eni - completion design manual

ARPO

ENI S.p.A.Agip Division

ORGANISINGDEPARTMENT

TYPE OFACTIVITY'

ISSUINGDEPT.

DOC.TYPE

REFER TOSECTION N.

PAGE. 1

OF 295STAP P 1 M 7100

The present document is CONFIDENTIAL and it is property of AGIP It shall not be shown to third parties nor shall it be used forreasons different from those owing to which it was given

TITLECOMPLETION DESIGN MANUAL

DISTRIBUTION LIST

Eni - Agip Division Italian DistrictsEni - Agip Division Affiliated CompaniesEni - Agip Division Headquarter Drilling & Completion UnitsSTAP ArchiveEni - Agip Division Headquarter Subsurface Geology UnitsEni - Agip Division Headquarter Reservoir UnitsEni - Agip Division Headquarter Coordination Units for Italian ActivitiesEni - Agip Division Headquarter Coordination Units for Foreign Activities

NOTE: The present document is available in Eni Agip Intranet (http://wwwarpo.in.agip.it) and aCD-Rom version can also be distributed (requests will be addressed to STAP Dept. inEni - Agip Division Headquarter)

Date of issue:

f

e

d

c

b Issued by M. Bassanini C. Lanzetta A. Galletta28/06/99 28/06/99 28/06/99

REVISIONS PREP'D CHK'D APPR'D

28/06/99

ARPO


IDENTIFICATION CODE PAGE 2 OF 295

REVISIONSTAP-P-1-M-7100 0

INDEX

1. INTRODUCTION 81.1. PURPOSE OF THE MANUAL 81.2. OBJECTIVES 121.3. FUNCTIONS OF A COMPLETION 131.4. MANUAL UPDATING, AMENDMENT, CONTROL & DEROGATION 13

2. RESERVOIR CONSIDERATIONS 142.1. INTRODUCTION 142.2. CHARACTERISTICS OF RESERVOIR ROCKS 14

2.2.1. Porosity 142.2.2. Permeability 142.2.3. Relative Permeability 152.2.4. Wettabilty 162.2.5. Fluid Distribution 172.2.6. Fluid Flow In The Reservoir 182.2.7. Effects Of Reservoir Characteristics 242.2.8. Reservoir Homogeneity 27

2.3. HYDROCARBON DATA 282.3.1. Oil Property Correlation 28

2.4. RESERVOIR/PRODUCTION FORECAST 292.4.1. Inflow Perfomance 312.4.2. Reservoir Simulation For IPR Curves 422.4.3. IPR Selection 442.4.4. Outflow Performance 462.4.5. Flow Rate Prediction 55

3. WELL TESTING 603.1. INTRODUCTION 60

3.1.1. Types of Tests 603.2. DST OBJECTIVE 633.3. DST STRING 643.4. RESERVOIR CHARACTERISTICS 69

3.4.1. Pressure Build-Up Analysis 693.4.2. Basics Of DST Operations 773.4.3. Common Test Tools Description 773.4.4. Tools Utilised With Permanent Packer Systems 803.4.5. Sub-Sea Test Tools Used On Semi-Submersibles 803.4.6. Deep Water Tools 813.4.7. Downhole Pressure Recording 82

ARPO




3.5. WELL PRODUCTION TEST OBJECTIVES 833.5.1. Periodic Tests 833.5.2. Productivity Or Deliverability Tests 843.5.3. Transient Tests 84

4. DRILLING CONSIDERATIONS 874.1. CASING DESIGN 87

4.1.1. Casing Profile 874.1.2. Casing Specifications 884.1.3. Casing Connections 89

4.2. WELL DEVIATION SURVEYS 894.3. CASING CEMENTING CONSIDERATIONS 90

4.3.1. Production Casing Cementing 904.3.2. Production Casing Cement Evaluation 91

5. WELL COMPLETION DESIGN 925.1. FACTORS INFLUENCING COMPLETION DESIGN 94

5.1.1. Reservoir Considerations 945.1.2. Mechanical Considerations 965.1.3. Safety Considerations 96

5.2. RESERVOIR-WELLBORE INTERFACE 975.2.1. Open Hole Completions 975.2.2. Uncemented Liner Completions 985.2.3. Perforated Completions 1005.2.4. Multi-Zone Completions 101

5.3. CASING-TUBING INTERFACE 1045.3.1. Packer Applications 1065.3.2. Packer-Tubing Interfaces 1075.3.3. Annulus Circulation 108

5.4. TUBING-WELLHEAD INTERFACE 1095.4.1. Tubing Hanger Systems 1095.4.2. Xmas Trees 1155.4.3. Metal-To-Metal Seals 115

5.5. FUTURE CONSIDERATIONS 1175.5.1. Stimulation 1185.5.2. Formation Management 1185.5.3. Well Servicing Techniques 119

5.6. OPTIMISING TUBING SIZE 1215.6.1. Reservoir Pressure 1235.6.2. Flowing Wellhead Pressure 1235.6.3. Gas-Liquid Ratio 1235.6.4. Artificial Lift 124

ARPO




6. CORROSION 1266.1. DEVELOPMENT WELLS 1266.2. CONTRIBUTING FACTORS TO CORROSION 1266.3. FORMS OF CORROSION 128

6.3.1. Sulphide Stress Cracking (SSC) 1286.3.2. Corrosion Caused By CO2 And Cl- 1356.3.3. Corrosion Caused By H2S, CO2 And Cl- 137

6.4. CORROSION CONTROL MEASURES 1386.5. CORROSION INHIBITORS 1396.6. CORROSION RESISTANCE OF STAINLESS STEELS 139

6.6.1. Martensitic Stainless Steels 1396.6.2. Ferritic Stainless Steels 1406.6.3. Austenitic Stainless Steels 1406.6.4. Precipitation Hardening Stainless Steels 1406.6.5. Duplex Stainless Steel 142

6.7. COMPANY DESIGN PROCEDURE 1426.7.1. CO2 Corrosion 1426.7.2. H2S Corrosion 142

6.8. MATERIAL SELECTION 1446.8.1. OCTG Specifications 1456.8.2. DHE Specifications 1466.8.3. Wellhead Specifications 147

6.9. ORDERING SPECIFICATIONS 152

7. TUBING DESIGN 1537.1. POLICIES 1537.2. THEORY 153

7.2.1. Mechanical Properties of Steel 1547.2.2. Temperature 1587.2.3. Tubing Movement/Stress Relationship 158

7.3. WELL DATA. 1607.3.1. Casing Profile/Geometry 1607.3.2. Tubing Data 1607.3.3. Bottom-hole Pressure 1607.3.4. Temperatures (Static and Flowing) 1607.3.5. Reservoir Fluids 1617.3.6. Completion Fluid 161

7.4. PRESSURE INDUCED FORCES 1617.4.1. Piston Effect 1627.4.2. Buckling Effect 1637.4.3. Ballooning Effect 1677.4.4. Temperature Effect 168

7.5. EVALUATION OF TOTAL TUBING MOVEMENT 169

ARPO




7.6. ANCHORED TUBING 1707.6.1. Tubing Permitting Limited Motion 1727.6.2. Packer Setting 174

7.7. TUBING LOAD CONDITIONS 1747.7.1. Pressure Testing 1747.7.2. Acid Stimulation 1757.7.3. Fracturing 1757.7.4. Flowing 1777.7.5. Shut-In 1777.7.6. Load Condition Summary 181

7.8. TUBING SELECTION 1817.8.1. Critical Factors 1827.8.2. Tubing Size And Weight 1827.8.3. Anchoring Systems 184

7.9. TUBING CONNECTIONS 1857.9.1. Policy 1857.9.2. Class of Service 1857.9.3. Selection Criteria 1867.9.4. NACE And Proximity Definitions 1897.9.5. CRA Connections 1907.9.6. Connection Data 190

7.10. TUBING STRESS CALCULATIONS 1907.10.1. Calculation Methods 1917.10.2. Safety Factor 1937.10.3. External Pressure Limit 1957.10.4. Packer Load Limits 1957.10.5. Example Manual Calculation 1967.10.6. Example Computation 205

8. SUB-SURFACE EQUIPMENT 2068.1. PACKERS 206

8.1.1. Selection Procedure 2078.1.2. Selection Criteria 2078.1.3. Well Classification 2098.1.4. Packer Selection For Single String Completion 2098.1.5. Single Selective Completion Packers 217

8.2. SUB-SURFACE SAFETY VALVES 2238.2.1. Policy 2238.2.2. Applications 2238.2.3. Wireline Retrievable Safety Valves 2238.2.4. Surface Controlled Sub-Surface Safety Valves 2248.2.5. Valve Type/Closure Mechanism Selection 224

ARPO




8.3. CONTROL/INJECTION LINE SELECTION 2258.3.1. Control Lines 2258.3.2. Injection Lines 2258.3.3. Tube Specifications 2268.3.4. Material Selection 2288.3.5. Fittings 2308.3.6. Protectors 2308.3.7. Encapsulation 2318.3.8. SCSSV Hydraulic Control fluid 2338.3.9. Control/Injection Line Selection Procedure Flow Chart 236

8.4. WIRELINE NIPPLE SELECTION 2378.4.1. Tapered Nipple Configuration 2388.4.2. Selective Nipple Configuration 239

9. PERFORATING 2409.1. SHAPED CHARGE PERFORATING 2409.2. GUN TYPES 241

9.2.1. Wireline Conveyed Casing Guns 2419.2.2. Through-Tubing Hollow Carrier Guns 2439.2.3. Through-Tubing Strip Guns 2439.2.4. Tubing Conveyed Perforating 243

9.3. GUN PERFORMANCE 2449.3.1. API And Performance Data 2449.3.2. Underbalanced Perforating 2469.3.3. Firing Heads 2479.3.4. Perforating Procedures 247

10. ARTIFICIAL LIFT 25010.1. GAS LIFT 251

10.1.1. Impact On Completion Design 25310.1.2. Common Problems 254

10.2. ELECTRICAL SUBMERISBLE PUMPS 25410.2.1. ESP Performance 25610.2.2. Impact On Completion Design 25910.2.3. Common Problems 259

10.3. HYDRAULIC PUMPING SYSTEMS 26010.3.1. Impact On Completion Design 262

10.4. ROD PUMPS 26210.4.1. Impact On Completion Design 265

10.5. SCREW PUMP SYSTEMS 26510.6. PLUNGER LIFT 26510.7. SUMMARY ARTIFICIAL LIFT SELECTION CHARTS 268

10.7.1. Design Considerations And Comparisons 26810.7.2. Operating Conditions Summary 27010.7.3. Artificial Lift Considerations 272

ARPO




11. USE OF UNDERBALANCE COMPLETION FLUIDS 27411.1. POLICY 27411.2. BARRIER PRINCIPLES 27411.3. APPLICATION 27411.4. RISK ASSESSMENT 275

11.4.1. Well Testing 27511.4.2. Completions 275

APPENDIX A - REPORT FORMS 276A.1. INITIAL ACTIVITY REPORT (ARPO 01) 277A.2. DAILY REPORT (ARPO 02) 278A.3. WASTE DISPOSAL MANAGEMENT REPORT (ARPO 06) 279A.3. PERFORATING REPORT (ARPO 07) 280A.4. GRAVEL PACK REPORT (ARPO 08) 281A.5. MATRIX STIMULATION/HYDRAULIC FRACTURE REPORT (APRO 09) 282A.6. WIRELINE REPORT (ARPO 11) 283A.7. PRESSURE/TEMPERATURE SURVEY REPORT (ARPO 12) 284A.8. WELL PROBLEM REPORT (ARPO 13) 285A.9. WELL SITUATION REPORT (ARPO 20) 286

APPENDIX B - NOMENCLATURE FOR TUBING CALCULATIONS 287

APPENDIX C - ABBREVIATIONS 289

APPENDIX D - BIBLIOGRAPHY 292

APPENDIX E - TUBING MOVEMENT/STRESS COMPUTER PROGRAMMES 294

ARPO



REVISIONSTAP-M-1-P-7100 0

1. INTRODUCTION

1.1. PURPOSE OF THE MANUALThe purpose of this manual is to guide experienced engineers of all technical disciplines,within the Eni-Agip Division and Affiliated Companies, in the completion design process andits importance on well productivity, well servicing capabilities and completion life. These inconsequence, have a large impact on costs and field profit.The Corporate Standards in this manual define the requirements, methodologies and rulesthat enable to operate uniformly and in compliance with the Corporate Company Principles.This, however, still enables each individual Affiliated Company the capability to operateaccording to local laws or particular environmental situations.The final aim is to improve performance and efficiency in terms of safety, quality and costs,while providing all personnel involved in Drilling & Completion activities with commonguidelines in all areas worldwide where Eni-Agip operates.The approach to completion design must be interdiscipline, involving Reservoir Engineering,Petroleum Engineering, Production Engineering and Drilling Engineering. This is vital inorder to obtain the optimum completion design utilising the process described in thismanual.The manual will provide the engineers within the various disciplines with a system to guidethem through the process with the objectives of helping them make the key decisions andobtaining the optimum design to maximise productivity and, hence profit.Many of the decisions made by the various disciplines are interrelated and impact on thedecisions made by other disciplines. For instance, the decision on the well architecture maysubsequently be changed due to the availability of well servicing or workover techniques.This does not mean that the process is sequential and many decisions can be made fromstudies and analysis run in parallel.The design process consists of three phases:

Conceptual Detailed design Procurement.

The process of well preparation and installation of completions is fully described in theCompletions Procedures manual.The activities in each phase are illustrated in figure 1.a, figure 1.b and figure 1.c.The conceptual design process guides the engineers through analysis and key questions tobe considered. During this phase, the user will resolve many of the dilemmas, raised by theinterrelated decisions, at an early time. The final conceptual design will be used as the basisfor the detailed design process.The conceptual design process begins at the field appraisal stage when a Statement OfRequirements (SOR) of the completion is produced. It is essential that this is an accuratestatement including all the foreseen requirements, as it has a fundamental effect on thefield final design and development.

ARPO




As more information is gleamed from further development wells and as conditions change,the statement of requirements need to reviewed and altered to modify the conceptualdesign for future wells. This provides a system of ongoing completion optimisation to suitchanging conditions, increased knowledge of the field and incorporate new technologies.

Figure 1.A - Conceptual Completion Design Process

ARPO




Figure 1.B - Detailed Completion Design Process

ARPO




Figure 1.C - Procurement Process

ARPO




1.2. OBJECTIVESThe fundamental objectives for a completion are:

Achieve a desired (optimum) level of production or injection. Provide adequate maintenance and surveillance programmes. Be as simple as possible to increase reliability. Provide adequate safety in accordance with legislative or company requirements

and industry common practices. Be as flexible as possible for future operational changes in well function. In conjunction with other wells, effectively contribute to the whole development

plan reservoir plan. Achieve the optimum production rates reliably at the lowest capital and

operating costs.

These may be summarised as to safely provide maximum long term profitability. This,however, in reality is not simple and many critical decisions are needed to balance longterm and short term cash flow and sometimes compromises are made.An expensive completion may derive more long term profit than a low cost completion butthe initial capital costs will be higher (Refer to figure 1.d).

Figure 1.D - Completion Design Versus Profitability

ARPO




On the other hand if the data available is not accurate, the estimate of some wellperformance and characteristics throughout the life of the well may be wrong and earlyworkover or well intervention operations will impact on well profitability.An inherent problem is that the Reservoir Engineering Departments objectives do notcoincide with the Completion Engineering Departments in that Reservoir Engineeringsobjectives are for the whole field performance whereas the Completion Groups is tooptimise for profit on a long term well by well basis which includes well servicing/workover.Reservoir and geoscience groups often have to set plans and objectives for the field on wellperformance based on limited information, in the early stages, but are not concerned aboutproduction problems, well maintenance or detailed operations.

1.3. FUNCTIONS OF A COMPLETIONThe main function of a completion is to produce hydrocarbons to surface or deliver injectionfluids to formations. This is its primary function, however a completion must also satisfy agreat many other functions required for safety, optimising production, servicing, pressuremonitoring and reservoir maintenance.These main functional requirements must be built into the conceptual design and include:

Protecting the production casing from formation pressure. Protecting the casing from corrosion attack by well fluids. Preventing hydrocarbon escape if there is a surface leak. Inhibiting scale or corrosion. Producing single or multiple zones. Perforating (underbalanced or overbalanced). Permanent downhole pressure monitoring.

1.4. MANUAL UPDATING, AMENDMENT, CONTROL & DEROGATIONThe Corporate Standards in this manual define the requirements, methodologies and rulesthat enable to operate uniformly and in compliance with the Corporate Company Principles.This, however, still enables each individual Affiliated Company the capability to operateaccording to local laws or particular environmental situations.The final aim is to improve performance and efficiency in terms of safety, quality and costs,while providing all personnel involved in Drilling & Completion activities with commonguidelines in all areas worldwide where Eni-Agip operates.

ARPO




2. RESERVOIR CONSIDERATIONS

2.1. INTRODUCTIONOil and gas wells are expensive faucets that enable production of petroleum reservoirs orallow injection of fluids into an oil or gas reservoir. As pointed out in section 1.1, acompletion conceptual design must take into account all the well objectives to produce theoptimum design to maximise profitability.The purpose of this section is to consider the characteristics of reservoir fluids and the flowof these in the area around the wellbore to allow these parameters to be tied into the wellcompletion design and well intervention/workover operational requirements.

2.2. CHARACTERISTICS OF RESERVOIR ROCKS2.2.1. Porosity

Porosity or pore space in reservoir rocks provides the container for the accumulation of oiland gas and gives the rock characteristic ability to absorb and hold fluids. Most commercialreservoirs have sandstone, limestone or dolomite rocks, however some reservoirs evenoccur fractured shale.

2.2.2. PermeabilityPermeability is a measure of the ability of which fluid can move through the interconnectedpore spaces of the rock. Many rocks such as clays, shales, chalk, anhydrite and somehighly cemented sandstones are impervious to movement of water, oil or gas even althoughthey may be quite porous. Darcy, a French engineer, working with water filters, developedthe first relationship which described the flow through porous rock which is still used today.Darcys Law states that the rate of flow through a given rock varies directly with permeability(measure of the continuity of inter-connected pore spaces) and the pressure applied, andvaries inversely with the viscosity of the fluid flowing.In a rock having a permeability of 1 Darcy, 1cc of a 1cp viscosity fluid will flow each secondthrough a portion of rock 1cm in length and having a cross-section of 1cm2, if the pressureacross the rock is 1 atmosphere.

pALqK

=

Eq. 2.A

ARPO




In oilfield units the linear form of Darcys Law for flow of incompressible fluid through a rockfilled with only one fluid is:

LB)pp(kA10127.1q 213

= Eq. 2.B

where:q = Flow rate, stb/dayk = Permeability, mdA = Flow rate, ft3 = Viscosity, cpL = Flow length, ftp1 = Inlet pressure, psip2 = Outlet pressure, psiB = Formation volume factor, res bbl/stb

2.2.3. Relative PermeabilityAs normally two or three fluids exist in the same pore spaces in a reservoir, relativepermeability relationships must be considered. Relative permeability represents the ease atwhich one fluid flows through connecting pore spaces in the presence of other fluids, incomparison to the ease that it would flow if there was no other fluid.To understand this, assume a rock filled with only with oil at high pressure where gas hasnot been able to come out of solution:

All available space is taken up by the oil and only oil is flowing. If reservoir pressure is allowed to decline, some lighter components of the oil will

evolve as gas in the pore spaces. Flow of oil is reduced but gas saturation is toosmall for it to flow through the pores.

If pressures to continue to decline, gas saturation continues to increase and atsome point (equilibrium gas saturation) gas begins to flow and the oil rate isfurther reduced.

With further increases in gas saturation, the gas rate continues to increase andless oil flows through the pores until finally only gas flows.

Significant oil may still occupy the pores but cannot be recovered by primaryproduction means as the permeability to oil has dropped to zero.

This same principle governs the flow of oil in the presence of water. The saturation of eachfluid present affects the ease of fluid movement or relative permeability.The gas-oil or water-oil relative permeability relationships of a particular reservoir rockdepend on the configurations of the rock pore spaces and the wetting characteristics of thefluids and rock surfaces. In an oil-water system, the relative permeability to oil is significantlygreater when the rock is water wet.

ARPO




Where two or more fluids are present, the permeability in eq. 2.b represents thepermeability of the rock to the desired fluid. This can be achieved by multiplying absolutepermeability of the rock by the relative permeability of the rock to the desired fluid.

LB)pp(Akk10127.1q

o

21roabs3

= Eq. 2.C

where:qo = Oil flow rate, stb/daykabs = Absolute permeability, mdkro = Relative permeability to oil

For a well producing both water and oil, the water cut or fraction of water in the total flowstream at standard conditions of temperature and pressure can be calculated by:

o

w

o

w

w

ow

BB

kk1

1f+

+

=

Eq. 2.D

where:ko = Relative permeability to oilkw = Relative permeability to watero = Viscosity of oil, cpw = Viscosity of water, cpBo = Formation volume factor for oil, res bbl/stbBw = Formation volume factor for water, res bbl/stb

2.2.4. WettabiltyMost reservoirs were formed or laid down in water with oil moving in later from adjacentzones to replace a portion of the water. For this reason, most reservoir rocks are consideredto be water wet. This means that the grains of the rock matrix are coated with a film ofwater permitting hydrocarbons to fill the centre of the pore spaces. The productivity of oil inthis condition is maximised.Although it is extremely difficult to determine wettability of cores due to the cutting andpreparing specimens for laboratory testing which alters the wettability characteristics, it isnot important as this characteristic is included in the permeability measurements.However, it is important when completing or servicing the well in that any foreign substancewhich may come into contact with the rock may alter its wettability characteristic and reducethe relative permeability to hydrocarbon fluids and cause emulsion which may block flow.

ARPO




2.2.5. Fluid DistributionThe distribution of fluids vertically in the reservoir is very important as the relative amountsof oil, gas and water present at a particular level determines the fluids that produced by awell completed at that level and also influence the relative rates of fluid production.In rock the capillary forces, which are related to water wettability, work to change the normalsharp interfaces between the fluids separated by density.From the point in a zone of the free water level upward to some point where watersaturation becomes constant is called the transition zone. Relative permeability permitsboth water and oil to flow within the transition zone. Water saturation above the transitionzone is termed irreducible water saturation or more commonly the connate watersaturation. Above the transition zone, only oil will flow in an oil-water system.Connate water is related to permeability and pore channels in lower rocks are generallysmaller. For a given height, the capillary pressure in two different pore sizes will be thesame, therefore the water film between the water and the oil will have the same curvature,hence more oil will be contained in larger pore spaces.The nature and thickness of the transition zones between the water and oil, oil and gas, andwater and gas are influenced by several factors: uniformity, permeability, wettability, surfacetension and the relative density differences between the fluids. These can be summarisedin three statements:

The lower the permeability of a given sand, the higher will be the connate watersaturation.

In lower permeability sands, the transition zones will be thicker than in higherpermeability sands.

Due to the greater density difference between gas and oil as compared to oiland water, the transition zone between the oil and gas is not as thick as thetransition zone between oil and water.

A well completed in the transition zone will be expected to produce both oil and water,depending on the saturations of each fluid present at the completion level. figure 2.asummarises oil, water and gas saturation in a typical homogeneous rock example.

ARPO




Figure 2.A - Example Fluid Distribution in a Uniform Sand Reservoir(Containing Connate Water, Oil and Gas Cap)

2.2.6. Fluid Flow In The ReservoirOil has little natural ability to produce itself into the wellbore. It is produced principally bypressure inherent in gas dissolved in oil, in associated free gas caps, or in associatedaquifers.

Pressure Distribution Around the WellborePressure distribution in the reservoir and factors which influence it are of great ofsignificance in interpreting well production trends caused by pressure characteristics.Pressure distribution around a producing oil well completed in a homogeneous zone willgradually drop from the reservoir pressure some distance from the wellbore until closer tothe wellbore where it will decline quite sharply. The wellhead pressure will be much lowerdue to the influence of hydrostatic pressure and tubing frictional effects.

ARPO




In a radial flow situation, where fluids move towards the well from all directions, most of thepressure drop in the reservoir occurs fairly close to the wellbore. As shown in figure 2.b, in auniform sand, the pressure drop across the last 15ft of the formation surrounding thewellbore is about one half of the total pressure drop from the well to a point 500ft away inthe reservoir. Obviously flow velocities increase tremendously as fluid approaches thewellbore. This area around the wellbore is the critical area and as much as possible shouldbe done to prevent damage or flow restrictions in this critical area.

Figure 2.B - Pressure Distribution Near Wellbore In Radial Flow

Radial Flow Around The WellboreSteady state radial flow of incompressible fluid is described by Darcys Law:

)r

r(n1B)pp(kh00708.0q

w

o

wo

=

Eq. 2.E

Corrections are required to account for the flow of compressible fluids and for turbulent flowvelocities.

ARPO




Figure 2.C- Units For Darcys Law Equation

For non-homogeneous zones, which is the usual case, permeablities must be averaged forflow through parallel layers of differing permeabilities.

321

332211

hhhhkhkhkk

++

++=

Eq. 2.F

Figure 2.D - Radial Flow In Parallel Combination of Beds

ARPO




Varying permeabilities around the well in series can be averaged as follows:

3

2

3

2

1

2

1

w

1

w

o

k

)r

r(n1k

)r

r(n1k

)r

r(n1

)r

r(n1k

++

=

Eq. 2.G

Figure 2.E - Radial Flow In Series Combination Of Beds

Linear Flow Through PerforationsIdeally perforating tunnels should provide be large and deep enough to prevent anyrestriction to flow. In cases where there may be sand problems and a gravel pack is used,the tunnels are packed with gravel to hold the formation in place, which will cause arestriction.Flow through perforating tunnels is linear rather radial and Darcys equation must becorrected as turbulent flow usually exists.Experiments have shown that pressure drop through gravel filled perforations comparedwith uncorrected linear flow Darcys Law calculations is substantial as shown in figure 2.fbelow. Curve A indicates that plugging with even high permeability (1 Darcy) sand gives alarge pressure drop. Actual test data with very high permeability sand, curve B, provesturbulent flow results in higher pressure drop than Darcys Law calculations, curve C,predict.Investigators have provided turbulence correction factors which can be applied to Darcysequation to permit calculation of pressure drop through perforating tunnels.

ARPO




Figure 2.F - Pressure Drop Versus Flow Rate Through Perforation

Causes Of Low Flowing Bottom-Hole PressureIn a well with uniform sand and fluid conditions, two factors may cause low flowing bottom-hole pressures. These are permeability and producing rate.With low permeability or excessive rate of production, pressure drawdown will beappreciable higher than normal thus reducing flowing bottom-hole pressures and causingthe well to be placed on artificial lift if higher productions rates are necessary.Low permeability is often caused by damage close to the wellbore through drilling,completion or intervention operations. This is particularly detrimental as the effect close tothe wellbore is greatly magnified.

ARPO




The existence of damage can be calculated by well test results analysing the pressurebuild-up periods. The skin effect (abnormal pressure drop) or the normal radial flowpressure drop can be calculated by:

skh

qB2.141ps

= Eq. 2.H

Other terms which are used to quantify formation damage are Damage Ratio and FlowEfficiency. Damage ratio calculation is:

a

t

qqDR=

Eq. 2.I

where:qt = Theoretical flow rate without damageqa = Actual flow rate observed

also:

swf

wf

actual

ideal

ppppp

JJDR

=

=

Eq. 2.J

Flow efficiency:

wf

swf

actual

ideal

ppppp

JJFE

=

=

Eq. 2.K

In multi-zone completion intervals, where transient pressure testing techniques may givequestionable results concerning formation damage, production logging techniques mayprovide helpful data. Flow profiling may highlight zones, in an otherwise productive interval,which are not contributing to the total flow. Non-contributing zones are likely to have beendamaged.

ARPO




2.2.7. Effects Of Reservoir CharacteristicsReservoir Drive MechanismsIn an oil reservoir, primary production results from existing pressure in the reservoir. Thereare three basic drive mechanisms:

Dissolved gas Gas cap Water drive.

Most reservoirs in actuality produce by a combination of all three mechanisms.In a dissolved gas reservoir, the source of pressure is principally the liberation andexpansion of gas from the oil phase as pressure is reduced.A gas drive reservoirs primary pressure source is the expansion of a gas cap over the oilzone.

A water drive reservoirs principle pressure source is an external water hydrostatic pressurecommunicated to below the oil zone.The effect of the drive mechanism on the producing characteristics must be evaluated in thecompletion design process, and also for later re-completions, to systematically recoverreservoir hydrocarbons. figure 2.g and figure 2.h, show typical reservoir pressures versusproduction trends and gas-oil ratio production trends for the three basic drive mechanisms.In a dissolved gas drive reservoir without any artificial pressure maintenance technique,pressure declines rapidly, gas-oil ratio peaks rapidly and then declines rapidly, and primaryoil recovery is relatively low. Re-completing would not reduce the gas-oil ratio.In a gas cap drive reservoir, pressure declines less rapidly and gas-oil ratios increase as thegas cap expands into the up-structure well completion intervals. Well intervention or re-completion to shut-off up-structure intervals may control the gas-oil ratio, therefore losepressure less rapidly.Water drive reservoirs pressure remains high and gas-oil ratios are lower but down-structure well intervals quickly begin to produce water. This is controlled by wellinterventions or re-completions to shut-off the water production or the well is shut-in.Gradually even the up-structure wells will water out to maximise oil recovery.Obviously many factors must be considered in developing a reservoir, however the mainfactors concentrate on the reservoir itself and the procedure used to exploit hydrocarbonrecovery. Well spacing, or well location, is fundamental and the cost of time, labour andmaterials consumed in the drilling are largely non-recoverable, therefore if developmentdrilling proceeds on the basis of close spacing before the drive mechanism is identified, theinvestment will have already been made.This does not usually present an insurmountable problem as a field of any considerablesize will require a minimum number of wells to be drilled in any case to define the reservoir,i.e. establish the detailed geological picture regarding zone continuity and locate oil-waterand gas-oil contacts. By careful planning when enough information is gained to determinethe well locations, these can be drilled at the appropriate spacing to maximise recovery withthe least amount of wells.

ARPO




Many case histories are available to show problems resulting from reservoir developmentwithout having sufficient information about the stratigraphy of the reservoir.

Figure 2.G - Reservoir Pressure Trends For Various Drive Mechanisms

Figure 2.H - Gas-Oil Ratios Trends For Various Drive Mechanisms

ARPO




With regard to drive mechanisms, some general statements can be made:

Dissolved gas drive reservoirs:Well completions in reservoirs with low structural relief can be made in a regularly spacedpattern throughout the reservoir and, provided the rock is stratified, can be set low in thereservoir bed.A regular spacing can also be used for dissolved gas reservoirs with high angle of dip.Again the completion intervals should be structurally low because of the angle of dip andthe exact sub-surface location would vary with well location on the structure. In this scenarioit would be expected that oil recovery would be greater with the minimum well investment asthe oil will drain down-structure through time. If this is recognised after drilling begins, thewell locations must be changed quickly to take full advantage of the situation.Due to the low recovery by the primary drive mechanism, some means of secondaryrecovery will almost certainly be required at some point in life of the reservoir and the initialwell completion design should take this into account.

Gas cap drive reservoirs:Wells are generally spaced on a regular pattern where the sand is thick, dip angle is lowand gas cap is completely underlayed by oil.Again completion intervals should be low in the structure to permit the gas cap to grow formaximum recovery and minimum gas production.Like the dissolved gas drive reservoir, the wells in thin sands with a high angle of dip islikely to be more efficiently controlled by having the completion irregularly spaced and low toconform to the shape of the reservoir. Regular spacing would place many completions toonear the gas-oil contact. Such reservoirs are common where multiple this sands are foundon a single structure and the oil column is only a fraction of the total productive relief.

Water drive reservoirs:Wells can be spaced on a regular pattern on a thick sand and low angle of dip.Completion intervals should be selected high on the structure to permit long production lifewhile oil is displaced up to the completion intervals by invading water from below.A water reservoir in a thin sand with high angle of dip may best be developed with irregularwell spacing because of the structural characteristics. Regular spacing of the wells maycause early water production and possible early abandonment in conjunction with reducingthe drive effectiveness through excessive water production.Significant levels of water production are unavoidable in later field life when maximisingproduction rates.

ARPO




2.2.8. Reservoir HomogeneityThe general procedures, as described in the previous section is to complete water drivereservoirs high and for dissolved gas drive reservoir low on the structure to obtain anadequate number of wells without excess. However this is only practical if the reservoir isuniform.Most sandstone reservoirs were originally laid down as stratified layers of varying porosityand permeability. Similar assumptions can be made for carbonate and even reef typereservoirs which results in reservoirs of a highly stratified nature. Fluids from such reservoirswill flow through the various layers at different restrictions to flow and often there areimpervious beds between the layers so that fluid cannot flow between the bed to bed. Thisis demonstrated in figure 2.i and figure 2.j.In thin beds or highly stratified beds, fingering of the free gas down from a gas cap, orwater from a water basin, is a distinct possibility, especially if the interval is short andproduction rates are high.If the reservoir is stratified, either by shale breaks or by variations in permeability, it willprobably be necessary to stagger the completion intervals in various members of thereservoir to be sure that each is drained properly. Vertical staggering of the completion canbe effected during development to obtain proportionate depletion of the various strata.Additional distribution of intervals in the various members can then be made during laterwell interventions on the basis of data obtained, experience and operating conditions.To maximise recovery, intervals should be produced independently wherever practical(usually determined by economics). Single string/single zone completions are preferred tofacilitate thorough flushing for higher recovery and flexibility of re-completion to controlreservoir performance. Completions with more than one zone are termed multi-zonecompletions and are required for long completion intervals for obtaining sufficient volumesof production.

Figure 2.I - Irregular Water Encroachment and Breakthrough

ARPO




Figure 2.J - High GOR Production by Encroachment of Gas

2.3. HYDROCARBON DATAThe practical approach to the study of reservoir fluid behaviour is to anticipate pressure andtemperature changes in the reservoir and at surface during production, and to measure, bylaboratory tests, the changes occurring in the reservoir samples. The results of these teststhen provide the basic fluid data for estimates of fluid recovery by various methods ofreservoir operations and also to estimate reservoir parameters through transient pressuretesting.Two general methods are used to obtain samples of reservoir oil for laboratory examinationpurposes, by means of subsurface samplers and by obtaining surface samples of separatorliquid and gas. The surface samples are then recombined in the laboratory in proportionsequal the gas-oil ratio measured at the separator during well testing.Information concerning the characteristics and behaviour of gas needed for gas reservoirs,depends upon the type of gas and the nature of the problem. If retrograde condensation isinvolved, it may require numerous tests and measurements. If the gas is wet with noretrograde condensation, or if dry gas, the information is less complex.

2.3.1. Oil Property CorrelationSeveral generalisations of oil sample data are available to permit correlations of oilproperties to be made (refer to the Compant Well Test Manual for sampling techniques).

ARPO




2.4. RESERVOIR/PRODUCTION FORECASTTo obtain the optimum performance from a well, it is first necessary to determine its fullpotential and which way this can be fully exploited within any technical or economicconstraints. The determination of the wells performance entails analysing the following:

In-flow performance Near wellbore performance and design Multiphase flow of tubing performance Artificial lift.

The process of this analysis is shown in figure 2.k which requires continuous repetitionduring field life to account for changing conditions.The inflow performance relationship (IPR) provides the flow potential of the reservoir intothe wellbore against the resistance to flow of the formation and near wellbore region. Thetheoretical IPR is an idealistic assumption of flow performance without pressure drop due toskin effect in the near wellbore region and governed only by the size, shape andpermeability of the producing zone and the properties of the produced fluids. The basictheory of this is described in this section along with some simplified IPR relationships fromobserved field data.Flow behaviour in the near wellbore region may cause a dramatic effect on the IPR curvewhich results in greatly reduced flow capability. This is characterised by a damaged IPRcurve and the amount of damage or skin effect, is mainly caused by the drilling andcompletion practices. Good drilling and completion practices can or may minimise thisdamage allowing use of the idealised IPR curve to be used for completion design.Some completion designs to deal with reservoir conditions, such as gravel packs forunconsolidated sands, will also cause reduced IPR curves which must be anticipated duringthe design phase. Two phase flow, velocity effects in gas wells, high rate or high GOR oilwells, in undamaged near wellbore regions also reduce the IPR curve. Alternatively,stimulation procedures which can provide a negative skin are desirable as this increasesproduction.Once the IPR is completed, the outflow performance can be determined which takes intoconsideration the relationship between the surface flowrate and pressure drop in the tubing.The prediction of this relationship is complicated by the nature of multi-phase fluid flow.Hence, analysis of the outflow performance requires predictions of phase behaviour,effective fluid density, friction losses and flowing temperatures.The results of the outflow performance analysis are usually produced graphically depictinghow bottom hole flowing pressure (BHFP), or pump intake pressure, varies with flowrateagainst a fixed back-pressure which is normally the wellhead or separator pressure. Thesecurves are termed tubing performance curves (TPC) and the point of intersection is thenatural flowing point as demonstrated earlier in figure 2.k.

ARPO




Figure 2.K - Process of Determining Optimum Well Performance

Selecting, or optimising, the tubing size is necessary to optimise the well performance overthe life of the well and should include the potential benefits of artificial lift systems and/orstimulation to reduce near wellbore skin effects.

ARPO




2.4.1. Inflow PerfomanceThis section addresses the fundamental principles of inflow performance for oil and gaswells. The use of IPRs generated from reservoir simulation models is also described as isthe technique for the applications of the various techniques for predicting inflowperformance. Essentially the less data which is available, the more appropriate it is to usetheoretical radial flow equation. As more data becomes available, an empirical expressioncan be validated and applied, however for larger projects, reservoir simulation is usuallyemployed.

Oil Well - Straight Line IPRThe simplest IPR equation assumes that inflow into a well is proportional to the pressuredifferential between the reservoir and the wellbore which is termed the drawdown.

wfR ppp = Eq. 2.L

where:p = Drawdown pressure, psipR = Reservoir pressure, psipwf = Bottom-hole flowing pressure, psi.

With a straight line IPR, the flow rate is directionally proportional to the drawdown. Thelinear relationship can be substantiated from theoretical arguments for a singleincompressible fluid (i.e. above the bubble point). However, it has been verified that thestraight line approach also provides the accuracy needed for well performance calculationsin situations which exceed the theoretical basis, e.g. low drawdowns and damaged wells.In situations which allow the use of a straight line IPR, the constant of proportionality istermed the productivity index (PI). PI defined as J by the API, is:

wfR ppqJ

=

Eq. 2.M

where:q = Total liquid flow rate at surface under stock tank conditions (14.7psia,

60oF)

ARPO




Figure 2.L - Straight Line IPR or Productivity Index J

The assumption of stable inflow performance relationship, or stabilised flow, is that well isproducing in pseudo-steady state or steady state flow conditions. Before this the wellproduces under transient conditions, as in most well tests, result in higher estimates ofproductivity than when under stabilised conditions.Productivity Index, J, also needs to be treated with caution as Production Engineers andReservoir Engineers assume different basis for J. Production Engineers relate J to grossliquid production (oil and water) whereas Reservoir Engineers relate it to oil productivity.J can be calculated directly from bottom-hole gauges in well test results or estimatedpressures from simulation studies. Oil PI, J, can also be derived theoretically from Darcysradial flow equation:

+

=

S75.0r

rn1B2.141

hkJ

w

eoo

oo

Eq. 2.N

where:h = Net pay thickness, ftko = Effective oil permeability, mdo Reservoir fluid viscosity, cpBo = Reservoir formation volume factor, bbl/stbro = Drainage radius, ftrw = Wellbore radius, ftS = Total effective skin, dimensionless (S = S + Dq)

ARPO




This assumes pseudo-steady state flow from a well in the centre of a circular reservoir andit is worth noting that ko is the effective permeability to oil for an oil PI. As water saturationincreases, Ko obviously decreases and as does Jo.Deviation from the theoretical ideal PI (i.e. S = 0) should be expected as a result ofadditional pressure losses in the near wellbore area due to damage, fractures, increasedgas saturation in oil wells, producing below the bubble point, changes in radial flowgeometry and non-Darcy pressure losses due to high flow velocities in gas wells, high rateor high GOR oil wells.Damaged wells with positive skins have straight line IPRs with PIs less than the ideal PI.Straight line IPRs with PIs greater than the ideal are typical of wells with negative skin suchas when they have been stimulated, have natural fractures or are highly deviated.The PI is very useful for describing the potential of various wells as it combines all rock andfluid properties as well as geometrical issues in a single constant making it unnecessary toconsider these properties individually.

Figure 2.M- Effect of Damage And Fractures on a Wells PI

ARPO




Oil Well - Vogels Two Phase Flow IPRThe previous straight line IPR does not hold with two phase flow (gas and liquid) in thereservoir.Once the BHFP falls below the bubble point pressure, gas saturation builds up around thewellbore which reduces the permeability to liquid which of course reduces well productivityat that particular drawdown compared to predicted by linear PI. This means the true IPR iscurved and, hence the PI J, decreases with increasing drawdown (slopes 1 and 2 in figure2.o). There may also be some non-Darcy gas flow effects in wells producing below thebubble point. Vogel used a computer programme to model a variety of solution gasreservoirs and developed a generalised IPR reference curve to account for the two phaseflow effects below the bubble point. He also presented an approximation using theexpression:

2

R

wf

R

wf

max pp8.0

pp2.01

qq

=

Eq. 2.O

where:pR = Reservoir pressure, psipwf = Bottom-hole flowing pressure, psiq = Liquid production, stb/dqmax = Maximum liquid production rate when pwf = 0, stb/d

Qmax is a theoretical value sometimes referred to as Absolute Open Flow (AOF) of the oilwell.

Figure 2.N - Typical IPR Curve for Saturated Reservoir

ARPO




Vogels equation has been validated through observed field data particularly on pumpedwells with high drawdowns where pwf approaches zero.The model used to develop Vogels reference curve did not include skin effects which wouldtend to straighten the IPR curve. Procedures to correct for skin are available.

Figure 2.O - Vogels IPR Reference Curve

ARPO




Where inflow relationship passes through the bubble point, a straight line IPR is drawnabove the bubble point and the curved IPR signifies the two phase flow below this point. Forthis, Vogels equation is combined with the PI to develop a general IPR equation. This hasbeen published by Brown. When the BHFP is above the bubble point use the normalstraight line equation:

( )wfRo ppJq = Eq. 2.Pand when it drops below the bubble point use the modified Vogel equation:

( )

+=

2

b

wf

b

wfbwfRo p

p8.0pp2.01

8.1JpppJq

Eq. 2.Q

where:pb = Bubble point pressure, psi

If water production is involved, it is dependant upon whether it is produced from the sameinterval or others. As oil is normally produced from a different zone to the water, thefollowing equations are applied:

( )wfRw ppJq = Eq. 2.R

=

2

R

wf

R

wfmaxoo p

p8.0pp2.01qq

Eq. 2.S

If oil and water both flow from the same zone then the Vogel equation is used for the grossflow rate:

( )

+=+

2

R

wf

R

wfmaxoowo p

p8.0pp2.01qqqq

Eq. 2.T

ARPO




Figure 2.P - Combined Straight Line IPR and Vogel IPR

Oil Wells - Generalised IPR CurvesAs described earlier, curvature of the IPR curve is not solely due to the reasons highlightedabove but also due to rate dependent skin. This is where Darcys law which is good formoderate to low flow rates is affected by high velocities. This non-Darcy flow, or turbulence,is sometimes the most dominant factor especially for gravel packs and high rate gas-liquidratio wells.Fetkovich recognised that many oil wells could be handled in the same way as gas wellsusing the curved IPR:

( )n2wf2Ro ppCq = Eq. 2.Uwhere:

C = Linear deliverability coefficientn = Deliverability exponent (0.5 to 1.0)

ARPO




Golan and Whitson showed how this relationship could be expressed in a similar form toVogels reference curve as:

n2

R

wf

max pp1

qq

=

Eq. 2.V

This equation is compared with Vogels reference curve in figure 2.q, for two values of theexponent, n. It is seen that when n = 1, the Vogel and Fetkovich IPRs are similar. It isrecommended that n be assumed to be 1 where no multi-rate data is available.n is considered as the means to account for non-Darcy flow but there is no theoreticaltechnique for finding it as it is a function of the rate used during testing. If multi-rate data isavailable then a log-log plot of q versus (pR2 - pwf2) will give a straight line with a slope of 1/n.

Figure 2.Q - Vogel And Fetkovich IPR Curve Comparisons

Use of this approach will provide better results than Vogels method, however it requiresfour points at widely different flow rates to maximise the benefit of this method. If such datais not available, n should be assumed as 1.Blount and Jones presented an alternative generalised IPR equation which was anextension to the Forcheimer equation to include the non-Darcy flow effects:

2wfR bqaqpp += Eq. 2.W

ARPO




The Darcy flow coefficient, a, can be determined theoretically for a well producing atpseudo-steady state flow in the middle of a circular reservoir:

+

= S75.0r

rlnkh

B2.141a

w

eooEq. 2.X

The skin term, S, is relative to all non-rate dependent skin contributions.The other non-Darcy flow coefficient, b, can also be found theoretically but requires aknowledge of the turbulence factor, , which is rarely measured in the laboratory. Similarly,it takes no account of completion non-Darcy effects such as inefficient perforating, etc.Again, if multi-rate test data is available, both a and b can be determined using a plot of (qR- pwf)/q versus q gives a straight line with a slope of b and an interception of a.In very high permeability wells, coefficient b can be much greater than b and perforatingefficiency (shots/ft and penetration) is a very important to productivity.

Oil Wells - Predicting Future IPRsEstimates of future IPR curves throughout the life of the reservoir are frequently required forproduction forecasting and planning artificial lift designs.The effects of increasing water influx on the gross PI, described earlier in Section 2.2, leadsto a significant increase in skin due to scaling, mobilisation of fines, skin damage duringremedial operations and reduced contribution from reduced pay through plugging back.In solution drive reservoirs, the reservoir pressure will decline against time, shifting the IPRcurve downwards resulting in a decline of the production rate and causing flow instability.The relative permeability to oil will also decrease due to increased gas saturation furthershifting the curve downwards. The liberation of gas also affects the oil fluid properties.Standing presented a method of predicting future IPR curves by the equation:

presentoo

ro

futureoo

ro

present

future

Bk

Bk

*J*J

=

Eq. 2.Y

and:

=

2

futureR

wf

futureR

wffutureRfuturefuture p

p8.0p

p2.01p*JqEq. 2.Z

where:J* = PI at minimal drawdown (i.e. where two phase flow effects are

negligible)

ARPO




J* at present conditions is established by carrying out a well test or theoretically. Relativepermeabilities and fluid saturations are determined from special core analysis data andreservoir material balance analysis (using either analytical calculations or a reservoirsimulation model). Fluid viscosities and volume are determined from PVT correlations.If data for Standings equation are not available, the simpler approach like Fetkovichrelation for predicting qmax in Vogels reference curve. Eickmeier first proposed anexpression based on Fetkovichs work, which in modified form is:

m

futureR

presentR

future.max

present.max

pp

qq

=

Eq. 2.AA

It may be shown theoretically that exponent m could vary between 1 and 3. An exponent of2.5 gives the best fit to the gas drive IPR curves by Vogel while values of 1.66 have beenfound in actual field studies by Eickmeir.

Gas Wells - Simplified Deliverability RelationshipRawlins and Schellardt developed a simplified gas well back-pressure equation whichrelates gas flow rate to the BHFP and is the well Known AOF equation;

( )n2wf2Rg ppCp = Eq. 2.BBThis equation was developed empirically using several hundred multi-rate gas well test dataand not by theory but satisfactorily describes the behaviour of the gas well tests considered.The exponent, n , in the equation must be estimated from one of a number of well testmethods (e.g. isochronal test) due to there being no accepted theoretical basis available. Alog-log plot of (pR2 - pwf2) versus q is conducted from which the slope gives the value of 1/n.This exponent can vary between 1.0 for laminar flow to 0.5 for fully turbulent flow. Obviouslyat low to moderate rates there is little turbulence and n is close to 1, however in high ratesthis is highly improbable and makes the IPR projections almost impossible and erring on theoptimistic side. It is, therefore, critical that well tests are conducted up to or above therate of intended production.The constant C is also found from the log-log plot and varies as a function of flow time untilit reaches a constant pseudo-steady state. In some instances C can be calculated fromreservoir parameters, using kh and S from build-up data but is only applicable if flow islaminar (n = 1). To obtain a value of n, it is normal to test the well at three rates at a fixedperiod of time followed by a single rate until stabilisation is reached to obtain C. Theproblem with this isochronal test is the time required to reach stabilised flow in tight gassands which could be months.While this method is widely used throughout the industry, it is not recommended forestimating IPRs as it lacks the theoretical basis and other rigorous equations are available.

ARPO




Gas Wells - Generalised DeliverabilityDue to the shortcomings of the back-pressure equation described above and sinceturbulence which is common in gas wells, it must be accounted for properly and atheoretical based method is more often used in modern engineering. The expression belowis based on the work of Forchemier and is:

2ggwfR AqAqpp += Eq. 2.CC

The Darcy and non-Darcy coefficients, A and B, are determined in a similar manner as thegeneralised IPR equation for an oil well, however the straight line plot is (pR2 - pwf2)/q versusq. It will be seen that the gas IPR is curved even when the non-Darcy term is 0.eq. 2.cc is not precisely correct since inherent in its derivation is an assumption that theproduct of and z is constant. For most gas compositions this is valid only at pressures lessthan approx 2,000psi or if drawdown pressure changes are small which is the case in highpermeability wells above 3,000psi when z is proportional to pressure, an equation similarto eq. 2.w can be used. Between 2,000psi and 3,000psi, there is curvature in the plot of zagainst p making neither approach applicable. In this range the correct inflow equation iswritten in terms of pseudo-pressures:

= ppb g dpzp2)p(mEq. 2.DD

where:g = Gas viscosity, cpz = Gas deviation factor

and where the integration limits are substituted with the pressure range being considered,normally pg and pwf for inflow calculations, hence:

2ggwfR BqAq)p(m)p(m += Eq. 2.EE

where:

A =

+

S75.0r

rlnhk

T1422w

e

g

B =ghk

TD1422

Here the results of the multi-rate test would be plotted as m(pg) - m(pwf)/q versus q to find avalue of B from the slope and to check the value of A from the intercept.The non-Darcy coefficient B can also be calculated theoretically but, as for oil wells,requires knowledge of the correct turbulence factor, . The non-Darcy skin is also frequentlyaccounted for by using:

++

= gw

e

g

gwfR DqS75.0r

rlnhkTq

1422)p(m)p(m Eq. 2.FF

ARPO




where:D is Derived from well testsqg = Gas flow rate, mscf/dT = Reservoir temperature, oFS = The sum of all non-rate dependent skinD = Rate dependent skinkg = Effective gas permeability, md

As modern test analysis use computer software, the pseudo-pressure values are readilyavailable, therefore there is a growing trend to use gas pseudo pressures for predicting gaswell IPRs at all pressure conditions although the pressure squared method has a use in thefield for convenience.

2.4.2. Reservoir Simulation For IPR CurvesReservoir simulation is commonly used in the development, planning and reservoirmanagement of many fields today. With the use of simulation the production engineer isable not only to predict pressures, WORs and GORs to obtain production targets, but alsoto generate IPR curves for determination of how current and future well IPRs will varyacross the field.To obtain the best use of simulation studies, a model needs to be set up by the reservoirengineer with input from the production engineer. Typically the following should beaddressed:

Assumptions on the minimum permissible value of Pwf as dictated by theoutflow performance altered by varying water-cut, artificial lift or use ofcompression.

Variations between the ideal IPRs and actual IPRs which may be expected fromthe undrilled well locations. This information is derived from well test results andis input into the models theoretical IPR equations as skin factor. Futurestimulation or any damaging effects need to be considered.

Long term effects from well interventions, workovers and movement of fines willhave on near wellbore performance causing changes of skin during the life ofthe project.

Using expected off takes, predict turbulence and two phase flow effects by theuse of total skin S inclusive of near wellbore and rate dependent skin effects.The value of D (Refer to eq. 2.ff) can also be directly entered into somesimulators.

If a PI is entered in rather than skin, well radius, etc., it will be necessary tocorrect it for the grid blocks size and shape.

Outflow performance curves should be derived from an accurate computerprogramme as some programmes are not rigorous in the handling of two phaseflow.

ARPO




The results from such field models will provide the reservoir pressure, production rates andwellbore saturations at various time steps, however judgement is required when using theseresults, in particular check:

Confirm if non-Darcy and multi-phase flow effects have been taken intoconsideration.

Input on skin is realistic for the period covered. Ensure that proposed completion effects on near wellbore performance, e.g.

gravel packing, partial completion, deviation, stimulation, etc. have beenconsidered.

If the reservoir pressure refers to grid block or to the drainage area. Whether rates have been modified for downtime due to maintenance, workover

or sales contracts, etc.

As the use of full field reservoir simulation requires many assumptions and simplificationsare made to manage the problem, therefore the predicted flow rates should not beconsidered as precise and the relevant reservoir engineer should be consulted to establishthe accuracy. They may also be able to advise on possible sudden changes in water cut orgas production due to conning or cusping.Often more reliable predictions in shape of the well IPR can be achieved by engineers usingsingle well models to study the probability of water or gas conning or to model transient welltest results. It is also used to determine the sensitivity of production to drawdown andoptimise perforating strategy.When and as new well data from log and RFT/DST results becomes available, it should beused to update the generalised IPR to reflect the actual pay interval, reservoir quality, skins,saturations, pressure and mechanical data. From this, revisions can be made to thecompletion designs, programmes and production forecast.After using measured IPR curves, the model needs to be updated to include actual log andtest results. Once this achieved, then the model can be used to evaluate the effect ofdepletion, water breakthrough and saturation changes on production and used for artificiallift studies. Care must be exercised, however, in extrapolating the shape of the IPR anddetermining the effects by well operations and production may have on skin.It is extremely important that production engineers understand that the uncertaintiesinvolved and do not give greater reliability on model studies than reasonably can beexpected.

ARPO




2.4.3. IPR SelectionIn developing representative IPRs for a field, the appropriate IPR model needs to beselected based upon the anticipated production conditions. These are summarised again inthe following table:

Type Of Well Producing Conditions Recommended IPR ModelUndersaturated oil Pwf > pb Linear PI or radial flow equationSaturated oil Pwf < pb Vogel or FetkovichDamaged saturated oil Pwf < pb

S > +3Standing or linear PI if verydamaged (S > 7)

Undersaturated oil at pR butsaturated at pwf

PR > pbPwf < pb

Composite Vogel and linear

Wells producing oil and water WC > 0 Use as above for theappropriate oil and linear PI orradial flow equation for water

Water zone WC > 90% Linear PI or radial flow equationHigh rate undersaturated oil q > 25stb/d/ft Blount - Jones or radial flow

equation with turbulenceHigh rate saturated oil q > 25stb/d/ft

Pwf < pbBlount - Jones

Gas wells Pseudo-pressure equation(m(pR) - m(pwf) = Aq + Bq2)Omit B if only single rate dataavailable

Table 2.A - IPR Selection Based on Reservoir Type

The appropriate technique will also depend on the reservoir data that is available which isfunction of the development stage. The selection of an IPR model based on this is given intable 2.b.

ARPO




Radial Flow Equation Reservoir ModelIPRs

Empirical IPRs

Technical EvaluationsProspect evaluationExploration well results

Guestimate potential.Extrapolate test results.

-

-

-

Validateinterpretation

Development PlanningConceptual design, largefield

Primary method. Identify variationsgeographically withtime.

Validate results.Highlight damagerisks.

Conceptual design,small field/single well

Primary method. - Validate results.Highlight damagerisks.

Development plan Validate results andskin assumptions.

Primary method. Validate results.Highlight damagerisks.

Detailed design, largefield

Validate results.Evaluate completionresults.

Primary method. Validate results.Highlight damagerisks.

Detailed design, smallfield/single well

Primary method.Evaluate completionmethods.

If available, use forfuture IPRs.

Validate results.Highlight damagerisks.

Optimising Operations/ WorkoverWell performanceassessment

Estimate skin anddetermine cause.

- Primary method.

Field studies (forecasts/artificial lift, lift/compression)

Primary method forpost workover IPR

Predict future IPR Primary method forcurrent IPRs.

Workover planning Primary method forpost workover IPR.

Predict future IPR Primary method forcurrent IPRs.

Revised developmentplan

Define model input Primary method. Validate reservoirmodel results.

Table 2.B - IPR Selection Based on Development Stage

ARPO




2.4.4. Outflow PerformanceTubing PerformancePredicting fluid flow behaviour in tubing involves combining the basic fundamentals of massmomentum and energy conservation with complex mass transfer phenomena for multi-component hydrocarbon mixtures. Application of these concepts, results in utilising thefollowing interrelated topics:

Phase behaviour. Flowing Temperature prediction. Pressure drop prediction.

The relationship between pressure and temperature drop in wells and PVT behaviour iscomplex. Pressure drop is determined using empirical and semi-empirical correlations andcarried out on computer software programmes. Refer to the following sections.The methods for predicting pressure and temperature drops are addressed in the followingsections.

PVT RelationshipsThere are two PVT methods used in the prediction of mass transfer between oil and gas,the black oil model and the compositional model.The black oil model assumes a constant composition for the liquid phase and accounts formass transfer using the parameters gas-oil ratio and formation volume factor. The variablecomposition model requires performing vapour-liquid equilibrium (VLE) or flash calculationsto determine the amount and composition of both the gas and liquid phases. Each modeluses differing methods to determine the densities and viscosities for each phase andinterfacial surface tension.In general the black oil model is easier to use than the compositional model.

Oil Well - PVT RelationshipsWith most modern software programmes there are four methods of obtaining PVTproperties for oil wells which are listed in order of preference. In the vast majority of casesthere are sufficient data to use the tuned black oil model correlation method.

Interpolate directly from experimental data. Interpolate from compositional simulation data. Tuned black oil model empirical correlations. Untuned black oil model empirical correlations.

ARPO




The approach adopted when choosing the appropriate method for each application shouldbe:

a) Use the black oil model flash vaporisation lab data if they are available. Do notuse differential separation data since it is not representative of the vaporisationthat occurs in the tubing.

b) Use the tuned empirical correlations for black oil model variables if theappropriate although limited experimental data are available.

c) Use black oil model parameters generated from results of compositionalsimulation if it has been performed for incidental reasons, e.g. reservoir orproduction reasons, but only if experimental data is not available.

d) Do not use untuned black oil model empirical correlations unless the dataavailable cannot justify a more rigorous method.

Gas/Gas Condensate Wells - PVT RelationshipsIn software programmes, PVT properties for gas and gas condensate wells must bedescribed with the compositional model. Black oil models parameters should never be usedto predict PVT properties for gas or gas condensate systems.

Temperature Drop CalculationPredicting the temperature loss in the wellbore as a function of depth and time is necessaryto determine PVT properties for use in calculating pressure drop. Some softwareprogrammes, temperature profiles may be specified in five ways:

Linear profile based on measured or assumed wellhead and bottom-holetemperatures.

Profile based on adiabatic heat transfer, i.e. constant temperature throughoutthe length of the string.

Profile based on a specified heat transfer coefficient. Profile based on conservation of energy that utilises complex wellbore heat

transfer calculations. Profile based on a simplified version of the complete rigorous calculation

involving correlating parameter for which there is unavailable information butwith data which are available.

The linear profile is the most widely used due to the complexity of heat transfer calculationsin conjunction with the lack of sufficient measured data. Although the linear approach isunrealistic, the error has been found to be less than 15% in overall temperature drop intypical wells. However, in gas wells it has amore significant effect.Some wells have produced fluids with special properties that are very sensitive totemperatures and more complex heat transfer calculations are required. These are:

Gas condensate wells with retrograde condensate. High pour point crude oil wells. Wells in which hydrate formation can occur.

ARPO




Pressure Drop CalculationCalculating pressure drop in tubing involve numerical integration of the steady-statepressure gradient equation over the entire tubing length. The equation consists of threecomponents and can be expressed as follows:

ACCFRHYD dLdp

dLdp

dLdp

dLdp

+

+

=

Eq. 2.GG

where:

cHYD gsingp

dLdp

=

Eq. 2.HH

is the pressure gradient caused by the hydrostatic head of potential energy of the multi-phase liquid.

Dg2vpf

dLdp

c

2

FR=

Eq. 2.II

is the pressure gradient caused by wall friction.

dLgvdvp

dLdp

cACC=

Eq. 2.JJ

is the pressure gradient caused by fluid acceleration.In multi-phase systems, the variables such as p and v in the pressure gradient equation arenormally averages for the gas and liquid phases present, therefore, the pressure issensitive to the relative amounts of gas and liquid present at any location in the tubing. Thehydrostatic head is the most predominant component of the pressure gradient in oil wells,often accounting for 90% of the pressure drop. The friction losses are the remainder of thepressure loss and are more significant in gas wells with acceleration effects being negligibleexcept when near to atmospheric pressure.Gas and oil phases normally flow at different speeds which is the phenomenon referred toas slippage. This slippage causes an additional accumulation of liquid in the tubing which istermed liquid hold up. The amount of slippage that occurs is dependent upon thegeometrical distribution of the gas and liquid in the pipe, referred to as the flow pattern orflow regime. Flow patterns are governed primarily by the flow rates of each phase, tubingdiameter and to a lesser extent PVT properties.

ARPO




Typical flow patterns are: Annular flow Churn flow Slug flow Bubble flow Liquid flow.

Considering the above, it is obvious that the pressure at each point in the well and,therefore, the total pressure drop is very dependent on flow pattern. Typical pressuregradients in wells for different flow patterns are:

Single phase oil = 0.36psi/ft Bubble flow = 0.25psi/ft Slug flow = 0.20psi/ft Mist flow = 0.1 - 0.2psi/ft

Hence, it is seen that prediction of pressure drop in multi-phase systems is complex andhas led to the development of different correlations to be used. Although many of thesehave been successful to some degree, no single method has been universally beenaccepted.The early developed correlations assumed the flow as homogeneous mixtures ignoringliquid hold up effects. Attempts were made to compensate for these errors in the equationsby single empirical derived friction factor. Subsequent correlations were developed topredict liquid hold up but most of these first required an empirical correlation or map topredict the flow pattern. The accuracy of existing correlations for predicting flow pattern,liquid hold up pressure gradient is limited by the ranges of data used in their developmentand no single method can be applied universally. More recent models developed based onflow mechanisms and conservation principles, referred to as mechanical models, offer morepotential for accurate predictions but these are not readily accepted as standard designmethods as yet.Some software programmes use all the correlations available and the more recentpromising mechanical models can be added.

Flow PatternsTransition between the various flow patterns, as listed in the previous section, can beidentified using flow pattern maps. The most common maps are empirically derived with co-ordinates based on dimensionless groups of variables that include volumetric flow rates,diameter and PVT properties.Although bubble, slug and churn floe predominate in oil wells, it is possible for oil and gaswells to include all flow patterns in addition to single phase liquid and gas.

ARPO




Classification Of MethodsPublished methods of multi-phase flow pressure gradients in wells can be placed into oneof three general categories based on the assumptions from which the method wasdeveloped:

Homogeneous flow correlations where slippage and flow pattern are notconsidered.

Slip flow correlations where slippage is considered but not flow pattern. Flow pattern dependent correlations where liquid hold up and flow pattern are

considered. Mechanised models where slippage, flow pattern and basic flow mechanisms

are considered.

Oil Well CorrelationsOil well correlations for predicting pressure gradients in oil wells have been published andthose most widely accepted in the Industry are:

Duns and Ros (1963) Hagedorn and Brown (1967) Orkiszewski (1967) Aziz, Covier and Fogarasi (1972) Beggs and Brill (1973).

As illustrated in figure 2.r and figure 2.s, these correlations predict different pressure dropsfor the same application, however any one of these may be successful in a given field.Validation and actual field data are the only means of choosing a pressure loss method butthis is not available at the time of designing the completions. Ansari recently performed anevaluation of the most widely used correlations and his own proposed mechanistic model.,performed using the TUFFP well databank consisting of 1775 flowing well surveys coveringa broad range of production variables and pressure loss methods were also evaluated foreach flow pattern. table 2.c presents the overall results below:

Method Average Error AbsoluteAverage ErrorStandardDeviation

Relative PerformanceFactor, RPC

Ansari 9.3 101.3 163.9 1.000Hagbr -28.5 102.8 178.4 1.132Dunros 33.4 110.9 177.7 1.178Aziz -20.8 116.6 190.4 1.198Begbril 41.3 134.9 207.9 1.404Orkis 12.2 151.3 273.3 1.597Mukbr 78.7 159.8 217.2 1.666

Table 2.C - Evaluation of Pressure Loss Methods Using TUFFP Well Databank

ARPO




Selecting the best prediction method from table 2.c is not appropriate as the best statisticalresults do not guarantee the best performance for a specific application. The choice mustbe made on experience. The applicability of the various methods is compared in table 2.d.

Figure 2.R - Comparison Lift Curves for High Gas-Oil Ratio Well

ARPO




Figure 2.S - Comparison of Lift Curves for Low Gas-Oil Ratio Well

ARPO




Method Category Accuracy Data Fluids Application/CommentsAnsari (TUFFP1963)

MechanisticModel

Good N/A N/A Appears a little conservative. Givesconsistent results for all flow patternsand TCP minimum. Needs to beverified through use.

Aziz et al (!972) Flow PatternDependent

Variabledependingon version

Laboratoryand field

Oil, water,gas

Optimistic. tends to under-predictpressure drop.

Beggs and Brill(1973)

Flow Pattern Poor Laboratory Air, water Developed for deviated wells buttends to significantly over-predictpressure drop. Should be avoidedunless well is highly deviated.

Beggs and Brill withPalmer

Flow PatternDependent

Fair Laboratory Air, water Developed for deviated wells buttends to over-predict.

Cornish (1976) Homo-geneous

Good insome flowpatterns

Field(annular

flow)Oil, Gas Does not predict a TPC minimum.

Usually not applicable for completiondesign.

Duns and Ros(1963)


Good Laboratory,

experimental plus

field data

Oil, gas,water

Conservative. Tends to over-predictpressure drop. Good where severalflow patterns exist.

Hagedorn andBrown (1965)

Slip Flow Good insome flowpatterns

Fieldexperiment

Oil, water,air

Does not predict a TCP minimum.Poor in bubble flow. Liquid hold upprediction can be less than for no slipflow. Should be used with caution.

Hagedorn andBrown with GriffithBubble andrestriction on holdup


Good Fieldexperiment

Oil, water,air

Optimistic. Tends to under-predictpressure drop. This is the preferredcorrelation in the absence of otherdata.

Kleyweg et alOccidental mod(1983)

Slip Flow Field Oil, water,Gas

Developed to optimise gas lift inhighly deviated wells (>70o) inClaymore field. Should not be usedexcept for similar conditions.

Orkiszewski (1967) Flow PatternDependent

Fair SomeHagedornand Browndata, field

Oil, water,gas

Conservative. Tends to over-predictpressure drop. can causeconvergence problems in computingalgorithm.

Table 2.D- Applicability of Pressure Loss Prediction Methods

Gas And Gas Condensate CorrelationsFor gas and gas condensate wells the following methods are frequently used:

Cullender and Smith Single phase gas with modified gravities Multi-phase flow correlations Gray correlation.

ARPO




As with oil wells, validation with field data is the only reliable method for determining themost appropriate correlation and, similarly, this is never usually available at the time that thecompletions are designed. If this is the case, the Gray correlation is generally recommendedalthough the Ansari model mat prove to be even more accurate since it includes a goodmodel for predicting pressure gradient in annular flow which is the most predominant in gaswells.Care is needed in the selection of tubing in that, even in low liquid rates, wells can quicklyload up over a few weeks if it is not correctly sized. Although any of the correlations can beused, the Gray correlation is recommended based on the work with Reinicke et al butresults should be used with caution.In gas wells, liquid loading can also be predicted using simplified methods presented withTurner et al which are independent of pressure drop calculations. These methods havebeen reviewed by Lea and Tighe. For wells producing high gas-water or gas-condensateratios, it is recommended that tubing size be assessed using these methods in addition tolift curve methods and that the most conservative approach be adopted.

Effect Of Deviation AngleNowadays most wells of interest to operators are directional or deviated wells. The accuracyof pressure drop calculations in these circumstances using correlations developed forvertical is obviously extremely questionable.Flow pattern and liquid hold up is very dependent on deviation angle. For wells withdeviations up to 45o from vertical, vertical correlations perform accurately enough for wellsgreater than 45o, accounting for deviation by simply using the sine in the hydrostaticcomponent of the pressure gradient equation may not be adequate in these cases, eitherthe Beggs and Brill correlation or a mechanistic model would be necessary.In any study, differing correlations should not be used for different deviations, as thedifference between the predicted pressure drops is generally greater than the effect of thedeviation itself.

Effect Of RestrictionsMost oil and gas wells contain some types of flow control devices in the completion whichchoke flow. The geometry of these restrictions varies from a simple reduced diameter axialflow path to a tortuous complex path. When a multiphase mixture flows through arestriction, the phase velocities dramatically increase. If these reach sonic velocity, criticalflow occurs.For critical flow, simple empirical correlations such as the Gilbert equation are sufficientlyaccurate. For sub-critical flow, behaviour is very dependent on geometry and a simpleBernoulli type equation with a discharge coefficient is recommended.

ARPO




Effect Of ErosionErosion in completions occurs when there are high velocities and if there are solids particlesin the flow stream. The most common points for erosion is where there are restrictionswhich cause increased velocities. The API have published a method in API RP 14E, todetermine the threshold velocities for erosion to occur in piping systems but the validity ofthis for all conditions is questionable.

2.4.5. Flow Rate PredictionFollowing the establishment of both the IPR and TPC, they must be presented in the sameplot from which the intersection of the lines can be used to predict the flow rate of a well atgiven set of stable flow conditions (Refer to figure 2.t ).Changing the system parameters like the tubing ID, reservoir pressure, GLR, etc., will effecteither or both the IPR and TPC and in consequence alters the production rate.Systematically varying the system parameters allows comparison of the incremental effectson production and these can, in turn, be forecast and analysed for cost/benefit of thecompletion options. Continuing in this manner provides information on which decisions canbe made on optimum well configuration or optimum operating conditions. This sectiondescribes this analysis.

Figure 2.T - Combining IPR and TPC Curves

ARPO




Natural Flow PointThe characteristic J shape of the TPC means there can be several possible intersectionswith the IPR as shown in figure 2.t through figure 2.v. The TCP, Pmin, occurs due to the gasand liquid phase velocities differ at low flow rates, i.e. slippage occurs. At low flow rates, thehydrostatic component in the total pressure drop predominates. As liquid velocities tendtoward zero, the gas escapes from the well and the hydrostatic gradient approaches thestatic pressure of the liquid. On the other hand, as the flow rate increases, the hydrostaticcomponent reduces due the gas lift effect while the friction component increases until theminimum is reached when the friction pressure drop exactly offsets the decrease inhydrostatic pressure drop.In figure 2.t, the IPR and TPC curves intersect well to the right of the minimum and, underthese conditions, the well will flow at a stable rate defined as the natural flow point. Theoptimum tubing size, or GLR, will give an intersection well to the right of the pmin and out ofthe flat portion of the TCP curve. but without

eni - completion design manual

Documents