Halliburton Wearing White Hats?
While it is in the Houston Chronicle‘s business section, it’s an AP story: Missing data causing rig reconstruction mystery.
I don’t like to pull quotes or anything else from the AP, but it covers some important points in the story and fills in a lot of background. While you have to skeptical that a lot of information is coming from someone who represents Halliburton employees, the reality is that Halliburton has the data to back up their claims, while the data from BP and Transocean after 3PM went “down with the ship”.
The basic point is that Halliburton had sensors on the well that transmitted data back to the shore. This is one of the services that Halliburton sells to clients, but wasn’t being used by BP or Transocean. Halliburton installed them to monitor their cementing of the well. BP and Transocean apparently did periodic batch dumps of data to shore.
Halliburton admits that they saw a pressure surge, which would endanger the cement job [many people have opined that the curing cement caused the surge], and told BP and Transocean. BP and Transocean say that they conducted another test and the pressure was normal, so they continued to replace the heavy drilling mud with sea water.
The problem for BP and Transocean is that this well had a history of “kicking”, sending high pressure bursts of gas up the riser for some time, and that will show up in the record. This was testified to at the Coast Guard/MMS hearing in Kenner by the crew members of the Damon B. Bankston.
What also comes out is that the Deepwater Horizon people would get a bonus if they finished early. That sounds like the reason for replacing the mud with sea water before the final cement plug was in place, rather than waiting and doing things in the normal order.
I don’t know the exact composition of the product that Halliburton was using, but I have poured some concrete over the years, and while it might looked solid after 24 hours, it takes a good deal longer to achieve full strength. This would have been some form of hydraulic cement, which is used for underwater applications. I’ve used it on concrete ponds to seal damage. It sets quickly, but you don’t subject it to unequal pressure for a few days or it will crumble. Twenty hours is a short time for cement to set, even Gorilla Glue tells you to wait 24 hours.
6 comments
Indeed, even “quick set” cement will be at less than 1/4th of its final strength at 20 hours… you want days, not hours, between doing a concrete pour and putting a load on it. When they built my house in Phoenix they waited a over a week between the pour and putting tension on the tendons (for the post-tensioning of the slab) and the foundation guys did a strength test sampling first before doing the tensioning and signing off on the slab. These BP guys were fuggin’ NUTS to start pulling mud at less than 20 hours with only a single plug in the hole…
— Badtux the Astounded Penguin
BTW, I worked for a competitor to Halliburton’s oilfield services. We hated each other, but Halliburton didn’t get to be the biggest name in oilfield services by not doing their job — they certainly aren’t the cheapest, that’s for sure, that’s the only way we stayed in business (by undercutting them in price). I’m inclined to believe what Halliburton is saying here, unless their oilfield services division has gone way downhill since my days in the oilpatch…
– Badtux the Rememberin’ Penguin
.-= last blog ..Out of kindness, I suppose =-.
The thing is, Halliburton knew the sensors were in place, meaning anything that happened would be recorded, so it wouldn’t be in their interest to try to lie about it.
It looks like the promise of possible bonuses trumped sanity.
Bryan, that sensor monitoring works via a radio connection that can “go on the blink” lickity split if the operator so desires. Some of the sensor monitors actually happens via regular old cell phone connections! My brother was called out to a rig to see why the data kept getting interrupted at random times. What he found was a box that had a cell phone modem in it… and that occasionally rig workers in the know would come to that box, unplug the cell phone modem, plug in their own handset, and call home to talk to the wifey or girlfriend or both (heh). Granted, this guy was too far out into the Gulf to be doing that, but point is that these radio connections are *not* reliable, and have a tendency to “go on the blink” whenever something is about to happen that the operator doesn’t want monitored…
Here Halliburton didn’t, err, “blink”, their connection because they were doing CYA. They’d told the drill crew that the well was burping. The drill crew ignored them. So be it. Been there, done that ;). (I.e., made sure everything was being recorded and that my objections were on file so that when the inevitable disaster happened, it was gonna be someone *elses* ass, not mine, that was fired).
– Badtux the Oily Penguin
BTW, reading the specs on quick set and reading the article you linked to, it appears that the Halliburton cement job was subjected to twice the amount of pressure that it could withstand at 20 hours, but if they’d waited for a week, they would have been in the clear. I.e., quick set at 24 hours has 1500psi of compressive strength, and the last pressure shown by Halliburton was 3500psi. No way, no how, was a 20 hour old cement job going to withstand that sort of pressure, and it wasn’t Halliburton’s decision to continue pumping drilling mud out of the well (and thus allow pressure to build). So yeah, looks like Halliburton is in the clear, if this data is real… they could have done the most perfect cement job ever, and there was just too friggin’ much pressure thanks to the morons who were in too big a rush to pump mud to care about the consequences of putting pressure on a just-placed plug.
So yeah, Halliburton doesn’t look like the culprit here… looks like they were doing real-time CYA, and have the data to prove it. Now if they could just quit electrocuting our soldiers in Iraq and Afghanistan ;). (To be fair, that’s their B&R division, not their oilfield services division, and their oilfield services division basically operates as a stand-along company with no real connection to any other part of the company other than ownership).
I assumed that this was CYA which is why they are letting the legal profession know they have the data, so don’t bother. I have filed more than a few memos and done a lot of recordings for exactly that purpose, i.e. to be sure that everyone knew what was about to happen wasn’t my idea, and I stated my objections.
The only time I did any software work for a law firm, audio tape was a major expense for the job.
As for broken data sets, it was a given that if the surveillance video magically failed during a “convenience store robbery” you would end up arresting the clerk. Yep, BP and Transocean just overlooked sending data to shore. I guess they were just too busy, what with the party going on below deck, and the well getting ready to blow to think about that.