The blowout preventer was supplied and installed by Transocean, but made by a separate American engineering company. Probably, elsewhere in the world Transocean would have a Briitsh, Dutch or Malayasian company make this item to a common specification. It weighs 140 tons and no-one wants to ship one too far if a competent manufacturer is close at hand. The competence of the manufacturer in this and other cases seems not to be at issue.

However, despite having about half of the drilling contracts in the Gulf of Mexico over the last few years, Transocean account for around three quarters of the accidents, and more than one of those has involved the partial failure of a blowout preventer. The Deepwater Horizon disaster is remarkable because the failure was complete, and it led to a cascade of problems on the rig itself, which is what killed the (11) human beings. Better rig design could have avoided these deaths, if not the oil spill, and Medawar is still somewhat concerned that the oil spill is seen as worse than the cost in human life.

Another fact, with multiple pertinencies for Deepwater Horizon, is that in 2006, Transocean was served with an Improvement Notice by the UK's Health and Safety Executive. This is to the HSE's page, explaining very clearly, exactly what an improvement notice is.

The HSE Inspector can only issue an improvement notice if:
The company is in some way breaking health and safety law (in a way that would allow it to be brought to a criminal trial if an accident had already happened.)
The Inspector can describe what it is the company must do, within a given time, to comply with the law and avoid prosecution.
It's a way of getting companies (and, incidentally, government departments and police authorities) to avoid accidents, before someone is hurt, rather than punishing them afterwards. Failure to comply with an improvement notice will lead to prosecution.

The notice in question related to Transocean's failure to test blowout preventers. Transocean appears to have complied with the notice within the time set by the Inspector and thus complied with UK law. The accident rate in the Gulf of Mexico suggests that this improvement in its performance was entirely local to UK jurisdiction. IE: the company now knew that its normal practice wasn't good enough for the UK, but took no steps to learn from the notice and make their American operation safer too.

I will get onto the technical implications in a moment, but there's an important pertinency of jurisdiction here: The relevant American authorities have been complaining, but not loudly, because Mr Obama wants BP to bear all the blame, that because Transocean is registered in Switzerland and its ships and rigs are registered as maritime vessels in the Marshal Islands, they had no power to inspect them. Actually, they had no duty to inspect them annually, but the US Coastguard as the same powers as the UK's Maritime and Coastguard Agency, to inspect any vessel, as the occasion demands, that is passing through their waters or visiting their ports, for basic seaworthiness. They have the same power to arrest any vessel, too, until defects are made good.

However, the transfer of responsibility for oil-rig inspections from the UK's Department of Energy, to the HSE, as recommended by the Cullen Inquiry into the Piper Alpha Disaster, gives the HSE complete jurisdiction over any rig or vessel carrying out work in UK waters, or on a licence issued by a UK authority. If American authorities had paid attention to the Piper Alpha disaster, other than to assure themselves that no American faced criminal proceedings over the 167 dead, as per the State Department's standard operating procedure, then they would have had nearly a quarter of a century's grace in which to ensure that some Federal or State authority had the same jurisdiction over the offshore drilling industry in American waters as the HSE does in UK waters. That they didn't, may owe something to "industry lobbying" but there is no evidence that BP was a party to such lobbying. BP are not on record as having opposed any regulatory effort to improve safety in the UK or Norway (they have sometimes suggested more constructive alternatives to some ideas) and Medawar suspects that they wouldn't have opposed such efforts in the USA, had they ever seriously transpired, which they did not.

Transocean, on the other hand, has links to a very heavy lobbying outfit run by a former Democratic Congressman, Bill Brewster. This investment seems to have paid off for them, so far.


Now for the pertinent technical implications:

The 2006 improvement notice was over a failure to test blowout preventers. This cannot have been a failure to test that they had been manufactured properly, or the HSE would have served the notice on the manufacturer and not Transocean! However, it is impossible for the manufacturer to ship a blowout preventer to its customers, in a fully-armed state where it would function immediately on arrival. It could not be transported without incident in the "armed" state and it certainly couldn't be safely handled on the rig and installed on the oil (or gas) well in that state.

The installer, Transocean or their sub-contractors, must prepare the equipment for installation when it arrives on the rig, test that they have done this correctly, and then install it. Then they must prepare it for operation after installation, and again, test that this has been done properly and that the device will work. Not carrying out this chain of actions properly was what earned Transocean the improvement notice in 2006.

In practice, what happens is that the equipment leaves the factory with covers fitted over every hole, and with some kind of pin or bolt immobilizing everything that can move.
Some parts may be packed with grease to protect them in transit, some fluid reservoirs may not be filled. It leaves the factory in a state which allows it to be transported and which will protect all its working parts from corrosion or incidental damage.

When equipment is on the rig, as much of this stuff is taken off as can be done safely, and any working fluids are loaded, protective grease is removed and where appropriate, replaced by whatever working lubricant is required. This procedure is documented in detail, and although complicated, is no more complicated than it needs to be. Someone signs a worksheet to show that everything required at that stage has been done. There will be an agreed test procedure that verifies, as far as possible, that everything that has been released so far, will work as intended.

Then the equipment can be lowered into the sea and installed, by divers or robots depending on how deep the ocean floor is. These days, robots are used more and more, even in relatively shallow waters, because there's no such thing as a completely safe deep sea industrial dive. Once the equipment is installed, it can be armed for operation and any remaining covers removed to allow pipes to be connected, etc. Then there's another test procedure before anyone can pack up and go home, or move the rig onto the next oilfield. (Transocean were apparently under pressure from their next customer to finish working for BP and move the Deepwater Horizon onto their oilfield.)

Because this happens towards the end of what may have been a months-long operation, even if the operation over-runs by just a few days, these final preparation and test procedures of the blowout preventer are more likely to be performed under pressure of time than the actual drilling work. This may explain why so many of Transocean's accidents have involved blowout preventers, not because these are necessarily flawed, but because they have to be installed with the end of the job in sight and management openly staring at its watch.

Now, the way in which the Deepwater Horizon blowout preventer failed, completely, in all its stages, isn't consistent with some of the normal causes, such as a foriegn body, defective parts, etc. There are four stages; two, or even three, defective parts do not cause all four stages to fail, and a foreign body would have to be many feet long to affect more than one stage.

If there was an environmental cause, such as temperature or pressure jamming the equipment completely, then this would have been discovered years ago when similar equipment was tested after proper installation at a similar depth. This is another reason why the HSE likes the tests to be done each time: to gather information on how far this kind of equipment can be relied upon in general, as well as to ensure that each installation is as safe as it can be.

Just about the only thing (short of highly competent sabotage) that can credibly immobilize all four stages of a blowout preventer, is something, the only thing, that is designed to immobilize all four stages: the safety devices (bolts) that are fitted at the factory to make it safe for transport and installation.

The blowout preventer failure was key to the whole disaster, and to any immediately effective remedy for the oilspill. Because BP had the whole power and fury of the US Government turned on them the moment they started to ask Transocean to account for this, BP have been working blind ever since, although there are ample circumstantial grounds for them to suspect, that if their supplier, Transocean, made an error, that was where it would be. BP may be held responsible, but it is deeply irrational to both deny them any right to inquire how other companies may have contributed to the disaster, and to make them fix it all none-the-less.

And even if BP have to guarantee that clean up costs will be met, it's not in the public interest to deny them the right to recover some of those costs from other companies that made mistakes, because that just gives those companies, however well represented they are on Capitol Hill, a free pass to commit mistakes and cause accidents in the future.

It is time for BP to act on suspicion, even if the White House is determined to deny them the evidence, and try, so far as is possible, to repeat the post-installation arming procedures for the blowout preventer, that Transocean should have done, regardless of what any surviving worksheets say was done or not. With the oilspill collection cap in place, not all of the stages are still accessible, but even if one stage could be activated, the amount of oil being spilled could be dramatically reduced.

And but for President Obama's heavy-handed, highly partisan and frankly bigoted intervention, that might have been possible a few weeks ago, which would have saved the vast majority of the oil that has been spilled, though sadly, none of the blood.

It is very obviously true that nobody in American politics now actually wants BP to succeed in stopping the spill, by doing something which they might have done a month earlier but for American politics.