

Provide an air purge connection just downstream of the steam supply shutoff valve for temporarily connecting a plant air hose to blow condensate out of the tracer to grade (this valve replaces the vent valve).

•

Provide high point vents and low point drains for draining condensate from the tracer.

The choice between these two depends on local maintenance practices.

Isolation Valves, Drains, and Blowdown Each instrument should be individually traced with its own shutoff and vent valves at the steam supply source and its own steam trap with condensate isolating valve. Provide a drain or blowdown valve immediately ahead of the steam trap so the tracer can be emptied to grade. This assures proper operation and protection for each instrument. It allows the tracer to be shutdown, drained, and cooled off in preparation for servicing the instrument without affecting the operation of other instruments. Locate the supply valves and the trap manifolds so they are easily accessible for maintenance.

Tracer Materials Use type K copper tubing made up with brass fittings for steam pressures below 175 psig and line temperatures below 400°F. Use only stainless steel tubing and fittings for higher pressures and temperatures.

Steam Pressure Considerations Select the steam pressure that will maintain the required lead temperature at the minimum expected ambient temperature. The steam will be at its saturation temperature. Check to see if light tracing is needed to avoid overheating at high ambient

July 1999

1500-60

Chevron Corporation



Fig. 1500-20 Steam Tracer Self-Draining Requirements (Adapted from API RP 550, I, 8, Figure 8-11, 1985, Courtesy of American Petroleum Institute)

KEEP TRACING FREE OF POCKETS, IF POSSIBLE. IF POCKETS EXIST, LIMIT TOTAL SUM OF HEIGHT OF POCKETS TO A + B + ......

= 10%

2.3(10.2 in SI)

OF HEATING STEAM HEADER PRESSURE, PSIG (KPa).

HEATING STEAM HEADER

A

B

FEET

FEET

(mm)

(mm)

INSULATION SEE FIG 8-10 FOR DETAILS

TRACER

INSULATED HOUSING

1/4" OR 3/8" TUBING USE RECOMMENDED FINNED HEATER OR HEATING COIL MOUNTED ON BOTTOM OR SIDE OF HOUSING. ANCHOR AND SHIELD TO PREVENT ACCIDENTAL CONTACT AND BURNS.

COIL MAY BE MADE IN

FIELD FROM COPPER TUBING. LARGER TUBING.

USE 3/8" OR

RADIATING SURFACE REQUIRED

VARIES WITH CLIMATE, SIZE OF HOUSING, HOUSING INSULATION AND STEAM PRESSURE.

STEAM TRAP OR USE LIQUID EXPANSION THERMOSTATIC VALVE FOR CONTROL OF HOUSING TEMPERATURE TO PROTECT CONDENSATE

INSTRUMENT FROM OVERHEATING.

DISPOSAL

IN

FINNED HEATING COIL RECOMMENDED FOR HOUSING TYPICAL METHOD OF STEAM TRACING

OUT

AND HOUSING FOR FIELD-MOUNTED PRESSURE INSTRUMENT FULLY SELFDRAINING TRACER.

temperatures. Consider electric tracing if a steam pressure low enough to assure no overheating is not practical. Low pressure tracer systems are more likely to freeze in cold weather. For ambient temperature at or below -20°F, at least 100 psig is recommended. Using 15 psig steam is not recommended. In addition to the danger that 15 psig will not prevent freezing, most plants do not have a collection system for 15 psig condensate so the condensate must be wasted to atmosphere with the accompanying risk of ice formation.

Chevron Corporation

1500-61

July 1999



Pressure at the trap, usually at grade, must be high enough to get the steam into the condensate collection system including the lift into elevated pipeways.

Identification of Tracers Use a nameplate to tag each instrument tracer at both the steam supply valve and at the steam trap. This is essential because these locations usually are not close enough to the instrument for a clear identification, and several supply valves or steam traps are likely to be grouped together. The nameplates should carry the tag number of the traced instrument, like this: 5-FT-1256 STM TRACER The nameplate should be durable and still legible when old and dirty. Use deeply engraved plastic with black letters on a white background or deeply engraved stainless steel. Tracers that are seasonal (used only in winter) should be identified by color-coded paint or special signs.

Steam Traps A thorough discussion of the design of steam traps and their manifolds is beyond the scope of this guideline. Only factors that apply to instrument steam tracing are mentioned here. See Figure 1500-21 for a sample steam trap manifold. This figure illustrates the following: the blowdown or drain valve ahead of the trap used for emptying the tracer for instrument maintenance; the block valve after the trap for isolation from the condensate system for both trap and instrument maintenance; and the strainer ahead of the trap to prevent scale and dirt from disabling the trap. Traps discharge into a condensate collection system. This requires a differential pressure across the trap. Always check to be sure the trap will work with the planned steam and condensate system pressures. Other factors to consider are the trap’s limitations on minimum and maximum pressure and temperature, mounting position for freeze protection, freeze protection provided on trap failure, ability to withstand surges, and maximum allowable hydrostatic test pressure. Traps with quick change trim and integral strainers are generally less expensive to install and less expensive to maintain so that the total evaluated cost is lower. They should be used wherever they suit the application and are otherwise acceptable. Sizing traps for instrument tracing service is generally not very difficult. Since the condensate loads are so small, one size usually will serve all applications. Help is available from trap manufacturers to complete the sizing that is required. Local standardization may determine the brand and type of trap to use on instrument tracing.

1558 Electric Tracing This section supplements the material on electric tracing in API RP 551, particularly with regard to the use of self-limiting cable for instrument tracing.

July 1999

1500-62

Chevron Corporation



Fig. 1500-21 Typical Steam Trap Manifold for Instrument Tracer (From Company Form EF-302-3)

Figures 1500-22 and 1500-23 illustrate some electrically traced installations. Standard Drawings GD-J1221 and GD-J1222 present Company practice and illustrate the installation of electric tracing and various electric tracing accessories. Electric tracing is the method of choice for instrumentation, and self-limiting heating cable is the preferred material to use for heating. This method is particularly well suited for applications involving temperature-sensitive process fluids where light tracing is called for. Heating cable can be selected that will self-limit below the “critical” temperature (boiling point, decomposition temperature, accelerated corrosion temperature, etc.). Heating cable is also well suited for protection against low ambient temperatures that occur infrequently or only a few weeks or months during the year. Water freeze protection and prevention of hydrate formation are typical applications because ambient temperature thermostats can be used to turn tracers on when they are needed.

Using Self-limiting Cable The construction and operating principle of this cable are described in the section entitled “Heating Cable” below. Self-limiting cable is the practical and cost effective way of electrically tracing instrumentation. It can easily be cut to length in the field. It operates on standard voltages. It does not require heat transfer cement. It is flexible enough to be spiraled around leads and wrapped around valves, valve manifolds, instrument bodies, and any other place where extra heat is needed. It generally does not require a thermo-

Chevron Corporation

1500-63

July 1999



Fig. 1500-22 Electrical Tracing: General Concept (Adapted from API RP 550, 1, 8, Figure 8-17, 1985, Courtesy of American Petroleum Institute)

COLD CONNECTING WIRING

RELAY

HEATING ELEMENT

POWER SOURCE, SWITCH AND END-OF-TRACER RUNNING LIGHT

THERMOSTAT-AMBIENT SHAPE AND ANCHOR HEATING ELEMENT

RELAY AND THERMOSTAT ARE NOT NEEDED FOR SOME APPLICATIONS

TO HEAT HOUSING USE TRACER AS SHOWN OR USE FINNED ELECTRIC HEATER

stat on instrument leads. It will not overheat and burn out when crossed over itself or when installed without contacting the pipe. Retrofitted field changes can be made easily when actual operating experience dictates. It can be easily monitored and checked to be sure it is in operating condition. Because self-limiting cable inherently limits the maximum temperature it can reach, the overheating and boiling problems that can occur with steam tracers can be avoided if the cable is properly selected and installed. Manufacturers make a variety of different types of self-limiting cables suitable for heat tracing instruments. They vary as to the maximum temperature they will maintain, the maximum temperature to which they can be exposed without permanent damage, the heat density (watts per foot) they can generate, and details of construction (heavy plastic outer sheath, metallic overbraid, etc.). Exercise great care in selecting cable to be sure it will be suitable for the application. In doing this do not forget upset and startup conditions, operation at maximum expected ambient temperature, and possible steamout or high temperature flush of the leads. Use manufacturers’ recommendations, charts, tables, and computer program(s), if available, when selecting type and amount of cable and designing installations. Take into account the insulation inside diameter, insulation thickness, insulation heat transfer factor, effect of wind velocity, and the thermal rating of the cable at the operating temperature of the cable. Installation drawings should be specific, showing the pitch (inches of pipe for each complete spiral wrap) on instrument leads

July 1999

1500-64

Chevron Corporation



Fig. 1500-23 Typical Electrical Tracing (From Drawing GD-J1221)

and any special wrapping on valves and instrument bodies that is needed to provide the required heat input. The recommended practice in determining the thermal rating of the cable is to assume that the cable temperature is at least 50°F above the process fluid temperature. This assumption provides a safety factor to allow for improper installation of the cable, such as poor contact with the lead, or for a poor thermal insulating job. An even higher temperature difference can result when oversize insulation is required and keeps the cable away from the pipe or tubing. Freezing of diaphragm seals has been a problem with self-limiting cable tracing. The cable should be wrapped more than once around the bottom and sides of the seal housing. The insulation should extend over the seal fluid end of the seal as well as the process end.

Temperature Considerations Always check the cable temperature to be sure that it does not exceed the manufacturer’s rating at the worst energized or de-energized conditions. Check for high

Chevron Corporation

1500-65

July 1999



temperature upsets, steamout, operation at high ambient temperatures, etc. If the temperature is too high, try to select a cable with a higher temperature rating. If that is not satisfactory (perhaps the higher temperature cable will self-limit above the boiling point), consider a different design, such as one that uses an ambient temperature thermostat, a high temperature shutoff, or a change to steam tracing.

Identification of Tracers Use a nameplate to identify each tracer at its disconnect switch on the power distribution panel for the tracer power. This is essential because these panels are usually not close enough to the instrument for a clear identification and several tracers will be supplied from the same panel. The nameplates should carry the tag number of the traced instrument and identify it as the tracer, like this: 5-FT-1256 ELECT TRACER The nameplate should be durable and still legible when old and dirty. Use deeply engraved plastic with black letters on a white background. Tracers that are seasonal (used only in winter) can be identified by color-coded paint or special signs.

Thermostats One of the attractions of electric tracing is the ease with which it can be controlled using thermostats. Line temperature thermostats are used with nonself-limiting tracer cable. They are seldom used with self-limiting cable because the cable inherently provides this function. Ambient temperature thermostats are frequently used with self-limiting cable. Sometimes a high temperature limit shutoff will be needed.

Ambient Temperature Thermostats Use these where protection is required against the effects of low atmospheric temperatures (water freezing, hydrates forming). Use on-off thermostats sensing ambient temperature to control groups of tracers in similar process fluid services. Obviously, this does not control temperature of the process lead or instrument body. Care must still be exercised to avoid overheating when the heater is on at ambient temperatures close to the thermostat setting. The recommended location for the thermostat or thermostats is at the distribution panel for the tracer(s) involved.

High Temperature Shutoff Though rarely needed, high temperature limit switches with the switch cemented to the heated line can be used to shut off the heater where there is serious danger from overheating if the primary control (probably an ambient temperature thermostat) fails to work properly.

Line Temperature Thermostats Line temperature thermostats are attached to the heated lead at the location where the highest temperature is expected. They control the temperature on applications using nonself-limiting cable, such as the constant heat density type discussed below.

July 1999

1500-66

Chevron Corporation



Monitoring Operation Monitoring devices should be provided to show that the heater is either operating or will operate when turned on. With the recommended self-limiting parallel cable this is relatively easy to do. Monitor the voltage between the bus wires at the remote end using a pilot light mounted in or at the distribution panel for the tracer. This pilot light will indicate whether the bus wires are intact for their entire length and whether the cable is energized. If the tracers involved are on ambient temperature control, the light will turn on only when low temperature has closed the thermostat contacts. To provide a means for checking the system at higher ambient temperatures, install a momentary contact switch that overrides the thermostat and turns on the tracers so that the pilot light will turn on.

Power Supply Considerations Electrically powered instrument tracers must be independent, protected from other power users, and arranged so that each instrument can be isolated. Provide a separate overload disconnect for each traced instrument. This configuration allows isolating individual instruments for maintenance and future modifications to the tracer. An electrical fault in one tracer will not interrupt the supply to the others. Proper operation of the tracer can be confirmed for each instrument individually. Instrument tracers should be grouped together and put on dedicated branch circuits with dedicated distribution panels. This arrangement minimizes the risk of inadvertent loss of instrument tracer power from activities and faults in other services and protects other services from activities and faults in the instrument tracer system. When ambient temperature thermostats are used, frequently several instrument tracers will have the same set point (e.g., water freeze protection). It is common practice to group several of these tracers together and put the group on a separate distribution panel. That way a single thermostat can control a load relay in the supply to the panel and a single push button can be used to monitor operation. The electrical supply system must accommodate the needs of the electrically traced instrument installations. Once the number of tracers, their locations, and the power demands have been established, the required electrical supply system can be defined. The electrical system must provide adequate current at the required voltage or voltages. It must provide the needed branch circuits, distribution panels, protective devices, and controls. The power supplying the instrument tracer should be isolated from other power users by being located on a separate main supply breaker or breakers. The circuit breakers and branch circuit wiring must be sized to accommodate the expected inrush current (which can be high) when the circuit is first energized and the tracer is at minimum ambient conditions. This precaution minimizes the risk of losing instrument tracing when faults occur in other services.

Heating Cable Of the various types of heating methods that are used for tracing lines and equipment, only the self-limiting parallel cable is really practical for instrumentation. Other methods are, however, briefly described below.

Chevron Corporation

1500-67

July 1999



Self-Limiting Parallel Cable. In this cable the resistance element is a blend of irradiated plastic material and graphite particles in the form of a ribbon between the bus wires. The bus wires are in continuous contact with the edges of this ribbon, and current flows across it. The entire assembly is insulated with plastic and has heavy plastic or metallic overbraid to provide mechanical protection. The resistance of the ribbon material increases rapidly with increasing temperature so that the cables are self-limiting as to maximum temperature. This feature has many advantages, among which are that it minimizes the probability of burnouts, eliminates the need for buried thermostats or high-limit temperature switches in most applications, and allows the use of a singe cable to trace complex arrangements where heat loss varies greatly at different points so that doubling back and overlapping the cable on itself is necessary. These cables are also flat enough for a single cable to fit under standard insulation. They give good heat transfer without the use of heat transfer cement and are flexible enough to be wrapped around valves and equipment. Consult manufacturers’ catalogs for more information on how the self-limiting parallel cable works, various optional types of construction including different types of protective armor, temperature limitations of the materials, the different temperature ranges that are available, many accessories and fittings that are available, and the suitability for use in electrically classified areas. Constant Heat Density Cable. This cable also is a parallel cable but it is not selflimiting with respect to temperature. This cable consists of two bus strips or wires with heating elements connected electrically in parallel between them. The typical construction is Nichrome wires wrapped around the insulated bus wires and connected to them at intervals of 1 foot or more. The heat density (watts per foot) at a fixed voltage is essentially independent of length and cable temperature. Thus, heat density is not affected when the cables are cut to length in the field to custom fit installations. However, because no heat is produced between the cut end and the last connection to the bus wires, cable lengths must be an integral multiple of the distance between connections to the bus wires. This is often difficult to do in the short runs needed for instrumentation. This characteristic, coupled with the fixed heat density and nonself-limiting behavior, make this type of cable undesirable for most instrumentation applications. It is not practical to customize cable and voltage for each instrument installation. However, there are some applications where constant heat density cable is the best choice, and sometimes the only choice. This type of cable can generate temperatures up to 1000°F. It has to be used where electric tracing is required and temperature or heat density requirements exceed the specifications of self-limiting cable. It can be used in place of steam tracing and will provide heavy tracing temperatures and heat input where steam is not available. Finally, constant heat density cable is suitable where temperature requirements exceed the temperature of available steam. Because it is not self-limiting, this cable can easily overheat process leads. If calculations show that overheating is likely, particularly at the higher ambient temperatures, then automatic temperature control is required. Thermostats can be used to control the electric power to the cable. This cable can be buried next to the process lead under the insulation along with the tracer. It should be at the location where the

July 1999

1500-68

Chevron Corporation



traced temperature is expected to be highest. The suppliers of electric heat tracing equipment can provide guidance in this area. Ambient temperature sensing thermostats are almost never used with constant heat density cable. The situations where this type of cable is used almost always require process lead temperatures above the highest expected ambient temperatures. Series Type Cable. In this type of cable the heating elements are insulated resistance wires that are connected together at the end remote from the power source so as to form a series path for the electric current. The required voltage for a given heat density (watts per foot of cable) is proportional to the length of the cable. Therefore, for a given cable, different voltages are required for different lengths, so changes in length cannot be made without substantially affecting the heat density unless the voltage is also changed. These features make it impractical for use in instrumentation where lengths are short, vary greatly, and cannot be determined in advance with any certainty. Internal Wire Impedance. Another method, well suited to long runs of piping but not for instrumentation, is the internal wire impedance system. In this system, a single insulated conductor is enclosed in a steel tube or conduit and is grounded to the tube at the far end. Current flows through the wire and returns through the tube. Heat is generated both by current flow in the wire and tube and by magnetic hysteresis and eddy currents induced by the alternating current. Net voltage on the entire length of the tube is very small because the voltage drop due to current flow in the tube is approximately equal to the voltage induced in the tube by transformer action from the current flow in the internal wire.

1559 Traced Tubing Bundles Definition Traced tubing bundles are standard factory fabricated assemblies made up of process lead tubing (one or two stainless steel tubes), a heat tracer, surrounding insulation, and a heavy black PVC plastic outer jacket. They are available using either self-limiting electric cable or steam tubing for the heat tracer. They are readily available in long lengths wound on wooden reels. Simple preinsulated tubing assemblies are made by eliminating the heat tracer. These are particularly well suited for steam supply and condensate return lines, for many analyzer sample lines, and occasionally for winterizing where heating is not required. The process tubing is usually 316 stainless steel with 0.035 inch wall thickness. A few vendors offer 0.049 inch wall tubing. Company preference for process applications is the 0.049 inch wall tubing.

Advantages of Traced Tubing Bundles for Process Applications Tubing bundles have some definite advantages: •

Chevron Corporation

The temperature maintained is more predictable.

1500-69

July 1999



•

Temperature difference errors are less likely in leads to differential pressure instruments.

•

Weather protection is more predictable and reliable.

•

Gaps in insulation are much less likely.

•

Total installed cost is usually lower.

•

Maintenance costs are generally lower.

•

Instrument housings are available designed specifically for use with tubing bundles.

Where to Use Traced Tubing Bundles With these advantages, tubing bundles should be used wherever possible. Typically this means with remote or grade-mounted instruments and transmitters. Tubing bundles are especially useful for analyzer sample systems. They are, however, totally impractical to use in short lengths such as on close-coupled instruments and line-mounted transmitters.

High Temperature Applications Electrically traced bundles using nonself-limiting cable are available for maintaining temperatures above the upper limits of self-limiting cable. Care should be exercised in obtaining and applying these cables because they may be custom built, and may not be suitable for the electrical area classification. Steam-traced bundles are usually used for such applications where steam is available at high enough pressures.

1560 Model Specifications, Standard Drawings, and Engineering Forms 1561 Standard Drawings

July 1999

GC-J1216

Steam Tracing Pressure Instruments

GC-J1217

Steam Tracing Level Instruments

GC-J1218

Steam Tracing Flow Instruments

GC-J1219

Connections for Line Mounted Pressure Gage in Outdoor Steam Service

GD-J1220

Electrical Heat Tracing Installation Details for Pipes, Valves and Fittings

GD-J1221

Electrical Heat Tracing Installation Details at Pipe Supports and Instrumentation

GD-J1222

Electrical Heat Tracing Installation Details for Accessories

1500-70

Chevron Corporation



1562 Engineering Forms ICM-EF-409

Requirements for Sealing, Purging, and Winterizing.

1570 References API Recommended Practice 551, Process Measurement Instrumentation

Chevron Corporation

1500-71

July 1999

1600 Field Instrument Installation Abstract This section covers recommended practices for the installation of field mounted instruments and provides information on the installation of pneumatic and electrical signal transmission systems. For details concerning the installation of specific process instruments (such as individual pressure, temperature, level, and flow devices) see the corresponding sections of this manual. Contents

Page

1610 Introduction

1600-3

1611 Process Connections and Piping 1612 Pneumatic Installations 1613 Electrical Installations 1614 Suggested Checklist 1620 Instrument Installation and Mounting

1600-4

1621 Field Instrument Installations 1622 Instrument Process Piping Material 1623 Instrument Field Supports and Mountings 1624 Instrument Accessibility 1625 Nameplates and Tags 1630 Signal Transmission Systems

1600-13

1631 Pneumatic Systems 1632 Pneumatic Raceways and Junction Boxes 1633 Electrical Instrument Installations 1634 Electrical Signal Transmission 1635 Electrical Cable Trays and Junction Boxes 1640 Model Specifications, Standard Drawings and Engineering Forms 1600-20

Chevron Corporation

1600-1

July 1999

1600 Field Instrument Installation


1641 Model Specifications 1642 Standard Drawings 1650 References

July 1999

1600-22

1600-2

Chevron Corporation



1610 Introduction This section provides general guidance on designing and reviewing instrument installations. To minimize design effort and manpower, a uniform approach to installation design should be established and followed for each facility. Installation design decisions and possible modifications must be made well before the construction stage. Otherwise, design omissions and modifications will cause scheduling constraints, encourage installers to improvise, possibly violate Standards and Codes, and compromise the quality of the installation. Instrument installations fall into three distinct categories: • • •

Process connections and piping Pneumatic connections Electrical connections and wiring details

1611 Process Connections and Piping Process connections and piping cover the physical installation and piping up of field instruments (sensors) that come in direct contact with process fluid. These instruments may get connected to piping, pressure vessels, or mechanical equipment depending on their function. Their installation, to ensure that they provide meaningful and reliable data is a primary concern. The designs must be economic, safe, and maintainable. To provide the most accurate readout, instruments should be mounted as closely as possible to their process connections. However, accessibility for maintenance and adverse ambient conditions at the process, e.g., vibration or high temperature, may require that the instruments be mounted remotely and connected by piping to their process connections. Also, instruments that tie into safety monitoring and/or shutdown systems must be installed so that they are testable without shutting down the process while they are being tested. Process installation design and review by the responsible instrument engineer include location and support of instruments, design of process piping and the test connections (if necessary), and assuring that the connecting piping stays functional and does not plug.

1612 Pneumatic Installations With the development of microprocessor based controls there is a continuing trend in the control industry to put less emphasis on pneumatic control systems. However, although pneumatic systems are slower than electronics, they are rugged and reliable and have been in use for a very long time. Operators, maintenance personnel, and many Company managers are very comfortable with pneumatic instruments. Even if the main control system may be electronic, almost all control valves and all field control loops will still be pneumatically actuated. In production operations, where environmental regulations permit and especially on offshore installations, natural gas is frequently used instead of instrument air to

Chevron Corporation

1600-3

July 1999



operate pneumatic instruments. Natural gas is a reliable source of actuating power and eliminates the capital expense to allocate space and to procure and install mechanical compressors for pneumatic systems, environmentally conditioned control rooms, signal distribution networks, emergency generators, and battery backups for electronic systems. The lost revenue of nonrecovered gas is more than offset by eliminating the expense to maintain such systems.

1613 Electrical Installations There are extensive applications for electrical and electronic control systems. Electrical instrument installations are very important to ensure signal isolation, system reliability, and conformance to codes.

1614 Suggested Checklist Review the following during the design phase: •

Proper sizes and rating of process connections on piping, vessels, and major equipment

•

Instrument installation details and piping isometrics

•

Heat tracings purges and winterizing of instruments, if applicable

•

Pneumatic installation details

•

Electrical installation details

•

Equipment-vendor-supplied instrumentation

This review should be given particular emphasis. Vendors are inclined to omit root valves and may try to substitute inferior instruments and installation material.

1620 Instrument Installation and Mounting 1621 Field Instrument Installations Location of Instrument Process Connections Locations of instrument process connections are determined either by the process equipment vendors or by the piping designers based on the applicable P&IDs. Instrument engineers should review the instrument process connections to ensure that piping for flow-meter runs and proper sized connections for relief and control valves have been provided. The exact locations of instrument process connections should appear on piping isometrics and on equipment and piping layout drawings. The size and type of material to be used for process connections are defined by the piping and equipment specifications published for the particular process condition.

July 1999

1600-4

Chevron Corporation



The instrument engineers and designers should review the location of the process connections against the P&IDs to ensure the following are correct: 1.

Orientation of instrument block and bleed valves is in compliance with piping and instrument specifications.

2.

Process connections of in-line mounted instruments match the connections provided by piping designers (e.g., flange size and rating), block valves (where required), and that adequate accessibility for servicing and for manual operation has been provided.

3.

Shutdown switches cannot be inadvertently blocked off while isolating other instruments for maintenance. Shutdown actuating devices should be installed on dedicated process connections, separately from other instruments.

Once the exact locations of instrument process connections are established, the instrument installation details and instrument location plan drawings should be finalized.

Instrument Installation Detail Drawings Instrument installation details should provide the instrument connection arrangement and tabulate the material required for the installation. When many instruments need to be installed, it is easier to separate the details by their function such as: • • • • •

Process connections Pneumatic connections Electrical connections Heat tracing and winterizing Miscellaneous (field supports, special fittings, etc.)

This approach is more efficient because the same process connection detail may apply to either pneumatic or electronic instruments, and heat tracing requirements and needs for special fittings can also be applied interchangeably. When craft labor is involved, this grouping also helps separate and define craft responsibilities.

Process Installation Review Process instrument installation drawings should be reviewed to ensure that the following conditions have been met:

Chevron Corporation

•

All process pressure connections have been provided with 3/4-inch root valves.

•

Designs include provisions for testing instruments in alarm and shutdown services.

•

Pressure instruments have a safe means of depressuring the installation for maintenance.

•

Instruments are mounted above the line, and the process lead slopes down to the process connection in condensing vapor services. If process temperature is above 150°F (except for steam flow meters), make sure that there is a siphon in the process lead (either purchased or field fabricated).

1600-5

July 1999


July 1999


•

Instruments are mounted below the line, and the process lead slopes up to the process connection in liquid services.

•

Remotely mounted instruments are installed using stainless steel tubing beyond the root valve. The tubing should be installed with a minimum slope of 1/2 inch per foot (1 to 2 inches per foot is preferable if space permits) and should be routed to avoid formation of pockets (high or low points where liquid or vapor can get trapped and introduce an error into the instrument reading).

•

Sufficiently long meter runs (upstream and downstream) have been provided for flow instruments that require meter runs, e.g., orifice meters, vortex shedding meters, or turbine meters.

•

Straightening vanes in lieu of long meter runs (to reduce turbulence in the flow) have been considered where space is limited.

•

Orifice meter runs in wet gas or vapor service should either be oriented vertically with downward flow, or have a weep hole in the orifice plate in a horizontal run, to prevent accumulation of liquid on the upstream face of the plate.

•

Differential pressure (d/p) flow instruments on orifice meters, both transmitters and bellows-type readouts and switches, have been provided with bypass manifolds at the instrument. (Three-valve manifolds are used in most manufacturing and chemical applications. Five-valve manifolds are used in production applications.) These manifolds are used by maintenance to zero and calibrate the meters.

•

On orifice meters in liquid service, the process taps are below the center of the line (but not in the bottom), and the instrument is installed below the line.

•

On orifice meters in steam service, process leads are extended above the line to exactly the same elevation before they are routed back down to the instrument mounted below the line. In steam service, steam condenses in the process leads and the condensate isolates the instrument from the high temperature of the steam. Running the leads up to the same elevation keeps the height of the condensate in the leads equal and balances out errors that would be introduced if the leads were at different elevations.

•

Seal pots are provided on bellows-type orifice meters in steam service. Bellows meters are high displacement instruments and if the flow changed, an error equal to the displacement of condensate in the leads would be introduced until the steam recondensed. Seal pots provide a surge volume in this application.

•

Seal pots are provided on all steam orifice meters in freezing climates. In freezing climates, steam meter leads have to be filled with anti-freeze. Seal pots identify the instruments that need to be winterized and state the type of antifreeze used.

•

All blind (nonindicating) transmitters and switches have been provided with pressure gages.

1600-6

Chevron Corporation



•

For instrument in toxic (e.g., H2S), corrosive, and high-pressure service, process connections include a restriction fitting in the instrument end of the root valve to limit accidental spill of process liquids. Vent and drain valves are piped to drains for personnel and environmental protection.

•

Capillaries for instruments with diaphragm seals and capillary tubing are supported in small channels to protect tubing from possible damage and to provide protection from sunlight. Excess tubing is neatly coiled at instrument end.

•

Instruments connected to process fluids that may solidify or increase in viscosity enough to impair measurement, have been heat traced or the process fluid isolated from the instrument by seal pots, diaphragm seals (gage protectors), purging, or other means. Refer to Section 1500, “Instrument Seals, Purges, and Winterizing,” of this manual and ICM-EF-409 “Seal, Purging, and Winterizing of Instruments.”

Standard Drawings GB-J1143 through GB-J1148 show typical installation details for pressure instruments. Standard Drawings GB-J1158 through GBJ1173 show typical installation details for level instruments. Standard Drawings GB-J1177 through GB-J1187 show typical installation details for flow meters. Standard Drawings GB-J1196 through GB-J1203 show methods for installing temperature sensing elements in process lines and equipment.

1622 Instrument Process Piping Material Instrument process piping is defined as all piping or tubing between an instrument and the root (block) valve nearest the process. Stainless steel tubing is preferred over rigid piping because it is more resistant to corrosion and is easier and more economical to install. With proper tools tubing can be bent to form slopes, pigtails and the required configuration for the process installation of instruments. Instrument process tubing should conform to ASTM A269. The tubing can be 304 or 316 stainless steel. Seamless, fully annealed and pickled tubing is usually preferred. Welded tubing is available, but there have been problems with the quality of the welds. ASTM A269 requires that the allowable working pressure for welded tubing be derated. Figure 1600-1 shows allowable working pressures for seamless tubing and specifies the ASTM derating requirements. Allowable working pressure at -20°F to 100°F is calculated as specified by ANSI B31.3 code.

Tube Fittings Tube fittings for process tubing should be the flareless compression type of 316 stainless steel and should conform to ASTM A269.

Chevron Corporation

1600-7

July 1999



Fig. 1600-1 Pressure Rating of Seamless 304 and 316 SS Tubing Wall Thickness (Inches) 0.028

0.035

0.049

0.065

Pressure Rating(1) Tube O.D., In.

psig

psig

psig

psig

1/8

8500

10900

—

—

3/16

5400

7000

10200

—

1/4

4000

5100

7500

10200

5/16

—

4000

5800

8000

3/8

—

3300

4800

6500

1/2

—

2400

3500

4700

(1) For welded tubing, multiply pressure rating by 0.80 for single welded tubing and 0.85 for double welded tubing.

Male threads should be rolled. Nuts should be pre-lubricated. Tests performed by Chevron Research Corporation and by several fitting manufacturers show that unless a compression fitting is improperly installed (mismatched fitting components, fitting not tightened properly, ferrule installed backwards, or the tube not inserted completely into the fitting before tightening), the tube will burst before a compression fitting will fail. Therefore, the allowable working pressure of the tubing is the limiting factor on tubing installations.

Instrument Hand Valves Instrument hand valves are defined as block, bypass, or bleed valves; remote or close coupled with instruments; mounted between process root valves and instruments. The instrument hand valves can be single valves, single valves with multiports, and three- or five-valve manifolds. The type of hand valves can be needle, ball, or globe design, with hard or soft seats for various temperature and pressure ratings. Needle valves are used for manual throttling of flow for purging, sampling, or just for straight shutoff. Multi-port gage valves provide extra pressure connections for bleed or test of pressure instruments. Three- or five-valve manifolds are used for differential pressure instruments where the block, bypass, and bleed valving combinations are required for servicing and calibration of instruments. Three-valve manifolds are also used for pressure instruments where an additional gage and bleed connection is required. Valve bodies, stems, and trim should be 316 stainless steel.

July 1999

1600-8

Chevron Corporation



Valves in services with pressures below 4,000 psig can have compression-type tubing connectors. Valves in services above 4,000 psig should have bar-stock bodies and threaded end connections. Teflon packing can be specified for services below 400°F (204°C). Grafoil packing should be specified for services above 400°F (204°C). Additional piping material like nipples and pipe fittings should be in accordance with applicable piping specifications.

1623 Instrument Field Supports and Mountings Most of the field transmitters, recorders, and controllers are supplied with 2-inch pipe mounting brackets. Surface mountings are available for mounting instruments remotely from the process. The required mounting must be specified by the instrument engineer. The 2-inch pipe mounting brackets are specified for most field mounted instruments.

Types of Standard Instrument Supports 1.

Free standing 2-inch pipe stanchions with flanged, concrete, or welded base

2.

Two-inch pipe saddle type supports (with U-bolt clamps to process pipe)

3.

Two-inch pipe supports with welded or flanged wall mountings

4.

Two-inch multiple pipe supports for two or three instruments

5.

Surface mounts supported by 2-inch pipe

6.

Target supports for field gages

See Standard Drawing GB-J1233, Field Mountings for Instruments.

Preferred Instrument Mountings: Line-mounted flow measurement d/p cells are usually mounted on 2-inch pipe saddles fastened by U-bolts or chain to the process line. Remote mounted flow and level d/p cells are mounted on free standing 2-inch pipe stanchions or wall mountings. Bellows type indicators, recorders, and switches are usually mounted on 2-inch pipe stanchions. Field recorders and indicating controllers are usually mounted on free standing 2inch pipe stanchions that are easily accessible for calibration and servicing. However, it is sometimes desirable to rack mount these instruments with surface mounts.

Chevron Corporation

1600-9

July 1999



Most field gages, switches, and all of the temperature elements are line-mounted (i.e., screwed directly into root valves or process connections). Temperature elements can also be flanged to process connections. If it is preferred that the readout be located mounted remotely from process connection, field gages can be ordered as surface-mounted.

Material for Supports Pipe for instrument supports should be steel 2 inches and a minimum of Schedule 40. Pipestand baseplate should be a minimum of 8-inches square and manufactured of 3/8-inch steel plate. Pipestands should be of all-welded construction, and the entire mounting assembly should be hot-dip galvanized.

Mounting of Field Instrument Supports Instrument supports should be welded (not bolted) to the steel floors, platforms, or structural steel. They can also be bolted to concrete floors and walls. Welding on process equipment or process piping is never allowed.

Instrument Location Plan Drawings Instrument location plan drawings furnish the following design and construction information: •

The approximate locations and mounting elevations of remotely mounted instruments (relative to some fixed point such as a pipe line, process connections, or final control element)

•

The approximate location of instrument air subheaders (relative to main air headers)

•

The exact locations and orientation of instrument junction and terminal boxes

•

The approximate routing of tubing runs

Review of location drawings When remotely mounted instruments are used, the instrument engineer should review the installation details and the instrument location plan drawings to ensure that the following conditions are met:

July 1999

•

Easy access is provided for maintenance, repair, and calibration.

•

Sufficient front and rear clearance is allowed for access to and removal of the instrument.

•

There is a clearance of at least 2 feet per 100°F (38°C) between equipment with surface temperatures in excess of 200°F (94°C).

1600-10

Chevron Corporation



•

Instruments are mounted so that the centerline of the instrument is either 4 feet or 4 feet and 6 inches above grade (or platform elevation) depending on local practices.

•

Instruments are located as close to the primary connection as possible, consistent with instrument accessibility and mounting requirements. Remote-mounted instruments should be mounted no more than 25 feet from the point of measurement. The shorter the lead the better the response.

•

Instruments in gas, vapor service and services where plugging is likely to occur, are installed above their process connections.

•

Instruments in clear liquid and steam services are installed below their process connections.

•

Indicating controllers or receiver gages from non-indicating transmitters are installed where their scales can be clearly visible from a point where manual control may need to be performed, e.g., at the by-pass valve around a control valve manifold.

•

Instruments are not mounted on handrails or on removable structures.

•

Instruments are not mounted in locations subject to leaks, spills, or damage from process equipment. Where such locations are unavoidable, a suitable shield or protective cover is provided.

•

Instruments are not attached to vibrating structures. Where such locations are unavoidable, antivibration mounts should be provided.

•

Instruments are not located in areas where the ambient temperature exceeds the limits specified by the manufacturer.

1624 Instrument Accessibility Accessibility to instruments, safety and ease of maintenance, and servicing are important. Adherence to the Company “Safety in Designs” publication should help ensure safe accessibility to instruments. Instrument accessibility must include access to the associated process block valve (i.e., root valve or the first shut-off valve remote from the root valve) and to the instrument air supply block valve. Types of instrument access are described below. Instrument access requirements for specific instruments are given in Figure 1600-2.

Chevron Corporation

•

Grade or Permanent Platform Access: Platforms should have either permanent ladders or permanent stairways.

•

Permanent Ladder Access: Access while standing on a permanent ladder.

•

Portable Stepladder or Rolling Platform Access: Access limited to 10 feet above grade or platform. Access from a portable stepladder or rolling platform at grade is subject to the following conditions:

1600-11

July 1999


– – –


The location is easily accessible by one person carrying a stepladder or moving a platform. The location is safe for placement of a stepladder or rolling platform. Access from a portable stepladder or rolling platform on an elevated platform is not allowed.

Fig. 1600-2 Access for Instruments Type of Instrument

Platform or Grade

Stepladder or Rolling Platform

Permanent Ladder

Transmitters (blind or indicating)

Yes

Yes

No

Field Controllers

Yes

No

No

Field Recorders and Indicators

Yes

No

No

Field Switches (alarm and control)

Yes

No

No

Routine Test Facilities for Alarms and Shutdowns

Yes

No

No

Control Valves and other Final Control Elements

Yes

No

No

Field Pressure Gages

Yes

Yes

Yes

Dial Thermometers

Yes

Yes

Yes

Thermocouples and Resistance Bulbs

Yes

Yes

Yes

Temperature Test Points

Yes

Yes

Yes

Level Gage Glasses

Yes

No

Yes

Analytical and Other Special Instruments(1)

Yes

No

No

Hydrogen Sulfide, Combustible Sensors, etc.

Yes

Yes

No

(1) This includes but is not limited to radioactive, chromatographic, capacitance, pH, boiling point, moisture, viscosity, oxygen, and specific gravity analyzers.

When instruments must be installed in overhead pipeways they should be grouped (if at all possible) to simplify access to the instruments. Access from portable stepladders or rolling platforms is acceptable for transmitters installed in the bottom level of overhead pipeways and mounted below the line (e.g., liquid or steam service). Vertical distance should be reviewed to make sure that a maintenance person can work on the instrument at chest height if at all possible.

July 1999

1600-12

Chevron Corporation



Process connections for instruments mounted in overhead pipeways above the line or in multilevel pipelines should be grouped and installed where they can be reached from a walkway across the pipeway. Transmitters should not be mounted between pipeway levels. Access to Flow Instruments in Pipe Trenches or Grade Level Pipeways: Where the orifice plate is located in a pipe trench or grade-level pipeway, a permanent platform or walkway should be provided for access to the instrument.

Access Requirements for Specific Instruments The types of access acceptable for specific instruments are outlined in Figure 1600-2.

1625 Nameplates and Tags All instruments should be labelled for identification. Field instruments should have 316 stainless steel nameplates permanently attached to the instruments with 316 stainless steel screws, rivets, or wire by the manufacturer. Nameplates should be stamped with the instrument tag number, purchase order number, model number, and serial number. Lettering size for nameplates should be a minimum of 3/16 of an inch. Nameplates for intrinsically safe instruments should include the following information. •

Manufacturer

•

Model

•

Serial number

•

Instrument tag number

•

Hazardous area class and group suitability

•

A statement indicating that the instrument is intrinsically safe

•

A statement indicating that any substitution of components may impair intrinsic safety

•

The agency certifying that the equipment is intrinsically safe

1630 Signal Transmission Systems 1631 Pneumatic Systems Although instrument air is the predominant medium used in pneumatic systems, natural gas is occasionally used by production facilities. Design considerations listed below for instrument air apply equally to gas.

Chevron Corporation

1600-13

July 1999



The main components of a pneumatic system are: • • • •

Instrument air supply and distribution (instrument air header and it’s laterals) Pneumatic supply to individual instruments Pneumatic signals between instruments Transmission of pneumatic signals to remote locations

Instrument Air Supply and Distribution Clean dry instrument air is needed for pneumatic instruments to operate. Instrument air is usually compressed to a pressure of 100 to 150 psig, dried to a dewpoint of at least 20°F lower than the lowest ambient temperature recorded at the facility, then piped throughout the facility to provide pneumatic power to instruments and final control elements. The primary system for distributing air to instruments in a facility is usually routed throughout the facility and is referred to as an instrument air header. Branch connections or laterals are piped from this header to provide air to instruments located in the same operating area. The size of these laterals is determined by the number of air users that the lateral has to supply. Branch headers and laterals should be sized as follows: Number of Instruments Supplied

Branch or Lateral Size

1 to 5

½ inch

6 to 10

¾ inch

11 to 25

1 inch

Full port ball valves are usually installed at each lateral and also at each end of the air header (for blowdown). Standard Drawing GB-J1214, Instrument Air Supply Header shows a typical air header layout.

Supply to Individual Instruments To minimize the air header size and to provide the needs of all final control elements, instrument air is usually distributed at a high pressure. Individual instruments operate at much lower pressures and cannot withstand the pressure at which air is available in the air header. To ensure that instruments operate properly and are not over-pressured, instrument air pressure is normally reduced at each instrument to match the rating of the instrument. Most transmitters and controllers require 18 to 22-psig instrument air, some controllers and all diaphragm control valves positioners require a 40-psig air supply, piston actuators and some other control elements require 60-psig and higher air supplies. Pressure gages on the outlet of the regulators (integral with or separate from the regulators) are used to set the air pressure. To satisfy individual instrument requirements, a ½ inch lateral is extended to the proximity of the instrument, a ball valve and a pressure reducing regulator are installed at the end of the lateral, and air is run through tubing to the instrument.

July 1999

1600-14

Chevron Corporation



¼-inch OD tubing is usually sufficient, but special applications may require 3/8-inch or ½-inch OD tubing. Standard Drawings GB-J1208 through GB-J1213 show typical instrument air supply piping to instruments.

Pneumatic Signals Between Instruments Pneumatic signals are usually transmitted through ¼-inch OD tubing. Bare copper, plastic coated copper, and stainless steel tubing can be used for pneumatic signals. Within control cubicles (panels) individual plastic tubes are sometimes used. When several pneumatic signals must be run in parallel, multitube bundles can be utilized economically. Plastic sheeted cables consisting of multiple plastic tubes can be run economically provided that they are protected from physical damage. Multitube cables can be installed in cable trays or pulled in conduit in overhead installations. They can also be installed in underground conduit. Multitube cables are available that are rated for direct burial — most Chevron Operating Companies avoid burying multitube cables without protection for fear of unforeseen physical damage, e.g., a backhoe cutting through a cable and shutting down a facility. Pneumatic transmission systems should be designed to limit transit time to about 3 seconds. If transit time is greater than 3 seconds, volume boosters should be provided to keep the transit time down.

Signal Tubing Installation Signal tubing should be straight run, securely fastened, and arranged to facilitate trouble shooting. Tubing should be installed with sufficient flexibility to allow for normal equipment movement.

Tubing Supports Single and multiple tubing runs should be supported every 3 feet of running length, both in and out of raceways. For support, single tubing runs should be placed in small steel channels or in galvanized thin wall steel conduits. Up to four tubes may be run in a single support conduit. Exposed ends of tubing should be not less than 18 inches nor more than 24 inches in length. Burrs and sharp edges should be removed from conduit ends to minimize damage to plastic coating of tubes.

Pneumatic Tube Routing Instrument tubing should be installed in raceways, on pipe racks, or otherwise supported to result in a plumb, level, vibration free and neat installation. Tubing should be routed so to avoid removal of the tubing when maintenance or repair of equipment is required.

Chevron Corporation

1600-15

July 1999



Tubing should not be routed through areas where a high risk of fire or mechanical damage exists (such as over vessels, compressors, or pumps), unless required by the application, e.g., emergency shutdown system, size loop, etc.

Aboveground Installation of Multitube Cables Cables should be installed in horizontal racks in the center of the pipeway as far from pump rows as possible. Plastic cable tubing should be installed away from any heat source and should be suitably protected in areas of likely mechanical damage. Preferred cable sizes are 4, 12, and 19 tubes per cable. Spare tubes should be provided in each cable; 2 in 4, 3 in 12, and 4 in 19.

Underground Installation of Multitube Cables Multitube cables should be pulled in conduits, one cable per conduit. Multitube conduits should be installed in the same manner as that specified for electrical conduits. Conduits should be sized for a maximum of 53% fill. The total degrees of bend should not exceed 360 degrees between pull boxes. Permissible bend radius, pulling length, pulling load, and handling method should comply with the manufacturer’s recommendations. Preferred multi-tube cable sizes for underground installation are 19 and 37. Spare tubings should be provided in each cable; 4 in 19 and 6 in 37.

1632 Pneumatic Raceways and Junction Boxes Pneumatic Tubing Trays, Channels, and Raceways Tubing lines should be supported with tubing trays or channels. The installed tubing trays should not present a hazard to personnel, block accessways, or prevent passage of equipment. Tubing raceways of heavy-duty reinforced fiberglass or hot deep galvanized steel (galvanized in accordance with ASTM A123, Specification for Zinc (Hot-Galvanized) Coatings on Products Fabricated from Rolled, Pressed, and Forged Steel Shapes, Plates, Bars, and Strip) may be required for some applications.

Pneumatic Junction Boxes Multiple pneumatic lines should terminate in outdoor junction boxes that are centrally located with respect to a group of field-mounted pneumatic instruments. Tubing should enter outdoor junction box through the sides or bottom only and only through a bulkhead union connector. Bulkheads should be installed in accordance with API RP 552, Transmission Systems. A breather/drain fitting should be provided in the bottom of the box.

July 1999

1600-16

Chevron Corporation



Junction boxes are available which are made of steel, stainless steel, aluminum, and fiberglass reinforced plastic materials.

1633 Electrical Instrument Installations Electrical installations for instruments should be coordinated with electrical engineers. Instrument engineers who are not electrical engineers should have a solid understanding of electrical installation techniques and wiring and should be thoroughly familiar with the National Electric Code (NEC). Electrical instruments must meet the following conditions: •

Installations must meet NEC requirements, particularly as related to hazardous (classified) areas.

•

Wire and cable construction and materials must be suitable for the environment, e.g., high temperature and contaminants.

•

Flexible conduit connection must be provided between the instruments and the rigid conduit.

•

Drains should be provided at low points of conduits.

Standard Drawing GD-J1234 and GD-J1235, Instrument Conduit Connections show typical conduit connection configurations for instruments.

Purging of Electrical Instrument Enclosures Instrument enclosures can be purged to reduce the classification within the enclosure: Type X: From Division 1 to nonhazardous. Type Y: From Division 1 to Division 2. Type Z: From Division 2 to nonhazardous. Power should be turned on in a purged instrument enclosure only after a fourvolume purge (ten volume for enclosures exceeding 10 cubic feet) has been completed and a minimum internal pressure of 0.1-inch water column has been established. Type X purging requires that all electrical power to the purged enclosure be de-energized in the event of purge failure. Purge failure should be monitored and an alarm actuated by a pressure switch.

1634 Electrical Signal Transmission This subsection provides an overview of wiring designs and installations for the instrument engineer. The installation of the electrical signal transmission systems should be coordinated with electrical engineers.

Chevron Corporation

1600-17

July 1999



Two-Wire Transmitter System Most two-wire transmitters are powered from the main or local control room panel. The most common supply voltage is 24 volts DC. The standard operating signal level is 4 to 20 mA.

Four-Wire Transmitter System Four-wire transmitters are powered from field power sources independent from main or local control panel mounted instrument(s). The most common supply voltages are 24 volts DC and 120 volts AC. The operating signal level is 4-20 MA.

Wiring and Cable Practices The size of the signal wire should be in accordance with ELC-MS-3551, “Instrument and Control Cable Signal and Multipair Construction.” The wire size requirements depend on the classification of the signal as defined by the NEC. Single-pair instrument signal lines are usually wired to field junction boxes with screw-clamp type terminal blocks. Multiconductor cables are run from field junction boxes to control panels. The size of multiconductor cables should be in accordance with ELC-MS-3551. The most commonly used cable for low voltage signal level is No.16 American Wire Gage (AWG), twisted, shielded pair. NEC defines signal level classes and defines wire sizes for each class.

Cable Routing Parallel runs as well as crossovers of control signal wiring and cables and power wiring and cables require separation to prevent induction of stray voltages into the low level signal wiring. Specification ELC-MS-1675 in the Electrical Manual tabulates separation requirements based on signal levels and power levels. Control signal cables should not be routed through areas where a high risk of fire or mechanical damage exists, e.g., pump rows.

Intrinsically Safe Systems Intrinsically safe systems are described in detail in Section 1400, “Intrinsic Safety,” of this manual. Basically, intrinsically safe equipment and wiring are incapable of releasing enough thermal or electrical energy to cause ignition of a specific hazardous mixture in its most easily ignited concentration. Intrinsically safe installations are much more cost effective in hazardous (classified) locations when armored cables and cable trays are used for signal routing instead of explosionproof fittings and conduit. For a system to be intrinsically safe, power and current going to the field devices must be limited, and all field components must be certified intrinsically safe. Intrinsically safe field devices must be so identified. Instruments that have been tested and approved for intrinsically safe applications are identified by a blue label that states that the instrument is certified as intrinsically safe. Intrinsically safe

July 1999

1600-18

Chevron Corporation



system wiring should be identified by either color coding all cables, wire and conduits with a light blue color or by identifying them by other means such as signs or tags. The Instrument Society of America (ISA) Recommended Practice ANSI/ISA RP12.6 - “Installation of Intrinsically Safe Systems for Hazardous (Classified) Locations” defines the installation requirements for intrinsically safe systems, including installation methods, identification of wiring and components, and separation of intrinsically safe system wiring from nonintrinsically safe wiring.

System Grounds Incorrect grounding of instrument components and systems can introduce stray voltages and inaccurate reference ground potentials into the signals generated and received by electronic instrumentation. Requirements for grounding and isolating instruments and systems are explained in detail in Specification ICM-MS-3651, “Installation of Digital Instruments and Computers.” Standard Drawing GF-J1236, “Typical Ground System for Digital Instrumentation and Process Computers,” which is referenced in ICM-MS-3651 and is also enclosed in this manual shows recommended grounding connections for various combinations of electrical and electronic instruments and systems. This drawing was produced as a result of troubleshooting many grounding related problems in control systems. This drawing has also been reviewed and is supported by most of the major electronic instrumentation and control system vendors.

1635 Electrical Cable Trays and Junction Boxes This section provides an overview of installing cable trays. For more specific details refer to the section entitled “Electrical Installation,” in the Electrical Manual.

Signal Isolation and Routing Cable trays should be installed in accordance with the following:

Chevron Corporation

•

Separate trays should be provided for instrument signal cables and power cables.

•

Overhead routing of instrument cable trays usually provides the best possible mechanical protection and isolation of cables from electrical interferences.

•

Instrument cable trays should be routed away from pump rows and other high fire hazard areas.

•

Cable tray installations should not present a hazard to operating and maintenance personnel.

•

Material for cable trays and accessories should be chosen in accordance with the guidelines in the Electrical Manual.

1600-19

July 1999



Electrical Junction Boxes for Instruments Electrical junction boxes for outdoor instrument signal wiring should be installed complete with mounting hardware, nameplates, terminal blocks, and wiring. In addition, the following conditions should be met: •

NEMA 4X enclosures are normally preferred for outdoor electrical junction boxes when they are not required to be explosionproof

•

Explosionproof junction boxes should conform to the NEMA 7 specifications

•

Entrances to junction boxes should be from the side or bottom. No holes are allowed in the top of any junction box.

•

High-density terminal strips should not be used in junction boxes

•

Terminal boxes should be mounted at an accessible height and should be provided with breather vents and drains

•

Terminal boxes should be provided with low surface temperature heaters where required

1640 Model Specifications, Standard Drawings and Engineering Forms 1641 Model Specifications ELC-MS-1675

Installation of Electrical Equipment

ELC-MS-3551

Instrument Wire and Cable

ELC-MS-3552

Thermocouple Extension Wire and Cable

ICM-MS-3651

Installation of Digital Instruments and Computer

ICM-MS-4815

Instrument Item List

1642 Standard Drawings

July 1999

ICM-EF-409

Seals, Purging and Winterizing of Instruments

GB-J1143

Pressure Gage installation - 3/4" Root Valve

GB-J1144

Pressure Gage with Diaph. Seal - 3/4" Root Valve

GB-J1145

Remote Mounted Pressure Instrument

GB-J1146

Remote Mounted Pressure Instrument — With Pressure Gage

GB-J1148

Draft Gage

GB-J1158

Level Gage With Screwed Process Connections

GB-J1159

Level Gage With 150# Flanged Process Connections

1600-20

Chevron Corporation


Chevron Corporation


GB-J1160

Level Gage With 300# Flanged Process Connections

GB-J1161

Level Switch, External Float Type With Screwed Process Connections

GB-J1162

Level Switch, External Float Type With 150# Flanged Process Connections

GB-J1163

Level Switch, External Float Type With 300# Flanged Process Connections

GB-J1164

Level Instrument, External Displacer Type With Screwed Process Connections

GB-J1165

Level Instrument, External Displacer Type With 150# Flanged Process Connections

GB-J1166

Level Instrument, External Displacer Type With 300# Flanged Process Connections

GB-J1167

Vessel Connections for Level Instruments and Gages With Screwed Process Connections

GB-J1168

Vessel Connections for Level Instruments and Gages With 150# Flanged Process Connections

GB-J1169

Vessel Connections for Level Instruments and Gages With 300# Flanged Process Connections

GB-J1170

Vessel Connections and Level Instrument Bridle (Strongback) Connections

GB-J1171

Differential Pressure Level Transmitter Mounted Below Lower Tap

GB-J1172

Differential Pressure Level Transmitter Mounted at Lower Tap

GB-J1173

Differential Pressure Level Transmitter Mounted At or Above Upper Tap

GB-J1177

Differential Pressure Flow Instrument, Gas Service - Instrument Above Taps

GB-J1178

Differential Pressure Flow Instrument, Liquid Service - Instrument Below Taps

GB-J1179

Differential Pressure Flow Instrument, Dry Gas Service Instrument Below Taps

GB-J1180

Differential Pressure Flow Instrument, Wet Gas Service Instrument Below Taps

GB-J1181

Differential Pressure Flow Instrument, Steam Service - Instrument Below Taps

GB-J1182

Differential Pressure Flow Transmitter, Gas Service - Instrument Above Taps

GB-J1183

Differential Pressure Flow Transmitter, Liquid Service - Instrument Below Taps

1600-21

July 1999



GB-J1184

Differential Pressure Flow Transmitter, Dry Gas Service Instrument Below Taps

GB-J1185

Differential Pressure Flow Transmitter, Wet Gas Service Instrument Below Taps

GB-J1186

Differential Pressure Flow Transmitter, Steam Service - Instrument Below Taps

GB-J1187

Differential Pressure Flow Throat Tap Connections

GB-J1196

Screwed Thermowell Installation in Piping

GB-J1198

Flanged Thermowell Installation in Piping

GB-J1200

Thermowell For Furnace Stack

GB-J1201

Furnace Tube Skin Point Thermocouple

GB-J1202

Reactor Skin Point Thermocouple

GB-J1208

Air Connections for Automatic Pump Start (Plastic Jacketed Copper Tubing)

GB-J1209

Air Connections for Automatic Pump Start (Stainless Steel Tubing)

GB-J1210

Air Connections for Instrument with Plastic Jacketed Copper Tubing

GB-1211

Air Connections for Control Valve with Plastic Jacket

GB-1212

Air Connections for Instrument with Stainless Steel Tubing

GB-J1213

Air Connections for Control Valve with Stainless Steel tubing

GB-J1214

Typical Instrument Air Supply Header

GD-J1233

Field Mountings for Instruments

GD-J1234

Instrument Conduit Connections

GD-J1235

Instrument Conduit Connections

1650 References API RP 14F, Design and Installation of Electrical Systems for Offshore Production Platforms API RP-551, Process Measurement Instrumentation API RP 552, Transmission Systems ASTM A123, Specification For Zinc (Hot-Galvanized) Coatings on Products Fabricated From Rolled, Pressed, and Forged Steel Plates, Bars and Ships

July 1999

1600-22

Chevron Corporation

1700 Wellhead Control Systems Abstract This section provides guidance for the design of a wellhead control system to be used to monitor the flowing conditions of the well flowline and to initiate a shutdown of the well. The various components of a wellhead shutdown system, their function in the system, and their operation are discussed. Shutdown systems for both surface controlled and subsurface controlled shutdown systems are included. Contents

Page

1710 Basic Principles

1700-2

1711 Shutdown Sensors 1712 Fail-Safe Design 1713 Surface Safety Valve (SSV) 1714 Surface-Controlled Subsurface Safety Valves (SCSSV) 1720 Design Considerations

1700-18

1721 Flowline Pressure Rating, Ten Foot Rule 1722 First Out Indication 1723 Testability and Maintenance During Operation 1724 Panels for Wellhead Controls 1725 Enclosures for Wellhead Control Systems 1726 Tagging and Nameplates 1727 Functional Requirements 1730 References

Chevron Corporation

1700-27

1700-1

July 1999

1700 Wellhead Control Systems


1710 Basic Principles The function of the wellhead control system is to monitor each well’s flowline pressure, interface with the facility Emergency Shutdown System (ESD), and control the surface safety valve (SSV). Like other shutdown systems described in Section 1300, a wellhead control system should be used to prevent the risk of injury or damage to personnel, the environment, or equipment. Wellhead control systems are designed to be “fail-safe.” When required, the wellhead control system also monitors the surface controlled subsurface safety valve (SCSSV) for the well. SCSSVs are required for offshore platforms. Regulations, such as API RP 14 C and the Minerals Management Service OCS Order No. 5, govern the requirements for these safety systems. SCSSVs have been installed on land wells in some facilities that are located near population centers or where the wellhead may be subject to physical damage. When hydraulically operated SCSSVs are required, the wellhead shutdown system must include a hydraulic reservoir and pump system to maintain pressure on the subsurface valves during normal operation. Almost all wellhead control systems are pneumatic for sensing and control of the surface safety valve (SSV). Pneumatics have been widely used for a long time and are accepted by the users. Pneumatics work well in the vicinity of wellheads, where they are subject to vibrations and fluids from drilling activities. On land, a separate wellhead control system is usually provided for wells operating under pressure flow conditions, when damage or injury to the environment, personnel, or equipment could occur. In temperate climates the controls may be mounted individually out-of-doors. In colder climates the controls are often mounted in a small panel, which may be housed in a building or shelter. On offshore platforms, the wellhead control systems are grouped on one or more panels. The control logic for each well is kept separate from the other wells so that wells may be easily added to or deleted as required. SCSSVs are hydraulically controlled and operated from these panels. A separate hydraulic system is usually furnished for every panel. Shutdown controls for each pneumatically operated SSV and its SCSSV are arranged together on the same control panel. When instrument air is available, as it is on many offshore platforms, it is the best source of gas pressure to operate these safety control systems. When instrument air is not available, process gas can be used. The source of gas pressure for the control system must be dry and filtered and free of contaminants. Nitrogen is often used on land as the source of gas pressure in the following circumstances: • •

July 1999

When only sour gas is available When hydrocarbons cannot be vented to the atmosphere

1700-2

Chevron Corporation


•


When a source of clean, dry gas is unavailable

In extreme cold conditions, hydraulics have been used for the surface control systems.

Electric Wellhead Control Electrical control systems have been developed and are very feasible for harsh climates or when handling toxic fluids in the flowline. One Company location in Wyoming, faced with an extremely high level of H2S, has successfully used electric pressure switches, hand switches, and solenoid valves on every wellhead in the entire field for years. They also included H 2S detectors as one of the shutdown sensors. When annunciation or indication of the shutdown sensor is important, electric shutdown systems would be more flexible, easier to implement, and more cost effective than pneumatic systems. Electric systems are far easier to interface with a Supervisory Control And Data Acquisition (SCADA) system for remote monitoring and control.

1711 Shutdown Sensors Generally, four types of shutdown sensors are used to send signals to each wellhead control panel: • • • •

Process and ESD shutdown pilot relays Fusible plugs on fire loop systems High- and low-pressure sensors from the flowline Sand probes in the flowline

SSVs are closed by any of the above devices whether on land or offshore. SCSSVs are shut down only by an ESD or fire signal and only after a time delay to allow the SSV to close first.

Process Shutdown Relays The pneumatic wellhead control system must be pressurized in order to operate. The interface with the process or platform shutdown system is accomplished by using a pilot relay. Pilot relay is short for a “pilot operated, three-way valve or relay.” The relay consists of a three-way block and bleed valve and a pneumatic piston or a pressure pilot on the other end. The relay is mounted inside the wellhead control panel. This pilot relay will automatically reset when the process shutdown system has been repressured or brought back to a normal situation. With a signal applied from the remote process or platform shutdown system, the three-way pilot valve allows free passage of the pneumatic holding pressure. When the remote signal is removed, the relay will switch, block in the supply pressure, and vent all the downstream holding pressure. This series of actions will initiate a wellhead shutdown by venting the pneumatic signal from a manually reset pilot relay that controls the surface safety valve (SSV).

Chevron Corporation

1700-3

July 1999



Manual Reset Relays A shutdown signal from a plant or platform ESD to the wellhead control system is usually caused by a possible dangerous or damaging condition existing downstream of the wellhead. It is important that the wellhead shutdown system be designed with a manual reset pilot relay. See Figure 1700-1. The manual reset pilot relay has a knob on one end as well as the pressure pilot on the other end of a three-way valve. The three-way valve will switch and shut down the surface safety valve either when the knob is pushed or the pilot is depressured. Fig. 1700-1 Pilot Relay with Manual Reset (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.) The Model 3300-A pilot relay valve is a normally closed, block and bleed three-way valve. The relay will stay locked in the normally closed position until the pull knob is pulled out. In this position, the spool valve is pinned with the automatic bypass holding the spool valve open, allowing actuator pressure to flow through. When instrument pressure is applied to the bottom of the relay piston, the spool valve is moved upward, automatically releasing the automatic bypass pin. As long as instrument pressure is present at the bottom of the relay piston, the relay will stay in operating position. Loss of instrument pressure returns the relay to its locked closed position and back-bleeds the downstream side. In the closed position, the relay is automatically locked closed and must then be manually reset for operation. FEATURES: 1. Fail safe. 2. Automatic bypass pin release. 3. Internal lock-closed device. 4. Manual shut-in of SSV by pushing in pull knob. 5. Can control multiple wells individually. 6. Pressure through relay from 0 to 250 psi. 7. Instrument pressure from 30 to 40 psi. 8. Materials meet NACE MR-01-75 specifications. 9. Viton seals. (Temperature range -20°F +400°F.) 10. Can be panel-mounted. Hole size is 1.5 inches. 11. All ports 1/4" NPTF

This means that when the shutdown signal is initiated, the system will trip and stay shut down. Clearing of the possibly dangerous or damaging condition will not automatically reset the system and open the surface safety valve. The only way the SSV or the SCSSV can be opened is for the production operator to manually reset this shutdown relay by pulling the knob. Pulling the knob sets a pin that latches the

July 1999

1700-4

Chevron Corporation



three-way valve until pressure is applied to the pilot. The pin will disengage when the pilot is pressured, thus activating this shutdown relay. This relay should not have a lockout feature that disengages the manual reset, which would allow the surface safety valve to automatically cycle shut and open. The design of this safety relay should require that it can only be latched in the open position while no instrument pressure is applied to the pilot diaphragm. Indicators are available to enable the operator to tell at a distance that a wellhead surface safety valve is shut down. On panels controlling multiple wells, indicators for the valve signals are very helpful to the operator. These indicators can be integral to the relay knob, or they can be a separate panel-mounted device. It is important to determine that the possibly dangerous or damaging condition has been corrected and to make sure that the operator is present at the wellhead to monitor the well while it is placed back into operation.

Pressure Sensors High- and low-pressure sensors are used to monitor the flowline pressure of the well downstream of the choke. The high-pressure sensor is used to protect both the final flowline segment and the downstream process equipment. The low-pressure sensor is used to detect a leak or flowline rupture. Requirements for pressure sensors are covered in API RP 14C Section A1 for offshore platforms and by established Company practices. The settings for these sensors must be carefully determined, reviewed, and documented. Because of the pressures normally encountered in flowline service and the proximity to the workover operations, the most commonly used pressure sensors are called “stick pilots.” See Figure 1700-2. These pressure sensors come in a wide selection of ranges, which can be easily adjusted. The sensors have 1/2 NPT connections and are typically color-coded to identify the spring range. See Figure 1700-3. The pressure sensors are usually connected in tandem. The holding circuit supply pressure is connected to the inlet of the high-pressure pilot. The low connection of the high-pressure pilot is connected to the inlet connection of the low-pressure pilot. The upper connection of the low-pressure pilot continues through the system holding circuit. A typical pneumatic hookup is shown in Figure 1700-4. Should the flowline pressure downstream of the choke exceed the preset limit, the high-pressure sensor internal piston will shift upwards, blocking its supply port and venting the holding circuit pressure, thus triggering a wellhead shutdown. Should the flowline pressure decrease below the low limit, the low-pressure sensor internal piston will shift down, blocking its supply port and venting the holding circuit pressure, thus triggering a wellhead shutdown. Pressure sensors are normally mounted on a manifold on the flowline and on the pneumatic signal tubing sent to the wellhead control panel. Some of the reasons for remote mounting are as follows.

Chevron Corporation

1700-5

July 1999



Fig. 1700-2 Typical Pressure Sensor (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.)

Fig. 1700-3 Typical Pressure Sensor Ranges (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.)

July 1999

1700-6

Chevron Corporation



Fig. 1700-4 Pressure Sensors in Tandem Service (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.)

•

To avoid high wellhead pressures and process fluids in the control panel

•

To reliably measure viscous liquid hydrocarbon pressures by having the sensors mounted close to the process

•

To minimize the risk of chloride or sulfide stress cracking when corrosive fluids are present

•

To avoid plugging problems in the tubing when paraffins are present

•

To avoid plugging problems in the tubing when the formation of hydrates is possible

•

To avoid freezing in long process leads in cold environments

•

To avoid mechanical damage during workover operations

Dual pressure pilot sensors, such as the Fisher Model 4660, have been directly mounted in the control panel in some locations.

Chevron Corporation

1700-7

July 1999



Sand Probes Sand probes are used on flowlines where erosion due to flowing conditions may be experienced. Under these conditions the probe will erode with the passage of sand and actuate the sensor. When properly placed in the flowline, the number of sand probe sensor failures can be valuable in helping to determine the erosion wear on the flowline. Therefore, the number and date of occurrence of sand probe failures should be carefully documented. One rule of thumb is to schedule for a wall thickness test (e.g., x-rays) after four sand probe failures. Sand probes should be inserted in a straight run of pipe at least ten (10) feet downstream of the well choke or any other change in piping direction. The pipe downstream of the probe should also be straight for another four feet. Probes should be selected for the line size of the flowline and should be purchased with 1/2 NPT connections. Figure 1700-5 shows a typical sand probe sensor. When erosion causes failure of the probe, the flowline pressure enters the sand probe sensor and the internal piston shifts upwards. The supply or instrument port is blocked, and the holding circuit pressure is vented, thus triggering a wellhead shutdown. The manual handle of the sand probe will give an instant indication that the sand probe has tripped. This manual handle may also be used to manually test the wellhead control system.

1712 Fail-Safe Design Two types of failure modes are described in Section 1300, “Process Alarm and Shutdown Systems.” One mode is de-energized-to-trip and the other energized-totrip. All wellhead control systems and components must use the de-energized-to-trip design, which means that the system will be a fail-safe design. As previously mentioned, the fail-safe design is accomplished by maintaining a pneumatic pressure holding circuit on all the components of the control system located at the surface of the wellhead and a hydraulic pressure on the SCSSV during normal operation. A typical simplified pneumatic circuit for a wellhead control of an SSV is shown in Figure 1700-6.

1713 Surface Safety Valve (SSV) The SSV is normally located as a wing valve on a high- or low-pressure wellhead Christmas tree. This name is derived from the tree-like appearance of the valves and fittings branching out from it. A manual valve must be installed between the SSV and the well to allow for maintenance. See Figure 1700-7. The type of valve used in the flowline for a shutdown valve is usually a reverse gate valve. This valve is well suited for this application due to its self-closing feature.

July 1999

1700-8

Chevron Corporation



Fig. 1700-5 Typical Sand Probe Sensor (Courtesy of the Cooper Cameron Corporation, owner of the W-K-M trademark.)

Chevron Corporation

1700-9

July 1999



Fig. 1700-6 Simplified Wellhead Control for SSV

The valve consists of a gate assembly that operates at ninety degrees to the pathway through the valve. The valve stem and gate rise to effect closure. This stem action is opposite the stem action of a normal gate valve. The internals of the valve are designed so that the body pressure generates a force on the gate and stem in the upward direction, always tending to drive the valve shut. A diaphragm- or piston-type of actuator is used with a reverse gate valve. The valve is opened by applying pressure above the diaphragm, which drives the stem down. To close the valve, the pressure is removed from the diaphragm. The flowline pressure drives the gate stem upward, closing the valve. A spring, located below the actuator diaphragm, will also close the valve when equal pressure is present on both sides of the valve. The actuator must be sized above the maximum anticipated operating pressure or for the maximum allowable working pressure (MAWP) of the valve itself. A safety

July 1999

1700-10

Chevron Corporation



Fig. 1700-7 Typical Wellhead Christmas Tree with SSV

factor of about 25% for wear and friction losses in the future should be added. Diaphragm actuators are presently used up to 15,000-psig design pressures. Manual overrides for an SSV can be provided on land but are not allowed offshore. Three types are available: • • •

Hydraulic Handwheel Lockout cap

Lockout caps should be furnished with a fusible insert so that the valve will close in case of a fire. A diaphragm-actuated valve is shown in Figure 1700-8. All surface safety valves should have a firesafe seal on the shaft and an external relief valve on the actuator housing. A fusible link has been installed on the pressure line to the actuator in some areas. Many wellhead SSVs require a quick bleed or quick exhaust valve on long actuator supply lines to ensure that the valve closes quickly enough. See Figure 1700-9. SSVs should be provided with some means to visually indicate to the operator whether the valve is open or closed.

1714 Surface-Controlled Subsurface Safety Valves (SCSSV) SCSSVs are specially designed wellhead shutdown valves that are held open by the maintenance of a constant hydraulic pressure. These valves are usually located hundreds of feet below the bottom of the sea or surface of the land. Sometimes more than one valve is required per well.

Chevron Corporation

1700-11

July 1999



The valve is installed in the wellhead tubing and the hydraulic control tubing is run between the tubing and the intermediate casing. One type of SCSSV is shown in Figure 1700-10. The number of wells being controlled by a hydraulic system depends upon local preference. Sometimes individual hydraulic systems are preferred. Usually the wells are grouped in logical “blocks” that enable good operator access and control in case of a problem. Normally, a limit of no more than 10 to 20 wells per hydraulic system will allow all the SCSSVs to be reset in under 5 minutes. Hydraulic pressure is supplied by a pneumatically driven pump with a second pump as a backup. The backup pump can be another pneumatically driven pump with a manual operator option or just a manually operated pump. A low-pressure sensor can be installed to monitor the hydraulic pressure and alert the operator. A relief valve is provided on the discharge of the pump to relieve excess back pressure back to the supply tank. The main pump is driven by approximately regulated 100-psig air or natural gas. Regulations such as API RP 14B and OCS Order No. 5 govern the requirements for these systems.

Control of SCSSVs Each well should have its subsurface controls located in the control panel, adjacent to the surface controls. A typical hydraulic circuit for an SCSSV is shown in Figure 1700-11. To operate the control system in order to open the SCSSV and the SSV, the following sequence occurs:

July 1999

1.

By pulling the knob, the operator manually resets the pilot relay for the hydraulic system (MR-1). The pin will hold the relay open.

2.

Instrument gas will flow through the ESD/Fire Loop pilot relay (R-2), and a.

Pressure up relay MR-1 and release its pin

b.

Pressure up the hydraulic system latching relay (R-3) and allow gas to flow to the hydraulic pumps

c.

Start the pneumatically driven hydraulic pump and

d.

Close the hydraulic dump valve (R-4)

3.

The selected SCSSVs will open and at the shut-in tubing pressure (SITP) the hydraulic low-pressure switch will allow instrument gas to flow to the hydraulic pressure pilot relay (R-5).

4.

Instrument gas will flow through R-5 and through the process equipment S/D pilot relay (R-6) and will pressure up the field flowline instrument tubing up to the pressure-switch-low (PSL) switch.

1700-12

Chevron Corporation



Fig. 1700-8 Typical SSV with Diaphragm Actuator (Courtesy of Axelson, Inc.)

Chevron Corporation

1700-13

July 1999



Fig. 1700-9 Quick Bleed or Exhaust Valve (Courtesy of Otis Engineering Division of Halliburton)

5.

After confirming that the first SCSSV is open, the operator pulls the knob to manually reset the pilot relay for the first SSV (MR-7-1). The pin will hold the relay open.

6.

100 psig instrument gas will flow through MR-7-1 to the quick exhaust valve and open the SSV.

7.

After the first well is sending crude oil to the process equipment, instrument gas will flow through the PSL switch and the sand probe and back to the SSV pilot relay (MR-7-1), where the pin will release.

8.

The second well is put onstream the same way by manually resetting the SCSSV selector and confirming that the SCSSV is open, then resetting the SSV pilot relay (MR-7-2), etc., until all the wells are flowing.

Hydraulic Pumps Most hydraulic pumps are air or natural gas driven reciprocating pumps. Electrically driven pumps are occasionally used. As shown in Figure 1700-12, the gas power end has a larger area than the liquid end. The pump reciprocates due to the action of a cycling gas supply. The maximum discharge pressure of the liquid at no flow is approximately equal to the pneumatic supply pressure times the area ratio plus the liquid suction pressure. When this balance of forces is reached, the pump stalls and ceases pumping without consuming any supply gas. The pump will automatically restart when the hydraulic pressure drops below 97% of the design pressure. Normally, a backup manually operated hydraulic pump is used for maintenance or emergency use to keep the SCSSV open.

Time Delays to Close SCSSV A remote emergency shutdown from the platform ESD system must shut down the SCSSV, but only after a nominal 2-minute time delay.

July 1999

1700-14

Chevron Corporation



Fig. 1700-10 Typical SCSSV (Courtesy of the American Petroleum Institute)

This delay will ensure that the SSV at the wellhead is closed first to allow it to absorb the wear and tear of opening and closing against a differential pressure at flowing conditions. The SSV is much easier and cheaper to repair than the SCSSV. The time delay, which is adjustable, is accomplished by adding a needle valve and volume bottle in the pneumatic ESD signal line going to a pilot relay that controls the air or gas to the hydraulic pump and dump valve. The orifice in the needle valve restricts the air or gas bleed to the piston of a pilot relay. After about 2 minutes, the three-way valve in the relay will shift and depressure the system. This action will open the hydraulic dump valve, which quickly allows the hydraulic pressure on the SCSSV to be relieved to the supply tank. Refer to Figure 1700-13.

Chevron Corporation

1700-15

July 1999


Chevron Corporation

Fig. 1700-11 Typical Hydraulic Circuit for SCSSV

1700-16


July 1999



Fig. 1700-12 Schematics of Basic Pump Types

Fig. 1700-13 Time Delay Circuit to Close SCSSV Slowly

Safety Interlock to Open SCSSV First The SCSSV must have a safety interlock to make sure that this valve is opened before the SSV. The SCSSV is extremely difficult to open when there is a maximum differential pressure across the valve, which would be true if the SSV were opened first. The addition of a pilot relay in the pneumatic circuit is the method usually used to provide the safety interlock. Pressure is thereby supplied to the hydraulic pump

Chevron Corporation

1700-17

July 1999



supply pressure, the hydraulic dump valve, and the supply port of the manual reset pilot relay that controls the SSV.

Other Safety Interlocks and Features Low- and high-level alarms on the main hydraulic reservoir tank are normally included to alert the operator that a hydraulic leak or backflow of process fluids from the well through the control tubing has occurred. The hydraulic fluid reservoir must be adequately vented to prevent any pressure buildup caused by returning fluid during an emergency shutdown of all the SCSSVs, or in case of backflow from the well through the control tubing. A low-pressure alarm on the automatic hydraulic pump discharge alerts the operator of a pump malfunction before the SCSSVs all close against operating wells. Exact requirements for surface-controlled systems for SCSSVs used on offshore platforms are covered in API RP 14B and shown in Figure 1700-14.

1720 Design Considerations 1721 Flowline Pressure Rating, Ten Foot Rule Flowlines transport hydrocarbons from the wellhead to the first downstream process equipment or vessel. Flowlines are divided into flowline segments. A flow-line segment is a portion of the flowline that has an assigned operating pressure from the other portions of the flowline. Thus, most wells have an initial and a final segment with an inline pressure reducing device or choke separating the segments. API RP 14C Section A1 describes the requirements for safety devices and their location for offshore platforms. These rules follow the Company requirements and should apply on land as well as offshore. The following discussion is an abbreviation of RP 14C. It assumes that the flowline has only one choke and therefore is divided into two segments. Refer to Figure 1700-15. Two important considerations determine the requirements for safety devices on wellheads and flowlines. •

Is the first choke device in the initial flowline segment less than 10 feet from the wellhead?

•

Is the maximum allowable working pressure of the final flowline segment (after the choke) greater than or less than the shut-in tubing pressure (SITP) of the well?

When the distance to the choke is less than 10 feet, then pressure sensors in the initial segment of the flowline upstream of the choke are not required. When the distance is greater than 10 feet, then API RP 14C only requires a low-pressure sensor to detect leaks and ruptures. High- and low-pressure sensors are always required in the final flowline segment downstream of the choke. When the maximum allowable working pressure (MAWP) of the final segment of the flow-

July 1999

1700-18

Chevron Corporation



line after the choke is greater than the shut-in tubing pressure (SITP), then both high- and low-pressure sensors are required to detect a blocked line or flow control failure as well as a leak or rupture. However, when the MAWP of the final flowline segment is less than the SITP, then a pressure relief valve as well as both high- and low-pressure sensors are required. This requirement follows the concept of an independent backup device discussed in Section 1300, “Process Alarm and Shutdown Systems.” In the previous case, where the MAWP of the final segment is less than the SITP, API RP 14C allows the substitution of a second shutdown valve and an independent high-pressure sensor in place of a pressure relief valve. If the flowline has no choke then the MAWP for the entire flowline must be greater than the SITP. Both high- and low-pressure sensors are required in this case where there is only one segment. Flowlines may have more than two segments. API RP 14C should be consulted to determine the required safety equipment. For all flowlines, the following safety devices should be included: • • • •

Check valve in the final flowline segment to prevent any backflow ESD shutdown Fire loop shutdown Downstream process equipment shutdown

In some locations fusible plugs have been installed in the pressure line to the SSV actuator and in other locations they have been installed inside every control panel.

1722 First Out Indication If three sensors are used to shut down the surface safety valve (e.g., high-pressure, low-pressure, and sand probe), then first out indication may be desirable. Most wellhead control panels in the Company have not required this feature in the past, but new installations are beginning to put them in. First out indicators can help during the testing of pressure sensors, especially when multiple wellhead control panels are installed. The sensors are mounted on the flowline and pilot relays are installed in the wellhead control panel. Panel indicators are connected to the outlet port of each of the pilot relays. From written instructions, the operator can determine which sensor tripped by observing the number of tripped indicators. The knob on the sand probe could also be used as an indicator, but because it is mounted remotely, it may not be visible from the panel. A better way to do the same thing is to install first out indicators (e.g., Amot Model 4400) which have internal porting and will only indicate the sensor that tripped. See Figure 1700-16.

Chevron Corporation

1700-19

July 1999



Fig. 1700-14 Schematic of a Control System for SCSSVs (Courtesy of the American Petroleum Institute)

July 1999

1700-20

Chevron Corporation



Fig. 1700-15 Recommended Safety Devices for Wellhead Flowlines (Courtesy of the American Petroleum Institute)

1723 Testability and Maintenance During Operation A three-way valve on the panel can be installed to bypass the high- and low-pressure pilots and the sand probes while sensors are being tested, calibrated, or replaced. For safety reasons, it should be very evident from a distance that the bypass valve has been switched by the use of panel-mounted indicators. Many facilities require an alarm when bypass switches are used. Sometimes individual bypass valves for each of the three sensors are used. A test connection in the field should be provided for checking the pressure sensors. Individual barstock root valves and test valves are normally used. When oil quality allows the pressure pilots to be panel-mounted, then the test connection can be panel-mounted also.

Chevron Corporation

1700-21

July 1999



Fig. 1700-16 Pilot Relay with First Out Indication (Courtesy of Amot Controls Corp.)

When a fault condition arises, the valve sensing that condition opens, causing a loss of pressure at the large end of the piston and allowing pressure on the small end to move the piston to the “Red” or tripped position. The OUT Port connects with the VENT Port through specially formed vent grooves, and all pressure downstream is released to the VENT as the IN Port is closed off from the OUT Port. This loss of pressure can be used to close fuel valves, actuate audible alarm devices or operate remote signal devices or switches. Any indications existing at that moment will be held indefinitely. The unique Red and White striped “Trip” tape, selected by optical specialists, can be clearly seen at a distance even in poor light and by those with impaired color vision. An operator can check the 4400 Relay panel at any time and tell immediately what caused the trouble.

Needle valves in the supply gas and hydraulic oil tubing runs should be provided to allow components to be replaced without requiring that an individual wellhead or all the wellheads be shut down. Some of the components which may require replacement are as follows: • • • •

Pressure regulators Pressure gages Indicators Hydraulic pumps and their regulators

1724 Panels for Wellhead Controls Several vendors specialize in building wellhead control panels. In order to be competitive, they will offer to supply the minimum number and lowest quality of logic and panel components. To get a panel that is reliable and maintainable, a drawing should be prepared that shows in general terms such things as the following. •

July 1999

Available space that the panel can occupy or the maximum acceptable panel dimensions

1700-22

Chevron Corporation



•

Accessibility to the panel for maintenance and the location of doors and bulkheads

•

General layout of the controls showing the wellhead control groupings, the location of subsurface controls for each well in relation to the surface controls, and the location of the hydraulic control unit and manual pump override

•

The minimum and maximum height of the controls from grade and how the panel is to be mounted

•

The amount of space required for expansion

•

Nameplate engraving details including the lines of text, the size of letters and nameplate, and color scheme, if any

•

Distances from the field devices to the panel

•

Indication requirements for such things as valve status and bypass switches

•

Bypass switch and test valve requirements

•

First out indication requirements for shutdown sensors

The drawing or an attached specification should include a list of the components and the acceptable manufacturers. To ensure quality and reliability, it is often necessary to include model numbers to avoid the problem when a vendor supplies the least expensive option. Model numbers also help in bid evaluation, during inspection of the finished panels, and to maintain standardized spare parts. Vendors should be allowed to offer reasonable substitutions in order to produce the lowest cost panel by avoiding unusual construction requirements.

1725 Enclosures for Wellhead Control Systems Enclosures are usually made of 316 stainless steel to withstand the corrosive environment of offshore platforms or the workover conditions around the wellheads. As a minimum, all the hardware must be made of 316 stainless steel. This includes hinges, latches, handles, bolts, and nuts. Enclosures may be made to withstand such things as windblown rain, sand, and dust; splashing water; hose-directed water; and spray from wellhead workover activities. External icing may also be a problem, but it is usually handled by the use of buildings or shelters. Specifications may be written to describe the enclosure construction required to withstand operating conditions, but the use of NEMA-type enclosures, developed for electrical equipment, can match the right cabinet or enclosure to the operating conditions. Some of the NEMA types are as follows. NEMA 13 — indoor use, protection against dust, spraying of water, oil, and noncorrosive fluids.

Chevron Corporation

1700-23

July 1999



NEMA 12 — indoor use, protection against dust, falling dirt, and dripping noncorrosive liquids. NEMA 3 — outdoor use, protection against windblown dust, rain and external icing. NEMA 3R — outdoor use, protection against falling rain, external icing, and only rust resistant NEMA 4 or 4X — outdoor use, protection against windblown dust and rain, splashing water, hose-directed water, and external icing. (The “4X” means corrosion-resistant, but “316 stainless steel” should be added to this description.) Following are some other important requirements for construction.

July 1999

•

Tubing runs between components should be designed to allow easy removal of components or in-place maintenance to replace “O” rings and gaskets.

•

Stacking or double layering of tubing runs should not be allowed.

•

Longer tubing runs should be clamped with solid stainless steel spacers.

•

All tubing should be reamed and blown clean with dry air before installation.

•

Pipe-to-tube fitting adapters should be used to connect components instead of pipe nipples and unions.

•

All penetrations or bulkhead fittings should be on the sides, back, or bottom. No penetrations should be allowed into the top of the enclosure.

•

No internal components should be mounted from the top or bottom of the enclosure.

•

When height or width exceeds 36 inches, then construction should be with at least 12-gage stainless steel. Otherwise 16-gage metal is satisfactory.

•

Gaskets should be made of oil- and water-resistant material such as neoprene.

•

Gaskets should be secured with an oil-resistant adhesive and supported by continuous stainless steel retainer strips.

•

The bottom should be sloped, with a 1/2-inch flush-mounted half-coupling for draining.

•

Lifting eyes with reinforced plates should be provided to support the entire weight of the completed panel for installation.

•

Mounting brackets or stands (including nuts and bolts) should be provided for small enclosures that are less than 48 inches tall.

•

Integral legs should be provided for free standing enclosures 48 inches or taller, including hardware to bolt the legs to the floor.

•

If earthquake requirements exist, brackets should be provided for securing tall enclosures.

1700-24

Chevron Corporation



•

Number and size of doors should be specified and should depend on the width of the enclosure. If it is wider than 36 inches, multiple doors are required, with a maximum width for each door of 30 inches.

•

If the type of door latch is not specified as NEMA 4X, three-point latches made of stainless steel are typically used.

•

Locks or other features should be provided for security.

•

Mounting and bracing of all internal components should be provided to prevent damage during shipment and installation and from operating conditions such as vibration. These supports and shelves are usually made from a corrosion-resistant material such as stainless steel.

•

All clips, clamps, straps, bolts, nuts, washers, screws, etc., should be made of 316 stainless steel, nylon, or some other durable, corrosion-resistant material.

•

It may be desirable to have a panel or even individual compartments to separate hydraulic systems from pneumatic systems.

•

When potentially hazardous or corrosive gasses are being handled, the design of the control system should prevent them from entering any panels or enclosures.

1726 Tagging and Nameplates Nameplates and tagging are very important for the factory checkout of the panel, the installation of the panel, safe operation of the controls, and maintenance troubleshooting of the wellhead control system. These are some important considerations:

Chevron Corporation

•

All internal components should be tagged with the instrument number or the identification number on the schematic drawing. If the component does not have permanent markings giving the make and model number, then this information should be added to the tag. Tags should be of stainless steel and wired or attached onto the components by a corrosion-resistant fastener. Panelmounted devices should have these tags attached to the device and visible from the inside of the panel.

•

All panel-mounted instruments including pressure gages, valves, indicators, pilot relays, etc., must have engraved Bakelite or plastic nameplates that clearly identify the device.

•

Engraved tags should have black letters in a white (or other color if a color scheme is used) background so that they can be easily cleaned. They should be attached with stainless steel screws.

•

Nameplates should be engraved from a nameplate schedule or front of panel drawing. They should include the tag number, a process function, and any operating instructions (e.g., “Pull To Reset”).

1700-25

July 1999



•

These tags are often color coded. Possible color groups are surface safety shutdown controls; subsurface safety controls including the hydraulic system; and bypass and ESD switches.

•

Panel title nameplate should have 3/4-inch letters at a minimum and should identify which wells in the area or wellbay are being controlled. The controls for each well should be clearly grouped and identified.

•

Each bulkhead connection should be identified with a stainless tag that is drilled and mounted under the bulkhead connection fitting. These tags should be mounted both on the outside and inside of the bulkhead. The tags should include the instrument number or schematic identification number.

•

Panel drawings, including a schematic and parts list, should be enclosed in a weatherproof envelope and put in a pocket in one of the doors. On the inside of the panel should be a stainless steel or phenolic tag with the supplier’s name, date of manufacture, vendor job number, Company purchase order number, and a list of the panel drawing numbers that could be useful to anyone troubleshooting the panel.

1727 Functional Requirements Project descriptions should detail the factory testing procedures, as well as who is responsible for the field installation of the panel, who is responsible for supplying the field devices, and who is responsible for the field commissioning of the panel connected to the field devices. A preliminary simplified schematic diagram should be prepared. The simplified schematic need only show the key devices in the control system, like the pressure regulators, pressure sensors, sand probes, pressure gages, SSV and SCSSV, manual reset pilot relays for the safety valves, and bypass switches for any sensors. For offshore installations, these requirements should be emphasized: •

Interlock for each well to make sure that the SSV cannot be opened before the hydraulic system is operational for opening the SCSSV valves

•

Interlock for each well to make sure the SCSSV closes two minutes after the SSV

•

Interlock to make sure the SCSSV closes only on an ESD or fire loop signal

•

Minimum capacity of the hydraulic reservoir, a level glass gage visible from the outside of the enclosure, and external fill and drain lines

•

When an SCSSV is opened, the design of the hydraulic system should prevent a pressure drop to any previously opened SCSSVs

•

Type of backup pump and manual override to the main hydraulic pump.

Optional functionality should then be described in text form. The description should include the following.

July 1999

1700-26

Chevron Corporation



•

Dual supply regulators and filters with isolating valves for each pressure system. (Each regulator should be capable of supplying the entire panel capacity.)

•

Pressure requirements for the system including pressure of the supply gas and a description (e.g., natural gas, air, or nitrogen); gas pressure used to operate the SSV; holding circuit pressure for the pilot relays, etc.

•

Number of bypass switches for the field sensors plus indication or alarm requirements for the bypass condition

•

List of types of panel devices that should be equipped with isolating valves so that they may be replaced without shutting down other wells

•

Size of the tubing and the wall thickness required for each pressure service

•

First out-type of indicating pilot relays

SCADA System Requirements Many wells are tied into a SCADA or production information system. Requirements for these electrical tie-ins should be reviewed during the panel design to make sure they are incorporated into the design, construction, and testing of the panel and field devices such as the wellhead SSV valves. Information sent to SCADA system usually includes: •

Open and closed status of wellhead valve from position or proximity switches

•

Operation of bypass switches around flowline pressure sensors

1730 References

Chevron Corporation

1.

API RP 14B, Recommended Practice for Design, Installation, and Operation of Subsurface Safety Valve Systems.

2.

API RP 14C, Recommended Practice for Testing of Basic Surface Safety System on Off-shore Production Platforms.

3.

API Spec 14D, Specification for Wellhead Surface Safety Valves and Underwater Safety Valves for Offshore Service.

4.

API RP 14F, Recommended Practice for Design and Installation of Electrical Systems for Offshore Platforms.

5.

Department of Interior, Minerals Management Service (MMS), OCS Order No. 5, Subsurface Safety Devices.

6.

Chevron Overseas Petroleum General Specifications GS 11.08-1, Alarm Systems.

7.

Chevron Overseas Petroleum Design Practice DP 11.08-1, Wellhead Controls.

8.

NEMA Standards Publication No. 250.

1700-27

July 1999

Appendix A. Orifice Design by Mainframe Computer

Abstract This appendix explains how to use the mainframe computer program named ORIFICE for orifice plate design. The appendix also provides the fundamental and working equations used and the limitations in calculations performed by the program. The remainder of this appendix is organized as follows. •

Section A1.0 describes what the program can do: the types of calculations it can perform, the kinds of orifice plates and taps it can be used to design, and the fluid flows and limits it can work with are discussed.

•

Section A2.0 is a “user’s guide.” It explains logon procedure, input data files, output reports, and error messages and provides examples of screen prints.

•

Section A3.0 documents the fundamental equations, constants, and working equations used in the program. It discusses orifice coefficients and their sources and limitations in order to help engineers better understand the basis of the calculations the program performs. Correction factors used in the calculations are also described in this section.

•

Section A4.0 provides test cases for users who want to test their PC-based orifice calculations against the mainframe ORIFICE.

•

Section A5.0 provides a list of technical references.

Contents

Chevron Corporation

Page

A1.0

Background Information on Orifice

A-2

A2.0

User’s Guide: Running the ORIFICE Program

A-6

A3.0

Equations, Coefficients, and Correction Factors

A-21

A4.0

Test Cases with Results by Orifice

A-30

A5.0

References

A-30

A-1

July 1996

Appendix A

A1.0


Background Information on Orifice A1.1

Background The orifice design calculations performed by ORIFICE are suitable for flow meters in process plants and producing facilities where the orifice meters do not require accuracy better than + 2% (e.g., plant applications). Although it provides design and flow rate calculation for orifice meters in gas service, ORIFICE is not suitable for natural gas custody transfer calculations. This is because the program does not have the provision, at present, to perform flow rate calculation using the current API MPMS, Chapter 14.3 (also published as AGA 3). The orifice coefficients (which is probably the most important empirical factor in orifice calculations) used in the program are taken from the following sources:

A1.2

•

The Stolz equation from ISO Standard 5167 for square edge orifice with flange, radius and corner taps.

•

An unpublished equation by Stolz for square edge orifice with pipe taps (Pipe taps not endorsed by ISO-5167).

•

Equations in “ASME Fluid Meter Applications” (6th Edition, Para II-III-15 to 19, Figure II-III-13) for small bore orifice meters.

•

Royal Dutch Shell (Delft, The Netherlands) Report 1312M used by L. K. Spink in his “Principles and Practice of Flow Meter Engineering” (9th Edition, 1972), Table 15.

Calculations ORIFICE performs three kinds of calculations:

July 1996

1.

Sizing of a new orifice by given standard differential pressure and full scale flow-rate.

2.

Re-ranging of an existing orifice by given orifice diameter and full-scale flowrate.

3.

Calculation of flow-rate (full-scale) by given differential pressure and orifice diameter.

A-2

Chevron Corporation


A1.3

Appendix A

Orifice Plates The following three types of orifice plates, which are commonly used in the United States, are included in the ORIFICE program. 1.

Square-edge orifice plates are the most commonly used for clean liquids, gases and low-velocity vapor (steam) flows in 2-inch or larger pipes.

2.

ASME small-bore orifice plates can be used for liquid, vapor (steam) or gas, usually with 1/2 to 1 1/2-inch (nominal) pipe size depending upon the type of taps used.

3.

Quadrant-edge orifice (or “quarter-circle”) plates are used for viscous liquids only. They are often used where the pipe Reynold’s number is low (usually below 10,000).

Conic-edge and integral orifice plates are not covered in the ORIFICE Program. Conic-edge orifice plates may have lower limit on the minimum Reynold’s number. They are more common in Europe, where they are used to replace quadrant-edge orifice meters. Integral orifices are used for clean fluids where the pipe size is 1/2 inch or smaller. Usually the supplier provides the method for performing sizing calculations.

A1.4

Orifice Taps Depending on the location of upstream and downstream pressure taps, the flow meter is referred to as a flange tap, a corner tap, a radius tap, a pipe tap, or a vena contracta tap orifice meter. Corner and radius taps are widely used in Europe; flange taps predominate in the United States. Pipe taps are sometimes used as bypass pump restrictors for natural gas, or where the other tapping arrangements require drilling too close to the plate. Vena contracta taps have been replaced in many places by radius taps because future changes in orifice bore will then require no tap relocation. The ORIFICE Program covers four types of pressure taps, defined as follows.

Chevron Corporation

1.

Flange taps: 1 inch upstream and 1 inch downstream of the orifice plate.

2.

Radius taps: 1 D (inside pipe diameter) upstream and 1/2 D downstream.

3.

Corner taps: Front and rear face of orifice plate.

4.

Pipe taps: 2 1/2 D upstream and 8 D downstream.

A-3

July 1996

Appendix A

A1.5


Fluid Flows ORIFICE is an orifice design calculation program for single-phase flow. Three kinds of fluid flows and their engineering units are as follows:

A1.6

1.

Liquid: Flow-rate units can be in barrels per day (BPD), barrels per hours (BPH), gallons per minute (GPM), or gallons per hour (GPH), all corrected to 60°F.

2.

Vapor (steam): Flow-rate unit should be in pounds per hour (PPH).

3.

Gas: Flow-rate units should be in standard cubic feet per hour (SCFH), i.e., at 60°F and 14.73 psia, which are the base conditions used by AGA-3.

Base Conditions Note that the European base, which is 59°F and 14.696 psia, is used in R. W. Miller’s book Flow Measurement Engineering Handbook (see Section A5.0, “References”). The definitions of “base conditions” are discussed in the API MPMS, Chapter 14.3, Part 1 (1990): “Historically, the flow measurement of some fluids, such as custody transfer and process control, have been stated in volume units at base (reference or standard) conditions of pressure and temperature. The base conditions for the flow measurement of fluids, such as crude petroleum and its liquid products, whose vapor pressure is equal to or less than atmospheric at base temperature are defined in the United States as pressure of 14.696 pounds per square inch absolute (101.325 kilopascals) at a temperature of 60.0°F (15.56°C). According to the International Standards Organization, base conditions are defined as a pressure of 14.696 pounds per square inch absolute (101.325 kilopascals) at a temperature of 59.00°F (15.00°C) For fluids, such as liquid hydrocarbons, whose vapor pressure is greater than atmospheric pressure at base temperature, the base pressure is customarily designated as the equilibrium vapor pressure at base temperature. The base conditions for the flow measurement of natural gases are defined in the United States as a pressure of 14.73 pounds per square inch absolute (101.560 kilopascals) at a temperature of 60.0°F (15.56°C). According to the International Standards Organization, base conditions are defined as a pressure of 14.696 pounds per square inch absolute (101.325 kilopascals) at a temperature of 59.00°F (15.00°C). For both liquid and gas applications, these base conditions can change from one country to the next, one state to the next, or one industry to the next. Therefore, it is necessary that the base conditions be identified for standard volumetric flow measurement.”

July 1996

A-4

Chevron Corporation


A1.7

Appendix A

Limitations The calculations in ORIFICE are based on the latest applicable industry standards and technical reports, which contain certain limits. These limits have been generalized for incorporation into the program (see Section A3.0, “Equations, Coefficients, and Correction Factors.”). The limits on pipe size (D), beta ratio (β, i.e., the ratio of orifice bore to inside pipe diameter) and pipe Reynold’s number (RD) are outlined below by type of orifice meter and type of tap. The asterisk (*) denotes that an error message (Section A2.5, “Error Messages”) will be displayed on the screen to inform the user that a limit has been exceeded. 1.

Square-Edge Orifice

Flange, Radius, and Corner Taps D:

2-in. to 36-in. pipe size D min. = 1.939 in. (2 in. Sched. 80 )* D max. = 35.5 in. (36 in. Sched. XS)

β

0.2 - 0.6 will have a min. error of ±0.6 % on the value of orifice discharge coefficient (C). 0.6 - 0.75 will have a min. error of ±β % on the value of orifice discharge coefficient (C). Min.= 104*

RD:

Max.= 107 Note The actual limitation on beta ratio in the ISO Standard varies depending upon the type of tap. The ISO Standard also limits the minimum orifice diameter to 12.5 mm (about 0.5 inches) and limits the maximum pipe Reynold’s number to 108. Pipe Taps No minimum or maximum limits are set on pipe size (D), beta ratio (β), and RD. In general, except for Dmin, the criteria on limits described above apply to square-edge orifice plates with pipe taps. 2.

ASME Small-Bore Orifice

Corner taps D:

0.5-in. to 1.5-in. pipe size D min.= 0.50 in.* D max.= 1.77 in. (1-1/2-in. 5S pipe size)*

Chevron Corporation

β:

0.1 - 0.8

RD:

>103*

A-5

July 1996

Appendix A


Flange Taps D:

1-in. to 1.5-in. pipe size D min.=1 in.* D max.=1.77 in. (1-1/2-in. 5S pipe size)*

β:

0.15 - 0.7

RD:

>103*

Radius taps and pipe taps are not allowed by the program when adopting ASME small-bore calculations. 3.

Quadrant-edge Orifice

Only flange taps are allowed by the program. D:

1-in. minimum pipe size*

β:

0.2 - 0.6 *. Within this range, the expected minimum error is +2%.

RD at 1/3 maximum (full scale) flow-rate: has an allowable range vs. beta ratio*

Beta Ratio

RD @ 1/3 Minimum

Maximum Flow-rate Maximum

0.2

670

17500

0.3

770

28000

0.4

630

45000

0.5

450

70000

0.6

320

85000

In any event, RD at 1/3 maximum (full-scale, or FS) flow-rate should not exceed 90,000*.

A2.0

User’s Guide: Running the ORIFICE Program A2.1

Accessing the ORIFICE Program in VM ORIFICE is a stand-alone engineering program in CRTC’s ENGR/MFG TECHNICAL LIBRARY. User can run ORIFICE directly from VM, or through the Plant Equipment Information System (PEIS). PEIS’s ORIFICE program provides different front-end (user-interface) from the VM ORIFICE program, but it shares the same calculations as the VM ORIFICE program.

July 1996

A-6

Chevron Corporation


Appendix A

Running ORIFICE on VM User can run ORIFICE by the following steps: 1.

Log on to one of the VM hosts, such as HOVMA.

2.

When the ready/time prompt is shown on the screen, type: ATT T9ENGR This entry can be in either upper or lower case.

3.

When the main menu of the ENGR/MFG TECHNICAL LIBRARY is displayed on the screen (Figure A-1), type: ORIFICE

Fig. A-1

Accessing ORIFICE via CRTC’s Technical Library

R; T=0.09/0.19 14:45:38 .attach t9engr DMSACC7231 B (102) R/0 T9ENGR attached as B DISK (XNOMAD2) —-ENGR/MFG TECHNICAL LIBRARY FOR VM PROGRAM LIST: FOR BATCH PROGRAM LIST: FOR SYSTEM INFORMATION: FOR ENGR/MFG ASSISTANCE:

Hit CARRIAGE RETURN KEY after next question Enter “BATCH” after next question Enter “INFO” after next question Enter “HELP” after next question

* Enter PROGRAM NAME or “QUIT” >

Running ORIFICE through PEIS At the time of this revision, PEIS - developed by Chevron Information Technology Company (CITC), is expected to be replaced by the Meridium system in AFIS (1996). PEIS users should contact their PEIS contact in CITC about the log on procedure of the ORIFICE subsection in PEIS. Note that PEIS customized some displays and reports for certain locations. Therefore, the screen displays and reports through PEIS/ORIFICE may be different from those described in this manual, which is based on the generic version of ORIFICE in VM.

A2.2

Running ORIFICE Once you are in ORIFICE, the program will guide you throughout. No step-by-step procedure is needed, but the following hints should be useful. 1.

Chevron Corporation

Always remember that you can “walk back” by “quitting.” Use Function Key PF3 or F3 to quit. “Quit” takes you back to the beginning of the program, or

A-7

July 1996

Appendix A


takes you out of the program, or takes you out of the Technical Library, depending upon where you are when you quit. Function Key PF2 or F2 refreshes the screen and brings you back to the beginning of the input data file.

A2.3

2.

When using a personal computer, find out how to use the key(s) to perform the function of certain VM terminal function keys that are not available to you. For example: combination of the CTRL and HOME keys on an IBM-PC keyboard will refresh the screen. The carriage return key on an IBM-PC provides the same function as the ENTER key on a VM terminal.

3.

The METER NUMBER in the input data file is your file name for the orifice meter. The file will be saved under this METER NUMBER if you press the ENTER (or the carriage return) key at any time. This is because the default automatically saves the file for your protection. Consider using the meter identification number, meter tag number, or instrument number assigned by the plant as the METER NUMBER (File Name).

4.

When you type in a file name, a routine in the ORIFICE program will check to see if you already have an existing orifice meter with this file name (METER NUMBER) under your VM user I.D. If this is the case, it will retrieve the existing file and display the file on screen.

Input Data General

July 1996

•

You may move the cursor to work on a new meter, or call up an existing meter number and modify it, or quit and leave the program. Just move the cursor to your choice and press the ENTER (or the carriage return) key.

•

If you chose to create a new file (new meter), you will first see the upper half of the screen. Once you provide the necessary information (i.e., kind of fluid flow and type of calculation desired), ORIFICE will display the lower half of the screen tailored for your application.

•

If you chose to modify an existing file, the design data for the meter will be displayed.

•

Use tab key or arrow keys to move the cursor. Sometimes you may wish to skip an entry and go back to it later. You may move the cursor to do this. Once you press the ENTER (or the carriage return) key, the information you have typed in will be stored under the METER NUMBER (i.e., the file name for this input data file), and ORIFICE will attempt to execute with its default values.

•

Note that the default on “Saving Input data in File?” is YES, as you can see at the bottom of the lower screen. When you press the ENTER (or the carriage return) key during editing, ORIFICE will try to save the file. You will be asked to press the key again to override the previous version of the file.

A-8

Chevron Corporation


Appendix A

Explanation of Input Data Figures A-2 through A-4 are examples of input screens. Fig. A-2

Main Menu ORIFICE DESIGN CALCULATIONS Enter new meter data Modify existing meter data Quit F1=HELP

Fig. A-3

F2=RESTART

F3=QUIT

Second (lower) Half of Screen Tailored for Calculation Defined in First (upper) Half of Screen

ORIFICE DESIGN CALCULATIONS Orifice Meter No.: FIGURE9

Date: 08/20/87

Plant/Location: Orifice Type: Q Tap Type: F Calculation Type: Service:

Fluid Type: L Material Type: 316SS Pipe Inside Diameter (INCHES): 10.20

Full Scale Flow Rate: 2000.00

Units: BPH

Normal Flow Rate: Flow Temperature (DEGF): 335. Sp. Grav. (flow): 0.8107 Kinematic Viscosity: 3.00000

Sp. Grav. (60F): 0.9088

Save input date in file: Y

Make calculations? Y

F1=HELP F2-RESTART F3=QUIT Fig. A-4

Example of a Help Screen: Use PF1 (or F1) Key to See Choices for Certain Parameters *****Help Table for TAP TYPE***** C - CORNER F - FLANGE P - PIPE R - RADIUS Table Search: Current Value: F F3=Quit F7=Bwd F8=Fwd F9=Top F10=Bot

Chevron Corporation

A-9

July 1996

Appendix A


The input data are explained below. For those items with an asterisk (*), the information is required; no default value is available. *Meter Number

Up to 8 characters (alphabetics or numbers, or both). As mentioned earlier, this is the permanent file name for the meter data.

Date

Any date can be entered. Default is today.

Plant Name and Location

Up to 40 characters. Any alphabetic description of the plant in which the meter is located. This entry is optional.

*Fluid Type

L

=

Liquid - uses volume rate of 60°F liquid.

V

=

Vapor (steam) - uses pounds per hours of vapor (steam).

G

=

Gas - uses standard cubic feet rate of gas at 60°F and 14.73 psia.

*Calculation Code OD =

Given pressure differential, find orifice diameter.

PD =

Given orifice diameter, find pressure differential.

FR =

Given orifice diameter and pressure differential, find full-scale flow-rate.

S

=

Square-edge - for RD at 1/3 F.S. flow < 10,000.

Q

=

Quadrant-edge - for RD at 1/3 F.S. flow < 90,000. Quadrant-edge orifice should not be used for either vapor (steam) or gas.

P

=

Program Choice - based on program calculation of RD at 1/3 full-scale flow. Program Choice is only available for re-ranging pressure differential or flowrate calculation.

F

=

Flange taps

C

=

Corner taps

R

=

Radius taps

P

=

Pipe taps

*Orifice Type

*Tap Type

*Correction Desired

Correction for liquid flow is not required. Choose N (= No correction), or default. Correction for vapor or gas: N

=

No correction

Y1 =

Upstream tap correction

Y2 =

Downstream tap correction

Note that the correction factor (Y) is for gas expansion at midscale flow-rate. *Material Type

316SS = 316 stainless steel 304SS = 304 stainless steel

July 1996

A-10

Chevron Corporation


Appendix A

OTHER = Other materials - user must provide Fa (Thermal expansion correction factor for steel) later. Service Information

Up to 16 characters to describe what service the orifice meter is for. This entry is optional.

*Pipe Inside Diameter

Inches to 3 decimal places. For limits on pipe size for various types of orifices and taps, see Section A1.7, “Limitations.”

*Orifice Diameter Inches to 3 decimal places. Mandatory information for PD or FR calculation for squareedge orifice. *Pressure Differential

Inches of water pressure differential across orifice taps. Mandatory information for OD or FR calculation for squareedge orifice.

*Full Scale Flow-rate

Up to 9 numbers for maximum flow-rate. Mandatory information for PD or OD calculation.

Normal Flow-rate

Up to 9 numbers for normal flow-rate. Default value is 0.707 of full-scale flow-rate.

*Flow-rate Units

Liquid: BPH = barrels per hour BPD = barrels per day GPM = gallons per minute GPH = gallons per hour Vapor (Steam): PPH = pounds per hour Gas:SCFH = standard cubic feet per hour

*Liquid Specific Gravity at Flow Temperature

Dimensionless numerical value for liquid flow only.

Liquid Specific Gravity at 60°F

Dimensionless numerical value for liquid flow only.

Viscosity

For use in RD calculation. Liquid - Kinematic viscosity in centistokes (Cs). Default value = 1.0 Cs (20°C water). Vapor (Steam) - Absolute viscosity in centipoise (Cp). Default value = 0.014 Cp (50 psig saturated steam). Gas - Absolute viscosity in centipoise (Cp). Default value = 0.011 Cp (low pressure natural gas @ 60°F).

Chevron Corporation

Vapor (Steam) Specific Volume

Cubic feet/pound for vapor (steam) flow only. Default valve is calculated by the program based on ASME Steam Tables (1987) for saturated and superheated steam.

*Gas Specific Gravity

Ideal specific gravity (G). Dimensionless, numerical value for gas flow only. Note that:

A-11

July 1996

Appendix A


G

=

MWgas /MWair = MWgas /28.963

=

GR x Z b

or G

where: GR =

real specific gravity and ratio of density of gas to air

Zb =

compressibility of gas at 60°F.

*Pressure

Absolute pressure in psia for vapor (steam) or gas only.

*Temperature

Temperature of flow in degrees F.

Ratio of Specific Heats

For vapor (steam) or gas only. The ratio (k) is defined as follows: k

=

Cp /Cv

where: Cp and Cv = specific heats. k is used in Y-correction factor calculation. Default value is 1.3. Thermal Expansion Correction (Fa)

This dimensionless factor assumes that orifice was made in shop at 68°F and operated at flow temperature specified in the input data. No input is required for 316SS or 304SS because ORIFICE calculates the Fa for them. You must enter a numerical value when selecting OTHER material.

Correction for Supercompressibility (Fpv)

Numerical value required only for gas flow. Note that:

Fpv =

(1/Zf)1/2

where: Zf

=

Compressibility of gas at flow conditions.

Default value = 1.0 Correction Factor You must specify if the correction is for upstream (Y1) or for Gas Expansion downstream (Y2). (Y) Note: ISO/Miller uses differential pressure to calculate Y. ORIFICE uses mid-scale flow rate to calculate Y. Gas Base Compressibility Factor (Zb)

July 1996

Numerical value required only for gas flow at base conditions (i.e., 60°F and 14.73 psia). Default value is 1.0.

A-12

Chevron Corporation


Appendix A

Input During Program Execution Input during program execution is required in two situations: 1.

For quadrant-edge orifice calculation

For the PD or OD calculations, the program displays a chart with several lines of calculation results. Each line is for a different value of quadrant-edge orifice thickness (Figure A-3). The user is instructed to “Select an orifice thickness and type in value.” The user may enter a value for orifice thickness or press the ENTER (or the carriage return) key for the alternate option: “Select meter differential and type in value.” User may enter a value for meter differential, or again press the ENTER (or the carriage return) key to repeat the original choice. For the FR calculation, the user is required to enter pressure differential and plate thickness or to enter orifice diameter and plate thickness. ORIFICE will then calculate directly the full-scale flow-rate. 2.

For ASME small-bore orifice calculation

If the pipe size falls below the ASME/Miller limit for square-edge orifices, the program will ask if you want to use the ASME small-bore calculation. Fig. A-5

Screen Print of Chart for Quadrant-Edge Orifice Calculation

Orifice Thickness

Orifice Diameter

Beta

Meter Press Diff

Reynolds At 1/3 Flow Min Reyn Max Reyn

0.2500 0.3125 0.3750

2.4758 3.0050 3.4802

0.247084 0.299897 0.347323

2738.099 1248.376 685.493

755 770 723

21738 27988 35098

0.5000 0.7500 1.0000

4.2474 5.2249 5.7415

0.423890 0.521448 0.573001

296.650 117.051 73.089

582 420 352

50734 73861 81760

1.2500 1.5000 1.7500

6.0099 6.1518 6.2505

0.599792 0.613955 0.623804

57.583 50.869 46.594

320 309 301

84975 85837 86428

Calculation options: D - Calculate Orifice Thickness and Orifice Diameter given Meter Differential T - Calculate Meter Differential and Orifice Diameter given Orifice Thickness Please enter the calculation type (D or T): VM READ HOVMA

A2.4

Output Reports Once the program completes its calculations, you may store the output under your VM personal ID permanently, and/or display it on the terminal. Figure A-4 is an example of the screen showing your options at this point.

Chevron Corporation

A-13

July 1996

Appendix A


Fig. A-6

Output Menu Displayed on Screen When Calculation Complete Report Options Enter the report option

F1=Help

I.

Save input data

P.

Print output report

S.

Save output report

V.

View outut report

X.

Exit

F3=Exit

F12=Cancel

Enter=Proceed

Figures A-7 through A-13 are examples of output reports. Fig. A-7

Sizing Square-Edge Orifice in Liquid Service

ORIFICE DESIGN CALCULATIONS

07/07/95

METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN PRESSURE DIFFERENTIAL, FIND ORIFICE DIAMETER TAP TYPE: FLANGE FLUID TYPE: LIQUID ORIFICE TYPE: SQUARE EDGE

: : : :

FULL SCALE FLOWRATE: NORMAL FLOWRATE: PIPE INSIDE DIAMETER: PRESSURE DIFFERENTIAL: SG AT FLOW TEMP: SG AT 60 DEG F: KINEMATIC VISCOSITY: TEMPERATURE: THERMAL EXPANSION (FA):

FE-001 SAMPLE-LIQUID SS

2000 1414 10.0200 100.00 0.8107 0.9088 1.0000 335 1.0049

BPH BPH INCHES INCHES OF WATER

CENTISTOKES DEG F

CALCULATED VALUES ---PIPE REYNOLDS NO. AT NORMAL FLOW: PIPE REYNOLDS NO. AT 1/3 MAX FLOW:

350241 165130

ORIFICE DIAMETER ORIFICE COEFFICIENT BETA:

July 1996

6.1577 0.2472 0.6145

A-14

INCHES

Chevron Corporation


Fig. A-8

Appendix A

Re-Ranging Existing Orifice Meter


07/07/95

METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN ORIFICE DIAMETER, FIND PRESSURE DIFFERENTIAL TAP TYPE: FLANGE FLUID TYPE: LIQUID ORIFICE TYPE: SQUARE EDGE

: : : :

FULL SCALE FLOWRATE: NORMAL FLOWRATE: PIPE INSIDE DIAMETER: ORIFICE DIAMETER: SG AT FLOW TEMP: SG AT 60 DEG F: KINEMATIC VISCOSITY: TEMPERATURE: THERMAL EXPANSION (FA):


2000 1414 10.0200 6.1570 0.8107 0.9088 1.0000 335 1.0049

BPH BPH INCHES INCHES

CENTISTOKES DEG F


350241 165130

PRESSURE DIFFERENTIAL ORIFICE COEFFICIENT BETA:

100.05 0.2471 0.6145

Chevron Corporation

A-15

INCHES OF WATER

July 1996

Appendix A

Fig. A-9


Calculating Orifice Diameter and Pressure Differential for Quadrant-Edge Orifice


07/07/95

METER NUMBER PLANT/LOCATION, QUADRANT EDGE SERVICE ORIFICE MATERIAL GIVEN ORIFICE THICKNESS, CALCULATE METER DIFFERENTIAL AND ORIFICE DIAMETER TAP TYPE: FLANGE FLUID TYPE: LIQUID ORIFICE TYPE: QUADRANT EDGE - FLUID TYPE MUST BE LIQUID

: : : :

FULL SCALE FLOWRATE: NORMAL FLOWRATE: PIPE INSIDE DIAMETER: SG AT FLOW TEMP: SG AT 60 DEG F: KINEMATIC VISCOSITY: TEMPERATURE: THERMAL EXPANSION (FA):


2000 1414 10.0200 0.8107 0.90887 3.0000 335 1.0049

BPH BPH INCHES

CENTISTOKES DEG F

CALCULATED VALUES ----

PIPE REYNOLDS NO. AT NORMAL FLOW: PIPE REYNOLDS NO. AT 1/3 MAX FLOW:

116747 55043

THICKNESS: ORIFICE DIAMETER: BETA: METER DIFFERENTIAL:

0.7500 5.2249 0.5214 117.05

MIN. REYNOLDS NO. AT 1/3 RATE: MAX REYNOLDS NO. AT 1/3 RATE:

July 1996

INCHES OF WATER

420 73861

A-16

Chevron Corporation


Appendix A

Fig. A-10 Re-Ranging Pressure Differential for Orifice Meter in Vapor (Steam) Service ORIFICE DESIGN CALCULATIONS

07/07/95

METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN ORIFICE DIAMETER, FIND PRESSURE DIFFERENTIAL TAP TYPE: FLANGE FLUID TYPE: STEAM ORIFICE TYPE: SQUARE EDGE

: : : :

FE-005 SAMPLEVAPOR/STEAM SS

FULL SCALE FLOWRATE: NORMAL FLOWRATE: PIPE INSIDE DIAMETER: SPECIFIC VOLUME: ORIFICE DIAMETER: PRESSURE: TEMPERATURE: THERMAL EXPANSION (FA): VISCOSITY:

3000 2121 3.0680 3.4979 1.5780 135 380 1.0058 0.0140

PPH PPG INCHES CU. FT./LB. INCHES PSIA DEG F CENTIPOISE


311889 147048

PRESSURE DIFFERENTIAL: ORIFICE COEFFICIENT: BETA: GAS EXPANSION FACTOR, Y:

99.08 0.1659 0.5143 1.0000

GAS EXPANSION FACTOR, Y1

= 0.9956

PRESSURE DIFFERENTIAL: ORIFICE COEFFICIENT: BETA: GAS EXPANSION FACTOR, Y:

99.97 0.1659 0.5143 0.9956

Chevron Corporation

A-17

INCHES OF WATER

(K = 1.30000) INCHES OF WATER

July 1996

Appendix A


Fig. A-11 Sizing Square-Edge Orifice in Gas Service ORIFICE DESIGN CALCULATIONS METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN PRESSURE DIFFERENTIAL, FIND ORIFICE DIAMETER TAP TYPE: FLANGE FLUID TYPE: GAS ORIFICE TYPE: SQUARE EDGE

07/10/95 : : : :

FE-006 SAMPLE-GAS SS

FULL SCALE FLOWRATE: NORMAL FLOWRATE: PIPE INSIDE DIAMETER: SPECIFIC GRAVITY: PRESSURE DIFFERENTIAL: PRESSURE: TEMPERATURE: THERMAL EXPANSION (FA): FPV: ZB: VISCOSITY:

400000 282843 10.2500 0.6000 100.00 60 60 0.9999 1.0000 1.0000 0.0110

SCFH SCFH INCHES INCHES OF WATER PSIA DEG F

CENTIPOISE


727289 342847

GAS EXPANSION FACTOR, Y: ORIFICE DIAMETER: ORIFICE COEFFICIENT: BETA:

1.0000 4.3972 0.1126 0.4290

GAS EXPANSION FACTOR, Y = Y1 = ORIFICE DIAMETER: ORIFICE COEFFICIENT: BETA:

0.9902 4.4179 0.1137 0.4310

July 1996

A-18

INCHES

(K = 1.30000) INCHES

Chevron Corporation


Appendix A

Fig. A-12 Calculating Flow-rate for Square-Edge Orifice Meter in Liquid Service ORIFICE DESIGN CALCULATIONS

07/07/95

METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN PRESSURE DIFFERENTIAL AND ORIFICE DIAMETER, FIND FLOW RATES TAP TYPE: FLANGE FLUID TYPE: LIQUID ORIFICE TYPE: SQUARE EDGE

: : : :

PIPE INSIDE DIAMETER: PRESSURE DIFFERENTIAL: ORIFICE DIAMETER: SG AT FLOW TEMP: SG AT 60 DEG F: KINEMATIC VISCOSITY: TEMPERATURE: THERMAL EXPANSION (FA):


10.0200 100.00 6.1569 0.8107 0.9088 1.0000 335 1.0049

INCHES INCHES OF WATER INCHES

CENTISTOKES DEG F


350187 165080

FULL SCALE FLOWRATE: NORMAL FLOWRATE: ORIFICE COEFFICIENT: BETA:

1999 1414 0.2471 0.6145

Chevron Corporation

A-19

BPH BPH

July 1996

Appendix A


Fig. A-13 Calculating Flow-rate for Quadrant-Edge Orifice Meter ORIFICE DESIGN CALCULATIONS

07/07/95

METER NUMBER PLANT/LOCATION SERVICE ORIFICE MATERIAL GIVEN ORIFICE DIAMETER, CALCULATE FLOWRATES AND ORIFICE THICKNESS TAP TYPE: FLANGE FLUID TYPE: LIQUID ORIFICE TYPE: QUADRANT EDGE - FLUID TYPE MUST BE LIQUID

: : : :

FE-004 SAMPLE-LIQUID, QUADRANT EDGE SS

PIPE INSIDE DIAMETER: PRESSURE DIFFERENTIAL: SG AT FLOW TEMP: SG AT 60 DEG F: KINEMATIC VISCOSITY: TEMPERATURE: THERMAL EXPANSION (FA):

10.0200 100.00 0.8107 0.9088 3.0000 335 1.0049

INCHES INCHES OF WATER

CENTISTOKES DEG F


107925 50876

THICKNESS: ORIFICE DIAMETER: BETA: FULL SCALE FLOWRATE: NORMAL FLOWRATE:

0.7500 5.2249 0.5214 1849 1307

MIN. REYNOLDS NO. AT 1/3 RATE: MAX REYNOLDS NO. AT 1/3 RATE:

420 73861

A2.5

INCHES INCHES BPH BPH

Error Messages Three types of error messages are provided by ORIFICE: 1.

July 1996

Fatal Error - Calculation will be aborted when the program detects a fatal error. Fatal errors include: a.

Quadrant-edge and program choice orifices not used for flange taps.

b.

Quadrant-edge orifice not used for vapor or gas.

c.

Radius or pipe taps not used on ASME small bore orifice.

d.

Pipe diameter outside allowable range for ASME small bore orifice.

e.

Illegal entry used in input data.

f.

Orifice diameter calculation not specified with program choice.

g.

Unable to calculate beta - Check your input data.

h.

S = 0 - Check for input data error.

i.

Program Choice option used for calculation other than liquid calculation.

A-20

Chevron Corporation


2.

3.

Appendix A

j.

Error found during default calculation or specific volume: Temperature is below calculated saturation value. There is less than 100% vaporization. The program cannot calculate two phase flow.

k.

Error found during default calculation or specific volume. Pressure is outside ASME steam table range (0.17811 PSIA to 3206.16 PSIA).

l.

Invalid units specified for type of liquid.

Non-Fatal Error - Calculation permitted to continue. User is warned that certain limits have been exceeded. a.

Pipe diameter below ASME/Miller limits (2" Sched. 80 pipe) for squareedge orifice.

b.

RD (1/3 max. flow) below ASME/Miller limits for square-edge orifice.

c.

RD (1/3 max. flow) above limits (90000) for quadrant-edge orifice. Calculation performed for square-edge orifice instead.

d.

RD (1/3 max. flow) beyond limits for quadrant-edge orifice.

e.

Beta ratio beyond limits (0.2/0.63) for quadrant-edge orifice. Calculation performed for square-edge orifice instead.

Missing or Inappropriate Data ORIFICE checks for missing data that are essential for calculation. An error message will be displayed to inform the user that data are missing. When performing the quadrant-edge calculations, the user may need to choose the orifice thickness (t) or pressure differential (h). If the selection is not in the Quadrant-edge Orifice Sizing Data File, a message will be displayed to inform the user of the inappropriate selection made. The program will then ask the user to make another selection. The Quadrant-Edge Orifice Sizing Data File is a look-up table from which the program takes the orifice coefficient for a given plate thickness (t) and pipe size (D) or for a given beta ratio.

A3.0

Equations, Coefficients, and Correction Factors A3.1

Definition of Symbols C = Orifice coefficient of discharge K = C(1-β4)-1/2, orifice flow coefficient based on orifice diameter S = Kβ2, orifice flow coefficient based on pipe diameter Q = Volume flow-rate, full-scale BPD, BPH, GPM, GPH corrected to 60°F for liquid flow SCFH corrected to 60°F, 1 Atm for gas flow

Chevron Corporation

A-21

July 1996

Appendix A


V = Velocity of flow in ft/second W = Mass flow-rate, full-scale, in lbs./hour N = Conversion factor used in liquid orifice sizing equation for different units of flow-rate: BPD BPH GPM GPH ------------- ------------- ------------- ------------194.3 8.095 5.666 340.0 D = Pipe inside diameter in inches d = Orifice diameter in inches β = d/D, the diameter ratio h = Pressure difference at orifice taps in inches of water G = Ideal specific gravity of gas = MWgas /MWair = GR x Zb; MWair = 28.963 GR = Real specific gravity of gas = ρgas /ρdry air = G/Zb Temperature and pressure of gas and air are at standard conditions. Note that ORIFICE equations use Ideal Specific Gravity. Gb = Specific gravity of liquid at a base temperature of 60°F Gf = Specific gravity of liquid at the flow temperature P = Absolute pressure of flowing gas in psia T = Absolute temperature of flowing gas in °R °R = °F + 459.67 v = Specific volume of steam or vapor at flow conditions in cubic ft/lb υ = Kinematic viscosity at flow conditions in centistokes µ = Absolute viscosity at flow conditions in centipoise g = Gravitational conversion constant = 32.17405 ρ = Density of flowing fluid - pounds/cubic foot Rd = Reynold’s Number based on orifice diameter RD = Reynold’s Number based on pipe diameter RD = β Rd B = Conversion factor used in liquid Reynold’s Number equation for different units of flow-rate: BPD BPH GPM GPH ------------- ------------- ------------- ------------92.24 2214. 3163. 52.71 k = Ratio of specific heats = Cp /Cv

July 1996

A-22

Chevron Corporation


Appendix A

Correction Factors Fa = Correction for thermal expansion of orifice meter at operating temperature. Base temperature = 68°F. Y = Gas expansion correction factor for midscale flow Y1 = correction factor based on upstream static pressure Y2 = correction factor based on downstream static pressure Fpν = Supercompressibility correction factor Fpν = (1/Zf)1/2 Zf = Compressibility factor at flow temperature and pressure Zb = Compressibility factor at 60°F, 14.73 psia Fc = Reynold’s number correction factor; not used by ORIFICE. RD is built into the orifice flow coefficients used by ORIFICE.

A3.2

Fundamental Equations Consider an orifice with diameter d in a pipe with inside diameter D P1 = upstream absolute pressure (in pipe) P2 = downstream absolute pressure (at orifice) Assume incompressible flow, constant density ρ Valid for vapor (steam) and gas when (P1 - P2) < P1 P = static pressure (lbs/ft2) ρ = density (lbs/ft3) Energy Balance Equation 2

2

V2 P2 V1 P1 ------- + ------ = ------- + -----2g ρ 2g ρ (Eq. A-1)

V1 = velocity (ft/sec.) in pipe V2 = velocity (ft/sec.) at orifice g = gravitational constant = 32.2 ft/sec2 Continuity Equation A1 V1 = A2 V2 (Eq. A-2)

A1 = area of pipe cross section = πd2/4 A2 = area of orifice = πd2/4

Chevron Corporation

A-23

July 1996

Appendix A


Combining and Simplifying 2g ( P 1 – P 2 ) 1 ⁄ 2 1 ---------------------------V 2 = --------------------------ρ 4 1⁄2 (1 – β ) (Eq. A-3)

β = d/D The Ideal Orifice Equation A2 × V2 = CFS (Cubic Feet per Second) 2 2g ( P 1 – P 2 ) 1 ⁄ 2 πd ----------------------------CFS = -----------------------------ρ 4 1⁄2 4(1 – β )

1 --------------------------- = velocity of approach factor 4 1⁄2 (1 – β ) (Eq. A-4)

The effects of non-ideal flow through the orifice are summarized into a single coefficient that is obtained experimentally. C is the coefficient of discharge. The Real Orifice Equation 2 2g ( P 1 – P 2 ) 1 ⁄ 2 πCd ----------------------------CFS = -----------------------------ρ 4 1⁄2 4(1 – β )

(Eq. A-5)

There are two definitions for orifice meter coefficients: C K = --------------------------4 1⁄2 (1 – β ) Cβ 2 S = --------------------------4 1⁄2 (1 – β ) Note that S = Kβ 2

July 1996

A-24

Chevron Corporation


Appendix A

The Real Orifice Equation Becomes: π 2g ( P 1 – P 2 ) 1 ⁄ 2 CFS = Kd 2 --- ----------------------------ρ 4 (Eq. A-6)

or 1⁄2 2 π 2g ( P 1 – P 2 ) CFS = SD --- ----------------------------ρ 4

(Eq. A-7)

Chevron has adopted this form as a standard. All of the “working equations” for sharp-edge orifices and quadrant-edge orifices are developed from this equation.

A3.3

Working Equations 1.

Liquid Flow COMPARE: NSD 2 Fa ( hG f ) 1 / 2 Q = -------------------------------------------Gb (Eq. A-8) Miller Table 9.23 eq. f.

BQG b R D = ---------------DµGf (Eq. A-9) Miller Table 9.20 eq. f

2.

Vapor (Steam) Flow h 1/2 W = 358.9 SD 2 FaY  ---  v (Eq. A-10) Miller Table 9.23 eq. a

6.316W R D = ------------------Dµ (Eq. A-11) Miller Table 9.20 eq. g

Chevron Corporation

A-25

July 1996

Appendix A


3.

Gas Flow hP 1 / 2 Q = 7708 SD 2 F a YF pv Z b  --------   TG  (Eq. A-12) Miller Table 9.23 eq. 1

.4832 QG R D = ------------------------D µ Zb (Eq. A-13) Miller Table 9.20 eq. 1

A3.4

ORIFICE Coefficients 1.

Square-edge orifice: D ≥ 2", close-up taps

Stolz equation ISO-5167-1980(E), 7.3.2.1 Flange taps: L1 = L2 = 1/D Radius taps (D, D/2 taps): L1 = 1, L2 = .47 Corner taps: L1 = L2 = 0 · B = .09L 1 if L 1 < .4333 · B = .039 if L 1 ≥ .4333 – .75 C = .5959 + .0312β 2.1 – .184β 8 + 91.7061β 2.5 R D

+Bβ 4 ( 1 – β 4 ) -1 – .0337L 2 β 3 (Eq. A-14)

Cβ 2 S = -------------------------( 1 – β4)1 / 2 (Eq. A-15)

July 1996

A-26

Chevron Corporation


2.

Appendix A

Square-edge orifice: D = 1" or 1-1/2", close-up taps From “ASME Fluid Meters: Their Theory and Application” (6th Ed. 1971). Sections II-III, 15 to 19.

Corner taps: .0044 .0175 S = β 2 .5991 + ------------- +  .3155 + ------------- ( β 4 + 2β 16 )  D D  1000 .00052 .00116 +  ---------------- – .000192 +  .01648 – ---------------- ( β 4 + 4β 16 ) ------------------ D  ( R )1 / 2  D  D (Eq. A-16)

Flange taps: 1000 2 S = β .5980 + .458 ( β 4 + 10β 12 ) + ( .00087 + 0081β 4 ) ------------------( RD )1 / 2 (Eq. A-17)

Note The ORIFICE program permits users to use the Stolz equation in ISO-5167 to calculate orifice flow coefficient(s). An error message will be displayed and printed to remind users of limit violation (D < 2"). 3.

Square-edge orifice: Pipe taps .039β 4 91.7061β 2.5 C = .5959 + .461β 2.1 + .48β 8 + ---------------- + ---------------------------· 1 – β4 R D .75 (Eq. A-18)

Cβ 2 S = -------------------------( 1 – β4 )1 / 2 (Eq. A-19)

4.

Quadrant-edge orifice: Flange taps r ---- = f ( β ) D (Eq. A-20)

S = f(β) (Eq. A-21)

Tabulated in file DQUADDATA

Chevron Corporation

A-27

July 1996

Appendix A

A3.5


Correction Factors (1) Fa = Correction factor for thermal expansion of orifice meter. 2 F a = 1 + --------------4- ( α Pe – β 4 α P ) ( T – 68 ) 1–β (Eq. A-22) (Miller eq. 9.51)

The following values were used: β = .5 αPe = expansion coef. for 3166SS orifice place = .0000096. αP = expansion coef. for carbon steel pipe = .0000067.

(Miller Table B.4)

yielding: Fa = 1 + .0000196 (T - 68) Notes: Base temperature is 68°F. Equation valid for 304 or 316 orifice plate/carbon steel pipe. The simplification that β = 0.5 produces a negligible error (i.e., less than the dimensional tolerance of the orifice diameter). (2) Y = Correction factor for gas expansion at midscale flow-rate. Note The fundamental orifice equation assumes the flowing fluid to be incompressible; that is, it assumes that the fluid density remains constant while flowing through the orifice. This is normally true for non-flashing liquids and approximately true for gases and vapors (steam) when ∆P/P is small. The Y correction factor is used for vapor (steam) or gas when ∆P/P gets too large. (a.) Y1 - Based on pressure, P measured upstream of orifice. .5h x = -------------27.7P x Y 1 = 1 – ( .41 + .35β 4 ) --- (for flange, radius, corner taps) k (Eq. A-23)

July 1996

A-28

Chevron Corporation


Appendix A

x Y 1 = 1 – ( .333 + 1.145 ( β 2 + 7β 5 + 12β 13 ) ) --- (for pipe taps) k (Eq. A-24)

A3.6

Method for Quadrant-edge Orifice Design - Liquid Flow The method is based on a table of S and r/D versus β. (See Figure A-6) Also see Table 15 in Spink and File DQUAD in the ORIFICE program. The ORIFICE calculation uses standard plate thicknesses in inches: .0625, .109 (USS 12 ga.), .125, .141 (USS 10 ga.), .1875, .250, .3125, .375, .500, .750, 1.000, 1.250, 1.500, 1.750, 2.000 Normal calculation - Given t, Calculate h & d 1.

Select orifice plate thickness, t. (Use standard thicknesses)

2.

Compute r/D Quadrant radius r = t, by definition

3.

Look up β and S in File DQUAD DATA

4.

Calculate orifice diameter: d = βD (Eq. A-25)

5.

Calculate differential pressure across taps: 2 QG b   - h =  --------------------------------------- SND 2 Fa ( G f ) 1 / 2 

(Eq. A-26)

6.

If h does not have a desirable value (e.g., 20-200 in. H2O) then change thickness and repeat steps 2, 3, 4, and 5.

Note

Thicker orifice plates result in lower values of h.

Alternate Calculation - Given h, Calculate t & d 1.

Select differential pressure across taps, h. (Use standard value: 200., 100., 50., 25., 10., etc.)

2.

Calculate S: QG b S = --------------------------------------ND 2 Fa ( hG f ) 1 / 2 (Eq. A-27)

Chevron Corporation

A-29

July 1996

Appendix A


3.

Look up β and r/D in File DQUAD DATA

4.

Calculate orifice diameter: d=βD (Eq. A-28)

5.

Calculate orifice plate thickness: t = r = (r/D) D (Eq. A-29)

A4.0

Test Cases with Results by Orifice Figures A-15 through A-17 provide calculations for various test cases for liquid, vapor (steam) and gas. Users who want to test their PC-based programs can compare the accuracy of their calculations.

A5.0

References Miller, R. W. Flow Measurement Engineering Handbook (1st edition, 1983). International Standard (ISO) 5167. Measurement of Fluid Flow by Means of Orifice Plates, Nozzles and Venturi Tubes Inserted in Circular Cross-Section Conduits Running Full (1st edition, 1980). ASME MFC-3M-1984. Measurement of Fluid Flow in Pipes Using Orifice, Nozzle, and Venturi. API. Measurement of Petroleum Measurement Standards, Chapter 14, “Natural Gas Fluids Measurement,” Section 3, “Concentric, Square-Edged Orifice Meters (Third Edition, Sept. 1990); GPA designation: GPA 8185-90, Part 1; AGA designation AGA Report No. 3. ASME. Fluid Meters; Their Theory and Application (6th edition, 1971). Spink, L. K. Principles and Practice of Flow Meter Engineering (9th edition, 1967).

July 1996

A-30

Chevron Corporation


Appendix A

Fig. A-14 Quadrant-Edge Orifice(s) and Thickness Ratio (t/D) versus Beta BETA S t/D BETA S 0.200 0.0305 0.0198 0.420 0.1407 0.205 0.0321 0.0203 0.425 0.1443 0.210 0.0338 0.0209 0.430 0.1480 0.215 0.0355 0.0214 0.435 0.1518 0.220 0.0372 0.0220 0.440 0.1556 0.225 0.0390 0.0225 0.445 0.1595 0.230 0.0408 0.0231 0.450 0.1635 0.235 0.0426 0.0236 0.455 0.1675 0.240 0.0445 0.0242 0.460 0.1716 0.245 0.0464 0.0247 0.465 0.1758 0.250 0.0484 0.0253 0.470 0.1801 0.255 0.0504 0.0259 0.475 0.1844 0.260 0.0524 0.0264 0.480 0.1888 0.265 0.0545 0.0270 0.485 0.1933 0.270 0.0566 0.0276 0.490 0.1978 0.275 0.0587 0.0282 0.495 0.2024 0.280 0.0609 0.0288 0.500 0.2070 0.285 0.0631 0.0294 0.505 0.2120 0.290 0.0654 0.0300 0.510 0.2170 0.295 0.0677 0.0306 0.515 0.2220 0.300 0.700 0.0312 0.520 0.2270 0.305 0.0724 0.0318 0.525 0.2320 0.310 0.0748 0.0325 0.530 0.2370 0.315 0.0773 0.0331 0.535 0.2430 0.320 0.0798 0.0338 0.540 0.2480 0.325 0.0824 0.0344 0.545 0.2540 0.330 0.0850 0.0351 0.550 0.2600 0.335 0.0877 0.0358 0.555 0.2660 0.340 0.0904 0.0364 0.550 0.2720 0.345 0.0931 0.0371 0.565 0.2790 0.350 0.0959 0.0378 0.570 0.2850 0.355 0.0987 0.0385 0.575 0.2920 0.360 0.1016 0.0392 0.580 0.2980 0.365 0.1045 0.0400 0.585 0.3050 0.370 0.1075 0.0407 0.590 0.3120 0.375 0.1106 0.0415 0.595 0.3190 0.380 0.1137 0.0423 0.600 0.3260 0.385 0.1169 0.0431 0.605 0.3330 0.390 0.1201 0.0439 0.610 0.3410 0.395 0.1234 0.0448 0.615 0.3480 0.400 0.1267 0.0456 0.620 0.3560 0.405 0.1301 0.0465 0.625 0.3640 0.410 0.1336 0.0474 0.630 0.3720 0.415 0.1371 0.0483 0.635 0.3800

Chevron Corporation

A-31

t/D 0.0492 0.0501 0.0511 0.0520 0.0530 0.0540 0.0550 0.0561 0.0572 0.0583 0.0595 0.0607 0.0619 0.0632 0.0646 0.0660 0.0675 0.0691 0.0708 0.0725 0.0743 0.0762 0.0781 0.0801 0.0822 0.0845 0.0869 0.0895 0.0922 0.0950 0.0980 0.1010 0.1040 0.1080 0.1130 0.1190 0.1250 0.1320 0.1410 0.1520 0.1640 0.1780 0.1920 0.2050

July 1996


Appendix A Orifice Design by Mainframe Computer

Fig. A-15 Orifice Sizing Test Cases (Liquid) Fluid Type

Liquid

Orifice Type

Square Edge

Flow Data for all test cases Pressure Differential

h

Specific Gravity at flow temperature

Gf

Specific Gravity at 60 º F

Gb

100 inches of Water 0.782 0.973

Kinematic Viscosity

ν

Pressure

P

85 PSIA

Temperature

T

243 º F

Fa

1.0032

Coeeficient for thermal expansion

2.0555 Centistokes

ISO-5167-Bore Orifice Test No.

Purpose

Tap Type

Pipe I. D.

Orifice Diameter

Pipe Reynolds No.

Full Scale Flow Rate

Normal Flow Rate

at Normal Flow (F)lange, (R)adius, (C)orner, (P)ipe F, R, C, P

D

d

inches

inches

Orifice Flow Coefficient based on pipe diameter

RD

Q

0.707*Q

BPD

BPD

S

Orifice Coefficient

Beta Ratio

Notes

of Discharge C

ß

1a

Dmin = 2 inches

F

2.245 (2" -Sch 5S)

1.2348

30578

1739

1229

0.1947

0.6135

0.5500

1b

Dmin = 2 inches

R

2.245 (2" -Sch 5S)

1.2348

30571

1738

1229

0.1947

0.6135

0.5500

1c

Dmin = 2 inches

C

2.245 (2" -Sch 5S)

1.2348

30505

1735

1227

0.1943

0.6122

0.5500

1d

Dmin = 2 inches

P

2.245 (2" -Sch 5S)

1.2348

37020

2105

1488

0.2358

0.7430

0.5500

2a

Dmax = 36 inches

F

35.376 (36" - Sch 10)

19.4568

474699

425338

300760

0.1918

0.6043

0.5500

2b

Dmax = 36 inches

R

35.376 (36" - Sch 10)

19.4568

475637

426178

301354

0.1922

0.6056

0.5500

2c

Dmax = 36 inches

C

35.376 (36" - Sch 10)

19.4568

474623

425270

300711

0.1918

0.6043

0.5500

2d

Dmax = 36 inches

P

35.376 (36" - Sch 10)

19.4568

578125

518010

366288

0.2336

0.7361

0.5500

3a

ßmin = 0.2

F

15.67 (16" -Sch 5S)

3.1340

26227

10410

7361

0.0239

0.5970

0.2000

3b

ßmin = 0.2

R

15.67 (16" -Sch 5S)

3.1340

26225

10409

7360

0.0239

0.5970

0.2000

3c

ßmin = 0.2

C

15.67 (16" -Sch 5S)

3.1340

26228

10410

7361

0.0239

0.5970

0.2000

3d

ßmin = 0.2

P

15.67 (16" -Sch 5S)

3.1340

26872

10665

7542

0.0245

0.6120

0.2000

4a

ßmax = 0.75

F

15.67 (16" -Sch 5S)

11.7525

446564

177239

125327

0.4074

0.5988

0.7500

4b

ßmax = 0.75

R

15.67 (16" -Sch 5S)

11.7525

453714

180077

127334

0.4140

0.6085

0.7500

4c

ßmax = 0.75

C

15.67 (16" -Sch 5S)

11.7525

445263

176723

124962

0.4062

0.5971

0.7500

4d

ßmax = 0.75

P

15.67 (16" -Sch 5S)

10.4440

447117

177459

125482

0.4079

0.8227

0.6665

5a

RDmin = 10^4

F

3.334 (3" -Sch 5S)

1.2860

21223

1792

1267

0.0910

0.6049

0.3857

1

5b

RDmin = 10^4

R

3.334 (3" -Sch 5S)

1.2860

21221

1792

1267

0.0910

0.6049

0.3857

1

5c

RDmin = 10^4

C

3.334 (3" -Sch 5S)

1.2860

21222

1792

1267

0.0910

0.6049

0.3857

1

5d

RDmin = 10^4

P

3.334 (3" -Sch 5S)

1.2336

21217

1792

1267

0.0910

0.6585

0.3700

1

6a

RDmax = 10^7

F

35.376 (36" - Sch 10)

26.5320

1004521

900067

636444

0.4060

0.5968

0.7500

6b

RDmax = 10^7

R

35.376 (36" - Sch 10)

26.5320

1022323

916019

647723

0.4132

0.6073

0.7500

6c

RDmax = 10^7

C

35.376 (36" - Sch 10)

26.5320

1003218

898899

635618

0.4054

0.5959

0.7500

6d

RDmax = 10^7

P

35.376 (36" - Sch 10)

23.5500

1004454

900008

636401

0.4059

0.8211

0.6657

7a

Dmin = .5 inch

C

0.71 (0.5" -Sch 5S)

0.3905

9849

177

125

0.1983

0.6248

0.5500

7b

Dmin = 1 inch

F

1.185 (1" -Sch 5S)

0.65175

16475

494

350

0.1988

0.6264

0.5500

8a

Dmax = 1.5 inches

C

1.77 (1.5" -Sch 5S)

0.9735

24022

1077

762

0.1940

0.6113

0.5500

8b

Dmax = 1.5 inches

F

1.77 (1.5" -Sch 5S)

0.9735

24519

1099

777

0.1980

0.6239

0.5500

9a

ßmin = 0.1

C

1.77 (1.5" -Sch 5S)

0.177

749

34

24

0.0061

0.6100

0.1000

9b

ßmin = 0.15

F

1.77 (1.5" -Sch 5S)

0.2655

1725

77

55

0.0139

0.6176

0.1500

10a

ßmax = 0.8

C

1.77 (1.5" -Sch 5S)

1.416

62333

2794

1976

0.5035

0.6045

0.8000

10b

ßmax = 0.7

F

1.77 (1.5" -Sch 5S)

1.239

47804

2143

1515

0.3861

0.6869

0.7000

11a

RDmin > 10^3

C

0.71 (0.5" -Sch 5S)

0.1867

2130

38

27

0.0429

0.6187

0.2630

2

11b

RDmin > 10^3

F

1.185 (1" -Sch 5S)

0.2417

2131

64

45

0.0257

0.6170

0.2040

2

ASME Small-Bore Orifice

Notes 1 RD at (1/3 of full scale flow rate) must be greater than or equal to 10000 2 RD at (1/3 of full scale flow rate) must be greater than or equal to 1000

Chevron Corporation

A-33

July 1996



Fig. A-16 Orifice Test Cases (Gas) Gas Square Edge

Fluid Type Orifice Type


h

100 inches of Water

Specific Gravity

G

0.5539

Absolute Viscosity

µ

0.02 centipoise

Pressure Temperature

P T

115 PSIA 68 º F

Compressibility Factor at flow

Zf

0.9849

Specific Heat Ratio

k

1.4


Fa

1.0000

ISO-5167/Miller Bore Orifice Test No.

Purpose

Tap Type

Pipe I. D.

Orifice Diameter

Pipe Reynolds No.


Normal Flow Rate

(F)lange, (R)adius, (C)orner, (P)ipe

D

d


at Normal Flow RD

Q

inches 2.245 (2" -Sch 5S)

inches 1.2348

0.707*Q

S

Orifice Coefficient of Discharge

Gas Expansion Factor for midscale flow

Beta Ratio

Notes

based on upstream static pressure

C

Y1

ß

SCFH

1a

Dmin = 2 inches

F, R, C, P F

198359

SCFH 47060

33277

0.1926

0.6069

0.9950

0.5500

1b

Dmin = 2 inches

R

2.245 (2" -Sch 5S)

1.2348

198315

47050

33269

0.1926

0.6069

0.9950

0.5500

1c

Dmin = 2 inches

C

2.245 (2" -Sch 5S)

1.2348

197893

46950

33198

0.1922

0.6056

0.9950

0.5500

1d 2a

Dmin = 2 inches Dmax = 36 inches

P F

2.245 (2" -Sch 5S) 35.376 (36" - Sch 10)

1.2348 17.6880

240162 2525798

56978 9442655

40290

0.2340

0.7373

0.9919

0.5500

6676965

0.1556

0.6026

0.9952

0.5000

2b

Dmax = 36 inches

R

35.376 (36" - Sch 10)

17.6880

2528184

9451577

6683274

0.1588

0.6150

0.9952

0.5000

2c

Dmax = 36 inches

C

35.376 (36" - Sch 10)

17.6880

2525586

9441863

6676405

0.1556

0.6026

0.9952

0.5000

2d 3a

Dmax = 36 inches ßmin = 0.2

P F

35.376 (36" - Sch 10) 15.67 (16" -Sch 5S)

15.9192 3.1340

2294218 171890

8576896 284647

6064781

0.1416

0.6848

0.9935

0.4500

201276

0.0239

0.5970

0.9954

0.2000

3b

ßmin = 0.2

R

15.67 (16" -Sch 5S)

3.1340

171874

284620

201257

0.0239

0.5970

0.9954

0.2000

3c

ßmin = 0.2

C

15.67 (16" -Sch 5S)

3.1340

171892

284650

201278

0.0239

0.5970

0.9954

0.2000

3d

ßmin = 0.2

P

15.67 (16" -Sch 5S)

3.1340

176185

291760

206305

0.0245

0.6120

0.9958

0.2000

4a

ßmax = 0.75

F

15.67 (16" -Sch 5S)

11.7525

2916470

4829623

3415059

0.4061

0.5969

0.9942

0.7500

4b

ßmax = 0.75

R

15.67 (16" -Sch 5S)

11.7525

2963427

4907382

3470043

0.4126

0.6065

0.9942

0.7500

4c

ßmax = 0.75

C

15.67 (16" -Sch 5S)

11.7525

2907921

4815466

3405049

0.4049

0.5951

0.9942

0.7500

4d

ßmax = 0.75

P

15.67 (16" -Sch 5S)

10.4440

2908084

4815735

3405239

0.4072

0.8212

0.9886

0.6665

5a

RDmin = 10^4

F

2.245 (2" -Sch 5S)

0.4715

27195

6452

4562

0.0264

0.5981

0.9954

0.2100

1

5b

RDmin = 10^4

R

2.245 (2" -Sch 5S)

0.4715

27194

6452

4562

0.0264

0.5981

0.9954

0.2100

1

5c

RDmin = 10^4

C

2.245 (2" -Sch 5S)

0.4715

27197

6453

4563

0.0264

0.5981

0.9954

0.2100

1

5d

RDmin = 10^4

P

2.245 (2" -Sch 5S)

0.4715

27946

6630

4688

0.0271

0.6139

0.9957

0.2100

1

6a

RDmax = 10^7

F

12.438 (12" -Sch 5S)

9.3285

2317164

3045748

2153669

0.4065

0.5975

0.9942

0.7500

6b

RDmax = 10^7

R

12.438 (12" -Sch 5S)

9.3285

2352668

3092415

2186668

0.4127

0.6066

0.9942

0.7500

6c

RDmax = 10^7

C

12.438 (12" -Sch 5S)

9.3285

2308617

3034513

2145725

0.4050

0.5953

0.9942

0.7500

6d

RDmax = 10^7

P

12.438 (12" -Sch 5S)

8.0847

2133182

2803917

1982669

0.3761

0.8068

0.9892

0.6500

ASME Small-Bore Orifice 7a

Dmin = .5 inch

C

0.71 (0.5" -Sch 5S)

0.3905

63449

4761

3366

0.1948

0.6138

0.9950

0.5500

7b

Dmin = 1 inch

F

1.185 (1" -Sch 5S)

0.6518

106705

13371

9454

0.1962

0.6182

0.9950

0.5500

8a

Dmax = 1.5 inches

C

1.77 (1.5" -Sch 5S)

0.9735

156043

29188

20639

0.1922

0.6056

0.9950

0.5500

8b

Dmax = 1.5 inches

F

1.77 (1.5" -Sch 5S)

0.9735

159262

29790

21065

0.1962

0.6182

0.9950

0.5500

9a

ßmin = 0.1

C

1.77 (1.5" -Sch 5S)

0.1770

4898

916

648

0.0060

0.6000

0.9954

0.1000

9b

ßmin = 0.15

F

1.77 (1.5" -Sch 5S)

0.2655

11084

2073

1466

0.0136

0.6043

0.9954

0.1500

10a 10b

ßmax = 0.8 ßmax = 0.7

C F

1.77 (1.5" -Sch 5S) 1.77 (1.5" -Sch 5S)

1.4160 1.2390

397766 310211

74402 58025

52610

0.4905

0.5889

0.9938

0.8000

41030

0.3823

0.6801

0.9945

0.7000

11a

RDmin > 10^3

C

0.71 (0.5" -Sch 5S)

0.0731

2131

160

113

0.0065

0.6127

0.9954

0.1030

2

11b

RDmin > 10^3

F

1.049 (1" -Sch 40)

0.1573

6592

731

517

0.0137

0.6087

0.9954

0.1500

2


Chevron Corporation

A-35

July 1996



Fig. A-17 Orifice Sizing Test Case (Vapor/Steam) Vapor(Steam) Square Edge

Fluid Type Orifice Type


h

100 inches of Water

Specific Volume

v

3.4979 cu. feet/lb.

Absolute Viscosity

µ

0.014 centipoise

Pressure

P

135 PSIA

Temperature

T

380 º F

Compressibility Factor at flow

Zf

0.9849

Specific Heat Ratio

k

1.3


Fa

1.0058

ISO-5167/Miller Bore Orifice Test No.

Purpose

Tap Type

Pipe I. D.

Orifice Diameter

Pipe Reynolds No.


Normal Flow Rate

at Normal Flow (F)lange, (R)adius, (C)orner, (P)ipe F, R, C, P

D inches

d


RD

inches

Q

0.707*Q

SCFH

SCFH

S

Orifice Coefficient of Discharge

Gas Expansion Factor for midscale flow

Beta Ratio

Notes

based on upstream static pressure

C

Y1

ß

1a

Dmin = 2 inches

F

2.245 (2" -Sch 5S)

1.2348

264851

1864

1318

0.1925

0.6065

0.9955

0.5500

1b

Dmin = 2 inches

R

2.245 (2" -Sch 5S)

1.2348

264791

1863

1318

0.1924

0.6062

0.9955

0.5500

1c

Dmin = 2 inches

C

2.245 (2" -Sch 5S)

1.2348

264227

1859

1315

0.1920

0.6050

0.9955

0.5500

1d

Dmin = 2 inches

P

2.245 (2" -Sch 5S)

1.2348

320815

2258

1595

0.2338

0.7367

0.9925

0.5500

2a

Dmax = 36 inches

F

35.376 (36" - Sch 10)

17.6880

3374519

374215

264610

0.1556

0.6026

0.9956

0.5000

2b

Dmax = 36 inches

R

35.376 (36" - Sch 10)

17.6880

3377707

374569

264860

0.1558

0.6034

0.9956

0.5000

2c

Dmax = 36 inches

C

35.376 (36" - Sch 10)

17.6880

3374237

374184

264588

0.1556

0.6026

0.9956

0.5000

2d

Dmax = 36 inches

P

35.376 (36" - Sch 10)

15.9192

3065618

339960

240388

0.1416

0.6848

0.9940

0.4500

3a

ßmin = 0.2

F

15.67 (16" -Sch 5S)

3.1340

229648

11281

7977

0.0239

0.5970

0.9958

0.2000

3b

ßmin = 0.2

R

15.67 (16" -Sch 5S)

3.1340

229627

11280

7976

0.0239

0.5970

0.9958

0.2000

3c

ßmin = 0.2

C

15.67 (16" -Sch 5S)

3.1340

229651

11281

7977

0.0239

0.5970

0.9958

0.2000

3d

ßmin = 0.2

P

15.67 (16" -Sch 5S)

3.1340

235381

11562

8176

0.0245

0.6120

0.9961

0.2000

4a

ßmax = 0.75

F

15.67 (16" -Sch 5S)

11.7525

3896306

191392

135334

0.4060

0.5968

0.9946

0.7500

4b

ßmax = 0.75

R

15.67 (16" -Sch 5S)

11.7525

3959061

194474

137514

0.4126

0.6065

0.9946

0.7500

4c

ßmax = 0.75

C

15.67 (16" -Sch 5S)

11.7525

3884880

190830

134937

0.4048

0.5950

0.9946

0.7500

4d

ßmax = 0.75

P

15.67 (16" -Sch 5S)

10.4440

3887287

190949

135021

0.4072

0.8212

0.9895

0.6665

5a

RDmin = 10^4

F

2.245 (2" -Sch 5S)

0.4715

36324

256

181

0.0264

0.5981

0.9958

0.2100

1

5b

RDmin = 10^4

R

2.245 (2" -Sch 5S)

0.4715

36324

256

181

0.0264

0.5981

0.9958

0.2100

1

5c

RDmin = 10^4

C

2.245 (2" -Sch 5S)

0.4715

36238

256

181

0.0264

0.5981

0.9958

0.2100

1

5d

RDmin = 10^4

P

2.245 (2" -Sch 5S)

0.4715

37328

263

186

0.0271

0.6139

0.9961

0.2100

1

6a

RDmax = 10^7

F

35.376 (36" -Sch 10)

26.5320

8778346

973470

688347

0.4052

0.5956

0.9946

0.7500

6b

RDmax = 10^7

R

35.376 (36" -Sch 10)

26.5320

8934434

990779

700586

0.4124

0.6062

0.9946

0.7500

6c

RDmax = 10^7

C

35.376 (36" -Sch 10)

26.5320

8766918

972202

687451

0.4047

0.5949

0.9946

0.7500

6d

RDmax = 10^7

P

35.376 (36" -Sch 10)

23.3480

8504294

943079

666857

0.3945

0.8152

0.9898

0.6600

ASME Small-Bore Orifice 7a

Dmin = .5 inch

C

0.71 (0.5" -Sch 5S)

0.3905

84644

188

133

0.1945

0.6129

0.9955

0.5500

7b

Dmin = 1 inch

F

1.185 (1" -Sch 5S)

0.6518

142542

529

374

0.1963

0.6185

0.9955

0.5500

8a

Dmax = 1.5 inches

C

1.77 (1.5" -Sch 5S)

0.9735

208322

1156

817

0.1920

0.6050

0.9955

0.5500

8b

Dmax = 1.5 inches

F

1.77 (1.5" -Sch 5S)

0.9735

212618

1180

834

0.1960

0.6176

0.9955

0.5500

9a

ßmin = 0.1

C

1.77 (1.5" -Sch 5S)

0.1770

6543

36

26

0.0060

0.6000

0.9958

0.1000

9b

ßmin = 0.15

F

1.77 (1.5" -Sch 5S)

0.2655

14783

82

58

0.0136

0.6043

0.9958

0.1500

10a

ßmax = 0.8

C

1.77 (1.5" -Sch 5S)

1.4160

530296

2942

2081

0.4894

0.5876

0.9943

0.8000

10b

ßmax = 0.7

F

1.77 (1.5" -Sch 5S)

1.2390

414146

2298

1625

0.3820

0.6796

0.9949

0.7000

11a

RDmin > 10^3

C

0.546 (0.5" -Sch 80)

0.0551

2126

4

3

0.0064

0.6286

0.9958

0.1009

2

11b

RDmin > 10^3

F

1.049 (1" -Sch 40)

0.1573

8786

29

20

0.0137

0.6087

0.9958

0.1500

2


Chevron Corporation

A-37

July 1996

Appendix B. Hand Calculation Method for Orifice Design

Abstract This appendix explains how to perform the calculations for orifice plate design by hand. For manual look-up of the orifice coefficient, S (given beta) or the diameter ratio, beta (given S), the following eight figures are provided: Figure B-1

Flange Taps - 6 Inch Pipe Size (6.056 in.); S vs. Beta for Different RD Values (Based on Data from ISO 5167)

Figure B-2

Flange Taps; Pipe Size Corrections for S Values (Based on Data from ISO 5167)

Figure B-3

Radius Taps; S vs. Beta for Different R D Values (Based on Data from ISO 5167)

Figure B-4

Corner Taps; S vs. Beta for Different R D Values (Based on Data from ISO 5167)

Figure B-5

Pipe Taps; S vs. Beta for Different RD Values (Based on Data from ISO 5167)

Figure B-6

ASME Small Bore with Flange Taps; S vs. Beta for Different RD Values—1 in. to 1-1/2 in. Pipe Size (Based on Data from "Fluid Meters: Their Theory and Application", 6th ed., 1971. Courtesy of ASME)

Figure B-7

ASME Small Bore with Corner Taps (.546 in.); S vs. Beta for Different RD Values—1/2 in. Schedule 80 Pipe Size (Based on Data from "Fluid Meters: Their Theory and Application", 6th ed., 1971. Courtesy of ASME)

Figure B-8

Quadrant — Edge Data; S and Thickness Ratio vs. Beta

Orifice calculation sheets have been prepared for guidance through the orifice calculation steps. Calculation sheets filled in with appropriate examples are attached. The four orifice calculation sheets are as follows: • • • •

LIQUID, square-edge orifice, Form ICM-EF-59B (Figure B-9) LIQUID, quadrant-edge orifice, Form ICM-EF-59C (Figure B-10) GAS, square-edge orifice, Form ICM-EF-59D (Figure B-11) VAPOR/STEAM, square-edge orifice, Form ICM-EF-59E (Figure B-12)

Note that in these calculation sheets, the term "sharp edge" is used instead of "square-edge." Results of the hand calculations agree closely with the computer program ORIFICE. The two types of calculations that can be performed are orifice sizing and orifice reranging, except for quadrant-edge.

Chevron Corporation

B-1

July 1999

Appendix B

B1.0


Calculation Procedure 1.

Fill in the STREAM PROPERTY input blocks. TEMPERAURE PRESSURE

--

degrees Fahrenheit for liquid and vapor/steam

--

degrees Rankine = °F + 460 for gas

--

PSIA for gas and vapor

LIQUID SPECIFIC GRAVITY

--

VAPOR/STEAM SPECIFIC GRAVITY -GAS SPECIFIC GRAVITY

Gb at 60 F (Gb for water = 1.00) cu.ft./lb.

--

for steam, use steam tables

--

Ideal sp. gr. is used = M.W. gas/M.W. dry air = Zb × Real sp. gr.

2.

3.

LIQUID KINEMATIC VISCOSITY

--

use centistokes

GAS AND VAPOR ABSOLUTE VISCOSITY

--

use centipoise

Fill in the FLOW RATE input blocks. UNITS:

For liquid use BPD, BPH, GPM, or GPH. For vapor use lbs/hour; for gas use SCFH.

FULL SCALE:

Flow that creates a dp across the orifice taps equal to the maximum value of the dp transmitter range.

NORMAL:

For square-edge and ASME small-bore this should be the mid-range dp value, which equals .707 × (full-scale flow rate)

1/3 FS:

For quadrant-edge, equal to .333 × (full-scale flow rate)

Fill in the ORIFICE DATA input blocks. TYPE OF ORIFICE PLATE

TYPE OF TAPS

--

Square-edge

--

ASME small-bore

--

Quadrant-edge

--

Flange, radius, corner or pipe for square-edge

--

Flange or corner only for ASME small-bore

--

Flange for quadrant-edge

PIPE INSIDE DIAMETER

--

Use 3 decimal places

ORIFICE DIAMETER

--

Entered for orifice re-ranging calculation

July 1999

B-2

Chevron Corporation


Appendix B

FULL SCALE DP

4.

5.

--

Use 3 decimal places

--

Entered for orifice sizing calculation

--

Standard value = 100 inches of water

Fill in the CORRECTIONS FACTORS. Fa

--

Correction for thermal expansion of orifice meter at flowing temperature found in ASME MFC-3M Tables or API MPMS. Chapter 14.3/AGA-3.

Y

--

Gas expansion factor for mid-scale flow Y1 = correction based on upstream pressure Y2 = correction based on downstream pressure Look up in table or chart - use mid-range differential pressure (dp)

Fpv

--

Supercompressibility correction factor Rarely used in refinery calculations (usually set = 1.0) Look up in table or chart when used.

Zb

--

Compressibility factor at 60°F, 1 Atm Rarely used in refinery calculations (usually set = 1.0) Look up in table or chart when used.

Calculate Pipe Reynold’s Number. Space is provided for values in equations. See examples on sheets provided.

6.

Calculate orifice size or dp range for square-edge or ASME small-bore orifice. Given h, Find d Fill in equations and solve for S. Look up beta ratio in table. Interpolation is necessary. Use Table in Figure B-2 to correct for pipe size other than 6-inches (2-12 inches). Use Reynold’s number column that is closest to value calculated in step 5. Given d, Find h Calculate beta ratio. Look up S in table. Interpolation is necessary. Use Reynold’s number that is closest to value calculated in Step 5. See examples on sheets provided.

7.

Chevron Corporation

Calculation for quadrant-edge orifice size.

B-3

July 1999

Appendix B


Select a plate thickness, look up values from Figure B-8, and calculate the differential pressure h. If h is too large, make plate thicker; if too small, make plate thinner, and repeat calculation. h should equal about 100 inches of water. The permissible range of Reynold’s numbers (at 1/3 full-scale flow) for a quadrant-edge orifice is a function of the beta ratio: beta ratio

0.2

0.3

0.4

0.5

0.6

min. RD

670.0

770.0

630.0

450.0

320.0

max. RD

17500.0

28000.0

45000.0

70000.0

85000.0

See example on calculation sheet provided.

July 1999

B-4

Chevron Corporation


Fig. B-1

Appendix B

Flange Taps - 6 Inch Pipe Size (6.056 in.); S vs. Beta for Different RD Values (Based on Data from ISO 5167)

Chevron Corporation

B-5

July 1999

Appendix B

Fig. B-2

July 1999


Flange Taps; Pipe Size Corrections for S Values (Based on Data from ISO 5167)

B-6

Chevron Corporation


Fig. B-3

Appendix B

Radius Taps; S vs. Beta for Different RD Values (Based on Data from ISO 5167)

Chevron Corporation

B-7

July 1999

Appendix B

Fig. B-4

July 1999


Corner Taps; S vs. Beta for Different RD Values (Based on Data from ISO 5167)

B-8

Chevron Corporation


Fig. B-5

Appendix B

Pipe Taps; S vs. Beta for Different RD Values (Based on Data from ISO 5167)

Chevron Corporation

B-9

July 1999

Appendix B

Fig. B-6

July 1999


ASME Small Bore with Flange Taps; S vs. Beta for Different RD Values—1 in. to 1-1/2 in. Pipe Size (Based on Data from "Fluid Meters: Their Theory and Application", 6th ed., 1971. Courtesy of ASME)

B-10

Chevron Corporation


Fig. B-7

Appendix B

ASME Small Bore with Corner Taps (.546 in.); S vs. Beta for Different RD Values—1/2 in. Schedule 80 Pipe Size (Based on Data from "Fluid Meters: Their Theory and Application", 6th ed., 1971. Courtesy of ASME)

Chevron Corporation

B-11

July 1999

Appendix B

Fig. B-8

July 1999


Quadrant — Edge Data; S and Thickness Ratio vs. Beta

B-12

Chevron Corporation


Fig. B-9

Appendix B

Orifice Calculation Sheet—Liquid, Square-edge Orifice

Chevron Corporation

B-13

July 1999

Appendix B


Fig. B-10 Orifice Calculation Sheet—Liquid, Quadrant-edge Orifice

July 1999

B-14

Chevron Corporation


Appendix B

Fig. B-11 Orifice Calculation Sheet—Gas, Square-Edge Orifice

Chevron Corporation

B-15

July 1999

Appendix B


Fig. B-12 Orifice Calculation Sheet—Vapor/Steam, Square-Edge Orifice

July 1999

B-16

Chevron Corporation

Appendix C. Guidelines for Meter Prover Design

Abstract This appendix explains the basics for meter provers used in custody transfer flow meter systems. It provides examples of various meter and prover systems, prover sizing and prover calculations. This guideline is an introduction to meter provers. More detailed information is available in the API Manual of Petroleum Measurement Standards, Chapters 4, 5, 6 and 12. Contents

Chevron Corporation

Page

C1.0 Principles: Custody Transfer Flow Metering

C-2

C2.0 Conventional Pipe Provers

C-7

C3.0 Small Volume Provers

C-18

C4.0 Calibrating Provers

C-21

C5.0 Correction Factors

C-23

C6.0 Example Proving Run Calculations

C-28

C-1

June 1989

Appendix C

C1.0


Principles: Custody Transfer Flow Metering C1.1

Flow Meter Basics Custody transfer flow meters measure volumetric flow either directly (PD meters) or inferentially (turbine meters) and transduce it into a pulse signal that can be counted. The number of pulses produced by a flow meter per unit volume throughput is called the K factor. Section C1.2 shows how the K factors for PD and turbine meters are related to the flow meter type and size. During a custody transfer operation, the pulses from a flow meter can be counted and displayed on a register. The total indicated barrels could then be hand-calculated by dividing the number displayed on the register by the flow meter K factor: Indicated Barrels = Meter Pulse Count/K factor (Eq. C-1)

Usually, however, the flow meter pulses to the register are scaled automatically so the register will read directly in barrels (See Figure C-1). The scaling is done by a module located between the flow meter and the register that will produce an output of 1 pulse for every n pulses of input. Fig. C-1

June 1989

Flow Meter Pulse Signal Scaling and Counting

C-2

Chevron Corporation


Appendix C

Unfortunately, the K factor has only an approximate value. Therefore, the indicated barrels determined from the K factor has, at best, an accuracy with respect to actual volume flow of 0.5%. A calibration factor called the meter factor (MF) is used to express the relationship between the actual volume flow and the indicated volume. MF = Actual Barrels/Indicated Barrels (Eq. C-2)

The MF will remain constant during the time interval of a custody transfer operation provided that flow conditions such as flow rate and liquid viscosity remain constant and the mechanical condition of the flow meter does not change. With this MF, the meter accuracy can be as good as 0.05% if the operating conditions are the same as those which prevailed during the proving. The MF should be determined at the beginning of each transfer as soon as operating conditions have been established. The MF needs to be determined again during the transfer only if the operating conditions (e.g., flow rate or temperature) shift beyond a specified range of tolerance. The determination of MF is performed on a very accurate volume calibrator called a “meter prover,” which is the topic of these guidelines.

C1.2

Types of Flow Meters Used Two types of flow meters are generally used for custody transfer of petroleum liquids: 1.

Positive Displacement (PD) meters: Best suited for low flow or high viscosity service such as crude oil and residual fuel oil. The Smith Meter Inc. rotary vane PD meter is commonly used by the Company in this kind of service. PD meters are also often used in low flow rate (less than 300 GPM) LPG service (propane through natural gas). A 16-inch PD meter (the largest size) has a maximum capacity of 12,500 BPH. The measuring element has a nominal displacement of 1 barrel per revolution. The meter gear train has a standard ratio of 1:1 so that the register shaft turns at a nominal rate of 1 revolution per barrel. The meter equipped with a photoelectric pulse transmitter (1000 pulses per revolution) has a K factor of 1000 pulses per barrel. A 4-inch PD meter has a maximum capacity of 600 GPM. The measuring element has a nominal displacement of 2 gallons per revolution. The meter gear train has a standard reduction ratio of 5:2 so that the register shaft turns at a nominal rate of 1 revolution per 5 gallons. The meter equipped with a photoelectric pulse transmitter (1000 pulses per revolution) has a K factor of 200 pulses per gallon or 8400 pulses per barrel.

2.

Chevron Corporation

Turbine meters: Best suited for high flow or low viscosity service such as refined products or light crude oil. Turbine meters are also used for ethane and all LPG in high flow rate service (more than 300 GPM).

C-3

June 1989

Appendix C


A safe rule of thumb: Turbine meters in crude oil service perform best when the meter size is 6 inches or greater and the operating viscosity is less than 20 centistokes. Equally important is that the oil be free of grass or debris that may become caught on the blades, and wax or other materials that may deposit on the meter internals. While PD meters measure volumetric flow by continual separation of flow into volumetric “packages,” turbine meters measure volumetric flow inferentially from momentum of the flowing stream, which causes the bladed rotor to spin. Output pulses are created by magnetic hubs embedded in the rim of the rotor at intervals along its circumference. A 20-inch turbine meter has a maximum capacity of 42,000 BPH. The K factor is typically 100 pulses per barrel. A 6-inch turbine meter has a maximum capacity of 4000 BPH. The K factor is typically 1000 pulses per barrel.

C1.3

Types of Meter Provers Used The most important kinds of meter provers are as follows: • • • •

Conventional pipe provers: Section C2.0 Small volume provers: Section C3.0 Tank provers Master meters

Custody transfer meters used for pipeline and marine transfers of crude and refined products (including LPG) are commonly proved with conventional pipe provers. Small volume provers are gaining in popularity but are not yet widely used for proving “sales” meters. There is particular incentive for their use in applications subject to space limitations such as offshore platforms and portable operations. API defines a “small volume prover” as any pipe or piston-type prover that produces less than 10,000 “unaltered” pulses from the flow meter being proved and uses pulse interpolation techniques to achieve the required 0.01% precision. Tank provers are used for proving smaller meters such as those used on truck rack loading facilities. This type of calibration is generally performed by contract inspectors. Master meters or portable provers are used to prove meters when stationary provers are not provided. These types of provings are sometimes performed by contract inspectors. These guidelines cover conventional pipe provers and small volume provers. Tank provers and master meters are not considered.

June 1989

C-4

Chevron Corporation


C1.4

Appendix C

Meter/Prover System Figure C-2 shows how the topics discussed above are implemented in a typical custody transfer flow meter/prover system. This generalized diagram applies to either PD meters or turbine meters. 1.

Meter Station The custody transfer metering station is a self-contained unit that is used to make volume measurements (for sales or accounting purposes) of liquid petroleum flowing in a shipping line. The optional, hand-operated control valve shown on the inlet line is used when pressure reduction is needed. (This control valve is not used in services where flashing can occur, such as LPG.) The pressure measurement shown at that location is needed for manual control. The control valve on the outlet line is used to maintain back pressure on the flow meters. This function is usually essential because reduced pressure on the meters would cause error due to cavitation. It is important to install a single control valve downstream of the prover rather than a valve on each of the individual meter runs. This valve location protects against vaporization in the prover and also holds the flow meter and prover at about the same pressure.

Fig. C-2

Generalized Diagram of Typical Meter Station for Custody Transfer for Petroleum Liquids

Chevron Corporation

C-5

June 1989

Appendix C


2.

Meter Runs The number of meter runs is not restricted. However, only two meter runs are shown on the diagram to avoid repetition. The block labeled “accessory equipment” would typically represent strainer and straightening vanes for turbine meters, and an air eliminator and a strainer for PD meters. Block valves are located on the inlet and outlet of each meter run. There is also a block valve from each meter run to the prover. The two valves used to select a particular meter for proving are shown with motor operators (MOVs). The motor operators are very convenient but not essential for local operation. They are, however, needed for remote proving. Double-block and bleed-type valves must be used between prover connections and on prover branches to avoid undetected leakage that would cause proving error.

3.

Measurements The small circles shown on the diagram indicate the locations of required measurements. The temperature and pressure required for flow meter corrections are measured at the flow meter outlet. The temperature and pressure required for prover corrections are the average of measurements at the prover inlet and outlet. A switch is provided for selection of the pulse signal to the prover pulse counter.

4.

Prover The prover is shown in three blocks: – – –

The flow metering section The prover controls The pulse counter

The function of these blocks is discussed in the sections that follow. However, one should understand the reason for high resolution pulse generators and counters. During the proving process, we must determine the oil volume registered by the meter. That volume must be accurate to 1 part in 10,000 and must be measured exactly as the displacer moves between the prover detector switches. Trying to read a mechanical counter during the proving process does not provide accurate results, whereas a gated electronic counter starts and stops precisely on detector switches. The need for a high-resolution pulse generator (commonly used pulse rate is 1000 pulses per barrel) is obvious if one uses an extreme example in which a pulse generator produces one pulse per barrel. In this case the prover volume must be 10,000 barrels to give an accuracy of 1 part in 10,000. The type of prover may be unidirectional, bidirectional, or small volume.

June 1989

C-6

Chevron Corporation


Appendix C

The setup procedure for proving a flow meter is straightforward and rather obvious: – –

– –

Make sure that the prover block valves are open. Open the proper set of MOVs to place the flow meter in series with the prover flow section. (Open the valve to the prover first; then close the block valve to the outlet header. This sequence will minimize the upset in the liquid flow.) Operate switch S-1 to direct the pulse signal from the selected flow meter to the prover pulse counter. Proceed with the proving operation as described next in Section C2.0 (Conventional Pipe Provers) and Section C3.0 (Small Volume Provers).

It important to understand that temperatures and pressures at the meter and prover must be read during each run and that valve body bleeds must be checked during each run. Thus, these items must be readily accessible and the paths between them must be free of obstructions that could cause tripping and other hazards.

C2.0

Conventional Pipe Provers C2.1

Operating Principle Figure C-3 shows a very elementary pipe prover. It consists of a straight piece of pipe with a calibrated measurement section between two ball detector switches. The ball is launched just upstream of detector #1. After the ball has been pushed through the measurement section by the liquid flow, it drops into the trap just downstream of detector #2. The measurement section has a base (60°F, 0 psig) volume that has been calibrated by a water draw method traceable to the National Bureau of Standards. (The water draw calibration method will be reviewed in Section C4.0.) During the proving procedure, the prover base volume must be corrected for the steel expansion/contraction at the operating temperature and pressure. Prover Operating Volume = Base Volume × Cts × Cps (Eq. C-3)

where: Cts = correction for temperature of steel Cps = correction for pressure on steel The flow meter being proved is located just upstream or downstream of the prover. The pulse count register is automatically scaled with the meter K factor so it will read directly in indicated barrels. An MF is determined by the proving operation as described below. The indicated barrels on the meter register multiplied by the MF gives the gross barrels with accuracy that can be as good as 0.05%.

Chevron Corporation

C-7

June 1989

Appendix C

Fig. C-3


Simplified Type of Meter Prover

The proving cycle is initiated by launching an elastomer ball into the pipe upstream of the calibrated measurement section. The ball has a diameter which is slightly larger than the pipe ID so it forms a seal as it is pushed along by the liquid flow. When the ball actuates the first detector switch the prover pulse counter starts counting pulses from the flow meter. The meter pulses continue to be counted while the liquid in the measurement section is being displaced. When the ball actuates the second detector switch the count is terminated. The proving cycle is ended when the ball drops into the ball trap. In this simple example it is assumed that the meter and prover are operating at the same temperature and pressure. (The more general case is covered in Section C6.0.) Therefore, the liquid volume passing through the meter during a proving run is exactly equal to the calibrated volume of liquid displaced by the ball as it passes between the detector switches (continuity in a liquid packed system): Actual Meter Throughput = Prover Operating Volume (Eq. C-4)

The pulse count from the flow meter divided by the K factor gives the meter indicated volume of the prover: Meter Indicated Volume = Pulse Count on Prover Counter/K factor (Eq. C-5)

Finally, the last step of the MF determination:

June 1989

C-8

Chevron Corporation


C2.2

Appendix C

Mechanical Aspects MF = Actual Meter Throughput/Meter Indicated Volume (Eq. C-6)

Because interchanging the ball manually from the trap to the launcher of the “simple” prover shown in Figure C-3 would be a very tedious job, this type of prover design is not very popular. Two types of pipe provers with automatic ball handling facilities are in common use: the unidirectional-type and the bidirectional type.

Unidirectional Prover—See Figure C-4 The heart of the unidirectional type prover is the ball interchange system, which provides the means for transferring the ball from the trap to the launcher, launching the ball, and then sealing off the path through the interchange unit with a double block-type seal. The integrity of the seal is verified by an indicating differential pressure sensor or a bleed valve. This type prover operates in exactly the same manner as the “simple” type described above except for the ball interchange system that has been added to make the operation automatic. As the name “unidirectional” implies, the ball always moves through the measurement section in the same direction during a set of proving runs.

Bidirectional Prover—See Figure C-8 The heart of this type of prover is the 4-way diverting valve, which operates at the start of the run and again at the mid-point to reverse the direction of flow through the measurement section, as the name “bidirectional” implies. The diverter valve has a double-block seal. The integrity of this seal is verified by an indicating differential pressure sensor or a bleed valve. The ball must make a pass through the measurement section in both directions—a round trip—to complete a run

C2.3

Construction Details Measurement Section The measurement section is constructed from carbon steel pipe ranging in size from 2 to 42 inches in diameter depending upon the flow capacity of the meter being proved. It is fabricated from straight pipe and long radius bends with a uniform inside dimension to within 1/32 inch on the vertical and horizontal diameters.

Ball Displacer The liquid flow forces an elastomer ball displacer to move along the length of the measuring section. The ball is inflated to a diameter slightly larger than the diameter of the pipe so that it provides a positive seal and “squeegees” the pipe wall clean as it passes. This type of displacer can even maintain its seal as it passes through the pipe bends. A piston-type displacer is often used in straight bidirectional provers but only a ball displacer can be used in folded bidirectional or unidirectional provers.

Chevron Corporation

C-9

June 1989

Appendix C

Fig. C-4

Unidirectional Prover: Plan View

Fig. C-5

Unidirectional Prover: End View

June 1989


C-10

Chevron Corporation


Appendix C

Fig. C-6

Unidirectional Prover Ball Trapping Sequence

Fig. C-7

Unidirectional Prover Ball Launching Sequence

Chevron Corporation

C-11

June 1989

Appendix C

Bidirectional Prover (Courtesy of Daniel Industrial, Inc.)

June 1989

Fig. C-8

C-12


Chevron Corporation


Fig. C-9

Appendix C

Bidirectional Prover Diverter Valve (Courtesy of Daniel Industrial, Inc.)

Detector Switches One detector is located at the beginning and another at the end of the measurement section to sense when the displacer passes. The detectors must have sufficient precision so that the measured displacement volume between the detectors can be repeatedly measured to within 0.01% when being calibrated by the water displacement method.

Pulse Counter The pulse signal from the meter being proved is input to a pulse counter that is gated on and off by the detector switches so that the number of pulses from the flow meter can be equated to the corrected volume between the detector switches.

Chevron Corporation

C-13

June 1989

Appendix C

C2.4


Unidirectional versus Bidirectional Designs In the past, the Company has preferred the unidirectional prover. This unit, when equipped with the Maloney-type ball interchange system, has been very easy to operate and maintain and has often been the better economic choice as well. However, the unidirectional prover may no longer be as cost competitive as it was in the past. In addition, certain advantages of bidirectional provers are becoming more important: •

A bidirectional prover for a particular flow rate can be supplied in a smaller size than a corresponding unidirectional prover.

•

A bidirectional prover is self-compensating for changes in detector switch adjustment. (For example, a switch that operates “sooner” will do so in both directions of ball travel so that while one pass is shortened, the other is lengthened.)

Unless a particular type of prover is required for specific reasons, it may be appropriate to solicit bids on both types of provers. When given an option, vendors usually quote the unit they can give the best price on. This results in a mixed variety bid, but that is of no consequence if either type is acceptable for the application.

C2.5

Design Practice The prover must be designed to meet a standard of repeatability for both calibration and operation. 1.

Two consecutive determinations of prover volume (between the detector switches) by the water draw calibration method shall agree within 0.02% of each other. This repeatability shall be achieved independently for both directions of bidirectional provers.

2.

Five consecutive runs at maximum flow rate and normal operating pressure shall not vary in pulse count by more than 0.05% between the maximum and minimum counts. The test liquid should be diesel fuel or jet fuel depending on the intended service. (Water may be used in LPG service.)

Ultimately, the minimum size of the prover must be based upon its ability to meet this standard. If the prover supplier agrees to provide a prover that will meet some size constraint imposed by customer space limitations, the prover must nonetheless meet the above repeatability standard. Remember, however, that while poor repeatability is an indication that a prover is not performing correctly, good repeatability does not guarantee good accuracy because there is always the possibility of an unrecognized systematic error. This type of error can sometimes be detected by forcing one of the water draw determinations to be made over a longer length of time so that any difference caused by leakage will be revealed.

June 1989

C-14

Chevron Corporation


Appendix C

The criteria listed below are guidelines and factors which affect the prover’s ability to meet the repeatability standards. In some cases, these criteria are more restrictive than the API standard.

C2.6

1.

A starting point to determine prover volume is 0.5% of the maximum BPH flow rate being proved.

2.

The volume requirement of item (1) may be relaxed when space restrictions limit the prover size. A smaller volume will force a greater precision from the rest of the prover to maintain the repeatability standard. In any case, the volume between detector switches must be sized to obtain at least 10,000 unaltered pulses from the flow meter being proved as needed for the ±.01% precision (i.e., the 1 part in 10,000 precision).

3.

The diameter of the prover pipe should be sized to obtain flow velocities not greater than 10 fps nor less than 2 fps over the range of flow rates being proved. The desired velocity is 5 to 7 fps.

4.

Some designers specify that the length of the prover pipe between detectors should be at least 30 feet, which is a sufficient length to make normal variations in switch operating points negligible. This is a conservative, though often more costly, approach.

5.

The detector switch repeatability shall be 0.01%.

6.

The prover manufacturer must provide sufficient prerun length (distance from the ball launcher to the first detector switch) so that the ball interchange or 4way valve seals before the ball reaches the first detector switch. Otherwise, fluid will bypass the calibrated section while the ball is within it. Without sufficient prerun length, the prover will not pass the repeatability test.

Application Examples The following examples illustrate the application of conventional pipe provers for specific flow meters. Examples 2 and 4 are hypothetical examples that illustrate designing for minimum size. Note that the other four examples, which are the data from actual Company installations, the meters have been supplied with (Prover Bbls)/(Meter BPH) = 0.5% or greater. 1.

16-inch PD meter—12,500 BPH maximum rated capacity Meter geared for 1 revolution of registration shaft per Bbl Pulse transmitter (photoelectric type)—1000 pulses per revolution Sizing criteria: Maloney’s standard size for 12,500 BPH flow rate Unidirectional prover—Pipe size: 20-in. diameter × .375-in. wall Ball velocity = 9.65 fps

Chevron Corporation

C-15

June 1989

Appendix C


Distance between detector switches = 173 feet Prover volume = 62.5 Bbls Pulse count per run = 62,500 pulses Prover Bbls/Meter BPH maximum = 0.5% This is a rather large prover (perhaps oversized) as evidenced by its length, volume, and number of counts per run. Contrast this with the minimum-sized prover in the next example. 2.

16-inch PD meter—12,500 BPH maximum rated capacity Meter geared for 1 revolution of registration shaft per Bbl Pulse transmitter (photoelectric type)—1000 pulses per revolution Sizing criteria: Ball velocity = 10 fps maximum Distance between detector switches = 30 feet min. Pulses per run = 10,000 min. Unidirectional prover—Pipe size: 20-in. diameter × .375-in. wall Ball velocity = 9.65 fps Distance between detector switches = 30 feet Prover volume = 10.8 Bbls Pulse count per run = 10,800 pulses Prover Bbls/Meter BPH maximum = 0.086% The sizing criteria for this example were selected to obtain a minimum size prover. The measurement section is much shorter than that of Example (1). Although it still meets the physical requirements of API Standards, it may take more operating and maintenance attention to get the required repeatability of 5 consecutive runs within a 0.05% spread. It is worth noting that repeatability problems with a minimum-sized prover such as this are usually caused by the meter and unsteady flow conditions and not by the prover. If the problem is in the prover, it is usually in the switches or is due to fluid bypassing the displacer or calibrated section.

3.

4-inch PD meter—857 BPH maximum rated capacity Meter geared for 5 gal. per revolution of registration shaft Pulse transmitter (photoelectric type)—1000 pulses per revolution 8400 pulses per Bbl Sizing criteria:

June 1989

C-16

Chevron Corporation


Appendix C

Daniel sizing for prover on an offshore platform Bidirectional prover—Pipe size: 8-in. diameter × .322-in. wall Ball velocity = 3.85 fps Distance between detector switches = 70 feet Prover volume = 4.33 Bbls Pulses per pass = 36,370 pulses Prover Bbls/Meter BPH maximum = 0.5% This prover is perhaps somewhat large but arguably so. 4.

20-inch turbine meter—42,000 BPH maximum rated capacity K = 100 pulses per Bbl Sizing criteria: Ball velocity = 10 fps maximum Distance between detector switches = 30 feet min. Pulses per run = 10,000 min. Unidirectional prover—Pipe size: 36-in. diameter × .375-in. wall Ball velocity = 9.67 fps Distance between detector switches = 84.0 feet Prover volume = 101.4 Bbls Pulses count per run = 10,140 pulses Prover Bbls/Meter BPH maximum = 0.24% The sizing criteria for this example were selected to obtain a minimum-sized prover. The prover volume turns out to be only half of the more conservative “1/2 percent of meter BPH” criterion. If the space is critical, a larger diameter, bidirectional prover would probably turn out to be more appropriate.

5.

3-inch turbine meter—930 BPH maximum rated capacity K = 4620 pulses per Bbl Sizing criteria: Maloney’s sizing for Casablanca Project Unidirectional prover—Pipe size: 6-in. diameter × .280-in. wall Ball velocity = 7.23 fps Distance between detector switches = 140 feet

Chevron Corporation

C-17

June 1989

Appendix C


Prover volume = 5.0 Bbls Pulses per run = 23,100 pulses Prover Bbls/Meter BPH maximum = .54% 6.

Kent P-NF 6-inch turbine meters—Ninian Platform 3930 BPH maximum rated capacity K = 790 pulses per Bbl Sizing Criteria: 1/2 percent of the Meter BPH rated capacity Bidirectional prover—Pipe size: 18-in. diameter × .375-in. wall Ball velocity = 3.78 fps Distance between detector switches = 74 feet Volume between detector switches = 21.43 Bbls Pulses per pass = 16,930 pulses (2 passes per run) Prover Bbls/Meter BPH maximum = 0.55%

C3.0

Small Volume Provers C3.1

Operating Principle By API definition, a small volume prover (SVP) is one whose calibrated volume is less than what is needed to produce 10,000 unaltered pulses from the flow meter being proved. According to this definition, an SVP could simply be an undersized conventional pipe prover. In this section, however, SVP refers only to one of the socalled “compact” type or “ballistic” type provers. The several different types of SVPs are distinctively designed (and patented) by their respective manufacturers. However, they have certain features in common: 1.

Measuring tube

The measuring section is a precision bore cylinder that is equipped with a piston displacer. A range of sizes is available, with diameters varying from 8 inches to 40 inches depending upon flow capacity. The maximum measuring volume is about 4 barrels, which is good for proving meters with flow rates up to about 25,000 BPH. The ratio of (Prover Bbls)/(Meter BPH) runs about 0.02%. 2.

Detector switches

The volume detector switches are mounted externally from the flowing liquid. They are generally optical, infrared-type switches that are actuated by flags or marks on a sensing shaft. The switches are very sensitive and have repeatabilities to the order

June 1989

C-18

Chevron Corporation


Appendix C

of ±0.0005% of the prover’s stroke. The use of low-temperature coefficient alloys such as Invar either corrects or minimizes ambient temperature effects. 3.

Pulse counting and interpolation

The SVP looks different and has more precision than the conventional unidirectional prover but it follows the same principle of operation. During the interval between detector switch actuations, a precise volume of liquid flows through the flow meter being proved and the pulses from the flow meter are gated to the SVP. There are N whole pulse periods plus one fractional pulse period. For the conventional prover (with greater than 10,000 pulses), the fractional pulse period is not significant because it is less than .01% of the total interval. However, for the SVP (with less than 10,000 pulses) the fractional pulse period is significant and must be accounted for. The SVP accomplishes this with a technique called “double chronometry,” which implements two clocks in a microprocessor: •

Clock No. 1 measures the gated time interval between detector switch actuations, TG.

•

Clock No. 2 measures the total time, TP, of the N integral pulse periods.

The calculations for determining MF for an SVP run are the same as those shown for the conventional pipe prover in Section C2.1. The calculations here are also simplified (as they were for the conventional pipe prover in Section C2.1) by assuming that the meter and prover are operating at the same pressure and temperature.

Prover Volume SVP Operating Volume = Base Volume × Cts × Cps (Eq. C-7)

Actual Meter Throughput = SVP Operating Volume (Eq. C-8)

Flow Meter Indication SVP Pulse Count = (Integral number of pulses, N) + (A fractional part of a pulse) = (TG/TP) × N (Eq. C-9)

Meter Indicated Volume = SVP Pulse Count/ K factor (Eq. C-10)

Meter Factor SVP MF = Actual Meter Throughput/Meter Indicated Volume (Eq. C-11)

Chevron Corporation

C-19

June 1989

Appendix C


(The corresponding set of steps for conventional provers in Section C2.1 are annotated with more detail.)

C3.2

Design Practice The design of conventional meter provers is described in the API Manual of Petroleum Measurement Standards, Chapter 4, Prover Systems. Although manufacturers fabricate these units using their particular selection of components such as detector switches, ball interchange units, etc., they follow the specifications of the industry standard. Each of the SVPs, on the other hand, is an invention that is designed and patented by each particular manufacturer. Although certain features of the SVPs have similarities as discussed above in Section C3.1, it is difficult to provide specific industry practice that covers their mechanical design. However, API has a new standard on SVPs scheduled for publication in late 1988. Repeatability is therefore adopted as the primary criterion for a prover’s acceptability. Remember that, whereas poor repeatability is an immediate indication that a prover is not performing satisfactorily, good repeatability does not necessarily indicate good accuracy because there is always the possibility of unknown systematic errors. Use of the SVP makes the meter/prover system more susceptible to this condition than the conventional pipe prover. As with conventional provers, the repeatability problem may be with the meter and unsteady flow conditions or with air in the system. If the problem is in the prover, it is usually in the switches or due to fluid bypassing the displacer.

C3.3

Application Examples Four examples of SVP application to specific flow meters are given below to illustrate the magnitude of certain important parameters such as the following. •

The number of times a PD meter rotor rotates per run

•

The number of pulses from a turbine meter per run

•

Prover Bbls per Meter BPH maximum

1.

16-inch PD meter—12,500 BPH maximum rated capacity Meter rotor displacement = 1 Bbl/rev. Pulse rate to prover = 1000 pulses /Bbl 36-inch × 120-gal. SVP; rated flow = 17,000 BPH Per run—2.857 rev. of meter rotor 2857 pulses from transmitter Prover Bbls/Meter BPH maximum = 0.023%

2.

June 1989

4-inch PD meter—857 BPH maximum rated capacity

C-20

Chevron Corporation


Appendix C

Meter rotor displacement = 2 gal./rev. Pulse rate to prover = 8400 pulses /Bbl 12-inch × 10-gal. SVP; rated flow = 1430 BPH Per run—5 rev. of meter rotor 2000 pulses from transmitter Prover Bbls/Meter BPH maximum = 0.028% 3.

6-inch turbine meter—3930 BPH maximum rated capacity K = 790 pulses per Bbl 18-inch × 30-gal. SVP; rated flow = 5000 BPH Per run—564 pulses from turbine meter Prover Bbls/Meter BPH maximum = 0.018%

4.

3-inch turbine meter—930 BPH maximum rated capacity K = 4620 pulses/Bbl 12-inch × 10-gal. SVP; rated flow = 1430 BPH Per run—1100 pulses from turbine meter Prover Bbls/Meter BPH maximum = 0.026%

PD meters often do not work well with SVPs due to gear slop and backlash. Turbine meters usually work well with SVPs.

C4.0

Calibrating Provers C4.1

Base Volume Definition (For conventional and small volume provers)

Unidirectional The base volume is that volume (at 60°F and 0 psig) displaced by one pass of the displacer between the detector switches. Consecutive water draw calibrations of the volume between the detector switches must agree within 0.02%.

Bidirectional The base volume is that volume (at 60°F and 0 psig) displaced by the displacer making a round trip between the detector switches. The volume displaced on consecutive round trips of the displacer between the detector switches must agree within 0.02% of each other. Consecutive water draw calibrations of each individual pass of the displacer between the detector switches should agree within 0.02%.

Chevron Corporation

C-21

June 1989

Appendix C

C4.2


Control of Conditions During Calibration 1.

The ball displacer must be inspected for roundness and cracks or cuts and inflated to the proper size.

2.

The seals must be inspected and tested as follows: – –

C4.3

Ball interchange section for a unidirectional prover. Four-way diverter valve for a bidirectional prover.

3.

Detector switches and pulse counters must be inspected and tested.

4.

Thermometers and pressure gages at inlet and outlet of measurement section must be calibrated to a standard traceable to the National Bureau of Standards. (Note: The name for the National Bureau of Standards is National Institute of Standards Technology, NIST.)

5.

The prover must be blinded off from the metering system. Any leaks from flanges, instrument connections, vents, drains, etc. must be eliminated.

6.

The prover must be cleaned to eliminate process deposits and process liquids, which would cause errors or contaminate the calibration water.

7.

The temperature of the prover should be uniform during the calibration. If weather conditions are variable, sun, wind or rain screens should be provided, especially if the prover is not insulated.

Water Draw Calibration Procedure Meters can be calibrated using provers because the volume between the detector switches can be measured very accurately. The process of measuring this volume is called water draw calibration. The procedure consists of displacing water from the prover into a series of cans that have accurately known volumes certified by the National Bureau of Standards. Figure C-10 shows a schematic drawing of how a meter prover is set up for the water draw calibration. A small pump is used to continuously circulate water through the prover. While the ball is repeatedly launched and allowed to pass through the measurement section, the vent and drain valves and the 3-way solenoid valve are manipulated to remove all air from the prover. When there is reasonable assurance that the prover and calibration system are packed solid with water, the water draw procedure is started. When the displacer trips the inlet detector switch, the 3-way solenoid valve is operated to divert the displaced water into the test measure. The displaced water continues to flow into the test measure until the outlet detector switch is tripped. Then the 3-way solenoid valve is reset to divert the flow of water back to the sump. The test measures have volumes that are certified by the National Bureau of Standards. The neck of the test measure contains about 2% of the volume of the measure and is equipped with a calibrated sight glass for accurate reading. The test measures are usually calibrated on a drain and fill basis so that they don’t have to

June 1989

C-22

Chevron Corporation


Appendix C

Fig. C-10 Schematic Drawing of Meter Prover Setup for Water Draw Calibration

be dried before each filling. The size of the test measures used and the sequence of filling are selected so that the last fill will come out just right, with the final level in the neck of the test measure where it can be read accurately. Measurements of the volume of water displaced between the detector switches are repeated until they agree repeatedly within 0.02%. At least two consecutive runs that agree within 0.02% are used to calculate the base volume of the prover.

C5.0

Correction Factors The correction factors discussed in Subsections C5.1 and C5.2 can be found in the following chapters of the API Manual of Petroleum Measurement Standards (API Manual):

Chevron Corporation

•

Chapter 12.2 — “Calculation of Liquid Quantities Measured by Turbine or PD Meters”

•

Chapter 11.1 — Tables 6A, 6B, 6C

•

Chapter 11.2 — “Compressibility Factors”

C-23

June 1989

Appendix C

C5.1


Temperature and Pressure Effects on Steel Equipment The standard (base) volume of a meter prover is based on a temperature of 60°F and a pressure of 0 psig. If the temperature or pressure increases, the internal volume of the prover increases. When the operating temperature and pressure are different from the base values, correction factors must be applied to calculate the actual volume.

Correction Factor for Temperature on Steel Since the volume of the prover at 60°F, V60, is known from the water draw calibration Then the volume at temperature T°F, VT, can be calculated as follows: VT = V60 × Cts (Eq. C-12)

where: Cts = Correction for temperature of steel Cts = 1 + (T - 60) × γ (Eq. C-13)

where: T = temperature of steel in °F γ = coef. of cubical expansion per °F 0.0000186 for carbon steel 0.0000265 for 316 stainless steel See the API Manual, Chapter 12.2, Tables A-1 and A-2.

Correction Factor for Pressure on Steel Because the volume of the prover at atmospheric pressure, Vatm, is known from the water draw calibration, the volume at pressure P psig, V P, can be calculated as follows: VP = Vatm × Cps (Eq. C-14)

where: Cps = Correction for pressure on steel Cps = 1 + (P × D)/(E × t) (Eq. C-15)

where: P = internal pressure in psig D = internal diameter in inches

June 1989

C-24

Chevron Corporation


Appendix C

E = modulus of elasticity for steel E = 30,000,000 lb/in.2 for carbon steel E = 29,000,000 lb./in.2 for 316 stainless steel t = wall thickness in inches See the API Manual, Chapter 12.2, Table A-3.

C5.2

Thermal Expansion of Petroleum Liquids It is customary to report liquid petroleum volume measurements at a standard temperature of 60°F. The liquid volume will expand as temperature increases or contract as temperature decreases. You can correct a volume of petroleum liquid at T°F to an equivalent volume at the base temperature of 60°F by using the following equation: V60 = VT × Ctl (Eq. C-16)

where: Ctl = Correction for temperature of liquid The correction factor Ctl has a value greater than 1.0000 when the operating temperature is less than 60°F since the liquid would expand if the temperature were increased to base temperature. Values for Ctl may be looked up in the API Manual, Chapter 11.1, as follows: Table 6A—Crude Oil

Ctl from °F and °API

Table 6B—Generalized Products

Ctl from °F and °API

Table 6C—Lube Oils

Ctl from α °F and °API

You must have the following information in order to use the tables: 1.

Class of liquid (e.g., crude, products, lube oil)

2.

API gravity at 60°F (rounded to nearest 0.5°API)

3.

Operating temperature (rounded to nearest 0.5°F)

The equations that are used to calculate the value of Ctl given in Tables 6A, 6B, and 6C are shown below. These equations will be useful to anyone who wants to computerize the correction factors. Go to Section C5.3 if you are not interested in that level of detail.

Chevron Corporation

C-25

June 1989

Appendix C


Equations Used for Calculating Tables 6A, 6B, and 6C (See Figures C-11 and C-12) Coefficient of Thermal Expansion, α60 α60 = (K0 + (K1 × k60))/k602 k60 = metric density in Kg/m3 = (141.5 × 999.012)/(API60 + 131.5) (Eq. C-17) Fig. C-11 Tables 6A and 6C (From Manual of Petroleum Measurement Standards, Ch. 11, Physical Properties, Sec 1 - Volume Correction Factors, 1980, Reaffirmed 1993. Courtesy of American Petroleum Institute) Class of Liquid

K0

K1

Crude Oils (Table 6A)

341.0957

0.0

Lube Oils (Table 6C)

144.0427

0.1896

Fig. C-12 Table 6B (From Manual of Petroleum Measurement Standards, Ch. 11, Physical Properties, Sec 1 - Volume Correction Factors, 1980, Reaffirmed 1993. Courtesy of American Petroleum Institute) Products: (Table 6B)

K0

K1

°API

Gasoline, Naphtha

192.4571

0.2438

52.1 — 85.0

Jet, Kerosene, Solvents

330.3010

0.0

37.1 — 47.9

Diesel, Fuel Oils

103.8720

0.2701

0.0 — 37.0

Transition Curve α60 = A + (B/k602) (Eq. C-18)

where: k60 = metric density in Kg/m3 = (141.5 × 999.012) /(API60 + 131.5) A

B

°API

1489.0670 (-.0018684) 48.0 — 52.0

Correction Factor for Temperature of Liquid Volume Ctl = (V60/V) = exp(-α60 × (T-60) × (1 + 0.8 × α60 × (T-60))) (Eq. C-19)

June 1989

C-26

Chevron Corporation


C5.3

Appendix C

Compressibility of Liquids It is customary to report liquid petroleum volume measurements at a standard pressure of 0 psig (atmospheric pressure) or at equilibrium vapor pressure for LPG. The liquid volume will decrease as pressure is increased. You can correct a volume of a petroleum liquid at a pressure P psig to an equivalent volume at the base pressure of 0 psig by using the following equation: V60 = VP × Cpl (Eq. C-20)

where: Cpl = Correction for pressure on liquid The correction factor Cpl has a value greater than 1.0000 when the operating pressure is greater than atmospheric pressure since the liquid would expand if the pressure were changed to base pressure. Cpl may be calculated from the following equation: Cpl = 1./(1. (P - Pe) × F × 0.00001) (Eq. C-21)

where: P = operating pressure—psig Pe = equilibrium vapor pressure at operating temperature; = psig 0 if less than 0 psig F = compressibility factor The compressibility factor, F, may be looked up in the API Manual, Chapter 11.2.1, or calculated from the equation given below. You must have the following information in order to use the table: • •

API gravity at 60°F (rounded to nearest 0.5°API) operating temperature (rounded to nearest 0.5°F)

The equation used to calculate the value for the compressibility factor, F, given in the table in Chapter 11.2.1 is shown below. F = exp(A + (B × T) + (C/k602) + (D × T)/k602) (Eq. C-22)

where: F = compressibility factor A = -1.99470 B = 0.00013427

Chevron Corporation

C-27

June 1989

Appendix C


C = 0.79392 D = 0.0023260 k60 = metric density in g/cc = (141.5 × 0.999012)/(API60 + 131.5) See the API Manual, Chapter 11.2.1 for Tabulation of F.

C6.0

Example Proving Run Calculations C6.1

Data from Proving Runs Assume that five proving runs have been performed. The following data have been recorded for each run: • • • •

Pressure and temperature at meter outlet Pressure and temperature at prover inlet Pressure and temperature at prover outlet Pulse count accumulated on prover counter

1.

Calculate the average temperature and pressure at the meter outlet.

2.

Calculate the average temperature and pressure in the prover. It is important to use the proper temperature or pressure at the prover (inlet vs. outlet). If the prover is downstream of the meter, the volume being measured in the prover is upstream of (behind) the displacer. In this case, we need the temperature and pressure at the prover inlet because that characterizes the fluid behind the displacer. Anything downstream of the displacer is not part of the measured volume. On the other hand, if the prover is upstream of the meter, the volume being measured in the prover is downstream (ahead) of the displacer and we need the temperature and pressure at the prover outlet.

3.

Round off values to nearest 0.5°F and to nearest measurable psig.

4.

Calculate average pulse count

The maximum or minimum pulse count must be less than .0005 times the average pulse count. The proving runs must be repeated until five consecutive runs (or sometimes any five out of six consecutive runs) meet the above criterion.

C6.2

Other Data Requirements Stream Data: API gravity rounded to nearest 0.5 unit Vapor pressure at operating pressure - psig

June 1989

C-28

Chevron Corporation


Appendix C

Flow rate: BPH Meter K factor: pulses/bbl Base Volume of Prover (round trip for bidirectional): Bbls

C6.3

Prover Volume Calculations Look up or calculate the prover correction factors: Cts, Cps, Ctl, Cpl Prover Operating Volume = Base Volume × Cts × Cps (Eq. C-23)

Corrected Prover Volume = Prover Operating Volume × Ctl ×Cpl (Eq. C-24)

C6.4

Metered Volume Calculations Look up or calculate the meter correction factors: Ctl, Cpl Meter Indicated Volume = Average Pulse Count/Meter K factor (Eq. C-25)

Corrected Metered Volume = Meter Indicated Volume × Ctl × Cpl (Eq. C-26)

C6.5

Meter Factor Calculation MF = Corrected Prover Volume/Corrected Metered Volume (Eq. C-27)

The MF is sometimes set in a flow computer but it is usually documented and used in ticket calculations.

C6.6

Total Flow Calculation Metered Indicated Volume = Total Pulse Count/Meter K factor (Eq. C-28)

Metered Gross Volume = Metered Indicated Volume × Meter Factor (Eq. C-29)

Metered Standard Volume = Metered Gross Volume × Ctl × Cpl (Eq. C-30)

Chevron Corporation

C-29

June 1989

Appendix C


where: Ctl and Cpl are based on flow weighted average temperature and pressure during transfer operation.

C6.7

Sample Calculation Service—Unleaded Gasoline 63.°API at 60°F (Rounded to nearest 0.5 °API) Flow rate = 1460 BPH Meter—8 in. Smith turbine meter K factor = 1000. Pulses/Bbl No temperature compensation Prover—16-in. diameter × 0.375-in. wall, unidirectional prover Base Volume = 19.50397 Bbls Note on rounding: Rounding of correction factors must be done at each step of the calculation. When the calculations are chained without rounding each factor, the final answer is not the same as when done in accordance with API.

Proving Run Average Pulse Count =

19,563 (rounded to nearest count)

Prover:

Temperature 53°F (rounded to nearest 0.5°F)

Pressure:

51 psig (rounded to nearest psig)

Meter:

Temperature 53°F (rounded to nearest 0.5°F

Pressure:

54 psig (rounded to nearest psig)

Prover Correction Factors (Rounded to 4 decimal places) Cts = 1. + (53. - 60.) × 0.0000186 = 0.9999 Cps = 1. + (51. × 15.25)/(30,000,000. × 0.375) = 1.0001 Ctl = 1.0049 from Table 6B F = 0.778 (from Chapter 11.2.1 of API Manual) Cpl = 1./(1. (51. × 0.778 × 0.00001)) = 1.0004

Meter Correction Factors Ctl = 1.0049 (Temp. at meter = Temp. at prover) Cpl = 1./(1. (54. × 0.778 × 0.00001)) = 1.0004

June 1989

C-30

Chevron Corporation


Appendix C

Corrected Prover Volume CPV = 19.50397 × 0.9999 × 1.0001 × 1.0049 × 1.0004 = 19.6073

Corrected Metered Volume CMV = (19,563/1000) × 1.0049 × 1.0004 = 19.6667

Meter Factor MF

Chevron Corporation

= 19.6073/19.6667 = 0.9970

C-31

June 1989

Appendix D. Flowmeter Selection Charts for Process Plant Meters

Abstract This appendix gives two oversized flowcharts to aid in the selection of flowmeters for measuring (1) natural gas and refinery fuel gas, and (2) steam flow in process plants. Explanatory notes are given for each of the two flowcharts. Contents

Page

D1.0 Meter Selection Flowchart for Measuring Natural Gas and Refinery Fuel Gas in Non-custody Applications D-2 D1.1 Using the Flowchart D1.2 Notes to Flowchart 1 D1.3 Literature Cited D2.0 Steam Meter Selection Flowchart for Measuring Steam Flow in Process Plant Applications D-5 D2.1 Using the Flowchart D2.2 Notes to Flowchart 2 D2.3 Literature Cited

Chevron Corporation

D-1

July 1996

Appendix D

D1.0


Meter Selection Flowchart for Measuring Natural Gas and Refinery Fuel Gas in Non-custody Applications D1.1

Using the Flowchart Flowchart 1 is intended for use in selecting new flowmeters and evaluating existing flowmeters used for measuring natural gas and refinery fuel gas in non-custody transfer applications. For applications where the flow meter is used in custody transfer measurement, refer to the Chevron Natural Gas Measurement Manual. For applications where the flow meter is used for calculating legal compliance with air quality regulations, some local air quality regulatory agencies require periodic inspection of the meter (e.g, orifice plate). Some may even require that the installation be made in accordance to API/AGA standards for custody transfer. The objective of selecting new fuel gas meters and evaluating existing meters is to close the balance around groups or meters to an uncertainty of two percent. Vital to this effort is periodic calibration of process variable transmitters and inspection of primary elements. This is especially true for the sharp edge orifice plate and the turbine meter. The sharp-edge orifice plate or the vortex-shedding meter are generally preferred for fuel gas measurement because they have no moving parts in the flowing stream, and are therefore less susceptible to damage by foreign matter entrained in the flowing stream. The measurement uncertainties of the sharp-edge orifice plate are well documented. Literally hundreds of papers documenting measurement uncertainties of this flow element have been published. An additional benefit of this device is that the inferential measurement transmitter, the d/P cell, is interchangeable among virtually all metering elements of this type, and among a dozen manufacturers. Vortex-shedding flowmeters are preferred for a high turndown between maximum and minimum flow rates. Flowmeter selection1 and evaluation generally assumes that pressure and temperature at the point of measurement are stable, or vary to a degree that induces an error of measurement at base conditions (e.g., 14.73 PSIA, 60 °F in Chevron USA) less than that inherent with the basic measurement. Accuracy of measurement (and therefore uncertainty in calculating energy balance around a set of meters) can be increased by compensating for variations in pressure, temperature, and Specific Gravity. Process variable transmitters and analyzers

1.

Multivariable transmitters which measure d/P, upstream pressure, and temperature and calculate flow at base conditions assuming a constant Specific Gravity and an ideal gas are under evaluation by Chevron Research and Technology Co. (The price of this multivariable transmitter will be less than that of a conventional d/P transmitter and pressure transmitter.)

July 1996

D-2

Chevron Corporation


Appendix D

common to a group of meters and flow correction algorithms resident in a monitoring computer can eliminate these uncertainties.

D1.2

Notes to Flowchart 1 The following notes are referenced to the numbered “decision diamonds” on the Meter Selection Flowchart. Where no note is listed, the ‘diamond’ is self-explanatory.

Chevron Corporation

A-2

Verify that block, vent, and drain valves used for isolation are functional prior to answering this question.

A-3

Although some manufacturers claim accurate measurement using 1/2" to 3/4" process lines and integral orifices, these installations are susceptible to large errors at flow rates below maximum flow. Therefore, limit process line size to 1" or greater when using integral orifice transmitters.

A-4 thru A-11

These steps determine whether multiple meters are required or whether a single thru orifice or vortex meter is acceptable. If multiple meters are required, each must be A-11 evaluated for type.

A-11

Velocity limits for new vortex-shedding meters are based on fuel gas with 0.6 < Sp. Gr. < 1.0, and are intended as a guideline for screening meter selection. Refer to Manufacturer’s Specification Sheets for exact limitations before final selection of a vortex-shedding meter for fuel gas measurement. (Note that piping system design guidelines suggest a gas velocity limit of 150 FPS to limit energy loss.)

B-1 thru B-4

These nested loops determine orifice size and generated differential for a sharp-edge orifice plate. Beta-ratio limits shown for selecting new orifice plate meters or evaluating existing orifice plate meters are based on minimizing uncertainty of measurement at reduced Reynolds number values (Rd < 100,000). AGA-3 [Reference 2] contains graphs quantifying meter uncertainty at beta-ratios above or below the limits shown on the Flowchart.

B-5

Some air quality regulatory agencies require periodic inspection of some or all components of a compliance meter. This includes the sharp-edge orifice plate which generates a differential for the inferential measurement of flow. Use of a Daniel Industries Inc. Senior Orifice Fitting (or equal) permits removal of the plate for inspection without removing the line from service.

B-6

At line size 10" or smaller, the downstream throat tap would be located on the weld-neck orifice flange or on its weld to process piping. This would induce an unpredictable error in measurement.

D-3

July 1996

Appendix D

July 1996


C-1

Secure concurrence from plant operations for use of a hot tap before answering this question.

C-2 thru C-11

These steps determine whether to use an insert vortex shedding meter or an insert turbine meter.

C-3

If line is not downstream of a knockout drum, excessive maintenance on insert turbine meters can be expected.

C-8

The 300 fps velocity limit for insert vortex shedding meters is significantly greater than the velocity limit for designing piping systems.

D-2

Refer to Miller [Reference 3] for determination of errors induced by changes in flowing pressure from reference conditions.

D-3

The Shell Flow Meter Engineering Handbook [Reference 4] lists errors arising from deviations from ideal meter installation for the sharp-edge orifice plate and related piping.

D-5

Beta-ratio limits shown for selecting new orifice plate meters or evaluating existing orifice plate meters are based on minimizing uncertainty of measurement and re-quantifying meter uncertainty at beta-ratios above or below the limits shown on Flowchart 1.

D-5 thru D-8

These nested loops set Beta-ratio for a single orifice plate and ranges for two transmitters measuring differential across this plate. If a single plate / two transmitter solution cannot be obtained, evaluate a low range meter using the existing orifice plate and an insert vortex shedding meter sufficient diameters upstream to comply with AGA-3 guidelines.

E-1

The limiting velocity above which permanent damage occurs is typically much higher than the 150 fps limit used for piping system design.

E-2 thru E-6

Consult Manufacturer’s technical publications to evaluate component and total inaccuracies for these conditions to determine probable inaccuracy of installation. Note that flow reading drops rapidly to zero as flow transits from turbulent to laminar flow regimes.

F-1 thru F-3

If meter is not downstream of a knockout drum, excessive maintenance on turbine meters can be expected. Errors from abnormal installation of a turbine meter generally all cause the device to read lower than actual flow.

F-4, F-5

Some relief can be obtained by varying the pitch of the turbine meter blading, where practical. Consult Manufacturer’s Specification Sheets and technical publications for details.

F-9

A ‘Slave’ meter (downstream, measuring flow to a consumer of the utility ) has less impact on energy balance than do the upstream ‘Master’ or ‘Sub-Master’ meters.

D-4

Chevron Corporation


D1.3

D2.0

Appendix D

Literature Cited 1.

Natural Gas Measurement Manual, Chevron Corporation - Chevron Research and Technology Company, Richmond, CA.

2.

Manual of Petroleum Measurement Standards, Chapter 14 - Natural Gas Fluids Measurement, Section 3, Concentric Square Edge Orifice Meters (Third Edition, 1990) ANSI / API-2530 (AGA-3) American Petroleum Institute Washington, D. C.

3.

Flow Measurement Engineering Handbook, R. W. Miller McGraw - Hill Book Company, New York - 1983

4.

Shell Flow Meter Engineering Handbook - 2nd Ed. McGraw - Hill Book Co. (UK), Ltd. London - 1985

Steam Meter Selection Flowchart for Measuring Steam Flow in Process Plant Applications D2.1

Using the Flowchart Flowchart 2 is intended for use in selecting new flowmeters and evaluating existing flowmeters used for measuring steam flow in process plant applications. The objective of selecting new steam meters and evaluating existing meters is to close the energy balance around groups of meters to an uncertainty of two percent. Periodic calibration checks of process variable transmitters and inspection of primary elements is vital to this effort. This is especially true for the sharp edge orifice plate and the turbine meter. The sharp-edge orifice plate is generally preferred for steam measurement because it has no moving parts in the flowing stream, and is therefore less susceptible to damage by entrained foreign matter. Flow nozzles are a good second choice, offering less sensitivity to damage by foreign matter in exchange for a higher first cost and greater difficulty in removing for inspection. Steam lines are subject to periodic expansion and contraction from thermal cycling, which can dislodge mill scale and internal corrosion byproducts. When these are combined with entrained boiler water chemicals and slugs of condensate, any device protruding into the flowing stream can be compromised. The measurement uncertainties of the sharp-edge orifice plate are well established. Literally hundreds of papers documenting measurement uncertainties of this flow element have been published. An additional benefit of this device is that the inferential measurement transmitter, the d/P cell, is interchangeable among virtually all metering elements of this type, and among a dozen manufacturers. The Shell Flow Meter Engineering Handbook [Reference 1] lists the magnitude and direction of errors induced by departures from recommended construction and installation standards.

Chevron Corporation

D-5

July 1996

Appendix D


Vortex-shedding flowmeters are acceptable for measuring steam flow within the turbulent flow regime, provided that the steam is dry or superheated. Significant amounts of entrained condensate can erode bluff body surfaces in the vortex-shedding meter, altering the frequency and distribution of vortices. Insert turbine meters should be considered only when a steam line cannot be isolated for installation of an orifice plate, and when the measuring element can be protected from entrained condensate. Averaging pitot tubes (Annubars) are not recommended for general steam flow applications, since they have a severely constrained envelope of application, and they can partially plug with no indication of this condition. Under high velocity conditions, averaging pitot tubes can experience vibration-induced fatigue failure, launching pieces of metal into downstream piping. Unless steam temperature and pressure are both controlled and stable at a known value, pressure and temperature should be measured and a compensation algorithm implemented for the meter station. ‘Smart’ microprocessor-based transmitters can perform this compensation in the field, or a common, Digital Control System (DCS) or programmable logic controller (PLC) - based algorithm can be shared by multiple steam metering stations, at the cost of increased processor time. Where steam quality is less than 100%, consider installing a throttling calorimeter to correct the flow measurement for moisture content. This device can measure steam quality down to 95%, at pressures up to 600 PSIG. Details of construction and installation can be found in Steam - Its Generation and Use [Reference 2].

D2.2

Notes to Flowchart 2 The following notes are referenced to the numbered “decision diamonds” on the Meter Selection Flowchart. Where no note is listed, the ‘diamond’ is self-explanatory.

July 1996

G-2

Verify that block, vent, and drain valves used for isolation are functional prior to answering this question.

G-3 thru G-5

These steps ensure that steam temperature is known. For steam quality less than dry, saturated, refer to text, above for a discussion on installing a throttling calorimeter. For steam pressures above or steam qualities below those stated in the text and referenced book, special techniques must be used to determine steam quality or total heat content.

G-4

Refer to Miller [Reference 3] for determination of errors induced by changes in flowing temperature from reference conditions.

G-6

Some publications describe use of small-bore orifice plates on 1" and 1-1/2" line sizes. These measuring elements are viable only of the user recognizes the gross errors induced by varying pipe roughness and internal diameter, and the reduced turndown ratio inherent with this specialty item.

D-6

Chevron Corporation


Appendix D

Integral orifice flow elements less than 1" nom. size should be avoided because of their severely restricted turndown ratio. G-8

a. Investigate downstream piping and equipment in making this decision. Generally, there is no benefit in installing a lowloss flow element, only to dissipate the saved pressure across a control valve immediately downstream. Since steam expands adiabatically through an orifice, there is no incentive for recovering pressure is the steam is being used solely for heating. b. Low-loss tubes other than a classic venturi can be used to recover pressure at a meter station. Refer to manufacturer’s specifications for a detailed evaluation.

G-9

The term shown, ReD, is pipe Reynolds number - sometimes expressed, RD or RD. Note that selection of a low-loss tube severely restricts the range of Reynolds numbers over which the meter can maintain accuracy.

G-9 & G-11

W is the weight of steam flow, expressed in pounds per hour. The upper limit of ReD far exceeds the economic limit for sizing piping systems, as discussed in the Chevron Piping Manual [Reference 4].

G-10 & G-12

This iterative process sizes dual flow elements for extended turndown. Each flow element and associated flow transmitter must be evaluated independently.

G-11

Prior to evaluating this decision, assume a Β of 0.55 and a line size measuring element.

G-11 & G-12

Chevron Corporation

The ReD low limit cited in Shell is : ReD>1260 Β2D, where D is stated in millimeters. the equivalent English limit becomes ReD>32,000 Β2 D, where D is stated in inches.

G-13 & G-14

Wmax and Wmin are defined as the maximum and minimum weight flows to be measured to the expected degree of accuracy for the flowmeter installation being designed or evaluated.

G-15

Below this value, minor changes in the elevation of impulse tubing can introduce gross measurement errors.

G-16 & G-18

The Β limits used for steam are those suggested for natural gas in AGA-3 [Reference 5]. There is no convincing literature indicating that the Β limits for other gases should be materially different for equivalent uncertainty limits.

G-17

This limits the measurement uncertainty to less than 0.5% for flows less that Wcalc for all β ratios.

D-7

July 1996

Appendix D


G-20

At line size 10" or smaller, the downstream throat tap would be located on the weld-neck orifice flange or on its weld to process piping. This would induce an unpredictable error in measurement.

H-1

Secure concurrence from plant operations for use of a hot tap before answering this question.

H-2 thru H-13

These steps determine whether to use an insert vortex shedding meter or an insert turbine meter.

H-4 thru H-6

These steps ensure that the specific weight / specific volume of the steam is known. By extension, total heat can be determined from steam tables.

H-5

If line is not downstream of a dropout leg and steam trap, excessive maintenance on insert turbine meters can be expected.

H-12

The 300 fps velocity limit for insert vortex shedding meters is significantly greater than the velocity limit for designing piping systems.

J-2 & J-3

If both answers are “No” return to the beginning of the Flowchart to determine whether a new flowmeter can be installed.

J-4

Recommended piping configurations in AGA-3 [Reference 5] are for natural gas flow measurement. There is no convincing literature indicating that the piping configurations should be any different for steam. The Shell Flow Meter Engineering Handbook [Reference 1] lists errors arising from deviations from ideal meter installation for the sharp-edge orifice plate and related piping.

July 1996

J-5 & J-6


J-7

Beta-ratio limits shown for selecting new orifice plate meters or evaluating existing orifice plate meters are based on minimizing uncertainty of measurement at reduced Reynolds number values (Rd < 100,000). AGA-3 [Reference 5] contains graphs quantifying meter uncertainty at beta-ratios above or below the limits shown on the Flowchart 2.

J-9

Below this value, minor changes in the elevation of impulse tubing can introduce gross measurement errors.

J-10

This limits the measurement uncertainty to less than 0.5% for flows less than Wcalc for all β ratios.

J-11

Limit turndown on analog transmitters without a square root extractor (internal, external, or embedded in display software) to 3 : 1. This places Wmin at 11% of full scale, and minimizes display errors across the range of measurements.

D-8

Chevron Corporation


D2.3

Chevron Corporation

Appendix D

K-1

The limiting velocity above which permanent damage occurs is typically much higher than the 150 fps limit used for piping system design.

K-2 & K-3

Consult Manufacturer’s technical publications to evaluate component and total inaccuracies for these conditions to determine probable inaccuracy of installation. Note that flow reading drops rapidly to zero as flow transits from turbulent to laminar flow regimes.

K-4 & K-5


L-1

If line is not downstream of a dropout leg and steam trap, excessive maintenance on insert turbine meters can be expected.

L-2

Consider special precautions (retracting insert turbine meter) to protect against damage by condensate slugs during plant warm-up and shutdown.

L-3

Errors from abnormal installation of a turbine meter generally all cause the device to read lower than actual flow. Yaw and pitch errors (turbine centerline not parallel with pipe center line) will also cause the device to read lower than actual flow.

L-4 & L-5


L-6 & L-7

Some relief can be obtained by varying the pitch of the turbine meter blading, where practical. Consult Manufacturer’s Specification Sheets and technical publications for details.

L-6 thru L-9

For this evaluation. Vmax is maximum velocity; Vnorm is normal velocity; Vmin is minimum velocity. Calculate these velocities at minimum expected line pressure at the flowmeter.

L-10

A ‘Slave’ meter (downstream, measuring flow to a consumer of the utility ) has less impact on energy balance than do the upstream ‘Master’ or ‘Sub-Master’ meters.

Literature Cited 1.

Shell Flow Meter Engineering Handbook - 2nd Ed. McGraw - Hill Book Co. (UK), Ltd. London - 1985

2.

Steam / Its Generation and Use - 39th Edition The Babcock & Wilcox Company. New York - 1978

3.

Flow Measurement Engineering Handbook, R. W. Miller McGraw - Hill Book Company, New York - 1983

D-9

July 1996

Appendix D

July 1996


4.

The Piping Manual, Chevron Corporation Chevron Research and Technology Company, Richmond, CA.

5.

Manual of Petroleum Measurement Standards, Chapter 14 - Natural Gas Fluids Measurement, Section 3, Concentric Square-Edge Orifice Meters (Third Edition, 1990) ANSI / API-2530 (AGA-3), American Petroleum Institute, Washington, D. C.

D-10

Chevron Corporation


Chevron Corporation

Appendix D

D-11

July 1996


Chevron Corporation

Appendix D

D-13

July 1996


Chevron Corporation

Appendix D

D-15

July 1996


Chevron Corporation

Appendix D

D-17

July 1996


Chevron Corporation

Appendix D

D-11

July 1996


Chevron Corporation

Appendix D

D-13

July 1996


Chevron Corporation

Appendix D

D-15

July 1996


Chevron Corporation

Appendix D

D-17

July 1996


Chevron Corporation

Appendix D

D-11

July 1996


Chevron Corporation

Appendix D

D-13

July 1996


Chevron Corporation

Appendix D

D-15

July 1996


Chevron Corporation

Appendix D

D-17

July 1996


Chevron Corporation

Appendix D

D-11

July 1996


Chevron Corporation

Appendix D

D-13

July 1996


Chevron Corporation

Appendix D

D-15

July 1996


Chevron Corporation

Appendix D

D-17

July 1996

Appendix E. Analyzer Drawings

Contents

Chevron Corporation

Page

Figure E-1 Typical Analyzer Wiring: 4-20 mA Loop with Alarms

E-2

Figure E-2 Typical Sample Probe

E-3

Figure E-3 Cylinder Rack Detail Analyzer House Outside Wall

E-4

Figure E-4 Typical Analyzer Shelter Vapor Vent System to Flare Line

E-5

Figure E-5 Typical Analyzer Shelter Liquid Vent System

E-6

Figure E-6 Typical Analyzer Shelter Vent System to Atmosphere

E-7

Figure E-7 Typical Gas/Liquid Analyzer Fast Loop Plate

E-8

E-1

July 1996

Appendix E

Fig. E-1

July 1996


Typical Analyzer Wiring: 4-20 mA Loop with Alarms

E-2

Chevron Corporation


Fig. E-2

Appendix E

Typical Sample Probe

Chevron Corporation

E-3

July 1996

Appendix E

Fig. E-3

July 1996


Cylinder Rack Detail Analyzer House Outside Wall

E-4

Chevron Corporation


Fig. E-4

Appendix E

Typical Analyzer Shelter Vapor Vent System to Flare Line

Chevron Corporation

E-5

July 1996

Appendix E

Fig. E-5

July 1996


Typical Analyzer Shelter Liquid Vent System

E-6

Chevron Corporation


Fig. E-6

Appendix E

Typical Analyzer Shelter Vent System to Atmosphere

Chevron Corporation

E-7

July 1996

Appendix E

Typical Gas/Liquid Analyzer Fast Loop Plate

July 1996

Fig. E-7

E-8


Chevron Corporation

Appendix F. Philadelphia FCC Control Specification

Appendix F is an example Control Specification to define Chevron’s requirements for advanced control technology applied to a specific process unit. The Philadelphia FCC unit is used as the example. This type of document has been used to solicit competitive bids from multiple vendors in order to decide which vendor to use for project implementation. Since 1994, Chevron has had an alliance agreement with AspenTech (formerly Dynamic Matrix Control Corporation). The alliance agreement to supply advanced control technology and services includes discounts for non-competitive bidding as well as for doing the bulk of our advanced control business with AspenTech. The parts of Chevron in the agreement include Chevron Corporation, Chevron Products Co., Chevron Chemical Co., Chevron Canada Ltd., and CRTC. Typically, advanced control projects using AspenTech services no longer use a control specification. Instead, meetings are held between Chevron and AspenTech to discuss the project requirements and control design. The savings due to Chevron foregoing specification preparation and AspenTech foregoing specification quotation are part of the benefits of the alliance. The format of the document is valuable as a guide for those parts of Chevron not covered by the alliance or for those projects that see potential benefits outside of the alliance. Contents

Page

F1.0 General Information

F-3

F1.1 The Project F1.2 Process Control System F1.3 Terminology F1.4 Dynamic Matrix Control F1.5 Intermediate Regulatory Control F1.6 Calculated Variables F1.7 Operating Environment F1.8 Format of Cost Estimates F2.0 Control Design Basis

F-5

F2.1 Process Considerations F2.2 DCS Data Base List and P&ID’s F2.3 Dynamic Matrix Control (DMC)

Chevron Corporation

F-1

May 1993

Appendix F


F2.4 Intermediate Regulatory Control (IRC) F2.5 Calculated Variables Attachment I-A

F-7

Honeywell TDC-3000 System Components Attachment I - B

F-8

Functional Requirements for Intermediate Regulatory Control Attachment II - A

F-9

On-line analyzers Attachment II - B

F-10

DMC-1: Reactor/Regenerator/Fractionator Attachment II - C

F-12

DMC-2: Gas Recovery Unit Attachment II - D

F-14

Intermediate Regulatory Control Attachment II - E

F-17

Vendor Calculated Variables / Predictors Attachment II - F

F-18

DCS Data Base List (not included) and P&ID’s (not included)

May 1993

F-2

Chevron Corporation


F1.0

Appendix F

General Information F1.1

The Project The Chevron Philadelphia refinery has reinstrumented the FCC, CO Boiler and Sulfur Recovery Unit, consolidating the operations into the existing FCC unit control house. A Honeywell TDC 3000 Distributed Control System (DCS) was placed in service 2Q92. The new control system will provide the foundation for the implementation of advanced control. The scope of this specification is to define the advanced control system which will be installed, with the objectives of increasing throughput and improving recovery of higher valued product. The implementation will begin in the fourth quarter of 1993. The Philadelphia refinery will assign one control engineer to assist in the implementation and provide the primary level of ongoing support.

F1.2

Process Control System The process control instrumentation consists of a Honeywell TDC 3000 Distributed Control System (DCS). The DCS system for the FCC uses 3 redundant Process Managers (PM). Honeywell smart transmitters are used and all critical loops have redundant analog output modules. The DCS includes one Application Module (AM) for use by the FCC, CO Boiler and Sulfur Recovery Unit. A diagram of the process control system is shown in Attachment I-A. Chevron’s VAX-based UNICORN computer system is available for monitoring and management information purposes only.

F1.3

Terminology For the purposes of this document, the following definitions will be used:

Chevron Corporation

•

Dynamic Matrix Control (DMC)

–

Synonymous for advanced control.

•

Intermediate Regulatory Control (IRC)

–

More complicated strategies than the Basic Regulatory Control, not including DMC.

•

Basic Regulatory Control (BRC )

–

Single loop control done in the Process Managers. (The scope is the same as the pneumatic controls prior to reinstrumentation).

•

Calculated Variables

–

Calculations used for advanced control or operator information. Typically, calculations are done in the Application Module.

F-3

May 1993

Appendix F


F1.4

F1.5

F1.6

May 1993

Dynamic Matrix Control •

DMC technology (licensed by Chevron from the DMC Corp.) will be used to do the advanced control. The latest version 5.0 will be used. (Chevron has a Corporate DMC license).

•

DMC will run on a dedicated computer (DEC VAX or Alpha) interfaced to the Honeywell TDC3000 via a DMCi/ABE. Displays to provide operators with DMC interaction will be implemented on the Honeywell universal stations. (Chevron will provide DMCi/ABE software and licenses).

•

The number of DMC model coefficients and the DMC execution frequency are to be determined during the model identification phase.

•

The Philadelphia refinery will provide analyzer signals and analyzer integrity status information to the DCS for use by the DMC controller. Where no analyzers are available, complete inferential predictors will be developed and installed by the contractor. For those streams with on-line analyzers, a simpler approach shall be used by the contractor: models between an appropriate temperature and the analyzer measurement will be developed and used on line. The analyzer measurement will automatically update the prediction when a new (valid) measurement is received.

Intermediate Regulatory Control •

The vendor has the responsibility for determining the engineering equations needed for Intermediate Regulatory Control (IRC).

•

The Honeywell TDC 3000 DCS running release 4.10 will be used for running IRC. No Intermediate Regulatory Control will run in the UNICORN system.

•

IRC operator displays will be graphic displays rather than faceplate displays. The proposed display standards will be reviewed by the Philadelphia refinery with the right of approval and modification. A separate document (Attachment I - B) defines the functional requirements for the IRC display standards.

•

IRC strategies will typically run in the Application Module (AM) except in special situations requiring higher reliability or faster processing. In these cases the IRC strategies will run in the Process Manager (PM). The refinery control system includes one AM.

Calculated Variables •

If DMC requires a manipulated variable such as heat duty or total flow rates, the vendor will implement the calculation. These variables will be calculated from existing measurements.

•

The vendor has the responsibility for determining the engineering equations needed for computing the calculated variables.

F-4

Chevron Corporation


F1.7

Appendix F

Operating Environment The objective of the Philadelphia FCC unit is to maximize throughput and recovery of higher valued products while meeting product quality constraints and using energy efficiently. If economic, the refinery can buy additional FCC charge stock if the FCC unit can process more feed than the Crude Unit (AVU 137) can provide.

F1.8

Format of Cost Estimates Please provide lump sum quotes itemized for each of the following items, from which we could sign a contract: •

FCC Unit Intermediate Regulatory Control

•

Inferential and Engineering Calculations

•

FCC Reactor, Regenerator and Main Fractionator DMC

•

FCC GRU (Absorber/Stripper, Lean Oil Still, 2 Debutanizers, Depropanizer, Deethanizer) DMC

Provide a Time and Expense Rate Schedule for this type of job.

F2.0

Control Design Basis F2.1

Process Considerations The Philadelphia FCC is a dual riser, slide valve operated unit with an associated reactor riser temperature of 965°F. The regenerator currently runs in a partial combustion mode. There are noticeable seasonal changes in operation due to hot /humid summers and cold/dry winters. This is reflected, for instance, in a throughput of 60 MBPD in summer when the air blowers are limiting and 70 MBPD in winter when the gas recovery unit limits the process. This poses a challenging control and optimization problem. The transition from a mainly pneumatic control configuration to a fully digital (DCS) system opens up many process improvement opportunities through application of intermediate regulatory and advanced control strategies.

F2.2

DCS Data Base List and P&ID’s The DCS Data Base List and the P&ID’s for the FCC are attached as Attachment II - F.

F2.3

Dynamic Matrix Control (DMC) One DMC controller is proposed for the reactor, regenerator, and main fractionator (DMC-1). A second DMC controller is proposed for the recovery side columns (DMC-2). The DMC design for the reactor/regenerator/fractionator is defined in Attachment II - B. The DMC design for the recovery side is defined in

Chevron Corporation

F-5

May 1993

Appendix F


Attachment II - C. These attachments include the controlled and constrained variables, manipulated variables, and feedforward variables. They are our best estimate of the variables for 1232’s advanced control. It is the Vendor’s responsibility to verify the adequacy of this information for their proposal. Today it is infeasible to have all the control in one DMC controller. The two controllers have been designed to encompass the important interactions. The size of the controllers are well within the capabilities of the microVAX computer and comparable to known successful installations. The DMC-1 controller is logically connected with DMC-2 through the 3 feedforward variables of DMC-2 (see Attachment II - C). These feedforward variables strongly help to decouple interactions (mainly one-way) between DMC-1 and DMC-2. Both DMC controllers incorporate a linear program (LP), which uses the predicted steady-state errors for each of the controlled/constraint variables, to calculate the optimum steady-state targets for all the manipulated variables. The LP will also determine optimum targets for the controlled/constraint variables. The controlled/constraint variables operate between an upper and lower limit in order to allow the greatest economic opportunity for the LP. The use of a composite LP, which coordinates multiple DMC controllers, shall be included in the bid response. If this package is sufficiently field proven, this approach should be used to coordinate DMC-1 and DMC-2. The composite LP will include compounded prediction vectors for both DMC controllers and will thus provide internally consistent control objectives for both controllers.

F2.4

Intermediate Regulatory Control (IRC) The specific IRC strategies are defined in Attachment II - D. The summary of IRC includes the strategy objective and briefly explains how it would be accomplished.

F2.5

Calculated Variables The required FCC calculations are listed in Attachment II - E.

May 1993

F-6

Chevron Corporation


Appendix F

Attachment I-A Philadelphia FCC Honeywell TDC-3000 System Components Fig. F-1

Honeywell TDC-3000 System Components (Courtesy of Honeywell Industrial Controls Division)

Chevron Corporation

F-7

May 1993

Appendix F


Attachment I - B Philadelphia FCC Functional Requirements for Intermediate Regulatory Control IRC Graphic Displays 1.

From the operator graphics display, clearly convey to the operator if the strategy is not functioning.

2.

From the operator graphics display: – –

Note 3.

provide an easy means for the operator to identify when a strategy is ready to be enabled. (All enabling requirements are satisfied.) provide an easy means to change the strategy setpoint (if applicable). Items 1 and 2 are not satisfied by a simple auto/manual switch.

Provide an easy means for the operator to find out: – – –

why a strategy is not working. what constraints have been encountered. what messages have been generated.

4.

Provide on-line facilities for the operator to review the strategy objective(s), inputs and outputs, operator interaction, interlock requirements, and frequency of execution. (A Help display.)

5.

Provide an automatic facility for strategy loading, scheduling, and error recovery such that operator involvement is not required.

6.

Provide an engineer’s support display to allow on-line tuning and adjustment of constraint limits without recompiling the code.

IRC Logic Design

May 1993

1.

Provide the necessary code to meet the objectives stated in the Functional Requirements for IRC Graphic Displays.

2.

Organize the code into an input processing/calculation section and an output section.

3.

Embed comments in the code.

4.

Check strategy inputs for validity.

5.

Check strategy outputs to avoid windup.

F-8

Chevron Corporation


Appendix F

Attachment II - A Philadelphia FCC On-line analyzers The following is a list of on-line analyzers that currently exist at the FCC. 1.

Continuous Regenerator flue gas O2 and CO (installation complete by approx. 10/93)

2.

Lt cat gas oil BP (95% PT) — not in good working order (see below)

3. a.

Lt cat gasoline (T9 bottoms) GC (% C4=) 1

b.

Debutanizer bottoms GC (% C4=)1

4.

Hvy cat gasoline BP (95% PT) (see below)

5.

Absorber stripper overhead GC (% C3=) 1

6. a.

Debutanizer T9 overhead GC (% iC5) 1

b.

Depropanizer bottoms GC (% C3= and % iC5)1

7.

Depropanizer overhead GC (% iC4) 1

8.

Deethanizer bottoms GC (% C2=) 1

(Labels a. and b. indicate that the particular GC analyzes two streams.) In addition, the following services will require complete inferential predictors:

1.

Lt cat gasoline (T9 Bottoms)

–

RVP

Lt cat gas oil

–

Flash Point

Lt cat gas oil

–

95% point

Hvy cat gasoline

–

95% point

This identifies the analyzer services requiring an inferential predictor. In these cases the simpler approach described in Section F1.4 can be used.

Chevron Corporation

F-9

May 1993

Appendix F


Attachment II - B Philadelphia FCC DMC-1: Reactor/Regenerator/Fractionator P&ID drawing numbers are indicated in [ ]

Controlled / Constrained Variables 1.

PDI103

Delta pressure across regenerated catalyst slide valve A [G-22708]

2.

PDI104

Delta pressure across regenerated catalyst slide valve B [G-22708]

3.

FC517

Suction Valve position on J101A Blower [G-22713]

4.

FC519

Steam Valve Position for Turbine Governor on J101B Blower [G-22713]

5.

TT16

Reactor fresh feed furnace: maximum stack temperature (950°F) [G-22712]

6.

TI662/3

Regenerator grid temperature [G-22708]

7.

TI58

Regenerator plenum temperature <1400°F (has not been a limit in the past) [G22708]

8.

PDI105

Delta pressure across spent catalyst slide valve [G-22708]

9.

FC130

Steam drum C-114A: Boiler feed water inlet (valve position) [G-22706]

10.

FC131

Steam drum C-114B: Boiler feed water inlet (valve position) [G-22706]

11.

AI625 A

Regenerator flue gas composition: O2 or ratio CO2/CO [G-22706]

12.

AI629

Hvy cat gasoline 95% PT [G-22727]

13.

AI643

Lt cat gas oil 95% PT [G-22717]

14.

Calc-N9

Lt cat gas oil flash point (inferential calculation)

15.

TC103

Trap tray reflux (valve position) [G-22720]

16.

TI14

Fractionator overhead minimum temperature (235°F) to prevent corrosion [G22720]

17.

PDR-N3

Fractionator delta pressure across entire column to not violate flooding constraint [G-22720]

18.

TI33

Fractionator bottoms temperature (TI91/2) to prevent coking [G-22720]

19.

TI-N1

Decanted oil and HGO to cutter tanks maximum temperature (proposed new DCS input) [G-22719]

20.

FC110

Fractionator distillate to absorber stripper valve position [G-22726]

21.

II23 A/B

Wet gas compressor maximum amperage [G-22725]

22.

PC523

Wet gas compressor suction (valve position) [G-22725]

23.

Calc-N11

Delta pressure across absorber stripper (high limit, cf. DMC-2) (calculation), linking DMC-1with DMC-2

24.

PC201

Absorber stripper tail gas valve position [G-22724]

25.

PC611

Deethanizer feed surge drum (V-2) pressure [G-23606]

26.

—

Levels deemed appropriate by contractor - Flow Smoothing

May 1993

F-10

Chevron Corporation


Appendix F

Attachment II - B (continued) Philadelphia FCC DMC-1: Reactor/Regenerator/Fractionator Manipulated Variables 1.

TC101/2

Riser A & B temperature [G-22708]

2.

FC104

Reactor stripping steam [G-22708]

3.

CCCMast

The New Blower Controls Master should change air flow setpoints - IRC Strategy#1

4.

FC101/2

Reactor total fresh feed flow (calculation) [G-22712]

5.

TRC104

Fresh feed furnace outlet temperature [G-22712]

6.

PC100

Regenerator pressure [G-22708]

7.

FC109

Lt cat gas oil flow [G-22717]

8.

FI424

Lt cat gas oil stripper steam flow or ratio (IRC Strategy #3) [G22719]

9.

TC103

Fractionator top temperature (TC103 - FC103 cascade); no handle on slurry loop [G-22720]

10.

Calc-N12

Fractionator HGO reflux heat duty (proposed IRC strategy #4)

11.

FC11

Fractionator emergency reflux [G-22722]

12.

FC259

Fractionator distillate flow to lean oil still [G-22727]

13.

PC523

Wet gas compressor suction K.O. drum pressure [G-22725]

14.

PC201

Absorber stripper tail gas K.O. drum pressure [G-22724]

Feedforward Variables

Chevron Corporation

1.

FC265

Waste oil feed to reactor [G-22708]

2.

FT581

Fresh catalyst flow rate [G-22713]

3.

Calc-N13

Fractionator intermediate reflux heat duty (calculation)

F-11

May 1993

Appendix F


Attachment II - C Philadelphia FCC DMC-2: Gas Recovery Unit Controlled / Constrained Variables

May 1993

1.

AR667

Absorber stripper overhead composition (% C3=) (calculationvalidated analysis) [G-22724]

2.

TI39

Absorber stripper overhead temperature [G-22726]

3.

TC632

Absorber stripper top chiller TIC203 valve position [G-22727]

4.

Calc-N11

Delta pressure across absorber stripper (high/high limit, cf. DMC-1) (calculation)

5.

Calc-N14

Ratio hvy cat gasoline/lt cat gas oil for lean oil still (calculation)

6.

Calc-N15

Lean oil still delta pressure (calculation) [G-22728]

7.

Calc-N16

Debutanizer delta pressure (calculation) [G-22729]

8.

FC210

Debutanizer bottoms flow valve position. New resized valve may no longer be a constraint [G-22729]

9.

AR633

Debutanizer bottoms composition (% C4=) (calculation-validated analysis) [G-22729]

10.

TC205

Debutanizer bottoms reboiler valve position [G-22721]

11.

PC203

Debutanizer pressure control valve position [G-22729]

12.

PDR153

T9 debutanizer delta pressure (calculation)

13.

AR641

T9 debutanizer overhead composition (% iC5) (calculation-validated analysis)

14.

AR640

T9 debutanizer bottoms RVP (lt gasoline) (calculation ) [G23602]

15.

TC206

Depropanizer reboiler valve position [G-22721]

16.

PDI599

Depropanizer delta pressure [G-22730]

17.

AR635

Depropanizer overhead composition (% iC4) (calculation-validated analysis)

18.

AR634

Depropanizer bottoms composition (% iC5) (calculation-validated analysis) [G-22730]

19.

AR634

Depropanizer bottoms composition (% C3=) (calculation-validated analysis) [G-22730]

20.

FRC206

Depropanizer bottoms level valve position [G-22730]

21.

PI651

Deethanizer feed surge drum (V-2) pressure [G-23608]

22.

AR638

Deethanizer bottoms composition (%C2=) (calculation-validated analysis)

23.

—

Levels deemed appropriate by contractor (flow smoothing level control)

F-12

Chevron Corporation


Appendix F

Attachment II - C (continued) Philadelphia FCC DMC-2: Gas Recovery Unit Manipulated Variables 1.

TC700

Absorber stripper bottoms temperature [G-22727]

2.

FC203

Absorber stripper sponge oil [G-22726]

3.

FC207

Lean oil still reflux flow [G-22728]

4.

PC202

Lean oil still column pressure (PC202 cascaded to FC212) [G22728]

5.

FC208

Debutanizer reflux flow [G-22729]

6.

PC203

Debutanizer column pressure [G-22729]

7.

TC205

Debutanizer bottoms temperature [G-22721]

8.

FC014

Debutanizer T9 reflux flow [G-23602]

9.

PC011

Debutanizer T9 column pressure (PC011 cascaded to FC016) [G-23603]

10.

TC695

Debutanizer T9 bottoms temperature (TC695 cascaded to FC012) [G-23602]

11.

FC209

Depropanizer reflux flow [G-22730]

12.

PC204

Depropanizer column pressure (PC204 cascaded to FC302)

13.

TC206

Depropanizer bottoms temperature [G-22721]

14.

FC653

Deethanizer bottoms temperature (TI665-FC653 Cascade — IRC strategy #6)

Feedforward Variables The following variables build a logical link between DMC-1 and DMC-2

Chevron Corporation

1.

FC110

L.P. distillate to absorber stripper [G-22726]

2.

FC259

L.P. distillate to lean oil still [G-22727]

3.

FC658

H.P. separator liquid flow [G-22724]

F-13

May 1993

Appendix F


Attachment II - D Philadelphia FCC Intermediate Regulatory Control Strategy List

May 1993

1.

Control total air flow to regenerator

2.

Control total fresh feed to reactor risers

3.

Control the ratio of stripping steam flow to lt cat gas oil flow

4.

Control the fractionator HGO reflux heat removal

5.

Lean oil still bottoms liquid level

6.

Control deethanizer temperature

F-14

Chevron Corporation


Appendix F

Attachment II - D (continued) Philadelphia FCC Intermediate Regulatory Control Strategy 1 Objective:

Control total air flow to regenerator.

Today this can be done by adjusting the atmospheric vents FRC118 and FRC661; the operator has to manage the blower suction flow on J101A (motor drive) via FRC517 and blower turbine governor on J101B (turbine driver) via FRC519. However, once the compressor control system is tied into the DCS, the DCS can automatically adjust in the most economic way the different valve positions in order to achieve the desired total air flow rate.

Strategy 2 Objective:

Control total fresh feed to reactor risers. FRC101

fresh feed “A” flow

FRC102

fresh feed “B” flow


Control the ratio of stripping steam flow to lt cat gas oil flow.

This strategy stabilizes the flash point of the LCGO as the LCGO flow rate varies. FC109

Lt cat gas oil flow

FI424

50# steam flow


Control the fractionator HGO reflux heat removal.

Heat removal can be inferred from reflux flow and draw and return temperatures. The reflux flow through the steam boiler is used to manipulate heat removal. HGO reflux heat duty: FC296

HGO reflux flow

TI30

draw temperature (cf. Attachment II-A, III.2)

TI-N6

return temperature (proposed new DCS point)


Lean oil still bottoms liquid level.

Use 2 controllers FRC201 and FRC202. Valve position range control on FRC202 (hvy cat gasoline).


Chevron Corporation

Control deethanizer temperature.

F-15

May 1993

Appendix F


Currently there is only a TI, but no control. Cascade TI665 to FRC653.

May 1993

F-16

Chevron Corporation


Appendix F

Attachment II - E Philadelphia FCC Vendor Calculated Variables / Predictors 1.

Inferential Calculations - validated by Analyzer Readings and self adjusting. The approach to be followed for these calculations is described in Section F1.4.

2.

3.

Chevron Corporation

a.

Absorber stripper overhead composition validated % C3= (AI667)

b.

Debutanizer bottoms composition validated % C4= (AI633)

c.

Debutanizer T9 overhead composition validated % iC5 (AI641)

d.

Debutanizer T9 bottoms composition validated % C4= (AI640)

e.

Depropanizer overhead composition validated % iC4 (AI635)

f.

Depropanizer bottoms composition validated % iC5 (AI634)

g.

Depropanizer bottoms composition validated % C3= (AI634)

h.

Deethanizer bottoms composition validated % C2= (AI638)

Inferential Calculations - Complete inferential predictors where no analyzer exists or it is unreliable. a.

Lt. Cat Gasoline (T9 Bottoms) - RVP

b.

Lt. Cat Gas Oil - Flash Point

c.

Lt. Cat Gas Oil - 95% Distillation BP

d.

Hvy Cat Gasoline - 95% Distillation BP

Other Calculations Required For DMC or Engineering Performance Monitoring a.

Fractionator intermediate reflux heat duty

b.

Fractionator HGO reflux heat duty

c.

Fractionator Slurry reflux heat duty

F-17

May 1993

Appendix F


Attachment II - F Philadelphia FCC DCS Data Base List (not included) and P&ID’s (not included)

May 1993

F-18

Chevron Corporation

Appendix G. Control Objectives Analysis

(Presented at the 1977 Computer Conference by Kirby L. Hadley, Chevron Research Company)

Summary Control engineers are having more and more influence on how refinery processes are operated by designing, building, and installing advanced control systems utilizing digital computers. New control systems will be installed which improve plant profitability providing that (1) the process control engineer clearly understands the problem to be solved and (2) operating management understands what to expect from the new control system. In the last two years, Chevron has developed a formal technique for insuring that this communication occurs without misunderstanding. We call it “Control Objectives Analysis.” The thesis behind “Control Objectives Analysis” is that it is always possible for a group representing operations, process engineering, and control engineering to define and agree on a list of “concise, precise, true-all-of-the-time” statements which define the operating objectives of a process. These objectives are met by manipulating the valves of the plant. This list of “Control Objectives” serves as the basis for the control system design, serves as a yardstick for measuring the success of a new control project, and provides a consistent basis for monitoring ongoing performance. Contents

Page

G1.0 The Game: The Analysis

G-3

G2.0 The Goal: The Control Objectives

G-3

G3.0 Rules of the Game

G-3

G4.0 Role of the Moderator

G-4

G5.0 Opening the Game

G-4

G5.1 Opening Step #1: The Diagram G5.2 Opening Step #2: The Overall Objective G5.3 Opening Step #3: Count the Valves G6.0 The Game of Defining Objectives

G-5

G7.0 Closing the Game

G-5

G7.1 Closing Step #1: Categorizing the Objectives G7.2 Closing Step #2: Back to the Overall Objective

Chevron Corporation

G-1

June 1989

Appendix G

June 1989


G8.0 Conclusions

G-6

G9.0 Credits

G-6

G-2

Chevron Corporation


G1.0

Appendix G

The Game: The Analysis A “Control Objectives Analysis” session is treated as a game. A minimum of four players is required: a moderator, a control engineer, a process engineer, and an operations representative. The game is won when the group reaches the Goal: defining a list of carefully stated “Control Objectives” equal in number to the number of control valves in the process area under consideration. In practice, we’ve made three observations after performing this Game many times:

G2.0

1.

The game is almost always won. Participants come away with a major feeling of accomplishment. There are no individual losers.

2.

It takes three or four hours to analyze a portion of a process containing about 15 valves.

3.

The analysis is an intense exercise in group dynamics. It’s best to limit sessions to half a day.

The Goal: The Control Objectives A properly defined “Control Objective” is a concise, precise, true-all-of-the-time sentence which defines one of the reasons valves are moved in a process. The three characteristics of properly defined objectives are important and need to be understood by the Players. Concise means that the objectives statement should be as short as possible yet still convey all of the necessary information. Excessive words may indicate a lack of understanding of the true objective. Precise means that the statement must actually say what the Players are thinking so clearly that everyone has the same understanding of what the objective means. True-all-of-the-time forces the Players to consider all of the conditions which may arise in the plant. This is particularly important if an advanced control system operating all-of-the-time is to be designed based upon the “Control Objectives.” As noted previously, the Game is over, the Goal is reached, when the number of agreed upon “Control Objectives” statements is exactly equal to the number of valves in the process area being considered.

G3.0

Rules of the Game Three rules have evolved which help keep discussions directed toward the Goal:

Chevron Corporation

1.

The existing control system is not discussed. It is best if a process diagram is available which shows only valves, no control loops. If the available diagram shows loops, Players should be instructed to totally ignore them.

2.

A new control system is not designed. This happens most frequently when a Player suggests an objective such as “adjust the reflux to hold the tower temper-

G-3

June 1989

Appendix G


ature at some target.” Linking the reflux and the temperature together in a single statement implies a loop configuration. 3.

Computers are not mentioned. The Analysis session should not become a forum for the discussion of computer control.

The Moderator of a Game should carefully review these Rules with the Players before the Game begins. We have noted that Players will begin correcting themselves and each other when a Rule is broken.

G4.0

Role of the Moderator An important part of this Game concept is that the Moderator is not dictating to the Players what the objectives of the process are. In fact, we have found it is not necessary for the Moderator to have any knowledge of the type of process under consideration. There are two requirements for a good Moderator: 1.

The Moderator must be able to determine when an objective meets the concise, precise, true-all-of-the-time criteria.

2.

The Moderator must develop the skill to extract from the Players their knowledge of the process.

We have found that the Moderator works best at a blackboard with chalk and eraser. Each objective proposed by a Player undergoes many revisions before it meets all three requirements, is consistent with all of the other objectives, and is approved by all Players.

G5.0

Opening the Game If the Players are familiar with the Game, the Moderator can move the group directly into the steps outlined below. If any Players are new, the Moderator should explain the Game’s Rules and describe the three criteria of a good objective.

G5.1

Opening Step #1: The Diagram A process diagram should be available, simplified as much as possible, which clearly shows the control valves under consideration. The Moderator should ask the Players to explain to him what the process does.

G5.2

Opening Step #2: The Overall Objective The Moderator asks the Players to define an overall objective for this portion of the process. A long rambling paragraph generally evolves which the Moderator writes on the board uncritically. This opening step is primarily intended to break the ice and allow everyone to talk. When the group agrees on the wording for the “overall objective,” it is copied down and set aside to be used again in the final Closing Step.

June 1989

G-4

Chevron Corporation


G5.3

Appendix G

Opening Step #3: Count the Valves The Moderator returns to the diagram and with the group counts the number of control valves. The Moderator then announces the goal: the definition of that same number of concise, precise, true-all-of-the-time Control Objectives.

G6.0

The Game of Defining Objectives Most of the time of an Analysis is spent defining and polishing the individual Control Objectives statements. The persons trained as Moderators at Chevron have developed individual styles. We have found, however, a more or less standard set of questions Moderators use to stimulate thinking. Here are some of them: “Is this proposed Objective precise? Is it really true?” For example, an Objective which says “Minimize fuel consumption” can best be met by shutting the plant down. “Is the proposed Objective met in practice in the plant today—or are there important constraints or overriding considerations which mean we sometimes relax the Objective?” A positive answer here has strong implications to the control engineer’s design of an advanced control system. “Do management signals change so that the plant shifts from one set of objectives to another?” “Can the existing valves achieve the objectives?” “Are there any valves which are normally totally open or totally closed and thus should not be included in the count of valves?” “Does the list of objectives provide enough information to specify the way in which valves should be moved and positioned?” “Is some valve on the diagram not covered by one or more of the objective statements?” “Can we measure the criteria defined by the objectives with measuring devices currently installed?” “New devices?” The questions above are not asked in any particular order. The skill of the Moderator is in directing the group’s thinking toward the completed goal by selecting the appropriate approach at each point through the discussion.

G7.0

Closing the Game When the number of agreed upon Control Objectives matches the number of valves, the Game is over. There are two final steps the Moderator may lead the group through to make the results more meaningful to the participants.

Chevron Corporation

G-5

June 1989

Appendix G

G7.1


Closing Step #1: Categorizing the Objectives It has been discovered that all single objectives statements fall into one of three categories: Type I:

Hold material balance, heat balances

Type II:

Hold operations at management-set targets

Type III:

Balance one objective against another; optimize; minimize or maximize

The group should reorganize the list of objectives and force each into one of these categories. Here are the types of understanding which can come out of this exercise:

G7.2

1.

Those objectives which we have no control over which are fixed by the requirements of the process are identified. These are the Type I objectives.

2.

The targets management should be issuing are identified by the Type II statements. This may lead to a re-evaluation of the methods of target setting.

3.

The Type III objectives are those which point toward engineering and economic studies.

Closing Step #2: Back to the Overall Objective As a wrapup, we have found it useful to return to that Overall Objective paragraph generated at the beginning of the exercise and compare the detailed Control Objectives with that beginning statement. This strongly contrasts preconceived ideas with the finished “Control Objectives List.” Participants frequently choose to rewrite the Overall Objective.

G8.0

Conclusions This simple technique has worked successfully for Chevron again and again. The Company is aggressively moving to implement advanced control projects based on this concept. Additionally, operating management has recognized a tool to clarify thinking and motivate operators in plants not scheduled yet for computer control projects. Control engineering as a discipline has increased its influence on company operations.

G9.0

Credits The “Control Objectives Analysis” technique evolved through the efforts of a number of individuals. Jim Bronfenbrenner and John Westmoreland of Chevron Research, in particular, developed many of the techniques described in this paper.

June 1989

G-6

Chevron Corporation

Appendix H. Safety Interlock System Guidelines

The following people are recognized for their effort in the preparation of this guideline:

Chevron Corporation

Larry Braun

Glen Perkins

Conrad Clark

Don Ogwude

John Coltart

Mike Ramseier

Jerry Current

Chris Rice

James Gray

Mark Schultz

Bill McMurran

Ed Skabowski

Rod Merkel

Parag Shah

Tim Montgomery

Tim Storrs

Jim Palomar

Mark Vaughn

H-1


Appendix H

TABLE OF CONTENTS 1.0 PURPOSE.............................................................................................................................................................. 4 2.0 SCOPE ................................................................................................................................................................... 4 3.0 DEFINITIONS ...................................................................................................................................................... 4 4.0 PROCESS HAZARDS ANALYSIS..................................................................................................................... 6 5.0 SAFETY OBJECTIVES ANALYSIS.................................................................................................................. 7 6.0 SOA VERIFICATION.......................................................................................................................................... 8 7.0 SAFETY REQUIREMENTS ............................................................................................................................... 8 7.1 GENERAL ............................................................................................................................................................. 8 7.2 VOTING STRUCTURES ........................................................................................................................................ 10 8.0 DETAILED DESIGN.......................................................................................................................................... 11 8.1 SENSORS ............................................................................................................................................................ 11 8.2 LOGIC SOLVER................................................................................................................................................... 13 8.3 FINAL CONTROL ELEMENTS............................................................................................................................... 15 8.4 SIS BYPASS ....................................................................................................................................................... 17 9.0 DESIGN VERIFICATION................................................................................................................................. 17 10.0 DCS & GRAPHICS .......................................................................................................................................... 17 11.0 FIELD WORK .................................................................................................................................................. 18 11.1 IDENTIFY SIS EQUIPMENT ............................................................................................................................... 18 11.2 POWER DISTRIBUTION ..................................................................................................................................... 18 11.3 WIRING ............................................................................................................................................................ 18 12.0 APPLICATION SOFTWARE ......................................................................................................................... 19 13.0 VALIDATION................................................................................................................................................... 19 14.0 DOCUMENTATION........................................................................................................................................ 20 14.1 DESIGN ............................................................................................................................................................ 20 14.2 OPERATING PROCEDURES ................................................................................................................................ 20 14.3 PERIODIC TEST PROCEDURES .......................................................................................................................... 21 14.4 MAINTENANCE ................................................................................................................................................ 21 14.5 MANAGEMENT OF CHANGE ............................................................................................................................. 21 15.0 TRAINING ........................................................................................................................................................ 22 15.1 OPERATORS ..................................................................................................................................................... 22 15.2 MAINTENANCE ................................................................................................................................................ 22 15.3 TECHNICAL SUPPORT ....................................................................................................................................... 23 16.0 PSSR................................................................................................................................................................... 23

H-2

July 1996


Appendix H

17.0 PERIODIC TESTING ...................................................................................................................................... 23 18.0 SIS SUPPORT ................................................................................................................................................... 25 18.1 SIS OWNER ..................................................................................................................................................... 25 18.2 GENERAL ......................................................................................................................................................... 25 18.3 SOFTWARE ....................................................................................................................................................... 25 19.0 MODIFICATION ............................................................................................................................................. 26 20.0 AUDIT................................................................................................................................................................ 27 CONTRACTORS...................................................................................................................................................... 27 REFERENCES.......................................................................................................................................................... 28 ATTACHMENT A - SIS LIFE CYCLE ................................................................................................................. 29 ATTACHMENT B - SIS LIFE CYCLE RESOURCE REQUIREMENTS ......................................................... 34

July 1996

H-3


1.0

Appendix H

PURPOSE A Safety Interlock System (SIS) is an instrument and control system used to prevent or mitigate hazardous events, protect people or the environment, or prevent damage to process equipment. A typical system measures process variables (e.g., flow, pressure, temperature) and responds to predetermined hazardous conditions (e.g., loss of flow) by initiating automated field actions (e.g., closing valves). As Chevron installs SIS on equipment we must ensure that the system will properly mitigate events. In addition, we must meet the requirements of OSHA’s Process Safety Management Rule 29 CFR 1910.119 (1). OSHA will look to ISA draft standard 84.01 (2) to define the practices required to design, install, and maintain SIS’s. Chevron will be expected to follow practices it has helped develop, such as the ISA-S84.01-1996. For these reasons, it is important for Chevron to develop a standard philosophy on applying SIS’s that meets the requirements of OSHA and ISAS84.01-1996. This document provides that philosophy.

2.0

SCOPE Many decisions which ultimately decide the reliability of the SIS and the safety of the plant are made throughout the life of the SIS. This document captures the practices that lead to a safe and reliable SIS. It contains design details and reviews that the Project Team should implement, operator documents and training to ensure proper usage of the system, and support issues for maintenance and the system owner. A typical project life cycle is presented and discussed in Attachment A. The guidelines presented in this document are categorized in accordance with the life cycle. This document presents concepts that follow those presented by the American Institute of Chemical Engineers Center for Chemical Process Safety (3, 4). These concepts do not conflict with the CRTC report "Protective Systems for Furnaces and Other Process Units: General Concepts"(5), but they do expand upon them. This document should be used for future SIS applications. In this document, "shall" refers to a requirement for compliance to OSHA, ISA, or Chevron Corporate Best Practices. Where applicable, the source of the requirement is shown in parentheses. Items that do not contain the word "shall" refer to good engineering practices learned from past projects. This guideline does not apply to rotating machinery protective systems.

3.0

DEFINITIONS Availability. The degree to which a system or component is operational and accessible when required for use. Typically, the ratio of system up-time to the total operating time. Diversity. Use of different people, design methods, software languages, functionality, measurement signals, or equipment to perform a common function with the intent to minimize common mode failures.

H-4

July 1996


Appendix H

Diagnostic Coverage. The percentage of faults detectable by embedded on-line diagnostics to the number of total faults. Periodic Test Interval. The time between routine tests performed on a system. These tests are performed to detect failures in the SIS so the system can be restored to an “as new” condition or as close as practical to this condition. Fail-safe Faults. A fault that immediately causes the system to go into a safe state or, in a redundant system, a fault which does not prevent proper and safe control of the process. A failsafe fault may cause a nuisance trip. Fail-safe. Fail-safe refers to the output action of a control system upon a failure. A fail-safe control system is one whose outputs operate in such a manner as to reduce the risk of an accident when a component or circuit failure occurs in the control loops associated with that failure. Fail-dangerous Faults. A fault that prevents the control system from responding to hazard warnings, or can cause a hazardous condition (e.g., Two pressure sensors must read high readings in order to trip the plant. Commonly, called two-out-of-two voting. If one transmitter fails such that it tracks the process, but can not send the high-high trip value, then the SIS is not capable of tripping the plant. For computer based transmitters, like the Honeywell Smart Transmitter, this could occur if the high order bit is stuck at zero). Independent Protection Layer (IPL). A protection layer or combination of protection layers qualify as an IPL when: 1. The protection provided reduces the identified risk by a large amount, that is, at least a 100-fold reduction. 2. The protective function is provided with a high degree of availability, that is, .99 or greater. 3. It has the following characteristics:

July 1996

•

Audit-ability: It is designed to facilitate regular validation of the protective functions. Functional testing and maintenance of the safety system is necessary.

•

Dependability: It can be counted on to do what it was designed to do. Both random and systematic failure modes are addressed in the design. Typically, random failures are caused by hardware component failures. Systematic failures may be caused by design errors, software errors, operator mistakes, and maintenance mistakes.

•

Independence: An IPL is independent of the other protection layers associated with the identified danger, that is, the IPL has no single points of failure, no common-mode failure points, electrical isolation, and physically separated.

•

Specificity: An IPL is designed solely to prevent or to mitigate the consequences of one potentially hazardous event (e.g., a runaway reaction, release of toxic material, a loss of containment, or a fire). Multiple causes may lead to the same hazardous event, and therefore multiple event scenarios may initiate action of one IPL.

H-5


Appendix H

Nuisance Trip. A fault that immediately causes the process to shutdown. Protection Layer. Protection layers typically involve special process designs, process equipment, administrative procedures, the basic process control system and planned responses to imminent adverse process conditions; and these responses may be either automated or initiated by human actions. For human actions, certain human-machine interactions must be considered such as, whether sufficient information exists to determine the process condition and whether sufficient time exists to deduce the process condition and mitigate it. Reliability. The probability that a device will function without failure over a specified time period or amount of usage. Note: Reliability measures the failure rate of components. Availability measures the failure of a protective function. They are not the same for redundant systems. Safety. The expectation that a system does not, under defined conditions, lead to a state in which human life, limb and health, economics or environment are endangered. Note: For system safety, all causes of failures which lead to an unsafe state shall be included; hardware failures, software failures, failures due to electrical interference, due to human interaction and failures in the controlled object. Some of these types of failures, in particular random hardware failures, may be quantified using such measures as the failure rate in the dangerous mode of failure or the probability of the protection system failing to operate on demand. The system safety also depends on many factors which cannot be quantified but can only be considered qualitatively. (IEC 65A(Secretariat)122, August 1991). Safety Integrity Level (SIL). A distinct level of performance expected from the Safety Interlock System. Both qualitative and quantitative distinctions are made. Validation. The process of evaluating a system or component at the end of the development process to determine whether it satisfies specified requirements. Verification. The process of evaluating a system or component to determine whether the products of a given development phase satisfy the conditions imposed by the previous phase.

4.0

PROCESS HAZARDS ANALYSIS 4.1

H-6

A Process Hazards Analysis (PHA) shall be conducted to identify hazardous events and the layer(s) of protection which exist to prevent or mitigate its occurrence (OSHA 1910.119). Typically, this is a Hazards and Operability Study (HAZOP). Each event shall be assigned a risk level by using the Chevron Refining Risk Ranking Guidelines (Contact the Fire Process Safety Management Team within CRTC for further details on HAZOP’s and risk ranking).

July 1996


5.0

SAFETY OBJECTIVES ANALYSIS 5.1

Following the HAZOP or some similar hazards analysis, a Safety Objectives Analysis (SOA) shall be conducted to define the SIS requirements. Refer to the CRTC SOA Report (6) for guidance on conducting this analysis.

5.2

The SOA shall address and document the functional requirements for the SIS (ISA-S84.011996). It shall include:

5.3

July 1996

Appendix H

•

A definition of the safe state of the process.

•

Identify and document all Independent Protection Layers (IPL) for each hazardous event.

•

Process variables used by the SIS to bring the process to a safe state (i.e., SIS inputs) and their trip points.

•

The final control elements used to bring the process to a safe state (i.e., SIS outputs) and their actions.

•

Functional relationship between process inputs and outputs (i.e., cause and effect matrices, startup permissives, etc.)

•

Requirements due to loss of energy sources.

•

Response time requirements for the SIS to bring the process to a safe state.

•

Response actions to be taken upon detection of a SIS failure.

•

Requirements for trip reset (i.e., field and control room).

•

Operator interface requirements (i.e., field panel, DCS graphics, alarm panel, etc.)

•

Special requirements for start-up, such as bypass conditions.

•

Requirements if nuisance trips are hazardous. A nuisance trip of the plant under study may cause hazards in other plants.

•

Identify the requirements for on-line testing and maintenance (i.e., type of tests and test interval).

The defined Safety Integrity Level is in accordance with ISA-S84.01-1996. It recognizes three levels of system performance. The quantitative goals are expressed here in Safety Interlock System availability and the average Probability of a Fail-dangerous fault (PFDavg): Safety Availability Range

PFDavg. Range

SIL 1

.9 - .99

10-1 to 10-2

SIL 2

.99 - .999

10-2 to 10-3

SIL 3

.999 - .9999

10-3 to 10-4

H-7


Appendix H

It is possible to calculate these numbers accurately for the hardware portion of the SIS, but no accurate calculation can be made for systematic errors such as software errors, design errors, operator errors, maintenance errors, poorly designed interfaces, etc. Therefore, this guideline uses a qualitative approach based on good engineering judgment and past experience. The required configuration, diagnostic coverage, and test interval to meet a given SIL has been calculated for a number of configurations and is presented in the respective sections of Sensors, Logic Solver, Final Control Elements. Guidelines to reduce systematic errors apply to all SIL. The Chevron Refining SIS classifications defined in the CRTC report (5) have the following translation in terms of ISA-S84.01-1996 Safety Integrity Levels:

6.0

Class A = SIL 2 or 3 (depends on design features not mentioned in report (5)).

•

Class B = SIL 1 or 2 (depends on design features not mentioned in report (5))

•

Class C = Below SIL 1 (not independent of the control system).

SOA VERIFICATION 6.1

7.0

•

One person should be responsible for resolving all SOA concerns and getting the appropriate reviews. Those most experienced with the process and responsible for the safe operation and financial health of the plant should review the SOA. This may include the Technical Manager, Process Engineers, Head Operators, Maintenance personnel, Shift Supervisors, Superintendents, and the CRTC SIS Expert.

SAFETY REQUIREMENTS 7.1

General 7.1.1

H-8

The Safety Interlock System includes all elements from the sensor to the final element, including inputs, outputs, power supply, and logic solvers. The user interface physical devices and portions of software (i.e., graphics and database) that provide SIS information are included as part of the SIS. The entire SIS must be considered when trying to meet a particular SIL. A Safety Requirements Specification or similar document should be written to capture the SIS design philosophy. This includes items such as: •

Independence from control system.

•

Diversity from control system.

•

Redundancy.

•

Software Reliability Considerations.

July 1996


July 1996

Appendix H

•

Architecture; hardware and software.

•

Periodic Test Interval.

•

Voting Configurations (e.g., 2oo3 sensor, 1oo2 logic solver, 1oo1 final element).

•

Failure rates and modes; single and common-cause.

•

Equipment Reliability.

•

User Interface.

•

Wiring Practices.

•

Documentation.

•

Training; maintenance and operators.

•

Diagnostics; system and application.

•

Power Sources.

7.1.2

Keep the functions as simple as possible and use the least number of functions to mitigate the incident. Over-designed systems increase the likelihood of introducing errors during design or while making modifications, and are more difficult to validate.

7.1.3

The Safety Interlock System shall be diverse and independent from the process control system.

7.1.4

SIS Best Practices developed by the Chevron Refining Best Practice teams shall be implemented (7).

7.1.5

Provide at least one manually actuated push button in the field or in the control room independent from the logic solver to bring the process to a safe state (ISAS84.01-1996).

7.1.6

The SIS shall be designed to move to the safe state on loss of energy (e.g. electrical power, instrument air), unless energy is required to bring the process to a safe state. Typically, the SIS will be deenergize-to-trip operation.

7.1.7

The SIS final control elements shall remain in their protective state after a trip until manually reset, even if the trip initiators return to their normal operating condition (ISA-S84.01-1996). The operator actuated trip reset can be field or console actuated.

H-9


7.2

Appendix H

7.1.8

If manual adjustments of equipment that manipulates trip parameters are required, warning signs shall be displayed. (e.g., manual adjustment of minimum flow bypass pressure settings can cause trips if low-low pressure trips are implemented in the SIS).

7.1.9

Where possible, split redundant inputs and redundant outputs to different modules. If only one module is used, connect the redundant inputs to different field termination assemblies if available (e.g., For TRICON systems, wire one redundant input to External Termination Panel [ETP] points 1 - 16 and the other to ETP points 17 - 32. ETP sections have separate power distribution and fuses).

7.1.10

Non-safety related functions are not allowed in the SIS (ISA-S84.01-1996).

7.1.11

The desired safe state of each component required for the safety function shall be defined (ISA-S84.01-1996).

7.1.12

The SIS shall operate within the time required to prevent the incident. The sum of the lag times for each component (i.e., measurement device, logic unit, final control element) shall be less than the allowable time delay for completing the protective function.

Voting Structures 7.2.1

The SIS voting structures shall be in accordance with Tables 1, 2, and 3 to meet the required SIL. For example, to implement a SIL2 interlock, select a SIL2 sensor configuration logic solver configuration and final control element configuration from the Tables

7.2.2

For 1oo1 systems that have similar process control transmitters, configure the process control system to alarm when unacceptable deviations are detected between the SIS transmitter and the process control transmitter.

7.2.3

When using two-out-of-two (2oo2) voting the following conditions shall be met: •

The SIS compares sensor values and alarms deviations.

•

Upon detecting a sensor failure the voting logic converts to 1oo1; or the availability of the SIS hardware is calculated and meets the desired SIL.

The concern is that 2oo2 voting dramatically increases the likelihood (i.e., greater than 1,000 fold) of a sensor having a Fail-dangerous fault. The benefit of 2oo2 voting is the likelihood of a nuisance trip is reduced (i.e., about three fold). 7.2.4

H-10

For 2oo3 voting systems, one of the transmitter process connections may be shared with the process control system. Isolation valves shall be provided so each transmitter can be maintained and tested separately.

July 1996


8.0

DETAILED DESIGN 8.1

July 1996

Appendix H

Sensors 8.1.1

Switches shall not be used to measure process trip parameters.

8.1.2

Each transmitter shall have process connections separate from the process control transmitter, unless redundant SIS transmitters are used.

8.1.3

Refer to the section "Standard Drawings and Forms" in Volume 1, part 2 of the manual for information regarding installation.

8.1.4

Process control transmitters and SIS transmitters measuring the same process variable shall have the same range and calibration.

8.1.5

Smart transmitters shall be configured to drive their output signal away from the SIS setpoint for self-diagnosed failures. The SIS shall then generate a SIS Problem Alarm to flag a faulty transmitter.

8.1.6

Consider using diverse measurements where practical to mitigate a SIL2 or 3 event. For example, use low fuel gas pressure and low fuel gas flow measurements to detect loss of fuel gas to the burner.

8.1.7

Redundant transmitters shall have separate root valves.

8.1.8

Flow. Common orifice plates may be shared to measure a SIS and a control system signal. Separate SIS process taps and transmitters shall be provided.

8.1.9

Temperature. Where practical, a separate thermowell should be used for SIS temperature measurement. However, redundant SIS thermocouples sharing the same thermowell are acceptable.

8.1.10

The results presented in Table 1 are the minimum requirements for sensor configurations, diagnostic coverage and periodic test interval to meet a given SIL and Nuisance Trip Rate. See reference (8) for calculations.

H-11


Appendix H

Table 1 - Sensor Requirements Safety Integrity Level (SIL)

1

2

3

Note:

Sensor Configuration

Diagnostic Coverage

Periodic Test Interval (Months)

Nuisance Trip Rate (Years)

Single

Low (60%)

9

5.5

Single

Medium (90%)

24+

4.6

Single

High (99%)

24+

4.4

Dual (2oo2)

Low

5

480

Dual (2oo2)

Medium

19

540

Dual (2oo2)

High

24+

4100

Single

Medium

4

4.6

Single

High

24+

4.4

Dual (2oo2)

Medium

2

4500

Dual (2oo2)

High

19

5000

Dual (1oo2)

Low

24+

4.4

Dual (1oo2)

Medium

24+

4.4

Dual (1oo2)

High

24+

4.4

Triple (2oo3)

Low

16

56

Triple (2oo3)

Medium

24+

150

Triple (2oo3)

High

24+

1400

Dual (2oo2)

High

2

30000

Dual (1oo2)

Low

8

4.4

Dual (1oo2)

Medium

24+

4.4

Dual (1oo2)

High

24+

4.4

Triple (2oo3)

Low

<5

160

Triple (2oo3)

Medium

20

170

Triple (2oo3)

High

24+

1400

1. Safety Integrity Level 3 systems are not expected at Chevron facilities. 2. For SIS’s with 2oo2 voting, diagnostic coverage is enhanced by comparing the SIS sensors to a third DCS transmitter and alarming deviations. 3. In general, Chevron SIS’s meet SIL 2 by using a 2oo2 configuration, 99% Diagnostic Coverage and quarterly periodic testing.

H-12

July 1996


8.2

Logic Solver 8.2.1

July 1996

Appendix H

The SIS logic solver shall be separate and diverse from the process control system. This requirement will help to: •

Minimize common mode failures between the process control system and SIS equipment.

•

Minimize access to the SIS to prevent erroneous changes.

•

Allow access to the process control system without compromising safety.

8.2.2

The SIS logic solver shall be installed in an environmentally controlled building which meets the manufacturers recommended requirements.

8.2.3

Sharing the logic solver to protect multiple pieces of equipment or plants shall be analyzed to understand the operational and maintenance requirements of the plant. Factors to consider include: •

Turnaround schedules for each piece of equipment.

•

Likelihood and acceptability of maintenance or operations shutting down all equipment connected to the logic solver.

•

Likelihood and acceptability of software errors introduced while programming the SIS for one piece of equipment causing the SIS for the other piece of equipment not to initiate a shutdown when needed.

8.2.4

When different SIL’s are implemented using the same logic solver, the logic solver design shall adhere to the highest SIL.

8.2.5

Security access (i.e., password levels, keys, etc.) shall be implemented to limit access to the logic solver software.

8.2.6

The results presented in Table 2 are the minimum requirements for logic solver configurations, diagnostic coverage and periodic test interval to meet a given SIL and Nuisance Trip Rate. See reference (8) for calculations.

H-13


Appendix H

Table 2 - Logic Solver Requirements (SIL)

1

2

3

H-14

Logic Solver Configuration (CPU - I/O Modules)

Diagnostic Coverage



Single - Single

Low (60%)

1

1.9

Single - Single

Medium (90%)

6

1.7

Single - Single

High (99%)

48+

1.6

Dual - Single

Low

1

3.5

Dual - Single

Medium

6

7.2

Dual - Single

High

48+

10

Dual - Dual (2oo2)

Medium

3

180

Dual - Dual (2oo2)

High

30

290

Dual - Dual (1oo2)

Low

16

1.0

Dual - Dual (1oo2)

Medium

36

0.8

Dual - Dual (1oo2)

High

36

0.8

Triple - Triple (2oo3)

Low

9

8.7


Medium

36

16


High

48+

120

Single - Single

High

6

1.6

Dual - Dual (2oo2)

High

3

1300

Dual - Single

High

6

10

Dual - Dual (1oo2)

Low

5

1.0

Dual - Dual (1oo2)

Medium

18

0.8

Dual - Dual (1oo2)

High

36

0.8


Low

3

25


Medium

12

30


High

48+

170

Dual - Dual (1oo2)

High

36

0.8


High

36

170

July 1996


8.3

July 1996

Appendix H

Final Control Elements 8.3.1

Applications which require redundant valves may share a valve with the process control system. The shared valve (e.g., process control valve) shall be suitable for both the control and the safety application.

8.3.2

The solenoid valve shall interrupt the air supply between the valve actuator and the valve positioner.

8.3.3

The solenoid shall have a manual reset.

8.3.4

Loss of air supply to the Chopper valve shall be detected. Upon loss of air the SIS logic shall require a complete start-up cycle before allowing the Chopper valve to open.

8.3.5

The results presented in Table 3 are the minimum requirements for final control element configurations, diagnostic coverage and periodic test interval to meet a given SIL and Nuisance Trip Rate. See reference (8) for calculations.

H-15


Appendix H

Table 3 - Final Control Element Requirements (SIL)

Final Control Element Configuration

Diagnostic Coverage



1

Single

Low (60%)

13

5.5

Single

Medium (90%)

24+

4.6

Single

High (99%)

24+

4.4

Dual (2oo2)

Low

6

400

Dual (2oo2)

Medium

24+

440

Dual (2oo2)

High

24+

4100

Single

Low

1

5.5

Single

Medium

5

4.6

Single

High

24+

4.4

Dual (2oo2)

High

24+

4100

Dual (1oo2)

Low

24+

4.4

Dual (1oo2)

Medium

24+

4.4

Dual (1oo2)

High

24+

4.4

Triple (2oo3)

Low

19

50

Triple (2oo3)

Medium

24+

150

Triple (2oo3)

High

24+

1400

Dual (2oo2)

High

2

30000

Dual (1oo2)

Low

10

4.4

Dual (1oo2)

Medium

24+

4.4

Dual (1oo2)

High

24+

4.4

Triple (2oo3)

Low

6

140

Triple (2oo3)

Medium

24

150

Triple (2oo3)

High

24+

1400

2

3

H-16

July 1996


8.4

9.0

SIS Bypass 8.4.1

All bypasses shall generate an SIS bypass alarm.

8.4.2

Bypassing of final elements to test or perform maintenance while on-line is acceptable when means are taken to provide safety coverage until tests or repairs are completed.

8.4.3

All bypasses shall be administratively controlled. The SIS shall be in bypass no longer than what is required for testing or to fix a problem.

DESIGN VERIFICATION 9.1

10.0

Appendix H

Those involved in operating and maintaining the SIS shall review the detailed design for accuracy and operational acceptability. A walk through of the design, and major safety and reliability decisions is often useful. Those that should attend the design would include Process Engineers, operators, Head Operators, Maintenance personnel, and the CRTC SIS Expert.

DCS & GRAPHICS 10.1 User interfaces which display SIS information to an operator (e.g., graphics, field mounted panels, push buttons, hardwired anunciators, and printers) shall be provided. 10.2 SIS graphics shall continuously display the following information: •

SIS measurements.

•

"First-out" indication for each final control element.

•

SIS alarms.

•

SIS bypasses.

10.3 Graphics shall be designed and configured using the sub-pictures from Chevron's Display and Alarm Guidelines for the Honeywell TDC 3000 Rev. 2.0 (9). The information shall be presented whether the user interface is Honeywell, Wonderware, or another type. 10.4 The following alarms shall be provided:

July 1996

•

SIS Bypass Alarm - any protective function is not capable of taking the desired action (e.g., safety valve is bypassed, SIS is bypassed).

•

SIS Problem Alarm - SIS failure has been detected by vendor supplied diagnostics or application program diagnostics.

H-17


Appendix H

•

Pre-Trip Alarm - any process condition that is approaching it's protective setpoint.

•

SIS Shutdown Alarm - SIS protective action has occurred.

•

SIS First-out Alarms - alarm which identifies which of the SIS inputs has initiated the protective action. Each final control element will have its own set of First-out alarms.

10.5 The process control system shall not write to the SIS except for limited periods of time during on-line testing.

11.0

FIELD WORK 11.1 Identify SIS Equipment 11.1.1

Clearly distinguish SIS devices from process control instrumentation (e.g., paint transmitter covers red, use wire with red sheath, use highly visible red tags to identify SIS equipment). This will reduce the chance of maintenance errors.

11.2 Power Distribution 11.2.1

The SIS shall have backup power supplied by an Uninterrupted Power Supply (UPS). Refer to CRTC UPS Guideline (10) for information regarding UPS selection and design.

11.2.2

Two independent feeders shall be provided to power the SIS logic solver and I/O.

11.2.3

Where practical, provide redundant power to field devices. Use 24 VDC output modules which provide continuous on-line diagnostics. Modules which use a “pulse” to determine continuity of the output wiring are preferred. 12O VAC modules can not provide this technology.

11.2.4

Redundant SIS elements (e.g., trip initiators, I/O modules, final control elements, solenoids) shall be powered from separate circuits.

11.3 Wiring

H-18

11.3.1

Where practical, use dedicated cables and junction boxes for SIS wiring. As a minimum, the cables and terminal strips shall be separated from the process control cables and terminal strips. This will help reduce maintenance errors and common mode failures.

11.3.2

All SIS cables, terminal strips, and junction boxes shall be clearly distinguished from others.

July 1996


12.0

Appendix H

APPLICATION SOFTWARE 12.1 The application program shall be formally structured. The structure shall be designed for a clear flow of major logic paths to enhance readability. The program shall be clear, concise, and the logic easy to follow. Program clarity shall not be sacrificed for program efficiency (e.g., do not use subroutine functions when a ladder logic rung will suffice, show all steps in a calculation even if more ladder logic rungs are required). 12.2 Time delays, analog trip points, and other designated variables shall be assigned and accessed via internal memory variables. These variables shall be assigned names and values at the beginning of the program. The variables shall be used in the program so that their value only need to be changed in one location in the application program. 12.3 All trip initiators shall be connected to time delay timers to minimize noise and temporary spikes. Proper time delays shall be determined by process dynamics. 12.4 The program shall provide "first-out" capture and indication for each final control element. 12.5 “GOTO” functions shall not be used. Typically, they cause the logic to serpentine up and down the program, leading to difficulty in understanding the program flow. 12.6 Use a consistent tagging convention throughout the program. 12.7 Each logical set of rungs shall contain comments that describe the function and purpose. 12.8 Redundant SIS transmitters shall be compared and unacceptable deviations shall generate an SIS problem alarm. 12.9 When maintenance bypasses are used, the application program shall be configured so that the SIS measurement alarm is not disabled. 12.10 See reference (11) for an example.

13.0

VALIDATION 13.1 For new systems, a Factory Acceptance Test (FAT) shall be conducted to verify that the integrated design meets the design requirements. The FAT shall use a written test procedure to at least test the following :

July 1996

•

Application logic.

•

Sensor variables’ range and calibration.

•

Redundancy features.

•

Operator interface, if available.

•

On-line testing features.

•

Diagnostics.

H-19


Appendix H

For each test within the FAT, the written procedure should describe the purpose of the test, the initial state of all inputs and outputs, the action to take, and the expected outcome of all inputs and outputs. All discrepancies between the expected outcomes and the test results shall be documented and resolved. 13.2 A Site Acceptance Test (SAT) shall be conducted to verify that the final installed design meets the design requirements. A test procedure, similar to the FAT procedure shall be written. All discrepancies between the expected outcomes and the test results shall be documented and resolved. The SAT may be a smaller set of tests than the FAT. However, all field instruments should be exercised at least once during the SAT.

14.0

DOCUMENTATION 14.1 Design 14.1.1

Documentation shall be developed and maintained which provides design information about the SIS. The design documentation shall include the following (OSHA 1910.119): •

P&ID's which show all SIS instrumentation.

•

List of critical values (i.e., pretrip, trip, maximum and minimum range, minimum fire trips).

•

Instrument Loop and Wiring Diagrams.

•

SIS design requirements (e.g., logic diagrams or cause and effect matrix with written functional description).

•

Written rationale of how SIL and SIS configuration was selected.

•

Logic solver configuration documents (e.g., Module configuration drawing; ladder logic listing; database listing of all points and their configuration; graphics printouts; and version number listing for all hardware, firmware, and software)

•

Modification history information.

14.2 Operating Procedures 14.2.1

H-20

Operating procedures describing the SIS functionality shall be developed and maintained (OSHA 1910.119). As a minimum, the operating procedures shall include procedures on: •

Initial start-up.

•

Normal operations.

•

Temporary operations (e.g., bypass mode).

July 1996


•

14.2.2

Appendix H

Emergency shutdowns. −

conditions of emergency shutdown.

−

assignment of shutdown responsibility.

−

verification that shutdown is executed safely and timely.

•

Normal shutdown.

•

Start-up after a shutdown.

•

Operating limits. −

consequences of deviations.

−

steps required to correct or avoid deviations.

Operating procedures shall be readily accessible and reflect current operating practices (OSHA 1910.119).

14.3 Periodic Test Procedures 14.3.1

Written procedures shall be provided to periodically test and inspect the SIS. All tests shall be recorded and kept. Each record shall include the following information (OSHA 1910.119): •

Date of inspection or test.

•

Name of person performing inspection or test.

•

Instrument number or other identifier.

•

Description of inspection or test performed.

•

Results of inspections or test.

14.4 Maintenance 14.4.1

Procedures shall be developed which describe how to perform maintenance on the SIS hardware and software. The procedure should include component data sheets and manufacturers suggested maintenance for all components of the SIS. It is acceptable for the procedure to reference a qualified person to contact in order to complete the maintenance activity (OSHA 1910.119).

14.5 Management of Change 14.5.1

July 1996

A Management of Change (MOC) procedure shall be written which is specific to the SIS. It should address the following (OSHA 1910.119): •

Purpose of change.

•

Impact on safety.

H-21


15.0

Appendix H

•

Documents to modify.

•

Testing and reviews required to validate new software or new hardware.

•

Time period required to modify SIS and operational concerns while SIS is being modified and tested.

•

Authorizations required to make the change.

TRAINING 15.1 Operators 15.1.1

The Operator training program shall include the following as a minimum (OSHA 1910.119): •

Purpose of the SIS.

•

SIS functions and the event they prevent.

•

Operating procedures for SIS-related tasks.

•

On-line test procedure.

•

Description of SIS equipment and quick visit to all equipment.

•

Review of operator interfaces’ (e.g., graphics, push buttons, field alarm panels) functionality and purpose.

•

Proper response to alarms.

•

Proper response to SIS failures.

15.2 Maintenance 15.2.1

H-22

The maintenance training program shall include the following as a minimum: •

SIS maintenance procedures.

•

On-line test procedure.

•

Description of SIS equipment and quick visit to all equipment.

•

Proper maintenance of SIS components (logic solver, transmitters, valves, etc.).

July 1996


15.2.2

Appendix H

Special training is required for the person or people responsible for performing software maintenance. This should include training in the following areas: •

SIS programming device commands.

•

SIS application program.

•

SIS user interface graphics.

•

SIS user interface database configuration.

15.3 Technical Support 15.3.1

16.0

Those responsible for the proper operation of the SIS should attend the maintenance or operator class. This may include Process Engineers, Control System Technicians, or Instrument and Control Technicians.

PSSR 16.1 The Pre-Startup Safety Review (PSSR) is a final safety review required for new facilities and for modified facilities when the modification is significant enough to require a change in the process safety information (OSHA 1910.119). The PSSR shall include the following:

17.0

•

Verification that the SIS was constructed and installed in accordance with design specifications and drawings.

•

Safety, operating, maintenance and emergency procedures pertaining to the SIS are in place and are adequate.

•

Process Hazard Analysis (e.g., HAZOP, What-If) recommendations that apply to the SIS have been resolved or implemented.

•

Operator, maintenance, and technical personnel training has been completed and those involved are confident they can operate and maintain the SIS.

PERIODIC TESTING 17.1 The SIS shall be designed to allow periodic on-line testing of the entire system from process measurement to the final actuating device without interrupting the process (ISA-S84.01-1996). On-line test facilities are not required for processes which shutdown more often then the test interval. For this case, testing may be completed at turnaround. The test interval shall be established by good engineering judgement based on general reliability data. 17.2 A form shall be developed to maintain records of the inspections or tests. The following shall be recorded and kept (OSHA 1910.119):

July 1996

H-23


Appendix H

•

Date of test or inspection.

•

Name of the person who performed the test or inspection.

•

Serial number or other unique identifier of equipment (e.g., loop number, tag number, equipment number, user-approved number, etc.) tested or inspected.

•

Results of test or inspection (i.e., "as-found" and "as-left" condition).

•

A description of test performed including required test interval.

17.3 The following terms may be used to describe the testing results: "Fail" - a SIS which is not able to provide its protective function. Redundant systems with failures but which are still able to perform their protective function shall not be classified as "Fail". "Pass" - a SIS which is able to provide it's protective function. 17.4 All equipment deficiencies found during periodic testing or otherwise shall be corrected before further use of the facility. In some cases, it may not be necessary that deficiencies be corrected before further use, as long as deficiencies are expeditiously corrected in a safe and timely manner, when other steps are taken to ensure safe operation (OSHA 1910.119). 17.5 An end-to-end functional test (i.e., test all safety functions from the sensor to the final control element for proper action) shall be completed. The test interval shall be more frequent than that specified in Table 1, 2 and 3 for the given SIL. If practical, the final element (e.g., Chopper valve, fan motor) or at least the final initiating element (e.g., solenoid valves) should be exercised. Note:

The logic unit can be used to expedite end-to-end functional testing. Typically, the SIS will have additional code to automate much of the test function. The operator initiates the test and verifies the results. Although this expedites testing, there are some concerns. The code required to implement automatic testing makes the SIS program much more complicated. In doing so, it is more likely that errors will be introduced into the program during design or modification. The other concern is writing from the operator interface to the SIS. This should be kept to a minimum. Automatic testing requires operator initiation, which is accomplished by writing from the operator interface to the SIS. CRTC has developed an automatic testing program called “QuickTest.” An analysis of the benefits and risk of QuickTest should be made before choosing to use it.

17.6 All Chopper valves shall be checked for tight shutoff at least once every quarter. Reliability data may be used to support longer test intervals. 17.7 Typically, the sensors are the most unreliable component in a SIS. It is good practice for operators to perform a visual check of all SIS sensor signals at the beginning of their shift.

H-24

July 1996


18.0

Appendix H

SIS SUPPORT 18.1 SIS Owner 18.1.1

SIS ownership by a single person or single group is required. The SIS Owner is responsible for completion of all SIS related tasks. Typical tasks include ensuring proper safety functions are selected and implemented, proper level of operator and maintenance training is achieved, periodic system testing is performed, all SIS documentation and procedures are accurate, and the current system represents the tested system (i.e., the current configuration of the system contains the exact version numbers of hardware and software that were last tested).

18.1.2

The SIS Owner shall be competent in using the logic unit and configuring the graphics. Training shall be provided to ensure this competence.

18.1.3

The SIS Owner is responsible for ensuring that the personnel involved in the modification are adequately trained and competent to perform their assigned tasks.

18.2 General 18.2.1

All safety interlock systems shall be included in the refinery I&E reliability program. The program shall include: •

A listing of each SIS, its SIL classification, and its planned test interval.

•

A management system to track testing schedule, testing results and repairs made to each system.

18.2.2

The SIS programming device shall not be left connected to the SIS during normal operation. It should be connected to perform maintenance and testing only, then immediately disconnected.

18.2.3

No points in the SIS shall be disabled ("forced") during normal operation. Points shall be forced only for maintenance, testing, or to circumvent an error in the interim of fixing the error. Once these activities are completed all points shall be enabled (ISA-S84.01-1996).

18.3 Software 18.3.1

July 1996

In this document, the term “software” refers to all software required to operate the SIS. This includes: •

The application software, such as ladder logic. Typically, written by Chevron or a contractor.

•

The database, such as point configurations. Typically, written by Chevron or a contractor.

H-25


19.0

Appendix H

•

The graphics. Typically, written by Chevron or a contractor.

•

The embedded software. For a SIS with an operator interface (e.g., Tricon/Honeywell or Siemens/Wonderware), this is the operating system, I/O module software, communication software, translators, etc. provided by the vendor.

18.3.2

Software maintenance shall not be performed on-line, unless it fixes a SIS error. However, it is acceptable to add software to an existing SIS if the software is independent of the software that currently exists in the SIS.

18.3.3

Embedded software (i.e., vendor developed software and firmware) shall not be changed after validation of the SIS, unless it fixes an error that impacts the safety performance of the SIS.

18.3.4

At least two copies of the latest program shall be archived on removable media. The backup copies shall be stored in two different locations (e.g., control room and office of person responsible for SIS’s).

MODIFICATION 19.1 All changes after the SIS is validated shall follow a Management of Change (MOC) procedure. 19.2 Changes designated as “replacement in kind” do not require a MOC. Things that are replacement in kind would be: •

Replacing I/O modules.

•

Upgrading firmware.

•

Replacing transmitters with the same or similar types (e.g., replacing a Honeywell Smart with a Rosemount).

Things that are not replacement in kind and therefore require a MOC are:

H-26

•

Significant changes in technology (e.g., upgrading a relay based SIS to a PLC based SIS).

•

Changes in safety functions. −

adding functions.

−

deleting functions.

−

changing a trip point.

July 1996


−

changing an input delay timer value.

−

See reference (12) for an example.

Appendix H

19.3 See Reference (12) for an example.

20.0

AUDIT An SIS Audit should be performed at start-up and once a year to verify that the SIS meets the requirements of this document. See CRTC SIS Checklist (13) for an example.

21.0

CONTRACTORS 21.1 Use contractors who specialize in SIS design. For further details on selecting an SIS engineering contractor see reference (14). 21.2 The Design Contractor shall provide field inspection during construction and equipment installation. They shall also participate in the SIS validation and debugging of SIS equipment during plant startup.

July 1996

H-27


Appendix H

REFERENCES 1. Process Safety Management of Highly Hazardous Chemicals; Explosives and Blasting Agents; Final Rule, OSHA 29 CFR 1910.119 , Occupational Safety and Health Administration, 1992. 2. ISA-S84.01-1996 Application of Safety Instrumented Systems for the Process Industries, Instrument Society of America ,1996. 3. Guidelines for Engineering Design for Process Safety, American Institute of Chemical Engineers, Center for Chemical Process Safety, September 1992. 4. Guidelines for Safe Automation of Chemical Processes, American Institute of Chemical Engineers, Center for Chemical Process Safety, 1993. 5. Leonard, Protective Systems for Furnaces and Other Process Units: General Concepts, Richmond, California: Chevron Research and Technology Company, Monitoring and Control Systems Unit, March 26, 1992. 6. A Method for Performing Safety Objectives Analysis, Richmond, California: Chevron Research and Technology Company, Monitoring and Control Systems Unit , December 2, 1994. 7. Best Practice For Safety Shutdown Systems, Sulfur Recovery Unit, Chevron U.S.A. Products Company, Sulfur/Amine Best Practice Team, July 19, 1993. 8. ISA-TR84.02, Draft 4 Electrical (E) / Electronic (E) / Programmable Electronic Systems (PES) for Use in Safety Applications - Safety Integrity Evaluation Techniques, Instrument Society of America, 1995. 9. Chevron's Display and Alarm Guidelines for the Honeywell TDC-3000 Rev. 2.0, Richmond, California: Chevron Research and Technology Company, Monitoring and Control Systems Unit, May 1993. 10. UPS Design and Application Guide, Richmond, California: Chevron Research and Technology Company, Materials and Equipment Engineering Unit, March 1992. 11. "QuickTest" Version 3.0, Richmond, California: Chevron Research and Technology Company, Monitoring and Control Systems Unit, February 1994. 12. Pascagoula Refinery, Change Procedure for Safety Interlock Systems, Pascagoula, Mississippi, March 2, 1995. 13. James V. Palomar, Safety Interlock System Checklist, Richmond, California: Chevron Research and Technology Company, September 1, 1995. 14. James V. Palomar, SIS Engineering Consultant Considerations, Richmond, California: Chevron Research and Technology Company, November 1, 1994.

H-28

July 1996


Appendix H

ATTACHMENT A - SIS LIFE CYCLE A typical SIS Life Cycle including duration and resources is shown in Figure 1. Each box represents a Life Cycle Phase. Following the figure is a detailed discussion of each phase. Attachment B discusses typical resource allocations for each phase of the life cycle. Process Hazard Analysis: HAZOP

Duration Phase Name Resources

Safety Objectives Analysis

A

Modify

2 days

Resources Available

All

PE OR MR PrE EC SIS SO

10 days SOA Verification All OK

Project Engineer Operation Rep. Maintenance Rep. Process Engineer SIS Engr. Contractor SIS Specialist System Owner

5 days Safety Requirements

PE, EC, SIS, OR

20 days Detailed Design PE, EC, SIS, OR

Modify

Design Verification

5 days All

OK

Documentation

Training

DCS Database & Graphics

Design Operating Periodic Test Maintenance MOC

Operators Maintenance Tech. Support

Application Software

Field Work

Validation

FAT

60 days All 4 days PE, EC, MR 4 days

Installation

Loop Checks

PE, EC, MR

4 days PE, EC, MR 4 days

SAT

PE, EC, MR 2 days PSSR All

Periodic Test

SIS Support

Modification

Audit

A

Figure 1 - SIS Life Cycle

July 1996

H-29


Appendix H

PHA OSHA requires hazard’s analysis on all facilities using hazardous material. Chevron has chosen to use Hazard and Operability (HAZOP) Studies to satisfy this requirement. The HAZOP report identifies the need for automated safety functions and is the primary input for a Safety Objectives Analysis (SOA). The deliverable for this phase is a HAZOP report.

SOA A Safety Objectives Analysis (SOA) is a formal process to determine if a Safety Interlock System (SIS) is needed. A typical SIS consists of sensors that measure process variables, a logic solver that is configured to recognize predetermined hazardous conditions and initiate safety actions, and final control elements (e.g., valves) which are driven by the logic solver to eliminate the unwanted process condition (e.g., over pressure, loss of flow). A SIS is usually the last layer of protection to prevent hazardous process conditions, damage to equipment, and to provide personnel safety. Applying non-SIS layers of protection (operating procedures, training, process control system, pressure relief devices, etc.) to reduce the level of risk to an acceptable level is desirable before applying a SIS. If an SIS is still required, the analysis continues. The protective functions to be implemented in the SIS and the availability required to meet the level of risk is defined. The level of SIS performance is referred to as the Safety Integrity Level (SIL). The deliverable for this phase is the Draft SOA report.

SOA Verification During this phase the SOA is reviewed by those not involved during the SOA meeting. This may be operators, operating assistances, superintendents, and other managers interested in the functionality of the SIS. Typically, the project engineer will produce a draft report for comments. After all comments are resolved and incorporated into the SOA, the report is complete. The deliverable for this phase is the Final SOA report.

Safety Requirements The objective of the Safety Requirements Specification (SRS) is to document detailed SIS data and decisions and to give sufficient information to the engineering contractor so they can continue with the detailed design. Typically, this package includes, assumptions, unanswered questions, items requiring further analysis, life cycle model, simplified P&ID’s (i.e., P&ID’s showing the SIS instrumentation only), Cause and Effect Matrix, I/O assignments, trip points, time delays, process ranges, pre-trip points, alarms, anunciators, process safe states, safety functions, speed of response requirements, start-up permissives, and the SOA Report. The deliverable for this phase is a Safety Requirements Specification.

H-30

July 1996


Appendix H

Detailed Design During this phase, the SIS design is further refined. Operating displays are produced for operator review. Logic diagrams may be produced for review by the SIS Expert and the Project Engineer. The field station containing controls and indicators is designed. Junction boxes and cable runs are chosen. Chopper valve manifolds are laid out. The pneumatic controls are laid out. This includes the Chopper solenoid valves, control valve pneumatic controls, and pressure switches. Sketches of special process connections for sensors, such as differential pressure sensors sharing an orifice plate are produced. Equipment layouts are sketched, including manual system trips (e.g., Big Red Push Button schematic), cabinets, rack room, sensor locations, and valve locations. The deliverable for this phase is a Draft Detailed Design Specification that includes detailed sketches of the items mentioned above.

Design Verification At this point the project team has completed the design and is ready for review. The detailed design should be checked against the SOA to verify that all functions are implemented, concerns are addressed, and no unintentional functions have been added to the SIS. The deliverable for this phase is the final Detailed Design Specification and drawings issued for construction.

Procedures Writing the procedures requires coordination between the engineering contractor, maintenance personnel, operations, and the project engineer. The SIS engineering contractor can provide the first draft of these procedures. The deliverables for this phase are the four procedures; Operating Procedures, Periodic Test Procedure, Maintenance Procedure, and MOC Procedure.

Training Training those involved in operating and maintaining the SIS is required by OSHA 1910.119. Typically, the engineering contractor works with operations and maintenance personnel to develop a course for each. The deliverables for this phase are two sets of course material and performing the required training. One course focuses on operators while the other is for maintenance. Other technical people (e.g., Process Engineers, Owner of the SIS, Project Engineer, etc.) should sit in on one of the classes.

DCS Database & Graphics After operations approval of the graphics is received, the DCS can be configured. The engineering contractor or the Plant Control System Technologist is capable of performing this phase. The deliverables for this phase are the SIS graphics and the configuration of the DCS points.

July 1996

H-31


Appendix H

Field Work This phase includes all the piping modifications, installation of sensors, wire, junction boxes, pneumatic tubing, and valves. Typically, the normal facility electrical, mechanical, etc., contractors are brought on board to complete this phase. The project engineer with the assistance of the SIS engineering contractor directs the installations. The deliverables for this phase are proper installation of field equipment and the appropriate drawings.

Application Software The Application Software phase includes specifying, purchasing and programming the PLC. The engineering contractor can perform the entire phase. The deliverable for this phase is PLC hardware and the first version of software.

FAT Typically, the SIS engineering contractor produces the procedure and maintenance personnel or operators perform the test with assistance from others. Those others may include the Project Engineer, the Process Engineer, the SIS Expert, and the SIS engineering contractor. The deliverable of this phase is an execution of the FAT without any discrepancies between the test procedure expected outcomes and the test results.

Installation After the FAT, the SIS is shipped to the plant and installed. The project engineer is responsible for proper installation. It is common for the SIS engineering contractor and SIS Expert to assist. The deliverables for this phase are proper installation of all SIS equipment and the all drawings marked up to reflect actual installation.

Loop Checks This effort requires coordination between the site contractor who installed the field equipment, the developer of the DCS graphics and database, and the PLC programmer. All loops are verified for proper installation, calibration, and range in the field, in the DCS database, on the DCS graphics, and in the PLC program. The deliverable for this phase is Loop sign-off sheets. Typically, the electrical contractor and the project engineer sign-off the sheets.

SAT As with the FAT, the SIS engineering contractor produces the procedure and maintenance personnel or operators perform the test with assistance from others. Those others may include the Project engineer, the Process Engineer, the SIS Expert, and the SIS engineering contractor.

H-32

July 1996


Appendix H

The deliverable of this phase is an execution of the SAT without any discrepancies between the test procedure expected outcomes and the test results, and all diagnostic alarms indicating no failures.

PSSR The Pre-Startup Safety Review is the last check that all project deliverables, installations, procedures, and training are adequate and complete. It provides a distinct point in time in which the SIS is handed over to an owner. The deliverable for this phase is a PSSR report signed by the appropriate parties, which assures all drawings, procedures, training, and installations are complete.

Periodic Test Either operations or maintenance will be responsible for conducting periodic testing. The testing continues for the life of the system. After the test all results are recorded. Any discrepancies are fixed immediately. The deliverables for this phase are written results of each periodic test. This phase continues for the life of the SIS.

SIS Support A single person or team should be designated as the owner of the SIS. They would ensure that the SIS is armed, working properly, documentation is updated, periodic tests are occurring and are documented, and be involved in all MOC’s to the SIS.

Modification OSHA 1910.119 requires a Management of Change procedure be written and enforced. Part of the MOC procedure is to update all documentation and train all involved about the change before implementation of the change. The deliverables for this phase are a signed MOC, updated documents, all involved trained, and a modified SIS.

Audit An audit of the SIS and it’s associated documentation should be completed periodically. Once a year is sufficient. A person independent of the facility should review all procedures, documentation, hardware, software, maintenance practices, and operation practices. Other SIS Owners around Chevron or the CRTC SIS Expert can be used as the independent auditor. The deliverables for this phase are a completed and signed off SIS Checklists.

July 1996

H-33


Appendix H

ATTACHMENT B - SIS LIFE CYCLE RESOURCE REQUIREMENTS The table below shows the resource requirements to implement a typical SIS project. The Phase names coincide with the Life Cycle model presented in Figure 1. The Duration column shows the number of hours required to complete the effort assuming all resources are available as needed. Each title under the Resources column represents one person. The % Effort column shows the percent of time the resource will be required for the given Duration.

Phase

Duration (Hours)

Resources

Effort (Person-hours)

20

Project Engineer

20

Operation Rep.

20

Maintenance Rep.

20

Process Engineer

20

SIS Expert

20

SIS Eng. Cont.

20

SIS Owner

20

Project Engineer

8

Operation Rep.

8

Maintenance Rep.

8

Process Engineer

8

SIS Expert

8

SIS Eng. Cont.

8

SIS Owner

8

Project Engineer

20

Operation Rep.

20

SIS Expert

4

SIS Eng. Cont.

40

SIS Owner

4

Project Engineer

40

Operation Rep.

40

SIS Expert

20

SIS Eng. Cont.

320

SIS Owner

20

HAZOP SOA

SOA Verification

Safety Requirements Specification

Detailed Design

H-34

80

40

160

July 1996


Appendix H

Phase

Duration (Hours)

Resources


Design Verification

80

Project Engineer

20

Operation Rep.

20

Maintenance Rep.

8

Process Engineer

4

SIS Expert

8

SIS Eng. Cont.

80

SIS Owner

8

Project Engineer

4

Operation Rep.

80

Process Engineer

4

SIS Eng. Cont.

80

SIS Owner

4

Project Engineer

4

Maintenance Rep.

40

SIS Eng. Cont.

40

SIS Owner

4

Project Engineer

4

Operation Rep.

40

Maintenance Rep.

40

SIS Expert

4

SIS Eng. Cont.

40

SIS Owner

4

Project Engineer

4

Operation Rep.

40

Maintenance Rep.

40

SIS Expert

4

SIS Eng. Cont.

40

SIS Owner

4

Procedures Operating

Maintenance

Periodic Tests

MOC

July 1996

80

40

40

40

H-35


Appendix H

Phase

Duration (Hours)

Resources


Training

40 to develop course (2-4 hour course)

Project Engineer

2

Operation Rep.

20

Process Engineer

2

SIS Expert

2

SIS Eng. Cont.

40

SIS Owner

2

Project Engineer

2

Maintenance Rep.

20

SIS Expert

2

SIS Eng. Cont.

40

SIS Owner

2

Process Engineers

Sit in on above courses.

Operators

Maintenance

Technical Support

40 to develop course (1-2 day course)

As needed

Control System Tech DCS Database & Graphics

Field Work

PLC

PLC Factory Acceptance Test (FAT)

H-36

60

200

500

40

Project Engineer

15

Operation Rep.

30

Process Engineer

4

SIS Expert

8

SIS Eng. Cont.

120

SIS Owner

20

Project Engineer

50

SIS Expert

10

SIS Eng. Cont.

100

SIS Owner

40

Electrical Cont.

800

Mechanical Cont.

800

Project Engineer

100

SIS Expert

10

SIS Eng. Cont.

500

SIS Owner

20

Project Engineer

40

Operation Rep.

40(optional)

Maintenance Rep.

40(optional)

SIS Eng. Cont.

40

SIS Owner

40

July 1996


Appendix H

Phase

Duration (Hours)

Resources


PLC Installation

40

Project Engineer

4

SIS Eng. Cont.

8

SIS Owner

8

Electrical Cont.

80

Mechanical Cont.

80

Project Engineer

4

Operation Rep.

8 (optional)

Maintenance Rep.

8 (optional)

SIS Eng. Cont.

80

SIS Owner

20

Electrical Cont.

80

Project Engineer

40

Operation Rep.

40

Maintenance Rep.

40

SIS Eng. Cont.

40

SIS Owner

40

Project Engineer

20

Operation Rep.

20

Maintenance Rep.

20

SIS Expert

8

SIS Eng. Cont.

20

SIS Owner

20

SIS Owner

1

Operation Rep.

10

Maintenance Rep.

10

SIS Owner

0.1

Operation Rep.

1

Maintenance Rep.

1

SIS Owner

0.1

Operation Rep.

1

Maintenance Rep.

1

SIS Owner

.1

Operation Rep.

15 min.

Loop Checks

Site Acceptance Test (SAT)

PSSR

40

40

20

Periodic Tests Turnaround

Functional

Tight Shutoff

Sensors

July 1996

10 hrs/turnaround

1 hour/quarter

1 hour/quarter

15 minutes/shift

H-37


Appendix H

Phase

Duration (Hours)

Resources


SIS Support

As needed for hardware failures or software errors.

SIS Owner

10

Maintenance Rep.

100

SIS Owner

20

Maintenance Rep.

200

SIS Expert

20 (to make program change)

SIS Owner

10

Operation Rep.

50

Maintenance Rep.

50

SIS Expert

100

Project Engineer

300

Operation Rep.

410 plus 2 hrs / qtr

Maintenance Rep.

624 plus 2 hrs / qtr

Process Engineer

40

SIS Expert

100

SIS Eng. Cont.

1600

SIS Owner

400

(Est. 100 hrs/year) Modification (MOC)

As needed due to new process equipment or software errors. (Est. 200 hrs per life of SIS)

Audit

TOTALS

20 hours/2 years

Control System Tech Training

H-38

Electrical Cont.

960

Mechanical Cont.

880

July 1996

159078182 447 Chevron Instrumentation[1]

Recommend Documents