House of Commons - UK Deepwater Drilling - Implications of the Gulf of Mexico Oil Spill

UK Deepwater Drilling - Implications of the Gulf of Mexico Oil Spill - Energy and Climate Change Contents

Annex 1—Chronology of the Deepwater Horizon Incident

The Macondo Well in the Gulf of Mexico

BP started drilling the Macondo well on 7 October 2009, using the Transocean owned Marianas platform. Hurricane Ida damaged this platform on 9 November 2009, and so BP and Transocean (who operated the platform under contract to BP) replaced the Marianas with the Deepwater Horizon, which began drilling on 6 February 2010. Transocean charged BP approximately $500,000 per day to lease the rig, plus contractor fees.[215] BP aimed for the drilling to take 51 days, with an estimated cost of $96 million. It was expected that the Deepwater Horizon would be leaving as early as 8 March 2010, but the Macondo well took longer to complete than anticipated. By 20 April—the day of the blowout that killed 11 workers and injured 17—the rig was 43 days late, which would have cost an extra $21million in lease fees alone.

BP owns a 65% interest in the Macondo well. The US-based company Anadarko Petroleum owns a 25% share, and the Japanese company Mitsui owns 10%. The Deepwater Horizon drilling platform (an exploration rig, not a production rig) was owned by Transocean, who also operated it for BP. The objective of the drilling operation was to "successfully evaluate any commercial hydrocarbon [oil and gas] [...] discovered".[216] The oil and gas reservoir was located at over 5,596m below the seabed, with the wellhead at a water depth of over 1,500m.

The Macondo project yielded one of the largest finds in the Gulf of Mexico, but the crew repeatedly struggled to maintain control of the well against powerful "kicks" of surging oil and gas. In evidence published by the US House of Representatives Energy and Commerce Committee, BP staff described Macondo as "a nightmare well that has everyone all over the place", just six days before the Deepwater Horizon platform exploded.[217]

The Deepwater Horizon Drilling Rig

The Deepwater Horizon was a semi-submersible, mobile offshore drilling rig built for Transocean by Hyundai (South Korea) in 2001. It flew a Marshallese Islands' flag of convenience. "Semi-submersible" rigs are kept afloat and upright by watertight pontoons located below the surface and beneath the waves, and are usually used in water depths greater than 200m where floating fixed structures are not practical. The Deepwater Horizon was dynamically positioned, which meant—rather than using chains or wire to anchor it in place during drilling operations—that its position was computer controlled using underwater thrusters.

The drilling rig was capable of operating in harsh environments and water depths of nearly 2,500m (upgradeable to over 3,000m). While drilling the Macondo well it was operating in just over 1,500m of water. In 2009, before work began on the Macondo well, Transocean crews working with BP discovered oil in the giant Tiber field in the Gulf of Mexico. At a total depth of approximately 10,685m (in an ocean depth of 1,259m) it was the deepest oil well in the world.

The Deepwater Horizon rig was due for a series of extensive maintenance checks late in 2010, with records indicating it was last checked thoroughly in 2005. Documents from Transocean's maintenance department indicated various asset deficiencies including "intermittent alarms [on the control panel] on unrelated functions when opening [a valve on the blowout preventer]", and "low pressure readings" in the hydraulic system.[218]

Blowout prevention devices are designed to handle a range of well control problems, and often come fitted with several different types of rams, giving engineers flexibility in their response. The blind shear ram is described as the ultimate fail-safe device, crushing and sealing the well pipe as a measure of last resort. The Deepwater Horizon had a single blind shear ram located inside the 15.5m tall blowout preventer stack at the wellhead on the seafloor. With a single blind shear ram, there is a risk that it could close on one of the extremely strong joints that connect the sections of drilling pipe, and be unable to collapse it. Such a risk is minimised by the use of two blind shear rams.

Transocean hired West Engineering to carry out a physical assessment of Deepwater Horizon's well control system, but they were unable to access the blowout preventer (BOP) as it was on the seafloor. This meant they were unable to verify whether the blind shear ram on Deepwater Horizon's BOP could shear through drill pipe and seal off the well while in deepwater. A 2009 industry study entitled Pull Your BOP Stack - Or Not? calculated the price of stopping operations to pull up a blowout preventer for repairs at $700 per minute.[219]

BP's internal investigation of the Gulf of Mexico oil spill culminated in the Deepwater Horizon Accident Investigation Report (the Bly Report), published 8 September 2010. The full details are not yet known, but it appears that gas and oil rushed up to the wellhead on the sea floor, and the blowout preventer (BOP) device was unable to contain the pressure. According to Dr Tony Hayward, Group Chief Executive of BP: "[the Deepwater Horizon incident] arose from an interlinked series of mechanical failures, human judgments, engineering design, operational implementation and team interfaces".[220]

Factors Identified as Contributing to the Incident

CASING

Deepwater wells are drilled in sections. The basic process involves drilling through rock, installing and cementing casing into place—casing lines the well—to secure the wellbore (well hole), and then drilling deeper and repeating this process. One day before the blowout—while preparing the well for future production at a later date—BP decided to install a single long-string casing from the top of the well to the bottom, rather than multiple individual casings with a seal (known as a "liner" with a "tieback"). Mr Richard Cohagan, Managing Director of the oil and gas exploration company Chevron UK, told us: "BP was designing the well so that they could use it as a production well and that's one reason that they had a long string in the well [...] We tend to have larger [...] and multiple physical barriers".[221] Mr Cohagan went on to argue that: "BP was trying to design the right well for their conditions".[222]

Multiple individual casings would have provided more barriers to the flow of gas up the well in the event of a blowout, but would have taken longer to install and been more expensive. A BP-plan review in mid-April recommended against the single casing as it would make the seal at the wellhead the "only barrier" in the event of a failure.[223] Dr Hayward told us: "The decision to run the long-string was actually based on long-term integrity [...] a liner with a tieback [...] is subject, over time, to degradation and can leak".[224]

When the final string of this single casing was installed, one key challenge was making sure the casing ran down the centre of the well bore. If this is not done properly, it becomes difficult to displace drilling fluid from the narrow open space around the casing, which in turn will lead to an inability to cement the casing in place properly. In such an instance, it is possible that channels will form in the cement that allow gas to flow up the open space around the casing. Centralisers are attachments that go around the casing to centre it in the borehole. Halliburton, the cementers, recommended using 21 centralisers on this final string of casing, but BP decided to use six. The Bly Report makes the case in Key Finding 2 that this decision is unlikely to have contributed to the incident.

BP aimed for the drilling of the Macondo well to take 51 days, at an estimated cost of $96 million. It was expected that the Deepwater Horizon would be leaving as early as 8 March 2010, but the Macondo well took longer to complete than anticipated. By 20 April—the day of the blowout that killed 11 workers—the rig was 43 days late, which would have cost an extra $21 million in lease fees alone.[225]

CEMENT

Despite Halliburton's and BP's own predictions of a gas flow problem caused by an incomplete cement job, BP decided not to run a 9-12 hour procedure known as a "cement bond log" to assess the integrity of the cement seal, dismissing the Schlumberger contractors who had been hired to undertake the test.[226] This acoustic test would have determined whether the cement had bonded to the casing and surrounding formations. If a channel that allows gas to flow up is found, the casing can be perforated and additional cement injected into the annular space to repair the cement job. Key Finding 1 of the Bly report discusses BP's belief that the cement mix designed by Halliburton was unfit for purpose. We were told by Dr Hayward: "we know the cement was not good because we had influx into the well".[227] Dr Hayward added: "I think we need to be cautious until we can complete [...] [an] analysis [of the cement] to understand why the cement failed".[228]

As Halliburton refused to provide samples for testing, the BP investigators had an independent laboratory analyse the design of the cement slurry.[229] BP noted that there was a high percentage of nitrogen found in the cement ingredients, making it difficult for the cement to form a stable "foam slurry".[230] The cement used was injected with nitrogen to make it into a lighter "foam".[231] This is done in order to avoid damaging the rock formation of the reservoir, which would make it more difficult to produce oil at a later date. BP says that when the independent laboratory tried to produce a representative cement sample—based on the slurry design—they could not demonstrate cement stability. Therefore, BP concluded that the foam slurry likely experienced "nitrogen breakout" resulting in channels forming that would have allowed oil and gas to flow through it.[232]

While drilling into high-pressure, high-temperature fields like the Macondo, the well is usually filled with heavy drilling fluid (known as "mud") while drilling to compensate for the upwards pressure of the oil and gas in the reservoir. It is recommended that this drilling mud is fully circulated from the top to the bottom before commencing the cementing process. This allows the mud to be conditioned—by removing any pockets of gas and other debris safely—so that the cement is not contaminated. BP decided against the full 12-hour procedure and only partially circulated the mud.[233]

The choice to use a single string of casing meant the Macondo well had just two barriers to gas flow up the annular space around the final string of casing: the cement at the bottom of the well and the seal at the wellhead on the sea floor. Insufficient centralisers also meant that there was a severe risk that the cement job would fail, and the lack of a cement bond log meant that BP were unable to check this. Finally, BP did not deploy the casing "lockdown sleeve" that would have prevented the seal from being blown out from below.[234] Even when cemented in the wellhead, under certain pressure conditions the casing can become buoyant and rise up, creating an opportunity for oil and gas to break through the wellhead seal and enter the riser to the surface. The lockdown sleeve prevents this.

NEGATIVE PRESSURE TEST

One of the Bly Report's key findings was that readings taken during the "negative pressure test" to determine well integrity indicated that there was a flow of oil and gas from the reservoir into the well even though the "Transocean rig crew and BP well site leaders" thought the test was a success and well integrity had been established.[235] Dr Hayward told us: "we know that with the benefit of hindsight that the negative test was erroneously interpreted".[236] This test simulates the temporary abandonment of the well after drilling and prior to production, when a proportion of the well is displaced to sea water. BP's Group Head of Safety and Operations, Mr Bly, added: "There are records of the information that would have been available, so we know that [information on the drill pipe pressure increasing, when it should have been decreasing] was there. We can't explain why they didn't see it".[237]

This series of decisions may have been driven by expense and time, as by 20 April—the day of the blowout—the rig was 43 days late, and would have cost BP at least an extra £21 million in lease fees alone. However, each decision and failure increased the risk of a blowout.

Exemplifying the industry's inability to take account of high-consequence, low probability events, Dr Hayward told us: "we weren't prepared".[238] Mr Cohagan, of Chevron UK, told us: "Deepwater Horizon gave us a new perspective on how bad things could be".[239] We are concerned that the offshore oil and gas industry has failed to prepare for what they had previously classified as worst-case scenarios.

BP's Attempts to Kill the Macondo Well

On 22 April, two days after the blowout and subsequent explosion that killed 11 workers, the Deepwater Horizon drilling rig sank. This bent the 1,500m "riser" pipe connecting the rig to the wellhead on the sea floor. Submersible robots discovered two leaks close to the seabed. Over the next few days, BP attempted to activate the single blind shear ram in the blowout preventer (BOP), a device located at the wellhead on the sea floor. The blind shear ram would have severed and sealed the pipe, but attempts to activate it failed.

On 2 May, BP began drilling relief wells, intending to intersect the existing well in order to send down heavy drilling mud and cement to stop the leak. By 7 May BP had constructed a 12m tall containment dome, known as "top hat". They attempted to lower the dome on to one of the largest leaks from the bent pipe, but it became clogged with an icy mix of gas and water (called gas hydrates). After several unsuccessful attempts, BP inserted a mile-long tube into one of the leaks on the broken riser pipe on 16 May, and succeeded in siphoning off some of the oil to a ship on the surface, collecting an estimated 22,000 barrels a day, over nine days. This siphon was cut off on 26 May as BP attempted its "top kill" and "junk shot" operations. "Top kill" attempted to overcome the pressure of the rising oil by pumping drilling mud into the top of the well, while "junk shot" attempted to clog up the BOP by injecting objects such as golf balls. These attempts failed.

On 31 May BP cut the damaged pipe away from the BOP and lowered a dome—connected to the surface by a new riser—on to the blowout preventer. Methanol and warm seawater pumped down the riser prevented the formation of icy gas crystals, and oil and gas were funnelled to a ship on the surface. An additional siphon supplemented the system on 16 June, pumping more oil to surface vessels. By mid-July BP had four vessels on site to collect and process retrieved oil and gas, collecting 62,000 barrels per day. On 10 July, BP removed this cap and replaced it with a new device, containing many of the same features as a BOP. The hydraulic rams on the new cap were closed on 14 July, and pressure sensors indicated that oil was not leaking from elsewhere on the seafloor.

BP began its "static kill" operation to plug the well on 4 August 2010. Drilling mud pumped from the surface forced the oil back down the well, and cement was then sent in through the top of the well to seal it off. The final "bottom kill" procedure—where cement was pumped through relief wells into the Macondo—took place successfully on 18 September.

BP's Gulf of Mexico Clean-up Operations

A team of experts assembled by the US National Incident Command (NIC) announced on 2 August that an estimated total of 4.9 million barrels of oil had been released from the Macondo well. The National Oceanic and Atmospheric Administration (NOAA) then determined what had happened to the oil. It is estimated that burning (5%), skimming (3%) and direct recovery from the well (17%) removed a quarter of the oil released. Another quarter naturally evaporated or dissolved, and just under a quarter was naturally (16%) or chemically (8%) dispersed. Dissolution is the process by which the oil dissolves into the water, whereas dispersion is the process by which larger volumes are broken down into smaller droplets. Residual oil made up just over a quarter of the oil spilt. Residual oil is a combination of categories all of which are difficult to measure or estimate, and includes oil: that is on or just below the surface as "light sheen" and "tar balls"; has been washed ashore or collected from the shore; or is buried in sand and sediments. It is thought that dispersed and residual oil will be naturally degraded. Response efforts addressed 33% of the oil spilled.

Dispersants are chemicals that can be used to break up and speed the natural degradation of oil on the surface. It is argued that they are less harmful than oil and biodegrade more quickly than untreated oil. In the Deepwater Horizon spill the dispersants were used underwater to prevent more oil from reaching the vulnerable marshes, wetlands and coastlines of the US Gulf states. BP was pre-authorised to use approved dispersants, according to the US Environmental Protection Agency, on spills no closer than three miles from the shore, but was required to get daily permission from the U.S. Coast Guard during the clean-up operations for this incident. Dispersants are usually used on the surface, but BP injected them into the oil as it flowed from the well. BP began by using the dispersant Corexit 9527a, and then switched to Corexit 9500. Both of these products were removed from the UK Marine Management Organisation's approved list in 1998, as they proved too toxic in instances where they might end up on rocky shorelines (although existing stocks could be used).[240]

Booms are temporary floating barriers used to contain oil by concentrating it into thicker surface layers. Exclusion booming is used to keep oil away from sensitive areas, while diversion booming is used to direct the flow of oil elsewhere. Containment booms are deployed in a "u" or "v" shape to direct the flow of oil to a recovery resource, such as a skimmer. "Skimmer" is a common name for any device (usually attached to a ship) used to remove oil (or an oil/water mixture) from the surface without using chemicals. In-situ burning is a method of burning freshly-spilled oil while it is floating on the water.

Environmental Impacts in the Gulf of Mexico

The Macondo well is estimated to have leaked 4.9 million barrels of oil, making it the largest marine spill in US history. The full extent of the impact on the environment is not yet known. As of 16 August 2010, more than 7,000 birds, sea turtles and dolphins have been found dead or debilitated in the Gulf of Mexico since the oil spill began.[241] While a majority of the dead were not visibly oiled, scientists have yet to determine why they died. However, it has been confirmed that more animals are dying than during the same time in previous years. Not all injuries or deaths were necessarily caused by the oil spill, and some of those found dead may have been oiled after death. The higher than expected numbers of animals found dead may have been an artefact of the increased monitoring of the area.

More than twice the number of stranded sea turtles have been found than normal at this time of year.[242] Of the nearly 500 found visibly-oiled, the majority were found alive. Of the nearly 600 found not-visibly-oiled, the majority were found dead. Some suspect that shrimp fishermen may be causing the increased deaths by not using devices that prevent turtles trapped in nets from drowning (whilst the federal agencies are distracted). More than 50% of one batch of turtle corpses analysed showed evidence of drowning.[243]

Of the more than 2,300 birds (mostly pelicans) found not-visibly-oiled, all were dead, compared to about half of the 3,800 found visibly-oiled.[244] When ingested or inhaled, oil can cause brain lesions, pneumonia, kidney damage, stress and death. There have also been reports of dolphins acting as if they were drunk, and it is suspected that disorientation caused by oil exposure is making them more susceptible to boat strikes.

215 Fleet Status Report, Transocean, 13 April 2010, www.deepwater.com Back

216 BP Macondo Prospect Well Information, September 2009, http://energycommerce.house.gov Back

217 BP Email, 14 April 2010, http://energycommerce.house.gov Back

218 Deepwater Horizon Rig Assessment, http://documents.nytimes.com/documents-on-the-oil-spill?ref=us#document/p19 Back

219 Jeff Sattler (West Engineering Services), "Pull Your BOP Stack - Or Not? A systematic method to making this multi-million dollar decision", SPE/IADC Drilling Conference and Exhibition - Amsterdam, 17-19 March 2009 Back

220 Q 89 (Hayward) Back

221 Q 251 Back

222 Q 252 Back

223 Letter to Tony Hayward, US House of Representatives, 14 June 2010 http://energycommerce.house.gov Back

224 Q 98 Back

225 Fleet Status Report, Transocean, 13 April 2010, www.deepwater.com

Back