Finding Facts or Fault

I devoted several hours yesterday to watching Senate hearings on the Gulf Coast oil spill. The Energy and Natural Resources Committee hosted two panels, one a technical panel featuring a former official of the Minerals Management Service--the agency that Interior Secretary Salazar has announced he intends to split in two--and a Professor of Petroleum Engineering from Texas A&M. The second, juicier panel was composed of senior executives from the three main companies involved in the spill, BP, Transocean and Halliburton. Despite the importance of these hearings in putting a face on this disaster and giving our elected representatives an opportunity to demonstrate their concern, I thought the panels served a useful educational purpose. And somewhat to my surprise, they also turned up at least one apparently new fact that might prove crucial in understanding what went wrong 5,000 feet below the Gulf of Mexico on April 20th.

I can't claim to be a great connoisseur of Congressional hearings. They offer some of the same morbid fascination as a car wreck: you know you shouldn't be watching, but you can't take your eyes off it. True to form, a few of the Senators treated the session as an opportunity to show their outrage and alignment with their constituents' concerns. Most, however, followed the tone set by the Chairman, Senator Bingaman (D-NM), in asking thoughtful, probing questions--though I couldn't help chuckling when one Senator seemed to imply that she had participated in the 1986 Space Shuttle Challenger hearings in that same room--alluding to their famous "O-ring" revelation--even though she would have just been elected to her state's legislature that year. Despite the obvious frustration of the committee members when the three executives deflected their efforts to pin the blame for the accident on each of them in turn, the discussion remained civil and the comments and questions mostly substantive.

I found two lines of questioning especially intriguing. The first related to the cause of the accident itself--as distinct from the subsequent leak--and whether it might have had something to do with the well having been cleared of drilling mud prior to setting the final concrete plug in the well. As I understand it, drilling mud is used to balance the pressure in the well between the higher reservoir pressures deep underground and the much lower pressure at the surface. Once the heavy mud was removed and replaced with lighter seawater, the barriers installed in the well (steel casing, cement, the first plug, and ultimately the blowout preventer, or BOP) would have had to withstand the full pressure in the reservoir, which Dr. F.E. Beck from the first panel estimated at 14,000 psi. Since this was apparently the last action performed by the drilling crew prior to the explosion, the sequence and timing of this step makes it an obvious candidate for one of the root causes leading to the explosion on the topsides of the Deepwater Horizon rig. Senator Sessions (R-AL), in particular, tried in vain to get any of the three witnesses to concur that it was contrary to normal practice for the mud to be displaced prior to the setting of the final cap.

The other fascinating exchange occurred later in the hearing, at about 1:24 into C-SPAN's archive video, when the ranking member, Senator Murkowski (R-AK), questioned Transocean's CEO, Mr. Newman, about reports that Deepwater Horizon's BOP had been modified. According to Mr. Newman, one of the five "ram" preventers on the BOP stack was converted "from a conventional well-bore-sealing ram preventer to a BOP test-ram", to "allow for more efficient testing of the BOP." He went on to explain the economic benefits of such a modification, which apparently has been done on other rigs, in reducing the cost and delays associated with testing the BOP. Unfortunately, although Senator Murkowski followed up with a question about whether any modified BOPs had experienced incidents, she didn't ask whether that modification had reduced the capability of the BOP to respond to a catastrophic failure of well control.

Perhaps I've misunderstood Mr. Newman's remarks, and the modification would have had no impact at all on the operation of the BOP. Or it's possible that one extra ram might have made no difference at all, in conjunction with the cascade of other failures necessary to produce a blowout of this magnitude. However, this certainly seems like a topic that should be examined in much greater depth during the full incident investigation that must follow.

I didn't have time to catch the afternoon hearings, in which the same executives were grilled by the Senate Environment and Public Works Committee, or the House hearings this morning. I'll be interested to see if any other new insights emerge, though a couple of things seem clear. First, and with due respect to the Senators and their staffs who clearly worked hard to get up the steep learning curve on this subject, they are simply not equipped to conduct an engineering investigation into an accident of the technical complexity involved in deepwater drilling. Moreover, the format and adversarial approach aren't well-suited to eliciting the necessary level of candor and cooperation from witnesses who've essentially been told they are auditioning for the role of chief villain in the piece. If anything, that understandable tendency to prioritize blame-apportionment over impartial fact-finding seems to have been amplified by the financial crisis and recession. But while it's easy to write such hearings off as political theater, they can still serve a useful purpose, because that same lack of technical knowledge on the part of these committees constrains the dialog to a level that the average American actually has a chance of understanding. That makes it all the more essential that the Congress should refrain from leaping to premature conclusions that could turn out to be wrong, but very hard to correct later with the public.

Disaster Scenario?

As the consequences of the ongoing oil leak in the Gulf of Mexico unfold, it's still not clear what we're facing. I've seen repeated requests by those affected to know what the worst-case scenario might be. I can't blame them, though when we don't even know how much oil is leaking, and with so much uncertainty surrounding the measures that BP and the US government are pursuing to plug the well and mitigate the spill, the range of possible scenarios becomes very wide. And if these near-term effects remain unpredictable, the potential longer-term implications for US energy policy are even more divergent. Until the well is capped and the full scope of the environmental and economic damage known, we can only guess at the future shape of this component of our energy supply.

In trying to imagine the range of outcomes, we must consider the rate at which the oil is flowing, how long it will flow, and the relative success of efforts to recover or break down the oil that has leaked before it reaches the shoreline, fisheries and other sensitive environments. A high-end projection of the volume of oil spilled might involve a leak that is actually well above the current 5,000 barrel-per-day (bpd) estimate, and that continues as long as it took to cap last year's Timor Sea leak: 10 weeks. Even at 5,000 bpd, that would eclipse the total spilled from the Exxon Valdez before it ended. And if oil were leaking much faster, as some estimates suggest, the result could rival the largest oil tanker spills, such as the Amoco Cadiz in 1978, while still falling short of the 1979 blowout of Pemex's Ixtoc-1 well farther south in the Gulf. We may never know the true extent of the spill, because there's no accurate way to measure the quantity of oil currently flowing from 5,000 ft. down in the Missississipi Canyon.

If that's the far extreme, what might a less dramatic scenario look like? As described in this morning's New York Times, BP is pursuing several approaches that could either shut off the well quickly, or at least contain the leakage until the well can be sealed by means a relief well drilled into the same formation to block the flow to the current well. The critical event following the explosion on the drilling rig was the apparent failure of the blowout preventer, which BP has been attempting to activate by remotely-operated vehicles (ROVs.) The BOP was supposed to cut off the errant flow--literally. The quickest solution would be to set a new BOP in place and activate it to crimp the riser and drillpipe and shear them off. From my limited understanding of the techniques involved, doing this at depths like these, using only ROVs, and with a well that might be blowing gas and oil at much higher rates once the bent riser and drillpipe were removed would be extremely challenging. BP's plan to use "domes"--essentially underwater cofferdams--to contain and siphon off the oil as it comes out of the well could be nearly as tricky to pull off, though with less downside if it failed. If any of these techniques worked, the total volume of the spill might be limited to something under 150,000 bbls, assuming that the well has already leaked 50-60,000 bbls. That would still qualify as a very large oil spill--much larger than early estimates projected--though far short of a true worst-case.

For now the efforts of all the oil and oil-service company personnel, the Coast Guard, and other military and civilian government personnel involved--along with those whose homes and livelihoods are affected--are properly focused on addressing the leak and its direct consequences. In the interim, the rest of us have had some time to think about what all this means in a broader context. Although I would argue that any permanent changes in policy would be premature, it's not too early to think about what should happen once the leaking well is capped. None of my readers will be surprised to learn that I disagree strongly with those calling for a permanent halt to offshore drilling anywhere in the US. At the same time, I believe most observers agree that the Deepwater Horizon accident raises serious questions about the technology and practices involved in drilling at such depths. This morning's Wall St. Journal cited a 2004 study questioning the efficacy of at least some of the blowout preventers that have been used in deepwater installations, and various reports have pointed to requirements by Brazil and Norway that offshore drillers install equipment enabling the BOP to be activated remotely, should the drilling vessel lose direct communication with it. Both issues should be revisited, in light of current events.

Until the Deepwater Horizon rig and the BOP on the well are ultimately recovered from the sea bottom and analyzed, we won't know exactly what caused the accident that led to the spill. That could take a year or more. Meanwhile, drilling continues on other rigs in the Gulf of Mexico. Even with President Obama calling a temporary halt to expanded drilling beyond the Gulf, that leaves a number of other blocks on the Gulf's Outer Continental Shelf that have been leased but not yet drilled. What standard should the government apply, when it receives applications for new drilling on these? (That could even include the lease encompassing the Macondo prospect, which is demonstrating its resource potential in the least-desirable manner imaginable.) Unfortunately, at this point we have nothing beyond the event itself and the previous, uneventful completion of thousands of similar wells (and many thousands of wells in shallower water) to gauge the probability of this ever happening again. If the administration opts for a hiatus in Gulf of Mexico drilling or an outright ban, that would have far-reaching economic and energy-security consequences that I will address in a subsequent posting.