Pressure Transient Testing Vol 1

•
PRESSURE TRANSIENT TESTING:
DESIGN AND ANALYSIS
Vol. I
• By
James T. Smith
0
COPYRIGHT
By
James T. Smith
P.O. Box 1990
• Cody, Wyoming 82414
Telephone: (307) 527-6494
E-Mail: jtsmith@180com.net
ALL RIGHTS RESERVED
This book, or any part thereof, may not be reproduced

in any form without permission of the author.
^ 0603
•
TABLE OF CONTENTS
Page
INTRODUCTION
Types of Tests . . . . . . . . . . . . . . . . . . . . . . . . 1-1

Information Available from Pressure
Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2
Purpose and Scope . . . . . . . . . . . . . . . . . . . . . 1-2
Organization .......... ..... ... .. . ... . 1-3
Nomenclature . .. . . ... ... ..... . . .. . .. . 1-4
References . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
II. FUNDAMENTALS OF RESERVOIR FLUID FLOW
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Primary Flow Variables . . . . . . . . . . . . . . . . . . 2-1
Types of Fluids in the Reservoir . . . . . . . . . 2-2
Number of Mobile Fluids in the Reservoir ... 2-2
Reservoir Geometry . . . . . . . . . . . . . . . . . 2-3
Time Dependence of Flow . . . . . . . . . . . . . 2-7
Flow Equations . . . . . . . . . . . . . . . . . . . . . . . 2-13
Darcy's Law . . . . . . . . . . . . . . . . . . . . . 2-14
Steady State Flow . . . . . . . . . . . . . . . . . . 2-15
Transient Flow . . . . . . . . . . . . . . . . . . . . 2-18
Late Transient Flow . . . . . . . . . . . . . . . . . 2-44
Pseudosteady State Flow . . . . . . . . . . . . . . 2-44
Productivity Index . . . . . . . . . . . . . . . . . . 2-52
Constant Pressure Outer Boundary . . . . . . . . 2-54
Principle of Superposition . . . . . . . . . . . . . . . . . 2-54
Superposition in Space . . . . . . . . . . . . . . . 2-57
Superposition of Flow Rates . . . . . . . . . . . . 2-63
Flow Equations in Dimensionless Form . . . . . . . . . 2-66
Type Curve Matching . . . . . . . . . . . . . . . . . . . 2-75
Summary . . .. . . . . . .. . . . . .. . . . . . . . . . . 2-84
References .. ........ . . .... . .. .. . . . . . 2-85
Nomenclature .. .... . . ...... ... .. . .. . . 2-86
iii
•
Page
Summary of Major Equations . . . . . . . . . . . . . . . 2-89

Problems ........................... 2-91
III PRESSURE DRAWDOWN TEST
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
Analysis of Transient Flow Data . . . . . . . . . . . . . 3-4
Determination of Permeability . . . . . . . . . . . 3-5
Boundary Effects . . . . . . . . . . . . . . . . . . . 3-7
Analysis of Pseudosteady State Data . . . . . . . . . . . 3-12
Reservoir Limits Test . . . . . . . . . . . . . . . . 3-12
Estimation of Drainage Shape . . . . . . . . . . . 3-14
Radius of Drainage . . . . . . . . . . . . . . . . . 3-18
Stabilization Time . . . . . . . . . . . . . . . . . . 3-20
Practical Limitations of Reservoir
Limits Test . . . . . . . . . . . . . . . . . . . . . . 3-21
Evaluation of Formation Damage . . . . . . . . . . . . 3-22
Modifications of Flow Equations for a
Skin Zone . . . . . . . . . . . . . . . . . . . . . . . 3-24
Effective Wellbore Radius 3-27
Effect of Skin on the Semilog Plot ........ 3-28
Calculation of the Total Skin Factor ....... 3-30
Calculation of the Damage Skin Factor ..... 3-33
Flow Efficiency . . . . . . . . . . . . . . . . . . . . 3-43
Damage Ratio and Damage Factor . . . . . . . . 3-45
Flow Improvement . . . . . . . . . . . . . . . . . . 3-45
Wellbore Storage . . . . . . . . . . . . . . . . . . . . . . 3-47
Causes of Wellbore Storage . . . . . . . . . . . . 3-47
Effect of Wellbore Storage on the Semilog
Plot 3-51
Effect of Wellbore Storage on Formation
Face Rate . . . ................. 3-52
Detection of Wellbore Storage . . . . . . . . . . . 3-56
Duration of Wellbore Storage . . . . . . . . . . . 3-59
Changing Wellbore Storage . . . . . . . . . . . . 3-68
Summary . . . . . . . . . . . . . . . . . . . . . . . . 3-69
iv
•
Page
Multiphase Flow . . . . . . . . . . . . . . . . . . . . . . . 3-75

Type Curve Analysis . . . . . . . . . . . . . . . . . . . . 3-77
Ramey Type Curve . . . . . . . . . . . . . . . . . 3-77
McKinley Type Curve . . . . . . . . . . . . . . . . 3-88
Gringarten Type Curve . . . . . . . . . . . . . . . 3-96
Derivative Type Curve . . . . . . . . . . . . . . . 3-104
Test Design . . . . . . . . . . . . . . . . . . . . . . . . . . 3-125
Summary ..................... ...... 3-126
References ....... . .......... . . . . . .. . 3-128
Nomenclature .............. .. . . . . . ... 3-133
Summary of Major Equations . . . . . . . . . ...... 3-136
Problems . . . . . . . . . . . . . . . . . . . . . ...... 3-141
IV. MULTIPLE RATE TESTING

^
Applications .. ..... .. ... .. ... . . . . . .. . 4-1
Method of Superposition . . . . . . . . . . . . . . . . . . 4-2
Two-Rate Test . . . . . . . . . . . . . . . . . . . . . . . . 4-3
General Multi-Rate Analysis . . . . . . . . . . . . . . . 4-13
References . . . ..... ... .. . ..... . . . . ... 4-26
Nomenclature .................. .. ... . 4-27
Problems .. ..... .. ........... . .. . .. . 4-30
V. PRESSURE BUILDUP TEST
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Horner Method of Analysis . . . . . . . . . . . . . . . . 5-5
Ideal Versus Actual Pressure Buildup
Behavior . . . . . . . . . . . . . . . . . . . . . . . . 5-8
Estimation of Average Pressure . . . . . . . . . . 5-13
Evaluation of Formation Damage . . . . . . . . . 5-29
Determination of tp . . . . . . . . . . . . . . . . . . 5-34
Wellbore Storage . . . . . . . . . . . . . . . . . . . 5-35
v
•
Page
End of Semilog Straight Line . . . . . . . . . . . 5-42

Detection of Faults and Other Flow
Barriers . . . . . . . . . . . . . . . . . . . . . . . . 5-53
Well Not Centered in its Drainage Area ..... 5-61
Effect of Producing Time on Buildup
Behavior . . . . . . . . . . . . . . . . . . . . . . . . 5-68
Multiphase Flow . . . . . . . . . . . . . . . . . . . 5-71
Effect of Rate Variation Prior to Shut-in .... 5-75
Miller-Dyes-Hutchinson Method of Analysis ...... 5-81
Evaluation of Formation Damage . . . . . . . . . 5-83
Average Pressure . . . . . . . . . . . . . . . . . . . 5-84
End of Semilog Straight Line . . . . . . . . . . . 5-85
Horner Versus Miller-Dyes-Hutchinson . . . . . 5-86
Other Applications . . . . . . . . . . . . . . . . . . 5-87
Extended Muskat Analysis . . . . . . . . . . . . . . . . . 5-94
Selection of Correct Straight Line . . . . . . . . . 5-97
Agarwal Method . . . . . . . . . . . . . . . . . . . . . . . 5-103
Total Skin Factor . . . . . . . . . . . . . . . . . . . 5-105
Type Curve Analysis . . . . . . . . . . . . . . . . . . . . 5-112
Ramey Type Curve . . . . . . . . . . . . . . . . . 5-112
McKinley Type Curve . . . . . . . . . . . . . . . . 5-114
Other Type Curves . . . . . . . . . . . . . . . . . . 5-114
Type Curves Versus Conventional Analysis ... 5-114
Other Problems of Interpretation . . . . . . . . . . . . . 5-115
Wellbore Phase Separation . . . . . . . . . . . . . 5-115
Interference from Adjacent Wells . . . . . . . . . 5-117
Stratification . . . . . . . . . . . . . . . . . . . . . . 5-118
Composite Reservoirs . . . . . . . . . . . . . . . . 5-120
Fractured Reservoirs . . . . . . . . . . . . . . . . . 5-120
vi
•
Page
Test Design . . . . . . . . . . . . . . . . . . . . . . . . . . .5-123

References .................... ...... 5-125
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . 5-130
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-137
VI. PRESSURE TRANSIENT BEHAVIOR IN

FRACTURED RESERVOIRS
Types of Fractured Systems . . . . . . . . . . . . . . . . 6-2

Infinite Conductivity, Single plane, Vertical
Fractures ........................... 6-3
Flow Regimes Associated with Fractures .... 6-4
Conventional Analysis of Pressure Data ..... 6-13
Permeability and Fracture Length ..... 6-13
Beginning of Pseudoradial Straight
Line . . . . . . . . . . . . . . . . . . . . . . . 6-17
Short Producing Times . . . . . . . . . . . . 6-19
Formation Damage . . . . . . . . . . . . . . 6-19
W el lbore Storage . . . . . . . . . . . . . . . 6-22
Average Reservoir Pressure . . . . . . . . . 6-24
Type Curve Analysis . . . . . . . . . . . . . . . . 6-32
Permeability and Fracture Length ..... 6-32
Limitations of Type Curve Analysis .... 6-34
Type Curve Versus Conventional
Analysis . . . . . . . . . . . . . . . . . . . . . 6-35
Finite Conductivity, Single Plane, Vertical
Fractures .......................... 6-39
Uniform Flux Fractures . . . . . . . . . . . . . . . 6-40
Finite Conductivity Fractures . . . . . . . . . . . 6-40
Naturally Fractured Reservoirs . . . . . . . . . . . . . . 6-44
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-46
References . . . . . . . . . . . . . . . . . . . . . . . . . . 6-49
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . 6-52
Problems .. ........... .. . .... . . . .... 6-55
vii
•
Page
LBF. BILINEAR FLOW - LIQUID RESERVOIRS
Flow Periods For A Vertically Fractured Well ... . LBF-1

Fracture Linear Flow . . . . . . . . . . . . . . . . LBF-1
Bilinear Flow . . . . . . . . . . . . . . . . . . . . . LBF-1
Formation Linear Flow . . . . . . . . . . . . . . . LBF-3
Pseudoradial Flow . . . . . . . . . . . . . . . . . . LBF-3
Bilinear Flow Equations . . . . . . . . . . . . . . . . . . LBF-3
Constant Formation Face Rate . . . . . . . . . . . LBF-3
Constant Formation Face Pressure . . . . . . . . LBF-4
Bilinear Flow Graphs . . . . . . . . . . . . . . . . . . . . LBF-5
Constant Formation Face Pressure . . . . . . . . LBF-7
End of Bilinear Flow . . . . . . . . . . . . . . . . . . . . LBF-9
Constant Formation Face Pressure . . . . . . . .
Analysis of Bilinear Flow Data . . . . . . . . . . . . . .
LBF-9
LBF-11 •
Liquid - Constant Rate . . . . . . . . . . . . . . . LBF- 11
Liquid - Constant Pressure . . . . . . . . . . . . . LBF-13
Effect of Flow Restrictions . . . . . . . . . . . . . . . . LBF-14
Effect of Wellbore Storage . . . . ....... . . . LBF-16
References ......................... . LBF- 18
NF. INTRODUCTION TO PRESSURE

TRANSIENT TESTING IN NATURALLY
FRACTURED RESERVOIRS
Warren-Root Model . . . . . . . . . . . . . . . . . . . . . NF-1

Idealized Model of Naturally Fractured
Reservoir . . . . . . . . . . . . . . . . . . . . . . . . NF-1
Assumptions in Warren-Root Model . . . . . . . NF-1
Primary Parameters . . . . . . . . . . . . . . . . . NF-2
Ideal Semilog Behavior . . . . . . . . . . . . . . . NF-3
Conventional Semilog Analysis . . . . . . . . . . NF-4
Type Curve Analysis . . . . . . . . . . . . . . . . . . .
Double Porosity Type Curves . . . . . . . . . . .
. NF-4
NF-4 •
viii
Page
Type Curve Matching . . . . . . . . . . . . . . . . NF-7

Type Curve Analysis . . . . . . . . . . . . . . . . NF-9
Type Curve Analysis When Total System
Behavior is Not Observed . . . . . . . . . . . . . NF-10
Analysis With Pressure Derivatives . . . . . . . . NF-11
Fissured Versus Multilayered Reservoirs .... NF-13
Example Problem . . . . . . . . . . . . . . . . . . . . . . NF-14
References ........ .. . ........ .... ... NF-24
VII. GAS WELL TESTING
Introduction . . . . . . . . . . . . . . . . . . . . . . . . .. 7-1
Pressure Drawdown Test . . . . . . . . . . . . . . . .. 7-3
p2-Method . . . . . . . . . . . . . . . . . . . . . .. 7-3
Limitations . . . . . . . . . . . . . . . . . .. 7-4
Permeability Determination . . . . . . . .. 7-5
Skin Factor . . . . . . . . . . . . . . . . . .. 7-6
Evaluation of Turbulence . . . . . . . . .. 7-6
Flow Efficiency . . . . . . . . . . . . . . .. 7-8
Pseudosteady State ............ .. 7-9
p-Method . . . . . . . . . . . . . . . ...... . .. 7-13
Limitations . . . . . . . . . . ...... . .. 7-13
Permeability Determination ...... . .. 7-13
Skin Factor . . . . . . . . . . ...... . .. 7-14
Flow Efficiency . . . . . . . . . . . . . . . . 7-14
Real Gas Pseudo-Pressure, m(p) . . . . . . . .. 7-14
Evaluation of m(p) . . . . . . . . . . . . .. 7-16
Pressure Drawdown Equation . . . . . . .. 7-22
Permeability Determination . . . . . . . .. 7-22
Skin Factor . . . . . . . . . . . . . . . . . .. 7-23
Flow Efficiency . . . . . . . . . . . . . . .. 7-24
Pseudosteady State Flow . . . . . . . . . .. 7-24
Advantages of m(p) Method ...... .. 7-24
Pressure Buildup Test ................. .. 7-29
p2-Method . . . . . . . . . . . . . . . . . . . . . . . 7-29
Horner Analysis . . . . . . . . . . . . . . . . 7-30
ix
•
Page
Permeability . . . . . . . . . . . . . . . 7-30
Skin Factor . . . . . . . . . . . . . . . 7-30
Average Pressure .......... . 7-31
Flow Efficiency ........... . 7-32
Miller-Dyes-Hutchinson Plot . . . . . . . . 7-38
Agarwal Method .............. . 7-39
p-Method . . . . . . . . . . . . . . . . . . . . . . . . 7-39
Horner Analysis . . . . . . . . . . . . . . . . 7-39
Permeability . . . . . . . . . . . . . . . 7-40
Skin Factor . . . . . . . . . . . . . . . 7-40
Average Pressure . . . . . . . . . . . 7-40
Flow Efficiency . . . . . . . . . . . . 7-40
Other Methods of Analysis . . . . . . . . . 7-40
Real Gas Pseudopressure, m(p), Method ... . 7-41
Horner Analysis . . . . . . . . . . . . . . . . 7-41
Permeability . . . . . . . . . . . . . . . 7-41
Skin Factor . . . . . . . . . . . . . . . 7-41
Flow Efficiency . . . . . . . . . . . . 7-42
Average Pressure . . . . . . . . . . . 7-42
Other Methods of Analysis . . . . . . . . . 7-42
Problems of Interpretation . . . . . . . . . . . . . . . . . 7-50
Bounday Effects . . . . . . . . . . . . . . . . . . . 7-53
Short Producing Time . . . . . . . . . . . . . .. . 7-53
Changing Gas Properties . . . . . . . . . . . .. . 7-53
Type Curve Analysis . . . . . . . . . . . . . . . . . .. . 7-55
Fractured Gas Wells . . . . . . . . . . . . . . . . . .. . 7-56
Conventional Method of Analysis . . . . . . . . . 7-57
Type Curve Analysis . . . . . . . . . . . . . . . . 7-59 .
Gas Well Deliverability Testing . . . . . . . . . . . . . 7-67
Flow-After-Flow Tests . . . . . . . . . . . . . . . 7-68
Empirical Method of Analysis . . . . . . . 7-70
Theoretical Method of Analysis . .... . 7-85
Summary . . . . . . . . . . . . . . . . . . . . 7-89
Isochronal Tests . . . . . . . . . . . . . . . . . . . 7-89
Empirical Method of Analysis . . . . . . . 7-93
x
•
Page
Theoretical Method of Analysis ..... . 7-93

Modified Isochronal Test . . . . . . . . . . . . . . 7-95
Analysis of Data When Pseudosteady
State Flow is Not Attained . . . . . . . . . . . . . 7-99
Use of Real Gas Pseudopressure, m(p),
in Gas Well Deliverability Analysis ...... . 7-101
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-102
References . . . . . . . . . . . . . . . . . . . . . . . . . . 7-103
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . 7-106
Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-114
GBF. BILINEAR FLOW - GAS RESERVOIRS
Bilinear Flow Equations . . . . . . . . . . . . . . . . . . GBF-1

^ Constant Formation Face Rate . . . . . . . . . . . GBF-1
Constant Formation Face Pressure . . . . . . . . GBF-2
Bilinear Flow Graphs . . . . . . . . . . . . . . . . . . . . GBF-3
Constant Formation Face Rate . . . . . . . . . . . GBF-3
End of Bilinear Flow . . . . . . . . . . . . . . . . . . . . GBF-6
Constant Formation Face Rate . . . . . . . . . . . GBF-6
Analysis of Bilinear Flow Data . . . . . . . . . . . . . . GBF-10
Gas - Constant Rate . . . . . . . . . . . . . . . . . GBF-10
Gas - Constant Pressure . . . . . . . . . . . . . . . GBF- 11
References . . . . . . . . . . . . . . . . . . . . . . . . . . GBF-13
VIII. INJECTION WELL TESTING
Pressure Falloff Test in Unit-Mobility Ratio

Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
Permeability . . . . . . . . . . . . . . . . . . . . . . 8-6
Skin Factor . . . . . . . . . . . . . . . . . . . . . . 8-6
Flow Efficiency . . . . . . . . . . . . . . . . . . . . 8-7
Miller-Dyes-Hutchinson Method . . . . . . . . . 8-7
xi
•
Page
Average Reservoir Pressure . . . . . . . . . . . . 8-8

Pressure Falloff Analysis in Non-Unit Mobility
Ratio Systems . . . . . . . . . . . . . . . . . . . . . . . . 8-14
Two-Bank System . . . . . . . . . . . . . . . . . . 8-15
Three-Bank System . . . . . . . . . . . . . . . . . 8-17
Pressure Falloff Analysis in Fractured
Reservoirs . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-24
Injectivity Test . . . . . . . . . . . . . . . . . . . . . . . . 8-33
Step Rate Testing . . . . . . . . . . . . . . . . . . . . . . 8-36
Introduction . . . . . . . . . . . . . . . . . . . . . . 8-36
Test procedure and Analysis . . . . . . . . . . . . 8-37
Pressure versus Rate Plot . . . . . . . . . . 8-37
Multirate Analysis . . . . . . . . . . . . . . 8-45
Factors Affecting SRT Analysis . . . . . . . . . . 8-68
Time Step Size . . . . . . . . . . . . . . . . 8-68
Rate Increment . . . . . . . . . . . . . . . . 8-74
Wel lbore Storage . . . . . . . . . . . . . . . 8-75
Changing Wellbore Storage . . . . . . . . . 8-80
Skin Effects . . . . . . . . . . . . . . . . . . 8-84
Selection of p; for SRT Analysis ...... 8-85
Step Rate Test on a Fractured Well ....... . 8-92
Two-Step Rate Test . . . . . . . . . . . . . . . . . 8-93
Analysis Methods . . . . . . . . . . . . . . . 8-94
Design Considerations . . . . . . . . . . . . 8-101
References . . . . . . . . . . . . . . . . . . . . . . . . . . 8-104
Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . 8-106
Problems ......................... . 8-110
IX. INTERFERENCE AND PULSE TESTING
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1
Interference Test . . . . . . . . . . . . . . . . . . . . . . . 9-4
Homogeneous Isotropic Reservoir . . . . . . . . . 9-7
Single Active Well . . . . . . . . . . . . . . 9-7
xii
Page
Multiple Active Wells . . . . . . . . . . . . 9-16

Homogeneous Anisotropic Reservoir . . . . . . . 9-24
Heterogeneous Anisotropic Reservoirs . . . . . . 9-36
Pulse Testing . . . . . . . . . . . . . . . . . . . . . . . . . 9-36
Introduction . . . . . . . . . . . . . . . . . . . . . . 9-36
Theoretical Background . . . . . . . . . . . . . . . 9-38
Homogeneous Isotropic Reservoirs . . . . . . . . 9-41
Test Design - Graphical Method . . . . . . 9-54
Test Design - Analytical Method . . . . . . 9-57
Test Analysis - Graphical Method .... . 9-60
Test Analysis - Analytical Method .... . 9-62
Homogeneous Anisotropic Reservoirs . . . . . . 9-64
Heterogeneous Anisotropic Reservoirs . . . . . . 9-65
Vertical Pulse Testing . . . . . . . . . . . . . . . . 9-65
Summary . . . . . . . . . . . . . . . . . . . . . . . . 9-66
References ......................... . 9-67
Nomenclature ....... .... . . .. .. . . . . . . . 9-70
X. DRILLSTEM TESTING
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1
Equipment and Procedures . . . . . . . . . . . . . . . . . 10-3
Analysis of Pressure Data . . . . . . . . . . . . . . . . . 10-11
Pressure Buildup Data . . . . . . . . . . . . . . . . 10-11
Flow-Period Data . . . . . . . . . . . . . . . . . . 10-14
Wireline Formation Tests . . . . . . . . . . . . . . ... 10-15
Interpretation of DST Pressure Charts . . . . . . . ... 10-16
References .. . .... .. ..... ... .. . . . . . .. 10-20.
Nomenclature ...... . .. .. . .... .. . . . . .. 10-22
Summary of Major Equations . . . . . . . . . . . . ... 10-23
APP. APPENDIX A: FLUID AND ROCK PROPERTY

CORRELATIONS
0
xiii
MAJOR REFERENCES: Pressure Transient Testing
0
1. Matthews, C.S. and Russell, D.G.: Pressure Buildup and Flow Tests
in Wells, Society of Petroleum Engineers, Dallas: Monograph
Series, Vol. I (1967).
2. Theory and Practice of the Testing of Gas Wells (third edition),

Energy Resources Conservation Board, Calgary, Alberta, Canada
(1975).
3. Earlougher, R.C., Jr.: Advances in Well Test Analysis, Society of

Petroleum Engineers, Dallas: Monograph Series, Vol. V (1977).
4. Ramey, H.J., Jr., Kumar, A., and Gulati, M.S.: Gas Well Test
Analysis Under Water Drive Conditions, American Gas Association,
Arlington, Virginia (1977).
5. Dake, L.P.: Fundamentals of Reservoir Engineering, Elsevier

Scientific Publishing Company, Amsterdam (1978).
6. Lee, W.J.: Well Testing, Society of Petroleum Engineers, Dallas:

Textbook Series, Vol. I (1982).
7. Slider, H.C.: Worldwide Practical Petroleum Reservoir Engineering

Methods, PenWell Publishing Company, Tulsa, Oklahoma (1983).
8. Streltsova, T.D.: Well Testing in Heterogeneous Formations, John

Wiley and Sons, New York ( 1988).
9. Stanislav, J.F. and Kabir, C.S.: Pressure Transient Analysis,

Prentice Hall, Englewood Cliffs, New Jersey (1990).
0
10. Sabet, M.A.: Well Test Analysis, Gulf Publishing Company, Houston
(1991). 40
11. Raghavan, R.: Well Test Analysis, Prentice Hall, Englewood Cliffs,
New Jersey (1993).
12. Dake, H.P.: The Practice of Reservoir Engineering, Elsevier,

Amsterdam (1994).
13. Horne, R.N.: Modern Well Test Analysis: A Computer Aided

Approach, Second Edition, Petroway, Inc., Palo Alto, California
(1995).
14. Lee, J. and Wattenbarger, R.A.: Gas Reservoir Engineering, Society

of Petroleum Engineers, Dallas: Textbook Series, Vol. 5 (1996).
0
•
Chapter 1
INTRODUCTION
One of the most important functions of the petroleum

engineer is to obtain reliable information about in situ
reservoir conditions. Detailed reservoir information is
essential to the reservoir engineer for the purpose of
analyzing and predicting the behavior of various types of
reservoirs during both primary and enhanced recovery
operations. Further, the production engineer must diagnose
the condition of production and injection wells in order to
optimize reservoir performance. The drilling engineer must
often determine if a hydrocarbon-bearing formation has
sufficient flow capacity to justify setting pipe. One of
the most accessible, economical and reliable sources of
information necessary to these engineering functions is the
pressure transient test.
• I. TYPES OF TESTS
A pressure transient test requires that the flow

rate of a well be changed, and that the resulting pressure
response of the reservoir be measured. The test can take
many forms depending upon the nature of the rate change, the
number of wells involved, and whether the test wells are
injectors or producers. The pressure transient tests most
commonly used in the petroleum industry are:
,Pn,e/,iJ.e^ c7Ginê,,_i • m.mîoh, P-Ussu.u- e^C^v^cae^- wL& te.n,v
jj^,e .Iu_S,u.Ft r-*- a. ^ vw na7b, - - l,i.-(JCv ^;v,,c&kaq,e,

a,-L, o(A.cuoua.e. (-Kdjc dA,ij &^^Wkk8. C-M)
•
F •
•
Pressure buildup
Pressure drawdown
• 11^1^- ^ ^,,^,t
• Multi-rate
• Falloff Ia^^ rn ^^• I^SDS(^^ ^.
• Injectivity
• Step -rate
• Interference
^ C^LuCcQ ^ct/J t,v^t
^
• Pulse a1lotên- w^bcv^.e^(S^
• Drill stem
L
II. INFORMATION AVAILABLE FROM PRESSURE TRANSIENT TESTS
It has long been recognized that the pressure behavior

of a reservoir following a rate change directly reflects
the geometry and flow properties of the reservoir, as well
as the properties of reservoir fluids. If a pressure transient
test is properly designed, run, and interpreted; it is possible
•
to completely describe a reservoir and its contained fluids.
Information available from a well test includes:'
• Effective permeability
• Detection and magnitude of formation
damage and/or stimulation
• Detection and approximate distances to
flow barriers and fluid contacts
• Volumetric average reservoir pressure
• Drainage pore volume
• Communication between wells
• Detection, length, and capacity of fractures # I
• Interwell rock properties, and directional

nature of properties
III. PURPOSE AND SCOPE
The purpose of this manual is to present those methods

•
of pressure transient analysis which are being used most
1-2
^tr M tvcu, e&" J4 .x-atJ, wto.c, -.- a,- a,,L c¢. u/VL S
^
commonly and successfully to characterize a reservoir. Test
design, as well as test analysis, is discussed. It is not
possible to understand pressure analysis without some
theoretical considerations; however, this manual stresses
the practical aspects of pressure analysis and design. All
methods of analysis are illustrated with the use of practical
examples.
This manual does not attempt to take the reader to the
frontiers of knowledge in well testing. Instead, those
methods of pressure analysis which, in the author's opinion,
have been most successfully applied are presented. References
are presented at the end of each chapter to assist the
engineer who needs to further pursue a particular subject.
Other publications which give excellent coverage of this
subject are the Society of Petroleum Engineers' Monographs
on well testing' '2, and Alberta Energy Resources and Conser-
vation Board's gas well testing manual3.
IV. ORGANIZATION
This manual is written so that the reader can develop

a logical and orderly understanding of the theory and
procedures necessary to design, conduct and analyze pressure
transient tests. Material is not necessarily presented in
the order of importance; instead, the manual is organized
so that the reader can develop a logical understanding of
pressure transient analysis by starting in Chapter 1 and
proceeding chronologically through the remaining chapters.
Within this framework, each chapter is, to the greatest
possible extent, written as an independent unit.
It is not possible to understand pressure transient
testing without first understanding the basic laws of fluid
flow which describe the relationship between reservoir
pressure, flow rate, and reservoir rock and fluid properties.
Chapter 2 presents the fundamentals of fluid flow which are
1-3
applied throughout the remainder of the book.
C:
Chapter 3 covers pressure drawdown testing which,
theoretically, is the most simple pressure transient test.
This test is not discussed first because it is the most
important test; instead, we cover it first because it
follows more directly and logically from the flow theory
in Chapter 2 and, therefore, is easier to understand than
other tests. Further, it is the only singZe-rate production
test that we run; all other tests we will discuss are multi-
rate tests and, basically, are variations of the drawdown
test. This is probably the most important chapter in the
manual. Most of the concepts used in pressure transient
testing are introduced in this chapter.
With the exception of the pressure drawdown test, all
pressure transient tests involve multiple rates. Chapter 4
will present a general approach to multirate testing.
Chapter 5 discusses the most commonly used test, the pressure
buildup test. Fractured reservoirs are becoming increasingly
important and an entire chapter, Chapter 6, is devoted to
the analysis of tests conducted in fractured reservoirs.
Discussion of well testing in Chapters 3-6 is restricted
to liquid systems. Chapter 7 shows how liquid theory can be
applied with minor modifications to the analysis of tests
from gas reservoirs. Gas deliverability testing is also
treated in this section.
Chapter 8 presents pressure testing techniques in
injection wells, a subject which is already important because
of waterflooding, but which will gain even more significance
with future emphasis on enhanced oil recovery. Chapter 9
discusses methods of well testing when more than one well
is involved in the test.
V. NOMENCLATURE AND UNITS
When possible, the standard symbols ado p ted b y the

Society of Petroleum Engineers of AIME''-6 are used in this
1-4
•
REFERENCES
Matthews, C.S. and Russell, D. G.: Pressure Buildup

and Flow Tests in Wells, Monograph Series, SPE, Dallas
1967T 1 . -
Earlougher, R. C., Jr.: Advances in Well Test Analysis,

Monograph Series, SPE, Dallas 1 77T S.
Theory and Practice of the Testing of Gas Wells, third

edition, Pub. ECRB-75-34, Energy Resources and-Conser-
vation Board, Calgary, Alberta (1975).
"Letter Symbols for Petroleum Reservoir Engineering,

Natural Gas Engineering, and Well Logging Quantities",
Society of Petroleum Engineers of AIME, Dallas (1965).
"Supplements to Letter Symbols and Computer Symbols for

Petroleum Reservoir Engineering, Natural Gas Engineering,
and Well Logging Quantities", Society of Petroleum
Engineers of AIME, Dallas (1972).
. "Supplements to Letter Symbols and Computer Symbols

for Petroleum Reservoir Engineering, Natural Gas
Engineering, and Well Logging Quantities", Society
of Petroleum Engineers of AIME, Dallas (1975).
1-6
•
Chapter 2
FUNDAMENTALS OF RESERVOIR FLUID FLOW
I. INTRODUCTION
A prerequisite to designing or analyzing a pressure

transient test is to have a computational model that
adequately describes the relationship between reservoir
pressure, flow rate, and reservoir rock and fluid properties.
This model may have many forms depending upon the geometry
of the reservoir, the number and types of fluids flowing, the
nature of pressure variation as a function of time, and the
flow rate behavior of production or injection wells. The
purpose of this chapter is to describe the variables which
affect pressure behavior and to present the equations which
serve as the basis of commonly used pressure transient tests.
II. PRIMARY FLOW VARIABLES
Petroleum reservoirs are very complex and the equations

needed to describe them will vary depending upon the
characteristics of the particular reservoir being studied.
The primary flow variables which must be considered when
developing or selecting an equation to analyze a pressure
transient test are:
• Types of fluids in the reservoir
^ • Number of mobile fluids in the reservoir
• Reservoir geometry
• Time dependence of flow
1. Types of fluids in the reservoir
•
The equations required to describe reservoir fluid flow
will vary depending upon the nature of fluids in the reservoir.
Fluids can be broadly categorized on the basis of their
L compressibility as:
A. Incompressible fluids
An incompressible fluid is one which experiences no
change in volume (or density) when pressure and/temperature
chang. Actually, there are no incompressible fluids; some
fluids, however, approximate this behavior and can be assumed
incompressible for certain types of engineering calculations.
The pressure behavior of a reservoir is very sensitive to
fluid compressibility; consequently, equations based on
incompressible flow are not adequate to design or analyze
pressure transient tests.
B. Slightly compressible fluids -- L i Q U I D 0

These are fluids which exhibit small changes in volume
with changes in pressure or temperature. All liquids fit
into this category although, for certain engineering appli-
cations, some liquids can be assumed to be incompressible.
C. Highly compressible fluids (;-AS

These are fluids which experience large changes in
volume as a function of pressure and temperature. All gases
fall into this category.
2. Number of mobile fluids in the reservoir

The equations required to analyze a pressure transient
test vary in form and complexity depending upon the number of
mobile fluids in the reservoir. Mathematical models are
generally referred to as single-phase, two-phase, three-
phase, etc., depending upon the number of fluids flowing.
The description of fluid flow and the subsequent analysis
2-2
0
of pressure data becomes more difficult as the number of
mobile fluids increases. A rigorous description of multi-
phase flow can generally be obtained only by the use of a
computer. Approximate solutions of multiphase problems •
which require only hand-calculations will be presented in
this manual.
3. Reservoir geometry
The shape of a reservoir has a significant effect on
its flow behavior. Most reservoirs have irregular boundaries
and a rigorous mathematical description of geometry is often
possible only with the use of computer simulators. Fortu-
nately, most pressure transient tests do not involve the
entire reservoir; instead, it is only necessary to describe
that part of the reservoir affected by the well(s) being
tested. This can usually be accomplished using simple
^ geometries. The flow geometries most useful in well test
work are:
0 A. Radial flow
In the absence of severe reservoir heterogeneities,
flow into or away from a wellbore will follow radial flow
lines for a substantial distance from the wellbore. Further,
the largest pressure gradients in a reservoir occur near
the wellbore where flow is radial. For these reasons, radial
flow is the geometry applicable to most pressure transient
work.
When in radial flow, reservoir fluids follow flow paths
which converge at the wellbore; moreover the cross-sectional
area to flow is a function of distance from the welibore.
The flow lines and isopotential lines for ideal radial flow
are illustrated by Fig. 2.1.
•
2-3
^y °rlT
^ ^,& •fourt^ ^ll s StahcJ
•
FLOW
LINES
PLAN
VIEW
ENT I AI,
ES
yv ^- WEI.LBORE
SIDE
VIEW •
• Fig. 2.1: Ideal radial flow into a welibore.
B. Linear flow
Linear flow occurs when flow lines are parallel and
flow is in a single direction. This can occur only if the
cross-sectional area to flow is constant. A common
application of linear flow equations is in the analysis of
pressure data from wells which have been hydraulically
fractured, or wells which intersect natural fractures. Wells
which fit this description exhibit linear flow immediately
following a rate change and the pressure data measured during
this time period must be analyzed using linear flow equations.
Figure 2.2 illustrates linear flow into a vertical hydraulic
fracture.
q
2-4
^ The analysis of pressure data from fractured reservoirs
will be discussed in detail in Chapter 6.
FRACTURE
---j --
ISOMETRIC
VIEW
WELLBORE
h .^!, • .•_., .. . ••'. .. •

;.• _.•. •... . ...
FLOW LINES
PLAN
VIEW WELL
FRACTURE
Fig. 2.2: Example of ideal linear flow

into a vertical hydraulic fracture.
C. Elliptical flow
Following linear flow in a fractured system, lines of
constant pressure become elliptical in shape. This is
illustrated by Fig. 2.3 which depicts a well producing from
a reservoir with a vertical fracture. The mathematical
equations required to rigorously describe elliptical flow
are difficult to handle. Fortunately, as will be shown in
Chapter 6, methods have been developed to analyze fractured
• reservoirs which do not require solution of these equations.
2-5
F
WELL
•
FLOj^
LINES
ISOPOTENTIAL
LINES
Fig. 2.3: Elliptical flow into a vertical hydraulic

fracture.
D. Spherical or hemispherical flow

Depending upon the type of welibore completion
configuration, it is possible to have spherical or hemi-
spherical flow near the wellbore. For example, a well with
a limited perforated interval could result in spherical
flow in the vicinity of the perforations. This is illustrated
by Fig. 2.4. A well which only partially penetrates the
producing formation, as depicted by Fig. 2.5, could result in
hemispherical flow. In both of these situations, flow becomes
radial as the pressure disturbance moves away from the wellbore
into the reservoir. While spherical and hemispherical
equations are useful in studying near-welibore effects, such
as the efficiency of perforations, it will be shown that
most well tests can be analyzed using less complicated models.
• 1
•
2-6
• WELL
SIDE
VIEW
Fig. 2.4: Spherical flow due to limited entry.
WELL
0 FLOW
SIDE LINES
VIEW
Fig. 2.5: Hemispherical flow in a partially

penetrating well.
4. Time dependence of flow

The behavior of a reservoir, or an individual well, is
dependent upon the length of time it has been producing.
Reservoirs are classified as steady state, transient, late
transient, or pseudosteady state,*depending on how pressures
within the reservoir change as a function of time.* Whereas
• a particular equation may be sufficient to describe flow

behavior at one time in the life of a reservoir, a different
2-7
equation will be necessary at another time. It is essential
that the well test analyst be able to recognize each of these
flow regimes in order to apply the correct equation or
analysis technique at the proper time. Failure to do so can
result in serious errors.
As noted previously, flow into or away from most wells
is adequately described by a radial flow model. Accordingly,
to aid our understanding of the effect of time on reservoir
behavior, consider a well centered in a circular reservoir of
radius re. It is assumed that this reservoir is homogeneous
and has a constant thickness. Further, the reservoir is
assumed to produce at constant rate into the well. This
system is illustrated in Fig. 2.6.
•
PLAN
VIEW
SIDE
: . : h VIEW
^ •• •, ..^^. • ^ -.^• =^^^' . ^

!
Fig. 2.6: Ideal radial flow model.
2-8
-(AAa,t VanA.&LPA4 d-zkeA.wu.n.c,
dUstcu6ocvti.c._, m o-w^ ?
• 4 vw^; a_6teQ 6jt^ *
A. Ste ady state flow PM (/L.tT'tuu 'L^ ^Ô&

Steady state flow requires that pressure be constant
with respect to time at every location in the flow system.
For this condition to exist, fluids produced from a reser'voir
^must be replaced at the same rate from an external source.
Very few reservoirs exist which can be described as
^ ^ steady state. Some reservoirs in which fluid production is
balanced by water influx from a connecting aquifer, or fluid
influx from injection wells, approximate this behavior. Since
the basic element of pressure transient testing is the
/-measurement of pressure versus time, it is obvious that
steady state models have limited application to pressure
transient analysis. pALC'9_1_^ ► ^^ uri{^, pa^-^^^
bust nvt ur^- ^-^
B. Transient flow _i, ^05t iMVQLfqMJ -}^ t^lla-
Consider the reservoir in Fig. 2.6. If this well were
opened to flow at a constant rate, a pressure disturbance
(transient) would be created at the formation face. The
pressure at the formation face would drop instantaneously
r^ • as the well was opened, and the pressure disturbance would
ZS ' ,
move away from the wellbore at a rate determined by the
formation diffusivity, n, defined as
tf'^
(2.1)
n ûct "t" -to 1;)'haur
1"zh^w 4 -^- ct
-V- where: k = formation permeability
= formation porosity
u= viscosity of reservoir fluid
ct = total compressibility of reservoir Croc^lt ^^l^.uo^^
It is important to observe that the producing rate of the]ti ^^J-wtf'

well has no effect on the velocity of the pressure 4c^S hnî
disturbance. Within the assumption that the reservoir is q^SS(Utit^ L^
homogeneous, the pressure disturbance would move radially
64-Lt
away from the wellbore.
MdVè'p-,
2-9
The reservoir pressure distribution caused by this
•
flowing well, when plotted as a function of time, will
appear as illustrated by Fig. 2.7. It is observed at time
tl that the pressure disturbance has moved a distance r 1
into the reservoir. At radii greater than rl, the reservoir
pressure remains equal to the initial reservoir pressure, pi..
It is important to note that the reservoir boundary (r=re)
is having no effect on the pressure behavior of the reservoir
at time tl. In fact, the pressure distribution in the
reservoir at t 1 is exactly the same as it would be if the
reservoir were infinite in size.
t=0 rl r2
Pi
tl •
t2
t3
n t4
y
t5
rw e
RadlLis
Fig. 2.7: Relationship between pressure distribution

and time in a closed cylindrical reservoir.
2-10
â-
"nc^, encc^ ,t,ô côû^
^ , hG^(Gttt;f,,p/1 GtM^C ^ ^-
These same comments can be made at time t2 when the
pressure disturbance has advanced to a radius r2; there is
still no effect of the boundary on the reservoir pressure
distribution, or the behavior of the producing well.
At time t3, the pressure disturbance reaches the
boundary. This causes the reservoir pressure behavior to
change. Until this time, the reservoir behaved as if it
were infinite in size. Beyond this time, it will behave
like a closed container.
These observations lead us to a definition of transient
flow: Transient flow is that time period during which the
boundary has no effect on pressure behavior in the reservoir.
For the pressure behavior depicted by Fig. 2.7, transient
flow occurs during the time interval o<t<t3. Equations
which can be used to predict the length of the transient
flow period will be presented in Chapter 3.
C. Late transient flow

When the pressure disturbance reaches the first
boundary, transient flow ends and a very complex flow behavior,
called late transient flow, is initiated. The pressure
behavior during late transient flow, as well as the length of
this flow period, depend upon the shape of the boundary and
the location of the well relative to the boundary. For a
well centered in a symmetrical geometry, such as the cylindrical
u v reservoir depicted in Fig. 2.6, the late transient period is
short; this is because the pressure disturbance simultaneously
reaches all boundaries and the pressure quickly stabilizes.
When the well is not centered in the drainage area, or when
• the drainage area is irregular in shape, this period of flow
can last a long time.
Because of the complexity of late transient flow, it
is very difficult to mathematically describe flow behavior
during this period of time. Fortunately, this period is
^ generally short compared to other types of flow. No effort
will be made to analyze pressure data measured during late
2-11
transient flow. It will be very important, however, to •
recognize late transient flow so that data measured during
this period are not incorrectly analyzed; methods to
recognize late transient flow will be presented in Chapter 3.
D. Pseudosteady state flow c5err.4-5;4eady Sfafe.)

After the reservoir recovers from the unstabilized
conditions of late transient flow, pseudosteady state flow
begins. Sometimes referred to a<quasisteady, semisteady,^
and incorrectly as steady state, this is a period of flow
when the pressures at all locations are changing at the same
rate as a linear function of time. This behavior is
illustrated by Fig. 2.8. yj '
%ugltS ^^^^, ÔÎJw^ ^fjaunoQahccS
pil- - - t=0 - - -
(PR)l^^ - - - - - - ^^
-+ A p1
•
Apl tl Apl
-- - -----î -------- p2
(PR) 2
t2
Ap2
I
^
_
t
s
_T
Ap1
^
^D2
01
r r
w e
Radius
q
Fig. 2.8: Relationship of pressure distribution
to time during pseudosteady state flow.
2-12
•^ p^ sû ru Cfiâô^c s bC.a. SoJvri-C dUIWA.^4 Prv"c,v^bve atve w-E^6,
Consider the pseudosteady state pressure distributions

at t1 and t2. During the time interval (t2 - t1), the
pressure changed by an amount Opl at every location in the
reservoir, including the formation face. Further, the
volumetric average reservoir pressure, pR, also changed by
Apl during this same time interval.
The volumetric average reservoir pressure, pR, is a
very important number in reservoir engineering calculations.
It is the pressure at which a reservoir would stabilize if
all wells were shut in and the pressures throughout the
reservoir were given time to equalize. The average pressure
in a closed system is related to the amount of material in
the system; accordingly, the amount of fluid in an oil or
gas reservoir at any time is related to the magnitude of PR
at that time. Further, the amount of fluid produced from a
formation during a given time interval is related to the
change in PR during that time.
^ It is not easy to determine PR' The most direct
method.would be to shut in all wells and measure PR after
pressures equalize throughout the reservoir. This is not
practical in most reservoirs, however, because of the
excessive shut-in time required and the accompanying loss
of production. Therefore, it is important to note from
Fig. 2.8 that the formation face pressure changes during
pseudosteady state at the same rate as average pressure;
consequently, changes in average reservoir pressure can be
determined by observing corresponding changes in the welibore
pressure. The importance of this observation will become
clear in subsequent discussions.
III. FLOW EQUATIONS
Depending upon the combination of variables present,

^ the equations required to describe flow in a reservoir can
take many forms. The purpose of this section is to present
2-13
those equations which will be useful in the design and analysis
of pressure transient tests. By combining the conservation of
mass equation with a transport equation ( Darcy's Law) and
various equations of state, the necessary flow equations can
be developed. Since all flow equations to be considered
depend upon Darcy's Law, it is important to consider this
transport relationship first.
1. Darcy's Law
This is an empirical relationship which shows that the
volumetric flow rate of a fluid through a porous medium is
proportional to the pressure gradient. For a horizontal
linear system, this relationship is '
u = g = - u ^ (2.2)
where: q = volumetric flow rate

k = effective permeability
A = cross-sectional area to flow
u = fluid viscosity
3p/3x = pressure gradient in direction of flow
u = apparent velocity (flux), flow rate
per unit cross sectional area.
For a horizontal radial system, this equation has the form
u = ar = k (,P) (2.3)
r
Ar u arr
where: qr = volumetric flow rate at radius r
A = cross sectional area to flow at radius r

r
(ap/ar)r = pressure gradient at radius r
ur = apparent velocity at radius r.
2-14
If more than one fluid is flowing in a formation, the flow
rate of a specific fluid can be computed from Eqs. 2.2 or
2.3 by substituting the appropriate effective permeability
and fluid viscosity.
It is very important to note that Darcy's law only
applies to Zaminar flow. When turbulent flow exists,
application of Darcy's law can result in serious errors.
Modifications for turbulent flow will be discussed in subse-
quent sections.
OLr
2. Steady state flow
The equations for steady state flow are based on
Darcy's Law. These equations are derived in numerous
publications and will be presented here only in their final
form. These equations have no application to pressure
transient testing; however, it is important that the reader
recognize these commonly published equations and not misuse
them in well test analysis.
A. Linear systems
A horizontal linear reservoir is illustrated by Fig.
2.9. Applications to reservoirs containing single-phase
incompressible and highly compressible fluids are considered.
P?
• ql
Fig. 2.9: Linear flow model.
2-15
•
a. Incompressible flow
ql = q2 = 0.001127 kA (p p) (2•4)
ULB 1 2
q = flow rate, STB/day

k = permeability, md
A = area, sq ft
u = viscosity, cp
L = length, ft
p = pressure, psia
B = formation volume factor, RB/STB
b. Highly compressible flok

Whereas inlet and outlet flow rates are equal for
incompressible fluids, the volumetric flow rate of a com-
pressible fluid is a function of pressure and will be
different at the inlet and outlet. Expressed in terms of
the pressure p2 at the outlet,
•
q2 = 0.003164 pkAL (p12-p22), cu ft/D (2.5)
If q2 is converted to standard conditions using the non-

ideal gas law,
kATsc(p1Z-p22)
a2sc = 0.003164 Tz uL , scf/D (2.6)
p sc 2
T = temperature at standard conditions,

sc
oR (°F+460)
psc = pressure at standard conditions,
U MGu^' trauaS' psia
T = reservoir temperature, °R
z2 = gas deviation factor at P21
T (Figs. A.1 and A.2)
Q
3 . , Q,^ hc,nr^
L_^, p7, 't ma"
2-16
^ B. Radial s y stems
Consider the radial reservoir in Fig. 2.6 which is
drained by a well at its center:
a. Incompressible fluid ', Pe-

)
p -P
q = 0.00708 hk e rwf STB/D (2. 7)
in e
r
w
pe = pressure at outer boundary, psia

pwf = flowing formation face pressure,
psia
me re = radius of reservoir, ft
• nc J^ r' ^- ^âE rw = wel lbore radius, ft
fU,^^lO^m-Ê (Wa-'^^`- Q ^
Another useful and equivalent form of this equation can be
written if the pressure drawdown is expressed in terms of
^ PR rather than pe , i . e., C P^ ho^f' 9uac1L4, îtinoLtSri UI^4,^0^' )
^ ^
hk pR pwf (2.8)
q= 0.00708
r
ln re - 0.5
w
PR = volumetric average reservoir pressure,

psia.
b. Highly compressible fluid

The gas flow rate expressed at conditions of reservoir
temperature and flowing formation face pressure is
z _ z
q= 0.01988 hk (pe Pwf), cu ft/D. (2.9)
g re
upwf ln r
w
If the formation face producing rate is expressed at standard

conditions of temperature and pressure, i.e., at Tsc and psc'
2-17
N ate, " ftme 6actu'r- -IA'- SS LItp,
^ .
La cczj.v'.^.
fl Vri2 .
C^Van-,O p"a- (^r^^-en
•
hkTsc(pe pwf) , scf/D (2.10)
agsc = 0.01988
re
upscTzwln r
w
3. Transient flow
Transient flow begins when the flow rate of a well is
changed; it ends when the pressure disturbance created by
the rate change hits a boundary.
The period of transient flow is very important in
pressure transient analysis. Every well tested experiences
transient flow and it is this flow period which typically
yields the greatest amount of information about a reservoir.
Consequently, this text will deal primarily with the
analysis of pressure data recorded during the transient
flow period.
A. Radial flow ^
When a pressure transient test is conducted in a well,
fluid flow near the wellbore, where pressure gradients are
largest, will have the greatest effect on test results. In
the absence of heterogeneities or discontinuities, flow near
the wellbore is essentially radial.
For purposes-of mathematical development, consider the
ideal cylindrical model depicted by Fig. 2.6. It is assumed
that the producing well has an^ôpen-hole completion, that it
fully penetrates the formation, and that the well is perpen-
dicular to the horizontal formation. It is also assumed that
no flow occurs in the vertical direction or the 0(circular)
* I
direction; thus, all flow is radial and can be described
using a small cylindrical segment of the reservoir which has
an inner radius, r, and an outer radius, r + Ar. This
reservoir segment is shown in Fig. 2.10 with fluid flowing
in the positive r- direction.
•
2-18
• N
^ (ou )
p_ r r+Ar
r
e
•
Fig. 2.10: Cylindrical segment of reservoir used
to derive transient flow equations.
For a small interval of time, At, a mass balance can

be written to describe flow into and out of the element
shown:
Rate of mass Rate of mass flow Rate of mass

accumulation in into the element _ flow out of
the element - during the time the element
during the time interval, At during the time
interval, At interval, At
assuming that only one fluid of density p is flowing,
R ate of massl = 2Trrh ( u^^ 2-'q) =4-
• into elementj r 6
^
2-19
[Rate of mass 1 = [2Tr(r+Ar)h] (pu )
out of elementf r r+Ar
•
Rate of mass 2^rorh[(^p)t+ot-(^p)ti
[accumulation] At
These terms substituted into the mass balance give
27rrArh[W)t+At- W)t]
= 27rrh(pur)r
- [27r (r+Ar) h] (Pur) r+Ar
Divide both sides of this equation by 21rrOrh to obtain
W) t+At - (OP) t (Pur) r+Or - (Pur) r

At Or
•
- r (Pu)
rr+Ar
Take the limit of both sides of this equation as Ar and At

approach zero to obtain
at (4^P) ° - ar (Pur) - r (Pur) r
or,
-rPur) a t (^DP) • (2. 11)

r ar (
Equation 2.11, which is simply a statement of-mas.s

conservation, is referred to as the continuity equation.
It is valid for any fluid flowing radially through any
medium. Before this equation can be used for practical
engineering calculations it must be combined with a •
2-20
• transport equation (Darcy's Law) and appropriate equations
of state.
If flow is restricted to a porous medium, the flux,
ur, can be described by Darcy's Law which was presented
previously as Eq. 2.3, i.e.,
k ^ (2.12)
ur
u ar
Whereas a positive sign was used in Eq. 2.3, a negative sign

is used here since flow in this model is away from the
wellbore. With this substitution for ur, Eq. 2.11 becomes
(2.1J)
r ar [r P-s ^ at (^P) •
It should be noted again that Darcy's equation and,
• consequently, Eq. 2.13 contain the assumption that flow

is laminar.
Equation 2.13 is equally valid for gases or -liquids.
However, to develop practical equations which can be used
to analyze pressure tests, these two cases much be
handled separately.. Slightly compressible fluids will be
considered first.
a. Slightly compressible fluids (liquids) ^- k,^; r^,

Q l^,vt-S-^.y
Q
Suppose the reservoir is saturated with a slightly
compressible liquid; if the reservoir is assumed to be
isothermal, changes in fluid volume (or density) can only
occur as a result of changes in pressure. The effect of
pressure on fluid volume under constant temperature
conditions is expressed by the isothermal coefficient of
compressibility.
The Lisothermal coefficient of compressibilit}of a
• liquid is defined as
?-?1
u.v" t15 ' c -- ^p SL
âfl^G^n ^ a.e ' c,^^
c^l crn, pte s S-^ bt
z (2.14)
c=- 1 av
vap G^ C (^ ^^ I^w U S
UL^ trU, C(.RC(

where v is the specific volume of the 1'^ quid. Since
specific volume is the reciprocal of density,
C = - P aP ( p
and
C = p aP (2.15)
Liquid density is, in general, a function of temperature

and pressure. However, since a reservoir is an isothermal
system, density in this situation is a function only of
pressure; thus, the partial derivative in Eq. 2.15 can be
written as an ordinary derivative, i.e.,
^
C = p p (2.16)
If c is assumed to be constant over the range of pressures

considered, Eq. 2.16 can be integrated as
P P
c dp = ( (2.17)
P
J -.
po po
The lower limit is an arbitrary reference pressure, po,

and the corresponding density, po, whereas the upper limit
is the pressure at which it is desired to obtain the
density. Integrating, Eq. 2.17 simplifies to
p = p ec(P-Po) (2.18)
0
•
2- 22
• The equation of state for density in Eq. 2.18 can be
substituted into Eq. 2.13 to obtain
1 a p ec(p-po)r Pêc(P-Po) (2.19)

r ar o aat o
r
Two additional assumptions are made at this point. First,

it is assumed that the"formation permeability is constant.
Second, it is assumed that thAiquid viscosity is not
dependent upon pressure; this is a good assumption for most
reservoir liquids for the range of pressures encountered
during a pressure transient test. With these assumptions,
Eq. 2.19 reduces to
1 a
r ar Le c (p-po) r ap ar k at
(P
[ eco) 1 .
• This expression can be differentiated and simplified to

obtain
2
1 + cr (^P) Z = k ^c ^P + a , (2.20)
r
1 3^ + r
ar2
(D 6"a^ w /t^m-t d-un,

Consider now the change in porosity due to formation
compressibility. Formation compressibility is generally
defined as the change in the pore volume of the rock per
unit pore volume per unit change in pressure. Mathematically,
cf = Vp
1 dp
(2.21)
where Vp is the initial pore volume of the reservoir.

Since Vn is the product of porosity, ^, and bulk volume, Vb,
•
2-23
1
cf = ^Vb
d(^V )
dp
b
=
1 do
dp
^ (2.22)
•
0
P
Using the chain rule of differentiation,
a^ __a! a = ap (2.23)
t ap at ^c f at '
With substitution of Eq. 2.23, Eq. 2.20 can be written as
a?P + 1 ^ + c ( D^P) 2 = ku c ^P + 1 ap ^P
r
or, + ( aP) 2= ku (c + c f) aI
a?P2 + r D r t
2-^p + 1 aP +,C
or,
ar2 r ar (^) 2 = k^ct ^ (2.24)
•
Ô Côi. t,cû^dS^
where ct = C + c f.
Equation 2.24 is a nonlinear partial differential

equation which des-cribes the pressure at any radius, r, at
any time, t, in a reservoir containing a liquid of constant
compressibility. Because of its nonlinearity, this equation
is very difficult to solve. However, if it is assumed that
the pressure gradient in the reservoir is small, the second-
order term can be neglected, and Eq. 2.24 reduces to a
. 1
linear equation which can be solved analytically, i.e.,
2 ^ ;^ c t ;p
+ 1 aP= (2.25)
a.. r 3r k 3t '
3r,
When written in customar y "oilfield units" , E q .

the form
2.25 has
•
2-24
• ?p } 1 3P W t aE ^ (2.26)
r ar 0.000264c at
ar2
-A
where: k = permeability, md ^SS^.vreS ^

^^SeYVa-z^r
r = radial position, ft • ore -F(uticQ i^n
p = pressure, psia
ct= total compressibility, psi
t = time, hrs
^ = porosity, fraction
u = viscosity, cp
When the reservoir contains more than one fluid,

total compressibility should be computed as
ct = c o S o + c S + c g S g + cf (2.27)
w w
V o I u rr\e.-tY i G 0..v Gr aq Z oP -^(^w dl S ^^r, r e Su- vo^ r

where c o , c w an d c g re 'f er to t h e compress ibilit y o f o il ,
water and gas, respectively, and S o , S w and S g refer to
the fractional saturation of these fluids. Please note
that the introduction of c t into Eq. 2.26 does not make
Eq. 2.26 applicable to multiphase flow; the use of ct, as
defined by Eq. 2.27, simply accounts for the compressibility
of any i mmobiZe fluids which may be in the reservoir with
the fluid that is flowing. SK11 (.LSSumeS oh(Y o ne- rno ô<<e -^'I(Jicl
-4( Equation 2.26 is referred to as the radial
diffusivity equation. It is one of the most important
equations in petroleum engineering. In particular, it is
the basis for most pressure transient tests. LAcknowledging
its importance, it is instructive to summarize the
assumptions and limitations to which it is subject:
1) horizontal, radial flow
2) laminar flow
3) isothermal flow
4) constant formation thickness - w/fî►^nî (^ ^^` ^4' ^^tu'4^ I
5) single phase flow
0 6) constant compressibility fluid
2-25
GLSûw^ tmS E t^, 2-26
7), constant fluid viscosity

8) constant permeability
9) small pressure gradients
10) open hole completion
11) fully penetrating well
12) well drilled perpendicular to formation
These assumptions appear to be very restrictive; however,

it will be shown that useful solutions can be obtained
from the diffusivity equation.
Solutions of the Radial Diffusivity Equation

We are interested in obtaining a solution for the
transient flow of a slightly compressible fluid in the ideal
reservoir depicted by Fig. 2.6 and described in the preceding
section. Before presenting the transient flow solution,
however, it is instructive to first consider the general
solution for the case where the reservoir is bounded with a
no-flow boundary.
(i) General solution for a bounded reservoir - Solution of

Eq. 2.26 requires that two boundary conditions and one
initial condition be specified. A useful solution is
obtained if it is assumed that ( 1) the formation produces
at a constant rate into the wellbore; ( 2)r^herP is no flow
across the outer boundary of the reservoir and (3)ltthe
reservoir is at a uniform pressure, pi, when production
begins. With these conditions, the pressure at any location
in the reservoir can be determinedl'2 as a function of
producing time according to the relationship
•
2-26
• p(r,t) = pi - 141.2 k qBll Z 2
r2
4+ tD)
reD - 1
reD in rD (3reD - 4reD in reD - 2r^D - 1)

2 - 1
reD 4(reD - 1) 2
00
ean2 tDJ2 reD) IJl (an) Y o (anreD)-Y 1 (an)1 0 (anri
+ Tr (2.28)
an [J (ctflrD) - Ji (an) J
n=1
where an are the roots of
(2.29)
J1(anreD)Y1(an) - Jl((xn)Y1((xnreD) = 0
• and:
rD
_
r
r
w
e
reD r
w
t = 0.0002637kt
D t
ûcrw
J1 = Bessel function of first kind of order one
Y1 = Bessel function of second kind of order two
p(r,t) = pressure at radius r and time t, psia
pi = initial reservoir pressure, psia
r = radial distance from producing well to
location where pressure is being computed, ft
re = radius of outer reservoir boundary, ft
t = length of time well has been producing, hrs
q = formation producing rate, STB/D
k = effective permeability of formation to
produced fluid, md
h = formation thickness, ft
• cp
u
= formation porosity, fraction
= viscosity, cp
e = base of natural logarithm
2-27
q
If only the pressure at the well is considered, i.e.,
r= rw and p(r,t) = p(rW) t) = p`ti,f, Eq. 2.28 reduces to:
u 2tD + l 3
pwf = Pi - 141.2 kh r2 n r eD 4
eD
°D 2
+ 2 ^ Dj2 (a nreD) (2. 30)

L_ân Jl (anreD) Ji (an
_
n-1 2^ ^
It is important to recognize that within the assumptions

and boundary conditions previously stated, Eqs. 2.28 and 2.30
represent exact solutions. These solutions apply both before
the pressure disturbance reaches the boundary (transient flow),
and after the boundary begins to influence the data (late
transient and pseudosteady state flow). Fortunately, it will
•
not be necessary for us to use these equations; instead, we
will show that approximate solutions are available which are
easier to use and which give sufficient accuracy for purposes
of well test analysis.
(ii) Infinite reservoir with line-source well - An adequate

approximation to Eq. 2.28 for the transient flow time regime
can be developed if it is assumed that (1) the reservoir is
infinite in size (recall that during transient flow, a
reservoir behaves as if it were infinite in size); (2) the
well.bore radius is zero; (3) the formation is producing at
constant rate and (4) the reservoir is at a uniform pressure,
pi, when production begins. With these assumptions, the
solution to Eq. 2.26 isz
•
2-28
AP Cr,{,)
Bu - 948ûctr2
p(r,t) = pi + 70.6 kh Ei (2.31)
kt
^ -u
where: Ei(-x) = - u du (2. 32)
B. = exponential-integral. function.
This solution, commonly referred to as the line

source, exponential-integral, or Theis3 solution, is not
an exact solution. A comparison of this solution to the
exact solution is shown in Fig. 2.11". It is observed
from this comparison that the exponential-integral solution
is essentially equal to the exact solution when tD/rn > 25.
When expressed in terms of producing time, this condition
means that t (hrs) > 9:48x104ûctr2/k. If tD/rD < 25 the
^ accuracy with which the exponential-integral solution
approximates the exact solution depends upon the values of
rD and tD. For example, when rD > 20, the exponential-
integral solution is accurate within one percent for
tD/rD > O.S. A more detailed analysis of the comparison
between the exponential-integral and exact solutions is
given by Mueller and Witherspoon". Practically, Eq. 2.31
is sufficiently accurate for most applications of pressure
transient analysis during the transient flow period.
Before Eq. 2.31 can be used, a logical question to ask
is: Given a number x, how do you evaluate Ei(-x)? It was
shown earlier that Ei(-x) is defined by Eq. 2.32; fortunately,
• numerical solutions of Eq. 2.32 already exist and it will not
be necessary to evaluate this integral. When x<0.01,
Ei(-x) can be approximated with less than 0.25 percent error
by:
Ei(-x) = in (1.781x) (2.33)
2-29
•
r--
r-,
.r.,
aa
^ •
r-^1
Fig. 2.11:
tD/rD
Comparison of exponential-integral solution

•
to the exact solution for a well located in
an infinite system. After Mueller and
Witherspoon".
Recognizing from Eq. 2.31 that x=948ûctr2/kt, this means

that Eq. 2.33 can be used when t>9.48x104ûctr2/k. For
0.01<x<10.9, Table 2.15 can be used to evaluate Ei(-x).
When x>10.9, Ei(-x) can be considered zero for practical well
test work.
Example 2.1: Calculation of reservoir pressure during

transient flow using the exponential-integral
solution
Problem. An oil well is producing at a constant rate of 250

STB/D under transient flow conditions from a reservoir with
the following rock and fluid properties:
2-30
•
Table 2.1: Values of the exponential-integral5,-Ei(-x)
- Ei (- x), 0.000<0.209, i nterva l -0.001

x _ 0 1 2 3 4 5 6 7 8 9
0.00 + o0 6.332 5.639 5.235 4.948 4.726 4.545 4.392 4.259 4.142
0.01 4.038 3.944 3.858 3.779 3.705 3.637 3.574 3.514 3.458 3.405
0.02 3.355 3.307 3.261 3.218 3.176 3.137 3.098 3.062 3.026 2.992
0.03 2.959 2.927 2.897 2.867 2.838 2.810 2.783 2.756 2.731 2.706
0.04 2.681 2.658 2.634 2.612 2.590 2.568 2.547 2.527 2.507 2.487
0.05 2.468 2.449 2.431 2.413 2.395 2.377 2.360 2.344 2.327 2.311
0.06 2.295 2.279 2.264 2.249 2.235 2.220 2.206 2.192 2.178 2.164
0.07 2.151 2.138 2.125 2.112 2.099 2.087 2.074 2.062 2.050 2.039
0.08 2.027 2.015 2.004 1.993 1.982 1.971 1.960 1.950 1.939 1.929
0.09 1.919 1.909 1.899 1.889 1.879 1.869 1.860 1.850 1.841 1.832
0.10 1.823 1.814 1.805 1.796 1.788 1.779 1.770 1.762 1.754 1.745
0.11 1.737 1.729 1.721 1.713 1.705 1.697 1.689 1.682 1.674 1.667
0.12 1.660 1.652 1.645 1.638 1.631 1.623 1.616 1.609 1.603 1.596
0.13 1.589 1.582 1.576 1.569 1.562 1.556 1.549 1.543 1.537 1.530
0.14 1.524 1.518 1.512 1.506 1.500 1.494 1.488 1.482 1.476 1.470
0.15 1.464 1.459 1.453 1.447 1.442 1.436 1.431 1.425 1.420 1.415
0.16 1.409 1.404 1.399 1.393 1.388 1.383 1.378 1.373 1.368 1.363
0.17 1.358 1.353 1.348 1.343 1.338 1.333 1.329 1.324 1.319 1.314
0.18 1.310 1.305 1.301 1.296 1.291 1.287 1.282 1.278 1.274 1.269
0.19 1.265 1.261 1.256 1.252 1.248 1.243 1.239 1.235 1.231 1.227
0.20 1.223 1.219 1.215 1.210 1.206 1.202 1.198 1.195 1.191 1.187
^ -Ei( -x), 0.00<x<2.09, interval =0.01

0.0 + co 4.038 3.335 2.959 2.681 2.468 2.295 2.151 2.027 1.919
0.1 1.823 1.737 1.660 1.589 1.524 1.464 1.409 1.358 1.309 1.265
0.2 1.223 1.183 1.145 1.110 1.076 1.044 1.014 0.985 0.957 0.931
0.3 0.906 0.882 0.858 0.836 0.815 0.794 0.774 0.755 0.737 0.719
0.4 0.702 0.686 0.670 0.655 0.640 0.625 0.611 0.598 0.585 0.572
0.5 0.560 0.548 0.536 0.525 0.514 0.503 0.493 0.483 0.473 0.464
0.6 0.454 0.445 0.437 0.428 0.420 0.412 0.404 0.396 0.388 0.381
0.7 0.374 0.367 0.360 0.353 0.347 0.340 0.334 0.328 0.322 0.316
0.8 0.311 0.305 0.300 0.295 0.289 0.284 0.279 0.274 0.269 0.265
0.9 0.260 0.256 0.251 0.247 0.243 0.239 0.235 0.231 0.227 0.223
1.0 0.219 0.216 0.212 0.209 0.205 0.202 0.198 0.195 0.192 0.189
1.1 0.186 0.183 0.180 0.177 0.174 0.172 0.169 0.166 0.164 0.161
1.2 0.158 0.156 0.153 0.151 0.149 0.146 0.144 0.142 0.140 0.138
1.3 0.135 0.133 0.131 0.129 0.127 0.125 0.124 0.122 0.120 0.118
1.4 0.116 0.114 0.113 0.111 0.109 0.108 0.106 0.105 0.103 0.102
1.5 0.1000 0.0985 0.0971 0.0957 0.0943 0.0929 0.0915 0.0902 0.0889 0.0876
1.6 0.0863 0.0851 0.0838 0.0826 0.0814 0.0802 0.0791 0.0780 0.0768 0.0757
1.7 0.0747 0.0736 0.0725 0.0715 0.0705 0.0695 0.0685 0.0675 0.0666 0.0656
1.8 0.0647 0.0638 0.0629 0.0620 0.0612 0.0603 0.0595 0.0586 0.0578 0.0570
1.9 0.0562 0.0554 0.0546 0.0539 0.0531 0.0524 0.0517 0.0510 0.0503 0.0496
2.0 0.0489 0.0482 0.0476 0.0469 0.0463 0.0456 0.0450 0.0444 0.0438 0.0432
2.0<x< 10.9, interval- 0.1

0 1 2 3 4 5 6 7 8 9
x
2 4.89x102 4.26x10Z 3.72x102 3.25x102 2.84x10 z 2.49x102 2.19x10-2 1.92x102 1.69x10'2 1.48x10-2
3 1.30x10'2 1.15x10'2 1.01x/0-2 8.94x10-3 7.89x10-3 6.87x10"3 6.16x10'3 5.45x10-3 4.82x10-3 4.27x10-2
4 3.78x10'3 3.35x10-3 2.97x10-3 2.64X10'3 2.34X10-3 2.07x10-3 1.84X10'3 1.64X10'3 1.45X10-3 1.29x10-3
5 1.15x10'3 1.02x10-3 9.08x10-4 8.09x10-4 7.19x10-4 6.41x10'4 5.71x10-4 5.09x10-° 4.53x10-4 4.04x10'4
6 3.60x10'4 3.21x10-4 2.86x10-'° 2.55x10-° 2.28x10-4 2.03x10-4 1.82x10'4 1.62x10'4 1.45x10'4 1.29x10-4
7 1.15x10'4 1.03x10-4 9.22x10-5 8.24x10'5 7.36x10-5 6.58x10-5 5.89x10'5 5.26x10-5 4.71x10-5 4.21x10-5
8 3.77x10-5 3.37x10-5 3.02x10'5 2.70x10-5 2.42x10'5 2.16x10-5 1.94x10'5 1.73x10-5 1.55x10-5 1.39x10'5
9 1.24X10-5 1.11X10'5 9.99X10-6 8.95x10'6 8.02x10'6 7.18X10-6 6.44x10-6 5.77X10-6 5.17x10-6 4.64X10'6
10 4.15x10-6 3.73x10-6 3.34x10-6 3.00 x10-6 2.68x10-6 2.41x10-6 2.16x10-6 1.94x10-6 1.74x10-6 1.56x10-6
•
2-31
B o = 1.32 RB/STB r w = 0.25 ft
•
u o = 1 . 2 cp k o = 80 md
14% h = 11 ft
c t = 16x10psi-1 p i = 3600 psia
a) Calculate pressures at radii of 0.25, 1, 5, 10, 100, 500,

1000, 2000 and 3160 feet, for t = 48 hours. Plot the results
as (1) pressure versus logarithm of radius and (2) pressure
versus radius.
b) What percent of the total pressure drawdown from the

drainage radius to the wellbore occurs between the wellbore
and the following radii: 1, 5, 10 and 100 ft? What part of
the reservoir drained by a well has the greatest effect on
the flow behavior of that well?
c) Repeat part ( a) for t = 1 hr and t = 6 hrs and plot

the results as pressure versus logarithm of radius.
Solution. Pressure can be computed as a function of radius

and time using Eq. 2.31:
Bu 948ûctrz
p(r,t) = p + 70.6 ^ Ei
1 kt
p(r,t) = 3600 + (70.6) (250) (1.32) (1.2)

( 80 )( 11 )
(948) (0. 14) (1.2) (16x10 6)r2

Ei
80t
r 3.185x1 0-Sr2^
- t
p(r,t) = 3600,+ 31.77 Ei
L
a) A summary of calculations for t = 48 hrs is presented
in the following table. These results are also presented
graphically in Figs. 2.12 and 2.13.
•
2-32
•
r, ft x Ei (-x) p(r,48),psia
0.25 4.18x10-8 -16.433 3078

1 6.63x10' -13.649 3166
5 1.66x10-5 -10.429 3269
10 6.63x10-5 - 9.044 3313
100 6.63x103 - 4.439 3459
500 1.66x10- - 1.378 3556
1000 6.63x10 - 0.404 3587
2
2000 2.65 - 2.065x10 3599
3160 6.63 - 1.770x104 3600
b) The total pressure drawdown in the reservoir at t = 48

hours is:
(AP)total = 3600-3078 = 522 psi
q The drawdown from a radius of one foot to the wellbore is:
(AP)r=1 = 3166-3078 = 88 psi
Therefore,
88
% Drawdown = (100) = 16.9%
522
Calculations at other radii are summarized in the following

table:
r,ft p(r,t),psia % Drawdown
1 3166 16.9
5 3269 36.6
10 3313 45.0
100 3459 73.0
•
2-33
370!
360
3S0
340
ca
.H
U)
330
w
320
310
300
OoL Y w 1 10 102 103 10"
r, feet
Fig. 2.12: Pressure distribution as a function of time for the reservoir

in Ex. 2.1.
• • •
• • •
3700
3600
n t = 1 hr
3500
A t = 6 hrs
• t = 48 hrs
3400
.r.,Cd
^
3300
3200
3100
3000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
r, feet
Fig. 2.13 Pressure distribution after 48 hours for the reservoir in Ex. 2.1.
E^.(-X) = 2, 303 111^ 'e cl•-?8^ x)
It is obvious from these calculations that most of the q

pressure loss occurs close to the wellbore and, accordingly,
near-wellbore conditions will exert the greatest influence
on flow behavior. Observing from Fig. 2.12 that the drainage
radius at t=48 hrs exceeds 1000 feet, and that 73 percent of
the pressure drop occurs in the first 100 feet around the
well, we see that 73 percent of the drawdown occurs over a
distance that represents less than 10 percent of the total
reservoir radius affected by production.
c) A summary of pressure calculations at t = 1 hour and

t = 6 hours is presented in the following table:
r,ft p(r,l),psia p(r,6),psia
0.25 3201 3144

1 3289 3232
5 3392 3335 q
10 3436 3379
100 3573 3523
S00 3600 3596
1000 3600 3600
2000 3600 3600
3160 3600 3600
These data are presented graphically in Fig. 2.12.
(iii) Log-approximation of the Ei solution - It was shown

in the previous section that Eq. 2.33 is an adequate
approximation of Ei(-x) when x<0.01. With this approximation,
Eq. 2.31 reduces to:
kt - (2. 34)
. p(r,t) = pi - 162.6 khu log 3.23 0
uctr2
2-36
• Like the exponential-integral solution, this equation can be
used to compute pressure as a function of time at any location,
within the stated assumptions.
In most well test applications, we are primarily
concerned about pressure behavior at the formation face of
the producing well. For this important situation, r=rw,
p(r,t) = p(rw,t) = pwf, and Eq. 2.34 becomes
r p = p - 162.6 ^ ( log kt z - 3.231 ( 2.35)

wf i /
Cuctrw
where: pwf = flowing formation face pressure of the

producing well, psia.
Example 2.2: Calculation of flowing formation face pressure

using the log-approximation to the exponential-
integral solution
Problem. a) For the reservoir described in Ex. 2.1,

compare the values of flowingPtormation face pressure calcu-
lated using the exponential-integral solution and the log-
approximation of the exponential-integral solution at the
following times: tl hour, 10 hours, 1 day, 10 days, 100 days.
What conclusion does this suggest concerning the accuracy of
E q. 2.35? Ir - rM,
b) At what minimum value of producing time can Eq. 2.35

be used to calculate the formation face pressure of this
well? What conclusion does this suggest?
c) Plot the pressures computed in (a) versus logarithm

of time.
Solution. The exponential-integral equation, Eq. 2.31 can

be written as:
2-37
Bu ûctr2 ^
p(r,t) = pi + 70.6 k Ei
L -948
kt
p(0.25,t) = 3600
+ 70.6 (250)(1.32)(1.2 )Ei - (948) (0.14) (1.2) (16x10-6) (0.25)21

( 80 ) ( 11 ) 80t J
p(0.25,t) = 3600 + 31.77 Ei - 1.99x10

t
The log-approximation of the exponential-integral solution

was previously presented as Eq. 2.35, i.e.,
kt Z- 3.23 ^
pwf = P. - 162.6 k qBp
L log ^ uctrw J
pwf = 3600
0
162,6(250) (1.32) (1.2) log 80t - 3.23 ^
-
77
0(11) (0.14) (1 . 2) (16x10 6 ) (0.2 5) 2
1
pwf = 3600 - 73.17 ( log t + 5.45)
a) The following table presents a comparison of the pressures

computed using the two methods:
t,hrs p(rw,t) pwf
1 3201 3201
10 3128 3128
24 3100 3100
240 3027 3027
2400 2954 2954
•
2-38
• It is obvious that Eq. 2.35 gives exactly the same results as
the exponential-integral equation, for the times considered,
when computing the flowing formation face pressure.
b) Equation 2.35 cannot be used until the flow time exceeds

the limit expressed by the relationship:
ûc r2
t > 9.48x10'` k
t > 9.48x10`' (0.14)(1.2)(16x10-6)(0.25)2

80
t > 1.99x10-4 hrs
t > 0.7 seconds.
^ Practically, this means that Eq. 2.35 can be used anytime

during the transient flow period to compute pressure at
the formation face of the flowing well. In fact, we will
find that this is true for essentially all wells.
c) A plot of pressure versus logarithm of time for the

results in part (a) is presented in Fig. 2.14.
In summary, we have three solutions of the diffusivity

equation which apply to the transient flow of a slightly
compressible fluid in a radial system:
(1) Exact solution (Eq. 2.28)
(2) Exponential-integral solution (Eq. 2.31) -

This is an adequate approximation to the exact
^ solution anytime tD/rD > 25, i.e., anytime
t > 9.48x10" ûctr2/k. When tD/rD < 25, the
accuracy of this solution depends upon the
2-39
3
CO 3
.H
W
2 800
1 10 102 103 10"
t, hrs
Fig. 2,14; Flowing formation face pressure versus time, Ex. 2.2.
0 • 0
q values of rD and tD as discussed on page 2.29.
(3) Log-approximation of the Ei solution (Eq. 2.34) -

This is an adequate approximation of the exact
solution anytime x < 0.01, i.e., when t > 9.48x10"
W tr2/k.
When considering pressures at the formation face of the

producing well, which is by far the most common problem we
face, the log-approximation ( Eq. 2.35) is almost always
adequate. When pressures must be calculated, or analyzed,
that are at a location away from the formation face (r » rw),
it will probably be necessary to use the exponential-integral
solution given by Eq. 2.31. The exact solution is seldom
required in practical well test work.
o b. Highly compressible fluids (gas)

When developing the radial diffusivity equation in the
G AS
^
^....J previous section, it was shown that the flow behavior of a
fluid of density p could be described by Eq. 2.13:
ivâss ââ^c^
(2.13)
r ar [r uk ar] = at p) •
4j- The density of gas can be expressed by the non-ideal gas

law as P V Z r) RT- - n..vn
Qv Z P-7
P (2.36)
zR
where: p gas density

v - P^ R
Z
M = molecular weight of gas
z = gas deviation factor
R = universal gas constant
T = reservoir temperature
With the substitution of this equation of state, Eq. 2.13

• becomes
2-41
,&,:?-, c. : o& L.. pc,rILI^ oY. ^c.e s swlu
_
P P
6F
pSeudO D YPSS(.kr2 :
f
^
(2.37)
^ r Tr ^ r ^ 7 at (z)
( nar li n c eu+ ^ cZU a^ti ati.^

where k and^ have been assumed constant. Expansion and
rearrangement of this equation yields:
âsguw,e- ne^► ^gible
1 2 + uZ
22^ + 32P2 +
r ar ar2 2 uZ [] 1 p{ ap }
[ 15 _ 1 aapz (2.38)
k at z '
The isothermal coefficient of compressibility was

defined by Eq. 2.14 as
c = - 1 ^V . (2.14) ^
V ap
^--^ S pe c,^ ^+ c trot^ me^

However, v can be expressed in terms of the non-ideal gas
law as
v = p = PMT (2.39)
so that (2.40)
L^± -V, Shang ^''u'c:G ► -m,. C-^ prcSSttir -&

If Eq. 2.40 is substituted into Eq. 2.38, and it is assumed 0
that the pressure gradient is small enough that gradient-
squared terms can be neglected, the gas diffusivity equation
becomes:
a?1)2 + 1 ^P2 = 1ucDPz (2.41)

r ;r k at
ar2
2-42
pî^. C^h tOY ( ^QUiO^: a?P- + L LP = ^^CT V I -
. u ar' r ar .oooZcq k at
Rock compressibility, as well as the presence of other fluids,
can be accounted for by introducing the total compressibility,
ct, into Eq. 2.41. Written in oilfield units, Eq. 2.41
becomes L* va-^mr. ôr 3aS,
a?p 2 + 1 ^P? = 0uCt DP?

(2.42)
r ar 0.000264k at
r2
Solutions and applications of this equation will be

presented in Chapter 7.
B. Linear flow
Equations which describe linear flow of liquids and
gases are derived from material balance considerations
using a procedure similar to that used to develop the
radial flow equations. The details of these derivations
• are available in the literature2 and will not be presented

here.
a. Slightly compressible fluids

The linear diffusivity equation for liquids is:
^+a+a=ûct ^p (2.43)
ax2 ay2 az2 0.000 2 6 4k at
This equation is limited by the same assumptions made in

deriving Eq. 2.25.
b. Highly compressible fluids

The gas diffusivity equation has the form:
2p2- + a?P2 + a?P? = ¢uCt 3P2 (2.44)

ax2 ay2 az2 0.000264k at
•
2-43
4. Late transient flow
•
Late transient flow begins when the pressure disturbance
created by placing a well on production reaches the nearest
boundary; it ends after the disturbance has reached all
boundaries defining the drainage area of the well and the
reservoir stabilizes into pseudosteady state behavior. As
noted previously in this chapter, the flow behavior is
generally too complex during the late transient period to
be described mathematically. Therefore, no attempt will
be made to present equations for this time regime.
n
5. Pseudosteady state flow ^îe âc^,,dL qLC bQur,oân,^¢,o
----- After a well has produced-at constant rate for a
sufficient period of time for the pressure disturbance to
reach all drainage boundaries, the reservoir will eventually
achieve a state of flow where pressures everywhere in the
drainage area are decreasing at the same rate. This behavior
was illustrated by Fig. 2.18.
A. Slightly compressible fluids (liquids)
a. Radial flow
The general solution for radial flow in a reservoir
bounded with no-flow boundaries was presented previously
as Eq. 2.28. Futhermore, it was found that for the special
case where r=rw, the flowing bottomhole pressure, Pwf'
could be computed using Eq. 2.30. When time is sufficiently
large in Eq. 2.30 for pseudosteady state to begin, the
summation term approaches zero and Eq. 2.30 becomes
r2
pwf = pi - 141.2 k t D + ln reD - 4 (2.45)
^rZeD
âj-_^^
{^,o Equation 2.45 is more useful if it is written in terms
^w d of the volumetric average reservoir pressure, PR- From a •
2-44
• material balance, the change in reservoir pressure which
will result from producing qt barrels of fluid from a
reservoir of pore volume Ah^ will be:
(q, STB)( D )(t,hrs)(5.615 ft3)(B, bbl

D 24hrs bbl STB)
Pi PR _
(Ah^,ft3)(ct'psi 1)
t B
Pi - PR = 0.2339 Ahq (2.46)
c
t
Introduction of Eq. 2.46 into Eq. 2.45 yields:
r
j pwf = PR - 141.2 k 11 n re -0.751
w J
Another useful form of the radial pseudosteady state

^ equation is obtained by solving Eq. 2.47 for the flow
rate, q:
q = 0.00708 kh pR-pwf 0 (2.48)

uB r
In re -0.75
w
It is interesting to observe that Eq. 2.48 is almost

identical in form to the incompressible, radial, steady
state equation previously presented as Eq. 2.8; the only
difference is in the constant which appears in the
denominator. The similarity of these equations, which
represent vastly different systems, has caused many
engineers to mistakenly conclude that steady state and
pseudosteady state are equivalent, or similar, systems.
h. Generalized geometry
During pseudosteady state flow, the shape of the
drainage area dictates the geometry of flow. Rather than
2-45
^{ 3 S^ o.^ 2= acJw
develop a separate equation for each geometry, we present

•
a general equation developed by Ramey and Cobb' which is
applicable to a large variety of drainage shapes:
_ 162.6 qBp 4A 0.2339 qBt

(2.49)
Pwf Pi kh Ah^ct
log 1.781C r2
A w
where: A = drainage area of well, ft2

CA = reservoir shape factor, dimensionless
The shape factor, CA, which depends upon the geometry of

the drainage area and the well location within the drainage
area, is presented' in Table 2.2 for various single-well
drainage areas.
Equation 2.49 is applicable to any drainage area for
which a shape factor is available. This equation can also
be expressed in terms of the average reservoir pressure
by introducing Eq. 2.46:
•
_ 162.6 qBp log 4A
(2 . SO)
PR pwf kh
g 1.781CArw
If Eq. 2.50 is applied to the familiar example of a

cylindrical reservoir of radius, re, where
A = 7r2
e
and CA = 31.62,
it reduces to
r
141.2 khu [ In re - 00..7 5 . " (2.51)
QS^ pwf = PR - w J
U'" •
Note that Eq. 2.51 is identical to Eq. 2.47 which was
developed from the general solution.
•
2-46
d; VAA^ VIM - We +10,VS^ u,,
^'"
Table 2.2; Shape factors f or =rolusîosed
^ //s/ingle-wel l drainage
• areas (after Ref. 7). +;,,,l fo

AARUnr»
," QcL..
(gT Ip6v.rriwt.l]
USE INFINITE SYSTEM

LESS THAN
2.2458 EXACT SOLUTION WITH LESS
CA -in CA 112 In ^ 1% ERROR
1/o
Cq FOR t0A> THAN 1% ERROR
FOR tpq>
IN BOUNOEO RESERVOIRS FOR It DA
O 31.62 3.4538 - 1:3224 0.1 0.06 0.10
0 31.6 3.4532 - 1.3220 0.1 0.06 0.10
27.6 3.3178 -1.2544 0.2 0.07 0.09

A
*
â 27.1 3.2995 -1.2452 0.2 0.07 0.09
21.9 3.0865 - 1.1387 0.4 0.12 0.08

1/3{
^ l4 0.098 -2.3227 +1.5659 0.9 0.60 0.015

31
• Fol
30.8828 3.4302 - 1.3106 0.1 0.05 0.09
12.9851 2.5638 -0.8774 0.7 0.25 0.03

FR
4.5132 1.5070 - 0.3490 0.6 0.30 0.025
® 3.3351 1.2045 -0.1977 0.7 0.25 0.01
21.8369 3.0836 - 1.1373 0.3 0.15 0.025

2
^1 10.8374 2.3830 -0.7870 0.4 0.15 0.025

2
^ 1 4.5141 1.5072 -0.3491 1.5 0.50 0.06

2
^ 1 2.0769 0.7309 +0.0391 1.7 0.50 0.02

2
q ®1 3.1573 1.1497 -0.1703 0.4 0.15 0.005

2
2-47
Table 2.2: CONT'D.
q
0.5813 -0.5425 +0.6758 2.0 0.60 0.02

®1
2
0.1109 -2.1991 *1.5041 3.0 0.60 0.005

^ 1
2
• 1 5.3790 1.6825 -0.4367 0.8 0.30 0.01
2.6896 0.9894 -0A902 0.8 0.30 0.01

4
0.2318 - 1.4619 +1.1355 4.0 2.00 0.03

4
• 0.1155 -2.1585 +1.4838 1.0 2.00 0.01

4
•
S
l 2.3606 0.8589 -0.0249 1_0 0.40 0.025 •
IN V£RT/CALLY-FRACT(JR£ORES£RW/RS USE {xe/x} 12 IN PLACE OF A/r,2, FOR FRACTURED SYSTEMS
+ =xF/xe 2.6541 0.9761 -0.0835 0.175 0.08 CANNOT USE
I t 2.0348 0.7104 +0.0493 0.175 0.09 CANNOT USE
I t 1.9986 0.6924 +0.0583 0.175 0.09 CANNOT USE
-05
[7
1.6620 0.5080 +0.1505 0.175 0.09 CANNOT USE
1 -+- 1.3127 0.2721 +0.2685 0.175 0.09 CANNOT USE e
1 0.7887 -0.2374 +0.5232 0.175 0.09 CANNOT USE

P
IN WAMR-ORIVE RESERVOIRS
19.1 2.95 -1.07 - - -

(D
IN RESERVOIRS OF UNKNOWN
P/iG10C/CTION CHARACTER
25A 322 -1.20 - - -

•
2-48
B. Hi ghly compressible fluids (gas)
Pseudosteady state solutions of the gas equation will
be presented in Chapter 7.
Example 2.3: Calculation of formation face pressure during

pseudosteady state flow
Problem. An interior well is producing from a reservoir,

developed on a 40-acre square drilling pattern which has the
following properties:
k = 175 md h = 27 ft
= 18% r w = 0.33 ft
u= 1.1 cp q= 720 STB/D
B = 1.37 RB/STB c t = 24.2x10 6psi-1
pi = 4,440 psia A = 40 acres
^
a) Assuming this well has been producing at constant rate
for a sufficiently long time to achieve pseudosteady state
flow, compute the flowing formation face pressure at the
following times: 20, 50, 100, 150, 200 and 250 hours. Plot
these results as pressure versus time on cartesian paper.
b) Based on this plot, what do you conclude about the

behavior of the flowing bottomhole pressure of a well
during pseudosteady state flow? How much did the volumetric
average pressure in the drainage area of this well decrease
from t=20 to t=250 hours?
Solution. The flowing formation face pressure during

pseudosteady state flow is described by Eq. 2.49:
_ 162.6 4A 0.2339 qBt

Bu lo
pwf = Pi kh g 2 Ah^ct
is L.781CArW
2-49
The reservoir shape factor, C A, is obtained from Table 2.2
as a function of drainage geometry and well location within
the drainage area. This well is producing from the interior
of a field developed on 40-acre square spacing; accordingly,
under ideal conditions, the well is centered in a square
drainage area which is 40 acres in size. The shape factor
for a well in the center of a square is 30.88. Therefore,
(162.6) (720) (1.37) (1.1) log (4) (40) (43,560)

4,440 -
p wf ( 175 ) (27 )
L
(l.781)(30.88)(0.33)2
_ (0.2339)(720)(1.37) t
(40) (43,560) (27) (0.18) (24.2x10 6)
0 pwf = 4213.5 - 1.126 t
a) The flowing formation face pressure is summarized as a

function of time in the following table:
t,hrs pwf,psia
20 4191
50 4157
100 4101
150 4045
200 3988
250 3932
These results are presented graphically in Fig. 2.15.
b) It is obvious from Eq. 2.49 and from the results depicted

by Fig. 2.15, that pressure decreases as a Zinear function
of time during pseudosteady state flow. This provides a
method of detecting when pseudosteady state flow is present
2-50
E 4200
4100
M
I-q
4000
SLOPE = - 1.126 psi/hr
3900
3800 L-
• 0 50 100 150
t, hours
200 250
Fig. 2.15: Flowing formation face pressure versus

time during pseudosteady state flow,
Ex. 2.3.
in a reservoir. It will be shown in the next chapter that

the rate of pressure decline during pseudosteady state flow
is directly related to the pore volume being drained by the
well.
The rate of pressure decline during pseudosteady state
is the same at all locations in the drainage area affected;
consequently, the average reservoir pressure must decline at
this same rate. Since the formation face pressure is declining
at a rate of 1.126 psi/hr in this well, it follows that the
change in average pressure from 20 to 250 hours is
A PR = (1.126 psi/hr)(250-20)
APR = 259 psi
2-51
6. Productivity Index 0
A commonly used measure of the productive potential

of an oil well is the productivity index. Defined by the
symbol J, the productivity index is the[production rate per
unit pressure drop that a well can produce J i.e.,
,
j = q (2.52)
PR pwf
The productivity index can be computed from theoretical

relationships but is generally measured during a production
test. If PR is known, a well can be produced at rate q until
it stabilizes, and the flowing formation face pressure can
be measured at that time. The productivity index is computed
using Eq. 2.52.
It is important to note that the productivity index is
a valid measure of a wells productivity only if the well is
flowing at pseudosteady state conditions. Therefore, in
order to accurately measure the productivity index of a well,
it is required that the flow rate and pressure drawdown be
measured after the well has produced for a sufficient length
of time to reach pseudosteady state. This is made clear if
we refer to Fig. 2.16; this figure shows a plot of pwf and
PR as a function of producing time for a constant rate well.
Note that the pressure drawdown, PR - pwf, is a continuously
varying quantity until pseudosteady state begins. Only after
pseudosteady state conditions are attained does the pressure
drawdown become constant. If the pressure drawdown is
measured before pseudosteady state begins, the value of
pwf' and the corresponding value of J, will vary depending
PR
upon what time the measurement is made. In particular, a
productivity test not run long enough will give a value of J
which is too large.
Recognizing that J must be defined in terms of pseudo-
steady state flow conditions, Eq. 2.50 can be substituted
into Eq. 2.52 to obtain
2-52
Se no^F, ^ylno^c , l.''tua' n ad^L , firam s^'ew"^ ^ s S ] l,^rw : pw^ vs -FZm e
h62.4 ^ kr slra^y.i Gm^

k ^ [ ^ `- i -r 3 • 23
^,µG rW ^
•
PR d4ap5 6-4
pR-pwf I^Vw 011-^
V • pwf
e'k PD-h e^1114IL3
^T 2rfs Start of pseudosteady
state flow
I, h 0+ Ccrn 5i'A^^ 4141,i•tn t In/ war n i n W a +^^' c,^ -f-a-o ^gi

a ---^ Meav^^^.g(es5 ^S +eS"(^l7 SGÔn.^ -b /tâG^n PS^
'fY af/V`SL (i1n^ ^ ( 6u1^ V
Fig. 2.16: Behavior of average reservoir
pressure and flowing well
t ^pk, pressure i n a bounded reservoir.
^Ghs+OLy,t 6ru-
PSS J = 0.0142
kh
STB/D/psi (2.53)
4
uB (ln A + in
1.781CA)
r2
w
For the most commonly used flow geometry, i.e., a circular

reservoir, this equation reduces to
kh , STB/D/psi. (2.54)
J = 0.00708
r
uB (ln re - 0.75)
w
• In many oil wells, pseudosteady state flow is

approximated during most of their productive life. The
productivity index is valuable as a method to predict
the future performance of these wells. Further, by
monitoring the productivity index during the life of a
well, it is possible to determine if a well has become
2-53
damaged due to workover, production or injection operations.
•
If it is recognized that a well is "damaged", an appropriate
stimulation treatment can be designed to remove the damage
and increase productivity. One of the most important
applications of a pressure transient test is to determine
the productive capacity of a well; by comparing this with
the ideal productivity index of the well, the need for well
stimulation can be evaluated.
7. Constant-pressure outer boundary

All of the solutions to the diffusivity equation
previously presented apply to a reservoir with a no-flow
outer boundary. Reservoirs subject to edge-water influx,
waterflooding, or enhanced recovery projects often have
boundaries that approximate a constant pressure.
The purpose of this section is to describe the pressure
behavior of a well producing a slightly compressible fluid
at constant rate from a reservoir with a constant pressure
is
boundary. Only the solution for the flowing formation
pressure, pwf, is presented. This solution is2
141.2 ^ [ln reD

pwf - pi -
co
esn tDw Jo(SnreD)
- 2 (2.55)
Y_ * S2 [Ji (Sn) - Jo (SnreD)1
n=1
/ .
r
where
reD = re
w
kt
t = 0.0002637
•
Dw 2
ûctrw
2-54
• and Sn is a root of
J1 (Bn) Y o (^nreD) - Y1 (Sn) Jo ( SnreD) = 0. (2.56)
The solution of Eq. 2.55 is presented 1 graphically in

Fig. 2.17.
Prior to the time the pressure disturbance reaches
the boundary, the constant pressure solution is identical
to the transient flow solution. After the pressure disturbance
reaches the boundary, pressure performance is determined
by the value of re/rw.
It is observed from Fig. 2.17 that the pressure
difference (Pi pwf) becomes constant at large flow times;
this means that the system has reached a steady state
condition. This can also be observed from Eq. 2.55; as t
becomes large in Eq. 2.55, the summation term approaches
zero and the equation reduces to:
• -
r
141.2 ^ ln re
lbQq„^
(2.57)
pwf = pi
w
p • -p
or q= 0.00708 µB 1 rwf (2.58)
ln e
rw
This system.achieves steady state when'
ûC r2
t > 4. 74x103 --F- e (2.59)
A discussion of how constant pressure solutions can be

applied to water drive reservoirs has been presented by
Ramey, et a18.
•
2-55
•
8.5
8.0
7.5
7.0
,-, 6.5
4-i
3
^•^
6.0
^
,C N
r- 5.5
r- •
5.0
4.5
4.0
3.5 103
104 fOS 106 107
0.0002637kt
z
ûcr
t w
Fig. 2.17: Pressure behavior of a well in the center of a

closed circular reservoir with constant external
pressure. After van Everdingen and Hurst'
2-56
•
IV. PRINCIPLE OF SUPERPOSITION
All of the equations presented so far are valid only

for a single well which produces at constant rate from time
zero onward. Unfortunately, most reservoirs contain several
wells which produce at varying rates. CThe objective of this
section is to show how the principle of superposition can be
used to generate equations which describe the pressure
behavior of reservoirs with multiple wells producing at
varying rates.1
MathematicJally, the principle of superposition states
that adding solutions of a differential equation results
in a new solution to that differential equation, but for
different boundary conditions. For purposes of well test
analysis this principle can be stated5 in the following way:
^ The total pressure change at any point in a reservoir is the
sum of the pressure changes at that point caused by flow
in each of the wells in the reservoir. We will consider
applications of superposition in space,and applications to
varying rates.
1. Superposition in space
Consider the three wells in Fig. 2.18 which began
producing at the same time from an infinite reservoir.
According to the principle of superposition, the total
pressure change at Well 1 is
(pi pwf)total at Well 1 (pi pwf)due to Well 1
+(pi p)due to Well 2
+(pi p)due to Well 3
•
2-57
^0.b0L, E. r^9^re ^P^Gp^z 10 •^0^^^ ^ •
âP^wû. - -7°• k ^8^'' ^^(-948

_^C^ ^
h J. ^
î rrii (a,, ^, w c,^-s 2 a.,,, dL 3 •

WELL 2
rZ ^. P(C,t)
r12
r2 3 r3
WELL 1
r13
WELL 3
Fig. 2.18: Three-well system in an infinite

reservoir.
à+,^.^ c'r..ôsc ptic sswu cl^ a^c^ ^(^04 pu ss wâ^ .

The pressure drawdown at Well 1 due to its own production
can be written using the log-approximation of the exponential-
integral solution presented in Eq. 2.35. The drawdown at
Well 1 due to production at Wells 2 and 3 must be written
in terms of the exponential-integral solution (Eq. 2.31)
because we are computing pressure a large distance from the
location of the rate change. Therefore,
QA.c..sswLL r.E•amo'^t a^^ cru ^^(â c,^rG2Q^ ^
q Bu
(pi pwf)total at Well 1- 162.6 -^ log Wctr2 - 3.23
t wl
q2Bu 948ûctrl2
- 70.6 kh Ei - kt
Bu 948ûc r2 ^
ktt 13
- 70.6 akh Ei (2.60)
2-58
• where ql, q2 and q3 refer to the respective producing rates
of Wells 1, 2 and 3.
This same concept can be used to compute the pressure
drawdown at Wells 2 or 3. Further, any number of wells
flowing at constant rate in an infinite system can be treated
in this manner.
A very important application of this principle is in
the analysis of interference tests. A typical interference
test is conducted by changing the rate at one well while
measuring the pressure response at one or more adjacent
wells. Proper analysis of the data from an interference test
can yield information about formation properties between
the wells. A modification of the interference test, called
a puZse test, is experiencing widespread use today. These
tests will be discussed in more detail in Chapter 9.
• Example 2.4: Application of superposition to constant rate

wells in an infinite system.
Problem. If the wells depicted by Fig. 2.18 are placed on

production at the same time, compute the flowing formation
face pressure at Well 1 after 72 hours of production.
ql = 80 STB/D = 16%
= 150 STB/D ct = 24. 6x10-6psi-1
q2
a3 = 210 STB/D h = 24 ft
rw2 = rw3 = 0.33 ft r12 = 800 ft
rwl =
B = 1.35 RB/STB 500 ft
r 13 =
W = 2.5 cp r23 = 600 ft
.
k = 56 md = 473 5 psia
pi
What assumption is made in these calculations concerning

the location of reservoir boundaries?
• Solution. The total pressure drop experienced by Well 1 is

the summation of pressure drops caused at that location by
2-59
each well:
0
/Y%^ (pi pwf)total at Well 1 ( ")Well 1 +( Ap)Well 2+('^p)Well 3
The pressure drop at Well 1 caused by its own production is

computed using Eq. 2.35:
q Bu
-W-
^ log kt2 - 3.23
(Ap)Well 1 = 162.6
ûctrwl
162.6 (80) (1.35) (2.5) log (56) (72)-6 - 3.23

=
56 24
(0. 16) (2. 5) (24.6xl0 )(0. 333) 2
(AP)Well 1 = 207.0 psi
Since Wells 2 and 3 are located a significant distance from

Well 1, the exponential-integral solution is used to compute
•
the pressure drop caused by production of these wells.
According to Eq. 2.31,
q Bu `ûctrl2
(AP)Well 2 = -70.6 kh Ei -948
kt
_ -70.6 (150) ( 1.35) (2.5) Ei 948(0.16) ( 2.5) (24.6x10-6) ( 800) `_ SI

5 ( 24 ) ( 56 ) ( 72 )
^
= -26.6 Ei(-1.48)
.
(AP)Well 2 = 2.7 psi.
The pressure drop caused by Well 3 is
_ -70.6 (210)(1.35)(2•5) Ei 948 (0•16)(2.5)(24.6x106)(500)2

(^p)
Well 3 5 TS- 6 )( 72 )
_10
= -37.2 Ei(-0.578)
(AP)Well 3 = 17.7 psi.

2-60
0
Superimposing these effects,
207•0 + 2'7 + 17'7

(pi pwf) total of Well 1
(pi pwf)total of Well 1- 227 psi
(pwf)Well 1 - 4735 - 227
(pwf) We 11 1 - 4508 psi.
These calculations assume that pressure disturbances

from all three wells have not reached the reservoir boundary,
i.e., flow is transient after 72 hours.
^ The principle of superposition can also he used to

predict the pressure behavior of a well in a bounded reservoir.
Consider for example the well in Fig. 2.19 which is located
a distance L from a no-flow boundary.
(1111 ^.ûI ^y^;crw ^60
NO FLOW
BOUNDARY
0 Fig. 2.19: Well near a no-flow boundary.
2-61
This situation occurs often in reservoirs when a well is
located near a sealing fault. It can be shown that two
identical wells producing at equal rates from an ideal
•
reservoir will always have a no-flow boundary equidistance
between the wells. Consequently, the effect on the well
of the boundary in Fig. 2.19 would be the same as the effect
of.an image well located a distance 2L from the actual well.
The use of an image well to simulate the no-flow boundary
is illustrated by Fig. 2.20.
ACTUAL IMAGE
WELL WELL
Qf--- L L --ti
NO FLOW
BOUNDARY
Fig. 2.20: Use of an image well to simulate a

no-flow boundary.
Mathematically, this problem reduces to one of

determining the effect of the image well on the actual
well. The total pressure drop at the actual well will be
the pressure drop due to its own production plus the
pressure drop caused by the image well:
[ log
kt 2- 3.23
Pi pwf - 162.6 a
ûctrw
948ûct (2L) 2
- 70.6 khu Ei (2.61)
kt
•
2-62
APC 206 2/01
âdarlot PROJECT COMPUTATIONS

Petroleum Corporation
NO. OF SHEETS
• DATE PROJECT JOB NO.
BY CHK.: SUBJECT -&L (V Y' 00 S^ ^ OYl
0
'
Noti ce that this equation assumes the reservoir is infinite
except for the indicated boundary.
The concept of image wells can be used to generate the
pressure behavior of a well located within a variety of
boundary configurations. Many of the methods considered
later in this manual are based on this principle.
2. Superposition of flow rates

All of the equations presented so far in this text
require that the wells being considered produce at constant
rate. Practically all wells produce at varying rates and,
consequently, it is important that we be able to predict
pressure behavior when rate changes. [For this purpose, the
principle of superposition states: The total pressure change
in a well which produces at multiple rates is the summation
of pressure changes caused individually by each change in
rate.
Ansider the production history of a well as depicted
by Fig. 2.21:
c^
^ Time
Fig. 2.21: Rate history of a variable rate well-.
2-63
The total pressure change at the formation face of the
producing well at time t will be the summation of the
pressure changes at time t caused separately by each rate
change.
AP)q3-q2
-0 + (pp) + (
(pi pwf)total (Ap)
The pressure change caused by increasing the rate from

zero to ql at time zero is given by Eq. 2.35 as
kt Z - 3.23
(Ap)ql-0 = 162.6 (ql khBu
0) [los
oucrw
t
It is important to observe in this equation that it is the

rate change (new rate minus old rate), not the new rate, that
enters the equation; it is the change in rate that causes
the pressure disturbance. Further, it should be noted that
time in this equation represents the length of time that
particular rate change has been in effect. Once a rate
change occurs, a pressure disturbance is created in the
reservoir which will last forever, and which acts independently
of pressure disturbances caused by other rate changes.
Therefore, the pressure change at time t caused by the rate
change (ql-0) which occurred at time zero will be the same
regardless of other rate changes which have occurred, or
will occur, in the reservoir. Using this same concept, the
pressure changes at time t caused by the other changes in
rate can be computed from Eq. 2.35 as:
(q2-ql)BU k(t-t1)
(Ap) = 162.6 log - 3.23
kh
q2-al ûc-tr2
(q..)Bu k(t-t )
3-q2 2- 3.23
(^p) = 162.6 log
q3'q2 ûctrW ^
2-64
^ By summing the indivi dual pressure changes, the flowing
formation face pressure of the producing well is determined
at time t to be:
B
pwf = pi -•162.6
a khu log kt Z- 3.23
ûctrw
(q2-c{1)BU k(t-tl)
- 162.6 log - 3.23
kh
P+o+ uj, ûctrw
(a3-q2)BU k(t-t2)
- 162.6 log - 3.23 (2.62)
kh 2
ûctw
^ By combining the concepts of superposition in space

and superposition of rates it is possible to describe the
pressure behavior of a reservoir with any number of wells
producing at variable rates. Superposition concepts are
required in the analysis of most pressure transient tests
conducted today. These applications will be evident
throughout this manual.
Example 2.5: Determination of flowing formation face

pressure in a variable rate well
Problem. Compute the flowing formation face pressure after

120 hours for a well which has the rate history depicted by
Fig. 2.21.
ql = 400 STB/D B = 1.21 RB/STB

q2 = 150 STB/D u = 3.1 cp
q3 = 230 STB/D h = 37 ft
• t1 = 48 hours
t2 = 96 hours
ct = 15x10-6psi-1
rw = 0.5 ft
k = 38 md pi = 3820 psia
26%
2-65
•
Solution. The flowing formation face pressure can be
computed directly using Eq. 2.62:
pwf = 3820
(162.6)(400)-(1.21)(3.1) log (38) (120) - 3.23]

(38)(37)
(0.26)(3.1)(15xl0-6)(0.5)2 I
(162.6) (150-400) (1.21) (3.1) lo _ (38) (120-48) - 3 23

(3
g (0.26)(3.1)(15x10-6)(0.5)2
(162.6) (230-150) (1.21) (3. 1) Flog (38) (120-96) _ 3.23

( 38 )( 37 ) ( 0.26)(3.1)(15x10-6)(0.5)`
pwf = 3820 - 1032 + 621 - 182

•
pwf = S`2^ psia
V. FLOW EQUATIONS IN DIMENSIONLESS FORM
Because of the large number of variables in the flow

equations presented thus far, it is very difficult to present
solutions in a compact form which is convenient to use. For
this reason, it is customary to present graphical or tabulated
solutions to flow equations in terms of dimensionless variables.
The purpose of this section is to introduce the concept of
dimensionless variables, define some commonly used dimension-
less variables, and present some of the more useful flow
2-66
r^
L equations in dimensionless form. First, we will show by the
following simple example how data presentation can be
simplified and compacted by judicious choice of the plotting
parameters.
Example 2.6: Illustration of how proper selection of

plotting parameters can simplify presentation
of data
Problem. Consider a reservoir for which the following

properties are known:
pi = 5000 psia = 22%

q = 400 STB/D k = 30 md
B = 1.30 RB/STB u = 1.7 cp
h = 20 ft ct = l2x10-6psi_1
• If you assume that only one well is producing from this

reservoir under transient flow conditions,
a) Compute the reservoir pressure at radii of 10 feet and

100 feet for flowing times of 0.1, 0.5, 1, 2, 5, 10, 20, 50
and 100 hours. Present this data on a graph of log [pi-p(r,t)]
versus log t
b) Present the data from (a) on a graph of log [pi-p(r,t)]

versus log (t/r2).
Solution. Reservoir pressures during transient flow can be

determined using Eq. 2.31:
•
2-67
P (r, t) = pi + 70.6
e
B
- Ei -948
t
uctrz
•
ûc r2
pi-p(r,t) = -70.6 k qB" Ei -948
kt
(1. 7) Ei ( 948) (0. 22) (301.7) (12x10- 6 ) r2

pi-p ( r , t) = - 70 .6 (400) 30 . 30)0
2
pi-p(r,t) = 104.0 Ei (-1.418x10 _a rt )
a) Reservoir pressure is presented as a function of radius

and time in the following table and in Fig. 2.22:
r = 10 feet r = 100 feet

, hr s
t/r2 Op,psi t/r2 Ap,psi •
0.1 1.0X10 -3 157 1.0X10-5 0
0.5 5.0x10 -3 314 5.0x10-5 2
1 1.Ox10 -2 384 1.Ox10 4 12
2 2.0x10 -2 455 2.0x10-4 38
5 5.0x10 -2 550 5.0x10-4 99
10 1.0x10 -1 622 1.0x10-3 157
20 2.0x10 -1 694 2.0x10-3 223
50 5.0x10 +1 790 5.0x10-3 314
100 1.0 862 l,Oxlo-2 384
•
2-68
0 3
10
,r ,l,3
102
.r.,
^
14
¢.
10
0.1 1 10 100
t,hrs
Fig. 2.22: Pressure drawdown as a function of time

for Ex. 2.6.
b) Reservoir pressure data are presented as log [pi-p(r,t)]

versus log (t/r2) in Fig. 2.23.
•
2-69
"L/1-2 Roriy,a,(A^S G^dL^Q.
t or e^-Fec#- o^ +'
103
r-I 102
^
10
1
10-5 10-4 10-3 10- 2 10-1 1
t/r2
Fig. 2.23: Pressure drawdown as a function of t/r2 for Ex. 2.6.
• • •
It is observed in this example that a plot of pressure
drawdown versus t gives two separate curves for the two
radii considered. If 100 different radii*were of interest,
100 different curves would be generated. An obvious question
arises in this situation: How do you present solutions in
compact form when a large number of variables are involved?
A clue to answering this question is provided by considering
Fig. 2.23. When the pressure drawdown is plotted versus t/rZ,
the data for both radii form a common curve. In fact, the
pressure drawdown for any reservoir radius will plot on this
exact same curve.
If radius and time are the only variables, all pressure
data can be plotted on a single curve in the manner indicated
by Fig. 2.23. Other variables in the flow equations, however,
include 0, u, ct, k, h, q, and B. Obviously, an infinite
number of solutions exist when all of these variables are
considered. f We will show in the following discussion that by
properly selecting the plotting parameters, the effect of all
these variables on pressure can be presented izi compact form.
The key to solving this problem is to use dimensionless
variables.I
Most dimensionless variables used in well testing arise
logically from the equations which describe reservoir fluid
flow. To introduce this dimensionless solution approach,
consider for example the radial, incompressible, steady
state solution presented previously as Eq. 2.7:
kh(pe-pwf)
q = 0.00708 (2.7)
r
Bu ln (re)
w
This equation can be rearranged to solve for the pressure

difference
r
pwf - 141.2 khu In re (2.63)
Pe
• w
2-71
or,
•
r
pe pwf = in re (2.64)
w
141.2 k qBll
It is obvious that the right-hand side of this equation

has no units, i.e., it is dimensionless; accordingly,
the left-hand side must be dimensionless. Since the left-
hand side is dimensionless, and (pe-pwf) has units of psi,
it follows that the term 141.2 qBu/kh has units of pressure.
The left-hand side, therefore, is the pressure difference,
pe pwf' in dimensionless form; in fact, any pressure
difference divided by 141.2 qBu/kh is a dimensionless
pressure. Equation 2.64 can now be written in dimension-
less form as
r
PD = in re (2.65)
w ^
where PD = pe pwf . (2.66)
141.2 ^
This same concept can be expanded to consider unsteady

state equations where time is a variable. For unsteady
state flow, we define
pi-p(r,t)
(2.67)
PD
D = 141.2 Bu
kh
+ t = 0.0002637kt (2.68)
D ûctr2
-^ rD = r (2.69)
w
•
2-72
• where: PD = dimensionless pressure
tD = dimensionless time
rD = dimensionless radius
The exponential-integral solution (Eq. 2.31), when expressed

in dimensionless form, can be written as ^xp in+câl ec^âôr
Nrtêr ir d^w^v^Sio^nltSS ô^m
r2
Ei (- D _k (2.70)
1 PD = - 2
4tD
It was shown earlier that when tD/rD > 25, the exponential-
integral solution can be approximated by Eq. 2.34 which, in
dimensionless form is
t
PD = 1 [in D + 0.80907]. (2.71)
rD
• A plot of the exponential-integral equation is presented in

Fig. 2.24. While the exponential-integral equation has an
infinite number of solutions, plotting the equation in
dimensionless form permits us to present all solutions on
a single curve. We will show in the next section that
this curve, often called a type curve, can be used to
analyze pressure transient data.
The pseudosteady state equation for a slightly
compressible fluid in a generalized geometry was presented
previously as Eq. 2.49. In dimensionless form, this
equation becomes
PD = 2TrtDA + 1 In `2 + ^ In 2.C458 (2 . 72)

r A
w
kt
where tDA = 0.0002637 (2.73)
t
•
2-73
^^ Pe Cur^ e
EO,r 2.^0
I 10 102 103 104

10-1
tD/ rD
Fig. 2.24: Dimensionless pressure for a single well. in an ideal infinite system, i.e.,
the exponential-integral solution.
0 • •
• It is also useful to present other flow equations
in dimensionless form. These equations will be presented
throughout the text as they are needed to accomplish specific
well testing objectives.
VI. TYPE CURVE MATCHING
It was shown in the previous section that the exponential-

integral solution can be plotted in dimensionless form as
Fig. 2.24. This graph, often called a type curve, can be used
to analyze well test data to obtain information about
reservoir properties through a process called type curve
matching. The purpose of this section is to explain the
• basis and procedure for type curve matching.

Consider again the definition of dimensionless pressure
presented in Eq. 2.67:
141^2qBu [Pi-P(r,t)] (2.67)

PD
Notice that PD is a product of two terms which includes the

pressure difference, pi-p(r,t), and the group kh/141.2qBp.
When a well is producing at constant rate it is important
to observe that kh/141.2qBU is a constant and, therefore,
PD and [pi-p(r,t)] differ only by a constant multiplier.
Likewise, dimensionless time is defined as
Gn'Sf a.*,t
t JO . 0002637k t
(2.68)
D 2
L uCr
t w
so that , tD ='0.00026371 t
(2.74)
rD
L pctr2
2-75
It is observed that tD/r2D is also a product of two terms
which includes t and 0.0002637k/ûctr2. When pressure is
measured at some distance r from the producing well
(interference test), it is significant that 0.0002637k/Cictr2
is a constant; accordingly, tD/rD and t differ only by a
constant multiplier.
If we take the logarithm of both sides of Eqs. 2.67
and 2 . 74 ,
log PD = log [pi-p(r,t)] + log 141k2qBu (2.75)
log tD = log t + log 0.0002637k (2.76)

rD ûctr2
Recognizing that kh/141.2qBu and 0.0002637k/ûctr2 are

both constants, it is obvious from Eqs. 2.75 and 2.76 that
a plot of log PD versus log tD/rD will have exactly the
same shape as a plot of log [pi-p(r,t)] versus log t; this
is illustrated by Fig. 2.25. Not only do these curves have
the same shape, but if they are moved relative to each
other until they match, the vertical and horizontal dis-
placements required to achieve the match are related to
the constants in Eqs. 2.75 and 2.76. In particular, the
vertical displacement required to achieve the match is
related to the value of kh/141.2qBU; further, the required
horizontal displacement is related to the value of 0.0002637k/
ûctr2. Once these constants are known, it is possible to
determine reservoir properties such as permeability and
porosity.
In order to accomplish this matching process, we must
know the relationship between PD and tD/rD and the relation-
ship between [pi-p(r,t)] and t. PD and tD/rD are related
theoretically by the exponenial-integral solution which is
plotted in Fig. 2.24. The relationship between [pi-p(r,t)]
and t can be determined in the field for a reservoir of •
interest by conducting an appropriate pressure transient test.
2-76
• t
102 l_03
0.1 1 10
I 102
I0 10
^
_p
10 J 1
I10 103 los
102 10"
• tD
r2
D
Fig. 2.25 Relationship between pressure-time plots

in dimensional and dimensionless forms.
UA``^ An example of a field test which can be conducted to

evaluate reservoir properties is the interference test.
Suppose we place a well on production at constant rate while
at the same time we measure the pressure, Pobs' versus time
at the formation face of an observation well located a
distance, r, from the producing well. The formation
properties between the two wells can be determined by the
following type curve matching procedure:
•
2-77
1. Obtain a large-scale copy of Fig. 2.24. This type
•
curve is an ideal prediction of how the pressure
should behave at the observation well.
2. Obtain a sheet of tracing paper and construct a
log-log scale having the same dimensions as the
type curve you are using; this is best accomplished
by placing the tracing paper on the type curve
and tracing the major grid lines.
3. Plot your field data on the tracing paper as Ap
versus t, where bp = Pi Pobs' This plot, often
referred to as a data curve, is illustrated by
Fig. 2.26.
1 0z
• •
• •
•
•
•
0
., 10 •
•
DATA
•
• CURVE
1
10 2 1 n3
0.1 1 10
t , hrs
Fig. 2.26: Data curve for interference test.
4. Overlay the data curve onto the type curve and,

keeping x- and y-axes of the graphs exactly .
parallel, slide the data curve over the type curve
until you find the position where the curves best
match. An illustration of curves in a matched
position is presented in Fig. 2.27.
•
2-78
^ DATA
\ CURVE
TYPE
MATCH
CURVE
POINT
------ ----- -I
^ ---Q ^ -- -
(PD) M (AP) M
( t)M
t
(tD/rD)^^ ^
^ D/rD
Fig. 2.27: Illustration of type curve and data

curve in matched position.
S. With the curves in a matched position, select a

"match point"; a match point is an arbitrary
point common to both graphs.
6. Corresponding to the match point, record (Ap) M
from the ordinate of the data curve and (PD)M
from the ordinate of the type curve. These
values are related by Eq. 2.67, i.e.,
_ kh
PD 141.2qBU ^P
so that,
•
2-79
•
kh = (PD) M
141.2qBU (AP)M
LAlial d_c. yCtCQ, rr-ew+ (02fiui•

k = 141.2qBU ( pD)M (2.77)
h Ap)M
7. Corresponding to the match point, record (t) M

from the abscissa of the data curve and (tD/rD2)M
from the abscissa of the type curve. These
values are related by Eq. 2.74, i.e.,
kt
t- = 0.0002637
ûctrz
rD
so that,
•
0.0002637k (tD/rD)M
Cuctr2 (t)M
0.0002637k (OM (2.78)

0
uctrZ (tD/rD)M
Example 2.7: The use of type curve matching to analyze an

interference test.
Problem. An oil well is produced at a rate of 427 STB/day.

A shut-in well located at a distance of 340 feet from the
producing well is being used as an observation well.
Pressure data recorded in the observation well are tabulated
along with other rock and fluid properties. Using the
exponential-integral type curve, estimate the reservoir
permeability and porosity. The stabilized reservoir pressure
•
(pi) at the beginning of this test was 3456 psia.
2-80
• h 23 ft rw = 0.5 ft
0.8 cp ct = 8.3x10-6psi_1
PO
Bo 1.12 RB/STB pi = 3456 psia
427 STB/day
q0
t,hrs pobs'psia pi pobs
0 3456 0
1.0 3454 2
1.5 3451 5
2.0 3449 7
3.0 3444 12
5.0 3435 21
10.0 3423 33
20.0 3410 46
^ 30.0 3402 54
40.0 3396 60
60.0 3386 70
80.0 3380 76
100.0 3376 80
150.0 3366 90
200.0 3360 96
Solution. The field data are presented as a plot of

(pi pobs) versus t in Fig. 2.28. This data curve was
placed onto the exponential-integral type curve to obtain
the match presented in Fig. 2.29.
Corresponding to the match point on Fig. 2.29,
(Ap)M = 10 psi
(pD)M = 0.22
(t) M = 10
(tD/rD)M = 1.6.
2-81
10z •
•
• •
• •
• DATA
CURVE
•
f:10
•
•
1
1 10 102 ln3
t, }1 r s
Fig. 2.28: Data curve for Ex. 2.7.

is
Reservoir permeability can be computed using Eq. 2.77:
k = 141.2qBii (pD)M
h Op M
k = (141.2) (427) (1.12) ( 0.8) 0.22

23 10
k = 51.7 md.
Porosity can be computed using Eq. 2.78:
0.0002637k (t)M
uctr2 (tD/rD)IN4
•
2-82
• • •
00
W
i(
10'1 I 10 102 103 104
tD/rD
Fig. 2.29: Type curve match for interference test in Ex. 2.7.
(0.0002637) (51. 7) 10
•
1.6
(0.8)(8.3x10-6)(340)2
= 0.111
= 11.1%
It should be observed that porosity is directly related to

total compressibility in this calculation. Since compressi-
bility is not known with a high degree of accuracy in many
reservoirs, porosity computed in this manner can be subject
to significant error.
The concept of type curve matching is very important

in pressure transient analysis. Many different kinds of
type curves exist in the petroleum industry; however, the
concept of using the curves will not change from that
presented in this section. Other type curves will be
introduced in subsequent chapters of this text.
VII. SUMMARY
This chapter has described the variables which affect

pressure behavior, and has presented the major equations and
mathematical concepts which serve as the basis of commonly
used pressure transient tests. This material will be applied
in the following chapters to the design and interpretation of
pressure transient tests.
•
2-84
0
REFEREN C E S
van Everdingen, A. F. and Hurst, W.: "The Application

of the Laplace Transformation to Flow Problems in
Reservoirs," Trans., AIME (1949) 186, 305-324.
Matthews, C. S. and Russell, D. G.: Pressure Buildu p

-' sciety of
and Flow Tests in Wells, Monograph Serie
Petroleum Engineers of-AIME, Dallas ( 1967) 1.
Theis, C. V.: "The Relation Between the Lowering of the

Piezometric Surface and the Rate and Duration of Discharge
of a Well Using Ground-Water Storage," Trans. AGU (1935)
519-524.
4. Mueller, T. D. and Witherspoon, P. A.: "Pressure

Interference Effects Within Reservoirs and Aquifers,"
J. Pet. Tech. (April, 1965) 471-474; Trans., AIME, 234.
5. Lee, W. J.: Well Testing, Textbook Series, Society of

Petroleum Engineers of AIME, Dallas (1982) 1, 3-4.
6. Ramey, H. J., Jr., and Cobb, W. M.: "A General Buildup

Theory for a Well in a Closed Drainage Area," J. Pet.
Tech. (Dec., 1971) 1493-1505; Trans. AIME, 251
7. Earlougher, R. C., Jr.: Advances in Well Test Analysis,

Monograph Series, Society of Petroleum Engineers of AIME,
Dallas (1977) 5.
8. Ramey, H. J., Jr., Kumar, A., and Gulati, M. S.: Gas Well
Test Analysis Under Water-Drive Conditions, American Gas
Association, Arl in
gton, Va. (1973T.
2-85
E
NOMENCLATURE - CHAPTER 2
A = drainage area of well, ft2

B = formation volume factor, reservoir volume/surface
volume
B = oil formation volume factor, RB/STB
0
c = compressibility, psi-1
cf = formation compressibility, psi
cg = gas compressibility, psi
c = oil compressibility, psi
0
ct = total compressibility, psi
cw = water compressibility, psi
e = base of natural logarithm
Ei(-x) = exponential integral
h = net formation thickness, ft
J = productivity index, STB/D/psi
Jo = Bessel function
Jl = Bessel function
k = permeability of reservoir rock, md
kg = effective permeability to gas, md
ko = effective permeability to oil, md
ln = natural logarithm, base e
log = common logarithm, base 10
L = distance to no-flow boundary, ft
L = length of linear flow system, ft
M = molecular weight
p = pressure, psi
p(r,t) = pressure at radial position, r, at time, t, psi
= dimensionless pressure = khAp/141.2qBu
PD
Pe = pressure at r e in a radial system, psi
pi = initial reservoir pressure, psi
^ Po = arbitrary reference pressure, psi
psc = pressure at standard conditions, psi
2-86
^
pwf = flowing formation face pressure, psi.
= volumetric average reservoir pressure, psi
PR
q = flow rate, STB/D
qg = gas flow rate, ft3/D
agsc = gas flow rate at standard conditions, scf/D
r = radial distance, ft
re = distance to outer boundary of cylindrical reservoir, ft
reD re/rw
rD = r/rw
rw = wellbore radius, ft
R = universal gas constant
Sg = gas saturation, fraction
so = oil saturation, fraction
Sw = water saturation, fraction
t = time, hours
tD = dimensionless time = 0.0002637kt/^pctrk
tDA = dimensionless time = 0.0002637kt/ûctA
T = reservoir temperature, OR
Tsc = temperature at standard conditions, OR
u = apparent velocity (flux) = q/A
v = specific volume, ft3/lbm
V = volume, ft3
Vb = bulk volume, ft3
Vo = pore volume, ft3
x = linear coordinate, ft
x = 948ûctr2/kt in exponential-integral equation
y = linear coordinate, ft
Y1 = Bessel function
Y2 = Bessel function
z = linear coordinate, ft
z = gas deviation factor, dimensionless
ct n = roots of Eq. 2.29
6n = roots of Eq. 2.56
TI = hydraulic diffusivity = k/buct
u = viscosity, cp
ug = gas viscosity, cp
2-87
•
uo = oil viscosity, cp
p = density, lbm/ft3
pg = gas density, lbm/ft3
Po = oil density, ibm/ft3
0
2-88
•
SUMMARY OF MAJOR EQUATIONS - CHAPTER 2
Equation Number Equation

in Text
aZp + 1 ap ûct ap
2'26 ar2 - - = 0.0002
r ar 64k at
2.27 ct = co So + cw Sw -* cg S9 + cf
qBu 948¢uctr2
2.31 p(r,t) = pi + 70.6 kh Ei
kt
2.34 p(r,t) = pi - 162.6 khu log kt 2- 3.23

ûctr
• qB^^ kt
2.35 pwf = pi - 162.6 kh log - 3.23
2
o^^r ^,^ ? 31 uctrw
4A
2.49 p = p. - 162.6 ^ log
wf 1 kh
1.781CArw
0.2339qBt
Ah^ct
2.50 - p = 162.6 ^ log 4A

PR wf kh
1.781CArw
r ^
2.51 = PR - 141.2 k qB'j [In re - 0.75]
p^^^f
w
2.52 J = a
PR pwf
•
2-89
•
2.53 J = 0.0142kh
4
uB [in_+in
2
1,781CA
rw
0.00708kh
2.54 J =
r
uB
C In re - 0.751
w J
kh[pi - p(r,t)]
2.67
PD = 141.2qBp
t = 0.0002637kt
2.68
D Z
trw
2.69 rD = r
w is
r2
2.70 PD = - ZEi 4D
D
2.71
C
PD = 2 In D+ 0.809071
rD J
A 1
2. 72 PD = 27rtDA + 2 In 2 +^ In [2. CA58
rw J
2.73 tDA = 0.0002637 ¢u^ A
t
•
2-90
• PROBLEMS
1. Determine the value of Ei(-x) from Table 2.1 and from

Eq. 2.33 for the following values of x: 0.01, 0.02,
0.1, 1, 10. What percent error is made at each value
of x if Eq. 2.33 is used as an approximation to Ei(-x)?
What do you conclude about the range of applicability
of Eq. 2.33?
2. The discovery well in a consolidated sandstone

reservoir is placed on production initially at a
constant rate of 4,820 STB/D. Six weeks after this well
began producing, another well. was completed 2000 feet
away in the same formation. Reservoir properties are:
= 23% Bo = 1.27 RB/STB

co = 18x10 6psi -1 h = 42 ft
cw = 3x10-6psi-1 Po = 0.8 cp
cf = 3.4x10 6psi-1 k0 = 115 md
Sw = 35% rw = 0.5 ft
Sg = 0 pi = 5,190 psia
(a) Assuming that flow is transient, estimate the flowing

sandface pressure of the nroducing well after 6 weeks
of production. What is the minimum producing time at
which Eq. 2.35 becomes a valid approximation of the
exponential-integral equation?
(b) What pressure will the second well encounter at

the time of its completion six weeks after the first
well started producing?
(c) Can the log-approximation to the exponential-

integral solution (Eq. 2.34) be used to compute the
2-91.
pressure in (b)? If not, how much error would you
make by using Eq. 2.34? What conclusion can you
make from this result about the use of Eq. 2.34 to
calculate pressure at large distances from the
producing well?
3. Compute the pressure at Well 3 after 72 hours of

production for the reservoir described in Example
2.4.
4. An oil well is placed on production in a reservoir

at a rate of 270 STB/day. Pressure is simultaneously
measured in an observation well located 440 feet away
from the producing well. Reservoir rock and fluid data
are presented along with pressure versus time data from
the test. You are requested to estimate reservoir
permeability and porosity from this data. The stabilized
reservoir pressure at the beginning of the test was
3875 psia.
h = 18 ft pi = 3875 psia
uo = 2.1 cp ct = 1l.1x10-6psi
Bo = 1.21 RB/STB
t,hrs pobs'psia t,hrs pobs'psia

1 3875 60 i'0 3799
2 3875 80 ?! 3794
5 3875 100 3782
10 3872 150 11 5 3760
15 3861 200 J3G 3743
20 3853 300 3719
30 3840 400 r7 3702
40 L7 3828 500 17 3688
50 3818
q
What assumptions are you making in these calculations
concerning the effect of reservoir boundaries, or other wells?
2-92
SOLUTIONS
1. When x = 0.01, it is determined from Table 2.1 that

Ei(-x) = - 4.038. According to Eq. 2.33
Ei(-x) = in (1.781x)
Ei(-0.01) = -4.028
The error incurred by using Eq. 2.33 is:
Error = -4.028 - (-4.038) (100)

-4.038
Error = -0.25%
Calculations at the other values of x are summarized

• in the following table:
E i (-x) from Ei(-x) from Percent-

error
X Table 2.1 Eq. 2.33
0.01 -4.038 -4.028 -0.25
0.02 -3.355 - 3.335 -0.60
0.1 -1.823 -1.725 -5.4
1 -0.219 0.577 -363.5
ln -^.15x10 2.880 - 6.9x10'
As indicated in the text, Eq. 2.33 can be used to compute

Ei(-x) with less than 0.25 percent error if x < 0.01.
It is noted, however, that the error increases rapidly
for larger values of x.
2-93
2. (a) The formation face pressure of the producing well
• can be estimated during transient flow using Eq. 2.35:
pwf = pi - 162.6 k log kt 2 - 3.23

ûcrw
t
0
ct = coSo + c w S w + V + cf
ct = (18x10-6)(0.65)+(3x10-6)(0.35)+3.4x10-6
ct = 16.15 x 10-6psi-1
pwf = 5,190 _ (162.6) (11520^(1.27) (0.8)
(115) (1008) - 3.23

log
• (0.23) (0.8) (16.15x10-6) (0.5)2
pwf = 3,877 psia
Equation 2.35 is a valid approximation of the exponential-

integral solution when x < 0.01, i.e., when
$uct r2
t > 9.48 x 104
k
t> 9.48 x 10" (0,23)(0.8)(16. 1SX10-6) (0 .5) 2

115
t > 6.1 x 10-4 hrs
t > 2.2 secs
^ (b) Pressure can be computed using the exponential-

integral equation, i.e., Eq. 2.31:
2-94
' 0 .
011c r2
p(r,t) = pi + 70.6 kqB" Ei 948
kt
•
p(2000, 1008) = 5,190 + 70.6 (4,820) ( 1.27) (0.8) •
(115) (42)
• Ei F _ (948) (0.23) (0.8) (16.15x10-6) (2000) 2

I 115 1008
= 5,190 + 71.58 Ei (-9.72x10-2)
= 5,190 + 71.58 (-1.848)
p(2000, 1008) = 5,058 psia
(c) Equation 2.34 can be used to make this calculation:
r2
t > 9.48x10" ûc
-^- •
t > 9.48x10" (0.23) (0.8) (16.15x10-6) (2000)2

115
t > 9,799 hrs
Since this time exceeds the producing time of 1008 hours,

some error will be introduced into the calculation if
Eq. 2.34 is used. For comparison, Eq. 2.34 yields the
result:
p(r,t) = pi-162.6 aBU log kt - 3.23

kh
W r 2
p(2000, 1008) = 5,190 - (162.6)(4,820)(1.27)(0.8) •

(11S)(42)
•
(115)(1008)
• _ - 3.23
[log (0.23) (0. 8) (16. 5x10 6) (2000) 2
2-95
• p(2000, 1008) = 5,065 psia
This ? psi difference represents a 0.6% error in

the calculated pressure drop at r=2000 ft. While in
this case the error is not large, it does indicate
that the log-approximation becomes less accurate as r
increases. As r increases, the minimum time at which
the log-approximation can be used will eventually exceed
the producing time of the well considered. CFor this
reason, Eqs. 2.34 and 2.35 are generally used to calculate
pressures only at the formation face of the test well,
whereas Eq. 2.31 is used to calculate pressures at radii
beyond the formation face.1
Since the majority of well tests involve pressure
measurements at the formation face of the producing well
(pressure drawdown, pressure buildup, pressure falloff,
etc.) Eq. 2.35 will apply to most pressure transient
• tests. However, some tests involve more than one well
such that the rate is changed in one well and pressure
is recorded in an adjacent well. These tests, called
interference tests, generally require use of the
exponential-integral equation since the pressures are
recorded a long distance from the producing well.
2-96
3. The total pressure drop at Well 3 is the summation of
pressure drops at that location caused separately by
• production of Wells 1, 2 and 3.
(AP)total at Well 3 (AP)Well 1+(AP)Well 2+(A P)Well 3
The effects of Wells 1 and 2 on the pressure behavior of

Well 3 can be determined using the exponential-integral
solution, i.e., Eq. 2.31. The pressure drop at Well 3
caused by Well 1 is:
q1BU 948ûctr1 ; I
(^P)Well 1 = -70.6 kh Ei -
kt
- 70.6 (80)(1.35)(2.5) •
TS-6T(24)
• Ei (948)(0.16)(2.5)(24.6xl0-6)(500)21
(56) 72 j
_ -14.2 Ei (-0.578)
(-14.2)(-0.475)
(AP)Well 1 6.7 psi
The pressure drop caused by Well 2 is:
q2Bu 948ûctr23
(AP)Well 2 = -70.6 kh Ei -
kt
_ - 70.6 (150)(1.35)(2.5) •
(56) (24)
• Ei - (948) (0.16) (2.5) (24.6x10-6) (600) `

(56)77 2
= -26.6 Ei (-0.833)
= (-26.6)(-0.293)
(Ap) Well 2 = 7.8 psi

0
2-97
The pressure drop at Well 3 caused by its own production
can be computed using the log approximation of the
exponential-integral solution (Eq. 2.35): •
q Bu
(AP)Well 3= 162.6 k-^ log k t 2 - 3.23
ûctrw3
162.6 (210) (1. 35) (2.5) •

=
( 56 )
(56) (72) - 3.23

• log
(0. 16) (2 . 5) (24 .6x10- 6 ) (0. 333) 2
(AP)Well 3 = 543 psi
Therefore,
(AP)total at Well 3 = 6.7 + 7.8 + 543 = 557.5 psi •

(pwf)Well 3 = 4735 - 557.5
(pwf)Well 3 = 4178 psia
These calculations are valid only if flow is still

transient after 72 hours.
•
2-98
APC 206 2/01
#1knadarlpt PROJECT COMPUTATIONS

NO. OF SHEETS
DATE I`^ ' ^.' PROJECT JOB NO.
BY CHK.: SUBJECT
_._._. .. . . . .
f O ^j
--- -
a. - .
7 - 0 2. 3
t " o Q
0 ^ L b
2- = f^ 4.41 rw b n 2.
- - - T- r _ r T-- ---I
_ ----- - - - ^ h,
X = 2 ` ,2 . 00 0 5 Z. = . 3 LIs
•
- - 3 -0-^ ._. ._._. _._.
= c S t- c s +
A p_C ^^' 2 .
G 4 ^ 2c^ ^. 2 8 0 ) 3 z
.5 2
= 51 90
- ^
^ = I 2« ^' - 3 2
__- - _
-- __. __ __
•
4. The pressure data obtained from the observation well
^ were used to prepare the plot of (Pi Pobs) versus t
shown in Fig. 2P.1. The data used to prepare this
graph are tabulated below:
t,hrs t,hrs
pi pobs pi pobs
1 0 60 76
2 0 80 81
5 0 100 93
10 3 150 115
15 14 200 132
20 22 300 156
30 35 400 173
40 47 S00 187
50 S7
i • '°
.'., • •
Ln •
•
•
n 10`
0 •
•
.^ •
•
•
10
10 102
103
t,hrs
• Fig. 2P.1: Data curve for interference

test in Problem 4.
2-99•
The data curve was matched with the exponential-integral
type curve to obtain the match depicted by Fig. 2P.2. •
From the match point,
(Ap)m = 100 psi
(PD) M = 0. 76
(t)M = 1000 hrs
(tD/rD)M = 16.0
Reservoir permeability can be computed from these

match values using Eq. 2.77:
k = 141.2qBU (pD)M
Q^ M
h
k = (141.2) (270) (1.21) ( 2.1) 0.76

18 100
k = 40.9 md
Porosity is computed according to Eq. 2.78:
0.0002637k (t)M
uctr2 (tD/rD)M
(0.0002637) (40.9) 1000

(2.1) (11. 1x10- 6)(440) 2 16.0
^ = 0.149
^ = 14.9%
The calculations assume that flow is transient

throughout the test, i.e., there is no effect from •
adjacent wells or reservoir boundaries.
2-100
• • ^
aQ
r.^
10 lot 103 to.

10-1 I
tD/rD
Fig. 2P.2: Type curve match for Problem 4.

APC 206 2/01
#0*nadav4V2 PROJECT COMPUTATIONS

NO. OF SHEETS
• PROJECT -,Â 2 P/-0 ^. LI JOB NO.

DATE 0^//a.^03
BY CHK.: SUBJECT
•
i i i
0
Chapter 3
PRESSURE DRAWDOWN TEST
I. INTRODUCTION
When a new well is placed on production, a pressure

drawdown occurs at the formation face of the well. This
results in the first pressure transient in the life of the
well. It seems logical, therefore that we begin our study
of pressure testing techniques with the pressure drawdown
• test. The drawdown test is not the most commonly used
pressure test; however, the wellbore pressure behavior which
occurs during a constant rate pressure drawdown test must be
fully understood before we can proceed to more commonly used
tests. L In fact, all other pressure transient tests conducted
in the petroleum industry are simply modifications of the
pressure drawdown test.1 This chapter will consider only the
behavior of oil wells producing from reservoirs which are
not fractured; fractured reservoirs will be considered in
Chapter 6, and gas reservoirs will be discussed in Chapter 7.
A drawdown test is a single-well test conducted by
producing a well at constant volumetric rate while continuously
recording formation face pressure as a function of,producing
time. Ideally, the test begins with uniform pressure in the
reservoir. A schematic of the ideal rate and pressure history
before and during a drawdown test is illustrated by Fig. 3.1.
The primary purpose of a drawdown test is to characterize
properties of a reservoir and its contained fluids. Primary
^ information available from a drawdown test is:
.
0
E
q
U
C[.
zw
CF
BEGINNING
cz SHUT-IN .-'ÔF TEST
0 0
w
0 t
TIME
* Pi
U
ts. W
• C V)
F_
(n
.2
Ce
O
LL.
0 t
TIME
Fig. 3.1: Ideal formation face rate and pressure

history for a pressure drawdown test.
Jo, Tâ.t jxo^, ( he ncc es5nrj ,c:wi+i o.Q, ,cLs pusc..w.r,)
• Volumetric average effective permeability
of reservoir to mobile fluids
• Total skin factor
• Flow efficiency (condition ratio) of well
• Drainage volume of well (reservoir limits)
• Detection of faults, and distances to faults
• Detection of fractures, and fracture length
•
3-2
•
The pressure drawdown test is particularly applicable
to:
• New wells
• Wells that have been shut-in sufficiently
long to achieve a reasonably uniform pressure
in the drainage area of the well
• Wells where loss of revenue necessary to run
a shut-in test (pressure buildup test) would
be unacceptable
• Wells in which the primary objective is to
establish reservoir limits
• Wells where equipment problems or reservoir
problems make it unadvisable to run a shut-in
test
Primary disadvantages of the drawdown test are:
• Difficult to maintain constant rate

• Cannot eliminate problems of wellbore storage
associated with production string as is
possible with pressure buildup test
• Problems of "clean-up" associated with new
wells, or wells following workover
It was shown in Chapter 2 that a well placed on production

at constant rate passes chronologically through three flow
regimes: Transient, late transient and pseudosteady state.
Pressure drawdown data measured during these flow periods
exhibit distinctly different behavior, and must be analyzed
using different equations. Late transient data is very
difficult to analyze and will not be considered; a discussion
of this analysis is presented in Ref. 1. The analysis of
transient and pseudosteady state data is presented in this
chapter.
•
3-3
• II. ANALYSIS OF TRANSIENT FLOW DATA
The flowing formation face pressure, pwf, that is -

measured during the transient flow period of an idealized
pressure drawdown test can be described by the logarithmic
approximation to the exponential-integral solution (Eq. 2.35):
pwf = pi - 162.6 kqBll (log kt - 3.23) (3.1)

ûcrt 2w
where: pwf = pressure measured at the formation

face during the drawdown test, psia
pi = stabilized reservoir pressure at the
time the well is placed on production,
psia
q = flow rate during test, STB/day
t = producing time during test, hrs
q k = volumetric average effective permeability
in drainage volume of test well, md
Equation 3.1 can be rewritten as:
pwf = pi - 162.6 Q B 11 (log t + log k 2- 3.23). (3.2)

4ûctrw
When this equation is applied to a particular well, the only

variables are pwf and t; consequently, it simplifies 'under
these conditions to
m log t+ b (3.3)
°^ pwf =
where: m = -162.6 kqBP . (3.4)
•
3-4
•
1. Determination of permeability
It is apparent from the form of Eq. 3.3 that[a plot of
pwf versus t should yield a straight line during the
transient flow period.] This semilog straight line is
illustrated by Fig. 3.2. It should be noted, however, that
this behavior is based on an ideal model; if any of the
assumptions or boundary conditions of this ideal model are
violated by the reservoir being tested, actual pressure
behavior will probably deviate from that predicted by Eq. 3.3.
Also, since Eq. 3.3 is based on transient flow, it should be
expected that this predicted behavior will change after the
pressure disturbance reaches a boundary.
S f,ncu oa.k f w^
av^- (AA- eV 3.21 SLOPE = m, PSI/LOG CYCLE
cv1 G^,^^ r,,,^ ►^ •
z5l"" 4r- •^^.__
• ^
^ fLQ ^^1QiV^-' END OF
U R^ ^ • ^
TRANSIENT FLOW
•
0.1 1 10 102
t, hrs
Fig. 3.2: Ideal semilog plot of pressure

drawdown data.
According to Eq. 3.3, the slope of the semilog

straight line is related to formation flow capacity, kh, by
the relationship
kh = - 162.6 9B,,
m
(3.5) •
3-5
• When formation thickness is known, permeability can be
computed as
k=- 16 2. 6 mh
qB" (3.6)
Example 3.1: Analysis of pressure drawdown test for

formation permeability
Problem. The discovery well in an undersatu rated oil

reservoir produced at a constant rate of 140 STB/D. The
net oil pay in the well is approximately 30 ft. Because
of an underlying salt water zone and a relat ively high
vertical permeability in the sand, only the top 15 feet of
the sand were perforated. Perforation densi ty was 4 shots/ft
with 90° phasing. A pressure drawdown test was run on the
well and the flowing formation face pressure was recorded as
shown. The static reservoir pressure measur ed prior to the
test was 5000 psia. Other reservoirs in the area experience
no water influx; hence, it may be assumed th at the reservoir
is bounded with a no-flow boundary. It is r equired that
the formation permeability be estimated from this data.
t, hours pwf,psia t, hours pwf,psia
1.0 4617.5 7.0 4584.4

1.25 4611.0 8.0 4580.6
1.50 4606.1 9.0 4580.0
1.75 4605.0 10.0 4578.1
2.0 4603.1 11.0 4575.8
2.5 4598.0 12.0 4574.0
3.0 4597.0 13.0 4572.1
3.5 4593.2 14.0 4570.5
4.0 4592.1 15.0 4568.0
^ 5.0 4590.2 16.0 4566.3
6.0 4584.3 17.0 4565.1
3-6
18.0 4563.2 24.0 4556.5
•
19.0 4562.3 26.0 4554.3
20.0 4561.0 28.0 4553.0
22.0 4558.6 30.0 4552.1
B 0 = 1.75 RB/STB
uo = 0.65 cp
so = 69%
S = 31%
w
11%
c o = 11.1 x 10
c w = 3.0 x 10-6
C f = 4.6 x 10-6
Solution. The pressure drawdown data are plotted in Fig. 3.3.

It is noted that transient flow lasts for about 10 hours.
The slope of the straight line is -36 psi/log cycle. The
formation capacity is calculated using Eq. 3.5:
162.6quB
kh m
(162.6) (140)6(0. 65) (1.75) = 719 md-ft.

kh
Thus, the average effective permeability of the reservoir

to the flow of oil is:
k = 719md-ft = 24 md.
30 ft
. 1
2. Boundary Effects
The straight line relationship predicted by Eq. 3.1 is
only valid for data recorded during transient flow. After
a well has produced sufficiently long for its drainage
3-7
• • •
plhr = 4614 PSIA
4620
SLOPE _ -36 PSI/CYCLF.

{
4600
1tmnfittrtl1rruttittt t r 1-' 1 r! I I I I I I I I ttttrt -1 I t t r t tt 11-ttt

APPROXIMATE END OF TRANSIENT FLOW
4580
^.
00 45fi0
4540
z 4 b 8 10 20 30 40
t, hours
Iîg. 3.3: Semilog plot of pressure drawdown data, Ex. 3.1.

yarncit^ un,u J.cTU.c-h'o-r,,
eafflhQ.&. (.w c'toO,) Gt^ , d ►net^
tllt fi^ 0./'LL Q. ^J"/hl. (iuC.O-
('ÛS+Gl,u C..&.. & ^t3LlVA".
boundary to affect the wellbore pressure, late transient

•
flow begins and it can no longer be expected that measured
formation face pressures will plot as a semilog straight
line. The effect of a boundary on the semilog plot is
illustrated by Fig. 3.2; wellbore pressures drop faster
under the influence o^f a bound4ry than during transient flow.
A..mi. -to e-^ FYawS+ r U^
The duration of transient flow can be estimated3 as
nj^
Ou c A
a^,,^ t etf = t 3^ (tDA)etf
3'7)
(3.7)
• fl ^ ^ µ , C - aleEtnvwtnt 4taw ^vrt- • ^(^Sâr, ^c. "_(xu st,4ed `t^n auc^G. W *

where: = time to the end of the transient
tetf
flow period, hrs
(t ) = dimensionless time to the end of
DA etf
the transient flow period."
Equation 3.7 can be used to predict the end of transient

flow in a drainage system of any geometry if the appropriate
value of (tDA)etf is obtained from the "Use Infinite
System Solution With Less Than 1% Error for tDA <" column
of Table 2.2. For a well centered in a circular reservoir,
(tDA)etf - 0•1; accordingly, Eq. 3.7 reduces to
380ûctA
(3. 8)
^ tetf = __T-
The boundary being referred to is the drainage boundary

of the test well. If the test well is the only producing
well in the reservoir, the drainage boundary is the physical
reservoir boundary. However, if other wells are producing
from the same formation, a boundary effect can be caused by
pressure interference from adjacent wells. This is illus-
trated by Fig. 3.4 which shows three wells producing at
equal rates from a homogeneous, constant thickness reservoir.
Superimposed on this schematic is a plot of reservoir
pressure along a line connecting the wells.
Consider, for example, the effect of Well 1 upon Well
2. As one moves away from Well 2 toward Well 1, the
3-9
• Well 1 Well 2 Well 3
,^ ^, . .\y ^
a^ =0^'• ^
a =0
ar
I
A B
Fig. 3.4 Schematic illustrating the occurence of

no-flow boundaries between producing wells.
reservoir pressure increases until it reaches a maximum at

point A, which represents the mid-point between the wells.
To the right of point A, pressure decreases toward Well 2
and, consequently, fluid in that part of the reservoir flows
toward Well 2. To the left of point A, pressure decreases
toward Well 1 and flow is toward Well 1. Thus, point A
represents a position in the reservoir where fluids divide
and flow toward Wells 1 or 2 depending upon the direction
of the pressure gradient. Note, however, that no fluid can
cross the dividing line denoted by A because it represents
the position of maximum pressure and [zero pressure gradientil
accordingly, it is a no-flow boundary which has the same
effect on fluid flow that an impermeable barrier would have.
Other wells adjacent to Well 2 will also result in an
interference boundary. For example, Well 3 causes a point
of "no-flow" to develop at position B. It is obvious,
therefore, that every well in a developed reservoir will
have a drainage boundary caused by the presence of other wells.
The shape of an interference, or pressure, boundary
depends upon the well pattern, the relative producing rates
of the wells, and reservoir properties. Consider for example,
3-10
a square well pattern in a constant thickness, homogeneous
•
reservoir where all wells are producing at equal rates. The
point of interference between adjacent producers will, in
this ideal system, occur half-way between the wells. Accordingly,
each well will be surrounded by a boundary in the shape of
a square. This is illustrated by Fig. 3.5. If a pressure
drawdown test was run in one of these wells, it would experience
transient flow until the pressure wave hit the drainage boundary
shown; after transient flow ended, pressures would be dictated
by the volume and shape of the square drainage area.
0 O O
NO-FLOW BOUNDARY
CAUSED BY ADJACENT
/ ^PRODUCE RS
♦
<
1]
O O
/
/
/
/
/ \
0
i
\ /
\ /
\ /
O O
Fig. 3.5: Schematic of the ideal no-flow interference

boundary in a square well pattern.
•
3-11
• Many different drainage patterns result from the
various development patterns encountered in petroleum
operations. Drainage geometries are further altered by
reservoir heterogeneities. Also, the drainage pattern
of a particular well will change if its producing rate
relative to adjacent producers, or vice-versa, is altered.
A knowledge of the drainage geometry is very important because,
as we will see in the next section, the pseudosteady state
pressure behavior of a well is dependent upon the size and
geometry of the area being drained.
III. ANALYSIS OF PSEUDOSTEADY STATE DATA
If a pressure drawdown test is run long enough for

pressures to stabilize throughout the drainage volume of the
test well, the formation face pressure can be described by
E q. 2.49, i. e., (^ aNUa oaem r^ ^^^ ►^c e L' c^ c,D L^''ê
^x
_ 162.6 Bu 4A 0.2339 Bt (3.9)
PSS pwf = pi W _ Ahoct
log [1.78Arj
where A is the drainage area of the test well, and CA is a

shape factor which is available from Table 2.2. Recognizing
that pwf and t are the only variables in Eq. 3.9, this
expression can be rewritten as
pwf = m't + b'. (3.10)
1. Reservoir limits test

^ Equation 3.10 suggests that a Cartesian plot of pwf
versus t will trace a straight line for pressures measured
during the pseudosteady state flow period. This is
3-12
illustrated by Fig. 3.6.
I^SS ^ e^t-c0ê^.
• End of transient
flow
•
• Beginning of
• pseudosteady
state low
b^ ^
c • •
Tranaient I
flow • Slope = m', psi/hr
°- f
^`
Late
transient I Pseudosteadv
flow state flow
, I
t, hrs ^
Fig. 3.6: Cartesian plot of pressure drawdown data.
Moreover, Eq. 3.10 indicates that the slope of the straight

line, m', is equal to
m, 0.2339q B (3.11)
Ah^ct
Accordingly, the drainage pore volume of the test well

can be computed as
.
Vp ` Ah^ 0'm3^9qB ft3. (3.12)

t
If h and ^ are known, it is obvious that Eq. 3.12 can be

solved for the drainage area of the test well.
Equation 3.12 is very important because it means the
pressure drawdown test can potentially be used to
3-13
^ ' î.v ►^,,^.-^S -^.¢.^-^
• ^
determine reservoir volume. A drawdown test run for this
purpose is commonly referred to as a reservoir limits test2.
2. Estimation of drainage shape

When pressure data are available from both transient and
pseudosteady state flow periods, it is possible to estimate
the shape of the area being drained by the test well.
Earlougher3''` has shown that the reservoir shape factor can
be estimated from
2.303 (p -
CA = 5.456 m, exp lmr Pint) (3. 13)
where: m = slope of semilog straight line, psi/log cycle

m' = slope of cartesian straight line, psi/hr
plhr = pressure at t = 1 hr from semilog straight
^ line, psia
pint = pressure at t = 0 from cartesian straight
line, psia
With CA known, Table 2.2 can be used to determine the drainage

configuration which has a shape factor in closest agreement
with the calculated value.
This is a rough and dirty method to determine the
drainage shape associated with the test well. However, if
the test well is an exploration well, this information
combined with geological information could give valuable
guidance in the selection of locations for step-out wells.
Example 3.2: Analysis of pressure drawdown test for drainage

volume and drainage configuration.
Problem. The well described in Ex. 3.1 was tested for 60

hours. Fhe pressure data for the first 30 hours were
^ recorded in that example. The remaining data are given here.
From the results of this pressure drawdown test, compute the
drainage area and drainage confi.guration associated with this
well.
3-14
•
t, hrs pwf, psia
32 4552.3
34 4550.7
36 4550.0
38 4549.2
40 4548.6
42 4548.5
44 4548.0
46 4547.2
48 4546.4
50 4546.2
52 4545.5
54 4545.0
56 4544.1
58 4544.0
60 4543.0
0
Solution. A Cartesian plot of the test data is presented in

Fig. 3.7. It is significant to note that the pressure data
become linear after about 30 hours of producing time. This
represents the start of pseudosteady state. The slope of
the Cartesian straight line is - 0.283 psi/hour. Also, the
total compressibility is:
ct = Soco + S w c w + cf
ct = (0.69)(ll.lxl0-6) + (0.31)(3.Ox10-6) + (4.6xl0 6)
ct = 13.2 X 10-6 psi -1
The drainage volume of the well is estimated from the pressure

data according to Eq. -).12, i.e.,
0.2339qB
J Ah^ =
m'ct
f 3-15
• • •
462
END OF
TRANSIENT FLOW
460
CTJ 458 START OF

.r.,
EA PSEUDOSTEADY STATE
GZ.
m' = -0.283 PSI

456
Pint - 4560 PSIA
4S4
4520r r_r__rTT-rm rrt-r-r-r-n r r-FTTTTrT-r-l r

0 10 20 30 40 50 60 70
t, hrs
Fig. 3.7: Cartesian plot of pressure drawdown data, Ex. 3.2.

Ah^ _
(0.2339)(140)(1.75)
•
(-0.283)(13.2x10-6)
Ah^ = 1.53 x 10' ft3
If the porosity and thickness are assumed constant, the

drainage area of the well is estimated to be
A = 1.53 x 10'
(30) ( 0.11 )
A = 4.64 x 106 ft2
A = 107 acres (1 acre = 43,560 ft2)
Thus, it appears that this reservoir is rather limited in

size.
From Figs. 3.3 and 3.7,
m = -36 psi/log cycle

•
plhr = 4614 psia
m' = -0.283 psi/hr
pint = 4560 psia
The reservoir shape factor is computed using Eq. 3.13, i.e.,
2 305(plhr pint)
C 5.456 m exp
A = m' m
CA = 5.456 [o.283j exp C (2.303)( 4614-4560)

-36
Ca = 21.9
Comparing this value to shape factors in Table 2.2, two

drainage configurations are possible: A well centered in
a 2•1 rectangle has a shape factor of 21.8 whereas a well
in an isosceles right triangle has a shape factor of 21.9. •
3-17
• More information about reservoir geology would be necessary
to decide which configuration is most likely.
St.c p 50 0 °^
fah- gfe,^ vralnl1.a'f,
3. Radius of drainage - .i
.cm^t^^ o o` ►
Many engineers use the concept of a radius of drainage
to aid in understanding and analyzing pressure transient
tests. Unfortunately, this is a term which is widely mis-
understood and, consequently, is often misused. Concep-
tually, the radius of drainage represents the distance which
a pressure disturbance has moved into a formation following
a rate change at a well. In practice, however, one must
decide what constitutes a pressure disturbance before its.
movement can be described.
When a well starts producing, a pressure disturbance
0 is created at the formation face; this disturbance moves

through the reservoir at the speed of sound. Theoretically,
we would expect the pressure throughout the entire drainage
area to change almost instantaneously as a result of this
production. Practically, it is found that the pressure
disturbance attenuates rapidly because of the system com-
pressibility; thus, after a given production time there is
some distance, the radius of drainage, beyond which there
has been no significant pressure change. The problem in
defining the drainage radius is to decide what constitutes
a significant pressure change. It is obvious that no
precise definition of drainage radius exists.
A variety of equations have been presented in the
literatures '6'''8 defining the radius of drainage. In SPE
Monograph V3, and in this text, it defines the radius of
a circular system within which pseudosteady state conditions
could develop within the producing time, t. Mathematically,
• rd = 0.029 u
[ct]
Z
(3.14)
3-18
According to Eq. 3.14, the drainage radius continues to
•
increase as t increases. However, the drainage radius
cannot extend beyond the drainage boundary of the well;
thus, Eq. 3.14 only applies until the time the pressure
disturbance reaches the drainage boundary.
Example 3.3: Calculation of radius of drainage for a

pressure drawdown test.
Problem. A 24-hour pressure drawdown test is run in an oil

reservoir with the following properties:
k = 10 md
= lSo
u = 2.7 cp
ct = 20 x 10-6psi
This well is completed on 40-acre spacing in a reservoir

•
with a square well pattern. When this data is plotted on
a semilog graph, should there be any boundary effects?
Explain.
Solution. If the well is centered within its square drainage

area, the distance to the nearest boundary should be
approximately 660 ft. The radius of drainage of the draw-
down test is, according to Eq. 3.14,
(10) (24) z
rd = 0.029
(015) (2.7) (20xl0 6)
0
rd = 158 ft.
Since rd < 660 ft, all of the drawdown data represent

transient flow and the semilog plot should not exhibit any
boundary effects.
•
3-19
PSS
-tAA_ aA-^C
h QaC^
.7 4. Stabilization time
Stabilization time has also been defined many ways in
the literature5. The most commonly used definition, and
the one used here, isCthe time required for pseudosteady
state flow to begin ^ An estimate of this time is
ûctA 2
t = t ) (3.15)
pss 0.000 263 c( DA pss
^tump cUa, wl^- 3•^)

where tpss = time required to reach pseudosteady
state, hrs I
(t DA)pss = dimensionless stabilization time
The dimensionless time at which pseudosteady state begins

has been determined theoretically for various drainage
shapes, and is presented in Table 2.2 under the column P 2-Ll`7
heading, "Exact for tDA > ". For a well in the center o^
a circle, as well as most other symmetrical drainage con-
• figurations, (tDA)pss = 0.1; accordingly,
380ûc A
(3.16)
pss
Equation 3.16 represents the minimum time to reach

pseudosteady state, since this condition is attained in a
circular system quicker than in any other drainage shape.
For non-symmetrical drainage configurations, or for layered
reservoirs, this time will be longer.
Example 3.4: Time required to achieve pseudosteady state

flow.
Problem. Calculate the time required to achieve pseudosteady

state conditions for the well described in Ex. 3.3.
^ Solution. From Table 2.2, (tDa)pss = 0.1 for a well in the

center of a square. Consequently, from Eq. 3.16,
3- 20
= (380) (0.15) (2. 7) (20x10 6) (40) (43 , 560) ^
tpss 10
t = 536 hrs
pss
^ S. Practical limitations of reservoir limits test

When a new reservoir is discovered, the first question
asked an engineer is: How large is the reservoir? If this
question could be answered upon completion of the discovery
well, it would be much easier to make decisions concerning
the drilling of additional wells. Often a reservoir can be
drained by one well; if this were known upon completion of
that well, the economic advantages are obvious. Therefore,
it is easy to get excited about a test which potentially
offers this kind of information.
While the pressure drawdown test offers the potential
of measuring reservoir volume, it is subject to several
•
practical limitations:
1) The test must be run sufficientl long to reach

pseudosteady state t roug out the drainage
volume of the test well. The minimum time
required to accomplish this can be estimated
using Eq. 3.16:
38ftuctA
(3.16)
(tpss)min ^
If the drainage shape is asymmetrical, the well

is off-center, or the reservoir is heterogeneous,
the time will be greater than predicted by Eq.
3.16.
The variables which have the greatest effect

on stabilization time are A and k. If the
drainage area is large, or if permeability is
small, the time required to run the test will be
prohibitive. If the variables in Eq. 3.16 can
be estimated, an engineer can check beforehand
if a reservoir limits test in a particular
q
reservoir is feasible.
3-21
• 2) The test often does not work well in large
drainage areas. The rate of pressure
change during pseudosteady state flow is
inversely proportional to the drainage pore
volume; consequently, the rate of pressure
change will be small and more difficult to
measure and interpret. Also, test times
required to achieve pseudosteady state flow
in large drainage areas will generally be
prohibitive.
3) It is difficult to maintain a constant rate

A.a&
during the long test time generally required
by a reservoir limits test.
4) The test will give erroneous results in any

reservoir with an active water drive. The
water drive maintains pressure at an
abnormally high level; this results in a slope
on the Cartesian plot which is too small, and
a calculated pore volume which is too large.
5) The calculated pore volume, according to Eq. 3.12,

is inversely proportional to the total reservoir
compressibility, and will be no more accurate than
^ our knowledge of the total compressibility. When
the reservoir contains free gas, the total
compressibility will be dominated by the gas
YhunA+ ^vt.aw +ut^ compressibility. Since we seldom know the gas
saturation, large errors can be made in determining
^^3t,cSS<^ji^"^ the pore volume of this type reservoir. Further,
when testing unconsolidated reservoirs, formation
compressibility can be large; this is apparent
from the data presented in Fig. A.6. Since
accurate values of compressibility are seldom
available for this type rock, reservoir limits
tests in unconsolidated reservoirs can be
difficult to interpret.
IV. EVALUATION OF FORMATION DAMAGE
[ One of the most useful applications of the pressure

drawdown test is to diagnose a well for possible formation
damage. lIt is not unusual during drilling, completion or
workoverJ operations, for materials such as mud filtrate
or clay particles to enter the formation and cause an
annulus of reduced formation permeability around
3-22
•
the wellbore. This region of altered permeability, commonly
referred to as a skin zone, can extend from a few inches to
several feet from the welibore. Many other wells are
stimulated by acidization or hydraulic fracturing; the skin
zone in this case effectively represents an area of increased
permeability near the wellbore. A schematic of a skin zone
is shown in Fig. 3.8.
SKIN UNDANLAGED
ZO\E ZONE
•
ti
Fig. 3.8: Schematic of skin zone near the wellbore.
The effect of formation damage is to alter the pressure

distribution around the wellbore; in particular, it causes
an additional pressure loss in the formation which was not
accounted for in the equations previously used to analyze
pressure drawdown data. This additional pressure loss, Aps,
All of the flow equations we have studied show that the
flow rate of a well is directly proportional to the pressure
drawdown in the well. It follows, therefore, that damage will
result in a decrease in production rate that is directly
related to the skin pressure loss, Aps. LThe objective of
this section is to show how a pressure drawdown test can be
used to ( 1) detect formation damage, (2) determine how much
3-23
APC 206 2/01
#14nadarApt
PROJECT COMPUTATIONS
• DATE 6(nÎ Q 103 PROJECT

NO.
JOB NO.
OF SHEETS
BY ^^ CHK.:. SUBJECT F92v►^Q^tCSL^ ^VvuGtq^
0
•
0
•
(pwf) Ideal
Damage
r r
w s
Radius
Fig. 3.9: Comparison of near-wellbore pressure

• distribution for a damaged and
undamaged well.
pressure loss is being caused by damage, and (3) determine

how much loss in production is being caused by damage. Once
the production loss due to damage is known, it can be determined
whether a stimulation treatment is economically viable.
1. Modification of flow equations for a skin zone

An approximate method to determine the skin pressure
loss was suggested by Hawkins9. If it is assumed that the
permeability in the skin zone, ks, is uniform, the additional
a pressure loss across the zone can be approximated by the
steady state radial flow equation presented previously as
Eq. 2.7.
r r
Aps = 141.2 k ln rs - 141.2 k qBp in rs (3. 17)
Ti -
s w w
•
3-24
or, r
C
dps = 141.2 kh^ k ln rs (3.18)
s w
It was shown previously in this chapter that the ideal

pressure change which occurs at the formation face of a well
during the transient flow period of a pressure drawdown test
is, from Eq. 3.1,
(log kt - 3.23)
(pi-pwf)IDEAL ° 162.6 k
Oucrt 2w
This equation assumes that a uniform permeability, k,

exists throughout the drainage area of the well. When an
altered zone of permeability ks and radius rs exists adjacent
to the wellbore, the total pressure drawdown will be increased
by an amount aps, i.e.,
pi-pwf (pi pwf)IDEAL + 'ps

J
or,
pi pwf = 162.6 k L log kt 2- 3.23^

OUctrw
r
+ 141.2 kh
qBP Ik - in rs
s w
pi pwf = 162.6 ^hu ^log kt - 3.23

L ûcrt 2w
. 1
r
+ 0.87 (k - 1) in rs (3.19)
s w
It is helpful at this point to define a skin factor,

s, such that
•
3- 25
^ r r
s= I k - 1^ in rs (3. 20)
L s w
With this definition, the skin pressure loss can be

written as
Aps = 141.2 khu S. (3. 21)
Further, the pressure drawdown equation, Eq. 3.19, becomes
pwf = pi - 162.6 ^ (log kt - 3.23 + 0.87s) (3.22)

$ucr
t 2w
Similarly, the pseudosteady state equation, Eq. 2.49, can

also be modified to include the skin factor:
^ p = p. - 162.6 ^ (log 4A + 0.87s)

wf 1
1.781CArw
_ 0.2339qBt (3. 23)

Ah^ct
What is the skin factor? Observe from Eq. 3.21

that
An
S _ . s _ (3.24)
141 .2qBv (Aps D
TE
It is obvious from this relationship that the skin factor

is the pressure loss due to skin expressed in dimensionZess
form. We can gain further insight into s by studying Eq.
3.20. If a well is damaged (ks < k), s will be a positive
number; further, the magnitude of s increases as the
contrast between k and ks increases, and as the depth of
^ damage (rs) increases. There is no upper limit on the
magnitude of s. If a well is stimulated (ks > k), s will
3-26
be a negative number. As the depth of stimulation
•
increases, the numerical value of s increases; however,
it is significant to note that a stimulated well rarely
has a skin factor less than -8, and this will occur only
in wells which have highly conductive fractures that
penetrate deep into the reservoir.
It is observed from Eq. 3.21 that a negative skin
factor will result in a negative value of Aps. This
simply means that a stimulated well will require Zess
pressure drawdown to produce at rate q than an equivalent
well with uniform permeability. A comparison of pressure
distributions near the wellbore for positive and negative
skin factors is illustrated by Fig. 3.10.
!J 5
Radius
Fig. 3.10: Effect of skin factor on reservoir

pressure distribution.
2. Effective wellbore.radius
all flow equations in 0
As an alternative to writing
terms of the skin factor, the skin zone around a well can
3-27
• be accounted for by defining an effective wellbore radius,
rwa, as
-S
r wa = r w e (3. 2S)
Conceptually, a well of radius rwa producing from an

unaltered formation would have the same drawdown as a well
of radius rw producing from an altered formation with a
skin factor s. This can be shown mathematically by con-
sidering Eq. 3.22
kt -
pwf = pi - 162.6 k qBP (log 3.23 + 0.87s)
z
0uctw
r
kt -
pwf = pi - 162.6 khu ( log 3.23)
Wtrwezs
• kt
pwf - pi - 162.6 khu (log - 3.23)
ûctrwa
All of the ideal radial flow equations presented in Chapter

2 can be modified for skin simply by substituting rwa for
r
w
3. Effect of skin on the semilog plot

Introduction of the skin factor into the flow equations
does not alter the previously described procedures used to
evaluate formation permeability and drainage pore volume.
The primary effect of skin is to shift all measured pressures
by an amount equal to Aps from the values which would have
been recorded in a reservoir of uniform permeability. There
• can also be a distortion of the early-time data because of

the permeability change near the wellbore; this effect is
3-28
often not apparent, however, either because it happens so
•
quickly, or because it is masked by other early-time
problems such as welibore storage. Figure 3.11 illustrates
the ideal effects of both negative and positive skin factors
on the semilog plot of pressure drawdown data.
SLOPE = m
's
START OF
BOUNDARY
/EFFECTS
G.•
^ps •I ``
•
^
0.1 1 10 10Z
t
Fig. 3.11: Effect of skin on semilog plot of

drawdown data.
Figure 3.11 indicates that the semilog plot will have

the same slope, m, at late times regardless of the value
of s. Since permeability is related to m by the relation-
ship
k = - 162.6 q
mh
this implies that the average reservoir permeability computed

from a pressure drawdown test will be the same regardless
of the skin factor. How can this be true when different
skin factors are reflecting different permeability
3-29
•
distributions? Recall that the slope of the semilog plot at

any time is a measure of the voZumetric average permeability
within the drainage radius of the well at that time. Referring
to Fig. 3.8, the volumetric average permeability can be defined
as
ksflVs + kAVud
(3.26)
AVs + AVud
where: AVs = volume of skin zone
AVud = volume of undamaged zone
The volume of the skin zone, AVE, is generally small;

however, AVud increases in proportion to the square of
the drainage radius. It is obvious, therefore, that soon
after a test begins, Eq. 3.26 reduces to
k = k
Practically, this means that, regardless of the value of

s, the slope of the pressure drawdown curve yields the
permeability of the unaltered formation.
Finally, it should be made clear that a skin zone near
a well affects only the pressure near that well, i.e., pressure
in the unaltered formation away from the wellbore will
not be affected by the skin zone. This means that any
equation used to compute the flowing formation pressure,
pwf, of a well with a skin zone should include the skin
• factor. Any equation used to compute pressure at a
location in the reservoir not affected by a skin zone, should
not include the skin factor.
4. Calculation of total skin factor

It was shown previously that the skin factor could be
computed in terms of ks and rs using Eq. 3.20. While
3- 30
Eq. 3.20 helps us gain insight into the meaning of the skin
•
factor, the reader is cautioned that this is a steady state
relationship and should not be used for quantitative
purposes. Instead, we will show that the skin factor can be
evaluated using measured drawdown pressures.
The skin factor can be calculated from transient flow
data by rearranging Eq. 3.22:
pi-pwf - log kt (3.27)

s = 1.151 + 3.23
16 ûctw
or, _
s= 1.151 pwf pi - lo kt + 3.23 (3.28)
M g ûctrW
Recall that pi is the stabil.ized pressure measured immedi-

ately before the test well is placed on production. The
value of(pwf is dictated by the value of t used in the
equation^; note, however, that pwf must be obtained from
the semiZog straight line. Traditionally, pwf is selected
at t = 1 hour (i.e., pwf plhr) and Eq. 3.29 becomes
plhr pi - log k + 3.23 . ( 3.29)

s = 1.1751
M
ûctrw
`-^ ►^^q0.^'iuC. ^ .
The proper determination of plhris illustrated by Fig.

3.12.
Once the skin factor is known, the pressure loss due
i 1
to skin can be computed using Eq. 3.21, i.e.,
aps = 141.2 ^ s (3.21)
or the equivalent form
0
Aps = -0.87ms. (3.30)
3-31.
•
•
• SLOPE = m
^^ •
^` ••
Plhr -------^^,^
0.1 1 10 100
t, hrs
Fig. 3.12: Determination of Plhr
• Example 3.5: Calculation of total skin factor and the

pressure loss due to skin.
Problem. Consider again the well described in Ex. 3.1.

Determine (a) the total skin factor for the well and (b)
the pressure loss caused by skin.
Solution. It was determined in Ex. 3.1 that
m = - 36 psi/cycle
plhr = 4614 psia
k = 24 md.
Other data from this well are:
= 0.11
uo = 0.65 cp
ct = 13.2x10-6p si 1
^
rw = 0.33 ft
3-32
•
(a) The skin factor is computed using Eq. 3.29:
s = 1.151 plhr pi log k + 3.23

m z
ûctrw
24
s = 1.151 4614-5000 - log
-36
(0.11) (0.65) (13.2x106) (. 33) 2
+ 3.231
s = 1.151 (10.72 - 8.37 + 3.23)
s =6.4. ok{. S (<i n
Therefore, the total skin factor indicates that this well is

poss^,b Zy damaged.
t_?), The ,Î'C'sSilI't' Ictiti L:Le t0 ^k 1n Ls CoT'.1r71:ted L'isli7(^ ^'.q . J.30:
Gps = - 0.87ms
pps = - (0.87) (-36) (6.4)
Lps = 200 psi
moyC. PY"5S ClmwdowV\ ti"av% k oleat wd

This means that measured formation face pressures are 200
psi less than they would be in the absence of skin.
5. Calculation of damage skin factor

When the concept of the skin factor was initially
published, it was reported that a positive value of s
indicates formation damage, whereas a negative value reflects
formation stimulation. Unfortunately, this is an over-
simplified interpretation of the skin factor which is still
3-33
R'e"mL' a.( ^, tk.e, c^^ rr ^z
-^, " tid..A.a.L c,-tdV _ c;n►^rP,v-e ^ .
• adhered to by many engineers, and results each year in

numerous incorrect engineering decisions.
If a positive value of s is calculated using Eq. 3.29,
this does not necessarily mean that the formation is
damaged. An^ fZo^v restriction near the wellbore of the test
well, which is not accounted for in the ideal radial flow
theory used to derive Eq. 3.29, will result in a calculated
skin factor greater than zero. Examples of other restrictions,
which can cause a positive skin are turbulence, perforations,
partial penetration and gas blockage. The total skin factor, Ac^'
therefore, is in reality a composite of several skin factors, °-r`(
i.e., ^&-nn
do-"n' a^
s= sd + sr + Sp + (3.31)
S t + Sf + Ssw
where
^ Sd = skin due to permeability alteration;
this includes formation damage
Sr = skin due to restricted entry
sp = skin due to perforations
St = skin due to turbulence
Sf = skin due to fracture
s sw = skin due to slanted well
If the purpose of a pressure test is to determine if

a formation is damaged, it is clear from Eq. 3.31 that sd,
not s, must be evaluated. The total skin, s, can be a
large positive number while sd is zero or negative. This
fact is often misunderstood and has resulted in many
unnecessary stimulation treatments.
The estimation of sd requires that other skin factors
in Eq. 3.31 be known; then,
sd = s - sr - sp - S t - s f - SSW (3.32)
0
3-34
The right-hand-side of Eq. 3.32 can be difficult to evaluate,
•
but methods are available to estimate these quantities.
A. Skin factor due to restricted entry, sr

The basic equations we have been using to analyze
pressure drawdown data assume that the entire productive
interval is open to the wellbore, i.e., that we have an
open-hole completion. For many reasons, however, it is
common to test wells where only a fraction of the productive
interval is open. This restricted entry to flow, depicted by
Fig. 3.13, causes an additional pressure loss which shows up
in the calculations as a positive skin effect.
h
0
•
hp^^
OPEti
h INTERVAL
P
1
r
e
Fig. 3.13: Formation with restricted entry.
Studies by Brons and Marting12 and Odeh13 have shown

that the magnitude of pressure loss caused by restricted
entry depends upon the fraction of the formation open to
the wellbore, the position of the open interval, net
formation thickness, and the relationship between vertical
•
3-35
E and horizontal permeability. These variables have been
combined mathematically by Odeh14 into a predictive equation
for s :
r
0.825
Sr = 1.35 1 h - 11 {ln (h kH/kv + 7)
P J
- [0.49 + 0.1 in (h kH/kv)] ln rwc - 1.95} (3. 33) o
where: Sr = skin factor due to restricted entry

hp = length of open interval, ft
kH = horizontal permeability, md
kv = vertical permeability, md
rwc= corrected wellbore radius, ft
0.2126 (z /h + 2.753)
= rwe m for h 0 > 0,
S
and rwc = rw for ho = 0
ho = distance between the top of the formation

and the top of the open interval, ft
zm = distance between the top of the formation
and the middle of the open interval, ft
The relationship between ho, hp, and zm is
zm = h 0 + hp/2
A partially opened interval can sometimes affect the

shape of the semilog plot. Kazemi and Sethls showed that
two straight lines may occur where the slope of the first
line represents kh of the open interval, and the second line
represents kh of the entire formation. This effect is
illustrated by Fig. 3.14. Recognizing that this character-
istic can occur in wells with limited entry can greatly
assist in the interpretation of field data; this is clearly
3- 36
•
EARLY TIME kh
REPRESENTS OPEN
INTERVAL
LATE TIME kh
REPRESENTS
ENTIRE
IINTERVAL
•
0.1 1 10 102
t, hrs
Fig. 3.14: Ideal behavior of a well with a

partially opened interval.
illustrated in a case history16 of pressure buildup analysis •

of Prudhoe Bay wells. It should be noted, however, that the
first straight line may not appear because of early-time
effects such as wellbore storage. Wellbore storage will be
discussed later in this chapter.
B. Skin factor due to perforations, sp

The basic drawdown equations assume an open hole
completion. When production is through perforations, the
perforation represents a small flow area which can restrict
flow and cause a positive skin factor. At the same time,
the perforation can penetrate a sufficient distance into the 9
formation to cause a stimulation effect. Thus, this skin

factor can be positive, negative or zero.
Studies l' 21 of productivity from perforated completions
indicate that the perforation skin factor depends primarily
upon the following variables:
•
3-37
^
^ perforation diameter
• depth of penetration
• shot density
• perforation pattern
• ratio of vertical to horizontal permeability
• radius of damage zone
Perforation penetration and shot density have the greatest

effect on productivity.
Correlations are available which present sp as a function
of the stated variables18'2o'21 Unfortunately, these
correlations are not very useful for our purpose because
much of the needed information is generally not available.
The correlations do show, however, that the value of sp
which results from normal perforation practices (reasonable
shot density, penetration, etc.) is generally close to zero;
^ this is because the stimulation effect of the perforation
channels tends to offset the restrictive effect of the small
flow area.
Practically, it is recommended that sp be neglected in
wells which produce at relatively low rates (most wells in
the United States fall into this category) unless unusual
perforating conditions exist. Additional care must be taken,
however, when working with high rate wells; even though the
skin factor associated with a high rate well may be small,
Eq. 3.21 shows that the pressure loss due to skin can be
large. Perforations are often the primary factor which
limits the productivity of high rate wells. A detailed study
of this problem is beyond the scope of this manual.
C. Skin due to turbulence

Due to the limitations of Darcy's law, the basic draw-
down equations assume that flow is laminar. At equivalent
flow rates, turbulent flow causes more pressure loss than
^ laminar flow; this additional pressure drop exhibits itself
as a positive skin. Further, turbulence increases with
3-38
n 1 ^^
V-1
increasing rate so that the skin due to turbulence is a
•
function of rate.
Turbulence is seldom a problem in oil wells. However,
turbulence is common in gas wells, particularly in low
permeability reservoirs. M ^thods of evaluating the
turbulence skin factor will be discussed in Chapter 7; these
methods can also be applied to oil wells.
D. Skin due to fractures, sf

The effect of a fracture is to stimulate a well; the
magnitude of stimulation will depend primarily upon fracture
length and fracture flow capacity. The presence of a
fracture near a test well will cause a negative skin factor.
It is possible when dealing with fractured systems to
calculate a negative total skin factor, while having
significant formation damage. This can be very misleading
if an engineer does not fully understand the concept of
skin factors.
Fractures not only exhibit their influence in the form
of a skin factor, but they alter the flow behavior of a
well. Consequently, an entire chapter, Chapter 6 will deal
with the analysis of fractured reservoirs.
E. Skin due to slanted-well , ssw

When a well penetrates a formation at an angle, the
amount of formation exposed to the wellbore is greater than
predicted by the basic drawdown equation. A correlation22
relating sSw to the angle a which the well makes with the
bedding plane of the formation, is presented in Fig. 3.15.
(This correlation assumes that horizontal and vertical
permeabilities are equal.) When horizontal and vertical
permeabilities are unequal, the following equation can be
used 22:
ssw (a,/41)2.06
- ( a'/56)1.865 log (hD/100) (3 ;4
• ^
3-39
•
V ^3 0
moo
-2 Jo
Â _^ Soo
-s
-6 ^Jo
4 6 • 2 3 4 6 • 2 3 4 6 •
I OY 103 104
h/r w
Fig. 3.15: Skin factor for a slanted

well. After Ref. 22.
('-bo 0;,n.pASfi,c tw ^Xa.c;c^Wz VV.,2)
where: k
a' = tan kv tan a
H
hD = ^ (kH/kv)0.5
w
F. Other skin effects

Any near-wellbore effect which was not accounted for
in the ideal mathematical model that serves as the basis
for pressure drawdown analysis can result in a skin effect.
In addition to the factors already discussed, positive
3-40
q
skin effects can result from the buildup of solution gas
near the wellbore in oil reservoirs, liquid buildup in
gas reservoirs, gravel packs, formation compaction near
the wellbore in unconsolidated formations, etc.
It is generally not possible to predict the skin effects
caused by all non-idealities in a well; accordingly, it will
not always be possible to compute an exact value of the
damage skin factor. However, even in these cases, an
engineer who is knowledgeable of his well, and who under-
stands the causes of skin effects, can make adequate common-
sense decisions.
Example 3.6: Determination of skin factor due to damage.
Problem. The total skin factor for the well described in

Ex. 3.1 was computed in Ex. 3.5 to be 6.4. This positive
value of skin does not necessarily mean the well is damaged
since other effects can contribute to the total skin factor.
Compute the skin factor caused by formation damage.
Solution. The damage skin factor can be written (Eq. 3.32)

as:
sd = s- sr - ^ -St
/ - s- s^
There is no indication that the well is fractured, and

turbulence should not be important in a low-rate oil well.
If perforation effects are neglected, this equation reduces
to
sd = s - sr
Thus, the problem is reduced to determining the skin effect

due to limited entry.
The skin due to limited entry can be estimated using 0
Eq. 3.33:
3-41
• h_-
sr = 1.35 [F
0.825
{1n (h kH/kV + 7)
P
-10.49 + 0.1 in (h kH/kv)] in rwc - 1.95}
From Ex. 3.1,
h = 30 ft
h = 15 ft
p
r = 0. 33 ft
w
kH/kv = 1 aS^cwêd,
h =0
0
Therefore,
• zm = ho + hp/2 = 0 + 15/2 (A.Krec
zm = 7.5 ft w w kr-n ^ o= O
f wc
rwc =-8:33 ft (Since ho
Substitute this information into Eq. 3.33 to obtain
0 .825
Sr = 1.35 i - 11 { in (30 ^+ 7)
^ J
- [0.49 + 0.1 in (30 v'-l)] in ( 0.33) - 1.951
Sr = 1.35 {3.61 + 0.92 - 1.95}
P Sr = 3.5.
is
3-42
Therefore, the damage skin factor is
•
-s sd = s - 4- 3.S
r 6 4
S d = 2.9
Although not damaged as seriously as indicated by the total

skin factor, it appears that the well does have some
formation damage and is possibly a candidate for some type
of stimulation treatment. A better feeling for this can
be obtained from the flow efficiency.
6. Flow efficiency (condition ratio)

C The skin factor, while it reveals the presence of
damage, does not indicate how much production is being
lost. When damage is significant, it is desirable to Is
stimulate the well if the increased production can be expected
to give a reasonable rate of return. Accordingly, it is
important that we be able to relate the skin factor to
production rate.
The flow efficiency, E, of a damaged well can be
defined in terms-of the productivity index as
E _ Jactual (3. 35)
Jideal
where Jactual is the productivity index of the damaged

well and Jideal is the productivity index which the well
would have if the damage were removed. From Eq. 2.52 we
can write
•
3-43
• _ Q (3.36)
actual
PR pwf
_ Q (3.37)
Jideal -
PR - pwf - Aps
Therefore,
E _( PR - pwf)- Aps (3. 38)

( PR pwf)
The value of PR in Eq. 3.38 can be assumed equal to

pi, the stabilized pressure at the beginning of the test.
The value of pwf is the intercept pressure obtained from
the cartesian plot of pwf versus t by extrapolating the
pseudosteady state straight line to t. = 0; this is
^ illustrated by Fig. 3.16. If the well is not flowed

sufficiently long to achieve pseudosteady state, the last
measured value of pwf can be used as an approximation.
CORRVALUE OF
PR FOR EQ. 3.38
p
PR
^ - \ pR Pwf
pin ^ -^
I
CORRECT VALUE OF Pwf I
Pwf FOR EQ. 3.38
TIME
• Fig. 3.16: Selection of pressures from the

cartesian plot for flow efficiency
calculations.
3-44
7. Damage ratio and damage fac or
•
Damage ratio and damage factor are also used as
indicators of the skin effect. The damage ratio, DR, is
the inverse of flow efficiency:
Piz pwf .
DR = 1 - (3.39
E - -
PR pwf _ Aps
The damage factor is obtained by subtracting the flow

efficiency from 1, i.e.,
A
DF = 1 - E = ps . (3.40!
PR pwf
8. Flow improvement
If the value of Aps used in Eq. 3.38 is based upon
the damage skin factor, i.e., Ops =-0.87msd, and if it
•
is assumed that this damage can be removed by a stimulation
treatment, the flow rate after stimulation should be
1 ^.^117
qafter E 'before (^
stimulation stimulation
Equation 3.41 assumes that pwf is the same before and

after stimulation.
The increase in rate predicted by Eq. 3.41 permits
an engineer to determine beforehand if a proposed
stimulation expenditure is justified.
Example 3.7: Increase in production rate due to removal of

formation damage.
•
3-45
Problem. If the damage detected in Ex. 3.6 were removed,
how much production increase could be expected?
Solution. From previous calculations,
pi =5,000psia (given in Ex. 3.1)

PR
pwf = 4,560 psia (this is the intercept pressure
from Fig. 3.7, Ex. 3.2)
m=- 36 psi/cycle (Fig. 3.3, Ex. 3.1)
sd = 2.9 (Ex. 3.6)
L-V A- f ewLO vc,J0 le "6 "in

The pressure loss caused by damage is computed using Eq. 3.30:
Aps = -0.87msd
= -0.87 (-36)(2.9)
Aps = 91 psi
Flow efficiency is computed using Eq. 3.38:
E = PR pwf Aps
PR pwf
E = 5000 - 4560 - 91
5000 - 4560
E = 0. 79
Assuming this damage cart be removed by stimulation, the flow

rate can potentially be increased to (Eq. 3.41)
_ 1
aafter E abefore
= T I -Tg (140 STB/D)
^
q after = 177) STB/D,
stimulation
3-46
^-.SUr^4Ce rCt-{"e
/ ._^-- •
/e5evvei,r Yo,'Ee
•
^4C°O
_L
Thus, an increase in production rate of 37 STB/D could
ideally be anticipated if the damage were removed.
V. WELLBORE STORAGE
The foregoing discussion deceivingly leads one to

believe that pressure drawdown analysis is simple; it is
simple, but only after the proper straight line on the semi-
log plot has been identified. Most wells tested are subject
to flow behavior which violates one, or more, of the
assumptions which were made when deriving the basic draw-
down equations. For example, all computations discussed
thus far assume the wellbore volume is negligible. This
results from the way the diffusivity equation was solved to
obtain the line source solution. This assumption is not
•
valid and, in many cases, the finite wellbore volume will
cause measured pressures to behave differently than pre-
dicted by ideal theory. When fractures intersect the test
well, their volume may also act as a p-art of the wellbore
and cause deviant pressure behavior.
1. Causes of wellbore storage

The theory used to analyze pressure drawdown data requires
that the flow rate at the formation face be constant throughout
the test; this ideal rate history is illustrated in Fig. 3.1.
Unfortunately, producing rate is controlled at the surface,
not at the formation face. [Because of the wellbore volume,
a constant surface rate does not insure that the formation
face rate is also constant.I
During the time a well is shut in prior to a drawdown
test, reservoir fluids accumulate in the wellbore; the volume
of stored fluids will depend upon the type of completion. •
3-47
• Subsequently, when the well is opened to flow at a constant
surface rate, q, part of the flow will be contributed by the
wellbore instead of the reservoir; i.e.,
q - qf + awb (3.42)
where: q = surface flow rate, STB/D

qf = formation flow rate, STB/D
awb = flow rate contributed by the wellbore, STB/D
As production time increases, the wellbore contribution

decreases, and the formation rate increases until it eventually
equals the surface flow rate. This is illustrated by Fig.
3.17. During the period when the formation rate is changing,
measured drawdown pressures will not exhibit the ideal
semilog straight line behavior which we expect during transient
^ flow. Accordingly, these data cannot be analyzed using the
methods previously discussed.
q -^
qwb
qf
STORAGE
EFFECT ENDS
0
0
t
Fig. 3. 17 Effect of wellbore storage on

reservoir production rate.
•
3-48
Prbduced wellbore fluids generally originate from two
•
sources: A falling fluid level in the casing-tubing
annulus, and expansion of fluids in the wellbore. Each of
these effects will be considered, and will be quantified in
terms of the welibore storage factor, C, which is defined as
C = AVwb (3. 43)

S. 6 ISAP
where: C = wellbore storage factor, bbl/psi

AV = change in volume of wellbore fluids at
wb
wellbore conditions, ft3
Ap = change in formation face pressure, psi.
When a drawdown test begins, the formation face

pressure decreases and causes a corresponding drop in the
annulus liquid level; this is illustrated by Fig. 3.18.
^
144:
C = 5.6
, ..
SHUT-IN FLOWING
Fig. 3.18: Weilbore storage caused by falling

annulus fluid level.
•
3-49
• If Aa is the cross-sectional area of the annulus, and p
is the average density of the fluid in the wellbore, the
wellbore storage factor can be computed as
(A, ft2) (144 in2)

C a9 ft
(p, lbm
^ lbf) (5.615 ft 3)
ft, ) (gc' lbm bbl
or,
144A
5.615p, bbl/psi (3.44)
C
Vo)'\OW This effect will be small if a packer is placed near the

S^ k
C60-11 w0^" producing zone. Note that C will be essentially constant
since p should not change significantly for the pressure
changes normally encountered during a test.
Wellbore fluids can also contribute to production as
a result of fluid expansion. In this case,
V..4, c,.,h
C = (3.45)
5.615
where: Vwb = total wellbore fluid volume, ft3

cwb = average compressibility of fluid
in the welibore, psi-1
It is the total storage factor, i.e., the sum of

Eqs. 3.44 and 3.45, which will govern pressure behavior
during a test. When working with liquids, the largest
storage effect will be caused by a changing annular f,luid.
level; the effect of fluid expansion will generally be
insignificant due to the small compressibility of liquids.
When working with gas wells, the primary storage effect
._,
will be due to gas expansion.
^ In future discussions it will be convenient to use the
wellbore storage factor in dimensionless form as defined
by the equation
3-50
C _ 5.615C 0.894C
(3.46)
•
D 27rh^ctrW 2 ohc t r2
2. Effect of wellbore storage on the semilog plot

The magnitude and duration of wellbore storage can
vary significantly from well to well depending upon the
completion configuration and reservoir characteristics.
In all cases, the effect of wellbore storage is to alter
the early-time pressure data so that these data cannot
be evaluated in the conventional manner to determine
information about the reservoir. The effect of storage
on the semilog plot is illustrated by Fig. 3.19.
C PRESSURE DEVIATION
C2
DUE TO STORAGE
•
CORRECT
nwf C1=0
STRAIGHT LINE
1 `
C3>C2>C1
BOUNDARY EFFECTS
O.i 10 102
1
t, hrs
Fig. 3.19: Effect of wellbore storage on a

semilog plot of pressure drawdown
data.
•
3-51
• When wellbore storage is significant, it must be
considered in the design of a test and in the analysis of
test data. Pressure data affected by wellbore storage
often plot as a straight line on semilog paper; this can
cause an engineer to misinterpret this data as transient
flow, and to make calculations and deductions which are
incorrect. Furthermore, it is possible in some wells for
wellbore storage to last longer than transient flow; when
this happens, the true semilog straight line will not be
present and, accordingly, it will not be possible to analyze
the data using the methods presented thus far. In order to
avoid these problems, it is necessary that we be able to
determine the presence and duration of wellbore storage
effects in pressure transient data, or the test must be
designed to avoid the occurence of wellbore storage.
^ 3. Effect of wellbore storage on formation face rate

During the time that wellbore storage effects are
present, the formation face rate, qf, is related to the
surface rate, q, and the flow rate contributed by the
wellbore, qwb, by Eq. 3.42:
. qf = a - awb, (3.47)
The purpose of this section is to show how the formation

face rate can be further related to the wellbore storage
factor, C.
A. Falling fluid level

Consider a well without a packer as depicted by
Fig. 3.20. During the time this well is shut-in prior
to conducting a pressure drawdown test, reservoir
pressure will support a column of fluid to some height.
When the well is placed on production, the fluid column
in the annulus will begin to fall and the first fluid
3-52
a
•
PC S--\ i: F,
A
a
:awb.l'.
.• piV f
q
f • ^
•
Fig. 3.20: Schematic of wellbore with
changing annular fluid level.
produced will be that stored in the.annulus. As the well

produces during a time interval (t2 - t1), the fluid level
in the annulus will fall from height z1 to z2; accordingly,
the rate at which fluid is produced from the annulus, qwb, is
24Aa(zl - z2)
qwb
5.615(t2 - t1)B
Or, in general,
24A
a dz (3.48)
awb 5.615B dt
•
3-53
• where:
qwb = flow rate contributed by the wellbore, STB/D
Aa = cross-sectional area of annulus, ft2
z = height of annular fluid column, ft
t = producing time, hrs
dz/dt = rate at which annular fluid level is
changing, ft/hr.
B = formation volume factor, RB/STB.
Substituting Eq. 3.48 into Eq. 3.47,
24Aa
dz (3.49)
qf 5 - 6 - l B dt
q + -
24Aa
dz _ (a - q) B. (3.50)
T-71-5 U t_ f
• The flowing formation face pressure, pwf, is
(3.51)
pwf pcs + 1
where:
PCs = surface casing pressure, psia
p = density of the fluid in the wellbore, lbm/ft3
Thus,
d(pwf pcs) _ p dz
(3.52)
dt 144 J T
Equation 3.52 can be combined with Eq. 3.50 to yield:
(24615p4) Aa d(pwf t pcs) = (qf- q)B. (3.53)
q The wellbore storage factor, C, was shown previously,

in Eq. 3.44, to be
3-54
144A
•
(3.44)
C = 5.615a p.
When this relationship is combined with Eq. 3.53,
24C d(pwf pcs) (3.54)

qf = a + B t '
If it is assumed that the surface casing pressure is

constant,
24C dpwf (3.55)

af= q + B ^t
This equation can be written in terms of dimensionless

pressure, PD) dimensionless time, tD, and dimensionless
storage factor, CD, as
q f dpD 0
= 1 - CD ^ (3.56)
q
D
0.00708kh(p. - p )
a ul wf (3 57)
where:
PD =
t = 0_.0002637kt (3.58)
D ûctrw
C = 0.894C (3.59)
D Octhrw
r 1
B. Fluid Expansion
Consider again the well depicted by Fig. 3.20 and assume
that the wellbore is initially filled with a single-phase
fluid. When the well is placed on production and the pressure
in the wellbore decreases at a rate dpwf/dt, the fluid stored
in the wellbore is going to expand and contribute to
production at a rate
3-55
24Vwbcwb dpwf (3.60)
awb 5.615B dt
where: V = volume of fluid initially in the

wb
wellbore, ft3
cwb = compressibility of wellbore fluid, psi.
Thus, from Eq. 3.47,
q + 24Vwbcwb dpwf (3.61)

qf 5.615B dt
In this case, the storage factor is given by Eq. 3.45, i.e.,
C = Vwbcwb (3.45)
S.615
so that,
= 24C dpwf ^ (3.62)

qf q + B dt
Notice that Eq. 3.62 for fluid expansion is identical to

Eq. 3.55 for a falling fluid level; the wellbore storage
factor has a different definition, however, in these two
equations. It follows that Eq. 3.56 will apply when
storage is caused by fluid expansion or a falling fluid
level.
4. Detection of wellbore storage

It is sometimes possible to recognize wellbore storage
from the semilog plot. However, this can be misleadiiLg
since several other things can cause the same behavior.
The most successful method of detecting storage can be
developed by considering Eq. 3.55 (or Eq. 3.62):
^ qf = 1 + 2 4C dpwf
(3.63)
q qB dt
3-56
At early flow times when wellbore storage dominates
•
pressure in the wellbore, i.e., when qf/q = 0,
24C dpwf
+ qB = 0,
__(Ft_
9
^ or
pwf t
2BC pwf = - f dt,
q I d 0
pi 5Av-Pc --V--
^ 2►-t G
so that
B P' Dwc
rX (3.64)
Pi pwf 24C' t b
^--^
.^ b=o °
Taking the logarithm of both sides of Eq 3.64, W-êh tF = o
• Lc^n s toAa41c.
log (pi pwf) = log 2 + log t âd^,, lo WAst•(3.65)
Equation 3.65 suggests a very important conclusion:

When early-time pressure drawdown data are dominated by
wellbore storage, a plot of log (pi - pwf) versus log t
will yield a straight line of unit sZope; this is
illustrated by Fig. 3.21-1 As wellbore effects decrease,
the pressure data deviate away from the unit slope line.
The time, t*, at which the unit slope line ends does not
represent the end of wellbore storage effects; instead,
it represents the time where complete controZ of the
pressure data by wellbore storage ends. It will be shown
in the next section that t* can be used to determine the
time at which wellbore storage effects end.
The unit slope line on the log-log plot can also be
used to determine the storage factor for a drawdown test;
simply select a point on this line and read the values of
pwf) and t which define the point. Substitute
(pi
these values into Eq. 3.64, which is the equation of the
3-57
yt„^.►.oê.t^ c^,^•.^.- ,
s.A
104r-
•
UNIT SLOPI
LINE
1O
•
.H •
•
a
•
10
'^ÊND OF COMPLETE
CONTROL BY
WELLBORE SÌ'ORACE
10
,
1
0.1 1 10 102 103
t, )lT'ti
^
Fig. 3.21: Effect of wellbore storage on the
log-log plot of pressure drawdown
data.
{traz ^. d^ # e C.E-c; ^
unit slope line, to obtain Q^ Vt
_^,^, n,'X•q^ C= Bt . (3.66

^" o^(i'_ v 24pi
pw f
^^'A
The value of C computed using this method is the totaZ

I rg ^^ . Q^^
storage factor and is independent of the type of storage
affecting the test. This value should be approximately
equal to the sum of the storage factors computed using
Eqs. 3.44 and 3.45; if not, a reason should be determined.
When written in dimensionless form, Eq. 3.64 has
the form
cDpD = tD. (3.67
3-58
Taking the logarithm of both sides,
•
log PD = log tD - log CD (3.68)
Equation 3.68 indicates that a log-log plot of PD versus

tD will also give us a straight line of unit slope for
the early-time data when qf = 0, i.e., for the data
completely controlled by wellbore storage. The importance
of this observation will be illustrated in the next
section.
S. Duration of welibore storage

The duration of storage effects increases as C increases,
i.e., as the magnitude of storage increases. The effect of
C on formation face flow rate is illustrated in Fig. 3.22.
•
1.0
C CC 3
qf
q
C }>C2 >C1
Fig. 3.22: Effect of the wellbore storage

factor upon the duration of
storage effects.
3-59
• In Chapter 2, we presented the solution to the radial
diffusivity equation for the case where the formation face
flow rate, qf, was assumed constant and equal to the surface
flow rate, q (i.e., zero wellbore storage). Within this
assumption, the formation face pressure during transient
flow was predicted to be (Eq. 2.35)
pwf = pi - 162.6 kqB" log kt 2 - 3.23^. (3.69)

L Ûcr
t w
According to Eq. 2.71, this equation can be further

expressed in dimensionless form as (notice that we are
considering the case where p = pwf and r = rw) so that
rD = 1)
PD = 2 (In tD + 0.80907) (3. 70)
q
where PD and tD are defined by Eqs. 3.57 and 3.58. When
modified to include the skin factor, Eq. 3.70 becomes
PD = 2(ln tD + 0.80907) + S. (3.71)
Equations 3.70 and 3.71 both assume zero wellbore storage,

i.e., CD = 0.
Detailed studies of the effects of wellbore storage
on pressure drawdown data were conducted in 1970 by Agarwal,
et a123 and Wattenbarger and Ramey24. In these studies,
the radial diffusivity equation was solved for the case
where the formation flow rate, qf, was permitted to vary
according to Eqs. 3.55 and 3.62 (or their equivalent
dimensionless form, Eq. 3.56); i.e., wellbore storage was
accounted for in the solutions. The results of these
studies are presented in Fig. 3.23, for the case where
s = 0
is
3-60
^63
PCrWF Iyi.Z
102
Q)
10
V) C^^ = 0 S= 0
U)
a^
^
^
a)
r--^
O 1
N
o\ r-
Q) 1 ,Ô
10-1
102 103 101 105 106 10' 108
Dimensionless time, tD
Fig. 3.23: Plot of dimensionless pressure versus dimensionless time

illustrating the effect of wellbore storage on pressure
drawdown data (Ref. 24).
y
^
The curve for CD = 0 on Fig. 3.23 is the familiar line
source solution. All of the curves for CD > 0 begin with
a unit slope line, as predicted by Eq. 3.68, then curve with
decreasing slope until they intersect the curve for zero
storage. The point of intersection of each storage curve
with the zero storage curve represents the time when wellbore
storage effects end; accordingly,it is the time at which the
correct transient flow straight line will begin on the semi-
log plot.
When the effects of both wellbore storage and skin are
included in the radial diffusivity equation, the solutionZ3'24
depicted by Fig. 3.24 is obtained. It is noted on this figure
that the skin factor has very little effect upon the time at
which wellbore storage effects end during a pressure drawdown
test. It will be shown in a subsequent section that this
curve, generally referred to as the Ramey Type Curve, can
be used to directly analyze pressure drawdown data.
^ Based upon the solution presented in Fig. 3.24, we can
develop the following rules of thumb to estimate where
storage ends:
1) Neglecting skin, wellbore storage effects will

end at a producing time equal to SOt*; t* is the
time where the unit slope line ends on a log-log
plot of (p. - pwf) versus t (See Fig. 3.21).
Stated in inother way, storage will end 1 to 1h
log-cycles beyond t*.
2) Considering skin, wellbore storage effects will

be negligible for
-}'p^'^Gb^ 3•5S^ G t > Cu(200 000 + 12 , 000s)

(3. 72)
`---=^ Low kk3 ===>-

Equation 3.72 can be valuable in the design of a LpN G, t
pressure drawdown test. The storage factor, C, can be ^4Maq,
4ctD
approximated from the well completion configuration using G
Eq. 3.44 and/or Eq. 3.45; if k, h and u can be approximated
from other sources, the test time required to overcome
3-62
io,
io
0
0
10,
10' 10' 10'
102 103 105 (o6
to
IIkI- - Fig. 3.24: Dimensionless pressure versus dimensionless time for radial flow in an
infinite system, including the effects of wellbore storage and skin
(Ref. 23).
0 • •
• storage can be predicted. A test must be run sufficiently
longer than this time in order for the semilog straight line
to develop. Note that, in addition to C, the length of
storage effects also depends upon reservoir transmissibility,
kh/u; in particular, storage will last longer in low
permeability formations.
Example 3.8: Design of a pressure drawdown test.
Problem. A pressure drawdown test is to be conducted in

the oil well depicted by Fig. 3.25 for the purpose of
determining the effective permeability of the formation and
the skin factor. Other data are:
Fig. 3.25: Well completion configuration for Ex. 3.8.
•
3-64
Depth to mid-point of perforations = 6530 feet
•
Average density of oil in wellbore = 43.7 lbm/ft3
Inside diameter of production casing = 5.012 inches
Inside diameter of tubing = 1.995 inches
Outside diameter of tubing = 2.375 inches
h = 20.1 feet ct = 18.2x10-6psi-1
= 14% rw = 0.333 feet
Uo = 1.7 cp k(from an adjacent well) = 30 md
co = 9.5X10psi 1 B 0 = 1.23 RB/STB
Recognizing that this well is developed on a square 40-acre

pattern, and assuming the welibore is completely filled
with oil at the time the drawdown test is initiated,
a) Compute the wellbore storage factor, C, for the

proposed test.
b) Express the storage factor from part (a) in

dimensionless form.
c) If this well is placed on production at a constant

rate, how long will it take for wellbore storage effects
to end?
d) How long will it take for transient flow to end?
e) How long should the proposed test be run in order to

obtain the desired information?
Solution.
^'^ a) The cross-sectional area of the annulus is

^
Tr ( 5. 012 ) 2 _^r ( 2. 3 7 5) Z
Aa 4 (144) 1
2
Aa = 0.106 ft .
3-65
• The storage factor caused by a falling fluid level in the
annulus can be computed using Eq. 3.44:
144A
C = a
5.615p
^L
C _ (144) (0.106)
.61 43. 7
C = 0.0622 bbls/psi.
The storage factor due to fluid expansion is computed using

Eq. 3.45:
C = Vwbcwb
5.615
The volume of fluid in the wellbore at the time the well

is placed on production is
Vwb = (0.106 ft2)(6,530 ft) + Tr(1.145) 2 (6,530)
Vwb = 834 ft3.
Assuming that only oil is present in the wellbore when the

well is placed on production
cwb = co = 9.5 x 10 6 psi -'
an d
C = (834)(9.5 x 10-6)
5.615
Finally, the total storage factor is
3-66
r , 65- ---^'j C = 0 . 0 6 2 2 + 0. 0 014
•
Practically, as is illustrated by these calculations,

the effect of fluid expansion can generally be neglected
in liquid systems.
b) The dimensionless wellbore storage factor is defined

by Eq. 3.46:
C = 0.894C
D 2
^hcrw
t
C = (0.894)(0.0636)
(18.2x10 ') (^?.3 33
^
CD = lo,Ono.
c) The time required for storage effects to cease is

determined using Eq. 3.72, i.e.,
= Cu(200,000 h + 12,000s)
t
The skin factor has a minor effect on this equation and,

since it is not known, is assumed to be negligible; thus,
t = (0.0636)(1.7)(200,000
(30) (20. 1)
--^t = 3S.9 hrs.
d) The time recuired for transient flow to end is given

by Eq. 3.-.
is
3-67
OuctA
tetf 00026 3 7 k
U-.- (tDA)etf'
Assuming that this well is producing from the center of

a square drainage area, we find in Table 2.2 that (tDA)etf
0.09; therefore,
t _ (0.14)(1.7)(18.2x10-6)(40)(43,560) (0.09)
etf 0.0002637 30
tetf - 85.9 hrs.
e) According to the calculations in part (c), the semilog

straight line required for permeability and skin calculations
will not develop until t > 35.9 hrs. Ideally, we would like
for the straight line to extend for one log cycle; however,
this would require that the test be run for 3S9 hrs. Since
transient flow (and the semilog straight line) will end at
approximately 86 hours, the test should not be run longer
k
0
than 86 hours.
6. Changing wellbore storage

The wellbore storage factor can change during a draw-
down test, as well as during other types of tests. For
example, consider a drawdown test in a pumping well. The
bottom hole pressure under shut-in conditions may be
sufficient to support the fluid level to a packer, or to
the surface. As pumping begins, we have an expansion ^ype
storage. After some time, the bottomhole pressure is
reduced sufficiently to cause the fluid level to start
dropping. The typeof storage then changes from expansion
to falling fluid level and the magnitude of storage
increases significantly. This change in storage will
affect the pressure data and the shape of the basic pressure-
time plots. In other tests such as buildup or falloff tests,
3-68
•
it is possible to have a decrease in storage.
A sharp increase in storage will affect the basic
plots, as shown in Fig. 3.26. In this illustration the
test started with C1 and then increased to C2 at tl. A
decrease in storage would cause similar changes except in a
different direction. These changes in shape are important
because they can be misinterpreted for other things. Since
the changes usually occur when the formation face pressure
corresponds to the hydrostatic head of the fluid column
involved, it should not be difficult to verify if wellbore
storage changes caused the deviant pressure behavior.
It is important to note that when the storage factor
changes, the second storage factor determines when the
correct semi-log straight line begins.
i. Summary wellbore storage

tidellbore storage is one of the most common, and most
serious, problems which the pressure analyst must deal with.
tiVellbore storage can cause several apparent straight lines to
form on the semilog plot and it is often difficult to decide
which line represents true reservoir behavior. The problem
is further complicated when skin effects, boundary effects,
fractures, faulting, or various other heterogeneities cause
abnormal pressure behavior. The engineer must also be
aware that the straight line being sought may not be present;
it is not unusual for wellbore storage effects to completely
mask the transient flow data.
If good early-time pressure data are available, the
log-log plot can often be used to identify that part of the
data affected by wellbore storage. It is cautioned, however,
that an accurate value of pi is needed to make this plot; if
pi is wrong, the value of (pi - pwf) used to prepare the
log-log plot will be correspondingly wrong. This error will
cause the shape and position of the curve to be incorrect.
For example, if pi. is too large, the storage data will plot
as a curve concave upward with slope less than unity, rather
3-69
q
C2
-%I _. _. C') >C1
Cl
pwf
C=0
7 t
CORRECT
^ STRAICHT LINE
- -
0.1 1 10 100
t , hrs
i•
100
C=0 ^- - - -
r ^
^
Ile
i /
An 10 C`>C1
^
Cl
'7 /
\,j /
1 ^ tl
0.1 1 10 100
t , hrs
Fig. 3.26: Effect of increasing wellbore storage

on the semilog and log-log plots.
3- 70
than a unit slope line. This could easily lead an engineer
•
to misinterpret the data.
Example 3.9: Pressure drawdown test with wellbore storage
^^S 4- TâS
Problem. An undersaturated oil well has been produced for
600 hours at a constant surface rate of 61 STB/day. The
bottomhole pressure data have been accurately recorded and
are tabulated as shown. Also, average reservoir rock and
fluid properties are presented. You are to estimate the
formation permeability and the total skin factor. The well
is drilled on an 80-acre square pattern.
pi = 3860 psia uo = 1.25 cp

qo = 61 STB/D = 15%
6 1
Bo = 1.23 RB/STB ct = 21.1X10 Psi
h = 37 ft rw = 0.316 ft
k = 5 md (core data)
•
t, hrs pwf, psia Ap=(pi-Pwf)'Psi
0.2 3855 5
0.4 3851 9
1.0 3837 23
2.0 3815 45
4.0 3775 85
6.0 3742 118
10.0 3677 183
20.0 3566 294
40.0 3452 408
60.0 3422 438
100.0 3362 498
150.0 3348 512
200.0 3334 526
535
300.0
400.0
500.0
3325
3314
3308
546
552
•
600.0 3305 555
3-71
• Solution. Semilog and log-log plots of the drawdown data
are presented in Figs. 3.27 and 3.28, respectively. It is
obvious from the log-log plot that the first four or five
points are completely controlled by wellbore storage. The
data begin to deviate from the unit slope line at t* = 3
hrs; accordingly, wellbore storage effects should end at
approximately 50t*, or 150 hrs. This corresponds to the
beginning of the straight line shown on Fig. 3.27. It is
still possible, however, that data after t = 150 hrs are
affected by the boundary. 'Core data indicate a permeability
of approximately 5 md; using this information, and assuming
that the well is centered in a square drainage area, the
estimated time for transient flow to end is, from Eq. 3.7,
ûctA
tetf - 0.0 (tDA)etf
From Table 2.2, (tDA)etf = 0.09; thus,

0
t (0.15)(1.25)(21. lX10_6)( 80) (43560) (0.09)

etf ' (0.0002637) 5
941 hrs.
tetf
Therefore, there should be no boundary effects, and the

indicated straight line on Fig. 3.27 should be correct.
From Fig. 3.27,
m = -65 psi/cycle
k = - (162.6) (61) (1.23) (1.25)

• and,
( -65 )
k = 6.3 md
^ The drawdown pressure at t = 1 hr is, from the

extrapolated straight line,
3-72
3900
3800
3700
3600
^
.r.,
pihr - 3483 psia
`'4 3500
3400
m = - 65 psi/cycle
3300
Fig. 3. 27: Semilog plot,
Ex. 3.9.
3200 11 1 1
0.1 1 10 102 103
t,hrs
• • •
• 11 •
103
UNIT SLOPE LINE
^ 102
-,
4A 3 hrs
° 10
1
10-' 1 10 102 103
t, hrs
Fig. 3.28: Log-log plot, Ex. 3.9.

plhr = 3483 psia.
•
The total skin factor is computed using Eq. 3.29:
s= 1.151 plhr pi - log k + 3.23

m ûctrw
s = 1.151 1 3483-3860
-65
6.3 + 3 .2
- log
(.15) (1.25) ( 21.1x10 6) (. 316)2
^
s = 2.1
•
^ VI. MULTIPHASE FLOW
The pressure drawdown theory described in previous

sections was developed for liquid flow and, theoretically,
is only appropriate for pressure surveys in undersaturated
oil reservoirs. Many tests, however, will likely be
conducted in reservoirs where the average pressure is below
the bubble point so that two phases, oil and free gas, will
coexist in the reservoir. Further, it is possible in some
reservoirs that a third phase, water, will be flowing. How
do we analyze pressure tests in these situations?
A rigorous approach to this problem requires that the
spatial distribution of each fluid be considered as a
function of time; the resulting equations require computer
solutions which are not practical for most engineering
applications. It has been found' 5 28 that a less rigorous
approach is satisfactory for most of the tests we analyze.
If the fluids in the reservoir are immiscible, and
saturations are uniformly distributed, the pressure draw-
10
3-75
S
down equation can be approximated as:
X t
162.6 Rh [log t 2- 3.23 + 0.87s^ (3.73)
pwf = pi - t ^ctw J
a
The total flow rate, qRt, can be determined as
qoBo + (1000 qgt - qoRs - qwRsw)Bg + q w B w (3.74)

qRt
where:
total production rate, RB/D
q Rt
total gas production rate, Mscf/D
qgt
water production rate, STB/D
qw
Rs solution gas-oil ratio, scf/STB
solution gas-water ratio, scf/STB
Rsw
gas formation volume factor, RB/scf
B9
water formation volume factor, RB/STB
^ Bw
The total mobility, at, is:
k k k
w
at = ouo
+ ug +Uw . (3.75)
Within the assumption that Eq. 3.75 is true, Perrine27

has shown that the permeability to each mobile phase can be
computed as
B
ko = - 162.6 °m0h° (3.76)
Bu
w
k w = - 162.6 q^`mh w (3.77)
162.6(1000 qgt
qoRs qwRsw)Bgug
kg = - mh (3.78)
Finally, the total skin factor is computed from the

relationship
3-76
^
•
s= 1.151 pihm pi - log t2 + 3.23 (3.79)
^ ^c t rw
VII. TYPE CURVE ANALYSIS
Another method which can be used to analyze pressure

transient data is the type curve. This technique, first used
by ground water hydrologists,29'3° was introduced in the
petroleum literature in 1970 by Ramey31 and Agarwal, et a123
Type curves, in somewhat different forms, were published
later by hicKin1ey32, Earlougher and Kersch33, Gringarten,
et al?' and Bourdet, et al!'
The basic principles of type curve matching were pre-
sented in Chapter 2. The objective of this section is to
show how the type curves of Ramey, et al ?3,31 McKinle Y32
Gringarten, et a13; and Bourdet, et a13; can be applied to
the analysis of pressure drawdown data.
0 1. Ramey type curve

The Ramey type curve is a graph of pressure versus time
plotted on log-log paper. Conveniently, both pressure and
time are plotted in dimensionless form. Dimensionless
pressure, pD, was defined previously in terms of real pressure
by Eq. 3.56, i.e.,
0.00708kh (pi-pwf)
(3.56)
PD aBu
and dimensionless time was defined by Eq. 3.58 as
t 0.0002637kt (3 . S8)
D Ouctrw
The Ramey type curve was presented previously as Fig.

3.24, and is presented again for convenience as Fig. 3.29.
Each curve on this graph represents the pressure transient
data for a specific magnitude of formation skin and wellbore
storage. Formation skin, as discussed previously, is
quantified in dimensionless form by the skin factor, s.
3-77
• • •
10`
C^
A
to QIP
0
CL
r
00 OQ^
10'14^51111[S1CZ1kIItW
10Z 103 104 10° 10' 10' 10"
tp
Fig. 3.29 Dimensionless pressure versus dimensionless time for radial flow in an
k
infinite system, including the effects of wellbore storage and skin (Ref. 23).
The magnitude of wellbore storage is expressed by the
•
wellbore storage coefficient, C, which in dimensionless
form, CD, was defined by Eq. 3.59 as
C = 0.894C (3.59)
D ^cthrW
Inspection of Fig. 3.29 reveals several interesting

features. First, all of the curves which represent non-
zero values of CD form a straight line with unit slope at
small values of tD. This observation, as discussed in a
previous section, serves as the basis for using the log-
log plot to detect wellbore storage. Further, it is noted
that after the unit slope line ends, the pressure data
fall below an extrapolation of the unit slope line; after
a sufficiently long flow time, the pressure data intersect
a curve for CD = 0 and a value of s which may be negative,
zero or positive. For example, consider the curve of pD
versus tD for CD = 103 and s = 10. For values of tD less
than about 1700, PD is a linear function of tD, and the
slope is unity; as noted previously, wellbore storage
completely controls this data. For values of tD greater
than 1700, PD deviates below the unit slope straight line.
Finally, at a tD of 9 x 10", the PD values intersect the
curve for CD = 0 and s = 10. This physically means that,
for values of tD > 9 x 104, wellbore storage effects are
negligible and the pressure data would yield the true
straight line if plotted on a semilog graph. Data in the
interval 1700 < tD < 9 x 104 are affected by both wellbore
storage and skin.
The Ramey type curve can be used to directly determin(-
permeability and skin. Each of the curves on Fig. 3.29
represents the theoretical pressure behavior which would be
expected if a pressure drawdown test was run in a reservoir
having the indicated skin and wellbore storage. It should
be noted that the Ramey curve applies only to the transient,
3- 79
•
radial flow of a single phase slightly compressible fluid.
If, when a drawdown test is run, we can determine which of
these theoretical curves the field data best match, it is
possible to compute the properties of the subject reservoir.
The required procedure for type curve analysis is:
1) Obtain a sheet of tracing paper and

construct a log-log scale having the
same dimensions as the type curve to
be used; this is best accomplished by
placing the tracing paper on the type
curve and tracing the major grid lines.
On the tracing paper, then, plot the
pressure drawdown data as Ap versus t,
where Ap = pi -pwf. This plot, referred
to as a data curve, is illustrated by
Fig. 3..30.
• c
1000
UNIT
SLOPE •
• •
• • ^ ^ • • • •
~ 100
.r.,
1(^
10-3 10 2 10-1 1 10 10`
t, hrs
Fig. 3.30: Data curve for use with Ramey type curve.
2) Check the early-time points on the data curve

to see if they form a unit slope straight line;
if so, select a point on this line and read the
corresponding values of Ap and t. Use this
information to compute the wellbore storage
coefficient, C, using Eq. 3.66, and the dimension-
• less storage factor, CD, using Eq. 3.59.
3- 80
•
3) Place the tracing paper onto the type curve
in such a way that the unit slope line on
the data curve overlays the unit slope line
on the type curve which has a value of CD
equal to the calculated CD.
4) Keeping the unit slope lines of the data
curve and the type curve matched, slide the
data curve over the type curve until the
data curve matches one of the curves on the
theoretical plot. An illustration of these
curves in the matched position is presented
by Fig. 3.31.
10`
10
cc ^1
ccr^
c
•
C
):
10,
102 10° 10' 10° 10` 10'
0.0002637kt
ûcrt 2w
Fig. 3.31: Illustration of data curve and Ramey

type curve in the matched position.
5) When the best match is determined, pick a

"match point"; a match point is an arbitrary
point common to both graphs. Corresponding
to this point, record values of the ordinate
and abscissa from the data curve, (Ap)M and
(t)M, and from the type curve, (pD)M and ( tp),,.
6) The ordinate of the type curve is
3-81
• 0.00708khOp
(3.56)
PD = qBp
By substituting values of the match point,

(Ap),,T into this relationship, the
and (PD)MI
permeability can be estimated as
k = 141.2qBu ( pD)ht
(3.81)
h
TA p7, 1
7) The abscissa of the type curve is
t = 0.0002637kt (3.57)
D z
ûctrw
Values of the match point, (t)h1 and (tD)M'

when substituted into Eq. 3.57, permit
the reservoir porosity-compressibility
product to Sc computed, i.e.,
_ 0.0002637k (t)hi (3.82)

4ct ur2 ^^
t
8) Each curve on Fig. 3.29 represents a specific

value of skin. Therefore, the value of s
is read directly from the theoretical curve
which the field data match.
The foregoing procedure has been photographically illustrated
by Earlougher3.
Example 3.10: Analysis of pressure drawdown test using the

Ramey type curve.
Problem. Use the Ramey type curve to analyze the pressure

drawdown data, nresented in Ex. 3.9, for permeability and skin.
^ Solution. Prepare a log-log plot of op versus t on tracing

paper having the same scale as the type curve to be used.
3-82
•
Figure 3.32 presents this data curve. It is noted that
the first five points on the data curve form a unit slope
line. The equation of this line is, from Eq. 3.66,
qBt
C =
pi pwf
l0Y
•• ••••
• •
•
•
.^ 1tj2
•
CD = 10'
c ln
1
10 10 10` 103
t , hrs
Fig. 3.32: Data curve, Ex. 3.10.
Select an arbitrary point on this straight line, i.e.,

Ap = 23 psi at t = 1 hr. Then,
•
3-83
• C = (61 STB/D)(1.23 RB/STB)(l hr)
(24) (23 psi)
C = 0.136 bbl/psi.
From Eq. 3.59,
C = 0.894C
D Octhr`
(0.894)(0.136)
CD = (0.15) (21.1xl0-6) (37) (0.316) 2
CD = 10,397 = 104
Overlay the data curve onto the type curve in such a

^ way that the unit slope line on the data curve coincides
with the straight line portion of the type curve for
CD = 104 . Keeping these straight lines matched, slide the
data curve over the type curve until the best match is
obtained. This match is illustrated by Fig. 3.33.
An arbitrary match point is selected. From the data
curve, (t)^^ = 10 hrs and (4p) N1 = 100 psi. Similarly,
from the type curve, (tD)M = 5.4 x 104 and (pD)M = 2.35.
According to Eq. 3.81,
k = (141.2)(61)(1.23)(1.2S) 2.35
37 x 100
k = 8.4 md.
It is noted from Fig. 3.33 that the drawdown data match a

theoretical curve with s = 5; this is the total skin for the
well. This compares with k = 6.3 md and s = 2.1 calculated
in Ex. 3.9 using the conventional semilog plot.
3-84
a
CL
W
CO
Ln
10= IOj 104 IOS 106 107 106
to
Fig. 3.33: Curve match for Ex. 3.10, (:h = 10"
• • 0
•
Two important observations can be made from Ex. 3.10:
First, because of the similar shapes of the type curves,
it is difficult to obtain a unique match. In fact, if a
unit slope line does not form on the data curve so that
CD can be computed, it is impossibZe to determine a unique
match. Suppose, for example, that CD had not been known
in Ex. 3.10, and the data had been matched on the curves
for C D = 103 instead of CD = 10', as shown by Fig. 3.34.
Although CD = 103 is incorrect, you can see from Fig. 3.34
that an exact match is still possible; obviously this
match would give incorrect answers. Because of this
difficulty in obtaining a unique match, it must be con-
cluded that the conventional semilog plot is a more
accurate method to evaluate pressure transient data.
It is also observed from Ex. 3.10 that the last 6
data points on Fig. 3.34 lie on the curve for CD = 0.
^ This means that these data points are not affected by
wellbore storage and, if plotted on a semilog plot, could
be expected to form the correct straight line. This
observation is confirmed if we observe the semilog plot of
this data which was presented in Ex. 3.9 as Fig. 3.27; it
is noted in Fig. 3.27 that points which form the straight
line are the same points which lie on the curve for CD = 0
on Fig. 3.33. This observation is very important because
it means the type curve will show us which data should form
the correct straight line on the semilog plot. Further, if
we look at Fig. 3.34 which shows a match of the test data
with the wrong storage curve, it is noted that these same
six points fall on the curve for CD = 0; i.e., even though
we don't know the correct storage factor, the type curve will
still show us which data are not affected by wellbore storage.
From these observations, we can make several conclusions
concerning the use of Ramey type curves;
^ 1) Because of the difficulty in obtaining a

unique match, type curve matching is not
as accurate as the conventional semilog
3-86
10t
I0
C
CL
00
I0'
10= 103 104 IOS 10s 107 101
to
Fig. 3.34: Curve match for Ex. 3.10, C D = 103
• 0 0
q
method. Therefore, type curves should
never be used for an analysis when
conventional methods can be applied.
2) The most important application of type
curves is to show which data are not
affected by wellbore storage and,
consequently, can be used to make a
conventional semilog analysis.
The only time a type curve should be used for the

analysis of nressure data is when all of the test data
are affected by storage. It was noted in a previous
section that data affected by storage will not form the
correct straight line on a semilog plot and, accordingly,
cannot be analyzed in this way. However, data affected
by storage can be analyzed using the type curve if a
unique match can be obtained.
^ 2. McKinley type curve

The McKinley type curve32 is a log-log graph of reaZ
time, t, versus a pressure group, 5.615 ApC/aB. This type
curve is illustrated by Fig. 3.35. McKinley's curve is
similar to Ramev's curve in that both were derived for
radial flow in an infinite reservoir. However, McKinley's
curve is different in the following ways:
1) Time is plotted on the ordinate rather than

the abscissa. Further, time is expressed
as real time, in minutes, rather than
dimensionless time.
2) Pressure drawdown, Ap = p-pwf , is plotted
on the abscissa rather than the ordinate,
and is included in the dimensional group,
5.615 ApC/qB.
3) The curves were computed usinj a constant
value of ûc t rw/k = 10-7 cp-ft /md-psi
4) The curves were generated, using finite
difference techniques, to describe
pressure buildup tests; however, they
can be applied to the analysis of
pressure drawdown data.
3-88
o ^
N
O
T •
^ ^--i 4--^
4-• ^
r J
v ^.
Gf,
sajnutw Ì
•
3- 89
• S) The curves do not include the effect of
a skin factor. Each curve on Fig. 3.35
is for an undamaged well with the para-
metric value of kh/5.61SuC shown on the
curve. The curves can be used, however
to detect the presence of a skin effect.
6) The 45° dashed straight line on the
right-hand side of Fig. 3.35 represents
100 percent wellbore storage. The equation
of this line, modified from Eq. 3.66, is:
t(minutes) = 1440 (pi pwf)C- (3. 83)

q
7) The dashed line to the left represents 1

percent wellhore storage. The approximate
equation of this line is
t(minutes, lo storage) =2'2 x 106 (3.84)

kh
5.61SuC
• Data above this line should plot as a

straight line on the conventional semi-
log plot.
A. Calculation of permeability
In order to analyze pressure drawdown data for
permeability using the McKinley curve, a curve matching
procedure is required. This procedure is:
1) Overlay the McKinley graph with a sheet

of transparent paper and draw horizontal
and vertical axes to match those on the
type curve.
2) Prepare the data curve on this paper by
plotting t in minutes on the ordinate,
and Ap on the abscissa. This data curve
3) Place the data curve on the type curve
in such a way that the time scales match.
Note that, once this is done, no vertical
movement of the two graphs relative to
• each other is required.
3-90
164
•
^
103
. F:
^ 10Z
10
1 10 102 103 104 l05
Qp = pi-pWf, psi
Fig. 3.36: Illustration of data curve required

for McKinley type curve matching.
•
4) Slide the data curve horizontally
over the McKinley curve until the
data coincide with one of the type
curves.
5) Record the value of (kh/5.615uC)r1 from
the curve which the data match.
6) Select an arbitrary match point
common to both graphs and record the
value of pressure, (Ap)M, correspond-
ing to this point. Also, record the
value of the pressure group, (5.615
dpC/qB)k, ,, defined by the match point.
i) Compute C according to the relationship 0 1
C = (5.615ApC B (3.85)
qB )M1f S. 615 Ap^1
8) From the value of (kh/5.615UC)M obtained

in step 5, calculate permeability as
•
3-91
E
kh 5.615aC
(3.86)
k = 5.15uCM ( h )
Example 3.11: Analysis of pressure drawdown test using the

McKinley type curve.
Problem. Use the McKinley type curve to analyze the pressure

drawdown data in Ex. 3.9 to determine permeability.
t, min ^p =(pi-p. , psi

pwf' psia
12 385S 5
24 3851 9
60 3837 23
120 .3815 45
240 3775 85
360 3742 118
600 3677 183
1200 3566 2194
2400 3452 408
3600 3422 438
6000 3362 498
9000 3348 512
12000 3334 526
18000 3325 535
24000 3314 546
30000 3308 552
36000 3305 555
Solution. A plot of the data curve is presented in Fig. 3.37.

This curve was combined with the McKinley curve to obtain the
match presented in Fig. 3.38. From the matched curves,
kh
^ (5.615uCM = 215
3-92
10`
•
104
ln3
4-J
102
10 •
1 10 102 10, 3 104
AP = pi-pf, psi
Fig. 3.37: Data curve for Ex. 3.11.
Also, corresponding to the match point on Fig. 3.38,
(AP)NI = 100
5.615ApC) = 0.93
â M
Substituting these match points into Eq. 3.85, along with ^

other reservoir data, we obtain
3-93
• • •
a^
4-J
5.616ApC/cIB, cu ft-day/RB
Fig. 3.38: Match of data curve with McKinley curve, Ex. 3.11.
= 5.615ApC B ^
^ ^ qB .615 Gp ,t
)M
C = (0.93) (51)i1.2100
C = 0.124 bbls/psi
Finally, from Eq. 3.86,
kh 5.615uC
k = (5.615^) ^t ( h )
k = ()15) (5.615) (1. 25) (0.124)

37
k = 5.1 md.
This permeability compares favorably with the values

computed in Exs. 3.9 and 3.10 using the conventional
method and the Ramey type curve, respectively.
It is noted that the last four points plot aho^-e the
1 percent storage line on Fig. 3.38; accordingly, these
points should lie on the correct transient flow straight
line on the semilog plot. This is in agreement with the
conventional analysis presented in-Example 3.9, and the
semilog plot in Fig. 3.27.
B. Effect of damage on McKinley curve

When a well has damage, the early-time data will match
a curve which has a small value of the parameter kh/5.615uC.
At later times, the data will match a larger value of
kh/5.615uC, reflecting the permeability in the undamaged
zone. While the skin factor cannot be determined from these
curves, McKinley has shown32'36'37 that the flow efficiency
of a well can be determined from matches of the early-time
and late-time data. Modifications of the McKinley curve
which account for restricted entry are also available3g.
3-95
• C. Potential errors in using McKinley type curves
Common errors made in using the McKinley type curve are:
(1) Value of ûctrw/k different from assumed value -
The McKinley type curve was developed using a constant value
of ûctrv/k = 1.0 x 10-7 . Figure 3.399 shows the relation-
ship between the true value of kh/u and the value which will
be computed using the McKinley type curve, as a function of
this parameter. Obviously, if the value of ûctr^/k for the
reservoir being tested varies significantly from the value
assumed by the McKinley type curve, a significant error can
result.
(2) Data out of wellbore storage region - The McKinley
curve should not be used unless data affected by wellbore
storage are present. If only data not affected by wellbore
storage are used in the type curve analysis, significant
error can occur in calculated results. When storage effects
^ are not present, the conventional method of analysis should
be used.
3. Gringarten type curve

The type curve developed by Gringarten, et al3', is a
plot of PD versus tD/CD where each of the curves on this
plot is characterized by a different value of CDe 2S. The
Gringarten type curve is presented by Fig. 3.40. Character-
istics of this type curve which should be noted are:
1) The ranges of CDe2S for various wellbore

conditions, i.e., damaged, zero skin,
acidized, and fractured, are indicated
on the curve.
2) The end of wellbore storage and the start
of the semilog radial flow straight line,
corresponding to a So approximation, are
shown by a dashed line on the curve. The
times indicated by the dashed line are
somewhat larger than those given by the
"50t rule" or by Eq. 3.72, but are believed
^ to be a good practical approximation to
determine where wellbore storage effects
end.
3-96
1.6
1 1 . 1 LL
1.4 fi l l Il y
W 1.Z
E-
^
1.0
E-
x f il l
w
0 f i ll
V r--^
.T'
0. 8
7T
[f il l H i ll
0.6
H il l
2.275 x 105ûctrw
Fi g. 3. 39 ; Correction factors for McKinley type curve.
0 16
0 0
• • •
CoVVMM O tiA FlDPE71101 .Wkn . FRANCE . AM rlpMt newvW .
B
YI^
00
DIMENSIONLESS TIME 'O = 0.000293 kh At

C°
Fig. 3.40: GriTigarten type curve for homogeneous reservoirs (Ref. 34).
3) All curves, except for very low CDe2s
•
values, merge into a single unit-slope
straight line at early times. Data
which plot on the unit-slope straight
line are completely controlled by
wellbore storage.
4) The axis on the right of the curve,
i.e., At/t , is for pressure buildup
analysis aRd will be discussed in
Chapter S.
5) The basic model used to construct the
curves for C D ezs > 0.S is identical to
that used to construct the Ramey type
curve23 and the Earlougher-Kersch type
curve33. Curves for C eZs < 0.5 are for a
well with an infinite 2onductivity
vertical fracture as developed in
Reference 40. The analysis of fractured
wells will be discussed in more detail
in Chapter 6.
The Gringarten type curve can be used to analyze

pressure drawdown data to obtain the permeability-thickness
product, the wellbore storage factor, and the total skin
factor. The procedure required to make these calculations is:
1) Overlay the Gringarten type curve with

a sheet of transparent paper and draw
horizontal and vertical log-cycle lines
to match those on the type curve.
2) Plot the pressure drawdown data on this
paper with Ap = p.-p on the ordinate
and t on the abscissWf This is the
same data curve required to use the
Ramey type curve and is illustrated by
Fig. 3.30.
3) Check the early-time points on the data
curve to see if they form a unit slope
straight line; if so, draw in the unit
slope line as it will be of assistance
in the curve matching procedure.
4) Place the data curve onto the type curve
in such a way that the unit slope line,
if present, overlays the unit slope line
on the type curve. Keeping the unit
slope lines of the data curve and type
curve matched, move the data curve
relative to the type curve until the
field data match one of the curves on
3-99
• If a unit slope
the theoretical plot.
line is not present on the data curve,
the matching procedure is not as con-
venient; in this case, the best match
must be decided upon without the control
imposed by the unit slope line.
S) When the best match has been determined,
select a match point and record the
corresponding values of (Ap) M and (PD)M
from the ordinate and the values of (t)N1
and (tD/CD)M from the abscissa.
6) Record the value of (CDeZS)N1 from the curve
which the data match.
7) Using the match values of pressure, compute
permeability as
k = 141h2aBU ( (P ) (3.87)
pP)^1
^q '
^ 8) If a unit slope straight line is present

on the data curve, a point on this curve
can be selected and the corresponding
values of Ap and t substituted into Eq.
3.66 to compute the wellbore storage
factor, C. Alternatively, or if the
unit slope line is not present, C can
be calculated using the match values of
time:
C = 0.000295 uh t(tCî (3.88)

D D h1
9) Compute the dimensionless wellbore

storage factor, CD, using Eq. 3.59, i.e.,
C = 0.894C (3.59)
D ^cthrW
10) Compute the total skin factor using the

match value of CDe2s:
(C e2s)
^ s = ^ ln DC NI . (3.89)
D
3-100
•
Example 3.12: Analysis of pressure drawdown data using the
Gringarten type curve.
Problem. Use the Gringarten type curve to analyze the

pressure drawdown data presented in Ex. 3.9 for permeability,
wellbore storage factor, and total skin factor.
Solution. The first step is to prepare a log-log plot of

pi pwf versus t on tracing paper having the same scale
Ap
as the type curve to be used. This plot, which is the same
as that required by the Ramey type curve matching procedure,
was presented previously in Ex. 3.10 as Fig. 3.32. Next,
overlay the data curve onto the type curve and, keeping the
unit slope lines on the two curves matched at all times,
determine which theoretical curve the field data best match.
This match is shown by Fig. 3.41.
From the arbitrary match point indicated on Fig. 3.41,
(AP) NI = 100
(pD)Nt = 2.25
(t) NI = 10
(tD/CD)%I = 5.2.
Also, from the data match,
(CDe2s)M = 108
Using the pressures corresponding to the match point,

permeability is computed using Eq. 3.87:
k = 141.2qBp (PD) NI
hi TIIht
^
3-101
• • •
OtpplVM 0 1WM F1OP£DIOL . YMkn . FlUNM. All dpBls nwved.
Ii;
1--+
a tl
103 104
2 3 4
kn At
DIMENSIONLESS TIME =o.oo0i95 -E-
cp a i. i x v
Fig. 3.41: Match of pressure clrawclown data on the Gringarten type curve, Ex. 3.12.
k = 141.2(61)(1.23)(1.25) 2.25
•
37 100
k = 8. 1 md.
The wellbore storage factor is computed by substituting

the match values of time into Eq. 3.88, i.e.,
kh (t)M
C= 0.000295 u tD CD ^1
C = 0.000295 (8.1)(37) 10
1.25 5.2
C = 0.136 bbls/psi.
This value is converted to dimensionless form using Eq . 3. 0-9

i.e.,
C = 0.894C
D Octhrw
C = (0.894)(0.136)
D (O.1S) (21.1 x 10-)(37)(0.316 )2
9
CD = 10,397.
With CD known from the previous step, the total skin

factor can be computed by substituting the match value of
CDeZS into Eq. 3.89:
s= 0.5 in
r (C e2s)
D
CD
M
0
108
s =
0.5 ln [lo397j •
3-103
•
These values of permeability, wellbore storage factor
and total skin factor compare favorably with the results
previously computed using other methods, i.e.,
Conventional semilog plot:
k = 6.3 md
s = 2.1
Ramey type curve:
k = 8.4 md
C = 0.136 bbls/psi
s = 5
McKinley type curve:
^ k = 5.1 ma
C = 0.124 bbls/psi.
4. Derivative type curve

All of the type curves previously presented utilize a
plot of pressure change, Ap, versus elapsed time, t, to
characterize a reservoir. These curves make possible the
identification of flow regimes associated with wellbore
storage and infinite-acting radial flow. Further, it is
possible by curve matching techniques to determine reservoir
characteristics such as permeability and skin factor.
Unfortunately, because of the similarity of curve shapes on
these type curves, it is often difficult to uniquely match
field data on one particular curve. If a unique match cannot
be determined, it follows that unique values of reservoir
parameters cannot be computed. An example of this problem
can be seen by considering the Gringarten type curve
3-104
presented in Fig. 3.40. At high values of COeZS on the
•
Gringarten curves, all of the curves have similar shapes;
consequently, if field data happen to plot on this portion
of the curve, it is difficult to obtain a unique match
and, accordingly, unique values of k, s and C.
Pressure transient data will often exhibit character-
istics which are more uniquely related to, and more sensitive
to, reservoir paramaters if the pressure derivative, rather
than pressure, is plotted versus elapsed time. Application
of the pressure derivative to conventional analysis has
been demonstrated for fractured wells' interference tests;2
and in the detection and location of sealing faults.°3 The
effect of plotting the pressure derivative rather than
pressure differential can be more generally observed,
however, by considering pressure transient solutions in the
form of a type curve.
The Gringarten type curve, Fig. 3.40, was replotted by
Bourdet, et a13; using the dimensionless pressure derivative,
pD, rather than dimensionless pressure, pD, where
d (PD)
PD = d(tll/CD (3.90)
The derivative type curve for a homogeneous reservoir is

presented by Fig. 3.42. You will observe that this type
curve is presented as a graph of pD(tD/CD) on the ordinate
versus tD/Cll on the abscissa. Also plotted on this graph
is pp versus tD/CD, i.e., the Gringarten type curve. Having
both pressure and pressure derivative plotted on the same
graph permits simultanecus matching of pressure data and
pressure derivative data, thereby utilizing the character-
istics of both to obtain a better solution.
The procedure required to use the derivative type
curve can best be understood by first considering the
mathematical basis of the curve. The relationship between
dimensionless pressure, pD, and pressure change, 'Ip = pi-pwf,
3-105
• E •
,
£^
c1o
dI-
Y ►
rQ
^.lJ
^
^
W
^
^
^
CD 1 i n 5 n^ q ra
µ C
Fig. 3.42: Derivative type curve for a homogeneous reservoir (Ref. 35).
•
was previously presented as Eq. 3.56, i.e.,
kh A(3.56)
PD = 1 .2qBp p'
This can be differentiated with respect to time as
_ AP T (3.91)
dtD 141k2qBu ddt^ 141k2qBu
By employing the 'chain rule of differentiation, Eq. 3.91

can be differentiated with respect to tD/CD:
D /C D)
(3.92)
tpDC d(t^ 141kZqBu ^p^'
D D
Introducing the definitions of t0 and CD from Ecls. 3.58

and 3.59, respectively,
tD _ 0.00029Skht (3_93)
CD uC
d(tD/CD) _ 0.000295kh (3,94)

and
dt uC
Equations 3.93 and 3.94 can be combined with Eq. 3.92

to yield
, tD _ kh tAp, (3.95)
pD CD 141.2qBp
Equation 3.95 indicates that the dimensionless

pressure derivative function pDtD/CD, which is the function
plotted on the ordinate of the derivative type curve, is
directly proportional to tAp'. This equation provides the ^
basis for permeability determination using the derivative
3-107
• type curve. Once the pressure transient data, with Ap't
plotted on the ordinate, is matched with the derivative type
curve, with pDtD/CD plotted on the ordinate, permeability
can be computed from Eq. 3.95 using any match point, i.e.,
k= CD )M
(PD C^)1^1
(3.96)
^p' t)^t h
The procedure required to analyze data using the derivative

type curve will be detailed in a subsequent discussion and
illustrated with an example problem.
Other important characteristics of the derivative type
curve are:
1) All derivative curves on Fig. 3.42, except

^ for those with small values of C e2s ,
converge to a single unit slope Yine at
early times. This unit slope line represents
data which are completely controlled by
wellbore storage. Those curves with small
values of CDe2s also form a unit slope line
but at times smaller than those shown on Fig.
3.42. The occurrence of a unit slope line on
the derivative type curve is predicted by
Eq. 3.67 which describes the pressure behavior
at early times when wellbore storage completely
controls the data:
tD
PD = CD • (3.67)
Therefore,
dpD
p'D = tD CD = 1
and
t t t
pD (CD) = (1) (CD) = D
D LD
log tD) = log (tD) (3.97)

(p ^D CD CD -
3-108
q
According to Eq. 3.97, a unit slope line
should be expected on a log-log plot of
pD(tD/CD) versus tD/CD when wellbore
storage completely controls the data.
2) When transient radial flow is reached,
i.e., when data would form a straight
line if plotted on a conventional semi-
log graph, the derivative curves become
horizontal at a value of pD(tD/CD) = O.S.
This is expected if we consider the
exponential integral equation (Eq. 3.71)
which describes transient radial flow,
i.e.,
PD = Z(ln tD + 0.80907) + S. (3.71)
However, since we know from Eq. 3.89 that

1 CDezs
s = 2 In C
D
Eq. 3.71 can be rewritten as

q
t
1 (3.98)
PD = 2(ln CD + 0.80907 + In CDe2S).
D
Taking the derivative of Eq. 3.98,
dpD _ 0.5
d-Et
CD tD CD
Therefore,
tD
PD CD = 0.S. (3.99)
3) When both wellbore storage data and i

transient radial flow data occur during
a test, a log-log plot of the test
data will exhibit two straight lines,
i.e., a unit slope line at early times
and a horizontal line at late times.
When the field data are compared to
the derivative type curve and the two
straight lines on the log-log plot of
the field data are overlayed onto the
corresponding straight lines on the
derivative type curve, only one match
0
3-109
• position will be possible. With this
control imposed on the matching process,
and recognizing that between the two
straight lines on the derivative type
curve the curves for different values
of CDe2s are distinctly different, it
is easy to identify the correct value
of. CDe2s once the match is obtained.
4) A log-log plot of test data using the
pressure derivative will not be
affected by any error in pressure
at the beginning of the test, i.e., pi.
An error in p, however, can signifi-
cantly affectlthe shape of a log-log
plot of Ap = pi-p f versus t. For
example, data co;6letely controlled by
wellbore storage should plot as a unit
slope straight line on a log-log plot
of p.-p versus t; this will only be
true, however, if p. is correct. It
is significant that la log-log plot
using the pressure derivative will
yield a unit slope straight line in
this situation regardless of any error
which might be present in pi.
A practical limitation associated with the use of the

pressure derivative is the ability to measure pressure
transient data with sufficient frequency and accuracy so
that it can be differentiated. Differentiation can be
very difficult and often inconclusive if inaccuracies due
to gauge resolution, vibration, etc. which are generally
present with mechanical gauges, are present, or if
pressures are not measured with sufficient frequency.
These problems are particularly difficult when working with
pressure data measured with acoustical fluid level devices
which are generally used on pumping wells. Consequently,
the pressure derivative can generally be applied only to
data measured with highly sensitive electronic bottomhole
gauges.
Another problem which must be considered is the
selection of a method of differentiation. Many differenti-
ation methods exist which vary from simple forward or
backward difference calculations that use only two points
3-110
to compute the derivative to very complex algorithms which
•
utilize several points on either side of the point in
question. It is sometimes necessary to try several different
methods in order to find one which best smooths the data.
A method which does an excellent job of smoothing data with-
out altering the shape of the derivative curve was published
in 1970 by Akima:'4 This method utilizes five points, two
before and two after the point of interest, to compute the
derivative. Its major disadvantage is that it requires a
computer solution. Other simpler differentiation schemes
were tested and published by Bourdet, et al"
The derivative type curve implies that pressure should
be differentiated linearly with time, i.e.,
d (Ap)
^p dt
_ dpD ^
p'D d tD CD)
Recent studies45 indicate, however, that it is best to

differentiate pressure with respect to the natural
logarithm of time. In this case, the pressure derivatives
take the form
, _ d(o)
I-pl n d in t
dpD
pDln
d [ln (tD/C D) ]
Taking the pressure derivative with respect to the natural

logarithm of time does not alter Fig. 3.42; however, it does
alter the definition of the pressure function on the ordi-
nate of this curve. From Eq. 3.56,
3-111
• _ kh
PD 141.2qBu Gp
and
dpD = kh d (Gp)
dt 141.2qBu dt
Using the chain rule of differentiation,
dpD'
d[ln(tD/CD)] _ kh d(G ) d(ln t)
141.2qBu d ln t dt
d[ln(tD/CD)] dt
This equation can be rearranged to the form
dpD kh d (L p )
d ln t)
• d[ln(tD/CD)] 141.2qBu
kh
or (3. 100)
pD1n = 141.2qBu "Pin
If the derivative type curve is to be used to analyze

pressure transient data where pressure is differentiated
with respect to the natural logarithm of time, the
dimensionless pressure derivative on the ordinate of Fig.
3.42 must be redefined according to Eq. 3.100.
The derivative type curve can be used to analyze
pressure drawdown data for permeability, wellbore storage
factor and total skin factor using the following procedure:
1) Redefine the dimensionless pressure

derivative on the ordinate of Fig.
3.42, according to Eq. 3.100, as
, kh ,
• pDln ^ 4 1.2qBU Apt
3-112
2) Differentiate the pressure drawdown
•
data using an appropriate method to
determine the value of Apin corre-
sponding to each measured pressure,
i.e.,
_ d(Ap) _ A(A P )
'Ap t A ln t)
In d(ln t)
3) Overlay the derivative type curve

with a sheet of transparent paper
and draw horizontal and vertical
log-cycle lines to match those on
the type curve.
4) Plot the pressure drawdown data
on this paper with dpin on the
ordinate and t on the abscissa.
5) Check the early-time points on
the data curve to see if they
form a unit slope line; if so,
draw in the unit slope line.
6) Check the late-time data to see
if they form a horizontal line
which indicates the occurence
•
of transient radial flow; if so,
draw in this horizontal line.
7) Place the data curve onto the
derivative type curve in such a
way that the unit slope line over-
lays the unit slope line on the
type curve, and the late time
horizontal line overlays the hori-
zontal line on the type curve
which corresponds to PDln - 0.5.
Notice that when both the unit
slope line and transient radial
flow straight line are present on
the data curve, only one match
position is possible and, according-
ly, the accuracy of the match is
greatly enhanced. If either, or
both, of the straight lines do not
appear on the data curve, move the
data curve relative to the type
curve until the best match is
determined.
•
3-113
• 8) After the match is obtained, select a
match point and record the correspond-
ing values of (Apin)M and ( pDln)M from
the ordinate and the values of (t).',1
and (tD/CD)^^ from the abscissa.
9) Record the value of (CDeZS)M from the
curve which the data match.
10) Using the match values of pressure
derivative, compute permeability from
Eq. 3.100 as
141. 2c^B (pDln) ^4

k = -^-- (3.101)
`^plnM
11) Using the match values of time, compute

the wellbore storage factor using
Eq. 3.88:
(t) M
0.000295 uh tD CD)^1 (^ 88)
7-
12) Compute the dimensionless wellbore

storage factor, CD, using Eq. 3.59:
C = 0.894C
(3.59)
D ^cthrw
13) Compute the total skin factor, s,

using the match value of CDe2s.
From Eq. 3.89,
1 (CDe2s):M I
s = 2 In (3 .89)
CD
The foregoing procedure can be significantly enhanced

if a simultaneous match of pressure and pressure derivative
is obtained. This is made possible because both the
Gringarten type curve and the derivative type curve are
^ plotted on the same graph. If both Ap and Apln are
plotted on the same data curve and the curves are simul-
3-114
taneously matched, both matches should correspond to the
0
same value of CDe2s. This procedure will simultaneously
take advantage of the characteristics of both type curves
and will result in greater confidence in the final solution.
This procedure will be demonstrated by the following
example:
Example 3.13: Analysis of pressure drawdown data using the

derivative type curve.
Problem. Use the derivative type curve to analyze the

following pressure drawdown data for permeability, wellbore
storage factor and total skin factor. Compare the values
of permeability and skin factor with those calculated using
conventional semilog analysis.
11 x 10 -' psi -i 0
P i = 3100 psia c=
1 _ 6 - 1
qo = 270 STB/D cw = 3 x 10 psi
qw = 0 cf = 4 x 10- psi
qg = 88 MSCF/D So = 0.6
R s = 340 SCF/STB S W = 0.4
B o = 1.15 RB/STB P O = 1.7 cp
0.-193 r w = 0.4 ft
h = 24.2 ft
t,hrs pwf,psia Ap,psi ^'p'ln

0.0167 3053.9 46.1 60.40
0.0333 3018.0 82.0 73.74
0.0500 2489.4 110.6 79.07
0.0667 2966.2 133.8 84.21
0.0833 2947.3 152.7 85.68
0.1000 2931.6 168.4 85.49
0.1167 2918.6 181.4 83.26
0.1333 2907.7 192.3 80.71
3-115
C 2898.4 201.6 77.18
0.1500
0.1667 2890.6 209.4 72.96
0.2000 2878.0 222.0 66.27
0.2333 2868.5 231.6 58.41
0.2667 2861.0 239.0 52.21
0.3000 2855.1 244.9 47.40
0.3333 2850.3 249.7 43.36
0.3667 2846.3 253.7 40.35
0.4000 2842.9 257.1 37.05
0.4333 2840.0 260.0 35.69
0.4667 2837.4 262.6 33.15
0.5000 2835.2 264.8 31.48
0.5833 2830.4 269.6 30.15
0.6667 2826.6 273.4 27.64
0.7500 2823.4 276.6 26.12
0.8333 2820.7 279.3 25.88
^ 0.9167 2818.3 281.8 24.52
1.0000 2816.1 283.9 24.04
1.1667 2812.4 287.6 23.76
1.3333 2809.3 290.7 22.97
1.5000 2806.6 293.4 22.88
1.6667 2804.2 295.8 22.68
1.8333 2802.1 297.9 22.35
2.0000 2800.1 299.9 21.68
2.5000 2795.3 304.7 21.39
3.0000 2791.4 308.6 21.40
3.5000 2788.1 311.9 21.38
4.0000 2785.3 314.7 21.11
4.5000 2782.8 317.2 20.98
5.0000 2780.6 319.4 20.94 °
5.5000 2778.6 321.4 20.95
6.0000 2776.8 323.2 20.58
6.5000 2775.2 324.8 20.55
7.0000 2773.7 326.4 20.40
• 7.5000 2772.2 327.8 20.18
8.0000 2770.9 329.1 20.10
3-116
8.5000 2r69.7 330.3 20.21
9.0000 2768.5 331.5 20.65
9.5000 2767.4 332.6 20.54
10.0000 2766.3 333.7 20.65
10.5000 2765.4 334.6 20.60
11.0000 2764.4 335.6 20.53
11.5000 2763.5 336.5 20.65
12.0000 2762.6 337.4 20.33
12.5000 2761.8 338.2 20.13
13.0000 2761.0 339.0 20.63
13.5000 2760.2 339.8 20.16
14.0000 2759.5 340.5 19.76
14.5000 2758.8 341.2 20.29
15.0000 2758.1 341.9 20.80
15.5000 2757.4 342.6 20.03
16.0000 2756.8 343.2 19.38
16.5000 2756.2 343.8 19.80
17.0000 2755.6 344.4 20.40
17.5000 2755.0 345.0 20.80
18.0000 2754.4 345.6 20.92
18.5000 2753.9 346.1 20.31
19.0000 2753.3 346.7 19.59
19.5000 2752.8 347.2 19.68
20.0000 2752.3 347.7 20.00
20.5000 2751.8 348.2 20.50
21.0000 2751.3 348.7 20.81
21.5000 2750.8 349.2 20.89
22.0000 2750.4 349.6 21.69
22.5000 2749.9 350.1 22.52
23.0000 2749.5 350.6 20.44
23.5000 2749.0 351.0 18.93
24.0000 2748.6 351.4 19.40
The pressure derivatives, Opin, presented in this table

wore computed using the method presented in Ref. 44.
3-117
•
Solution. A log-log plot of the pressure derivative, GPln'
as well as the pressure differential, dp = pi-pwf, is
presented in Fig. 3.43. It is observed from the curve which
represents pressure derivative that a unit slope line is not
present; however, the late time data appear to form a
horizontal line which indicates the probability of transient
radial flow. These data curves were overlayed onto the
derivative type curve in Fig. 3.42 to obtain the match shown
by Fig. 3.44. Notice that the late-time horizontal line on
the pressure derivative data curve overlays the horizontal
line on the derivative type curve corresponding to PDln 0'S'
this control was of significant assistance in obtaining the
final match. Observe also that both data curves match type
curves which have the same value of CDe2s.
From the match point indicated on Fig. 3.44,
('^pln)M = 100
(PDln)M = 2.5
(t) M = 10
(tD/CD)Ni = 850.
Also, from the data match,
(CDe2$)M = 104
Permeability can be computed by substituting the pressure

derivatives corresponding to the match point into Eq. 3.101:
^
k = 141.2QBU PDln M
h Apln M
k = (141.2) (270) (1.15) (1. 7) 2.5

0 24.2 100
k = 77 md.
3-118
103
• A p
n 4p ln
• ••••N
• • • •••• • •••
•••
•• 'M • • ••••
• ••
•
•
•
a •
102
^ , nn nnn
n
n
n n
w a. nn
• n
n
nn
n
n^ nn
am n nnnnn nnnn n nnnn %
10 _2
10 10° 101 102
t,hrs
Fig. 3. 43 Data Curves, Ex. 3.13.
0 0 0
• • •
Copyright © 1Y09 RLPETROL , MNun. FRANCE. All rIpft ►Mw1^d.
-I
PQ
.x •
I`r
c^.
pq
6r
.>~ N
.`^ •
^1.
t^^/C^ = 0.000295 ^h t
Fig. 3.44: Derivative type curve inatch, Ex. 3.13.

The wellbore storage factor is computed by substituting
•
the match values of time into Eq. 3.88, i.e.,
(t)
C = 0.000295 uh (tD CD î
(77)(24.2) 10
C = 0.000295
1.7 850
This value is converted to dimensionless form using Eq.

3.59, i.e.,
C = 0.894C
D ^cthrW
where ct = cOSO + cL`.5`.. +

CF 0
ct =(11 x 10 - ")(0.6) + (3 x 10-6) (0. 4) + 4 x 10-6
ct = 11.8 x 10-,Psi-1
Therefore,
(0.894) (0.0038)
=
cD (0.193) (11.8 x 10-6)(24.2)(0.4)2
CD = 385.
• 1
The value of CD from the previous calculation can be

substituted into Eq. 3.89 along with the match value of
CDeZS to compute the total skin factor:
(CDe2s)^t
s = ^ ln
CD . ^
3-121
•
1 10`'^
S =
[
Z in 385
s = 1.6.
It appears from the match of the test data with the

Gringarten type curve, as well as the occurence of a hori-
zontal line on the derivative curve, that several pressures
measured at late time represent transient radial flow. It
follows that a conventional semilog analysis can be applied
to these data. a semilog plot of the test data is presented
by Fig. 3.45. From the transient radial flow straight line
on this plot,
m = -46 psi/cycle
= 2813 psia.
plhr
0
Permeability can be computed from the slope using Eq. 3.6:
k 162.6qBU
mh
k = (162.6)(270)(1.15)(1.7)
(-46) 24.2)
k = 77md.
The total skin factor is, according to Eq. 3.29,
s= 1.151 plhr pi - log k + 3.23

m ûctrw
12813-3100 _ l0 77
s = 1.151
-46
g (0.193)(1.7)(11.8x10 6)( 0.4)2
^
+ 3.23]
s = 1.6
3-122
}. .
• -! . .i ., . 1+1^1 1 t •. ^•-+-^^.
3050 . .. , ^ ^ r } + , ^
.. . r
t .^.i
...•. . - _t +
.Î ^ N'h!
. . . . . . . ....: .... .. . ._. L
, _. j.-., . .. . ^ .. .. .^ ^. ^'..... . .. ^ .
., ^ . a
• .. '
. . . . .
..:..:,_.
.. ...... .. . -
_. .: : .... »a.:... »... :^ ..,, ..1
y ... . . _ 1 - 1 , i -..
.{ .. ^ . _.. ; .. .. ' l .
... j . . : .... ». ._.. . L ..,^ .. ..' ... . . . .^. . .. ±
, , r . _+`{
_ i l-^ , .r•w ^
.
n»+--
3000
. . . . .. .
,. _ .... . .. t Lrt k.+1.. ,-rh..^».^.,.,i. ^ ,.
. . .. 1... ..:.^,. ....- «._ . .^ .... .. . ..i. ., . ...I...._._ - ... .... . . ' , ^. .: ._^
4- { ++ f
.. .. + ^ ^ . ..y..: J 1. + rl ._ ^ . .. t . ,j. ..^1 ^ $ ^t11 1 7Ir ^ • ,r, ll h 1-
2950 . .l .Ll..l , ^ I.
H
_ .i ... . ..^
. .. . ^
W ... . _. . .. . ^ ,.., ^ ;
^ ^.__ ...:. .. . ._: . ,•. ^ 4-1 ,
`
^'
_ ... .... ..... ..»_ . .. .. ....
. . .... ._ _._..... _..... ...^ .u. ..... . . - .._ _ ..:. . . ,' . ^ d r^
. . . . . . ,- -- --. _ . .. ... .... ...... _ .... .. " ..,.- .^ .
, ^, ,:. ^. .,
I
t ..^
m 2900 .. . .. .i ^ j:
.. . ... . . . ...... .......^ .^ . _. : .j_ + t .`+^ ^ ;
. . . . .. .. _ ^ .._. . ..... i._... .. ..- ... ... ._. . . ... . _. . , .. ... _.. .. .,_, _^ + . ,:. . .i , „i . _,,.
..... ............_.^. . f.... .... .. ; . . _.....4 .
2850 ^ . ...... + .... ^ . .. îi ^y i
. - . . . j_ _ . . . ... _ _ ; ... '. :;. . _ y 4 ^^,
,. • :. »
i
' •.
^ j
JI ,
. _ .... .. ^ .... ..... .. ...... .j......^. ^ , .^ ....{ y .^ . . .... . . _.._....I ..,. . ^; _ .,_i... . .....,..:: . :::' ... ^' ._ , ^ • ^ __
^ ^ _ ^.-.-.W..____ _- • ^' . ^
y ' ^^ , ! . .
m SLOPE - 4 6 PSI/C Y C L E
:::.......::......:..._ t . ^.^ ... . -,
2800 ^..
1 : t.
.^ .
^•1
. 1 ..,
.
,. . .. . . .. ' . ^_ , ..^ . _
'
. . . . . . . . . ...... .... . . _ ,. . . ._. _. .a . . .. . . . . . ....
+ =
•• 2 , 813 p S IA ----r-
. . P 1HR ^
.........
....
..........._w.__ .... _» .. . .; .
.. ............ . f . t
2750 .. ..., ..
.. ... .,. .. : ^
:... . . .... ........
._..__ ...... .....•^-. ... .
, .. ,f , . .^ 4 l
. ........ ......_......_....... . .. ... ^_. ^ , . . .. .. .. .: .. .. . . .. . ^, . . ... ;_.: .: .. . _ .. ,1.._ t ,. .:
...,. . ^.. , -i , + ' ., .,, . .. .. ... G l .: .^..
. . . ...... _ ........... ...: .. ( . ., ^ .^ ^ ff

.. ._...... , i.
_ . . .... .........._ ,.._ ^ , „ .. . . . . . . . . ... . . __. ..^_^ . . :: ..{ _ , . - ^_
_... ^^ .... , t: r*.,- ^-^
2700
10-2 10-1 10° 101 102
t, hrs
Fig. 3.45: Semilog plot, Ex. 3.13.
0 0 9
^
A comparison of the permeability and skin factor
computed using the derivative type curve and the
conventional semilog plot shows that the two methods are
in excellent agreement.
The derivative type curve offers two major advantages

when compared to type curves based on pressure differential,
Ap: First, when both wellbore storage and transient radial
flow data are present, the two straight lines which occur
at early and late times impose a control on the matching
process which makes it easier to obtain a unique match.
Second, the derivative curves are more sensitive to values
of CDezS and, accordingly, cause curves for different
values of CDeZS to have more distinctive shapes; this makes
it easier to determine a unique match and results in more
0 accurate values of permeability and skin factor.
Derivative type curves have also been found to have
application to the analysis of pressure transient tests
from wells in heterogeneous reservoirs. Heterogeneities
which have been studied include infinite conductivity and
uniform flux fractures" finite conductivity fracture S46
and dual porosity systems such as naturally fractured
reservoirs ``s'``''`à These types of systems are beyond the
scope of this chapter, but will be discussed in subsequent
chapters.
•
3-124
VIII. TEST DESIGN
•
If satisfactory results are to be obtained from
pressure drawdown data, it is important that the test be
properly planned. While test design considerations will
vary depending upon the objectives of the test, the
following factors should be considered when planning a
test:
1) Record all pressures downhole; surface

pressures are not adequate for most
calculations. If a pressure gauge
cannot be run in the hole, which is the
case in many pumping wells, the next
best alternative is to shoot fluid
levels at short intervals of time.
2) Measure the initial pressure, p.,
before the test begins. 1
3) Maintain a reasonably constant flow
rate during the test. If flow rate
varies significantly, the data should
be analyzed using techniques discussed
in Chapter 4.
4) Record early-time data at very short
intervals of time; this data will be
useful in the detection of-wellbore
storage effects.
5) For permeability determination, the
test must be run long enough to insure
that wellbore storage and skin effects
have dissipated. The time required
for wellbore storage effects to end
can be approximated using Eq. 3.72:
t Cu(200,000 + 12,000s)
>
kh
The skin factor has a small effect on

this time and can be assumed zero for
design purposes.
6) For reservoir limit testing, the test
must be run long enough to achieve
pseudosteady state. This time can be
approximated using Eq. 3.15 if the
drainage geometry is known. When
3-12S
• geometry is not known, the minimum
time required to achieve pseudosteady
state can be computed using Eq. 3.16:
380ûctA
pss minimum k
Further, the flow rate must be

sufficiently large to give a
measurable slope on the cartesian
graph.
7) Use a pressure bomb with sufficient
sensitivity to accurately record
the pressure changes to be encountered.
For example, if the expected pseudo-
steady state rate of pressure change
is 0.5 psi/day, a pressure measurement
accurate within 5 percent will require
a gauge sensitive to pressure changes
of + 0.025 psi.
8) On tests which are difficult to run,
^ unusually expensive, or of particular
importance, run two bombs to insure
against equipment malfunctions. This
is a good practice on any well.
A good discussion of test design considerations is presented
by Earlougher3.
IX. SUMMARY
We have shown in this chapter how pressure drawdown

data can be analyzed to determine permeability, skin, and
reservoir volume. The basic procedures required to analyze
a drawdown test are simple; however, many practical problems
such as wellbore storage and boundary effects can make it
difficult to decide which part of the test data to analyze.
Fortunately, procedures are available to detect wellbore
storage and boundary effects which, when applicable, and
properly used, can give us a high degree of confidence in
our analysis.
3-126
Pressure drawdown tests can yield other information
•
which was not discussed in this section. For example,
this test can be used to detect fractures, estimate
fracture length, detect faults, and estimate the distance
to faults. These capabilities will be discussed in
subsequent chapters, and procedures for making these
calculations will be presented. Unfortunately, there are
other practical problems which can further complicate
pressure analysis; these problems, and their effects on
pressure transient data, will also be discussed.
•
3-127
•
REFERENCES
1. Matthews, C. S. and Russell, D. G.: Pressure Buildup

and Flow Tests in Wells, Monograph Series, Society of
Petroleum Engineers ofAIbiE, Dallas (1967) 1.
2. Jones, P.: "Reservoir Limit Test," Oil and Gas J.

(June 18, 1956) 184-196.

Monograph Series, Society of Petroleum Engineers o
AIME, Dallas (1977) S.
4. Earlougher, R. C., Jr.: "Estimating Drainage Shapes

From Reservoir Limit Tests," J. Pet. Tech. (Oct., 1971)
1266-1268; Trans., AIME, 251.
5. van Poolen, H. K.: "Radius-of-Drainage and Stabilization-

Time Equations," Oil and Gas J. (Sept. 14, 1964) 138.
6. Odeh, A. S. and Nabor, G. W.: "The Effect of

Production History on Determination of Formation
Characteristics From Flow Tests", J. Pet. Tech.
(Oct., 1966) 1343.
7. Kazemi, H.: "Pressure Buildup in Reservoir Limit

Testing of Stratified Systems", J. Pet. Tech.
(April, 1970) 503.
8. Gibson, J. A. and Campbell, A. T., Jr.: "Calculating

the Distance to a Continuity From D.S.T. Data", paper
SPE 3016 presented at the SPE-AIME 45th Annual Fall
Meeting, Houston, Oct. 4-7, 1970.
9. Hawkins, M. F., Jr.: "A Note on the Skin Effect",

Trans., AIME (1956) 207, 356-357.
10. van Everdingen, A. F.: "The Skin Effect and Its

Influence on the Productive Capacity of a Well",
Trans., A1ME (1953) 198, 171.
11. Hurst, W.: "Establishment of the Skin Effect and Its

Impediment to Fluid Flow into a Welibore", Pet. Eng.
(Oct., 1953) 25, B-6.
12. Brons, F. and Marting, V. E.: "The Effect of Restricted

Fluid Entry on Well Productivity", J. Pet. Tech. (Feb.
1961) 172-174; Trans., AIME, 222.
13. Odeh, A. S.: "Steady State Flow Capacity of Wells

With Limited Entry to Flow", Soc. Pet. Eng. J.
(March 1968) 43-51; Trans., AIME, 243.
3-128
•
14. Odeh, A. S.: "An Equation For Calculating Skin Factor
Due to Restricted Entry", J. Pet. Tech. (June 1980)
964-965.
15. Kazemi, H. and Seth, M. S.: "Effect of Anisotropy

and Stratification on Pressure Transient Analysis of
Wells With Restricted Flow Entry", J. Pet. Tech.
(May 1969) 639-647; Trans., AIME, 2T-6.
16. Brown, M. E. and Ming-Lung, M.: "Pressure Buildup

Analysis of Prudhoe Bay Wells", J. Pet. Tech. (Feb. 1982)
387-396; Trans., AIME, 273.
17. "API Recommended Practice: Standard Procedure for

Evaluation of Well Perforators", API Div. of Production,
RP 43, 2nd ed. (Nov. 1971).
18. Harris, M. H.: "The Effect of Perforating on Well

Productivity", J. Pet. Tech. (April 1966) 518-528;
Trans., AIME, 237.
19. Klotz, J. A., Krueger, R. F., and Pye, D. S.: "Effect

of Perforation Damage on Well Productivitv", J. Pet. Tech.
(Nov. 1974) 1303-1314; Trans., AIME, 257.
• 20. Hong, K. C.: "Productivity of Perforated Completions

in Formations With or Without Damage", J. Pet. Tech.
(Aug. 1975) 1027-1038; Trans., AIME, 259.
21. Locke, S.: "An Advanced Method for Predicting the

Productivity Ratio of a Perforated Well", J. Pet. Tech.
(Dec. 1981) 2481-2488; Trans., AIME, 271.
22. Cinco, H., Miller, F. G., and Ramey, H. J., Jr.:

"Unsteady State Pressure Distribution Created by a
Directionally Drilled Well", J. Pet. Tech. (Nov. 1975)
1392-1400; Trans., AIME, 259.
23. Agarwal, R. G., Al-Hussainy, R. and Ramey, H. J., Jr.:

"An Investigation of Wellbore Storage and Skin Effect
in Unsteady Liquid Flow: I. Analytical Treatment",
Trans., AIME (1970) 249, 279.
24. Wattenbarger, R. A. and Ramey, H. J., Jr.: "An

Investigation of Wellbore Storage and Skin Effect in
Unsteady Liquid Flow: II. Finite Difference Treatment",
Trans., AIME (1970) 249, 291.
25. Martin, John C.: "Simplified Equations of Flow in Gas

Drive Reservoirs and the Theoretical Foundation of
Multiphase Pressure Buildup Analyses", Trans., AIME
• (1959) 216, 309.
3-129
•
26. Miller, C. C., Dyes, A. B., and Hutchinson, C. A., Jr.:
"The Estimation of Permeability and Reservoir Pressure
From Bottom Hole Pressure Buildup Characteristics",
Trans., AIME (1950) 189, 91.
27. Perrine, R. L.: "Analysis of Pressure Buildup Curves",

Drill. and Prod. Prac., API (1956) 482.
28. Earlougher, R. C., Jr., Miller, F. G., and Mueller, T. D.:

"Pressure Buildup Behavior in a Two-well Gas-Oil System",
Soc. Pet. Eng. J. (June, 1967) 195; Trans., AIME, 240.
29. Papadopulos, I. S. and Cooper, H. H., Jr.: "Drawdown

in a Well of Large Diameter", Water Resources Res. (1967)
3, No. 1, 241.
30. Cooper, H. H., Jr., Bredehoeft, J. D., and Papadopulos,

I. S.: "Response of a Finite-Diameter Well to an
Instantaneous Charge of Water", Water Resources Res.
(1967) 3, No. 1, 263.
31. Ramey, H. J., Jr.: "Short-time Well Test Data

Interpretation in the Presence of Skin Effect and
Wellbore Storage", J. Pet. Tech. (Jan., 1970) 97;
32.
Trans., AIME, 249.
McKinley, R. M.: "Wellbore Transmissibility From

•
Afterflow-Dominated Pressure Buildup Data", J. Pet.
Tech. (July, 1971) 863; Trans., AIME, 251.
33. Earlougher, R. C., Jr., and Kersch, K. M.: "Analysis

of Short-Time Transient Test Data by Type-Curve
Matching", J. Pet. Tech. (July,^ 1974) 793; Trans.,
AIME, 257.
34. Gringarten, A. C., Bourdet, D. P., Landel, P. A., and

Kniazeff, V. J.: "A Comparison Between Different
Skin and Wellbore Storage Type-Curves for Early-Time
Transient Analysis," SPE 8205, Presented at the 54th
Annual Fall Technical Conference and Exhibition of the
Society of Petroleum Engineers of AIME, held in Las
Vegas, Nevada, September 23-26, 1979.
35. Bourdet, D., Whittle, T. M., Douglas, A. A., and Pirard,

Y. M.: "A New Set of Type Curves Simplifies Well Test
Analysis," World Oil (May, 1983) 95.
36. McKinley, R. M.: "Estimating Flow Efficiency from

Afterflow - Distorted Pressure Buildup Data", J. Pet.
Tech. (June, 1974) 696-697.
•
3-130
• 37. McKinley, R. M. and Streltsova, 1'. D.: "Early-Time
Pressure Buildup Analysis for Prudhoe Bay Wells",
SPE 10266, Presented at the 56th Annual Fall Technical
Conference and Exhibition of the Society of Petroleum
Engineers of AIME, held in San Antonio, Texas, October
5-7, 1981.
38. Streltsova, T. D. and McKinley, R. M.: "Early Time

Buildup Data Analysis for a Complex Reservoir", J. Pet.
Tech. (May, 1982) 1145-1155.
39. Provided by R. M. McKinley.
40. Gringarten, A. C., Ramey, H. J., Jr., and Raghavan, R.:

"Unsteady-State Pressure Distributions Created by a
Well With a Single Infinite-Conductivity Vertical
Fracture," AIME (1974) 257, 347.
41. Puthigai, S. K. and Tiab, D.: "Application of p^

Function to Vertically Fractured Wells - Field
Cases," SPE 11028, presented at the 57th Annual Fall
Technical Conference and Exhibition of the Society
of Petroleum Engineers of AIME, held in New Orleans,
Louisiana, September 26-29, 1982.
42. Tiab, D. and Kumar, A.: "Application of the PD Function

to Interference Analysis," J. Pet. Tech. (August, 1980)
1465-1470; 'Trans., AIME (198Uj 269, 1465.
43. Tiab, D. and Kumar, A.: "Detection and Location of Two

Parallel Sealing Faults Around a Well," J. Pet. Tech.
(October, 1980) 1701-1708; Trans., AIME T_1980) 269, 1701.
44. Akima, H.: "A New Method of Interpolation and Smooth

Curve Fitting Based on Local Procedures," J. Assn. Comp.
Mach. (Oct., 1970) 17, 589-602.
45. Bourdet, D., Ayoub, J. A., and Pirard, Y. M.: "Use of

Pressure Derivative in Well Test Interpretation,"
SPE 12777, presented at the 1984 California Regional
Meeting, held in Long Beach, California, April 11-13,
1984.
46. Wong, D. W. and Harrington, A. G.: "Application of the

Pressure Derivative Function in the Pressure Transient
Testing of Fractured Wells," SPE 13056, presented at
the 59th Annual Technical Conference and Exhibition of
the Society of Petroleum Engineers, held in Houston,
Texas, September 16-19, 1984.
47. Bourdet, D., Alagoa, A., Ayoub, J. A., and Pirard, Y. M.:
^ "New Type Curves Aid Analysis of Fissured Zone Well
Tests," World Oil (April, 1984) 111-124.
3-131
48. Bourdet, D., Ayoub, J. A., Whittle, T. M., Pirard,
•
Y. M., and Kniazeff, V.: "Interpreting Well Tests
in Fractured Reservoirs," World Oil (Oct., 1983).
•
3-132
•

Aa = cross-sectional area of annulus, ft2
b' = intercept of cartesian plot of pwf versus t, psi
B = formation volume factor, reservoir volume/surface volume
Bg = gas formation volume factor, RB/scf
Bo = oil formation volume factor, RB/STB
Bw = water formation volume factor, RB/STB
cf = formation compressibility, psi-1
cg = gas compressibility, psi i
co = oil compressibility, psi 1
ct = total compressibility, psi 1
cw = water compressibility, psi 1
cwb = compressibility of fluids in wellbore, psi 1 ^
C = wellbore storage factor, bbls/psi
= reservoir shape factor, dimensionless
CA
CD = dimensionless wellbore storage factor
DF = damage factor, dimensionless
DR = damage ratio, dimensionless
E = flow efficiency, dimensionless
ho = distance between the top of a producing formation and
the top of the open interval, ft
hp = thickness of formation open to the wellbore, ft
J = productivity index, STB/D/psi
Jactual = productivity index of a well under actual
reservoir conditions, STB/D/psi
Jideal = productivity index of a well under ideal

conditions, STB/D/psi
k = volumetric average effective permeability of the
reservoir in the drainage volume of the test well, md
kg = effective permeability to gas, rr:d
kH = horizontal permeability, md
3-133
•
ks = permeability of skin zone, md
kv = vertical permeability, md
kw = effective permeability to water, md
m = slope of semilog plot of pwf versus t, psi/log-cycle
m' = slope of cartesian plot of pwf versus t, psi/hr
surface casing pressure, psi
p cs =
= dimensionless pressure = kh(pi-p f)/141.'.'qBu
PD w
pD = dimensionless pressure derivative, dpD/d(tD/CD)
dimensionless pressure derivative, dpD/d[ln(tD/CD)J
pDln-
p. = stabilized reservoir pressure at beginning of drawdown
1 test, psi
intercept pressure on cartesian plot of pwf versus t,
pint
psi
= volumetric average reservoir pressure, psi
PR
flowing formation face pressure, psi
pwf =
pressure from semilog straight line at t= 1 hour, psi.
plhr
qf = formation face flow rate, STB/D
qgt = total gas production rate, Mscf/D
qo = oil production rate, STB/D
total production rate of gas, oil and water, RB/D
qRt =
qk, = water production rate, STB/D
flow rate contributed by the wellbore, S'I'B/D
qwb =
rd = drainage radius, ft
re = distance to outer boundary of cylindrical reservoir, ft
rs = radius of skin zone, ft
r wa = effective wellbore radius, ft
r wc = corrected wellbore radius, ft
Rs = solution gas-oil ratio, scf/STB
R sw = solution gas-water ratio, scf/STB
s = total skin factor, dimensionless
sd = skin due to permeability alteration, dimensionless
sf = skin due to fracture, dimensionless
sp = skin due to perforations, dimensionless
^ S = skin due to restricted entry, dimensionless
r
3-134
•
ssw = skin due to slanted well, dimensionless
st = skin due to turbulence, dimensionless
So = oil saturation, fraction
Sw = water saturation, fraction
t = producing time, hrs (t has units of minutes only
when using McKinley type curve)
tD = dimensionless time
(tDA)etf = dimensionless time to end of transient flow
(tDA)pss = dimensionless time to beginning of pseudosteady
state
tetf = time to end of transient flow, hrs
tpss = time to beginning of pseudosteady state, hrs
Vwb = volume of wel3bore in communication with the producing
formation, ft
z = length of fluid column in annulus, ft
zm = distance from top of producing formation to the middle
of the open interval, ft
a = angle of deviation between wellbore and bedding plane
of formation, degrees
©p = pressure differential, pi pwf' psi
Ap' = pressure derivative, d(Ap)/dt
pressure derivative, d(Ap)/d(ln t)
"pln
^ps = pressure loss caused by skin zone, psi
(Aps)D = dimensionless pressure loss due to skin zone
AVS = volume of skin zone, ft3
AVUD= volume of undamaged zone, ft3
AV"wb= change in volume of wellbore fluids, ft3
a = total mobility of reservoir fluids, md/cp
viscosity, cp
p = density of fluid in wellbore, lbm/ft3
= porosity, fraction
•
3-135
E
Equation Number Equation

in Text
3.1 p = 162.6 ^(log kt - 3 23)

wf pi z
ouctrw
3.3 pwf = m log t + b
3.4 m = - 162.6 k qB^j
3.6 k = - 162.6 ^
s.7
ûct
• i
' tetf 0.0002637k (tDA)etf
380ûctA
3.8
tetf k
3.9 p = pi - 162.6 ^ log 1 4A 1

wf
L1.781Car` j
0.2339clBt
Ah^ct
0 1
3.10 pwf = m1t + bl
3.11 mt = - 0.2339^c. B
AhTc t
•i
3-136
• 3.12
ph^ _ - 0.2339qB ,
m,c
fts
[2.303 (p )
2. 313 (Plhr pint
Pint
3.13 CA = 5.456 11
kt ^Z
3.14 r d = 0.029
[4ctj
0uCtA
3.15 tpss 6.00026371(t DA pss
3.21 ^ps = 141.2 k qBp s
3.22 qBp (log

pk f= pi - 162.6 k kt - 3.23
ûer
t 2w
• + 0.87s)
-S
3.25 rwa = rw e
1.151 F^lhr pl - log k + 3.23

3.29 s =
2
ûetrw
3. 30 Ops = - 0.87 ms
3.31 s = sa + sr + sp + st + S + S
0.825
3.33 sr = 1.35 r
h - 1 fln (h kH/kv + 7)
[0.49 + 0.1 in (h kH/kV)]ln rwc-1.95}
•
3-137
3 . 34 s S w _(a' /41) 2' 06- (( a'/56) 1865 log(hD/100)
•
3.38 E = pR pwf Aps
PR pwf
3.39 DR = 1= Piz pwf

E
PR pwf - Aps
Ap
3.40 DF = 1 - E = s
PR pwf
3.41 = 1
aafter E ' before
144A
3.44 a
C = 5.615 p
3.45 C = Uwb cwb

•
5.615
0.00708kh(pi
3. 57 = pwf)
PD c{Bu
3.58 t = 0.0002637kt
D 2
ûcrw
t
3.59 c = 0.894C
D ^ cthr w
3.66 C = Bt
24 pi - pwf)
3 . 72 t > Cp(200,000 + 12,000s)

kh
•
3-138
i• xt
3. 73 pwf = P. - 162.6 Rh [log t Z
t ^crw
t
- 3.23 + 0.87si
3.74 qRt = qoBo + (1000qgt - q°Rs - awRsw)Bg
+ qwB
w
k k k
3.75 a = ° + 9 + w
t uo ug uw
B
3.76 ko = - 162.6 q°mhuo
i• 3.77 k
w
= - 162.6 awBwuw
mh
7 - 162.6(1000qgt
a oRs q wRsw)BQug
3.,8 kg
mh °
plhr
3.79 s 1.151 pi - log 't + 3.25
m ^ctr^,
RAMEY TYPE CURVE
k = 141.2c{Bu (pD)M
3.81
h (AP) M
McKINLEY TYPE CURVE
C_ F5.615 ApC qB
3.85
^ qB ]%1 5.615 (AP) M
3-139
•
r kh ^ 5.615pC
3 . 86 k
L5.615^C ^^ h
GRINGARTEN TYPE CURVE
141.2qBu (PD)N1
3.87 k =
h
'-PT^
(A I
(t)
3.88 C = 0.000295 uh tD CD)M
1 ! (C De2s)\1
3.89 s= 2 In ^ C
D
DERIVATIVE TYPE CURVE
, tD = kh ,
3.95
PL CD 141.2qBu tAp
•
, kh
3.100
pDln 141.2qBu ^p,In
k = 141.2aBu ( pDln),%t
3.101
h (^pl^til
•
3-140
PROBLEMS
•
l. The discovery well in a consolidated sandstone

reservoir has been produced at a carefully regulated
rate of 95 STB/day. During the flow period, bottom-
hole pressures were recorded as tabulated below.
Other rock and fluid properties are listed. You are
requested to estimate formation permeability and well
drainage area. What conclusion does the drainage area
suggest?
pi = 3830 psia rlti, = 0.25 ft

h = 16 ft co = 14 x 106 psi -1
q = 95 STB/day cw = 3 x 10-6 psi-1
B
u
= 1.2 RB/STB
= 16%
= 0.65 cp
So = 78%
Sl^ = 22% •
t, hrs pwf, psia t, hrs ptiti f, psia
0.20 _ 3605 22.90 3493

0.50 3587 30.55 3486
1.00 3569 45.83 3477
1.50 3559 61.11 3469
2.30 3550 76.38 3464
3.06 3542 84.03 3461
3.82 3537 91.66 3458 ! I
5.35 3529 99.30 3455

6.11 3524 114.58 3450
7.63 3518 122.22 3447
11.45 3509 137.50 3441
15.27 3502 152.77 3434
19 . 10 3497 •
3-141
• • •
3550
3525
17-
_
^
^.. . . . , .. ....- .__ - - -- .{_._: ^:;
3500 .- --- - -- , ._,....^ . ;...
__ . - -j - -
- ;, .. .: .. _ -. - ---^-
77
ca
=-^1^?^ - ^;^ --^ - ^ - --_
3475 - r ^ - -- - - - -
=
w 3
t-F•-
- ^_ -^--T
._`_ - F: - • ----
• } _;
^
- J
_.^ 4
T.
.- . - . , . . - .-.-. 4-
3450
; .
3425
---'-}-
L T
J , rl _t
i
r i
_L_
- . .1 ., _
L L
1-7
3400
0 20 40 60 80 100 120 140 160
" t, hrs
Fig. 2P.2: Cartesian Plot, Problem 1.

3700
^ I^ ! 41rr,^^: " .J 1 I. ^
Ti II II„ i n^
(^ •{I --. _ ^ - :}J .. 1
PROBLEM 3 .1
I ^ +
i I I^ ..r .
1 1
•^, f a l f,ll
1 ^' .l: l , ÎI^^ ^ II^ ,î• a;;
, '^ I l , .^
't' ,!il ' , 1 ^l, ^^? -•i ^!^++{
-
^ ^-
- }
4r
-
3650 L' { l'„ , i II
y 1t L l 7 r
. l_ ^t , ..i .w; ;-
14- - r- I 4
3600
(. h !!+ 1 Ir __ ..
I. .. . :. ,^.
^ . I. I.
: ' '^ „^ • . . ^ ^ . ^ .- - .. . ^.
.^ l ,+ I .. .
I ^ ' __-•
I .:7 _ I ^.. ^^
'
, y I . . . i:. ^ . . ...^ • I ^ {
+ + .L..-.. ^
T ' II I I ,, -^. I
3550 I
. , Î^^ ^ . .•
{ ` . .. L^^ . . .,' ! -.-^-:., L:.'.- I^ . ^ ^., ^.._ _.... ._.. . ,,.
;+ _. ..:
., I . : : ^_ ^,
- .. .'', 4 '
., _' ' ^ ..: . . ^ ...Î : . _ - •^ ^.I ^ ,.. . ^ , I
. I. .
(.n1 ^
.•. { ^ f ;,i ,^{'
*
', ^^^^ _ ^^ ^ I ' :...^ . j I . Iy ' t
1 ^
^_ - ^ ^ I .. _
-{^
w •
J r :...... ._ .^^
j, ll^ ,. _ __. -._ ^ ,Î ' Î I f} ' 1^ ^ ^
{ .,
l:... ..
3500
'ti7
7. ' ., Tt . .tl .
11
.^:11... ..
I 1
t
.
I
:. I I ...I :_. 11„
44 ^
: , LI { •I ^ 4 'll' h+ J I ^.
I '' I' Ii ^' t t, i t:t
T * 1i? i i
4t lili '
3450 2I
, i.
J {; , ^ tl i°rv'1 -t p ri
4- 1
.: ,.
, _ ..
3400 I I{I
10-1 1 10 102 103
t, hrs
• • •
^
2. A well completed in a consolidated formation is placed
on production at a rate of 303 STB/day. You are given
the fluid,rock, and pressure drawdown data. Use your
knowledge of transient analysis and, if possible,
estimate the permeability and drainage area of the well.
A semilog plot of the drawdown data accompanies this
problem.
q = 303 STB/day = 12.9'.

Bo = 1.116 RB/STB rw = 0.32 ft.
6
uo = 0.88 cp co = 6.9 x 10 psi
So = 7S% c`V = 3.2 x 10- 6 psi 1
S = 25% h = 8 ft
w
t, hrs pwf, psia t, hrs pwf, psia
^ 0.167 4714 1.66 4699

0.25 4708 2.50 4698
0.33 4705 3.16 4697
0.50 4703 6.67 4693
0.67 4703 10.0 4689
0. 83 4702 13. 3 4686
1.0 4701 20.0 4683
1.167 4701 23.3 4677
1.33 4700 30.0 4667
1.50 4700 36.7 4659
3-144
4720
+j t} {i rl 1
' r• i ^
- . t'- •^. • - r - ^ .
. ^ I
r-, + PROBLEM 2
4710
ti.^ ,
- .. - I ^ ;. -
, . . . . . . .^ .^
-+^ -, -
1- ; --I- ^
_ _ --t ^ - 'f^-r - ---+- ^ - 2 N -rt«+ • ^-r-r i m*,^*r-- ^ - ..^ .
-- _ T i._•
r i J1 1 i
4700
^._ - •r^ . L.
..- ^•^ -- 1"' :^-
- - ' ;,
I• ^
y ___ _ I , , . . ^' - -•'
^ .^ ..-^ -.-•. . A .. I
^ _- t • ^ ^ _`l
-^l" ^ ^^ _•-^_
"'"
^^ T
__ ^ ^ - , ^ __ ^ _..- • _ _
- - •;- .- -•,• , - -r---- • --^ T4 ,^, yl+...rl.e-.µ
-^-- • - -- ; -- -1 - - -
4690 •; .^ _ 3 I
--^^ _ ---
- - - -, --- -- ----- _ -- -
-- ^ `!-- - --f-^- ---^ -•-, - --- ^ -^---^-^-
- ^- . . .. . - -
--- -+------^ ^ _ -,, ' ' -T----!- - , -• • .
w - --_ --- _ _- - • - - -
- - -- --^^ -••---- .._-_---i.^: _-.- ---- ..... . __r ,
4680 '
.{.
,
46 70
^"^ - - - ^' ,^_ ^ •
_ ^ ^ ' _++^ ^.. . _ . _ . î: l ,. ; ^
-- - -{ I ^1 .
-- ; ------} - -- ? ^
-- ^ - ~3I-±-^- ^
4660 -.
- ^^ - - - - • -- - - a. ^ .. _ . - :^: ^,
. _ .. ... . ^^
,_ . . • - --_^ _ __ . .-.. . _ . - ._,. _ .- '^' - 1 ^_ . ' ; - ::

t,
4650
0.1 1 10 10z
t, hrs
s • s
• ^ ; ± # ^
,
{ (
{! t': lr f
i
^+ r _ }. , y { ' •i r Î' f ^ ,I r , . f^ . . , ; 1
I , 1 1 1 , ! ttt . ., . : .,: ;. f
Ll t I+t- , J { ' 1:; r{ . t:'^ + ;'+
.l ^ ^ , tt } {
{ ^ t- ^ r
^
^ ^ i .i ^ -,,. ` ! .1 1-• ^ • ^+' ^ ^ lii' t f' ,l^ ! ^ I^' +t:f rf^l I I ^Î
^.il , l • ^ ^^ ^ # r.# i ,+{' ,Î^ 'i ' ^ at j:, tr ! r f- "+ t t;^f I ^ ^ ^ 'I} ,` , . ` '1
T Tf r
^. . + 1tf^ r
,• -7 , '^^^ l.
11 i1T 'j,- ` ^ Y I }^ , •{ t r•
.I . ^ „ .1 - t7
< <y r ^ T f r • t Tt.^ lt 'I r Î t: + ; , tr t ^ . ,

Iî ^{ , ?i,^ I r I ! ^ ; $1 ^
} 4 + ^ ^ 11 t
1 „ 'li 4' J^ l } }f^
•^ 1 1 i
1 ,
r ^
z .` r *
q ^^:? 1 l ^ l y l I
f+t-
f
^ I :I i. '^^ IHj
q
r 1 ; ,r ^ , f + } r^ ;. +
a•
rc^
,
;
U - 77
Z 1 4^1 411 j t
!
W
l7 ^ !j ` .r
N - r •+ t^ ^r i i^`t^ r-^ .^^lt r I r i= I i,î ^ î { ;l i
r
TTT
W
j ; + f 'I ff il ^
q L r^ ,^
4T
. .
r ! 7 !
I r .I Ifî, .14. +: ^-. -;+•
' i t ^ ,1 , . ,. r^^ ,
,r^; r., , , } ^ :f ' ^► if ^( ,^ ! ' ^ ;- .r. ^ ? f
• r
' ;fl;
,J '
, i t^
.
f • 7^
!
, ^(
1
I
^
'i
- ^ 1
,t^
t.
^ : , ( ':' ;!
^:
^
;1.,11:^
r
^f: f, ^
ir
rf
7'
i ^ 1} ' i' 4`^' ' ; ^^^!^ ' î ( ^^f, +*, ; ^^ ^ ;
.
^ ^:I
:
: +
. ^
i l. } ii
rr t
r
`( ^
•
rÎ . ;
7 t{ II + i i T { J + l+ i, {I . r{ •t
^ } t ^^^: ^ ^ ^. ^ ; ' -ii1(lr Y, ^ r ^ • ^ I ^ll' ^ i j^r ff
^ -^ ! ^ ' (: J ^ : ^ t'I• ^ ^ ^ ! !
i^+^
i
? I
, ; 11 =r r} }rr i y; -+!
u {` ^ r' i }}
^ ^ I^ 1 , -- 1
+^
1,f!
?'
I
a i t
a < U_ l
Z !}
7t + , ^ I+ 1 : ^
y.. t
^f, 'jr. t r i (' I ^ , .I . . ^ 4 •} r ' , = yi ^-r ^1, I---i, '-+-+--Y
Z a
W
^ i
G a
N q
W N
11 !: '
t r l ^
C i
ox
qq .^ ,- ' +^ •' •, ^. I' I L '
:1: +
: j - :l ?_ r• ; ,+ ,r.,
• E :11 ^
N N
q
} # t t 1 { +1 ^ T
] ^ • r 1"' ' - I . i- I 2 .^ .; IL rrr i1 4
_ I ! +
i lfr :irr t ^^
{
+ ^ f^ !!, 'l +rt?
,,
I r ` t 111?f
^ ^ .I,
, I
I
,
r
7i
I
• ^`+
ir
^ r^r `h-(r
I{
?
7 .
.« ! t F
-• t r r
471
t+ 1 T7T
.?t
m
;1 ;:++ t +^ f.;i I (1: 'tÎ
{ F 1
{j. ri t+` HI r
j
I {_' r l1 7 i.I
* -'.. - L
,: + ' :' j .}: '' •,.^^' ^-1: ^' :,:
f ^1~ ; _.r.
.I. i•i J µ rl. ^+îr` ?l.
•^ r• (. i t . t: }t?
i
r i ^ l{' i ,, 11 { 1 1 }^^ i f : {
: fj f : ( _
^
_ y # r=
^{ .1 1: 1- ^ ^ j1^
' rf F» ♦ 1:
Y } + y. . 1 ..
^^ ^' , t +Y + i
1 :^;^
r ^ +^f ` ^ • T.. 11 4 , ' ^ 1 t^
.
I r{ , j ,1; r { ^^ + F^
3-146
3. An oil well is located in a field developed on 40-acre
•
spacing. A 24-hr pressure drawdown test is run on the
well. When the pressure data from this test are plotted,
would you expect to see any boundary effects? Explain
the reasoning behind your answer. Other reservoir
information is given below:
k = 22 md
= 12.6%
2.3 cp
-s i
ct = 1.86 x 10 psi
4. Given below is the rock and fluid properties data for

the discovery well in a gas reservoir. Many of the
reservoirs in the vicinity of the discovery well
are small (less than 500 acres), whereas the normal
well spacing is 320 acres. Accordingly, your company
would like to pursue the possibility of running a
reservoir limits test in this well to see if the
reservoir is large enough to justify drilling another
well. Based on the data given, would you recommend
such a test? Explain the reason for your answer.
8.2%
h = 27 feet
k = 2.2 md
u = 0.0156 cp
ct = 2.5 x 10 4 psi 1
5. An extended pressure drawdown test was run on a new

oil well to determine the interwell permeability, and
to determine if the well experienced formation
damage during drilling and completion. The subject
formation was completed by perforating the top 12 feet
with a density of 2 shots per foot. The perforation
penetration is estimated to be 8 inches. The well is
3-147
•
drilled on 80 acre spacing. The vertical permeability
is approximately two-thirds of the horizontal perme-
ability. Pressure data, as well. as rock and fluid
data, are presented. You may neglect turbulence
effects, and you may assume the reservoir is not
fractured. Does this well need a stimulation treatment?
If so, what increase in production can be expected?
(Note: The early time data in this test are affected
by near-wellbore non-idealities)
q = 200 STB/D
u = 3.5 cp
= 0.15
h = 20 ft
B = 1.20 RB/STB
= 5500 psia (measured)
pi
^ rw = 4 inches
c t = 33.94 x 10 ^ psi
t, hrs p«f, psia
0.1 5041
0.2 4880
0.5 4758
1.0 4718
2.0 4686
3.0 4669
5.0 4647
7.5 4630
10.0 4618
20.0 4585
30.0 4579
50.0 4554
^ 70.0 4538
100.0 4530
3-148
5100
5000
4900
4800
4 700
4600
4 500
4400
0.1 1.0 10.0 100.0
t, hrs
•
6. An oil well was placed on production at a constant
surface production rate of 45 STB/D. The pressure
decline at the formation face was recorded and
tabulated. A semilog plot of these data accompanies
this problem. Other reservoir rock and fluid
properties are given. Estimate the formation
permeability.
pi = 3400 psia
q = 45 STB/D
u = 2 cp
ct = 13.2 x 10-6 psi-1
B = 1.1 RB/STB
h = 15 ft
rw = 0.316 ft
0.10
•
t, hrs pwf, psia
0.1 3399.1
0.2 3398.1
0.4 3396.3
1.0 3390.7
2.0 3381.4
4.0 3362.3
10.0 3308.3
20.0 3267.5
•
3-150
3450
PROBLEM 6
1
3400
3350
^
2
4-
Ln 3300
3250
3200
0.1 1 10 102
t, hrs
^ 0 0
•
.o
•
c
L
^ W
C J
^ U Y
f U
7 1- M
X
J a W
J U
]
vi
a
z
3-15.2
7. The pressure drawdown data for an oil well are presented
•
along with other pertinent information. It is required
that the wellbore stor<ag'e ^factor, formation permeability
and total skin factor be computed. A semilog plot of the
data is presented.
t, hrs 4 = pi-pwf, psi

pwf, psia
0 3700 0
0.I0 3640 60
0.15 3611 89
0.20 3583 117
0.25 3556 144
0.30 3S30 170
0. 40 34 86 214
0.50 3440 260
0.60 3401 299
0.80 3332 36S
1.0 3273 427
1.5 3159 541
2.0 3069 631
3.0 29 39 761
4.iT 2892 808
5.0 2857 863
6.0 2805 895
8.0 2762 938
10.0 2736 964
12.0 2730 970
15.0 2690 1010 .
20.0 2662 1038
p = 3700 psia 1.3 cp

i
ct = 14.0 x 10 fi psi-1 c^ = 115 STB/D
= 18.6% B = 1.73 RB/STB
r w = 0.25 ft h = 8 ft
3-153
• • •
3700 ! , .! 14
!
3600 t' 1 , +
r*; 7 1- 17 t _4
Ilk PROBLEM 7 t.^
3500 +T WI
=Y a:.i ^
! I I t 1 I
I
3400 _
i I
T ;
' "
! „ , ; ,' „ • f
t, ! 7,
3300
r+ ~=- ^ ^ " .'
,
3200
...;^._..:-
- ^ - ---, ---^- ^ _ ---
LA ---^- ~_ 1 ..... -
3100
^ ^ _} - - ^ ^' -^---r---
- -
w `'•'
.i } ,____^^^ _•-.^._{-. .{ .._ y^? 1 `_- ^
3000
cn
4-
1 - --T-t-- --r , -^ ,- t - ,-; ^;' -- • ^--r-- ^- -._ -^-. , ----^ - i . - , '-- -.-4--- ! *---,
, _ _ , - . _^---- __ ! • -- . ., --1 ^ - -_ _^__L_ .-,-'
' _ ^>. _ ... .1 . $ . t ^l ^: -^---}--+.} -- ,-^' _ ^,-. ^ E, . - .
2900 , ------ r ' -i --- ^ -; -i - -
+ ' . ^ ' ' , -. . --^- - - ^-'^ - . . . ^ - ' - ---r-- + - - - - -' -i --? T^ l ;
_^_ -T
, ^ - t! . . ..- , .. . _.. _ ...-_ -- ^ - ` r ' ` 1_• , ---`^-^ -+^r^--• Y
^ i
2800
^ ^ -i . . ---°^-'-+ - -
-^ _ ^ • _ j. - . ^ .^ _j._ } - - .-1_ ' I , .-..-}-.J.' -: _ .-_. ^-a-- I

=----•--r--__.-_1 -. .l --
^} a!+-^^. {1't _'^ .
^S •-^•-'-_ --1--•^ ± ! ^
. ^ .
2700 . . . I `
^+-
^ ._-'_..1- +J . 'u' i '_ Y .
_-. '^ 7 1. : '. I'
,
:-J-•^ .^. ! _ ^} . + -. - t'^
.^ +
.. .
^ ^ ^.. . y ^ "'.i . . , ^+ I• t'f . " i
^ ^ I
! . ^ .^J .1 ,^
-•--.• ^ . .._.^--^.^ ^ '^-I'^'
---1. _^t ^• _.-- T'•--- '^ . , . ."r-r-^
^ Î -1- -..?
--I--T ^^---•-r-^---.-T--
2600 - - .. - ,-1_ ^--+ - + r- . i.. `.. } - .'. -. . - _ ^ -' T --• _ . T' _^-^ `_
. _ __ _ T.-.- T , , __ i '. . ' _ -• _ • + -- : -. - ^ .'-__
-` -^-^-t-1-^-f-•.- _ .
_`-' t._ ^ ^^-F ^ _ 'a ._. -^-+-'_'^ -t . "^-iT-- ^
_ I , . T-"_ _ . .. . .*.. .
_ _. -• . -r^- , '__
2500
0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0 8.0 10 20 40
t, hrs
NO. 340-L22 DIETZGEN GRAPH PAPER EUGENE DIETZGEN CO.
LOGARITHMIC MADE IN U. S. A.
2 CYCLES X 2 CYCLES
iJ 41 A tA G J m 4D 0 t^ m h W 6i J CD mO
t^ . . • , i'^ r{,
i- ^, Ih + ,-r,+ ,- +. ', -'+-- r •. -i ^--+"^-^--_ 1 t+^ i.
+4 4 -s.: -#•{ ' '" rfi ,{ +.-
t
''1
, ,
pil
'
i ^. r$ « ru,
r r}^ + }
+ 3 s ^} s
f,^fi
F Lkt
Id
} _.• - ^ ^- ^ {# ' ^ +^
ti 1 "+1- +++F - --1
r t + 4
t ^1: -.-r-• y-î: . 1 . .', 1 I ' • +- ^ .. 4t : . î ^
I-+ -4 .... . ... ..
T.
^^ ::f _ ..: i- t . 1.. •i . ^ -'r}'_ + ^ -^ t

v T^ ^^^ ^ ' - { + - ^ ^
_ . - r - ---- - . ... - ^
, ^- - : .: «-^. =3s"_'.-
. ^ . ..:c . ....'^..! .-. .•: ..,.^ .i .. .t ..i .. ..: . ._. i ' -ì: .M
G - .;-..
• 1
J _ ._.. ^_'y."
t_,1 _ . ..... . . .. l .I
.. , .. ^
.::::
.......
1--+ m _^ ..---'--.
U^
u, o ,_
-`--- ^ ._ 1
^^ t r t ii:'.t: }.,... a..r .:,-: .. _
0
+ 41,
^
r77
=•. - I ^
{ I r 1 ; t t
-1-'-{-^ -^
.
,
f rt+^ j {
^ ^ f. J 1 !
,
... . ! .? 44
r 11 1
_ rj. ^ i^^ ^ ^ i}
! +r; t Y ^^^ ^ ^• I ^ Prt It}
..
_l a ._....-.-.-- i ^1-^ ^ a r J- ^ - 1 {T_'
11• ^ i l .1.
•1-... ^' 1 '
_ J.^ - r
W . . . ..:
1^
f..l, } ^ + t t a ^ 1 ,I1• ^
^. .I
r ( i.
y.- ^ -..- { Y 1 t I f .I^ + .il
'-C^'1--.'
'$t 117 ^L' 1^ ^^ T ^ 1 A • -t t Y t.
Zj ~-4 ^ I ]-': -i 1 .I -.3 Y. } ItS '1 - ^ ]' 1 '7 tI
-^. . ..._.^. '

+4-r
A
: { t 1 1 { a^
'
;+H 7 T_
7 -
a 0
... . _ :
-,+_.-._' - ._., ^...
_T ^7-. ] - -:_ 1«.
.. .
-''_.:..^...::.. .-37• ..ti. _ . t «.I.: _.. ._ -^.t. . -1 . 1 ^..!__ t-^xZ$ _ ' T1.
-}
m
_. '. .. 1 . . . . '
. ^
^::rz =-_ r ._.....:; , . ..., T_"., . __:+ ..,. ^. ..._ _ r._ ++^^,---' . ..: .. ........_.....{ ..,. ... ...... i :^ +-
J
,
1
-- Pit I l .-{
H
O^ t
t^ . ,
Lt____._:41^_ ^_.4 ____^____ ^ i ^^^.^._ _..,_..._._^. _ ^-- ^^-'1 ^• ^ . c _._ . i ... __J
^
0 0 •
• 8. Use the Ramey type curve to calculate permeability
and total skin factor for the data in Prob. 7.
Compare your answers to those obtained in Prob. 7.
9. Use the McKinley type curve to calculate permeability

for the data presented in Prob. 7. Compare your
answer to the values of permeability determined in
Probs. 7 and 8.
10. Use the Gringarten type curve to analyze the data

in Prob. 7 for permeability, wellbore storage factor
and total skin factor. Compare your results to those
obtained in Probs. 7, 8 and 9.
11. You are given simulated pressure drawdown data for a
well which, at the time it was placed on production,
had a stabilized shut-in pressure of 6017.3 psia. You
are requested to analyze this test for permeability,
• wellbore storage factor and total skin factor by

simultaneously matching the data on both the derivative
and Gringarten type curves. Compare your values of
permeability and skin factor with those computed using
the conventional semilog analysis.
Pi = 6017.3 Bo = 1.17 RB/STB

h = 80 ft uo = 1.3 cp
cp = 0.08 T = 178°I'
S o = 0. 74 rw = 0.290 ft
S « = 0.26 q 0 = 100 STB/D
c o = 12 x 10-6 psi 0
x 10 - 6 qw
cW = 3 psi- i cf = 5 x 106 psi 1
•
3-156
•
t, hrs pwf, psia Ap, psi
apln
1.00 x 10 -4 6014.5 2.8 3.65

_4
2.60 x 10 6010.0 7.3 7.83
_ 4
5.00 x 10 6003.6 13.7 10.80
1.00 x 10 -3 5990.6 26.7 23.12
-3
3.50 x 10 5933.5 83.8 66.81
-3
7.00 x 10 5869.9 147.4 113.40
-z
1.04 x 10 5821.1 196.2 139.35
2.10 x 10 -2 5719.7 297.6 140.90
4.20 x 10 -2 5626.4 390.9 124.17
_2
8.00 x 10 5564.8 452.5 73.93
0.12 5536.8 480.5 62.32
0.17 5515.8 501.5 55.45
0.30 5486.1 531.2 47.08
0.40 5472.7 544.6 47.3S
0.60 5453. 5 563. 8 47. _-) () 0
1.00 5429.3 588.0 47.34
1.50 5411.1 606.2 43.96
2.00 5398.6 618.7 44.36
2.50 5388.6 628.7 45.52
3.00 5380.3 637.0 45.54
4.00 5367.2 650.1 45.55
•
3-157
• • •
6000 -rt
PROBLEM 11
5900
T-+r
--r.
5800
^
.r.,
570 0
^
5600
v, +t
co
5500
5400
5300
10 10-2 10-1 10° 101 102
t, hrs
SOLUTIONS
•
1. Semilog and cartesian plots of the drawdown

data are presented in Figs. 3P.1 and 3P.2. From Fig.
3P. 1,
m = - 56.3 psi/cycle.
The permeability is computed using Eq. 3.6:
k = - 162.6 mh
qB"
k = (162.6) (95) 1.2) (0.65)

-56. 3 16
k = 13. 4 md.
•
From the cartesian plot, Fig. 3P.2, m' = -0.37
psi/hr. Also, from Fig. A.4, cf = 4 x 10-6 psi-1.
The total compressibility is computed using Eq. 2.27,
i.e.,
ct = coSo + c w S w + cf
ct = (14xl0-6) (0.78) + (3x10 6) (0.22) + 4x10-6
-6
ct = 15.6 x 10 psi
The drainage area is computed using Eq. 3.12:
A = - 0.2339aB
m' cth^
• A (0.2339) (95) (1.2)

(-0.37) (15.6x10 6) (16) (0.16)
3-159
A = 1.8 x 106 ft2
•
A = 41 acres.
Since the test well is the only well in the

reservoir, the calculated drainage area is the total
reservoir area.
•
3-160
E 3700
• •
3650
SLOPE = m = -56.3 PSI/CYCLE
3600
cd
.r{
V)
3550
w
APPROXIMATE END OF
TRANSIENT FLOW
3500
3450
3400 1--
0.1 1 10 102 103
t, hrs
Fig. 3P.1: Semilog plot, Problem 1.

+{.
l ^
t
,j ..^^ , r I•' ^„ tf f .fi t-1-1 ^ l r
l--• -rT'^}'^"^
`^
w+
r^
:;
f
• I ' , ,' tt^ j t t! t .•« 1}
,a
YT "
^ l
` 1^.^
Îi^ ^{{I
^ tr ' t ;^ ++rt r .'r i

r
^t
•
0
^$ , { „ fî -f I +^^ ^
^ ff 1 } ^ M it • ? r, '` r C' 1} ; ^ r^: r^-{
1, h1I .^ I, -, 1 ^, , ,t , . , I ' ,;i• _, ,, t

t ,,t r I r ^ ^
. ^^ .,. ., ,^ r.,
1
^ V .r .i•, 1 V
,F. C)
iiij ,, (( ^
rr
1
-' r ` ^^f ' •« ^ r^, ^^

O
, ^) O
i_. ^ ' ri
^1' 1ir1 _ f '' }} r^ i, ) •'^ ,«';^ ^(^ zi^; ; ^r „ 1 t i.;
t^ , .• , , h ' }.^$..,, . . . . , ^ ., , ^ '. , ,^ .^1 ^,
Jî -
-T- ^
I 1 o
Q H ^ 00
+i . r +r t r
+ I)i ;t + t r- -^T r • r :
I f^ ^ ! ^(^ îj' ff? I l fi
'
ra
t1 }t ^ r i /^
1^
_^ l l 1' I )• 1 ^{t
^
I 1
^
i); ,;r- Y 1 Ilj f ,
:+
^ •
'r„ ^
^, 1 ^f r„ ^'i. ! ;•; i ^ r I•_
{
l'
,r
tf!
i r^^ ^
/^ I i ^ f i •
r : :^ ,
o C:^ 0
, 1, ) r 1
' 1 ` ^,^ #^ ^ t $ l i!f i
^ :- ^ I•' _; r^+f ¢af I~ : ^ ^ :_ «f ^ ^ {ir j ' ;; .^, ') r ilj^ ;^:; 4J

'+ rt` ;,
âO
tl IT 0
V)
.^,^ _• _, , :-^ ; t r ';f r^ ; ;^ ,,,,' } i ? 1I Ô^ ^ ^, 1 4J
^^ t a^
f-•r
r t i ^f#; it{; t U
1^ ^j tI I 1' ^ l I rl f ^1 11G
i ttrf rrl 4 )+ h ^
!1 . •, ^ ^lr^ "
$ ^'- r ^' ; c. ^ - 1-1' ^ • - ^-{ { r F ^'1` ,j: i ^[ ) {) {.) . :^ 1- O
1 - t - 1L +; ^ 'f Ji { , ' J 'f 'Î : h ' t1 I I ^ +^ ^ y ! . }`rY.
+
14 ; Y^ i ; . ) ) , i- , r/t' 1 'r I I •. ^ Ir^ ; IF +.).}
00
•1 ^
, h ' i^ i ' . .. . .
0
C) Ln o C) Ln 0
u7 N O Uy
Ln
O
Ln l!)
tt
M M M M M M M
Etsd 1^ Md
Is
3-162
APC 206 2/01
#0*nadarlV2 PROJECT COMPUTATIONS

NO. OF SHEETS
•
DATE _"-SLI.PROJECT P"3- 1 JOB NO.
BY Ald-1 CHK.: SUBJECT P A.CS S urc.t.. ^=.rk, _^t/3^
0
2. Instead of forming one semilog straight line, as
• predicted by the ideal theoretical model, it appears
from the semilog plot that the drawdown data form four
possible straight lines. The obvious question is:
Which line represents transient flow and can be used to
compute permeability?
Additional insight into this problem is gained
from the cartesian plot. Inspection of Fig. 3P.3
suggests that pseudosteady state probably begins some-
where between 4 and 8 hours. If this is true, it is
evident thtLt lines 3 and 4 on the semilog plot are
apparent straight lines which do not represent transient
flow; accordingly, we must now decide between lines 1
and 2. Notice that line 1 ends very quickly (t = 0.3 hrs) ;
we will show in a subsequent section that this is also
an apparent straight line which is being caused by an
early-time non-ideality such as wellbore storage or
formation damage. Thus, it is concluded that line 2
• represents transient flow, and can be used to compute
0
permeability.
The slope of line 2 is m = -8.3 psi/cycle.
From Eq. 3.6,
k = - 162.6 mBu
k (162.6) (303)(l.116) (0.88)

(-8.3) (8)
k = 729 md.
Since this rock is consolidated, the formation

compressibility can be approximated from Hall's
Correlation, Fig. A.4, to be 4.4 x 10-6 psi-1. Thus,
from Eq. 2.27,
• •
3-163
V 0L L-I
4^:b A 1^ 4-
C^ CN O V V J
tli O^ V 00 CD O F-+ N
O O O O O O
O
f
^1
1. 11
^wf Psia Aii

• , {, {, ' w
1 ^l-^t t I• } ,t }1 „ }'
^1 F^ t`(1 t* 11r ,
,r
, + t . {:I 1 l'^' `ti ^_^;, 1. , ,1 ,. I +1i1 , •, : $ ; ^ ^ _ 1 '+ `^ ' ^ ^ 1^
! r^ ^ . ^
f . ! n 1,
^ t il: r ^ !t !^ t t '' It , .,• I
,I Ir ', I t, I! i r. , { I- 1. . • I Il
i ^^ 1^ , 4 ( t"I
1^ 1
1^- Fly-i.
1 1 t ; n Ti
00
.
1 t 1.;^
, , I,
i.t^ . 1 1'j Jt 11 t . ^
^,
+ ^ , '7 '
; ,:', !1: 1 , {` t1 1 - i •^ ; 1 f# - 1}l+
1
11 +1' . ^}1
^
TI t
N
I. ,i{
tl1 ^` I
; +r { l i l t . ' +_ ^ . ;; •^ j
.i :
,,
$
1 ;t#
+-
O^
t. ,+,
$+l ,.^ ' i 14i i {
•N
0
,I
t+
.
1
, :-.t
.
.
I T
-a - _ a
,
J
t
' ^
T lk. 1 ^
a
i
1
♦-
•
I $ # ^ 1p;
;
1 ,k `^ 1 ;#t ; u;; +,i fi1; #; nil
1 iv-
1 ► ;; ' 1 1; ^ ; 1; ^*' ±{1; ! ^ ,{ ^ 44;
,
11 f i ll
1. ,, +
N
Itt $t { ;+
t ,, t
N ; :
i -!^ •-i-^{ ., 11±'{ { - I l^:i a.j.

" :}-! t ^^ ^
t ♦ ^-- 1 • ^~{ j± 1 :' ^ ` ^^ ^. .1 l..l. ':1 ^ -,+, ^11

W
N
-11 { ^ , . ^1 , ^+lf 1{ 1 1 ^^ • ^ 11 1 ^ i,. ^ ,;^1 $ ^f-^.:

,1 ^ !t I ' 1 ^- ,
7 - 77
^J
1;
•W
^
{^ ,
~
!
^
{
f
I
1 ' 1 ' ^ 1
t
t+
,
(
-1
^11
1
^
.
^,
.
^
^
t
I
11:.
1
,1
,
'i1
+11t
111
' ,
-
{It
:t
+a
I T
I
11.
1
-r,,t
t+,
,-
{ t.
l
t 31
tî
1
^
L
•i^'
i ♦ -
r+`
l 1 ,, rf+
It
+
I
I ^,^1
i
!
,
r
?'
{,i
1
•
1 t I t !} t + i{ 1* ^L,. ,
i .
r A l } ild: ,
rti t:. '

1
. , { :
^ '{
!
l t; ?1; ^1 t '
1.
T H lot ,+ ^j i ; {i ! jt ' , x+!' ^# till I rll }' 1;
ct = coSo + cwSw + cf
•
ct = (6.9x10 6) (0. 75) + (3.2x10-6) (0. 2S) + 4.4x10-6
-6
ct = 10 . 4 x 10 psi
The drainage area is computed using Eq. 3.12:
A = - 0 m?
The slope, m', of the cartesian plot is -1.10 psi/hr.

Therefore,
(0.2339) (303) (1.116)

A
(-1.10) (10.4x10-6) (8) (0 .129)
• A = 6.7 x 106 ft2
A = 154 acres.
This problem illustrates two very important points:

First, reservoirs are complex and seldom behave in the
manner predicted by ideal mathematical models. Second,
it is possible on most graphical presentations of pressure
data to identify multiple straight line segments; unless
the cause of each line is defined, it is easy to select
the wrong data for an analysis. Determining which part
of the data to analyze is the most difficult problem
encountered in pressure transient analysis.
•
3-165
• 3. In the absence of any knowledge about the drainage
geometry, or the well location within the drainage area,
it will be assumed that the well is centered in a square
drainage area. From Table 2.2 we find that this system
should behave like an infinite system for dimensionless
times less than 0.09, i.e., (tDA)etf ^ 0.09. Therefore,
from Eq. 3.7, transient flow should last until
ûc ta
tetf - 0.00 0 26-)7k (tDA)etf
_s
__ (0. 126) (2. 3) (1.86x10 40) (43560
(0.09)
t etf 0.0002637) 22)
tetf - 145.7 hrs.
This is significantly greater than the test time and,

consequently, there should be no boundary effects in the
• data.
E
3-166
4. The real problem in a tight reservoir such as this
relates to the length of time it takes to attain pseudo-
steady state flow. C Dnsider, for example, the time
required for the well to stabilize if the drainage area
were 320 acres. From Eq. 3.16,
380ûctA
tpss k
-4
t _ (380)(0.082)(0.0156)(2. SX10 ) (320) (43560)
pss 2.2
tpss = 770 hrs
tpss = 32.1 days.
Since this calculation is based upon the behavior of a

well centered in a circular drainage pattern, 32.1 days
represents the minimum time required to achieve pseudo-
steady state. If the drainage area is larger than 320
acres, the required test time will increase proportion-
ally; further, if the drainage geometry is not circular,
or if the well is not centered, the time will be larger
than indicated.
It will be expensive to run a test this long;
^ further, it could be difficult to maintain a constant
rate for the required period of time.
•
3-167
V
5. From the semilog plot of the test data, Fig. 3P.4,
I*
m = -94 psi/cycle.
The permeability can be calculted using Eq. 3.6:
k 162.6quB = (162.6) (200) (3.5) (1.20)

mh (-94) (20)
k = 72.7 md.
s= 1.151 plhm pi - log k Z+ 3.23 .

ûctrw
From the semilog straight line, Plhr = 4713 psia. Thus,
4713-5500
• s = 1.151
[ -94
72.7 6 • + 3.23
- log
(.15) (3. 5) (33.94xl0 )(0.333) 2
s = 1.151(8.37 - 7.57 + 3.23)
s = 4.6.
Neglecting any effects due to turbulence, fracturing,

or perforations, the skin factor due to permeability
alteration can be written, based upon Eq. 3.32, as
sd = s - sr
The pseudo-skin due to restricted entry can be evaluated

using Eq. 3.33:
3-168
5100
5000
4900
4800
^
w
cD m = -94 PSI/CYCLE
4700
p1HR = 4713 PSIA

4600
i--^
a--Y
4500
Fig. 3P.4: Semilog plot, Prob. S.
4400
0.1 1.0 10.0 100.0
• t, hrs
• 0 i.
4w-
0.825
825
^ sr = 1.35 (h kH/kv + 7)
[_.- 1 1
- [0.49 + 0.1 In (h kH/kv)] In rwc - 1.95}.
Based upon the perforation configuration,
h = 0.
0
Also,
zm = ho + hp/2
zm = 0 + 12/2
zm = 6 ft.
• The corrected wellbore radius, rwc, is equal to rw

since ho = 0. Therefore, the skin due to restricted
entry is
0.825
sr = 1.35 [i^ - 11 {in (20 3/2 + 7)
- [0.49 + 0.1 In (20 V3/2)] In (0.333) - 1.95}
Sr = 2.3.
Subtracting Sr from the total skin factor gives the skin

due to damage:
sd = s - sr = 4.6 - 2.3
• sd = 2.3.
3-170
-4
The pressure loss due to damage is computed using

Eq. 3.30: is
Aps = - 0.87 msd
Aps = -0.87(-94) (2.3)
Aps = 188 psi.
The flow efficiency can be computed using Eq. 3.38:
E = PR pwf AP s
PR pwf
Since the last measured pressure is on the semilog

straight line, it appears that this well did not reach
pseudosteady state; consequently, we cannot determine
the value of pwf theoretically required by Eq. 3.38.
The best value of pwf to use in this situation is the
last measured pressure, i.e. , pWf = 4530 psia; thus
E = 5500 - 4530 - 188

5500 - 4530
E = 0.81.
If the damage indicated by this flow efficiency can be

removed by stimulation, the expected flow rate after
stimulation can be predicted using Eq. 3.41:
_ 1
aafter E "before
= 0
1 .81 (200 STB/D)
aafter
= 247 STB/D.
•
stimulation
3-171
^ Thus, the potential for a 47 STB/D increase in
production is indicated if the damage in this well
is removed.
is
3-172
7. Log-log and semilog plots of the pressure data
• are presented in Figs. 3P.6 and 3P.7, respectively.
From the log-log plot it can be seen that the first
few data points lie on a straight line of unit slope.
These points are completely controlled by wellbore
storage. From the straight line we can pick any
arbitrary point and read the corresponding values of
Lp and t. For example
Ap = 300 psi
t = 0.51 hrs.
The value of C can be computed using Eq. 3.66
C = Bt
24 pi - pwf)
C = (115) (1 . 73) (0 . S1)

(24) ( 300 )
.
C = 0.0141 bbl/psi.
The approximate time at which the unit slope line

ends on the log-log plot is t* = 0.20 hrs. Therefore,
using the rule of thumb that wellbore effects will have
dissipated by t = SO t'k, the true straight line should
begin at approximately 10 hrs. This line, shown on
Fig. 3P.7, has a slope
m = -235 psi/cycle.
Therefore,
kh = - 162.6quB
m
kh = - (162.6) (115) (l. 3) (1. 73)

-235
1*
3-176
-Y
1000 . 1
77 7
800 A f #^-
r +I - r _ ^*
77
rTT t Yiî--
600
17
UNIT SLOPE LINE
Fi T ,
400
t4
.^ + 1 ; , r , +.-` r{ , r 1 r{
} r
f F ^ 1 f t Tr
1#1 ti
-47 -# :7 ^ j * t i t, -t
_
; ,
. ;! ^r !t ,1
+Y ,
200 i^ ^ ,r ;r ^ •r . ,_ ,.^ „ ^ •
^# !
.^ fi^r
_ . T ^1 .i,' ?. .;. ^ ::...:. ti.^
.:: .. :.. } : :;i# f : .^r' ^ _ I' . ` ^ } t
•r
tt _ Y-{ r 1^_ i } i L 1
^ ft t ^
1+»- .-li-,•
^y^ r^ } ^ -
- TI t r
♦ E 1 ?^
YTI G-}r L [
l j 4 rr!
^ t , i ,' ± t r{ +^ `^ r ;+ rr }}
,, ,• } f
. .^ L- 1:. j f .t r, lt If ^ Y - f i1 1 .f
-t1^^ }^ • .. . ^
r ^t ^^^ } -1 • ^♦ ^ T } ^, ^ 1 I f+
'
1 f
,
100 1 7: ir t
i 7
80 I^tr rr! # i
1 ^ I
^ r,
1
t , ^ r- 1 rI
' } . ^ -. .♦ , + 1 ^ ^.... . ^ tk^ ^.^ +++'^{5 'Yr r f ,r .i{ `
1:.. ^'r'
60
^ ^r^
F
r LI
'
.}^. -
71 1
♦f
r
f .I. t ^1^ -
r
l^^YL.,
_
•
` ^! l 1
''H rr + t
^
'+
.
r
^ -}
r ^TTT Î^^ I
• j_ 1 -}"^1 { r 1 S_t j^ t•r ` ' S 1!
. . I..I^
r
^ t x t i, - rt i.
^ } ^
40
^ ^.^^ ♦•' - . • ^' ^ ^ ' .^ ^ ^ .._^ F
{ t . 1
^1 # - t
♦ - -r + +1 ^ j.
^.. • r ! • -} rY r ^'7-1 -f'^
r I ry
4 ^•
.,.
^} !1 1 f^1 t,
1 1: -r-^ y ^- i ^} t ^ } # ^
}- + Y f r l.
} t i
z i^ ^: j
L( , {I t }rl .i r $ t r ^f t t i
T
1 ^ ,
^t-^T{ --` LLt 4-1 I. rI t ^_ I ^ Iryl I i ^-'^
3-^ j
t
^ I i 't # r r ' I , r r l { {, ti l Z _ t r ` ^ #i i i{I l i i I ^ I + 1 f ^ ; ^ {
1 ._rt^ ; ^ I 1i 1 ^
: I ^ ^ t t Ir r^ 1 I,
x
20 tjt7+
r
[
4. '
..
+`^ CI i h,ft 11. ^ f^ ! ., #11 1

^^ } ^ r+++ ^}a, ^Y^^ ! • 1 '1 ,
^^^1 }/ r ^ '^$ 1 ^ r- r }1 ^ -;
r
- r
}
{,}.
t-^ ^ +## I{
} }^
.,.
t,
r„
n
.
if
*1^# I I l î t* t , n, ^ ,r t 1' r !. `+r ^^ + 'L' F1 i ^^ l,r
-}• t,ir,^} Lf '$tÎttîrl^ 1^t1•^+ r
- # tC} Ir+1 ,t jl^ r t i ^ h +f} } ^ } ^ t r 'I •^. , , ^ 1.
- - = 1 ' } ^r l ^^
• rrr+++jj il',Ir f t#. I(t} i'i }^ .h ^+ ^
r Tr 1 ilt^J} { •.
10 ^ •rr++ i
0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0
t, hrs
Fig. 3P.6: Log-log plot, Problem 7.

.
3-177
i 3700
• • -
IF I
, .i .. a ^ . •^
'_,..t { l:«F Y • 1I ' + 1 .
1 'rt T^T
3600 I I
41
3500 ,.^
- •T +. - .N .
^
^ -^ ' •a •^ :>.. , 111 d^ . ,I - ,----^---^--- ,
r .^:
3400 I• ^ I I '^ __1..^_ ^: ^ ^ ^1; .
ti_1I î,
^, I ^
' •• ^ '^ ` - ^ ' I.,. ^ t I , l i , -
_^. .- ^ ' i.:.
," I •
^ i^ ^ '_" ' - . ^^ ^^ •
^
.
.1 . •
^. . . . ^
- ^ ^ _ i- 1 ^ ^^ ., . _
3300
_ .. -^- . ^^ ^_ --L--^'-. ^--^^^^ - __+...

- ' Y '- ` -. I . ' ^•-i^ +-
-T I . "1 ^ "_1^ . ., •
3200 _..-^ J
^ , I . .y .^. . i . . .
_y .,- _
.r., ' T r. 't' '.'_-
.tI ^'
-^--1--F^ . ^j.+. ° ..+-: .{ t _ ^ .. 1 ^ ^ -^ ' .^--i ,--r---:-^•--^. -____
T^t'-'-'-' i .
-- j--- • : -^-t._.: _ . ,- ^L. -,-^ ± a i;•T ^- - ,-'
-T - 7
3100 -T f
4-i .
- - . } - -1 - i ^ - --^ - - t --} t - -- -_ _ _ .
W ¢3
_.
---
3000
cc _ _^ __ ^ ^ . .. _, . _.. . - -- - ' - 1 -'-• r- ^_ . _-__ L1___^ T_ .
t-
-__^_r_-^
--+
:
^.
2900
i 1T_^ __c.. ^ II r :.^_^ ^^ t + L i_ -^ ._ ^-- , T^ I , ^r'^ t- ' -, ,- ^ --r---t- + ^ ^-•+
' -;-. . ' ' : . r^^ :
-._ ... .... . . .
---- ^ . _._, • • -- -•- T-- - - ! • - -- -- -+---^-+ - - - -
2800 y- ' I -. ^^ I • Il l I- Î I
^^.^ 1 . .-
' -
^
.î
-
Y
^':._^_ _ _ •-t_r W^ • • ^ -t ^.-l._-l.i^l
---t-1-•-•-- ♦. { '^
l y
!I _
--•- J _^ --•-.-,+ ^ _, ._r-^_ , TT.^ .,^-^_t_._ • ,+,.._.. -. _^ --, --- -1 --f^ -- i ' _' ., * ._
T
2700 m = 235 PSI/CYCLE
. _ . T^
- ^+. 1 -
^' 1 '1. ^ ^ ^ .i - a-+.i^--' ' .. ' i 111 ^t _ Y. ^ ^ ^^ _' •.1 , -•.^
- -•-
_- ^
i -^-- ^ --
.. - . . ^ ..
T +--::I-._. ^ ' _ ^ -+ ^-_-_ _ ^T • a-4- t ., â
.
---i--. ^ .J-; - ^ - . -_ ^.^
; ^. ; t • , ----^ -^-^ -^ =- } ---^-- + •- i --^ - ^--^---s ^ ;-{ -
.-.
-j
. i ^., . 1.;. ,-.. a f ^. ^ .^ , ._
:• ,` _^.. _ .; ^ Tr ^^
2600
1,__.;f , ; r _ ^. ':, - - ^-^^_ {^ ^ ^ ` ._ . , _ :^ , + ; • :^
._ ^ . . ' ... .. .-. ♦ ' . t _. .t., .^ L ^-^ ^ ♦ _.. __, , _. ^- .-.`..•...
---y--'__Y ^ . _ . _ __._ - - ' - -, _ • t-_ "Y_ --.-
2500
0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0 8.0 10 20 40
t, hrs
Fig. 3P.7: Semilog plot, Problem 7.
kh = 179 md-ft
k = 189 = 22.4 md.

•
The skin factor is computed using Eq. 3.29:
s= 1.151 plhm pi - log k 2+ 3.231

ûctrw I
s = 1.151 2968-3700
-23S
- log 22.4 -6 + 3.23

(0.186) (1.3) (14x10 ) (.25) Z
S = -1.9.
From the total skin factor, it appears that this well •

is slightly stimulated.
•
3-179
8. A data curve is presented in Fig. 3P.8. It is
observed that the first four points on the curve
•
define a unit slope straight line. This line was used
in Problem 7 to compute the wellbore storage factor C
as
C = 0.0141 bbl/psi.
This factor can be expressed in dimensionless form

using Eq. 3.59:
C _ 0.894C
D ^cthrw
_ (0.894)(0.014l)
CD (0.186) (14x10 6) (8) (.25)2
CD = 9,682 = 104
• A match of the data curve with the Ramey type

curve is presented in Fig. 3P.9. From the match
point on this plot,
(PD) hI = 1.0
(Ap)N1 = 143 psi.
Therefore, from Eq. 3.81
k_ 141 2 Bu (pD) h9
( 11 p ^1
k=(141.2)(115)(1.73)(1.3) X 1.0
8 143
k = 31.9 md.
to
3-180
•
104
CD = 10"
103
• • •
.H •
•
C].
•
4-^
•
•
°
1.02
•
10
0.1 1 10 102
t, hrs
Fig. 3P.8: Data curve, Problem 8.
•
3-181
• ^ •
10= 103 104 105 106 10T 106
tp
Fig. 3P.9: Matched curves, Problem 8.

According to the matched position of the data,
S = 0.
LI
These answers compare to k = 22.4 md and s = -1.9
which were computed in Problem 7 using the conventional
method. Because of the difficulty in obtaining a unique
match with type curves, the conventional analysis is
regarded as the most accurate.
•
3-183
I
9. The data curve for the McKinley plot is presented
• in Fig. 3P.10. This curve was matched with the
McKinley graph to obtain Fig. 3P.11.
10 4
103
Ln
a)
102
+-+
•
• 10
•
1
10 102 103 10" IO5
Ap, ps i
Fig. 3P.10: Data curve, Prob. 9.
It is noted on Fig. 3P.11 that the field data

match a type curve having a value
kh
• = 2500
5.15pC
3-184
^
^
00
Ln
5.61SOpC ft3-day
qB ' RB
• Fig. 3P. 11: McKinley Curve match, w blem 9.

• ^
Further, from the match point,
E ^5.615ApC1
qB = 3.0
JM
(pp)M = 1 x 104 psi.
Using these values in Eq. 3.85,
C = 5.615LpC B
qB
IM 5615M
(115) (1. 73)

C = (3.0)
(5.615)(1x10")
C = 0.0106 bbl/psi.
Then, according to Eq. 3.86,
•
k = [5.615cj1L5.6h5uC1
k = (2500) (5.615) (1.3) (0 .0106) •

8
k = 24.2 md.
This compares to k = 22.4 md and s = -1.9 calculated in

Problem 7 using the conventional method. The Ramey
type curve match gave k = 31.9 md and s = 0.
•
3-186
• • •
PTA^:7: Chapter 3, Problem 7 - Oil Well Drawdown
Storage: unit slope; linear low: half slope; type curve matching
2*1OJ
Le gend
o TEST DATA
z
10v 0 d(P)/d(In(T)) DER5G,-
-- -
i o
^
a
a - -^-- - E - - I CID
1 - ^ - ®
---^ -
E*10' - ^ ---- -' - ^ -- ^ - , ► r ^^ - - - ; ----; - ^

10 2*101
10- 10'
At (houl-S)
4,
10. The data curve for the Gringarten type curve match
• is presented in Fig. 3P.12. This curve was overlayed
onto the Gringarten type curve to obtain the match
presented by Fig. 3P.13.
10''
UNIT SLOPE LINE
103 • • •
^ •
t^. •
•
•
•
•
•
102
10
0.1 1 10 102
t, hrs
Fig. 3P.12: Data Curve, Problem 10.
•
3-187
4r•M ► O WO RorEtwoX. MOM, MANN.rpnr.
Al ^w...a .
s
s
cl^^
F-+
00
^ x
a
I
2 3 4 ^ b / tl v
. fp 5 , r-. kh At
DIMENSIONLESS TIME 0.0004Y5 - ^ 1 3 • s • , e 7 i 4 567a9
o N
Fig. 3P.13: Gringarten type curve match, Problem 10.
• • ^ .i
The value of CDeZs determined from the match
• depicted by Fig. 3P.13 is
(CDe2$)M = 104
Also, match values of pressure and time corresponding

to the arbitrarily selected match point shown on Fig.
3P.13 are
(Ap) M = 1000
(PD) M = 7.0
(t) M = 1.0
(tD/CD)M = 4.5.
Permeability can be computed by substituting

is
the match values of pressure into Eq. 3.87:
k = 141.2qBp (PD)NI
h (A P) NI
k = (141.2) (115) (1.73) (1.3) 7.0

8 1000
k = 32 md.
The match values of time can be substituted into

Eq. 3.88 to calculate the wellbore storage factor:
(t) ^1
C = 0.000295 kh
u tD CD M
• C = (0.000295) ( 32) (8) 1.0

1.3 T-.5
3-189
C = 0.0129 bbls/psi. .
The storage factor can be expressed in dimensionless

form, according to Eq. 3.59, as
C = 0.894C
D ^cthrW
C _ (0.894)(0.0129)
D (0.186) (14x10 6) (8)(0.25 )z
CD = 8,858.
The total skin factor can be determined, according

to Eq. 3.89, as
(C e2s), i
s = 0.5 in
D CD 1 •
s= 0.5 in [
88 1
s = 0.06.
These values compare to the following results

which were obtained in Probs. 7, 8 and 9:
Conventional semilog plot:
k = 22.4 md
s = -1.9
•
3-190
Ramey type curve:
0
k = 31. 9 md
C = 0.0141 bbls/psi
s = 0
McKinley type curve:
k = 24.2 md
3-191
APC 206 2/01
tiiadarIîg PROJECT COMPUTATIONS

NO. OF SHEETS
• DATE PROJECT %-c,o 6 3j JOB NO.
BY CHK.: SUBJECT j]jjo & CCNLVQ.-,,
- - - -
_ 2 L
C D ZS ^. = D
or?^ --
00 o 9a 8 .l
-, ^
-- ^^ ^^h . , 6 ► _ ^'^ .^Y 2s.

•
C' _ q^ ^ ^
2 ^
_. ^
-7111
-0
11. A log-log plot of both Ap and opin versus time

is presented in Fig. 3P.14. These data were matched
simultaneously with both the derivative and Gringarten
type curves; this match is presented in Fig. 3P.15.
From the match point on Fig. 3P.15,
(Apln)M = 1000
(pDln)M = 11.5
(t)M = 0.01-
(tD/CD)M = 3.4.
Also from the match,
CDe2$ = 103.
Permeability is computed by substituting the match

values of derivative pressure into Eq. 3.101:
141.2qBu ( pDln)M
k = ,
h
"pn) NI
k = (141.2) (100) (1.17) (1.3) 11. 5

80 1000
k = 3.1 md.
Using the match values from the time scales of

the data curve and type curve, the wellbore storage
factor is determined from Eq. 3.88 as
C = 0.000295 kh (t) M
tD CD N1
3-192
10
• •^• •
• • • • •
• •
•
• n
102 n n
• •
n ^
n n
a n n n n nnn n
u.7
^.
•
cp
101
^
Op
APT ln
n
10°
10-4 10-a 10-2 10-1 10° 101
t , hrs
Fig. 3P.14: Data Curve, Prob. 11.
• • r
.
• ^ •
copyAqM © 1Be3 FIOPFInoL,MMw,,FRANCE.AII rwM.ls.rr.a.
The use of this type-curve I. M^cAOW In WorW 041 - Wy 1M:
I~
S^.
P1
^
.^ •
r-i
-Q
à a
aa
Cr
.y. •
'--1
r-1
W 10' 1U - ^p^
'w lo-, ' 10 10 10 10 U
t
^L = 0.000295 kh t
D C
Fig. 3P.15: Derivative type curve match, Prob. 11.

C = 0.000295 (3.1)(80) 0.01
1.3 3.4
C = 1.65 x 10-4 bbls/psi.
This value can be expressed in dimensionless form by

Eq. 3.59:
C = 0.894C
D ^cthrw
The total compressibility is:
ct = coSo + cwSW + cf
ct = (12 x 10-6)(0.74) + (3 x 10-6)(0.26) + 5 x 10-
ct = 14.7 x 10-6 psi-1
^
Therefore,
-4
C _ (0.893)(1.65 x 10 )
D (0.08)(14.7 x 10- 6) (80)(0.2 9) Z
CD = 18.6.
1
S = 2 (CDe2s)M
1n
CD
= 1 ln r103
s Z
L18.6J
s = 2.0.
0
3-195
The type curve match depicted by Fig. 3P.15
indicates that the last few points should be in
transient radial flow and, accordingly, should form
a semilog straight line. A semilog plot of the
pressure drawdown data is presented in Fig. 3P.16.
From this plot,
m = -106 psi/cycle.
This slope can be substituted into Eq. 3.6 to compute

permeability:
k = - 162.6 m
k = -162.6 (100) (1.17) (1. 3)

(- 10 6) ( 80 )
k = 2.9 md.
0
s= 1.151 plhm pi - log k 2+ 3.23 .

ûctrw
From the semilog straight line on Fig. 3P.16,
5,430 psia.
plhr =
Therefore,
= 1.151 5430-6017
s
-106
- log - 2.9 -6 + 3.23

(0.08) (1.3)(14.7x10 )( 0.29) 2 J
3-196
6000
5J00
S800
5700
5600
w
P1HR - 5,430 PSIA

SLOPE = m = -106 PSI/CYCLE
5500
5400
5300 "llL
10-3 10-2 10-1 10° 10, 102
t, hrs
Fig. 3P.16: Semilog p1ot, Prob. 11.
0 9 0
s = 1.6
0
Comparing these values of k and s to those determined
by type curve analysis, it is found that they are in
excellent agreement.
•
3-198
r
C7 • •
Prob. 3-11: Typ e Curve Match

Match Results
Radial homogeneous
Infinitely acting
K = 3.0994 md
Cs = 1.701 e-004 bbUpsi
d = 19.2075
S = 1.9762
1000
^
n
...
n.
CO
aD
o .
100
•^ Pressure
Radial Deriv
10
0.001 0.01 0.1 1 10
Elapsed Time hours
This is a match of the pressure drawdown data from Problem 3.11, Smith, using the Bourdet, et al., derivative type curve for homogeneous
reservoirs. It should be noted that s is the total skin factor, total skin can include the effects of many non-idealities in addition to formation
damage.
1000 Prob. 3-11: Log-Log Quick Match Results
Quick Match Results
Radial homogeneous
Infinitely acting
Cs = 1.701 e-004 bbUpsi IF-
K = 3.0994 md
S = 1.9762
o.
..
CL 100
Ca
...
(D
0
a TP1 Pressure
M Radial Deny
- Genera ted
- Gen dedYatIME
10
0.001 0.01 0.1
Elapsed Time (hours)
This plot shows a comparison of the acutal test data (points) to the theoretical behavior (lines) that should occur in an infinite-acting
homogeneous reservoir with the parameters shown in the results box. The theoretical results were generated using Quick Match. If the
reservoir model, welibore storage model, and boundary model are correct, and if the parameters determined from the test analysis are correct,
the theoretical test behavior generated using these models and parameters should match the test data. In this case, the theoretical model is in
excellent agreement with the measured data, thereby confirming that this is a reasonable solution.
11 • •
• • •
6100 Prob. 3-11: Semilo Analysis
6000
5900
5800
.m
^
n
"
5700 M Results
:3
tal homogeneous
InfinRety acting
K = 2.9674 md
= 237.3905 md.R
5600
Rim = 80.8052 ft
FE -0.759
S = 156.6804 psi
S = 1.7319
5500
5400
5?^
0.0001 0. 001 0 . 01 0. 1 1
Elapsed Time (hou rs)
This is the conventional semilog analysis of the test data. The straight line used in this analysis was fit to those data that exhibit a horizontal
derivative on the log-log plot. It should be noted that S represents the total skin factor, not the damage skin factor, and that the flow efficiency,
FE, and the pressure loss caused by skin, dpS, are computed using the total skin factor. It should also be noted that the radius of investigation,
Rinv, is calculated using the total length of the test; if boundaries are encountered during the test, this number can be misleading.
PROBLEM 3-11B : SIMULATED BOUNDARY EFFE CTS,
• PRESSURE DRAWDOWN TEST
Pressure drawdown data for a well in a homogeneous reservoir are

presented in Prob. 3-11, Smith. The analysis of this test is presented in
PanSystem file PTA3-11B. Use the data in this file to:
(a) Simulate the behavior this well would have if it were located in
the center of a square drainage area with no-flow boundaries
located 50 feet away. Present your results on both semilog and
log-log plots and summarize any diagnostic characteristics that
you observe for this type of boundary.
(b) Repeat part (a) for a well in the center of a square drainage area
with constant pressure boundaries located 50 feet away.
0
t
• • •
PTA3-11 B: Square - No Flow Boundaries

1000
Quick Match Results
Radial homogeneous
C4osed system -(L:L:L:L]
Cs = 1.701 e-004 bbUpsl
K = 3.0994 md
= 1.9762
L =50 It
N
n
a 100
a^
0
. . ..
•. TP1 Pressure
H Radial Deriv
- Generated
- Gen derivative
10
0.001 0.01 0.1 1 10
The solid line on this plot illustrates the pressure behavior that would have occurred if the well had been located in the center of a square
drainage area with no-flow boundaries located 50 feet away. After the onset of boundary effects, the well approaches pseudosteady state flow; it
is observed that the derivative forms a unit-slope straight line at late time after pseudosteady state is reached.
6100 PTA3-11 B: Sq uare - No Flow Boundaries
6000
5900
5800
5700
^
N
02 5600
a
Quick Match Results
R a di a l h omogeneous
5500 losed system - IL: L: L:LI
= 1.701 o-004 bbUpsl
K = 3.0994 md
-1.9762
L -50 ft
5400
n
n
5300
S2nn
0.0001 0.001 0101 0.1 1 10
draingage area with no-flow bounaries located 50 feet away. After the onset of boundary effects, the reservoir approaches pseudosteady state
flow. Pseudosteady state flow does not result in any diagnostic characteristics that are recognizable on the semilog plot.
• • •
3
,^ v_.
• • •
1000 PTA3-11 B: Square - Constant Pressure
Quick Match Results
Radial homogeneous
Closed system - IL:L:L:L)
Cs = 1.701 e-004 bbUpsl ^
K = 3.0994 md
S = 1.9762
L =50 ft
=
n.
a 100
^
0
. , , ..
nn TP1 Pressure
H Radial Deriv
- Generated
- Gen derivative
10
0.001 0.01 . 0.1 1 10
Elapsed Time hours
drainage area with constant pressure boundaries located 50 feet away. After the onset of boundary effects, the well approaches steady state
flow; while not shown by this plot, the derivative becomes zero upon reaching steady state flow.
6100 PTA3-11 B: Sq uare - Constant Pressure
6000
5900
5800
N
SL'
..
5700
^
(U
a
Quick Match Resufts
5600 Radial homogeneous
Closed system - (L:L:L:Lj
s t 1.701 e-004 bbVpsl
K = 3.0994 md
S = 1.9762
5500
L =50 ft
5400
n
5-Ann
0.000i 0.001 1101 0.1 1 l
draingage area with constant pressure baunaries located 50 feet away. After the onset of boundary effects, the formation face pressure
approaches a constant value as the reservoir approaches steady state.
0 0 0
• PROBLEM DD NDAHl : R ESE RVOIR LIMITS TEST
This is a drawdown test from the discovery well in a sandstone reservoir

which contains a volatile oil. Reservoir pressure was above the bubble point
at the time of the test. The primary purpose of the test is to determine
reservoir limits.
Amerada mechanical gauges were used to conduct the test. Because of
the length of the test, it was necessary to change gauges approximately once
per week; some of the scatter in the data was caused by this process. Flow
rate was maintained constant during the test within a tolerance of ±5%.
You are requested to estimate the reservoir area.
PanSystem File: DDNDAHl.PAN Y° = 53.1 °API
Test Length (t) = 1122.5 hrs Yg = 0.739
• =
q° 595.4 STB/D B° = 2.55 RB/STB
r,„ = 0.3646 ft µ°=0.11cp
h=18ft
^=0.152
T=270°F
p; = 5,115 psia
SW 0.36
ct = 28.6 x 10' psi-'
^
q
0
Ar
Solution: DDNDAH1
Attached are the log-log and cartesian plots of the test data. Early data on the log-log plot indicate the
presence of wellbore storage effects. It appears that wellbore storage effects may have masked the transient
flow period and that boundary effects were already present when storage ended. Since no information is
available regarding the position of the well relative to reservoir boundaries, no attempt was made to model the
possible geometries which could have caused the behavior depicted by this test.
The objective of this test was to estimate the size of the reservoir. The log-log plot of the derivative shows
that the late date form a unit-slope line; this is characteristic of pseudosteady state flow. A straight line through
• the pseudosteady state data on the cartesian plot was used to estimate the drainage area to be:
A = 293 acres
NOTE: The shape factor, Ca, reported on the cartesian plot has no meaning II Computation of the shape
factor requires knowledge of the skin factor and the slope of the semilog straight line; the porrect values of
these parameters are not available for this test since no semilog analysis was made. Also, the value of R
reported on the cartesian plot has no physical meaning; it represents the radius a circular reservoir would have
if it contained the reservoir area, A, computed from this test.
0
DDNDAHI: Lo -Lo Plot
.•
+..
•• .
. n
,-. ••
. .
a100
a n ^ •
ô n
•
a^
0 • n •
• ;
• •
%
n
n • n •
• n ' n
. n
10 0.1 1 10 100 1000

The derivative data which plot as a unit-slope line on this plot are believed to represent pseudosteady state flow. These data also plot as a
straight line on cartesian paper, the slope of the cartesian straight line can be used to estimate the area of the reservoir.
0 0 • _ *I
^ r
• •
5200 DDNDAH1: Cartesian Plot
Model Resutts
Radial homogeneous
baed system
5100 = 292695 acres
R = 2014.5433 ft
Ca = 9.`'ii
5000
4900
,-.
co
V
CL
ED 4800
^
^
Q.
4700
' ••••
••^
•
•
4600
•••
••
4500
44000
200 400 600 800 1000 1200
The data used to construct the pseudosteady state straight line on this graph correspond to those derivative data which plot as a unit-slope
line on the log-log graph.
NOTE: The value of the shape factor. Ca, reported on this graph has no meaning. Calculation of the shape factor requires knowledge of the
skin factor and the slope of the semilog straight line; these values are not known for this test since no semilog analysis was made. Also, the
value of R reported on this graph has no physical meaning; this is the radius that a circular reservoir would have if it contained the area, A,
computed from this test.
-40
0
• PROBLEM DDNDAH2: RESERVOIR LIMITS TEST
This is a drawdown test from the discovery well in an oil reservoir.

Reservoir pressure was above the bubble point at the time of the test. The
primary purpose of the test is to determine the size of the reservoir.
The test was conducted by running Amerada mechanical gauges in and
out of the hole on an approximate one-week schedule. Although the test is
unusually long, the production rate was reported to be constant within a
tolerance of ±5% of the initial rate. Scatter in the data is caused by changing
the gauges periodically, and by paraffin problems during the test. Problems
were also encountered near the end of the test by the apparent drop-out of salt
near the wellbore.
You are requested to determine the reservoir area.
PanSystem File: DDNDAH2.PAN
• Length of Test (t) = 1169.3 hrs
q° = 23 8.1 STB/D
rW= 0.3646ft
h= 12ft
0. 169
p; = 5,458 psia
T = 262 OF
ct = 18.32 x 10' psi-'
lp B° = 1.73 RB/STB
•
0
DDNDAH2
The purpose of this test Is to determine the size of the reservoir. This is the discovery well, and is the only
well in the reservoir. This test is considered to be of high quality; although the test was almost 50 days long, a
major effort was successfully made to keep the rate constant during this entire period. Scatter in the data are
due primarily to running gauges In and out of the well every week, and to treatment for paraffin problems.
Problems were encountered toward the end of the test due to the apparent drop-out of salt near the wellbore.
Fluid properties are from a laboratory PVT study. It Is believed that this is a closed reservoir with a
solution-gas-drive recovery mechanism.
The early data are complex due to wellbore storage effects and what appear to be multiple boundary effects.
No geological information is available regarding the location of boundaries; accordingly, no attempt was made
• to model boundary effects using the well test data.
The derivative data form a unit-slope line on the log-log plot after approximately 200 hours. These data,
which are believed to represent pseudosteady state flow, also form a straight line on the cartesian plot. From
the slope of the cartesian straight line, the reservoir area was determined to be
A = 51 acres.
NOTE: The shape factor, Ca, reported on the cartesian plot has no meaning!! Computation of the shape
factor requires knowledge of the skin factor and the slope of the semilog straight line; the correct values of
these parameters are not available for this test since no semilog analysis was made. Also, the value of R
reported on the cartesian plot has no physical meaning; It represents the radius a circular reservoir would have
if it contained the reservoir area, A, computed from this test.
0
1000
100
10
1000
The unit-slope line on this graph is believed to represent pseudosteady state flow. The data which form the unit-slope line are analyzed on the
attached cartesian plot to determine reservoir area. Scatter in the data are caused primarily by running gauges in and out of the well on
approximate one-week intervals, and by paraffin problems encountered during the test. Problems were also encountered toward the end of the
test due to the apparent drop-out of salt near the weilbore. J
0 0 0
• • •
5600 DDNDAH2: Cartesian Plot
Model Results
5400 Radial homogeneous
Closed system
5200 = 50 .8587 acres
= 839.7533 ft
500 0 Ica = 10.241
4800
4600
4400
-420
y
0
^4000
a^
3800
•
a 3600 a A.
^.^.«
3400
3200 ^ ..
^ .► •
3000
•.^• •
2800
4^ ^* , •
2600 •..
2400
22000
200 400 600 800 1000 120
The straight line on this plot is believed to represent piseudosteady state flow. The line was selected to correspond to those derivative data
which form a unit-slope iine on the attached log-log plot.
NOTE: The shape factor, Ca, reported on this graph has no meaning!! Calculatlon of the shape factor requires knowledge of the skin factor
and the slope of the semilog straight line; the correct values of these parameters are not available since no semilog analysis was made. It
should also be noted that the value of R reported on this graph represents the radius a circular reservoir would have if it contained the area, A,
determined from this test; this does not imply that this reservoir is radial.
•
0
^ PROBLEM DESIGN-DD: Design of Drawdown Test, Closed System, with
Wellbore Storage and Skin
The objective of this problem is to introduce the test design capabilities of

PanSystem. You are requested to predict the pressure behavior of a well located in
the center of a closed square with no-flow boundaries a distance of 400 feet from
the well. The rate of the well is to be 250 STB/D, and the prediction is to be made
for a flow time of 50 days.
Additional Data:
Initial Reservoir Pressure = 4400 psia
Reservoir Temperature = 212 OF
h = 40 feet
(^=22%
SW=27%
Sg = 0
cf=12x10' psi`
y°=34°API
yg=0.72
is
Producing GOR = 675 SCF/STB
k=52md
s=5
C = 0.02 bbl/psi
rW = 0.33 feet
After predicting the behavior of this test, analyze the test to determine k, s, C, and
drainage pore volume. Explain any potential problems that might be encountered in
analyzing a test conducted in this manner. How could you change the design of the
test to eliminate these problems?
0
•
0
Computation of the Pseudoskin Factor
Caused by a Restricted-Entry Well
Completed in a Multilayer Reservoir
N. Yeh,' SPE, and A.C. Reynolds, SPE, U. of Tulsa
Summary. This work considers flow to a restricted-entry well in a single- or multilayer reservoir where crossflow occurs between
reservoir layers. A simple algebraic expression for approximating the pseudoskin factor caused by restricted entry is presented. It is
shown that, for practical purposes, this equation yields sufficiently accurate estimates of the pseudoskin factor regardless of the location
of the open interval. This work also extends single-layer results of interest in well testing to the multilayer case and shows that i f sufficient
data are available, the results of this work can be used to prepare a complete analysis of well testing pressure data obtained at a restricted-
entry well.
Introduction
Wells are frequently completed over only a fraction of the productive ,0t =0, and c,t =cf. The dimensionless wellbore pressure and
zone to delay water and/or gas coning. A well completed in this dimensionless time are defined by.
way is referred to as a restricted-entry or partially penetrating well. Pwli-(kh(P!-Pwj))1(141.2qµ) ......... .. ............ (3)
Two questions of practical importance are (1) what is the produc-
tivity loss caused by restricted entry and (2) should the well be stimu- and 1D-(2.637x10-4 kt)l(¢crprw) . ......: .. ..........(4)
lated. Answering these questions requires estimates of the pseudo- where q-total sandface production rate in RB/D. All definitions
skin factor resulting from restricted entry and the real skin (me-
are given in terms of oilfield units. For flow to a restricted-entry
chanical skin) caused by damage, or stimulation, as well as an es- well in a layered reservoir, pseudorad'1al flow will eventually occur,
timate of reservoir permeability. Although well testing has been provided that the well's drainage area is sufficiently large. 1.2.8. 10
advocated for estimating these parameters, 1.2 pressure data ob- During pseudoradial flow, the dimensionless wellbore pressure drop
tained from restricted-entry wells will usually exhibit only one
is given by
semilog straight line, the one corresponding to pseudoradial flow.
Semilog analysis of such pressure data yields estimates of the total PwD ° 1.1511og(4rD/e7)+st . . . . . . . . . . . . . . . . . . . . . . . . . . (5)
flow capacity and the total skin factor. Because the total skin factor
where st =total skin factor.
is a linear combination of the real skin factor and the pseudoskin
We let kw denote the'thickness-averaged horizontal permeability
factor, additional information is required to obtain estimates of the
(based on hw) adjacent to the open interval. In general, kw is given
individual skin factors. The standard procedure is to use correlations
by
to estimate the pseudoskin factor (see Refs. I through 6). Once the u'o
• total and pseudoskin factors have been estimated, the real skin factor
can be calculated directly.
Refs. 1 through 7 discuss procedures or solutions that can be used
to estimate the pseudoskin factor caused by restricted entry. With
kw-(l/h.) D k1hw/ . ............................ (6)
1^ô
where the sum is over all layers that are at least partly opt:n to flow.
the exception of Ref. 2, all results assume a single-layer reservoir. For example, for both cases shown in Fig. 1, no - 1, mo =2, and
The results of Ref. 2 are restricted to a two-layer reservoir where kwa(kthwt+k2h„r1)/hw, where hw,t and h,,,z are as shown in Fig.
the open (perforated) Interval is adjacent to one of the layers and 1. Note that for Case A, h„2-hz, whereas for Case B, h,,,t -h t.
the length of the open interval is exactly equal to the thickness of In general, h1 is the height (length) of the part of the open'interv al
the reservoir layer open to flow. The results of this work remove that is adjacent to Layer J. For cases where the height (length) of
these restrictions and apply specifically to a restricted-entry well the open interval, hw„ is equal to the thickness of one individual
in a multilayer reservoir with an arbitrary location of the open in- layer with thetop of the open interval adjacent to the top of that
terval. Moreover, we present a simple equation that can be used -layer, kw equals the permeability'of the open layer. Figs. 2, and
to compute the pseudoskin factor and show that this equation yields 3 depict such situations. For the geometries of Figs. 2 and 3,
results that are sufficiently accurate for practical purposes. In this -k 1. Note that in Fig. 3, we have designated the layer open to
work, we also extend the well test analysis procedures of Bilhartz flow (the middle layer) as Layer 1.
and Ramey' to the multilayer case. On the basis of the results dis- The dimensionless flow capacity of the open interval is denoted
cussed in this work, Refs. 8 and 9 present type.curves for analyzing by ft and is defined by
well test pressure data obtained at a restricted-entry well.
fl =(khw)I ( kh) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (7)
Deflnltions and Background
where h=total reservoir thickness. When only one layer is open
For the layered-reservoir problem considered here, it is appropriate to flow, the open la yer is always designated as Layer 1. In such
to define the thickness-averaged horizontal permeability and the cases, ft m(kth,,,t)/(lrJr). For cases where the height (length) of the
thickness-averaged porosity-compressibility product by open interval equals the thickness of the open layer, hwl aht and
m ft-(ktht)/(kh) (see Figs. 2 and 3). For single-layer cases, Eqs.
k=(1/h) E kjhl ..................................(1) 6 and 7, respectively, reduce to kwsk and it ab, where b is the
J-1 penetration ratio and is defined by
M
b =hw/h.• .............. . .......................... (8)
and cr =(1/h) E Ojc,/h/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . (2) With this notation, the results of Refs. 1, 2, 8, and 10 indicate that
1-1 the total skin factor that appears in Eq. 5 may be approximated by
The subscript j refers to the layer index; e.g., kj =horizontal per-
• meability of Layer j. For the single-layer case, m-l, ktak,

'Now at Amoeo Production Co.
Copyright 7989 Society of Parolaum EnQln••n
sr=sb+(s^ft) . . . . . .. . . . ... . .... . . . . . . . .. . ... . . .. . . (9)
where sb- pseudoskin factor caused by restricted entry and s=real
skin factor resulting from damage or stimulation. The validity and
interpretation of Eq. 9 are discussed in detail later.
SPE Formation Evaluation, June 1989
253
•
0
e PARTIAL PENETRATION - Spherical & Hemispherical Flow
Reference: Stanislav, J.F. and Kabir, C. S.: Pressure Transient Analysis, Prentice
Hall, Englewood Cliffs, New Jersey ( 1990), 109-121.
Spherical Flow - Perforations in Middle of Layer

If perforations are located near the middle of the layer, i.e., away from the top
and bottom boundaries, spherical flow will occur following a short radial flow
period. A drawdown test during spherical flow will define a straight line on a plot
of p,,, versus l/tln with slope m, defined as
(^µct)li2
M = 2453 Bµ
e q k 3n ^l)
S
where
kQ = k 2' kZ ^ (2)
• The vertical permeability can be computed from the slope of the spherical flow
straight line, ms, as
kZ =^ µ ct 2453 qBµ (3)

mBk
The horizontal radial permeability, k, can be determined from early radial flow data
before spherical flow begins, or from late radial flow data after spherical flow ends.
2. Hemispherical Flow - Perforations Near Top or Bottom of Layer

When perforations extend to the top of the layer, as is common when the
producing zone is underlain by an aquifer, of if they extend to the bottom of the
layer, which is common if the layer is overlain by a gas cap, hemispherical flow will
occur. Hemispherical flow will cause drawdown data to define a straight line on a
plot of p,,,f versus 1/0. In this case, the vertical permeability can be computed using
• Eq. 1 if the constant is changed to 4906.
•
0
REFERENCES: Multiphase Well Test Analysis
•
Perrine, R. L.: "Analysis of Pressure Buildup Curves," Drilling and
Production Practices, API (1956) 482.
2. Martin, J. C.: "Simplified Equations of Flow in Gas Drive Reservoirs and

the Theoretical Foundation of Multiphase Pressure Buildup Analyses,"
Trans. AIME (1959) 216, 309.
3. Weller, W. T.: "Reservoir Performance During Two-Phase Flow,

Journal of Petroleum Technology (Feb. 1966) 240.
4. Raghavan, R.: "Well Test Analysis: Wells Producing by Solution Gas

Drive," Society of Petroleum Engineers Journal (Aug. 1976) 196.
5. Ayan, C. and Lee, W. J.: "The Effects of Multiphase Flow on the

Interpretation of Pressure Buildup Tests," SPE Formation Evaluation (June
1988) 459.
6. Chu, W-C., Reynolds, A. C. and Raghavan, R.: "Pressure Transient

Analysis of Two-Phase Flow Problems," SPE Formation Evaluation (April
^ 1986) 151.
7. Al-Khalifah, A-J. A., Aziz, K. and Horne, R. N.: "A New Approach to
Multiphase Well Test Analysis," paper SPE 16743 presented at the 62nd
Annual Technical Conference and Exhibition of the Society of Petroleum
Engineers, Dallas, Texas, Sept. 27-30, 1987.
Jones, J. R. and Raghavan, R.: "Interpretation of Flowing Well Response

in Gas Condensate Wells," SPE Formation Evaluation (Sept. 1988) 578.
Vo, D. T., Jones, J. R. and Raghavan R.: "Performance Predictions of

Gas Condensate Reservoirs," SPE Formation Evaluation (Dec. 1989) 576.
10. Boe, A., Skjaeveland, S. M. and Whitson, C. H.: "Two-Phase Pressure

Test Analysis," SPE Formation Evaluation (Dec. 1989) 604.
11. Raghavan, R.: "Well-Test Analysis for Multiphase Flow," SPE Formation
Evaluation (Dec. 1989) 585.
12 Camacho-V., R. G. and Raghavan, R.: "Performance of Wells in

Solution-Gas-Drive Reservoirs," SPE Formation Evaluation (Dec. 1989)
611.
0
13. Serra, K. V., Peres, A. M. M. and Reynolds, A. C.: "Well-Test Analysis
for Solution-Gas-Drive Reservoirs: Part 1 - Determination of Relative and
Absolute Permeabilities, " SPE Formation Evaluation (June 1990) 124.

for Solution-Gas-Drive Reservoirs: Part 2 - Buildup Analysis," SPE
Formation Evaluation (June 1990) 133.

for Solution-Gas-Drive Reservoirs: Part 3 - A Unified Treatment of the
Pressure-Squared Method," SPE Formation Evaluation (June 1990) 141.
16. Raghavan, R., Chu, W.C. and Jones, J.R.: "Practical Considerations in
the Analysis of Gas-Condensate Well Tests," SPE Reservoir Evaluation &
Engineering (June 1999) 288.
L^
0
I
• DERIVATIVE ALGOR ITH MS
Many algorithms exist which can be used to compute the derivative of

pressure transient data. Since all field data are influenced by noise, a
compromise must be reached between achieving a derivative which is
sufficiently smooth to be usable in well test interpretation and distorting the
character of the derivative response. Three algorithms which are commonly
used in the petroleum industry are presented.
Lagranian (three-point) Method

This is the method which provides the least amout of smoothing. When
taking the derivative, the three points which are used are the point of interest
and the points immediately before and after this point. This is illustrated by
Fig. 1. The value of the derivative at the point of interest is computed as
A1RmL + AtLmR
M=
OtL + AtR (1)
AP
Fig. 1: Lagranian (three-point) algorithm.
0 -1-
1
where AtL is the spacing between the point of interest and the point to the left, •
and mL is the slope of the line connecting the point of interest and the point to
the left; AtR and mR are the corresponding spacing and slope, respectively, to
the right.
Akima (five-point) Method'

The Akima method is a five-point slope averaging method which
computes the derivative using two points immediately before and two points
immediately after the point of interest; this is illustrated by Fig. 2. The value
^
^ 1 0
i m4 i
M, 0
mi
1+Z
MI
l+l
AP
01
o
0 ^
i-2
0 Data points
• Points used to
compute derivative
•
t
Fig. 2: Akima (five-point) algorithm.
of the derivative at the point of interest, i.e., at point i, is computed as
m=
m2 - ml I ms + 1 r"a - m3 1 m2
(2)
Ima - mJ + Im2 - mil
where m,, m2, m3, and m4 represent the slope of the line segments connecting
-2-
•
• the points. For example, the slope of the line connecting points i-2 and i-1
is computed as
m1 = Ap;-i - AA-2 (3)

ti-1 - ti-2
where t is the time function on the x-axis.
Noise Reduction Method2

Rather than using a fixed number of points before and after the point of
interest to compute the derivative, this method uses a fixed distance before
and after the point of interest. This distance, S, is called a smoothing
parameter. Figure 3 shows several data points on a plot of Ap versus t (time
function) and illustrates the parameters needed to compute the derivative at
the point of interest. Depending upon the value chosen for S, the number of
data points that exist within the interval before and after the point of interest
can vary from zero to a large number. If there are no points within the
•
0
0 0
ML
Op 0
oil
^
0 1 I
0 o Data points
etL AtR
• Points used to
compute derivative
Fig. 3: Noise reduction algorithm.
9 ,.
interval, the time function interval, At, is measured from the point of interest
to the first point outside the interval. If only one point is within the interval,
the time function interval will be measured from the point of interest to this
point. If more than one data point exists within the interval, the time function
interval will be measured from the point of interest to the farthest point in the
interval. Using the three points defined in this manner, the derivative is
computed using Eq. 1.
The advantage of this method is that the smoothing parameter, 8, can be
increased as the amout of noise in test data increases. Care must be taken,
however, that S is not made so large that the character of the data is changed.
As a general rule, you should use the smallest smoothing parameter that will
result in a derivative curve that can be interpreted.
References
1. Akima, H.: "A New Method of Interpolation and Smooth Curve Fitting
Based on Local Procedures," Journal Association of Computing
Machinery (Oct. 1970) 17, 589-602.
2. Bourdet, D., Ayoub, J.A. and Pirard, Y.M.: "Use of Pressure Derivative
in Well Test Interpretation," SPE Formation Evaluation (June 1989)
293-302.
-4-
•
• PROBLEM PTAX3-13: BUILDING A PANSYSTEM DATA FILE
FOR A DRAWDOWN TEST; ANALYSIS
OF DRAWDOWN DATA FROM A
HOMOGENEOUS RESERVOIR WITH
WELLBORE STORAGE AND SKIN
This problem has two objectives: First, to show how to build a

PanSystem data file for an oil well pressure drawdown test when pressure
data must be imported from a file and, second, to show how to perform
conventional and type curve analysis of a drawdown test using PanSystem.
You are requested to build a Pansystem data file using the data given
below. Reservoir rock and fluid properties, and flow rate, must be entered
manually. Pressure data should be imported from the file PTAX3-13.TPR.
After completing the data file, you are requested to analyze the test
using both conventional and type curve methods. Use QUICK MATCH to
confirm that your analysis is correct,
0 qo = 270 STB/D co = 11 x 10' psi-'
µo1.7cp cW=3x 10' psi''
Bo = 1.15 RB/STB ce=4x 10' psi`
rW = 0.4 ft ct = 11.8 x 10-6 psi-'
0. 193 So = 0.6
h=24.2 ft SW = 0.4
p; = 3100 psia
^
•
0
0 • •
0 0 •
1000
C14
100
10
0.01 0.1 1 10
0 0 •
PTAX3-13: Quick Match

n • • • Pressure #1
• ^ ^ • Pressure *1 Derivative
Match Results
Radial hwr4geneotm
1000 Infinitely acting
K =77 md
Cs = 0.0039 bbUpol
_ 395.2147
S = 1.6
c3.
'-' 100
6?
10
0.01 0.1 1 10
0 • 0
310(
PTAX3-13: Quick Match
305(
300(
295(
2901
285( ) •
2801
Quick Match Results
R ad i a l homogeneous
Infinitely acting
Constant compressibility
2751
Cs = 0.0039 bbUpsi
K =77 md
S = 1.6
Pi = 3100 psia
270^ P
.01 0.1 1 10 1
Elansed Time (hours)
• • •
1000 DESIGN-DD: Log-Log Plot
.
. •
. ^
. •
. ' .
.
. •
.
. •
n • • • . . . . • • • • ' .
. •
100 . .
.
• ' .
.
. • ^ • . ^
rA
.
^
A . ^ • •
a • . . ^
ed.+
.
^ • • .
A •
: • •
10 + • •
s • .
• ^
s • • • .
•
1 •
0.001 0.01 0.1 1 10 100 1000

Elapsed Time hours
4500
DESIGN-DD: Semi-lo Plot
4400 n n • . . . . . . . •
^ • . .
n .
4300 ^ • . ^
. a
n n n • • n ' • • • .
4200
. .
4100 •
4000
vî •
3900
^
3800 ^
3700
3600
3500 .
3400
0.001 0.01 0.1 1 10 100 1000
is 0 is
RJeYevl ce \_ee1 3 . w,,,,^ 1n)4^ênbarj er t 'R A . . (; aS
Re.cevvoir Er, t,n^^Yir^ I S PE , ^t^.llas ^
^ Tex^-^DOI^ C.r es Vol
Appendix F
Determining Pressure Derivatives
The derivative at a point is determined by finding a weighted mean TABLE F-1-DATA FOR DERIVATIVE CALCULATION
of the slopes to a preceding point and a following point (Fig. F-1).
The parameter L defines the minimum abscissa distance to these at GP Derivative
points, smoothing out "noise" in the neighborhood of the central (hours) (psi) In(At) Window
point. Thus, L can be defined as G(ln t) for a flow test or as 11.99 830 2.4841
&(]n at,) for a buildup test. Experience indicates that 0.1.-!-.L!.- 15.98 839 2.7713
0.3 often is a satisfactory compromise between being too far from 2.8764
the central point so that detail is lost and being too near the central 19.97 845 2.9942
. point so that a great amount of noise is introduced. However, trial 23.96 850 3.1764 3.1764
and error may indicate that other values of L are more appropriate 29.93 859 3.3989
in a given situation. 3.4764
The pressure derivative calculation procedure is best illUNIrated 39.88 864 3.6859
with an example taken from field data. Suplx).w that we want in 49.32 869 3.8983
determine the derivative -dpw,fld Int=d(Ap)ld In(W) at at=
23.96 hours from the drawdown data in Table F-1 and that we
choose L=&(ln t)=0.3. We first calculate In(At) for all test times, 0
obtaining ln(23.96)=3.1764 at Ata23.96 hours. Then 3.1764-L=
3.1764-0.3=2.8764, and 3.1764+L=3.1764+0.3=3.4764, as
0
shown in the last column of Table F-1, creating a "window" around
the central point. We use the data points just beyond this window
r
o^ ^ --°-------
dPR
to calculate the pressure derivative, mp: 0
AIL = 3.1764 -2.7713 - 0.4051.
II Ii LIIII1I
APL (850-839) 0 ^1 L ^ L
a 27.1538 s mL. 0
AtL 0.4051
atR =3.6859-3.1764=0.5095.
0
i dtL ^ dtR ^
A PR (864-850)
27.4779 smg.
AtR 0.5095
mLAtR +mRGtL
mp=
AtL +AtR
0 - DATA POINTS
(27.1538)(0.5095) +(27.4779)(0.4051)
_ =27.30_ )( - POINT TO BE DERIVED
^ (0.4051 +0.5095)
q - POINTS USED FOR THE CALCULATION
S1 Metric Conversion Factor
psi x 6.894 767 E+00 = kPa M P . VALUE OF DERIVATIVE
Fig. F-1-Calculation of pressure derivative.

•
0
REFERENCES: Application of Pressure Derivatives in Well Test Analysis
1. Tiab, D. and Crichlow, H. B.: "Pressure Analysis of Multiple Sealing

Fault System and Bounded Reservoirs by Type Curve Matching,"
Society of Petroleum Engineers Journal (Dec. 1979) 378.
2. Tiab, D. and Kumar, A.: "Application of the pD Function to Interference

Analysis," Journal of Petroleum Technology (Aug. 1980) 1465.
3. Puthigai, S. K.: "Application of pD Function to Vertically Fractured

Wells - Field Cases," paper SPE 11028 presented at the 57th Annual fall
Technical Conference and Exhibition of the Society of Petroleum
Engineers, New Orleans, Louisiana, Sept.26-29, 1982.
4. Bourdet, D., Whittle, T. M., Douglas, A.A. and Pirard, Y. M.: "A
New Set of Type Curves Simplifies Well Test Analysis," World Oil (May
1983) 95.
• 5. Bourdet, D., Ayoub, J. A., Whittle, T. M., Pirard, Y. M. and Kniazeff,

V.: "Interpreting Well Tests in Fractured Reservoirs," World Oil (Oct.
1983) 77.
6. Bourdet, D., Alagoa, A., Ayoub, J. A. and Pirard, Y. M.: "New Type
Curves Aid Analysis of Fissured Zone Well Tests," World Oil (April
1984) 111.
7. Clark, D. G. and Van Golf-Racht, T. D.: "Pressure-Derivative

Approach to Transient Test Analysis: A High-Permeability North Sea
Reservoir Example," Journal of Petroleum Technology (Nov. 1985) 2023.
8. Wong, D. W., Harrington, A. G. and Cinco-Ley, H.: "Application of

the Pressure Derivative Function in the Pressure Transient Testing of
Fractured Wells," SPE Formation Evaluation (Oct. 1986) 470.
9. Wong, D. W., Mothersele, C. D. and Harrington, A. G.: "Pressure

Transient Analysis in Finite Linear Reservoirs Using Derivative and
Conventional Techniques: Field Examples," paper SPE 15421 presented
at the 61st annual Technical Conference and Exhibition of the Society of
Petroleum Engineers, New Orleans, Louisiana, Oct. 5-8, 1986.
a
10. Ozkan, E., Vo, D. T. and Raghavan, R.: "Some Applications of

Pressure Derivative Analysis Procedure," paper SPE 16811 presented at
the 62nd Annual Technical Conference and Exhibition of the Society of
Petroleum Engineers, Dallas, Texas, Sept. 27-30, 1987.
11. Onur, M. and Reynolds, A. C.: "A New Approach for Constructing
Derivative Type Curves for Well Test Analysis," SPE Formation
Evaluation (March 1988) 197.
12. Tiab, D. and Puthigai, S. K.: "Pressure-Derivative Type Curves for

Vertically Fractured Wells," SPE Formation Evaluation (March 1988)
156.
13. Yeh, N-S. and Agarwal, R. G.: "Development and Application of New
Type Curves for Pressure Transient Analysis," paper SPE 17567
presented at the SPE International Meeting on Petroleum Engineering,
Tianjin, China, Nov. 1-4, 1988.
^
14. Bourdet, D., Ayoub, J. A. and Pirard, Y. M.: "Use of Pressure
Derivative in Well Test Interpretation," SPE Formation Evaluation (June
1989) 293.
15. Duong, A. N.: "A New Set of Type Curves for Well Test Interpretation
With the Pressure/Pressure-Derivative Ratio," SPE Formation Evaluation
(June 1989) 264.
16. Onur, M., Yeh, N. and Reynolds, A. C.: "New Applications of the
Pressure Derivative in Well-Test Analysis," SPE Formation Evaluation
(Sept. 1989) 429.
17. Mattar, L. and Zaoral, K.: "The Primary Pressure Derivative (PPD) - A
New Diagnostic Tool In Well Test Interpretation," Journal of Canadian
Petroleum Technology (April 1992) 63.
0
•
Chapter 4
MULTIPLE-RATE TESTING
It was shown in Chapter 3 that the pressure drawdown

test requires a constant rate; this requirement is also
imposed on other commonly used pressure transient tests.
It is impossible, or at least impractical, to maintain a
constant rate throughout the duration of many tests.
Accordingly, it is desirable to develop a method which can
be used to analyze pressure transient data which are
^ measured under variable rate conditions.
I. APPLICATIONS
The primary applications of the multi-rate testing

methods developed in this chapter are:
• Analyze pressure transient data from wells

with a variable rate history;
• Obtain pressure transient data without
discontinuing production;
• Use when operational problems prevent
shutting in a well;
• Use when phase segregation prevents use
of the pressure buildup test;
• Use when changing wellbore storage makes
analysis of other pressure transient tests
difficult or impossible.
1]
U U II. METHOD OF SUPERPOSITION
0
The principle of superposition, introduced in Chapter 2,

statesl'2'3 that adding solutions to a linear differential
equation results in a new solution to that differential
equation, but for different boundary conditions. For example,
Eq. 2.35 is a solution of Eq. 2.26 for a single well producing
at constant rate from an infinite reservoir. This is the only
set of conditions for which this solution is valid. however,
by using the method of superposition, it is possible to apply
this equation to multiple well systems, to wells producing at
multiple rates, and to reservoirs with finite boundaries.
We are concerned in this chapter with applications to wells
with a multiple rate history.
Practically, the principle of superposition means that
every rate change in a well results in a pressure response
which is independent of the pressure responses caused by
rate changes at other times. Accordingly, the total pressure
change which has occurred at any time in the life of a well
is the summation of p ressure chan g es caused se p arately by
each rate change in the life of the well. For example,
consider the rate history, depicted by Fig. 4.1, where a
well flows at rate ql for time tl, and then decreases to q2.
The total pressure change at time t ( t > tl) is
t t
(4.1)
(pi .pwf) total (pi pwf)ql] o+[(PiPwf)q2q1] t'
1
This is illustrated graphically by Fig. 4.2.

The best way to visualize the method of superposition
illustrated by Fig. 4.2 is to imagine that two separate wells
coexist at the same location; one of the wells produces at
rate ql from t = 0 to t, and the other well produces at rate
(q2-ql) starting at tl and extending until t. The net rate
for times greater than ti is ql +(q2-ql), or qZ, and the
net pressure change is the summation of the individual
4-2
•
q1
^
q 2 al
Q2 -------^
0
0 t, t
TIME
Fig. 4.1: Typical two-rate flow history.
pressure changes which have occurred at that time. The net

^ pressure response is depicted by the pressure-time plot at
the bottom right-hand side of Fig. 4.2. This procedure can
be extended for as many rate changes as is necessary.
<* III. TWO-RATE TEST
The simplest multi-rate test is the two-rate test. It

can be-run by either increasing the rate or decreasing the
rate, but experience indic_atP^ rha the rate will restabilize
faster if the test is conducted by decreasing the rate'''S.^
In either case, it is important that the rate change be
sufficiently large to produce a measurable pressure response.
Figure 4.3 illustrates the rate and pressure history of
a well in which the rate is decreased to generate a pressure
response. This well flows at a constant rate ql for time tl,
after which the rate is changed to q2; pressures are
measured as a function of test time, At, following the rate
change. It is noted that the rate does not change instan-
4-3
•
pi
al
0 0
`
t
0
(a2-(1l) ------
pi
tii
al pi
a2
0
tl t tl
E
Fig. 4.2: Graphical illustration of the principle of
superposition applied to a multiple-rate well.
4-4
CA/v- ô^^ . 6k,
,,i,w,rds.#, (e-^ u,n.
•
PRI/SSURE RESPOVSE DUE
Pi TO DI:CRL'ASED RATE
pwf
pwf(At=0) ^ -^ ^^-
0 ^
TIME
r^.<<„ ^-„ Inh r L
Qqru,.S
GL S U
,1^. s4Ow4&v,&c
A.a)b, tt'avq
• 6,,t- wtt,c.b41A
4
0 tl (At=o) tl+1A
TIME
Fig. 4.3: Two-rate history with decrease

in rate.
taneously from ql to q2; instead a finite period of time is

required for the rate to restabilize because of the effects
of wellbore storage. It is commonly believed that wellbore
storage can be eliminated by running a multi-rate test.
This is not true; ^wellbore storage lasts approximately the
same length of time , during a two-rate test as it does in a
• normal drawdown or buildup test.^ The primary advantage of

the multi-rate test is that permeability and skin can be
4-5
evaluated without shutting in the well.
•
In order to analyze data obtained with the two-rate
test, it is necessary to develop an equation that can
predict pressures at any time At after the rate change
occurs, i.e., at -t = tl + At. This is accomplished by
applying the principle of superposition to the two-rate
test depicted by Fig. 4.3, i.e.,
_ _
(pi pw f) t1+At (pi_pwf) q1Jt1+ At
o
+ (Pi Pw f) a2 -all tl+At (4.2)

L Jti
•
The rate ql begins at t=0 and, at the test time tl + At,
has been in effect for a total time t = t1 + At. Accordingly,
the pressure change caused by rate ql can be predicted from
Eq. 3.22: new^ - e-^ ^
t+At _ 162 .6q1BU log k(ti^-At)

^^.^
[(PiPwf)q1]1
kh ^ uc r z
^ 0 t w
Al K) OP
" n^ - 3.23 + 0.87s . (4.3) •
]
Vd_ U ^
The rate changes at t1 by an amount ( q2-ql). This rate
change extends from tl to ( t1+At), for a total time At.
The change in pressure caused by this rate change is also

•
4-6
predicted by Eq. 3.22, i
and is s
equal to
t +At 162.6(q - q )Bu k

2 1
[ (PiPwf) q2q1J 1 =
t trw
1
- 3.23 + 0.87s]. (4.4)
Substitution of Eqs. 4.3 and 4.4 into Eq. 4.2 yields
162.6q Bu k(t +At)

E pi-pwf^ _ kh 1
'Tb7ptL
1 Z
ûctrw
- 3. 23 + 0.87s
[log
c4p)(0 11-% + CaP>2V4'^t +

162.6(q - q) Bu
k2 1
log kAt _ 3.23 + 0.87s (4.5)
ûcr
t 2w
•
Simplifying,
162.6q1Bu
[log (t1+Ot) - log OtI
pi pwf kh
162.6q Bu
+ kh 2 log At + log 2- 3.23
ûcrw
tk
+ 0.87s] (4.6)
mTR X
or,
162.6q 1Bu t l +At
+ a
q 2 log At
pwf = kh At
1
r16:
2.6q,B^;
q2 k
log 2 3. 23 + 0.87s (4. 7) ^
ql ûctrw
4-7
vl
so that, •
^ pwf mTRX + bTR
(4. 8)
where, 162.6q1Bu
(4.9)
m'I'R kh
+ q? lo k - 3.23 + 0.87s (4.10)

bTR pi mTR ql
g0uctrw
X = log
tl + ^t + a2
Q log At. (4.11)
At 1
According to Eq. 4.8, a plot of pwf versus X on

cartesian paper should yield a straight line with slope
mTR and intercept bTR. This plot is illustrated by Fig.
4.4. The permeability, according to Eq. 4.9, is
•
162.6q1Bu
k = - • (4.12)
m TR
Moreover, the skin factor can be estimateds as
s= 1.151 ql (pwf Ot=Oplhr)- log k
ql a2 mTR ûctrw
+ 3.23] . (4.13)
The false pressure, pcan be determineds from the

intercept of the two-rate plot as
q2
P = bTR - ql_2 [pwf(Ot=O) - Plhr1' (4.14)
•
4-8
• INTERCEPT = bTR
q 2 < q 1
•
• SLOPE = mTR
• ^
RETURN TO
pwf PRESSURE DRAWDOWN
RATE RESTABILIZATION--5-6^4
EFFECT •
•
•
A INCREASING At
tl + ^t a2
log + log At
At
1
• Fig. 4.4: Cartesian plot of two-rate pressure

data where q 2 < ql.
It will be shown in Chapter 5 that p* can be used to

estimate the volumetric average reservoir pressure within
the drainage area of the test well.
Example 4.1: Analysis of a two-rate test for an oil well
Problem. An oil well has produced for seven days at an average

rate of 78 STB/day; the rate was then decreased to 64 STB/day
for the purpose of conducting a two-rate test. The flowing
formation face pressure immediately before the rate change
40 was measured to be 1,963 psia. Pressure data recorded follow-
4-9
ing the rate change are given along with other reservoir
•
data. What is the effective permeability to oil and the
total skin factor?
At, hrs pwf, psia At, hrs pwf, psia
0.165 2007 1.50 2182

0.245 2028 1.72 2187
0.325 2047 2.0 2190
0.41 2062 2.5 2194
0.49 2085 3.0 2198
O.S8 2098 4.0 2200
0.67 2126 6.0 2201
0.75 2127 8.0 2204
0.83 2136 10.0 2205
0.90 2148 12.2 2206
0.99 2154 14.0 2206
1.09 2158 21.0 2210
1.15 2166 25.0 2212
1.32 2176 27.0 2213
A = 160 acres
rw = 0.265 ft
= 20%
h = 20 ft
uo = 0.39 cp
Bo = 1.322 RB/STB
ct = 1.38 x 10-4 psi-1.
Solution. To properly analyze the two rate test, it is

t +^t q
necessary to graph pwf versus (log At + a2 log At) on
1
t +4t .
cartesian paper. Let Y=(log l^t + a? log At), and ti =
al
(7) (24) = 168 hr.
4-10
• The calculations necessary to prepare the data plot are
summarized in the following table:
t1+ot
t1+°t a2 X pwf, psia
At, hrs log log At
At At
1
0.165 1019 3.01 -0.64 2.37 2007
0.245 687 2.84 -0.50 2.34 2028
0.325 519 2.71 -0.40 2.31 2047
0.41 411 2.61 -0.32 2.29 2062
0.49 344 2.54 -0.25 2.29 2085
0.58 291 2.46 -0.19 2.27 2098
0.67 252 2.40 -0.14 2.26 2126
0.75 225 2.35 -0.10 2.25 2127
0.83 203 2.31 -0.07 2.24 2136
0.90 188 2.27 -0.04 2.23 2148
^ 0.99 169 2.23 0.00 2.23 2154
1.09 155 2.19 0.03 2.22 2158
1.15 147 2.17 0.05 2.22 2166
1.32 128 2.11 0.10 2.21 2176
1.50 113 2.05 0.14 2.19 2182
1.72 99 1.99 0.19 2.18 2187
2.0 85 1.93 0.25 2.18 2190
2.5 68 1.83 0.33 2.16 2194
3.0 57 1.76 0.39 2.15 2198
4.0 43 1.63 0.49 2.12 2200
6.0 29 1.46 0.64 2.10 2201
8.0 22 1.34 0.74 2.08 2204
10.0 17.8 1.25 0.82 2.07 2205
12.2 14.8 1.17 0.89 2.06 2206
21.0 9.0 0.95 1.08 2.04 2210
25.0 7.7 0.89 1.15 2.03 2212
27.0 7.2 0.86 1.17 2.03 2213
* A cartesian plot of the test data is presented in

Fig. 4.5. From the straight line on this plot, mTR =-124
4-11
2250
bTR = 2,462 PSIA AT X=0
m TR = -124 PSI/CYCLE
2200
2150 P1hr = 2,187 PSI

ro
H
•
2100
2050
2000 LLÛ
2.0 2.1 2.2 2.3 2.4 2.5
x M
Fig. 4.5: Plot of two-rate test data, Ex. 4.1.
•
4-12
^
psi/cycle. Thus, from Eq. 4.12,
162.6q1Bu
k = -
mT Rh
k (162.6) (78) (1. 322) (0.39)

-124 20
k = 2.6 md.
ql (pwf(At-0) plhr)
s = 1.151
[q-q2 mTR
- log k + 3.231
CuctrW J
^
78 1963-2187
s = 1.151
78-64 ( -124 )
2.6
- log
(. 2) (.39) (1. 38x10-4) (.265) 2
+ 3.231
s = 7. S.
IV. GENERAL MULTI-RATE ANALYSIS
The equations and procedures developed for a two-

rate test can be extended to describe any multiple rate
history. Consider, for example, the general multi-rate
4-13
history illustrated in Fig. 4.6. The flowing pressure at
•
time t can be calculated using Eq. 3.22 if the method of
superposition is applied to this rate history, i.e.,
ql
q3
qN
q4
0 q.2 ^ I^'•.• qN 1
I I I I q
0 tl t2 t^ La tN-2 tN-i t
TIME
^ Fig. 4.6: General multiple-rate history.
162.6q1B11 [iou kt - 3.23 + 0.87s

h
pwf(t) pk ûc tr2w
162.6(q2-ql)BU
log k(t-t1) - 3.23 + 0.87s
- kh
ûctrw
162.6(q Nk l-q N-2) Bu log k(t- tN2

h-
t
ûcrw
2)
•
-3.23 + 0.87s]
4-14
• 162.6(q N -q N- 1)Bp k(t-tN-1)
kh log $uct r2w
- 3.23 + 0.87s]. (4.15)
Expanding,
162.6q1BU [iou kt
3
2- 3.23 + 0.87s
pwf(t) kh
ûctrw
162.6q2Bu k(t-t1)
- log - 3.23 + 0.87s
^ Ouctrw
162.6q Bu k(t-t )
+ 1 log t 1- 3.23 + 0.87s
ch
0 162.6qN-1BU
ûcr2w
k(t-tN-2)
-•••- log
kh
Ouc r2
t w
- 3.23 + 0.87sI
162.6qN BU k(t-tN-1)
- 3.23 + 0.87s
kh 2
[log Ouctrw
+ 162.6QN-1BU lo k(t-tN-1)
g
k ûctw
- 3.23 + 0.87s (4.16)
•
4-15
Collecting terms,
•
162.6q1Bu t 1
[iou
Pwf(t) pi kh t-tll
J
162.6q2BU [ t tl
kh log t-t2
162.6qN-1Bu t-tN-2
kh 1og t-tN-1
162.6qNBu k(t-tN-1)
kh log 2
$uctrw
(4.17)
- 3.23 + 0.87s]
0
Factoring qN,
162.6qNBu ql/qN
t
[iou ( )
pwf(t) pi kh t-tl
t-t 1 q2/qN
+ log (t-t2)
t-t NqN-1/qN
+...+ log (t-t-2
N-1
162.6qNBU k(t-tN-1)
-
.kh ûc r2
[log t w
- 3.23 + 0.87s] (4. 18)
•
4-16
• Expressing in a more general form,
N-1 qj/c{N
Pi pwf(t) _ 162.6Bu ^-1

log ( t - t. )
qN kh
^ )
j=1
+ 1.6kh6Bu log k(t-tN1) - 3.23 + 0.87s ^ (4.19)

t 2
^lacr w
^îQ.êcQ
safe-,
Or, we could leave it in a form similar to Eq. 4.15, -.e. , 'ft',mt 0-3 - ^
`û.iva.le,►.^.^ fv►^
N
pi pwf(t) 162.6Bu [ ii l 1 qN kh aN og(t tj-1)
j =l
^ m mR
• + 162khBU
log k 2- 3.23 + 0.87s ( 4.20) •
where qo and to are by definition equal to zero, ^-^- bM2

Equation 4.20, commonly referred to as the Odeh-Jones6
equation, is the most conveniently used form of the multi-
rate equation. It can be expressed in the form of a straight
line as
Pi pwf (t) ^ q 1
^N = mMR (4. 21)
Z log (t t j-1) + bMR
j=l
where the slope mrlR is
_ 162.6Bu
mMR kh (4,2j)
and the intercept bMR is
•
4-17
•
2- 3.23 + 0.87s . (4.23)
bMR 'MR log
ûctrw
Equation 4.20 indicates that multi-rate test data

should form a straight line when plotted as
pi
qN
pwf
vs F
q" q' 1
> qN> lo â (t-t ^. -1
j=1
on cartesian paper; such a plot is illustrated by Fig. 4.7.
.-, INTERCEPT = bMR

•
4J
Z
U,
SLOPE = mMR
•î
- •
N
qi qj -1
q log (t - tj-1
)
ti
.
j=1
Fig. 4.7: Typical plot of multi-rate

test data.
•
4-18
• It is important when making this plot to understand that
each measured pressure is associated with the rate
occurring when that pressure was measured.
The permeability and skin factor can be computed by
rewriting Eqs. 4.22 and 4.23, respectively, as
k = 162.6Bp (4.24)
m MR F
and
MR - log k + 3.23 . (4.25)
b s= 1. 51 m
MR 0uctr 2w
Example 4.2: Analysis of a multi-rate test (after Ref. 3)
^ Problem. During a drawdown test on an oil well, the

producing rate declined from 1,580 to 983 STB/D. Rate and
pressure data recorded during the test are tabulated.
Other reservoir data are:
pi = 2,906 psia
Bo = 1.27 RB/STB
p = 0.6 cp
h = 40 ft.
t q pwf t q pwf
(hours) (STB/D) (psia) (hours) (STB/D) (psia)
1.00 1,580 2,023 9.6 1,370 -
1.50 1,580 1,968 10.0 1,300 1,815
1.89 1,580 1,941 12.0 1,300 1,797
2.40 1,580 - 14.4 1,260 -
3.00 1,490 1,892 15.0 1,190 1,775
3.45 1,490 1,882 18.0 1,190 1,771
3.98 1,490 1,873 19.2 1,190 -
4-19
4.50 1,490 1,867 20.0 1,160 1,772
•
4.80 1,490 - 21.6 1,160
5.50 1,440 1,853 24.0 1,137 1,756
6.05 1,440 1,843 28.8 1,106
6.55 1,440 1,834 30.0 1,080 1,751
7.00 1,440 1,830 33.6 1,080
7.20 1,440 - 36.0 1,000
7.50 1,370 1,827 36.2 983 1,756
8.95 1,370 1,821 48.0 983 1,743
Solution. A conventional analysis of this drawdown data is

not recommended because of the large variation in rate during
the test. Consequently, the data are analyzed as multirate
data using a plot of
N
Pi pwf a' - q'-1
a vs - a ^ log (t - t j
N N ^
j=1
The calculations necessary to make this plot are summarized

in the following table:
Pi pwf
q
t pwf pi pwf aN
(hours) (STB/D) N (psia) (psi) (psi/STB/D)
1.00 1,580 1 2,023 883 0.5589 0.0000
1.50 1,580 1 1,968 938 0.5937 0.176
1.89 1,580 1 1,941 965 0.6108 0.277
2.40 1,580 1 - - - -
3.00 1,490 2 1,892 1,014 0.6805 0.519
3.45 1,490 2 1,882 1,024 0.6872 0.569
3.98 1,490 2 1,873 1,033 0.6933 0.624
4.50 1,490 2 1,867 1,039 0.6973 0.673
4.80 1,490 2 - - - -
5.50 1,440 3 1,853 1,053 0.7313 0.787
4-20
• 6.05 1,440 3 1,843 1,063 0.7382 0.819
6.55 1,440 3 1,834 1,072 0.7444 0.849
7.00 1,440 3 1,830 1,076 0.7472 0.874
7.20 1,440 3 - - - • -
7.50 1,370 4 1,827 1,079 0.7876 0.974
8.95 1,370 4 1,821 1,085 0.7920 1.009
9.6 1,370 4 - - - -
10.0 1,300 5 1,815 1,091 0.8392 1.124
12.0 1,300 5 1,797 1,109 0.8531 1.154
14.4 1,260 6 - - - -
15.0 1,190 7 1,775 1,131 0.9504 1.337
18.0 1,190 7 1,771 1,135 0.9538 1.355
19.2 1,190 7 - - - -
20.0 1,160 8 1,772 1,134 0.9776 1.423
21.6 1,160 8 - - - -
24.0 1,137 9 1,756 1,150 1.0114 1.485
28.8 1,106 10 - - - -
• 30.0
33.6
1,080
1,080
11
11
1,751
-
1,155
-
1.0694
-
1.607
-
36.0 1,000 12 - - - -
36.2 983 13 1,756 1,150 1.1699 1.788
48.0 983 13 1,743 1,163 1.1831 1.800
As an example of how the summation in the last column

is evaluated, consider the data measured at t = 7.5 hrs.
At 7.5 hrs, the rate q = 1,370 STB/D is the fourth rate to
be observed during the test; thus, N = 4, and
(qj -aj -1) log (t - tj -1)

N
j=1
= 1 ^(al-Qo) log (t-to) + (a2-al) log (t tl)

^ Q4
+ (q3-a2) log (t-t2) + (a4-c{3) log (t-t3)]
4-21
S
= 1 [(1580 - 0) log (7.5 - 0)
1370
+ (1490-1580) log (7.5-2.40)
+ (1440-1490) log (7.5-4.80)
+ (1370-1440) log (7.5-7.20)1
= 0.974.
Therefore, the abscissa value on the plot corresponding to

t = 7.5 hrs is 0.974.
A plot of the pressure data is presented in Fig. 4.8.
It is noted that the data form two straight lines. Without
further information about the reservoir, it is not possible
to conclude definitely what caused the increase in slope of
the data resulting in the second line. Several possibilities ^
include faulting, a decrease in permeability or mobility
away from the wellbore, or a transition to pseudosteady state
flow. Regardless of the cause of the second line, the slope
of the first straight line can be used to estimate the
permeability. From Eq. 4.24,
_ (162.6) 1.27 RB/STB)(0.6 c p)

k [0. 227 psi (STB D/cycle1 (40 ft)
k = 13.6 md.
It is interesting to compare this solution with the

solution obtained using the conventional drawdown analysis.
You will recall that the conventional analysis assumes that
the sandface flow rate is constant throughout the test.
Fig. 4.9 presents a graph of pwf vs log t for the test data.
Based on a slope of ml = -160 psi/cycle from this plot, the
calculated permeability is 28.1 md. Compared to the multi-
rate analysis, the conventional analysis gives a permeability
that is in error by 107 percent.
4-22
• • •
1.
(mMR)2 = 0.513 psi/(STB/D)

r^
pa
^
.^,
^
0.
^, z
^
W c}^ I:^L A
(mMR) 1= 0. 227 psi/ (STB/D)
iiI^ ^^^
0.
bMR = 0.557 psi/(STB/D)
0.
I.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
N
(aj - aj _ 1)
x log (t - tj 1)
q
j=1
Pig. 4.8: Plot of multirate data for Ex. 4.2.

•
2050
2000
1950
ml = -160 psi/cycle
ct 1900
^
w i :
1850 n
^l I
1800
m2 = -63 psi/cycle
1750
0
1700
1 10 60
t, hrs
Fig. 4.9: Conventional semilog plot of test data

in Ex. 4.2.
0
4-24
•
The conventional analysis is also misleading in other
ways. For example, the data on Fig. 4.9 show a decrease in
slope at a test time of approximately 12 hours. This would
indicate a possible increase in permeability or mobility
away from the wellbore which is just the opposite of the
effect shown by the more accurate multi-rate analysis.
Actually, the decrease in slope on the conventional plot is
caused by the changing flow rate; accordingly, this analysis
could be very misleading.
This example clearly illustrates the danger of using
constant-rate theory to analyze tests where significant
changes in rate occur.
The equations and procedures developed herein have

widespread application to many testing situations. It
will be shown in the next chapter that the pressure buildup
test is a special type of two rate test.
•
4-25
•
0
•
RE FE I,, EN CES
1. van Everdingen, A. F. and Hurst, W.: "The Application

of the Laplace Transformation to Flow Problems in
Reservoirs," Trans., AIME (1949) 186, 305.
2. Collins, R. E.: Flow of Fluids Through Porous Materials

Reinhold Publishing Corp. , New York ( 1961)_
.

Monograph Series, Society of Petroleum Engineers o
AIME, Dallas (1977) S.
4. Matthews, C. S. and Russell, D. G.: Pressure Buildup

Petroleum Engineers of AIME, Dallas (1967) 1.
5. Russell, D. C.: "Determination of Formation Characteristics

From Two-Rate Flow Tests," Trans., AIME (1963) 228, 1347.
(^. Odeh, A. S. and Jones, L. G.: "Pressure Drawdown

^ Analysis, Variable Rate Case," Trans., AIME (1967)
234, 960-64.
•
4-26
•
0
•
bMR = intercept of general multi-rate plot, psi/(STB/D)

bTR = intercept pressure on two-rate plot, psia
ct = total compressibility, psi-1
k = effective permeability, md
mMR = slope of general multi-rate plot, psi/(STB/D)/log-cycle
mTR = slope of two-rate plot, psi/log-cycle
p^ = false pressure from two-rate test, psia
p. = initial reservoir pressure, psia
pwf = flowing formation face pressure, psia
pwf (At=0) = flowing pressure immediately prior to the rate
change in a two-rate test, psia
plhr = pressure at At = 1 hour from straight line of two-rate
plot, psia
t = producing time, hrs
X = plotting function for two-rate plot
cp = porosity, fraction
At = test time during two-rate test, hrs
u = viscosity, cp
•
4-27
•
0
•
Equation Number
in Text Equation
162.6q Bu t +^t q
4.7 [1og + a2 log ^t
pwf = pi - kh 1 At
1
162.6q1Bu q2 r k
- log - 3.23+0.87s]
kh al L ûotrw
4.8 pwf mTRX + bTR
162.6q1Bu
4.9 MTR = - kh
• 4.10 bTR = pi + mTR a2 rlog k - 3.23 + 0.87s^

1 [1 ûor Z t w
tl + At q2
4.11 X = log At + a log At
1
162.6q1Bu
4.12 k = -
mT Rh
1 (pwf (At-ÔR plhr

4.13 s = 1.151
[q 1- q2
- log k + 3.23]
z
ûotrw
q2
-4.1PbTRql2[pwf(At=O) plhrl
•
4-28
N
Pi pwf qj Q^-1 ^
4.21 aN = mMR log (
t-t)-1)+byiR
j=1
m _ 162.6Bu
4.22
MR TE-
4.23
bMR = mMR L01- k Z - 3.23 + 0.87s^
t
ûorw
k = 162.6Bu
4.24
MMR
4.25 s = 1.151
r MR - log ( k 2 + 3.23
m IR ^ U ot r w
•
4-29
• PROBLEMS
MULTIPLE-RATE TESTING
1. Given below are the results of a two-rate test conducted

on a well producing off the coast of Trinidad. This
well produced at a rate of 1,841 STB/D for 28.4 hours
before the rate was increas ed to 3,523 STB/D. Estimate
the formation permeability and total skin facto r.
At, hours pwf, psia
0 1419
0. 25 1400
0.5 1384
0. 75 1358
^ 1.0 1335
1.25 1321
1.5 1310
1.75 1304
2 .0 1300
2.5 1286
3.0 12 80
3. 5 1274
4.0 1270
.4.S 1265
5.0 1261
6.0 1255
7.0 1249
8.0 1245
9.0 1241
10.0 1237
15.0 1219
20.0 1206
26.1 1200
4-30
•
L]
Z
W W
a
N0
yF1 N
p X
•
4-31
• Additional reservoir information:
uo = 1.63 cp
Bo = 1.63 RB/STB
h = 108 ft
^ = 30%
ct = 302 x 106 psi
rw = 0.365 ft
2. The discovery well in an oil reservoir had an initial

pressure of 4,000 psia. The well was then flowed
according to the rate history illustrated below.
100
Ln
50
0'-
0 6 12 18 24
t, hrs
Other reservoir data are:
Bo = 1.2 RB/STB ^ = 0.20

k = 100 md ct = 20 x 10-6 psi-1
h = 25 ft rw = 0.25 ft
u = 2.0 cp s = 10
•
4-32
a) Develop an equation which can be used to compute
the flowing formation face pressure for any time t
where t>18 hrs.
b) Use the equation from (a) to calculate the
expected formation face pressure at t = 24 hrs.
3. The following pressure-time data are from a step-rat

test on a water injection well. Prior to the test, the well
was shut in and the pressure was stabilized at 1000 psia.
a) Use the pressure-time data recorded during the
first three steps of the test to determine the permeability
and total skin factor. NOTE: Injection rate should be
treated as a negative producing rate in all calculations.
b) Replot the test data as in (a) but include the
pressure-time data measured during Step 4. Explain why the
data during Step 4 exhibit a different behavior than the
data in Steps 1-3. What additional information can be
determined by including the data from Step 4?
p; = 1000 psia
h = 80 ft
15%
c, = 6 x 10-6 psi-'
ru, = 0.3 ft
0 I
µw, = 0.7 cp
BW = 1.0 RB\STB
•
4-33
q
Point t piwi
No. hours psia
1 0 1000
STEP 1: q = 1000 bpd
2 1.000E-03 1159.7
3 2.000E-03 1192.2
4 3.162E-03 1208,5
5 5.012E-03 1223.7
6 1.000E-02 1245.9
7 1.585E-02 1260.6
8 2.512E-02 1275.3
9 5.012E-02 1297.3
10 0.100 1319.3
11 0.200 1341.4
12 0.283 1352.7
13 0.366 1361.2
14 0.450 1368.0
15 0.500 1371.4
16 0.501 1531.2
17 0.502 1563.8
18 0.504 1587.8
19 0.506 1602.9
20 0.510 1617.9
21 0.520 1640.6
22 0.550 1671.9
23 0.600 1696.8
24 0.700 1723.8
25 0.783 1738,8
26 0.908 1755.6
^
4 - 3 3A
E
Point t piwi'
Na. hours Psia
27 1.000 1765.4
28 1.001 1925.2
29 1.002 1957.9
30 1.003 1974.2
31 1.005 1989.6
32 1.010 2012.3
33 1.020 2035.3
34 1.040 2059.2
35 1.063 2076.1
36 1.100 2094.0
37
38
1.200
1.325
2123.8
2148.2
•
39 1.408 2160.8
40 1.500 2172.6
S'I'I?P 4: <I _ 4000 bpd
41 1.501 2041.7
42 1.502 2066.2
43 1.503 2081.0
44 1.505 2095.6
45 1.510 2117.7
46 1.520 2140.5
47 1.540 2164.7
48 1.563 2181.9
49 1.600 2200.5
50 1.700 2232.3
51 1.825 2258.9
52
53
1.908
2.000
2273.0
2286.5 •
4 -3313
a
SOLUTIONS
• MULTIPLE-RATE TESTING '
1. The calculations necessary to prepare the two-rate plot

are summarized in the following table:
q
At, hrs tptt log t^tt al log At x pwf, psia
0.25 114.6 2.06 -1.15 0.91 1400

0.5 57.8 1.76 -0.58 1.18 1384
0.75 38.9 1.59 -0.24 1.35 1358
1.0 29.4 1.47 0.00 1.47 1335
1.25 23.7 1.37 0.19 1.56 1321
1.5 19.9 1.30 0.34 1.64 1310
1.75 17.2 1.24 0.47 1.71 1304
2.0 15.2 1.18 0.58 1.76
•
1300
2.5 12.4 1.09 0.76 1.85 1286
3.0 10.5 1.02 0.91 1.93 1280
3.5 9.11 0.96 1.04 2.00 1274
4.0 8.10 0.91 1.15 2.06 1270
4.5 7.31 0.86 1.25 2.11 1265
5.0 6.68 0.82 1.34 2.16 1261
6.0 5.73 0.76 1.49 2.25 1255
7.0 5.06 0.70 1.62 2.32 1249
8.0 4.55 0.66 1.73 2.39 1245
9.0 4.16 0.62 1.83 2.45 1241
10.0 3.84 0.58 1.91 2.49 1237
15.0 2.89 0.46 2.25 2.71 1219
20.0 2.42 0.38 2.49 2.87 1206
26.1 2.09 0.32 2.71 3.03 1200
4-34
6
A plot of pwf vs X is presented in Fig. 4P.1.

Since the second flow rate is higher than the first •
rate, this test behaves like a drawdown test. It is
observed that the early time data exhibit the effects
of wellbore storage whereas the late time data trace
a straight line with a slope equal to -78.3 psi/log-cycle.
Using this slope, the effective permeability to oil is
computed using Eq. 4.12:
k = - (162.6) (1841) (1.63) (1.63)

- .3 1 8
k = 94.1 md.
The total skin factor is computed using Eq. 4.13,

i.e. ,
al (pwf(At=0) plhr)
s = 1.151
al a2 mTR
•
- log k + 3.231
ûctr 2
w
1841 1419-1315
s = 1.151
1841-3523 ( -78.3 )
- log + 3.231
(.3) (1.63) (302x10
94.1 -6) (.365) Z
s = - 2.0.
4-35
• • •
bTR = 1431 PSIA
1400
cl
.r.,
V)
U, ¢' 1300
plhr = 131S PSIA
mTR = -78.3 PSI/LOG-CYCLE
1200
0 1 2 3
log "At + q2 log At

At
q1
Fig. 4P.1: Plot of two-rate test data, Prob. 1.

,
---------------
q„-o
41aw
ouIe- Cv ►ti
"lXb
(D
Vln d êDn cs- J•w a"àQ^',o
^
2. a) This equation can be written using basic super-
position principles, or it can be obtained from
40 Eq. 4.20 by taking the indicated summation. Since we
are interested in predicting t during the third flow
rate, N = 3. Therefore, from Eq. 4.20,
pi q pwf = 16kh6BU (aa ao) log (t-t 0)

3 3
+ (qq al) log (t-tl) + (qa a2)log (t-t2)

3 3
+ 16kh6Bu [log k - 3.23 + 0.87s^.

ûctw
This can be simplified to:
_ _ 162.6Bu
al log t+(q2-ql) log (t-tl)
• pwf pi kh
+ (q3-q2) log (t-t2)1
162.6q3BU [log k.
2 - 3.23 + 0.87s^ .
- kh
Cµcrw
t
b) When t = 24 hrs,
4000 -(162 6) (1.2) 2.0) [(50) log (24)

pwf (100) (25
+ (100-50) log (24-6) + (50-100) log (24-18)

J
_ (162.6) (50) (1.2) (2.0)
(100) (25)
C log
-
100 -6 - 3.23+(.87) (10
(.2) (2.0) (20x10 ) (.25) 2
0
pwf = 3878 psia.
4-37
•
0
SOLUTION: PROB. 4-3, SMITH
SOLUTION:
(A) From the attached general superposition plot,
k=19md
s=0
• (B) The downshift in pressure-time data from Step 4 is caused by formation parting which resulted when the
injection rate was increased from 3000 to 4000 bbl/d. Formation parting causes a reduction in the skin factor
which, as shown by the Intercept term in Eq. 4.20 (Smith), results In a line with a smaller intercept than was
obtained using the data from Steps 1-3. The last pressure which was recorded before this downshift occurred,
i.e., p = 2172.6 psia at t = 1.5 hrs, is a good estimate of the formation parting pressure.
11
4-38
0. 4 Prob . 4-3: Gen . Superposition Plot .
•^ TP1 Pressure
no TP2 Pressure
n. TP3 Pressure
0.35
cc
0.3
7
A CU
W
tô
0-0.25
^
Model Results
Radial homogeneous
Infinitely acting
K = 19.0161 md
= 1521.2844 md.ft
S --0.0301
0. 2
0.1r,
0.001 0.01 . 0.1
Equivalent Time (hours)
This is a generalized superposition plot, referenced back to the initial pressure, of the first three steps of a step-rate test on a water injection
well. Since all three steps are below the formation parting pressure, all of these data represent transient radial flow and, accordingly, plot on a
common semilog straight line.
• is •
• • •
0.4 Prob . 4-3: Gen . Su p erp osition Plot .
TP1 Pressure
m TP2 Pressure
nn TP3 Pressure
°° TP4 Pressure n
•^
n
0.3 5 •
n
.-,
T
N ^n
n
M n
I °
^
h
Q °
(Y 0 . 3
cc
°
t ^ 0
a ^
0 IL
cm °
0
a
^
0
0.25
0.2
0.01 0.1
This is a generalized superposition plot, referenced back to the initial pressure, of the first four steps of the step-rate test. Since injection
pressures during the first three steps are below the formation parting pressure, all these data represent transient radial flow and, accordingly, plot
on a common semilog straight line. The increase in injection rate from Step 3 to Step 4 caused the formation to part and, because of the
decrease in total skin factor (or increase in effective wellbore radius), caused a downward shift of the data during Step 4. This shift in data at the
time of formation parting provides a method of measuring the formation parting pressure during a step-rate test.
•
0
•
Chapter 5
PRESSURE BUILDUP TEST
I. INTRODUCTION
The pressure buildup test is the most commonly used

pressure transient test. This test requires that a pro-
ducing well be shut in and the resultin^ increase in
formation face pressure be measured as a function of shut-in
time In addition to the assumptions made in deriving the
^ line source solution of the diffusivity equation, the basic
theory used to analyze pressure buildup data assumes that
the test well was produced at constant formation face rate
for a time, tp, prior to being shut in. Shut-in time is
denoted by the symbol At. The ideal rate history of a well
subjected to a buildup test is illustrated by Fig. 5.1.
The primary objectives of this chapter are to show how
the pressure buildup test can be designed and analyzed to
evaluate permeability, ^formation damage, average reservoir
pressure, andreservoir limits such as faulting. Common
problems of interpretation such as wellbore storage, boundary
effects, phase separation, and stratification will be discussed.
The use of both conventional and type curve methods of analysis
are presented and compared.
The pressure buildup test is generally preferred to
other testing methods for the following reasons:
0
•
0
a5
i• kr 6;wv, 5,c,^ 6e .tAtt- cetau'y 'ov- wtP,c. ^L0-o,^.,^

2. Ra1C. c^c,^, i^ w, SFc^,,,^.â^v^c.auo CJ
• 2- Czate, P Bt)
FLOWING
^Q q
U
z :â
q F
t 'A t
p
q
IT
0 SHUT-IN
0 t_(At=0) t+ A t
r'
TIME
Fig. 5.1: Ideal history of formation

face rate during a pressure
• ^ •
buildup test.
The problem of rate control, which is the

greatest disadvantage of flowing tests, is
eliminated since the well is shut in during
the test;
• Wellbore storage can be reduced, or eliminated,
by using a bottomhole shut-in device;
• Aarerag^nresst^r^ within the drainage volume of
the test well can be determined with a short
shut-in period;
• The test can be used on wells with certain types
of artifical lift where subsurface pressure
measurements would be difficult to obtain under
flowing conditions.
.
Primary disadvantages of the test are:
• Loss of production during the test;

• Redistribution of fluids in the wellbore during
shut-in can make analysis of some data difficult,
or impossible, if a bottomhole shut-in device is
not used;
• Well can sand up, or experience other mechanical
problems, during shut-in;
5-2
• Requires a reasonably constant rate for a period
of time prior to shut-in;
• The pressure buildup test is a two-rate test;
accordingly, superposition methods must be
used to evaluate the data.
The pressure buildup test is a two-rate test;.this can

be observed from the rate history presented in Fig. 5.1. This
means that pressure changes measured during shut-in are
influenced not only by shutting in the well, but by the flow
period prior to shut-in. Consider, for example, the pressure
behavior of a well which at the time of shut-in had been
producing sufficiently long for the boundary pressure to be
less than the initial pressure, pi. This is depicted in
Fig. 5.2 by the curve, Ato. When this well is shut in for a
buildup test, the pressure at the formation face will begin
to increase and a pressure disturbance will be propagated away
from the wellbore at a rate dictated by the formation diffusivity
and the nature of the flow period preceding shut-in. After
some shut-in time, Gtl, the shut-in transient will have moved
to a radius, rl. Between the formation face and rl, pressure
in the reservoir is increasing; beyond rl, the reservoir
has not been influenced by the buildup test and pressures are
still declining as a result of the flow period prior to shut-
in. The buildup test is said to be in transient, or infinite-
acting, flow at At l; however, it should be obvious that
pressures at the formation face are being influenced by the
boundary as a result of the flow period prior to shut-in.
The test will remain in transient flow until the shut-in
transient reaches the boundary; after this time, pressures
throughout the drainage volume of the well begin to equalize
and, given enough time, will equalize at the volumetric
average reservoir pressure, PR' Pressure equalization is
depicted by the shut-in time At3.
Figure 5.3 depictes the pressure behavior of a well
which at the time of shut-in, oto, was still in transient
flow. In this example, there is no influence of the boundary
5-3
pi Ato
Atl
At2
PR dt3
^
^
:n4
pwf(At=0)
rw r1 re
r2
DISTANCE FROM WELL
Fig. 5.2: Pressure buildup behavior of a

reservoir which has undergone
significant depletion at the
time of shut-in.
on the flowing pressure disturbance until shut-in time At3;

accordingly, the boundary will not affect the shut-in
transient, or the pressure at the formation face, until the
shut-in time is significantly greater than At3.
L By considering Figs. 5.2 and 5.3, it is obvious that
pressure behavior during a buildup test depends not only
upon the shut-in period, but also upon the flow period prior
to shut-in.] In particular, it will be shown that the length
of the flow period prior to shut-in, as well as the relative
lengths of flowing and shut-in periods, are important
parameters in determining the information that can be obtained
from a test, and the method which should be used to analyze
the test.
5-4
Atn Atl At2 At-
pi
PR At4
At
V
,,
a pwf(At=0) rw r1 r2 r3
re
DISTANCE FROM WELL
Fig. 5.3: Pressure buildup behavior of a

reservoir which is in transient
flow at the time of shut-in.
II. HORNER METHOD OF ANALYSIS,+

,,.-^.,.-..
The radial diffusivity equation is difficult to solve

analytically for the rate history depicted by Fig. 5.1.
However, we already have a solution for the constant rate
case; this solution was presented in Chapter 2 as Eq. 2.35,
i.e.,
2 kh Bu log kt 2- 3.23 + 0.87s . (5. 1)

pwf = pi - 16
ûctw
•
5-5
• You will recall that Eq. 5.1 applies to a constant rate well
producing from a reservoir which is infinite-acting, i.e., in
transient flow. This expression does not apply directly to
the variable rate history shown in Fig. 5.1; however, it can
be used along with the method of superposition to develop a
relationship that describes the pressure behavior to be
expected during a buildup test.
According to the superposition principle, the pressure
change, pi pws' which occurs at the formation face during a
shut-in period, At, is equal tol:
pressure drop caused by rate change, q, for

^ pi pws =(
time, tp + At) + (pressure change
caused by rate change, - q, for time, At)
or,
• pi pws (pi pwf)q + (pi pwf)-q

(5.2)
where: pws = formation face pressure at shut-in time

At, psia.
Figure 5.4 graphically illustrates this superposition process.

Observe that pws and pwf are actually the same pressure
since they both represent the formation face pressure of the
test well. Different symbols are used, however, to differ-
entiate the pressures before and after shut-in.
Equation 5.1 applies separately to each of the terms
on the right-hand side of Eq. 5.2; thus the total pressure
change which will have occurred after a shut-in time, at
is
162 kh qBu k (t +At)

l
log p 2 - 3.23 + 0.87s
pi pws -
ûctrw
• + 162.6(-q)Bu
kh
log kLt
2 3.23 + 0 87s (5.3)
Ouc trw
Vf,
5-6
q
•
-q
•
Q
C--^
tp (At=0)
TIME
Fig. 5.4: Graphical illustration of the

superposition principle applied
to the rate history of a pressure
buildup test.
5-7
0 A/.\. At -+
ar ^
& •
This reduces to: ^ ^ ^ ^-^- ^^ WOW 04^+ 01'
JWh9 •P'r n,,, SS^,IIc(, W.9.c,^,Q-t(-
+ At ( ^
_ _ 162.6qBu lo t p lpu;^ld ^ P^ (5.4)
pi pws kh g At ^
or,
t +At
162.6qBU P .X_ (5.5)*
pws pi kh log at
( (b ^ -Y -^ `----Y--^'
VVII
where ^s = shut-in formation face pressure, psia
pw
t = time well flowed before shut-in, hrs
P
At = shut-in time, hrs
q = rate before shut-in, STB/D.
Equation 5.5 can also be developed by setting q2=0 in the

general two-rate equation which was presented in Chapter
4 as Eq. 4.7.
The pressure buildup equation, Equation 5.5, was first
developed by ground water hydrologists2. It was introduced
to the petroleum industry, however, by D. R. Horne r3, and is
commonly referred to as the Horner equation.
1. Ideal versus actual pressure buildup behavior

The form of Eq. 5.5 suggests that a plot of pws versus
t +Ot
log ( ^ ) will yield a straight line of slope, m, where
162.6qBu (5.6)
m =
kh
0
This plot, commonly referred to as the Horner pZot, is
illustrated by Fig. S.S.
Note on the Ho rner plot that the scale of time ratio
increases from left to right. Because of the form of the
ratio, however, the shut-in time, At, actually increases
from right to left. Mathematically, the slope is a negative
number, and must be used as a negative number in all
5-8
Pi
SLOPE = m, PSI/CYCLE
^
^
^n
l.,^,^v At Y""°an
1 10 102 103 10"
t +At•
p
At
Fig. 5.5: Ideal Horner Plot of pressure

buildup data.
calculations.
It is observed from Eq. 5.5 that pws = pi when the time
ratio is unity. Graphically, this means that the initial
reservoir pressure, pi, can be obtained by extrapolating the
Horner straight line to (tp+At)/pt = 1. This is illustrated
by Fig. S.S. It must be remembered, however, that Eq. 5.5
assumes the reservbir to be infinite in size; if infinite
reservoirs did exist, then at some point in the reservoir
the pressure would always be equal to the initial pressure,
and the Horner straight line would always extrapolate to
that pressure. In practice, reservoirs are finite and, soon
after production begins, fluid removal will cause a pressure
decline everywhere in the system. Under these conditions,
the straight line will not extrapolate to the initial
piêssure. Instead, the pressure obtained at a unit time
ratio will be a false pressure, p*. This is illustrated by
Fig. 5.6. The false pressure has no physical meaning, but
•
5-9
^w etv,_ 1/4A d, Mtt
CJ jyko- 6'u4`'^ oU
UA'ta ! ^ UAA11
M2-
• no <a,^
p* FOR OLD WELLS

^-` FOR NEW WELLS IN
Pi A NEW RESERVOIR
ô • • ^•^
^
(A
a PR
102 103 104

1 10
t +^t
At
Fig. 5.6: Horner plot for limited re-servoir.
it can be used to determine the volumetric average pressure,

PR; this will be discussed in subsequent sections. [The only
....-^ .
time p* is equal to the initial reservoir pressure is when a
new well in a new reservoir is tested ]
After a well has been shut in for a period of time
sufficient for the shut-in transient to be affected by the
boundary, pressures will deviate away from the ideal straight
line as indicated by Fig. 5.6. Pressure behavior following
the end of transient flow will depend upon the drainage
geometry associated with the test well, the position of the
well relative to the drainage boundaries, and the length of
time the well produced prior to shut-in. If the well is
located in a reservoir with no other wells producing, the
shut-in pressure would eventually become constant and equal
to the volumetric average reservoir pressure, pR.
5-10
^
According to Eq. 5.5, pressures measured during the
transient flow period of a buildup test will plot as a
straight line on the Horner graph with a negative slope
equal to
m = - 162.6 ^
The measured slope can be used to determine formation

capacity as
kh = - 162.6 ^. (5.7)
Further, if the net pay is known, the effective permeability

to the producing fluid can be computed as
k = - 162.6 m . (5.8)
Example 5.1: Horner analysis of a new well.
Problem. The discovery well of an oil reservoir was flowed

during initial test at a constant rate of 150 STB/D while
producing a cumulative volume of 278 STB of oil. The well
was shut in for a buildup test and bottomhole pressure was
recorded for 18 hours as shown. Using the rock and fluid
property data given, estimate the effective permeability to
oil and the initial reservoir pressure.
h = 16 ft
= 13%
pwf = 290 psia ( measured immediately before shut-in)
rw = 0.25 ft
Bo = 1.67 RB/STB
PO = 0.7 cp
5-11
4
At, hrs pws, psia
0.5 925
1.0 1380
2.0 1800
3.0 1815
4.0 1820
5.0 1824
6.0 1830
8.0 1837
10.0 1843
12.0 1847
14.0 1852
16.0 1853
18.0 1856
Solution. The producing time prior to shut-in is:

is
t _ 278 STB 24 hrs

= 44.4 hrs.
day ]
p 150 STB
ay-
The measured pressures and corresponding Horner time ratios

are summarized in the following table:
At, hrs tp+At pws, psia

At
0.5 89.8 925
1.0 45.4 1380
2.0 23.2 1800
3.0 15.8 1815
4.0 12.1 1820
5.0 9.9 1824
6.0 8.4 1830
8.0 6.55 1837
10.0 5.44 1843
S-12
12.0 4.70 1847
•
14.0 4.17 1852
16.0 3.78 1853
18.0 3.47 1856
A Horner graph of these data is shown in Fig. 5.7.

Although not shown, the first two points fall below the
correct semi-log straight line because of near-wellbore
effects. The slope of the semilog straight line is
m = - 66 psi/cycle.
The effective permeability to oil is computed using Eq. 5.8:
k 162.6qBu
0 mh
k (162.6)(150 STB/D)(1.67 RB/STB)(0.7 cps)

0
(-66 psi/cycle) (16 £t) 0
ko = 27.0 md.
Extrapolation of the Horner straight line to a time

ratio of unity gives the initial pressure, i.e.
P4 = 1891 psia.
It is significant that the well in this example is a new well

in a new reservoir. This procedure will not give the initial
reservoir pressure except under these restricted conditions.
3. Estimation of ave a pressure in the drainage pore volume

o a test well
After a reservoir has produced for a period of time, it
is frequently desirable to determine the volumetric average
5-13
• • •
1900
p* = pi = 1891 PSIA
1880
1860 111 lltftffi i 11

SLOPE = m = -66 PSI/CYCLE
^
^
In
1840
u, ^
1820
1800
1780 L
1 2 4 6 8 10 20 30
t + ^t
P
At
Fig. S.7: Horner plot, Ex. 5. 1.

uc.(,Q. eJ
M6 Nwu -f^ + 0, to &o v,-t t,- !;C- 4,
pressure, pR, in the reservoir. This pressure is required

•
for many engineering calculations such as material balance
studies, water influx calculations, and reservoir performance
predictions. One way to obtain this pressure is to shut in
211 wells producing from the reservoir and wait for pressures
to equalize throughout the system; the stabilized pressure,
pR, could then be measured. Obviously, such a procedure is
not practical. Fortunately, the same information can be
obtained from a pressure buildup test. The purpose of this
section is to show how the average pressure in the drainage
volume of a particular well can be estimated from a buildup
test.
A. Matthews-Brons-Hazebroek (MBH) Method

It was noted previously that extrapolation of the Horner
straight line to a time ratio of unity yields a false pressure,
p*. It was shown by Matthews, Brons and Hazebroek" that p*
can be related to PD if the geometry of the area drained by
the test well and the location of the well relative to the
•
drainage boundaries are known. In particular,
_ * m (5.9)
PR P + 2.303 PDMBH
where PDMBH is the MBH dimensionless pressure determined at

the dimensionless ^.^.
p roducing time, tDA'
000 k 41
0.0002637kt ' WA S i- VAIQkAj' 'tDA
i&p- (5.10)
DA (ALt*X& (
kic,ca ba/14;o Cfh,wo,-tk- 7oA )

The relationship between PDMBH and tDA is presented in Figs.
5.8 through 5.11 for various drainage geometries. The pro-
cedure for using these curves to obtain pR,is:^
1) Prepare a Horner graph of the buildup data
2)
and identify the correct straight line;
Compute k from the slope of the Horner
straight line;
•
5-15
• • •
6
O
rn 3
N
U-7
11
10 10 - 1 1 10
tnA -- tArFG, ^^
• Fig. 5.8: h1atthews-13rons-Hazebroek dimensionless pressure for a well in the center

of equilateral drainage areas ( After Ref. 4).
6
3
M
O
M
U^
i fd
r-'
1 II 2
.i..
_1
ln 2 1()-1
tDA
Fig. 5.9: Alatthews-13rons-flazebroek dimensionless pressure for different well
locations in a square drainage area ( After Ref. 4).
• • •
7 • •
6 I LL 1 1. 1 } ^+ jfq ,Îl
lip
- . j k I} ^ ^} { t 1 l r. t i,
^ 1! i
^ _, , l .
#' { ,• t, ,lj '
1 1r-
; 1 ^^ I i
; TIT" Im 1_ f _` Ir L^!t I rt ` ^ j ,t r
. .
5 r ^ Fr !l^
.,
i ; T ^ `.
a ,.
^ y4, r ^ } ^ r 1Ij r^_
r i+ y?l+
1 l + r ^ I !f
t 1 i I f 44 ^ ^,
., ^^` L- ^ *rt
T ^' ^ll t:; ,' } ^^:.
- - r t ,^11 L lf 1^ r^c J I ^ - t-^ 1^ F^ f' E1 -- ^J •
4 ^ - t Fl1^ :
1I
^t «
It
t i!+
t
^ t }^- I^; t , ^
t t Ô !
+ j . ► r ^[_ ^^^^
Yr ^ r , E ^t
1 r 1 l^ ii y^
T I+ ^ ,• r ,
3
!r, - fl t
v
t ili t
1}
4
^
0 ^ t- ^
(.n _ ! . . h iti •l -r l -- * . -. _
-
^
a ^
rt i ^ •{ I
l rf l^l _^ l ^! I - t
!--'
00 ^ ^ ,
II 2 - - - - t I l,t - rt 44 f V
•• ^} :^ I ! ..* '^ -I^. ^ ^- - !,1
^, I . t. r . -i -
"O
O
or, +: y
+ P^
P,^ 1 ,, 1 „• j } t
^. ,t } - 1- ^ a ^ ^ ^ I,
- - 1 ^ If f^ ^ ^-! , = +
- _
Alt t . . ^ ^^ lj tl _- _ I 4ir i f ^ •.t ^{. , ,
1
- ^ + 1, j ^ t- - f`t ►-^^G ^ j ^, • ^ '` 1 'il' ^;^;'! lii
- - ! „! r
r "
i L6 :, + .^ _ I ;
- - tt
tl ' it
If I ' lit
1 r' :'t J^ i r I, l4 l l,
r
0
-^ ^^ ^ ^L T 1 1 `t • - ♦ -- ;_ j ^_ ^ ^ t ^ r7: *^to^^ P SS ,, I t ! , ^ ,-;
-1 ^ LT V ;,I til
I I
a^ 4 a a 1 9 t ^ a 4 • • 7 8 s t a a e a 7• s
_ 2
10 10- 1 I 10
t I)A
^ Fig. S. 10 Matthews-Brons-Ilazebroek dimensionless pressure for different well locations

in a 2.1 rectangular drainage area (After Ref. 4).
4
^
2
M 1.
(V
Ln
N II
lD
,z. 0
G-1
^D
-1
-2
10-^ 10-i 1 10
tI)A
Fig. 5.11: Matthews-Brons -tiazebroek dimensionless pressure for different well locations
in 4:1 and 5:1 rectangular drainage areas (After Ref. 4).
• 0 0
• 3) Determine p*;
4) Estimate the drainage configuration
associated with the test well; the
well pattern will give a clue to this
configuration. In the absence of other
information, assume the well is centered
in a square drainage area;
5)- Estimate the drainage area associated
with the test well. Use the area based
on well pattern for developed fields;
6) Compute tD q according to Eq. 5.10. Make
sure that the value of t used to prepare
the Horner Plot is used p to compute t DA;
7) Enter the appropriate curve and
determine PDMBH'
8) Compute PR using Eq. 5.9.
It is noted in Fig. 5.8 that the MBH curves essentially

overlay each other for wells centered in equilateral drainage
^ areas. Fortunately, many wells tend to be in the center of
their respective drainage areas and, even if the wrong shape
is assumed, the calculated pressure will be approximately
correct.
Of all the methods we will discuss, the MBH method is
the most flexible. It can be used to compute PR for a large
variety of drainage shapes and well locations, and is valid
regardless of how long the well was produced before shut-in.
If the well was produced to pseudosteady state before shut-
in, Pinson5 has shown that the method is independent of the
value of tp used to construct the Horner plot and, conse-
quently, to determine PR' This can be important because, as
we will discuss in a subsequent section, it is often difficult
to decide what value of tp should be used.
The primary disadvantage of the MBH method is the amount
of data required. In addition to tp, this method requires
that k, 0, u, ct, A, drainage shape, and well location be
known. If a free gas saturation exists, it is often difficult
^ to compute an accurate value of ct because of the uncertainity
in the gas saturation. Further, if the well location within
5-20
the drainage area is not accurately specified, significant
errors can result; this is apparent from Fig. 5.9 which shows
the effect of well location on the MBH dimensionless pressure
in a square drainage area.
Example 5.2: Determination of average reservoir pressure

using the Matthews-Brons-Hazebroek Method
Problem. A well producing from a reservoir developed on a

sguare pattern is believed to be located near the center of
a closed square drainage area. The well has been produced
at an approximately constant rate of 135 STB/day. An 18-
hour buildup test was run and the pressures were recorded
as shown. You are to determine the effective permeability
to oil and the average pressure, PR, in the drainage area
of the well at the instant of shut-in. The well has pro-
duced 26,325 STB of oil since it was last shut in.
Drainage area = 80 acres

^ = 15%
Po = 0.9 cp
ct = 13x10-6psi
B o = 1.63 RB/STB
h = 22 ft
r w = 0.25 ft
t +Ot
At, hrs pws, psia
P At
0.1 2504 46801
0.2 2668 23401
0.3 2769 15601
0.5 2882 9361
0.75 2953 6241
1.0 2980 4681
5-21
• 1.5 3006 3121
2.0 3019 2341
3.0 3030 1561
4.0 3037 1171
5.0 3043 937
6.0 3048 781
8.0 3054 586
10.0 3059 469
12.0 3066 391
15.0 3068 313
18.0 3071 261
Solution. The buildup data are presented on a Horner graph

in Fig. 5.12; the slope of the Horner straight line is -54
psi/log cycle. The formation permeability is, from Eq. 5.8,
k = 162.6 mqBW
^
k = - (162.6) (135) (1.63) (0.9)
-54 (22)
k = 27.1 md.
Extrapolation of the straight line to a time ratio of

unity gives a p* value of 3,203 psia. The producing time is:
t_ 26,325 STB (24 hr) - 4680 hrs.

p 135 STB-FD- D
This value of tp was used to prepare the Horner plot. The

dimensionless producing time, from Eq. 5.10, is:
0.0002637ktp
tDA ûctA
(0.0002637) (27.1) (4680)

t = 5.47.
DA
^ p^ (0.15) (0.9) (13xl0 ') (80) (43560)
5-22
1 10 102
3250
3200
31S0
Cd
.r.,
U)
3100
tn
W ^
3050
3000
IIJa Ma v-.,
2950
102
' 10 3 10"
Fig. 5.12: Horner plot, Ex. 5.2.
• • •
APC 206 2/01
Ansadariq3t PROJECT COMPUTATIONS

•
NO. OF SHEETS
DATE 0612)105 PROJECT .___ JOB NO.
BY CHK.: SUBJECT (O^C^'«.^P c{^ 3.ll,.^
0
•
•
0
The MBH dimensionless pressure correspondinga to this v a lu e
is, from Fig. S.91 pDMBH = 5.12; thus, from Eq. 5.9,
m pDMBH
PR = p* + -2.303
PR = 3203 + (5.2? 30354)
PR = 3083 psia.
'JYWK10d i4- Ie,u, ^t.cawae.- cCe-x&rcd a.,•^ uraô2 ^^t.^

• B. Odeh-A1 Hussainy Method
This method6 differs from the MBH method in that it does
not require knowledge of porosity, permeability, viscosity,
compressibility, or the drainage area. However, it does
require that the value of average pressure, pi, be known at
^ the beginning of the flow period preceding the buildup test.
Moreover, it assumes that the drainage volume of the test
well has not changed between the time pi was determined and
the time of the current buildup test. Factors which could
cause the drainage volume to change would include a change
in the producing rate of the test well relative to the rates
of adjacent wells, and the completion of additional wells
in the vicinity of the test well since pi was determined.
If the drainage volume has changed, significant error could
result in the computed value of pR. Finally, this method
assumes that the test well flowed sufficiently long before
shut-in to reach pseudosteady state.
Figures 5.13 and 5.14 present the necessary correlations
to determine PR- The procedure for using these curves is:
1) Determine p* from the Horner plot;

2) Compute the value of (p*-pi)/m;
3) Enter the appropriate curve-with the value
from Step 2 and obtain (PR-Pi) /M;
4) Compute PR'
5-24
6.0
•
5.0
4.0
•
3.0
2.0
1.0
0 4.0 4.5
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
6.0
^
5.0
4.0
• 3. 0
2.C
!.C
s 0
0 .5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.
S-26
•
Example 5.3: Determination of average reservoir pressure
using the Odeh-Al Hussainy Method
Problem. A well is completed in a reservoir developed on a

square drilling pattern and is assumed to behave as if it is
located within the center of a closed square. The well is
shut in for a pressure buildup test. The well has produced
for 1,548 hours since it was last shut in, when the average
pressure was 2,150 psia. You are given the buildup data,
and are asked to determine the current average pressure, PR'
The flowing bottomhole pressure at the time of shut-in, pwf,
was measured to be 1,699 psia.
t +Ot
At, hours pws, psia P At
0.133
0.267
2027
2097
11,600
5,770
•
0.40 2108 3,850
0.53 2113 2,910
0.67 2116 2,300
1.2 2122 1,285
2.0 2125 775
3.0 2127 515
4.0 2127.8 386
5.0 2128.5 309
6.0 2129.0 258
7.0 2129.6 221
8.0 2130.0 193
9.0 2130.5 172
10.0 2131.0 155
Solution. A Horner graph of the buildup data is presented in

Fig. 5.15. The straight line has a slope of -7 psi/cycle and
extrapolates to p* = 2,146 psia at a unit time ratio. The
5-27
• • 0
2150
2140
2130
a
2120
^
W
2110
2100
2090
1 10 102 103 10"
t' + At
P
At
Fig. 5.15: Horner plot, Ex. 5.3:

abscissa value of the Odeh-Al Hussainy correlation is
•
^_-
p pi _ 2146-21S0 = 0.57.
m -7.0
From Fig. 5.13, the ordinate value is read as
PR -Pi = 1.53;
m
therefore,
PR = 2150 + (1.53)(-7.0)
PR = 2139 psia.
C. Ramey-Cobb Method
•
Ramey and Cobb' showed that PR can be read directly from
the Horner straight line if the well reached pseudosteady
state before shut-in. The time ratio at which PR is read from
the Horner graph is:
t +At
(5.11)
At - = CAtDa.
pws-pR
In addition to the information required to compute tDA, the

size and shape of the drainage area must be known in order
to determine C A from Table 2.2.
4. Evaluation of Formation Damage
A. Total skin factor

The total skin factor, s, does not appear in the basic
pressure buildup equation previously presented as Eq. 5.5.
•
5-29
q
However, it is possible to develop a buildup equation which
includes the skin factor. The flowing pressure at the
formation face at the time the test well is shut in is,
according to Eq. 3.22,
ktp
_ 162.6qBp
pwf(^t=0) = Pi log
kh 2
Cuctrw
- 3.23 + 0.87s1 (5.12)
Subtract this expression from Eq. 5.5 to obtain:
kt
khgBU
pws - pwf(At=0) = 162.6 L Z- 3.23
log
ûctrw
t +At
+ 0.87s - log P (5.13)
• or,
Pws
Pwf(At=0) =-m log
ktpOt
2
Cuctrw (tp+At)
- 3.23 + 0.87s1. (5.14)
Equation 5.14 can be rearranged to obtain the skin factor:
p (Ot=O) - p"'s kt At
s = 1.151 ^'^f m - log 2p (5.15)
ûctrw(tp+At)
+ 3.231.
If At is small compared to tp, so that (tp+At) = tp, Eq.

5.15 reduces to
P wf (Ot=O)
m
-p
s 1.151 ws - log kAt2 + 3.23 , (5.16)
• ûcrw
t
5-30
LJ
Any value of At, and the corresponding value of pws
can be used in Eq. 5.16 to compute s; however, pws must be
obtained from the Horner straight Zine, not from the
measured data. Since these values are arbitrary, it is
common practice to select ^t = 1 hr; hence, the correspond-
ing pws is called plhr' With these substitutions, Eq.
5.16 takes the most commonly used form of the skin equation:
pwf(At-0) plhr k
s = 1.151 - log + 3.23 (5.17)
m ûctrw
It is again emphasized that plhr must be read from the Horner

straight line, or its extension, at a time ratio correspond-
ing to At = 1 hr.
B. Damage skin factor

The skin factor defined by Eqs. 5.15 through 5.17 is a
totaZ skin factor. As previously discussed in Chapter 3, s
is a composite of many skin factors as defined by Eq. 3.31:
•
s = sd + sr + sp + st + sf + ssw (5.18)
Accordingly,
sd = s- sr - sp - S. - s f - sSw (5.19)
The methods discussed in Chapter 3 to evaluate the skin

factors in Eq. 5.19 can also be applied to pressure buildup
tests.
C. Flow Efficiency
The flow efficiency was previously defined by Eq. 3.38
as:
E = PR pwf Aps
PR pwf
(5.20)
•
5-31
If PR is not available, it can generally be replaced with p*
without much loss in accuracy, i.e.,
p* pwf Aps
E _ (5 . 21)
p pWt
Equation 5.21 should not be used unless PR cannot be obtained.

The pressure loss due to skin required by Eqs. 5.20 and
5.21 can be computed as
Aps = - 0.87 ms. (5. 22)
If the purpose of evaluating the flow efficiency is to

predict the potential increase in production from a stimu-
lation treatment, ^ps should be calculated using the damage
skin factor, sd.
• D. Damage ratio and damage factor

Damage ratio and damage factor have the same definition
as for pressure drawdown tests, i.e.,
DR = 1 = PR pwf (S.23)
E -
PR - pwf - Aps
^ps
DF = l - E = (5 . 24)
PR pwf
If PR cannot be obtained, p* can be used as an approximation

in Eqs. 5.23 and 5.24.
.
E. Flow improvement
If the value of Qps is computed using the damage skin
factor, and if it is assumed that this damage can be removed
by a stimulation treatment, the flow rate after stimulation
should be
•
5-32
_ 1 0
(5.25)
aafter E abefore
fiquation 5.25 assumes that pwf is the same before and after
5.timulation.
Example 5.4: Evaluation of skin factor and flow efficiency

from pressure buildup data
Problem. Estimate the total skin factor and flow efficiency

for the well described in Ex. 5.2.
Solution. From Ex. 5.2,
m = -54 psi/cycle
3,005 psia
plhr =
PR
k
= 3,083 psia
= 27.1 md
•
pwf(At=0) = 2,260 psia
= 0.15
uo = 0.9 cp
ct = 13 x 10 6 psi
rw = 0.25 ft
[Pwf(t0) - lhr _ log k

= 1.151 + 3.23
m Ûctrw
= 1.151 r2260-3005
s
L -54
2 7. 1 6 + 3. 23]
- log
( 0 . 1S) (0 .9) (13x10 ) (0 .25) Z
s = 9.9. •
5-33
• The pressure loss due to this total skin is obtained from
Eq. 5.22:
Aps = -0.87ms
Aps = -0.87(-54)(9.9)
Aps = 465 psi.
Finally, the flow efficiency is computed using Eq. 5.20:
E = PR Pwf ©ps
PR pwf
E = 3083-2260-465
3 .5-226
E = 0.43.
^ It appears that this well has rather severe damage.

However, determination of sd, rather than s, would be
necessary to insure that the positive skin is not due to
some other restriction.
5. Determination of tp for the Horner analysis

The theory used to analyze pressure buildup data
assumes the test well flowed at constant rate for a time t
p
before being shut in. This ideal rate history was depicted
by Fig. S.I. [In practice, however, the flow rate of a well
varies throughout its life; further, a well may be shut in
a number of times for various reasons.I When these situations
occur, what value of tp should be used to analyze the data?
Horner indicated that tp could be approximated by
dividing the cumulative production by the flow rate
immediately before shut-in. In practice, it is generally
desirable to compute tas:
5-34
^
ative volume duced since last j ation(5.26)
t =
---* P constant rate iust before shut-
Some practical rules5ie for selecting tp under various

conditions are:
1) Permeability and skin calculations

a. When tp > tpss, correct values
of k and s can be obtained using
practically any value of tp to
prepare the Horner plot. A longer
straight line, however, will be
obtained by using tp = tpss to
prepare the Horner graph.
b. When tp < tpss, use the value of tp
computed using Eq. 5.26.
2) Average pressure calculations

a. MBH Method - If t > t ss, theoretically
any value of tp can bepused for the
•
analysis. Practically, a more accurate
value of PR will be obtained by using
tp = t ss. If tp < t pS s, use the value
of tp based on Eq. 5.26.
b. Odeh-Al Hussainy Method - Use the
value of tp computed with Eq. 5.26;
otherwise, serious errors will result.
c. Ramey-Cobb Method - Use the value of tp
computed using Eq. 5.26.
• A good summary of the literature dealing with the effect of

tp is presented by Earlougher9 and Dake10.
6. Wellbore storage
Wellbore storage is one of the most common and severe
problems encountered in the analysis of pressure buildup
data. The purpose of this section is to explain the cause
•
5-35
0
of storage, show its effect upon buildup data, detail how
it can be detected, present methods to predict the duration
of storage, and discuss ways of minimizing storage through
test design.
A. Cause of wellbore storage

The ideal theory used to interpret pressure buildup
data assumes that a well is shut in at the formation face
and that no fluid enters the wellbore after shut-in. In
practice, although downhole shut-off tools are available,
most wells are shut-in at the surface, and fluid continues
to enter the wellbore after shut-in. A comparison of ideal
and actual rate histories for a typical buildup test is
depicted by Fig. 5.16.
•
Q
w
4 r ACTUAL SHUT-IN
HISTORY
IDEAL SHUT-IN
HISTORY
0 tp(At=o)
TIME
Fig. 5.16: Actual versus ideal shut-in

history for a typical buildup
test.
•
5-36
This afterfZora of fluid is possible because, in effect, the
•
wellbore acts as a storage tank; fluids enter the wellbore
by causing the liquid level to rise or, if the wellbore is
full, by compressing the fluids in the wellbore. Wellbore
storage will continue after shut-in until the pressure in
the wellbore is sufficiently large to stop formation flow.
B. Wellbore storage factor

The effects of wellbore storage upon a buildup test can
be quantified in terms of the wellbore storage factor, C,
which was previously defined for pressure drawdown tests in
Chapter 3 as
SAV
C = (5.27)
^lbSap
where: C wellbore storage factor, bbl/psi

AV = change in volume of wellbore fluids
wb
at wellbore conditions, ft3
Ap = change in formation face pressure, psi.
When a well is shut in, the formation face pressure

increases and causes the fluid level in the wellbore to
rise; this is depicted by Fig. 5.17 for a well without a
packer. The wellbore storage factor for a rising fluid level
can be computed according to the relationship
144A
(5.28)
C = 5.6l5p . bbl/psi
where: A = cross-sectional area of wellbore

wb
affected by rising fluid level, ft2
p = average density of fluid in the wellbore,
lbm/ft 3.
For a well without a packer, Awb will be the sum of the

annulus area and tubing area. When a packer is present,
Awb will include only the cross-sectional area of the
5-37
•
FLOWING
Fig. 5.17: Wellbore storage caused by a

rising fluid level.
• tubing.
The compression of fluids within the wellbore can also
contribute to wellbore storage. The wellbore storage factor
caused by this effect is
Vwbcwb bbl/psi (5.29)

C_ 5.615'
where: Vwb = total wellbore fluid volume, ft3

cwb = average compressibility of fluid in
the wellbore, psi-.
.
The behavior of a buildup test will be dictated by the
total storage factor, i.e., the sum of Eqs. 5.28 and 5.29.
The primary storage effect in most oil wells will be caused
by a rising fluid level; however, when the gas-oil ratio in
the producing stream is large, gravity segregation of oil
• and gas in the wellbore can cause special problems which will
be discussed in a subsequent section. When working with gas
5-38
•
wells, the primary storage effect will be due to gas
compression.
The use of some type curves to analyze buildup data
requires that the wellbore storage factor be expressed in
dimensionless form. The dimensionless storage factor is
defined as
C = 0.894C (5.30)
D 2
t
^hcrw
C. Effect of storage on the semilog plot

Because the well continues to flow after shut-in,
pressures measured during the storage period will be less
than those anticipated in an equivalent well without
storage. The effect of storage on the Horner plot is
depicted by Fig. 5.18. The duration of storage, and the
deviation in pressure from the ideal straight line, will
increase as the wellbore storage factor increases.
D. Detection of wellbore storage

During early buildup times when wellbore storage
completely controls pressure behavior in the wellbore, it
can be shown, using an analysis similar to that presented
in Chapter 3 for pressure drawdown tests, that
P ws P wf (At=0) = qBAt (5.31)

24C
This equation assumes that the wellbore storage factor

remains constant during the buildup test.
Equation 5.31 suggests that data controlled by wellbore
storage will trace a straight line of unit sZope when plotted
as [Pws Pwf(At=0)] versus At on log-log graph paper. This
plot is illustrated by Fig. 5.19. The shut-in time, clt*, at
which data begin to deviate from the unit-slope line denotes
the end of complete control by wellbore storage.
5-39
9ums'tt.'vru.o, 1ti113 Aw.a'c^^ A,^r W" .
•
SLOPE = m
.^
Cs+
EFFECT OF
WELLBORE STORAGE
0 • •
10 102 103 104

t +At
p
At
• Fig. 5.18: Effect of wellbore storage on the

Horner plot.
After the unit-slope line has been identified, it can

be used to compute the storage factor of the test. Since
Eq. 5.31 is the equation of the unit-slope line, it can be
solved for C as
C= qBAt , bbl/psi. (5. 32)

24 [pws-pwf(Ot=O)]
Any values of At and [pws-Pwf(At-0)] which define a point on

the unit-slope line can be substituted into Eq. 5.32 to
compute C.
E. Duration of wellbore storage

The following guidelines can be used to approximate
the shut-in time at which wellbore storage effects become
• negligible:
5-40
D(a,wcxwm
CP; ' pwrp^ vs.
sQ',,.,e- tK,;j ... a p vs4,

•
10
10
^ 10
0.1
0.1 1 lU lU lU •
At
Fig. 5.19: Log-log plot of pressure buildup

data showing effect of wellbore
storage.
1) Neglecting skin, wellbore storage can be

considered negligible for shut-in times
greater than 50At*; at* is the shut-in
time where the unit-slope line ends on
the log-log plot of [pws pwf(At=0)] versus
• At (see Fig. 5.19).
2) Considering skin, wellbore storage
will be negligible for
0 •14s
At > 170,000uCe (5. 33)
kh
•
5-41
^ Equation 5.33 can b e ver y useful in the design of a
buildup test. The value of C can be estimated from the well
completion configuration; if values of k, h, u and s can be
estimated from other sources, the shut-in time required to
overcome storage effects can be estimated. If s is not known,
substituting s=0 into Eq. 5.33 will yield a minimum time for
storage to become negligible.
A comparison of Eqs. 5.33 and 3.72 suggests that the
skin factor has a larger effect upon the duration of storage
in buildup tests than it does in drawdown tests.
F. Design of test to minimize wellbore storage

Wellbore storage effects last so long in many buildup
tests that useable data are impractical, or impossible, to
obtain. In this situation, the use of a bottomhole shut-off
tool should be considered.
Several tools are available which permit a pressure
gauge to be run into the hole in conjunction with a tubing
pack-off assembly. This allows the well to be shut in at
the bottom of the hole and prevents fluid from rising in the
tubing during shut-in. Depending upon the remaining volume
in communication with the test interval, this procedure will
minimize, or eliminate, the effects of wellbore storage.
The potential to eliminate wellbore storage with a bottomhole
shut-off tool is a major advantage of the pressure buildup
test.
7. End of semilog straight line

The equation which predicts that buildup data will trace
a straight line on the Horner plot is no longer valid after
the reservoir boundary begins to affect the data; accordingly,
after some shut-in time, measured pressures will deviate from
the correct straight line. If a straight line is incorrectly
drawn through this boundary-affected data, serious errors can
• result. It is desirable, therefore, to predict the time when
5-42
boundary effects begin, i.e., when the correct Horner straight
•
line ends.
Cobb and Smith" developed correlations which relate the
shut-in time at which boundary effects begin, atesll to the
length of time the well produced prior to shut-in; this time
is
û ctA
(5.34)
Atesl 0 .0002 6 3 7 k (AtDA)esl.
The dimensionless time, (AtDA)esl' is obtained from Fig.

5.20. The procedure for using this correlation is:
1) Estimate k from core data, from

previous pressure transient tests,
from an adjacent well, or from
other available sources.
2) Compute tDA, i.e.,
0.0002637ktp
0
tDA ûcta
3) Enter the appropriate curve on

Fig. 5.20 and read (dtDA)esl'
4) Compute Atesl using Eq. 5.34;
the Horner straight line should
end at a time ratio corresponding
to this shut-in time.
It is observed from Fig. 5.20 that (AtDA)esl becomes

smaller as the producing time increases. If the producing
time is long, (AtDA)esl approaches an asymptotic value of
0.0042 for all drainage configurations; this is a fortunate
circumstance since the drainage configuration, nor the
parameters required to compute tDA, are accurately known for
many wells. When producing time is short, however, a good
estimate of drainage configuration and tDA is required to
compute an accurate time for the end of the semilog straight
line.
•
5-43
• • •
Table 5.1: Drainage shapes used in
Fig. 5.20 (After Ref. 9) .
CURVE
NUMBER SHAPE
^1
Î
,1 10'1 •
Cl) 3 1
Cn 4)
2
^
4 I
IO':
^ri-, 14 4
IO-3 10-= 1O"1 I
e I
t DA
r- -,
Fig. 5.20: Dimensionless time to
end of Horner straight 7 1 I
I I
line for shapes identi- L- _J
fied in Table 5.1.
NO FLOW
Data of Cobb and Smithll
------ CONSTANT PRESSURE
and Kumar and Ramey12
(After Ref. 9).
•
Example 5.5:* Pressure buildup design - wellbore storage and
boundary effects
Problem. A pressure buildup test is to be run in a well

which, based upon geologic data, is located between two
faults and is approximately centered in a 2 x 1 rectangular
drainage area. The drainage area is estimated to be 65
acres.
Depth to mid-point of
perforated interval = 8,493 ft
Average density of oil
in the wellbore = 40.3 lbm/ft3
Inside diameter of tubing = 1.995 inches
Cumulative oil production = 18,300 STB
Packer depth = 8,460 ft ^
h = 33 ft ct = 24.7x10psi 1
= 18.5% rw = 0.333 ft
u= 2.4 cp Bo = 1.34 RB/STB
co = 11.2x10-bpsi q = 235 STB/D
k = 56 md (from test on an adjacent well)
a) Compute the wellbore storage factor, C, for this well;
b) Compute the dimensionless wellbore storage factor, CD;
c) Estimate the length of time the well must be shut-in

before wellbore storage effects become negligible;
d) Estimate the shut-in time when boundary effects will

cause transient flow to end, i.e., when the Horner straight
line will end.
Solution. Because this is an oil well, the effects of fluid ^

compressibility upon wellbore storage will be small;
5-45
^
accordingly, only the rise in fluid level will be considered.
Further, since a packer is located immediately above the test
interval, the only volume affecting wellbore storage is the
tubing volume. The cross-sectional area of the tubing is
A = 7T(1.995 2 = 0.0217 ft2

wb (4)(144
a) The wellbore storage factor, considering only the rising

fluid level, is computed using Eq. 5.28:
144A wb
C = 5615 p
C - (144) (0.0217)
T-
5. 15 40 .
C = 0.0138 bbl/psi.
b) The dimensionless wellbore storage factor is computed

using Eq. 5.30:
C _ 0.894C
D
Ohctrw
(0.894) (0.0138)
CD = (0.185)(33)(24.7x106)(0.333)2
CD = 738.
c) The time required for storage effects to become

negligible is, from Eq. 5.33:
0•1VS
At = 170,00011Ce
Since the skin factor is not known, it will be assumed

zero; thus,
S-46
At = (170,000) (2.4) (0.0138)
•
( 56 ) ( 33 )
At = 3.0 hrs.
d) Boundary effects will begin to influence the data at

a time, according to Eq. 5.34, equal to
ûctA
^tesl 0.0002637k (^tDA)esl
Based upon cumulative production, the producing time is
(18,300 STB) (24 hrs)

tp 235 STB/D D = 1868.9 hrs.
The dimensionless producing time is:
0.0002637ktp
tDA ûctA
(0.0002637) (56) (1868.9)

tDA (0.185) (2.4) (24.7x10-6) (65) (43,560)
tDA = 0.89
From Fig. 5,20,
(AtDA)esl = 0.0042.
Therefore,
_ (0.185) (2.4) (24.7x10-6) (65) (43,560

^tesl - 0.0002637) 56
AteSl = 8.8 hrs
Notice that, because of the long producing tinie, the drainage 0

geometry has no effect on the length of transient flow in this
well.
5-47
•
Example 5.6: Pressure buildup test with wellbore storage
and boundary effects
Problem. The results of a pressure b uildup test for an oil

well completed on 80-acre spacing are presented along with
associated rock and fluid properties. This is an interior
well in a field developed on a square drilling pattern. The
objective of the test is to determine the effective permea-
bility to oil and the skin factor.
q = 95 STB/day
t = 505 hours
p
r = 0.25 ft
w
h = 15 ft
Bo = 1.24 RB/STB
ct = 15 x 10-6psi-1
uo = 0.85 cp
0 = 15%
k = 21 md (data from adjacent well)
t +At
At, hrs pws, psia p
At
0 pwf=3041 -
0.00158 3064 3.19 x 105
0.00315 3087 1.60 x 105
0.00631 3134 7.99 x 10"
0.0158 3222 3.19 x 10"
0.0315 3331 1.60 x 104
0.0631 3444 7.99 x 103
0.158 3533 3.19 x 103
0.316 3560 1.60 x 103
0.631 3580 8.0 x 102
1.58 3603 3.2 x 10z
3.16 3620 1.6 x 10 Z
5-48
6. 30 36 36 81. 0
15.8 3655 32.9
31.6 3668 17.0
63.1 3677 8.99
158.0 3681 3.19
Solution. A Horner graph of the buildup data is plotted in

Fig. 5.21. Analysis of the data requries identification of
the correct straight line. However, at least three straight
lines are present. The log-log plot, presented in Fig.
5.22, indicates a unit-slope line which ends at approximately
At = 0.007 hrs; multiplying this time by 50 according to the
SOAt rule, the effects of storage should end at approximately
0.35 hrs, i.e., a Horner time ratio of 1.44 x 103. This time
ratio corresponds to the end of line I; thus, it is concluded
that line I is caused by wellbore storage.
Test data from an adjacent well indicate a permeability
of 21 md. This value can be used to check for boundary effects.
The dimensionless producing time is
0.0002637ktp
tDA ûctA
t = 0.0002637(21)(505) = 0.42.
DA
(0.15) (0.85) (15x10-6) (80) (43560)
Entering this value on Fig. 5.20, the corresponding ordinate

value is (otDA)esl - 0.005; therefore, from Eq. 5.34,
ûctA
^tesl 0.0002637k (^tDA)esl
_ (0.15) (0.85) (15x10 6) (80) (43,560) (0.005)

At
esl 0.0002637) 21
Atesl = 6.0 hrs
•
5-49
• • •
3900
3800
I II. II'1
+ III
3700 1 inuiuniu
II
3600
V)
3500 m = -56 PSI/CYCLE
U,
0
I
3400
3300
3200 I t f f I l l l l lll lllllllllliitllll'NIINNfIIIIIIttiIIIIINIII^`N I I I I l l l l l lÌIIIIIiIIIIIIIIIItlNNINlllllllllmimifwl^ 111111111 lluuuuu iilumnnnn^!mu ulmnn^mn u

102 10 3 10'` 105
1 10
t + At
p
At
Fig. 5.21: Horner graph, Ex. 5.6.
IV
103
.14
V)
r-,
0
1
+) 102
NI 4-4
til UNIT SLOPE LINE
10
10 3 10-2 10 1
At, hrs
Fig. 5.22: Log-log plot, Ex. 5.6.
.7 • •
^ t +At _ 505 + 6.0
= 85.2.
( ^ )esl 6.0
Referring to Fig. 5.21, this time ratio corresponds

approximately to the beginning of line III; accordingly,
it is concluded that line III is due to boundary effects.
By elimination, line II must be the correct straight line.
The slope of line II is
m = - 56 psi/cycle;
k = - 16.6) (95) (1.24 (0.85)

thus, = 1g.4 md.
(15) -56
The pressure at Af = 1 hr from line II is
3,592 psia.
plhr -
^ Therefore, from Eq. 5. 1; ,
P wf ("t=') - p lhr _ log

s= 1.151 k + 3.23
m 2
ûcrw
t
= 1.151 3041-3592 lo 19.4

s
L -56
g (0.15)(0.85)(15x10-6)(0.25)z
+3,231
s = 5.6.
In summary, it has been shown that with the aid of the

log-log plot and the Cobb-Smith correlations, the correct
Horner semilog straight line can be identified. Thus, a
higher degree of confidence can be placed in the calculated
results.
5-52
8. Detection of faults and other flow barriers •
When a well is located near a sealing fault, or other
flow barriers, pressure behavior during a buildup test can
be significantly altered from that previously discussed.
If the pressure behavior for this situation is not under-
stood, an incorrect interpretation of the data will likely
be made. However, if the pressure analyst understands this
behavior, it is not only possible to detect the flow barrier
but, in many cases, the distance to the barrier can be
estimated. Figure 5.23 depicts a well located a distance,
L, from a sealing fault.
TEST
WELL
(,}." L
SEALING
FAULT •
Fig. 5.23: Well near a sealing fault.
The pressure behavior of a well near a sealing fault

can be predicted3 using the method of images; this super-
position technique was introduced in Chapter 2 for a flowing
well. In using the method of images, the effect of the
fault is simulated by assuming that a well identical to the
test well is located a distance L on the opposite side of
the fault; this well is called an image weZZ. The effect
of the image well is to cause a no-flow boundary equidistance
•
5-53
• between the wells, i.e., at the position of the fault. The
relationship of the test well and image well is depicted by
Fig. 5.24.
TEST IMAGE
WELL WELL
^- L L -=^
• Fig. 5.24: Method of images to simulate presence

of a sealing fault.
The pressure behavior of the test well can be predicted

by superimposing upon it the effect of the image well located
a distance 2L from the test well. It will be assumed in
^ this analysis that the fault is the only discontinuity in an
otherwise infinite svstem. It was shown in Chapter 2 that
the pressure change a distance L from a constant rate well
can be predicted, assuming infinite-acting flow, using Eq.
2.31:
948ûc L2
Ap = pi - p(L,t) = - 70.6 kghu Ei - kt t . (5.35)
When both the test well and image well are flowing at the
same rate prior to shut-in, the total pressure change at
^ time tp in the test well is, by superposition,
'^ Vâ w^ o êti 44^ Q,t^vt^. ^ ^t1a

a
5-54
^ (Pi Pwf)TEST + (Pi Pwf)IMAGE.
(5.36)
•
(Pi Pwf)TOTAL AT
TEST WELL WELL WELL
Therefore, the flowing pressure at the test well after

producing time, tp, will be
948ûctrw 2]
pwf = pi + 70.6 qB"
kh Ei kt
p
948ûc (2L)2
+ 70.6 k Ei - ktt . (5.37)
P
If both wells are shut in for a buildup test, superposition

can be applied to Eq. 5.37 to predict the expected buildup
pressure after a shut-in time At:
,
^ 948ouctrW
p ws = pi + 70.6 kh Ei k tp+At)
948ûctrW 948ûct(2L)2
- Ei - + Ei k tp + At)
kAt
948ûct(2L)2
- Ei (5.38)
kAt
For all practical test times, it was shown in Chapter 2

that the exponential integral terms which represent the test
well (r=rw) can be approximated by the relationship.
Ei(-x) = ln (1.781x).
With this substitution, Eq. 5.38 reduces to
t +4t
pws = pi - 70.6 khu {ln At
•
5-55
^
3792¢uc L` 3792ûc L2
- Ei - + Ei - t } (5.39)
k(tp+^t) kAt
large;^
it^,,.^, ^ Ss.
The values of L2 and t will generally be quite
thus, for early buildup times when At is relatively small,
the first Ei-term in Eq. 5.39 will be approximately constant.
Further, the second Ei term will be essentially zero. Under
these conditions,
_ 162.6qBU log tp+At + (5.40)

At constant.
pws pi - kh
According to Eq. 5.40, the early-time data will form a

straight line on the Horner plot with a slope identical to
that in a reservoir without a fault.
At longer shut-in times, when At is large, all of the
Ei terms in Eq. 5.39 can be replaced by the logarithmic
^ approximation, so that
t +At t +ot
_ 70.6 Bu (5.41)
pws pi - khq ln pt + In At
162.6qBU tp+At (5.42)

or, pws = pi - 2 log
kh At
According to Eq. 5.42, late-time buildup pressures will

form a Horner straight line with twice the slope of the
early data.
Comparing Eqs. 5.40 and 5.42, it is seen that a fault
causes two straight lines, the second having a slope twice
the first. This is illustrated by Fig. 5.25. This two-
slope behavior provides a means of recognizing the presence
of a fault from pressure buildup data.
The permeability and skin factor can be computed in
the normal manner using the slope of the first straight
line. The value of p* for use in computing p is obtained
by extrapolating the second line to a unit time ratio.
S -S6
^^n tv^^
•
.^
1 10 102 103 104
t +At
i?
At
Fig. 5.25: Theoretical Horner plot for

a faulted system. •
When tp >> At, the distance from the well to the fault
can be calculated using the following equation suggested by
Gray13:
k1,t
L = 0.01217 uct • (5.43)
The shut-in time, Atx, represents the point of intersection

of the two straight lines; this is illustrated by Fig. 5.25.
Equation 5.43, strictly speaking, applies only to drawdown
and injectivity tests. It is, however, an adequate approxi-
mation for buildup tests, although more rigorous methods
have been suggested9i.
Example 5.7: Pressure buildup test in a faulted reservoir
5-57
Problem. An oil well produced 4,410 ST B o f oil a nd was shu t
in for a buildup test. The average production rate was 105
STB/day prior to the test. The well is located in an area
where a fault is suspected. Given the rock, fluid, and build-
up data presented, estimate the distance to the discontinuity,
the effective permeability to oil,and the total skin factor.
qo = 105 STB/day
rw = 0.33 ft
h = 19 ft
= 160
Po = 0.87 cp
Bo = 1.27 RB/STB
ct = 18.4 x 106psi
t +^t
At, hrs P pws, psia
0 -- 2665
0.1 10,081 2940
0.2 5,041 2970
0.5 2,017 3009
1.0 1,009 3040
2.0 50S 3070
5.0 203 3111
10.0 102 3148
20.0 51 3192
50.0 21.2 3258
100.0 11.1 3311
200.0 6.04 3361
Solution. Given the Horner graph in Fig. 5.26, it is observed

that m2/m1 = 1.90 = 2.0, which suggests the presence of a
fault. From the intersection of the two straight lines,
•
5-58
3400
3300
3200
.^,
N
u,
3100
3000
2900
1 10 102 10 3 10"
t + At
P
At
• 0 0
• t
^
+Ot
= 73;
x
thus,
Atx = 14 hours.
Also,
162.6q oBouo
ko = - m1h
k (162.6) (105) (1.27) (0.87)

o ( -100 ) ( 19)
k = 9.9 md.
0
The distance to the discontinuity is computed using

Eq. 5.43, i.e.,
• L = 0.01217
k^t x
^^ct
(9•9)(14)
L = 0.01217
(0 . 16) (0 .87) (18. 4x10- 6)
L = 90 ft.
The total skin factor can be computed using Eq.

S.17:
= 1.151 rPwf(to)In lhr

loo k 2 + 3,23
ûcrw
t
9.9
s= 1.151 r2665-3040 - log
-100 (0.16) (0.s7) (1a.4x10-6) (0.33) 2-
+ s,231
s = -0.7.
5-60
Several studies14-'-a have been reported in which the
effect of parallel faults, and faults intersecting at various
angles, have been studied. Theoretically, it is possible to
compute the distance to each of the faults in a multiple
fault system, and the angle of intersection of the faults.
Practically, however, only the distance to the nearest fault
can generally be computed.
If formation permeability is low, or if a fault is
located a large distance from the test well, the fault may
not be detected because of the long shut-in time required
for the pressure transient to reach the fault. In wells
where the producing time prior to shut-in is short, it is
possible that a fault will not be detected by a buildup test
regardless of the shut-in time; this problem will be discussed
in a subsequent section.
9. Well not centered in its drainage area

It has long been recognizedl9-21 that a well which is
not centered in its drainage area°can exhibit abnormal
behavior. Fig. 5.27 illustrates pressure buildup curves19
for several closed reservoir configurations where the test
well is off-center. In each case, the correct semilog
straight line is evident at early shut-in times. At long
shut-in times, however, the pressures deviate in a manner
that is unique to the drainage configuration. In most cases,
the slope tends to increase as a result of the boundary
effect, resulting in shapes that could easily be misinterpreted
as faulting, fracturing, stratification, or other reservoir
heterogeneities.
A well can be off-center in its drainage area for a
variety of reasons which include assymetrical development
pattern, unequal production rates, variable pay thickness,
and reservoir heterogenities. In general, unequal producing
rates are the primary cause of assymetrical drainage
configurations. Under pseudosteady state flow conditions,
5-61
•
4
2
6
3 .^ ^. ^
• ^^^ ,q ,
. .)
4 ^^\ hq ^ •9
^ N
..^ .
1` • 0
04
.^ ^ 1
10 ^ 1 '4
12 3
1
• 14
4
F1 1
1 10 iu 2 1u 3 lo"
t +nt
P
At
Fig. 5.27: Typical behavior of wells not

centered in their drainage area
(Ref. 21) .
the volume drained by a well is approximately proportional' 22

to that well's flow rate, i.e.,
VP
Vpi = 1 (5.44)
qt
Likewise, if formation thickness is constant, the area
•
drained by a well is proportional to the flow rate of the
well.
5-62
The drainage configuration of a well in a complex
reservoir can be difficult to estimate without the use of
computer simulation studies. As a rough estimate, the
p"bsition of the drainage boundary between two wells will be
proportional to the relative rates of the wells. For
example, if two wells separated by a distance, d, are pro-
ducing at rates ql and q2, the approximate distance from
Well 1 to the boundary will be dql/(q1+q2). By considering
the relative rates of all adjacent wells, this method can
be employed to obtain an estimate of the drainage geometry
of the test well. While more rigorous studies of pressure
distribution and drainage geometry are availab1e1, 22, it is
generally difficult to apply these to field situations.
It is not necessary to know the drainage configuration
to compute permeability and skin factor, provided the correct
semilog straight line can be identified. However, an
accurate determination of average pressure using any of the
methods previously discussed requires that both drainage
configuration and drainage area be known. Observation of
the Matthews-Brons-Hazebroek curves, presented in Figs. 5.8-
5.11, shows clearly that an error in the drainage configu-
ration can result in a significant error in the calculated
value of average pressure.
Example 5.8: Pressure buildup behavior of a well off-center

in a closed drainage area
Problem. A pressure buildup test was simulated on the com-

puter for a well which produced at constant rate of 135 STB/D
for 10 days prior to shut-in. The well produces from a
square drainage area, but is located near one of the drainage
boundaries as shown in Fig. 5.28. The buildup data and other
reservoir information are presented. Compute the permeability
and average reservoir pressure.
5-63
^ ^r-- 933' ---^
233
Fig. 5.28: Drainage geometry
• for Ex. 5.8.
q = 135 STB/D rW = 0.33 ft

0
Bo = 1.26 RB/STB ct = 12 x 10psi
Po = 0.63 cp ^ = 11%
h = 16 ft A = 80 acres
t +At
At, hrs pws, psia
P At
0 pWf = 1511 -
0.095 2082 2.53x103
0.167 2094 1.44
0.23^ 2102 1.01
0.334 2110 7.2x102
0.447 2118 5.38
0.716 2128 3.36
0.955 2134 2.52
• 1.19.
1.62
2139
2146
2.03
1.49
5-64
2.14 2153 1.13
2.63 2157 9.23x10
3.44 2163 7.08
4.39 2170 5.57
6.20 2178 3.97
7. 16 2183 3.45
8.11 2186 3.06
9.55 2191 2.61
11.93 2197 2.11
14.32 2202 1.78
16.70 2206 1.54
19.09 2210 1.36
21.48 2213 1.22
23.86 2216 1.11
24.82 2217 1.07
28.64 2221 9.38x10°
31.03 2223 8.73
35 .8 0 2226 7 . 70
42.96 2231 6.59
47.73 2233 6.03
71.60 2241 4.35
95.46 2245 3.51
143.19 2250 2.68
190.92 2251 2.26
Solution. A Horner plot of the buildup data is presented

in Fig. 5.29. It is noted that two straight lines occur.
The line of slope m1 is the correct semilog straight line;
this line can be used to determine k, s and p*. The
second line of increased slope is caused by the proximity
of the test well to the nearest boundary. This is typical
of the behavior of any well which is closer to one boundary
than to the others. Without knowledge of the drainage
geometry, it would be easy to misinterpret this behavior
as faulting, fracturing, or a variety of heterogeneities
which will be discussed in subsequent sections.
5-65
• • •
2300
2200
cz
.^,
u, ^
o3
o, a
2100
2000
1 10 102 103 104
t t'At
P
At
_ .
From the first straight line,
m1 = - 53 psi/cycle
k (162.6) (135) (1.26) (0.63)

(-53)(16)
k = 20.5 md.
From an extrapolation of the first straight line to a

unit time ratio,
p* = 2,262 psia.
The dimensionless time, tDA, required to determine p is:
0.0002637ktp
tDA ^^;ctA
_ 0.0002637)( 20.5 )( 10 ) ( 24
(0.11) (0.63) (12x10-6) (80) (43,560)
)tD`^(
tDA = 0.45.
From Fig. 5.9, corresponding to the drainage shape in

Fig. 5.27,
-2.303(p*-pR)
= 0.44
0.44m1
PR = p + 2.303
PR = 2,262 + (0.4430351)
PR = 2,252 psia.
•
5-67
^
This is in agreement with the last buildup pressure which,
from Fig. 5.29, appears to be close to stabilizing at PR'
It is noted that the ratio of slopes, m2/ml, of the

two lines in Fig. 5.29 is 1.3. Since we expect a fault to
cause the slope to double, it would seem that this ratio
would discount the possibility of faulting. It must be
remembered, however, that a fault causes the slope to double
in an infinite-acting system, i.e., where the fault is the
only boundary affecting the well. In reality, a sealing
fault is nothing more than a no-flow boundary and, conse-
quently, its effect on pressure data will be the same as any
other drainage boundary. Therefore, a well with a fault,
but with other drainage boundaries, will exhibit the same
buildup behavior as a well near the drainage boundary in a
^ non-faulted system. This means that a fault, in the presence
of other boundaries, can cause a slope ratio significantly
different from two; likewise, a doubling of slope does not
necessarily indicate the presence of faulting. It must be
concluded, therefore, that pressure transient tests cannot
detect faults with a high degree of certainity unless there
is other evidence, such as seismic data, geologic data, or
production performance, to support the conclusion.
10. Effect of producing time on buildup behavior

It was shown in previous sections that certain types of
reservoir boundaries will cause predictable and recognizable
characteristics in pressure buildup data. For example, a
fault located in an otherwise infinite system will cause two
semilog straight lines where the second (late-time) line has
a slope twice the first. Unfortunately, most of the theory
^ developed to date concerning pressure buildup in wells
located near reservoir discontinuities assumes that production
5-68
times prior to buildup were long enough for the discontinuity
to influence the pressure drawdown behavior of the well.
Many well test conditions, however, do not permit long pro-
ducing times; this is particularly true in drill stem tests,
'and in new wells where storage facilities are limited and
pipeline connections have not yet been made. The objective
of this section is to discuss the effect of short flow times
on the buildup behavior of wells located near reservoir
discontinuities.
Recent studies18,2s,26 clearly show that the buildup
behavior of a well located near reservoir discontinuities is
strongly dependent upon the length of the flow period
preceding shut-in. If a well in the presence of a discon-
tinuity is flowed sufficiently long before shut-in for the
effect of the discontinuity to be fully developed in the
drawdown behavior, then the buildup data, providing the
well is shut in long enough, will exhibit characteristics
which are indicative of the type of discontinuity present.
However, if the well is not produced sufficiently long for
the discontinuity to fully influence the drawdown behavior,
two possibilities exist: First, the discontinuity may have
no effect on the buildup data, regardless of how long the
well is shut in or, second, the buildup data will exhibit
anomalous behavior which is not indicative of true reservoir
characteristics, and which can lead to an erroneous interpre-
tation.
In a study of buildup behavior of wells near a linear
no-flow boundary in an otherwise infinite system, Streltsova
and McKinleyl$ showed that the second straight line of slope
2m would not appear on the Horner plot, regardless of the
shut-in time, unless the well produced for a dimensionless
time before shut-in equal to
0.0002637kt
tDL = 2 p = 10. (5.45)
û c tL
0
5-69
This means that the r a dius of investi gation, R1., of the flow
period would have to meet the criterion that
kt _'Z
R. = 0.029 u^ = 6.32L. (5.46)
t
Equations 5.45 and 5.46 represent the conditions required

for the line of slope 2m to begin. Practically, it takes
almost a cycle of data to recognize accurately the presence
of a straight line. In order for the line of slope 2m to
span one log cycle, the minimum criterion would be
tDL = 100, R. = 20L (5.47)
Similar results were found in a study of faulted systems

by Earlougher26.
Streltsova and McKinley18 also considered the effect
of producing time on buildup behavior in other types of
heterogeneous systems. Systems studied included linear
constant pressure boundaries, parallel boundaries of infinite
extent, radial discontinuities, uniformly distributed
heterogeneities, and vertical stratification. In all cases,
the criteria established by Eqs. 5.45 and 5.46 had to be
met before the true nature of the discontinuity could be
determined.
In summary, flow time and shut-in time, not shut-in
time alone, determine if a discontinuity can be detected and
accurately characterized by a buildup test. In general, the
flow time must be long enough so the radius of investigation
of the flow period ;s several times longer than the distance
to the discontinuity. This problem has only been studied for
simple geometric systems; it is clear that additional work
needs to be done in this area.
Short producing times are also a problem when using
^ drawdown type curves to analyze buildup data. This problem
will be discussed in a subsequent section which covers the
subject of type curve analysis.
5-70
11. Multiphase flow
•
The pressure buildup equations can be modified23,24 to
account for multi-phase flow. The Horner equation is
tp+At
pws = pi - 162.6 a.h log (5.48)
At
t
where qRt, the total flow rate, is defined as
q Rt = c1oBo +(1000qgt - qoRs - qwRsw)Bg + awBw (5.49)
where: qRt = total production rate, RB/D

dgt = total gas production rate, Mscf/D
= water production rate, STB/D
qw
Rsw = solution gas-water ratio, scf/STB
•
Bw = water formation volume factor, RB/scf.
The total mobility, .1t, can be determined from the slope

as
at = -162.6 m'F
a (5.50)
k k k
Since a = -o + g + W (5 . 51)
t ug Ow
PO
it follows that the permeability to each mobile phase

can be computed as:
a oB ouo (5 .52)
k = -162.6
o mh
k = -162.6 awBwuw (5.5.s)

w mh
(1000qgt - qoRs - awRsw)Bgug
kg = -162.6 (5.54)
mh
0
5-71
The total skin factor can be computed as
pwf(At=0)-Plhr log + 3.23 (5.55)

s = 1.151 -
m ^ctrW
Example 5.9: Multiphase buildup test analysis
Problem. Rock, fluid and well properties are given for a

well which produced oil, water and gas prior to being shut
in for a buildup test. The flowing pressure at the time of
shut-in, pwf(At=0), was measured to be 1,778 psia. From
the transient flow straight line on the Horner plot the
slope, m, was measured to be -61 psi/cycle, and the pressure
from the semilog straight line at At = 1 hour, Plhr' was
determined to be 2,216 psia. Estimate the total mobility,
^
^..^ the permeabilities to oil, water and gas, and the total skin
factor.
S o = 0.54 B = 1.33 RB/STB

0
S g = 0.10 Bw = 1.02 RB/STB
S w = 0.36 Bg = 1.25x10-3RB/scf
cw = 3.5x10 -6psi-1 = 19%
c o = 22xl0- 6psi-1 rw = 0.33 ft
c = 420x10 -6psi-1 h = 31 ft
C9 = 4.3x10 -6psi-1 qo = 278 STB/D
uo = 1.3 cp qw = 67 STB/D
uw = 0.6 cp qg = 720 Mscf/D
ug = 0.026 cp Rs = 705 scf/STB
R 0
sw
Solution. Calculation of the total mobility requires that

we first know the total flow rate. From Eq. 5.49,
^
5-72
qRt = qoBo + (1000qgt - qoRs)Bg + QwBw
•
aRt (278) (1.33)+[(1000) (720) -(278) (705)] (1.25x10-3)
+ (67) (1.02)
aRt = 1093.1 RB/D.
The total mobility, at, is computed using Eq. 5.50:
at = - 162.6 ami
(162.6) (1093.1
^`t -61 ( 31 )
at = 94 md/cp.
From Eq. 5.52, the effective permeability to oil is: •

k = - 162.6 aoBouo
0 mh
(162.6)(278)(1.33)(1.3) _
(-61) (31)
ko = 41.3 md.
From Eq. 5.53, the permeability to water is:
k = - 162.6 q wBwuw
w mh
(162.6) (67) (1.02) (0.6)

( -61) ( 31)
kw = 3.5 md.
•
5-73
• The permeability to gas is determined from Eq. 5.54 as
162.6 (1000q gt - qoRs)B u

k = -
g
k = - 162.6 [(1000 20)- (278) (705)] (1.25x10- 3) (0.026

g -oi ) Tm S
kg = 1.5 md.
The total compressibility is needed before the skin

factor can be computed, i.e.,
ct = coSo + c w S w + cgSg + cf
ct = (22x10-6) (0.54) + (3.5x10-6) (0.36)
+ (420x10-6) (0.10) + 4.3x10-6
• ct = 59.4x106psi
From Eq. 5.55,
pwf (Qt-0) plhr log + 3.23

s = 1.151 -
m ^ctrw
s= 1.151 L1778-2216 _ log 94

-61 (0.19) (59.4x10-6) (0.33) 2
+ 3.231
s = 2.9.
5-74
^
1 J
1^. Effect of rate variation prior to shut-in

The Horner method of analysis assumes that the flow
rate prior to shut-in is constant. Most wells tested,
,-however, have had a variable rate history at the time of
the test. Consider, for example, the rate history depicted
by Fig. 5.a0. How should the pressures measured during the
shut-in period be analyzed?
ql
`1 N - 1
q
-3:
•
q=0
0
,
tl t2 tV-2 tN-1 tN
TIME
Fig. 5.30: Rate variation preceding a pressure

buildup test.
The most commonly used approach to analyzing buildup data

preceded by a variable rate history is to ignore the rate
changes and compute a pseudo-producing time according to
Eq. 5.26. This producing time is used to prepare the Horner
plot, and is used along with the producing rate at the time
of shut-in to make all calculations. This method, generally
referred to as the Horner approximation, is an adequate
approximation provided the well has been flowing at constant
5-75
• rate prior to shut-in for a sufficient length of time
that pressure changes caused by previous rate changes
are negligible during the shut-in period.
An obvious question is: How long must the well be
flowed at constant rate prior to shut-in in order for
the Horner approximation to be sufficiently accurate. Two
guidelines will handle most testing situations: First,
if the last rate is maintained constant sufficiently long
for the radius of investigation (Eq. 3.14) achieved at this
rate to reach the drainage radius of the well, this method
will always be adequate27. Second, if a well undergoes a
rapid series of rate changes immediately before shut-in,
this approximation will generally be adequate if the last
rate is maintained constant for a minimum of three times as
long as the length of the previous rate. Referring to Fig.
5.30, the requirement of this second condition is that
• tN - tN-1 > 3(tN-1 tN-2 ).

(5.56)
When rate variations occur and the producing time is

shorter than the shut-in period, the method of Odeh and
Selig28 is preferred to the Horner approximation. This
method is similar to the Horner method in that a modified
production time is defined as
N
L qj(t^ - t2 1)
tOS = 2tp - N (5.57)
aj (tj - tj-1)
j=1
and a modified flow rate as
qOS = tl qj (t j- tj (5.58)
OS
• j=1
5-76
The value of tp in Eq. 5.57 is the actual flow time, not the
•
Horner time. The regular Horner analysis is made using t0S
and qOS in place of tp and q.
Finally, the most rigorous analysis would be to treat
the test as a general multi-rate test. In this case, Eq.
4.20 can be modified to express the shut-in pressure, pws, as
N
tN tj 1+ At
pws pi + m n
L log ( t ^ tJ + At (5.59)
L
j=1
Equation 5.59 indicates that a plot of pws versus the

summation on the right-hand side should yield a straight
line with slope m, where m is defined by Eq. 5.6.
Permeability can be computed using Eq. 5.8, and s de-
termined from Eq. 5.15 where tp is replaced by (tN - tn-1)'
•
Example 5.10: Buildup test analysis with rate variation
before shut-in.
Problem. Odeh and Selig28 presented the following data for

a pressure buildup test on an oil well.
At, hrs pws, psia
2 2813
3 2838
5 2872
7 2895
9 2910
11 2919
13 2930
15 2955
17 _94?
0
5-77
• Bo = 1.0 RB/STB
uo = 0.6 cp
The rate history prior to the test is presented in the

following table:
j ti, hrs qj, STB/D
0 0 0
1 3 478.5
2 6 319.0
3 9 159.5
Compute the capacity, kh, of the formation tested using the

general multirate analysis.
^ Solution. The following multirate analysis was presented

by Earlougher9. The summation in Eq. 5.59 can be written
as follows for the data presented (N=3):
N j t. -1 + At
(t N ^ .
- 10
qN g tN - t j + At )
j =1
478.5 9-0 + At
- 159.5 log (9-3 + Ot
319.0 9-3 + At
+ 159.5 log (9-6 + Ot
159.5 9-6 + At
+ 159.5 log (9-9 + At
= 3 log (g + At + 2 log (6 + At + log (3 + At

6+ At 3+ At At
^ The following table summarizes the summation as a function of
5-78
At and pws
•
At, hrs pws, psia
2 1.2210 2813
3 1.0280 2838
S 0.7949 2872
7 0.6533 2895
9 0.5563 2910
11 0.4851 2919
13 0.4305 2930
15 0.3871 2935
17 0.3517 2942
A plot of the buildup data is presented in Fig. 5.31.

The slope in Fig. 5.31 is m=-153 psi/cycle.
From Eq. 5.7,
0
kh = - 162.6 qB^'
m
kh = - (162.6) (159.5) (1.0) (0.6)

-153
kh = 102 md-ft.
Odeh and Selig28 reported that their method resulted

in a value of 97 md-ft and the conventional Horner plot
gave 77 md-ft. The correct value of kh for this simulated
data is 106 md-ft. It is clear in this example that the
rate variation before shut-in had to be taken into account
in order to obtain a correct analysis. Further, it is
observed that the Odeh-Selig method and the general multi-
rate analysis are in good agreement.
•
5-79
^ 2950
2920
m = -153 PSI/cycle
2890
^
Ln
;7-
2860
2830
is
2800
0.2 0.4 0.6 0.8 1.0 1.2 1.4
N log ( t N tJ-1++ At
> aN tN tj t
j=1
Fig. 5.31: Plot of buildup data, Ex. 5.10.
•
5-80
_k III. MILLER-DYES-HUTCHINSON ( MDH) METHOD OF ANALYSIS
•
Another widely used method of analyzing pressure buildup
.data was suggested by Miller, Dyes and Hutchinson29. This
method is based upon the solution of equations which describe
the pressure behavi^r ^n a f^n^tP rese^voir. The pressure
drop (pi-pws) at the wellbore at any time after shut-in is
obtained by superimposing the pressure drop due to a flow rate
q, for a time tp+At, upon the pressure drop caused by a flow
rate -q acting for the shut-in time At. This is the same
procedure used to obtain the Horner equation except we are
now considering a finite reservoir. Therefore, by superposition,
pi pws pa]tp+At + [Apq]tp+At (5.60)

tp
0
If it is assumed that a well is produced to pseudosteady
state prior to shut-in, the pressure change due to the rate
change q can be written in terms of Eq. 3.9:
_ rp_p tp+Ot _ 162.6qBU log 4A

Ap
q i wf^ o kh
L L 1.781cA rWi
_ 0.2339 Bt (5.61)
^ct
The pressure change caused by shutting in the well,

assuming that shut-in times are sufficiently short that
the shut-in pressure transient is not affected by a boundary,
can be written as
_ t +At _ 162.6 (-q) Bu kAt

A [Pipwf]t pkh log
p-q c r2
t utw
p ^
- 3.23 + 0.87s] . (5.62)
5-81
^ Substituting Eqs. 5 . 6 1 and 5.62 into Eq. 5.60 yields
p - p = 16^ log ( 4A + 0.87s

1 ws h
1.781Cqrw
+ 0.2339qB(tp+At)
162.6qBU log kAt
Ah^ct kh
ûc t w2
- 3.23 + 0.87s]. (5.63)
It is assumed that the producing time is sufficiently long

that t p + At ~ t p ; with this assumption, Eq. 5.63 can be
written as
pws = pi + 162khqBU log At
^ + 162.6qBU
log k - log 4A - 3.23
kh ûctrw 2 1.781CArW
0.2339qBtp ^Q
(5.64)
Ah^ct
Recognizing that pws and At are the only variables in

Eq. 5.64, it follows that a semilog plot of pws versus At
should yield a straight line of slope m where
m = 162.6 ^ . (5.65)
It is observed, by comparison,that the MDH and Horner plots

should yield semilog slopes of the same magnitude; however,
the MDH slope is positive, whereas the Horner slope is
negative. A schematic of the MDH plot is illustrated by
Fig. 5.32.
5-82
•
ca
I-q
tA
0.1 1 10 102 103
At, hrs
Fig. 5.32: Miller-Dyes-Hutchinson plot

of buildup data.
•i
After the correct semilog straight line has been
identified, permeability can be computed as
k = 162.6 ^ (5.66)
V^
2. Evaluation of formation damage

The skin factor equation for the MDH method is developed
in the same manner as was presented for the Horner method;
• this equation is
P lhr 'P wf (At=0) - log

s 1.151 k + 3.23 (5.67)
L m ûctrw
Recall that m is a positive number on the MDH plot;

accordingly, it must be used as a positive number in 0
5-83
• Eq. 5.68, as well as in other MDH equations to be presented.
The pressure loss due to skin is
Aps = 0.87ms. (5.68)
Equations for flow efficiency, damage ratio and damage

factor, are identical to those presented for the Horner
method.
3. Average pressure in drainage pore volume of test well

The amount of reservoir fluid expansion which occurs
in reducing the average reservoir pressure from pi to pp is:
AQ = (Ah^) ( pi-PR) ct, ft3. (5.69)
If this fluid is produced at an average rate, q, over a
• period of time, t, we can also write:
AQ = (qBtp)(5.615 bbl) (24dhrs ft 3. (5. 70)
Equating Eds. 5.69 and 5.70,
ah^ct(pi-pR) = 0.2339qBtp (5. 71)
0 . 2 339qBtp
or, PR = pi (5.72)
Ah;pct
Subtracting Eq. 5.64 from 5.72 yields
p = 162.6qBU 3.23 + log 4A

PR ws kh
1.781CArw2
k4t (5 .73)
- log
• z
ûctrw
5-84
It is now possible to determine the value of At to
•
which the MDH straight line must be extrapolated to obtain
Since we want to know when pws
PR' PR' PR can be
substituted for pws in Eq. 5.73 to obtain
ûct
(at) PR = 3792 kC (5.74)
A
The value of pR can be read directly from an extrapolation

of the MDH straight line at a value of shut-in time given
by Eq. 5.74. This result, known as the Dietz Method',' is
only valid if the well was produced to pseudosteady state
prior to shut in. The value of CA in Eq. 5.74 can be
obtained from Table 2.2.
4. Wellbore storage
Wellbore storage affects the N1DH plot in the same
manner it affects the Horner plot. Wellbore storage can
be detected using the log-log plot, and the duration of
storage is predicted using the same equations presented
for the Horner method.
5. End of semilog straight line

The time to the end of transient flow, i.e., the
time to the beginning of boundary effects, can be predicted
using Eq. 5.34:
ûctA
(5.34)
Atesl 0.0002637k (AtDA)esl'
The dimensionless time, ( AtDA)esl' is obtained from Fig.

5.33. Each curve in Fig. 5.33 represents a different
drainage configuration; the drainage geometry corresponding
to each curve is defined by, 'Fable 5.1.
5-85
•
-4
V)
i0-3 10-= 10-1 i
tDA
0 Fig. 5.33: Dimensionless time to

end of MDH straight
line for shapes identi-
fied in Table 5.1. Data
of Cobb and Smith" and
Kumar and Ramey12 (after
Ref. 9).
A comparison of Fig. 5.33 for the MDH plot to Fig.

5.20 for the Horner plot shows that the Horner plot will
yield a straight line for a longer shut-in time when the
producing time is short. For long producing times,
boundaries will affect both plots at the same time.
6. Horner versus Miller-Dyes-Hutchinson

Because of the different mathematical assumptions
involved in the Horner and MDH methods, it is commonly
believed that the Horner method should only be used in new
wells in new reservoirs, and the MDH method applies only to
^ old wells in fully developed reservoirs; this is not true:
The Horner method can be applied to any well regardless of
5-86
the producing time or the stage of reservoir develo p ment.
The MDH method, however, should not be used unless the
well has flowed sufficiently long to be in pseudosteady
s-tate at the time of shut-in. The time required to achieve
pseudosteady state flow can be predicted using Eq. 3.15.
A study by Cobb and Smith" concluded that, in general,
the Horner method is superior to the MDH method because it
is applicable over a greater range of producing times. In
particular, it was shown that for short producing times,
i.e., t p < t pss , that the Horner plot will give a straight
line over a longer interval of shut-in time than the MDH
plot for the same data. This can be very important when
wellbore storage or boundary effects are present; the
straight line on the MDH plot might be masked by these
effects, whereas, on the Horner plot the same data would
yield a straight line.
The MDH plot is probably used more often than the
Horner plot because it is easier to prepare. The author
prefers the Horner plot because of its application to a
wider range of test conditions.
7. Other applications
While equations will not be repeated, it should be
stated that modifications to the MDH method for multiphase
flow are identical to those previously presented for the
Horner method. Also, the criteria for detecting faults,
the effects of wells not centered within their drainage
area, and the procedures for handling variable rates before
shut-in, are the same as previously presented.
Example 5.11: Miller-Dyes-Hutchinson buildup analysis
Problem. The following pressure buildup data were presented •

by Lee27 for an oil well producing above the bubble point.
5-87
At the time of shut-in, the well had been producing for an
effective time, tp, of 13,630 hours.
At, hrs pws, psia pws-pwf(At=0), psi
0.15 3,680 146

0.2 3,723 189
0.3 3,800 266
0.4 3,866 332
0.5 3,920 386
1 4,103 569
2 4,250 716
4 4,320 786
6 4,340 806
7 4,344 810
8 4,350 816
^ 12 4,364 830
16 4,373 839
20 4,379 845
24 4,384 850
30 4,393 859
40 4,398 864
SO 4,402 868
60 4,405 871
72 4,407 873
(At=0) = 3,534 psia

pwf
= 250 STB/D
Qo
= 0.8 cp
Po
0 = 0.139
B = 1.136 RB/STB
ct = 17x10-6psi-1
rw = 0.198 ft
A = 160 acres (square drainage pattern)
".' o = 53 lbm/ft3
A wb
= 0.0218 ft2
h = 69 ft
5-88
•
a) Determine permeability, total skin factor, average
pressure and flow efficiency using the MDH method.
b) Compute the wellbore storage factor from the log-log

plot and compare with the value calculated based upon well
completion data.
c) Compute the time required for storage effects to end,

using Eq. 5.33,and compare to the actual performance of the
pressure data.
Solution. Log-log and semilog plots of the buildup data are

presented in Figs. 5.34 and 5.35, respectively. It appears
from Fig. 5.34 that the unit slope line ends at approximately
t* = 0.2 hours. Using the rule of thumb that storage ends
at 50t*, the semilog straight line should begin at approxi-
mately 10 hours. The semilog straight line selected on Fig.
5.35 is consistent with this result.
103
'--1
+^+
^
UNIT SLOPE LINE
C_
10Z
0.1 1 10
At, hrs
Fig. 5.34: Log-log plot for Example 5.11.
•
5 -89
• • •
4600
4400
4200
•^
^n
u, 4000
3800
3600
0.1 1 10 102
At, hrs
Fig. 5.35: Miller-Dyes-Hutchinson plot for Example 5.11.

a) From the semilog straight line, m= 70 psi/cycle-
•
Permeability is computed using Eq. 5.66:
k = 162.6 m
qBP
k = 162.6 (250) (1.136) (0.8)

(70 ) ( 69 )
k = 7.6 md.
k + 3.23
s = 1.151 - log
LPlhrPwf(At0)
m ûctrw
1.151 [4284-3534 _ log 7.6

s =
70 (0. 139) (0 .8) (17xl0- 6)(0.198 ) 2
+ 3.23] •
s = 6.8.
The average pressure, according to the Dietz method, can

be read directly from the semilog straight line at a shut-in
time given by Eq. 5.74:
ûctA
(^t)pR = 3792 kCA
= 3792 (0.139) (0.8) (17x10 6) (160) (43,560)

(7.6) (30.9)
(At)p = 213 hrs.

R
The pressure on the extrapolated straight line at At = 213

hours, although not shown on Fig. 5.35, is
•
5-91
• PR = 4,449 psia.
The flow efficiency is computed using Eq. 5.20, i.e.,
pR pw f(At=0) - Aps
E
pR pwf(At=0)
where
Ops = 0.87ms.
Therefore,
Aps = 0.87(70) (6.8)
Aps = 414 psi
^ and,
E = 4449 - 3534 - 414
4449 - 3534
E = 0.55.
Thus, based upon the drawdown before shut-in, this well is

only producing at S5o of the rate it could produce if it
were an ideal well with no restrictions.
b) The wellbore storage factor can be computed from the unit

slope line on the log-log plot using Eq. 5.32:
C = aBAt 0 1
24[p ws-pwf(At=O)]
At At = 0.15 hrs. on the unit slope line of Fig. 5.34,

[pws pwf(At=0)] = 146 psi; therefore,
C = (250) (1.136) (0.15)

0 (24) 1
5-92
q
C = 0.012 bbls/psi.
The storage factor can also be computed based upon the well
completion. Since this is an oil well, it is assumed that
the effects of fluid expansion are negligible and that the
dominant storage effect is caused by a rising fluid level.
The storage factor caused by a rising fluid 1, Dvel is given by
Eq. 5.28:
144A
wb
C =
5.615p
C = (144) ( 0.0218)
(5.615 53
Thus, the storage factors computed by independent methods

are in good agreement.
c) According to Eq. 5.33, wellbore storage effects should

end at
0.14s
At = 170,000uCe
kh
(0.14) (6. 8)
(170 , 000) (0. 8) CO. 0 12) [e
( 7.6 ) ( 69 )
At = 8 hrs .
This time is in excellent agreement with the log-log plot and

with the beginning of the semilog straight line on the MDH
plot.
•
5-93
•
IV. EXTENDED MUSKAT ANALYSIS
The oldest method of buildup analysis was proposed in

1937 by Muskat31. This method requires a plot of log
(pR pws) versus L%t. Several studies'^12 have indicated that
this method should be used with caution, and that it should
be applied onZy to Zate-time data. It is not considered a
good method for analyzing buildup data for wells producing
from closed drainage areas because it takes too long for the
required straight line to form. However, it appears12 to
have more application to analyzing pressure data from water-
drive reservoirs and waterfloods which have achieved fillup
because of the longer duration of the required straight line.
^ I. Average reservoir pressure in the drainage volume of the

test well
The most useful application of the Muskat method is in
the determination of pR This method requires that a plot
of log (PR pws) versus At be prepared for several estimates
of pR The value of PR which yields a straight line for
late-time data is the correct average pressure. A schematic
of the required plot is depicted by Fig. 5.36. It is noted
that an assumed value of PR too large will cause the plot to
be concave upward; when PR is too small, the plot will be
concave downward. Practically, it is recommended that you
start with values of PR which are too small, and increase
PR until the straight line is obtained.
Permeability can be computed as'
141.2qBî
k = (5^ . 'S)
pDMint
"s int
hR
5-94
•
103
PR TOO LARGE
^
102
CORRECT pR
(PR-Pws) int
)
PR TOO SMALL
10
At, hrs
Fig. 5.36: Extended Nluskat plot of

buildup data.
•
where (pR pwsint is the intercept of the correct straight
line at At=0. The dimensionless pressure, PDNiint' is
presented as a function of dimensionless producing time,
tDA, in Fig. 5.37 for a well in the center of a closed-
square drainage system', and for a square with a constant
pressure boundary12. If a well in a closed-drainage-
boundary square produces long enough before shut-in to
achieve pseudosteady state, it is observed from Fig. 5.37
that
PDMint (tDA > 0.1) = 0.67, (5. 76)
•
5-95
•
1.4
/ -- 1
1.2
I • I
L___J
CONSTAN T
1.0 PRESSURE
/
0.8
/
/
0.6
•
0.4
NO FLOW
0.2
0
2 4 e 2 4 6 e 2 4 8 e
1 O-s io-e 1O-1 i
• tDA
Fig. 5.37: Muskat dimensionless

intercept pressure for
a well in the center of
a closed or constant-
'pressure-boundary square.
Data of Ramey and Cobb'
and Kumar and Ramey12
(After Ref. 9).
This same condition in the constant-pressure-boundary square

occurs at tDA = 0.25, so that
pDMint (tDA > 0.25) = 1.34. (5. 77)
It has been reported' that for a closed-boundary-circular

system, pDMint(tDA > 0.1) = 0.84. This large difference
between the closed-square system is suprising and is cause
^ for concern. This indicates that the Muskat method is
very sensitive to geometry and, in view of the difficulty
5-96
^
in specifying an accurate geometry for a test well, raises
very serious questions about the accuracy of permeability
calculations from the Muskat method.
3. Selection of correct straight line

Caution must be used when applying the Muskat method
because only the late-time data are valid. It may be
possible to find a value of pR which will cause the early-
time data to form a straight line but it will be the
wrong line and will result in incorrect values of pR and
k. There is also a limit on the late-time data which can
be used but this is generally not a problem. The beginning
or end of the correct straight line can be estimated 7,12
as
ûcta
At 0.0002637k (^tDA)sl
(5. 78)
•
where L^tDA is presented by Fig. 5.38 for a well in the
center of a square. The time to the end of the straight
line is not indicated by this correlation for the
constant -pressure -boundary square; this line will end at
At DA=0 512
•
5-97
• 0.09
0.08
END OF MUSKAT
TRAIGH T LINE
0.07
NO
FLOW
0.06
.--^
4-J START OF MUSKA T

4 0.05 S TRA IGHT LINE
0.04
0.03 CONSTANT ^ ^
PRESSURE
0.02
! • • n L . • • ^ ^ o
10-6 10-3 10-^ 10-1
• Fig. 5.38:
tDA
Time to beginning and

end of Muskat straight
line for a well centered
in a square drainage area
(After Refs. 7 and 12).
Example 5.12 Extended Muskat analysis of pressure

buildup data
Problem. The following pressure buildup data were measured

in an oil well producing from the San Andres formation in .
Yoakum County, Texas.
•
5-98
•
At, hrs pws, psia At, hrs pws, psia
10 579 140 1422

20 802 150 1443
30 950 160 1460
40 1052 170 1478
50 1124 180 1495
60 1180 190 1510
70 1227 200 1524
80 126S 210 1538
90 1300 220 1552
100 1330 230 1566
110 1355 240 1578
120 1380 250 1590
130 1401 260 1601
h = 18 ft
Bo = 1.2 RB/STB
PO = 1.3 cp
qo = 330 STB/D
A 80 acres (wells drilled on square pattern)
tp = 14 days
kair = 5.0 md (core data)
^ = 0.10
_6 _1
ct = 18x10 psi
Estimate the permeability and average pressure within the

drainage area of this well using the Extended Muskat method.
Solution. The following table presents a summary of the data

required to prepare the Muskat plot for assumed values of PR
equal to 1,620, 1,700, and 1,775 psia.
•
5-99
•
PR= 1,620 PR= 1,700 PR= 1,775
At, hrs pws, psia pR pws pR pws pR pws
10 579 1041 1121 1196

20 802 818 898 973
30 950 670 750 825
40 1052 S68 648 723
50 1124 496 S76 651
60 1180 440 520 595
70 1227 393 473 548
80 1265 355 435 510
90 1300 320 400 475
100 1330 290 370 445
110 1355 265 345 420
120 1380 240 320 395
130 1401 219 299 374
140 1422 198 278 353
150 1443 177 257 332
160 1460 160 240 315
170 1478 142 222 297
180 1495 125 205 280
190 1510 110 190 265
200 1524 96 176 251
210 1538 82 162 237
220 1552 68 148 223
230 1566 54 134 209
240 1578 42 122 197
250 1590 30 110 185
260 1601. 19 99 174
These data are presented in Fig. 5.39.

The value of PR which causes the late-time data to
plot as a straight line is
^ PR = 1,775 psia
5-100
103
•
102
q
10
0 40 80 120 160 200 240 280
At, hrs
Fig. 5.39: Extended Muskat plot for Ex. 5.12.
Therefore, this is considered to be the correct value of

average pressure. While we show curves for only three
assumed values of PR on Fig. 5.39, other values were also
plotted to arrive at this conclusion.
Extrapolation of the straight line to At=0 yields
800 psi. The dimensionless producing time
(pR pws)int
based on the value of k from core data is:
•
S-101
• tDA
0.0002637kt
^Pc
_ (0.0002637) (5.0) (14) (24)

tDA (0.10) (1.3) (18x10-6) (80) (43,560)
tDA = 0.054.
Corresponding to this value, we find from Fig. 5.37 that

pDMint - 0.6. Therefore, from Eq. 5.75,
k = 141.2dBu
pDiNtint
Yl (pR pivs) int
k = (141.2) (330) (1.2) (1.3) (0.6)

(18) ( 800 )
k = 3.0 md.
Since this value of k does not agree with that used to

determine pDMint' it is necessary to iterate to a correct
solution. If k = 2.3 md is assumed, tDA = 0.025 and pDMint
0.46; then,
k = (141.2)(330)(1.2)(1.3) (0.46)
(18) 800
k = 2.3 md.
Since this agrees with the assumed value of k, it is concluded

that this is the correct answer.
While not presented here, the Horner plot gives k = 4.6
md and, using the Matthews-Brons-Hazebroek correlations, PR =
1757 psia. The values of average pressure from these two
methods are in good agreement; however, the Muskat permeability
is significantly less than values obtained from core data
and from the Horner method.
5-102
Finally, we want to verify that we used the correct
portion of the data to make this analysis. Using tDA = 0.025,
Fig. 5.38 indicates that the correct straight line should
begin at (AtDA)esl = 0.044; therefore, from Eq. 5.78,
ûctA
At 0.0002637k (AtDA)esl
At = (0.10) (1.3) (18x10 6) (80) (43,560) (0.044)

0.0002 37 ( 2.3 )
At = 592 hrs.
Since this test was only run for 260 hours, this clearly
indicates that even though the Muskat plot appears to have
a straight line, this line is incorrect. This also explains
why the calculated permeability is not in agreement with
either the Horner analysis or the core data. The fact that
PR agrees with the Horner method is probably accidental.
We conclude, therefore, that this data should not be
analyzed using the Muskat method.
0 V. AGARWAL METHOD
A novel method of analyzing pressure transient data

was introduced by Agarwa132 in 1980. L The primary advantage
of the Agarwal method is that it allows drawdown type curves
to be used in the analysis of pressure buildup data when
producing time is short;j this application will be discussed
in a subsequent section. However, this method can also be
used to perform conventional semilog analysis of buildup
data. While the method was developed for liquid, radial
flow, it appears32 that the method also works well for gas
5-103
• wells and for fractured reservoirs.
The basis of the Agarwal method is a new time function,
referred to as equivaZent time, defined as
t At
Ate = t p At (5.79)
- p
This function appears similar to the Horner time ratio,

but is much more general in its applications. It is
noted that when tp is large, Ate becomes equal to At
and this method reverts back to the MDH method.
When a well is shut in, pressures during the infinite-
acting period can be expressed in terms of equivalent time
as
pws = pwf(At=0) + 162.6 ^ [log Ate

.
^
+ log k - 3.23 + 0.87s^. (5.80)
2 J
ûctw
Equation 5.80 suggests that a semilog plot of pws versus

Ate will yield a straight line with slope m defined as
m = 162.6 a. (5.81)
This plot is illustrated by Fig. 5.40.
.
After the correct semilog straight line has been
identified, permeability can be computed from Eq. 5.81 as
k = 162.6 mhu . (5.82)
Notice that the slope, m, of the Agarwal plot is positive
5-104
•
P
SLOPE = m
.'.,
V)
• I
•
• •"'``-tVELLBORE tp
• STORAGE
0.1 1 10 102
A te, hrs
Fig. 5.40: Schematic of Agarwal semilog

plot.
•
and must be used as a positive number in all calculations.
2. Total skin factor
The skin factor can be computed as
[Pws(Ate 1hr) - p (At0)

wf - log k
s= 1.151
m Ûctrw 2
+ 3.23]. (5. 83)
The value of pws(Ote=lhr) is obtained from the semilog

straight line, or its extension, at Ate=lhour.
♦
5-105
3. Average pressure in the drainage pore volume of the
test well
The initial reservoir pressure, pi, or the false
pressure, p*, can be read directly from the semilog straight
line where Ate is equal to the producing time tp. This is
equivalent to reading the pressure from a Horner plot at a
Horner time ratio equal to unity. After p* is determined,
it can be used to determine PR using correlations previously
discussed for the Horner method.
4. Wellbore storage
Wellbore storage affects the Agarwal plot in the same
manner as it affects the Horner and MDH plots. The log-log
plot is again used to determine the length of wellbore
storage effects; in this case, however, [Pws pwf(At=0)] is
plotted versus At e rather than At. This plot is illustrated
^ by Fig. 5.41. If the unit slope line ends at At, the semi-
log straight line on the Agarwal plot should begin at
approximately 50Ate. Equations for predicting the duration
of storage (Eq. 5.33), and for computing the storage factor
(Eq. 5.32) are also applied as before by substituting Ate
for At. Use of equivalent time in the log-log plot accounts
for producing time in the solution; the primary advantage of
this plot will be illustrated in the next section which
discusses type curve analysis.
S-106
:i
•
10
UNIT SLOPE •
LINE •
r-, •
10 2
•
END OF COMPLETE
CONTROL BY
WELLBORE STORAGE
10
Ate
1
0.1 1 10 102 103
Ate, hrs
Fig. 5.41: Log-log plot for Agarwal method.

•
Example 5.13: Agarwal buildup analysis
Problem. A well producing above the bubble point pressure

has produced 14,138 STB of oil at a stabilized rate of
1,214 STB/D. The well was shut-in for a buildup test and
the pressure data presented in the following table were
recorded. Determine k, s and p* using the Agarwal method
of analysis.
•
5-107
•
t at
At, hrs Ate=tP+ot , hrs Pws, psia pws-pwf(At=0), psi
P
0.05 0.05 3291 162

0.1 0.10 3454 325
0.2 0.20 3689 560
0.3 0.30 3841 712
0.5 0.50 4017 888
1 1.00 4229 1100
2 1.99 4432 1303
3 2.97 4530 1401
4 3.94 4584 1455
S 4.91 4621 1492
6 5.87 4628 1499
8 7.78 4640 1511
^ 10 9.65 4649 1520
12 11.5 4657 1528
14 13.3 4664 153S
16 15.1 4674 1545
20 18.7 4680 1551
24 22.1 .4685 1556
pwf(At=0) = 3,129 psia

qo = 1,214 STB/D
ct = 18x10-6psi-1
rw = 0.5 ft
Bo = 1.119 RB/STB
h = 2S ft
= 17%
Po = 0.6 cp
Solution. The producing time, tp, is:
(14 ,138 STB) (24^'^rs)

= 279.5 hrs.
tp 1,214 STB/D D
5-108
This time was used with shut-in time to compute the equivalent
•
times shown in the table of input data. Log-log and semilog
plots of the buildup data are presented in Figs. 5.42 and 5.43,
respectively.
It is observed from Fig. 5.42 that a unit slope line
lasts until Ate ~ 0.1 hours. Applying the 50Ate rule of
thumb, the semilog straight line should begin at a shut-in
time of approximately 5 hours. This is consistent with the
semilog straight line shown on Fig. 5.43.
The slope of the semilog straight line is m = 100 psi/cycle.
Permeability is computed using Eq. 5.82:
k = 162.6 mh
k = 162.6 (1214) (1.119) (0.6)

(100) 25
k = 53 md.
The total skin factor can be estimated using

Eq. 5.83, i.e.,
pws (Ate=lhr) - Pwf (At=0)

s = 1.151
m
- log k + 3.231 .
z
ûctrw
The value of pws(Ate=lhr) is found from an extrapolation of '

the semilog straight line at Ate=1 to be 4,552 psia. Therefore,
= 1.151 4552-3129 - log 53

s
100
(0 .17) (0 .6) (18x106 ) (0 .5) 2
+ 3.23
1 0
s = 10.8.
5-109
• q E
104
IT SLOPE
H LINE
V)
r-,
i-^ 103
LH
tn
V)
F--+
i--^
102
10-2 10-1 10
Ate, hrs
Fig. 5.42: Log-log plot of buildup data in Ex. 5.13.
9
4800
plhr = 4,552 PSIA
4600
m = 100 PSI/CYCLE
4400
.^,
^
4200
400C
380C
10 1 10 102
Ate, hrs
Fig. 5.43: Semilog plot of buildup data in Ex. 5.13.
9 9 0
•
The false pressure, p*, is read directly from the
semilog straight line at Ate = tp. A lthough not shown
on Fig. 5.43, the value of pressure at Ate = 279.5 hours
from an extrapolation of the straight line is
p* = 4,798 psia.
This pressure could now be used in the conventional manner

to determine average reservoir pressure.
VI. TYPE CURVE ANALYSIS
The use of type curves to analyze pressure data

^ was introduced in Chapter 2, and applications to pressure
drawdown tests were detailed in Chapter 3. Most type curves
which have been developed and published to date are based on
drawdown solutions and, accordingly, are best suited for
analyzing drawdown data. However, these curves can also be
applied to buildup test analysis provided the producing time,
t p , prior to shut-in is sufficiently long compared to shut-in
time, At, that log (t p + At) = log t p . For cases where the
producing time is short and this criterion is not satisfied,
the use of drawdown type curves to analyze buildup data can
result in significant errors. Short producing times can
occur in any test situation, but are most common in drill
stem tests and in new wells where the absence of storage or
pipeline facilities limit the amount of fluid which can be
produced.
1. Ramey type curve

^ The Ramey type curve, previously presented as Fig. 3.24,
is a graphical presentation of the solution to the pressure
5-112
drawdown equation for the transient fZow of a singZe-phase
•
Ziquid in a radiaZ flow system. This curve, which accounts
for the effects of both skin and wellbore storage, is pre-
sented as a plot of dimensionless pressure, pD, versus
dimensionless time, tD, where PD and tD were previously
defined for drawdown tests by Eqs. 3.57 and 3.58, respectively.
In order to apply the Ramey curve to the analysis of
buildup tests, it is necessary to redefine PD and tD as
p _ 0.00708kh [p -p (At=O)] (5. 84)

D qBU ws wf
t = 0.0002637k at (5. 85)

D 2
ûcrw
t
The procedure for type curve analysis, which requires that

test data be plotted on log-log paper and matched with the
Ramey curve, is the same as presented for drawdown tests.
The form of the log-log plot required for curve matching
will vary, however, depending upon the length of the
producing time, tp, prior to shut-in.
A. Long producing times.

If t p is sufficiently long compared to At that log
(tp+At) = log tp, the required log-log plot is made in the
conventional manner, i.e., as [pws pwf(At=0)] versus At.
B. Short producing time

When producing time is short, use of the Ramey curve in
the conventional manner will result in significant errors.
• In this situation, the log-log plot should be prepared as
[pws pwf(At=0)] versus equivalent time, Ate. Equivalent time
was defined previously by Eq. 5.79. Agarwa132 has shown that
use of Ate, rather than At, eliminates the errors associated
with short producing time. Effectively, use of equivalent
time makes the buildup solution equivalent to the drawdown
solution, thereby permitting use of drawdown type curves to
5-113
• analyze buildup data, regardless of the producing time.
Although other methods have been proposed 33-35 which account
for short producing time in type curve matching, the Agarwal
method seems to be the easiest and most effective method
available.
2. McKinley type curve

The McKinley type curve, previously presented as Fig.
3.35, was originally developed for pressure buildup tests.
However, the model used to develop the curves assumed that
producing time was much longer than the longest shut-in
time3z-33 Therefore, when producing time is long, the
McKinley method can be applied using the same procedure
outlined in Chapter 3 for drawdown testing. When producing
time is short, this method is subject to significant errors
of analysis and should be used with caution.
• 3. Other type curves

Other type curves are available which can be used to
analyze buildup data. The curves of Earlougher and Kersch9,36
and Gringarten37 have been particularly useful in the analysis
of non-fractured wells. Type curves for fractured wells will
be discussed in Chapter 6.
4. Type curves versus conventional analysis

Type curves offer an alternative method of analysis to
conventional semilog plots. In general, however, type curve
matching procedures for non-fractured reservoirs are not as
accurate as conventional methods of analysis. Type curves
should not be used for the final analysis of data if the
correct semilog straight line can be identified; instead,
type curves should be used as an aid in identifying the
correct semilog straight line. Final analysis of the data
using type curves should only become necessary if the correct
0
5-114
straight line is masked by wellbore storage or skin effects.
•
VII. OTHER PROBLEMS OF INTERPRETATION
The conventional methods of analysis discussed in this

chapter are based on the premise that the correct semilog
straight line can be identified; this is seldom an easy task:
The theory which predicts that transient flow data will plot
as a semilog straight line is based on the ideal reservoir
model described in Chapter 2. Any deviation from this ideal
model can cause pressure to behave differently than predicted
by theory. The effects of some non-idealities, such as
wellbore storage, skin, faults, and assymetrical drainage
configurations, have already been discussed. The objective
of this section is to present a brief discussion of other
commonly encountered problems which make identification of
the correct semilog straight line a difficult task. The
reader is referred to the list of references for detailed
analysis of these problems.
1. Wellbore phase separation

The problem of wellbore phase separation occurs in a
well where g_as and liquid are flowing simultaneously-in-.the
wellbore at the time of shut-in. It has been shown 38,39
that when a well of this type is shut in at the surface,
gravir„ forces cause the liquid to fall and the gas to rise
within the wellbore. Since the liquid has a small compressi-
bility, and the gas is unable to expand within the fixed
volume wellbore, this redistribution of fluids causes an
abnormal increase in pressure within the wellbore. At early
shut-in times, it is possible for the pressure in the wellbore
to exceed the pressure in the formation; in this situation,
fluids are forced back into the formation resulting in a
decrease in bottomhole pressure. This humping effect is
S-115
• illustrated by Fig. 5.44. Given enough time, the formation
and wellbore will achieve an equilibrium condition and
pressure will again increase in the wellbore.
•
• • • •• •
Ln • •
• •• • •
•
•
G:. •
•
•
•
•
• 1 10 102 103 104
t {'At
p
At
Fig. 5.44: Horner plot showing the effect of

phase separation.
The effect of phase separation, like wellbore effects

in any form, is to mask the early-time transient flow data.
Given enough time, the effect will end and measured for-
mation face pressures will again reflect true reservoir
• 1
behavior. Unfortunately, in most wells where this effect
is evident, it lasts so long that test times required to
obtain the semilog straight line are impractical, or
boundaries begin to affect the data before the straight
line can be observed. Even when the straight line does
• appear, identification of the correct line is made difficult

by the abnormal behavior.
5-116
Type curves have been developed 40 which model the
•
pressure behavior caused by phase separation. These
curves offer the potential for analyzing data affected
by phase separation, but are limited by the assumption
that no additional gas enters the wellbore after shut-in.
When the problem of phase separation is present,
one of the following approaches should be used:
1) L,âve the wPll shut in long enough for the

effect to end. The only problem with this
approach is that test times may be impractical,
or transient flow may end before wellbore
effects become insignificant.
2) Use a bottomhole shut-off tool. -}r
3) Run a different type of test. This problem

does not occur in flowing tests.
4) Attempt to analyze the data affected by phase

separation using available type curves40.
•
2. Interference from adjacent wells
This can occur when other wells are producing from the
same reservoir while the well is being tested. If the test
well is shut in long enough, the bottomhole pressure will
begin to decrease due to drainage caused by adjacent wells.
The Horner plot for a well subject to this effect will appear
as shown in Fig. 5.45. The straight line can be used in the
conventional manner to compute reservoir properties. This
plot may sometimes appear similar to data being affected by
phase separation, but the effect generally occurs at later
times in the case of interference. Whereas pressures will
eventually build up again in a well affected by phase
separation, pressures in this case will continue to decline
as long as adjacent wells are producing. If the test well
happens to be adjacent to an injection well, pressure at
late times will increase rather than decrease.
•
5-117
•
SLOPE = m
• 00
•
•
• \
• \\
•
•
0 • •
1 10 102 10 3 104
t '}At
p
At
• Fig. 5.45: Horner plot showing effect of

interference.
3. Stratification
The theory used to analyze pressure buildup data
assumes that production is from a single homogeneous
reservoir. The_pressl^re hehavi ^r nf a reservoir w; th w^
or more distinct layers with communication only at the
wellbore will behave quite differently from a single layer.
Multi-layered systems can exhibit abnormal behavior and in
many cases are difficult to analyze. Significant studies
have been made only for two-layered systems.
A theoretical 1Iorner plot for a two-layered reservoir
with no crossflow is shown in Fig. 5.46. Early-time data
will yield the correct semilog straight line with a slope
that is proportional to the average flow capacity of the
two layers. This line may not last long, however, and can
easily be masked by wellbore storage and skin effects. The
straight line is followed by a transition period caused by
changing fractional flow rates from each layer; the data at
S-118
•
5 ,4 3 ? 1 1. Storage and skin
?. Proper straight
• line
•• 3. Transition
• 4. Flattening
• • • port ion
•
5. Final rise
1 •
SLOPE = m •
•
•
•
• •
•
1 10 102 103 104 105
+At
t^
At
Fig. 5.46: Horner plot for two-layered system

with communication only at the wellbore.
•
this time appear to be stabilizing toward the average
reservoir pressure. This is followed, theoretically, by
a flattening of the data, and then a final rise before
pressure stabilizes at the average pressure. A number of
studies`'1-`'s of layered systems have shown that many
variations of the behavior depicted by Fig. 5.46 can occur,
depending upon the contrast in properties between the
layers, the geometry of the system, and the length of the
producing time. In particular, several studies`'2-"s show
that the flattening of the data may be absent or insignifi-
cant such that the final rise occurs immediately after the
straight line ends. This behavior is common to many other
types of heterogeneities and can easily be misinterpreted.
When a reservoir is layered but has crossflow between
the layers, it is found46 that pressure behavior will be
the same as in a homogeneous system. •
5-119
q
4. Com osite reservoirs
A composite reservoir is one in which rock or fluid
properties vary in a stepwise fashii as you move radiallZ
away from the wellbore. Composite systems can occur
naturally but most are induced as a result of damage,
stimulation, or some type of injection project. Since the
hydraulic diffusivity, defined by Eq. 2.1, controls the rate
of pressure change in a formation, it is obvious that any
change in k, p, ^, or ct encountered by a pressure transient
through the reservoir, will alter the rate of pressure change.
Many studies4'-52 have been conducted to consider
pressure behavior in various types of composite systems. The
serious pressure analyst should review these studies in order
to gain a qualitative understanding of how various changes in
the diffusivity can affect the shape of a buildup plot. For
example, a stepwise change in k/U away from the wellbore - a
common problem in injection projects such as waterflooding,
will cause a change in slope on the semilog plot, as shown by
Fig. 5.47. The nature of the slope change will depend upon
the manner in which k/p changes. In this case, the slope of
the first straight line can be used to compute k/p of the
zone adjacent to the test well. However, because of the
complex interrelationship of k, p, and ct in the two zones,
the slope of the second line will yield k/p of the zone away
from the wellbore only under very restrictive conditions.
This problem will be considered in detail in Chapter 8 which
discusses injection well testing.
5. Fractured reservoirs
Serious errors can he made in the analysis of buildup
data from fractured reservoirs if methods developed for non-
fractured systems are indiscriminately applied. All of the
methods discussed thus far assume radial flow. When a for-
mation is fractured, the flow geometry can be drastically
altered. For example, a hydraulically fractured well will
5-120
•
m2
CHANGE IN SLOPE m
DUE TO MOBILITY •
INCREASE
•
•
••
1 10 102 103 10"
t +at
p
Ot
Fig. 5.47: Schematic of Horner plot for

increase in mobility away from
the wellbore. •
exhibit linear, rather than radial, flow during early build-
up times, followed by elliptical flow, and finally at late
times, radial flow may occur. The Horner plot in this
situation would have the typical appearance depicted by
Fig. 5.48. If it is not recognized that a well is fractured,
this behavior could be misinterpreted for a number of common
heterogeneities such as a stratified reservoir, lateral change
in properties, fault, etc. A detailed analysis of fractured
systems will be presented in the next chapter.
In summary, many different types of heterogeneities,
boundaries and wellbore problems exist which can cause
pressure-time data to deviate from the behavior predicted by
ideal models. Unfortunately, many different reservoir
situations can cause pressure data to exhibit the similar
characteristics. Further, it is also possible for a par-
ticular reservoir heterogeneity to exhibit different pressure
behavior with a change in test conditions. This means that
•
5-121
•
•
•
•
•
.^, •
^ •
•
•
^ •
•
•
•
•
• •
• • • •
l04
1 10 102 103
t +Qt
P
At
• Fig. 5.48: Schematic of Horner plot for

a hydraulically fractured well.
when we observe a particular pressure behavior, there are

generally several different interpretations which can be
made. Unless an engineer has a good understanding of
pressure transient behavior, and utilizes all available
information related to the well completion, reservoir
properties, and reservoir history, a misinterpretation of
the data can easily be made. Again, it is emphasized that
all available information about a well and the formation
being tested must be considered when analyzing a pressure
transient test if an accurate and unique interpretation is
to be made:
•
5-122
•
VIII. TEST DESIGN
The first, and perhaps the most important, criterion

required for accurate well test analysis is proper test
design. Since test design will vary depending upon the
information desired from the test, it is important to clearly
specify test objectives before designing the test. Factors
which should be considered when planning a test are:
1) Record all pressures downhole; surface pressures

are not adequate for most calculations. When a
pressure gauge cannot be run in a well, the next
best alternative is to monitor fluid levels at
short intervals of time using an accoustical
device. With knowledge of fluid gradients in
the wellbore, fluid levels can be converted to
bottomhole pressure.
2) Select a gauge which has the accuracy, sensitivity,

pressure range, and temperature range necessary
to achieve test objectives.
3) Measure the flowing pressure, pwf(At=0) ,

immediately before shut-in.
4) Establish a constant rate prior to shut-in.

The length of the constant rate period will
vary with test objectives and with the rate
history preceding the test. Criteria for
establishing the length of the flow period
were presented for various test situations.
5) Record early-time data at short intervals of

time; this data is required for the detection
and analysis of wellbore storage effects, and
for type curve analysis.
6) For permeability, skin, and average pressure

determination using conventional methods, it
is required that a well be shut in long enough
for wellbore storage and skin effects to
become negligible. The time required for
storage effects to end can be approximated
using Eq. 5.33:
0.14s
At > 170 ,OOOCue .
kh
5-123
q
For purposes of design, setting s=0 in
this equation will yield a minimum value
of At. If storage is predicted to last
so long that the length of the test
becomes impractical, or there is the
danger of boundaries affecting the data
before storage ends, a bottomhole shut-
off tool should be considered.
7) The time at which boundaries cause the

semilog straight line to end varies
with producing time,t ,, and the drainage
configuration. This tpime can be predicted
for a large variety of drainage configu-
rations using Eq. 5.34:
ûcta
Atesl 0.0002637k ^^tDa)esl'
The dimensionless time, (AtDA)esl can be

obtained from Figs. 5.20 and 5.33. Knowing
when to expect boundary effects is especially
important when wellbore storage effects last
is a long time.
8) When the required length of a test is

uncertain, or when test results are of
sufficient importance to justify the
additional expense, use a pressure gauge
with a surface recorder. With this
equipment, pressures can be analyzed as
they are measured; consequently, it can
be assured that test objectives have
been satisfied before the test is
terminated.
A good discussion of test design considerations is presented

by Earlougher9.
•
5-124
REFERENCES
•
1. Matthews, C. S., and Russell, D. G.: Pressure Buildup
Petroleum Engineers of AIME, Dallas (1967) 1.
2. Theis, Charles V.: "The Relationship Between the

Lowering of the Piezometric Surface and the Rate and
Duration of Discharge of a Well Using Ground Water
Storage," Trans., AGU (1935) 519.
3. Horner, D. R.: "Pressure Build-up in Wells," Proc.,

Third World Petroleum Congress, The Hague (1951 Sec.
II, 503. Also Reprint Series, No. 9-- Pressure
Methods, Society of Petroleum Engineers of AIME,
Dallas (1967) 25.
4. Matthews, C. S., Brons, F., and Hazebroek, P.: "A

Method for Determination of Average Pressure in a
Bounded Reservoir," Trans., AIME (1954) 201, 182.
5. Pinson, A. E., Jr.: "Concerning the Value of

Producing Time Used in Average Pressure Determinations
From Pressure Buildup Analysis," J. Pet. Tech. ( Nov., ^
197/2) 1369.
6. Odeh, A. S., and Al-Hu ssainy, R.: "A Method for

Determining the Static Pressure of a Well from
Buildup Data," Trans., AIME (197/1) 251, 621.
7. Ramey, H. J., Jr., and Cobb, W. M.: "A General Buildup

Theory for a Well in a Closed Drainage Area," Trans.,
AIME (1971) 251, 1493.
8. Kazemi, H.: "Determining Average Reservoir Pressure

From Pressure Buildup Tests," Trans., AIME (1974) 257,
55.
9. Earlougher, R. C., Jr.: Advances in Well Test Anal sis,

Society of Petroleum Engineers of AIME, Dallas 1977) S.
10. Dake, L. P.: Fundamentals of Reservoir Engineering,

• Elsevier Scientific u lishing Co., Amster am 1978).
11. Cobb, W. M., and Smith, J. T.: "An Investigation of

Pressure Buildup Tests in Bounded Reservoirs," Trans.,
AIME (1975) 259, 991.
12. Kumar, A., and Ramey, H. J., Jr.: "Well Test Analysis
for a Well in a Constant-Pressure Square," Soc. Pet.
Eng. J. ( April, 1974) 107.
0
5-125
• 13. Gray, K. E.: "Approximating Well-to-Fault Distance
From Pressure Buildup Tests," J. Pet. Tech. (July, 1965)
761.
14. van Poollen, H. K.: "Drawdown Curves Give Angle Between

Intersecting Faults," Oil and Gas Journal ( Dec. 20, 1965)
71.
15. Prasad, R. K.: "Pressure Transient Analysis in the

Presence of Two Intersecting Boundaries," 'Frans., AIME
(1975) 259, 89.
16. Tiab, D. and Kumar, A.: "Detection and Location of Two

Parallel Sealing Faults Around a Well," paper SPE 6056
presented at the SPE Slst Annual Fall Technical Conference
and Exhibition, New Orleans, Oct. 3-6, 1976.
17. Tiab, D. and Crichlow, H. B.: "Pressure Analysis of

Multiple-Sealing Fault Systems and Bounded Reservoirs
by Type-Curve Matching," Soc. Pet. Eng. J. (Dec., 1979)
378.
18. Streltsova, T. D. and McKinley, R.M.: "Effect of

Flow Time Duration on Buildup Pattern for Reservoirs
with Heterogeneous Properties," paper SPE 11140
presented at the SPE 57th Annual Fall Technical
0 Conference and Exhibition, New Orleans, Sept. 26-29,
1982.
19. Ramey, H. J., Jr., and Earlougher, R. C., Jr.: "A

Note on Pressure Buildup Curves," J. Pet. Tech.
(Feb., 1968) 119-120.
20. Earlougher, R. C., Jr., Ramey, H. J., Jr., and Mueller,

T. D.: "Pressure Distributions in Rectangular Reservoirs,"
Trans., AIME (1968) 243, 199.
21. Denson, A. H., Smith, J. T., and Cobb, W. M.: "Determining

Well Drainage Pore Volume and Porosity from Pressure
Buildup Tests," Trans., AIME (1976) 261, 209.
22. Matthews, C. S. and Lefkovits, H. C.: "Studies on

Pressure Distribution in Bounded Reservoirs at Steady
State," Trans., AIME (1955) 204, 182.
23. Perrine, R. L.: "Analysis of Pressure Buildup Curves,"

Drill. and Prod. Prac., API (1956) 482-509.
24. Martin, J. C.: "Simplified Equations of Flow in Gas

Drive Reservoirs and the Theoretical Foundation of
Multiphase Pressure Buildup Analysis," Trans., AIME
(1959) 216, 309-311.
5-126
25. Tipple, D. B. and van Poollen, H. K.: "Effect of
•
Reservoir Discontinuities Upon Buildup Behavior
Following Short Flow Times," paper SPE 6756 presented
at the SPE 52nd Annual Fall Technical Conference and
Exhibition, Denver, Oct. 9-12, 1977.
26. Earlougher, R. C., Jr.: "Practicalities of Detecting

Faults from Buildup Testing," J. Pet. Tech. (Jan.,
1980) 18-20
27. Lee, W. J.: Well Testin , Textbook Series, Society

of Petroleum Engineers of AIME, Dallas (1982) 1, 18.
28. Odeh, A. S. and Selig, F.: "Pressure Build-Up

Analysis, Variable Rate Case," Trans., AIME (1963)
228, 790.
29. Miller, C. C., Dyes, A. B., and Hutchinson, C. A., Jr.:

"The Estimation of Permeability and Reservoir Pressure
from Bottom-Hole Pressure Build-up Characteristics,"
Trans., AIME (1950) 189, 91.
30. Dietz, D. N.: "Determination of Average Reservoir

Pressure from Build-Up Surveys," Trans., AIME (1965)
234, 955.
31. Muskat, M.: "Use of Data on the Build-Up of Bottom-

hole Pressures," Trans., AIME (1937) 123, 44-48.
32. Agarwal, R. G.: "A New Method to Account for Producing

Time Effects When Drawdown Type Curves Are Used to
Analyze Pressure Buildup and Other Test Data," paper
SPE 9289 presented at the SPE 55th Annual Fall Technical
Conference and Exhibition in Dallas, Sept. 21-24, 1980.
33. Raghavan, R.: "The Effect of Producing Time on Type

Curve Analysis," J. Pet. Tech. (June, 1980) 1053-1064.
34. Raghavan, R., Reynolds, A. C., Jr., and Meng, H. Z.:

"Analysis of Pressure Buildup Data Following a Short
Flow Period," J. Pet. Tech. (April, 1982).
35. Soliman, M. Y.: "Analysis of Buildup Tests with Short

Producing Times," paper SPE 11083 presented at the SPE
57th Annual Fall Technical Conference and Exhibition
in New Orleans, Sept. 26-29, 1983.
36. Earlougher, R. C., Jr., and Kersch, K. M.: "Analysis of

Short-Time Transient Test Data by Type-Curve Matching,"
J. Pet. Tech. (July, 1974) 793-800.
37. Gringarten, A. C., Bourdet, D. P., Landel, P. A., and

Kniazeff, V. J.: "A Comparison Between Different Skin
and 1dellbore Storage Type-Curves to Early-Time Transient
Analysis," paper SPE 8205 presented at the SPE 54th Annual
Fall Meeting, Las Vegas, Sept. 23-26, 1979.
5-127
• 38. Stegemeier, G. L. and Matthews, C. S.: "A Study of
Anomolous Pressure Build-Up Behavior," Trans., AIME
(1958) 213, 44.
39. Pitzer, S. C., Rice, J. D. and Thomas, C. E.: "A

Comparison of Theoretical Pressure Build-Up Curves
with Field Curves Obtained from Bottom-Hole Shut-in
Tests," Trans., AIME (1959) 216, 416.
40. Fair, W. B., Jr.: "Pressure Buildup Analysis with

Wellbore Phase Redistribution," paper SPE 8206
presented at the SPE 54th Annual Fall Technical
Meeting, Las Vegas, Sept. 23-26, 1979.
41. Lefkovits, H. C., Hazebroek, P., Allen, E. E., and

Matthews, C. S.: A Study of the Behavior of Bounded
Reservoirs Composed of Stratified Layers," Soc. Pet.
Eng. J. (March, 1961) 43-58.
42. Kazemi, H.: "Pressure Buildup in Reservoir Limit

Testing of Stratified Systems," Trans., AIME (1970)
249, 503.
43. Cobb, W. M., Ramey, H. J., Jr., and Miller, F. G.:

"Well-Test Analysis for Wells Producing Commingled
40 -ones," Trans., AIME (1972) 253, 27.
44. Raghavan, R., Topaloglu, H. N., Cobb, W. M., and Ramey,

H. J., Jr.: "Well-Test Analysis for Wells Producing
From Two Commingled Zones of Unequal Thickness," Trans.,
AIME (1974) 257, 1035.
4S. Earlougher, R. C., Jr., Kersch, K. IN4., and Kunzman,

W. J.: "Some Characteristics of Pressure Buildup
Behavior in Bounded Multiple-Layer Reservoirs Without
Crossflow," Trans., AIME (1974) 257, 1178.
46. Russell, D. G. and Prats, M.: "The Practical Aspects

of Interlayer Crossflow," J. Pet. Tech. (June, 1962)
589-594.
47. Loucks, T. L. and Guerrero, E. T.: "Pressure Drop in

a Composite Reservoir," Trans., AIME (1961) 222, 170. .
48. Carter, R. D.: "Pressure Behavior of a Limited Circular

Composite Reservoir," Trans., AIME (1966) 237,328.
49. Odeh, A. S.: "Flow Test Analysis for a Well With Radial
Discontinutiy," Trans., AIME (1969) 246, 207.
50. Bixel, H. C. and van Poollen, H. K.: "Pressure Drawdown

and Buildup in the Presence of Radial Discontinuities,"
Trans., .aIMP (1967) 240, 188.
5-128
5l. Kazemi, H., Merrill, L. S., and Jargon, J. R.:
•
"Problems in Interpretation of Pressure Fall-Off
Tests in Reservoirs With and Without Fluid Banks,"
J. Pet. Tech. (Sept., 1972) 1147-1156.
52. Merrill, L. S., Kazemi, H., and Gogarty, W. B.:

Pressure Fall-Off Analysis in Reservoirs With Fluid
Banks," Trans., AIME (1974) 257, 809.
•
5-129
• REFERENCES: Wellbore Phase Redistribution
1. Stegemeir, G.L. and Matthews, C.S.: "Study of Anomalous

Pressure Build-Up Behavior, " AIME (1958) 213, 44.
2. Pitzer, S.C., Rice, J.D., and Thomas, C.E.: "A Comparison of

Theoretical Pressure Build-Up Curves With Field Curves Obtained
From Bottom-Hole Shut-In Tests," Trans., AIME (1959) 216, 416.
3. Fair, W.B. Jr.: "Pressure Build-Up Analysis With Wellbore Phase

Redistribution," SPEJ (April 1981) 257.
4. Thompson, L.G., Jones, J.R., and Reynolds, A.C.: "Analysis of

Pressure Buildup Data Influenced by Wellbore Phase
Redistribution," SPEFE (October 1986) 435.
^ 5. Rushing, J.A. and Lee, W.J.: "Use of an Automatic History-

Matching Technique To Analyze Pressure Buildup Data Affected by
Wellbore Phase Segregation: Case Histories," paper SPE 18837
presented at the 1989 SPE Production Operations Symposium,
Oklahoma City, Oklahoma, March 13-14.
6. Fair, W.B. Jr.: "Generalization of Weilbore Effects in Pressure

Transient Analysis," paper SPE 24715 presented at the 1992 SPE
Annual Technical Conference and Exhibition, Washington, DC,
October 4-7.
7. Hasan, A.R. and Kabir, C.S.: "A Mechanistic Approach to

Understanding Wellbore Phase Redistribution," paper SPE 26483
presented at the 1993 SPE Annual Technical Conference and
Exhibition, Houston, Texas, October 3-6.
8. Hegeman, P.S., Hallford, D.L., and Joseph, J.A.: "Well-Test

Analysis With Changing Wellbore Storage," SPEFE (September
• 1993) 201.
9. Raghavan, R.: Well Test Analysis, Prentice Hall, Englewood Cliffs,
New Jersey ( 1993) 314-323. 0
10. Hasan, A.R. and Kabir, C.S.: "Modeling Changing Storage During
a Shut-In Test," SPEFE (December 1994) 279.
0
• REFERENCES: Well Testing in Layered Reservoirs
Larsen, L.: "Wells Producing Commingled Zones With Unequal Initial

Pressures and Reservoir Properties," paper SPE 10325 presented at the
1981 Annual SPE Technical Conference and Exhibition, San Antonio,
Texas, Oct. 5-7.
2. Larsen, L.: "Determination of Skin Factors and Flow Capacities of

Individual Layers in Two-Layered Reservoirs," paper SPE 11138
presented at the 1982 Annual SPE Technical Conference and Exhibition,
New Orleans, Louisiana, Sept. 26-29.
3. Prijambodo, R., Raghavan, R. and Reynolds, A. C.: "Well Test

Analysis for Wells Producing Layered Reservoirs With Crossflow,"
Society of Petroleum Engineers Journal (June 1985) 380.
4. Kuchuk, F., Karakas, M. and Ayestaran, L.: "Well Testing and

Analysis Techniques for Layered Reservoirs," SPE Formation Evaluation
^ (Aug. 1986) 342,
5. Joseph, J., Bocock, A., Nai-Fu, F. and Gui, L. T.: "A Study of
Pressure Transient Behavior in Bounded Two-Layered Reservoirs:
Shengli Field, China," paper SPE 15418 presented at the 1986 Annual
SPE Technical Conference and Exhibition, New Orleans, Louisiana, Oct.
5-8.
6. Kuchuk, F., Shah, P. C. and Ayestaran, L.: "Application of Multilayer

Testing and Analysis: A Field Case," paper SPE 15419 presented at the
1986 Annual SPE Technical Conference and Exhibition, New Orleans,
Louisiana, Oct. 5-8.
7. Gao, C.: "Determination of Parameters for Individual Layers in

Milltilayer Reservoirs by Transient Well Tests," SPE Formation
Evaluation (March 1987) 43.
8. Camacho-V., R., Raghavan, R. and Reynolds, A. C.: "Response of

Wells Producing Layered Reservoirs: Unequal Fracture Length," SPE
Is Formation Evaluation (March 1987) 9.
9. Ehlig-Economides, C. A. and Joseph, J.: "A New Test for
Determination of Individual Layer Properties in a Multilayered
Reservoir," SPE Formation Evaluation (Sept. 1987) 261.
10. Shah, P. C., Karakas, M., Kuchuk, F. and Ayestaran, L. C.:

"Estimation of the Permeabilities and Skin Factors in Layered Reservoirs
With Downhole Rate and Pressure Data," SPE Formation Evaluation
(Sept. 1988) 555.
11. Larsen, L.: "Similarities and Differences in Methods Currently Used to

Analyze Pressure Transient Data From Layered Reservoirs," paper SPE
18122 presented at the 63rd Annual SPE Technical Conference and
Exhibition, Houston, Texas, Oct. 2-5, 1988.
12. Raghavan, R.: "Behavior of Wells Completed in Multiple Producing

Zones," SPE Formation Evaluation (June 1989) 219.
13. Olarewaju, J. S. and Lee, W. J.: "Pressure Behavior of Layered and

Dual-Porosity Reservoirs in the Presence of Wellbore Effects," SPE
Formation Evaluation (Sept. 1989) 397. •
14. Gao, C., Jones, J. R., Raghavan, R. and Lee, W. J.: "Interpretation of
Responses of Commingled Systems With Mixed Inner and Outer
Boundary Conditions Using Derivatives," paper SPE 22681 presented at
the 66th Annual SPE Technical Conference and Exhibition, Dallas,
Texas, Oct. 6-9, 1991.
15. Agarwal, B., Chen, H-Y. and Raghavan, R.: "Buildup Behaviors in
Commingled Reservoir Systems With Unequal Initial Pressure
Distributions: Interpretation," paper SPE 24680 presented at the 67th
Annual SPE Technical Conference and Exhibition, Washington, DC,
Oct. 4-7, 1992.
16. Ehlig-Economides, C. A.: "Model Diagnosis for Layered Reservoirs,"

SPE Formation Evaluation (Sept. 1993) 215.
17. Chen, H-Y., Raghavan, R. and Poston, S. W.: "Average Reservoir

Pressure Estimation of a Layered Commingled Reservoir," paper SPE
26460 presented at the 68th Annual SPE Technical Conference and
Exhibition, Houston, Texas, Oct. 3-6, 1993. is
•

Awb = cross sectional area of wellbore in communication
with formation being tested, ft2
B = formation volume factor, reservoir volume/surface volume
Bo = oil formation volume factor, RB/STB
Bw = water formation volume factor, RB/STB
cf = formation compressibility, psi-1
cg = gas compressibility, psi-'
co = oil compressibility, psi
ct = total compressibility, psi
cw = water compressibility, psi
cwb = compressibility of fluids in wellbore, psi i
C = wellbore storage factor, bbls/psi
CD = dimensionless wellbore storage factor
DF = damage factor, dimensionless
DR = damage ratio, dimensionless
E = flow efficiency, dimensionless
Ei(-x) = exponential-integral
k = volumetric average effective permeability of the
reservoir in the drainage volume of the test well, md
kg = effective permeability to gas, md
kw = effective permeability to water, md
in = natural logarithm, i.e., logarithm to base e
log = common logarithm, i.e., logarithm to base 10
L = distance to fault, or other boundary, ft
m = slope of semilog plot, psi/log-cycle
p* = false pressure obtained from buildup plot at
infinite shut-in time, psia ^
= dimensionless pressure
PD
5-130
•
pDMBH = dimensionless Matthews-Brons-Hazebroek pressure
dimensionless Muskat intercept pressure
pDMint
= initial reservoir pressure, psia
pi
pi = average reservoir pressure at time of last shut-in,
psia
PR = volumetric average reservoir pressure, psia
pwf = flowing formation face pressure, psia
pwf(At=0) = flowing formation face pressure immediately
preceding shut-in of well for buildup test,
psia
pws = formation face pressure during shut-in, psia
(pR pws)int = intercept pressure difference on Muskat plot,
psi
plhr = pressure from semilog straight line at At = 1 hour,
psia
q*gt = total gas production rate, Mscf/D
qo = oil production rate, STB/D
qos = Odeh-Selig modified flow rate, STB/D
^
qRt = total production rate of oil, gas and water, RB/D
= water production rate, STB/D
qw
rà = wellbore radius, ft
Ri = radius of investigation, ft
Rsja = solution gas-water ratio, scf/STB
sd = skin due to permeability alteration, dimensionless
sf = skin due to fracture, dimensionless
sp = skin due to perforations, dimensionless
sr = skin due to restricted entry, dimensionless
ssw = skin due to slanted well, dimensionless
st = skin due to turbulence, dimensionless
So = oil saturation, fraction
= water saturation, fraction
SW
^ tD = dimensionless time based on rw
tDA = dimensionless time based on A
5-131
tDL = dimensionless time based on L
•
tOS = Odeh-Selig modified time, hrs
tp = time well produced before shut-in, hrs
Vp = pore volume, ft3
Vwb = wellbore volume in communication with producing
formation, ft3
Aps = pressure loss due to skin, psi
At = shut-in time, hrs
Ate = Agarwal equivalent time, hrs
Atesl = shut-in time to end of semilog straight line, hrs
(AtDA)esl = dimensionless shut-in time to end of semilog
straight line
AtX = time of intersection of straight lines caused by
faults, hrs
AVwb= change in volume of wellbore fluids, ft3
at = total mobility, md/cp
u = viscosity, cp
p = density, ibm/ft3
= porosity, fraction
•
5-132
Equation Number Ecluat ion

in Text
HOILVER METHOD (m is a negative number
tp+Ot
16 2 . 6qBU log
5.5
pws pi kh At
m = - 162.6qBu
5.6
kh
k = _ 162.6qBp
5.8
mh
^, in
5. 9
5.10
PR P + 2. 30 3 pDhlBH
0.0002637ktp
• '
tDA = WtA
pwf(At=0) - plhr k
5.17 s = 1.151 - log
m
ûorw
t 2
+ 3.231
E = PR pwf Ap s
5.20
PR - Pwf
5.22 Aps = - 0.87ms
144A^^b
5.28
C = 57615p
5.29 C = Vwbewh
5.615
•
5-133
• S.30 C = 0.894C
D ^hcrw
t
qBAt
5.32 C=
24 [pws-pwf(At=0)]
0.14s
At > 170 000uCe
5.33
kh
ûctA
5.34
Atesl 0.0002637k(AtDA)esl
/kAtx
0uC
5.43 L = 0.01217
ct
tp+dt
5.48 pws = pi - 162.6 ^ t Rh log
• At
5.49 qRt = qoBo +(1000q t

g a oRs awRsw)B g
+ B
aw w
5.50 xt 162.6 mh t
5.52 k = - 162.6 °°°

0 mh
5.53 11-W =- 16 2. 6 aw B
mh ^w
(1000q ^ t - qoRs - c{wRsw)B g u g

5.54 k g = - 162.6
mh
P wf (Ot=O) - p
• S.55 s 1.151
m
lhr - log t
trw
Z
+ 3.231
5-134
MILLER-DYES-HUTCHINSON METHOD ( m is a positive number)
5.64 pws = Pi + 162khqB^ log At
+ 162.6qBU lo 4A
g k - lo g
kh 1.781CArW
ûctr^,
0.2339qBt p
- 3.23] -
Ah ^c t
5.65 m = 162.6 kh
qBV
5.66 qBw
k = 162.6 M11
Plhr Pwf(At=0) k
5.67 s = 1.151 - log
m ûctrw
+ 3, 231
5.68 Aps = 0.87ms
ûctA
5.74 (At)PR = 3792 kC A
EXTENDED MUSKAT NIETHOD
141 . 2cIBu
5.75 k = PDMint
h(PR Pws) int
•
5-135
• AGARWAL METHOD ( m is a positive number)
t At
S. 79 Ate = t p +At
p
S.80 pwf (At=0) + 162 . 6 qBll

pws [log Ate
+ log k - 3.23 + 0.87s

Ciotrw
5.81 m = 162.6 kh
qBp
• 5.82 k = 162.6 mh
qB"
5.83 s = 1.151 l
L
e) - pwf(At=0)
m
- log k + 3.23
$uctrw 2
•
5-136
PROBLEMS
•
PRESSURE BUILDUP TEST
1. A consolidated sandstone reservoir containing an

undersaturated oil is located along the Gulf Coast.
The reservoir contains 14 wells drilled on 80-acre
spacing in a square pattern. The development wells
were drilled while awaiting a court settlement
involving the laying of a pipeline to transport the
produced oil. After a favorable settlement, all
wells were placed on production at about the same time
at approximately constant rates. One particular inner
well was shut in after 74 hours of production for a
pressure buildup test. The well produced at a constant
rate of 460 STB/day prior to shut in. The buildup
pressure and other rock and fluid data are tabulated
below. You are to estimate k, s, PR) and flow
efficiency. These data are not influenced by wellbore
storage or near-wellbore effects; accordingly, you may
assume that the transient flow straight line starts
at the beginning of the test.
At, hours pk,s, psia
0.092 4033
0.23 4040
0.46 4045
0.92 4050
2.3 4056
4.6 4060
9.2 4062
23 4064
•
5-137
0 q = 460 STB/D A = 80 acres
B = 1.46 RB/S'I'B = 26%
u = 0.75 cp c o = 20xl0-6psi 1
h = 18 ft c w = 3xl0-6psi-1
r = 0.295 ft S = 22%
w w
pw f(4t=0) = 3,803 psia
2. The following pressure buildup data, published by Lee27,

were obtained from a well producing above the bubble
point. At the time of the test, the well had produced
141,980 STB's of oil and was flowing at a stabilized
rate of 250 STB/D. The well is an interior well of a
reservoir developed on a 160-acre square drilling
pattern. The well is producing from 2-inch tubing with
a cross-sectional area of 0.0218 ftZ. A packer is
located immediately above the perforations. Other
reservoir data arc:
• q 0 = 250 STB/D rw = 0.198 ft

PO = 0.8 cp = 53 lhm/ft3
Po
= 0. 0 39 pr-b- wirol"^ Awb = 0.0218 ft2
B 0 = 1.136 RB/STB h = 69 ft
ct = 17x10-'psi-1 tp = 13,630 hrs
p`^f(At=0) = 3,534 psia Depth to midpoint of
perforations = 11,233 ft
t + at
At, hrs pot Pws' psia Ap, psi Apt
ln
0.15 90,900 3,680 146
0.2 68,200 3,723 189
0.3 45,400 3,800 266 208.7
0.4 34,100 3,866 332 228.9
0.5 27,300 3,920 386 247.3
1 13,600 4,103 S69 238.0
2 6,860
• 4 3,410
4,250
4,320
716
786
156.5
68.8
5-138
NO. 340-L310 DIETZGEN GRAPH PAPER OIETZ6EN CORPORATION
SEMI-LOGARITHMIC wôc
3 CYCLES X 10 DIVISIONS PER INCH
CJl
F-
W
1^0
• • ^
q
6 2,270 4,340 806 52.9
7 1,950 4,344 810 39.9
8 1,710 4,350 816 38.2
12 1,140 4,364 830 33.1
16 853 4,373 839 32.5
20 683 4,379 845 28.5
24 569 4,384 850 28.0
30 455 4,393 859 24.0
40 342 4,398 864 21.8
50 274 4,402 868
60 228 4,405 871
72 190 4.407 873
A Horner plot and log-log plot of the buildup data
are presented..
(a) Compute the wellbore storage factor expected based

on the well completion.
(h) Using the log-log plot, compute the wellbore

storage factor based on the measured data.
(c) Using the log-log plot of pressure data, pressure

derivative data, and type curve matching, determine
where wellbore storage ends and justify your selection
of the semilog transient flow straight line. Calculate
permeability and total skin factor using conventional
semilog analysis.
(d) Use the Bourdet derivative type curve to determine

permeability, wellbore storage factor and total skin
factor.
3. The following pressure data are from a consolidated

reservoir located offshore Trinidad. The test well
is the only well in a fault block which, based upon
seismic data and well performance, is believed to have
the following approximate drainage configuration:
•
5-140
1000
PROBLEM 2
100
F-+
F-+
A Ap
^ Apln
10
0.1 1 10 100
At, hrs
• • •
• • •
4600
PROBLEM 2
4400
4200
.^,
^
^n
4000
3800
3600
102 103 10`` 105
t +At
p
At
SEALING FAULT •
1
Using the data given,
(a) Confirm the existence of the fault from the buildup

test and estimate the distance to the fault;
(b) Compute the effective permeabilities to all fluids

flowing;
(c) Compute the total skin factor;

•
(d) Estimate the volumetric average reservoir pressure.
6 ]
A = 81 acres cf = 3.3 X 10 psi
tp = 43,869 hours So = 0.40
CIO = 817 STB/D Sg = 0.15
qw = 2,327 STB/D Sw = 0.45
q gt= 1,512 NIscf/D B 0 = 1.19 RB/STB
uo = 0.57 cp Bg = 1.96 x 10 -3 RB/scf
ua = 0.0142 cp Bw = 1.02 RB/S'I'B
uw = 0.45 cp Rs = 330 scf/STB
co = 31.1 x 10 6 psi1
- h = 110 ft
cw = 5.3 x 10 psi i 26%0
cg = 7.S x 10-4psi-' pwf(At=0) = 1,012 psia
r = 0.356 ft
w
•
5-143
•
t +At
t hr P n
0.083 526,640 1064 52

0.167 263,162 1077 65
0.250 175,477 1087 75
0.333 131,740 1097 85
0.417 105,278 1104 92
0.50 87,739 1110 98
0.75 58,493 1130 118
1.00 43,870 114S 133
1.25 35,096 1158 146
1.50 29,247 1167 155
1.75 25,069 1172 160
2.0 21,936 1175 163
2.5 17,549 1178 166
3.0 14,624 1190 178
3.5 12,535 1197 18S
4.0 10,968 1200 188
4.5 9,750 1202 190
5.0 8,775 1203 191
5.5 7,977 120S 193
6.0 7,313 1.205 193
7.0 ^ 6,268 1208 196
8.0 5,485 1210 1.98
9.0 4,875 1213 201
10.0 4,388 1215 203
15.0 2,926 1221 209
20.0 2,194 1230 218
•
25.0 1,756 1235 223
30.0 1,463 1243 231
35.0 1,254 1248 236
40.0 1,098 1253 241
45.0 976 1257 245
48.0 915 1259 247
A Horner plot of the buildup data is presented.
5-144
1300
1250
1200
.^,
1150
U^
4-
cn
1100
1050
1000
102 103 10" 105 106
t Tl +At
At
0 is 0
^ NL7 3•1tJI C93 DIETZGEN L;RAPH PAPER ^ DIETZGEN CORPORATION
LOGARITHMIC -DE IN u.e.A. ^.-.,^ •il^...,^
3 CYCLES X 3 CYCLES • , _}
' N 41 p lp a J a o^ N 4i pIA a v m° - N W P (A 01 V m° 0
.i . 4
' t . ' rl { .• 1 1 ..I ^ ^ ^
- -
r . I ^{ .
i .I f 1 I :^ , l,; } •1
N - ^ . ^ ..F i f . •r I• t = 1^ = - - - .
r # { :l l 1! tt 1 ,I f
AR
a = Y i '{ ^: j }1 4 {_ t. c l
. . . 1,1
u
: . `. ^ ^ 1: {r r r i ^ ^
^ -- t t I I V : ^^ 3'_ ti':
^
_ i- .. F 14 :1
ti• .
I:T 4 ffli I F hill
.. i ^ . .
- - I 1 _- -_ -
M UM,
'lll ^
; i..:; ^ .; ^ • •^ ^ - ,^, f^
f
ft ,, ^ ^ ^:. , _ ,I 1, j, I ^.
r i . i1 1 , Lf ., . , } _ i` }: i .i I ^• . Gr ' t tj ^ t
N „^ ,:q 'i^, 1,
i ^ {' f•
^ 7
4 f ^r
W # '
4-
^ :: ^ t ` ^ 3 i ;_ ` ^ ' t j'' ` '" ^1'; : ^ ^1 ;' - -^ ♦ 1 ' J{
i i• . _ .FuL'
"t:j i+ .L }'' u: _
u , • _ - -_
a - ^- - ^ - i-^ ^ ^ r ^^^ - ^ `^ :.^ ., 1" - i *

- • ^ } -^ t' i; • . i i, l l•
a _ t i f= _ E
t ^ : { r {
° r jit ^! ^ ^' r I {
{!
: I
1 ;,. . . . .
L^ . . ,^ ^ . . , .:
I` i I• . ^^
N _ . ,.t i' ._ _ il t - - •f ."1 ^•

f
W T
.? y 4^tl4t 4!:.I
: 1 fi r r
,
^ i ^ C f , i
_ ^} f l
u f'_ i , • i
47
1. t î: »:j sE
.. ;t-.
3 Uh i
4. An oil well is located i n the center of a closed
!
square drainage pattern. The well produced at a
constant rate of 210 STB /day for 50 hours. The well
was shut-in, and a press ure buildup test was conducted.
Reservoir rock and fluid properties are listed along
with the buildup data. The effects of wellbore storage
are negligible. Analyze the data by the Miller-Dyes-
Hutchinson method and es timate permeability, average
pressure and total skin factor. Reservoir pressure is
greater than the bubble point pressure.
A= 80 acres uo = 0.70 cp
^ = 22% r w = 0.17 ft
c t = 14 x 10-6psi-1 B o = 1.37 RB/STB
h = 12 ft pwf(At=0) = 2,360 psia
,^t , hours pws, p sia
0.100 2709
0.165 2716
0.331 2726
0.497 2731
0.628 2735
1.00 2741
1.65 2747
2.48 2753
3.31 2756
4.97 2761
6.62 2764
8.27 2766
9.93 2768
5. Analyze the buildup test described in Problem 2 for

permeability, total skin and p* using the Agarwal
method. Compare the per meability and skin factor to
those found in Problem 2 using the Horner method.
S-147
NO. 340-L310 DIETZGEN GRAPH PAPER DIETZGEN CORPORATION
SEMI-LOGARITHMIC Mnoc IN u-e-A-
iû. utiu-û^ ui^^^. ^^ .vr^ r.rrcrt UIÎtuC1^1 t.UFtt'tIKAl1U1V
LOGARITHMIC -DK IN U.S. A.
K,,`^^ 3 CYCLES X 3 CYCLES ., ^^;
(Jl
}--1
0 • •
• NO. 340-L310 DIETZGEN GRAPH PAPER
SLMI LOGARITHMIC
^ OIETZDEN
MAU[
CORPORATION
IN U. ^. A.
,•
(Jl
I--'
vl
6. The following pressure buildup data were obtained from
•
a well in a West Texas carbonate reservoir. The test
was conducted using an acoustical well sounding device.
Prior to the test, the well had produced 98 days since
last shut-in at an average rate of 6.6 STB/D. Determine
the effective permeability to oil and total skin factor
using the following methods:
(a) Ramey-Agarwal type curve

(b) Bourdet derivative type curve
(c) McKinley type curve
(d) Conventional semi-log analysis
h = 32 ft ( 2SPF) ct = 24 x 10 6psi 1
Bo = 1.15 RB/STB P o = 2,5 cp
= 12% rw = 0.33 ft
Well spacing = 40 acres (square drilling pattern)

PR = 1200 psia (material balance estimate)
•
5-151
•
At, hrs At e , hrs
tp+nt
p ws , psia Op, psi Ap'l n
At
0 0 165 00 - -
0.5 0.5 175 4705 10 -
1.0 1.0 191 2353 26 -
2.0 2.0 221 1177 56 60
3.0 3.0 248 785 83 79
4.0 4.0 274 589 109 99
5.0 5.0 29S 471 130 114
6.0 6.0 317 393 152 131
7.0 7.0 339 337 174 143
8.0 8.0 358 29S 193 160
9.0 9.0 376 262 211 174
10 10.0 395 237 230 185
11 10.9 413 215 248 196
12 11.9 429 197 264 205
^ 14 13.9 464 169 299 224
16 15.9 496 148 331 237
18 17.9 525 131.7 360 248
20 19.8 551 118.6 386 259
22 21.8 577 107.9 412 257
24 23.8 601 99,0 436 262
30 29.6 662 79,4 497 277
36 35.5 714 66.3 549 296
42 41.3 749 57.0 584 299
44 43.2 759 54,5 594 301
46 45.1. 763 52.1 598 305
48 47,0 782 50.0 617 301
52 50.9 815 46.2 650 297
56 54.7 844 43.0 679 282
60 58,5 869 40,2 704 280
64 62.3 890 37.8 725 271
68 66.1 908 35.6 743 263
72 69.9 922 33.7 757 271
^ 76 73.6 937 31.9 772 267
5-152
80 77.4 951 30.4 786 -
84 81.1 962 29.0 797 -
88 84.8 974 27.7 809 -
92 88,5 981 26.6 816 -
96 92,2 985 25.5 820 -
97 93.2 987 25,2 822 -
98 94.1 990 25.0 825 -
100 95.9 994 24.5 829 -
104 99.6 1001 23.6 836
108 103.2 1008 22.8 843 -
112 106.9 1017 22.0 852 -
116 110.6 1023 21.3 858 -
120 114.2 1026 20.6 861 -
126 119.6 1036 19.7 871 -
132 125.0 1043 18.8 878 -
133 125.9 1051 18.0 886 -
5-153
•
• • •
1200
PROBLEM 6
1000
300
Ln 600
^• c^.
400
200
10 102 103 104
t {-At
p
At
10`
PROBLEM 6
<
102
vn
;r a
.I--
• Pressure
n Pressure Derivative
10
10 102 103
Ot, hrs
0 • •
•
SPE 9289 SPE
SOcêtb of PeU'dam Efêgè Of AIME
A NEW METHOD TO ACCOUNT FOR PRODUCING TIME EFFECTS WHEN DRAWDOWN TYPE
CURVES ARE USED TO ANALYZE PRESSURE BUILDUP AND OTHER TEST DATA
by Ram G. Agarwal, Amoco Production Co.
©Copyright 1980, American Institute of Mining. Metallurgical, and Petroleum Engineers, Inc
This paper was presented at the 55th Annual Fall Technical Conference and Exhibition of Ihe Society of Petroleum Engineers of AIME, held in Dallas. Texas, September 21 -24, 1980.
The material is subject to correction by the author. Permission to copy is restricted lo an abstract of not more than 300 words. Write 6200 N Central Expwy., Dallas, Texas 75206.
INTRODUCTION
ABSTRACT
Type curves have appeared in the petroleum lit-
Currently, type curve analysis methods are erature since 1970 to analyze pressure transient
being commonly used in conjunction with the conven- (pressure drawdown and pressure buildup) tests taken
tional methods to obtain better interpretation of on both unfractured and fractured wells. The
majority of type curves' 8 which have been developed
•
well test data. Although the majority of published
type curves are based on pressure drawdown solu- and published to date were generated using data
tions, they are often applied indiscriminately to obtained from pressure drawdown solutions and obvi-
analyze both pressure drawdown and buildup data. ously are most suited to analyze pressure drawdown
Moreover, the limitations of drawdown type curves, tests. These drawdown type curves are also commonly
to analyze pressure buildup data collected after used to analyze pressure buildup data. The applica-
short producing times, are not well understood by tion of drawdown type curves in analyzing pressure
the practicing engineers. This may often result in buildup data is not as bad as it may first appear.
an erroneous interpretation of such buildup tests. As long as the producing time, t, prior to shut-in
While analyzing buildup data by the conventional is sufficiently long compared topthe shut-in time,
semi-log method, the Horner method takes into At (that is (t +pt)/t - 1], for liquid systems,
account the effect of producing time. On the other it is reasonablg to analyze pressure buildup
hand, for type curve analysis of the same set of data using drawdown type curves. However, for cases
buildup data, it is customary to ignore producing where producing times prior to pressure buildup
time effects and utilize the existing drawdown type tests are of the same magnitude or only slightly
curves. This causes discrepancies in results larger than the shut-in times [that is, (t +At)/t
obtained by the Horner method and type curve >> 1], the drawdown type curves may not bepused tg
methods. Although a few buildup type curves which analyze data from pressure buildup tests. The above
account for the effect of producing times have requirement on the duration of producing times is
appeared in the petroleum literature, they are the same for the conventional semi-log analysis. If
either limited in scope or somewhat difficult to pressure buildup data obtained after short producing
use. time are to be analyzed, the Horner methodlo is
recommended over the MDH ( Miller-Dyes-Hutchinson)
In view of the preceding, a novel but simple method.9 The MDH method is generally used to ana-
method has been developed which eliminates the lyze buildup data collected after long producing
dependence on producing time effects and allows the times, whereas the Horner method is used for those
user to utilize the existing drawdown type curves obtained after relatively short producing times.
for analyzing pressure buildup data. This method Although pressure buildup tests with short producing
may also be used to analyze two-rate, multiple-rate times may occur often under any situation, they are
and other kinds of tests by type curve methods as rather more common in the case of drill stem tests
well as the conventional methods. The method and pre-fracturing tests on low permeability gas
appears to work for both unfractured and fractured wells. -
wells. Wellbore effects such as storage and/or
damage may be taken into account except in certain Thus, there is a need for generating buildup
cases. type curves, which account for the effects of pro-
The purpose of this paper is to present the new

method and demonstrate its utility and application
by means of example problems.
ducing time.
this regard.
Some limited work has been done in
McKinley" has published type curves
for analyzing buildup data for a radial flow system.
However, his buildup type curves were generated on
•
the assumption of long producing times; and these
References and illustrations at end of paper. type curves are therefore very similar to drawdown
NEW METHOD TO ACCOUNT FOR PRODUCING TIME EFFECTS WHEN DRAWDOWN TYPE
CURVES ARE USED TO ANALYZE PRESSURE BUILDUP AND OTHER TEST DATA SPE 9289
• type curves and are obviously unsuitable for cases

where producing times prior to shut-in are rela-
BASIS OF DRAWDOWN AND BUILDUP TYPE CURVES
tively short. Crawford, et al., 12 pointed out the A type curve is a graphical representation of a
above limitations for McKinley type curves in ana- mathematical solution (obtained analytically or
lyzing pressure buildup data from the DST tests. numerically) for a specific flow type. The solution
They also presented buildup curves for short pro- is normally plotted, in terms of dimensionless vari-
ducing times. Since their curves deal with specific ables, on log-log graph paper. The graph thus pre-
values of real producing times prior to shut-in, pared becomes the type curve for the specific flow
they are limited in scope and utility. Recently, problem with given inner and outer boundary condi-
the effect of producing time on analysis of pressure tions. Depending on the type of solution (drawdown
buildup data using drawdown type curves has been or buildup), drawdown and buildup type curves are
discussed by Raghavan.13 His study clearly points generated.
out the limitations of drawdown type curves for ana-
lyzing buildup data collected after small producing Drawdown Type Curves
times. A family of buildup type curves is presented
both for unfractured and fractured wells with pro- As the name implies, these type curves are
ducing time as a parameter. Although these type based on the drawdown solutions. The pressure draw-
curves offer a definite advantage over the existing down solution for a well producing at a constant
drawdown type curves, they are difficult to use rate as a function of flowing time, t may be written
because of the multiplicity of type curves. In a as
recent paper, Agarwa119 also discussed the limita-
tions of using drawdown type curves for analyzing
buildup data obtained after small producing times kh[pi-pwf(t))
but no details were given. These limitations are
141.2 qBp - pwD(CD) (1)
discussed here in this paper. Recently Gringarten,
et a1.15, presented drawdown type curves, plotted
in a slightly different form, and suggested some
guidelines regarding the portions of buildup data where,
which may be analyzed by drawdown type curves.
Although these guidelines may be useful in certain
cases, the basic problem still remains.
t _ 2.634 x 10-4kt (2)
• To overcome the above-mentioned difficulties

and to eliminate dependence on producing time, a new
method has been developed. This method should pro-
vide a significant improvement over the current
D O(Nct)i rw2
methods because (1) this permits us to account for Eq. (1) is a general solution and is not meant
the effects of producing time, and (2) data are nor- to be restricted to any particular drainage shape or
malized in such a fashion that instead of utilizing well location. The majority of the published type
a family of type curves with producing time as a curvesl 8 for both unfractured and fractured wells
parameter, the existing drawdown type curves may be are based on pressure drawdown solutions for liquid
used. This concept appears to work for both unfrac- systems. Examples of pressure drawdown type curves
tured and fractured wells. Wellbore storage effects for unfractured wells are those presented by
with or without damage may also be taken into Agarwal, et al.,1 Earlougher and Kersch4 and
account provided that producing time prior to Gringarten, et al.ls In another publication Grin-
shut-in is long enough to be out of such wellbore garten, et al.,S-presented type curves for verti-
effects. cally fractured wells with infinite flow capacity
and uniform flux fractures. Type curves for finite
This method has been extended to include anal- flow capacity fractures were provided by Cinco
ysis of data from two-rate tests8i1s'17 and multiple et a1.,6 and.Agarwal, et al.7 More regarding
rate tests8r17'1s by type curve methods. Although the use of above type curves for analyzing buildup
not shown, it appears to have a potential for data will be said later.
applying type curve methods to other kinds of
testing. Buildup Type Curves
This new method, although originally conceived To obtain pressure buildup solutions, superpo-
for type curve analysis of buildup data, is quite sition may be applied in the normal manner to pres-
suitable for the conventional semi-log analysis. It sure drawdown solutions. This provides buildup
is similar to the Horner method because it includes pressures at shut-in times, At after a producing
the effects of producing time, and may be used to time, t . Fig. 1 shows a schematic of pressure
determine formation flow capacity, skin factor and builduppbehavior obtained following a constant rate
the initial reservoir pressure. However, it has an drawdown for a production period, t . Flowing pres-
added advantage. It allows the plotting of pressure sures pwf(t) are shown as a functioR of flowing
buildup data, with and without producing time time, t up to a production period, t, when a
effects, on the same time scale as the graph paper. buildup test is initiated. Buildup Fressures,
This enables a better comparison of data using the p s(t +pt), are shown as a function of shut-in time,
• MDH and Horner type graphs.
Although the new method will be developed using

the solutions for liquid systems, its applicability
At pInstead of taking a buildup test, if the well
was allowed to produce beyond time, t flowing pres-
sures as shown by p f(t +At) would ha^re been
obtained. Note thaYthg flowing pressure at the end
to gas wells will also be indicated. of the production period which is denoted by pwf(t
is same as the buildup pressure at the instant of p
3
RAM G. AGARWAL
^
SPE '9289
shut-in which is shown as p(Ot=O). Superposition

pwD[(tp)D]-PwD[(tp+At)D] = 0 (9)
when applied to drawdown solutions provides the fol-
lowing.
On Fig. 1, the above difference has been shown
as the cross-hatched area and may be defined as
kh[pi Pws(tp+dt)]
141.2 qBu
(10)
(AP)difference - Pwf(tp)-Pwf(tp+At)
= Pwp[(tp+Ot)D]-PwD[(6t)DI (3)
or (11)
0P)difference ' pws(At-0)-Pwf(tp+At)
The flowing pressure, pwf(t ) at the end of
producing period, tp is given byP
As producing period t gets smaller or At gets
larger, the difference sho& by Eqs. (9) through
kh[pi pwf(tp (11) can no longer be ignored and the use of draw-
(4) down type curves to analyze pressure buildup data
141.2 qBp PwD[(tp)D] becomes invalid. The impact of the assumption shown
by Eq. (9) will be discussed first in a generalized
fashion followed by its impact on type curves for
Subtracting Eq. (3) from Eq. (4) and substituting specific flow regimes. Finally, the new method will
pWS(At=0) for pwf(tp), we obtain be discussed which accounts for producing time
effects for analyzing pressure buildup data.
kh[pws(tp+At)-pws(At=0)] Fig. 2 schematically shows pressure buildup

behavior obtained following a constant rate drawdown
141.2 qBu but at the end of three succesbsivel^ incrasing pro-
ducing periods, t such that p3 > p > p1. The
= PwD[(tp)D]-PwD[(tp+Ot)D]'{'PwD[(at)D] (5) cross-hatched area shown at the end oi each produc-
tion period denotes the difference between
(AP)drawdown and (AP) buildup represented by Eq. (10)
Eq. (5) provides a basis for buildup type curves and
has been utilized in this paper.
Before discussing the new method, let us review

or (11). Note that (AP)difference gets smaller as
the length of the producing period increases.

•
the simplified version of Eq. (5) which has been
commonly used in the past and has provided the basis This can be better shown by means of Fig. 3
of utilizing drawdown type curves for analyzing where (Ap) drawdown vs. flowing time, t has been com-
pressure buildup data. If producing time, t , is
significantly larger than the shut-in time, PAt, it pared with (AP)buildup vs. shut-in time, At with
is reasonable to assume that [(t +At)/t ] = 1.
Although approximate, this also Ymplicspthat producing period t as a parameter. Although sche-
(tp+Ot) = tp, or PwD[(tp+at)D] = PwD[(tp)D]. matic, Fig. 3 cleaily indicates that there is a
significant difference between (aP)drawdown and
Thus, Eq. (5) can be simplified as a pressure
buildup equation as shown below:
(AP)buildup for small producing periods. However,
kh[pws(tp+At)-pws(At=0)] this difference gets smaller as the length of the

141.2 qBy. = PWD[(ot)D] (6) producing period increases. Also note that for a
given producing period, the difference between the
two (Ap)s is small at early shut-in times but it
gets bigger as shut-in time, At, increases. Fig. 3
A comparison of pressure buildup equation (6) clearly indicates the limitations of using drawdown
and the pressure drawdown equation (1) indicates type curves for analyzing pressure buildup data
that they are similar at least for cases where pro- where producing period, t , prior to shut-in is
ducing period, t is significantly longer than the relatively small. P
shut-in time, At? It also implies that (6p)
vs. flowing time, t is equivalent to (pp) drawdvsn
Next we will examine the impact of this differ-
shut in time, At, where buildup
ence on type curve analysis for the specific flow
regimes (radial flow, linear flow, etc.) and discuss
the new method which accounts for producing time
(AP) drawdown - Pi Pwf(t) (7) effects.
UNFRACTURED WELL
(AP) buildup - Pws(tp+At) - pws(At=0) (8)
Since Eq. (6) has been derived from Eq. (5)

based on the assumption of long producing period,
tp, the difference
Infinite Radial System (s=0;
Let us first consider the pressure drawdown

solution for a well producing at a constant rate in
•
a radial system.
• pwD(tD) = Z [ln(tD) + 0.80907] (12)

t D=- corresponds to the pressure drawdown solution
gpiven by Eq. (12). Since data are plotted on
semi-log graph paper (for t > 100), the pres-
sure drawdown solution on Fpg 4 is a straight line.
This figure clearly points out the limitations of
Eq. (12) is based on the assumption that well using pressure drawdown solution in a conventional
bore effects (storage and skin) are negligible and or type curve analysis mode to analyze pressure
the dimensionless time, tD > 100 such that the buildup data obtained after short producing times.
log approximation applies to the 6.-solution. Sub However, buildup curves may be utilized for type
stitution of Eq. (12) in Eq. (1) provides curve matching purposes as discussed by Raghavan.13
The obvious disadvantage is that this requires the
use of a family of type curves.
kh[pi-pwf(t)] 1
[ln(tD) + 0.80907] (13) To overcome the above difficulty, a new method
141.2 qBp -
has been developed which should allow us to analyze
pressure buildup data by means of pressure drawdown
type curves. This new method may also be used to
Eq. (13) is a pressure drawdown solution for a perform the conventional analysis.
radial system which also forms the basis for
semi-log straight line on a semi-log graph paper.. NEW METHOD
If Eq. (12) is substituted in Eq. (3), the well
known Horner pressure buildup equation is obtained. Eq. (5), presented earlier as a pressure
buildup solution, forms the basis for this new
method. Substitution of Eq. (12) in Eq. (5)
kh[pi-pws(tp+At)] 1 (tp+At)D
= 2 [ln AtD ] (14)
141.2 qBp
kh[pws(tp+At)-Pws(At=0)]
141_2 qBN
In Eq. (14) the subscript D may be dropped if

desired. The above equation also takes into account 1 (tpD x AtD)
producing time effects. Unfortunately Ap = = 2 [In (t + At) + 0.80907] (16)
[p-p (t +pt)] on the left hand side of Eq. (14) p D
• requires pa knowledge of initial reservoir pressure,
pi
which is generally not known. Consequently,
Eq. (14) is not suitable for the purposes of type
curve matching. However, (AP)buildu defined by
Eq. (16) thus becomes a pressure buildup solu-
tion for the infinite radial system. A comparison
Eq. (8) is generally known and is noprmally used for of pressure drawdown, Eq. (13) with Eq. (16), indi-
type curve analysis. If Eq. (12) is substituted in cates that pressure drawdown curves generated by
Eq. (6), the simplified pressure buildup equation is Eq. (13) should be the same as the pressure buildup
obtained. curves obtained by Eq. ( 16). To demonstrate this,
the family of buildup curves shown in Fig. 4 and
plotted as a function of dimensionless shut-in time,
kh[Pws(t p +At)-p (c1t=0)] were replotted as a function of the time group,
c.s At D'
(t D x AtD)/(tP + At)D. Results are shown in
141.2 qBp
F. 5.
Note that it is possible to normalize a family

= 2 [ln(AtD) + 0.80907] (15) of buildup curves into a single curve. Moreover,
this single curve is the same as the pressure draw-
down curve. The preceding suggests that for ana-
lyzing pressure buildup data by drawdown type
Eq. (15) is the familiar MDH (Miller-Dyes- curves, (AP) buildup data should be plotted as a
Hutchinson) equation for pressure buildup and
assumes that the producing period prior to shut-in function of (t x At)/(t + At) rather than just the
is sufficiently long such that transients during the shut-in time, Pat. P
flow period do not affect the subsequent pressure
buildup data. Thus, it should be obvious that For the sake of simplicity and brevity let us
Eq. (15) is not suitable for pressure buildup anal- define the new time group as an equivalent drawdown
ysis (conventional or type curve) when producing time, Atedt or further abbreviated as Ate, where
times prior to shut-in are small. However, Eq_ (5)
may be used, as was done by Raghavan, to generate a
family of pressure buildup curves with dimensionless t x At
producing time, t D, as a parameter. Fig. 4 pre- Ate = tP(17)
+ At
sents such result^ for an infinite radial system. P
Dimensionless pressure change, p w^ during buildup
has been plotted as a function of eimensionless
shut-in time, fltD, with dimensionless producing In a dimensionless form, Eq. (17) may be
time, tpD, as a parameter. Note that PwDs will be expressed as
defined later as Eq. (19) and is the same as the
left hand side of Eq. (5). The buildup curve with
SPE 9289 RAM G. AGARWAL
^teD
(tpD x AtD)
(t + At) D
(18)
Infinite Radial System (s=0; •
To study the effect of storage on buildup type
curves, data presented by Agarwal, et al.,1 were
The dimensionless pressure change during utilized. Pressure drawdown data pw p vs. tD data
buildup or a rate change may be defined as13 for s=0 and CD=1000 were taken from Table 3 of the
above paper. Eq. (5) was used to generate the pres-
sure buildup data for a number of producing times as
_ kh[pws(tp + At) - Pws(At=0)] was done by Raghavan.13 Both pressure drawdown data
pwDs (19) and pressure buildup data are plotted on Fig. 6
141.2 qBp
(semi-log graph paper) as a function of t n and AtD
respectively. Note that a family of buildup curves
is obtained with producing time, t p ( 103 to 106) as
In establishing the new method, it was previ- a parameter. These data are also FIotted on log-log
ously assumed that wellbore effects (such as storage graph paper as shown by Fig. 7. These figures
and skin) are negligible. It appears that skin further emphasize the limitations of using pressure
effect, s, may be considered in the development of drawdown curves to analyze pressure buildup data
this method. obtained after short producing periods. Fig. 7
shows that unit slope lines for buildup data are
Infinite Radial System (s#0; shifted to the right of the drawdown curves. If
dimensionless storage, CD, is computed using the
buildup data, the computed value of C n will be erro-
If skin effect, s, is introduced in the pres- neously high. Moreover, if pressure buildup data
sure drawdown solution given by Eq. (12), we obtain are forced to match the pressure drawdown type
curve, the computed value of formation flow capacity
(kh) will be erroneously optimistic. The magnitude
of error will increase with decreasing producing
PwD(tD) = 2 Iln(tD) + 0.809071 + s (20) period.
Figs. 8 and 9 are the replots of pressure

buildup solution on semi-log and log-log graph
If we go through the same steps as we did for
an infinite radial system (s=0; CD=0) and instead of
utilizing Eq. (12) we use Eq. (20), it will be
obvious that the new method is equally valid when
s#0. Pressure drawdown Eq. (13) and the new pres-
papers utilizing the new time group. Fig. 8 indi-
cates that almost all pressure buildup curves are
normalized except two which correspond to dimension-
less producing period, t D equal to 103 and 104.
Although these curves donot seem to appear bad on
•
sure buildup Eq. (16) will respectively become the semi-log paper, they look rather poor on the
Eqs. (21) and (22) as shown below: log-log graph in Fig. 9. The reason for this may be
obvious if we inspect the following equation19 which
provides the time for storage effects to become neg-
kh[Pi-pwf(t)] ligible.
141.2 qBp
For s=0
= Z fln(tD) + 0.80907] + s (21)

tD > 60CD (23)
kh[pws(tp+At)-Pws(At=0)] If Eq. (23) is used for the subject problem,

141.2 qBp C=1000, the minimum producing time required for the
sporage effects to become negligible will be equal
to 6 X 104. Since the producing periods in the two
1 (tpD x AtD) cases were only 103 and 10q respectively, pressure
2 ^ln (t + pt) + 0.80907] + s (22) buildup data could not be normalized. Based on a
p D number of cases studied, it appears that it,is pos-
sible to normalize the pressure buildup curves pro-
vided that the producing time, t D, is at least
The above equations establish the validity of equal to or greater than that giîen by Eq. (23).
using pressure drawdown type curves for pressure
buildup analysis even when skin is present. Infinite Radial System (s#0; CD#0)
Based on the encouraging results obtained thus

far, we wanted to apply this concept of the equiva- Agarwal, et al.,' data were taken for a
lent drawdown time, At to other wellbore effects number of cases for non-zero values of CD and s.
and also to other floweregimes.
•
The attempt was Although not shown in this paper, results indicate
made to establish the validity of this concept for that pressure buildup curves are normalized when the
the above situations by graphical means rather than new method is used. The lower limit of the pro-
the mathematical solutions. Let us first consider ducing time for s#0 is determined by the following
the infinite radial system with wellbore storage equation.20
effects.
• tpD > (60 + 3.5 s)CD (24) CONVENTIONAL ANALYSIS USING THE NEW METHOD
Although the new method was originally con-

The preceding discussion establishes the val- ceived as to be used for type curve analysis pur-
idity of using Agarwal, et al'sl pressure drawdown poses, it also appears useful for analyzing pressure
type curves (radial flow with storage and skin buildup data by the conventional semi-log analysis
effect) for analyzing pressure buildup data provided methods. This can be seen by rewriting Eq. (22) in
that the new method is used. the following familiar form:
APPLICABILITY OF NEW METHOD TO OTHER TYPE CURVES

IPws(tp+^t)-Pws(At=0)] = 162kh qBu
It appears that it is possible to extend this
method to other drawdown type curves which have
appeared in the petroleum literature. Two sets of (t x At)
p k
type curves will be considered: (1) Earlougher and
[log (tp + At) + log ûc r 2 -
Kersch,4 and (2) Gringarten et a1.15
tw
Earlougher and Kersch Type Curves`' 3.23 + 0.87 s] (26)
These type curves are based on pressure draw-

down solution and are applicable to an infinite or
radial system with wellbore storage and skin. They
are basically the same type curves as that of
Agarwal et al.1 because both use the same solu- [APbuildup] = mjlog(Ate) +
tion. However, they are distinctly different in
appearance because data are plotted differently. A
schematic of their drawdown type curves is shown in log k 2 3.23 + 0.87s] (27)
Fig. 10, where (pwDC )/tD has been plotted as a
ûctrw
function of t/C wi?h C el as a parameter. where
D D
CD is the dimensionless storage coefficient, and is
defined as where, m is the slope per log cycle, s is the usual
skin effect and Ate was defined earlier by Eq. (17)
• C
D
= 5.615 C 2
2n^hctrw
(25)
m = 162.6 qµB
kh
(28)
Although no proof is demonstrated, their type

curves may be converted for analyzing pressure (AP)buildup is the left hand side of Eq. (22) and
buildup data if dimensionless time, t, appearing
both in y-axis and x-axis is replaced Dby At , and was defined earlier by Eq. (8). The form of
p on the y-axis is replaced by p as shown in Eq. (26) or Eq. (27) suggests that a graph of
Fyg. 11. In performing type curvewaD nalysis, the buildup pressure, pws or ('P)buildup vs- At, should
basic steps as outlined by Earlougher and Kersch4
remain the same except for some minor changes in the be linear on a semi-log graph paper. This will be
preparation of the data plot (Apbuildup )/Ate vs. Ate shown later by means of Fig. 16. The slope of
the line should provide the value of formation flow
should be plotted instead of plotting 6p/t vs t. capacity, kh. Note that the graph utilizing At is
Limitations on the lower limits of producing time, similar to the Horner graph because it also takes
t D, given by Eqs. (23) and (24) should also apply into account the effect of producing time, t.
iî this case. Moreover, this graph appears more general thpan the
Horner graph because the value of At increases with
15 the increasing value of shut-in time, At as opposed
to the Horner time group (t + At)/At where it
Recently, Gringarten et a1.15 presented decreases as at increases. PThis permits plotting of
another set of type curves in a form different than buildup data on the same time scale using At and ete
that of Agarwal et al.1 and Earlougher and so that the effect of including or excluding the
Kersch4. Pressure drawdown data have been plotted producing time can be compared. Eq. (17) also indi-
as p D vs. (tD/C ) with CDel as a parameter. Their cates that for long producing times, when
draw`ôwn type cuKes are schematically shown in (tP + At)/tp = 1. Eq. (17) reverts back as
Fig. 12. To convert their type curves for pressure
buildup data, p D is changed to and the param-
eter tD/CD on tîe abscissa should be replaced by
At e /CD. In using their type curves for buildup, Ate = At (29)
(Ap9 buildup vs. Ate should be plotted on the data
• plot.
same.
Steps for type curve matching remain the
The limitations on the lower limit of pro-
ducing time as discussed earlier should also apply
in this case.
Eq. (29) also provides the basis for making a
MDH plot for long producing times. Eq. (26) may be
solved for the skin effect, s as
SPE 9289 RAM G. AGARWAL
•
pws(Ate = I hr) - pws(4t = 0) Fig. 14 shows a comparison of drawdown data
s = 1.151[ [plotted as (Ap)drawdown vs. t and shown by trian-
m
gles] with the pressure buildup data [(np)buildu
vs. At ](shown by solid circles). The comparison
e
between the two plots is excellent. (Ap)ûi^d
log - k 2 + 3.23] (30) data plotted as a function of conventiona s u^Pin
0 ctrw time, At, are also shown by a dotted line with open
circles. Note that between 200 and 250 minutes
there is a departure between the conventional
Note that in Eq. (30), p(At = 1 hr) should buildup curve and the other two curves. Gringarten
be read on the semi-log straight line or its exten- et al.ls observed a similar departure between the
sion. conventional buildup curve and the drawdown curve at
about 250 minutes and concluded that the buildup
The initial reservoir pressure, p., or a false data beyond this time should not be analyzed by
pressure, p'^, can be directly read from the straight drawdown type curves. However, the modified buildup
line portion of the semi-log graph [p
ws vs. Ate] plot does not suffer from the above restriction. If
where At is equal to producing time, t. Inspec- the new method is used, the majority of data may be
tion of eEq. (17) indicates that this cgrresponds to type curve matched. Fig. 14 also suggests that in
the Horner time ratio, (t +At)/At equal to unity or this case, the real shut-in time, At, of 4281
shut-in time, A close to ?nfinity. The estimation minutes is only equal to (1347 x 4281)/(1347 + 4281)
of initial reservoir pressure by this method will be = 1025 minutes in terms of equivalent drawdown time,
illustrated later by means of a field example and Ate
will be shown on Fig. 16.
Next (Ap)huildu vs. Ate data were type curve
FIELD EXAMPLE matched using Gringagten at al'sls type curve as
shown in Fig. 15. Note that a very satisfactory
Pressure Buildup Analysis Using New Method match has been obtained. Computations for type
curve analysis are shown in Table 2 and results sum-
A field example taken from Gringarten marized in Table 3. It is also possible to read the
et al.'sis paper will be utilized to illustrate initial pressure directly from the log-log plot. To
the application of the new method to analyze pres- accomplish this, read (Ap) at At = t . This
sure buildup data taken on an acidized well. Both
conventional and type curve methods will be used to
analyze the data. Results will be compared with
those of Gringarten et al.'s To maintain the con-
provides p. = p(At=O)+(pp)buildup
pi = 3251 psi, as shown onbPiga95.
In ^.his Ease
To demonstrate applicability of the new method

•
tinuity, part of the information appearing in their to conventional semi-log analysis, buildup pres-
paper will be reproduced here. sures, p were plotted on a semi-log graph paper
both as asfunction of conventional shut-in time, At
Table 1 lists the pertinent reservoir and well (shown by open circles) and the equivalent time, At
data, along with pressure-time data both during the (shown by solid circles). This is shown in Fig. 16e
drawdown and buildup periods. Fig. 13 is a graph As expected, there is a significant difference
showing well pressures both during the constant rate between the plots. In a way it is similar to com-
drawdown [vs. flowing time, t] and during the paring a MDH plot with a Horner plot. However, the
subsequent buildup [vs. (t + pt)]. Buildup data new method is better because data can be compared on
were replotted using the n9w time group, an equivalent time scale. It also provides a rea-
(tp x At)/(tp + At) or Ate. On Fig. 13, they sonable straight line, whose slope was used to com-
pute formation flow capacity, kh, as given by
are plotted as a function of [t +pt ]. Note that Eq. (28). Eq. (30) was used to compute the skin
there is a significant differenge between the two effect, s. It is also possible to directly read the
buildup curves. Data plotted simply as a function initial pressure from the semi-log straight line or
of shut-in time, At (shown by open circles) appear its extension where At = t . Results of both con-
flatter compared to the second buildup curve plotted ventional semi-log andetypepcurve analyses are
vs. At and shown by solid circles. This is to be listed in Table 3. For comparison purposes, anal-
expected because the first curve does not take into ysis results obtained by Gringarten et al.'s are
account the effect of producing time. More will be also shown in Table 3.
said about this later.
Note that excellent agreement has been obtained
Although initial reservoir pressure was not between the conventional semi-log and type curve
known for this problem, it is possible to estimate methods when the new method is used. Moreover,
it by means of an expanded plot (not shown here) of results also agree very well with that of Gringarten
early time buildup and drawdown data and recognizing et al.ls when they used the Horner method for the
that a graph of [(Ap) buildup vs. Ote] is equivalent semi-log analysis and the desuperposed data for type
curve matching purposes. Obviously, their MDH type
to [(pp drawdown vs. flowing time, t). Using the
results shown in Table 3 obtained by ignoring the
known value of (Ap)build up at a given Ote, the cor- effect of producing time (using semi-log or type
responding (Ap) drawdown at the same value of

flowing time, t, was estimated. This provided p. _
3251 psi, which was used to compute (Op)
flowing time, t.
drawdown
vs.
curve method) will be wrong, as expected. This was
also pointed out by Gringarten at al.'s In regard
to the desuperposition principle, it should be
pointed out that it is not always possible to desu-
perpose the buildup data because it requires a
•
knowledge of pressure vs. time data from the
preceding flow period. If the new method is used,
8 CURVES ARE USED TO ANALYZE PRESSURE BUILDUP AND OTHER TEST DATA SPE 9289
• desuperposition of data may not be necessary.

Although not shown, the new method may be used to
[Pwf2(tl + At) - Pwfl(tl)] vs.
desuperpose pressure buildup data. ql

^ql - q2
EXTENSION OF NEW METHOD TO OTHER KINDS OF TESTING
I • At
It appears that the new method may be extended [(l+At)
to analyze other types of testing such as two-rate
and multiple rate tests in an infinite radial
system. Once this method is used, data may be ana- obtained during the second rate testing, Also note
lyzed by both the conventional semilog method and that Eq. (33) reverts back to the pressure buildup
type curve matching techniques. Eq. (22) when q is set equal to zero. This is to
2
be expected in view of the fact that a pressure
Two Rate Testinga'16'17 buildup test is a special case of two rate tests
with q2 = 0
A schematic of two rate testing with rate and
pressure history is shown in Fig. 17. This type of For a two rate testing, let us define equiva-
testing consists of flowing a well at a constant lent drawdown time, IIte2 as
rate, ql, for time, t1, when the rate is changed to
q2 during the incremental time, At. The flowing
pressure, pwfl(tl) at the end of the first flow ql
rate, q1, can be obtained from Eq. (21) as
ql - q
t
•Ot (34)
Ate2 = (11 At
kh[Pi - pwfl(t1)]
141.2 NB
Substituting Eq. (34) in Eq. (33) and
expressing it for conventional semilog analysis, we
= 2 ql[ln(tI)D + 0.80907 + 2s] (31) obtain
• Note that the above equation assumes the log

approximation. By applying the superposition prin-
ciple, we obtain the following equation for the =
[Pwf2(tl + At) - Pwfl(tl)]
162.6 (qI - q2)BN
kh - [1og(Ate2) +
flowing pressure, pwf2(tl + pt) during the second
flow rate, q2.
log k 2 - 3.23 + 0.87s) (35)

kh[pi-pwf2(tl + At)] ONctrw
141.2 NB
Eq. (35) suggests that a graph of [p 2(t +
Ate2ws6ould
At) - pwfl(t )J or pwf2(tl + At) vs.
= 2 ql[1n(tI + pt)D + 0_80907 + 2s] be linear with slope, m. Formation flow capacity
is computed as
- 1 (ql-q2)[ln(OtD) + 0.80907 + 2s] (32) 162.6(qI -q2)Bp

2 kh = m md-ft (36)
Subtracting Eq. (32) from Eq. 31, we obtain

and skin effect, s, is computed using
kh[Pwf2(t1 + at) - Pwfl(tl)]

141.2(q1 - q2)NB s= 1.151[Pwf2(Ate2 = 1 hr) - pwfl(t1)
m
(ql
l q2 - log k 2 + 3.231 (37)
= 2 [ Ln ^t + ^t (At)D ^Nctrw
0 I + 0.80907 + 2s] (33) An example will be shown later for multiple rate
testing.
Comparison of the above equation with Eq. (21)

indicates that a plot of [p l . - p (t)J vs. t during
a pressure drawdown test isequivalent to plotting
of data as
SPE 9289
Multiple Rate Testing8i17'ls

RAM G. AGARWAL
and Eq. (40) is expressed for the skin effect, s,

as follows:
9
•
A schematic of multiple rate testing is shown
in Fig. 18. This type of testing consists of
flowing a well at a constant rate q l for time, tl,
s = 1.151[pwfn(Ate n-1 hr) - pwfn-1(tn-I)
at rate q2 for time t] to t2 and so on. Say the
m
final rate is q for lime t to any incremental
time, At. Although not shown; pressures are denoted
as p (t ), p f (t2).... and p (t ) at the end
timewfn-1 n-1
of first , Isecond2and t periods. p (At) - log k 2 + 3.231 (42)
are the pressures during2the final (nth) period. If
^Nctrw
the steps similar to those shown for two rate
testing are followed for multiple rate testing, the
following equation is obtained. Next a computer simulated example will be con-
sidered to illustrate the application to multiple
rate data.
kh[pwf.n(Ot) - pwfn-1(tn-1)1
SIMULATED EXAMPLE
141.2 (qn-1 - qn)BN
Multiple Rate Analysis Using New Method
qj-1
n-1 To demonstrate the application, a computer gen-
qjn- 1 -
q n
erated example will be utilized. Table 4 lists the
1 tn-1 - tj-1
2{ln At)D reservoir and well data for a gas well, where (Nc )
^t + tn-1 - t 1 product is kept constant to eliminate the effectsgof
J
j=1
(Nc ) variations as a function of pressure. Rate
andgpressure histories are shown on Fig. 19. Note
+ 0.80907 + 2s} (38) that the gas well was produced at three different
flow rates (15, 10 and 5 MMCF/D) with intermediate
buildup periods. Each flow period was simulated to
where t= 0; q = 0 and n> 2. Eq. (38) is very be 1/2 day long whereas each of the first two
general°and should apply to any number of flow and
•
buildup periods was 1 day long. The third and the
buildup periods, provided that the system is final buildup period was 2 1/2 days long. Pressure
behaving like an infinite radial system and log vs. time data for each flow period are shown on
approximation is valid. Eq. (38) also suggests that Fig. 19 by means of open symbols, whereas for each
multiple rate test data during any flow or buildup subsequent buildup period by means of corresponding
period may be analyzed using drawdown type curves. solid symbols. Multiple rate test Eq. (40) was
For multiple rate testing, the equivalent drawdown expressed in terms of real gas pseudo pressure19,
time may be defined as m(p) and rearranged slightly for the purpose. In
Fo* n eaae aha.h9ts practical gas units, the following equation is
obtained:
At =
en
ji
( __ m[pwfn(4t)1 - m[pwfn-1(tn-1)}
n- l
qn I qn
(qn-1 - q n )
tn-1 - tS-1
.At (39)
TT ^t + tn-1 - j - 1
j=1 k
17T[log(Aten)
kh + log ^(Nct
)irw2
For conventional semi-log analysis, Eq. (38) - 3.23 + 0.87s] (43)

may be written as
Eq. (43) was used to process the pressure vs.

time data obtained for each test period. Results
[pwfn(At) pwfn-1(tn-1)]
are shown on Fig. 20 using the same symbols as shown
162.6(qn-1 on Fig. 19. The left hand side of Eq. (43) denoted
qn)BN
[log(Ote1) + by [dn(p)/Aq] vs. the equivalent drawdown time, At
kh
have been plotted on a semilog graph paper for eacîn
test. Note that it was possible to normalize all
test data on the semi-log straight line obtained
log k 2- 3.23 + 0.87s1 (40) using single rate drawdown data. Although details
are not shown, the slope of the semi-log straight
^Nctrw
line provided the value of formation flow capacity
(kh) which was consistent with the kh value entered
Eq. (40) also suggests a linear relationship on a
semi-log paper with slope, m where
162.6(qn-1
into the program, as expected. The preceding
example was used to demonstrate the application and
establish the validity of the new method for mul-
tiple rate testing data.
•
qn)BN
kh = md-ft (41)
m
r 1
Other Kinds of Testing Finite Flow Capacity Fracture
Although not shown in this paper, it appears The new method was next applied to data for a
that the new method may be applied to other kinds of vertically fractured with finite flow capacity frac-
testing methods such as interference, constant pres- ture. Constant rate pressure drawdown data of
sure testing, etc. Agarwal, et al.,7 were used to generate a family
of buildup type curves with producing time as a par-
VERTICALLY FRACTURED WELL ameter. This had to be done for each value of
dimensionless fracture flow capacity. These results
The new method was next applied to vertically were replotted using the new method. Once again, it
fractured wells with both infinite and finite flow was possible to normalize the majority of buildup
capacity fractures. Results are discussed below. data on drawdown type curves. For the sake of
Eq. (5) again fgrms the basis of this study. Dimen- brevity, results are not presented here. However,
sionless time, Dx f for a fractured well is defined it should be suffice to say that constant rate pres-
as follows: sure drawdown type curves of Agarwal, et a1.7 and
Cinco, et al,s, may be utilized to analyze pres-
sure buildup data. Requirement is that (Op)b ldup
data are plotted as a function of Ate rather ^îan
t = 2.634 x 10-4 kt
(44) the conventional shut-in time, At.
Dxf 2
O(Nct)i xf ANALYSIS OF GAS WELL BUILDUP DATA
The development of the new method, for ana-

Where x is the fracture half-length in feet_ lyzing pressure buildup data, has been discussed
Definitions of real and dimensionless drawdown and mainly utilizing solutions for liquid systems. How-
buildup pressures were kept the same. ever, it appears that the method may be extended to
include the analysis of data from gas wells, if real
Infinite Flow Capacity Fracture gas pseudo-pressure m(p) of Al-Hussainy, et a1.19,
is used and variations of (pc ) vs. pressure are
Gringtrten et al.'s5 pressure drawdown data accounted for. The latter ma§ be accomplished if
(p D vs. Dxf for the infinite reservoir case were real times in the new time group are replaced by
ta^Cen from their Table 1. Eq. (5) was used to gen- real gas pseudo-time, t(p) of Agarwa114. For
•
erate a family t of pressure buildup curves with pro- example, if pressure buildup data collected after
ducing times, pDx , as a parameter. Results short producing time from an t9iF gas well are to be
similar to those of Raghavan are presented in analyzed by drawdown type curves, the following
Fig. 21. Since Raghavan13 adequately discussed the procedure is recommended.
limitations of using pressure drawdown curves for
analyzing pressure buildup data for fractured wells,
only certain key points will be re-emphasized- Graph {m[pws(t + At)] - m[pws(At=0)]j vs.
P
(i) Computed formation flow capacity will be (tap x Ata)
optimistiC.
(t + Ota) on
ap
(ii) Computed value of fracture length will be
pessimistic.
the data plot, utilizing the appropriate type curve.
(iii) The characteristic half slope line may Steps outlined in Ref. 14 for type curve matching
not appear on the log-log paper. remain the same. In the above time group, t and
At represent flowing time, t, and shut-inaFime,
Fig. 22 shows the replot of pressure buildup Atapexpressed in terms of real gas pseudo-time. If
data utilizing the new method. Data are plotted as variations of (pc ) vs. pressure during the test
period appear to gbe small, instead of using pseudo-
pw s vs. ^€Dxf or (tpDxf xAtDxf)/(tpDxf +atDxf). time, real times may be used.
Note that eDxf is the equivalent drawdown time,
expressed in the dimensionless form, for a verti- CONCLUDING REMARKS
cally fractured well. The majority of buildup data
have been normalized on the drawdown curve. It was 1. A new method has been developed to analyze
rather a surprising observation in view of the fact pressure buildup data by pressure drawdown type
that a time group developed for the radial system curves. It provides a significant improvement
should also be applicable for a fractured well which over the current methods because
is normally associated with linear, elliptical and
radial flow regimes. (i) the effects of producing time are
accounted for;
Although not included in this paper, the pres-
sure drawdown data of Gringarten, et al.,s for the (ii) data are normalized in such a way that
uniform flux fracture case were also considered. instead of using a family of buildup
Pressure buildup data were generated and plotted curves, the existing drawdown type curves
• using the new method. Once again it was possible to

normalize the majority of buildup data on the draw-
down type curve. The plot was very similar to that
shown in Fig. 22.
may be used;
(iii) wellbore storage and damage effects may

be considered except under certain condi-
tions.
SPE 9289
2. This method can also be used to perform

RAM G. AGARWAL
conventional semi-log analysis to estimate for-

m = slope of the semilog straight line,
psi/log cycle (kPa/log cycle)
11
•
mation flow capacity, kh, skin effect, s, and
initial pressure, p.. It appears similar to m(p) = real gas pseudo pressure, psiz/cp
the Horner method because both methods take (kPa2/Pa•s)
into account producing time effects. However,
this method is more general and has the advan- = initial pressure, psi (kPa)
Pi
tage that
= dimensionless pressure drop [see
pwD
(i) both the MDH plot and the plot using the Eq. (1)]
new method utilize the common time scale,
which permits comparing the two plots and
determining the effects of including or
pwDs = dimensionless pressure rise or
excluding the producing time; change ( see Eq. (19)]
(ii) it provides a relationship between the = wellbore flowing pressure, psi

pwf
flowing time, t, during a drawdown test (kPa)
to an equivalent time, At , during a
buildup test. e
pwfn-1 = pressure at the end of test period,
tn-1, psi (kPa)
3. For long producing periods, the new method
.reverts back to the MDH method.
Pwfn = pressures during nth test period,
psi (kPa)
4. A field example is included to demonstrate the
application of the new method and point out the = shut-in pressure, psi (kPa)
pws
utility.
= pressure change during buildup, psi
(AP)buildup
5. This method has been extended to include the (kPa) [see Eq. (8)]
analysis of two rate and multiple rate test
data by both type curve and conventional = pressure difference, psi (kPa) [see
(AP)difference
methods. This was shown both theoretically and Eq. (10) and (11(]
6.
also by means of example problem.
Although originally developed for radial sys-

tems, this method appears to work well for ver-
tically fractured wells with infinite and
(Ap)drawdown
q
= pressure change during drawdown,
psi (kPa) [see Eq. (7)]
= flow rate, STB/D or MCF/D

•
finite flow capacity fractures. ("standard" m3/D)
7. The method, although developed using liquid q = flow rate during the final flow
solutions, should be applicable to data from n period, STB/D ("standard" m3/D)
gas wells, as shown in the paper.
rw = wellbore radius, ft (m)
8. Finally, it appears that the new method may be
applied to a variety of testing methods such as s = skin effect
interference testing and constant pressure
testing to name a few. t = flow time, hours
NOMENCLATURE t n-1 = total time up to (n-1) test period,

hours
B = formation volume factor RB/STB
(res m3/stock tank m3) ^ tD = dimensionless time based on well-
bore radius (see Eq. 2)
cg = gas compressibility, psi' (kPa 1)
= real gas pseudo time (see Ref. 14)
ct ta(p)
= total system compressibility, psi 1
(kPa 1) t = real gas pseudo-producing time
ap
C = storage coefficient, RB/psi (res. tDx f = dimensionless time based on half
m3/kPa) fracture length (see Eq. 44)
CD = dimensionless storage coefficient t = producing period, hours

[see Eq. (25)] p
t 1 = producing period during the. first
FCD = dimensionless fracture flow p test,, hours.
capacity (see Ref. 7)
k
= formation thickness, ft (m)
= formation permeability, md
t pD
At
= dimensionless producing period
= shut-in time or incremental time

during the final flow test period,
hours
•
log = logarithm to base 10
At = real gas pseudo shut-in time
a
ln = natural logarithm
• t ap = real gas pseudo producing period 2. Wattenbarger, Robert A. and Ramey, H. J., Jr.:
"An Investigation of Wellbore Storage and Skin
at e = equivalent drawdown time, hours Effect in Unsteady Liquid Flow: II. Finite
(see Eq. 17) Difference Treatment," Soc. Pet. Eng. J. (Sept.
1970) 291-297; Trans., ATME, 249.
AtDxf. = dimensionless shut-in time based on
fracture half length 3. Ramey, H. J., Jr.: "Short-Time Well Test Data
Interpretation in the Presence of Skin Effect
At = equivalent drawdown time for two and Wellbore Storage," J. Pet. Tech. (Jan.
e2
rate test, hours (see Eq. 34) 1970) 97-104; Trans., AIME, 249.
At = equivalent drawdown time for mul- 4. Earlougher, Robert C., Jr., and Kersch, Keith
en
tiple rate test, hours (see Eq. 39) M.: "Analysis of Short-Time Transient Test
Data by Type Curve Matching," J. Pet. Tech.
T = reservoir temperature, °R (July 1974) 793-800; Trans., AIME, 257.
xf = fracture half length, ft (m) 5. Gringarten, Alain C., Ramey, Henry J., Jr., and
Raghavan, R.: "Pressure Analysis for Fractured
N = viscosity, cp (Pa • s) Wells," paper SPE 4051 presented at the SPE-
AIME 47th Annual Fall Meeting, San Antonio,
(Nc t ) i = viscosity-compressibility product Tex., Oct. 8-11, 1972.
at initial condition, cp
6. Cinco-L, Hever, Samaniego-V., F. and Dominguez-
^ = formation porosity, fraction A., N.: "Transient Pressure Behavior for a
Well With a Finite-Conductivity Vertical Frac-
it = constant, 3.14159 ture," Soc. Pet. Eng. J. (Aug. 1978) 253-264.
Subscripts 7. Agarwal, R. G., Carter, R. D., and Pollock,

C. B.: "Evaluation and Prediction of Perform-
a = adjusted or pseudo ance of Low Permeability Gas Wells Stimulated
by Massive Hydraulic Fracturing," J. Pet. Tech.
CD = dimensionless flow capacity (March 1979) 362-372; Trans., AIME, 267.
q D = dimensionless 8. Earlougher, Robert C., Jr.: Advances in Well

Test Analysis, Monograph Series, Society of
Dxf = dimensionless, based on x f Petroleum Engineers, Dallas (1977) 5.
e = equivalent 9. Miller, C. C., Dyes, A. B., and Hutchinson,

C. A., Jr.: "The Estimations of Permeability
eD = equivalent, dimensionless and Reservoir Pressure from Bottom-hole Pres-
sure Build-Up Characteristics," Trans., AIME
e2 = equivalent for two rate test (1950) 189, 91-104.
en = equivalent for multiple rate test 10. Horner, D. R.: "Pressure Build-Up in Wells,"
Proc., Third World Pet. Cong., The Hague (1951)
f = fracture II, 503-521.
g = gas 11. McKinley, R. M.: "Wellbore Transmiscibility

From Afterflow-Dominated Pressure Buildup
i = initial Data," J. Pet. Tech. (July 1971) 863-872;
Trans., AIME, 251.
j = index
12. Crawford, G. E., Pierce, A. E., and McKinley,
n = index R. M.: "Type Curves for McKinley Analysis of
Drill-Stem Test Data," paper SPE 6754, pre-
p = producing sented at the SPE 52nd Annual Fall Meeting,
0
Denver, Colo., Oct. 9-12, 1977.
t = total
13. Raghavan, R.: "The Effect of Producing Time on
w = wellbore Type Curve Analysis," J. Pet. Tech. (June,
1980) 1053-1064.
14. Agarwal, Ram G.: "Real Gas Pseudo-Time - A New

1. Agarwal, Ram G., Al-Hussainy, Rafi, and Ramey, Function for Pressure Buildup Analysis of MHF
H. J., Jr.: An Investigation of Wellbore Sto- Gas Wells," paper SPE 8270 presented at SPE
rage and Skin Effect in Unsteady Liquid Flow: 54th Annual Fall Meeting, Las Vegas, Nev.,
• I. Analytical Treatment," Soc. Pet. Eng. J.

(Sept. 1970) 279-290; Trans., AIME, 249.
Sept. 23-26, 1979.
SPE 9289 RAM G. AGARWAI. 13
15. Gringarten, A. C., Bourdet, D. P., Landel, 18. Odeh, A. S. and Jones, L. G.: "Pressure Draw-
P. A., and Kniazeff, V. J.: "A Comparison down Analysis, Variable-Rate Case," J. Pet.
Between Different Skin and Wellbore Storage Tech. (Aug. 1965) 960-964; Trans., AIME, 234.
Type-Ruves to Early-Time Transient Analysis"
paper SPE 8205 presented at SPE 54th Annual 19. Al-Hussainy, R., Ramey, H. J., Jr., and Craw-
Fall,Ueeting, Las Vegas, Nev., Sept. 23-26, ford, P. B.: "The Flow or Real Gases Thrqugh
1979. Porous Media," J. Pet. Tech. (May 1966)
624-636; Trans., AIME, 237.
16. Russell, D. G.: "Determination of Formation
Characteristics From Two-Rate Flow Tests,"
J. Pet. Tech. (Dec. 1963) 1347-1355; Trans.,
AIME, 228.
17. Matthews, C. S. and Russell, D. G.: Pressure

Buildup and Flow Tests in Well, Monograph
Series, Society of Petroleum Engineers of AIME,
Dallas (1967) 1, Chapter 5.
is
0
• TABLE 1
TABLE 2
RESERVOIR AND WELL DATA
(Field Example - Ref. 15) TYPE CURVE ANALYSIS
(Field Example - Ref. 15)
Match Point
Formation thickness, h 30 ft
Formation porosity, ^ 0 .15 fraction PV
[6p)M = 10 psi [PwD) = 0.17
Wellbore radius, r 0 .3 ft
1 .0 cp M
Fluid viscosity, Nw
System compressibility, ct 10 X 10-6 psi-1
[6te] t, = 10 min [D)
cL = 0.64
1 .25 RB/STB D M
Formation volume factor, 8
Production rate, q 800 STB/D [cDe2s[
1347 minutes = 1.0
Producing period, t p
M
(i) Formation flow capacity, kh

Drawdown Buildup (t p X At)
fAt)-p ws(pt=0) ♦ A [PD] (141.2) qBN
ati n pws pws (t
(min) ( psi Z (m ( ps1) P'Jps i) (t p
kh = M [^) md-ft
H
45 3198 3 3105 8 2.99 kh = (0.17)(141.2)(800)(1.25)(1)
85 3180 5 3108 11 4.48 10
192 3154 9 3115 18 8.64
kh = 2400 md-ft
297 3141 16 3125 28 15.61
417 3130 30 3139 42 29.74
654 3116 40 3146 49 38.6
(ii) Skin effect, a
993 3196 66 3159 62 62.6
1347 3097 100 3171 74 92.9 (^t e[
138 3180 83 125.0
C = (0.000295) h It /c i Res Bbl/psi
252 3195 98 210.5
106 269. 0 D M
334 3203
423 3208 111 321. _ (0.000295)(2400)(10/60)
119 396. Res Bbl/psi
574 3216 (1)(0.64)
779 3222 125 499.
612. C = 0.184 Res Bbl/psi
1092 3228 131
1674 3234 137 748.
•
Eq. (2S) is used to compute
2186 3238 141 842.
2683 3242 145 898. C = (C)(5.615) _ (0.184)(5.615)
3615 3246 149 962. = 40600
D 2rtphctrw2 2n(.15)(30)(IX10-5)(.3)2
4281 3246 149 1036.
De2s)
S= 2 An [(cc MJ = 2 en [40821)
D
= -5.31
TABLE 3
SUMMARY OF RESULTS
(Field Example - Ref. 15)
Analysis Using This Method TABLE 4
Semilog Type Curve RESERVOIR AND WELL DATA

(Simulated Example)
kh (md-ft) 2323 2400
a -5.15 -5.31
C (Res Bbl/psi) -- 0.184 Initial reservoir pressure, pt . 500 psi
pi 3253 3251 Reservoir temperature, T 720°R
Formation permeability, k . 5 ad
Formation thickness, h
40 feet
Grinaarten et al.'s Analysis Hydrocarbon porosity, ^
5%
Viscosity-compressibility product, (ye)
t 3.93 X 10-6 cp/psi
Semilog Type Curve Wellbore radius, r
w 0.25 feet
Horner MDH Desuperposition HDH
kh (md-ft) 2274 4279 2259 4095

s -5.0 -4.0 -5.1 -4.0
C (Rea Bbl/psi) -- -- 0.19 0.25
pi 3253 3230 --
•
• '
Pi
1 ^ I)drar,davn pWS(tp+
pwf(t) ^ """"P I PRESSURE

1^^ P) dIfferl
pwt ltp)•pWS I ^t-0
p^(tp+ 6t)
STANT RATE DRAWDOWN BUILDUP
t 6t
TIME TIME
FIGURE l: SCHEMATIC OF PRESSURE BUILDUP BEHAVIOR FOLLOWING

A CONSTANT RATE DRAWOMN FOR A PRODUCTION PERIOD, tP
FIGURE 2: SCHEMATIC OF PRESSURE BUILDUP BEHAVIOR FOLLOWING CONSTANT RATE

DRAYOOYN OF SUCCESSIVELY INCREASING FLOW PERIODS, tp
8
• .
7
4
p)drawdo7
a 6 DRAWDOWN
------^ 1;
m SOLUTION
-_-__ 4 5 _ --•--- - -- 104
+p Q -R'" R- ^
----- 103
4 __.
o-- "----
^ ------
^ 3 2
^p)buildING
-' 0
, tp BUILDUP SOLUTION /p
2a 2
TIME, t or pt
Tn ^> I
10-° 26t
DIMENSIONLESS BUILDUP TIME, ptD • 2.634 x
0N ct rW
FIGURE 4: BUILDUP TYPE CURVES FOR VARIOUS PRODUCING TIMES- SEMI-LOG GRAPH
(INFINITE RADIAL SYSTEM, CO • 0 8 S. 0)
FIGURE 3: SCHEMATIC SHOWING COMPARISON BETWEEN [(DP)drawdown vs t]
AD [(dP)bulldup vs 6t]
•
E
8 7
DIMENSIO
PRODUCING
7 6
4 a,^ tPD
6 DRAWDOWN ^ 5
a m SOLUTION
BUILDUP DATA
5 FOR VARIOUS ^ 4 sA
PRODUCING TIMES, t 9
pD
4 3
J tn
104
5• N 3 ° 103 2
2a^ 2 102 1 1 SA
c
1 0
102 10j 104 105 t x6 t 106
DIMENSIONLESS EQUIVALENT TIME, OteD • I t'+ ^D ) DIMENSIONLESS BUILDUP TIME, 6t0
pD D
FIGURE 6: NORIMLI2ED BUILDUP TYPE CURVE-SEMI-LOG GRAPH FIGURE 6: BUILDUP TYPE CURVES FOR VARIOUS PRODUCING TINES-SENI-LOG GRAPH
(INFINITE RADIAL SYSTEN, Cp 0; s. 0) (INFINITE RADIAL SYSTEN, Cp 1000; s•O)
•
10
-.------- • •
DRAWDOWN ^105
SOLUTION 10
V1
^W 5 BUILDUP DATA
1 SA , a--------°---°------ ^03 OC FOR VARIOUS
a
^ ^. 4 sA PRODUCING TIMES, tPD
_ ` s
m 3 •^• • 106
(n 0. 1 Vf 105
4A
^ 2 X 104
^BUILDUP ^ DIMENSIONLESS
fA SOLUTION PRODUCING TIME In ° 103
9 A
01 C •16 S °
102 103 104 105 7

102 103 104 105 106 ltl
DIMENSIONLESS BUILDUP TIME, 6tD
DIMENSIONLESS
DIMENSIONLESS EQUIVALENT TIME, "'OD
FIGURE 8: N004ILI2ED BUILDUP TYPE CURVE ( SEMI-LOG GRAPH)

(INFINITE RADIAL SYSTEM, Cp 1000: 5-0)
FIGURE 7: L06-LOG BUILDUP TYPE CURVES FOR VARIOUS PRODUCING TIMES
(INFINITE RADIAL SYSTEM, Cp 1000; 5b)
•
N10 '
9
DRAWDOWN m
SOLUTION r--
7
PRODLICING
U
c
COeZ6 • 101,
C6 a^
C6 SA CL
1014
9
GO
10$-,
tj 0.1 106104 ---"

e 10Z J
U 10
3 ^
a
0.0
10' 103 104 105 10b
DIMENSION LESS EQUIVALENT TIME, AID
1 D 1 'teD
0.000295 C OR 000295 C
D dr^dam D buildup
FIGURE 9: NORMALIZED BUILDUP TYPE CURVES - LOG-LOG GRAPH
FIHIRE 10: A GEIIERALIZED SCHEMATIC OF EARLOUONER AND KERSCH
•
(INFINITE RADIAL SYSTEM, Cp 1000. S-0)
DRAYOONN TYPE CURVES4
U 1
NIP
C e 2s.
aIQ
0 10-1
N 0
^ rY
t6 4
O
10
10_2
11
or teD (kh (^t or ntB md.R hr

[. CD ] `Y^ \ c ^ Cp ' bbllpsi
FIGURE 11: A SCHEMATIC OF EARLOUfiIER AND KE0.5pl4 DINNOOi01 TYPE CURVES FOR A WELL WITH STORAGE AMD SKIN
(INFINITE RADIAL SYSTEM)-By PERMISSION Of MARATHON OIL CC1hJ+Y
•
!
1
4--
N
iL[
rî
j 6
43 O
N
W
N
N
y,^ C.'
G.
K
a
N 0
Q
N
00
C TIME MINUTES)
DIMENSIONLESS TIME 1 t0or Otel • 0. 000295 ^^^ ^t or
CD
FIGURE 12: A SCHEMATIC OF GRINGARTEN Lt Li115 DPAYDOYN TYPE CURVES FOR A WELL WITH STORAGE AND SKIN FIGURE 13: WELL PRESSURES Vs TIME DURING PRESSURE DRAHDOid AND BUILDUP PERIODS
(BY PERMISSION OF FLOPETROL) ( FIELD EXAMPLE - REF. 15)
• 200
180
EXTRAPOLATED (0 pi - 154 psi
AT pte•tp•1347Min.
ORpi •3097+ 154•3251psi
160
ew
^ 140 ^.i
2 i" K ♦ "I iy''Ì ^t _
100
120
r ------ ---^
c 10D HI_ -
U DEPARTURE BETWEEN
80 TWO CURVES
pl ^ ' DPibuiidup Vs ^ t-
e ,
N
3^ • i
W 60
a 6 P)drawdown vs. t 10 ^.. -I
a 40 1 • (Lp)buildup vs. ^te 1./^' , ,
0 N - - 1 C D ^^ CDO^
--L----__- ,-.
,
1
P) buildup vs. 6t
0.1
20
W
10
d
TfYY1 91YY1 ^(YI(1 d(YY1 500(1

TIME (MINUTES) 1
1 10 100 1000 2000
MATCH POINT pte MINUTES)
0 1
( n p)M 10 psi; (pwD)M • 0.17
1 o VM - 10 min; ItDICD)M-0"64
(COeZs)M• 1.0
FIGURE 14: COMPARISON BETWEEN DRAYDOMN AND BUILDUP DATA USING NEW METHOD FIGURE 15: APPLICATION OF NEW METHOD USING GRINGARTEN Et Al TYPE CURVES
(FIELD EXAMPLE - REF. 15) (FIELD EXAMPLE - REF. 15)
q
ttF
•
Ke ql
33A) 9 q2
3300 • PWs Vs . o te
^ 3280 ^ pWS Vs. dt
EXTRAPOLATED PRESSURE FLOW TIME, I, HR.
3M Pi • 3253 psi^
USING NEWMETH00
3240 ^ INITIAL PRESSURE - PRESSURE DURING
kh - 2323 md-ft o°o ^ __,E- SECOND RATE
a ^
^ 3n0 s - 5,31 ' v+ -
SLOPE 00 K P,.,n Ra')
3200 o
70 ps ilcycl e
5 ^ G*
3180
9 3160 9
m 3140 6t e -t P 1341 min.
3120 ^ FLOW TIME, t, HR.
310D ^
3080
1 10 1 103 10
SHUT IN TIME, At or pt e (MINUTES)
FIGURE 16: APPLICATION OF NEW METHOD TO CONVENTIONAL SEMILOG-ANALYSIS

(FIE LO EXAMPLE - REF. 15) FIGURE 17: SCHEMATIC RATE AND PRESSURE HISTORY FOR A TWO-RATE TESTING, q> i
ti N l^1
n a
3-9-3-9-3 9
9 ^ 9 ^ 9 S ^
15
q5
q qn 10
q2 ^ j
q_ u`c
5
d
q3 D
o<e
3
1 I
LL I I
1 I
q4 1 I ^t Q
^
a
e-0 tl t2 t3 t4 t5 n-2 tn-1 W
TIME, t, HOURS ^
^
d
FIGURE 1e: SCHEMATIC OF MULTIPLE RATE TESTING
0 1 2 3 4 5 6
TIME (DAYS)
FIGURE 19: RATE AND PRESSURE HISTORY FOR S114JLATED EXAMPLE
0
•
..,...I . . . ..... ^ , , , .,.
X
o FLOW I
"7-
_21L • BUILDUP 1
e 5 a FLOW 2
LJ ^ BUILDUP 2
e FLOW 3
20 & BUILDUP 3
TIME (DAYS)
-- ^
INFINITE RESERVOIR ( xeJX M)
FIGURE 20: N0P}IALIZED MULTIPLE RATE TEST DATA (SIMULATED EXAMPLE)

DRAWDOWN
SOLUTION
6S
i•
a _
-------- -
9
lU - T BUILDUP SOLU7IQAL-
? G
co
INFINITE RESERVOIR lxelxt • W)
^
z
ce 0
N
1 DRAWDOWN
n^. SOLUTION
CL i .i
9 I in
DIMENSIONLESS BUILDUP TIME, pt ^
f
Ln 10 i
LH
z
0
^
FIGURE 21: BUILDUP TYPE CURVES FOR VARIOUS PRODUCING TIMES-LOG-LOG GRAPH
(VERTICALLY FRACTURED WELL WITH INFINITE FLOW CAPACITY FIWCNRE)
1D -2
10 10 1 10 102
DIMENSIONLESS EQUIVALENT TIME, 'teD
FIGURE 22: NORIMLIZED BUILDUP TYPE CURVES ( LOG-LOG GRAPH)

(VERTICALLY FRACTURED WELL WITH INFINITE FLOW CAPACITY FRACTURE)
0
r'+
1:'
^^
• RESERVOIR ENIGINEERING
How wellbore dynamics affect pressure

transient analysis
L. MATTAR and M. SANTO
Fekete Associates Inc.
Calgary, Alberta
ABSTRACT theoretical than practical. In this presentation the authors will ad-
There are numerous technical papers on pressure transient analy- dress some of the many issues under the over-all umbrella of
sis which address both simple and complex reservoir-related "Wellbore Dynamics". These reflect the experience with "practi-
phenomena. These papers all make the implicit assumption that cal" welltest interpretation, which stems from years of specializa-
reservoir pressure can be measured directly. It is a fact of life, tion in this field, (more than 10 000 tests have been analyzed in
however, that the pressure recorder is located in a wellbore, not detail by the authors and their colleagues).
in the reservoir. The weltbore is an intrinsic link between the reser-
voir and the recorders. Although the pressures recorded in the well-
bore are normally representative of the pressures in the reservoir,
Weiibore Dynamics
they can also be affected by a number of wellbore related The topics which will be addressed in this paper will obviously
phenomena. deal with wellbore phenomena. However, not all of these well-
The many wellbore related phenomena that can have a signirl- bore phenomena will be studied. For example, we will not be
cant effect on the measured pressure have not been addressed in reviewing the effects of temperature on wellbore fluids or pres-
• the literature, except for wellbore storage and the classical phase sure recorders; nor will we address such topics as gas/oil solu-
redistribution hump. tion/liberation or retrograde condensation. Moreover, some of the
This paper presents several examples of tests that have been qf- ideas discussed do not take place "in" the wellbore, but in the
fected by wellbore dynamics, and shows that these could easily immediate vicinity of the wellbore, and often, their effects are in-
have been misinterpreted as complex reservoir phenomena (dual separable from weUbore effects. The "Wellbore Dynamics" that
porosity, etc.) instead of wellbore effects. These effects are often we shall illustrate will be grouped into the following topics:
accentuated by the (semilog) derivative which is traditionally used Liquid Influx/Efflux, Phase Redistribution, Wellbore (and near-
to diagnose reservoir characteristics. The fact that these are well- wellbore) Clean-up, Differences between Drawdown and Build-
bore and not reservoir effects must be recognized to prevent mis- up, Plugging, Recorder Effects, Mysterious Effects.
diagnosis, Often this can only be done from examination of the Many, but not all, of these "wellbore" transients occur at very
test data other than the recorded pressure-time trace. early time. Their detection and documentation has been facilitat-
ed by the advent of electronic pressure recorders. These recorders
are so accurate that they will detect these wellbore effects very
Introduction clearly, and can easily be misinterpreted as "Reservoir" effects.
There have been more than 1000 papers published on Pressure Indeed, the whole thrust of our field of engineering is to interpret
Transient Analysis. These publications address in great detail such the "Reservoir", by using pressures recorded in a "Wellbore".
fundamental topics as homogeneous infinite reservoirs, fractures, This often places the Reservoir and the Wellbore in a tug-of-war,
convolution, dual porosity, multi-layers, or such esoteric concepts and we find that, very often,"Wellbore" Transients will distort,
as super-bilinear-equivalent-pseudo-time or integral type curves. or even, dominate our "Reservoir" Transients. This is not sur-
With the progress made in both analytical and numerical solu- prising in view of the fact that in this tug-of-war, the Wellbore
tions to reservoir problems, and with the ready availability of has a distinct advantage over the Reservoir because this highly sen-
powerful computers, virtually every imaginable reservoir descrip- sitive recorder is sitting in the Wellbore and not in the Reservoir.
tion can be modelled and the solution presented in graphical form
(type curve).
In contrast, there are very few papers that deal with wellbore Liquid Influx/Efflux
effects. The only two classical wellbore topics that have been In this section, the authors will illustrate that the pressure trend
studied in any detail, are the concepts of Wellbore Storage0) and observed with the recorders does not necessarily reflect the pres-
Phase Redistribution(2). Even then, the treatment has been more sure in the reservoir. All the well test theory treats pressure as
measured at the sandface. In many cases, it is impossible to land
recorders at the sandface or even at the mid-point of perforations
Keywords: Wellbore dynamics, Liquid movement, Phase redistribution, (MPP). In some cases the recorder run depth (RRD) is above MPP
Recorder, Cleanup, Drift, Practical, Transient, Well testing. as shown in Figure 1, or at MPP in Figure 2. In the wellbore
Paper reviewed and accepled for publication by the Editorial Board of The Journal of Canadian Petroleum Technology.
32 The Journal of Canadian Petroleum Technology

WELLBORE SCHEMATIC 1
YELL80RE SCHEMATIC 2
•
RRD ABOVE MPP
RRO - MPP
L FIGURE 1. RRD above MPP. FIGURE 2. RRD at MPP.
schematic of Figure 3, the RRD is at MPP, but because of the

"U-tube" effect caused by the tubing being below the perfora-
tions, the same sandface pressure trend could appear very d{Jfer-
ently on the recorders of Figure 2 and Figure 3. This will happen
every time that there is more than one phase in the wellbore, which,
MELLBORE SCHEMATIC 3
•
in the testing of oil and gas wells, is often the case.
For the reservoir engineering equations to be applicable, the
"recorded" pressures must obviously be converted to "sandface"
(reservoir) pressures when the recorders are set either above or
below the producing interval. This correction is performed by add-
ing or subtracting the appropriate hydrostatic head of fluid (gas,
oil or water). In the case of tubing below the perforations, Figure
3, it is very important to recognize that the fluids inside and out-
side the tubing may be different, and separate corrections must
be done inside and outside the tubing. In performing the pressure
correction from RRD to MPP, the effects of friction have been
ignored to simplify the presentation, though in high rate wells,
Tubing BELOW FlPP
this may cause a significant error.
(U-tube)
In the simplest case of a single phase fluid and a wellbore
schematic, as in Figure 1, the correction from RRD to MPP is:
Pressure (MPP) = Pressure (RRD) + P (Hydrostatic)
P (Hydrostatic) = H x gradient
where H is the vertical distance between RRD and MPP
gradient is 1 Y,Pa/m for gas
5 kPa/m for oil (live)
10 kPa/m for water
In this situation, as seen in Figure 4, the pressure response at
RRD is parallel to that at MPP.
When there is more than one phase present in the wellbore, and
the liquid interface is moving up or down, there can be signifi- FIGURE 3. Tubing below MPP.
cant differences in the pressure trends observed on the recorder
(RRD) as compared to what is really happening at the sandface r^
(MPP). We shall illustrate this using the simple case of a "Chang-
ing Liquid Level" in a gas well with water in the wellbore. The sure at the sandface (MPP) is increasing in a straight line, as shown
well is producing gas and water to surface, and is shut-in for a by the line labelled MPP in Figures 5, 6 and 7. The liquid (water)
build-up test. Assume (for the sake of illustration) that the pres- level is above RRD and is falling, which will occur at different
February 1992, Volume 31, No. 2 33

^ >t>,
Ir^crur^ (1r1) • Irc,c^n (NIID)

CHANGING LIQUID LEVELS
Gas = 1 kPa/m Liquid = 10 kPalm RRD = t MPP = n
n ► (W^nftae^cl
7000
^ Q +-r eNo
fr11
♦
kPa MPP
c1 (Rycnftltlc) - N I In11rK ^ 6000
1000 kPa
rrul^nt . 1 t111a • Pi ^
{-L uv1 RRD
. I vrP - oil
r )O Vth • wtrr
5000
ti t2 t3 ^
Y..
_- A p=100x10=1000kPa
1(RIID)
JI'^ ' t
1111D ^
FIGURE S. Liquid above RRD.
n,
7000
r,(Y/ ► 1 ♦
kPa
MPP 55500 a
1kP
ti. 6000
FIGURE 4. Pressure correction from RRD to MPP - single
phase. [RRD
5000
t1 Q t3
rates depending on permeability and damage ratio. At time t1, it
reaches RRD, at time tz it is halfway between RRD and MPP ep•(50x1)+(50x100)=550kPa
(Fig. 6) and at time t3 the liquid level has reached, or is below
MPP. In Figure 5, the pressure difference between MPP and RRD FIGURE 6. Liquid halfway between RRD and MPP.
is a constant, equal to the hydrostatic head of water. It stays con-
stant until the liquid level reaches the RRD. From then on, the
difference between MPP and RRD varies and equals the sum of 7000
♦
an increasing gas column plus a decreasing water column. Figure
6 illustrates when the liquid is halfway between MPP and RRD. kPa
MPP
When the liquid level is at, or below MPP, the correction from 100 kPa
MPP to RRD is the hydrostatic head due to a column of gas. M
It is obvious, as is illustrated in Figure 7, that whereas the
response recorded on the pressure recorder at RRD shows a sig- RRD
nificant deviation from linearity, and could be (mis)interpreted as
5000
a reservoir phenomenon, the truth is, that the pressure trend at
MPP was perfectly straight, and the recorded effect was a well- ti t2 t3
bore phenomenon, pure and simple. A rising liquid level would
exhibit the opposite trend to that shown in Figure 7. Ap^100x1- 100 kPa
Is this "ivory tower" thinking, or is the phenomenon real?
From the thousands of well tests that we have analyzed, these FIGURE 7. Liquid at MPP.
effects have been observed on approximately 75% of well tests
where more than one phase is flowing. Here are a few exam-
ples of very typical situations:
1. Oil wells flowing oil and gas to surface. At the end of a 2-week "efflux" causing the liquid level in the tubing to fall. Note that
shut in for a build-up test, a static gradient is conducted and (due to density segregation), during the first 5 days, only water
shows no liquid in the wellbore. Where did all the oil in the effluxes from the wellbore (into the reservoir?), and after that,
weilbore go? only oil is effluxing. Figure 9, on the other hand, shows that dur-
2. Gas wells producing at high water-gas ratio: Once again, at ing the shut in of an oil well, only water "influxed" into the well-
the end of a build-up test, the static gradient shows not a bore] ('The first and last static gradients are questionable-the first
drop of water above the perforations. one was taken a few hours into the build-up and the gas/oil/water
3. Acoustic well surveys (AWS): These very often indicate both levels are still unsettled).
rising and falling liquid levels. Admittedly, AWS are subject Figure 10 illustrates a gas well which produced at a high water-
to a lot of interpretation problems, but the over-all trends gas ratio. The point at which the water level falls below the record-
of rising or falling liquid levels are clearly illustrated (quite ers is clearly seen at 100 hours, and at 108 hours the liquid level
independently of such problems as collapsing foam columns, -has reached the sandface. The pressure difference (approximate-
misinterpreted reflections, etc.). ly 300 kPa) corresponds to the distance between RRD and MPP
Even though we have developed "rules of thumb" respecting multiplied by the difference between a water gradient and a gas
• the phenomena to be expected, it is very hard to predict, a priori,

whether, for a particular test the liquid level will rise and/or fall,
or both, at what rate. Figures 8 and 9 show measurements of liquid
levels obtained with static gradient surveys conducted every 3 or
gradient.
How do these effects appear when performing pressure tran-
sient analysis? The type curve and derivative plot of a typical falling
liquid level test is shown in Figure 11. The derivative has the shape
4 days on two different oil wells. Figure 8 clearly illustrates liquid of the classical "Dual Porosity" reservoir, yet is caused only by

600 500
750 1^j
800 - 1000
,. 1250
-E, E
> >
^ 1000 1500
Cr
-' 1750
1200 2000
2250
E
1400 2500
r M CD a) N t0
1 4 8 12 14 r r
Time (days) Time (days)

0 Gas Oil 0 Water Gas Oil M Water
FIGURE 8. Falling liquid level - Ertlux. FIGURE 9. Rising liquid level - Infiux.
liquid level changes in the wellbore. Figure 12 illustrates the full ing liquid level. It also illustrates that, sometimes, the subsequent
set of pressure data on the same well as Figure 11, while Figure reservoir pressure response does give analyzable semilog straight
13 illustrates the parallel semilog straight lines often seen with such lines.
tests. Figure 17 shows the first seven hours of a build-up in an oil
well. The two tracks shown are for two electronic pressure record-
ers set for 30-second interval readings. Except for the slight off-
Phase Redistribution set in time, the recorders match exceptionally well, confuming that
The phenomenon of phase redistribution also known as "hum- some event is actually taking place and what is being recorded
ping"has been well documented in the reservoir engineering liter- is not a spurious aberration. It is the authors contention that what
ature(2,3). It is distinct from the liquid influx/efflux situation we are observing is phase redistribution on a smaller scale than
described previously. It represents pressure changes (sometimes sig- the humping effect - gas bubbles, coalescing into slugs and bub-
nificant and dramatic) caused purely by the redistribution of the bling through the oil column below the recorders. Whatever the
gas and liquid phases in the wellbore. An interesting explanation true nature of this wellbore dynamic, we do not believe that what
of the cause of the "humping" effect is given by S ►ider(4), to we are observing is a reservoir effect. Rather, we believe it is purely
which text the interested reader is referred. In this section, we will a wellbore effect which has totally masked the reservoir response.
provide three examples of phase redistribution effects, Figures 14,
15 and 16. These illustrate that the phase redistribution hump can
be of the same magnitude (Fig. 14) or even significantly larger Wellbore and Near Wellbore Cleanup
(Fig. 15) than the subsequent reservoir pressure response. Figure Consider the flow and build-up test shown in Figure 18. We have
16 shows the combined effect of phase redistribution and chang- two tests on different reservoirs in different parts of Alberta, that

144!
-,An 41-1- 11- M

.
`. . .
a•aaiaa
FIGUItE 13. Parrallel line of failing liquid level.
- .. . a..
• rc...^..t aw.• o...
•
.w •
01 ar1.HIn
.,.
a .... '
.»
10,
e
...
'." . ^. . . . ,. » ^ » »
K. M
10
FIGURE 14. Phase redistribution - small hump.
lo" io'• lo• lo' to
FIGURE 11. Dual porosity or 7
sm
^,. I I I^ ,.I-. li l l I i l I il l li ,l. I l l .l I l i !.

FIGURE 15. Phase redistribution - large hump.
-^.
:^}!
dient survey at the end of the build-up showed a dry wellbore.
FIGURE U. Falling liquid level of Figure 11. The pressure depletion of some 300 kPa was purely a wellbore
effect. The reservoir pressure at MPP before and after the test
did not change at all.
Iliere are some effects that take place in the immediate neigh-
looked like Figure 18. They both show a depletion of some 300 bourhood of the wellbore that are often observed during tests.
kPa between the initial pressure and the extrapolated (Homer) Even though these are not truly wellbore effects, they distort the
.,^
build-up pressure. One of them was a "good news" story, the reservoir pressure data in the same way and can easily lead the
other was "bad news"! The "bad news" well was obviously deplet- analyst to a misdiagnosis of the reservoir response. The "bad
ing during test and hence had very small reserves. The "good news" case in Figure 18 clearly shows depletion, hence a limited
news" well had the recorders landed some 30 m below the perfo- reservoir. Yet, during the drawdown neither the rate nor the pres-
rations. Initially, there was some liquid (completion fluids) in the sure is declining; in actual fact, the flow rate was increasing. The
wellbore - as determined by the static gradient survey before the drawdown information is contradicting the build-up results. The
,; .
test. During the test this liquid was blown out, and the static gra- explanation is that during the drawdown, the near-wellbore for-

•
f IM .M. )l -- . 0t, r4.^

n N^/Cos .^1.^ 2 --• !) W/u.Kr
6 00 0
IvN
p 000
190 -
tID00
3 ^000
7C00
Ii70
rer ywr,.^
sow N ' ^r
r aa ^, • 0.^ rwcrd
"M - - - OTT F1-=
AL
um
00 w q RO IUO 200 7+0 700 370 340
t1110 _j Iw l1r)
L-
11 L
tO L
lf7 11 1^ W Ql QI Q7 Q4
,1.. ^ ►^
FIGURE 17. Phase redistribution - gas bubbling through oil? FIGURE 20. Increasing wellbore damage.
T
-------- - -- - -
,^c ...
^ - t- - -
GOOD NEWS I
BAD NEWS T
a^
r^
a0w
^ r loe w w ar s^ a.
...1 c.....,
FIGURE 18. Depiction or what?
a steeply declining pressure in spite of a decreasing gas flow rate.

This drawdown behaviour is usually caused by either a very tight
mation is continuously cleaning-up and the corresponding decrease reservoir or a depleting system, but the build-up in this case shows
in skin is masking the anticipated pseudo-steady pressure decline. that neither is the case. In actual fact the drawdown is reflecting
A more dramatic example is shown in Figure 19 in which the rate the increasing skin caused by water coning in the near-wellbore
and flowing pressure are both increasing! This near-wellbore area. The logs showed 1 m of gas over water and the production
to
cleanup can last several weeks. In six tests on wells that were frac'd showed a 25 fold increase in water-gas ratio.
and tagged with identifying tracers, frac fluids were still being
produced after three weeks of flow.
The reverse phenomenon has also been observed as shown in Differences: Drawdown and Build-up
Figure 20. In this example the build-up data indicate a good permo• The previously described cases will obviously show a distinct differ-
ability well with no depletion, while the drawdown data exhibit ence in analysis between the drawdown and the build-up. Figures

i. 3 .i,-U H .,-) y lY/JUn ..., 7
.F'.r
• ,a+
+° +
^+ f
IQ- m - - l.° ,a+ +o +o w•

^daoi•>
FIGURE 22. Build-up data. FIGURE 24. Response shown on 4th recorder.
^-^
r--,... •.
rr^ rr^ r^ rr^ ra^ rr^ 2C r C. nV. W. r_

ra• ^,âr.^
FIGURE 23. Response shown on 3 recorders.
21 and 22 are examples of a drawdown and build-up from the Plugging

same test. The data and derivative curves are very different in For one particular gas well test, there were four pressure record-
character. The drawdown shows a permeability of 19 mD and a ers placed in a carrier at the same depth. Three of the pressure
skin of 31 while the build-up shows a permeability of 2 mD and charts looked like Figure 23 and the fourth onb looked like Figure
a skin of -3. During this test, the well was producing frac sand 24. These charts were alike throughout the test, except for the last
and frac fluid left over from a stimulation treatment. Due to the flow period in which throe recorders show an increasing pressure
frac sand production the choke was washing, thus causing a gradu- while the fourth recorder showed a decreasing pressure. Our final
al increase in choke size throughout the drawdown. At the same analysis of the wellbore configuration showed that the three record-
time, the cleanup of frac fluids was causing a continuous change ers were "inside" recorders while the fourth recorder was an "out-
in the skin. In this instance, the drawdown analysis is not valid, side" one.
and the build-up analysis provides the best estimates of the reser- In fact, the throe inside recorders had been affected by hydrates
voir flow characteristics. plugging during the last flow period, and the only true reservoir
Build-ups are often analyzed using drawdown typecurves with response was from the "outside" recorder. We have observed simi-
equivalent time or using some form of superposition (convolu- lar situations in waxy oil wells, in which a recorder port hole gets
tion) of the drawdown equations. For oil wells, one should ex- plugged with wax.
pect the early time response in drawdown and build-up to be
significantly different because of the different wellbore conditions.
In a drawdown, one can often start the flow from the condition Recorder Effects
of the wellbore full of a single phase fluid (hence a particular well- With the ever increasing use of electronic pressure recorders, the
bore storage coefficient). However, in a build-up, in the majority authors have observed phenomena that can only be ascribed to
of cases, the wel,lbore will contain a significant amount of free recorder (mis)behaviour. Figure 25 shows a sudden jump in pres-
?ti gas (and hence have a completely different wellbore storage coeffi- sure when the sampling rate is changed - its companion record-
cient). Moreover, the phenomena taking place in the wellbore are er did not show this effect. Figure 26 depicts two recorders (I in
6 completely different - in drawdown we have expansion of well- apart) located in an observation well during an interference test.
bore liquid, in build-up we have gas being compressed and (pos- Note the unexplained significant pressure difference between the
sibly) going into solution throughout the build-up. (This is like two recorders. There are numerous examples of tandem record-
squeezing toothpaste back into the tube!) In build-up tests for oil ers (1 in apart) not tracking each other, some diverging, others
wells one would expect wellbore effects to last for a lot longer crossing each other, while others tend to have a sinusoidal wave
^ period. Drawdown tests, on the other hand, are affected by gas- pattern. Electronic recorder drift is recognized but has not been
liquid slugging and variable liquid holdup and often, the data is quantified, and it is suspected that in some high permeability reser-
not analyzable. Because of these significant differences between voirs the drift of electronic recorders is of the same order of mag-
drawdown and build-up wellbore behaviour, tests of short dura- nitude as the reservoir response. This means that a recorder
tion are particularly susceptible to misinterpretation. characteristic can be analyzed as a reservoir characteristic.

Mysterious Effects ----.._-._.._.._..-----^-------.._.._.._..--------
The two recorders shown in Figure 27 are sitting in an observa- 1,
tion well, and show a very erratic performance. When we looked /M b0
at the first recorder, we thought that it was malfunctioning. 3„

However when we examined the second (and third and fourth) 61« ,.^
recorders, they showed an identical behaviour. These two record-
ers are not of the same type. One is a strain gauge, the other quartz w e
crystal. What are all these spikes that we see? Note the pressure ar t^ o â
spike around 400 hours. There are several hundred data points "«
down one side and up the other side of this spike, and the magni- ,, ^
tude of this pressure aberration is 40 kPa (6 psi), 50 to 100 times
the resolution of the gauges. This "detour"in pressure takes place
„« wl
over some 20 hours. We have no idea what causes these effects
^,r.r W* ,asr :x.• zt+.r ^:,.- .. -.:^.
•
and can only guess that they might be micro-seismic effects (we
,^ ^-Itsl fi^t CN^1a WYCYW MM.rynl,ilry^ 4^
could not find any documented earthquakes at the 400-hour time
frame!). We have observed geo-tidal effects in other situations;
FIGURE 23. Mysterious pressure effects.
they have a regularity to their pattern and do not look like Figure
26. In Figure 28 we present one of two recorders (identical be-
haviour on both) during a flow and build-up test in an oil well.
Notice the "cilia" (hair-like attachments) during the drawdown response, the wellbore has the unfair advantage of having (a) in-
and the build-up, except between points A and B.The regularity finite transmissivity and (b) immediate proximity of the pressure
of this pattern is fascinating. The size of a typical spike is 30 kPa recorder. This means that in most situations wetlbore effects will
(5 psi) and the interval between spikes is approximately 30 minutes. dominate over the reservoir effect.
The sampling frequency was such that there were some 20 data This state of affairs is unfortunate in that wellbore dynamics
points for each of the spikes. We have no idea what causes this, can mask or completely obliterate the reservoir response, and
cilia behaviour, nor why it is absent between points A and B, nor as a result, can easily lead to a misdiagnosis of the response.
why it appears so consistent during the drawdown which is ex-
pected to be in turbulent (hence erratic) flow. (As we write this,
it occurs to us that this cilia behaviour could easily be caused by Conclusion
a casing gas regulator - a back pressure valve that opens and The presence of Wellbore Transients must be recognized and taken
closes intermittently to release casing gas!). into consideration when performing a (Reservoir) Pressure Tran-
sient Analysis. All too often, analysts work with a set of pressure-
time data, oblivious of the practicalities of test operations. It is
Tug-of-war Wellbore vs Reservoir our hope that, in this paper, we have illustrated some of the
Of the many wellbore related effects that we have observed, only anomalies that can easily be (mis)interpreted as reservoir charac-
a few have been presented in this paper. Many of the effects have teristics. To avoid these pitfalls, the analyst must be familiar with
logical explanations while others are still a mystery. Many of these the operation of surface and bottom hole test equipment, field
effects occur at early time (Fig. 17), but they are certainly not con- conduct and practices, and details of the test performance. We
fined to early time (Fig. 26). We have documented cases of chang- estimate that approximately 50°/o of the well test analyst's time
ing liquid level effects a full 2 weeks after shut-in. should be spent examining the raw data and conducting validity
The primary function of welltest analysis is to determine the checks and reconciliation of all the data. The following procedures
reservoir characteristics, and not to study the wellbore (even though must be performed routinely, before or simultaneously with, any
some wellbore coefficients can be calailated). The best way to study pressure transient analysis:
the reservoir is to place the recorder inside the reservoir. Unfor- 1. examine wellbore configuration - packer, sliding sleeve,
tunately that is not possible, and the pressure recorder must be recorder depth, tubing depth, perforation depths, etc.
placed inside a wellbore that is in communication with the reser- 2. compare all the pressure recorders for inconsistencies; do not
voir. This means that any wellbore pressure effects will be felt in- rely on one recorder when one or more backup recorders are
stantaneously by the recorder and will be accentuated in used.
comparison with the more diffuse reservoir response. In the con- 3. adjust pressures to mid-point of perforations based on static
tinuous tug-of-war between wellbore response and reservoir gradient surveys conducted before and after the test -

adjustments may vary with flow rate and time. ous effects as far as reservoir analysis is concerned) and wc may
be at the point of not being able to "see the forest for the trees".
4. compare bottom hole pressure behaviour with surface tubing
and annulus pressures.
NOMENCLATURE
5. examine flow periods for constancy and accuracy of: flow rates,
gas-oil ratio, water-oil ratio, gas-water ratio. H = Vertical distance between RRD and MPP
^
MPP = Mid-point of Perforations
6. if a large capital investment depends on the results of a test; RRD = Recorder Run Depth
calibrate the recorders before and after the test to confirm the P = Hydrostatic pressure difference between RRD and MPP
integrity of the recorders.
7. maintain constant dialogue with the field supervisor to resolve REFERENCES
any anomalies in the data (leaks, operations, not reported in 1. AGARWAL, R.G., AL-HUSSAINY R., and RAMEY, H.J., JR.,
the field notes). 1970, An Investigation of Wellbore Storage and Skin Effect in Un-
8. conduct pressure transient analysis taking into consideration steady Liquid Flow: 1. Analytical Treatment; Soc. Pet. Eng. Jour-
all of the above practical aspects in conjunction with all of the nal, Yol. 10, pp. 2 79-290.
relevant reservoir engineering/geological factors. 2. FAIR, W.B., JR., Pressure Buildup Analysis with Wellbore Phase
Redistribution Effects; Soc. Pet. Eng. Journal, pp. 259-270, April
Wellbore Dynamics have been ignored for far too long. They
1981.
must now be fully incorporated into any analysis. With the arrival 3. STEGEMEIER, G.L. and MATTHEWS, CS., 1958, A Study of
of extremely sensitive pressure recorders, we may be approaching Anomalous Pressure Build-up Behaviour; Trans., AIME, 213, pp.
the point of "information overload" - 30 000 to 50 000 data 43-30.
points per recorder per test is not uncommon. The recorders are 4. SLIDER, H.C.. Worldwide Practical Petroleum Reservoir Engineering
so sensitive that the response is dominated by "noise" (extrane- Methods, PcnnWell Books, Tulsa, Oklahoma, pp. 204-211.
Corrections
Corrections on two errors in the paper, "A Type-Curve 0.32
Matching Procedure for Material Balance Analysis of + p6 .................................(14)
Production Data from Geopressured Gas Reservoirs", by
Anil K. Ambastha, published in the September-October 1991
issue of the JCPT, are as folbws: 2. Equation (A-3) should read as:
1. Equation ( 14) should read as:

p [ 1 - C (p, - P}] = ^ ^ ° , G
... (A-3)
Z zt r; G
0.066
B=(0.62-0.23 Ta)Pa+r - 0.037,P2
`(Ta-0.86) ^
• Explosive record at Suncor

Suncor Inc., Oil Sands Group set new records
in October 1991, when its blasting crews suc-
vate the oil sands. Blasting is especially help-
ful ci0.uing the wirder montin bocaave it looeens
drilled that were 48 in. in diameter and 85 ft
deep. Sunoor's blasting crews were assisted by
cessfully shot what is believed to be the larg- ground that would otherwise be bound by their spetialty contractors, Mining Services In-
est singk charge in a blasthok on a commercial frost. ternational and McGinnis Rath* Drilling Ltd.
scale. Using the company's normal blasting proce- The work was completed successfully,
The record was set when a charge of 21 970 dures, only the top 55 feet of oil sands can Suncor Inc., Oll Sands Group uses mined
pounds was used in order to loosen an un- be blasted, but Sunoor needed to loosen an oil sand to produce a daily average of 63 000
usually deep section of oil sands. Normally, unusually deep deposit which varied in depth barrels of synthetic crude oil. The company's
a charge of 5815 pounds is used. from 75 ft to 95 ft. The company's standard blasting program has been instrumental in im-
Suncor uses blasting because it improves practice is to drill holes that are 30 in. in diam- proving mine efficiency and reducing excava-
mining efficiency by making it easier to exca- eter and 55 ft deep; for this job, holes were tion costs.
Alberta Research Council offers introduction to

neural networks for industry
The Advanced Computing and Enginoering plications. Those attending will gain an program the relationships in the examples ex-
Department of the Alberta Research Council understanding of neural networks and be able plicitly.
will offer an introductory seminar on neural to evaluate application of this technology to The registration fee of $95.00 includes
networks, Monday, February 25, 1992, from specific problems in their industry. course materials. To register by phone or for
9:00 am. - 12.-OQ noon at 6815 - 8 Street N.E., Serial computers are very good at tasks that more information, call Chantal Preston at
use clear step-by-step algorithm or logical (403) 297-7577.
"Neural Networks for Industry" introduces rules. However, intelligent reasoning includes The Advanced Computing and Engineer-
the structure of neural networks and how they perception of the real world and common ing Department of the Alberta Research Coun-
work, as well as some practical applications sense - the brain is very good at this. There cil is the largest applied research group in
in data analysis, pattern recognition and proc- is a growing belief that artificial neural net- Canada working in artificial intelligence, auto-
ess control. The seminar also covers future works also may be able to solve the oomputa- mation and robotics.
directions, communities of expert networks tional problems involved. New techniques for For more information, contact: Ken
• and hardware implementations.

This seminar is intended for researchers and
managers in industry who would like an in-
troduction to neural networks and their ap-
multilayered non-linear neural networks in-
elude procedures to learn from examples.
These procedures automatically generate in-
ternal representations, removing the need to
Gamble at (403) 297-7580, or Lynn Suther-
land, (403) 297-7572.

• • •
3600 MNL60-1: Cartesian Plot
3500 -- -
3400
3300
•^^-, 3200
CD
^ liytb t9- S e^ V6v`t^
N
tN
"W 3100 °
Q- 0
3000
2900
2800
270A
200 220 240 260 280 300;
Time (hours) __ _ _
I This plot shows the effect of fluid egression during a buildup test where the fluid level dropped below the position of the pressure gauge.
3520 MNL60-1: Cartesian Plot
3500
3480
-3460
a
Q)
N
N
Q)
CL 3440
3420
3400
338^
7265 270 275 280 285 290
__ Time (hours)
This is a magnified view of the effect of fluid egression upon the gauge pressure. This is a welibore effect caused when the fluid level drops
below the position of the gauge during the test.
49 0 0
SOLUTIONS
0 PRESSURE BUILDUP TEST
1. A Horner graph of the pressure buildup data is

presented in Fig. 5P.1. The data used to prepare this
graph are presented in the following table:
t +At
At, hrs pWS, psia
p At-
0.092 805 4033

0.23 323 4040
0.46 162 4045
0.92 81.4 4050
2.3 33.2 4056
4.6 17.1 4060
9.2 9.0 4062
• 23.0 4.2 4064
The slope of the Horner straight line is -16.7 psi/cycle,

i.e.,
m = -16.7 psi/cycle.
Therefore, from Eq. 5.8,
k = -162.6 mh
k = - (162.6) (460) (1.46) (0.75)

18 -16.7
k = 272 md.
•
5-155
4090
p* = 4,082 PSIA
4080
4070
4060
^
^ m = -16.7 PSI/CYCLE
4050
P1HR = 4,051 PSIA
4040
4030
1 10 10 2 103
t +At
P
At
Fig. SP.1: Horner graph, Pr6 1.

0
The total skin factor can be computed using
Eq. 5.17; first, however, the total compressibility is:
•
ct = coSo + c w S w + cf
cf = 3.3x10 6psi-1(Fig. A.4)
ct = (0.78) (20x10-6) + (0.22) (3x10-6) + 3.3x10-6
ct = 19.6 x 10-6psi .1
Also, plhr = 4051 psia (Fig. 5P.1).
Thus,
= 1.151 3803-4051
s
-16. 7
• - log - 272 6
(0.26) (0 . 7S) (19.6x10 ) (. 295) Z
+ 3.23
s = 10.6.
This positive skin suggests formation damage.

Average pressure is computed using the MBH
method. Because the reservoir is developed using a
square drilling pattern, it is assumed that the test
well is centered in a square drainage area. The
dimensionless producing time is
0.0002637ktp
tDA ûctA
_ (0.0002637) (272) (74)

(0.26) (0.75) (19.6x10-6) (80) (43,560)
• tDA = 0.40.
5-157
From Fig. 5•9' pDMBH - 2.53, so that
•
2 . 30 3 (pR-p*)
= 2.53.
m
Also, from Fig. 5P.1,
p * = 40 82 .
Therefore,
m
PR = p* + 2.53 (2. 303)
PR = 4082 + 2.53 (-16.7/2.303)
PR = 4064 psia.
It is observed that this value of PR is equal to the •

last measured value of pws; this agrees with the fact
that the data on Fig. 5P.1 appear to be stabilizing
at a constant pressure.
The pressure loss due to skin is calculated
using Eq. 5.22:
Aps = -0.87ms
= -0.87 (-16.7) (10.6)
Aps = 154 psi.
From Eq. 5.20, the flow efficiency is
E = PR pwf Aps
PR pwf
•
5-158
E _ 4064 - 3803 -154
• 40 64 - 3803
E = 0.41.
Thus, there is significant potential to increase the

rate of this well if the skin reflected by this flow
efficiency can be removed. However, further analysis
would have to be made to determine what part of the
skin effect is caused by formation damage.
•
5-159
!
2. (a) The wellbore storage factor caused by a rising

fluid level can be computed based on the wellbore
area exposed to the rising fluid level. In this case,
the annulus is packed off and fluid can only rise in
the tubing. Therefore, from Eq. 5.29,
144A
wb
C 5.615p
C = (144) (0.0218)
5.615 53
C = 0.011 bbls/psi.
Since the well primarily produces oil, it is assumed that

the wellbore storage factor due to fluid compression is
negligible in comparison to the rising fluid level.
Note that this calculation represents the required
method of estimating the wellbore storage factor when
• designing a test. After a test has been completed, the
storage factor used in analyzing the test should be
determined using the methods presented in (b) or (d).
(b) Log-log and Horner plots of the buildup data are

presented in Figs. 5P.2 and 5P.3. The wellbore storage
factor can be determined from the unit-slope straight
line on Fig. 5P.2 using Eq. 5.32, i.e.,
C= qBA t
24[pws-pwf(At=0)j
where At and [pws-pwf(At=0)] define any point on the

unit-slope line. The point selected has coordinates
of
At = 0.82 hours
pws pwf(At=0) = 800 psi.
5-160
1000
100
^ a
10
0.1 1 10 100
At, hrs
Fig. 5P.2: Log-log plot, Prob. 2.
0 0 0
• •
4600 + ^ „^ + i Vl : f :{
...
, + i1.,1 ,li '1,
;, r r• }t t
rT 1+ }
,. _
{, i m= -68 PSI/CYCLE
I
I+ `7 -U
4400 ; tr; .t}}1 },,;
^,;, ! , ,^.., ^ ti ,+•1l
. .. . . ., .,. .,..,{ ^ ....,. ..
,
;. } 1r
'1 ' {lit` i"' i1'
^.. l. i „C tf':, f ^ I^ I^`
1 T«
l r•'
^; Î
jlilr,^;^ , ^ ,. ,t}+
l 'i:''t^ t,' j• ,r ` t
IT'
t -^ t
. . .. ... . ' r i • • 1 ;1 , , , ., ! , , .I ^ l + , i . ,i f j ^ f ^, I ^ ^'+ ^

. . . ' . . .{, .1 -t• ^' 7 i , i , ^rj.; . ^ •^ , i, ^ i V , ^ { il ,l jii , ^^'• i ,•^^ •ri'r ! , i .,. .
^, , ^ 'i'f^lllf
111-.r..;#.., Iq ^ri ti l l,. i,.i .;
4200 4,293 PSIA
m I,; t, 1. ti I• p1^^ R
^ 1•+,. i it { {^ ±+s î tt 1 t '''
tt
: .. FG 1
^ ^ {i I
f ,lir- },^ , , jl +, I, }; 1'I t{I i rî• }' î I f^„ ^^ t } { t ±; + {r
^ , ,
;^ ^ , ,. ; .^. h
, •+ • ,, ^ , ^ , ^; i: t ^ ft
4000
„!
N +I { 1 t
. F• . I ^ ^ f ^•i 1 ^ ^ r ... . , ,• ; j t i , ,{ ^ I ^ r ft .:. 1,, ^ •r . ,, .
_I ' ^ - .
--^ -ra--^ ' +•»'^ r
. . • ^ ^ t f^ ^ ^,1 1^ I f i i{ {, ^ ., ^ , ', { +
...., j j,l ^ . 1 , ?
. ^ , t'rr , ^•,•^'1 ti• f ,; -$^ t t1, 1^ T' ^r tl 1 ! - - - ' 1 i
..... . ^, t „I , ^•^ C I ,^ ,
, _^
t ^r $' •,, ^ I ^ ^ -^ ^1^t ^-
^ I ^ i{. ^•I}
I
3800
-, . ; . }' : i ._ .I ^ , • t ,l r -.1 ' :;. ^ 1 ! ' + } ^' ^ r
f. r : , } { ,;•
... :a , ; :{ : .', . ' 1++ ; } ^^•, i, f ,^ ,
M• .++« 1 +} + i , Y r
}I t : '} ' {f ! ^ ^•• 1..'
f
^ . . : . ,. , ; ^
;t^ -+ +
{
3600
lOz 103 10`' i05
t +At
p
At
Fig. SP.3: Horner plot, Prob. 2.

Therefore,
C = (250) (1.136) (0.82

•
(24) ( 800 )
C = 0.012 bbls/psi.
This is in excellent agreement with the storage factor

in (a) and is a good indication that storage is being
caused primarily by the rising fluid level.
(c) From Fig. 5P.2, it appears that the unit-slope

line through the pressure data ends at approximately.
0.2 hours. Therefore, using the "50 At rule", storage
effects should end at a shut-in time of approximately
10 hours, or a Horner time ratio of 1,364. This is in
agreement with the derivative curve on Fig. 5P.2 which
appears to become horizontal at approximately 10 hours.
Further confirmation is given by Fig. SP.3A which shows
a match of the data on the Bourdet derivative type curve;
note that the derivative data approach the horizontal
derivative (p 61n = 0.5) on the type curve at approximately
10 hours, thereby indicating the end of wellbore storage
and the beginning of transient flow.
Consistent with the preceding observations, the
transient flow straight line is shown on Fig. 5P.3 with
a slope
m = -68 psi/cycle.
Thus, from Eq. 5.8,
k = -162.6 mh
qBl'
k = -162.6 (250)(1.136)(0.8)
k = 7.9 md.
•
5-163
WELL WITH WELLBORE STORAGE AND SKIN
INFINITE ACTING RESERVOIR WITH HOMOGENEOUS BEHAVIOR
We of this type-curve is described in World Oil-May 1983 : A NEW SET OF TYPE
nu.r,.a,a...a... •
^..i..:.
^ 9hsZ FOR OIL - PD kh Tw 2 P dp . ....c^
- PD GAS
' 141. LP 5.030 104q T Pu „ U(P)Z(P)
11
0
n
0
U
O
0
¢
c^
W to
¢
w
a
w
¢
Z)
N
U)
W
¢
a
tli o
z
W aa
w
¢
^
rn
w
¢ 1
a
2 '
î
Z
O
rn
z
w
0
10't
DIMENSIONLESS TIME GROUP (D/CD U`
Fig. 5P.3A: Type curve match, Prob. 2

The total skin factor is e^•aluatecl using Eq. S.17: •
P wf (At=0) -p lhr _ log

s = 1.151 k + 3.23
m ûctrw
s = 1.151 [3534 8 4293
7' 9
- log
(0.039) (0. 8) (17x10- 6) (0. 198) 2
+ 3.23]
s = 6. 7.
(d) Fig. 5P.3A presents a match of the buildup data on

the Bourdet derivative type curve. The match values are: •
zs
CDe M = 108
(dp)M = 100 (At) M = 10
(pD)hT = 1.5 (tp/CD)M = 180
Permeability is calculated using Eq. 3.87, i.e.,
k = 141. 2qBu (PD) M

h ,,
^ ^p TM
k = (141.2) (250) (1.136) (0.8) (1.5

69 100
k = 7.0 md.
0
5-163B
This is in good agreement with the value of 7.9 md
• computed from the semilog plot.
The wellbore storage factor, from Eq. 3.88, is
kh (A
C = 0.000295 u(tD -
CD -FM
C=(0. 000295) ( 7. 0) (69) 10

0.8 180
C = 0.01 bbl/psi.
From Eq. 5.30, the dimensionless storage factor is
C = 0.894C
D ^cthrW
_ (0.894)(0.01)
CD (0.039) (17x10- 6)(69)(0.198)'
• CD = 4,985.
The total skin factor can now be determined using Eq.

3.60:
1
S = Z (CDe2s)M
ln C
D
1 101
s = ln
[4,985 ]
s = S.
This compares with a skin factor of 6.7 from the

conventional analysis.
•
5-163C
• 0 •
1000 Prob. 5-2: Lo -Lo Plot
TF
a 100
(D
Quick Match Results

Radial homogeneous
Infinitely acting
Cs = 0.0105 bbUpsi
K = 8.1 md
S = 7.24
Pi = 4575.7167 psia
0
0.1 1 10 100
Equivalent Time (hours) - Tp=13630.0
This plot shows a history match of the test using an infinite-acting model. It is observed that the late-time data deviate from this model; it is
possible that these data are being affected by adjacent wells.
a-100
Ca
. *
Quick Match Resutts
Radial homogeneous
Ckased system - (L:L:L:Lj
Cs = 0.0105 bbUpsi
K =8.1 md
S -7.24
L = 1200 ft
Pi = 7846.8753 psia
0 0.1 100
1 10
Equivalent Time (hours) - T-13630.0
For the reservoir and wellbore parameters shown in the results box, this plot shows the theoretical behavior that should occur if this well were
centered in a square drainage area with no-flow boundaries at a distance of 1200 feet to each of the boundaries. Based upon the well spacing of
160 acres, it is expected that the boundaries would be approximately 1320 feet from the well; the behavior of the well implies, however, that the
boundaries are closer than expected and that the drainage area is less than 160 acres. This could be caused by heterogeneities, or by higher
producing rates at adjacent wells.
0 0 0
r
• • •
The purpose of this plot is to show that test data can be matched with a reservoir model that is incorrect. This well is surrounded by other
producing wells and, since there is no gas cap or water drive, it is known that all drainage boundaries are no-flow. This test was matched,
however, by assuming the well to be centered in a square with constant pressure boundaries at a distance of 660 feet from the well. THIS
SOLUTION IS INCORRECT!!! It is important to recognize that most well tests can be history matched with more than one model. Just because
you find a model that matches the behavior of your test does not necessarily mean that the model accurately describes your reservoir. Only
when your model is supported by information external to the test, i.e., seismic data, production data, production performance, etc., can you have
a high level of confidence in the results.
PROBLEM IPR-OIL: Oil Deliverability Prediction using:
0 (1) Vogel Method, and (2) Pseudosteady
Equation
The objective of this problem is to show how the production

performance of an oil well can be predicted using Pansystem after the analysis
of a pressure transient test has been completed. The analysis of the buildup
test described by Prob. 5.2, page 5-138, in your course manual is presented in
file IPR-OIL.PAN. Access this file and predict the inflow performance of
this well at the current average reservoir pressure using: (1) Vogel Method,
and (2) Pseudosteady State Equation. IMPORTANT: Make sure you
understand the assumptions made in each of these methods before using the
results of these predictions.
Additional Data:
Bubble Point Pressure = 3,280 psia

0
q 250 STB/D @ p f= 3,534 psia (from production period prior to buildup)
A 160 acres (based on well pattern and well spacing)
CA = 30.88 (based on well pattern) •
Estimated average reservoir pressure = 4,400 psia
0
PROBLEM IPR-OIL
SOLUTION:
The "IPR-Prod Test" curve, which is computed using Vogel's inflow performance relationship, is based on one
production point, I.e., the flow rate and flowing bottomhole pressure at the time the well was shut in for the
buildup test (if available, up to three production points can be used). It is assumed that the well had flowed
sufficiently long to be in pseudosteady state at the time this point was measured. It is also assumed that the
Vogel relationship is a valid model to describe the inflow performance of a well. The advantage of this method is
that it is based upon actual production performance, and does not require any Information about reservoir fluid or
• rock properties, reservoir size or geometry, or reservoir thickness. The disadvantage is that it 1s subject to
potential errors in the production test; for example, if the production test is not run sufficiently long to reach
pseudosteady state, this method will give optimistic results. Also, this test is based upon a semi-empirical
relationship by Vogel which may not accurately describe the production behavior of this reservoir.
The "IPR-Trans Test" curve is not based on a production test, but is computed using the theoretical
pseudosteady state equation. This method assumes that the reservoir conforms to the assumptions made in this
equation, i.e., homogeneous formation, constant compressibility liquid, constant viscosity, constant permeability,
etc. It further assumes that the permeability and skin factor used in the equation, which came from analysis of
the pressure buildup test, are correct, and that the drainage area, A, and shape factor, CA, which are based on
well pattern and spacing, are correct. Fluid properties and formation thickness used in the equation are also
assumed to be correct. The advantage of this method is that no production data are required. The disadvantage
is that the theoretical pseudosteady state equation must accurately describe the reservoir, and the rock
properties, fluid properties, drainage area, shape factor, and formation thickness used in the equation must be
known with reasonable accuracy.
0
Page 1 of 2
Deiiverabili Plot - IPR
4500 - IPR-Trans Test
IPR-Prod Test
4000
3500
3000
,-.
1 2500
^ 2000
a
1500
1000
500
0 0 100 200 300 400 500 600 700 800 900

Flow Rate STB/da
"IPR-Trans Test" , shown in the legend on this plot, refers to flow rates calculated using the pseudosteady
equation. The curve identified as "IPR-Prod Test" was computed using the Vogel equation.
Layer I Deliverability Data •

Layer 1
Layer pressure 4400.0000 psia
Bubble-point pressure 3280.0000 psia
Dietz shape factor 30.8800
Drainage area 160.0000 acres
Permeability 8.10 md
Skin factor 7.24
Productivity Index 0.2823 STB/day/psi
Absolute open flow potential (Oil) 830.1824 STB/day
Test Pressure 1 3534.0000 psia
Test Rate 1 250.0000 STB/day
Estimated Shut-in Pressure 4400.0000 psia
Test J 0.2887 STB/day/psi
Test AOF 848.893 STB/day
is
Page 2 of 2
gr
• PROBLEM: Pressure buildup test, Egypt
The following pressure buildup survey was conducted

June 13,_148?, on a well in Egypt. This well was originally
_, and had.produced 27, 707
^28P.199
placed on production Mav
stock tank barrels of oil at the time of the test. The well
is believed to be isolated within a fault block. The
reservoir was undersaturated at the time of the test. It
should also be noted that this well has a tapered flow string
and, accordingly, it is possible that the wellbore storage
factor will increase during the buildup test if the'fluid
level in the tubing rises above the point where the tubing
size increases.
You are requested to analyze this test to determine
permeability, total skin factor and flow efficiency. Is
there any evidence of a fault or other boundaries?
• q0 = 3,366 STB/D q'1 Awl.

^---^-.
h = 45 ft
= 23%
S = 10%
w
Bo = 1.666 bbl/STB C= ^w^
ljo = 0.28 cp
C f = 3.3 x 10-6 psi-1
cw = 2.8 x 10-6 psi-1
co = 5.6 x 10-6 psi -1 C = Vwb L^P` /z.d^d^^
rw = 0.2615 ft 5C^25
Depth of Amerada gauge = -9073 (s.s)
Datum = -9073 (s.s.)
• M
Ate'
1# i
At ate t + ^t A d A S
pWS p
d lnot e ^
hrs hrs At psia psi
0 0 - 3262 0 -
0.017 0.017 11857.0 3311 49 -
0.033 0.033 5929.0 3367 105 -
0.050 0.050 3953.0 3572 310 288.18
0.067 0.067 2965.0 3723 461 227.47
0.083 0.083 2372.2 3729 467 166.91
0.100 0.100 1977.0 3729 467 106.43
0.117 0.117 1694.7 3729 467 39.54
0.133 0.133 1483.0 3729 467 62.87
0.150 0.150 1318.3 3729 467 69.43
0.167 0.167 1186.6 3738 476 73.48
0.183 0.183 1078.8 3753 491 75.49
0.200 0.200 989.0 376S 503 87.12
0.217 0.217 913.0 3777 515 97.87
0.233 0.233 847.9 3789 527 102 . 29
0.250 0.250 791.4 3795 533 114.46
0.283 0.283 698.4 3807 545 106.06
0.333 0.332 593.8 3819 557 92.18
0.383 0.382 516.5 3843 581 79.40
0.433 0.432 457.0 3849 587 71.13
0.500 0.499 396.2 3852 590 65.42
0.583 0.581 339.7 3862 600 57.64
0.667 0.665 297.4 3867 605 41.45
0.750 0.747 264.5 3873 611 40.35
0.833 0.830 238.1 3879 617 42.82
1.000 0.995 198.6 3883 621 34.17
1.167 1.160 170.4 3889 627 33.66
1.417 1.407 140.S 3896 634 27.69
1.667 1.653 119.6 3898 636 27.21
2.000 1.980 99.8 3904 642 28.94
2.333 2.306 85.7 3907 645 29.11 •
2.833 2.793 70.7 3911 649 28.07
I.
3.333 3.278 60.3 3919 657 28. 86

^ 4.167 4.081 48.4 3925 663 33.45
6.000 5.823 33.9 3935 673 27.48
7.667 7.381 26.8 3946 634 31.17
9.333 8.912 22.2 3949 687 34.49
12.500 11.756 16.8 3954 692 35.99
15.833 14.658 13.5 3964 702 37. 79
20.000 18.162 10.9 3976 714 40.26
25.000 22.192 8.9 3982 720
31.667 27.293 7.2 3994 732
38.333 32.105 6.2 4000 738
42.950 35.281 5.6 4012 750
0
10,
U
4-J
102
'd
C^.
10
10-2 10-1 1 10 102
dte, hrs
0 0 0 .
^ • •
420
410
400
390(
^
380(
n^ 370C
3600
3500
3400
3300
1 10 102 103 10"
t + L^ t
At
3. a,b) Horner and log-log plots of the data are
presented in Figs. 5P.4 and 5P.5, respectively. The
log-log plot does not show a unit slope line; accordingly,
using the shut-in time corresponding to the first data
point, the maximum time that storage could last is
(50) (0.083) = 4.15 hours
which corresponds to a Horner time ratio of 10,572.

This is consistent with the observation that the first
straight line begins at a time ratio of 10,000.
It is observed that two straight lines are present
with slopes
m1 = -37 psi/cycle
m2 = -78 psi/cycle
or m2/m1 C^ 2.
•
This is consistent with our expectation that a fault is
near the well and seems to confirm the existence of the
fault.
The total reservoir flow rate is computed using
Eq. 5.49:
=o
aRt q0B0 + (1000 qgt - aoRs
^w)Bg + awBw
= (817) (1.19) + [(1000) (1512)
- (817) (330)]0.00196 + (2327) (1.02)
aRt = 5780.9 RB/D.
5-164
1300
m2 = -78 PSI/CYCLE
1250
p11iR = 1,177 PSIA

ml = -37 PSI/CYCLE
1200
11 50
^n
^ ^.
a^
fJ,
1100
t +At
p At x = 3000
1050 x
1000 1 1 1j
102 103 10'` 105 106
t +At
P
At
Fig. 5P.4: Horner plot, Prob. 3.
0 0 0
• • •
103
UNIT SLOPE LINE
0
n
4-) 1() 2
10
10 2 10-1 1 10
At, hrs
Fig. SP.S: Log-log plot, Prob. 3.

The total mobility is, according to Eq. 5.51,
aRt
xt = -162.6
mh
Using the slope of the first straight line, ml = -37

psi/cycle,
5780.9
at = -162.6 (-37) 110
at = 231 md/cp.
The distance to the fault can be estimated using

Eq. 5.43:
^
L = 0.01217
[ kAt x l 2
ûct
atlltx z ^
or, L = 0.01217
Ht J
The time ratio at which the two straight lines intersect
is
t +At
0pt x = 3000 ;
x
therefore,
AtX = 14.6 hrs.
The total system compressibility is:
ct = coSo + cgSg + c w S w + cf
_ [(31.1) (0.40) + (750) (0.15) + (3.3) (0.45)
+ 3.3] x 10-6
5-167
ct = 129.7 x 10 6 psi-1
0
Thus,
(231)(14.6) z
L = 0.01217
(0.26)(129.7x10-6
L = 121.7 ft.
(b) The effective permeabilities to oil, water, and gas

can be computed using Eqs. 5.52-5.54, respectively. The
permeability to oil is:
k = -162.6 q°Bouo
0 mh
_ - 162.6 (817) (1.19) (0 .57)

(-37) ( 110)
^ k° = 22.1 md.
The permeability to water is:
kw = -16 2. 6 awûw
_ -162.6 (2327) (1.02) (0.45)

( -37 ) ( 110 )
kw = 42.7 md.
The gas permeability is:
(1000 qgt - q°RS - qsw) Bgug

kg = -162.6
mh
[(1000) (1512) - (817) (330)] (0 .00196) (0.0142)

-162.6
(-37)(110)
kg = 1.4 md.
5-168
(c) The total skin factor can be computed using Eq.
5.S5:
pwf(At=0)-P1hr
s = 1.151
m
^t + 3.231.
- log
^ctrw J
= 1.151 1012-1177
s
37
^
231
- log
(0 .26) (129. 7x10-fi) ( 0. 356) Z
+ 3,231
s = -0.05.
Based on the total skin factor, it appears that this

•
well is not damaged.
(d) The average reservoir pressure will be estimated

using the Matthews-Brons-Hazebroek Method. From an
extrapolation of the îrst straight line to (tp + At)/at = 1 ,
p* = 1,349 psia.
Notice that p* is obtained from an extrapolation of the

first straight line because this is a bounded system.
The dimensionless producing time is, according to Eq. 5.10,
0.0002637ktp
tDA ûctA
It
tDA _
0.0002637attp
¢ctA
•
5-169
(0 .0002637) (231) (43, 869) _
tDA _ (0.26)(129.7x10-6)(81)(43,560)
40
tDA = 22.46.
From an extrapolation of the appropriate curve on Fig.

5.10,
pDMBH = 4.24.
From Eq. 5.9,
,^ m
PR = p + T.-TUT pDMBH
= 1349 + (-37)(4.24)
2.303
PR = 1,281 psia.
•
5-170
-
aappa
Engineering
SAPHIR V1.30 Methodology 3.7
Effect of errors in the rate history

During pre-processing of the data, an inaccurate definition of the test history introduces errors in
the resulting analysis curves: they are distorted, and can be mis-interpreted. In the following, we
will discuss some of the possible errors of rate, describe their influence on the analysis curves,
and show how to identify and to correct them.
When preparing build-up data, it the production time used for superposition is too small
compaied to the actual-well production history, the build-up correction will be excessive. On a
semi-log scale, the build-up curve is over compressed, and the straight line slope, if any, will be
over estimated. On the log-log derivative plot, the same error produces an upward deviation of
the response above the 0.5 line. This behavior is similar to responses corresponding to a
decrease of mobility, such as boundary effects.
t Conversely, when the production time is over-estimated, the build-up correction is insufficient.
On log-log derivative plot, the curve drops below the 0.5 line, as if mobility was increasing.
The definition of the well production history is not always a simple task. Some flow rates
measurements, such as during the clean-up period for example, are frequently missing and
have to be estimated. In other cases, fluid has been Injected into the well and this has to -be
considered in the history. The well can also have been shut-in for a while before the test, after a
production period: is it possible to neglect the well history prior to the long shut-in?
The influence of the different rate history simplifications can be illustrated on the following
simple example: the well has been produced at the same rate during two periods separated by a
shut-in, and the final build-up is analysed. The first production is 100 hours, the intermediate
shut-in period 50 hours, the final drawdown 20 hours. If the last build-up is analysed with 'the
superposition of the complete rate history, the derivative response is correct. If the equivalent
Horner time method is used, the production time is assumed to be 120 hours, and the
intermediate shut-in period is ignored: tp is too large and the derivative curve drops below 0:5.
On the other hand, if the production prior to the 50 hours shut-in is ignored, the production time
used for the superposition calculation is too small (20 hours only), and the derivative deviates
upwards [Fig.3.5].
In practice, it is recommended to check the validity of, the different assumptions performed
during the definition of the rate history. After the interpretation, a simulation of the complete well
pressure response can help in adjusting the assumed flow rates if necessary.
For long and complex production histories, the flow rate curve can be simplified, in particular for
rates produced a long time before the analysed period. The equivalent Horner time concept,
based on cumulated production figures, may then used. The closer the production data is to the
test period to be analysed, the more accuratly it has to be defined.
Before deciding to neglect part of the production history, it is recommended to first evaluate the
possible influence of this data on the final build-up. This is done by comparing the superposition
piotq,, including this production data or not.
S.ArHIR V1.30 Methodology 3.8
ie
ieZ
.,..
i . '^ ... •••
to
i
ie e-z 18-1 lee lei 1e2 to ^
Figure 3.5
to
Cft
le2
;
^
^
te1
+
Cft
e
le 1e"Z
Figure 3.6
,
i
aappa
Engineering
Methodology 3.9
SAPHIR V1.30
.7
Effect of errors In the initial time and pressure
Shut-In time errors
When estimating the elapsed time at for a build-up response, the time of start of the period,
^.^.7 when the well was closed, can be in error. If the shut-in time is too soon (earlier than actual shut-
in time), the resulting elapsed times will be too large, and both the pressure and the derivative
curves are compressed at early times (Fig.3.6J.
On the other hand, if the shut-in time is taken as later than the actual value, the resulting elapsed
times are too small, and the curves are expanded in the early time region [Fig.3.71.
The influence of time errors is frequently a function of the sampling rate of the pressure gauge.
If an electronic gauge is used, points are recorded every few seconds, and a fraction of a minute
difference on the shut in time can produce a significant distortion. With mechanical gauges,
errors smaller than one minute are not really significant.
Shut-in pressure errors
When an error is made on the last flowing pressure before shut-in, pressure changes, Ap, -will
then be wrong, but only the pressure curve Is distorted, and not the derivative, since the
pressure error is a constant (Fig.3.8 & Fig.3.9J.
le
1e2
+
>I 1e1
4
1a tp-2 ,p-1 to
II M^
Figu ►e 3.7
i
.
a+1 " 6 1a: 1 V I.JV
^wCUûauôgy 1U
•^ '
to
ieZ
Ell
iel
++ En
a
to 1©-z
19-1 iee lei 102 to
Figure 3.8
to
i
ieZ +++
iei
8
19 18-Z
iee iei 1e2 to •
Figure 3.9
• ^.
• • •
Log-Log Plot: BUEGYPT
Quick Match Results

Rad i o! h omogeneous
Int•rs•cfnp fads (45) - [L:L
Cwstwt conVressbft
Cs =0.0089 bbUpsi
-0.0064 bb,/psi
c -0.05 hr
• K =90 md
= 4.5
LNF =850 ft
PI -0 psia
^ ••
100 - •
^ i• •
QI •
^ •
y •
'Po
•
• •
• •
• • •
•
• • • • •
10
0.01 0.1 1 10
Equivalent Time(bours) - Tp=197.62
This is not a final analysis of the test but illustrates the output when using the "time stepped" wellbore storage model. The stepwise increase in the
wellbore storage factor in this test results when the rising fluid level encounters an increase in tubing size. A continual increase in the wellbore storage
factor is also occurring during the test because of gas compression in the weilbore and gas going back into solution in the wellbore oil. PanSystem
V2.6 is not able to describe this combination of changing wellbore storage.
0 0 0
• • •
1000 - Prob. 5-3: Lo -Lo Plot, no corrections
. •
• •
^.
N •
g •
a-1oo
CO
I ^, •
o •
.
.
. n
10 0.01 0.1 1 10 10
Equivalent Time (hours) - Tp=43869.2188 _^
This is a log-log plot of the test data with no corrections to the initial pressure or time. The raw data, from an Amerada mechanical gauge,
has a lot of scatter and a large smoothing parameter of 0.5 is being used.
• • •
1000 Prob. 5-3: Lo -Log Plot -.25 hr corr. _
• ..
.
, .. .
. •
^. • .
a
^ 100
^
0
• . .
.
. ^ n . .
. . •
. .
10 101
o.1 10
I This is a log-log plot of the test data with a correction of -.25 hours to the starting time of the test.
r • • •
Prob. 5-3: Type Curve Match

Match Results
Radial homogeneous
1000 Infinitely actkV
K = 24.8654 md
s = 1.0097 bbVpsi
= 1914.8909
= 0.8265
^ ♦♦
'N100
Q
n n
CD n
^ n n
n n n n
n n
10
0.1 1 10 10
Equivalent Time (hours) - T-43868.9688
• • •
13a Prob. 5-3: Semilog Pl ot _
oa .^ Resucs
^^
fm* I I I I a l l-
1 = 21.317 md
= 2344.Q711 md.R •
=1.3297 md ^
an
= 41.0559 and ^ • •
a = 222-Z76D mdkP
= 516.5775 R
E = 1.0261 •
_ ^.92®9 Pal
a _.q2675
` 1 ^-gqg^ pain
1 =127.6887 R
1100
10oioo 10 00 1 0000 100000

Homer Time Function - Tp=43868.9688
• • •
1GO Prob. 6-3: Lo -Lo Plot _

Quick Match aeadta
ROXW homogeneous
SkVW fVA
= 1.0097 bbVpd
K z 21.317 md
-02G75
=127.7 R
Pi =1484.4556 psia
♦♦ ♦♦
♦ ♦
a 100
00
m
0
• . .
. •
.
n
. n . .
.. n
. .
10 0 1
1 10 100
The theoretical model on this plot (solid line) was generated using the solution from the semilog plot. It is obvious that this solution does not
provide a good history match of this test.
0 • •

Match Results
am homopsr►sais
nierseCUnD tat^lEs (90) - ILL]
=1.15 bbUpd
= 31.S44 md
3 = 25&51
-163.418 ft
=1552.4928 psia
Cq
a 100
cc
.. .
.
, n n
W7
.. •
10 0.1 1 10 I
Equivalent Time (hours) - T- 3868.9688
• PROBLEM BUNDAH2: PRESSURE BUILDUP TEST ON A WELL
WITH ACID STIMULATION
File BUNDAH2.PAN contains the data for a buildup test on a well

which has been stimulated with 2000 gallons of 15% HCL. You are requested
to analyze the test for permeability and skin factor, and to determine if the
stimulation treatment was successful.
9
• • •
1 Prob. 6-3: Semilo Plot _
Mg" R.
p«
taults (90) - ILL)
2 00 bbUpsi ♦
944 md
51
418 R
2.4928psia
11
100^00
1 000 1 0000 1 00000
Homer Time Function - T p=43868 .9688
0 • •
__ Prob. 5-3: Lo -Lo Plot _

1000
Match Psadts
rêne^ huks C.io} - lLtl
= 1.15 b6Upei
_ ^.s427 me
= 7.1227
n 153.568 ft
^ 1786.2324 pals
CL 100
cc
n r n n n
n n
10 0.1 1 10 1
Equivalent Time hours - Tp=43868.9688
• • •
Prob. 5-3: Semiio Plot _

13 00
Ma" RAUft
saw homogeneous
^ô)-(^:y
1 = 115 bbUpsl
= 49.8427 md
= 7.1227 •
= 153.556 ft
1786.2324 psia
II
10°ioo 1000 1 0000 1 00000

Home r Time Function - T p= 43 86 8 .9688
PROBLEM GRIF-AUS: OIL DRILLSTEM TEST, HIGH
0 PERMEABILITY SAND,
INTERSECTING FAULTS
This is a drillstem test on an oil well located between two

intersecting faults which are believed to be sealing. A map showing the
location of the well is provided. You are requested to analyze the final
buildup in this test for permeability and skin, and to determine if the test
confirms'the geological interpretation shown on the map. Can you find
other boundary models that will match this data?
PanSystem File: GRIEF-AUS.PAN
rW = 0.35 ft
0 0.174
•B. 1.361 RB\STB
h 40 ft
!^'0 0.294 cp
ct 18.66 x 10', psi-1
0
. ^^
ONL'%? PRo,»c ► Nb Ws--- LL

AT -['1e'tE or- DsT
^ ^.
^ KILOWETERS 1 - '
^ ^ s -^ •
t
SOLUTION: GRIF-AUS
SOLUTION:
Attached is the analysis of the test data using two different models: (1) Intersecting fault model with a 30
degree angle of intersection, and (2) U-shaped fault model. In both cases, the well was assumed to be equal
distance from the faults. The U-shaped fault model gives a better history match of the test, but the intersecting
fault model is in better agreement with the geological interpretation. Other models will also match the test but
have no geological basis.
From the intersecting fault model,
• k=2,700md
s = 10.6
Distance to faults = 450 ft
Angle of intersection = 30 degrees.
While the intersecting fault model does not result in a perfect match of the test data, it Iis important that the
behavior of the test is consistent with the geological interpretation of this reservoir.
•
GRIF-AUS: Manual T ype Curve Match
Match Results
Radial homogeneous
Infinitely acting
K = 2493.5371 md
Cs = 0.0386 bbl/psi
= 2166.4456
= 7.6725
0.01
^ ^ ^ n n nnnnnn
n •
ca
10
CO
W ♦
^. ♦
C7
CU
=
(D
^
a 0.001
cc
a)
0
.
♦N♦
♦♦
♦
♦♦
0 . 00ni
0.0001 0.001 0.01 0.1 1 ,0
This graph shows the results of a manual type curve match where the minimum in the derivative curve was assumed to represent radial
transient flow. The purpose of this plot is to show the behavior of the test data relative to the ideal behavior of a homogeneous radial infinite
system (i.e., the Bourdet, et al., derivative type curve); there is no gurantee that this match gives the correct solution.
0 0 0 .
• • •
GRIF-AUS: Intersecting Fault Model
0.01
T n ©
W ♦
♦
H ♦
^
CL
U Quick Match Results
^ 0.001 Radial homogeneous
0 Intersecting faults (30)
^ n s = 0.03 bbUpsi
K = 2700 md ♦
cc S = 10.6
Q L1 ft N♦
_ 450 ♦
L2 = 450 ft • ♦♦
Pi = 3864.1236 psia ♦ ♦ ^•
♦ ♦ •
0.0001
0.001 0.01 0_1 1 10
This is a history match of the test data using an intersecting fault model where the angle of intersection is 30 degrees. A better match can be
obtained with a smaller angle of intersection, but 30 degrees is the smallest angle permitted by PanSystem.
The early data appear to be influenced by changing wellbore storage; this could be caused by phase segregation, gas compression, etc. No
attempt was made to model the changing storage.
The intersecting fault model does not adequately model the late data; this is probably caused by the presence of other boundaries, or because
the geometry depicted by the geological map provided with the test data is not correct. Other models such as a U-shaped fault (this model is
attached) give a better match of the test, but without other data to support these models, it is dangerous to conclude that these models correctly
describe this reservoir simply because they give a good history match.
GRIF-AUS: History Match, U-Shaped Fault
0.01
^
. '
Quick Match Results

a di al h omogeneous
shaped faults - [L:L:L]
s = 0.03 bbUpsi
' • K -2700 md
^
T ♦ ♦ = 10.6
cc = 530 ft
pp Pi = 3882.4931 psia
C7
cc
0.001
0
CL n •
cc ♦ ^
a^
o ♦
•
• ♦
0 . 0001
0.001 0.01 0.1 1 10
This is a match of the test data using a U-shape fault model. This does not agree with the geological interpretation of intersecting faults, but
gives a better match of the test data than is obtained with the intersecting fault model.
0 is 0 ,
^ PROBLEM 84-35X: Processing and Analysis of Raw Data,
Multirate Drillstem Test, High Permeability
Oil Reservoir
This test consisted of the following flow/shut-in sequence:
n 6.75 hours `cleanup' flow at 3300 STB/D

n 1.214 hours shut-in
n 14.483 hours `cleanup' at 3742 STB/D
n 25.172 hours shut-in to run bottomhole pressure gauges
n 2.59 hours flow at 1120 STB/D
n 0.35 hours shut-in to increase choke size
n 43.5 hours main flow period at 2794 STB/D
n 73.12 hours main shut-in period
Pressure data from the lower gauge located at a depth of 5587.7 ft-KB TVD
• are presented in File 84-35X.TPR.
You are requested to build a PanSystem data file for this test, and to
analyze the final buildup for permeability, total skin, and p*. You are also
requested to comment on possible boundaries or heterogeneities that are
indicated by the test.
Additional Data:
rw = 0.354 ft ct = 1.28 x 10-5 psi-'
h = 147.7 ft B° = 1.336 RB/STB
0.228 PO 0.69 cp
SW 0.213 T= 187°F
Y° = 32 °API
0
• • •
2800 84-3 5X: PRESSURE AND RATE HISTORY

3816.84
2600 a
3038.5
2400
2200
ca
2260.17 m
a 2000 m Pressure ^
m a - Rate Change Q ^ U)
`, a p
^ a !! ^
y 1800 0 0
n- a a 1481 . 83 c
1600 o a 0
aa
a a
S aa
1400 a 0
g o 703.496
a a
1200 8 ^
^
100-5 0 4.84
0 -25 0 25 50 75 100 125 15
Time (hours)
This is a plot of the raw pressure data before any data reduction, and the rate history.
Problem 84-35X: Rate Change Table
is
LI
2700 84-35X: Test Simulation Overla
2849. 88
2675 2486.66
0
o m BH Pressure
2650 _ T
^^ n 2123.44
°
2625 1760.22 ^
o n m
m 2600 1397
Quick Match Results cc
Rsdalharagsnsom
a ParoM fauft - ILL)
2575 Cs = 0.02 bblipsi 1033 . 78 _
u
K = 2eg.05 md ° O
S = 13.6577
L -548 ft
2550 Pi = 2^7 . 1129 p sia 670 . 56
2525 307.34
2500 45.88
0 25 50 75 100 125 1
Time (hours)
i
1; is C J
V
• • •
2700 84-35X: SQUARE ROOT PLOT LINEAR FLOW
2675
2650
2625
a.
2600 a
Model Results
Radial homogeneous
P ara" fa u lts
2575 w = 11 rr.2631 ft
L1 = 433.6464 ft
Sc = 6.3596
K = 289.05 md
S = 13.6577 a
P` = 26?^_8G'^^' psai
25503L -
4 5 6 7 8
Square-root Su sition Time Function
This figure illustrates the straight line linear flow analysis on the square root plot. The straight line is constructed through those points
which form a half-slope on the log-log derivative curve. The slope of the linear flow straight line yields the distance, W, between the
parallel faults. Data on this plot are from the main shut-in period of the drillstem test.
270 84-35X: SEMILOG QU IC K MATCH RESULTS
an TP1 Pressure
- Gen
mat
2675 a
"NOUN
2650
n
^ o a
L42 2625
2s0o
Quick Match Results

Radial homogeneous
Parallel faults - [L:L]
2575
Cs = 0.02 bbl/psi
K = 289.05 md
S = 13.6577
L = 548 ft
I N = 2697.1129 psia
2550
1 10 1 00 1 000
Superpos ition Ti me Function
This plot shows a match of test data from the main shut-in period to the type curve generated using the model and results displayed in
the results box.
is 0 0 I
IZ
• • •
01 84-35X: LOG-LOG QUICK MATCH RESULTS -
cc
:2
m TP1 Prenure
a • .. RsdW Deriv
...
Q 0.01
C'°" pr°as<x°
cc - Gen derNaWw
CD
0
0 • •
d •
•
Q 0
Quick Match Results

Radial homogeneous °• • • •
PafiKEi hKft - [L:Ll • ° °
Cs -0.02 bbil ^ e • 0 • .
K -269.05 md • ^ ê^' • ^
0
S = 13.65T7 ° • ^ ^
L = 548 ft e ° • •" ss
Pi = 2697.1129 psfa s
0.0V
01 0.1 1 10
Equivalent Time ihours
This plot shows a match of test data from the main shut-in period to the type curve generated using the model and results displayed in
the results box
2700 84-35X: STRAIGHT LINE SEMILOG ANALYSI S
2675
aft4 walw aaftaw

2650 70 al l
C
.T
v O O
2625
2600
Model Results
Radial homogeneous
Parallel faults
K = 289.0499 rtx!
kh = 47115.1367 md.
2575
Rinv = 2878.2573 ft
FE -0.3906
dpS -105.1856 pal
S = 13.6577
f" _ 2678.1826 psia
2550
1 10 100 1 000
Superposition Ti me Functi on
This is the conventional straight line semilog analysis of the main shut-in period. The straight line is constructed through those data
which correspond to the horizontal derivative on the Log-Log plot.
is is 0
• • •
0. 1 84-35X: LOG-LOG PLOT
o a o coca
0 0
a
a
m ,
H
U) e
[ TP1Pressur.l
CL ° .• Radial Detiv
(30.01
0 . .
e
n.
c4 e ^
o 00
e
e, .
. •° °
'°°
^ • °
°°.a • as
. . 4 . . ^• e
e , . , 0 : •0. e
. ,
0 • °• ^
0-OT
01 0.1 1 10
Equivalent Time (hours) __
84-35X
Main Shutin
r
• 4. The buildup data are presented on a MDH plot in

Fig. 5F.6. The early time data form a straight line
which has a slope
m = 31.1 psi/cycle.
From Eq. 5.66,
k = 162.6 m qBP
k = (162.6) (210) (0.70) (1.37) = 87.7 md.

( 31.1 ) (12)
The minimum time required to reach pseudosteady

state in this well is
380ûctA
tpss
i0 t = (380) (0.22) (0.7) (14x106) (80) (43,560)

pss 87. 7
t = 32.6 hrs.
pss
Since tp >tpss, it is assumed that pseudosteady state

was reached before the well was shut in. Therefore,
the Dietz method can be used to estimate PR- The MDH
straight line can be extrapolated to PR at a shut-in
time which, according to Eq. 5.74, is
ûc tA
(Ot)pR = 3792 kC A
= 3792 (0.22) (0.7) (14x10-6) (80 (43,560)

(At) -
(87.7)(-30.88)
PR
(At)p = 10.5 hrs.

R
5-171
2770
2760
27 50
^ 2740
^
V N
N j
i^.
2730
2720
2710
2700
10 1 10 • 102
• At, hrs
-4-
From Fig. 5P.6 at At = 10.5 hrs,

0
PR = 2,773 psia.
The skin factor is computed using Eq. 5.67:
[PlhrPwf(At0) - log
s= 1. 151 k + 3.23
m ûctrw
= 1.151 [2741-2360
s
3 1.1
87'7
- log
(0. 22) (0.7) (14x10-6) (0.17) 2
+ 3. 231
i• s = 7. 3.
5-173
5. Data required to construct the Agarwal semilog
and log-log plots are presented in the following table:
t At
Ate tp
At pws pws pwf(At=0)
(hours) (hours) (psia) (psi)
0.15 0.15 3,680 146
0.2 0.2 3,723 189
0.3 0.3 3,800 266
0.4 0.4 3,866 322
0.5 O.S 3,920 386
1 1 4,103 569
2 2 4,250 716
4 4 4,320 786
6 6 4,340 806
7 7 4,344 810
8 8 4,350 816
• 12 12 4,364 830
16 16 4,373 839
20 20 4,379 84S
24 24 4,384 850
30 29.9 4,393 859
40 39.9 4,398 864
50 49.8 4,402 868
60 59.7 4,405 871
72 71.6 4,407 873
Log-log and semilog plots of pws versus Ate are

presented in Figs. 5P.7 and 5P.8.
From the log-log plot, it is observed that the
unit-slope line ends at approximately 0.15 hours.
According to the 50 At rule, therefore, wellbore
storage effects should be negligible for equivalent
^ times greater than 7.5 hours. This is consisent with
the beginning of the semilog straight line shown on
5-174
Fig. 5P. 8.
• The slope of the semilog straight line is
m = 74 psi/cycle.
Permeability is computed using Eq. 5.82, i.e.,
k = 162.6 ^
k = 162.6 (250) (1.136) (0.8)

(6 9 )
k = 7.2 md.
Total skin factor, according to Eq. 5.83, is
= 1.151 [Pws(Ate = lhr) - Pwf (At=0)

s
m
is
- log k + 3. 231 .
ûctrw.
From the semilog straight line at Ate = 1 hours,
pws(Ate=lhr) = 4,280 psia.
Therefore,
= 1.151 4280-3534
s
74
7' 2
- log -
(0 .039) (0. 8) (17x10-6) (0.198) Z
+ 3.23
•
J
s = S.5.
5-177
4600
pws(Ate=1HR) = 4,280 PSIA

4400
4200
m = 74 PSI/CYCLE
3 ^ -^--
^ 4000 -+
-^,
o^ l- t
;-+-
3800
3600
10-1 1 10 102
Ate, hours
Fig. 5P.8: Agarwal semilog plot, Prob. 5.
• 0 0
• 11 •
10"
UNIT SLOPE LINE
0
4--) 103
(4-+
F--' V)
10 2
10 1 10 102
Ate, hours
Fig. 5P.7: Agarwal log-log plot, Prob. S.

PROBLEM WD73BU : BUILDUP TEST WITH WELLBORE
0 PHASE REDISTRIBUTION
This is a buildup test from Louisiana that illustrates the effect of

phase redistribution in the wellbore. The well was producing with gas lift
prior to the test. There is also an apparent boundary effect at the end of
the test but, since no information is available on possible boundaries, it is
recommended that no attempt be made to model this anomaly. You are
requested to analyze this test using three different wellbore models: (1)
Constant storage factor, and (2) Fair changing storage factor model, and
(3) Hegeman, et al., changing storage factor model. Which of these
models best describes this test?
PanSystem File: WD73BU.PAN
rW=0.4115ft
^
h=21ft
0 = 0.30
B0 = 1.238 RB/STB
µo = 1.377 cp
ct = 9.9 x 10' psi-1
tp = 2280 hrs
p,,,Ât=Q) = 2304 psia
q = 155 STB/D
0
WD73BU
SOLUTION:
An attempt to match this test on the constant storage factor type curves shows that the wellbore storage
factor is continually decreasing during the early part of the test. Decreasing wellbore storage can be caused by
compression of wellbore fluids, gas going into solution, or phase segregation. Since the "hump" in the early
data indicates the probability of phase segregation, it is believed that this Is the primary cause of the changing
wellbore storage.
An attempt to match the data using the Fair changing storage model was not successful. The derivative
decreases more rapidly than is explained by the Fair model.
It was found that the data are best matched using the Hegeman, et al., model. From this match,
k = 489 md
s=210.
No information is available about the reservoir or well completion that explains the very large skin factor.
It is not obvious that radial transient flow exists in this test. Accordingly, there is considerable uncertainity in
the placement of the horizontal derivative in the type curve match which is presented, and in the semilog
analysis which is presented. The final solution does give a reasonably good history match of the test data,
however, and it is this observation that permits us to have a reasonable level of confidence in the final solution.
LJ
3200 WD73BU: Cartesian Plot
3100 • • • • • • • • • • • • • • • • • • • • • • • • • • •
^
3000 •
.
•
•
•
•
-290
CU
Q2800
... •
2700
0-
2600
2500
240&1
23001, 5 10 15 2
Ela psed Time (hours)
After building up for a short time, this plot shows that pressure then began to decrease because of wellbore phase redistribution; pressure
again increased toward the end of the test. This "humping" effect is a common feature of buildup tests that undergo changing wellbore storage
due to phase segregation within the wellbore; it should be noted, however, that phase redistribution does not always cause the pressure to
decrease.
9 0 0
• • •
3120 WD73BU: Semilo Plot =
3100 • .
• • •
• • • . . • • • . . .
308 a
3060
CL
CD
a 3040
3020,
3000
2980
10
1 Equivalent Time (hours) - Tp=2280.0
This is a magnified view which shows the decreasing pressure from approximately 0.65 hrs to 5 hrs; this is caused by phase redistribution.
Pressure continues to increase from 5 hrs to the end of the test.
This is a manual type curve plot of the test data. It appears from this plot that the wellbore storage factor was decreasing during the early
part of the test. This could be caused by (a) compression of the wellbore fluids, (b) gas going back into solution, or (c) wellbore phase
redistribution. While all of these may be occurring to some extent, it is believed that wellbore phase redistribution is the primary cause. This is
a gas lift well and phase redistribution is a common problem in other gas lift wells in this area. Also, the cartesian and semilog plots clearly show
the "humping" effect that is often associated with phase redistribution.
The late data indicate a possible boundary effect. Scatter in the late data, shown by the derivative curve, is possibly due to poor resolution in
the mechanical gauge being used.
0 0 0
i • •
1ô WD73BU: Lo -Lo Plot
O NE n.n
Aw-
100
...
a 10
CO
7FD
0
n Y
n n
1 Q uick Match Results

Radial homogeneous
IT
Infinitely acting
Cs = 0.0026 bbl/psi
Cphi = 379 psi
Tau = 0.42 hr
K = 489 md
0.1
S
Pi
0.01
= 210
= 3104.4521 psia TFITI
0.1 1 10
E,uivalent Time (hours) - T=2280.0
This is a match of the data using the Hegeman wellbore phase redistribution model. This match was obtained by trial and error using Quick
Match
3200 WD73BU: Semilo Plot -
.. TP1 Pressure
Generated
3100 OT .
3000 •
2900
az
Q2800 4 ick Match Results
y homogeneous
Infinitely acting
Cs = 0.0026 bWpsi
(n 2700
pN = 390 psi
Tau = 0.42 hr
K = 489 md
2600 S =210
Pi -3104.4521 psia •
2500 • '
2400 •
230 n
100 1000 10000 100000
Homer Time Function - T p=2280.0
This semilog plot shows the model generated by Quick Match superimposed on the test data.
40 0 0
• ! 0
3200 WD73BU: Semilo Anal ysis
3100
••
3000 •
•
2s0a •
.
.
a 2800 Model Resutts • •
Radial homogeneous •
Infinitely acting ♦
K = 471.382 md
o`'i 2700
= 9899.0225 md.R •
Rinv = 1321.9075 ft •
FE = 0.0488
2600 = 758.8402 psi •
SS = 201.3361 •
P* = 3101.7341 psia •
•
2500
2400,
230j00
1000 10000 100000
Homer Time Function - Tp=2280.0
1. The straight line used in this analysis was constructed through those data which form a horizontal derivative on the log-log plot.

Pressure Transient Testing Vol 1

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Pressure Transient Testing Vol 1

Încărcat de

Drepturi de autor:

Formate disponibile

•

PRESSURE TRANSIENT TESTING:

DESIGN AND ANALYSIS

ALL RIGHTS RESERVED

This book, or any part thereof, may not be reproduced

Types of Tests . . . . . . . . . . . . . . . . . . . . . . . . 1-1

II. FUNDAMENTALS OF RESERVOIR FLUID FLOW

Summary of Major Equations . . . . . . . . . . . . . . . 2-89

III PRESSURE DRAWDOWN TEST

Multiphase Flow . . . . . . . . . . . . . . . . . . . . . . . 3-75

IV. MULTIPLE RATE TESTING

V. PRESSURE BUILDUP TEST

End of Semilog Straight Line . . . . . . . . . . . 5-42

Test Design . . . . . . . . . . . . . . . . . . . . . . . . . . .5-123

VI. PRESSURE TRANSIENT BEHAVIOR IN

Types of Fractured Systems . . . . . . . . . . . . . . . . 6-2

LBF. BILINEAR FLOW - LIQUID RESERVOIRS

Flow Periods For A Vertically Fractured Well ... . LBF-1

NF. INTRODUCTION TO PRESSURE

Warren-Root Model . . . . . . . . . . . . . . . . . . . . . NF-1

Type Curve Matching . . . . . . . . . . . . . . . . NF-7

VII. GAS WELL TESTING

Theoretical Method of Analysis ..... . 7-93

GBF. BILINEAR FLOW - GAS RESERVOIRS

Bilinear Flow Equations . . . . . . . . . . . . . . . . . . GBF-1

VIII. INJECTION WELL TESTING

Pressure Falloff Test in Unit-Mobility Ratio

Average Reservoir Pressure . . . . . . . . . . . . 8-8

IX. INTERFERENCE AND PULSE TESTING

Multiple Active Wells . . . . . . . . . . . . 9-16

APP. APPENDIX A: FLUID AND ROCK PROPERTY

2. Theory and Practice of the Testing of Gas Wells (third edition),

3. Earlougher, R.C., Jr.: Advances in Well Test Analysis, Society of

5. Dake, L.P.: Fundamentals of Reservoir Engineering, Elsevier

6. Lee, W.J.: Well Testing, Society of Petroleum Engineers, Dallas:

7. Slider, H.C.: Worldwide Practical Petroleum Reservoir Engineering

8. Streltsova, T.D.: Well Testing in Heterogeneous Formations, John

9. Stanislav, J.F. and Kabir, C.S.: Pressure Transient Analysis,

12. Dake, H.P.: The Practice of Reservoir Engineering, Elsevier,

13. Horne, R.N.: Modern Well Test Analysis: A Computer Aided

14. Lee, J. and Wattenbarger, R.A.: Gas Reservoir Engineering, Society

One of the most important functions of the petroleum

A pressure transient test requires that the flow

jj^,e .Iu_S,u.Ft r-*- a. ^ vw na7b, - - l,i.-(JCv ^;v,,c&kaq,e,

It has long been recognized that the pressure behavior

• Interwell rock properties, and directional

III. PURPOSE AND SCOPE

The purpose of this manual is to present those methods

This manual is written so that the reader can develop

V. NOMENCLATURE AND UNITS

When possible, the standard symbols ado p ted b y the

Matthews, C.S. and Russell, D. G.: Pressure Buildup

Earlougher, R. C., Jr.: Advances in Well Test Analysis,

Theory and Practice of the Testing of Gas Wells, third

"Letter Symbols for Petroleum Reservoir Engineering,

"Supplements to Letter Symbols and Computer Symbols for

. "Supplements to Letter Symbols and Computer Symbols

FUNDAMENTALS OF RESERVOIR FLUID FLOW

A prerequisite to designing or analyzing a pressure

II. PRIMARY FLOW VARIABLES

Petroleum reservoirs are very complex and the equations

B. Slightly compressible fluids -- L i Q U I D 0

C. Highly compressible fluids (;-AS

2. Number of mobile fluids in the reservoir