Hacker News Reader

Comment by GiraffeNecktie | Hacker News Reader

I don't understand why BP's people think that you can just shove some heavy gunk into the hole and expect it to stay there. Common sense would suggest that you'd have to start by anchoring something inside the hole which you could use to build a progressively denser matrix. How the hell you do that at 30,000 feet underwater I don't know. Depending on the state of the hole perhaps you could put a hollow, expandable cylinder into the hole. The cylinder also would have a series of hinged blades that could be released and the pressure of the oil would force the blades together and close the leak. Hard to explain without drawing a diagram. Since it's an obvious solution I would guess that there's a good technical reason why it can't be done.

View on HN

Getting a cylinder into the hole seems difficult. An earlier attempt called the 'top hat' involved putting a capped cylinder on top of the hole but apparently methane hydrates, which are lighter than water, started accumulating inside it and it kept lifting off the seabed.

So my idea is halfway between the two: place a large, weighted concrete cylinder with no end caps above the hole - a hollow top hat, if you like. It needs to be big, like 25' x 25' or larger; height is more important than diameter. This in itself does nothing to stem the flow - but nor does it obstruct it. So the cylinder can then be sealed around the bottom and fixed to the seabed without interference from the pressure of the outrushing oil.

Then one can begin filling the cylinder with gravel or whatever is suitable - smallest material first, working up to rocks. As we begin covering the gusher, the pressure will obviously push some of the filler material out of the way...but as it's inside a cylinder, the filler is just going to get pushed against the side of the cylinder, where its weight is going to increase the downward pressure on the material at the bottom, which has nowhere to go except in towards the center. When the aggregate pressure at the bottom of the cylinder from the sides into the center exceeds the pressure of the oil, it will pinch off the flow, like a valve. This of course assumes perfect packing of the filler material; in reality the oil will diffuse through it, but in doing so the pressure will be distributed across the diameter of the cylinder, which will make it much easier to cap.

Or we could just go with Andy Borowitz's suggestion to plug the whole with BP executives....

but as it's inside a cylinder, the filler is just going to get pushed against the side of the cylinder

You lost me there. The cylinder is open at the top, right?

Ever try to push your finger into a garden hose?

I have no idea with how much pressure the oil is coming out there, but it's probably more than enough to push any gravel out of the way before it makes it into the cylinder.

The idea is build a 1 mile long column of dense liquid. Just the pressure of 1 mile of heavy liquid was hopefully enough to balance out the pressure of the oil.

It turned out not to be, but that was the hope.

I clearly didn't explain this idea very well. OK, let's imagine we have a mains pipe of 3 inches in diameter sticking out of the ground on a building site gushing water which we can't shut off, but fortunately it's at least fixed in place. That's our high pressure source of unwanted liquid, analogous to the oil gusher.

As you say, sticking something into it or even over the end of it to create a seal is not very practical. But suppose we took an oil barrel, which is about 2 feet in diameter, and drilled a hole in the bottom big enough to fit over the pipe. Getting in place would be a bit of a hassle but once you got the hole lined up with the pipe it wouldn't be too bad, right? OK water is still gushing but now it appears to be gushing in a narrow jet from the center of the barrel resting on the ground. It's still coming out of the same 3 inch pipe, rather than the whole mouth of the barrel.

We anchor the barrel with steel cables or whatever we can. Maybe we seal around the bottom of the barrel with concrete, so it's well and truly fixed in place. Water is gushing out the whole time, but money's no object here. When we are sure the barrel is sturdily anchored, we start shoveling gravel or lead weights it as fast as we can. As it fill instead of gushing in a narrow jet, the water will start (quickly) filling up the barrel - it's the same amount of liquid at the same pressure, but in order to get through the much heavier lumps of rock it's now flowing through a much wider pipe, at a rate proportional to the circumference of the original pipe/the circumference of the barrel.

If we work fast, we can get some kind of plug - perhaps on with an open valve on it - into the end of the barrel before it fills. OK, so then the liquid comes gushing through the open valve. But that means there is not too much pressure around the edges of the barrel, so we seal them with more concrete (I am of course assuming that the walls of the barrel can handle whatever pressure of water is coming out of the original pipe).

Come to think of it, we might as well just drop the barrel with the open valve on top of the thing. The valve prevents pressure buildup inside the barrel while we work on making a seal between the bottom edge and the ground. It just seems a bit easier to me to use filler material to step down the flow from a narrow jet to a wider tube.

Now we still have a gusher, but instead of coming out of a small hole in the ground it's coming out of a big faucet which we have built around the flow without making too much effort to halt it. At which point we close the valve at the top and hope the seal between the ground and the barrel is stronger than the pressure of the oil.

the pressure will be distributed across the diameter of the cylinder

Okay, I was skeptical about this thread when I first saw it, but now I see the awesome potential for teaching physics here.

You can't "diffuse the pressure". At every point along the wall of a sealed vessel, the pressure is constant. So if, e.g., the pressure of the hole when sealed would be 100 PSI, then to stop it every square inch of the seal needs to withstand that 100 PSI.

If the top of your concrete cylinder has, say, 100 times the area of the hole, then the flow of oil may end up evenly distributed over that area. But if you then try to seal the top you must now provide the same strength of seal -- it must withstand the same pressure -- over a much larger area, which is probably harder to accomplish.

This seems counterintuitive, just as the lever is counterintuitive. But we use this principle all the time to lift things like elevators and cars. You push on a tiny-diameter cylinder with your arms (or with a little electric motor), and it pumps fluid into a much larger-diameter cylinder under the car, and the car rises. Of course, you have to pump up and down dozens or hundreds of times to lift the car one inch.

up voted you because the opportunity to learn more about this situation was one of the main reasons I asked this question

I should also note that the idea of blocking the hole with gravel isn't exactly a bad one. You're right to point out that if you use big pebbles for your gravel the oil will just squeeze through the abundant empty spaces. So... use finer gravel!

The pressure will try to push the gravel out. So maybe make it really sticky gravel, so that it glues itself together.

Now there are three things to say about this plan:

(a) we call this sticky, very fine gravel concrete and we use it all the time;

(b) a fun educational point: Even with really fine gravel, oil can still diffuse through it. Oil diffuses through "solid" rock all the time. The oil underground is actually trapped inside rock. The rock that holds the oil is relatively porous; the oil is kept underground because above that rock is a layer of denser rock that oil can't diffuse through very well.

(c) In case it isn't obvious yet: The reason your idea still doesn't work very well is that the first cupful of wet concrete you pour onto the hole will promptly get flushed away by the pressure of the leaking oil. ;) As will every successive cupful of wet concrete. Unfortunately, a giant slab of solid concrete must necessarily start out life as a thin layer of wet concrete, so the only way to put a concrete seal over a hole that is actively leaking is to cast it someplace else and then drop it on the hole.

Well this was why I didn't suggest pouring concrete into it...I was assuming that the cap would be cast at the same time as the cylinder :)

The only point of the gravel (rather than sand) is to disperse the flow sufficiently that maneuvering is easier - same way you could wade across a stream but would be knocked over by the same volume of water aimed at your from a firehouse.

Actually an interesting point. What would happen if they cast a stadium sized block of concrete and dropped it on the hole.

It would go around.

"we call this sticky, very fine gravel concrete and we use it all the time;"

I wonder if Nanotech could come up with some kind of super concrete. Ie particles that stick together extremely well?

I understand that, and apologize for my loose terminology. There are two reasons I suggest employing a larger size.

The first is to do with flow, as you mention. While all of a large cap needs to be able to withstand high pressure, that is not in itself an especially tricky engineer problem. But size doesn't hurt; in my experience it's rather easier to stop up a large pipe than a small one with the same pressure, simply because it's a less delicate maneuver. In the case of a domestic plumbing crisis, it's nice to be able to exert your arm rather than just your fingers.

The other reason is simply to do with bulk. The larger the cylinder, the greater it weight when finally sealed, the more room you have to anchor it, and the farther from the (presumably fragile) ground in the immediate vicinity of the gusher the circumference will be, if one is trying to bore into the ground with restraining pegs.

The surgical approach has not been very successful so far. Given the urgent nature of the problem, I feel it's worth trying the crude but frequently effective approach of large mass + gravity.

>You can't "diffuse the pressure".

I must be missing something, the parent is talking about wrapping the 21" (inch) diameter pipe in a 25' (foot) outer pipe, this is effectively 200 times the area. The force applied currently by the oil coming up is going to provide 1/200th of the pressure when applied over the end of the pipe.

It's like blowing out through a straw - you can feel the pressure easily. Versus blowing out through a dustbin with a hole in the bottom.

Mind you it's 4:30am ... perhaps I should go to bed and think about this another time.

You can't diffuse static pressure, but when you deal with fluid flowing through pipes, you need to take the Bernoulli effect into account. Narrowing a pipe with a fast flow can sometimes decrease side pressure.

Although sealing the top of this concrete cylinder may be impossible due to pressure, would it not make it easier to attach a pipe and start collecting the oil?

Yes, it would have to be a big pipe, but is that possible? Or, could you not attach a cone to the top of the cylinder to condense the flow into a smaller (pipe size flow)?

Having slept on the idea and read some more news this morning, I'm much less optimistic about it :-(

I'm struck by the fact that there are other oil plumes nearby on the seabed, eg 400 feet away from the big one. If the borehole itself is damaged and oil is also leaking out through cracks in the ocean floor, then successfully capping the wellhead will just increase the outward pressure to other leaks, and we'll end up playing whack-a-mole. Depending on how porous or fractured the ground is in the area around the well, attempting to just cap it may be futile.

Indeed, it seems as if the latest strategy is the mentioned here of dropping a giant funnel on it and sucking up as much oil as possible through a pipe rather than allowing it to diffuse into the water, while we try to get an/other rig(s) in place to drill relief wells. And the estimate of having that done by August is based on the time it took to perform a similar task in much shallower water. The explosion in April took place about 3 months after drilling began, and while BP began drilling a relief well around the first week of May, there's no guarantee that any individual drill will tap into the same pocket on the first attempt, which I presume is why the government has ordered they get a second one going ASAP. The Ixtoc I oil spill in the 1970s, which was also in the gulf, took 10 months to bring under control. Although our drilling and seismic imaging technology has improved significantly since then, they're also applied to wells at much greater depths.

Because often you can ram a bunch of gunk down the hole and have it stay there. This is how they have been controlling wells and stopping gushers since the beginning.

The problem with asking non petroleum engineers this sort of question is that you get incorrect and useless answers based on very limited "intuition and common sense" and not based on actual knowlege of the discipline.

That was kind of my point in creating this thread. While there are some complex elements in play here and it is a problem with many technicalities to it, there aren't THAT many elements to worry about at I understand it. You have water pressure, the pressure of the oil excreting, the depth and breadth of the hole, and how deep it is under water.

Further more its not like the solutions that they have tried so far have been particularly high tech (put a cap on it and hope the water pressure with seal it, stuff the hole with crap and hope the water pressure will seal it, put fluid in the hole and hope the water pressure will seal it).

There are a lot of smart people on HN, so if nothing else it is good mental masturbation to make suggestions and then explain why those suggestions are or are not feasible.

> and hope the water pressure will seal it

No. The hope was that the weight of a column of heavy mud a mile long would be enough to counteract the pressure.

People spend years just studying fluid dynamics alone. People spend years studying mechanical engineering by itself. People spend years studying chemistry by itself. People spend years coming to grips with the operational difficulties of undersea operations.

What those people would see in this thread is like what we'd see from a bunch of laypeople designing the back-end of a Web app.

It's a smart idea; but I think the failure point there is the pressure. My mech engineering is shaky but I suspect getting a turn to expand under that pressure is hard. Also the hinges sound like a possible weak point.

It's "only" ~5,000ft of water.

Reply on news.ycombinator.com