investigation of lost circulation materials impact on

Scholars' Mine Scholars' Mine

Doctoral Dissertations Student Theses and Dissertations

Fall 2015

Investigation of lost circulation materials impact on fracture Investigation of lost circulation materials impact on fracture

gradient gradient

Mortadha Turki Alsaba

Follow this and additional works at: https://scholarsmine.mst.edu/doctoral_dissertations

Part of the Petroleum Engineering Commons

Department: Geosciences and Geological and Petroleum Engineering Department: Geosciences and Geological and Petroleum Engineering

Recommended Citation Recommended Citation Alsaba, Mortadha Turki, "Investigation of lost circulation materials impact on fracture gradient" (2015). Doctoral Dissertations. 2437. https://scholarsmine.mst.edu/doctoral_dissertations/2437

This thesis is brought to you by Scholars' Mine, a service of the Missouri S&T Library and Learning Resources. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact [email protected].

https://library.mst.edu/

https://library.mst.edu/

https://scholarsmine.mst.edu/

https://scholarsmine.mst.edu/doctoral_dissertations

https://scholarsmine.mst.edu/student-tds

https://scholarsmine.mst.edu/doctoral_dissertations?utm_source=scholarsmine.mst.edu%2Fdoctoral_dissertations%2F2437&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/245?utm_source=scholarsmine.mst.edu%2Fdoctoral_dissertations%2F2437&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarsmine.mst.edu/doctoral_dissertations/2437?utm_source=scholarsmine.mst.edu%2Fdoctoral_dissertations%2F2437&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

INVESTIGATION OF LOST CIRCULATION MATERIALS IMPACT ON

FRACTURE GRADIENT

by

MORTADHA TURKI ALSABA

A DISSERTATION

Presented to the Faculty of the Graduate School of the

MISSOURI UNIVERSITY OF SCIENCE AND TECHNOLOGY

In Partial Fulfillment of the Requirements for the Degree

DOCTOR OF PHILOSOPHY

in

PETROLEUM ENGINEERING

2015

Approved by:

Runar Nygaard, Advisor

Baojun Bai

Shari Dunn-Norman

Mingzhen Wei

Arild Saasen

3

2015

Mortadha T. Alsaba

All Rights Reserved

iii

ABSTRACT

Lost circulation is a challenging problem to be prevented or mitigated during

drilling. Lost circulation treatments are widely applied to mitigate losses using a

corrective approach or to prevent losses using preventive approaches, also known as

“wellbore strengthening”. The disagreement among the different wellbore strengthening

theories and the lack of understanding the strengthening mechanism resulted in the

absence of a standardized method to evaluate the effectiveness of lost circulation

materials (LCM) for wellbore strengthening application.

An extensive experimental investigation was performed by constructing a high

pressure LCM test apparatus to investigate the effects of different parameters on the

sealing efficiency of LCM treatments. In addition, hydraulic fracturing experiments,

which simulates downhole conditions, were carried out to evaluate the impact of LCM

addition on enhancing both; breakdown and re-opening pressure.

The results showed that the sealing efficiency of LCM treatments is highly

dependent on the fracture width and the particle size distribution (PSD). Carefully

selected LCM blends can seal fractures up to 2500 micron and certain unconventional

squeeze LCM can seal wider fractures. A particle size distribution selection criterion for

LCM treatments was developed based on a statistical analysis of the experimental results

states that D50 and D90 should be equal or greater than 3/10 and 6/5 the fracture width,

respectively. The addition of different LCM blends enhanced the breakdown pressure up

to 18% and the re-opening pressure up to 210%. Comparing the fractures created by the

experiments with analytical models, only one model estimated similar fracture widths.

iv

ACKNOWLEDGMENTS

First of all, thanks to Allah (God) for granting me the blessings, health and

strength to complete this research “Alhamdulillah”. Second, I would like to express my

sincere gratitude to my PhD advisor Dr. Runar Nygaard for his invaluable support

throughout my research. He has been my mentor, motivator, supporter, my older brother,

and more since the first day I joined his group. I would also like to thank my committee

members, Dr. Baojun Bai, Dr. Shari Dunn-Norman, Dr. Mingzhen Wei, and Dr. Arild

Saasen for their valuable advice and recommendations. Dr. Wisner, and Dr. Wronkiewicz

for their help with the microscope work, Dr. Galecki for his help with the PSD analysis.

Many thanks to Jeff Heniff, Mike Bassett, and John Tyler for their valuable assistance in

manufacturing the testing apparatuses. I am very grateful to our research group members,

especially Mohammed Al Dushaishi for his help in the statistical analysis, Reza Rahimi

for his help in the analytical modeling, and Montri Jeennakorn for his help in the lab.

I would also like to acknowledge the funding received from Det Norske

Oljeselskap ASA under research agreements #37709 and #45343.

A Special thanks to my family for their love, support, encouragement, and prayers

throughout my study. For my parents, Turki and Maryam, who brought me to this life and

raised me under their watchful eyes.

Last, but not least, I would like to thank my lovely wife Zahra for her love,

understanding, and the years of patience. I would never have been able to complete this

work without her support for the last 8 years. She was and will be always there for me.

v

TABLE OF CONTENTS

Page

ABSTRACT ....................................................................................................................... iii

ACKNOWLEDGMENTS ................................................................................................. iv

LIST OF ILLUSTRATIONS ............................................................................................. ix

LIST OF TABLES ........................................................................................................... xiv

NOMENCLATURE ........................................................................................................ xvi

UNIT CONVERSION ................................................................................................... xviii

SECTION

1. INTRODUCTION ...................................................................................................... 1

1.1. LOST CIRCULATION TREATMENTS ........................................................... 3

1.1.1. Corrective Treatments. ............................................................................. 4

1.1.2. Preventive Treatments .............................................................................. 5

2. LITERATURE STUDY ............................................................................................. 6

2.1. PREVIOUS EXPERIMENTAL INVESTIGATIONS ....................................... 6

2.1.1. LCM Performance Evaluation.................................................................. 6

2.1.2. Hydraulic Fracturing Experiments. .......................................................... 9

2.2. WELLBORE STRENGTHENING THEORIES .............................................. 11

2.3. ANALYTICAL MODELING OF FRACTURE WIDTH ................................ 13

2.4. REVIEW OF BRIDGING THEORIES ............................................................ 17

2.5. CRITICAL REVIEW OF THE LITERATURE ............................................... 18

2.6. RESEARCH OBJECTIVES ............................................................................. 23

3. METHODOLOGY ................................................................................................... 24

vi

3.1. LCM CHARACTERIZATION ........................................................................ 25

3.1.1. LCM Classification. ............................................................................... 26

3.1.2. Particle Size Distribution Analysis......................................................... 26

3.1.3. Optical Microscopy. ............................................................................... 27

3.1.4. Scanning Electron Microscopy. ............................................................. 28

3.1.5. Swelling. ................................................................................................. 28

3.2. LCM PERFORMANCE EVALUATION ........................................................ 29

3.2.1. Fluid Used. ............................................................................................. 30

3.2.2. Conventional LCM Used........................................................................ 30

3.2.3. Unconventional LCM Used.................................................................... 33

3.2.4. Fracture Width Variation........................................................................ 34

3.2.5. Low Pressure LCM Testing. .................................................................. 35

3.2.6. High Pressure LCM Testing. .................................................................. 36

3.3. HYDRAULIC FRACTURING TEST .............................................................. 40

3.3.1. Core Preparation. .................................................................................... 41

3.3.2. Test Procedures. ..................................................................................... 41

3.4. DATA ANALYSIS ........................................................................................... 42

3.4.1. Statistical Methods. ................................................................................ 42

3.4.2. Analytical Fracture Models. ................................................................... 45

4. RESULTS ................................................................................................................. 46

4.1. LCM CHARACTERIZATION ........................................................................ 46

4.1.1. Updated LCM Classification. ................................................................. 46

4.1.2. Particle Size Distribution Analysis......................................................... 49

vii

4.1.3. Optical Microscopy. ............................................................................... 50

4.1.3.1 LCM particles. ............................................................................50

4.1.3.2 Formed seals. ..............................................................................52

4.1.4. Scanning Electron Microscopy. ............................................................. 55

4.1.5. Swelling. ................................................................................................. 58

4.2. LOW PRESSURE LCM TESTING ................................................................. 59

4.3. HIGH PRESSURE CONVENTIONAL LCM TESTING ................................ 61

4.3.1. Effect of LCM Type and Fracture Width. .............................................. 64

4.3.2. Effect of Concentration. ......................................................................... 65

4.3.3. Effect of Temperature. ........................................................................... 66

4.3.4. Effect of Particle Size Distribution. ....................................................... 68

4.3.5. Effect of Base Fluid. ............................................................................... 72

4.3.6. Effect of Density. ................................................................................... 73

4.3.7. Effect of Weighting Material. ................................................................ 75

4.3.8. Effect of LCM Shape. ............................................................................ 75

4.4. HIGH PRESSURE UNCONVENTIONAL LCM TESTING .......................... 76

4.5. HYDRAULIC FRACTURING TEST .............................................................. 79

4.5.1. Test # 1 with EDC95-11 Base Fluid. ..................................................... 79

4.5.2. Test # 2 with 11 lb/gal OBM without LCM. .......................................... 80

4.5.3. Test # 3 with 11 lb/gal OBM + 20 ppb Graphite and Nutshells. ............ 82

4.5.4. Test # 4 with 11 lb/gal OBM + 30 ppb Graphite and Sized Calcium

Carbonate. .............................................................................................. 84

4.5.5. Test # 5 with 11 lb/gal OBM + 55 ppb Graphite, Sized Calcium

Carbonate, and Cellulosic Fiber............................................................. 85

viii

4.5.6. Hydraulic Fracturing Results Comparison. ............................................ 87

4.6. DATA ANALYSIS ........................................................................................... 88

4.6.1. Statistical Analysis. ................................................................................ 88

4.6.2. Analytical Fracture Models. ................................................................... 92

5. DISCUSSION .......................................................................................................... 95

5.1. LCM PERFORMANCE EVALUATION ........................................................ 95

5.2. HYDRAULIC FRACTURING EXPERIMENTS .......................................... 104

5.3. PARTICLE SIZE DISTRIBUTION SELECTION CRITERIA ..................... 109

5.3.1. Development of PSD Selection Method. ............................................. 111

6. CONCLUSIONS AND RECOMMENDATIONS ................................................. 115

6.1. CONCLUSIONS............................................................................................. 115

6.2. RECOMMENDATION AND FUTURE WORK ........................................... 118

APPENDICES

A. PREVIOUSLY REPORTED LABORATORY RESULTS .................................. 120

B. HYDRAULIC FRACTURING TEST PREPARATION AND

PROCEDURES ..................................................................................................... 124

C. CROSS-REFERENCING TABLES FOR AVAILABLE LCMS UNDER THE

NEW CLASSIFICATION ..................................................................................... 130

D. PARTICLE SIZE DISTRIBUTION ANALYSIS ................................................. 138

E. PREDICTIVE LINEAR MODEL ......................................................................... 149

F. PSD MODELS COMPARISON ........................................................................... 151

G. PUBLICATION FROM THIS DISSERTATION ................................................. 154

BIBLIOGRAPHY ........................................................................................................... 156

VITA .............................................................................................................................. 165

ix

LIST OF ILLUSTRATIONS

Figure Page

2.1. HPHT Fluid Loss Apparatus ........................................................................................ 7

2.2. Particle Plugging Apparatus (Fann Instrument Company, 2011) ................................ 7

2.3. Schematic of the Impermeable Fracture Test Apparatus (Van Oort et al. 2011) ........ 8

2.4. Corrugated Aluminum Platens (Hettema et al. 2007) .................................................. 8

2.5. Schematic of the Permeable Fracture Test (Hettema et al. 2007) ................................ 8

2.6. Cross Section View of the Fracture Cell (Kaageson-Loe al. 2009) ............................. 8

2.7. Symmetric Bi-wing Fracture around Wellbore.......................................................... 13

3.1. Dry Sieve Analysis Setup .......................................................................................... 27

3.2. Sphericity and Roundness Chart (Powers, 1953) ...................................................... 28

3.3. Schematic of the Swelling Test Apparatus ................................................................ 29

3.4. Wedge Foam of Three Sizes; small (S), medium (M), and coarse (C) ...................... 34

3.5. Schematic of Tapered Discs ....................................................................................... 35

3.6. Set of Tapered and Straight Slotted Discs ................................................................. 35

3.7. (a) Low Pressure LCM Testing Apparatus (LPA), (b) Snug-fit Spacer, (c) Tapered

Disc on the Spacer ..................................................................................................... 36

3.8 Schematic of the High Pressure Testing Apparatus .................................................... 37

3.9. Pressure vs. Time Plot Generated from HPA Tests for Conventional LCMs ........... 38

3.10. Pressure vs. Time Plot Generated from HPA Tests for Unconventional LCM ....... 39

3.11. Schematic of the Hydraulic Fracturing Apparatus ................................................... 40

4.1. Graphite Particles at 60x Magnification (left) and 200x Magnification (right) ......... 50

x

4.2. Calcium Carbonate Particles at 60x Magnification (left) and 200x Magnification

(right) ......................................................................................................................... 51

4.3. Nutshells Particles at 60x Magnification (left) and 200x Magnification (right) ....... 51

4.4. Cellulosic Fiber Particles at 60x Magnification (left) and 200x

Magnification (right) ................................................................................................. 51

4.5. Optical Microscopy Images of Graphite and Sized Calcium Carbonate (30 ppb)

Seal Retrieved from 1000 microns Fracture Width ................................................... 52

4.6. Optical Microscopy Images of Graphite and Nutshells (20 ppb) Seal Retrieved

from 1500 microns Fracture Width ........................................................................... 53

4.7. Optical Microscopy Images of Graphite and Nutshells (40 ppb) Seal Retrieved

from 2000 microns Fracture Width ........................................................................... 54

4.8. Optical Microscopy Images of Graphite, Sized Calcium Carbonate, and

Cellulosic Fiber (55 ppb) Seal Retrieved from 1500 microns Fracture Width .......... 55

4.9. Nutshells # 1 (50 ppb) SEM Images of Formed Seal Retrieved from

2000 microns Fracture Width .................................................................................... 56

4.10. Graphite and Sized Calcium Carbonate #1 (80 ppb) SEM Images of Formed

Seal Retrieved from 1500 microns Fracture Width ................................................. 57

4.11. Graphite and Nutshells # 1 (40 ppb) SEM Images of Formed Seal Retrieved

from 2000 microns Fracture Width ......................................................................... 57

4.12. Graphite, Sized Calcium Carbonate, and Cellulosic Fiber # 1 (55 ppb) SEM

Images of Formed Seal Retrieved from 1000 microns Fracture Width ................... 58

4.13. Nutshells Swelling (24 hr Water Soak) @ Room Temperature and 180 ˚F ............ 59

4.14. Effect of LCM Type on the Sealing Pressure for (a) Single LCM Treatments

and (b) LCM Mixtures ............................................................................................. 65

4.15. Effect of LCM Concentration on the Sealing Pressure for (a) Single LCM

Treatments and (b) LCM Mixtures .......................................................................... 66

4.16. Effect of Temperature on the Sealing Pressure........................................................ 67

4.17. Effect of PSD on the Sealing Pressure using 1000 microns Fracture

Width (TS1) for (a) Single LCM Treatments and (b) LCM Mixtures .................... 68

xi

4.18. Effect of PSD on the Sealing Pressure using 1500 microns Fracture

Width (TS2) for (a) Single LCM Treatments and (b) LCM Mixtures .................... 69

4.19. PSD vs Sealing Pressure for 1000 microns Fracture Width .................................... 70



4.22. Effect of Base Fluid on Low and High Concentration LCM Mixtures ................... 73

4.23. Effect of Increasing the Mud Density in OBM and WBM ...................................... 74

4.24. Effect of Weighting Materials Type in OBM and WBM ........................................ 75

4.25. Wedge Foam in WBM Formed Plug using 5 mm disc; (left) Side View,

(right) Bottom View ................................................................................................ 78

4.26. Wedge Foam in OBM Formed Plug using 1500 microns disc; (left) Side View,

(right) Bottom View ................................................................................................ 78

4.27. Damaged Slotted Disc.............................................................................................. 79

4.28. Pressure vs. Time for Test # 1 ................................................................................. 80

4.29. The Resulting Fracture under Microscope (200x Magnification) for Test # 1

Showing Fracture Widths Ranging between 15 – 22 microns ................................ 80


4.31. Fractured Core Showing Two Vertical Fractures and a Cross-sectional View of

the Fractured Core for Test # 2 ................................................................................ 81




4.34. Fractured Core Showing Two Vertical Fractures and a Cross-sectional View

of the Fractured Core for Test # 3 ........................................................................... 83




xii

4.37. Fractured Core Showing Two Vertical Fractures and a Cross-sectional View

of the Fractured Core for Test # 4 ........................................................................... 84


Showing Fracture Widths Ranging between 40 – 145 microns .............................. 85


4.40. Fractured Core Showing Two Vertical Fractures for Test # 5 ................................. 86

4.41. The Resulting Fracture under Microscope (200x magnification) for Test # 5

Showing Fracture Widths Ranging between 50 – 140 microns .............................. 86

4.42. Leverage Plot Showing the Effect of Fracture Width on Sealing Pressure ............. 89

4.43. Leverage Plot Showing the Effect of Density on Sealing Pressure ......................... 89

4.44. Leverage Plot Showing the Effect of D90 on Sealing Pressure............................... 89

4.45. Leverage Plot Showing the Effect of LCM Blend on Sealing Pressure .................. 89


4.47. Leverage Plot Showing the Effect of Base Fluid on Sealing Pressure .................... 90




4.51. Leverage Plot Showing the Actual Sealing Pressure versus the Predicted

Sealing Pressure using the fit model ........................................................................ 91

4.52. Residual Plot of Sealing Pressure versus the Predicted Sealing Pressure ............... 91

5.1. Comparison between Fluid Loss Values from the LPA and the Calculated

Average Fluid Loss per Cycle from the HPA ............................................................ 95

5.2. High Concentration of Nutshells (Left) and Graphite (Right) Treatments Screened

out at the Fracture Surface ......................................................................................... 96

5.3. Plugged Fracture tip of a 2000-micron Tapered Disc showing Stuck Nutshells

Particles ...................................................................................................................... 97

xiii

5.4. Sealing Pressure vs. Average Fluid Loss per Cycle using 1000 microns Fracture

Width (TS1) ............................................................................................................... 98


Width (TS2) ............................................................................................................... 98


Width (TS3 and TS4) ................................................................................................. 99

5.7. Particle Size Distribution for the used Barite and Hematite using Laser

Diffraction Technique .............................................................................................. 102

5.8. Effect of Sieving Barite in OBM ............................................................................. 103

5.9. Pressure vs. Time for the Control Sample (CS) without Nanoparticles .................. 106

5.10. Pressure vs. Time for the Optimum Concentration of Iron-based Nanoparticles

Blend (DF1) ........................................................................................................... 107

5.11. Pressure vs. Time for the Optimum Concentration of Calcium-based

Nanoparticles Blend (DC6) ................................................................................... 107

5.12. (a) Cross-section of Fracture Plane near Wellbore, (b) Seal at Fracture Mouth,

and (c) Cross-section of Fracture Plane at the Fracture End

(Contreras et al. 2014b) ......................................................................................... 108

5.13. EDX of Seal Cross-section at 500x magnification where the Green Color

Indicates Iron Particles Distribution on the Formed Seal

(Contreras et al. 2014b) ......................................................................................... 108

5.14. Fractures at Core End (left) and 3D Representation of the Core End

Containing the Fractures (right) (Contreras et al. 2014b) ...................................... 109

5.15. Leverage Plot Showing the Actual Sealing Pressure versus the Predicted

Sealing Pressure using the Simple Linear Fit Model for the Subset Dataset ........ 112

xiv

LIST OF TABLES

Table Page

2.1. Summary of the Previous LCM Testing Apparatuses and their Limitations ............. 19

2.2. Summary of the Competing Wellbore Strengthening Theories ................................. 20

2.3. Summary of Current Selection Criteria ..................................................................... 22

3.1. Drilling Fluid Properties as Measured in Accordance with API Standards

(API- RP 13B) ........................................................................................................... 30

3.2. Single LCM Treatment Formulation ......................................................................... 31

3.3. LCM Mixture Formulation ........................................................................................ 32

3.4. WEDGE-SET Formulation for Specific Fracture Widths ......................................... 34

3.5. Tapered and Straight Slotted Discs Specifications .................................................... 35

4.1. Summary of the Particle Size Distribution Analysis ................................................. 49

4.2. Low Pressure Apparatus (LPA) Fluid Loss Results in (ml) for Single LCM

Treatments ................................................................................................................. 60

4.3. Low Pressure Apparatus (LPA) Fluid Loss Results in (ml) for LCM Mixtures ....... 61

4.4. Selected Blends for High Pressure Apparatus (HPA) Testing................................... 61

4.5. Summary of HPA Results for Single LCM Treatments using WBM ........................ 62

4.6. Summary of HPA Results for LCM Mixture Treatments using WBM ..................... 63

4.7. Summary of HPA Results for LCM Mixture Treatments using OBM ...................... 63

4.8. HPA Results for Foam Wedges Based System.......................................................... 77

4.9. Summary of the Hydraulic Fracturing Experiments for the Concrete Cores............. 88

4.10. Effect of Different Parameters on the Sealing Pressure and Model Fit ................... 92

4.11. Analytical Models Input Parameters ........................................................................ 93

xv

4.12. The Estimated Fracture Width from the Analytical Models and the Measured

Width under Microscope for the Five Fractured Core Samples .............................. 94

5.1. Particle Size Distribution Analysis Comparision for Barite and Hematite .............. 102

5.2. Description of the Different Fluid used in the Fracturing Experiments .................. 104

5.3. Summary of the Hydraulic Fracturing Experiments for the Shale Cores

(Contreras et al. 2014b) ........................................................................................... 105

5.4. Effect of Different Parameters on the Sealing Pressure and Model Fit for the

Subset Dataset .......................................................................................................... 111

5.5. Comparison of Percentage Match for the Different PSD Selection Criteria ........... 114

xvi

NOMENCLATURE

Symbol Description

LCM Lost Circulation Material

LOT Leak Off Test

XLOT Extended Leak Off Test

PSD Particle Size Distribution

LPM Loss Prevention Material

G Graphite

SCC Sized Calcium Carbonate

NS Nutshells

CF Cellulosic Fiber

HPA High Pressure Apparatus

LPA Low Pressure Apparatus

HPHT High Pressure High Temperature

PPA Particle Plugging Apparatus

NC Non-Controlled Fluid Loss

WBM Water-Base Mud

OBM Oil-Base Mud

SBM Synthetic-Base Mud

TS Tapered Slot Disc

SS Straight Slot Disc

NPs Nanoparticles

CS Control Sample

xvii

DC Blends Containing Calcium-Based Nanoparticles

DF Blends Containing Iron-Based Nanoparticles

PB Breakdown Pressure

T0 Tensile Strength

σH Maximum Horizontal Stress

σh Minimum Horizontal Stress

Wc Fracture Width

L Fracture Length

E Young’s Modulus

v Poisson’s Ratio

Pf Pressure along the fracture surface

pw Wellbore Pressure

rw Wellbore Radius

x Distance from the fracture mouth

p0 Pore Pressure

H0 Null Hypothesis

H1 Alternative Hypothesis

α Type I Error

Si Sample Variance

b0 Constrained Sealing Pressure under the Hypothesis

b Least Square

λ Lagrangian Multiplier

xviii

UNIT CONVERSION

To Convert From To Multiply By

bbl m3 0.1589873

micron mm 0.001

ppb kg/m3

2.85

psi Pascal 6894.76

lb/gal kg/m3

119.826

inches cm 2.54

1

1. INTRODUCTION

With the significant increase in oil demand, a huge number of conventional

hydrocarbon resources are being depleted. As a result, more challenging drilling

operations are required. When drilling challenging wells, such as extended reach wells or

deep water wells, the operational mud weight window narrows (Charlez, 1999). The

lower limit is increased due to higher collapse pressure in deviated wells (Aadnoy and

Chenevert, 1987). While the upper limit, controlled by the fracturing gradient, is reduced

due to higher equivalent circulation density (ECD) in extended reach wells, damaged

wellbores, natural fractures, and lower overburden gradient or as a result of wellbore

deviation (Whitfill and Wang, 2005; Fjaer et al. 2008).

Lost circulation events, defined as the loss of drilling fluids into the formation

(Messenger, 1981), are known to be one of the most challenging problems to be

prevented or mitigated during the drilling phase. The severity of the consequences varies

depending on the loss severity. The loss severity could be classified based on loss rate in

bbls/hr into; seepage (1-10 bbls/hr), partial (10 to 500 bbls/hr) and severe (>500 bbls/hr)

(Nayberg, 1987). By identifying the loss severity, proper remedial action can be taken to

mitigate or stop the losses.

In addition, lost circulation events, because they hinder further drilling, are well-

known for their contribution toward increased cost of drilling operations as a result of

non-productive time (NPT). Approximately 1.8 million bbls of drilling fluids are lost per

year (Marinescu, 2014). This huge number demonstrates the operational challenges

caused by losses while drilling.

2

Lost circulation events are often encountered when drilling into cavernous,

vugular, high permeability, naturally fractured formations or as a result of drilling

induced tensile failures. The loss mechanism differs for each candidate formation.

Drilling fluid losses into natural fractures, cavernous, vugular and high permeability

formations are initiated as soon as the drilling fluid pressure exceeds the pore pressure.

Losses into induced fractures are initiated when the drilling fluid pressure exceeds the

formation breakdown pressure (Howard and Scott, 1951) or in other words, when the

drilling fluid density (lb/gal) exceeds the formation breakdown pressure expressed in

equivalent density (lb/gal).

Assuming a linear poro-elastic stress distribution around the wellbore, the

formation breakdown pressure (PB) occurs when the tangential stress around the wellbore

exceeds the formation tensile strength (T0) (Eq. 1) (Fjaer et al. 2008).

𝑃𝐵 = 3𝜎ℎ − 𝜎𝐻 − 𝑃𝑃 + 𝑇0 (1)

Leak off tests (LOT) or extended leak off tests (XLOT) are often used to estimate

the formation breakdown pressure (fracture gradient) in order to construct the operational

mud weight window, which is very important to select mud weight limits and design

casing programs. However, different wellbore conditions affect the interpretation of

LOT/XLOT measurements (Nygaard and Salehi, 2011). Therefore, misinterpretation of

LOT/XLOT might result in inadequate estimation of formation breakdown pressure and

consequently improper mud weight selection, which could result in initiating fluid losses

into the formation.

3

Nygaard and Salehi (2011) provided LOT interpretation guidelines that takes into

consideration the effect of wellbore condition. In an intact wellbore, with no fractures,

fluid losses will occur whenever the drilling fluid pressure exceeds the formation

breakdown pressure defined in Eq. 1.

However, in the presence of a small fracture at the wellbore wall, losses will start

when the drilling fluid pressure surpasses the effective tangential stresses only (Eq. 2)

since the formation has already lost its tensile strength.

𝑃𝐵 = 3𝜎ℎ − 𝜎𝐻 − 𝑃𝑃 (2)

When a larger fracture exist, the effective tangential stresses goes to the minimum

horizontal stress that is perpendicular to fracture plane and the losses will occur as soon

as the drilling fluid pressure exceeds the minimum horizontal stress. In worst case

scenarios where a large fracture is intersecting with a network of fractures, losses will be

triggered when drilling fluid pressure exceeds the pore pressure.

One of the early efforts to cure losses by adding granular materials to the drilling

fluid was introduced by M T. Chapman in 1890 (White, 1956). Since then, lost

circulation materials (LCM) have been widely used to stop or mitigate drilling fluid

losses into the formation.

1.1. LOST CIRCULATION TREATMENTS

The way that lost circulation treatments are applied could be classified based on

the time when these treatments are implemented. It can be either before (preventive) or

after (corrective) the occurrence of the lost circulation event:

4

1.1.1. Corrective Treatments. Corrective treatments can be defined as any

method that is applied after the occurrence of the losses (Kumar and Savari, 2011). In this

approach, lost circulation treatments are either added continuously to the drilling fluid or

spotted as a concentrated pill in order to mitigate the losses.

Conventional LCM’s which can be described as physical discrete particles

suspended in the drilling fluid to prevent losses, are often classified based on their

appearance or physical properties into four main categories; as fibrous, flaky, and

granular or a blend of all three (Howard and Scott, 1951; White, 1956; Canson, 1985).

Conventional LCM treatments are often used to mitigate seepage or partial losses;

however, there is no industry common approach to deal with severe losses. Different

solutions to prevent/mitigate severe losses have been introduced and can be categorized

in two main categories; LCM solutions and mechanical solutions.

LCM solutions includes but not limited to concentrated LCM pills, cement

(Messenger and McNiel, 1952; Messenger, 1981; Morita and Fuh, 1990; Fidan et al.

2004), chemically activated cross-linked pills (CACP) (Bruton et al. 2001; Caughron et

al. 2002), cross-linked cement (Mata and Veiga, 2004), deformable-viscous-cohesive

systems (DVC) (Whitfill and Wang, 2005; Wang et al. 2005; Wang et al. 2008), nano-

composite gel (Lecolier et al. 2005), gunk squeezes (Bruton et al. 2001; Collins et al.

2010), and concentrated sand slurries (Saasen et al. 2004; Saasen et al. 2011).

Mechanical solutions include expandable liners (Filippov et al. 1999; Lohoefer et

al. 2000; Perez-Roca et al. 2003; Abouelnaaj et al. 2013), and casing/liner drilling

(Shepard et al. 2002; Warren et al. 2004; Karimi et al. 2011). The mechanical solutions

often come at the end of a solution list due to the high cost associated as well as the

5

required logistics limitation. Therefore, LCM solutions are often applied in the first place.

Even though often these LCM treatments were successful, the underlying mechanisms

are not well quantified. As a result, what worked in one well might not work for other

wells.

1.1.2. Preventive Treatments. Preventive treatments can be defined as those

treatments/solutions which are applied prior to entering lost circulation zones in order to

prevent the occurrence of losses. The overall objective of this method is to strengthen the

wellbore (Whitfill, 2008). The concept of wellbore strengthening can be defined as “a set

of techniques used to efficiently plug and seal induced fractures while drilling to

deliberately enhance the fracture gradient and widen the operational window” (Salehi and

Nygaard, 2012).

The overall objective of preventive LCM treatments is to widen the mud weight

window or in other words, to enhance the fracture gradient. The strengthening effect by

means of LCM addition was previously investigated experimentally (Morita et al., 1990;

Van Oort, 2011; Mostafavi et al. 2011; Salehi, 2012). In the field, the addition of LCM to

drilling fluids has increased the fracture gradients effectively above previously drilled

wells in permeable rocks (Fuh et al. 1992; Dupriest et al., 2008). Extensive experimental

investigations took place in order to fully understand how these techniques could possibly

increase the fracture gradient. Different models that explain wellbore strengthening

phenomena have been also introduced (Fuh et al. 1992; Alberty and McLean, 2004;

Aadnoy and Belayneh, 2004; Dupriest, 2005; Van Oort et al. 2011).

6

2. LITERATURE STUDY

The main objective of this section is to review the previous experimental

methods, wellbore strengthening theories, bridging theories and analytical modeling

efforts used to estimate fracture widths followed by a critical review of the literature to

identify the current gaps and limitations.

2.1. PREVIOUS EXPERIMENTAL INVESTIGATIONS

2.1.1. LCM Performance Evaluation. Evaluating the performance of LCM’s is

an important step for an optimized LCM treatment prior to field application. Different

parameters are measured and used as an indication of the LCM performance.

One of the early testing apparatuses to evaluate the effectiveness of LCM was

introduced in the early 1980’s by Sandia National Laboratories (Kelsey, 1981). This test

evaluates the effectiveness of LCM in mitigating losses based on the maximum pressure

that an LCM plug will seal across different slot sizes. The apparatus can be operated at

temperatures up to 300 F˚ and pressures up to 1000 psi.

HPHT filter press (Figure 2.1) and Particle plugging apparatus (PPA) (Figure 2.2)

are currently the most common tests used to evaluate LCMs effectiveness in controlling

fluid loss through hardened filter paper, slotted (SS), and tapered discs (TS) that simulate

natural/induced fractures or ceramic discs that simulate a porous formation (Collins et al.

2010; Kumar and Savari, 2011; Savari et al. 2013). The main difference between PPA

and HPHT is that for the PPA, the pressure is applied from the bottom. The Drilling fluid

containing LCM is forced to pass through these discs under a constant pressures and

7

temperatures. The results are evaluated based on both the total fluid loss within 30

minutes and the time it takes to form a tight seal.

Figure 2.1. HPHT Fluid Loss Apparatus

Figure 2.2. Particle Plugging Apparatus

(Fann Instrument Company, 2011)

A testing apparatus was designed to investigate the efficiency of different LCMs

in sealing impermeable fractured formations (Hettema et al. 2007; Van Oort et al. 2011).

The apparatus (Figure 2.3) uses three pumps and two accumulators to control and

monitor both the fluid and the fracture tip pressure while recording the fluid loss volume

across the fracture tip. The impermeable fracture was simulated by an opposed piston that

uses two matched 2.5-in.-diameter uneven aluminum platens where the fracture width is

set using three screws (Figure 2.4).

Drilling fluids containing LCMs are pumped through the open fracture at a

constant flow rate while keeping both fracture tip and fracture closure pressures constant

8

and monitoring the sealing pressure. A similar testing apparatus (Figure 2.5 and Figure

2.6) was also developed to determine the sealing efficiency in permeable formations

(Hettema et al. 2007).

Figure 2.3. Schematic of the Impermeable

Fracture Test Apparatus (Van Oort et al. 2011)

Figure 2.4. Corrugated Aluminum

Platens (Hettema et al. 2007)

Figure 2.5. Schematic of the Permeable

Fracture Test (Hettema et al. 2007)

Figure 2.6. Cross Section View of the

Fracture Cell (Kaageson-Loe al. 2009)

9

A modified PPA, which can operate at a higher pressure rating, was recently

introduced by Mostafavi et al. (2011). This test measures the performance of LCMs in

terms of the maximum sealing pressure achieved before the seal fails.

Experimental investigations (Hinkebein, et al. 1983; Hettema et al. 2007; Tehrani

et al. 2007; Kaageson-Loe et al. 2009; Savari et al. 2013) were conducted using the

testing apparatuses discussed above to investigate the effect of different parameters on

either the amount of fluid loss or the sealing pressure. These parameters include LCM

types, LCM concentrations, fracture width, and base fluids.

2.1.2. Hydraulic Fracturing Experiments. The drilling engineering association

(DEA-13) fracturing experiments, conducted on 30 inch x 30 inch large blocks, were one

of the very first efforts to investigate the loss circulation phenomena (Morita et al., 1990).

The study revealed that lost circulation pressure is highly dependent on the formation

Young’s modulus, which is in agreement with linear elastic fracture mechanic theories

(Griffith, 1920; Irwin, 1954; Anderson, 1995), as well as wellbore size, and type of

drilling fluid. It was observed that drilling fluids (containing solids) tend to seal/bridge

narrow fractures, which will results in the development of a stabilized fracture and the

breakdown will not occur until the fluid starts to penetrate into these fractures again. The

existence of a narrow fracture tip zone that prevents drilling fluid invasion was observed

in addition to a dehydrated (de-fluidized) mud zone that forms a plug behind the fracture

tip which also helped in sealing the fracture pressure. The filled fracture tip, or in other

words the formed plug size is dependent on the spurt loss and fluid loss, which are

strongly related to the drilling fluid properties. In addition, it was observed that the

fracture reopening pressure depends on the amount of mud cake left on the wellbore wall

10

(Morita et al., 1990, Onyia, 1994 Morita et al., 1996a, and Morita et al., 1996b). It was

also observed that water-base muds (WBM) will have higher reopening pressures than

oil-base muds (OBM) due to the tendency of developing a larger mud cake in WBM. The

theory of fracture initiation and propagation around borehole, where drilling fluids

enhances the stability, has been developed and validated with experiments and field

evidence (Morita et al., 1990).

An increase in the fracture propagation pressure was observed from the

experiments carried out by Fuh et al. (1992) when lost circulation materials (LCM’s)

were used and compared with experiments with untreated drilling fluid. Field trials

showed an increase in the fracture propagation pressure in the range of 3-6 lbm/gal and it

was concluded that this was due to the isolation of the induced fracture tip (Fuh et al.

1992).

The Global Petroleum Research Institute (GPRI) Joint Industry Project (JIP) took

place in the late 1990's to replicate the DEA-13 experiments but on a smaller scale using

4 inch (diameter) x 6 inch (height) cylindrical cores (Dudley et al., 2001 and Van Oort,

2011). The main objective of this project was to investigate the effectiveness of different

LCM’s in increasing the fracturing pressure. It was observed that, by using a specific type

and size of synthetic graphite, the fracturing pressure could be increased significantly. It

was also concluded that WBM would yield a higher increase in the fracturing pressured

compared to synthetic-based muds (SBM). In addition, it was found that the existence of

hydraulically conductive fractures could lower the ideal fracture reopening pressure to

the confining pressure or the minimum horizontal stress.

11

To investigate the effect of including nano sized particles in the LCM blend, , a

set of hydraulic fracturing experiments on 5 7/8 inch (diameter) x 9 inch (height)

cylindrical cores, were conducted on both permeable and impermeable rocks to

investigate the feasibility of using nanoparticles (NP) for wellbore strengthening

applications (Nwaoji et al. 2013). Sandstone cores were used to simulate permeable

formation while concrete cores were used to simulate impermeable formation.

The results indicated that iron hydroxide NP combined with graphite increased

the fracture pressure in permeable rocks by 70% when used in WBM. In addition, the

calcium carbonate NP combined with graphite produced a 36% increase in OBM in

permeable rocks. When these NPs were tested on impermeable samples (concrete), iron

hydroxide NP combined with graphite achieved a 25% increase. The blend of calcium

carbonate NP and graphite increased the fracture pressure by 22%. The results showed

how the size of LCM could results in increasing the fracture pressure due to optimum

sealing of fractures.

2.2. WELLBORE STRENGTHENING THEORIES

The concept of wellbore strengthening can be defined as “a set of techniques used

to efficiently plug and seal induced fractures while drilling to deliberately enhance the

fracture gradient and widen the operational window” (Salehi and Nygaard, 2012). This

approach depends on propping or sealing the fractures using LCMs (Salehi and Nygaard,

2011).

Different models have been introduced to address and explain wellbore

strengthening. Based on a theoretical approach and field trials, Fuh et al. 1992 presented

12

an approach to inhibit the initiation and propagation of fractures while drilling by the

addition of “loss prevention materials” (LPM), or in other words LCMs. The fracture

pressure inhibitor model suggests an increase in both the formation breakdown and

fracture propagation pressures as a result of LCMs screen out at the fracture tip, which

will form a good seal that prevents further fracture growth.

Based on a linear elastic fracture mechanics model, the stress cage model was

presented by Alberty and Mclean (2004). This model suggests that the hoop stress around

the wellbore is increased as a result of propping the fracture by LCMs set at the fracture

mouth to form a seal.

Fracture Closure Stress (FCS) model was introduced by Dupriest (2005) to

explain how LCM’s could increase the fracture gradient (FG) when added to drilling

fluids as a remedial solution for loss circulation. FCS is defined as “the normal stress on

the fracture plane keeping the fracture faces in contact”. The increase in FCS is achieved

by widening the fracture and sealing the fracture tip thus, compressing the adjacent rock

which will result in changing the near wellbore hoop stresses.

The Elastic-Plastic Fracture model presented by Aadnoy and Belayneh (2004)

explains how the FG could be increased above the theoretical Kirsch model value by

means of fracture healing. This model suggests that the fracturing resistance could be

improved as a result of the mud cake plastic deformation that builds up an effective

barrier at the fracture mouth. In addition, it was observed that the fracturing pressure is

sensitive to the type of LCM and concentration used.

Fracture propagation resistance (FPR) is another explanation introduced by Van

Oort et al. 2011 that was based on the DEA-13 experiments (Morita et al., 1990). This

13

work suggested that the increase in the fracture gradient resulted from enhancing the

fracture propagation pressure by means of tip isolation using LCM.

Salehi (2012) investigated the hypothesis of wellbore hoop stress enhancement

when sealing fractures using a three-dimensional poro-elastic finite-element model to

simulate fracture initiation, propagation and sealing in the near wellbore region. The

simulation results contradict that sealing fractures will increase the hoop stresses above

the ideal state of an intact wellbore, which can also be defined as the Kirsch solution.

Based on the simulation results, it was concluded that effective fracture mouth sealing

could restore the hoop stresses around the wellbore.

2.3. ANALYTICAL MODELING OF FRACTURE WIDTH

Different analytical models (Hillerborg et al. 1976: Carbonell and Detournay,

1995; Alberty and McLean, 2004; Wang et al. 2008; Morita and Fuh, 2012) based on

linear elastic fracture mechanics are used to estimate both; the fracture width and the

stress intensity factor at the tip of fracture. These models assume a symmetric fracture

around the wellbore as shown in Figure 2.7.

Figure 2.7. Symmetric Bi-wing Fracture around Wellbore

14

The LCM seal/plug is assumed to either form at the mouth of fracture or at some

distance from the fracture mouth. A brief description of fracture analytical models used in

LCM analysis and fracture width estimation is given below.

An analytical solution (Eq. 3) to estimate the fracture width (Wc) was presented

by Hillerborg et al. (1976). The solution was developed by combining both; fracture

mechanics principal and the finite element method.

𝑊𝑐 =2√2𝐿(𝑝𝑓−𝜎ℎ)

√√(𝑝𝑓−𝜎ℎ)4

+𝐸2𝜎ℎ

2

𝜋2(1−𝜐2)2−(𝑝𝑓−𝜎ℎ)

2

(3)

Where (pf) is the pressure along the fracture, (σh) is the minimum horizontal

stress, and (L) is the fracture length. Assuming a linear crack, the crack propagation

decreases when the crack width increases. The model was developed for aggregated

material like concrete where the fracture width is almost equivalent to the maximum size

of aggregate particles in the order of few millimeters. The fracture width solution is

primarily a function of fracture length and the pressure along the fracture, which is

assumed to be constant.

Carbonell and Detournay (1995) developed a semi-analytical solution to estimate

the fracture propagation pressure using a dislocation based approach. Their solution was

based on the original singular solution of a finite edge dislocation on the boundary of

circular wellbore in an infinite elastic plane derived by Warren (1982). Both, the stress

intensity factor at the tip of fracture and the fracture width could be determined using

Guass-Chebyshev (Erdogan and Gupta, 1972) integration formula to the singular integral

15

of dislocation density function (Eq. 4) where m, h, and tk are Chebyshev polynomials and

Gauss-Chebyshev integration parameters.

𝑊𝑐 =𝜋(𝐿)

2𝑚

4(1−𝜐2)

E∑ ℎ(𝑡𝑘)𝑚

𝑖=0 (4)

The fracture propagation pressure could be calculated using the Newman’s

solution (1971) and the calculated stress intensity factor. Although, the model gives a

solution for the fracture width, fracture length is an input parameter. In addition, the

pressure inside the fracture is assumed to be constant.

A 2-D line crack solution (Eq. 5) was presented by Alberty and McLean (2004) to

compare the results of the fracture width with the result from a finite element analysis.

𝑊𝑐 =4(1−𝜐2)

E(𝑝𝑤 − 𝜎𝐻)√(𝐿 + 𝑟𝑤)2 − 𝑥2 (5)

Where (pw) is wellbore pressure and (rw) is the wellbore radius. The original

solution was for a small line crack (Continuum Fracture Mechanics). Alberty and

McLean (2004) modified the original analytical solution by including the wellbore radius

to make it applicable for wellbore analysis which they compared to the developed finite

element analysis (FEA). The fracture width could be estimated at any distance from the

fracture mouth; however, the estimation is dependent on the fracture length similar to

Hillerborg et al. (1976) and Carbonell and Detournay (1995).

Simplifying an earlier model (Deeg and Wang, 2004), which was based on three

distinct symmetrical pressurized regions to two pressurized regions, Wang et al. 2008

presented a semi-analytical solution for fracture length (Eq. 6).

16

𝑊𝑐 =8(1−𝜈2)𝐿

𝜋𝐸

{𝜋

2𝑎 (𝑝0−𝜎ℎ)+(𝑝𝑤−𝑝0)(𝑎 𝑏 + ∑ sin(2𝑛 𝑏)(2𝑎 cos(2𝑛 𝑏)+

𝑟𝑤 sin(2𝑛 𝑏)

𝐿𝑛)∞

𝑛=1 )}

(2𝑛−1)(2𝑛+1) (6)

Where:

𝑎 = √1 − (𝑟𝑤

𝐿)

2

(7)

And

𝑏 = sin−1 (𝑟𝑤

𝐿) (8)

The crack length is calculated by comparing the stress intensity factor of fracture

to the fracture toughness of rock using an analytical solution for the stress intensity

factor. Similar to the previous models, the fracture length is the dominant parameters to

determine the fracture width.

Based on a 2-D boundary element model and assuming a small fracture around

wellbore, a closed form solution was developed to estimate the fracture width (Morita

and Fuh, 2012). The analytical model (Eq. 9) determines the fracture width (Wc) as a

function of rock properties (E and v), in-situ stress (σh), the length of fracture (L), and the

wellbore (pw) and pore pressure (po), however, in the described workflow (Morita and

Fuh, 2012) to calculate fracture propagation pressure (Pf), the fracture width is an input

parameter to find the fracture length.

𝑊𝑐 =4(1−𝜐2)

𝐸𝐿(𝜎ℎ − 𝑃𝑓) (9)

The fracture propagation pressure came from the linear elastic fracture mechanics

of solids (Tada et al. 2000), which is based on the known toughness at the tip of the

17

crack. Since the plugging material (LCM) might seat somewhere along the fracture

(possibly close to tip of fracture) rather than mouth of fracture, the alternate closed form

solution was provided.

The main limitation of Morita and Fuh’s (2012) model is the use of fracture

geometry as an input parameter due to the fact that there is cumbersome to accurately

measure it. The model is also based on an assumed fracture length of 6 inches, which is

not a verified assumption. Overall, the model is highly sensitive to fracture geometry,

which might results in unrealistic estimated fracture propagated pressure according to the

authors.

2.4. REVIEW OF BRIDGING THEORIES

Bridging can be defined as the buildup of solids within the formation pore spaces

to restrict fluid losses into the formation. Different bridging theories are used as a

guideline to enhance the bridging capabilities of drill-in fluids (i.e. drilling through

reservoir sections), thus minimizing the formation damage caused by fluid losses into the

formation.

Two rules to select an optimum PSD and concentration that is required to

minimize formation damage caused by mud invasion into the formation were proposed

by Abrams in 1977. The first rule suggests that the average particle size (D50) of

bridging materials should be equal or slightly larger than one-third the formation average

pore size. The second rule suggests that the selected bridging materials concentration

should be at least 5% by volume of total solids in the mud formulation.

18

Another criterion for selecting PSD based on laboratory and field application was

suggested by Smith et al. 1996. It was later recommended by Hands et al. (1998) based

on field observations where formation damage was reduced and production rates were

increased. This rule suggests that the D90 of PSD of bridging materials should match the

pore size in order to minimize the fluid invasion into the formation.

To enhance the bridging effectiveness by having a wider range of optimized PSD,

two more target fractions were suggested by Vickers (2006). This method suggests that

both D25 and D75 should be included in the selection criteria, in addition to D10, D50,

and D90, based on a set of laboratory studies. The five D-values selection criteria is a

combination of both Abrams and Hands rules where the D90 should be equal to the

largest pore throat, D75 is less than two-third of the largest pore throat, D50 equal or

slightly larger than one-third, D25 equal to one-seventh of the mean pore throat and D10

greater than the smallest pore throat.

To overcome the uncertainty in the expected fracture widths and optimize PSD, a

new method, which optimize PSD based on fracture width rather than pore throat sizes,

was proposed by Whitfill (2008). This method proposed that the D50 value should be

equal to the fracture width to ensure the formation of an effective seal or plug.

2.5. CRITICAL REVIEW OF THE LITERATURE

The literature study of LCM performance evaluation revealed that there are

pressure, temperature, and fracture width limitations for the different testing apparatuses.

Some of the previous tests emphasized the amount of fluid loss (Savari et al. 2013) while

other tests focused on the sealing pressure (Hettema et al. 2007; Tehrani et al. 2007;

19

Kaageson-Loe et al. 2009; Van Oort et al. 2011; Mostafavi et al. 2011). HPHT filter press

and PPA tests might give a good indication of how much of the drilling fluid might be

lost into the formation. However, these tests were conducted at a constant pressure, which

is not always the case while drilling.

From the permeable and impermeable fracture test (Hettema et al. 2007; Tehrani

et al. 2007; Van Oort et al. 2011), particle size distribution was found to be a critical

parameter for fracture sealing. However, the maximum particle size that can be tested

using these two apparatuses must be less than 1 mm to avoid blocking the tubing in the

apparatus. This limitation resulted in using a maximum fracture width of 1000 microns. It

was also concluded that higher LCM concentration of larger particle sizes would improve

the sealing capabilities. The variation in base fluid showed an impact on the fracture

sealing pressure where higher sealing pressure and lower fluid loss values were obtained

when LCMs were added to OBM compared to WBM and SBM. A summary of the

previous LCM testing apparatuses and their limitation is tabulated in Table 2.1.

Table 2.1. Summary of the Previous LCM Testing Apparatuses and their Limitations

Apparatus Measured

Values

Max.

Press.

(psi)

Temp.

˚F

Filtration

Medium

Fracture

Width (mm)

LCM Tester Sealing

Efficiency 1000 300

API Slots

TS

1 - 5 mm (SS)

2 – 12.7 (TS)

HPHT Fluid Loss Filtration

Characteristic

1500-

2000

350-

500

Filter paper

Ceramic discs N/A

PPA Filtration/

Bridging

4000-

5000 500

Ceramic discs

API/ TS

2 mm (TS)

5 mm (SS)

Impermeable

Fracture Test

Sealing

Efficiency 1250 N/A

Uneven

aluminum

platens

0.3 - 1 mm

20

Table 2.1. Summary of the Previous LCM Testing Apparatuses and their Limitations

(Cont’d)

Permeable Fracture

Test

Sealing

Efficiency 6000 N/A Porous plates 0.25 – 1 mm

Modified PPA

Filtration/

Bridging

Characteristic

8700 N/A Fabricated

Steel Fractures 0.3 - 0.7 mm

Even though a large dataset was developed in these studies, only a limited amount

of the laboratory results were reported in these studies (Appendix A). There are still

limited published results on how fracture width, permeability of the rock, type of base

fluid, mud density, temperature, and particle size distribution could affect the

performance of different LCM’s.

Based on the hydraulic fracturing experiments and the strengthening theories

reviews, we can see that all the previous studies agreed on the effect of LCM addition in

terms of increasing the fracture gradient but no agreement has been achieved on how

LCM size and strength could affect the strengthening. Table 2.2 summarizes the

discussed wellbore strengthening theories and their emphasis on the material size and

strength.

Table 2.2. Summary of the Competing Wellbore Strengthening Theories

Wellbore Strengthening Model Material

Size

Material

Strength Authors

Fracture Pressure Inhibitor Important Important Fuh et al. 1992

Stress Cage Important Important Alberty and McLean,

2004

21

Table 2.2. Summary of the Competing Wellbore Strengthening Theories (Cont’d)

Fracture Closure Stress (FCS) Not

Important

Not

Important Dupriest, 2005

Fracture Healing Important Important Aadnoy and Belayneh,

2004

Fracture Propagation Resistance

(FPR) Important

Not

Important Van Oort et al. 2011

Fracture Mouth Sealing Important Important Salehi et al. 2015

The lack of understanding has resulted in a controversy about the models that

explains the strengthening mechanism. Some of these models (Alberty and Mclean,

2004) suggested that the strengthening effect could be achieved by sealing the fracture

mouth while the others (Fuh et al. 1992; Dupriest, 2005; Van Oort et al. 2011) suggested

sealing the fracture tip. In addition, some models (Alberty and Mclean, 2004; Dupriest,

2005) suggested that sealing fractures would result in hoop stress alteration while others

(Salehi and Nygaard, 2015) suggested that hoop stresses could only be restored and not

increased.

All the analytical fracture models considered in this study estimates the fracture

width primarily as a function of rock properties, fluid pressure within fracture, in-situ

stresses, and fracture length. One of the main limitations of analytical models is using the

length of fracture as input parameters since the measurement or estimation of the fracture

length could be not practical. Some models are assuming fracture length close to 6 in

(Carbonell and Detournay, 1995; Morita, et al. 2012) which might not be realistic for

field applications. Assuming a constant fluid pressure within the fracture is another

22

simplification, which might not be true due to the fact that the pressure beyond the

plugging material might decrease gradually for both permeable and non-permeable

(shale) formations. The shape of the induced or natural fractures might not necessarily be

a line crack and the width of fracture might change along the fracture.

The current bridging theories and their selection criteria (Table 2.3) were

developed to optimize PSD for drill-in fluids based on the pore size distribution, which

can be estimated from thin section analysis, scanning electron microscopy, mercury

injection, or derived from permeability measurements (Dick, 2000; He and Stephens,

2011). The pore sizes dimensions are relatively small compared to fracture widths and

therefore, using these criteria might not be applicable for fracture sealing. Whitfill (2008)

had the only proposed selection criteria based on fracture sealing application; however,

there were no experimental validation presented.

Table 2.3. Summary of Current Selection Criteria

Method Selection Criteria Authors

Abrams Rule D50 ≥ 1/3 the formation average pore size Abrams 1977

D90 Rule D90 = the formation pore size Smith et al. 1996

Hands et al. 1998

Vickers Method

D90 = largest pore throat

D75 < 2/3 the largest pore throat

D50 ≥ 1/3

D25 = 1/7 the mean pore throat

D10 > the smallest pore throat

Vickers et al. 2006

Halliburton Method D50 = fracture width Whitfill, 2008

23

2.6. RESEARCH OBJECTIVES

The literature review in the previous section shows that there is an agreement that

plugging existing fractures by means of LCM resulted in either restoring or increasing the

field experienced fracture gradient (i.e. when losses occur). It can also be concluded that

the sealing efficiency of different LCM’s have a strong effect on wellbore strengthening

by allowing drilling to continue after losses has occurred. However, no single study has

shown how the type and concentration of LCM’s could contribute toward the

enhancement of the fracture gradient. The mechanism of how different LCM’s

characteristics affect the fracture initiation or propagation pressure is still not well

understood.

The main objective of this dissertation is to investigate the feasibility of

effectively sealing wide fractures, with specific application to overburden shales, using

conventional LCM and potentially enhance the fracture gradient. The main objectives can

be broken down into four sub-objectives as follows:

1. Investigate how LCM properties such as shape, PSD, LCM concentration,

types and mixtures impact sealing efficiency.

2. Determine limitations of conventional LCM for sealing existing wide fractures.

3. Investigate the enhancement of the fracture gradient of intact wellbores by the

addition of LCMs.

4. Evaluate the ability to estimate fracture width analytically.

The ability to predict fracture widths and optimize PSD based on the predicted

widths should help in designing effective LCM treatments that would mitigate or prevent

losses, which will be very beneficial in both planning and drilling challenging wells.

24

3. METHODOLOGY

This chapter outlines the different research methodologies used to address the

research objectives listed in chapter 2. In addition, procedures, experimental setups used,

materials as well as the different methods of data analysis will be presented in details.

Due to the large number of up to date available LCM’s and their different

applications, LCM’s will be re-classified into different categories. Dry sieve analysis

techniques were used to analyze the particle size distribution (PSD) of LCM treatments in

order to understand how PSD could affect both the fracture sealing and the sealing

pressure for a known fracture width. PSD analyses should answer the following

questions; what is the role of PSD in effective fracture sealing? Did the effective blends

follow a specific selection criterion? What is the best criterion to select PSD with respect

to fracture width for an effective sealing?

Optical microscopic analysis were undertaken for LCM particles in order to

investigate how LCM shape, in terms of sphericity and roundness, could affect the

performance of LCM in forming an effective seal. Formed seals within the fractures were

examined using a conventional scanning electron microscope (SEM) to investigate how

the particles shape affected the permeability of the formed seal.

Evaluating the performance of LCMs consisted of two steps using two developed

apparatuses. First, the low pressure apparatus (LPA) was used as a quick indicative

measurement that served as a screening phase of the different conventional treatments,

and based on the results; further evaluation of LCM performance at higher pressure was

conducted using the high pressure apparatus (HPA).

25

The effects of varying LCM type, formulation, concentration, fracture width

particle size distribution, base fluid, density, weighting material, and temperature were

studied with respect to differential pressure and fluid loss volume. The objectives are to

establish a better understanding of how these parameters could affect the sealing

efficiency of LCMs and identify their limitations in sealing fractures.

Hydraulic fracturing experiments were carried out using concrete core sample.

These experiments were used to: a) measure the fracture breakdown and re-opening

pressures for OBM, b) investigate the strengthening effect as a result of adding LCM to

OBM, c) investigate whether the sealing pressure measured using the high pressure LCM

tester correlate with the enhancement in breakdown or re-opening pressure.

Statistical analysis of the LCM evaluation results was performed to define the

parameters with the highest effect on the sealing pressure. The knowledge of the

dominant parameter will help in improving in designing an effective LCM treatment.

Finally, the analytical models were used to estimate the fracture widths and the

results were compared to the fracture widths measured under optical microscopy in order

to determine the applicability of the analytical models in estimating fracture widths.

3.1. LCM CHARACTERIZATION

Optical microscopy was used to look at both LCM and the formed plug while a

Scanning Electron Microscopy (SEM) was used to examine the formed plugs. LCMs

PSD were analyzed using dry sieve analysis techniques. In addition, the swelling of one

LCM was quantified at high temperature.

26

3.1.1. LCM Classification. Conventional LCM’s has been classified based on

their appearance as fibrous, flaky, and granular or a blend of all three (Canson, 1985).

LCM’s have different physical and chemical properties and therefore a proper LCM

selection is a key factor for a successful lost circulation treatment.

Howard et al. 1951 classified LCM’s based on their physical properties into four

groups: fibrous, granular, lamellated and dehydratable. Robert J White (1956) modified

the previous classification by replacing the dehydratable category with mixture of LCM

category.

A review of available LCM’s that are currently used in drilling operations was

conducted in order to re-classify LCM into more representative classification (Alsaba et

al. 2014a).

3.1.2. Particle Size Distribution Analysis. Two methods were considered to

determine particle size distribution (PSD); sieve analysis and laser diffraction. However,

since the provided D50 values (from manufacture) for some of the used LCMs in this

study are 2000 microns and due to the limitation of the laser diffraction techniques in

measuring particles larger than 2000 microns (Zhang et al 2015), dry sieve analysis was

used instead to analyze PSD for LCMs. For the weighting materials, the particle size

distribution was measured with MicroTrac S3500 laser diffraction particle size

distribution analyzer.

Samples of LCM treatments were sieved through a series of stacked sieves using

a Gilson mechanical tapping sieve shaker (Figure 3.1). The cumulative weight percent

retained for each sieve size was calculated from the measured weight retained in each

sieve and then plotted versus the sieve sizes.

27

The main five parameters of interest for the different blends that sealed different

fractures width, which were obtained from the resulting plot, are the D10, D25, D50,

D75, and D90, measured in mm and converted to microns where 1 mm = 1000 microns.

In addition, random blends that failed in creating a seal or resulted in a high fluid loss

during the screening phase were analyzed to understand the reason behind their failure in

forming a seal within the fracture.

Figure 3.1. Dry Sieve Analysis Setup

3.1.3. Optical Microscopy. Optical microscopic images were taken for LCM

particles using the INSIZE digital measuring microscope (ISM-PM200) with

magnification up to 200X. The formed plugs were examined under a polarized light

microscopy (Nikon Eclipse E400 POL) that was connected to a digital camera.

28

The sphericity and roundness chart presented by Powers (Figure 3.2), was used to

provide a qualitative visual inspection of LCM particles.

Figure 3.2. Sphericity and Roundness Chart (Powers, 1953)

3.1.4. Scanning Electron Microscopy. Formed seals within the fractures were

analyzed under scanning electron microscope (SEM) using a Hitachi S-570 SEM at the

S&T’s advanced materials characterization laboratory (AMCL). The formed seal samples

were prepared by removing the seal gently from the tapered discs after running the high

pressure apparatus test. After that, the seals were left to dry prior to inspection.

3.1.5. Swelling. A swelling test apparatus (Figure 3.3) using a linear variable

differential transformer (LVDT) was developed following the design and procedure used

to quantify the swelling of shale (Fann, 2015). The apparatus consist of 5 main

components labeled on Figure 3.3 as follows:

1- Adjustable stand to hold LVDT

2- LVDT

3- Mold to hold the sample for compaction and then running the test

29

4- Plate inserted on top of the sample to ensure distributed swelling measurement

5- Data logging (swelling versus time)

The selected LCM sample was first sieved to remove larger particles in order to

have a well-sorted sample. Then the sample was compressed inside the sample holder (3)

and covered with a thin plate (4). The LVDT (2) is lowered and the test was run for 24

hours and the change in the sample height represents the percentage of swelling. To

quantify the effect of temperature on swelling, two tests were conducted; first at room

temperature and secondly at 180 ˚F.

Figure 3.3. Schematic of the Swelling Test Apparatus

3.2. LCM PERFORMANCE EVALUATION

This section presents the preparation and formulation of the evaluated LCM

treatments as well as the drilling fluids used in this study.

(1)

(3)

(5)

(4)

(2)

30

3.2.1. Fluid Used. Unweighted water-based mud (WBM) containing 7%

bentonite was used to eliminate negative or positive effects, if any, of drilling fluid

additives such as weighting materials and fluid loss reducers on LCM performance. In

addition, a pre-mixed low-toxicity mineral oil-based mud (OBM) was used to investigate

the effect base fluid on the performance of LCM. Rheological parameters for the two

fluids used are summarized in Table 3.1.

The effect of density was investigated by repeating some tests at different fluid

densities up to 16.5 lb/gal. In addition, the effect of weighting material was studied by

repeating tests using fluids weighted with barite and hematite.

Table 3.1. Drilling Fluid Properties as Measured in Accordance with API Standards

(API- RP 13B)

Drilling Fluid Properties WBM OBM

Density (lb/gal) 8.6 11

Gels 10s (lb/100 ft2) 18 7

Gels 10m (lb/100 ft2) 35 10

Plastic Viscosity (cP) 10 36

Yield Point (lb/100 ft2) 19 14

3.2.2. Conventional LCM Used. Four commonly used conventional LCMs

having different physical properties, including graphite (G), cellulosic fiber (CF), sized

calcium carbonate (SCC), and nutshells (NS), were used in this investigation to formulate

LCM treatments at different concentrations.

31

To represent background treatments, a total concentration of 15 ppb was used to

formulate single LCM blends. For single LCM concentrated pills, a total concentration of

50 ppb was used. Table 3.2 shows the percentage of total concentration of each LCM to

formulate any individual LCM blend.

Table 3.2. Single LCM Treatment Formulation

Blend Name G Blends SCC Blends NS Blends CF Blends

Blend # 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3

LCM

Type

D(50)

microns % of Total Concentration if used Individually

Gra

ph

ite (

G) 50 20 14 0 0 -- -- -- -- -- -- -- -- -- -- --

100 20 20 20 0 -- -- -- -- -- -- -- -- -- -- --

400 30 26 40 50 -- -- -- -- -- -- -- -- -- -- --

1000 30 40 40 50 -- -- -- -- -- -- -- -- -- -- --

Siz

ed

Ca

lciu

m

Ca

rb

on

ate

(S

CC

)

5 -- -- -- -- 16 6 0 0 -- -- -- -- -- -- --

25 -- -- -- -- 16 6 0 0 -- -- -- -- -- -- --

50 -- -- -- -- 16 13 0 0 -- -- -- -- -- -- --

400 -- -- -- -- 16 21 33 20 -- -- -- -- -- -- --

600 -- -- -- -- 18 27 33 27 -- -- -- -- -- -- --

1200 -- -- -- -- 18 27 34 53 -- -- -- -- -- -- --

Nu

t S

hell

s

(NS

)

620 -- -- -- -- -- -- -- -- 33.3 0 0 0 -- -- --

1450 -- -- -- -- -- -- -- -- 33.3 50 100 0 -- -- --

2300 -- -- -- -- -- -- -- -- 33.3 50 0 100 -- -- --

Cell

ulo

sic

Fib

er

(CF

)

312 -- -- -- -- -- -- -- -- -- -- -- -- 50 100 0

1060 -- -- -- -- -- -- -- -- -- -- -- -- 50 0 100

For example, to mix 1 bbl. Of drilling fluid containing a 15 ppb Graphite Blend

#2 (Table 3.2), the formulation will be; 15 ppb x 14% = 2.1 ppb (D50 = 50 microns), 15

32

ppb x 20% = 3 ppb (D50 = 100 microns), 15 ppb x 26% = 3.9 ppb (D50 = 400 microns),

and 15 ppb x 40% = 6 ppb (D50 = 1000 microns).

Three recommended LCM mixtures (Aston et al. 2004; Hettema et al. 2007,

Kumar et al. 2011) were used in this study to benchmark the results (Table 3.3). Graphite

and sized calcium carbonate blends # 1 and 2 were investigated at two different

concentrations, 30 ppb and 80 ppb to follow the recommendations by Aston et al. (2004).

Graphite and nutshells blends were tested at the concentrations of 20 ppb and 40 ppb as

suggested by Hettema et al. (2007). A 55-ppb blend combining graphite, sized calcium

carbonate, and cellulosic fiber was investigated following Kumar et al. (2011). Graphite

and sized calcium carbonate blends # 3-9 were evaluated at high concentration of 105

ppb as recommended by Det Norske Oljeselskap ASA for field applications (A. Saasen,

2015, Pers. Comm.).

Table 3.3. LCM Mixture Formulation

Blend Name G & SCC Blends G, SCC &

CF Blends

G & NS

Blends

Blend # 1 2 3 4 5 6 7 8 9 1 2 1 2

LCM

Type

D(50)

microns % of Total Concentration if used in combinations

Gra

ph

ite (

G) 50 10 6.7 -- -- -- -- -- -- -- 3.6 2.4 10 6.5

100 10 10 -- -- -- -- -- -- -- 3.6 3.6 10 10

400 15 13.3 33.3 -- -- -- -- 33.3 -- 5.5 4.8 15 13.5

1000 15 20 -- -- -- -- -- -- -- 5.5 7.3 15 20

Siz

ed

Ca

lciu

m

Ca

rb

on

ate

(S

CC

)

5 3.3 -- -- -- -- -- -- -- -- 4.4 0 -- --

25 3.3 -- -- -- -- -- -- -- -- 4.4 0 -- --

40 -- -- 33.3 -- -- -- -- 16.7 --

50 6.7 -- -- -- -- -- -- -- -- 9.5 0 -- --

400 10 16.5 -- -- -- -- -- -- -- 15.3 24 -- --

600 13.3 16.5 -- -- -- -- -- -- -- 19.6 24 -- --

33

Table 3.3. LCM Mixture Formulation (Cont’d)

Siz

ed

Ca

lciu

m

Ca

rb

on

ate

(SC

C)

1200 13.3 17 -- -- -- -- -- -- -- 19.6 24.7 -- --

1400 -- -- 33.3 -- 33.3 50.0 33.3 16.7 37.5 -- -- -- --

2400 -- -- -- -- -- -- 33.3 33.3 37.5 -- -- -- --

Nu

t

Sh

ell

s

620 -- -- -- -- -- -- -- -- -- -- 16.5 16.5

1450 -- -- -- -- -- -- -- -- -- -- 16.5 16.5

2300 -- -- -- -- -- -- -- -- -- -- 17 17

Cell

ulo

sic

Fib

er

(CF

)

312 -- -- -- -- -- -- -- -- 4.5 4.5 -- --

1060 -- -- -- -- -- -- -- -- 4.5 4.5 -- --

Pro

prie

tary

G&

SC

C

Ble

nd

500 -- -- 100 66.7 50 33.3 -- 25 -- -- -- --

3.2.3. Unconventional LCM Used. A wedge foam based system was included in

the study as an unconventional LCM. This system is used to address fracture size and

shape uncertainties when experiencing losses (Wang 2011). This system consists of two

main components: foam wedges at different sizes that are highly deformable (Figure 3.4)

and micron-sized particles.

This high deformability allows the foam wedges to be compressed and forced into

openings of different sizes and shapes. Once the openings are filled with the foam, they

will form a highly permeable filtration bridge for the second component. The second

component consists of high fluid loss fine particles that will form a solid plug on the

filtration bridge.

34

Figure 3.4. Wedge Foam of Three Sizes; small (S), medium (M), and coarse (C)

(Taken from Wang, 2011)

Table 3.4 shows the formulation used to mix unweighted pill for specific fracture

widths. The coarse WEDGE-FOAM was only used with a 10000-micron fracture width

while all tests were mixed with WEDGE-FOAM (M).

Table 3.4. WEDGE-SET Formulation for Specific Fracture Widths

Material

Fracture Width

Up to 8000 microns 10000 micron

Fracture Width

WEDGE-SET 100 ppb 100

WEDGE-FOAM (M) 0.25 ppb 0.5 ppb

WEDGE-FOAM (C) -- 1 ppb

3.2.4. Fracture Width Variation. A set of tapered disc that have a larger

opening (fracture aperture) and tapers down to a smaller opening (fracture tip) as shown

in Figure 3.5 were manufactured at Missouri S&T to simulate sealable fracture widths. In

addition, two commercially available straight slotted discs were used. The tapered and

straight slotted discs (Figure 3.6) specifications are tabulated in Table 3.5.

35

Figure 3.5. Schematic of Tapered Discs

Figure 3.6. Set of Tapered and Straight

Slotted Discs

Table 3.5. Tapered and Straight Slotted Discs Specifications

Disc

#

Diameter

(inches)

Thickness

(inches)

Fracture Aperture

(microns)

Fracture Tip

(microns)

Fracture

Angle

(Degrees)

TS1 2.5 0.25 2500 1000 6.73

TS2 2.5 0.25 3000 1500 6.73

TS3 2.5 0.25 4000 2000 8.95

TS4 2.5 0.25 5000 2000 13.3

TS5 2.5 0.25 4500 2500 8.95

TS6 2.5 1.0 5000 3000 8.95

TS7 2.5 1.0 7000 5000 8.95

TS8 2.5 1.0 18000 10000 8.95

SS1 2.5 0.25 3000 3000 0

SS2 2.5 1.0 5000 5000 0

NOTE: Tapered Discs (TS) Straight Slot (SS)

3.2.5. Low Pressure LCM Testing. The low pressure apparatus (LPA) shown in

Figure 3.7 (a) was manufactured based on the standard API filter press design (API RP

Fracture Aperture

Fracture Tip

36

13B). A snug-fit spacer, shown in Figure 3.7 (b), was manufactured to serve as a holder

for the tapered discs (Figure 3.7 (c)) to prevent plugging the cell outlet.

The test is run by filling the cell with the simple WBM containing LCM’s and

then slowly apply a 100 psi to force the fluid containing LCM to flow through the tapered

disc until no more fluid is coming out. The most important parameter here is the fluid

loss volume. The fluid loss volume could tell if the concentration and PSD are able to

seal a specific fracture width.

Figure 3.7. (a) Low Pressure LCM Testing Apparatus (LPA), (b) Snug-fit Spacer, (c)

Tapered Disc on the Spacer

3.2.6. High Pressure LCM Testing. The effectiveness of LCM treatments was

investigated using a high-pressure LCM testing apparatus, which was designed and

(c)

(b)

(a)

37

manufactured to evaluate the formed seal integrity. The seal integrity, or in other words

the sealing pressure is defined here as the maximum pressure at which the formed seal

breaks and fluid loss resumes. Figure 3.8 shows a schematic drawing of the experimental

setup.

A plastic accumulator (1) used to transfer the drilling fluids to the metal

accumulator (2) prior to pressurizing the fluids containing LCM treatments inside the

testing cell (3) through the tapered discs (4) using an IscoTM

pump (DX100) (5) to

provide injection pressure, which was connected to a computer to record pressure versus

time. The testing cell could also be heated up to 300˚F using an insulated heating blanket,

which was used to investigate the effect of temperature by repeating some tests at a

higher temperature (180˚F) and compare it with the results at ambient temperature.

Figure 3.8 Schematic of the High Pressure Testing Apparatus

(4)

(2)

(5) (3)

(1)

38

The conventional LCM test is run by injecting fluids containing LCMs at a

constant flow rate of 25 ml/min until a rapid increase in the injection pressure is observed

(Figure 3.9). This increase indicates the formation of a seal. Once the seal has formed, the

drilling fluid containing no LCMs is injected continuously until the maximum sealing

efficiency is reached. A significant drop in the pressure is observed as a result of breaking

the seal. However, the test is repeated as cycles until no further seal can be formed. For

tests at high temperature, fluids containing LCM were poured into the high pressure cell

and heated up to 180oF before running the test.

Figure 3.9. Pressure vs. Time Plot Generated from HPA Tests for Conventional LCMs

To evaluate the unconventional LCM, a hesitation squeeze technique was used.

The main objective of this technique is to ensure placing the LCMs (WEDGE-FOAM) in

the fracture and forming the seal within the fracture. Continuous cycles of injection,

0

100

200

300

400

500

100 120 140 160 180 200 220 240 260

Se

ali

ng

Pre

ss

ure

(p

si)

Time (s)

Seal Broken Seal Formation

39

followed by a shut in time, were applied during the test to simulate the hesitation

squeeze. Continuous injection with no shut in period caused the foam wedges to pass

through the fractures before forming a seal. Once the injection stops, the remaining

pressure in the system will push the fluid into the fracture but at a reduced pressure.

Different injection rates were used to determine the optimum injection rate that would

form the seal.

The testing procedures can be summarized into three main steps as follows: (a)

seal initiation, (b) plug development, and (c) plug integrity evaluation (Figure 3.10). The

seal is initiated by injecting fluid at a constant flow rate for 40 seconds and then injection

stopped for 60 seconds. The step is repeated until pressure starts to increase. This process

is continued to develop the plug until significant pressure increase up to 100 psi is

observed. Finally, the plug is further developed and its integrity is tested by injecting

fluid until the pressure goes up to 500 psi and then injection stops for 60 seconds. This

step is repeated and pressure is increased by 500 psi each time until the seal breaks.

Figure 3.10. Pressure vs. Time Plot Generated from HPA Tests for Unconventional LCM

0

500

1000

1500

2000

2500

3000

100 1100 2100 3100 4100 5100 6100Seali

ng

Pre

ssu

re (

psi)

Time (s)

(a) (b) (c)

40

3.3. HYDRAULIC FRACTURING TEST

Hydraulic fracturing experiments were performed on five concrete cores using

five different drilling fluid formulations (OBM). The tests were conducted using a triaxial

pressure cell, which was designed and manufactured at Missouri S&T.

Figure 3.11 shows a schematic drawing of the hydraulic fracturing apparatus.

Two pumps are used to apply confining and injection pressure. Hydraulic hand pump is

used to apply overburden stress on the core sample. A metal accumulator was used to

inject fluids into the core. Injection pressures was recorded using LabVIEW© software.

Figure 3.11. Schematic of the Hydraulic Fracturing Apparatus

41

3.3.1. Core Preparation. Cement core samples were prepared using Portland

cement (Class H) to simulate impermeable formations. The cement core preparation steps

are summarized as follows:

1- Class H cement was mixed with API recommended water requirements of 38%

by weight of cement in a large batch following the standard mixing procedures to

ensure the same physical properties of the fractured cores.

2- The cement mixture was poured into 5 7/8 inch (diameter) x 9 inch (height)

molds and left to cure for at least 7 days.

3- A ½ inch wellbore was drilled in the cement cores using a drill press.

4- Top and bottom caps was attached to the cement cores using epoxy.

3.3.2. Test Procedures. After preparing the cement cores, the hydraulic

fracturing test procedures can be summarized as follows:

1- The core was placed in the fracturing cell

2- The pressure lines were connected to both the cell and the core sample

3- 400 psi overburden and 100 psi confining pressure were applied

4- Filled up accumulator with drilling fluid (without LCM)

5- Wellbore was filled up with drilling fluid containing LCM

6- Injection was started to build pressure inside the wellbore until reaching the

breakdown pressure where an increase in the confining pressure will be observed

due to fluid pushing against the rubber sleeve.

7- The injection was stopped to allow for 10 minutes before running the re-opening

cycle.

8- The test was stopped after the second cycle when an increase in the confining

42

pressure is observed, indicating that fluid already propagated through the fracture.

Detailed steps for the core preparation and test procedures can be found in Appendix B.

3.4. DATA ANALYSIS

3.4.1. Statistical Methods. A statistical analysis was conducted using JMPTM

statistical analysis software to understand the relationship between the different

investigated parameters such as fracture width, LCM type/blend, base fluid, and PSD on

the performance of LCM in terms of the sealing pressure.

First, a regression analysis was conducted by performing a multiple linear

regression analysis to model a relationship between 9 explanatory variables and the

sealing pressure response. The 9 variables used are fracture width, LCM type/blend, base

fluid, fluid density, and the five D-values obtained from PSD analysis; D10, D25, D50,

D75, and D90.

The probability test (F-test) was performed to test the influence of each parameter

on sealing pressure. The F-test provides a P-value, where the P-value is basically a

statistical probability that the predicted value (in this case the F-value) is similar or very

different from the measured value, assuming a true null hypothesis (H0) that proposes no

influence of a specific variable on the sealing pressure (Montgomery, 2001). With a

confidence interval of 95% and type I error (α) of 0.05, P-values less than 0.05 suggests

rejecting the null hypothesis (H0) and accepting an alternative hypothesis (H1). The

alternative hypothesis suggests that the sealing pressure is influenced by a specific

variable. The F-test is calculated based on the variance of the data as:

43

𝐹 =𝑆1

2

𝑆22 (10)

Where 𝑆12 is the variance of the first sample and 𝑆2

2 is the variance of the second

sample. The variance can be defined as the average squared difference from the mean.

Leverage plots for general linear hypothesis, introduced by Sall (1990), were

plotted for each of the 9 variables (predictors) to show their contribution to the predicted

sealing pressure. The Effect Leverage plot is used to characterize the hypothesis by

plotting points where the distance between each point to the fit line shows the

unconstrained residual while the distance to the x-axis shows the constrained residual by

the hypothesis. The constrained sealing pressure for each parameter under the hypothesis

can be written as:

𝑏0 = 𝑏 − (𝑋′𝑋)−1𝐿′𝜆 (11)

Where b is the least square, (𝑋′𝑋) is the inverse matrix (the transpose of the

matrix data being D-values and other parameter, used to enforce orthogonality), and λ is

the Lagrangian multiplier for the hypothesis constraint (L). The residual constrained by

the hypothesis (r0) is:

𝑟0 = 𝑟 + 𝑋(𝑋′𝑋)−1𝐿′𝜆 (12)

Where the Lagrangian multiplier is defined as:

𝜆 = (𝐿(𝑋′𝑋)−1𝐿′)−1𝐿𝑏 (13)

44

The residuals unconstrained by the hypothesis (r) are the least squares residuals

defined as:

𝑟 = �́� − 𝑋𝑏 (14)

The Leverage plot is constructed by plotting vx on the x-axis (Eq. 15) versus vy on

the y-axis (Eq. 16). vx values represents the difference in the residuals caused by the

hypothesis, which is the distance from the model fit line to the x-axis while vy values are

vx plus the unconstrained residuals.

𝑣𝑥 = 𝑋(𝑋′𝑋)−1𝐿′𝜆 (15)

𝑣𝑦 = 𝑟 + 𝑣𝑥 (16)

The sealing pressure residuals are regressed on all predictors except for the

variable of interest while the x-residuals (variable of interest) are regressed on all other

predictors in the model. The mean of the sealing pressure, without the effect of variable

of interest, is plotted as well as a least square fit line and confidence interval for easier

interpretation of the results. The upper and lower confidence interval could be plotted

using Eq. 17 and Eq. 18, respectively. The least squares fit line slope is a measure of how

the tested variable affects the sealing pressure i.e. a non-zero slope implies that the tested

variable will affect the sealing pressure (Sall, 1990).

𝑈𝑝𝑝𝑒𝑟 (𝑥) = 𝑥𝑏 + 𝑡𝛼/2𝑠√𝑥(𝑋′𝑋)−1𝑥′ (17)

45

𝐿𝑜𝑤𝑒𝑟 (𝑥) = 𝑥𝑏 − 𝑡𝛼/2𝑠√𝑥(𝑋′𝑋)−1𝑥′ (18)

Where x = 1 x is the 2-vector of regressors.

3.4.2. Analytical Fracture Models. The test parameters used for the hydraulic

experiments were applied as the input values for the five analytical fracture models

presented in chapter 2.3. The analytical models were used to predict the fracture width.

The estimated fracture widths were compared to the fracture width measured under

optical microscopy after the sample was retrieved from the fracturing apparatus.

46

4. RESULTS

4.1. LCM CHARACTERIZATION

4.1.1. Updated LCM Classification. A new classification of LCM was

performed including new LCM technologies, based on both the physical, chemical

properties, and their application. The physical properties include the shape and the size of

these particles while the chemical properties include material solubility in acids,

swellability, and reactivity with other chemicals to activate the blend. LCM’s are re-

classified into eight categories: granular (1), flaky (2), fibrous (3), LCM’s mixture (4),

acid soluble (5), high fluid loss LCM’s squeezes (6), swellable/hydratable LCM’s (7),

and nanoparticles (8).

1. Granular. Granular materials are defined as additives that are capable of forming a

seal at the formation face or within the fracture to prevent the losses into the formation

(Howard and Scott, 1951; Nayberg, 1987). They are available in a wide particle size

distribution. Due to their rigidity, this type of materials is used often for wellbore

strengthening applications. Granular materials have higher crushing resistance than other

LCM types and some of them could be classified as granular and at the same time acid

soluble such as calcium carbonate. Granular materials include graphite, nutshells, sized

calcium carbonate, glisonite, course bentonite, asphalt, and perlite.

2. Flaky. Flaky materials are defined as “A type of LCM that is thin and flat in shape,

with a large surface area” (SLB Oil Field Glossary). This type may or may not have any

degree of stiffness and they are capable of forming a mat over the permeable formation

face (Howard and Scott, 1951; Nayberg, 1987). Flaky materials include cellophane, mica,

cottonseed hulls, vermiculite, corn cobs, and flaked calcium carbonate.

47

3. Fibrous. Fibrous materials can be defined as “A type of LCM that is long, slender and

flexible and occurs in various sizes and lengths of fiber” (SLB Oil Field Glossary). This

type of materials may have low stiffness and will form a ‘’mat-like’’ bridge when used to

reduce the losses into fractures or vugular formations (Howard and Scott, 1951). The

ability to form a “mat-like” bridge serves as a filtration medium for smaller particles in

the drilling fluids to deposit and form a seal (Nayberg, 1987). Fibrous materials comes in

a wide range of grades/sizes and some types of fibrous material are acid soluble such as

Magma Fiber. Fibrous materials are often used in both WBM and OBM but some of

these materials have some limitations when used in OBM. Fibrous materials include

cellulose fibers, nylon fibers, mineral fibers, saw dust, and shredded paper.

4. Mixture of LCM’s. Several studies have shown that mixing two or more LCM’s

together will yield a better performance in mitigating losses due to the different

properties and particle sizes of the mixed blend compared to individual loss circulation

materials (Nayberg, 1987; Whitfill, 2008; Van Oort et al. 2011; Kumar and Savari, 2011).

Engineered LCM’s blends are available for different lost circulation scenarios. These

blends contain optimized types and particle size distributions that have been proved

experimentally to seal a wide range of fracture sizes.

5. Acid soluble/water soluble. Conventional LCM’s have the disadvantage of damaging

the formation when used in the reservoir section; as a result the development of non-

damaging LCM’s has increased (Sweatman and Scoggins, 1988). Acid/water soluble

LCM’s are considered as non-damaging LCM’s that could be used to cure losses in

reservoir sections. Acid soluble materials include calcium carbonate and mineral fibers.

Water soluble LCM’s include sized salts.

48

6. High fluid loss LCM’s squeeze. This type of LCM’s combination is often used to

cure severe losses when encountering fractured or highly permeable formations. The

filtration process will form a plug that seals the losses zone. These treatments often

require special procedures in order to squeeze them into the losses zone and it is usually

performed as a “hesitation squeeze” (Lost Circulation Guide, 2014).

7. Swellable/hydratable LCM’s combinations. Settable/Hydratable treatments are

basically a blend of LCM’s with a highly reactive material such as polymers. These

treatments will be activated either by chemical reagents or whenever they contact drilling

fluids or formation fluids; as a result, a plug will be formed within the losses zone. These

types of treatments often require special placement procedures (Lost Circulation Guide,

2014).

8. Nanoparticles. Current applied nanoparticles (NPs) include silica, iron hydroxide and

calcium carbonate (Friedheim et al. 2012; Hoelscher, et al. 2012; Nwaoji et al. 2013;

Contreras et al. 2014a). NPs can be prepared by either ex-situ or in-situ procedures. Ex-

situ stands for the preparation of NPs that occurs into an aqueous solution that is later

added to the mud. In-situ involves the addition of the precursors that forms the NPs

directly to the mud. Other types of NPs obtained from aluminum and titanium have been

investigated for permeability reduction in presence of WBM (Chenevert and Sharma,

2009). Carbon black NPs have been used to reduce mud cake thickness to mitigate

differential pipe sticking (Paiaman and Al-Anazi, 2008). Ballard and Massam (2008)

investigated the use of barium sulfate NPs as a weighting agent. Zinc NPs application for

lubricity improvement was reported by Griffo and Keshavan (2007) when used as a

drilling bit lubricator in the presence of other additives including silica gel.

49

A comprehensive summary that includes most of the available LCMs are

tabulated in cross-referencing table using the new classification to be used as a reference

table for operators and drilling engineers (Appendix C). The tables list the generic name,

trade name and the recommended application for each LCM. Majority of LCM comes in

different grades; fine, medium, and coarse to suit different losses scenarios.

4.1.2. Particle Size Distribution Analysis. Table 4.1 shows a summary of the

blends particle size distribution, which was analyzed using the dry sieve analysis

technique. Graphs of the cumulative % of finer particles passing versus the logarithmic

sieve size in mm for the tested blends can be found in Appendix D.

Table 4.1. Summary of the Particle Size Distribution Analysis

LCM Blend Total Concn.

(ppb)

Particle Size Distribution (microns)

D10 D25 D50 D75 D90

G # 1 15 60 85 320 800 1300

G # 1 50 60 95 340 800 1300

NS # 1 15 180 400 1000 1600 2000

NS # 1 50 180 400 1000 1600 2400

SCC # 3 50 250 360 680 950 1200

CF # 1 15 90 140 220 700 1400

CF # 1 50 90 150 220 800 1400

G & SCC # 1 30 80 100 460 900 1300

G & SCC # 1 80 80 120 480 900 1300

G & SCC # 3 105 65 90 420 1100 1400

G & SCC # 4 105 60 150 500 700 900

G & SCC # 5 105 90 400 700 1200 1400

G & SCC # 6 105 100 500 900 1250 1400

G & SCC # 7 105 170 650 1300 1900 2600

G & SCC # 8 105 100 250 1000 1800 2400

G & SCC # 9 105 300 800 1400 1800 2200

50

Table 4.1. Summary of the Particle Size Distribution Analysis (Cont’d)

G & NS # 1 20 65 180 500 1300 1900

G & NS # 1 40 80 180 580 1400 2000

G,SCC, & CF # 1 55 55 100 450 850 1200

NOTE: Graphite (G) Nutshells (NS) Sized Calcium Carbonate (SCC) Cellulosic Fiber (CF)

4.1.3. Optical Microscopy.

4.1.3.1 LCM particles. Figure 4.1 - Figure 4.4 show the optical microscopy

images for four of the LCMs used in this study. The images were analyzed using the

sphericity and roundness chart by Powers (1953). The graphite consist of (Figure 4.1) of

mostly sub-rounded particles with high sphericity. Calcium carbonate particles can be

classified as sub-angular with high sphericity (Figure 4.2). Nutshell particles (Figure 4.3)

could be described as angular with low sphericity. The cellulosic fiber (Figure 4.4) can be

described as nonspherical particles with some elongated and flat particles.

Figure 4.1. Graphite Particles at 60x Magnification (left) and 200x Magnification (right)

1 mm 1 mm

51

Figure 4.2. Calcium Carbonate Particles at 60x Magnification (left) and 200x

Magnification (right)

Figure 4.3. Nutshells Particles at 60x Magnification (left) and 200x Magnification (right)

Figure 4.4. Cellulosic Fiber Particles at 60x Magnification (left) and 200x Magnification

(right)

1 mm 1 mm

1 mm 1 mm

1 mm 1 mm

52

4.1.3.2 Formed seals. Figure 4.5 - Figure 4.8 show the formed seal of different

blends. The formed seals were retrieved from the tapered discs carefully after the high

pressure testing and left to dry prior to examination under microscopy.

Figure 4.5 shows the formed seal using a 30 ppb graphite and sized calcium

carbonate blend retrieved from a 1000 microns fracture width test. The image shows that

the sized calcium carbonate particles are relatively large compared to the graphite

particles. Despite the low average fluid loss per cycle (4.1 ml/cycle) caused by the

presence of very fine particles, which contributed in sealing smaller voids between the

larger particles, the seal broke at relatively low sealing pressure of 487 psi. High

sphericity and roundness of both graphite and calcium carbonate particles (Figure 4.1 and

Figure 4.2) may have resulted in lower friction between the particles and the tapered slot.

Thus, less pressure was required to force these particles through the fracture tip, which

can explain the low sealing pressure of this blend.

Figure 4.5. Optical Microscopy Images of Graphite and Sized Calcium Carbonate (30

ppb) Seal Retrieved from 1000 microns Fracture Width

40X 20X

Calcium carbonate

Fine graphite

Calcium carbonate

Fine graphite

53

Figure 4.6 shows the formed seal of 20 ppb graphite and nutshells blend at two

different magnifications, which was retrieved from a 1500 microns tapered slot test. The

image at 20x magnification shows more presence of the coarse nutshells particles while

graphite particles were very small compared to nutshells. It can be also seen that the

irregularity of the nutshells resulted in a better arrangement of the particles inside the

tapered disc. This seal was able to sustain higher sealing pressure of 805 psi at wider

fracture compared to the graphite and calcium carbonate seal (Figure 4.5), which was

tested with a 1000-micron fracture width. The higher sealing pressure can be explained

by the irregularity of nutshell particles, which resulted in higher friction between the

particles that required higher pressure to break the seal.

Figure 4.6. Optical Microscopy Images of Graphite and Nutshells (20 ppb) Seal

Retrieved from 1500 microns Fracture Width

Figure 4.7 shows the retrieved formed seal using 40 ppb graphite and nutshells

when tested at 2000 microns fracture width. The image shows the presence of fine

graphite particles among the formed seal with a very wide distribution (poor sorting) of

60X 20X Coarse nutshells

Irregular shape of nutshells

Fine graphite

54

nutshells particles. The formed seal was only able to hold a maximum pressure of 295 psi

before fluid flow through the tapered disc, which resulted in higher average fluid loss per

cycle (32 ml/cycle). The lack of large particles, seen in both the optical microscope

picture (Figure 4.7) and the PSD analysis (Table 4.1), may have contributed to the low

sealing pressure of this blend.

Figure 4.7. Optical Microscopy Images of Graphite and Nutshells (40 ppb) Seal

Retrieved from 2000 microns Fracture Width

Figure 4.8 shows the retrieved seal using 55 ppb graphite, sized calcium

carbonate, and cellulosic fiber when tested with a 1500 microns fracture width. The

image shows clearly the presence of fine graphite particles, relatively larger calcium

carbonate particles, and less presence of poorly sorted cellulosic fibers. The seal was

able to withstand a maximum pressure of 391 psi with an average fluid loss per cycle of

14.1 ml/cycle. The low sealing pressure is attributed to the lack of particles that are large

enough to bridge inside the tapered disc, where the D90 obtained for this blend from the

60X 20X

Fine graphite Wide PSD of nutshells Wide PSD of nutshells

Fine graphite

55

dry sieve analysis (1200 microns) relatively small compared to the fracture tip (1500

microns).

Figure 4.8. Optical Microscopy Images of Graphite, Sized Calcium Carbonate, and

Cellulosic Fiber (55 ppb) Seal Retrieved from 1500 microns Fracture Width

In general, it was observed that the finer particles helped in sealing the voids

between other coarser particles. The low particles sphericity and angularity of nutshells

improved the formation of a tighter seal higher sealing pressures for nutshell blends.

4.1.4. Scanning Electron Microscopy. Figure 4.9 - Figure 4.12 show SEM

images of the formed seals. When analyzing SEM images, the light is dense while dark is

less dense so the dark spots represent void spaces between the particles.

Figure 4.9 shows the formed seal of a 50 ppb nutshells retrieved from 2000

microns fracture width test. This seal was sustained a maximum sealing pressure of 755

psi with an average fluid loss per cycle of 18.3 ml/cycle. The SEM image shows that the

low sphericity and the irregular shapes nutshells particles (30x magnifications) with the

rough surface (100x magnifications) resulted in a good packing inside the tapered disc

40X 20X

Fine graphite Cellulosic fiber

Calcium carbonate

Cellulosic fiber

Fine graphite

Calcium carbonate

56

with higher internal friction and thus higher sealing pressure. This blend has a wide PSD

(poor sorting) of nutshell blends, with D10 of 180 microns and D90 of 2400 microns

(Table 4.1). The wide PSD resulted in sealing the 2000 microns fracture with low fluid

loss due to the presence of fine nutshell particle that filled void spaces between larger

particles.

Figure 4.9. Nutshells # 1 (50 ppb) SEM Images of Formed Seal Retrieved from 2000

microns Fracture Width

Figure 4.10 shows formed seal using an 80 ppb graphite and calcium carbonate,

which was retrieved from 1500 microns fracture width. This seal sustained lower sealing

pressure (224 psi) at a smaller fracture width compared to the nutshells seal (Figure 4.9).

The low sealing pressure is attributed to the high sphericity of both graphite and calcium

carbonate particles (Figure 4.1 and Figure 4.2), which resulted in more void spaces within

the formed seal.

100X 30X

Fine particles

Surface Roughness of Nutshells

57

Figure 4.10. Graphite and Sized Calcium Carbonate #1 (80 ppb) SEM Images of Formed

Seal Retrieved from 1500 microns Fracture Width

Figure 4.11 shows the 40 ppb graphite and nutshells seal retrieved from a 2000

microns fracture width test. This seal was able to hold a maximum sealing pressure of

242 psi, which is relatively low. The image (30x magnification) shows clearly the

irregular shape of nutshell particles as well as the presence of fine particles between the

larger particles. The image at higher magnification (100 x) illustrates the difference in the

surface roughness between graphite (lighter) and nutshells (darker). It can be seen that the

nutshell particles has relatively rougher surfaces compared to graphite particles.

Figure 4.11. Graphite and Nutshells # 1 (40 ppb) SEM Images of Formed Seal Retrieved

from 2000 microns Fracture Width

300X 180X

Void Spaces Void Spaces

100X 30X

Nutshells Graphite

Irregular shapes Void space

Void space

58

Figure 4.12 shows the formed seal using graphite, calcium carbonate, and

cellulosic fiber blend retrieved from 1000 microns fracture width. The formed seal was

capable of holding a maximum pressure of 1011 psi. The SEM image at 35x

magnification shows fine particles filling the gaps between larger particles. At 450x

magnification, the surface roughness of cellulosic fiber could be clearly seen, which

enhanced the sealing pressure due to the higher friction among the particles.

Figure 4.12. Graphite, Sized Calcium Carbonate, and Cellulosic Fiber # 1 (55 ppb) SEM

Images of Formed Seal Retrieved from 1000 microns Fracture Width

4.1.5. Swelling. Swelling test was conducted for nutshells to correlate their

sealing pressure enhancement when tested at higher temperature. The increase in the

sealing pressure for nutshells blend was experimentally confirmed to be due to the

swelling of nutshell particles at higher temperature. Figure 4.13 shows the swelling of

nutshells for 24 hours at room temperature (dotted line) and at 180 ˚ F (dashed line). The

solid line represents the temperature profile.

35X 450X

Fine particles Cellulosic fiber

59

Figure 4.13. Nutshells Swelling (24 hr Water Soak) @ Room Temperature and 180 ˚F

The results show a 3% swelling (water soak) over 24 hours at room temperature.

At higher temperature (180˚F), nutshells exhibited approximately 13% swelling as shown

in Figure 4.13. Based on these tests, temperature is another factor that could improve, or

reduce, the seal integrity for some LCM treatments. Therefore, it is recommended to

evaluate the seal integrity of LCM treatments at higher temperatures.

4.2. LOW PRESSURE LCM TESTING

A total of 160 tests were conducted using the low pressure apparatus (LPA) for

LCM blends at different concentrations. A cut off value of 100 ml was used in all the

testing, i.e. if fluid loss value exceeds 100 ml, the fluid loss is considered as non-

controlled (NC). All screened LCM treatments failed to establish a seal in fractures wider

than 2000 microns. Fluid loss values for single LCM treatments are tabulated in Table

4.2. For LCM mixtures, the results are tabulated in Table 4.3.

0

20

40

60

80

100

120

140

160

180

200

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12 14 16 18 20 22 24

Tem

p (

F)

Sw

ellin

g %

Time (hr)

Swelling % @ Room Temp

Swelling % @ 180 Deg F

Temp F

60

From the screening tests results, blends were selected for further evaluation using

the HPA based on the fluid loss. Table 4.4 shows the selected blends and the fracture

width to be tested using the HPA. Nutshells blends were the only LCM capable of sealing

the 2000 microns fracture width (TS3 and TS4) when used individually and therefore, all

the other individual LCM treatments were excluded from being tested using the disks

TS3 and TS4. Because none of the LCM’s mixtures were able to seal the disks TS3 and

TS4 except the one with nutshells at 40 ppb, all other mixtures were also excluded from

being evaluated using the disks TS3 and TS4.

Table 4.2. Low Pressure Apparatus (LPA) Fluid Loss Results in (ml) for Single LCM

Treatments

LCM

Type

Blend # 1 2 3 4

Concn.

(ppb) 15 50 15 50 15 50 15 50

Gra

phit

e (G

) 1000 (TS1) 9 7 11 6 19 5 35 NC

1500 (TS2) 41 35 69 19 38 6 52 NC

2000 (TS3) NC NC NC NC NC NC NC NC


Nu

tsh

ells

(N

S)

1000 (TS1) 4 2 15 2 12 3 19 33

1500 (TS2) 12 7 21 4 21 17 58 NC

2000 (TS3) NC 31 45 15 NC NC NC NC

2000 (TS4) NC 30 38 15 NC NC NC NC

Siz

ed C

aCO

3

(SC

C)

1000 (TS1) NC 64 NC 10 48 13 29 19

1500 (TS2) NC NC NC NC NC 13 NC 26



Cel

lulo

sic

Fib

er

(CF

)

1000 (TS1) 12 0 67 23 34 58 -- --

1500 (TS2) 39 4 NC NC 32 37 -- --

2000 (TS3) NC NC NC NC NC NC -- --

2000 (TS4) NC NC NC NC NC NC -- --

61

Table 4.3. Low Pressure Apparatus (LPA) Fluid Loss Results in (ml) for LCM Mixtures

LCM mixtures G & SCC

(Aston et al. 2004)

G & NS

(Hettema et al. 2007)

G, SCC, & CF

(Kumar et al.

2011)

Blend # 1 2 1 2 1 2

Concn. (ppb) 30 80 30 80 20 40 20 40 55 55

Fracture

Width

(micron)

1000 (TS1) 8 2 5 4 4 5 4 4 7 2

1500 (TS2) NC 22 68 8 8 7 11 7 13 19

2000 (TS3) NC NC NC NC 87 56 NC 54 NC NC

2000 (TS4) NC NC NC NC NC 63 NC 33 NC NC

Table 4.4. Selected Blends for High Pressure Apparatus (HPA) Testing

LCM

Type/Mixture

Blend

#

Concn.

(ppb)

Fracture Width Used (microns)

1000

(TS1)

1500

(TS2)

2000

(TS3)

2000

(TS4)

G 1 15 x x x

G 1 50 x x x

NS 1 15

NS 1 50

SCC 3 50 x x

CF 1 15 x x x

CF 1 50 x x

G & SCC 1 30 x x x

G & SCC 1 80 x x

G & NS 1 20 x x

G & NS 1 40

G, SCC & CF 1 55 x x

4.3. HIGH PRESSURE CONVENTIONAL LCM TESTING

A total of 75 tests were conducted using conventional LCM at different testing

conditions where 45 tests were done with WBM systems and 30 tests with OBM systems.

62

The results are summarized in Table 4.5 - Table 4.7. The tables list the following

for each test: base fluid, fluid density in lb/gal, type of weighting material, LCM blend

type and number, total LCM concentration in ppb, tapered disc used, the maximum

measured sealing pressure in psi and the average fluid loss per cycle. The fluid loss per

cycle was calculated as the total fluid loss collected divided by the number of cycles. A

cycle is here defined as any drop of the pressure that is equal to or greater than 100 psi.

Some tests noted with (R) were repeated to evaluate results repeatability. Results noted

with (HT) are tests that were repeated at higher temperature.

Table 4.5. Summary of HPA Results for Single LCM Treatments using WBM

Test # Base Fluid Density

(lb/gal)

Weighing

Material LCM Blend

Total

Concn.

(ppb)

Fracture

Width

(microns)

Sealing

Pressure

(psi)

Avg. FL

(ml/cycle)

1 WBM 8.6 None G # 1 15 1000 (TS1) 414 8.3

2 WBM 8.6 None G # 1 50 1000 (TS1) 449 7.6

3 WBM 8.6 None NS # 1 15 1000 (TS1) 984 3.5

4 WBM 8.6 None NS # 1 50 1000 (TS1) 2202 2

5 WBM (R) 8.6 None NS # 1 50 1000 (TS1) 2010 27

6 WBM 8.6 None NS # 1 15 1500 (TS2) 1754 11.8

7 WBM 8.6 None NS # 1 50 1500 (TS2) 2027 19.2

8 WBM (R) 8.6 None NS # 1 50 1500 (TS2) 2266 27.6

9 WBM 8.6 None NS # 1 15 2000 (TS3) 347 31

10 WBM 8.6 None NS # 1 15 2000 (TS4) 441 24.5

11 WBM 8.6 None NS # 1 50 2000 (TS4) 755 18.3

12 WBM (HT) 8.6 None NS # 1 50 2000 (TS4) 1164 17.3

13 WBM 8.6 None NS # 1 50 2000 (TS3) 2237 95

14 WBM (R) 8.6 None NS # 1 50 2000 (TS3) 2242 92

15 WBM 8.6 None SCC # 3 50 1000 (TS1) 589 13.3

16 WBM 8.6 None SCC # 3 50 1500 (TS2) 162 31.3

17 WBM 8.6 None CF # 1 15 1000 (TS1) 800 11.8

18 WBM 8.6 None CF # 1 50 1000 (TS1) 2160 0.5

19 WBM 8.6 None CF # 1 50 1500 (TS2) 1129 4.4

63

Table 4.6. Summary of HPA Results for LCM Mixture Treatments using WBM


(lb/gal)

Weighing

Material LCM Blend

Total

Concn.

(ppb)

Fracture

Width

(microns)

Sealing

Pressure

(psi)

Avg. FL

(ml/cycle)

20 WBM 8.6 None G & SCC # 1 30 1000 (TS1) 487 4.1

21 WBM 9.5 Barite G & SCC # 1 30 1000 (TS1) 1205 3.8

22 WBM 11 Barite G & SCC # 1 30 1000 (TS1) 901 7.6


24 WBM 14.5 Barite G & SCC # 1 30 1000 (TS1) 1037 4..4


26 WBM 9.5 Hematite G & SCC # 1 30 1000 (TS1) 861 4.6

27 WBM 12.5 Hematite G & SCC # 1 30 1000 (TS1) 1507 5



30 WBM 8.6 None G & SCC # 1 80 1000 (TS1) 584 2.5

31 WBM (HT) 8.6 None G & SCC # 1 80 1000 (TS1) 537 2.9



34 WBM 8.6 None G & SCC # 1 80 1500 (TS2) 224 8.4


36 WBM 8.6 None G & NS # 1 20 1000 (TS1) 2372 3.1

37 WBM 8.6 None G & NS # 1 40 1000 (TS1) 2892 2.4

38 WBM 8.6 None G & NS # 1 20 1500 (TS2) 805 10.1

39 WBM (HT) 8.6 None G & NS # 1 20 1500 (TS2) 718 7

40 WBM 8.6 None G & NS # 1 40 1500 (TS2) 1713 3

41 WBM (HT) 8.6 None G & NS # 1 40 1500 (TS2) 1190 3

42 WBM 8.6 None G & NS # 1 40 2000 (TS3) 295 32

43 WBM 8.6 None G & NS # 1 40 2000 (TS4) 242 14.2

44 WBM 8.6 None G,SCC, & CF # 1 55 1000 (TS1) 1011 4.7

45 WBM 8.6 None G,SCC, & CF # 1 55 1500 (TS2) 391 14.2

Table 4.7. Summary of HPA Results for LCM Mixture Treatments using OBM


(lb/gal)

Weighing

Material LCM Blend

Total

Concn.

(ppb)

Fracture

Width

(microns)

Sealing

Pressure

(psi)

Avg. FL

(ml/cycle)

46 OBM 8.6 Barite G & NS # 1 20 1000 2398 6.8

47 OBM 8.6 Barite G,SCC, & CF # 1 55 1000 1505 20

48 OBM 8.6 Barite G & SCC # 1 30 1000 737 18.8

49 OBM 9.5 Barite G & SCC # 1 30 1000 1049 6.9

50 OBM 11 Barite G & SCC # 1 30 1000 1050 4

51 OBM 12.5 Barite G & SCC # 1 30 1000 1191 3.3

64

Table 4.7. Summary of HPA Results for LCM Mixture Treatments using OBM (Cont’d)

52 OBM 14.5 Barite G & SCC # 1 30 1000 1214 8.75

53 OBM 16.5 Barite G & SCC # 1 30 1000 1238 5.7

54 OBM 12.5 Hematite G & SCC # 1 30 1000 1269 3.5

55 OBM 14.5 Hematite G & SCC # 1 30 1000 1305 4

56 OBM 16.5 Hematite G & SCC # 1 30 1000 1283 6.7

57 OBM 12.5 Sieved

Barite (C) G & SCC # 1 30 1000 1073 4.4

58 OBM 12.5 Sieved

Barite (M) G & SCC # 1 30 1000 1134 6.4

59 OBM 12.5 Sieved

Barite (F) G & SCC # 1 30 1000 1249 4.4

60 OBM 8.6 Barite G & SCC # 1 80 1500 776 120

61 OBM 11 Barite G & SCC # 3 105 1000 1569 4

62 OBM 11 Barite G & SCC # 3 105 1500 205 47

63 OBM 11 Barite G & SCC # 4 105 1000 117 NC


65 OBM 11 Barite G & SCC # 5 105 1000 1489 4.4


67 OBM 11 Barite G & SCC # 7 105 1500 1708 13.5

68 OBM 11 Barite G & SCC # 7 105 2000 1669 15

69 OBM 14.5 Barite G & SCC # 7 105 2000 2471 6.2

70 OBM 16.5 Barite G & SCC # 7 105 2000 2751 4.4

71 OBM 12.5 Barite G & SCC # 7 105 2000 3097 11.4

72 OBM 12.5 Barite G & SCC # 7 105 2000 3132 11

73 OBM 11 Barite G & SCC # 7 105 2500 1534 17.5



4.3.1. Effect of LCM Type and Fracture Width. To investigate the effect each

individual parameter has on the sealing pressure, the results are plotted and described

with respect to one variable at a time. Four colors are used to define the tapered discs:

blue for 1000 microns (TS1), red for 1500 microns (TS2), orange for 2000 microns (TS3)

with a fracture mouth of 4000 microns, and purple for 2000 microns (TS4) with a fracture

mouth of 5000 microns for Figure 4.14 - Figure 4.18. From the plot of the three high

concentrations (50 ppb) single LCM treatments (Figure 4.14a), nutshells treatments

65

resulted in higher sealing pressures at the TS1 and TS2 fracture widths (up to 2200 psi)

compared to graphite (449 psi) and sized calcium carbonate (589 psi).

Figure 4.14. Effect of LCM Type on the Sealing Pressure for (a) Single LCM Treatments

and (b) LCM Mixtures

For LCM mixtures (Figure 4.14b), treatments containing nutshells and graphite

(40 ppb) exhibited higher sealing pressures up to 2800 psi compared to the other blend

containing graphite and sized calcium carbonate at a higher concentration (80 ppb) using

the two different fracture widths (589 psi and 224 psi for TS1 and TS2, respectively). In

general, increasing the fracture width for both single and LCM mixtures resulted in lower

sealing pressures.

4.3.2. Effect of Concentration. Figure 4.15 gives nutshells LCM concentration

versus sealing pressure for nutshells and the graphite nutshells blend for different fracture

widths. Increasing the concentration of nutshells from 15 ppb to 50 ppb increased the

sealing pressure up to 500% (Figure 4.15a). For graphite and nutshells blends (Figure

(a) (b)

66

4.15b), the sealing pressures were increased when tested using the TS1 (1000 microns)

and the TS2 (1500 microns) by 22% and 113%, respectively. Figure 4.15 shows that

increasing the LCM concentration is increasing the sealing pressure for both single LCMs

and mixtures.

Figure 4.15. Effect of LCM Concentration on the Sealing Pressure for (a) Single LCM

Treatments and (b) LCM Mixtures

4.3.3. Effect of Temperature. Figure 4.16 shows a comparison between four

tests, which were conducted at both room temperature and at an elevated temperature

(HT) of 180 ˚F.

The sealing pressure of an 80 ppb graphite and sized calcium carbonate # 1 was

not affected significantly by the elevated temperature, where a negligible reduction from

584 psi to 537 psi (8%) was observed. For both tests using a 20 and a 40 ppb graphite and

nutshells (G&NS # 1 blend), a reduction in the seal integrity was observed at 180 ˚F by

(b) (a)

67

11% and 30.5%, respectively. However, when a mud with 50 ppb nutshells added was

tested at 180 ˚F, an increase in the seal integrity from 755 psi to 1164 psi (54%) was

observed (Figure 4.16).

Figure 4.16. Effect of Temperature on the Sealing Pressure

The variation in the sealing pressure of the different blends when tested at higher

temperature suggests that temperature has a significant effect on LCM performance. The

reduction in the sealing pressure for graphite and nutshell blends could be attributed to

the increased graphite lubricity at high temperature, since graphite is known for acting as

a solid lubricant (Alleman et al. 1998; Savari et al. 2012).

The increase in the sealing pressure for nutshells blend is linked to the swelling of

nutshells at higher temperature, which was experimentally confirmed above (Figure

4.13).

0

500

1000

1500

2000

Sealin

g P

ressu

re (

psi)

LCM Type

68

4.3.4. Effect of Particle Size Distribution. In order to understand the effect of

PSD on the sealing pressure, the first step was to examine test results for individual

fracture widths separately. The sealing pressures for each fracture width are plotted on

the left y-axis versus the particle size distribution obtained from the dry sieve analysis on

the right y-axis as shown in Figure 4.17 and Figure 4.18.

Figure 4.17 shows the sealing pressure versus the particle size distribution for

single LCMs and LCM mixtures using the 1000 microns fracture width disc. For blends

containing graphite, calcium carbonate or a mixture of the two, the sealing pressures were

around 500 psi at both low and high concentration. For blends with nutshells, the

measured sealing pressures were 984 psi and 2200 psi for the 15 ppb and the 50 ppb

blends, respectively. When graphite was combined with nutshells (Figure 4.17b), the

sealing pressures were increased to 2800 psi.

Figure 4.17. Effect of PSD on the Sealing Pressure using 1000 microns Fracture Width

(TS1) for (a) Single LCM Treatments and (b) LCM Mixtures

(a) (b)

69

The particle size distribution of the different blends (Table 4.1) shows that

nutshells particles are coarser than both graphite and sized calcium carbonate particles.

The PSD of the single LCM nutshell treatments ranged between 180 – 2400 microns

while the PSD of graphite and sized calcium carbonate ranged between 60 – 1300

microns and 250 – 1200 microns, respectively.

Figure 4.18 shows the sealing pressures versus the particle size distribution for

single LCM treatments and LCM mixtures when tested at 1500 microns fracture width.

Both calcium carbonate blend and the mixture of graphite and sized calcium carbonate

resulted in sealing pressures lower than 250 psi. For blends containing nutshells only, the

sealing pressures were 1754 psi and 2027 psi for the 15 ppb and the 50 ppb blends,

respectively. The 20 ppb and the 40 ppb graphite and nutshell mixtures gave sealing

pressures of 805 psi and 1713 psi, respectively.

Figure 4.18. Effect of PSD on the Sealing Pressure using 1500 microns Fracture Width

(TS2) for (a) Single LCM Treatments and (b) LCM Mixtures

(a) (b)

70

The presented results in both Figure 4.17 and Figure 4.18 imply that both single

LCMs and LCM blends with coarser materials gave higher sealing pressure. In other

words, the wider the range of particle sizes increases the ability in sealing fractures and

improves the sealing pressure. The tests that contained nutshells exhibited higher sealing

pressures which indicate that the type of material is also influencing the results.

To qualitatively understand the overall trend of how PSD could affect the sealing

pressure with respect to fracture width, all sealing pressure results in chapter 4.3 were

plotted (black dots) regardless of base fluid, concentration, or LCM type and sorted with

the highest pressure to the right (Figure 4.19 - Figure 4.21). The black dashed line is

plotted as a reference line that represents the fracture width in microns. The quadratic

regression fit lines for the five D-values as well as the sealing pressure were plotted in

different colors, where D90 (red solid line), D75 (orange dashed line), D50 (blue dashed

line), D25 (green dashed line), and D10 (red dashed line).

Figure 4.19. PSD vs Sealing Pressure for 1000 microns Fracture Width

3000

2500

2000

1500

1000

500

0

3000

2500

2000

1500

1000

500

0

Se

ali

ng

Pre

ssu

re

(p

si)

Part

icle

Siz

e D

istr

ibu

tio

n(m

icro

ns)

Sealing Pressure (R 2̂ = 93.8%)

D10 (R 2̂ = 8.6%)

D25 (R 2̂ = 27.9%)

D50 (R 2̂ = 20%)

D75 (R 2̂ = 40%)

D90 (R 2̂ = 35%)

71

Figure 4.19 shows the sealing pressures (left y-axis) versus PSD (right y-axis) for

1000 microns fracture width. It is noticeable that higher sealing pressures are achieved

with wider PSD; however no clear correlation that suggests the effect of a specific D-

value could be observed.

Figure 4.20 and Figure 4.21 show the sealing pressure results versus PSD for

1500 microns and 2000 microns fracture widths, respectively. In general, the sealing

pressure results tend to increase as the particle size distribution gets coarser, i.e. wider

PSD. In other words, blends with higher D-values resulted in higher sealing pressures.

The D-values fit lines shows an upward increase with the pressure.

Even though the observed PSD fit lines shows an increase with the pressure, no

clear conclusion could be made on the significance of each D-value contribution toward

enhancing the sealing pressure from the results.


3000

2500

2000

1500

1000

500

0

3000

2500

2000

1500

1000

500

0

Sea

lin

g P

res

su

re (

ps

i)

Pa

rtic

le S

ize

Dis

trib

uti

on

(mic

ron

s)


D10 (R 2̂ = 39.6%)

D25 (R 2̂ = 44%)

D50 (R 2̂ = 56.1%)

D75 (R 2̂ = 64.3%)

D90 (R 2̂ = 78.1%)

72


4.3.5. Effect of Base Fluid. In this subsection, different blends mixed with water

and oil base fluid were analyzed to investigate the effect of base fluids on LCM

performance. Figure 4.22 shows the effect of base fluid for different blends with a

variation in the total concentration between 20 – 105 ppb. All blends were tested at 1000

microns fracture width (TS1) using both; OBM and WBM except for graphite and sized

calcium carbonate # 7, which was tested at 1500 (TS2).

The sealing pressure of both blends; graphite, sized calcium carbonate, and

cellulosic fiber # 1 and graphite and sized calcium carbonate # 1 were increased by

approximately 50% when used in OBM. However, graphite and nutshells # 1 blend was

not affected by the change of base fluid. For the three LCM mixtures at relatively high

concentration (105 ppb), graphite and sized calcium carbonate blends # 3, 5, and 7,

higher sealing pressures were observed when used in WBM. An increase of 30%, 92%

and 50% was observed for graphite and sized calcium carbonate blends # 3, 5, and 7,

when using WBM.

3500

3000

2500

2000

1500

1000

500

0

3000

2500

2000

1500

1000

500

0

Se

ali

ng

Pre

ssu

re

(p

si)

Part

icle

Siz

e D

istr

ibu

tio

n(m

icro

ns)


D10 (R 2̂ = 67.7%)

D25 (R 2̂ = 74.8%)

D50 (R 2̂ = 77%)

D75 (R 2̂ = 73%)

D90 (R 2̂ = 83.7%)

73

Figure 4.22. Effect of Base Fluid on Low and High Concentration LCM Mixtures

The difference in the effect of base fluid on low LCM mixtures versus the high

concentration LCM mixtures is attributed to the difference in rheological properties of

WBM and OBM. WBM tested has a yield point (YP) equal to 19 lb/100 ft2 (Table 3.1)

which makes is it more capable of suspending higher LCM concentrations. LCMs were

suspended in WBM and plugged the slot continuously. However, the OBM had a lower

YP of 14 lb/100 ft2 compared to WBM, which could have possibly resulted in LCM

sagging at the bottom of the cell.

4.3.6. Effect of Density. To understand the effect density, the 30 ppb graphite

and sized calcium carbonate blend # 1 was used in all tests for this step at different

densities. The density of drilling fluids was increased using barite for both WBM and

OBM. Since the pre-mixed OBM is already weighted to 11 lb/gal, the density was

reduced by adding the base fluid (EDC95-11).

Figure 4.23 shows the effect on the sealing pressure when varying the density for

the 30 ppb graphite and sized calcium carbonate blend # 1 when tested with the 1000

0

500

1000

1500

2000

2500

3000

3500

G & NS(20 ppb)

G,SCC,CF # 1(55 ppb)

G & SCC # 1(30 ppb)

G & SCC # 3(105 ppb)

G & SCC # 5(105 ppb)

G & SCC # 7(105 ppb)

Sealin

g P

ressu

re (

psi)

LCM Blend

WBM OBM

74

microns fracture disc using both WBM and OBM and for the 105 ppb graphite and sized

calcium carbonate blend # 7 when tested at 2000 microns fracture disc using OBM.

Figure 4.23. Effect of Increasing the Mud Density in OBM and WBM

For the 1000 microns fracture tests, the sealing pressure in OBM was increased by

approximately 68% from 737 psi to 1238 psi when the fluid density was increased from

8.6 lb/gal to 16.5 lb/gal. A higher increase of the sealing pressure by 175 % was observed

for the same blend mixed in WBM when the fluid was weighted from 8.6 lb/gal to 16.5

lb/gal.

For the 2000 microns using the graphite and sized calcium carbonate blend # 7 in

OBM, the sealing pressure was also increased by approximately 65% from 1669 psi for

the 11 lb/gal mud to 2751 psi when the fluid density was increased to 16.5 lb/gal.

Therefore, increasing the density of the drilling fluid caused the sealing pressure to

increase.

0

500

1000

1500

2000

2500

3000

3500

8 10 12 14 16 18

Sealin

g P

ressu

re (

ps

i)

Mud Density (lb/gal)

WBM (1000 microns)OBM (1000 microns)OBM (2000 microns)

75

4.3.7. Effect of Weighting Material. The 30 ppb graphite and sized calcium

carbonate blend #1 was tested using 1000 microns fracture width and with drilling fluids

weighing from 9.5 lb/gal to 16.5 lb/gal using either barite and hematite as weight

material.

Figure 4.24 shows the effect of the weighting material type on the sealing

pressure for OBM and WBM. The results show that increasing the density improved the

sealing pressure for both WBM and OBM. Using hematite as the weight material resulted

in higher sealing pressures compared to barite, especially for WBM.

Figure 4.24. Effect of Weighting Materials Type in OBM and WBM

4.3.8. Effect of LCM Shape. In this paragraph the variation in sealing pressure

results is evaluated based on the qualitative analysis of LCM particle shapes (Chapter

4.1.3.1). The angularity and low sphericity of nutshells resulted in higher sealing

pressures (up to 2800 psi) due to higher shear strength of the formed plug. Angularity and

0

200

400

600

800

1000

1200

1400

1600

1800

2000

8 10 12 14 16 18

Sealin

g P

ressu

re (

psi)

Mud Density (ppg)

Barite (OBM)

Hematite (OBM)

Barite (WBM)

Hematite (WBM)

76

sphericity was found to affect the angle of internal friction, thus affecting the degree of

mechanical interlocking among particles (Shinohara et al. 2000, Rong et al. 2013). The

higher angularity of nutshell particles seems to create higher mechanical interlocking

between the particles at the fracture tip, thus higher pressures were required to break the

formed seal. While the high sphericity of both graphite and calcium carbonate resulted in

lower sealing pressures at low concentrations due to the higher tendency of spherical

particles to flow through the fracture tip.

The variation in nutshell particle sizes enhanced the sealing capabilities compared

to graphite and calcium carbonate, which are moderately sorted. The asphericity of

cellulosic fiber also helped in a better sealing pressure compared to graphite and sized

calcium carbonate at the same testing conditions, i.e. fracture width and concentration

(See Test # 1, 2, 15, 17, 18, and 19 in Table 4.5.)

4.4. HIGH PRESSURE UNCONVENTIONAL LCM TESTING

A total of 10 tests were conducted using the unconventional LCM at different

testing condition where 8 tests with WBM and 2 tests with the low-toxicity mineral

OBM. The results are summarized in Table 4.8.

The tables list the following for each test: base fluid, fluid density in lb/gal, type

of weighting material, tapered disc used, injection rate used to initiate the seal, injection

rate used to evaluate the formed plug, the maximum measured sealing pressure in psi and

the total fluid loss collected. While conventional LCMs weren’t able to seal fractures

wider than 2500 microns, this unconventional system was able to seal fracture widths up

to 5000 microns.

77

Table 4.8. HPA Results for Foam Wedges Based System

Test

#

Base

Fluid

Density

(lb/gal)

Weighing

Material

Fracture

Width

(microns)

Injection

Rate

before

Seal

Initiation

(ml/min)

Injection

Rate

after

Seal

Initiation

(ml/min)

Sealing

Pressure

(psi)

Fluid

Loss

(ml)

1 Water Unweighted None 1500 (TS2) 5 10 2571 180




5 Water Unweighted None 3000 (SS1) 25 25 340 190




9 EDC95-11 12.5 Barite 1500 (TS2) 5 10 78 325

10 EDC95-11 12.5 Barite 5000 (TS7) 5 10 69 350

The seal is formed by slow filtration of the carbonate particles through the wedge

foams trapped in the fracture. A slower process (i.e. hesitation squeeze) is required for

initiating a seal. A high flow rate can break the seal before it becomes strong enough to

hold an increasing differential pressure. Lower sealing pressures were observed for

injection rates higher than 5 ml/min (Table 4.8). Tests with injection rate of 5 ml/min

showed higher seal integrity and therefore, an injection rate of 5 ml/min was used for the

tests. The effect of the injection rate after initiating a good seal was not significant. There

is only a certain amount of fluid that can be squeezed out of the slurry for each test. The

lower fluid loss values at the lower injection rate indicates that the seal started to form

earlier within the fracture.

78

Figure 4.25 and Figure 4.26 show the formed seals when used in WBM and

OBM, respectively. The unconventional LCM performed good in WBM with sealing

pressures up to 2800 psi using 3000 microns straight slotted disc, which was damaged as

a result of the high pressure Figure 4.27. However, when EDC95-11 (Low-toxicity

mineral oil) was used as the base fluid, the system failed to seal both 1500 and 5000

microns fracture widths. For this reason, the system wasn’t evaluated at wider fracture

width with the OBM.

Figure 4.25. Wedge Foam in WBM Formed Plug using 5 mm disc; (left) Side View,

(right) Bottom View

Figure 4.26. Wedge Foam in OBM Formed Plug using 1500 microns disc; (left) Side

View, (right) Bottom View

79

Figure 4.27. Damaged Slotted Disc

4.5. HYDRAULIC FRACTURING TEST

Five concrete cores were fractured with different LCM blends and the pressure

versus time was recorded and breakdown and re-opening pressure interpreted. The first

test was conducted using EDC95-11, which is a solid free clear base fluid used to prepare

the pre-mixed OBM. The second test was conducted with the pre-mixed OBM without

LCM to serve as a control sample. Tests # 3, 4, and 5 were conducted using the pre-

mixed OBM in addition to 3 different LCM mixtures.

4.5.1. Test # 1 with EDC95-11 Base Fluid. Figure 4.28 shows the pressure

versus time plot for the first hydraulic fracturing test, which was conducted using a solid

free base fluid (EDC95-11) with no LCMs. The blue line represents the first injection

cycle, which is used to estimate the breakdown pressure. The red line represents the

second injection cycle, which is used to estimate the re-opening pressure after the 10

minutes fracture healing period. The breakdown pressure was found to be 843 psi while

the re-opening pressure was 571 psi. The fracture widths were measured under optical

microscope (Figure 4.29) and the width ranged between 15 - 22 microns.

80

Figure 4.28. Pressure vs. Time for Test # 1

Figure 4.29. The Resulting Fracture under Microscope (200x Magnification) for Test # 1

Showing Fracture Widths Ranging between 15 – 22 microns

4.5.2. Test # 2 with 11 lb/gal OBM without LCM. Test 2 was conducted with

the pre-mixed OBM and no LCMs were added. The breakdown pressure was 2008 psi

and the re-opening pressure was 592 psi (Figure 4.30). Two vertical fractures were

observed (Figure 4.31). The fractured core sample was split to take a cross-section view

of the core. The measured fracture widths ranged between 27 - 36 microns (Figure 4.32).

0

200

400

600

800

1000

1200

1400

1600

0 500 1000 1500 2000

Pre

ssu

re (

psi)

Time (s)

1st Cycle

2nd Cycle

Breakdown @ 843 psi

Re-Opening @ 571 psi

1 mm

Up Up

1 m

81


Figure 4.31. Fractured Core Showing Two Vertical Fractures and a Cross-sectional View

of the Fractured Core for Test # 2

0

500

1000

1500

2000

2500

0 500 1000 1500 2000

Pre

ssu

re (

psi)

Time (s)

1st Cycle

2nd CycleBreakdown @ 2008 psi


82



4.5.3. Test # 3 with 11 lb/gal OBM + 20 ppb Graphite and Nutshells. In this

test, 20 ppb graphite and nutshells # 1 was added to the pre-mixed OBM. The breakdown

and re-opening pressures were 2199 psi and 1334 psi, respectively (Figure 4.33).


0

500

1000

1500

2000

2500

0 500 1000 1500 2000

Pre

ssu

re (

psi)

Time (s)

1st Cycle



1 mm

Up

83

Figure 4.34 shows the resulting two vertical fractures and a cross-section of the

fracture core. The measured fracture widths ranged between 40 - 85 microns (Figure

4.35).





1 mm

Up

84

4.5.4. Test # 4 with 11 lb/gal OBM + 30 ppb Graphite and Sized Calcium

Carbonate. Figure 4.36 shows the pressure versus time for test # 4 where 30 ppb

graphite and calcium carbonate was added to the pre-mixed OBM. The breakdown

pressure was 2309 psi and the re-opening pressure was 1834 psi. Figure 4.37 shows the

resulting two vertical fractures and a cross-section of the fracture core in test # 4. The

measured fracture widths ranged between 40 - 85 microns (Figure 4.38).




0

500

1000

1500

2000

2500

3000

-100 400 900 1400

Pre

ssu

re (

psi)

Time (s)

1st Cycle2nd Cycle

Breakdown @ 2309 psi


85



4.5.5. Test # 5 with 11 lb/gal OBM + 55 ppb Graphite, Sized Calcium

Carbonate, and Cellulosic Fiber. Figure 4.39 shows the hydraulic fracturing results

using 55 ppb graphite, sized calcium carbonate, and cellulosic fiber # 1. The breakdown

and re-opening pressures were 2372 psi and 1717 psi, respectively.


0

500

1000

1500

2000

2500

3000

0 500 1000 1500

Pre

ssu

re (

psi)

Time (s)

1st Cycle



Up Up

1 mm 1 mm

86

The fractured core is shown in Figure 4.40, where two vertical fractures were

observed. The measured fracture widths under the microscope (Figure 4.41) were ranging

between 50 – 140 microns.

Figure 4.40. Fractured Core Showing Two Vertical Fractures for Test # 5

Figure 4.41. The Resulting Fracture under Microscope (200x magnification) for Test # 5


1 mm

Up Up

1 mm

87

4.5.6. Hydraulic Fracturing Results Comparison. The effect of adding LCMs

on the breakdown pressure and the re-opening pressure was calculated as percentage

increase based on the results obtained from the control sample of pre-mixed oil based

drilling fluid (Test # 2). The percent (%) increase in the breakdown and re-opening

pressures were calculated based on the control sample (Test # 2) breakdown pressure of

2008 psi and re-opening pressure of 592 psi as shown in Table 4.9.

When comparing the solid free oil base fluid (EDC95-11) from Test # 1 with the

pre-mixed OBM control sample (Test # 2), the breakdown pressure was reduced to 58%

and re-opening was reduced by 4%. The reduction in the breakdown pressure for the

solid free oil base fluid is caused by to the lack of formation of a filter cake around the

wellbore.

For Test # 3 with the 20 ppb graphite and nutshells blend enhanced the

breakdown pressure by approximately 10% while the re-opening pressure was enhanced

by 125% compared to the control sample (Test # 2). The enhancement in both the

breakdown and the re-opening pressure is attributed to the formation of a stronger filter

cake with the presence of LCMs. The increase suggests also that LCMs sealed the first

fracture, which occurred in the first cycle, and improved the wellbore integrity resulting

in higher re-opening pressure in the second injection cycle.

The 30 ppb graphite and sized calcium carbonate blend in Test # 4 resulted also in

enhancing both the breakdown and re-opening pressure by 15% and 210%, respectively.

For Test # 5 which contained 55 ppb graphite, sized calcium carbonate, and cellulosic

fiber resulted in the highest enhancement in the breakdown pressure by 18%. The re-

opening pressure increase also by 190% compared to the control sample (Test # 2).

88

Table 4.9. Summary of the Hydraulic Fracturing Experiments for the Concrete Cores

Test

# LCM Blend

Concn.

(ppb)

Breakdown

Pressure

(psi)

% Inc. in

Breakdown

Pressure

Re-opening

Pressure

(psi)

% Inc. in

Re-Opening

Pressure

1 None 0 843 -58 571 -4

2 None 0 2008 N/A 592 N/A

3 G&NS 20 2199 10 1334 125

4 G & SCC 30 2309 15 1834 210

5 G, SCC, & CF 55 2375 18 1717 190

4.6. DATA ANALYSIS

4.6.1. Statistical Analysis. Figure 4.42 - Figure 4.50 show the Effect Leverage

plots for each parameter with the resulting P value. The blue dashed line represents the

mean sealing pressure, the solid red line represents the fitted model, and the dashed red

line represents the confidence interval (5% confidence level). If the mean sealing

pressure is inside the confidence interval envelope the parameter does not have any

significant effect on sealing pressure. If the confidence interval crosses the mean pressure

at a high angle, it has a significant contribution to sealing pressure.

Figure 4.42 shows the effect of fracture width in the sealing pressure. The effect

of fracture width is very significant since the confidence interval curves crossed the mean

pressure with a high slope. The P-value of 0.0003 also indicates that the fracture width

influenced the predicted sealing pressure. The effect of fluid density (Figure 4.44) was

also pronounced since confidence curve crossed the horizontal line with a P-value that is

less than 0.05 suggesting a good correlation.

89

Figure 4.42. Leverage Plot Showing the

Effect of Fracture Width on Sealing

Pressure


Effect of Density on Sealing Pressure

The effect of D90 was the third significant parameter to affect the prediction of

sealing pressure. The Effect of D90 can be clearly seen (Figure 4.44) with a P-value that

is slightly larger than 0.05. The variation in LCM blends showed also a clear effect with

P-value 0.1147 (Figure 4.45)


Effect of D90 on Sealing Pressure


Effect of LCM Blend on Sealing Pressure

90

The Effect Leverage plots (Figure 4.46, 4.48, 4.49, 4.50, and 4.51) for D75, base

fluid, D25, D50, and D10, respectively shows less effect on the sealing pressure with P-

values ranging between 0.2863 and 0.9817. However, the less significance of these

parameters might be due to the outliers in the analyzed data set.




Effect of Base Fluid on Sealing Pressure





91

Figure 4.50. Leverage Plot Showing the Effect of D10 on Sealing Pressure

A predictive linear fit model (equation given in Appendix E) is shown in Figure

4.51 shows a good correlation with an R2 of 80% and a P-value less than 0.05. The

residual plot (Figure 4.52) showed the data being randomly distributed around x-axis,

verifying that a linear model was appropriate for the collected data. Table 4.10

summarizes the P-values for the different variables as well as the model fit R2.


Actual Sealing Pressure versus the

Predicted Sealing Pressure using the fit

model

Figure 4.52. Residual Plot of Sealing

Pressure versus the Predicted Sealing

Pressure

92

Table 4.10. Effect of Different Parameters on the Sealing Pressure and Model Fit

Parameter Unit P-Values

Fracture Width (microns) 0.00028

Density (lb/gal) 0.00267

D90 (microns) 0.05957

LCM Blend N/A 0.11473

D75 (microns) 0.28628

Base Fluid (WBM/OBM) 0.30786

D25 (microns) 0.60427

D50 (microns) 0.75352

D10 (microns) 0.98169

Model Fit

R2 0.8

From the statistical analysis, the sealing pressure was highly dependent on the

different parameters in the following order: fracture width, fluid density, D90, LCM

blend/type, D75, base fluid, D25, D50, and D10. Out of these parameters, the fracture

width cannot be controlled and the fluid density should be designed based the mud

weight window.

4.6.2. Analytical Fracture Models. In this section, the five analytical fracture

models presented in subchapter 2.3 were used to estimate the fracture width based on the

hydraulic fracturing test results for the different blends and the mechanical properties of

concrete cores. The estimated fracture widths are then compared to the measured fracture

width from the microscopic images of the fractured cores.

Table 4.11 shows the input parameters used to estimate the fracture width from

the five analytical models. The applied confining pressure in the hydraulic fracturing test

was used for both the maximum and minimum horizontal stress. Both Young’s modulus

93

and Poisson’s ratio were measured previously (Weideman, 2014) for a cured class-H

Portland cement mixed and hardened following the procedure outlined in subchapter

3.3.1. The radius of the fractured core samples was used as the fracture length since the

fracture reached the end of the cores. The pressure along the fracture was set to zero

similar to previous studies (Morita et al. 2012).

Table 4.11. Analytical Models Input Parameters

Parameter Value Unit

Minimum horizontal stress σH 100 psi

Maximum horizontal stress σh 100 psi

Young's Modulus E 1450377 psi

Poisson’s ratio v 0.25

Fracture length L 2.65 inches

Wellbore radius rw 0.275 inches

Pressure along the fracture Pf 0 psi

The calculated fracture widths using the five analytical models and the range of

the measured fracture width under the microscope are tabulated in Table 4.12. It is clear

that both Hillerborg’s and Wang’s models gave unrealistically small fracture widths

compared to the actual measured widths. Alberty’s model gave relatively large fracture

width compared to the measure widths. Carbonell’s model gave fracture widths that are

comparable to the measured widths. The estimated fracture widths by Morita’s model

were constant for all different cases since the model is not a function of the wellbore

pressure.

94

Table 4.12. The Estimated Fracture Width from the Analytical Models and the Measured

Width under Microscope for the Five Fractured Core Samples

Fracture Estimation Method Fracture Width (microns)

Test # 1 Test # 2 Test # 3 Test # 4 Test # 5

Measured under Microscope 15 - 22 27 - 36 40 - 85 40 - 145 50 - 140

Hillerborg et al. (1976) 2.54 7.62 8.1 8.5 8.77

Carbonell and Detournay (1995) 22.86 29.46 30.48 31.31 31.7

Alberty and McLean (2004) 144.78 368.3 403.86 424.18 436.88

Wang et al. (2008) 0.082 0.2 0.21 0.21 0.23

Morita and Fuh (2012) 19.94 19.94 19.94 19.94 19.94

Based on the measured fracture width for the tests using the pre-mixed OBM

containing LCM with the PSD analysis of the tested blends (Chapter 4.1.2), we can see

that the fracture widths are within the range of the D10 and D25.

95

5. DISCUSSION

In this chapter, the experimental results and observations are discussed in details.

The results were also compared with previous results. The discussion is divided into 3

subchapters as follows; LCM performance evaluation, hydraulic fracturing experiment

results, and PSD selection criteria.

5.1. LCM PERFORMANCE EVALUATION

Figure 5.1 shows the low pressure (LPA) fluid loss results (circle) of the 26 tests

using the selected blend (Table 4.4) plotted with the average fluid loss per cycle (dash)

values calculated for the same blends under high pressure.

Figure 5.1. Comparison between Fluid Loss Values from the LPA and the Calculated

Average Fluid Loss per Cycle from the HPA

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Flu

id L

oss (

ml)

Test #

Fluid Loss (ml) @ Low Pressure

Fluid Loss (ml) Per Cycle @ High Pressure

96

The majority of the LPA results in Figure 5.1 were comparable to the average

fluid loss per cycle, which could be attributed to a fixed amount of fluid lost prior to the

formation of a seal within the fracture. However, some fluid losses per cycle values

(circled on Figure 5.1) were significantly higher than the LPA values. This is believed to

be due to the continuous injection of drilling fluid through a permeable seal that did not

seal the fracture effectively. The high seal permeability resulted from the formation of a

thick plug due to the high concentration of larger LCM particles, which screened out at

the fracture surface (Figure 5.2) rather than plugging the slot. This comparison also

highlights the effect of higher LCM concentrations on the fluid loss. This observation

suggests that selecting the right LCM concentration is critical to avoid screen out at the

fracture openings, however further investigation is required to define the critical

concentration of LCM.

Figure 5.2. High Concentration of Nutshells (Left) and Graphite (Right) Treatments

Screened out at the Fracture Surface

After running HPA tests and when cleaning the used tapered discs, it was hard to

remove the formed seals when nutshells were used because the particles were stuck at the

97

fracture tip (Figure 5.3). This indicates a degree of deformability and strength of nutshells

particles that resulted in higher sealing pressure compared to other LCM treatments at the

same concentration.

Figure 5.3. Plugged Fracture tip of a 2000-micron Tapered Disc showing Stuck Nutshells

Particles

The sealing pressures versus the average fluid loss per cycle for the 26 tests

mentioned above were plotted in Figure 5.4, 5.5, and Figure 5.6 for 1000 microns, 1500

microns and 3000 microns fracture width, respectively. The results are plotted and circled

in groups from high to low sealing pressure.

From Figure 5.4 for tests using 1000 microns fracture width (TS1), it can be

observed that the results were distributed in three distinguished regions. In region (a),

higher sealing pressures were observed at lower fluid loss values <5 ml at small fracture

widths. While test (b) shows an evidence of LCM screen out at the fracture mouth, which

resulted in higher fluid loss but with high sealing pressure. Region (c) shows lower

sealing pressures with relatively higher fluid loss values compared to region (a).

98

Figure 5.4. Sealing Pressure vs. Average Fluid Loss per Cycle using 1000 microns

Fracture Width (TS1)

Figure 5.5 shows the sealing pressures versus the average fluid loss per cycle

using the 1500 microns fracture width (TS2).


Fracture Width (TS2)

0

500

1000

1500

2000

2500

3000

3500

4000

0 5 10 15 20 25 30

Sealin

g P

ressu

re (

psi)

Avg. Fluid Loss per Cycle (ml)

50 ppb CF 15 ppb CF 40 ppb G & NS20 ppb G & NS 50 ppb NS 15 ppb NS55 ppb G, SCC & CF 50 ppb SCC 80 ppb G & SCC30 ppb G & SCC 50 ppb G 15 ppb G

(b)

(a)

(c)

0

500

1000

1500

2000

2500

3000

0 5 10 15 20 25 30 35

Sealin

g P

ressu

re (

psi)


50 ppb NS 15 ppb NS 40 ppb G & NS

50 ppb CF 20 ppb G & NS 55 ppb G, SCC & CF

80 ppb G & SCC 50 ppb SCC

(b)

(a)

(c)

(d)

99

The figure above (Figure 5.5) shows 4 distinguished regions, where region (a)

contains high sealing pressures in conjunction with low fluid loss values similar to the

one observed in Figure 5.4. Region (b) shows high sealing pressure but with high fluid

loss values that are higher than 10 ml/cycle, while region (c) exhibited moderate sealing

pressures and fluid loss values. Test (d) shows a very low sealing pressure as well as high

fluid loss per cycle.

Figure 5.6 shows the sealing pressure versus the average fluid loss per cycle for

tests conducted using 2000 microns fractures (TS3 and TS4). In this figure, only two

regions were observed. Region (a) containes moderate sealing pressures with higher fluid

loss values compared to the results at smaller fractures, while region (b) exhibted high

sealing pressures with a very high fluid loss values.


Fracture Width (TS3 and TS4)

0

500

1000

1500

2000

2500

3000

0 20 40 60 80 100

Sealin

g P

ressu

re (

psi)


50 ppb NS @TS3 50 ppb NS @TS4

15 ppb NS @ TS4 40 ppb G & NS @ TS4

15 ppb NS @ TS3 40 ppb G & NS @TS3

(b)

(a)

100

Based on these observation of sealing pressures versus fluid loss per cycle at the

different fracture widths, it can be concluded that lower fluid loss values will result in

higher sealing pressures at fractures up to 1000 microns. However, the higher fluid loss

values and higher sealing pressures at fractures wider than 1000 microns suggest that de-

fluidization of the LCM treatments was required to form a strong seal.

The results of graphite and sized calcium carbonate mixtures and graphite and

nutshells mixtures were compared with the results published in Aston et al. 2004 and

Hettema et al. 2007. Aston et al. (2004) showed that both the 30 ppb and 80 ppb were

successful for field application in terms of enhancing the fracture gradient. The 80 ppb

graphite and sized calcium carbonate enhanced the break down pressure by

approximately 850 psi when applied as a stress cage treatment in a shale formation

(Arkoma basin). Due to the limited information about their laboratory test methodology,

it is difficult to establish a comparison between their findings and this study.

From the results published by Hettema et al. (2007) for the permeable fracture test

using the 1000 microns open fracture tip (Table A.2), which is comparable to the test

conditions in this study, higher sealing pressure were observed up to 4420 psi. Only one

test (Test # 10 in Table A.2) using the 40 ppb graphite and nutshells LCM blend in a

synthetic based mud (SBM) showed a comparable sealing pressure (2037 psi) compared

to our results, where 2398 psi and 2372 psi were observed in OBM and WBM,

respectively (Test # 36 in Table 4.6 and Test # 46 in Table 4.7). For the 20 ppb graphite

and nutshells blend, no tests were conducted at 1000 microns fractures by Hetema et al.

(2007) and therefore the results cannot be compared.

101

Hettema et al. (2007) concluded based on two tests using WBM and synthetic-

base mud (SBM) (Test # 6 and 7 in Table A.2) that the base fluid is largely independent

factor that does not have an effect on the sealing pressure. Our investigation showed that

graphite and nutshells blend was not affected by the base fluid; however, the effect of

base fluid on the sealing pressure for blends using graphite and sized calcium carbonate

was significant (Figure 4.22).

The effect of density on the sealing pressure was also studied by Hettema et al.

(2007) where 3 tests were repeated at 3 different densities. It was concluded that higher

density will reduce both fluid loss and sealing pressure. However, a different effect was

observed in our study (Figure 4.23 and Figure 4.24) where higher densities resulted in

improving the sealing pressure.

The laboratory investigation by Kumar et al (2011) was conducted on the particles

plugging apparatus (PPA) using one tapered disc size (1000 microns) where LCM

performance was based on the ability of LCM blends in forming an impermeable seal in

the tapered disc. However, no information about the applied pressure was provided.

Therefore, no comparison for the sealing pressures could be established with their results.

It was concluded from their study that sized calcium carbonate and graphite blends were

not very successful in arresting fluid loss which is in agreement with what was observed

in this study for similar blends.

To investigate the reason behind the better performance of hematite compared to

barite (Subchapter 4.3.7), particle size distribution analysis was conducted for both barite

and hematite using laser diffraction technique to investigate the PSD differences.

102

Figure 5.7 shows the PSD analysis for barite (black) and hematite (red). It can be

seen that hematite has slightly coarser distribution compared to barite, which could have

resulted in improving the sealing pressure. Table 5.1 shows the PSD analysis of the used

weighting materials compared to the PSD conducted by Tehrani et al. 2014 for API size

barite and hematite, which is relatively comparable.

Figure 5.7. Particle Size Distribution for the used Barite and Hematite using Laser

Diffraction Technique

Table 5.1. Particle Size Distribution Analysis Comparision for Barite and Hematite

PSD (microns)

Weighting Material Barite Hematite

Measured

D10 1.8 6

D50 13 23

D90 50 70

Taken from

Tehrani et al.

2014

D10 3 3

D50 21 27

D90 64 68

1001010.1

100

90

80

70

60

50

40

30

20

10

0

Size (microns)

% F

ine

r

D 90

D 50

D 10

Barite

Hematite

103

To investigate the effect of sieving barite on the sealing pressure, a sample of

barite was sieved into three grades; fine (F), medium (M), and coarse (C). Cutoffs for

defining the three ranges of sizes were made up as follows:

a) Particles equal to or smaller than 50 microns were considered fine.

b) Particles > 50 microns < 90 microns were considered medium.

c) Particles larger than 90 microns were considered coarse.

The sieved samples were used to weight up OBM containing 30 ppb graphite and

sized calcium carbonate blend # 1 to 12.5 lb/gal. Figure 5.8 shows the sealing pressure

when using different grades of the sieved barite. The fine barite resulted in increasing the

sealing pressure by only 5% while using the cores and medium barite resulted in reducing

the sealing pressure by approximately 10% and 5%, respectively. The reduction in the

sealing pressure as a result of using course and medium grade (as per the definition

above) barite highlights the effect of finer particles in improving the sealing pressure.

Figure 5.8. Effect of Sieving Barite in OBM

950

1000

1050

1100

1150

1200

1250

1300

No Sieve Coarse Medium Fine

Sealin

g P

ressu

re (

psi)

Sieved Barite

104

5.2. HYDRAULIC FRACTURING EXPERIMENTS

From the hydraulic fracturing experiments on concrete cores, it was observed that

in general the addition of LCM to fracturing fluids enhanced both the breakdown and re-

opening pressures compared to the results obtained from the same fluid that has no LCMs

(Table 4.9). However, when comparing the results of the fractured core with the solid

free clear fluid to the one that has been pre-mixed and weighted with barite, only the

breakdown was significantly enhanced by 138% while the re-opening improved slightly

by 4%.

Similar hydraulic fracturing experiments were conducted previously to investigate

the effect of adding nanoparticles and graphite to an OBM on the breakdown and re-

opening pressure of shale cores (Contreras et al. 2014b). The same test condition such as

the overburden pressure, confining pressure, and injection rate were applied to fracture

Catoosa shale cores. Two injection cycles were also applied allowing 10 min for fracture

healing after the first cycle.

A summary of 7 hydraulic fracturing tests results, using different fluid

formulation (Table 5.2), with the % increase in breakdown and re-opening pressures as a

result of adding nanoparticles is tabulated in Table 5.3.

Table 5.2. Description of the Different Fluid used in the Fracturing Experiments

LCM

Blend Description

CS Control sample without NPs or graphite

G1 Graphite blend # 1

G2 Graphite blend # 2

105

Table 5.2. Description of the Different Fluid used in the Fracturing Experiments (Cont’d)

DF1 Graphite and Iron-based NPs blend # 1

DF2 Graphite and Iron-based NPs blend # 2

DC4 Graphite and Calcium-based NPs blend # 4

DC6 Graphite and Calcium-based NPs blend # 6

Table 5.3. Summary of the Hydraulic Fracturing Experiments for the Shale Cores

(Contreras et al. 2014b)

Oil Based Mud Results

LCM

Blend

Breakdown

Pressure

(psi)

% Inc. in

Breakdown

Pressure

Re-opening

Pressure

(psi)

% Inc. in

Re-Opening

Pressure

% Inc. in

Re-Opening

Pressure

Above the

Breakdown

Pressure

CS 520 N/A 498 N/A N/A

G1 522 0.4 533 7 3

G2 520 0 538 8 3.5

DF1 524 0.8 626 26 20

DF3 524 0.8 552 11 6

DC4 520 0 659 32 27

DC6 503 -3.3 674 35 30

The % increase in the reopening pressure using an oil based mud was calculated

based on the breakdown (520 psi) and re-opening (448 psi) pressures for the control

sample. In addition, the % increase in the re-opening pressure when compared to the

breakdown pressure was calculated.

106

The results in Table 5.3 showed a different enhancement behavior compared to

the results shown in Table 4.9. The breakdown pressures were not enhanced when

nanoparticles were added compared to the control sample. However, an increase in the

re-opening pressures not only above the control sample re-opening pressure but also

above the breakdown pressure of the control sample. An increase of up to 30% above the

original breakdown pressure was observed when calcium-based NPs blend (DC6) was

used. The same was observed with iron-based NPs blend (DF1) where the re-opening

pressures were enhanced by 20% above the original breakdown pressure.

The pressures vs. time plots obtained from the hydraulic fracturing experiments

with the optimum nanoparticles concentrations for both; iron- based and calcium-based

NPs are shown in Figure 5.9 - Figure 5.11.

Figure 5.9. Pressure vs. Time for the Control Sample (CS) without Nanoparticles

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50

Pre

ssu

re (

psi)

Time (min)

1st Cycle (CS)

2nd Cycle (CS)

Breakdown @ 520 psi Re-Opening @ 498 psi

107

Figure 5.10. Pressure vs. Time for the Optimum Concentration of Iron-based

Nanoparticles Blend (DF1)

Figure 5.11. Pressure vs. Time for the Optimum Concentration of Calcium-based

Nanoparticles Blend (DC6)

The detailed analysis of the fractured shale cores under the SEM (Figure 5.12)

and the energy dispersive X-ray spectroscopy (EDX) (Figure 5.13) showed that the

fractures were completely sealed from wellbore to tip.

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50

Pre

ssu

re (

psi)

Time (min)

1st Cycle (DF1)

2nd Cycle (DF1) Re-Opening @ 626 psi

Breakdown @ 524 psi

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50

Pre

ssu

re (

psi)

Time (min)

1st Cycle (DC6)

2nd Cycle (DC6)

Breakdown @ 503 psi


108

Figure 5.12. (a) Cross-section of Fracture Plane near Wellbore, (b) Seal at Fracture

Mouth, and (c) Cross-section of Fracture Plane at the Fracture End (Contreras et al.

2014b)

Figure 5.13. EDX of Seal Cross-section at 500x magnification where the Green Color

Indicates Iron Particles Distribution on the Formed Seal (Contreras et al. 2014b)

This observation suggests that the nanoparticles penetrated through the resulting

fracture and formed the seal inside the fracture. However, the observed fractures in the

concrete cores were not sealed with LCMs. LCMs were observed within the formed filter

cake around the wellbore and present inside the fracture in the immediate vicinity close to

the fracture mouth. The calcium-based and iron-based nanoparticles have an average size

(a) (c) (b)

(b)

109

(D50) of 60 and 30 nm, respectively (Zakaria, 2013), which makes it easy for these

nanoparticles to penetrate inside the fractures while the D10 values of the conventional

LCM treatments used on the concrete cores ranged between 55 and 80 microns (Table

4.1).

The average fracture width measure after retrieving the shale cores form the

experiment was found to be approximately 20 microns (Figure 5.14), which can be

compared to the measured fracture width of the concrete core that was fractured with

clear fluid. With this small width and the size of nanoparticles, the nanoparticles

penetration inside the fracture plane is possible.

Figure 5.14. Fractures at Core End (left) and 3D Representation of the Core End

Containing the Fractures (right) (Contreras et al. 2014b)

5.3. PARTICLE SIZE DISTRIBUTION SELECTION CRITERIA

From the qualitative analysis of the effect particle size distribution (PSD) has on

sealing pressure (Chapter 4.3.4), the improved performance of treatments containing

(b) (c)

110

nutshells could be attributed to the fact that the D90 values of theses blends ranged

between 1900 – 2400 microns. In general, all LCM treatments having a D90 smaller than

the fracture width failed to establish a seal within the fracture. This suggests that the D90

value should be selected such that it matches the fracture width to ensure an effective

fracture sealing. In addition, the statistical analysis (Chapter 4.6.1) showed that the

sealing pressure is highly affected by the D90 and D75; hence proper selection of the D-

values is required to ensure effective fracture sealing. However, the current selection

methods were developed to enhance the bridging capabilities for drill-in fluid not to seal

fractures. Therefore, using these selection criteria might be somehow irrelevant for

fracture sealing applications.

In this section, the development of a method to select PSD of LCM treatments for

a known fracture width is presented. The method was developed based on a statistical

analysis of a subset dataset (Table F. 1) containing 31 tests to determine the relationship

between the effectiveness of different LCM treatments in terms of the sealing efficiency

and PSD.

The main reason for running the statistical analysis on a subset dataset is

eliminate the effect of other parameters such as the drilling fluid density and LCM

particles with highly irregular shapes. Since cellulosic fiber particles were described as

nonspherical particles with some elongated and flat particles, PSD analysis of blends

containing cellulosic fiber was found to be unrepresentative. For this reason it was

decided to exclude their results from the statistical analysis and limit it to granular

materials only. In addition, all the results for tests run with higher drilling fluid densities

111

were excluded since the density was found to have a strong effect on the sealing pressure

(Table 4.10).

5.3.1. Development of PSD Selection Method. In normal overbalanced drilling

operations (i.e. drilling fluid pressure higher than formation pressure), a minimum static

overbalance pressure of 150-300 psi is required to prevent formation fluid influx (Jahn et

al. 2008; Rehm et al. 2012; The Drilling Manual, 2015). Therefore, the formed seal

within the fracture should withstand at least the minimum overbalance pressure without

failing. In the statistical analysis, the sealing efficiency results of the different blends

were grouped into two categories; weak and strong seal. The seals that failed to withstand

an overbalance of 500 psi to (selected to account for the static and dynamic overbalance

pressure) are considered weak seal and the ones that survived 500 psi or more are

considered strong seal regardless of LCM type, concentration and fracture width.

A simple linear model was fitted using fracture width, fluid base, total LCM

concentration and the five D-values. The Leverage plot and P-test was performed on each

parameter individually to get an overall understanding of how each parameter affect the

sealing pressure (Table 5.4). The overall model fit resulted in a good correlation with R2

of 74% and a P-value less than 0.05 (Figure 5.15)

Table 5.4. Effect of Different Parameters on the Sealing Pressure and Model Fit for the

Subset Dataset

Parameter Unit P-Values

Fracture

Width (microns) 0.00001

D90 (microns) 0.08116

D50 (microns) 0.26377

D10 (microns) 0.44148

112

Table 5.4. Effect of Different Parameters on the Sealing Pressure and Model Fit for the

Subset Dataset (Cont’d)

D75 (microns) 0.53233

Total LCM (ppb) 0.53522

D25 (microns) 0.85566

Base Fluid (WBM/OBM) 0.90849

Model Summary Fit

R2 0.743

Figure 5.15. Leverage Plot Showing the Actual Sealing Pressure versus the Predicted

Sealing Pressure using the Simple Linear Fit Model for the Subset Dataset

In order to develop a generalized model that predicts the sealing pressure for

different fracture widths, statistical partition technique was used based on the obtained

information from the previous model. The statistical partition was performed using the

four significant parameters; D90, D75, D50 and D10 as a function of fracture width and

taking the total LCM as a weighted parameter. Sensitivity analysis was then performed

by first taking the three parameters (D90, D50 and D10) to predict sealing pressure and

113

second by taking only the first two parameters (D90 and D50). The model using D90 and

D50 to predict sealing pressure for a given fracture width gave a model fit similar (3%

difference) to the model with the four parameters. Thus, the model using D90 and D50

was chosen, where the model have a better fit than the previous model with an R2 of

74%. The statistical partitioning was then converted into a mathematical relationship that

predicts a good sealing pressure, which was defined earlier as any pressure >500 psi, as a

function of the selected PSD D-values. The obtained mathematical relationship suggests

the following:

D50 should be ≥ 3

10 the fracture width

D90 should be ≥ 6

5 the fracture width

In order to validate the proposed selection criteria and compare it with the

previous selection criteria discussed in subchapter 2.5 (Table 2.3), a comparison table

was constructed (Table F.2). The table lists the test number, the fracture width used, the

measured seal integrity, and the predicted seal integrity.

The measured seal integrity is based on the measured sealing pressure, i.e. if the

measured sealing pressure is larger than 500 psi, then the measured seal integrity is

considered strong. If the measured sealing pressure is less than 500 psi, then the

measured seal integrity is considered weak. In the comparison, it is assumed that

following the selection criteria should result in a strong seal that could withstand

pressures higher than 500 psi. The predicted seal integrity column is based on comparing

the suggested D-Value by the selection criteria with the actual measured PSD. If the

measured PSD is equal or larger than the suggested value by the selection criteria, the

114

predicted seal integrity is considered strong. If the measured PSD is less than the

suggested PSD, the predicted seal integrity is considered weak. The last step is to check

if the measured seal integrity matches with the predicted one and calculate the overall

match percentage. The overall match percentage was calculated for the selected 31 tests

using the previous selection criteria as well as the proposed criteria as shown in Table

5.5.

Table 5.5. Comparison of Percentage Match for the Different PSD Selection Criteria

Proposed

Method

Abrams

Rule (1977)

D90 Rule

(Smith et al. 1996,

Hands et al. 1998)

Halliburton

Method

(Whitfill, 2008)

Vickers’s

Method

(2006)

90% 68 % 77 % 55 % 45%

The verification showed a good match between the predicted seal integrity and the

measured seal integrity pressure with an overall match of 90% for the proposed selection

criteria. Vickers method showed the lowest percentage match due to the number of D-

values that needs to be satisfied. Halliburton’s method (Whitfill, 2008) also showed a low

predictive match. Abrams and D90 rules had higher percentage match compared to

Vickers and Halliburton’s methods. Since the validation was performed on the same

dataset, additional testing would be beneficial to independently validate the proposed

criteria. However, the result is in agreement with the observation in subchapter 4.3.4

were a wide particle size distribution gave higher sealing pressure since smaller particles

filled the voids between the larger particles in the formed plugs.

115

6. CONCLUSIONS AND RECOMMENDATIONS

6.1. CONCLUSIONS

In this dissertation, an experimental study of the factors affecting the effectiveness

of LCM treatments in sealing fracture was carried out with a characterization of the used

LCMs. Two apparatuses were developed to evaluate the capability of LCM treatments in

sealing fractures, as well as measuring the maximum pressure that can be achieved before

the seal/plug breaks. The investigated factors include; fracture width, LCM type, LCM

concentration, temperature, PSD, base fluid, fluid density, weighting material, and LCM

shape. Experimental results and statistical methods were used in this study to provide a

better understanding of the reasons behind the variation in LCM performance. Hydraulic

fracturing tests were conducted to investigate the effect of LCM addition to drilling fluid

for wellbores strengthening application and the results were used to validate how realistic

available analytical models in estimating fracture widths.

An updated classification of LCM into eight categories, based on their appearance

and field application, was presented to include the recent technologies. In addition, a

comprehensive list of available LCM, which includes the generic name and description of

each LCM as well as the trade name and it field application, was constructed to serve as a

cross-reference table for operators, service companies and drilling engineers.

Based on this work the following conclusions can be drawn;

Widening the fractures (increasing slot size) reduced sealing pressures.

Lower fluid loss values resulted in higher sealing pressures for fractures up to

1500 microns. For wider fractures, de-fluidization was required to initiate a strong

plug for fractures.

116

Blends containing nutshells showed the best performance in terms of sealing

fractures at both low (15 - 20 ppb) and high concentrations (40 – 50 ppb).

Graphite and sized calcium carbonate blends were only effective at a very high

concentration (105 ppb). This can be attributed to the effect of LCM particle

shape, in terms of sphericity and roundness, which showed a significant effect on

the overall sealing pressure where higher angularity resulted in higher sealing

pressure and higher sphericity resulted in lower sealing pressure.

The sealing pressure increased with higher LCM concentrations; however, there is

a critical maximum concentration at which LCM will not penetrate the fracture

and no tight seal will be formed.

The variation in the sealing pressure at high temperatures suggests that LCM

treatments should be evaluated at the expected formation temperature.

The increase in the sealing pressure for nutshells by 54% when used in single

LCM treatment at high temperature is attributed to the swelling of nutshell

particles, which was quantified experimentally.

An updated selection criterion for PSD selection for a known fracture width was

developed stating that D50 and D90 should be equal or greater than 3/10 and 6/5

the fracture width, respectively.

The proposed criterion was verified with the lab results and it gave a 90% match

between the actual and predicted performance.

There is no clear trend on how different base fluids affect the sealing pressure.

Increasing fluid density increased the sealing pressure for both WBM and OBM.

The finer the PSD of the weighting materials, the higher the sealing pressures.

117

Conventional LCMs used in this study were able to seal fractures up to 2500

microns. For fractures wider than 2500 microns, the unconventional LCM

treatments were able to seal fractures up to 5000 microns using WBM. However,

the unconventional LCM failed to initiate a strong seal at 1500 microns with

OBM.

The lowest breakdown and re-opening pressure for the fractured concrete samples

was observed with the solid free clear base fluid (EDC95-11).

The addition of different LCM blends enhanced the breakdown and re-opening

pressure up to 18% and 210%, respectively, compared to the control sample that

contained no LCM. Only tests with nanoparticles were able to increase reopening

pressure above breakdown pressure.

The fractured core using base fluid only had the smallest fracture widths

compared to the other cores, which were fractured with fluid containing solids.

Of the five analytical fracture models evaluated, only one model (Carbonell and

Detournay, 1995) estimated comparable fracture widths to the measured widths

on the fractured cores. For the other models, two models underestimated fracture

width, one model overestimated fracture width, and the last model had a constant

fracture width.

As a final note, this experimental study shows that carefully selected LCM

treatments can successfully seal drilling induced and natural fractures up to 2500 microns

for any base fluid. This result is higher than any previously reported studies. For larger

fractures there is still need for unconventional LCM materials or conventional methods

such as setting cement plugs. It is also seen that introducing LCM particles in the drilling

118

fluid is causing the fracture breakdown pressure and the fracture reopening pressure (i.e.

propagation pressure) to be increased. This shows that the experienced safe mudweight

window can be manipulated with including LCM materials in the drilling fluid.

6.2. RECOMMENDATION AND FUTURE WORK

Even though a large data set was developed from the high pressure LCM

evaluation, there are some experimental limitations with the current apparatus. The fluid

loss values were measure at the end of each test and the average fluid loss per cycle was

calculated based on the number of cycles for each test. The high pressure tests were

conducted at low injection rate, which does not simulates the flow rates seen in a

wellbore. Fluid loss values should be recorded with respect to time in order to provide an

accurate measurement of fluid loss through fractures. The testing apparatus could be

improved by installing a flow loop that circulates drilling fluids across the fractures.

Different factors that might have an effect on the sealing pressure were not

addressed in this study. Only one fracture length (i.e. slot thickness) was investigated for

conventional LCM evaluation. The designed tapered slots do not simulate the actual

formation roughness, which might have an effect on the overall sealing effectiveness. The

slot fracture angle also might have an effect on the packing of LCM inside the fracture. A

parametric study of how the variation in fracture length, roughness, and angle is

suggested to investigate how slot design will affect the results.

Only one unconventional LCM treatment (WEDGE-SET) was used in this study

due to the difficulty in getting other unconventional treatments from services companies.

Further investigation on unconventional treatments is encouraged.

119

The effect of increasing the temperature was investigated at one temperature;

however further investigation using a wider range of temperatures is required to establish

an understanding of how LCM is affected.

Five hydraulic fracturing experiments were carried out in this study using OBM

on concrete core samples, where 3 tests were performed using LCMs. Future work should

include further investigation using permeable and impermeable rocks using WBM and

different LCM treatments to identify the relationship between the maximum sealing

pressure and the enhancement in the breakdown and re-opening pressure.

The fracture widths of the fractured cores were measured after removing the

applied overburden and confining stresses. Therefore, the measured width represent the

relaxed fracture width rather than the actual stressed width. For a better measurement,

future work should consider in-situ fracture width estimation.

120

APPENDIX A

A. PREVIOUSLY REPORTED LABORATORY RESULTS

121

Table A.1. Summary of Results from The Impermeable Test using Different LCMs and

Fracture Widths (Tehrani et al. 2007)

Test

# LCM Used

Concn

(ppb)

Base

Fluid

Fracture

Width

(microns)

Fluid

Loss

(ml)

Sealing

Pressure

(psi)

1

SCC (C/M) 20

WBM 530 12.8 73 SCC (F) 10

CF (F) 20

2

SCC (C/M) 20

WBM 848 3.3 998 SCC (F) 10

CF (F) 20

NS (F) 10

3 G 40

OBM 555 4.4 425 CF (F) 10

4

G 40

OBM 637 3.2 876 CF (F) 10

NS (F) 10

5

G 3.5

OBM 742 4.7 900 SCC (C/M) 11.5

SCC (F) 14.7

SCC (C) 10.2

6

G 20.5

OBM 650 3.5 685 SCC (M/F) 16

SCC (F) 3.5

7

G 7

OBM 555 14.9 491 SCC (M/F) 17.5

SCC (F) 17.5

Note. Coarse (C) Medium (M) Fine (F)

122

Table A.2. Summary of the Results from the Permeable Test using Different LCMs and

Fracture Widths (Hettema et al. 2007)

Test # LCM

Used

Concn.

(ppb)

Base

Fluid

Fracture

Width

(microns)

Fluid

Loss

(ml)

Sealing

Pressure

(psi)

1 None 0 OBM 250 66.5 679

2 None 0 SBM 250 78.1 1746

3 None 0 SBM 250 53.3 2733

4 SCC 27

SBM 500 45.5 988 G (A) 13

5 SCC 27

SBM 500 6.7 1547 G (A) 13

6 G (B) 20 SBM 500 13.9 1962

7 G (B) 20 WBM 500 35.14 2041

8 G (B) 20 OBM 500 8.8 3836

9 NS 10

SBM 500 19.5 6132 G (B) 10

10 NS 20

SBM 1000 11.7 2037 G (B) 20

11 NS 20

SBM 1000 2.5 3611 G (B) 20

12 NS 20

SBM 1000 1.8 4420 G (B) 20

123

Table A.3. Summary of Results from the Permeable Test using Different LCMs and

Fracture Widths (Kaageson-Loe et al. 2009)

Test # LCM

Used

Concn.

(ppb)

Base

Fluid

Fracture

Width

(microns)

Fluid

Loss

(ml)

Sealing

Pressure

(psi)

1 G (R)

* 20

SBM 250 N/A 1490**

NS (R)

* 20

2 G (R)

* 20

SBM 250 N/A 1710**

NS (R)

* 20

3 G (R)

* 10

SBM 250 N/A 1970 NS (R)

* 10

4 G (R)

* 20

SBM 500 N/A 1190 NS (R)

* 20

5 G (R)

* 20

SBM 500 N/A 1520**

NS (R)

* 20

6 G 20

SBM 500 N/A 3900 NS 20

7 G 20

SBM 500 N/A 6131 NS 20

8 G 20

SBM 500 N/A 6132 NS 20

9 G 20

SBM 1000 N/A 1560**

NS 20

10 G 20

SBM 1000 N/A 2037 NS 20

11 G 20

SBM 1000 N/A 2803 NS 20

12 G 20

SBM 1000 N/A 3611 NS 20

* Resized LCMs ** Test conducted at closed fracture tip

124

APPENDIX B

B. HYDRAULIC FRACTURING TEST PREPARATION AND PROCEDURES

125

CORE PREPARATION

1- Mix cement in a large container to ensure the consistency of the cores (Figure B.

1.1)

2- A 94 lbs sack of Portland cement Class-H was mixed with 38% of water (35 lbs

of water) following the recommended mixing procedures.

3- Apply grease or any kind of lubricants inside the molds to easily extract the

samples after the dry.

4- Pour the cement mixture into 5 7/8 inch (diameter) x 9 inch (height) molds

(Figure B. 1.2) while hitting on the molds outside to get rid of the air bubbles.

5- Let the concrete to cure for at least 7 days.

6- Drill ½ inch wellbore in the cement cores using the drill press (Figure B. 1.3)

7- Screw casings into the top and bottom cap.

8- Attach top and bottom caps to the cement cores using epoxy (Figure B. 1.4) and

let the epoxy dry for 24 hrs. Note that each cap needs to be attached on a different

day.

9- Grind the core sample or use sanding papers to remove excess dried epoxy and

ensure a smooth surface to avoid damaging the confining rubber sleeve.

10- Screw the injection nipple into the top cap (Figure B. 1. 5)

126

Figure B. 1. Concrete Core Sample Preparation Steps

(1) Mix Cement (2) Pour cement in molds

(4) Attach top and bottom cap (3) Drill wellbore

(5) Screw Injection Nibble

127

TEST PROCEDURES

1- Raise the pressure cell and remove the clevis pins after removing the safety cutter

pins.

2- Lower the pressure cell (Figure B. 2).

3- Place an O-ring inside the pressure cell at the bottom.

4- Place the prepared core in the fracturing cell carefully.

5- Place the three top spacers with O-rings between them on the core sample.

6- Raise the pressure cell and place the clevis pins back again with the safety cutter

pins.

7- Once the clevis pins are in place, drop the cell on the clevis until the hoist cables

are no longer in tension.

8- Close the cell exit valve.

9- Fill up wellbore with drilling fluid containing LCM

10- Connect the injection line into the injection nipple.

11- Connect the confining line to the cell and close the confining exit line.

12- Apply 400 psi overburden on the core, which reads 7000 psi in the upper gauge,

using the hydraulic hand pump.

13- Apply 100 psi confining pressure using the Isco Pump.

14- Fill up accumulator with drilling fluid (without LCM) using the upper plastic

accumulator.

15- Open the Isco pump software and assign a name to file and take the injection

pump into the remote control.

128

16- Change the flow rate to 5ml/min and start recording the data (Logging ON) and

hit the start button to start injection.

17- Continue injection until a dramatic decrease in the injection pressure is observed,

which indicates the sample is fractured. Note that the confining pump will show

an increase in the pressure above too.

18- Stop the injection and open the mud exit valve to relief wellbore pressure

19- Close the mud exit valve and start timing (10 minute for fracture healing) before

the second cycle.

20- Refill injection pump if needed.

21- Start the second injection cycle until another drop in the injection pressure is

observed.

22- Stop the pump and stop the data logging, at this point the test is complete.

23- Open the mud exit valve again to release the wellbore pressure.

24- Remove the overburden pressure and open the confining exit valve to remove

confining pressure.

25- Unscrew both injection and confining lines from the pressure cell.

26- Raise the pressure cell and remove the clevis pins after removing the safety cutter

pins.

27- Lower the pressure cell.

28- Remove the three top spacers.

29- Pull the core sample out carefully.

30- Clean the system, including the cell inside.

129

31- Clean both plastic and metal accumulator before the next test if different fluid will

be used.

Figure B. 2. Hydraulic Fracturing Apparatus; (a) Metal Accumulator, (b) Plastic

Accumulator, (c) Hydraulic Hand Pump, (d) Overburden Gauge, (e) Pressure Cell, (f)

Confining Pump, (g) Injection Pump, (h) Clevis Pin

a

b

c

d

e

f

g

h

130

APPENDIX C

C. CROSS-REFERENCING TABLES FOR AVAILABLE LCMS UNDER THE NEW

CLASSIFICATION

131

Table C. 1. Examples of Granular LCMs

Granular LCM's

Generic Name /

Description Trade Name Provider Application

Ground/Sized walnut

shells

WALL-NUT Halliburton Used as concentrated

pills or high filtration

squeeze. It is used to

cure seepage, partial and

total loss based on the

selected grade.

MIL-PLUG Baker Hughes

NewPlug NEWPARK

WALNUT HULLS GEO Drilling Fluids

NUTSHELL Anchor Drilling Fluids

MESUCO-PLUG Messina Chemicals

Resilient, angular,

dual-carbon based,

sized graphite

STEELSEAL Halliburton Can be used as a

background treatment or

as a concentrated pill

depending on the losses

rate.

Used as wellbore

strengthening material.

G-SEAL MI SWACO

C-SEAL MI SWACO

LC-LUBE Baker Hughes

NewSeal NEWPARK

A proprietary natural

loss prevention

material (LPM)

SURE-SEAL Drilling Specialties Can be used as a

preventive treatment or

as a concentrated pill.

TORQUE-SEAL Drilling Specialties

A blend of acid

soluble particulates EZ-PLUG Halliburton

Can be used as a

background treatment,

for seepage losses or

severe losses

Sized-ground marble

BARACARB Halliburton Used as a bridging agent

for lost circulation

problems.

Used as wellbore

strengthening material.

SAFE-CARB MI SWACO

NewCarb NEWPARK

FLOW-CARB Baker Hughes

MIL-CARB Baker Hughes

W. O. 30 Baker Hughes Used for severe losses.

132

Table C. 2. Examples of Flaky LCMs

Flaky

Generic Name /


Cellophane

MILFLAKE Baker Hughes Used in conjunction with

other LCM's depending

on the severity of the

losses.

MESUCO-

FLAKE Messina Chemicals

Sized grade of Mica

MILMICA Baker Hughes Used as preventive

measures for seepage

losses. MESUCO-MICA Messina Chemicals

Flaked Calcium

Carbonate SOLUFLAKE Baker Hughes

Used for seepage or

severe losses based on

the selected grade.

133

Table C. 3. Examples of Fibrous LCMs

Fibrous

Generic Name /


Natural cellulose fiber

BAROFIBRE Halliburton

Can be used as a


concentrated pills to cure

seepage to severe losses

based on the selected

grade.

M-I-X II MI SWACO

VINSEAL MI SWACO

CHEK-LOSS Baker Hughes

MESUCO-

FIBER Messina Chemicals

CyberSeal NEWPARK

FIBER SEAL GEO Drilling Fluids

A proprietary micro-

cellulosic fiber for use

in water base muds

DYNARED Drilling Specialties Used as normal

treatment to cure

seepage losses or as a

concentrated pill for loss

circulation.

A proprietary micro-

cellulosic fiber for use

in oil base muds

DYNA-SEAL Drilling Specialties

Shredded cedar fibers.

M-I CEDAR

FIBER MI SWACO Can be used as a

preventive treatment,

concentrated pill or high

fluid-loss squeezes.

FIBER PLUG Anchor Drilling Fluids

PLUG-GIT Halliburton

MIL-CEDAR Baker Hughes

Acid soluble extrusion

spun mineral fiber

N-SEAL Halliburton Can be used as a



CAVI-SEAL-AS Messina Chemicals

MAGMA FIBER GEO Drilling Fluids/

Anchor Drilling Fluids

134

Table C. 4. Examples of LCMs Combinations

LCM's Combinations

Generic Name /


A combination of

different LCM types

and wide range of

particle sizes.

STOPPIT Halliburton Used as a concentrated

pill. PRIMA SEAL GEO Drilling Fluids

STOP-FRAC S Halliburton

WELL-SEAL Drilling Specialties

Used to cure minor to


the selected grade.

BARO-SEAL Halliburton Can be used as a


concentrated pills to cure

seepage to severe losses

based on the selected

grade.

STOP-FRAC D Halliburton

M-I SEAL MI SWACO

MIL-SEAL Baker Hughes

CHEM SEAL Anchor Drilling Fluids

KWIK-SEAL Messina Chemicals

MESUCO-SEAL Messina Chemicals

A blend of acid

soluble particulates. EZ-PLUG Halliburton

Can be used as a



severe losses.

A proprietary

particulate blend

designed to be used

with foam wedges.

QUIK-WEDGE Sharp-Rock

Technologies, Inc.

Can be used as a pre-

treatment or as a

concentrated pill.

A proprietary

particulate blends

that includes

modified natural

materials and other

additives.

STRESS-SHIELD Sharp-Rock

Technologies, Inc.

Can be used as a pre-

treatment or as a

concentrated pill

135

Table C. 5. Examples of Acid Soluble LCMs

Acid Soluble/ Biodegradable/Water Soluble

Generic Name /


A blend of acid soluble

particulates EZ-PLUG Halliburton

Can be used as a



severe losses

A non-damaging, cross

linkable water soluble

polymer blended with

selected sized

cellulosic fibers.

N-SQUEEZE Halliburton Used as a settable

squeeze for severe losses

Acid soluble extrusion

spun mineral fiber

N-SEAL Halliburton Can be used as a



CAVI-SEAL-AS Messina Chemicals

MAGMA FIBER GEO Drilling Fluids/

Anchor Drilling Fluids

Sized and treated salts BARAPLUG Halliburton

Use as a temporary seal

in high permeability

formations

Acid Soluble Sized-

Calcium Carbonate

BARACARB Halliburton

Used as a bridging agent

for lost circulation

problems

SAFE-CARB MI SWACO

NewCarb NEWPARK

FLOW-CARB Baker Hughes

MIL-CARB Baker Hughes

W. O. 30 Baker Hughes Used for severe losses.

Flaked calcium

carbonate SOLUFLAKE Baker Hughes

Used for seepage or


the selected grade.

Nontoxic fibrous

powdered

polysaccharide,

biodegradable and acid

soluble lost circulation

material.

HOLE-SEAL-II Messina Chemicals

Can be used as a

pretreatment or as a

concentrated pill.

136

Table C. 6. Examples of High Fluid Loss LCMs Squeezes

High Fluid Loss LCM's Squeeze

Generic Name /


High fluid loss squeeze GEO STOP

LOSS GEO Drilling Fluids

Used as a high-fluid loss

squeeze.

High-solids, high-

fluid-loss reactive lost

circulation squeeze

DIASEL M Drilling Specialties

A specially formulated

high-solids high fluid

loss squeeze.

DIAPLUG Messina Chemicals

A proprietary blend of

granular and fibrous

materials.

X-Prima NEWPARK

A blend of granular

and fibrous materials. NewBridge NEWPARK

Micro-sized cellulosic

fiber combined with a

blend of organic

polymers

ULTRA SEAL GEO Drilling Fluids Used as concentrated

pills.

A blend of fine

particles to promote

high fluid loss and

other additives in

addition to highly

compressible and

permeable foam rubber

chunks.

WEDGE-SET Sharp-Rock

Technologies, Inc.

Used as a high-fluid loss

squeeze.

A combination of both

resilient graphitic

carbon and malleable

components

DUO-SQUEEZE Halliburton

Can be used as a high

fluid loss squeeze or as

concentrated pill.

137

Table C. 7. Examples of Settable/Hydratable LCM’s

Table C. 8. Examples of Nanoparticles LCM’s

Nanoparticles

Generic Name /


Iron Hydroxide NP Iron Hydroxide NP nFluids Inc. Used as a background

treatment to seal micro

fractures and wellbore

strengthening

applications.

Calcium Carbonate NP Calcium Carbonate NP nFluids Inc.

Settable/Hydratable LCM's Combinations

Generic Name /


A combination of

swelling polymer along

with engineered

combinations of

resilient graphitic

carbon and other

materials

HYDRO-PLUG Halliburton

Used as a hydratable pill

to plug vugular,

fractured, and cavernous

formations.

Dry powdered/granular

material with synthetic

polymers, inorganic

minerals, chemical

reagents and stabilized

organic filler.

SUPER-STOP Messina Chemicals Used as a swellable pill

for sever losses.

A non-damaging, cross

linkable water soluble

polymer blended with

selected sized

cellulosic fibers.

N-SQUEEZE Halliburton Used as a settable

squeeze for sever losses.

138

APPENDIX D

D. PARTICLE SIZE DISTRIBUTION ANALYSIS

139

Figure D. 1. Particle Size Distribution for 15 ppb G # 1 Blend

Figure D. 2. Particle Size Distribution for 50 ppb G # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 1300 D75 = 800 D50 = 320 D25 = 85 D10 = 60

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 1300 D75 = 800 D50 = 340 D25 = 95 D10 = 60

140

Figure D. 3. Particle Size Distribution for 15 ppb NS # 1 Blend

Figure D. 4. Particle Size Distribution for 50 ppb NS # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 2000 D75 = 1600 D50 = 1000 D25 = 400 D10 = 180

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 2400 D75 = 1600 D50 = 1000 D25 = 400 D10 = 180

141

Figure D. 5. Particle Size Distribution for 50 ppb SCC # 3 Blend

Figure D. 6. Particle Size Distribution for 15 ppb CF # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 1200 D75 = 950 D50 = 680 D25 = 360 D10 = 250

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 1400 D75 = 700 D50 = 220 D25 = 140 D10 = 90

142

Figure D. 7. Particle Size Distribution for 50 ppb CF # 1 Blend

Figure D. 8. Particle Size Distribution for 30 ppb G & SCC # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 1400 D75 = 800 D50 = 220 D25 = 150 D10 = 90

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 1300 D75 = 900 D50 = 460 D25 = 100 D10 = 80

143



1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 1300 D75 = 900 D50 = 480 D25 = 120 D10 = 80

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 90

D 75

D 50

D 25

D 10

D90 = 1400 D75 = 1100 D50 = 420 D25 = 90 D10 = 65

144



1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 900 D75 = 700 D50 = 500 D25 = 150 D10 = 60

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 90

D 75

D 50

D 25

D 10

D90 = 1400 D75 = 1200 D50 = 700 D25 = 400 D10 = 90

145



1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 1400 D75 = 1250 D50 = 900 D25 = 500 D10 = 100

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 90

D 75

D 50

D 25

D 10

D90 = 2600 D75 = 1900 D50 = 1300 D25 = 650 D10 = 170

146



1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 2400 D75 = 1800 D50 = 1000 D25 = 250 D10 = 100

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 90

D 75

D 50

D 25

D 10

D90 = 2200 D75 = 1800 D50 = 1400 D25 = 800 D10 = 300

147

Figure D. 17. Particle Size Distribution for 20 ppb G & NS # 1 Blend

Figure D. 18. Particle Size Distribution for 40 ppb G & NS # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 10

D 25

D 50

D 75

D 90

D90 = 1900 D75 = 1300 D50 = 500 D25 = 180 D10 = 65

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r

D 90

D 75

D 50

D 25

D 10

D90 = 2000 D75 = 1400 D50 = 580 D25 = 180 D10 = 80

148

Figure D. 19. Particle Size Distribution for 55 ppb G, SCC, & CF # 1 Blend

1010.1

100

90

80

70

60

50

40

30

20

10

Seive Size (mm)

% F

ine

r D 90

D 75

D 50

D 25

D 10

D90 = 1200 D75 = 850 D50 = 450 D25 = 100 D10 = 55

149

APPENDIX E

E. PREDICTIVE LINEAR MODEL

150

Equation E.1 gives the linear predictive model presented in subchapter 4.6.1.

𝑆𝑒𝑎𝑙𝑖𝑛𝑔 𝑃𝑟𝑒𝑠𝑠𝑢𝑟𝑒 (𝑝𝑠𝑖) = −12006.89 + 𝐹𝑙𝑢𝑖𝑑𝐶𝑜𝑒𝑓𝑓. + (122.4 × 𝜌𝐹𝑙𝑢𝑖𝑑 ) +

𝐿𝐶𝑀𝐶𝑜𝑒𝑓𝑓. + (−0.9 × 𝑊𝑐 ) + (−1.85 𝐷10) + (8.93 𝐷25) + (−7.28 𝐷50) +

(9.69 𝐷75) + (2.45 𝐷90)

Where:

Fluid Coefficient

OBM -87.5

WBM 87.5

LCM Blends Coefficient

CF # 1 2881.496

G # 1 2995.117

G, SCC, & CF # 1 4251.583

NS # 1 -3153.86

SCC # 3 2418.024

LCM Blends Coefficient

G & SCC # 1 3207.762

G & SCC # 3 1298.886

G & SCC # 4 5353.874

G & SCC # 5 364.3878

G & SCC # 6 -1011.73

G & SCC # 7 -6158.57

G & SCC # 8 -4532.94

G & SCC # 9 -5785.11

151

APPENDIX F

F. PSD MODELS COMPARISON

152

Table F. 1. Selected Subset Data for PSD Selection Criteria Development

Test No.

Sealing

Pressure

(psi)

Fracture

Width

D10

(microns)

D25

(microns)

D50

(microns)

D75

(microns)

D90

(microns)

Measured

Seal

Integrity

1 414 1000 60 85 320 800 1300 Poor

2 449 1000 60 95 340 800 1300 Poor

3 487 1000 78 100 460 900 1300 Poor

4 584 1000 80 120 480 900 1300 Good

5 224 1500 80 120 480 900 1300 Poor

6 1569 1000 65 90 420 1100 1400 Good

7 117 1000 60 150 500 700 900 Poor

8 1489 1000 90 400 700 1200 1400 Good

9 205 1500 65 90 420 1100 1400 Poor

10 0 1500 60 150 500 700 900 Poor

11 164 1500 100 500 900 1250 1400 Poor

12 1708 1500 170 650 1300 1900 2600 Good

13 1669 2000 170 650 1300 1900 2600 Good

14 110 3000 100 250 1000 1800 2400 Poor

15 0 3000 300 800 1400 1800 2200 Poor

16 589 1000 250 360 680 950 1200 Good

17 162 1500 250 360 680 950 1200 Poor

18 984 1000 180 400 1000 1600 2000 Good

19 1754 1500 180 400 1000 1600 2000 Good

20 347 2000 180 400 1000 1600 2000 Poor

21 441 2000 180 400 1000 1600 2000 Poor

22 2202 1000 180 400 1000 1600 2400 Good

23 2027 1500 180 400 1000 1600 2400 Good

24 2237 2000 180 400 1000 1600 2400 Good

25 755 2000 180 400 1000 1600 2400 Good

26 2372 1000 65 180 500 1300 1900 Good

27 805 1500 65 180 500 1300 1900 Good

28 2892 1000 80 180 580 1400 2000 Good

29 1713 1500 80 180 580 1400 2000 Good

30 295 2000 80 180 580 1400 2000 Poor

31 242 2000 80 180 580 1400 2000 Poor

153

Table F. 2. PSD Model Effectiveness Comparison

Proposed

Criterion Abrams

(1977) D90 Rule

(1996, 1998) Vickers

(2006) Halliburton

(2008)

Test

#

FW

(microns) Measured P M P M P M P M P M

1 1000 W S N W Y S N W Y W Y

2 1000 W S N S N S N W Y W Y

3 1000 W S N S N S N W Y W Y

4 1000 S S Y S Y S Y W N W N

5 1500 W W Y W Y W Y W Y W Y

6 1000 S S Y S Y S Y W N W N

7 1000 W W Y S N W Y W Y W Y

8 1000 S S Y S Y S Y W N W N

9 1500 W W Y W Y W Y W Y W Y

10 1500 W W Y S N W Y W Y W Y

11 1500 W W Y S N W Y W Y W Y

12 1500 S S Y S Y S Y W N W N

13 2000 S S Y S Y S Y W N W N

14 3000 W W Y S N W Y W Y W Y

15 3000 W W Y S N W Y W Y W Y

16 1000 S S Y S Y S Y W N W N

17 1500 W W Y S N W Y S N W Y

18 1000 S S Y S Y S Y W N S Y

19 1500 S S Y S Y S Y W N W N

20 2000 W W Y S N S N W Y W Y

21 2000 W W Y S N S N W Y W Y

22 1000 S S Y S Y S Y W N S Y

23 1500 S S Y S Y S Y W N W N

24 2000 S S Y S Y S Y W N W N

25 2000 S S Y S Y S Y W N W N

26 1000 S S Y S Y S Y W N W N

27 1500 S S Y S Y S Y W N W N

28 1000 S S Y S Y S Y W N W N

29 1500 S S Y S Y S Y W N W N

30 2000 W W Y W Y S N W Y W Y

31 2000 W W Y W Y S N W Y W Y

Overall Match % 90% 68% 77% 45% 55%

Note: Fracture width (FW) Predicted (P) Match (M) Strong (S) Weak (W) Yes (Y) No (N)

154

APPENDIX G

G. PUBLICATION FROM THIS DISSERTATION

155

PUBLISHED PAPERS

Alsaba, M., Nygaard, R., Hareland, G., and Contreras, O. 2014a. “Review of Lost

Circulation Materials and Treatments with an Updated Classification”. AADE-14-

FTCE-25, AADE Fluids Technical Conference and Exhibition, Houston, USA, 15-16

April.

Alsaba, M., Nygaard, R., Saasen, A., and Nes, O. 2014b. “Lost Circulation Materials

Capability of Sealing Wide Fractures”. SPE 170285, SPE Deepwater Drilling and

Completions Conference, Galveston, TX, USA, 10–11 September.

Alsaba, M., Nygaard, R., Saasen, A., and Nes, O. 2014c. “Laboratory Evaluation of

Sealing Wide Fractures Using Conventional Lost Circulation Materials”. SPE

170576, SPE Annual Technical Conference and Exhibition, Amsterdam, the

Netherlands, 27–29 October.

Contreras, O., Hareland, G., Husein, M., Nygaard, R., and Alsaba, M. 2014b.

“Experimental Investigation on Wellbore Strengthening In Shales By Means of

Nanoparticle-Based Drilling Fluids”. SPE 170589, SPE Annual Technical

Conference and Exhibition, Amsterdam, the Netherlands, 27–29 October.

PAPERS IN REVIEW

Alsaba, M., Nygaard, R., Saasen, A., and Nes, O. “Experimental Investigation of Fracture

Width Limitations of Granular Lost Circulation Treatments”. Journal of Petroleum

Exploration and Production Technology. (Submitted)

Contreras, O., Hareland, G., Husein, M., Nygaard, R., and Alsaba, M. “Experimental

Investigation on Wellbore Strengthening In Shales By Means of Nanoparticle-Based

Drilling Fluids”. SPE Journal. (Submitted)

156

BIBLIOGRAPHY

Aadnoy, B. S. and Belayneh, M. 2004. Elasto-Plastic Fracturing Model for Wellbore

Stability Using Non-Penetrating Fluids. Journal of Petroleum Science and

Engineering, Vol. 45 (3), pp. 179-192.

Aadnoy, B. S., and Chenevert, M. E. 1987. Stability of Highly Inclined Boreholes. SPE

Drilling Engineering, Vol. 2 (4), pp. 364 374.

Abrams, A. 1977. Mud Design to Minimize Rock Impairment Due to Particle Invasion.

Journal of Petroleum Technology, Vol. 29 (5), pp.586-592.

Alberty, M. W. and McLean, M. R. 2004. A Physical Model for Stress Cages. SPE-

90493-MS, SPE Annual Technical Conference and Exhibition, Houston, USA, 26-29

September.

Alleman, J., Owen, B., Cornay, F., and Weintritt, D. J. 1998. Multi-functional Solid

Lubricant Reduces Friction/Prevents Mud Loss. World Oil, Vol. 219 (9), pp. 87-90.

Anchor Drilling Fluids USA, Inc. Products Data Sheets,

http://intranet.anchorusa.com/msds-sheets. Accessed 24 January, 2014.

Anderson, T. L. 1995. Fracture Mechanics: Fundamentals and Applications. Second

edition, CRC Publications, United States.

API Recommended Practice 13B, 1974. Standard Procedure for Testing Drilling Fluid.

American Petroleum Institute, Washington, D.C., Section 10.

Aston, M. S., Alberty, M. W., McLean, M. R., Jong, H., and Armagost, K. 2004. Drilling

Fluids for Wellbore Strengthening. SPE-87130-MS, IADC/SPE Drilling Conference,

Dallas, Texas, 2-4 March.

Baker Hughes. 2010. Specialty Products: Drilling Fluids Solution,

http://www.bakerhughes.com/news-and-media/resources/brochures/specialty-

products-brochure. Accessed 24 January, 2014.

Ballard, A., and Massam, J. 2008. Precipitated Weighting Agents for use in Wellbore

Fluids. WO2008033838A2, 20 March.

Bruton, J. R., Ivan, C. D., and Heinz, T. J. 2001. Lost Circulation Control: Evolving

Techniques and Strategies to Reduce Downhole Mud Losses. SPE-67735-MS,

SPE/IADC Drilling Conference, Amsterdam, the Netherlands, 27 February -1 March.

Canson, B. E. 1985. Lost Circulation Treatment for Naturally Fractured, Vugular, or

Cavernous Formations. SPE/IADC 13440, SPE/IADC Drilling Conference, New

Orleans, USA, 5-8 March.

http://intranet.anchorusa.com/msds-sheets

http://www.bakerhughes.com/news-and-media/resources/brochures/specialty-products-brochure

http://www.bakerhughes.com/news-and-media/resources/brochures/specialty-products-brochure

157

Carbonell, R., and Detournay, E. 1995. Modeling Fracture Initiation and Propagation

from a Pressurized Hole: A Dislocation-Based Approach. ARMA-95-0465, the 35th

U.S. Symposium on Rock Mechanics (USRMS), Reno, Nevada, 5-7 June.

Caughron, D., Renfrow, D., Bruton, J., Ivan, C., Broussard, P., Bratton, T., and

Standifird, W. 2002. Unique Crosslinking Pill in Tandem with Fracture Prediction

Model Cures Circulation Losses in Deepwater Gulf of Mexico. SPE-74518-MS,

IADC/SPE Drilling Conference, Dallas, Texas, USA, 26-28 February.

Charlez, P. A. 1999. The Concept of Mud Weight Window Applied to Complex Drilling.

SPE-56758-MS, SPE Annual Technical Conference and Exhibition, Houston, TX, 3-

6 October.

Chenevert, M., and Sharma, M. 2009. Maintaining Shale Stability by Pore Plugging. US

Patent 0314549. 24 December.

Chevron Phillips Chemical Company. 2014. “Lost Circulation Guide”,

http://www.cpchem.com/bl/drilling/en-us/Documents/LostCirculationGuide.pdf.

Accessed 24 January, 2014.

Chevron Phillips Chemical Company. 2014. Products Data Sheets,

http://www.cpchem.com/bl/drilling/en-us/Pages/Products.aspx. Accessed 24 January,

2014.

Collins, N., Whitfill, D., Kharitonov, A., and Miller, M. 2010. Comprehensive Approach

to Severe Loss Circulation Problems in Russia. SPE-135704-MS, SPE Russian Oil

and Gas Conference and Exhibition, Moscow, Russia, 26-28 October.

Contreras, O., Hareland, G., Husein, M., Nygaard, R., and Alsaba, M. 2014a.

Application of In-House Prepared Nanoparticles as Filtration Control Additive to

Reduce Formation Damage. SPE-168116-MS, SPE International Symposium and

Exhibition on Formation Damage Control, Lafayette, USA, 26-28 February.

Contreras, O., Hareland, G., Husein, M., Nygaard, R., and Alsaba, M. 2014b.

Experimental Investigation on Wellbore Strengthening In Shales By Means of

Nanoparticle-Based Drilling Fluids. SPE-170589-MS, SPE Annual Technical

Conference and Exhibition, Amsterdam, the Netherlands, 27–29 October.

Cutler, R., Enniss, D., Jones, A., and Swanson, S.1985. Fracture Conductivity

Comparison of Ceramic Proppants. Society of Petroleum Engineers Journal. Vol. 25

(2), pp. 157-170.

Deeg, W., and Wang, H. 2004. Changing Borehole Geometry and Lost-Circulation

Control. ARMA-04-577, the 6th North America Rock Mechanics Symposium,

Houston, Texas, June 5-9, 2004.

http://www.cpchem.com/bl/drilling/en-us/Documents/LostCirculationGuide.pdf

http://www.cpchem.com/bl/drilling/en-us/Pages/Products.aspx

158

Dick, M. A., Heinz, T. J., Svoboda, C. F., and Aston, M. 2000. Optimizing the Selection

of Bridging Particles for Reservoir Drilling Fluids. SPE-58793-MS, SPE

International Symposium on Formation Damage Control, Lafayette, Louisiana, 23-24

February.

Dudley, J., Fehler, D. F., and Zeilinger, S. 2001. Minimizing Lost Circulation Problems

with Synthetic Muds, GPRI Project 2000 DC 3, December.

Dupriest, F. E. 2005. Fracture Closure Stress (FCS) and Lost Returns Practices. SPE-

92192-MS, SPE/IADC Drilling Conference, Amsterdam, Netherlands, 23-25

February.

Dupriest, F. E., Smith, M. V., Zeilinger, C. S. and Shoykhet, I. N. 2008. Method to

Eliminate Lost Returns and Build Integrity Continuously with High-Filtration-Rate

Fluid. SPE-112656-MS, SPE/IADC Drilling Conference, Orlando, Florida, USA, 4-6

March.

Erdogan, F. and Gupta, G. D. 1972. On the Numerical Solution of Singular Integral

Equations. Quarterly of Applied Mathematics, Vol. 29, pp. 525–534.

Fann Instrument Company (2011) Particle Plugging Apparatus Product Information.

http://www.fann.com/public1/pubsdata/Brochures/PPA_Paper.pdf. Accessed 3

September 2015.

Fann Instrument Company (2015) Linear Swell Meter 2100 Instruction Manual.

http://www.fann.com/public1/pubsdata/Manuals/LSM_2100_Manual_102114531.pdf

. Accessed 31 July 2015.

Fidan, E., Babadagli, T., and Kuru, E. 2004. Use of Cement as Lost-Circulation Material:

Best Practices. PETSOC-2004-090, Canadian International Petroleum Conference,

Calgary, Alberta, 8-10 June.

Filippov, A., Mack, R., Cook, L., York, P., Ring, L., and McCoy, T. 1999. Expandable

Tubular Solutions. SPE-56500-MS, SPE Annual Technical Conference and

Exhibition, Houston, Texas, USA, 3-6 October.

Fjaer, E., Holt, R., Raaen, A., Risnes, R., and Horsud, P. 2008. Petroleum Related Rock

Mechanics, 2nd Edition. Elsevier, Oxford, UK.

Friedheim, J. E., Young, S., De Stefano, G. Lee, J., and Guo, Q. 2012. Nanotechnology

for Oilfield Applications – Hype or Reality?. SPE-157032-MS, SPE International

Oilfield Nanotechnology Conference and Exhibition, Noordwijk, Netherlands, 12-14

June.

Fuh, G. F., Morita, N., Byod, P. A., and McGoffin, S. J. 1992. A New Approach to

Preventing Lost Circulation While Drilling. SPE-24599-MS, SPE Annual Technical

Conference and Exhibition, Washington D.C., USA, 4-7 October.

http://www.fann.com/public1/pubsdata/Brochures/PPA_Paper.pdf



159

GEO Drilling Fluids, Inc. 2010. Product Descriptions,

http://www.geodf.com/index.cfm/fuseaction/Pages.Page/id/302. Accessed 24

January, 2014.

Griffith, A. A., 1920. The Phenomena of Rupture and Flow In Solids. Philosophical

Transactions of the Royal Society of London, A Vol. 221 (4), pp. 163–198.

Griffo, A., and Keshavan, M. 2007. High Performance Rock Bit Greased. US Patent

20070254817 A1. November.

Halliburton. 2014. Products Data Sheet, http://www.halliburton.com/. Accessed 24

January, 2014.

Hands, N., Kowbel, K., Maikranz, S., and Nouris, R. 1998. Drill-in Fluid Reduces

Formation Damage, Increases Production Rates. Oil & Gas Journal, Vol. 96 (28), pp.

65-69.

He, W., and Stephens, M. P. 2011. Bridging Particle Size Distribution in Drilling Fluid

and Formation Damage. SPE-143497-MS, SPE European Formation Damage

Conference, Noordwijk, The Netherlands, 7-10 June.

Hettema, M., Horsrud, P., Taugbol, K., Friedheim, J., Huynh, H., Sanders, M. W., and

Young, S. 2007. Development of an Innovative High-Pressure Testing Device for the

Evaluation of Drilling Fluid Systems and Drilling Fluid Additives within Fractured

Permeable Zones. OMC-2007-082, Offshore Mediterranean Conference and

Exhibition, Ravenna, Italy, 28-30 March.

Hillerborg, A., Modeer, M., and Petersson, P. 1976. Analysis of Crack Formation and

Crack Growth in Concrete by Means of Fracture Mechanics and Finite Elements.

Division of Building Materials Lund Institute of Technology, Lund, Vol. 6, pp. 773–

782, Pergamum, Sweden.

Hinkebein, T., Behr, V., and Steve L. Wilde. 1983. Static Slot Testing of Conventional

Lost Circulation Materials, SAND82-1080, January.

Hoelscher, K. P., De Stefano, G., Riley, M. and Young, S. 2012. Application of

Nanotechnology in Drilling Fluids. SPE-157031-MS, SPE International Oilfield

Nanotechnology Conference and Exhibition, Noordwijk, Netherlands, 12-14 June.

Howard, G. C., and Scott, P. P. 1951. An Analysis and the Control of Lost Circulation.

Journal of Petroleum Technology, Vol. 3 (6), pp. 171-182.

Irwin, G. R., and Kies J. A. 1954. Critical Energy Rate Analysis of Fracture Strength.

Welding Journal, Vol. 33, pp. 193-198.

Jahn, F., Cook, M., and Graham, M. 2008. Hydrocarbon Exploration & Production, 2nd

Edition. Elsevier Science, Oxford, UK.

http://www.geodf.com/index.cfm/fuseaction/Pages.Page/id/302

http://en.wikipedia.org/wiki/Alan_Arnold_Griffith

http://www.cmse.ed.ac.uk/AdvMat45/Griffith20.pdf

160

Kaageson-Loe, N. M., Sanders, M. W., Growcock, F., Taugbol, K., Horsrud, P.,

Singelstad, A. V., and Omland, T. H. 2009. Particulate-Based Loss-Prevention

Material - The Secrets of Fracture Sealing Revealed!. SPE Drilling & Completion,

Vol. 24(4), pp. 581-589.

Karimi, M., Petrie, S. W., Moellendick, E., and Holt, C. 2011. A Review of Casing

Drilling Advantages to Reduce Lost Circulation, Augment Wellbore Strengthening,

Improve Wellbore Stability, and Mitigate Drilling-induced Formation Damage. SPE-

148564-MS, SPE/IADC Middle East Drilling Technology Conference and

Exhibition, Muscat, Oman, 24-26 October.

Kelsey, J. R. 1981. Geothermal Drilling and Completion Technology Development

Program Quarterly Progress Report. SAND81-1020, June.

Kumar, A., and Savari, S. 2011. Lost Circulation Control and Wellbore Strengthening:

Looking Beyond Particle Size Distribution. AADE-11 NTCE-21, AADE National

Technical Conference and Exhibition, Houston, Texas, USA, 12-14 April.

Kumar, A., Savari, S., and Jamison, D. E., and Whitfill, D. 2011. Application of Fiber

Laden Pill for Controlling Lost Circulation in Natural Fractures. AADE-11 NTCE-

19, AADE National Technical Conference and Exhibition, Houston, Texas, USA, 12-

14 April.

Lecolier, E., Herzhaft, B., Rousseau, L., Neau, L., Quillien, B., and Kieffer, J. 2005.

Development of a Nanocomposite Gel for Lost Circulation Treatment. SPE-94686-

MS, SPE European Formation Damage Conference, Sheveningen, The Netherlands,

25-27 May.

Lohoefer, C. L., Mathis, B., Brisco, D., Waddell, K., Ring, L., and York, P. 2000.

Expandable Liner Hanger Provides Cost-Effective Alternative Solution. SPE-59151-

MS, IADC/SPE Drilling Conference, New Orleans, Louisiana, USA, 23-25

February.

Marinescu, Pavel. 2014. “Loss Circulation Challenges – Current Approach and What are

the Next Solutions?”. SPE Lost Circulation Workshop, Dubai, 20-22 May.

Mata, F., and Veiga, M. 2004. Crosslinked Cements Solve Lost Circulation Problems.

SPE-90496-MS, SPE Annual Technical Conference and Exhibition, Houston, Texas,

USA, 26-29 September.

Messenger, J. U. 1981. Lost Circulation. PennWell Corp, Tulsa, Oklahoma.

Messenger, J. U., and McNiel, J. S. 1952. Lost Circulation Corrective: Time-Setting Clay

Cement. Journal of Petroleum Technology, Vol. 4 (3), pp. 59-64.

Messina Chemicals. 2012. Products Data Sheet,

http://www.messinachemicals.com/index.php/products/drilling-chemicals/lost-

circulation-seepage-control-materials. Accessed 24 January, 2014.

161

Montgomery, D. 2001. Design and Analysis of Experiments, 5th Edition. John Wiley &

Sons, Inc., New York, NY.

Morita, N., and Fuh, G-F. 2012. Parametric Analysis of Wellbore-Strengthening Methods

from Basic Rock Mechanics. SPE Drilling & Completion, Vol. 27(2), pp. 315-327.

Morita, N., Black, A. D., and Fuh, G-F. 1990. Theory of Lost Circulation Pressure. SPE-

20409-MS, SPE Annual Technical Conference and Exhibition, New Orleans,

Louisiana, USA, 23-26 September.

Morita, N., Fuh, G-F., and Black, A. D. 1996a. Borehole Breakdown Pressure with

Drilling Fluids-I. Empirical Results. International Journal of Rock Mechanics and

Mining Sciences & Geomechanics, Vol. 33 (1), pp. 39-51.

Morita, N., Fuh, G-F., and Black, A. D. 1996b. Borehole Breakdown Pressure with

Drilling Fluids-II. Semi-Analytical Solution to Predict Borehole Breakdown Pressure.

International Journal of Rock Mechanics and Mining Sciences & Geomechanics, Vol.

33 (1), pp. 53-69.

Mostafavi, V., Hareland, G., Belayneh, M. and Aadnoy, B. S. 2011. Experimental and

Mechanistic Modeling of Fracture Sealing Resistance with Respect to Fluid and

Fracture Properties. ARMA 11-98, the 45th US Rock Mechanics Symposium, San

Francisco, USA, 26-29 June.

Nayberg, T. M. 1987. Laboratory Study of Lost Circulation Materials for use in Both Oil-

base and Water-base Drilling Muds. SPE Drilling Engineering, Vol. 2 (3), pp. 229-

236.

Newman, J. C. 1971. An Improved Method of Collocation for the Stress Analysis of

Cracked Plates with Various Shaped Boundaries. NASA Technical Notes. TN D-

6376. National Aeronautics and Space Administration. Washington D.C.

Newpark Drilling Fluids. 2012. Products Data Sheet,

http://www.newpark.com/capabilities/newpark-drilling-fluids/united-states/about-

us/product-bulletins. Accessed 24 January, 2014.

Nwaoji, C.O., Hareland, G., Husein, M., Nygaard, R., Zakaria, M.F. 2013. Wellbore

Strengthening-Nano-Particle Drilling Fluid Experimental Design Using Hydraulic

Fracture Apparatus. SPE-163434-MS, SPE/IADC Drilling Conference and

Exhibition, Amsterdam, the Netherlands, 5 – 7 March.

Nygaard, R. and Salehi, S. 2011. A Critical Review of Wellbore Strengthening: Physical

Model and Field Deployment. AADE-11-NTCE-24, AADE National Technical

Conference and Exhibition, Houston, USA, 12-14 April

Onyia, E. C. 1994. Experimental Data Analysis of Lost-Circulation Problems during

Drilling with Oil-Based Mud. SPE Drilling and Completion, Vol. 9 (1), pp. 25-31.

http://www.newpark.com/capabilities/newpark-drilling-fluids/united-states/about-us/product-bulletins

http://www.newpark.com/capabilities/newpark-drilling-fluids/united-states/about-us/product-bulletins

162

Paiaman, M., and Al-Anazi, D. 2008. Using Nanoparticles to Decrease Differential Pipe

Sticking and its feasibility in Iranian Oil Field. Oil and Gas Business.

Perez-Roca, E., Andrews, S., and Keel, D. 2003. Addressing Common Drilling

Challenges Using Solid Expandable Tubular Technology. SPE-80446-MS, SPE Asia

Pacific Oil and Gas Conference and Exhibition, Jakarta, Indonesia, 9-11 September.

Powers, M. C. 1953. A New Roundness Scale for Sedimentary Particles. Journal

Sedimentary Research, Vol 23 (2), pp. 117-119.

Rehm, B., Haghshenas, A., Paknejad A. S., Al-Yami, A., and Hughes, J., 2012.

Underbalanced Drilling: Limits and Extremes. Gulf Publishing Company, Houston,

TX.

Rong, G., Liu, G., Hou, D., and Zhou, C. 2013. Effect of Particle Shape on Mechanical

Behaviors of Rocks: A Numerical Study Using Clumped Particle Model. The

Scientific World Journal, Vol. 2013.

Saasen, A., Godøy, R., Breivik, D.H., Solvang, S.A., Svindland, A., Gausel, E., and

Frøyland, K. 2004. Concentrated Solid Suspension as an Alternative to Cements for

Temporary Abandonment Applications in Oil Wells. SPE-SA-34, SPE Technical

Symposium of Saudi Arabia Section, Dhahran, Saudi Arabia, 15-17 May.

Saasen, A., Wold, S., Ribesen, B.T., Tran, T.N., Huse, A., Rygg, V., Grannes, I., and

Svindland, A. 2011. Permanent Abandonment of a North Sea Well Using

Unconsolidated Well Plugging Material. SPE Drilling & Completion, Vol. 26 (3), pp.

371-375.

Salehi, S. 2012. Numerical Simulations of Fracture Propagation and Sealing: Implication

for Wellbore Strengthening. PhD dissertation, Department of Geological Sciences

and Engineering, Missouri University of Science and Technology.

Salehi, S. and Nygaard, R. 2011. Evaluation of New Drilling Approach for Widening

Operational Window: Implications for Wellbore Strengthening. SPE-140753-MS,

SPE Production and Operations Symposium, Oklahoma City, USA, 27-29 March.

Salehi, S. and Nygaard, R. 2012. Numerical Modeling of Induced Fracture Propagation:

A Novel Approach for Lost Circulation Materials (LCM) Design in Borehole

Strengthening Applications of Deep Offshore Drilling. SPE-135155-MS, SPE Annual

Technical Conference and Exhibition, San Antonio, USA, 8-10 October.

Salehi, S. and Nygaard, R. 2015. Developing a Full Fluid-Solid Finite-Element Model to

Simulate near Wellbore Fractures for Wellbore Strengthening Applications. ASME

Journal of Energy Resources Technology, Vol. 137 (1).

Sall, J. 1990. Leverage Plots for General Linear Hypotheses. The American Statistician,

Vol. 44 (4), pp. 308-315.

163

Savari, S., Kumar, A., Whitfill, D. L., Miller, M., Murphy, R. J., and Dale E. Jamison.

2013. Engineered LCM Design Yields Novel Activating Material for Potential

Application in Severe Lost Circulation Scenarios. SPE-164748-MS, SPE North

Africa Technical Conference and Exhibition, Cairo, Egypt, 15-17 April.

Savari, S., Whitfill, D. l., and Kumar, A. 2012. Resilient Lost Circulation (LCM): A

Significant Factor in Effective Wellbore Strengthening. SPE-153154-MS, SPE

Deepwater Drilling and Completions Conference, Galveston, Texas, USA, 20-21

June.

Schlumberger. 2013. Products Data Sheet,

http://www.slb.com/services/miswaco/services/drilling_fluid/df_products/lost_circula

tion_material.aspx. Accessed 24 January, 2014.

Schlumberger. 2014. Oil Field Glossary, http://www.glossary.oilfield.slb.com. Accessed

24 January, 2014.

Sharp-Rock Technologies, Inc. 2014. Product Data Sheet, www.sharp-rock.com.

Accessed 2 February, 2014.

Shepard, S. F., Reiley, R. H., and Warren, T. M. 2002. Casing Drilling: An Emerging

Technology. SPE Drilling & Completion, Vol. 17 (1), pp. 4-14.

Shinohara, K., Oida, M., and Golman, B. 2000. Effect of Particle Shape on Angle of

Internal Friction by Triaxial Compression Test. Powder Technology, Vol. 107 (2), pp.

131-136.

Smith, P. S., Browne, S. V., Heinz, T. J., and Wise, W. V. 1996. Drilling Fluid Design to

Prevent Formation Damage in High Permeability Quartz Arenite Sandstones. SPE-

36430-MS, SPE Annual Technical Conference and Exhibition, Denver, Colorado, 6-9

October.

Sweatman, R. E., and Scoggins, W. C. 1988. Acid-Soluble Magnesia Cement: New

Applications in Completion and Workover Operations. SPE-18031-MS, SPE Annual

Technical Conference and Exhibition, Houston, USA, 2-5 October.

Tada, H., Paris, P., Irwin, G. 2000. The Stress Analysis of Cracks Handbook, 3rd

Edition.

ASME Press, New York.

Tehrani, A., Cliffe, A., Hodder, M. H., Young, S., Lee, J., Stark, J., and Seale, S. 2014.

Alternative Drilling Fluid Weighting Agents: A Comprehensive Study on Ilmenite

and Hematite. SPE-167937-MS, IADC/SPE Drilling Conference and Exhibition, Fort

Worth, Texas, USA, 4-6 March

Tehrani, A., Friedheim, J., Cameron, J., and Reid, B. 2007. Designing Fluids for

Wellbore Strengthening – Is It an Art?. AADE-07-NTCE-75, AADE National

Technical Conference and Exhibition, Houston, Texas, USA, April 10-12.

http://www.slb.com/services/miswaco/services/drilling_fluid/df_products/lost_circulation_material.aspx

http://www.slb.com/services/miswaco/services/drilling_fluid/df_products/lost_circulation_material.aspx

http://www.glossary.oilfield.slb.com/

http://www.sharp-rock.com/

164

The Drilling Manual, Fifth Edition, 2015. CRC Press, Boca Raton, FL.

Van Oort, E., Friedheim, J., Pierce, T., and Lee, J. 2011. Avoiding Losses in Depleted

and Weak Zones by Constantly Strengthening Wellbores. SPE Drilling &

Completion, Vol. 26 (4), pp. 519-530.

Vickers, S., Cowie, M., Jones, T., and Twynam, A. J. 2006. A New Methodology that

Surpasses Current Bridging Theories to Efficiently Seal a Varied Pore Throat

Distribution as Found in Natural Reservoir Formations. AADE-006-DF-HO-16.

AADE Fluids Conference and Exhibition, Houston, Texas, USA, 11-12 April.

Wang H, Sweatman R, Engelman B, Deeg W, Whitfill D, Soliman M, Towler BF. 2008.

Best Practice in Understanding and Managing Lost Circulation Challenges. SPE

Drilling & Completion, Vol. 23 (2), pp. 168-175.

Wang, H. 2011. Is It Really Possible to Efficiently Form A Strong Seal inside

Subterranean Openings without Knowing Their Shape and Size?. AADE-11-NTCE-

25, AADE National Technical Conference and Exhibition, Houston, USA, 12-14

April.

Warren, T., Tessari, R., and Houtchens, B. 2004. Casing Drilling with Retrievable

Drilling Assemblies. OTC-16564-MS, Offshore Technology Conference, Houston,

Texas, USA, 3-6 May.

Warren, W. E. 1982. The Quasi-Static Stress Field around a Fractured Wellbore.

International Journal of Fracture, Vol. 18 (2), pp. 113–124.

Weideman, B. 2014. Investigation of Cased Wellbore Integrity in the Wabamun Area

Sequestration Project. MSc Thesis, Department of Geological Sciences and

Engineering, Missouri University of Science and Technology.

White, R. J. 1956. Lost-Circulation Materials and Their Evaluation. API-56-352.

Drilling and Production Practice, New York, New York, USA, 1 January.

Whitfill, D. 2008. Lost Circulation Material Selection, Particle Size Distribution and

Fracture Modeling with Fracture Simulation Software. SPE-115039-MS, IADC/SPE

Asia Pacific Drilling Technology Conference and Exhibition, Jakarta, Indonesia, 25-

27 August.

Zakaria, M. 2013. Nanoparticle-Based Drilling Fluids with Improved Characteristics.

Ph.D. Dissertation. Department of Chemical and Petroleum Engineering. University

of Calgary, Calgary, AB, Canada.

Zhang, K., Chanpura, R. A., Mondal, S., Wu, C-H., Sharma, M. M., Ayoub, J. A., and

Parlar, M. 2015. Particle-Size-Distribution Measurement Techniques and Their

Relevance or Irrelevance to Wire-Wrap-Standalone-Screen Selection for Gradual-

Formation-Failure Conditions. SPE Drilling & Completion, Vol. 30 (2), pp. 164-174.

165

VITA

Mortadha Alsaba was born the 17th

of April 1985, in Dhahran, Saudi Arabia. He

received a BEng. Degree in Petroleum Engineering with first class honours from London

South Bank University, London, UK, in 2007. He also received a MSc. in Petroleum

Engineering from Missouri University of Science and Technology, Rolla, MO, in 2012

and a PhD in Petroleum Engineering from Missouri University of Science and

Technology in 2015. He has published and presented several papers on lost circulation

prevention and mitigation.

Alsaba is a member of several professional associations, including Society of

Petroleum Engineering (SPE), American Association of Drilling Engineers (AADE),

American Association of Petroleum Geologist (AAPG) and the American Society of

Mechanical Engineers (ASME). He has served as a secretary for the AADE student

chapter from 2013-2014.

His research interests are in prevention and mitigation of lost circulation events,

laboratory evaluation of wellbore strengthening techniques, drilling fluids design and

optimization.