ANALYSIS:MOST REPORTING of BP's Gulf of Mexico catastrophe has concentrated on the oil slick rather than its technical causes.
Last February, the 8,000-ton rig Drilling Horizon floated into position 80km off the US coast to drill exploration well Macondo in 1,522m of water. Designed for waters far too deep for anchors, the rig kept its position using satellite positioning and computer-controlled multidirectional thrusters. Including helicopters, boats, fuel and specialist services, the operation cost BP a cool million dollars a day. By April, Macondo had found oil and gas some 5,500m beneath the sea. The rig was therefore to be disconnected pending availability of production facilities.
Wells are drilled using rotating pipes which circulate a heavy “mud” that controls pressures and removes cuttings. As the well progresses, a succession of concentric tubes, “casing”, is lowered into the hole, with cement pumped into the annulus between it and the hole to secure upper layers. The next section is drilled at a smaller diameter, secured with a narrower casing, and so on, telescope fashion.
At ground level a “wellhead” provides additional sealing of the annuluses between the concentric casings. During drilling, a blowout preventer (BOP) is bolted to the wellhead to control unwanted oil or gas influx entering the well.
Weighing 200 tons and standing 12m high, it has up to six hydraulic valves which, to shut off the well, clamp around – and can slice apart – any pipe sticking through it. Typically, with the BOP sealing off the annulus, heavier mud is pumped into the well to overcome the pressure below until it can safely be opened and operations resumed.
In deep water, the BOP, activated from surface, sits atop the wellhead on the seabed. The floating rig kilometres above connects to the BOP by a 50cm-diameter “riser”, which provides access to the well.
On Macondo, the final 12cm-diameter casing was run to bottom, cemented by contractor Halliburton through a non-return valve at its foot and up the annulus, and hung off in the wellhead. This would leave the well ready for another rig to reconnect later.
On April 20th, once the cement had set, the BOP was closed to pressure test the cement, casing, wellhead and BOP. Only two steps remained before disconnection: to pump a plug of cement into the well as an additional safety barrier against untoward pressure from below (it is unclear whether this was done) and to replace the heavy mud in the riser with seawater before removing the riser.
Once full of seawater, the 1km-long riser exerted 200 atmospheres less pressure on the top of the well than when full of heavy mud.
When the BOP was opened, this reduction allowed an enormous bubble of high-pressure gas that had inexplicably accumulated to burst into the riser, expelling the huge column of seawater out of the riser and 70m into the air. This lowered the pressure in the riser even further, which sucked in ever greater quantities of first gas then oil.
With no wind to disperse it that calm night, the gas quickly spread across the rig until inevitably a spark somewhere ignited it. A huge fireball ensued and although 115 people managed to evacuate, 11 men died.
Fuelled by uncontrolled gas and oil, the conflagration raged for 36 hours until the rig sank. In the immediate chaos, no one got to the control panel to close the BOP. The riser and control lines broke free of the rig and fell haphazardly onto the seabed where the riser continues to spew the oil that is causing such environmental alarm.
BP’s efforts are now focused on using unmanned submarines to try to close the BOP, funnelling the escaping oil into a cofferdam, containing the slicks on the surface and drilling a relief well to intersect Macondo at reservoir level.
How had gas built up in the wellbore in an apparently sealed casing? Faulty casing? Poor cement? Lack of plug? Defective equipment? Inadequate procedures? Insufficient expertise? Organisational dysfunction? Human error?
These are big questions which only a formal inquiry can answer.
Tony Allwright is an engineering and industrial safety consultant. He blogs at tallrite.com/blog.htm