Você está na página 1de 140

UNIVERSITY OF OKLAHOMA

GRADUATE COLLEGE

FRACTURE PRESSURE ANALYSIS OF DIAGNOSTIC PUMP- IN TESTS

OF RED FORK SANDS IN WESTERN OKLAHOMA

A THESIS

SUBMITTED TO THE GRADUATE FACULTY

in partial fulfillment of the requirements for the

degree of

MASTER OF SCIENCE

By

SANDEEP JANWADKAR

Norman, Oklahoma

2004
FRACTURE PRESSURE ANALYSIS OF DIAGNOSTIC PUMP- IN TESTS
OF RED FORK SANDS IN WESTERN OKLAHOMA

A THESIS APPROVED FOR THE


MEWBOURNE SCHOOL OF PETROLEUM AND GEOLOGICAL
ENGINEERING

BY

Chair:________________________________

Dr. SUBHASH SHAH

Member:______________________________

Dr. CHANDRA RAI

Member:______________________________

Dr. SAMUEL OSISANYA

ii
©Copyright by SANDEEP JANWADKAR 2004
All Rights Reserved

iii
ACKNOWLEDGEMENTS

To achieve any kind of success only personal capabilities are not sufficient.

There are several factors that contribute towards achieving the end goals and

objectives. Some of these factors are inspiration, guidance, moral support,

motivation, initiative and encouragement. While working towards completion of this

master’s thesis several individuals have made valuable and significant contributions.

I would like to express my sincere gratitude to all of those individuals. It would not

be possible to name each individual but I definitely would like to mention some of

them. I apologize to all those wonderful individuals whom I acknowledge but have

missed out.

Dr. Subhash Shah, my graduate advisor and Chair of my thesis committee.

Right from the beginning, throughout my work and till the end it was Dr. Shah’s

guidance, untiring support, and constant encouragement without which this thesis

would never have been possible. Thanks a lot Dr. Shah.

Dr. Samuel Osisanya, my thesis committee member who always provided me

with inspiration and built a strong desire in me to work hard to succeed. Thank you

Dr. Sam.

Dr. Chandra Rai, my thesis committee member whose support and help have

proved to be very valuable. Thank you Dr. Rai.

Mr. David Roddy, Operations Manager, Crawley Petroleum, who inspired me

throughout my thesis work. Thank you Dave.

iv
I would like to take this opportunity to express my gratitude to all the staff and

students of Well Construction Technology Center of The University of Oklahoma for

all their help, support and good wishes.

Several organizations have made significant contributions and I would like to

thank them:

Crawley Petroleum Corporation for providing the data required for this thesis

and also for granting the permission to publish their data, Pinnacle Technologies Inc

for the use of their software FracPro PT 10.2 which was used to run the simulations.

From the bottom my heart I would like to say a BIG THANK YOU to all the

people and organizations that made this difficult task worthwhile. I hope this work

will make a contribution in expanding the existing knowledge of the Petroleum

industry and that means so much to me.

Last but not the least I would like to mention my parents and my family who

have always supported me during all the ups and downs of life, their contribution in

achieving my objectives and goals are the most significant.

Sandeep Janwadkar

v
TABLE OF CONTENTS

LIST OF TABLES............................................................................................ viii

LIST OF FIGURES………………………………………………………..… x

ABSTRACT……………………………………………………...................... xiv

Chapter 1 INTRODUCTION………………………………………….... 1
1.1 The Origin and Growth of Hydraulic Fracturing
Technology…………………………………………….. 1
1.2 Hydraulic Fracturing: Concept and Application……….. 1
1.3 Characteristics of the Red Fork Formation…………….. 3
1.4 Evolution of the Red Fork Fracturing Technology…….. 4
1.5 Research Goals and Outline of Work………………..… 5

Chapter 2 CONCEPTS OF HYDRAULIC FRACTURING……………. 7


2.1 Closure Pressure…………………………………...…… 7
2.2 Fluid Loss……………………………………………… 11
2.3 Mechanics of Hydraulic Fracturing……………………. 13
2.4 Hydraulic Fracture Models…………………………..… 17

Chapter 3 MINI-FRACS AND DIAGNOSTIC PUMP-IN TESTS…… 25


3.1 Mini-Fracs……………………………………………… 25
3.2 Diagnostic Pump-In Tests……………………………… 26
3.3 Comparison of Mini-Fracs with Diagnostic
Pump-In Tests………………………………………….. 27

Chapter 4 FRACTURE PRESSURE ANALYSIS AND G-FUNCTION 29


4.1 Generalized Expression for Fluid Loss………………… 29
4.2 Cumulative Fluid Loss Volume………………………... 30
4.3 Newtonian Filtrates ……………………..…………...… 32
4.4 Non-Newtonian Filtrates …………………………….… 34
4.5 Typical Pressure Profile of Fracture Treatment………... 35
4.6 Basic Definitions……………………………………..… 35
4.7 Equations for Pumping Stage………………………….. 37
4.8 Log-Log Plot …………………………………...........… 39
4.9 Pressure Derivative Plot……………………………...… 43
4.10 Near WellBore Pressure Losses………………………... 45
4.11 Rate Step-Down Test………………………………...… 48
4.12 Dimensionless Fluid Loss Function ………………...… 49
4.13 Interpretation of G-function plot……………………..… 52
4.14 Summary of G-Function...…….……………………..… 58

Chapter 5 BREAKDOWN TESTS AND SIMULATIONS……….……. 60


5.1 Test Methodology and Data Analysis………………….. 60

vi
5.2 Simulation Model Options...…………………………… 62
5.3 Log Analysis and Reservoir Layers….………………… 65
5.4 Simulator Input.………………………………………... 69
5.5 Breakdown Test Analysis……………………………… 69

Chapter 6 SIMULATION RESULTS AND INFERENCES...……….… 73


6.1 Interpretation of Simulation Plots…………………….... 73
6.2 Inferences from Simulations…………………………... 76
6.3 Comparison of PKN Model Results with GDK Model... 80

Chapter 7 WELL PRODUCTIVITY CORRELATIONS…....……….… 82


7.1 Well Productivity Factors…………………………….... 82
7.2 Statistical Analysis and Correlations (PKN Model).…... 84
7.3 Statistical Analysis and Correlations (GDK Model)…... 89

Chapter 8 CONCLUSIONS AND RECOMMENDATIONS…………... 94


8.1 Summary……………...………………………………... 94
8.2 Conclusions…………...………………………………... 94
8.3 Recommendations……………………………………… 95

Nomenclature………………………………………………………………… 96

References…………………………………………………………………..... 101

Appendix A WELL AND CASING PERFORATION DATA……………. 106

Appendix B WELL PRODUCTION DATA………………………………. 110

Appendix C LITHOLOGY OF RESERVOIR LAYERS…………………. 113

Appendix D SIMULATION PLOTS……...………………………………. 115

vii
LIST OF TABLES

Table 4.1 Analytical equations for g (∆t D , α , θ ) and g 0 (α , θ ) …………….. 33

Table 4.2 Fracture Net Pressure – Fracture Geometry Relations…………... 38

Table 4.3 Fracture Net Pressure – Time Relations…………………………. 38

Table 4.4 Fracture width relations………………………………………….. 39

Table 4.5 Interpretation of log-log plot slopes……………………………… 40

Table 4.6 Values of α1 (Upper bound of area exponent)…………………… 50

Table 4.7 Values of CL (Leakoff coefficient)………………………………. 52

Table 5.1 Simulator Model Options………………………………………… 62

Table 5.2 Shale Index of Well-4 Reservoir Layers ………………………… 67

Table 5.3 Values of t1/t2 for Seven Wells………...………………………… 68

Table 6.1 Data from Simulation Results (PKN Model)…………………..… 75

Table 6.2 Data from Simulation Results (GDK Model)…….................…… 76

Table 6.3 Average rate of pressure decline ………….……………………... 77

Table 7.1 Height of Producing Interval for Each Well……………………... 83

Table 7.2 Well Production Data……………………...……………………... 84

Table A1 Well Configuration Data for Well – 1…………………………… 106

Table A2 Well Configuration Data for Well – 2…………………………… 106

Table A3 Well Configuration Data for Well – 3…………………………… 107

Table A4 Well Configuration Data for Well – 4…………………………… 107

Table A5 Well Configuration Data for Well – 5………………………….... 108

Table A6 Well Configuration Data for Well – 6…………………………… 108

viii
Table A7 Well Configuration Data for Well – 7…………………………… 109

Table B1 Production data for Well – 1……………………………………... 110

Table B2 Production data for Well – 2……………………………………... 110

Table B3 Production data for Well – 3……………………………………... 110

Table B4 Production data for Well – 4……………………………………... 111

Table B5 Production data for Well – 5……………………………………... 111

Table B6 Production data for Well – 6……………………………………... 112

Table B7 Production data for Well – 7……………………………………... 112

ix
LIST OF FIGURES

Figure 2.1 Plot of Fracture Pressure versus Fracture Width................... 7

Figure 2.2(a) Step Rate Test (Injection rate Increments)……………….… 8

Figure 2.2(b) Step Rate Test (Estimating Closure and Extension Pressure) 8

Figure 2.3 Shut-In Decline Test……………………………………..… 9

Figure 2.4(a) Influence of Rate on Flowback Test……………………..… 10

Figure 2.4(b) Estimating closure pressure from Flowback Test………..… 10

Figure 2.5 Regions of fluid loss……………………………………….. 11

Figure 2.6 Cross Section of the Bore Hole…………………………….. 14

Figure 2.7 Pressure Profile v/s Time during a Fracture Treatment……. 14

Figure 2.8 PKN Fracture Model……………………………………….. 19

Figure 2.9 GDK Fracture Model………………………………………. 22

Figure 4.1 Pressure Profile of a Fracture Treatment…………………... 35

Figure 4.2 Log-Log Plot ……………………………………………….. 40

Figure 4.3(a) T-Shaped Fracture………………………………………….. 42

Figure 4.3(b) Height growth through Pinch Point.……………………….. 42

Figure 4.3(c) Fissure Dilation…………………………………………….. 42

Figure 4.4 Estimation of closure pressure……………………………... 43

Figure 4.5 Magnification of fracture pressure events………………….. 44

Figure 4.6 Fracture Reorientation……………………………………... 46

Figure 4.7 Perforation Misalignment………………………………….. 47

Figure 4.8 Rate Step-Down Test………………………………………. 49

x
Figure 4.9 G-Function versus Pressure and Pressure Derivative……… 51

Figure 4.10 G-Plot - Normal Leakoff…………………………………… 53

Figure 4.11 G-Plot - Pressure Dependent Leakoff -Fissure Opening…... 55

Figure 4.12 G-Plot – Fracture Tip Extension…………………………… 56

Figure 4.13 G-Plot – Fracture Height Recession……………………….. 57

Figure 5.1 Typical Well Casing Diagram……………………………... 60

Figure 5.2 Log for Well-4 (gamma ray and neutron porosity)………... 66

Figure 5.3 Reservoir Layers built from Logs for simulations (Well-4).. 68

Figure 6.1 ISIP Plot for Well – 1………………………………………. 74

Figure 6.2 Square Root Plot for Well – 1……………………………… 74

Figure 6.3 G-Function Plot for Well – 1 (PKN Model).................……. 74

Figure 6.4 G-Function Plot for Well – 1 (GDK Model)..…………..…. 75

Figure 7.1 County Map of Oklahoma State ..…………………………. 82

Figure 7.2 Plot of 30-day Cumulative Production versus


Height of Producing Interval………………………………. 86

Figure 7.3 Plot of 60-day Cumulative Production versus


Height of Producing Interval………………………………. 86

Figure 7.4 Plot of 30-day Cumulative Production versus


G-Function at Closure (PKN Model)………………………. 87

Figure 7.5 Plot of 60-day Cumulative Production versus


G-Function at Closure (PKN Model)………………………. 87

Figure 7.6 Plot of 30-day Cumulative Production versus


(H/G-Function at Closure -PKN Model)…………………… 88

Figure 7.7 Plot of 60-day Cumulative Production versus


(H/G-Function at Closure - PKN Model)……………….…. 88

Figure 7.8 Plot of 30-day Cumulative Production versus


G-Function at Closure (GDK Model)…………………….... 90

xi
Figure 7.9 Plot of 60-day Cumulative Production versus
G-Function at Closure (GDK Model)…...…………………. 91

Figure 7.10 Plot of 30-day Cumulative Production versus


(H/G-Function at Closure -GDK Model)………………...… 91

Figure 7.11 Plot of 60-day Cumulative Production versus


(H/G-Function at Closure - GDK Model)……………….…. 92

Figure C1 Layers of Well – 1…………………………………….….… 113

Figure C2 Layers of Well – 2 (Zone 1)…………………………….….. 113

Figure C3 Layers of Well – 2 (Zone 2)………………………………... 113

Figure C4 Layers of Well – 2 (Zone 3)…………………………….….. 113

Figure C5 Layers of Well – 3………………………………………….. 114

Figure C6 Layers of Well – 5……………………………………….…. 114

Figure C7 Layers of Well – 6………………………………………….. 114

Figure C8 Layers of Well – 7………………………………………….. 114

Figure D1 ISIP Plot for Well – 2 (Zone 1)…………………….………. 115

Figure D2 ISIP Plot for Well – 2 (Zone 2)………………………….…. 115

Figure D3 ISIP Plot for Well – 2 (Zone 3)………………………….…. 116

Figure D4 ISIP Plot for Well – 3…………………………………….… 116

Figure D5 ISIP Plot for Well – 4………………………………….…… 116

Figure D6 ISIP Plot for Well – 5…………………………………….… 117

Figure D7 ISIP Plot for Well – 6…………………………………….… 117

Figure D8 ISIP Plot for Well – 7………………………………….…… 117

Figure D9 Square Root Plot for Well – 2 (Zone 1)…………………..... 118

Figure D10 Square Root Plot for Well – 2 (Zone 2)………………...….. 118

xii
Figure D11 Square Root Plot for Well – 2 (Zone 3)……………………. 118

Figure D12 Square Root Plot for Well – 3…………………………….... 119

Figure D13 Square Root Plot for Well – 4…………………………...…. 119

Figure D14 Square Root Plot for Well – 5…………………………...…. 119

Figure D15 Square Root Plot for Well – 6…………………………….... 120

Figure D16 Square Root Plot for Well – 7…………………………….... 120

Figure D17 G-Function Plot (PKN Model) for Well – 2 (Zone 1)……... 120

Figure D18 G-Function Plot (PKN Model) for Well – 2 (Zone 2)……... 121

Figure D19 G-Function Plot (PKN Model) for Well – 2 (Zone 3)……... 121

Figure D20 G-Function Plot (PKN Model) for Well – 3……………….. 121

Figure D21 G-Function Plot (PKN Model) for Well – 4.………………. 122

Figure D22 G-Function Plot (PKN Model) for Well – 5……………….. 122

Figure D23 G-Function Plot (PKN Model) for Well – 6……………….. 122

Figure D24 G-Function Plot (PKN Model) for Well – 7……………….. 123

Figure D25 G-Function Plot (GDK Model) for Well – 2 (Zone 1)……... 123

Figure D26 G-Function Plot (GDK Model) for Well – 2 (Zone 2)……... 123

Figure D27 G-Function Plot (GDK Model) for Well – 2 (Zone 3)……... 124

Figure D28 G-Function Plot (GDK Model) for Well – 3……...……….. 124

Figure D29 G-Function Plot (GDK Model) for Well – 4.………...……. 124

Figure D30 G-Function Plot (GDK Model) for Well – 5……………..... 125

Figure D31 G-Function Plot (GDK Model) for Well – 6…………...….. 125

Figure D32 G-Function Plot (GDK Model) for Well – 7………………. 125

xiii
ABSTRACT

Estimation of a well’s productivity prior to a fracture stimulation job is the

key to a successful economic analysis of a fracture treatment. Methods to correlate

various factors with the production response are available. Case studies have proven

that there exists a relationship between productivity of a well and parameters such as

pump-in derived permeability, leakoff coefficient, and net pressure development

during hydraulic fracture, etc.

The Red Fork formation in western Oklahoma is a tight sandstone gas

reservoir. Due to low permeability these wells typically do not produce economically

without fracture stimulation. Performing build up tests in order to determine the

economics of fracture stimulation is time consuming. The decision of selecting

candidate wells for fracturing is difficult and at times very tricky one. The existing

procedures for selecting wells as candidates for fracture stimulation are not reliable

which sometimes do not result in an increase in the productivity of the wells after the

fracture job.

The objective of this study is to obtain a correlation between post-fracture

productivity of the Red Fork sands and parameters obtained from the pump-in

diagnostic tests. In developing this correlation this study considers the G-function at

closure obtained from the pump-in diagnostic test.

Utilizing production data from seven wells in the Hammon field of western

Oklahoma and data obtained from diagnostic pump-in tests performed on these wells

were utilized to perform simulations on commercially available stimulation design

xiv
software, to develop and establish the correlation of the G-function at closure with the

post fracture well productivity.

This correlation can be implemented in the future wells to predict the

productivity prior to fracture stimulation. With the help of this correlation the

economics of future fracturing jobs in the Red Fork formation (western Oklahoma)

can be evaluated. If the well productivity cannot justify the cost of expensive fracture

stimulation jobs then the decision of fracture stimulation needs further careful review.

xv
Chapter 1

Introduction

1.6 The Origin and Growth of Hydraulic Fracturing Technology

Hydraulic fracturing technology was developed in the late 1940s. The first

hydraulic fracture job was performed in 1947 on the well “Klepper No. 1” in the

Hugoton gas field located in Grant County of western Kansas1. Since then this

technology has continuously been improved, developed and applied to the oil and gas

industry. Statistics indicate that approximately 50 percent of natural gas wells and 30

percent of oil wells employ hydraulic fracturing to improve hydrocarbon recovery2.

In a time span of 15 years from 1984 to 1999 more than 100,000 wells have

been hydraulically fractured in the U.S., which has been responsible for an increment

of more than 7 billion barrels of oil and 600 trillion cubic feet of natural gas2. The

above facts and figures clearly demonstrate the significance of hydraulic fracturing in

the oil and gas industry.

1.2 Hydraulic Fracturing: Concept and Application

The recovery of hydrocarbons from the subsurface formations involves the

flow of oil and gas from the reservoir into the wellbore and then to the surface. This

phenomenon is similar to a person walking through a dense forest without any roads

or rivers to follow and trying to locate a nearby town. In the event that a road existed

through the forest that connects the towns then the person would have to travel a

small portion of his journey through the forest till he reaches the road and then reach

1
the town in a much faster and easier way. The concept of hydraulic fracturing is to

induce a crack in the formation to facilitate the flow of oil and gas through the

formation. The fracture in the formation acts like a road through the forest which

speeds up the flow of fluids through the reservoir.

The process of hydraulic fracturing consists of pumping a fluid into the

wellbore in order to break down the formation and to create a fracture. A proppant is

added to the fluid in order to keep the fracture propped open after the pumping of

fluid has stopped. The fluid used has high enough viscosity to suspend the proppant

during pumping and after pumping has ceased. Fracturing fluids are designed to

breakdown after the fracturing job is completed so that they can easily be recovered.

Hydraulic fracturing has several applications; some of the major ones are

described below1:

1. Fracturing to Overcome Wellbore Damage

Fluids and solids in the drilling mud invade the formation and reduce its permeability.

Limited penetration hydraulic fracturing is performed to overcome the near wellbore

damage.

2. Deep Penetrating Fractures

Deep fractures help in recovering large volumes of hydrocarbons at an accelerated

rate. They are meant to increase the drainage areas in formations with low

permeability.

3. Improve Secondary Recovery Operations

2
Hydraulic fracturing is performed in the injecting well to increase capacity of a well

to accept fluid in water injection. Additionally fracturing the producing well

increases the overall efficiency of waterflooding.

4. Disposal Wells

With the help of fracturing, large volumes of fluids and industrial waste can easily be

disposed off.

1.3 Characteristics of the Red Fork Formation

This research study is focused on the Red Fork formation. The total overall

thickness of the Red Fork formation is approximately 1500 feet and is located in the

Anadarko basin of western Oklahoma. Consisting of a sequence of sandstones and

shales the Red Fork formation is of Middle Pennsylvanian Age (Des Moinesian)3.

Red Fork formation consists of fine sand deposited in a deltaic complex3. It is

primarily a sandstone formation with major component of quartz (70%) and a

moderate quantity of feldspar (15%); other clay minerals (kaolinite and illite) and iron

minerals (chlorite and siderite) exist in the range4 of 3 to 6%.

The Red Fork formation has four main producing zones: Cherokee and the

Lower, Middle and Upper Red Fork. Porosity of the Red Fork formation ranges4

from 6 to 18%. The Young’s modulus ranges from 4.9 to 8.5 x 106 psi and the acid

solubility varies5 from 4 to 15%. The amounts of primary and secondary porosity

determine the permeability of the Red Fork formation, which usually exists in the

range3 from less than 1 µd to 0.5 md. Generally speaking the Red Fork formation is

3
classified as tight gas bearing sandstone with low permeability and low to moderate

porosity.

1.4 Evolution of the Red Fork Fracturing Technology

The early fracture treatments on the Red Fork formation were performed

during the late 70’s. The first treatments were performed using sintered bauxite as a

proppant (less than 2 ppg) with borate cross-linked hydroxypropylguar (HPG) fluids.

The trend was to increase the fluid and the proppant volumes4. The production

results were not as expected. One of the possible reasons for low productivity after

fracture treatments was due to the settling phenomenon of the high-density proppant

used in conjunction with the shear sensitive cross linked fluids. Due to the higher

cost of bauxite as a proppant and settling of the proppant, subsequent fracture

treatment designs used sand with non-cross linked gels3. Concerns were raised for

regaining the permeability, surfactants were used to reduce surface tensions and

increase the treatment recoveries6. A 2% KCl water solution was found to be

compatible with the formation and was considered as a fluid. The results obtained

with linear gel and sand showed substantial improvement due to which the fluid and

proppant volumes were reduced3.

The use of CO2 as a fracturing fluid started emerging. Although the density of

CO2 is similar to water, CO2 has a lower surface tension due to which a faster clean up

can be achieved. Early treatments with CO2 faced several problems due to high

tubing friction pressures and unreliable equipment coupled with the lack of

knowledge of the physical properties of foams and liquid CO2. A new procedure of

4
flowing back the fluid as soon as possible was adopted. This helped the prevention of

proppant settling as the fracture closed and trapped the proppant before substantial

settling3.

Average productive interval in Red Fork formation is 20 to 35 feet. High

viscosity fluids resulted in excessive fracture heights beyond the pay zone, as fracture

height is directly proportional to the fluid viscosity. The strategy adopted to

overcome this problem was to reduce gel concentration and lower the injection rates;

due to this the effective treatments grew smaller in size and proppant concentrations

used went up to 4 ppg. The next trend in the industry was to use pre-cured resin-

coated proppant and fluids using delayed cross linkers. This resulted in economically

priced intermediate strength proppants and fluids with better transport characteristics.

The treatments resulted in longer propped fractures and increased the ability of the

fluids to transport proppant.

The current trend is to use delayed cross-linked HPG fluids along with

intermediate-strength proppant. Higher injection rates can be obtained because cross

linkers are used with lower gel concentration. Fracturing fluids of low viscosity

result in less fracture height and longer fracture lengths as fracture length is inversely

proportional to the fracturing fluid viscosity3. Further improvements in the fracture

treatments were obtained with the help of limited-entry perforating and calibration

treatments.

1.5 Research Goals and Outline of Work

The step-by-step scope of work is outlined below:

5
• To review the history of fracture technology

• To review the characteristics of the Red Fork formation in Western Oklahoma

and the evolution of the fracture technology as applicable to the Red Fork

formation

• To summarize the theoretical concepts of hydraulic fracturing

• To present a critical review of fracture pressure analysis

• To compare the procedures of mini-fracs versus diagnostic pump-in tests

• To collect production data from seven producing wells in the Hammon field

after the fracture stimulation jobs have been carried out

• To collect data from diagnostic pump-in tests (breakdown tests) for the above

wells

• To enter data from diagnostic pump-in tests (breakdown tests) to FracProPT

version 10.2 and run breakdown simulations

• To prepare Instantaneous Shut In Pressure Plot (ISIP), square-root plot and G-

function plot (The G-function is discussed in detail in Chapter 4)

• To analyze simulation results

• To develop a correlation to predict post-frac productivity

6
Chapter 2

Concepts of Hydraulic Fracturing

2.3 Closure Pressure

The width of the fracture is directly proportional to pressure in the fracture as

depicted7 in Fig. 2.1. As the fracture pressure increases the fracture width increases

along the solid line in Fig. 2.1. The plot is a straight line except for the initial portion,

which is non-linear. Figure 2.1 is valid for all fracture models (PKN, GDK etc).
Fracture
Width

Non-Linear Portion

σ min
Pc

Fracture Pressure

Figure 2.1- Plot of Fracture Pressure versus Fracture Width7, fracture closure
pressure corresponds to the pressure when width reduces to zero

As the fracture pressure is reduced it results in a reduction of the fracture

width. The reversal of this plot (reduction of pressure) follows a straight line7. The

pressure at which the fracture width reduces to zero is the fracture closure pressure

7
Pc. This plot is a straight line and the non-linear portion is replaced by the dotted line

as illustrated in Fig. 2.1. The measurement of changes in fracture width with respect

to pressure is not practical in the field and therefore the estimation of fracture closure

pressure is accomplished using several other techniques as discussed below.

The fluid used in the step rate test depends on the permeability of the

reservoir. In low permeability reservoirs the step rate test is performed with the

completion fluid like treated water, whereas for high permeability reservoirs polymer

fluids are used to control fluid loss8. The closure pressure can be estimated by

performing a step rate test9. The test is conducted by pumping fluids at different

(increasing or decreasing) rates and recording the corresponding pressures.


Bottomhole pressure at
Injection Rate

step end

Pc
Extension Pressure

0 0 Injection Rate
Elapsed Time

Figure 2.2a- Step Rate Test (Injection rate Figure 2.2b- Step Rate Test
increased in fixed time intervals)9 (Estimating closure pressure
and extension pressure)9

A plot of the bottomhole pressure versus the injection rate provides an

estimate of the closure pressure Pc as well as the fracture extension pressure as shown

8
in Figs. 2.2a and 2.2b. The fracture extension pressure as obtained can be considered

as the upper bound of the closure pressure.

The shut-in decline test is performed by initiating a fracture, shutting-in the

well and observing the pressure decline (Fig. 2.3). The bottomhole pressure during

the shut-in decline period is inversely proportional to the square root of time elapsed,

which is discussed in detail in Chapter 4. A change in slope on the pressure decline

curve is an indicator of closure pressure. After the fracture closure the pressure

decline is gradual, normal or fast depending on the rate of fluid leak off as seen in

Fig. 2.3.
Bottomhole
pressure

Pc
s
litie
ibi
ss
Po

0
Square root of time elapsed since injection

Figure 2.3- Shut-In Decline Test9

The closure pressure can also be estimated by using the G-plot for the shut-in

decline test. The G-plot is the bottomhole pressure plotted versus the G-function, this

is discussed in detail later in Chapter 4. A derivative plot is used to magnify the

changes to easily identify a change in slope. Estimating the closure pressure from the

shut-in decline test is not very reliable due to the following two reasons7:

9
• The shut-in decline plots may not indicate any change in slope at all.

• There could be more than one change in slope and the change in slope

could be due to several reasons, some of these reasons are as follows7:

o Fracture closure

o Height recession from bounding layers

o Fracture extension and recession transition

o Reservoir linear or radial flow

The closure pressure can also be obtained from the data of flowback test10.

After the fracture is created the well is flowed back at a constant rate. The rate of

flowback is important to estimate the closure pressure accurately. Figure 2.4(a)

shows the influence of flowback rate on the pressure response. The three curves

shown in Fig. 2.4(a) represent the pressure response for too low, correct and too high

flow rates. The correct flow rate is usually 1/6 to ¼ of the last injection rate7.

Flowback
Pump-in Flowback

Rate too Low Intersection of tangents


Bottomhole

Bottomhole
pressure

pressure

(closure pressure
estimate)

Correct rate

Rate too high

0 0
Time Time

Figure 2.4(a)- Influence of Rate Figure 2.4(b)- Estimating closure


on Flowback Test10 pressure from
Flowback Test10

10
The closure pressure is estimated by drawing two tangents on the pressure

response curve as shown in Fig. 2.4 (b). The intersection of these two tangents is the

estimated closure pressure.

2.4 Fluid Loss

The total fluid loss is characterized by two coefficients:

1. Fluid-loss coefficient, CL

2. Spurt-loss coefficient, Sp

Spurt loss occurs for wall building fluids till the filter cake develops11. Making an

assumption of linear flow fluid loss, Carter12 in 1957 developed the following

equation for the rate of fluid loss, qL.

2CL A
qL = …………………………………………………………..……………(2.1)
t −τ

where A is an element of the fracture area created at a time τ and t is the time from

the start of injection. The fluid loss has three distinct regions11 (filter cake, invaded

zone, reservoir control region) as shown in Fig. 2.5.

Figure 2.5- Regions of fluid loss11

11
The fluid-loss coefficient CL can be split up into three coefficients as follows 13:

1. Wall building fluid loss coefficient, Cw

2. Viscosity controlled fluid loss coefficient, Cv

3. Compressibility controlled fluid loss coefficient, Cc

The wall building property of the fracturing fluid defines Cw. The fracturing fluid

filtrate penetrates the formation as is shown in the invaded zone of Fig. 2.5. The

filtrate effect is controlled by Cv. The factors on which Cv depends are as follows11:

a) Relative permeability of the formation to the fracturing fluid filtrate.

b) Pressure difference (pressure inside the fracture minus the reservoir

pressure).

c) Viscosity of the fracturing fluid.

As the fluid leaks off it has to displace the formation fluid, this reservoir effect

defines Cc. The factors on which Cc depends are as follows11:

a) Pressure difference (pressure inside the fracture minus the reservoir

pressure).

b) Permeability of the formation to the mobile formation fluids.

c) Viscosity of the formation fluid.

d) Total system compressibility for the reservoir, ct.

The three coefficients Cw, Cv and Cc vary depending on the factors as mentioned

above and collectively contribute to the total leak-off coefficient, CL.

12
2.5 Mechanics of Hydraulic Fracturing

Figure 2.6 shows a cross section of a borehole with a wellbore fluid pressure

of Pi. The in-situ maximum and minimum horizontal stresses are σx and σy

respectively. The radius of the wellbore is ‘a’. Regional stresses present in the

formation induce stresses around the wellbore14.

The stresses at any point which is at a distance of ‘r’ from the wellbore center

and at an angle ‘θ’ measured from point ‘A’ can be expressed as:

Radial Stress:

 σ x + σ y  a 2   σ x − σ y  4a 2 3a 4     a 2 
σ r =  1 − 2  +  1 − 2 + 4 (cos 2θ ) +  Pi  2  …. (2.2)
 2  r   2  r r     r 

The last term in Eq. 2.2 is that component of stress which is induced around the

borehole due to internal fluid loss.

Tangential Stress:

 σ x + σ y  a 2   σ x − σ y  3a 4     a 2 
σ θ =   
1 + 2  −  1 + 4 (cos 2θ ) −  Pi  2  …….….. (2.3)
 
 2  r   2  r     r 

Shear Stress:

 σ x − σ y  2a 2 3a 4 
τ rθ = − 1 + 2 + 4 (sin 2θ ) ……………………………...……..….. (2.4)
 2  r r 

Figure 2.7 shows the profile of pressure vs. time for a fracture treatment. The

rock initially breaks-down at a pressure equal to Pc1, the instantaneous pressure when

the pumps are shut down is equal to Ps and the pressure at which the existing fracture

propagates further is equal to Pc2.

13
B
r

A' Pi
θ A σ h,max = σ x

B'

σh,min =σy

Figure 2.6- Cross section of the borehole


showing stress distribution14

Breakdown Pressure
Pc1
(Pc1 - Pc2) Fracture Propagation
Pressure
Pc2
Pressure

Ps

Resume Pumping
Stop Pumping (Instantaneous
Shut-in Pressure)

0
Time

Figure 2.7- Pressure Profile versus Time during a Fracture Treatment14

14
In rocks the fracture propagates in a plane perpendicular to the least principal

normal stress. If the tensile strength of the rock is equal to -To then the condition for

a vertical fracture to occur is that at points A and A’:

3σ h,min − σ h,max − Pc1 = −To …………………………………………...………..... (2.5)

The fracture propagation pressure is less than the breakdown pressure; the tensile

strength of the rock can be expressed as:

Pc1 − Pc 2 = To ……………………………………………...…………………..…... (2.6)

The average stress in the horizontal direction is equal to σ h and is defined as follows:

σ h ,min + σ h ,max
σh = ……………………………………...………….…………..... (2.7)
2

K , K and N are ratios that are defined as follows:

σ h ,max σ x
K= = ………………………………...…………………….………..... (2.8)
σ h ,min σ y

σh
K= ………………………………………...……………………………........ (2.9)
σv

σ h ,min
N= ………………………………………..…………………………....... (2.10)
σ h ,max

The conditions for a vertical fracture to take place are:

1  N +1 
N> and K < 
3  6N − 2 

which can alternatively be expressed as Eqs. 2.11 and 2.12.

1 
σ h ,min >  σ h ,max  ……...……………………………………………...……...... (2.11)
3 

15
 σh 
K <   ..……………………………………………………....…..(2.12)

 3σ h , min − σ h , max 

It can be proved that the breakdown pressure is given by the following expression:

 υ (3 − K )  
Pc1 =  (γZ − Pp ) + To ……………………………………..…….......... (2.13)
 (K − υ )  

where υ , γ , Z and Pp are the Poisson’s ratio, fracture gradient, depth and pore

pressure respectively.

In the special case where σ x = σ y = σ h the value of K = 1 and Eq. 2.13 reduces to

Eq. 2.14.

 2υ  
Pc1 =  (γZ − Pp ) + To ………………………………..…………….......... (2.14)
 (1 − υ )  

The average horizontal fracture gradient to create a vertical fracture G f is given by:

 υ (K + 1)   K + 1  1 − 2υ  
G f =   S v +   α p (PG )  …………...…………........ (2.15)
 2(K − υ )   2  K − υ  

where α p is the poro-elastic constant with a range from 0 to 1, PG is the pore pressure

gradient and Sv is the vertical stress gradient due to overburden.

In the special case of α p = 1 , Eq. 2.15 reduces to Eq. 2.16.

 υ (K + 1)   K + 1  1 − 2υ  
G f =   S v +    (PG )  ……………...…………….... (2.16)
 2(K − υ )   2  K − υ  

In the special case of K = 1 , Eq. 2.16 reduces to Eq. 2.17.

 υ   1 − 2υ  
G f =   S v +   (PG )  ………………...…………...……..….... (2.17)
 (1 − υ )   1 − υ  

16
2.6 Hydraulic Fracture Models

Hydraulic fracturing is a complicated process because it involves four

different types of mechanics: fracture mechanics, solid mechanics, fluid mechanics

and thermal mechanics15. For proper design and analysis of hydraulic fracturing it is

necessary to build models. The four major reasons for the development and usage of

fracture models can be summarized as follows15:

1 Economic optimization.

2 Design a pump schedule.

3 Simulation of the fracture geometry and proppant placement obtained from a

pump schedule.

4 Fracture treatment evaluation.

Some of the early work on fracture modeling was reported by Sneddon16,

Sneddon and Elliot17. They proved that for a fracture of fixed height and infinite

extent the maximum width and net pressure can be given as:

2 pnet h f
w= …………………………………....…………….……...…….......... (2.18)
E'

E
where, E ' = is the plane strain modulus.
1 −υ 2

They assumed that the shape of the fracture is elliptical, therefore the average width is

π
given as: w = w.
4

In 1955 Khristianovich and Zheltov18 developed a model considering the

fracture mechanics with simple assumptions about the fluid flow. In 1957 Carter12

built a model by concentrating his work on fluid leakoff and neglected effects of fluid

viscosity and solid mechanics. In 1961 Perkins and Kern19 concentrated their work

17
on fluid flow and neglected the role of fracture mechanics. Geertsma and de Klerk20

incorporated fluid loss in 1969 and developed an extension to the work of

Khristianovich and Zheltov and provided a model known as the GDK model. In 1972

Nordgren21 added leakoff and storage width to the work of Perkins and Kern and

developed a model, which is now known as the PKN model. The GDK model

considers the fracture mechanics aspects of the fracture tip. The assumption in this

model is that the flow rate in the fracture is constant and that the pressure in the

fracture could be approximated by a constant pressure in the majority of the fracture

body, except for a small region near the tip with no fluid penetration, and hence no

fluid pressure. The concept of fluid lag is an important element of the mechanics of

the fracture tip. Warpinski22 validated the concept of fluid lag at the field scale in

1985. Brady et al23 published the results of their work in 1993. They utilized

simulators to study the details of fracture initiation and near-well complexities of

three-dimensional fracture models.

The basic underlying assumption made by Perkins and Kern is that every

vertical cross section acts independently. The pressure at a section is dominated more

by the height of the section than by the length of the fracture15. This is true where the

length is much larger than the height. The PKN model neglects the effect of fracture

tip and fracture mechanics and focuses on fluid flow and their pressure gradients19.

Other assumptions of the PKN model are that the height of the vertical fracture is

constant and does not exceed the pay zone19. The cross section of the fracture is

assumed to be elliptical. At any cross section the maximum width is proportional to

the net pressure at that point and independent of the width at any other point19. The

18
fracture width at any point is proportional to its height. The in-situ stresses are

assumed to be homogeneous and the PKN model utilizes the Sneddon16,17 width

equation.

Figure 2.8 shows a diagrammatic representation of a PKN fracture model24.

z
L

hf /2
V(x)
x

b(0,t) w(x,t)

hf /2

Figure 2.8- PKN Fracture Model24

Perkins and Kern neglected the effects of fluid leakoff and storage resulting from

width increase and developed the following expressions for net pressure and width19:

1
16µqi E '3 L  4
pnet =  …………………………………....…………...…….......... (2.19)
 πh f
4


 q µ (L − x )  4
1

w( x) = 3 i  ………………...……………………….…...…….......... (2.20)
 E' 

Equation 2.20 when expressed in oilfield units results in Eq. 2.21 where ww is the

width at the wellbore ( x = 0 ).

19
In Eq. 2.21 the flow rate, width, viscosity µ and the plain strain modulus E’ are

expressed in bbl/min, inches, centipoise and psi respectively.

1
 q µL  4
ww = 0.38 i '  ………………...………………………….......………......... (2.21)
 E 

Nordgren21 considered leakoff and storage width and obtained Eq. 2.22

 ∂ 2 w4  E '  ∂w   8CL 
 2  = +  .………………….........………........ (2.22)
 ∂ x  128 µh f  ∂t   ( )1
 π t − t f ( x) 
2 

where tf is the time of fracture opening and initial fluid exposure.

Nordgren21 defined a dimensionless time term and solved Eq. 2.22 numerically to

obtain approximation for two cases.

2
 64CL 5 E 'h f  3
td =   t ……………..……………………….….........…….......... (2.23)
 π 3 µq 2 
 i 

Case 1: Storage dominated or high-efficiency (td < 0.01)

In this case the fluid efficiency approaches one (η → 1 ) and the length and width of

the fracture are given by:

1
 E 'qi 3  5
L(t ) = 0.39(t )  4  ……………..……………………….….....…….......... (2.24)
4
5
h µ
 f 

1
 qi 2 µ  5
ww = 2.18(t )  '  …………..……………………….…...........…….......... (2.25)
1
5
Eh 
 f 

Case 2: High Leakoff (td > 1)

In this case fluid efficiency approaches zero (η → 0 ) and the length and width of the

fracture are given by:

20
1  qi 
L(t ) = (t ) 2  ……………..……………………….….....……….…....... (2.26)
 2πh C 
 f L 

1
 qi 2 µ  4
ww = 4(t )
1
8   ……………………………….…...........…...….......... (2.27)
 π 3 E 'C h 
 L f 

For any point at a distance x from the wellbore the following are the approximations

for the PKN model14:

1
  x  4
w = wmax 1 −    …………………………………….…...........….……........ (2.28)
  L 

π
w= wmax ………………………………….…................................……........... (2.29)
5

Figure 2.9 shows a diagrammatic representation of a GDK fracture model24.

The GDK model is based on the assumption that the fracture width is proportional to

its fracture length, the fracture height is constant, the fracture has an elliptical cross-

section in the horizontal plane, there is slippage between layers, fluid does not act on

the entire fracture length and the cross section in the vertical plane is rectangular

(fracture width is constant along its height)14.

21
Area of Highest
Flow Resistance
L
w(x,t)
xL
Vx Approximately Elliptical
Shape of Fracture
w(0,t)

Rw

hf

Figure 2.9- GDK Fracture Model24

Geertsma and de Klerk15 incorporated fluid loss into the work of Khristianovich and

Zheltov and developed Eq. 2.30, Eq. 2.31 and Eq. 2.32:

4
ww = (Pnet L ) ………………………………………………..........…...…........ (2.30)
E'
1
 ' 3 21qi µ  4
Pnet ,w = E  …………………………………….…......…...….......... (2.31)
 64πh L2 
 f 

1
 1  84qi µL2  4
ww =  '   …………………………..…….…...........…...….…..... (2.32)
 E  πh f 

The length and width can be expressed as a function of time for two different cases as

follows15:

Case 1: No Leakoff (where L, t, E’, qi, hf and µ are in ft, min, psi, bbl/min, ft and cp)

1
 E 'qi 3  6
L(t ) = 0.38(t ) 3  3  ……………..……………………….….....…….......... (2.33)
2

h µ
 f 

22
1
 q 3µ  6
ww = 1.48(t ) 3  i' 3  …………..……………………….….............……....... (2.34)
1

Eh 
 f 

Case 2: High Leakoff (where L, t, qi, hf and CL are in ft, min, ft3/min, ft and ft/min1/2)

1  qi 
L(t ) = (t ) 2  ……………..……………………….….....………..…...... (2.35)
 2πh C 
 f L 

In case of high leakoff no explicit equation for width has been provided by Geertsma

and de Klerk.

For any point at a distance x from the wellbore Eqs. 2.36 and 2.37 are an

approximation for the GDK model14:


1
  x  2
w = wmax 1 −    …………………………………….…...........…...….......... (2.36)
  L 

π
w= wmax ………………………………….…................................…...…........ (2.37)
4

This model assumes that the fracture propagates in a uniform stress medium

without any vertical containment, the shape of the fracture is circular and the

maximum width is at its center. Other assumptions made in this fracture model are

that the fracture propagates radially by the same distance, pressure drop at the same

distance from the wellbore in all directions is equal, no boundary containment or

interface slippage and the width decreases in all directions14.

The maximum width of the radial fracture can be expressed14 by Eq. 2.38:

 8P R 
wmax =  net'  ……………………………….................................…...…........ (2.38)
 E 

23
Geertsma and de Klerk20 developed the equations for a radial fracture. The

width and radius approximation for different cases are summarized below:

Case 1: No Leakoff (where ww, t, E’, qi, R, and µ are expressed in inches, min, psi,

bbl/min, ft and cp)

1
 µ 2 qi 3t  9
ww = 2.17   ……………………….….................................…...…........ (2.39)
'2
 E 

1
 E ' qi 3  9 4 9
R = 0.52   t …………………….…....................................…...…........ (2.40)
 µ 

Case 2: High Leakoff

1
1 q t 
2 4
R =  i 2  …………………….…..............................................…...…........ (2.41)
π  CL 

24
Chapter 3

Mini-Fracs and Diagnostic Pump-In Tests

3.1 Mini-Fracs

To ensure the success of a fracture treatment, a careful design and planning of

the fracture treatment job in advance is a must. The design of the fracture treatment

requires several parameters. A mini-frac is performed to obtain some of these

parameters. The parameters that are obtained from the mini-fracs are closure

pressure, in-situ elastic rock properties, fluid-loss coefficient, fluid efficiency,

fracture geometry, and fracture height. The mini-frac essentially consists of two

parts: the stress test and the calibration treatment24,25. Brief outline of these tests is

given below25.

Potassium chloride treated water (water with 4 to 6 percent KCl as an

additive) without proppant is pumped till the formation breaks down; subsequently a

step rate injection test is carried out. In the next step a flowback or a pressure decline

analysis is performed from which the closure pressure is determined. One of the most

critical parameters in any fracture treatment design is the closure pressure. The

closure pressure can be defined as equal to the minimum horizontal in-situ stress. In

addition to the closure pressure the stress test also determines the fracture extension

pressure, perforation and fluid friction pressure25.

A calibration treatment is performed after the stress test. In the calibration

treatment the fluid pumped is generally the same that will be used later in the actual

fracture treatment; but without the proppant24,25. Fluid without proppant is pumped at

25
the same rate as that is planned for the fracture treatment. The pressure decline is

closely monitored until closure is attained. Information obtained from the pump-in

period is: the type of fracture propagation, identification of rapid height growth,

detection of fissure opening25. The other critical parameters obtained from the

pressure decline are the in-situ fluid efficiency and fluid loss co-efficient for a

particular fluid in the well25,26,27. Fluid efficiency is defined as the ratio of the volume

of the fracture created to the volume of the fluid pumped. The information obtained

is then utilized in optimizing the fracture treatment.

3.2 Diagnostic Pump-In Tests (Breakdown Tests)

In known areas where several wells have been drilled and fractured in the near

vicinity the preference is to perform a Diagnostic Pump-In Test, also known as a

Breakdown test. The breakdown test is described below.

The fluid in the casing is displaced to KCl treated water. The zone of interest is

perforated. Potassium chloride treated water is pumped down the casing with the

well shut-in. The pressure inside the casing keeps rising until the fracture gradient of

the formation is reached. Further pumping of the fluid results in the formation being

fractured, this is confirmed by a drop in the pressure. Subsequently balls made of

synthetic material are dropped in the wellbore. The number, size and the

specifications of the balls are selected based on the number and size of perforations.

The well is shut-in and the balls are allowed to fall to the bottom of the wellbore.

Pumping of fluid is resumed and a ‘Ball off” operation is performed. The ‘Ball off”

operation ensures that the balls seal off the open perforations, and with fluid being

26
pumped the closed and partially open perforations begin to open up. When all

perforations have been opened and all restrictions cleared up the pumping of fluid is

stopped. The balls are then surged off the perforations and allowed to fall to the

bottom of the wellbore. A step rate test is performed at different pressures and the

rates and pressures are recorded. The details of the step rate test are described in

Chapter 2. The well is the shut-in and the shut-in decline data monitored. At the end

of the shut-in decline the pressure is bled-off and the well flown back. The shut-in

decline data is analyzed to determine the G-function values at various pressures,

closure pressure, fracture geometry, leak-off coefficients and pump-in derived

permeability. The G-function is described in detail in Chapter 4.

3.3 Comparison of Mini-Fracs with Diagnostic Pump-In Tests (Breakdown

Tests)

The major differences between mini-fracs and the breakdown tests are as

follows:

• The Stress test portion of the Mini-Frac is performed with KCl treated

water and the calibration treatment portion of the mini-frac is

performed with the same fluid that will be used in the main fracture

treatment. The breakdown test is performed only with KCl treated

water.

• A mini-frac uses the same blending equipment as those planned for

main fracture treatment. It is due to this reason that the mini-frac is

performed immediately prior to the main-frac job. A breakdown test is

27
performed several days prior to the main-frac job, this provides

sufficient time to make a decision whether to perform the main

fracture treatment or not.

• Upon review of the results of the mini-frac, last minute changes in the

main fracture treatment design are made. The main-fracture treatment

is conducted after the mini-frac. After analyzing the results of the

breakdown test one decides whether the main-frac job must be carried

out or not.

• Mini-Fracs require more time and equipment (blenders); therefore

mini-fracs are more expensive and time consuming as compared with

breakdown tests.

• Due to the additional equipment, the number of personnel required for

a mini-frac is more, which also adds to the total cost.

• Conducting mini-frac is not critical in known areas.

It is for these reasons that breakdown tests are becoming more common in the field.

28
Chapter 4

Fracture Pressure Analysis and G-Function

4.1 Generalized Expression for Fluid Loss

Once the formation breaks down fracture is initiated. Further growth of the

fracture results in an increase of the fracture area. If the fracture area at any time τ is

equal to a and this fracture area changes to A at time t then the ratio of the fracture

areas at these two different times is expressed7 in terms of Eq. 4.1:

α
a τ 
=   ……………………………....…………...……….………...……........ (4.1)
A t

where, α is the fracture area growth exponent

To consider Newtonian as well as non-Newtonian fluid-loss filtrate a generalized

expression for fluid loss flux (uL) can be expressed in terms of the fluid loss rate ( q L )

as follows7:

qL 2C L
uL = = ….………………....………………….…….…..……........ (4.2)
A (t − τ )1−θ

where, θ is the fluid loss exponent. By integrating Eq. 4.2 with respect to time the

specific fluid loss volume vL is obtained as follows:

t
 2C θ
vL = ∫ u L dt = −  L (t − τ (a ))  ………….......……………….……....……........ (4.3)
0  θ 

where, vL is defined as the fluid loss volume per unit leak off area.

If the flow behavior index of the power law fluid model of the fracturing fluid

filtrate invading the formation is given by nf, then it can be proven that28:

29
nf
θ= …………………………….......………………….…….…...……........ (4.4)
nf +1

1
For a Newtonian fluid filtrate the value of n f = 1, θ = and Eq. 4.2 reduces to Eq.
2

2.1 which is the Carter’s equation for fluid loss7. Similarly when the fluid filtrate is

1
pseudoplastic non-Newtonian7 then the value of n f < 1, and θ < . For viscoelastic
2

filtrate29 the value of θ is greater than ½ and less than 1, the values of θ ≈ 1 for high

values of nf (of the order of 10-15).

4.2 Cumulative Fluid Loss Volume

The cumulative fluid lost can be expressed as:

VL = VL ,C + VL ,S …………...………….......………………….………....……......…(4.5)

where, VL,C is the Carter’s fluid loss volume (CL component) and VL,S is the spurt loss

fluid volume.

To obtain an expression for the fluid loss volume a dimensionless shut-in

time ∆t D , dimensionless area parameter ξ and dimensionless time tαD are defined as

follows:

t − tp ∆t
∆t D = = …………………….….……..…………………...……......……(4.6)
tp tp

a t
ξ= and tαD = ……………...….......………………….……...……......…...(4.7)
Af tp

where, Af is the fracture area at the end of injection and tp is the injection time. The

term A in Eq.4.1 is the fracture area at any intermediate time t. Substituting the terms

30
of Eq. 4.7 into Eq. 4.3 and integrating over the area, the volume of the Carter

component of fluid loss VL ,C (CL component) can be obtained as follows7:

(
V L ,C = 2 rp A f C L t p
θ
) g (∆t D, α , θ ) ………...…………………….……...……......….(4.8)

where, rp is the ratio of permeable area of the fracture to the total fracture area and the

dimensionless fluid loss volume function g (∆t D , α , θ ) is defined as7:

1 θ
g (∆t D , α , θ ) = ∫ 1 + ∆t D − ξ α  dξ
1 1
∆t D > 0 ………………….......……(4.9)
θ 0 

where, ∆tD, α, θ and ξ are the dimensionless shut-in time, fracture area growth

exponent, fluid loss exponent and dimensionless area parameter. At the end of

injection, substituting ∆t D = 0 in Eq. 4.9, the following Eq. 4.10 is obtained:

θ1
1
g (∆t D = 0, α , θ ) = g 0 (α ,θ ) = ∫ 1 − ξ α  dξ
1
∆t D = 0 …….….........(4.10)
θ 0 

In 1989 Meyer and Hagel30 proved that the G-function in Eq. 4.9 for Newtonian

fluids can be expressed as follows:

α Γ(α ) π
g 0 (α ) = ……….......……………….………………….…….…....….(4.11)
3 
Γ + α 
2 

where, g 0 (α ) is equal to the fluid loss volume function g (∆t D , α , θ ) when ∆t D = 0

1
and θ = (Newtonian fluids). The term Γ(α ) in Eq. 4.11 is the Euler gamma
2

function and α is the fracture area growth exponent. Alternate expression of the fluid

loss volume function was given by Valko and Economides31 in terms of a

hypergeometric function F [a, b; c; z ] as follows:

31
 1 −1  
4α ∆t D + 2 1 + ∆t D  F  , α ; 1 + α ; (1 + ∆t D )  
g (α , ∆t D ) =  2 
……........…........ (4.12)
1 + 2α

The hypergeometric function F [a, b; c; z ] is available in tabular form32 or through

computing algorithms.

Some of the fracturing fluids exhibit a wall building characteristic. The

volume of fluid lost prior to the cake build up is known as the spurt fluid loss ( VL ,S )

and is expressed in7 Eq. 4.13, where SP is the spurt loss co-efficient:

VL ,S = 2rp S P A f ………………………….…….......……………….……........…..(4.13)

4.3 Newtonian Filtrates

Fracturing fluids that have a wall building characteristic produce a Newtonian

1
filtrate ( n f = 1 and θ = ) after the wall cake has been deposited. The fracture area
2

1
growth exponent α as defined in Eq. 4.1 has bounding values given by: < α < 1.
2

The simplified expressions for the fluid loss volume function ‘g’ for the bounding

values of α are obtained by substituting the values of α and θ into Eq. 4.9 and 4.10

and integrating the resulting equations8,9. Simple interpolation can be used to obtain

the values of ‘g’ for other values of α lying between ½ and 1. Table 4.1 is obtained

from Eqs. 4.9 and 4.10.

32
Table 4.1: Analytical equations8,9 for g (∆t D , α , θ ) and g 0 (α , θ )

g0 (∆t D = 0) g (∆t D )

1 1 π 1 + ∆t D sin −1 (1 + ∆t D )
−1
+ ∆t D
1
α= , θ= 2 2

2 2 2

α = 1, θ =
1
2
4
3
4
3
(
(1 + ∆t D ) 2 − ∆t D 2
3 3
)

The CL component of fluid loss at the end of the injection period is obtained

1
by substituting ∆t D = 0 and θ = in Eq. 4.8 as follows:
2

 1
A f g 0 α ,θ = 
1/ 2
V L ,C = 2 rp C L t p ………….…...………………...…........……(4.14)
 2

The total fluid loss at the end of the injection period is obtained by substituting Eqs.

4.13 and 4.14 into Eq. 4.5 and simplifying as follows7:

VL , P = 2rpκC L t p A f g 0 α ,θ = 1 ( 2
) ……………………………...….…..…….. (4.15)

where, the spurt factor κ which provides for increase in fluid loss over no-spurt

condition is given by the following expression:

SP
κ = 1+
(
g 0 α , θ = 1 CL t p
2
)
………….……………………………....……....….(4.16)

The fluid loss during the shut-in period is obtained by subtracting Eq. 4.15

from Eq. 4.8 as follows27:

(
VLS (∆t ) = 2rp CL t A f ) [g (∆t ) ( )]
α ,θ = 1 2 − g 0 α ,θ = 1 2 ………...……......….(4.17)
D,

33
4.4 Non-Newtonian Filtrates

When fracturing fluids that have a power law based fluid rheology and do not

build up a filter cake; the filtrate that leaks off is a non-Newtonian fluid. Integrating

Eq. 4.10 an expression for g 0 (α ,θ ) is obtained7:

Γ(1 + θ )Γ(1 + α )
g 0 (α ,θ ) = ……………………………………...…...……......….(4.18)
θ Γ(1 + α + θ )

where, Γ(x ) is the gamma function.

For the upper bound, value of α = 1 and the fluid loss volume function is expressed

as7:

g (∆t D , α = 1, θ ) =
1
θ (1 + θ )
[ ]
(1 + ∆t D )1+θ − (∆t D )1+θ ……………...……..……....….(4.19)

In cases where the value of θ approaches 1 the fluid loss volume function is

expressed as7:

1
g (∆t D , α , θ = 1) = + ∆t ……………...….…………….…….…….....…(4.20)
(1 + α ) D
The fluid loss at the end of the injection period is given as follows7:

1
VLp = 2rp C Lt p A f g 0 (α ,θ )
θ
θ≠ ………….……...………......…..(4.21)
2

The fluid loss during the shut-in period is obtained by subtracting Eq. 4.19

from Eq. 4.8 as follows7:

(
VLs (∆t ) = 2rpCLt θ A f ) [g (∆t α ,θ ) − g 0 (α ,θ )]
D, θ≠
1
2
………...(4.22)

34
4.5 Typical Pressure Profile of Fracture Treatment

The pressure profile of a typical fracture treatment9 is shown in Fig. 4.1. The

bottomhole pressure pw increases as proppant-laden fluid is injected until the fracture

is initiated. Following the injection stage the well is shut-in and the bottomhole

pressure declines until the fracture closes on the proppant at the closure pressure, pc.

The net pressure in the fracture is equal to pw - pc. The pressure keeps reducing in the

fracture until the bottomhole pressure reaches the reservoir pressure.

Figure 4.1- Pressure Profile of a Fracture Treatment7

4.6 Basic Definitions

Some of the basic terminologies used in subsequent sections are defined

below.

The fracture treatment efficiency (η) can be defined7 as the ratio of the volume

of the fracture created at the end of pumping (Vfp) to the total volume of fluid injected

(Vi).

V fp
η= ……………………………....…………...……….……...…….….......... (4.23)
Vi

35
During the fluid pumping stage the pressure within the fracture varies from

the maximum pressure pw at the wellbore to pc the closure pressure that is a short

distance away from the fracture tip. The pressure gradient7 within the fracture is

characterized by a factor β which is defined as the ratio of the average net pressure

∆ p f in the fracture to the net pressure at the wellbore pnet.

∆pf
β= …………………………....…………...……….……...…….….......... (4.24)
pnet

pnet = pw − pc …………………………....…………...……….……….….......... (4.25)

∆ p f = p f − pc …………………………....…………...………...……...…........ (4.26)

where p f is that constant internal pressure that would produce the same average

width as that when a pressure gradient exists.

The pressure gradients during injection and shut-in are different, which

implies that the value of β is different for these two stages. For injection stage as

well as shut-in stage the values of β are selected at the end of the respective stage and

are designated as β p and β s respectively.

The fracture width averaged over its length and height known as w is directly

proportional to its net pressure. The proportionality constant c f is the fracture

compliance and depends on the factor β , the plane strain modulus E’, and the fracture

geometry model being considered7.

w = c f p net ……………………………....…………...………...…….…........... (4.27)

36
4.7 Equations for Pumping Stage

The equations relating fracture pressure, fracture geometry and time during

fluid injection stage have been developed by Nolte33,34 and are given in the following

sections.

The fracture net pressure is directly proportional to the term as given in Table

1
4.2. The exponent e = , where n and K are the power law behavior exponent
2n + 2

and consistency index of the fracturing fluid. As can be seen from Table 4.2,

Nolte33,34 concluded that for all models the fracture net pressure depends on

parameters like n, K, qi and fracture extension (L or R). In case of PKN model the net

pressure is directly proportional to the fracture extension L, whereas in case of GDK

and radial fracture model the net pressure is inversely proportional to L and R

respectively. From Table 4.3 it can be concluded that a plot of net pressure versus

time on a logarithmic scale would yield a straight line with slope equal to the

exponent in Table 4.3. The value of the flow behavior index can then be calculated

from the actual data.

37
Table 4.2: Fracture Net Pressure – Fracture Geometry Relations33,34

Model Net Pressure

proportional to

PKN  L 
e

(E ' 2 n+1
K qi )
n e
 3n+1 
 h f 

GDK  1 
e

(E ' 2 n +1
K qi )
n e
 n 2n 
 h f L 

Radial
(E )
e
' 2 n +1 n e  1 
K qi  R 3n 

Similar to the relations for fracture net pressure Nolte33,34 also developed

relations for the fracture width and are summarized in Table 4.4.

Table 4.3: Fracture Net Pressure – Time Relations33,34

Model Net Pressure

proportional to

↓ ↓

PKN (t )1 / 4(n+1) for η → 0 (t )1 / (2 n+3) for η → 1


GDK (t )− n / 2(n+1) for η → 0 (t )− n / (n+2 ) for η → 1
Radial (t )−3n / 8(n+1) for η → 0 (t )− n / (n+2 ) for η → 1

38
Similar conclusions can be drawn from Table 4.4; Nolte33,34 recognized that

for all models the fracture width depends on parameters like n, K and qi. In case of

PKN model the width is directly proportional to the fracture extension L and the

fracture height hf, whereas in case of GDK the width is directly proportional to the

fracture extension L and inversely proportional to the fracture height hf. In case of the

radial fracture model the width is directly proportional to the fracture extension R.

Table 4.4: Fracture width relations33,34

Model Fracture width

proportional to

PKN e
 Kqi n  1−n e
 '  hf L
 E  [ ]
 

GDK  Kqi n 
e
 L2 
e

 '   n
 E 
   h f 

Radial e
 Kqi n  2−n
 '  R
 E  [ ]e

 

4.8 Log-Log Plot

The slope of a log-log plot of net pressure versus time provides a lot of

information about the fracture propagation mode. Figure 4.2 is a log-log plot

showing various fracture propagation modes. Table 4.5 provides the interpretation of

the various slopes on the log-log plots.

39
Log net pressure
IV Va
III
Vb
II
Ia, Ib

Log time

Figure 4.2- Log-Log Plot showing various fracture propagation modes7

Table 4.5: Interpretation of log-log plot slopes7

Section Slope of Section Slope of Section Interpretation from


of Plot (for n = 0.5) (for n = 0.25, Slope
typical for frac
fluids)
Ia −1 to − 1 − 1 to − 1 GDK model propagation
6 5 10 3
Ib −1 to − 1 −3 to − 1 Radial model
8 5 40 9 propagation
II 1to 1 1 to 2 PKN model propagation
6 4 5 7
III Less than section Less than section Controlled height
II II growth or Stress
sensitive fissure
IV 0 0 Height growth through
pinch point or Fissure
dilation or T-shaped
fracture
Va ≥ 1 ≥ 1 Restricted Extension
Vb Negative Negative Uncontrolled height
growth

The ratio of perforated interval to the reservoir thickness determines the initial

section (Ia/Ib) of fracture propagation. In case of the length of perforated intervals

equal to the length of reservoir the fracture propagation approximates a GDK model

40
(Section Ia). Alternatively in the case of limited fluid entry interval (low ratio of

perforation length to reservoir length) the fracture propagates in a radial mode

(Section Ib). The slope in this section is negative indicating that the fracture is

propagating in an unrestrained manner with a reducing net pressure7.

The section II in Fig. 4.2 is observed when the fracture height growth is

confined, this happens when the reservoir is bounded by higher stress zones above

and below. The fracture propagation approximates a PKN fracture model. The net

pressure rises as the fracture extension increases and the resulting slope of the plot is

positive. The slope of the plot in section Ia, Ib and II depends on the value of n and

η (fracture treatment efficiency), the relationships given in Table 4.3 are utilized to

calculate the slope for different values of n and η (fracture treatment efficiency).

Section II progresses till the fracture net pressure approaches a value of one

half the stress difference between the stress of the formation being fractured and the

stress of the adjoining barrier. When this happens the fracture begins to penetrate the

adjoining barrier. This controlled height growth is shown as section III in Fig. 4.2.

Fissures are very fine cracks that occur naturally in the earths crust. Some of these

fissures are stress sensitive. During fracturing the increase in the stress causes these

fractures to open up and dilate. The dilation of stress sensitive fissures is another

possible cause of the section III in Fig. 4.2. Propagation of the fracture from a low

stress formation to a high stress formation causes the fracture width to abruptly

reduce, this phenomenon produces a pinch point as seen in Fig 4.3(b). A horizontal

fracture immediately above a vertical fracture is known as T shaped fracture. The

causes of Section IV (usually a flat section) as seen in Fig. 4.2 are either height

41
growth through a pinch point or fissure dilation or a T-shaped fracture. Figure 4.3

shows the typical shape of a T-shaped fracture, height growth through a pinch point

and dilation of stress sensitive fissures.

Pinch Point Fissures

Fracture width

T-Shape

width
width
(c)

(a) (b)

Figure 4.3 (a)- T-Shaped Fracture


(b)- Height growth through Pinch Point
(c)- Fissure Opening

Section Va shows a sharp shoot-up in the net pressure as well as the slope of

the plot rises ( > 1). The possible cause for this type of behavior is restricted

extension (screenout). The slope of this section is less if the restriction exists at the

fracture tip (Tip screenout) as compared to the slope of the section when the

screenout occurs at the wellbore.

42
Section Vb in Fig. 4.2 is a characteristic indication of uncontrolled height

growth. This could happen when the fracture propagates vertically beyond the barrier

intervals into a low stress region.

4.9 Pressure Derivative Plot

Fracture pressure diagnosis and analysis become easier when the pressure

derivative is plotted in addition to the net pressure on the same graph using a

logarithmic scale. Ayoub et al35 showed that by plotting the pressure derivative we

can estimate the closure pressure or confirm closure pressure obtained from other

techniques as illustrated in Fig. 4.4.

10000
Underestimated pc
Net Pressure (psi)

ct p c
Corre

pc
1000 Overestimated

ativ e
deriv
re s sure
P

100
1 10 100
Time (min)

Figure 4.4- Estimation of closure pressure35

pnet = pw − pc = At b …………………….…………...………...…….….............. (4.28)

The terms b, A and t are defined as:

43
• b is the slope of net pressure and pressure derivative plots on a log-log scale

• A is a constant

• t is time

The net pressure and the pressure derivative plots are separated by a value of 1/b on a

log-log scale.

Differentiating Eq. 4.28 with respect to time t and multiplying both sides of the

resulting equation by t we get,

 dp w 
t dt  = Abt ……………………….…………..……………...…….….......... (4.29)
b

 

Nolte34 showed that a pressure derivative plot gives a magnification of

fracturing events (eg. Introduction of proppant, Tip screenout) as illustrated in Fig.

4.5.
Net Pressure (psi)

Injection time (min)

Figure 4.5- Magnification of fracture pressure events34

The pressure derivative plot is much more sensitive than the net pressure plot

due to which the fracturing events are easily recognized. The fracture height growth

into the barrier zones can be quantified and tip screenouts are detected much earlier34.

44
4.10 Near WellBore Pressure Losses

The fracture pressure analysis can give erroneous results if the near wellbore

pressure losses are not accounted for. The near wellbore pressure losses occur due to

three main reasons as given below7:

When the fluid is injected, kinetic energy gets dissipated as it flows with a

high velocity through the small opening of the perforation. This pressure loss could

be high if the number of perforations is insufficient or there was a poor perforation

cleanup or if the breakdown procedure was ineffective. One way to identify high

perforation friction, ∆p pf is when the bottomhole pressure measurement displays an

instantaneous change during shut-in7. The perforation friction pressure can be

quantified from the change in the bottomhole pressure at shut-in or by performing a

rate step-down test as explained subsequently in the section on rate step down test.

The number of perforations N that are accepting fluid can be estimated by using Eq.

4.30, this equation is derived from the standard orifice equation:

 q 2ρ 
∆p pf = 0.2369 2 i 4 2  ……………………...……………...…….…........ (4.30)
N D C 
 p 

where the flow rate, density of fluid, diameter of perforation and the discharge

coefficient are denoted by qi, ρ , Dp and C respectively. In case the value of ∆p pf is

very high that would restrict the value of qi; re-perforating with larger holes or a

perforation clean up can be performed. For non-Newtonian fluids El-Rabba et al36

developed correlations to estimate the coefficient of discharge used in Eq. 4.30.

45
The connection between the wellbore and the main fracture body could be

straight or it could be a highly convoluted path as illustrated in Fig. 4.6. This latter

case is defined as near wellbore fracture tortuosity15.

Figure 4.6- Fracture Reorientation15

Tortuosity can have a major effect on the fracture treatment when the wellbore

and the stress fields are misaligned37. The effect produced is a reduction in the width

of the fracture. In extreme cases there is excessive reduction of the width at the

wellbore which may result in near wellbore screenouts.

The effects of tortuosity reduce as the treatment progresses. One method to

reduce the pressure drop caused by tortuosity is to increase the fluid viscosity38.

Increasing the pad volume is another way to reduce effects of tortuosity and prevent

near wellbore screenouts. The side effect produced by these methods is that it can

affect the height confinement, permeability in the fracture and proppant placement.

46
Some of the perforations are well aligned with the preferred fracture plane

whereas others may be misaligned to as much as 90 degrees. When the fracture does

not initiate at the perforation then the fluid communication between the perforation

and the fracture takes place through narrow channel on the side of the casing39. This

is illustrated in Fig. 4.7. This width restriction results in an increase of treating

pressures and in certain cases the proppant forms a bridge at the pinch point which

may result in a screenout.

Figure 4.7- Perforation Misalignment15

Unlike tortuosity the effects produced by the pinch points increase as the

pressure increases. The fluid at the pinch point has a high velocity which has an

erosive effect. This phenomenon takes place where proppant slugs are pumped to

erode the pinch points and thereby reduce the restriction37.

47
4.11 Rate Step-Down Test

Rate step-down test is performed to identify the existence of near wellbore

pressure loses (perforation friction and losses due to tortuosity)40. These losses can

be estimated and the approximate number of blocked perforations can be identified.

The procedure for the rate step-down test is as follows40:

1. Following a calibration treatment or a breakdown injection test, reduce the pump

rate in intervals of 20 to 30 percent of the maximum rate.

2. At each interval measure the stabilized pressure (preferably bottomhole) and the

rate. The change in the two constant rates should be done quickly.

3. Calculate the drop in pressure ∆p at each interval of the rate step down test. The

perforation pressure from Eq.4.30 :

2
∆p pf = k pf qi ………………………...…………………………...…….….......... (4.31)

 ρ 
where k pf = 0.2369 2 4 2  ………………....……………......…….…........ (4.32)
N D C 
 p 

The pressure drop due to tortuosity is given as37:

1
∆pt = kt qi 2 ……………...…...………………………………......…….….......... (4.33)

where kt is a proportionality constant. The pressure drop, ∆p for each interval of the

rate step down test is substituted in Eq. 4.34 and the values of kpf and kt are

determined using a dual fit algorithm.

1 2
∆p = kt qi 2 + k pf qi ……………...…………………….……......…….…............ (4.34)

A plot of ∆p , ∆p pf and ∆pt versus the rate is prepared on the same graph to evaluate

the severity of perforation friction and tortuosity friction as illustrated in Fig. 4.8.

48
Figure 4.8- Rate Step-Down Test (Adapted from Wright40)

It can be seen from Fig. 4.8 that for this particular case the perforation friction losses

are higher as compared to the tortuosity losses. Based on the results obtained we can

then take the appropriate action to rectify the problem.

4.12 Dimensionless Fluid Loss Function G (∆t D )

The fracture area growth exponent α has a lower bounding value of α 0 when

the fluid efficiency η → 0 (high fluid loss) and an upper bounding value of α1 when

the fluid efficiency η → 1 (negligible fluid loss). Equation 4.35 defines7 the lower

bounding value of α .

α 0 = 1 − θ ……………………………....…………...……….……...……........ (4.35)

nf 1
where, θ = ; in case of a typical value of θ = , the lower bounding value of
nf +1 2

the area exponent α 0 equals ½ ( as nf = 1 for Newtonian fluids).

49
Table 4.6: Values of α1 (Upper bound of area exponent)7

Model α1

PKN 2n + 2
2n + 3

GDK n +1
n+2

Radial 4n + 4
3n + 6

The upper bounding value of the area exponent α1 depends on the fracture model and

the flow behavior index of power law fracturing fluid and is given7 in Table 4.6

The value of α can be calculated by interpolating between the lower and upper bound

values as shown below7:

α = α 0 + η (α1 − α 0 ) …………………....…………...……….……...……........... (4.36)

Table 4.1 gives the analytical equations of g (∆t D , α ) and g 0 (α ) with θ =


1
and for
2

upper and lower bounds of α . Values of g (∆t D , α ) at other values of α are

obtained by interpolation7 with Eq. 4.37, where, α1 is the upper bound of the fracture

area growth exponent α and η is the fracture treatment efficiency:

 1   1 
g (∆t D , α ) = g  ∆t D , α =  + η ((2α 1 ) − 1) g (∆t D , α = 1) − g  ∆t D , α =  ……... (4.37)
 2   2 

Nolte41 introduced a function G (∆t D ) defined as follows:

4
G (∆t D ) = [g (∆tD ) − g0 ]……………..............…………...……….………........ (4.38)
π

50
It can be proven7 that the bottomhole pressure pw during the shut-in period can be

expressed by Eq. 4.39, where pws is the bottomhole pressure at shut-in.

  π rpCL t p  
pw = −    [G (∆t D )] + pws .............……...……….…………........ (4.39)
  2c f  
 

Equation 4.39 indicates that a plot of pw versus G (∆t D ) has a negative slope equal

π rp C L t p dP
to as illustrated in Fig. 4.9. The derivative (pressure with respect to
2c f dG

G-function) is also shown in Fig. 4.9.

Figure 4.9- G-Function versus Pressure and Pressure Derivative7

The G-function at the closure pressure pc is defined as Gc. Fluid efficiency is defined

as the ratio of the volume of the fracture created to the volume of the fluid pumped.

If the fluid leaks off at a high rate then the volume of the fracture created is relatively

smaller then the volume of the fluid pumped. A low fluid efficiency indicates a high

51
rate of fluid leakoff and a high fluid efficiency indicates a low rate of fluid leakoff.

The fluid efficiency η can be expressed7 in terms of the G-function at closure and the

spurt factor κ (defined in Eq. 4.16) as shown below in Eq. 4.40 :

Gc
η= ...........…………………………………….....……….………........ (4.40)
2κ + Gc

The leakoff coefficient CL can be calculated from the slope m as given in Table 4.7,

where β s is the pressure gradient in the fracture during shut-in as was defined earlier.

The expressions in Table 4.7 can be utilized to calculate the leakoff coefficient

provided the values of rp and hf (for PKN models), L (for GDK models) and R (for

Radial models) are known.

Table 4.7: Values7 of CL (Leakoff coefficient)

Model CL

mβ s
PKN
[ ]
hf
rp t p E '

GDK mβ s
[2 L]
rp t p E '

Radial mβ s  32 R 
rp t p E '  3π 
2

4.13 Interpretation of G-function plot

Subsequent to the work done by Nolte41 and Castillo27, Barree and

Mukherjee42 developed techniques for identifying pressure dependent leakoff and its

52
effect on fracture geometry. The technique developed by Barree and Mukherjee42 is

dP dP
to plot pressure, pressure derivative , and the superposition derivative G
dG dG

versus the G-function. Their research work was done on five hypothetical case

studies in which unique behavior of the plot for pressure dependent leakoff, fracture

extension, fracture recession and height changes were identified. They validated their

work by using a fully 3-D fracture geometry model43. The technique developed by

Barree and Mukherjee42 was applied to field case studies by Craig et al 44


. The G-

function plots were utilized to identify the type of leakoff, fracture closure stress and

fracture geometry behavior. The G-function plots for four common leakoff types in
44
low permeability sandstones from Craig et al are described in the following

sections.

Figure 4.10 shows the G-function plot for a normal leakoff.

Figure 4.10- G-Plot - Normal Leakoff (from Craig44)

53
The requirements for a normal leakoff to occur are a constant fracture area

during shut-in and a homogeneous rock matrix. The two characteristic signatures for

normal leakoff are44:

dP
1. A constant pressure derivative .
dG

dP
2. The superposition derivative G lies on a straight line that passes
dG

through the origin.

dP
The fracture closure point is identified on the superposition derivative G
dG

curve when the curve starts to deviate from the straight line as shown in Fig. 4.10.

This point is then projected vertically on to the pressure curve and the pressure at this

point is the fracture closure pressure. A small initial portion of the pressure decline

might be pressure dependent, but the majority (more than 80%) of the pressure

decline is a normal leakoff.

Dilation of fractures or fissures during injection results in a pressure

dependent leakoff during shut-in. Figure 4.11 shows the G-function plot for a

pressure dependent leakoff due to fissure opening.

54
Figure 4.11- G-Plot - Pressure Dependent Leakoff -Fissure Opening
(from Craig44)

The characteristic signatures for a pressure dependent leakoff are44:

dP
1. A large hump in the superposition derivative G .
dG

2. Subsequent to the hump, the pressure decline exhibits a normal leakoff.

3. The portion of normal leakoff lies on a straight line passing through the

origin.

4. The hump lies above the straight line that passes through the origin.

The pressure at the point where the straight line begins is the fissure opening

pressure. The superposition derivative then exhibits a decline from the straight line

indicating a fracture closure. The fracture closure point is identified on the

dP
superposition derivative G curve when the curve starts to deviate downwards
dG

from the straight line as shown in Fig. 4.11. This point is then projected vertically on

to the pressure curve and the pressure at this point is the fracture closure pressure.

55
Figure 4.12 shows the G-function plot for a pressure decline in which fracture

tip extension occurs. Fracture tip extension is said to occur if the fracture continues

to grow after shut-in. Fracture tip extension is a typical phenomenon occurring in

low permeability reservoirs, in most of the cases it results in poor well

productivity45,46. Leakoff is very low in low permeability reservoirs; the energy

which cannot be released through leakoff then results in fracture-tip extension after

shut-in.

The characteristic signatures for a fracture tip extension are44:

dP
1. The superposition derivative G curve lies on a straight line.
dG

2. This straight line when extrapolated intersects the Y-axis above the origin.

In Fig. 4.12 fracture closure has not yet taken place as the superposition derivative

dP
G does not deviate downwards from the straight line.
dG

Figure 4.12- G-Plot – Fracture Tip Extension (from Craig44)

56
Figure 4.13 shows a G-function plot where fracture height recession occurs.

The layers adjoining the fracture zone are usually high stress and having relatively

low permeability. During injection if the fracture propagates through the adjoining

impermeable layers then during shut-in the fracture first closes on the impermeable

layers and subsequently closes on the permeable layers.

Figure 4.13- G-Plot – Fracture Height Recession (from Craig44)

During shut-in initially the pressure declines at a constant rate which is

relatively low, this happens because only the permeable area causes fluid leakoff in

the entire fracture. Subsequently the impermeable fracture area starts closing (height

recession), during this period rate of pressure decline increases. At the end the

fracture consists of only permeable area, during this period the rate of pressure

decline is once again constant (normal leakoff) but at a higher value than the initial

leakoff. The three characteristic signatures for height recession during shut in are44:

57
dP
1. The superposition derivative G curve lies below a straight line
dG

extrapolated through the normal leakoff data.

2. Concave down pressure curve.

dP
3. Increasing pressure derivative curve .
dG

4.14 Summary of G-function

The theory and equations pertaining to the G-function have been discussed in

detail in the previous sections. The same can be summarized as given below:

G-Function is a function of:

ƒ Fracture Area Growth Exponent

ƒ Dimensionless Time

ƒ Dimensionless Area Parameter

ƒ Fluid Loss Exponent

Some of the important equations pertaining to the G-function which have been

discussed in detail in the previous sections have been summarized below:

α
a τ 
=   ……………………………....…………...……….………...……........ (4.1)
A t

t − tp ∆t
∆t D = = …………………….….……..…………………...……......……(4.6)
tp tp

a t
ξ= and tαD = ……………...….......………………….……...……......…...(4.7)
Af tp

1 θ
g (∆t D , α , θ ) = ∫ 1 + ∆t D − ξ α  dξ
1 1
∆t D > 0 ………………….......…….(4.9)
θ 0 

58
1 θ
1  α 
g (∆t D = 0, α , θ ) = g 0 (α ,θ ) =
1

θ ∫0 
 1 − ξ  dξ ∆t D = 0 ……..............(4.10)

4
G (∆t D ) = [g (∆tD ) − g0 ]……………..............…………...……….…….…........ (4.38)
π

59
Chapter 5

Breakdown Tests and Simulations

5.1 Test Methodology and Data Analysis

A total of seven wells were considered in this study. Figure 5.1 shows a

typical well casing diagram for the wells under consideration. The details about

casing and perforation depths for each of the seven wells are given in Appendix A.

20" set @ 100'

13 3/8" set @ 600'

9 5/8" set @ 5000'

4 1/2" Casing, P-110, 11.6 ppf


up to 11,000' to 11,500'

4 1/2" Casing, P-110, 13.5 ppf


set @ 12,000' to 13,000'

Figure 5.1 Typical Well Casing Diagram

60
A pump-in diagnostic test was carried out for seven wells under consideration.

All of these wells are located in the Hammon field in western Oklahoma. These

wells are gas wells producing from the Red Fork formation. The details regarding

the diagnostic pump-in tests (breakdown tests) performed on each of the wells and

the simulation results are given in the following sections. The breakdown tests

were performed using KCl treated water. The well was shut-in and the pressure

decline was observed for approximately two to three hours after breakdown. The

pressures during the shut-in period were monitored and recorded.

The breakdown test data recorded consists of the following:

• Time in one second interval

• Surface pressure in psi

• Fluid flow rate in bpm

The data obtained from the service company was in an ASCII (American Standard

Code for Information Interchange) text format and an ASCII Excel comma separated

value spreadsheet format. The time interval of the data recorded was one second;

therefore data recorded over a period of four hours had approximately 14,400 rows of

data per well. The first step done was to ensure that the data for every breakdown test

was complete, evenly spaced at one-second-time interval and was free of errors. Data

for a total of ten wells were analyzed. Three wells which had missing data or out of

range readings were not considered in this analysis. The fracturing simulator requires

that the data file be entered in a binary format. Commercially available data

conversion software was utilized to convert the ASCII format files into binary format

(database files).

61
5.2 Simulation Model Options

For each well an input file for the fracture simulator was created and the options

selected for the simulations are summarized in Table 5.1 and subsequently

discussed in detail:

Table 5.1: Simulator Model Options

Model/Option Selection for Selection for

Simulations Simulations

(Set 1) (Set 2)

Fracture Model PKN GDK

Leakoff Model Lumped Parameter Lumped Parameter

Backstress Option Backstress Ignore Backstress Ignore

Fracture Growth Option Allow Growth after Shut- Allow Growth after

in Shut-in

Fracture Orientation Vertical Vertical

Option

Heat Transfer Option Heat Transfer Ignore Heat Transfer

Ignore

Iteration Option General Iteration General Iteration

Fracture Model: The options available for the fracture model are PKN,

GDK, Radial, Tip Dominated and Conventional. The tip-dominated model predicts

very high net pressures; radial and conventional models predict low net pressures47.

The assumptions of the PKN model are that the height of the vertical fracture is

62
constant and does not exceed the pay zone19. The cross section of the fracture is

assumed to be elliptical. At any cross section the maximum width is proportional to

the net pressure at that point and independent of the width at any other point19. The

fracture width at any point is proportional to its height. The fracture model utilized

for the simulator for the first set of simulations was ‘PKN’ as the assumptions of the

PKN model are more appropriate for breakdown tests. Additionally field experience

with the breakdown tests have proven that the PKN fracture model for the simulator

gives reliable results in comparison to the other models available.

The simulations were repeated for all the wells with the GDK fracture model

for comparison with the results obtained from the PKN model.

Leakoff Model: The two models available for leakoff were ‘lumped

parameter’ and ‘grid based’. The ‘lumped-parameter’ leakoff model was selected for

the simulations. This model is formulated for speedy computations and is accurate in

low permeability reservoirs. The grid-based leakoff models are slow and are more

suited for reservoirs with high permeability (atleast 100 mD) and reservoirs having

large variations in permeability47.

Backstress Option: There are two options available; ‘backstress model’ and

‘backstress ignore’. Reservoirs having a high fluid leakoff increase the pore pressure,

which results in a change in formation closure stress known as backstress. For gas

reservoirs selecting the backstress model option complicates and slows the

computations. The Red Fork formation being a low permeability gas reservoir has

relatively low fluid leakoff, due to this reason the ‘backstress ignore’ option was

utilized for the simulations47.

63
Growth After Shut-In Option: Accurate interpretation of the pressure-

decline requires that growth after shut-in be considered47. Among the two options

‘freeze dimensions’ and ‘allow growth after shut-in’ the later option was selected for

the simulations.

Fracture Orientation Option: Among the two options (‘vertical’,

‘horizontal’) for fracture orientation the ‘vertical fracture orientation’ was selected for

the simulations as vertical orientation is the most commonly occurring fracture

orientation.

Heat Transfer Effect Option: For Heat transfer effects there are two options

available; ‘heat transfer model’ and ‘heat transfer ignore’. The volume of the fluid

pumped in a breakdown test is very small as compared to the main fracture treatment.

Neglecting the heat transfer effects will increase the computation speed and not have

any significant change in the accuracy of the results. The ‘ignore heat transfer’

option was selected for the simulations.

Iteration Option: Among the two options of ‘general’ and ‘simplified’

iteration, the ‘general’ iteration option was selected for the simulations. Simplified

iterations are much faster but the results obtained are less accurate. The ‘general

iteration’ option takes longer computation time but the results are more accurate and

reliable.

The next step performed was to input the binary format database file into the

simulator. Subsequently the wellbore configuration was created for the simulations.

The data entered to complete the wellbore configuration are as follows:

• Drilled Hole Data (various hole sizes and corresponding depth)

64
• Casing Data (various casing strings diameter, corresponding setting

depths, casing grade, casing weight and casing inner diameter)

• Survey Data (inclination and azimuth)

• Perforation Data (depth, diameter of perforations, number of perforations)

5.3 Log Analysis and Reservoir Layers

Logs for the seven wells were analyzed to determine the lithology of the

formation. Gamma ray and neutron porosity logs were analyzed. The estimated

lithology was used to create reservoir layers in the commercial fracturing software.

Figure 5.2 shows the gamma ray and neutron porosity logs for Well – 4. For every

reservoir layer a shale index Ish is calculated from the log using Eq. 5.1:

I sh = (γ log − γ min )/ (γ sh − γ min ) ………………………………………………(5.1)

where γlog is the log response for the layer being evaluated, and γsh and γmin are the log

responses in shales and zones of minimum radioactivity48.

65
Feet
0 Gamma Ray (gr) 130 30 Neutron Porosity (cnl) -10
(API) (pu)

Figure 5.2 Log for Well-4 (gamma ray and neutron porosity)49

Table 5.2 summarizes the values of γlog for the individual layers and the

corresponding values of the shale index Ish calculated using Eq. 5.1. The log has a

scale of 0-130 API gamma ray units. The values of γsh and γmin are 110 and 52 API

gamma ray units. A lithology of shale, sandy-shale, shaley-sand and sandstone was

modeled for Ish ranges of 0.9 - 1, 0.6 – 0.89, 0.25 – 0.59 and 0 – 0.24 respectively.

66
Table 5.2: Shale Index of Well-4 Reservoir Layers

Layer top Layer γlog Ish Lithology

depth (feet) thickness (feet) API

units

104 0.90 Shale


12600 26 98 0.79 Sandy-Shale
12626 22 65 0.22 Sandstone
12648 7 78 0.45 Shaley-Sand
12655 10 57 0.09 Sandstone
12665 6 96 0.76 Sandy-Shale
12671 18 59 0.12 Sandstone
12689 4 72 0.34 Shaley-Sand
12693 22 58 0.10 Sandstone
12715 8 78 0.45 Shaley-Sand
12723 12 58 0.10 Sandstone
12735 41 91 0.67 Sandy-Shale
12776 7 64 0.21 Sandstone
12783 32 90 0.66 Sandy-Shale
12815 10 99 0.81 Sandy-Shale
12825 105 0.91 Shale

From Table 5.2 it can be concluded that the total thickness of all the sandstone

layers (t1) and all the non-sandstone (shaly-sand and sandy-shale) layers (t2) for Well-

4 are 91 and 92 feet respectively. The ratio of t1 to t2 for Well-4 is 0.98.

The depth and the rock type (shale, sandstone, shaly-sand, sandy-shale etc.)

were entered into the simulator for all layers of the reservoir. Figure 5.3 shows the

reservoir layers built for Well-4.

67
Figure 5.3 Reservoir Layers built from Logs for simulations (Well-4)

Following similar procedures layers were built for six other wells. The

lithology inferred in these wells is shown in Appendix C. Additionally, the total

thickness of sandstone layers (t1) and the total thickness of non-sandstone layers (t2)

for all other wells were calculated. Table 5.3 summarizes the ratio of t1 to t2 for the

seven wells. Wells 2, 5 and 7 indicate a very high thickness of sandstone layers as

compared to the sum of shaley-sand and sandy-shale layers. The Well production

data given in Appendix B indicates that the cumulative production for Wells 2, 5 and

7 were the highest among the seven wells.

Table 5.3: Values of t1/t2 for lithology of seven wells

Well t1 / t2

1 1.5
2 5.1
3 0.29
4 0.99
5 5.71
6 0.35
7 6.59

68
5.4 Simulator Input

The type of fluid and proppant need to be specified prior to running the simulator.

Potassium chloride treated water without proppant was specified for the

simulations.

The treatment schedule details that were specified for the simulations were the

stage type (water injection stage, shut-in stage etc), stage length in minutes and the

type of fluid for water injection stage. The treatment schedule information entered

was obtained from the service company breakdown report and was verified with the

ASCII format breakdown data file to ensure accuracy of simulations.

The final simulation input that need to be specified to the simulator are the

start time, end time and the time step for performing calculations. Start time is

selected in order to skip over any initial glitches in pump rate or data that is collected

prior to actually starting the job. The end time selected is usually a few minutes more

than all the stages of the treatment schedule. The start time and end time are obtained

from the test reports and verified from the ASCII format data files to eliminate

possible errors in simulations. The time step determines the frequency of calculations

performed by the simulator. The time step recommended47 for conducting Minifrac

analysis is between 0.1 and 0.5 minute. The time step selected for performing the

simulations was 0.05 minute.

5.5 Breakdown Test Analysis

Simulations were executed and from the results a breakdown test analysis was

performed for each well. The simulator calculates several parameters from the

69
database file to create plots required for breakdown test analysis; the methodology is

as given below:

Instantaneous Shut-In Pressure Plot (ISIP): This is a plot of surface

pressure and bottomhole pressure versus time. The surface pressure and time are

obtained from the database file which was input to the simulator. The bottomhole

pressure during the shut-in period is calculated by using Eq. 5.2:

p w = (0.052 * Z * ρ ) + p s …………………………………………………..(5.2)

where pw is the bottomhole pressure in psi, Z is the depth from the surface to the

center of the perforations in feet, ρ is the density of the KCl treated water in ppg and

ps is the measured surface pressure in psi. A vertical line is drawn on the plot at the

time when pumping stopped; this vertical line is the beginning of the shut-in decline

period. The ISIP plots in Chapter 6 and Appendix D are drawn to have the Y-axis as

the end of pumping which is also the beginning of shut-in decline period. A tangent

is drawn to the bottomhole pressure decline curve at the start of the decline. The ISIP

should be the reflection of the pressure in the main body of the fracture. Near

wellbore effects make it difficult to pick the true ISIP. The tangent should be

positioned in such a way that the initial effects have dissipated. The intersection of

the tangent with the end of the injection line (Y-axis for plots in Chapter 6 and

Appendix D) gives the ISIP.

Square-root Plot: The square-root plot shows surface pressure, bottomhole

pressure and the derivative of the surface pressure plotted versus the square root of

time. The scale on the ‘X-axis’ of this square-root plot is non-linear similar to the

non-linear scale of a log-log plot. As previously discussed in Chapter 2, a change in

70
slope of the bottomhole pressure on a square-root time plot indicates fracture closure.

The surface area available for fluid leak-off changes at fracture closure and this

results in a change of slope on the pressure decline curve. A tangent is drawn to the

bottomhole (BH) pressure curve; the point at which the BH pressure curve starts to

deviate from the tangent is the bottomhole closure pressure. The time at this point is

the closure time. The surface pressure for the plot is obtained from the database file;

the bottomhole pressure is calculated using Eq. 5.1. The simulator calculates the

values of dps/dt (derivative of surface pressure with respect to time).

G-Function Plot: G-Function plot consists of the surface pressure,

bottomhole pressure, derivative of the surface pressure and the superposition

derivative of the surface pressure (G dP/dG) versus the G-function time. A tangent is

drawn to the superposition derivative of the surface pressure (G dP/dG) curve passing

through the origin; the point at which the curve starts to deviate from the tangent is

the closure point. The G-function time (dimensionless) at this point is the G-function

time at closure. The bottomhole pressure at this closure point (G-function time at

closure) is the bottomhole closure pressure.

The equations utilized to generate the G-function plot are as follows:

• The values of t (instantaneous time) and tp (end of injection time) are

obtained from the breakdown test data. Equation 4.6 is utilized to calculate

the ∆tD (dimensionless time)

• A value of α = 4/5 is assumed for the simulations with the PKN fracture

model and α = 2/3 is assumed for the simulations with the GDK fracture

model. As discussed in Chapter 3 for Newtonian filtrates the lower and

71
upper bounds for the fracture area growth exponent α are ½ and 1. The fluid

used for the breakdown tests being KCl treated water implies that ½<α<1.

Additionally it has been proven that for the PKN fracture model the value of

α = 4/5 is a very good approximation50 . Similarly the value of α = 2/3 is a

very good approximation for the GDK model. Equation 5.3 is utilized for

estimating the G-function50 for the simulations performed with the PKN

fracture model.

2 3
g (4 / 5, ∆t D ) = 1.415 + 1.376∆t D − 0.388∆t D + 0.097∆t D
4 5 6
……………….…(5.3)
− 0.01407∆t D + 0.00105∆t D − 0.000031∆t D

Equation 5.3 is a very good approximation to obtain the G-function values for a PKN

fracture model with KCl treated water as the fracturing fluid (without proppant) used

for the breakdown tests. The value of g0 = g(4/5, ∆tD) = 1.415 for PKN fracture

model.

The G-function for the simulations performed with the GDK fracture model is

obtained31 by substituting the value of α = 2/3 in Eq. 5.4.

 1 −1  
4α ∆t D + 2 1 + ∆t D  F  , α ; 1 + α ; (1 + ∆t D )  
g (α , ∆t D ) =  2 
……........…........ (5.4)
1 + 2α

The hypergeometric function F [a, b; c; z ] is available in tabular form32 or through

computing algorithms.

72
Chapter 6

Simulation Results and Inferences

6.1 Interpretation of Simulation Plots

The ISIP and square-root plots for Well-1 are shown in Figs. 6.1 and 6.2

respectively. The G-function plots for Well-1 with PKN and GDK fracture

models are shown in Figs 6.3 and 6.4 respectively. Similar plots for Well-2

through Well-7 are illustrated in Appendix D as Figs. D1 through D32. The

instantaneous shut-in pressure, bottomhole closure pressure and the G-function at

closure for each well is estimated from the simulation plots utilizing the procedure

described in Chapter 4.

Data obtained from simulation plots for PKN and GDK fracture models are

summarized in Tables 6.1 and 6.2 respectively. The average of the closure pressure

obtained from the square-root plot and the G-function plot is denoted by ‘Average

pc’. The closure time is denoted by ‘tc’ and the dimensionless G-function at closure is

denoted by ‘Gc’. Well-2 has three zones; the break down test for each of these zones

was performed separately; the values of ‘average pc’, ‘tc’ and ‘Gc’ given for Well-2 in

Table 6.1 and Table 6.2 are the weighted average values for all the three zones.

73
Meas'd Btmh (psi) Surf Press [Csg] (psi)
10000 5000

ISIP = 9,632 psi


Tangent to Bottomhole
9600 4600
pressure curve

9200 4200

8800 3800

8400 3400

8000 3000
22.0 42.2 62.4 82.6 102.8 123.0
Time (min)

Figure 6.1- ISIP Plot for Well – 1

(d/dt) Surf Press [Csg] (psi)


200 Meas'd Btmh (psi) Surf Press [Csg] (psi)
5000
10000

120
9600 4600

40 4200
9200

-40
3800
8800

-120
8400 3400

-200
3000
80000.0 20.0 40.0 60.0 80.0 100.0
Time (min)

Figure 6.2- Square-root Plot for Well – 1

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
2000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 2000
10000 5000

1600 1600
9600 4600

1200 1200
9200 4200

800 800
8800 3800

400 400
8400 3400

0 0
8000 0.000 0.780 1.560 2.340 3.120 3.900 3000
G Function Time

Figure 6.3- G-Function Plot (PKN Fracture Model) for Well – 1

74
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
2000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 2000
10000 5000

1600 1600
9600 4600

1200 1200
9200 4200

800 800
8800 3800

400 400
8400 3400

0 0
8000 0.000 0.800 1.600 2.400 3.200 4.000 3000
G Function Time

Figure 6.4- G-Function Plot (GDK Fracture Model) for Well – 1

Table 6.1: Data from Simulation Results (PKN Fracture Model)

Well ISIP (psi) Average pc (psi) tc (min) Gc

1 9,632 9,190 44.7 2.45

2 11,100 10,789 62.71 2.36

3 10,010 9,064 101.7 4.30

4 9,609 9,536 24.6 3.96

5 10,430 9,039 98.1 3.38

6 10,513 9,954 101.4 5.00

7 10,345 9,812 74.4 1.70

75
Table 6.2: Data from Simulation Results (GDK Fracture Model)

Well ISIP (psi) Average pc (psi) tc (min) Gc

1 9,632 9,223 40.17 2.13

2 11,100 10,805 40 1.60

3 10,010 9,111 84.2 3.99

4 9,609 9,547 21.7 4.03

5 10,430 9,083 89.9 3.19

6 10,513 9,968 100 4.85

7 10,345 9,830 75.2 1.74

6.2 Inferences from Simulation Plots

ISIP Plots: The average rate of pressure decline during the shut-in period can

be obtained from the ISIP plot (the total pressure drop divided by the shut-in period

time). Table 6.3 summarizes the average rate of pressure decline for seven wells

obtained from the ISIP plots (Figs. 6.1 and D1 through D8). Well-2 has the highest

rate of pressure decline (44 psi/min) and Well-4 has the lowest rate of pressure

decline (1.42 psi/min). The rate of pressure decline is an indication of the rate of

fluid leakoff.

76
Table 6.3: Average rate of pressure decline

Well Average rate of pressure

decline (psi/min)

1 6.32

2 44

3 6.824

4 1.42

5 10

6 5.06

7 6.7

As discussed in Chapter 2 the fluid leakoff is characterized by the spurt-loss

coefficient and the total fluid loss coefficient. Potassium Chloride treated water does

not build a filter cake which implies the spurt-loss is not applicable. The total fluid-

loss coefficient depends on wall building fluid loss coefficient, viscosity controlled

fluid loss coefficient and the compressibility controlled fluid loss coefficient. For

KCl treated water the only factors that affect the fluid loss are viscosity controlled

fluid loss coefficient and the compressibility controlled fluid loss coefficient. The

factors that affect the fluid leakoff rate in the break down tests with KCl treated water

are as follows:

1. Relative permeability of the formation to the fracturing fluid filtrate.

2. Pressure difference (pressure inside the fracture minus the reservoir pressure).

3. Viscosity of the fracturing fluid.

77
4. Permeability of the formation to the mobile formation fluids.

5. Viscosity of the formation fluid.

6. Total system compressibility for the reservoir, ct.

The viscosity of the fracturing fluid and the formation fluid being similar for

all the wells are not the contributing factors for the difference in the rate of pressure

decline. Relative permeability of the formation to the fracturing fluid and the

formation fluids are the most significant factors affecting the leakoff rate. A high rate

of fluid leakoff is an indication of high formation permeability which implies a high

probability that the well will be a good producer. Production data in Appendix B

indicates that Well-2 had the highest cumulative production in the first sixty-day

period as compared to all other wells. Well-6 had the lowest cumulative production

among the seven wells; the average rate of pressure decline for Well-6 was observed

to be 5.06 psi/min which was the second lowest among the seven wells.

Square-Root Plots: Closure pressure is determined from the square-root plot.

Occasionally a change of slope cannot be observed on the bottomhole pressure curve.

Alternatively there could be more than one change in slope and the change in slope

could be due to either fracture closure, height recession from bounding layers or from

fracture extension/recession transition.

In situations where the closure point cannot be easily identified, the pressure

derivative curve is utilized. The pressure derivative magnifies the changes in

pressure and helps in identifying the closure point. In case of Well-2 (Zone-2) the

closure point is easily identified by the change of slope in the bottomhole pressure

curve (Fig. D10). In case of Well-2 (Zone-1) the change of slope of the bottomhole

78
pressure curve cannot be easily identified (Fig. D9). The pressure derivative curve

helps to identify the closure point in Fig. D9.

A combination of the square-root plot and the G-function plot can help

eliminate the ambiguity in determination of the closure pressure. An example where

the G-function plot is utilized to identify the closure point is Well-4. In Fig. D13 the

closure point cannot be identified from both the bottomhole pressure curve as well as

the pressure derivative curve. The G-function plot for Well-4 (Fig. D21) is utilized to

identify the closure point.

The closure pressure obtained from the square-root plot differs slightly from

the closure pressure obtained from the G-function plots. Well-2 (Zone-2) has a high

pressure derivative slope indicating a high rate of fluid leakoff as seen in Fig. D10.

The cumulative production of Well-2 was the highest among all the wells.

G-Function Plots (PKN Model): The characteristic signatures of the G-

function plot are utilized to determine the leakoff mechanism as mentioned in Chapter

4. The characteristic signatures for a normal leakoff, pressure dependent leakoff,

fracture tip extension and fracture height recession were elaborated earlier in Chapter

4. Analysis of Figs. 6.3 and D17 through D19 for Well - 1 and Well – 2 (Zones 1, 2

and 3) indicate a normal leakoff as the pressure derivative (dP/dG) is constant and the

superposition derivative (G[dP/dG]) lies on a straight line that passes through the

origin. Analysis of Figs. D20 through D22 for Well–3 through Well-5 respectively

indicate a pressure dependant leakoff due to the following reasons:

1. A large hump in the superposition derivative (G dP/dG).

2. Subsequent to the hump, the pressure decline exhibits a normal leakoff.

79
3. The portion of normal leakoff lies on a straight line passing through the origin.

4. The hump lies above the straight line that passes through the origin.

G-Function Plots (GDK Model): Analysis of Figs. 6.4 and D25 through

D27 for Well 1 and Well – 2 (Zones 1,2 and 3) indicate a normal leakoff. Analysis of

Figs. D28 through D30 for Well–3 through Well-5 respectively indicate a pressure

dependant leakoff.

6.3 Comparison of PKN model results with GDK Model

The ISIP and the square-root plots for both the models are identical as these plots

depict the pressure decline data, which is recorded at the surface and is

independent of the fracture model selected.

Well-1 has a normal leakoff with PKN as well GDK fracture models, similarly

Well-4 has a pressure dependent leakoff in both the models. For every well the

G-function plots for the simulations performed with the PKN model and GDK

model indicate an identical leakoff mechanism. The reasoning for this similarity

is that the leakoff mechanism (normal or pressure dependent) is controlled by the

reservoir properties and is independent of the geometry of the fracture created.

Pressure dependent leakoff occurs due to opening of natural fissures and is an

inherent property of the reservoir.

The closure pressure and closure time obtained from the G-function plots for the

PKN model are slightly higher than those obtained from the G-function plots of

the GDK model. The reasoning for this phenomenon is that the PKN model

predicts fractures with more length and less width in contrast to the GDK fracture

80
model that predict fractures with less length and more width. The actual range of

the fracture geometry as generated by the simulator for PKN and GDK models

were as follows:

PKN Model:

Fracture half length range: 190 – 210 feet

Fracture width range: 0.04 – 0.08 inch

GDK Model:

Fracture half length range: 85 – 110 feet

Fracture width range: 0.11 – 0.15 inch

The range of the fracture half length generated with the PKN models is almost

twice that of GDK model; whereas the range of fracture width generated for GDK

model is approximately 50 percent more than the PKN model.

81
Chapter 7

Well Productivity Correlations

7.1 Well Productivity Factors

The productivity of a well depends on several factors; some of the major ones

are listed as follows:

• Reservoir Pressure

• Well Stimulation (Acidizing/Fracturing)

• Height of Producing Interval

• Porosity

• Permeability

All seven wells considered in this study are in the Hammon field, which is

located in two adjoining counties namely Roger Mills and Custer in western part of

Oklahoma. Figure 7.1 shows the location of the Counties.

Figure 7.1 County Map of Oklahoma State


(Source: http://www.state.ok.us/osfdocs/staemap.html)

82
Subsequent to the diagnostic injection test these wells have been stimulated

and all of them have a very similar fracture stimulation job performed. The

producing formation for all of the above wells is ‘Red Fork’ and the heights of the

producing intervals are given in Table 7.1. Production data for the seven wells are

provided in Appendix B. The 30-day (Q30) and 60-day (Q60) cumulative gas

production for the seven wells is summarized in Table 7.2.

The 60-day cumulative production data for Well-3 was not available, as this

well has been completed recently. Due to this reason Table 7.2 and Figs. 7.3, 7.5,

7.7, 7.9 and 7.11 have one data point less corresponding to the cumulative production

of the sixty-day period.

Table 7.1: Height of Producing Interval for Each Well

Well Height of producing

Interval (feet)

1 130

2 16 (Zone 1)

2 52 (Zone 2)

2 32 (Zone 3)

3 60

4 90

5 200

6 38

7 128

83
Table 7.2: Well Production Data

Well Q30 Q60


(Mscf) (Mscf)
1
42200 70500
2
71500 128653
3
25000 n/a
4
38890 71600
5
56210 116872
6
17337 27253
7
69891 154976

7.2 Statistical Analysis and Correlations for PKN Fracture Model

A statistical analysis of the data in Tables 6.1, 7.1 and 7.2 indicated that no

correlation exists between the well production data and three other variables ISIP,

closure pressure pc and closure time tc. The data, however indicates a correlation of

the well productivity with the G-function at closure.

Figures 7.2 and 7.3 are log-log plots of the height of the producing interval of

each well in feet versus the cumulative production in MCF for the first thirty-day and

sixty-day period respectively. A good correlation could not be established that fitted

all the points on Figs. 7.2 and 7.3. The reason for this is attributed to the variation of

reservoir properties as can be seen in the lithology of the layers in Figs. 5.3 and C1

through C8. The ratio of the total thickness of sandstone layers to the total thickness

of the non-sandstone layers was calculated and is summarized in Table 5.3. Wells 2,

5 and 7 have the highest values of the ratio t1/t2 which is in the range of 5.1 to 6.59.

84
The corresponding values of the ratio t1/t2 for the Wells 1, 3, 4 and 6 were in the

range of 0.29 to 1.5. The production data in Table 7.2 indicates that Wells 2, 5 and 7

have the highest cumulative production. The higher value of the ratio t1/t2 for the

Wells 2, 5 and 7 is considered as one of the reasons for higher well productivity.

As discussed in Chapter 6 one of the reasons for high cumulative production

of a well is the high rate of pressure decline during the shut-in period. Analysis of the

data in Table 6.2 (Rate of Pressure Decline) indicates that Well-2 has the highest rate

of pressure decline (44 psi/min). A high rate of pressure decline indicates a high rate

of fluid leakoff. The high rate of fluid leakoff implies that the permeability of the

well is high. High permeability is considered as one of the reasons for the high

cumulative production of Well-2.

A detailed review of the logs of the seven wells indicated that the porosity of

Well-7 to be the highest and was in the range of 11 to 12.5 percent. For all other

wells the porosity was found to be in the range of 8 to 10 percent. The high

cumulative production of Well-7 can be explained due to the high porosity.

Wells 2 and 7 have the highest cumulative production and the corresponding

points for these wells lie prominently apart from all the other points in Figs. 7.2 and

7.3. Wells 1, 3 and 4 through 6 when considered as a group show a good correlation

on the plots of Figs. 7.2 (96.79 percent fit) and 7.3 (94.89 percent fit). The curve on

Figs. 7.2 and 7.3 was fitted with the points corresponding to Well 1, 3 and 4 through

6.

85
100000

2 7

5
30 Day Cumulative Production (MCF)

1
0.7052 4
y = 1407.8x
2
R = 0.9679

10000
10 100 1000
Height of Producing Interval (feet)

Figure 7.2 Plot of 30-Day Cumulative Production versus


Height of Producing Interval

1000000
60 Day Cumulative Production (MCF)

7
0.841
y = 1349.8x 2
2
R = 0.9489
5
100000

4 1

10000
10 100 1000
Height of Producing Interval (feet)

Figure 7.3 Plot of 60-Day Cumulative Production versus


Height of Producing Interval

Figures 7.4 and 7.5 are semi-log plots of the G-function at closure of each

well versus the cumulative production in MCF for the first thirty-day and sixty-day

period respectively. It can be seen that in general the G-function at closure (PKN

86
Model) is inversely proportional to the thirty-day and sixty-day cumulative

production.

100000

7 2

5
30 Day Cumulative Production (MCF)

1 4

6
2
y = -2161.6x - 1177.2x + 76015
2
R = 0.7927

10000
0.00 1.00 2.00 3.00 4.00 5.00 6.00
G-Function at Closure

Figure 7.4 Plot of 30-Day Cumulative Production versus


G-function at Closure (PKN Model)

1000000
60 Day Cumulative Production (MCF)

7
2
5

100000
4

2
y = -1822.9x - 19837x + 177507
2
R = 0.6936

10000
0.00 1.00 2.00 3.00 4.00 5.00 6.00
G-Function at Closure

Figure 7.5 Plot of 60-Day Cumulative Production versus


G-function at Closure (PKN Model)

87
Figures 7.6 and 7.7 are semi-log plots of the Height/G-function at closure for

PKN model of each well versus the cumulative production in MCF for the first thirty-

day and sixty-day period respectively.

100000

7 2

5
30 Day Cumulative production (MCF)

4 1

0.5603
y = 5770.1x
6 2
R = 0.9459

10000
1 11 21 31 41 51 61 71 81 91
(Height/G-function at Closure)

Figure 7.6 Plot of 30-Day Cumulative Production versus


H/(G-function at Closure-PKN Model)

1000000
60 Day Cumulative production (MCF)

5 2
100000

4
1

0.6499
y = 7710.6x
2
R = 0.8892

10000
0 10 20 30 40 50 60 70 80 90
(Height/G-function at Closure)

Figure 7.7 Plot of 60-Day Cumulative Production versus


H/(G-function at Closure-PKN Model)

88
Equation 7.1 gives a fit of 94.59 percent as seen from Fig. 7.6:

(0.5603 )
H 
Q30 = 5770.1  ……………………………………………………..……..(7.1)
 Gc 

Equation 7.2 gives a fit of 88.92 percent as seen from Fig. 7.7:
(0.6499 )
H 
Q60 = 7710.6  ……………………………………………………………(7.2)
 Gc 

where Q30 and Q60 is the cumulative production in MCF of gas produced for the first

30-day and 60-day periods respectively, H is the height of the producing interval in

feet and Gc is the G-function at closure (dimensionless) for PKN model.

Figure 7.6 indicates that there exists a very good correlation between the ratio

Height/G-function at closure for PKN model and thirty-day cumulative production.

Whereas Fig. 7.7 indicates that there exists a reasonably good correlation between the

ratio Height/G-function at closure for PKN model and sixty-day cumulative

production.

Wells 2 and 7 in Figs. 7.6 and 7.7 do not stand prominently apart as they did

in Figs. 7.2 and 7.3. The reason for this is that the G-function takes into account

properties of the formation.

7.3 Statistical Analysis and Correlations for GDK Fracture Model

Similar to the PKN fracture model a statistical analysis was performed

utilizing the data obtained from the simulations with the GDK fracture model shown

in Table 6.2. No correlation exists between the well production data and three other

variables ISIP, closure pressure pc and closure time tc. The simulation results confirm

89
that similar to the PKN model the GDK model indicates a correlation of the well

productivity with the G-function at closure. Figures 7.8 and 7.9 are semi-log plots of

the G-function at closure (GDK Model) of each well versus the cumulative

production in MCF for the first thirty-day and sixty-day period respectively. Similar

to the PKN Model it can be seen that the G-function at closure for the GDK Model is

inversely proportional to the thirty-day and sixty-day cumulative production.

100000

2
7

5
30-Day Cumulative Production (MCF)

2
y = -2185.9x - 1018.2x + 75790
2 6
R = 0.7922

10000
0 1 2 3 4 5 6
G-function at Closure

Figure 7.8 Plot of 30-Day Cumulative Production versus


G-function at Closure (GDK Model)

90
1000000

60-Day Cum. Production (MCF)

7
2

5
100000

1 4

6
2
y = -1862.1x - 19567x + 177110
2
R = 0.6927

10000
0 1 2 3 4 5 6
G-function at Closure

Figure 7.9 Plot of 60-Day Cumulative Production versus


G-function at Closure (GDK Model)

Figures 7.10 and 7.11 are semi-log plots of the Height/G-function at closure

for GDK model of each well versus the cumulative production in MCF for the first

thirty-day and sixty-day period respectively.

100000

0.5788
y = 5700.4x
2
R = 0.856 2 7

5
30-Day Cumulative Production (MCF)

1
4

10000
0 10 20 30 40 50 60 70 80
Height/G-function at Closure

Figure 7.10 Plot of 30-Day Cumulative Production versus


H/G-function at Closure (GDK Model)

91
1000000

0.6791
y = 7447.7x
2
R = 0.833
60-Day Cumulative Production (MCF)

7
2

5
100000

4
1

10000
0 10 20 30 40 50 60 70 80
Height/G-function at Closure

Figure 7.11 Plot of 60-Day Cumulative Production versus


H/G-function at Closure (GDK Model)

Equation 7.10 gives a fit of 85.6 percent as seen from Fig. 7.6:

(0.5788 )
H 
Q30 = 5700  ……………………………………………………..…...…..(7.3)
 Gc 

Equation 7.11 gives a fit of 83.3 percent as seen from Fig. 7.7:
(0.6791)
H 
Q60 = 7447.7   ……………………………………………………………(7.4)
 Gc 

where Q30 and Q60 is the cumulative production in MCF of gas produced for the first

30-day and 60-day periods respectively, H is the height of the producing interval in

feet and Gc is the G-function at closure (dimensionless) for GDK model.

The PKN model gives better correlations than those obtained with the GDK model.

This indicates that the fracture geometry actually created are better modeled with the

PKN fracture model. The reasoning for this is that the fluid utilized for the

breakdown tests is KCl treated water which is a low viscosity fluid. The length of the

92
fracture created is inversely proportional to the viscosity of the fracturing fluid;

therefore KCl treated water creates fractures that are longer and sammler in width.

This geometry is more suited with a PKN fracture model.

93
Chapter 8

Conclusions and Recommendations

8.1 Summary

1. Breakdown test data and post-frac production data for seven wells from Hammon

field in Western Oklahoma were analyzed for performing simulations.

2. Wireline logs of the above wells were analyzed and lithology layers created for

the reservoir.

3. Input files for the simulator were created from well data and lithology layers.

4. Breakdown test data and input files were utilized to perform simulations for PKN

and GDK fracture models.

5. Specialized plots were obtained from the simulations.

6. Statistical analysis was performed on the data obtained from these specialized

plots.

7. Correlations were developed to estimate the post-fracture well production from

simulation results obtained from the breakdown test data.

8.2 Conclusions

1. Pump-in diagnostic testing can be utilized to develop correlations for predicting

post-frac production response.

2. The cumulative production of the well was found to be proportional to the ratio of

height of the producing interval divided by the G-function at closure.

94
3. Simulations with a PKN fracture model give better results than those with a GDK

fracture model.

4. A high rate of pressure decline during the shut-in decline period indicates that the

well is likely to be very productive.

5. The probability of the well to be a good producer was found to be directly

proportional to the ratio of the total thickness of sandstone layers to the total

thickness of the non-sandstone layers in the pay zone.

8.3 Recommendations

1. The correlations developed to estimate well productivity should be used as a

guideline to estimate well productivity. Offset well productivity and field

experience are significant factors to be considered.

2. Other factors like porosity, permeability and reservoir pressure should be

taken into consideration in addition to the G-function at closure and the height

of the producing interval.

3. Data for additional wells should be considered to improve the reliability of the

correlations developed.

4. Permeability derived from pump-in data can be considered an additional

variable for the improving the correlations.

95
Nomenclature

a = intermediate area of one face of fracture, ft2

A = area of one face of fracture, ft2

Af = area of one face of fracture at the end of injection, ft2

AL = leakoff area, ft2

cf = fracture compliance

C = discharge coefficient of perforation

CL = leakoff co-efficient, ft/min1/2

Dp = diameter of each perforation, in

E = Young’s Modulus, psi

E’ = Plane strain modulus, psi

go = fluid-loss volume function

g (∆t D ) = dimensionless fluid-loss volume function

G (∆t D ) = dimensionless fluid-loss time function

Gf = average horizontal fracture gradient, psi/ft

hf = fracture height, ft

hp = permeable height of fracture, ft

K = consistency index for power law fluids, lbf-secn/ft2

L = fracture half length, ft

n = flow behavior index of power law fracturing fluid

N = number of perforations

nf = power law exponent of filtrate

96
Pc1 = breakdown pressure of the rock, psi

Pc2 = fracture propagation pressure, psi

Pi = wellbore fluid pressure, psi

Pnet = net pressure, psi

Pp = pore pressure, psi

Ps = instantaneous shut-in pressure, psi

∆pf = average net pressure, psi

∆p = total near wellbore pressure loss, psi

∆p pf = perforation friction pressure loss, psi

∆pt = tortuosity friction pressure loss, psi

pf = constant fracture internal pressure, psi

pc = closure pressure, psi

pw = bottomhole pressure, psi

pws = bottomhole pressure at shut-in, psi

PG = pore pressure gradient, psi/ft

qi = injection rate, bbl/min

qL = fluid leakoff rate, bbl/min

rp = ratio of surface area available for leakoff to total fracture

area

R = fracture radius, ft

Rf = radial fracture radius, ft

Sp = Spurt loss coefficient, gal/ft2

97
Sv = vertical stress gradient due to overburden, psi/ft

t = time, min

td = dimensionless time

te = time at end of injection, min

tf = time of fracture opening and initial fluid exposure, min

tp = injection time, min

-To = tensile strength of the rock, psi

-To’ = tensile strength of the rock in the vertical direction, psi

uL = leak off velocity, ft/sec

vL = specific fluid loss volume

Vfp = volume of fracture created at end of pumping, bbl

Vi = total volume of fluid injected, bbl

VL = cumulative fluid volume leaked off, bbl

VL,C = fluid loss volume (Carter’s component), bbl

VL,S = fluid loss volume (Spurt loss component), bbl

VLp = fluid loss volume (at the end of injection), bbl

VLs (∆t ) = fluid loss volume (during shut-in), bbl

w = fracture width, in

w = fracture width averaged over length and height, in

wmax = maximum fracture width, in

ww = fracture width at the wellbore, in

w = average fracture width, in

xf = fracture half length, ft

98
Z = depth, ft

α = fracture area growth exponent

σx = stress in ‘x’ direction, psi

σy = stress in ‘y’ direction, psi

σh = horizontal stress, psi

σv = vertical stress, psi

σr = radial stress, psi

σθ = tangential stress, psi

τ rθ = shear stress, psi

υ = Poisson’s ratio

γ = fracture gradient, psi/ft

αp = poro-elastic constant (range from 0 to 1)

µ = viscosity, cp

κ = spurt factor

τ = time, min

θ = fluid loss exponent

η = fracture treatment efficiency

β = fracture fluid pressure gradient factor

∆t D = dimensionless shut-in time

ξ = dimensionless area parameter

Γ(x ) = gamma function

99
Γ(α ) = Euler’s gamma function

ρ = density, lbm/ft3

100
References

1. Howard G.C., and Fast C.R.: “Hydraulic Fracturing,” SPE Monograph, Vol. 2, 8,
1970

2. Interstate Oil and Gas Compact Commission.: “Hydraulic Fracturing: A Primer,”


July, 1999

3. Cornell, F.L.: “Engineering Improvements for Red Fork Fracturing,” JPT (Feb.
1991) 132-137

4. Hickey, J. W., Brown, W. E., and Crittenden, S. J.: “The Comparative


Effectiveness of Propping Agents in the Red Fork Formation of the Anadarko
Basin”, SPE No. 10132, presented at the 56th Annual Fall Technical Conference
of SPE, San Antonio, Texas (Oct. 5-7, 1981)

5. Rose, R. E.: “New Techniques Lead to Better Completions in Red Fork and Other
Low-Permeability Formations”, SPE No. 17305, presented at the SPE Permian
Oil and Gas Recovery Conference, Midland, Texas (March 10-11, 1988)

6. Penny, G.S., Soliman, M.Y., Conway, M.W. and Briscoe, J.E.: “Enhanced Load
Water-Recovery Technique Improves Stimulation Results”, SPE No. 12149,
presented at the 1983 SPE Annual Technical Conference and Exhibition, San
Francisco, CA, (October 5-8, 1983)

7. Gulrajani S.N. and Nolte K.G.: “Fracture Evaluation Using Pressure Diagnostics,”
Chapter 9 of “Reservoir Stimulation, Third Edition – 2000, edited by
Economides, M.J. and Nolte, K.G.”

8. Smith, M.B.: “Stimulation Design for Short, Precise Hydraulic Fractures,” paper
SPE 10313, SPE Journal (June 1985), 371.

9. Nolte, K.G.: “Principles for Fracture Design Based on Pressure Analysis,” paper
SPE 10911, SPE Production Engineering (February 1988c) 3, No.1, 22–30.

10. Plahn, S.V., Nolte, K.G., Thompson, L.G. and Miska, S.: “A Quantitative
Investigation of the Fracture Pump-In/Flowback Test,” paper SPE 30504, SPE
Production & Facilities (February 1997) 12, No. 1, 20–27.

11. Smith, M.B. and Shlyapobersky, J.W.: “Basics of Hydraulic Fracturing” Chapter
5 of “Reservoir Stimulation, Third Edition – 2000, edited by Economides, M.J.
and Nolte, K.G.”

12. Carter, R.D.: “Derivation of the General Equation for Estimating the Extent of the
Fractured Area,” Appendix I of “Optimum Fluid Characteristics for Fracture

101
Extension,” Drilling and Production Practice, G.C. Howard and C.R. Fast, New
York, NY, USA, American Petroleum Institute (1957), 261–269.

13. Howard, G.C. and Fast, C.R.: “Optimum Fluid Characteristics for Fracture
Extension,” Drilling and Production Practice, New York, New York, USA,
American Petroleum Institute (1957) 24, 261–270. (Appendix by E.D. Carter)

14. Shah, S.N.: “Advanced Stimulation - PE 5423 – University of Oklahoma - Class


Lecture Notes”, Fall 2003

15. Mack, M.G., and Warpinski N.R.: “Mechanics of Hydraulic Fracturing” Chap 6
of “Reservoir Stimulation, Third Edition – 2000, edited by Economides, M.J. and
Nolte, K.G.”

16. Sneddon, I.N.: “The Distribution of Stress in the Neighbourhood of a Crack in an


Elastic Solid,” Proc., Royal Soc. London (1946) 187, Ser. A., 229–260.

17. Sneddon, I.N. and Elliot, A.A.: “The Opening of a Griffith Crack Under Internal
Pressure,” Quarterly of Appl. Math. (1946) 4, 262–267.

18. Khristianovich, S.A. and Zheltov, Y.P.: “Formation of Vertical Fractures by


Means of Highly Viscous Liquid,” Proc., Fourth World Pet. Congress, Rome
(1955) 2, 579–586.

19. Perkins, T.K. and Kern, L.R.: “Widths of Hydraulic Fractures,” paper SPE 89,
Journal of Petroleum Technology (September 1961) 13, No. 9, 937–949.

20. Geertsma, J. and de Klerk, F.: “A Rapid Method of Predicting Width and Extent
of Hydraulic Induced Fractures,” paper SPE 2458, Journal of Petroleum
Technology (December 1969) 21, 1571–1581.

21. Nordgren, R.P.: “Propagation of a Vertical Hydraulic Fracture,” paper SPE 3003,
SPE Journal (August 1972) 12, No. 8, 306–314.

22. Warpinski, N.R.: “Measurement of Width and Pressures in a Propagating


Hydraulic Fracture,” SPE Journal (February 1985) 25, No.1, 46-54.

23. Brady, B.H., Ingraffea, A.R. and Morales, R.H.: “Three-Dimensional Analysis
and Visualization of Wellbore and the Fracturing Process in Inclined Wells,”
paper SPE 25889, presented at the SPE Rocky Mountain Regional/Low
Permeabililty Reservoirs Symposium, Denver, Colorado, USA (April 12-14,
1993)

24. Gidley et al: “Recent Advances in Hydraulic Fracturing,” SPE Monograph Vol.
12, 1990 Edition, SPE, Richardson, Texas.

102
25. Thompson, J.W., and Church D.C.: “Design, Execution, and Evaluation of
Minifracs in the Field: A Practical Approach and Case Study”, SPE No. 26034,
presented at the Western Regional Meeting held in Anchorage, Alaska, USA (26-
28 May, 1993)

26. Warpinski, N.R.: “Hydraulic Fracturing in Tight, Fissured Media”, JPT (Feb.
1991) 146-209.

27. Castillo, J.L.: “Modified Fracture Pressure Decline Analysis Including Pressure-
Dependent Leakoff”, SPE No. 16417 presented at the 1987 SPE DOE Low
Permeability Reservoirs Symposium, Denver, May 18-19.

28. Parlar, M., Nelson, E.B., Walton, I.C., Park, E. and Debonis, V.: “An
Experimental Study on Fluid-Loss Behavior of Fracturing Fluids and Formation
Damage in High-Permeability Porous Media,” paper SPE 30458, presented at the
SPE Annual Technical Conference and Exhibition, Dallas, Texas, USA (October
22–25, 1995).

29. Chaveteau, G.: “Fundamental Criteria in Polymer Flow Through Porous Media
and Their Relative Importance in the Performance Differences of Mobility
Control Buffers,” Water Soluble Polymers, J.E. Glass (ed.), Advances in
Chemistry Series (1986), 213.

30. Meyer, B.R. and Hagel, M.W.: “Simulated Mini-Frac Analysis,” Journal of
Canadian Petroleum Engineering (September–October 1989) 28, No. 5, 63–73.

31. Valkó, P. and Economides, M.J.: “Fracture Height Containment with Continuum
Damage Mechanics,” paper SPE 26598, presented at the SPE Annual Technical
Conference and Exhibition, Houston, Texas, USA (October 3–6, 1993b).

32. Abramowitz, M. and Stegun, I.A. (eds.): Handbook of Mathematical Functions,


ninth edition, New York, New York, USA, Dover (1989).

33. Nolte, K.G.: “Determination of Proppant and Fluid Schedules from Fracturing
Pressure Decline,” paper SPE 13278, SPE Production Engineering (July 1986b)
1, No. 4, 255–265.

34. Nolte, K.G.: “Fracturing Pressure Analysis for Non-Ideal Behavior,” paper SPE
20704, presented at the SPE Annual Technical Conference and Exhibition, New
Orleans, Louisiana, USA (September 23–26, 1990); also in Journal of Petroleum
Technology (February 1991) 43, No. 2, 210–218.

35. Ayoub, J.A., Brown, J.E., Barree, R.D. and Elphick, J.J.: “Diagnosis and
Evaluation of Fracturing Treatments,” SPE Production Engineering (February
1992a), 39.

103
36. El-Rabba A.M., Shah S.N. and Lord D.L.: “New perforation Pressure-Loss
Correlations for Limited-Entry Fracturing Treatments,” SPE Production and
Facilities (February 1999), Vol. 14, No. 1, 63-71.

37. Cleary, M., Johnson, D., Kogsboll, H., Owens, K., Perry, K., de Pater, C., Stachel,
A., Schmidt, H. and Tambini, M.: “Field Implementation of Proppant Slugs to
Avoid Premature Screen-Out of Hydraulic Fractures with Adequate Proppant
Concentration,” paper SPE 25892, presented at the SPE Rocky Mountain
Regional/Low Permeability Reservoirs Symposium, Denver, Colorado, USA
(April 12–14, 1993).

38. Aud, W.W., Wright, T.B., Cipolla, C.L., Harkrider, J.D. and Hansen, J.T.: “The
Effect of Viscosity on Near-Wellbore Tortuosity and Premature Screenouts,”
paper SPE 28492, presented at the SPE Annual Technical Conference and
Exhibition, New Orleans, Louisiana, USA (September 25–28, 1994).

39. Nolte, K.G.: “Application of Fracture Design Based on Pressure Analysis,” paper
SPE 13393, SPE Production Engineering (February 1988a) 3, No. 1, 31–42.

40. Wright C.: “Rate step-down test analysis – a diagnostic for fracture entry,”
Sidebar 9E of “Fracture Evaluation Using Pressure Diagnostics,” Chapter 9 of
“Reservoir Stimulation, Third Edition – 2000, edited by Economides, M.J. and
Nolte, K.G.”

41. Nolte, K.G.: “Determination of Fracture Parameters from Fracturing Pressure


Decline,” paper SPE 8341, presented at the SPE Annual Technical Conference
and Exhibition, Las Vegas, Nevada, USA (September 23–26, 1979).

42. Barree, R.D. and Mukherjee, H.: “Determination of Pressure Dependent Leakoff
and Its Effect on Fracture Geometry,” paper SPE 36424, presented at the SPE
Annual Technical Conference and Exhibition, Denver, Colorado, USA (October
6–9, 1996).

43. Barree, R.D.: “A Practical Numerical Simulator for Three-Dimensional Fracture


Propagation in Heterogeneous Media,” paper SPE 12273, presented at the SPE
Reservoir Simulation Symposium, San Francisco, California, USA (November
15–18, 1983).

44. Craig D. P., Eberhard M. J. and Barree, R.D.: “Adapting High Permeability
Leakoff Analysis to Low Permeability Sands for Estimating Reservoir
Engineering Parameters,” paper SPE 60291, presented at the SPE Rocky
Mountain Regional/ Low Permeability Reservoirs Symposium, Denver, Colorado,
USA (March 12–15, 2000).

45. Rollins, K. and Hyden, R.E.: “Pressure-Dependent Leakoff in Fracturing—Field


Examples from the Haynesville Sand,” paper SPE 39953 presented at the 1998

104
SPE Rocky Mountain Regional/Low Permeability Reservoirs Symposium,
Denver, 5-8 April 1998.

46. Craig, D.P., et al.: “Case History: Observations From Diagnostic Injection Tests
in Multiple Pay Sands of the Mamm Creek Field, Piceance Basin, Colorado,”
paper SPE 60321 presented at the 2000 SPE Rocky Mountain Regional/Low
Permeability Reservoirs Symposium, Denver, Colorado, 12-15 March 2000.

47. FracProPT 10.2 User’s Guide, 2003.

48. Bassiouni, Z.: “Theory, Measurement and Interpretation of Welllogs”, SPE


Textbook Vol. 4, 1994 Edition, SPE, Richardson, Texas.

49. Well Drilling Data Files, Crawley Petroleum Corporation, 800 N. Hudson, High
Tower Building, Oklahoma City, OK, 2003.

50. Craig D.P. and Brown T.D.: “Estimating Pore Pressure and Permeability in
Massively Stacked Lenticular Reservoirs Using Diagnostic Fracture Injection
Tests,” paper SPE 56600 presented at the 1999 SPE Annual Technical Conference
and Exhibition, Houston, Texas, 3-6 October 1999.

51. Well Production Data Files, Crawley Petroleum Corporation, 800 N. Hudson,
High Tower Building, Oklahoma City, OK, 2003.

105
Appendix A

Well Casing and Perforation Data49

Table A1: Well Configuration Data for Well – 1

Well TD 13,100’

Casing Surface to – 11,556’, 4.5” csg, 11.6 ppf

11,556’ to – 13,093, 4.5” csg, 13.5 ppf

Perforations Total 260 perforations of 0.35” - 2 SPF

Perforated Intervals 12635’ to 12,765’

Table A2: Well Configuration Data for Well – 2

Well TD 13,100’

Casing Surface to 11,574’ -- 4.5”, 11.6 ppf

11,574’ to 13,098’ – 4.5”, 13.5 ppf

Perforations Total 202 perforations of 0.35”

Perforated Intervals 12,462’ – 12,468’ at 4 SPF

12,578’ to 12,582’ at 4 SPF

12,596’ – 12,602’ at 4 SPF

12,634’ – 12,686’ at 2 SPF

12,752’ – 12,784’ at 2 SPF

106
Table A3: Well Configuration Data for Well – 3

Well TD 13,000’

Casing Surface to 11,470’ -- 4.5”, 11.6 ppf

11,470’ to 12,981’ – 4.5”, 13.5 ppf

Perforations Total 114 perforations of 0.42” at 2/4 SPF

Perforated Intervals 12,486 -- 12,496 at 4 SPF

12,560 – 12,564 at 4 SPF

12,612 – 12,626’ at 2 SPF

12,708’ – 12,740’ at 2 SPF

Table A4: Well Configuration Data for Well - 4

Well TD 12,995’

Casing Surface to 11,963’ -- 4.5”, 11.6 ppf

11,963 to 12,995’ – 4.5”, 13.5 ppf

Perforations Total 90 perforations of 0.27” at 1 SPF

Perforated Intervals 12,628’ – 12,648’

12,656’ – 12,664’

12,672’ – 12,688’

12,694’ – 12,714’

12,724’ – 12,734’

12,776’ – 12,782’

12,804’ – 12,814’

107
Table A5: Well Configuration Data for Well – 5

Well TD 12,910’

Casing Surface to 11,911’ -- 4.5”, 11.6 ppf

11,911’ to 12,908’ – 4.5”, 13.5 ppf

Perforations Total 202 perforations of 0.30” at 1 SPF

Perforated Intervals 12,510’ – 12,680’

12,700’ – 12,730’

Table A6: Well Configuration Data for Well – 6

Well TD 12,909’

Casing Surface to 11,374’ -- 4.5”, 11.6 ppf

11,374’ to 12,889’ – 4.5”, 13.5 ppf

Perforations Total 76 perforations of 0.30” at 2 SPF

Perforated Intervals 12,602’ – 12,640’

108
Table A7: Well Configuration Data for Well – 7

Well TD 12,950’

Casing Surface to 11,415’ -- 4.5”, 11.6 ppf

11,415’ to 12,946’ – 4.5”, 13.5 ppf

Perforations Total 376 perforations of 0.42”

Perforated Intervals 12,452’ – 12,460’ at 4 SPF

12,472’ – 12,530’ at 2 SPF

12,546’ – 12,556’ at 4 SPF

12,574’ – 12,600’ at 2 SPF

12,606’ – 12,622’ at 2 SPF

12,626’ – 12,636’ at 4 SPF

12,640’ – 12,660’ at 2 SPF

12,664’ – 12,670’ at 4 SPF

109
Appendix B

Well Production Data51

Table B1: Production data for Well – 1

Cumu.
Month Number of Hours Net Prod CumulativeProduction Prod.
Days Down Days Days MCF MCF
Dec-03 7 7 6.71 6.71 12788 12788
Dec-03 8 62 5.42 12.13 7422 20210
Dec-03 7 31 5.71 17.83 7474 27684
Jan-04 8 9 7.63 25.46 10064 37748
Jan-04 8 0 8 33.46 9994 47742
Jan-04 8 0 8 41.46 5235 52977
Jan-04 7 0 7 48.46 5234 58211
Feb-04 29 0 29 77.46 31236 89447

Table B2: Production data for Well – 2

Cumu.
Month Number of Hours Net Prod Cumulative Production Prod.
Days Down Days Days MCF MCF
May-03 20 29 18.79 18.79 47748 47748
Jun-03 30 23 29.04 47.83 61526 109274
Jul-03 31 3 30.88 78.71 49186 158460
Aug-03 31 0 31 109.71 40668 199128
Sep-03 30 0 30 139.71 34748 233876
Oct-03 31 72 28 167.71 31289 265165

Table B3: Production data for Well – 3

Cumu.
Month Number of Hours Net Prod CumulativeProduction Prod.
Days Down Days Days MCF MCF
Jan-04 5 45 3.13 3.13 3688 3688
Jan-04 14 0 14 17.13 13892 17580
Feb-04 29 0 29 46.13 15934 33514

110
Table B4: Production data for Well – 4

Cumu.
Month Number of Hours Net Prod Cumulative Production Prod.
Days Down Days Days MCF MCF
Apr-02 7 5 6.79 6.79 6916 6916
May-02 31 87 27.38 34.17 37720 44636
Jun-02 30 0 30 64.17 31317 75953
Jul-02 31 0 31 95.17 24730 100683
Aug-02 31 26 29.92 125.08 21447 122130
Sep-02 30 0 30 155.08 18815 140945
Oct-02 31 0 31 186.08 17487 158432
Nov-02 30 0 30 216.08 15587 174019
Dec-02 31 0 31 247.08 14989 189008
Jan-03 31 0 31 278.08 13901 202909
Feb-03 28 8 27.67 305.75 11743 214652
Mar-03 31 0 31 336.75 12497 227149
Apr-03 30 0 30 366.75 11695 238844
May-03 31 0 31 397.75 11751 250595
Jun-03 30 0 30 427.75 11277 261872
Jul-03 31 0 31 458.75 11234 273106
Aug-03 31 0 31 489.75 10903 284009
Sep-03 30 0 30 519.75 10527 294536
Oct-03 31 76 27.83 547.58 10236 304772

Table B5: Production data for Well – 5

Cumu.
Month Number of Hours Net Prod Cumulative Production Prod.
Days Down Days Days MCF MCF
Nov-02 6 98 1.92 1.92 2560 2560
Dec-02 31 24 30 31.92 57319 59879
Jan-03 31 24 30 61.92 60890 120769
Feb-03 28 12 27.50 89.42 46364 167133
Mar-03 31 0 31 120.42 46325 213458
Apr-03 30 4 29.83 150.25 35617 249075
May-03 31 40 29.33 179.58 32847 281922
Jun-03 30 0 30 209.58 30109 312031
Jul-03 31 0 31 240.58 26457 338488
Aug-03 31 0 31 271.58 24477 362965
Sep-03 30 0 30 301.58 20953 383918
Oct-03 31 99 26.88 328.46 18873 402791

111
Table B6: Production data for Well – 6

Cumu.
Month Number of Hours Net Prod Cumulative Production Prod.
Days Down Days Days MCF MCF
Mar-03 6 57 3.63 3.63 4017 4017
Apr-03 30 22 29.08 32.71 14689 18706
May-03 31 0 31 63.71 9709 28415
Jun-03 30 0 30 93.71 7668 36083
Jul-03 31 0 31 124.71 7027 43110
Aug-03 31 0 31 155.71 6611 49721
Sep-03 30 0 30 185.71 6090 55811
Oct-03 31 121 25.96 211.67 5211 61022

Table B7: Production data for Well – 7

Cumu.
Month Number of Hours Net Prod Cumulative Production Prod.
Days Down Days Days MCF MCF
Oct-03 3 7 2.71 2.71 3001 3001
Nov-03 8 15 7.38 10.08 16177 19178
Nov-03 8 24 7 17.08 16209 35387
Nov-03 8 0 8 25.08 20939 56326
Nov-03 6 0 6 31.08 16081 72407
Dec-03 8 0 8 39.08 20535 92942
Dec-03 8 0 8 47.08 19975 112917
Dec-03 8 0 8 55.08 17755 130672
Jan-04 7 0 7 62.08 14339 145011
Jan-04 8 0 8 70.08 15338 160349
Jan-04 8 0 8 78.08 14381 174730

112
Appendix C

Lithology of Reservoir Layers

Fig. C1 Layers of Well-1 Fig. C2 Layers of Well-2 (Zone 1)

Fig. C3 Layers of Well-2 (Zone 2) Fig. C4 Layers of Well-2 (Zone 3)

113
Fig. C5 Layers of Well-3 Fig. C6 Layers of Well-5

Fig. C7 Layers of Well-6 Fig. C8 Layers of Well-7

114
Appendix D

Simulation Plots

Meas'd Btmh (psi) Surf Press [Csg] (psi)


12500 8000

ISIP = 11,322 psi


12000 7400

11500
Tangent 6800

11000 6200

10500 5600

10000 5000
20.0 44.6 69.2 93.8 118.4 143.0
Time (min)

Fig. D1 ISIP Plot for Well – 2 (Zone 1)

Meas'd Btmh (psi) Surf Press [Csg] (psi)


12000 7000

ISIP = 11,322 psi


11400 Tangent 6400

10800 5800

10200 5200

9600 4600

9000 4000
111.1 115.0 118.9 122.7 126.6 130.5
Time (min)

Fig. D2 ISIP Plot for Well – 2 (Zone 2)

115
Meas'd Btmh (psi) Surf Press [Csg] (psi)
11000 6000

10400 5400

9800 4800

9200 4200

8600 3600

8000 3000
39.0 63.2 87.4 111.6 135.8 160.0
Time (min)

Fig. D3 ISIP Plot for Well – 2 (Zone 3)

Meas'd Btmh (psi) Surf Press [Csg] (psi)


11000 7000

10200 6200

9400 5400

8600 4600

7800 3800

7000 3000
326.5 356.0 385.5 415.0 444.5 474.0
Time (min)

Fig. D4 ISIP Plot for Well – 3

Meas'd Btmh (psi) Surf Press [Csg] (psi)


10000 4900

9800 4700

9600 4500

9400 4300

9200 4100

9000 3900
6.00 17.80 29.60 41.40 53.20 65.00
Time (min)

Fig. D5 ISIP Plot for Well – 4

116
Meas'd Btmh (psi) Surf Press [Csg] (psi)
12000 7000

11200 6200

10400 5400

9600 4600

8800 3800

8000 3000
28.3 58.8 89.4 119.9 150.5 181.0
Time (min)
Fig. D6 ISIP Plot for Well – 5

Meas'd Btmh (psi) Surf Press [Csg] (psi)


11000 6000

10600 5600

10200 5200

9800 4800

9400 4400

9000 4000
23.9 48.1 72.3 96.6 120.8 145.0
Time (min)

Fig. D7 ISIP Plot for Well – 6

Meas'd Btmh (psi) Surf Press [Csg] (psi)


11000 6000

10400 5400

9800 4800

9200 4200

8600 3600

8000 3000
75.0 94.2 113.4 132.6 151.8 171.0
Time (min)

Fig. D8 ISIP Plot for Well – 7

117
(d/dt) Surf Press [Csg] (psi)
100 Meas'd Btmh (psi) Surf Press [Csg] (psi)
7000
12000

60
11600 6600

20 6200
11200

-20
5800
10800

-60
10400 5400

-100
5000
100000.0 24.5 48.9 73.4 97.8 122.3
Time (min)
Fig. D9 Square Root Plot for Well – 2 (Zone 1)

(d/dt) Surf Press [Csg] (psi)


500 Meas'd Btmh (psi) Surf Press [Csg] (psi)
8000
12000

300
11400 7000

100 6000
10800

-100
5000
10200

-300
9600 4000

-500
3000
90000.0 3.9 7.7 11.6 15.4 19.3
Time (min)

Fig. D10 Square Root Plot for Well – 2 (Zone 2)

(d/dt) Surf Press [Csg] (psi)


500 Meas'd Btmh (psi) Surf Press [Csg] (psi)
7000
12000

300
11200 6200

100 5400
10400

-100
4600
9600

-300
8800 3800

-500
3000
80000.0 24.0 48.1 72.1 96.1 120.1
Time (min)

Fig. D11 Square Root Plot for Well – 2 (Zone 3)

118
(d/dt) Surf Press [Csg] (psi)
500 Meas'd Btmh (psi) Surf Press [Csg] (psi)
12000
12000

300
9600 9600

100 7200
7200

-100
4800
4800

-300
2400 2400

-500
0
0 0.0 29.6 59.3 88.9 118.5 148.1
Time (min)

Fig. D12 Square Root Plot for Well – 3

(d/dt) Surf Press [Csg] (psi)


800 Meas'd Btmh (psi) Surf Press [Csg] (psi)
6000
11000

600
10400 5400

400 4800
9800

200
4200
9200

0
8600 3600

-200
3000
80000.0 11.7 23.5 35.2 47.0 58.7
Time (min)

Fig. D13 Square Root Plot for Well – 4

(d/dt) Surf Press [Csg] (psi)


1500 Meas'd Btmh (psi) Surf Press [Csg] (psi)
12000
12000

900
9600 9600

300 7200
7200

-300
4800
4800

-900
2400 2400

-1500
0
0 0.0 31.0 62.0 93.0 124.0 155.0
Time (min)
Fig. D14 Square Root Plot for Well – 5

119
(d/dt) Surf Press [Csg] (psi)
1000 Meas'd Btmh (psi) Surf Press [Csg] (psi)
7000
12000

760
10600 5600

520 4200
9200

280
2800
7800

40
6400 1400

-200
0
50000.0 24.2 48.4 72.7 96.9 121.1
Time (min)

Fig. D15 Square Root Plot for Well – 6

(d/dt) Surf Press [Csg] (psi)


500 Meas'd Btmh (psi) Surf Press [Csg] (psi)
6000
11000

300
10400 5400

100 4800
9800

-100
4200
9200

-300
8600 3600

-500
3000
80000.0 19.1 38.2 57.2 76.3 95.4
Time (min)

Fig. D16 Square Root Plot for Well – 7

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
1000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 1000
12000 7000

800 800
11600 6600

600 600
11200 6200

400 400
10800 5800

200 200
10400 5400

0 0
10000 0.000 1.020 2.040 3.060 4.080 5.100 5000
G Function Time

Fig. D17 G-Function Plot (PKN Model) for Well – 2 (Zone 1)

120
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
12000 8000

2400 2400
11400 7400

1800 1800
10800 6800

1200 1200
10200 6200

600 600
9600 5600

0 0
9000 0.000 0.420 0.840 1.260 1.680 2.100 5000
G Function Time

Fig. D18 G-Function Plot (PKN Model) for Well – 2 (Zone 2)

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 6000

2400 2400
10400 5400

1800 1800
9800 4800

1200 1200
9200 4200

600 600
8600 3600

0 0
8000 0.000 1.180 2.360 3.540 4.720 5.900 3000
G Function Time

Fig. D19 G-Function Plot (PKN Model) for Well – 2 (Zone 3)

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 7000

2400 2400
10200 6200

1800 1800
9400 5400

1200 1200
8600 4600

600 600
7800 3800

0 0
7000 0.000 1.140 2.280 3.420 4.560 5.700 3000
G Function Time

Fig. D20 G-Function Plot (PKN Model) for Well – 3

121
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
50.00 Meas'd Btmh (psi) Surf Press [Csg] (psi) 100.0
10000 4500

40.00 80.0
9800 4300

30.00 60.0
9600 4100

20.00 40.0
9400 3900

10.00 20.0
9200 3700

0.00 0.0
9000 0.000 1.500 3.000 4.500 6.000 7.500 3500
G Function Time

Fig. D21 G-Function Plot (PKN Model) for Well – 4

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 6000

2400 2400
10400 5400

1800 1800
9800 4800

1200 1200
9200 4200

600 600
8600 3600

0 0
8000 0.000 0.900 1.800 2.700 3.600 4.500 3000
G Function Time

Fig. D22 G-Function Plot (PKN Model) for Well – 5

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
4000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 4000
11000 6000

3200 3200
10600 5600

2400 2400
10200 5200

1600 1600
9800 4800

800 800
9400 4400

0 0
9000 0.000 1.100 2.200 3.300 4.400 5.500 4000
G Function Time

Fig. D23 G-Function Plot (PKN Model) for Well – 6

122
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 4000
11000 6000

2400 3200
10400 5400

1800 2400
9800 4800

1200 1600
9200 4200

600 800
8600 3600

0 0
8000 0.000 0.380 0.760 1.140 1.520 1.900 3000
G Function Time

Fig. D24 G-Function Plot (PKN Model) for Well – 7

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
1000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 1000
12000 7000

800 800
11600 6600

600 600
11200 6200

400 400
10800 5800

200 200
10400 5400

0 0
10000 0.000 1.020 2.040 3.060 4.080 5.100 5000
G Function Time

Fig. D25 G-Function Plot (GDK Model) for Well – 2 (Zone 1)

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
12000 8000

2400 2400
11400 7400

1800 1800
10800 6800

1200 1200
10200 6200

600 600
9600 5600

0 0
9000 0.000 0.420 0.840 1.260 1.680 2.100 5000
G Function Time

Fig. D26 G-Function Plot (GDK Model) for Well – 2 (Zone 2)

123
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 6000

2400 2400
10400 5400

1800 1800
9800 4800

1200 1200
9200 4200

600 600
8600 3600

0 0
8000 0.000 1.180 2.360 3.540 4.720 5.900 3000
G Function Time

Fig. D27 G-Function Plot (GDK Model) for Well – 2 (Zone 3)

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 7000

2400 2400
10200 6200

1800 1800
9400 5400

1200 1200
8600 4600

600 600
7800 3800

0 0
7000 0.000 1.140 2.280 3.420 4.560 5.700 3000
G Function Time

Fig. D28 G-Function Plot (GDK Model) for Well – 3

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
50.00 Meas'd Btmh (psi) Surf Press [Csg] (psi) 100.0
10000 4500

40.00 80.0
9800 4300

30.00 60.0
9600 4100

20.00 40.0
9400 3900

10.00 20.0
9200 3700

0.00 0.0
9000 0.000 1.560 3.120 4.680 6.240 7.800 3500
G Function Time

Fig. D29 G-Function Plot (GDK Model) for Well – 4

124
(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 3000
11000 6000

2400 2400
10400 5400

1800 1800
9800 4800

1200 1200
9200 4200

600 600
8600 3600

0 0
8000 0.000 0.900 1.800 2.700 3.600 4.500 3000
G Function Time

Fig. D30 G-Function Plot (GDK Model) for Well – 5

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
4000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 4000
11000 6000

3200 3200
10600 5600

2400 2400
10200 5200

1600 1600
9800 4800

800 800
9400 4400

0 0
9000 0.000 1.100 2.200 3.300 4.400 5.500 4000
G Function Time

Fig. D31 G-Function Plot (GDK Model) for Well – 6

(d/dG) Surf Press [Csg] (psi) (G·d/dG) Surf Press [Csg] (psi)
3000 Meas'd Btmh (psi) Surf Press [Csg] (psi) 4000
11000 6000

2400 3200
10400 5400

1800 2400
9800 4800

1200 1600
9200 4200

600 800
8600 3600

0 0
8000 0.000 0.420 0.840 1.260 1.680 2.100 3000
G Function Time

Fig. D32 G-Function Plot (GDK Model) for Well – 7

125

Você também pode gostar