Better BOP Design ??? 1

[docellen]

Starting a new thread as suggested by Windward

DrillerNic, Thanks for your excellent information on the BOP Emergency Disconnect. The best I am able to find on the Internet is a drawing of the BOP linked at

http://energy.gov/open/oilspilldata.htm.

I'm surprised that BP doesn't make more of this info public.

So I'm guessing what happened is that nobody had time to hit the Big Red Button when the fit hit the shan, and their backup plan was trashed in a pile of twisted metal. What I'm thinking of is something that will work even in that circumstance. No doubt you can think of even better ideas, but here are my suggestions:
1) Separate quick-connect nipples on the side of the BOP that go directly to every actuator in the system.
2) Clamps instead of bolts on all flanges. These should be easily operable by ROVs. The clamps should also have a well-calibrated breaking strength, so we never have to worry about extreme forces breaking something not designed to break.
3) Multiple wellheads. I'm still not convinced that this won't work. With an extra outlet to relieve the pressure, it won't take much to hold off whatever is coming straight up. You might even put the normal riser on the side connection, allowing emergency access straight down the center pipe.

If they had all this in the current situation, the action would have been: Pop the BOP, grab the drill pipe, and blow some mud as far down as it will reach. If that just blows mud in your face, pull the broken drill pipe and insert a pipe with some kind of plugs that grab the inside of the well casing, maybe one every 500 feet through the zone where the gas and oil is flowing in.

Question: How much force, worst-case does the casing and cement have to hold back? I've heard the pressure at the wellhead could be as high as 15,000 psi. On a 36" diameter wellhead, that could produce enough lift to make those steel pipes stretch like taffy, ripping loose from the cement in a thin zone moving quickly down the entire length of the pipe.

 

[waross]

Every time the design is declared to be foolproof, a better fool comes along.
Why didn't the weight of the mud hold back the pressure?
What mud?
Why didn't the annular seal work when it was activated BEFORE the blowout?
What annular seal?
And after the event, how about the head in the sand insistance that the leak was only 5000 barrels a day?
NEVER NEVER underestimate the power of a fool and the havoc that he can wreak.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[BigInch]

According to my whistle blower's network, a new BOP design is NOT NEEDED. What I predict what will eventually come out in this investigation, is that BP and TransOcean both knew unquestionably that the BOP system was not in complete working order and, only to save money and time, in that order, "somebody" purposely chose not to stop and correct the problem. They crossed their fingers, stared at the <$££$> mandala and lost all touch with reality, while they continuously repeated Sarah Palin's mantra ... something about drilling.

Oh... cat's already out of the bag.

http://www.mirror.co.uk/news/top-st...lt-weeks-before-rig-exploded-115875-22350381/

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[BigInch]

OK, make it bigger, thicker and tougher, but a complete redesign... forget it. This will be a human error root cause (or purposeful misdirection), as are most big disasters, outside of volcanoes, tsunamis, hurricanes and earthquakes, humans are an important trigger. All you really have to do is just to get humans to follow the procedures, by the numbers. Now to paraphrase Waross, Fools: Enter stage left.

As things get more complex, its apparent that the future will be wrestled away from the dynamic optimum path selectors by the simplest bots capable of connecting only the nearest of dots. Long live the cockroaches.


"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[DrillerNic]

In theory, docellen, the LMRP has just such a quick connect- a hydralic connection, so that the riser can be disconnected with the BOP shut in (if there's a big storm coming, for example).

The Deepwater Horizon's BOP was rated to 15,000psi. to work out the force, know the area.

The pore pressure in Macondo is reported to be 12.4ppg (pound per gallon) at 18190ft MDWLM (I'm not sure what that stands for- we do love acronyms in the oil patch- but I guess that's using mean sealevel as the datum) on the DOE documents.

So that's 12.5 * 18190 * 0.52 = 11823psi at reservoir depth

Assuming a full column of gas with a pressure gradient of 0.1psi/ ft (which is the worst case) from reservoir depth, we have a maximum pressure at the wellhead at 5067ft MDWLM of 11823 - (5067 * 0.1) = 11316.8psi

Well within the rating of the BOP. It's an 18-3/4" BOP (ie the maximum bore is 18-3/4"), so the area that pressure would act on is 276 sq in (the actual exposed area of a closed preventer is bit more, but for this exersise this number will do)

So maximum upward force at the BOP would be 11316 * 276 = 3.1million pounds force.

In reality, the fluid column in the well isn't just gas, but oil with a pressure gradient of about 0.3- 0.35 psi/ft, so the numbers are a bit smaller.

The standard subsea connector is the H4 from Vetco, the 18-3/4" model is rated to up to 7.5 million pounds force, depending upon the model and the connector/ wellhead configuration.

 

[shawnpeter]

DrillerNic - MDWLM is Measured Depth Wire Line Measured...

Docellen - Cameron made the BOP. I thought I might have a BOP manual/parts list for the same CAM unit but all I have is Hydril stuff.

For every big idea there is always a big enough gun to shoot holes in it. Had this been a well on land we could have killed the well in a matter of a few days tops. Moses parting the gulf of mexico would have helped.

 

[CH5OH]

I assume they left a few good brain cell engineer make the dessicions:
if I leave deep out of deep sea exploitation, exploitation becomes cheaper.
if things go wrong i just ignore them and they'll go away, they allways do?

 

[Windward]

I am not qualified to comment on BOP design and function, but I believe they should have a new feature: a simple way of attaching a new riser to direct flow to the surface when the BOP fails. Design of such a device is the subject of the thread "use gas lift pump on BP Macondo blowout?"

 

[DrillerNic]

shawnpeter- thanks for the info.. that's another one to add to MDBRT, MDRKB, MDBGL, MDMSL, MDBML....!

If the well had been onshore, it would almost certainly have caught fire... less ground pollution perhaps, but more air pollution! But with the fire out, it would have been much quicker to bolt a new BOP ontop of the old one... say 4 weeks tops (for example there's only ever been 1 onshore blowout in the UK, which burnt for 2 weeks and flowed for 3 weeks before it was capped).

 

[BigInch]

http://tahoe-is-walking-on.blogspot.com/2010/05/alternative-to-deep-sea-bop-operation.html

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[docellen]

I'm surprised to see much of the same beliefs here as I see on anti-oil forums. Maybe the reasons are entirely different, but both seem to be saying: A safe design is impossible.

I would agree that a perfect design is impossible, and the weakest link will always be in oversight (regulation), but from everything I am learning, and my assumption that deep water drilling will continue, it looks to me like we need some critical improvements in design, not just the BOP, but in the planning on what to do in case the BOP fails. We need a design which will "work" even if we assume worst-case behavior. That means not just incompetence, but willful destruction that might be caused by a disgruntled employee.

An acceptable design might be one that reduces the chance of a blowout to less than one in 1000, and the chance of a blowout going beyond 24 hours to less than 1 in 1000. That would allow us to drill a million wells and expect less than one disaster. The design must anticipate a situation where there hasn't been an accident in many years, and the regulated industry has gained control of the regulating agency. That means we can't count on any design feature requiring active oversight. We can't expect the overseers to be looking over everyone's shoulder, making sure they didn't fake a test report, etc.

DrillerNic, good data, but I don't think the BOP itself will be where the pressure causes the first failure. We need to look not at the area of the BOP bore, but the total area of the casing near the wellhead. 11 kpsi in a 36 inch diameter pipe will produce about 11 million pounds total tension in the wall of the pipe, regardless of what is mounted on the end of the pipe. If that wall is 100 square inches of steel (a really thick pipe wall), the tensile force is about 110 kpsi, which will certainly stretch the pipe, even if it doesn't make it flow like taffy. The part I can't calculate is how much tension in the pipe will break the bond with the cement around the pipe. There must be some experimental data on this, or even some experience with casings pushed out of wells.

Unless there is some really good argument that it can't happen, a safe design must assume it will, and not count on shutting off the flow at the BOP once a blowout has started. So I think multiple wellheads will be a necessary part of any design. If we can't stop it, the only alternative is to divert it.

 

[BigInch]

1000!? its probably already under 1:10,000 if not 100,000
Maybe you're thinking 1MM, or 10MM

I think it might be more prudent to assume you're going to have 1 and figure out what you're going to do about it (other than punt).

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[DrillerNic]

Docellen- fluid forces on the inside of tubulars due to flow are well understood and a standard part of tubing design, and to be honest these forces are low: for example, fluid friction pressure losses due to 30,000bpd of water in a 5-1/2" brand new tubuing string are about 50psi per 1000ft. To convert this to a force, multiply the total pressure loss due to friction by the cross sectional area of the inside of the pipe and the length of the pipe. For 5000ft of 5-1/2" tubing, 30.000bpd of water, we get 4229lbs force- not very much (unforunately the lookup tables I have are for tubing sizes rather than riser sizes!) The frictional pressure loss would be even less for a 21" marine riser as it is so much bigger. Much higher forces are imposed by diameter changes, plugs and the poisson effect.

The casing is always designed to the worst case it is likely to see- usually total evacuation to gas from the next section TD. As you say the problem comes when a casing string is exposed to a pressure it wasn't designed for- so the 36" conductor is never designed for 11kpsi from the reservoir! But even now, I very much doubt the Macondo 36" conductor is seeing 11kpsi- there are several strings of casing and the wellhead between the well pressure and the conductor. It was concerns about intemediate casing strings and shallower formation fracture pressures that were probably behind stopping the top kill- concerns that the intermediate casing string(s) might see a pressure greater than they were designed for and the blowout could become an underground blowout (where the flow goes underground rather than to surface) and this underground blowout could fracure the near surface sediments and flow to surface outside the wellhead- at that point everything has really got to ratsh1t and we're all doomed....

But really, until we know why the Deepwater Horzion's BOP failed, we can't make recommendations about BOP designs. Remember, this is the first deepwater blowout in 30 years or so, and thousands of wells, and heaven knows how many well control incidents and the first blowout in the GoM for 31 years and hundreds of thousands of wells, so the current design may already be at the 1 in a million or whatever failure rate.

I think the oil industry is secretly hoping that the deadman system was not maintained properly (which is probably Ttanocean's nightmare?) and then that's the answer.

What we can do now, is make sure all the designs and solutions for dealing with the blowout are disseminated (I expect to see a lot of SPE and OTC papers about this in the next couple of years- maybe even a dedicated SPE conference), and the kit: booms, skimmers, centrifuges, are held at a couple of key locations where you go deepwater drilling- Sao Paulo, Kabinda, Singapore and somewhere on the Gulf Coast.

 

[BigInch]

I wouldn't have thought 100s x 1000 offshore wells. OK, maybe worldwide, but does anyone have an actual estimate of how many offshore wells have been drilled say, since 1990. And how many of those were deeper than 1000 ft water depth. I'll bet that's a long way from 10,000.

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[docellen]

DrillerNic, Sorry I wasn't more clear about my calculation. The stretching of the well casing I'm concerned about is due to the static pressure (11 kpsi we are assuming). This will stretch the wall of the pipe in both a tangential and a longitudinal direction. The same calculation applies regardless of the diameter of the pipe. If the pipe we are considering is 12 inches instead of 36, and the wall of the pipe is 1/4 inch, instead of 3/4, the same stretching will occur. If the cement is rigid, all that stretching could result in enormous shear forces in a thin band between the upper section, which has already broken loose, and the lower section, which is still bonded to the cement.

Anyway, regardless of the calculation, it looks like we will never be able to rely on a cemented casing to take pressures of this magnitude. Therefore, we cannot rely on a BOP to simply stop the flow. It is becoming ever more clear that we must have a way to divert the flow.

I'm wondering now, what might be the worst-case flow we will ever have to handle. Imagine drilling into a reservoir of gas in a zone that had a recent earthquake. What is the maximum reservoir pressure, assuming the reservoir is supporting the entire weight of the rock above? How can we handle such an event?

 

[docellen]

BigInch, The one in 1000 number is actually incalculable, except in some artificial situation that would not be useful in reality. What I was trying to get across was the idea that if we have systems that are truly redundant, no common points of failure (probably no real system we can actually get in operation), then the overall probability of failure is the product of some small numbers, which can be made truly small. Now throw in one disgruntled worker, a connection to Al-Queda, and all your calculations go down the toilet.

The other way to calculate some estimate of the probability of disaster is to look at the number of deepwater wells that haven't ended in disaster. That might lead us to conclude that we will get something like 1000 good wells for every disaster. The only way to lower that number will be to change the way business and government work, and change it in a way that can never slip back to complacency and short-cuts.

Where engineering can help is in devising systems that work in spite of complacency, seat-belt buzzers that keep nagging when you don't do the right thing. We can also use more redundancy. I don't mean yet another shear ram. I mean something that will work even if we can't use the BOP at all.

 

[DrillerNic]

Docellen- we rely on cemented casing to cope with well pressures of thousands of psi every day, while we drill them, and produce them.... we know what teh compressive strgnth of cemetn is (we lab test it before each cement job).

There are concerns about the effect of ballooning on the casing/ cement/ rock bond- some people (including me) don't like it as we believe it produces a micro annulus due to the different modulus of elasticity of the different materials, so we try to do the casing pressure test before it has set; others argue that doing a casing pressure test after teh cement has set does not cause micro annuli...

The other thing to consider is that the measured burst stress of the production casing: the DOE documents say it is 7", 32ppf, Q125 x 9-7/8" 62.8ppf Q125. From my casing tables, the 7" pipe has a burst resistance of 14,155psi, and the 9-5/8" pipe (9-7/8" pipe simply deosn't exist) has a burst resistance of 13851psi- well above the reservoir pressure of 11,823psi I estimated above....

The concern is that if the flow is up the backside of this casing, the other casings that might see the wellhead pressure include 16" 97ppf P110, with a burst rating of about 7000psi.... (but I've already shown that the well diagram is inaccurate, so I'm not sure if it's really 16" 97ppf P110 casing)....

 

[docellen]

Another concern that just occurred to me, in thinking about multiple wellheads, is that letting the pressure at the wellhead drop to 2250 (the seawater pressure) even for a few minutes, may produce such extreme pressures at the bottom that the rock fractures and the flow increases beyond any chance of control. Is it possible that even a relief well may not be able to inject mud fast enough to stop the flow?

So forget about "pop the BOP", we need a design that allows a more controlled sequence of events, with a valve at the alternate wellhead, so the flow can be redirected without losing the 4400 psi pressure at the top of the well. Once the flow to the alternate wellhead is established, we could put a new valve on top of the old BOP, and shut it off completely.