Former NASA Investigator Offers Two More Ways for BP to Stop the Oil Spill

Fast Company has received hundreds of ideas about how to fix the gulf oil spill, but this one is getting a bit more attention because it's author earned masters and Ph.D. degrees from MIT in mechanical engineering and worked in the MIT Fluid Mechanics Laboratory. He started his career at Rockwell International, designing auxiliary power units for the hydraulic systems on the space shuttle. He has been Principal Investigator on several NASA Small Business Innovative Research grants. Below, his two proposals for fixing BP's leak.


A way to capture all of the flow and stop the spill is outlined in the Figure A below. The advantages of this solution over BP's planned solution are:

  1. It will capture all of the leaking oil, not "most" or "much" of the leaking oil.
  2. It can be implemented now.
  3. It does not require a new "clean cut." Cutting the pipe may be undesirable because it will increase the leak at least temporarily.

Oil Spill Solution
FIGURE A: Oil Spill Collector

Referring to Figure A above, the leaking oil is pulled from the leaking pipe through a Coupling that is fabricated to conform to the shape of the pipe adjacent to the leak. The oil then pulled through a Flexible Inlet Duct and into the Pump.

The Pump is a key component of this design. The Pump provides the pressure to drive the oil one mile to the surface.

The Pump speed is adjusted based on the pressure measured by the Inlet Pressure Sensor. This pressure sensor measures pressure relative to the local ambient pressure in the sea; it measures the differential or "gauge" pressure. If this inlet pressure is positive, the pressure in the duct is higher than the sea pressure, so the pump is made to work harder to pull in more oil and thereby reduce the inlet pressure. Alternately, this inlet pressure becomes too negative, the pump slows down to bring this pressure back to near zero. In this way the pressure sensor is used to adjust the pump speed appropriately to match the flow being spilled. The speed of the Pump adjusts itself to keep the inlet pressure to a specified pressure set point.

The pressure set point is best set slightly negative so that there is negative pressure (suction) in the inlet Duct. This suction helps secure the coupling onto the leaking pipe. The suction guarantees that any minor leakage in the coupling will result in sea water being drawn in inlet, rather than the oil leaking into the sea. The system works much like a vacuum cleaner with a flexible hose.

Another aspect of the design is that it is better equipped to handle leakage in the coupling. Coupling the new pipe to the leaking pipe should be a straightforward plumbing operation. However, if the joint turns out to be leaky, a small suction pressure inside the pipe will allow sealant to be drawn in to fill in the gaps for an oil-tight seal.

Methane Hydrates formed through the limited contact with sea water during the connection process are pumped out through the pump.

In contrast, all of the systems designed by BP collect only part of the leaking oil. The inlet pressure is positive, which is causes most of the oil to spill into the ocean through the leaky couplings. This was


A more complicated way was developed for the spill when the leak includes too much gas to directly pump. This solution is outlined in Figure B.

Oil Spill Solution
FIGURE B: Oil and Gas Spill Collector

The leaking gas-oil mix is pulled from the leaking pipe through a Coupling that is fabricated to conform to the shape of the pipe adjacent to the leak. The oil-gas mix is then pulled through a Flexible Inlet Duct and into the Separator Box. Gravity causes the heavier oil to fall to the bottom and the lighter gas to rise to the top where it is vented to the sea. A Sump Pump draws oil from the bottom of the separator box and pumps it through a pipe to the surface. In contrast, the two solutions BP used pulled the flow from either the top or the side, mixing gas and oil and preventing the devices from effectively siphoning the oil.

The Sump Pump is a key component of this design. The Sump Pump provides the pressure to drive the oil one mile to the surface. The Sump Pump is controlled by a sensor that determines the level of the oil so that it pumps just enough oil to match the leak rate and maintain the set level of oil. In contrast, the current BP collector has no pump or control system and instead leaks most of the oil to the sea through a leaky coupler.

Another key component of this design is the Gas Vent. The gas vent as shown vents just enough gas to equalize the internal pressure of the collection system with the ambient pressure of the sea. This equalized pressure prevents an internal buildup of overpressure that could cause the Coupling to pop off the pipe. This equalized pressure greatly simplifies the most difficult problem, attachment to the leaking pipe. Without a pressure difference across the wall of the inlet duct, the duct can be flexible and facilitate easy connection to the leaking pipe. Without a pressure difference across the coupling, the coupling can be attached to the leaking pipe without the need for a great force to hold the components together.

The gas vent could be made more sophisticated than the simple flap shown. The vent could include an adjustable pressure relief valve, to control the internal pressure to be a positive pressure slightly above ambient or a suction pressure slightly below the ambient pressure. The suction pressure could be used, for example, to help hold the coupling to the leaking pipe through suction. The amount of available suction is determined by the height of the vent above the coupling through the principles of hydrostatics.

The Sump Pump could also be made more sophisticated than the simple device shown. The Sump Pump maintains a fixed liquid oil level in the tank, based on a sump level sensor located in the tank. However, it is not strictly necessary to locate the pump at the tank. The pump may be at the surface to deliver up to 14.7 PSI pumping force, or submerged a short distance below the surface to deliver additional pumping pressure as needed. Submerging the pump provides additional pumping pressure of approximately ½ PSI for each foot under the surface the pump is submerged.

This design can be implemented in many variations. For example, if the natural gas is depleted before the design is implemented, the separator is not needed and the device can be much simpler. The only essential feature is that it pumps the all of the oil to the surface and thereby stops the spill.

Add New Comment


  • Jim Fiske

    It seems quite clear that your pump system would work, but it seems to me the pump itself is unnecessary. If air were pumped down a pressure hose from the surface and released into a riser pipe, it would produce more than adequate suction (the pressure inside the riser will be lower than the pressure outside, since the riser will be partially filled with air instead of water). The precise amount of suction can be varied by controlling the amount of air pumped into the riser. This technique is commonly used in reverse-circulation drilling, where the suction produced is strong enough to lift rock cuttings to the surface.

    There is another, even simpler technique with some potential (but may be too late, now). When BP placed a cofferdam over the oil leak, gas produced by the well mixed with cold seawater and froze into methane hydrate ice, which clogged the cofferdam and prevented oil from flowing up through the riser pipe. So “clogging the pipe” was considered a failure. On the other hand, BP attempted to “clog the pipe” quite intentionally, inside the BOP, with their unsuccessful “junk shot”. However, if they were to inject cold seawater into the BOP, it would mix with gas and create methane hydrates inside the BOP. This might also “clog the pipe”, but in this case that would signify success, not failure. The only question is whether the immense pressure inside the BOP would force the ice out along with the oil. But even if it did, the ice would emanate in a stream, perhaps allowing it to travel up the riser without clogging it. This approach would have worked better before they cut off the bent riser.

    Jim Fiske

  • Edward Bullister

    The air is an interesting alternative. The one advantage of the pump is its relative high frequency response. The pump impeller can speed up in a couple of seconds to increase the flow in response to a higher pressure inside the cap. If you let the pressure get too high, the cap will pop off like a champagne cork. Filling all or part of the 1 mile riser I think may take more time.

    If there were a fast response adjustable choke near the bottom that might fit the bill. The buoyant column gives the driving energy, the choke provides the control. There may already be enough buoyance in the oil and natural gas in the pipe. I'd need to run the numbers.

    Unfortunately the force of the oil in the well is in the range of 1-2 million pounds, so I don't see any way to plug it with ice.

  • Jason Irwin

    I enjoyed reading about your idea. I had a similar idea in that the "top hat" should be hooked to pumps to suck the oil out instead of letting the pressure alone carry it to the surface. It almost seems so obvious that something in the plan must not be possible?

    As for the pumps and if it was possible to find one big enough or not. If you can't find one then use 10. At this point it's not an issue of economics or ease of design. They need to get the job done and stop creating plans that may work and changing their mind. Too much PR and not enough action.

  • Rene Sugar

    Is there enough pressure in the methane, in Plan B, to drive a turbine to power the sump pump?

    Or, you could use a separate pipe bring the methane to the surface to combust the methane in a turbine on the surface. Capstone Turbine Corporation makes turbines for oil and gas operations among other applications.

  • Edward Bullister

    There may be enough pressure energy in the methane but it would be too complicated to extract it. The pumps are electrically driven.

    The methane is not as significant a pollutant as the oil. You could release it at the bottom or pipe it to the top. The main issue is collecting the oil.

  • Pat Howard

    Mr. Bullister,
    Would adding a second duct with pump (or even three or more additional ones) reduce the pressure and make the whole idea less subject to leaks sprung from the high pressure? I know the pump does reduce the pressure; I just had a thought that perhaps more than one device as you describe would reduce the pressure even more. By having say two ducts it would also make installation easier as you could mitigate the pressure on the one device by opening the other until the first is ready and vice versa. I am not a mathematician, so am unsure about this; but thought I would throw this out for your perusal.

  • Edward Bullister

    There may be value in redundancy, but one pump and duct should do it. The main issue is to equalize the pressure inside and outside of the duct. Without a pressure difference to drive a leak, the leakage should stay near zero.

  • Vlad

    I see only one problem with this idea: the sump pump. It needs to be powerful enough to pump thousands of gallons an hour gushing from the well at 6,000+psi all the way to the surface while remaining a slightly negative pressure at the pipe joint. Is there such a device and can it be lowered to the bottom of the ocean, supplied with enough power and operated there?

  • Edward Bullister

    Good question. Yes, there are such pumps. The pump does not need to pump a large differential pressure. The oil collumn has a specific gravity of about 0.8 - it is lighter than sea water, so the column could be bouyant and drive itself. The pump's main function is to CONTROL the pressure, and ram through any methane ice. I would need some of the details of the pipe, methane fraction, etc. to run the numbers. A couple of tens of PSID pressure should do it. A flow of 20,000 barrels/day is only 580 GPM or 36 liters/sec - this does not have to be a huge pump.

    I know there are electric pumps available (I think form Schlumberger) for oil drilling that operate down to 2 km (more than deep enough). I don't have any model numbers, but if anybody in the industry does please send to eb -at- If operating at 5000 ft is a logistical problem, the pump could be near the surface on the top end of the pipe. This makes it a little slower to respond to correct the pressure at the bottom (1 mile of semi-compressible fluid lies between), but it should be OK. At the surface the pumping pressure is limited to 14.7 PSI, but you can submerge it a small depth and gain an additional half a psi per foot depth.

    Thanks for the comment!

  • David Martin

    Will a column of oil float to the surface if it is contained within a closed vertical pipe? It appears to me that the route from the sub-surface oil reservoir, through the well bore, sump, pump, and finally the riser pipeline to the surface represents a closed system isolated from the bouyant effects of the ocean water and would NOT drive itself.

    This would mean that at the base of the riser your pump is facing 5000 feet of static head which must be overcome to push the oil up the pipe to the surface. At 0.8 specific gravity that's about 2000 psi. If the well is providing 6500 psi (w.r.t. surface atmosphere?) of head into the riser pipe then no additional head required from the pump, but now the problem is to maintain a partial vacuum at the well head.

    At half a psi per foot for water there is only about 2500 psi of static water pressure at the leak. With 6500 psi of pressure in the well, and thus 4000 psi of pressure difference driving the oil through any gaps, the pump in your solutions needs to suck at 4000 psi and blow at 2000 psi, providing 6000 psi of additional head at whatever volume the well is producing. That is a very powerful pump and may in fact be beyond the capacity of available pumps?

    The hoop stress on the riser pipe is certainly balanced by the (enormous) external water pressures but unless I am missing something I think it wrong to speak of the oil being floated up to the surface within the riser pipe unless the sea is allowed to enter the base of the pipe (as it was with the top hat solution). If the sea is allowed to enter the pipe, then the formation of hydrates must be controlled, and the injection of solvents from the surface needs to be provided for. These too would need to be pumped down at great pressure, being bouyant and open at the sea floor.

  • Edward Bullister

    That is true. There is a static head. However, the pressure at the bottom is more that adequate to drive it.

    The system tries to keep the pressure in the base near zero LOCAL GAUGE, that is, about 2500 PSIA. There is more than enough pressure at the well. An LA Times article estimated the wellhead pressure (if you tried to stop the flow completely) at 6500 PSI LOCAL GAUGE (or 9,000 PSIA!!). If you let it flow from the pipe freely, the flow rate will adjust itself to a steady state in which there is zero local gauge at the exit, and the underground frictional pressure drop through the long bore hole will make up the pressure difference.

    This pump would NOT try to stop the flow from the well (the "top kill" approach). Whatever flwo the well want to spew out is fine with the pump; the pum will react by pumping that amount to the surface.

    It may desirable to keep the pressure at exactly zero local gauge (2500 PSIA), as this would mininize leak flow if you can't get a good seal at the pipe. If this is the case, then the lower density of the oil and gas in the riser may indeed drive the flow buoyantly as the pressure in the pipe from weight of the oil and gas column would be less then the 2500 PSIA, and the column would push itself.

    So itis possible that the pump would be unnecessary and an adjustable choke would be enough to throttle the flow to control the local pressure to zero gauge. However, the pump might be handy for startup (with sea water in the pipe). In any case a pump would cover all bases, as the pump can be run at low enough speed to actually impede the flow. I use a pump in the design becasue there are several critical variables that BP has not disclosed, so the necessity of the pump is not certain.

    There is no reason to let water into the mix to make hydrates. The plumbing between the spewing pipe and the pump inlet should be good enough to keep a watertight seal, especially if there is zero pressure across the seal. I cannot understand why BP can't make a strong, watertight and oil tight connection to a pipe. With such a good connection, you could use the pressure from the well which would drive anything out of the pipe, with or without a pump.

  • David Martin

    Thank you, Dr Bullister, for the clarification. I now see where the 6500 psig of well pressure is being used up, and while I still feel challenged using the term "bouyant" to describe the oil in the riser, I had missed a key point: when the oil pressure at the well connection is equalized at 2500 psia there is then sufficient head within the riser pipe for the oil to move towards the surface. I expect you would still require the pump to manage friction losses in the riser pipe and keep the pressure balanced (or slightly negative) at the connection as I can see the extra 500 psi being used up on the trip to the surface.

    I was certainly not advocating the introduction of water into the system, but was wondering if open access to the static pressure of the sea water was a requirement when speaking of the oil being "bouyant" and floating up to the surface.

  • Ben Blackburn

    @ David Martin.
    Thank you for doing most of the writing for me!

    If a supposed top scientist missed the basics of pressure differentials and that buoyancy isn't a factor in a closed pressurized system, I think all his ideas should be looked at with serious skepticism.

    Remember that Knowledge can be had without Wisdom! :)

  • Richard Fontana

    I know Ed from graduate school and working with him. He is brilliant, creative, and qualified. As usual for Ed, when he thinks hard about something, he comes up with a good solution. At this point, they have cut and capped the pipe and there is still a leak. This is not surprising since they had already said there was probably a failed connection between the pipe and the rock. Now the oil/gas is probably eroding that connection. They should implement Ed's configuration ASAP to capture essentially all of the oil and reduce the leakage/erosion of the joint. They can then use the pressure control from Ed's configuration to facilitate repairs. Thanks Ed for working on this...

  • Chris Reich

    To me, whether this exact plan is the perfect solution or not is irrelevant. What matters is the source of the idea. The Obama administration has operated from a false assumption from day one: BP is most qualified to handle the fix. Untrue.

    Consider an example. A fire starts in a large car repair garage. The fire department arrives and decides, since the owner of the garage knows the most about cars, the shop and the contents therein, he should be in charge of putting out the fire.

    This makes no sense. The administration should have seized the asset immediately as a crime scene as they would in any other situation in which people died. 11 people did die. Remember? Then experts from various disciplines should be assembled including representatives from BP's engineering department to begin work. The government should have appointed a chair. No, not some senator who lost his seat, a qualified leader.

    BP's management, whose interest is PR, liability protection and stop loss should NOT have been making any decisions after the accident. I don't care how nice a guy Tony Hayward is, his motivation is tainted. BP had NDAs prepared before they had any other plans in place. They covered their legal ass before the leak. They still don't want the press snooping around.

    Next, a clear path from idea to reject or refinement to implementation has to be established. This mitigates the "Drunkard's Walk" of valuable ideas through the bumps of egos and conflicts of self-interest. (Brownian motion of thought)

    In a simple line: the thought process has been wrong from the moment the accident occurred.

    I hope we learn from that.

    I hope your business is learning from this. I work with companies every day who make bad decisions, costly decisions because the thought process is flawed by Brownian motion.

    There are now thousands of ideas on the table. Many of those ideas have merit. Pieces of many of those ideas have merit. But still there is a hit and miss approach in place. We're still hearing things like "this has a 70% chance of success". We should not be doing anything with a "chance" of success.

    We should be solving problems.

    Chris Reich

  • Ben Blackburn

    Although I do think that there is too much politics being included in the decision making process, both by BP and the other agencies involved in the spill, I have to point out that between BP, Haliburtan, and the other companies involved (including a couple of the top wild well control companies), they have some of the best and most experienced people working on this.
    There are a lot of ideas floating around the web ranging from nuking it to using pumps that use more energy than a small city, but they are mostly coming from people that have no idea of the basics of oil well control or equipment.

    I think that the blowout experts should be put in charge of the well control, and the spill cleanup experts should be put in charge of the spill, and let the politicians go shake babies and kiss hands!

  • Edward Bullister

    Mr. Baker- I think the below column is a little off topic, but thanks.

  • Ben Blackburn

    Off Topic???

    If you only think it is off-topic, then you need to go back and read up on your high-school physics! (I think you should anyway, based on the problems with your idea! ;)

    Mr. Baker is deluded as to the function of gravity, but then you also seem to be a little fuzzy on this subject, as Mr. Martin pointed out above.

  • Lawrence Baker

    Dr. Bullister, thank you for contributing your ideas. Although your design doesn’t address stopping the oil flow, it does demonstrate a workable model of a solution encountered on the surface where you would need to pump oil up a mile.
    BP, Obama and the DOE could stop the oil flow permanently tomorrow with deep water pile driving technology if they wanted to; they are more interested in propaganda, genocide and the New World Order.
    One of the quick and best ways is to lower a 4 ft. in diameter pipe over the wellhead and drive the pile down 200 ft. into the sea floor sealing the wellhead and bore hole. The pipe is then sealed from the surface with cement and the pipe is cut off far below the surface. It would take 83 sections of welded 60’ pipe and 80 valves to get to 5,000 feet in depth.
    Fluid dynamics is one of my favorite subjects so let us revisit Atmospheric Pressure and Hydrostatic Pressure which are the key components in this particular deep water problem.
    Deep water Hydrostatic science is something that the DOE and Big Oil do not want people to know about because it is the greatest renewable energy resource on earth- and it’s free. Think of the ocean as a large pressure container, the deeper you go, the more pressure there is. At 5,000 feet below the water surface (147 Atmospheres) the pressure is 2,161 lbs. per. Sq. inch. On the surface there is only 14.7 lbs per sq in of pressure. The weight of a column of water in a pipe reaching to the bottom is neutral until it reaches the surface. How high the column of water will rise above the surface in the pipe and how much pressure the water will have; is the difference in pressure minus the weight of the water ABOVE the surface in the column. The high volume and pressure of the water coming out of the pipe is free renewable energy in the form of hydro-electric power. This energy can also be used to suck up and separate spilled oil in the water at great depth. Not much use for an oil man and big bankers but a great source of free renewable energy for Humanity.
    Here are two simple experiments that anyone can perform to prove and demonstrate Hydrostatic Energy and the neutrality of the weight of water in a column of water below the surface.
    The hydrostatic pressure concept can be demonstrated with a small milk carton and two straws. Fill the carton up with water (or half way) and seal it with duct tape. One straw is inserted and sealed with duct tape to provide pressure. The other straw is lowered anywhere that the straw is under water and the hole in the carton is sealed. As we blow on the straw, we pressurize the vessel and the water will rise in the other straw because the end of the straw is open to the outer atmosphere and at lower pressure. The water in the straw was neutral until the vessel was pressurized. The same is true if we lower a pipe down into the layers of pressure in the ocean; the water will rise in the pipe because the open top pressure of the pipe is lower than the pressure at the open bottom end. The volume of water in the pipe has a neutral weight as it is the same weight as the outside water until it reaches the surface. The height and pressure that the water will rise above the surface in the pipe is determined by the difference between the bottom atmospheric pressure and the top atmospheric pressure minus the weight of the water in the pipe above the surface.
    The weight neutrality of water, in water, can be demonstrated with a bucket of water. On the surface the bucket of water has weight but as we lower the bucket of water below the surface; the bucket will float away because the weight of water in the bucket is the same as the surrounding water and has weight only when it rises above the surface.
    Dr. Bullister, my purpose in posting here is not to criticize, but to add solutions to solve this intransigent problem.

  • Ben Blackburn

    As Paul states, your water pressure in the pipe ideas won't work due to something know as gravity.
    The pressure in a column of water is the same for any amount of surface area. Thus 5000 feet of ocean will have the same pressure at the bottom as a 5000 foot long vertical garden hose. If you lower the hose to the bottom of the ocean, the water will rise in the hose until it matches the sea level outside the hose. Ask Mr. Boyle about his Law, he can give you the details.

    As to the idea of the piling into the seafloor, there are a couple of issues with it.

    1. It would need to be a lot bigger than 4'! The BOP is a huge monstrosity. I do not know the exact diameter needed to clear it, but it would be large.

    2. We are dealing with 6000+ PSI coming out of the hole. If you simply drove the pipe into the seafloor mud (which is all you could do if you were driving it), the oil would displace the mud and go under the piling.