TABLE OF CONTENTS INTRODUCTION AND STATEMENT OF PURPOSE 6 SUMMARY OF EDUCATIONAL BACKGROUND AND WORK EXPERIENCE 6 EXECUTIVE SUMMARY 9

Transcription

1

2 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 2 TABLE OF CONTENTS INTRODUCTION AND STATEMENT OF PURPOSE 6 INTRODUCTION 6 STATEMENT OF PURPOSE 6 SUMMARY OF EDUCATIONAL BACKGROUND AND WORK EXPERIENCE 6 EDUCATION 6 WORK EXPERIENCE 7 EXECUTIVE SUMMARY 9 SUMMARY OF OPINIONS 11 DEEPWATER DRILLING AND BP PRESSURE INTEGRITY TESTS 11 HISTORICAL PERSPECTIVE 12 EVOLUTION OF ROLES AND RESPONSIBILITIES 13 OVERBURDEN STRESS 20 FRACTURE GRADIENT 21 PRESSURE INTEGRITY TESTS 23 DRILLING MARGINS 28 SAFE DRILLING MARGIN 33 BP S PRESSURE INTEGRITY TEST OF THE 9-7/8 INCH LINER SHOE 42 EFFECT OF SPACER ON NEGATIVE PRESSURE TEST INTERPRETATION 48 NEGATIVE TEST NOMENCLATURE 49 FLUID DISPLACEMENT 54 MI-SWACO SPACER 54 NEGATIVE PRESSURE TEST 59 WELL MONITORING FAILURE 64 APPENDIX A 69

3 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 3 APPENDIX B 70 APPENDIX C 71 APPENDIX D 90

4 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 4 LIST OF FIGURES Figure 1: Deepwater Drilling Units and Production Facilities 13 Figure 2a: Pore pressure in hydrostatic equilibrium with surface water 18 Figure 2b: Pore Pressure is controlled by Mud Pressure 18 Figure 2: What is Pore Pressure? 18 Figure 3: What is Mud Weight? 19 Figure 4: Overburden Stress calculated from Macondo Well Logs 21 Figure 5: What is Fracture Gradient 22 Figure 6: Comparison of Calculated and Observed Leak-off Test Results 23 Figure 7: Typical Leak-Off Test Plot 26 Figure 8: Typical Pressure Integrity Test Plot 27 Figure 9: Approximate effect of Water Depth on Fracture Gradient at 3500 feet 28 Figure 10: What is a Kick? 29 Figure 11: Shoe Strength protects Weaker Sediments Above 31 Figure 12: What is an Underground Blowout? 32 Figure 13: Pressure Integrity Test at 13-5/8 inch Liner Shoe 41 Figure 14: Pressure Integrity Test at 9-7/8 inch Liner Shoe 45 Figure 15: Common Assumptions in Fracture Modeling 48 Figure 16: Schematic of Macondo Well after Circulating Cement 50 Figure 17: Planned Configuration of Macondo Well with Temporary Cap 51 Figure 18: Use of Equivalent Mud Weights in Negative Test Nomenclature 53 Figure 19: Displacement Planned by MI-Swaco Drilling Fluid Specialist 55 Figure 20: Plot of Real Time Data Recorded during Negative Pressure Test 61 Figure 21: U-tube analogy for Negative Test being conducted (pressures equal) 64 Figure 22: Real Time Data Showing Kick Indications that were not detected 66

5 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 5 LIST OF TABLES Table 1 Summary of Events during the Initial Displacement 58 Table 2 Summary of Actions Taken during the First Negative Pressure Test 60 Table 3 Summary of Actions Taken during the Second Negative Pressure Test 62

6 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 6 INTRODUCTION AND STATEMENT OF PURPOSE INTRODUCTION The BP Exploration & Production Inc. Macondo Well (OCSG ) located offshore Louisiana in Mississippi Canyon Block 252 was in the process of being temporarily capped when a tragic accident occurred in which eleven people lost their lives and 17 others were injured. The rig Deepwater Horizon, sank, and the well continued to flow oil into the Gulf of Mexico for 87 days before it was successfully plugged. The well had been successfully drilled, the production casing had been run, and cement circulated at the time of the blowout. However, an event of this magnitude invariably raises questions about the drilling operations as well as the temporary capping operations. STATEMENT OF PURPOSE Upon the request of counsel representing BP, I have reviewed the daily operational summaries and reports, reports of other experts, and other materials provided. The list of those materials is included in Appendix D. I was asked to review various aspects of the drilling operations for the Macondo well. This report summarizes my opinions reached based on my reviews and analysis. It also provides schematics illustrating well conditions during various operations. These opinions are based on the information I have reviewed at this time. My work is ongoing and I reserve the right to adjust my opinion as new information is obtained. SUMMARY OF EDUCATIONAL BACKGROUND AND WORK EXPERIENCE EDUCATION B. S. in Petroleum Engineering, Cum Laude, May 28, 1966 Louisiana State University M. S. in Petroleum Engineering, January 23, 1968 Louisiana State University Ph.D. in Petroleum Engineering, January 24, 1970 University of Texas at Austin

7 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 7 WORK EXPERIENCE I have more than 45 years of experience in the oil and gas industry, particularly in Louisiana and the Gulf of Mexico. My work experience in the oil and gas industry began through participation in summer/co-op programs while in college. I worked for Mobil Oil Company three months as an onshore roustabout and three months as an offshore roustabout. After reaching senior status at Louisiana State University (LSU), I worked three months as an Engineering Assistant involved with offshore drilling and well work-over planning. After receiving my B.S. Degree and prior to entering graduate school, I worked three months for Texaco as an Assistant Drilling Engineer involved with offshore field operations, well planning, and drilling optimization. My training for this position included working as a floor hand on the first semi-submersible rig, the Ocean Driller. After entering graduate school, I worked three months for Chevron at their research laboratory in La Habra, California and three months for Conoco at their research laboratory in Ponca City, Oklahoma. In 1969, after completion of my course work at the University of Texas at Austin (UT- Austin), I joined Conoco in Houston as a Senior Systems Engineer in their Production Engineering Services Group. There, I participated in several drilling and production projects including an offshore drilling project involving real-time drilling data acquisition and estimation of formation pore pressure. In 1971, I joined LSU as an Assistant Professor. For the next 29 years, I worked in the undergraduate, graduate, and continuing education programs of the LSU Petroleum Engineering Department and in administration of the College of Engineering. I had primary responsibility for the drilling engineering and drilling fluids laboratory courses, but taught production engineering and reservoir engineering courses as well. I served as Chairman of the Petroleum Engineering Department from 1977 to At the time of my retirement in December of 1999, I was the Campanile Charities Professor of Offshore Mining and Petroleum Engineering and Dean of the College of Engineering. I have been especially active in the area of blowout prevention. Soon after joining LSU in 1971, I began assisting Professor Bill Hise, one of the more senior LSU faculty members, in teaching industry short-courses on well control for onshore and bottom-supported offshore drilling rigs. Under the sponsorship of the International Association of Drilling Contractors (IADC), Professor Hise had reworked an abandoned well on the LSU Campus for use in hands on well control training including the safe circulation of a threatened gas blowout. Building on this foundation, LSU founded the first blowout prevention training program with open enrollment. The program was enthusiastically received by the industry and several hundred industry participants per year attended the program during the 1970s. Discussions held with a

8 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 8 wide cross-section of industry participants provided me valuable insight into the complications that can arise during well control operations. I became particularly interested in complications associated with deepwater drilling operations with the blowout preventer at the seafloor. A list of publications, training courses and other relevant items is included in Appendix C. Starting in 1979, I guided the development of a multi-million dollar research and training well facility at LSU to support work on deepwater well control and to complement the older training well. The newer facility was funded through the combined support of 13 major oil companies, 40 service companies, and the Minerals Management Service (MMS) (now the Bureau of Ocean Energy Management, Regulation and Enforcement or BOEMRE), and became operational in The facility was initially centered around a 6000-foot well specially configured to model the full-scale well control flow geometry of a floating drilling rig in 3000 feet of water. Extensive surface equipment provided for hands on training as well as highlyinstrumented well control experiments. Gas could be injected into the bottom of the well to initiate the conditions of a threatened blowout. The goal of the research was the development of improved well control procedures and training for deepwater drilling operations. The facility, which still operates today, was later expanded to include additional wells and model diverter components for experimental study of flow erosion and pressures seen during diverter operations. This research was aimed at reducing the incidence of failures in diverters used to handle a shallow gas flow that could not be safely shut-in. Under sponsorship of Amoco (now BP) and the Drilling Engineers Association (DEA), the facility was further expanded to include an additional 6000-foot well to study kick detection and other potential well control complications associated with gas solubility in oil base muds. Between 1981 and my retirement in 1999, I supervised the graduate research of 19 MS theses and 12 PhD dissertations on various well control topics of interest to industry and the MMS (now BOEMRE). Numerous Well Control Research Workshops were held at LSU during this period and were well attended by both MMS and industry personnel. The research has resulted in more than 50 publications related to well control and including formation pore pressure estimation, fracture gradient correlations, Leak-off test data, modeling well control and relief well operations, and improved procedures for safe removal of a gas influx. During this period I also organized and helped to teach specialized deepwater well control schools for Amoco, Exxon, Shell, Conoco, Phillips, and Zapata as well as numerous open enrollment schools. I am the lead author of the Society of Petroleum Engineers (SPE) Drilling Engineering Textbook, entitled Applied Drilling Engineering which was developed for petroleum engineering college curriculums. This textbook is widely accepted and has been a top seller for SPE since it was first published in I have also written Drilling Practices, a chapter in

9 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 9 the Encyclopedia of Chemical Processing and Design and Shallow Gas Blowouts, a chapter in Firefighting and Blowout Control. I have also written several chapters in a well control manual used in LSU s well control schools. I have served as chairman of the SPE reprint series on Pore pressure and Fracture Gradient Determination and also for another reprint series on Well Control. I am a past recipient of the SPE Distinguished Achievement Award for Petroleum Engineering Educators and have received the SPE Drilling Engineering Award for distinguished contributions to petroleum engineering in the area of drilling technology. In 1990, I was selected as a Distinguished Member of SPE. In , I was selected as a Distinguished Lecturer by the SPE and gave lectures at about 30 locations in the U.S., Europe, and Middle East. During 1998 I also served on a steering committee of the International Association of Drilling Contractors (IADC) that coordinated the development of a manual on well control practices for deepwater drilling operations. I have also served on an ad hoc committee of SPE to review the exam leading to professional registration of Petroleum Engineers. Upon my retirement from LSU, I was recognized by the House of Representatives of the State of Louisiana in House Concurrent Resolution No. 33 of the First Extraordinary Session, 2000, commending me for achievements in scholarly research and writing in the field of petroleum engineering and for highly significant contributions to higher education in Louisiana. In December of 2001, I was recognized as a Distinguished Graduate of the College of Engineering at the University of Texas at Austin. In 2006, I was inducted into the LSU Engineering Hall of Distinction. I am currently President of Bourgoyne Enterprises, Inc., which offers Petroleum Engineering consulting services to the Oil and Gas Industry. I have consulted extensively with Pennington Oil and Gas, LLC in their drilling and completion of deep, high-temperature, high-pressure wells in the Tuscaloosa Trend Area of Louisiana. I have also served as an expert witness on blowout and well control matters. EXECUTIVE SUMMARY The blowout prevention practices and equipment used in the oil and gas industry are designed to maintain redundant and verifiable barriers to an uncontrolled release of formation fluids. For a blowout to occur there generally must be a perfect storm series of both equipment failures and human failures (mistakes) to circumvent all of the intended barriers. The Deepwater Horizon tragedy was not an exception to this rule. As many of the publicly available reports have concluded, the equipment failures required for the blowout to occur included (1) the failure of the cement to seal the annular space through the oil and gas zone, (2) the failure of the cement to seal the shoe track of the production casing, and (3) the failure of the blowout

10 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 10 preventers to stop the flow at the seafloor. The human failures (mistakes) required for the blowout to occur included (1) the failure to correctly interpret a critical negative pressure test of the cement seals, (2) the failure to detect the influx of formation fluids into the casing, and (3) the failure to use the diverter system to vent overboard the oil and gas that reached the surface. The human failures are especially hard to understand because there are multiple groups on the rig tasked with constantly monitoring the well during all operations, including the Transocean rig crew and the Sperry-Sun mud loggers. The ability to detect an influx of formation fluid of less than 30 barrels is an expected normal response, and numerous drills are generally conducted while drilling to test this response. An influx of over 600 barrels was required for the formation fluids to reach the seafloor undetected and begin entering the marine riser. It is unknown why standard well control training concepts were not followed that night. If the situation was such that rig pumps were shut down at 9:30, standard well control would dictate that the well be checked for flow, and shut in if flowing, before taking any other steps. Both a driller and a toolpusher were evaluating the well at that time. Yet, on that night, the next action taken was to bleed the drill pipe. Additionally, negative pressure tests involve simple hydrostatic pressure concepts that form the basis of many rig site calculations and the needed hydrostatic pressure concepts are always included in well control training. The parties interpreting the negative pressure test made a mistake in determining that the negative pressure tests run by Transocean s rig crew were successful. I reviewed the BP Daily Operational Summaries for the Macondo Well, real time data collected during drilling and other rig operations, deposition transcripts, discovery materials, expert reports issued by other parties and the reports of investigative bodies issued in regard to the tragic accident that occurred. See Appendix D. I saw no evidence that BP drilled this well in a careless or reckless manner with disregard for safety in order to save money and drill fast. There are always advantages and disadvantages to various design options considered when drilling and completing a well. On any given issue, it is normal for members of a drilling team in a large organization to have differences of opinion as to which option is better. In my opinion, all of the final choices made in regard to planning fell within the range of normal drilling and completion practices. A critical review of the planning decisions must recognize that the well was successfully drilled, the production casing was successfully run to bottom, and cement was circulated into the well as planned.

11 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 11 SUMMARY OF OPINIONS I have reached the following opinions to a reasonable engineering certainty in this matter: First, this well was drilled to total depth safely and successfully and consistent with industry practices. The Deepwater Horizon crew and BP s team followed industry practices in conducting formation integrity tests, interpreting those tests and properly managing their mud weights to safely drill each interval. Second, the use of the Loss Circulation Material (LCM) spacer during the temporary abandonment procedures on April 20, 2010 likely resulted in fluids being displaced in a manner that the rig crew did not fully understand. Third, the negative pressure test was not properly understood, but the parties involved allowed the displacement to continue. While everyone involved had every incentive to get it right and had no incentive to cut corners, they did not properly interpret the test. Fourth, the rig crew failed to properly monitor the well on April 20, 2010 and allowed a kick to go unnoticed until it turned into a blowout. Even after recognizing that signs of a kick were present, the Transocean rig crew failed to shut in the well immediately. DEEPWATER DRILLING AND BP PRESSURE INTEGRITY TESTS Several issues related to normal drilling practice have arisen in this litigation. These issues are best understood after a discussion of how the practice evolved as the industry moved from land and shallow water operations offshore to floating drilling operations in deep water. A background discussion of my perspective on these issues is provided regarding: Historical Perspective Evolution of Roles and Responsibilities Pore Pressure Overburden Stress Fracture Gradients Pressure Integrity Test Drilling Margins Safe Drilling Margins Negative Pressure Test

12 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 12 HISTORICAL PERSPECTIVE 1 Offshore drilling began in California in 1887 in an area where wells drilled nearest to the ocean tended to be better producers. Wharfs were constructed that served as the first platforms for offshore drilling, the longest extending almost a quarter mile offshore. In the early 1930s along the Louisiana and Texas inland waters, rigs built on steel barges were constructed that could be towed to a well s location and sunk to rest on bottom with the top of the barge still above water. At the end of the job, the derrick was lowered, the barge pumped out to a floating position, and the rig towed to another location. The use of barge rigs was gradually extended from inland waters in the coastal areas to shallow waters offshore. To extent the water depth ranged to about 40 feet, shell mats were deposited at the well location and the barge was sunk onto the shell mat. The barge rigs evolved into submersible rigs and jackups to further extend the water depth. The first submersible rig was basically an upper and lower barge type structure connected by steel columns. The jackup design was a barge with long legs that could be raised when moving the rig and lowered to jack the barge up to a safe height above the waves. For development wells, offshore platforms were designed to allow a rig to be put on the platform and used to drill multiple directional wells. The submersible rig evolved into the semisubmersible rig, which could be anchored over the well location and the well drilled from a floating position. The semisubmersible Ocean Driller, launched in 1963, was the first drilling vessel designed to drill in a floating position. Since that time, a variety of mobile offshore drilling units and production facilities have been designed (Figure 1). These designs have allowed the search for oil and gas to extend beyond the continental shelf into the deep water of the continental slope. Anchoring becomes more difficult as water depth increases. Dynamically positioned drillships and semisubmersibles were developed capable of keeping the drilling vessel on location over the well without the use of anchors. Tension Leg Platforms, Compliant Towers, and Spars are used for field development with multiple wells drilled directionally through a template at the seafloor. Deepwater wells are generally capable of producing at very high flow rates because of the characteristics of the turbidite sands found in deepwater. Deepwater drilling in the Gulf of Mexico offers one of the best opportunities for increasing domestic oil production. 1 This background section is based on my professional experience as summarized in Appendix C, my drilling book, (footnote continued)

13 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 13 Figure 1: Deepwater Drilling Units and Production Facilities EVOLUTION OF ROLES AND RESPONSIBILITIES I have seen the roles and responsibilities for companies involved in drilling operations change during my 46 years of association with the oil and gas industry. In the earlier portion of my career, the oil and gas operators did most of the engineering work. The oil companies were in the final stages of shifting from owning the drilling rigs and manning the rigs with their employees to contracting for rigs and drilling services with drilling contractors. Some oil companies still maintained a few rigs but had sold most of their equipment during the bust of the 1950s that followed the boom after World War II. Petroleum engineering enrollments had dwindled to a trickle as jobs were scarce during this period of contraction. Petroleum engineering enrollments hit bottom in 1963 and began growing again over the next decade. During this period, almost all of the petroleum engineering graduates went to work for major oil companies. Oil company employees did almost all of the engineering work associated with drilling, completing, producing, and managing the oil and gas reservoirs. Most of the major oil companies had active research and development organizations and my papers on petroleum engineering manpower supply and demand.

14 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 14 and almost all of the technical papers written and presentations on new technology made at technical conferences and meetings of the Society of Petroleum Engineers (SPE) were by employees of oil companies and a few large service companies. Work done by service companies was primarily at the field technician level. Procedures used in the field were developed largely by the well operators and both engineering and operational supervision was provided almost exclusively by the oil company. The Arab oil embargo created a domestic oil crisis and oil and gas activities quickly ramped up during the 1970s in response to increasing oil prices. During the boom period that followed, national petroleum engineering enrollment grew from about 2,000 in 1975 to over 10,000 in 1982 in response to high demand and starting salaries. However, the boom ended with another bust before many of these students could graduate. The enrollment fell just as fast as it went up as major changes again took place in the way the oil companies did business. Oil companies became more international and began contracting out more of their engineering work to service companies. While I was serving on the board of directors of SPE in the mid 1980 s, the society made major changes towards becoming a truly international technical society in response to these changes in the industry. During the 1990s, service companies grew, conducted more research and development, and achieved leadership positions in many technical fields associated with oil and gas drilling. Gradually, the number of technical papers and presentations on new developments at SPE conferences and meetings that were made by engineers employed by service companies increased and became dominant. Service companies increased their training programs and began hiring new graduates as well as experienced engineers. The number of major domestic oil companies decreased and the independent oil companies began to play a major role in domestic operations. This trend has continued through the 2000s with mergers and acquisitions of both major oil companies and service companies as they positioned themselves for competition in a more global economy. The latest boom in oil prices began in Drilling for oil and gas in deep water requires teamwork from many specialists. Service company personnel have offices in oil company facilities and work together with oil company personnel in planning and coordinating field operations. The well operator has overall responsibility for drilling, completion, production, and reservoir management. However, the operator transfers some of these responsibilities to service companies when they hire them to use highly specialized procedures and equipment on their behalf. The shared responsibility is controlled by contracts, work agreements and applicable law. The operator s representative on the rig generally does not have the specialized knowledge to supervise all of the work. The operator s representative coordinates the implementation of the procedures provided to him,

15 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 15 helps to coordinate logistics, sees that company policy is followed, and reports on the progress being made. BP s representatives on the rig are called Well Site Leaders. They are trained in company safety policies, well control policy and approved operational procedures. General planning documents are generally discussed in meetings (Spud Meetings) held prior to the well starting. Training is provided so that procedures given the Well Site Leaders in outline or abbreviated form can be implemented according to BP s Drilling and Well Operations Practice (DWOP) manual, Well Control manuals, and other safety policies. This is a common industry practice. The drilling contractor is responsible for the operation of the drilling vessel and the work done by the rig crew. They have a responsibility for the safety of those aboard their vessel. They too are trained in well control and safety procedures appropriate for their level of job responsibility. 2 The driller and his rig crew are highly trained in recognizing the signs of a kick and closing the blowout preventers to stop the flow with the influx still near the well bottom. They are also trained in the use of a diverter to divert flow overboard and away from the rig when hydrocarbons enter the marine riser above the blowout preventer at the seafloor. 3 Two diverter lines are available to always allow downwind diversion. The chain of command from the Well Site Leader to the rig crew is through the drilling contractor s Offshore Installation Manager and Senior Toolpusher. 4 The Offshore Installation Manager and Senior Toolpusher have the responsibility to stop work when asked to perform an unsafe operation. 5 Any member of the crew can call for work to be stopped if an unsafe condition is seen. 2 Transocean s policies and procedures are consistent with this general industry standard. Examples include MDL Ex. 1454, Transocean s Well Control Handbook 1.3 at p. 2; MDL Ex. 1452, Transocean s Field Operations Handbook 3, 1.3 at pgs. 1-3; MDL Ex Transocean Deepwater Horizon Emergency Response Manual, Vol Many Transocean witnesses testified to these responsibilities as well, including Mr. Newman (S. Newman 9/30/11 Dep. Tr. at I am also familiar with the IADC Well Cap training that Transocean uses as part of its training program. MDL Ex. 1454, Transocean Well Control Handbook, 1.4. TRN-MDL Examples of Transocean s policy on this point include: (i) TRN-INV On page 811, it provides that Currently accredited IADC Well Control programs are taught in Transocean, Introductory, Fundamental, and Supervisory. All three provide instruction on how to use these [MGS and Diverter] systems for well control, (ii) MDL Ex. 1454, Transocean Well Control Handbook at 1.4, and (iii) the Driller and Assistant Driller On the Job Training Module there are tasks (Driller Task 45,46 & Assistant Driller Task 35, 36, 37) covering the diverter that must be completed and signed by a supervisor to be deemed competent. TRN-MDL (Driller Module) and TRN-MDL (Assistant Driller Module). Additionally, the Well Control Handbook states At any time, if there is a rapid expansion of gas in the riser, the diverter must be closed, if not already and the flow diverted overboard. MDL Ex at , TRN-MDL MDL Ex. 1453, Transocean s Deepwater Horizon Emergency Response Manual at TRN-MDL ; MDL Ex Transocean Management System - HSE Management at TRN-MDL Stop the Job authority is something that is common among rig personnel they should know that if something is unsafe, they stop the job and correct it. See MDL Ex. 1449, Transocean s HSE Policies and Procedures Manual at (footnote continued)

16 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 16 Mud logging services are generally provided by a service company specializing in rig monitoring and drilling data acquisition and transmission. Mud loggers monitor the entrained gas and rock cuttings that are circulated in the drilling mud to the surface as well as monitoring real time data. 6 Mud loggers are responsible for monitoring the well at all times, for being aware of the status of the well at all times, for recognizing the signs that formation fluids are entering the well and alerting the crew to shut-in the well. They provide a redundant level of eyes and ears for watching the well in addition to the driller and rig crew. Cementing services and equipment are generally provided by a service company specializing in oil and gas well cementing. This is a complex and highly technical service that requires a large organization to carry out. Cementing companies have a research group to develop improved proprietary cement compositions and placement techniques, highly trained engineers to properly design the cement job, trained technicians for performing the needed testing and quality control function, and trained field technicians for carrying out the cement placement into the well using their specialized equipment. Numerous other service companies provide specialty services regarding well logging, downhole equipment and subsea equipment. All of the service companies can have some responsibility in accidents caused by failure of their equipment. It is now common industry practice for oil companies, such as BP, to work together with specialized, expert contractors in order to successfully drill deepwater wells in the Gulf of Mexico. PORE PRESSURE Well design and casing depth placement is largely controlled by how formation pore pressure and fracture pressure vary with increasing depth. The depths of casing strings are initially established based on predicted pore pressure and fracture gradient values, which then are reevaluated as drilling operations are conducted and actual pore pressure measurements are established. If pre-drilling pore pressure predictions are off (which is common in the Gulf of Mexico), then the casing strings may be set deeper or shallower than originally planned and additional casing strings may be added to the well design. TRN-MDL Steve Newman also noted the stop the job authority for all personnel on the rig in his deposition. S. Newman 9/30/11 Dep. Tr. at I conducted independent review and analysis of much of the real time data from the rig, including the data monitored by the Sperry-Sun mudloggers.

17 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 17 The term pore pressure is used to refer to the pressure in the pore space of a rock. The sediments that we drill into in the Gulf of Mexico have been eroded from the North American Continent, carried to the sea by ancient rivers like the present Mississippi river, and laid down in the ancient gulf waters over geologic time. Pore pressure is easier to visualize if we think in terms of a sand bottom offshore from a sand beach (Figure 2a). The pore space is the space between the sand grains and the pore pressure is the pressure of the sea water in the pore space. Since water can flow freely between the grains in this situation, the magnitude of the pore pressure is controlled by the depth of the water column above the point of interest and the density of the seawater. The pressure gradient in a fluid in psi per foot can be obtained by multiplying the density in pounds per gallon by Seawater has a density of 8.5 pounds per gallon so that the pressure increases by 0.44 psi for each foot of depth. Note in Figure 2a that the pore pressure at points A, B, and C are all equal to 4.4 psi since all three points are at a depth of 10 feet. If we removed the sediments by drilling a hole at point B, pore fluid would fill the hole. If we refilled the hole with sand, the pressure in the pore space at the bottom of the hole would remain unchanged. This is because the pore fluid can move through the pore space and be in hydrostatic equilibrium with the ocean. The pore pressure is said to be normal when it is in hydrostatic equilibrium with the surface. The average pore water density in the Gulf of Mexico area is a little higher than the density of today s seawater because of additional dissolved minerals so that the normal formation pore pressure increases by psi per foot of depth. In south Louisiana, the prevalent formation type is sandstone down to a depth of approximately 10,000 feet and the sediments are nearly in hydrostatic equilibrium with the surface. Thus, the normal formation pore pressure gradient is psi/ft and the pore pressure expected in this area at 10,000 feet is 4,650 psi. The porosity of rock is the fraction of the total rock volume that is pore space. Permeability is a measure of how easily pore fluid can flow through the rock. A sandstone formation made from the deep burial of sand grains such that one would find on a beach would have a high porosity, typically on the order of 25%, and a high permeability. Such a rock would be an excellent reservoir rock if it contained oil and gas because it could contain a large oil volume that could flow easily into a well. Porosity decreases with depth of burial in normal pressure formations as pore water is squeezed out towards the surface.

18 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 18 Figure 2a: Pore pressure in hydrostatic equilibrium with surface water Figure 2b: Pore Pressure is controlled by Mud Pressure Figure 2: What is Pore Pressure?

19 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 19 When very fine grained clay is converted into shale rock by the elevated temperature and pressure of deep burial, it has an extremely low permeability such that pore fluid can flow only with extreme difficulty. Water is bound up in the crystalline structure of the clays as well as being present as free pore water. Shale can form a good seal so that when thick layers of shale are present, the hydrostatic equilibrium with the surface is interrupted. It becomes more difficult for pore water to escape as sediments are buried more deeply. In addition, more free water is released from the conversion of clay mineral structures from one form to another, e.g., conversion of montmorillonite clay to illites, chlorites, and kaolinite clays. The pore pressure increases and is said to be abnormal pore pressure. As one moves farther offshore into deep water, the top of abnormal pressure is encountered at more shallow depths. Figure 3: What is Mud Weight?

20 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 20 Pore pressure at any given vertical depth is often expressed as a pore pressure gradient by dividing the pore pressure by the depth. The pore pressure gradient is also commonly expressed as an equivalent density. The pore pressure gradient in pounds per gallon is obtained by dividing the pore pressure gradient in psi per foot by This tells the rig crew that a mud (Figure 3) having at least this density is needed to prevent pore fluids from flowing into the well. The mud weight at the surface is measured with a mud balance to the nearest 0.1 pounds per gallon. The mud weight in the well varies slightly with depth due to the effect of increasing temperature and pressure on the density of the synthetic oil and emulsified water in the mud. The annular mud weight can also be slightly affected by the rock cuttings and pore fluids being carried to the surface by the mud. OVERBURDEN STRESS Just as the density of a fluid determines the increase in pressure exerted by the fluid with increasing depth, the density of the sediments controls the vertical pressure or overburden stress within the sediments. This concept is more easily visualized if we go back to the example of beach sand. If one is buried under a foot of slightly submerged beach sand, the pressure that is felt is due to the combined density of the water and the sand grains. This average density, called the bulk density, is weight per unit volume of sediment with the pore water in the sediments. The bulk density of the sediments is not a constant but increases with depth because the grains rearrange and get closer together as they are buried deeper and subject to higher overburden stress. Water is squeezed out towards the surface as the porosity decreases. Overburden stress cannot be measured directly and must be calculated from sediment density values provided using well logging tools. The well logging tools do not make a direct measurement of sediment density. Density values are computed from measurements such as sonic travel time through the rock and neutron absorption. Figure 4 is a plot of bulk density and overburden stress versus depth estimated from well logs run in the Macondo Well. Since the rock grains or matrix tend to interlock and resist moving in a direction perpendicular to the overburden stress, the horizontal stress is usually less than the overburden stress in a tectonically relaxed region like the Gulf of Mexico. The horizontal stress can vary with horizontal directions and is expected to have a minimum value in a direction perpendicular to local fault lines. However, near salt domes or other intrusions, the horizontal stress can become higher than the overburden stress.

21 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 21 Drilling a borehole through the sediments can create local stress concentrations near the borehole wall. After the borehole is drilled, the stress in the sediments close to the borehole increases to accommodate load that was previously borne by material removed. Figure 4: Overburden Stress calculated from Macondo Well Logs FRACTURE GRADIENT If the pressure in the borehole is increased significantly above the minimum horizontal stress so that rock tensile strength and local stress concentrations are exceeded, the rock of the borehole wall will crack vertically along the wall parallel to the axis of the hole. The crack will begin to open (Figure 5) and mud will enter the crack. The pressure at which the crack begins to open and take mud is called the fracture pressure. The fracture gradient in units of psi per foot is the fracture pressure in psi divided by the vertical depth of the fracture below the rig floor in feet. The fracture gradient can be expressed in pound per gallon by dividing by This tells the drill crew that a static mud having this density or higher could fracture the formation.

22 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 22 Figure 5: What is Fracture Gradient There are a number of correlations that have been developed for estimating the formation fracture gradient. However, because of the wide variability found in nature, there is often considerable disagreement between the predicted and observed fracture gradients. There is no foolproof way of determining the fracture pressure of a section of hole. The best available method is to intentionally create a small fracture at the top of section and measure the pressure required to do so. This operation is called a Leak-off Test (LOT). Figure 6 is a comparison of predicted fracture pressure and Leak-off test results made by one my students using Leak-off test data and presented at the 1994 LSU/MMS Workshop. 7 If the measured Leak-off test value for 7 Rocha, L.A. and Bourgoyne, A.T., Session II, Presentation 8, LSU/MMS Well Control Workshop, March 30-31, 1994.

23 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 23 fracture gradient, shown in Figure 6 as a point, agrees with the calculated value of fracture gradient, shown as a dashed line, the point would fall on the line. Note that the measured values for fracture gradient often disagreed with the calculated value by as much as two pounds per gallon. Three different published methods for calculating the fracture gradient were used in the comparison. The Leak-off test data was from the Green Canyon Area of the Gulf of Mexico. Figure 6: Comparison of Calculated and Observed Leak-off Test Results In some cases, the formation fracture pressure is higher than needed to finish the next section of hole and the pressure integrity test is stopped before creating a small fracture. The Pressure integrity test does not always go all the way to Leak-off. PRESSURE INTEGRITY TESTS Pressure integrity tests provide information about the effectiveness of the cement seal at the bottom of the casing and about the resistance of the formation just below the casing to hydraulic fracturing. This is generally done after setting each casing string. Leakage past the cement is indicated by a non-linear, slower than expected pressure build-up that does not reach the minimum expected pressure. If leakage past the cement outside of the casing is detected, cement is squeezed into the leak and allowed to harden. The cement is again drilled out and the Pressure Integrity Test is repeated. This sequence is repeated until leakage past the cement is no longer seen. Running multiple pressure integrity tests is common. These tests involve two steps. First, after the cement has hardened, a casing pressure test is made to insure that there are no leaks in the casing. A recorded plot of pressure versus time is made during the casing pressure test. The blowout preventer is closed and mud is pumped into the well at a low pump rate (typically 0.25 or 0.5 barrels per minute) in order to impose the desired pressure on the casing. After reaching the desired test pressure, the pump is stopped and pressure continues to be recorded for at least 30 minutes. The pressure during the shut-in can change slightly due to temperature changes in the mud. MMS (now BOEMRE) limits the

24 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 24 observed decline in pressure to no more than 10% during a 30 minute period. The onsite operator s representative signs and dates the casing pressure test as being correct. The driller s report includes the time, date, and results of the test. Second, following a successful casing pressure test, the hardened cement is drilled out of the bottom portion of the casing and then a short interval of formation (typically 5 to 10 feet onshore and 10 to 50 feet offshore) is drilled below the casing. After circulating cuttings out of the well, the blowout preventer is closed and mud is pumped into the well at a low pump rate (typically 0.25 or 0.5 barrels per minute). BP s procedure calls for pumping down both the drill pipe and the casing to minimize the effect of the frictional pressure loss between the surface and the casing seat. 8 This also reduces the effect of mud gelation when starting and ending the test. The pump pressure and volume pumped is recorded at even time increments (typically after each minute of pumping). If the cement seal around the bottom of the casing is good, a straight line trend of increasing pressure with volume pumped is seen. If the pressure reaches a value corresponding to the formation strength needed to drill the next section of hole, the test may be terminated prior to creating a small fracture. If the fracture initiation pressure is to be determined, the test is continued until two or three points fall significantly to the right of the straight line trend indicating that mud is beginning to enter a small fracture. BP s procedure calls for continuing a Leak-off test until the same pressure or a lessor pressure is seen at the previous value recorded. BP s procedure is consistent with industry practices and my understanding is that BP informed the MMS (now BOEMRE) of their testing procedures in a meeting in February Upon fracture initiation, the rock tends to crack vertically, parallel to the axis of the borehole. The borehole acts as a notch in the rock from which the fracture grows. Fracture resistance is determined by creating a small vertical fracture and then stopping the test. Fracture growth is limited so that it does not extend upward past the casing seat into weaker rock. When the pressure is allowed to bleed off, the small fracture will close. Usually the rock will not be weakened significantly because natural cracks and imperfections are already present and only have to be opened by the mud pressure. Once it is decided that the pressure integrity test will be terminated, the pump is stopped and pressure readings are then taken at a regular time interval. BP s procedure calls for recording the 8 MDL Ex FIT/LOT Worksheet and instructions. 9 MDL Ex. 4734; MDL Ex. 4735; MDL Ex. 4736; M. Saucier 7/27/11 Dep. Tr. at ; D. Trocquet 9/23/11 Dep. Tr. at

25 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 25 shut-in pressures for a minimum of 10 minutes if performing a Formation Integrity Test (FIT). 10 The rate of decline in pump pressure is affected by the permeability of the zone exposed, by the size of the fracture created (if any), and by the rate of increase in mud gel strength. If the test is conducted with only low permeability shale exposed below the casing, the Leak-off rate after the pump is stopped will be much slower that if sand is exposed, and will be negligible if a fracture was not initiated. It is common practice to record the volume of mud bled back at the end of the test when the pressure is released. This volume can then be compared to the volume pumped during the test to estimate the volume leaked-off during the test. A graph of pump pressure versus volume pumped is generally prepared to assist in the interpretation of a Leak-off test. The casing pressure test is often plotted as a reference for system compressibility expected. This helps in interpreting the test in regard to whether or not the cement seal at the bottom of the casing is leaking. Typical pressure test results to be expected are shown in Figure 7 for a Leak-off test (LOT) and Figure 8 for a Formation Integrity Test (FIT) not taken to Leak-off. Two cases are illustrated in Figure 7. The upper plot is for the case in which a smooth borehole is drilled into impermeable shale that does not have any natural cracks or imperfections and has a significant tensile strength. Rocks are generally weak in tension but previously unbroken rock often has a tensile strength of about 5% of it compressive strength. When interpreting a Leak-off test, the first few data points after pumping begins can fall off trend and are ignored because any air trapped in the surface piping needs to be compressed sufficiently before it has a negligible effect on the trend line. This can cause the few points to fall to the left of the subsequent straight line trend. After sufficient air compression and before initiation of leakage into a formation fracture or past the cement sheath around the bottom of the casing, the points will fall approximately on a straight line. After the leak begins, the pressure vs. volume pumped plot will begin to depart from a linear relationship, i.e., the points will begin to fall to the right of the straight line trend. The leak may be past the cement sheath around the bottom of the casing or into a fracture that is being initiated. After the fracture begins to grow, the pump pressure will reach a maximum value and in some cases will begin decreasing with increased volume pumped as the fracture is extended away from the borehole. 10 Although nomenclature varies somewhat in industry practice, an FIT in this context is taken to be a Pressure Integrity Test (PIT) that does not go all the way to Leak-off.

26 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 26 Figure 7: Typical Leak-Off Test Plot The maximum pressure reached is called the breakdown pressure when the shape of the test is similar to upper plot of Figure 7. This is also done by some operators when the shape of the test is similar to the lower plot of Figure 7. Industry practice varies on whether the leak initiation pressure or the breakdown pressure measured at a low pump rate is used to compute the formation integrity expressed as a maximum equivalent mud weight. The maximum pressure seen when the shape of the test is similar to the lower plot of Figure 7 can be sensitive to how fast the mud is being pumped during the test. The leak initiation pressure is not as sensitive to the pump rate used during the test. The computation of fracture gradient using either the fracture initiation pressure or a breakdown pressures measured at a low pump rate clearly falls within normal industry practice. To my knowledge, MMS does not take a position on what constitutes the best available technology for running and interpreting a Pressure Integrity Test and will accept either approach described above MDL Ex. 4731; MDL Ex. 4023; M. Saucier 7/27/11 Dep. Tr. at ; F. Patton 7/13/11 Dep. Tr. at

27 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 27. Figure 8: Typical Pressure Integrity Test Plot Points in the table are most often obtained every ¼ barrel while pumping at ¼ barrels per minute. However, when the hole-volume is large the test time required becomes unnecessarily large and points are obtained every ½ barrel while pumping at ½ barrels per minute or every barrel while pumping at one barrel per minute. Pump rates higher than one barrel per minute during a Leak-off test are unusual. When the points are too far apart, it becomes more difficult to determine the pressure at which the departure from linearity occurred. A common practice is to draw a straight line through the points on the linear trend and also draw a straight line though the first two or three points that fall to the right of the linear trend. The leak initiation pressure is then determined from the intersection point of these two lines. When pumping is stopped after pumping only a small volume into the fracture, the fracture will close quickly and the Initial Shut-In Pressure (ISIP) will be only slightly higher than the pressure to re-initiate the fracture. This will often yield the same results as the point of departure from the straight-line trend. Also, with the pump off, there is no frictional pressure loss down the drill string. The interpretation of leak of tests is not an exact science and I have observed significant differences in interpretations among practicing professionals attending my well control class. During the mid-1990 s, my graduate students collected Leak-off test data from a number of offshore operators in an effort to develop improved fracture gradient prediction methods. It was clear that there was variability in the selection of the formation breakdown pressure from the Leak-off test results. Data from softer, less consolidated formations are sometimes more

28 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 28 difficult to interpret. Experience in some areas has shown that small leakage rates can be tolerated or controlled with loss circulation material. Dr. Alan Huffman has concluded that in drilling the Macondo Well, BP departed from the standards that would be met by any prudent operator. This was in part based on his analysis of the pressure integrity tests conducted at the bottom of each casing or liner. In some cases Dr. Huffman focuses on reporting errors made in submissions to MMS (now BOEMRE) and in other cases he questions the validity of the tests conducted. As discussed below, I disagree with Dr. Huffman s assertions. Figure 9: Approximate effect of Water Depth on Fracture Gradient at 3500 feet DRILLING MARGINS For conventional drilling in deepwater, there is a narrow margin between the mud weight needed to control pore pressure and the mud weight that would cause an exposed formation to fracture. Figure 9 is a comparison of fracture gradients below casing set with a sediment penetration of 3500 ft. Note the drastic decrease in fracture gradient for a casing string cemented 3500 feet into the sediments as water depths increase. The narrower margin requires the use of more casing strings than for the same well in shallow water. This occurs because the open hole must support a column of mud that is heavier than seawater and that extends far above the sea floor to the floating drilling vessel at the surface.

29 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 29 Every operator in deepwater wells encounters narrow drilling margins. The way to manage narrow drilling margins is to carefully monitor the well at all times and set more casing strings. The Macondo well was not unusual in that regard. Figure 10: What is a Kick? As shown previously in Figure 2b, the mud weight should be sufficient to maintain the pressure within a borehole at any given depth above the pore pressure of a permeable formation exposed to the borehole at that depth. If this is not done, fluids from the rock pores can flow into the well. An influx of formation fluid into the well is called a kick (Figure 10) because in an extreme case it can kick the mud out of the well onto the rig floor. The rig crews are trained to recognize a kick early when signs are much less obvious and to close the blowout preventer. Formation fluids entering the well cause the mud to flow out of the well faster than it is being

30 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 30 pumped and cause the fluid level in the pits to rise. Pit drills, in which a false indication of a pit level increase is created, are routinely conducted to test the response of the crew. 12 Kicks are not desired but happen on many wells drilled in the Gulf of Mexico because of the presence of abnormal pressure. Mud logging service companies such as Sperry-Sun display a number of indicators that show when the pore pressure could be increasing. None of this technology is 100% accurate, so occasionally a well will become underbalanced and formation fluids will enter the well. Normally the influx of fluid is detected quickly from either a pit level increase or an increase in the flow returning from the well and the well is shut-in with only a small pit gain. The required mud weight (kill mud weight) is calculated from the observed shutin drill pipe pressure. The mud service company prepares the kill mud and it is circulated around with the blowout preventers closed and flow from the well taken through an adjustable choke. Formation fluid is circulated out of the well and kill mud is circulated to the surface. After it is verified that the well is dead, normal operations can be resumed. Redundant well control systems and good training practices have been established for many years. Such an event can be handled in a routine manner. When pulling the drill pipe out of the well to trip for a new bit, the pressure at the bottom of the well decreases slightly below that of a static well. Thus, the mud weight must be maintained slightly above the pore pressure gradient in order to prevent a kick. The additional mud weight maintained over the pore pressure gradient is called the trip margin. The mud weight should also be less than the fracture gradient at the bottom of the casing, which is referred to as the casing seat or the casing shoe or sometimes just shoe. As shown in Figure 11, it is important to prevent the development of a very large fracture that could grow upward outside of casing into weaker sediments previously protected by the casing. This can cause an inability to keep the hole full of mud and lead to an underground blowout, (Figure 12). Underground blowouts are very expensive to control and usually require sealing off and losing at least the bottom portion of the well. In some cases, the entire well is lost. If the casing penetration is not very deep (less than about 2500 feet) and there are no shallow sands or shell beds to take the flow, the sediments can be broached all the way to the surface and cause a large crater to develop. This is dangerous for bottom supported rigs, since they are not easily moved. 12 Transocean s procedures require pit drills to be conducted weekly, along with other well control drills on the rig. See MDL Ex. 1454, Transocean Well Control Handbook at TRN-MDL

31 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 31 When circulating the mud, or when running drill pipe or casing into the well, the pressure at the bottom of the well increases slightly above that of a static well. Thus the mud weight should be maintained below the fracture gradient to prevent loss of mud to a formation fracture. The rig crew is trained to quickly detect when mud is being lost and to take appropriate measures to stop the loss. There have been numerous advancements in the design and use of loss circulation material to heal the fracture so that normal operations can be resumed. Figure 11: Shoe Strength protects Weaker Sediments Above

32 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 32 Figure 12: What is an Underground Blowout? When a kick is taken and the blowout preventers are closed, the surface drill pipe pressure increases by the amount that the pore pressure exceeds the mud pressure in the well. The casing pressure increases higher than the drill pipe pressure by an amount equal to the difference in hydrostatic pressure of the mud in the drill pipe and the kick contaminated mud in the annulus. When the kick is detected quickly and the kick volume is small, the casing pressure is only slightly higher than the drill pipe. The shut-in pressure causes the downhole pressures to be higher than static downhole pressure when no kick is present. A higher mud weight must be circulated into the well at a low pump rate and under pressure with the blowout preventers closed before normal operations can be resumed.

33 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 33 SAFE DRILLING MARGIN The initial drilling plan must be made before starting the well. For the first well drilled in a new prospect area, prediction of how formation pore pressure and fracture gradient will vary with depth is especially difficult and is usually not accurate. This is especially true when the normal trend of increasing pore pressure and fracture gradient with increasing depth is not present. The bottom part of the Macondo well had a reversal of the normal trend in which both formation pore pressure and fracture gradient decreased with increasing depth. 13 Changes in the casing program based on information learned while drilling is a common occurrence. The estimation of pore pressure and fracture gradient becomes more accurate after data from previously drilled nearby wells are available. For example, BP was able to use data obtained drilling the original Macondo well to drill a nearby relief well with no delays. 14 When planning a section of hole in a well below casing, it is clearly desirable for the operator to plan to use a mud weight below the fracture gradient to allow for circulating, running pipe, or taking an unexpected kick. Since the fracture gradient is known at the casing seat and fracture gradient normally increase with depth, normal practice is to base the planned margin between fracture pressure and mud weight on the Pressure Integrity Test results. The reference to safe drilling margin is in 30 C.F.R covering the requirements for pressure integrity tests. The fracture gradient included in the Application for Permit to Drill (APD) provided BOEMRE (previously MMS) must be included in a single plot containing estimated pore pressures, formation fracture gradients, proposed drilling fluid weights, and casing setting depths in true vertical measurements. The regulations do not specify how the estimate is made or what safety margins to use. Approval of the plan is at the discretion of the District Manager of MMS based on the information presented. Once the plan is approved, and drilling commences, the plan in the APD defines the approved safety margin. For example, if the planned maximum mud weight is 0.5 pounds per gallon less that the estimated fracture pressure for the hole section, then the approved safety margin is 0.5 pounds per gallon. It is recognized that the approved APD is based on estimates and that data obtained while drilling must be recorded in the daily drilling reports as the well is being drilled. The regulations state: 13 MDL Ex. 1 BP s Deepwater Horizon Accident Investigation Report, September 8, 2010 at 16-17; MDL Ex. 986 Chief Counsel s 2011 National Commission on the BP Deepwater Horizon Oil Spill and Offshore Drilling Report at 59-61; MWD/LWD real time data recorded on the rig and cited in Appendix D. 14 Daily Drilling Reports for the relief wells are identified in Appendix D.

34 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 34 You must use the pressure integrity test and related hole-behavior observations, such as porepressure test results, gas-cut drilling fluid, and well kicks to adjust the drilling fluid program and the setting depth of the next casing string. You must record all test results and hole-behavior observations made during the course of drilling related to formation integrity and pore pressure in the driller's report. 15 This recognizes that mud weights must be allowed to deviate from the APD based on what is seen while drilling. MMS does not specify how to interpret the information that you must record, but they want the ability to be able to inspect the available information. In my opinion, it is extremely important for the team on the rig to be able to change mud weights when needed for proper well control and safety without having to wait for approval. Dr. Alan Huffman expresses the opinion that if mud is lost while drilling, this event is equivalent to an open-hole pressure integrity test and requires immediate recalculation of the drilling margin. He states that regulations require that if the mud weight cannot be safely reduced within the approved safe drilling margin (as re-calculated based on loss return incident), then the operator must cease drilling immediately and get approval from BOEMRE to drill ahead with a reduced drilling margin or to set casing immediately. He further indicated that BOEMRE would not allow drilling to continue if the recalculated drilling margin was less than 0.3 pounds per gallon and would require casing to be set immediately. I do not agree with any of this for the following reasons: 16 In deepwater drilling, the margin between pore pressure and fracture gradient is often so small that permeable formations encountered while drilling will initially fracture at mud weights significantly lower than the shoe strength. The fracture resistance of these zones can usually be raised to that of the confining shale by treatment with modern loss circulation material. When the use of loss-circulation-material allows full circulation to be regained, the mud weight that the well will hold can often be gradually increased. There is no way to calculate what the final fracture gradient will be after multiple treatments. Geo-Tap tools can measure pore pressure but do not measure fracture pressure. BP has specialists for calculating fracture gradient from other observations, but these are just estimated and are often inaccurate. The drilling team leader uses this group to provide input, but he must weigh their input against other potential problems and operational risks C.F.R

35 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 35 In my opinion, normal drilling practice is to drill through a sandstone interval whenever circulation can be established, even if some mud losses continue. Setting casing in the middle of a sandstone formation would be extremely unwise because the weak zone would still be exposed and it would be much more difficult to get a good cement job on the casing. Furthermore, the formation pore pressure gradient within a permeable formation cannot increase with additional penetration to the bottom of the sandstone formation. The mud weight in use will be sufficient to control formation pore pressure. Care will have to be taken when pulling pipe out of the hole, but again, there are standard drilling practices for not allowing the bottom-hole pressure to fall when pulling pipe out of the hole. This is especially true on modern rigs which allow mud to be pumped when pulling pipe. Efforts to heal the loss circulation zone would continue while drilling through the permeable zone and there would be no basis for accurate re-determination of the final fracture gradient. When the use of loss-circulation-material allows full circulation to be regained, the mud weight that the well will hold can often be gradually increased. I would expect the District Manager of BOEMRE (previously MMS) and experienced field inspectors to recognize normal practice in such situations. If regulation enforcement was a simple enforcement of pre-determined drilling margins as indicated by Dr. Huffman, then this would be written into the regulations. There would be a many more issuances of non-compliance made in the Gulf of Mexico region if every case of drilling ahead with hole-ballooning seen in a field inspection resulted in one being issued. If casing had to be run every time loss circulation or hole-ballooning was experienced, there would be very few wells successfully drilled in the Gulf of Mexico area. Dr. Huffman found examples of reporting errors made in documents submitted to BOEMRE (previously MMS). In my opinion, the few errors found in the large amount of material submitted to BOEMRE (previously MMS) do not warrant the leap Dr. Huffman makes to a conclusion that BP intentionally hid information so that they could operate in an unsafe and imprudent manner. 16 Transocean recognized many of these aspects of deepwater drilling as well, as reflected in their Drilling Deepwater Wells powerpoint presentation. TRN-MDL

36 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 36 Dr. Huffman also based his conclusions regarding purported imprudent practices in part on BP not reporting Pressure Integrity Test results to the nearest hundredth of a pound per gallon. In my opinion, normal engineering practice is not to exceed expected accuracy when reporting test results. Mud weights are measured to the nearest tenth of a pound per gallon. BP used a Pressure While Drilling (PWD) tool above the drill bit on the Macondo well. This tool is not required and the use of this tool is an indication of using the best available technology and incurring extra costs to improve safety. The accuracy of a very good pressure gauge seldom is better than one percent and it is difficult to rely on a PWD tool to be accurate to a hundredth of a pound per gallon when measuring downhole conditions. In my opinion, it is normal practice to record mud weights and equivalent mud weights to the nearest tenth of a pound per gallon. Also, it is my understanding that the MMS s (now BOEMRE) electronic ewells system requires data to be reported in tenths. 17 Dr. Huffman expressed an opinion that the pressure integrity tests that were conducted below the 13-5/8 inch liner and the 9-7/8 inch liner were not valid tests. I was asked to do independent analyses of these pressure integrity tests and to provide an opinion as to whether or not these tests were valid. Both tests were valid pressure integrity tests. BP S PRESSURE INTEGRITY TEST PROCEDURE BP s written procedure for conducting a Pressure Integrity Test is as follows: PRE-JOB CONSIDERATIONS: Graph casing pressure test versus volume pumped and use as a baseline for the PIT test. Use cement unit and cement unit gauges. Ensure sufficient mud supply to the cement unit. Calculate estimated Leak-off Test (LOT)/Formation Integrity Test (FIT) pressure and test lines to 1000 psi over expected LOT/FIT pressure. OPERATIONS 1. Drill out the cement shoe track, cleanout rat-hole, and drill 10-ft of new formation or as otherwise specified in well specific program Technical Note: MMS regulations require a minimum of 10-ft md and a maximum of 50-ft md of new formation. 17 F. Patton 7/13/11 Dep. Tr. at 271:2-6 ( Q. Mr. Patton, the ewell submission, the computer submission, you can only submit data in tenths, correct? A. That is correct, yes. )

37 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE Circulate cuttings out of the well or at least above the BOPs. If minimal cuttings have been drilled, circulate a minimum of 1,000-ft above the Bottomhole Assembly. Ensure consistent mud weight with surface samples with pressurized scales. 3. Pull bit back into shoe and space out to close the appropriate BOPs. 4. While keeping drill string still, circulate up consistent downhole equivalent static densities (ESDs) from the pressure while drilling (PWD) tool to achieve < 0.05 pound per gallon consistency. 5. Displace the appropriate choke or kill line to fresh mud. Technical Note: The test will be performed pumping down the drill pipe and down the annulus simultaneously via the appropriate choke or kill line to reduce friction pressure. 6. Rig up to pump down the drill pipe and down choke or kill line with the cement unit. 7. Break circulation down the drill pipe and down the choke or kill line. 8. Close the appropriate surface valve on the drill pipe side, and the appropriate choke/kill line valve at the BOP stack. Test lines to 1,000 psi over anticipated maximum LOT or FIT surface pressure. 9. Bleed off test pressure but do not completely drain lines. Open valves and break circulation down drill pipe and down the choke or kill line. 10. Shutdown and re-zero pressure gauge at cement unit to account for hydrostatic between cement unit and rig floor. 11. Close appropriate BOP and monitor return line to ensure no returns during PIT. 12. Perform LOT/FIT, pumping a maximum of ½ barrel per minute (bpm). 13. Record volume pumped and surface pressure consistent with pump rate. For example, if pumping ½ bpm, record data every ½ bbl. If pumping ¼ bpm, record data every ¼ bbl. 14. Graph data on a Surface Pressure versus Volume Pumped/Time graph. Technical Note: Note 1: If performing an FIT, once desired surface pressure is achieved, shutdown and monitor pressure for a minimum of 10 minutes. Note 2: If performing a LOT, continue pumping until the pressure flatten or decreases (pump until the subsequent pressure is equal or less than the last pressure). This is the shut-down point.

38 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 38 BP S PRESSURE INTEGRITY TEST OF THE 13-5/8 INCH LINER SHOE The Pressure Integrity Test on the 13-5/8 inch liner shoe was conducted on March 21, 2010 with R. Sepulvado listed the Well Site Leader supervising the test. 18 The measured depth of the hole was 13,150 feet, which corresponded to a True Vertical Depth of 13,140 feet. The bottom of the liner was at a measured depth of 13,145 feet which corresponded to a True Vertical Depth of 13,135 feet. The well was vertical on bottom (inclination angle of zero). The top cement plug was at the float collar at 01:30 hours on March 21, The casing was pressure tested at about 17:00 hrs. The volume of 12.5 pound per gallon mud pumped during the casing test to achieve a test pressure of 2415 psi was 14.0 barrels and the data recorded during the pressure build-up was as follows: Casing Test Volume Pumped (bbl) Pump Pressure (psi) The data underlying this discussion comes from Daily Operations Reports for March 20 and 21, 2010 (BP-HZN- MBI ); IADC Drilling Report for March 20 and 21, 2010 (BP-HZN-2179MDL , ); Daily Geological Reports for March 20 and 21, 2010 (BP-HZN-2179MDL , 16-18), and MDL Ex FIT/LOT Worksheet and the realtime downhole MWD/PWD data referenced in Appendix D.

39 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE The pressure after 30 minutes was 2370 psi, which was about a 2% decline in 30 minutes and thus much less than the 10% allowed for a successful test. The volume bled back after the test was 14 barrels. By 10:30 hours the next day, the drilling assembly had been run to bottom and cement was drilled out from 13,100 feet to 13,150 feet. It was reported that the bit exited the shoe at 13,145 feet and cement was drilled in the rat hole from 13,145 feet to 13,150 feet. Ten feet of new formation was drilled from 13,150 feet to 13,160 feet (measured depth). The cuttings were pumped from bottom to the wellhead. A static bottom bole pressure equivalent to pounds per gallon was reported with a surface mud weight of 12.5 pounds per gallon when circulating at a measure depth of 13,140 feet. The Pressure Integrity Test was performed at about 13:00 hours using a pump rate of 0.5 barrels per minute. The midpoint of the new formation drilled was a measured depth of 13,155 feet and a true vertical depth of 13,145 feet. The following data was recorded: Pressure Integrity Test Shut-in Data Pump Volume Pump Time (min) Pressure Pumped Pressure (min) (psi) (bbl) (psi)

40 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE The volume of mud bled back at the end of the test was 9.25 barrels. Note that 9.5 barrels was pumped into the well when increasing the pressure. This indicates that 0.25 barrels of mud and mud filtrate leaked into the formation during the test. The results of my analysis are shown graphically in Figure 13. Basing the fracture gradient on the observed fracture initiation pressure of 1480 psi, the fracture gradient is calculated to be equivalent to a surface mud weight of 14.7 pounds per gallon. On this test, the fracture initiation pressure was the breakdown pressure. This is consistent with the 10 feet of formation being exposed consisting of an impermeable smooth borehole with no pre-existing defects or cracks and significant tensile strength and stress concentration near the borehole wall. The fracture pressure at 13,145 feet calculated using the equivalent mud weight on bottom as measured by the Pressure While Drilling (PWD) tool was 10,175 psi. The overburden stress at a measured depth of 13,155 feet (true vertical depth of 13,145 feet), due to the weight of the sediments above, was calculated from well logs to be 9,632 psi. Thus the Leak-off test gave a value for fracture pressure that was about 543 psi higher than the calculated overburden stress. This is a reasonable value for the tensile strength for impermeable rock that was not previously fractured. The shut-in pressure versus time plot is indicative of a fracture that closed slowly over a two minute period after pumping was stopped. The fracture closure pressure appears to be about 1,200 psi, which is equivalent to a surface mud weight equivalent to 14.3 pounds per gallon. However, fracture closure pressure generally cannot be accurately measured in mud because the mud starts to gel as soon as pumping is stopped. Dr. Huffman indicated that Pressure Integrity Test was a bad test because of the following problems : The slope of the pressure build-up curve for the Pressure Integrity Test was very close to the slope of the casing pressure test build-up curve.

41 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 41 The leak-off pressure and the propagation pressure was nearly the same. An initial shut-in pressure that is lower than Dr. Huffman would expect relative to the apparent leak-off pressure. An anomalously steep decline in pressure during the shut-in period. The fracture gradient indicated by the test was higher than expected and higher than the overburden stress computed from well logs. Figure 13: Pressure Integrity Test at 13-5/8 inch Liner Shoe I do not agree that the above observations indicate that the test was bad. I find the test results to be consistent with 10 feet of borehole drilled below casing into low permeability rock that was not previously fractured. A point by point response to Dr. Huffman s list follows. The volume being compressed for the casing and leak-off tests are almost the same. Adding 60 more feet to the 13,100 feet of well depth does not add much to the volume that has to be compressed. The length of drill string in the hole was less for the casing test. Thus, for low permeability rock, the slope would be expected to be nearly the same.

42 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 42 I do not find it strange for the breakdown and propagation pressure to be the same. Sometimes the breakdown is very sharp and the first point after breakdown is lower than the breakdown pressure as shown in Figure 7. In this case we have only one point and not a continuous curve. The Initial Shut in Pressure falls 75 psi below the final pumping pressure on the casing test and 64 psi on the Leak-off test. I find that this is consistent with opening a very small fracture and leaking only 0.25 barrels of mud and mud filtrate into the fracture. 19 I did not find the decline in pressure after shut-in to be anomalous. I find it to be consistent with a small fracture not penetrating all the way through the region of higher stress concentration surrounding a borehole drilled through rock not previously fractured. It is true that the test result was higher than expected and was likely not to be representative of the entire hole-section. However, it did indicate that the shoe strength was strong enough to proceed with drilling the next section, which was the main purpose of the test. BP eventually adjusted the fracture gradient curves to calculated values more representative of the hole section. BP S PRESSURE INTEGRITY TEST OF THE 9-7/8 INCH LINER SHOE The Pressure Integrity Test on the 9-7/8 inch liner shoe was conducted on April 2, 2010 with Lee and Price listed as the Well Site Leaders supervising the test. 20 The measured depth of the hole was 17,183 feet, which corresponded to a True Vertical Depth of 17,173 feet. The bottom of the liner was at a measured depth of 17,168 feet which corresponded to a True Vertical Depth of 17,158 feet. The well was nearly vertical on bottom (inclination angle of 0.5 degrees). The cement was in place at 11:00 hours on March 31, The casing was pressure tested at about 13:00 hours on April 1, The volume of 14.1 pound per gallon mud pumped during the 19 The volume pumped into the well was 9.5 barrels and the volume bleed back after the test was 9.25 barrels. 20 The data for this discussion comes from the Daily Operations Report for April 2, 2010 (BP-HZN-MBI ), the IADC Drilling Report for April 2, 2010 (BP-HZN-2179MDL ), Daily Geological Reports for April 2, 2010 (BP-HZN-2179MDL ), MDL Ex FIT/LOT worksheet, and the real time downhole MWD/PWD data referenced in Appendix D.

43 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 43 casing test to achieve a test pressure of 914 psi was 5.75 barrels and the data recorded during the pressure build-up was as follows: Casing Test Volume Pumped (bbl) Pump Pressure (psi) The pressure after 30 minutes was 887 psi, which was about a 3% decline in 30 minutes and thus much less than the 10% allowed for a successful test. The volume of mud bled back after the test was 5.5 bbl. By 05:00 hours on April 2, 2010, the drilling assembly had been run to bottom and cement was drilled out from 17,000 feet to 17,173 feet. Ten feet of new formation was drilled from 17,173 feet to 17,183 feet (measured depth). The cuttings were pumped from bottom to above the blowout preventers. A static bottom bole pressure equivalent to pounds per gallon was reported with a surface mud weight of 14.3 pounds per gallon when circulating at a measured depth of 17,183 feet. The Pressure Integrity Test was performed at about 09:00 hours using a pump rate of 0.5 barrels per minute at the liner shoe at 17,168 feet (measured depth). The midpoint of the new formation drilled was a measured depth of 17,178 feet and a true vertical depth of 17,173 feet. The following data was recorded: Pressure Integrity Test Shut-in Data Pump Volume Pump Time (min) Pressure Pumped Pressure (min) (psi) (bbl) (psi)

44 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 44 Pressure Integrity Test Shut-in Data The volume of mud bled back at the end of the test was 10.0 barrels. Note that 10.0 barrels was pumped into the well when increasing the pressure. This test did not go to Leak-off. The equivalent mud weight on bottom as measured by the Pressure While Drilling (PWD) tool was reported to be equivalent to pounds per gallon. The results of my analysis are shown graphically in Figure 14. Using the pressure of 1,520 psi when the test was terminated, the shoe strength is calculated to be equivalent to a surface mud weight of 16.0 pounds per gallon. The overburden stress at a true vertical depth of 17,173 feet, due to the weight of the sediments above, was calculated from well logs to be 13,772 psi. The pressure measured with the PWD tool was pounds per gallon or 14,470 psi. Thus the Formation Integrity Test gave a strength value that was 698 psi above the calculated overburden stress. This is a reasonable number for 10 feet of formation being exposed consisting of an impermeable smooth borehole with no pre-existing defects or cracks and significant tensile strength and stress concentration near the borehole wall. Dr. Huffman has suggested that this was not a valid test because of its high value and BP was not a prudent operator because they did not continue testing. I saw no evidence that this was true. There is often a wide variation in shoe test results when comparing results obtained on different wells and when comparing observed results to calculated results. Calculation results

45 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 45 also vary depending on the correlation used in the calculation. An example of variations on the order of two pounds per gallon was shown previously in Figure 6. Figure 14: Pressure Integrity Test at 9-7/8 inch Liner Shoe Dr. Huffman pointed out that the slope of the buildup curve was nearly the same as the casing test and compared the test to the test at the 16 inch shoe which had more difference in slope. The decrease in slope of the straight line buildup curve (between the casing test and the formation test in a deep well with 10 feet of open borehole) depends primarily on the rate at which mud filtrate is being lost during the test. If the 10 feet of borehole is good shale with a very low permeability, then one would expect the slopes to be nearly the same. The permeability of shale is in the nanodarcy range, 21 which is about a million times less permeable than 21 Discussion of shale permeability in Shale Water as a Pressure Support Mechanism in Gas Reservoirs having Abnormal Formation Pressure, Journal of Petroleum Science and Engineering, 3 (1990)

46 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 46 sandstone, and would be expected to behave much like steel pipe from the standpoint of allowing fluid to enter the borehole wall when no cracks are present. Dr. Huffman pointed out internal s within the BP organization discussing the high shoe strength. Various reasons were suggested including the possibility that the driller did not know what depth he was at when he drilled out the shoe. 22 This issue was brought up in regard to both tests that produced unexpectedly high fracture gradients. BP s team members that supervised the test recognized that the results were higher than expected and BP even repeated a portion of the test. 23 It is not plausible that they would not have checked for possible pipe tally errors and that the driller was off by more than 10 feet. Another speculation is that not all of cement in the five feet of rat hole below the casing had been drilled out. This speculation was partially based on a report of abundant cement mixed with shale cuttings when drilling new formation below the 13-5/8 inch shoe. When cementing with a short rat hole below the casing, the cement is not expected to displace all of the mud out of the rat hole, but sometimes good cement fills the entire rat hole. Good cement in the entire rat hole results in more cement cuttings being circulated to the surface when drilling ahead. I have examined the daily operational reports, cuttings evaluations, penetration rate data, and gamma ray logs and found the speculation about the tests being conducted in casing or cement not being drilled out to be unfounded for both tests. The data is consistent with the depths reported in the operational records and confirmed that both tests were conducted in newly drilled formation below previously drilled larger hole. It should also be pointed out that cement is often squeezed at the shoe to repair a leaking cement seal. The drilled out cement after a squeeze job is not strong enough in tension to interfere with the Pressure Integrity Test conducted after the squeeze job. Sometimes multiple squeeze jobs must be done to repair the cement seal around the bottom of the casing. The fact that BP has on its staff a number of scientists that specialize in pore pressure and fracture gradient calculations is an indication to me that it was not trying to cut corners to save money but instead was employing the best available technology to support and advise its experienced team leader. Filings by BP and the number of individuals focused on pore pressure prediction and estimates show a strong emphasis by BP towards obtaining as accurate as possible data related to pore pressure and estimated fracture gradients. The specialists often see only their side of an issue and do not recognize the risk versus benefit of taking additional data. If drilling was stopped to do a Leak-off test in the extended borehole, the risk of getting stuck 22 MDL Ex. 1343, M. Albertin April 2, discussing FIT test. 23 MDL Ex. 3734, BP-HZN-2179MDL

47 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 47 would increase. If the test was carried to a high value, the risk of unnecessarily causing a leak at the liner top had to be considered. Even if the test gave a value close to what was calculated by the specialists, this would not necessarily be representative of formations not yet drilled. The test had accomplished its main objective of demonstrating that the shoe was strong enough to proceed with drilling the next section of hole. Additionally, it is my understanding that the author of the quoted by Dr. Huffman concluded that the test was valid after he sent the The resistance of a rock to fracture is controlled primarily by the overburden stress in the sediments caused by the weight of sediments above and other local tectonics. The overburden stress always increases with depth so that the weaker sediments exposed by an open borehole will usually be near the top of the open borehole. However, rock fracture resistance is also affected by pore pressure, permeability, mineral composition, and cementation. Mother Nature is not uniform and large variability in fracture resistance is to be expected. Figure 15, taken from one of my old lectures, shows the common assumptions made to do fracture mechanics modeling on a picture of what Mother Nature has to offer. Fortunately, the technology for healing downhole fractures has been improved. When drilling in deep water, care must be taken to keep a good supply of loss circulation material on hand. 24 M. Albertin 7/13/11 Dep. Tr. at (discussing FIT test and stating that after discussion, BP believed it was a valid test of rock stronger than they expected); G. Vinson 6/23/11 Dep. Tr. at (same).

48 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 48 Figure 15: Common Assumptions in Fracture Modeling (Homogeneous Layers, Isotropic Media, Elastic Behavior) EFFECT OF SPACER ON NEGATIVE PRESSURE TEST INTERPRETATION The discussion of the previous section was regarding the Pressure Integrity Test conducted after setting each casing string or liner required to complete the drilling operations. Once the production casing is run and cement is circulated as planned, additional pressure tests are conducted before completing the well to put it on production or to temporarily cap the well. A temporary cap allows the drilling rig to be released to drill other wells and a smaller rig moved on the well to complete the well and make it ready for production. An overarching principle of well control that industry follows in drilling and completing wells is to always maintain two independently verified barriers to an uncontrolled release of formation fluids from the well. 25 When drilling ahead in deep water, the two barriers are the hydrostatic pressure of the mud and the blowout preventer stack at the seafloor. The blowout preventer stack has redundant blowout preventers so that it can be closed on any size or shape of pipe in the hole or when nothing is in the hole. Provisions are also made for multiple blowout preventers for the same size range of pipe and for pipe to be stripped into the well under pressure. The hydrostatic pressure of the mud is verified to be a barrier simply by keeping the well filled with mud and observing that it does not flow. The blowout preventers are pressure tested on a routine schedule to verify their ability to function as a barrier. After cementing casing, the casing and cement work together and isolate the rock formations from the inside of the well. However, the casing/cement system must be pressure tested to verify that it serves as a barrier. A positive pressure test of the production casing is conducted in the same manner that the other casing strings and liners are tested. The positive pressure test verifies that the casing is intact and does not leak from the inside of the casing to the outside of the casing from the wellhead to the Top Plug landed on the float collar at the bottom of the casing. However, before the hydrostatic head of the mud in the well can be reduced below formation pressure, a Negative Pressure Test is conducted in which the hydrostatic head is temporarily removed in a controlled manner to observe whether or not fluid can leak past the cement around the outside of the casing and the cement remaining in the bottom portion of the casing. The casing joints below the float collar are called the shoe joints. The purpose of the 25 Transocean s well control handbook includes a discussion of multiple barriers, consistent with industry practice. See Ex. 1454, Transocean Well Control Handbook at TRN-MDL

49 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 49 negative test is to verify that formation fluids cannot enter the well from the outside at the negative pressure difference that was imposed in the test. Once the negative test is passed, the mud that was in the marine riser and in the top part of the well can be replaced with seawater. At this time the two verified barriers would be the tested casing/cement system and the blowout preventer stack. A cement plug could then be set in the top part of the casing. 26 The cement plug would have no pressure differential across it and could be verified by tagging it with pipe to insure the cement had hardened and was in place. BP was planning to test the top plug in the Macondo well to 1000 psi, although there was no regulatory requirement to do so. Once the upper plug was in place, the Marine Riser could be pulled back and laid down on the drilling vessel and the temporary well cap put in place. There would be three barriers in place after the temporary cap was installed. Figure 16 is a schematic showing the Macondo well after circulating cement. The left side of the schematic shows an overview of the well and various casing strings and liners. The right side of the schematic is an enlargement of the bottom part of the production casing and the sediments penetrated below the 9-7/8 inch liner. Figure 17 is a schematic of how the well would have been configured at the end of the temporary capping operation. Note that the change in fluid density occurs only above the depth of the plug (8,367 feet), where the 14 pound per gallon mud is replaced by seawater having a density of about 8.5 pounds per gallon. Everything else remains the same. NEGATIVE TEST NOMENCLATURE It is customary to reference a negative pressure test in terms of the depth to which the hydrostatic pressure will change or the depth of the equipment being tested. For the negative pressure test needed for the temporary capping operation, the negative test had to be conducted to an equivalent mud weight of 8.5 pounds per gallon at the plug depth of 8,367 feet. As shown in Figure 18, this is also equivalent to an 11.5 pound per gallon mud at the top of the shoe track at a vertical depth of approximately 18,100 feet. 26 Regulations require that the plug be at least 100 feet in length and be set no more than 1000 feet below the seafloor. However, BP had sought and received permission to set the plug about 3,400 feet below the seafloor, but would use a 300 foot plug. MDL Ex (BP-HZN-2179MDL ). Setting the plug deeper would facilitate using more weight of pipe to set the casing lockdown device in the subsea wellhead at the seafloor.

50 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 50 Figure 16: Schematic of Macondo Well after Circulating Cement

51 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 51 Figure 17: Planned Configuration of Macondo Well with Temporary Cap I know of no industry standard regarding Negative Pressure Tests. The most common way to conduct a negative pressure test is to pump water down the drill pipe to a depth sufficient to cause the desired reduction in well hydrostatic pressure. To do this, the needed depth and volume required to fill the drill pipe to that depth is first calculated. The rig crew then pumps the calculated water volume down the drill pipe. When pumping is stopped, the surface drill pipe

52 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 52 pressure will be equal to the desired reduction in hydrostatic pressure. The blowout preventers are then closed and the surface drill pipe pressure is bled down to zero while the volume of water bled back is measured. When the drill pipe pressure reaches zero, the volume bled back should be the expansion volume of the fluids in the well corresponding to the pressure decrease. The system is then left static for the duration of the test. For a 30 minute negative test, this period would be 30 minutes. If no fluid flows from the drill pipe, the negative test is passed and reported as a good test. If the drill pipe pressure cannot be bled to zero or bleeds down and then increases, or water continues to flow from the drill pipe, the negative test is said to have failed. As shown in Figure 21 and as will be discussed in more detail later in this report, the system is easily understood using a U-tube analogy with the drill pipe being the high pressure side of the U-tube and the annulus being the low pressure side of the U-tube. The rig crew should know not to let a significant influx of fluid into the well. The test is terminated quickly if the well is flowing. At the end of the test, the pump is started and the drill pipe pressure is increased back to the initial value seen when the blowout preventer was closed. The pressure on the drill pipe is slowly bled through a choke while the annulus kept full of mud by filling through a choke of kill line. This removes the water or synthetic base oil from the drill pipe and returns the well to its initial condition before the test. On a deepwater rig with the blowout preventer at the seafloor, the luxury of additional piping (the choke and kill lines) down to the seafloor is present. Water or some other low density fluid such as synthetic oil can be pumped down the choke line instead of the drill pipe. This option is often preferred because it simplifies the bleed back part of the test. However, the depth of the choke line is fixed, so the hydrostatic pressure reduction achieved during the test is limited by the lowest density fluid available for pumping down the choke line. The rig crew does not have to be told how to run a negative test. This should be a routine operation that fits within their training. A typical procedure sent to the rig would say simply to conduct a negative test to an equivalent density of so many pounds per gallon at a depth of so many feet. The procedure could also say to conduct a negative pressure test at a depth of so many feet to a pressure differential of so many psi. The operator s representative and the rig crew would be expected to know the proper procedure for running the test. I have reviewed several of Transocean s written procedures for Performing Negative Flow Tests. These included tests conducted by pumping a low density fluid down the drill string 27 and 27 The procedure labeled TRN-INV is basically the general procedure described above in which low density fluid is pumped down the drill pipe to create the desired differential.

53 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 53 tests conducted by pumping a low density fluid down the choke line. They also included a negative test run in conjunction with displacing the top part of the well from mud to seawater (separated by a spacer) as was being done on the Macondo Well at the time of the blowout. 28 I reviewed several written Transocean procedures for conducting a Negative Pressure Test after pumping synthetic oil down the choke line and/or kill line. 29 Transocean s Task Specific THINK procedure for the Deepwater Horizon for performing a Negative Flow Test using the choke and kill lines is a variation of this procedure that uses both the choke line and kill line as the high pressure side of the U-Tube. 30 One of the hazards warned about in the THINK procedure is that a closed flow path can give a false indication of no flow. Figure 18: Use of Equivalent Mud Weights in Negative Test Nomenclature 28 The procedure labeled TRN-MDL uses a spacer between the seawater and mud and pumps the spacer above the BOP stack before closing the Blowout Preventer and performing a Negative Test. 29 The procedure labeled TRN-INV does not use the drill pipe and annulus as the two sides of the U- tube, but instead uses the choke line and kill line. The procedure labeled TRN-INV is a slight variation of this procedure that uses the annular Blowout Preventer instead of the blind/shear rams. The procedure labeled BP- HZN-2179MDL used the choke line and the marine riser as the two sides of the U-tube. 30 TRN-MDL

54 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 54 FLUID DISPLACEMENT The removal of the mud from the top part of the well and from the marine riser requires what is called a fluid displacement. This is accomplished by placing the drill string at the depth to be displaced. The replacement fluid is pumped from mud tanks on the rig into the drill pipe while taking return flow of existing well fluid to different mud tanks on the rig. The displacement of a heavy oil base mud by sea water is complicated by the fact that mixing between the two fluids will take place during the displacement and contamination of the mud by sea water can cause the mud to become unstable. If this happens, the mud solids dispersed in the oil phase can start clumping together in the seawater. In order to avoid this mud chemistry problem, the drilling fluid specialist must design a spacer fluid to pump between the mud being removed from the well and the seawater being pumped into the well. The spacer must be compatible with the mud on the leading edge and the seawater on the trailing edge. The drilling fluid specialist is generally in charge of the displacement. 31 He is trained in this operation and the operator depends on him to make sure that the displacement goes well. 32 The displacement of oil base fluids by water base fluids or vice versa have been known to go bad and create a highly contaminated well that is hard to clean. MI-SWACO SPACER MI-Swaco provided drilling fluid and drilling fluid waste management services on the rig for BP. The MI-Swaco drilling fluid specialist wrote procedures and supervised displacement of various fluids into the well. 33 At the end of drilling operations, there was a significant volume of loss circulation material that had been mixed as a contingency in case there was another event of loss circulation. The spacer for the fluid displacement was composed of a blend of two lost circulation material (LCM) pills left over from the drilling operations plus some additional 31 B. Billon 6/23/11 B. Billon Dep. Tr. at 109:11-21 (MI-Swaco personnel in most cases, always prepare a [displacement] procedure ); B. Billon 6/24/11 Dep. Tr. at (discussing M-I Swaco responsibilities); L. Lindner 9/14/11 Dep. Tr. at (discussing development of M-I Swaco displacement procedure). 32 MDL Dep. Ex. 567; B. Billon 6/23/11 Dep. Tr. at 109:11-21; L. Lindner 9/15/11 Dep. Tr. at 549: In addition to the M-I Swaco personnel preparing the displacement procedure and calculating the pumps necessary, the Transocean rig crew would also have evaluated the displacement procedure. Wyman Wheeler s interview notes reflect the collaborative process involved in displacement when he answered the question about displacement figures and said that displacement figures are evaluated by the Transocean rig crew and that the Assistant Driller, the Driller and the Toolpusher all have to agree before the procedure goes forward. W. Wheeler interview notes, TRN-INV Leo Lindner also testified that he provided the displacement procedure to the mudloggers, Transocean crew, MI-Swaco personnel on shore and BP s company man to review. L. Linder 9/14/11 Dep. Tr. at L. Lindner 9/14/11 Dep. Tr. at (discussing M-I Swaco internal discussions regarding the spacer and Lindner testifying that he got the idea from Doyle [Maxie of M-I Swaco] ); L. Lindner 9/14/11 Dep. Tr. at (discussing development of M-I Swaco displacement procedure); MDL Ex

55 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 55 weight material and viscosity enhancer. 34 MI-Swaco had not previously used LCM pills for this purpose but believed the mixture could be used effectively and at the same time better manage waste by providing a beneficial re-use of material. 35 The mixture chemistry was adjusted by not adding the cross linking chemical so that it would not function in the same manner that it would if used to stop loss circulation. Without the cross linking chemical, the polymers present could not link and make the spacer too thick. The fibers and ground limestone that were in the mixture were inert materials that are routinely circulated in the mud without problems. MI-Swaco personnel has circulated this idea within their organization and also told BP that this was their plan. 36 Figure 19: Displacement Planned by MI-Swaco Drilling Fluid Specialist The basic displacement that was planned is shown in Figure 19. Note how the spacer separates the seawater from the 14 pound per gallon mud. In the drill pipe, the thick 16 pound 34 MDL Ex. 1, BP Deepwater Horizon Accident Investigation Report at Appendices P and Q. 35 MDL Dep. Ex. 567; L. Lindner 9/15/11 Dep. Tr. at 583:22-25; B. Billon 6/24/11 Dep. Tr. at 522: B. Billon 6/24/11 Dep. Tr. at (discussing M-I Swaco internal s regarding displacement); MDL Ex (internal M-I Swaco s regarding the spacer); L. Lindner 9/15/11 Dep. Tr. at

56 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 56 per gallon spacer displaces the mud below it. The spacer is about 14% heavier than the mud, but little mixing would take place at the interface because the spacer is so thick. The water displaces the spacer from the drill pipe and since water is only half as dense as the spacer, there would be little mixing between the water and the spacer in the drill pipe. In the annular space outside of the drill pipe, the spacer efficiently displaces the mud upward because it is heavier than the mud and more viscous. Note that the spacer gets shorter when it enters the large-diameter Marine Riser. When the water pushes the spacer up the Marine Riser, a less efficient displacement takes place and there is some mixing of the water and the spacer. However, the spacer contains only non-toxic material that could be released to the environment along with the seawater. The final instructions sent to the rig regarding the temporary capping procedure included the following steps: Test Casing per Application for Permit to Drill to a low pressure test of 250 psi and a high pressure test of 2500 psi. 2. Run in the hole to 8,367 feet. 3. Displace to seawater from there to above the wellhead. 4. With seawater in the kill line, close the annular blowout preventer and do a negative test to a pressure differential of approximately 2350 psi. 5. Open annular blowout preventer and continue displacement. 6. Set a 300 foot balanced cement plug with 5 barrels in the drill pipe. 7. Pull out of hole about 100 to 200 feet above the top of cement and drop a nerf ball. Then circulate a drill string 38 volume. 8. Spot corrosion inhibitor in the open hole. 9. Pull out of the hole to just below the wellhead or above with the 3-1/2 inch stinger 39. (If desired, wash with the 3-1/2 inch stinger but do not rotate. A separate run will not be made to wash since the displacement will clean up the wellhead. 10. Pull out of the hole and make Lead Impression Tool/Lock-Down Sleeve runs MDL Ex The term drill string is used to refer to all of the drill pipe sections and bottom-hole assembly (if any) as a whole. In this case it was composed of a section of 6-5/8 drill pipe on top, 5-1/2 drill pipe in the middle and 3-1/2 pipe in the bottom section. 39 A section of small diameter pipe on bottom is called a stinger. It is best to set a balanced cement plug using a small diameter pipe to reduce disturbing the slurry when pulling the stinger out of the cement after it is placed. 40 A Lead Impression Tool is used to provide an imprint of the area in the wellhead where the top of the production casing has been landed and the seals installed. The impression made into the soft lead block can confirm that the wellhead at the seafloor is properly configured to accept the lock down sleeve that will lock the top of the casing so that upward movement is prevented.

57 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE Test casing and surface plug to 1000 psi with seawater as required by BP s Drilling and Well Operations Practice (DWOP). a. Confirm volume in barrels needed to pressure up on original casing test versus volume in barrels to test the surface plug. Less volume should be required to test the surface plug because of differences in well fluid volume and compressibility. (seawater versus synthetic oil bas mud). b. Plot on chart and send to Houston for confirmation. The negative test procedures included in this plan appear very similar to what was used previously by Transocean on the Kodiak Well in Mississippi Canyon 727 #2. 41 After cementing, the drill pipe was pulled to the planned depth of the displacement. The Boost line, Choke Line, and Kill Line were displaced to seawater. They then pumped water base spacer down the drillpipe and displaced the drill pipe and hole with seawater to the top of the BOP Stack to perform the negative test. The differential pressure was put back using the Halliburton Pump, the annular preventer was opened, and the displacement to seawater was completed. This negative test, which was run on January 28, 2010, used essentially the same procedure as used on the Macondo Well. MI-Swaco s drilling fluid specialist included the negative test of Step 4 in the displacement procedure that he prepared. This procedure is given as Appendix P in BP s Deepwater Horizon Accident Investigation Report issued September 8, The procedure called for first displacing the boost line, 43 choke line, and kill line 44 to seawater and then closing the lower valve where these lines enter the Marine Riser. The MI-Swaco Procedure called for pumping: 73 barrels of seawater or 579 strokes to displace the Boost Line, 8 barrels of seawater or 63 strokes to displace surface lines plus 100 barrels or 794 strokes to displace the choke line (total of 857 strokes), 100 barrels of seawater or 794 strokes to displace the kill line, 425 barrels of 16 pound per gallon spacer or 3373 strokes into the drill pipe, Follow spacer with seawater until 775 barrels or 6150 strokes were pumped. 41 TRN-MDL MDL Ex. 1 at Appendix P; MDL Ex The boost line is attached to the outside of the Marine Riser and enters the Marine Riser above the Blowout Preventer Stack. By pumping down the boost line with the boost pump, the upward fluid velocity in the Marine Riser can be increased. Boosting the fluid velocity in the Marine Riser can help carry the cuttings to the surface in this large diameter pipe. 44 The choke line and kill line are attached to the outside of the Marine riser and provide two high pressure fluid paths to or from the well when the blowout preventers are closed.

58 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 58 At this point in the displacement, the spacer was intended to be above the upper annular in the blowout preventer with only seawater below the blowout preventer down to the planned cement plug depth of 8,367 feet. 45 The blowout preventer would then be closed to run the Negative Pressure Test. Implementation of the procedure by the Transocean Rig Crew began at about 3:00 pm on April 20, An overview of the first part of the displacement is shown in Table 1. Activity Displaced mud from Booster, Choke, and Kill lines Event No. Action Description Time 1 Displaced boost line mud to seawater Pump 1 used to displace boost line from 14-ppg mud to seawater. 15:03-15:15 2 Tested surface lines Pump 2 used to test the surface lines. 15:15-15:21 3 Displaced choke line mud to seawater Pump 2 used to displace choke line from 14-ppg mud to seawater. 15:21-15:38 4 Displaced kill line mud to seawater Pump 2 pumped used to displace kill line from 14-ppg mud to seawater. 15:38-15:55 Displaced mud from Drill pipe and Riser 5 6 Pumped spacer through drillpipe into well Pumped seawater through drillpipe into well Pumps 3 and 4 used to pump about 454 bbls of 16-ppg spacer from Pit 5 into the well displacing 14-ppg mud. Pumps 3 and 4 used to pump 353 bbl of seawater behind the spacer to place the theoretical spacer/water interface above the BOP. Pumps are shut down. 15:55-16:27 16:27-16:53 Table 1 Summary of Events during the Initial Displacement Sperry-Sun Real Time Data indicated the following: 616 strokes of seawater was pumped (Boost Line), 872 strokes of seawater was pumped (Surface line and Choke Line), 845 strokes of seawater was pumped (Kill Line), 3607 strokes were pumped down drill pipe from the tank containing the spacer, Pumping continued with seawater down drill pipe until total stroke count of 6408 (807 barrels) were pumped. At this point, the pumps were stopped and the blowout preventer was closed to conduct the negative test. 45 Leo Lindner, a MI-Swaco Drilling Fluid Specialist testified that he had calculated that the spacer would be 17 feet above the Blowout Preventer. L. Lindner 9/14/11 Dep. Tr. at See also MDL Ex. 1 at page 84, BP Deepwater Horizon Accident Investigation Report at 84 and Appendix Q (spacer calculated to be 12 feet above BOP).

59 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 59 The intent of the procedure was to place the entire spacer above the upper annular in the blowout preventer. The actual position of the bottom of the spacer at the time pumping was stopped to do the negative pressure test cannot be accurately determined. If the spacer tank contained 454 barrels of spacer, 46 the interface would be below the bottom of the blowout preventer. In addition, the pump pressure observed when pumping was stopped indicated that the effective height of the spacer was greater than expected for all of the spacer to be above the blowout preventer. I would expect mixing at the spacer seawater interface during the displacement because the spacer was almost twice the density of the water. Once pumping stopped, the rate of settling of solids from the spacer into the seawater below would have increased. Spacers are designed to separate fluids for circulating conditions and not for static conditions. In my opinion, the procedure followed did not place all of the spacer above the blowout preventer and spacer material was in and below the blowout preventer. NEGATIVE PRESSURE TEST A summary of the actions taken by the rig crew during the first Negative Pressure Test is shown in Table 2. A plot of data recorded in real time and transmitted to shore bases by the Sperry-Sun well monitors is shown in Figure 20. The first action was to close the blowout preventer and bleed fluid from the drill pipe. The Transocean Rig crew conducting the negative pressure test should have expected the drill pipe pressure and choke line pressure to both bleed to zero after bleeding about 4 barrels of water from the drill pipe. Estimates of the initial amount of fluid bled back were in the range of 15 to 25 barrels and the pressure did not bleed to zero. This was an indication that the Negative Pressure Test had failed, but it also could be explained by fluid leaking past the closed blowout preventer. With some additional checking and bleeding, it was seen that the annular preventer was leaking and allowing fluid from above the blowout preventer to leak into the casing that was intended to contain only seawater. The leak was stopped by increasing the closing pressure to the annular in the blowout preventer. An estimate of the amount of fluid that was required to refill the casing was in the range of barrels. 47 This is an indication that the barrels of seawater had been bled from the well before the leak was found. This would have moved additional spacer material down into the area where the choke line and kill line connect with the 46 The MI-Swaco Drilling Fluid Specialist testified that the volume of spacer in the tank was measured to be 454 bbl. L. Lindner 9/14/11 Dep. Tr. at 151, MDL Ex. 1 BP Deepwater Horizon Accident Investigation Report at 84; Macondo Well Incident Transocean Investigation Report, Vol. 1 at 96.

60 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 60 well and substantially further down into the drill pipe casing annulus inside the narrower diameter production casing. Activity Event No. Action Description Time 7 Closed BOP The annular BOP is closed. 16:53-16:54 8 Bled drill pipe The drill pipe is opened and an estimated 25 bbl of fluid are bled from the drill pipe, and the drill pipe pressure decreases to 1,250 psig. The drill pipe is then closed. 16:54-16:57 9 Opened subsea kill The subsea kill valve is opened. The drill pipe pressure increases to 1400 psig valve and the kill line pressure decreases to 680 psig. 16:57-16:58 10 Bled drill pipe and Resume bleeding drill pipe and bleed kill line. The kill line pressure bleeds to 0 kill line psig while the drill pipe pressure decreases to approximately 340 psig. 16:58-16:59 Crew conducts NPT on Drill pipe Bled drill pipe; Fluid leaked through BOP; Spacer moved downward Closed drill pipe; Annular BOP sealed Continue bleeding drill pipe. The drill pipe pressure decreases to approximately 240 psig before increasing. Discovered fluid level in riser has fallen indicating leak of annular BOP and downward movement of the spacer. The drill pipe is closed and the drill pipe pressure increases rapidly to 830 psig, more slowly to 1,250 psig, and then decreases to 1,200 psig. Early in this period the annular preventer is sealed preventing any further movement of 16 ppg spacer into BOP. Based on the volume used to refill the riser, approximately 50 bbl of spacer passed the BOP during the leak. 16:59-17:05 17:05-17:26 13 Bled drill pipe The drill pipe is opened and approximately 15 bbl bled. Compressibility calculations indicate that only 3.5 bbl of the fluid bled can be attributed to liquid expansion, suggesting an cement seal is leaking. The drill pipe pressure decreases to approximately 0 psig. 17:26-17:32 14 Well becomes underbalanced to formation; hydrocarbons leak into well until pressure builds up to equalize with formation pore pressure. ~17:30 Table 2 Summary of Actions Taken during the First Negative Pressure Test The Transocean rig crew was responsible for placing the fluids in the correct position and for maintenance of the blowout preventer. After refilling the well, it should have been obvious that the spacer was not correctly positioned for the planned negative test using the Kill Line. I would have expected that the situation would have been discussed with the MI-Swaco mud specialist and that allowance would have been made for mixing of seawater and spacer to have taken place while the leaking blowout preventer was being diagnosed and the well refilled. The well had been static so long, it is questionable that a valid 8.5 pound per gallon negative test at 8,367 feet could have been run using the Kill line without first displacing the well back to mud and starting over. Unfortunately, nothing to rectify the situation was done; there was no attempt to circulate the 65 barrels that was known to have leaked below the blowout preventer back to the correct position above it.

61 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 61 Figure 20: Plot of Real Time Data Recorded during Negative Pressure Test After stopping the leakage past the annular blowout preventer, the rig crew bled an additional approximately 15 barrels from the drill pipe to bring the pressure to zero. This was three times the volume that should have been required and should have caused the test to be interpreted as a failed test. The Well Site Leader decided to re-test using the kill line since the kill line was specified in the procedure approved by MMS (now BOEMRE). 48 Table 3 shows a summary of the actions taken during the second Negative Pressure Test. After making the appropriate valve changes to line up for testing using the kill line, the kill line was bled to zero. Estimates of the volume bled were in the range of 3-15 barrels. The surface kill line valve was closed and the drill pipe pressure slowly increased to 1400 psi. This was a direct indication that the second negative test had also failed and that formation fluid was leaking past the cement seals into the well. The leak stopped after the pressure in the well became in balance with the formation pore pressure and the observed drill pipe pressure. This is indicated by the stabilized drill pipe pressure readings. 48 MDL Ex (outlining temporary abandonment plan).

62 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 62 Activity Event No. Action Description Time 15 Lined up to monitor test with kill line System lined up to perform negative pressure test on kill line. Pressure likely had built-up on the drill pipe due to leakage past cement seals. 17:32-17:52 16 Bled drill pipe and kill line A valve is opened exposing a gauge to the drill pipe pressure. After rapidly increasing to 770 psig, the pressure is again bled off. Kill line is opened and witnesses report that between 3 and 15 bbls of seawater flowed from the well, and that the flow spurted before being shut in. Crew change time for certain individuals. 17:52-18:00 17 ~18:00 18 Closed drill pipe; Opened kill subsea valve The drill pipe is closed and the drill pipe pressure gradually increases to 1,200 psig. A slight pressure increase is recorded on the the kill line, suggesting the subsea kill valve was opened. 18:00-18:35 Crew conducts NPT on Kill Line 19 Well potentially brought into balance; Drill pipe pressure stabilizes The drill pipe pressure increases to 1,400 psig and the kill line pressure increases to 140 psig, suggesting the well is in communication with the formation. The drillpipe pressure becomes relatively stable at this value throughout the duration of the negative pressure test. This drillpipe pressure conicides with the well near balance with the formations. 18:35-18:40 20 Pumped down kill line; Bled kill line Fluid is pumped into the kill line and the pressure spikes to 490 psig, suggesting the kill line was full. The kill line pressure is then bled off. A slight reduction in the drill pipe pressure is noted on two occasions with no corresponding response on the kill line pressure. The kill line pressure increases while drill pipe pressure decreases suggesting the columns of fluid in drill pipe and kill line equalized. 18:40-19:07 21 Bled Kill line The kill line valves at BOP and surface are opened and the pressure is bled off, returning a small volume of liquid (0.2 bbl). At least 3 bbl of spacer in BOP is potentially drawn into 150 ft of kill line. 19:07-19:15 The negative test is conducted by monitoring the kill line for flow. The drill pipe is stable at 1,400 psig 22 19:15-19:54 and the kill line is open with no flow. Table 3 Summary of Actions Taken during the Second Negative Pressure Test The Transocean Toolpushers and BP Well Site Leaders had considerable discussions in an effort to resolve the conflicting information. They appeared to be working together to reach a consensus as to the reason for the conflicting test data. Witness testimony 49 indicated that the Toolpusher proposed that the drill pipe pressure was caused by a bladder effect that he had seen before. This explanation was accepted by everyone and a conclusion was reached that the Negative Pressure Test had passed, when in actuality it was not. I find this misinterpretation by experienced people trained in well control extremely hard to understand. Figure 21 is a U-tube analogy that is always taught to be used to understand differences in surface pressures and how surface pressure is related to downhole pressure. 50 In this case, if the well was configured as intended with seawater everywhere in the hydraulic path 49 See MDL Ex. 1 at pgs. 25 and 86; L. Lambert 5/9/11 Dep. Tr. at (testifying regarding discussions about negative test); L. Lambert 5/10/11 Dep. Tr. at 444 (same). 50 Rig workers are taught in Well Control training to use a U-tube analogy where the drill pipe is one side of the U- Tube or manometer and the annular fluid path is the other side of the U-tube.

63 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 63 between the surface drill pipe pressure and the surface kill line pressure, both the drill pipe pressure and the kill line pressure would have to be zero for the negative test pressure to be good. As shown in Figure 18, the hydrostatic pressure computed at the top of the shoe joint using a surface mud weight of 14 pounds per gallon is 10,784 psi. Adding a stabilized surface drill pipe pressure of 1,400 psi to the hydrostatic pressure gives a pressure of 12,184 psi at the top of the shoe joint. This is equivalent to a pressure gradient, which is close to the measured formation pore pressure. This calculation does not account for low density oil that may have occupied an unknown height above the Shoe Track. The effect of the oil would be to slightly lower the equivalent density calculated and make it even closer to the measured formation pore pressure. This calculation also does not account for the potential effect of the fluid column height of spacer below the blowout preventer, which could also impact the pressure reading. The reason that the kill line did not flow during the 30 minute test cannot be completely resolved. It could have been caused by a sufficient effective height of spacer material 51 on the kill line side of the U-tube, by a plugged kill line due to settling of barite and loss of circulation material, or a closed valve. We know that a pressure relief valve actuated later when trying to start pumping down the kill line. This could have been caused by a plugged kill line or a closed valve. There were other instances of pressure relief valve actuations reported in the drilling records. If plugging did occur due to spacer, in my opinion, the plugging would be related more to the placement of the spacer than to the inclusion of loss circulation material in the spacer. Barite was by far the predominant solid in the spacer. A 16 pound per gallon spacer with no loss circulation material will contain about 30% barite by volume. The barite is about twice as dense as loss circulation material and would settle readily from a 16 pound per gallon mud that is mixing freely with seawater in a static condition. Barite is a good plugging agent and is sometimes intentionally used to plug off an underground blowout. If a 16 pound per gallon spacer had been used that had no loss circulation material included, placement of the spacer opposite the kill line prior to a negative test would have still been undesirable. 51 Some spacer material would have to be in the kill line and below the closed blowout preventer to have sufficient hydrostatic head.

64 EXPERT REPORT OF ADAM T. (TED) BOURGOYNE, JR., PH.D., P.E. PAGE 64 Figure 21: U-tube analogy for Negative Test being conducted (pressures equal) WELL MONITORING FAILURE The blowout preventer upper annular was opened after the negative test and the hydrostatic head of the mud and spacer above the preventer returned the well to an overbalanced condition in which the well could not flow. However, by proceeding with the displacement of mud from the riser, it was just a matter of time until the well began flowing again. Once the barrier of mud weight was removed, normal kick detection practices became extremely important. It was clearly the responsibility of the Transocean Rig Crew and the Sperry-Sun mud loggers to do so. The Deepwater Horizon was a top of the line rig with approximately 20 mud tanks that could have been configured to properly monitor the well. This would have required organizing the active system 52 so that the volume of mud returning from the well could always be compared to 52 Well monitoring systems allow the total volume in the mud tanks designated as included in the circulation system to be constantly monitored. The active system includes the tanks from which fluid is being pumped into the well and the tanks to which fluid is being returned from the well.