A Theory About What Went Wrong
Given my operational background, I’m intrigued by the Gulf of Mexico disaster, because I know that it is an event that should have never occurred, particularly if all proper procedures had been followed. I’ve read and seen all sorts of theories about the cause from various experts and even a few nutjobs (the most interesting being that BP had actually drilled into an alien refinery buried beneath the sea). Setting the nutjobs aside, there are lots of sound ideas about what went wrong; with most of them, including my own early thoughts, centering on a catastrophic BOP failure.
However, after reading several internal BP documents and Transocean’s account of the event, I now think the BOPs did work briefly; that is, they tried to close, but they couldn’t because there was 9 7/8 inch casing blocking them, a pipe size for which they were not configured to handle.
BP had made several changes to the well design in the days leading up to the disaster, including changes to the casing program - making the decision to change from a liner which would be set from about midway down the hole to the bottom, to a long string of pipe running from the bottom of the well all the way to the well head on the ocean floor. Though the new design would be cheaper and faster to set, it did have additional risks in that, because it was made up of two different sizes of pipe, 9 7/8″ and 7″, and because of poor bottom hole conditions that had previously caused them to lose circulation, it was going to increase the risk of not getting a successful cement job. In fact, while contemplating the new design, BP personnel had exchanged a number of emails expressing concern that they may not be able to get the cement to flow up behind the pipe (in the annulus) to a level that it would completely cover the producing formation. Without this coverage, the formation would be exposed, allowing hydrocarbon flow that would be creating building pressure behind the casing. That pressure would then be pushing against the seal at the wellhead where the 9 7/8″ casing was connected, called the wellhead hanger seal assembly. Casing is essentially “hung” in the wellhead, meaning that it locked in the wellhead under tension, and that tension keeps it in place and also applies force that seals the joint where the casing and the wellhead meet.
I think what ultimately happened with this well is that the cement job failed to fully cover the hydrocarbon producing formation, and may have even failed to adequately set-up at the bottom of the hole because the decision was made to hang the casing off without allowing sufficient time for the cement to cure. According to Transocean documents, the the nitrogen foamed cement that BP used (because of the difficult downhole conditions) may have needed up to 48 hours to set. Records from the rig indicate that BP allowed only 18 hours before they set the casing in the wellhead hanger - meaning after they pulled tension on the casing and locked it in the wellhead, it may have broken free from the cement down hole, relieving some or all of the tension that had been applied. Adding to the problem, BP never ran a cement bond log which would have told them that the cement job was bad. They did have contingency plans to run the log if there were indications the cement job wasn’t successful, but because the cement job appeared to have proceeded pretty much as they planned it, they made the assumption that it had worked.
Once BP ordered the crew to start circulating the mud out of the hole, the seawater that was displacing the mud essentially lightened the casing, allowing it to float and start to move up hole (think of it as a very long bottle floating in a sea of mud - if the bottle if full of the same mud, it will sink. But, if you remove some of the heavy mud and replace it with lighter water, the bottle will start to float). Once the casing started to “float”, it released from the wellhead hanger, breaking the seal and allowing the hydrocarbons from the exposed formation to start flowing out of the well. In fact, I think the casing may have broken loose from the hanger with enough force that it was propelled up into the BOP, blocking or jamming the rams that could have shut in the well. At this point, the last line of defense would have been the annual BOP’s that set atop the BOP stack. However, for some reason yet unknown, they also failed to close, either due to mechanical failure or just lack of time. If the casing had suddenly released from the wellhead hanger, the flow up the riser pipe to the rig would have been almost explosive, allowing the rig crew little time to react.
If my theory is correct, it could be very difficult to kill this well. In order to kill a blowout via a relief well as BP is currently working toward, you have to ensure that you intersect the blown out well at the right point with the relief well. If you are too low, below the producing formation, you are not going to be successful. I believe that ideally, BP wants to intersect right near the top of the producing formation, allowing them to redirect the pressure up the new well and plug the old well from the top; or if its not possible to get into the old well due to trash or collapsed pipe, the can pump (under high pressure) large volumes of mud into the well and hope to overcome the formation pressure. However, if the well is flowing behind the casing, its going to be more difficult. If thats the case, I’m unclear how they would do it, but it would no doubt require massive volumes of mud pumped under high pumped pressure in order to kill this well.
BP is currently drilling two relief wells. I originally thought the second well a contingency - if something went wrong with the first well, they would still have a chance to kill the well as soon as possible with the second relief well. However, it now appears to me that BP is also concerned that the well could be flowing behind the casing and intend the other relief well to intersect the well bore up hole, ensuring that they will have a chance to kill the well regardless of whether the flow is coming up the casing from the bottom of the well, or from behind the casing, up hole from the bottom of the well.
We won’t really know what happened until the blowout is controlled and BP can re-enter the well to determine the cause. I do hope my theory is wrong because if I’m right, this could be a very difficult blowout to deal with.