Advanced Pulverised Coal Power Plants EPRI 2007 208p

CoalFleet Guideline for Advanced Pulverized Coal Power Plants Version 1 1012237

CoalFleet Guideline for Advanced Pulverized Coal Power Plants Version 1

1012237 Technical Update, March 2007

EPRI Project Managers J. Wheeldon D. Dillon

ELECTRIC POWER RESEARCH INSTITUTE 3420 Hillview Avenue, Palo Alto, California 94304-1338 PO Box 10412, Palo Alto, California 94303-0813 USA 800.313.3774 650.855.2121 [email protected] www.epri.com ▪

▪

▪

▪

▪

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT. ORGANIZATION(S) THAT PREPARED THIS DOCUMENT CoalFleet Advanced PC Guideline Working Group (see Citations) Electric Power Research Institute

NOTICE: THIS REPORT CONTAINS PROPRIETARY INFORMATION THAT IS THE INTELLECTUAL PROPERTY OF EPRI, ACCORDINGLY, IT IS AVAILABLE ONLY UNDER LICENSE FROM EPRI AND MAY NOT BE REPRODUCED OR DISCLOSED, WHOLLY OR IN PART, BY ANY LICENSEE TO ANY OTHER PERSON ORORGANIZATION.

This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of continuing research, a meeting, or a topical study. It is not a final EPRI technical report.

NOTE For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or e-mail [email protected]. Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE FUTURE OF ELECTRICITY are registered service marks of the Electric Power Research Institute, Inc. Copyright © 2007 Electric Power Research Institute, Inc. All rights reserved.

CITATIONS This document was prepared by

Alstom Power Glen Jukkola Babcock & Wilcox Co. Kevin McCauley Bechtel Power Corporation Paul Kochis Ram Narula Bob Nicolo Harvey Wen Bevilacqua-Knight, Inc. Rich Myhre Eric Worrell Consultants Janos Beer Carl Bozzuto CPS Energy John Kosub

EPRI Ralph Altman Tony Armor Kent Coleman Chuck Dene Des Dillon Tony Facchiano George Offen Vis Viswanathan John Wheeldon E.ON US Doug Schetzel Exelon Daniel Wusinich Great River Energy Charles Bullinger

MHI David McDeed Midwest Generation (EME) Kent Wanninger TXU Corp. Ronald Hagen U.S. Departmet of Energy Robert Romanosky WorleyParsons Group, Inc. Gary Grubbs Bruce M. Kautsky Don Leininger Paul K. Shewchuk Richard E. Weinstein

Lincoln Electrical System Tom Davlin

This document describes research sponsored by the Electric Power Research Institute (EPRI). This publication is a corporate document that should be cited in the literature in the following manner: CoalFleet Guideline for Advanced Pulverized Coal Power Plants: Version 1 , EPRI, Palo Alto, CA, 2007. 1012237.

iii

ABSTRACT The CoalFleet Guideline for Advanced Pulverized Coal Power Plants provides an overview of state-of-the art and emerging technologies for pulverized coal-fired generating units along with lessons learned for current plants worldwide. The Guideline aims to facilitate the timely deployment of reliable, next-generation generating units that incorporate: •

Higher steam conditions for higher efficiency and reduced generation of pollutants

•

Advanced environmental controls for reduced emissions and environmental impacts

•

Techniques for CO2 capture, or for future retrofit of CO 2 capture, that minimize impacts on efficiency and capacity

This Guideline represents the first step in an ongoing collaborative effort by the CoalFleet Advanced PC Working Group, which includes more than 30 participants from CoalFleet member companies, EPRI staff, and expert consultants. The Guideline reflects information from EPRI, DOE, power producers, equipment suppliers, plant designers, and engineering, procurement, and construction (EPC) companies. Version 1 features a summary of worldwide history with supercritical steam cond itions for pulverized coal power plants. Data are provided on current and planned units with supercritical and more advanced “ultra-supercritical” steam conditions. A review of current design trends addresses unit size, major component types and maximum sizes, furnace design, cycling of supercritical steam generators, fuel properties, use of materials with improved high-temperature strength and corrosion resistance to enable higher efficiency, use of coal drying to improve efficiency, environmental control technologies for SO2 and SO3, and multi-pollutant control technologies. Future versions of the Guideline will update and expand upon these topics to include control of NOX, mercury, and fine particulate emissions, and technologies for carbon dioxide capture and compression. Subsequent versions will also add lessons learned from power industry experience with new advanced pulverized coal power plants and technology development pilot projects. The state-of-the-art and emerging technologies covered in the Guideline provide a viable path to coal-based power generation that meets economic, environmental, and security criteria. The focus is on advancements in supercritical generating units, although the advanced environmental control technologies are applicable to conventional pulverized coal units as well.

v

EPRI Proprietary Licensed Material

CONTENTS 1 INTRODUCTION .................................................................................................................... 1-1 Overview of Guideline Development Approach and Content............................................... 1-1 Guideline Topic Areas.......................................................................................................... 1-3 2 ADVANCED PULVERIZED COAL REFERENCE PLANT GUIDELINES APPROACH ........2-1 Assumed Generation Planning Decisions............................................................................ 2-2 Future Generations of Reference Plants.............................................................................. 2-3 3 STATE OF THE ART FOR ADVANCED PULVERIZED COAL POWER PLANTS ............... 3-1 Supercritical Steam Technology Deployment History .......................................................... 3-1 Drivers for SC and USC Technology Evolution.................................................................... 3-3 Economic Factors .......................................................................................................... 3-3 Environmental Factors ................................................................................................... 3-4 Lessons Learned from 50 Years of Supercritical Technology.............................................. 3-4 World Market Trends for Advanced Pulverized Coal Units: Supercritical and Ultra-Supercritical Plants...................................................................................................... 3-6 World Market for Supercritical Steam Generators.......................................................... 3-6 Planned Units in China................................................................................................... 3-8 Planned Units in Europe ................................................................................................ 3-9 Planned Units in the United States .............................................................................. 3-11 Major Equipment Supplier Experience with Supercritical and Ultra-Supercritical Steam Power Plants ...................................................................................................................... 3-13 4 CURRENT DESIGN TRENDS AND ISSUES ......................................................................... 4-1 Unit Size and Scale.............................................................................................................. 4-1 General Capital Cost Considerations............................................................................. 4-1 Construction and Schedule Considerations ................................................................... 4-6 Cost of Redundancy and Reliability versus Replacement Power .................................. 4-6 Technical Risk................................................................................................................ 4-6 Steam Generator Design Issues and Trends....................................................................... 4-6 Furnace Design.............................................................................................................. 4-6 Designing Supercritical Steam Generators for Low Minimum Load Capability and Continuous Duty Minimum Load Cycling ..................................................................... 4-13 Design Provisions for Higher Peak Power Rating........................................................ 4-14 5 ISSUES RELATED TO FUEL QUALITY................................................................................ Coal Rank ............................................................................................................................ Coal Analysis ....................................................................................................................... Grindability ..................................................................................................................... Ignition and Flame Stability ............................................................................................ Unburned Carbon...........................................................................................................

vii

5-1 5-1 5-2 5-3 5-3 5-4


Emissions....................................................................................................................... Ash Properties and Deposition Behavior ....................................................................... Coal Blending....................................................................................................................... Blend Impact on Coal Grinding ...................................................................................... Blend Impact on Combustion and Deposition ................................................................ Blend Impact on Emissions............................................................................................

5-4 5-5 5-6 5-6 5-7 5-7

6 IMPROVING PLANT EFFICIENCY WITH ADVANCED STEAM CONDITIONS ................... 6-1 Designing for High Steam Pressure: >3750 psi (>260 bar) ................................................. 6-1 Designing for High Steam Temperatures: 1050–1150°F (565–620°C) ................................ 6-2 Steam Generator Components ...................................................................................... 6-4 Superheater and Reheater Design ................................................................................ 6-7 Headers and Piping...................................................................................................... 6-13 7 IMPROVING PLANT EFFICIENCY WITH COAL DRYING.................................................... 7-1 Conventional Coal Drying in Pulverized Coal Units ............................................................. 7-1 Advanced U.S. Coal Drying Technologies ........................................................................... 7-3 Great River Energy Lignite Dryer ................................................................................... 7-3 AMAX Coal Dryer........................................................................................................... 7-7 Rosebud Coal Dryer....................................................................................................... 7-9 Advanced International Coal Drying Technologies ............................................................ 7-11 Mechanical Thermal Expression Drying System .......................................................... 7-11 RWE WTA Fluidized-Bed Dryer ................................................................................... 7-13 8 AIR EMISSIONS CONTROL .................................................................................................. 8-1 Environmental Regulations .................................................................................................. 8-1 Annual Emissions................................................................................................................. 8-3 9 WET FGD SYSTEMS FOR SO2 CONTROL........................................................................... 9-1 Equipment and Process for Limestone-Based Open Spray System.................................... 9-1 Gypsum Processing ....................................................................................................... 9-3 Limestone Preparation System ...................................................................................... 9-6 Alternative Designs .............................................................................................................. 9-7 Lime-Based FGD Systems............................................................................................. 9-7 Other Wet FGD Technologies.............................................................................................. 9-9 Jet Bubbling Reactor...................................................................................................... 9-9 Dual Contact Absorber ................................................................................................. 9-10 Alstom .......................................................................................................................... 9-11 Babcock & Wilcox (B&W)............................................................................................. 9-11 Babcock Power Environmental Inc. (BPEI) .................................................................. 9-12 Ammonia FGD System ...................................................................................................... 9-12 Current Design Issues........................................................................................................ 9-14 10 DRY SO2 CONTROL TECHNOLOGIES............................................................................. 10-1

viii


Lime Spray Drying Absorption ........................................................................................... 10-1 Lime Preparation System............................................................................................. 10-3 Recycle Slurry System ................................................................................................. 10-3 Lime Spray Drying Absorption Process........................................................................ 10-4 Current Issues for Lime Spray Drying .......................................................................... 10-5 Materials of Construction ............................................................................................. 10-7 SDA Vessel Size .......................................................................................................... 10-7 Single versus Multiple Atomizers ................................................................................. 10-7 By-product Disposal ..................................................................................................... 10-7 Other Dry SO2 Control Technologies ................................................................................. 10-8 Dry Sorbent Injection Process...................................................................................... 10-8 Circulating Dry Scrubber (CDS) Process ..................................................................... 10-9 Flash Dryer Absorber (FDA) Process ........................................................................ 10-11 11 SO3 CONTROL TECHNOLOGIES ..................................................................................... 11-1 SO3 and Acid Mist Formation in Coal-Fired Boilers............................................................ 11-2 Sorbent Injection Control Technologies ............................................................................. 11-3 Injection Methods ......................................................................................................... 11-4 Sorbent Properties ....................................................................................................... 11-5 Wet Electrostatic Precipitators (WESP) ............................................................................. 11-9 Horizontal Flow WESP ............................................................................................... 11-10 Tubular WESP ........................................................................................................... 11-11 Materials of Construction ........................................................................................... 11-14 Emerging Technologies for SO3 Control .......................................................................... 11-14 Membrane WESP ...................................................................................................... 11-14 Plasma-Enhanced WESP .......................................................................................... 11-14 Lime Spray Drying for SO3 Removal .......................................................................... 11-15 Power Plant Applications of Sorbents for SO3 Control ..................................................... 11-15 Power Plant Applications of WESP for SO3 Control......................................................... 11-17 12 MULTI-POLLUTANT CONTROL SYSTEMS ..................................................................... Powerspan ECO Process, .................................................................................................. Three-Step Processing of Flue Gas............................................................................. Collection of Liquid Streams ........................................................................................ By-product Recovery.................................................................................................... Performance Data and Other Considerations .............................................................. Other Multi-Pollutant Processes......................................................................................... ReACT Process ........................................................................................................... Airborn Process............................................................................................................ Mobotec ROFA/ROTAMIX Process .............................................................................

12-1 12-1 12-1 12-2 12-3 12-3 12-3 12-3 12-4 12-4

A TERMINOLOGY, ABBREVIATIONS, AND ACRONYMNS.................................................. A-1

ix


LIST OF FIGURES Figure 2-1 EPRI Reference Plant Evolution in the Family of Advanced PC Guidelines ............2-4 Figure 3-1 State of the Art in Worldwide Pulverized Coal Installations...................................... 3-1 Figure 3-2 Steam Conditions and Key Material Selections for State-of-the-Art Pulverized Coal Plants................................................................................................................................. 3-3 Figure 3-3 Supercritical and USC Units Commissioned 1995–2004 with Main Steam at 1050°F (565°C) or Higher....................................................................................................................... 3-6 Figure 3-4 Worldwide Pulverized Coal Units with Main Steam above 1050°F (565°C) Installed from 1995 to 2005...................................................................................................................... 3-8 Figure 3-5 Ultra-Supercritical Steam Generator Units Planned in Europe for Commissioning in 2006–2012 ........................................................................................................................... 3-10 Figure 3-6 Ultra-Supercritical Power Plant Units Announ ced in the United States for Construction Start in 2006–2014 ............................................................................................. 3-12 Figure 4-1 Trend in Cost versus Unit Gross Output Rating ....................................................... 4-1 Figure 4-2 Furnace Circuit Recirculation with Separate Recirculation Pump .......................... 4-11 Figure 4-3 Furnace Circuit Recirculation without Separate Recirculation Pump ..................... 4-12 Figure 6-1 Comparison of Allowable Stresses of Ferritic Steels for Boiler.............................. 6-16 Figure 6-2 Comparison of Allowable Stress for Various Metals............................................... 6-17 Figure 6-3 Vallourec & Mannesmann Hot Neck P91 Fitting .................................................... 6-18 Figure 6-4 P91 Superheater Outlet Headers for Dayton Power and Light, Stuart Station....... 6-20 Figure 6-5 Relative Rupture Strength of High Temperature Steels ......................................... 6-21 Figure 6-6 Comparison of Piping Wall Thickness for Candidate Ferritic Steels....................... 6-21 Figure 6-7 Typical P91 to P22 Weld — Vulnerable to Cracking at Junction between B9 Filler and P22........................................................................................................................... 6-22 Figure 6-8 Joint Geometries of Concern — Highlighted Transition Indicates Weakest Part of Weld ............................................................................................................................. 6-23 Figure 6-9 Correct Weld Profile for P91 to P22 Welds ............................................................ 6-23 Figure 7-1 Coal Drying and Grinding with Pressurized Preheated Air ....................................... 7-2 Figure 7-2 Coal Drying and Grinding with Furnace Gases and Air (Exhauster Mill).................. 7-2 Figure 7-3 Simplified Schematic of Great River Energy Dryer................................................... 7-4 Figure 7-4 Reduction of Moisture at Great River Energy’s Coal Creek Station ......................... 7-5 Figure 7-5 AMAX Coal Dryer Schematic ................................................................................... 7-8 Figure 7-6 Rosebud Coal Dryer Schematic ............................................................................. 7-10 Figure 7-7 Schematic of Mechanical Thermal Expression Coal Drying Process ..................... 7-12 Figure 7-8 WTA Dryer Schematic with Sample Flow Calculations .......................................... 7-14 Figure 9-1 Wet FGD Spray Tower Configuration....................................................................... 9-2 Figure 9-2 Schematic of Typical Wet Flue Gas Desulfurization System—Absorber and Reagent Mixing .......................................................................................................................... 9-4 Figure 9-3 Schematic of Typical Wet Flue Gas Desulfurization System—Gypsum Processing System ....................................................................................................................................... 9-5 Figure 9-4 Typical General Arrangement for Wet Limestone Grinding Systems .......................9-7 Figure 9-5 Schematic of Jet Bubbling Reactor Internals.......................................................... 9-10 Figure 10-1 Typical Dry FGD Process Flow Diagram .............................................................. 10-2 Figure 10-2 Dual Fluid Nozzle Atomizer (Left) and Rotary Atomizer (Right) ........................... 10-3 Figure 10-3 Typical Dry Injection System ................................................................................ 10-9 Figure 10-4 Schematic of Circulating Dry Scrubber System (Lurgi Lentjes North America).. 10-10 Figure 10-5 Alstom FDA Process .......................................................................................... 10-11 Figure 11-1 Visible Results of SO3 Control Using Sorbent Injection ........................................ 11-2 xi


Figure 11-2 Preferred Injections Points for Various Sorbents.................................................. 11-5 Figure 11-3 Side Cut-Away View of Horizontal Flow Wet Electrostatic Precipitator .............. 11-11 Figure 11-4 Example of Tubular WESP Installation above FGD ........................................... 11-13 Figure 11-5 Magnesium-Enhanced Lime SO3 Control Process with Bleed Stream Oxidation and Mg(OH)2 Recovery .......................................................................................................... 11-16 Figure 11-6 Power Plants Using the Thiosorbic® Magnesium-Enhanced Lime FGD Process .................................................................................................................................. 11-17 Figure 11-7 AES Deepwater After WESP Installation............................................................ 11-18 Figure 11-8 Xcel Sherco Station ............................................................................................ 11-20 Figure 11-9 Coleson Cove Shown Prior to the Installation of Wet FGD and Wet ESP Systems ................................................................................................................................. 11-22 Figure 11-10 Wet ESP Arrangement for Coleson Cove Station ............................................ 11-23

xii


LIST OF TABLES Table 3-1 Solutions to Reliability Issues Encountered in Early U.S. SC and USC Plants .........3-5 Table 3-2 Summary of Planned U.S. Supercritical Capacity Additions.................................... 3-13 Table 3-3 Supercritical Steam Generators Supplied by Alstom ............................................... 3-14 Table 3-4 Supercritical Steam Turbines Supplied by Alstom ................................................... 3-18 Table 3-5 Supercritical Steam Turbines Supplied by Ansaldo Energia ................................... 3-19 Table 3-6 Supercritical Steam Generators Supplied by Babcock & Wilcox (B&W)..................3-20 Table 3-7 Supercritical Steam Generators Supplied by Burmeister & Wain Energy (BWE) ....3-26 Table 3-8 Supercritical Steam Generators Supplied by Foster-Wheeler ................................. 3-27 Table 3-9 Supercritical Steam Generators Supplied by Hitachi Power Systems .....................3-28 Table 3-10 Supercritical Steam Turbines Supplied by Hitachi Power Systems .......................3-30 Table 3-11 Supercritical Steam Generators Supplied by Ishikawajima-Harima Heavy Industries (IHI) ......................................................................................................................... 3-31 Table 3-12 Supercritical Steam Generators Supplied by Doosan Babcock (formerly Mitsui Babcock)........................................................................................................................ 3-33 Table 3-13 Supercritical Steam Generators Supplied by Mitsubishi Heavy Industries (MHI) ..3-34 Table 3-14 Supercritical Steam Turbines Supplied by Mitsubishi Heavy Industries (MHI) ......3-36 Table-3-15 Supercritical Steam Turbines Supplied by Siemens-Westinghouse......................3-37 Table 3-16 Supercritical Steam Turbines Supplied by Toshiba ............................................... 3-38 Table 6-1 Temperature Limits for Materials Proven in High-Temperature Applications ..........6-11 Table 6-2 Evolution of Four Generations of Ferritic Steels ...................................................... 6-15 Table 6-3 Composition of Advanced Steels, including Tungsten-Containing P92, P122, and E911.................................................................................................................................. 6-19 Table 6-4 Summary of the Availability and Use of Grade 91 and Other Advanced Ferritic Steels ........................................................................................................................... 6-25 Table 6-5 EPRI Documents Related to Forming and Welding P91 in Fossil Plants ................ 6-31 Table 6-6 Specification Example for Main Steam Piping for Supercritical Steam Conditions ................................................................................................................................ 6-32 Table 6-7 Specification Example for Hot Reheat Piping for Supercritical Steam Conditions ................................................................................................................................ 6-32 Table 6-8 Specification Example for Main Steam Piping for Ultra-Supercritical Steam Conditions ................................................................................................................................ 6-33 Table 6-9 Specification Example for Hot Reheat Piping for Ultra-Supercritical Steam Conditions ................................................................................................................................ 6-33 Table 7-1 Maximum Grinding Mill Exit Temperatures for Different Coal Types ......................... 7-3 Table 7-2 Improved Unit Performance at the Coal Creek Station (With Just One of Seven Pulverizers Receiving Dried Coal) .................................................................................. 7-6 Table 8-1 Emission Limits from the Latest Revision to 40CFR60, Subpart D............................ 8-2 Table 8-2 Worksheet for Expected Annual Air Emissions for a PC Plant .................................. 8-4

xiii


1

INTRODUCTION The CoalFleet for Tomorrow program aims to accelerate the deployment of clean, efficient, advanced coal power systems by addressing technical and economic challenges to reduce risk. This guideline is intended to help CoalFleet members expedite the technology selection, permitting, and design processes for advanced coal plants. Rather than serving as a comprehensive specification, the Guideline aims to identify key areas, technology changes, and lessons learned that should be addressed by engineers developing such specifications, with emphasis placed on technologies and issues unique to advanced PC plants.

Overview of Guideline Development Approach and Content Compilation of Lessons Learned

The intent of the Guideline is to assemble proven approaches and lessons learned while identifying areas of inadequate knowledge requiring further RD&D. Source materials used in preparation of the Guideline includes: •

Experience of the expert team developing the Guideline

•

Input from EPRI CoalFleet program members

•

Information from published EPRI, DOE, and industry studies

Future versions of the Guideline are expected to include non-proprietary information from sitespecific design studies conducted by Early Deployment Project owners and their EngineerProcure-Construct (EPC) companies and technology suppliers. Although this initial version of the Guideline conc entrates on 60 Hz plants using North American coals, the Guideline draws on experience of power generators in Africa, Asia, Australia, and Europe. Content Specific to Advanced PC Plants

The Guideline focuses on issues related to advanced pulverized coal plant technology, which EPRI considers to include the following categories:

1-1


•

technologies for once-through steam generators (“boilers”), steam turbines, and associated balance-of-plant equipment used in pulverized coal, Rankine-cycle generating units utilizing 1 supercritical or ultra-supercritical steam conditions

•

high-temperature materials for SC and USC boilers and steam turbines

•

state-of-the art emission control systems

•

design of pulverized coal plants to accommodate future retrofit of CO 2 capture process equipment, including CO2 steam cleanup, drying, and compression for on-site geologic injection or transfer to pipeline

For the purpose of this Guideline, an “advanced” PC plant is defined by its use of one or more of the above categories of technology. The Guideline generally skips the much wider range of issues and technologies related to building any pulverized coal unit, except where those topics are useful for understanding advanced or state-of-the-art PC technology or may not be familiar to CoalFleet members. Content Responsive to Varying User Needs and Backgrounds

The Guideline recognizes that various CoalFleet members will approach the design of advanced PC plants with a broad range of prior experience and a diverse set of needs and constraints. Therefore, the Guideline content and organization aims to satisfy the needs of a variety of users, including: •

•

•

•

Engineers, managers, generation planners, and financial personnel responsible for initiating and/or monitoring the development of advanced coal power plants. Owner’s engineers experienced in the development of subcritical PC plants who are now charged with guiding specification and selection of key advanced plant parameters EPC/CM contractors and OEMs who oversee the development of advanced supercritical plants. Engineers experienced with development of supercritical fossil plants who need the latest information on best practices and current and developing technologies for ultra-supercritical and advanced low-emissions plants.

Communicating and Advancing the State of the Art

The Guideline provides CoalFleet members with: •

An overview of the best current information relevant to technology selection and plant design decisions, such as: – Materials for higher steam conditions – State-of-the-art environmental controls

1

For the purposes of this Guideline, ultra-supercritical (USC) steam conditions are defined as having final main steam temperatures greater than 1100°F (593°C) and pressures greater than 3625 psia (250 bar). Although supercritical (SC) steam conditions are defined by pressure and temperature above the critical point of water (3200.1 psia (220.6 bara) and 705.1°F (373.9°C)), supercritical steam cycles typically have main steam pressures of about 3500 psia (240 bar) and main steam and reheat temperatures of about 1050°F (565°C). This allows the expanding steam to remain superheated throughout most of the steam turbine.

1-2


– – –

•

•

•

•

Trade-offs between design features for factors such as maximum heat rate, lowest cost of electricity, operating flexibility, etc. Explanation of design features which may be dependent on site conditions Pre-engineered allowances for mid-life changes effecting areas such as: · emissions limits · duty cycles · fuel selection · water quality, availability, and discharge requirements · requirements for CO2 capture

comparative reference of existing and planned technology implementations (i.e., fuel specification, size, performance, and technology selections for specific plants) OEM specification and operating history (e.g., performance, reliability, and availability data) for different technologies and locations Identification of knowledge gaps where better understanding of material behavior or system dynamics is needed Identification of technology gaps where known challenges require better solutions

Guideline Topic Areas Topics to be addressed in this and future versions of the Guideline include the following (italics indicate future material): •

•

•

Defining the state of the art for advanced PC power plants – Reference plant approach – International experience – World market trends – Status and experience of major suppliers – Current design trends – Fuel quality issues Improving plant efficiency through advanced steam conditions – Higher temperatures – Higher pressures – Single versus double reheat Reducing environmental impact – Current and future regulations – SOX reduction – NO X reduction – CO, VOCs reduction – Mercury and other HAP reductions – Water use and liquid wastes – Solid wastes

1-3


– Steady state versus startup and changing loads – Improving plant efficiency via other methods – Reliability availability and maintenance •

Operations

•

Controls and monitoring

•

Construction considerations

•

Project schedules

•

Safety issues

1-4


2

ADVANCED PULVERIZED COAL REFERENCE PLANT GUIDELINES APPROACH The reference plant approach provides an initial configuration using standardized components that can be individually modified to accommodate site-specific design requirements. The reference plant is not based on a specific boiler type or manufacturer (i.e., boiler type could be spiral wall or vertical wall, wall-fired or corner-fired). Other variations developed from the reference plant configuration may include changes to: •

•

•

Unit size Various components experience different types of impacts as a result of reducing unit size to as low as 600 MW or increasing unit size to as much as 1000 MW (or higher). Steam temperatures and pressures Although there are significant changes to furnace and convective pass dimensions, the materials used for high-temperature piping, headers, and tubing are the greatest area of concern when temperatures are increased. The primary materials for current state-of-the-art plants are 9-chrome and 12-chrome ferritic alloys. Austenitic and/or high-nickel alloys may be required for the “second generation” (1200°F, or 650°C) reference plant. High-nickel alloys are almost certain to be the primary high-temperature materials used in a “third generation” (1300°F, or 700°C) reference plant. Fuels and fuel blends The reference plant may be used as a starting point for configurations firing single fuels, multiple fuels, or blended fuels. Significant design variations result from the significant variations of constituents (carbon, volatile matter, ash, moisture, nitrogen, sulfur, chloride, etc.) and properties (heating value, ash fusion temperature, etc.) between and within such fuel types as bituminous coal, subbituminous coal, lignite, petroleum coke, cofired biomass, etc. Necessary modifications address: – Sizing and arrangement of the steam generator (boiler dimensions, materials, surface area distribution, etc.) – Burner design and control (low-NOX burners, degree of staging, etc.) – Waterwall corrosion and mitigation strategies, including limiting staging with greater NOX removal in the selective catalytic reduction (SCR) reactor, coatings, reagent injection, limiting steam temperatures based on fuel constituents – Air quality control equipment selection and design for reliable operation to meet permitted air emissions. Areas addressed include: · SCR or hybrid SCR/selective non-catalytic reduction (SNCR), electrostatic precipitator (ESP) or fabric filter (FF; often called a “baghouse”), wet or dry flue gas desulfurization (FGD), sorbent injection or a wet ESP (WESP) for SO 3 control, and multi-pollutant control · Mercury control as a function of fuel, the suite of air quality control (AQC) equipment, activated carbon injection (ACI) with a baghouse, and fuel additives

2-1


· • •

•

•

• •

•

Impacts on fly ash sales if ACI is used; available carbon removal technologies

Condenser backpressure Pure sliding pressure versus hybrid sliding pressure operation (with corresponding impacts on equipment and effect on ramp rate and unit response) Baseload or cycling operation, which may be implemented with initial operation of the unit or expected to occur at an undefined future time Normal or fast startup (with corresponding impacts on specifications for turbine bypass, inclusion of an auxiliary boiler, fatigue-resistant design, chemical treatment, etc.) Type and size of turbine bypass system Plant cooling method (cooling pond, wet mechanical draft cooling tower, wet/dry mechanical draft cooling tower, etc.) Access to plant by barge or rail; impact on the level of modularization that can be achieved

Assumed Generation Planning Decisions The reference plant approach used in the Guideline assumes that a power producer has already established the need for new generation and selected a location for the plant. In making these determinations, the power producer would have already considered the following items, which are not explicitly addressed in the Guideline: •

Existing generation capability

•

Load growth projections

•

Size of new unit(s)

•

Loading profiles for existing units, new unit(s), and future units in the generating system

•

Proposed plant site(s)/location(s)

•

Available space on-site or off-site for landfilling by-products that cannot be sold

•

Proximity to rail or barge service

•

Proximity to and availability of water

•

Quality of water

•

Ability to discharge treated wastewater (i.e., Zero Liquid Discharge not required)

•

Proximity to gas for startup versus on-site oil storage

•

Proximity to existing transmission lines

•

Capability of existing transmission system

•

Schedule established based upon when generation is needed and the time to permit, design, procure, deliver, construct, startup, and commission the n ew unit(s)

In addition, it is assumed that the power producer has already established key design criteria, including: •

The new unit(s) will be pulverized coal-fired with supercritical (or USC) steam conditions. [Note: Although recommendations in the Guideline may apply more broadly, it is also generally assumed that the unit will employ single reheat (although double reheat should be 2-2


considered where the highest efficiencies are sought), sliding-pressure operation, and a heater above the reheat point (HARP) cycle with eight feedwater heaters (four low-pressure, a deaerator, and three high-pressure).] •

•

•

•

Solid fuel specifications (coal and petroleum coke design range, design blend ratios, and design basis for guarantees) Minimum quality requirements for by-products (identified in the Guideline as a range of properties). Typically, fly ash, bottom ash, and gypsum will be sold to the greatest extent practical and therefore must meet minimum requirements. If a dry scrubber is used, fly ash may not be saleable. Other specifications will typically include: – Fly ash: maximum carbon content – Bottom ash: maximum carbon content – Gypsum: maximum chlorides and maximum moisture content SCR reagent type and specifications. Considerations for choosing anhydrous ammonia, aqueous ammonia, solid urea, or liquid urea include purchase and transportation cost, safety, availability, and O&M requirements. FGD reagent type and specifications. Considerations for choosing lime or limestone include purchase and transportation cost, availability, and O&M requirements.

•

Air emissions targets

•

Liquid discharge targets

•

Level of accommodation to be made for CO2 capture. This may include: – Space allocation for future equipment, considering likely technology and reagent choice – Economic evaluation and design accommodation for impacts to the low-pressure (LP) turbine(s). This may assume steam is extracted from the intermediate pressure (IP)-to-LP crossover or that provisions are made for other (typically LP ) steam extraction points – Consideration of auxiliary power requirements. This may include modifications in the design of the auxiliary power system to include spare capacity in the initial design or to facilitate its addition at a later date. – Other minimal pre-investment options that could potentially avoid substantial future rework – Review of available technologies and an assessment of technology trends so that future facility needs can be estimated

Future Generations of Reference Plants Ultimately, EPRI envisions a reference plant approach that uses progressively more advanced design criteria. The three “generations” shown in Figure 2-1 incorporate sequentially advancing cycle conditions, improved environmental controls, and increasing considerations for CO2 capture and compression. Version 1 of the Guideline is built around the state-of-the-art reference plant design elements that are now commercially available. The nominal 1200°F (650°C) and 1300°F (700°C) design criteria represent expected future steps for plant designs that leverage newer technologies as they become ready for commercial application.

2-3


VERSION 1

FUTURE VERSIONS

1150F FINAL STEAM TEMPERATURES 3750 PSIG FINAL STEAM PRESSURE

1200F FINAL TEMPERATURES 4 000 PS IG FIN AL S TE AM PRE SSU RE

LARGER TO 1000 MW NET

860 MW GROSS 800 MW NET 1050-1150F

1300F FINAL TEMPERATURES 4500 PS IG FI NA L STE AM PR ES SURE


(SAME STEAM FLOW)

SMALLER TO 600 MW NET FERRITIC

ANTHRACITE BITUMINOUS SUBBITUMINOUS (BASE?) LIGNITE PET COKE BLEND

870 MW GROSS 810 MW NET 1150-1200F


INCREASED STEAM FLOW TO OFFSET CO2

1000 MW GROSS 800 MW NET 1200-1300F

CAPTURE LOAD

SMALLER TO 600 MW NET

SMALLER TO 600 MW NET

AUSTENITICS, HIGH NICKEL

HIGH NICKEL

ANTHRACITE BITUMINOUS SUBBITUMINOUS (BASE?) LIGNITE PET COKE BLEND

STATE OF THE ART AIR EMISSIONS

IMPROVED AIR EMISSIONS CONTROLS

AIR QUALITY CONTROL EQUIPMENT BASED ON FUEL SELECTION


CONSIDERATION GIVEN TO SPACE ALLOCATION AND STRATEGIC PREINVESTMENT IN FACILITIES FOR FUTURE CO 2 CAPTURE

CONSIDERATION GIVEN TO SPACE ALLOCATION AND STRATEGIC PREINVESTMENT IN FACILITIES FOR FUTURE CO 2 CAPTURE

ANTHRACITE BITUMINOUS SUBBITUMINOUS (BASE?) LIGNITE PET COKE BLEND NEAR ZERO EMISSIONS


Figure 2-1 EPRI Reference Plant Evolution in the Family of Advanced PC Guidelines

2-4

CO 2 CAPTURE DESIGNED IN


3

STATE OF THE ART FOR ADVANCED PULVERIZED COAL POWER PLANTS Supercritical Steam Technology Deployment History Supercritical technology was pioneered in the United States in the late 1950s. American Electric Power put the Philo supercritical unit in service in 1957 (retired 1979) and Philadelphia Electric Power followed in 1960 with Eddystone Unit 1, a double reheat, USC unit, which is still in operation, albeit with a slight derate from original specifications. To this da y, Eddystone 1 remains the unit with the highest operating steam conditions in the world, with main steam at 5000 (345 bar) and 1135°F (613°C). The two reheats are at 1050°F (565°C). Many supercritical units were built in the United States in the 1960s and 1970s. Most of these units employed single reheat with main steam conditions of about 3500 psi and 1000°F and with the reheat also at 1000°F (240 bar/538/538°C). For a time, supercritical technology fell out of favor for new plants as a result of technical problems, including materials degradation and the need for overly complex operating procedures. Many U.S. power producers selected subcritical drum-type boilers thereafter, believing that supercritical technology had limited operating capability, complex maintenance issues, lower availability, and lower-than-expected plant efficiency.

Figure 3-1 State of the Art in Worldwide Pulverized Coal Installations

3-1


The problems experienced at the early U.S. plants have largely been remedied and these units are now achieving good performance with availabilities and operating costs similar to those of subcritical plants. Nonetheless, leadership in supercritical plant development moved overseas, with power producers in Denmark building units with steam temperatures exceeding 1050°F (565°C) in the 1990s. This trend was followed by Japanese power producers, who built a large number of units that would be classified as ultra-supercritical by EPRI’s definition (with temperatures reaching 1110°F or 600°C). Today, Germany, Italy, and China all have projects under way that will increase substantially the world’s installed base of generating units with ultra-supercritical steam conditions. Figure 3-1 illustrates USC PC technology trends by plotting maximum steam temperature versus year of initial commercial operation. The plot of recently announced plants for the United States shows the lag of this market behind others. An upward turn in recent years shows a growing trend toward adopting higher steam conditions with U.S. coals. Figure 3-2 shows the steam conditions and materials used for a selection of leading USC plants.

3-2


LANSHAN (Pressu re 4420 psi , 305 bar)

REHEAT 1 TEMP, 1112°F, 600°C

920

940

960

980

JAPAN 2002


940

960

980


940

960

980


940

960

980

940

960

980

ITALY 2006

P91 TP347FG ----Toshib a 12Cr

High Temperature Materials

Headers: SH tubes: RH Tubes: ST Rotors :

MAIN STEAM TEMP, 1112°F, 600°C REHEAT 1 TEMP, 1130°F, 610°C

920

High Temperature Materials:

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200

TORREVALDAL IGA (Pressu re 3625 psi , 250 bar)

900

JAPAN 1998

Headers: SH tubes: RH Tubes: ST Rotors:

MAIN STEAM TEMP, 1100°F, 593°C

920


1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200

NANAO-OHTA (Pressur e 3698 psi , 255 bar)

900

JAPAN 2000

Headers: P122 SH tubes: Super 304H-HR3C RH Tubes: ----ST Rotors: Toshib a 12Cr


920

P122 Super 304H T122 COST 501E

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200

TSURUGA (Pressu re 3698 psi , 255 bar)

900


Headers: SH tubes: RH Tubes: ST Roto rs:


920

-----------------

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200

ISOGO (Pressur e 3857 psi, 266 bar)

900


Headers: SH tubes: RH Tubes: ST Rotors :


900

CHINA 2009

P92 Super 304H Super 304H -----

1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200

Figure 3-2 2 Steam Conditions and Key Material Selections for State-of-the-Art Pulverized Coal Plants

Drivers for SC and USC Technology Evolution Economic Factors The economic benefits offered by today’s supercritical technology (and, by extension, ultrasupercritical) include the following:

2

CoalFleet Database of Advanced Pulverized Coal Plants and Development Projects

3-3


•

Reduced coal consumption, and therefore lower fuel costs per unit of electricity generated

•

Better part-load efficiency and operating flexibility

•

Excellent availability—comparable to that of existing subcritical plants

•

•

Reduced use of consumables such as ammonia for SCR and limestone based sorbents for SO2 capture Reduced CO2 production which may reduce potential future costs for: – retrofit for post-combustion CO 2 capture should the plant need to be retrofitted – purchase of CO2 offsets – taxes based on CO2 emissions

Environmental Factors The environmental benefits offered by supercritical technology include reductions of the following per unit of electricity generated: •

Emissions of NOX, SO2, particulates, and mercury

•

CO2 production

•

Impacts of coal mining, transportation, and handling coal

•

Ash production and disposal

•

Water consumption for condenser cooling

Lessons Learned from 50 Years of Supercritical Technology In hindsight, the operation and maintenance problems experienced by older U.S. plants have been primarily attributed to three major design issues: 1. Constant pressure operation Early supercritical units used constant-pressure operation and required the boiler to remain at constant pressure throughout startup and the entire load range. Constantpressure operation requires a complicated system startup, with longer startup times and higher minimum load than for sliding pressure units. The startup valves must endure large pressure differences during bypass operation, resulting in faster erosion and frequent valve maintenance. More recent supercritical units use sliding-pressure operation to mitigate these types of issues. 2. Slagging problems attributable to inadequate furnace size Furnaces of the early units in the 1960s were relatively small in size compared with those of newer units. A trend toward increased furnace size was a direct result of slagging problems, experienced with U.S. coals, which led to low availability and reliability. 3. Inappropriate water treatment chemistry Once-through boilers and supercritical steam generators are more susceptible to internal scaling of tube walls than are natural-circulation boilers, which use liquid blowdown from the steam drum and mud drum to limit concentrations of dissolved and suspended solids. If internal scale prevents cooling of the tube wall, increased metal temperature can lead to failure of waterwall and superheater tubes. In extreme cases, thick scale can

3-4


increase pressure drop and reduce flow, further reducing cooling of the tube walls. Oncethrough units must use very pure feedwater and a carefully balanced addition of water treatment chemicals to prevent corrosion and subsequent re-deposition of dissolved solids (scaling) on the interior of tube walls. Table 3-1 summarizes some of the design improvements developed and implemented to overcome the problems found in early units. Research and development worldwide has led to improved reliability, fuel flexibility, and wider load range operation. Building on these successes, supercritical technology is much more attractive to U.S. power producers than it was 20 years ago. Table 3-1 3 Solutions to Reliability Issues Encountered in Early U.S. SC and USC Plants Problem

Cause

Countermeasures

Erosion of startup valves

High differential pressure due to constant pressure operation and complicated startup systems

Sliding-pressure operation, simplified startup systems, and low-load recirculation systems

Long startup times

Complicated startup systems and operations (ramping operation required; difficulty matching steam and metal temperatures, etc.)

Sliding-pressure operation; simplified startup systems; low load recirculation systems

Low ramp rates

Rapid temperature change during constant pressure operation causes high thermal stresses in the HP turbine

Sliding-pressure operation

High minimum stable operating load

Bypass operation and pressure ramp-up operation required

Sliding-pressure operation; low-load recirculation systems

Slagging

Undersized furnace and inadequate coverage by sootblower system

Design of adequate plane area heat release rate and furnace height without division walls. Provisions of adequate system of sootblowing and water blowers

Circumferential cracking of water wall tubes

Metal temperature rise due to inner scale deposit and fireside wastage

Oxygenated water treatment (OWT). Protective surface in combustion zone of furnace for high-sulfur coal (e.g., thermal spray or weld overlay).

Frequent acid cleaning required

Inappropriate water chemistry

Application of OWT

Lower efficiency than expected

High air in leakage due to pressurized furnace. RH spray injection required due to complications of RH steam temperature control in a double reheat cycle configuration.

Tight seal construction. Single reheat system with high steam temperature control by parallel damper gas biasing.

Low availability

All the above

All the above

3

“US Revisits Supercritical Systems: CBEC 4 Leads the way,” Modern Power Systems, 8th April 2004.

3-5


World Market Trends for Advanced Pulverized Coal Units: Supercritical and Ultra-Supercritical Plants World Market for Supercritical Steam Generators In total, there are around 600 supercritical and ultra-supercritical generating units operating worldwide, with the vast majority classified as supercritical, not ultra-supercritical. The 4 combined capacity of these units totals more than 300 GW. Not all SC and USC units are coal-fired. For example, of 170 such units in the United States, 115 are coal-fired. Thirty-five of the 100 units in Japan are coal-fired. The International Energy Agency’s “Coalpower 5” database, updated in 2006, lists nearly all of the 60 SC and USC units in western Europe as coal-fired, whereas many of the 240 units in the former Soviet Union and eastern European countries are oil-fired. In Asia, China ha s about 21 coal-fired units in 5 operation; South Korea has 22. Figure 3-3 shows the total number and capacity of supercritical and ultra-supercritical power plants commissioned between 1995 and 2004. During this 10-year period, Japan and Korea dominated the new plant market, while China began to show signs of rapid growth. In the United States, the last supercritical unit built was in 1989 (Rockport). MidAmerican’s Council Bluffs Unit 4, planned for startup in 2007, will break an almost 20-year hiatus.

Figure 3-3 Supercritical and USC Units Commissioned 1995–2004 with Main Steam at 1050°F (565°C) or 6 Higher

4

M.R. Susta, IMTE, “Supercritical and Ultra Supercritical Power Plants - SEA vision or Reality,” Powergen Asia 2004. 5 IEA, “Coalpower 5” Database, 2006. 6 Source: IMTE AG.

3-6


Figure 3-4 compares the main steam temperatures of pulverized coal units commissioned from 1995 to 2005. The final main steam pressures of these units ranges from a high of 4305 psi (297 bar) for Avedore Unit 2 to a low of 3494 psi (241 bar) for Matsuura Unit 2. The present day market for supercritical boilers is dominated by the rapidly expanding market in 7 China, which accounts for about 90% percent of all supercritical orders placed worldwide. With some 46 supercritical steam generators ordered each year, the total annual addition equates to roughly 28 GW of electrical generation. Due to the significant demand for new generation in China, many international steam generator manufacturers have established agreements with Chinese boiler fabricators allowing manufacturing and technology transfer. Notable observations about this time frame include: •

•

•

•

•

7

Japan deployed 11 pulverized coal units with supercritical or USC steam conditions. Other countries deploying SC/USC units included Denmark, Italy, and China, with 3 units each; South Korea with 2 units; and Germany, Canada, and Australia, with 1 unit each. Japan leads the way in deploying high-temperature steam conditions. However, final steam pressures for these units (3494–3857 psi, or 241–266 bar) are not the world’s highest. Since the mid 1990s, Japan has been continually building the world’s largest capacity supercritical units for firing market-traded coals with less than 1% sulfur content. Eight units of about 1000 MW (net) capacity each are currently operating within Japan. Moderate-size units are not obsolete, as several (6 x 700 MW) units were also commissioned within this time frame. Danish power companies now operate nine supercritical units, three of which have steam temperatures above 1050°F (565°C). These plants, which were built in the late 1990s, each produce about 400 MW of electricity along with ~450 MW t for district heating systems. The three Danish plants with final steam temperatures abov e 1050°F (565°C) feature higher final main steam pressure (4305 psi, or 297 bar) than do Japanese designs.

A.J. Minchener, “Market Perspectives of Clean Coal in Asia,” IEA Clean Coal Centre.

3-7


MW (net) 1100

1050

n i e m s u s s i u M a r e d e i N

1000

950

n a n i k e H

) 2 ( n a w a n a b i h ) ) 1 ( c 2 ) ( 2 i a a ( k T h a a r c n u a i u m h c s a t a a r a t i M H H

Millmerran Genesee (3) Wangqu isogo Changshu

900

850

800

750

1 a g i l a d l a v e r r o T

700

650

u q g n a W

600

u h s o o g n g n a o r o h i s C H i

2 a g i l a d l a v e r r o T

3 a a g g u r i l a u s d T l a v e r r o T

a t h o o a n a N

) 3 ( k e a b r e a k ) S 4 ( a m u s t A h o t a m o T

) 2 ( e r o d e v A

) 3 ( d n a l l y j d r o N

1 o d g n u h g n o Y

2 o d g n u h g n o Y

Hirono Wangqu Torrevaldaliga 1 Torrevaldaliga 2 Torrevaldaliga 3 Tsuruga Nanaoohta Tomatoh-Atsuma (4) Skaerbaek (3)

u q g n a W

Aved ore (2) Nordjylland (3) Yonghungdo 1

550

500

450

Yonghungdo 2

) 3 ( e e s n e a n e r r G e m l l i M

Niederaussen Misumi Hekinan Matsuura (2)

400

Haramachi (2) Hitachinaka (1)

350

Tachibana-wan (2)

300

Figure 3-4 Worldwide Pulverized Coal Units with Main Steam above 1050°F (565°C) Installed from 1995 to 2005

Planned Units in China China’s first supercritical units (2 x 600 MW net) were built in the early 1990s at Shanghai Shidoukou No. 2 power plant. Since that time, over 20 imported supercritical units with a total capacity of 6000 MW have been commissioned.

3-8


In 2003, 26 GW of supercritical boilers were ordered by China (~43 x 600 MW). With similar numbers of orders placed between 2004 and 2006, this vast requirement for new generating 8 capacity is expected to continue well into the future. Estimates suggest that ~22 GW of new supercritical capacity will be installed annually in China for the next 10 years. Units ordered to date are generally 600 MW in capacity and employ well established steam 9 conditions (3510 psi/1050/1050°F, or 242 bar/565/565°C). However, there has recently been a noticeable leap to very large plant capacities, with supercritical plants in the 900–1000 MW range. This is exemplified by the 2 x 900 MW Waigaoqiao plant commissioned in 2002. Several Chinese demonstration projects are adopting ultra-supercritical steam conditions. The Huaneng Yuhuan plant, (4 x 1000 MW units) is planned to commence operation by 2009, with steam conditions of 3625 psi/1112/1112°F (250 bar/600/600°C). The Lanshan plant, which is also due for commissioning in 2009, is set to become one of the world’s foremost USC plants with steam conditions of 4420 psi/1112/1112°F (305 bar/600/600°C) while using smaller unit sizes (4 x 660 MW net). Lanshan represents a significant milestone in China’s energy development, placing China as the leading nation in deployment of ultra-supercritical technology, surpassing Japan, Italy, and Germany before the end of the decade.

Planned Units in Europe Planned capacity additions in Europe during the next five years include a significant number of SC and USC plants, although not to the same extent as is planned for China. Figure 3-5 shows the names and capacities of planned European plants.

8

A. Minchener, “Market perspectives for clean coal in Asia,” IEA Presentation. Z. Zongrang, TPRI, “Development and Application of Supercritical Coal-Fired Units and CFB Boilers In China,” 26 Jan 2005. 9

3-9


MWa (net) 1200

e k a l v s a a M

1150 1100

F h t 4 a r n u l e e t N t a D

G h t a r u e N Maritza 1

1050

Maritza 2

1000 950 900 850 800 750 700

R g r e b x o B

650

y a r e d e c a v i r o v e n d c n e o a L P H

1 e l l o T o t r o P



5 e n r e H

0 1 m u s l a W g r u b s i u D

1 n e l a f t s e u r W b l e t t i M

2 n e l a f t s e W

8 K D R e h u r 2 l s w r a 1 2 t o K g a r g r h u u c b b l e m m B a a H H

Boxberg R Hannover Ledvice Poceraday Porto Tolle 1 Porto Tolle 2 Porto Tolle 3 Herne 5 Duisburg Walsum10 Mittelbrun Westfalen 1

600

Westfalen 2 Hamburg 1

550

Hamburg 2 500

Belchatow2

450

Karlsruhe RDK 8 Maasvlake

400

350 300

1 a z t i r a M

Datteln 4

2 a z t i r a M

Neurath F Neurath G

250 200

Figure 3-5 Ultra-Supercritical Steam Generator Units Planned in Europe for Commissioning in 2006–2012

The following should be noted: •

•

The vast majority of the planned European units adopt ultra-supercritical steam conditions (i.e., steam temperatures above 1100°F, or 593°C, and main steam pressure above 3625 psi, or 250 bar). The chief exception is Polish lignite plants, where steam temperatures are about 1030°F (554°C). The steam conditions for plants planned for the Czech Republic and Bulgaria had not been announced as of early 2007. Germany is set to invest some 60 billion euro in new power stations and electric transmission networks. The trend within the German market appears to be toward larger output units of 700–1000 MW net, primarily firing lignite but with some use of bituminous coal. Three lignite units larger than 1000 MW are due to be completed by 2010. The two units planned for completion at Neurath in 2008 are to have the largest steam generators and highest steam

3-10


temperatures realized to date for lignite. Eight units of ~700 MW are due to be commissioned by 2012. A further two units of ~600 MW have been announced for the same time frame. •

Italy has three units planned at about 660 MW net.

•

The Netherlands has one 1100 MW unit due for completion by 2012.

•

Several eastern European countries have lignite units in development. These include Poland (one 833 MW unit due for completion by 2012), Bulgaria (two 330 MW units due for completion by 2010), and the Czech Republic (two 660 MW units due for completion by 2010).

Planned Units in the United States As of early 2007, 49 supercritical plants had been announced for construction in the United States beginning in 2006–14. As noted, MidAmerican’s Council Bluffs Unit 4 will be the first supercritical unit commissioned in the United States in almost 20 years. TXU has announced planned additions of supercritical units in Texas. FPL has announced plans to build two 980 MW (net) units in Glades County, Florida, by 2012 and 2013, respectively. Plans for these units feature ultra-supercritical steam conditions of about 1118°F (603°C) and 3800 psi (262 bar). Longview Power (a subsidiary of GenPower) has announced a 600 MW ultra-supercritical plant intended to be operational in West Virginia by 2011. These announcements signify a renewed confidence in supercritical technology within the U.S. marketplace and will place the United States among the front runners of ultra-supercritical technology deployment by the early 2010s. As shown in Figure 3-6 about two-thirds of these announced plants are expected to have net outputs greater than 700 MW, with ratings ranging from 700 to 950 MW. The remaining plants will have capacity ratings in the 400 to 700 MW net output range.

3-11


1050

1000

950

900

850

800

750

700

650

600

550

500

450

400

2 ) 1 k a a c d d i o r i r r l o l o d F F e r ( L L r P P e F F ) ) n 4 4 o & & o ) S 3 3 n ( ( 7 i ( a e e k ) ) 4 ) 3 ) t s s e ( ( 4 n 3 u u e e ( ( k u o o r l o k e n ) l a e o H H C m r w r ( n e g g 4 c L C o e r t n i n y a t e i n B 2 1 1 2 2 i t e g n e I d d l n r r s l k 1 2 s e e n r a r a a o o a a i g u s v v a s s u p t t T T V M y f M M L B o u u p r o r I a m n f u 1 2 e p p m l n G G a a u e e t o k k m m C o B d l d a a C i i s a a i ) C C d d s s c e O C f r n f ) 3 O e f f ( e e e i i r r o 3 m l l i u ( e t t b b l y C L C o a t a h h a e h t l g C ) S S g o c n u u T 2 n ( a e e i o 1 2 r i r r o r i y m m t o o a h h o e r a r h h n 1 2 1 2 c c u C P P T T a S a l n l o y y n r o t o r g g t C e e 2 r r g g e n n G e 2 e n l n n i i G u b j o E E e m m r m g 1 2 n a i y l y r u l r a a C n b b o T H E E F F g i m m b 1 2 ) o r B 2 o 2 c o k 1 c l ( ) d d l 2 3 n k e e o N o a a e ( s w e e r r o o H H g e k t o e a n a n C C R o e O a k k R S t m m a a l l d S D g n r O n g e i E E O a i e w o T M S B P ) 4 ( n o t s e W

) 4 ( e l l i v r e g n i r p S

350

300

Figure 3-6 Ultra-Supercritical Power Plant Units Announced in the United States for Construction Start in 2006–2014

Table 3-2 breaks down the tally of planned units into several size ranges. This summary suggests a clear trend in planned pulverized coal unit capacity toward orders for larger units of 800 MW net output and above, namely:

3-12


•

•

• •

About 2 in every 3 units currently planned in the United States are around 800 MW in net output capacity. About 1 in every 3 units currently planned in the United States is in the 600 MW net output capacity range. The average net output of the 49 units planned is 740 MW. Only two of the 49 planned supercritical pulverized coal units are smaller than 600 MW net output.

None of the planned units are smaller than 400 MW net output. Table 3-2 Summary of Planned U.S. Supercritical Capacity Additions Size Range MW (net)

<500 MW

500–700 MW

700–900 MW

>900 MW

Number of Units Planned

1

16

26

3

(2%)

(32%)

(60%)

(6%)

(Percent of Total)

The majority of the planned units (~70%) will be fired on subbituminous coal. About on e-fourth will utilize bituminous coal, and only 4 units planned for the United States will be lignite-fired. Although steam conditions have not yet been announced for many of these proposed plants, developments outside of the United States provide indication of the pressures and temperatures likely to be adopted by some of these units.

Major Equipment Supplier Experience with Supercritical and Ultra-Supercritical Steam Power Plants The following tables list basic parameters for supercritical steam generators and steam turbines supplied by various major manufacturers. This partial listing provides insight into the breadth of the experience base and trends in unit size and steam conditions. [Editor’s Note: As of early 2007, EPRI had not yet completed data collection activities. Thus, not all suppliers are represented in the following tables, and the data for listed suppliers may not be complete. Parameters shown are as-reported by suppliers or in published literature. EPRI has not conducted independent data verification activities. For various reasons, owners may modify operating steam pressures and/or temperatures and plant output during the life of a unit; such modifications may not be reflected below. It should also be noted that given the pace of mergers and acquisitions in the power industry, plant owner names may not be current. Future versions of the Guideline will strive to provide more comprehensive supplier experience compilations.]

3-13


Table 3-3 Supercritical Steam Generators Supplied by Alstom Unit Net Output

Main Steam Pressure

Main Steam Temperature

Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Exelon Power (original ratings)

Eddystone 1

354

5000

345

1200

649

1050

566

Bit.

1959

1050

566

1050

566

Bit.

1959

Lignite

2004 2006? 2003 2003 2000 1996 1995 1992

Reheat #2

Exelon Power (current ratings)

Eddystone 1

325

4500

310

1150

620

Reheat #2

1050

566

Elektrim Megadex Hua Yang EPC Hua Yang EPC Hua Yang EPC Vattenfall VEAG VEAG Vestkraft

Patnow Houshi 7? Houshi 6 Houshi 5 Lippendorf S Shwarze pumpe IV 2? Shwarze pumpe IV 1 Vestkraft PS 3

450 600 600 600 933 800 800 400

4206 4103 4103 4103 4061 4032 4032 4002

290 283 283 283 280 278 278 276

1011 1006 1006 1006 1031 1004 1004 1040

544 541 541 541 555 540 540 560

1054 1050 1050 1050 1031 1022 1022 104

568 566 566 566 555 550 550 560

Shanghai Municipal Elect Company

Waigaoqiao 3

1000

3916

270

1112

600

1112

600

RWE (Germany) Public Power Corp Shanghai EPB Shanghai EPB Intergen Korea EPC Korea EPC

Niederaussem Florina 1 Waigaoqiao II 1 Waigaoqiao II 2 Millmerran Yunghung 1 Yunghung 2

1000 330 900 900 400 800 800

3844 3800 3626 3626 3597 3568 3568

265 262 250 250 248 246 246

1076 1009 1000 1000 1053 1051 1051

580 543 538 538 567 566 566

1112 1008 1040 1040 1105 1051 1051

600 542 560 560 596 566 566

Lignite Lignite — —

Xcel Energy (Colorado USA)

Comanche 3

750

3565

244

1050

566

1100

593

PRB

Due 2009

Exelon Power*

Eddystone 2

325 (354)

3500

241

1050

565

1050

565

Bit.

1960

3-14

Bit. Bit. Bit.

Lignite Lignite —

2009

2002 2001 2004 2004 2002 2002 2002


Unit Net Output

Main Steam Pressure


Owner

Unit

MW

psi

bar

°F

°C

Formosa Plastics Formosa Plastics Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Grosskraftwerk Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC Korea EPC HIPDC HIPDC GKW Mannheim Texas Utilities Public Services Okla

Reheat #2 Mai-Liao FP-1 5 Mai-Liao FP-1 4 Hadong 5 Hadong 6 Tangjin 3 Tangjin 4 Tangjin 2 Franken II 3 Tangjin 1 Hadong 4 Hadong 3 Hadong 2 Hadong 1 Shamchonpo 5 Shamchonpo 6 Taean 3 Taean 4 Taean 1 Taean 2 Poryong 5 Poryong 6 Poryong 3 Poryong 4 Shiongkou II 1 Shiongkou II 2 Mannheim 18 Sandow 4 Northeastern 4

600 600 500 500 500 500 500 600 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 600 600 480 591 473

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

3-15

Reheat Temperature °F

°C

1050

565

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — —

Fuel

— — — — — — — — — — — — — — — — — — — — — — — — — — Lignite

Start Year

2000 1999 1999 1999 1999 1999 1999 1998 1998 1998 1998 1997 1997 1997 1997 1996 1996 1995 1995 1994 1994 1993 1993 1991 1991 1982 1981 1980


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Public Services Okla Texas Utilities Georgia Power Co Texas Utilities Texas Utilities Georgia Power Co Salt River Project Georgia Power Co Salt River Project Texas Utilities Texas Utilities Alabama Power Co Georgia Power Co Salt River Project Colombus Southern Pennsylvania Power South Carolina Gen Alabama Power Co Georgia Power Co Texas Utilities Alabama Power Co Edison International Georgia Power Co Penn Elect Co Pennsylvania Power Potomac Electric Texas Utilities Duke Power Edison International

Northeastern 3 Martin Lake 3 Wansley 2 Martin Lake 2 Martin Lake 1 Wansley 1 Navajo 3 Bowen 4 Navajo 2 Monticello 2 Monticello 1 Gaston 5 Bowen 3 Navajo 1 Conesville 4 Montour 2 A.M. Williams 1 Gorgas 10 Bowen 2 Big Brown 2 Barry 5 Mohave 2 Bowen 1 Conemaugh 2 Montour 1 Morgantown 2 Big Brown 1 Marshall 4 Mohave 1

473 793 952 793 793 952 803 952 803 593 593 952 952 803 842 819 533 789 789 593 789 818 806 936 806 626 593 648 818

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

— — — — — — — — — — — — — — — — — — — — — — — — — — — — —

3-16

Fuel

Lignite Lignite Lignite

Lignite Lignite

Lignite

Lignite

Start Year

1979 1979 1978 1978 1977 1976 1976 1975 1975 1975 1975 1974 1974 1974 1973 1973 1973 1972 1972 1972 1971 1971 1971 1971 1971 1971 1971 1970 1970


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Penn Elect Co Potomac Electric Dairyland Power Duke Power Penn Elect Co Pennsylvania Power Kansai Electric Monongahela Power Penn Elect Co Powergen Tenn Valley Auth Durr-Werke AG The Illuminating Co

Conemaugh 1 Morgantown 1 Genoa 3 Marshall 3 Keystone 2 Brunner Island Himeji II 4 Fort Martin 1 Keystone 1 Drakelow C 3 Bull Run 1 Frankin Avon Lake

936 626 350 648 936 790 450 576 936 375 950 110 215

— — — — — — — — — — — — —

— — — — — — — — — — — — —

— — — — — — — — — — — — —

— — — — — — — — — — — — —

— — — — — — — — — — — — —

— — — — — — — — — — — — —

3-17

Fuel

Start Year

Lignite

1970 1970 1969 1969 1968 1968 1967 1967 1967 1967 1967 1962 1959


Table 3-4 Supercritical Steam Turbines Supplied by Alstom Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Start Year

RWE

Neurath

1050

4279

295

1112

600

1121

605

Due 2008

ELSAM

Nordjylland 3

411 (+450 MWt)

4134

285

545

582

1076

580

1998

Reheat #2 Patnow Schkopau Houshi Houshi Houshi

1076

580

Elektrim Megadex Kraftwerk Schkopau Hua Yang Electric Power Corp Hua Yang Electric Power Corp Hua Yang Electric Power Corp

460 450 600 600 600

4205 4133 4103 4103 4103

290 285 283 283 283

1011 1010 1005.8 1005.8 1005.8

544 545 541 541 541

1054 1040 1050 1050 1050

568 560 566 566 566

2006 1995 2004 2005? 2006

Veag Vestkraft Nuon Intergen Fortum & TVO

Lippendorf S Vestkraft 3 Hemweg 8 Millmerran Meri Pori

933 400 630 400 550

4061 4002 3770 3597 3583

280 276 260 248 244

1031 1040 500 1053 471.2

555 560 540 567 540

1031 1040 1054 1105 1034

555 560 568 596 560

2000 1992 1994 2002 1994

3-18


Table 3-5 Supercritical Steam Turbines Supplied by Ansaldo Energia Unit Net Output Owner

Unit

Energy E2

Avedore 2

Main Steam Pressure


Reheat Temperature

MW

psi

bar

°F

°C

°F

°C

Start Year

390 (+570 MWt)

4134

285

1076

580

1112

600

2001

3-19


Table 3-6 Supercritical Steam Generators Supplied by Babcock & Wilcox (B&W) Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

AEP-Ohio Power

Philo 6

125

4550

314

1150

621

1050

566

Cyc

1953

1000

538

Forest Grove 1 Monticello 3

775 775

3850 3850

265 265

1005 1010

541 543

1005 1005

541 541

Coal Coal

1973 1973

1957 (ret 1979) — 1977

Cardinal 3

650

3850

265

1005

541

1005

541

Coal/Oil

1972

1977

Andrus 1 De Cordova 1 Tradinghouse Creek 2 Baxter Wilson 2 Tradinghouse Creek 1

750 775

3850 3850

265 265

1005 1005

541 541

1005 1005

541 541

O&G Gas

1970 1970

1974 1973

775

3850

265

1005

541

1005

541

Gas

1968

1972

783

3850

265

1005

541

1000

538

O&G

1967

1972

565

3850

265

1005

541

1005

541

Gas

1966

1969

Rockport 2

1300

3845

265

1010

543

1000

538

Coal

1979

1989

Rockport 1

1300

3845

265

1010

543

1000

538

Coal

1979

1984

AEP-Appalachian Pwr. AEP-Ohio Power AEP-Ohio Power

Mountaineer Gavin 2 Gavin 1

1300 1300 1300

3845 3845 3845

265

1010 1010 1010

543 543 543

1000 1000 1000

538

265 265

538 538

Coal Coal Coal

1970 1969 1969

1980 1975 1974

AEP-Appalachian Pwr.

Amos 3

1300

3845

265

1010

543

1000

538

Coal

1969

1973

Entergy Gulf States Entergy Gulf States Entergy Gulf States

Sabine 4 Willow Glen 4 Roy S. Nelson 4

575 575 592

3840 3840 3840

265 265 265

1005 1005 1005

541 541 541

1000 1000 1005

538 538 541

Gas O&G Gas

1969 1967 1966

1974 1973 1970

Reheat #2

TXU Electric TXU Electric AEP-Ohio Power/Buckeye Power Entergy Mississippi TXU Electric TXU Electric Entergy Mississippi TXU Electric AEP-Indiana & Michigan Power AEP-Indiana & Michigan Power

3-20


Unit Net Output

Main Steam Pressure


Reheat Temperature

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

Moss Landing 7

739

3830

264

1005

541

1005

541

O&G

1963

1968

Moss Landing 6

739

3830

264

1005

541

1005

541

O&G

1963

1967

La Cygne 1

844

3825

264

1010

543

1005

541

Cyc

1968

1973

Canal 1

543

3825

264

1005

541

1001

538

Oil

1964

1968

Hatfield Ferry 3

575

3810

263

1005

541

1010

543

Coal

1967

1972

Hatfield Ferry 2

575

3810

263

1005

541

1005

541

Coal

1966

1969

Hatfield Ferry 1

575

3810

263

1005

541

1005

541

Coal

1966

1969

Sibley 3 Fort Martin 2

419 576

3810 3810

263 263

1005 1010

541 543

1005 1010

541 543

Cyc Coal

1965 1965

1968 1968

Bailley 8

422

3810

263

1005

541

1005

541

Cyc

1964

1968

Dayton P&L/Cincinnati G&E/AEP-Columbus Southern Pwr.

J.M. Stuart 4

600

3805

262

1005

541

1005

541

Coal

1969

1974


J.M. Stuart 3

610

3805

262

1005

541

1005

541

Coal

1967

1972


J.M. Stuart 1

610

3805

262

1005

541

1005

541

Coal

1965

1971


J.M. Stuart 2

610

3805

262

1005

541

1005

541

Coal

1965

1970

Owner

Duke Energy Moss Landing LLC Duke Energy Moss Landing LLC Kansas City Power & Light Co./Kansas Gas & Electric Co. Southern Energy Allegheny Energy Services Allegheny Energy Services Allegheny Energy Services Utilicorp United Monongahela Power Northern Indiana Public Service

3-21


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

Detroit Edison Co.

Monroe 4

800

3800

262

1006

541

1002

539

Coal

1969

1973

Detroit Edison Co.

Monroe 2

800

3800

262

1005

541

1005

541

Coal

1967

1972

Detroit Edison Co.

Monroe 1

800

3800

262

1005

541

1005

541

Coal

1967

1971

PG&E National Energy

Brayton Point 3 Reheat #2 Muskingum 5

643

3800

262

1005

541

1030

554

Coal

1965

1969

1055

568

591

3800

262

1000

538

1025

552

Coal

1965

1968

Cardinal 2

590

3800

262

1000

538

1025

552

Coal

1964

1967

1050

566

1025

552

Coal

1964

1966

1050

566

AEP-Ohio Power AEP-Ohio Power/ Buckeye Power

Reheat #2

AEP-Ohio Power/ Buckeye Power

Cardinal 1 Reheat #2

Public Service Co. of Oklahoma Public Service Co. of Oklahoma Public Service Co. of Oklahoma NRG Energy FirstEnergy Orion Power Midwest LLC FirstEnergy Unidentified Customer Compania Electrica de Langreo SA Potomac Electric Power

590

3800

262

1000

538

Riverside 2

450

3792

261

1005

541

1005

541

Gas

1970

1976

Riverside 1

450

3792

261

1005

541

1005

541

Gas

1969

1974

Northeastern 2

473

3792

261

1005

541

1005

541

Gas

1966

1970

Eastlake 5 W.H. Sammis 7

680 600

3785 3785

261

1005 1005

541 541

1005 1005

541

261

541

Coal Coal

1968 1967

1972 1971

Avon Lake 9

680

3785

261

1005

541

1005

541

Coal

1966

1969

W.H. Sammis 6 Midwest USA

623 500

3785 3775

261 260

1005 1085

541 585

1005 1085

541 585

1965 2004

1968 2008

Lada 4

350

3760

259

1005

541

1005

541

Coal Coal Coal & Gas

1975

1981

Chalk Point 2

365

3750

259

1000

538

1050

566

1961

1965

1000

538

Reheat #2

3-22

Coal


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

Potomac Electric Pwr.

Chalk Point 1

365

3750

259

1000

538

1050

566

Coal

1961

1964

1000

538

Reheat #2

Ente Nazionale per l'Energia Elettrica Ente Nazionale per l'Energia Elettrica Ente Nazionale per l'Energia Elettrica Ente Nazionale per l'Energia Elettrica Reliant Energy Inc. Reliant Energy Inc.

Torrevaldaliga 4

660

3700

255

1004

540

1004

540

O&G

1979

1985

Torrevaldaliga 3

660

3700

255

1004

540

1004

540

O&G

1974

1984

Torrevaldaliga 2

660

3700

255

1004

540

1004

540

O&G

1974

1983

Torrevaldaliga 1

660

3700

255

1004

540

1004

540

O&G

1974

1982

P.H. Robinson 4 Cedar Bayou 2

750 750

3700 3700

255

1010 1010

543 543

1005 1005

541

255

541

Gas Gas

1969 1967

1973 1972

Reliant Energy Inc.

Cedar Bayou 1

750

3700

255

1010

543

1005

541

Gas

1967

1972

Detroit Edison Co. Zhejiang Power Bureau Zhejiang Power Bureau Zhejiang Power Bureau Zhejiang Power Bureau Hebei Electric Power Admin Hebei Electric Power Admin

Monroe 3

800

3692

255

1002

539

1002

539

Coal

1968

1973

Lanxi 4

600

3683

254

1059

571

1056

569

Coal

2003

2009

Lanxi 3

600

3683

254

1059

571

1056

569

Coal

2003

2008

Lanxi 2

600

3683

254

1059

571

1056

569

Coal

2003

2008

Lanxi 1

600

3683

254

1059

571

1056

569

Coal

2003

2007

Xibaipo Ill-2

600

3683

254

1009

543

1056

569

Coal

2003

2007

Xibaipo Ill-1

600

3683

254

1009

543

1056

569

Coal

2003

2006

Northern States Power

Allen S. King 1

574

3675

253

1005

541

1005

541

Cyc

1964

1968

Ente Nazionale per l'Energia Elettrica

LaSpezia

600

3675

253

1000

538

1025

552

Coal/ O&G

1962

1967

1050

566

Reheat #2

3-23


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

Los Angeles Dept of Water & Power

Haynes 6

343

3665

253

1000

538

1025

552

O&G

1963

1967

1050

566

538

1025

552

O&G

1963

1966

566

Reheat #2

Los Angeles Dept of Water & Power

Haynes 5

343

3665

253

1000

Arizona Public Service

Four Corners 5

800

3660

252

1008

542

1050 1008

542

Coal

1965

1970

Arizona Public Service

Four Corners 4

800

3660

252

1008

542

1008

542

Coal

1965

1969

Duke Power Company

Belews Creek 2

1100

3650

252

1007

542

1000

538

Coal

1969

1975

Duke Power Company Tennessee Valley Authority Tennessee Valley Authority Tennessee Valley Authority

Belews Creek 1

1100

3650

252

1007

542

1000

538

Coal

1969

1974

Cumberland 2

1300

3650

252

1003

539

1003

539

Coal

1967

1973

Cumberland 1

1300

3650

252

1003

539

1003

539

Coal

1967

1972

Paradise 3

1150

3650

252

1003

539

1003

539

Cyc

1965

1969

Reliant Energy Inc.

P.H. Robinson 2

477

3650

252

1005

541

1005

541

Gas

1964

1967

Reliant Energy Inc.

P.H. Robinson 1

477

3650

252

1005

541

1025

1005

Gas

1964

1966

Northern Indiana Public Service Northern Indiana Public Service Georgia Power Georgia Power Tokyo Electric Power Public Service Electric & Gas

Rollin M. Schahfer

520

3635

251

1005

541

1005

541

Cyc

1970

1976

Michigan City 12

500

3635

251

1005

541

1005

541

Cyc

1967

1974

Harllee Branch 4 Harllee Branch 3 Anegasaki 1

490 480 640

3625 3625 3625

250 250

538 538 543

1000 1000 1055

538 538

250

1000 1000 1010

568

Coal Coal Oil

1966 1965 1964

1969 1968 1967

Hudson 1

420

3625

250

1000

538

1025

552

Cyc

1960

1964

1050

566

Reheat #2

Reheat #2

3-24


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Order Year

Start Year

AEP-Indiana & Michigan Power

Tanners Creek 4

580

3625

250

1000

538

1025

552

Cyc

1959

1964

1050

566

1050

566

Coal

1956

1960

1050

566

1050

566

Cyc

1956

1960

1000

538

Reheat #2

AEP-Appalachian Power

Sporn 5 Reheat #2

AEP-Indiana & Michigan Power

Breed 1 Reheat #2

Ameren (Union Electric) Ameren (Union Electric) TXU Electric Consumers Energy Constellation Power Source Generation Inc.

AES AES AES AES Intergen Intergen

450

3625

250

1050

566

450

3625

250

1050

566

Sioux 2

489

3620

250

1000

538

1000

538

Cyc

1962

1968

Sioux 1

489

3620

250

1005

541

1005

541

Cyc

1962

1967

Valley 2 J.H. Campbell 2

559 385

3600 3600

248

1005 1005

541 541

1005 1005

541

248

541

Gas Coal

1964 1964

1967 1967

H.A. Wagner 3

322

3600

248

1005

541

1005

541

Coal

1962

1966

1005

541

Reheat #2 Redondo Beach 8

Redondo Beach 7 Alamitos 6 Alamitos 5 Millmerran 2 Millmerran 1

480

3599

248

1005

541

1000

538

O&G

1964

1967

480

3599

248

1005

541

1000

538

O&G

1964

1966

480 480 420 420

3599 3599 3596 3596

248 248 248 248

1000 1000 1054 1054

538 538 568 568

1000 1000 1105 1105

538 538 596 596

O&G O&G Coal Coal

1963 1963 1999 1999

1966 1965 2002 2002

3-25


Table 3-7 10 Supercritical Steam Generators Supplied by Burmeister & Wain Energy (BWE) Unit Net Output Owner

Unit

Dragon Power

Lanshan

Energy E2

Avedore 2

ELSAM

Skaerbaek 3

Main Steam Pressure


Reheat Temperature

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

4 x 660 390 (+570 MWt) 411 (+450 MWt)

4424

305

1112

600

1112

600

Coal

Due 2009

4424

305

1080

582

1112

600

Gas/Oil

2001

4205

290

1080

582

1076

580

Nat Gas

1997

1076

580 Bit./Oil

1998

Reheat #2

10

ELSAM

Nordjylland 3

411 (+450 MWt)

E.ON

Staudinger 5

500 gross

3800

Elsam

Fyns 7

400 gross

3626

4205

290

1080

582

1076

580

262

1013

545

1044

562

250

1004

540

1004

540

Utility Steam Boilers List of References, Burmeister & Wain Energy A/S BWE-20-0004rev.3 http://www.bwe.dk/usc-ref.html

3-26

Coal/ Oil Coal/ Oil

1992 1991


Table 3-8 Supercritical Steam Generators Supplied by Foster-Wheeler Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Genpower

Longview

695

3626

250

1112

600

1112

600

Bit.

2011

AEP

Mitchell 1

760

3500

241

1000

537

1025

552

Bit.

1971

1050

565

1025

552

Bit.

1971

1050

565

1025

552

Bit.

1971

1050

565

1025

552

Bit.

1971

1050

565

Reheat #2

AEP

Mitchell 2

AEP

Amos 1

Reheat #2 Reheat #2

AEP

Amos 2 Reheat #2

760

3500

241

1000

537

3500

241

1000

537

760 760

3500

241

3-27

1000

537


Table 3-9 Supercritical Steam Generators Supplied by Hitachi Power Systems Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Tohoku Electric Power Co. China Huaneng Group Co.

Noshiro 1 Zouxian 7

600 1000

3868 3807

267 262

1008 1121

542 605

1053 1117

567 603

Coal Coal

1993 2007

China Huaneng Group Co.

Zouxian 8

1000

3807

262

1121

605

1117

603

Coal

2008

MidAmerican Energy Co.

Council Bluffs 4

853

3803

262

1057

569

1103

595

PRB

2007

We Energies

Elm Road 1

677

3800

262

1055

568

1055

568

Coal

2009

We Energies

Elm Road 2

677

3800

262

1055

568

1055

568

Coal

2010

Electric Power Develop. Co.

Tachibanawan 2

1050

3754

259

1121

605

1135

613

Coal

2000

CS Energy, Ltd.

Kogan Creek

750

3754

259

1009

543

1045

563

Coal

2006

Tohoku Electric Power Co.

Haramchi 2

1000

3683

254

1119

604

1116

602

Coal

1998

Tokyo Electric Power Co. China Huaneng Group Co. China Huaneng Group Co. China Huaneng Group Co. China Huaneng Group Co. China Huaneng Group Co.

Hitachinaka 1 Qinbei 1 Qinbei 2 Shantou Taicang 1 Taicang 2

1000 600 600 600 600 600

3683 3683 3683 3683 3683 3683

254 254 254 254 254 254

1119 1060 1060 1060 1060 1060

604 571 571 571 571 571

1116 1056 1056 1056 1056 1056

602 569 569 569 569 569

Coal Coal Coal Coal Coal Coal

2003 2004 2005 2006 2007 2007

China Huaneng Group Co. China Huaneng Group Co.

Yahekou 1 Yahekou 2 Torrevaldaliga Nord 4 Torrevaldaliga Nord 3 Torrevaldaliga Nord 2

600 600

3683 3683

254 254

1060 1060

571 571

1056 1056

569 569

Coal Coal

2007 2007

660

3655

252

1119

604

1134

612

Coal

2008

660

3655

252

1119

604

1134

612

Coal

2008

660

3655

252

1119

604

1134

612

Coal

2009

450

3626

250

1006

541

1029

554

Oil

1970

Enel Power Enel Power Enel Power Kansai Electric Power Co.

Kainan 1

3-28


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Kansai Electric Power Co. Tokyo Electric Power Co. Tokyo Electric Power Co.

Reheat #2 Kainan 2 Anegasaki 1 Anegasaki 2

450 600 600

3626 3626 3626

250 250 250

1009 1009 1009

543 543 543

1054 1054 1054 1054

568 568 568 568

Tokyo Electric Power Co.

Anegasaki 3

600

3626

250

1009

543

1054

568

Tokyo Electric Power Co. Kansai Electric Power Co.

Kashima 4 Kainan 4

600 600

3626 3626

250 250

1009 1009

543 543

1054 1054

568 568

Oil Oil Oil O&G, LNG, LPG Oil Oil

Chugoko Electric Power Co.

Tamashima 3

500

3626

250

1009

543

1006

541

Oil

3-29

Fuel

Start Year

1970 1991 1976 1974 1972 1973 1974


Table 3-10 Supercritical Steam Turbines Supplied by Hitachi Power Systems Unit Net Output Owner

Unit

We Energies

Elm Road

MidAmerican Energy J-Power EPCOR Tohoku Electric Power Co. Hokkaido Electric Tokyo EPCo Kyushu Electric Chubu Electric Soma Kyodo Power

Main Steam Pressure


Reheat Temperature

MW

psi

bar

°F

°C

°F

°C

Start Year

2 x 615

3815

263

1051

566

1051

566

Due 2009– 2010

Council Bluffs 4

790

3684

254

1049

565

1099

593

Due 2007

Isogo 2 Genesee 3

600 450

3626 3626

250 250

1112 1058

600 570

1150 1054

621 568

2009 2005

1000 gross

3553

245

1112

600

1112

600

1998

700 gross

3553

245

1112

600

1112

600

2002

1000 gross 700 gross 700 gross 1000 gross

3553 3495 3495 3495

245 241 241 241

1112 999 999 999

600 537 537 537

1112 1049 1049 1049

600 565 565 565

2003 1989 1992 1994

Haramachi 2 Tomato – Atsuma 4 Hitachinaka 1 Matsuura 1 Hekkinan 2 Shinichi 1

3-30


Table 3-11 Supercritical Steam Generators Supplied by Ishikawajima-Harima Heavy Industries (IHI) Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Kobe Steel Ltd. Tokyo Electric Power Tokyo Electric Power Kansai Electric

Shinko Kobe 1 Kashima 2 Anegasaki 4 Himeji II 6 Reheat #2 Sendai 1 Toyama Shinko 1 Nishi Nagoya 6 Anan 4 Buzen 1 Kudamatsu 3 Hirono 1 Toyama Shinko 2 Higashi Niigata 2 Nakoso 9 Chita II Tomatoo Atsuma 2 Sendai 2 Shin Onoda 1 Shin Onoda 2 Akoh 2 Higashi Ogishima 2 Nankoh 3 Hirono 4 Shiriuchi 2 Hekinan 3 Noshiro 2 Reihoku 1 Nano-Ohta 2 Hekinan 4

700 600 600 600

3495 3626 3626 3626

241 250 250 250

1006 1006 1006 1006

541 541 541 541

2004 1971 1972 1973

3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626 3626

250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250 250

1006 1006 1009 1006 1006 1006 1006 1006 1006 1006 1009 1006 1006 1006 1006 1006 1008 1006 1008 1006 1009 1058 1058 1107 1060

541 541 543 541 541 541 541 541 541 541 543 541 541 541 541 541 542 541 542 541 543 570 570 597 571

568 568 568 554 568 540 540 541 568 540 568 568 568 568 568 569 568 568 568 568 568 568 569 568 568 569 596 568 595 596

Coal Oil Oil Oil

500 500 500 450 500 700 600 500 600 600 700 600 500 500 500 600 1000 600 1000 350 700 600 700 700 1000

1054 1054 1054 1029 1054 1004 1004 1006 1054 1004 1054 1054 1054 1054 1054 1056 1054 1054 1054 1054 1054 1054 1056 1054 1054 1056 1105 1054 1103 1105

Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil Oil LNG LNG Oil Oil Coal Coal Coal Coal Coal

1974 1974 1975 1976 1977 1979 1980 1981 1983 1983 1985 1985 1985 1986 1987 1988 1991 1991 1993 1998 1993 1994 1995 1998 2001

Kyushu Electric Hokuriku Electric Chubu Electric Shikoku Electric Kyushu Electric Chugoku Electric Tokyo Electric Hokuriku Electric Tohoku Electric Jooban Kyoodo Chubu Electric Hokkaido Electric Kyushu Electric Chugoku Electric Chugoku Electric Kansai Electric Tokyo Electric Power Kansai Electric Tokyo Electric Hokkaido Electric Chubu Electric Tohoku Electric Kyushu Electric Hokuriku Electric Chubu Electric

3-31


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Chubu Electric Sumitomo Metal Ind. Electric Power Dev. Callide Power Callide Power Tarong Energy Corp. Hokkaido Electric Electric Power Dev. Electric Power Dev.

Hekinan 5 Kashima 1 Tachibana- Wan 1 Callide 3 Callide 3 Tarong North 1 Tomato Azuma 4 Isogo 2 Isogo 1

1000 507 1050 420 420 450 700 600 600

3626 3626 3742 3742 3742 3742 3757 3945 3989

250 250 258 258 258 258 259 272 275

1060 1008 1121 1056 1056 1056 1117 1121 1121

571 542 605 569 569 569 603 605 605

1105 1054 1135 1054 1054 1056 1116 1153 1135

596 568 613 568 568 569 602 623 613

Coal Coal Coal Coal Coal Coal Coal Coal Coal

2002 2007 2000 2001 2001 2003 2002 2009 2002

3-32


Table 3-12 Supercritical Steam Generators Supplied by Doosan Babcock (formerly Mitsui Babcock) Unit Net Output Owner

Unit

Nuon (Netherlands) (China)

Hemweg 8 Changshu

(Finland)

Meri-pori

(Hong Kong) (Hong Kong) (China)

Castle Peak “B” (Subcritical OnceThrough) Castle Peak “A” (Subcritical OnceThrough) Yaomeng (vertical wall repowering)

Main Steam Pressure


Reheat Temperature Order Year

Start Year

568 569

—

1992 2005

1040

560

—

1993

541

1002

539

1981

—

—

—

—

—

—

—

—

—

—

—

MW

psi

bar

°F

°C

°F

°C

680 600 630 (gross)

3770 3755

260 259

1004 1056

540 569

1054 1056

3465

239

1004

540

2425

167

1006 —

4 x 680 4 x 350 —

3-33


Table 3-13 Supercritical Steam Generators Supplied by Mitsubishi Heavy Industries (MHI) Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Taiwan FPCC Taiwan FPCC Taiwan FPCC Tohoku EPCO Chugoku EPCO Taiwan FPCC Taiwan FPCC China China China China Tokyo EPCO China China China China China China China China China China Joban Joint EPCO Kyusyu EPCO Tokyo EPCO Chubu Hokuriku EPCO Soma Joint EPCO Hokuriku EPCO KOBELCO

FP-1 1 FP-1 2 FP-1 3 Haramachi 1 Misumi 1 UP-1 A UP-1 B CP-1 1 CP-1 2 CP-1 3 CP-1 4 Hirono 5 Yuhan Yuhan Yuhan Yuhan Kanshan Kanshan Yingkou Yingkou Taizhou Taizhou Nakoso 8 Matsuura 1 Hirono 3 Hekinan 1 Tsuruga 1 Shinchi 2 Tsuruga 2 Kobe 1

600 600 600 1000 1000 600 600 600 600 600 600 600 1000 1000 1000 1000 600 600 600 600 1000 1000 600 700 1000 700 500 1000 700 700

3740 3740 3740 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3553 3500 3500 3500 3500 3500 3500 3500 3500

258 258 258 245 245 245 245 245 245 245 245 245 245 245 245 245 245 245 245 245 245 245 241 241 241 241 241 241 241 241

1000 1000 1000 1050 1112 1000 1000 1000 1000 1000 1000 1112 1112 1112 1112 1112 1112 1112 1112 1112 1112 1112 1000 1000 1000 1000 1050 1000 1100 1000

538 538 538 566 600 538 538 538 538 538 538 600 600 600 600 600 600 600 600 600 600 600 538 538 538 538 566 538 593 538

1050 1050 1050 1100 1112 1050 1050 1050 1050 1050 1050 1112 1112 1112 1112 1112 1112 1112 1112 1112 1112 1112 1050 1050 1050 1050 1050 1050 1100 1050

566 566 566 593 600 566 566 566 566 566 566 600 600 600 600 600 600 600 600 600 600 600 566 566 566 566 566 566 593 566

Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal/Oil Coal Oil/Coal Coal Coal Coal Coal Coal

1999 1999 2002 1997 1998 2000 2002 2000 2000 2002 2002 2004 2007 2007 2007 2007 2007 2007 2007 2007 2007 2007 1983 1989 1989 1991 1991 1995 2000 2002

3-34


Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Fuel

Start Year

Kyusyu EPCO Kansai EPCO

Reihoku 2 Maizuru 1

700 900

3500 3500

241 241

1100 1100

593 593

1100 1100

593 593

Coal Coal

2003 2004

3-35


Table 3-14 Supercritical Steam Turbines Supplied by Mitsubishi Heavy Industries (MHI) Unit Net Output Owner

Electric Power Dev. Co Xcel Energy (Colorado USA) Chubu Hokuriku J-Power Chugoku EPCO Hokuriku EPCO Kyusyu Tokyo EPCO Kansai

Main Steam Pressure


Reheat Temperature

Unit

MW

psi

bar

°F

°C

°F

°C

Tachibanawan 2

1050

3641

251

1112

600

1130

610

Comanche 3

750

3565

244

1050

566

1100

593

600 700 500 1000 1000 700 700 600 900

3553 3495 3495 3495 3495 3495 3495 3495 3495

245 241 241 241 241 241 241 241 241

999 999 999 999 1112 1099 1099 1099 1099

537 537 537 537 600 593 593 593 593

1049 1112 1112 1112 1112 1099 1099 1099 1099

565 600 600 600 600 593 593 593 593

FP 15 Hekinan 3 Nanao – Ohta 1 Matsuura 2 Misumi 1 Tsuraga 2 Reihoku 2 Hirono 5 Maizuru 2

3-36

Fuel

Start Year

2001 PRB

Due 2009 2001 1993 1995 1997 1998 2000 2003 2004 2004


Table-3-15 Supercritical Steam Turbines Supplied by Siemens-Westinghouse Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

Start Year

Shanghai Municipal Elect Company

Waigaoqiao 3

1000

3916

270

1112

600

1112

600

2009

Vattenfall

Lippendorf

3873

267

1029

554

1081

583

2000

RWE Energy (China) J-Power Shanghai Municipal Elect Company Shanghai Municipal Elect Company

Niederaussem Yuhuan Isogo 1

930 + heat 965 4 x 1000 600

3844 3807 3641

265 263 251

1069 1112 1112

576 600 600

1112 1112 1130

600 600 610

2002 2008 2002

Waigaoqiao 2

900

3626

250

1000

538

1040

560

2004

Waigaoqiao 1

900

3626

250

1000

538

1040

560

2004

Kogan Creek Mitchell 1 & 2

750 2 x 760

3626 3495

250 241

1040 999

560 537

1040 1026

560 552

2007 1971

AEP

3-37


Table 3-16 Supercritical Steam Turbines Supplied by Toshiba Unit Net Output

Main Steam Pressure


Reheat Temperature

Owner

Unit

MW

psi

bar

°F

°C

°F

°C

KCP&L Hokuriku EPCO Hokuriku Kyusyu Kansai

Iatan 2 Tsuruga 2 Nana Ohta 2 Reihoku 2 Maizuru Tsuruga 1 Noshiro 2 Harmachi 1 Tachibanawan 1 Hekinan 4 Hekinan 5

850 700 700 700 900 500 600 1000 700 1000 1000

3688 3495 3495 3495 3495 3495 3495 3495 3495 3495 3495

254 241 241 241 241 241 241 241 241 241 241

1099 1099 1099 1099 1099 1049 1049 1049 1049 1049 1049

593 593 593 593 593 565 565 565 565 565 565

1099 1099 1099 1099 1049 1099 1099 1099 1099 1099

593 593 593 593 565 593 593 593 593 593

Tohoku Electric Tohoku Shikoku Chubu Chubu

3-38

Start Year

2010 2000 1998 2003 2010 1991 1994 1997 2000 2001 2002


4

CURRENT DESIGN TRENDS AND ISSUES Unit Size and Scale Issues of engineering design that should be considered when selecting unit size include: •

Capital cost

•

Constructability

•

Project schedule

•

Redundancy and other design for reliability versus replacement power cost risk

•

Technical risk

General Capital Cost Considerations For the normal range of pulverized coal plant sizes (i.e., 200 MW to 1300 MW), steam generator economies-of-scale (and overall plant costs) generally follow a rough rule-of-thumb trend: For a completely installed and operational plant, the cost ratio is equivalent to the size ratio raised to an exponent of about 0.7. Figure 4-1 illustrates the general relationship of capital cost versus generating unit gross output. Project Cost versus Unit Size 2000 1800

s 1600 n o i l l 1400 i m $ , t s 1200 o C t 1000 c e j o r P 800 600 400 0

200

400

600

800

1000

1200

1400

1600

Unit Size, MW gross

Figure 4-1 Trend in Cost versus Unit Gross Output Rating

In practice, the cost versus size curve is not completely smooth and linear (i.e., note the “kinks” in the Figure 4-1 plot). Step changes in cost occur at various points due to size limitations for major components. When unit size crosses these threshold points or “breakpoints,” it becomes

4-1


necessary to use multiple components rather than a single larger component. It is advantageous to be aware of the unit sizes at which these step changes occur. For some size increments, this is especially important, as several systems are subject to step ch anges within a relatively narrow range of size increase. Factors that may influence the unit output ratings at which cost breakpoints occur for a given project may include fuel type, steam conditions, boiler inlet air and outlet gas temperatures, and atmospheric conditions. The lists below are based on assumptions for all these factors, with notes on how component limitations and cost breakpoints may be affected by changes in the assumed factors. It is also assumed that redundancy and reliability approaches would be consistent across the range of unit output capacity. A change in approach to system component redundancy is likely to also change the cost breakpoints. There will be many component sizes and quantities that change as unit size increases; listed below are only the major components that would result in more significant step chang es in cost.

Assumed Basis The basis for the discussion below is as follows: •

•

•

Fuel: typical eastern U.S. bituminous, with a higher heating value of 12,500 Btu/lb (29,000 kJ/kg) Steam conditions: 3500 psig (241 barg) and 1050°F (565°C) steam at the high-pressure turbine inlet, and 1050°F (565°C) steam at the intermediate-pressure turbine inlet Air heater designed for about 100°F (40°C) inlet air and about 300°F (150°C) outlet flue gas temperature

•

Atmospheric conditions: 95°F (35°C) dry bulb, 80°F (27°C) wet bulb

•

Single train back-end components up to the maximum loads possible

•

Single main boiler feedwater pump

Component Limitations/Cost Step Changes Turbine-Generator •

•

Current maximum available generator size for 60 Hz applications is 1200 MVA, which corresponds to a unit gross output limit of about 1080 MW from a single generator (assuming a 0.90 power factor). Turbines designed for 60 Hz grid systems are inherently limited to smaller maximum sizes (and smaller mass flow and power output) than those for 50 Hz grids. This is because the higher rotational speed used in 60 Hz machines applies greater centrifugal force to the blade root and turbine disks (for a given blade size). This influence must be considered along with material strength limits imposed by increasing temperature and pressure and tradeoffs of turbine material costs versus the benefits of larger unit size. For unit output greater than 1080 MW (60 Hz), a cross-compound unit will be required on two spindles rather than tandem compound steam turbine on one shaft.

4-2


•

•

Cross-compound designs result in a step increase in turbine-generator capital cost and associated pedestal, isophase bus, turbine building area, and maintenance costs. There are no significant changes in component size limitation or cost breakpoints with different assumptions for fuel, steam conditions, air and gas temperatures, or atmospheric conditions.

Boiler Feedwater Pumps (BFP) •

•

•

•

Current maximum available BFP size will support 1300 MW (one known supplier) or 1000 MW (multiple suppliers). Switch from one to two BFPs will increase equipment cost plus foundations, electrical, and controls; however, reliability will be improved. With more advanced steam conditions, the breakpoint in unit size regarding BFP size will be at slightly higher loads. Selection of BFP size and driver type is also influenced by relationships between steam production, steam turbine ratings, auxiliary and cogeneration steam flows, and transmission system characteristics. For some combinations, steam turbine drivers are preferred whereas electric drives are preferred in other cases. Plant layout may also influence this selection.

Other Pumps and Drivers •

As with boiler feedwater pumps, breakpoints in size selection for c ondensate pumps and cooling water circulating pumps are influenced by unit size and available pump sizes along with a variety of plant-design-specific factors. Auxiliary equipment may use variable speed, fixed speed, or two speed electric motor drivers, or steam turbine drivers, each of which is preferred for different size ranges.

Feedwater Heater Trains •

•

Switch from one to two high-pressure (HP) heater trains at about 500 MW due to valve size limitations at supercritical pressures. This breakpoint will be somewhat lower for cycles with higher steam pressures.

Cooling Towers •

•

For mechanical draft systems, no limit is imposed solely by unit size. Cell size and count are fully customizable. For natural draft towers, a one tower to two tower transition occurs at about 1000 MW. – If ambient dry bulb or wet bulb temperatures are higher than assumed basis, transition from one to two towers will be at a lower unit design load. – If steam conditions are higher, breakpoint from one to two towers will be at a higher unit design load.

High-Energy Piping •

For main steam, the one lead to two lead transition occurs at about 700 MW; the two lead to four lead transition occurs at about 1100 MW.

4-3


•

•

For cold and hot reheat lines, the transition from one to two leads occurs at 700 MW. Four leads are required above 1100 MW. If steam conditions are higher, the transition breakpoints will occur at lower unit sizes.

Single- and Dual-Fireball Design • •

•

• •

•

Applies to tangentially fired units For typical eastern U.S. bituminous-fired boilers, the transition point from one to two fireballs (i.e., a divided furnace) ranges from 700 MW to 1100 MW. For divided furnace (dual fireball) designs, capital costs will increase in step changes with increases in the number of burners, coal pipes, and burner penetrations in waterwalls. Maintenance costs increase due to the increased number of components in the firing system. For lower rank coals (subbituminous and lignite fuels), the transition breakpoint will be lower. For higher steam conditions, the transition breakpoint will generally be higher.

Vertical-Tube and Spiral-Tube Furnace Design •

•

Spiral-tubed furnaces are often used in the design of variable-pressure supercritical units to: – Achieve adequate mass flows and velocities for tube cooling during startup and low load operation – Provide more even heating of steam in furnaces where heat flux may vary across the width of the walls There is some commercial experience and a growing commercial interest in offering straight, vertical-wall furnace designs that address tube cooling and heat flux imbalances through use of rifled tubing, orificing, and carefully designed low-load recirculation systems.

Ash Removal •

•

•

Wet hydraulic sluice bottom ash removal: – There is no theoretical limit or major cost increments; howe ver, it is necessary to design for capability to convey ash from two or more points simultaneously. – The largest size installed to date is 1300 MW. Wet drag chain bottom ash removal: – 50 tph (45 tonne/hr) is the nominal limit of drag chain sizing. – With typical sizing at twice the ash production rate, this would allow up to 2000 MW unit size when firing coal up to 20% ash and allowing up to 4 hours of storage time. Dry bottom ash removal: – Limits are similar to those for wet drag chain if the dry system is based on a flight conveyor. – For a system with a vibratory conveyor, the limit is much higher (>1500 MW unit size). – The largest dry bottom ash removal system installed to date is 900 MW.

4-4

EPRI Proprietary Licensed Licensed Material

Main Coal-Feed Conveyor •

• •

One pair of (redundant) main feed conveyors, using conventional 96 inch (~2.5 m) belts, will feed up to 8 units at 750 MW each, 6 units at 1000 MW each, and 4 units at 1500 MW each. For lower rank coals, unit size or quantity supported by conventional belts will be less. For higher steam conditions, unit size or quantity supported by conventional belts will be greater.

Sootblower Lance Lengths •

Above 1300 MW, sootblowers may require special design for extending over unusually long spans.

Rotary Air Heaters •

• •

•

•

The largest air heaters that are currently available, from the two international manufacturers, will support unit size up to 700 MW with one train; 1400 MW with two trains (bituminous coals). Additional ductwork (and cost) is required to accommodate ac commodate a second train of air a ir heaters. Higher inlet air temperature or lower exit gas temperature will reduce the unit size that can be served by the largest available air heater. Use of lower rank coals will reduce the unit size serviceable by the largest available air heaters. Higher steam conditions will increase the unit size that can be serviced by the largest available air heater.

Induced Draft (ID) Fans •

•

The largest centrifugal fans that are currently available will support maximum unit size of about 460 MW gross with a single fan. The largest currently available axial fans will support maximum unit size of about 700 MW gross with a single fan.

•

Lower rank coals will result in lower unit ratings with maximum-sized fans.

•

Lower exit gas temperatures will result in higher unit ratings with maximum-sized fans.

•

Higher steam conditions will result in higher unit ratings with maximum-sized fans.

Particulate Removal • •

The number of dry ESP or wet ESP modules will be proportional to the number of ID fans. Step changes increase capital costs co sts for separate structures and foundations, additional insulation, etc.

Flue Gas Desulfurization •

Wet Scrubber: Step change from one absorber module to two at about 1100 MW – For lower rank coals, the maximum single module size will correspond to a lower MW level

4-5


– –

For higher steam conditions, the maximum single module size will correspond to a higher MW level For lower exit gas temperatures, the maximum ma ximum single module size will correspond to a higher MW level

Construction and Schedule Considerations

Construction Issues With larger plants, impacts of equipment sizing on transportation and erection must be considered. If transportation modes or access are limited, it may not be possible to ship larger modules and preassembled components of larger-sized units. For the largest unit designs, if site location, site conditions, or other factors limit the use of very large cranes, an alternate erection approach must be identified.

Project Schedule Larger unit sizes may decrease overall implementation schedules for larger power projects. For example, the labor-hours required to construct two 500 MW units will generally be greater than that required for a single 1000 MW unit.

Cost of Redundancy and Reliability versus Replacement Power Along with the issue of grid stability when a larger unit is forced off-line, there is also an issue concerning replacement power availability and cost. Replacement power cost is obviously greater for loss of a very large unit. For this reason, it may be prudent to build higher reliability into larger units through redundancy and other design measures.

Technical Risk Larger unit designs present a slightly higher technical risk, mostly because the level of experience with the largest size units is relatively limited. Although there are no theoretical limits or obstacles in the scale-up of supercritical unit designs, history has shown tha t there are sometimes unforeseen design problems. Even with the relatively limited amount of experience with the largest unit sizes, the level of risk should be considered to be relatively low with designs up to 1050 MW, especially for fuels that have already been used at these sizes.

Steam Generator Design Issues and Trends Furnace Design

Combustion Considerations Pulverized Coal Units

In a properly designed furnace, the combustion process is substantially completed before the product gases and particulates enter the upper furnace pendant sections and the back pass of the steam generator. Regardless of whether the steam conditions c onditions are subcritical, supercritical, or ultra-supercritical, the properties of the fuel must be given primary consideration during the

4-6


design of the steam generator. Key coal properties include moisture content and the amount and properties of non-combustible mineral content (ash). The furnace dimensions must accommodate the mass and volume flow of the combustion products by providing sufficient gas velocity to carry the particulates while also providing sufficient radiant heat transfer time for the ash to cool and solidify before it contacts the superheater tubing. For lower rank coals and higher slagging potential coals, furnace size must be increased, in some cases substantially. In the case of U.S. Northern Plains lignite, for example, the higher ash quantity, lower ash melting temperature, and higher mass flow (from the high moisture content) result in a very large furnace volume for a given megawatt output. Compared with an eastern U.S. bituminous coal, Northern Plains lignite requires nearly 2.4 times the overall furnace volume. Many of the problems experienced with earlier generations of supercritical units in the United States resulted from a poor understanding of the relationships between coal quality, ash characteristics, and furnace size. As a result of being undersized, many furnaces experienced high slagging, high corrosion, and high NOX production. Some of these conditions were aggravated by the higher surface temperatures of supercritical furnace walls compared with subcritical boilers. Today these relationships are much better understood and the problems are generally avoided in both subcritical and supercritical designs. Successful designs include adequate sizing of the furnaces, spacing of the burners, and spacing of radiant and convective con vective heat transfer surfaces at the furnace exit. Cyclone Burner Units

Another solution to furnace slagging and corrosion, convective pass fouling, and convective pass erosion is to remove the ash in the form of molten slag at the bottom bo ttom of the furnace. The cyclone cy clone furnace boiler design was specifically developed in the 1940s to combust coals co als with low ash fusion temperatures. In cyclone-fired boilers, the coal burner discharges tangentially into a cylindrical combustion chamber (typically 7–10 ft or 2–3 m in diameter) that acts as a cyclone c yclone separator. The coal is crushed (<0.25 inch or <6 mm) rather than pulverized (typically 70% passing 200 mesh), and is fired at very high heat release rates to liquefy the ash. Typical heat release rates are in the range 2 of 450,000–800,000 Btu/hr per square foot of cyclone barrel surface area (1400–2500 W/m ), as 2 opposed to pulverized coal boiler burner zone heat release rates of 100,000–400,000 Btu/hr-ft 2 (300–1200 W/m ). About 60–70% of the ash a sh is removed as liquid slag through the floor of the furnace. Because of the high heat release rates utilized in cyclone boilers, they produce high NOX levels, typically 1.2–2.5 pound per million Btu of input fuel heating value (lb/MBtu), or 0.52–1.1 kg/GJ, depending upon fuel type and boiler heat release rate. With U.S. passage of New Source Performance Standards for utility boilers in the 1970 Clean Air Act, fabrication of cyclone boilers discontinued. However, significant progress has been mad e to operate these boilers at or below the 1990 Clean Air Act Amendments Title IV NOX emission level of 0.86 lb/MBtu (0.37 kg/GJ). In some cases, NOX has been reduced to levels comparable to uncontrolled pulverized coal boilers (i.e., 0.40 lb/MBtu or 0.17 kg/GJ).

4-7


Fuel Delivery: Cyclone Delivery: Cyclone boilers utilize a crushed fuel rather than pulverized fuel; thus there is no need for a pulverizer with its associated costs, maintenance, and safety issues. ea sily burn Fuel Flexibility: Because Flexibility: Because of the combustion intensity within a cyclone barrel, it can easily lower quality coals (high ash, low fusion temperature), as well as a variety of specialty fuels (biomass, refuse, chars, petroleum coke, etc.). Cyclone boilers liquefy most of the ash, which is then removed as slag in the Ash Partitioning: Partitioning: Cyclone furnace bottom. This reduces the amount of ash material passing through the convection section and the particulate control device and exiting the stack. Corrosion: There Corrosion: There does not appear appea r to be a significant difference in corrosion potential in cycloneequipped boilers compared with pulverized coal boilers. In fact, corrosion in the furnace of a cyclone boiler may actually be reduced because of the removal of about 60–70% of the coal ash in the slag tap alleviates the vast majority of deposition/corrosion species. Reliability: Reliability: Progress is being made to improve cyclone barrels, which in turn, will increase cyclone furnace availability and should render the overall reliability of these furnaces comparable to that of pulverized coal units. NO X : Research conducted by EPRI was able to substantiate the performance and cost effectiveness of overfire air (OFA) operation in cyclone boilers. In all instances, cyclone boiler NOX emissions have been reduced to below the 1990 Clean Air Act Amendments Title IV NOX limits. Currently, 70% of the cyclone boiler capacity has installed overfire air, with NO X emissions reduced to a range of 0.40–0.85 lb/MBtu (0.17–0.37 kg/GJ). Particulate: Because baghouses and electrostatic precipitators are essentially constant efficiency devices, the less material that enters the device, the lower the total mass of particulate exiting. Because cyclone boilers remove about 60–70% of the ash as liquid slag in the bottom of the furnace, there is less material entering the particulate pa rticulate control device compared with pulverizedcoal boilers. Mercury: Utilizing Mercury: Utilizing the unburned carbon from combustion systems is a method to remove mercury from flue gas. Advanced cyclone furnace designs may be able to strategically optimize the size distribution of the crushed coal for to balance competing demands for mercury level reduction and maximum carbon-in-ash in a manner more effective than that possible with advanced pulverized coal units. Cost: Preliminary Cost: Preliminary studies indicate that a cyclone boiler with SCR can compete with a PC unit designed to meet the same environmental regulations.

Furnace Waterwalls Furnace walls, often referred to as waterwalls, are usually the first heat absorbing circuit following the economizer in the water/steam path of the steam generator. In a drum-type subcritical pressure unit, the waterwalls operate at a fairly constant an d relatively low temperature (e.g., <700°F, or 370°C) because beca use they contain boiling water at the operating pressure of the unit. Also, the boiling action generally assures that the internal temperature cannot rise above the boiling point, so good heat transfer is maintained from the tube surface to the internal fluid, and metal temperatures toward the exterior of the tubes remain relatively low. This allows carbon steel to be used for the most part.

4-8


At supercritical pressures, however, the fluid is not boiling. With the once-through design, tube temperatures continuously increase as the supercritical H2O proceeds up the furnace tube-walls. Steam temperatures at the exit of the furnace wall tubes of a supercritical once-through unit are in the range of 790–815°F. Average metal temperatures, on the tube exterior at the top of the furnace, approach 1000°F. In addition, there can be significant variations in the temperatures of these tubes due to steam side flow unbalances and fireside absorption imbalances. These temperature variations present mechanical design problems for all of the various support structures, openings, and attachments to the waterwalls. Differential thermal expansion is a key consideration and the resulting movement of the walls must be accounted for. The higher waterwall temperatures for supercritical furnaces typically require the use of alloy tubing in the furnace walls (e.g., T2, T11, T12, T13, T22, or T23). Experience has shown that although T2 and T11 materials have allowable stresses at temperatures up to 1000°F (538°C), they do not fare well in the high-temperature corrosive environment of the furnace, even with lower sulfur fuels. Despite corrosion resistance, one of the reasons that austenitic steels are prohibited from use in the waterwalls is because their coefficient of expansion is 30% higher than that for ferritic materials. Inconel materials have a coe fficient of expansion closer to that of ferritic materials. Thus, protective weld overlay coatings often have compositions mimicking Inconel materials. Fixed Pressure versus Variable Pressure

Most of the development efforts for once-through supercritical units over the past 20 years has occurred in the European and Japanese markets. For these markets, in which cycling and reduced load operation account for a considerable portion of their operating time, a variable pressure cycle has been found to improve heat rate at low loads, increase allowable ramp rates, and reduce stresses experienced by the steam turbine (and thereby extend the life of turbine components). Because the fixed-pressure, multi-pass furnace designs typical of older, baseloaded supercritical units in the United States are not compatible with the two-phase flow regimes experienced in variable pressure operation, a single pass of furnace circuits was developed. However, the single pass once-through design results in very low mass flow rates at reduced loads. For variable pressure units, which operate below the critical pressure at loads below about 70% of the maximum continuous rating (MCR), this means there can be departure from nucleate boiling (DNB) as load is reduced, which can result in localized tube overheating where the furnace circuits dry out. These issues have led designers of modern units to use a spiral arrangement of tubes in the lower part of the furnace to reduce the number of tubes carrying fluid and thereby increase mass flow rate. Some manufacturers also have a design that uses straight vertical tube orientation throughout the furnace and relies on either complex circuit orificing and/or specially rifled tubes to maintain sufficient turbulence and/or mass flow in the tubes to avoid overheating. At the lowest loads, usually below 30–40% MCR, these units revert to a recirculation mode in which the furnace circuitry functions much like a drum boiler. Spiral Tube versus Vertical Tube Furnace

Both spiral wound and vertical water walls have been used successfully in supercritical steam generators. In Japan and Europe, plants are often designed for sliding pressure control in which

4-9


the main steam passes through the critical point as load decreases. This enhances unit heat rate by eliminating throttling losses as it is no longer necessary to throttle steam flow to reduce load. In the spiral-wound design, flow from the inlet headers is directed to tubes that spiral at about a 20 degree slope and form the furnace enclosure. The spiral goes to an elevation just below the furnace arch where the flow collects in a header. From here the flow is redistributed into vertical tubes that form the upper furnace enclosure. This arrangement provides uniform temperatures at the fluid outlet as the tubes spiral around all four walls evening out heat flux variations in the lower furnace where the coal is introduced. There are only limited flux variations in the upper furnace. In the vertical tube design, flow from the inlet headers is directed to vertical tubes that travel the full height of the furnace. Orifices are located at the inlet of each tube to ensure even flow distribution and to even out the effect of in-furnace heat flux variations. Overall, it can be concluded that both spiral wall and vertical wall arrangements are technically sound approaches for variable pressure supercritical units. Either can be utilized for an advanced cycle supercritical unit with nominal steam temperatures of about 1100°F (600°C). Owners should evaluate differences in erection, maintenance, and operating costs, as well as the experience levels for the two designs. Baseload versus Cycling Mode of Operation

Although variable pressure does not necessarily provide advantages for baseload applications, virtually all design development for large supercritical PC units in the United States over the past 20 years has focused on the variable pressure cycle. Thus, it represents the state-of-the-art technology and will almost assuredly be the design that current suppliers will propose. However, it should be noted clearly in the steam generator specifications whether the unit is intended primarily for baseload operation or whether load cycling or on-off cycling will be necessary. Significant savings in full-load operating cost can be realized by designing the furnace circuits for lower mass velocity and therefore lower pressure drop. The trade-off is that a higher minimum recirculation point will be necessary, thereby raising operating costs at reduced loads, but this will be of little consequence if full load or near full load is the primary operating objective. Costs can also be saved in startup equipment if baseload duty is the primary objective. Instead of using separate recirculation pumps for low-load ope ration, the design can use the main boiler feedwater pumps to accomplish recirculation at lower loads. The disadvantage to this approach occurs only during recirculation mode, when heat from the furnace circuits is lost to the condenser rather than conserved within the circuit. Sketches illustrating the two recirculation options are provided in Figure 4-2 and Figure 4-3.

4-10


Superheater

To Steam Turbine

Separator

Furnace Circuits

Economizer Recirculation Pump

HP Heater

Figure 4-2 Furnace Circuit Recirculation with Separate Recirculation Pump

4-11

Boiler Feed Pump


Superheater

To Steam Turbine

Separator Furnace Circuits Flash Tank Condenser

Economizer

HP Heater Boiler Feed Pump

Figure 4-3 Furnace Circuit Recirculation without Separate Recirculation Pump

Power producers should evaluate the anticipated likelihood of baseload versus cycling operation for different periods during the lifetime of the new unit. Operating scenarios may change as a result of future additions of large blocks of baseload power (e.g., nuclear) or unpredictable power (e.g., wind power). The recirculation system can be changed at relatively moderate retrofit cost whereas the full-load furnace circuit mass flow rates will be set by tube sizing and pitch of the spiral walls. Once the furnace is designed and built, it is not practical to change these parameters to accommodate a different mode of operation. If future modes of operation are uncertain, a compromise design approach could be to specify a moderately high minimum recirculation point, say 40–50% MCR, and use the main boiler feedwater pumps for recirculation during initial operations. If future duty requirements change to include more cycling, the recirculation system could be modified to include separate pumps and be operated more frequently (i.e., every time load drops below 40–50%). Duty cycle modes that may need consideration include: •

Moderate daily and weekly cycling

•

Daily cycling to minimum load

•

Daily on/off cycling

•

Deep load cycling

4-12


Along with the mechanical and hydraulic design features to optimize control and heat rate during cycling, design for cycling modes must also consider: •

•

•

The number and type of cycles and their influence on fatigue, corrosion, corrosion fatigue, creep fatigue, and other factors that impact component life The physical configuration of piping, headers, tubing, pipe supports, and other components that may introduce stress concentrations or lead to condensate accumulation and thermalshock-inducing flow events The impact of control modes, such as attemperator and bypass operation, on these damage mechanisms

Startup Systems and Turbine Bypass Considerations The traditional use for turbine bypass systems has been for matching steam and turbine metal temperatures during startup; to keep the boiler on-line during load rejection; and for cooling the reheater when necessary in certain types of steam generator designs. In Europe, bypass systems have also been used to eliminate safety valves. With modern steam generator design, the bypass system should generally be large enough to be able to synchronize the turbine generator and pick up minimum load steam flow. The nominal size of the bypass system is established as a percentage of full load main steam flow at full load main steam pressure and temperature. Additional capacity beyond that needed for minimum turbine flow may be necessary to meet the temperature characteristics of the steam generator. The maximum bypass capacity must be limited to prevent overpressure of the condenser. When the turbine is tripped, a condenser, sized for full load LP turbine exhaust, can accept approximately 50% of the design boiler heat load from the bypass system, with an acceptable temporary rise, in circulating water temperature and con denser pressure. Accepting this rise to have somewhat higher values, avoids the need for additional condenser surface area. Startup system variations include: •

Partial flow turbine bypass

•

100% flow turbine bypass

•

Total bypass systems

•

IP and LP turbine bypass or LP turbine (only) bypass

•

Steam generator superheater bypass

Designing Supercritical Steam Generators for Low Minimum Load Capability and Continuous Duty Minimum Load Cycling To achieve very low minimum continuous loads, several steam-generator-related systems must be addressed. These systems include the boiler itself, pulverizers and burners, draft fans, air heaters, flue gas desulfurization equipment, particulate removal equipment, boiler feedwater pumps, and the turbine-generator. Consideration must be given to: •

Extended extreme low-load supercritical plant operation

4-13


•

Minimum load firing considerations

•

Daily start stop service (dss)

•

Weekend start stop service (wss)

•

Loading rate

•

Total number of cycles

•

Turbines and feedwater heaters for cycling service

Design Provisions for Higher Peak Power Rating The economics of contemporary electric system operation often make it desirable to include capability for higher power output during periods of peak demand. With supercritical generating units, this can be accomplished to some extent by: •

Increasing firing rate to maximize steam production and turbine mass flow

•

Operating with feedwater heaters out of service

•

Maximizing cooling tower airflow and condenser water flow

These measures in effect entail acceptance of temporary heat rate penalties in exchange for added output. They also may increase stresses and temperatures that accelerate creep and fatigue damage mechanisms.

4-14


5

ISSUES RELATED TO FUEL QUALITY The principal feedstock for a pulverized coal power plant is typically bituminous coal, subbituminous coal, or lignite (also known as “brown coal” outside of North America). Firing systems typically are designed to use oil and gas for startup and may allow full-load operation on these fuels. Some plants blend significant fractions of petroleum coke (a low-volatile, high-rank solid fuel that is a by-product of petroleum refining), mining waste fuels (e.g., anthracite culm and bituminous gob), or biomass fuels with the feed coal. This Guideline does not address wastetype alternative fuels as they are generally better utilized in circulating fluidized-bed units. The fuels used in an advanced PC unit can have a significant impact on materials selections, steam generator design, emissions control system design, and cost. Therefore, care must be taken to understand the properties of the fuel(s) that the plant will use and how the composition of fuel(s) might vary over the life of the plant. To maintain consistent fuel properties and costs, some plant owners enter long-term contracts with fuel providers. Such contracts designate specific coa l mines and/or detailed coal quality specifications to assure fairly consistent throughout the life of the contract. Other plant owners purchase coal on the spot market and are likely to obtain coals with a much wider range of properties. Even the generating units which are sourced from a single supplier should expect variations in composition, trace elements, and ash characteristics as different regions of the coal reserves are mined.

Coal Rank In the United States, coal is classified by rank using a system established by the American Society for Testing and Materials (ASTM D 388). Coal rank generally correlates with the age of the coal, with coal class ranging from oldest to youngest in the following order: •

Class I: Anthracitic

•

Class II: Bituminous

•

Class III: Subbituminous (including Powder River Basin coals)

•

Class IV: Lignitic (brown coal)

In general, the heating value of coal decreases with decreasing rank (increasing Class number), while moisture and ash content increase. Fixed carbon and volatile matter form the primary basis for differentiating between higher rank coals whereas Btu analysis (on a moist, mineral free basis) sets the criteria for lower rank coals. Although much of the coal mined in a limited geographical region tends to have the same rank, coals of a given rank can exhibit significant variation in constituents from mine to mine and even within a given seam of a mine. Historically, sulfur content has been a parameter of particular interest.

5-1


Although the coal rank to be used for a generating unit provides a rough basis for general design parameters and economic performance forecasts, a much more detailed analysis of the specific fuel(s) to be used is required to establish criteria for boiler design details.

Coal Analysis In specifying a steam generator, it is important to provide a complete, individual analysis for each of the various coals that may be fired. It is not sufficient to simply state a range for each of the constituents. Slagging, fouling, and corrosion potential are partly determined by the interrelation between individual constituents of a coal. Specifying only the range of each of the constituents may lead to over-design. For example, to size pulverizers, the designer may look at the maximum moisture content in the range along with the minimum grindability and heating values in those respective ranges. However, it may be very unlikely that the maximums and minimums of concern will occur together to create a worst-case scenario. For the most reliable and cost-effective steam generator design, the following complete analysis is recommended for each coal considered for the project: •

Source Information – Region – Mine – Seam

•

Proximate analysis (moisture, volatile matter, fixed carbon, and ash)

•

Ultimate analysis (moisture, carbon, hydrogen, sulfur, nitrogen, oxygen, and ash)

•

Heating value

•

Ash fusion temperatures, in both oxidizing and reducing atmospheres, including initial deformation temperature (IDT), hemispherical deformation temperature (HDT), spherical deformation temperature (SDT), and fluid deformation temperature (FDT)

•

Chlorine content

•

Ash chemistry analysis

•

Forms of sulfur (pyritic, organic, sulfate)

The heating value of the fuel determines the quantity of fuel needed, and therefore influences the size of fuel handling equipment and the quantity of ash, sulfur, and other non-beneficial constituents. Ultimate analysis and proximate analysis help reveal the coal combustion characteristics. Along with ash chemistry analysis, they provide information on the different ash constituents that influence the ash properties. The ash content derived from mineral matter in the fuel is generally one of the most important factors in the overall design of the boiler. The ash fusion temperature is also a crucial piece of information because the slagging characteristic of the ash is typically the most significant property. High ash fusion temperatures allow the ash to freeze more quickly once the fuel is burned and the resulting heat is radiated from the product gases the waterwalls. Low ash fusion temperatures require design modifications to limit potential problems with both slagging (accumulation of molten ash) and fouling

5-2


(accumulation of material that is not molten, although molten material may have initiated the sticking). The buildup of these materials changes the heat absorption properties of different tube surfaces and, in the worst case, can lead to “slag falls” that cause impact damage at the bottom of the furnace. Design considerations include maintaining oxidizing conditions, which increases melting temperatures by about 200°F (110°C), keeping molten ash away from the walls, and increasing boiler size to provide residence time for the ash to freeze before it comes in contact with tubing. Chlorides can cause corrosion problems throughout the back end of the unit (including in the ESP and SO2 scrubber). Along with various sulfur compounds, they combine with ash to cause corrosion problems on high temperature surfaces, particularly when reduc ing zones are created beneath slag or accumulations of ash.

Grindability The Hardgrove Grindability Index is calculated from the particle size reduction of a coal sample in a vertical shaft laboratory grinding mill. This index provides a measure of grindability in comparison with coals used as standards. Its applicability is limited, however, if a designer is considering a blend of coals of greatly different rank. For example, when admixing low-rank, high-moisture coal with a bituminous coal, the feed rate would more likely be limited by the 11 drying capacity of the existing mill rather than by its grinding capacity.

Ignition and Flame Stability Generally, coals with higher volatile matter content ignite mo re easily and provide better flame stability. Additionally, the devolatilized char is of higher reactivity, with favorable characteristics for combustion completion. Proximate analysis of volatile matter content, however, is a poor indicator of furnace performance, partly because the volatile yield at flame temperature is significantly higher than at the ASTM test temperature, and because volatiles of higher rank coals consist mainly of combustible hydrocarbons. In contrast, lower rank coals contain large amounts of moisture and oxygen in their volatiles. Also, decomposition of carbonates in high-ash coals produces CO2 that appears as a volatile in the analysis. To properly assess the combustion behavior of a blend, it is necessary to know the volatile yield and composition, for the respective coals, at flame temperature. The relative ignition behavior of different coals can be determined using Differential Thermal 12 Gravimetric Analysis (DTGA). DTGA provides a continuous plot of the rate of weight loss of a pulverized coal sample as the furnace temperature is gradually raised over time. From the graphical plot called burning profile, the ignition temperature of the volatiles and of the partially devolatilized char can be derived.

11

J.R. Gunderson, S.J. Selle, and N.S. Harding, Engineering Foundation Conference on coal blending and switching of low sulfur western coals, Snowbird, UT, ASME (1994), pp. 11-130. 12 A.K. Chambers et al., Effect of blending on the combustion properties of dissimilar Alberta coals, (Report [Alberta Research Council] no. CHP 89/17), prepared for Alberta Office of Coal Research and Technology (1988), p. 92.

5-3


Unburned Carbon The volatiles and the high reactivity char of a low-rank coal readily ignite and burn in the early flame, which also increases the yield of volatiles from the high rank component. The combustion of these volatiles and the burning of the low-rank char consume their share of the available oxygen and leave the burning of the low reactivity bituminous char in a localized atmosphere of 13 depleted O2 concentration, resulting in increased unburned carbon. A combination of coal blending and staged combustion to reduce NOX formation can accentuate this effect because of the deactivation of the devolatilized char in the high-temperature, fuel-rich atmosphere of the first combustion stage. But there may be a compensating factor: because of the increased volatile yield from the high-rank coal, and hence increased conversion of the fuel-N to N2, it is possible to reduce the depth of combustion staging and to produce a char with higher reactivity.

Emissions

Sulfur Oxides (SO2 + SO3) One reason for coal blending is to reduce SO2 emissions. The sulfur content in coal is additive and hence the sulfur input can be calculated as the weighted average of the sulfur content of the coals blended. This, however, can overestimate emissions, because the high concentrations of 14 organically bound calcium in western U.S. coals favor sulfur capture. Also, a small portion of the SO2 forms SO3, part of which is reabsorbed in boiler deposits and in fly ash. Although coal blending was considered a less expensive alternative to installing FGD (in initially complying with acid rain regulations), it is questionable whether it can now, on its own, produce compliance with increasingly stringent SO2 emissions restrictions. It should, however, remain a valuable tool for incrementally reducing sulfur emissions from high-sulfur bituminous coals.

Oxides of Nitrogen (NO + NO 2) Nitrogen oxides form in reactions of molecular nitrogen (N 2) with O atoms in high-temperature oxidizing atmospheres (thermal NO); with hydrocarbon radicals in the early flame (prompt NO); and by oxidation of organically bound nitrogen in the coal (Fuel-N). In modern pulverized coal combustion, about 80% of the NO forms from Fuel-N. In unstaged combustion, Fuel-N evolved with coal volatiles readily oxidizes to NO X. In the hot portion of a fuel-rich (i.e., oxygen-depleted) flame, however, it can be converted to N2 and hence rendered innocuous for NO formation. This is the basis of staged combustion as a method of reducing in-furnace NOX production. It is also known that the nitrogen in coal volatiles converts to N2 with higher efficiency than the nitrogen that remains in the devolatilized char. So, for staged combustion, the steep rise in temperature, following the early ignition of the low-rank coal, increases the volatile yield of the bituminous coal, thereby improving the efficiency of NOX reduction. Although NOX emissions increase with the fraction of high-volatile coal in the blend during unstaged combustion, this trend is reversed for staged combustion, where NOX emission is reduced as the fraction of low-rank coal is increased. 13

J.P. Smart and T. Nakamura, Journal of the Institute of Energy, 66 (467); (1993) pp. 99-105. S.G. Kang et al., Engineering Foundation Conference on coal blending and switching of low sulfur western coals, Snowbird, UT, ASME, (1994), pp. 341-354. 14

5-4


Mercury For mercury emissions control, unburned carbon in fly ash can partially or fully replace activated carbon injection. In particular, mercury can be chemisorbed on the surfaces of unburned carbon in fly ash, as with activated carbon, and then captured in a downstream particulate control device. By readjusting staged combustion conditions, the amount of unburned carbon can be 15 raised to achieve increased mercury reduction. “Oxidized” mercury (i.e., at a +2 valence state, even though the actual compound may be HgCl2, HgS, or another mercuric compound, in addition to HgO) can also be captured in a wet FGD system. Increased halogen concentration is one key to the “oxidation” of mercury. Although the mercury content of western U.S. subbituminous coals is about half that of eastern U.S. bituminous coals, these coals also have low chloride content. By blending bituminous coals with subbituminous coals, chloride content is boosted and mercury capture is improved.

Particulates ESP performance is strongly affected by the electrical resistivity of fly ash. Fly ash with high 11 resistivity (>10 ohm/cm) limits the current that can pass through the collected ash layer, 8 whereas very low resistivity ash (<10 ohm/cm) can lose electric charge, get separated from the 16 plate, and become re-entrained by the gas flow. In general, eastern U.S. bituminous coals have lower resistivity ash and western U.S. subbituminous coals have higher resistivity ash. The amount of SO3 absorbed by the fly ash has a strong influence on its resistivity. Some operators of units firing low-sulfur coals have adde d SO3 to flue gas to lower resistivity and boost ESP collection efficiency. With respect to fly ash chemical composition, a substantial reduction in Fe2O3, K2O, and Na2O, or enrichment in CaO, MgO, and SiO2 will increase resistivity. High carbon in ash can reduce resistivity to the extent that it makes collection in an ESP difficult. Conversely, re-entrained carbon reacts preferentially with SO3, degrading the benefits of conditioning by SO3 injection. Experience in power stations with low-sulfur, low-rank coals show that the increased resistivity, finer particle size of the fly ash, and reduced residence time of the of flue gas in the ESP due to the increased water vapor, can increase emissions. High rapping forces and frequent rapping are generally needed to maintain ESP performance with high resistivity fly ashes. Fabric filter performance is less affected when using a lower sulfur content coal blend.

Ash Properties and Deposition Behavior A variety of indices are used to estimate the slagging and fouling characteristics of coal ash. These indices have limited use for coal blends because deposition characteristics cannot be reliably predicted by interpolation between those of the component coals. There are significant synergistic effects between the blended coals. For example, in the common case of a blend of low-rank western U.S. coal and eastern U.S. bituminous coal, fly ash of the low-rank coal with a 15

A. Capri, Journal of Water, Air & Soil Pollution, Volume 98, Nos. 3-4, (2004) pp. 241-254. P.P. Bibbo, Engineering Foundation Conference on coal blending and switching of low sulfur western coals, Snowbird, UT, ASME, (1994), pp. 301-321. 16

5-5


low softening temperature enters the convective pass at elevated temperature, due to the higher combustion temperature of the bituminous coal. This would result in more severe fouling of the superheaters. The high concentrations of organically bound Ca, Na, and Mg in low-rank coal ash form reactive fluxing agents upon combustion that reduce the viscosity of liquid phases and produce sticky layers on furnace walls and superheater surfaces. These sticky tube surfaces capture impacting 17 particles with high efficiency, leading to enhanced slagging and fouling. Predictive indices and models in the technical literature can give guidance about the slagging and fouling behavior of coal blends. A combination of Computer Controlled Scanning Electron Microscopy (CCSEM) and mathematical modeling is capable of predicting the fouling tendency 18 of a blend relative to that of a design coal. Experience in power plants shows that by keeping the furnace walls clean of the highly reflective slag formed from the low-rank coal ash, increase in furnace exit gas temperature can be avoided and the fouling of the superheaters reduced. Superheater fouling is managed by increased frequency of sootblowing, but boilers with finned economizers may have fouling problems.

Coal Blending Blending of different coal types is a popular method to control emissions, save fuel costs, or improve the combustion behavior of the “design” coal. As noted above, coal blending may also have significant positive or negative effects on boiler operation: •

Coal preparation: drying and grinding of coal

•

Combustion properties: coal devolatilization, ignition, flame stability, carbon burn out

•

Formation and emission of pollutants: NOX, SOX, Hg, PM2.5

•

Ash deposition: slagging of furnace walls; fouling in the convective pass of the boiler

•

Ash quality: carbon and alkali content of the fly ash

•

ESP performance

Plant experience highlights the need for extra attention to operating practices if the advantages of coal blending are to be realized. This care extends from controlling coal blend drying and grinding to specification in the mills, to maintaining a homogeneous feed to the burners, and to keeping furnace walls and convective heat exchange surfaces clean by increased frequency sootblowing.

Blend Impact on Coal Grinding Coal blending can be carried out at different points along the transport chain: at coal mines, at coal cleaning plants, or at power plants. At the power plant, blending can take place in stock piles, bins, bunkers or silos, on belt conveyors, or downstream of separate grinding mills.

17

T.F. Wall et al., Engineering Foundation Conference on coal blending and switching of low sulfur western coals, Snowbird, UT, ASME, (1994) pp. 453-463. 18 J.M. Beér, A.F.Sarofim, and L.E. Barta, Journal of Institute of Energy, 65 (462); (1992) pp. 655-62.

5-6


Although the choice of blending method is situation- or site-specific, all methods require care to ensure blending accuracy. This is especially important because blend evaluation by standard analytical tests is complicated by the non-additive nature of some coal properties. For optimal choice of coal type and blend fractions, standard analytical methods should be supplemented with more sophisticated analytical techniques such as CCSEM, computational modeling, pilot plant experiments, and a review of local conditions of coal drying and grinding. Ultimately, tests using the actual coal blend and mill may be required to determine feasibility and optimal mill settings. Coal blending options may be limited by safety measures and other considerations. For example, when increasing the blend fraction of lower rank, higher moisture coals, issues with dust or volatiles may limit the use of higher drying air temperature. At the same time, extra drying to enable finer grinding may not be necessary as the lower rank coal may perform adequately with coarser particle size than is necessary for bituminous co al.

Blend Impact on Combustion and Deposition Combustion and fouling are areas where important coal properties are non-additive for the prediction of properties of coal blends. For example, NO X formations may be reduced as a result of synergistic interactions between component properties. Others factors, such as unburned carbon in fly ash, slagging and fouling of furnace-and superheater tube surfaces, and ESP performance, are prone to becoming worse than weighted mean values of coal properties would indicate. Experience with coal blending in PC plants suggests that the addition of 25–30% low-rank coal to bituminous coal can improve burner performance (not withstanding a necessary reduction in mill exit temperature). Homogeneity of the blend is important; “slugs” of low-rank coal reaching the burner produce rapid changes in flame length and can cause flame failure when the low-rank coal fraction drops, potentially leading to damaging pressure pulses in the furnace as another slug appears.

Blend Impact on Emissions An important driver for coal blending may be reduction of air emissions, most notably SO2, but potentially also mercury, NOX, and particulate matter. As mercury control gains importance, synergistic interactions between constituents in the different coals becomes an important consideration for blend evaluation, along with reassessment of the desirable level of unburned carbon in fly ash. Chloride content in one coal may help with reduction of emissions resulting from high mercury levels in another. Blend and grind specifications may increase or decrease the amount of mercury that is absorbed by carbon in the fly ash versus activated carbon injected for mercury control. Along with fine particulate matter, the ultimate destination of adsorbed mercury is determined by ESP performance, which is influenced by ash resistivity that in turn is a function of carbon-inash and other constituents derived from the coal blend. The quantities and species of fuel NOX and SOX produced in the furnace are determined by both content of these species in the fuel blend and the combustion properties of the blend. Ultimate emissions of these pollutants are further influenced by chloride content, which can increase

5-7


oxidation of SO2 to SO3 in the SCR catalyst, and other constituents that can degrade SCR catalyst performance.

5-8


6

IMPROVING PLANT EFFICIENCY WITH ADVANCED STEAM CONDITIONS The efficiency of a Rankine-cycle generating unit is a strong function of steam temperature and pressure at the turbine inlets. Increasing the maximum operating pressure and/or the maximum operating temperature of a supercritical steam generator can provide significant improvement in the net efficiency of the unit. Being able to increase steam temperature is especially important; an incremental boost in either final superheat or reheat temperature provides almost five times the cycle efficiency improvement of an equal percentage boost in main steam pressure. Cost-effective implementation of higher steam conditions requires an increased standard of care by the designers, fabricators, and operators of the steam generator, steam turbine, and associated balance of plant equipment. The materials selection, component geometry, fabrication techniques, component integration, control system design, and operating practices all contribute to achieving required standards for capital and operating costs, reliability, availability, and maintainability. This chapter primarily addresses material properties and aging mechanisms for advanced ferritic steel alloys for use in ultra-supercritical steam generators with nominal 1100°F (600°C) final steam temperatures. Steam generator materials of construction must be selected carefully and installed correctly. The worldwide experience base is still fairly limited for alloys which are suitable for use at ultra-supercritical steam conditions. For some of these materials, experience has shown that rigorous welding practices are critical for ach ieving required resistance to aging mechanisms such as creep, fatigue, and corrosion. For advanced USC materials, such as nickelbase alloys, the cost of the materials themselves can be a high proportion of capital cost, and availability of materials must be considered when planning construction schedules.

Designing for High Steam Pressure: >3750 psi (>260 bar) In general, to increase plant efficiency, increasing steam pressure is not as effective as increasing steam temperature. As pressure is raised, component stresses rise proportionately, requiring either thicker walls for tubing and headers or a switch to a higher strength steel. Both of these options add cost to the steam generator. Further costs are inherent in the increased structural strength requirements for the heavier components. Adding thickness to components can also increase thermal fatigue and maximum metal temperature at the tube OD. Nonetheless, if a very corrosive coal is to be used, it may be more economical to raise steam pressure rather than steam temperatures. This can avoid the considerable expense incurred when it is necessary to use exotic alloys in the superheaters and reheaters. For less aggressive coals, and where the highest efficiencies are sought, raising both steam pressure and temperature should be considered.

6-1


19

Designing for High Steam Temperatures: 1050–1150°F (565–620°C)

Development of stronger and more corrosion resistant high-temperature steels constitutes the main enabling technology needed to allow plant designers to employ higher steam parameters. This is particularly true for critical pressure parts of the steam generator such as piping, headers, and superheater tubes. These applications call for not only creep strength but also for resistance to fireside and steamside corrosion and ease of welding, fabrication, and coating. Even with these challenging, multi-faceted requirements, many new alloys have been developed that appear capable of operation under severe steam conditions. These include ferritic/martensitic steel alloys with 9-12% chrome, austenitic stainless steel alloys, and nickel-base alloys. These alloys have been developed through over two decades of worldwide activities focused on hightemperature material. Significant new information continues to be developed for the most prevalent of these high-strength alloys, the modified 9Cr-1Mo-V steel with the ASTM designations P91 and T91. T91 was first applied in fossil power boiler superheater tubing in the early 1980s. P91 was first used in fossil power plant headers in the late 1980s. Continuing development and study is aimed at better meeting the key requirements for critical components for ultra-supercritical plants as outlined below: •

•

•

High-pressure steam piping and headers – High creep strength – Thermal fatigue strength – Resistance to steamside oxidation and exfoliation – Weldability, including pre- and post-weld heat treat requirements Superheater/Reheater tubing – High creep strength – Thermal fatigue strength – Resistance to fireside corrosion – Resistance to steamside oxidation and exfoliation – Weldability, including pre- and post-weld heat treat requirements Waterwall tubing – High creep strength – Thermal fatigue strength – Resistance to steamside corrosion and cracking – Resistance to fireside corrosion and cracking – Weldability, including pre- and post-weld heat treat requirements

High-energy steam piping carries high-pressure, high-temperature steam from the steam generator to the turbine. Like high-energy piping, waterwall (evaporator), superheater, and reheater headers are fabricated with large diameters and thick walls. They also have numerous

19

All temperatures cited in this section are steam temperatures unless otherwise specified. For headers and piping, metal temperature is nearly equal to the steam temperature. For final superheater tubing, the maximum mean metal temperature is generally up to 100°F (55°C) higher than the average bulk steam temperature.

6-2


penetrations where the heat exchange tubes are attached. The geometry, materials, and welding procedures used for these components must provide: •

•

Sufficient creep strength to allow them to operate a t high temperatures and pressures for their desired design life Resistance to fatigue from thermally and mechanically induc ed stresses. – Thermal fatigue is a particular concern for thick-walled piping and headers, which may experience high temperature differentials across the wall thickness caused by rapid steam temperature changes or condensate flow events during start-up, shutdown, load changes, and attemperator operation. – Calculations for creep life and fatigue life must also consider interactions between these two mechanisms (creep-fatigue). – Units that experience wet steam conditions with significant cycling operation or attemperator operation may also experience corrosion-fatigue.

Ferritic/martensitic steels are preferred because of their lower coefficient of thermal expansion and higher thermal conductivity compared with austenitic stainless steels. Many of the early problems in the U.S. supercritical plants were traceable to the use of austenitic steels prone to thermal fatigue. Research during the last two decades has, therefore, focused on developing costeffective, high-strength ferritic/martensitic and high-nickel steels that can be used in lieu of austenitic steels for higher temperature applications. This has resulted in ferritic steels capable of operating at metal temperatures up to 1150°F (620°C) with good weldability and fracture toughness while retaining a relatively low coefficient of thermal expansion. Superheater and reheater (SH/RH) tubing applications call for high creep strength, thermal fatigue strength, weldability, resistance to fireside corrosion, and resistance to steamside oxidation and oxide exfoliation. Along with cost considerations, thermal fatigue resistance would dictate the use of ferritic/martensitic steels. Unfortunately, the strongest of these steels, on the basis of creep rupture strength for metal temperature up to 1150°F (620°C), also experiences fireside corrosion that limits their use to maximum metal temperature of about 1100°F (593°C). This corresponds to an average steam temperature of about 1000°F (540°C), because maximum mean temperature in SH/RH tube walls can exceed the average bulk steam temperature by as much as 100°F (55°C). Excessive corrosion of ferritic steels, caused by liquid iron-alkali sulfates in the tube deposits, is an acute concern in the United States where high-sulfur, corrosive coals are used more frequently than elsewhere. As a result, high-strength ferritic/martensitic steels, such as T91, are infrequently used for final superheater/reheater applications in the United States. In conventional pulverized coal plants, the standard practice is to use T22 for the lower temperatures and SS304H or SS347 for the higher temperatures. For waterwall tubing, the concern is twofold: •

The first concern is metal temperature—supercritical pressures result in a monophase fluid which increases in temperature continuously up the height of the furnace. The supercritical fluid does not provide the superior cooling effect and even temperature distribution seen with nucleate boiling in the waterwall tubes of subcritical drum boilers. The higher temperatures, especially at higher elevations, require use of materials with higher creep strength.

6-3


•

The second concern is corrosion that results from the combination of high tube crown temperatures, reducing conditions at the wall surfaces, and higher sulfur/iron coals. Furnace wall tubes with this exposure can experience excessive wastage rates and cracking. Reducing or eliminating excessive corrosion requires use of weldable, high-strength alloys with high 20,21 chrome alloy cladding or overlays.

Steam Generator Components

Furnace Wall Corrosion Waterwall tubing in once-through supercritical boilers is subject to significant potential for corrosion and for cracking associated with corrosion bec ause of its exposure to high temperatures and to coal ash and combustion chemistry. This vulnerability is seen with both spiral and vertical tube arrangements. An accumulation of research, spanning over 60 years, has identified numerous mechanisms and contributors to corrosion and cracking of waterwall tubing. Some of these mechanisms are applicable to all types of boilers whereas a few are nearly exclusive to supercritical waterwalls. Not all supercritical furnace walls experience corrosion. For those that do, it has generally been observed that high rates of corrosion are influenced by one or more of the following factors: •

Coal and ash chemistry

•

Presence of a reducing or oxidizing atmosphere at the tube surface

•

High temperature at the tube surface

•

Tube metal composition

•

Unit loading profile, especially frequent on-off or deep load cycling

•

Incorrect or poorly controlled feedwater chemistry

Corrosion mechanisms involve some form of damage to the oxide layer of furnace wall tubing, which will generally protect the base metal as long as it is stable. The oxide layer will remain stable as long as an oxidizing atmosphere is maintained at the tube surface. The oxide layer will be lost if reducing conditions are present because the metal oxide will be reduced to elemental iron and oxygen (in a manner similar to a blast furnace). These protective oxide products are lost from the tube and carried away with the ash and flue gases. As oxidizing conditions are again restored, the oxide layer is regenerated, at the expense of the base metal. The most severe form of wastage occurs with conditions that alternate frequently between oxidizing and reducing, which occurs mostly in boilers with low-NOX burners and frequent load cycling. This phenomenon leads to alternate destruction of the oxide layer and regeneration of the layer by further oxidation of the base metal. Significant wastage rates can occur even in the absence of other corrosive agents. More severe wastage occurs when this mechanism is combined with those agents. 20

R. Viswanathan, R. Purget, and U. Rao, “Materials Technology for Advanced Coal Power Plants,” 2003, Technology Review sponsored by USDOE and Ohio Coal Research Office. 21 R. Viswanathan and W.T. Bakker, “Materials for Ultra Supercritical Coal Power Plants, Boiler Materials: Part 1,” and “Turbine Materials: Part 2,” Journal of Materials Engg. and Performance, ASM, Vol. 10(1), (Feb 2001), pp. 81-100.

6-4


To meet ever tightening NOX emission requirements, manufacturers seek to reduce NOX from the boiler to the greatest degree possible to control SCR catalyst expense and ammonia consumption. This often leads to the introduction of deeply staged combustion systems, in which the air/fuel ratio is significantly less than 1.0, and additional combustion air is added above the burners via overfire air ports. Several boilers in the United States retrofitted with such systems have reported severe c orrosion of low alloy steel waterwalls, with metal losses in the 40–120 mil/yr (1–3 mm/yr) range. Supercritical units are generally more severely affected than subcritical units due to the higher tube surface temperatures. Severe corrosion is generally limited to coals with more than 1%S. However, above 1% sulfur there is no strict correlation between S and corrosion rate. The highest corrosion losses are found in regions where H2S rich substiochiometric flue gas mixes with air from the overfire air ports. Laboratory studies indicate that the high corrosion rates cannot be explained by the presence of H2S and CO in the flue gas alone. Work by Kung has shown that corrosion rates in gas mixtures actually found in boilers containing 500–1500 pm H2S and 5–10% CO are generally less 22 than 20 mils/yr (0.5 mm/yr) at 840°F (450°C). More recently it was shown that the presence of FeS deposits can greatly increase the corrosion rate, but only under alternating oxidizing/reducing conditions or oxidizing conditions alone. Sulfidation is another key mechanism for wastage of waterwalls. Sulfides of hydrogen, sodium, and potassium can form in furnace wall deposits, for example, which can form low melting point eutectics with the tube metal. These eutectics can be very corrosive above a critical temperature, even taking the form of intergranular crack formation. Sulfur can be more of a problem for waterwalls when reducing conditions are experienced. Although much more prevalent around the burner zone, with very deep and delayed fuel and air staging it is possible to get reducing conditions at the waterwalls even in the upper furnace. Sulfidation of the waterwalls in the presence of hydrogen sulfide is a key mechanism for wastage of waterwalls. Other mechanisms and influences include metal chloride attack resulting from HCl and chloride salts and corrosion fatigue at the steamside rear wall of waterwall tubing.

Furnace Wall Cracking The mechanism specifically identified as “supercritical waterwall cracking” presents with multiple, parallel cracks, which are often V-shaped and oxide coated, on the fireside tube surface 23 or membrane. The cracking and spalling of the tube surface is some times referred to as “alligator cracking,” “cross cracking,” or “elephant skin.” This mechanism may result from wall thinning with fireside corrosion, increased tube metal temperatures resulting from excessive internal deposits, and/or thermal cycling caused by slagging and deslagging. The phenomenon has been observed in many earlier U.S. supercritical steam generator designs. Two thermal fatigue mechanisms appear to be responsible for initiation of the cracks, which may then be accelerated by corrosion mechanisms. The first fatigue mechanism is caused by the preferential heating of only one side (furnace side) of the tubes. Because the tubes are

22

S.K. Kung, “Prediction of Corrosion Rate for Alloys Exposed to Reducing/Sulfidizing Combustion Gases,” Corrosion 97, NACE (1997), pp 97-136. 23 Boiler Condition Assessment Guideline, Fourth Edition. EPRI, Palo Alto, CA, 2006. 1010620.

6-5


constrained to remain straight, the one-sided heating induces a compressive stress on the fire side of the tube which can eventually lead to irreversible compressive strain (set). When the tube cools at lower loads and during outages, the surface is put in tension. Multiple occurrences of this cycle with frequent cycling of the unit or shut downs can lead to fatigue cracking. The second fatigue mechanism is the shedding of ash or slag from wall tubes. When an insulating layer of slag is shed or removed from a tube surface, a spike in the heat flux to the tube surface raises its temperature and restrained thermal expansion puts compressive stress on the material. As with the stresses developed by more generalized heating, this may result in a small amount of compressive set. As a new coating of slag forms, the heat flux gradually reduces, the temperature falls, and restrained thermal contraction places tension on the surface. These cycles may be repeated many times per day as slag is shed by its own weight, by sootblowing, and by load fluctuations. Although stresses from this slagging/shedding cycle are typically less than those from load changes or shutdowns, the cycles are more frequent. Once cracks are initiated by these fatigue mechanisms, repeated cycles can cause them to propagate quickly. Crack propagation may expose vulnerable base metal to corrosion mechanisms which further accelerate crack growth. Irregular stresses and strains and/or cracking can also cause spalling of protective oxide coating. This allows corrosion and/or reoxidation of the unprotected tube surface. Repeated cycles of cracking, spalling, corrosion and wastage can lead to rapid deterioration of the tube walls. The mechanisms may be accelerated by higher metal temperatures and uneven tube temperatures. This wastage can be made worse by reducing conditions at the furnace walls, caused by slag composition or by deeply staged combustion. Spalling can further accelerate the wastage, mentioned previously, that occurs when load cycling exposes the tube surface to alternating strong reducing conditions and oxidizing conditions. Tubes with weld overlay may also experience these fatigue mechanisms. The overlay is typically a high chrome-nickel alloy which has a slightly higher coefficient of thermal expansion than the base metal. Although the high alloy is more resistant to corrosion, it may also be more subject to thermal fatigue cracking as a result of this higher coefficient of thermal expansion. This vulnerability may be increased by stress concentrations created by the overlay weld profile.

Furnace Wall Metals Selection The issue of appropriate waterwall metals selection for adva nced supercritical units was 24 addressed early on by Blum. In boilers operating at 4640 psi/1157°F (320 bar/625°C), maximum midwall temperatures can be as high as 900–980°F (500–525°C) if there are scale deposits on the inside surface of the tube wall. At the higher temperatures, the creep resistance of standard low alloy ferritic steels such as T11 is not adequate. T91 steel has been the only suitable material for this application, with fabrication of this material into waterwalls demonstrated under 25 the European COST program. However, welding of P91 material must be followed by a wellcontrolled post-weld heat treatment procedure, which is difficult to perform in the field.

24

R. Blum, “Materials Development for Power Plants with Advanced Steam Parameters: Utility Point of View,” Proc. Materials for Advanced Power Eng., 3-6 Oct. 1994, Liege, Belgium, Kluwer Ac. Publ. Dordrecht, The Netherlands, 15. 25 C.J. Frankling and C. Henry, “Materials Development and Requirements for Advanced Boilers,” Ref. 31, p 89.

6-6


Two steels developed by Sumitomo and MHI, containing 2.5% and 12% Cr, respectively, appear to be more promising for waterwall fabrication because they do not require preheat or post-weld 26, 27,28 heat treatment. Both steels have creep strength in the same range as T91 and use similar precipitation strengthening mechanisms. The 2.5% Cr steel appea rs to be particularly suitable and has recently been approved, as T23, by the ASME boiler code committee. Test panels are now in service in various boilers.

Superheater and Reheater Design

Superheater and Reheater Corrosion For designs with higher final steam temperatures, the final superheater and reheater elements are among the most critical areas of the steam generator. These elements experience the highest metal temperatures and therefore require high resistance to creep, oxidation, and liquid phase ash corrosion. Although the reheater will always be at a lower operating pressure, with less stress on the metal, concerns for ash corrosion and oxidation remain. Ash constituents can be particularly aggressive at these locations. Many of the early U.S. attempts at higher steam temperatures, including Eddystone Unit 1, were frustrated by unexpectedly high corrosion rates of superheater elements. Although metals were found with adequate strength at the very high temperatures, they had very little resistance to ash corrosion. Ultimately, the final temperatures of the Eddystone unit had to be lowered to reduce corrosion rates to acceptable levels with the fuels that the plant owner wished to fire. Coal and coal ash chemistry are primary factors that must be known for predicting high temperature ash corrosion potential. Although sulfur and chlorine are often named as primary agents, it is the combination of certain coal constituents that determines the corrosion potential. In fact, some constituents (e.g., Mg and Ca) are sometimes credited for being mitigators of ash corrosion. Numerous opinions and theories attempt to explain the mechanisms and constituents responsible for high temperature ash corrosion. It is generally accepted that rapid corrosion rates of superheater and reheater tubing may be facilitated by alkalis, chlorides, sulfur, iron, and other low melting constituents (lead and zinc). Alkalis are the elements in the first column of the periodic table, with sodium and potassium being the most common. Salts and oxides of these elements melt at relatively low temperatures, in the range of 1100–1400°F (590–760°C), to form compounds and eutectics such as alkali-iron trisulfates. In the molten state, these compounds are very aggressive in attacking tubing materials. For final steam temperatures in the range of 1100– 1200°F (590–650°C), the outside tube surfaces temperatures, which are 50–150°F (28–83°C) above3 the steam temperatures, are in the “sweet spot” for the aggressive molten phase of these alkali sulfates. The Borio Index is one of the most widely used tools for predicting the interrelated corrosion effects of relevant constituents in coal. Borio developed a nomograph to predict ash corrosion potential of eastern U.S. coals, based on the following constituents: 26

Blum, op. cit. Frankling and Henry, op. cit. 28 F. Masuyama, Y. Sawargi et al., “Development of a Tungsten Strengthened Low Alloy Steel with Improved Weldability,” Reference 31, p 173. 27

6-7


•

Acid Soluble K2O

•

Acid Soluble Na2O

•

Fe2O3

•

CaO

•

MgO

Interestingly, this index does not directly consider the sulfur in coa l. In these eastern U.S. coals, sulfur is usually related very closely to iron content (occurring as Fe 2S). In the test burns that were used to develop the nomograph, iron content appeared to be more indicative of higher corrosion rates than sulfur content. This was perhaps because the sulfur in pyrite contributed more to corrosive reactions by way of slag deposits on the walls. The Borio Index also does not account for the chlorine in coal. The role of chlorine with respect to corrosion is a subject of some controversy. Based on U.S. experience, some cite alkali chlorides as significant agents in waterwall corrosion. Experience in the United Kingdom has also contributed to this conclusion, as boilers using high-chloride, high-alkali coals have experienced notoriously high corrosion rates. Other opinions hold that chlorides facilitate the volatilization of alkalis that would normally remain bound in clays and that these freed alkalis abet the alkali-iron trisulfate corrosive mechanism. These parties believe that higher chlorine levels in U.S. fuels will not necessarily contribute to higher corrosion rates unless there are also significant sources of alkali in the coal. It has been suggested that chlorides can be detrimental throughout the steam generator and backend equipment, all the way to the stack. One mechanism for tube wall wastage is the formation of iron chloride salts that vaporize at operating temperatures. As this corrosion product is carried away by the gas flow, the vulnerable tube wall may continue thinning until it fails. Chlorides may also have the ability attack at the location of any weakness in the metal grain structure. Often a pit will form and cause a failure even though the rest of the tube is sound. Chlorides in the fuel form hydrochloric acid (HCl) in the flue gas if there is sufficient moisture from the air, fuel, or combustion products. By itself, gaseous HCl is not terribly corrosive. However, being a strong acid, it displaces other anions in salts to form metal chlorides, which can be aggressive. The combination of alkalis and chlorides may be particularly damaging. As with many chemical reactions, the corrosion rate due to chloride mechanisms generally rises as temperature rises. Advancing steam temperatures therefore heighten concerns about chloride induced attack along with alkali trisulfate attack. Due to these concerns, steam ge nerator designers/manufacturers generally use the following guidelines to establish relative corrosiveness of fuels based on chlorine content: •

<0.2% Cl: Low corrosion potential

•

0.2–0.3% Cl: Moderate corrosion potential

•

>0.3% Cl: High corrosion potential

Oxidation and Exfoliation of Tube Metals Steam-side oxidation of tubes with subsequent exfoliation of the oxide scale is a well known cause of solid-particle erosion damage to the turbine as well as plugging of superheater and

6-8


reheater tubes. This problem is expected to be more severe in advanced steam plants, because the much higher steam temperatures are likely to cause more rapid formation of oxide scale.

Superheater and Reheater Metals Selections As higher steam temperatures are considered, metals must be selected to accomplish the following objectives in the highest temperature sections of the superheater and reheater: •

Adequate strength at the temperatures and pressures of service

•

Low oxidation rates at the temperature of service

•

Adequate corrosion resistance for the coal and ash characteristics expected at the temperature of service

Creep Rupture Strength

In high temperature superheater and reheater design, it is noted again that the average mean metal temperature can be higher than the steam temperature by as much as 50–150°F (28–83°C). Thus, tubes made of T22 should be limited to maximum steam temperature of 1000°F (538°C). Alloys T91, HCM12, EM12, HCM9M, and HT91 are limited to maximum steam temperature of 1050°F (565°C). Alloys T92, P122, and E911 are limited to maximum steam temperature of 1100°F (593°C), or metal temperature of about 1150°F (620°C). Under corrosive conditions, however, even the best ferritic steel may be limited to about 1125°F (610°C). S ome austenitic steels are suitable for SH and RH tube applications with metal temperatures in the range of 1150–1250°F (620–675°C). Above 1250°F (675°C), nickel-base alloys are needed. For convenience, austenitic steels can be classified as those containing less than 20% Cr and those containing more than 20% Cr. Alloy modifications based on the 18Cr-8Ni steels, such a s TP304H, 316H, 347H, and alloys with lower chromium and higher nickel contents, such as 17– 14 CuMo steel, fall into the classification of steels with less than 20% Cr. The allowable tensile stresses for steels in this class are intermediate between ferritic steels and high Cr austenitic stainless steels. Some high-creep-strength alloys containing more than 20% Cr, such a s HR3C (TP310H Nb-N), have been developed and accepted by ASME Code Case. These alloys offer low-cost alternatives to Incoloy 800 for use in the temperature range of 1150–1250°F (620– 675°C). Oxidation Resistance

Very limited data are available regarding the steam-side scale-growth characteristics of the ferritic tubing alloys. The following study results provide useful information: •

•

29

In a study by Sumitomo Metal Industries, the oxide growth in steam for alloys T22 (2-1/4Cr-1Mo), T9, HCM9M, and T91 (modified 9Cr-1Mo) were compared using 500-hour tests. Results showed the superiority of the T91 alloy over the other alloys. Masuyama et al. compared alloys HCM12, HCM9M, 321H, and 347H in field tests in the 30 temperature range 1020–1155°F (550–625°C) over a period of one year. Samples were

29

Properties of Super 9Cr Steel Tube (ASTM) A 213-T 91, Report 803 F-No. 1023, Sumitomo Metal Industries, (July 1983).

6-9


inserted in the tertiary and secondary superheaters and reheaters. Based on results, they concluded that the resistance to steam oxidation of HCM12 is superior to those of 321H and HCM9M and comparable to that of fine-grained 347H for exposure to the high-temperature region of the reheater. Subsequent monitoring over a period of three years has borne out the 31 earlier conclusions. •

•

•

•

•

In addition to the inherent resistance of HCM12M steel to steam-side oxidation, Masuyama et al. suggest that the tendency toward exfoliation of oxide scale would also be less for this 32,33 alloy than for austenitic steels. Additional improvements in 9–12% Cr steels may be possible by extending the 34,35 36 chromizing and chromate conversion treatments that currently are applied to lower-alloy steels. Grain refinement during heat treatment has been shown to be clearly beneficial as well. Internal shot blasting is also known to improve the steam oxidation resistance of 300 series stainless steels by enhancing chromium diffusion. It is therefore anticipated that these steels would be used in the fine-grain and shot-peened conditions. Results of steam oxidation tests at 1200°F (650°C) for times up to 2000 h have been reported 37 for several austenitic steels. Steamside oxidation results on Ni base alloys are not available. The upper temperature limits for these alloys have not been defined.

Corrosion Resistance

Because resistance to fire-side corrosion increases with chromium content, the 9–12% Cr ferritic steels are more resistant than the 2-1/4Cr-1Mo steels currently used. The 12% Cr steel in turn shows better corrosion resistance than 2-1/4% Cr steel and 9% Cr steel. Stainless steels and other superalloys containing up to 30% Cr represent a further improvement. Increasing the chromium content beyond 30% results in a saturation effect on the corrosion resistance (according to

30

F. Masuyama et al., “Development and Applications of a High Strength, 12% Cr S teel Tubing with Improved Weldability,” Technical Review, Mitsubishi Heavy Industries, Ltd., Japan, (Oct. 1986), pp 229-237. 31 F. Masuyama et al., “Three Years of Experience with a New 12% Cr Steel in Superheater,” Advanced Materials Technology for Fossil Fuel Power Plants, R. Viswanathan and R.I. Jaffee, Eds., American Society for Metals, Metals Park, OH, (1987), pp 259-266. 32 F. Masuyama et al., “Development and Applications of a High Strength, 12% Cr Steel Tubing with Improved Weldability,” Technical Review, Mitsubishi Heavy Industries, Ltd., Japan, (Oct. 1986), pp 229-237. 33 F. Masuyama et al., “Three Years of Experience with a New 12% Cr Steel in Superheater,” Advanced Materials Technology for Fossil Fuel Power Plants, R. Viswanathan and R.I. Jaffee, Eds., American Society for Metals, Metals Park, OH, (1987), pp 259-266. 34 A.J. Blazewicz and M. Gold, “Chromizing and Turbine Solid Particle Erosion,” ASME Paper No. 78, JPGC PWR7, Joint ASME/IEEE/ASCE Power Generation Conference, Dallas, (Sept 1978). See also: P.L. Daniel et al., “Steamside Oxidation Resistance of Chromized Superheater Tubes,” CORROSION 80, NACE Conference, Chicago, (May 1980). 35 P.L. Daniel et al., “Steamside Oxidation Resistance of Chromized Superheater Tubes,” CORROSION 80, NACE Conference, Chicago, (May 1980). 36 J.M. Rehn et al., “Controlling Steamside Exfoliation in Utility Boiler Superheaters and Reheaters,” Paper No. 192, CORROSION 80, NACE Conference, Chicago, (May 1980). 37 K. Kubo et al., “Application of Boiler Tubing Tempalloy Series to the Heat Exchanger of Advanced Coal-Fired Boilers,” in Proceedings of the First International Conference on Improved Coal-Fired Power Plants, A.F. Armor, W.T. Bakker, R.I. Jaffee, and G. Touchton, Eds., Report CS-5581-SR, EPRI, Palo Alto, CA , (1988), pp 5-237 to 5-254.

6-10


laboratory experiments). For practical purposes, when corrosive conditions are present, fine distinctions between ferritic steels may be academic, as it is usually necessary to use austenitic steels containing chromium in excess of 20%. Extensive field experience at EPRI and TVA has shown fireside corrosion is very localized, even with very corrosive coals. Thus, 20% Cr alloys are generally suitable as the primary material of construction. Specific areas where severe corrosion is predicted by combustion modeling or found after initial operation can then be made more corrosive resistant by applying high chromium weld overlays. Austenitic steels are generally better candidate alloys than ferritic steels for tubing transitions from Ni alloys used in high vulnerability areas. Considerable development and commercialization work is being done to provide materials which address thee issues of high temperature strength, oxidation resistance an d corrosion resistance. Commercially available and proven materials for high temperature applications which have been included in ASME Section 2, or have been accepted by Code Case (CC), are presented in Figure 6-1. These currently commercial materials are sufficient to allow designs up to 1150°F (620°C) for many coal types. For some of the more corrosive coals, it may be desired to arrange the superheater circuitry with flow concurrent with the gas flow. This is less efficient for surface utilization, but will allow lower maximum tube surface temperatures which will lower the corrosion potential for these problem coals. For the most aggressive coals, even HR3C (SA-213 TP310H-Nb-N) arranged in concurrent flow may not provide adequate corrosion resistance when final steam temperatures are as high as 1150°F (620°C). In these situations, the Owner/Engineer may decide lower steam temperatures will be necessary. If there is a strong impetus for the highest temperatures, however, a preventive inspection/maintenance program may be feasible to monitor corrosion areas and prevent forced outages due to tube failures. Alternately, coating with a corrosion inhibiting metal spray can be considered, although such coatings will themselves likely require an inspection and maintenance program.

Table 6-1 Temperature Limits for Materials Proven in High-Temperature Applications Material Specification

Nominal Composition

Max. Temperature Limit at O.D. for corrosion resistance, °F (°C) for Sulfur (S) in Fuel, wt % S<1.2

1.2
S>1.8

SA-213 T22

2 ¼ Cr, 1 Mo

1100 (593)

1100 (593)

1075 (579)

SA-213 T23 (CC: 2199)

2 ¼ Cr, 1.6 W, V, Nb

1100 (593)

1100 (593)

1075 (579)

SA-213 T91

9 Cr, 1 Mo, V

1125 (607)

1125 (607)

1100 (593)

SA-213 T92 (CC: 2179)

9 Cr, 2W, Mo, V, Nb, N

1125 (607)

1125 (607)

1100 (593)

SA-213 T122 (CC: 2180)

12 Cr, 1.5 W, Mo, V, Nb, N, Cu

1150 (621)

1150 (621)

1125 (607)

SA-213 TP304

18 Cr, 8 Ni

1400 (760)

1150 (621)

1125 (607)

6-11


“Super 304” (CC: 2328)

18 Cr, 9 Ni, 3 Cu, Nb, N

1400 (760)

1150 (621)

1125 (607)

SA-213 TP347

18 Cr, 10 Ni, Nb

1400 (760)

1150 (621)

1125 (607)

SA-213 TP347 HFC (CC:2159)

18 Cr, 10 Ni, Nb, fine grain

1400 (760)

1150 (621)

1125 (607)

SA-213 TP310H Nb N (HR3C) (CC: 2115)

25 Cr, 20 Ni, Nb, N

1500 (816)

1250 (677)

1225 (663)

Recommendations for High Steam Temperature Applications Supercritical steam generators are likely to require larger furnace v olumes when steam conditions are increased to improve heat rate, and as lower NOX combustion is implemented. Larger furnace volume provides several advantages for optimizing furnace materials selection, performance and reliability. For the same level of heat input, the higher furnace volume provides a lower volumetric heat release rate, resulting in a lower mean bulk temperature. This lower mean bulk temperature tends to reduce the tendency for molten or sticky ash to reach the waterwalls and form deposits or running slag. As lower temperatures promote less NO X formation, less severe staging of combustion air is required. Further, reducing mean bulk temperature reduces the intensity of radiation heat flux to the walls and the peak metal temperatures on the crowns of the tubes. These reduced metal temperatures translate into lower corrosion potentials and reduced likelihood of ash adherence. The larger furnace volume also results in a higher gas residence time. This increased residence time allows the gases to reach a lower average temperature leaving the furnace, which, in turn, promotes resolidification of molten or sticky ash particles. The maintenance of consistently oxidizing conditions, at the waterwalls, is particularly important in mitigating ferric oxide reduction, sulfidation and alkali-chloride attack. The methods employed by equipment suppliers to accomplish this include curtain air, boundary air, offset air, and concentric air. Although it is desirable to have an oxidizing atmosphere at the furnace wall, it is not desirable to have fuel combustion or unburned carbon deposits at the wall. Such a condition may further contribute toward corrosion mechanisms due to the generation of the reducing gas constituent, CO. It is important to design burners or fuel nozzles which have good jet penetration into the furnace and avoid flame “lick-back” on the furnace walls. To assure the success of a furnace wall air blanketing system, it is necessary to verify that the furnace manufacture has adequate experience for similar size units with similar firing systems burning similar coals. Steam generator suppliers should be asked to provide a history of development and experience with their systems that demonstrates adequate understanding of this design area. One good way to verify effectiveness of a given system is to observe improvements in corrosion rates resulting from a retrofit of a system. For example, a degree of confidence can be established if a retrofit system has been demonstrated to show marked reductions in wastage/corrosion rates in a unit firing an aggressive coal (e.g., high sulfur, chorine, alkali, etc.), even if the system has not been used for a new unit firing aggressive coals. An alternative or supplementary approach for addressing furnace wall corrosion and superheater/reheater tube corrosion with more aggressive coals and/or higher steam temperatures is to provide metal surfaces which are very resistant to corrosion. The first step is to move to

6-12


base tube metals which are more resistant (e.g., the T22, T23 metals which have a little higher chromium). T91 tubes have more chromium still, but present significant difficulty with field welding due to unforgiving heat treatment requirements. Some available high chromium a lloys with improved corrosion resistance for superheater and reheater surfaces are listed below. Furnace wall coatings (e.g., flame spray, metallizing), including plasma sprays and high velocity oxy-fuel (HVOF), have been steadily improving and are very effective while the coatings remain on the tubes. These coatings have been applied to superheater and reheater surfaces, as well, with varying success. Vendors should be queried about recent experience and longevity of their offered coating systems. Some economic analysis of initial cost versus maintenance cost of the coating will be necessary. One of the most effective corrosion deterrents is weld overlay. A high chromium/nickel alloy has been proven to be very effective at arresting even severe corrosion situations. Candidate overlay weld metals include In617 (22Cr, 44.5 Ni), In72 (44Cr, balNi), and In671 (48Cr, balNi). Austenitic TP 309 has also been used as an overlay material in the past, but has the disadvantage of a higher thermal coefficient of expansion compared with the base furnace tube metals. The high chrome-nickel alloys have good corrosion resistance and have a thermal coefficient of expansion close to that of ferritic tubing. However, with the slight difference in expansion coefficient, and because the overlay makes for an overall thicker metal wall, overlay is sometimes more subject to thermal fatigue cracking than bare tubes. Extensive development in strengthening of 9–12% ferritic steels (T91/P91, T92/P92, T122/P122) as well as advanced austenitics (Super 304, TP347 HFG, HR3C) have resulted in temperature/pressure capabilities well over the conventional framework of about 1000°F (540°C) for the steam. Nearly two dozen plants have been commissioned worldwide with main steam temperatures of 1080–1112°F (580–600°C) and pressures of 3400–4200 psi (235–290 bar). To assure a design which will provide long life and high reliability, it is recommended that specific materials and temperature limits be imposed in the steam generator specifications according to the characteristics of the coal. The steam generator supplier should be asked to demonstrate how his design will adhere to the temperature limits, including expected upsets in steam flow and uneven absorption rates.

Headers and Piping

Choices of Materials Material-property requirements for headers and steam piping are likely to be similar. Hence they have been grouped together. Some differences exist which may affect material selection. The steam temperature is likely to be more uniform in piping whereas differential heating in the attached tubing makes headers more subject to time-dependent and location-dependent

6-13


fluctuations. Self-weight-induced stresses are less important for headers than for steam piping. The location of attemperators, drains, and other connections may influence thermal or mechanical stresses seen by headers or piping. Depending on the selection of materials for the superheater/reheater tubes and the header piping, dissimilar-metal welded joints may be required. Headers generally have thicker walls because of their larger diameters and to provide reinforcement for penetrations provided for tube and branch connections. The many welded attachments to inlet stub tubes from reheaters and superheaters and to outlet piping add further sources of stress. The thicker walls, stress concentrations created by pen etrations, stresses created by attached tubes and piping, and non-uniform temperatures all increase vulnerability to thermal fatigue. In the United States, headers and piping have traditionally been made from low alloy steels such as P11 and P22. For headers, especially, use of lower strength steels increases the required wall thickness and increases the vulnerability to thermal and mechanical fatigue, especially in units with significant load cycling. A number of (mostly conventional) units have required expensive weld repairs and header replacements when fatigue cracks have propagated to the point where the headers had to be removed from service. A common failure mode is the cracking of the 38 ligaments between the tube boreholes. Without proper selection of materials, design, and fabrication techniques, the use of higher temperatures and pressures can further increase the problem. With design thicknesses closer to their stress limits, combined with self-weight, main steam piping and hot reheat piping are more likely to be vulnerable to creep or creep-fatigue interaction. Several high profile incidents have emphasized the need to monitor and maintain or replace these piping systems in conventional and supercritical plants. In many cases, seamwelded P11 and P22 piping have been replaced with seamless P91 or other higher strength steels. Use of higher strength steels significantly reduces thermal stress mechanisms in thick-walled piping and headers. Previous attempts to use austenitic steels have not been successful due to high thermal expansion of these steels. As shown in Table 6-2, ferritic steels can be used up to the temperature limits indicated. The most creep resistant steels, P92, P122, and P911 can be used for heavy-section applications up to 1150°F (620°C), although steamside oxidation may limit their applicability at higher temperatures. At temperatures exceeding 11 50°F (620°C) but below about 1250°F (675°C), austenitic steels may be needed. Beyond 1250°F (675°C) but below about 1450°F (788°C), nickel-base alloys may be used.

38

R. Viswanathan, “Damage Mechanisms and Life Assessment of High-Temperature Components,” ASM International, Metals Park, OH, (1989).

6-14


Table 6-2 39 Evolution of Four Generations of Ferritic Steels 5

Generation

Era

Alloy Modifications

Strength 10 hr Creep Rupture Achieved MPa (ksi) at 600°C (1112°F)

Example Alloys

Maximum Metal Use Temperature* °C (°F)

1

1960–70

Addition of Mo or Nb, V to simple 12Cr and 9Cr Mo steels

60 (8.7)

EM12, HCM9M, HT9, Tempaloy F9, HT91

565 (1049)

2

1970–85

Optimization of C, Nb, V

100 (14.5)

HCM12, T91, HCM2S

593 (1100)

3

1985–95

Partial substitution of W for Mo

140 (23.2)

P92, P122, P911 (NF616, HCM12A)

620 (1148)

4

Emerging

Increase of W and addition of Co

180 (26.1)

NF12, SAVE12

650 (1202)

5

*Based on 100 MPa/10 h 40

Figure 6-1 shows a plot of the allowable stress at various temperatures for ferritic steels. The figure clearly shows the enormous advances in the materials technology that have been made in the last 20 years. Especially at the higher temperatures, the most advanced steels show allowable stresses that are nearly 2.5 to 3 times that of 2-1/4Cr-1Mo steel (P22), the “workhorse steel” in conventional plants. The layering of the alloys into the different generations as noted above is also evident. HCM12A (P122), NF616 (P92), and E911 emerge as the three highest strength alloys suitable for ultra-supercritical plants up to 1150°F (620°C), followed by T 91, HCM12, EM12, and HT91 suitable for intermediate temperatures up to 1100°F (593°C), followed by T22 for use up to 1050°F (565°C). NF12 and SAVE12 are still developmental. Alloy HCM2S has much higher strength than P22 and is weldable, making it suitable for application as a replacement for P22. Fujita has reported on a modified version of NF12 with aluminum content below 20 ppm and Ni content below 0.1%. This alloy has better creep properties than NF12 and 41 is believed to have adequate strength for 1200°F (650°C) applications.

39

R. Viswanathan, R. Purget, and U. Rao, op. cit. F. Masuyama, “New Developments in Steels for Power Generation Boilers,” in A dvanced Heat Resistant Steels for Power Generation, R. Viswanathan and J.W. Nutting, Eds.; IOM Communications Ltd., London, (1999), pp 33-48. 41 R. Viswanathan, R. Purget, and U. Rao, op. cit. 40

6-15


Figure 6-1 42 Comparison of Allowable Stresses of Ferritic Steels for Boiler

Some additional design considerations in applying the advanced ferritic steels are as follows: •

•

•

•

•

The advanced alloys NF616, HCM12A and E911 (P92, P122, E911) have essentially the same high temperature strength as low-end austenitic alloys. However, they have less oxidation resistance than the austenitic alloys. This parameter of advanced 9Cr to 12Cr alloys must be more fully evaluated prior to application to high temperature parts. Post-weld heat treatment (PWHT), along with careful attention to welding procedures, is always required for welded joints of advanced 9Cr to 12 Cr alloys. This ensures minimal stress and optimal ductility. To control production and P WHT costs, designs should maximize shop welding and minimize field connections to reduce field heat treatment as much as possible. In the weldment of dissimilar alloys, material selection must be based on consideration of PWHT temperature. The creep rupture strength of welded joints is the most important consideration for high temperature piping and headers with longitudinal (seam) welds. The 9Cr-1Mo alloy and 1Cr-0.5Mo steels would generally not be acceptable materials where these joints are required. A further concern is the apparent susceptibility of ferritic steel welds to Type IV cracking, which occurs at the edges of fine-grained heat-affected zone (HAZ) material adjacent to unaffected parent material. Susceptibility to this has been clearly demonstrated for 1/2CrMoV, 2-1/4Cr-1Mo and 9Cr-1Mo (T91) steels. Safety margins of 10–20% are sometimes adopted to provide for this mechanism. Because the problem in girth welds is primarily associated with bending stresses, the problem can be overcome by proper design and maintenance of piping geometry and pipe supports.

42

F. Masuyama, “New Developments in Steels for Power Generation Boilers,” in Advanced Heat Resistant Steels for Power Generation, R. Viswanathan and J.W. Nutting, Eds.; IOM Communications Ltd., London, (1999), pp 33-48.

6-16


Figure 6-2 Comparison of Allowable Stress for Various Metals

Figure 6-2 is a plot of the allowable stresses versus temperatures for comparing the temperature 43 capabilities of ferritic alloys with austenitic steels and Ni base alloys. The figure also shows the actual stresses at several steam pressures. The Ni base alloys are superior to the austenitic steels, which, in turn are better than ferritic steels. The nickel-base alloys Inco740, Haynes 230, IN625, IN617, HR6W, and HR120 have much higher temperature capability, in decreasing order as listed compared with austenitic steels, followed by ferritic steels. Purely from the creep strength point of view, at a pressure of 5500 psi (379 bar) for a 2 inch (50 mm) x 0.5 inch (13 mm) tube (at a stress of 8.6 ksi or 60 MPa), ferritic steels are useful up to about 1150°F (620°C) (metal temperature), and austenitic steels up to about 1250°F (675°C). At metal temperatures higher than about 1250°F (675°C), nickel-base alloys are needed. The alloy Inco740 appears capable of reaching 1450°F (788°C). Because the thick walled components are used over a range of conditions, all of the above categories of materials might be used at different locations in an advanced ultra-supercritical plant.

Advanced Ferritic Materials: Specific Issues Related to Grade 91 Steel Fossil power plants that are subject to frequent cycling operation can experience early failure due to fatigue and creep-fatigue damage. To protect against long-term reliability issues it is advisable, particularly for higher temperature supercritical units, to deploy strong ferritic alloy steels. For more than 20 years, the 9% chrome material known as Grade 91 has been used in fossil plant applications. Originally formulated in the United States as a material for the breeder reactor by Combustion Engineering and the Oak Ridge National Laboratory, the material then 43

R. Viswanathan, R. Purget, and U. Rao, op. cit.

6-17


known as super 9-chrome was standardized in the United States in 1983–84 and first found its niche in fossil plants as a design option for advanced supercritical units. Recently, a few cases of cracking in Grade 91 components during operation have been reported.

Figure 6-3 Vallourec & Mannesmann Hot Neck P91 Fitting

Fossil plant components, such as headers, steam lines, and boiler tubing, have increasingly been successfully retrofit to existing plants to improve reliability. Figure 6-3 shows a hot neck P91 fitting from Vallourec & Mannesmann. Today, particularly for superheater tubing and for steam lines and headers, Grade 91 steel is generally showing its resilience in the fossil power plant industry, validating its e arly promise as a breakthrough material of great strength. Lately the search for even stronger ferritic steels has led to evolutions of the grade 91 material formulation (e.g., Grades 92, 122) using small additions of tungsten to push the strength limits (Table 6-3).

6-18


Table 6-3 Composition of Advanced Steels, including Tungsten-Containing P92, P122, and E911 EM12

X20

T91/P91

T92/P92

HCM12A (T122)

E911

C

max. 0.15

0.17-0.23

0.08-0.12

0.07-0.13

0.07-0.13

0.09-0.13

Si

0.20-0.65

max. 0.50

0.20-0.50

max. 0.50

max. 0.50

0.10-0.50

Mn

0.80-1.30

max. 1.00

0.30-0.60

0.30-0.60

max. 0.70

0.30-0.60

P

max. 0.030

max. 0.030

max. 0.020

max. 0.020

max. 0.020

max. 0.020

S

max. 0.030

max. 0.030

max.0.010

max. 0.010

max. 0.010

max. 0.010

Ni

max. 0.30

0.30-0.80

max. 0.40

max. 0.40

max. 0.50

0.10-0.40

Cu

—

—

—

—

0.30-1.70

—

Cr

8.50-10.50

10.00-12.50

8.00-9.50

8.50-9.50

10.0-12.5

8.50-9.50

Mo

1.70-2.30

0.80-1.20

0.85-1.05

0.30-0.60

0.25-0.60

0.90-1.10

W

—

—

—

1.50-2.00

1.50-2.50

0.90-1.11

V

0.20-0.40

0.25-0.35

0.18-0.25

0.15-0.25

0.15-0.30

0.18-0.25

Nb

0.30-0.55

—

0.06-0.10

0.04-0.09

0.04-0.10

0.06-0.10

Al

—

—

max. 0.040

max. 0.040

max. 0.040

max. 0.040

N

—

—

0.030-0.070

0.030-0.070

0.040-0.100

0.050-0.090

B

—

—

—

0.001-0.006

max. 0.005

0.0005-0.005

Background on P91 in the Fossil Industry In the early 1980s, EPRI led teams of boiler and turbine equipment manufacturers and steel producers from around the world in the search for materials for advanced supercritical fossil plants. Work under EPRI’s RP1403 project on super 9-chrome led to its being selected as an ASME standardized material in 1984. Clearly stronger than the commonly accepted alloys P11 and P22, applications of P91 to fossil plants began in the late 1980s. Early U.S. Applications

In 1991, Lower Colorado River Authority replaced a P11 secondary superheater outlet header 44 that had shown signs of swelling, with a header made of P91 steel. In 1992, Mannesmann supplied P91 headers as replacements for P22 headers at Dayton Power and Light’s Stuart plant 45 in Aberdeen, Ohio. In 1993, San Diego Gas and Electric replaced P11 main steam line fittings 46 with P91 in Encina units 4 and 5, when evidence appeared of circumferential weld cracking. In 1993, Consumers Power replaced at Unit 2 of the J.H. Campbell station two P22 secondary

44

LCRA Demonstrates Advanced Material for Replacement Header, EPRI Innovators IN-100457, November 1992. C.P. Bellanca, J. Infield, et al., ASME Pressure Vessel and Piping Conference, New Orleans, LA, 1992. 46 “San Diego Gas and Electric Increases Plant Safety by Using F91 Steel in Wye Fittings,” EPRI Innovators IN-103316, November 1994. 45

6-19


superheater outlet headers with a single P91 header, subsequently reported in 1997 to be 47 operating successfully with no evidence of cracking. The American Electric Power units, retrofitted in 1994–95 with P91 material and reported above, were operating as of June 2003 48 with no evidence of cracking.

Figure 6-4 P91 Superheater Outlet Headers for Dayton Power and Light, Stuart Station

Costs and Benefits of Grade 91 The benefits of using P91 steel over standard alloys such as P11 or P22 can be put into three categories: •

Greater strength, that permits increased safety margins in existing units.

•

Significantly longer component life for given creep and fatigue duty.

•

Reduced wall thickness for tubing and piping for the same design conditions, leading to lower thermal storage and less thermal stress.

An indication of the greater strength of P91 over a 100,000 hour lifetime is seen in the rupture strength curves in Figure 6-5. As for operating lifetime comparisons they depend on the creep and fatigue duty. At 1110°F (600°C) and 14.5 ksi (100 MPa), P91 reaches about 80–90,000 hours of operation before rupture, whereas P22 reaches less than 1000 hours before rupture.

47

“Secondary Superheater Outlet Header Material and Design Upgrade,” Information Document, Babcock and Wilcox Company, 1997. 48 Personal Communication, Tom Andress (AEP) to Tony Armor (EPRI), June 23, 2003.

6-20


High Temperature Steels: Relative Rupture Strength P92 P91 2.25 CrM o 1CrMo 450

500

550

600

650

Maximu m Tem per ature Deg. C for 100,000 Hour s at 100 MPa

Figure 6-5 Relative Rupture Strength of High Temperature Steels

The potential reductions in wall thickness can be seen graphically in Figure 6-6, which shows the relative piping wall thickness of P91, P22, and other ferritic steels for the same design conditions.

Figure 6-6 Comparison of Piping Wall Thickness for Candidate Ferritic Steels

Issues in Welding and Forming P91 As noted, obtaining desired material life from Grade 91 components requires careful attention to welding and fabrication procedures. Cracking of P91 tubes and pipes that has occurred on some installations appears to be due to improper procedures for welding and post-weld heat treatment.

6-21


Other concerns include cold forming requirements, stress corrosion cracking, and making a reliable transition joint between Grade 91 and other alloys. This latter issue, of current concern, is affected by the thickness of the weld metal, the filler material, and the post-weld heat treatment temperature. Figure 6-7 shows a typical, incorrectly made joint between P22 and P91 piping. Figure 6-8 diagrams two types of errors made for this type of joint. Figure 6-9 diagrams the correct weld profile. A list of EPRI reports relating to c orrect practices in welding P91, and in PWHT approaches for different weld metals, are provided in Table 6-5.

Figure 6-7 Typical P91 to P22 Weld — Vulnerable to Cracking at Junction between B9 Filler and P22

6-22


B9 Filler 30o max

Grade 22

Grade 91

B3 Filler 30o max

Grade 22

Grade 91

Figure 6-8 Joint Geometries of Concern — Highlighted Transition Indicates Weakest Part of Weld

B9 Filler Grade 22

30o max

Grade 91

Figure 6-9 Correct Weld Profile for P91 to P22 Welds

Use of Grade 92 and Grade 122 Materials The ASME Boiler and Pressure Vessel Code has approved Code Cases for the use of Grade 92 (Code Case 2179-3) and Grade 122 (Code Case 2180-2) material for construction in ASME Sections I (Power Boilers) and VIII (Unfired Pressure Vessels). However, ASME B31.1 Code for Power Piping does not currently list Grade 92 or Grade 122 material as Listed Materials within the B31.1 Code, and consequently there are no allowable stress values listed in the B31.1 Code for these materials. As of early 2007, a tentative revision in progress in the B31.1 Code Committee (Ballot 06-412) to permit the acceptance of ASME B&PV Section I Code Cases for use within the ASME B31.1 Code. Timing for the possible acceptance and publication of this revision to the ASME B31.1 Code is unknown. Because Grade 91 is a Listed Material in the B31.1 Code, established requirements exist for design, fabrication, and installation. This is not the case for Grade 92 and Grade 122, which are

6-23


presently considered by ASME B31.1 Code as Unlisted Materials. Should the Owner desire to use Grade 92 or Grade 122 for piping systems external to the boiler (such as for main steam or hot reheat), a number of steps must be secured, including: •

•

•

The designer shall document the Owner’s acceptance for use of the Unlisted Material. The applicable jurisdiction within the State shall be informed of this request to utilize an Unlisted Material for work covered by the ASME B31.1 Code, and the State must provide its acceptance for use of the Unlisted Material on the Project. The Owner’s Authorized Inspection Agency who will sign-off on the ASME Data Forms (P-4 and P-4A Forms) for ASME B31.1 Boiler External Piping shall be advised of both the Owner’s and State’s acceptance of the Unlisted Material and the AI’s acceptance/agreement to sign-off the forms shall be secured.

How to Deploy Advanced Steels: A Summary Over 20 years of operational experience in the use of Grade 91 steel has developed fairly comprehensive knowledge and expertise for how and when to deploy this steel to best effect. There is a more limited experience base for other advanced ferritic steels. Avoiding failures of tubes and pipes, particularly in advanced supercritical units, depends on careful attention to field fabrication methods, welding techniques, and optimum operational procedures. Table 6-4 tabulates strategies to use to achieve successful applications of these materials that minimize the 49 risks of premature damage accumulation.

49

“Issues/Risks in Designing for Final Steam Temperatures Over 1100°F,” WorleyParsons. Personal communication Paul Shewchuk (WorleyParsons) to Tony Armor (EPRI), February 7, 2007.

6-24


Issue/Risk

Discussion

Mitigation Strategies ash corrosion.

Rapidly falling allowable stresses for materials at high temperatures reduces margins of material strength during periods of upset.

ASME allowable stresses at temperatures over 1100°F (593°C) are on a rapidly declining curve. This means that minor temperature excursions translate into large changes in allowable stress.

Confirm that boiler suppliers are providing adequate margin by requesting temperature/design curves with proposal and contract. Specify pipe with enough margin to account for anticipated changes in allowable stress with temperature fluctuations. Restrict boiler temperature change rates and cyclic operation to reduce temperature fluctuations.

Limited availability of high alloy pipe, valves, and fittings.

Both Grade 91 and Grade 92 pipe are available from the pipe manufacturers. Current delivery time of Grade 91 and Grade 92 is 15 to 17 months after receipt of order. Fabrication time adds approximately 24 weeks after material receipt and delivery to the jobsite. Fittings for Grade 91 (forgings for safety valve nozzles, heavy wall laterals or “Y” fittings, and 2 and under connections) are available with deliveries ranging from 8 to 9 months at present. ʺ

Both forged steel valves (ASTM/SA 182 F91) and cast steel valves (ASTM/SA 217 WC12a) for Grade 91 are readily available with current deliveries between 32 to 38 weeks after receipt of an order. No large bore cast steel valves are presently included in the design of the main steam and hot reheat piping systems. Grade 91 weld filler metal is available and a number of suppliers are carrying quantities in inventory. Current delivery of Grade 91 weld filler metal is not a problem with an average price of $20.00 per pound. Fittings for Grade 92 (forgings for safety valve nozzles, heavy wall laterals or “Y” fittings, and 2 and under connections) are available with deliveries ranging from 12 to 14 months at present. ʺ

Grade 92 valves are not readily available as most manufacturers in the United States have not started the process for the production of Grade 92 valves. As of early 2007, there is no specification for cast steel valves in the Grade 92 material in ASME B16.34. The only specification for Grade 92 valves is for forged valves (ASTM/SA A182 F92). Grade 92 weld filler metal is very difficult to obtain and has very long lead times. It

6-29


Issue/Risk

Discussion

Mitigation Strategies

appears at present that Grade 92 weld filler metal may be available in large mill run lot quantities, but stock inventories are extremely limited and located outside the United States. Price for Grade 92 filler metal is averaging $40–50 per pound.

6-30


Table 6-5 50 EPRI Documents Related to Forming and Welding P91 in Fossil Plants TR-101394

Thick-Section Welding of Modified 9Cr-1Mo (P91) Steel

TR-103617

P91 Steel for Retrofit Headers -- Materials Properties

TR-104845

Creep Behavior of Modified 9% CrMo Cast Steel for Application in Coal-Fired Steam Power Plants

TR-105013

Material Considerations for HRSGs in Gas Turbine Combine Cycle Power Plants

TR-106856

Properties of Modified 9Cr-1Mo Cast Steel

TR-108971

Review of Type IV Cracking in Piping Welds

TR-111571

Advanced Heat Resistant Steels for Power Generation

TR-114750

Materials for Ultra-Supercritical Fossil Power Plants

1001462

Advances in Material Technology for Fossil Power Plants—SWANSEA

1006299

Conference on 9Cr Materials Fabrication and Joining Technologies

1006590

Guideline for Welding P(T)91 Materials

1004516

Performance Review of P91/T91 Steels

1004527

Development of Advanced Methods for Joining Low-Alloy Steel

1004702

Optimal Hardness of P91 Weldments

1004703

Post Forming Heat Treatment of P91 Materials

1004915

Normalization of Grade 91 Welds

1004916

Development of Advanced Methods for Joining Low-Alloy Steel

1009758

Evaluation of Filler Materials for Transition Weld Joints between Grade 91 to Grade 22 Components

1011352

Effect of Cold-Work and Heat Treatment on the Elevated-Temperature Rupture Properties of Grade 91 Material

1009757

Temperbead Repair Welding of Grade 91 Materials

1009758

Evaluation of Transition Joints between Grade 91 and Grade 22 Components

Examples of Current Costs of P91 versus P92 Pipe for Main Steam and Hot Reheat Pipe at Different Design Conditions The design conditions for temperature are established by taking the turbine throttle valve temperature requirement, adding 5°F (3°C) for temperature drop from boiler to turbine and then adding a 10°F (6°C) operating margin. For example, a 1075°F (579°C) turbine throttle valve requirement plus 5°F (3°C) equals 1080°F, (582°C) plus 10°F (5°C) margin equals a 1090°F (588°C) design temperature for a main steam pipe system.

50

Productivity Improvement for Fossil Steam Power Plants, 2005: One Hundred Case Studies , EPRI Report 1012098, July 2005; List of papers provided by Kent Coleman, EPRI.

6-31


NOTE: The minimum wall thickness for P92 pipe is based upon the anticipated reduction of the allowable stress values by 18% at the higher temperatures versus the allowable stress values presently shown in ASME Code Case 2179-3. Table 6-6 Specification Example for Main Steam Piping for Supercritical Steam Conditions MAIN STEAM 4 LEADS — 850 MW NET Design Conditions: 1090°F @ 3950 psig (588°C @ 272 bar) ba r) Operating Conditions: 1075°F @ 3800 psig (579°C @ 262 bar) at Turbine ASTM/SA P91

ASTM/SA P92

20 (508)

20 (508)

(73.9) 2.908 (73.9)

2.723 (69.2)

(kg/m) Weight – lb/ft (kg/m)

(835) 561 (835)

518 (771)

Average Cost – US$/lb (US$/kg) (US$/kg)

4.00 (8.8) (8.8)

4.80 (10.6)

2,241 (7,352) (7,352)

2,486 (8,156)

Outside Pipe Diameter – in. (mm) in. (mm) Minimum Wall Thickness – in. (mm)

Average Cost of Pipe – US$/ft (US$/m)

Table 6-7 Specification Example for Hot Reheat Piping for Supercritical Steam Conditions HOT REHEAT 2 LEADS — 850 MW NET Design Conditions: 1115°F @ 690 psig (602°C @ 48 bar) Operating Conditions: 1100°F @ 670 psig (593°C @ 46 bar) b ar) at Turbine) ASTM/SA P91

ASTM/SA P92

34 (864)

34 (864)

1.198 (30.4) (30.4)

1.086 (27.6)

Weight – lb/ft (kg/m) (kg/m)

491 (731) (731)

409 (609)


6.00 (13.2) (13.2)

7.20 (15.9)

2,946 (9,665) (9,665)

2,945 (9,662)



6-32


Table 6-8 Specification Example for Main Steam Piping for Ultra-Supercritical Steam Conditions MAIN STEAM 4 LEADS — 850 MW NET Design Conditions: 1127°F @ 3950 psig (608°C @ 272 bar) ba r) Operating Conditions: 1112°F @ 3800 psig (600°C @ 262 bar) at Turbine ASTM/SA P91

ASTM/SA P92

20 (508)

20 (508)

3.604 (91.5) (91.5)

3.183 (80.8)

Weight – lb/ft (kg/m) (kg/m)

660 (982) (982)

587 (874)


4.00 (8.8) (8.8)

4.80 (10.6)

2,640 (8,661) (8,661)

2,818 (9,245)



Table 6-9 Specification Example for Hot Reheat Piping for Ultra-Supercritical Steam Conditions HOT REHEAT 2 LEADS — 850 MW NET Design Conditions: 1145°F @ 690 psig (618°C @ 48 bar) Operating Conditions: 1130°F @ 670 psig (610°C @ 46 bar) at Turbine ASTM/SA P91

ASTM/SA P92

34 (864)

34 (864)

1.614 (41.0) (41.0)

1.285 (32.6)

(kg/m) Weight – lb/ft (kg/m)

(830) 558 (830)

481 (716)

(US$/kg) Average Cost – US$/lb (US$/kg)

(13.2) 6.00 (13.2)

7.20 (15.9)

3,348 (10,984) (10,984)

3,463 (11,361)



6-33


7

IMPROVING PLANT EFFICIENCY WITH COAL DRYING To a greater or lesser extent, moisture and non-combustible minerals in coal reduce thermal efficiency and otherwise impact the operation of coal-fired boilers. Although there has been a trend toward minimizing the effort and expense invested in coal processing—both at the mine and the power plant, recent RD&D projects have shown that cost-effective coal drying methods can improve the net heat h eat rate of PC units using high-moisture coal. This heat rate improvement translates into lower air emissions per MWh generated and provides other benefits. Coal fed to the bunkers bunke rs from a power plant’s coal pile may contain moisture from a number of sources, including: •

•

•

Inherent moisture that was present in the organic and mineral sediments that formed the coal (by definition, as noted in Chapter 5, lower rank coals have increasingly higher inherent moisture content) Atmospheric moisture (a relatively mineral-free source) that condensed or precipitated onto, or was absorbed by, the coal during mining, shipping, and storage Surface or process water (which may have high dissolved solids content) that was applied to the coal to help with processing or for dust suppression and fire prevention

Except in very dry climates, c limates, preparation of coal for firing in a pulverized c oal unit includes some amount of drying. This is primarily for the purpose of removing surface moisture that might cause the coal to stick to surfaces in conveyors, chutes, mills, feed nozzles, and other equipment. For coals with high moisture content and/or certain types of mineral content, further drying may be necessary to improve the grindability of the coal. Otherwise the coal may ma y deform or form a paste rather than fracturing into the proper particle p article size. Drying may be necessary to assist ignition and flame stability. Although drying may also improve the thermodynamic efficiency of the generating unit, this alone has not typically been a criterion for specifying the extent or mechanics of drying.

Conventional Coal Drying in Pulverized Coal Units Coal drying is typically accomplished by introducing preheated air or hot furnace gases at (or before) the grinding mills. For high-rank, low-moisture bituminous or anthracite coal, the drying agent in the mill is usually air supplied by the boiler’s FD fan and preheated in the regenerative (Ljungstrom) air heater (see Figure 7-1). 7-1). The mill is pressurized to the extent necessary n ecessary to convey the coal through the mill, classifier, coal flow pipeline, and burner (i.e., in balance with the flow resistance of the coal-laden flow).

7-1


Figure 7-1 Coal Drying and Grinding with Pressurized Preheated Air

For high-moisture subbituminous coal or lignite, which can produce highly reactive dust when dried, it is preferable to withdraw furnace gas from the boiler for use as a drying agent. For this arrangement, the gas pressure in the mill is reduced by an exhauster fan situated between the mill classifier and the burner (see Figure 7-2). The maximum exit gas temperature from the mill depends upon the type of coal and the drying agent (see Table 7-1). Coal drying with furnace gas that is recirculated to the boiler improves the grindability of the coal and may assist its ignition, but it does not increase the plant efficiency.

Figure 7-2 Coal Drying and Grinding with Furnace Gases and Air (Exhauster Mill)

7-2


Table 7-1 Maximum Grinding Mill Exit Temperatures for Different Coal Types

Drying Agent Temperature Furnace gas Air (O2 max =4%)

Coal Type Subbituminous and Lignite Bituminous coal (depending on volatile content)

210°F (100°C)

390°F (200°C)

210–390°F (100–200°C)

390°F (200°C)

Advanced U.S. Coal Drying Technologies EPRI estimates that, in many cases, the benefit of a drying system to reduce the water content of high-moisture coal in terms of improved performance, reduced emissions, and increased availability will significantly exceed the cost of the d rying system, especially for plants burning U.S. lignite and subbituminous coals. Specific benefits may include: •

Increased plant efficiency

•

Increased net generating capacity of units that burn high-moisture coal

•

•

Reduced emissions, particularly from plants based on lignite and Powder River Basin subbituminous coal Increased value of lignite and subbituminous coal reserves

Reducing the total moisture content of lignite by as little as 5% has been shown to have a significant impact on the operational cost and efficiency of a power plant. In addition, as a consequence of improved availability and a lower flow rate of flue gas, net electrical generation capacity will increase. For advanced pulverized coal units, particularly for lignite and subbituminous coals, additional coal drying before combustion is now becoming more common (see following subsections).

Great River Energy Lignite Dryer At its Coal Creek Station in Underwood, North Dakota, Great River Energy captures and reuses unit waste heat to supply hot water and hot air to a fluidized bed lignite dryer (see Figure 7-3), rather than burning additional fuel to generate heat. This method is proving to be a commercially 51 viable means of thermal coal drying.

51

C. Bullinger, M. Ness, and N. Sarunac, “Coal Creek Prototype: Fluidized Bed Coal Dryer,” 31st International Conference on Coal Utilization and Fuel Systems, Clearwater, FL, May 21–25, 2006.

7-3


Figure 7-3 Simplified Schematic of Great River Energy Dryer

Initial results from the Coal Creek Station indicate that the system could be the first to demonstrate that pre-drying high-moisture coals before they are fed to a power plant’s boiler offers a practical and economical way to generate more power from a lower quantity of coal. Until now, however, the cost of thermally drying these coals has often outweighed any potential gains in the plant’s operational performance. Drying the coal increases its heating value, which means that less coal is needed to generate the same amount of energy. Less flue gas is also emitted, which reduces boiler stack gas heat loss and the workload on other equipment in the plant, such as fans. Based on prototype dryer results shown in Figure 7-4 and Table 7-2, if all of the coal fed to the Coal Creek boiler was dried using this process, the result would be an estimated increase in efficiency of about 5%—a very significant improvement in plant performance and costs. Reduction of moisture at Coal Creek is aimed at about 10 percentage points for the wet lignite feed, though the heat source is capable of delivering considerably more drying capacity.

7-4


Figure 7-4 Reduction of Moisture at Great River Energy’s Coal Creek Station

The first dryer is now supplying processed lignite to one of the seven pulverizers for one of the two lignite units at Coal Creek. Early estimates suggest that with just one pulverizer using dried coal, the stack flow rate from the unit decreased 1%, boiler efficiency increased 0.3 percentage points, pulverizer power consumption decreased 4.5%, sulfur oxide emissions fell 2%, oxides of nitrogen emissions decreased 8.5% (because drier coal allowed adjustments to burner air flows that lowered NOX production), and CO2 emissions decreased 0.34% (see Table 7-2).

7-5


Table 7-2 Improved Unit Performance at the Coal Creek Station (With Just One of Seven Pulverizers Receiving Dried Coal)

Parameter

Units

Coal Dryer In Service

Coal Dryer Out of Service

Change

Gross Power Output

MW

589

590

–

Throttle Steam Temperature

deg F

988

989

–

deg C

531

532

–

deg F

1002

1002

–

deg C

539

539

–

klb/hr

46

52

-6.4

kg/s

5.8

6.6

-0.81

klb/hr

953

972

-2.02%

kg/s

120

123

% of Total

14.62

0.00

ksfcm

1611

1626

Nm /s

760

767

kJ/klb

4.09

4.29

J/kg

9.01

9.46

kW

4057

4206

-3.53%

lb/hr

1345

1470

-8.52%

g/s

170

185

lb/hr

3618

3692

g/s

456

465

deg F

353

362

-8.6

deg C

178

183

-4.8

Air Heater 22 Gas Exit Temperature

deg F

368

377

-9.3

deg C

187

192

-5.2

Stack Temperature

deg F

180

184

-4.2

deg C

82

84

-2.3

Reheat Steam Temperature

SHT Spray Flow

Total Coal Flow Rate

Dried Coal Stack Flow Rate

3

Specific Pulverizer Work

Total Pulverizer Power NOX Mass Emissions SOX Mass Emissions Air Heater 21 Gas Exit Temperature

-0.96%

-4.65%

-2.00%

The full set of lignite dryers for both Coal Creek units, scheduled to be built in 2007, will operate through 2008 to generate data that could be applied to other high-moisture, coal-burning power plants that operate primarily in western coal regions of the United States, and in other countries. Subsequently, the dryers are expected to continue operating as part of the power station’s commercial operations.

7-6


Along with Great River Energy, partners in the project include EPRI, Lehigh University (Bethlehem, PA), Barr Engineering (Minneapolis, MN), Heyl & Patterson (Pittsburgh, PA), Headwaters Energy Services (South Jordan, UT), and Falkirk Mining and Couteau Properties (Underwood, ND).

AMAX Coal Dryer The McNally/AMAX coal drying process (which along with the Rosebud process is an example 52 of “higher temperature” processes focused on reduced transportation cost ) was constructed at the Belle Ayr Mine in Gillette, Wyoming (see Figure 7-5). This plant was built for $23 million without any government financing, and initially started in 1990. Of the 100,000 tons (about 91,000 metric tonnes) of upgraded coal produced through 1992, most were shipped to Fremont Utilities Wright Station in Nebraska.

52

Upgraded Low-Rank Coal User Guidelines, EPRI report TR-103221.

7-7


Figure 7-5 AMAX Coal Dryer Schematic

7-8


The raw coal fed at a rate of 215 ton/hour (195 tonne/hr) is initially crushed to a one-inch (25 mm) top size. Because the coal drying rate and its product moisture content are size dependent, coal crushing is required to allow sufficient drying to lower the product moisture to approximately 10% (from 30% in the raw coal). The crushed coal is then passed into a coal dryer where it comes in contact with 900°F (482°C) hot gas for three to four minutes to evaporate the coal moisture, allowing the dried coal product to reach 190°F (88°C). The fine particles are carried in the hot gas stream through a cyclone to remove the coarser fraction of these particles. This coarse fraction is then combined with the larger particles that passed though the dryer. The product stream is then sent to a cooling chamber to reduce the temperature of the final product to 100°F (38°C). Because the drying process causes the coal to break into a much finer, dustier product—creating the tendency of dried coal to reabsorb moisture—the final product is aggregated to make a stable transportable product by adding No. 6 fuel oil (at a rate of 2–4 gallons, or 8–15 liters, per ton). Approximately 165 ton/hour (150 tonne/hr) of dried coal product was produced at a yield of 77% in the AMAX plant. About 40% of the dryer hot gases coming from the cyclone pass into a dryer baghouse, where the remaining entrained fine dust is removed at a rate of about 7 ton/hour (6 tonne/hr). This fine dust is used as fuel for the combustor to create the hot gases used for drying. The combustor is designed for 160 MBtu/hr (170 GJ/hr), using 60% of the dryer chamber hot gases for makeup air. In summary, this coal drying process is carried out at the mine, uses a gas/oil-fired combustor for drying, adds a small amount of oil for transportation, and requires a much higher drying temperature than the GRE process. However, the capital cost is relatively modest and few modifications to the generating plant are needed.

Rosebud Coal Dryer As noted, high-temperature processes generally operate at 400–800°F (200–430°C). One example, the Rosebud SynCoal Advanced Coal Conversion Process (ACCP), has been operating since 1992 at a 300,000 ton per year (270,000 tonne/yr) demonstration plant in Colstrip, Montana (see Figure 7-6). This demonstration plant was built for $69 million, with 50% DOE Clean Coal Technology Program funding. Raw coal fed at 1640 ton/day (1490 tonne/day) is converted into two products—a coarse coal product at 988 ton/day (900 tonne/day) and briquetted coal fines at 240 ton/day (220 tonne/day)—with a total product yield of 75%. Through 1992, the limited production of upgraded coal was used by Montana Power’s Colstrip and Corette generating stations.

7-9


Figure 7-6 Rosebud Coal Dryer Schematic

Coal is initially screened to provide a 1 x 1/4 inch (25 x 6 mm) feed for the process. The minus 1/4 inch (<6 mm) fraction is shipped directly to the adjacent Colstrip power plant. By not processing the fine coal fraction, the fines content of the upgraded SynCoal product should be reduced. The plus 1/4 inch (>6 mm) coal is passed through a two-stage drying process. In the

7-10


first stage, coal is heated to about 220°F (100°C) using recirculated combustion gases to remove the coal surface moisture. In the second stage, the coal is heated to near 800°F (430°C) using superheated steam to remove the water trapped in coal pores and causing decarboxylation. The second stage dryer must be carefully controlled, because residence time and temperature significantly influence final product quality. After the dryers, the coal is cooled to 150°F (65°C) by a combination of quench coolers (adding water) and vibratory coolers (using cool inert gas). Because the local raw coal cannot meet a target limitation of 1.2 lb of SO2 produced per million Btu of fuel heat content input to the boiler (0.52 kg/GJ), the SynCoal process added a coal cleaning step of gravity separators after the product coolers to remove 95% of the pyritic sulfur content, lowering the final product SO2 emissions potential to less than 1.0 lb/MBtu (0.43 kg/GJ). In summary, this is a high-temperature (800°F, or 430°C) coal drying process that has produced a product for two Montana Power stations. It uses multi-stage drying with natural gas as the heat source. Produced close to the mine, it is separate from the generating plants, and results in briquetted coal fines as well as minus quarter inch (<6 mm) coal. A separate coal cleaning process removes a significant amount of sulfur ahead of combustion.

Advanced International Coal Drying Technologies Brown coal (lignite) drying has been explored in Europe, particularly in Germany, and in Australia, where very high moisture lignite deposits are found in Victoria. Two drying schemes of particular note have emerged: the WTA system (a German acronym for “fluidized-bed dryer with integrated waste heat recovery”) and the Mechanical Thermal Expression (MTE) system (Germany/Australia).

Mechanical Thermal Expression Drying System Deregulation of the Australian electricity supply industry has led to an increase in competition between all power generators (hydroelectricity, natural gas, black and brown coal) across the national grid. The need to reduce greenhouse gas emissions has necessitated improvements in brown coal utilization for power generation. The competitiveness of brown coal power generation will in large part depend on the implementation of energy efficient drying technologies for brown coal boilers, which currently burn coal with a moisture content of 60%. Three major drying technologies—Steam Drying (SD), Hydrothermal Dewatering (HTD) and Mechanical Thermal Expression—were investigated in batch and bench scale units. Steam drying is an evaporative drying technology that utilizes superheated steam to remove water from high moisture coals at much lower temperatures than HTD and MTE; however, the metal ash 53 forming elements remain in the final product.

53

David J.Allardice, “The Utilization of Low Rank Coals,” Allardice Consulting, Brian C. Young, Envirosafe International, Eighteenth Annual International Pittsburgh Coal Conference, Dec. 3-11, 2001, Newcastle, New South Wales, Australia.

7-11


A recent option to emerge for brown coal drying is the MTE process, initiated by the University of Dortmund, Germany. In this process (see Figure 7-7), water is expressed from brown coal in a particle-board-type hydraulic press for a few minutes at up to 870 psi (60 bar) pressure, after steam heating the coal to 300–390°F (150–200°C). Some energy can be recovered from the expressed water.

Figure 7-7 Schematic of Mechanical Thermal Expression Coal Drying Process

The MTE dryer from Germany uses pressure to expel moisture from the coal feed. Early studies in Victoria on ambient temperature press dewatering of brown coal were abandoned as impractical because of the high pressures required for residence times of 20 minutes or more. The higher temperature used in MTE decreases the pressure and residence time required to manageable levels. The MTE process can reduce the moisture content of the brown coal to below 25%, with very little energy consumption and a substantial reduction in greenhouse gas emissions from even conventional brown coal power generation. MTE demonstrated the process at 11 ton/hr (10 tonne/hr) scale at Rheinbraun and is now the preferred drying process of the Australian Cooperative Research Centre (CRC) for Clean Power from Lignite and the focus of a significant proportion of the CRC’s current research program. The CRC for Clean Power from Lignite is developing MTE as a practical concept for retrofitting to existing boilers or to pre-dry the feed coal for an IGCC plant. CRC has concluded that MTE is less expensive and provides greater efficiency improvements in these applications than HTD or Steam Fluid Bed Drying. On this basis, CRC is planning a Latrobe Valley demonstration plant for the technology using additional funds from the state government.

7-12


It is important to note that any dewatering or drying process that can remove water from brown coal in liquid form (i.e., without the need to supply the evaporative energy to dry the coal) has the potential to reduce, by up to 25%, the greenhouse gas emissions arising from the combustion of these coals.

RWE WTA Fluidized-Bed Dryer Developed by Rheinbraun (now RWE Power), the WTA fluidized-bed dryer has been tested since 1993. The first prototype in Frechen, Germany, operated for over 13,000 hours, and dried 54 55,000 tons (50,000 tonnes) of raw brown coal. Since 1999, a new prototype, based on the finegrain WTA technology, has demonstrated improved capital and operating costs. In 2007, this technology will be implemented at commercial scale, when one initial unit of about 120 ton/hour (110 tonne/hr) dry lignite throughput will be installed and connected to the Niederaussem power plant. The WTA process schematic is shown in Figure 7-8. Brown coal is inserted into the dryer following pre-heating at 150°F (65°C) within a heat exchanger using condensed water from the drying process of the previous charge. It is fluidized at 210°F (100°C) under the influence of slightly superheated steam. Much of the heat needed for moisture evaporation is provided by pressurized steam from the drying of the previous charge in a heat exchanger. The steam, which is used for the fluidization, also contributes to the evaporation of fuel moisture. After being cleaned in an E-filter, the evaporated moisture is fed into a steam compressor, where its temperature is increased from 210°F (100°C) to 300°F (150°C) by pressuring it in multiple steps up to 60–75 psi (4–5 bar). A small part of this evaporated moisture is used as a fluidization medium. The condensed water from the heat exchanger is used for cooling the steam compressor as well as for pre-heating of raw brown coal. The residence time of the brown coal in the dryer depends on the moisture content and is typically 60–90 minutes.

54

Prof. Emm. Kakaras et al., Concept study for a 700°C power plant: using poor quality brown coal with ultra supercritical PF boiler , National Technical University of Athens, VGB contract, Final Report, Oct. 2006.

7-13


Figure 7-8 WTA Dryer Schematic with Sample Flow Calculations

As an alternative to this “close cycle” concept where the evaporated steam from the coal is reutilized as a heating medium for the drying process after being compressed, an “open cycle” concept has been developed. It avoids the installation of a compressor by the use of lowtemperature steam extracted from the low-pressure part of the steam turbine. The evaporated steam from the coal is afterward utilized at the first water pre-heater step of the power plant. The “open cycle” concept was chosen to be implemented in the first WTA prototype, which will be integrated in the Niederaussem power plant. The lack of a mechanical compression step, with its own difficulties and failure characteristics, simplifies the overall design and improves the expected availability of the plant.

7-14


8

AIR EMISSIONS CONTROL Environmental Regulations Prior to the start of construction, developers of new power plants are required to obtain air permits to ensure that emissions do not significantly impact ambient air quality or visibility standards. The U.S. EPA’s New Source Review (NSR) process includes an analysis of air quality and visibility impacts, along with a determination of ne cessary emission control requirements. At a minimum, emissions from new generation plants will be required to meet applicable New Source Performance Standards (NSPS). In addition to meeting federal NSPS, a new power plant in the United States may be required to meet more stringent state-imposed emission limits. Typically, this entails working with the appropriate state and/or local environmental control agencies to determine how federal and state emission regulations will be applied within the specific air qua lity control jurisdiction where the new plant will be located. If emissions from a new or modified source are expected to cause local air quality to exceed Prevention of Significant Deterioration (PSD) thresholds, the facility will most likely be required to implement Best Available Control Technology (BACT), which generally is the technology that has been determined to be most technically and economically feasible for limiting emissions from that type of facility. Even more stringent permit requirements may be applied for projects that will significantly impact ozone nonattainment areas or federal “Class I” areas (national parks, wilderness areas, etc.). The necessity and scope of more stringent requirements are determined on a project-specific basis. A BACT analysis for an individual project considers three primary factors in determining the required emission limits: 1) Historic BACT determinations for similar pulverized coal steam plants 2) The technical feasibility of available emission controls 3) The cost-effectiveness of each control option The inclusion of both technical and economic criteria is a fundamental characteristic of BACT analysis. Rather than prescribing specific technology, this approach gives a power producer the ability to work with the permitting agency to determine the overall technology to be selected and the emission rate that is applicable using that technology. 40CFR60, Subpart D, for electric utility steam generating units, including pulverized coal steam plant units, is the applicable regulation for the size of units addressed in this Guideline. The U.S. EPA definition of BACT is contained in the following excerpt: “…an emissions limitation (including a visible emission standard) based on the maximum degree of reduction for each pollutant subject to regulation under the Clean Air Act which would be emitted from any proposed major stationary source 8-1


or major modification which the Administrator, on a case-by-case basis, taking into account energy, environmental, and economic impacts and other costs, determines is achievable for such source or modification through application of production processes or available methods, systems, and techniques, including fuel cleaning or treatment or innovative fuel combustion techniques for control of such pollutant. In no event shall application of best available control technology result in emissions of any pollutant which would exceed the emissions allowed by any applicable standard under 40 CFR Parts 60 and 61. If the Administrator determines that technological or economic limitations on the application of the measurement technology to a particular emissions unit would make the imposition of an emissions standard infeasible, a design, equipment, work practice, operational standard, or combination thereof, may be prescribed instead to satisfy the requirement for the application of best available control technology.” Table 8-1 lists the input- and output-based limits contained in the NSPS for facilities that began construction after February 28, 2005. For comparison purposes, the output-based limits in lb/MWh (based on gross output) have been converted to the more familiar input-based limits in pounds per million British Thermal Units (lb/MBtu). Table 8-1 Emission Limits from the Latest Revision to 40CFR60, Subpart D Pollutant

NSPS Limit on Gross Energy Output Basis

NSPS Limit on Heat Input (1) Basis

Particulate Matter (PM)

0.14 lb/MBtu

0.015 lb/MBtu

Sulfur Dioxide (SO 2)

1.4 lb/MWh

0.15 lb/MBtu

Oxides of Nitrogen (NOX)

1.0 lb/MWh

0.11 lb/MBtu

Mercury (Hg)

20 x 10 lb/MWh

-6

(2)

(3)

2.15 lb/TBtu

Notes: 1. Based on estimated gross heat rate of 9300 Btu/kWh calculated from PM emission limits 2. For coal-fired electric utility steam generating unit that burns only bituminous coal 3. lb/TBtu – pounds per trillion Btu

The “gross heat rate” requirement presents a challenge when comparing different technologies, but is required by the EPA. Although the plant output is sold on a net megawatt output basis, and emissions have been more typically calculated on a gross input basis, the NSPS for pulverized coal steam plant facilities are based on gross energy (generation) output. To effectively choose between different technologies with the same net output, the power plant developer will need to use all of these different bases for evaluating emissions, determining the required emission control performance, and determining the cost and feasibility of purchasing emissions offsets, if needed, to install a desired block of new generation in a specific location. The following explanatory notes may also be helpful for understanding emissions control performance requirements applicable to PC plants: •

NOX – steam generating unit stack emissions are limited to 1.0 lb/MWh (0.45 kg/MWh) or lower in the flue gas based on the use of SCR.

8-2


•

•

SO2 – steam generating unit stack emission levels are limited to 1.4 lb/MWh (0.63 kg/MWh) or lower in the flue gas based on the use of FGD. Particulate Matter (PM) – Filterable PM, without condensables, collected from the front half of sampling train, is limited to 0.14 lb/MWh (0.063 kg/MWh) or lower. – PM10 (e.g., particulate matter 10 microns or less in size) is assumed to be 0.03 lb/MBtu (0.013 kg/GJ) or less when condensables are included from back half of the sampling train, emission rate. Additional PM10 may result from the use of SCR, which is likely to produce additional ammonium bisulfate and sulfate particles.

•

CO – the current BACT assumes no CO catalyst is used

•

VOC – the current BACT assumes no CO catalyst is used

•

• •

-6

Mercury – steam generating unit stack emission levels are limited to 20 x 10 lb/MWh (or 9 mg/MWh) lower based on cumulative performance achievable using SCR, ESP or fabric filter, and FGD. Ammonia Slip (from SCR) – current BACT is 2 ppmv (parts per million by volume). Offset Credits – for some plant locations, it will not be possible to permit a plant without offsetting emission through purchase of credits that are created by the reduction of emissions from other sources. In some cases, the high cost or limited availability of credits will make it cost effective to use emissions performance design criteria that are muc h more stringent than indicates by specific numbers in the NSPS. This may further be influenced by the ability to reduce the quantity of one type of credit by purchasing another type, typically with a multiplier, or to purchase a larger quantity of credits from a more distant source.

Annual Emissions Power producers will generally need to complete charts similar to that shown as Table 8-2 to support their environmental permit application.

8-3


Table 8-2 Worksheet for Expected Annual Air Emissions for a PC Plant

Source

Annual Operation

hr/y

SO2

ton/y

lb/ MBtu

NOX

lb/ MWh

ton/y

Permanent Sources Boiler Coal Handling Ash Handling SCR Reagent Handling FGD Reagent Handling FGD By-product Handling Fugitive Emissions Cooling Tower Startup Sources Auxiliary Boiler Intermittent and Upset Sources Emergency Diesel Generator(s) Emergency Firewater Pump(s) ANNUAL TOTAL

8-4

lb/ MBtu

PM filterable

lb/ MWh

ton/y

lb/ MBtu

lb/ MWh

PM condensable

ton/y

lb/ MBtu

lb/ MWh


Sulfuric Acid Mist

Heavy Metals

CO

Source

VOC ton/y

lb/ MBtu

lb/ MWh

ton/y

lb/ MBtu

lb/ MWh

ton/y

lb/ MBtu

lb/ MWh

Halogens

Hg

Pb

Other

Fl

lb/y

lb/y

lb/y

lb/y

Cl lb/y

Permanent Sources Boiler Coal Handling Ash Handling SCR Reagent Handling FGD Reagent Handling FGD By-product Handling Fugitive Emissions Cooling Tower Startup Sources Auxiliary Boiler Intermittent and Upset Sources Emergency Diesel Generator(s) Emergency Firewater Pump(s) ANNUAL TOTAL Notes: lb/MWh based on gross MW output lb/MBtu based on fuel HHV thermal input to the boiler Hours of operation consistent with those to support 8760 hours per year of operation of the boiler. Fugitive emission may include, but not limited to: wind erosion of coal storage piles, limestone storage piles, FGD by-product storage piles, truck traffic on paved and unpaved facility roads. Specific emission sources can vary by project and pulverized coal steam plant technology. Some of the sources or streams listed above may be combined and vented to a common control device, or recycled in the process.

8-5


9

WET FGD SYSTEMS FOR SO 2 CONTROL The majority of wet flue gas desulfurization systems installed at coal-fired power plan ts utilize alkaline slurry to remove SO2 contained in the flue gas. The alkaline slurry is forced into close contact with SO2 laden flue gas, in an absorber vessel, where the SO2 is absorbed to the liquid phase and reacts with dissolved alkaline reagent. The reagents used in wet FGD systems include limestone, lime, enhanced lime, ammonia, magnesium oxide, and sodium carbonate. Limestone based systems, which account for approximately 85% of the installations, are used as the basis for the bulk of this write-up. Current systems have demonstrated SO 2 removal efficiencies approaching 98%. Systems based on limestone with forced oxidation (LSFO) and magnesiumenhanced lime (MEL) with forced oxidation also generate gypsum by-product that can be sold for use by several industries. The absorber may take various forms (open spray tower, tray absorber, jet bubbling reactor, and double-contact flow absorber), depending on the manufacturer and desired process configuration. The absorber application with the most U.S. operating experience is the counter-flow, vertically oriented, open spray tower.

Equipment and Process for Limestone-Based Open Spray System For most installations, the vertically oriented, counterflow, spray tower configuration, using a limestone slurry reagent, has been found to achieve the necessary SO2 removal performance with the best balance of capital and operating cost. The desired process endpoint is accomplished through a complex series of time-dependent reactions that occur in the gas, liquid, and solid phases of the absorption and post-absorption treatment processes. A vertical flow spray tower consists of an inlet nozzle, a spray tower with an integral reaction tank located in the base of the tower, slurry spray headers and nozzles, mist eliminators, mist eliminator wash system, and optional perforated trays (see Figure 9-1, Figure 9-2, and Figure 9-3). The hot flue gas (250–350°F, 120–180°C) enters the lower portion of the tower through the inlet nozzle and flows upward through multiple levels of interspatial spray nozzles operating at pressures of between 5 and 20 psig (0.3–1.4 barg). The flue gas flow is countercurrent to the flow of scrubbing slurry, which is distributed throughout the tower as a fine spray by specially designed nozzles. The flue gas entering the absorber is first cooled through adiabatic saturation as it makes contact with SO2-rich slurry. SO2 is progressively absorbed as the flue gas passes each bank of slurry spray nozzles, with the most of the absorption occurring in close proximity to the nozzles. On exiting the absorber, the flue gas passes through mist eliminators which capture fine particles and spray that are carried by the flue gas.

9-1


Figure 9-1 55 Wet FGD Spray Tower Configuration

The slurry containing the absorbed SO2 falls into the reaction tank located at the base of the spray tower. By the time the spray falls reaches the reaction tank, much of the SO2 has combined with CaOH and water to form hydrated calcium sulfite ( CaSO3•½H2O). In LSFO systems, oxidation air blowers introduce compressed air into the reaction tank where oxygen completes the scrubbing reactions and oxidizes the calcium sulfite to form gypsum (hydrated calcium sulfate— CaSO4•2H2O). Limestone slurry, from a limestone preparation system, is added to the reaction tank to neutralize and regenerate the scrubbing slurry. A slurry recycle system recirculates the regenerated slurry from the reaction tank to the spray nozzles. The oxidation air system consists of air compressors, valves, transport piping, air lances, and agitators. The air is transported to the reaction tank by the transport piping and is introduced into the reaction tank through multiple air lances installed in the reaction tank. The air lances are positioned in front of horizontal agitators that ensure the air is mixed well with the slurry, as well as keeping the slurry solids in suspension.

55

Courtesy of Babcock & Wilcox.

9-2


A gypsum bleed system removes the appropriate amount of slurry to maintain process equilibrium and transports this slurry along with waste solids (sludge) to the gypsum processing area. The absorber module is typically equipped with a vertical flow, two-stage, chevron-type mist eliminator system that uses inertial contact to remove carryover mist from the cleaned flue gas. The primary stage captures large particles and droplets carried over from the absorber; the secondary stage captures wash-water droplets and finer particles. The mist eliminators are kept free of slurry deposits by using a water wash system, which directs process makeup water through spray nozzles to the upstream and downstream faces of the first stage mist eliminators and the upstream face of the second stage mist eliminators.

Gypsum Processing Figure 9-3 outlines a typical gypsum processing system, which includes primary and secondary dewatering and wastewater feed systems. The dewatering system, which is typically a common system for all units at a plant, is designed to continuously receive slurry from the absorbers and produce gypsum conforming to the end user’s needs. Primary dewatering is accomplished with hydrocyclones whereas a vacuum-belt-filter system is used for secondary dewatering. System blowdown is removed as a sidestream from the hydrocyclone overflow and processed through a secondary (wastewater) hydrocyclone before discharge to the wastewater treatment system. Gypsum bleed pumps feed the gypsum slurry to the multi-element dewatering hydrocyclones, which remove a first cut of filtrate from the suspended gypsum solids. A valve network is used to adjust performance by isolating flow to individual cyclones. The thickened hydrocyclone underflow is routed to a filter feed tank. The hydrocyclone overflow is transferred to the filtrate tank where it is returned to the process for reuse. A blowdown sidestream, prior to the filtrate tank, is directed to a wastewater hydrocyclone feed tank. The chloride level in the absorber is maintained below its design maximum at all times through adjustment of this blowdown, the gypsum slurry bleed rate, the settings on the wastewater hydrocyclones (underflow is returned to the filter feed tank), and makeup water flow.

9-3


Figure 9-2 Schematic of Typical Wet Flue Gas Desulfurization System—Absorber and Reagent Mixing

9-4


Figure 9-3 Schematic of Typical Wet Flue Gas Desulfurization System—Gypsum Processing System

9-5


Filter feed pumps deliver thickened slurry to the vacuum belt filter trains, where cake formation occurs very quickly after vacuum is applied and gypsum is dewatered to meet end user requirements. Makeup water is used to wash the filter cloth and to wash gypsum cake to displace dissolved contaminants as the cake is being dewatered. The source of cake wash water may be either recycled filter cloth wash water or makeup water. The filtration cycle ends when the vacuum is released and the filter cloth and dewatered cake pass over a small radius discharge roll that separates the dewatered cake from the cloth. The cake falls from the filter cloth into the discharge chute which directs it to the gypsum handling system for transport to the final destination. The wastewater hydrocyclone feed tank (not shown) receives the chlorides bleed (blowdown) streams from each of the gypsum dewatering hydrocyclones or from the wastewater bleed pumps and discharges it to the wastewater hydrocyclone feed pumps. Solids in this stream are recovered via the underflow from the wastewater hydrocyclone cluster; the underflow is pumped back to the filter feed tank by the dewatering area sump pumps. The overflow is pumped to the FGD wastewater treatment system. Variations on the gypsum dewatering system design are typically specific to the client and sometimes are site specific. Drum filters may be substituted for the vacuum b elt filters for applications requiring less moisture removal from the gypsum. Other gypsum dewatering systems use vertical basket centrifuges instead of vacuum belt filters. Some co mpanies are now pumping the gypsum bleed slurry directly to the vacuum belt filter, eliminating the hydrocyclones. Optimal system design requires an evaluation of end user requirements and sitespecific issues.

Limestone Preparation System The limestone preparation system is normally installed as a common system for all units at a plant. Dry limestone is fed from a storage silo onto a weigh belt filter, which controls the feed of the limestone into the ball mill. Water is added with the limestone to assist in grinding and to slurry the limestone. Figure 9-4 shows a grinding system of the type used for most wet FGD applications. This system uses a horizontal, wet overflow discharge type mill with steel grinding balls. The tumbling motion of the balls pulverizes the limestone as the mill rotates at low speed. The ball mill produces slurry containing 60–65% solids, which is then discharged to the mill slurry tank. Additional water is added to this tank to produce the 30% solids slurry used in the FGD system. Feed pumps transport slurry from the mill slurry sump to hydrocyclone classifiers, where centrifugal force separates coarse particles from the slurry and gravity flow returns them to the ball mill for further grinding. The slurry containing finer particles exits the hyd rocyclone overflow with sufficient head for delivery to the reagent storage tank.

9-6


Figure 9-4 Typical General Arrangement for Wet Limestone Grinding Systems

Alternative Designs Lime-Based FGD Systems The lime process is the second most common wet FGD system utilized in the United States. This process uses hydrated lime slurry in a countercurrent spray tower. Most of these systems use a magnesium-enhanced lime (MEL) process where the reagent is magnesium-enhanced lime (typically 5–8% magnesium oxide) or dolomitic lime (typically 20% magnesium oxide). Although lime is a more costly reagent than limestone, MEL is also much more reactive and can achieve high SO2 removal efficiencies in significantly smaller absorber towers and with a much lower liquid-to-gas ratio than is required for limestone scrubbers. The gypsum processing system used with MEL is more challenging to design and operate than for limestone slurry processes because the waste solids from MEL are more difficult to dewater. As with limestone-based FGD, forced oxidation can be used in MEL to improve the quality of the solids to produce commercial grade gypsum. However, this requires a separate tank external to the absorber.

9-7


Lime Preparation In a process called slaking, quicklime or a lime/magnesium oxide mixture is added to water, creating a vigorous reaction with the release of heat, and obtaining calcium hydroxide or a mixture of calcium hydroxide [Ca(OH)2] and magnesium hydroxide [Mg(OH)2] as the end product for use as a reagent for the FGD system. The lime slaking system is normally installed as a common system for all units at a plant. The lime/magnesium oxide mixture is fed from a storage silo onto a weigh belt feeder, which controls the feed of the lime into the lime slaker. The end product from the slaker is a slurry, which is pumped to lime slurry storage tanks from which it is added to the FGD system as required. Several types of slakers are available: •

Detention or slurry slakers

•

Paste slakers

•

Batch slakers

•

Horizontal ball mill slakers

•

Vertical ball mill slakers

All of these have been used in utility applications to process the lime slurry for wet FGD systems.

MEL Process

56

In the absorber, most of the SO 2 reacts with calcium hydroxide to precipitate calcium sulfite (CaSO3•½H2O). The portion of SO2 that reacts with magnesium hydroxide forms soluble salts as wells as insoluble magnesium sulfite and magnesium bisulfite [MgSO3, Mg(HSO3)2]. The formation of these soluble salts allows the scrubbing action to proceed at a much more rapid rate than is possible with less soluble calcium reagents. Magnesium sulfite also provides a buffering action, which prevents a sharp decrease in pH as the quantity of absorbed SO2 increases, and helps reagent utilization approach 100%.

Gypsum Processing As noted, limestone with forced oxidation (LSFO) systems oxidize hydrated calcium sulfite to hydrated calcium sulfate (gypsum) in the reaction tank. This process is called in situ oxidation. Due to process limitations, the MEL-based FGD system product canno t be oxidized in situ. Oxidation is accomplished with ex-situ oxidation using compressed a ir in an exterior bubble tower. When bleed slurry is forced into contact with compressed air in the external forced oxidation bubble tower, hydrated calcium sulfite (CaSO3•½H2O) is converted to gypsum (hydrated calcium sulfate—CaSO4 •2H2O), and magnesium sulfite is oxidized to magnesium sulfate. The gypsum precipitates while magnesium sulfate remains in solution. Liquid containing magn esium sulfate is returned to the oxidation tower from the gypsum dewatering area. A portion of magnesium sulfate is recirculated to the absorber where it replenishes magnesium sulfite, which buffers the 56

Wet Flue Gas Desulfurization Technology Evaluation, PN 11311-000, prepared by Sargent and Lundy for National Lime Association, January 2003.

9-8


MEL reaction. The gypsum slurry bleed exiting the oxidation tower is pumped to a typical gypsum processing system as described above.

Other Wet FGD Technologies The following sections present a general discussion of the unique design features for the systems offered by suppliers. It is not meant as a detailed description of their offerings.

Jet Bubbling Reactor The Chiyoda CT-121 FGD system utilizes a jet bubbling reactor (JBR) to remove the SO 2 contained in the flue gas. This process utilizes fan pressure to force the flue gas into the reaction tank containing slurry (see Figure 9-5). An open spray tower utilizes pumps to allow contact between the flue gas and slurry. The Chiyoda CT-121 FGD system consists of two sections: gas cooling and SO2 removal. The flue gas enters the inlet gas cooling section where it is cooled and saturated with a mixture of makeup water and process slurry. The flue gas then enters an enclosed plenum chamber in the JBR formed by the upper deck plate and lower deck plate. Sparger tube openings in the floor of the inlet plenum force the inlet flue gas below the level of the slurry reservoir in the jet bubbling zone (froth zone) of the JBR. After bubbling through the slurry, where the reactions occur, the gas flows upward through large gas riser tubes that bypass the inlet plenum. Entrained moisture and slurry in the cleaned gas disengages in the outlet plenum, located above the upper deck plate, due to a drastic velocity reduction. The cleaned gas passes to the two-stage, horizontal-flow mist eliminator before entering the chimney.

9-9


Figure 9-5 Schematic of Jet Bubbling Reactor Internals

Dual Contact Absorber Advatech offers a Double Contact Flow (DCF) wet FGD system, with two design options available. In the first design, flue gas enters the lower portion of the absorber horizontally, turns and flows vertically upward through fountain sprays where SO2 is removed. The fountain sprays consist of a single layer of multiple spray headers located immediately above the inlet nozzle. Nozzles located on the headers spray fine droplets of slurry upward in a fountain-type spray. The spray initially is co-current to the flue gas flow, then becomes countercurrent as the slurry droplets reach their pinnacle and are drawn downward by gravity. The fresh slurry droplets first contact flue gas with reduced SO2 concentration, as they travel upward, then pass through flue gas with increasingly higher concentration, as they fall downward, eventually passing below the level of the spray nozzles and the tower inlet nozzle and falling into the reaction basin. This trajectory promotes good gas-liquid contact and extends its duration to enhance SO2 removal efficiency. The second design uses a horseshoe-shaped absorber and a two-step process to remove SO2 from the flue gas. The flue gas enters the upper portion of the absorber, flows vertically downward where SO2 is removed. The flue gas then flows horizontally around the bend, and vertically upward, where additional SO2 is removed. Water spray is used to quench the flue gas as it enters the top of the absorber. Fountain sprays, with a single layer of multiple spray headers, are installed in the lower portion of the first stage of the absorber. As nozzles located on the headers spray fine droplets of slurry upward in a fountain-type spray, the slurry first has countercurrent contact with the downward flowing flue gas. After the slurry reaches its pinnacle, it falls by gravity and contacts the flue gas again, this time co-currently.

9-10


The second stage of the absorber acts like the first DCF design, with fountain sprays contacting flue gas co-currently at first, then reversing direction and sustaining countercurrent contact with the upward-flowing flue gas. The two “double passes” of fresh slurry through the flue gas enhance SO2 removal efficiency The internal headers are supplied slurry by a single header with multiple recycle pumps feeding the header. This allows various combinations of recycle pumps to be used, depending on the boiler load and sulfur content in the fuel. The cleaned flue gas is then discharged to the outlet duct through horizontal flow mist eliminators.

Alstom Alstom offers a patented design feature called performance enhancement plates (PEPs), or wall rings, for enhancing the performance of the open spray tower absorbers. These radial baffles project inward from the spray tower wall to redirect flue gas back into the spray zone and thereby prevent it from traveling up the wall and bypassing the spray nozzles. PEPs can reduce capital and operating costs by improving SO2 removal efficiency and/or by reducing liquid-togas ratio. Alstom also offers the Flowpac Absorber, which is capable of achieving high SO 2 removal efficiency without the use of the large slurry recycle pumps required by other absorber types. The Flowpac Absorber takes advantage of the density difference between aerated and nonaerated limestone slurry. Oxidation air is used to ex pand the volume of the slurry in the reaction tank, allowing it to flow up and over the top of a sieve plate. Flue gas is introduced below the sieve plate and flows upward through the frothy layer of limestone slurry on the plate. The gas and slurry come into intimate contact, resulting in high SO 2 removal efficiency. The slurry travels across the sieve plate and is discharged from the tray over a weir. The de-aerated slurry in the reaction tank quiescent zone descends and can be recirculated in the air-lift zone.

Babcock & Wilcox (B&W) Babcock & Wilcox (B&W) offers a tray tower design to enhance the SO2 capabilities of its wet FGD system. The flue gas flows upward through the absorber where it is quenched by absorber slurry from the undertray spray headers and by scrubbing slurry falling from the perforated absorber tray. The quenched flue gas is then forced into contact with slurry as it passes through the perforated absorber tray. The tray is sectioned into compartments by baffles to promote a more even distribution of liquid on top of the tray and more even flow through the tray. Above the tray, the flue gas encounters the upper spray zone where it flows counter currently to the absorber slurry from the upper level absorber spray headers. B&W offers several performance enhancements as part of their design. One set of enhancements has allowed the tray tower to operate with higher superficial flue gas velocity—15 ft/sec (4.6 m/s) instead of the conventional 10 ft/sec (3.0 m/s). This has enabled a reduction in tray tower diameter, which reduces the capital costs for the absorber. The higher tower gas velocity was obtained by optimizing the B&W tray to uniformly distribute the tower gas flow and by utilizing proper high-velocity mist eliminators.

9-11


B&W has also reduced capital costs by utilizing a patented interspatial spray header design to shorten tray tower height. This design doubles the number of spray nozzles at each level, and reduces the number of spray levels, by locating the main header outside the tower and penetrating the tower with branches at several locations around the tower perimeter. Another design enhancement modifies the tower inlet by redesigning the awning that controls the formation of solids buildup at the wet/dry interface in the absorber inlet. The awning directs the contact between the gas and the liquid into a location in the middle of the tower away from the inlet flue.

Babcock Power Environmental Inc. (BPEI) BPEI offers dual flow up/down slurry nozzles in the spray tower. Essentially, each level of spray header can provide twice the limestone slurry to the flue gas. This approach minimizes nozzles, headers, associated pumps, and power consumption, which, in turn, reduces capital and operating costs. BPEI can include internal wall baffles to prevent untreated flue gas from flowing up the walls. These rings are located at multiple elevations in the absorber to create a physical barrier that drives the untreated flue gas into the slurry spray zone for effective treatment. BPEI also offers a dual loop system, which couples the scrubber with quench, oxidation and absorption, absorbent dosing, and gypsum slurry bleed. The flue gas enters the first loop (called the quench loop) from the bottom and passes the quench spray bank level into the second loop (called the absorber loop). This loop absorbs most of the SO2. The cleaned flue gas exits the scrubber at the top after passing through the mist eliminators. An intermediate collection bowl separates the two loops and collects the scrubbing liquid from the absorber loop. A return pipe connects the collection bowl with the external sump tank. BPEI cites the following as main benefits of dual loop system: •

•

Superior performance for removal of higher SO2 concentrations from the flue gas. Testing has demonstrated that when burning high-sulfur coal and/or with the requirement of SO2 control efficiencies of greater than 97%, the d ual loop system is sometimes superior to a single loop system. The system can use lower quality limestone and limestone with larger particle size. Higher pH value in the absorber loop allows use of a lower total L/G than in a single loop system, resulting in a reduction in power con sumption. When high SO2 control efficiency is required, the lower operational cost more than compensates for the additional cost of an external sump tank.

Ammonia FGD System The ammonia-based wet FGD system utilizes ammonia as the reagent to remove SO2 from the flue gas. The system generates ammonium sulfate, a fertilizer, as a product. The conventional ammonia FGD system utilizes the same basic design as the limestone-based FGD systems. Flue gas enters the vertical counterflow spray tower through an inlet nozzle above the reaction tank located in the base of the tower. The design of the slurry spray headers and nozzles, mist eliminators, and mist eliminator wash system is similar to LSFO designs. Hot flue gas enters the lower portion of the tower and flows upward through multiple levels of spray 9-12


headers and associated nozzles. Flue gas flow is countercurrent to the flow of scrubbing slurry, which is sprayed downward by the nozzles. The flue gas is cooled through the adiabatic saturation temperature as it makes contact with the scrubbing slurry. Ammonium sulfite and ammonium sulfate are formed as SO2 is absorbed into the scrubbing slurry. The ammonium sulfate partially crystallizes as water is evaporated from saturated ammonium sulfate solution before it falls into the reaction tank at the base of the spray tower. Oxidation air blowers introduce compressed air into the reaction tank to complete the scrubbing reactions and oxidize the ammonium sulfite to ammonium sulfate. In some systems, the oxidation air also serves as a vaporizer and carrier for makeup ammonia, which is injected into the air transport piping, in anhydrous or aqueous form, upstream of the distribution control valves. In other systems, aqueous ammonia, or vaporized anhydrous ammonia, is introduced to the reaction tank in a separate stream. The oxidation air system consists of air compressors, valves, transport piping, air lances, and agitators. The air is transported to the reaction tank by the transport piping and introduced into the reaction tank through multiple air lances installed in the reaction tank. The air lances are positioned in front of horizontal agitators. The agitators ensure that the air and ammonia are mixed well with the slurry and that the solids in the reaction tank remain in suspension. Recycle pumps take suction from the reaction tank and pump the refreshed scrubbing slurry to the headers and nozzles in the spray tower. A bleed system removes the appropriate amount of ammonium sulfate slurry to maintain process equilibrium and transport the slurry to the processing area. The bleed system may take a sidestream from the recirculation piping or may use separate pumps that take suction from the reaction tank. The absorber module is typically equipped with a vertical flow two-stage mist eliminator system that removes carryover mist by inertial contact. The primary stage of the mist eliminator captures large particles and droplets. The secondary stage captures wash water droplets and finer particles. A water wash system flushes the mist eliminators to prevent buildup of slurry deposits. Makeup water is directed to the upstream and downstream faces of the first stage mist eliminators by spray nozzles and the upstream face of the second stage mist eliminators.

Ammonia Supply System Makeup ammonia is supplied to a wet FGD system by a storage and pumping system that consists of an ammonia storage tank farm, ammonia forwarding pumps, and associated piping, valves, and controls. Anhydrous ammonia is stored under pressure in multiple tanks. Forwarding pumps deliver the ammonia from the storage tanks to the reaction tank by either a controlled injection to the oxidation air system, or through a vaporizer to a separate injection point. If aqueous ammonia is used, it may be injected into the oxidation air system or pumped directly to the reaction tank.

Ammonium Sulfate Dewatering The dewatering system for the ammonia FGD system is slightly different than the limestonebased FGD system. The bleed slurry is fed to a set of dewatering hydrocyclones that increases the slurry density from approximately 10% by weight solids to 35% solids. The higher density

9-13


slurry (underflow) is directed to a centrifuge feed tank. The lower density slurry (overflow) is directed back to the absorber. The slurry from the centrifuge feed tank is pumped to vertical basket centrifuges where additional water is removed. The 35% by weight solids is increased to 95–97% by weight solids. The centrifuges then discharge the material into a rotary drum dryer where the material is dried to less than 1% moisture. The material is then transported to a storage silo. Depending on the objectives of the end user, the ammonium sulfate product may be sold without further processing, or may require additional processing. If the product can be sold as is, the storage silo will be equipped with truck loading facilities. If further processing is required, the product will be fed to a compaction system, where it is mixed with recycled materials, converted from sugar-like crystals to larger granular crystals, and fed to the co mpactors. The compactors compress the material at a high force, producing a large “flake” of ammonium sulfate. The flake is then ground in sizing mills and fed to a sizing screen, which separates the on-size material from the fines. The fines are recycled back to the feed of the compaction process. The on-size product passes through a dryer and then a cooler to ensure that the product meets industry quality requirements.

Current Design Issues

Higher Removal Efficiencies As SO2 emission limits become more stringent, the demand for increased removal efficiencies grows. Although SO2 removal efficiencies of 95–97% were the norm in the early 2000s, current wet FGD systems are being designed for 98% SO2 removal efficiency, and pressure to achieve guaranteed higher efficiencies as quickly as possible is mounting. FGD systems must continue to meet industry standards for reliability and availability, even when higher removal efficiencies place new demands on the systems. Several approaches are being used to increase efficiency from new and existing absorbers. These include: •

•

Increasing contact between the reagent and the SO2-laden flue gas (several methods are described below) Modifying the slurry with chemical additives that increase the SO2 absorption rate and capacity (see the subsection on additives)

Increased contact between the reagent and the SO2-laden flue gas can be accomplished in several ways, each with a different impact on capital and/or operating costs. These include: •

Adding an extra spray level in a vertical spray tower

•

Using higher pressure spray nozzles to decrease droplet size

•

Installing wall rings to prevent untreated flue gas from byp assing the spray nozzles and to reentrain slurry that is flowing down the walls of the absorber

A typical spray tower, designed for 98% SO2 removal efficiency, has a fixed number of operating spray headers with one available as a spare, to be used if there is a problem with plugging, excessive nozzle wear or slurry reactivity, or to catch-up after a system upset to maintain annual

9-14


average emissions. If the system were designed for a higher SO2 removal efficiency, it might have additional operating spray headers. The number of operating headers is based on the amount of sulfur in the coal and the percent removal required. Adding a spray header increases capital costs, not only for the additional header, but for a new pump, piping, and valves. Further capital costs would be incurred for increasing the absorber height and upgrading structural supports, foundations, and electrical equipment. There would be a small percentage increase in the cost of reagent. Electric power consumption would increase by a greater percentage, due both to increased flow rate and to the head requirement for the header added at higher elevation. Higher pressure spray header nozzles are used to decrease the slurry droplet size and thereby increase the surface area in contact with the flue gas. Although this method alone does not achieve the total increase in the SO2 removal efficiency, it is used as a step to achieve increased removal efficiency. Although SO2 removal efficiency is increased by the smaller droplet size, the smaller droplets are more easily entrained in the gas flow. The mist eliminators are more likely to plug because the volume of entrained solids has increased while the eliminators are being designed with smaller passages and sharper direction changes to trap the finer droplets and particles. Increased operating costs result from increased pump head requirements and, potentially, from increased spray nozzle wear rates. As noted in the Alstom and BPEI subsections, installation of wall rings can enhance the performance of spray tower absorbers by decreasing the amount of untreated gas that can bypass each bank of spray nozzles and by re-entraining slurry that has begun flowing down the wall. Radial baffles projecting inward from the spray tower wall redirect flue gas, which might have bypassed the spray nozzles by traveling up the wall, back into the spray zone. Trays or wall rings can help reduce capital and operating costs by substantially improving SO2 removal efficiency and/or reducing the required liquid-to-gas ratio.

Materials of Construction Wet FGD systems present a significant challenge with respect to selecting materials of construction and design features that balance capital costs with long-term operating and maintenance costs. Other substances removed from the flue gas and incorporated in the FGD slurry (along with SO2) include chlorides, fluorides, beryllium, and other halides. The chlorides and other halides form dissolved salts that remain in the process slurry loop. As the chloride concentration continually increases, the slurry becomes progressively more acidic and more corrosive to the materials of construction of the FGD system. A high chloride concentration can also have a negative effect on the system’s SO2 removal capability. Until recently, the typical industry for maximum process c hloride concentration ranged from 40,000 to 60,000 ppm. This appears to be the maximum concentration before SO2 removal is impacted. Very expensive materials of construction are req uired to resist corrosion at this concentration of chlorides. These measures include high-nickel alloys, flake-glass-lined carbon steel, or ceramic tile linings. The current trend for new installations is toward lower chloride concentrations, in the range of 12,000 to 20,000 ppm, which allows use of less-expensive alloys and linings.

9-15


Achieving a lower process chloride concentration requires an increase in the rate of blowdown (the purge stream required to remove the chlorides from the process). The result is a larger wastewater stream that contains more dilute concentrations of chlorides and other impurities and that potentially entails significantly higher treatment costs. An evaluation and optimization study should be undertaken to recommend a chloride concentration and materials of construction that give an appropriate balance between the cost of materials and decreased performance at high chloride levels and the cost of wastewater treatment at lower levels.

Wastewater Impurities enter a wet FGD process with the flue gas and fly ash created by coal combustion, with limestone or other reagents, and with makeup water. Blowdown from the FGD system is required to remove chlorides and other impurities that may influence the reliability and performance of the system, and to control the fines content of the slurry. This blowdown is typically considered to be wastewater and must be treated before being discharged to surface water or recirculated back to the scrubber process. FGD wastewater presents a challenge to treat because of the following characteristics: •

High concentrations of TDS and TSS

•

High buffering capacity

•

Supersaturation in sulfates

•

High temperature

•

Potentially high organic concentration from dibasic acid (DBA) addition

•

•

Ammonia, from the ammonia slip for ESP conditioning and NOX control; nitrites and nitrates (if they’re produced by the SCR or by the ammonia treatment) Miscellaneous regulated heavy metals and trace constituents that vary by coal type (i.e., arsenic, mercury, selenium, boron, etc.).

Due to the complex nature of wet FGD wastewater, several stages are required to treat the wastewater to meet the National Pollutant Discharge Elimination System (NPDES) Permit requirements for surface water discharge. This treatment may include the following steps: •

Calcium sulfate desaturation

•

Primary solids removal

•

Trace metals precipitation

•

Secondary solids removal

•

Filtration

•

Biological treatment

•

Sludge dewatering

Wet FGD wastewater is typically combined with other water discharges from the power plant (wet bottom ash handling, cooling water, steam condensate, etc.). This mixing further defines exact wastewater treatment requirements. Wastewater characterization is required before treatment in order to tune these treatment processes for the many variables that are present at each site, and after treatment to determine if the wastewater is acceptable for discharge. These parameters may be affected by the ESP, baghouse, SCR, FGD; additives like organic acids; coal 9-16


and limestone types and makeup water constituents. Typical parameters include biochemical oxygen demand (BOD), total suspended solids (TSS), heavy metals (i.e., selenium, arsenic, boron), temperature, pH, and total dissolved solids (TDS). Testing (bench and pilot scale) is critical to define wastewater treatment system requirements and performance, especially when dealing with problematic pollutants like selenium, arsenic, mercury, BOD, etc. Many unanswered questions persist about the application of treatment technologies to FGD wastewater, especially treatment of problematic constituents. Many of the primary and tertiary wastewater treatment technologies being constructed by the power industry today have not been demonstrated at full-scale using actual FGD wastewater.

Single Absorber Early vintage FGD installations included multiple absorbers, even for smaller power plants, with at least one complete spare module. This resulted in the installation of inlet and outlet dampers to isolate the absorbers. Corrosion of the dampers and nearby duct was prevalent. Scale and slurry build-up on damper blades would impair their operation. The reliability of the absorbers was compromised. As the reliability of the absorbers increased and confidence their capability grew, smaller power plants started using single absorbers and stocking spare equipment and materials for maintenance. This eliminated the need for isolation dampers and their inherent problems. As the technology continued to mature, improved larger single absorbers were developed. Several years ago, the industry was utilizing a single absorber for 500 MW (or flue gas flow equivalent) or smaller. Larger power plants utilized two absorbers, again without a spare absorber. Suppliers are now offering a single absorber for 1000 MW (or flue gas flow equivalent) with the appropriate spare equipment to achieve very high availability and reliability.

Flue Gas Superficial Velocity A key factor in SO2 removal efficiency for a wet FGD system is contact time between the reagent slurry and the flue gas. Longer contact time enables higher SO2 removal. Flue gas velocity (typically measured as superficial velocity, neglecting the influence of vessel internals and slurry volume) is one of the major factors to achieving optimum contact time. Lower flue gas superficial velocities allow more contact time between the slurry and the flue gas. In some of the first wet FGD systems, absorbers had flue gas superficial velocities as low as 8 ft/sec (2.4 m/s). Because lower velocities require larger diameter absorbers, modifications to allow use of smaller absorber diameters were pursued as material costs rose. Superficial velocities increased to 10–11 ft/sec (3.0–3.4 m/s) as the technology developed and improvements were made. Current designs have been able to meet demands for higher SO2 removal efficiencies while also using higher superficial flue gas velocities in the 10–14 ft/sec (3.0–4.3 m/s) range. Testing and pilot testing have been conducted in an attempt to achieve superficial flue gas velocities as high as 20 ft/sec (6 m/s).

9-17


Additives Chemical additives have found widespread use in the industry, both for their ability to improve poor performance on existing FGD systems, and for their ability to allow a well-performing system to conform to strict annual emissions limits despite occasional upsets. For units that are unable to meet performance requirements, additives may be the least-cost approach for improving performance. The additives of choice have been organic acids, including dibasic acid (DBA) and formic acid. Historically, newly purchased wet FGD systems have not included organic acid addition in the original design, reserving that option for use if original g uarantees were not met or system upsets occurred. This also preserves the option of using organic acid to meet future, stricter limits without the capital expense required to increase absorber size and associated components. The interest in additives has been renewed as stricter SO 2 emission limits have required higher removal efficiencies, with organic acid enhanced wet limestone scrubbing now being considered for new wet FGD systems. The benefits of organic additives include: •

Reduced liquid/gas ratio in wet scrubbers

•

Less power consumption by slurry pumps

•

Reduced limestone consumption

•

Flexibility to over-scrub

•

Accumulation of additional SO2 credits for use or sale

•

Allows for flexibility in fuel selection

•

Improved ability to handle swings in inlet SO 2

•

Equipment downsizing for new scrubbers

Additives are also being considered as an alternative to installation of a spare spray header for ensuring system reliability and availability, and increasing the reliability of the FGD system while reducing capital costs. A small skid-mounted additive system and additive storage tank can be installed for a much lower cost than an installed spare header and associated piping, valves, nozzles, and pumps. The additive system would only be used in the event of the failure of one of the operating headers or other system upset. Again, the costs of the additive should be compared to the capital costs for the header and associated equipment. If organic acids are used in a wet FGD system, the gypsum and wastewater will contain small concentrations of the acid. This may influence the ability to sell the gypsum for some end uses and may change wastewater treatment system design or operation.

Flue Gas Distribution and Moisture Collection Flue gas distribution is probably the most overlooked parameter in the design of wet FGD systems. As higher SO2 removal efficiencies are required, design optimization for uniform gas flow in the absorber is important to the performance of the FGD system. Mal-distribution can allow gas to flow through the absorber with minimal contact with the sprays and corresponding reduction in system performance.

9-18


Another important flue gas distribution issue is the location of the wet/dry interface between the hot flue gasses entering the absorber and the wet saturated flue gasses in the absorber. This is the most corrosive zone in the wet FGD system and therefore requires the highest grade materials; typically, a high-nickel alloy is used. If the wet/dry interface extends into the inlet duct, which is normally carbon steel, rapid corrosion may occur. The flue gas exits the absorber as a saturated gas with some entrained moisture droplets. In the past, reheat was commonly used to evaporate the droplets and to decrease gas density to provide buoyancy for gas dispersion above the stack exit. However, this required considerable energy and created another wet/dry interface, with resultant material failure. Later wet FGD designs eliminated the reheat system. Current systems use a wet stack design, with saturated flue gas from the absorber entering the stack downstream of the FGD absorber. This can lead to “rain out” if condensation droplets remain entrained after forming in the stack. Flow modeling of the downstream duct and chimney liner is essential for determining the shape of the duct and position moisture collection devices and drains to remove moisture that accumulates on the duct and chimney liner walls. The moisture is then drained to the absorber. Flow model testing provides another benefit by helping the stack designer locate turns and other internal features so as to reduce flue gas pressure drop.

Recycle Pump Sizes A wet FGD slurry recycle system requires specially designed slurry pumps that are resistant to erosion, corrosion, and plugging. These pumps use rubber lining or alloy construction. They gain further resistance to erosion by operating at very low speeds and using belt drives or gear boxes for speed reduction. Until recently, the largest pumps offered for recycle service were capable of pumping approximately 35,000 U.S. gallons per minute or 2200 liters per second. The latest pumps offered by the manufacturers are capable of pumping approximately 70,000 gpm (4400 l/s). This increase in capacity means that fewer pumps and associated equipment are required. FGD system reliability has not been affected.

Process Control With the ever tightening regulations and lower SO 2 emission rates, process control becomes more important. Reliable process monitoring is essential for maintaining the required removal efficiencies. Typical process parameters that are continuously monitored include: •

Flue gas flow

•

Flue gas inlet SO2

•

Flue gas outlet SO2

•

Inlet flue gas temperature and pressure

•

Outlet flue gas temperature and pressure

•

Process slurry pH

•

Process slurry density 9-19


•

Process slurry pressures

•

Makeup water flow

•

Limestone slurry density

•

Limestone slurry flow

•

Bleed slurry density

•

Bleed slurry flow

•

Gypsum moisture content

FGD system parameters are further tracked through periodic sample testing and wet chemical analysis. Some of this sample analysis serves as a check on the accuracy of the automatic monitoring. Other parameters, which are not automatically monitored, help identify process trends that may indicate equipment malfunction or impending process upsets. Parameters evaluated by sampling include: •

Various process densities

•

Process pH

•

Limestone analyses

•

Dissolved solids

•

Suspended solids

•

Particle sizes

•

Sulfate and sulfite content

•

Additive content

Process Upsets Process upsets can result from failure of mechanical equipment, instrumentation, or control systems, or from changes in the chemical properties of the coal, reagents, additives, or process water. Several of the most common upsets are addressed here. Limestone Blinding

Limestone blinding can cause a fairly rapid decrease in wet FGD process reactions. This phenomenon occurs when another substance precipitates on the surface of the limestone particle, forming a thin coating that acts as a barrier and stops dissolution of the limestone. The coating prevents interaction between the solid CaCO3 and the sulfurous acid formed by the absorption of SO2 into the water surrounding the particle. The major contributor to the blinding phenomena has been identified as formation of soluble aluminum fluoride complexes (AlFx). Coal combustion is the primary source for both the aluminum and the fluoride ions. HF vapor, formed after combustion, travels in the flue gas before being absorbed by the limestone slurry. The aluminum content is released from fly ash that has been entrained in the FGD liquor and deposited in the reaction basin. Limestone blinding has become less of an issue as fly ash removal has improved with the increasing efficiencies of particulate removal devices (ESP or baghouse) installed upstream of

9-20


the FGD unit. Less fly ash reaches the FGD system, lowering the likelihood that sufficient aluminum to cause blinding will be present. Limestone blinding has also been observed in forced-oxidation systems when insufficient oxidation air is provided. Blinding results from the increasing concentration of dissolved sulfite and disappears quickly when sufficient oxidation air is introduced. Gypsum Quality

Gypsum generated by wet FGD processes is used in wallboard, cement, concrete, and, to a lesser extent, in agriculture industries. The gypsum processing utilizes hydrocyclones and vacuum filters or centrifuges to dewater the gypsum bleed slurry from the FGD process. For some uses, the gypsum is washed to remove some of the impurities. Gypsum quality must be maintained to ensure use by industry. Quality issues include: •

Moisture content

•

Crystal size

•

Chloride content

•

Mercury content

•

Heavy metal content

Oxidation Air

Oxidation air is introduced into the reaction tank of a wet FGD system to supply oxygen that completes the scrubbing reactions and oxidizes the calcium sulfite contained in the reaction tank to form gypsum (calcium sulfate). The most common method for introducing oxidation air into the reaction tank is the lance system. This system utilizes a series of vertical down flow air pipes to direct the compressed air to the jetting zone of side entry agitators installed in the reaction tank. The air is then distributed throughout the reaction tank. Another method used is the sparge grid. This system utilizes a multiple header arrangement with even spacing of bubble stations across the reaction tank plan. The performance of the lance system relies on the energy of the fluid jet (mixer power) and the submergence depth (compressor power). The performance of the sparge grid is less dependent on the mixer power and is, to a much greater degree, dependent on submergence depth. A newer method for introducing oxidation air directs part of the discharge from the slurry recycle pumps through an eductor that entrains oxidation air, which then enters the reaction tank through an injection tube. This method eliminates the air compressor, and reduces the capital and operating and maintenance costs associated with the compressor. At higher SO2 inlet loading, which requires more oxidation air, an air compressor will be required to supplement the edu ctor system.

9-21


10

DRY SO2 CONTROL TECHNOLOGIES Dry flue gas desulfurization (dry FGD) technology is normally u sed in plants burning low-sulfur (typically less than 2%) feed coal and in locations where limited space makes it difficult to install gypsum recovery and wastewater treatment systems. Dry FGD is a also a viable choice when insufficient water supply is available for wet FGD or when it is more advantageous to dispose of a dry waste product in a landfill. The lime spray drying absorption (LSDA) process is the most commonly used design. In this semi-dry scrubbing process, lime slurry is injected at the top of a scrubber located downstream of the air heaters. With slurry flow co-current with the flue gas flow, the slurry dries while adding moisture to the flue gas and absorbing up to 95% of the SO2. Part of the dry waste product is collected in the bottom of the Spray Dry Absorber (SDA) while the rest is collected in a particle collector. The Dry Sorbent Injection (DSI) process also produces a dry mixture of fly ash and reaction products. The application of lime-based DSI to coal-fired boilers is generally limited to lowsulfur fuels where an SO2 removal efficiency of 30–50% or less is required. In the DSI process, dry hydrated lime is introduced uniformly into the flue gas via a grid of duct injection lances. As with the LSDA process, the waste products are collected in a particulate control device. Experience with DSI indicates that SO2 removals rarely exceed 25–30% when using hydrated lime alone. With simultaneous humidification of the flue gas, SO2 removals of up to 50% are achievable.

Lime Spray Drying Absorption In the lime spray drying process represented schematically in Figure 10-1, flue gas is treated in an absorber by mixing atomized lime slurry droplets with the gas stream in co-current flow. As the water in the spray droplets evaporates, near the top of the absorber, the flue gas is cooled from 300°F (150°C) or higher to as low as 30°F (17°C) above the adiabatic water saturation temperature. The droplets absorb SO2 from the flue gas and the SO 2 reacts with the lime in the slurry.

10-1


Figure 10-1 Typical Dry FGD Process Flow Diagram

The lime spray drying process produces a dry mixture of fly ash and reaction products, which is typically captured in a baghouse. In most LSDA systems, a portion of the waste products is mixed with process water and recycled back to the atomizers as recycle slurry. The application of the lime spray dryer FGD process to coal-fired boilers is generally limited to fuels with less than 2% sulfur when high SO 2 removal efficiency is required. To optimize contact in the spray drying absorption process, the slurry is atomized through rotary atomizers or through dual fluid nozzles, as shown in Figure 10-2. Flue gas is introduced into each SDA module by means of gas dispersers, which distribute the incoming flue gas symmetrically around the atomizer unit(s) at a velocity and direction appropriate for achieving optimum absorption of the acids contained in the flue gas. The gas dispersers can be designed to introduce the flue gas into the absorber from the roof or from the roof and the center of the vessel. Guide vanes are installed at the outlet of the dispenser and are constructed of abrasion-resistant material. Careful control of gas distribution, slurry flow rate, and droplet size assures that the droplets evaporate fully prior to contacting the internal walls of the SDA module.

10-2


Figure 10-2 Dual Fluid Nozzle Atomizer (Left) and Rotary Atomizer (Right)

Lime Preparation System The lime preparation system combines process water and pebble lime to prepare hydrated lime slurry at approximately 20–25% (by weight) suspended solids concentration. Lime is discharged from each storage silo through a weigh feeder and is fed to an individual lime slaking system. The lime slurry product is screened of oversized particles in a vibrating grit screen and discharged from each slaker train, and either pumped or gravity fed to a common lime slurry storage tank. A tank agitator keeps ke eps the lime solids in suspension. This lime slurry storage tank can act as the sole source for all absorbers, or slurry can be pumped to separate tanks feeding individual absorbers. Lime slurry feed pumps draw suction from the slurry feed tanks and discharge to the lime slurry feed loops, which supply slurry to the SDA atomizer head tank. Constant pump speeds and pipeline velocities are maintained to prevent solids from settling or caking up within the lime slurry feed loop.

Recycle Slurry System In some systems, recycle slurry is mixed with fresh lime slurry in the lime storage tank to form the final slurry that is sprayed into the SDA module. mod ule. In other systems, recycle slurry is delivered to common or dedicated injectors with a separate system of pumps and piping. The recycle slurry enhances utilization of the lime reagent as well as promoting droplet drying in the SDA chamber. In the recycle system, FGD by-product solids from the SDA and particulate collector are conveyed pneumatically by the ash handling system to the recycle solids silo. Recycle material is discharged from the recycle solids silo through a fluidized outlet and flows to the recycle slurry preparation system. The recycle solids are initially combined with process water during discharge to the recycle mix tanks where additional water is added to prepare recycle slurry at up to 45% (by weight)

10-3


suspended solids. The recycle slurry is then delivered to the recycle slurry storage tank, entering the tank through a vibrating grit screen that removes oversized particles. The storage tank has an agitator to keep the solids from settling. Centrifugal pumps are used for the recycle slurry feed service. Constant pump speeds and pipe line velocities are maintained to eliminate settling or caking within the dedicated recycle slurry supply line to the atomizer level on top of the SDA.

Lime Spray Drying Absorption Process The LSDA feed slurry is injected into the flue gas as a fine spray using atomizing nozzles, which are typically either a rotary design or an air-atomized, two-fluid nozzle design, as shown in Figure 10-2. 10-2. The finely atomized feed slurry mixes with the flue gas, resulting in the evaporation of water and the removal of the SO2 via chemical reaction with the slurry. Recycle solids slurry may be mixed with fresh hydrated lime slurry before or during injection, or may be injected separately. Slurry is gravity fed from the atomizer head to the atomizers. The flow of slurry is controlled by maintaining an absorber outlet gas temperature. As the gas temperature gets higher than the set point, more total slurry flow is fed to the absorber. The flow of fresh lime slurry into the atomizer head tank is controlled to maintain an absorber outlet SO 2 level. As the SO2 level gets above the set point, more fresh lime slurry is fed into the head tank. The water content and droplet size of the slurry spray are precisely controlled so that the water evaporates completely as the flue gas and spray travel together through the absorber vessel. SO2 absorption takes place primarily while the flue gas is cooled adiabatically by this evaporation of the water contained in the atomized spray. The difference between the temperature of flue gas leaving the SDA and the adiabatic saturation temperature of the flue gas is known as the approach temperature. Reagent stoichiometry, recycle ratio, residence time, and approach temperature are the primary variables that control the SO 2 removal efficiency in the SDA system. The primary product of the reaction between the hydrated lime, Ca(OH)2, component of the feed slurry and the SO2 is hydrated calcium sulfite, according to the following relationship: SO2 + Ca(OH)2  CaSO3•½H2O + ½H2O (1) A smaller portion of the sulfur dioxide may also a lso react with oxygen in the flue gas to produce the secondary product of calcium sulfate dihydrate by the following reaction:

•

Ca(OH)2 + SO2 + H2O + ½O2  CaSO4•2H2O Sulfur trioxide is also found in the flue gas in small amounts. The sulfur trioxide reaction produces additional calcium sulfate dihydrate by the following:

•

(2)

SO3 +Ca(OH)2 + H2O  CaSO4•2H2O (3) Hydrochloric and hydrofluoric acids are also typically typica lly found in the flue gas in even smaller amounts. Calcium chloride (CaCl2) and calcium fluoride (CaF2) are produced as a result. High levels of calcium chloride can have a marked impact on the SDA performance.

•

Most of the water added to the lime in the initial hydration process is evaporated in the absorber. There are no wastewater streams exiting the absorber. The degree of reaction depends on the

10-4


amount of liquid present, the approach to the adiabatic saturation temperature, and the residence time for particle drying. As the flue gas and feed slurry mixture passes through the SDA module, the spray drying and initial SO2 removal processes are completed. The flue gas and entrained reaction products, unreacted reagent, and fly ash exit the SDA module and flow into the particulate collector. Additional SO2 removal takes place in the particulate collector as the un-reacted lime accumulates on the bags and comes in contact with the flue gas still containing SO2. The SDA module is designed to ensure that most of the particulate can be entrained in the flue gas and carried to the particulate collector. The larger, coarser particulate that is not entrained in the flue gas exiting the SDA module may be discharged from the bottom of the SDA module hopper for disposal. The reaction products, un-reacted reagent, and fly ash collected in the particulate collector hoppers is then conveyed by the ash handling system to either the recycle ash storage silo for reuse or the waste ash storage silo for disposal. Upon exiting the particulate collector, cleaned flue gas is directed to booster fans that discharge to the stack.

Current Issues for Lime Spray Drying

Higher Removal Efficiencies The demand for dry FGD systems to increase removal efficiencies has heightened as SO2 emission limits continue to be lowered. As recently as the early 2000s, SO 2 removal efficiencies of 90% were the norm. Current LSDA systems are being designed for SO2 removal efficiency as high as 95%. Very few methods are currently practical for increasing SO 2 removal efficiency in LSDA systems. The systems must continue to meet industry standards for reliability and a vailability after any process or equipment modifications are made to obtain higher removal efficiencies. Available options include: • •

Operating at closer approach to adiabatic saturation temperature Increasing effective contact time between the reagent and the SO2 laden flue gas by passing the flue gas through filter cake accumulated by a barrier filter

•

Increasing slurry reactivity by increasing chloride content in the slurry

•

Increasing contact time by increasing the height and/or diameter of the SDA vessel

•

Increasing contact proximity by increasing the fluid to gas ratio

•

Increasing slurry reactivity by reducing recycle and increasing makeup with fresh lime.

Reducing Approach to Saturation Temperature In dry FGD applications, SO2 removal is closely correlated to the flue gas adiabatic saturation temperature. SDA operation at outlet temperatures close to the saturation temperature prolongs drying time and increases SO2 removal and reagent utilization. The impact of increasing SO2 removal by operating at close approach temperatures is the added capital cost for protective coatings for the SDA walls, wa lls, SDA hoppers, and the particulate collector co llector inlet duct and hoppers. More aggressive by-product removal systems such as screw or drag chain conveyors may be required. Higher costs would also be incurred for the additional particulate

10-5


collector equipment, structural supports, and foundations. It is expected that reagent costs would decrease slightly.

Increasing Contact in a Barrier Filter The use of a barrier filter, typically a fabric filter, for particle collection is the most cost effective, and therefore the most common method for increasing the contact time for the reagent and flue gas in order to achieve high SO2 removal efficiency. The fabric filter can be sized conservatively and/or operated at higher pressure drops to maximize the contact between the filter cake and the SO2 remaining in the SDA outlet flue gas. This increase in contact time reduces reagent cost by improving the removal efficiency for a given consumption lime reagent. This saving must be weighed against increased fan power consumption.

Effect of Chlorides in Fuel and Makeup Water Another method to increase SO2 removal performance is to increase the chloride concentration in the flue gas or the makeup water. This will create more CaCl2 in the by-product, which will lead to longer particle drying times and higher highe r SO2 removals. The chloride levels in the SDA byproduct can be increased by blending with higher chlorine coals or through direct addition of chlorine compounds to the makeup water. The impact of increasing SO2 removal by operating at higher chloride levels is the added capital costs for protective coatings for the SDA walls, SDA ho ppers, and the PJFF inlet duct and hoppers, as well as more aggressive by-product removal systems such as screw or drag chain conveyors. Operation at higher approach temperatures will offset much of the CaCl2 effect. It is expected that reagent costs will decrease when operating at low approach temperatures.

Gas Residence Time In the design of early spray dryer d ryer absorber processes, SDA gas residence time was on the order of 8–10 seconds. This may have been adequate for low-sulfur western U.S. coals where SO2 removal requirements were minimal and chloride levels were low. As the size of SDA modules has increased, so has the number of applications on low- to medium-sulfur eastern U.S. bituminous fuels. As the effect of chlorides on drying d rying time and by-product characteristics has become better known, most suppliers have increased the minimum gas residence time to 12–15 seconds. In most applications, this assures that a dry product is discharged from the SDA.

Use of Recycle The recycle slurry enhances utilization of the lime reagent as well as promoting droplet drying in the SDA chamber. Typically, lime usage is reduced by the use of recycle slurry in the SDA, where un-reacted reagent may remove more SO2 as it is recycled back through the spray drying process. Although saving on reagent cost, the use of recycle increases capital cost and other operation and maintenance costs. Additional mixing tanks, pumps, valves, and piping are required. O&M cost impacts include power consumption and increased wear and tear on the atomizers. Many operators of SDA units on western U.S. coals have ceased using recycle due to increased maintenance issues, including atomizer pluggage and control valve wear.

10-6


Although some systems offer capital savings by mixing recycle with fresh slurry for common injection, others use separate systems that can quickly change the balance of fresh slurry and recycle slurry to respond more quickly to upsets and to events such as sootblowing that cause rapid increases in fly ash loading.

Materials of Construction Nearly all of the SDA units in service to date use A36 carbon steel for the absorber construction. These materials hold up well as long as the SDA outlet temperature is well controlled at more than 30°F (17°C) above the adiabatic saturation temperature. Instances where problems with back end corrosion have occurred can usually be traced to improper operation, too close to the saturation temperature, in an attempt to reduce reagent cost and increase SO 2 removal. Other instances of corrosion have included units where flue gas chemistry was not well known and chloride levels in the flue gas or makeup water were higher than expected.

SDA Vessel Size In an attempt to use single SDA modules for generating units up to 400 MW, SDA vessels have increased in diameter from 48–50 ft (about 15 m) to 60–70 ft (18–20 m). Use of a single SDA module saves significantly on capital costs and requires less space than would be needed for multiple SDA modules for the same unit. The successful use of multiple atomizers per SDA module has made the use of large SDA modules possible.

Single versus Multiple Atomizers State regulations regarding SO2 averaging times can impact the use of single SDA vessels and single atomizers. Atomizers can be maintained on-line, but a lower level of SO2 removal is usually the result of atomizer maintenance. This impact can be mitigated by the use of multiple atomizers, where the effect of the removal of one atomizer for maintenance will be minimized. Where SO2 removal requirements are on a 24-hour or 30-day rolling average, use of overscrubbing during low load operation can also help the plant meet compliance levels.

By-product Disposal Within the U.S markets, dry FGD waste has limited commercial application and will most likely be sent to landfill. Excess waste solids, that are not recycled in the scrubbing process, are pneumatically conveyed to the by-product or waste storage silo. This material is discharged from the silo through a pin mixer, where it is wetted to control dusting, and is dropped into a dump truck. Large trucks are typically used to haul the material to a landfill for disposal. In Europe, recycling and utilization of dry FGD waste products has advanced significantly, as a result of almost two decades of efforts by European industry leaders and authorities. As much as 86% of dry FGD by-product generated within the European Union is registered as being reclaimed for beneficial uses (ECOBA statistics for 1998). This reuse still faces considerable difficulties, as the annual amount produced at each plant is relatively small and utilization options are limited by the complexity of the chemical composition, which is compounded by process specific variations from unit to unit. The most important current uses for dry FGD wastes include:

10-7


•

•

Use as stowing material and mining mortar component for coal mines. This accounts for approximately 75% of the dry FGD waste utilization in Europe. Commercial applications of dry FGD/fly ash mixtures as fill and sub-base material for storage areas, parking lots, embankment material for roads, sealing layers, etc.

Use of dry FGD product as reagent in wet FGD and conversion to gypsum. In Denmark, this is now a large-scale commercial use. Developing applications include: •

•

•

Thermal conversion to anhydrite, for use as a setting agent in cement and as a component in bricks and aggregates of different kinds. This has been demonstrated technically, but only on a small-scale basis. Use as Ca-S fertilizer. The market interest for dry FGD produ ct as fertilizer continues to increase after successful research in the 1990s.

Other Dry SO2 Control Technologies Alternatives to the lime spray drying process for dry SO 2 control technologies include dry lime injection, circulating dry scrubbers, and flash dryer absorbers.

Dry Sorbent Injection Process The Dry Sorbent Injection process produces a dry mixture of fly ash and reaction products (similar to LSDA). The application of DSI using lime, for SO 2 removal in coal-fired boilers, is generally limited to low-sulfur fuels where an SO2 removal efficiency of 30–50% or less is required. In the DSI process, dry hydrated lime is introduced into the flue gas by means of a grid of injection lances inserted in the ducting downstream of the air heater. The purpose of the grid is to uniformly distribute the lime in the flue gas to assure optimum absorption of the acids contained in the flue gas. The hydrated lime reacts with the incoming SO 2, as in equation (1) in the subsection on the LSDA process, and is captured downstream by a particulate collector. The DSI reaction proceeds at a slower rate than with lime spray drying because there is a minimum of water present. As such, DSI with hydrated lime typically achieves SO2 removals of only 25–30%. With simultaneous humidification of the flue gas, SO2 removal of up to 50% may be achieved. The DSI process produces an inherently dry by-product and does not directly reduce flue gas temperature. Flue gas temperature may be reduced if simultaneous humidification using water spray is implemented.

10-8


Figure 10-3 Typical Dry Injection System

Circulating Dry Scrubber (CDS) Process The CDS process, illustrated in Figure 10-4, uses a reaction similar to the LSDA process and produces a similar dry mixture of fly ash, reaction products, and unreacted reagent. By using a high recirculation rate, this process is able to use a small reactor size. Because the reagent makes 35 to 50 passes in a circulating fluidized bed, obtaining close contact and extended residence time of thirty minutes or more, the CDS process is not constrained to a maximum fuel sulfur content as are most other dry FGD processes. Higher SO2 loadings can be addressed by increasing the reagent content in the bed, though this can result in significant increase in reagent cost.

10-9


Figure 10-4 57 Schematic of Circulating Dry Scrubber System (Lurgi Lentjes North America)

The CDS process is reported to be capable of achieving 95–98% removal efficiency using dry 58 powdered lime reagent. Performance is maximized by using spray water to moisten the bed material or to reduce the temperature of flue gas entering the scrubber from around 300°F to about 30°F (17°C) above the adiabatic water saturation temperature. This technology uses an ESP or fabric filter to capture and recycle particulate carryover, which is 90% lime. Accumulation of spent reagent and fly ash is limited by disposing of a fraction of the captured material, normally to landfill. ESP is normally preferred for particulate control because high dust loading tends to plug fabric filters. ESP sizing is critical for controlling particulate emissions. To date, there is little experience with this techn ology in the United States.

57

Best Available Retrofit Technology Analysis for Sherburne County Generating Plant, Units 1 and 2, Submission to Minnesota Pollution Control Agency, Xcel Energy, October 2006 . 58

Ibid.

10-10


Flash Dryer Absorber (FDA) Process The FDA process, illustrated in Figure 10-5, uses a reaction similar to the LSDA process and produces a similar dry mixture of fly ash, reaction products, and unreacted reagent. The application of the FDA using lime to coal-fired boilers is generally limited to low- to mediumsulfur fuels, in most cases where an SO 2 removal efficiency of 94–95% is required.

Figure 10-5 Alstom FDA Process

The FDA process uses a much smaller reactor size and a much higher recycle rate and flue gas velocity than are used in the LSDA process. Water and fresh lime are added to fly ash, reaction products, and un-reacted reagent recovered from the downstream particulate control system, which may be an ESP or a fabric filter. This mixture is re-injected into the upward flowing flue gas in the dust collector. With the high solids-to-water ratio, evaporation occurs rapidly, cooling and humidifying the flue gas while flash drying the particulate. Water is injected in quantities similar to that required for a semi-dry SDA, to reduce the flue gas temperature to within 30–50°F (17–28°C) of the adiabatic saturation temperature. The hydrated lime reacts rapidly with the incoming SO2. The waste by-product produced in the CDS or FDA is similar to the SDA by-product. However, experience with CDS or FDA with hydrated lime injection indicates that SO2 removals of 95% are achievable, especially downstream of a circulating fluid bed boiler system, where limestone 10-11


is injected for SO2 control. The un-reacted lime remaining in the boiler outlet flue gas helps with SO2 removal in the CDS or FDA. For this application, the downstream particulate collectors are specifically designed to handle the extremely large circulating particulate loads.

10-12


11

SO3 CONTROL TECHNOLOGIES Although sulfuric acid (H2SO4) aerosol emissions have not traditionally been regu lated, most permits for new plants contain limits for such aerosols and/or for condensable particulate, which are in the broader category of PM2.5. Because sulfur trioxide (SO3) is the precursor to H 2SO4 aerosols, considerable effort has been devoted to the development of SO3 control measures. The amount of SO3 that eventually reaches the back end of a coal-fired boiler (i.e., the target of controls) varies with coal properties, boiler thermal profile, and operating conditions. The increased concern results, in part, from increased flue gas concentrations following the conversion of SO2 to SO3 in SCR systems that are now being installed reduce NOX emissions. At the same time, flue gas temperatures are reduced, and additional water vapor is made available, by wet FGD systems that are installed to remove SO 2. When the increased concentration of SO3 combines with water vapor, it may raise the acid dew point of the flue gas above the stack exit temperature. This leads to formation of a condensed sulfuric acid mist that becomes apparent as a visible trailing plume emitted from the chimney. Although the plume is typically blue, plumes that appear brown or other colors have been attributed to H2SO4 in varying atmospheric conditions. Visible plumes have been observed with H2SO4 concentrations as low as 5 ppmv and also may be influenced by the presence of ammonium bisulfate, some nitrous compounds, and other substances. Figure 11-1 contrasts chimney plumes created without and with SO3 control.

11-1


Figure 11-1 59 Visible Results of SO 3 Control Using Sorbent Injection

Various methods have been evaluated to reduce the sulfuric acid mist concentrations in the flue gas, especially from plants burning eastern U.S. bituminous coal. Although both wet and dry FGD systems remove some SO3, a larger fraction is able to bypass these systems. The most highly developed technologies today use an alkali sorbent in combination with specification of SCR catalyst with low SO2-to-SO3 conversion. Increasing numbers of permits for new plants are requiring the Owners to install a wet ESP as a final polishing step following the wet FGD system. Coal cleaning can be part of an SO3 reduction strategy, as it lowers incoming sulfur (hence SO 2 and proportionally SO3). An equally important benefit of coal cleaning is reduction of iron compounds that make the fly ash serve as an SO2-to-SO3 oxidation catalyst. Ongoing EPRI research aims to understand the mechanisms of SO3 formation and depletion in the boiler in order to be able to reduce boiler outlet SO3 through operational means

SO3 and Acid Mist Formation in Coal-Fired Boilers There are numerous factors that affect the formation and depletion of SO3 in the boiler, and the amount of SO2 generated in the boiler that is oxidized to SO3 and then hydrolyzed to H2SO4 is quite variable. For typical eastern U.S. medium- to high-sulfur coal, about 1% of the SO2 is oxidized to SO3 by the time the flue gas exits the economizer. At the other extreme, western U.S. coals, especially Powder River Basin, produce virtually no SO3. Their highly alkaline fly ash removes the bulk of any SO3 that is generated. 59

P. Nolan, “Developments in Sorbent Injection Technology for Sulfuric Acid Mist Emissions Control,” U.S. Department of Energy Environmental Controls Conference, Pittsburgh, Pennsylvania, May 18, 2006.

11-2


SO3 is formed in the boiler furnace during combustion of the coal and downstream of the furnace as a result of catalytic oxidation. In the furnace, most of the sulfur in the coal combines with oxygen in the combustion air to form SO2. Some of this SO2 is further oxidized to SO 3 through a combination of homogeneous reactions and heterogeneous reactions in which the fly ash (suspended in the flue gas as well as deposited on the tubes) acts as a catalyst. In boilers that have SCR systems for post-combustion NOX control, the SCR catalyst may oxidize additional SO2 and significantly increase the SO3 concentration in the flue gas. The terms “SO3,” “H2SO4,” and “sulfuric acid mist” represent different forms of the same pollutant and are often used interchangeably. At the low flue gas temperatures downstream of the air heater, the SO3 hydrolyzes (combines with water vapor) to form H 2SO4 vapor. Sulfuric acid mist (or aerosol) is formed when the flue gas temperature drops below the acid due point (which is higher than the water dew point), in or downstream of the FGD system or shortly after the stack exit. This condensation forms very fine droplets that are only minimally removed by impact in the wet FGD system or its mist eliminators. The amount of SO2 converted to SO3 by the SCR catalyst is a function of the catalyst properties. The first SCR catalysts installed on coal-fired power plants in the United States typically converted 0.5–1.5% of the incoming SO2 to SO3. Suppliers are now claiming they can provide catalysts that oxidize as little as 0.1% of the SO2 across each catalyst layer. In the example below, it is assumed that 1% of the SO2 entering the SCR reactor will be converted to SO3. The following example outlines the relative concentrations and conversions that may be observed at a plant. Sulfur in coal:

2.5%

SO2 in the flue gas in the boiler:

1737 ppm

SO3 generated in the boiler:

17 ppm (assuming 1% conversion of SO2)

SO3 generated in the SCR

17 ppm (1% conversion in the SCR)

Total entering the wet FGD

34 ppm

Nominal capture by the FGD

10–15 ppm

Total SO3 (H2SO4) emitted

19–24 ppm

In this example, the flue gas H2SO4 concentration of 34 ppm would result in a visible plume. Plants burning western U.S. coals, whose sulfur content is less than 1%, might chose to use a spray dryer and baghouse (fabric filter) for SO 2 and particulate control. This fuel produces very little SO3, and the spray dryer/baghouse combination efficiently captures any remaining SO3. Outlet SO3 emissions should be <1 ppm.

Sorbent Injection Control Technologies SO3 has been successfully controlled by duct injection of several types of alkaline sorbents containing calcium, magnesium, sodium, or ammonia. These sorbents typically are injected upstream of the primary particulate control device (ESP or baghouse) in a manner that causes

11-3


them to mix well with the flue gas. The sorbents and SO 3 react to form end products that are solid compounds that can be removed in a particulate control device. Sorbents successfully used in the control of SO 3 include magnesium oxide, calcium hydroxide, sodium bisulfate, and sodium sesquicarbonate (trona). Ammonia is also effective in removing SO3 but it can prevent the fly ash from being sold for use elsewhere. Sorbents can be added to the fuel or injected in the boiler at different locations: in the ductwork between the SCR and air heater; between the air heater and the particulate emission control device. Figure 11-2 shows injection points for different sorbents.

Injection Methods Proper reagent distribution, mixing, and residence time are important for achieving high SO3 removal efficiencies. The injection system and the injection points should be chosen to optimize these variables for maximum SO3 removal. Injecting sorbents downstream of the air heater, with the inherent temperature gradients at this location, presents a unique design consideration. It is important to recognize gas stratification in the design of the sorbent injection system. Sorbents that are injected as dry powdered solids are typically delivered by rail tank car or tanker truck and stored in silos. Pneumatic conveying systems are normally used to transfer dry powder to the silo and from the silo to injection points. In-line milling systems may be used to ensure that the sorbent reaching the injection point is in fine particle form. Sorbents that are injected in slurry form are usually delivered as a concentrated slurry by rail tank car or tanker truck and transferred by pump to a slurry storage tank. Batches of concentrated slurry are transferred from the storage tank to a “day storage tank” where additional water is added to dilute the slurry to the desired concentration. From the day tank, the slurry is pumped to the injection points where atomizing nozzles distribute it into the flue gas stream. Dual flow nozzles, with air injection, may be used produce a fine spray to ensure evaporation and reaction with the SO3. Depending on the settling and caking properties of the sorbent, the day tank may require agitation to ensure that a consistent concentration of slurry is delivered for injection. Wet slurry injection generally requires longer contact time in the flue gas as the moisture must be evaporated to realize the proper SO3 removal reaction. Poor atomization of the slurry may result in solids build-up on the boiler internals and ductwork.

11-4


Figure 11-2 60 Preferred Injections Points for Various Sorbents

Sorbent Properties

Magnesium Hydroxide Two types of magnesium hydroxide (Mg(OH)2), also known as hydrated magnesium oxide, can be injected in the boiler furnace to remove SO3 generated during combustion. One is a byproduct from a magnesium lime-based wet FGD system. The other is a commercially available product. Both are delivered as a slurry containing 50% solids. The concentrated slurry is diluted on-site to approximately 15% solids prior to injection into the furnace. Removal of up to 90% of SO3 generated in the furnace has been demonstrated with magnesium hydroxide injection. Magnesium hydroxide has not been effective in controlling SO3 formed in an SCR reactor, as the flue gas temperature downstream of the SCR unit is not high enough for the reactions to occur. Magnesium-based compounds have been used successfully for decades for capture of SO3 in oilfired units. Their use in coal-fired units has been limited. Magnesium oxide is effective because it reacts directly with SO3 to form magnesium sulfate. Magnesium sulfate is water-soluble, so it is unlikely to form hard-to-remove deposits on equipment and ductwork. In coal-fired units, it is believed that magnesium oxide can form a film on furnace heat transfer surfaces and prevent catalysis of SO2 to SO3 by iron oxides. Also, magnesium oxide is known to

60

Figure from NETL report, Economic Comparison of SO3 Control Options for Coal-fired Power Plants .

11-5


modify coal ash fusion temperatures and slag properties, so that injection of magnesium oxide can reduce furnace slagging or make slag deposits more friable. However, compared with other sorbents, this option appears to have relatively high operating and maintenance costs. The delivered cost of the sorbent (due to the weight of water in the slurry) and the molar ratio required to achieve high SO3 removal rates tend to be high relative to other sorbents. Also, in systems where a significant amount of S O3 is formed in the SCR unit, magnesium hydroxide might need to be used in conjunction with another SO3 control method.

Sodium Bisulfate (SBS) A patented technology injects sodium bisulfate (Na2HSO4) in solution form upstream or downstream of the air heater. It is necessary to have sufficient length of duct, free of internal obstructions, to provide about two seconds of residence time downstream of the injection point. This provides necessary drying and reaction time and helps prevent deposition of un-reacted or moist solids on duct surfaces. The preferred injection location is upstream of the air heater because higher temperatures promote faster drying and slightly higher reaction rates, as well as offering maximum corrosion protection by reducing the SO3 concentration before the air heater. In addition to decreasing the chance that corrosive sulfuric acid will form deposits on the cold-end surfaces of the air heater, this allows for lower flue gas temperatures at the air he ater outlet, improving the plant heat rate. In a typical power plant burning medium-sulfur coal, the air heater outlet temperatures can be lowered by approximately 25°F (14°C) and still remain above the acid dew point temperature. Testing has shown that a molar ratio of Na-to-SO 3 of 2:1 is required to achieve >85% SO 3 removal. Reducing SO3 concentrations upstream of the air heater also helps minimize the potential fouling effects of ammonia slip from the SCR. Specifically, the combination of ammonia with SO3 can form ammonium bisulfate (NH4HSO4), which is a leading contributor to air heater fouling and plugging. For SBS injection upstream of the air heater, residence time must also be sufficient to dry the reaction products so that sodium bisulfate and sodium pyrosulfate do not contribute to air heater fouling.

Trona Another patented SO3 mitigation technology uses injection of trona (Na2CO3•NaHCO3•2H2O). Like other sodium-based sorbents, trona is very reactive and effective in removing SO 3. The SO3 reacts with the trona to form solid sodium sulfate and sodium bisulfate. Trona is injected in dry form, which both minimizes the capital investment needed to deliver the sorbent to the flue gas and the required sorbent preparation equipment. Dry trona is also relatively benign, minimizing safety concerns. A possible supplemental benefit of trona injection is improved ESP performance. The sorbent cost is low, but the transportation cost from the mine in Wyoming can be high. O&M costs are minimal relative to other sorbent technologies. Trona is delivered as a solid and is conveyed to storage silos before being pneumatically injected into the ductwork downstream of the air heater and upstream of the particulate control device. As with any solid, material handling systems must incorporate proper silo design and control of temperature and moisture to maintain free flow.

11-6


Proper nozzle and lance design are imperative for reliable operation. Flue gas temperature is an important consideration in preventing deposition in the ductwork, which could occur if the trona and/or SO3 reactants become liquid. Reported test results indicate that the trona reactions do not proceed with the same efficiency as SBS injection and that trona-to-SO3 molar ratios of 5:1 are needed to achieve removal efficiencies >85%. The slower reaction rate is one factor in the solids deposition problems that have occurred with trona injection upstream of the air heater. Therefore, trona is injected downstream of the air heater, in a preferred temperature window of 275–340°F (135–170°C), where trona is rapidly calcined to sodium carbonate, which neutralizes the acid gases. Removal efficiencies approaching 80% have been reported with molar ratios less than 2.5:1. At temperatures above 360°F (180°C), the sorbent can become sticky and adhere to downstream surfaces. Particle size plays a large role in the efficiency of SO 3 removal by the trona injection process. Trona is typically delivered as T-200 (23 microns), but use of smaller trona particles leads to significant increases in SO3 removal along with reduced trona consumption. Obtaining the more finely ground trona requires off-site or on-site grinding because S olvay Chemicals, the current U.S. supplier, does not offer the smaller particle sizes.

Hydrated Lime Hydrated lime, Ca(OH)2, has been successfully used as a sorbent to mitigate SO 3. It is injected on a dry basis after the air heater and ahead of the particulate control device. Test data indicate that high-surface-area lime hydrate provides the best performance and creates less maintenance issues compared with low-surface-area forms. Hydrated lime injection has been observed to reduce ESP performance at some units whereas other others have had no ESP performance issues. Specifically, sites with marginally sized ESPs have reported a reduction in ash collection due to calcium buildup and increased resistivity, which may in turn be a result of the increased particle loading to the ESP.

Ammonia Ammonia is injected in vapor form between the air heater and the particulate control device. The reaction produces a particulate, ABS, which is removed in the particulate control device. Injection of ammonia upstream of the air heater is not feasible for SO3 control because the ABS changes phase and causes fouling as it is cooled in the air heater. This application uses NH3-to-SO3 molar ratios in the range of 1:1 to 2:1. A molar ratio of 1:1 will produce mostly ammonium bisulfate, whereas a 2:1 ratio will yield mostly ammonium sulfate. Most plants operate at the higher molar ratios to avoid ash handling problems caused by ammonium bisulfate and because ammonium sulfate is the more desired product. Low ratios are used to weight the reaction toward ABS production in power plants that produce little SO3. In these cases, NH3 injection improves ESP performance by increasing the cohesiveness of the fly ash. With the injection point downstream of the air heater, the ammonium bisulfate will be liquid and/or a sticky solid at the flue gas temperatures in the duct.

11-7


The molar ratio should not exceed 2:1 because this can increase reaction with SO2 and cause formation of ammonium bisulfite. In this environment, ammonium bisulfite forms submicron particles that can not be absorbed by the FGD system and can produce a visible plume. Where allowed, ammonia is typically delivered by rail or truck as anhydrous ammonia. The ammonia is stored on-site in pressurized tanks. The equ ipment includes an ammonia control/dilution skid. The anhydrous ammonia is fed to an ammonia flow control station. The ammonia is diluted with heated air and injected downstream of the air heater. As in any process requiring a control system that is designed for proper operation, an ammonia injection system must be designed, maintained, and operated such that adequate measures are taken to prevent over-injection. When using anhydrous ammonia, the plant also must follow significant protocols to protect environmental and public safety.

Particulate Control Impact SO3 content in the flue gas can have positive impact on the ESP performance. Units burning lowsulfur coals frequently inject small quantities of SO 3 for ash conditioning to improve ESP particulate removal efficiency. Removal of all SO3 would negate this positive effect. SO 3 is not injected when higher sulfur coals are used. Sodium injection enhances the ESP collection efficiency by reducing the fly ash bulk resistivity. This has been shown to offset the negative impact of removing the SO3. Sorbent injection for SO3 removal increases loading of the particulate collector. Because ESPs essentially perform as constant efficiency devices, this may lead to a small rise in particulate emissions. Although this increase is expected to be small, because the sorbent is injected at single digit molar ratios to the SO 3 (i.e., a few tens of ppm), the resulting ESP outlet emissions could increase enough to trigger U.S. “New Source Review” rules (25 TPY for total particulate or 15 TPY for PM10). ESPs that are sufficiently large and/or have operating margin (i.e., can increase power input above normal current levels) should be capable of removing the additional particulate and avoid emission increases. However, if the ESP is already marginal or undersized, the additional particulate loading will very likely cause emission increases. Fabric filter particulate collectors should be capable of handling the additional particle load without negative impact to removal efficiency. They should be evaluated, however to ensure that total flow area and automatic cleaning systems are adequate for preventing development of excessive pressure differential. Ammonia is sometimes injected upstream of an ESP specifically to enhance its performance. The ammonia improves ESP performance by increasing fly ash cohesiveness.

Fly Ash Quality Many power plants market the fly ash that is produced at their facility. Most of the fly ash is used in the production of concrete and cement products. The products formed by the SO3 removal process utilizing sorbent injection can directly impact the ash marketability. If the fly ash cannot be sold, it is landfilled as a waste product. The standard specification used in the concrete industry (ASTM C-618) places limits on various constituents in fly ash to be used by the industry. Close attention should be give to the alkali

11-8


content limit which is 1.5% maximum, expressed as Na 2O. Injection of sorbents can result in fly ash approaching or exceeding this limit. It has been reported that injecting SBS can result in 50–80% of total sodium allowed in the fly ash by the ASTM standard, with additional alkali contributed by the mineral content of the fly ash. This should be considered before SBS is selected as the control method for SO3. Trona-based sodium products have similar impacts on fly ash quality which are more pronounced than for SO3 control using the SBS technology. With the high molar ratio required when using trona, it is very likely that the limit of 1.5% alkali in fly ash will be exceeded. This should be considered before trona is selected as the control method for SO 3. The most prevalent issue associated with ammonia injection for SO3 control is the strong ammonia odor that can result even at very low concentrations of ammonia in the fly ash. With 2 ppmv ammonia slip from the SCR, the ammonia content of the fly ash can be equivalent to 100– 300 ppmw (parts per million by weight) in the fly ash. This is a small fraction of available ammonia that is seen with SO3 control by ammonia injection. The total ammonia content of ash from a system using ammonia sorbent could be as high as 1000–7000 ppmw. Because the reaction products—solid ammonium sulfate and ammonium bisulfate—are readily soluble in water, the solids could readily dissolve into their constituents and liberate ammonia into the air with just the moisture from humidity in the air. Fly ash with high amounts of ammonia will be unacceptable for use in concrete due to odor problems associated with the mixing, pouring, and curing of the concrete. However, some ash marketing firms offer additives or other beneficiation processes that fix or remove the ammonia from the fly ash.

Wet Electrostatic Precipitators (WESP) Wet electrostatic precipitators are used in incinerators and some industrial processes to control acid mists, sub-micron particulate, metals, and dioxins/furans as the final polishing devices downstream of other air pollution control systems. Removal efficiencies >90% for PM2.5 and SO3 have been reported, along with near-zero opacity. Detailed information on WESP operation, configurations, and design considerations can be found in EPRI report 1009775, “Update of Particulate Control Guidelines, A State-of-the-Art Report for Utility Wet Electrostatic Precipitators.” New developments in WESP technology offer the promise of reduced cost (membrane WESP) and enhanced mercury control (plasma-enhanced WESP).

Although the basic technology is over 100 years old, power plant experience with WESP is extremely limited. The majority of installations are at industrial sites. In Japan there are a few installations downstream of FGD units at power plants in that b urn coal with less than 1.2% sulfur. In North America, several power plant installations are beginning to accumulate significant operating time. Still, industrial experience provides most of the av ailable experience base from which “lessons learned” can be drawn to establish good designs for power plant applications. As there are potential pitfalls when a technology is transferred from one industry to another, power companies should use caution and implement conservative designs to minimize the impact of potential unknowns. A WESP is designed and operates much like a dry ESP, but uses wetted plates to capture and hold the particulate. The collectors are washed continuously by flooding from the top and/or by 11-9


spraying liquid, usually water, from the top and/or ga s path entry. The plate washwater retains the particles that are attracted to the collecting surface, thereby avoiding re-entrainment emissions (typically the source of ~50% of the emissions from an advanced dry ESP). As the water flows down the plates, it carries with it all collected material (fine particles, aerosols, etc.) and is then collected in troughs located at base of the WESP. It is then bled to a waste water treatment system to be neutralized and cleaned of the collected material, before being recycled to the WESP. At installations where the WESP follows a wet FGD, the WESP water cycle is usually integrated with the scrubber water cycle to optimize water usage. Resistivity of the collected material is generally not a major factor in removal performance in the WESP because the liquid on the plates typically has a low resistivity. Because resistivity does not constrain the power that can be applied to a WESP, these units usually incorporate very aggressive discharge electrode designs to simultaneously maximize the applied current and the product of the applied voltage and current. The high power levels allow a WESP to have higher collection efficiencies than a dry ESP of comparable size. It should be noted that even though power densities in a WESP can be double the power density in a dry ESP, the total power consumed by a WESP is still modest compared with power consumed by other pollution control systems such as FGD systems.

Horizontal Flow WESP One possible layout for a horizontal flow WESP following an FGD is two parallel casings with half of the flue gas being treated in each box. The two layouts most often proposed are over-andunder and side-by-side arrangements. The over-and-under design typically has a lower capital cost and is normally the proposed arrangement unless a particular configuration is specified by the power producer. The inlet duct arrangement from the FGD system to the WESP is less complex in this arrangement and requires less duct material. However, due to the low flue gas temperature and associated acidic conditions, the ductwork must be fabricated of corrosionresistant materials. The disadvantage to this layout is constrained operability and maintainability. Removal and installation of the transformer/rectifier (TR) sets for the lower box will be difficult due to the low overhead caused by the upper box. Further, if the WESP should require a rebuild, the over-andunder design may make it difficult to gain access to the internals of the lower casing. The over-and-under design can have increased risk of deterioration if any corrosive washwater drips from valves, flanges, or access doors onto the top of the lower box where the TR sets are typically located.

11-10


Figure 11-3 61 Side Cut-Away View of Horizontal Flow Wet Electrostatic Precipitator

WESP units require a source of washwater to be sprayed near the top of the collector plates either continuously or at timed intervals. The water flows with the collected particles into a series of troughs from which the fluid is pumped or drained. A portion of the fluid may be recycled to reduce the total amount of water required. The washwater may be neutralized and reused in the process. The remainder is pumped to the FGD system to be used as makeup.

Tubular WESP The tubular design is most likely to be used when the WESP is located on top of an FGD absorber to serve as a combined mist eliminator and polishing step for fine particulate, aerosols, etc. Some major WESP manufactures are considering the use of plates in their vertical flow design to simplify design and maintenance. However, this approach has not yet been demonstrated at full scale. In a tubular WESP, the exhaust gas flows vertically through conductive tubes, with many tubes operating in parallel. Typically, these systems will use two or more stages in series because the high SO3 loading causes space charge effects in the first stage. The tubes may be formed as a circular, square, or hexagonal honeycomb. Square and hexagonal pipes can be packed closer

61

Figure from Alstom Wet Electrostatic Precipitators Presentation, August 2005.

11-11


together than cylindrical pipes, reducing wasted space. Pipes are generally 3–12 inches (80–300 mm) in diameter and 3–12 feet (1–4 m) in length. The high voltage discharge electrodes are long wires or rigid “masts” suspended from a frame in the upper part of the WESP that run through the axis of each tube. Both an upper and lower frame generally support the rigid discharge electrodes. Sharp points are added to the electrodes, either at the entrance to a tube or along the entire length in the form of stars, to provide additional ionization sites. Some major WESP manufactures are considering the use of plates in their vertical flow design to simplify design and maintenance. However, this new approach has not been demonstrated at full scale. Figure 11-4 shows a side view of a tubular WESP installed on top of a wet FGD system.

11-12


Figure 11-4 62 Example of Tubular WESP Installation above FGD

The power supplies for the WESP convert industrial AC current to pulsating DC current in the range of 20,000–100,000 volts, as required. Each electrode-pipe combination establishes a charging zone through which the particles must pass. Either conventional 60 Hz power supplies or the newer high-frequency supplies can be used. The high-frequency supplies have the advantage of being more energy efficient. Tubular WESP units also require a source of washwater to be sprayed near the top of the collector pipes either continuously or at timed intervals. When used as mist eliminators on top of the FGD, the water flows with the collected particles into the underlying wet FGD.

62

Figure from Alstom Wet Electrostatic Precipitators Presentation, August 2005.

11-13


Materials of Construction The collection section of a WESP must be conductive, but the material has to be resistant to the corrosive conditions caused by acid mist in the flue gas. Historically, WESP have been constructed of conductive fiberglass, carbon steel, stainless steels, and various high-alloy materials, depending on the duty. However, this experience comes from industrial applications. Power plant designs must reflect the conditions of coal-fired boiler flue gas, which is extremely variable. Material selection should be based on worst-case scenarios in order to protect the equipment against upset conditions and minimize the chances for a unit shutdown due to failure of the WESP. Materials being proposed for WESP units range from relatively mild alloys such as SS316 to 317LM, Alloy C-276, Alloy 2205, AL6XN, and even non-metallic fabric membranes. The WESP water treatment system should be considered in the material selection process as this system plays a large roll in determining the lifetime of the materials in a WESP. As the technology matures, materials may be selected for the corrosive resistance required in specific zones of the WESP. In addition, linings may come into use at appropriate locations in the WESP.

Emerging Technologies for SO3 Control Membrane WESP The membrane WESP is a new type of wet precipitator, in which fabric membranes replace the metal collecting electrodes. Testing indicates that membranes made from materials that transport liquid (primarily water) by capillary action are effective collection electrodes. Ca pillary flow promotes well-distributed water flow both vertically and horizontally which is necessary for particle collection, removal and transport. Membranes may avoid the tendency experienced in some WESP units of “beading” of the flushing water as it flows down the plate surface due to both surface tension effects and surface imperfections. Because beading causes the flushing liquid to be non-uniformly distributed over the surface, with channeling, dry spots are formed on the collecting plates. This inhibits current flow and results in increased emissions. It is hoped that membrane WESP will more reliably maintain rated performance.

Plasma-Enhanced WESP With parallel growth of interest in using WESP technology for mercury control, there may be a need to consider design tradeoffs to simultaneously optimize removal of SO 3 and removal of mercury. The plasma-enhanced WESP may be may be one tool used for this optimization, although current emphasis is on mercury removal. Mercury adsorbed onto particulate and mercury in oxidized forms is readily removed from the flue gas using a wet FGD system. However, the remaining elemental mercury vapor passes through the air pollution control devices. Most mercury research has focused on sorbent injection followed by a baghouse for plants without an FGD system (typically plants burning low-sulfur coal). Tests of WESP technology at the Bruce Mansfield pilot plant indicate that a WESP can remove both oxidized and particulate forms of mercury at levels similar to those for SO 3. To increase elemental mercury removal efficiencies using WESP technology, a new technology is being developed that uses the central electrode of a standard WESP to inject a mercury oxidizing reagent gas directly into the flue gas stream. As the reagent gas passes through this 11-14


corona discharge, reactive species are formed that subsequently oxidize elemental mercury into fine particulate. The mercury particulates acquire a negative charge in the electrical field and are attracted to positive collecting electrodes where they are adsorbed into the liquid layer and are removed from the process in the wash cycle.

Lime Spray Drying for SO 3 Removal A slight modification of the lime spray drying absorption (LSDA) p rocess is under investigation as a process for SO3 removal. This application differs from the LSDA process used for SO 2 removal by varying the degree of flue gas cooling. The spray dryer for SO2 removal lowers the flue gas temperature to as low as 30°F (17°C) approach to the adiabatic saturation temperature water. The lime spray drying process for SO 3 removal only lowers the temperature to approximately 300°F (150°C), which makes for good conditions for SO3 removal (and HCl removal), but minimizes SO2 removal.

Power Plant Applications of Sorbents for SO3 Control Most experience with injection of alkali sorbents for removal of SO3 from power plant flue gases is with units firing high-sulfur oil where trace metals, mainly vanadium, increase oxidation of SO2. Furnace injection of Mg(OH)2 (magnesium hydroxide) has been popular for this use because it reacts directly with SO3 to form magnesium sulfate, a water-soluble compound that is unlikely to form hard-to-remove deposits on equipment and ductwork. As noted above, a drawback of this process is the high cost of commercial magnesium hydroxide. Duke Energy’s 1400 MW Zimmer Unit 1 provides a notable case study for application of this SO3 control method on a coal-fired plant. Zimmer’s conventional, natural oxidation MEL (Thiosorbic®) FGD process was updated in 2000 with the addition of bleed stream oxidation and additional process equipment for recovery of Mg(OH)2 by-product. This process, shown in Figure 11-5, reduces the cost of Mg(OH)2 by roughly 70% compared with purchase of commercial magnesium hydroxide. The process was invented by Carmeuse Lime and developed in a pilot project at Zimmer in 1995. In 2004, in conjunction with addition of an SCR system, Zimmer began using its Mg(OH)2 by63 product for reduction of furnace-generated SO3. This followed successful full scale tests of the application, by DOE/NETL, at FirstEnergy’s 800 MW Bruce Mansfield Unit 3 and AEP’s 1300 64 MW Gavin Unit 1. The tests at Gavin and Mansfield showed capture of 90% of furnacegenerated SO3 and 70% of total SO3 for a plant equipped with SCR. At Gavin, testing found a total of about 2.3% of coal sulfur converted to SO3, with conversion of 1% occurring in the SCR catalyst.

63

L.B. Benson, K.J. Smith, R.A. Roden, E. Loch, and J. Potts, “Control of S ulfur Dioxide and Sulfur Trioxide Using By-Product of a Magnesium-Enhanced Lime FGD System,” Paper #130, Carmeuse North America Technology Center (http://www.carmeusena.com/Corporate/resources.asp?body=white_papers). 64 Furnace Injection of Alkaline Sorbents for Sulfuric Acid Removal, Final Report Cooperative Agreement No.: DE-FC26-99FT40718, (http://www.netl.doe.gov/coal/E&WR/pm/pubs/40718final.PDF).

11-15


Figure 11-5 Magnesium-Enhanced Lime SO3 Control Process with Bleed Stream Oxidation and Mg(OH)2 65 Recovery

At Zimmer, Mg(OH)2 slurry is injected into the furnace through air atomizing nozzles in the front wall of the furnace. The slurry is about 20% suspended solids with 65% of that being magnesium hydroxide. The design injection rate was based on an Mg-to-SO3 molar ratio of 8:1 assuming 1% conversion of SO2 to SO3. To obtain 90% removal of SCR-generated SO3, Zimmer also implemented injection of hydrated lime, as dry powder, downstream of the air heaters. The design rate for this injection was based on a Ca(OH)2-to-SO3 molar ratio of 7.7:1 assuming 1% conversion of SO2 to SO3 by the SCR catalyst and 4.1% sulfur in coal. As noted above, this process is comparatively economical to install on plants that already use variations of the Thiosorbic® MEL FGD process. This comprises about 17,700 MW of coalfired electric generating capacity in the United States, as shown in Figure 11-6.

65

Benson, op. cit.

11-16


Figure 11-6 66 Power Plants Using the Thiosorbic® Magnesium-Enhanced Lime FGD Process

It should be noted that the DOE tests applied four different calcium and magnesium-based sorbents in full-scale utility boilers. In addition to the magnesium hydroxide derived as a byproduct of the magnesium enhanced lime wet FGD process, DOE tested three commercially available sorbents: dolomite, pressure-hydrated dolomitic lime, and commercial magnesium hydroxide. At Bruce Mansfield, the dolomite was injected as a powder, whereas the other three reagents were injected as slurries in the upper furnace. The commercial Mg(OH)2 performance was nearly identical to the by-product Mg(OH)2. Dolomite and dolomitic lime required excessively high molar ratios to obtain desired performance at this injection location.

Power Plant Applications of WESP for SO 3 Control In the United States and Canada, as of the early 2000s, wet electrostatic precipitators were used 67 to reduce opacity and control sulfur trioxide at four power plants. AES Deepwater, firing petroleum coke, has the longest operating hours of these four. The largest of these four WESP installations, and the only WESP installation on a coal-fired plant in the United States, is at the

66

Ibid. R.C. Staehle et al., “The Past, Present and Future of Wet Electrostatic Precipitators in Power Plant Applications,” presented at the Combined Power Plant Air Pollutant Control Mega Symposium, DOE/EPRI/EPA/AWMA, May 1922, 2003, Washington, DC. (Report BR-1742, http://www.babcock.com/pgg/tt/pdf/BR-1742.pdf). 67

11-17


Xcel Energy Sherburne County Generating Station (Sherco). NB Power Dalhousie Station was converted for Orimulsion firing in 1995 and added a single-field WESP for SO 3 control in 2000. NB Power Coleson Cove Station added WESPs to its three 350 MW units during conversion in 2004 to allow Orimulsion firing.

AES Deepwater—WESP with 100% High-Sulfur Petroleum Coke and Future SCR AES Deepwater (Figure 11-7) is a 155 MW petroleum coke-fired cogeneration plant located on the Houston Ship Channel in Pasadena, Texas. This plant utilizes a dry electrostatic precipitator to limit the levels of particulates and unburned carbon entering the limestone-based, gypsum producing wet FGD system. A wet venturi scrubber removes additional particulate prior to the wet FGD.

Figure 11-7 68 AES Deepwater After WESP Installation

Deepwater commissioned the WESP in 1986 to reduce SO3 and opacity. The decision to add the WESP after the scrubber developed from the following observations: •

•

68

Although particulate limits are necessary for regulatory compliance, control of unburned carbon is required to prevent contamination of salable by-product gypsum. The petroleum coke fuel has a high vanadium content, which results in relatively high level of SO3 entering the wet FGD where the gas quenching action completes the formation of sulfuric acid mist. Only about 20–30% of the mist is captured in the FGD system because of the fine particle size of the mist droplets. The inlet concentration of SO 3 varies from 35 to

Source: Babcock & Wilcxox.

11-18


100 ppm at 3% O2 depending on furnace operating conditions and vanadium content in the petroleum coke. •

•

The limit for total particulates including sulfuric acid and con densables was set at 0.005 grains/scfd, thus requiring a collection level of >90% on sulfuric acid alone. SO3 and particulate control across the wet ESP is typically 95–97%. Additional SO3 removal occurs upstream of the WESP in the wet FGD. Additional particulate is collected by the dry ESP and the wet venturi scrubber ahead of the FGD.

In 2005, Deepwater was controlling opacity to about 8% while firing 100% on petroleum coke 69 with 5% sulfur content. The scrubber upstream of the WESP accomplishes about 7% of total SO3 removal. Sorbent injection, using MgO or sodium bisulfite, was considered prior to the decision to install the WESP. This evaluation determined that residence time in the existing ducting would be inadequate for the targeted SO3 reduction. The current system comprises are twelve WESP modules, of which eleven are needed for complete performance. Maintenance requirements keep one module out of service on a nearly continuous, rotating basis. Maintenance includes removal of scale build-up, adjustment of water flow distribution, correction of electrode and plate distance correction, and remediation of leakage. When the planned installation of SCR and low-NOX burners is completed in 2007, Deepwater will become the only plant in the U.S. that uses SCR while firing 100% petroleum coke (the AES Somerset plant in upstate New York uses SCR while firing 20% petroleum coke). After the SCR system commences service, WESP performance will be monitored closely to determine if an upgrade is needed, or if SO3 reduction can be accomplished with more frequent than normal catalyst change-out.

Sherco—Two 750 MW Bituminous PC Units with WESP Controlling Opacity At Xcel Energy’s Sherburne County Generating Station (Figure 11-8), 12 wet ESP modules were installed in 2000–01, on each of the two 750 MW units. Stack opacity has been controlled as intended during several years of successful operation of the wet ESP modules have. The project cost of just over $50 million is just over half of the estimated $90 million cost of a baghouse for the same application. The driving force behind the installations on Sherco Units 1 and 2 was the need to limit stack opacity below 20%. Sherburne County burns subbituminous coals with enough calcium oxide in the flue gas to absorb what SO3 exists. With no dry ESPs, WESPs were installed to address a problem of opacity that the combined particulate/wet scrubbers were not completely able to handle. Of the twelve 75 MW modules on each unit, only 10 modules are needed for full load.

69

Personal Communication, Leon Ballard, AES Deepwater Station to Tony Armor, EPRI, May 23, 2005.

11-19


Figure 11-8 70 Xcel Sherco Station

The following observations were documented during the initial years of WESP operation: •

•

•

•

•

•

Wet ESPs have limited stack opacity to about 10–12% (compared to pre-WESP levels of 30– 40%). Particulate removal efficiency has exceeded 90% with residence time of one second available within the wet ESP’s treatment zones. Due to the high calcium content of the fly ash, material scaling occurs in the bottom portion of the first field collector tubes. The first fields primarily capture the re-entrained droplets and carryover from the wet FGD thus allowing for relatively stable electrical operation for fine particulates capture in the second fields. Periodically, each module receives a thorough, off-line manual high pressure washdown to remove scale. Unit 1 modules have averaged about 6 months between cleanings whereas Unit 2 modules have averaged about one year. The difference is thought to result from the lowNOX burners on Unit 2 that produce an easier-to-clean softer ash. Normal daily operation includes water flushing of ea ch module with its power supply deenergized.

70

Energy Success Stories: Partners For Affordable Energy, Dave Heberling, General Superintendent for Environmental and Plant Services, Sherco, Ron Elsner, Project Manager, www.powerofcoal.com/environmental_success/details.

11-20


The conditions that require more frequent cleaning for Unit 1 also lead to some other maintenance issues. To avoid corrosion related to “arcing and sparking” on the lower field electrodes and collector tubes, the Unit 1 wet ESP modules will be equipped with two-pass demisters. The upper fields have not had this problem and go 4-5 years without cleanings. Also, installation of 316 SS sleeves over the original 304 SS electrodes has helped with corrosion control. It was originally thought that the alkaline nature of fly ash and absence of sulfuric acid 71 mist in this application allowed the use of 304L stainless steel for the WESP internals. Other lessons learned include: •

•

Wet ESPs are able to overcome difficult conditions for dust build-up through scheduled wash downs. Wet ESPs are able to efficiently collect non-acid particulate, as well as sulfuric acid mist.

Dalhousie—WESP with Orimulsion-Firing A single-field wet ESP system was installed on the 312 MW Dalhousie plant in 2000. Wet FGD was added as part of the conversion for Orimulsion firing completed in 1994. Dalhousie began firing 100% Orimulsion in 1995. Orimulsion 400 is an emulsion of 30% water and 70% natural bitumen extracted from the Orinoco belt in Venezuela. On an HHV basis, it has roughly 50% more sulfur than No. 6 fuel oil. Like heavy fuel oil, Orimulsion has high vanadium content that catalyzes conversion of SO2 to SO3 in the furnace.

Coleson Cove—Three 350 MW Oil-Fired Units Experience at Dalhousie provided the background for modifying three 350 MW units at NB 72 Power’s Coleson Cove Station to fire Orimulsion. As part of this refurbishment, two 525 MW FGD absorbers were installed to control SO2 emissions from the three 350 MW units. Integrated, upflow wet ESP modules were installed on top of the scrubbers to control SO3 emissions to below 5 ppmvd @ 3% O2, limit fly ash particulates below 0.015 lb/MBtu (6.5 g/GJ), and control opacity. After a fuel supply agreement could not be reached, the plans for Orimulsion were discontinued and the plant reverted to firing No. 6 fuel oil from Venezuela with 1.5–2.5% sulfur.

71

Personal Communication, Bob Henningsgard, Xcel Energy Sherburne County Station to Tony Armor, EPRI, May 25, 2005. 72 Coleson Cove Orimulsion Conversion Project , R. Telesz, K. MacLean, R. Ojanpera, Combined Power Plant Air Pollutant Control Mega Symposium, EPRI/DOE/EPA/AWMA, May 19-22, 2003, Washington, DC. (Report BR 1740: http://www.babcock.com/pgg/tt/pdf/BR-1742.pdf).

11-21


Figure 11-9 73 Coleson Cove Shown Prior to the Installation of Wet FGD and Wet ESP Systems

Considerations for the Coleson Cove refurbishment and WESP installation included the following: •

•

•

•

•

73

The plant’s original burners (which had uncontrolled NOX emissions of 0.9 lb/MBtu or 0.39 kg/GJ) were replaced with low-NOX burners and reburning. The WESP units use three electrical fields in series and four independently energized high voltage bus sections across each wet ESP electrical field. This conservatively sectionalized, twelve bus section design allows for small sections to be d e-energized during periodic water flushings while maintaining overall emissions within design levels. To achieve the desired level of control on sulfuric acid at all times, collection efficiency requirements had to exceed 90%. Figure 11-10 shows a schematic of the layout for the Coleson Cove plant for each of the two wet FGD absorbers with a three-field upflow type wet ESP mounted at the top of each. This design is similar to the one for AES Deepwater. Scrubbed flue gas enters the inlet field of the wet ESP after exiting the wet FGD mist eliminator. The gas exits the top of the wet ESP through a final mist eliminator section which captures any re-entrained droplets that may be present during flushing cycles, and transitions directly into the stack through an outlet hood. This design lowers the balance of plant costs compared with the wet ESP as a stand-alone system outside of the wet FGD vessel.

The Coleson Cove plant is located on a shoreline, with laydown space at a premium. This also favored the integrated arrangement of the wet FGD absorber and wet ESP.

Ibid.

11-22


Figure 11-10 74 Wet ESP Arrangement for Coleson Cove Station

As the WESP units were being prepared for final acceptance testing in 2005, the stack was reported to be “dramatically cleaner,” no longer producing its prior SO3 plume, while SO3 at the 75 economizer outlet was 23 ppm at full load and 40 ppm at low load. One reported problem was the fit up around the beams supporting the massive 56-foot-diameter wet WESP. This allowed the ceramic insulators to be exposed to dirt. Continuous purge air was introduced to keep the components clean.

74

R.C. Staehle et al., op. cit. Personal Communication, Keith Maclean, New Brunswick Power to Tony Armor, EPRI, May 27, 2005.

75

11-23


12

MULTI-POLLUTANT CONTROL SYSTEMS Multi-pollutant control systems seek to reduce costs relative to emissions control approaches using independent systems for each pollutant. Expected savings are in the range of 10–25% compared with separate controls for NOX, SOX, and Hg. The sale price of by-products is often the key to favorable economic performance with these processes. Multi-pollutant control systems generally use complimentary chemical reactions and physical phenomena for multiple ends. The term generally implies greater integration of processes than is seen with current efforts to leverage removal of a pollutant through relatively minor modifications to other environmental processes (for example, selection of an SCR catalyst with increased tendency to oxidize mercury to form solid mercuric oxide). Among the 50 plus processes in development, a few leaders are far enough advanced to be considered for plants needing to add SCR, FGD, and Hg controls in the next two to four years. 76,77

Powerspan ECO Process

Powerspan’s Electro-Catalytic Oxidation (ECO®) process is one example of an integrated multipollutant control technology that is nearing commercial viability. The description provided here shares many common elements with other multi-pollutant technologies. The ECO process achieves reductions in emissions of NO X, SO2, fine particulate matter, and mercury from the flue gas of coal-fired power plants. The technology also reduces emissions of other air toxic compounds such as arsenic and lead, as well as acid gases such as hydrochloric acid. The ECO process is installed downstream of the air heater and the particulate collector (ESP or fabric filter) that collects fly ash and other particulate. ECO was successfully pilot tested and demonstrated at FirstEnergy Corporation’s R.E. Burger Plant. The pilot plant has been operating since 2002. The 50 MW commercial demonstration unit has been operating since 2004.

Three-Step Processing of Flue Gas ECO treats flue gas in three steps using a barrier discharge reactor, an ammonia-based scrubber, and a wet electrostatic precipitator. The reactor and the scrubber bring about a series of reactions that occur at relatively low temperatures (150–350°F, or 65–180°C) to absorb or adsorb gaseous pollutants with solid and liquid compounds that can be physically removed from the flue gas stream. The scrubber and the WESP perform this removal.

76

C. McLarnon and M. Jones, “Electro-Catalytic Oxidation Process for Multi-Pollutant Control at FirstEnergy’s R.E. Burger Generating Station,” Presented at Electric Power 2000, Cincinnati, April 5, 2000. 77 C. McLarnon et al., “Final Technical Report for Mercury Removal in a Non-Thermal, Plasma-Based MultiPollutant Control Technology for Utility Boilers,” Final Report to U.S. DOE, December 2004.

12-1


Oxidation The barrier discharge reactor generates a non-thermal plasma in which energetic electrons create hydroxyl (OH) and atomic oxygen (O) radicals that oxidize gaseous pollutants to higher oxides: • •

•

Nitric oxide (NO) is oxidized to nitrogen dioxide (NO 2) and nitric acid (HNO3) A small portion of sulfur dioxide (SO2) is converted to sulfuric acid (H 2SO4) at higher moisture levels and lower temperatures Elemental mercury (Hg) is oxidized to mercuric oxide (HgO)

Scrubbing A dual-loop ammonia scrubber follows the barrier discharge reac tor. As flue gas travels upwards, after entering near the bottom of the scrubber tower, the lower circulation loop quenches it to saturation. The upper loop removes SO2 and oxides of nitrogen and neutralizes the acids, forming ammonium sulfate and ammonium nitrate, which are high value fertilizer products. High levels of NOX and SO2 removal are enabled by the conversion of NO to NO2 in the reactor and the synergy between SO2 scrubbing and NO2 capture in the scrubber. Without these two effects, NO and NO2 would pass through the scrubber.

Capture of Aerosols and Particulates Some of the particulates that have bypassed the ESP, along with acid aerosols created in the reactor, are captured by the sprays in the scrubber absorber tower and carried to the basin at the bottom of the tower. At the same time, the upward flow of flue gas entrains droplets of reagent solution, along with particulates formed or captured in the solution, and carries them toward the WESP located above the scrubber in the absorber tower. Before the flue gas reaches the WESP, mist eliminators remove larger droplets and particulates. The WESP then captures most of the fine acid aerosols produced by the discharge reactor, aerosols generated in the scrubber, and fine particulate matter, which includes oxidized mercury.

Collection of Liquid Streams A separator tray captures most of the reagent and reaction product circulating in the upper loop and returns it to the Upper Loop Recycle Tank. The quench sprays and some carryover from the upper loop fall into the Reaction Basin at the bottom of the tower. Another tray captures the runoff of makeup/flush water and solids from WESP along with recycle slurry used to flush the mist eliminators. Except for a small amount that escapes through the WESP, all liquids and solids that enter the absorber tower ultimately end u p in the reaction basin, which is also the recycle tank for the lower quench loop. This includes: •

All material removed from the flue gas, which is in the form of dissolved ammonium sulfate and nitrate salts, dissolved and suspended mercury and other metals, and captured fine particulate matter. The concentration of solid material in the liquid is very low and the liquid is basically clear.

•

Reagent and makeup water introduced to the upper loop

•

Makeup water used to flush the WESP

12-2


Air is blown into the reaction basin to complete the oxidation of any remaining sulfite or nitrite and to keep mercuric oxide from being reduced back to elemental mercury. Agitators help distribute the air and keep the small amount of solids in suspension. The solution in the tank is kept near the concentration at which ammonium sulfate saturates and begins to crystallize by the simultaneous evaporation of water, during quenching, and removal of a liquid bleed stream, which is sent to the product processing system.

By-product Recovery The bleed stream of concentrated, clear liquid presents a very convenient opportunity to remove constituents not desired in the fertilizer by using simple filters and ab sorbent beds. The relatively high concentration of the material in a clear liquid stream makes this processing significantly easier and cheaper than could be done with solids or slurries. After the unwanted compounds are removed, the ammonium sulfate and ammonium nitrate are crystallized, separated from the liquid flow by centrifuge, and granulated to produce a commercially valuable fertilizer. The ammonium nitrate is not separable from the ammonium sulfate; it adds to the value of the fertilizer because it increases the nitrogen content. There is no liquid discharge from an ECO system. The concentrate is returned to the reaction basin. The only streams leaving the process are the granulated co-product stream and the low volume of concentrated solids removed from the liquid before it is processed into fertilizer.

Performance Data and Other Considerations Reported results from 6 months of testing at the 50 MW demonstration facility include: •

90% NOX reduction

•

98% SO2 reduction

•

85% Hg reduction

•

~0.01 lb/MBtu (10 mg/Nm ) particulate

3

Ongoing investigation is addressing the tendency of some reaction products to reverse the oxidation of mercury accomplished in the barrier discharge reactor. The steady-state mercury concentration must be maintained at a low level to minimize the rate of mercury reduction and re-emission. A CO2 capture add-on module is under development.

Other Multi-Pollutant Processes ReACT Process The Regenerative Activated Coke Technology (ReACT) process is a second generation Bergbau Forschung (Mitsui) process marketed worldwide by J-Power EnTech Inc. Upgrades to the process, which uses ammonia injection with a moving bed of activated coke, include use of multiple coke flowpaths and a form of coke pellets with reduced attrition. The original and enhanced versions of the process have been used in Japan on several industrial facilities, including the 350 MW Takehara AFBC plant and the 600 MW Isogo Unit 1 PC plant. Reported

12-3


performance includes Hg reductions up to 98%, SO2 reductions of 90–98%, and NOX reductions of 50–80%. NOX reduction performance is reported to be best when fuel sulfur content is below 2%. In the ReACT process, ammonia is added to the flue gas before it enters an adsorber vessel. The NH3 reacts with SO2 to form sulfuric acid, ammonium sulfate, and ammonium bisulfate. These products are subsequently adsorbed onto the bed material along with mercury. NOX is catalytically reduced, on the coke surface, reacting with NH3 to form nitrogen and water. Coke is circulated between the adsorber and regenerator by conveyor. After the coke is transferred to the regenerator vessel, it is heated indirectly to 750–930°F (400–500°C). This releases SO2, which is processed into sulfur, sulfuric acid, or other products. The coke is cooled and screened to remove fines and captured fly ash before being returned to the adsorber vessel. The mercury is retained on the coke until the coke is removed as fines or during a changeout every few years. A 2.50 MW equivalent test to investigate the performance of the ReACT process with U.S. bituminous and subbituminous coals is scheduled to begin in spring of 2007 on one of the ~250 MW units at Idaho Power and NRG Energy’s Valmy Generating Station in Nevada.

Airborn Process The Airborn Process uses sodium bicarbonate injection and wet sodium scrubbing with an oxidant. Airborn Technologies Inc. (ATI) developed this enhancement of the long-established wet sodium scrubbing process by using ammonia to regenerate sodium sulfate. This produces high purity (+98%) sodium bicarbonate for reuse and (+99%) ammonium sulfate, which is sold as fertilizer to improve the overall economics of the process. The Airborn Process is also able to capture NOX and oxidized mercury. Work is in progress to add CO2 capture. Targeted performance includes SO2 removal >99.5%, SO3 removal >99.5, NOX removal >98%, and oxidized Hg removal >98%. In experiments using different oxidants, ATI has reported near 100% capture of NOX and near 100% capture of elemental mercury, which otherwise does not dissolve and passes through the process. Data available to date are from small scale tests at CANMENT Energy Technology Center in Ottawa and a 5 MW pilot at Kentucky Utilities’ Ghent Station. Oxidant screening tests were carried out at the Energy and Environmental Research Center (EERC) in North Dakota.

Mobotec ROFA/ROTAMIX Process The ROFA/ROTAMIX process aims to enhance combustion and reagent performance through optimized mixing. The Rotating Overfire Air (ROFA) process uses off-center injection of high velocity air to produce high turbulence and rotating swirl. Enhanced mixing reduces thermal NOX production by creating more even combustion. The mixing also enables more complete combustion, which improves carbon burnout, reduces carbon monoxide, and allows use of lower excess air levels. Rotamix injects reagent/sorbent through the ROFA ports—ammonia or urea for enhanced NOX control, limestone or trona for SO X control, and MinPlus for Hg control. The ROFA process has been installed on thirteen coal-fired PC boilers with 2470 MW cap acity in the United States. The combined process is being tested on Dynegy’s 89 MW Vermilion Unit 1 and Carolina Power and Light’s 174 MW Cape Fear Unit 6. Early field results show Hg

12-4


reductions up to 65–90%, SO2 reductions of 65–70%, SO3 reductions of 90%, and NOX reductions up to 75%. Addition of SCR is expected to reduce NOX further. Particulate control also appears to improve due to better distribution across the ESP with the improved mixing in the furnace.

12-5


A

TERMINOLOGY TERMINO LOGY,, ABBREVIATIONS, ABBREVIATIONS, AND A ND ACRONYMNS ABMA

American Boiler Manufacturers Association Standards

AWS

American Welding Society Standards

ABS

ammonium bisulfate

AWWA

ACI

American Concrete Institute

American Water Works Association Standards

AGMA

American Gear Manufacturers Association

BACT

best available control technology

AISC

American Institute of Steel Construction

barg

bars, gauge

BEC

bare erected cost

ALS

ammonium lignosulfonate

BOD

biological oxygen demand

AMCA

Air Movement and Control Association Standards

boiler

ANSI

American National Standards Institute

in subcritical plants, the equipment that produces the steam for power production; the steam generator.

API

American Petroleum Institute

Btu

British thermal unit

ARI

Air Conditioning and Refrigeration Institute

CCPI

Clean Coal Power Plant Initiative of the United States DOE

ASA

Acoustical Society of America Standards

CEMS

continuous emissions monitoring system

ASCE

American Society of Civil Engineering

CF

capacity factor

ASHRAE

American Society of Heating, Refrigerating and AirConditioning Engineers

CFB

circulating fluidized bed

CFR

Code of Federal Regulations

COE

levelized busbar cost of electricity, often expressed in $/MWh

ASHRAE

American Society of Heating, Refrigeration, and Air Conditioning Standards

CTI

Cooling Technology Institute

ASL

site elevation above mean sea level

DAF

dry and ash-free; synonym for MAF

ASME

American Society of Mechanical Engineers

DCS

digital control system

ASNT

American Society of Nondestructive Testing

DNB

departure from nucleate boiling

DI

de-ionized

ASSE

American Society of Sanitary Engineering Standards

DOE

United States Department of Energy

ASTM

ASTM International, originally known as the American Society for Testing and Materials

dscf

dry standard cubic feet

dscm

dry standard cubic meter

EAF

equivalent availability factor

A-1


EIA

Electronic Industries Association

EJMA

Expansion Joint Manufacturers Association

a bomb calorimeter at 77°F (25°C) is included) HI

Hydraulics Institute

EPA

United States Environmental Protection Agency

HMI

Hoist Manufacturers Institute Standards

EPC

engineering procurement and construction contract or contractor

HMI

human/machine interface

HP

high pressure

HVAC

heating ventilation and air conditioning

IBC

International Building Code 2000

ID

Induced Draft (fan)

IDC

Interest during construction

IEEE

Institute of Electrical and Electronics Engineers

IP

intermediate pressure

I.R.

infrared radiation

ISA

Instrumentation, Systems, and Automation Society

EPRI

Electric Power Research Institute

FAA

Federal Aviation Administration

FAC

flow accelerated corrosion

FD

forced draft (fan)

FDA

flash dryer absorber

FM

Factory Mutual Insurance Company

FOF

forced outage factor

FOR

forced outage rate

fps

feet per second

FT

fluid temperature

ISBL

Inside Battery Limits

FW

Foster Wheeler

ISO

GADS

Generating Availability Data System

International Organization for Standardization

ISO

International Standards Organization

K.O.

knockout drum

kW, kWe

kilowatt electric

kWt

kilowatt thermal

LAER

lowest achievable emissions rate

LBtu

low Btu content coal

LHV

lower heaving value (fictitious heating value of fuel if the heat of condensation of water vapor is ignored; i.e., if the water in the combustion products were assumed to remain in the vapor state in the bomb calorimeter)

LOI

loss on ignition

LP

low pressure

MAF

moisture and ash-free; synonym for DAF

MCR

maximum continuous rating

GE

the General Electric Company

generator

a device for producing electricity; the piece of rotating machinery that produces electricity by inducing electrical current in stationary coils by turning a rotating electromagnet within them. Not to be confused with “steam generator,” a device that produces steam.

GTAW

gas tungsten arc welding

HAP

hazardous air pollutants

HDPE

high density polyethylene

HEI

Heat Exchange Institute Standards

Hg

mercury

HHV

Higher Heating Value (heating value of fuel if heat of condensation of water vapor from its combustion products in

A-2


6

MBtu

10 Btu, (million Btu)

ppmv

parts per million by volume

MPTA

Mechanical Power Transmission Association

ppmvd

ppmv, dry basis

ppmvw

ppmv, wet basis

MS

main steam

PRB

Powder River Basin

MSS

Manufacturers Standardization Society Standards

PSD

Prevention of Significant Deterioration

MTPD

metric tons/day

psia

MW, MWe

megawatt electrical

lb/square inch (14.696 psia = 1 atm)

MWt

megawatt thermal

psid

NA, N/A

not applicable

lb/square inch difference, used for pressure drop

NAAMM

National Association of Architectural Metal Manufacturers

psig

lb/square inch gauge [ (psia) – (local atmospheric pressure in psia) ]

NAAQS

national ambient air quality standards

PSM

Process Safety Management

PVC

polyvinyl chloride

NAVFAC

Naval Facilities Engineering

Q

heat

ND

not detected

RBD

Reliability Block Diagrams

NEMA

National Electrical Manufacturers Association

RCRA

Resource Conservation and Recovery Act

NERC

North American Electric Reliability Council

RH

reheat

RMA

Rubber Manufacturers Association

RMP

Risk Management Plan

RMS

root-mean-square

RO

reverse osmosis

NETL

DOE National Energy Technology Laboratory

NFPA

National Fire Protection Association

NOX

nitrogen oxides

NPDES

National Pollutant Discharge Elimination System

S

sulfur content of fuel

SAMA

New Source Performance Standards

Scientific Apparatus Manufacturers Association

scf

standard cubic feet

NSR

New Source Review

SCR

O&M

Operating and Maintenance

selective catalytic reduction (for NOX control)

OEM

original equipment manufacturer

SMACNA

OHSA

Occupational Health and Safety Administration

Sheet Metal and Air Conditioning Contractors' National Association

SNCR

P&ID

piping and instrumentation diagram

selective non-catalytic reduction (for NOX control)

SO2

sulfur dioxide

PC, pc

pulverized coal Pipe Fabrication Institute

SOX

sulfur oxides

PFI

SSPC

PHA

Process Hazard Analysis

The Society for Protective Coatings

PM

particulate matter

ST

steam turbine

NSPS

A-3


steam generator the equipment in a subcritical or supercritical power plant that produces the superheat and reheat steam for power generation from feedwater; in a subcritical plant, also provides for the evaporation (boiling) of water; synonymous with “boiler” for subcritical plants. Supercritical plants do not “boil” water, there is no phase change under supercritical pressure, so “steam generator” is the preferred term for this equipment in supercritical units. The steam generator is a device for producing steam, and should not to be confused with a generator, which is a device for producing electricity. SWS

sour water stripper

t

short ton (2,000 lbs)

T&D

transmission and distribution (electrical)

t/h, tph

short tons per hour (2000 lb/h)

t/y,tpy

short tons per year (2000 lb/y)

T250

the temperature at which the slag viscosity is 250 poise

TBD, tbd

to be determined

TCLP

toxicity characteristic leaching procedure

TEMA

Tubular Exchanger Manufacturers Association

TG

turbine-generator, (turbine-generator)

TOC

total organic carbon

ton

short ton, (2000 lb)

tonne

metric ton, (1000 kg or 2205 lb)

TPC

total plant cost

UL

Underwriters Laboratory

US, U.S.

United States

USACE

United States Army Corps of Engineers

USC

ultra-supercritical steam plant

USD, US$

United States dollar

A-4

USDOE

United States Department of Energy

USEPA

United States Environmental Protection Agency

UTS

Universal treatment standards

VHP

very high pressure turbine section

VOC

volatile organic compounds

VWO

valves wide open

y, yr

year

ZLD

zero liquid discharge

Advanced Pulverised Coal Power Plants EPRI 2007 208p

Recommend Documents