A Petroleum Engineer's Explanation

reply to post by Mike6158


While PLCs do have some advantages with diagnostics, response times are slower, and failure rates are higher because the systems are more complicated.

What the tech describes is overspeed, and those systems do react very quickly.

If they put explosion rated doors, why wouldn't they use explosion rated electrical enclosures? At the very least they would have sealed the enclosures, so that they wouldn't be flooded with explosive gas. Why pay for one, and not the other?

As I explained earlier, the generator enclosure with the explosion rated door, so an environment created for such a hazard, would have had a limited amount of outside air brought into the room. This should have given the sensors time to react. Multiple sensors are typically put into place in consideration that some sensors might be flooded.

The LEL discussion really is moot, what the tech describes sounds like the generator was overspeeding, and the overspeed detectors should have shut the generators down.

Why not a mile long tube from the leak to the surface and then a a suction boat filtering all the oil there?

Originally posted by poet1b

While PLCs do have some advantages with diagnostics, response times are slower, and failure rates are higher because the systems are more complicated.

Typically the scan time for a modern plc is in the 1-2 ms range for an "average" sized program. The "response time" is moot. Failure rates are not higher. I wish I could give you a link to prove it but there is no link to paper. It might be out there on the web somewhere but I don't know where it is. There's a book and ISA class called "The Design and Implementation of Safety Instrumented Systems". It's an extremely good class that covers this topic in great detail for all kinds of applications. The Oil and Gas industry is the least demanding of the industries represented in the class that I took. Mechanical devices fail with no warning. Meaning that the switch could freeze and not actuate a year later (which is why frequent testing should be but is rarely performed). There is a term for this. PFD. Probability of Failure on Demand. The PFD calc applies to the end device, the control device (PLC, relay, whatever), and the final element (valve, controller, etc). Testing frequency, diagnostic capability, all of these things also come into play. In the end, the PLC will trump a mechanical SD system (assume sound design for both) any time. PFD and MTBF all come into play when determining the SIL or Safety Integrity Level of a system.

SD system design is not rocket science. They are far easier to design and implement than feed forward control of a process.

Originally posted by poet1b
What the tech describes is overspeed, and those systems do react very quickly.

If they put explosion rated doors, why wouldn't they use explosion rated electrical enclosures? At the very least they would have sealed the enclosures, so that they wouldn't be flooded with explosive gas. Why pay for one, and not the other?

As I explained earlier, the generator enclosure with the explosion rated door, so an environment created for such a hazard, would have had a limited amount of outside air brought into the room. This should have given the sensors time to react. Multiple sensors are typically put into place in consideration that some sensors might be flooded.

The LEL discussion really is moot, what the tech describes sounds like the generator was overspeeding, and the overspeed detectors should have shut the generators down.

I would agree except as someone pointed out in an earlier post, once there was gas in the intake air system, even if a door closed, there would be enough residual fuel to cause the engine to run away... honestly... we are both speculating and other than having an entertaining discussion there isn't much point. We don't know what caused the ignition. We don't know the type and distribution of the gas sensors. We don't know if they were stand alone (I really think that most offshore applications require stand alone detectors with hard wired outputs. Ie no PLC) I think we can be pretty sure of the source of the gas but that's about it.

reply to post by Mike6158


You're right that unless we know the specifics, we really can't say.

PLCs have improved, and specifically designed systems, that are not programmed by the user are very effective, which is why I tend to be leery of PLC safety controls. Independent stand alone systems are still quicker, but only useful for very specific applications. These would be electro-mechanical-pneumatic systems. When I think of mechanical safety systems I think of sheer pins, melt point links, and other kinds of last resort safety mechanisms.

Yeah, safety systems do need to be tested regularly, when they can be tested, and should be designed with testing procedures in mind. Admittedly, I come from a military and pharmaceutical background, so everything is highly regulated, when you have an admin doing its job. Without reg enforcement the testing is always quickly scaled back.

Figures, the oil industry which should be the most heavily regulated is the least regulated, and the most slack, when they should be the most regulated. It certainly would create jobs for controls people.

The main thing which made me go HUH!, is the over speed description. With fuel injection systems, I thought that shutting off the fuel supply and the air supply would shut down the diesel engine quickly, as the far more efficient systems require less fuel, and would have less fuel still in the system. It should have been one ramp up, and then a die down.

What can I say, I like failure analysis, and it is something I am good at, and made a career of solving problems that others could not. I enjoy looking into these things.

What can I say, I like failure analysis, and it is something I am good at, and made a career of solving problems that others could not. I enjoy looking into these things.

I can't fault you there. I am the same way. It's part of the job.

Originally posted by poet1b
Figures, the oil industry which should be the most heavily regulated is the least regulated, and the most slack, when they should be the most regulated. It certainly would create jobs for controls people

When the motorman said he wasn't aware of any testing for the overspeed system, and he'd been there for 7.5 years, that raised my eyebrows. I wouldn't be surprised if one of the outcomes of the investigation is more industry-wide routine testing of safety systems. And some of the existing safety systems may need to be updated. By the way doesn't the generator use little carbon brushes on the commutator that makes tiny little sparks that could have ignited the gas even if the overspeed control DID work?

Originally posted by polar
Why not a mile long tube from the leak to the surface and then a a suction boat filtering all the oil there?

We talked about that earlier in the thread. One thing that's unclear is the buoyancy of the water where the gas is coming up. And if a big enough pocket of frozen methane melts and pops up at once, it could sink a ship, even a supertanker.

I think it was molecularPhD who said they need to get all the smaller boats out of there and get some big supertankers in there to suck it up and I agree, they are using the smaller boats as an excuse for why they can't bring supertankers in.

But they do plan to do just what you said:

news.yahoo.com...

BP PLC is readying another patchwork attempt as early as Wednesday, this one a cut-and-cap process to put a lid on the leaking wellhead so oil can be siphoned to the surface.

[edit on 31-5-2010 by Arbitrageur]

reply to post by polar


A mile long pipe...I Guess the solidified hydrates might mess with this idea. A heated pipe maybe? Also I guess the big box they tried was the equivalent of the pipe idea. Why couldn't they heat the hydrates in the box? Guess it's just not an off-the-shelf kind of technology.

Originally posted by mrwiffler
reply to post by polar

 

A mile long pipe...I Guess the solidified hydrates might mess with this idea. A heated pipe maybe? Also I guess the big box they tried was the equivalent of the pipe idea. Why couldn't they heat the hydrates in the box? Guess it's just not an off-the-shelf kind of technology.

Heating the pipe would take a lot of btu's / hour. The ocean depths are cold and water conducts heat pretty well.

Hydrates are funny things. Under pressure they are very difficult to vaporize. Take the pressure off of them and they go away on their own.

If Billyjack is right, by the way, great movie, that the reservoir pressure, and so the pressure coming out of the pipe, is 20,000 PSI, and the pipe inside diameter is 10", then we are looking at about 75^2 inches, so about 1.5 million pounds of pressure. That is a lot of force pushing upwards, and if the ID is bigger, then the force increases exponentially.

As someone mentioned earlier, couldn't that pull the pipe apart down below in the casing and create a far worse leak. It seems they would have to run pipe down to the cap, and let oil pump up to the surface, hopefully to a tanker up above.

At the flow rate, it will take a huge quantity of heat to warm the pipe. The problem that they had with the first "cone" thing was that sea water was being drawn up the line and was turning into slush, sort of like when you pop the top off an almost frozen coke and it flashes into ice when the gas expands. If they can get this one sealed to the stub pipe at the well so that it don't draw in a lot of water, then the warm oil will help keep the system from freezing up. The temperature of the earth goes up with depth after the first half mile or so. I would imagine that the oil is probably pretty warm coming out of the hole before the gases begin to expand.
I wouldn't want to be handling the top end of that pipe when they connect it to the well. I get this mental picture of it thrashing around like a runaway fire hose on steroids.

Originally posted by poet1b
couldn't that pull the pipe apart down below in the casing and create a far worse leak. It seems they would have to run pipe down to the cap, and let oil pump up to the surface, hopefully to a tanker up above.

Couldn't what pull the pipe apart?

Yes there's a lot of pressure in the reservoir relative to the surface, but different sections of casing are exposed to only incrementally different pressures as the depth changes.

But there's no doubt that a blowout pressure could have a dramatic effect on something that's not cemented in place like the drill pipe.

The story on the drilling rigs I was on was that 1.5 million pounds of pressure or whatever was enough to take a miles long section of drill pipe and shoot it into the air like a giant strand of spaghetti, and that drill pipe would be falling everywhere. I had my doubts if that would actually happen but when you calculate the pressures like you just did, then it doesn't seem as unbelievable.

But we knew about this danger on the rig and the importance of keeping the mud heavy enough to prevent it, which is why I find it so shocking that the BP exec would overrule the toolpusher who wanted to keep the column of heavy mud in place. I'd like to know why the toolpusher felt he needed to do that. Apparently he was right.

You guys are just talking about static pressure too, not including the kinetic shock load that would have to be overcome as the flow of the liquid was slowed to a halt. The numbers you are talking about, if from reliable sources, are just the pressure that would exist if the oil were not moving, but capped off.
My understanding is that somewhere along the event timeline, the cement blew out somewhere down the hole, allowing a problem between sections of differing sizes of casing pipes, and causing a loss of drilling mud. Were they not trying to cement when the event began to occur? Maybe this is part of the reason that they are having such a bad time with the top kill method. The oil may be going between the pipes and coming up along the bore outside the casing, or through fissures in the gulf bottom.

reply to post by MelonMusketeer

I don't think the pressure in the reservoir has changed much, the reservoir has only been depleted 2% or so at the most. So if there was 20kpsi there to begin with, most of that pressure is probably still there and would be more or less the same if the well is blowing out or is capped. It's probably gone down slightly but not much.

But yes that's what billyjack described, the cement didn't seal the casing like it was supposed to. Perhaps they didn't have the mud heavy enough, or circulate it long enough, prior to the cementing operation. If the mud is heavy enough and they circulate it long enough, it's supposed to actually force mud particles into the sidewall of the hole and along with the weight of the column that forms borehole pressure to keep it in place, it forms sort of a seal, while they pump the cement. That didn't work right. It happens sometimes, but there's no need for it to be catastrophic if they do other things right.

This is soo painful to watch. I'm hoping some other people saw that huge claw thing they had earlier. Why can't they use that, but instead of it having sharp blades replace them with something dull and flat so that it crimps the end of the pipe closed.

Even if it doesn't work, which I really think it would. All they would have to do is cut the bolts on that flange and there would be a nice flat surface to mount to lower riser package to.

It's makes me sick to watch, because I think they are throwing away there only chance to seal this leak.

reply to post by billyjack

Very informative and knowledgeable post.

Is the leaking pipe 7" or 21" wide as I have read in another site?

I have a simple plan using anchors from small Navy boats for securely plugging up the broken pipe and need your experienced input.

caferush

Originally posted by Mogwomp
It's makes me sick to watch, because I think they are throwing away there only chance to seal this leak.

I think their only chance to seal it is the relief well which is still weeks out.

All they can do until then is collect what is leaking. They're going to try to do that Wednesday.

The relief wells are months away, there saying August, which will turn into Sep. Or Oct. Atleast with the crimping idea that could be easily tried and proven successful or not within a couple of days. It's like they waited a month doing very little and within a day they already have most of the riser cut off. BP needs to slow down I think a little caution goes a long way/

I'm watching it live. Let me guess they are going to cut the riser right above the BOP first. Then they will use the big yellow claw to then cut the riser into a managable section that can be easily removed only there won't be any oil flowing through it so we won't be able to see how the flow might be reduced by the use of the yellow claw.

I thought these guys were scientific. Why not cut a piece of the riser right above where it is flowing out that other leak we have been seeing the one that is horizontal. There would be nothing to loose and I know it would reduce the flow. It seems that would be the logical scientific thing to do. That way we could see how the yellow claw effects the flow. I know that claw could easily be modified to smash together rather than cut.

If it could be modified to clamp the end of the riser and stay there, thus reducing the flow. I think a top kill could be tried again with the increased back pressure it might just work.

With every statement made about how the oil spill "will take longer to fix" I find myself more and more involuntarily drawn to the idea that this is the event described in Revelations where the sea turns as the blood of a dead man.

Originally posted by Mogwomp
The relief wells are months away, there saying August, which will turn into Sep. Or Oct. Atleast with the crimping idea that could be easily tried and proven successful or not within a couple of days. It's like they waited a month doing very little and within a day they already have most of the riser cut off. BP needs to slow down I think a little caution goes a long way

So are they taking too long or do they need to slow down? I am confused by what exactly it is you want them to do.
They have been cautious and thinking this out but then they get the hate because they are taking too long.