1:23:09 PM - 1 -

GENERAL PRINCIPLES
OF
BOTTOM-HOLE PRESSURE TESTS

BHP BALANCE



ACHIEVED OBJECTIVE

BHP TEST
Field Operators
Test Analyst
Proposal
Writer






Prof. Mike Obi Onyekonwu
Petroleum Engineering Department
University of Port Harcourt, Nigeria.




ACHIEVED OBJECTIVE
2
PREFACE
Bottom-hole pressure test is one of the most economical methods of obtaining
information required for reservoir management. The test involves measuring the
bottom-hole pressure at the sandface under specified rate conditions. The bottom-
hole pressure contractors (wireline operators) are responsible for running the test
while the reservoir engineers analyze the tests.

For the past twelve years, we have been analyzing pressure tests for companies in
Nigeria. The analysis involves using sophisticated software and models to obtain
reservoir parameters that will not only match obtained pressure, but that are in
agreement with known information about the reservoir. From our experience, we
observed that more than fifty percent of the errors in BHP analysis could be due to
ignorance on the part of the field operators responsible for measurements. Therefore,
there is need for everyone involved with BHP test, to understand the purpose of the
test and factors that can affect the test. This is what we call the total concept approach
to BHP tests and analysis.

The total concept approach requires that all parties involved in BHP operations - BHP
proposal writer, BHP contractor, production staff, reservoir engineers and users of
BHP test data - need to understand the principles of BHP tests. This book provides
required general information for all the parties involved in BHP test and analysis.
Chapter 1 of the book is on general introduction and definitions while Chapter 2 is on
field practices. The field practices cover gauges, running procedure and procedure for
data quality check.

Chapter 3 is on basis for BHP analysis. Subjects such as units, flow phases and flow
regimes are discussed. Chapter 4 is on analysis models while Chapter 5 dwells on
skin factor and concepts. Chapter 6 is on analysis methods. The conventional and
type-curve methods are covered. Chapter 7 is on sensitivity on factors affecting BHP
tests. Factors such as leaks, rate variation and gauge problems are considered.
Chapter 8 is on field cases. Both good and bad tests are presented. The bad tests
could be a result of man-made problems and problems resulting from malfunction of
equipment. For the bad ones, solutions are proffered on how to avoid future mistakes.

Chapter 9, on the theory of welltesting, is for the academic minded group that is not
only interested in how, but why. The fundamental equation, diffusivity equation, is
derived and solutions presented for different conditions. Concepts like superposition
are also discussed. Finally, there is analogy between liquid and gas flow equations.
Chapter 10 is on bottom-hole pressure analysis in horizontal wells. Several equations
are presented and procedures for using the equations outlined. Finally some field
cases on horizontal wells are presented.

Materials for this book were taken from manuals that we have used successfully in
teaching three different short courses Practical Aspects of Bottom-Hole Pressure
Tests, General Principles of Bottom-Hole Pressure Tests and Theory of Welltesting.
These courses were organized for participants with different backgrounds. We are
therefore certain that anyone who needs some knowledge on bottom-hole pressure
tests will find this book useful. In this era of multidisciplinary approach to problems,
this book will give reservoir and non-reservoir engineers a good working knowledge
of bottom-hole pressure tests.

Mike Onyekonwu
3
ACKNOWLEDGEMENT
For me, writing a book is simply a way of expressing my understanding of a subject.
We therefore owe our teachers, students and collegues a lot in contributing to our
knowledge on the subject. We are grateful to Late Professor H. J. Ramey (Jr) of
Stanford University, California the father of Well Testing for teaching us a
difficult subject in a simple way. We are also grateful to Steve Rice and Jim Willetts
of Shell Petroleum Development Company who challenged us and helped us acquire
much needed practical experience on this subject. Tom McAllister who took over
from Steve and Jim has continued to encourage us. We appreciate his contributions to
our continued growth. We are also particularly grateful to petroleum and production
staff of Shell Petroleum Development Company, Warri, for their support and use of
our services.

I also thank my good friend, Dr G. A. Okpobiri, who was my co-consultant when we
taught the first Well Testing course. I appreciate the immense contributions made by
Laser Engineering Consultants staff and particularly Mr. Obi Ekeh and Mr. Cornel
Udoh for being part of the analysis team. Mrs. Margaret Anele and Mr. Sam Jumbo
of Laser Engineering did a good job at typesetting the book. I am equally indebted to
my students both in the oil industry and University of Port Harcourt for allowing us
use them as the guinea pigs. Their contributions are indeed invaluable.

I also thank Professor G. K. Falade who taught us the first course on Well Testing
at University of Ibadan. I am equally indebted to Professor Chi Ikoku for creating a
nice working environment at the University of Port Harcourt and therefore made it
possible for me to continue practicing engineering. Finally, I thank my dear wife,
Tina, for always standing by me. May God reward everyone. This is actually our
book.

Mike Onyekonwu

4
ABOUT THE AUTHOR
Include photgraph!
Dr Mike Onyekonwu has a B.Sc. (First Class Honours) in Petroleum Engineering
from University of Ibadan, Nigeria. He also has an MS and Ph.D. degrees in
Petroleum Engineering from Stanford University.

Dr Onyekonwu is a Senior Lecturer and former Head of Petroleum Engineering
Department, University of Port Harcourt. He is a member of University Senate. Dr
Onyekonwu is the founder and Managing Consultant of Laser Engineering
Consultants, Nigeria. Dr Onyekonwu worked for Shell Petroleum Development
Company Nigeria and Stanford University Petroleum Research Institute, California.

Dr Onyekonwu is a registered engineer and a member of different professional
bodies. He consults for Shell, Mobil, Elf, NNPC, Agip and other oil operating and
service companies. His area of specialization includes welltest analysis, reservoir
simulation, recovery methods and computer applications.



























5

TABLE OF CONTENTS

PAGE

PREFACE ............................................................................................................................ ii
COURSE ADVERTISEMENT ............................................................................................ iii

1 INTRODUCTION .......................................................................................... 1
1.1 Objectives of BHP Survey .. .............................................................................. 1
1.2 Uses of BHP Derived Information..................................................................... 1
1.3 Common Types of BHP Tests ........................................................................... 2
1.4 Ideal Conditions and Information Derived from Test........................................ 3
1.5 Importance of Sticking to Ideal Condition for Test ........................................... 4
1.6 Uses of Information Derived from BHP Tests................................................... 6
1.7 Well Test Equipment ......................................................................................... 7
1.8 Electronic Gauges and Problems ...................................................................... 11
1.9 Flowing Gradient/Buildup/Static Gradient(Fg/Bu/Sg) Survey.......................... 11
1.10 Flowing Gradient/Buildup/Static Gradient Survey Proposal............................. 19
1.11 Useful Hints on Proper Testing of Wells........................................................... 23
1.12 Practical Hints.................................................................................................... 23
1.13 Gauge Quality Check Procedure........................................................................ 27
1.14 Roles of Field Staff In BHP Survey................................................................... 28
1.14.1 Roles of Production Staff...................................................................... 32
1.14.2 Roles of BHP Contractor Staff ............................................................. 33


2. BASIS OF ANALYZING BOTTOM HOLE PRESSURE TESTS ............................ 34
2.1 Flow Phases ...................................................................................................... 34
2.2 Features of Different Phases ............................................................................. 35

3. ANALYSIS OF BOTTOM HOLE PRESSURE TESTS ........................................... 46

4. EFFECT OF CERTAIN FACTORS ON ANALYSIS OF SIMULATED DATA ..... 64
4.1 Analysis of Ideal BHP Data .............................................................................. 64
4.2 Effect of Gauge Accuracy and Datum Correction ............................................ 73
4.3 Effect of Noise .................................................................................................. 78
4.4 Effect of Gauge Sensitivity ............................................................................... 98
4.5 Effect of Rate Variation .................................................................................... 98
4.6 Effects of Leaks ................................................................................................ 112
4.7 Effect of Interference/Leak ............................................................................... 112

5. FIELD CASES ........................................................................................................... 117
5.1 Good Test .......................................................................................................... 117
5.2 Effect of Gauge Movement ............................................................................... 117
5.3 Effect of Gauges Off-Depth .............................................................................. 125
5.4 Effect of Reporting Wrong Rates ..................................................................... 125
5.5 Effect of Ineffective Shut-in/Well not flowing before Shut-in.......................... 131
5.6 Effect of Leak ................................................................................................... 132
5.7 Effect of Gauge Oscillations/Sensitivity Problems ........................................... 135
5.8 Effect of Gas Phase Segregation ....................................................................... 140
5.9 Effect of Liquid Interface Movement ............................................................... 144
5.10 Effect of Gaslift ................................................................................................ 144
5.11 Effect of Short Buildup or Flow Period ............................................................ 144

6. CLASS DISCUSSION ............................................................................................... 154

7. FLOWING AND STATIC GRADIENT SURVEYS ................................................. 165
8. SKIN FACTOR
8.1 What is Skin Factor............................................................................................ 172
6
8.2 Causes of Skin ................................................................................................... 174
8.3 Classification of Pseudoskins ............................................................................ 174
8.4 Calculation of Pseudoskins................................................................................ 176
8.5 Relationship Between Total Skin and Pseudoskins ........................................... 180
8.6 Pressure Change Due to Skin............................................................................. 181
8.7 Relationship Between Skin and WIQI ............................................................... 181
Exercises ............................................................................................................ 182

APPENDIX A: TYPICAL PROPOSAL .............................................................................. 184

APPENDIX B: PROPOSAL WITH WRONG INSTRUCTION ......................................... 192

APPENDIX C: MISCELLANEOUS MATERIAL...............................................................










































7
1. INTRODUCTION

Oil well tests are made for numerous reasons and the type of test required depends on
the objective of the test. Common well tests include:
(a) Potential test
(b) Gas-oil ratio test
(c) Productivity test
(d) Bottom-hole pressure test

Potential test involves measurement of the amount of oil and gas a well produces
during a given period (normally 24 hours or less) under certain conditions fixed by
regulatory bodies. The information obtained from these tests is used in assigning a
producing allowable of the well. The gas-oil ratio test is made to determine the
volume of gas produced per barrel of oil so as to ascertain whether or not a well is
producing gas in excess of permissible limit.

Bottom-hole pressure test involves measurement of sandface pressure and flowrate
variation with time. Such tests are quite economical to run and they yield valuable
information about the reservoir characteristics and well characteristics. Hence,
bottom-hole pressure tests are usually referred (Earlougher, 1977; 1982) to as
welltests.

Productivity tests are made on oil wells and include both the potential test and the
bottom-hole pressure (BHP) test. The purpose of this test is to determine the effects
of different flow rates on the pressure within the producing zone of the well and
thereby establish producing characteristics of the producing formation. In this
manner, the maximum potential rate of flow can be calculated without risking
possible damage to the well which might occur if the well were produced at its
maximum possible flow rate.

In this book, the term welltest will be used for bottom-hole pressure test unless
otherwise stated. In this chapter, the purpose of well testing, types of well tests and
well test equipment are discussed. In addition, other practical aspects of BHP tests
such as test procedure and equipment problems are discussed.


1.1 OBJECTIVES OF BHP SURVEY
Bottom-hole pressure tests are conducted to obtain data that can be used for the
following purposes:

Determine Well Parameters
- Skin
- Productivity Index
- Wellbore storage constant
- Fluid distribution in wellbore
- Flowing pressures in wellbore
- Static gradients

Determine Reservoir Parameters
- Average pressure in the drainage area
- Permeability
- Distance to boundaries
8
- Vertical/Horizontal permeability
- Gas/oil contacts

Determine Dynamic Influence of other Wells/Aquifer

Assess Changes Since Previous Survey
- Changes in datum pressure
- Changes to damage skin
- Changes in drainage area (from a drawdown test)
- Confirm boundaries

1.2 USES OF BHP DERIVED INFORMATION
Results obtained from BHP tests are used for the following purposes:
* Reservoir Surveillance
* Determination of Stimulation Candidates
* Gaslift Optimisation
* Input for Reservoir Simulation
* Material Balance Calculation

Examples of benefits from BHP test compiled by a client are given in Table 1.1. The
benefits were realised by using results from BHP tests for good well and reservoir
surveillance.

1.3 COMMON TYPES OF BOTTOM-HOLE PRESSURE (BHP) TESTS
Common types of bottom-hole pressure tests include the following:
(a) Drawdown test
(b) Injectivity test
(c) Buildup test
(d) Falloff test
(e) Interference/pulse tests
(f) Others

The definitions of the tests and the rate and pressure profiles during the test are as
follows:

Table 1.1: Benefits from BHP Surveys
ACTIVITIES SAVINGS ($ million)
Well Surveillance

Stimulation (abort 5 jobs, contribute to finding 5 more)
Gaslift Optimization (10% improvement of target at
$1/bbl)

0.8
1.0
Reservoir Surveillance
Sand F4.0/F4.1X Production (3 Mbopd) 1.0
Sand-X Block (new well cancelled) 3.0
Dump Creek (10% of the 6 fewer wells required) 3.0
Well -11 (sidetrack raise trajectory) 1.0
Sand D5.0X Development
(horizontal well changed to recompletion) 3.0
(10% of 8 well campaign) 4.0
Total 16.8

9
1. Drawdown Test: Involves measuring the variation of sandface pressure with
time while the well is flowing. For a drawdown test, the well must have been shut in
to attain average pressure before production commences for the test. The rate and
pressure profiles during drawdown test are in Fig 1.1.

P
wf
0 time time
q
0

Fig. 1.1: Rate and Pressure Profiles During Drawdown Test

2. Injectivity Test: This is the counterpart of a drawdown test and involves
measuring the variation of sandface pressure with time while fluid is being
injected into the well. The rate and pressure profiles during drawdown test are in
Fig 1.2.

q P
wf


0 time
Injection (-ve q)

time

Fig. 1.2: Rate and Pressure Profiles During Injectivity Test

3. Buildup Test: Involves measuring the variation of sandface pressure with time
while well is shut-in. The well must have flowed before shut-in. Figure 1.3 shows
the rate and pressure profiles during the flow and buildup periods. Buildup tests are
more common and will be the main subject of our discussion.

P
w
0 time 0 time
q
Shut-in Time
0
Drawdown
Buildup
Drawdown
Buildup

Fig. 1.3: Rate and Pressure Profiles During Drawdown and Buildup

4. Falloff Test: This is the counterpart of buildup test and it involves measuring the
variation of sandface pressure with time while well is shut-in. In this case, some
fluid must have been injected into the well before shutting. Figure 1.4 shows the
rate and pressure profiles during the injection and falloff periods.

10

q P
w


0 time
Injection


time

Fig. 1.4: Rate and Pressure Profiles During Injectivity and Falloff Test

5. Interference Test: Unlike the first four tests (drawdown, injectivity, buildup,
falloff) which are tests involving only one well (single well tests), the interference
test involves the use of more than one well (multiple well test). During
interference tests, pressure changes due to production or injection or shut-in at an
active well is monitored at an observation well. The active well and the
observation well are shown in Fig 1.5. Only one active well is required, but there
could be more than one observation well.

Interference tests are primarily used to establish sand continuity between the
active and observation wells. In situation where more than one observation well is
used, interference test can be used to find (Ramey and ) maximum and minimum
permeability and their directions.

Active Well Observation Well
Sand continuity
Gauge
q > 0
q = 0
q < 0
q = 0

Fig. 1.5: Active and Observation Wells in an Interference Test

1.4 IDEAL CONDITIONS AND INFORMATION DERIVED FROM TEST
If possible, BHP tests should be run under the stated ideal conditions as this makes
interpreting such tests easy. The ideal conditions for running different BHP tests and
information that can be obtained from the tests are given in Table 1.2

Table 1.2: Types of Well Tests and Derivable Information
Type of Test Ideal Conditions for Test Information Derived from Test
Drawdown 1. Constant rate production
2. Well shut-in long enough
before test to attain uniform
pressure in reservoir.
1. Permeability
2. Skin factor
3. Reservoir drainage volume
4. Flow efficiency
5. Distance to linear no-flow
barrier
Buildup 1. Constant rate production
before shut-in.
1. Permeability
2. Skin factor
Injectivity
Falloff
Shut-in
11
3. Flow efficiency
4. Average Pressure
5. Distance to linear no-flow
barrier
Interference 1. Constant rate production or
injection at the active well.
1. Permeability
2. Storativity
3. Anisotropic permeability
values and orientation
4. Sand continuity

1.5 IMPORTANCE OF STICKING TO IDEAL CONDITION FOR TEST
In this section, we shall discuss the importance of sticking to the ideal condition for
any test. Two factors considered are constant rate production and not shutting well
for long period before a drawdown test.

(1) Constant Rate Production: Rate variation makes tests difficult to analyse because
effect of rate changes last until well is shut-in and builds up to average pressure. Rate
changes are modelled using the concept superposition illustrated in Fig. 1.6. Figure
1.6 shows that a rate which occurred at where at time, t, will continue to affect
pressure response until time, t + t. In a laymans language, wells do not forget rate
changes that occurred in them unless they are shut in to build up to average pressure.

q
1
q
2
q
1
- q
1
q
2
+
t

t + t t t

Fig. 1.6: Effect Rate Variation

Causes of Rate Variation
Some tests like the potential tests are designed such that the rates in the wells are
varied. Such rate variations create no problem during analysis because the rates are
measured and therefore can be considered during analysis. Situations that create
problems include cases where the rates are varied and not measured. Such situations
may occur under the following conditions:

(a) Partly closing wing value to lower tools
(b) Slow shutting of well at end of flowtest
(c) Not allowing for rate stabilization. Surface indications of well stabilization
include
(i) Constant wellhead flowing pressure
(ii) Constant gas production rate
(iii) Constant fluid production rate.


12
(2) Long shut-in Requirement: The rate and pressure profiles for wells shut-in for
long and short times are shown in Fig. 1.7 and Fig. 1.8.

Short Shut-in Period
q
2
= 0
q
3
q
1
Rate
Time
Time
Pressure

Fig. 1.7: Rate and Pressure Variation During Short Shut-in Period
Long Shut-in Period
q
2
= 0
q
3
q
1
Rate
Time
Time
Pressure

Fig. 1.8: Rate and Pressure Variation During Long Shut-in Period

For the case of short shut-in period, the well did not build up to the average pressure
before the drawdown test was started. In this case, analysis of the test will involve
using three rates.

However, for the case of long shut-in time before the drawdown, the well reached the
average pressure during the buildup. Hence, analysis of the drawdown will simply
involve a single rate. A single rate test is usually simpler to analyse than a three rate
test.

1.6. FLOWING GRADIENT (FG) AND STATIC GRADIENT (SG) SURVEYS
We regard the flowing gradient and static gradient surveys as auxiliary surveys that
complement bottom-hole pressure tests. These tests and their uses are described.

Flowing Gradient (FG): The flowing gradient survey involves measuring flowing
pressure at different depths in the well while the well is flowing. Results from this
test are used for gaslift optimization. Figures 1.9 shows cases where the flowing
13
pressures measured along the traverse of the well reveal examples of optimized and
non-optimized gaslifting.





Depth






Fig. 1.9: Optimized and non optimized gaslift.

In the non-optimized case, we may have a U tube effect in which injected gaslift
gas is simply re-circulated giving rise to lower flowing pressure gradient in the upper
part of the tubing. This is shown in Fig.1.10.
Gas Gas

Fig. 1.10: Non-Optimized Gaslift.

Flowing gradient surveys also provide flowing pressures which can be used to
determine the appropriate correlation for modelling flow along the wellbore. Such
models are used for gaslift optimization. In all cases during the flowing gradient
survey, the depth where pressure was measured is important.
Static Gradient Survey: In this case, we measure pressure at different depths in the
well while the well is shut in. This implies that it can be run in well that has not been
flowing. Usually, before a static gradient survey is run, the well must have been shut
in for sufficient time to allow the pressure to stabilize. At every static gradient stop
along well, the gauges must be left for a minimum of 15 minutes so that pressures will
be steady.

The static gradient survey is used to determine the fluid distribution in the wellbore.
This information is required for pressure correction and locating the depth for the
operating gaslift valves. In a well that is closed in, the static gradient survey is a
good source of pressure data that can be used in calculating the datum pressure with
no oil deferred.

The basis for determining fluid gradients using static gradient survey is that fluid
gradients depend on the density of the fluid. Therefore, pressure gradient in the gas
zone is small because gas has the smallest density of the wellbore fluids. Figure 1.11
shows fluid gradients determined using results from static gradient survey in a
wellbore that contains gas, oil and water.

Pressure Pressure
Depth
Optimized
Non-optimized
14

Water (0.433 psi/ft)
Pressure, psi
Gas ( 0.07 psi/ft)
Oil ( 0.35 psi/ft)
Depth, ft

Fig. 1.11: Wellbore Fluid Distribution Determined with Static Gradient Survey

With the static gradient survey, we can determine the gas-oil contact which is
important in selecting the depth of the gaslift operating valve. An operating valve that
is above the gas-oil contact as shown in Fig. 1.10, will result to a non-optimized
gaslift operation.

To ensure that correct wellbore fluid contacts are determined using static gradient
survey, a minimum of two static survey stops must be taken in each phase as the
gauge moves through the fluid phases. Generally, it is recommended that in gaslifted
wells, there should be two stops around the region where the gaslift mandrels are
installed. This helps determine fluid contacts (if any) around the gaslift mandrels.

In bottom-hole pressure (BHP) tests, it is required that we measure pressure at the
sandface (mid-perforation). However, in some situations it is not possible for the
gauge to be lowered to the sandface. In such situations, the static gradient survey
provides the fluid gradient required for obtaining the pressure at the mid perforation.
The equation for calculating the pressure at mid perforation is as follows:

P
mid perf.
= P
gauge
+ (Fluid Gradient x z) 1.1

Where
P
mid perf = Pressure at the mid perforation
P
gauge = Pressure recorded by gauge at the last stop

z = Distance between mid perforation and last gauge stop
Fluid Gradient = Wellbore fluid gradient in the interval z

A graphical interpretation of Eq. 1.1 is shown in Fig. 1.12 for a case where the last
gauge depth is the XN nipple. From this, it is obvious that we need to be careful in
reporting the gauge depth. An error of 5 ft with water near the perforations (gradient
of 0.433 psi/ft) means a 2.1 psi error and this is more than the absolute accuracy of the
crystal gauges.







15




Depth







Pressure
Fig.1.12: Extrapolating Pressure to top of Perforation

1.7. COMPLETE BOTTOM-HOLE PRESSURE TESTS PROFILES
A complete buildup or drawdown survey requires that both flowing and static gradient
surveys should also be taken. Typical pressure profiles for such tests are as follows:

Flowing Gradient/Buildup/Static Gradient (FG/BU/SG): Typical pressure profile for
this test is shown in Fig. 1.13.















Time
Fig 1.13: Pressure Profile for FG/BU/SG Survey

The sequence of events performed during the test that resulted to the pressure profile
shown in Fig 1.13 are as follows:

Event Description
A B Gauge in lubricator reading atmospheric pressure as there is no
communication yet with the well
B C Increasing pressure due to running in hole
C D Flowing gradient stops
D E Running hole to final survey depth
E F Flowtest with gauge at final survey depth
F G Buildup period
G H Static Stops near the final survey depth
H I Pulling gauge out of hole
I J Static stops in the upper part of tubing for liquid level detection
J K Pulling gauge out of hole to lubricator
XN
Water
Oil
Extrapolation
to mid perf
Increasing
pressure
A
B
C
D
E F
G
H
I
J
K
Flowing Gradient Buildup Static Gradient
16

Although events in Fig 1.13 are typical, but there may be variations. For example, at
the end of buildup, the gauge may be pulled out about 200 ft and then moved down
about 200 ft to original survey depth. The profile for this case is shown in Fig. 1.14
with events GH and HI representing the pull out and run back respectively. This
could be used for checking the accuracy of depth measurement as pressure at the same
depth in a well that has been shut in for sufficiently long time must always be the
same.

Another variation is that of where well is shut in while the gauge is run in hole. With
the gauge on the bottom, the well is then opened for a flowtest and then shut-in again
for a buildup. A typical profile for this case is shown in Fig.1.15.



















Fig 1.14: Pressure Profile for another FG/BU/SG Survey


















Fig 1.15: Pressure Profile for Buildup Survey

Drawdown
Buildup
Gradient stops
Time
Increasing
pressure
Running in hole
Pulling out of hole
Increasing
pressure
A
B
C
D
E F
G

I
J
Flowing Gradient Buildup Static Gradient
Time
H
17
If this test is properly run, there will be the advantage of obtaining both buildup and
drawdown data that can be analyzed. Also as the well is shut in while the gauge is
run in hole, the problem of lowering gauge especially in high flowrate wells will be
eliminated.

The problem with this type of test is that the duration of the shut-in while the gauge is
run in hole may not be adequate for the drawdown and buildup tests to be easily
analyzed without using superposition. That is, the duration may be too short for the
system to attain average pressure, a condition required prior to good drawdown test.

Static Gradient/Drawdown/Flowing Gradient (SG/DD/FG): This is the complete test
sequence in a situation where the well is just programmed for a drawdown. Typical
pressure profile for the tests is shown in Fig. 1.16.















Fig 1.16: Pressure Profile for SG/DD/FG Survey

1.8 USES OF INFORMATION DERIVED FROM BHP TESTS
Most petroleum engineers already know how parameters derived from BHP tests are
used. To our readers who are non-petroleum engineers, this section will help them
understand the importance of some parameters derived from BHP tests.

A. Permeability k: This is a measures of the ability of a formation allow fluid flow
through it. Permeability is one of the parameters required for rate prediction as
shown in the following equations used for rate prediction:

Linear: q (STB/D) = 1.127 x
10
3
KA p
B L

Radial: q (STB/D) =
7.08 x 10 Kh p
In
r
r
-3
e
w

B


B. Skin Factor: A measure of the efficiency of drilling and completion practices
used. The skin factor can be used in calculating additional pressure drop around
the wellbore caused by drilling and completion practices. Skin factor is discussed
in detail in Chapter ????


Static Gradient Drawdown Flowing Gradient
Time
Increasing
pressure
18
The pressure drop may be caused by the following:

1. Alteration of permeability around the wellbore caused by invarion of drilling
fluid, dispersion of clay, mud cake and cement, acidizing etc. In the case of lower
permeability around the wellbore, the skin in this case can be likened to the extra
fuel spent in driving through a bad road.

2. Partial well completion as shown in Fig 1.17.

Fully completed Partial compleion

Fig. 1.17: Flow Streamlines in Fully and Partially Completed Wells

In the case of a partially completed well, the skin could be likened to the extra energy
lost at the door when many people want to go out (assuming there is fire outbreak in
the room) of a room at the same time.

The pressure drop due to skin is wasted because it does not contribute to the useful
drawdown. The skin simply causes an additional pressure drop at the well as shown
in Fig. 1.18.
P
wf
if no skin
P
wf
if there is skin
p
s
r
w
r
e
useful
drawdown

Fig. 1.18: Pressure Drop due to Skin

Pressure drop due to skin, p
skin
, and efficiency are related as follows:

Flow Efficiency (FE) =
P
P
- P - p
- P
wf skin
wf

Note that the skin and permeability are determined by the amount of pressure rise and
the rate at which pressure rises with time. This is shown in Fig. 1.19. This implies
that if pressure does not rise, there will be no pressure rise and skin and permeability
cannot be obtained from the test.

19
Time
Pressure
Skin
permeability
P*

Fig. 1.19: Pressure Rise During Buildup

C. Reservoir Drainage Volume: This is the volume of the reservoir drained by test
well. Drainage volume is required in choosing adequate well spacing and reservoir
management. Note that wells drain reservoir volumes in proportion to their rate. This
is shown in Fig 1.20.

2 q
.
q
.
q
.

Fig. 1.20: Relationship between Drainage Area and Rate.

D. Porosity : A measure of void spaces in the reservoir. Porosity can be obtained
from interference test. Volumetric calculation of initial oil-in-place requires porosity
as an input parameter. This is shown in the equation:

N =
7758 A S
oi

B
oi
.

E. Average Pressure: This is a measure of depletion as the amount of fluid in
reservoir is related to average pressure in the reservoir. Average pressure is used in
material balance calculations.
















20
2.0 FIELD PRACTICES

Over the years, emphasis has been on developing sophisticated models for BHP
analysis. However, Fig 2.1 show that achieving the objective of BHP test requires
that the parties involved the proposal (programme) writer, field staff (BHP
contractors and production staff), test analyst and test result users must play their
roles correctly.













Fig. 2.1: BHP Balance

In this chapter, we shall discuss basic BHP equipment, procedures for running BHP
tests and proposals. We shall also highlight roles of production staff and BHP
contractor during BHP surveys.

2.1 WELL TEST EQUIPMENT
The main equipment used for bottom-hole pressure surveys is as follows:
(1) Pressure recorders
(2) Lubricator (Wireline BOP)
(3) Wireline unit
(4) Christmas tree with hydraulically operated value

Figure 2.2 taken from Dake (1978) is a schematic of the equipment while Figure 2.3
shows detail of some of the equipment. Most of the equipment are also used in
wireline operations and will not be discussed in this test. We shall only discuss the
pressure gauges, which are the most important equipment for bottom-hole pressure
tests.

Fig 2.2 and fig 2.3 here???
Leave a page for
Figure 2.3: Wireline Surface Equipment
(Example of an Arrangement)

2.1.1. Pressure Gauges
Different types of gauges are used for measuring bottom-hole pressure. The
sensitivity and the accuracy of the gauges vary. The accuracy of a gauge is
principally concerned with systematic errors, often attributed to the calibration of the
gauge. For example, if the accuracy of a gauge is 5 psi and the gauge reads 1000 psi,
this implies that the correct readings lie in the range (1000 - 5) psi to (1000 + 5) psi.

BHP BALANCE
Field
Operators
Proposal Writer
Test Analyst
And Users
ACHIEVED
OBJECTIVE
21
The sensitivity or resolution of a gauge is described as the smallest pressure change
that can be reliably measured by the gauges. Table 2.1 shows bottom-hole pressure
measured in a Niger Delta well with an insensitive gauge. The constant pressure
readings after shut-in time of 3 minutes is not necessarily due to stabilization, but the
gauge could not discern the pressure changes with time.

Table 2.1: Buildup Data from Niger Delta Well
Shut-in Time, min Shut-in Pressure, psi
1 3488
2 3531
3 3539
5 3539
10 3539
20 3539
30 3539
40 3539
50 3539
60 3539
90 3539
120 3539

The types of gauges, their principle of operation, accuracy and sensitivity are given in
Table 2.2.

Table 2.2: Types of Gauges and Operation Principles
Type of Gauge Principle of Operation Accuracy Sensitivity
Amerada
Strain Gauge
Quartz Crystal
(Electronic)
Bourdon tube (Mechanical)
Change in resistivity
Change in frequency
0.2% FSD
0.05% FSD
0.035% R
0.05% FSD
0.0025% FSD
0.0001% FSD
FSD = Full Scale Deflection, e.g. 5000 or 10000 psi
R = Reading, i.e. the measured pressure

The implications of information in Table 2.2 for a 5,000 psi and 10,000 psi rated
gauges are shown in Table 2.3.

Table 2.3: Sensitivity and Accuracy of 5000 psi and 10000 psi Rated Gauges
Type of Gauge FSD = 5000 psi FSD = 10,000 psi
Accuracy Sensitivity Accuracy Sensitivity
Amerada 10 psi 2.5 psi 20 psi 5.0 psi
Strain Gauge 2.5 psi 0.125 psi 5.0 psi 0.25 psi
Quartz Crystal
(Electronic)
1.75 psi 0.005 psi 3.5 psi 0.01 psi

For the electronic gauge, the calculated sensitivity is the maximum because we
assumed that the measured pressure is equal to the Full Scale Deflection (FSD).
Some deduction from Tables 2.2 and 2.3 are as follows:
22

1. The electronic gauges are more sensitive and accurate than the strain gauge
while the strain gauge is more sensitive and accurate than the Amerada gauge.

2. If a 5000 psi rated gauge can do the job, do not use a 10,000 psi rated gauge
because the 10000 psi rated gauge has lower accuracy and sensitivity. This
deduction is more relevant when Amerada gauges are used.

Ideally, we recommend that electronic gauges be used. However, it costs more than
other gauges, but it is worth it.

2.1.1.1 Amerada Gauges
Until 1994 about 80% of bottom-hole pressure tests in Nigeria were run with
Amerada gauges. Now many companies do not use Amerada gauges. However, for
historical reasons, we need to discuss the Amerada gauge because it clearly shows the
components of any gauge: a clock, pressure sensor and recorder. Figure 2.4. taken
from Dake (1978) is a schematic of the Amerada gauge. The continuous trace of
pressure versus time is made by the contact of a stylus with a chart that has been
specially treated on one side to permit the stylus movement to be permanently
recorded. The chart is held in a cylindrical chart holder, which in turn is connected to
a clock that drives the holder in the vertical direction. The stylus is connected to a
bourdon tube and is constrained to record pressures in the perpendicular direction to
the movement of the chart holder. The combined movement is such that, on removing
the chart from the holder after the survey, a continuous trace of pressure versus time is
obtained as shown in Fig. 2.4b, for a typical pressure buildup survey.

Fig 2.4 here ??????

2.1.1.2 Electronic Gauges and Problems
For BHP surveys, the electronic gauges have now replaced the Amerada gauges. The
electronic gauges are more sensitive and accurate, but they are also more delicate and
require regular calibrations. Figures 2.5 to 2.12 show pressures measured with
electronic gauges and problems associated with the measurements. The problems
include gauge shifts, vibrations, synchronization, failure, etc. Discussion of some
of the measurements and problems are as follows:

a) Good Measurement: Figure 2.5 shows good pressure measurements from the
lower and upper gauges taken with electronic gauges during a buildup test in a
well in the Niger Delta region. The bottom graph in Fig. 2.5 is the pressure
difference between the lower and upper gauges. The implications of the pressure
difference will be discussed in the section on quality control of measured pressure.

23

Fig. 2.5: Good Measurements with Electronic Lower and Upper Gauges

b) Pressure Shift: Pressure shifts are drastic changes in pressure not cause by
pressure changes in the environment of measurement. Figure 2.6 shows an
example of pressure shift in the lower gauge at about 20.7 hours. Pressure shifts
are common with electronic gauges and are mainly caused by change in the
sampling frequency of the gauges.

In most cases, the pressure shifts are less than 2 psi, but such pressure shifts cause
distortion of the pressure derivative and this could lead to misinterpretation of the
test.

To minimize the effect of pressure shifts, it is recommended that sampling
frequencies of should not be changed during the critical periods of the test. For
tests in high permeability formation, the critical periods are 30 minutes prior to a
rate change and 1hour after the rate change. Generally, the sampling frequency of
the gauge should not be changed during the transient state phase.

24

Fig. 2.6: Pressure Shift in the Lower Gauge

c) Pressure Vibration: This is erratic variation in pressure readings, which are
caused by known and unknown factors. Figure 2.7 shows two situations where
pressure variations occurred during buildup tests. In the situation shown in Fig.
2.7, the variations occurred in the two gauges and we suspect that the variations
were caused by vibration or local phenomena in the wellbore. We believe that it
is not a reservoir response.

Pressure vibration could completely mar a test if it occurred during the critical
periods of the test. This will be discussed in Chapter ? In some of the wells,
pressure vibration always occurred irrespective of the time of test and gauges used
for the test. This implies that some unknown factors may be responsible for the
vibration.

Fig 2.7 here ????
Fig. 2.7: Pressure Vibration in the Lower and Upper Gauges

d) Gauge Sensitivity: Although electronic gauges are sensitive, but we have seen
cases where electronic gauges could not precisely discern the pressure correctly.
Figure 2.8 shows a case where the upper gauge had sensitivity problem while Fig
2.9 shows case where both gauges had the problem. Sensitivity problem is purely
a gauge problem and could be remedied by calibration.

Put Fig 2.8 and 2.9 here here ?????

d) Gauge Synchronization: It is recommended that two gauges be run in tandem
during any test. Originally, this procedure was adopted because the second gauge
will serve as a backup if one of the gauges failed. In addition to this, pressure
25
readings with two gauges are used for quality check and also detecting wellbore
phenomenon such as phase segregation effects.

Gauge synchronization simply means that clocks used in both gauges must give
same readings every time. Figure 2.10 shows pressure readings when the clocks
are not synchronized. Same events are recorded to have occurred at different
times by the gauges. Pressures in Fig. 2.10 are from a surging well and the
peaks and lows of the surge occur at different times. This is due to the non-
synchronization of the times and this will affect the results of the quality check.
Unsynchronized pressure readings could be corrected with welltest analysis
software.


Fig. 2.10: Unsynchronized Pressure Readings

e) General Gauge Failure: A gauge is considered to have failed if the pressure
recordings are not realistic. In addition to the enumerated gauge problems, there
are other problems that could result in gauge failure. Figure 2.11 shows an
example where the upper gauge failed and therefore gave pressures that are higher
than that of the lower gauge. Figure 2.12 shows case where one of the
temperature elements of the gauges failed and measured temperatures are erratic.

In many cases, gauge failure can be remedied by re-calibrating the gauges. The
BHP contractors are therefore advised to perform a quality check on the BHP data
after each test to ascertain whether the gauges can be used for the next test. Even
in situations where the gauges have performed well, it is generally required that
gauges be re-calibrated after every six months.

26

Fig. 2.11: Gauge Failure Upper Gauge Giving Higher Pressures than the Lower
Gauge


Fig. 2.12: Temperature Element Failure

2.2 GAUGE QUALITY CHECK PROCEDURE
Checking the quality of gauge measurements is one of the best things that have
happened in welltesting. Gauge quality check (QC) now enables us to do the
following:

a) Display the entire readings of the gauges and determine which of the gauges
obtained a more reliable pressure data. Some of the gauge anomalies have been
shown.
b) Determine whether gauge readings are consistent.
c) Determine wellbore phenomena such as phase segregation, fluid interface
movement, etc.
27

Quality check can easily be performed on data obtained with digital instruments such
as quartz gauges or strain gauges because the measured values are now known
precisely. This is not so for values read from Amerada chart as such values may
depend on the person that read the Amerada chart.

2.2.1 Quality Check Procedure: There are software that can be used for quality
check. The procedures for the QC are summarized as follows:
a) Load pressure data from both gauges.
b) Synchronize the data if they are not synchronized.
c) Using the zoom option, enlarge the different sections of the test so that you can
see fine details. We are able to see abrupt pressure changes of about 0.1 psi
caused by changing the sampling frequency of the gauges!
d) Take readings of lower and upper gauges corresponding to specific events.
The events that you may choose in a buildup are a flow period, early buildup, mid
buildup, late buildup and a static stop.
e) Plot the pressure difference between the lower gauge and the upper gauge.
f) Interpret pressure difference plot
g) Repeat procedure for temperature data from both gauges. Some of the
pressure anomalies may be caused by temperature changes.
h) Fill the quality check report sheet. Table 2.4 is a sample copy of QC report sheet.

Table 2.4 here ?????

Inferences that could be made from the pressure difference plot for two gauges that
were placed 4 ft apart are as follows:

1. The pressure differences will reflect the fluid in the 4-ft. column between the
gauges.
2. If the 4-ft column was filled with gas, the pressure difference will be about 0.4 psi
and if the column between the gauges was filled with oil, the pressure difference
between the gauges will be about 1.5 psi
3. Ideally the maximum pressure difference will be 4 ft x 0.433 psi/ft (water
gradient) = 1.732 psi.
Getting pressure differences as high as 3 psi in a situation where the gauges are 4 ft
apart shows that something is wrong with one of the gauges or both gauges. Note that
some companies run their gauges 6-ft apart. This implies that the expected maximum
pressure difference will be about 2.6 psi.

2.2.2. Quality Check Plots: Some quality check graphs are discussed in this section.
The graphs show effects of phase segregation, liquid interface movement, leak and
unrealistic pressure readings.

Gas Phase Segregation: Figures 2.13 and 2.14 show pressure and temperature
differences from two tests. In Fig.2.13, the pressure difference between the upper and
lower gauge was constant and this implies that there was no gas segregation. In Fig.
2.14, there is evidence of gas segregation. The pressure difference was initially low
and later increased as gas bubbled out leaving a denser fluid. Detecting this
phenomenon helps while analyzing tests because non-reservoir responses will not be
interpreted.

Fig 2.13 and 2.14 here?????

28
Liquid Interface Movement: Figure 2.15 shows quality check in which fine details
are revealed. In this case, there was an initial buildup followed by a pressure drop and
then a final buildup. Movement of liquid interface in the wellbore caused the pressure
drop. This is discussed in detail under field cases. Such phenomenon can cause
misinterpretation of test.

Fig 2.15 here?

Leak: Part of quality check of pressure data involves enlarging different sections of
the data to view fine detail. Such fine detail easily reveals things like leaks. Figure
2.16 is a quality check plot showing the effect of leak.

Fig 2.16 here?

Unrealistic Pressure Differences: Figure 2.17 is a quality check plot showing
pressure difference of about 10 psi. This is unrealistic for gauges that are just 4 ft
apart. With this detection, we now proceed to determine the source of the error before
choosing the data to be analyzed.

Fig 2.17 here?

2.3 SAMPLE BHP SURVEY PROPOSAL
The proposal writer is the first man in the relay race to achieving the desired BHP
objective. This implies that without proper education, the proposal writer can spoil the
race. In most companies, writing BHP proposals is not considered as a job that
requires skill. Hence, trainee engineers are assigned the job of writing BHP proposals
using already existing templates. This may be a problem if they just change the test
dates without going through well files to see changes that have occurred in the wells
since the last test. They may also forget to review the success of the last test, which is
a guide to the changes required for the new test to succeed.

A typical FG/BU/SG survey proposal is in Appendix A. Some of the information in
the proposal and reasons for including such information are discussed. We shall
discuss Page A.1 to Page A.8.

Page A.1: Contains the following information:
(a) Objectives of test,
(b) Types of tests required
(c) Depth reference data (DFE, CHH).

The objectives of a tests and types of tests required are related. For example, if one of
the objectives is to optimize gaslift, a flowing gradient survey must be included as one
of the test. Also, for us to calculate false pressure, P*, skin and permeability, both
buildup and static gradient surveys must be included. Always match the type of
survey with the objectives of the survey.

Different depth references are used in the field. The drillers use the derrick floor as
his reference and therefore his depths are stated as xx ftDFE. This means xx ft below
derrick floor elevation (DFE). The BHP contractor, in most cases, gets to the well
when the derrick floor is no more there. Hence, his depths are measured with respect
to the casing head housing. That is xx ftCHH, which means xx ft below the casing
head housing.
29

The status diagram depths used in determining survey depths are stated with respect
to the DFE. Therefore the depth references must be supplied in the proposal. The
depth reference (DFE and DFE - CHH) shows the basis for calculating depths with
respect to the casing head. The BHP contractor should ensure that stated depth
references agree with what is in the status diagram.

Note that both the driller and the BHP contractor use along hole (ah) depth, which is
different from vertical depths used by reservoir engineers. The vertical depths are
measured with respect to the sub-sea (SS) level. The different depths and their
interpretations are as follows:

XX ahftDFE: XX ft along hole below the derrick floor
YY ahftCHH: YY ft along hole below the casing head housing
ZZ ftSS: ZZ vertical ft below the sub-sea level

Page A.2: Main information on this page of the proposal is as follows:
(a) Production rate
(b) Perforated Interval
(c) Location of sleeve and nipple

The rate supplied in the proposal is the flowrate at the time of writing the proposal.
The BHP contractor should always check with the flowstation staff to confirm that the
stated rate is the current rate. With this, we can know the current state of the well.
There is no point running a flowing gradient and buildup surveys in a well that is not
flowing. Even in flowing wells, the contractor should check the rate stated in the
proposal and ascertains that the gauges can be run in hole while the well is flowing.
The proposal states on Page A.4 that if the gauges cannot be run hole at the current
rate, the rate should be adjusted (i.e. change bean size) 24 hours before test. Changing
bean size requires that the BHP contractor contacts the production staff. We expect
that the BHP contractor will be guided by his experience

The rate stated in the proposal is just for information and does not mean that during
the flowtest, the bean size has to be adjusted to obtain the rate in the proposal. The
flowtest preceeding the buildup provides the rate required for analysis of test.
Therefore the rate must be measured during the flowtest. Measuring the flowrate
correctly in a BHP test is as important as the pressure measurements. We therefore
recognize the important role that flowstation staff plays in giving us accurate rate
data.

Perforated interval is an important data because it gives the interval of interest in the
survey. Also, we are interested in the pressure at the middle of the perforation. In
tests run in the long string, the gauges can be lowered, in most cases, to the top of the
perforation. This is not possible for tests run in the short strings because of the
Amerada stops, which will not allow gauges pass through them.

In all cases where the gauges can be lowered to the top of the perforation, the BHP
contractor is advised to do so because doing so will minimize phase segregation
effects and also errors in correcting to the top of the perforation. In situations where
the gauges cannot be lowered to the perforations, static gradient stops should be taken
at short intervals so that type of fluid at bottom can be determined. We need this
information for correcting pressure to the top of the perforation.

30
The BHP contractor should note the positions of the nipple and sleeves, as they are
needed for depth control. Depths in the proposal must be verified with the status
diagram. Also, the speed at which the wireline is lowered should be reduced in areas
where there are restrictions in the well to avoid hitting the gauges against objects.
Gauges are sensitive and can easily be damaged.

Page A.3: This page gives information on the following:
(a) Gauge specification
(b) Sampling rate
(c) Depth Control procedure

The BHP contractor should ensure that their gauges meet stated specifications.
Usually, we recommend that two gauges be used in any survey. This must be adhered
to because readings from both gauges can be used to detect wellbore phenomenon and
also for data quality check. Also, in a situation where one gauge fails, we can still
rely on data from the other gauge.

Using the sampling rate given in the proposal in important. Many changes in
sampling rate during a survey is not welcome because we have observed that
changing sampling frequency could cause pressure shifts that are as much as 3 psi.
Such pressure shifts cause discontinuities that may make analysis of such test difficult
and hence obtained results may be unrealistic.

We wish to encourage BHP contractors to consider depth control as a very important
issue because the depth at which pressure was measured is as important as the
pressure measurement. The contractors are advised to follow the recommended
procedure and document locations of sleeve and nipple if they are different from what
is in the status diagram. If the BHP analyst knows the measured depths he can correct
the recorded pressure to the reference depth.

Page A-4: This page contains information on the following:
(a) Well conditioning
(b) Gauge Quality Control

Well conditioning is required in some cases to ensure that well produced at stable
conditions before shut-in. Normally, no time should be spent on well conditioning if
rate was not changed while the gauges were run in hole or during the flowtest. This is
the reason why all required rate changes should occur at least 24 hours before test
commences.

Quality checks on gauge measurements by BHP contractors are required so that
anomalies can be detected early and faulty gauges removed. This implies that BHP
contractors should purchase quality check software. In addition, operating companies
should also perform their own quality check.


Page A-5: This page shows the flowing gradient stops and the duration of the stops.
There must be at least two stops around the gaslift mandrels if the flowing pressures
will be used in vertical lift performance studies and determining correct gaslift valve
locations.

Page A-6: This page contains information on the following
(1) The duration of the flowtest.
31
(2) The duration of the build up test.
(3) Data to be gathered during the flowtest.

The duration of the flowtest does not include whatever time it takes to run in hole and
perform the flowing gradient survey. The gauges must be at the final survey during
the flowtest. If the gauges are moved, the flowtest should be repeated.
Measurements to be taken during the flowtest include the THP, THT, GOR and
flowrate. These parameters should be measured every 15 minutes during the flowtest.

In Page A-6, there is also instruction on the need to secure the well with the gauges at
fixed location in cases where the well will be shut-in overnight. This is to ensure that
there is no slippage or tampering.

Page A-7: This gives information on the location of the static gradient stops and the
duration of the stops. It is important to note the following:
(a) The distance between stops should be smaller around the final survey depths.
This is important because we need to accurately determine the fluid gradient
required for pressure correction.
(b) There must be, at least two stop around the gaslift mandrels so that the gas/oil
contact (if any) can be determined accurately.
(c) Measurement of the static oil gradient is useful for estimating the oil density in
the reservoir.

Class Exercise and Discussion
A sample of the FG/BU/SG survey proposal that has a problem is enclosed in
Appendix B for discussion. The problem with this proposal is that gauges were
moved after the flowtest. This was because of confusing instruction in the proposal.

2.4 PROCEDURE FOR RUNNING FG/BU/SG TEST
A summary of the procedure for running FG/BU/SG survey is as follows:
1. Perform dummy run to determine hold-up-depth (HUD). Locate accessory closest
to the interval of interest (eg. XN-Nipple) and mark the wireline. Why?.
2. Run in hole both gauges to specified depths while well is flowing. Use wireline
mark as depth control.
3. Suspend gauges at final survey depth and allow flow to continue for the specified
period which is usually between 2 to 6 hours. Obtain the flowrate figures from the
flowstation staff.
4. Shut in well for buildup survey. In case of gas-lifted wells, shut off gas supply
prior to shutting in.
5. At the end of the specified buildup period, start the static gradient survey, which
involves measuring static pressures at specified depth.

2.5 USEFUL HINTS ON PROPER TESTING OF WELLS
(a) The gauges must be in good conditions to record accurate pressure.
(b) Depths where the pressure measurements were taken must be known.
(c) Correct stable flowrate must be known during flowing gradient survey and prior
to shutting in.
(d) Sufficient gas must be available for reliable stable flow of gaslifted intervals.
(e) Survey programme must be correct, understood and followed.


32
2.6 PRACTICAL HINTS
1. Report events such as leaks, gauge movements that occur during the tests. The
golden rule is that it is better to report than to cover up! There is no blame. It just
means that we can interpret the test with the actual information rather than being
puzzled by an inconsistency. Figure 2.18 and Table 2.4 show example of truthful
reporting while Fig. 2.19 and Table 2.5 show strange pressure response which does
not agree with test sequence.
2. Report activities in nearby wells if you are aware of them.
3. Once you have started the flowtest prior to shutting in, do not move the gauges
even if you just realized that you are not at the correct depth. Just record the correct
depth of the gauges when the flowtest started.
4. Do not change rate during the test. Any rate adjustment should occur 24 hours
before test.
5. Avoid activities that will cause unnecessary vibration of gauges.
6. Ensure that test well is correctly hooked on to the test separator during the flowtest.

Fig 2.18 and table 2..4/ 2.19 and 2.5 ????????? ask

2.7 ROLES OF FIELD STAFF IN BHP SURVEY
Field staff involved in BHP survey is production staff and BHP contractor staff. The
roles played by both are vital for good data to be obtained. Also, both production and
BHP contractor staff need to understand the principles involved in BHP test and
analysis. This will help them understand the importance of their roles.

The test and analysis principles are illustrated in Fig. 2.20.



Fig. 2.20: Test and Analysis Principles

The test principle involves allowing some known rate changes to occur in the
reservoir and measuring the resulting pressure changes. Note that some
characteristics of the reservoir such as the permeability are not known.

Analysis principle involves applying the same rate changes to a mathematical model
whose characteristics are known and observing the resultant pressure changes. The
model characteristics can be changed until the pressure changes from the model
become equivalent to pressure changes obtained when the same input (rate changes)
were applied to the reservoir. We now conclude that the model characteristics are
equivalent to the reservoir characteristics.

The implications of the test and analysis principles are as follows:
1.) Rate changes (input) are needed to create pressure changes (output).
Reservoir
k? s?
Pressure change
(Output)
Same
Rate change
Rate change
(Input)
(Input)
(Output)
Pressure change
Test Principle
Analysis Principle
Model
k, s, etc known
33
2.) Correct rate (input) applied to the reservoir must be known.
3.) Unrecorded rate changes will render test useless in some cases.
4.) Correct pressure changes caused by rate changes must be measured.
5.) Factors such as leak, gauge movement, etc. cause pressure changes not
associated with rate changes.

Based on deductions from test and analysis principles, the roles of production staff
and BHP contractors are as follows:
Roles of Production Staff
Production staff benefits from results obtained by analyzing BHP tests because such
tests are used for good well and reservoir management. We are sure everyone is
happy when wells are producing at optimum rate. The production staff can contribute
very much to the success of any BHP survey by doing the following:
1.) Ensuring that a full flowtset is conducted prior to shutting in the well for
buildup. Note that the BS&W and GOR are also required during test analysis.
2.) Ensuring that the correct well is hooked on for test.
3.) Reporting everything that could introduce errors in recorded production rate
(e.g. zero rate tests, changes to gaslift availability, surging).
4.) Telling the BHP contractor the correct status of well before survey starts.
5.) Taking more interest in BHP surveys. After all, we are all partners in
progress.
Roles of BHP Contractor Staff
The success of BHP survey depends largely on the field staff that runs the test. A few
things the field staff should be aware of are summarized as follows:
1.) A well that is not flowing or a well that cannot be shut in will not produce the
required pressure change needed for a good buildup test.
2.) Moving gauges when they are required to be at a certain position will produce
pressure changes that will distort the correct pressure changes induced by
applied rate changes.
3.) Due to the relationship between depth and pressure, good depth control is
needed so that we can associate pressure with correct depths.
4.) Leaks are unwelcome events because they affect pressure changes.
5.) Good gauges are always needed for correct pressure measurements.
6.) Mistakes give rise to wrong interpretations and wasted resources.
7.) Cooperation with flowstation staff is necessary.
8.) Accurate reporting of all deviations from programme. If the test analyst has
accurate information, he may be able to compensate for deviations.














34
3. BASIS OF ANALYZING BOTTOM HOLE
TESTS

In this section, we shall discuss some concepts that form the basis for analyzing
bottom-hole pressure tests. The concepts include the following:
Graphical presentation of Data
Units and conversions
Flow phases and identification
Flow geometry
Typical models

3.1 Graphical Presentation of Data
Graphs are one of the preferred methods for presenting information and yet they are
often poorly made. In welltesting, many graphs of bottom-hole pressure versus time
are used to deduce desired information. In this section, we shall discuss how to make
good graphs and common types of graphs used in welltest analysis.

3.1.1 Making Good Graphs
To make a good graph involves the following:

(a) Correct Labelling: The axis of the graph must be correctly labelled with units.
For example, time (hr); pressure ( psi); Dimensionless pressure, etc.

(b) Proper Scaling: Choosing the scale of a graph is important. The scales should
be chosen such that unnecessary subdivision of the grids in the graph paper is
avoided. It is convenient to choose the grid spacing such that each unit represents 1, 2,
5, 10, etc.

(c) Data Points: Data points on any graph should be made conspicuous using
symbols such as , , , etc. The practice of representing data points with dots
should be discontinued because once a line passes through such points, the position of
the point may be obscure. We are more interested in the points and not in the line. In
addition, each data set should be represented using the same type of symbol. Different
data sets should be represented with different symbols.

(d) Title of Graph: Every graph must have a title, which briefly explains what the
graph is all about. It is now common practice to put the title at the bottom of the
graph. In some cases, some information may be put on the graph in the form of legend
or something to make the graph understandable.

3.1.2 Types of Graphs
The three common graphs used in well testing are the Cartesian, semilog and log-log
graphs. The Cartesian, semilog and log-log graphs are included as Specimen A, B and
C respectively. Further discussion on these graphs follows:

Leave 3 pages for the specimens ????
35
36
37
38
(a) Cartesian Graph Paper: This is used for graphing data set (x,y) of the form

y = a + m x 3.1

Equation 3.1 is simply the equation of a straight line where a is the intercept while
m is the gradient (slope). The intercept is the value of y when x = 0 while the
gradient is calculated as change in y (y) divided by change in x (x). The units of
the intercept a, and gradient m are

a = intercept (units of y) and m = gradient
unitsof y
unitsof x
.


(b) Semilog Graph Paper: This is used for graphing data set (x,y) of the form

y = a + m log x 3.2

The advantage of using the semilog graph for plotting data set that satisfy Eq. 3.2 is
that on the semilog paper, the abscissa (x-axis) is already in log scales. Hence the data
points can just be plotted without finding the logarithm of x. From Eq. 3.2, it can be
shown that the value of a corresponds to the value of y when x is equal to one. This
follows from the fact that log 1 is zero. The gradient m, in Eq. 3.2 is defined as

m = change in y per cycle of x

Mathematically,


1 2
1 2
log log x x
y y
m

=

Generally, x
2
and x
1
are chosen so that they are one cycle apart. In that case, the
denominator is unity.

The major grids on a log scale are scaled to the powers of ten. That is, 10
-2
, 10
-1
, 10
o
,
10
1
, 10
2
, etc. The interval between 10
x
and 10
x+ 1
, where x is an integer is known as a
cycle. Note that a data set that satisfied Eq. 3.2 may still be graphed on a Cartesian
graph to obtain a straight line. In that case, the transformation

X = log x 3.3

is required. Hence the final equation becomes:

y = a + m X 3.4


(c) Log-log Graph Paper: This type of graph is used for data set of the form:

log y = a + m log x 3.5

The gradient m, in Eq. 3.5 is given in cycles of y per cycle of x. The gradient may be
calculated using the equation:

39

m =
1 2
1 2
x log - x log
y log - y log
3.6

The points (x
1
, y
1
) and (x
2
, y
2
) are taken arbitrarily from the straight line. The log-log
graph paper is already divided into log scales and the data points are graphed directly
without finding the logarithm of anything. By suitable transformation, the data set
that satisfy Eq. 3.6 may be graphed on a Cartesian graph to get a straight line but the
slope of the straight line will be different from that obtained from Eq 3.5.


3.2 SYMBOLS, UNITS AND CONVERSIONS
In this section, we shall discuss symbols, units and unit conversions.

a) Symbols and Units
The three unit systems used in welltesting are the CGS, Oilfield and SI units.
Although the Oilfield units are more common, but the preferred unit is the SI unit.
Some of the welltest parameters, symbols and units are shown in Table 3.1. For more
symbols and units, refer to SPE Metric Standard (1982). We assumed that you are
already familiar with basic definitions of reservoir properties.

Table 3.1: Symbols and Units
Parameter Symbol CGS Units Oilfield Units Practical S.I. Units
Liquid Flow Rate q cc/sec STB/day
m
3
/day
Permeability k Darcy millidarcy
(mD)
millidarcy (mD)
Time t seconds (s) hours (hr) seconds (s)
Liquid Viscosity centipoise centipoise millipascal-second
Compressibility c
atm
-1
psi
-1
kpa
-1

Wellbore Storage C
s
cc/atm bbl/psi
m
3
/kpa
Porosity fraction of bulk volume same as CGS same as CGS
Saturation S fraction of pore volume same as CGS same as CGS
Pressure p atm psi kpa
Thickness h cm ft m
Skin Effect s dimensionless dimensionless dimensionless


b) Conversion Factors
Some basic conversion factors used in this welltesting are as follows:

1 atmosphere (atm) = 14.7 psi = 1.01325 x 105 kpa
1 cp = 1 x 10
-6
kpa-s
1 barrel (bbl) = 1.589873 x 10
-1
m
3

1 bb1/day = 1.840131 x 10
-6
m
3
/s
= 1.840131 cc/s

More conversion factors are given in Table 3.2 taken from Earlougher (1977).


Leave two pages for conversion factors ??????

TABLE 3.2-CONVERSION FACTORS USEFUL IN WELL TEST ANALYSIS.
40

SI conversions are in boldface type. All quantities are current to SI standards as of
1974. An asterisk (*) after the sixth decimal indicated the conversion factor is exact
and all following digits are zero. All other conversion factors have been rounded.
The notation E + 03 is used in place of 103, and so on.

41
42
43
c) Unit Conversion
The Society of Petroleum Engineers (SPE) recommends the use of S.I. units. Articles
that appear in SPE journals are now written using both the SI units and the Oilfield
units. The Oilfield units are also referred to as English units.

Conversion from one set of units to another is easy. Two cases will be discussed here.
The first case considers conversion of a quantity expressed in one unit to another unit.
The second case considers conversion of an equation with parameters expressed in
some units to an equivalent equation with parameters given in different units. In both
cases, what is required is a conversion factor. Consideration for given cases follows:

Case 1: Conversion of a given quantity from one unit to another

Example 1
(a) ft inches

3ft = 3ft x
12 inches
ft 1



old conversion
units factor
(first)

= 36 inches

Example 2
bb1 ft
3

5 bb1 = 5 bb1 x
5.614583
b b l
ft
3
( 5.61 x 5) ft
3


Note that in both cases, if the units in the numerator and denominator are cancelled,
the resultant unit (unit to remain uncancelled) is the new unit.

Case 2: Conversion of units in an equation

Example 1

q =
K A p
L

3.7

Equation 3.7 is the steady state form of Darcys law for a linear system. The
parameters in Eq. 3.7 are in Darcys units.


That is:


(cm) L (cp)
(atm) p (cm) A (D) k
=
s
cc
q ) (

3.8

Equation 3.8 can be converted to Oilfield units. That is:
44

q
) (
s
cc
q (
STB
day
)

k (D) k (mD)

A (cm
2
) A (ft
2
)

p (atm) p (psi)

(cp) (cp)

L (cm L (ft)


Note that while the flowrate in Darcys units is expressed at reservoir conditions, in
Oilfield units, it is expressed at stock tank conditions. Conversion of Eq. 3.8 to
Oilfield units is done by replacing all terms in the original equation (Eq 3.8) with
other terms and combining to get the overall conversion factor. The terms in Eq. 3.8
and their replacements are as follows:

Parameters in Original Equation Replacements

k (D) K (mD) x
1 (D)
1000 (mD)


A (cm
2
) A (ft
2
) x
30.48 (cm
(
2
)
1 ft )
2


p (atm) p (psi) x
1
147
( )
. ( )
atm
psi


(cp) (cp)

L (cm) L (ft) x
30.48 (cm)
1 (ft)



q(cc/s) q (STB/day) B
o
bbl
STB
x
1 day
(24x3600)sec
158987.3 cm
1 (bbl)
3

(
x
Substituting,

q (
STB
day
x B (
bbl
STB
x
1 day
(24 x 3600) secs
o
) ) x
0.158973m
bbl
x
100 cm
m
3 3 3
3


= k(mD) x
D
1000mD
x A(ft ) x
(30.48) cm
ft
) x
30.48cm
ft
2
2 2
2
(cp) L (ft

x p (psi) x
(atm)

14 7 . psi
3.9
45

Evaluating, Eq. 3.9 becomes

qB
STB
day
x 1.84013 = 0.002073469
k (mD) A (ft p (psi)
(cp) L (ft)
2
o

(
)

3.10



and therefore,

q (
STB
day
x 10
K(mD) A(ft p (psi)
(cp) B
(bbl)
(STB)
L(ft)
-3
2
o
) .
)
= 1 127

. 3.11

The conversion factor 1.127 x 10
-3
should be familiar to those who have taken basic
reservoir engineering courses.


Example 2

Dimensionless pressure is defined in Darcys unit as:

P
D
=

q
P] - [P 2
i
kh
3.12

The term P
D
in Eq. 3.12 may for now be considered to be just a dimensionless number
(pressure). Equation 3.12 is in Darcys unit. That is:

P
D
(dimensionless) =
) (
(S)
(cc)
q
(atm) p (cm) h (D) 2
cp
k

3.13

Conversion of Eq. 3.13 to Oilfield units is done as follows:

P
D
(dimensionless) = 2k(mD) x ( )
.48 1
1000
30 D
mD
h ft x
cm
ft
x

x
p psi psi
q
STB
day
xB
bbl
STB
x
day
x s
x
o
atm
14.7
( ) ( )
( )sec 24 3600
01589873 10
6 3
. x cm
bbl
x cp 3.14

P
D
(dimensionless) = 2 x 0.0011268
K mD h ft p psi
q
STB
day
B
bbl
STB
cp
o
( ( )
( )

=
kh p
q B
o

1412 .
3.15

46
This implies that dimensionless pressure, defined in Darcys unit as

P
D
=
2

kh p
q

, 3.12

is given in oilfield units as

P
D
=
kh p
qB
o

1412 .
3.16

Similarly, it can be shown that dimensionless time t
D
is defined in Darcys units as:

t
D
=
kt
c r
t w

2
3.17

is given in Oilfield units as

t
D
=
0 000264
2
. kt
c r
t w

3.18

The time, t, in Eq. 3.18 is in hours. If it is in days, the conversion factor will be
0.00634. Note that in converting from Darcys to Oilfield units, nothing should be
done to the dimensionless parameters P
D
. Why? They are dimensionless.

In Case 2 where the parameters in an equation are converted from one set of units to
another, if the units in the numerator and denominator are cancelled, the old units will
remain. This is the difference between Case 1 and Case 2.

d) Rule of Thumb for unit conversion
Welltest equations usually contain dimensionless groups. The ability to recognize
these groups and their equivalent in the different unit systems forms the basis of the
rule of thumb used in unit conversion. Table 3.3 shows the groups and their
equivalent in different unit systems.

Table 3.3: Definitions of Dimensionless Groups in Different Unit
Systems
Parameter CGS Oilfield SI
Dimensionless
Time, t
D

kt
c r
t w

2

0.000264
2
kt
c r
t w

2
6
10 557 . 3
w t
r c
kt x

Dimensionless
Time Based on Area, t
DA

kt
c A
t

0.000264 kt
c A
t

A c
kt x
t

6
10 557 . 3

Dimensionless Pressure, P
D
2

kh p
q


kh p
141.2qB

qB x
p kh
3
10 866 . 1


Dimensionless Distance r
r
w

r
r
w

r
r
w



The use of the rule of thumb is illustrated with examples.
47

Example 1
Convert the equation

q =
2

kh p
r
r
e
w

ln


from Darcys unit to oilfield units.

Solution
Rearrange the equation into recognizable dimensionless groups as follows

ln
r
r
kh p
q
e
w
=
2

(Form of Dimensionless Distance) (Dimension Pressure)

Therefore, the equation in oilfield unit is

ln
r
r
kh p
qB
e
w
=

141.2


Rearranging

q =
7.08 10
3
x kh p
B
r
r
e
w

ln


Example 2

Convert the following equation in Darcys unit to Oilfield unit.

P = p
i
+ )
4
(
4
2
kt
r c
E
kh
q
t
i

Solution
Rearranging to bring out the dimensionless group,

2 x
[ ] 2
4
2

kh p p
q
E
c r
kt
i
i
t

=
|
\

|

Recognizable dimensionless groups and their oilfield equivalents are as follows


[ ] [ ] 2
141.2qB

kh p p
q
kh p p
i i

c r
kt
c r
kt
t t
2 2
0.000264


48
Substituting, the oilfield equivalent of the equation is


[ ] 2
141.2qB 4x0.000264
2
kh p p
E
c r
kt
i
i
t

=
|
\

making p the subject gives

p = p
i
+
70.6qB
0.001056
2

kh
E
c r
kt
i
t

|
\

|

3.3 Dimensionless Forms
Many dimensionless parameters are defined in petroleum engineering. The
dimensionless parameters make it possible to cast fluid flow equations into
dimensionless forms. The advantages and disadvantages of the dimensionless forms
are given in this section. Also, the definitions of some of the dimensionless
parameters are given.

a). Advantages of Dimensionless Forms
(i) Ease of comparing solutions
(ii) Makes it possible for results to be generalized.

For example, the pressure at any point in a single well reservoir produced at constant
rate q, is written in a general form using dimensionless pressure as:

P
i
- P(r, t) =
1412 .
, , ,
qB
kh

[P (t r C geometry ) + s]
D D D D
3.19

Equation 3.19 is in Oilfield units and the dimensionless pressure, P
D
, is a function of
other dimensionless parameters, r
D
, t
D
and C
D
.

(iii) With pressure expressed in dimensionless form, It becomes easier to apply
superposition concept to handle varying flowrates and pressure drop in multiwell
systems.

(iv) Dimensionless form aid in presentation of results in a more compact form.
Also, the results are invariant in form, irrespective of the unit system used.

b) Disadvantages of Dimensionless Forms
(i) With dimensionless form, the engineer may loose a sense of magnitude of the
quantity.

For example, a time of 24 hours may correspond to dimensionless time of 300 for a
tight reservoir and a dimensionless time of 1000 for a highly permeable sand. This
follows from the way dimensionless parameters are defined.

3.3.1 Definition of Dimensionless Parameters In Darcy Units

Dimensionless Time
t
D
=
kt
c r
t w
2

3.20

49
Dimensionless Time Based on Area
t
DA
t
=
kt
c A
3.21

Dimensionless Radius
r
D
=
r
r
w
3.22

Dimensionless Pressure
P
D
=
2 kh
q
(P - P (r, t) )
i

3.23

P
D
=
P - P (r, t)
P - P
i
i wf
3.24

The form of dimensionless pressure, P
D
, used in any problem depends on the type of
boundary condition used at the well. Use Eq 3.23 if well produces at constant
pressure and Eq 3.24 for well producing at constant rate.

Dimensionless Cumulative Production

q
D
=
q (t)
2 kh (P - P
i wf

)


Q
D
o
t
DW
= q dt
D DW
3.25

Other Dimensionless Parameter

kt
c r r
t D
r
=
kt
c r

r
=
t
2
t w
2
w
2
D
2 2
3.26


3.4 FLOW PHASES
Just like a man goes through phases in life, a well undergoing a test also passes
through phases with each phase revealing some information about the well or
reservoir. It is therefore necessary that we understand the phases and how to identify
them to avoid problem of obtaining required information from a wrong phase.

In this section, we shall describe the phase and methods of diagonizing the phases.

3.4.1. Flow Phases in Drawdown and Buildup Tests
Figure 3.1 shows the phases that a well undergoes during a drawdown or buildup test.
In case, there is a wllbore storage phase, transient phase and late-time phase. The
late-time phase depends on the nature of the boundary and test (drawdown or
buildup).




50
Wellbore Storage
A B C D E
Transient State Phase Late-Time Phase
Increasing Flow Time

Figure 3.1: Flow Phases in Buildup or Drawdown Test.

Some deductions from Fig 3.1 are as follows:
1. The wellbore storage phase is an independent phase and can occur concurrently
with the transient state. This happened in Interval BC.
2. The wellbore storage phase with no concurrent transient phase is known as the
strong wellbore storage. This is represented by Interval AB.
3. If wellbore storage and transient state occur concurrently, the pressure-time data
acquired will be polluted and cannot be analyzed using conventional method of
analysis.
4. The wellbore storage phase may last so long that all the transient state phase may
be polluted. Tests must be designed so that this type of pollution does not
occur as the most valuable information from such tests are obtained from the
unpolluted transient state phase.
5. The transient state phase without concurrent wellbore storage phase is the good
transient state phase. This is represented by Interval CD.
6. The transient state phase and late-time phase are dependent. The transient state
must end before the boundary effect phase is reached.
7. For the buildup test, The late-time phase represents the stabilization phase.
During this period, the well builds up to average pressure if interference from
other wells is minimal.
8. For drawdown test, the late-time phase represents the pseudo-steady state phase if
the reservoir boundary is closed to inflow (bounded reservoir). On the other hand,
if the reservoir boundary is open to inflow, the late-time phase will represent the
steady state phase.
9. Information derived from the pressure time data obtained during the different
phases is shown in Table 3.4.

Table 3.4: Information from Different Phases
Phases Derived Information
1. Strong Wellbore Storage Wellbore Storage Constant, Cs
2. Good Transient State (1) Permeability, k
(2) Skin Factor, s
3. Pseudo-Steady State Drainage Volume, Vd
4. Stabilization Phase Average Pressure

3.4.2 Definition of the Phases
The function of the pressure gauge is simply to measure pressure irrespective of the
flow phase occurring in the well. Depending on the information we need from the
test, we need to ascertain that the acquired pressure-time data used in the analysis
correspond to the phase of interest.

51
In this section, we shall define the phases and the characteristic features of pressure-
time data obtained during the phases.

Wellbore Storage Phase: This occurs early in the life of the test well. The pressure
changes that occur during this phase are caused by fluid stored in the wellbore or
stored fluid produced from the wellbore. This is caused by the fact that wells are
opened or shut at the surface during tests and also reservoir fluids are compressible.

Figure 3.2 shows a test well that has just been shut in. Note that the production rate
(q) is zero at the surface, but the reservoir is still producing (q
sf
0) and the produced
fluid is stored in the well. The duration in which the surface rate is zero and the
sandface rate is not zero is the wellbore storage phase for a buildup test.


q = 0
q = 0
q
sf
= 0
q
sf
0
Wellbore Storage
phase
End of wellbore Storage
phase
Increasing Shut-in time

Figure 3.2: Surface and Reservoir Production During Shut-in.

Figure 3.3 shows wellbore storage phase in a drawdown test. In this case, on opening
the well for a drawdown test, the initially produced fluid is the fluid stored in the
wellbore. Thai is, q = q
wb
. Wellbore storage phase ends in this case when the total
production is equal to the fluid produced by the reservoir. That is, q = q
sf.




52
q = q
wb
+ q
sf
q = q
sf
q
sf
= q
q
wb
= 0
q
sf
= 0
or
q
sf
< q
Wellbore Storage
phase
End of wellbore Storage
phase
Increasing production time
0
q
wb

Figure 3.3: Surface, Wellbore, and Reservoir Production in an Opened Well.

In Figure 3.3, q
wb
is the production from fluid stored in the well. Note that if q
sf
= 0
for drawdown or q
sf
= q in a shut-in well, we have strong a wellbore storage phase.
The manner in which q, q
sf
and q
wb
vary during buildup and drawdown tests are
shown in Figure 3.4.

q = q
sf
+ q
wb
q
0 time
t
s
t
e
t
e
q
wb
q
sf
q
sf
Drawdown
Buildup
Wellbore
storage
Wellbore
storage

Figure 3.4: Rate Variation During Buildup and Drawdown Tests.

In Figure 3.4,
t
e
= end of wellbore storage phase for drawdown test
t
s
= shut-in time
t
e
= End of wellbore storage phase for buildup test.

53
The wellbore storage phase is a nuisance because it pollutes the transient phase
from which we can get useful information about our reservoir. The duration of the
well storage phase must be reduced or eliminated if it is possible.

To reduce the duration of wellbore storage phase will require knowledge of the
factors that can affect the phase. The factors are mainly compressibilty of the
wellbore fluid and wellbore volume.

(a) The compressibility of the fluid in the well: The higher the compressibility of the
fluid, the longer the duration of the wellbore storage phase. The compressibility
of the wellbore fluid depends on the gas-oil ratio (GOR). Wells with high GOR
has high wellbore fluid compressibility . Can the GOR be reduced?

(b) The volume of the well that communicate with the tubing: This is shown by the
volume of the shaded region in Fig. 3.5. Figure 3.5a represents a case with a
packer while Fig 3.5b represents a case without a packer or a non-sealing packer.
The duration of the wellbore storage phase increase with the increase in volume
communicating with the tubing.


(c) Production Rate: The production rate of the well also affects the duration of the
wellbore storage phase. The higher the production rate, the smaller the duration
of the wellbore storage phase.


(a) Sealing packer (b) No packer or non sealing packer

Figure 3.5: Volume communicating withTtubing


Exercise: Explain how the following will affect the duration of the wellbore storage
phase:
(i) Test well was shut at the flow station because the wing valve was faulty.
Draw the phase box diagram showing situation where this could make it
difficult to calculate permeability and skin.
(ii) Test well was shut in with a downhole shut-in tool. Also, show the phase box
diagram in this case.

The practical implications of these are as follows:
54
(a) Do not shut the well at the flowstation as that will increase the volume
communicating with the tubing and thus increase the duration of the wellbore
storage phase.
(b) In wells with high gas-oil ratio (GOR) the duration of the wellbore storage phase
is long because of the high compressibility of the fluid in the wellbore.
(c) Use a downhole shut-in tool in situations with unusually long wellbore storage
duration.
(d) In buildup tests, if there are leaks, the wellbore storage phase may not end because
the sandface production, q
sf
, will not be zero.
(e) For wells producing less than 500 STB/D, the duration of wellbore storage phase
should cause some concern.

Transient State Phase: This is the most important phase because important reservoir
parameters (permeability, skin, etc.) are deduced from pressure-time data obtained
during this phase. The useful part of the pressure-time obtained during this phase is
the part not polluted by the wellbore storage phase. Due to the usefulness of this
phase, the following guidelines must be followed during tests.

(1) Design and run test so that not all parts of the transient state phase will be
polluted by the well bore storage phase.
(2) Test duration must be such that the transient state phase must be reached before
the test is stopped.

The transient state phase occur when the pressure changes at the wells are not
influenced by the nature of the boundary. For example, if you drop a little stone into
a bowl containing water, concentric waves will move outwards as shown in Figure
3.6.


Figure 3.6: Waves illustrating Transient State Period.

The waves will continue to move outwards until they hit the side of the bowl. The
waves get distorted and become less orderly. The period during which the waves
have not hit the boundary can be likened to transient state phase because the effect of
the boundary has not been felt. The mathematicians describe this phase as period
when the rate of pressure change with time is neither zero or constant. All systems go
through the transient state irrespective of the nature of the boundaries.

The duration of the transient state is affected by many factors which includes the
following:

(a) Permeability of the Formation: The higher the permeability, the shorter the
duration of the transient state phase (waves move faster). For Niger Delta
55
formation with permeabilities greater than 1000 mD, the transient state phase has
a short duration. This is a problem because the short transient state duration could
easily be marred by the wellbore storage phase.

(b) Location of Test Well: The location of the test well with respect to reservoir
boundary affects the duration of the transient state. Wells that are closer to the
boundary will have a shorter transient state period compared to wells that are
farther from the boundary. Figure 3.7 shows two cases.

Longer TS phase
Shorter TS phase

Figure 3.7: Effect of Well Location on Duration of Transient State Phase.


Late-Time Phase: The nature of the late-time phase depends on nature of test. For
buildup test, the well will build up to average pressure at late-time if there is just one
well in the reservoir. In case where we have many wells, the well will build up and
then drops due to interference effect. In a situation where the drainage area of the test
well is large, the pressure may build up to average pressure before dropping. The
pressure profile in such case is shown in Fig 3.8.














Fig. 3.8: Effect of Interference

For drawdown, the nature of the late-time phase depends on the type of the outer
boundary condition of the reservoir. If the boundary is open to inflow (water influx),
a steady state will be attained during the late-time phase. During steady state,
pressure in the system will no longer be changing with time. If the reservoir boundary
is closed to flow, a pseudo-steady state will be attained at late time. During the
pseudo-steady state period, pressure in the system will be changing, but the rate at
which the pressure will be changing everywhere in the system will be constant. The
value of the is related to the drainage volume of the test well and this is the basis of
reservoir limit test. Figure 3.9 shows the transient state, steady state and pseudo-
steady state phases during a drawdown test



Shut in
Pressure
Shut in Time
No Interference

Interference
56

.














For steady state to be attained there must be an adjoining aquifer providing the source
of water influx and the aquifer permeability must be large to permit ease of flow of
water into the reservoir. Reservoirs in highly faulted environment are more likely to
be sealed off and less likely to be in contact with such an aquifer.


3.5 Flow Geometry
Depending on the reservoir and nature of perforations in well, flow could occur
linearly, radially, spherically or elliptically. In some cases, the flow geometry could
change with time from one form to another. In this section, we shall discuss
equations governing the different flow geometries and parameters that may be
deduced from the flow geometry.

3.5.1 Linear Flow
The plan and elevation in situations where linear flow occur is shown in Figure 3.10









Plan Elevation
Fig. 3.10: Linear Flow Geometry

The linear flow described here is the one-directional flow that occurs in Cartesian co-
ordinates. The practical situations where linear flow could in a reservoir are as
follows:

a. Single Vertical Fracture Intersecting Well
This case is illustrated in Figure 3.11. The permeability in the fracture zone is much
greater than the permeability of the formation (infinite conductivity). Therefore, fluid
flows linearly into the fracture and the wellbore. Such flow occurs at early time and
will explained later.
Transien
t
Steady State
Pseudo-Steady State
Log (Flowing Time)
P
w

Fig 3.9: Transient, Steady and Pseudo-Steady States
57








Fig 3.11: Vertically Fractured Well

Linear flow is also observed at early time in situations where the flow per unit area of
fracture is constant (uniform flux fracture).

b. Horizontal Well
At some period, linear flow could occur in horizontal wells. The flow streamlines in a
horizontal well during linear flow is shown in Figure 3.12.



?????????????? Cut and paste

Fig. 3.12: Linear Flow in Horizontal Wells

c. Reservoirs with Strong Partial Influx
In a reservoir where a segment is open to strong water influx, flow will be
dominant in the direction of the influx. Depending on the size of segment, the flow to
the wellbore could be linear.

In situations where linear flow occur, the flow could be modelled with the equation:

2
2
x
ct
k t
=
3.27

Equation 3.27 is also a diffusivity equation. Hence, all assumptions relating to
diffusivity equation will hold.

The solution to Equation 3.27 is of the form

p = (At) ---------------------------------------------------------------------------
3.28

Taking log of both sides,

Log p = log t + B-------------------------------------------------------------
3.29

In Equations 3.28 and 3.29, p is the pressure change while A and B are constants
that depend on fluid and rock properties.
Reservoir
Boundary
Well

Fractur
e
x
f
58

The implication of Equation 3.29 is that a graph of p versus t (time) on a log-log
paper will give a straight line with slope . This is distinguishing feature of all forms
of linear flow irrespective of where they occur.

A similar relationship existing between dimensionless pressure and dimensionless
time for a well intersecting a single vertical fracture is given as


Dxf D
t P =
3.28

where t
Dxf
is dimensionless time based on fracture half length and is defined in
Oilfield units as

2
000264 . 0
f t
Dxf
x c
kt
t

=
3.29

Equation 3.28 also shows that a graph of P
D
versus t
Dxf
will give a slope of 0.5.

3.5.2 Bilinear Flow
This is a form of linear flow observed in some wells with single vertical features. In
this case, the permeability in the fracture is not much greater than the permeability in
the formation (finite conductivity fracture). Hence, there is linear flow to the fracture
and another linear flow from the fracture to the wellbore. This is illustrated in Figure
3.13.








Plan showing Well and Fracture Elevation showing the Well
Fig. 3.13: Bilinear Flow Geometry

For the bilinear flow, the relationship between pressure change and time is

p = A t

----------------------------------------------------------------------------
3.30

Taking log,

Log p = log t + B ------------------------------------------------------------
3.31

59
Equation 3.31 implies that a graph of p versus t (time) on a log-log graph has a
unique slope of .

3.5.3 Radial Flow
This is the most common flow geometry and occurs when the flowing fluid surrounds
the wellbore and flow streamlines come from distances that are large compared to the
size of the wellbore. Streamlines converge towards a central point in each plane.
Figure 3.14 shows the plan and elevation during radial flow.








Plan showing Well and Streamlines Elevation showing Well and Strealines
Fig 3.14: Radial Flow

Irrespective of what it is called (pseudo-radial, late-time radial, early-time radial, etc.)
the relationship between pressure and time in all forms of radial flow is

p = A log t + B -----------------------------------------------------------------
3.32

In Figure 3.14, p is pressure change, (P
i
p), t is time (or some form of time
function), and A and B are constants that depend on the reservoir characteristics.

The implication of Equation 3.32 is that a graph of p (or pressure) versus log t gives
a straight line. This is shown in Figure 3.15 for a drawdown test.













Fig 3.15: Semilog Straight Line due to Radial Flow

Radial flow could occur virtually in all system including horizontal well. In a
reservoir with a single vertical fracture, a form of radial flow, pseudo-radial flow,
occurs at the end of the linear flow. Figure 3.16 shows the pseudo-radial flow
occurring in a system with a single vertical fracture.


0.1 1 10 100
1000
200
0
250
0
300
0
Pressur
e
Time
60









Fig 3.16: Pseudo-radial Flow occurring in well with a Single Vertical Fracture

In Fig 3.16, the streamlines come from far distances and the fracture behaves as a
point source.

A form of radial flow (pseudo-radial) also occurs in horizontal wells and this is shown
in Figure 3.17.
???????????? cut and paste
Fig 3.17: Pseudo-radial Flow occurring in Horizontal Well

A form of radial flow, hemiradial flow, occurs when the well is close to a boundary as
shown in Fig. 3.18. Equation 3.32 also holds for a hemiradial flow with constants, A
and B defined appropriately.










Fig 3.18: Hemiradial Flow

3.5.4 Spherical Flow
For this case, flow occurs from all directions towards the wellbore. This is shown in
Figure 3.19.








Plan Elevation
Fig 3.19: Spherical Flow

Boundary
61
Spherical flow could occur in thick reservoirs if the perforated interval is small.
Moran and Finklea (1962) used spherical flow equations for analysing pressure
transient data. Raghavan (1975) found an expression for vertical permeability in a
partially penetrating well using spherical flow equations.

The spherical flow equations at early and long times are of the forms:

Early Time
p = At
1/2
--------------------------------------------------------
3.33

Long Time
p = 1 -
1
1
2
Bt
-------------------------------------------------
3.34

In Equations 3.33 and 3.34, A and B are constants. Graphical implications of
Equations 3.33 and 3.34 are obvious. Onyekonwu and Horne (1983) published detail
on the equations. Equation 3.34 is generally recognized as the spherical flow
equation and hence a plot of pressure versus the inverse of the square root of time
yields a straight line.

3.6 Flow Geometry and Phases
In this section, we illustrate using box diagrams the phases and geometries that occur
during transient in some well and reservoir systems.

a) System with Wellbore Storage and Skin in Homogeneous Reservoir








Fig. 3.20: Radial Flow in a Homogeneous Reservoir

b) System with Well on a Single Vertical Fracture










Fig. 3.21: Flow Geometry and Phases in a Well on a Single Vertical Fracture
Wellbore Storage Phase
Radial Flow
Transient State Phase Late-Time Phase
Increasing Time
Wellbore Storage Phase
Transient State Phase
Late-Time Phase
Increasing Time
Linear Flow Pseudo-Radial Flow
62

An implication of Fig 3.21 is that linear flow can easily be marred by wellbore storage
phase.

c) Horizontal Well in a Homogeneous System













Fig. 3.22: Flow Geometry and Phases in a Well on a Horizontal Well

Depending on the conditions, horizontal wells could exhibit other flow regimes as
explained by Du and Stewart (1992) and Kuchuk (1995). Figure 3.22 shows the
main flow regimes published by Lichtenberger (1994). Note that there is usually a
transitional flow between the flow geometry.

3.7 Distinguishing the Phases and Geometry
The pressure gauges simply measure pressure and time irrespective of the flow phase
and geometry. However, information derived from the pressure-time data depends on
the flow phase and geometry. This implies that we must be able to distinguish
pressure-time data gathered during each phase and geometry.

Until 1980, pressure plots were solely relied on for diagnosing flow phases and
geometry. This is simply based on equation relating pressure and time. Bourdet et al
(1983) introduced the concept of pressure derivative, which has been found to be
more unique and reliable in diagnosing flow phases and geometry.

The pressure derivative is defined as

)) ( ( t f dIn
dp
P = 3.35

where f(t) is a time function which may be defined as follows:

f(t) = t (drawdown test)
f(t) = t (buildup test)
f(t) = (t
p
+ t)/ t (buildup test)

The derivative of pressure drop could also be used and in that case, the pressure term
in Eq. 3.35 is replaced with p.

Wellbore
Storage Phase
Transient State Phase
Late-Time
Phase
Increasing Time
Linear Flow Pseudo-Radial Flow Radial Flow
63
As pressure values during test are obtained at discrete times, the pressure derivative is
obtained numerically. Horne (1990) published an algorithm that can be used in
obtaining the derivative as follows:

i i
t
p
t
t
p
|

\
|

= |

\
|

ln
3.36
=
( )
( ) ( )
k i j i i j i
j i k i i
t t t t
p t t
+ +
+

/ ln / ln
/ ln
+
( )
( ) ( )
k i i i j i
i i k i j i
t t t t
p t t t
+
+

/ ln / ln
/ ln
2
-
( )
( ) ( )
k i j i k i i
k i i j i
t t t t
p t t
+
+

/ ln / ln
/ ln
3.37

In Eq. 3.36 and 3.37, the time function is assumed to be time. The constraints on the
time in Eq. 3.37 are as follows:

In t
i+j
- In t
i
0.2 and In t
i
- In t
i-k
0.2

The value of 0.2 (known as differentiation interval) could be replaced by smaller of
larger values (usually between 0.1 and 0.5), with consequent differences in the
smoothing of the noise in the pressure data. Higher values yield a more smoothened
derivative.

Pressure derivatives characteristic shapes published by Gringarten (1987) are shown
in Fig. 3.23.

????? leave 3 inches

Fig 3.23: Characteristic Shapes of Pressure Derivatives

Table 3.5 summarizes the feature and the characteristics of the derivative

Table 3.5: Features and Characteristics of the Derivatives
Features Characteristics of Derivative
Wellbore storage and skin Hump with derivative attaining a maximum
Stimulated or fractured well No hump, No maximum
Heterogeneous behaviour A valley
Fault Upward turn
Closed boundary (drawdown) Upward turn
Closed boundary (buildup) Downward turn
Constant pressure
(due to influx or gas-cap)
Downward turn
Infinite-acting radial flow (IARF) flat (zero gradient)
Linear Flow Gradient with slope of 0.5
Spherical flow Gradient with slope of - 0.5
Strong wellbore storage Gradient with slope of 1 (unity)
Hemiradial flow flat (zero gradient at a higher level)

Table 3.5 will form the basis for using pressure derivative for diagnosis.

3.7.1 Diagnostic Plots for Distinguishing Flow Phases
In this section, we shall show the diagnostic features of the phases using pressure (or
pressure drop) and pressure derivative. Discussion now follows:
64

Diagnosing Wellbore Storage Phase: Wellbore storage is distinguished on a log p
versus log t (or log t) plot. The following features help us determine pressure
response obtained during this phase.

(1) Pressure responses obtained when wellbore storage is strong lie on a unit slope
line. This corresponds to time ending at t* on Figure 3.24. This follows from the
fact that

log(p) = log t + C. 3.38

For buildup, t* is replaced by t*.

(2) Pressure responses obtained when wellbore storage is not strong will have a slope
that is in the range of 0 < m < 1. The duration of this part of wellbore storage
phase is in time range of 10t* < t
ewb
< 50t* (1 to 1.5 cycle rule) where t* is the time
when strong storage effect ended. The times are shown in Figure 3.24.

(3) A typical log-log plot is shown in Figure 3.25.
Wellbore Storage
Transient
Late Time
time t*
t
esl
(Figure not to scale)
t
ewb

Figure 3.24: Phases and Duration of Wellbore Storage Phase.

















Figure 3.25: Log-log Showing Wellbore Storage Influenced Data
0.1 1 10 100
Time, t or Shut-in Time, t
1
10
100

p
45
o

t* t
ewb

65

Parameters in Figure 3.25 are defined as follows:
p = Pi - p
wf
(Drawdown)
p = P
ws
- p
wf
(t
p
) (Buildup)
t = flowing time (drawdown)
t = shut-in time (Buildup)
P
i
= initial pressure
p
wf
= flowing pressure
t
p
= total flowing time
p
wf
(t
p
) = flowing pressure at shut-in time
P
ws
= shut-in pressure.

The wellbore storage phase is more easily distinguished on a log-log plot of the
pressure derivative. On this plot the wellbore storage phase forms clearly defined
hump as shown in Fig.3.26.

Derivat ive
Wellbore St orage Hump
45
Time
Fig. 3.26: Derivative Plot - Diagnostic Plot for Wellbore Storage Phase


Since the advent of the derivative plot, I have always relied on it when detecting
pressure responses affected by wellbore storage phase. In most cases, both the
pressure and derivative plots are displayed in one graph. This is shown in Figure
3.27.

Derivat ive
and Pressure
Wellbore St orage Hump
45
Time
Fi g. 3. 27: Log-l og Plot of Pressure and Derivati ve.
Pressure
Derivat ive



Diagnosing Transient State Radial FlowPhase: Pressure responses obtained during
the good transient state radial flow phase or infinite-acting radial flow phase (IARF)
will fall on a straight line when pressure is plotted against log of time (semilog plot).
The time in this case is defined as follows:
time = flowing time for drawdown
66
time = shut-in time for buildup (MDH Plot)
time =
t t
t
p
+

for buildup (Horner Plot)

This follows from the fact that during the good transient,

p = A + m[ log t(ime) + C]. 3.39
where A and C are constants. Also, m is a constant that depends on fluid and rock
properties. The semilog plots are shown in Figure 3.28, 3.29 and 3.30.












Figure 3.28: Semilog Plot (Drawdown Test)






















Fig 3.30: MDH Plot

Log (time)
P
w
t
ew
Good
Transient
Log t
P
ws

Good
TraTransie
Log (t +t/t) 1
P
ws

Stabilizin
g at
average
pressure
Fig. 3.29: Horner
Plot
67
The good transient state phase is the transient state not influenced by wellbore storage
phase. It occurs in the time range t
ewb
< t < t
esl
. The term t
esl
is the time when
transient ends and it is shown in Figure 3.24.

Many factors affect the semilog straight line obtained during the transient state phase.
Figure 3.31 taken from Mathews and Russell (1967) show how these factors affect the
semilog straight line.

????? ask for Fig 3.31.

The good transient state phase can also be clearly discerned on a derivative plot. The
derivative plot in this case falls on a horizontal line (flat) after the end of the
wellbore storage hump. This is shown in Figures 3.32.



















Fig 3.32: Flow Phases Discerned on Pressure Derivative

Transient state is more uniquely identified on a dimensionless pressure-time plot. A
graph of dimensionless pressure versus log of dimensionless time has a unique slope
with a value of 1.151. This is shown in Fig 3.33.













Fig 3.33: Dimensionless Plot Showing Transient State

Derivativ
e
Log p
Log (time)
m= 1.151
Log t
D

P
D

Wellbore
Storage
Good
Wellbore
Storage
Transient
State Phase
Pressure
Unit slope line
68
The unique slope follows from the fact that

P
D
= 1.151 [log t
D
+ C] 3.40

Equations that can be used for diagnosing other flow geometries during the transient
state have been discussed under flow geometries.

Diagnosing Late Time Phase: The diagnostic plot for the late time phase depends on
the nature of the outer boundary and the type of test. We shall not go in detail, but a
summary of the different plots is as follows:

Drawdown Test (Closed Boundaries): For drawdown test in wells with closed outer
boundaries, the system attains pseudo-steady state at late time. Cartesian plots of
flowing pressure versus time during the late time phase yields a straight line. The
slope of the straight line is related to the drainage volume of the well. Figure 3.34
shows the diagnostic plot.














Fig. 3.34: Cartesian Plot Showing Pseudo-Steady State

The basis of the plot shown in Fig 3.34 follows from the equation that holds during
pseudo-steady state:

P
wf
= A + mt 3.41

Where A and m are constants. The value of m is related to the drainage volume and
this forms the basis of reservoir limit test.

The dimensionless plot is a more unique plot for detecting pseudo-steady state
because a graph of dimensionless pressure versus dimensionless time based on area
has a slope of 2 during pseudo-steady state. Figure 3.35 shows the dimensionless
plot.








Time
Wellbore
storage and
transient state
Pseudo-steady
Flowin
g
Slope = m
69


















Fig. 3.35: Dimensionless Plot Showing Pseudo-Steady State

The basis of Fig 3.35 is Eq 3.42 which is as follows:

P
D
= 2t
DA
+ C 3.42

Where C is a constant.

Drawdown Test (Open Boundaries): For a drawdown test with boundaries open to
flow, the system will attain steady state (pressure remains constant) if the influx
through the boundaries is sufficient to stop further pressure decline.

A semilog graph for systems with closed boundaries and constant pressure boundaries
is shown in Fig 3.36.



















Fig. 3.36: Semilog Plot of Drawdown Data with Different Boundary
Conditions
Slope = 2
Pseudo-steady state
P
D

Dimensionless Time based on Area, t
DA

Pseudo-steady
state
Wellbore
storage
Transient Late-time
Log
(time)
Steady
state
P
w

70
Buildup Test: For a well in a closed system that is shut-in, the pressure builds up to
average pressure during the late time phase. If there are many wells in the system,
pressure in the test well will build up and later drop due to interference effect. If there
is a large gas-cap, the system attains a constant pressure due to the gas-cap.

The derivative plots for the different late-time conditions are discussed in Chapter 4.














































71
4. ANALYSIS MODELS

Pressure time data obtained from BHP tests are normally analyzed using
mathematical models. Figure 4.1 shows both the test and analysis principles.


Reservoir
k? s?
Pressure change
(Output)
Model
k, s, etc known
Same
Rate change
Rate change
(Input)
(Input)
(Output)
Pressure change
Test Principle
Analysis Principle


Fig. 4.1: Test and Analysis Principles

The test principle involves perturbing the reservoir by applying some input (usually
rate changes) and measuring the resulting pressure. The analysis principle involves
perturbing a mathematical model using similar input applied to the reservoir and
comparing the resulting pressure with actual pressures obtained during the test. The
mathematical model is fine-tuned until the actual pressures obtained during the test
agree with pressures obtained with the mathematical model. It is then inferred that
properties of the mathematical model are similar to that of the reservoir. The
uniqueness of the result is not the subject of discussion now.

The choice of mathematical model is not arbitrary because pressure and pressure
derivatives obtained during the test contain signatures that reveal the nature of the
type of model to be used in the analysis. Therefore understanding the models will
help in relating to the reservoirs.

Each model used in test analysis may consist of 3 sub-models: well model, reservoir
model and boundary model. Options in the sub-models are shown in Table 4.1. The
term reservoir model is used here to represent the behaviour of the reservoir during
transient state phase. Any well model can be used with any reservoir model and any
boundary model to make up the analysis model. However, in some cases, the
mathematical models for the chosen sub-models may not be available or physically
feasible. Available choices show that a number of mathematical models are available.
The boundary model will not be used if the late-time was not reached.

Although it is not possible to describe all analysis models, but the following models
deserve some mention:
a. Wellbore storage and skin well in homogeneous reservoir
b. Wellbore storage and skin well on single vertical fracture
c. Wellbore storage and skin well in a double porosity reservoir.
Discussion on these follows:




72
Table 4.1: Analysis Models
Well Models Reservoir Models Boundary Models
Wellbore storage and skin Homogeneous Closed
Changing Wellbore storage Single Fracture Constant Pressure
Limited entry well Double Porosity Fault
Horizontal well, etc. Composite Leaky Fault
Well on a Fracture

4.1 WELLBORE STORAGE AND SKIN WELL IN HOMOGENEOUS
RESERVOIR.
A homogeneous reservoir is one whose properties (permeability and porosity) are
invariant in the direction of flow. This is the most common model and many high
permeability formations (eg. Niger Delta) are homogenous. Typical profiles for a
well with wellbore storage and skin in a homogenous reservoir are shown in Figures
4.2.





Figure 4.2: Log-Log and Semilog Profiles from Well with Wellbore Storage and Skin
in Homogenous Reservoir

Note that no boundary model was used because test ended during the infinite-acting
radial flow (I.A.R.F) phase.

Parameters that can be deduced with this model are as follows

C
s
= wellbore storage constant
s = skin factor
73
k = permeability


4.2 WELL ON A SINGLE VERTICAL FRACTURE
It is not unlikely that a well is located on a single vertical or horizontal fracture as
shown in
Fig 4.3. The fracture, which was caused by faulting or fracturing, becomes a fast
track for fluids getting to the wellbore.

Parameters that can be deduced with this model include the following:
x
f
= fracture half length
k = permeability
C
s
= wellbore storage constant
s = skin factor

Three forms of this model depending on the flow in the fracture exist. They are as
follows:

(a) Infinite conductivity fracture
(b) Finite conductivity fracture
(c) Uniform flux fracture





Fig. 4.3a: Horizontal Fracture


















74



Fig. 4.3b: Vertically Fractured Wells

Gringarten et al (1974) published solutions for these models. Features of the models
are discussed.

4.2.1 Infinite Conductivity Fracture
This represents the case where the fracture permeability, k
f
, is much greater than the
permeability of the matrix, k (k
f
>> k). The fracture is considered to have an infinite
permeability and therefore there is no pressure drop during flow in the fracture. The
pressure profile in this case is shown in Figure 4.4.

This reservoir goes through a linear flow, followed by a pseudo-radial flow before the
boundary effect. However, no pseudo-radial flow will appear if x
f
/x
e
= 1. This is
shown in Figure 4.5.

4.2.2 Uniform Flux Fracture
In this case, fluid enters the fracture at uniform flowrate per unit area of fracture face
so that there is a pressure drop in the fracture. Features of uniform flux fracture are
similar to the infinite conductivity case shown in Figures 4.4 and 4.5.
4.2.3 Finite conductivity fracture
In this case, fluid flows within the fracture and there is a pressure drop along the
length of the fracture. Features of this case are shown in Figure 4.6.


75


Fig. 4.4: Pressure and Derivative for Well on A Single Vertical Infinite Conductivity Fracture.




Fig. 4.5: Effect of x
f
/x
e
on Well on a Single Vertical Infinite Fracture
76




Fig 4.6: Log Log Graph of Data from Well with Finite Conductivity Fracture


4.3 DOUBLE POROSITY RESERVOIR
The double porosity reservoir is simply a fissured (naturally fractured) reservoir and is
shown in Figure 4.7.




Fig. 4.7: Fissured Reservoir


77
The main feature of this reservoir is that the pore space is divided into two distinct media:
the matrix, with high storativity and low permeability, and the fissures with high
permeability and low storativity. Flow between the fissure and matrix can occur under
pseudo steady state on transient state. However, the former is more common.

In addition to fissured reservoirs, double-porosity models can also represent layered
reservoirs in which one layer has a permeability that is much higher than the other. This
is shown in Fig 4.8. Fluid essentially reaches the wellbore through the layer with higher
permeability.












Fig. 4.8: Layered Reservoir

Layered reservoirs are also modeled with double permeability model (Bourdet, 1985), but
the features of double permeability model are similar to the double porosity model.

Warren and Root (1963), de Swaan (1976), Bourdet and Gringarten (1980) and
Gringarten (1984) published double porosity solutions. Informations that may be
deduced with the double porosity model are as follows:

k = permeability
s = skin factor
Cs = wellbore storage constant
= ratio of the storativity in the most permeable medium to that of the total
reservoir
k
2
>> k
1

k
1

78


= Inter porosity flow coefficient


Some parameters in the equations are defined as follows:

k
m
= permeability of matrix or least permeable layer
k
f
= permeability of fracture or most permeable layer
V = ratio of the total volume of one medium (matrix or fissure) to the
bulk volume
= characteristics of the geometry of the interporosity flow.

Figure 4.9 shows features of a double porosity reservoir with pseudo-steady state
interporosity flow. Features in Fig 4.9 are explained as follows:

1. At early time, only the fissures are detected. A homogeneous response corresponding
to the fissure storativity and permeability may be observed.




( )
( ) ( )
=
VC
VC + VC
t
fissure (f)
t
(f)
t
(m)

fissure matrix
=
r k
k
w
2
m
f

79
Fig. 4.9 Features of Double Porosity Reservoir

2. When the interporosity flow starts, a transition period develops. This is seen as an
inflection in the pressure response and a valley in the derivative.

3. At the end of the transition, the reservoir acts as a homogeneous medium, with the
total storativity and fissure permeability.

A few things to note while analyzing test with double porosity models are as follows:

(a) Wellbore storage may mask all indications of heterogeneity.
(b) The depth of the transition valley is a function of . When decreases (low fissure
storativity) the valley is more pronounced and the transition starts early.
(c) The time the transition ends is independent on .
(d) The time when transition occurs is a function of . When increases (higher k
m
/k
f
),
the transition occurs earlier. The time the transition ends is proportional to 1/ only.
(e) The value of can be less than or equal to one. The double porosity system
degenerate to single porosity system when = 1.
(f) The values of are usually small ( 10
-3
to 10
-10
). If is larger than 10
-3
, the level of
heterogeneity is insufficient for dual porosity effects to be important. The system
then acts as a single porosity reservoir.
(g) If inter porosity flow is transient, the valley is less evident.

4.3.1 Practical Hints on Double Porosity Models
The double-porosity model can either be used for a fissured reservoir on a multi layer
reservoir with high permeability contrast between the layers. As a result, it is not
possible, from the shape of the pressure versus time curve alone, to distinguish between
the two possibilities. The following practical hints will be of help in distinguishing the
systems. Common features are also highlighted.
(a) If well is damaged, an increase in C, after an acid job and the resulting high value of
wellbore storage constant are characteristic of fissured formation. This is because
when the well is damaged, most of the fissures intersecting the wellbore are plugged
and do not contribute to wellbore volume. On the other hand, there is no significant
change in the wellbore storage constant following an acid job in a multi layer
reservoir.

(b) Double-porosity reservoirs have skin value for non-damaged well that is lower than
zero. In reality, double-porosity reservoirs exhibit pseudo skins, as created by
hydraulic fractures. A skin of 3 is normal for non-damaged wells in formations with
double-porosity behaviour. Acidized wells may have skins as low as 7, whereas a
80
zero skin usually indicates a damaged well. A very high wellbore-storage constant
and a very negative skin should suggest a fissured reservoir, even if the well exhibits
homogeneous behaviour.

(c) The parameters and may change with time for the same well depending on the
characteristics of the reservoir fluid. The reason is that and both depend on fluid
properties, not just on rock characteristics. The parameters and will definitely
change as pressure falls below bubble point.

4.4 BOUNDARY MODELS
It is impossible to cover all possible boundary models. Appendix C taken from Middle
East Well Evaluation Review shows features of different analysis models including
different boundary models.

Although it is not absolutely correct to use models derived during drawdown for buildup
analysis, but for practical purposes, it is accepted. Pressure and pressure derivative
obtained during buildup show similar features seen in drawdown tests. Figure 4.10
published by Economides (1988) for different models shows this.

4.5 MODEL SELECTION
Two key steps in the process for estimating reservoir properties from pressure/production
data are as follows:

1. Selection of an appropriate reservoir model.
2. Estimation of parameters with the chosen model.

The selection of an appropriate model requires selecting appropriate set of material and
energy balances for the physical processes involved, as well as the fluid properties and
reservoir geometry. The problem in choosing the most appropriate reservoir model is
that several different models may apparently satisfy the available information about the
reservoir. That is, the models may be consistent with available geologic and
petrophysical information and seem to provide more or less equivalent matches of the
measured pressure/production data.





81



Fig. 4.10: Different Models used for Buildup Analysis

Watson et al (1988) suggested a method of model selection, which is summarized as
follows:
1. Select candidate models that are consistent with all available information about the
reservoir. A pool of candidates may be formed as a hierarchy of models as shown in
Figure 4.11. The number of independent parameters to be estimated from the models
is also shown.

2. Using a parameter estimation (automatic history matching) method, estimate the
independent parameters.

3. Using the calculated independent parameters, calculate the expected pressure and
production data.

82
4. Compare the calculated data with actually obtained data. The correct model is the
one that minimizes the difference between calculated and actual data in the least
square sense. Weighting factors can be included.

Although the model can be chosen at the end of Procedure 4, there is still need to find out
whether a simpler model can be used. This is because pressure and production data are
not known with certainty. Also with a simpler model, fewer numbers of unknowns are
calculated. A model that has too many parameters for the given set of data will often
result in parameter estimates that have large errors associated with them. The reason for
this is that the estimation process using models with too many parameters tends to be
poorly conditioned in that many different set of parameter values tend to give essentially
equivalent fits to the data. Consequently small measurement errors may result in large
errors in parameter estimates.

In deciding whether a simpler model can be used, Watson et al (1988) suggest using an
F-test to find out whether the estimated parameters are very different from known values
of such parameters for the simpler model. For example, if at end of Procedure 4, a
double-porosity model is chosen, calculated values of and are compared with 1 and 0
which correspond to a simpler single porosity model. A hypothesis test is done for chosen
level of significance.












Fig 4.11: Model Hierarchy

Single Porosity
With Skin
Dual Porosity
Dual Porosity with
Skin
Dual Porosity
Infinite Acting
Single Porosity, Infinite Acting
Single Porosity, Infinite Acting,
With Skin
Dual Porosity
Infinite Acting
With Skin
83
EXERCISES

1. State the uses of pressure derivative plots.
2. Assuming reservoir is infinite, sketch the pressure derivative for the following
models:
i ) Wellbore and skin model in homogeneous reservoir
ii) Fractured model with well intersecting a single vertical fracture
iii) Double porosity model

3. Using buildup pressure and pressure derivative, show the distinguishing features
between the following systems:

i) Well producing near a single fault and a well producing between two parallel faults.
ii) Well producing in a reservoir with gas-cap and a well producing in a well without
gas-cap.
iii) Well producing in a truncated channel and a well in a channel that is not truncated.
iv) Well producing from an infinite system and well producing in a rectangle with
boundaries close to flow.

4. Categorise the following into well, reservoir and boundary models:
Wellbore storage and skin, Single fracture, leaky fault, horizontal well, double
porosity, composite, constant pressure.
























84
5. SKIN FACTOR, s
T

The skin factor is useful in evaluating the well condition, but it is usually applied
wrongly. In this chapter, we shall discuss the basic things about skin and how to use the
skin factor properly.

What is Skin Factor? Factor used in BHP analysis to account for extra-ordinary
pressure changes caused by flow around the wellbore

Example of Normal (Ordinary) Pressure Change Around Wellbore
Open Hole Completion of Entire Interval (no restriction)

h
t

Fig.5.1: Unrestricted Flow at Wellbore

p p
qB
r
r
x kh
w e
e
w
=

ln
. 7 08 10
3
(steady State) 5.1

p p
qB
r
r
x kh
w w
e
w
=

_
ln .
.
0 75
7 08 10
3
(pseudo-steady state) 5.2


Example of Extra-Ordinary Pressure Change Around Wellbore
Partially Completed Well

h
t
h
p

Fig.5.2: Restricted Flow at Wellbore

Due to the stampede at the wellbore, the equation for bottom-hole flowing pressure
becomes
85
p p
qB
r
r
s
x kh
w e
e
w
c
=
+

ln
. 7 08 10
3
(steady state) 5.3

p p
qB
r
r
s
x kh
w w
e
w
c
=
+

_
ln .
.
0 75
7 08 10
3
(pseudo-steady state) 5.4

where s
c
is the skin factor due to partial completion. Generally,

p p
qB
r
r
s
x kh
w e
e
w
T
=
+

ln
. 7 08 10
3
(steady state) 5.5

p p
qB
r
r
s
x kh
w w
e
w
T
=
+

_
ln .
.
0 75
7 08 10
3
(pseudosteady state) 5.6

where s
T
is the total skin due to all factors that cause extra-ordinary pressure drop around
the wellbore. The total skin is calculated from BHP data.
Note:
1) Given the same drawdown and positive skin, well with normal pressure around
wellbore will produce more than well with skin.
2) If the average pressure in the drainage areas are the same, both wells can produce at
the same rate if the flowing pressure in well with skin is lower by an amount equal to
p
qB
kh
s
qB
x kh
s
s T T
= =

1412
7 08 10
3
.
.

. 5.7
This is shown in the Fig. 8. 3.
P
wf
if no skin
P
wf
if there is skin
p
s
r
w
r
e
useful
drawdown

Fig.5.3: Effect of Total Skin on Wellbore Pressure (van Everdingen Model)





86
5.2 Causes of Skin
a) Invasion of region around wellbore by drilling fluid and fines
b) Dispersion of clay
c) Presence of mud cake and cement
d) Stimulation
e) Sand Consolidation
f) Presence of high gas saturation around wellbore
g) Partial well completion
I) Limited Perforation
j) Slanting of wells
k) Gravel Pack
l) Sand in wellbore if flow occurs through it
m) Non-Darcy effects (caused by turbulence especially in gas wells)
n) etc.

5.3 Classification of Pseudoskins
Factors that cause skin are classified under different pseudoskins as follows:
1) Pseudoskin due to Damage, s
d
: Includes Factors a to f and other factors that result to
alteration (relative to the native permeability of the formation) of permeability around
the wellbore.
2) Pseudoskin due to Completion (also called Mechanical Skin) s
c
: Includes Factor g.
3) Pseudoskin due Perforation, s
p
: Includes Factor i.
4) Pseudoskin due to Slanted, s
sw
: Includes Factor j
5) etc.

Each pseudoskin creates its own extra-ordinary pressure around the wellbore. This is
shown in the figure for cases where the pseudoskins are positive.

p
T
r
w
r
e
useful
drawdown
p
sc
p
sd
p
sp
p
wf
p

Fig.5.4: Effect of Pseudoskins on Wellbore Pressure (van Everdingen Model)
Note:
1) The total pressure drop p
T
caused by the total skin is the sum of pressure drop
caused by individual pseudoskins
2) Stimulation can only reduce pressure loss caused by pseudoskin due to damage
87
3) Total skin may not be a good yardstick for determining stimulation candidates. It is
better to used pseudoskin due to damage, but preferably (personal experience) use
ratio defined as follows:
R
p
P p
sd
wf
=

_
5.8
where p
sd
is the pressure loss caused by skin due to damage and P p
wf
_
is the
drawdown. Table 5.1 . shows the pressure loss due to skin and calculated R ratios.

Table 5.1: Skin and Pressure Losses from Wells
WELL TEST NO
1 2 3 4 5 6 7
Total Skin 18.4 3.5 26.4 192 6.1 196 63.6
p
T
, psi 57.86 174.2 216.6 723 9.9 119 367
Damage Skin 3.34 9.1 16.6 45 -17.6 66 3.15
p
sd
, psi 30.72 135.2 204 677 -17.5 115 190
R 0.34 0.71 0.75 0.91 - 0.95 0.47

Inferences from this table are as follows:
1) Using the magnitude of total skin for ranking the priority for stimulation, the wells
will be ranked as follows: 6, 4, 7, 3, 1, 5, 2. Note that Well 5 does not require
stimulation because the skin due to damage is negative.

2) Using the damage skin or R ratio, the wells will be ranked as follows: 6, 4, 3, 2, 1,7,
5. The added advantage of using R ratio is that it takes values from 0 to 1 and from
experience we do not recommend stimulation if R ratio is less than 0.6. Tables 2 and
3 comparing pre-stimulation and post-stimulation surveys from two wells illustrate
this.

Table 5.2: Results of Pre-Stimulation and Post-Stimulation Tests (Well A)
PARAMETER PRE-
STIMULATION
POST-
STIMULATION
Rate, STB/day 568 570
Drawdown, psi 68.16 168.72
Total skin,s 9.82 4.85
p due to total skin, psi 32.45 59.7
Mechanical skin 2.2 2.2
Damage skin 7.57 1.3
p
damageskin,
psi 25.02 32.3
R -
p
damageskin
/Drawdown
0.367 0.191
Oil PI, STB/day/psi 2.37 3.08


88

Table 5.3 : Results from Pre- and Post-Stimulation Analysis (Well B)
Parameter Pre-Stimulation Post-Stimulation
Rate, STB/day 66 539
Total skin, s 46.74 8.58
Mechanical skin 0.2 0.2
Damage skin 46.57 8.4
Drawdown, psi 365.63 212.18
p
damageskin,
psi 259.53 102.2
R -
p
damageskin
/Drawdown
0.71 0.48
Oil PI 0.15 2.14

From the R ratios calculated from pre-stimulation, Well A was not a clear stimulation
candidate (R < 0.6) while Well B was a good candidate (R > 0.6). The post-stimulation
oil gains also confirm this.

5.4 Calculations of Pseudoskins

a) Pseudoskin due to Damage, s
d

r
w
r
1
r
e
k k
1

Fig.8.5: Damaged and Undamaged Zones

s
k
k
r
r
d
w
=

(
1
1
1 ln 5.9

if k > k
1
, s
d
> 0 (damage) and if k < k
1
, s
d
< 0 (stimulation)

Some implications of Equation. 5.9 that will give you a sense of magnitude is given in
Table 5.4.

89
Table 5.4: Effect of Permeability Alteration on Pseudoskin Due to Damage
k1 (% of k) Depth of Damage, r1, in (r
w
= 6 in)
8 24 60 120
90 0.03 0.15 0.25 0.33
50 0.29 1.39 2.30 3.00
10 2.59 12.47 20.7 26.96
5 5.46 26.34 43.7 56.9

Note that in high permeability formation, permeability reduction can be large if fines
block the pore throat of the porous media.

The problem with the equation for calculating pseudoskin due to damage is that depth of
invasion and permeability of the damage zone are not known. Hence, we now estimate
the damage skin from the total skin. This will be discussed.
An equation for determining the pseudoskin due to consolidation is given as:
s
r r
h
k k
k
r
r
dc
c w
p
h c
c
c
w
=

|
\

(
1 0 2 . ln 5.10
Comparing Equations 8.9 and 8.10, r
c
= r1, k
h
= k and k
c
= k. The term hp is the
perforated interval

b) Pseudoskin due to Partial Completion (Mechanical Skin), s
c
: This depends on
i) ratio of completed interval h
p
to formation thickness, h
T
(most important parameter)
ii) location of completion relative to total thickness
iii) ratio of vertical to horizontal permeability

h
p
h
T

Fig8 6: Effect of Partial Completion on Flow

Table 5.5 gives a sense of magnitude of how s
c
varies with h
p
/h
T
.


Table 5.5: Effect of Penetration Ratio on Mechanical Skin
h
p
/h
T
0.1 0.2 0.3
s
c
36.9 16.4 9.6

90
There are many sources for obtaining s
c
, but Odehs equation is preferred. Photocopy
containing his article and equation is enclosed. Saidikowskis (SPE 8204 1979 SPE
Conference) equation is simpler and gives reasonable answer for practical purposes. The
equation is
s
h
h
h
r
k
k
c
T
p
T
w
H
v
=
|
\

|
|
\

`
)
1 2 ln 5.11

c) Pseudoskin due to Slanting of Well, s
sw
: This is always negative as slanting of wells
will generally enhance productivity because more wellbore length will be exposed to
flow. This is shown in Figure . The limiting case is a horizontal well.
Length available
for fluid entry

Fig. 5.7: Length Available to Flow - Vertical and Slanted Well

The pseudoskin due to slanting of well depends on the angle of slant, , and ratio of total
thickness to wellbore radius, h
T
/r
w
.. It can be approximated using the equation published
by Cinco, Miller and Ramey. The equation is

s
h
r
sw
T
w
=
|
\

|
|
\

|
|
\

|

41 56 100
2 06 1 865 . .
log 5.12
and it is valid for 0 < < 75
o
and h
T
/r
w
> 40 (i.e. about h
T
> 12 ft).
Typical values of ssw

for different h
T
/r
w
and angles of slant are given in Table 5.6.




Table 5.6: Pseudo Skin Due to Slanting of Well
h
T
/
r
w

h
T
(ft)
r
w
= 0.4 ft
ANGLE (Degrees)
20 30 40 50 60 75
50 20 -0.18 -0.43 -0.79 -1.26 -1.85 -2.95
100 40 -0.23 -0.53 -0.95 -1.51 -2.19 -3.47
200 80 -0.27 -0.62 -1.11 -1.75 -2.53 -3.99
300 120 -0.30 -0.67 -1.21 -1.89 -2.73 -4.29
400 160 -0.32 -0.71 -1.27 -1.99 -2.88 -4.51

91
Note: If k k define
h
r
h
r
k
k
H v
T
w
T
w
H
v
,
A graph for estimating pseudoskin due to slanting of well is enclosed.

d) Pseudoskin due to Perforation, sp: This depends on many parameters, but more
importantly on perforation penetration and perforation density. Figure illustrates how
this contributes to skin.


Fig.5.8: Flow Convergence to Perforation
The skin due to perforation can be positive (for shallow penetration) or negative (for deep
penetration). The enclosed chart by Harris will be used to have a sense of magnitude.
The best value is about -3 and the worst value is about 10.

In calculating the pseudoskin due to perforation, most models (including Harris) neglect
the pressure drop due to flow in the perforation tunnels. This pressure drop could be
substantial and contribute to skin if tunnel is filled with sand or gravel. This is because
there could be turbulent flow within the tunnels. Deductions from studies that consider
pressure losses within the perforation tunnels are as follows:

i) There is optimum perforation length for minimum skin. This is shown in Fig 5.7. The
productivity ratio used in the figure is defined as

PR
r
r
r
r
s
w
w
p
=
+
ln
ln
1
1
5.13
Based on this, the skin caused by perforation tunnel being filled with gravel can be
between 13 to 30.

ii) Minimum skin due to perforation is smallest for highest shot density.
iii) Minimum skin depends on gravel type, viscosity, rate openhole/casing diameter, etc.

8.5 Relationship between Total Skin, s
T
and Pseudoskins

1) Full Penetration of Entire Interval: s
T
= s
c
+ s
d
+ s
p
+ s
sw
+ .. 5.14
2) Partial Penetration
a) Shallow Damage Skin (Damage restricted to perforated interval)
92

[ ]
s s
h
h
s s s s
T c
T
p
d p sw i
= + + + + 5.15

b) Deep Damage Skin (Damage covers entire total skin)

[ ]
s s s
h
h
s s s
T c d
T
p
p sw i
= + + + + 5.16

If total skin is known (from BHP test result) and other pseudoskins calculated, the
pseudoskin due to damage can be deduced from these equations.


8.6 Pressure Change Due to Skin
The pressure change caused by skin can be calculated as follows:

1) Total Skin, s
T
:

p
qB
kh
s
s
T
T
=
1412 .
5.17

2) Pseudoskin Due to Partial Penetration (Mechanical Skin), s
c
:

p
qB
kh
s
sc
T
c
=
1412 .
5.18


3) Pseudoskin Due to Damage,s
d
:
a) Shallow Damage Skin:
p
qB
kh
s
sd
p
d
=
1412 .
5.19
b) Deep Damage Skin:

p
qB
kh
s
sd
T
d
=
1412 .
5.20

4) Other Pseudoskins, s
i
:

p
qB
kh
s
si
p
i
=
1412 .
5.21

Note that 7.08 x 10
-3
is the inverse of 141.2.


93
8.7 Relationship Between Skin and Well Inflow Quality Indicator (WIQI)

WIQI
PI
PI
actual
ideal
= 5.22
where:
PI
x kh
B
r
r
s
actual
e
w
T
=
+

7 08 10
0 75
3
.
ln .
5.23
and
PI
x kh
B
r
r
s
ideal
e
w
c
=
+

7 08 10
0 75
3
.
ln .
5.24

The mechanical skin is included in defining the ideal productivity index (PI) because it is
permanent and can only changed if the completion is changed. From Equations 5.22
to 5.24, we have
WIQI
r
r
s
r
r
s
p p
p
e
w
c
e
e
T
wf sd
wf
=
+
+

ln .
ln .
0 75
0 75

5.25

Exercises
1) A well was completed with a 7 in diameter casing diameter as shown in the Fig. 5.9
50 ft
150 ft

Fig. 5.9: Partially Completed Well
The casing has a perforation density of 4 shots/ft and 120
o
phasing. It is estimated that
the formation penetration is 15 inches and total skin determined from BHP analysis is 25.
Considering only the pseudoskin due to partial completion and skin due to perforation,

Calculate the following:
a) Skin due to damage assuming damage is shallow.
b) Skin due to damage assuming damage is deep.
c) Pressure change due to total skin.
d) Pressure change due to damage skin due to damage assuming damage is shallow
e) Pressure change due to damage skin due to damage assuming damage is deep
f) Is there any difference between answers from d and e.
94
(Use k = 1000 md, q = 1000 STB/D, B = 1.10 rb/STB, = 0.6cp)
2) Repeat Exercise 1 for case where well is slanted at angle 60
o
.

3. Five pages of a report on a test considered unreliable are enclosed. The calculated
parameters were not entered in the provided slots, but just written down. Even
though the results are considered unreliable, but someone still wants to know the skin
due to damage, pressure loss due to it and WIQI. Your job is to calculate required
parameters. All data required can be found in the enclosed report.

Skin Correction: In some analysis, it is difficult to match the correct flowing pressure
before shut-in. This is usually due to rate or time problem. The case where the real
flowing pressure was not matched is shown in the following simulation plot.
P
real data
simulated profile
(Analysis results)
p = analysis p
wf
- actual p
wf
Time
flowrate


In a situation like this, the analysis skin was calculated with the following equation:

s
P p t
m
k
c r
hr wfa
t w
=
=

|
\

| +

`
)
1151
0
323
1
2
.
( )
log .

where:
p
wfa
(t=0) = flowing pressure just before shut-in (used for analysis)
m = semilog straight line slope=
162 6 .
/
qB
kh
psi cycle

.

As the analysis flowing pressure is greater than the actual flowing the actual skin will be
s + s where

s
p t p t
m
wfa wfr
=
= = |
\

| 1151
0 0
.
( ) ( )

and
p
wfr(t=0)
= real flowing pressure just before shut-in


??????????????????

95
6. ANAYSIS OF BOTTOM-HOLE PRESSURE TESTS

There are three methods of analyzing bottom-hole pressure tests: conventional, type-
curve and regression (automated type-curve) methods. To the layman the methods can
be likened to three ways of getting an overall work dress for a worker. The conventional
method is like going to a tailor who takes measurements at three points of the body and
with just three measurements, the overall is made. Note that with the exception of the
three measurements, other dimensions used in making the dress are extrapolations or
interpolations. Also, if the tailor uses the dimension of the waist to be that of the
shoulder, the dress will not fit. This underscores the importance of where the
measurements were taken.

The type-curve method can be likened to just walking into a shop and choosing from
already made dresses. The problem here is that the exact size may not be available and
one has to manage with the next lower of higher size. In high permeability-high skin
Niger Delta formations, we have observed that available type-curves do not match the
pressure data.

The regression method is simply an automated type-curve method where the exact size
that takes into account all dimensions is made. The problem here is that if you are not
standing properly while the measurements are taken, the wrong overall will invariably be
made. This implies that wrong pressure data will create problems during analysis with
regression method if such data are not removed. The advantage of the regression method
is that the correct-sized dress will be made and therefore there is no problem with the
high permeability-high skin Niger Delta formations.

In this section we shall discuss methods of analyzing drawdown, buildup and interference
tests. Emphasis will be on the conventional and type-curve methods. However, we shall
show that in an ideal situation, the three analysis methods give comparable results.


6.1 Conventional Method
Bottom-hole pressure tests can be analyzed using the conventional method. The
conventional analysis is based on locating the straight lines that are characteristic of the
different flow phases. The desired parameters are then calculated from the slopes of the
straight lines. This implies that if you put the wrong straight lines, you will get the wrong
answers.

The procedure for analyzing drawdown and buildup test using conventional method is as
follows:

6.1.1 Drawdown Test
The transient state equation during a drawdown test is given as follows:

96
(
(

+
|
|

\
|
+ = s
r c
k
t
kh
qB
P P
w t
i wf
87 . 0 23 . 3 log log
6 . 162
2

6.1

Equation 6.1 forms the basis of calculating permeability and skin from drawdown test
using data not influenced by wellbore storage effects. The procedure for the analysis is
oultined.

Procedure for Analyzing Drawdown Test
1. Locate pressure influenced by wellbore storage effect and calculate wellbore
storage constant. How?
(a) Make a graph of (P
i
-p
wf
) versus t on a log-log paper
(b) Locate data with strong wellbore storage effect on the unit slope line.
(c) Calculate the wellbore storage constant, Cs, with p= (Pi -Pwf) and t taken
from a point on the unit slope line. Use the following equation.
Cs =
qBt bbl
psi 24
(rb/psi) 6.2

(d) Locate the data not strongly influenced by wellbore storage effect. Use the
gentle slope rule or the 10t* to 50t* rule.

2. Locate data obtained during the transient state phase and calculate permeability
and skin. How?
a) Plot p
wf
versus log t. Use a semilog paper with time graphed on the log scale axis.

b) With the knowledge of the time when wellbore storage effect has died down
completely, put the correct straight line that represents the good transient state
behaviour.

c) Determine the slope of the semilog straight line, m, and calculate the permeability
using the equation

k (mD) =
mh
qBu 6 . 162
6.2

Note that the slope, m, is negative and therefore permeability, k, is positive.

d) Calculate the total skin using the following equation:

s = 1.1513
( )
)
`

232 . 3 log
2
1
w t
i hr
r uc
k
m
P P

6.3
97

where:

P
1hr
is the pressure taken from the straight line portion when flowing time is equal to one.

Exercise 6.1:. The log-log and semilog plot of the pressure-time obtained during a
drawdown test are shown in Fig 6.1 and Fig 6.2 (leave space for these figures ???? ) Data
from the test are shown in Table 6.1 ??? (ask me). Draw the necessary straight lines and
calculate Cs, k and s. Other information required for the analysis are as follows:

r
w
= 0.5 ft h = 50ft B
o
= 1.125
c
t
= 20 x 10
-6
psi
-1
= 0.6 cp = 0.25.
q = 1000 STB/D


Table 6.1: Drawdown Data
Time, hr Flowing Pressure, psi Time, hr Flowing Pressure, psi
.00000000E+00 3183.763 .10176000E+01 3123.157
.10000000E-03 3183.245 .12576000E+01 3122.951
.80000000E-03 3179.758 .14976000E+01 3122.781
.20000000E-02 3174.302 .17376000E+01 3122.638
.48000000E-02 3163.727 .19776000E+01 3122.513
.96000000E-02 3150.896 .22176000E+01 3122.402
.12000000E-01 3146.317 .24576000E+01 3122.303
.18240000E-01 3138.186 .28176000E+01 3122.172
.21600000E-01 3135.356 .31776000E+01 3122.056
.27840000E-01 3131.839 .35376000E+01 3121.953
.32400000E-01 3130.196 .38976000E+01 3121.860
.39600000E-01 3128.530 .42576000E+01 3121.776
.44400000E-01 3127.814 .46176000E+01 3121.698
.55680000E-01 3126.796 .49776000E+01 3121.626
.60000000E-01 3126.553 .53376000E+01 3121.559
.88800010E-01 3125.731 .56976000E+01 3121.496
.11040000E+00 3125.446 .60576000E+01 3121.438
.13440000E+00 3125.224 .64176000E+01 3121.383
.17760000E+00 3124.928 .67776000E+01 3121.330
.24960000E+00 3124.572 .71376000E+01 3121.281
.37440000E+00 3124.154 .74976000E+01 3121.234
.53760000E+00 3123.788 .79536000E+01 3121.177
.77760000E+00 3123.421


Although you have gone through the exercise manually, there are computer programmes
that do the graphing and calculations in a twinkle of an eye. I used the programme in
98
analyzing this test and the results obtained are as follows: Cs = 0.93 x 10-2, k = 991.2,
and s = 24.72.

In a situation where the drawdown test is run long enough, the volume drained by the test
well can be determined. The equation that forms the basis of this analysis is as follows:


(

+ + + = s
C r
A
kh
qB
t
hAc
qB
P P
A w t
i wf
2
2458 . 2
ln ln
2 . 141 234 . 0
2

6.4

The implications of this equation are as follows:
(a) A graph of P
wf
versus t gives a straight with a slope,

t
hAc
qB
m

234 . 0
= (psi/hr) 6.5

(b) From the slope, the drainage volume, hA, is calculated.

Exercise 6.2: Table (6.2 ??? ask) shows pressure data from a drawdown test in a well in
the center of a square which was tested long enough to reach pseudo-steady state. Using
the conventional method and data in Table 6.3, analyze the test and calculate the
following:
a) Formation permeability b) Skin factor c) Wellbore storage constant d) Drainage area

table 6.2 here ??

Table 6.3: Parameters used for Analysis
Parameter Value
Wellbore radius, r
w
, ft 0.5
Total compressibility, c
t
, psi
-1
20 x 10
-6

Porosity, 0.25
Oil formation volume factor, B
o
, rb/STB 1.125
Oil viscosity, , cp 0.6
Production rate, q, STB/D 1000
Formation thickness, h, ft 50

Solution: Plots from the analysis and results are shown in Fig 6.3. ????? ask




6.1.2 Buildup Test
Two forms of transient state equation can be used in the analysis of buildup test. The
first equation was developed by Horner (1951) by assuming that the test well produced
99
for a while (late-time state was not reached) before the well is shut in. The second form
of the equation was developed by Miller, Dyes and Hutchinson (1950) by assuming that
the test well produced for a long time (late-time state was reached) before the well was
shut in.

Based on the assumptions used in developing the equations, their applications were
limited. For example, the Horner equation was used for new wells while Miller, Dyes
and Hutchinson (MDH) equation was used for old wells. However, Ramey and Cobb
(1971) have shown that the Horner equation is superior even when it is used for old
wells. On the other hand Agarwal (1980) has shown that MDH equation can be used for
new wells by using an equivalent shut-in time defined as follows:

|
|

\
|

+

=
p
eq
t
t
t
t
1
6.6

Parameters in Eq. 6.6 are defined as follows:

t
eq
= equivalent shut-in time to be used for analysis
t = actual shut-in time
t
p
= production time before shut in

Discussions on the Horner and MDH equations and procedure for using them follow:

Horner Equation
The Horner equations for infinite and developed systems are given as follows:

|
|

\
|

+
=
t
t t
kh
qB
P P
p
i ws
log
6 . 162
for infinite reservoir 6.7

and
|
|

\
|

+
=
t
t t
kh
qB
P P
p
ws
log
6 . 162
*

for finite or developed reservoir 6.8

Note that as shut-in time, t, increases, the Horner time {(tp + t)/ t} decreases. At
infinite shut-in time, Horner time is unity (1).

MDH Equation
The MDH equation is given as follows:


100
(
(

|
|

\
|
+
|
|

\
|
+ + =

A
r C
r c
k
t
kh
qB
P P
w A
w t
ws
2
2
log
000264 . 0
log log
6 . 162

6.9

Unknown parameters in Eqs. 6.7 to 6.9 are defined as follows:

P = Average pressure
P* = false pressure
P
i
= initial pressure
C
A
= shape factor or geometric factor
A = drainage area

The shape factor depends on the location of the well with respect to the boundary.
Typical shape factors are in Table 6.4.

Table 6.4: Reservoir Geometries and Shape Factors
Geometry and Well location Shape Factor



Well in center of circle

31.62




Well in center of square

31.6




2
Well in the center of 2 to 1 rectangle

21.8369


1

2
Well in a quadrant of 2 to 1 rectangle

2.0769

Shape factors for other reservoir geometries and well location are given by Earlougher
(1977).

Irrespective of the equation used for analysis, the procedure for calculating wellbore
storage constant and skin is the same. We shall therefore present a general procedure for
analyzing buildup test and highlight the variations resulting from the use of any of the
equations (MDH or Horner)

1
101
Procedure for Analyzing Buildup Test

Calculation of Wellbore Storage Constant
1. Locate pressure influenced by wellbore storage effect. How?
(a) Make a graph of p =
( )
[ ]
p p t
ws wf p
versus t on a log-log paper.
(b) Locate data with strong wellbore storage effect on the unit slope line.
(c) Calculate the wellore storage constant, Cs, using p and t from a point on the
unit slope line and the following equation:

Cs =
qB t
p

24
rb/psi 6.10

(d) Locate the data not strongly influenced by wellbore storage effect. Use the
gentle slope rule or the 10t* (1 cycle) to 50t* 1.5 cycle) rule.

Calculation of Permeabilty
MDH Plot
2. Plot p
ws
versus log t. Use a semilog paper with time graphed on the log scale axis.
3. With the knowledge of the time when wellbore storage effect has died down
completely, put the correct straight line that represents the good transient state
behaviour.
4. Determine the slope of the semilog straight line, m, and calculate the permeability
using the equation.
k (mD) =
162 6 . qBu
mh
6.11

At the end of this step, move to Step 8 to calculate skin factor.

Horner Plot
5. Plot p
ws
versus
(

+
t
t t
p
log . Use a semilog paper with Horner time graphed on the
log scale axis. The graph can be made with Horner time increasing from left to right
to left. The later is usually preferred as it gives a graph with shut-in time increasing
from left to right.
6. With the knowledge of the time when wellbore storage effect has died down
completely, put the correct straight line that represents the good transient state
behaviour.
7. Determine the slope of the semilog straight line, m, and calculate the permeability
using the Eq. 6.11.
8. Extrapolate the straight line to Horner time equal to 1 (infinte shut-in time) and read
off P* or Pi. They are needed for estimating the average pressure in a well that was
not shut in long enough to reach average pressure.

102
Calculation of Skin
9. The correct equation is given as

(
(

+
|
|

\
| +
+
|
|

\
|

= 23 . 3
1
log log
) (
1513 . 1
2
1
p
p
w t
p wf hr
t
t
r c
k
m
t P P
s

6.12

However, if production time before shut-in, t
p
>> 1, Eq. 6.12 degenerates to


(
(

+
|
|

\
|

= 23 . 3 log
) (
1513 . 1
2
1
w t
p wf hr
r c
k
m
t P P
s

6.13


In Eqs. 6.12 and 6.13, P
1hr
is the pressure taken from the straight-line portion of the
semilog plot at shut-in time, t = 1 hour. Note that shut-in tome of 1 hr corresponds to
Horner time of (t
p
+ 1).


Note on Production Time before Shut-in, t
p

For a well that produced at constant rate (unusual) before shut in, t
p
, is the actual
production time before shut in. This is illustrated in Fig 6.4.




Rate



Time
Fig.: 6.4: Actual Production Time Before Shut in Equal to t
p


If a well did not produce at constant rate (usually the case) before shut in, t
p
can be
calculated from cumulative production provided that a stable flow rate was attained
during the flowtest prior to shutting in. The equation for calculating t
p
is given as
follows:

s
p
p
q
N
t

= 6.14

where N
p
= Cumulative production since last shut in
q
s
= stabilized rate prior to shutting in

This case is illustrated in Fig 6.5.
q
t = t
p

103










Fig.: 6.5: Actual Production Time Before Shut inis not Equal to t
p


Experience has shown that for wells that produced for a long time (reached pseudo-
steady state or steady state) before shut in, t
p
, can be replaced by a time that is at least the
time to reach pseudo-steady state, t
pss
. The Horner time will simply be defined as as [(t
pss

+ t)/ t]. For high permeability formation obtained in the Niger Delta, we normally use
1000 hours.

Exercise 6.3: Figures 6.6 to 6.8 log-log and semilog plots of pressure-time data obtained
during a buildup test. Draw the necessary straight lines and calculate Cs, k and s. Other
information required for the analysis are as follows:

r
w
= 0.5ft h = 50ft B
o
= 1.125
C
t
= 20 x 10
-6
psi
-1
= 0.6 cp and = 0.25.
Q = 1000 STB/D t
p
= 1000 hr.

Answers: k = 1000 mD, s = 25 and Cs = 0.009 rb/psi


Calculation of Average Pressure
For wells that were not shut-in long enough to average pressure, the average pressure can
be obtained if the transient state was attained. An example of such a well is shown in
Figure 6.9.














q
s

Rate
Time
t t
p

Good
Transient
Log (t +t/t) 1
P
ws

P*
104






The average pressure is obtained based on the fact that P* is related to the average
pressure by the equation

) (
303 . 2
*
PDA DMBH
t P
m
P P =

6.15
where

P = Average pressure
P* = False pressure
P
DMBH
= Dimensionless pressure defined by Mathews, Brons and Hazebroek (1954)
m = Slope of the semilog straight line
tpDA = Dimensionless (based on area) production time before shut-in

A c
kt
t
p

000264 . 0
= Oilfield units.

The P
DMBH
values depend on geometry of reservoir, well location and dimensionless
production time based on area. The values for some cases can be obtained from Fig 6.10.

Leave a page for this figure ????

Fig 6.10: Mathews, Brons and Hazebroek Dimensionless Pressure, P
DMBH



Note that Horner plot and Mathews, Brons and Hazebroek dimensionless pressure, P
DMBH

Were used to obtain the average pressure. The Dietz (1965) method which involves the
used of MDH plot in obtaining average pressure will not be discussed.

The values of P
DMBH
can also be calculated at certain periods using the following
equations:

1. Early production time before shut-in (t
p
< t
pss
)

P
DMBH
= 4t
DA 6.16


2. Long Production time before shut-in (tp t pass )

P
DMBH
= In C
A
t
pDA 6.17


Fig. 6.9: Horner Plot
105
Confirm the correctness of these equations with Fig 6.10.

Some facts about average pressure are as follows:

(a) For infinite system, p = p* = p
i


(b) If cumulative production is not sufficient to reduce the average pressure
significantly, p* is still a good estimate of average pressure.

(c) Average pressure decreases with increase in production time before shut in. This
is shown in Fig 6.11

leave space here???? ask

Fig 6.11: Relationship between Average Pressure and False Pressure


(d) The average pressure calculated from a test well is the average pressure in the
drainage area of the test well. The reservoir average pressure is then calculated as
follows:

reservoir P =
p q
i
i=1
N
i
t
q

6.18

where
p
i
= average pressure from Well i.
q
i
= production rate from Well i.
q
t
= total production from all the wells in the reservoir.

In situation where all the wells were not tested, the average pressure in the untested well
can be estimate using the p - q relationship from the tested well. You have to interpolate
or extrapolate.


6.1.3 Interference Test
The equation that forms the basis of analyzing interference is as follows:

(
(

|
|

\
|
+ = 23 . 3 log log
6 . 162
) , (
2
r c
k
t
kh
qB
P t r P
t
i

6.19

106
Equation 6.19 predicts the transient state pressure at an observation point that is r
distance from the active well. This equation does not include wellbore storage and skin
in either the active or the observation well.

The procedure for using Eq 6.19 for interference test analysis is as follows:
(a) Graph P(r,t) (pressure recorded in the observation well) versus time, t, on a
semilog graph.
(b) The transient state data will fall on a straight line. Permeability can be calculated
with straight line slope, m, with the equation:

k (mD) =
162 6 . qBu
mh
6.20

(c) Porosity or storativity can then be calculated from the equation
23 . 3 log
1
+

=
|
|

\
|
m
P P
c
k
hr i
t

6.21

In Eq 6.21, P
1hr
is the pressure at time, t = 1 hr, taken from the straight-line portion of the
curve.


6.2 Type-Curve Method
Gringarten (1987) defined a type curve as a graphic representation of the theoretical
solution of the fluid flow equation used in representing the test well and the reservoir
being tested. For a constant-pressure test, the theoretical solution is presented in the form
the change in production rate with time. This is usually used in decline curve type
curves. For a constant-rate test, the theoretical solution is presented as change in pressure
at the bottom of the well as a function of time. This form is normally used in bottom-hole
pressure test analysis. Other types of response are also used, such as the time derivative
of the bottom-hole pressure.

Type curves are derived from solutions to the flow equations under specific initial and
boundary conditions. For the sake of generality, type curves are usually presented in
dimensionless terms, such as a dimensionless pressure vs. a dimensionless time. A given
interpretation model may yield a single type curve or one or more families of type curves,
depending on the complexity of the model.

107
Type-curve analysis involves finding a type curve that matches the actual response of
the well and the reservoir during the test. The reservoir and well parameters, such as
permeability and skin, can then be calculated from the dimensionless parameters defining
that type curve.

The match can be found graphically, by physically superposing a graph of the actual test
data with a similar graph of the type curve(s) and searching for the type curve that
provides the best fit. Alternatively, an automatic fitting technique involving a linear or
nonlinear regression can be used.
Figure 6.12 taken from Gringarten (1987) gives an example of a graphic type-curve
match. The graph of the data is positioned over the graph of the type curves, with the
axes kept parallel, so that the test data match one of the type curves. Reservoir parameters
are calculated from the value of the dimensionless parameter defining the type curve
being matched and from the x and y axis shifts.

Fig 6.12 here ask ????

Fig 6.12: Graphical Type-Curve Matching Process

There are many kinds of type-curves, but we shall only discuss the Theis (1935) type-
curve used for analyzing interference test and Bourdet et al (1983) type-curve used for
analyzing buildup and drawdown tests. The type curves are shown in Fig. 6.13 and Fig
6.14. The Theis type-curve is also known as line source solution.

Leave a page for this fig 6.13

Fig 6.13: Theis Type-Curve

Leave a page for this fig 6.14

Fig 6.14: Bourdet Type-Curve
108

6.2.1 Basis of Type-Curve Analysis
The basis of type-curve analysis will be illustrated using the interference type case and
application of type-curve will be extended to buildup tests.

Interference Test: The basis of type-curve analysis is the relationship between
dimensionless parameters and the non-dimensionless parameters used in interference test
type-curve. The relationships are shown for dimensionless pressure and time as follows:

P
D
=
kh[ p]
141.29B

6.22


t
r r
D
D
2 2
=
0.000264kt
c
t

6.23

Taking log of Eqs 6.22 and 6.23 gives

log P
D
= log p + log
qB
kh
2 . 141
6.24

log
t
r r
D
D
2 2
= log t + log
0.000264k
c
t

6.25

Note that log(kh/141.2qB) and log(0.000264k/c
t
r
2
) are constants and therefore
regarded as shifts.

The implications of Eqs 6.24 and 6.25 are as follows:
(1) On a log-log paper, P
D
is directly related to p and
t
r
D
D
2
is also directly related to t.
The parameters of interest in the analysis constitute the shifts in the relationships.

109
(2) A graph of p versus t on a log-log paper will look exactly like that of P
D
versus
t
r
D
D
2
on a similar log paper.
(3) p - t plot can then be overlain on the P
r
D
D
-
t
D
2
type curve. This is known as type-
curve matching process. Normally, the field data are plotted on transparent paper so
that the match can easily be made. Details on the process of type-curve matching
will be demonstrated in the class
On matching the curves, any point on the type-curve ( P
D
and t
D
/r
D
2
) and the
corresponding point on the field data (p and t ) is known as a match point. With values
of the match point, the desired parameters can be calculated with the following equations.

K =
1412 . qB
h
p
m

.
P
D

(
(
6.26

C
r
t
D
m
=
0.000264k
r

t
t
2
D
/
2

(
6.27

Noe that Equations 6.26 and 6.27 are simply rearranged forms of dimensionless
parameters defined by Eqs. 6.22 and 6.23.

Buildup Test (Pressure Match): Using Bourdets type-curve, a match could be obtained
using the pressure type-curve or the pressure derivative type-curve. Both type-curves are
in Fig. 6.14. For pressure match, dimensionless parameters defined by Bourdet et al
(1983) for analysis of buildup test are as follows:

P
D
=
Kh
141.2qB
p

6.28


t
C
D
D
= 0.000295
Kh

t
C

6.29
110

C
hr
D
w
=
0.8936 C
C
t

2
6.30

The procedure for using the type-cure is as follows:
1. Graph p = P
ws
P
wf
(t
p
) versus t on a transparent paper with same log cycle
dimensions as the type-curve.
2. Match the real graphed data with a type-curve
3. Obtain the match point values (p, t, P
D
t
D
/C
D
) and the C
De
2s
of the representative
type-curve
4. With the match point values, calculate
a k from Eq. 6.28
b C from Eq. 6.29
c C
D
from Eq. 6.30
d s using C
D
and C
De
2s
of the representative type-curve.

Buildup Test (Derivative Match): For this case, we give you the opportunity to outline
the procedure.


6.3 Comparison of Analysis Methods
Under ideal conditions, all analysis method will yield close results. The ideal conditions
includes good data acquisition, test run long enough to reach required phases and also
accurate identification of the phases. Table 6.5 show results obtained by analyzing
drawdown test using conventional and type-curve methods. Data for this test are in
Table 6.1. In this case, both analysis methods gave close results.

Table: 6.5 DrawDown Analysis Results
Parameter Conventional Method Type-Curve Method
Permeability, k, md 991.2 1000
Skin, s 24.72 25
Wellbore Storage, Cs, rb/psi 0.93 x 10
-2
0.9 x 10
-2


Some comments on the analysis are as follows:
1. Both analysis methods gave close results, but the type-curve method is more accurate.
We know about this because the analyzed pressure data were generated with k =
1000md, s = 25 and Cs = 0.9 x 10
-2
rb/psi. This implies that the type-curve method
gave exact answers.
111
2. The test was run long enough as the transient state is clearly shown on the derivative
plot.

Table 6.6 shows results obtained by analyzing drawdown test using conventional and
type-curve methods. The derivative for this analysis is shown in Fig 6.15. Both analysis
methods did not give close results because the transient state is not apparent. We cannot
say which resulta are correct because data are actual field data.

Table 6.6: Buildup Analysis Results

Parameter Conventional Method Type-Curve Method
Permeability, k, md 65 293
Skin, s 12 75
The test well whose data are shown in Fig 6.15 was producing 129 (STB/day) and the
gas-oil ratio (GOR) was 882 SCF/STB. With such a low production and high GOR, what
do you expect?. Wellbore storage effect will last for unusually long period and the
transient state phase will be masked.

Leave space for fig 6.15 ask

Fig 6.15: Derivative from Test Data


Note that with the type-curve method, we do not look for straight lines corresponding to
the different phase. Rather, we do a curve-fit to match the pressure responses with the
ideal responses for the entire phases. When an acceptable curve-fit is obtained, the
desired reservoir parameters can be calculated.


Problem Set 6
1 Pressure-time data from a buildup test are given in Table 6.7. The well and reservoir
data are as follows:
r
w
= 0.5 ft, h = 50 ft, q = 1000 STB/day, c
t
= 20 x 10
-6
psi
-1
, = 0.25,
o
= 0.6 cp, B
o

= 1.125 rb/STB, t
p
= 1000 hr

A. Using the Horner plot determine the following:
a) effective permeability to oil
b) total skin factor
c) wellbore storage constant
d) average pressure assuming that well is in the centre of a square with a drainage area
of 80 acres

112
Table 6.7: Buildup Data
t, hr p
ws
, psi t, hr p, psi
0 3183.76 0.0888 3241.80
0.0001 3184.28 0.1776 3242.60
0.0008 3187.77 0.3774 3243.37
0.0020 3193.22 0.5376 3243.74
0.0048 3203.80 0.7776 3244.10
0.0120 3221.21 1.0176 3244.37
0.0278 3235.69 1.2576 3244.57
0.0557 3240.73 1.9776 3245.01

2) Confirm results obtained in Question 1(a), (b) and (c) using the MDH method.

3) The pressure drop, p, and the derivative of the pressure drop, p, from a buildup test
are given in Table 6.8. The well and reservoir data are as follows:
r
w
= 0.5 ft, h = 50 ft, q = 1000 STB/day, c
t
= 20 x 10
-6
psi
-1
, = 0.25,
o
= 0.6 cp, B
o

= 1.125 rb/STB, t
p
= 1000 hr.

A. Using the Horner plot determine the following:
a) effective permeability to oil
b) total skin factor
c) wellbore storage constant

B. Repeat the analysis using type-curve analysis and pressure derivative data

Table 6.8: Buildup Data
t, hr p, psi p t, hr p, psi p
0 0 - 0.0888 58.032 1.668
0.0001 0.5180 - 0.1776 58.835 1.200
0.0008 4.005 4.646 0.3774 59.609 1.195
0.0020 9.4610 9.087 0.5376 59.975 1.169
0.0048 20.036 15.461 0.7776 60.34 0.987
0.0120 37.446 18.078 1.0176 60.605 0.979
0.0278 51.923 11.773 1.2576 60.811 -
0.0557 56.967 3.984


4). An interference test was run in shallow water sand. The active well produced 466
STB/D and pressure was measured in the observation well, which was 99 ft from the
active well. Estimated rock and fluid properties are as follows:
w
= 1.0 cp, B
w
= 1.0
bbl/STD, h = 9 ft, r
w
= 3 in, c
t
= 27.4 x 10
-6
psi
-1
Match points obtained from type -
curve matching with the pressure response are as follows:
t = 128 minutes,
t
r
D
D
2
= 10, p = 5.1 psi, and P
D
= 1.0
Using supplied information calculate permeability and porosity.
113


5) An interference test is run between two oil wells. One is produced at 477 STB/D and
the pressure response is measured at a nearby (263ft) well with a high precision guage.
Pressure and reservoir data are given below. Using the semilog analysis method and type
curve method, find the effective permeability to oil, md, and the effective in place
porosity, fraction of bulk volume.

h = 15ft r
w
= 0.275ft

o
= 0.89cp c
t
= 13.5 x 10
-6
psi
-1

B
o
= 1.10 RB/STB

Table 6.9: Pressure Data at Observation Well
t, hrs p, psi
0 0
25 57.5
37.5 66.3
50 73.8
62.5 78.8
75 82.5
87.5 86.6
125 95.0
























114
7. SENSITIVITY ON FACTORS AFFECTING BHP
ANALYSIS
In this chapter, we shall discuss the effect of certain factors on analysis of bottom-hole
pressure (BHP) tests. Factors considered include leaks, rate variation, sampling
frequency, interference effect, wrong datum correction, and gauge sensitivity and
accuracy. The effects of these factors were simulated in ideal BHP data and results from
analysis of such data compared with the ideal case. We present the ideal data case and
results from analysis of the data. We shall use the buildup test, which is the commom
test, in illustrating these factors affect results obtained from BHP analysis. The ideal
BHP test data are used as the standard because fluid and reservoir parameters used in
generating such tests are known.

7.1. Analysis of Ideal BHP Data
The ideal BHP data were obtained by simulating the buildup test using parameters in
Table 7.1. The simulation is also called test design. Generated pressure-time data are in
Table 7.2.

Table 7.1: Parameters for Test Design
Parameter Design Value
Wellbore radius, r
w
, ft 0.5
Total compressibility, c
t
, psi
-1
20 x 10
-6

Formation thickness, h, ft 50
Porosity, 0.25
Oil formation volume factor, B
o
, rb/STB 1.125
Oil viscosity, , cp 0.6
Production rate, q, STB/D 1000
Production time before Shut-in, t
p
, h
r
1000
Permeability, k, md 1000
Total skin, s, 25
Wellbore storage constant, C
s
0.009

The ideal buildup data were analyzed using the conventional and type-curve methods.
The analysis procedure is as follows:

A Conventional Analysis Method
(i) Graph on log-log paper, p versus t and calculate Cs (see Figure 7.1)
(ii) Graph on semilog paper, Pws versus
t t
t
p
+

(Horner plot). Calculate k and s (see

Fig. 7.2)
(iii) Use values calculated in Steps (i) to (ii) to generate pressure profiles. Plot the
profiles with solid lines and compare with ideal data. This comparison is shown in
Figs. 7.3. What can you conclude from the comparison?
115
Table 7.2: Designed Buildup Data
Shut-in Time, hr Pressure, psi Shut-in Time, hr Pressure, psi
************************************************************************
0.00000 3183.763 1.01760 3244.368
0.00010 3184.281 1.25760 3244.574
0.00080 3187.768 1.49760 3244.743
0.00200 3193.224 1.73760 3244.887
0.00480 3203.799 1.97760 3245.011
0.00960 3216.630 2.21760 3245.122
0.01200 3221.209 2.45760 3245.220
0.01820 3229.340 2.81760 3245.352
0.02160 3232.170 3.17760 3245.467
0.02780 3235.686 3.53760 3245.569
0.03240 3237.330 3.89760 3245.662
0.03960 3238.996 4.25760 3245.746
0.04440 3239.712 4.61760 3245.824
0.05570 3240.730 4.97760 3245.895
0.06000 3240.973 5.33760 3245.962
0.08880 3241.795 5.69760 3246.024
0.11040 3242.080 6.05760 3246.082
0.13440 3242.302 6.41760 3246.137
0.17760 3242.598 6.77760 3246.189
0.24960 3242.953 7.13760 3246.238
0.37440 3243.372 7.49760 3246.285
0.53760 3243.738 7.95360 3246.341
0.77760 3244.104


B. Type curve Analysis Method
We used the automatic type curve analysis method. Results obtained were used in
generating simulated profiles. The simulated profiles are compared with the ideal data
and shown in Fig. 7.4. Table 7.3 is a summary of calculated results and the expected
results.

Table 7.3 Buildup Analysis Results
Parameter Calculated Results Corrected Results
Conventional Type Curve
K, md 1000 1000.29 1000
S 25.01 25.01 25
Cs rb/psi 1.03 x 10
-2
0.8996 x 10
-2
0.9 x 10
-2

P* 3251 3251 3251

From Table 7.3, we conclude that for the ideal case, both analysis methods will give the
expected results. We now investigate the effects of some factors on the results.


116
7.2 Effect of Gauge Accuracy and Datum Correction.
The effect of gauge accuracy was investigated by introducing consistent errors in the
ideal BHP data. Two cases were considered. In the first case we assumed that the gauge
readings were consistently lower by 15 psi. This case may also represent a case where
the pressures were measured at depth that is 40 ft shallower than the datum.

Datum
Measurement depth
Gauge
Top of perforation
Z/2
h

Figure 7.1: Gauge position and Datum.

As expected, the pressure profiles obtained from the analysis of this first case are similar
to what we obtained with the idea data. The only difference is the shift in pressure level.
This implies that P* and average pressure obtained in this analysis will be lower by 15
psi. Table 7.4 shows a summary of the results.

Table 7.4: Effect Gauge Accuracy and Datum Correction-Case 1
Parameter Calculated Results Corrected Results
Conventional Type Curve
K, md 1000 1005 1000
S 25.01 25.01 25
Cs rb/psi 0.962 x 10
-2
0.901 x 10
-2
0.9 x 10
-2

P* 3236 3236 3251

In the second part of this investigation, we introduced 5psi, 7psi and 10psi errors over
certain ranges. In real situation, this may be likened to wrong calibration over certain
pressure ranges. Figures 7.5 and 7.6 show the plots of pressure data for this case. Table
4.5 shows the results obtained from the analysis of this case.



Table 7.5: Effect of Gauge Accuracy -Case 2
Parameter Calculated Results Corrected Results
Conventional Type curve
K, md 1002 995.8 1000
S 24.02 24.02 25
Cs rb/psi 0.94 x 10
-2
0.93 x 10
-2
0.9 x 10
-2

P* 3229 3229
117

The staggering of the pressure profile in this case is embarrassing, but we are pleased
that the calculated results are close to the correct results. This does not mean that good
gauge calibration is not important. Who knows the magnitude of the error you will
introduce by not calibrating your gauge properly.

7.3 Effect of Noise
By noise, we mean erratic pressure perturbations caused by gauge malfunction or erratic
flow in the wellbore. The effect of noise was simulated by randomly introducing errors
in the pressure data. Figures 7.7 to 7.8 show the pressure data after introducing the errors.
Table 4.6 gives a summary of the results for this case.

Table 7.6: Effect Gauge of Noise
Parameter Calculated Results Corrected Results
Conventional Type curve
K, md 1003 995 1000
S 35.42 35.42 25
Cs rb/psi 0.32 x 10
-2
0.32 x 10
-2
0.9 x 10
-2


Some comments on the results are as follows:
1 Noise may cause overestimation or under-estimation of parameters.
2 Due to noise, the derivative plot may exhibit features, which may be wrongly
misinterpreted as boundary effect

7.4 Effect of Gauge Sensitivity.
We simulated the effect of gauge sensitivity by setting the pressure readings in some
regions to be the same. This implies that the gauge could not discern the difference in
pressure in that region. Figure 7.9 shows data for this case while Table 7.7 shows the
summary of the results from the analysis.






Table 7.7: Effect Gauge of Sensitivity
Parameter Calculated Results Corrected Results
Conventional Type curve
K, md 1003 987 1000
S 24.9 24.9 25
Cs rb/psi 1.42 x 10
-2
1.4 x 10
-2
0.9 x 10
-2



7.5 Effect of Rate Variation
We simulated the bottom-hole pressure using design rate shown in Table 7.8. We also
analyzed the test using the rate shown in Table 7.8. Both the design and analysis rates
118
are also shown in Fig. 7.10. Table 7.8 and Figure 7.10 represent a case where a well
flowed at 1000 STB/D for 999 hours and then just an hour prior to shut in the rate was
reduced to 500 STB/D. During the analysis, a rate of 500 STB/D was used and this
implies that the complete rate history was not used. The rate variation simulated here
could be caused by reducing the choke size to drop the tool. A tool whose size could
obstruct flow reasonably could also cause it.


Table 7.8: Rate Schedule for Design and Analysis
Time, hrs Design Rate, STB/D Time, hr Analysis Rate, STB/D
0 1000 0 500
999 500 1000 0
1000 0


Design Rate
Analysis Rate
1000
1000
1000 999
500
500
time
time
0
0
q
q


Figure 7.10: Design Rate and Analysis Rate.

The pressure profiles in this case are shown in Fig. 7.11 and Fig 7.12. The figures and
analysis look reasonably, but the results shown in Table 7.9 are not correct. Table 7.9
also shows results obtained by analyzing the test using the correct rate history. Figures
7.13 and 7.14 show the match obtained using the wrong rate history and correct rate
history respectively.



Table 7.9: Effect of Rate Variation


Parameter
Calculated Results
(Wrong Rate)

Calculated Results
(Correct Rate)



Correct Results
Conventional Type Curve Conventional Type
Curve

k, md 743.9 736.9 958.6 1000 1000
s 17.24 17.24 23.68 25 25
Cs, rb/psi 1.01 x10
-2
1.01 x10
-2
0.92 x 10
-2
0.91 x 10
-2
0.9 x 10
-2


Figure 7.11 to 7.14 here ???

From these results, we make the following inferences:
119

1 Analyzing a bottom-hole pressure test with incorrect rate schedule will give wrong
answers. Even the Horners approximation for handling multirate test as a single
rate test did not give satisfactory results.

2 Avoid things that will cause unnecessary rate changes during test. If rate changes
occur, record it properly and let the one who will analyze the test be aware of it.

3 Rate variation could lead to the beginning of another wellbore storage phase and
this could be misinterpreted as a boundary. This is shown in Fig 7.15 for actual
data.

Fig 7.15 ????? here

7.6 Effect of Leaks
Leaks may result in the gauge measuring pressure that may be lower than the actual
pressures especially when the leaking system is at higher pressure. For a buildup test,
leaks may result to continuous flow and hence the sandface rate not being zero (q
sf
0).
In this section, we shall consider leaks caused by the sandface rate not equal to zero. The
following may cause such leaks:
(a) Leaking lubricator
(b) Leaking packer, which may result in communication between two sand intervals.

We simulated the bottom-hole pressure test with a leak of 1% of the original rate during
the buildup stage and analyzed the test with the assumption that there was no leak (q
sf
=
0) during the buildup period. The results of the analysis are shown in Table 7.10.




Table 7.10: Effect of Leak
Parameter Calculated Results Correct Results
Conventional Type Curve
k, md 944 964 1000
s 22.94 23.84 25
Cs, rb/psi 1.13 x 10
-2
0.96 x 10
-2
0.9 x 10
-2


Concluding, leaks of any form are not welcome. Both permeability and skin may be
underestimated or overestimated.

7.7 Effect of Interference /Leak
Producing wells that are around the test well can cause interference. These wells cause
some pressure drop at the test well because the wells are draining the same reservoir and
each well behaves as if it does not know that the other well is present. We simulated this
effect by deducting the pressures due to interference from the ideal pressure data. The
resulting data could also represent a case with leaks.
120

Figures 7.16 shows the log-log plot for this case and results obtained are in Table 7.11.

Table 7.11: Effect of Interference
Parameter Calculated Results Corrected Results
Conventional Type curve
K, md 1043 1316.7 1000
S 26.18 34.6 25
Cs rb/psi 1.016 x 10
-2
0.917 x 10
-2
0.9 x 10
-2


Deductions from the analysis are as follows:
(1) From the derivative plot, the effect of interference could be misinterpreted as constant
pressure boundary effect.
(2) Both analysis methods gave wrong answers, but the type-curve results are worse.
This is because during the type curve analysis, every data point is given equal
weighting while the conventional analysis considers only the segment of interest.


Concluding, we should not allow leaks and if there are serious interference effects, the
effects must be accounted for before analyzing your data. For the real tests, all these
factors may be present during one test and results obtained from analyzing the test will be
wrong. We therefore advice that you avoid as much as possible any of the factors that
may adversely affect BHP tests.























121
8. FIELD CASES
In this chapter, we shall present some of the tests we have analyzed and problems
encountered. We shall first present analysis of a test that was properly run to show that
tests can actually be run properly.

8.1 Good Test
A good test requires the following:
(a) The gauges must be calibrated properly so that gauge readings will be reliable.
(b) Accessories must be tagged for effective depth control.
(c) Test programme must be correct, understood and followed.

Figures 8.1 and 8.2 show the readings of the upper and lower gauges for a good test.
Table 8.1 is a worksheet that shows the temperature and pressure readings at different
times during the test. From the worksheet we infer that the readings are consistent as the
expected pressure differences between the gauges agree with the calculated pressure
differences based on the fluid gradient between the gauges. The same deduction could be
made from the temperature differences.

???

Fig. 8.1: Good Measurements with Upper and Lower Gauges During Buildup Period

??????

Fig. 8.2: Good Measurements with Upper and Lower Gauges During Entire Test Period

Table 8.1: Worksheet For Gauge Quality Check
General Information
Field: Niger Well: 1 Reservoir: E 1.0
Top Gauge:WX1 Depth 7713 ft/CHH Bottom Gauge: WX2 Depth 7717 ft/CHH
Test Date: 01/01/95 Gross (STB/D): 340 BSW (%): 0 GOR (SCF/STB): 1236

Quality Check Analysis
Differential Pressure Analysis
Event Time
(hr)
P
lower

(psi)
P
upper

(psi)
p (psi) Fluid Pressure
Gradient
(psi/ft)
Calculated
p (psi)
Difference in
p (psi)
Flowing
Period
14.5 3082.69 3081.54 1.15 oil 0.30 1.2 -0.12
Early
Buildup
14.91 3240.73 3239.14 1.59 oil 0.30 1.2 -0.92
Mid
Buildup
20.46 3246.29 3244.15 1.59 oil 0.30 1.2 -0.86
Late
Buildup
26.43 3245.73 3244.15 1.58 oil 0.30 1.2 -0.88
Static
Stop
27.82 3005.19 3005.24 0.95 oil 0.30 1.2 -1.17
Average Difference in calculated p (psi): - 0.722

122
Differential Temperature Analysis
Event Time
(hr)
T
lower

(C)
T
upper

(C)
T (C) Fluid Temp.
Gradient
(C/m)
Calculated
T (C)
Difference
in T (C)
Flowing
Period
14.5 159.28 158.45 0.83 oil 0.01 0.04 0.136
Early
Buildup
14.91 159.85 159.05 0.80 oil 0.01 0.04 0.045
Mid
Buildup
20.46 159.77 158.74 0.83 oil 0.01 0.04 0.066
Late
Buildup
26.43 159.82 158.99 0.83 oil 0.01 0.04 0.034
Static
Stop
27.82 156.15 155.39 0.76 oil 0.01 0.04 0.103
Average Difference in calculated T ( C) : 0.083

Comments: The pressure gauges gave consistent readings. The average pressure offsets of -0.722 psi can
be accepted as the pressure rise during the buildup was about 1565 psi. The pressure difference plot shows
no evidence of phase segregation. This is clearly shown on the pressure difference plot in Fig 8.2.

Figures 8.3 to 8.5 show the plots used for the analysis. The derivative plot shows the
wellbore storage phase, infinite-acting radial flow phase and late-time data influenced by
the constant pressure boundary. Note that in this analysis, the pressure profiles simulated
with the calculated results (solid lines) match the real data very well. In this test there
were no leaks and fluid segregation effects.


???????
Fig. 8.3: Log-Log Plot of Data from Good Test

???????
Fig. 8.4: Semi-Log Plot of Data from Good Test


???????

Fig. 8.5: Simulation Plot of Data from Good Test
8.2: Effect of Gauge Movement
From experience, moving gauges during flowtest prior to shutting in or during buildup
renders tests useless. This is also true even if the test analyst knows that the gauges were
moved. This implies that once you start the flowtest, do not move the gauges even if you
have just realized that gauges are not at specified depth. Just state in your report the
exact location of the gauges. On the other hand, if you must move the gauges (say to the
perforations to minimize fluid interface movement) you must start the flowtest all over
again on reaching the new depth.

123
Figure 8.6 shows an example in which gauges were moved prior to shutting in the well.
The gauge was moved further down after the flowtest. In this case, the flow period at the
new depth was not sufficient.

????
Fig 8.6: Gauge Movement Prior to Shutting in

The problem with analyzing this test is that two different bottom hole flowing pressure
before shut-in could be used in the analysis. The first bottom-hole flowing pressure is the
one obtained prior to the gauge movement. Analysis plots obtained by using this bottom-
hole pressure are shown in Figures 8.7 and 8.8. Obtained results are shown in Fig 8.8. A
second bottom-hole pressure before shut in that could be used in the analysis is the one
prior to shutting in. That is, after the gauge was moved. The analysis plot for this case is
shown in Fig 8.9. The second option looks more reasonable, but may be difficult to
recognize. It may be regarded as an artefact as pressure has started building up. We
recognize it because the BHP contractor was instructed to move the gauge after flowtest
a wrong instruction.

An important issue in moving gauges prior to shutting in is that the correct shut-in
pressure may not be used for the analysis. This is an important parameter in the
calculation of skin and also in type-curve matching. Also, gauge movements can yield a
dome that may be misinterpreted as wellbore storage phase as shown in Figure 8.8.
This is because if you are running in the gauges at constant speed, pressure will vary
linearly with time. During the wellbore storage phase, pressure also varies linearly with
time. Therefore, do not complicate the analysis with unnecessary movement of the
gauges.

8.3 Effect of Gauges off Depth
We are not just interested in measured pressure, but we also want to know the depth
where the pressure was measured. If depth is off by 50ft from datum in a reservoir that
produces dry oil of gravity 0.35psi/ft, the error in datum pressure will be 17.5psi. This
may seem small, but many reservoirs may not have that much depletion in one year. This
implies that a 1-year error has been introduced in reserve calculations!

Depth control is important and that is why we insist on tagging accessories, marking
wire, etc., but many operators still do not understand the importance. Poor depth control
is a source of error in calculating wellbore fluid gradient. The equation for calculating
the fluid gradient is as follows:

1 2
1 2
x x
P P
Gradient Fluid

= 8.1

where
P
1
= Pressure measurement at Poin1
x
1
= Vertical depth at Point 1 (along hole depth corrected for deviation)
P
2
= Pressure measurement at Poin2
124
x
2
= Vertical depth at Point 2 (along hole depth corrected for deviation)
Figure 8.10 shows an example where we detected poor depth control. Both the upper and
lower gauge readings are shown. After correcting for deviation, the calculated static
gradients in Column 4 gave unrealistic results. For example, in some cases, a lighter fluid
is found below a denser fluid. This is not physically possible and was likely caused by
depth error.

?????? Leave a page
Fig. 8.10: Incorrect Static Gradient Due to Poor Depth Control

The magnitude of the depth error can be estimated by introducing shifts in the depths
until a satisfactory fluid gradient trend is obtained. Column 5 shows the shifts and the
final fluid gradients are in Column 8. The shifts are considered as the suspected depth
errors. This should be brought to the notice of the BHP contractor so that he will be
careful with his depth measurements.


Also, leaks could cause faulty static gradients. This is shown in Fig 8.11. In this case,
there is no remedy as both the numerator and denominator in Eq 8.1 may be affected.

?????? Leave a page
Fig. 8.11: Incorrect Static Gradient Due to Leak


We have observed some contradiction between the claimed gauge position and resulting
effect on pressure. For example, with the gauge at mid perforation we do not expect any
effect on pressure due to liquid interface movement.



8.4 Effect of Reporting Wrong Rates
Flowrate data is as important as the pressure data. Without the flowrate data,
permeability, the most important parameter, cannot be calculated from BHP test. The
relationship between flowrate and permeability is given in Oilfield units by the following
equation:

K
qB
mh
=
162 6 .
8.2
where:
K = permeability, md
q = flowrate, STB/day
m = slope obtained from semilog plot of BHP data, psi/cycle
B = formation volume factor rb/STB
125
= Viscosity, cp
h = drainage thickness, ft

A consequence of using wrong rate is that permeabilities calculated from different tests
on the same well will be inconsistent. This is not expected unless there is a process
altering the permeability in the drainage area. Figure 8.12, shows the expected feature.
The slope of the semilog plot is constant and therefore permeability is constant, but
average pressure in the drainage area changes time.

????? Here

Fig 8.12: Consistency in Permeability and Analysis Rates

Flowrate is also used in computer programs used in BHP analysis for calculating
wellbore storage constant as shown for a buildup by the following equation:

C
qB t
p
s =

24
8.3

where:
C
s
= wellbore storage constant.
p = P
ws
-P
wf
(tp) from unit slope of log-log plot of wellbore storage phase.
t = shut-in time corresponding to p.

The wellbore storage constant can also be estimated with equations that do not involve
rate but depends on storage mechanism (fluid expansion, falling liquid, etc.), fluid
compressibility in the wellbore, wellbore and connected volume. The wellbore storage
constant calculated with these equations should agree within reasonable limits with
wellbore storage constant calculated with Equation 8.3. In situations where they do not
agree, it is likely that the wrong rate was used in the analysis. For example, in one of our
analysis, the wellbore storage constant calculated using Equation 8.2 was 0.00176
STB/psi while estimated wellbore storage constant depending on whether tubing
communicates with casing or not lies between 0.00556 to 0.043 STB/psi. This implies
that the rate supplied for the analysis is off by a factor of at least three. In another case,
we calculated a wellbore storage constant of 0.0071 STB/psi using Equation 8.3 while the
estimated wellbore storage constant is 0.00724 STB/psi. In this case, we believe that the
rate used for the analysis was correct. This could be used as a quality check for rate
measurement if measured pressures are reliable.

To overcome the problem of reporting the wrong rate, we recommend the following:

1. The BHP contractor should be interested in obtaining correct rate data as much as he
is interested in obtaining good pressure data.
2. The flowstation supervisors should cooperate with the BHP contractor and both
should ensure that the test well is correctly hooked on to the test separator so that
correct flowrate data will be obtained.
126
3. We are aware that the flowstation supervisor has scheduled dates for testing each
well, but he should utilize the opportunity created by the BHP to perform a complete
test on the test well. That is, obtain rate, BSW, GLR, etc.

8.5 Effect of Ineffective shut-in / Well not flowing before Shut-in
If there are no speed limits, the state of many roads can be determined by how fast
vehicles move on the roads. In the same manner, important reservoir parameters can be
determined from BHP test from the rate at which pressure drops (as in drawdown) or
rises (as in buildup) with time. This implies that during a good buildup test, pressure
must rise. This will occur if the well was flowing and then effectively shut in.

The correct pressure rise during a buildup test can be achieved by flowing the well until
stabilization and then completely shutting the well to flow. The implications of the last
sentence are as follows:
1. During a buildup, the well must be flowed until stabilization. The stabilization
condition is removed if well is surging. However, in that case, well must be flowed
for at least 6 hours.
2. If well cannot be completely shut-in due to faulty valve or wax or any other reason,
the buildup part of the test should be aborted.

Table 8.2 shows the production data recorded by a BHP contractor before the well was
shut for buildup. It is apparent from the flowrate data that the well was not flowing prior
to shutting in. The little production he observed when he opened the well could have
come entirely from the wellbore and not the reservoir. Do not forget that fluid in the
wellbore is compressed and this could cause the wellbore fluid to be produced on opening
the well initially.

Table 8.2: Production Data Prior to Shut-in
OIL
TIME Meter Reading(bbl) Obser. Vol. (bbl)
12:45:00 205.70 67
13:00:00 210.00 4.30
13:15:00 248.40 38.40
14:00:00 336.70 2.90
14:15:00 355.00 18.30
14:30:00 372.50 17.50
15:30:00 391.70 0.00
15:45:00 391.80 0.10
16:30:00 391.90 0.00
16:45:00 391.90 0.00
17:00:00 391.90 0.00
17:15:00 391.90 0.00
18:15:00 391.90 0.00
18:30:00 391.90 0.00
The well was shut-in at 18:30:00
127

Figure 8.13 is a pressure plot of a well that could not be shut in or the well was not
flowing prior to shutting in. In this case, there was no pressure rise and the test is useless
if the objective of the test is to calculate permeability and skin.

???
Fig 8.13: Ineffective Shutting in of Well

It is easy to know whether a well flowed or could not be shut in during buildup test. In
this case, the flowing gradients (misnomer for a dead well) and the static gradients are
same. This is shown in Fig 8.14. Figure 8.15 is the static and flowing gradients for a
well that was flowing and properly shut in during the buildup. Note that for this case, the
static pressure at a given depth is higher than the flowing gradient at the same depth.

????? Leave a page
Fig 8.14: Static and Flowing Gradients for a Well that is not Flowing


Fig 8.15: Static and Flowing Gradients for a Live Well

In addition to our interest in ensuring that wells are effectively shut during buildup, we
are equally interested in your shutting the well as fast as possible. Pressure profiles in
Fig. 8.16 show a well that was slowly shut in. Slow shutting in of wells makes tests
difficult or impossible to analyze.

??????
Fig 8.16: Slow Shutting in of Well

8.6 Effect of Leak
Using the road analogy, a leak could be likened to a detour on the road. Such a detour
will make it difficult to ascertain the condition of the road from the speed of the vehicles.
Leaks make it impossible to analyze BHP tests because leaks cannot be quantified. Also,
when there are leaks, calculated fluid gradients are wrong as shown in Fig 8.11.

During BHP tests, leaks could occur through
(a) leaking lubricator.
(b) communication between long string and short string or between a string and casing.
This could occur through gas lift ports or non-sealing packers.
(c) not completely shutting the well, etc.

At early shut-in time (wellbore storage phase), the effect of leak may not be detected
from pressure data because during that period, the rate at which pressure rises is usually
greater than the rate at which it leaks. Therefore, a net pressure rise is observed. But,
during the infinite-acting radial flow phase in formation with high permeability as in the
Niger Delta, the rate at which the pressure rises will in most cases be lower than the rate
at which pressure leaks. A net pressure drop is recorded and such a test cannot be
128
analyzed. Figure 8.17 shows pressure plots from a test with leak. Note that pressure
difference between lower and upper gauge may be constant if there is leak.

?????
Fig 8.17: Pressure Drop Caused by Leak

Figures 8.18 and 8.19 show the semilog and log-log for a case where there was an intial
leak through the lubricator. The leak was stopped by pumping the lubricator before
pressure started to rise again. It is always better to avoid such problems by using tested
lubricators.

?????? page 107

Fig 8.18: Semilog Plot of Data with Initial Leak from Lubricator


??????
Fig 8.19: Log-log Plot of Data with Initial Leak from Lubricator

In conclusion, make sure your lubricator is not leaking and that wells are completely shut
in. Leaks due to communication between strings and casing can be detected by recording
the surface pressures of all the strings and casing using a surface chart pressure recorder.
??? Figures 8.20 and 8.21 show the pressure chart using in recordind surface pressure for
two cases. Figure 8.20 shows that there was communication between the test string (Well
15S) and casing during the test. Pressure from this string merged with pressure in the
casing after a while. Fig 8.21 is an example of a case with no communication between
test string and casing. The pressures from both strings and casing followed their own
trend. We recommend the use of these charts.

??? Leave a page
Fig. 8.20: Surface Pressures of Short String, Long String and Casing (Case with Communication)


??? Leave a page
Fig. 8.21: Surface Pressures of Short String, Long String and Casing (Case with no Communication)
???????????

8.7 Effect of Gauge Oscillations/Sensitivity Problems
We have observed strange oscillations in pressure data measured with electronic gauges.
In most cases, the oscillations occur in both gauges as shown in Fig. 8.22. Definitely,
such oscillations are neither due to well nor reservoir responses. We believe that the
oscillations may be due to the procedure used while running the test. Example of
procedures that could produce such oscillations include jarring the gauges and allowing
the gauges hit objects in the well. The oscillations could also be caused by electronic
problems that we do not understand.

??????
129
Fig 8.22: Gauge Oscillations


Another factor that could cause milder oscillations is poor gauge sensitivity or
temperature effects. This is shown in Fig. 8.23. The gauge gives readings that are about
the correct value. An example of the effect of such oscillation on pressure data graphed
on Horner plot is shown on Fig. 8.24. These oscillations cause discontinuities on the
pressure derivative as shown in Fig. 8.25. The implication of this is that the different
flow phases may not be clearly discerned.


?????
Fig 8.23: Gauge Oscillations Due to Sensitivity Problem


????
Fig 8.24: Effect of Gauge Oscillations on Horner Plot


????
Fig 8.25: Effect of Gauge Oscillations on Pressure and Derivative on Log-log Plot



Depending on when and how the oscillations occur, oscillations can make test analysis
impossible. Also, in cases where BHP tests with oscillations in data are analyzed, the
oscillations cause a large band of uncertainty for the calculated parameters.

8.8 Effect of Gas Phase Segregation
Liquid phase is usually produced with dissolved gas. Hence when a well is shut-in, the
dissolved gas comes out of solution and moves to the top of the wellbore creating an
unusual pressure rise (or hump) in the wellbore. This pressure hump is illustrated in
Fig. 8.26 and Fig 8.27.
????
Fig 8.26: Gas Phase Segregation Hump on a Semilog Plot

????
Fig 8.27: Gas Phase Segregation Hump on a Simulation Plot

Gas phase segregation is more common in wells with high gas-oil ratio(GOR) and can be
identified using the following:
(a) Pressure difference (i.e. lower gauge pressure minus upper gauge pressure) between
the gauges increases with time because as gas comes out of solution, liquid between
the gauges become denser.

130
(b) The pressure changes due to phase segregation cause a dip in the pressure
derivative. The dip in the pressure derivative may be erroneously interpreted as
double porosity feature. Figure 8.28 illustrates this.











Fig: 8.28: Pressure Derivatives Showing Dips Due to Phase Segregation and Double
Porosity

Figure 8.29 shows an example of a real case of the effect of gas phase segregation on
pressure derivative.

??????
Fig 8.29: Effect of Gas Phase Segregation on Pressure Derivative

(c) On semilog plot of the pressure data, phase segregation effects cause a hump. This
is shown on Fig 8.26.

Generally, gas phase segregation is unlikely to prevent correct interpretation of test
because the effect dies down quickly.


8.9 Effect of Liquid Interface Movement
Quite like gas, oil and water will separate in a well that is shut in. Depending on the
position of the gauges and amount of water produced, recorded pressure data can be
affected by oil/water interface movement. The oil/water interface can fall below the
gauges or rise above the gauges. A schematic from Lingen (1995) showing the effect on
pressure for case of rising and falling contacts are in Figs. 8.30 and 8.31. Figure 8.30
shows a case where the oil/water contact was initially below the gauges and later rose
above the gauges. In this case, pressure builds up, drops and then rises again. This is
likely to occur in wells producing with BS&W of about 45% or more if the gauges are
not close to the perforation. Figure 8.31 shows a case where the oil/water contact was
initially above the gauges and letter dropped below the gauges. The pressure rises
normally and then more rapidly before the normal pressure rise continues. This is more
likely to occur in wells with gas with the gas pushing down the oil/water contact.

?????
Log t
Log p
Dip due to Phase
Segregation
Dip due to Double
Porosity
Log t
131
Fig 8.30: Effect of Rising Oil/Water Contact

?????
Fig 8.31: Effect of Falling Oil/Water Contact


These effects can be explained by the fact that the density of the fluid in the well and
between the gauges affect the rate of pressure change during buildup. Unlike the case of
gas phase segregation, pressure distortion caused by liquid interface movement can last
for a much longer time and therefore can make the test uninterpretable.

Figures 8.32 and 8.8.33 show examples of a real case of rising liquid interface. Figure
8.32 is on a Cartesian plot while Fig. 8.33 is the same phenomenum on a semilog log
plot. Liquid interface movement is an unwanted phenomenon. To eliminate or minimize
it, the gauges must be at the perforations. This is especially important if the BS&W is
high (BS&W > 30% - not a magic number).

????
Fig 8.32: Liquid Interface Movement on a Cartesian Plot

????
Fig 8.33: Liquid Interface Movement on a Semilog Plot


8.10 Effect of Gaslift
Distortions on pressure data caused by gaslift effects are unpredictable. Intermittent
opening and closing of gaslift valves during the buildup cause the distortions. The
distortions could be worse if the gas supply to the well was not turned off.

Distortions caused by gaslift malfunction make flow phase indetification difficult. Hence
analysis will yield wrong results. Figures 8.34 and 8.35 show two interpretations of a test
with gaslift distortion. The log-log and semilog plots used in the interpretation are shown.
Figures 8.34a and 8.34b show one form of interpretation and Fig 8.35a and 8.35b show
another form of interpretation of the same test. Results from both interpretations are
also shown. Note that the gaslift effects makes it difficult to simulate the correct pressure
profiles and result obtained from analyzing such test may depend on the assumed profile.

????????
Fig 8.34 a and b
Fig 8.35 a and b


8.11 Effect of Short Buildup or Flow Period
We have discussed the different flow phases that a well that is shut-in goes through. We
also discussed the information that may be derived from each phase. For example,
132
accurate values of skin and permeability are derived from the infinite-acting radial flow
phase.

In this section, we present results showing the effect of test duration. In one case, the test
was run long enough to reach the infinite acting radial flow (IARF) phase while in
another case, the some test data were removed to simulate a case where the test was not
run enough to reach (IARF) phase. The two cases were analyzed and the results
compared.

In the first case where the (IARF) phase was reached, analysis was easy. Both
conventional and automated type-curve methods could be used. Figure 8.36a and b show
the log-log and semilog plots used in the analysis. Figures 8.37a and b show plots used in
the analysis in the case where the (IARF) phase was not reached and the the skin
constrained. That is, the skin was not varied during the autmatic type-curve matching.
Figures 8.38a and b show plots used in the analysis in the case where the (IARF) phase
was not reached and the skin not constrained. That is, the skin was varied during the
autmatic type-curve matching. Results obtained from analyzing the different cases are
shown in Table 8.3.

Table 8.3: Effect of Buildup Period on Calculated Results.
Parameter Calculated Results
Long Shut-in Short Shut-in
Case 1
(contrained skin)
Case 2
(uncontrained skin)
Permeability, md 1840 708 2090
Skin 32.4 9.09 37.8
C
s
, STB/psi 0.0248 0.0237 0.025


Figure 8.36 to 8.38 here ??????

Deductions from the analysis are as follows:
1. Tests run long enough to reach a clearly defined infinite-acting radial flow phase will
yield correct values of permeability and skin.
2. For a test that is not run long enough (e.g. infinite-acting radial flow phase not
reached) the calculated permeability and skin will be wrong irrespective of the
analysis method.
3. In these cases where the wellbore storage phase was observed, the calculated
wellbore storage constants are close.
4. The semi-log plots show that if the infinite-acting radial flow phase is not reached test
cannot be analyzed using the conventional method.

In addition to our being interested in the duration of the buildup period, we are also
interested in the duration of the flow period prior to shutting in of well. A flow period of
about 2 to 6 hours (good for Niger Delta high permeability formations) with the gauges at
the survey depth where the well will be shut is necessary for the following reasons:
133
1. To ensure that there is a flow period that can be used for matching simulated profile
with real data.
2. To ensure that the correct flowing pressure at shut-in is used. This affects calculated
skin and productivity index.

8.12 Class Discussion
It is not possible to present and discuss every type of problem encountered in BHP test
and analysis. In some cases, we are not able to figure out what caused the problem. In
such cases, we give probable cause of the problem.

In this section, we shall present some cases for discussion. Every one will participate in
trying to find out what happened during the tests or the characteristics shown by the test
data.

Case 1: Figures 8.39a and b are the log-log and semilog plots of a test we regard as a
mystery test. The mystery if the sudden rise in pressure of about 300 psi that occurred at
about 3.00 am while the well was shut for overnight. The gauge depth was less than 150
ft from the maximum well depth. Now explain what caused the problem.

???
Fig 8.39: Log-log and Semilog Plots of the Mystery Test


Case 2: Figures 8.40a and b are the log-log and semilog plots of a test with a problem.
Explain what the problem is.

?????
Fig 8.40: Log-log and Semilog Plots of a Test


Case 3: Figures 8.41a and b are the log-log and semilog plots of a test with a problem.
Explain what the problem is.

?????
Fig 8.41: Log-log and Semilog Plots of a Test

Case 4: Figures 8.42a and b are the log-log and semilog plots of a test with a problem.
Explain what the problem is. Varying kh ?, Composite system? Etc.

?????
Fig 8.42: Log-log and Semilog Plots of a Test


Case 5 : Figures 8.43 shows the log-log plot of a test with ran with an Amerada gauge.
Explain the trend.

134
?????
Fig 8. 43: Log-log of a Test ran with Amerada Gauge


Case 6 : Figures 8.44 shows the Cartesian plot of a test with a final shut-in pressure that
is lower than the flowing pressure. Suggest the cause of the problem.

?????
Fig 8.44: Cartesian Plot of a Test with a Problem


Case 7 : Figures 8.45a and b show the semilog plots from tests in two adjacent wells
draining the same reservoir. The permeability calculated from the semilog plots differ by
a factor of about 10. Explain the big unrealistic difference.


Fig 8.45: Semilog Plots of Tests in Adjacent Wells Draining the same Reservoir

One deduction from this case is the need for system approach to welltest analysis. Tests
from wells in the same reservoir must be compared to determine unrealistic results. This
comparison will help show the local (well) and global (reservoir) effects.

























135
9. THEORETICAL CONCEPTS

We have given a lot of equations without showing how they were obtained. Readers who
are mathematically inclined will like to know the origin of the equations. Therefore, for
completeness, we decided to include a chapter on theoretical concepts of bottom-hole
pressure test analysis. In this chapter, we shall derive some of the equations in Darcy
units, outline method of prediction pressure for test design and throw a little light into the
concept of superposition which forms the basis for generating multi-rate flow equations.

9.1 DERIVATION OF FLUID FLOW EQUATIONS
To understand and correctly analyze well test data require:

(a) Understanding of physical processes involved. For example, the fluid flow
processes the effect of reservoir geometries and heterogeneties.

(b) Selection or production of proper mathematical equations or models for analysis
of results.

To satisfy conditions (a) and (b) will require understanding the mathematics of fluid flow
in porous medium. Discussion on the derivation of some of the fluid flow equations
follow.

The basic concepts involved in the derivation of fluid flow equations are:
(a) Conservation of mass equation
(b) Transport rate equation (e.g. Darcys law)
(c) Equation of state

The use of these basic concepts in derivation of liquid and gas flow equations in porous
medium is discussed. Flow in cylindrical coordinates is considered with flow in angular
and z-directions neglected. The equations are given as follows:

9.1.1 Liquid Flow Equations
A. Conservation of Mass Equation:

Mass rate in - Mass rate out = Mass rate storage 9.1

B. Transport Rate Equation

q = -
k
A
dp
dr
9.2

By conserving mass in an elemental control volume shown in Fig. 9.1 and substituting
the transport rate equation, the following equations can be obtained:

136
=
r
p

rhk 2

r (

r r
r
p

rhk 2
+ (

+ 2 r r h
t
( )

9.3

Expanding the first term on the right hand side of Eq. 9.3 about r using Taylors series
expansion gives:


1
r

r
{
rkp

p
r
} =
t
( )

9.4













Figure 9.1: Schematics of the Reservoir and Control Volume.

Equations 9.1 to 9.4 apply to both liquid and gas. For each fluid, the density or pressure
term in Eq. 9.4 may be replaced by the correct expression in terms of pressure or density.
For slight but constant compressibility liquid,

=
sc
e
c[P - p
sc
]
9.5

Substituting for pressure in Eq. 9.4, the final equation in terms of density is:

2
+
1
r

p
=
(c + c
k

p
t
r
p
r r
2
)
9.6

In terms of pressure, Eq. 9.4 is:

2
r
+
1
r

p
+ c(
p
r
=
(c + c

p
t
p
r r k
2
2
)
)
9.7

If pressure gradient is small everywhere in the system, Eq. 9.7 becomes:

2
+
1
r

p
=
C
k

p
t
t
p
r r
2

9.8

r
r+
r r+
137
Equation 9.8 is the diffusivity equation. Some deductions about the form of the derived
equation are:

1. The partial differential equation in terms of density is linear (Eq. 9.6).

2. The partial differential equation in terms of pressure (Eq. 9.7) is non-linear but if
the pressure gradient is small, the resulting equation (Eq. 9.8) becomes linear.

The assumptions inherent in the derivation of the diffusivity equation are as follows:
(a) The formation is homogeneous, horizontal and of uniform thickness.
(b) Flow is radial, with gravity and capillary effects negligible.
(c) The fluid is considered to be of slight and constant compressibility.
(d) The pressure gradient in the reservoir is considered to be small.
(e) No reaction between fluid and formation matrix occurs.
(f) Reservoir is at isothermal condition.
(g) Flow is laminar.

9.1.2 Gas Flow Equations
Equations 9.1 to 9.4 also hold for gases. In addition, the equation of state for gases is:

=
PM
ZRT
9.9

For real gases, isothermal compressibility, c, is defined as:

c =
1
P
T
-
1
z

dz
dp

(
9.10

For ideal gases,

c =
1
P
9.11

Combining Eq. 9.4, Eqs. 9.9 and 9.10, the different forms of gas flow equations will be:

A. Pressure-Squared Form (p
2
Form)



. [
kpM
ZRT
] =
cpM

p
t

p
RT
9.13

Differentiating Eq. 9.13 and assuming that permeability is constant gives:


1
Z
-
1
Z
Z)
dp
( p =
c
kZ

p
t
2
2
2
2

p
d (
2
2
2
) 9.14

Hence,
138



2 2
p -
d
dp
[ ln ( Z)] ( p =
c
k

p
dt
2
2
2
2

) 9.15

Under certain conditions, Eq. 9.15 simplifies to :


dt
p

k
c
= p
2
2 2

9.16

Equation 9.16, which is the commonly used p
2
- form of gas flow equation follows from
Eq. 9.15 and assuming that the pressure gradient is small everywhere in the system. For
ideal gases, Eq. 9.16 holds but the assumption of small pressure gradient is not necessary.
This is because for ideal gases, the gas compressibility factor z is unity and ideal gas
viscosity is not dependent on pressure. The viscosity of ideal gas only depends on
temperature. Hence, the second term on the left-hand side of Eq. 9.15 will always be
zero.

B. Pseudo -Gas Pressure Form (m (p) Form)
Equation 9.13 still holds. That is



. [
1
z
p =
c
kZ

p
t
2
2

] 9.13

The pseudo-gas pressure is defined as:

p
P
p
z
2pdp
= m(p)

9.17
Thus,

m (p) =
1
z
p
2

9.18

Substituting Eq. 9.18 into Eq. 9.13 gives:

2
m(p) = 9.19

Equation 9.19 holds for both real gas and ideal gas with no assumptions on the magnitude
of pressure gradient term. Equation 9.19 appears linear but it is not as and z are
dependent on pressure. The resemblance between the gas flow equation (Eq. 9.19) and
liquid flow equation (Eq. 9.8) form the basis of adapting liquid flow solution to gas flow
problems. Summary of the equations of state and resulting flow equations are give in
Table 9.1.

139

Table 9.1: Summary of Flow Equations


Continuity Equation: ) (
t
}
r
p

rk
{
r
1
=

r



Equation of State

Resulting Flow Equation

Liquid

] P - c[P
sc
sc
e =

2
p =
ct
k

p
t

Gas

=
PM
ZRT




2
p -
d
d p
[ln Z] ( =
ct
k

p
2
2
2
2

p
t
)

2
m(p) =
c
k

t

m(p)

9.1.3 Relationship between m(p), p
2
and p

The real gas pseudo-pressure is defined as:

=

p
P
b
z
2pdp
= m(p)

9.17

At low pressure (p < 2000 psi)

z =
i
z
i
= constant 9.20a
Hence,

m (p) =
P
i
2
z
i
9.20b

for P
b
= 0

Substituting Eq. 9.20 into Eq. 9.19, it can be shown that the m(p) form of the gas
equation is exactly the same as the p
2
form at low pressure.

At high pressure (p > 3000 psi)
140


p
z
i

=
P
z
i
i
9.21

Hence

m(p) =
2 p
z
P
i
i

i
9.22

for P
b
= O

Substituting Eq. 9.22 for m(p) into Eq. 9.19, it can be shown that the m(p) form of the
equation degenerates to the p-form. That is, the gas flow equation will be same as the
liquid flow equation (diffusivity equation).

The relationships in Eqs. 9.20a and 9.21 are shown in Figure 9.2. Note that in Figure 9.2,
the assumptions that z is constant at low pressure and
z
P
is constant at high pressure
may be acceptable for practical purposes.
141

Figure 9.2: Variation of Gas Deviation Factor-Viscosity Product for a Real Gas (after
Wattenberger and Ramey, 1968)

Calculation of Real Gas Potential, m(p)
The m(p) can be calculated graphically or read from Tables. The graphical method
requires that p, , z be given and m(p) calculated from area under a curve. This is
illustrated below for a case where the trapezoid method is used in calculating the area
under the curve.
Table 9.2: Graphical Method of Calculating m(p)
P z 2P/( z) Area m(p)
P
0

0
z
0
2P/(z)
0
= Y
0

P
1

1
z
1
2P/(z)
1
= Y
1
A
1
= 0.5(Y
0
+Y
1
)( P
1
-P
0
) A
1

142
P
2

2
z
2
2P/(z)
2
= Y
2
A
2
= 0.5(Y
1
+Y
2
)( P
2
-P
1
) A
1
+ A
2

P
3

3
z
3
2P/(z)
3
= Y
3
A
3
= 0.5(Y
2
+Y
3
)( P
3
-P
2
) A
1
+ A
2
+ A
3


P
n

n
z
n
2P/(z)
n
= Y
n
A
n
= 0.5(Y
n-1
+Y
n
)( P
n
-P
n-1
) A
1
+ A
2
+
A
3
++ A
n


The graphical method is illustrated in Figure 9.3.

















Figure 9.3: Graphical Method of Obtaining m(p)

Tables 9.3 and 9.4 can be used in obtaining m(p). Note that the data in the tables were
generated using some assumed typical gas properties. Both tables give reduced pseudo-
pressure defined as:
2
1
2
c
r
P

= 9.22a
where

= m(p)

1
= initial viscosity
P
c
= critical pressure

From Eq 9.22a, m(p) can be calculated.

The differences between values in both tables are simply due to the use of different bases.
Table 9.3 uses reduced pressure of 0.2 as base while Table 9.4 uses reduced pressure of 0
as base. However, the values can be reconciled with the knowledge that

2p/(z)
p
m(p
)
P
b

143

2 . 0
0 0 2 . 0
y y y
p p
9.22b


Note that at low pressure, m(p) can simply be calculated with Eq 9.20b.

144
Table 9.3: Reduced Pseudo-Pressure Integral ( r) With 0.2 as Base
Pseudo-
Reduce
d
Pressur
e


Values of integral for Pseudo-Reduced Temperature T
pr
of
P
pr
1.05 1.15 1.30 1.50 1.75 2.00 2.50 3.00
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
2.10
2.20
2.30
2.40
2.50
2.60
2.70
2.80
2.90
3.00
3.25
3.50
3.75
4.00
4.25
4.50
4.75
5.00
5.25
5.50
5.75
6.00
6.25
6.50
6.75
7.00
7.25
7.50
7.75
8.00
8.25
8.50
8.75
9.00
9.25
9.50
9.75
10.00
10.50
11.00
11.50
12.00
12.50
13.00
13.50
14.00
14.50
14.00
0.0257
0.0623
0.1102
0.1698
0.2418
0.3264
0.4236
0.5326
0.6546
0.7903
0.9484
1.1444
1.3671
1.5828
1.7924
1.9959
2.1926
2.3821
2.5649
2.7424
2.9147
3.0825
3.2464
3.4066
3.5633
3.7169
3.8679
4.0165
4.3788
4.7278
5.0653
5.3938
5.7144
6.0276
6.3347
6.6368
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.0229
0.0553
0.0971
0.1485
0.2105
0.2835
0.3678
0.4631
0.5691
0.6855
0.8126
0.9503
1.0980
1.2546
1.4191
1.5883
1.7595
1.9321
2.1071
2.2841
2.4619
2.6399
2.8172
2.9937
3.4683
3.3403
3.5094
3.6766
4.0876
4.4874
4.8766
5.2579
5.6367
6.0088
6.3897
6.7235
7.0706
7.4124
7.7495
8.0821
8.4099
8.7330
9.0520
9.3670
9.6786
9.9876
10.2936
10.5963
11.1935
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.0198
0.0477
0.0839
0.1283
0.1810
0.2419
0.3111
0.3889
0.4755
0.5707
0.6734
0.7838
0.9020
1.0277
1.1606
1.3001
1.4457
1.5966
1.7526
1.9138
2.0791
2.2473
2.4186
2.5935
2.7710
2.9504
3.1320
3.3153
3.7771
4.2400
4.7052
5.1693
5.6277
6.0822
6.5308
6.9714
7.4044
7.8304
8.2497
8.6632
9.0711
9.4731
9.8703
10.2635
10.6531
11.0398
11.4223
11.7998
12.1731
12.5433
12.9102
13.2735
13.6340
13.9925
14.3483
14.7011
15.3996
16.0892
16.7703
17.4427
18.1069
18.7642
19.4147
20.0588
20.9676
21.3318
0.0170
0.0409
0.0716
0.1091
0.1532
0.2037
0.2608
0.3246
0.3954
0.4734
0.5579
0.6484
0.7449
0.8473
0.9558
1.0703
1.1906
1.3464
1.4474
1.5838
1.7253
1.8712
2.0214
2.1758
2.3341
2.4957
2.6612
2.8308
3.2685
3.7223
4.1897
4.6678
5.1539
5.6459
6.1412
6.6377
7.1355
7.6343
8.1338
8.6336
9.1326
9.6297
10.1249
10.6185
11.1091
11.5957
12.0794
12.5615
13.0416
13.5194
13.9939
14.4644
14.9322
15.3980
15.8609
16.3205
17.2313
18.1318
19.0212
19.8976
20.7640
21.6238
22.4762
23.3216
24.1596
24.9921
0.0145
0.0348
0.0609
0.0927
0.1303
0.1734
0.2221
0.2763
0.3358
0.4004
0.4702
0.5452
0.6255
0.7114
0.8025
0.8983
0.9988
1.1042
1.2146
1.3293
1.4498
1.5744
1.7034
1.8370
1.9751
2.1169
2.2626
2.4123
2.8038
3.2178
3.6504
4.0997
4.5638
5.0406
5.5280
6.0234
6.5252
7.0320
7.5449
8.0622
8.5836
9.1085
9.6364
10.1665
10.6973
11.2279
11.7587
12.2897
12.8211
13.3532
13.8858
14.4187
14.9513
15.4834
16.0146
16.5447
17.6030
18.6590
19.7090
20.7507
21.7858
22.8166
23.8434
24.8616
25.8642
26.8596
0.0126
0.0303
0.0530
0.0807
0.1132
0.1505
0.1927
0.2397
0.2915
0.3483
0.4098
0.4758
0.5461
0.6209
0.7001
0.7840
0.8724
0.9653
1.0624
1.1636
1.2687
1.3777
1.4904
1.6068
1.7268
1.8504
1.9778
2.1094
2.4534
2.8178
3.2016
3.6049
4.0268
4.4663
4.9203
5.3860
5.8621
6.3472
6.8412
7.3442
7.8551
8.3739
8.8993
9.4298
9.9647
10.5034
11.0452
11.5897
12.1377
12.6897
13.2440
13.7993
14.3558
14.9128
15.4700
16.0274
17.1463
18.2662
19.3931
20.5120
21.6135
22.7156
23.8144
24.9057
25.9948
27.0862
0.0100
0.0241
0.0421
0.0640
0.0898
0.1194
0.1529
0.1902
0.2312
0.2761
0.3248
0.3773
0.4335
0.4932
0.5566
0.6235
0.6940
0.7679
0.8454
0.9264
1.0111
1.0994
1.1912
1.2862
1.3846
1.4864
1.5915
1.6998
1.9849
2.2896
2.6119
2.9516
3.3077
3.6788
4.0649
4.4664
4.8825
5.3130
5.7575
6.2150
6.6844
7.1643
7.6544
8.1543
8.6633
9.1808
9.7064
10.2398
10.7812
11.3308
11.8872
12.4497
13.0182
13.5926
14.1700
14.7499
15.9178
17.0928
18.2738
19.4614
20.6575
21.8627
23.0724
24.2820
25.4964
26.7197
0.0083
0.0200
0.0349
0.0532
0.0747
0.0993
0.1271
0.1580
0.1920
0.2292
0.2695
0.3129
0.3594
0.4090
0.4616
0.5173
0.5760
0.6378
0.7025
0.7700
0.8401
0.9144
0.9907
1.0700
1.1522
1.2373
1.3232
1.4159
1.6550
1.9144
2.1841
2.4731
2.7782
3.0994
3.4357
3.7865
4.1511
4.5286
4.9194
5.3241
5.7413
6.1699
6.6104
7.0633
7.5283
8.0049
8.4921
8.9884
9.4932
10.0062
10.5281
11.0583
11.5962
12.1421
12.6952
13.2545
14.3923
15.5560
16.7372
17.9315
19.1388
20.3556
21.5858
22.8246
24.0719
25.3268

145
Table 9.4: Reduced Pseudo-Pressure Integral ( r) as a Function of Tr and Pr.
146
9.2 SOLUTION OF LIQUID FLOW EQUATIONS
Solutions of the liquid flow equation (diffusivity equation) are discussed in this section.
Three cases - steady state, pseudo-steady state and the unsteady state (transient state) -
are considered. Conditions under which they are attained are also given.

9.2.1 Steady State Solution
Steady state condition is attained when the rate of change of pressure is zero in the entire
reservoir. That is,

0 =
t
p

9.23

The condition expressed in Eq. 9.23 can be attained in a reservoir undergoing depletion if
there is influx of some material (e.g. water) at the external boundary to recharge the
reservoir and hence prevent the pressure from dropping as a result of the production at
the well. This implies that for a reservoir to attain steady state condition, its boundary
must be open to flow. However, the entire boundary does not necessarily need to be open
to flow. As long as the water influx from any part of the boundary is strong enough to
keep the pressure in the reservoir from changing, the system will attain steady state
condition. The diffusivity equation, inner and outer boundary conditions for the steady
state case are given as:

Diffusivity Equation

0 = )
r
p
r (
r r
1
r
p

r
1
+
2
2

r
p
9.24

Inner Boundary Condition

r
dp
dr
= = constant

(
r
q
kh
w

2
9.25

Outer Boundary Condition

p = p
e
at r = r
e
9.26

The solution to Eq. 9.24 subject to the inner boundary condition is:

p - p =
q
2 kh
ln
r
r
w
w

9.27

Applying the condition at the outer boundary (r = r
e
, p = p
e
), Eq. 9.27 becomes:
P - P =
q
2 kh
ln
r
r
e w
e
w

9.28
147

Rearranging, Eq. 9.28 gives

q =
2 kh

[P
e
- P
w

r
e
r
w

]
ln
9.29

The volumetric average pressure for any system is given as follows:

p =
pdV
V
o
V
T
T

9.30

where

V = (r - r
2
w
2
)h 9.31

And

V h
T
= (r - r
2
e
w
2
) 9.32

From Eq. 9.13,

dV = 2rhdr 9.33

We can use Eqs. 9.30 to 9.33 to determine the expression for average pressure for a
system at steady state. Substituting Eqs. 9.32 and 9.33 into Eq. 9.30 and integrating gives:

p
r r
e e
= p +
q
2 kh

r
- r
1n r - 1n r -
r
- r
1n r -
1
2

w
e
2
w
2 w w
w
2
w
2 w

2 2

(
9.34

For re >> r
w
, Eq. 9.34 becomes

]
2
1
-
r
[1n
kh 2
q
+ p =
e
w
w
r
p

9.35

Rearranging Eq. 9.35,

q
r
w
=
2 kh [p - p ]
[1n
r
-
1
2
]
w
e

9.36

148
=
]
r
0.606 1n [
] p - p [ kh 2
e
w
w
r

9.37

Combining Eqs. 9.28 and 9.36 gives:
p
e
- p = (p - p - (p - p =
q
4 kh
e w w
) )

9.38

Equation 9.38 shows the relationship between the pressure at the external boundary and
the average pressure for a system that attained steady state. Equation 9.37 is not a very
useful equation because of the difficulty in obtaining the average pressure. When a well
that attains steady state is shut in, the pressure does not build up to the average pressure.
The pressure builds up to the initial pressure because of influx from contiguous aquifers.
And for a system that attains steady state, p
e
= p
i
.

9.2.2 Pseudo-Steady State Solution
A reservoir attains pseudo-steady state if the rate of pressure decline with time is a
constant. At that state, the mass rate of production is equal to mass rate of depletion.
Further discussion on pseudo-steady state follows. Pseudo-steady state is also referred to
as semi-steady state, quasi-steady state and in some cases, wrongly referred to as steady
state.

The conditions necessary for this state to be attained are as follows:
(a) Reservoir outer boundary must be closed to flow.
(b) Well producing from finite drainage volume due to long production time as
shown in Figure 9.4. The finite drainage volume will be in proportion to the
production rate of that particular well in the drainage area. That is, a well producing
at a rate of 2q will drain twice the drainage volume of a well producing at rate q.








Figure 9.4: Reservoir Depletion under Pseudo-Steady State

Mathematically, at pseudo-steady state condition,

p
t
=
dp
dt
= constant 9.39

2q
q
q
149
Ideally, only reservoirs with closed outer boundaries should attain pseudo-steady state but
partially closed reservoir with very limited influx may attain a state, which for practical
purpose may be considered to be pseudo-steady state.

9.2.2.1 Pseudo-Steady State Equation
Equations for pseudo-steady state are given in this section. First, the value of the constant
(

p
t
= constant) is determined. Secondly, the solution to the pseudo-steady state
equation is given and lastly, the expressions for the volumetric average pressure and other
related equations are derived.

A. Derivation of Value of Constant
When pseudo-steady state condition is attained,

Mass production rate = rate of mass depletion

That is,

q
w
= V
d
dt
( )
T
9.40

= V c
dp
dt
T t


Substituting for the total volume gives:

q
w e
2
w
2
t
= - [r - r ] h c
dp
dt
9.41

Therefore,


dp
dt
=
- q
(r - r ) h c

e
2
w
2
t
w

From the equation of state for a constant compressibility fluid

w
= e [p - p ]
-c
w
9.42

Substituting Eq. 9.42 into Eq. 9.41 gives


dp
dt
=
- q
[r - r ] h c
e
e
2
w
2
t
-c [ p - p
w

]
9.43

150
For liquid of small compressibility,

e
c [

p - p
w
1
]


Hence


dp
dt
=
- q
[r - r ] h c
=
p
t
e
2
w
2
t

9.44

=
q
Vc
t
= constant 9.45

Equation 9.45 is the basis of reservoir limit test. This equation implies that if the
pressure-time data obtained during pseudo-state is graphed on a Cartesian paper, a
straight line whose slope is related to the wells drainage volume is obtained. Equation
9.45 can be derived from a volume balance (a degenerate form of mass balance, see Craft
and Hawkins, page 286), but such derivation does not explicitly show that an assumption
was made about the fluid compressibility.

B. Pseudo-Steady State Solution
The governing diffusivity equation is given by Eq. 9.8. At pseudo-steady state, the rate of
pressure change is given by Eq. 9.44. Substituting Eq. 9.44 into Eq. 9.8, the diffusivity
equation at pseudo-steady state becomes:


1
r

r
(
r p
r
=
- q
kh (r - r
e
2
w
2

)
)
9.46

Using the inner boundary condition given by Eq. 9.25 and the fact that the pressure at any
point, r, in the reservoir is p, the solution to Eq. 9.46 is

(
)
( )
)
P - P =
q
2 kh (r - r

1
2
r - r + r ln r / r
w
e
2
w
2 w
2 2
e
2
w

(
9.47

For r
e
>> r
w
, Eq. 9.47 becomes

P - P =
q
2 kh
[ln r / r -
1
2
r
w w
2

/ ] r
e
2
9.48

When r = r
e
, p = p
e
, then Eq. 9.48 becomes

P - P =
q
2 kh
[ln
r
r
-
1
2
e w
e
w

] 9.49

151
Hence
q =
2 kh [p - p
[ln
r
r
-
1
2
e w
e
w

]
]
9.51

C. Volumetric Average Pressure
Using the same principles as in steady-state case, the volumetric average pressure when
pseudo-steady state condition is attained is:

p = p +
q
2 kh [r - r
[
r
2
- r ln r
w
e
2
w
2
w
2
e
2
w

]
] -


+
q
kh [r - r
{r ln r -
r
4
ln +
r
4
(r
e
2
w
2 e
2
e
e
2
w
2
e
4

]
( ) )}
2
2 2 4
1
2
1
2
1
8
r r r
e w w

9.52
For r
e
>> r
w
, Eq. 9.52 becomes

p = p +
q
2 kh
[ln
r
r
-
3
4
w
e
w

] 9.53

Hence

q =
2 kh [p - p
[ln
r
r
w
e
w

]
]
3
4


=
2 kh [p - p
ln 0.472
r
r
w
e
w

]
9.54
Some deductions from the pseudo-steady state equation are:

1. The wellbore pressure p
w
, average pressure p , and the pressure at the external
boundary P
e
in Eq. 9.54 are dependent on time. This is not explicit in Eq. 9.54.

2. When a system with closed outer boundaries is shut-in, the pressure builds up to the
average pressure. Hence Eq. 9.54 is a useful equation.

3. Note the similarities between Eq. 9.37 derived for steady state case and Eq. 9.51
derived for the pseudo-steady state case. The equations are summarized in Table 9.5.

152
4. The term 0.472r
e
in Eq. 9.54 is usually replaced with r
d
called the drainage radius.
This is a misnomer because it gives the impression that just a part of the reservoir is
being drained. Actually, the whole reservoir is drained.


Table 9:5: Summary of Stabilized Flow Equations in Darcy Units
Steady State Pseudo-Steady State (r
e
> >> >> >> >r)

General relationship between p and r

p p
w
=
q
2 kh
ln
r
r
w

p p
r
w
e
=
q
2 kh
[ln
r
r
-
r
w
2

2
2
]


Flow equations when p = p
e
and r = r
e


p =
q
2 kh
ln
r
r
e
w
e w
p

p =
q
2 kh
ln
r
-
1
2
e
e w
w
p
r

Flow equations in terms of reservoir average
pressure

p - p =
q
2 kh
ln
r
r
-
1
2
w
e
w

(



p - p =
q
2 kh
ln
r
r
-
3
4
w
e
w

(




D. Other Useful Pseudo-Steady State Equations
By combining the pseudo-steady state equations, the relationship between the reservoir
average pressure, the initial pressure P
i
and the pressure at the external boundary P
e
during pseudo-steady state are obtained as follows:
P
i
- p =
qt
[r - r
h c
e
2
w
2 t
]
9.55a

p
i
p = (p
i
- p) + ( p - p
w
) - (p - p
w
)

=
q
2 kh
[2
r
r

kt
c r
+ ln
r
r
-
3
4
+
1
2
(
r
r
]
w
e t w
2
e

(

2
2
)
e
9.55b

(p
e
- p) = (p
e
- p
w
) - ( p - p
w
) =
q
8 kh

9.56
153

p
i
- p
e =
(p
i
- p) - (p
e
- p) =
q
2 kh
[
2kt
c r
-
1
4
t e
2

] 9.57

9.2.3 Unsteady State (Transient State) Solution
During the transient state, the rate of pressure change in the reservoir has a value which is
neither zero nor a constant. Transient state behavior occurs in every system when the
boundary effects are not yet felt. For example, a reservoir that is infinite (let us for now
assume that there is something like that) will always be in transient state. Also, reservoirs
that are finite (the boundary may or may not be closed to flow) will undergo transient
behavior at early time when the boundary effects or interference due to production from
and/or injection into other wells are not yet felt. The early time when the transient state
behavior persists is generally known as the infinite- acting stage.

In this section, a simple analytical solution of the transient state equation is given. The
solution may be obtained by Boltzman transformation discussed in Matthew and Russel
(1967). The governing equation and associated boundary and initial conditions are:

Diffusivity Equation

2
+
1
r

p
r
=
c
k

p
t
t
p
r
2

9.58

Inner Boundary Condition (Constant Production)


kh 2
q
= ]
r
p
[
0
r

w
r
9.59

Outer Boundary Condition

Lim p = p
i
as r 9.60

Initial Condition

P = P
i
, when t = O 9.61

Equation 9.59 specifies that the well has a vanishing wellbore radius. That is the wellbore
is considered to be a line and hence the solution to the diffusivity equation obtained with
such inner boundary condition is called the line source solution. If a finite wellbore inner
boundary condition is considered, a solution can be obtained by Laplace transformation.

154
Although the outer boundary condition (Eq. 9.60) indicates that the system is infinite but
this is just a mathematical representation of systems where the boundary effects or effects
due to interference have not been felt.

Solution to diffusivity equation subject to the given boundary and initial condition is:

p(r, t) = p
i
+
q
kh
r
kt

4 4
2
E - (
c
i
t

) 9.62

For x < 0.0025,

E
i
(- x) ln (x) = ln x + 0.5772 9.63

The factor, , is the Eulers constant and it is equal to 1.781. E
i
(- x) is called the
exponential integral function and is defined as

E
i
(- x) = -
e
u
du
-u
x

9.64

Applying the approximation (generally called log approximation) given as Eq. 9.62, for


c r
kt
t
2
4
< 0.0025. That is (
kt
c
t
r
> 100)
2
,

the transient state solution becomes

p(r, t) = p
i
+
q
4 kh
ln
( c r
t
2

)
4kt
9.65

At the well
p(r
w
, t) = P
i
+
q
4 kh
ln
( c r
t
2

)
4kt
9.66

The above solution is the line source solution also called the exponential integral solution
or Theis solution. Equation 9.65 makes it possible to predict the pressure at any point r, in
the reservoir due to constant rate production at the well. This equation forms the basis of
interference test.
Equation 9.66 makes it possible to predict the pressure at the well due to constant rate
production at the well. This equation forms the basis of drawdown tests

The interest in the transient state solution is due to the fact that every well, at early time,
goes through the transient state. That is, the well behaves like a single well in an infinite
reservoir. At later times, the effects of other wells, reservoir boundaries, aquifer influence
cause it to deviate from the infinite acting behaviour. Figure 9.5 is a semilog graph of
155
bottom-hole flowing pressure versus time showing that every system irrespective of the
boundary condition initially goes through the transient state phase.



Figure 9.5: Plot of P
wf
versus log t for Reservoirs with Different Outer Boundary
Conditions

Determination of E
i
Function
For practical purposes, the log approximation to the E
i
function holds for
kt
c
t
r
2
> 5 with about 2 percent error.

When the log approximation does not hold, the function can be read from Table 9.6.


Table 9.6: Values of the E
i
Functions














156

























9.3 SUMMARY OF TRANSIENT, PSEUDO-STEADY AND STEADY STATE
FLOW EQUATIONS
The equations describing flow during the transient, pseudo-steady (semi-steady) and
steady state periods have been given. A brief review of the different types of flow is
given in the paper published by Matthews (1986).

Further discussion on transient, pseudo-steady, and steady state flows is given by Ramey
(1975), Matthew and Russel (1967) and Earlougher (1977). A summary of the governing
equations during the flow periods is given in Table 9.7. The equations are given in both
Oilfield and Darcys Units. In the equations in Table 9.7, the skin effect is neglected.

Table 9.7: Summary of General Flow Equations
Flow Type: Transient

Darcy Units:
P = P +
q
4 kh
E
- ct r
kt
i i
2

(
4


Oil Field Units: (

kt 0.00264 x 4
r ct -
E 70.6 + P =
2
i i

kh
qB
P
Practical SI Units
(

kt 0.00264 x 4
r ct -
E 10 x 9.33 + P =
2
i
2
i

kh
qB
P
157

Flow Type:

Pseudo-Steady State

Darcy Units:
P = P +
q
2 kh

r
r
-
3
4
w
e
w

ln

(


Oil Field Units:
P = P + 141.2
qB
kh

r
r
-
3
4
w
e
w

ln

(

Practical SI Units
(

4
3
-
r
r
ln
kh
qB
10 x 1.866 + P =
w
e 3
w

P

Flow Type:

Steady State

Darcy Units:
P = P +
q
2 kh
ln
r
r
w
w

Oil Field Units:
P = P + 141.2
qB
kh

r
r

w
w

ln

(


Practical SI Units
(

r
r
ln
kh
qB
10 x 1.866 + P =
w
3
w

P


9.4 LIQUID FLOW SOLUTIONS IN DIMENSIONLESS FORMS
In this section, the dimensionless forms of the diffusivity equation, steady state,
pseudosteady state and transient state solutions are given

Diffusivity Equation:

2
D
D
D
D
D
+
1
r

p
r
=
p
t
P
r
D
D
2
9.66

Steady State Solution

p
t
D
D
= O

(


q =
2 kh

[P - P

r
r
e w
e
w

]
ln
9.67

and P
e
= P
i


Then in dimensionless form

158
P =
2 kh
q

Dw
[P
i
- P
w

]
= ln
r
e
r
w
9.68

Generally,


r
r
q
p P kh
P
e i
D
ln
] [ 2
=

9.69

Pseudo-Steady State Solution,
( )

p
t
D
D
= constant
P - P =
q
2 kh
2
r
r

kt
c r
+ ln
r
r
-
3
4
+
1

r
i
w
e
2
t w
2
e

|
\

|
\

(
(
2
2
r
e
9.70


In dimensionless form,

P =
2 kh
q
(P - P
Dw
i
w

) =
2t
r
+ ln r -
3
4
+
1
2
r
D
De
2 De De
2
9.71
Unsteady State (Transient State) Solution (Line Source Solution)

P = P +
q
4 kh
E
(- cr
i i
2

)
4kt
9.72

In dimensionless form

P = -
1
2
E
(-r
D i
D
2
4
)
t
D
9.73

For
r
t
D
D
2
4
< 0.0025, the log approximation holds. That is,

P = -
1
2
ln
( r
D
D
2
)
4t
D
9.74


(

+ = =
303 . 2
80907 . 0 t
log 1.15 ] 0.80907 +
r
t
[1n
2
1

2
D
2
D
D
D
r
9.75

Figure 9.6 is a schematic of dimensionless time for an infinite system, closed outer
boundary system and constant pressure outer boundary system.


159

















Figure 9.6: Dimensionless Pressure and Time for Different Reservoir Outer Boundary
Conditions

9.5 COMPARISON BETWEEN LINE SOURCE SOLUTION AND FINITE
WELLBORE SOLUTION
The line source solution also called the exponential integral solution is very easy to
obtain using the Boltzman transformation technique. The finite wellbore solution can be
obtained in the Laplace space using Laplace transform technique. Solution in real space
may be obtained numerically using Stefhest algorithm. The comparison between the
exponential integral solution and finite wellbore solution is shown in Fig. 9.7.

Deductions that may be made from Fig. 9.7 are:

1. When
t
r
D
D
2
25 (for all r
D
) the exponential integral solution and the finite wellbore
solution are same
2. Also at locations for which r
D
20 both solutions are also the same when
t
r
D
D
2
0.5.

160

Figure 9.7: Line Source Solution and Finite Wellbore Solution

9.5 GENERALIZED FLUID FLOW EQUATIONS
So far, we have considered a single well draining a circular reservoir. The dimensionless
pressures for such case have been given for different conditions. The drainage area in
many reservoirs may not be circular. Hence, the reservoir geometry must be accounted
for. This is done by the introduction of Dietz shape factors, which allows a general form
of inflow equation to be developed for a wide range of geometries of drainage area and
positions of the well within the boundary.

The pressure drop at any point in a single-well reservoir being produced at constant rate
q, is then described with the general solution:

( )
{ }
P - p =
q
2 kh
P
D
t geometry
i D

, , , r C
D D

9.76

The dimensionless pressure in Eq. 9.76 is dependent on dimensionless time,
dimensionless radius, wellbore storage dimensionless constant, and reservoir geometry.
The skin factor is added to the dimensionless pressure only when calculating the pressure
at the well. That is

( ) { } P - p =
q
2 kh
P t geometry + S
i w D D

, , , r C
D D

9.77

Practically speaking, dimensionless pressure is just a number that may be given by an
equation, read from a table or a graph for different systems. The equations for
calculating dimensionless pressure for some systems are given.

161
A. Infinite (Infinite Acting) System without Wellbore Storage and Skin

P
D i
D
2
D
= -
1
2
E
r
4t
|
\

|
9.78

5 >
r
t
For
2
D
D


P
r
D
D
=
1
2
ln
t
+ 0.80907
D
-
-
2
|
\

9.79

Equations 9.78 and 9.79 should be familiar.


B. Closed System (General Case)
For all closed systems undergoing transient flow, the dimensionless pressure is given
within 1 percent as:

P
wD
=
1
2
[ln t + ln
A
r
+ 0.80907 ]
DA
w
2

9.80

where 0.000025 < t
DA
< t`
DA
and t`
DA
is the time to use infinite system solution with
less than 1 percent error. Values of t`
DA
for different reservoir geometries and well
locations are given in the last column of Table 9.8 taken from Earlougher (1975). Does
Eq. 9.80 imply that the transient state equation depends on drainage area? No. The area
term in Eq. 9.80 cancels out if t
DA
is expressed as a function of t
D
and A. Equation 9.80
is exactly the same as Eq. 9.79 for r
D
= 1 (wellbore).
162
Table 9.8: Shape Factors for Various Closed Single - Well Drainage Areas
163
164
At long times, every closed system attain pseudo-steady state and the dimensionless
pressure is given as:

P
wD
= 2 t +
1
2
ln
A
r
+
1
2
ln
2.2458
C
DA
w A
2

9.81

Equation 9.81 applies for different systems when t
DA
is greater than the time under the
column with heading, exact for t
DA
> given in Table 9.8. With a maximum of 1%
error, the time when pseudo-state will start can be read from the column with the heading
less than 1 percent error for t
DA
>. This makes it possible to use the pseudo-state
equation to calculate pressure in a period that ideally is considered a transition period
(transition from transient to pseudo-steady state).

The C
A
in Eq. 9.81 is the geometric or shape factor, which is also given in Table 9.8 for
different drainage systems and well locations. From Eq. 9.81, a Cartesian graph of
dimensionless pressure versus dimensionless time based on area gives a straight line with
slope of 2 when pseudo-steady state condition is reached.


C. Infinite System with Single Vertical Fracture (No Storage and Skin)
(i) Uniform Flux Fracture: For uniform flux fracture, fluid enters the fracture at a
uniform flowrate per unit area of fracture face so that there is a pressure drop in the
fracture. Uniform flux fractures are closer to natural fractures. The dimensionless
pressure for uniform fracture is:

P = t erf
1
1
-
1
2
E -
1
4t
wD Dxf i
Dxf

2
t
Dxf
|
\

|
9.82

where dimensionless time is based on half-fracture length defined as

t = t
r
x
Dxf D
w
f

(
2


The function erf is the error function defined as:

erf (x) =
2

e du
-u
2
o
x

9.83

For t
Dxf
> 10, Eq. 9.82 becomes

165
P
wD
=
1
2
[ln t + 2.80907]
Dxf

9.84

With less than 1 percent error and for t
Dxf
< 0.1,

P
D
= t
Dxf

9.85

Taking the log of Eq. 9.85 gives

log P
D
=
1
2
log [t + ]
Dxf

9.86

Equation 9.86 implies that a log-log plot of P
D
versus t
Dxf
gives a straight line with a
slope of 0.5. This implies that at short times, flow into fracture is linear. Linear flow is
characterized by a slope of 0.5 on log-log graph of P
D
versus t
Dxf or (p versus t)
.

(ii) Infinite Conductivity Fracture: In this case, the fracture is considered to have
infinite permeability and therefore uniform pressure in the fracture. The dimensionless
pressure for infinite conductivity fracture systems is given by Earlougher (1977) as:

P
D
=
1
2
t [erf(
0.134
t
+ erf (
0.866
t
]
Dxf
Dxf Dxf
) )

-0.067
Dxf
i
Dxf
t
(-0.750)
E 0.433 - )
t
0.018 -
(
i
E
9.87

But when t
Dxf
> 10,

P
D
=
1
2
[ln t + 2.20]
Dxf

9.88

Also, with less than 1 percent error and for t
Dxf
< 0.01

P
D
= t
Dxf

9.89

Equation 9.89 indicates that at short times, flow into infinite conductivity fracture is also
linear.
166
D. Closed System with Single Vertical Fracture (No Storage and Skin)
Equations given for the infinite system with single vertical fracture apply but when the
system attains pseudo-steady state.

P
x
x C
D
e
f A
= 2 t +
1
2
+
1
2
DA
ln ln
.

(
2
2 2458

9.90

where x
e
is half-length of the side of the system parallel to the fracture.

9.6 WELLTEST DESIGN
In new fields or special circumstances, it is important that tests be designed before they
are run. The objectives of such test design are as follows:

1. To determine the duration of the flow phases
2. To determine the expected pressure changes.
3. To ensure that correct equipment (especially gauges) are used.

During test design, values of well and reservoir parameters are assumed while equations
are used to calculate the required parameters. This will be illustrated with an example.

Example: Table 9.9 shows assumed and estimated well and reservoir parameters. Using
the values in Table 9.9 and assuming well in the center of a 2:1 rectangle and 40 acre
drainage area calculate the following:

(a) The duration of the wellbore storage phase.
(b) Duration of the transient state period
(c) The time when pseudo-steady state will start
(d) Draw the box diagram showing the duration
(e) The maximum pressure change at the end of strong wellbore storage effect
(f) The pressure change at the end of transient
(g) The pressure change at the beginning of pseudo-steady state

Table 9.9: Parameters for Test Design
Parameter Value
Wellbore radius, r
w
, ft 0.5
Total compressibility, c
t
, psi
-1
20 x 10
-6

Porosity, 0.25
Oil formation volume factor, B
o
, rb/STB 1.125
Oil viscosity, , cp 0.6
Production rate, q, STB/D 1000
Formation thickness, h, ft 50
Permeability, md 1000
Skin Factor 25
Wellbore Storage Constant, rb/psi 0.009
167
Solution
A. For drawdown, wellbore storage ends at time

t
kh
ws
=
(200000 + 12000S)C
s
/
= 0.054 hrs
B. For well in the center of 2:1 rectangle, transient ends at a time read from Table 9.8
as t
DATs
= 0.025. Converting t
DATs
to ordinary time gives

t
A
k
t
Ts DATs
=
c
t

0000264 .
. = 0.495hrs

C. For this system pseudo-steady state starts at time read from Table 9.8 as t
DA
PSS
=
0.3. In absolute time, this is

t
A
K
=
c
t
t
DApss

0000264 .
= 5.94 hrs

D. Box Diagram Showing Duration for Drawdown Test












O t(hr)


E. Assuming wellbore storage effect died down completely at 50t*, therefore

t * =
t
ws
50


t
r
D
w
* =
0.000264 kt *
c
= 380.16
t

2


C
c hr
D
t w
=
0.8936 C
= 128.68
s

2


0.054
0.495
Transient phase
Wellbore storage phase
Wellbore storage phase
Pseudo-steady state
5.94
168
Dimensionless pressure at end of strong wellbore storage is
P
C
D
D
=
t *
D

The pressure change now becomes

[ ] p
B
=
141.2q
kh
P + S
D

= 1.9062 [2.95 + 25] = 53.28 psi

F. Dimensionless time at end of transient is

t
c r
D
t w
=
0.000264kt
= 1.74 x 10
Ts 5

2


Dimensionless pressure at end of transient is

P
D i
= -
1
4t
1
2
D

|
\

|

or

[ ]
P In t
D
=
1
2 D
+ 080907 . if log approximation holds.

Log approximation holds, for
t
r
D
D
2
> 5 , therefore, P
D
= 6.44

Pressure drop at the end of transient is

[ ] p
B
=
141.2q
kh
P + S
D

= 1.9062 [25 + 6.44] = 59.93 psi

G. Dimensionless pressure at the beginning of pseudo-steady state is

P
D pss
= 2 t + In
A
r
+ In
2.2458
C
= 8.626
DA
1
2
w
2
1
2
A


Pressure drop at the beginning of pseudo-steady state is then

[ ] S +
kh
141.2qB
=
D
P p

= 1.9062 [8.626 + 25] = 64.10 psi

9.7 SUPERPOSITION
So far, we have considered the well to be produced at constant rate and the effect due to
only one well in a given reservoir system. In this section, we shall use superposition
concept to account for production at different rates and also the effect of having more
than one well in a given reservoir system.
169
Simply put, superposition generally means that any sum of solutions to a linear
differential equation is also a solution. In Petroleum Engineering, this implies that the
total pressure drop at any point in a reservoir that contains more than one well is the sum
of the pressure drop due to each well acting as if it is in the reservoir alone. Hence
superposition concept can be used for generation of changing flowrate or pressure
boundary condition in pressure buildup analysis and water-drive material balance
methods. Also superposition can be used for determining the effect of many wells in a
reservoir.

In this section, the application of superposition will be illustrated with examples.
Superposition in both space and time will be considered. Discussion on these follows.


9.7.1 Superposition in Space
Considering a reservoir containing two wells as shown in Figure 9.8, the pressure drop
due to both wells at point (x,y) in the reservoir is given using superposition in space as:

Pressure drop at x, y = Pressure drop at x, y due to Well 1 +Pressure drop at x, y due to
Well 2 9.91

















Figure 9.8: Superposition in Space


Using the notations in Figure 9.8, Eq. 9.91 implies:

P
i
- P = (P - P + (P - P
x, y, t i a, t i b, t
) )
9.92

Well 2
P(x, y)
b
a
d
Well 2
q
2
q
1
170
Generally, (P - P =
q
2 kh
P (r t
i r, t D D D
) , )

9.93

Therefore,
) t , (b P
2
q
+ ) t , (a P
2
q
= t) y, (x, P -
D D D D D D
2 1
kh kh
P
i

9.94

At Well 1, Eq 9.94 becomes

) t , (d P
2
q
+ ] ) (t [P
2
q
= P -
D D D 1 D D wf1
2 1
kh
s
kh
P
i

+ 9.95

The second term on the right hand side of equation 9.95 is the interference effect due to
Well 2 at Well 1.

At Well 2, Eq 9.94 becomes

] ) (t [P
2
q
+ ) t (d P
2
q
= P -
2 D D D D, D wf2
2 1
s
kh kh
P
i
+

9.96

The first term on the right hand side of equation 9.95 is the interference effect due to Well
1 at Well 2.

For q
1
= q
2
= q, Eq. 9.94 becomes

2kh
qu
t
D
P - P (x, y, t) = P (a + P (b t ( )
i D D D D D
, ) , ) 9.97

At Well No. 1, Eq. 9.97 is:


2kh
qu
P - P = P (1, t + S + P (d t ( )
i wf1 D D 1 D D D
) , ) 9.98

and at Well No. 2, Eq. 9.97 is:


2kh
qu
P - P = P (d , t + P (1 t + S ( )
i wf2 D D D D D 2
) , ) 9.99

Note that at the wells, the skin factor terms are added. Also, for superposition in space,
the wells may have different production rates but for all wells, production must
commence at the same time. Signs are changed (q is negative) for injectors.
171

Implications of Superposition in Space
1. The pressure distribution for one well producing at constant rate q located at a
distance d/2 from a linear flow barrier, or one well in a semi-infinite reservoir can be
simulated by super-imposing solutions of two wells. This is shown in Figure 9.9. The two
wells must either be injectors or producers.











Figure 9.9: Simulation of No-Flow Boundary

2. Similarly, a single well producing near a constant pressure barrier can be simulated
with two wells. In this case, one of the wells must be an injector while the other must be a
producer. This is shown in Figure 9.10.

















Figure 9.10: Simulation of Constant Pressure Boundary

3. Single well producing in closed square can be generated by considering an infinite
array of wells as shown in Figure 9.11.




d/2
d/2
d/2
= =
172













Figure 9.11: Simulation of a Well in a Closed Square
For Figure 9.11, if we want to compute the pressure at the well in the center of a closed
square with dimension A r
w
/ = 2000 and at time t
DA
= 0.1, the procedure is as
follows:
(i) Draw the closed square with the well at center. Then, draw a few of the infinite
arrays of wells around the well of interest. A few of the infinite array of wells is required
because the further the wells are from region of interest, the lesser their contribution to
the total pressure effect. The array of wells is shown in Figure 9.12.

(ii) Of the array of wells around the well of interest, locate wells that are of equal
distance from the well of interest. This was done by assigning the same code number to
wells at equal distance from well of interest as shown in Figure 9.12.

















Figure 9.12: Schematic of Array of Wells.
= =
5
5
4
3
4
5
1
4
2
4
4 3
2
3 4

1
4
5
4
4
2
1
4
3
173
(iii) Calculate the distance from the well of interest to array of wells surrounding it. The
square has an area of A and each side of the square has a length of A . Also the
following relationship holds


t
r
D
D
2
=
kt
cr

2

9.100

The above relation is important because the dimensionless pressure due to each well is a
function of t
D
/r
D
2
. Hence the contribution of each well is easily found from the
exponential integral solution. Note that each well will act as if it is in the reservoir alone.
Table 9.10 shows the calculated distances and dimensionless pressures.

Table 9.10: Calculated Distances, Time and Dimensionless Pressures

Assigned
No.

No. of
Wells

Distance From
Well to Well of
Interest
t r
D D
/
2
=
kt
cr
2

0.1) = (
) / (
2
DA
D D
t
For
r t P


1 4
A
t
DA
.012458

2 4
2A
t
DA
2

.000574


3 4 2 A t
DA
4

.0000025


4 8
5A
t
DA
5

0
(iv) Applying the superposition principle,


2
2 4 5
2
2 2 2
kh
q
P - P = P
D
(t + S + 4 P
D

t
= t
4 P
t
=
t
+ 4 P
t
=
t
+ 8 P =
t
[
i wf
D
D
DA
D
D DA
D
D DA
D
DA

] )
r
r r
t
r
D
D D
D
D
|
\

|
|
+
|
\

|
|
\

|
|
\

|
9.101

The first term on the right hand side of Eq. 9.101 is the contribution of the well of
interest. This contribution is calculated as follows:
t
D
=
kt
cr
w
2

=
kt
cA
.
A
r
D
2

= t .
A
r
DA
D
2
9.102

For the case of interest, it is given that

174
A r
w
/
2
= 4 x 10 and t = 0.1
6
DA


Therefore

P (t = P (0.1 x 4 x 10
D D D
6
) ) = 6.8542

Neglecting the skin factor S, and substituting values in Eq. 9.101 gives

P
D
= 6.8542 + 4 x 0.012458 + 4 x 0.000574 + 4 x 0.0000025 + 0 = 6.90634

This compares favourably with a value of 6.9063 calculated by Earlougher et al (1968).
The same procedure may be used to obtain dimensionless pressures at any point in the
well.

9.7.2 Superposition in Time
Superposition in time is used for handling flow rate that varies with time at a given well.
A schematic of how this is done is for a two-rate case is shown in Figure 9.13.

















Figure 9.13: Superposition in Time (2-Rate Case)
In words, the pressure drop due to a well producing in an infinite reservoir at rate q
1
for
time t, and then for rate q
2
for a period of t is equivalent to the sum of pressure drop
due to the well producing at rate q
1
for time t + t and the same well producing at rate
(q
2
- q
1
) for a period of t.

That is:

[ ] )
D D D
1 2
D
1
i
t , r P
2
) q - (q
+ ) ( , r P
kh 2
q
= ) t (r, P - P (
D
+
kh
t t
D

9.103
q
2
-q
1

t
t +t t
t
q
2

q
1

q
q q
+
=
175

At the well, r
D
=1, P(r,t) = P
wf
and the skin term will be added to give

(P - P =
q
2 kh
P [1, (t + t) +
(q - q
2 kh
P (1, t +
q
2 kh
S
i wf
1 2 1 2
D D D D
) ]
)
)

9.104

If q
2
is zero, Eq. 9.104 becomes

{ } ] [ ] [ ) (
D D D D
1
ws i
t 1, P - t) + (t , 1 P
kh 2
q
= P - P

9.105

Note that the skin term has disappeared and we have changed notation from p
wf
(pressure
well flowing) to p
ws
(pressure well shut-in). If we assume transient and log
approximation,

] 80907 . 0 ) [ln( 5 . 0 ] ) ( , 1 [ + + = +
D D D
t t t t P 9.106

and

] 80907 . 0 [ln 5 . 0 ] , 1 [ + =
D D D
t t P 9.107

Substituting Eq 9.106 and 9.107 into Eq 9.105 gives


)
`

+
t
t t
ln
kh 2
q
= P - P
1
ws i
) (

9.108

Equation 9.108 is the basis of the buildup test.

We have derived the equations for a two-rate case. Figure 9.14 shows how superposition
is done for a three-rate case. You should derive the equations for this case as an exercise.












=
q
t
1
+ t
2
+ t
2

t
1
t
2
t
q
1

q
2

q
3

176











Figure 9.14: Superposition in Time (3-Rate Case)


EXERCISES
1. Estimate the pressures at the well located in the center of a 2:1 rectangle after the
well has produced 800 STB/day of dry oil for 15 minutes and 15 days. Other data are:

P
i
= 3265 psi, k
o
= 900 md,
o
= 6.2 cp, B
o
= 1.02 RB/STB, h = 47 ft
= 0.17, c
t
= 2 x 10
-5
psi
-1
r
w
= 0.50 ft A = 40 acres, s = 7

2. The dimensionless pressure at pseudo-steady state well in the center of a circular
reservoir is:

P
r r r
i
e w e
- P =
q
2 kh
2
r

kt
c
+ ln
r
-
3
4
+
1
2

r
w
w
t
r
w
2
e w

|
\

|
|
\

(
(
2 2


And the generalized dimensionless pressure for any closed system with any geometry is



where C
A
is the geometric factor. Assuming r
e
>>r
w
, find the geometric factor for a well
in the center of a circular reservoir by relating both equations.

3. Ramey and Cobb (1971) have shown that for a well situated at the center of a
regular shaped drainage area, (for instance, a circle square or hexagon) the transient
period is of extremely short duration. Under these circumstances, it is possible to equate
the transient flow equation and pseudo-steady state equation to determine the time at
+
+
q
2
q
1

q
3
q
2

177
which pseudo-steady state condition starts. Find an expression for t
DA
at which this
occurs for a well in the center of a circle?

4. Neglecting skin, the pseudo-steady state equation for flow in circular reservoir is:


|
|

\
|
4
3
-
r
ln
kh 2
q
= P -
e
wf
w
r
P

The generalized flow equation at pseudo-steady state is


(

2
w A
wf
r C
4A
ln
2
1

kh 2
q
= P -

P

where C
A
is the shape factor. By comparing the two equations, estimate the shape factor
for a well in the centre of a circular reservoir.






























178
10. BOTTOM-HOLE PRESSURE TESTS IN
HORIZONTAL WELLS

The objectives of bottom-hole pressure tests in vertical and horizontal wells are similar.
Actually, a horizontal well can be viewed as a vertical well with infinite conductivity
vertical fracture or a highly stimulated vertical well. However, BHP tests in horizontal
wells are more difficult to analyze for the following reasons.

1. The horizontal wells are not perfectly horizontal as assumed by the analysis models.
The wells may be snake-like as shown in Fig. 10.1.














Fig. 10.1 Snake-Like Nature of some Horizontal Wells

The consequence of the snake-like nature of horizontal wells is that waves emanating
from different sections of the horizontal well at the same time may hit different
boundaries. This complicates pressure response.

2. The total drilled length may not all contribute to the producing length. Some drilled
length may intersect non-productive interval. The unknown non-producing length is
erroneously accounted for as skin factor.

3. There are many possible flow regimes which depending on the reservoir and well
conditions may or may not be properly discerned from pressure tests.
In this chapter, we shall present basic information about horizontal wells and how to
identify the flow regimes obtained during horizontal welltests. We shall also present flow
equations and how to analyze horizontal welltests.


Vertical
Depth
Horizontal Distance
179
10.1 Introduction
A horizontal well is a well drilled parallel to the reservoir bedding plane while a vertical
well is drilled perpendicular to the reservoir plane. This is shown in Fig 10.2. With
horizontal well, we can enhance reservoir contact by well and therefore hence enhance
well productivity. This implies that the drainage area of the horizontal well is bigger than
that of a vertical well. Joshi (1988) showed two methods of calculating the drainage area
of horizontal well based on knowledge of the drainage area of a vertical well. Joshis
procedures are presented.












Fig. 10.2: Vertical and Horizontal Wells.
Drainage Area of Horizontal Well
If the drainage area of a vertical well is:

A = r
ev
2
10.1
Where r
ev
is shown in Fig 10.3.








Fig. 10.3: Drainage Area of Vertical Well


The two method for the calculating the drainage area of horizontal well are as follows:
Horizontal Well
Vertical Well
r
ev

180

Method 1: This assumes that each end of the horizontal well drains a semi-circle drainage
area while the length drains a rectangular drainage area as shown in Fig 10.4.



r
ev




Fig. 10.4: Rectangular/Semi-Circle Drainage Area


Assuming that h r
ev,

H
WDA
= ( ) { }
r
ev
2
+ L 2r
ev
10.2

Where

H
wDA
= Drainage area of the horizontal well
L = Productive length of the horizontal well

Method 2: We assume an elliptical drainage area in the horizontal plane, with each end
of a well as a foci of drainage ellipse as shown in Fig 10.5.





Fig. 10.5: Elliptical Drainage Area

In this case, the drainage area is,

H
WDA
= ab 10.3

where
L
L
181
a = half major axis of ellipse =
L
2
+ r
ev

b = half minor axis = r
ev


The calculated drainage areas (using different methods) are not the same. Therefore, the
average of the two can be used as the effective drainage area.
Applications of Horizontal Wells
(a) Some naturally fractured reservoirs
(b) Reservoir with gas/water coning problems
(c) Thin reservoirs
(d) Reservoirs with high vertical permeability
(e) In EOR projects with injectivity problems
(f) In fields (e.g. offshore) requiring limited wells due to cost or environmental
problems.
Advantages of Horizontal Wells
(a) Horizontal well productivities are 2 to 5 times greater than that of unstimulated
vertical well. Actually, the performance of horizontal wells depends on the
effective length of the horizontal section in the formation. Higher productivity may
result to early payout.
(b) Horizontal well may intersect several fractures or compactments and help drain
them effectively
(c) Reduce coning tendencies
(d) As injectors can improve sweep efficiency in EOR projects
Disadvantages of Horizontal Wells
(a) Ineffective in thick (500ft to 600ft) low permeability reservoirs.
(b) Cannot easily drain different layers
(c) Technological limitations
(d) Cost more (1.4 to 2 times) than cost of drilling a vertical well.

Dimensionless Parameters used in Horizontal Wells
Dimensionless parameters are also used in horizontal wells. Figure 10.6 shows
dimensions and coordinates in horizontal wells.










182













Fig. 10.6: Dimensions in a Horizontal Wells

Table 10.1 shows the different dimensionless parameters in Darcy unit.

Table 10.1: Dimensionless Parameters used in Horizontal Wells
Dimensionless Parameters Equations

Dimensionless Pressure in terms of h and k
r
P
Dh
=
k h p
B
r
s o o

141.2q


Dimensionless Pressure in terms of L and k k
r z
P
DL
=
k h L p
B
r z
s o o

141.2q


Dimensionless Time in terms of L/2 t
D
=
0.0002637k
2)
2
r
t
t
c (L /


Dimensionless Time in terms of h and k
z
t
Dz
=
0.0002637k
2
z
o t
t
c h


Dimensionless Time in terms of y
e
and k
y

t
Dy
=
0.0002637k
2
y
o t e
t
c y


Dimensionless Time in terms of h and y
e

t
Dhy
=
0.0002637k
y
o t
t
c hy
e



Dimensionless Time in terms of and r
w

t
Drw
=
0.0002637
2
k k t
c r
r z
o t w


h
x
z
y
z
w

L/2
183

Derivative in terms of p
DL

( )
p
dp
d t
DL
DL
D
'
ln
=

Derivative in terms of p
Dh

( )
p
dp
d t
Dh
Dh
D
'
ln
=
Dimensionless x, y, z Coordinate x
D
= 2(x-x
w
)/L

y
D
= 2(y-y
w
)/L

z
D
= z/h
Dimensionless Wellbore Location

x
wD
= 2x
w
/x
e

y
wD
= 2y
w
/y
e

z
wD
= z
w
/h
Dimensionless x, y-direction boundary width x
eD
= 2x
e
/L
y
eDL
= 2y
e
/L
Dimensionless Wellbore Length
L
D
=
L
k
k
z
r
2h

Dimensionless Wellbore Radius r
wD
= 2r
w
/L

10.2. Flow Regimes during BHP Tests In Horizontal Wells
During bottom-hole pressure tests in a horizontal well, the followoing flow regimes could
be discerned: linear, radial, hemiradial, and pseudo-radial. Figure 10.7 is a box diagram
showing some of the flow regimes and the order in which they occurred.

Some parameters in Table 5.1 are defined as follows:
Z
w
= vertical distance measured from bottom of payzone to the well
X
w
, Y
w
, Z
w
= Well location co-ordinates
X
e,
Y
e
= reservoir boundaries in x and y directions
L = Length of horizontal well
k
r
= k
h
= permeability in horizontal place
= k k
x y

k
z
= k
v
= vertical permeability








184











Fig. 10.7: Typical Flow Regimes in a Horizontal Well


The pressure and pressure derivative for the case shown in Fig 10.7 is shown in Fig 10.8
taken from Lichtenberger (1994).




Fig. 10.8: Pressure and Derivative for a Typical Horizontal Well BHP Test

Details on the flow regimes are as follows:

Early Radial Flow: Figure 10.9 shows the early time radial flow period in a vertical
plane, which develops, when the well is put initially on production. The well acts as
though it is a vertical well turned sideways in a laterally infinite reservoir with thickness,
L. This flow period ends when the effect of the top or bottom boundary is felt or when
flow across the well tip affects pressure response. This flow regime may not develop
(Kuchuk , 1995) if the anisotropic ratio, k
H
/k
V
is large.




Transient State
Wellbore Storage
Radial Linear Pseudoradial Boundary
Effect
185












Fig. 10.9: Early Radial Flow in a Horizontal Well

Many authors (Ozkan et al, 1989; Goode and Thambynayagam, 1987; Odeh and Babu,
1990; Du and Stewart, 1992; Lichtenberger, 1994 and Kuchuk, 1995) published
equations for identifying the different flow regimes. Inferences from their publication
show that the early radial flow can be identified using the pressure derivative or semilog
plot if it is not marred by wellbore storage effects. The pressure derivative gives a zero
slope as shown in Fig 10.8 while a graph of pressure versus log of time yields a straight
line. The basis for the straight line is the equation:

P
i
- p
wf
=

+
|
|

\
|
A s
r c
k k
L k k
qB
w t
v y
y v
87 . 0 23 . 3
t
log

6 . 162
2

10.4

where s = skin due to damage/stimulation. If it is positive, it is denoted as S
m
,
mechanical damage due to drilling and completion.
Also, A is a constant given by Lichtenberger (1994) as

A = 2.303 log
k
k
y
v
4 +
k
k
v
y
4
|
\

|
|
10.5

The equation implies that a graph of P
wf
versus log t gives a straight line with slope

m
1
=
162 6 . qB
k k
v y

L
10.6

From this, the equivalent permeability in the vertical plane, k k
v y
, can be calculated.
186
k k
m L
v y
=
qB 162 6
1
.
10.7

The skin equation for the first radial flow period is

S = 1.151
( )
Pi
r
- P 1 hr
m
- log
k k
c
+ A + 3.23
1
y v
t w
2

|
\

|
|

10.8

Note:
1. In arealy isotropic reservoir, k
x
= k
y
= k
h

2. If effective reservoir permeability k k
v y
is known, the given equation can be used
in determining the effective producing length.
3. The early-time flow regime can be short and may be completely marred by
wellbore storage effect. Use of downhole shut-in tool is useful here. Time to end
of wellbore storage is given by Lichtenberger (1994) as:


/
) 240 4000 (
L k k
C s
t
v h
m
Eus
+
= 10.9
where
t
Eus
= time for end of wellbore storage effects, hours
s
m
= skin factor
C = wellbore storage constant, rb/psi
L = effective producing length of well, ft
= viscosity of oil, cp
k
h
= horizontal permeability =
y x
k k
k
v
,k
x
,k
y
= permeability in the vertical, x and y directions respectively.

Early radial flow ends when either of the following will occur:
a. Effect of bottom or top boundary is felt.
b. Flow across well tips affects pressure response

Mathematically, different authors gave time to the end of early radial flow, t
e1
, as follows:

Goode and Thambynayam (1987)

t
e1
=
v
w z
k
d
t
095 . 0 095 . 2
c r 190

10.10
where d
z
is the distance of the well to the closest boundary (top or bottom)

Odeh and Babu (1990) and Licthenberger (1994)

187
t
e1
= min
d C
k
C
k
z
2
t
v
t
y
1800
125
2

10.11
The first equation represents the time when the effect of the boundary (top or bottom)
will distort early radial flow while the second equation represents time when flow across
the tip of the well will distort early radial. Lichtenberger assumed areal isotropy in his
equations.

DU and Stewart (1992)
t
e1
= min
C d
k
C d
k
t z
2
v
t x
2
x
947
947

10.12

where d
x
is nearest distance from well point to the boundary normal to the well length
axis. Other parameters are as defined earlier.

Although the equations by different authors are not exactly the same, but they are similar
and therefore can be used as a guide.

Hemiradial Flow : When the wellbore is closer to any of the no-flow boundaries, hemi-
radial (or hemi-elliptical) flow may develop. This produces slope doubling on the
semilog plot and p
1
(pressure derivatives) will plateau at twice the radial flow value.

The flow equation during the hemi-radial flow period is given by Kuchuk (1995) and
Lichetenberger (1994) as

P
wf
= P
i
- 2 x
162 6
3 23
0 87
2
2
.
log .
. qB
k k
k k
c r
S
A
H v
H v
t w

L
t |
\

| +

10.13
where
A = log
(
(

|
|

\
|
w
z
r
d
v
H
k
k
+ 1
d
z
= distance to the nearest boundary (top or bottom)

Lichtenberger (1994) gave the time for the end of the hemiradial flow as

188
t
Ehrf
=
1800 d c
k
z
2
t
v

10.14
Implications of the Eq 10.13 are as follows:
a. A plot of p
wf
versus log t is a straight line with slope m

b. k k
mL
H v
= 2
qB 162 6 .
|
\

| 10.15

c. The skin equation is

S = 2.30
( )

(
(

|
|

\
|
|
|

\
|
w
z i
r
d
r
P
V
H
2
w t
v H
k
k
+ 1 log + 3.23 +
c
k k
log -
m
hr 1 P -

10.16

Intermediate Linear Flow: This flow regime may develop after the effects of upper and
lower boundaries are felt at the well. Figure 10.10 shows the streamlines during the
intermediate-time linear flow. This flow regime develops if the well length is sufficiently
long compared with reservoir thickness and there is no constant pressure boundary.










Fig. 10.10: Intermediate Linear Flow

The flow equation during the linear flow is given as follows:

P
i
- P
wf
= ( )
8128 . qB
Lh
t
c k
S S
t y
z

+
141.2 qB
L k k
y v
+ 10.17

where S
Z
is the pseudo-skin factor caused by partial penetration in the vertical direction
and is given by different authors as:

189
Odeh and Babu (1990)

S
Z
= ln
h
r
w
|
\

|
|
\

|

|
\

|
|
\

|
+ 0.25 ln
k
k
Z
h
y
v
w
ln sin .
180
1838 10.18

Lichtenberger (1994)

S
Z

(
(

|
|

\
|
+
h
d
sin 1
h
z w

y
v
k
k r
10.19

Kuchuk (1995)
S
Z
= ln
(
(

|
|

\
|
h
d
sin
k
k
+ 1
h
w
H
v w
r
10.20

Implications of the flow equations are as follows:
a. A graph of (P
i
-P
wf
) versus t is a straight line
b. Slope of line

m
2
=
8128 . qB
Lh c k
t y

10.21

Therefore

L
2
k
y
=
8128
2
2
. qB
hm c
t
|
\

10.22

c. p ( )
t o
y v
z
L k k
S S
=
+ =
141.2 qB
10.23

where p
t=0
is the pressure drop at time equals zero. The skin due to damage, S, can
therefore be calculated as S
Z
is known (calculated from Equation 10.18, 10.19 or 10.20)

Time to end of early linear flow is given by different authors as follows:

Goode and Thambynayagam (1987)

t
e2
=
20 8 . c L
k
t
2
x

10.24

Du and Stewart (1992)

190
t
e2
=
16 c L
k
t
2
x

10.25

Odeh and Babu (1990) and Lichtenberger (1994)

t
e2
=
160 c L
k
t
2
x

10.26

Odeh and Babu (1990) also gave the time to the start of early linear flow period as

t
S2
=
180 D c
k
z
2
t
v

10.27

where D
z
(= h-d
z
) is the maximum distance between the well and the z-boundaries (top or
bottom boundary)

Pseudoradial Flow: In sufficiently large reservoir, pseudo radial flow will develop
eventually as the dimensions of the drainage areain the horizontal plane becomes much
larger that the effective well length. Figure 10.11 is a schematic showing the streamlines
during the pseudoradial flow. This is similar to what happens in a horizontal with a
vertical fracture.









Fig 10.11: Pseudoradial Flow

The flow equation during the pseudoradial flow period is given as follows:

P
i
P
wf
= ( ) S S
k k L
qB
A
L c
t k
k k
qB
Z
v y t
x
y x
+ +

|
|

\
|

2 . 141
log
h
6 . 162
2
10.28

where A is a contant given by many authors as follows:
A = 2.023 (Goode and Thambynayagam, 1987)
A = 2.5267 (Kuchuk, 1995)
A = 1.76 (Odeh and Babu, 1990)
A = 1.83 (Lichtenberger, 1994
191

Implications of Eq 10.28 are as follows:
a. Graph of P
wf
versus log t is a straight line
b. Slope,

m
3
=
162 6 . qB
k h
x

k
y
10.29

Therefore
k k
qB
m h
x y
=
162 6
3
.


The skin equation is

S =
( ) 1151 1
3
2
.
log
L
h
k
k
Pi P hr
m
k
c L
A S
v
x
x
t
Z

`
)

Note:
a. Pseudo radial develops if L>>h (h
D
2.5)
b. If top or bottom boundary is maintained at constant pressure, no pseudo-radial flow
period will occur. Instead, there is steady state flow at late time.

Time for beginning of pseudoradial flow, t
s3
, is given by different authors as follows:

t
S3
=
1230 L c
k
2
t
x

(Goode and Thambynayagam

t
S3
=
1480 L c
k
2
t
x

(Odeh and Babu, Lichtenberger used 1500)

t
S3
=
2841 c L
k
t
2
x

(Du and Stewart)

Lichtenberger (1994) gave the time for the end of the pseudoradial flow as follows:

t
Eprf
= min
( )
1650
2000 4
2
2

C D
k
C L D
k
t x
H
t w y
H
/ +

10.30

where D
x
and D
y
are the lateral distances of the reservoir in the x and y directions
respectively.

Late Linear Flow Period: After, the pseudoradial flow, it is possible that a late-time
linear flow period develops. The flow equation for this phase is
192

P
i
- P
wf
= ( )
8128
2
. qB
x
t
k c
S
eh y t
x

|
\

|
|
+
141.2qB
L k k
+ S + S
y v
z
10.31
where
2x
e
= width of reservoir
S
x
= pseudo skin due to partial penetration in the x direction.

Implications of Eq 10.31 are obvious.

Skin in Horizontal Well
The skin factor will serve the same purpose in horizontal well as it does in vertical wells.
The dominant pseudoskins in horizontal wells are the pseudoskin due to damage and
pseudoskin due to convergence in the z-direction. The pseudoskin due to damage is
dominant because of more fluid losses resulting from larger area contacted by the well.

The pressure loss due to skin is defined with respect to the formation thickness in vertical
wells and well length in horizontal wells. This is shown in the following equations:

Pressure loss due to skin in vertical wells:

p
qB
kh
s vertical
S =
1412 .
10.32
Pressure loss due to skin in horizontal wells:

p
qB
kL
s Horizontal

S
=
1412 .
10.33

From Eqs 10.32 and 10.33, we infer that the pressure loss due to skin in horizontal well is
much smaller than the pressure loss due to skin in vertical wells because the horizontal
well length is usually longer than the formation thickness.

The small pressure loss due to skin in horizontal wells does not imply that skin has small
effect on horizontal well productivity because the drawdown in horizontal wells is also
small. The remedial factor, R, used in vertical wells should also be used in horizontal
wells to quantify the effect of skin.
Problem Set
A 2100ft long well is completed in a 100 ft thick formation with closed top and bottom
boundaries. The estimated average horizontal permeability from several vertical well test
is 1500md while the vertical permeability is 300md. The horizontal well has a diameter
of 8 in and is located 30ft from top of the sand. Other parameters are as follows:

= 20%, = 0.65cp, c
t
= 20 x 10
-6
psi
-1
and s = 5. Also, assume d
x
= L/2

193
(a) Assuming that the early radial flow will not be distorted by wellbore storage effects,
determine the time when the early radial flow will end. Also, determine when
wellbore storage effect will end if wellbore storage constant, c = 0.025 rb/psi.
(b) Calculate the time to start and end of the early linear flow period for the horizontal
well whose parameters have been given
(c) For given reservoir and well data, calculate time required to start a pseudo-radial
flow.
(d) Using the following additional data calculate the pressure change in a horizontal
well. S = 25, q = 4000 STB/D and Bo = 1.05 rb/STB
Tabulated Solution to Problem Set


Authors
End of early
radial flow
tel x10
-3

(hrs)
End of well-
bore storage
effect tews
x10
-5
(hrs)
Start of early
linear flow
ts2 x10
-3

(hrs)
End of
early
linear flow
te
2
(hrs)
Start of
Pseudo
Radial Flow
ts
3
(hrs)
Skin due to
partial pent.
In vert. dir.
Sz
P @ the start
of Pseudo
Radial flow
P (psi)
Goode et al 2.560 N/A N/A 0.159 9.402 N/A N/A
Odeh and Babu 14.040 N/A 7.644 1.223 11.313 3.996 12.11
Du and Stewart 7.387 N/A N/A 0.122 21.717 N/A N/A
Lichtenberger N/A 5.998 N/A N/A 11.466 2.15 x 10-4 10.833
N/A implies author did not give required equation

10.3 Detecting Flow Regimes Using Pressure Derivative
The pressure derivative is the best diagnostic tool for detecting flow regimes. This
follows from the characteristic slopes of the derivatives obtained for different flow
regimes on a log-log plot. Figure 10.12 shows the characteristic slope for the first radial,
hemiradial and pseudoradial flow regimes in a horizontal well. Note that the derivatives
have zero slopes at different levels. For linear flow, the derivative has a slope of 0.5.








Log







Log (time)

Fig 10.12: Prssure Derivatives for the Different Radial Flow Regimes
Pressure
Derivative
Early Radial
Hemiradial
pseudoradial
194


Figure 10.13 taken from Kucuk (1995) shows the pressure derivative for cases with well
and reservoir parameters shown in Table 10.2.



Fig 10.13 : Derivatives for Cases shown in Table 10.2

Table 10.2: Reservoir Parameters for Examples Shown in Fig 10.13
Example h, ft k
H
, md K
V
, md L
w
, ft z
w
, ft r
wD

1 100 100 10 500 20 0.00146
2 100 100 1 500 20 0.00389
3 100 100 5 500 5 0.00194
4 40 100 5 500 20 0.00197
5 200 200 1 500 20 0.00530

(
(

+
|
|

\
|
=
V
H
w
w
wD
k
k
L
r
r 1
2


Deductions from the derivatives are as follows:
(a) The first radial period can be seen in all cases.
(b) In Example 3, the well is close to the boundary (5ft) and therefore, hemiradial flow
occurred after a short duration early radial flow.
(c) In Example 4, a linear flow regime manifested because the well length is much
greater than formation thickness
(d) In all cases, the pseudoradial flow developed

Du and Stewart (1992) quantified the effect of parameters on the flow regimes. In their
work, they defined dimensionless parameters as follows:

P
DL
=
2

k k
p
z r
L
q
. 10.34
195
P
DL
1
=
dp
d ln t
DL
DZ
(Pressure Derivative) 10.35
t
DZ
=
k
z
t
c h
t
2

10.36

Values of P
DL
1
for different flow regimes are as follows:
P
DL
1
= 0.5 vertical radial flow
P
DL
1
= 1 vertical hemi-radial flow
P
DL
1
= L
D
pseudo radial flow

Du and Stewart (1992) concluded that parameters z
WD
(dimensionless well location in the
z-direction) and L
D
(dimensionless wellbore length) have the dominant effect on flow
regimes obtained in horizontal wells. These parameters are defined as follows:

z
wD
= z
w
/h and
H
v
D
k
k
h
L
L
2
=

Figure 10.14 shows effect of L
D
on flow regimes for infinite reservoir with no flow top
and bottom. Horizontal well is in the center of the formation (Z
WD
= 0.5).




Fig. 10.14: Effect of L
D
in Homogeneous Laterally Infinite Reservoir With Sealed Top and Bottom

Inferences from Fig. 10.14 are as follows:
(i) For L
D
3 the flow regimes are: Vertical radial flow (VRF: P
DL
1
= 0.5) + transition
+ early linear flow opposite the completed section (ELF: Gradient = ) + transient
reservoir pseudo radial flow (PRF: P
DL
1
= L
D
).

196
(ii) For L
D
< 3, the flow regimes are: VRF + vertical spherical flow (VSF)
Gradient = - + transition + PRF. The smaller L
D
, the longer the duration of VSF.

(iii) The smaller the L
D
, the shorter the duration of VRF and longer the length of PRF.

(iv) For L
D
0.1, no VRF at all. For a 100ft thick formation and K
z
/ K
r
= 0.2, this
implies a minimum well length of 45 ft.
Figure 10.15 shows effect of Z
WD
. From Fig 10.15, we infer that when Z
WD
0.1 (well
close to one of the boundaries), there is a hemi-radial flow (HRF), between VRF and ELF
with a transition in between.

Figure 10.16 shows the effect of Z
WD
in a situation with gas cap. The dimensionless well
length, L
D
, is large.




Fig. 10.15: Effect of Z
wD
in Homogeneous Laterally Infinite Reservoir With Sealed Top
and Bottom




197
Fig. 10.16: Effect of Z
wD
in Homogeneous Laterally Infinite Reservoir with Bottom
Constant Pressure Boundary and Top No Flow Boundary

Inferences from the graph are as follows:
(i) When Z
WD
< 0.5 (well closer to the bottom no-flow boundary) the flow regimes are
as follow: VRF + transition + HRF + transition + constant pressure effect (rollover).

(ii) When Z
WD
0.5 (well nearer the constant pressure boundary) the flow regimes are
VRF + transition + rollover (constant pressure effect).

(iii)The nearer the well to the constant pressure boundary, the stronger the constant
pressure boundary effect.
The effect of the boundaries is similar to what obtains in a vertical well. A no - flow
boundary causes stabilization at higher level with respect to the infinite reservoir case.
Figure 10.17 shows the effect of lateral boundaries, which are parallel to the direction of
well length.




Fig. 10.17: Effect of Lateral Boundaries in Reservoir with Sealed Top and Bottom

10.4 Field Cases
In this section, we shall discuss some field examples of bottom-hole pressure tests. The
objective is to see what may actually obtain in real life. Table 10.3 shows basic
parameters or the different tests.

Table 10.3: Parameters for Different Field Cases
Parameters Values for Different Cases
1 2 3 4
198
Formation thickness, ft 73 95 19 123
Well length, ft 1984 1387 232 1330
Well location, ft 7 7.4 8 16.1
Oil rate STB/D 3948 4144 685 4951
Porosity, % 29 29 24 29
Oil Viscosity, cp 1.97 1.97 0.34 2.23
Formation volume factor, rb/STB 1.128 1.120 1.539 1.298
Wellbore radus, ft 0.4 0.4 0.4 0.3
Calculated Permeability, md 15363 16320 1070 3820
Calculated Skin 35 23 0 -1.4
Drawdown, psi 15 10 9 34

Discussions on the field cases follow:

Case 1: Figure 10.18 shows the pressure and pressure derivative for this case.




Fig. 10.18: Pressure and Pressure Derivative for Case 1
The pressure derivative shows wellbore storage, early vertical radial flow, early linear
flow, hemiradial flow and rollover due to a constant pressure boundary. The
hemiradial flow was inferred because of the nearness of the well to the bottom boundary
and the fact that p (VRF) 2p (HRF). The constant pressure effect was caused by
gas-cap. Figure 10.18 shows that mathematical models give us good insight unto pressure
and flow regime obtained during actual BHP tests.

199
Case 2: Figure 10.19 shows pressure data for Case 2. Case 2 and Case 1 are from the
same reservoir in western Niger Delta. The pressure and pressure derivative for the two
cases exhibit similar characteristics. The exception is the scatter in Case 2 data during
the wellbore storage phase. The scatter was due to gauge shift.

Case 3: Data for this case were obtained from the eastern Niger Delta reservoir. Figure
10.20 shows the pressure and derivative for this case. The identified flow regimes are as
follows: wellbore storage phase + early vertical radial flow + vertical spherical flow +
pseudo radial flow + rollover due to lateral constant pressure boundary.







Fig. 10.19: Pressure and Pressure Derivative for Case 2









200




Fig. 10.20: Pressure and Pressure Derivative for Case 3

The vertical spherical flow regime resulted because the well length is small and therefore
L
D
< 3 (actually L
D
2). This is in agreement with the finding of Du and Stewart (1992).

Case 4: Data for this case were obtained from the first horizontal well in eastern Niger
Delta reservoir. Figure 10.21 shows the pressure and derivative for this case.



Fig. 10.21: Pressure and Pressure Derivative for Case 4
We believe that the distortions in pressure and derivative were caused by the snakelike
nature of the horizontal part of the well. The distortions made it difficult to clearly
discern the flow regimes.
201

In field situations, there could be problem resulting because the gauge may not get to the
horizontal part.

10.4 Analysis Procedure
Analysis of bottom-hole pressure test in horizontal wells, requires the following
a Identifying boundaries and main features such as faults, fractures, etc. from flow
regimes analysis.
b Estimating well/reservoir parameters and refining the model that is obtained from
flow regime analysis.

The graphical type curve procedure is practically impossible for the analysis of horizontal
welltest data because of the many unknowns (k
H
, k
V
, s, C, L
w
, h, d
z,
) even in the case of
a single-layer reservoir. Thus, along with the flow regime analysis, non-linear least-
square techniques are usually used to estimate reservoir parameters. In applying these
methods, one seeks not merely a model that fits a given set of output data (pressure,
flowrate, and/or their derivatives) but also knowledge of what features in that model are
satisfied by the data.

A flowchart showing recommended procedure for test analysis in horizontal well is
shown in Fig 10.22.






















Fig. 10.22: Flow chart of Procedure for Test Analysis in Horizontal Well.


Regimes
Clear
Diagnose
Flow Regimes
No
Any
Early radial?
Test cannot
be Analyzed
Analyze Tests
using the 3
methods
Yes
Analyze test with
Regression
Time
constraints ?
No Results not
accepted
Yes
Simulate profile
and compare
Accept results
with best
confidence