Deepwater Horizon: a scientific response

Newly published papers document how the science community responded to a disaster of unprecedented magnitude in the Gulf of Mexico

Alan Bailey

Petroleum News

With countless thousands of words having already been written about the Deepwater Horizon disaster in the Gulf of Mexico, it may be tempting to think that anything that could be said about this calamitous event has already been put into print. But a series of scientific papers published in early December in the Proceedings of the National Academy of Science puts a particular scientific slant on the events following the April 20, 2010, blowout of BP’s Macondo well, and the response to the subsequent spewing of oil into the waters of the Gulf.

A scientific perspective can perhaps put some objectivity around events that inevitably trigger high levels of emotion.

One of the papers in the Proceedings overviews the scientific findings and experience in the Deepwater Horizon response. This paper is authored by officials from several federal agencies, including U.S. Geological Survey Director Marcia McNutt and Energy Secretary Steven Chu, with Jane Lubchenco, the then administrator of the National Oceanic and Atmospheric Administration, or NOAA, as lead author.

Unprecedented, unprepared

“The situation of the Macondo blowout was unprecedented, with the oil spewing forth into an extreme ocean environment — deep, cold and high pressure — but rapidly spreading to mid-waters, the surface and the atmosphere,” the paper says. “Experience and response methods applicable for other oil spills in many cases proved either impossible to apply or ineffective.”

In particular, BP’s government-approved spill response plan did not take account of the presence of deep, suspended microscopic oil droplets in the seawater, even although the formation of these droplets had been predicted as likely to occur, the paper says.

And the evolving response to the out-of-control well involved cross-agency cooperation between government scientists and the vital involvement of scientists from academia and private institutions, the paper says.

Oil flow rates

“The lack of reasonable estimates of flow rate early on was problematic from the perspectives of both communications and response, but the lack was caused by real uncertainty rather than any attempt to hide information or underestimate numbers,” the paper says. “It is true that much of the response did not depend on knowing the exact rate, but some of it did, particularly the capacity to capture oil directly from the well.”

As the response proceeded, new methods of estimating flow rates emerged. For example, NOAA and academic scientists developed a method of determining which components of the oil escaping from the well actual reached the surface of the sea, thus enabling an estimation of well flow rates by the detection of oil components escaping into the air.

Airborne, surface and subsurface chemical measurements ultimately led to a consistent picture of the dynamics of oil flow, indicating that only about half of the oil and none of the methane gas escaping from the well ever reached the sea surface, the paper says. And echo-sounder imaging of oil droplets in the water, carried out by a surface ship, provided an additional means of estimating oil flow rates.

Ultimately, the scientists estimated an initial oil flow rate of about 62,000 barrels per day, declining to around 53,000 barrels per day that the time the well was shut in.

Final tallies for volumes of recovered oil indicated that 5 percent of the spilled oil was burned, 3 percent was skimmed and 17 percent was recovered directly from the riser pipe from the well, the paper says.

Fate of the oil

Repeated sampling of offshore waters showed that within 19 days of eventually shutting in the well, oil in the water had dissipated to background levels, the paper said. However, sediment sampling revealed grounded oil in deep areas around the wellhead; in deep-water sites to the northeast and southwest of the well; in many shallow coastal areas around oiled marshes; and near some beaches.

The assessment of oil contamination in deep-water animals also pointed to some significant accumulation of oil on sediments, while coral communities, mostly within 20 kilometers of the well, were also impacted.

Weathered oil samples in beach and nearshore environments showed 86 to 98 percent depletion of polycyclic aromatic hydrocarbons, with further depletion to 20 percent of current levels anticipated within five years, the paper says. According to information published by the Environmental Protection Agency, or EPA, polycyclic aromatic hydrocarbons are a component of weathered crude oil that can be toxic, depending on which hydrocarbons are present, and on the level of the contamination.

Dispersants

But people have questioned the potential toxicity of the dispersant chemicals. And the accelerated bacterial action on the oil can reduce oxygen levels in the water, perhaps adversely impacting water-living creatures.

Factors leading to a decision to inject dispersants directly into the escaping oil included a view that this method of dispersant application would require less dispersant than other methods while maximizing the exposure of oil to the chemicals before the weathering and emulsification of the oil occurred. And there would be less exposure of response workers to dispersant chemical and to organic compounds from the oil, the paper says.

On the other hand there were concerns about a lack of understanding of the potential consequences of the dispersant application, the possibility of severe hypoxia in the seawater and the potential for dispersed oil and dispersants to damage subsea flora and fauna.

“Balancing these tradeoffs was not easy, but the potential for more rapid degradation of hydrocarbons was compelling,” the paper says.

Monitoring results

As dispersant application proceeded, repeated water sampling showed a drop in oxygen levels, but not to levels considered hypoxic. And assessments of dispersal effectiveness through the measurement of oil droplet sizes pointed to an increase in the estimate of the volume of oil dispersed from 8 percent to 16 percent of total oil, the paper says.

Subsequent tests on water and sediment samples from nearshore and offshore locations for the most part failed to find dispersant chemicals at detectable levels, and no samples contained chemical concentrations above benchmarks set as acceptable for aquatic life, the paper says.

No biological impact

However, “additional studies are required before a complete understanding of the tradeoffs with the use of dispersants is known, including potential impacts of dispersants, dispersed oil and oil alone on the plethora of other species in the Gulf, especially plankton and juvenile stages,” the report says.

Seafood safety

Of biggest concern was a dispersant chemical called dioctyl sodium sulfosuccinate. New analytical techniques developed during the Deepwater Horizon response made it possible to determine how much of this chemical was present in seafood gathered from the Gulf, thus ensuring that seafood in re-opened areas posed no health risk. And as part of the response a new rapid method of testing for aromatic hydrocarbons was developed. In total more than 8,000 seafood specimens were tested.

“This extraordinary effort to protect the integrity of seafood seems to have been successful: No tainted seafood was reported to have reached the market,” the paper said. “An independent assessment arrived at the same conclusion.”

However, after the sight of oil and gas flowing from the Macondo well and, with images of oil covered shores and birds appearing for weeks on end, many people had difficulty in believing that oil was disappearing from open waters, that fish could metabolize aromatic hydrocarbons and that seafood testing was reliable, the paper says.

Role of science

In the case of the response to Deepwater Horizon, federal and local government officials overseeing restoration efforts decided on a policy of openness and transparency, allowing public access to data that was collected, the paper says.

Although it may take several years for all of the effects of the oil spill from the Macondo well to become apparent, there have already been new scientific discoveries as a consequence of the disaster. For example, the discovery of microbes and sea conditions that lead to the rapid decomposition of hydrocarbons in the water, the paper says.

Recommendations

The paper also says that, with knowledge of oil flow rates being so important to the planning and execution of response strategies, devices that can provide oil flow rates should be installed on any equipment used for the extraction of oil.

And adequate spill response preparation is a key to successfully dealing with an oil spill emergency.

“The importance of preparedness cannot be overstated,” the paper says. “Despite significant advances in technology that allowed drilling in deep waters, comparable progress had not been made in devising methods that would have enabled us to stop the flow from deep wells or deal with a spill of the magnitude seen in Deepwater Horizon. Both could and should have been anticipated.”

The science behind Macondo well capping

After the worst oil spill in U.S. history, BP’s Macondo well was finally capped on July 15, 2010. But the decision process leading to that well capping proved far from easy, given the potential for the capping operation to result in a seafloor oil leak that would have proved much more difficult to deal with than the leak from the well itself.

A paper published in early December in the Proceedings of the National Academy of Science by scientists involved in the well capping decision describes the science behind the decision that stopped the flow of oil from the well.

Leading contender

In May 2010 an operation called “top kill” involved pumping drilling mud down the well. And when this operation failed to stop the blowout BP engineers postulated the possibility of a wellbore breach having allowed mud to escape from the well — the existence of such a breach would likely provide a route for oil to escape from the well following a capping operation, thus increasing the probability of a catastrophic leak.

To evaluate the possibility of an subsurface oil blowout, BP and a well integrity team consisting of scientists and engineers from government agencies and academia agreed on a test involving the temporary capping of the well to enable well pressure monitoring, with a procedure to re-open the well within a fairly short time if pressures remained below a specified level.

In the event, when in mid-July the cap was applied, the oil pressure in the well did climb above a level below which there would have been a clear indication of a well breach. But unfortunately when the pressure subsequently stopped climbing the pressure was still too low to completely rule out the possibility of a breach being present.

So, to ensure that no subsurface blowout would occur, government regulators ordered that the wellbore should be re-opened after 24 hours, the paper says.

Meantime, to account for the observed maximum well pressure attained after the capping operation, scientists plugged reservoir data supplied by BP into a U.S. Geological Survey computer model originally designed to simulate the flow of groundwater through subsurface aquifer rocks. By simulating what would happen were there no well breach, given estimated oil flow rates from the well, the scientists were able to determine that the observed lower-than-expected well pressures following capping were likely to have resulted from oil depletion in the subsurface oil reservoir following the well blowout.

And, given this explanation for the observed pressures, the government allowed the capping operation to continue beyond 24 hours, but with continuous monitoring of well pressure and geophysical surveillance data, and with a re-evaluation of the well capping decision at regular intervals.

The geophysical surveillance included the use of seismic surveys, conducted as frequently as four times per day, to seek early evidence of any flow of oil and gas from the well bore through the surrounding rocks.

With the shut-in of the well extending over several days, new pressure data from the well enabled the near-continuous updating of the reservoir model used to assess the possibility of a well breach. And, with the scientists also refining the assumed reservoir geometry used in the model, the well pressures predicted from the model turned out to be a close match with the pressures measured in the well. It appeared that the well had maintained its integrity following the capping operation.

It subsequently became possible to keep the stacking cap in place until Aug. 2, at which time the use of a relief well enabled the Macondo well to be fully sealed off and cemented.

Success in the capping operation can be attributed to collaboration between the many scientists, engineers and emergency response officials involved; clear protocols for data requests through a well-defined chain of command; the very rapid analysis of diverse datasets; the co-location of government scientists with BP staff; continuous access to required expertise and training; and excellent access to BP’s data and mitigation plans, the paper says.

—Alan Bailey