United States District Court, E.D. Louisiana
[REVISING PHASE ONE FINDINGS OF FACT AND CONCLUSIONS OF LAW]
CARL BARBIER, Judge.
The Court has made three minor revisions to its September 4, 2014 Findings of Fact and Conclusions of Law (Rec. Doc. 13355), which are listed below. These revisions do not alter the substance of the document. A revised copy of the Findings of Fact and Conclusions of Law is attached to this order.
Page 21, ¶ 74: "which willd depart" changed to "which will depart"
Page 59, ¶ 233: "12:30 a.m. on April 19, 2010" changed to "12:30 a.m. on April 20, 2010"
Page 82, ¶ 323: "Transocean's flow out sensor was not bypassed, and therefore the Transocean drill crew could still monitor flow out" changed to "Transocean's flow out sensor was not bypassed; therefore, the Transocean drill crew could still monitor flow out."
IN THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF LOUISIANA
FINDINGS OF FACT AND CONCLUSIONS OF LAW PHASE ONE TRIAL
Pursuant to Federal Rule of Civil Procedure 52(a), the Court enters these Findings of Fact and Conclusions of Law relative to the Phase One trial. If any finding is in truth a conclusion of law, or if any conclusion stated is in truth a finding of fact, it shall be deemed so.
The Court has also issued simultaneously with these Findings of Fact and Conclusions of Law a separate order ruling on various motions pertaining to the Phase One trial.
I. INTRODUCTION AND PROCEDURAL HISTORY
1. Figure 1 illustrates the DEEPWATER HORIZON, the HORIZON's marine riser and blowout preventer ("BOP"), and the Macondo well prior to the blowout on April 20, 2010.
Figure 1 the MODU DEEPWATER HORIZON (sometimes referred to as the "HORIZON") as it was in the process of temporarily abandoning a well, known as Macondo, it had drilled on the Outer Continental Shelf off the coast of Louisiana.
3. Eleven men died tragically in the incident: Jason Anderson, Dewey Revette, Aaron (Dale) Burkeen, Donald Clark, Stephen Curtis, Roy (Wyatt) Kemp, Karl Kleppinger, Shane Roshto, Adam Weise, Keith Blair Manuel, and Gordon Jones. At least seventeen others were injured. The survivors evacuated to the M/V DAMON BANKSTON, a supply vessel that was near the HORIZON when the explosions occurred.
4. The explosions and/or fire should have triggered the automatic function on the HORIZON's BOP, but that function either failed to activate the BOP or the BOP otherwise failed to shut in the well. Subsequent attempts to operate the BOP with remotely operated vehicles also failed to stop the blowout.
5. Several vessels responded to the distress calls and attempted to extinguish the fire with their monitors (water canon). Despite these efforts, the HORIZON burned continuously until mid-morning on April 22, when it capsized and sank into the Gulf of Mexico.
6. As the rig descended, the marine riser-the approximately 5, 000 feet of pipe that connected the rig to the BOP-collapsed and broke. Millions of gallons of oil discharged into the Gulf of Mexico over the next 87 days. number is the "pdf" page number (i.e., page "1" is the very first page of the exhibit), which may be different from the page number appearing on the exhibit. discharge halted on July 15, 2010. In mid-September, a relief well intercepted the Macondo well and permanently sealed it with cement.
8. It was not long after the initial explosions that the first lawsuits were filed. Since that time, approximately 3, 000 cases, with over 100, 000 named claimants, have been filed in federal and state courts across the nation. These suits asserted a wide array of claims including wrongful death and personal injury due to the explosion and fire, post-incident personal injury resulting from exposure to oil and/or the chemical dispersants used during the oil spill response, damage to property or natural resources, and economic losses resulting from the oil spill.
9. On August 10, 2010, the United States Judicial Panel on Multidistrict Litigation transferred most federal cases to this Court as Multidistrict Litigation no. 2179 ("MDL 2179").
10. This Court adopted a phased trial proceeding that ultimately focused on two cases within MDL 2179: In re Triton Asset Leasing GmbH, et al. (Civ. A. No. 10-2771) and United States v. BP Exploration & Production Inc., et al. (Civ. A. No. 10-4536). Both cases are before the Court for all purposes,  and both are proceedings in admiralty under 28 U.S.C. § 1333(1) and Federal Rule of Civil Procedure 9(h). Consequently, these cases may be tried by this Court without a jury.
11. In re Triton Asset Leasing GmbH is a limitation action filed by several of the Transocean entities pursuant to 46 U.S.C. § 30501, et seq., commonly referred to as the Shipowner's Limitation of Liability Act. Thousands of claims were filed in that action. Pursuant to Federal Rule of Civil Procedure 14(c), the Transocean entities impleaded other parties who it alleged were partially or wholly liable. The Rule 14(c) parties then counterclaimed against Transocean and crossclaimed against one another. Although Transocean is technically the "plaintiff" or "petitioner" in a limitation action, for simplicity the Court will sometimes refer to it and the impleaded 14(c) parties as "Defendants." Likewise, claimants in the limitation proceeding who typically would occupy the position of plaintiffs will be referred to as "Plaintiffs."
12. United States v. BP Exploration & Production Inc. concerns the United States' claims for civil penalties under Section 311(b) of the Clean Water Act, 33 U.S.C. § 1321(b), and for a declaratory judgment of liability under the Oil Pollution Act of 1990, 33 U.S.C. 2701, et seq., and the Declaratory Judgment Act, 28 U.S.C. § 2201. The United States sued BP Exploration and Production, Inc., Anadarko Exploration & Production LP, Anadarko Petroleum Corporation, MOEX Offshore 2007 LLC, various Transocean entities, and QBE Underwriting Ltd., Lloyd's Syndicate 1036. Other parties were added as third-party defendants.
13. The "Phase One" trial commenced on February 25, 2013, and concluded on April 17, 2013. Known as the "Incident Phase, " it addressed fault determinations relating to the loss of well control, the ensuing explosion and fire, the sinking of the DEEPWATER HORIZON, and the initiation of the release of oil from the well. Phase One also considered issues related to Transocean's limitation defense, as well as the various cross-, counter-, and third-party claims between the several defendants.
14. "Phase Two" commenced on September 2013, and concluded on October 18, 2013. This phase was divided into two segments: "Source Control" and "Quantification." The former concerned issues pertaining to the conduct or omissions relative to stopping the release of hydrocarbons. The latter segment pertained to the amount of oil actually released into the Gulf of Mexico, which is an important factor for determining the amount of civil penalties under the CWA.
15. After each phase the parties submitted memoranda, responses, and proposed findings. Phase One post-trial memoranda and proposed findings were submitted on June 21, 2013; response memoranda were submitted on July 12, 2013. Phase Two post-trial memoranda and proposed findings were submitted on December 20, 2013; response memoranda were submitted by January 27, 2014.
II. PARTIES TO THE PHASE ONE TRIAL
i. The BP Entities
16. BP Exploration & Production, Inc. was the primary leaseholder of the Macondo site. BP Exploration & Production, Inc. is also the only BP entity that was sued by the United States in United States v. BP Exploration & Production Inc., et al.
17. BP America Production Company contracted with Transocean Holdings LLC to drill the Macondo well.
18. BP Exploration & Production, Inc. and BP America Production Company are direct or indirect wholly-owned subsidiaries of BP p.l.c.
19. BP Exploration & Production Inc. and BP America Production Company are sometimes collectively referred to as "BP."
ii. The Transocean Entitiesx
20. Triton Asset Leasing GmbH was the owner of the DEEPWATER HORIZON.
21. Triton Asset Leasing GmbH bareboat chartered the HORIZON to Transocean Holdings LLC. Transocean Holdings LLC was also the contracting party with BP for the Macondo well.
22. Transocean Deepwater Inc. employed the crew of the HORIZON.
23. Transocean Offshore Deepwater Drilling Inc. employed the HORIZON's supervisory and managerial employees onshore.
24. Triton Asset Leasing GmbH, Transocean Holdings LLC, Transocean Deepwater Inc., and Transocean Offshore Deepwater Drilling Inc. are all subsidiaries of Transocean Ltd.
25. Triton Asset Leasing GmbH, Transocean Holdings LLC, Transocean Deepwater Inc., and Transocean Offshore Deepwater Drilling Inc. are sometimes collectively referred to as "Transocean."
26. Halliburton Energy Services, Inc. ("Halliburton") was contracted by BP to provide cementing services and mudlogging services, the latter of which was performed by Halliburton's Sperry division.
iv. Cameron and M-I
27. Cameron International Corporation, f/k/a Cooper Cameron Corporation ("Cameron"), manufactured the HORIZON's blowout preventer.
28. During the Phase One trial, Cameron moved for judgment on partial findings pursuant to Federal Rule of Civil Procedure 52(c). The Court orally granted this motion and dismissed all claims against Cameron, including any counter-claims, cross-claims, and third-party claims.
29. M-I, LLC contracted with BP to provide goods and attendant services related to drilling fluids at Macondo.
30. M-I, LLC also moved for judgment on partial findings during the Phase One trial, which the Court granted.
31. The Phase One plaintiffs include the United States, the States of Louisiana and Alabama, and numerous private individuals, businesses, or other entities who have filed claims in Transocean's limitation action, Civ. A. No. 10-2771.
C. Non-Parties to Phase One Trial
32. Due to various settlements, rulings, and/or stipulations, BP's co-lessees-the MOEX entities and the Anadarko entities-were not parties to the Phase One trial. Furthermore, prior to the Phase One trial the Court granted motions for summary judgment by defendants Weatherford U.S., L.P., Weatherford International, Inc., and Dril-Quip, Inc., and dismissed all claims against them.
III. SUBSTANTIVE FINDINGS OF FACT
A. The DEEPWATER HORIZON
33. On December 9, 1998, predecessors to Transocean Holdings, LLC and BP America Production Company entered into a contract for the construction, use, and operation of the DEEPWATER HORZION.
34. The HORIZON entered service in 2001. It was capable of drilling up to 35, 000 feet deep at a water depth of 10, 000 feet. Prior to Macondo, the HORIZON had successfully drilled approximately 50 wells. All of these were in the Gulf of Mexico, and all except one were for BP.
35. The HORIZON was 396 feet in length and 256 feet in breadth. It floated on two massive pontoons. Four large columns rose from the pontoon and supported the main deck, drill floor, derrick, bridge, engines, living quarters, helipad, cranes, etc.
36. Figure 2 depicts the DEEPWATER HORIZON prior to April 20, 2010.
37. The HORIZON was a mobile offshore drilling unit or "MODU." which is a general category of drilling vessel. The HORIZON was a self-propelled, dynamically-positioned, semi-submersible MODU.
38. "Semi-submersible" refers to the fact that the HORIZON would partially submerge itself during drilling operations, which increased the rig's stability.
39. "Self-propelled" refers to the fact that the HORIZON used its eight large thrusters to move from place to place.
40. "Dynamically-positioned" refers to the fact that the HORIZON also used its thrusters to keep itself relatively stationary over a well, as opposed to relying on anchors or some other attachment to the seafloor. Thus, one or more of the HORIZON's thrusters were active nearly all of the time, even when the rig was not making actual headway.
41. The HORIZON had a master, chief mate, dynamic positioning operators, bosuns, able-bodied seamen, and ordinary seamen. These Transocean employees were commonly referred to as the "marine crew" and were responsible for, among other things, the MODU's navigation function and keeping the MODU "on station" with the dynamic positioning system. There were other Transocean "crews" aboard the HORIZON. Notably, the "drill crew" was primarily responsible for the MODU's drilling function and consisted of the Offshore Installation Manager, toolpushers, drillers, roustabouts, and others.
B. MC252 and the Macondo Well
42. On March 19, 2008, BP Exploration and Production Inc. acquired a lease from the United States of 5, 760 acres of property on the Outer Continental Shelf comprising Mississippi Canyon Block 252 ("MC252").
43. BP was the "operator" of the Macondo well under MMS regulations, meaning it was "the person the lessee(s) designates as having control or management of operations on the leased areas or a portion thereof." As operator and primary leaseholder, BP's responsibilities included assessing the geology of the site, engineering the well design, obtaining regulatory approvals for well operations, retaining and overseeing the project's contractors, and working on various aspects of the well and drilling operations.
44. BP America Production Company contracted with Transocean Holdings, LLC to drill the Macondo well in MC252.
45. The Macondo well was drilled in approximately 5, 000 feet of water, considered "deepwater" in the current oil and gas industry. It was located approximately 50 miles south of the Louisiana coast. The initial well plan called for a total depth of 20, 200 feet, measured from the deck of the MODU.
46. The DEEPWATER HORIZON was not initially chosen to drill the Macondo well. Instead, the MARIANAS, another Transocean-owned semi-submersible, spudded the well on October 6, 2009.
47. The MARINAS left the Macondo well in November 2009 after it was damaged by Hurricane Ida. At that point the MARIANAS had drilled the well to approximately 9, 000 feet deep.
48. The DEEPWATER HORIZON was chosen to resume drilling the Macondo well. It left the Kodiak well-another BP well-and arrived onsite on January 31, 2010. A few days later the HORIZON latched its BOP onto the wellhead. Drilling resumed on February 11, 2010.
C. Drilling the Macondo Well
i. Some Offshore Drilling Concepts
49. In simplified terms, drilling the Macondo well involved a repeated sequence of steps: drill a certain distance, stop drilling, set "casing" to reinforce the wellbore, cement the casing in place, drill further, stop, set more casing, and so forth.
50. Casing is essentially large diameter pipe that is placed inside a drilled-out section of the well. Once properly cemented in place, casing isolates the adjacent geologic formation (i.e., the rock) from the well.
51. The top of each successive casing string was set inside the previous one. Thus, the diameter of the well decreased slightly with each new piece of casing. Setting casing also takes time. Consequently, there are economic incentives for a well operator to set as few casing strings as possible.
52. As a well is drilled, it encounters different layers of rock, some of which contain fluids-e.g., hydrocarbons or brine-within the pore spaces of the rock. These fluids are under pressure and, if the rock has sufficient permeability, will flow into an area of lower pressure. When formation fluids unintentionally flow into the wellbore, it is called a "kick." Unchecked, a kick can develop into a "blowout, " an uncontrolled flow of formation fluids into the wellbore and possibly to the surface. Kick events involving oil or gas are particularly dangerous given the flammable nature of the hydrocarbons.
53. During normal drilling operations, the primary means of preventing kicks and keeping the well under control is by maintaining an "overbalanced" state; i.e., pressure in the wellbore is greater than the "pore pressure" of any exposed formation(s). This is typically achieved by pumping a dense drilling fluid, commonly known as "drilling mud" or simply "mud, " into the well. Generally speaking, if the "mud weight" (a combination of mud density and the length of the mud column) is greater than the pore pressure, the well will be overbalanced and formation fluids should not migrate into the well.
54. If the only concern was ensuring that formation fluids did not flow into the well, then a well operator would maintain a very high mud weight. This is not the only concern, however. If the mud weight exceeds the "fracture gradient" of the exposed formation, the rocks will fracture. If this happens, the mud may escape into the formation-what is called "lost returns." Lost returns can lower the pressure inside the well to the point it becomes "underbalanced, " resulting in a kick. Large fractures can also cause "underground blowouts, " where there is an uncontrolled flow of formation fluids through the well from one zone to another.
55. Thus, deepwater drilling requires a delicate balance between pore pressure, mud weight, and fracture gradient. Mud weight must be kept above the pore pressure but below the fracture gradient. In fact, federal regulations required well operators to maintain a certain "safe drilling margin" between their mud weight and the fracture gradient. As the United States' expert, Dr. Alan Huffman, explained, the drilling margin is intended to provide a sort of cushion for drilling operations. If the well encounters an area with higher pore pressure, the operator will be able to increase the mud weight (to overpower the pore pressure and suppress a kick) without fracturing the exposed rock. If an operator cannot maintain the specified safe drilling margin, federal regulations require it to "suspend drilling operations and remedy the situation."
ii. Drilling Operations at Macondo
56. Drilling the Macondo well did not go smoothly. Some called it the "well from hell."
57. Many of the problems at Macondo stemmed from the fact that the well encountered increasingly fragile sandstone. This contributed to a narrow window between pore pressure and fracture gradient, particularly as the well got deeper. BP was aware of this issue, but did not always manage it properly.
58. On October 26, 2009, while drilling with essentially no margin between the mud weight and the fracture gradient-much less the safe drilling margin required by the MMS-the MARIANAS experienced a kick at a depth of 8, 970 feet. The well was shut in (i.e., one or more of the sealing elements in the BOP were used to seal the well) and well control operations were initiated, which succeeded. BP decided to drill 100 more feet so it could set casing in hard shale rather than in the delicate sandstone. As BP contemporaneously acknowledged, its decision to drill another 100 feet with no drilling margin came with a risk that it might encounter "another overpressured sand package that would initiate a potentially uncontrollable well control event." The well was successfully drilled forward and BP was able to set the casing in shale. Still, the casing was set significantly shallower than BP had planned. As Transocean's expert, Calvin Barnhill, concluded, "Simply put, the well had run out of drilling margin requiring the 18" drilling liner to be set 917' high."
59. The HORIZON experienced another kick on March 8, 2010, at a depth 13, 250 feet, when it drilled into a higher-than-anticipated pore pressure area. The well was successfully shut in; however, the formation had collapsed around the drill pipe, resulting in a "stuck" pipe. The pipe was severed with explosives and a cement plug was placed around the severed piece of pipe at the bottom of the well. The well was then sidetracked; i.e., the HORIZON drilled around the stuck pipe and then continued downwards.
60. The cause of the March 8th kick was BP's decision to drill faster than its geologists could analyze the data from the well, BP's decision to ignore the information it did have, or both.
61. The well also experienced multiple lost returns incidents.
62. Notably, on April 4, 2010, the well lost total returns at a depth of 18, 260 feet and was shut in. Losing total returns means that all of the mud pumped from the rig escaped through open fractures in the formation at the bottom of the well.
63. BP decided to spend the next 5 days pumping wellbore strengthening treatments in an effort to repair the formation. BP also reduced the mud weight. These efforts succeeded insofar as the well stopped losing returns, but the formation remained in a very fragile state.
64. "At this piont, [BP] was faced with a touch decision." At 18, 260 Feet the well had encountered the primary reservoir sands, but BP needed to drill another 100 feet to ensure that the well was through the entire primary reservoir package and to be able to conduct wireline evaluation operations and completion procedures. However, as BP's Geological Operations Coordinator observed, drilling further meant it would be drilling with "minimal, if any, drilling margin."
65. On April 9, 2010, BP decided to drill the extra 100 feet. Dr. Huffman, whom the Court found to be credible, viewed this decision as "one of the most dangerous things [he] had ever seen in [his] 20 years' experience" and "totally unsafe." The Court agrees that the decision was dangerous and further finds that it was motivated by profit.
66. After drilling the extra 100 feet BP called total depth. The drilling operation concluded at that point.
67. At 18, 360 feet the well had reached its primary objective sands, but not the deeper, secondary objectives. Indeed, the original planned depth for the Macondo well was 20, 200 feet. BP called total depth early because it "had simply run out of drilling margin." BP's Geological Operations Coordinator stated, "Drilling ahead any further would unnecessarily jeopardize the wellbore.... At this point it became a well integrity and safety issue."
68. The United States' expert, Dr. Huffman, and BP's expert, Dr. Adam Bourgoyne, disagree over whether the decisions BP made relative to drilling were safe, in accordance with federal regulations, and consistent with industry standards.
69. Although the Court found both experts well qulified, on the whole, the Court agrees with Dr. Huffman's conclusions.
70. Dr. Bourgoyne points out that the blowout occurred during the temporary abandonment procedure, days after the drilling phase concluded. He concludes, then, that BP's decisions during drilling had nothing to do with the blowout, explosion, or oil spill.
71. Dr. Bourgoyne's point is not without merit, but the Court does not entirely agree. Although drilling operations concluded without major catastrophe, the decision to drill the last 100 feet of the well with little or no margin left the wellbore in an extremely fragile condition. This resulted in the presence of a large amount of debris in the well when the production casing was set in the well a few days later. As will be explained, this debris compromised the production casing, which led to the incorrect placement of cement, which in turn permitted hydrocarbons to enter the well on April 20, 2010. Therefore, BP's decision to drill the final 100 feet was the initial link in a chain that concluded with the blowout, explosion, and oil spill.
iii. Post-Drilling Operations: Production Casing and Temporary Abandonment
72. As of April 9, 2010, BP was $60 million (60-70%) over budget and 54 days behind schedule on the Macondo well. For each additional day the HORIZON remained at the Macondo well, BP lost approximately another $1 million. Moreover, the HORIZON was under pressure to get to the Nile well, and then to the Kaskida well, which BP needed to spud by May 16th or face losing the lease.
73. After calling total depth, BP planned to set production casing and then temporarily abandon the Macondo well.
74. The production casing is discussed below. Temporary abandonment is the process by which a well is secured so the operator can safely leave the well before returning to begin completion operations. It involves placing cement or mechanical barriers in the well to replace the barriers formerly provided by the drilling mud and the BOP, which will depart with the MODU. It was intended that another rig would eventually come to Macondo, drill through the temporary barriers, perform completion operations, and turn Macondo into a producing well.
75. The temporary abandonment procedure ultimately selected for the Macondo well involved a series of steps. These included (1) setting a cement plug at the bottom of the well (known as the "production casing cement"), (2) displacing some of the mud in the well to seawater and performing a negative pressure test, (3) displacing all of the mud above 8, 367 feet to seawater, and (4) placing a second cement plug between 8, 067 feet and 8, 367 feet.
76. When the blowout occurred on the evening of April 20, 2010, steps (1) and (2) were complete and step (3), known as "final displacement, " was underway.
D. Production Casing
i. Long String Casing vs. Liner With Tieback
77. After a well has been drilled to total depth, additional tubing is typically installed that will allow hydrocarbons to be moved from the target formations to the surface. These lengths of tubing are known as production casing or production liners.
78. BP selected a long string production casing for the Macondo well, which extends continuously from the wellhead at the seafloor to the bottom of the well.
79. Figures 3 illustgrartes a long string production casing.
80. Other options were available, such as a production liner. A production liner attaches to the bottom of the previous casing and extends down to the bottom of the well. The liner is later "tied back" to the wellhead with separate casing.
81. Figure 4 illustrates a production liner (left) and the liner with tie-back (right).
82. The Court heard competing opinions regarding the reasonableness of BP's decision to use a long string production casing instead of a production liner with tieback. Several parties proposed that BP's decision was unreasonable and primarily or entirely driven by a desire to cut costs.
83. The weight of the evidence does not show that BP's decision to use a long string production casing was unreasonable. It is clear that BP's engineering and operational personnel thoroughly debated this issue. There were pros and cons associated with each option, such that one did not appear significantly more advantageous over the other at the time BP made its decision. The initial cost savings associated with the long string production casing did enter the analysis, but it was one of several factors considered.
84. Furthermore, using a long string production casing insted of a production liner did not cause or contribute to the blowout. As the Court understands the issue, the argument that BP should have used a liner is premised on the theory that hydrocarbons flowed up the outside of the production casing. At trial, however, the parties generally agreed that hydrocarbons entered the production casing near its bottom and flowed up the inside of the casing. Therefore, the production casing versus production liner issue appears to be of little relevance.
ii. Running the Production Casing 85. The operation to run the production casing into the Macondo well commenced on
April 18 at approximately 3:30 a.m. and concluded a day-and-a-half later on April 19 at around 1:30 p.m.
86. The evidence reflects that there was a significant amount of debris in the well when the production casing was run on April 18 and April 19. This was due to the fragile wellbore.
87. As the production casing was run down the well, some of the debris flowed up inside the casing.
88. The bottom of the production casing was set into debris at the bottom of the well. This applied substantial compressive force to the production casing, causing it to buckle.
89. The significance of the previous two findings will be explainer later.
E. Overview of Cement Issues
90. As mentioned above, a major step in the temporary abandonment procedure involved setting a cement plug in the bottom of the well; i.e., the production casing cement job. The purpose of this cement was to achieve zonal isolation, i.e., to isolate the hydrocarbon-bearing zones in the formation from the well and prevent hydrocarbons from migrating into the well. It was intended that the production casing cement, once pumped into place and sufficiently hardened, would take the place of the drilling mud as the primary barrier to hydrocarbon influx.
91. No one, not even the cement contractor Halliburton, disputes that the cement failed to achieve zonal isolation. However, there are differing opinions as to why the cement failed.
92. There are two dimensions to the cement issue. One concerns whether the cement pumped down the Macondo well was stable, and if not, whether this was due to an inherent vice in the cement or some outside factor. This focuses on, among other things, the composition of cement slurry; or as the United States' expert Glen Benge put it, the "chemistry" behind oil well cementing. The other dimension concerns how the cement was placed in the well; specifically, whether the cement was placed across the hydrocarbon bearing sands as intended. Mr. Benge referred to this as the "physics" of oil well cementing.
93. Mr. Benge explained that proper cementing takes a combination of chemistry and physics. In other words, the right type of cement must be placed in the right place in order to create a barrier to flow. If either component fails, then the cement will not achieve zonal isolation.
94. The Court will discuss cement placement first and cement composition second.
F. Cement Placement
95. The production casing cement job at Macondo was designed to place "foamed" cement in the narrow space between the production casing and the formation (known as the "annulus") and across the hydrocarbon bearing zones. To do this, cement on the rig had to be pumped down the production casing, out the "reamer shoe, " and then up the annulus.
96. Before cement could be pumped, however, a mechanical device known as a "float collar" had to be converted from a two-way valve to a one-way valve.
i. The Weatherford M45AP Float Collar
97. At Macondo, the bottom 189 feet of the production casing was called the "shoe track." The float collar was located at the top of the shoe track.
98. A float collar is essentially a valve. Its purpose is to prevent unset cement, once pumped into the annulus, from "u-tubing" and flowing back into the casing. The float collar also serves as a landing profile for plugs later sent down the well as a part of the cementing operation.
99. The specific model float collar used at Macondo was the Weatherford M45AP float collar. The M45AP float collar contained two spring-loaded flapper valves. In "unconverted" mode, an internal sleeve called an "auto-fill tube" holds these valves open, which allows fluid to flow through the float collar in either direction (up or down). Shear pins hold the auto-fill tube in place across the valves. When the float collar is "converted, " the auto-fill tube is ejected from the valves, allowing them to close. When properly converted, fluid from above can push the valves open and flow down through the float collar; however, the valves prevent fluid from flowing back up through the float collar.
100. Inside the auto-fill tube was a 2-inch diameter ball. As designed, the ball was free to float the length of the auto-fill tube but could not exit the tube at either end. As the production casing was run down the well in unconverted mode, drilling mud would flow up through the bottom of the auto-fill tube and push the ball to the top tube. Three fingers would keep the ball from exiting the tube, and mud would flow around the ball and out the top of the tube. Once casing was set, the ball would fall to the bottom of the auto-fill tube and land on a ball seat, blocking the relatively large opening at the bottom of the auto-fill tube. Smaller ports in the side of the auto-fill tube would remain open, however. At this point, mud pumped down through the auto-fill tube should exit the tube through these smaller ports. These ports restricted flow through the tube, but flow should not be blocked entirely. The intended result is that the restriction would create a pressure differential across the auto-fill tube; i.e., there will be greater pressure inside and above the auto-fill tube (before the restriction) than there is below the auto-fill tube (past the restriction). As the pump rate increases, so does the pressure differential, until the shear pins holding the auto-fill tube give way and the tube is ejected from the valve openings.
101. Weatherford's specifications stated that converting the M45AP float collar required circulating drilling mud at a rate of 5 to 8 barrels per minute ("bpm") to create a flow-induced pressure differential of 500 to 700 psi.
102. Figures 5 depicts a M45AP float collar being properly converted:
ii. The Attempted Conversion of the Float Collar
103. Successful conversion of the float collar is a prerequisite to commencing cementing operations.
104. BP was responsible for determining whether the float collar converted and whether the cement job should commence.
105. The operation to convert the float collar at Macondo started on April 19, 2010 around 2:30 p.m., after the production casing was run into the well.
106. The float collar could have been converted prior to running the casing down the hole. In fact, BP's own "best practices" called for converting the float collar before running it across any hydrocarbon bearing zones. However, BP was concerned that running the production casing with the float collar converted might create high surge pressures, that could further damage the fragile rock formation. Consequently, BP decided not to convert the float collar until after the production casing was fully run.
107. The decision to run the production casing with the float collar in unconverted mode was not unreasonable given the circumstances, at least not when viewed in isolation. In fact, the M45AP float collar is specifically designed for "[p]ressure sensitive formations and close tolerance annuli, where surge reduction or fast running speeds are desirable." However, running the casing with the float collar in unconverted mode increased the risk that debris from the well would flow into and plug the reamer shoe, the float collar, or both. BP could have mitigated this risk by using a shoe filter, which is designed to filter cuttings and other debris that could possibly plug the float equipment. In fact, Weatherford's specifications stated that the M45AP float collar "should be run with a Weatherford Mud Master filter shoe." BP did not use any type of shoe filter.
108. When BP first attempted to convert the float collar, the rig crew could not circulate mud at all-mud was pumped down the well but there was no return mud at the surface. Pressure on the casing also started increasing during the attempted conversion.
109. The only explanation given at trial as to why mud would not circulate is that debris blocked the flow path. However, the parties disagree over exactly where the blockage or blockages occurred. The evidence reflects that on April 19, 2010, BP personnel involved with the Macondo well also believed debris blocked circulation, although they would not have known for certain how many blockages existed or exactly where the blockages were located.
110. The Court finds that debris blocked the flow parth at two points, if not more: the float collar and the reamer shoe. Debris likely flowed up and around the auto-fill tube while the production casing was being run down the hole in unconverted mode. As Halliburton's expert Dr. Gene Beck pointed out, the smallest hole in the flow path, and therefore the easiest to clog, was the auto-fill tube. The reamer shoe's ports were likely clogged when the production casing was pressed into debris at the bottom of the hole.
111. When mud circulation could not be initially achieved, BP directed the rig crew to repeatedly increase and then bleed off the pressure in the well-a process BP called "rocking"- which was intended to clear the debris.
112. Nine attempts were made over the course of two hours to clear the blockage and convert the float collar. Each attempt, with the exception of the sixth, used greater pressure than the last. During the sixth attempt the pressure was kept the same as it was for the fifth attempt, but the pump rate was increased from 1 bpm to 2 bpm.
113. During these nine attempts the pump rate never rose above 2 bpm. BP chose to keep pump rates low because it was concerned about damaging the fragile rock formation.
114. Circulation was achieved on the ninth attempt with 3, 142 psi and a flow rate of 1 bpm. This is approximately five times the pressure and less than one quarter the flow rate called for in Weatherford's specifications.
115. Even after circulation broke, BP never directed the rig crew to pump at more than 4 bpm.
116. When circulation broke on the ninth attempt, there was a rapid depressurizatiob from 3, 142 psi to about 150-200 psi. After circulation broke, the circulating pressure was significantly lower than predicted.
117. After noticing the rapid depressurization and/or the low circulating pressure, BP personnel expressed concern:
118. A Halliburton cement engineer, while standing on the rig floor, overheard BP Well Site Leader Bob Kaluza state "I need to make a phone call. We may have blown something higher up in the casing. "
119. In a post-incident interview with BP investigators, Bob Kaluza was recorded as saying "[m]y opinion is that after it sheared the flow came back real quick. I said Wow look at how much fluid we got back.' Halliburton had modeled that at 4bbls/min pressure should be 570 psi. Ramped up in 1 bbl increments slowly to 4bbls/min at 350 psi. I said that is odd you guys this is very low. .'... Switched pumps from number 3 to number 4 took 205 psi to break over then at 4 bbls/min had 390 psi. That was an anomaly. I discussed it with [BP Wells Team Leader] John Guide and Keith Dagle. John said pump cement."
120. Brian Morel, a BP drilling engineer who was on the HORIZON at the time, wrote in an e-mail dated April 19, 2010, "Yah, we blew it at 3, 142, but still not sure what we blew yet. "
121. Mark Hafle, BP's senior drilling engineer in Houston, wrote in an e-mail dated April 19, 2010, "Shifted at 3, 142 psi. Or we hope so. We are [circulating] now."
122. BP did take some steps to investigate these anomalies.
123. BP directed Transocean to circulate mud with a different mud pump, in case the low circulating pressure was due to an issue with the pump. Circulating pressure increased after switching pumps, however, it still remained significantly lower than expected. Bob Kaluza's statement above confirms that the circulating pressure was still viewed as an anomaly even after switching pumps.
124. BP also considered the possibility that there was a leak in the system. BP thought that the diverter tool might be leaking, but was able to confirm that it was closed and not leaking.
125. After the diverter tool was eliminated as a possible leak, BP consulted with M-I about the circulating pressures. M-I's response to BP's inquiry is reflected in an e-mail dated April 20, 2010 at 3:34 p.m. (the day after the attempted float collar conversion), where M-I's Doyle Maxie wrote:
I have gone through my inputs for VH [virtual hydraulics] for the modeling I did for circulating prior to cementing casing. I have tried several different inputs, and the closest I can get is 480 psi and that is taking out the fann 70 data which is not giving a reasonable estimate of true pressure. I have had several individuals double check and critique my inputs and still cannot explain the difference. Looking to predictions from the past modeling the pressure is never quite the same when we are drilling so I would not expect them to be the same for this model. Pressure is one of the hardest numbers to correlate. I would be interested to see what Landmark would predict as circulating pressures. I am open to sit down and discuss the inputs with the team. John and I went through some scenario[s] this morning and we could not do any better than 480 psi. 
Mr. Maxie's e-mail was sent to numerous BP personnel involved with the Macondo well, including John Guide (Wells Team Leader), Brett Cocales (Senior Operations Engineer), Mark Hafle (Senior Drilling Engineer), and Brian Morel (Drilling Engineer).
126. Although Mr. Maxie's e-mail notes that modeled and actual pressures are not always the same and that pressure is "one of the hardest numbers to correlate, " the fact remains that the actual circulating pressure was lower than the modeled pressure. Furthermore, although several individuals reviewed Mr. Maxie's calculations, no one could "explain the difference" between the modeled and actual circulating pressures. Therefore, M-I's response to BP's inquiry did not eliminate the possibility that there was a leak. Moreover, BP did not have Mr. Maxie's response until the afternoon of April 20, after BP determined the float collar had converted and instructed Halliburton to pump the cement.
127. BP concluded on April 19 that the float collar converted based on the fact that circulation was established. When asked on direct examination why he believed on April 19 that the float collar converted, BP Wells Team Leader John Guide responded, "By the fact that we could circulate."
128. Under the circumstances, achieving circulation was consistent with converting the float collar. However, achieving circulation did not establish that the float collar actually converted.
129. BP never verified whether or not the float collar actually converted.
130. After the cement job was pumped, a "float check" was performed, but it did not verify whether the float collar actually converted. During a float check, pressure in the well is bled to zero and fluid from the well monitored for a period of time. If the flow does not stop, it is an indication that the valves in the float collar have not converted and the heavier fluids in the annulus are flowing back into the casing. However, a reliable float check requires sufficient differential pressure in the annulus to lift the plug that has been forcibly set on top of the float collar. All of the experts who considered the topic agreed that the differential pressure during the float check was too low to lift the plug.
131. As the United States' expert Glen Benge explained, BP could have verified whether the float collar converted by having the drill crew attempt to reverse circulate; i.e., pump mud down the annulus and up the casing. If the float collar had converted, the drill crew would not be able to reverse circulate, because the closed flapper valves would prevent mud from flowing up through the float collar.
132. BP did not attempt to reverse circulate the well.
133. Without verifying whether the float collar converted and without resolving whether something was "blown" or why the circulating pressure was low-other than to conclude, the next day, that the predicted pressures must have been incorrect-BP instructed Halliburton to commence the cement job.
iii. The Float Collar Did Not Convert
134. As explained above, the auto-fill tube must be ejected from the valve openings in order to convert the float collar.
135. Halliburton's expert, Dr. Gene Beck, opined that the float collar experienced a mechanical failure when circulation was achieved at 3, 142 psi. Dr. Beck testified that debris inside the production casing probably settled in and around the float collar, packing off the auto-fill tube. This held the auto-fill tube in place across the valves and prevented the shear pins from shearing. According to Dr. Beck, on the ninth conversion attempt the ball seat at the bottom of the auto-fill ruptured and the ball inside the tube ejected, but the auto-fill tube remained across the flapper valves. This would allow fluid to pass freely through the float collar, but the flapper valves would remain open; i.e., unconverted.
136. Bill Ambrose, who was in charge of Transocean's incident investigation, testified that the Transocean investigation team similarly concluded that the float collar was clogged with debris and that the float collar never converted.
137. Dr. Beck's opinion relies in part on testing performed during Transocean's post-incident investigation. This testing reflects that the auto-fill tube could be ruptured and the ball ejected at pressures substantially less than 3, 142 psi. During one test the ball shot out the bottom of the auto-fill tube at 1, 477 psi. During a second test the ball shot out of the tube at 1, 840 psi.
138. Dr. Beck's theory also relies on the fact that the pressure dropped rapidly from 3, 142 psi after circulation was achieved. According to Dr. Beck, the rapid depressurization was a signature of mechanical failure, not of debris unplugging. Dr. Beck explained that the pressure response for clearing blockage is a gradual change from high to low pressure as debris is cleared and circulated out. This description is consistent with testimony from Transocean's expert, Calvin Barnhill, when he testified about events that occurred on April 20.
139. BP's counsel pointed out that some of the components on the float collar Transocean tested were different from those used at Macondo. BP also contends that Transocean's test did not allow the auto-fill tube's shear pins to activate. BP concluded that Transocean's test was not representative of actual conditions. Bill Ambrose essentially responded that the internal components of the float collar were the same and/or the differences were irrelevant for purposes of the test.
140. Brent Lirette, a Waterford employee, testified that it is not possible for the ball inside the auto-fill tube to be ejected while the auto-fill tube remained in place.
141. BP also conducted tests on the float collar after the incident. These tests reflect that the M45AP float collar would convert at 3, 142 psi without damaging the float collar. The float collar tested on behalf of BP used the same components as the float collar used at Macondo.
142. Halliburton points out that BP's tests did not account for the effects of debris packed inside the float equipment. Dr. Beck testified that "[O]nce you plug things up... all bets are off on the performance of shear pins.... When you have debris around mechanical devices... the performance of that device is not always as designed.... I've had this happen before where shear pins don't shear when they're supposed to when... you're working in an unclean wellbore environment."
143. Data from the negative pressure test that was conducted on the afternoon of April 20, 2010, tips the balance in favor of the theory that the float collar did not convert. By the end of the Phase One trial there was general agreement that hydrocarbons entered the casing at some point below the float collar and then flowed up the casing. Although Weatherford's float collar is not marketed as a barrier to flow, the M45AP float collar, once converted, is rated to withstand 5, 000 psi differential pressure from below. Weatherford's Brent Lirette testified during cross examination that he would expect the converted float collar to withstand 5, 000 psi differential pressure from below after the float collar converted. Transocean's post-incident tests showed that the flapper valves, when properly converted, could hold at least 3, 000 psi from below. During the negative pressure test, 1, 400 psi of pressure registered on the drill pipe, which was above the float collar. Given that hydrocarbons entered the casing below the float collar, the 1, 400 psi pressure during the negative pressure test must have communicated through the float collar in order to register on the drill pipe. Therefore, if the float collar had converted, it is unlikely 1, 400 psi would have registered on the drill pipe.
144. The Court finds that the float collar did not fully and properly convert on April 19, 2010, or anytime thereafter. On the ninth conversion attempt, the float collar experienced a mechanical failure. The most likely point of failure was the ball seat at the bottom of the auto-fill tube.
iv. The Shoe Track Breached During the Attempted Float Collar Conversion
145. As discussed above, Dr. Beck believes the rapid drop in pressure from 3, 142 psi is a signature of mechanical failure, not of debris unplugging. Dr. Beck opined that two mechanical failures occurred. The first failure was the float collar, discussed above. Dr. Beck also testified that a second failure occurred in the shoe track (i.e., the casing below the float collar).
146. As mentioned above, the circulating pressures were lower than expected after circulation was achieved at 3, 142 psi.
147. Dr. Beck opined that the rapid depressurization and lower-than-expected circulating pressure, when viewed together, demonstrate that rather than circulating mud through the three small ports at the bottom of the reamer shoe, the rig was actually circulating mud through a larger breach or opening in the shoe track below the float collar.
148. This opinion is consistent with the testimony of Bill Ambrose, who stated that the lower-than-expected circulating pressure indicated that "something has opened up in a larger geometry" in the path of circulation. Similarly, Mr. Barnhill proposed in his report that the low circulating pressure indicated "[p]ossible damage to casing shoe track."
149. Dr. Beck's opinion is also consistent with testimony by United States' expert Glen Benge, who stated:
[I]t was acknowledged that whenever that 3, 100-odd psi - when it suddenly released, that's a big concern - that's a big risk because at that point you do not know where in the well you're circulating. You've got a sudden pressure surge, and pressures were much lower, showing that some restriction that you used to have isn't there anymore.
[I]f you broke something and the circulating pressure is lower, you're not really sure of where you're circulating.
Operationally you could have had a break in that well somewhere, and you're circulating at a much higher point.
150. Notably, Bob Kaluza's statement on April 19, "We may have blown something higher up in the casing" shows that the Well Site Leader was concerned that the casing may have breached on the ninth attempt. Brian Morel's statement on April 19, "Yah, we blew it at 3, 142, but still not sure what we blew yet, " arguably reflects a similar concern.
151. Returning to Dr. Beck's theory, Dr. Beck believes that, up until the float collar mechanically failed, the debris blockage in the float collar/auto-fill tube prevented pressure above the float collar from communicating below it. When the float collar failed on the ninth attempt, the 3, 142 psi of pressure that had been isolated at and above the float collar, suddenly transmitted to the casing below. Dr. Beck explained that the rapid depressurization from 3, 142 psi indicates that a sudden pressure surge or shock wave was imparted on the casing below the float collar at this time. This is consistent with Glen Benge's testimony quoted above, particularly his reference to pressure "suddenly releas[ing]" and "a sudden pressure surge."
152. In response to questions during cross examination about how it was possible for 3, 142 psi to breach casing that was rated to withstand substantially higher pressures, Dr. Beck explained that the casing was already in a state of extreme stress when the pressure surge occurred. Dr. Beck testified that when the production casing was run down the wellbore and set into debris,  as much as 140, 000 pounds of compressional force was applied to it. This caused the casing to buckle. Dr. Beck explained that buckling magnified the stress due to compression by a factor of 1.5 to 2.5, because the casing is bent in addition to being compressed.
153. Dr. Beck's testimony about buckling is consistent with e-mails dated April 16, 2010, between BP engineer Brian Morel and Halliburton employee Preeti Paikattu. There Mr. Morel asked at what point the production casing would buckle if it were landed on top of an obstruction. Ms. Paikattu responded that the casing would buckle if 30, 000 pounds of compressional force was applied to the casing, and that this buckling would occur low in the casing.
154. Dr. Beck concluded that the sudden pressure surge or shock wave that was released on the ninth conversion attempt, combined with the extreme stress already imposed on the casing, was sufficient to breach the shoe track below the float collar.
155. The Court agrees with Dr. Beck's opinion and finds that a breach or opening occurred in the shoe track during the ninth attempted conversion. consistent with, and thus provides an explanation for, the data recorded after circulation was achieved-most notably the lower-than-expected circulating pressures. Contrariwise, the Court finds BP's apparent explanation for the low circulation pressures-that the predicted pressures must have been incorrect-unpersuasive.
157. Furthermore, a breach in the shoe track is also consistent with, or at least is not incompatible with, other evidence and testimony regarding cement placement and how hydrocarbons entered the well, as explained below.
v. Cement Was Pumped Through the Breach in the Shoe Track and Placed Improperly; Hydrocarbons Later Entered the Well Casing Through the Breach in the Shoe Track
158. Pumping the production casing cement began at 7:30 p.m. on April 19, 2010, and finished around 12:30 a.m. on April 20, 2010. Approximately 60 barrels of cement were pumped, 48 barrels of which were "foamed" cement.
159. One purpose of the shoe track was to assist in placing cement in the annulus during the production casing cement job.
160. The production casing cement job was designed to pump unfoamed "cap" cement down the well first, followed by foamed cement, and then unfoamed "tail" cement. As planned, the cap cement and the foamed cement would travel down through the float collar and shoe track, exit through the three holes in the reamer shoe, and then head up the annulus between the casing and formation.
161. If properly conducted, foamed cement should have been placed in the annulus across the hydrocarbon bearing zones. The cap cement would be in the annulus above the foamed cement. A small amount of tail cement would also be in the annulus below the foamed cement. Tail cement would fill the shoe track.
162. Figure 6 illustrates the intended placement of cement fluids in the annulus and shoe track and across hydrocarbon-bearing zones (yellow) and brine-bearing zones (blue):
163. Dr. Beck testified that if there was a tiny breach or opening in the shoe track, a significant amount of the cement would exit the casing through that point rather than through the reamer shoe as intended. If the breach was greater than one square inch in diameter, virtually all of the cement would pass through the rupture point.
164. Dr. Beck further testified that cement pumped through the rupture would immediately turn up the annulus; no cement would be placed below the rupture point. Dr. Beck also testified that a breach in the shoe track would prevent cement from being placed in the shoe track.
165. Dr. Beck believes that cement was pumped through the breach in the shoe track, rather than through the holes in the reamer shoe. Consequently, cement was not placed across the hydrocarbon-bearing sands below the rupture point. Therefore, the hydrocarbon-bearing sands below the breach in the shoe track were left exposed to the well and had unrestricted access into the casing through the breach in the shoe track.
166. The Court agrees with Dr. Beck's theory.
167. As already noted, a rupture in the shoe track is consistent with the sudden drop in pressure from 3, 142 psi and the lower-than-expected circulating pressures.
168. Dr. Beck also testified that the blowout occurred relatively quickly, which indicates that there was not much restriction to flow. Dr. Beck's theory regarding cement placement and the flow path of oil is consistent with the rapid blowout and little restriction to flow.
169. Dr. Beck pointed out that his theory is also consistent with evidence from the relief well, which did not encounter hydrocarbons when it intercepted the upper annulus of the Macondo in September of 2010. He testified that the absence of hydrocarbons in the upper annulus shows that the cement placed above the rupture point did set up and provide a barrier to flow. Dr. Beck further explained that, if cement had been properly placed in the annulus and the shoe track, one would not expect some of the annular cement to set up while the rest of the lower annulus and shoe track did not. The Court agrees that the absence of hydrocarbons in the upper annular section is certainly consistent with Dr. Beck's theory. However, the Court notes that this is not as persuasive as the other evidence supporting Dr. Beck's theory. For instance, the relief well evidence is also consistent with the competing theory that hydrocarbons entered the casing through the ports in the reamer shoe.
vi. The Court Is Not Persuaded by BP's Theories Regarding Float Collar Conversion, Cement Placement, and Flow Path
170. BP's position is that the float collar converted properly, the shoe track was not breached, and cement was properly placed in the annulus and shoe track. Furthermore, BP (and the private plaintiffs, Alabama, and Louisiana) believe hydrocarbons traveled through the foamed cement in the annulus, down the annulus, entered the casing through holes in the reamer shoe, through the tail cement in the shoe track, through the converted float collar, and up the casing.
171. BP believes that the foamed cement was unstable, which permitted hydrocarbons to migrate from the formation through the cement, down the annulus, and up into the casing through the reamer shoe. BP further proposes that the unfoamed, tail cement placed in the shoe track failed to stop the influx either because it was contaminated by the nitrogen from the foamed cement or because the hydrocarbons encountered the shoe track cement before it had an opportunity to harden. BP believes the float collar had converted, but failed to stop the hydrocarbons because a float collar is not intended to provide a barrier to hydrocarbon flow.
172. BP's theory that hydrocarbons entered through the reamer shoe is primarily based on the work of its expert, Morten Emilsen. Using sophisticated computer software, Mr. Emilsen created a model of the well based on certain known data. Mr. Emilsen then ran hundreds of computer simulations that made different assumptions about unknown data, such as flow path. The modeled results were compared with certain recorded data from the well (e.g., drill pipe pressure). If the modeled results for a particular scenario did not match the recorded data, then Mr. Emilsen concluded that the scenario likely did not occur. When Mr. Emilsen assumed that the flow path was through the holes in the reamer shoe and that 13 to 16.5 feet of the formation was exposed to the well bore-what he termed "Case 7"-the modeling closely matched much of the known data. Based on this match, Mr. Emilsen concluded, among other things, that the flow path was "through a leaking casing shoe and up through the inside of the casing."
173. Despite running hundreds of simulations, Mr. Emilsen never ran a simulation that assumed the flow path was through a breach in the shoe track below the float collar. Because he had not investigated that scenario, Mr. Emilsen could not state whether a model that assumed flow through a breach in the shoe track would also produce results matching the recorded well data. However, Mr. Emilsen did indicate that flow through the shoe track would probably have little effect on at least some variables. Moreover, Mr. Emilsen could not exclude the breached shoe track as a plausible theory, given that he did not run this scenario through his modeling program.
174. Mr. Emilsen made a number of assumptions about the conditions at the bottom of the well. These assumptions cause the Court to question the accuracy of his conclusion insofar as it concerns the precise location and manner hydrocarbons entered the casing.
175. In short, Mr. Emilsen's work does little to discredit the theory that hydrocarbons entered the casing through a breach in the shoe track. Mr. Emilsen's testimony does not prove that hydrocarbons entered through the reamer shoe, nor does it disprove that hydrocarbons entered through a breach in the shoe track.
176. BP's theory regarding flow path, etc., ignores or dismisses the data recorded after circulation was achieved on the ninth conversion attempt, particularly the lower-than-expected circulation pressure, which, as noted, indicated that a larger opening was created in the flow path. BP's theory also requires a succession of failures. While there is ample evidence to support the idea that the foamed cement was unstable, and therefore probably would have failed even if properly placed in the annulus, there is less evidence to support the notion that the tail cement in the shoe track also failed, and even less evidence that the float collar, if converted, would not have stopped pressure from communicating during the negative pressure test. Yet all of these things must occur in order for BP's theory to be correct.
177. To be fair, Dr. Beck also proposes what could be viewed as an unlikely chain of failures (production casing buckled, float collar clogged, auto-fill tube ruptured, shoe track breached). This comes as little surprise, though, given that BP's entire explanation for this tragedy and overarching theme at trial was that there was a series of failures. The difference is that there is more support for Dr. Beck's chain of failures than there is for BP's chain of failures.
178. To conclude, the Court finds that when cement was pumped on April 19 and April 20, most of it exited the casing through a breach in the shoe track below the float collar, rather than through the reamer shoe. Consequently, no cement was placed in the annulus below the breach point, and little or no cement was placed in the shoe track below the breach point. When hydrocarbons later entered the casing on April 20, they flowed through the breach in the shoe track.
179. Given this conclusion, the court further finds that BP's cementing program violated 30 C.F.R. § 250.420, which required it to, inter alia, "[p]revent the direct or indirect release of fluids from any stratum through the wellbore into offshore waters."
vii. Cement Bond Log
180. After a cement job is pumped, a cement evaluation technique such as a cement bond log ("CBL") can be used to evaluate whether zonal isolation was achieved. A CBL uses acoustic signals and associated software to ...