Flow Assurance Guide

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002 Page 1 of 128

ULTRA DEEPWATER PROJECT

FLOW ASSURANCE GUIDE Flow Management

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Table of contents

1.

PRESENTATION ..........................................................................................................................................5 1.1 GENERAL INTRODUCTION................................................................................................................5 1.2 THE CHALLENGES OF FLOW ASSURANCE ..................................................................................................7 1.3 DEVELOPING A FLOW ASSURANCE STRATEGY (FILE) ...............................................................................8 1.3.1 Data .................................................................................................................................................9 1.3.2 Risk assessment & Pre-sizing...........................................................................................................9 1.3.3 Management.....................................................................................................................................9 1.3.4 Sensitivity Studies...........................................................................................................................10 1.3.5 Project Studies ...............................................................................................................................10 1.3.6 Start-up and Production ................................................................................................................10 1.3.7 Flow Assurance file........................................................................................................................10 1.4 KEY PERSONNEL INVOLVED IN THE FLOW ASSURANCE GUIDE ................................................................12

PART 1 – PHYSICO-CHEMICAL....................................................................................................................13 2.

HYDRATES .................................................................................................................................................15 2.1 THE ISSUES .............................................................................................................................................15 2.2 DATA REQUIRED .....................................................................................................................................18 2.3 METHODOLOGY ......................................................................................................................................19 2.3.1 Formation Prediction (thermodynamic range) ..............................................................................19 2.3.2 Evaluating the Risks of Plugging ...................................................................................................21 2.4 SOLUTIONS .............................................................................................................................................23 2.4.1 Prevention ......................................................................................................................................23 2.4.2 Hydrate solution and limitations for Dry Trees associated with wellhead platform .....................26 2.4.3 Monitoring .....................................................................................................................................27 2.4.4 Curative Methods...........................................................................................................................28

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3.

ORGANIC DEPOSITS................................................................................................................................32 3.1 WAXES .................................................................................................................................................32 3.1.1 The issues .......................................................................................................................................32 3.1.2 Data Required................................................................................................................................33 3.1.3 Methodology ..................................................................................................................................36 3.1.4 Solutions.........................................................................................................................................41 3.2 ASPHALTENES....................................................................................................................................44 3.2.1 The issues .......................................................................................................................................44 3.2.2 Data Required................................................................................................................................45 3.2.3 Methodology ..................................................................................................................................46 3.2.4 Solutions.........................................................................................................................................48 3.3 NAPHTENATES...................................................................................................................................49 3.3.1 The issues .......................................................................................................................................49 3.3.2 Data Required................................................................................................................................49 3.3.3 Methodology ..................................................................................................................................50 3.3.4 Solutions.........................................................................................................................................52

4.

SCALE ..........................................................................................................................................................55 4.1 4.2 4.3 4.4

5.

THE ISSUES .............................................................................................................................................55 DATA REQUIRED .....................................................................................................................................57 METHODOLOGY ......................................................................................................................................58 SOLUTIONS .............................................................................................................................................61

MISCELLANEOUS PROBLEMS .............................................................................................................64 5.1 RHEOLOGY..........................................................................................................................................64 5.1.1 Definitions......................................................................................................................................64 5.1.2 The issues .......................................................................................................................................64 5.2 RHEOLOGY OF CRUDE OILS ...........................................................................................................65 5.2.1 Data required.................................................................................................................................65 5.2.2 Methodology ..................................................................................................................................65 5.2.3 Solutions.........................................................................................................................................68 5.3 RHEOLOGY OF EMULSIONS............................................................................................................69 5.3.1 Data required.................................................................................................................................69 5.3.2 Methodology ..................................................................................................................................69 5.3.3 Solutions.........................................................................................................................................71 5.4 SEPARATION OF EMULSIONS .........................................................................................................73 5.4.1 General ..........................................................................................................................................73 5.4.2 Required Data................................................................................................................................73 5.4.3 Methodology ..................................................................................................................................73 5.4.4 Solutions.........................................................................................................................................73

PART 2 - MULTIPHASE TRANSPORT ..........................................................................................................75 6.

HYDRODYNAMICS AND PRODUCTIVITY .........................................................................................77 6.1 THE ISSUES .........................................................................................................................................77 6.1.1 Productivity Calculations : ............................................................................................................77 6.1.2 Stability issues................................................................................................................................78 6.1.3 Flow enhancement .........................................................................................................................79 6.2 REQUIRED DATA ...............................................................................................................................79 6.3 METHODOLOGY ................................................................................................................................80 6.3.1 Steady State calculations ...............................................................................................................81 6.3.2 Stability studies ..............................................................................................................................83 6.3.3 Flow enhancement .........................................................................................................................86 6.4 SOLUTIONS .........................................................................................................................................87

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.

THERMAL MANAGEMENT....................................................................................................................96 7.1 THE ISSUES .........................................................................................................................................96 7.2 REQUIRED DATA ...............................................................................................................................97 7.3 METHODOLOGY ................................................................................................................................97 7.3.1 Under permanent flowing conditions.............................................................................................98 7.3.2 Under shutdown and restart conditions.........................................................................................99 7.4 SOLUTIONS .......................................................................................................................................101 7.4.1 Passive Insulation ........................................................................................................................102 7.4.2 Active Heating Systems ................................................................................................................105 7.4.3 Temperature monitoring ..............................................................................................................110 7.4.4 Hydrate prevention ......................................................................................................................111

8.

EROSION / CORROSION .......................................................................................................................113 8.1 TERMINOLOGY ......................................................................................................................................113 8.2 POSITION OF THE PROBLEM ...................................................................................................................113 8.2.1 Main materials concerned ...........................................................................................................114 8.2.2 Main equipment concerned..........................................................................................................114 8.2.3 Main fluid effects..........................................................................................................................115 8.3 DATA REQUIRED ...................................................................................................................................116 8.4 METHODOLOGIES ..................................................................................................................................117 8.4.1 Risk assessment ............................................................................................................................117 8.4.2 Erosion-corrosion prevention ......................................................................................................122 8.4.3 Erosion-corrosion monitoring .....................................................................................................123

9.

WELL METERING ..................................................................................................................................127 9.1 9.2

THE ISSUES .......................................................................................................................................127 SOLUTIONS .......................................................................................................................................127

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Presentation


1. PRESENTATION

1.1 GENERAL INTRODUCTION This guide has been drawn up within the framework of the activities of the PRODUCTION OPERATION Part of the Ultra Deepwater (UGF) R&D Project. It is based on experience acquired within the R&D activities (UGF and Maîtrise des Dépôts) and within the Deepwater field development projects. The document has involved specialists from TDO/EXP, TDO/TEC, SCR/RD and SCR/ED. This guide is intended for anyone seeking information on Flow Assurance related to Deep Offshore developments, especially peoples involved in Field development, Design, Construction and Production phases. The aim of this guide is to highlight concerns related to multiphase flow production, to give the level of knowledge and understanding of phenomenon, explain and detail the studies to be carried out to handle and quantify associated risks and finally describe solutions to be implemented in order to prevent or remediate the problems. As subsea production is a key element to deepwater production, this document focus on problems and solutions developed for subsea systems. More traditional developments with dry trees and direct access meet similar constraints and solutions as those identified in this document. In the same way, the document is more dedicated to oil dominated system because of our deepwater history. As deepwater production for oil dominated systems is more stringent in terms of Flow Assurance risks and solutions, general strategy can therefore be transposed to gas systems. Other Flow Assurance documents are available as: Ø

Flow Assurance Design guide developed by DEEPSTAR phase IV.

Ø

Internal study “Etude Architecture 3000” identifying technology gaps.

Ø

TECHNOSCOOP magazine N° 23 August 2000 “Flow Assurance”.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



For greater clarity and easy reading, the document is divided into two main parts: PART 1 – PHYSICO-CHEMICAL Part I deals with physical and chemical aspects of flow problems, broken down into the four following themes: Ø HYDRATES. Ø ORGANIC DEPOSITS (waxes, asphaltenes and naphtenates). Ø SCALE DEPOSITS (carbonates, sulphates and halite). Ø MISCELLANEOUS PROBLEMS (rheology, gels and emulsions). PART 2 – MULTIPHASE TRANSPORT Part II covers MULTIPHASE TRANSPORT of fluids, discussed around four themes: Ø HYDRODYNAMICS and PRODUCTIVITY (friction losses, instabilities and artificial lift), Ø THERMAL MANAGEMENT Ø EROSION – CORROSION Ø WELL METERING & MONITORING We have tried to use the same presentation in our coverage of the different themes in both parts: - an introduction to the ISSUES, aimed at defining the nature of the problem, - an inventory of the DATA needed to conduct the study, in order to understand the phenomena and identify the problems. - a presentation of the general METHODOLOGY, explaining how the studies proceed and what resources are implemented, - and finally, a description of the SOLUTIONS that can be considered, to prevent or remedy the different problems.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



1.2 The challenges of Flow Assurance Specific characteristics for Deep offshore and Ultra Deep offshore Subsea production is seen as a key element to develop deepwater fields from large stand alone to small remote single well production. The ability to transport the fluids (oil, water and gas) from the wells to the host facilities is the major concern for Flow Assurance. Deep offshore and, to an even greater extent, Ultra Deepwater developments (1500-3000m), are characterised by: Ø Very low ambient temperatures, around 4°C from depths of 500 m down and even less than 2°C around 3,000 m. Ø High pressures in the production lines (flowlines & risers). Ø Laterally extensive, shallowly buried reservoirs. Ø Harsh environment, which prevents easy access into the system for remediation.. The economic considerations of these projects mean we can only take an interest in wells that are highly productive. Accumulations that are widely spread one from another have to be developed together, whereas onshore they would be covered by individual developments. As a result, the subsea evacuation system is an extensive, complex network, combining several high productivity wells. In this context, any incident is likely to lead to significant losses in production, requiring interventions that will be difficult, sometimes even impossible and in all events extremely expensive. Flow Assurance objective The emergence of flow assurance as a discipline is driven by the combination of the hostile environment, challenging fluid properties, and system reliability targets associated to deepwater production systems. The overall objective of flow assurance is simply to “keep the flow path open”, to insure uninterrupted flow of production from subsea reservoirs to host facilities at minimum capital and operating costs. Deepwater production systems must be conservatively designed with high front-end capital equipment to overcome all perceived flow assurance problems to maintain production. The Operator's main concern is therefore to focus on “Flow Assurance”. This means taking major problems such as the following into account:

Ø Formation of hydrates, Ø Deposits, Ø Gelling point, Ø Productivity and Pressure drops Ø Flow Stability. Deposition and associated plugging by both hydrates or waxes in lines are the largest risks currently limiting oil extended reach subsea tieback (in excess of 25-30 Km). Other potential deposits such as scale or asphaltenes are also to be addressed but generally at a lower priority. Associated with the production of fluids, cooling and reduction in pressure from the reservoir to the host facility are inescapable. This may induce phase changes including the separation of gas and

TotalFinalElf E&P

DGEP/SCR/RD/UGF



liquid, the phase slippage or the formation of solids. Multiphase transport of fluids presents number of challenges to be managed in order to assure flow and ensure the integrity of the production system. These issues tend to be fluid and system specific. Flow Assurance studies strategy Flow Assurance studies consider hydrodynamic and thermal behaviour, together with the thermodynamic, chemical and physical characteristics of the effluents. They therefore entail thorough analysis of the fluid stream, from the reservoir to the surface installations. Their results are used to identify potential problems to be tackled as early as possible in the development design phase, and to propose solutions in line with the project technical and cost objectives, such as: -

Quantifying and managing production-related risks to avoid costly interventions on wells and subsea equipment

-

Choosing the most appropriate subsea multiphase technology

-

Proposing or developing back-up solutions and curative techniques

-

Optimising the design specifications for fluid management.

This approach is initiated in the pre-project phase and continues well beyond the construction and installation phases throughout the duration of the field's operation. The preferred current strategy available for dealing with hydrates and waxes is still to insulate and possibly heat the fluids to prevent them from cooling into the solids formation region during normal operation and for limited shut-in times. Remote subsea production in deepwater rely on a combination of prevention and remediation solutions to guarantee uninterrupted production, flow assurance strategies involve an association of equipment design & selection, operational methodologies and chemical treatments. Flow Assurance studies focus on developing prediction tools as well as identifying and developing prevention and curative solutions, in order to better quantify the risks of plugging a production system and in qualifying or improving the performance of prevention, inhibition and remediation technologies. Flow Assurance mission is to identify, promote and develop new technologies that insure the uninterrupted flow of production from subsea reservoirs to process facilities at minimum capital and operating costs without undue risks to environment, personnel and installation. For the current near term (3-5 years) the objective is to optimise current design practices and operation, to improve prediction software by analysing acquired data during operation and in parallel identify, evaluate and improve prevention and remediation technologies. Long term strategy has to provide technologies to allow for the minimal CAPEX/OPEX production system as single un-insulated flowline. Associated with this minimal design the ability to quickly and safely remediate any potential restriction is crucial (low cost intervention system).

1.3 Developing a Flow Assurance Strategy (File) Flow Assurance concerns are not unique to deepwater environments but this discipline is recognised as a critical technical activity, present in all the phases of project execution, operation and production of deepwater developments. Flow assurance analyses originate during exploration activities and continue through the entire life cycle of deepwater fields exploitation.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



As a result flow assurance must be integrated into the overall subsea systems engineering effort, to integrate flow assurance with system/equipment selection, elaboration of operational guidelines and development of blockage remediation strategies.

1.3.1 Data The first stage involves listing and assembling the data essential for the different studies. Obtaining reliable data is one of the keys to evaluation of the risks associated with produced fluids, reservoir behaviour, production profile, but also with the adjacent functions like water and gas injection.

1.3.2 Risk assessment & Pre-sizing Using the analysis of fluid samples and preliminary data (reservoir, environment), the Flow Assurance study will first: Ø evaluate, from the physical and chemical standpoints, the risks of organic and mineral deposits and plugging, Ø preliminary sizing , in hydrodynamic terms, the lines under steady state conditions, on the strength of preliminary data. Once the major risks have been identified and the lines pre-sized, we then describe the operational constraints. The operating range will be delimited by identifying the nominal and downgraded operating modes (i.e. arrival temperature range).

1.3.3 Management Using the different operating modes (normal and downgraded), Management studies then seek solutions regarding: Ø

Flow Management for steady state and transient flow, dealing with, • unstable conditions (hydrodynamics and severe slugging), • shut-in and restart conditions, • artificial lift based on the well eruptivity limit, • erosion-corrosion, • well metering.

Ø

Thermal Management of Behaviour of the streams. The different possibilities (passive insulation and heating) are assessed technically and economically. The thermal signature for the lines (flowlines & riser) is defined for steady state and transient phase (shutdown and re-start) over the entire operating range.

Ø

Deposit Management. Mineral and organic deposits are quantified for the different operating modes. Solutions for prevention, and possible curative means are described in detail.

Ø

Corrosion management. The different solutions (material selection, prevention, inhibition) are evaluated technically and economically. Monitoring and inspection means are defined.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



1.3.4 Sensitivity Studies Sensitivity studies are conducted to assess the adaptability of the selected system to variations in certain production conditions predicted or not by the reservoir studies. As an example, the repercussions of velocity and cooling changes induced by gas or water breakthrough have to be taken into account.

1.3.5 Project Studies In accordance with operating philosophy and based on design of wells, pipes, subsea and surface equipment, project studies include: •

Verification of the capacity of wells, pipes, subsea and surface equipment to handle prevention and remediation,

•

Definition of monitoring to ensure detection of any changes as early as possible,

•

Basic and detailed engineering of the production system (subsea and surface) in connection with flow assurance, such as selection of chemicals, chemicals injection system, sampling points…

•

Establishment of procedures for normal and downgraded operating modes.

Throughout the project phases, each Project Technical Review (PTR) includes a Flow Assurance evaluation.

1.3.6 Start-up and Production During the production start-up phase, fluid samples are collected and analysed. The results are compared to the previously determined characteristics. The thermal and hydrodynamic signatures of lines are monitored and checked and compared to forecasted performances. Following comparison of actual and forecasted performances, corrective actions are determined if needed, and procedures are updated. During the production phase and throughout field life, continuous monitoring data are collected and analysed. Trends are noted and used to update the Flow Assurance strategy. Disruptive and corrective actions are recorded and analysed.

1.3.7 Flow Assurance file According to the different development phases, in order to consolidate and facilitate feedback all the data, studies and information obtained in the course of the above parts should be collected and classified in the “Flow Assurance File”. The figure here after shows a synthetic diagram of the work process related to Flow Assurance.

TotalFinalElf E&P

DGEP/SCR/RD/UGF


Exploration

Pre-Project

Project

Start-up Production Exploitation


•

Data (Fluids, Reservoir, production profiles…)

•

Risk assessment

•

Solutions (Deposits, Thermal, Hydraulic)

•

Operating Logic

•

Sensitivity studies

•

Consistency Equipment / Risks

•

Monitoring issues

•

Detailed Engineering

•

Procedures for Normal & Degraded modes

•

Sampling

•

Thermal behaviour

•

Hydrodynamic Behaviour

•

Monitoring …

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 12 of 128

Presentation

1.4 Key personnel involved in the Flow assurance Guide The editorial board, chaired by C FOUILLOUT (UGF), comprises the following members, all specialists taking part in the different specialities of Flow Assurance. SPECIALITY

Name

Entity

HYDRATES

JL PEYTAVY

TDO/EXP

ORGANIC DEPOSITS Waxes, Asphaltenes, Naphtenates Asphaltenes only Naphtenates only

JL VOLLE C. SCHRANZ H. ZHOU C. HURTEVENT

TDO/EXP SCR/ED TDO/EXP TDO/EXP

JL VOLLE

TDO/EXP

C. HURTEVENT

TDO/EXP

D. LARREY P. BRANJONNEAU

SCR/ED TDO/EXP

D. LARREY

SCR/ED

P. BRANJONNEAU

Ex. TDO/EXP

EROSION – CORROSION

M. BONIS

TDO/TEC

WELL METERING & MONITORING

L. THOMAS

Ex. SCR/RD/UGF

MISCELLANEOUS PROBLEMS Rheology, Gels, Emulsion SCALE DEPOSITS Carbonates, Sulphates, halite HYDRODYNAMICS and PRODUCTIVITY Pressure losses, Instabilities, Artificial lift THERMAL MANAGEMENT

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 – Physico-chemical

Ref. : UGF/ Rev. : 0 – 01 / 2002 Page 13 of 128

FLOW ASSURANCE GUIDE Deep Offshore Flow Management

Part 1 – Physico-chemical Ø

HYDRATES.

Ø

ORGANIC DEPOSITS vWaxes vAsphaltenes vNaphtenates

Ø

SCALE DEPOSITS vCarbonates vSulphates vSalts

Ø

MISCELLANEOUS PROBLEMS vRheology vEmulsions vSeparation

FA Guide rev 02-10-01

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 Physico-chemical



Hydrates

TotalFinalElf E&P

DGEP/SCR/RD/UGF



2. HYDRATES

2.1 The issues Natural gas hydrates are solid crystals that resemble compact snow or porous ice. They are made of crystalline cages of water molecules entrapping gas molecules. They belong to the clathrates family. Inside production lines, 3 conditions have to be fulfilled for hydrates to form: ♦ The presence of water, "free", dissolved or emulsified ♦ The presence of light hydrocarbons (from methane to butane), or acid gases (CO2 and H2S), or Nitrogen, "free", dissolved or liquefied ♦ Low temperatures or relatively high pressures e.g. 4°C and 12 bar or 20°C and 100 bar. Their formation being exothermic, the dissociation of hydrates, conversely, is endothermic, thus ice formation may occur during hydrate dissociation at low ambient temperature.

Fig 5.1 A General issue: The major difficulty in controlling the risk represented by hydrate formation stems from the fact that only the thermodynamic conditions governing their existence can be satisfactorily predicted. We are extremely short of knowledge on the kinetics of their formation and growth, especially in the presence of oil, leading to a plug forming. ♦

The first practical consequence is that, when a prevention system fails, we can no longer control the formation of hydrates and are thus reduced to counting on huge injections of thermodynamic inhibitor (methanol or glycol) and plenty of good luck!

♦

While the formation of hydrates systematically entails a risk of plugging, this does not necessarily mean that plugging will actually occur. The presence of geometrical or thermal irregularities is in this respect fairly critical.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 16 of 128

Part 1 Physico-chemical ♦

Nor is there any direct relation between the quantity of hydrates formed and the plugging speed. In a flowline several inches wide, the plugging process may be fairly slow (taking several hours and even several days). A substantial amount of material is needed for a plug to form. On the other hand, a relatively small quantity of hydrate may be enough to instantly plug a valve (e.g. during re-opening of an SCSSV, if a gas pocket, formed upstream during shutdown, comes into contact with water segregated downstream).

♦

Finally, as far as prevention methods are concerned, the economic issue hinges above all on the produced water flow rate. Thus, we can clearly see (figure below) that the continuous injection of chemicals beyond a certain rate of water flow will become crippling. This water flow limit has to be defined on a case-by-case basis, but, a priori, we can consider: Ø gas fields (with a low liquid content) where continuous injection of chemicals will be economically viable, and Ø oil fields, where the water flow rate to be inhibited will, sooner or later, become totally unacceptable; especially when water injection is used to ensure reservoir pressure maintenance.

OPEX of hydrate inhibitors (continuous injection)

OPEX (US$ per bbl of produced oil)

5 4.5 4

Methanol -20°C subcooling

3.5

Methanol 3

-10°C subcooling

2.5 2

new GHI (KI or AA) -10°C subcooling

1.5 1 0.5 0 0

1

2

3

4

5

6

7

8

9

10

11

watercut (%)

Fig 2.1 B

12

13

14

15

16

17

18

19

20

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Associated risks: the following merit consideration: ♦

Rapid or even instantaneous plugging of flowlines, leading to significant production losses due to the difficulties of intervention.

♦

Rupture of the line subsequent to a projectile effect when the plug dissociates with under-balanced pressure (velocity of 18 to 80 m/s).

♦

Bursting of the flowline with local overpressures when evacuation of the gas during dissociation (180 v of gas / v hydrates) is not possible.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 18 of 128

Part 1 Physico-chemical In Deep offshore contexts

For oil fields, despite preventive measures based on thermal insulation, the major risk that subsists lies in the re-start phase after a long unscheduled shutdown. At shutdown, the risk of a plug forming is near to zero. On re-start, a massive, even explosive formation of hydrates may occur, leading to extremely rapid plugging of flowlines.

M Remediation means are limited and time consuming in deepwater context. After plugging the line with an hydrate plug that has cooled down to ambient (4°C), dissociation time can extend into several months if the pressure cannot be reduced to very low value (< 4 bar). Graph here after shows the predicted plug dissociation times for an oil system.

Hydrate Plug dissociation time (cf Shell Global Solution) 12'' PIP with hydrate plug cooled down to 4°C prior starting depressurization 600

dissociation time (days)

500 400 300 200 100 0 0

50

100

150

200

250

downstream pressure (psi)

Associated costs from production losses and intervention are high. For gas fields, the difficulty is to implement continuous inhibition economically and reliably. Plugging does not have the same consequences in curative terms due to possibilities of depressurising the lines.

2.2 Data Required To predict hydrate formation and associated risks, we need the following information: Ø

the molar composition of the effluents, at least C1, C2, C3, iC4, nC4, N2, CO2 and H2S,

Ø

the phase range provided by the PVT study,

Ø

the amount and salinity of the actual produced water,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 19 of 128

Part 1 Physico-chemical Ø

the pressure and temperature profiles of the evacuation stream (wells, flowlines and risers) from OLGA type modelling, under steady and transient phases.

Ø

The maximum sub-cooling expected along the lines will be deduced from these data (hydrate curve and operating range).

Ø

For oil fields, the feasibility of hydrate transportation in slurry needs to be assessed on a specific pilot rig.

2.3 Methodology

2.3.1 Formation Prediction (thermodynamic range) The only possible prediction is the pressure and temperature range for hydrate formation. This has to be compared with the P, T transient and steady operating conditions.

Pressure Bar

EXAMPLE OF HYDRATE FORMATION CURVE

200 180 160 140 120 100 80 60 40 20 0

Hydrate formation No Hydrate

0 g/l NaCl 60 g/l NaCl 0

4

8

12

16

20

24

Temperature °C

It is important to remember that the problem has to be dealt with in the same way, whatever the state of the effluent and the phases present, and regardless of whether the phases are "free", dissolved, liquid, or emulsified. There are several software applications capable of calculating the P, T, and hydrate formation region, among them: ♦ Equiphase (integrated in Pro II), ♦ CSMHYD (developed by D. Sloan's team at the Colorado School of Mines), ♦ Aqua*sim (integrated in SHG),

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 20 of 128

Part 1 Physico-chemical

♦ FHYD developed with the help of Professor Peneloux (University of Marseilles) 4 00 3 75

H yd rate fo rm a tio n cu rves S h a llo w ca se N o S a lt

3 50 3 25 3 00 2 75

Pressure (bara)

2 50 2 25 2 00 1 75 1 50 1 25 S h allo w S h allo w S h allo w S h allo w

1 00 75 50

A q ua sim E q uiph as e C S M H YD F hyd

25 0 0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

T e m pe ra tu re (°C )

These software applications are calibrated against experimental measurements and generally deliver similar results. However, we have observed some differences (see curves above), and it is best to systematically compare the calculation results of at least 2 of them. Where a substantial interval persists (> 2 - 3°C), it is advisable to perform experimental PVT cell measurements to define the formation range.

For operating temperatures above Ti, fluids will be outside the hydrate region, whatever the pressure.

80

Pressure (bars)

Remark: The intersection Pi, Ti of the hydrate curve (see curve here aside) and the bubble curve (quadruple point) must be determined carefully when located in the operating range conditions.

70

hydrate curve

60 50

Quadruple point

40 30

hc bubble curve

20 10

Pi

0 0

4

8

12

Temperature (°C)

Ti

16

20

TotalFinalElf E&P

DGEP/SCR/RD/UGF



2.3.2 Evaluating the Risks of Plugging To identify the problem areas, we have to compare the hydrate formation region and the operating pressure-temperature pairs. For gas field application, the risks of plugging by hydrates can be determined reasonably accurately with simulation models. For oil field application, the risk of line plugging is not only related to thermodynamic conditions. The kinetics of hydrate growth and agglomeration also depend on many factors not yet fully understood. Recent works on test loops show that some oil/water systems contain natural surfactant components that play the same role as anti-agglomerant additives. Thanks to this property, hydrate crystals can be transported in the form of slurry, and installations can be re-started without the lines becoming plugged. To date, two types of oil have been identified as having that property, at least up to a certain level of water content: Ø

Asphaltenic crude oils. The presence of resins and asphaltenes gives them the interfacial properties to stabilise an emulsion of water in oil. The slurry is thus stabilised when the droplets of water turn into hydrates. Examples of asphaltenic crudes include FROY, TROLL and ALBACORA, which can transport a hydrate slurry with up to 30% water.

Ø

Acidic crude oils. The combined presence of naphtenic acids in the oil and sodium and/or calcium salts in the water, under certain pH conditions, generates naphtenates at the oil/water interface. These components are well known for their surfactant properties and capacity to stabilise the slurry. Among acidic crudes, we could mention that during loop test with HEIDRUN oil we didn’t plug the test line up to 30% water.

To conclude, for oil fields, in addition to the thermodynamic study, an experimental evaluation of the transportation potential needs to be performed, either on the CSTJF pilot rig or on the IFP loop in Solaize (see photo here after).


TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 Physico-chemical IFP loop in Solaize

Pilot laboratory unit


TotalFinalElf E&P

DGEP/SCR/RD/UGF



2.4 Solutions

2.4.1 Prevention Various means of prevention can be envisaged. They are classified into four categories: Ø

Dehydration of effluents (gas application), usually by a glycol process. By eliminating water, we remove any risk of hydrate formation. This method is reliable and has been proved, and is used mainly to protect gas export or injection pipelines.

Ø

Keeping the temperature of produced fluids above the hydrate formation threshold. The effluent flows from the wellhead at a temperature higher than that conducive to hydrate formation. To maintain this condition in subsea pipelines, with an ambient temperature at 4°C, we need to limit heat exchanges. That means insulation and sometimes heating of the lines (for more details see § 7 Thermal Management). During a long shutdown, the temperature is maintained above the hydrate formation threshold for a limited period (passive insulation) and even permanently (heating device). But insulation performances have to be carefully checked before implementation, a number of previous experiences having shown them to be poorer than expected due to thermal losses, particularly on singular points with a different thermal signature (manifold, valves, connectors, etc.). Passive thermal insulation alone is not sufficient for prolonged shutdowns. To avoid plugs, which can take a long time and be costly or even impossible to dissociate (several days to several months), we have to consider additional means: 1.

Replacing the effluents by dead oil. Start-up then proceeds by circulating (loop arrangement) the stock-tank oil to warm up the lines and, further on, by opening the wells while massive shots of methanol are injected. Any parts of the system not located on the production circulating loop, must be flushed with methanol.

2.

Remaining outside the hydrate formation region during shutdown by depressurising the lines assisted by riser gas-lift, insofar as this is possible. The restart phase is achieved by massively injecting methanol at the wellhead while controlling the pressure in the lines to avoid any incursion into the hydrate region.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 24 of 128

Part 1 Physico-chemical 3.

Maintaining the fluid temperature outside the hydrate region by permanent heating activated during shutdown (case A). T e m p e r a tu r e

F lo w in g T e m p .

P R O D U C T IO N

S t a r t e le c tr ic H e a tin g

C r itic a l te m p .

2 0 °C / 7 6 °F

H y d r a te R e g io n S e a te m p .

4 °C / 4 0 °F

T IM E

D e fa u lt

SHUTDO W N Case A Maintaining the temperature outside the hydrate region by electrical heating of lines 4.

After a total black-out, if all power generation is lost, the fluid will cool down to ambient temperature, and then when power generation will be available re-heating the effluents to quit the hydrate region before any start-up of producers(case B). Temperature

Fluids Stabilization

Flowing Temp.

Re-Start Production Well opening

PRODUCTION

Stop Heating Critical temp.

20°C Start Heating

Hydrate Region

∆Τ°

4°C Required duration

TIME Default

SHUTDOWN Case B Heating fluids before start-up of wells These last two active solutions can only be considered for lines equipped with an heating system (electrical, hot fluid, etc.).

TotalFinalElf E&P

DGEP/SCR/RD/UGF


Ø


Continuous injection of chemicals:

Continuous injection of chemicals is nowadays only economically viable for fields with low water flow rates (a few hundreds of barrels/day) i.e. gas fields. When the water rate to be inhibited represents several thousands of barrels/day and the sub-cooling to be overcome is around 15-20°C, the continuous use of any sort of chemical inhibitor becomes unrealistic. There are 3 families of chemical inhibitors, with different modes of action at different stages of industrial development: v 1- Thermodynamic inhibitors: like (non-recycled) methanol and (in principle, regenerated) glycols. These are antifreeze type products. They act by displacing the hydrate formation region. This is therefore by far the most reliable method and the one most often used. The quantity of methanol and glycol inhibitor that must be present in the water can be predicted by the models mentioned above and will depend directly on the desired drop in temperature with respect to hydrate formation. Remarks: - the total quantity of methanol to be injected must take into account losses in the gas and oil phases, - the presence of salts in the water to be inhibited, although favourable from a thermodynamic point of view, can lead to 2 types of limitations: ♦ when methanol is used, the solubility of the salts will be considerably reduced. For example, the solubility of NaCl dwindles from 270 to 150 g/l when the methanol content increases from 0 to 30% weight. The risk of precipitation of salts can thus increase significantly (salt bridges), thereby impairing inhibition of hydrate formation, ♦ the fouling, or even corrosion, of boilers with the presence of salts often makes the re-generation of glycol almost impossible.

v 2- Low Dosage Hydrate Inhibitors (LDHI): These new additives are not currently used yet in deepwater conditions (subcooling > 15°C or BSW > 50%). Although the active material is relatively expensive as compared with methanol, their low dosage may result in significant cost cuts (Fig 2.1 B). Remark: before deciding to use these additives, experimental work is always needed to optimise the active material formulation and quantify the injection rate. The LDHI additives are of 2 types: Þ kinetic Inhibitors (KHI), hydrosoluble polymers, injected at low dosage (generally 0.5% weight of active material versus water content). They retard the crystal growth and prevent hydrate formation for the time needed to transport effluents under hydrate region conditions. Their performance is limited, at least for the time being, to relatively mild cases of sub-cooling (< 10°C). An inhibitor of this type, developed by BP/TROS, was tested industrially in 1997 on a 12″ pipeline between Pont d’As and Lacq, in mild conditions (sub-cooling ≤ 6°C). The results were positive as far as the additive's capability to inhibit hydrate formation was concerned. However, problems of emulsion and foaming were


TotalFinalElf E&P

DGEP/SCR/RD/UGF



observed on the downstream facilities. Before any use, we need to carefully assess this issue. Despite many other tests on fields, these inhibitors are still seldom used in continuous injection. BP has been using them since 1997 on a North Sea line covering 70 km from the HYDE field to the Easington terminal to replace the methanol planned in the initial design. On the MUNGO field (ETAP field run by BP), a kinetic inhibitor is injected to prevent hydrate formation when there are long scheduled shutdowns. To date, performance is not fully satisfactory since the subcooling is around 10°C. An instance of plugging was reported in 2001. Þ Anti-Agglomerants (AA) or dispersant additives, hydrocarbon soluble polymers, also injected in low dosage (generally 1% weight of active material versus water content). These do not prevent the formation of hydrate crystals, but limit their size and keep them in suspension in the liquid phase as slurry. These additives are still at the development stage. Their scope is limited, at least for the time being, to moderate BSW (< 30 to 40%). The presence of a liquid hydrocarbon phase is necessary i.e. these chemicals do not control hydrtates in a gas-water system. They may become economically applicable for many cases once the subsea water separation technique is available. An additive of this type (Emulfip 102 B), developed by TFE and IFP, was recently tested on the CANADON Alfa field in Argentina. The relative failure of the test can be attributed to extremely severe conditions (laminar flow, temperature sometimes lower than 0°C and sub-cooling of around 20°C). Other tests need to be conducted before concluding as to the feasibility of this prevention process. An AA additive ( RE 4394 HIW from Baker Hughes) is presently under test on Shell Popeye field, in a subsea flow line which is operated at 12 °C subcooling. Shell is preparing an other test with an AA chemical Armoclear 2550 on a satellite of SEAN ( North Sea) This will be the first industrial application of AA.

v

3- The natural oil surfactant properties: This way of controlling hydrates obviously needs to be validated in the field. This is the goal of current JIPs. But a priori, we can say that it has one sure advantage in that it is free! Its application, as for AA additives, is limited by the BSW, but here too, the development of subsea water separation could render this method viable in a very large number of cases without incurring extra costs.

2.4.2 Hydrate solution and limitations for Dry Trees associated with wellhead platform The main concern for dry tree risers in deepwater is the hydrate formation leading to plug the production tubing during shutdown or associated restart phase. The classic design for dry tree riser (isolation by nitrogen in the inner annulus) can provide a very short cooldown (delay) before the fluid enters in hydrate domain. This cooldown time depends upon the pressure of the nitrogen in the annulus (2 hours and lower).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



In mid water depth, operations of WHP do not lead to hydrate plugging controversially in deepwater, low seabed temperature and high shut in pressures can result to get severe incursion into the hydrate region (on the order of 20° C in some cases). As the SCSSV is located as deep as to operate out of hydrate domain, depending on water cut and GOR, depressurisation of riser after closing SCSSV may not reduce significantly the pressure at the mud line out of hydrate formation. On the other hand, remediation of hydrate plug can be handled « easily » with coiled tubing technique. But in the case of multiple wells (more than 10), using coiled tubing to restart after shutdown is not realistic due to economic impact. Riser insulation and or active heating techniques have to be improved to give sufficient cooldown for preventive actions. Production restart with dry trees at high water cuts is also a main issue. During the restart period MeOH injection will prevent hydrate formation. This thermodynamic inhibition injection is maintained until a sufficient temperature is reached to give required cool down in the case of new shutdown. The design of completion limits the inhibitor injection line diameter an subsequently methanol or LDHI injection rate to less than 10 gal/mn. (Matterhorn < 5 gal/mn w/ 9/16’’ injection line). At high water cuts, this limitation in injection will impact the production rate at which the well can be restarted. Lower flowrates during restart phase result in longer warm-up times and consequently in large volumes of methanol.

2.4.3 Monitoring As mentioned above, a distinction needs to be drawn between hydrate formation and plugging by hydrates. But detection of the presence of hydrate crystals in effluents should provide warning. At present, there is no one universal way of detecting the formation of a hydrate plug. The following four observations should, however, act as a warning and so trigger or increase the injection of inhibitor as a preventive measure: -

Pressure measurement (pressure drops) and temperature measurement to evaluate the operating range as opposed to the hydrate formation region. Temperature measurement throughout the liquid stream by a non-intrusive optical fibre (DTS Distributed Temperature System) provides monitoring of fluid temperature all along the lines (see for more details in § 7.4.3).

-

the presence of hydrate crystals in the pig receiver,

-

reduction in the water flow at the separator on the receiving installations (gas applications),

-

gradual increase in pressure drops, especially if punctuated by sudden drops (sign that a plug is being formed and partly tearing free from time to time).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



2.4.4 Curative Methods Theoretically, there are four methods that can be used either together or separately: −

1- Injection of thermodynamic inhibitors. This method is generally a long and painstaking matter but is reliable provided the inhibitor actually manages to make contact with the plug (with, as a corollary, the problem of multiple plugs!). It therefore can only be used if the plug is perfectly located and the inhibitor can migrate to reach it. Remark: due to its density of 800 kg/m3, methanol may not traverse an oil layer by the effect of gravity to reach the sedimented water part.

−

2- Depressurisation. This should be performed, whenever possible, on both sides of the plug. It is the most commonly used method, as it is simple and effective and you do not need to know the precise plug location. However, if you only depressurise on one side, there is a risk of making a plug behave like a cannonball, likely to damage the pipeline system. A model on radial dissociation has just been developed by the Colorado School of Mines (D. Sloan) within the framework of the DEEPSTAR JIP. This is the first model of its kind and is claimed to be capable of predicting the time needed for radial dissociation of the hydrate plug by depressurisation. One limitation to this method emerges in Deep Offshore oil production. The residual static pressure after depressurisation can be higher than the minimum pressure required to quit the Hydrates region. Furthermore, in a thermally insulated pipe excessively long dissociation times are to be expected because of the endothermic nature of hydrate dissociation.

−

3- Heating: This method is not recommended when it is impossible to evaluate the length of the plug (which is often the case in subsea pipes), heating must be banned as a precaution. Indeed, if evacuation of the gas released cannot be ensured due to the plug sealing effect, the pressure may increase exponentially with temperature (heating mode) and ultimately cause rupture of the pipeline. Conversely, if the pipeline is within the hydrate region during prolonged shutdown but no plug has formed, the heating approach can be recommended to get effluents out of the hydrate region BEFORE re-starting production (see § 7.4.2 active heated systems). However, recent study performed by SHELL provided a detailed analysis on hydrate dissociation by electrical heating demonstrating that it is technically feasible, operationally safe and efficient means for plug remediation. During dissociation pressure increases is relatively modest compared to the pipe burst pressure and a very small gap (pipe wall and hydrate plug) is required to discharge a very small mass of gas at large pressure difference. Hydrate dissociation process which is endothermic, will quickly consume heat transferred from the pipe wall. Pipe wall temperatures do not approach the value required to support pressures near the pipe burst limit. Based on the conclusion of this study Electrical Heating will be used (installation in 2003) for Nakika field (WD: 1900m) for hydrate remediation on the different flowlines (10’’ x 16’’ PIP length 2 to 12.8 Km).

−

4- Mechanical, by drilling or jetting effect. This method can be associated with the injection of thermodynamic inhibitors to combine both (mechanical and chemical) effects.

•

Remediating for oil dominated system by the use of intrusive system as an Extended Reach Coiled Tubing Technique going gradually along the riser and the flowline up into contact with the hydrate plug(s) to dissociate them by circulating tank battery oil and/or methanol. This method is now limited by the size of the coiled tubing available on the market, its weight, tensile limits and mechanical stability, and its ability to slide into bends. Usual diameters range from ½’’ OD up to 3 ½”OD.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Classic drums available have lengths of between 450 and 600m of steel tubing, the limit being due to the maximum diameter and weight. One limit to the range of action for CT stems from the buckling effect, which can cause spiral winding on the internal wall of the pipe after too high a thrust. The buckling effect will depend on the ratio between the internal diameter of the pipe and the outer diameter of the Coiled Tubing. To use this technique, you need to take into account geometrical specifications in terms of bending radius and change in direction during the design phase. Recommended curve radii range from around 6 to 18 m. Recent works conducted over 1997-1998 within the framework of the DEEPSTAR JIP highlighted the advantage of combining a crawler and hydraulics (Jet and piston effect developed by RADOIL providing 15T traction for 70 bars ∆P at the CT head to pull the coiled tubing rather than push it. A skate system developed by AMBAR ensures partial contact between the CT and the pipe wall to limit friction. After tests combining the crawler and skates, the maximum range for this technique has been extended to about 8 – 10 kilometres. Other crawlers are also available from WELLTEC, ASTEC and DACON. Another means of improvement is to use composite coiled tubing. Here, the main advantages are reduced weight (1/4 to 1/3 the weight of steel CT) and mechanical resistance to fatigue. The FIBER SPAR Company offers this type of equipment (OD from ½ to 6’’). Typical functional requirements for hydrate plug removal systems (HPRS): 1. Deploy in the production riser a means of circulating a chemical inhibitor (methanol) associated with a jetting effect. The hydrate plug has to be destroyed and dissociated by overbalance in order to avoid cannonball effect. 2. Deploy in the production riser a depressurising means (gas or nitrogen lift) to re-start production without any risk –pressure & temperature) of a new hydrate plug. 3. Run-in and out without causing any damage to production equipment. 4. All parts of the HPRS have to be “fishable” in order to retrieve any loosen part. A generic study on “Hydrate Removal in flowline & riser” has been generated by UGF R&D project and is performed by Halliburton Deepwater, conclusions should be available beginning 2002. The aim of this study is first to establish the feasibility of hydrate removal in deepwater oil production lines (Risers & Flowlines) and then define the general specification to incorporate in the development design phase to facilitate hydrate removal interventions. Also the study will identify today limitation (technology gaps) and will propose development plan to achieve the goal. This study will be completed by beginning 2002. Hydrate plug removal guidelines: This set of guidelines is an extension of recommendations listed in the Offshore Hydrate Engineering Handbook (Sloan 1998). 1. Hydrate plugs are always dissociated, but the time scale is usually days to weeks. Deliberate changes and patience are required. Hourly changes are ineffectual.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



2. Multiple hydrate plugs should always be assumed and treated as a safety hazard. 3. Many hydrate plugs are porous and transmit pressure easily while acting to obstruct flow. Some plugs are permeable to gas, but less so to condensate or black oil. This concept controls many aspects of hydrate dissociation, including radial depressurisation, Joule-Thomson cooling through the plug, and the fact that depressurisation may cause the temperature downstream of the plug to decrease below the hydrate equilibrium temperature. 4. Methods are not well defined for locating hydrate plugs and determining their length. However, knowledge of the precise location and length of a plug would be a vital help in dissociation. The current model does not require a plug length, but this information is necessary to determine how the pipeline can be depressurised safely. 5. Attempts to "blow the plug out of the line" via high pressure upstream always result in a larger, more compacted plug. 6. Depressurisation from both sides of hydrate plugs is the preferred method of removal, from both safety and technical viewpoints. This implies access points at both plug ends through dual production lines, service lines, etc. 7. If the pressure is decreased too much, the hydrate plug will rapidly form an ice plug which may actually be easier to remove than the original hydrate plug. 8. In a deepwater line, a liquid head on a hydrate plug may be sufficient to prevent depressurisation. Liquid head removal is a current challenge to flow assurance. 9. In some cases, depressurisation from one side of the plug has been safely achieved. 10.Heating is not recommended for hydrate plugs without a means for relieving the excess gas pressure when hydrates dissociate. 11.Coiled tubing represents the primary mechanical means for dissociating hydrates. 12.Usually methanol or glycol is injected into plugged flow lines, but this is seldom effective due to the difficulty of bringing the inhibitor into contact with the face of the plug. This method is most effective when used in conjunction with depressurisation and the hydrate plug has had a chance to recede sufficiently from the pipe wall to allow the flow of inhibitor around the remaining plug. 13.Inhibitor injection and de-pressurising techniques are available for system shut-in and start-up – two potentially hazardous periods as regards the formation of hydrate plugs. 14.Insulated pipelines can represent a serious hindrance to hydrate dissociation. Modelling indicates that the resistance to heat transfer through the pipe wall increases the hydrate dissociation time by several orders of magnitude. Care should be taken when designing an insulated pipeline. e.g. A typical overall heat transfer resistance is about 1 BTU/hr/ft/°F (Sloan 1998) This heat transfer resistance corresponds to a Biot number of about 0.02 The time it takes for the hydrate plug to completely dissociate is about 100 times greater. At an ambient temperature of 4°C, the dimensionless dissociation time changes from 75, with no resistance to heat transfer, to 660 at a Biot number of 0.02.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Organic Deposits


TotalFinalElf E&P

DGEP/SCR/RD/UGF



3. ORGANIC DEPOSITS

3.1 WAXES

3.1.1 The issues Wax deposition can be a potential issue even on crude oils or condensates with low wax content, due to low ambient temperatures. Waxes are chemically well-identified components, alkanes, the simplest representative of which is methane. These are hydrocarbon chains that are mostly linear (n-paraffins) with a few slightly ramified components (iso-paraffins). Waxes with high molar masses are in solution in the crude oil but crystallise when the temperature drops. This can have two consequences: ♦

A deposit of waxes on the wall. As soon as the deposit forms on the wall, a process will begin whereby its thickness will increase over time, with an impact on the pressure profiles (increase in pressure drops inherent in the reduction in cross-section) and temperature (improvement of thermal insulation due to the deposit). The restriction in the passage may be highly substantial, to the extent of all but plugging the pipeline.

Example of wax deposit from Main Pass Field GOM - Elf Explo Inc. -

♦

However, operational problems are likely to be encountered much earlier than any significant restriction or sticking. The waxy layer indeed can be rough, increasing the surface roughness of the pipe wall. The resulting increase in friction pressure can effectively choke the system, resulting in higher pumping costs or reduced throughput.

♦

Gravity settling in lines can occur during shutdown, but in this case the deposit remains very soft and is easily removed under flowing conditions when the line is restarted.

♦

Congealing occurs in the form of gel when production is stopped due to an increase in viscosity. This phenomenon is quite different to the previous one but can also lead to obstruction of the liquid flow in the event of prolonged shutdown. This aspect is explained in greater detail in section § 2.4 of this guide.

The risks associated with wax deposits on the wall are as follows: ♦ Plugging of flowlines, especially during pigging operations. ♦ Plugging of tapping points (pressure measurement).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



The position and the speed of the wax deposits in the lines depend on the properties of the fluid, the thermal conditions (wall temperature & gradient) and the flow rate. For Single-phase liquid systems, the mechanisms involved are almost fully explained . They are less completely understood for multiphase systems. In all cases, the ageing effect (change in consistency) and influence of internal pipe surface are never taken into account. Deposition is a lengthy phenomenon. The speed of growth is generally slow and must be considered as an equilibrium related to the steady flow. Deposition under short transient situations is never an issue as the return to steady state conditions will eliminate its consequences. The proportion of waxes deposited and adhering to the walls is small in comparison with the waxes that crystallise in the crude oil. For there to be a deposit, the wall temperature needs to be lower than both the Wax Appearance Temperature (WAT) and the flowing temperature. This is a requisite condition but is not sufficient insofar as some lines operated in this situation show no deposits. The phenomenon has not yet been explained. Three mechanisms can contribute to the transfer of waxes to the pipe wall: ♦ Molecular diffusion ♦ Turbulent diffusion of crystals (shear diffusion) ♦ Sedimentation. On the other hand, turbulence in the flow regime produces shearing stresses likely to tear away the wax deposits. As a first approximation, the rate of deposition in a pipe will depend mainly on molecular diffusion and thus be closely related to the thermal gradient on the wall. In addition: Ø

a low temperature gradient will lead to a hard deposit forming over a long period while,

Ø

a high gradient will lead to a rapidly accumulated soft deposit.

3.1.2 Data Required Sampling A representative sample of the fluids in reservoir conditions. It is important to make sure that the well is cleaned before sampling to avoid polluting samples with drilling mud. For surface sampling (test separator), a drop in the temperature of the fluid can result in crystallisation and the formation of wax deposits in the separator. The fluid sampled at the outlet will be depleted of part of the waxes that have precipitated, and give rise to an erroneous WAT value and crystallisation curve. Sampling is the key to good prediction. On the first Girassol well, samples from an MDT (modular drilling test) gave a WAT value of 25°C. Subsequently, the WAT values determined from surface samples taken during clean-up phase fluctuated over time from 29°C to 38°C.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



We concluded that both surface and downhole samples had been polluted by the drilling mud (oil based).

Paraffin distribution 0.3

0.25

% Weight

0.2

0.15

0.1

0.05

0 Carbon number

MDT sampling: MDT samples can be polluted by oil-based mud. The extent of the pollution of the oil sample can be assessed from a sample of the mud itself. As an example, the graph above shows the paraffin distribution obtained on a Dalia 5 MDT sample. We know that the Dalia crude contains no paraffin. The left part of the graph represents the paraffin distribution of the base oil used for the drilling. The right part represents the distribution of heavy paraffin, probably a deposit from previous drilling on a well where the crude contained a significant amount of paraffin. The WAT measured on the sample was 26°C instead of the usual 7-10°C value for the Dalia crude. Surface samples are acceptable, providing the clean-up phase is sufficient to eliminate the mud pollution.

Analysis In addition to "conventional" PVT analyses, the evaluation of the wax potential of a crude oil requires additional information:

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 35 of 128


• detailed compositional analysis by Gas Chromatography (GC) up to C10+ or C20+ (about 300 crude oil components are identified), • distribution of n-alkanes (n-paraffins) from C15 to about C50 by HTGC (High Temperature Gas Chromatography). The general appearance of the HTGC curve shown below allows a first evaluation of the risks related to crude oil. A crude oil with a lot of "light" paraffins (< C30) is likely to give rise to rheological problems that can go as far as gelling in the shut-down phase. A crude oil with few light paraffins but a significant number (more than 0.5% weight) of "heavy" paraffins (> C30) will be more likely to form deposits. n-Paraffindistribution Nparaffin wt % 0,7

The distribution of) super heavy nparaffins (up to C80) can, in most cases, be calculated by applying an exponential correlation in order to reproduce the general distribution trend given by measurements (see calculated curve on the right). Sometimes the n-paraffin distributions do not exhibit an exponential decay with increasing carbon number ( i.e. AKPO crude). In that case extrapolation is not possible.

0,6 % w e i g h t

0,5 0,4 0,3

fig 1

%weight calculated

0,2 0,1 0 13 15 17 19 21 23 25 27 29 31 33 35 37 carbonnumber

• The measurement of the WAT, is currently under procedure of normalisation. • WAT and the crystallised fraction are measured by Differential Scanning Calorimetry (DSC). These values are obtained in the laboratory on stock tank hydrocarbons. Results can be extrapolated to actual pressurised fluids with appropriate models (e.g. TU-WAX). The WAT measurements are influenced by the kinetic. Therefore, it is strongly recommended to determine the WAT using the lowest possible cooling rate. Other technique are available( Cross Polar Microscopy, Filtration…). WAT is determined with an accuracy of +/- 2°C. • Viscosity versus temperature and shear rate curves • Temperature profile of the production stream.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 36 of 128


WAX FORMATION EVOLUTION Pressure (Bar)

An example of WAT evolution vs. pressure, given by model, is shown (right). Note: WAT value is lowest at the bubble point.

180 160 as 140 120 100 80 60 40 20 0 32

Wax No Wax

34 Bubble

36

38

WAT °C 42 (°C)

40

SOLID WAX CONTENT AS A FUNCTION OF TEMPERATURE Solid Wax Fraction %wt

18 16

One useful measurement is the amount of wax precipitated as a function of temperature (see right).

14

Waxy Light

12 10

This type of information is used in wax deposition models to predict rates.

8

The measurement is performed by DSC, applying a specific, internally developed correlation.

4

6

2 0 -25 -2

-5

15

35

55

75

95

115

135

Temperature °F

3.1.3 Methodology Based on the previous data recovered, evaluation of the wax deposition risk generally involves two steps: 1. Use of a thermodynamic model to determine the WAT and the crystallised fraction in relation to temperature. The model is generally based on a measurement at atmospheric pressure and allows the prediction to be extended to all pressures and compositions. However, the validation of input and output data is still a developing science. It is therefore recommended to seek expert advice. 2. Use of a kinetic model to evaluate the thickness and location of the deposit in the pipelines (cold wall deposits) versus time. It has to be stressed that this type of model is more providing a trend of profile and magnitude of the problem.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 37 of 128

Part 1 Physico-chemical Models

Most market models are based on the assumption that the deposit builds up by molecular diffusion. The software used in studies is: • An integrated (thermodynamic + thermal + multiphase flow + deposition kinetics) model, TUWAX, developed within the framework of a JIP by the University of TULSA. This is a two-phase oil and gas model and does not take the water phase (assimilated to oil) into account. From a given composition, it will determine: 1 the phase envelope, to be tuned on PVT data, 2 the change in viscosity vs. temperature, to be tuned with laboratory measurements, 3 WAT, to be tuned on the value given by DSC, 4 After validation / calibration of the 3 measurements, the thickness of the deposit versus time along the production line.

Kinetic models only concern cold wall deposits. Sedimentation occurring in stock tanks is considered to be systematic when the crude oil is stored at a temperature lower than its WAT. This assumption is likely to be extremely conservative. Theory on wax deposition The equation representing the Molecular diffusion is shown below:

λ dF dT dQ = dt υ dT dy c

where:

– dQ/dt – λ – µ – Fc – T – dy

represents the deposit kinetics diffusion coefficient crude oil viscosity the crystallised fraction temperature boundary layer limit

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 38 of 128


Laminar Boundary Layer

Turbulent core

Dissolved wax

Validations performed show that for a single-phase fluid, molecular diffusion is actually the driving force behind deposits. The mechanism for deposition on the wall can be compared with the formation of condensation on a glass panel, which only settles when there is a thermal gradient.

Heat Loss

P I P E W A L L

Velocity profile

Rippling only plays a preponderant role when shearing is significant, that is generally the case when the pipeline is close to plugging. TU-WAX makes allowance for this effect with a “shear factor”. But it should be noted that the shear influence has never been validated by experimental data. The practical consequences of these assumptions are as follows: Ø

Wax deposition only occurs when wall temperature is below WAT,

Ø

Deposition will not occur if the external temperature is greater than or equal to the crude oil temperature,

Ø

If the wall temperature is close to that of the crude oil, the kinetics for deposits will be relatively insignificant (case of an insulated pipe),

Ø

Plugging is theoretically impossible with a pipe in continuous operation (because of the increase of the shear effect). Any plugging usually arises from transient operation (pigging, significant change in flow conditions, change in temperature, etc.).

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 39 of 128


12

Thickness (mm)

These curves show that wax deposits mainly accumulate along the first kilometres owing to the fact that the temperature of the wall is less than the WAT.

6 hours 1 day 2 days 3 days 4 days 5 days 6 days 7 days 6h

Inlet T°C=54°C

10

An example of a simulation result is given (see on right):

8 Inlet T°C=25°C 6 4 2 0 0

5000 10000 15000 20000 25000 30000 35000 Distance fromwells (m)

Distribution of Carbon Number in Wax deposit at different pipewall temperatures

10°C

1,5

At a given point, provided the fluid temperature remains below the WAT, the molecular distribution of the waxes deposited will depend on the temperature of the pipe-wall (see diagram on right).

20°C

30°C

1

0,5

% weight

0 30

35

40

Carbon Number

The higher the temperature (but < WAT), the greater the molecular weight of the deposit will be. Since Paraffin deposits as a solid solution, the composition of the deposit at high temperature will be different from that for deposits generated at temperatures close to ambient temperature.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Shearing effects due to flow turbulence will also be preponderant in the formation of deposits. The mechanical forces generated by flow act on the pipe wall and affect the quantity of waxes that adhere and remain in the deposit layer. Under a turbulent regime, the most "tenacious" waxes will adhere to form a hard deposit. However it is highly important to note, for practical operation, that commercial software does not provide any information on the quality of wax deposits. Finally, the flowrate influences the thermal gradient and the shear effect. The general trend can be represented as below: Deposition rate

Effect of flowrate on Deposition rate

Turbulent Laminar

Flow/ Shear Rate

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3.1.4 Solutions Wax is generally assumed to be easily prevented or removed by a combination of thermal and mechanical solutions. As pointed out above, the driving factor in wax deposition is pipe wall temperature, and its kinetics are mainly dependent on existing gradient and shear effect. Prevention 2 types of action can be implemented to prevent wax deposits on cold walls: ♦ chemical ♦ thermal. Chemical solutions Chemical additives act by limiting the agglomeration and growth of crystals (dispersing effect). They are never completely effective and are mostly used to assist mechanical removal (pigging or scraping). Their cost is not to be underestimated, due to the high concentrations that have to be injected. Another chemical means worthy of mention is film-forming additives (Anticor, CECA), which work on principles similar to those of a corrosion inhibitor, forming a film on the pipe wall deterring the wax crystals from adhering to it. This type of product is currently being evaluated. Suppliers claim near to absolute efficiency but this remains to be proved. Today, due to the limited efficiency of present additives, chemical treatment cannot be envisaged as sole prevention method. Chemical means are mainly injected in order to decrease the curative treatment frequency. Thermal solutions Simple (passive) or more complex thermal insulation is used to maintain temperature above the WAT. There are a variety of thermal insulation solutions, like bundles, pipe in pipe, pipe coated with insulating foam, etc. Monitoring In-situ monitoring of wax deposits has so far not been industrially proved. However, different parameters can be monitored and analysed to deduce how the deposit will evolve: ♦ pressure losses in the different sections, from monitoring of discrete pressure measurements (well head, manifold, riser and top side arrival). ). It has to be highlighted that pressure losses could be mainly affected by the increase in roughness which will dominate the frictional pressures long before any consequent reduction in pipe diameter. ♦ temperature along the line, which can be measured using an optical fibre in contact with the pipe wall ♦ spot measurements of thermal flux through the pipeline (under development).

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 42 of 128


Curative treatment 1- Thermal To melt deposits, rapid cleaning of the lines with heat is not realistic because it requires much higher temperatures (generally around 70°C) than the WAT. However, heating the wall to a temperature slightly higher than the WAT can change the structure of the edges of the deposit, rendering it more prone to tear-off by turbulence. This solution is being evaluated within the framework of a JIP aimed at qualifying a PIP insulation system, equipped with an electrical heating device. 2- Mechanical Pigging has two functions: Ø

mechanical destruction and elimination of wax and any other deposits that have settled in the low points (i.e. corrosion products)

Ø

separation of fluids for hydrate preservation procedures applied with respect to hydrates by circulation of stock-tank oil.

Pigging frequency can be determined from simulation results and data acquired during pigging operations (delta P, quantity recovered at the pig receiver, number and length of production shutdowns, etc.).

Optimisation of operating costs

Costs

Cost of Intervention incl. lost production

Plugging Risks

Cost of lost production due to reduced productivity Optimal 0

3

6

9

12

Intervention No.

One absolute requisite for Deep Offshore pigging is control over the speed and position of the pipeline scraper in the lines. Cleaning may be improved by the jet effect (Spider nose type bypass system), by control over the pig speed (Vmin. 1.3 to 1.5 m/s) and by the choice of a suitable design (two-way, detectable and destructible).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Two configurations can be considered in this context: ♦

Loop arrangement, with both departure and arrival trap located on topsides and two production lines connected together to form a loop. During the pigging phase, one of the two branches produces in reverse direction and some wells are either shut down or have their flow rate reduced. A fluid is pumped from the surface to provide energy.

♦

Arrangement in a single line equipped with a subsea pig launcher (SPL). To reduce operating costs for installation and removal of the SPL, several pigs are stored ready for use in a cassette (5 to 10) on the pig trap. SPL have been selected by BP on ETAP (GD engineering concept) . Same approach has been adopted by Shell on two GOM satellite fields: Serrano and Oregano (Oceaneering concept) tied-back on AUGER.

3- Chemical The use of solvents is not very effective unless considerable quantities are injected, as heavy waxes do not readily dissolve in most conventional solvents. Aromatic solvents (toluene and xylene) are the “least bad”. They can be doped with dispersants to improve the dissolution kinetics. Commingled chemical and thermal cleaning (Petrobras SGN process) is to be kept for desperate cases. The major difficulty is injecting the chemicals close to the plug location. A chemical/mechanical combination is sometimes needed to limit risks of plugging. The injection of a batch of solvent upstream of a pig “softens” the deposit. Conclusions The most frequently adopted solutions are: ♦

Non-piggable zones (wells, jumpers, manifold) are kept outside the deposition temperature range with appropriate insulation for normal and specified low flow conditions.

♦

Flow lines and risers are likely to be in the same situation. Nevertheless, the production loop arrangement or single line associated with sub-sea pig launcher system can tackle any issues, especially when deposition occurs in certain downgraded conditions such as very low flow-rate or insulation deficiency.

♦

Continuous chemical injection should be considered only as an ancillary means needed to lower pigging frequency.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3.2 ASPHALTENES

3.2.1 The issues Unlike to waxes, asphaltenes are not chemically identified components, and cannot be dissolved by adding light solvent or by heating fluids. They are defined as the solid fraction of hydrocarbon that precipitates when a light alkane (pentane or heptane) is added to the crude oil. They are not in a true solution in the crude oil but in the form of colloidal particles. Their stability is characterised, firstly, by the type of asphaltenes present and their propensity to be “soluble” and, secondly, by the crude oil's solvent quality. Flocculation is like a coalescence of colloidal particles to form a compact phase that sediments or adheres to the walls (rock, tubing, pipe, etc.). Flocculation and precipitation can occur under different conditions: • naturally, with the pressure losses associated with fluid flow in the tubing or pipes. It always occurs above the bubble point. With depletion of the reservoir, flocculation can reach the near wellbore area and cause total plugging (HASSI MESSAOUD). Flocculation can occur with light crude oils containing very few asphaltenes, given the poor solvent quality of these fluids. Conversely, very heavy asphaltenic crude oils (like those from VIC BILH, ROSPO MARE or SINCOR) can be perfectly stable due to their solvent capability.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



• when mixing two fluids, usually an asphaltenic crude oil with a paraffinic hydrocarbon, or sometimes with the injection of gas. • During CO2 flooding. • Some cases of asphaltene floculation in the near wellbore area are also reported during stimulation jobs (well acidification). Asphaltenes floculation can deposit in reservoirs, on wall of tubing or flowline, and decant in separators. Deposits can interrupt partially or totally the Production due to the formation of plugs. Flocculated asphaltenes take a very long time to re-dissolve by re-pressurisation. The simple re-dispersion of asphaltenes flocculated and decanted in a tank is almost impossible.

3.2.2 Data Required A representative sample of the reservoir fluid in down-hole conditions is the key point. The sampling pressure must be maintained in the sampling bottle with a Nitrogen chamber in order to compensate variation due to the temperature drop. These specific bottles are currently available in service companies. The pressure is further on kept stable during all necessary transfer up to the laboratory level As already mentioned for wax sampling, the well has to be carefully cleaned up before sampling takes place, especially in the case of oil-based mud. Analyses In addition to "conventional" PVT analyses:

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 46 of 128

Part 1 Physico-chemical • Characterisation of asphaltenes by “dilution lines”.

• Distribution of “saturates, aromatics, resins and asphaltenes”, (SARA), on heavy fractions (C15+), • detailed compositional analysis by Gas Chromatography (GC) up to C10+ or C20+ (about 300 crude oil components are identified), • Pressure and temperature profile of the stream from the reservoir up to the separator, • Change in the reservoir pressure over time, pressure profile in the near wellbore area. • Floculation point can be, in some cases detected using a light transmission measurement, this method is internally available for atmospheric conditions and need specific procedures for some fluids.

3.2.3 Methodology Risk prediction Natural flocculation is difficult to predict. The existing thermodynamic models are poorly suited to this. A rough evaluation of the “asphaltene risk” is however possible using the characterisation data for asphaltenes, the crude oil and the production parameters for a given field. At this stage, if a possible risk has been identified, it will then be necessary to perform laboratory experiments under Operating P & T conditions, on a downhole sample as defined above.

HPLC column High Pressure Liquid chromatography

filter fluid

fluid

Sapphire window

∆P Injection point

TotalFinalElf E&P

DGEP/SCR/RD/UGF



The bottomhole sample is depleted in the PVT laboratory by transferring the fluid from 2 cells. The transfer line is equipped with a filter, a sapphire window and video microscopy. Flocculation is detected by a combination of: ♦ Video microscopy observation, ♦ Analysis of the pressure drop on the filter edges, ♦ Quantification of the asphaltenes remaining in the filtrate, ♦ Characterization by chromatography HPLC (high pressure liquid chromatography). ♦ Another detection has been developed based on laser detection. This method is not internally used. It should be noted that most known asphaltene problems are discovered a posteriori during field operation. The same approach must also be followed when reservoir gas injection is performed. It may lead to asphaltenes deposits either in the reservoir or in the near well bore of the production well. In particular, if the injected gas is CO2 or a concentrated acid gas Asphaltenes flocculation is widely enhanced. However, when the flocculation occur far from the near well bore it does not seem to damage the well productivity.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3.2.4 Solutions Preventive chemical treatment in the well bore is reserved for special cases. Squeezes of solvent + dispersant mix are used to clean the near wellbore area (HASSI MESSAOUD) but their efficacy for use in prevention is dubious due to the limitations of adsorption on the rock. They are not very realistic for Deep Offshore fields. Chemicals have to be improved for preventive application. They are poorly adsorbed on the formation rock and conversely need to be present at high dosage in the fluid to be efficient. Reservoir management (pressure), when possible, is the only available method to prevent deposits near the wellbore area. Increasing or decreasing the well bore pressure shifts the location of asphaltene formation to inside the reservoir where it has no influence, or to the tubing where it can be chemically controlled. If flocculation occurs inside the tubing, down hole chemical injection is the best practicable solution. Additives for this come in two types: ♦

The first type is intended to retard flocculation or reduce the quantity precipitated.

♦

The second type (the most frequently used and most effective) is intended to disperse the asphaltene flocculates and keep them in suspension so they can be carried away (case of LAGRAVE).

Mechanical methods can be used to remove periodically asphaltenes deposits. Asphaltenes can be more brittle and harder to remove than typical wax deposits. If they form in the near wellbore area mechanical removal is impossible.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3.3 NAPHTENATES

3.3.1 The issues Some biodegraded crude oils contain acids called naphtenic acids. Under certain conditions, these pose problems of emulsion or deposits in the production stream from the well to the surface installation. Such problems generally arise under low water content flow conditions . Desalting units are typically in this condition whatever the production water cut. Problems in the wells may also arise at the beginning of water production (very low water cuts < 1%). In reservoir conditions, naphtenic acids are soluble in the crude oil. They have a molecular weight distribution between 130 and 500+ g/mole, generally centred on C17C18. (See graph: Mass distribution of organic acids present in oil). Naphtenic acids with low molecular weight are relatively soluble in water as soon as the pressure drops. Their passage from the crude oil into the water generates sodium or calcium naphtenates (salts). These salts are surfactants (creation of stable emulsions) and may create naphtenate deposits (mainly with calcium). As yet, no models are available for predicting the behaviour of these crude oils. Experimentation remains necessary. However, the transfer of acids into the water is related to the increase in pH which results from CO2 degassing during the pressure drop. Thus, the richer the reservoir in CO2 the greater will be the scale of the problems. Finally the formation of naphtenate deposits or of stable emulsions is related to: ; the calcium content of the water in the reservoir, ; the organic acid content, ; the bicarbonates present in the water.

3.3.2 Data Required To evaluate the risk, we need the following data: • TAN (Total Acid Number) Measurement of naphtenic acids is obtained by standardised measurement of the TAN expressed in milligrams of KOH (Potassium hydroxide) per gram of crude oil • Distribution of naphtenic acids in distillation cuts (TAN), • Distribution of naphtenic acids in crude oil (GCMS gas chromatography mass spectroscopy), • Standard analysis of the water,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 50 of 128

Part 1 Physico-chemical • Partial pressure of CO2 under reservoir condition, • Reservoir P & T

3.3.3 Methodology The main factors affecting the formation of naphtenates are: Ø

The calcium content of the water, expressed in milli-equivalents per litre (meq/l),

Ø

The water bicarbonate content. This value is calculated from the partial pressure of CO2 in the hydrocarbon, expressed in meq/l,

Ø

The TAN converted to meq/l for evaluation of the potential risk.

To evaluate the naphtenate risk, known as PHA (Potentially Hazardous Acids), we retain the lowest of the values (meq/l) for these 3 parameters. This value represents a qualitative definition of the level of risk (see table below).

-

Field

TAN

Acids (meq/l)

HCO3 (meq/l)

Calcium (meq/l)

PHA (meq/l)

DALIA Prod

1.4

100

18

150

18

DALIA Inject

1.4

100

21

48

21

GRYPHON

4.5

321

12

93

12

HEIDRUN

2.8

200

11

55

11

ALBA

2

143

10

29

10

IME

2.1

150

29

4

4

AFIA

1.2

86

51

4

4

Following previous definition, DALIA is, to date, the most severe case, but the naphtenates risk is limited by the 18-meq/l bicarbonate content, despite an acid content of 100 meq/l (TAN = 1.4). This can be compared with the least severe case of IME which, despite an acid content of 150 meq/l (TAN = 2.1), has a limiting factor of 4 meq/l due to the water calcium content. Evaluation of the risks of deposits is still at the R&D stage. We are currently studying the influence of the molecular weight of acids on problems of production (deposits or emulsions). We are performing three types of tests: •

A detailed distribution of organic acids present in oil

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 51 of 128


Mass distribution of organic acids present in oil •

A distribution of TAN values in the different distillation cuts.

Below, we can see the differences in distribution between four different samples of oil.

ACID DISTRIBUTION 1

0.9

0.8

0.7

0.6 TA N

Dalia 3 Dalia2

0.5

Bilondo Orchidea

0.4

0.3

0.2

0.1

0 0-150

150-230

230-350

350-400 cut

•

Decantation test as a function of pH.

400-550

550+

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 52 of 128


By studying the decantation of water in a Bottle Test for different pH values (adjusted with Sodium hydroxide), it is possible to calculate the quantity of acid transferred into the water phase and the pH value at which the emulsion became stable, and thus determine the ideal pH value at which to perform water-oil separation. The graph below shows decantation performances vs. pH values. As an example, for DALIA 1, when pH is greater than 6.2 – 6.4, the emulsion becomes very stable causing separation difficulties.

very stable emulsion

100

TANc=0.86

TANc=0.22

TANc=0.29

90

TANc=1.4

TANc=0.11

TANc=1.1

% non decanted water at 48h

80 TANc=0.045

70

TANc=0.23

60 TANc=0.33

DALIA2

50 DALIA1

40 30

TANc=0.032

TANc=0.086

ORQUIDEA

TANc=0.56

unstable emulsion

20 10

TANc=0.022 TANc=0.014

TANc=0.062 TANc=0.024

TANc=0.25 TANc=0.01

0 4

5

6

7

8

9

10

11

12

13

14

final pH

3.3.4 Solutions The pH can be controlled either by adding acetic acid or by increasing the separation pressure that limits the release of CO2. 1- The easiest way to deal with the naphtenate problem is to deliberately lower the pH of the crude oil by acid injection. The most often used is acetic acid (low cost), but hydrochloric acid (handling problems) or formic acid (better performance but high cost) can also be used. Acid injection has to be carried out with care since asphaltene floculation/precipitation can arise. The dosage can vary from 100 to 500 ppm (for fields in Nigeria and Cameroon, 300 to 400 ppm predicted for the Miocene reservoirs on Block 17). 2- Since the pH of the solution is directly related to the pressure, the formation of naphtenates can be eliminated by separating water out under pressure (10 to 20 bar).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



3- For naphtenate deposits forming at a high pressure (> 20 bar), demulsifier can be injected as a dispersant to prevent deposits in the production stream. 4- As additional back-up, the following curative treatment can be implemented: Ø to fit the separator with an interfacial collector able to draw out the accumulated emulsion pad Ø to clean the naphtenate soap decanted on the separator/tank bottom using heated water jetting. Ø To inject batch of acid.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Scale

TotalFinalElf E&P

DGEP/SCR/RD/UGF



4. Scale

4.1 The issues Mineral deposits come about as the result of the precipitation of mineral components present in the waters, while paraffin and asphaltene are deposits formed by crude oil components. Due to crystallisation and growth mechanisms, mineral deposits can eventually clog or even completely plug parts of production installations like the well bore area, lines (tubing, flowlines and risers), valves, chokes and topside equipment. Mineral deposits in oil production are generally inorganic salts like calcium carbonates and calcium, strontium or barium sulphates. They may also be composed of iron salts such as sulphides and carbonates (derived from corrosion). The main forms have two origins: Ø

The water produced undergoes changes in temperature and pressure throughout the production stream. The solubility of certain inorganic components of water can then decrease and form deposits by precipitation.

Ø

When two incompatible waters are blended, precipitation of inorganic salt(s) may occur. This is the case with reservoir water containing barium mixed with sea-water (injection water) containing sulphate. It may also occur when reservoir water containing calcium is mixed with a completion or stimulation fluid made from sea-water or containing acid (HF or HCl).

•

CARBONATE

Carbonate deposition results from the variations in solubility in relation to pressure and temperature. When pressure drops, the CO2 dissolved in the water tends to be released, thereby decreasing the solubility of the carbonates. Unlike most mineral deposits, carbonates have inverted solubility, meaning that when the temperature rises, their solubility decreases. Conversely, when the temperature drops, their solubility increases. Generally, reservoir waters are in equilibrium with calcite, the most common carbonate. Risks of carbonate deposits will depend mainly on the pressure and temperature conditions the fluid will encounter. When pressure drops, the risk increases, but when the temperature drops the risk decreases. A fluid that is rapidly decompressed while its temperature is maintained will thus produce carbonate deposits (e.g. ESP). Conversely, a fluid that cools quickly will not form deposits. For example, thermal insulation of deep water pipes to avoid wax deposits may lead to carbonate deposits that would not have formed under a conventional development, as the temperature in this case drops more slowly than the pressure. The main carbonate deposits are calcite and aragonite (calcium carbonate deposits). Siderite (iron carbonates) may also be encountered, generally resulting from corrosion. Other, rarer forms exist, such as witherite (barium carbonate), encountered in HP-HT or vaterite, which is a rarer form of calcium carbonate.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 Physico-chemical •


SULPHATES

Sulphate deposits are most often due to incompatibility between the reservoir water and another fluid such as injection water or stimulation fluids. Reservoir waters may contain significant quantities of calcium, strontium and barium. Calcium sulphate is highly soluble and deposits are therefore seldom encountered. The most common forms of calcium sulphate are gypsum and anhydrite. Some reservoir waters have a very high calcium content (e.g. Persian Gulf oil production, Torpille in Gabon or N'Kossa in Congo). In this case, deposits of calcium sulphate may form by mixing with injected sea-water (reservoir pressure maintenance). One may also come across deposits of calcium sulphate following acidification in a reservoir containing carbonate or dolomite, if the residual fluid encounters sulphated water, either sea-water or the reservoir water itself. For example, some dolomitic reservoirs have a reservoir water highly charged with sulphate. The solubility of strontium sulphate is intermediate between those of calcium and barium sulphate. It is often combined with barium sulphate scale. The solubility of barium sulphate is negligible, around a few ppm. Any contact between a reservoir water containing barium with a water containing sulphate results in the formation of barite (barium sulphate). The incompatibility of sea-water, which is generally used to maintain pressure with reservoir barium laden waters, is a serious problem as soon as the barium content exceeds a few tens of ppm. It becomes a critical problem as soon as this barium content rises above 30 ppm (i.e. 30 – 40 ppm starts to be a problem on Brent fields). Above 300-400 ppm of barium, a feasibility and economical study should be carried out to recommend the most cost effective treatment. •

SILICA

Silica deposits are not encountered in oil production on conventional developments. However, this phenomenon may occur when re-injecting produced water from a sandstone reservoir at high temperature. This can happen indeed, with EOR (steam injection) but also with deep HP/HT production. Sandstone reservoirs are in equilibrium with quartz. At high temperature (T> 130°C), the silica content of the production waters will be such that the if the temperature drops below 20°C, amorphous silica may be precipitated. SALT BRIDGES Salt deposits may occur when liquid reservoir water evaporates in the gas phase under bubble point pressure or for very salty waters due to variations in solubility associated with the drop in temperature. The most common deposits due to evaporation are Halite (sodium chloride) and anhydrite (calcium sulphate). Note also that adding methanol (hydrate prevention) reduces the solubility of salt (NaCl). On waters close to saturation (salinity around 300g/l) the injection of methanol may lead to massive precipitation of salt.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



4.2 Data Required Sampling To evaluate the risks of mineral deposits, representative samples of formation/injection waters are essential. The impact of production water treatments on investments for Deep Offshore developments is becoming a major factor. The lack of data on waters is likely to mask certain risks of plugging or clogging production lines and lead to unacceptable excess costs. The samples must be obtained as early as possible during the exploration phase or, failing this, at the start of the appraisal phase. Water samples obtained by MDT generally allow a correct risk evaluation. However, they are most of the time polluted by mud (even oil based). Calcium carbonate prediction risk still remain possible. Baryum or Strontium sulfate deposit prediction risks becomes difficult. Sampling must be done after production of a significant quantity of water from the aquifer. Knowledge of the mud composition will make it possible to better evaluate the pollution of the MDT sample. A DST sample of the aquifer is ideal to obtain a reliable prediction of deposits. However, this type of sample should be reserved for exceptional cases, especially when the barium content identified by the MDT (to be carried out during appraisal campaign) is extremely high (greater than 400 ppm). In addition to the water sample, a PVT study is needed to evaluate the CO2 and H2S content of the gas. The results are expressed in partial pressure to feed the simulation models. Where water injection is used to maintain pressure, core samples or mineralogical analyses from the producing zones are needed to evaluate risks and define prevention means. This technique allows to determine the water equilibrium with the reservoir rock. Sometimes precipitation may occur in the reservoir itself. Analyses The needed data are: Ø

Sodium, chlorides, calcium, magnesium, potassium, barium, strontium, silica, bicarbonates, sulphates, sulphides, organic acidity, iron, pH and TDS (Total Dissolved Solid) contents.

Ø

Lithium, boron and bromine (useful to validate the samples).

Ø

Certain isotope ratios that may also help in validating samples, such as hydrogen / deuterium, oxygen, strontium and boron. Analysis are generally used by geochimists.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 58 of 128


4.3 Methodology •

Prediction

CARBONATES Based on available water analysis, prediction is conducted with the in-house model: SCALE 2000. Its validity range has been extended to temperatures higher than 200°C and high salinity (300 g/l). The model was developed in co-operation with the BRGM (Bureau de Recherche Géologique et Minier). At a given pressure and temperature with water and gas composition, the model determines the Saturation Index (SI) of the water.

Example of isosaturation index curves Calcium Carbonate 3,00

Po & To condition at reservoir location 2,50 IS=0 2,00

IS=0,5 IS=1 IS=2

1,50

Real profile

T/To 1,00

0,50

Downhole 0,00 0,00 -0,50

0,10

Surface

0,20

0,30

0,40

0,50

0,60

0,70

0,80

0,90

1,00

P/Po

-1,00

When SI < 0, scale formation is impossible. SI = 0 means equilibrium concentration of calcium carbonate is equal to water solubility limit. If SI > 0, the water is super-saturated and calcium carbonate can be deposited. SI =1 means carbonate concentration in water is 10 times higher than the solubility limit. SI=2 means carbonate concentration in water is 100 times higher than the solubility limit.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 59 of 128


On the graph above, the brown curve represents the evolution of T/To and P/Po pairs from down-hole to surface facilities (real profile). Po and To correspond to reservoir conditions. Other curves represent different iso saturation index profiles: SI = 0 - 0.5 - 1 and 2. The relative positioning between these curves and the real profile gives the scaling potential all along production stream. The SCALE 2000 model includes a kinetic module to determine precipitation quantities versus time and thereby obtain a deposit thickness for a given water production.

g of deposit / Kg water Calcite 0.045

0.04

Precip. (+)/Dissolv. (-) mass, (g/kg H20)

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0 0

5

10

15

20

25

Time (minutes)

30

35

40

45

Time

The graph above shows the quantity of "precipitable" deposits versus time. The quantity of deposits is calculated according to the flow-rate and the velocity values in the lines (tubing, flowlines, risers). SULPHATES When sea-water is injected, sulphate from the sea-water will be introduced into the reservoir. If the reservoir water contains barium and/or strontium and calcium, the risk of sulphate deposition (barium, strontium and calcium) is maximum when a certain ratio of reservoir to sea water is reached (see graphs here below). Sulphate deposits in the near well bore area and inside the bottom part of the tubing will strongly affect well productivity. To determine the sulphate deposition risk, we use the same predictive model: Scale 2000. At a given pressure and temperature, the model computes the super-saturation index (figure1) and quantity of deposits for different ratios of reservoir to sea waters (figure 2).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Production water

Figure 1: Supersaturation index as a function of ratio sea water / reservoir water This graph shows the difficulty of treatment linked to super-saturation amplitude. Inhibitor dosage is related not to deposit quantities but to the super saturation index.

Figure 2: Precipitable mass as a function of ratio sea water / reservoir water Precipitable mass indicates the amplitude of the scale issue. The risk of deposit is predicted by analysing the compatibility between sea and reservoir waters, which delivers a pessimistic approach of the associated risk.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Another, more realistic method takes into account the ionic equilibrium reached when the sea water contacts either the reservoir matrix or the reservoir water: Ø Under certain conditions, seawater can exchange its sulphates with carbonates from the matrix. This phenomenon significantly decreases the sulphate content of the injected water on arrival in the near well-bore area and consequently reduces the associated risk of deposition. Ø

The precipitation of sulphate in the reservoir when the sea-water comes into contact with reservoir water can be evaluated by introducing a precipitation module in dedicated models (ECLIPSE UTCHEM).

A combination of the two approaches gives a more realistic estimation of sulphate deposit risks in the near well-bore area. • SALT BRIDGES Salt bridges normally form under conditions of high salinity reservoir water and high reservoir temperature. The composition of the deposit depends on the solubility of all the salts present in the produced water (generally halite NaCl sometimes anhydrite CaSO4) Salt bridge formation in the well is a consequence of three different phenomena: Ø

Water vaporised in the gas phase while pressure decreases,

Ø

The decrease in salt solubility when temperature drops,

Ø

The tendency of soluble water in liquid hydrocarbon to separate when temperature drops.

While the first two tend to favour the formation of salt bridges, the last phenomenon generally eliminates the risk. Thermodynamic models are used to simulate the separation of the water into the gas phase, the oil phase and as a separate aqueous phase. Modelling generally indicates a risk when the total water cut does not exceed a few percent of water.

4.4 Solutions CALCIUM CARBONATE deposits Carbonate deposits are generally controlled by continuous downhole injection of a suitable scale inhibitor. Due to the low seabed temperature and associated high viscosity of the chemical, a solution is made with MEG, thereby serving to inhibit hydrate formation in the injection line as well. Where such prevention fails, remediation is usually possible by acidifying the tubing through the service line.

SULPHATE scaling

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Downhole chemical injection for carbonate scale is usually fairly efficient in curbing sulphate scale, providing the chemicals are injected at sufficient dosage and frequency. This can be used to protect tubing and production stream from deposition. However, if risks have been identified in the well-bore area, other solutions must be considered to prevent deposition: • Water other than sea-water may be used when possible, providing its sulphate content is suitable • Gas injection can replace water injection, thereby obviating any incompatibility • For deepwater wells, sulphate removal will generally be the wisest choice. The residual sulphate content should be around 40 ppm, in order to considerably reduce the risk in the tubing and eliminate the risk in the near well-bore area. • Even though performing squeezes in deepwater wells is a difficult challenge, this solution must be examined, especially for satellite fields. It can prove economical compared to sulphate removal. This is the case when water breakthrough occurs late in the field life, and when squeeze efficiency is demonstrated on the concerned wells. Downhole injection of chemicals below the perforations may also be an answer. It has never yet been applied, due to technological limitations and reduced confidence in its reliability. •

When a risk of calcium sulphate deposits has been identified, reservoir management is the only preventive method. Clearly, it is not recommended to produce a level with sulphated water at the same time as a level producing water containing calcium. The quantity of deposit precipitated can be so vast that no preventive or removal methods will be economically feasible.

•

Curative treatment , as back-up solution, can be implemented: Ø Chemical treatment on singular points (valves, choke, SCSSV…) Ø Mechanical removal (milling or jetting). This intervention in deepwater context are very costly and can be consider only as ultimate back-up.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Miscellaneous Problems


TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 64 of 128


5. Miscellaneous Problems

5.1 RHEOLOGY

5.1.1 Definitions Under laminar flow, the fluid layers slide one against the other. The shear rate is expressed by: γ

=

dv , where dy is the distance between the layers. dy

A surface element in a layer requires a force df to slide along the adjacent layer. Shearing stress is known as the value τ such that τ

=

df . ds

dS

v dy

df

For a fluid said to be Newtonian, the ratio

τ γ

is constant and termed viscosity µ.

For a non-Newtonian fluid, the shear stress can be any function of the shear rate. Nevertheless, by analogy with Newtonian fluids, one terms apparent viscosity µ a the ratio

τ γ

for any pair

γ

and

τ

. The

apparent viscosity therefore varies with shear rate variation. In concrete terms, when a crude oil has non-Newtonian properties, its apparent viscosity depends on the flow rate or the diameter of the pipe. Finally, some non-Newtonian fluids do not flow until a yield value of the shear stress has been applied to them. This is specifically the case of gelled paraffinic crude oil.

5.1.2 The issues There are four types of issues involved:

• Paraffinic crude oils are non-Newtonian at low temperature. The usual hydrodynamic models such as Olga, Prosper, Tacite and others, cannot handle such fluids. • At a temperature close to or below their Pour Point, the rheology of the paraffinic fluids shows a yield value. This means they do not flow in a pipe until a certain pressure has been applied. The

TotalFinalElf E&P

DGEP/SCR/RD/UGF



relevant pressure may be high and may exceed pressure related to the reservoir pressure or the maximum admissible service pressure of the pipe. • Even generally Newtonian asphaltenic or naphthenic fluids are highly viscous, especially at low temperature. The evacuation scheme must be studied in this respect. • Emulsions have a higher viscosity than that of their component fluids. The viscosity of a crude oil containing between 50 and 70% water can be multiplied by a factor of 20. They may be considered as Newtonian in the range of shear rates applied on standard pipes. Even if they present a yield value (mayonnaise is an example of emulsion), it remains negligible in all the cases encountered in oil production. • In a pipe transporting water and oil, an equivalent viscosity can be associated with the fluid if the dispersion of one phase into the other is entirely completed. In that case, the dispersion is equivalent to a single phase. It is pointless, highly inaccurate and even dangerous to try to tune viscosity in a multiphase fluid flow code if the phases are even slightly separated.

5.2 RHEOLOGY OF CRUDE OILS

5.2.1 Data required • Two types of measurements are typically taken: • Viscosity of live oil at reservoir conditions, • Apparent viscosity of stock tank oil under atmospheric pressure as a function of varying temperature and shear rate. • The viscosity of live oil is used for reservoir simulations. It is seldom of any use when considering multiphase transportation in tubing and flow lines, or single-phase flow on stabilised fluids. Moreover, since the measurement is performed in a capillary tube, the results are valid only for Newtonian fluids. This is most probably the case of all fluids at reservoir conditions. Extrapolation to surface conditions should be avoided.

5.2.2 Methodology

5.2.2.1 Measurement principles Viscosity can be measured using various different principles, each specific in nature, and applied according to the product you are working on. In the oil industry we use:

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 Physico-chemical Ø


capillary viscometers with imposed flow: to measure the apparent viscosity of the reservoir fluids at reservoir conditions.

The crude oil circulates in a capillary tube at a fixed flow rate while the pressure drop is measured. The relationship between viscosity and the pressure drop can be expressed using the Poiseuille relation. Shearing cannot be controlled, so they are only suitable for Newtonian fluids, which is generally the case for high temperature reservoir fluids. Ø variable flow capillary or "Canon Fenske" viscometers: for commercial transactions on crude oil. These are U-shaped tubes. One of the branches of the U is filled with the crude oil. The time needed for the levels to balance is a function of viscosity and density. Specific charts are then used to relate this time to viscosity. The viscosity is then termed “kinematic” viscosity to take into account the dependence on density. Shearing is not checked at all, and the measurement is therefore not reliable for non-Newtonian crude oils. In practice, commercial transactions sometimes neglect this problem. Ø coaxial cylinder (or cone-plan) viscometers involve introducing the crude oil between a rotor and a stator. A rotation speed is then imposed on the rotor and the torque needed to ensure this speed is measured. As the value of the gap is known accurately, the shearing can be reliably measured. The relation between the torque and the rotation speed is equivalent to that between the shear stress and the shear rate, thereby enabling apparent viscosity to be measured. The value obtained is independent from the density and is termed “dynamic” viscosity. Consequently, dynamic viscosity is equal to the ratio of Kinematic viscosity to density. This is the only principle that can provide the apparent viscosity value for non-Newtonian fluids at known shearing rates. Finally, models for extrapolating atmospheric measurements to pressurised fluid are under construction but are far from being accurate enough. The major difficulties arise from the necessity to take into account both the influence of the pressure itself and the light components. Naturally, these models can only simulate viscosity for Newtonian fluids. To the best of our knowledge, any ongoing research to date aimed at including non-Newtonian behaviour on pressurised fluids is pure speculation.

5.2.2.2 Paraffinic crude oil Paraffinic crude oils are perfectly Newtonian at temperatures higher than their WAT and generally not very viscous. At low temperatures (below their WAT), they are non-Newtonian, and their rheology depends on their thermo-mechanical history. In simple terms, a flowing crude oil at low temperature can be compared to a suspension of crystals in a hydrocarbon matrix. Its rheological behaviour will thus depend on: Ø

the viscosity of the matrix,

Ø

the number, size and shape, interactions and orientation of the crystals.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



The viscosity of the matrix, considered to have constant composition, will depend only on temperature. On the other hand, the state of the crystalline fraction will be related to the temperature, but also to the way the crude oil has been cooled. For example, by analogy with mineral crystals, we may expect slow cooling to cause the formation of large crystals in small numbers, while conversely, quick cooling will lead to crystallisation in the form of many small crystals. A list of the thermo-mechanical conditions affecting apparent viscosity is given below. It is not intended to be exhaustive, but corresponds to phenomena that have been observed in the laboratory during research projects. It helps account for the extreme variety of parameters to be taken into account to obtain a rheological "model" for paraffinic crude oil. The relevant parameters are: Ø

cooling speed,

Ø

presence or absence of shearing close to the WAT,

Ø

presence or absence of shearing below the WAT,

Ø

shearing value,

Ø

the time when shearing is applied,

Ø

presence or absence of organic or mineral nucleons (water, salt crystals, sediments, etc.),

Ø

presence of additive,

Ø

initial crude oil heating temperature.

Each of these conditions may change the apparent viscosity of the crude oil at low temperature by a magnitude of one or more. Further, significant interactions between these conditions have been noted. As an example, a crude oil subjected to shearing is normally thixotropic (viscosity decreases with time when the crude is sheared) at low temperature but can become rheopectic (viscosity increases with time when the crude is sheared) when doped by a wax inhibitor. Finally, the best recommendations to follow when investigating the rheological behaviour of a paraffinic crude are: Ø

Cover as much as possible of the range of temperature and shear rate needed to represent the conditions of the whole produced stream.

Ø

Determine apparent viscosity as a function of these parameters.

Ø

Use these values in the flow simulators to calculate the average shear rate in the line considered. This last point is of course highly tendentious but should be considered for the moment as affording the best possible guess.

Ø

Assess the yield value if the crude is likely to remain in a flow line at a temperature close to or below its Pour Point. In this aim, , to measure the yield value during lab studies, we have to reproduce as much as possible the the thermo-mechanical history of the fluid.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



5.2.2.3 Asphaltenic or Naphtenic crude oils As already pointed out, although very viscous, these fluids are generally Newtonian. In fact when they are found to be non-Newtonian it is always due to some residual paraffin content. Consequently, any principle for measuring viscosity will be suitable at atmospheric pressure, viscosity being only a question of temperature and following a simple Arhenius law. Two measurements at two temperatures are sufficient to extrapolate to any other temperature. Since these crude oils are often transported blended with a diluent (see § Solutions below), it is sometime useful to assess the viscosity of the blend for different compositions. Some mixing rules have been devised for commercial refinery products. Direct measurements are also possible; the blends are always Newtonian.

5.2.3 Solutions

5.2.3.1 Paraffinic crude oils (Gelling remediation) There is of course no solution for eliminating the uncertainties that arise in trying to assess the apparent viscosity of a paraffinic crude oil. This means that at development stage, it is almost impossible to predict the pressure drop on a flowing pipe or the restart pressure of a stopped pipe if the external temperature is below the WAT or the Pour Point of a paraffinic crude oil. That is why most of the solutions adopted at the production phase are highly conservative. They are generally: Ø

Insulation of the main lines,

Ø

Heat tracing of all the small secondary lines likely to be frequently stopped,

Ø

Chemical injection to reduce apparent viscosity and lower the Pour Point.

Combinations of these solutions are often used. Finally, especially when chemical injection is used, it is strongly recommended to examine the possibilities of reducing the injection rate as soon as production starts up. The chemicals are expensive and often injected at high concentration. The test procedures must be adapted case by case, but should include, whenever possible, iso-pressure sampling, avoiding high shear rate in the sampling valve.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



5.2.3.2 Asphaltenic or Naphtenic crude oil The only, albeit major, issue with these crudes is their high viscosity. To tackle this problem, three methods can be used: Ø

Blending with a light cut or crude oil. Compatibility must be checked before when contemplating asphaltenic crude dilution. Asphaltene flocculation is the major drawback of such a technique

Ø

Insulation or heat tracing of the lines, or re-heating station on long distance pipes

Ø

Transport in emulsion with extraneous water.

There are no chemical solutions. Even conventional Drag Reducer Agents cannot be used since, with such high viscosity, the flow regime is not turbulent. Production of heavy asphaltenic crude oil in deepwater condition should be a severe technical challenge.

5.3 RHEOLOGY OF EMULSIONS An emulsion is a mixture of two non-miscible liquids one of which is dispersed in the form of droplets in the other. The dispersion is stabilised by emulsifying agents at the interface. The liquid dispersed is known as the internal phase, the other as the external phase. On oil fields, emulsions comprise oil and water. Conventionally, an oil emulsion in water is called "direct", while a water emulsion in oil is called "invert". Starting from any type of emulsion, if small quantities of internal phase are progressively added, there comes a point, referred to as the inversion point, where the internal phase becomes the external phase. The emulsifying agents most often encountered are polar components contained in crude oils, like asphaltenes or resins. Certain solids can also play this role (sand, clays, paraffins or corrosion products, etc.).

5.3.1 Data required. The water cut at the inversion point and the associated maximum possible apparent viscosity are the two items of data to be acquired for a new development.

5.3.2 Methodology The apparent viscosity of an emulsion is easy to measure in the laboratory at atmospheric pressure on a coaxial cylinder viscometer. It has been standardised for water cuts from 0 to 100% in a single test. Close to the inversion point, the emulsion is often not stable enough to avoid oil, emulsion and free water segregation into the coaxial viscometer. In fact, what is actually measured is the viscosity of a complex mixture of free oil, free water and only a part of emulsion. That behaviour is shown is the following curve.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 70 of 128


Of course, if the oil itself is non-Newtonian, it is safer to take the measurement at the required shear rate. Some people use empirical laws to assess the viscosity of an emulsion according to its components and their volume fraction. The best known, used in SHG internal code is Eilers' law:

aΦ ( ) µ = e 1−bΦ µ0

, where:

Ø

µ, is the viscosity of the emulsion,

Ø

µ0, is the viscosity of the external phase at the considered temperature,

Ø

a and b are constants dependent on particle size and distribution,

Ø

Φ is the volume fraction of the dispersed phase.

In theory, viscosity can increase infinitely for values corresponding to the more compact stacking of the drops, such that the smaller the size of the dispersed phase particles, the greater the coefficient “a” will be. In practice, laboratory measurements show that the coefficient for increase does not exceed the value 20. Using standard formulae to calculate the viscosity of emulsions often proves a highly conservative approach. We generally consider that their validity range is limited to water contents of the order of 35 to 40%.

VISCOSITY OF THE EMULSIONS 45 40 35 30

µ/µ°

µ/µ° model. µ/µ° meas.

25 20 15 10 5 0 0

20

40

60

BS&W

80

100

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 71 of 128


5.3.3 Solutions Chemicals shift the inversion point towards low BS&W (see DALIA curve below). With constant composition, the viscosity of the emulsion will thus approach that of water.

INFLUENCE OF CHEMICALS (DALIA curve) µ/µ°

8 7

INVERSION POINT

6 5

WITHOUT CHEMICALS

4

µ/µ° meas. µ/µ° meas.

3 2

WITH CHEMICALS

1 0 0

20

40

60

80

100

BSW

The pressure drop in two-phase oil/water lines is thus strongly reduced by injection of demulsifier. Tests performed by ELF CONGO and ELF GABON show that injections in wells by gas-lift can have a highly beneficial action on their productivity. The gains observed (see graph here after) vary according to the case by between 20 and 50% of the well production. Some cases (Tchibouéla) seem to be even more favourable.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Behaviour of well TBM 210 with and without injection of demulsifier via gas-lift The ideal situation is with paraffinic crude oil, BS&W 40 to 60%, and a temperature at wellhead lower than the WAT. Indeed, not only does the product improve the viscosity of the fluid in the tubing but this action also stabilises the fluid flow while avoiding slugs. The effectiveness of G-L is thus significantly improved. The precautions to be taken are as follows: Ø

Check that the product can be injected into the gas-lift (evaporation of the solvent may cause the formation of gums and plugging of the annular pressure response safety valve or even the G-L valve).

Ø

In most cases, the product can be injected diluted at 50% in gas oil, a heavy alcohol or a glycol, thereby completely eliminating the risk of gum formation. This technique also allows for better dosage of the product.

Ø

Inject a plug of solvent to saturate the annular space before introducing the demulsifier

Ø

Systematically stop injection at well shutdown.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



5.4 SEPARATION OF EMULSIONS

5.4.1 General Even though not directly related to deepsea production, emulsion separation is often included in a Flow Assurance study because of its strong relation with the other potential issues, and its clear interaction with the development process scheme.

5.4.2 Required Data Laboratory decantation test results are the only available data. Even if these tests can hardly reproduce the field conditions, they give indications of what will be the degree of difficulty of the water separation. They must be interpreted carefully by expert.

5.4.3 Methodology Two types of tests are generally performed: Ø

Static bottle tests to compare the oil to be produced to already known cases

Ø

Dynamic decantation tests on a pilot rig to assess the efficiency of decantation in conditions as close as possible to those of the actual field situation.

5.4.4 Solutions Decantation issues can seldom be prevented. The only known case concerns crudes containing naphthenic acid. It has been proved that pressurised separation to keep the acidity of the water below a critical pH is sometimes highly beneficial. Most of the time, once the risk is established, the installation has to be specially designed. There are numerous devices devoted to water separation. Most common are: Ø Ø Ø Ø Ø Ø Ø Ø Ø

3 phase separator FWKO, Continuous separation tank, Batch separation tank, Wash tank, Treater heater, Electrical dehydrator Centrifuge, …

In addition to these physical devices, different solutions can be envisaged, alone or combined to increase their efficiency. Basically, the solutions are: Ø

Thermal - temperature enhances the water/oil separation by reducing viscosity,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 1 Physico-chemical Ø


Chemical - by injecting additives upstream of the separator. In this case, even though additives can be roughly pre-selected in the laboratory, it is better to opt for on-site testing.

Ø A number of more ‘exotic’ solutions may be possible for individual cases such as re-circulating the decanted water in order to separate a direct emulsion (oil in water). This drastically reduces the viscosity of the external phase thereby decreasing decantation time. This possibility is particularly well suited to the separation of heavy viscous crudes.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 75 of 128

Part 2 Multiphase Transport


Part 2 - Multiphase Transport Ø HYDRODYNAMICS and PRODUCTIVITY vfriction losses vinstabilities vartificial lift Ø THERMAL MANAGEMENT Ø EROSION – CORROSION Ø WELL METERING

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 2 Multiphase Transport



Hydrodynamics and Productivity

TotalFinalElf E&P

DGEP/SCR/RD/UGF



6. HYDRODYNAMICS AND PRODUCTIVITY

6.1 THE ISSUES The optimal development architecture for many deepwater fields involves the extensive use of multiphase flow lines to reduce infrastructure costs. The aim of the hydrodynamic studies is to design the production lines (flowlines & riser) and associated receiving facilities to evaluate and quantify the flowing performances of various subsea concepts, which are operated in multiphase conditions. At the early design stage theses studies give a general overview or a screening of: •

The productivity,

•

The stability for different configurations

•

the need for Flow Enhancement (Artificial lift).

Then, at a preliminary stage, parametric flow studies are performed, with sensitivities to the following basic data: • Environmental data • Fluid properties oil, gas and water, and flowrates • Reservoir pressure and temperature, Productivity Index • Geometry and mechanical design of wells, mud-line and risers • Development scheme phasing, number of producers. • Process parameter of the topsides. The number of parameters affecting the flow behaviour requires a systematic parametric simulation approach which can be eased by the use of efficient simulation management technique, allowing to run and interpret hundreds of sensitivity cases in a limited amount of time.

6.1.1 Productivity Calculations : For these first, integrated studies, the fluid flow between the reservoir and the host facility is simulated. using an integrated sub-surface/surface/process simulation tool. The pressure losses from reservoir to surface facility occur through the draw-down in the well-bore area, the hydrostatic height and friction in the well, in the sea lines, and in the riser, pressure drop at the chokes (wellhead and top of the riser) and arrival pressure (first stage separation). The reservoir pressure and temperature and the well productivity index are the boundary conditions at the inlet of the transport system. The arrival pressure (first stage separation) is the boundary condition at the outlet of the transport system. The parameters studied are of different nature: • Architecture of wells Architectural: − Architecture of wells (tubing diameter, Productivity Index). − Diameter of flow-lines and risers (6’’, 8’’, 10’’, 12’’).

TotalFinalElf E&P

DGEP/SCR/RD/UGF



− Length and geometrical profile of flow-lines. − Height and geometry of risers. • Reservoir conditions: − temperature − depletion versus efficient water injection. • Fluid characteristics: − GOR. − Water-cut. •Operating conditions − Arrival pressure @ topside inlet. − Choking, topside vs. subsea. The results of simulations performed with the different parameters determine: • the diameter of flow-lines and riser suitable for the expected production rate and according to different lay-outs, • the limit of natural flow and ability to re-start production. Flow-line diameters have to be optimised to cater simultaneously for low pressure drops, sufficient velocities, flowing temperature, temperature autonomy for preservation and respect of erosioncorrosion limit. The results are sensitive to the physical properties of the fluid which are pressure and temperature dependent. Integrated hydrodynamic and thermal calculations are therefore required, based on a prescreening of the overall heat transfer coefficient profile along the various parts (well, mud-line, riser) of the production system. Refer to the Thermal Management theme.

6.1.2 Stability issues For Deepwater developments, the length and geometry of the flow-lines and risers will cause hydrodynamic disturbances outside of the usual range. For such applications, the consequences in the event of severe slugging will cause major problems on the surface installations with a high risk of shutdown, but also in the wells owing to the steep pressure variations , leading eventually to damage on sand control equipment and to well killing. Severe slugging is a major issue in flow integrity. Severe Slugging is a transient phenomenon. Its analysis obviously requires transient simulators. Among which OLGA is the most widely used. The complexity of such dynamic simulator requires a careful selection of the options adapted to each problems and application of design margin due to the actual accuracy of the thermal/hydrodynamic models implemented in the simulator. These transient simulations are used to determine regions of unstable flow and the time-variations of the flow parameters, among which the outlet instantaneous liquid and gas velocities and liquid and gas flowrates, which impact the design and the operation of the topside facilities. The results are quite sensitive to the following parameters: • Slope of flow-lines (mud-line topography) at riser base, manifolds, sleds… • Inside diameter of flow-lines and riser • Water-cut • the GOR • Arrival pressure at Host facility.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



• The physical properties of the fluid which are both pressure and temperature dependent. As for the productivity calculations, integrated hydrodynamic and thermal calculations are required, based on a pre-screening of the overall heat transfer coefficient profile along the various parts (well, mud-line, riser) of the production system. Refer to the Thermal Management theme.

6.1.3 Flow enhancement In addition to previous estimations, a fine analysis of the flow is required for optimisation purposes. Hydrodynamic studies show that two main parameters govern the magnitude of the pressure drop in the evacuation (or production) network (from downhole to host facility): • The vertical elevation governing the gravity losses, in which the flow regime and the watercut play a major role • Friction losses in mud-lines, where gas interaction is predominant compared to hydrostatic head. Artificial Lift is used to speed up production or when natural flow ceases due mainly to the increase of water-cut. Different artificial lift scenarios can be examined: Gas Lift at the riser base is usually implemented, and could be associated or not with Subsea Multiphase Technology such as Multiphase pumping or Water Separation.

6.2 REQUIRED DATA Fluid composition according to: ♦ Compositional model ♦ Sampling analysis ♦ Equation of State tuned to fit measured data

♦ Extrapolation to possible or future conditions (GOR evolution) The fluid model is most generally based on the Peng-Robinson Equation Of State. Well Inflow Ø Estimation of productivity, taking into account a range of Productivity Indexes Ø Evolution of Reservoir pressure. If Pressure maintenance is planned by water or gas injection assumption have to be made for reservoir pressure and gas breakthrough. Then a sensitivity to GOR changes should be performed. Ø Well profile, including consideration of deviation and kick-off depth Ø Production tubing size. Ø Production tubing U-value profile. Production parameters Ø Flowrates (oil, gas water)

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Ø Water injection rate Ø Gas injection rate, composition and injection conditions (P & T°) Ø Process capacity (Liquid, water, gas) Ø Arrival pressure (minimal) Field Architecture and preliminary operating principles Ø Wet / Dry trees Ø FPSO, FPU or tieback to existing facilities Ø Well offset, central / satellite development Ø Single line or loop according to hydrates preservation and pigging issues. Layout distance and profile: Flowline& Riser configuration Ø Length and geometrical profile of lines from wells to the central complex (FPSO, WHP, …). Ø Possible diameters (fabrication & installation constraints) for flowlines and risers. In flow studies, line diameters are defined only by the internal diameter; external diameter depends on technology choices (mechanical design: internal pressure, collapse resistance, corrosion allowance). Ø Insulation performances (U value range and thermal inertia) Ø heating technology (electrical or hot fluid).

6.3 METHODOLOGY Simulation tools ♦

Steady-state simulations are performed with a simulator including a compositional model, a mechanistic hydrodynamic model (typically TACITE two-phase Hydro model or OLGA two-phase hydro model), a rigorous heat transfer model. In case of steady-state flow studies carried out internally, the following software can be used, all including TACITE Hydro model: PROII-US14 (recommended), SHG, PipePhase. Simulation software based on empirical correlations are prohibited for predictive multiphase flow calculations

♦

Transients as Stability and Cooldown behaviour are evaluated with OLGA 2000.

♦

In case of Steady-state and transient simulations combined in the same study, for example for «quick evaluation» of developments, OLGA 2000 is used through out.

♦

In case of segregated flow conditions, the slippage between the oil or condensate and the water phase plays a important part and a three-phase flow approach must be considered. The Water option of OLGA 2000 is then used.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 81 of 128

Part 2 Multiphase Transport 6.3.1 Steady State calculations Line sizing and eruptivity limit

The objective of these calculations is to evaluate productivity constraints in order to size the diameters of the flowlines. Line pressure drop and well performances are calculated using: -

Thermodynamic module: Peng-Robinson or equivalent

-

Mechanistic Hydraulic module: TACITE or OLGA.

-

Thermal module: Steady-state heat transfer between the effluent and the ambient media are calculated with a given U-value (W/m2.K) profile along the various parts of the network (well tubing, flowline and riser (see Thermal Management theme). Note that flow simulators usually refer the U-value to the inner steel pipe-wall (noted Ui ), whereas software used for pipeline mechanical design refer the U-value to the outer steel pipe-wall (noted Uo). Moreover the later do not include the internal convection factor in the Uo calculation and assume an infinite external convection factor. To summarize, the thermal resistance of the insulation layers only is accounted for the evaluation of Uo. It is therefore of some importance to clearly state the reference surface (inner vs outer pipe wall surface) when mentioning U-values and to indicate whether this U-value accounts for external/internal heat convection.

An example of inflow and outflow performances is given on the graphs here after for water cuts of 0 and 40%.

400 375

Wellhead Pressure / Inlet Pressure (bara)

350

GOR = 150 m3/m3 Water cut = 0%

325

6 " ID

300

8" ID 10" ID

275

11" ID IP = 50 INFLOW Performances

250

IP = 75 IP = 200

225

IP = 380

200 175 OUTFLOW NEEDS for a given length

150 125 100 0

5000

10000

15000

20000

25000 Liquid Flowrate (bopd)

30000

35000

40000

45000

50000

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 82 of 128


400 375

GOR = 119 m3/m3 Water cut = 40%

350

6 " ID

Wellhead Pressure / Inlet Pressure (bara)

8" ID 10" ID

325

11" ID IP = 25

300

IP = 50 IP = 100

275

IP = 140 250 225 200 175 150 125 100 0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

Liquid Flowrate (bopd)

The line diameter is chosen according to these results, taking into account the decrease in pressure losses and associated increase in flow transport capacity. The flowline diameter is selected as the minimum diameter providing the required flow capacity and compatible with the available wellhead pressure. Given the present incertainty of the hydrodynamic models, in particular in high riser applications, it recommended to apply some design margin (tentatively, 10% in case of well defined fluid and well known topography, to 20% otherwise) on the calculated pressure drops. We have to consider also stability and erosion/corrosion issues (see § 6.3.2 and § 8 Erosion/Corrosion). Erosion Corrosion) using a common criterion (API RP14E with a C factor of 130 to 170 in absence of solid particles). If a gas breakthrough occurs, erosion/corrosion could become a governing factor and consequently lead to choosing a larger diameter line. From a productivity standpoint, a deeper configuration with a sufficient increase in water depth is beneficial compared to similar developments with a shallower water depth: ♦ at low water cut, pressure losses are mainly due to the column weight, and a higher GOR is quite beneficial ♦ at high water cut, a gas-lift injection point at the riser base is more efficient compared to a shallow injection location.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



6.3.2 Stability studies On a general basis, the stability study would not include 3-phase simulations with OLGA. It is considered that 2-phase simulations are pertinent for preliminary design stage. This however shall be confirmed on a case by case analysis. Prior to any simulation with OLGA 2000, either steady-state or dynamic, the selection of the appropriate version, including the version of the thermodynamic package used to generate the physical properties of the fluid (e.g. PVTSIM, PVTOL or PROII) as well as the appropriate modelling (geometrical grid) and calculation options (time step, slug initiation parameters), shall be discussed with DGEP/SCR/ED/ECP, which is , within TFE, the technical Dept in charge of the follow-up of the OVIP project. •

Prediction of zones with unstable flow

The objective of the prediction is to evaluate potential production problems caused by unstable flow regime in the riser and which may lead to severe slugging, and also to determine and quantify the flowing parameters involved in this phenomenon. It should be noted that transient simulation results are fairly sensitive to many inputs. Therefore an stability study should not be based on a fixed set of inputs but should deliver sensitivity to the inputs. The distribution of slug characteristics (lengths, hold-ups, velocities, frequencies) is of a statistical nature. It is not reasonable to size the topside slug handling capacities on the larger predicted slug volume. Only the impact of most probable ones should be taken in account. Biggest ones should be controlled with adapted procedures. It should be noted frequent small slugs can also give rise to operational difficulties. Unlike onshore or shallow water field development, deep offshore is characterised by a significant uncertainty regarding the mud-line topography. In order to the define the minimum stable flowrate and the effects of topography and arrival pressure, the stability study proceeds in 3 steps: 1. Sensitivity studies to determine the effects of OLGA 2000 options (use of slug tracking according to the flow regimes identified by the steady state study, see above). The standard OLGA dynamic simulator can provide a preliminary estimate of riser-induced slugging. But the largest terraininduced slugs are either not predicted or underestimated. The use of OLGA slug tracking option allows further assessment and usually extends the region of flow instability. 2. Evaluation of the slug characteristics :lengths, frequencies, volumes or hold-ups, velocities. 3. Stability mapping and slugging severity: various slugging regions are mapped as a function of flowrate, GOR, Water Cut, mud-line topography (up-hill or down hill) and arrival pressure as shown on graph here-after.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 84 of 128


WORST RATE / AVERAGE RATE

Severity index Severe slugging

2.5 2.0

Hydrodynamic slugging 1.5

Arrival pressure

1.0

35 bara

0.5 5000

10000

Arrival Pressure

Stability Zone

25 bara 20000 RATE [ bopd] 30000 15 bara 40000

Flowrate

Note that this predictive flow stability mapping , carried out during the project phases is only indicative. A precise quantification can only be achieved the dynamic simulation model has been tuned against insitu measurements. The actual behaviour of the flow will be very sensitive to the details of the flowline geometrical profile and to the process control on the topside receiving facilities. The interaction between the flow behaviour and the process regulation is an important issue that can be investigated by coupling OLGA 2000 with a dynamic process simulator, using the OLGA server interface. This should be contemplated only at an advanced stage of the design, once the detailed features of the process equipment and associated regulation are specified. The slug tracking option determines the minimum flowrate which gives total stability and the transition region between severe slugging and stable flow with smaller slugs. The following graph shows typical results of simulations performed to identify the flow-rate threshold at which flow instabilities appear. Note however that this remains a theoretical approach since in actual operating conditions, the arrival pressure would not be fixed but controlled through the process regulation loops, allowing smooth reaction time, as is usually the case for such service. Moreover the severe instabilities will induce some fluctuations of the back-pressure to the well bore, which in turn may impact the productivity. In fact in the instability zone, unlike what is simulated in the case presented on the graph, the flow boundary is not steady. Therefore some margin (tentatively 10 to 20% on the amplitudes of the fluctuations) must be taken, when using such results as design basis for the slug control system and for developing operating guidelines.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 85 of 128


8" production Flowline ; 5km long & -2% slope; BSW = 0% ; Arrival Pressure 10 bara 7000

Flow rate (m3/d)

6000

5000

4000

Bottom hole Production FPSO Production (Sm3/d)

Q = 2500 m3/d : flow is stable

3000

2000

Severe slugging appears for Q = 2400 m3/d 1000

0 0

2

4

6

8

10

12

14

16

18

20

Time (hours)

Similar generic simulations can be performed for various flowline lengths, diameters or slopes as shown in the graph here after, in which the instability region can easily be identified. Stability in the riser

Minimum flowrate for stability (m3/d)

Effect of the slope on severe slugging

4500

4000

3500

3000

2500

6" ; 10 bar ; 0% 8" ; 10 bar ; 0% 10" ; 10 bar ; 0% BSW

2000 Stable production 1500

1000

Severe slugging

500

0 -2

-1

0

1

2

Slope of the flowline (%)

The graph above shows the impact on flow stability due to downward slopes just before the riser.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 86 of 128

Part 2 Multiphase Transport 6.3.3 Flow enhancement

To evaluate the benefits of the different concepts on a specified case, a detailed flow study from the reservoir to the host surface production facility will return production profiles allowing for the various production constraints. These simulations are conducted by using a Reservoir model (ECLIPSE) and an in-house network simulator PROII, which incorporates the TACITE Hydrodynamic flow model (Steady state two phase mechanistic) and based on boundary conditions from the Reservoir model. This module can track the composition of the production mixture as it actually evolves all along the flow path. The graph below represents an example of flow study results in terms of cumulative production, and enables us to rank the different artificial lift concepts in term of flow performances. cumulative production evolution as a function of the activation type

NP (10^3Sm3) 2.00E+03

1.80E+03

1.60E+03

1.40E+03 riser GL multiphase pump eruptive

1.20E+03

multiphase pump + riser GL

1.00E+03

sep subsea eruptive sep subsea+riser GL

8.00E+02

6.00E+02

4.00E+02

2.00E+02

0.00E+00 0

1000

2000

3000

4000 date (days)

5000

6000

7000

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 87 of 128


6.4 SOLUTIONS SLUG MANAGEMENT Instabilities in flow-line and riser systems are phenomena which have been addressed in investigations or experiments for shallow water developments and still create substantial operational difficulties. Where deepwater developments are concerned, the length and geometry of the flow-lines and risers will cause hydrodynamic disturbances outside the usual ranges. For such applications, the consequences in the event of severe slugging will be major problems not only in the surface installations, but also in the wells owing to the steep pressure variations induced in the subsea transfer lines. The use of active flow instability control systems is crucial in such contexts. Such instabilities arise during the restart phases and under certain low flow rate conditions, and can be aggravated by the geometry of the production sea-lines installed on the seabed, potentially leading to riser-base induced slugging. Different control methods can be applied, based on: • riser head choking, • gas injection at the riser base, • combined riser head choking and riser base injection, • management of incoming flow in the network, Riser base gas lift to prevent riser induced slugging: Gas lift has to be simulated to estimate the injection rate needed to stabilise flow. The graph here after shows the stabilisation effect following gas-lift injection associated to choking down at riser head. Stabilisation with gaslift +valve closure 3

TOTAL MASS FLOW FPSO [kg/s]

Gaslift=0 m /d

50 45

Gaslift=50000 m3/d

Gaslift=100000 m3/d

40 35 30 25 20 15 10 5 0 -5 0

50000

100000 Time [s]

ELF Astrid/Anton Thermal Study - Gas lift study, downward slope

150000

TotalFinalElf E&P

DGEP/SCR/RD/UGF



On the basis of the TFE analysis of the different methods, management of flow instabilities should call on the following methods: • Riser head choking, which limits slippage between the liquid and gas phases and smoothes down the liquid slugs which reach the topside facilities with adverse effects stemming from the back-pressure applied to the fluid stream, • Gas injection at the riser base which lightens the hydrostatic column, increases flowing velocity and breaks down the liquid build-up in the lower part of the riser where the initiation of the Slugging phenomena occurs, • Management of inflow to the subsea production line (flow-lines + riser), manifolding of the wells onto the subsea network provides the flexibility in some cases, of shifting the operating point in a stable flow zone, by an adequate routing of the well production. • Capability for the surface facility to handle limited liquid slugs by reducing liquid retention time. A correct design should allow for adequate slug handling capacity in the inlet separator acting as a slug catcher, from which the liquid drainage rate shall not exceed the design rate of the downstream process, unless excess liquid can be transferred to a dedicated off-spec storage tank prior to be re-circulated in the stabilisation process. In such a case due account of gas blow-by through the emergency drain should be taken care of in the design of the relief system, which may become a difficult governing case. • Production well choking (bean down) to a step-back position to avoid propagation of pressure variation into the tubing and protect down-hole sand control equipment. This operation has to be performed in conjunction with gas injection at the riser base. In addition to the various control methods above, prediction (“real time” simulation code) and detection (sensors) means should be used to anticipate to the maximum the disrupting effects caused by the flow instabilities and increase the possibilities of eliminating them. Control of flow instabilities must not penalise overall well production but, on the contrary, constantly seek the optimum in relation to the capacities of the wells and the installations. From the analysis of the advantages and drawbacks of each method, the wise choice would appear to be the combination of all means, so as to mitigate the shortcomings of each and obtain a consistent overall control system. As regards implementation and feedback, there are little or no interesting offers for command-control software to be combined with the control system. In this field, our in-house experience based on application of well control systems (FCW) has no identified equivalent and its extrapolation to the control of deepwater fluxes, integrating the well, flow lines and risers, would appear to be the essential path to follow in controlling slugging effects. The Overall Control of Wells and Risers developed for deepwater application, executes all commands on actuators from X-mas-tree to production and test manifold, including chemical injection treatment. The main purpose of OCWR for oil production is to reach a maximum or target flow-rate, while: - avoiding or limiting flow instabilities in risers and maintaining them in acceptable ranges, - preserving the near well-bore area by avoiding substantial pressure variations, - limiting general production shutdowns due to surface installations, by anticipation, - avoiding hydrate formation and wax deposition in the network, - protecting subsea equipment. The Overall Control of Wells and Risers operates with more sophisticated external calculation tools used as input, i.e.: - On-line: well and riser production rate estimation calculated by Well Monitoring system,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Rev. : 0 – 01 / 2002


Page 89 of 128

Part 2 Multiphase Transport -

Ref. : UGF/2001-135

Off-line: assistance for decision and determination of OCWR parameters, such as network calculation for optimal routing of wells. OCWR ARCHITECTURE FOR LOOP ARRANGEMENT

RCHK

RISER

Méthanol

Hot Oil

P r o d u c t i o n

T e s t

GLFCV

G a s GLFR

W ELL

L i f t

Service line

CHK

LinRig e ht Pro d .

The main measurements used for the wells and riser control are : Wells and subsea manifold Main sensors and calculations

§

Well bottom flowing pressure

§

Well Head Pressure and Temperature

Riser head §

Gaslift valve downstream pressure (calculation of riser bottom pressure)

§

Riser Head Pressure

§

Choke downstream pressure

§

Riser Liquid Flow Rate (Well Monitoring)

§

Sand production detector

§

§

Manifold Production Pressure and Temperature

Riser Gas Flow Rate (Well Monitoring)

§

Riser Bottom Pressure (Internal OCWR))

§

Estimated well liquid rate (Well Monitoring)

Main actuators Well Choke

§

Riser pressure gradient (Internal OCWR)

Riser choke Gaslift flow control valve

Topside Oil system : §

Separators level

Gas system : §

Separators pressure

§

“ FPSO ” temperature

Gaslift : §

network pressure

§

compressor intake rate

§

compressor status

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 90 of 128


Stable

OCWR general principles are presented in the following simplified diagram:

Well choke to max

Riser stabilization

Wells and Riser TARGET

Well rate=target?

Well Monitoring no

Stability

Branch stability ?

Riser bottom pressure Riser head pressure

Wells Opening Naturally flowing riser start-up

Modulation on selected risers

Bottlenecks load ? yes

Separator system Gas handling

Unstable

Gaslifted riser Unload Start-up of wells

Shut-in

Load shedding on selected risers

Production target

START-UP PHASE PRODUCTION PHASE

Following table sum-up ways and means to reach the OCWR main principles during start-up and production phases for oil production : Objective

Actions during start-up phase

Action during production phase

Production

Start-up sequence : riser kickoff

Well target liquid rate established through : - Well target choke (objective) - Riser target bottom pressure (objective) - routing (prevention) - comparison between real and target liquid rate (regulation) Maximum rate regard to topside installations: - modulation of selected risers regard to bottlenecks - load shedding of selected risers and well choke to beaned according to bottlenecks - flow stability

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Rev. : 0 – 01 / 2002


Page 91 of 128


Stability

Instability prevention made by : - routing of wells on riser - start-up sequence: riser and well stabilisation through gaslift injection and riser head choke

Ref. : UGF/2001-135

Instability prevention made by : - routing of wells on riser - riser and well “target” choke Detection : - riser bottom pressure checking - riser head pressure checking Stability recovering : - gaslift injection - riser head choke

Well bore damage prevention

Prevention by : - well and riser chokes ramp - stability

Prevention by : - target choke - stability

Detection : - high riser head and bottom variations survey - too low riser head and bottom pressure survey - backflow risks survey - sand detector survey

Detection : - too low riser head and bottom pressure survey - backflow risks survey - sand detector survey

Action - choke ramp suspension - well choke beaned

Anti-hydrates

Action - alarm - well choke beaned

Prevention : Prevention : - routing of wells according to minimum - continuous methanol injection flowrate per line - riser head temperature survey and - automatic injection of methanol before automatic injection of MeOH if required well opening -

well head choke ∆P limitation to avoid local cooling

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 92 of 128

Part 2 Multiphase Transport FLOW ENHANCEMENT / MULTIPHASE TECHNOLOGY

Due to eruptivity limitation, flow enhancement can be applied at different locations by providing mechanical energy or lightening the produced fluid: • In the wells by gas-lift or pumping (ESP) • in the riser by gas-lift • at the mud-line near the wellheads or near the riser base by using multiphase pumps or a water separation system. • Artificial lift in the wells For deep-water applications, possible artificial lift systems include down-hole equipment. All the techniques for boosting energy in deepwater subsea wells have to contend with the major constraint of keeping risky and costly intervention to a minimum. Our experience of wells gas-lifted and especially of lift with ESP has shown that these techniques need intervention in the wells for maintenance or repair of down-hole components. The complexity of down-hole equipment limits its reliability and entails costly well intervention and associated production losses. This option is to be evaluated, but only for wellhead platforms (dry tree architecture), mainly due to intervention consequences in subsea condition. • Gas lift at riser base High pressure gas from (export or re-injection network) the compressors is injected at the riser base and enhances flow making use of a well known artificial lift technique used to improve well performance with low associated CAPEX. The efficiency of riser gas-lift increases with the water depth and takes advantage of riser height. Dedicated studies provide riser base flowing pressure versus liquid rates for different ratios of injected gas (VLP curves) and for different water cuts. The results of simulations enable an appropriate GLRi to be chosen, providing a nominal gas-lift flow rate per riser for the design conditions. P re ssure e volution @ rise r ba se 140

Riser basePressure(bar)

120

100 WLR % 80

0 60 80

60

40

20

0 0

500 Inje cte d Gas

1000 KSm 3/j

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Special attention must be paid to erosion due to gas expansion causing high velocity in the upper part of the riser (refer § 6.3.1). In addition, before any need for continuous lift, the kick off phase during the re-start phase can be handled easily by gas-lift. Gas injection can also be used when low flow conditions prevail to control or prevent slugging phenomena in the risers. This technique can be advantageously combined with other emerging subsea techniques such as multiphase pumping (Topacio development) or water separation (Troll C development). Gas-Lift is a reliable lift technique to use in deepwater condition because it needs very simple and reliable “static” subsea equipment. Subsea multiphase pumping Multiphase pumping is installed on the seabed where it accelerates oil production from the outset (maximise draw-down when associated with gas-lift), and can shorten the recovery time considerably. For satellite development, this technique can extend the step-out distance to existing infrastructures, provided that the temperature preservation of produced fluids is not a bottleneck (see § 7 Thermal management) To cope with the expected range of total flow produced by the wells, flexibility can be provided by a modular arrangement of 2 or more pumps in parallel. In addition, this configuration minimises maintenance weight for the pump unit and, in the event of failure of one pump, the other(s) can reliably stand in to maintain sufficient flow. A multiphase pump arrangement module is a balanced choice between: • Flow energy needs related to rising hydrostatic pressure brought about by increased watercut, and the consequent reduction in the gas volume fraction, • Flexibility to handle variations in gas content, • Technical limitations related to high power demand and its distribution, • Required level of redundancy and reliability to satisfy, • Maintenance criteria and associated limitations due to intervention vessel availability and performances. Installation of multiphase pumps at the mud line offers many advantages such as limited gas volume fraction (better efficiency of multiphase pump) or early energy addition to the evacuation system. But integration in a deepwater environment requires a qualification process for such components as electric motor (RPM, sealing, cooling, power demand) or power distribution (umbilical, frequency converter, transformer, wetmateable connector). The reliability of such technique is a key point in responding to deepwater application challenges. Topacio field in Equatorial Guinea (Exxon-Mobil) application (2 x 1,1 Mw pump units). The multiphase pumping units use the helico-axial technology and are driven by electrical motor 1Mw @ 5000 rpm wxith an expected differential pressure of 35 bar. The pump and the motor are retrievable according to a cartridge design. The tie-back distance is 8,5 kilometres to Zafiro host facility. This application should give us the maturity of such technique in 500 meters water depth. Pump module was installed in august 2000. To date this implementation is the only subsea application of the multiphase pumping technology. Subsea Separation (emerging technology) Subsea separation should consider two alternatives: water / hydrocarbons separation and gas/ liquid separation to be deployed either near the wellheads or at the riser base. Subsea water separation would prevent the increase of the hydrostatic pressure stemming from watercut evolution, and lower the liquid volume to transport through the sub-sea line pipe.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



The main advantages are: • to lighten the head of hydrostatic column thus maintaining natural flow productivity, • to de-bottleneck liquid or water capacity limitation, • to modify Hydrates management, • to release evacuation capacity allowing connection of in-fill wells. However, water separation induces a significant decrease in total liquid flow-rate in the riser which, for high water-cuts, can lead to instabilities like severe slugging in the risers, unless riser head choking is performed, with a detrimental effect on the recovery, or a well management procedure is implemented to connect an adequate number of producers per line. Eliminating water decreases the heat capacity rate with consequences on the fluid temperature which can induce viscosity effects, wax deposition on pipe walls or increase hydrate formation risks. This technique is based on water separation efficiency in deepwater conditions and also on high reliability of the water disposal system (Injectivity Index Maintenance) to accept on a long term basis an oily water with solids. The main questions are sand handling and de-oiling of the separated water. The first approach is that water separation can be subsea deployed if natural separation is quite good. Subsea separation can be foreseen in deepwater with deployment at the riser base or at the wellhead depending on the targeted benefits. Subsea water separation is potentially efficient (hydrodynamic benefit) once the production becomes hydrated after a few years. However, it entails the transfer of active process components to the mudline. This technology is still maturing but represents a major advantage for deep-water oil field applications where water drive depletion is predominant.

The first application in subsea water separation has been performed by Norsk Hydro. A unit has been installed by 350 meters water depth at Troll C field The pilot operates as a separator and re-injection station for 8 wells with a capacity of 10 000 m3/day. For water re-injection the solution consists in a variable speed water pump (2 Mw) with injection rate 6000 m3/day. The unit was started in may 2000 with an early failure on the electrical wet meteable connector (MECON connector). After retrieving the th electrical motor, the connector was repaired and the unit was restarted the 25 of august 2001. For gas dominated systems: Gas-liquid separation allows under certain condition a better valorisation of the gas, it will also help the gas transfer over huge distance.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Thermal Management

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7. THERMAL MANAGEMENT Note : This theme is closely integrated with the Hydrodynamic theme as both use common simulation tools and flow and thermal parameters interact with each other

7.1 THE ISSUES Thermal management of the production system is of primary importance for coping with the following requirements: Ø avoid hydrate formation on a permanent basis during normal operation, then during shutdown/restart phases or during downgraded flowing conditions, Ø minimise the pigging requirement (lower the frequency) for wax removal, Ø improve productivity as lower temperature can entail higher viscosity, Ø avoid blockages when production restarts in the case of a high Pour Point value (gelling effect), Ø decrease heating requirements on topsides as the process requires the fluid to be heated at high temperature to reach the commercial specifications. This however, should be considered more as a “by product” of the subsea thermal management, than as a design requirement. It is far easier and more of a normal practice to include heating facilities on the topsides, adequately sized for the whole range of operating conditions, than to go for extreme subsea thermal insulation or heating techniques. Thermal requirements are of several types and depend on the system, i.e. active heating or passive insulation. They can be divided into two categories: Ø Ø

Flowing conditions, common to all solutions (passive or active) as heating should be minimised while producing, Shutdown conditions for which the requirements differ according to the selected solution with respect to actual risk of solid deposition.

The thermal requirements for both flowing and shutdown conditions are specified for various cases intended to cover nominal and transient modes likely to be anticipated during field life.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.2 REQUIRED DATA •

Normal flowing conditions: Ø Liquid flowrate: nominal, min. and max. values for normal and downgraded conditions Ø Fluid composition, inlet temperature Ø GOR evolution or range Ø Water Cut profile or range Ø Gas-Lift injection (rate and gas temperature and composition) Ø Arrival pressure at surface facilities Ø Arrival temperature specification if any (wax constraint) Ø Geometry of wells, flowlines & risers (length, topography).

•

Shutdown & restart conditions Ø Hydrate/wax prevention strategy for the different sections of the production system and for scheduled and unscheduled shutdowns Ø General basis for cool-down time related to operation philosophy (topsides operations, preservation procedures fluid replacement, depressurisation of lines, use of active heating during shutdown, restart sequences…) Ø Passive insulation or active heating system Ø Geometry of flowlines (length, topography).

7.3 METHODOLOGY THERMAL BEHAVIOUR: INSULATION REQUIREMENTS Thermal calculations are performed, at the parametric stage of study with the purpose of assessing the overall thermal insulation requirement (U-value), with PROII/US14 simulator in steady state flowing conditions and with OLGA to investigate the shut-down and restart conditions. In case of thermal study sub-contracted to a Service or Engineering company, or in case of study covering both normal flowing conditions and shut-down conditions, OLGA shall be used throughout. Temperature drops are mainly due to heat exchanges with the cold sea-water and, and to Joule Thomson effect due to gas expansion mainly in the riser section. The aim of these calculation is to determine the required Overall Heat Transfer Coefficient (U value) for the different parts of the production network ( well tubing, flowline and riser ) in order to ensure a minimal arrival temperature, and an adequate cooldown delay to avoid hydrates risk in shut-down conditions.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 98 of 128


This thermal evaluation has to take into account heat exchange arising from conduction, free convection and possible active heating device (electrical or fluid transfer). ). The understanding and modelling of the thermal transfers in deepwater architectures are generally more complex than for shallow water or onshore applications, due to the non-cylindrical/non-symmetrical geometry of the systems such as bundles or other complex insulation concepts. Therefore the use of CFD tools like FLUENT or CFX may be required to calculate U values.

7.3.1 Under permanent flowing conditions In most severe case the key point is the wax constraint (minimum arrival temperature). Thermal calculations are performed with a network simulator. The results give the temperature for over the whole range of normal flowing conditions (Minimum to Nominal flowrates and taking care of the GOR, and water cut variation over the field life) chiefly with respect to wax or hydrates concerns. Note that simulators combining hydrodynamic and heat transfer simulations usually refer the U-value to the inner steel pipe-wall (noted Ui ), whereas software used for pipeline mechanical design refer the Uvalue to the outer steel pipe-wall (noted Uo). Moreover the later do not include the internal convection factor in the Uo calculation and assume an infinite external convection factor. To summarise, the thermal resistance of the insulation layers only is accounted for on the evaluation of Uo. It is therefore of some importance to clearly state the reference surface (inner vs outer pipe wall surface) when mentioning U-values and to indicate whether this U-value accounts for external/internal heat convection. Wellhead temperatures are estimated for various flowrates, assuming different heat transfer coefficients obtained with the definition of insulation in the annulus and in the tubing. The effects of pressure drops at the production choke have to be fully integrated in thermal modelling. For a given wellhead temperature, heat transfer coefficient simulations of various pipe geometry can be run to identify the limitation on arrival temperature as shown in the graph below. Dalia flow studies

Temperature on the FPSO ; BSW = 0 45

40

Temperature on FPSO (°C)

35

30 6" 8" 6" 8"

25

20

15

10

5 Insulation : 50 m m of polyurethane (λ λ =0.11 W / °C)

0 0

500

1000

1500

2000 liquid flow rate (m 3/d)

2500

3000

3500

4000

L=5km L=5km L=10km L=10km

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 99 of 128


Similar simulations can be performed for various insulation levels with a given geometry as shown in the following graph.

Arrival temperature on FPSO versus liquid flowrate 2 for various insulations (Ui in W/m °C) - 10" line 5 km long 42

FPSO arrival temp. (°C)

40 U=2.5 38

U=2 U=1.5 U=1

36

U=0.5 34

32

30 1000

2000

3000

4000

5000

6000

7000

8000

9000

oil flowrate (m3/d)

7.3.2 Under shutdown and restart conditions The key point is the preservation of the subsea production network from hydrates formation. After a shutdown, the distribution of gas and liquid inside the production flowlines and risers will depend mainly upon their geometry, but also upon the settle out pressure profile resulting from the sequence of the shut-down. A line going upward to the riser base could, depending on the fluid GOR, be predominantly liquid-filled at the dispatcher end and gas-filled further along the line up the riser base. However, due to local topography and undulations on the seabed, there will be horizontal or local high point sections, which will be predominantly gas-filled and may come in contact with sufficient water to form a hydrate plug. In light oil (high GOR) field development, the pressure profile and the hold-up profile may also depend on the packing status of the line prior to and associated to the shutdown sequence. The Flow line (mudline & riser) insulation requirement is therefore dictated by cool-down of gas filled parts of the network, with respect to the operating parameters prior and during the shutdown sequence. The thermal capacitance (heat capacity and mass) of the various materials surrounding the flow-lines is the parameter governing the performance of the insulation during transient conditions. Simulations to assess this performance are performed with OLGA. The results give the time it will take in hours after a shutdown to reach a given temperature determined essentially with respect to hydrate concerns. The graph below shows response time at different locations along the flow path for different heat transfer coefficients.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 100 of 128

Part 2 Multiphase Transport Dalia Flow studies - Temperature calculations

Case : 5 km x 8" flowline ; 2000 m3/d ; Wellhead temp. = 48°C ; BSW = 0 ; 10 bar on FPSO Time after shut down before 20°C is reached

45

40

35

Time (hours)

30 well Head mid flow line bottom of riser mid of riser (gas)

25

20

15

10

5

0 0

1

2

3

4

5

6

7

8

Pipe thermal transfert coefficient (W/m²°C)

Sensitivity studies In order to better analyse the previous results, various sensitivity studies are carried out: Ø

Line Geometry according to different Slope value, and including dips and bumps as relevant.

Ø

Closure sequences between well valves and surface choke with associated depressurisation,

Ø

Fluid composition and GOR variation reproducing gas breakthrough.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.4 SOLUTIONS To improve thermal performances during the cool-down associated with shutdown phase, three means can be applied: 1. Limit heat exchange (losses) with good insulation (low U values) 2. Heat the lines (electrical or hot fluid transfer by circulation). 3. Increase the thermal capacitance of the system (higher m.Cp values) The state of the art thermal management solutions can be divided into two categories, passive insulation and active heating. Ø

In passive insulation, the industry has advanced traditional materials such as syntactic foams for application in water depth up to 3000 m. Some new insulation concepts are vacuum insulated pipe, jacketed insulation with microporous insulation with or without vacuum, aerogel beads, and area gel vacuum panels. A new concept of Phase Change Material is developed, which uses latent heat to extend the cooldown time.

Ø

In active heating, current methods include hot fluid heating and electrical heating. Hot fluid heating is the most reliable option and may be also more economical. Drawbacks are the temperature limit, ,non-uniform heating along the flowline, and large Electrical heating includes SECT heating, low power electrical heat traced, and direct electrical heating. The first two are under development. The direct heating methods have been used for 2-3 fields, with apparent successes. Electrically heated systems have issues of voltage, reliability of electrical connections, interference, corrosion, and possibility of broken electrical cables.

Thermal tests Full-scale thermal tests can be expensive, but is necessary to ensure the performance of the installed flowline system is as designed. The expense of thermal tests is also very small when compared with the remediation (pigging issues or hydrate removal) cost due to insulation failure. Thorough thermal tests should be performed on new thermal management (insulation, active heating etc) concepts, to prove the concepts and to develop proper analytical tools for use in design. For actively heated systems have many more issues to cover than tests of passive insulation systems, due to the addition of system function and components. Natural convection in insulation is the major cause of many past insulation problems. For open-cell polyurethane foams under pressure, the pre-cast insulation allows significant natural convection to occur and degrade the thermal performance by a very large extent. Similar problems are expected in insulation with porous and permeable insulation such as fiber glass and rockwool, if high pressure is applied. For wet insulation system such as external syntactic foam insulation, pre-cast modules result in gaps in insulation and between insulation and flowline/riser. Strong natural convection effect was observed in thermal tests. One method to minimize the natural convection effect is to cast the insulation material in place. For syntactic foams, this eliminates the natural convection in the insulation. For pressurized open-cell polyurethane foams, there are still natural convection in the cells, but the effect on OHTC is 20% or less. There are two types of thermal management solutions: passive insulation and active heating. This section reviews the state-of-the-art thermal management solutions of both types, particularly in issues related to testing of these solutions. The solutions will be discussed in the context of offshore developments.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



Passive insulations currently in use or under development include the following: Ø· External Insulation Ø· Jacketed Insulation (pipe in pipe or bundle) Ø· Burial as supplemental insulation Actively heated systems include Ø· Hot water/liquid heating Ø Electrical Heating by Skin Effect Current Tracing Ø· Low power electrical heating by multiple heat traced cables Ø· Direct electrical heating using the natural electrical impedance of pipes

7.4.1 Passive Insulation

7.4.1.1 External Insulation With external insulation method for offshore developments, the pipeline is insulated by materials of sufficient strength to withstand the hydrostatic pressure at the pipeline depth. The materials also should have sufficiently low levels water absorption. Commonly used materials include solid polypropylene, syntactic foams (polypropylene, urethane or epoxy based), and elastomers. For deepwater applications, the strength of syntactic foams can be increased by additives in the form of spheres. However the additives tend to increase the thermal conductivity of the foam and reduce the thermal performance. Some typical thermal conductivity values are listed below: Ø· Solid polypropylene, 0.2 – 0.22 W/m/K Ø· Polyurethane syntactic foam for 900 m and below: 0.08 W/m/K Ø· Syntactic for 3000 m water depth: 0.12 W/m/K Ø· Carazite (syntactic foam with glass spheres): 0.16 W/m/K A recent development in external insulation is the possibility of reeling insulated pipelines. Compared with jacketed insulation, external insulation is less costly for shallow water depth and mild insulation requirement such as those achievable by up to 3 inches of insulation. With large water depth or high insulation requirement, external insulation becomes either too expensive or technically not feasible. A key point in ensuring the success of external insulation is to eliminate the flow paths from the flowline to the ambient water. Even small paths of flow can cause the thermal performance to be much worse than designed due to the natural convection by seawater. In general, pre-cast insulation modules should not be used. If used, care should be taken to remove all gaps. Testing of the thermal performances of these systems should include pre-design tests of material properties at the operating conditions and tests of full- or reduced-scale models simulating the operating conditions. Since the insulation material is under hydrostatic pressure during the field life which can be over 20 years, methods should be developed to determine the long term behaviour of water absorption, hydrostatic creep, thermal conductivity. The industry does not have much experience in these areas, but some efforts are being made. (IFP studies) plans to evaluate about 15 materials (syntactic foams with different resins, elastomers, etc), up to 300 bar hydrostatic pressure, and up to 10,000 hours of ageing period.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.4.1.2 Jacketed Insulation with Air or Pressurized Gas For high insulation requirement, or for large water depth, jacketed insulation systems can be used. The low thermal conductivity but pressure sensitive insulation materials on the flowline are protected by the jacket pipe. Because of the additional pipe, jacketed insulation is usually quite costly in material and installation. However, recent progress in installation capabilities made the cost much lower than a few years ago. Polyurethane foams (PUF) are frequently used in jacket insulation because of their low cost (less than 20% of the volume is the polyurethane material) and low thermal conductivity. Closed cell polyurethane foam (CCPUF) has a thermal conductivity of 0.03 W/m/K and is used in pipe-in-pipe configuration where the jacket pipe takes the full hydrostatic pressure. Open cell polyurethane foam (OCPUF) with open cell percentage at least 90% is usually used in bundles and sometimes in pipe-in-pipe systems. The open cells allow pressurised gas (normally nitrogen) to permeate through without damage. The pressurisation enables the jacket pipe to be designed to a lower pressure, thus reduces its weight and cost. Due to natural convection inside the cells, the effective thermal conductivity of OCPUF increases with pressure. With nitrogen pressure, thermal conductivity measured 0.04 W/m/K at 0 psig increased to 0.063 W/m/K at 1500 psig. Application method of the foam is extremely important. Through the research following the BP Troika project, it is clear that any free flow paths for the gas from the flowline to the jacket pipe must be eliminated for the success of OCPUF insulation. The BP Nile project used reeled pipe-in-pipe system with 24 mm thick layer of Wacker WDS, which were packaged microporous fumed silicam insulation, from German company Wacker Chemie GmbH. The effective thermal conductivity of the package was about 0.02 W/m/K. Thermal Tests The thermal tests for these insulation systems should simulate the operating conditions as closely as possible, and be accompanies by extensive analysis.

7.4.1.3 Jacketed Insulation with Vacuum Several types of jacketed insulation with vacuum, for even higher level of insulation or for smaller jacket size, have been developed. Vacuum Insulated Pipe One is the extension of vacuum insulated tubing concept to pipeline. A vacuum is pulled in the annulus between the flowline and the jacket pipe. Due to the reduction of convection and conduction heat transfer by air, the insulation is highly effective. However, there has not been field application of the vacuum insulated pipe concept. Three main issues still need to be resolved: (1) the risk of lost vacuum and the subsequent total loss of insulation, (2) the ability to create the vacuum after installation, (3) The maintenance of vacuum. Izoflex by ITP Izoflex insulation is a microporous insulation patented by ITP InterPipe. With vacuum in the annulus, the effective thermal conductivity can be reduced to below 0.01 W/m/K. On TCHIBELI field (production started in April 2000) , offshore Congo a PIP ITP is used, proving that 2 the overall heat transfer coefficient for the 25 km length 10”/12’’ flowline was 0.6 W/m °K with 15 mm of Izoflex insulation. Aerogel Beads and Aerogel Vacuum Package Aerogel is a relatively new material based on silica and has very low effective thermal conductivity, about 60% of the value for Wacker Vacuum Insulation Panels (0.02 W/m/K) as beads and 25% of the value for Wacker VIP with 1 mbar Vacuum Insulated Panel. According to the manufacturer, the panels have a service life of 20 years, after which the external gas can penetrate the panel sealing and increases the effective thermal conductivity. For long service life

TotalFinalElf E&P

DGEP/SCR/RD/UGF



applications, the pipe-in-pipe annulus where the panels are applied can be vacuumed to a moderate level. Thermal Tests The issues in thermal tests of these insulation systems are the same as those of OCPUF insulated systems, except that the influence of vacuum levels needs to be determined by the tests.

7.4.1.4 Jacketed Insulation with Phase Change Material (ILS) Bouygues Offshore and IFP have recently developed an insulation system using liquid/solid phase change material, for the purpose of extending flowline cooldown time after production shutdown. The ILS system is housed between the pipe-in-pipe annulus. During normal production, the ILS temperature increases and becomes amorphous. In the process it absorbs heat from the flowline. Since the ILS does not become liquid, there is no natural convection in the annulus so the effective thermal conductivity can be relatively small, about 0.12 – 0.17 W/m/K. After production shutdown, the flowline and the ILS temperature drop with time. When the phase transition temperature (5-10 C above the hydrate formation temperature) is reached, the ILS gradually crystallises and produces heat. The heat generated helps the flowline temperature to be kept above the hydrate formation temperature for a longer time than conventional insulation of similar thermal conductivity. The tests for insulation systems using ILS should include the following areas: Ø· Effective thermal conductivity at a range of temperature Ø· Material stability, particularly with ageing and cycles of phase changes Ø· Degradation in the phase change process with ageing and change cycles Ø· Determination of latent heat and phase change temperatures Ø· Extent of increase in flowline warm-up time.

7.4.1.5 Insulation Consideration for Risers Deepwater risers pose particular challenges to thermal management. In addition to the need to reduce riser weight and the resulted limits in insulation method and material selections, the Joule-Thompson cooling and potential energy effects reduces the fluid temperature significantly. Due to the long vertical length, natural convection in open cell insulation (even if foamed-in-place), or other insulation where natural convection flow paths exist, may be more severe than in that in the flowline on the sea floor. Insulation for risers should be tested separately from the flowlines, particularly to determine the natural convection effect. Obviously full-scale tests with vertical risers are much more complicated and more expensive than flowline tests, which can be done either in air or water tank. Extensive analysis should be done to guide the thermal tests. Some possibilities of facilities for thermal tests of risers are Ø Full size but short (10 m or so instead of 25-50 m) vertical test pieces with good end insulation. The test piece can be exposed to air, or inside a jacket for ambient temperature control by coolant. Ø Near horizontal full-scale and long test pieces with a range of inclination angles, to an angle as high as practical. The data are used to validate the computational heat transfer models which the design is based on. Ø Reduced size flowline and insulation test piece in vertical orientation to validate the computational heat transfer models which the design is based on.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.4.2 Active Heating Systems Common heating options are hot water (or other hot fluids) heating and electrical heating The following electrical heating methods will be discussed below. Ø· Skin Effect Current Tracing (SECT) Ø· Multiple cable heat traced methods Ø· Direct electrical heating (DEH)

7.4.2.1 Hot Fluid (Water) Heating Experience and Advantages Hot water heating system is the most reliable option for offshore installations, since it only consists of pipe and pipe connections subsea. The concept also has been used numerous times for offshore developments. The heating and pumping facilities are located on the host platform. Frequently hot water heating is also the most economical because the platform exhaust/waste heat can be used to heat the water, so the only additional power requirement is for pumping. There are two types of hot water heating. One is indirect heating in which the water pipe is attached to the flowline pipe (tracing). The other is direct heating in which the water pipe encases the flowline pipe and water flows through the annulus. Recent projects that use or can use the indirect heating concept include Enserch MC387, Enserch MC 441, and Amerada Hess Penn State, all in the Gulf of Mexico at over 600-m water depth. Statoil Gulfaks flowline bundles also used indirect heating (Britania development). An example of the direct heating method is the BP King development, at 1670-m water depth the Gulf of Mexico (PIP flowlines system tieback to Marlim TLP). Drawbacks The drawbacks of hot water heating are primarily the result of two factors: ØThe non-uniformity of heating over the pipeline/bundle length. Unlike electrical heating, which produces the same heat output and temperature along the whole length, the temperature of the hot water reduces while traveling due to heat loss to the ambient and to the bundle (the latter occurs before an equilibrium is reached). The inlet temperature of the hot water is higher than the designed minimum flowline temperature, which occurs at the end of the flowline. For very long tieback lengths, the required inlet temperature may be too high (close to 100 o C or even higher) unless water flowrate is increased. However the increase of flowrate will increase the pump discharge pressure requirement, which in turn leads to higher power and equipment cost, as well as thicker walls of the hot water pipe. ØLocal heating efficiency issues due to heat loss to the environment. In indirect heating, there is only a small contact between the water pipe and the flowline. A significant amount of heat that the hot water produces at a section along the flowline may be lost to the ambient. Direct heating eliminates the contact factor, but includes a large heat transfer area to the ambient. Direct heating systems also needs many spacers in the annulus (one about every 6 m) which may significantly increase the pump discharge pressure. In the three electrical heating method, local heating efficiency is expected to be higher than the hot water heating. SECT heating has similar problems as indirect hot water heating but the efficiency may be higher because the conductor tube may be smaller than the hot water pipe. Both of these drawbacks can be effectively addressed by increasing the insulation of the system, including around the hot water pipe. However, this increases the total cost through more materials and potentially larger bundles. Heated Risers A recent development for heating deepwater risers is the Integrated Production Bundle (IPB) from CSO. IPBs are heated flexible risers. The concept uses a series of small heating tubes around the flowline. Instead of heating up the produced fluids, the most important function of the heating tubes is to create a thermal shield between the flowline and the ambient, and thus minimises the heat loss by

TotalFinalElf E&P

DGEP/SCR/RD/UGF



the fluids. This philosophy is based on the observation that, due to the short resident time of produced fluids in the riser during normal production, heating in the riser to raise the fluid temperature is usually not effective. Thermal Tests Thermal test of actively heated flowlines should cover a much wider scope than passive insulation, because of the added system component and function. First, thermal performance without heating needs to be determined. The issues and methods are the similar to passively insulated flowlines. The steady state test methods are directly applicable. But cooldown tests with lumped thermal mass data reduction method cannot be directly applied because of the presence of heating pipe and the heating fluid, both having substantial thermal mass. To use the cooldown tests, a new data reduction methodology must be developed. Second, heating performance needs to be determined. The heat transfer rates between the heating fluid and the produced fluid, at various heating flowrate and heated fluid properties (and velocities if continuous heating), are a key factor determining the heating performance. However this is one of most uncertain areas in prediction and design. Thermal tests need to be carefully performed so that the system performance can be assured for the expected ranges of operating conditions. These tests should also provide data for improving prediction tools for future designs. Another aspect of the heating performance is the heating efficiency, i.e., how much of the heat lost from the heating fluid is used to heat up the fluid and how much is used to heat up the insulation or lost to the ambient. Hot water heated flowline systems frequently have spacers or other components that can degrade the insulation or heating performance. These should be included in the analysis and determination by thermal tests.

7.4.2.2 Skin Effect Current Tracing (SECT) SECT heating utilises a ferromagnetic heating tube and a heat resistant electrical heat conductor cable inside it. In this section we briefly describe the SECT heating. With SECT heating, alternating current (AC) electrical power is transmitted to the far end through the cable, which is connected to the heating tube at the far end. The tube completes the electrical circuit and is the primary heating element. Because of the skin effect phenomenon and the attendant proximity effect, the current flow in the heating tube is concentrated in a layer of a couple of millimetres thick at the inner surface. There is nearly no voltage on the outer surface. Key facts in SECT heating are ØThe system can be small and economical ØPower factor is about 90% ØNormally one tube is needed for up to 14” pipe size ØThe voltage range is 0.3-0.7 Volt/m ØSkin thickness about 1 mm ØNo track record in offshore installations, although some designs have been done and some companies such as Thermon are working on issues related to offshore applications of SECT heating. ØMaintenance/risk includes break of the insulated wire, which can mean costly remediation for offshore installations SECT heating will need several kilovolts or higher for flowline length over a few miles unless multiple circuits are used. Multiple circuits need subsea electrical connections, the reliability of which for a typical service life requirement of over 20 years has not been proven. Thermal Tests Same as for hot water heated flowlines, thermal performance without heating needs to be determined. The steady state test methods for passive insulation systems are directly applicable. Since SECT heating tubes are generally small and welded onto the flowline, they are at about the same temperature as the flowline in non-heated situations. Cooldown tests with lumped thermal mass data reduction method thus can be directly applied, as long as the lumped thermal mass includes that for the heating tubes and cables. Temperature of the tubes should be measured and compared with that

TotalFinalElf E&P

DGEP/SCR/RD/UGF



of the flowline. If large differences exist, the lumped thermal mass data reduction method cannot be used. Heating performances also need to be determined. SECT heating tubes are welded on to the flowline and thus the heat resistance is expected to be small. However the relationship between the heating output and the heat gained by the produced fluid needs to be determined for a range of heating rate, flow velocities, and fluid properties. This determines the heating power requirement, fluid warm-up time, as well as heating efficiency of the system. Due to the localized nature of the heating tubes, the temperature at the tubes can be quite high. The thermal test should determine the temperature during various operating conditions. This helps determine whether the insulation material has sufficiently high temperature rating, and whether potential problems exist in accelerated aging of the insulation materials due to the higher temperature.

7.4.2.3 Low power Electrical Heat Traced Methods In recent years, CSO and ITP Interpipe have in parallel developed low power electrical heating methods. These methods utilise multiple heat traced cables around the flowline pipe, inside high insulation levels to reduce heat loss to the environment and thus reduces the total power requirement (see results here after § Thermal tests) Both methods have been development in the application of pipe-in-pipe insulation systems, but can be readily adapted to bundle insulation. In the CSO method, several groups (2-4 depending on the flowline size and heating requirement) of three to four cables each are placed around the flowline. The cables in a group are connected using a three star connection system and form a complete circuit. The ITP method uses evenly distributed insulated electrical wires (heating elements) around the flowline. Electrical power is supplied through these wires at one end. The power distribution system can be open loop or closed loop. In the open loop system, the electrical wires are grounded in the seawater at the other end of the pipeline. In this type of system, corrosion of the external pipe and nearby structures needs to be considered. In the closed-loop system, the current is brought back to the surface facility through (possibly) a separate electrical cable. The differences of the two methods are in electrical power supply and connection methods and the related issues. In bundle fabrication and installation they do not have significant differences with each other. Thermal Tests The issues in thermal testing of these actively heated systems are the same as SECT heated systems. ITP Electrical heat traced flowline PIP 6’’ x 9’’ was tested (JIP GPRI Texaco Humble year 2001). The results shows that the fluid temperature can be maintained above 25°C with a power supply of 7 watts/m with an ambient temperature of 4°C. If the fluids reach sea floor temperature (i.e. 4°C) the EHTF can raise oil and water temperatures above hydrate region (25°C) within 24 hours by applying 20 watts/m. In the PIP annulus a Fibre Optic was inserted by pumping with water through a 1/8’’ tubing. The tests demonstrated that the FO can be used as an approximation of fluid temperature all along the flowline (see § 9.4.3).

7.4.2.4 Direct Electrical Heating (DEH) Electrical heating has been studied extensively by SHELL and STATOIL in the last several years. DEH methods use the natural electrical impedance of the flowline pipe instead of separate heating elements to heat the flowline. This has the advantages of a simpler system than other electrical heating method. The cost can be lower with the elimination of the heating cables/wires. The risk of broken wires, for the flowline part of the electrical loop, is eliminated. DEH systems can be open loop or closed loop. In an open loop system, electricity is supplied to the end of the flowline through subsea cables while electric current flows through both the flowline and

TotalFinalElf E&P

DGEP/SCR/RD/UGF



seawater (in parallel). Only AC current can be used because DC current in seawater will cause corrosion on all nearly structures. In a closed loop system uses a separate cable or the steel jacket pipe (if present) as the return conductor. Both AC and DC currents can be used in closed systems. While the DEH concept has a much longer history in the industry, only recently have field tests and installations been done: Ø

DeepStar IV engineering study and full scale field demonstration. This tests driven by Shell (Ron Bass) were performed in Norway and operated by Alcatel Norge with the support of Thermotite (electrical & thermal insulation). Reports are available.

Ø

Statoil Sleipner (North Sea WD: ? – installation 1996) used induction system on a single pipe 20’’ 12.6 Km long for temperature maintenance.

Ø

Statoil Asgard (North Sea WD: 400m – installation 2000) used DEH for the first time. Open loop systems were used on a total of six flowlines 8’’. The Heating cable lengths were 6 & 8.5 km long. The maximum system voltage required was 5.3 KV and the maximum system current was 1520 A, which occurred in the case of heating from complete cooldown state. The system was designed for hydrate prevention, and was only to be used during shutdown or warm-up (maintenance at 27°C). 70 – 80 watts/m of heat is lost to the surroundings. Heating requires a supply of about 105 W/m.

Ø

Statoil Huldra (North Sea WD: 400m – installation 2001) development. Open loop system was to be used on the 16 km x 8’’ condensate pipeline. The system DEH will be used similarly as Statoil Asgard except for it also will be used for wax prevention and removal, and should be used continuously during the last four years of field life.

Ø

Shell Serrano/Oregano (GOM WD: 1000m – installation 2001) subsea developments (Gulf of Mexico). The development used 6” in 10” pipe-in-pipe insulation and closed system DEH Mid line active heating to tieback the two fields to Auger TLP, with distances about 6 miles (10 km) and 7.5 miles (12 km) respectively. At the platform end, the power supply is connected to the inner pipe in an electric isolation joint. At the steel bulkhead at the PLEM, the two pipes are connected so the AC electrical current can flow back to the platform through the outer pipe. Due to skin effect, the current is concentrated within a thin layer at the OD of the inner pipe, and a thin layer at the ID of the outer pipe. Through tests Shell proved that electrical leakage to seawater is negligible due to this phenomenon.

Ø

As mentioned in HYDRATES theme (§ 2.4.4) Shell plan to use (installation in 2003) Direct Electrical Heating system for hydrate remediation (back-up solution) on Nakika field (WD: 1900m) on the different flowlines (10’’ x 16’’ PIP length 2 to 12.8 Km).

The most challenging issue in design was the possible effects on corrosion and signal interference from impressing high-voltage AC over DC cathodic protection. However details of the solution are not available. Short circuit is a major risk aspect in the system. In installation, water egress in the pipeline must be strictly avoided and every joint of pipe was tested before the next was installed. In the current configuration, the DEH method is limited to 15 miles due to annulus material breakdown issues. Increasing to 30 miles can be done but involves subsea isolation joint and electrical connection. Due to high electrical impedance of flowlines, DEH methods have a common shortcoming of high voltage, possibly close to 0.5-1 V/m. Thermal Tests For the DEH systems, the tests for thermal performances of the insulation while no heating is applied are the same as those for passive insulation system. There are no additional cable or heating pipe. The steady state and cooldown test methods can be used in the same way. As with other actively heated systems, heating performances should be determined, in all of the designed operation modes such as warming-up before production begins,

TotalFinalElf E&P

DGEP/SCR/RD/UGF



temperature keeping during production shutdown, or continuous heating during normal production. Direct heating is expected to be efficient since the heating surface is in contact with the fluid. The heating performances to determine are the power requirement in the designed operations at the design criteria. The expected range of fluid properties should be covered. The temperature of the flowline is expected to be close to that of the fluid during heating, but measurement should be made to confirm.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 110 of 128


7.4.3 Temperature monitoring Distributed Temperature System with Fibre Optic technology Temperature monitoring can lessen the risks related to hydrate or wax plugging by a better monitoring of thermal behaviour of the flowlines and risers. Anticipating the problem can allow the operator to take appropriate prevention or preservation action and limit the number of interventions required. DTS developed by SENSA measures temperature along the whole lenght of an optical fibre in real time. The FO is installed in a protective housing, typically a control line attached to the flowline. The fibre optic is installed afterward by fluid drag ( by pumping) taking the FO cable from surface through a turn around spool and then back to surface.

Principle of FO temperature measurement: A laser light pulse is sent down a multimode fibre optic wave-guide. Molecular vibration which is directly related to temperature creates reflected signals (amplitude gives the temperature). As the laser light pulse travels along the wave guide temperature-induced molecular vibrations cause a very weak reflected signal to travel back up the fibre to the source and is analysed by an optoelectronic unit. The time differential determines the location of the temperature measurement as the speed of light in the FO is constant and known.

Directional Coupler

Raman Scattering

Laser Source

Brillouin

Brillouin

Anti-Stokes Signal Raman Band Processing

Stokes Raman Band Wavelength

SENSA is the leader company for DTS using Raman spectroscopy (Reflection Raleigh + Stokes & anti-Stokes). Temperature limitation is around 150°C, the maximum length is: Ø

< 10 kilometres with temperature (+/- 0.5°C) measurement every metre using multimode fibre. Ø 10 to 12 kilometres with temperature (+/- 0.5°C) measurement every 2 metres using multi-mode fibre. Ø < 30 kilometres with temperature (+/- 1°C) measurement every 10 metres to 30 metres using mono-mode fibre. There is a limitation with the number of measurements (< 10 000 points) per fibre, and the sampling time is around 15 minutes. To date SENSA claims 130 DTS application. Among them we can mention: Gulfaks (2 DTS deployed on 15 Km by Statoil), Wytch Farm (DTS on 11 Km by BP), Tern & Bellridge (Shell), Coalinga (well temperature monitoring on steam flooding Chevron California), Marlim (monitoring of well temperature following collapse problem with production casing on a wellhead platform, BP)

TotalFinalElf E&P

DGEP/SCR/RD/UGF



7.4.4 Hydrate prevention For the hydrate prevention strategy under shutdown conditions, two solutions have to be considered: Ø

Passive insulation system allowing enough time to apply preservation procedures. During this period, the produced fluids in the subsea production system are replaced by pumping storage oil

Ø

Active insulation system by heating the subsea lines to compensate for thermal losses during shutdown in order to maintain line temperature outside the hydrate region, or re-heat fluids prior to production restart (see Hydrate Theme 2).

Other areas: ü

Valve location can be a problem due to water or gas accumulation, in order not to foster hydrate blockage during closing / opening phases.

ü

For a non-heated system (or one in which only the riser will be heated), the lines must allow access by coiled tubing at any point in the subsea system. This will be useful for removal in the event of hydrate blockage. The CTU (coiled tubing unit) located on the surface utilities allows circulation of inhibitors or/and storage oil necessary to restore fluid circulation,

ü

All “weak” points in terms of thermal insulation, mainly connectors, must be placed at high points in the system where gas pockets may form after a shutdown. This will allow water to drain down to a warmer location,

ü

Instrumentation points, such as temperature and pressure transmitters, must be non intrusive, and placed on the upper third of the pipe, as near as practical to the 12 o’clock position,

ü

Dead legs namely: piping of the pigging module, piping downstream of the dispatching valves (subheaders), piping between header isolation valves and sub-headers must be minimised. Their shapes must also be such as to avoid water accumulation,

ü

Gas-lift line geometry, if non-heated, must allow water drainage within the production line. As liquid may enter these lines on shutdowns, their shape must allow the water to drain out of them by gravity so as to minimise the hydrate formation risk.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Erosion / Corrosion

TotalFinalElf E&P

DGEP/SCR/RD/UGF



8. Erosion / Corrosion

8.1 Terminology The term "erosion-corrosion" word is a very general one, incorporating a very wide range of phenomena, from erosion-induced degradation (i.e. with almost no corrosion effect) to corrosion degradation aggravated by increased ionic diffusion (i.e. without any mechanical effect). For reasons of clarity, 2 basic considerations and 4 distinct terms will be used: Basic considerations: 1. The term "Erosion" will always refer to mechanical effects, produced by either liquid or solids, and either on the metal itself or on protective layers formed on the metal, 2. The term "Corrosion" will always be used when the final degradation is brought about by the electrochemical reactions involved in corrosion phenomena. Terms used: 1. "Erosion" (Er): Degradation resulting exclusively from a mechanical effect of the fluid on the metal itself, without requiring any corrosion reaction (e.g. sand erosion in dry gas or in oil containing no water). Both "solid erosion" and "liquid erosion" could be used, but in practice liquid erosion phenomena (without any solid or any corrosive liquid) are not of any real concern in oil and gas production applications. 2. "Solid erosion-corrosion" (S E-C): Final degradation results from corrosion reactions, but solid particles contribute to the mechanical removal of a protective layer such as a passive film or an oxide - carbonate –sulphide… protective layer. 3. "Liquid erosion-corrosion" (L E-C): Final degradation results from corrosion reactions, but liquid flow contributes to the mechanical removal of a protective layer (by fatigue cracking, shear stress effect…). No solid particles involved. 4. "Flow-enhanced corrosion" (F-E C): Flow effect is exclusively chemical and electrochemical, without any mechanical effect (flow only enhancing diffusion of species which favour electrochemical reactions and/or dissolution of the protective layer). No solid particles involved. The overall term of "erosion-corrosion" is thus only used as a general one. It currently incorporate only the 3 first phenomena (no mechanical effect of flow velocity in F-E C). However, in practice, it is difficult to draw any clear distinction between F-E C and S E-C (corrosion layer removal frequently resulting from a combination of mechanical and chemical effects which can hardly be distinguished). As a consequence, although F-E C is no longer mentioned, it is currently covered by limits used for L E-C.

8.2 Position of the problem Because of high flow velocities and 3-phase transportation, all the erosion-corrosion phenomena defined above are potentially enhanced in deepwater applications. In addition, most of the reservoirs involved are unconsolidated sandstone ones, whence a significant risk of solid particle transportation from the reservoir.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 114 of 128

Part 2 Multiphase Transport 8.2.1 Main materials concerned

Material

Possible erosion-corrosion phenomena

Key factors

Carbon steel

All

Flow velocity, corrosive behaviour of produced fluid, sand particles…

Stainless steel and nickel alloys

Er and S E-C

Sand particles (amount and size), flow velocity

Hard materials (stellite, tungsten carbide, diamond…)

Er

Sand particles (amount and size), flow velocity

8.2.2 Main equipment concerned

Equipment

Usual materials

Possible erosion-corrosion phenomena

Screens

Stainless steels

Erosion

SCSSV and Well-head equipment and flowlines

Carbon steel

Solid and liquid E-C

Stainless steel or Ni alloys

Almost no risk (HP, hence moderate flow velocities)

Chokes (sub-sea and surface)

Hard materials

Erosion

Subsea equipment with irregular geometry (tees, short radius elbows, gas mandrels, connectors…)

Stainless steels

Erosion

Carbon steels

All

Top of risers

Stainless steels

Erosion, solid E-C

Carbon steels

All

Stainless steels

Erosion, solid E-C

Carbon steels

All

Topside flowlines and manifold

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 115 of 128

Part 2 Multiphase Transport 8.2.3 Main fluid effects Fluid characteristic

Erosion-corrosion phenomena involved

Material involved

Presence of water in contact with the metal surface


Carbon steel

Corrosive behaviour (CO2 content, pH, )


Carbon steel

Solid particles

"Erosion" and "solid E-C"

All, although hard, erosionresistant materials resist better

Inhibitor


Carbon steels

Based on these considerations, the main issues of a deepwater development regarding erosioncorrosion management are to: Ø

Assess all possible erosion-corrosion risks throughout the expected lifetime of the installations, and define practical limits for the occurrence of such risks (flow velocity limitations, particle size limitations, allowable sand flow rates…)

Ø

Curb sand particle production as soon as erosion risks are predicted,

Ø

Define an overall material selection/corrosion management policy adapted to the effective erosioncorrosion risks, by the optimum combination of an adequate design (sufficient diameter, avoidance of irregular geometry…), the use of corrosion-resistant alloys, a consistent flow management, the application of corrosion inhibitors…

Ø

Implement monitoring tools, methods, programmes and evaluation criteria aimed at verifying the successful application of the above corrosion management policy.

A complete “Erosion-Corrosion management” evaluation is strongly recommended for all deepsea applications.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 116 of 128


8.3 Data required Requisite and useful data for erosion-corrosion evaluations are summarised in the following tables:

Type

Evaluation concerned

Usefulness

Expected lifetime

All

N

BHT and BHP

L&S EC

U

Main composition (sandstone, carbonates…)

Er.

Accuracy/ other comments

Reservoir P: ± 20%, T ± 5 °C

U

Production data Flow rate of each phase on all evaluation points (or equivalent data such as Oil flow rate + GOR + BSW)

All

R

Order of magnitude (+/- 20%)

Nominal pressures and temperatures

All

R

P: ± 20%, T ± 5 °C

Predicted solid production. Permanent or accidental?

S EC, Er

R

Evolution of all above data with time (prod profiles, water evolution, T and P variation)

All

U

Internal diameters

All

R

Specific profiles

S E-C and Er

U

L & S E-C

R

L & S E-C

R

Geometrical data:

Produced water: Detailed water analysis Produced gas : 2

CO and H2S content Complete analysis

U

Preferably with organic species

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 117 of 128

Part 2 Multiphase Transport Crude: Bubble pressure

L & S EC

U

Viscosity

Er

R

Main chemical characteristics

Er

U

Type (clays, sand…)

S EC, Er

U

Average diameter

S EC, Er

R

Granulometric profile

S EC, Er

U

Solid particles:

Very useful

* : Usefulness: R = Required (for risk assessment), U = Useful, but not mandatory. The evaluation may be less precise if the data is not available, but without any critical impact in most cases.

8.4 Methodologies

8.4.1 Risk assessment

8.4.1.1 Liquid Erosion-corrosion L E-C risks on carbon Steel are evaluated according to the 3 following steps : 1. Presence of water in contact with the metal: Although it is very common to find water present somewhere on the metal surface in oil and gas installations (particularly at high flow velocities where centrifugal effects enhance water separation), there may be cases for which it is not certain this contact is possible (dry gas, stable emulsions…) and in which, L E-C risks are consequently nil. 2. Occurrence of L E-C: When the actual flow velocity is higher than a “limiting flow velocity”. 3. Corrosion rate of L E-C: When water is in contact with the metal and the limiting flow velocity is exceeded, the L E-C rate is calculated as the “Potential corrosivity” of the produced water. “Potential corrosivity” is the uniform corrosion rate of bare carbon steel, when there is no protection by any corrosion layer and when no increased corrosion by localised corrosion occurs (as is typically the case for L and S E-C). The Liquid Erosion-corrosion rate therefore necessarily requires a corrosive fluid, so it may remain negligible despite very high flow rates as long as the fluid is not significantly corrosive (dry gas, dry oil…)

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 118 of 128

Part 2 Multiphase Transport 8.4.1.1.1 Water presence in contact with the metal

A difficult and very specific matter, for which there are not as yet any general criteria available. To be analysed on a case-by-case basis, if this lack of water is considered likely to last any significant time.

8.4.1.1.2 Limiting flow velocities E

The most common equation for calculating flow velocities is the API RP 14 equation: Vcr = C / √ρ, Where C is a constant and ρ is the average density of the fluid considered. This equation is currently used by process people in terms of ρ V limiting values (e.g. a C value < 100 2 in US units is almost equivalent to a ρ V < 15000 in SI units) 2

Although this simple equation is not fully acknowledged to cover the wide variety of effects and parameters involved in critical velocity determination, it is recognised as a quite valid design and engineering criterion, as long as constant values (C parameters) are adjusted to the type of media and service conditions. As an example, C values in the range of 100 to 250-300 are indicated in the literature for different types of fluids (gas, oil, water, mixture of these 3 phases…), different materials, presence of corrosion inhibitors… An internal procedure indicating C values (or limiting velocities in m/s) for different production cases is included in the in-house document Technical Corrosion Practices. These values are to be used for design and engineering purposes. If the case under evaluation is not included in this procedure, or if higher values than the reference limiting ones are calculated, corrosion specialists must be contacted to conduct an in-depth evaluation (including not only flow velocity but also flow regime, duration of high flow conditions, combination of favourable and adverse parameters…). The in-depth evaluation is then based on more detailed criteria included in the in house LIPUCOR calculation tool.

8.4.1.1.3 Liquid Erosion-corrosion rate The “Potential Corrosivity” parameter, which gives the worst L E-C corrosion rate, is an electrochemical parameter, which indicates the corrosive behaviour of the water on bare steel. It is calculated with the in house CORMED calculation programme.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



8.4.1.2 Solid Erosion-Corrosion Solid Erosion-Corrosion risks are typically based on similar criteria to those used for L E-C (presence of water, flow velocities exceeding limiting values and sufficient corrosion rate). Both the presence of water and the corrosion rate evaluation are evaluated with basically the same criteria as for L E-C. See paragraphs above. However, limiting velocity criteria are not as clearly defined as for L E-C, because of: •

the increased difficulty (more influencing parameters, in particular solid flow rate, particle size and shape, fluid composition…),

•

less field experience (solid particle production being almost systematically prevented, it is only experienced following accidental failure of sand prevention methods)

•

The practical difficulty of distinguishing the pure Erosion effect (direct metal removal) from the Solid Erosion-Corrosion effect on failed equipment.

In practice, more conservative C values than for L E-C are used. These values are selected on a case-by-case basis, a particular attention being paid to the amount and size of solid particles expected. More precise predictive criteria (and even models) are expected within a few years. They are not yet at a sufficient level of development, validity and qualification (with respect to field experience) to be used as design and engineering tools.

8.4.1.3 Erosion The risk of erosion damage is of particular concern for deepwater developments since most of the reservoirs are unconsolidated ones, i.e. massive sand production may occur if no bottomhole prevention is applied, or if it fails. In addition, high flow velocities are expected and flow limitation, because of sand presence, is of course a costly method when significant production loss is involved. As opposed to L and S E-C, erosion damage no longer involves any electrochemical corrosion phenomenon. Consequently, neither water presence nor corrosive behaviour is necessary for erosion damage, although combined or even synergic effects between erosion and corrosion damage may occur in practice. Erosion prediction models are based on a simple Erosion Rate prediction equation as follows: n

2

ER = k * M * V / d where: • M is the mass flow rate of sand production (in g/s or kg/d...) • V is the mixed flow velocity of the fluid, i.e. the sum of liquid and gas superficial flow velocities in 2 or 3 Φ flow (m/s) • d is the pipe ID (mm or m or ") • k is a constant. • n value is given in the range of 1.7 to 2.5, depending on the models used.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



The M parameter may also be replaced by the mass sand concentration parameter (usually in ppm). n+1 When the latter is used, the ER is then proportional to V , i.e. to an exponent of about 3 on the velocity (M being proportional to both the sand concentration and the fluid flow rate, hence the fluid velocity). st

There are various models available based on this 1 equation, and which have been improved and adapted to different types of fluids (k constant being dependent on a large number of parameters). In particular, initial models were dedicated to gas fluids (with almost no liquid), and then to liquid fluids, whereas present models combine the 2 previous approaches for application to 3-phase flowing conditions (Oil + gas + water). The most adequate models for preliminary evaluation are: 1. Salama-Venkatesh 1983, for gas applications, with almost no liquid (typically gas-condensate fluids), 2. Salama 1998 (OTC papers nos. 8898 and 8900) for liquid and 3-phase applications. More complex models (Tulsa, DnV, AEA…) exist but their limits of application are not yet fully known (to be evaluated in 2001-2002). They should therefore only be used on a case-by-case basis. These models are designed for calculating Erosion Rates ER (in mm/yr) versus sand production, flow velocity and other contributing factors (the more complex the model, the higher the number of parameters included). As a consequence, limiting flow velocities can be evaluated for a specified max. Erosion rate. The following qualitative considerations must also be taken into account: •

Even though vendors claim their prediction models are accurate, the effective accuracy of the predicted values must still be considered as quite low. In particular, whatever the accuracy of the model itself, the data used for calculation are generally not very accurately known. In particular, how can the simultaneous evolution over time of pressure, BSW, flow rate, sand production rate and particle size distribution be accurately predicted? It is thus recommended to use these results mainly in 3 wide categories, e.g. low risk when ER< 0.1 or 0.05 mm/yr, high risk when ER> 1 or 2 mm/yr, and medium risk between these 2 extreme cases.

•

As long as bottomhole sand prevention is implemented (with screens and gravel-packs), two distinct cases of sand production must be separately considered: normal residual sand production through gravel-pack and screens (with a very low amount of sand produced and a low particle size of 10 to 30 microns) and accidental massive sand production, up to the max. particle size of the reservoir, in case of accidental or progressive gravel-pack/screen damage.

•

For oil production, i.e. with a significant amount of liquid produced, erosion risk is almost negligible under the usual flow velocities (15 to 30 m/s) as long as particle size is below 50 microns. The normal residual sand production through undamaged screens consequently has no significant impact on erosion risks.

•

Erosion damage, hence erosion prediction models, first concern particular geometries where sand particles preferentially impact the pipe wall. As a consequence, chokes, short radius elbows, Tees, restrictions… are particularly concerned, whereas erosion along straight lines (either horizontal or vertical) is very rare.

•

The 2 more critical areas where erosion risks merit consideration are usually: Ø

The subsea flow-line, downstream of the choke, because of the sudden pressure loss at the choke level, the relatively low diameter and the irregular geometry of the flowline,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

FLOW ASSURANCE GUIDE Part 2 Multiphase Transport Ø •


The top of riser, because of the lower pressure level and higher gas content than at higher pressure.

At present, there is no adequate predictive erosion model for chokes, i.e. for the most severe erosion conditions of all production equipment (very high flow velocities, highly irregular geometrical profile, specific materials).

8.4.1.4 Main deliverables The main deliverables of an erosion-corrosion risk assessment study are: 1. The effective predicted risks for all erosion-corrosion phenomena, among the range of production conditions anticipated for the complete production period. It should be emphasised that all adverse parameters must not be simultaneously set at their most critical level, since it is exceptional to find the highest flow rates combined with the lowest pressures, the highest BSW… 2. Limiting C values or velocity values to be used for further design/ engineering studies, 3. The most sensitive parameters which might modify the prediction already performed, 4. In addition to these deliverables, the graph below provides an example of velocity limitations given for the Girassol case in terms of acceptable GLR versus liquid flow rates, with different erosion and erosion corrosion criteria.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 122 of 128


Production limitations for SPS equipment Figure 2 : Accidental massive sand production (100 microns, 500 kg/d) Capi = 130 (80 bars)

Capi= 130 (110 bars)

Capi= 180 (80 bars)

Capi= 180 (110 bars)

Medium erosion risk (80 bars)

Severe erosion risk (80 bars)

GLR (Sm3/ m3)

1000

100

10 0

5000

10000

15000

20000

25000

30000

Liqu. Flow Rate (bpd)

35000

40000

45000

50000

8.4.2 Erosion-corrosion prevention The various possible erosion-corrosion prevention methods applicable are summarised in the following table: Action on:

Prevention method

E-C phenomena concerned

Comments

Material selection

Use of stainless steel

L and S E-C

Very efficient as long as no Er risk.

Use of “hard materials” (stellite, tungsten carbide…)

All, but mainly used for Er prevention

Mainly for choke and specific injection devices

Well design

Bottom-hole sand prevention by screens, gravel-packs

S E-C and Er

The key factor for E-C prevention when sand production is expected.

Flowline/ Piping design

Diameter sufficient to avoid exceeding velocity limitations

All, with different limiting factors

General design of

Minimizing geometrical

All

Avoid impact areas, maximise curvature radius of elbows,

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 123 of 128

Part 2 Multiphase Transport production circuits

anomalies/irregularities

uniformity of internal diameters…

Produced fluids

Corrosion inhibitor injection

L and S E-C

Also affected by flow velocity limitations, but higher that without inhibitor

Production management

Careful start-up of wells

S E-C and Er

To minimise the risk of screen and gravel-pack damage.


Preferential choking at the top of riser, rather than at subsea wellhead level

All

Keep the subsea circuits at the highest pressure level compatible with production requirements, in order to minimise flow velocities.


Flow rate optimisation between production pipelines

All

In order to remain below specified limiting velocities, hence corresponding flow rates

Almost all the methods mentioned above must usually be implemented within a complete project on the various components of the production circuits from bottomhole to the top of the risers.

8.4.3 Erosion-corrosion monitoring

8.4.3.1 Liquid Erosion-corrosion monitoring Much the same methods as for usual corrosion monitoring are used for L E-R (coupons, ER probes, FSM, Wall thickness measurements…) since both corrosion failure modes and corrosion prevention methods are similar (use of inhibitors, flow velocity management…) The main specificity of L E-R monitoring resides in the location of the monitoring tools, which must be adjusted to the worst locations where this corrosion is likely to appear if not properly prevented. The most likely spots are on the outlet side of elbows, and in impact areas. The monitoring locations must be precisely defined on PFDs during the basic Engineering stage of projects, and then checked on detailed isometric drawings at the Detailed Engineering phase.

8.4.3.2 Solid E-R and Erosion monitoring As long as solids are present in produced fluids, monitoring must include 2 distinct objectives: 1. Early detection of solid particles being produced from the producing wells, 2. Detection of possible damage to producing equipment due to S E-R or Er. There are dedicated tools and methods for each of these 2 objectives. Table x provides some indications on sand detection methods, while table y is dedicated to damage detection (not to be considered as an exhaustive review)

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 124 of 128

Part 2 Multiphase Transport st

As for L E-R, the 1 key point is the location of the tools, which must be placed in the most critical areas. These need to be precisely defined and checked later during the Engineering phases of the Project, when the geometry of the equipment is known. In addition, as none of these monitoring solutions is fully reliable or fully indicative of all possible erosion effects, a combination of several methods is necessary.

Table x: Basic methods to be used for sand detection Monitoring method

Methodology

Main advantages

Main drawbacks and limitations

Acoustic sand detection

Detection of the impact of solid particles on the pipe wall

Non intrusive measurement

To be located on an appropriate area, where particle impacts are effective.

Both topside equipment and subsea/ deep sea equipment available

Detection capability limited to impacting particles

1st important location: subsea wellhead outlet, i.e. in deepsea conditions. Precise measurement requires calibration with injection of a limited volume of sand. Not an easy challenge in a deepwater environment.

Solid content measurement on liquid samples

Filtration, drying and weight measurement

Choke performance monitoring

Periodic control of choke response to flow and ∆P (the choke usually being the 1st damaged component in the event of massive sand production)

Easiness

Only possible on topsides, i.e. not dedicated to one well Discontinuous measurement è No immediate alert

No specific equipment required (only data management)

Limited sensitivity to sand production (the choke being designed for sand resistance)

Table y: Basic methods to be used for erosion monitoring Monitoring method

Methodology

Main advantages

Main drawbacks and limitations

Stainless steel ER probe

Wall thickness measurement of a thin sensitive element. No sensitivity to corrosion, since the material used is SS

Quick standard measurement (used for corrosion monitoring, with CS probes).

Intrusive probe.

Field Signature Monitoring

Periodic wall thickness measurement over a portion of pipe by Electrical measurement from the outside (non intrusive system)

Available in both topsides and deepsea versions

Expensive, particularly in deepsea applications

Monitoring of a significant portion of pipe (0.2 to 1 m) of any size and shape (straight

Poor sensitivity to localised degradation

Thin wall for measurement, hence quick degradation in case of erosion è inadequate for subsea installations.

TotalFinalElf E&P

DGEP/SCR/RD/UGF

Ref. : UGF/2001-135 Rev. : 0 – 01 / 2002


Page 125 of 128

Part 2 Multiphase Transport section, elbow, weld area …)

Limited reliability up to now

Measurement on the whole pipe thickness Non-intrusive system

Instrumented ultrasonic wall thickness measurement

Instrumented periodic US monitoring probes, with several probes for covering a significant area

Applicable on deep sea equipment

Unknown reliability of existing tools

Well known, precise measuring technique

Limited surface covered, hence the risk of missing the most critical locations

Non-intrusive

Sacrificial piping element

Sacrificial element, less thick than the piping itself, included in a pressure-containing envelope. Erosion detected by pressure detection in the envelope.

Very limited instrumentation (pressure control)

No early detection, since a large thickness has already been consumed by the time the inner sacrificial element leaks.

TotalFinalElf E&P

DGEP/SCR/RD/UGF




Well Metering

TotalFinalElf E&P

DGEP/SCR/RD/UGF



9. WELL METERING

9.1 THE ISSUES In order to maximise the Production in accordance with the Reservoir management strategy, all produced fluids have to be monitored. The main objective of the Well Monitoring system is to satisfy the following requirements: •

to provide, at least once a month, an estimate of each of the well flow-rates (oil, gas and water) with an accuracy in line with the reservoir management requirements,

•

to detect “well events” such as water or gas breakthrough as early as possible,

•

to provide continuous well flow-rate estimates as input to the Control System (OCWR for example).

The financial requirements and technical constraints require development schemes in which several wells produce through a single flowline. Obviously, these same reasons prevent the use of a dedicated production metering flowline. Consequently, any metering method envisaged is more complex and costly than usual (e.g. production loss, high methanol consumption, etc.) and the risk of failure or incidents is increased. From a technical aspect, several potential solutions can be implemented using: ÄA test separator, ÄMultiphase meters, ÄFlowrate calculation tools based on flow modelling, ÄA combination of these techniques.

9.2 SOLUTIONS 1. Direct testing through a surface test separator. In this case, testing the well requires either shutting-in the other wells connected to the same flowline or switching them to another available flowline, if field architecture permits (loop layout). The main disadvantages include significant production losses, huge consumption of methanol to inhibit the resulting “dead branches”, gas-lift to stabilise flow or a number of subsea valve operations. The main advantage of this approach resides in its precision, which is the same as the separator metering methods usually available . Moreover, this method is not affected by failures in subsea gauges. An alternative to direct well testing through the separator is to test “n” linear combinations of “n” wells. This methodology is called “permutation”. In practice, it is very difficult or even impossible to apply when the number of wells per branch exceeds 3.

TotalFinalElf E&P

DGEP/SCR/RD/UGF



2. Subsea Multiphase metering There are not many subsea applications around. Successful cases (Fluenta at -500m on Albacora (Petrobas), Framo at -100m on West Brae (Marathon Oil)) alternate with setbacks (Aasgard (Gamma densitometers deteriorated by He), ETAP (2 interventions on a gamma meter)). Even though multiphase meters remain good candidates for subsea utilisation, there is still room for improvement in reliability or operability if they are to be used beyond -1500 m. When properly designed, installed and operated, their advantages reside in their ease of use, their metering capabilities on many types of fluids, and their ability to read flowrates in real time. Their disadvantages are: the radioactive source, their unsuitability to GLRs exceeding 85%, their lack of auto-adaptability when significant variations occur in fluid properties.

3. Metering by flow modelling. Using hydrodynamic, thermal and thermodynamic laws, pressure and temperature measurements are taken along the entire fluid vein (bottomhole, upstream and downstream choke, manifold, riser…) and converted into flowrates. A mathematical solver renders the flowrates calculated consistent with each other by taking into account uncertainties ascribed to measurements, in order to obtain continuous values (approximately every minute) for water, oil and gas flowrates in each well. If implemented successfully, this technique can provide flowrate values with an uncertainty of less than 10%, without inducing any production loss, with no methanol consumption and more generally, without affecting the flow in any way. There are, however, difficulties in its adjustment and implementation, which require regular calibrations against another metering method (separator, MPM…). Three new tools will soon be (or have recently been) put on the market: Idun by LOKE, Wellocate by ABB and another product resulting from joint R&D by TOTALFINAELF and RSI, which will be available in 2002 and should offer other advantages over rival tools, such as greater robustness with regard to gauge defects or failure.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Flow Assurance Guide

Recommend Documents