NUMERICAL STUDY OF WATER CONING CONTROL WITH DOWNHOLE WATER SINK (DWS) WELL COMPLETIONS IN VERTICAL AND HORIZONTAL

WELLS

A Dissertation Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy in The Department of Petroleum Engineering

by Solomon Ovueferaye Inikori B. Engineering, University of Nigeria, Nsukka-Nigeria, 1988 M. Engineering, University of Benin, Benin City-Nigeria, 1993 August 2002

ACKNOWLEDGEMENTS The author expresses special gratitude and appreciation to professor Andrew K. Wojtanowicz for his encouragement, patience and guidance as research advisor and the chairman of the examination committee. A deep appreciation is also extended to other members of the examination committee, Dr. Chris White, Dr. Zaki Bassiouini, Chair of the Craft and Hawkins Department of Petroleum Engineering, Dr. Dandina Rao and Dr. Dimitris Nikitopoulos for their valuable review, guidance and assistance throughout this research. The author would also like to acknowledge the members of the Joint Industry Panel (JIP) of the Downhole Water Sink Technology Initiative (DWSTI) for the financial support and the supply of relevant field data for this research. The author wishes to also acknowledge Dr. John McMullan, Dr. Ephim Shirman, Dr. Omowunmi Iledare of the Petroleum Technology Transfer Council at the Louisiana State University, Dr. Sam Ibekwe of Mechanical Engineering Department at the Southern University, Rev. Chibuike Azuoru, (CPA), of Southern University, Mrs. Mary Azuoru of the Louisiana State Department of Education, members of the African Christian Fellowship and Christ the King Christian Center, Baton Rouge, my colleague, Djuro Novakovic and the staff of the Petroleum Engineering Department for all the support and encouragement. Finally, the author is grateful to his wife, Evarista Omoyoma Inikori and three sons, Erosuo Godson Inikori, Ofejiro Hubert Inikori and Rukevwe Nicholas Inikori for all their support, patience, advise, prayers, understanding and encouragement throughout this doctoral program. To God be the glory. Great things He has done. ii

TABLE OF CONTENTS ACKNOWLEDGEMENTS ............................................................................................. ii LIST OF TABLES .......................................................................................................... vii LIST OF FIGURES ....................................................................................................... viii NOMENCLATURE........................................................................................................ xii ABBREVIATIONS ...................................................................................................... xviii ABSTRACT xxi CHAPTER 1 INTRODUCTION......................................................................................1 1.1 Background and Purpose .....................................................................................1 1.2 Statement of Research Problem ...........................................................................3 1.3 Significance and Contribution of this Research...................................................5 1.4 Research Method and Approach..........................................................................6 1.5 Work Program Logic ...........................................................................................8 CHAPTER 2 LITERATURE REVIEW .......................................................................10 2.1 Water Coning in Vertical Wells.........................................................................10 2.2 Other Industry Solutions for Water Coning in Vertical Wells ..........................16 2.2.1 Perforation Squeeze-off and Re-completion..............................................17 2.2.2 Conformance Control Technology ............................................................17 2.2.3 Downhole Oil-Water Separation Technology ...........................................18 2.2.4 Total Penetration Approach .......................................................................19 2.3 Water Coning Control with Dual Completions in vertical wells .......................19 2.4 Capillary Pressure and Relative Permeability Hysteresis in DWS Water Coning Control ..................................................................................................22 2.5 Horizontal Well Technology..............................................................................23 2.6 Horizontal Well and Water Influx Control ........................................................26 2.7 Friction Pressure Loss Consideration in the Horizontal Wellbore ....................29 2.8 Water Cresting Control in Horizontal Wells .....................................................32 2.9 Flow Regimes of Oil-Water Flow in Horizontal Pipes/Wellbore......................33 2.10 Friction Pressure Loss of Oil-Water Flow in Horizontal Wells ........................35 2.11 Numerical Simulation of Water Coning in Wells..............................................36 2.12 Current Methods of Controlling Water Cresting in Horizontal Wells...............37 2.12.1 Horizontal Well Completion With Stinger ................................................37 2.12.2 Water Cresting Control With Variation of Perforation Density .......................................................................................................38 2.12.3 Water Cresting Control with High Angle Wells........................................40 2.13 Summary............................................................................................................41

iii

CHAPTER 3 DWS VERTICAL WELL PERFORMANCE WITH VARIABLE WATER SATURATION...................................................43 3.1 Pressure Stabilization and Steady-state Water Cone Simulation.......................45 3.2 Verification of Results with Analytical Model IPW..........................................52 3.3 Capillary Pressure Transition and Simulation of DWS Well ............................54 3.3.1 Capillary Pressures.....................................................................................54 3.4 Relative Permeability Hysteresis and the DWS Technology Modeling............62 3.4.1 Fractional Flow Analysis ...........................................................................65 3.4.2 Imbibition and Drainage Relative Permeabilities ......................................66 3.4.3 Simulation of Relative Permeability Hysteresis in Eclipse .......................69 3.4.4 Analysis of Results from the Relative Permeability Hysteresis and Capillary Pressure Studies.................................................71 3.5 DWS Well Productivity Performance Studies...................................................74 3.5.1 DWS Well Oil Recovery Performance Study............................................75 3.5.2 DWS Well Productivity Index Comparison ..............................................76 3.5.3 Watercut Performance Evaluation .............................................................81 3.6 DWS Well Design Recommendations for Old Wells........................................82 CHAPTER 4 DWS WELL DELIVERABILITY WITH OIL-FREE DRAINAGE WATER..............................................................................84 4.1 E & P Produced Water Regulations and Requirements.....................................85 4.2 Overview of DWS Environmental Performance ...............................................86 4.3 Capillary Transition and the IPW and Clean Water Drainage...........................88 4.4 Modeling Clean Water Performance of DWS in Old Fields. ............................89 4.4.1 Model Description. ....................................................................................90 4.4.2 Capillary Pressure Transition Evaluation. .................................................91 4.4.3 Relative Permeability Curves. ...................................................................92 4.5 Presentation and Discussion of Results .............................................................92 4.6 Design Considerations for DWS Wells With Oil-free Drainage Water ............94 4.7 Effect of Location of the Water Sink in the Capillary Transition Zone ............97 4.8 Recommendations for Designing DWS Well and Production Schedule ...........98 CHAPTER 5 EFFECT OF FLOW FRICTION ON WATER CRESTING IN HORIZONTAL WELLS.......................................................................100 5.1 Limitations of Water Coning Control With Horizontal Wells Field Evidence ..........................................................................................................101 5.2 Model of Oil-Water Flow in Horizontal Wells................................................104 5.2.1 Flow Regimes and Conditions in Horizontal Wellbore...........................104 5.2.2 Properties of Oil-water Mixture in Horizontal Wellbore.........................107 5.2.3 Frictional Pressure Loss in Horizontal Wellbore.....................................114 5.2.4 Generalized Friction Factor Correlation for Horizontal Wells and Pipes........................................................................................120 5.2.5 Verification of Generalized Friction Factor (GFF) Relation ....................................................................................................121 5.3 Numerical Study of Water Cresting in Horizontal Wells With Compound Frictional Pressure Loss Model ....................................................125 iv

5.3.1

Changes in Horizontal Wells Water Crest Development Due to New Model ................................................................................127

CHAPTER 6 WATER CRESTING CONTROL IN HORIZONTAL WELLS WITH DOWNHOLE WATER SINK TECHNOLOGY....................134 6.1 Feasibility of Water Cresting Control with Dual Completion Technology......................................................................................................134 6.1.1 Drilling and Completion of the Tail-pipe Multi-lateral Well Option ......................................................................................................137 6.1.2 Drilling and Completion of the Bilateral Water Sink Well Option ......................................................................................................140 6.2 Water Cresting Control Performance of the Tail-pipe and Bilateral Water Sink Technology...................................................................................144 6.2.1 Effective Length of Bilateral Water Sink Horizontal Well .....................147 CHAPTER 7 COMPARISON OF DWS VERTICAL WELLS AND HORIZONTAL WELLS IN BOTTOM WATER DRIVE RESERVOIR..........................................................................................149 7.1 Equivalent Length/Drainage Area of Horizontal Wells...................................152 7.1.1 Method 1: Two Half-circle and Rectangle Approach..............................153 7.1.2 Method 2: Ellipse Shaped Drainage Area Method ..................................153 7.2 Simulation Grid Properties of Horizontal Well and Vertical Well..................155 7.3 Comparison of Watercut Performance of Vertical and Horizontal Wells .......156 7.3.1 Oil Production Comparison Study...........................................................157 7.3.2 Watercut Performance Evaluation ...........................................................159 7.4 Comparison of Vertical Wells with DWS Technology and Horizontal Wells ......................................................................................161 7.4.1 Two Vertical DWS Wells Vs Horizontal Well Oil Production Analysis .................................................................................161 7.4.2 Two Vertical DWS Wells Vs Horizontal Well Watercut Analysis....................................................................................................163 7.5 Economic Evaluation of the DWS Technology Vs Horizontal Well ..............164 CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS...............................167 8.1 Conclusions......................................................................................................167 8.1.1 Relative Permeability Hysteresis and Capillary Pressure Transition Zone on DWS Well Performance...........................................168 8.1.2 Oil-Free Water Production Capacity of DWS with Capillary Pressure Zone...........................................................................169 8.1.3 Water Cresting in Horizontal Wells and DWS Technology....................170 8.1.4 Vertical Wells with DWS and Conventional Horizontal Wells ........................................................................................................171 8.2 Recommendations............................................................................................172 REFERENCES ..........................................................................................................173 v

APPENDIX A CAPILLARY PRESSURE/RELATIVE PERMEABILITY HYSTERESIS DATA DECK ...............................................................183 APPENDIX B WATER CONE REDUCTION STUDY ON WELL LVT ...........188 B.1 Executive Summary.........................................................................................188 B.2 Objective ..........................................................................................................188 B.3 Input Data for Eclipse Simulator .....................................................................189 B.4 Approach to Study ...........................................................................................190 B.5 Discussion of Results.......................................................................................193 B.5.1 Scenario A Watercut Reduction Without the Effect of Capillary Pressure. ...................................................................................193 B.5.2 Scenario B Watercut Reduction Performance with Capillary Pressure Effect. ........................................................................194 B.5.3 Scenario C Watercut Reduction Performance with Leaking Cement. ......................................................................................195 B.5.4 Scenario D Watercut Reduction Performance with high rate at top Completion..............................................................................195 B.6 Appropriate Watercut Level for Deployment of DWS Technology................196 B.7 Conclusion of Studies ......................................................................................197 APPENDIX C DATA DECK FOR BILATERAL WATER SINK (BWS) CONCEPT FOR WATER CRESTING CONTROL .........................198 APPENDIX D DATA DECK FOR TAIL-PIPE WATER SINK (TWS) CONCEPT FOR WATER CRESTING CONTROL .........................205 APPENDIX E - DWS WELL COST DATA SPREADSHEET.................................212 APPENDIX F SAMPLE DATA OF MODIFIED PIPE ROUGHNESS...............214 VITA ..........................................................................................................216

vi

LIST OF TABLES Table 3-1 Craig's Rules-of-Thumb for Determining Rock Wettability............................ 64 Table 4-1 Requirements on Hydrocarbon Contamination of Produced Water................. 85 Table 4-2 Hydrocarbon Contamination of Drainage/System Water................................. 87 Table 4-3 PAH Contamination of Drainage /System Waters ........................................... 87 Table 4-4 LVT Reservoir Input Data................................................................................ 90 Table 5-1 Experimental Values for the Constants, a, b and Cn....................................... 118 Table 6-1 Oil Recovery for Conventional Horizontal Well with Fanning Friction Factor...............................................................................................................144 Table 6-2 Improved Oil Recovery in Horizontal Well with DWS Technology............. 145 Table 7-1 Reservoir/Rock Properties of Horizontal Well Vs Vertical Well Water Cresting/coning Performance Studies ............................................................ 155 Table 7-2 Drilling and Completion Cost for 2-Vertical Wells and 1 Horizontal Well... 165

vii

LIST OF FIGURES Figure 2-1 Schematic Representation Water Coning in Vertical Wells ........................... 11 Figure 2-2 Cone Stability Analysis................................................................................... 13 Figure 2-3 Schematic of the dual completion for the Nebo Hemphill project.................. 21 Figure 2-4 Oil-Water Flow Regimes in Horizontal Pipes ................................................ 34 Figure 2-5 Horizontal Well Completion with a Stinger.................................................... 38 Figure 2-6 Schematic of Typical Variable Perforation Density ....................................... 39 Figure 2-7 Schematic of High-Angle Well with tip Close to the OWC ........................... 40 Figure 3-1 Typical Inflow Performance Window (IPW) and its Regions ........................ 44 Figure 3-2 Steady-state Numerical Simulation Models................................................... 46 Figure 3-3 Unstable Flowing Bottom-hole Pressure at Top Completion ....................... 47 Figure 3-4 Unstable Wellbore Pressure Distribution with time...................................... 49 Figure 3-5 Creeping Watercut for Unstable Flowing Bottomhole Pressure................... 50 Figure 3-6 Overlap of Pressure at Different times Indicate a Stable Steady-state Flow Process ...................................................................................................................... 50 Figure 3-7 Watercut Profile for Stabilized Bottomhole Pressure (Plateau)...................... 51 Figure 3-8 Comparison of IPW - Numerical Simulator and Analytical Model................ 52 Figure 3-9 Typical Capillary Transition Pressure Zone ................................................... 58 Figure 3-10 Capillary Pressure Hysteresis........................................................................ 59 Figure 3-11 Effect of Capillary Transition Zone on DWS IPW....................................... 62 Figure 3-12 Relative Permeability Curves for Drainage and Imbibition......................... 68 Figure 3-13 Relative Permeability Hysteresis Curve with Scanning Option ................... 70 Figure 3-14 Negligible Hysteresis Effect Due to use of Same End-point Saturations for both Imbibition and Drainage. .................................................................................. 72 viii

Figure 3-15 Combined Effects of Capillary Pressures and Relative Permeability Hysteresis.................................................................................................................. 73 Figure 3-16 Oil Recovery Performance Study of DWS Well........................................... 75 Figure 3-17 Field liquid PI Comparison between DWS Well and Conventional Well .... 77 Figure 3-18 Oil Productivity Index (OPI) Comparison .................................................... 78 Figure 3-19 Oil Recovery Plot for Same Drawdown Pressure ......................................... 79 Figure 3-20 Improved Oil Productivity Index with Rate.................................................. 80 Figure 3-21 Watercut and Cumulative Oil Recovery Performance.................................. 81 Figure 4-1 Reduction in Size of IPW Due to Capillary Transition Zone ......................... 88 Figure 4-2 Shows Zone of Concurrent Production of Contaminated Fluid in both Completion................................................................................................................ 89 Figure 4-3 Capillary Pressure Plot from Leverett J-function............................................ 91 Figure 4-4 Relative Permeability Curves from Cores....................................................... 92 Figure 4-5 Transition Zone Enlargement around Conventional Well .............................. 93 Figure 4-6 Transition Zone Enlargement around Dual-completed (DWS) Well ............. 93 Figure 4-7 IPW Comprising Transition Zone Effect (LVT reservoir); Critical Rate lines Intercept and Diverge at Higher Rates Enveloping Region 4 of Twophase Inflow at the Top and Bottom Completion......................................................95 Figure 4-8 IPW for LVT Reservoir With WC Isolines..................................................... 96 Figure 4-9 Optimization of Oil Production Rate for DWS Well With Oil-free Water Drainage.................................................................................................................... 96 Figure 4-10 Oil Contamination in Water sink Completion placed within Capillary Transition Zone......................................................................................................... 98 Figure 5-1 Uniform Flux Model of Horizontal Well shows Uniform Water Crest Development along the Length of the Well............................................................ 102 Figure 5-2 Finite conductivity Model of Horizontal Well (Non-uniform Crest Development).......................................................................................................... 102 ix

Figure 5-3 Oil-Water mixture Viscosity Evaluation....................................................... 110 Figure 5-4 Effect of Mixture Viscosity Model on Reynold's Number ........................... 112 Figure 5-5 Effect of Viscosity Model on Fanning Friction Factor ................................. 113 Figure 5-6 Effect of Perforation Without Fluid influx on Wellbore Friction Factor...... 116 Figure 5-7 Effect of Perforation and Fluid Influx on Wellbore Friction ........................ 119 Figure 5-8 Friction Factor Plot from 'Generalized Equation' Vs Yuan's Experimentally Derived Correlations (5SPF).................................................................................. 121 Figure 5-9 Friction Factor Plot from 'Generalized Equation' Vs Yuan's Experimentally Derived Correlations (10SPF)................................................................................. 122 Figure 5-10 Performance of 'Generalized Equation' With in-situ pipe Roughness Vs Blasius Equation for Smooth Pipe. ......................................................................... 123 Figure 5-11 Performance of 'Generalized Equation' Vs Blasius Equation for smooth pipe..............................................................................................................124 Figure 5-12 Friction factor comparison (Model Vs weighted average) in oil-water flow with phase inversion ............................................................................................... 126 Figure 5-13 Early Water Breakthrough in Horizontal Wells.......................................... 128 Figure 5-14 Early Water Breakthrough but gradual Development of Watercut............. 128 Figure 5-15 Friction Pressure Loss Effect on Horizontal Well Bottom Hole Pressure .. 130 Figure 5-16 New Water Crest Development - Skewed toward the Heel of the Horizontal Well........................................................................................................131 Figure 5-17 Weighted Average Approach shows Symmetrical Water Crest ................. 131 Figure 6-1 Tail-Pipe Water Sink Completion Schematic ............................................... 135 Figure 6-2 Bilateral Horizontal Well Water Sink Option............................................... 136 Figure 6-3 A level 5 Multi-lateral Schematic for Tail-pipe Water Sink......................... 137 Figure 6-4 Multi-lateral Well Project in Purdue Bay, Alaska (Courtesy, Bakerhughes website) ............. 141 Figure 6-5 Dual Completion Multi-lateral Well in Saudi Arabia (Al-Umair, A.N., 2000) ............... 142 x

Figure 6-6 Cumulative Oil Recovery Comparison, Conventional Horizontal Well, Horizontal Well with Tail-pipe Water Sink and Horizontal Well with Bilateral Water Sink .............................................................................................................. 146 Figure 6-7 Evaluation of Effective Length of Bilateral Water Sink Well ...................... 148 Figure 7-1 Productivity Comparison between Vertical and Horizontal Wells ............... 150 Figure 7-2 Horizontal Well Drainage Area .................................................................... 153 Figure 7-3 Elliptical Drainage Area for Horizontal Well ............................................... 154 Figure 7-4 Oil Production Rate Performance Evaluation ............................................... 157 Figure 7-5 Cumulative Oil Recovery Performance Analysis ......................................... 158 Figure 7-6 Watercut Performance Evaluation of Vertical and Horizontal Wells ........... 160 Figure 7-7 Oil Production with DWST Vs Horizontal Well .......................................... 161 Figure 7-8 Cumulative Oil Recovery Performance of Horizontal Well Vs Two DWS Vertical Wells ......................................................................................................... 162 Figure 7-9 Watercut Performance of Horizontal Well and Two Vertical DWS Wells .. 164 Figure B-1 Capillary Pressure Plot for LVT Well .......................................................... 189 Figure B-2 Relative Permeability Curve for LVT Well Studies..................................... 190 Figure B-3 LVT Well Reservoir Grid Model ................................................................. 191 Figure B-4 Watercut Reduction Performance of DWS Technology .............................. 191 Figure B-5 Watercut Forecast of LVT Well without Capillary Pressure ....................... 192 Figure B-6 Watercut Reduction Forecast with and without Capillary Pressure............. 193 Figure B-7 Watercut Performance with High Rate at Top Completion ......................... 195 Figure B-8 Watercut Performance of DWS in a Developing Cone Scenario................. 196

xi

NOMENCLATURE a A B Bo Bw Cn D d dp dX dr F f f Ms f MRI fo ftp fT fw fw fw,INV fw,lim exponent [dimensionless] cross-sectional area [L2] constant [dimensionless] Oil formation volume factor [L3/L3] water formation volume factor [L3/L3] constant [dimensionless] main flow pipe inner diameter [L] pipe diameter [L] perforation diameter (siz) [L] change in height of water cone, [L] change in circumference of cone arc, [L] dimensionless function friction factor (Fanning) [fraction] Moody pipe static friction factor [fraction] Moody perforated pipe friction factor with influx [fraction] fractional flow of oil [fraction] two-phase flow friction factor [dimensionless] total friction factor [dimensionless] wall friction factor [dimensionless] fractional flow of water [fraction] fractional flow of water at the inversion point [fraction] limiting water fraction at which the relative viscosity is 100 xii

gc H h ht ho hcb hct hp J(Sw) K k ke kg kh ko (ko)wc kv kr krnw kro kow (kw)or L

acceleration due to gravity [L/T2] height of transition zone above the oil-water contact [L] formation thickness [L] height of total formation thickness, [L] average oil column thickness, [L] height of completion bottom from top of formation, [L] height of completion from top of formation, [L] perforation thickness, [L] Leverett capillary function [dimensionless] Richardson constant [dimensionless] absolute permeability, [L2] effective permeability [L2] effective permeability to gas[L2] horizontal permeability, [L2] effective permeability to oil [L2] effective permeability to oil at connate water saturation [L2] vertical permeability, [L2] relative permeability [fraction] non-wetting phase relative permeability [fraction] relative permeability to oil [fraction] effective permeability to water [L2] effective permeability to water at residual oil saturation [L2] length of horizontal well [L] xiii

NRe NRem P Pc Pct pe pw PHydrostatic Preservoir Pwellhead DP PIH PIV Q Qani. Qiso. Qc QcH Qcv QD qi Qo Q(o,w)

Reynolds number [dimensionless] mixture Reynolds number [dimensionless] pressure [M/LT2] capillary pressure [M/LT2] threshold capillary pressure [M/LT2] boundary pressure [M/LT2] wellbore pressure [M/LT2] hydrostatic pressure [M/LT2] reservoir pressure [M/LT2] wellhead Pressure [M/LT2] differential pressure [M/LT2] productivity index of horizontal well [L4T/M] productivity index of vertical well [L4T/M] main flow rate, [L3/T] anisotropic reservoir flow rate, [L3] isotropic reservoir flow rate, [L3] critical rate, [L3/T] critical rate of horizontal well, [L3/T] critical rate of vertical well, [L3/T] dimensionless flow rate influx through perforation [L3/T] oil flow rate, [L3/T] oil or water flow rate [L3/T] xiv

qw r rc reH reV re rw Rw rwH rwV (S*w)drainage (S*w)imb Sr Sw Swc Swc(drain) Swc(imb) Swc(imb) Swc(scan) Swi tBT tD tDBT

water flow rate [L3/T] effective radius, [L] radius of capillary (L) drainage radius of horizontal well, (L) drainage radius of vertical well, (L) well drainage radius, [L] wellbore radius, [L] wellbore flow resistance [dimensionless] wellbore radius of horizontal well, (L) wellbore radius of vertical well, (L) drainage case water saturation [fraction] imbibition case water saturation [fraction] residual saturation of the non-wetting phase [fraction] water saturation [fraction] connate water saturation [fraction] drainage connate water saturation [fraction] imbibition connate water saturation [fraction] imbibition connate water saturation [fraction] scanning curve connate water saturation [fraction] irreducible water saturation [fraction] time-to-water break through, [T] dimensionless time dimensionless time-to-water break through, [T] xv

(t D )BJ BT

dimensionless breakthrough time, (Bournazel & Jeanson) dimensionless breakthrough time, (Sobocinski & Cornelius) velocity [L/T] total velocity [L/T] mixture superficial velocity [L/T] oil superficial velocity [L/T] water superficial velocity [L/T] location of a constant pressure boundary (L) cone height, [L] change in cone height, [L] dimensionless cone height

(t D )SC BT
u ut vm vso vsw XA Z dz ZD

Greek Letters: a adip a a b D d e F friction factor exponent [dimensionless] dip angle [degree] dimensionless factor for water breakthrough transformed dimensionless factor constant [dimensionless] Difference or increment increment pipe roughness [L] Potential or datum corrected pressure [M/LT2] xvi

dF f j m mm mo mw mr p q ro rw s sm so swo sw

change in potential, [M/LT2] porosity, [fraction] perforation density [L-1] viscosity [M/LT] mixture viscosity [M/LT] oil viscosity [M/LT] water viscosity [M/LT] relative viscosity [dimensionless] constant = 3.142 contact angle [degree] oil density, [M/L3] water density, [M/L3] surface tension [M/T2] mixture surface tension [M/T2] oil surface tension [M/T2] water-oil interfacial tension [M/T2] water surface tension [M/T2]

xvii

ABBREVIATIONS Anal.: Avg: bbl: BLPD: BOPD: Bt.: btm: BWS CAP: Compln: Conv.: DBD: DL: DVWCON: DWS: DWSVW: E&P: EPA: FWL: HW: HWCON: HWWFIP: Analytical Average barrel Barrel Liquid Per Day Barrel Oil Per Day Break Through bottom Bilateral Water Sink capillary Completion Conventional Dual Bore Deflector Detection Limit Dual Vertical Wells CONventional Downhole Water Sink Downhole Water Sink Vertical Well Exploration and Production Environmental Protection Agency Free Water Level Horizontal Well Horizontal Well CONventional Horizontal Well With Friction and Infinite Porosity xviii

IMPES: IPW: JIP: LSU: mD: Mod.: ND: NPDES: OC: OCS: OOIP: OPI: O/W OWC: PAH: PI: ppb: PV: RCRA: RF: Sim.: Stb: SVWCON:

IMplicit Pressure Explicit Saturation Inflow Performance Window Joint Industry Panel Louisiana State University milliDarcy Modified Not Detectable National Pollutant Discharge Elimination System Oil Cut Outer Continental Shelf Original Oil In Place Oil Productivity Index Oil in Water Oil-Water Contact PolyAromatic Hydrocarbons Productivity Index pounds per barrel Pore Volume Resource Conservation and Recovery Act Recovery Factor Simulator stock tank barrel Single Vertical Well CONventional xix

TWS UK: US: Vs: WC: WHR WOR W/O Wt.:

Tailpipe Water Sink United Kingdom United States Versus Watercut Water Hydrocarbon Ratio Water Oil Ratio Water in Oil Weighted

xx

ABSTRACT Approximately 2.5 billion dollars is spent annually to solve the problem of produced water in oil and gas wells. Downhole Water Sink (DWS) technology is one industry solution to control water coning in oil wells. DWS technology involves the segregated production of oil and water through separate completions with zonal isolation packer. However, several problems have been experienced in the application of the technology in watered-out oil wells. This study identified two factors that could aid in a better modeling of the technology in old vertical wells inclusion of capillary transition pressures and relative permeability hysteresis. It also identified a pressure enhanced capillary transition zone enlargement around the wellbore as responsible for the concurrent production of contaminated fluid from both completions. Another widely recommended industry solution to the problem of produced water is horizontal well technology. However, field reports indicates that water breakthrough into horizontal wells could be quite dramatic and tend to erode the merit of high deliverability. This study analyzed the problem of water cresting in horizontal wells and developed a generalized compound friction pressure loss relation for horizontal wells and pipes. The new relation includes factors such as perforations, oil-water emulsions, and radial influx of fluid into the wellbore as well as phase inversion. It also shows the results of the application of this relation in the modeling of water cresting in horizontal wells subject to bottom water drive. These results reveal an asymmetrical distribution of water influx skewed toward the heel in line with field observations. xxi

Finally, the study presents two innovative dual-completion concepts for controlling water cresting in horizontal wells adapting the principles of the Downhole Water Sink technology. The results of the initial studies shows that oil recovery could be improved by as much as 7 percent over conventional horizontal wells.

xxii

CHAPTER 1 INTRODUCTION 1.1 Background and Purpose In 1998, produced water from 40 counties in the state of Colorado was estimated as 220.6 million barrels in comparison to only about 22.46 million barrel of oil produced over the same period, while 2001 production stood at 360 million barrels of water against 25.5 million barrels of crude oil (www.oil-gas.state.co.us/statistics, 1998-2001). In 1993,the U.S. produced about 2.5 billion barrels of crude oil as against about 25 billion barrels of associated unwanted water for the same year (Wanty,1997,). Similar problem of produced waters exists in the North Sea and in the Niger Delta Basin, as well as in the Middle East. The extent to which produced water problem is a big nuisance in the oil and gas industry is reflected in the fact that unwanted production of water has been estimated to cost the petroleum industry about $45 billion a year (www.halliburton.com, 2001). These costs according to Halliburton include the expense to lift, dispose or re-inject produced waters, as well as the capital investment in surface facility construction and to address other environmental concerns. In fact Kimbrell (2001) asserted that, Produced water is a fact of life in Louisiana. The largest volume of waste associated with oil and gas production operations in Louisiana, as well as nationally, is produced water. The amounts of produced water are overwhelming compared to the amounts of hydrocarbons produced. In 1993, over 1.2 billion barrels of produced water was generated compared to less than 200 million barrels of oil and condensate and a little over 200 million BOE (barrel oil equivalent) of gas

produced in the same period. From 1990 to 1993, the statewide water-hydrocarbon ratio (WHR) averaged approximately 3.2 (Kimbrell, 2001). Several research efforts have been directed at understanding the mechanism of water coning in vertical oil wells or water cresting in horizontal wells. These research efforts have led to the calculations of the critical oil rate to avoid water coning; the time to water breakthrough at production rates above the critical rates, and prediction of watercut behavior with time after water breakthrough. Unfortunately, the proffered critical oil rates are usually uneconomic for any practical purposes. The need for a practical solution that allows operators to produce oil at reasonable economic rate and yet live pleasantly with water production issues is paramount. This need requires advancement of Petroleum Engineering practice from sheer prediction of what is happening in the reservoir to solving the problem of cash flow, which is traditionally tied to oil production rate and ultimate recovery. This research presents a solution to this need. The purpose of this research is to evaluate the performance of the DWS technology in old watered out vertical wells. The production of disposable oil-free water at the water sink completion in these old vertical oil wells with severe water coning history is also addressed. Further, the research reviews the problem of water cresting in horizontal wells and investigates the feasibility of extending the DWS principle to controlling water cresting in horizontal wells and to recover some of the bypassed oil. Thus, the research represents an advancement of Petroleum Engineering practice from merely predicting water influx parameters to controlling water influx into producing oil wells at economic production rates and at the same time recovers some of the bypassed 2

oil.

It involves the concurrent production of the oil and water through separate

completions with zonal isolation packer. The technology has been tested in vertical wells and the principles are, here, recommended for the control of water cresting in horizontal wells. At the same time, this research work reviews the performance of this concurrent fluid production technology in watered-out vertical oil wells with severe water coning history. 1.2 Statement of Research Problem The two recurring problems of critical concerns that require remedial solution and re-completion in the oil and gas industry are: (1) water shut-off repairs in bottom or edge aquifer support reservoirs; and (2) sand control repairs in unconsolidated deltaic deposits. A very significant development has been made in the area of sand production control in vertical and deviated wells. It has been established that effective planning and execution of well completions and well management could potentially eliminate the irritation of sand production in the industry. On the other hand, the problems associated with produced water, which include the surface-handling problem, and the ensuing bypass of recoverable oil still remains a challenge to the industry. Oil reservoirs with bottom water drive have the advantage of a high oil recovery due to energy support from the aquifer. However, the encroachment of the water into the producing oil well caused by the pressure drawdown around the wellbore creates problems over time. These problems include water handling and bypass oil. The production of oil and/or gas from a well causes pressure around the wellbore to drop. This creates a pressure gradient around the wellbore vicinity. A counteracting gravitational force, due to the difference between the oil density and water density, 3

causes the oil-water contact interface to remain stable. However, at a certain production rate, the viscous forces due to the pressure gradients around the wellbore become greater than the gravitational forces making the oil-water interface unstable. Consequently, the contact rises up until water breakthrough into the producing well. The shape and nature of the cone development depends on several factors ranging from production rate, mobility ratio, horizontal and vertical permeability, well penetration as well as the viscous forces. After the initial water breakthrough in the well, water production increases rapidly and oil production decreases. Studies indicate that there is a critical rate of oil production that mitigates water coning into the well usually designated Qcrit (Joshi, 1991, Bournazel and Jeanson, 1971, Chappellar and Hirasaki, 1975). However, this rate is generally too low for any economic significance. Joshi (1991) presented several calculations of critical oil rate for vertical and horizontal wells using different correlations and obtained results ranging from 57 stb/day to 84 stb/day for vertical wells and from 190 stb/day to 470 stb/day for a 1640 ft long horizontal well. The capital outlay of oil production operation requires high oil production rate to ensure early capital recovery. Thus operators have to come to grips with the problem of producing water and oil. Over the passed two decades, horizontal drilling technology has been widely recommended as a solution to the development of these reservoirs experiencing high water coning problems. While vertical wells act like point source concentrating all the pressure drawdown around the bottom of the wellbore, horizontal wells act more like a line sink and so distribute the pressure drawdown over the entire length of the wellbore. 4

The result is reduced pressure drawdown around the wellbore and delayed water influx (cresting) compared to vertical wells. The facts, however, remain that the reduced pressure drawdown is not a total solution to water coning. However, recent evidence indicates that horizontal wells are, themselves, not free the problem of water production. Horizontal wells, by their nature and geometry, have inherent problems. One such problem is that the increased contact with the reservoir, which is an advantage in terms of fluid production rates, actually becomes a disadvantage when water breaks through into the wellbore. Thus, horizontal well water cresting has increasingly become a concern to the industry. Existing solutions to the problem of water coning in vertical wells or water cresting in horizontal wells have either recommended uneconomic oil production rates or have not adequately addressed the recovery of bypassed oil due to water influx. Further, other technologies commonly employed to minimize water production problems have not solved the problem of contaminated water handling or addressed the problem of bypassed oil in the reservoir, successfully. Conversely, the Downhole Water Sink (DWS) technology have proved to be very successful at recovering some of the bypassed oil while encouraging high oil production rates with reduced water handling in vertical wells. 1.3 Significance and Contribution of this Research The significance of this research is five folds. First, the research presents an advancement of the modeling of the DWS in vertical wells using existing commercial numerical simulators with all the capabilities of modeling several complex reservoir phenomena such as capillary pressure transition zone and relative permeability hysteresis. 5

Second this research presents a design and optimization criteria for clean drainage water production from the water sink completion in old oil wells due the presence of the capillary transition zone.

Third, the research presents a more robust friction pressure loss relation for horizontal wells and pipes (generalized friction pressure loss relation) incorporating factors such as perforations and perforation density, fluid influx into the wellbore, two-phase oil-water flow effects and the formation of emulsion, as well as phase inversion from oil dominated flow to water dominated flow.

Fourth, the application of the generalized relation for friction pressure loss in horizontal wells has led to a more practical modeling of the problem of water cresting in horizontal wells showing the severity of the water crest at the heel of the well and bypassed oil at the toe.

Finally, the research presents an extension of the downhole water sink technology principle to the control of water cresting in horizontal wells evolving two variants with the acronyms: (1) Tail-pipe Water Sink Technology (TWS); and (2) Bilateral Water Sink (BWS) technology.

1.4

Research Method and Approach This research used existing commercial numerical simulator (Eclipse 100/200)

and a radial grid model to study the introduction of complexities such as capillary pressure transition zone and relative permeability hysteresis in DWS technology in watered out vertical wells. It assumes a steady-state process with adequate water influx from the aquifer. Thus, the first step was to evaluate and adopt an adequate steady state model for the commercial simulator. Two options were found adequate for most of the 6

studies; fluid re-injection and the introduction of an infinitely large pore space at the bottom of the reservoir to simulate a bottom-water-drive reservoir. For most of the studies on DWS performance in watered-out wells, the Inflow Performance Window (IPW) was used as an adequate evaluation tool. The IPW is a plot of fluid withdrawal rate at the top oil zone completion on the x-axis and fluid withdrawal rates at the bottom water zone completion on the y-axis. For horizontal wells, a Cartesian grid model with finer grids around the wellbore was adopted in the numerical simulator to model water influx into the wellbore. However, due to limitations of existing commercial simulators in modeling horizontal well production, the research focused, in part, on the development of an effective tool for evaluating friction pressure loss in horizontal wells. The research developed a generalized compound friction pressure loss relation for horizontal wells incorporating all of the variables of perforations, water influx, two-phase flow, and flow regimes. An equivalent pipe roughness function is then calculated from a re-arranged Colebrook friction pressure loss relation. This equivalent pipe roughness function is input into the commercial numerical simulator to give the same pressure loss effect in the wellbore. This new compound friction pressure loss relation is then applied to model the performance of water cresting in horizontal wells. The application of the generalized friction pressure loss relation showed an increase in the water cresting around the heel of the well and bypassed oil at the toe. Thus, a fluid flow re-distribution technique that focuses on reducing the water influx at the heel has been developed to counter the crest development in this region and reduce the severity of oil bypass at the toe. 7

1.5

Work Program Logic This study is divided into eight chapters. The introduction chapter presents an

overview of the problem of water influx into producing oil wells and the enormity of the problems associated with this unwanted phenomenon. It also presents a concise statement of the origin of water influx and ends with a presentation of the relevance of this study and approach. Chapter two presents a literature review of scientific research into the problem of water coning and water cresting and industry solutions to mitigating these dual phenomena. Chapter three gives an overview of the application of downhole water sink technology to the control of water coning in vertical wells with specific focus on application in watered-out wells and reservoirs with capillary transition zone. Chapter four then takes a look at the environmental attraction of the technology in these old wells in terms of production of disposable oil-free water at the sink. Chapters five and six are ground breaking areas of this research presenting a more robust tool for evaluating compound friction pressure loss in horizontal wells considering two phase oil-water flow and the concept of phase inversion. The friction pressure loss model presented in chapter five is easily adaptable to three-phase oil-water-gas flow by use of appropriate liquid-gas viscosity. Chapter six ends with the recommendation of two techniques for controlling water cresting in horizontal wells. Chapter seven reviews current industry practice of comparing the performance of a single vertical well with a horizontal well concluding that it does not compare apple for apple. It also presents a comparative study of oil recovery with DWS vertical wells and

horizontal wells in bottom water drive reservoirs. Chapter eight provides conclusions from this research work, including recommendations for future research.

CHAPTER 2 LITERATURE REVIEW A survey of literature shows that tremendous amount of research work has been done ranging from experimental studies to analytical and numerical simulation studies in order to understand and predict water coning and cresting in vertical and horizontal wells respectively. Several correlations have been developed as a result of these research efforts to predict the time for water breakthrough, the critical oil rate to avoid water breakthrough and the post breakthrough behavior of the water influx at supercritical rates of production. Several research efforts and solutions have also been developed to reduce the level of severity of the post water breakthrough performance of the well. An overview of research findings on the subject of water coning/cresting will be presented in this chapter. 2.1 Water Coning in Vertical Wells Production of oil through a well that partially penetrates an oil reservoir underlain by water creates a differential pressure between the wellbore and the reservoir. This differential pressure causes the oil/water interface to deform into a cone shape (Figure 21). As production rate is increased, the cone height above the initial oil/water contact (OWC) also increases until water breaks through into the wellbore. The breakthrough occurs when the cone shaped profile becomes unstable due to the high-pressure drawdown around the wellbore. Mathematical models have been developed to predict the performance of oil wells with water coning problems after water breakthrough. Muskat and Wyckoff (1935) and 10

Muskat, (1949) observed that coning is a rate-sensitive phenomenon, which develops only after certain equilibrium conditions are unbalanced by increasing the pressure differential beyond critical limits. The equilibrium condition necessary to avoid water breakthrough requires that the vertical pressure gradient in the oil zone caused by viscous forces be balanced by the differential-gravity gradient in the underlying water zone. When this occurs, the water cone is termed stable.

oil

W ater

Figure 2-1 Schematic Representation Water Coning in Vertical Wells Muskat and Wyckoff (1935) also determined the critical oil rate analytically solving Laplace equation for single-phase flow and for a partially penetrating well. However, Wheatly (1985) observed that Muskat and Wyckoff relation over predicts the 11

critical oil production rate because it did not account for the presence of the cone when calculating the pressure distribution in the reservoir. In effect, they assumed that the pressure distribution in the reservoir is the same with or without water coning. Wheatly, also observed that the value of the well radius does not significantly affect the critical rate of production. This is contrary to Schols (1972) observation. The difference in observation could be due to the fact that Wheatly studied a partially penetrating well while Schol studied a fully penetrating well. Guo and Lee (1993) demonstrated that the existence of the unstable water cone depends on the vertical pressure gradient beneath the wellbore. If the vertical pressure gradient is higher than the hydrostatic pressure gradient of the water, the unstable water cone can be observed and the associated critical oil rate exists. That is, the upward dynamic force resulting from wellbore pressure drawdown causes the bottom of the oil zone to rise to a height where the dynamic force due to pressure difference is balanced by the weight of water beneath this point. As the radial distance from the wellbore increases, pressure drawdown decreases and so does the cone height above the original oil-watercontact creating a locus of cone-shaped balance-points called the water cone as shown in Figure 2-1. An important result of Guo and Lees study is that the maximum water-free production rate (critical rate) does not occur at zero wellbore penetration as one might expect from intuition, but at a wellbore penetration about one-third the total oil-zone thickness for an isotropic reservoir. Archer and Wall (1986) observed that the maximum oil potential gradient possible for a stable cone could be written in terms of the cone height, z, and the potential, F, (or datum corrected pressure). 12

Using the illustration in Figure 2.2, they concluded that at the critical oil rate, the following condition holds:
dF dF = dr dz

2-1

Where, z, is the vertical height of the cone above the OWC and, r, is the length of the cone arc along its locus. In Figure 2-2a, the length of the cone arc is greater than the height of the cone above the OWC (dF/dz > dF/dr) and so the cone is stable (no water breakthrough). In Figure 2-2b, however, the length of the cone arc, dr, is equal to the height of the cone, dz, and dF/dz = dF/dr and the cone becomes unstable. Summarizing, in field units, Archer and Wall concluded that the viscous to gravity balance equation for cone height is given by:
z= DF 0.433( r w - r o )

2-2

(where r, is in specific gravity, z, is the cone height in ft, and F is in psi)

(a) Low Rate: Stable

F2
dr dz Water Oil OWC

(b) Critical Rate

F
dz dr

Figure 2-2 Cone Stability Analysis 13

Chappellear and Hirasaki (1975), developed an expression for the critical oil production rate as follows:

Qc =

2pht k h k ro Drg (ho - hcb ) 887.2m o ln r 'Bo

2-3

ho - hcb r ' = 4ho k h / kv h -h o ct

2-4

ln r ' =

ln[re /(rw + r ' )] 1 2 2 2 1 - (rw + r ' ) / re

2-5

(where r, re, rw are in ft, ho, hcb, hct, are in ft, kh, kv are in mD, mo is in cP, Qc is stb/day, r is in psi/ft, and Bo, Bw is in bbl/stb) Bournazel and Jeanson (1971), however, proposed a fast water coning evaluation relation for the critical oil rate for isotropic reservoir as:
5.14 E (-05) * k h ho Drg (1 - h p / ho ) mo
2

Qc =

2-6

And for anisotropic reservoir, they simply recommended that: Qani. = Qiso.* (kv/kh). (Equations 2-6 to 2-11, are in field units, Q = stb/day; kh = mD; Dr = psi/ft; g = ft/s2; mo = cP; and ho, hp = ft) For time-to-water breakthrough evaluation, tBT, several researchers have also proposed different correlations. Sobocinski and Cornelius (1965) as well as Bournazel and Jeanson (1971) proposed three-step correlations using dimensional analysis. The correlations are almost the same except for the evaluation of the dimensionless time. For the purpose of completeness, the correlations are listed below:
14

 h 3.69 E - 04 * ( r w - r o )kh ho2 1 - p ho ZD = (in field units) m oQo Bo

2-7

(t )

SC D BT

2 Z 16 + 7 Z D - 3Z D = (Sobocinski & Cornelius, 1965) 4 7 - 2Z D

2-8

ZD (t D ) BJ BT = (Bournazel & Jeanson, 1971) 3 - 0.7 Z D

2-9

Thus the time to water breakthrough is evaluated with Equation 2.10 using any of the dimensionless time evaluated from either Equation 2.8 or 2.9.
t BT =
m ofh(t D ) 1.645 E - 04 * ( r w - r o )kv (1 + M a ' )

2-10

a', = 0.5 for M < 1.0; and 0.6 for 1 < M < 10.0
M = m o (k w ) or m w (k o ) wc

2-11

M is the water-oil mobility ratio. (kw)or is the effective permeability to water at the residual oil saturation and (ko)wc is the effective permeability to oil at the connate water saturation (Joshi, 1991). Several significant variations exist in the results of the numerous studies on water coning predictions. These variations arise mainly due to the approximating assumptions used to simplify the complex mathematical solutions of the two- or three-phase flow in porous media. Joshi (1991) evaluated the results of the various correlations for critical oil rate and time to water breakthrough. Variations in the magnitude of 90 percent were recorded for the critical rate and as high as 140 percent for the time-to-water breakthrough (Joshi, 1991).
15

However, the common points of agreement of all the studies on vertical well water coning could be summarized as follows: 1. High pressure gradient/high fluid velocity (production rate) stimulates water coning. 2. Completion at the topmost part of the oil zone (preferably the top 45 percent) away from the oil-water-contact delays the time to water breakthrough. 3. Critical oil rates for water-free oil productions are typically uneconomical for all practical purposes. 4. After water breakthrough in the wellbore, water production increases rapidly and significant resources and efforts have to be put into water disposal and oil/water separation facilities. 5. The development of the water cone creates bypass oil problem and consequently shortens the producing life of the well. 6. A corollary is that more money is spent on infill drilling to recover bypassed oil reducing the economic gains of the field.
2.2 Other Industry Solutions for Water Coning in Vertical Wells

Several practical solutions have been developed to try to minimize the level of severity of this water coning in vertical wells. The approach, basically, is to delay the time to water breakthrough such as increasing the distance between the bottom perforation and the original oil water contact or minimize the amount of water in the wellbore thus reducing the hydrostatic head. Some of the methods are briefly described below.

16

2.2.1 Perforation Squeeze-off and Re-completion

In cases where the reservoir deposition is such that shale barriers are inter-bedded with the sandstones as in laminated sands, the shale barriers could form effective seal/separation between the sand layers. High permeable sand layers in contact with the water zone are often times responsible for the high water influx and these could be isolated by cement squeeze operation during workover to minimize the level of water production. In some cases, an entire perforation is completely squeezed off and the well re-completed higher up the structure away from the oil-water-contact (OWC). Cement squeeze operation may not be feasible or effective if adequate zonal isolation is not possible due to absence of shale barrier streaks. In such cases, application of fluid mobility modifiers could be considered in the form of conformance technology, described next.
2.2.2 Conformance Control Technology

Reservoir conformance is the process of applying various methods and technologies to a reservoir or wellbore to reduce or control unwanted water or gas production so that recovery efforts are efficiently enhanced and operator profitability improved (www.halliburton.com, 2001, Matthews, et.al., 1996). Several water production control chemicals have been developed for different applications. In some cases, matrix fracture sealants are injected into the formation, if the production of unwanted water is traceable to fracture paths. Other applications include selective relative permeability modifier that is capable of reducing relative permeability to water without affecting relative permeability to the oil and/or gel treatment. While several publications have appeared giving case histories of field applications of these technologies, the long17

term effect on reservoir properties and overall well performance remains a controversy to industry operators. Ehlig-Economides et. al (1996) observed gel treatment to build a coning barrier has been found to be ineffective and uneconomical in certain fields.
2.2.3 Downhole Oil-Water Separation Technology

Another industry solution to the problem of unwanted water production is the development of Downhole Oil-Water Separation technology. This technology uses hydrocyclone separators and special design downhole pumps installed in the completion (production) string to separate the oil/water mixture within the wellbore. The elimination or reduction of water hydrostatic column by this method means more energy is available to the reservoir in terms of drawdown pressure. The result is increase in oil production rate. To effectively understand the concept of downhole water separation technology, the basic nodal analysis equation is listed as follows:
Pwellbore = Preservoir - DP = PHydrostati c + Pwellhead

2-12

The above equation rearranged gives:

DP = Preservoir - Pwellhead - PHydrostati c

2-13

Assuming a constant reservoir pressure and constant wellhead pressure, Equation 2.13 indicates that a reduction in the hydrostatic pressure gradient in the wellbore translates to an increase in the drawdown pressure in the reservoir. Following Darcy equation (2.14), the increased pressure drawdown translates to an increase in oil production, assuming all other reservoir properties remains constant.
Q= 0.00708kh ( DP ) mB ln(re / rw )

2-14

18

Q is in stb/day; k is in mD; h is in ft; Dp is in psi; m is in cP; re, rw are in ft, and B is bbl/stb. The technology provides reduced surface water handling. The fundamental problem of water interference with oil production within the reservoir creating bypass oil still remains unresolved with this technology. In other words, the problem of bypassed oil by the water development has not been solved by this technology.
2.2.4 Total Penetration Approach

In this approach the perforation interval is extended to cover the entire oil zone and into the bottom water zone. The aim is to maintain radial flow of fluid and so avoid cone development and the attendant oil bypass. Production of water starts immediately and so requires increased water handling facilities. As production increases, over time, the tendency for cone development is inevitable (Ehlig-Economides, et. al., 1996). Also, the co-mingled production of high water volume and oil in one string could create unwanted environmental problem caused by the disposal of the contaminated water. All or most of the above solutions addresses only one aspect of the two-prong problem of water coning namely; (1) increased watercut and water handling problems and; (2) bypassed oil in the reservoir as a result of water coning/cresting around the wellbore.
2.3 Water Coning Control with Dual Completions in vertical wells

Pirson and Mehta (1967) presented the results of studies performed using numerical simulators. The studies investigated several solutions to the problem of water coning in vertical wells. One of the solutions is the re-injection of produced oil into the reservoir below the oil zone perforations to suppress the development of the cone. This 19

technique known as the Oil Doublet Model was not attractive economically. Another option investigated by Pirson and Mehta is the selective production of oil and water from their respective zones using dual completions. They concluded that, though the approach reduces the growth of the cone, it gives the same overall produced water-oil ratio at all times. Widmyer (1955) patented a water coning control technique in which he proposed a dual completion string to separately produce the oil and water zones and thus counter the cone development. He suggested perforating the top and bottom of the oil zone and producing each perforation with a production string. Considering the cost of the dual completion string, Driscoll (1972) suggested a variant of the dual perforation technique. He suggested two perforations one in the oil zone and one in the water zone below the original oil-water contact. Fluid from both perforations is to be co-mingled in one production string with the possible use of a packer and adjustable flow choke to control oil/water rates. The demerit of this approach is the reduction in oil rate as a result of the increased hydrostatic head of the co-mingled fluid. The problem of separation of the produced fluid and disposal of the contaminated water makes this technology not so different from the conventional completion. It, however, solves the problem of bypass oil in the reservoir. Ehlig-Economides et. al (1996) observed that the concept of critical rate is a misnomer as water is bound to be produced in any reservoir with strong bottom water drive. They also observed that total penetration and dual penetration method of completion yields the most of oil production and recovery but at a cost of handling high rate and volume of water production.

20

In the dual completion technology with downhole water sink (DWS), the top completion is placed as high as possible, within the top 20 percent of the oil zone and the second perforation placed at some depth below the oil water contact. As will be seen in chapter 4, the location of this bottom completion (the water sink) is a very important determinant of the environmental and production efficiency of the technology. Fisher et. al (1970) used a numerical simulator to study the application of the dual selective production completion in the Bellshill lake, Blairmore Pool in Canada and concluded that it can reduce the water cone development and even eliminate it completely in certain cases. Wojtanowicz, Xu and Bassiouni (1991) and Wojtanowicz and Shirman (1997, 1998), conducted theoretical studies on the hydrodynamics of water coning control using a production well with tailpipe water sink (DWS). The model used flow potential distribution generated by two constant-terminal-rate sinks located between the two linear boundaries of the oil-water contact (OWC) and no-flow top boundary in the oil zone and constant-pressure outer radial boundary representing a steady state process.

Water Oil PCP 18 3 Oil Water 7 5 64 OWC

Figure 2-3 Schematic of the dual completion for the Nebo Hemphill project

21

The studies indicated that oil recovery could be improved by efficient water production from the bottom completion. It also shows that the bottom completion can produce oil-free water. The first field application of the technology was performed in 1994 at the Nebo-Hemphill Field located in LaSalle Parish, in Louisiana (Swisher, 1995). The result of the field application indicated that the technology did not only prevent water coning, but was actually capable of reversing a water cone after breakthrough. The well performance showed a remarkable improvement in oil production rate and total recovery over conventional wells. Bowlin et. al (1997) reported the result of the installation of the technology by Texaco Inc. as the in-situ gravity segregation in the Kern County, California. Shirman and Wojtanowicz (1998) reported that more oil could be produced with less water in the top completion while producing oil-free water in the lower completion. A limitation of these studies is that they lack the ability to analytically represent the oil/water transition zone (capillary transition) and related effects of water saturation changes on oil contamination in the produced water. Thus they could not effectively model the performance of old oil wells with severe water coning history.
2.4 Capillary Pressure and Relative Permeability Hysteresis in DWS Water Coning Control

Researchers have largely ignored capillary pressure transition and relative permeability hysteresis effects on the phenomenon of water coning. For most of the analytical studies, incorporating these effects complicates the mathematical solutions. Letkan and Ridings (1970) and Mungun (1975) included the effect of capillary pressures in their numerical coning model but did not report on the effect of the inclusion or non22

inclusion in the models. Reed (1984) investigated the mechanism of water production in a high permeability sandstone reservoir with a capillary transition zone comparable to the formation thickness for a conventional well. He concluded that watercut is principally dependent on the average fractional flow of water and thus the average water saturation in the drainage area of the well. Effectively, the average water saturation is governed by the amount of mobile water and hence the height of the transition zone. Miller and Rogers (1973) studied the performance of oil wells in bottom water drive reservoirs. They observed that the vertical permeability determines, to a great extent, the time to water breakthrough and the rate of decline of oil production after water breakthrough. They also concluded that the shape of the relative permeability curve has no significant effect on the performance of the wells. Kurban (1999) evaluated the performance of DWS wells in the presence of capillary transition zone and relative permeability hysteresis and concluded that both concepts have significant effect on the performance of the technology. He concluded that the presence of a capillary transition zone increases the amount of produced water. Kurban (1999) did not report on the mechanism of the effect of these parameters on DWS well performance.
2.5 Horizontal Well Technology

Several researchers have recommended horizontal well technology as a solution for the development of reservoirs with water coning problems. While vertical wells act like point source concentrating all the pressure drawdown around the bottom of the wellbore, horizontal wells act more like a line sink and so distribute the pressure drawdown over the entire length of the wellbore. The result is reduced pressure drawdown around the wellbore. The nature of fluid influx into a horizontal well remains 23

a challenge to analysts. Researchers have presented three models to explain the flow pattern in a horizontal wellbore. They are: infinite conductivity, uniform flux and finite conductivity. Infinite conductivity assumes there is no pressure loss in the wellbore. Uniform flux assumes that the influx of fluid into the wellbore from the reservoir in constant along the length of the horizontal well. The finite conductivity assumption is more encompassing as it incorporates the effect of friction pressure loss in the wellbore and changes in the distribution of fluid flux along the wellbore. The most controversial and least accepted of all three is the uniform flux assumption. Babu and Odeh (1988) developed a generalized expression for evaluating the productivity of horizontal wells assuming uniform influx of fluid along the wellbore. Several authors, Gilman (1991); Peaceman (1991); Suprunowicz (1992, 1993) reacted to this paper questioning the validity of the uniform flux model. They argued that the productivity calculated from the uniform flux model is an underestimation. Proponents of the infinite conductivity model argue that the effect of friction pressure loss along the wellbore is small compared to the reservoir pressure drawdown and so is negligible. While this may be true under certain conditions, such as low permeability large wellbore conditions, the ratio of wellbore friction pressure loss to reservoir pressure drawdown could be quite significant in some other conditions that its effects cannot be neglected. Research efforts have led to the development of mathematical equations for the evaluation of the performance of horizontal wells. Geiger (1984) developed a correlation for evaluating the productivity index of single-phase oil production or the pre-water breakthrough oil production rate in an isotropic reservoir (Equation 2-15). Geiger (1984)

24

also gave the productivity index of a vertical well (Equation 2-16) in the same paper in order to compare the performance of vertical and horizontal wells.

' k L 1 PI H = 0.0145* m 1 + 1 - (L / 2r ) 2 eH ( L / h) ln L /(2reH )

] + ln

h 2p * r wH

2-15

2pkh 1 PI v = m ln rev rwv

of vertical and horizontal wells of length, L, in an isotropic reservoir as:
reV ln r PI H wV = 2 PIV 1+ 1- L 2reH h h ln + ln L L 2prwH 2reH

2-16

Consequently, Geiger (1984) developed a correlation to compare the performance

2-17

(where, reH and reV are the respective drainage radius of a horizontal and vertical well, rwH and rwV are the respective wellbore radius of the horizontal and vertical wells, L is the length of the horizontal well and h is the reservoir net thickness). Equation 2-17 has been widely accepted by reservoir engineers in evaluating the performance of a horizontal well and the need or lack of it in field development planning.

25

2.6

Horizontal Well and Water Influx Control

The reduced pressure loss in the wellbore causes a delay in water influx in horizontal wells compared to vertical wells. Another advantage of horizontal wells frequently advanced by analysts is that the standoff from the bottom water can be maximized with horizontal well than for vertical wells. In studying the effect of coning toward vertical and horizontal wells in anisotropic formation, Chaperon (1986) developed a correlation for the prediction of critical oil rate in horizontal wells as:

QcH = 4.89 E - 04 *

kH h 2 L DP m B F XA o o

2-18

Where 1 a < 70 and 2XA < 4L

a" =

X A kV kH h

2-19

For the dimensionless function F, Chaperon generated a table of data for different scenarios. However, Joshi (1991) developed a correlation, with a maximum error of 44 percent for the function F as follows: F = 3.9624955 + 0.0616438(a) 0.000540(a)2
2-20

Like Muskat, Chaperon neglected the influence of the cone shape on the flow pattern. Geiger (1989) and Karcher et. al. (1986) also presented useful correlations for evaluating the critical oil rate of horizontal wells in a steady-state process. Papatzacos et. al. (1989) developed an analytical method, assuming a moving boundary approach, for predicting cone evolution toward the horizontal well in the transient state. To develop a correlation for the time to water breakthrough, they assumed a constant pressure on the

26

moving boundary. The result of their analysis yielded an expression for the dimensionless time to water breakthrough as: tDBT = 1/6QD
2-21

where, QD, the dimensionless flow rate is given by Equation 2-22 and the dimensionless time to water break through is given by Equation 2-23: QD = 39.12 and
t DBT = 0.2284kV (r w - r o ) t BT hfm o
2-23 m oQo Bo hL( r w - r o ) kV k H 2-22

Papatzacos (1991) later developed correlations for critical rate calculations in horizontal wells using the gravity equilibrium assumption on the moving boundary. Both solutions refer to an infinite acting reservoir. Yang and Wattenberger (1991) developed several correlations from the result of numerical simulation studies for predicting water-oil-ratio (WOR) as well as the critical rate and time to water breakthrough for horizontal wells considering a closed outer boundary reservoir in transient state. To reduce the complexity of the mathematical solutions, all the analytical equations presented in Equations 2-15 through 2-23 assumed an infinite conductivity concept. In other words, they neglected the friction pressure loss in the wellbore. Nevertheless, they are useful tools in analyzing the performance of horizontal wells. Reviewing field performance of horizontal wells, Sherrad et. al. (1986), Broman et. al. (1990) and Cunningham et. al. (1992) reported on the success of the application of horizontal wells in the re-development of the Prudhoe Bay field in Alaska. Murphy

27

(1988) also reported the success of horizontal well in producing the Helder Field in Netherlands. Several other countries, Chen (1993) in Oman, Lien et. al (1990, 1991) on the Troll Field in Norway amongst others, have used the horizontal well technology to improve oil recovery from reservoirs with strong bottom water drive. Peng and Yeh (1995) also concluded that the use of horizontal wells is a proven technology for reducing coning problems and improving recovery in reservoirs underlain by water. However, most of these reports have been written quite early in the life of these fields developed with horizontal wells. Recent field evidence shows that the fact of reduced pressure drawdown is not a total solution to water influx. Horizontal wells, by their nature and geometry, have inherent problems. One such problem is that the increased contact with the reservoir, which is an advantage in terms of fluid production rates, actually becomes a disadvantage when water breaks through into the wellbore causing a very rapid increase in watercut. Like their vertical well counterpart, the typical critical oil production rates to avoid water influx to the wellbore are also too low for any economic purpose. Thus typical production rates are usually higher than the critical rates for these wells. The result is that high mobility bottom water invades the oil zone and ultimately moves toward the well. In other words, although critical rates are typically higher for horizontal wells than for vertical wells and the time to water breakthrough is also longer for horizontal wells than for vertical wells, the typical high production rates expected of horizontal wells soon creates the problem of unwanted water influx. The nature of water influx into horizontal wellbore remains a challenge to researchers. The uniform flux model assumes that bottom water rises uniformly and 28

forms a symmetrical crest along the well length. Physical evidence from production logs in horizontal wells with water cresting problem show that this is not the case. More of horizontal production (over 80 percent) occur within the first 50 percent of the well length indicating that water breakthrough will first occur at the heel and spread gradually toward the toe of the well (Lien et. al., 1991). Friction pressure loss in the wellbore have been adduced as a reason for the skewness of horizontal well flow profile toward the heel (Guo and Lee, 1993).
2.7 Friction Pressure Loss Consideration in the Horizontal Wellbore

In order to evaluate the effect of the friction pressure loss component, researchers and industry operators have been faced with another challenge of developing an adequate correlation for the friction pressure loss. Most reservoir simulation tools neglect the friction pressure loss component or at best approximate it with friction pressure loss in horizontal pipes using the Fanning friction factor or Moody friction factor (Eclipse Manual, 2000, VIP, 2000). However, several research projects conducted in the past 10 years have shown that approximating horizontal wells with horizontal pipe flow is grossly inadequate (Su and Gudmundsson, 1993, Yuan, 1997, 1999, Yula et. al., 2000). Yuan (1997) recommended that the regular pipe friction factor correlations should not be used to predict the pressure drop in a horizontal well. The pipe roughness can be much greater in horizontal wells than for regular pipe. This is quite understandable considering the effect of perforations and the fluid influx along the well length. Chaperon (1986) was the first to consider the effect of pressure drop along the wellbore. Chaperon evaluated the case of laminar flow in a pipe and considered the pipe

29

inner bore, r, to behave as a medium of permeability where the pipe bore equivalent permeability is given by:
k = 1.15 E10 * r 2

(k is in mD)

2-24

The pressure drop in the wellbore, (DP in psi) could be obtained by re-arranging the Poiseuilles law for total flow rate through capillary as in Craft and Hawkins (1991):
DP =

QBo m o L 1.30(10)10 pro4

2-25

where Q is in stb/day, L is in ft, p1, p2 are in psi and ro is in ft. For a well of length 1500 ft and ro = 0.417 ft, Bo = 1.125 bbl/stb, and mo = 1.45cP. For a flow rate of 5,000 stb/day,
DP calculated from the above relation (Equation 2-25) is 0.01 psi.

Joshi (1991) performed several calculations to try to show that friction pressure loss in the wellbore is small and, therefore, negligible. These calculations, however, used correlations developed for pressure loss in pipe flow that are considered inadequate for horizontal wellbore flow conditions. Guo et. al. (1993) observed that the line sink theoretical approximation of the horizontal well is inadequate as it assumes the infinite conductivity model. They

observed that the inclusion of wellbore friction pressure loss causes a pressure gradient in the pressure drawdown in the formation along the wellbore at the downstream end of the horizontal well. The consequence is that critical rate predicted by the link sink approximation method will be overestimated. However, they did not validate their observation. On the other hand, mathematical models used for evaluating friction pressure loss in pipe are inadequate in horizontal wells. Several parameters such as perforations 30

roughness, fluid influx into the wellbore and oil production rate affects the flow geometry of the produced liquid which in turn affects the level of pressure loss in horizontal wells. The consequence is that pressure loss in horizontal wellbore is of a higher magnitude than in conventional pipe flow. Dikken (1990) observed that laminar flow assumptions for horizontal wells are inaccurate when the radial influx of fluid is considered. He observed that the flow geometry is rather turbulent in the wellbore. In order to accommodate the increased friction pressure loss proposition, Dikken assumed that the effect of turbulence and increased pressure loss could be approximated by varying the factor, a, in the typical Blasius equation for friction pressure loss in pipe from between 0.0 to 0.15 as against 0.25 in smooth pipe flow. The Blasius equation for friction pressure loss in smooth pipes is presented in equation 2-22.
f = 0.0791 a N Re

2-26

Asheim et. al (1992) proposed a correlation for friction pressure loss in horizontal wellbore using a Blasius-type pressure loss equation. They incorporated an expression for the ratio of fluid influx to main wellbore flow. The use of the Blasius type correlation for smooth pipe neglects the in-situ pipe roughness. Typical commercial pipes used in horizontal wells have manufacturer recommended in-situ pipe roughness that poses a resistance to the flow of fluid, especially when the high production rates of horizontal wells is taken into consideration. Su and Gudmundsson (1993) developed a correlation to evaluate the extent of increased pipe roughness due to perforation and consequent increase in pressure loss. They concluded that pressure loss increment in the magnitude of 15 percent or more is 31

possible. Ozkan et. al (1993, 1999), Yuan (1994, 1997), Yuan et. al. (1998, 1999), Gonzalez-Guevara and Camacho-Velazquez (1996) and Yula (2000) performed experimental studies on the effect of pressure loss along the wellbore on the productivity of single-phase fluid in horizontal wells. They concluded that pressure loss in the wellbore of horizontal wells could be quite significant compared to that of pipe flow. McMullan and Larson (2000) also performed numerical simulation of the horizontal well productivity incorporating friction pressure loss along the wellbore. The result, using a 2-inch diameter pipe was quite significant. Unfortunately, the commercial numerical simulator employed in McMullan and Larson studies uses the Fanning friction pressure loss correlation for fluid flow in pipe, which has been considered inadequate for effective evaluation of friction pressure loss in horizontal wells. The summary of the above literature review is that friction pressure loss in horizontal wellbore cannot be adequately represented by the correlations for pressure loss in conventional pipe flow and that friction pressure loss in horizontal wells could be quite significant.
2.8 Water Cresting Control in Horizontal Wells

The subject of control of water cresting in horizontal wells is yet to receive much attention as in vertical wells. This is, perhaps, due to the fact that the technology is relatively new. However, the experience of water cresting in the horizontal wells over the past few years have prompted some level of research into the development of possible solutions. Chugbo et. al. (1989) observed that the euphoria of horizontal well technology for the development of thin oil columns in an unconsolidated sand deposit in a deltaic 32

environment may be erroneous and dangerous. Unfortunately, they did not report the result of further studies to validate this claim. Permadi et. al. (1997) investigated the horizontal well completion with a stinger introduced into the perforated pipe to redistribute the crest profile along the wellbore. They concluded that inserting a stinger about 0.25 times the length of the horizontal well provides a better pressure drawdown distribution and therefore more uniform flux along the wellbore. Asheim and Oudeman (1997) recommended an optimal perforation scheme that reduces perforations at the heel and increases perforation density toward the tail end and so redistribute the fluid influx model to possibly achieve uniform influx pattern. Ehlig-Economides (1996) proposed that a dual horizontal well completion, one in the oil zone and one in the water zone could reduce water cresting problem in horizontal wells. She did not present the results of research to validate the proposition. Renard et. al. (1997) recommended drilling of multilateral wells instead of several horizontal wells to economically increase and accelerate the overall recovery of oil in reservoirs underlain by an active bottom aquifer.
2.9 Flow Regimes of Oil-Water Flow in Horizontal Pipes/Wellbore

In order to understand the properties of the evolving mixture fluid, a better understanding of the flow regimes for oil-water flow in horizontal pipes is necessary. Six different flow regimes have been identified for oil-water flow mixtures in horizontal pipe (Trallero, et. al., 1996). The six regimes are summarized in Figure 2-4. Evaluations of the various flow patterns indicate that at a certain fractional flow of water, the flowing fluid mixture changes matrix from water-in-oil (W/O) flow to oil-in-water (O/W) flow. This point of flow regime transformation has been termed the inversion point. It is

33

characterized by a sudden change in the mixture viscosity that in turn affects the frictional pressure loss.
Oil-Water Flow (a) Segregated Flow (a-i) Stratified Flow with some mixing (a-ii) Stratified Flow (b) Dispersed Flow (b-i) Water Dominated Flow (b-i-1) Dispersed oil in water (b-i-2) Emulsion of oil in water (b-ii) Oil Dominated Flow (b-ii-1) Emulsion of water in oil (b-ii-2) Dual dispersion

Figure 2-4 Oil-Water Flow Regimes in Horizontal Pipes

In the flow conditions prevalent in horizontal oil wells with the fluid influx models, however, the flow process is rather dominated by unstable emulsion due to turbulence. It has also, been observed that friction pressure loss effect is predominant in the W/O flow regime than in the O/W regime. In the latter case, the dominant forces become that of slippage between the flowing mixture components. Considering the radial influx of fluid along the length of the well, it is reasonable to conclude that the presence of a stratified flow in horizontal wellbore is almost impossible but rather, the flow regimes will be that of various level of dispersion of one phase in the other. Thus, oil-water mixture viscosity is more appropriately described by emulsion viscosity relations.

34

2.10 Friction Pressure Loss of Oil-Water Flow in Horizontal Wells

Most pressure loss relations have been developed for single-phase flow or for 2phase gas-liquid flow conditions. 2-phase oil-water flow has not received commensurate attention as in gas-liquid flow. Studies indicate that the basic equation for the evaluation of the Reynolds number is all that is affected. The density and fluid velocity are evaluated as weighted average of the flowing mixture as a function of the fractional flow. Most commercial numerical simulators have adopted weighted average approach for the mixture viscosity also. However, the mixture viscosity component is more complex and cannot be evaluated as the weighted average. The flow of oil in the presence of increasing watercut creates interfacial drag forces that tend to change the mixture viscosity. Arirachakaran et. al. (1989) explained that the relation for flowing oil-water mixture viscosity is a complex one that cannot be represented with the weighted average concept. They observed, that a flowing oil-water mixture behaves more like a non-Newtonian fluid. In fact their experiments indicates the mixture behavior follows that of a power-law fluid. However, the viscosity relationship depends on the continuous phase. Broughton and Squires (1937) presented several correlations for the evaluation of changing viscosity of oil in the presence of increasing watercut in a turbulent emulsionlike flow regime. Richardson (1950) also developed a correlation for the evaluation of emulsion viscosity in oil-water flow systems. Martinez et. al. (1988) studied the dispersion viscosity of oil-water mixture flow in pipes. They observed two broad

classification of oil- water flow systems namely oil dominated and water dominated flow with a phase inversion point. They also proposed that oil-water mixture exhibits a nonNewtonian rheological behavior and so apparent viscosity of the mixture cannot be 35

calculated using weighted mixing rules. Arirachakaran et al. (1989) developed a mathematical model for the evaluation of the phase inversion point. They observed that in the oil-dominated flow regime, friction pressure loss is critical, whereas slip consideration is critical in the water-dominated flow. Schramm (ed.) (1992) stated that the mixture of two immiscible liquids flowing under turbulence forms an emulsion when one of the phases becomes a collection of droplets dispersed in the other phase. Rnningsen (1995) studied eight different North Sea crude and developed a Richardsontype correlation for evaluating emulsion viscosity of oil-water flow in pipe.
2.11 Numerical Simulation of Water Coning in Wells

Several researchers have employed numerical simulators to study the complex phenomenon of water coning/cresting in vertical and horizontal wells respectively. MacDonald and Coates (1970) performed water coning studies using implicit pressure-explicit saturation (IMPES) analysis with the production term treated explicitly and a fully implicit analysis. They concluded that fully implicit model accepts larger time increment sizes and is more efficient for problems involving high capillary forces but requires more computer time. They also recommended radial model with fine grid around the wellbore for vertical well conceptual studies. Chappelear and Hirasaki (1974) also used radial model for studies of water coning in vertical wells. Yang and Wattenbarger (1991) used radial model with logarithmic grid distribution for vertical wells and a 3-Dimensional Cartesian model for horizontal well studies with finer grid distribution around the wellbore and coarser grid away from the wellbore. Several other authors including Wu et. al (1995) and McMullan and Larson (2000) used a 3-Dimensional Cartesian model with finer grid in the oil zone and coarser 36

grid in the water zone together with implicit type commercial numerical simulators for horizontal well studies. Bowlin et. al. (1997) and Kurban (1999) also employed radial grid systems to evaluate the performance of Downhole water sink technology in vertical wells. Thus the use of radial model with logarithmic grids for modeling vertical well conceptual water coning studies and a 3-dimensional Cartesian grid model with fine grid distribution at the oil zone for horizontal well is employed in an implicit function commercial numerical simulator for this study.
2.12 Current Methods of Controlling Water Cresting in Horizontal Wells

Various methods have recently been recommended by researchers to restore the productivity of horizontal wells. A brief overview of each of the current technology is presented in the following section.
2.12.1 Horizontal Well Completion With Stinger

This method involves redistributing the pressure losses along the wellbore by inserting a piece of pipe of a smaller diameter into the completion/production liner (Figure 2.5). This smaller piece of pipe is typically called the stinger. Permadi et.al. studied this approach both numerically and with a Hele-Shaw physical model and concluded that a stinger length of 0.25 times the producing length of the horizontal wellbore is adequate to redistribute the pressure drawdown and have a more uniform flux along the wellbore. They also reported that the rate reduction due to the insertion of the un-perforated stinger is relatively small and is overcome by the much longer delay in the time to water breakthrough (Permadi, et.al., 1997). They recommended a stinger outside diameter-to-liner inside diameter ratio of 0.667 as most appropriate for optimum 37

combination of annulus pressure loss and production rate increase. For most of the pipe length studied, they observed that inserting the stinger was capable of delaying the time to water breakthrough in horizontal wells by as much as four times.

Figure 2-5 Horizontal Well Completion with a Stinger

Brekke and Lien reported that the stinger completion method, in combination with reduced perforation density, could provide up to 25 percent increase in well productivity during the first part of the production life. They recommended that the completion method be used in horizontal wells where frictional pressure losses along the perforated section of the well restrict the production performance (Brekke, and Lien, 1992). Thus, the completion of the horizontal well with stinger creates an improved sand face pressure profile by controlling the inflow of fluid along the wellbore. This process helps to redistribute the frictional pressure loss along the perforated part of the well.
2.12.2 Water Cresting Control With Variation of Perforation Density

Another scheme commonly recommended for controlling water cresting in horizontal wells is variation of perforation-density to uniformly distribute the influx of fluid (Figure 2.6). In other words, the perforation density is less at the heel than at the toe of the horizontal well.

38

Asheim and Oudeman investigated the option of varying the perforation density in order to create a uniform production and injection profile along the wellbore. They observed that the uniform inflow design may involve some marginal reduction in productivity compared to uniform perforation density method. However, the sweep efficiency is higher and there is attendant delay in water breakthrough time and improved ultimate recovery (Asheim, and Oudeman, 1997).

Heel
Figure 2-6 Schematic of Typical Variable Perforation Density

Toe

Marett and Landman (1993) observed that in cases where water or gas cresting is anticipated, the perforation strategy of most interest is uniform fluid inflow from the reservoir into the well. To accomplish this type of optimization, the well is partitioned into small number of uniformly perforated segments. The perforation density in each segment is varied (either by varying size of perforation or by varying the number of perforation for the same hole sizes) until the optimum distribution is achieved. For the case of water cresting control, the optimum distribution is when each segment has the same inflow. However, the perforation density is least at the heel of the well and increases toward the toe.

39

2.12.3 Water Cresting Control with High Angle Wells

Another technology that has been extensively used to develop bottom water drive reservoirs with severe water coning problems is the high-angle well technology (Figure 2.7). In a bottom water drive reservoir, the tip end of the well tend to be placed closed to the plane of the oil-water contact while the heel end is placed closed to the top of oil sand far away from the oil water contact.

Oil

OW C

W ater

Figure 2-7 Schematic of High-Angle Well with tip Close to the OWC

Typically, slanted wells are classified by the angle of inclination of an axis through the wellbore and a vertical line intersecting the axis. For high-angle wells, this angle is usually higher than about 80 degrees. Results of field experience demonstrate that water breakthrough in high angle wells can occur at two points close to the downstream/heel end and at the toe of the well, but it always occur close to the heel end in horizontal wells. Thinner oil leads to bottom water invading oil around the toe after breakthrough and results in increased displacement efficiency (Permadi, P, 1996). In a Hele-Shaw model experiment on the performance of high-angle wells, Permadi reported that oil recovery tended to be higher 40

for thicker oil columns and higher well angles. However, water went through the whole part of the well length of high-angle wells in a relatively shorter time while some part of horizontal well length in the upstream region was never invaded by water even at high watercut (Permadi, 1996). An important observation is that each of the three methods enumerated above involves some form of rate restriction and specifically in the case of high-angle well, the extent of reach of the well is limited by the thickness of the oil-bearing reservoir.
2.13 Summary

Water influx into an oil well is caused by the high pressure gradient around the wellbore during oil production. In order to avoid this problem of water influx, the oil well should be produced below the critical oil rates. However, these critical oil rates are usually too low for any economic reason. Consequently, researchers have developed correlations for evaluating the time to water breakthrough into the well and the performance of the well after water breakthrough. Several solutions have been developed to minimize the impact of the concurrent production of unwanted water in oil wells. Downhole Water Sink technology is one industry technique that allows for production rates above the critical oil rates at the oil zone with reduced contamination with water. It involves segregated production of the oil and water with a dual completion string with zonal isolation. However, most of the analytical models developed to evaluate the performance of oil wells subject to water influx (coning) have made several simplifying assumptions to reduce the complexity of their solutions. Phenomena such as capillary pressure transition zones and relative permeability hysteresis have often been ignored in these models. 41

Another industry technique for developing reservoirs subject to water coning is horizontal wells. However, horizontal wells are themselves not free from water influx problems (termed water cresting). Like the vertical well counterpart, typical critical oil rates to avoid water influx into horizontal wells are too low for any economic purpose. Most of the analytical solutions for horizontal wells neglect the friction pressure loss in the wellbore or assume that it is negligible compared to the reservoir pressure drawdown. Also, several researchers have recommended various approach to redistribute the wellbore pressure of the horizontal well to minimize the level of water influx at the heel of the well. Finally, the review of literature reveals that detailed studies of water coning in vertical wells using numerical simulators involves radial grid with a fully implicit models while studies on horizontal wells use 3-D Cartesian grids with local grid refinement around the wellbore also in a fully implicit model.

42

CHAPTER 3 DWS VERTICAL WELL PERFORMANCE WITH VARIABLE WATER SATURATION

The application of Downhole Water Sink (DWS) completion in the control of water coning in vertical wells involve the segregated production of water and oil using a dual-type completion with a zonal isolation packer. The basic idea is to perforate both the oil zone and the water zone and produce each fluid with a separate completion string. The production of water from the water zone creates a pressure sink and counters the development or progression of the water cone toward the oil zone completion. Typically, vertical reservoir permeabilities are lower that horizontal permeabilities by a magnitude of 2 to 10 in most cases. This fact becomes an advantage to the DWS technology. The technology enhances radial (horizontal) flow of the reservoir fluid and so involves a less drawdown pressure than in the conventional well with water coning. The consequences is that, for the same amount of fluid (oil and water), pressure drawdown in the conventional well is greater than the DWS well. The Department of Petroleum Engineering at the Louisiana State University conducted theoretical studies on the DWS technology (Wojtanowicz, et al, 1991) and the result of the theoretical studies showed that the hydrodynamic isolation of the bottom and top completions is the key factor to effective control over the production performance of the dual completion. Hunt Petroleum performed the first successful field trial of the DWS completion technology in the Nebo Hemphill field in Louisiana in 1994. Subsequent studies led to the development of a speadsheet-based software (SCON) and an innovative technique (the Inflow Performance Window IPW) to evaluate the performance of the 43

technology. The IPW is a 2-dimensional plot of fluid production rate at the top completion on the x-axis and fluid production rate from the bottom completion (the water sink) on the y-axis. Four zones are distinguishable from the window as shown in Figure 3.1. Inside the triangular-shaped window is the region of segregated fluid production where oil production at the top completion is water-free and water production from the bottom completion is oil-free. Below, this window is the domain of clean water production from the sink (bottom completion) and watercut oil production from the top completion.

120 Water Rate (bbl fluid/day) 100 80 60 40 20 0 0

ugh etion o r h akt ompl e r C B Oil ottom in B

n egio r p -flo ity Flip nstabil of i

te ega r g Se
10

n gio e wR Flo

ugh o r h kt rea pletion B m ter Wa op Co in T

30 40 50 60

20

Oil Rate (bbl fluid/day) S:Oil Bt Line S:Water Bt Line

Figure 3-1 Typical Inflow Performance Window (IPW) and its Regions

Above the segregated fluid region, oil breaks through into the bottom completion (the sink) and at the extreme end of the window is a line of high cone instability where production of segregated fluid is almost impossible. The cone could be a water cone

44

breaking through at the top completion or change quickly to an inverse oil cone breaking through at the lower completion with very little variation in production rates. As in most coning studies on reservoirs with strong bottom water drive, this study assumes a steady-state flow process. Consideration of the fluid flow mechanisms of capillary transition pressures and relative permeability hysteresis complicates the mathematical analysis and thus, the studies were performed with numerical simulators. The numerical simulation of DWS performance employs a conceptual model, which assumes that the bottom aquifer creates enough pressure support to achieve a steady state flow behavior. The concept of steady-state flow process also pre-supposes that, for a particular production, the pressure drawdown becomes stabilized after a very short period of time. The consequence is that for a particular flow rate, there exist a stabilized limiting watercut and a stabilized flowing bottom hole pressure. The steady state simulation approach was also necessary to effectively evaluate the results of the analytical model developed by Wojtanowicz, et al. (1991), and Shirman (1997, 1998) while at the same time verifying the results of the simulation with known theoretical solutions. The analytical model calculates steady state pressure distribution around the well with single or two-phase production at the top and bottom completions.
3.1 Pressure Stabilization and Steady-state Water Cone Simulation

Numerical simulation of water cone development is very sensitive to little changes in pressure. As an example, a pressure differential as small as 1.0 psi can create a cone height of as much as 15.0 ft for a water/oil model where the water density is about 62.4 Ibm/ft3 and the oil density is about 53.04 Ibm/ft3 respectively. The consequence of the above observation is that effective simulation of steady-state water cone development 45

requires a high level of accuracy. Researchers have recommended three approaches for the simulation of a steady-state production process with commercial simulators. They are: (1) Creation of a large aquifer with adequate enchroachable water, (2) Use of infinitely large porosity in last radial grid cell creating steady-state boundary pressures, or at the lowest bottom grid in the aquifer zone to simulate a bottom water drive reservoir; and (3) Fluid re-injection approach. Each of the above solutions is capable of generating adequate reservoir energy in the aquifer zone to simulate a steady-state process.
T H R E E B A S IC M O D E L S O F S T E A D Y S T A T E IN F L O W S IM U L A T IO N

1 . L a r g e b o tto m w a t e r d r iv e

2 . P r o d u c e d F lu id R e - in j e c t io n

3 . I n fi n it e P o r o s it y m o d e l.

L a r g e B o tt o m a q u if e r s u p p o r t to m a i n ta in P r e s su r e

R e - in j e c ts P rodu ced F lu id s (m in im u m b o tto m a q u if e r )

L a s t w a te r c e ll h a s in f in it e p o r e volu m e

Figure 3-2 Steady-state Numerical Simulation Models

However, the first option requires effective material balance solution to establish appropriate enchroachable water. This is time consuming considering the amount of simulation that is required to create enough data points to generate an inflow performance window (IPW). The second option of infinitely large porosity at the radial boundary was employed in all previous work on this project. However, it was found to be inadequate as 46

it stabilized pressure at the boundary but failed to stabilize pressure around the wellbore. This option creates a drawdown pressure profile at the near wellbore region different from that at the radial boundary for different time frames. In order to effectively simulate a steady-state bottom water drive system with stable pressure at all points, the infinite porosity grid should be the lowest set of grid blocks in the water zone and not the last radial grid. This ensures that the energy supply from the aquifer is uniform through out the drainage area of the reservoir. Figure 3-3 shows a typical pressure profile generated as a function of time using the radial boundary infinite porosity model. The graph shows a different pressure within the first five years of simulation forecast and the remaining fifteen years.

IN F IN IT E P O R O S IT Y M O D E L A N D D E L A Y E D S T A B IL IZ A T IO N
D e la y e d S ta b iliza tio n C r e a te s tw o P re s s u re P ro f ile s (In finite p o ro s ity m o d e l) 1988 1987 1986 1985 1984 1983 1982 1981 1980 1979 0

Top BHP (psi)

10 T im e (y e a r s )

15

20

Figure 3-3 Unstable Flowing Bottom-hole Pressure at Top Completion

This development of differential pressure gradient was considered inadequate for water coning studies since the water cone development is a near wellbore phenomenon. To effectively present the above argument, it is necessary to present the equation of Darcy for a steady-state fluid flow process. This equation is produced with a slight 47

modification for the 2-phase oil-water flow incorporating the relative permeability function to account for the component of oil and water in the fluid flow, assuming no slippage (field units, Q= stb/day; pe, pw = psi; k = mD; Bo, Bw = bbl/stb; mo, mw = cP, h = ft and re, rw = ft).

Qo = 0.00708 *

kkro h * ( pe - pw ) re mo Bo ln r w

3-1

Qw = 0.00708 *

kkrw h * ( pe - pw ) re m w Bw ln r w

3-2

The above equations show that for a steady-state process with constant pressure boundary and constant flow rate, the wellbore pressure should be constant with time. The results presented in Figure 3-3 shows a differential pressure profile with time. Thus it was necessary to further investigate this abnormality. To understand the cause of the unstable pressure distribution with time, it was necessary to evaluate the radial pressure profile from the near wellbore region to the outer reservoir boundary. The pressure distribution at mid-perforation grid blocks is shown in Figure 3-4. While the pressure at +/- 350 ft away from the wellbore converges as expected for steady-state fluid flow process, it diverges at the near wellbore region over time. This difference in near wellbore pressure profile over time was attributed to the time lag in the infinite porosity steady-state option.

48

Well P1 BHP vs radius plot (Infinite Porosity model)

1987.50 1987.00

Well P1 BHP (psia)

1986.50 1986.00 1985.50 1985.00 1984.50 1984.00 1983.50 1983.00 0

15 Years Section B-B

3 Years Section A-A

50

100

150

200

250

300

350

400

Pressure Radius (ft)

Figure 3-4 Unstable Wellbore Pressure Distribution with time

The conclusion from this study is that the infinitely large pore space option at the radial grid boundary is inadequate for the pressure sensitive studies of water cone development. Another tool employed to evaluate the effectiveness of the radial boundary infinite porosity model is the watercut profile over time. It is expected that the watercut should reach a stable level called limiting watercut over time for a steady-state process. Figure 3-5 shows that the watercut never reaches the expected limiting value for this infinite porosity option but rather creeps. Kurban reported that the inability to stabilize the flowing bottom hole pressure means that limiting watercut could not be achieved (Kurban, 1999). Consequently, previous research efforts adopted two acronyms called engineering approximation and asymptotic extrapolation in order to define a cut off point for determining the critical rates needed to generate the IPW. 49

Field WaterCut vs Time (Creeping case)

0.9 0.8 Water Cut, Fraction 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 200 400 600 800 Time, Days 1000 1200 1400 1600

Figure 3-5 Creeping Watercut for Unstable Flowing Bottomhole Pressure

The last model of steady-state simulation in commercial numerical simulator is the fluid re-injection method. By this approach, the produced fluid (oil and water) is reinjected into the reservoir to maintain the reservoir energy and thus simulated a steadystate flow process. Figure 3-6 is made with the fluid re-injection approach.

S T A B IL IZ E D P R E S S U R E A T B O U N D A R Y A N D A T W E L L B O R E

W e ll P 1 BHP vs r a dius plot (Ne w m ode l - m id-uppe r c om ple tion)

1 9 8 8.00 1 9 8 6.00 1 9 8 4.00 1 9 8 2.00 1 9 8 0.00 1 9 7 8.00 1 9 7 6.00 1 9 7 4.00 1 9 7 2.00 1 9 7 0.00 0.00 0

Well P1 BHP (psia)

T im e 15- ye a r s ( B - B s e c t io n - t h in lin e )

T im e 3 - y e a r s ( A - A s e c t io n - t hic k lin e )

2 0 0 .0 00

4 00 .0 0 0

6 0 0 .0 00

80 0 .0 0 0

1 00 0 .0 0 0

P re ssu re R a d iu s (ft)

Figure 3-6 Overlap of Pressure at Different times Indicate a Stable Steady-state Flow Process

50

This approach appears to be effective in stabilizing flowing wellbore pressures with accuracy of 0.01 psi both at the wellbore and at the outer radial boundary as shown in Figure 3-6. The foregoing exercise has been an attempt to minimize the time expended in generating a typical IPW using commercial numerical simulators. The ability to determine a stable limiting watercut from the results of the simulator reduces the time and efforts needed to determine each of the critical rates of the two completions. On a typical IPW, a minimum of eight data points representing the critical water production rates of the bottom completion for several given oil rates at the top completions are needed. Thus, a simplified approach to determining these points also increases its attraction to industry operators. Operator personnel need few hours of training to generate the Inflow Performance Window using in-house commercial simulators.

Figure 3-7 Watercut Profile for Stabilized Bottomhole Pressure (Plateau)

51

Figure 3-7 shows the watercut profile for the fluid re-injection model generated from the graph software. The limiting watercut stabilizes over time as shown in the plateau. It is, thus, now possible to obtain the critical flow rates necessary to generate the IPW from the results of the reservoir simulation runs without applying the concept of asymptotic extrapolation or engineering approximation. Consequently, fluid re-injection method was adopted as adequate for simulations performed on vertical wells. A disadvantage of this approach is that oil production follows a straight-line profile over time with no distinct decline since the oil volume in the reservoir remains constant with time.
3.2 Verification of Results with Analytical Model IPW

In the absence of field data in form of history match, it was determined that the bench-mark to verify the results of the commercial simulator would be the results from the analytical model.
100 90 80 70 60 50 40 30 20 10 0 0 5 10 15 20 25 30 35 40 45 Oil Rate (bbl fluid/day) A:Oil Bt(Anal.) A:Water Bt(Anal.) S:Oil Bt(Sim.) S:Water Bt(Sim.)

Figure 3-8 Comparison of IPW - Numerical Simulator and Analytical Model

Water Rate (bbl fluid/day)

52

The logical step was to compare the Inflow Performance Windows from both numerical simulation and analytical model for the same reservoir and fluid properties. Figure 3-8 is a superimposed plot of IPW from numerical simulator and a similar IPW from the analytical model previously developed for DWS studies. The solid line plots represent the IPW generated from numerical simulator while the hatched line plots represent an IPW generated from the analytical model for the same reservoir and fluid data. The plot shows a close agreement in the size and shape of the two windows. The variation in the area of the two IPW is about 0.050 or 5 percent. A primary difference between the result of the analytical model and that of the numerical simulator, neglecting the effect of capillary transition, is that the IPW for the numerical simulator appears to close faster than that of the analytical model. This is attributable to the fact that the analytical model did not incorporate the dynamics of relative permeability variations with water saturation changes. The analytical model uses only end-point mobility considerations and so neglects changes in relative fluid mobility due to increasing water saturation. It therefore tends to overestimate the size of the domain of segregated fluid production. Swisher and Wojtanowicz (1995) indicated that as the oil rate increases, the stability of the oil-water contact (OWC) range decreases so that there is a maximum oil rate above which the drainage-production system is inherently unstable. The term, unstable as used by the DWS technology refers primarily to the deformation of the cone profile causing concurrent production of water and oil from both completions. Having verified the performance of the numerical model, the next step was to study the performance of the DWS wells under certain fluid flow mechanisms such as 53

capillary transition zone characterized by the presence of mobile oil and water as well as incorporate the directional properties of relative permeability termed hysteresis effects. Old oil wells with a history of severe water coning typically presents this watered out zone of co-mingled mobile oil and water in a manner similar to the presence of a capillary transition zone due to low reservoir permeability. The goal is to investigate how the DWS technology will perform in these old oil wells with severe water coning history as compared to the performance in a new oil well with distinct oil-water contact.
3.3 Capillary Pressure Transition and Simulation of DWS Well

To clearly understand the effect of capillary forces on water coning and hence the DWS technology performance, the following definitions are needed:
3.3.1 Capillary Pressures

Capillary pressure is simply the pressure exerted by inter-molecular attractive force between immiscible fluids, solids and gases in the reservoir. This capillary pressure in the reservoir is a function of the chemical composition of the rock and fluids, the pore size distribution of the sand grains in the rock and the saturation of the fluids in the pores. The reservoir pressure gradient must be such as to overcome the effect of these attractive inter-molecular forces before fluid mobility can be achieved. Consequently, when reservoir energy is not high enough to overcome these inter-molecular forces the oil is trapped as residual oil. In field units, capillary pressures are given by (Craft and Hawkins, 1991):
Pc = 9.519(10) -7 s wo cosq rc

3-3

The radius of capillary is related to reservoir permeability by: 54

k = 1.15(10)13 rc

3-4

Where, Pc is in psi; swo is in dynes/cm; rc is in ft; q is in degrees; and k is in mD. Combining the two equations yield an expression for capillary pressures as a function of reservoir permeability: Pc = 3.228s wo cos q k
3-5

The above relation has been confirmed by experiments (Hassler and Brunner, 1944) for consolidated sands where Pc is proportional to 1/k0.5. For unconsolidated sands and limestones, Pc vs. k has no definite relationship but Pc is rather a function of k, r, and Sw. The angle, q, measured through the wetting phase fluid against the pore walls is known as the contact angle. This angle is influenced by the tendency of one of the two immiscible fluids to preferentially spread on or adhere to the pore wall surface. This preferential spreading or adhering capacity is called the wettability of the rock pore wall and the fluid that spreads/adheres more on the surface of the wall is called the wetting phase. The other fluid that does not spread on the pore wall, but rather occupy the inner space is called the non-wetting phase. The degree of wettability depends on the chemical compositions of the two fluids, particularly the asphaltene content of the oil, as well as the nature of the pore wall. An increase in the saturation or content of the wetting phase is called an imbibition process while an increase in the concentration (saturation) of the non-wetting phase is called a drainage process. Equation 3.5 shows that capillary pressure effects are less severe in high permeability reservoirs than in low permeability reservoirs. Consequently, for normal 55

recovery, the effect of capillary pressures is neglected if the reservoir permeability is high in the magnitude of thousands of millidarcies. Capillary pressure is typically measured in the laboratory using air-mercury system. The laboratory measurements are then up-scaled to reservoir condition using the Young-Laplace equation:

Pc1 = Pc 2 *

(s cos q )1 (s cosq ) 2

3-6

Typically, the reservoir rock pore space is initially occupied by the wetting phase fluid (water for sandstone). The migration of hydrocarbon into this pore space causes the water to be displaced. A minimum threshold pressure, capillary threshold pressure, is required to displace the fluid in the largest pore space. Thereafter, smaller pore spaces are filled up by the migrating hydrocarbon until an irreducible level of water (wetting phase) is reached called the connate water saturation or irreducible water saturation. At the point of irreducible wetting phase saturation, a further increase in the displacement pressure produces no further increase in the hydrocarbon saturation. The region between the threshold capillary pressure and the capillary pressure due to the irreducible water saturation is known as the capillary transition zone or region. Thus, the capillary transition relates to the distribution of water saturation as a function of height above the free water level (FWL) in a reservoir. The threshold capillary pressure, Pct, is thus, the minimum pressure required to overcome the intermolecular forces between the surfaces of the two immiscible fluids in order to move the wetting phase fluid (water) from the pore space by the non-wetting phase (oil). 56

This threshold capillary pressure found in reservoir rocks is proportional to the height above the free water level (FWL) datum, where a region of 100 percent water saturation is found. The free water level is thus a property of the reservoir system, while an oil-water contact observed on a particular well in the reservoir will depend on the threshold pressure of the rock type present in the vicinity of the wellbore.(Archer and Wall, 1986). The relationship between height above FWL and capillary pressure is derived from consideration of the gravity-capillary pressure force equilibrium and is expressed in field unit as(Archer and Wall, 1986):
Pc ( S W ) = H ( SW ) * ( r w - r o ) 144

3-7

Where, Pc is in psi, H(Sw) is in feet, and rw, ro are in Ibm/ft3. The equation 3.7 can be used to evaluate the threshold capillary pressure when the height of the oil-water contact above the field FWL is known together with the densities of the oil and water. Alternatively, if the OWC is known, a drainage test can be conducted on a core sample from the reservoir to determine the threshold capillary pressure from which the depth of the FWL can be determined. The experimental measurement of capillary pressure transition is rarely a straight line. Rather, it is curved line as shown in Figure 3.9. However, where the capillary pressure information is not available, researchers have used the Equation 3.7 to extrapolate the value of capillary pressure with height above the oil-water contact. (Kurban, 1999), (Weinstein, 1986), (Willhite, 1986), and (Yokoyama, 1981). 57

clean oil

Pc, a to Height above FWL

Top Transition Zone

Pct

Observed Oil Water Contact

Free Water Level, FWL, Pc = 0 Swirr Wetting Phase Saturation, Sw

Figure 3-9 Typical Capillary Transition Pressure Zone

This linear extrapolation approach assumes the OWC as the zone of 100 percent water saturation. Where, capillary pressure data are available from adjacent wells or reservoirs with similar properties, the equation of Leverett can be used to evaluate a dimensionless parameter called the Leverett J-function. This dimensionless value can in turn be used to evaluate the capillary pressures for the reservoir rock in question. It expresses a relationship between capillary pressures and permeability as follows:

J (SW ) =

Pc ( S W ) k s cos q f
58

3-8

3.3.1.1 Capillary Pressure Hysteresis

The capillary pressure that must be overcome to displace the wetting phase from the pore space by the non-wetting phase differs from that required to displace the nonwetting phase from the pore space by the wetting phase. This difference creates separate paths for the capillary pressure effect as shown in Figure 3-10. When the wetting phase (water) saturation increases by displacing the non-wetting phase from the pore space, the imbibition threshold pressure, called the suction pressure does not coincide with the 100 percent wetting phase (water) saturation point. Rather it stops short of the 100 percent point and defines the residual saturation of the non-wetting phase.

Capillary Pressure, Pc, psi

Drainage Imbibition

Pct

0 Swirr
Figure 3-10 Capillary Pressure Hysteresis

1.0 Wetting Phase Saturation, Sw

59

The threshold pressure of the drainage phase almost always coincides with the 100 percent wetting phase saturation point. This deviation in the path of the two processes, drainage and imbibition is termed the capillary pressure hysteresis as shown in the Figure 3.10. However, the experimental difficulties in determining the definition of the imbibition capillary pressure curve, combined with the difficulties of using the information in reservoir simulation models, has led to the assumption by many reservoir engineers that, for practical purposes, capillary pressure hysteresis does not exist (Archer and Wall, 1986). For the reservoirs studied, imbibition capillary pressure data were not available. However, the drainage capillary pressure data were available in some cases and in other cases, the linear extrapolation approach was adopted necessitating the use of only one table of capillary pressures in the simulation. The primary objective was to incorporate a capillary transition zone in the DWS model (i.e. a zone of varying water saturation) and this was effectively achieved with the inclusion of the drainage capillary pressure data.
3.3.1.2 Capillary Transition Pressure and the DWS Technology in Old Oil Wells

The concept of capillary pressure transition, however, is very significant in modeling the performance of the dual completion water sink performance, especially in old oil wells with severe water coning history. The prolonged production of water causes a zone of co-mingled variable fluid saturation to develop. This zone of mobile comingled oil and water is termed the transition zone. The presence of this transition zone of varying water saturation indicates that the intermediate relative permeabilities, as will be seen later, play prominent roles in fluid mobility and hence oil recovery. 60

Most water cone studies have neglected the effect of capillary transition in water cone development. A few authors (Henley, 1960; Reed, 1984) have simulated the effect of capillary pressures in water production and concluded that the effect of capillary transition zone was small and could be ignored. This conclusion is, probably due to the fact that the studies did not incorporate the effect of the directional properties of permeability to the various fluid phases namely, the imbibition and drainage properties. The studies applied only one set of relative permeability curves - the imbibition curve. However, when the both imbibition and drainage phenomenon are involved as in the DWS technology, these effects cannot be ignored. Neglecting the capillary pressure transition effect will not only overestimate the amount of moveable oil, it will underestimate the amount of water required to stabilize the oil-water contact or at least avoid water breaking through into the top completion. The results of inclusion of the capillary transition effect in modeling of DWS completion technology for old oil wells with severe water coning history are quite significant. Water breakthrough times for thin reservoirs showed little or no change with inclusion of capillary transition zone. The reason for this response is because water breakthrough times are usually very small, almost zero, for thin reservoirs and thus capillary pressure effects will not be easily noticeable. This agrees with the result of previous research in this type of reservoirs (Kurban, 1999). However, for extensive reservoirs, over 50-ft of oil sand, the effect becomes very pronounce depending on the extent of the transition. For an increased capillary pressure transition height, the Inflow Performance Window reduces drastically. The critical water rate is essentially unchanged, however, the critical oil rate is greatly affected. The 61

gradient of the water breakthrough line increases with increasing size of the capillary transition zone.

Water Rate (bbl fluid/day)

350 300 250 200 150 100 50 0 0 50 100 150 200 250

Oil Rate (bbl fluid/day)

30' C AP :Oil B t 10' C AP :Wa te r B t 30' CAP :Wa te r B t 0'C AP : Wa te r B t" 10' CAP :Oil Bt "0'C AP :Oil B t"

Figure 3-11 Effect of Capillary Transition Zone on DWS IPW

The consequence is a reduction in the size of the domain of segregated fluid production. The relevance of the above observation is that the analytical model tends to overestimate the performance of the DWS technology for these old wells. Figure 3.11 show the reduction in size of the IPW for different level of capillary transition region.
3.4 Relative Permeability Hysteresis and the DWS Technology Modeling

Another fluid flow mechanism of interest to the modeling of DWS completions is that of relative permeability hysteresis. During the water coning process, the wetting phase saturation increases representing an imbibition process. When the water sink is activated to counter the cone development, the saturation of the non-wetting phase increases representing a drainage process. 62

The concept of relative permeability has been conveniently used in reservoir engineering to relate absolute permeability of the porous media (100% saturated with a single fluid) to the effective permeability of a particular fluid in the system when that fluid occupies a fraction of the total rock pore volume. ke =k * kr where: ke is the effective permeability of the phase, kr is the relative permeability of the phase, k is the absolute permeability of the porous medium. For two-phase flow in porous media, the relative permeability to each phase is expressed as a function of saturation. The minimum possible level of the wetting phase saturation, water in our studies, is usually designated as the connate water saturation or Irreducible water saturation (Swc or Swirr). The maximum wetting phase saturation Sw,max. is reached when the non-wetting phase, oil, reaches its residual saturation level, i.e oil mobility becomes zero either due to high capillary pressure trapping as earlier discussed or due to decreasing reservoir energy. Relative permeability curves are dependent, to a great extent on the rock-fluid interactions or wettability. Wettability refers to the preference of reservoir rocks to be coated or wetted by a particular fluid type in multi-phase system. This preference depends on the rock fluid composition and interaction. Wettability of a surface depends on the threshold capillary pressure term scosq". This term indicates that capillary pressure controls the sequence of pore saturation changes and hence relative permeability so long as viscous forces are not controlling the flow process. For petroleum engineering studies we observe that rocks could be oil wet, water wet, intermediate wet or mixed-wet. 63

Relative permeability curves are currently generated for either oil wet or water wet rocks using published correlations (Archer and Wall, 1986). Some rules-of-thumb has been proposed by researchers in order to characterize the wettability of reservoir rocks. One such rules-of-thumb is that proposed by Craig (Craig, 1993). Table 3-1 list Craigs rules-of-thumb for determining wettability from twophase relative permeability curve (Craig, 1993). It has been quite valuable for water flooding projects and as a benchmark for reservoir simulation experts in analyzing results of experimental/laboratory data prior to the development of reservoir simulation models.
Table 3-1 Craig's Rules-of-Thumb for Determining Rock Wettability Water wet Oil Wet

Irreducible or connate water saturation

Usually greater than 20 to 25 % PV

Generally less than 15% PV. Frequently less than 10%

Saturation at which oil and water relative permeabilities are equal. Relative permeability to water at the maximum water saturation based on effective oil permeability at reservoir connate water saturation

Greater than 50% water Less than 50% water saturation Generally less than 30% saturation. Greater than 50% and approaching 100%

The table indicates the need to also establish the wettability of the reservoir rock system before deciding whether DWS technology is applicable to the reservoir of interest

64

or not. The research results presented in this dissertation have assumed a water wet sandstone reservoir for all the models.
3.4.1 Fractional Flow Analysis

Relative permeability curves dominate the fractional flow equation applied in numerical simulators to determine the flow rate of each phase in the reservoir. Thus, an understanding of fractional flow equations and how they depend on the fluids saturation is necessary. Fluid permeability characteristics are frequently presented in terms of permeability ratio kw/ko. This is because the fractional flow equations are expressed in terms of relative permeability ratios. However, the generalized fractional flow equation often neglects the effect of capillary pressure transition but rather assumes a sharp fluid interface. A review of the basic fluid flow equations in porous media as they apply to the effect of capillary pressures and relative permeabilities will be undertaken at this point. The first and most important flow equation is the Darcy equation. Leverett, (1941) first presented the concept of fractional flow of fluid for an oil-water system in terms of the relative permeabilities and the fluid properties:

1+ fw =

kkro ut m o

Pc - gDr sin a dip x m k 1+ w o mokw

3-9

Civan and Donaldson (1987), developed a more detailed expression (Equations 310 and 3-11) for evaluating the relative permeabilities of the wetting phase and the nonwetting phase incorporating the effect of capillary pressures in the solution of the fractional flow equation as follows: 65

 f w kA Pc S w dPc 1 . . k rw m q S x f o dS dS w w w w w swx=0 f w kA Pc S w dPc S w dS w - x L + Q dQ 1 - m q . S . x k rw f o dS w w w w + Lqw m w 1 d Dp =0 . - Dp - Q kA k rw dQ

-1

sw

-1

3-10

And

k rw m o f w k rw kA Pc 1. . = . m w k ro m fo q x w w

3-11

Civan and Donaldson indicated the above two equations can be solved for the respective values of relative permeabilities as functions of water saturations. The expression, however, neglects the directional properties (hysteresis) of relative permeabilities.
3.4.2 Imbibition and Drainage Relative Permeabilities

Like, capillary pressure, relative permeability exhibits directional properties leading to the concept of hysteresis. In effect, the relative permeability curve for an imbibition process is different from that for a drainage process for the same fluids. Inspite of the wide variety of pore structure in reservoir rocks, of preferential wettabilities between fluids and rock surfaces, and in fluid properties, normalized plots of relative permeability, ko/k, kg/k, kw/k against saturation, exhibit general similarities of form. It is, then, attractive to attempt to formulate theoretical semi-empirical, or purely empirical relationships, to assist in smoothing, extrapolating, extending (or even dispensing with) experimental measurements of effective permeabilities (Archer and 66

Wall, 1986). For some of the fields studied, relative permeability data were not available and so published correlation presented in sections 3.4.2.1 to 3.4.2.4 (Archer and Wall, 1986) were used. In other fields, relative permeability and capillary pressure data from cores were supplied by Joint Industry Panel (JIP) members.
3.4.2.1 Wetting Phase Drainage Relative Permeability
* k rw = S w

( )

, where a = 4.0 (Corey Model, used for this study)

3-12

(S )

* w drainage

S w - S wi , 1 - S wi

Swi = irreducible water saturation

3-13

3.4.2.2 Wetting Phase Imbibition Relative Permeability

* k rw = S w

( )

imb

(a = 4.0),

3-14

And

(S )

* w imb

* = Sw

( )

drainage

-1

(S ) 2

* 2 w drainage

3-15

3.4.2.3 Non-wetting Phase drainage Relative Permeability

* * k rnw = 1 - S w 1 + 2S w

)(
3

) (where S

* w

is the drainage case)

3-16

3.4.2.4 Non-wetting Phase imbibition Relative Permeability

k rnw

S - S wi = 1 - w 1 - S wi - S r

3-17

67

With these normalized equations, all that is required is information on some basic properties such as irreducible water saturation, residual oil saturation and the respective end-point effective permeabilities at these saturations and the absolute permeability in order to generate complete relative permeability curves.
1.0000 0.9000
Relative Permeability(fraction)

0.8000 0.7000 0.6000 0.5000 0.4000 0.3000 0.2000 0.1000 0.0000 0.00 Blue = Imbibition curves Red = Drainage (Corey Exponent=4.0)

0.20

0.40

0.60

0.80

1.00

Water Saturation (Sw - fraction)

Figure 3-12 Relative Permeability Curves for Drainage and Imbibition

The typical set of imbibition and drainage relative permeability curves generated for this study is shown in Figure 3-12. As will be observed later, the position of the end-point saturations plays a significant role in the fractional flow of the two immiscible fluids especially in cases where there is a sharp interface between the water and the oil zones, that is, if there is no capillary transition zone. However, if there is capillary transition zone as in old oil field with severe water coning history, not only the end-points but the entire relative permeabilities at different saturations within the hysteresis loop plays a paramount role in determining the fractional flow of each phase. For most of the reservoirs studied, only 68

one value of connate water (irreducible water) saturation and one of residual oil saturation were provided from the industry sponsors, thus the end-points of the drainage and imbibition curves are the same.
3.4.3 Simulation of Relative Permeability Hysteresis in Eclipse

The application of the dual completion in DWS wells causes frequent reversal in the direction of saturation change creating a relative permeability hysteresis effect. Some simulators have the capability to sense saturation reversals and introduce hysteresis effects into the relative permeability functions automatically. Other simulators require that the relative permeability curves be changed manually when saturation reversals occur (Mattax and Dalton, 1989). The Geoquest Eclipse 100/200 requires that the relative permeability curves be changed manually when there is saturation reversal. The option of relative permeability hysteresis is used to specify different tables of saturation functions for both the drainage process of increasing saturation of non-wetting phase and the imbibition process of increasing saturation of wetting phase. Two user-supplied tables of saturation Vs relative permeability are specified using the key words SATNUM and IMBNUM in the REGIONS section of the data deck. First, the keyword HYSTER is used to turn on the hysteresis function in the RUNSPEC section. A further keyword, EHYSTR in the PROP section sets the values of the two parameters that determine the scanning curves for capillary pressure hysteresis and non-wetting phase relative permeability hysteresis, and selects one choice of models for relative permeability hysteresis (Schlumberger Geoquest, 1997). The scanning curves help to determine the path of flow should the process be reversed

69

abruptly at some intermediate point. Two methods are generally available for scanning curve generation, the Carlson method and the Killough method.

1
Relative Permeability (Fraction)

Drainage Curve

4
Imbibition Curve

Swc

5
Swc(imb) Swc(scan)

2
Swc(drain.)

Increasing wetting phase (water) saturation

Figure 3-13 Relative Permeability Hysteresis Curve with Scanning Option

The Killough method has been selected for this study to ensure that the end point saturation of the scanned curves always lie between the end-points of the critical drainage and imbibition curve saturations. Figure 3-13 will be used to illustrate the scanning curve option in numerical simulators and its effect. In Figure 3-13, consider a drainage process (decreasing water saturation when the water sink is turned on) starting at point 2, if the full drainage process 70

is carried out, the end point will be at point 1. However, if the drainage stops midway say at point 4, (when the production from the top completion is increased) such that water saturation begins to increase (representing an imbibition process), the simulator can either retrace the curve back to point 2, if the RETRACE option is specified or create a new scanning curve from path 4 to5 as shown above. For the emphasis of reproducibility of results, the retracing option has been adopted for this research. Also, the Killough option as described earlier ensures that, should the scanning option be adopted, the end-point saturation of the scanned curve lies between the critical saturations of the drainage and imbibition processes specified in the two saturation tables. This critical saturation at the end-point of the scanned curve is the trapped critical saturation. It is a function of the maximum oil saturation reached in the simulation run process. Also, in order to simplify the results of this study, the hysteresis option for capillary pressures has been ignored. Again, for most of the studies in this chapter, the critical saturation for the imbibition and drainage flow process have been set to coincide in order to effectively evaluate the improved oil production capability of DWS wells and to maintain the same value of residual oil saturation and connate water saturation. This is necessary to maintain equal amount of displaceable oil. Appendix A is a sample data deck for DWS well simulation with permeability hysteresis and capillary pressure transition zone.
3.4.4 Analysis of Results from the Relative Permeability Hysteresis and Capillary Pressure Studies

The result of simulation studies on the effect of relative permeability hysteresis function, neglecting the presence of a capillary transition zone, is presented in Figure 371

14. This result indicates that the two windows were essentially the same. Analysis of this result showed that the end points relative permeabilities plays a very prominent role. As long as the connate water saturation is the same for both imbibition and drainage relative permeabilities and the residual oil saturations are also the same for both imbibition and drainage cases, the IPW windows will practically over lay each other in reservoirs with distinct oil-water contact as in Figure 3-14. Thus, only the end point relative permeabilities are necessary for a step-function fluid flow model.
Effect of Relative Permeability Hysteresis on DWS IPW Window for Extensive reservoir without capillary transition 400 350 300 Water Rate (bbl fluid/day) 250 200 150 100 50 0 0 50 100 Oil Rate (bbl fluid/day) DNG ZERO CAP:Oil Bt IMB ZERO CAP:Water Bt DNG ZERO CAP:Water Bt HYS ZERO CAP: Water Bt IMB ZERO CAP:Oil Bt HYS ZERO CAP: Oil Bt 150 200 250

Figure 3-14 Negligible Hysteresis Effect Due to use of Same End-point Saturations for both Imbibition and Drainage

In Figure 3-14, DNG ZERO CAP represents a drainage process neglecting the effects of capillary pressures; IMB ZERO CAP also represents an imbibition process neglecting the effects of capillary pressures and HYS ZERO CAP represents the hysteresis option also neglecting the effect of capillary pressures. In effect, the Figure 314 results essentially tests the effects of relative permeability hysteresis in reservoirs without capillary transition pressures. 72

This result agrees with Miller and Rogerss (1973) conclusion of a similar study. It also explains why some industry analysts (operators) recommends the use of straightline relative permeability curves for high permeability reservoirs with distinct oil-water contact. They arguing that only the end-point saturations and relative permeabilities are useful.
Bottom completion production rate (stb/day)
350 300 250 200 150 100 50 0 0 50 100 150 2 00 250

T o p C o m p letio n p ro d u c tio n rate (stb /d a y )

H Y S 0' CA P :O il B t H Y S 30' CA P :W a t e r Bt H YS 0' CA P :W a te r Bt H YS 30' C A P : O il B t

Figure 3-15 Combined Effects of Capillary Pressures and Relative Permeability Hysteresis

However, Figure 3-15 show the performance of the DWS technology in a field with water saturation gradient such as an old oil field with severe water coning history or low permeability reservoirs with a capillary transition zone. HYS 0 CAP shows the segregated flow domain of the IPW for a case of zero capillary transition while HYS 30 CAP shows the segregated flow domain for the same 50 feet oil pay zone reservoir with 30 ft or 60 percent of the oil zone occupied by the capillary transition zone. For this case, the effect of the intermediate points of the various relative permeability curves becomes important. The mobility of fluid at different saturation is different for the case of increasing water saturation (imbibition) when the cone is developing and for decreasing 73

water saturation during cone reversal. Straight-line approximation of relative permeability data will, therefore, not adequately account for the mobility of fluid in the saturation transition zone. The effective way to model the performance of such wells with a zone of variable saturation is to incorporate the effects of capillary pressures with relative permeability hysteresis. The cone reversal process of the water sink creates a temporal drainage displacement when oil displaces the water cone. On the other hand, when the production of liquid from the top is high, water saturation increases and the cone begins to develop and the flow path is that of imbibition. The result of Figure 3-15 shows that the inflow performance window is smaller in reservoirs with capillary pressure transition zone. This indicates that the region of segregated fluid production becomes smaller for such reservoirs. Combining the results of the Figures 3-11 and 3-15, it could be observed that the critical factor in modeling the performance of the DWS well completion technology in old oil wells with severe water coning history is the size of capillary pressure transition zone (Inikori, and Wojtanowicz, 2001(A)). It is important that capillary pressure measurements from cores be performed in candidate wells and the effect incorporated in the design model. The result of the study also show that the concept of Inflow

Performance Window presents a powerful tool for understanding and quantifying the effects of capillary pressure transition zone in oil-water flow in porous media.
3.5 DWS Well Productivity Performance Studies

Following the results of effects of capillary pressure and relative permeability hysteresis on the performance of DWS technology in these old wells presented in the preceding sections, it was necessary to evaluate the oil productivity performance of the 74

technology incorporating these effects. Three main aspects of oil production performance of the technology in these old oil wells were considered, namely, oil recovery performance evaluation; oil productivity index analysis; and watercut performance evaluation.
3.5.1 DWS Well Oil Recovery Performance Study

In a study of DWS well oil recovery performance using a fluid re-injection model, it was observed that the technology demonstrates superiority in oil recovery over conventional completion as shown in Figure 3-16. Several oil production rate s at the top completion were evaluated varying the fluid withdrawal rates at the water sink completion. Figure 3-16 shows a plot of oil recovery over a period of twenty years for various oil production rates.
5,0 0 0 ,0 0 0 4 ,50 0 ,0 0 0 4 ,0 0 0 ,0 0 0

Oil Recovery (Stb)

3 ,50 0 ,0 0 0 3 ,0 0 0 ,0 0 0 2 ,50 0 ,0 0 0 2 ,0 0 0 ,0 0 0 1,50 0 ,0 0 0 1,0 0 0 ,0 0 0 50 0 ,0 0 0 0 10 0 200 300 Sink Dra iang e R at e, b b l/ d a y 2 0 0 B OPD 10 0 B OPD 3 0 0 B OPD 4 0 0 B OP D 6 0 0 B OP D 400 50 0 600

Figure 3-16 Oil Recovery Performance Study of DWS Well

The intercept of the various lines with the oil recovery axis represents the typical oil recovery of conventional wells. The upward sloping nature of the various curves indicates that the technology is capable of improving recovery of oil even in old wells 75

with water coning history. The counter-cone development around the wellbore means reduction in amount of bypassed oil and consequent improvement in oil recovery. The Figure shows that oil recovery increases with increasing drainage rate at the water sink completion. This shows that the technology counters the cone development making the bypass oil available to the well. The upward slope of the recovery curves demonstrates that oil recovery can be improved by increasing the water rate from the sink. Another merit of the technology that could be observed from the plot in Figure 316 is the flexibility in oil production rate. This is accomplished by varying the drainage water rate at the sink. This allows operators to improve oil recovery over a shorter time and accelerated investment recovery. The improved recovery represents the level of bypassed oil due to water coning in typical conventional completions. This result agrees with the performance of the field trial of the technology in the Nebo-Hemphill well in La Salle Parish in Louisiana (Swicher and Wojtanowicz, 1984).
3.5.2 DWS Well Productivity Index Comparison

The DWS technology incorporates two completions in one well and the target is optimum oil production at the top completion. Consequently, evaluation of its performance compared to conventional single completion presents some difficulty. One approach compares the total fluid (Combined Oil and Water production) productivity index of the DWS well with an equivalent fluid productivity index of a conventional well completion. In effect, we compare the productivity index of the combined fluid production from the two completions in a DWS well with the PI of equal fluid production from a conventional well producing both oil and water. Figure 3-17 demonstrates the

76

results of this approach. In this figure, the field total fluid withdrawal capacity of a DWS well is compared with that of a conventional well. In order to obtain the field productivity index, the total liquid production rate is divided by the field pressure drawdown caused by production from the dual completion. This is compared with PI due to the liquid production and pressure drawdown in the conventional well. The results show that the DWS completion does not necessarily deplete reservoir energy faster than conventional wells for the same liquid production rates. The reason for the above result/observation could be due to the value of vertical permeability of the reservoir. In most reservoirs vertical permeabilities are much smaller that horizontal permeability due to overburden load and compaction effects. Fluid flow involving vertical component of permeability is enhanced in conventional wells than in DWS wells.

Field Productivity Index (stb/day/psi)

8 7 6 5 4 3 2 1 0 0 1000 2000 3000 4000 Time (Days) 5000 6000 7000 8000

Phc.1065-935dws

Phc.C2000

Figure 3-17 Field liquid PI Comparison between DWS Well and Conventional Well

77

The withdrawal of fluid in conventional, partially penetrating, well causes the OWC to move up over time. This upward movement combined with the typically low value of vertical permeability creates increased reservoir pressure drawdown and attendant low fluid production. This causes increased dissipation of reservoir energy for the same fluid rate for the conventional wells as in DWS wells. Also, for the particular reservoir studied, watercut increases rather more dramatically for the conventional well than for the DWS well. The limiting watercut is also greater for conventional wells. In practice, increased watercut is accompanied by a reduction in fluid rate and reduced wellhead pressure. However, for the simulator, the wellhead pressure is kept constant as well as the liquid rate. Thus, it is possible to observe that the effect of increased water production translates to increased drawdown pressure in both cases.

Oil Productivity Index (Stb/Day/psi)

1000

2000

3000 DW S Oil PI

4000
Tim e (Days)

5000

6000

7000

8000

Conv. W ell Oil PI

Figure 3-18 Oil Productivity Index (OPI) Comparison

78

In another more pragmatic approach, the productivity index of oil production (OPI) is compared (Figure 3-18). This approach assumes that the water produced from the sink is environmentally friendly and could be disposed overboard at little or no handling cost. It also assumes that reservoir energy is sufficiently strong. That is, reservoir has strong water drive. Consequently only the valuable oil production should be considered. Thus the oil productivity index is the ratio of oil production rate and pressure drawdown from the top completion compared with the ratio of oil production rate and pressure drawdown for the conventional well. Again, the DWS completion shows tremendous superiority over the conventional completion as shown in Figure 3-18. The corollary of the above results is that, for the same pressure drawdown, the DWS well is capable of delivering more oil than the conventional well without necessarily depleting the reservoir energy much faster.

Total Oil Production (Million stb)

20 18 16 14 12 10 8 6 4 2 0 0 100 200 300 400 500 600 700 Pressure Drawdown, psi Conv.Oil DWS-Oil

Figure 3-19 Oil Recovery Plot for Same Drawdown Pressure

79

In a similar study of cumulative oil recovery over a period of twenty years the DWS well proved to be capable of delivering more oil than the conventional well for the same pressure drawdown in the top completion as in the conventional well. The Figure 3-19 demonstrates the superior oil recovery capacity of the DWS well over conventional completion. The heavy line shows the performance of the DWS well and the thin line shows that of the conventional completion. The diverging trend of the two lines with increasing pressure drawdown indicates that the DWS well performs even better at higher production rates. The two lines also, indicate that total oil production over the period of twenty years (cumulative oil recovery) was consistently better for the DWS well for this particular field. This result further emphasizes a more attractive aspect of the technology, which is the ability to improve oil productivity index by increasing the rate of liquid production at the top. Figure 3-20 shows a plot of stabilized oil productivity index as a function of liquid withdrawal rate at the top completion.
Stabilised PI (stb/day/psi)

7 6 5 4 3 2 1 0 0 500 1000 1500 2000 2500 Fluid Flow Rate, (stb/day) Conv. Compln DWS Compln

Figure 3-20 Improved Oil Productivity Index with Rate

80

The flexibility of being able to increase the rate of oil production by increasing liquid rates from both the top completion and the water sink is shown to improve the stabilized oil productivity index of the DWS well as in Figure 3-20. Again, the diverging trend of the DWS well PI curve over the conventional well curve demonstrates the superior performance of the technology.
3.5.3 Watercut Performance Evaluation

A final parameter used to measure the performance of the DWS water coning control technology is the evaluation of watercut in the top completion compared with a conventional well. In Figure 3-21, the Y-shaped curve represents the cumulative oil recovery with and without the application of the DWS technology and the other prongshaped curve show the reduction in watercut at the top completion due to the deployment of the DWS technology. In this study performed on LVT Well for a JIP member (Joint Industry Panel), the DWS technology demonstrated tremendous potential in reducing watercut in the oil zone completion by a significant level over time.
1 2000000

0 .9

1800000

0 .8

1600000

0 .7

1400000

0 .6

1200000

0 .5

1000000

0 .4

800000

0 .3

600000

0 .2

400000

0 .1

200000

0 0 1000 2000 3000 T im e (D a y s ) D W S W e ll W a te r C o n v . W e ll W a te r C o n v . O il R e c . D W S O il R e c . 4000 5000 6000

0 7000

Figure 3-21 Watercut and Cumulative Oil Recovery Performance

81

This parameter is most useful in cases where water handling capacity of surface facilities is limited. In a typical study, requested by another of the Joint Industry Panel members, surface water handling facilities have reached their designed capacity and watercut from the top oil zone completion must be reduced while producing clean disposable water at the bottom completion. By varying the rate of water production from the sink, the technology is capable of increasing the cumulative oil recovery while decreasing the limiting watercut from the top completion to a manageable level for the surface water-handling facilities. Figure 3-21 represents the results of studies on the LVT well in La Victoria Oil field. The results show that by varying the water production rate at the sink completion, the cumulative oil recovery could almost double while the watercut could be reduced by over 20 percent. This reduction in watercut oil reduces water separation costs and the problem of overload of water treatment/handling facilities. Other studies on the LVT-34 Well in La Victoria Field are summarized in Appendix B.
3.6 DWS Well Design Recommendations for Old Wells

The design and modeling of DWS well completions for old oil wells with

severe water coning history should incorporate the dual effects of capillary pressure transition zone and relative permeability hysteresis. 2 Further, numerical reservoir performance simulators should be used to

effectively evaluate the expected performance rather than relying on semi-analytic methods developed for new wells with sharp oil-water interface. The complexities of capillary pressure transition, relative permeability hysteresis, dual completion with zonal isolation in addition to reservoir heterogeneities and anisotropy can be modeled more accurately using numerical simulators. 82

Where possible, core samples should be obtained to have representative

data, especially capillary pressure, relative permeabilities, porosity, oil-water contact as well as rock and fluid compressibilities. 4 The location of the water sink below the oil water contact is also

important. To minimize reverse coning problems, especially where the produced water must be oil-free for environmental reasons, the water sink should be located at a depth of +/- 20 percent of the total thickness of the water zone below the OWC. 5 Finally, the water sink production rates should be optimized. For a new

well without capillary transition effects, employing the water sink from start of production does not necessarily give the best performance. The sink could be turned on when the water cone is still developing (around 50 to 60 percent watercut) for optimum performance. The rate of drainage water at the sink should also be graduated in steps over time.

83

CHAPTER 4 DWS WELL DELIVERABILITY WITH OIL-FREE DRAINAGE WATER

Several years of oil production in reservoirs underlain by water creates a zone of saturation transition between the original oil-water contact and the zone with connate water saturation as has been discussed in chapter 3. This zone contains mobile oil and water that is capable of flowing co-mingled into either the oil well (top completion) or the water sink (bottom completion) or both completions concurrently. The concurrent production of oil and water in both completions defeats one of the objectives of the DWS technology of producing disposable oil-free drainage water. At the same time, it minimizes the watercut in the top oil zone completion. Production of oil-free water is attractive to the petroleum industry. It can allow operators to meet environmental regulatory requirements for re-injection and overboard disposal for offshore operations. Water quality regulations vary from one country to another and even from state to state. However, the goal of all environmental regulations is that only oil-free water should be disposed overboard or re-injected. The objective of this chapter is to present guidelines for the design and operation of DWS wells in watered out reservoirs (reservoirs with saturation transition) to achieve production of oil-free drainage water. A brief overview of environmental regulations governing the quality of produced water in oil and gas industry is presented in section 4.1.

84

4.1

E & P Produced Water Regulations and Requirements

The Federal Environmental Protection Agency (EPA) believe that the large volume of special waste from E & P operations are lower in toxicity than other wastes regulated as hazardous wastes by the Resource Conservation and Recovery Act (RCRA). Thus, in 1988, the EPA issued a regulatory guideline stating that control of E&P wastes under the RCRA Subtitle C regulations of hazardous wastes is not warranted (EPA530K-95-003). One of the E&P wastes exempted from the hazardous classification is produced water. However, removal of produced water from the hazardous waste classification does not preclude it from emission control. Produced water has to conform to certain minimum requirements. These requirements regulate the level of oil and hydrocarbon contamination of produced water. Table 4-1 gives an overview of the requirements for maximum concentration of hydrocarbons.
Table 4-1 Requirements on Hydrocarbon Contamination of Produced Water

Parameters

Requirements as at 1980 Requirement as at 1993 (Maximum Concentration) (Maximum Concentration) 29 mg/liter

Oil and Grease Concentration (monthly average) Oil and Grease Concentration (Daily average)

48 mg/liter

72 mg/liter

42 mg/liter

The National Pollutant Discharge Elimination System (NPDES) regulates the requirements regarding maximum hydrocarbon contamination for offshore produced 85

water disposal in the United States. This produced water disposal requirements depend solely on the level of hydrocarbon contaminants as in the OCS (Outer Continental Shelf) of the Gulf of Mexico and in the Niger Delta Basin. Average oil content of produced water in Norway in 1998 was 23 mg/l (21 ppm), giving an oil discharge of about 2100 tonnes from about 100 million tonnes of total annual produced water. Aromatic compounds like Benzene and solubles such as Phenols make up additional 1200 tonnes. Polyaromatic Hydrocarbons (PAH) and Alkylphenols cause most worry, amounting up to 49 tonnes in 1998 (Wolff, 2000). Various governments around the world have discharge limits for residual oil in produced water that range from 23-40 mg/l (about 21 40 ppm). Some have also imposed zero discharge limits, which in practice, demands re-injection of produced water. Other state regulations in the US target organic carbons, benzenes, ethylbenzene or toluene.
4.2 Overview of DWS Environmental Performance

The major environmental merit of the DWS technology is the ability to produce oil-free water and water-free oil with the separate completion. This technology not only permits production of environmentally friendly fluids, it counters the cone height growth around the wellbore overcoming the problem of bypassed oil in the reservoir. In offshore environments where environmental regulations are most stringent, clean produced water is disposable overboard saving substantial costs. In some cases, the produced water could be re-injected. The produced water may be injected into the lower part of the producing reservoir for pressure maintenance. The first field application of DWS technology was in the Nebo-Hemphill field located in LaSalle Parish in Louisiana. The environmental performance of the technology 86

was outstanding as shown in tables 4-2 and 4-3 (Swisher, and Wojtanowicz, 1995). An analysis of produced water from the sink and water from conventional completions in the field after one year of production indicated that the level of contamination of the produced water from the sink was below detection limit.
Table 4-2 Hydrocarbon Contamination of Drainage/System Water

Parameter Total Dissolved Solids Oil and Grease

Unit mg/l mg/l

Detection Limit (DL) 1.0 2.0

Drainage Water1 63,300 <DL <DL <DL <DL <DL

System Water1 69,100 484 <DL <DL <DL <DL

Bezene mg/l 1.0 Ethylbenzene mg/l 1.0 Toluene mg/l 1.0 Xylene mg/l 1.0 1 Average of two measurement; 2End point system water

Table 4-3 PAH Contamination of Drainage /System Waters

Component Naphtalene Phenanthrene Fluorene Dibenzothiophene Anthracene Pyrene Other PAH Total PAH ppb ppb ppb ppb ppb ppb ppb ppb

Unit

Effluent from Free Water Knockout; 2Endpoint system water

Drainage Water 11.32 ND ND ND ND ND ND 11.32

System Water Mid-point1 End-point2 536.61 450.38 34.74 26.35 6.66 6.11 12.70 9.54 0.17 ND 0.25 0.23 1.48 0.33 592.61 492.94

System water refers to water samples taken from the separator facilities from conventional wells in the field. This ability to produce oil-free water makes DWS extremely attractive in offshore operations as no water processing is needed prior to overboard discharge. In land operations, oil-free water may be injected into the same well

87

below the drainage completion or lifted to the surface and disposed into injection wells without processing. However, field tests and some studies have also shown that sustainable drainage of oil-free water becomes somewhat difficult as the two completions (top and bottom) may receive co-mingled inflows of the two fluids. The phenomenon indicates that there is an enlargement of transition zone (with mobile oil and water) due to the dynamic production process to a level much larger than that explained by the static capillary pressure effect.
4.3 Capillary Transition and the IPW and Clean Water Drainage

In the previous simulations performed using the linear approximation of capillary transition (Chapter 3), the size of the IPW was found to be smaller for the reservoir with capillary transition zone. The shaded region in Figure 4-1 represents the reduction in the size of the IPW due to capillary transition zone.
A
400 300 Water Rate (bbl/day) 200 100 0 0 50 100 150 200

A
250

O il R a te (b b l/d a y ) H Y S 3 0 ' C A P :W a te r B t ZERO CAP, NO H YS H Y S 3 0 ' C A P : O il B t ZERO CA P, NO HY S

Figure 4-1 Reduction in Size of IPW Due to Capillary Transition Zone

88

A section across the apex of the shaded region, A-A showed concurrent production of contaminated fluid in both completions (Inikori and Wojtanowicz, 2001(B)).

0.2500 0.2000 0.1500 0.1000 0.0500 0.0000 0 200 400 600 Fluid Production at Sink (BLPD)

0.0450 0.0400 0.0350 0.0300 0.0250 0.0200 0.0150 0.0100 0.0050 0.0000 800

Contaminated Fluid Production Belt

WC

OC

Figure 4-2 Shows Zone of Concurrent Production of Contaminated Fluid in both Completion

This observation elicited the need for a better understanding of the mechanism of this concurrent fluid production in both completions. JIP members supplied relevant field data to initiate a numerical study of this phenomenon and to recommend guidelines for oil-free drainage water production.
4.4 Modeling Clean Water Performance of DWS in Old Fields

In order to effectively model the performance of old field with long production history, the inclusion of capillary transition zone to account for the zone of changing water saturation is necessary.

89

Oil Cut in Bottom Compleetion (Fraction)

Water Cut in top completion (Fraction)

4.4.1

Model Description

A 40 x 1 x 23 radial grid model of an actual reservoir, LVT, was used for this study. Reservoir thickness is 85ft (40 ft net oil thickness and 45 ft water zone thickness). The anticline reservoir structure is slanted on either side of the well creating a dome shaped profile. In order to represent this shape in the radial grid model, certain grids were assigned zero porosity and permeability (Figure B-3 in Appendix B).
Table 4-4 LVT Reservoir Input Data INPUT DATA Reservoir pressure Thickness of oil/gas column Thickness of water column Depth of OWC Position and length of production perforations Position and length of water zone perforations Horizontal permeability in oil column Vertical permeability in oil/gas column Horizontal permeability in water column Vertical permeability in water column Water density at temperature Water viscosity at temperature Reservoir temperature Porosity in oil column Porosity in water column Oil formation volume factor Oil gravity / gas gravity Oil/gas viscosity at temperature (@BHP Completion diameter/hole size Unit psi ft ft ft ft ft mD mD mD mD Ib/ft2 cP O F Fraction Fraction Rb/stb API cP inches LVT Reservoir 4200 40 45 9589 9540 - 9552 9596 - 9604 1475 220 1489 225 71.76 0.50 185 0.25 0.27 1.11 32 1.20 7

Also, the reservoir has a strong bottom water drive and high water coning history. Thus, the model assumed a steady-state flow by assigning an infinitely large pore space to the bottom radial grids. Table 4-4 shows reservoir data employed for this study. Capillary pressure transition was also represented in the numerical model to account for the severe water coning history. 90

4.4.2

Capillary Pressure Transition Evaluation

Capillary pressure data from adjacent field with similar reservoir properties were supplied by the operator. Using the Levereth J-function, a capillary pressure transition profile was developed (Figure 4-3). The capillary pressure data from cores presents a useful tool in model a more practical performance of the DWS well for the oil-free drainage water production.

3.00 2.50 2.00 1.50 1.00 0.50 0.00 0.000

0.200

0.400

0.600

0.800

1.000

Water Saturation (Sw, Fraction)

Figure 4-3 Capillary Pressure Plot from Leverett J-function

An alternative approach could be to use linear approximation equations to generate different sizes of capillary transition in order to quantify the effect of varying sizes on water coning performance (Weinstein, et al., 1986, Willhite, 1986, and Yokoyama, 1981). It has been adjudged suitable by researchers for conceptual evaluation studies as enumerated in chapter 3. The results presented in section 4.41 employed the linear approximation approach to obtain various sizes of capillary pressure transition zone. 91

4.4.3 Relative Permeability Curves

Measured relative permeability data from cores in adjacent wells were also provided by the operator (Figure 4-4). The data indicate a water wet sandstone reservoir with a connate water saturation of 27 percent and a residual oil saturation of 19.2 percent. The data supplied represent an imbibition relative permeability measurement from cores in adjacent reservoirs. Consequently, hysteresis effects were not included in this model.

1.200 1.000 0.800 0.600 0.400 0.200 0.000 0.000 -0.200

0.200

0.400

0.600

0.800

1.000

Water Saturation, (Sw Fraction)

Figure 4-4 Relative Permeability Curves from Cores 4.5 Presentation and Discussion of Results

For conventional wells with one completion, fluid inflow to the well expands the size of the capillary transition zone (or swept zone for old oil wells) around the wellbore. The (capillary) transition zone is smallest at the reservoir boundary but increases toward the wellbore as shown in Figure 4-5. 92

Figure 4-5 Transition Zone Enlargement around Conventional Well

The numbers 1 through 3 show an enlargement of the capillary transition zone size from the static OWC toward the wellbore. The expansion is pronounced around the wellbore where the size approximately doubles. Most of the expansion occurs between the wellbore and a few hundred feet.

3 2

Figure 4-6 Transition Zone Enlargement around Dual-completed (DWS) Well

The enlargement effect of the transition zone increases for the DWS completion due to the dual pressure drawdown and diffusion effects (Figure 4-6). Like in the physical model, this diffusion causes a cross-flow expansion of the two fluids to concurrently 93

break through into the two completions. The level of expansion depends on the wellbore drawdown pressure as well as the reservoir flow capacity. An important corollary from this observation is that piston-like displacement concepts cannot be applied to low permeability reservoirs with capillary transition zone nor to old oil wells with severe watered-out zone.
4.6 Design Considerations for DWS Wells With Oil-free Drainage Water

The primary goal of this study was to understand the principles governing the concurrent production of oil and water in both completions and to formulate a design procedure for operating the DWS wells with oil-free water production at the sink. The study addressed the DWS Inflow Performance Window (IPW) under the conditions of diffusive transition zone. For conventional wells without capillary transition zone, the IPW typically consists of four regions (Figure 3-1). A triangular-shaped envelope represents the domain of segregated fluid production (water- free oil production from the top completion and oil-free water production from the water sink completion). This is the target region for environmentally sensitive areas. Below the envelope is the domain of water breakthrough into the oil completion. Above the envelope is the domain of inverse oil coning with oil breakthrough into the water sink completion. A fourth region beyond the apex of the envelope displays a flip-flop line representing a region of instability with some level of contamination in fluid production on either side. The DWS production schedule could also be implemented in the water breakthrough zone. This permits maximum oil production at the top completion with low watercut oil and at the same time produce oilfree water at the sink. 94

For the reservoir with capillary transition zone, this fourth region or the flipflop line representing line of cone instability is changed into an envelope of concurrent fluid production (Region 4 in Figure 4-7). It is characterized by a cross over of the water breakthrough and oil breakthrough lines. Thus, beyond the apex of the segregated fluid production triangle, both oil and water flow concurrently into both completions. A cleanwater zone, however, still exists below the oil breakthrough line (Region 3 in Figure 4-7).
1000 900 800 700 600 500 400 300 200 100 0 0

Water Sink Liquid Rate (BLPD)

Wa ter B r

eak thr

oug h

line

Region 4 A A Region 3
200 300 400

Region 1
2 ion Reg
100

th Oil Break
500

rough line

600

700

Top Liquid Rate (BLPD)

Figure 4-7 IPW Comprising Transition Zone Effect (LVT reservoir); Critical Rate lines Intercept and Diverge at Higher Rates Enveloping Region 4 of Two-phase Inflow at the Top and Bottom Completion

Thus, the oil breakthrough line represents the maximum rate of fluid withdrawal at the water sink completion to ensure production of disposable oil-free water. The task is to identify, a condition for maximum oil rate at the top completion while operating along the oil breakthrough line, i.e., producing oil-free water. An additional plot of watercut isolines has been superimposed on IPW in Figure 4-8. The plot shows increasing watercut with increasing rate of liquid production at the 95

top completion along the oil breakthrough line. Thus, there should be an optimum liquid rate at the top that should maximize oil production rate at optimum watercut along the section A-A of the Oil breakthrough line in Figure 4-7.
1200 Water Sink Liquid Rate (BLPD) 1000 800 600 400 200 0 0 100 200 300 400 500 600 700 Top Liquid Rate (BLPD)

WC, % =

0,

1,

7,

21,

47,

57,

67

Figure 4-8 IPW for LVT Reservoir With WC Isolines

250 Oil Rate (BOPD 200 150 100 50 0 0 100 200 300 400 500 600 700 Top Liquid R ate (BLP D )

A A

Figure 4-9 Optimization of Oil Production Rate for DWS Well With Oil-free Water Drainage

96

A plot of oil production rate Vs total liquid rate at the top completion is shown in Figure 4-9. It gives an optimum top liquid rate at the point of maximum oil rate. For this reservoir, a top liquid rate of 275 BLPD and a bottom water sink completion rate of 285 BWPD would be recommended in order to maximize oil production with minimum watercut at the top completion and, at the same time, deliver oil-free water at the water sink completion.
4.7 Effect of Location of the Water Sink in the Capillary Transition Zone

In another application of the technology by a JIP member, the initial oil-water contact position for the well was unknown. The prolonged history of water coning implies a zone of co-existing mobile oil and water. The task is to investigate the performance of the water sink location in terms of liquid production. The studies showed that locating the water sink completion in the swept zone causes the two completions to concurrently produce contaminated fluid. This is shown in Figure 4-10. The negative values in the y-ordinate represent the location of the water sink (Lower completion) within the zone of saturation transition above the supposed initial oil-water contact. The positive values of the y-ordinate represents location of the water sink below the initial oilwater contact. The result in Figure 4-10 shows that concurrent fluid production is immediately experienced in the two completions. The contamination of water with oil in this case does arise from inverse coning of oil downwards. Rather it arises due the location of the sink with a region of mobile oil and water. The conclusion from this investigation is that the water sink location should be below the original oil-water contact if environmental concerns are paramount.

97

-20
Location above or below the OWC (ft)

-15 -10 -20000 -5 0 0 5 10 15 20 20000 40000 60000 80000 100000

Total oil production (stb)

Figure 4-10 Oil Contamination in Water sink Completion placed within Capillary Transition Zone 4.8 Recommendations for Designing DWS Well and Production Schedule

As explained above, the design of DWS well with oil-free water drainage is entirely controlled by capillary pressure data and the initial size of capillary pressure transition zone. The data constitute initial condition for the equilibrium represented by the IPW. Another control parameter is the placement of water sink completion associated with a risk of drainage water contamination with oil. The following comments arose from the studies: 1. Adequate field data and production logs should be run to understand the extent of water saturation transition development over time and the possible location of the original oil-water contact. 2. A good understanding of field history from start of production and location of original oil-water contact is necessary.

98

3. Capillary pressure data from core analysis within the field or correlation fields should be used to derive suitable capillary pressure data from the Leverett J-function correlation for the pre-installation studies. 4. In the absence of core data, capillary pressure information could be obtained from electric resistivity log responses using a capillary profile match (Ibrahim et. al., 1994). 5. In the last option, linear approximations could be used together with information from the production logs. 6. To avoid early inverse oil cone and oil breakthrough (initial inverse oil cone), the water sink location should be as deep as the limit of water handling capacity allows. The sink should not be installed a few feet below the oil-water contact or in the transition zone where mobile oil can easily flow into the water sink. A height of about 20 percent of the total thickness of the oil or water zone (whichever is smaller) could be used as a rule-of-thumb to determine the depth of the water sink below the original oil-water contact.

99

CHAPTER 5 EFFECT OF FLOW FRICTION ON WATER CRESTING IN HORIZONTAL WELLS

Horizontal drilling technology has been widely recommended by several researchers as a solution for the development of bottom water drive reservoirs experiencing water coning problems (Nzekwu, 1992; Hansen, and Verhyden, 1991). However, several field experiences indicate that horizontal wells do not eliminate the problem of water coning in such reservoirs (Nzekwu, 1992; Tehrani, 1992). In some cases water cresting in horizontal wells erodes the merit of the technology (Heysel, 1992). Modeling of horizontal well inflow performance and fluid flow processes remains a challenge to the industry. Reservoir simulation tools and several theoretical evaluation techniques are largely inadequate, as they tend to neglect the complex flow geometry of horizontal wells and its implication on fluid influx. Most analytical work on horizontal wells has ignored the effect of pressure loss in the wellbore or at most treats it as insignificant (Papatzacos et. al., 1991). Recent research efforts that incorporate the effects of pressure drop in the wellbore stopped at the single-phase flow regime or rather the prewater breakthrough period (Yuan, et. al., 1999; Yuan, 1997; Weipeng, et. al., 2000; Ozkan, et. al., 1993, 1999; and Tang, 2000). Modeling of fluid flow behavior and pressure drop in horizontal wells in bottom water drive reservoirs requires the incorporation of several important factors. These include: (1) effect of two phase oil-water flow in the horizontal well after water 100

breakthrough, (2) understanding of the flow regimes of immiscible liquid-liquid flow in horizontal wells and the formation of emulsion, (3) the effect of emulsion viscosity on the pressure drop along the well, (4) the effect of perforations/slots along the well and implication on pipe roughness, (5) the implication of axial influx of fluid into the wellbore. In this chapter, the analysis of the complex problem of water coning in horizontal wells, (also known as water cresting because of the development along the well length) will be presented along with all of the critical fluid and well length parameters that affect the phenomenon.
5.1 Limitations of Water Coning Control With Horizontal Wells Field Evidence

One of the earliest references to the problem of water cresting in horizontal wells underlain by water was the performance test by Elf Aquitaine (Geiger, Reiss and Jourdan, 1984). The well Lacq 90, (in the Lacq Superieur field) produced at 19 BOPD in 4 years with a watercut of 99 percent. This report indicates that horizontal wells are not a complete solution to the problem of water cone development around producing oil wells. Guo et. al (1992) observed that neglecting pressure loss in the wellbore due to friction as contained in most theoretical analysis is a major limitation of theory in field application. The frictional pressure loss causes a gradient in the pressure drawdown in the formation along the horizontal wellbore and the implication of this is that the water crest is higher at the down-stream (heel) end than at the upstream (toe) end. Thus, water breaks through first at the heel and further oil production causes the water breakthrough to advance towards the toe. The implication is that the infinite conductivity model of

101

horizontal well analysis does not conform to practical evidence and the critical oil rate predicated from such analysis is usually over-estimated. Figure 5-1 is a cartoon of the profile of water crest when the effect of pressure loss in the wellbore is neglected and assuming a uniform flux model. It shows that water crest development is uniform along the well length.

Figure 5-1 Uniform Flux Model of Horizontal Well shows Uniform Water Crest Development along the Length of the Well

Figure 5-2, shows another cartoon of a more realistic profile of the water crest in horizontal wells. The water crest profile is skewed toward the heel and so water breaks through at the heel and then advances toward the toe.

Figure 5-2 Finite conductivity Model of Horizontal Well (Non-uniform Crest Development)

102

In the first major test of application of horizontal well in the Troll field, in the Norwegian sector of the North Sea, the producing horizontal section was completed with an 8-inch pre-packed slotted liner to reduce wellbore friction pressure loss. The productive section of the well length was 1640 ft in a reservoir with permeability range of between 3,000 10,000 mD. The cleanup production test indicated a PI above 6000 stock-tank m3/day/bar (2600 stb/day/psi) and a kV/kH ratio of 0.15. The result of production logging test showed that 80 percent of the well was initially open to flow and that 75 percent of this inflow took place in the first half of the completed interval (Lien, Seines, Havig, and Kydland, 1991). Lien et. al. believed, however, that the unproductive 20 percent was caused by insufficient cleanup in certain intervals. Several other research efforts have shown that horizontal well productivity is highest at the heel than at the toe. The Obagi field experience with laminated thin reservoirs indicates that horizontal wells may not be successful in such reservoirs with bottom water drive (Chugbo, Roux and Bosio, 1989). Mobils experience with a horizontal well to mitigate water coning in the Ness field, UK sector, was reported to be unsuccessful. In a field where average watercut from conventional vertical wells was 40 percent, the completed horizontal well showed rapid rise in watercut averaging nearly 65 percent despite all careful well planning efforts. They, however, attributed the performance to geological uncertainties (Koonsman and Purpich, 1991). Whole oil fields developed with horizontal wells are increasingly experiencing severe water cresting problems and mechanical intervention in these wells is currently very limited compared to vertical or deviated wells. In summarizing this section, it is important to state that the use of horizontal wells in the development of reservoirs with bottom water drive does not eliminate the problem 103

of water influx and that at the same time, the water crest development in horizontal wells is rather skewed toward the heel of the well and not uniformly developed. These important observations form a critical basis for solving the problem of water crest development in horizontal wells.
5.2 Model of Oil-Water Flow in Horizontal Wells

Effective modeling of the performance of producing horizontal oil wells subject to water cresting requires evaluating the following concepts: (1) flow regimes and reservoir conditions of producing horizontal oil wells, (2) effect of the flow of oil-water mixture on the fluid properties under reservoir conditions, (3) effect of axial fluid influx into the horizontal well with fluid flow along the wellbore and the nature of wellbore friction pressure loss in horizontals wells.
5.2.1 Flow Regimes and Conditions in Horizontal Wellbore

Presently, no characterization work has been done on the phase distribution of oilwater flow in horizontal wells. All of the existing research on phase characterization relates to liquid-liquid flow in horizontal pipes. Horizontal wells have the added dimension of perforations and radial fluid influx creating increased turbulence to the main flow. Mixture of Oil-water flow in horizontal pipes has been characterized into two broad categories, namely, segregated flow and dispersed flow. Segregated flow shows a continuity of both phases in the axial direction while dispersed flow shows a disruption in the continuity of any of the phases. Another broad characterization of oil-water flow is the definition of the flow pattern as either oil-dominated or water-dominated. Trallero et. al. identified six flow patterns for oil-water flow in pipes under the two broad categories 104

of segregated flow and dispersed flow (Trallero, et. al., 1996). They observed that segregated flow could be either stratified flow or stratified with some mixing at the interface. Dispersed flow was divided into two headings: (1) water dominated dispersed flow made up of, (a) dispersion of oil in water, and (b) emulsion of oil in water, and (2) oil dominated dispersed flow comprising of (a) emulsion of water in oil, and (b) dual dispersion. For the specific experimental set up, an average mixture velocity of 0 7.65 ft/sec, produced stratified flow and between 7.65 ft/sec and 13.12 ft/sec or higher produced dispersed flow (Trallero, et. al., 1999). These flow velocities corresponds to the laminar and turbulent flow regimes for the experimental setup. Flores et al (1999) reported that no segregated flow exists in deviated pipes with inclination greater than 33 degrees. The experiment did not represent the actual profile of horizontal wells as fluid influx was neglected in the property measurement section. Dikken observed that the typical flow regime encountered in horizontal oil wells are more likely to be turbulent than laminar considering the typical production rates of such wells (Dikken, 1990). An inversion point is reached at a certain concentration of water where the flow pattern changes from oil-dominated to water-dominated regime. This point called phase inversion has been typically characterized by a change in mixture viscosity as will be discussed later. Another important aspect of the phase inversion phenomenon is that friction pressure loss in the pipe is critical in the oil-dominated flow period while slippage between the liquids is critical in the water-dominated flow regime. Arirachakaran et. al. developed a correlation for fluid phase inversion from oil dominated flow to water dominated flow in the fully laminar region (Arirachakaran, S., et. al., 1989): 105

f w, INV = 0.500 - 0.1108 log m

5-1

The evaluation of the above relation agrees with the observation of Soleimani et. al. that a phase inversion seems to occur at a watercut of between 35 and 46 percent from oil-dominated flow to water-dominated flow (Soleimani, A., et. al., 1999). However, the phase inversion and general characteristics of the oil-water flow regimes depend on several other factors such as oil-water mixture viscosity, fluid mixture velocity, and watercut, amongst others. Combining all of the above analysis with the additional agitation created by the radial fluid influx in typical reservoir conditions, the author has assumed a dispersed flow model for oil-water flow in horizontal wells as more representative of field experiences. Also, typical reservoir flow conditions are elevated temperatures in the magnitude of over 175o F. The combination of the reservoir conditions and the axial influx of fluid into the flowing wellbore create the conducive environment for the formation of unstable emulsion. In fact, Schramm observed as follows: If two immiscible liquids are mixed together in a container and then shaken, examination will reveal that one of the two phases has become a collection of droplets that are dispersed in the other phase; an emulsion has been formed(Schramm, (ed.), 1992). The emulsion formed by the

agitation process of the two immiscible fluids under horizontal well conditions (oil and water) are, however, unstable and can separate into their various individual components when the emulsifying conditions are removed. Stable crude oil emulsions may form in systems containing mixtures of crude oil and formation water, either as a result of sudden pressure drop across a choke valve on the wellhead or as a result of turbulent mixing in the wellbore or in transport pipeline (Ronningsen, 1995). This basic principle has been 106

widely acknowledged in the petroleum industry but has not been applied in the mathematical model developments. The flow of oil and water mixtures under reservoir conditions in horizontal wells with axial influx of fluid creates a high level of turbulence conducive for the formation of emulsions. The turbulence created by the influx of fluid coupled with the typical high production rates of horizontal wells encourages the dispersion of water in oil below the inversion point and also reduces the droplet sizes of the dispersed phase a condition necessary for increased mixture viscosity. Thus oilwater flow in horizontal wells is best modeled as unstable emulsion and is very different from the flow of oil and water in pipes at surface conditions. Consequently, the results of this study have assumed that the oil-water mixture in horizontal wells behave like unstable emulsion.
5.2.2 Properties of Oil-water Mixture in Horizontal Wellbore

The fluid mixture properties have been typically evaluated using weighted averages of the individual phase properties. Thus the mixture density, mixture surface tension, and the mixture velocity are calculated as a weighted average of the concentration of each phase. This assumes, a no-slip condition. Thus mixture density is:
r m = r w * f w + r o * (1 - f w ) 5-2

Oil-Water mixture velocity:

vm = vsw * f w + vso * (1 - f w )
5-3

where superficial velocities are: Q(o,w)/A Oil-Water mixture interfacial tension: 107

s m = s w * f w + s o * (1 - f w )

5-4

However, the relationship for the mixture viscosity is not as simple.

5.2.2.1 Oil-Water Mixture Viscosity Martinez et. al. observed that accurate prediction of the apparent mixture viscosity

of oil-water mixture is a difficult one because the mixture exhibits non-Newtonian rheological behavior and so their apparent viscosities can not be simply calculated by using standard weighted average rules (Martinez, et. al., 1988). They developed a rather specific relation for evaluating the mixture viscosities for the range of oils studied. The viscosity of an emulsion generally depends on: 1 Volume fraction of the dispersed phase 2 Viscosity of the continuous phase 3 Shear rate (if non-Newtonian) 4 Temperature 5 Viscosity of the dispersed phase 6 Nature and concentration of the emulsifying agents 7 Average droplet size and size distribution 8 Presence of solids in addition to dispersed liquid phase 9 For water-in-crude oil emulsions the viscosity of the dispersed phase (formation water) can be regarded more or less constant. . A series of experiments conducted in the present study (Ronningsens) in fact indicate a fairly small effect (of droplet size distribution) even at 50% watercut. The viscosity of an emulsion is directly proportional to the viscosity of the continuous (oil) phase. The most important factor that influences the emulsion viscosity, the volume fraction of the dispersed phase (water). (Ronningsen, 1995) Other researchers in the chemical industry have performed extensive research on the viscosity of oil-water mixtures (Broughton, and Squires, 1937; Richardson, 1932; Schramm, (ed.), 1992). The modified Richardson equation (Richardson, 1952) is widely used in the chemical industry.

m m = mo * e b + K * f w
108

5-5

m is viscosity, K is the Richardson constant, b is also a constant, which depends

on the system and fw is the concentration of the dispersed phase in the continuous phase (water fraction for water in oil emulsion). Rnningsen performed extensive research on viscosity of oil/water emulsions for samples of North Sea crude and developed a Richardson-type relation for the crude samples studied (Rnningsen, 1995). However, the more generalized and more recent relation in the chemical industry for oil-water mixtures is the Pal and Rhodes relation (Schramm, (ed.), 1992).
f w / f w,lim mo mr = m = 1 + 1.187 - ( f / f w w ,lim m
2.49

5-6

where the fw,lim. is the dispersed phase concentration at which the relative viscosity becomes 100. The Pal and Rhodes relation correlates both Newtonian and nonNewtonian emulsions and the normalization of the watercut with the limiting watercut takes into consideration the effects of other forces other the hydrodynamic forces. Figure 5-3 presents an overview of the divergence between mixture viscosity predicted by the Pal and Rhodes relation and the current industry approach of weighted average. The hatched line (tagged Ronningsen in the legend) is the correlation derived from the experimental data on a wide spectrum of North Sea crude. The thin line (Pal & Rhode in the legend) is the mixture viscosity from the Pal and Rhodes equation and the heavy line (tagged Wt Avg in the legend) is the mixture viscosity from the weighted average approach.

109

1000.0 Mixture Viscosity (cP) 100.0 10.0 1.0 0.1 0.00

mo = 1.34 cP mw = 0.5 cP

0.20

0.40

0.60

0.80

1.00

Watercut in Oil (fraction)

Wt. Avg P al & Rhode Ronningsen

Figure 5-3 Oil-Water Mixture Viscosity Evaluation

The Figure show the dramatic increase in oil-water mixture viscosity with increasing watercut using the Pal and Rhodes relation compared to the current weighted average approach employed by most commercial numerical simulators and researchers (Schlumberger, Geoquest, 1997; Seines, et. al., 1993). As a check, the performance of mixture viscosity with increasing watercut as measured by Rnningsens experiment for the North Sea oil with the same properties has been included for comparison. The results of the experimentally derived correlation compares more closely with that of the Pal and Rhodes relation than the current weighted average approach widely used by the industry. Thus, the results presented in this research have incorporated the Pal and Rhodes relation.

110

5.2.2.2 Effect of Mixture Viscosity on Reynolds Number/Friction Factor The effect of the mixture viscosity will be presented in terms of the impact on

Reynolds number and the friction factor. Three equations have been widely used by the petroleum industry in evaluating friction factor in pipe flow as well as horizontal wells. They are the Colebrook relation, the Haalands equation and the Blasius equation for smooth pipes. The Colebrook equation involves iterative processes and requires more computer time. The Haalands relation incorporates the in-situ pipe roughness function and is a one-step solution. The Blasius equation is a one-step solution but it is developed for smooth pipes and for Newtonian fluids. Most commercial pipes used in the petroleum industry have in-situ roughness. Neglecting this pipe roughness function reduces the accuracy of the calculated friction factor with increasing flow rate. Thus, most commercial simulators and researchers have found the Haalands equation more useful and potable. The Colebrook, relations is given as (Bourgoyne, (Jr.), et.al., 1991):
1
1.255 = -4 log 0.269 e + D f N Re f

5-7

The equation involves the friction factor on both sides and so requires an iterative process for its evaluation. The Haalands equation, on the other hand does not involve iteration. Thus numerical simulators have found it more efficient (VIP-Executive Reference, 1996).

6.9 e 10 9 1 = -1.8 log + N Re 3.7 D f

5-8

111

The VIP executive reference manual also stated that the pipe absolute roughness,
e, can represent the effect of turbulence created by inflow through perforations.

The Blasius equation represents a straight-line approximation of the Colebrook function on a log-log plot (Bourgoyne, (Jr.), 1991).
f =
0.0791 0.25 N Re
300000 Reynold's Number (Dimensionless) 250000 200000 150000 100000 50000 0 0.00 0.20 0.40 0.60 0.80 1.00 1.20

5-9

Q = 5,000 stb/day; rw = 8.58 ppg; ro = 6.67 ppg; mw = 0.5cP; mo = 1.34cP

Water Cut (Fraction)

Weighted Average Emulsion, Pal & Rhodes

Figure 5-4 Effect of Mixture Viscosity Model on Reynold's Number

All of the above functions assume a turbulent flow model for the typical oil rates encountered in horizontal wells. As an example Figure 5-4 presents the effect of correct application of the mixture viscosity on the Reynolds number. The effect of Reynolds number on friction factor is in turn presented for smooth horizontal pipe, using equation 5-9. The Figures compare the results of the Pal and Rhodes relation for mixture viscosity with the weighted average approach as a function of increasing watercut. 112

The results show how dramatic the friction pressure loss can be when the appropriate model of liquid-liquid mixture viscosity is used. As expected the decreasing trend of the Reynolds number in Figure 5-4 translates to an increase in the friction factor and consequently, an increase in the pressure drop along the wellbore as watercut increases.

0.01 200

FanningFriction Friction Factor Fanning

0.01 000 0.00800 0.00600 0.00400 0.00200 0.00000 0.00 0.1 0

Q = 5,000 stb/day; rw = 8.58 ppg; ro = 6.67 ppg; mw = 0.5cP; mo = 1 .34cP

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1 .00

W ater Cut (Fraction) W eighted Average Emulsion, Pal &Rhodes

Figure 5-5 Effect of Viscosity Model on Fanning Friction Factor

The continuous line represents the friction plot with increasing watercut using the Pal & Rhodes equation and the hatch line represents the expected friction factor with increasing watercut using weighted average considerations. The diverging trend confirms current observation by researchers that increasing watercut increases the mixture viscosity until the inversion point where the continuous 113

phase changes from oil to water. The observation, made here, about the inadequacy of the weighted average concept of mixture viscosity is also confirmed by Arirachakaran et. al. in a study on oil-water flow in horizontal pipes (Arirachakaran, et. al., 1989). The concept of phase inversion from oil-dominated flow regime to waterdominated flow at about 35 50 percent watercut, however, reduces the effect of the increasing friction factor since friction pressure loss gives way to slippage between the fluids in the water-dominated flow regime. This effect of increased slippage is typically not included in most computer models as these models assume homogeneous fluids.
5.2.3 Frictional Pressure Loss in Horizontal Wellbore

In general, pressure loss in pipes/wells consists of several components (Brill, 1999).

dP dP dP dP + + = dL Total dL accelerati on dL elevation dL friction

5-10

For horizontal pipes/wells, the elevation term is usually neglected (equal to zero).

rvdv dP = dL acc g c DL

is the kinetic energy term,

5-11

This term representing the acceleration of flow due to the fluid influx has frequently been evaluated and incorporated in the friction pressure loss term.
fr v 2 dP , represents the (wall) friction component, = dL fric 2 g c D

5-12

For single phase flow, the friction factor, f, is a function of pipe roughness, e, and the Reynolds number, NRe. For two-phase flow, other factors such as fluid interaction may become important beside pipe roughness and Reynolds number. 114

Thus, introducing the subscript, m, for mixtures, the equation for the mixture friction pressure loss becomes:
2 f tp r mvm dP = 2 gc D dL fric

5-13

In field units, equation 5-13 is written as:

fr v 2 dP ( field .units ) = dL fric 25.8 D

5-14

Until recently, analysts have maintained that friction pressure loss in horizontal well is very small compared to the reservoir pressure drawdown and could be ignored (Joshi, 1991). The reason for this conclusion is because the mathematical models developed for horizontal pipes do not consider other effects such as perforations and axial influx of fluid into the wellbore. These additional factors could increase the friction factor by over 500 percent as will be seen from the results of this research presented in the succeeding sections of this chapter. More importantly, result from field experiences of horizontal well performance does not support the fact that friction pressure loss is negligible. Su and Gudmundsson (1993) investigated the effect of perforation pressure loss in pipes for single-phase fluid. They did not include the effect of fluid influx. However, their results indicated that perforation effect could increase the friction pressure loss by as much as 15 percent or more. Su and Gudmundsson developed a three-step mathematical model for evaluating the friction pressure loss in perforated pipe (Su, and Gudmundsson, 1993). The author employed the relation of Su and Gudmundsson to evaluate the effect of perforation on friction factor.

115

0.01 200 0.01 000 0.00800 0.00600 0.00400 0.00200 0.00000 0 20000 40000 60000 80000 1 00000 1 20000 2 spf; D = 4-1 /2-inch; d = 0.5-inch; j=1 .34 cP ; m w = 0.5 cP; mo = 1 r o = 6.67ppg; r w = 8.59ppg;

Fanning Friction Factor

Reynold's Number (Dimensionless)

Su, Gudmunsson B lasius (Smoo th pipe)

Figure 5-6 Effect of Perforation Without Fluid influx on Wellbore Friction Factor

The result show an increase of about 35 percent compared with the Blasius relation for smooth pipe. This result is presented in Figure 5-6. Dikken performed an extensive research on the effect of pressure drop along the wellbore on horizontal well performance. He developed a mathematical expression for flow resistance, Rw, which incorporates friction factor for turbulent flow:
pm D 8 r Rw = 0.316 p 2 D5 4r
a

5-15

Dikken analyzed the pressure drop along the wellbore as:

pm D 8r d 2 -a * 2 5 Q ( x) 2 Pw ( x) = RwQ( x) = 0.316 dx 4Qr p D
a

5-16

Expressing the equation 5-16 in terms of the Blasius expression for Moody friction factor gives: 116

0.316 8r 8r d 2 -a * 2 5 Q ( x ) 2 = f M * 2 5 Q( x ) 2 Pw ( x) = RwQ( x) = a ( N Re ) p D p D dx

5-17

fM is the Moody friction factor, D is the pipe diameter, r is the density of the fluid mixture, Q is the flow rate. While Dikken did not incorporate an expression for the fluid influx, he observed that the Reynolds Number exponent, a, in the Blasius equation for friction factor could vary from 0.15 for perforated pipe to 0.25 for smooth pipe (Dikken, 1990). This translates to an over three fold (three hundred percent) increase in friction factor for perforated pipe over smooth pipe. The result of Dikkens research further corroborates the authors observation that the equations of friction pressure loss in pipes are quite inadequate in evaluation of horizontal well performance. Ashiem and Kolnes developed a correlation, from experimental studies, for friction pressure loss in horizontal wells. The correlation accounted for the influx of fluid into the wellbore. The expression assumes a fixed value for the ratio of fluid influx to the main flow rate (Ashiem, et. al., 1992).
2 0.16 qi D qi + + 4 D 2 0.19 Q n Q N Re

fT =

5-18

qi is the fluid influx rate, Q is the main flow rate, D is the pipe diameter and fT, is the total friction factor, and n is the number of perforation. The expression for wall friction follows the Blasius-type relation for smooth pipe and so neglects the effect of the pipe in-situ roughness. Recent research efforts have included the effect of fluid influx and perforation for single-phase fluid (Yuan, et. al., 1999; Yuan, 1997). 117

f T = f w + Cn * 2 D * j *

qi Q

5-19

The authors used a Blasius-type expression for the wall friction factor as:

a b N Re

5-20

Table 5-1 Experimental Values for the Constants, a, b and Cn

No. 1 2 3* 3# 4 5* 5# 6 7* 7# 8 9 SPF 5 5 5 5 10 10 10 10 20 20 20 20 PHA 90 180 360 360 90 180 180 360 90 90 180 360 a 0.873 0.622 0.641 0.590 0.159 0.363 0.368 0.755 1.297 0.590 4.532 1.078 b 0.341 0.287 0.312 0.312 0.155 0.266 0.271 0.308 0.421 0.339 0.505 0.346 Cn 2.344 2.028 2.169 2.300 1.489 2.200 2.300 3.901 2.200 2.300 2.332 2.694

# Data from Yuans correlation and * represents data from Yuans experiments Table 5-1 represents values for the constants a, b and Cn presented in equations 519 and 5-20 (Tang, et.al., 2000). Evaluation of the results of the experiment indicates that pressure the friction factor could increase by as much as four fold or more depending on the ratio of fluid influx to main flow rate (Figure 5-7). Tang and Yuan research work used a Blasius-type relation to evaluate the pipe wall pressure loss and does not incorporate an expression for pipe roughness function. 118

0.035000 0.030000

Fanning FrictionFactor Factor FanningFriction

0.025000 0.020000 0.01 5000 0.01 0000 0.005000 0.000000 0 20000 40000 60000 80000 1 00000 1 20000 Reynold's Number (Dimensionless) Y uan, et. al. Blasius (Smooth pipe) /2-inch; d =0.5-inch; mo =1 .34 j=5 spf; D =5-1 cP; mw=0.5 cP; ro =6.67ppg; rw=8.59ppg; PHA= 3600 , qi/Q =1 /200

Figure 5-7 Effect of Perforation and Fluid Influx on Wellbore Friction

Again, the Blasius relation for friction pressure loss is applicable to Newtonian fluids only. Also, the correlations developed by Yuan et. al.(1999), did not quite produce the constants obtained from the experimental data (Yuan, 1997). Finally, the experimentally derived correlations are case specific and require that a specific set of constants be applied for different flow conditions. One task of this research therefore, is to develop a generalized compound friction pressure loss relation that effectively models the performance of horizontal wells and reduces to the normal horizontal pipe flow relations for both commercial and smooth pipes.

119

5.2.4 Generalized Friction Factor Correlation for Horizontal Wells and Pipes

The first step was to modify the Su-Gudmundsson relation to accommodate the effect of fluid influx. This was achieved by introducing the influx to main flow ratio concept from the Asheim and Yuan experimental data as well as a numerical constant. The influx to main flow ratio is actually not constant but varies along the length of the wellbore. The development of an adequate mathematical expression that effectively models the ratio of influx to main wellbore flow remains a research area. However, the influx to main flow ratio at the ultimate production rate is typically used for evaluation. The evolving relation requires a three-step solution as follows: 1. Employing Haalands equation for friction factor, determine the Moody friction factor due to the in-situ pipe roughness, fMs.
1.11 6.9 e = - 1.8 log + N 3 . 7 D Re

-2

MS

5-21

2. Employing, fMs, from step 1 above calculate a constant, Av, in the universal velocity distribution law for smooth wall.
N 8 Re Av = 2 . 5 ln ms 2 f f Ms 8 + 3.75

5-22

3. Finally, by iteration, evaluate the Moody friction factor for the perforated pipe with fluid influx using the following equation.
N qi f MRI Re m + A - 7.0d p - 3.75 = 82.5 ln + 4.0 Dj v Q D 2 8
-2

MRI

5-23

The Fanning friction factor could be calculated by dividing the Moody friction factor from step 3 above by the factor of 4.0. (f (Fanning) = fMRI/4). The component on

120

the right hand side of the plus sign in equation 5-23 has been added to the Su and Gudmundsson relation by the author.
5.2.5 Verification of Generalized Friction Factor (GFF) Relation

The next step involves verifying the accuracy of the generalized correlation. The experimentally derived correlations of Yuan and the constants from her experiment, (Table 5-1) presents the most recent and detailed analysis and were considered an excellent calibration tool to verify the performance of the correlation for producing horizontal wells. The results show a very good fit with a maximum error margin of +/7.0 percent. Two plots of the verification exercise are presented below in Figures 5-8 and 5-9.

Factor Fanning FrictionFriction Fanning Factor (Dimensionles s)

0.0350 0.0300 0.0250 0.0200 0.01 50 0.01 00 0.0050 0.0000 0 20000 40000 60000 80000 1 00000 1 20000 Reynolds Number (Dimensionless) Yuan,Sarica Mod. Su,Gudmunsson /2-inch; d = 0.5-inch; j= 5spf; D = 5-1 = 1 .34 cP; mo mw = 0.5 cP; PHA =360o = 6.67ppg; /1 00 ro rw = 8.59ppg; qi/Q = 1 Max. Deviation = + /- 3 %

Figure 5-8 Friction Factor Plot from 'Generalized Equation' Vs Yuan's Experimentally Derived Correlations (5SPF)

121

The hatched line represents the plot of friction factor from the experimentally derived correlation of Yuan (Yuan, 1997) and the solid line represents the results of the generalized friction pressure loss relation for the same fluid properties and flow conditions. Figures 5-8 and 5-9 also show that beyond a certain level of turbulence the dependence of friction pressure loss of the flowing fluid on the Reynolds number decreases. The Fanning friction factor plot becomes almost horizontal indicating that friction pressure loss depends less on the fluid rate and more on the pipe roughness. Both plots also show that the friction factor is greater than 0.025 for all turbulent flow rates for the particular combination of fluid properties and wellbore geometry.

Fanning Friction Factor (Dimensionless)

0.0350 0.0300 0.0250 0.0200 0.01 50 0.01 00 0.0050 0.0000 0 20000 40000 60000 80000 1 00000 1 20000 Reynolds Number (Dimensionless) Yuan,Sarica Mod. Su,Gudmunsson 0spf; D = 5-1 /2-inch; d = 0.5-inch; j= 1 = 1 .34 cP; = 80o mw 0.5 cP; PHA =1 mo /200 ro = 6.67ppg; rw = 8.59ppg; qi/Q = 1 Max. Deviation = +/- 3 %

Figure 5-9 Friction Factor Plot from 'Generalized Equation' Vs Yuan's Experimentally Derived Correlations (10SPF)

122

Again, the plot in Figure 5-9 show a good fit within a maximum error margin of +/- 3.0 percent. Most of the error occurs in the low Reynolds number region indicating that there is more perfect fit with increasing turbulence. The close fit with the experimentally derived correlations and the generalized nature of the relation developed in this research makes it more attractive. The third step involves testing the generalized equation with the Blasius equation for smooth pipe in the absence of perforations and fluid influx. Two steps were taken. The first involves checking the generalized friction factor relation for commercial pipes with the Blasius relation for smooth pipes (The Blasius equation neglects the effect of the in-situ pipe roughness). The second step involves testing the performance of the GFF for a smooth pipe with zero in-situ pipe roughness.

Fanning Friction factor (Dimensionless)

0.0100 0.0080 0.0060 0.0040 0.0020 0.0000 0

D = 5-1/2-inches, rw = 8.59 ppg, ro = 6.67 ppg, mw = 0.5 cP, mo = 1.34 cP, e/D = 0.0004

50000

100000

150000 Blasius eqn

Reynold's Number (Dimensionless) Mod. Gudmunsson (With In-situ pipe roughnes)

Figure 5-10 Performance of 'Generalized Equation' With in-situ pipe Roughness Vs Blasius Equation for Smooth Pipe

123

The result in Figure 5-10 show the amount of error introduced in the evaluation of friction pressure loss in commercial pipes when the Blasius equation for smooth pipe is used (as much as 10 15 percent error). This error increases as the fluid flow rate is increased. This is demonstrated by the diverging trend toward the right. In the next Figure, the performance of the generalized equation (also known as modified Gudmundsson equation) is compared with the Blasius equation for smooth pipes. The plot shows an excellent fit for both cases. In Figure 5-11, the result from the equation is compared with the Blasius equation for smooth pipe. The result shows a close agreement (within +/- 3 percent error) for smooth pipe consideration. Thus the relation could be applied to both horizontal wells and pipe flow.
Fanning Friction factor (Dimensionless)

0.0100 0.0080 0.0060 0.0040 0.0020 0.0000 0

D = 5-1/2-inches, rw = 8.59 ppg, ro = 6.67 ppg, mw = 0.5 cP, mo = 1.34 cP

50000

100000

150000 Blasius eqn

Reynold's Number (Dimensionless) Mod. Gudmunsson (No pipe roughnes)

Figure 5-11 Performance of 'Generalized Equation' Vs Blasius Equation for smooth pipe

124

An important observation from the foregoing analysis is that while the friction factor in horizontal wells with perforations and fluid influx effects range from 0.025 to 0.032, the friction factor for smooth pipes range from 0.004 to 0.008 representing an increase of over 400 percent to 625 percent. This shows the magnitude of error introduced when friction factor relations for smooth pipes are adopted for the modeling of horizontal wells.
5.3 Numerical Study of Water Cresting in Horizontal Wells With Compound Frictional Pressure Loss Model

The essence of the studies on effective evaluation of friction pressure loss in horizontal wells, incorporating all of the above effects, is to effectively model horizontal wells subject to bottom water drive. In this section, the results of application of the new compound friction factor in modeling water cresting will be presented. Users cannot edit the codes of most commercial numerical simulators. Therefore, the approach adopted was to back-calculate an equivalent pipe roughness that will give the same friction pressure loss effect as the compound friction loss model. This new pipe roughness function is then input into the numerical simulator. The Colebrook equation (Bourgoyne, (Jr.), et. al., 1991) is re-arranged to solve for the new value of pipe roughness for use in the simulator.
e = 3.72d 10 -1 / 4 1.255 N Re f

5-24

Typical equivalent pipe roughness of the order of 0.015 to 0.1525 ft were calculated compared with conventional values of 0.000065 ft. Appendix F shows sample

125

values of equivalent pipe roughness calculated from the compound friction pressure loss relation. Also, to accommodate the dual concepts of phase inversion and mixture viscosity, the new compound friction factor is calculated with the modified Reynolds Number for mixture viscosity up to the point of inversion. Beyond the inversion point (Water dominated flow regime) mixture viscosity is calculated with the conventional weighted average method. The phase inversion concept is modeled in the numerical simulator by re-completing the well at the watercut corresponding to the phase inversion point with a new value of pipe roughness. The results in Figure 5-12 show the tremendous increase in friction factor between the new approach and the conventional weighted average used by the commercial numerical simulators.
0.06000 0.05000 Fanning friction factor 0.04000 0.03000 0.02000 0.01000 0.00000 0.00
Q=9,000 stb/day;Do=5-1 /2-in, Di = 4.92-in, qi/Q =1 /200; qi/Q =1 /1 00 (2), j = 1 0 spf; rw = 8.59 ppg, ro = 6.67 ppg, mw = 0.5 cP, mo = 1 .34 cP, d(perf.) = 0.5-in Lw = 1 500 ft

0.20

0.40

0.60 Water Cut (Fraction)

0.80

1.00

1.20

Modified Factor (qi/Q=1/200)

Eclipse Wt. Avg

Modified Factor (qi/Q=1/100

Figure 5-12 Friction factor comparison (Model Vs weighted average) in oil-water flow with phase inversion

126

Also the Figure 5-12 shows that the inclusion of perforation effect and axial fluid influx introduces a tremendous increase in equivalent friction pressure loss in the horizontal wells. These dual effects should not be ignored. The friction factor increase could be as much as six fold depending on the ratio of influx to main flow. Beyond the phase inversion point, the friction factor for both cases employed the weighted average approach. The wide margin, therefore in Figure 5-12 between the Eclipse weighted average plot and the compound friction weighted average plot with perforations and fluid influx beyond the point of phase inversion shows the components introduced by the dual concepts of perforation and axial fluid influx. Also evident from the Figure is the fact that the concept of phase inversion effect is over-shadowed by the dual effects of perforation and axial fluid influx in horizontal wells.
5.3.1 Changes in Horizontal Wells Water Crest Development Due to New Model

To present a clear picture of the effect of friction pressure loss on water cresting in horizontal wells, the compound friction model was applied to the numerical simulation of water cresting. The application of the new model affected three primary aspects of water crest development (Inikori, and Wojtanowicz, 2002). A fourth point, which is a corollary of the crest profile, is the level of bypassed oil in horizontal wells due to water cresting. These effects will be presented in the subsequent sections.
5.3.1.1 Time to Water Breakthrough Previous analytical tools that ignored the effect of pressure loss along the

wellbore or used the friction pressure loss relations for horizontal pipes seems to overestimate the time to water breakthrough. As shown in Figures 5-13 and 5-14, the time to water breakthrough for the reservoir studied reduced from 150 days to 70 days. 127

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 200 400 600 Time (Days) HW-0.1100-5000 HW-0.00065-5000 800 1000 1200

Figure 5-13 Early Water Breakthrough in Horizontal Wells

150 days has been predicted using the Eclipse Fanning friction pressure loss relation for pipe flow as well as the weighted average option for fluid mixture viscosity. Figure 5-14 is a highlighted version of Figure 5-13.
0.6 Water Cut (Fraction) 0.5 0.4 0.3 0.2 0.1 0 0 50 100 150 200 250 300 350 Tim e (Day s ) HW -0.1100-5000 HW -0.00065-5000

Figure 5-14 Early Water Breakthrough but gradual Development of Watercut

Water Cut (Fraction)

128

It shows that water tends to breakthrough earlier when adequate friction pressure correlation is used. The consequence of this observation is that the period of oil plateau production typically evaluated with commercial numerical simulators will be overestimated. The change in time to water breakthrough represents a 50 percent reduction at 5,000 stb/day production rate, for this particular reservoir. However, the water development is rather gradual. This is because water first breaks through at the heel and spreads gradually toward the toe of the well. As is expected, the ultimate watercut is a sole function of the reservoir properties and this is demonstrated by the convergence of the two curves in Figure in Figure 5-13. In the infinite conductivity and uniform flux models, the water crest rises symmetrically and uniformly toward the wellbore. Thus, the water crest height above the Oil-Water Contact (OWC) is the same at all points along the length of the well. This causes a dramatic increase in watercut as soon as the water breaks through into the well. This concept is not supported by current field evidence.
5.3.1.2 Effect on Flowing Bottomhole Pressure

Another effect of the increased friction pressure loss on horizontal wells is in the reservoir performance. The effect of increase pressure loss in the wellbore translates to an increase in reservoir pressure drawdown (Figure 5-15). The result of evaluation of drawdown pressure with conventional Eclipse simulation and the new equivalent pipe roughness function shows an increase in drawdown pressure. The effect is even more pronounced at increased oil production rate as shown by the diverging trend. Some researchers isolate the pressure loss in the wellbore from the pressure drawdown in the reservoir due to fluid withdrawal. 129

60 Pressure Drawdown (psi) 50 40 30 20 10 0 0 5000 10000 15000 20000 25000 L iq u id Ra te ( s tb /Da y ) Ec lip s e Fr ic tio n e =0 .0 0 0 6 5 Mo d if ie d Fr ic tio n e = 0 .1 5 0 0

Figure 5-15 Friction Pressure Loss Effect on Horizontal Well Bottom Hole Pressure

. Such analysts use the infinite conductivity equations to calculate the reservoir pressure drawdown and the smooth pipe flow equations to calculate the friction pressure loss (Jiang, et. al., 2000). The obvious conclusion from such analysis is that the effect of friction pressure loss in horizontal wells is insignificant. Ozkan has developed a more complete equation for reservoir pressure drawdown. (Ozkan, et.al., 1999). This equation is capable of accommodating any friction pressure loss correlation and could be used with the relation presented in this research.
5.3.1.3 Friction Pressure Loss and Water Crest Development/Profile Figures 5-16 and 5-17 show two different profiles of water crest development in

horizontal wells. Figure 5-16 was generated using the modified pipe roughness function while Figure 5-17 was made using Eclipse weighted average and Fanning friction factor relations.

130

HWWFIP2, 5-1/2, 5000stb/day, 0.15000,

Figure 5-16 New Water Crest Development - Skewed toward the Heel of the Horizontal Well

H W W FIP2, 5-1/2, 5000stb/day, e = 0.000065,

Figure 5-17 Weighted Average Approach shows Symmetrical Water Crest

In Figure 5-16, the water crest development is skewed toward the heel of the horizontal well. When water breaks through at the heel and spread gradually toward the 131

toe, it creates a poor sweep efficiency and bypassed oil at the toe of the well. Industry reports of horizontal wells water cresting supports this result. Thus, the diagram in Figure 5-16 shows a more practical representation of water crest development in horizontal wells. Figure 5-17 shows a symmetrical profile of water crest development along the wellbore for the weighted average approach and in-situ pipe roughness. The profile indicates that water crest develops symmetrically toward the heel and the toe of the well with the peak at the middle section. Consequently, bypassed oil should be at both the toe and the heel end (if any). This picture indicates an efficient sweep by horizontal wells even with water cresting. Field reports of horizontal well water cresting from production logs do not seem to support this result.
5.3.1.4 Water Cresting and Bypassed Oil in Horizontal Wells Another important aspect of water cresting in horizontal wells is the problem of

bypassed oil. Field evidence have shown that, often times, a lot of oil is left unproduced at the toe (upstream end) of horizontal wells due to water cresting (Permadi, 1995). In some cases, horizontal wells experiencing up to 80 percent watercut are yet producing 100 percent oil at the toe (Lien, S.C., et.al., 1990). Figure 5-16 adequately represnts the profile described above. In Figure 5-16, oil production at the toe is yet 100 percent while the well watercut is about 84 percent. Secondary mechanical intervention is not yet well developed for horizontal wells. If horizontal well cone water at the toe, the solution is an easy one run mechanical bridge plugs and isolate the toe section and continue production. However, when water breaks through at the heel, often times, the watercut increases rather rapidly requiring that the well be abandoned and in-fill drilling may be required to drain bypassed oil. 132

In concluding this chapter, the author recommends that adequate multidisciplinary field studies be performed prior to the development of whole reservoir subject bottom water drive with horizontal wells.

133

CHAPTER 6 WATER CRESTING CONTROL IN HORIZONTAL WELLS WITH DOWNHOLE WATER SINK TECHNOLOGY

The previous chapter argues that pressure gradients in horizontal wellbore are varied and so create non-uniform drawdown and fluid influx (Brekke, and Lien, 1992, Permadi, et.al., 1997). The consequence is that water break through occurs at the heel end of the well and gradually extends to the toe. Thus, a considerable length of horizontal well upstream section is never contacted by bottom water at even very high watercut (Permadi, et.al., 1995). In this chapter, the author presents an overview of several solutions to the problem of water cresting in horizontal wells. The chapter ends with the presentation of an extension of the principle of the downhole water sink technology for water coning control in vertical wells to water cresting in horizontal wells.
6.1 Feasibility of Water Cresting Control with Dual Completion Technology

The petroleum industry has made tremendous technological progress in the area of multi-lateral well drilling and completions. Re-entry of existing wells for the purpose of drilling multi-lateral wells is also growing. The technology of drilling multi-lateral sections from existing wellbore is even more attractive in offshore operations where the cost of adding one more slots to the subsea template could be expensive. Two variants of the multi-lateral technology have been recommended and extensively studied for mitigation of water crest development in horizontal wells. They are: (1) a vertical well extension into the water zone and an upper section horizontal well 134

targeted at the top of the oil pay for optimum oil recovery. This concept will be called the Tail-pipe Water Sink (TWS) option; (2) Two horizontal wells drilled laterally on top of each other with the upper section targeted at the top of the oil zone and the lower section targeted a few feet below the original oil-water contact. This approach will be called the Bilateral Water Sink (BWS) technology option. The recommendation of the two approaches are based on the field evidence of the profile of water crest development in horizontal wells which indicates that the water first breakthrough at the heel and then extends along the length of the wellbore. The

concurrent production of water from the water zone using the vertical tail pipe well or the lower horizontal well (in the case of the bilateral option) performs a dual role: (1) redistibution of the pressure profile along the wellbore but more specifically around the heel, and (2) reduction of water cresting in the upper oil zone horizontal well.

Water

Oil

PCP 18 Oil OWC

Water 64

Figure 6-1 Tail-Pipe Water Sink Completion Schematic

135

The segregated production of oil-free water at the lower horizontal well or the vertical tail-pipe (similar to its variant for vertical wells called the Downhole Water Sink Technology) essentially creates a pressure sink around the heel of the horizontal well. Figures 6-1 show schematic representation of the tail-pipe water sink completion. The vertical tail-pipe is simply an extension of the original vertical hole section drilled to test the properties of the reservoir prior to kick-off to drill the horizontal section. Figure 6-2 shows the tandem horizontal well option for controlling water production in the horizontal wells (the Bilateral Water Sink technology).

Water

Oil

PCP 18 Oil OWC

Water 64

Figure 6-2 Bilateral Horizontal Well Water Sink Option

The segregated production of water from the lower horizontal section creates a redistributive pressure effect around the heel of the primary oil-zone horizontal well and so reduce the level of water influx around the heel and reduce the incidence of bypassed oil around the toe of the well. 136

6.1.1

Drilling and Completion of the Tail-pipe Multi-lateral Well Option

Perhaps the easier of the two technologies in terms of drilling and completion is the tail-pipe option. A level 5 multi-lateral well completion technology is recommended for the tail-pipe water sink. The level 5 multi-lateral well is defined as a system offering hydraulic isolation of the junction through the use of additional straddling completion equipment Figure 6-3 (Emerson, et. al., 1998).

Figure 6-3 A level 5 Multi-lateral Schematic for Tail-pipe Water Sink

A variant of level 4 multi-lateral well could also be used in cases where the oil zone production will be accomplished through the tubing-casing annulus while the water production is accomplished through the main-bore completion tubing. Level 4 multi137

lateral well is defined as having a cased and cemented main bore with a cased and cemented lateral liner in order to provide full mechanical support at the junction. The variation involves replacing the top permanent production packer with a dual bore bottom hole oriented completion guide. The main bore completion tubing is used to produce the water zone while oil zone production is accomplished through the annulus.
6.1.1.1 Project Planning Options for Tail-pipe Water Sink For a project plan that incorporates the implementation of the tail-pipe water sink

option as a remedial solution, the extension of the original rat hole into the water zone is accomplished with a whip stock guide. An oriented whipstock guide is set just below the kick-off point of the original horizontal well for the case of remedial application. A stiff drilling assembly is then run through the whipstock guide oriented away from the azimuth of the horizontal section. This permits re-drilling of the cement/mechanical plug of the original rat hole into the water zone. The completion involves using the downhole split packer and a multi-lateral junction expandable seal. The recent development of the expandable composite liner makes this option very attractive. In the case of pre-planned tail-pipe water sink well a premilled casing kick-off window option could also be used as in the first multi-lateral well in Saudi Arabia (AlUmair, 2000). The vertical section is drilled into the water zone and protected with a primary casing string. The vertical section is usually drilled with sufficient rat hole over 200 ft of rat hole to permit drop off of lost retrievable accessories since hole cleaning usually constitutes a major problem for multi-lateral wells. The primary casing string contains a pre-milled casing joint carefully spaced out across the desired kick-off depth to land the horizontal section at the top-most part of the oil zone. A whipstock packer set 138

across the depth of the pre-milled window guides the drilling bit through it to drill the horizontal section.
6.1.1.2 Operation Sequence and Equipment for Pre-planned Tail-pipe Water Sink The mainbore is drilled, cased and cemented. A Liner Hanger Packer or

TorqueMaster Packer is run in conjunction with whipstock System to create the casing exit window. After drilling the lateral, the lateral casing is run and cemented in place with the top of the liner extending back through the casing exit and into the mainbore portion of the well creating a level 4 system with good mechanical junction integrity. The lateral bore can then be perforated, stimulated, and completed as required.

After completing the lateral, a washover assembly tool is used to wash over and retrieve both the portion of the lateral liner extending into the mainbore and the whipstock and anchor assembly. Additional completion equipment is installed to create the hydraulic integrity required for a Level 5 multilateral system. First, a Scoophead Diverter and Anchor System are run. The anchor system latches into and orients against the Packer positioned below the window. Once the Scoophead Diverter is landed, the lateral production string is run through the Scoophead Diverter and sealed off in a previously run production packer set in the lateral bore. The final step in the multilateral process is dependent upon the type of production desired. For the segregated production required of the water sink technology, isolated fluid production is desired. A standard Dual bore Packer is run directly above, and tied into, the Scoophead Diverter. Selective re-entry into either bore is possible with coil tubing or wireline tools.

139

6.1.1.3 Application of Tail-pipe Water Sink Technology in Old Vertical Wells

The descriptions of the operation procedures outlined above indicates that the tailpipe water sink technology could be extended as a remedial option for fields developed with vertical wells where severe water coning has created substantial amount of recoverable but bypassed oil. For such applications, the old oil zone perforations are first squeezed off with cement. Then, the old vertical well is extended into the water zone as the water sink completion. A whipstock milling and drilling operation is then targeted at the top section of the oil zone to produce the bypassed oil. Both wells are completed with sufficient mechanical and hydraulic isolation as level 5 multi-lateral well completion and segregated fluid production could be implemented.
6.1.2 Drilling and Completion of the Bilateral Water Sink Well Option

The bilateral water sink (BWS) concept requires that the two horizontal well be drilled laterally above each other. The top horizontal well targets the top of the oil zone while the lower horizontal section targets the water zone. The two completions should have zonal isolation integrity between them to avoid co-mingling of the oil and water in the wellbore. A level 5 multi-lateral well completion is also recommended for adequate zonal isolation and segregated fluid production. Multi-lateral technology has advanced tremendously such that lateral wells could be drilled a few feet from each other with the help of geo-steering tools and anti-collision magnetic surveys. In a typical project in Alaska, a watered out horizontal well was re-entered and a top horizontal well drilled and landed just 2 feet away in true vertical depth (Figure 6-4).

140

Figure 6-4 Multi-lateral Well Project in Purdue Bay, Alaska (Courtesy, Bakerhughes website)

The pre-planned option could use the proven application of dual completion well in Saudi Arabia (Figure 6-5). The project involves the selection of a pre-milled casing window joint. The window joint is made of composite wrap type material permitting easy drillout and minimal debris problems. At the same time, the pre-milled window contained an internal pressure sleeve to tolerate differential pressures during cementing operations. This pressure sleeve is retrievable after the drilling and cementing operations. A typical sequence of operation is as follows:

Drill to the production casing point at the top of the oil water contact or any shale point between the oil and water zone.

Run and cement the production casing including a landing collar, pre-milled window joint, casing annular packer and a port collar for two-stage cementing.

Retrieve an inner sleeve from the pre-milled window joint 141

Set a drilling whipstock at the window Drill the upper lateral out of the window joint Retrieve the drilling whipstock Drill out the lower lateral in the water zone. Set a production packer in the mainbore lateral (also known as the water sink lateral)

Set a dual bore deflector (DBD) in the production casing latch coupling with a lower seal assembly stung into the lower packer.

Figure 6-5 Dual Completion Multi-lateral Well in Saudi Arabia (Al-Umair, A.N., 2000)

Set a retrievable hydraulic dual packer in the production casing with a seal assembly stung into the dual bore deflector, and tail pipe assembly run into the upper lateral (oil zone completion). 142

Figure 6-5 shows the schematic of the dual completion multilateral for the Saudi Arabian project that could be perfectly adapted for the bilateral water sink technology option. The remedial option for horizontal wells with water cresting problems follows the same approach as the tail-pipe option. The vertical rat hole is re-entered. The cement and mechanical bridge plug is drilled out with sufficient rat hole below the water zone. A new cement plug is then set prior o kick off. A mechanical-set multi-lateral hanger/packer is installed as datum for deploying a window master retrievable whipstock. The mechanical hanger/packer also provides a reference for the final completion. The multi-lateral sections could also be completed with slotted liners. In cases where borehole stability is a major concern due to unconsolidated sandstone reservoir, pre-packed slotted liners could be used in the multi-lateral sections to mitigate the problem of sand production. The lower water sink section could be completed with a permanent packer and screens. The upper lateral (oil zone completion) could be completed with scoophead divertor system and the junction adequately isolated and the production tubing sealed into a permanent packer previously set in the lateral. Finally, a dual string packer is used in the mainbore so that each completion (zone) could be produced independently, a primary requirement for the segregated fluid production in the water sink technology. Similar application for two zone reservoirs has been implemented in the OB1 project offshore Brunei (Bakerhughes website, 2001). This level 5 multi-lateral well posed very little problem from drilling to completion and the technology is quite adaptable for the bilateral water sink completion option to control water cresting in horizontal wells. 143

6.2

Water Cresting Control Performance of the Tail-pipe and Bilateral Water Sink Technology

The benefits of the application of multi-lateral well technology in controlling water cresting in horizontal wells was evaluated with the compound friction pressure loss model and compared with the conventional eclipse 100/200 Fanning friction factor and in-situ pipe roughness and weighted average mixture properties. Sample data deck for Eclipse models of the horizontal wells with DWS Technology are contained in Appendices C and D. In the conventional case with regular Fanning Friction Factor, the application of tail-pipe or bilateral water sink contributed very little to improving cumulative oil recovery over time. As the Figure 5-17 showed, the sweep efficiency for the symmetric crest is high and bypassed oil is small.
Table 6-1 Oil Recovery for Conventional Horizontal Well with Fanning Friction Factor Well completion type Cum. Oil Prod. Recovery (stb) (In 20 years) Factor (RF) 3,456,980 0.6094 Percent Increase % 0

Conventional Horizontal Well Horizontal well with Tail-pipe DWS Horizontal well with Bilateral DWS

3,542,071 3,490,449

0.6244 0.6153

1.5 0.59

Table 6-1 was generated for a conventional pipe roughness of 0.000065 ft and weighted average properties for the fluid mixture as modeled by the Eclipse 200 reservoir simulator. Well curvature radius was less that 200 ft and so the well could be classified as a short radius well. The poor recovery performance is attributable to the fact that water crest development is nearly symmetrical along the well length as the effect of friction 144

pressure loss is minimized. Thus water table moves up symmetrically toward the horizontal well as oil production progresses. For this model the water breakthrough is rather dramatic. After initial water breakthrough, watercut rises to over 60 percent in less than four months. Using the modified friction pressure loss model, the water crest development becomes rather asymmetrical with water breaking through first at the heel and gradually spreading to the toe of the well. Thus any technology that can effectively redistribute the pressure around the heel should be able to control water production and ultimate oil recovery. In another horizontal well with similar reservoir properties the modified friction factor is applied. The table showed an improvement in recovery with the concepts when the water crest is greatest at the heel.
Table 6-2 Improved Oil Recovery in Horizontal Well with DWS Technology Well completion type (Modified Friction Cum. Oil Prod. Recovery Factor) (stb) (In 15 years) Factor (RF) Conventional Horizontal Well 2,762,463 0.799 Percent Increase % 0

Horizontal well with Tail-pipe DWS Horizontal well with Bilateral DWS

2,917,122 3,013,089

0.844 0.872

4.5 7.3

The well has start of perforation about 400 feet away from the vertical axis of the surface location. Table 6-2 and Figure 6-6 show cumulative oil recovery over a period of 15 years. The water crest development showed an increase at the heel with a gradual spread toward the toe. Original Oil In Place (OOIP) is 3.457 million stock tank barrels.

145

3200000 3000000
Cumm. Oil Prod. (stb)

2800000 2600000 2400000 2200000 2000000 2000 2500 3000

85ft-1k-0

3500 4000 Time (Days)

85ft-1k-12k+tp

4500

5000

5500

.33BL-85ft-1k-12k+

Figure 6-6 Cumulative Oil Recovery Comparison, Conventional Horizontal Well, Horizontal Well with Tail-pipe Water Sink and Horizontal Well with Bilateral Water Sink

The reservoir studied in Figures 6-6 and 6-7 has 85 ft total thickness with 40 ft net oil sand and 45 ft water zone. The top completion liquid rate is 1,000 stb/day (1k) and the bottom completion liquid rate is 12,000 stb/day for DWS well and zero for the conventional horizontal well. In the legend, 85k-1k-0 represents a conventional horizontal well delivering 1,000 stb/day liquid rate. 85ft-1k-12k+tp represents the tailpipe water sink completion with liquid rates of 1,000 stb/day and 12,000 stb/day at the top and bottom completions respectively. .33BL-85ft-1k-12k+ represents a bilateral water sink completion delivering 1,000 stb/day liquid at the top completion and 12,000 stb/day liquid at the water sink completion. The length of the water sink completion is 33 percent (one-third) the length of the oil zone completion.

146

The graph shows an improvement of about 300,000 stock barrel of oil recovery when the bilateral water sink completion concept for water cresting control is implemented. The Figure also shows that the bilateral water sink is superior to the tail pipe water sink in controlling water cresting and hence oil recovery. This result could be attributed to the fact that the lateral drainage influence of the vertical tail pipe is limited compared with the bilateral water sink that is another horizontal well. The tail pipe option promises to be more effective in short radius as well drain hole type of horizontal wells. However, for drain holes, the well lengths are usually small compared to conventional horizontal wells and the effect of friction pressure loss is negligible. As a result of the superior performance of the bilateral water sink, the next section presents the results of the performance evaluation study of the BWS concept.
6.2.1 Effective Length of Bilateral Water Sink Horizontal Well

With a water crest development maximum at the heel, the pertinent question that follows is; what effective length of the bilateral water sink well is adequate to control water cresting in the oil zone horizontal well? Various lengths of the lower lateral water sink were investigated. The parameter to evaluate the effect of different lengths of the water sink completion is the oil recovery over time. A period of 15 years as was selected as in previous simulations. The results shown in Figure 6-7 indicate that only a third of the length of the upper horizontal well is required to effectively redistribute the water crest along the length of the upper oil zone horizontal well. The cumulative oil recovery in a 15 year period differ only by 18,162 stb or a fraction of 0.60 percent.

147

3200000 3000000
Cumm. Oil Prod. (stb)

2800000 2600000 2400000 2200000 2000000 2000

2500

3000

3500

4000

4500

5000

5500

Time (Days) 85ft-1k-0 85ft-1k-12k+BL1 .33BL-85ft-1k-12k+

Figure 6-7 Evaluation of Effective Length of Bilateral Water Sink Well

Thus this study recommends a well length of 33 percent (or one third) of the length of the oil zone horizontal well as adequate to redistribute the pressure along the wellbore of the upper horizontal well.

148

CHAPTER 7 COMPARISON OF DWS VERTICAL WELLS AND HORIZONTAL WELLS IN BOTTOM WATER DRIVE RESERVOIR

Several researchers and industry operators have recommended the application of horizontal wells as a suitable solution for developing reservoirs experiencing water coning problems. The discussions in chapters 5 and 6 reveal that horizontal wells are, themselves, not free from water influx (water cresting) problems and that in some cases, the influx of water into horizontal wells could be so dramatic that it tends to erode its merit. In this chapter, the research focuses on comparing the performance of vertical wells and horizontal wells in developing fields with severe water coning problems. To undertake this study, certain assumptions and observations have to be made as follows:

Commercial considerations are not included in this evaluation Comparing the performance of one vertical well to one horizontal well is inappropriate from reservoir engineering viewpoint as each well drain different areas of the reservoir.

Drainage area consideration is used in the evaluation of the performance of the two type of wells

Thus, two vertical wells are equivalent to one horizontal well of length 1,500 ft. 149

Horizontal wells, are quite superior to vertical wells because of the high production rates at low drawdown pressure. The low drawdown pressure is as a result of the extended reach nature of the horizontal well within the reservoir. The major factor responsible for coning in conventional wells is the pressure drawdown around the wellbore. Due to its extended exposure with the reservoir, a horizontal well usually has less drawdown pressure than a vertical well for equal rates of production. Moreover, the standoff from the bottom water can be maximized with a horizontal well located high in the oil pay zone. For over 15 years, these specific advantages of horizontal well have been the basis of their intensive use in place of vertical wells in developing reservoirs subject to bottom aquifer or a gas cap. Infact, it can be stated that horizontal wells have become the standard industry procedure to develop these reservoirs.

7.000 6.000 5.000 4.000 3.000 2.000 1.000 0.000 0 1000 2000 Length of Horizontal well (ft) 3000 4000

Figure 7-1 Productivity Comparison between Vertical and Horizontal Wells

150

The plot in Fig. 7.1 shows the superiority of horizontal well PI over that of a vertical well for various lengths of the horizontal wells for single phase flow and in an isotropic and homogeneous reservoir using equation 2-17 (Geiger, 1984). As the Figure shows, a horizontal well of length 1,500ft has a productivity index three and half times greater than that of a single vertical well in the pre-water breakthrough period. Also comparing critical rates of vertical and horizontal wells, Chaperon developed a simplified relation as follows:
Qch L F = * Qcv X A qc

7-1

Note that for, F = 4; and, q*c 1.0 , Eq.(4.6) gives: Qch/Qcv = 4L/XA, and XA is the location of a constant pressure boundary or abscissa where the interface level is known to be, h, for steady-state flow; h is the initial oil thickness or interface elevation at XA. Thus for XA = 750 ft and length, L, = 1,500 ft, the critical oil rate should be about 8 times higher for horizontal well than for the vertical well. An inherent shortcoming of all of the above equations is that they were developed assuming friction pressure loss in the wellbore is negligible. The discussion in chapters 5 and 6 indicates that these equations tend to overestimate the performance of horizontal wells due to the assumption of a frictionless flow. This research has shown that friction pressure loss in the wellbore could be quite significant. As an example, the ratio of PI of a 1,500ft horizontal well and a vertical well has been estimated to be 3.5 using the equation 2-17 and Figure 7-1. However, these ratio could be reduced to as low as 1.5 when the effects of friction pressure loss in the wellbore is considered. 151 Another

shortcoming of the above PI relation is that it assumes single-phase flow. It is, thus, applicable to the pre-water breakthrough production period. It also failed to consider the concept of equivalent drainage area. A more pragmatic evaluation of the performance of horizontal wells over vertical wells in reservoirs with water coning/cresting problems should be to consider the concept of equivalent drainage area for both types of wells. One frequently advanced advantage of horizontal wells over vertical wells in field development practice is that fewer wells are required to drain the same reservoir for horizontal wells than for vertical wells. This advantage needs a more scientific analysis. It implies that both type of wells do not drain the same area of the reservoir. If the technology of horizontal wells is to be compared with that of vertical wells considering the overall reservoir performance, then the more scientific comparison should be on the basis of equivalent drainage area. The next section presents an equivalent drainage area tool for comparing the performance of horizontal wells and vertical wells.
7.1 Equivalent Length/Drainage Area of Horizontal Wells

Researchers have agreed that horizontal wells drain larger area of the reservoir due to its extended reach. Consequently, several studies have been initiated to evaluate an effective drainage area for the horizontal well. Two mathematical considerations have been advanced for estimating the drainage area of a horizontal well. One approach treats the drainage area of the horizontal well as a combination of two half-circles at each end of the well length and a rectangle along the length of the well. The width of the rectangle is equal to twice the length of the drainage radius of the conventional vertical well and the length of the rectangular section is equal to the length of the horizontal reach (length 152

of the perforated section). The other method considers the drainage area of the horizontal well as an ellipse with minor radius equal to the drainage radius of conventional vertical well (Joshi, 1991).
7.1.1 Method 1: Two Half-circle and Rectangle Approach

Drainage area in method 1 is calculated as, re2 + (2re*L) = Area (Drainage)

7-2

L a b
Figure 7-2 Horizontal Well Drainage Area 7.1.2 Method 2: Ellipse Shaped Drainage Area Method

This method approximates the drainage area of a horizontal well with an ellipse assuming the length of the minor axis b to be equivalent to the drainage radius of the vertical well and the major axis a equal to half the length of the well plus the drainage radius. The drainage is, A(ft2) = (L/2 + re)*re
7-3

153

a b
Figure 7-3 Elliptical Drainage Area for Horizontal Well

It should be emphasized that the two methods give different solutions to the size of the drainage area of the horizontal well. Joshi (1991) takes the average of the two as the equivalent drainage area of the horizontal well. Joshis approach has been adopted for this research. Joshi average-area approach is presented as follows: For a vertical well of drainage radius re = 750 ft, the drainage area is (*(750)2)/43560 = ) 40.57 acres. The area for two such vertical wells is 81.14 acres. The length of a horizontal well that would drain the same area is given by the following sample calculations:

1 L * p * + a * a + p * a 2 + (L * 2 * a ) = 81.14 Acres 2 2

7-4

Substituting values in the above equation where a = 750 ft and solving for L gives a horizontal well with an equivalent drainage area of two vertical wells. The horizontal well length is L = 1320 ft. Consequently, affective comparison of the performance of vertical wells and horizontal wells in reservoirs with water coning problems, devoid of economic considerations will be performed on the basis of one horizontal well of length 1,500ft to two vertical wells.

154

The results presented in the subsequent sections are derived from numerical simulation of two vertical wells and one equivalent horizontal well in a reservoir with same grid sizes and properties.
7.2 Simulation Grid Properties of Horizontal Well and Vertical Well

The reservoir considered in this section has initial thickness of the oil zone 40 ft and the thickness of the water zone is 45 ft. The modeling is performed with the Geoquest Eclipse software.
Table 7-1 Reservoir/Rock properties of Horizontal Well Vs Vertical Well Water Cresting/coning Performance Studies
Input Data Properties Unit

Reservoir Pressure Thickness of oil/gas column Thickness of water column Depth of OWC/GWC (static) Length of oil production perforations (Horizontal) Length of water perforations (Horizontal Bilateral) Position and length of Oil zone perforations (Vertical) Position and length of water perforations (Vertical) Horizontal permeability in Oil/Gas column Vertical Permeability in Oil/Gas column Horizontal permeability in water column Vertical Permeability in water column Water density at temperature Water viscosity at temperature Reservoir temperature Porosity in oil column Porosity in water column Oil/gas formation volume factor Oil gravity Oil viscosity at BHT Water compressibility Oil compressibility Rock compressibility Completion diameter/hole size

2500 40 45 6194 1500 500 8000 800 8000 800 64.2 0.50 165 0.31 0.35 1.15 35 1.34 3.0E-06 1.5E-05 4.0E-06 9-5/8-in csg

psi ft ft ft ft ft mD mD mD mD Ib/ft2 cP O F fraction fraction rb/stb O API cP sips sips sips

155

The Eclipse model consists of a 75 x 20 x 12 Cartesian grid (about 18,000 grid blocks) with block centered geometry and implicit formulation. Comparison is performed using the Joshi equivalent drainage area and length method. Two vertical wells with drainage radii of +/- 750 ft are considered; Equivalent length of a horizontal well in the same area is 1500 ft. The thickness of each grid block from top of sand to about 15 ft below Oil Water Contact (OWC) is 5 ft giving 11 blocks. The last row of grid blocks have been made large compared to the others (30 ft thickness) and assigned infinitely large porosity to simulate a steady-state flow process. The friction option of the Eclipse has been activated for this study to include the effect of pressure variation along the length of the wellbore necessary for the finite conductivity assumption. The modified equivalent pipe roughness function of 0.1100 ft, is used for this study. Table 7.1 shows the reservoir and fluid properties employed for this study. The horizontal oil well completion is placed within the first 30 percent of the oil pay following good industry practice for reservoirs with bottom water drive. The water sink is modeled (1) as a vertical tail pipe originating from the heel of the horizontal well, and (2) a bilateral well with water sink running parallel below the oil completion in the water zone. Locations of both water sinks are varied to obtain optimum performance and minimize initial inverse oil coning.
7.3 Comparison of Watercut Performance of Vertical and Horizontal Wells

To effectively evaluate the effective performance of each of the technologies, three parameters of utmost importance to this study will be analyzed namely oil production rate performance of each well type, cumulative oil production over time and watercut performance analyses. 156

7.3.1

Oil Production Comparison Study

Horizontal wells have undoubted superiority over vertical wells in producing oil reservoir subject to water coning problems. As could be seen in Figure 7-4, the low drawdown pressure capability of horizontal wells is responsible for the longer period of plateau production prior to water breakthrough.
3500 3000 2500 2000 1500 1000 500 0 0 500 1000 1500 Time (Day s)
HWCON DVWCON SVWCON

Oil Rate (stb/day )

2000

2500

3000

Figure 7-4 Oil Production Rate Performance Evaluation

The thin continuous line (tagged HWCON in the legend) represents the performance of a horizontal well producing at 3,000 stb/day. The medium line (DVWCON in the legend) represents the performance of two single wells delivering a combined total of 3,000 stb/day (1,500 stb/day each) and the thin broken line (SVWCON in the legend) shows the performance of a single vertical well delivering 3,000 stb/day liquid rate.

157

As shown in the diagram, the single vertical well experiences water breakthrough rather early and so the production of oil declines almost immediately. However, comparing the performance of two vertical wells with a combined total oil rate equal to the oil rate of the horizontal well shows the field wide performance of horizontal wells may be comparable to that of equivalent number of vertical wells. This picture is made clearer by comparing the cumulative oil recovery of the three types of field drainage options over time as shown in Figure 7-5.

5.00E+06 4.50E+06

Oil Recovery (Stock Tank Barrels, stb)

4.00E+06 3.50E+06 3.00E+06 2.50E+06 2.00E+06 1.50E+06 1.00E+06 5.00E+05 0.00E+00 0 1000 2000 3000 4000 Time (Days) 5000 6000 7000 8000

HWCON

DVWCON

SVWCON

Figure 7-5 Cumulative Oil Recovery Performance Analysis

In terms of cumulative oil recovery over a period of twenty years, a horizontal well would perform better than one vertical well. However, for the equivalent drainage area concept with two vertical wells the cumulative oil recovery for the same combined rates as the horizontal well show that the cumulative oil recoveries converges (Fig. 7-5). The gain in oil recovery in the pre-water breakthrough period, arising from the long oil production plateau period of horizontal wells is eroded by the dramatic increase in 158

watercut in the horizontal well post water breakthrough whereas the low initial recovery by the dual vertical wells due to early water breakthrough is improved by the gradual rate of increase of watercut post breakthrough.
7.3.2 Watercut Performance Evaluation

The superior performance of a horizontal well over a single vertical well is in the period of delayed water breakthrough. After water breakthrough the watercut in horizontal well increases rapidly and approaches that of a single vertical well producing at the same rate. However, from the equivalent drainage area evaluation, the watercut in the horizontal well rapidly increases and exceeds the combined watercut of the two equivalent vertical wells over a short time after water breakthrough. This result indicates that horizontal wells may not be the preferred optimum solution for developing bottom water drive reservoirs with high permeability, if long-term production of oil is a critical consideration. Figure 7-6 is a plot of watercut performance of single vertical well delivering same oil rate as the horizontal well. Also included in the plot is the watercut of two vertical wells draining thee same equivalent reservoir area. In the Figure 7-6, the thin continuous line represents the watercut performance of one horizontal well. The medium line represents the watercut performance of two equivalent vertical wells and the thin broken line represents the performance of a single vertical well. The result presented in the plot shows watercut for the horizontal well overtaking the two equivalent vertical wells. It underscores the need for industry operators to carefully analyze and evaluate the added advantage of horizontal wells over vertical wells in developing reservoirs subject to water coning problems, especially in high permeability reservoirs. It also further 159

emphasizes the fact that water cresting in horizontal wells could be so dramatic that it may erode the merit of initial high oil rate.

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 0 500 1000 1500 2000 2500 3000 3500

Time (Days)

SVWCON

HWCON

DVWCON

Figure 7-6 Watercut Performance Evaluation of Vertical and Horizontal Wells

The ultimate oil recovery from reservoirs with bottom water drive and high deliverability performance may be less if such fields are developed with horizontal wells than for vertical wells as the field tend to be abandoned pretty early at high watercut. This analysis of watercut performance of horizontal wells combined with the difficulty of secondary mechanical intervention in horizontal wells after water breakthrough compared with conventional vertical wells or even high angles wells makes horizontal wells less attractive as a solution to developing these reservoirs experiencing severe water coning.

160

7.4

Comparison of Vertical Wells with DWS Technology and Horizontal Wells

In this section, the performance of two vertical wells completed with the Downhole water Sink (DWS) technology will be compared with that of an equivalent horizontal well. As in the previous section, three parameters will be used to qualify the superiority of one technology over the other namely, oil production rate over time, cumulative oil recovery and finally watercut performance in oil zone completion.
7.4.1 Two Vertical DWS Wells Vs Horizontal Well Oil Production Analysis

The application of the Downhole water sink (DWS) technology in vertical wells, not only reduces the water production in the top completion, it makes otherwise bypass oil accessible to the top completion. More oil is produced using DWS technology in vertical wells. Figure 7-7 and 7-8 shows the improvement in oil production rate and cumulative oil recovery for the two vertical wells with DWS and the single equivalent horizontal well.
3500 3000 2500 2000 1500 1000 500 0 0 1000 2000 3000 4000 Time (Days) HWCON DVWCON DWSVW 5000 6000 7000 8000

Figure 7-7 Oil Production with DWST Vs Horizontal Well

161

The heavy line graph shows the oil production rate performance of the two vertical wells completed with the DWS technology. The thin broken line indicates the performance of the two vertical wells without the DWS technology and the thin continuous line shows the performance of the horizontal well. The first spike in the heavy line indicates the point where drainage water production from the sink was initiated and the second spike shows a point where the rate from the sink was increased to optimize oil recovery. The performance of the cumulative oil recovery presents a clearer picture of the superiority of DWS technology in vertical wells with water coning problems. It shows that the technology can recover more oil than the horizontal well. The initial high oil production rate merit of the horizontal well is soon marred by the influx of water and the vertical wells with DWS soon out perform the horizontal well.

Oil Recovery (Stock Tank Barrels, stb)

5.00E+06 4.50E+06 4.00E+06 3.50E+06 3.00E+06 2.50E+06 2.00E+06 1.50E+06 1.00E+06 5.00E+05 0.00E+00 0 1000 2000 3000 4000 5000 6000 7000 Time (Days) HWCON DVWCON DWSVW

Figure 7-8 Cumulative Oil Recovery Performance of Horizontal Well Vs Two DWS Vertical Wells

162

In Figure 7-8, the heavy line (DWSVW in the legend) represents the performance of two vertical wells with the DWS completion. The thin broken line (DVWCON in the legend) represents the performance of the two vertical wells without the DWS technology and the thin continuous line (HWCON in the legend) represents the performance of a single equivalent horizontal well. The horizontal well initially performs better than the vertical wells even with the DWS (Figure 7-8). This is due to the low drawdown pressure and improved period of plateau production of oil from the horizontal well. However, after water break through, the advantage of the horizontal well is eroded and the two vertical wells with DWS technology outperform the horizontal well. This is shown by the heavy line with a cumulative oil recovery about 300,000 stb more than the conventional horizontal well.
7.4.2 Two Vertical DWS Wells Vs Horizontal Well Watercut Analysis

The application of the DWS technology in two vertical wells delays the time to water breakthrough (Figure 7-9). The horizontal well shows the longest time to water breakthrough due to the distribution pressure drawdown along the wellbore. On the other hand, application of DWS in vertical wells also improves the time to water breakthrough. This delayed water break through time improves oil recovery performance of the DWS completions. Initially, the horizontal well outperforms the equivalent two vertical wells due to the delayed water breakthrough time. However, after water breakthrough watercut in the horizontal well increases rapidly and soon overtakes the vertical wells. The result is that the duo-vertical wells with DWS completion produces less water in the oil zone than the horizontal well.

163

1 0.9 0.8 0.7

Watercut (Fraction)

0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 0 500 1000 1500 2000 2500 3000 3500

Time (Days) DWSVW HWCON DVWCON

Figure 7-9 Watercut Performance of Horizontal Well and Two Vertical DWS Wells

The preceding analysis in section 7.4 reveals that vertical wells with DWS installations could recover more oil when employed in field development than horizontal wells.
7.5 Economic Evaluation of the DWS Technology Vs Horizontal Well

The foregoing analysis shows the superior technical performance of the Downhole Water Sink technology in reservoirs with water coning problems. DWS in two equivalent vertical wells outperforms one horizontal well of length 1,500 ft. However, is the DWS technology economically justified? In this section, a simple economic evaluation will be presented. This analysis uses estimates for the cost of installation of two new vertical wells each equipped with dual production strings for the oil and water (DWS type completion, table 7-1). The wells have a true vertical depth of 8,000ft and are assumed to have high deliverability. The horizontal well producing length is 1,500ft, 164

the drainage area equivalent of two vertical wells. The estimated drilling time for one vertical well is 20 days and 27 days for the horizontal well. The estimated time for installing the completion strings is 10 days for the dual-string vertical wells and 7 days for the single string horizontal wells. Cost data are obtained from a shallow offshore well drilled in the Niger Delta Basin. The well project was planned and supervised by a team of engineers and they supplied the spreadsheet for this cost evaluation. The detailed spreadsheet for the drilling and completion cost is attached as appendix E.
Table 7-2 Drilling and Completion Cost for 2-Vertical wells and 1 Horizontal Well Description 2 Vertical Wells 1 Horizontal Well

Drilling Operation and Associated Services Completion Operation and Associated Services Casing Strings (Drilling) Completion strings Wellhead Equipment Subsurface Equipment Miscellaneous/Contingencies
Total Costs

4,242,000 1,105,600 1,547,000 320,000 180,000 250,000 80,800

$7,725,400

3,312,350 556,950 825,500 120,000 90,000 125,000 40,400

$5,070,200

It costs less to drill one horizontal well than two DWS vertical wells. However, the ultimate oil recovery (Figure 7-8) indicates that the two DWS wells outperform the single horizontal well after water breakthrough (about three years). An additional 150,000 to 300,000 stb of crude oil is recovered above that of horizontal well. This improved recovery may justify the extra cost of installation besides several other merits 165

of the technology enumerated in chapters 4 through 6. These factors include reduced water handling costs. Another disadvantage of horizontal well technology is the difficulty of secondary mechanical intervention for the purpose of well repairs and maintenance. In cases where water breaks through at the toe (very rare cases), the toe section of the well could be isolated. However, in most cases, water breakthrough occurs at the heel making it difficult for any meaningful intervention. Finally, the rapid increase in watercut level of horizontal wells with water cresting problems may cause rapid overload of water separation facilities leading to break down and high maintenance costs. Thus, the economic justification of the application of DWS technology in oil field development should not be based only on the initial sunk cost of drilling and completing the wells but on other critical factors during the producing life of the field.

166

CHAPTER 8 CONCLUSIONS AND RECOMMENDATIONS 8.1 Conclusions

The control of water coning/cresting with Downhole Water Sink (DWS) technology in both vertical and horizontal wells has been analyzed in this study. The application of the technology in controlling water coning creates a hysteresis effect. The study ague that relative permeability curves for water and oil differ during the water cone development process (imbibition process) and during the cone reversal process (when the DWS water sink is implemented). This hysteresis, coupled with the development of a saturation transition region for wells with severe water coning history have been studied in this research. The inclusion of relative permeability hysteresis and capillary transition zone in the design of DWS completions in watered out wells offer the necessary solution to the desire to produce oil-free disposable drainage water while reducing watercut in the oil zone and recovering bypassed oil with the technology This study also reviewed friction pressure loss in horizontal wells and agued that adequate representation of wellbore friction pressure loss provides a tool for effective understanding and modeling of water cresting in horizontal wells. It presented a generalized compound friction pressure loss relation for horizontal wells and horizontal pipes. The results of application of the compound friction pressure loss relation indicate that friction factor could increase by the order of magnitude of over 400 percent. Finally the study presented to concepts of the Downhole Water Sink Technology for controlling water cresting in horizontal wells. The following summarizes the findings of this research. 167

8.1.1 Relative Permeability Hysteresis and Capillary Pressure Transition Zone on DWS Well Performance

The conclusions derived from this study on effect of relative permeability hysteresis and capillary transition pressure on water coning control performance of the DWS technology are:

Cone reversal process by DWS involves capillary pressure and relative permeability hysteresis effects and these concepts should be included in the model development.

Relative permeability hysteresis without significant capillary pressure transition zone has no effect on DWS performance if the end-point-residual saturations are the same for both imbibition and drainage curves.

Drawdown pressure around the wellbore creates an enhancing effect that causes enlargement of the capillary transition zone around the wellbore. This pressureenhanced capillary transition zone enlargement is further made worse by the presence of the two pressure sinks in DWS wells causing expansion of the saturation transition zone until both fluids break through into the two completions.

The capillary pressure transition reduces the size of the IPW and thus reduces the domain of the water-free oil production. The practical implication of this effect is increased water pumping rate from the sink completion and increased pressure drawdown is needed for a required production of oil.

Consequently, recompletion of conventional wells with severe water coning history with DWS technology would not be as beneficial as DWS completions of new wells.

168

8.1.2

Oil-Free Water Production Capacity of DWS with Capillary Pressure Zone

With respect to the environmental merit of the technology in old wells the study has the following conclusions:

The results presented in chapter 4 show that, though the presence of capillary transition could result in concurrent production of contaminated fluid in both completions, oil-free water production at the water sink can be achieved by incorporating a capillary transition zone in the model and design procedure.

Transition zone enlargement effect occurs in conventional wells due to diffusion resulting from pressure distribution around the well. For DWS wells, however, the effect is not only more pronounced but it also alters the IPW plots a basic tool for design. There is a commingled inflow envelope (Region 4 in Figure 4-8) in addition to the envelope of segregated inflow (Region 2).

The study also shows that liquid rate at the bottom water sink completion is limited by the oil breakthrough line. The optimum rate of liquid withdrawal at the top (oil zone) completion is determined by the point of maximum oil rate. This could be obtained from a graph of oil rate against liquid rate at the top completion.

The study of IPW shows isolines of increasing watercut to the right along the oil breakthrough line indicating that for oil-free water drainage design, maximizing liquid production rate at the top completion typical for DWS design would not be effective. Not only would it result in smaller oil rate, it might also lead to lesser recovery with DWS potential subject of future studies.

Finally, locating the water sink in the swept zone or the zone of saturation transition could cause concurrent production of contaminated fluids in both completions. Thus, 169

it is important that the water sink completion be located below the original oil-water contact.
8.1.3 Water Cresting in Horizontal Wells and DWS Technology

The study also evaluated the problem of water cresting in horizontal wells. As part of establishing an adequate representation of water cresting performance of horizontal wells, a new generalized relation for evaluating compound friction pressure loss in the wellbore of horizontal wells and in pipes have been developed. This generalized equation is a modification of the equation earlier developed by Su and Gudmundsson. To accommodate the limitations of existing commercial numerical simulators, a new equivalent pipe roughness function has been developed that creates the same friction pressure loss effect in the wellbore using the conventional Fanning friction factor approach. The study also recommended the adoption of the Pal and Rhodes equation for evaluating the flow viscosity of oil-water mixture as opposed to the weighted average approach. Using this approach, a more practical numerical modeling of water cresting performance of horizontal wells have been presented and the following conclusions made:

The water crest profile in horizontal wells is skewed toward the heel. Water breakthrough occurs first at the heel end of the well and spreads gradually toward the toe (upstream end of the well).

The time to water breakthrough evaluated with most of the current numerical or analytical models are optimistic and not representative of current field results. 170

According to the viscosity and friction pressure loss model used in this study, the pressure drawdown in the wellbore is actually higher than is currently assumed and the difference between current analytical models and the new approach diverges with increasing rate of production.

The application of a simple Fanning friction pressure loss relation for fluid flow in pipes to horizontal wells is probably inadequate. In recommending water cresting control option with the DWS technology, the

study evaluated the application of two innovative concepts namely tail-pipe water sink (TWS) well and bilateral water sink (BWS) well for redistributing the pressure along the wellbore. The results show that both the vertical tail pipe and the bilateral well concepts can redistribute the pressure profile along the wellbore and reduce the bypassed oil at the toe of the well. However, for long radius wells, the bilateral well option shows more promise in improving oil recovery. The study showed that only a third of the length of the oil-zone horizontal well is required for the water sink to adequately redistribute the pressure profile along the wellbore. Improvements in oil recovery (as much as 7 percent) have been predicted with the bilateral well technology.
8.1.4 Vertical Wells with DWS and Conventional Horizontal Wells

Finally, the study also compared the performance of vertical wells with DWS technology and horizontal wells in the development of reservoirs underlain by water. The results indicate that two vertical wells with DWS completion are capable of improving ultimate oil recovery more than horizontal wells in reservoir where water coning/cresting is critical. It also shows that water contamination of the produced oil could be less with the DWS completions than with horizontal wells. The result is reduced water handling 171

costs. Thus, the application of DWS in vertical wells may improve the ultimate oil recovery of reservoirs with strong water drive compared with the performance of horizontal wells in such fields.
8.2 Recommendations

Studies of oil recovery performance of DWS wells with capillary pressure transition zone with oil-free drainage only would it result in smaller oil rate, it might also lead to lesser recovery with DWS potential subject of future studies.

Further research work should be initiated to develop a relation for the ratio of fluid influx to the main wellbore flow as a function of the length of the horizontal well.

A study to evaluate the extension of the pressure loss relation developed in this dissertation to three-phase (oil-water-gas) flow should be performed.

Analytical evaluation of the effect of the increased pressure loss in the wellbore on reservoir pressure drawdown should be investigated.

A Rule-Of-Thumb relation should be developed to help industry operators in determining when to use or not to use horizontal wells in developing reservoirs subject to bottom water drive.

Field application of the bi-lateral well technology and the tail-pipe option in watered out reservoirs with substantial bypassed oil should be vigorously pursued.

Finally, the performance of the bi-later and tail pipe techniques for water cresting control in reservoirs with capillary transition zone should be investigated. 172

REFERENCES AL-Umair, N. A., 2000, The First Multi-Lateral/Dual Completion Well in Saudi Arabia, IADC/SPE Paper 62771, IADC/SPE Asia Pacific Drilling Technology Conference, Kuala Lumpur, Malaysia, September 11-13.

Archer, J.S. and Wall, C.G., 1988, Petroleum Engineering principles and practice, Graham and Trotman, London, pp. 92-119 Arirachakaran, S., Oglesby, K.D., Malinowsky, M. S., Shoham, O., and Brill, J. P., 1989, An Analysis of Oil/Water Flow Phenomena in Horizontal Pipes, Paper SPE 18836, SPE Production Operations Symposium, Oklahoma City, OK, March 1314 Asheim, H., Kolnes, J., and Oudeman, P., 1992, A flow resistance correlation for completed wellbore, Journal of Petroleum Science and Engineering, 8, Elsevier Science Publishers, Amsterdam, pp 97-104 Asheim, H., and Oudeman, P., 1997, Determination of Perforation Schemes To Control Production and Injection Profiles Along Horizontal Wells, Paper SPE 29275, SPE Drilling and Completion, March pp 13-17. Babu, D.K., and Odeh, A.S., 1988, Productivity of Horizontal Well, Paper SPE 18334, 63rd Annual Technical Conference and Exhibition, Houston, TX, October 2-5. BaherHughes, 2001, Multilateral level definitions and multilateral case histories www.bakerhughes.com/bot/Multilateral.htm November. Bourgoyne, A.T. (Jr.), Millheim, K.K., Chenevert, M.E., and Young, F.S. (Jr.), 1991, Applied Drilling Engineering, Society of Petroleum Engineers, Richardson, TX., pp144-147. Bournazel, C. and Jeanson, B., 1971, Fast Water-Coning Evaluation Method, Paper SPE 3628, 46th Annual Fall Meeting of Society of Petroleum Engineers of AIME, New Orleans, LA, October 3-6. Bowlin, K.R., Chea, C.K., Wheeler, S.S., and Waldo, L.A., 1997, Field Application of In-situ Gravity Segregation to Remediate Prior Water Coning, Paper SPE 38296, SPE Western regional Meeting, Long Beach, CA, June 25-27. Brekke, K, and Lien, S.C., 1992, New and Simple Completion Methods for Horizontal Wells Improve the Production Performance in High-Permeability, Thin Oil Zones, Paper SPE 24762, 67th Annual Technical Conference and Exhibition, Washington DC, October 4-7. Brill, J.P.(ed.), 1999, Multi-phase flow in Pipes, SPE Monograph Volume Richardson, TX, pp 5 20. 173 ,

Broman, W.H., Stagg, T.O., and Rosenzweig, J.J.,1990, Horizontal Well Performance Evaluation at Prudhoe Bay, Paper CIM/SPE 90-124, Joint International Technical Meeting of the Petroleum Society of the CIM and the SPE, Calgary, Canada, June 10 13. Broughton, G., and Squires, L., 1937, The Viscosity of Oil-Water Emulsion, Journal of Colloid Science 5, Academy Press, New York, NY, pp 253-263. Chappelear, J.E., and Hirasaki, G.J., 1974, A Model of Oil-Water Coning for TwoDimensional, Areal Reservoir Simulation, Paper SPE 4980, 49th Annual Fall Meeting, Houston, October 6-9. Chen, H., 1993, Performance of Horizontal Wells, Safah Field, Oman, Paper SPE 25568, SPE Middle East Oil Technical Conference and Exhibition, Bahrain, April 3-6. Chaperon, I., 1986, Theoretical Study of Coning Toward Horizontal and Vertical Well in Anisotropic Formations: Subcritical and Critical Rates; paper SPE 15377, 61st Annual Technical Conference and exhibition, New Orleans, LA, October. 5-8 Chugbo, A.I., Roux, G.D., and Bosio, J.C., 1989, Thin Oil Columns: Most People Think Horizontal Wells, Obagi Field Case Suggests the Contrary, Paper SPE 19599, SPE 64th Annual Technical Conference and Exhibition, San Antonio, TX, October 8-11. Civan, F and Donaldson, E.C.,1987, Relative Permeability from Unsteady State Displacements With Capillary Pressure Included SPE 16200, SPE Production Operations Symposium, Oklahoma, March 8-10. Colorado state, 2001, Production and Sales by County Monthly, www.oilgas.state.co.us/statistics.html Craft, B.C. and Hawkins, M.F., Terry , R.E., 1991,Applied Petroleum Reservoir Engineering Second Edition, Prentice Hall, Englewood Cliffs, New Jersey. Craig, F.F., (Jnr.), 1993, Reservoir Engineering Aspects of Waterflooding , Monograph Vol. 3, SPE Textbook Series, Richardson, TX, p76. Cunningham, A.B., Jay, K.L., and Opstad, E.,1992, Unconventional Prudhoe Bay Wells Benefit From MWD Use, World Oil, May, p43-52. Dikken, B.J., 1990, Pressure Drop in Horizontal Wells and Its Effect on Production Performance, Paper SPE 19824, Journal of Petroleum Technology, November., pp1426-1433. 174

Driscoll, V.J., 1972, Multiple Producing Intervals to Suppress Coning, US Patent No. 3,638,731, Feb. 1. Ehlig-Economides,C.A., Chan, K.S., and Spath, J.B.,1996, Production Enhancement Strategies for Strong Bottom Water Drive Reservoirs, Paper SPE 36613, Annual Technical Conference and Exhibition, Denver, Colorado, October 6-9. Emerson, A. B., Jones, R. C., and Lim, S. B., 1998, Multi-Lateral Case Histories in the Thailand, Malaysia and Brunei Area, Paper SPE 47810, IADC/SPE Asia Pacific Drilling Conference, Jakarta, Indonesia, September 7-9. EPA (United States Environmental Protection Agency), 1995, Crude Oil and Natural Gas Exploration and Production Wastes: Exemption from RCRA Subtitle C Regulation, Solid Waste Emergency Response 5305, EPA530-K-95-003, May Fisher, W.G, Letkeman, J.P., and Tetreau, E.M., 1970, The Application of Numerical Coning Model to Optimize Completion and Production Methods To Increase Oil Productivity in Bellshill Lake Blairmore Pool Journal of Canadian Petroleum Technology, Oct. Dec., p33-39. Flores, J. G., Chen, X.T., Sarica, C., and Brill, J.P., 1999, Characterization of Oil-Water Flow Patterns in Vertical and Deviated Wells, Paper SPE 56108, SPE Journal of Production and Facilities, May, p102-109. Geoquest Schlumberger, 1997, Eclipse 100/200 Technical Description Manual, Schlumberger Technology Corporation Giger, F.M., Reiss, L.H., and Jourdan, A.P., 1984, The Reservoir Engineering Aspects of Horizontal Drilling Paper SPE 13024, SPE 59th Annual Technical Conference and Exhibition, Houston, TX, September 16-19. Giger, F.M., 1989, Analytical Two-Dimensional Models of Water Cresting Before Breakthrough for Horizontal Wells, Paper SPE 15378, SPE Reservoir Engineering Journal, November, p409-416. Gilman, J.R., 1991, Discussion of Productivity of a Horizontal Well, Paper SPE 21610 SPE Reservoir Engineering Journal, February, p147-148. Gonzalez-Guevara, J.A., and Camacho-Velazquez, R., 1996, A Horizontal Well Model Considering Multiphase Flow and The Presence of Perforations, Paper SPE 36073, Fourth Latin American and Caribbean Petroleum Engineering Conference, Port Spain, Trinidad and Tobago, April 23-26. Guo, B.and, Lee, R.L.H., 1993, A Simple Approach to Optimization of Completion Interval in Oil/Water Coning Systems, Paper SPE 23994, SPE Reservoir Engineer Journal, November, p249-255. 175

Guo, B., Molinard, J.E., Lee, R.L., 1992, A General Solution of Gas/Water Coning Problems for Horizontal Wells, Paper SPE 25050, SPE European Petroleum Conference, Cannes, France, November 16-18. Halliburton, (2001), Reservoir Conformance Technology Maximizing Your Reservoir Value Through The Management of Unwanted Water and Gas, (www.halliburton.com) , Document No. H01027, Hansen, K.L. and Verhyden, M., 1991, Horizontal Well Solves Water Coning, Petroleum Engineer International, Houston, TX, September 16. Hassler, G.L. and Brunner, E.,1944, Measurement of Capillary Pressures in Small Core Samples AIME, Los Angeles Meeting, October. Henley, D.H., Owens, W.W., and Craig, F.F., Jnr, 1960, A Scale-Model Study of Bottom-Water Drives Paper SPE , 35th Annual Meeting Heysel, M., 1992, Horizontal Well Performance in the Dina Sandstone in the Provost Area of Alberta, Paper CIM 92-93, Annual Technical Conference of Petroleum Society of the CIM, Calgary, Alberta, Canada, June 7-10. Ibrahim, A., Desbrandes, R., and Bassiouni, Z., 1994, Derived Capillary Pressure From Well Logs, Petroleum Engineer International, New York, July, pp 38-41 Inikori, S.O., and Wojtanowicz, A.K., 2001(A), Assessment and Inclusion of Capillary Pressure/Relative Permeability Hysteresis Effects in Downhole Water Sink (DWS) Well Technology for Water Coning Control, ETCE2001-17100, 23rd ASME Energy Sources Technology Conference and Exposition, Houston, February 5-7. Inikori, S.O. and Wojtanowicz, A.K., 2001(B), Contaminated Water Production in Old Oil Fields With Downhole Water separation: Effects of Capillary Pressures and Relative Permeability Hysteresis Paper SPE 66536, SPE/EPA/DOE Exploration and Production Environmental Conference, San Antonio, TX, February 26-28. Inikori, S.O., and Wojtanowicz, A.K., 2002, Effect of 2-Phase Oil-Water Flow and Friction Pressure Loss in The Wellbore on Water Cresting in Horizontal Wells, ETCE2002/Prod-29165, 25th ASME Annual Engineering Technology Conference on Energy, Houston, TX, February 4-5. Jiang, W., Sarica, C., Ozkan, E., and Kelkar, M., 2000, Investigation of the Effects of Completion Geometry on Single Phase Liquid Flow Behavior in Horizontal Wells, ETCE2000/Prod-10061, Energy Sources Technology Conference and Exhibition, New Orleans, February 14-17. 176

Joshi, S.D., 1991, Horizontal Well Technology, Pennwell Publishing Company, Tulsa, OK., pp 389-420 Karcher, B.J., Giger, F.M., and Combe, J., 1986, Some Practical Formulas To Predict Horizontal Well Behavior, Paper SPE 15430, 61st Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, New Orleans, LA, October 5-8. Kimbrell, C. W., 2001, Produced Water in Louisiana: Analyzing the Magnitude of the Problem, Proceedings of PTTC Symposium on Water Production, Baton Rouge, LA, April 5. Koonsman, T.L., and Purpich, A.J., 1991, Ness Horizontal Well Case Study, Paper SPE 23096, SPE Offshore Europe Conference, Aberdeen, Scotland, September 36. Kurban, H., 1999, Numerical Simulation of Downhole Water Sink Production System Performance, M.Sc. Thesis, Louisiana State University and A & M College, Baton Rouge, LA, December Landmark, 1996, VIP- Executive Reference Landmark Corporation, Houston, TX Letkeman, J.P., and Ridings, R.L., 1970, A Numerical Coning Model, Paper SPE 2812, Second Symposium on Numerical Simulation of Reservoir Performance, Dallas, TX, February 5-6. Leverett, M.C., 1941, Capillary behavior in Porous solids, Transactions of AIME 142, pp152 - 169. Lien, S.C., Seines, K., and Havig, S.O., 1990, 1991, The First Long-Term HorizontalWell Test in the Troll Thin Oil Zone, Paper 20715, SPE Annual Technical Conference and Exhibition (JPT, August 1991, p914-974), New Orleans, LA, September. MacDonald, R.C., and Coats, K.H., 1970, Methods for Numerical Simulation of Water and Gas Coning, Paper SPE 2796, Second Symposium on Numerical Simulation of Reservoir Performance, Dallas, TX, February 5-6. Mattax, C.C and Dalton, R. (ed.), 1989, Reservoir Simulation, Monograph Vol. 13, SPE Textbook Series, Richardson, TX, p135. McMullan, J.H., and Larson, T.A.,2000, Impact of Skin Damage on Horizontal Well Flow Distribution, Paper SPE/Petroleum Society of CIM 65507, SPE/Petroleum Society of CIM International Conference on Horizontal Well Technology, Calgary, Canada, November 6-8.

177

Marett, B.P., and Landman, M.J., 1993, Optimal Perforation Design for Horizontal Wells in Reservoirs With Boundaries, Paper SPE 25366, Asia Pacific Oil and Gas Conference and Exhibition, Singapore, February 8-10. Martinez, A.E., Arirachakaran, S., Shoham, O., and Brill, J.P., 1988, Prediction of Dispersed Viscosity of Oil/Water Mixture Flow in Horizontal Pipes, Paper SPE 18221, SPE 63rd Annual Technical Conference and Exhibition, Houston, TX, October 2-5. Miller, R.T. and Rogers, W.L., 1973, Performance of Oil Wells in Bottom Water Drive Reservoirs Paper SPE 4633, 48th Annual Meeting, Las Vegas, NV, September. 30-October. 3 Muskat, M., 1949, Physical Principles of Oil Production, McGraw-Hill Book Co. Inc., New York City, p. 226-236. Muskat, M, and Wyckoff, R.D., 1935, An Approximate Theory of Water-Coning in Oil Production AIME Transaction, Vol. 114, p144-163. Mungan, N., 1975, A Theoretical and Experimental Coning Study, Paper SPE 4982, SPE-AIME 49th Annual Fall Meeting, Houston, TX, October 6-9. Murphy, P.J., 1988, Performance of Horizontal Wells in the Helder Field, Paper SPE 18340, SPE European Petroleum Conference, London, U.K., October 16-19. Nzekwu, B., 1992, Case Histories of Worldwide Application of Horizontal Wells For The Abatement of Water/Gas Coning, 13th IEA Collaborative Project on Enhanced Oil Recovery Symposium on Abatement of Water and Gas Coning, Banff, Alberta, September 30. Ozkan, E, Sarica, C., and Haci, M., 1993, 1999, Influence of Pressure Drop Along the Wellbore on Horizontal Well Productivity, Paper SPE 25502 and 57687, SPE Journal 4 (3), September. Papatzacos, P., Herring, T.R., Martinsen, R., and Skjaeveland, S.M., 1989, 1991, Cone Breakthrough Time for Horizontal Wells, Paper SPE 19822, SPE Annual Technical Conference and Exhibition (Reservoir Engineering Journal, August 1991), San Antonio,TX., October 8-11. Peaceman, D.W., 1991, Further Discussion of Productivity of a Horizontal Well, Paper SPE 21611, SPE Reservoir Engineering Journal, Richardson, TX, p149 Peng, C.P., and Yeh, N, 1995, Reservoir Engineering Aspects of Horizontal Wells Application to Oil Reservoirs with Gas or Water Coning Problems, Paper SPE 29958, International Meeting on Petroleum Engineering, Beijing, PR China, November 14-17. 178

Permadi, P., Lee, R.L., and Kartoatmodjo, R.S.T., 1995, Behavior of Water Cresting Under Horizontal Wells, Paper SPE 30743, Annual Technical Conference and Exhibition, Dalls, TX, October 22-25. Permadi, P., 1996, Water Cresting Behavior Under High Angle Wells: An Experimental Investigation, Paper SPE 35715, Western Regional Meeting, Anchorage, Alaska, May 22-24. Permadi, P., Wibowo, W., Alamsyah, Y., and Pratomo, S.W., 1997, Horizontal Well Completion With Stinger for Reducing Water Coning Problems, SPE Production Operations Symposium, Oklahoma City, OK, March 9-11. Pirson, S.J., and Mehta, M.M., 1967, A Study of Remedial Measures for Water-Coning By Means of a Two-Dimensional Simulator, Paper SPE 1808, 42nd Annual Fall Meeting of the Society of Petroleum Engineers of AIME, Houston, TX, October 14. Reed, R.N., and Wheatley, M.J., 1984, Oil and Water Production in a Reservoir With Significant Capillary Transition Zone, Paper SPE 12066, Journal of Petroleum Technology, September, p1559-1566 Renard, G., Gadelle, C., Dupuy, J.M., and Alfonso, H., 1997, Potential of Multilateral Wells in Water Coning Situations, Paper SPE 39071, Fifth Latin American and Caribbean Petroleum Engineering Conference and Exhibition, Rio De Janeiro, Brazil, August 30 September 3. Richardson, E.G., 1932, Kooloid Z, . (65) 1933 p.32 Richardson, E.G., 1959, The Formation and Flow of Emulsion, Journal of Colloid Science 5, Academy Press, New York, NY, pp 404-413. Ronningsen, H.P., 1995, Correlation for predicting Viscosity of W/O-Emulsions based on North Sea Crude Oils, Paper SPE 28968, SPE International Symposium on Oilfield Chemistry, San Antonio, TX, February 14-17. Schols, R.S., 1972, An Empirical Formula for the Critical Oil Production Rate, Erdoel Erdgas,Z, Vol. 88, No.1 , January, p6-11 Schramm, L.L. (ed.), 1992, Emulsions Fundamentals and Applications in the Petroleum Industry Advances in Chemistry Series 231, American Chemical Society, Washington, DC, pp 1-62, 141-159. Seines, K., Aavatsmark, I., Lien, S.C., and Rushworth,P., 1993, Considering Wellbore Friction Effects in Planning Horizontal Wells, Journal of Petroleum Technology (JPT), October pp 994 1000 179

Sherrard, D.W., Brice, B.W., and MacDonald, D.G., 1986, Application of Horizontal Wells at Prudhoe Bay, Paper SPE 15376, 61st Annual Technical Conference an Exhibition, New Orleans, LA, October 5-8. Shirman, E.I., and Wojtanowicz, A.K., 1998, More Oil with Less Water Using Downhole Water Sink Technology: A Feasibility Study, Paper SPE 49052, Annual Technical Conference and Exhibition, New Orleans, LA, October 27-30. Sobocinski, D.P., and Cornelius, A.J., 1965, A Correlation for Predicting Water Coning Time, Journal of Petroleum Technology, May, p594-600. Soleimani, A., Lawrence, C.J., and Hewitt, G.F., 1999, Spatial Distribution of Oil and Water in Horizontal Pipe Flow, Paper SPE 56524, SPE Annual Technical Conference and Exhibition, Houston, TX, October 3-6. Su, Z. and Gudmundsson, J.S., 1993, Friction Factor of Perforation Roughness in Pipes, Paper SPE 26521, 68th Annual Technical Conference and Exhibition, Houston, TX., October 3-6. Suprunowicz, R, and Butler, R.M., 1992, Discussion of Productivity of a Horizontal Well, Paper SPE 25295, SPE Reservoir Engineering Journal, November, p453454 Suprunowicz, R, and Butler, R.M., 1993, Further Discussion of Productivity of a Horizontal Well, Paper SPE 26262, SPE Reservoir Engineering Journal, May, p160 Swisher, M.D., and Wojtanowicz, A.K., 1995, New Dual Completion Eliminates Bottom Water Coning, Paper SPE 30697, SPE Annual Technical Conference & Exhibition, Dallas, TX, October 22-25. Swisher, M.D., and Wojtanowicz, A.K., 1995, In-situ Segregated Production of Oil and water A Production Method With Environmental Merit: Field Application, Paper SPE 29693, SPE/EPA Exploration and Production Environmental Conference, Houston, TX, March 27-29. Tang, Y., Ozkan, E., Kelkar, M., Sarica, C., and Yildiz, T., 2000, Performance of Horizontal Wells Completed with Slotted Liners and Perforations, Paper SPE/PS - CIM 65516, SPE/Petroleum Society of Cim International Conference on Horizontal Well Technology, Calgary,Alberta,Canada, November 6-8. Tehrani, A.D.H., 1992, An Overview of Horizontal Well Targets Recently Drilled in Europe, Paper SPE 22390, SPE International Meeting on Petroleum Engineering, Beijing, China, March 24-27. 180

Trallero, J.L., Sarica, C., and Brill, J.P., 1996, A Study of Oil-Water Flow Patterns in Horizontal Pipes, Paper SPE 36609, SPE Annual Technical Conference and Exhibition, Denver, CO., October 6-9 Wanty, R. B., 1997, USGS Research on Saline Waters Co-Produced With Energy Resources, Fact Sheet FS-003-97, http://greenwood.cr.usgs.gov/energy/factshts/003-97/FS-003-97.html Weinstein, H.G., Chappelear, J.E., and Nolen, J.S., 1986, Second Comparative Solution Project: A Three Phase Coning Study, SPE paper. Journal of Petroleum Technology, March. Widmyer, R.H.: Producing Petroleum from Underground Formations, U.S. Patent No. 2,855,047, Oct. 3, 1955 Willhite, G.P.,, 1986, Waterflooding, Third Printing, SPE Textbook Series, Richardson, TX, pp. 2 & 5. Wheatley, M.J., 1985, An Approximate Theory of Oil/Water Coning, Paper SPE 14210, 60th Annual Technical Conference and Exhibition, Las Vegas, NV, September 22-25. Wojtanowicz, A.K., and Shirman, E.I., 1994, An In-situ Method for Downhole Drainage injection of Formation Brine in a Single Oil-Producing Well, Proc. Int. Symposium on Scientific and Engineering Aspects of Deep Injection Disposal of Hazardous and Industrial Wastes, Lawrence Berkeley Laboratory, Berkeley, CA, May 10 13. Wojtanowicz, A.K., Hui Xu, and Bassiouni, Z., 1991, Oilwell Coning Control Using Dual Completion With Tailpipe Water Sink, Paper SPE 21654, SPE Production Operation Symposium, Oklahoma City, OK, April 7-9. Wolff, E.A., 2000, Reduction of Emissions to Sea by Improved Produced Water Treatment and Subsea Separation Systems, Paper SPE 61182, SPE International Conference on Health, Safety, and Environment in Oil and Gas Production, Stavanger, Norway, June 26-28. Wu, G., Reynolds, K., and Markitell, B., 1995, A field Study of Horizontal Well Design in Reducing Water Coning, Paper SPE 30016, International Meeting on Petroleum Engineering, Beijing, PR China, November, 14-17. Yang, W. and Wattenbarger, R.A., Water Coning Calculations for Vertical and Horizontal Wells; Paper SPE 22931, 66th Annual Technical Conference and exhibition, Dallas, TX, October. 6-9

181

Yokoyama, Y, and Lake, Larry W., 1981, The Effect of Capillary Pressure on Immiscible Displacements in Stratified Porous Media, Paper SPE 10109, 56th Annual Fall Technical Conference and Exhibition of Society of Petroleum engineers of AIME, San Antonio, TX, October 5-7. Yuan, H, 1994, Investigation of Single Phase Liquid Flow Behavior in a Single Perforation Horizontal Well, M. Sc. Thesis, The university of Tulsa, Tulsa, OK., Yuan, H, 1997, Investigation of Single Phase Liquid Flow Behavior in Horizontal Wells, PhD. Dissertation, The university of Tulsa, Tulsa, OK., Yuan, H. J., Sarica, C., and Brill, J.P., 1999, Effect of Perforation density on SinglePhase Liquid Flow Behavior in Horizontal Wells, Paper SPE 57395, SPE Journal of Production and Facilities, Richardson, TX., August, pp 203-209.

182

APPENDIX A CAPILLARY PRESSURE/RELATIVE PERMEABILITY HYSTERESIS DATA DECK

-- ============================================================== -- DONWHOLE WATER SINK TECHONOLOGY SIMULATION RUN ON ECLIPSE 100 -- IPW WINDOW VERIFICATION ON HILAL'S PROJECT DATA - AUGUST 1999 -- PERMEABILITY HYSTERESIS STUDY ON DWS IPW WINDOW (30 FT CAPILLARY) -- ================================================================ RUNSPEC TITLE DWS oil/water coning test run -- SOLOMON O INIKORI DIMENS 31 1 20 / RADIAL NONNC OIL WATER -- gas -- disgas FIELD WELLDIMS 4 20 1 4 / AQUDIMS 1 1 1 1 1 40 / SATOPTS 'HYSTER' / TABDIMS 2 / START 1 'AUG' 1999 / NSTACK 100 / MESSAGES 8* 5000 2*800 / UNIFOUT GRID -====================================================================== -SPECIFY INNER AND OUTER RADIUS OF IST AND LAST GRID BLOCK IN THE RADIAL DIRECTION INRAD 0.292 / -- drv -0.093 0.123 0.163 0.215 0.283 0.374 0.493 0.651 0.859 -1.133 1.495 1.973 2.604 3.436 4.534 -5.984 7.896 10.420 13.751 18.147 23.948 31.602 -41.704 55.035 72.628 95.843 126.480 166.909 220.263 290.670 200 / OUTRAD 850 EQUALS

183

'DTHETA' 'PERMR' 'PERMZ' 'PERMR' 'PERMZ' 'PORO' 'PORO' 'DZ' / 'DZ' /

360 / 1000 500 1000 500 0.30 0.37

1 1 1 1 1 1 5.0 5.0

31 31 31 31 31 31

1 1 1 1 1 1 1 1

1 1 1 1 1 1 31 31

1 1 11 11

10 10 20 20 1 11 10 20 1 1

/ / / / / / 1 11 10 20

1 1

'TOPS' 4720 1 31 1 1 1 1 / / RPTGRID 'DX' / INIT GRIDFILE 1 / PROPS ===================================================================== -NO RELATIVE PERMEABILITY DATA SUPPLIED. EMPIRICAL DATA FROM -COREY'S CORRELATION USED SWOF -- Sw Krw Kro Pc -0.20 0.0000 1.0000 1.428 0.30 0.0002 0.8374 1.250 0.40 0.0039 0.6328 1.071 0.55 0.0366 0.3337 0.803 0.70 0.1526 0.1187 0.536 0.80 0.3164 0.0391 0.357 1.00 1.0000 0.0000 0.000 / TABLE 1 -Sw 0.20 0.30 0.40 0.55 0.70 0.80 1.00 Krw 0.0000 0.0029 0.0128 0.0605 0.1898 0.3547 1.0000 Kro 1.0000 0.6944 0.4444 0.1736 0.0278 0.0000 0.0000 Pc -1.428 1.250 1.071 0.803 0.536 0.357 0.000

/ TABLE 2

EHYSTR 1* 4

1.0

0.1

'KR'

'RETR'

PVTW -WATER PVT DATA (REF. PRESSURE, WATER FM VOL. FACTOR, VISCOSITY -COMPRESSIBILITY, ETC -Pref Bw(ref) Cw Visw Viscosibility 1788 1.02 3.0e-6 0.46 0 / PVCDO

184

-OIL PVT DATA FOR DEAD OIL WITH CONSTANT SOLUTION GAS (Bo AND Co NOT GIVEN-ASSUMED) -Pref Bo Co Viso Viscosibility 1788 1.26 1.5e-5 1.25 0 / RSCONST -DATA FOR THIS SECTION ALSO NOT SUPPLIED. -Rs Pb 0.379 1000 / ROCK -DATA ALSO NOT SUPPLIED. -Pref Cf 1788 4.0e-6 / DENSITY -OIL WATER GAS 53.9 62.47 0.0702 / REGIONS =================================================================== SATNUM 620*1 / IMBNUM 620*2 /

SOLUTION ================================================================== -DATUM DATUM OWC OWC GOC GOC RSVD RVVD SOLN -DEPTH PRESS DEPTH PCOW DEPTH PCOG TABLE TABLE METH EQUIL 4770 1788 4770 0 1000 0 / AQUFET -- 1 2 3 4 5 6 7 8 9 10 11 12 13 -- ITEM 14 - INITIAL SALT CONCN. WILL BE RUN AS A SENSITIVITY LATER 4770 1788 10E06 7.0E-6 20 1 1 31 1 1 10 20 'K+' / RPTSOL 'RESTART=2' 'FIP=2' / SUMMARY ==================================== WOPR 'P1' WWPR 'P1' WGPR 'P1' WWCT 'P1' FOPR FOPT FWPR FWPT FLPR 'P2' 'P2' 'P2' 'P2' / / / /

185

-- WELL BOTTOM HOLE PRESSURES WBHP 'P1' 'P2' / FWCT FPR FOE WOPT 'P1' 'P2' / WWPT 'P1' 'P2' / -RPTONLY -PRESENT DATA IN TABULAR FORM FOR GRAF PURPOSE IN PRINT FILE -REQUEST RUNSUM OUTPUT TO GO TO A SEPARATE RSM FILE RUNSUM SEPARATE SCHEDULE ======================================= -- SET MAXIMUM NUMBER OF LINEAR ITERATIONS IN THE NEWTON ITERATION TUNING / / LITMAX 2* 100 / RPTSCHED 'RESTART=2' 'FIP=2' / RPTRST 'BASIC=4' / -- INTRODUCING THE WELLS -- WELL GROUP -- NAME NAME PHASE WELSPECS 'P1' 'G' 1 1 4720 'P2' 'G' 1 1 4780 'i1' 'G' 31 1 4720 'i2' 'G' 31 1 4780 LOCATION I 'LIQ' 'LIQ' 'oil' 'water' J / / / / BHP PREF DEPTH

/ -SPECIFYING COMPLETION DATA -- WELL - LOCATIONOPEN WELL -NAME I J K1 K2 COMPDAT 'P1' 1 1 1 1 'P2' 1 1 13 14 'i1' 31 1 1 1 'i2' 31 1 13 14 / COMPIMB -WELL NAME 'P1' I-LOCN 4* J-LOCN 2 /

SAT SHUT 'OPEN' 'OPEN' 'OPEN' 'OPEN' TAB 1* 1* 1* 1*

CONN FACT 1* 1* 1* 1* DIA 0.585 0.585 0.585 0.585 / / / /

K-UPR

K-LWR

IMB TABLE #

186

'P2' /

4*

-- PROD. WELL CONTROLS WCONPROD -1 2 3 -- WELL OPEN/ CNTL -- NAME SHUT MODE 'P1' 'OPEN' 'P2' 'OPEN' /

DRAINAGE RATE IS VARIED AT FIXED OIL RATE. 4 5 OIL WATER RATE RATE 'LRAT' 'LRAT' 6 7 GAS LIQUID RATE RATE 3* 110 1* 3* 180 1* 8 RES RATE 1* 1* 9 BHP / /

wconinje -- WELL INJT/ OPEN/ CNTL/ 5 6 7 -- NAME TYPE SHUT MODE 'I1' 'OIL' 'OPEN' 'GRUP' 3* / 'I2' 'WATER' 'OPEN' 'GRUP' 3* / / GCONINJE 'FIELD' 'OIL' 'REIN' 2* 1.00 / 'FIELD' 'WATER' 'REIN' 2* 1.00 / / -SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT -- WECON -WELL MIN MIN MAX MAX MAX WKVR -NAME OIL GAS WATER GOR GWR PRCDR -RATE RATE CUT -- 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT FOUR YEAR INTERVAL FOR 20 YEARS TSTEP 48*30.4 TSTEP 48*30.4 TSTEP 48*30.4 TSTEP 48*30.4 TSTEP 48*30.4 / / / / /

END RUN

END

187

APPENDIX B WATER CONE REDUCTION STUDY ON WELL LVT B.1 Executive Summary

A watered-out well LVT had been selected as a candidate for re-completion, installation, and field implementation of the Downhole Water Sink (DWS) technology for water coning control. Prior to watering out, the well had been produced for several years. As a result of water coning a sizable zone of high water saturation has developed around the well. The estimated watercut of the well before it was shut down was between 94 and 98 percent. The LVT well has 37 ft of oil sand underlain by 87 ft of water. The reservoir structure represents an edge water drive system with strong aquifer support. Initially, prior to the movement of the Oil-Water Contact (OWC), the well had 45-foot long perforated section covering the entire length of the oil bearing sand. Later, these perforations were squeezed off and the top 15 ft re-perforated (9540 9555 ft). A bottom section of water sink completion had been recently added within the interval: 9596 9606 ft. Also, an isolation packer has been set around the OWC to isolate fluids produced from the top and bottom completion through separate tubing strings.
B.2 Objective

Objective of this study was to determine the approximate length of time required to bring watercut in the top production from current level of 94-98 percent to a manageable level. Also, the study shall verify the effect of prolonged process of watering out wells on reinstating their productivity to oil with DWS. In addition, the study shall evaluate the effect of capillary pressure and leaking cement on the process of water saturation drainage around the well with DWS. 188

B.3

Input Data for Eclipse Simulator

Capillary pressure data was supplied by the Joint Industry Panel (JIP) member. The data had been collected from another well in the M2 reservoir with permeability much lower than that in LVT well. The data was then used to construct the Leveretts Jfunction which, in turn, was used to generate an estimated capillary pressure correlation for LVT well. The correlation/curve is as shown in Figure B-1,

8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 0.00

Capillary Pressure (psi)

0.20

0.40

0.60

0.80

1.00

Water Saturation (Sw - Fraction)

Figure B-1 Capillary Pressure Plot for LVT Well

Also, a slight linear interpolation of the relative permeability data from M2 was used to generate estimates for LVT well calculations (Figure B-2). The value of contact angle of 140 degrees was ignored in the generation of the Leverett function, as it does not represent a water-wet reservoirs. Also, direction of measurement was not specified so we could not use the complementary angle. 189

Relative Permeabilities (Fraction)

1.000 0.900 0.800 0.700 0.600 0.500 0.400 0.300 0.200 0.100 0.000 0.000

0.200

0.400

0.600

0.800

1.000

Water Saturation (Sw - fraction)

Figure B-2 Relative Permeability Curve for LVT Well Studies

Other modification of the provided input data was the increase of the rock compressibility value to solve simulators instability problem resulting from rapid changes in production rates. Also, in order to obtain a good history match of the well oil and water production profile, the vertical permeability was reduced to 1/100th of the horizontal permeability. All other data for this study were same as for previous work by the LSU DWS team.
B.4 Approach to Study

To create a conceptualized model of the reservoir with edge water drive, certain grids in the radial model were assigned zero porosity and permeability representing shaled out areas. The nature of the reservoir employed for this study is shown in Figure B-3. Prior to forecasting watercut reduction in LVT well efforts were made to simulate the actual behavior of the well from the beginning of initial production. The approach was necessary to determine the extent of water invasion and to create a more realistic picture of the current status of this well. 190

Figure B-3 LVT Well Reservoir Grid Model

1.20 7000 st b/ day 100 st b/ day

1.00

Nat ural Cone Collapse at Low rat e 100BOPD

Watercut (Fraction)

0.80

0.60

Cone Collapse w/DWS / Cap. transition Channel ef fect Cone Collapse wit h DWS No Cap.

0.40

0.20

0.00 0 1000 2000

Channel ef fect 3000 4000 5000 6000 7000

Time (Days)
No Sink Prod.-No Cap DWS wit h Channel Wit h Sink Prod.-No Cap DWS- FNC W/ Sink&Mod.Cap

Figure B-4 Watercut Reduction Performance of DWS Technology

191

The next step was to simulate reduction of watercut due to the activation of DWS drainage under several scenarios (Figure B-4). The study also investigated the rate of natural collapse of water cone around the wellbore due to a reduction in the drawdown pressure simply by a rate reduction. This is a typical industry practice as a solution to reducing the adverse effect of water cone development. The top completion was put at 100 stb/day production without the DWS. The performance shown in Figure B-5 as a plot marked watercut without DWS Low rate and no capillary Transition or,

Natural Cone Collapse indicates that it will take a life time for the cone to naturally reduce to an appreciable level even at the very low oil production rate once the cone is fully developed around the wellbore. It shows that fully developed water cone in an oil well may not retract quickly even if the well is shut-in for a considerable length of time.

1.20

Watercut Performance (Fraction)

1.00 0.80 0.60

Watercut without DWS - Low rate - No Capillary Transition

Watercut with DWS and Capillary Transition

0.40 0.20 0.00 0 1000 2000 3000 4000 Time (Days) from start of production with/out DWS
DWS No Cap(1 00 B LP D to p; 1 0,000 B LP D btm) DWS w/Cap (1 00 -1 0,000 B LP D) No DWS No Cap(to p 1 00 B LP D)

Watercut w/DWS No Capillary

Figure B-5 Watercut Forecast of LVT Well without Capillary Pressure

192

B.5

Discussion of Results

Several scenarios have been investigated. They include watercut performance with or without capillary pressures using a top rate of 100 stb/day and a bottom drainage rate of 10,000 stb/day.
B.5.1 Scenario A Watercut Reduction Without the Effect of Capillary Pressure

The top completion was put on production at a rate of 100 stb/day. Prior to production, the well had been shut-off for 30 days representing the period of workover when the well was not producing. The water sink production was initiated at a rate of 10,000 stb/day. The results showed several stages of watercut decline corresponding to the extent of cone development. At the initial stage, water production from the sink did not affect the watercut at the top completion appreciably. This is because, water cone had developed over several years creating and extensive radius of watered out zone around the wellbore.

Watercut Performance (Fraction)

1.20 1.00 0.80 0.60 0.40 0.20 0.00 0

Watercut w ith DWS and Capillary Transition

Watercut w ithout DWs - Low rate - No Capillary Transition

Watercut w /DWS No Capillary Transition

200

400

600

800

Time (Days) from start of production with/out DWS

DWS No Cap(1 00 B LP D to p; 1 0,000 B LP D btm) DWS w/Cap (1 00-1 0,000 B LP D) No DWS No Cap (to p 1 00 B LP D)

Figure B-6 Watercut Reduction Forecast with and without Capillary Pressure

193

To reverse such a large cone the radial spread of the cone has to be minimized to few feet around the well before any appreciable decrease in watercut could be noticed. Figure B-6 is a highlighted plot of Figure B-5. The Figure B-6 show watercut reduction of only about 1.5 percent during the first fifty days of drainage. Also, it took about 270 days of continuous pumping water from the bottom completion to reduce the watercut to 80 percent. The results indicate that reduction in watercut would not be noticeable in a couple of days or perhaps a few weeks. Results of this study correspond qualitatively to those obtained by another JIP member in their field test of DWS. They reported steady reduction of watercut of 0.02 percent/day. The rate of watercut reduction could increase if the vertical permeability is greater than that used in the simulation.
B.5.2 Scenario B Watercut Reduction Performance with Capillary Pressure Effect

This case is marked, with capillary transition, in Figure B-6. Modified capillary transition effect was incorporated in the study. To achieve 37 ft of oil sand the OWC was moved to the original position. Well production history matching shows an increased limiting watercut of about 98 percent which corresponds better to its current value in LVT Well. Inclusion of capillary effects delays any noticeable change in watercut from the top completion. Over three years of water pumping is required to reduce the watercut at the top completion from 98 percent to 80 percent. Similar to the previous case, a dramatic reduction in watercut occurs after the cone size has been reduced to a few feet around the wellbore. Also, the minimum watercut achievable with capillary transition is about 27 194

percent as compared to 10 percent without the capillary transition effect. It might seem that the effect of DWS technology is negligible in the case with capillary transition zone looking at Figure B-6. However, this is for a period of about three years. Beyond this period, the effect of the DWS water sink completion becomes impressive in reducing watercut. This could be seen in Figure B-4.

B.5.3 Scenario C Watercut Reduction Performance with Leaking Cement

This case is marked, channel effect. in plots in Figure B-4. This scenario was included to predict the behavior of the cone/watercut at top completion in a case where there is no cement integrity behind the casing. The simulation run was made for 100 stb/day top rate and 10,000 stb/day water sink rate. The results showed that cement channeling would actually enhance watercut reduction. In the presence of cement channel, watercut would reduce to about 80 percent within three weeks of pumping.
B.5.4 Scenario D Watercut Reduction Performance with High Rate at Top Completion
1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0 1000 2000 3000 4000 Time (Days from start of DWS)

Figure B-7 Watercut Performance with High Rate at Top Completion

Watercut (Fraction)

195

This case is depicted in B-7. Watercut reduction was simulated with 800 stb/day from top and 10,000 stb/day from bottom water sink completion for the case without the effect of capillary pressure. The results show that the watercut reduction is very gradual. Watercut reduces to about 80 percent after about one year of continuous pumping. Also, the minimum achievable value of watercut is about 27 percent the value similar to the case with low top rate and capillary pressure effect.
B.6 Appropriate Watercut Level for Deployment of DWS Technology

In this section the study evaluated the performance of the DWS technology in reducing watercut in the oil zone completion in a case where the water cone is at the development stage. In effect, the lateral extent of the water cone around the wellbore is limited to a few tens of feet.
0.90 0.80 0.70

Watercut (Fraction)

0.60 0.50 0.40 0.30 0.20 0.10 0.00 -0.10 0 1000 2000 3000 4000 5000 6000 7000 Time (Days)

Figure B-8 Watercut Performance of DWS in a Developing Cone Scenario

The result of the study shown in Figure B-8 indicated that the initial delay experienced when the Downhole Water Sink technology is implemented to collapse the 196

cone is absent. This confirms the initial preposition that the initial delay is due to the full development of the water cone over time. The watercut in top completion shows immediate reduction as soon as the water sink in activated.
B.7 Conclusion of Studies

The study demonstrates the effect of well production history on watercut control with DWS. As the well LVT had been seriously watered-out over number of years it might be necessary to allow much longer time for pumping to reduce watercut and restore oil productivity of this well. Also, the results show that the DWS technology does have the capability of reducing the watercut over time but the time frame may be slow depending on the extent of coning severity.

197

APPENDIX C DATA DECK FOR BILATERAL WATER SINK (BWS) CONCEPT FOR WATER CRESTING CONTROL
-- ============================================================== -- DONWHOLE WATER SINK TECHONOLOGY SIMULATION RUN ON ECLIPSE 200 -- SIMULATION OF HORIZONTAL WELL APPLICATION OF DWS - JAN. 2002 -- ============================================================== RUNSPEC TITLE DWS OIL/WATER CONING STUDIES ON HORIZONTAL WELL -- SOLOMON O INIKORI DIMENS -- NX NY NZ 75 20 12 / NONNC OIL WATER -- gas -- DISGAS FIELD WELLDIMS 4 150 1 4 / START 1 'JAN' 2002 / FRICTION 2 1 / NSTACK 100 / MESSAGES 8* 5000 2*200 / UNIFOUT GRID -====================================================================== DXV 40 DYV 60 EQUALS 'DZ' 'DZ' 'PERMX' 'PERMZ' 'PERMX' 'PERMZ' 'PORO' 'PORO' 'PORO' 5 30 1 1 75 75 1 1 1 1 1 1 1 1 1 75 75 75 75 75 75 75 20 20 1 1 1 1 1 1 1 1 12 20 20 20 20 20 20 20 11 12 1 1 9 9 1 9 12 / / TOTAL 85' 8 8 12 12 8 11 12 / / / / / / / STEADY STATE 60 60 60 60 10*60 60 60 60 60 60 / total 1200' 40 40 40 40 65*40 40 40 40 40 40 / total 3000'

8000 800 8000 800 0.31 0.35 1e10

198

'TOPS' / COPY

6154

75

20

'PERMX' 'PERMY' / / INIT GRIDFILE 1 / PROPS ===================================================================== -NO RELATIVE PERMEABILITY DATA SUPPLIED. USED EMPIRICAL CORRELATION DATA SWOF -- Sw Krw Kro Pc -0.10 0.00 0.98 0 0.20 0.00 0.93 0 0.30 0.05 0.60 0 0.40 0.10 0.35 0 0.55 0.18 0.18 0 0.70 0.35 0.06 0 0.85 0.60 0.00 0 1.00 0.99 0.00 0 / PVTW -WATER PVT DATA (REF. PRESSURE, WATER FM VOL. FACTOR, VISCOSITY -COMPRESSIBILITY, ETC -Pref Bw(ref) Cw Visw Viscosibility 2500 1.02 3.0e-6 0.50 0 / PVCDO -OIL PVT DATA FOR DEAD OIL WITH CONSTANT SOLUTION GAS -Pref Bo Co Viso Viscosibility 2500 1.15 1.5e-5 1.34 0 / RSCONST -DATA FOR THIS SECTION ALSO NOT SUPPLIED. -Rs Pb 0.379 1000 / ROCK -DATA ALSO NOT SUPPLIED. -Pref Cf 2500 4.0e-6 / DENSITY -OIL WATER GAS 50.0 64.2 0.0702 /

SOLUTION ================================================================== -DATUM DATUM OWC OWC GOC GOC RSVD RVVD SOLN -DEPTH PRESS DEPTH PCOW DEPTH PCOG TABLE TABLE METH EQUIL 6154 2500 6194 0 1000 0 / RPTSOL

199

'RESTART=2'

'FIP=2'

'PRES'

'SWAT'

SUMMARY ==================================== WOPR 'P1' 'P2' / WWPR 'P1' 'P2' / WGPR 'P1' 'P2' / WWCT 'P1' 'P2' / WOPT 'P1' 'P2' / WWPT 'P1' 'P2' / WBHP 'P1' 'P2' / FOPR FOPT FWPR FWPT FLPR FWCT FPR FOE / -RPTONLY -PRESENT DATA IN TABULAR FORM FOR GRAF PURPOSE IN PRINT FILE -REQUEST RUNSUM OUTPUT TO GO TO A SEPARATE RSM FILE RUNSUM SEPARATE SCHEDULE ======================================= -- SET MAXIMUM NUMBER OF LINEAR ITERATIONS IN THE NEWTON ITERATION TUNING / / LITMAX 2* 100 1* 50 / RPTSCHED 'RESTART=2' RPTRST 'BASIC=4' 'FIP=2' / PREF PHASE 'LIQ' 'LIQ' / / /

-- INTRODUCING THE WELLS -- WELL GROUP LOCATION BHP -- NAME NAME I J DEPTH WELSPECS 'P1' 'G' 20 11 6157 'P2' 'G' 20 11 6201 / --SPECIFYING COMPLETION DATA WELL - LOCATIONOPEN/

SAT

CONN

WELL

200

-NAME I J K1 COMPDAT 'P1' 22 11 1 1 'P1' 23 11 1 1 'P1' 24 11 1 1 'P1' 25 11 1 1 'P1' 26 11 1 1 'P1' 27 11 1 1 'P1' 28 11 1 1 'P1' 29 11 1 1 'P1' 30 11 1 1 'P1' 31 11 1 1 'P1' 32 11 1 1 'P1' 33 11 1 1 'P1' 34 11 1 1 'P1' 35 11 1 1 'P1' 36 11 1 1 'P1' 37 11 1 1 'P1' 38 11 1 1 'P1' 39 11 1 1 'P1' 40 11 1 1 'P1' 41 11 1 1 'P1' 42 11 1 1 'P1' 43 11 1 1 'P1' 44 11 1 1 'P1' 45 11 1 1 'P1' 46 11 1 1 'P1' 47 11 1 1 'P1' 48 11 1 1 'P1' 49 11 1 1 'P1' 50 11 1 1 'P1' 51 11 1 1 'P1' 52 11 1 1 'P1' 53 11 1 1 'P1' 54 11 1 1 'P1' 55 11 1 1 'P1' 56 11 1 1 'P1' 57 11 1 1 'P1' 58 11 1 1 'P1' 59 11 1 1 'P2' 22 11 10 10 'P2' 23 11 10 10 'P2' 24 11 10 10 'P2' 25 11 10 10 'P2' 26 11 10 10 'P2' 27 11 10 10 'P2' 28 11 10 10 'P2' 29 11 10 10 'P2' 30 11 10 10 'P2' 31 11 10 10 'P2' 32 11 10 10 'P2' 33 11 10 10 / -- WPIMULT -- 'P1' 0.5

K2 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN'

SHUT 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1*

TAB 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667

FACT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1*

DIA 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /

201

-- 'P2' 0.5 / REPRESENTS UNDAMAGED WELL SINCE SYSTEM IS 1/2 OF WHOLE -- / WFRICTN -- WELL DIAM ROUGH FLOW SCALE 'P1' 0.41667 0.1100 1 / -- I J K Tlen1 Tlen2 Dirn RangeEnd Diam 22 11 1 1* 1* 'I' 59 1* / / WFRICTN 'P2' 0.41667 0.1100 1 / 22 11 10 1* 1* 'I' 33 1* / / -- PRODUCTION WELL CONTROLS - DRAINAGE RATE IS VARIED AT FIXED OIL rATE WCONPROD -1 2 3 4 5 6 7 8 9 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES BHP -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* / 'P2' 'OPEN' 'LRAT' 3* 0 1* 1* / / -SPECIFY WELL WECON -- WELL MIN -- NAME OIL -RATE 'P1' 1* -- SPECIFY REPORT / TSTEP 24*15.2 TSTEP 24*15.2 ECONOMIC LIMIT AS 98 PERCENT WATERCUT MIN MAX MAX MAX WKVR GAS WATER GOR GWR PRCDR RATE CUT 1* 0.98 1* 1* 'WELL' AT ONE YEAR INTERVAL FOR XX YEARS / / 6 7 8 GAS LIQUID RES RATE RATE RATE 3* 3000 1* 1* 3* 0 1* 1* 9 BHP / / END RUN 'Y' /

WCONPROD -1 2 3 4 5 -- WELL OPEN/ CNTL OIL WATER -- NAME SHUT MODE RATE RATE 'P1' 'OPEN' 'LRAT' 'P2' 'OPEN' 'LRAT' /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR -NAME OIL GAS WATER GOR GWR PRCDR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 /

END RUN

202

TSTEP 24*15.2 TSTEP 24*15.2 TSTEP 24*15.2 / 5 6 OIL WATER RATE RATE 'LRAT' 'LRAT' 7 8 9 GAS LIQUID RES RATE RATE RATE 3* 3000 1* 1* 3* 0 1* 1* WCONPROD -- 1 2 3 4 -- WELL OPEN/ CNTL -- NAME SHUT MODE 'P1' 'OPEN' 'P2' 'OPEN' / / /

BHP / /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR -- NAME OIL GAS WATER GOR GWR PRCDR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / WCONPROD -1 2 3 4 5 6 7 8 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* 'P2' 'OPEN' 'LRAT' 3* 0 1* 1* /

END RUN

9 BHP / /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR END -- NAME OIL GAS WATER GOR GWR PRCDR RUN -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / WCONPROD -1 2 3 4 5 6 7 8 9 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES BHP -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* /

203

'P2' /

'OPEN'

'LRAT'

3*

1*

1*

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX -- NAME OIL GAS WATER GOR GWR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / END

WKVR PRCDR

END RUN

204

APPENDIX D DATA DECK FOR TAIL-PIPE WATER SINK (TWS) CONCEPT FOR WATER CRESTING CONTROL
-- ============================================================== -- DONWHOLE WATER SINK TECHONOLOGY SIMULATION RUN ON ECLIPSE 200 -- SIMULATION OF HORIZONTAL WELL APPLICATION OF DWS - JAN. 2002 -- ============================================================== RUNSPEC TITLE DWS OIL/WATER CONING STUDIES ON HORIZONTAL WELL (TAIL PIPE STUDY) -- SOLOMON O INIKORI DIMENS -- NX NY NZ 75 20 12 / NONNC OIL WATER -- gas -- DISGAS FIELD WELLDIMS 4 150 1 4 / START 1 'JAN' 2002 / FRICTION 2 1 / NSTACK 100 / MESSAGES 8* 5000 2*200 / UNIFOUT GRID -====================================================================== DXV 40 DYV 60 EQUALS 'DZ' 'DZ' 'PERMX' 'PERMZ' 'PERMX' 'PERMZ' 'PORO' 'PORO' 'PORO' 'TOPS' / 5 30 1 1 75 75 1 1 1 1 1 1 1 1 1 1 75 75 75 75 75 75 75 75 20 20 1 1 1 1 1 1 1 1 1 12 20 20 20 20 20 20 20 20 11 12 1 1 9 9 1 9 12 1 / / TOTAL 85' 8 8 12 12 8 11 12 1 / / / / / / / STEADY STATE / 60 60 60 60 10*60 60 60 60 60 60 / total 1200' 40 40 40 40 65*40 40 40 40 40 40 / total 3000'

8000 800 8000 800 0.31 0.35 1e10 6154

205

COPY 'PERMX' 'PERMY' / / INIT GRIDFILE 1 / PROPS ===================================================================== -NO RELATIVE PERMEABILITY DATA SUPPLIED. USED EMPIRICAL CORRELATION DATA SWOF -- Sw Krw Kro Pc -0.10 0.00 0.98 0 0.20 0.00 0.93 0 0.30 0.05 0.60 0 0.40 0.10 0.35 0 0.55 0.18 0.18 0 0.70 0.35 0.06 0 0.85 0.60 0.00 0 1.00 0.99 0.00 0 / PVTW -WATER PVT DATA (REF. PRESSURE, WATER FM VOL. FACTOR, VISCOSITY -COMPRESSIBILITY, ETC -Pref Bw(ref) Cw Visw Viscosibility 2500 1.02 3.0e-6 0.50 0 / PVCDO -OIL PVT DATA FOR DEAD OIL WITH CONSTANT SOLUTION GAS -Pref Bo Co Viso Viscosibility 2500 1.15 1.5e-5 1.34 0 / RSCONST -DATA FOR THIS SECTION ALSO NOT SUPPLIED. -Rs Pb 0.379 1000 / ROCK -DATA ALSO NOT SUPPLIED. -Pref Cf 2500 4.0e-6 / DENSITY -OIL WATER GAS 50.0 64.2 0.0702 /

SOLUTION ================================================================== -DATUM DATUM OWC OWC GOC GOC RSVD RVVD SOLN -DEPTH PRESS DEPTH PCOW DEPTH PCOG TABLE TABLE METH EQUIL 6154 2500 6194 0 1000 0 / RPTSOL 'RESTART=2' 'FIP=2' 'PRES' 'SWAT' /

206

SUMMARY ==================================== WOPR 'P1' 'P2' / WWPR 'P1' 'P2' / WGPR 'P1' 'P2' / WWCT 'P1' 'P2' / WOPT 'P1' 'P2' / WWPT 'P1' 'P2' / WBHP 'P1' 'P2' / FOPR FOPT FWPR FWPT FLPR FWCT FPR FOE / -RPTONLY -PRESENT DATA IN TABULAR FORM FOR GRAF PURPOSE IN PRINT FILE -REQUEST RUNSUM OUTPUT TO GO TO A SEPARATE RSM FILE RUNSUM SEPARATE SCHEDULE ======================================= -- SET MAXIMUM NUMBER OF LINEAR ITERATIONS IN THE NEWTON ITERATION TUNING / / LITMAX 2* 100 1* 50 / RPTSCHED 'RESTART=2' RPTRST 'BASIC=4' 'FIP=2' / PREF PHASE 'LIQ' 'LIQ' / / /

-- INTRODUCING THE WELLS -- WELL GROUP LOCATION BHP -- NAME NAME I J DEPTH WELSPECS 'P1' 'G' 20 11 6157 'P2' 'G' 20 11 6201 / -SPECIFYING COMPLETION DATA -WELL - LOCATIONOPEN/ -NAME I J K1 K2 SHUT COMPDAT

SAT TAB

CONN FACT

WELL DIA

207

'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P1' 'P2'

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 22

11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 1 11 10

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 10

'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN' 'OPEN'

1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1*

1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1*

0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667 0.41667

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1* 1*

'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X' 'X'

/ / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / /

/ -- WPIMULT -- 'P1' 0.5 / -- / WFRICTN -- WELL DIAM ROUGH FLOW SCALE 'P1' 0.41667 0.1100 1 / -- I J K Tlen1 Tlen2 Dirn RangeEnd Diam 22 11 1 1* 1* 'I' 59 1* / / -- PRODUCTION WELL CONTROLS - DRAINAGE RATE IS VARIED AT FIXED OIL rATE WCONPROD -1 2 3 4 5 6 7 8 9 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES BHP -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* /

208

'P2' /

'OPEN'

'LRAT'

3*

1*

1*

-SPECIFY WELL WECON -- WELL MIN -- NAME OIL -RATE 'P1' 1* -- SPECIFY REPORT / TSTEP 24*15.2 TSTEP 24*15.2

ECONOMIC LIMIT AS 98 PERCENT WATERCUT MIN MAX MAX MAX WKVR GAS WATER GOR GWR PRCDR RATE CUT 1* 0.98 1* 1* 'WELL' AT ONE YEAR INTERVAL FOR XX YEARS / / 6 7 8 GAS LIQUID RES RATE RATE RATE 3* 3000 1* 1* 3* 0 1* 1* 9 BHP / / END RUN 'Y' /

WCONPROD -1 2 3 4 5 -- WELL OPEN/ CNTL OIL WATER -- NAME SHUT MODE RATE RATE 'P1' 'OPEN' 'LRAT' 'P2' 'OPEN' 'LRAT' /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR -NAME OIL GAS WATER GOR GWR PRCDR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / WCONPROD -- 1 2 3 4 -- WELL OPEN/ CNTL -- NAME SHUT MODE 'P1' 'OPEN' 'P2' 'OPEN' / 5 6 OIL WATER RATE RATE 'LRAT' 'LRAT' 7 8 9 GAS LIQUID RES RATE RATE RATE 3* 3000 1* 1* 3* 0 1* 1*

END RUN

BHP / /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR

END

209

-- NAME OIL GAS WATER GOR GWR PRCDR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / WCONPROD -1 2 3 4 5 6 7 8 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* 'P2' 'OPEN' 'LRAT' 3* 0 1* 1* /

RUN

9 BHP / /

-SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX WKVR END -- NAME OIL GAS WATER GOR GWR PRCDR RUN -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP 24*15.2 / WCONPROD -1 2 3 4 5 6 7 8 9 -- WELL OPEN/ CNTL OIL WATER GAS LIQUID RES BHP -- NAME SHUT MODE RATE RATE RATE RATE RATE 'P1' 'OPEN' 'LRAT' 3* 3000 1* 1* / 'P2' 'OPEN' 'LRAT' 3* 0 1* 1* / / -SPECIFY WELL ECONOMIC LIMIT AS 98 PERCENT WATERCUT WECON -- WELL MIN MIN MAX MAX MAX -- NAME OIL GAS WATER GOR GWR -RATE RATE CUT 'P1' 1* 1* 0.98 1* 1* 'WELL' 'Y' / -- SPECIFY REPORT AT ONE YEAR INTERVAL FOR XX YEARS / TSTEP 24*15.2 / TSTEP 24*15.2 / TSTEP

WKVR PRCDR

END RUN

210

24*15.2 TSTEP 24*15.2 TSTEP 24*15.2 TSTEP 24*15.2 END

/ / / /

211

APPENDIX E - DWS WELL COST DATA SPREADSHEET

Louisiana State University (Petroleum Engineering Department)

Drilling Cost Estimate For

Downhole Water Sink Well (Vertical)

PROSPECT/FIELD NAME LEASE/WELL NO. FIELD COUNTY/STATE X DESCRIPTION OF WORK: DWS Vertical Well MOB/DEMOB DRYHOLE COMPLETION TOTAL Development Drilling $0 2,937,400 925,300 $3,862,700

8000 'MD 8000 'TVD 8000 'MD 8000 'TVD DRILL "S" SHAPED PILOT HOLE TO TEST GREEN & RED HORIZONS. SET SIDETRACK PLUG. DRILL HIGH ANGLE SIDETRACK TO GREEN HORIZON @ 85. COMPLETE W/ 7" PERFORATED LINER DAYWORK DIRECTIONAL SERVICES & SURVEYS MOBILIZATION/DEMOBILIZATION MUD MOTOR INSURANCE RIG MANAGEMENT COMPANY SUPERVISION - ONSITE QUARTERING AND CATERING POWER, FUEL, WATER & LUBE RIG SUPPLIES & SERVICES CEMENTING COSTS BITS, SCRAPER & REAMERS MUD & CHEMICALS TESTING, CORING & WIRELINE MUD LOGGING OPEN HOLE LOGGING MWD/LWD CASING CREW & TOOLS RENTAL EQUIPMENT PERFORATING GRAVEL PACK PUMPING CASED HOLE LOGGING TRANSPORTATION - AIR TRUCKING & HAULING TRANSPORTATION - MARINE CLOSED LOOP COMMUNICATIONS P&A EXPENSE ENVIRONMENT, REGULATORY & SAFETY MISCELLANEOUS (CONTINGENCIES) SHORE BASED FACILITIES CONTINGENCY TOTAL INTANGIBLES DAYRATE 35000 8500 8400 5000 3500 1000 4000 500 1000 28500 12500 1000 15000 12500 3000 1200 30000 10000 2750 300 9000 600 1000 1500 2500 1500 DAYS/QTY 20 8 0 0 20 20 20 20 20 20 4 20 20 5 0 15 20 5 0 20 20 25 20 20 20 20 30 DRILLING COMPLETION 700,000 280,000 100,000 70,000 28,000 20,000 8,000 80,000 32,000 10,000 4,000 20,000 8,000 114,000 250,000 20,000 75,000 45,000 24,000 9,600 150,000 30,000 55,000 22,000 6,000 2,400 225,000 72,000 12,000 4,800 20,000 8,000 30,000 12,000 50,000 20,000 45,000 12,000 2,121,000 552,800 TOTAL COST 980,000 100,000 98,000 28,000 112,000 14,000 28,000 114,000 250,000 20,000 75,000 45,000 33,600 150,000 30,000 77,000 8,400 297,000 16,800 28,000 42,000 70,000 57,000 2,673,800

8 8 8 8 8

0 0 8

3 8 8 8 8 8 8 8 8

212

DRIVE PIPE CONDUCTOR CASING SURFACE CASING INTERMEDIATE CASING & DRILLING LINER PRODUCTION LINER TUBING WELLHEAD EQUIPMENT FLOAT SHOE AND CENTRALIZERS SUBSURFACE EQUIPMENT LINER HANGER FLOWLINE AND HOOKUP MISCELLANEOUS EQUIP. (CONTINGENCIES) TOTAL TANGIBLES TOTAL COSTS

DESCRIPTIO DEPTH Unit Cost Length (ft) 30" x 1" wa 450 350 20" x 133 Ib 105 1500 13 3/8" x 68 Ib 45 4500 9 5/8" x 47 Ib 32 8000 7" 26 0 3.5 10 16000

157,500 157,500 202,500 256,000 25,000 15,000 2,900 816,400 2,937,400

160,000 65,000 18,000 125,000

4,500 372,500 925,300

157,500 157,500 202,500 256,000 160,000 90,000 33,000 125,000 7,400 1,188,900

3,862,700

TOTAL ESTIMATED COST FOR TWO VERTICAL DWS WELLS

$7,725,400

213

APPENDIX F SAMPLE DATA OF MODIFIED PIPE ROUGHNESS

Q (stb/d)

d -in

r w (ppg)

r o (ppg)

m w (cP)

m o (cP)

j (spf)

q i/Q

9000

3.92

8.59

6.67

0.5

1.34

0.0200

5000

3.92

8.59

6.67

0.5

1.34

0.0200

1000

3.92

8.59

6.67

0.5

1.34

0.0200

WC 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

f 0.01753 0.01760 0.01767 0.01772 0.01778 0.01740 0.01738 0.01735 0.01733 0.01731 0.01729 0.01778 0.01789 0.01800 0.01809 0.01818 0.01756 0.01752 0.01748 0.01744 0.01741 0.01737 0.01944 0.01979 0.02010 0.02037 0.02062 0.01871 0.01856 0.01842 0.01829 0.01815 0.01801

e (ft) 0.01563 0.01575 0.01586 0.01595 0.01605 0.01541 0.01537 0.01534 0.01530 0.01527 0.01523 0.01605 0.01624 0.01641 0.01657 0.01671 0.01568 0.01561 0.01555 0.01549 0.01540 0.01533 0.01949 0.01843 0.02002 0.02092 0.02161 0.01777 0.01753 0.01728 0.01703 0.01678 0.01652

Q (stb/d)

d -in

r w (ppg)

r o (ppg)

m w (cP)

m o (cP)

j (spf)

q i/Q

9000

3.92

8.59

6.67

0.5

1.34

0.0050

5000

3.92

8.59

6.67

0.5

1.34

0.0050

1000

3.92

8.59

6.67

0.5

1.34

0.0050

WC 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

f 0.00943 0.00950 0.00957 0.00963 0.00968 0.00930 0.00928 0.00926 0.00923 0.00921 0.00919 0.00968 0.00980 0.00990 0.00999 0.01008 0.00946 0.00942 0.00938 0.00935 0.00931 0.00928 0.01134 0.01169 0.01200 0.01228 0.01252 0.01061 0.01047 0.01033 0.01019 0.01005 0.00991

e (ft) 0.00312 0.00316 0.00320 0.00324 0.00328 0.00303 0.00302 0.00300 0.00299 0.00298 0.00297 0.00328 0.00335 0.00342 0.00349 0.00354 0.00313 0.00311 0.00308 0.00306 0.00300 0.00297 0.00535 0.00348 0.00514 0.00594 0.00650 0.00416 0.00405 0.00393 0.00381 0.00368 0.00355

214

Q (stb/d)

d -in

rw (ppg)

ro (ppg)

mw (cP)

mo (cP)

j (spf)

qi/Q

9000

3.92

8.59

6.67

0.5

1.34

0.0010

5000

3.92

8.59

6.67

0.5

1.34

0.0010

1000

3.92

8.59

6.67

0.5

1.34

0.0010

WC 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

f 0.00728 0.00734 0.00741 0.00747 0.00752 0.00714 0.00712 0.00710 0.00708 0.00705 0.00703 0.00752 0.00764 0.00774 0.00783 0.00792 0.00730 0.00726 0.00722 0.00719 0.00715 0.00712 0.00918 0.00953 0.00984 0.01012 0.01036 0.00845 0.00831 0.00817 0.00803 0.00789 0.00775

e (ft) 0.001283 0.001299 0.001315 0.001329 0.001343 0.001253 0.001247 0.001242 0.001238 0.001233 0.001229 0.001343 0.001372 0.001399 0.001425 0.001449 0.001289 0.00128 0.001271 0.001262 0.001207 0.001192 0.002866 0.00064 0.002339 0.003098 0.003608 0.001884 0.001834 0.001768 0.001694 0.001616 0.001536

215

VITA

Solomon Ovueferaye Inikori, Son of chief Johnson E. Inikori (Late) native of Eku, and Madam Regina Okpalefe, native of Ohrerhe - Agbarho, was born on March 27, 1964 in Warri, Delta State, Nigeria. He received his bachelor of engineering degree in mechanical engineering from the University of Nigeria, Nsukka in September 1988. In April 1993, he received his master of engineering degree in mechanical engineering from the University of Benin, Benin-City, Nigeria. He has eight years of full-time and over three years of part-time industry experiences in petroleum engineering practice with both independent and major oil companies. His job experiences range from drilling and completions engineering to reservoir engineering practice. He enrolled in the doctoral degree program in Petroleum Engineering at the Craft and Hawkins Department of Petroleum Engineering in August 1998. The doctoral degree will be conferred on him at the August 2002 commencement.

216