BP's High Risk Solution
Inserting a pipe into the BP/Horizon riser would be in the right direction to stopping oil flow. Failing that, we fear that BP is being forced to cut all the debris away from the top of the BOP and install another BOP atop the first, with suitable transition. Frankly, we believe that should have been done much sooner, but the risk is that to do so would actually increase oil flow. Only a Monday morning quarterback would second-guess BP in this decision.
The Geometry is the Problem, Stupid!
For the idea of diverting the flow using the riser, the problem is that the riser is a tortuous 20 inch (i.d.) pipe with 1" walls. Worse, the 7" production pipe is still in the riser for its whole length and well beyond. A very small flow of oil/gas through the 7" pipe was stopped earlier by having ROVs install a block valve on the end of that pipe. Now, the best BP has been able to do is to insert a smaller 4" line into the riser with flexible collars or diaphragms attached, but this "siphon" has produced a fraction of the leaking oil/gas, probably a bigger share of gas than of the noxious oil. Simple hydraulics shows that the siphon is very limited because it cannot seal against the wall of the riser, with our pesky 7" pipe in the way. Where is Kafka when we need him?
Back at the top of the BOP, the riser is twisted so badly near where it sits on top of the BOP that several leaks occurred (it looks like as many as 4, now). This part of the riser is weakened. If we succeeded in inserting a new pipe tightly into the "top" end of the riser, we would be concerned that these cracks would widen under any increase in differential pressure. Even the use of somewhat abrasive drilling mud in the "top kill" runs the risk of widening these cracks and increasing the rate of leakage. Our idea has that covered.
But the best bet of stopping flow before August seems to us to be to wrap or weld over the 4 upstream leaks, either before or after we solve the geometry problem at the top end of the riser, and either stop the flow at the top end with a block valve or (preferably) insert a sealed "pipe" and direct the well product to the surface through any of many ways available to BP. The latter is the clear winner since it does not run the risk of rupture of the riser, creating very little extra back-pressure on the weakened walls.
Inserting a Flexible Pipe with Wall Sealing
There are 3 principals which we will employ which allow this system to work. First, we have developed various ideas about making deep water piping systems out of very high performance braided fiber (we call it superfiber) cylinders, coated with polymeric systems which protect and seal the resultant tubes. Second, we use adhesives and stitching of fabric-to-fabric to create seams which work very well at pressures encountered in deep water E&P operations. Third, we employ clever design and the principals of Judo to actually use the power of the high pressures in the well to fight the leak.
Our plan is in stages.
Reinforce or even stop the 4 leaks near the BOP (can be done later)
Cut the 7' line near the end of the riser: reinstall block valve (optional)
The two images below shows how we build and install an expandable pipe to grip the riser's inner surface and transition to either a block valve or the surface. We braid a pair of cylinders of super-fiber like Kevlar® and coat them with marine-grade, flexible urethane such as used on marine hawsers. Outer cylinder is (say) 30 feet long and diameter sized to fit against the 20" inside diameter of riser. Outer surface should have embedded grit or even "thumbtacks" through the fabric to provide grip to inner surface of riser.
The inner cylinder is slightly smaller ... gap is not important ... and we use stitching and adhesive to connect inner cylinder to the outer cylinder at each end. Here's an important innovation: we glue and stitch the two cylinders together longitudinally at several points around the circumference. This forms an elongated torus with space enough between the inner and outer walls. And there are several longitudinal pressure chambers between the plies. Install a flange around the upstream end to protect this double-walled pipe as we thread it into the riser around the pesky 7' pipe. There is a bigger flange around the downstream end, where it will hook to a block valve. In the annular space between the pipe layers, we install some small diameter plastic rods or pipes to make the flexible pipe rigid in the longitudinal direction without preventing it expanding to fit within the riser. This allows the ROVs to thread the flexible, double-walled pipe as far as possible into the riser, surrounding the 7" pipe.
How Does This Work?
Pressure at this level is approximately 151 atmospheres. Inside the riser is somewhat higher. Pressure relieve valves (PRVs) are installed in the inner cylinder walls, set to open at pressures above (say) 155 atmospheres. This will allow material and pressure to enter the annular chamber. When in place, we use a moderate head seawater pump to increase the pressure insider the annulus to (say) 200 atmospheres, which will increase the outer diameter because we have designed the braided, coated cylinder to expand under such differential pressures. The grit or high tech thumbtacks will try to enter the walls of the riser, thereby increasing the friction required to displace the pipe from the riser. Remember, our pesky 7" pipe is actually inside the annular pipe. When expanded to meet the walls, our pipe will divert all flow into itself.
At the other end of our annular pipe is a traditional metal flange; we have invented proprietary means of transitioning securely from the flexible pipe to the flange. Next comes a suitable block valve--preferably low pressure drop. Then the pipe can connect to BP's various ways to pipe to the surface.
Here is where we draw upon the principals of judo, allowing the flowing well pressure to help us defeat the well flow. With PRVs installed in the inner wall of our annular tube, any higher pressure inside the riser will allow gas or oil (or even drilling mud or water) to enter the annular space, equilibrating the pressure. The flexible tube will try to expand even further. The higher pressure will act to push our designer thumbtacks through the wall of the super fiber tube, assuring that the pipe is not displaced. Thus, the higher the internal pressure provided by the oil/gas formation, the better our seal.
While BP cannot be accused of being overly informative, we have heard that measured well pressures like 10 or 11 Kpounds/SqIn existed before the accident. If true, we know that if we closed our block valve without any more pressure reduction from drilling mud (top kill is under way as this is written), the worst case scenario is for the pressure in the riser to approach 10,500 psi (about 715 atmospheres). With 151 atm outside, the differential pressure across the riser would approach 564 atmospheres in this case. Once the 20" diameter flexible pipe was outside the riser, it would experience that same differential pressure; we can design for this but would rather not, as it will increase the lead time greatly. Therefore, we have allowed for using this approach and the block valve to shut off flow and do a top kill. With no net flow of oil/gas upward, this would be a relatively easy proposition. But we do not like the risk/reward ratio.
How Would We Use the Product to Stop the Leak Soonest?
Our preferred method is to design the flexible pipe for much lower differential pressure, install it with an open block valve, prebolted to a line to the surface, with extra long flange bolts, allowing a gap after the block valve. When the expandable flex-pipe is snaked into the riser by the ROV, the oil gas flow will be mostly diverted through the tube, through to low delta P block valve and leak out around the still open flange. Once the tube has been inserted as far as feasible and the high head seawater pump pressures it up to seal and hold, our ROVs would torque down the flange bolts downstream of the open block valve and virtually all flow would be diverted to the waiting surface ships without putting any great strain on the rise from added pressure.
At that point, BP and the Feds are back in the driver's seat with the ability to evaluate the risk/reward of closing the block valve and doing a top kill vs. continuing to route the oil/gas to the surface normally while continuing the relief well. Our instinct says that if the flow can be diverted with little risk of reverting before August, BP will elect to continue to relief well (do they still need 2?) and not risk a top kill which could damage the somewhat fragile riser. Isn't it ironic that a 20 inch diameter seamless tube with 1 inch walls is considered fragile?
Ability to Execute
We believe that our affiliates can design and braid the tubes we describe. Materials are readily available from DuPont and generic competitors. While some of this design work has been done, some remains. Most of the rest of this design is off-shelf and proven.
We have not even SWAGed a cost, but know that it would be among the least expensive of actions facing BP. My German craftsman grandfather, who would be amazed by the audacity of this idea, always taught me to spend my attention on getting the right tool for every job. This is the right tool.
Now, let's demand that some of this same kind of thinking be applied to how to prevent a mess like this from recurring.