Failsafes in Offshore Drilling

Mankind is not infallible. So there can be no expectation that his machines will be. Nothing can be done to make it fail safe. Sooner or later something unexpected comes along. The devices don't work then we work to see it doesn't happen again by improving on the devices. We can think of all the possible scenarios in which might happen. Plan for it the best we can in the designs our devices. There will always be something that we didn't think of.

In relation to this a better question is the risk worth the reward? Is the need for oil worth the environmental damage when it does happen?

Fail/Safe, operator control & input are such tricky subjects.

Remember that Three Mile Island would not have happened if the operator had allowed the safety system to do it's job.

Inspection and frequent operation is too often put to the back burner in this RBI and Profit Making day and age. I do not know of all of the gages, controls, and training that comes into play either, but it sure needs to be addressed.

Operator control and input: ...training:

Inspection and frequent operation:

I don't know anything about these systems either, but you definitely are making key points here. I wondered whether the BOP is regularly deployed (ever, or how regularly) in the process of drilling. Ideally the operator would be involved in the frequent inspection and operation of the device.

In the Gulf of Mexico spill, it seems like something beyond operator control, a contingency that wasn't planned for : the explosion of unknown cause (at least, I haven't heard anything about it yet).

Check this quote taken verbatim from another forum dedicated to the industry

In what stands to be one of the biggest oil spills in the history of the United States, the possible cause of the spill now appears to be an unauthorized modification of the blow out prevention (BOP) valve. BP's Deep Water Horizon oil rig exploded and sank off the cost of Louisiana last week. Eleven rig workers are missing, which was operated by Swiss-based Transocean Ltd, the largest independent driller in the world. Meanwhile, the BOP failed to stop the flow of oil as it should have after the explosion and every day 200,000 gallons is pouring into the Gulf of Mexico. Fingers are being pointed at BP, even though Transocean was the sub-contractor. New evidence shows BP will eventually be exonerated. But first we will all have to play a game of cover our asses. From the subcontractor who reportedly modified the BOP without the knowledge or permission of BP, to the government regulators who okayed the modified BOP, to the Obama administration who wants to look tough on BP even thought the company had no culpability, everyone but BP is running for cover. Internal records show that Cameron built the BOP and delivered it to Transocean. Before the BOP was put into service, it was reportedly altered by Transocean with no modification approval or notes. The BOP was then installed, and the modifications are assumed to have prevented the part from operating properly.These modifications were discovered by remote operated vehicles, whose pictures transmitted to engineers trying to establish why the BPO didn't activate, showed the part had been altered. Yesterday, top BP executives met with Secretary of the Interior Ken Salazar and the U.S. Coast Guard and told them about the tampered BOP and the fact that BP had no knowledge the device had been altered. Menahwile back in D.C., in an effort to boost his sagging approval ratings, President Barack Obama is trying to make this into political theatre while his people know this was not BP's fault. The slow leak of information to the media about the history of the rig and the health and safety record is an attempt to shift blame away from the agency's who regulated the rig, to the operator. Everyone is trying to cover their own ass. BP continues to work around the clock and continues to re-assign people to the region to assist with the needs of locals. Lets hope we can get to the bottom of why Transocean modified the part, if they did, and why Federal regulators signed off on a BOP that failed to stop the flow of oil.

Well this certainly looks ominous. Installing three, four or five of these modified BOPs (A) would likely make no difference. But one unmodified BOP (B) is what is required. If this is not a convenient rumor, it will be interesting to see what excuses come up to explain the modification and approval.

Since these guys are all Inspection hands that have worked on and around this equipment most of their career, I will definitely pass on info from the technical front, not the media. I sense Transocean is the culprit, not BP.

But everyone is covering their own behinds right now and rumors are flying.

Most of us have contracted or consulted to BP through the years and want the facilities to be safe.

You'd think a manual override of the valve would be employed, perhaps a ball valve with a lever that's rotated 90 degrees. Maybe a patent search of the Cameron products would divulge the valve innards. Transocean may only have incorporated external fitting adapters. It would be unlikely that Transocean would have made internal changes unless they were authorized by Cameron.

I did a quick search on my hard drives for some Specs or other info on Cameron for the valves, but no luck. I have done a few things for them thru the years, mostly compressors, but no valves. There website does give some insight into BOPs though.

http://www.c-a-m.com/content/dps/drl/index.cfm

BOPs are listed under subsea systems.

The BOP that i am familiar with were on land based rigs. They were hydraulic and operated from a station away from the rig. The BOP did have manual controls. Hand wheels on the sides to crank them closed. Little hard to do in 5000 feet of water. Even if the manual controls are there they may not be accessible if when the rig blew up the structure dropped on top of the well head.

in this case, the bops should be 'normally closed' safety position, with hydraulic fluid required to keep them open. If the rig crashes, and the hydraulic fluid stored leaks or drops pressure, the bop's close. (yes, including the shears)

so if you accidentally shear off a mile of drill pipe.. it is still cheaper to spear the pipe and pull it out later, than a billion dollar clean up, and the loss of the 750 million dollar rig.

Chris

Well, as BillR suggested, the odds of preventing such an event by having a triple failsafe would be much improved.

In terms of a risk/cost/reward analysis, I'm willing to bet that the cost of two extra BOP's and a wireless control for backup would be a heck of a lot less costly than the damage that's threatening the coast today. I never want to hear of another case like this, ever. The failsafe odds have to be improved as much as possible.

Before that risk should be taken, there should be an answer to every "what if" scenario, in terms of the failsafe mechanisms. What if there's an earthquake? What if there's an explosion onboard. Etcetera.

Interesting subject. It would seem if three BOP's were stacked one above the other, the odds would be high that one of the three would work. A low frequency LF/VLF communications interface, such as used with submarines, would provide the commands to shutdown. A ship or ships, designed to transmit high power LF/VLF commands, could be employed to penetrate the ocean depths to communicate with the valves. If there were multiple stacked valves independent of each other, then they could be individually tested for closing and opening. I've designed VLF navigation receivers in the past, so I have a good feeling with regards to that technology being able to work. High tech rechargeable batteries would be useful and perhaps the oil flow through the valves could employ a path for oil to spin a vane that spins a generator to charge the batteries.

Increasing the odds by backing up the failsafe is a good point. Three BOP's, your odds are much better that at least one will work.

An excellent point. Redundancy alone is not sufficient: diverse mechanisms are a real backup against failure of the design.

I have to say that, drilling a relief well is not a very convincing backup plan for the BOP, since this takes way too long. The backup mechanism should be immediately applicable by design.

I agree that the time frame for drilling a relief well at this point seems futile, but if 8 weeks from now oil is still spewing, then we will all be glad that someone already decided to get started. In hindsight, wouldn't it have been nice if the relief port was drilled when the rig was built as a required "diverse safety mechanism?"

I agree, drilling the relief well in advance could be a good security - also a good thing they are proceeding with it while other measures are attempted.

I don't know whether an advance relief well would be agreed to, though, because of the expense, which would probably make offshore drilling not economically viable. Being a realist, I hope that there is some cost effective measure to improve the present failsafes - something we can ask for and expect it to be agreed.

I'm sure that I'm not the first person to ever consider that if a relief port is the only effective counter-measure to a catastrophic failure, and that the amount of time required to drill the relief port is measured in weeks or months, then the only way to avert a complete disaster which will cost hundreds of millions if not billions of dollars is to have the hole pre-drilled.

I also have no doubt that you are absolutely correct in saying that the oil companies will go to great lengths to avoid the cost of drilling two holes. In fact the last time they needed a "cost-effective" solution to present to regulators they came up with a great idea for a remote-operated blow out prevention (BOP) valve.

I agree... perhaps such a plan can be embedded in an API performance criteria for [undersea] well drilling.

But what if the third system of a "different approach" makes for a lousy backup because it is poorly made (less reliable than the first two valves). As a good system engineer, I would pick the best proven system with the best operating history and use it for all three valves. That's the proper approach for a statistical analysis. If the "different approach" is statistically better then it should be used for all three valves in the third place. Trust me, I've been there. Bill

No, statistically you are wrong! Ignoring the property of coupled random processes does not make them go away. Review the fundamentals of Bayesian inference for your answer. The fallacy in your logic comes from ignoring that approach "A" did not work at this location after two attempts. To paraphrase Einstein, repeatedly doing the same thing but expecting a different result is insane.

Look at it this way, given that process A will work at a particular location and scenario 75% of the time and A has a successful execution in doing that task of 90%. Then when doing A twice and neither attempt works then there is only a 5% chance that you are at a location and scenario that A will work here but both attempts were the 10% execution failures. So 95% of the time when A1 and A2 do not work no amount of A attempts will work. If the lousy B procedure works only 40% of the time anywhere, this is a much higher probability of working than the 5% chance A has at a location that A has already failed twice.

Einstein was a scientist and not an engineer; there's a big difference. Those of us that have worked around a lot of scientists understand the difference. Number 1: they are most often not designers. A scientist might have a great idea, but often is not a good implementer of the idea. So I'm not enthusiastic about listening to the opinion of a scientist with regards to product development engineering.

Problem with your approach to thinking is it doesn't provide what's best in the real world. So you are saying it's OK to use two great valves and a third mediocre valve? I don't even need to use numbers to explain that one ... just use common sense. There's only one "best" valve.

Have you ever designed something that worked much better than anything else available? I can think of two things I've worked on (a commercial airline Omega navigation receiver front-end for Litton Aero Products and a 3kW TWT power supply that operated between -55 to 95 degrees C for Raytheon ESD). The competition hardware was available for study in both cases. So it wasn't difficult to reverse engineer the products and determine whose product was more reliable.

Are you willing to use a substandard product? And I'm defining "substandard" as any product proven to be less reliable. It could be the oil guys are using a lousy valve, but who really knows. Regardless, there's only one choice of valve; it's the one that is proven to be the best. And it's easy to determine the best valve by employing rigorous testing; such as used for military products. I'll fly on the super magne-transporter that's the best made; not something that is made better by number manipulation.

I've just given you a GA. As much as I dislike statistics, I understood your point (You have got to start using bigger words if I can understand ), and I agree qualitatively (I've not run the numbers, but I'm assuming you're right since you know this sort of thing pretty well).

I guess a simple way to say it is (for an OF like me) to remember Mickey Mantle, possibly the greatest player in baseball (when his knees were good and he was sober, he was better than Ruth or DiMaggio). But, the Mick couldn't field a ground ball to save his life. So, had you had three Mantles lined up, a fair number of ground balls would still get past them. But, if you had put one Mantle and one Brooks Robinson in a line, all balls - ground, fly, whatever - were gonna get caught, even though Robinson wasn't as good as Mantle.

Hello Bill and others,

I'm amazed that the BOP hydraulic systems did not perform correctly. If, as suggested, all the BOP systems was modified by a single person or a team operating under a common paradigm, that would definitely lower the system reliability.

The Space Shuttle had two contractors writing flight software to ensure that design and paradigm flaws are not propagated across all flight control computers. IBM and NASA wrote the flight software and Rockwell wrote the backup flight software.

A series cascade of components that have a 10% chance of not opening can be taken to arbitrary levels of reliability. Twenty 90% reliable relays in series designed to open the circuit all operated at once have a probability of failure of 0.1 ^20. The Soviet Union used this technique in their launch vehicles.

Incidentally, I'm a Senior Scientist/Engineer.

Hi Fire Stranger. Welcome to CR4.

From what I have learned from people who know the BOP systems, all are amazed at this failure. The fact that it did fail is an indication that there is room for improvement, and this is necessary especially for those deep water situations where the opportunity to fix a problem after the fact is severely limited.

I heard on the news last night, that Chevron intends to go ahead with a much deeper drilling project here in the North Atlantic, where weather and sea conditions are also much more severe than the Gulf of Mexico. I'm assuming that Chevron's analysts have insight or information about the present BOP failure that rules out pressure due to the extreme depth as a factor in this case, or they would not be announcing a go-ahead on 2km deep drilling....

I may be wrong since I am not a Senior Scientist/Engineer only a senior Engineer but as far as I know the functional probability of a "n" components in series, each one having an operational probability of "P" (P<1) will be P(n)= P(1)^n thus the probability of right function for 20 elements in series with a unit probability of 0.9 is 0.9^20=0.1216 the failure probability being 1-0.1216= 0.8784. The functional probability can be increased by a parallel use of elements not by series, again as far as I know. For a parallel connection the functional probability is P(n)= 1-(1-P(1)^n) in your example P(n)= 1-(1-0.9)^20) =1-0.1^20 ≈ 1, but of course this result is only based on what I know (incidentally I have a Russian book from 1968 where this is clearly explained).

This one aspect of reliability of elements connected in series lead to failures in different programs. I may doubt that the soviets used a redundancy of 20. In usual airborne systems the redundancy (which means parallel connection of components with same function) goes up to 5 or 7 in very sensitive situations. The number is always not even (3;5...)to allow a majority voting procedure which will not be possible with an even number. The degree of redundancy depends on several factors as risk, expected operational time & cost among others.

I think engineers who design BOPs consider cost-to-make and install. If I had the job of designing a BOP, I would over-design it. Something as important as this, operating at such great depths that are virtually impossible to access, service or repair can be made fail safe by over-design. As engineers on land and sea, we are taught to design using the properties of the materials available. I can remember way back 50 plus years ago, ships of the U.S. Navy were over-designed with safety factors of 5 and more. With better knowledge of structural analysis and with the aid of computers, safety factors have dropped to much lower levels in order to keep it lighter and less costly. This design mentality may be OK for buildings, towers, bridges and ships, but not in a BOP. Look at the Brooklyn Bridge for example and compare it to a modern day bridge. There are places where one must not skimp. Over-design in this case is erring on the side of caution.

Yes, the BOP must be as perfect as it can be. Considering that it's non-accessible, in my opinion it is probably a very elegant design of simplicity; something that gets the job done and that's all. No bells and whistles; definitely jam-proof as can be and no material cost constraints. And more product reliability testing than we can imagine.

One suggestion by marcot in another thread is to predrill a relief well on every well. Then the failsafe is there in advance (instead of three months later to get this in place).

The oil is spewing out with a considerable pressure. Drilling a next "relief" well, given that the diameter is the same, the volume also. Theoretically if 50 % will be delivered by the new well. The spill will be half of what it is now. Practically: oil output will probably also be function of the used length of the pipe so the spill remain more than 50%. As long as we do not "PUMP OUT", there will still be an overpressure that spews the oil out. Pumping will - after (un) reasonable time reduce the pressure and when it equals the water column pressure the leak will stop. But what is happening there in our mother earth's stomach?

Well, I talked to my brother today who used to work in the industry. He has seen the BOP and described it to me, and it appears there is redundancy already built into the device: layers and layers of ram shears (to twice the height of a man) that close with enough force to cut the drilling pipe. He also told me that the BOP is used and tested on a regular basis during drilling operations. Lastly, he suggested that the idea of drilling a relief well on every well would be prohibitively expensive, so that's probably not a viable suggestion for backup.

With so much redundancy built into the BOP itself, an outright mechanical failure doesn't seem possible: I'm assuming that the mechanics of each layer is independent therefore creating a failsafe against physical failure of one set of shears. It seems that an electrical or signalling failure is the thing that would cause the entire device to fail.

Go to Google Patents and search for Cameron's Patent Number 4,460,151 (Annular Blowout Preventer). The invention described appears to be an all-mechanical device.

Instead of relying on my personal unrelated engineering training and philosophy I finally decided to examine a BOP manufacturer's web site. Under their sales definition the Sunnda sales information states:

(BOPs) are used to control blowout. The crew usually install several blowout preventers (BOP stack) on top of the well, with an annular blowout preventer at the top and at least one pipe ram and one blind ram blowout preventer below. Also, some well control techniques require both the annular and the ram blowout preventers.

This manufacturer recognizes that more than one BOP design can should be applied in critical locations.

There is also another highly unlikely but plausible scenario that meets the known facts that has yet to be discussed. What if the blowout path is not through the BOP?

I don't understand the relevance of your post? please explain.

In my understanding of things, when you are a mile underwater, there is no sure thing. Even a dumb cone guided by a pipe from the surface, or without, can go wrong, and there you are.. unable to move the thing you spent weeks and a 100g's making.

When drilling with 'surface mounted' bop equipment, you can always use explosives or other brute force methods to cap the well. (call Red Adare) and while burning wells may pollute the air, a well gone wrong doesn't mean the extinction to any species typically.

As a guide, the higher the technology you employ (eg, moving parts), the more it can go wrong. What you need is a larger infrastructure with huge underwater submarines like ginormous lobsters that can work at that pressure, and move around reasonably quickly to pinch off and control a rogue well. Perhaps they need to rethink the cementing procedures entirely, as that seems to cause a great deal of the issues.

Chris

I believe its correct to state the following:-

The BOP had an illegal modification. The modification has caused the malfunction.

Does anyone know what the modification was(is)?

Andy, I don't think it's safe to assume that information source is true and correct. I wouldn't take that as true, until there is official word of it.

Thats EXACTLY why I was asking in such a manner.......trying to get the REAL story....

Thank you. You have exactly answered the OP and I vote you GA.

I would like to know whether redundancy in electrical/power supply and in signalling is also standard for the BOP, as you described the sensible way to backup.

Secondly, I wonder about the gas pressure and how a failsafe can be designed for the hydraulics. Any failure of that pressure system and the device is useless. Pressure requires containment, so a physical breach would prevent pressure building up as required. Gas lines are also vulnerable to a physical breach. You almost need a backup that does not depend on hydraulics at all. For example, a valve that defaults to the closed position whenever power is cut off.

Your comments make the most sense, since you have a complete understanding of the BOP design. Without that inside knowledge, we can only talk in very general terms and of course miss the mark with assumptions that are purely speculative (and even funny).

In terms of the OP, how to improve failsafes for the BOP, we know that the ROV's have the capacity to send the signal at any amplitude necessary, so that doesn't appear to be the problem. However if there's a power failure and insufficient energy for the pilot valve actuator as you indicated, what sort of backup system would improve the design?

For example, if it was part of the ROV backup plan, that the vehicle could plug in and power the device if necessary, that would cover power failure for the BOP.

I agree that statistics is often a very poor and frequently misleading method to persuade an opinion. All to often a perfectly executed but incorrect analysis approach can lead people astray. But an unstated statistical analysis was posed as the root to revelation. On this I do agree but with the caveat that a correct statistical analysis must be done that recognizes all of the degrees of freedom and actual combinations available. (Yeah, that should be enough big words to be both accurate and cryptic. )

Since this is a single event condition under discussion, statistics should not be applied in the first place. For no matter how rare or unlikely an event was perceived to happen, it has happened. And as anyone should know, statistics is the mathematical study of many discreet events or continuous conditions that can and do occur.

With this in mind and with your partial familiarity with at least land based drilling, would you, nick name, please comment on my unlikely suggestion that the seal between bedrock and drilling pipe could be the problem. It seems to me to be the one scenario that no amount of blow out protectors would have worked and that a slant drill to intercept at a lower point on the pipe might be the only solution.

Nice picture you found from the BBC Chris.

I'm tempted to make some untrained conclusions and comments about each of the three proposed methods.

Short-term:

Success with this method will/would permit reuse of this well head with the least amount of expense. I'm certain this is everyone's preferred scenario to be successful. This requires either an operational or repairable BOP in place or the possibility of installing another BOP on top of the permanently failed BOP.

Medium-term:

As the title implies this will take some time. This method does not rely on any part of the well head n working order. The drawback of this method though is that this effort must be a temporary effort for it will not permit easy recovery for use the oil. It may not be quick enough though to prevent significant enviormental impact.

Long-term:

This method will take the most amount of time. It does not require any part of the old well head to be functional. This will permit installing a new well head so that oil may still be drawn.

What makes them think long term stops the leak? Only when they create an under pressure in the outflowing feed, the old pipe will stop spewing. This requires enormous pumping in the direct vicinity of the old well. Fact is there is maybe 200 bar pressure or more. To pump that down over 1500 m?? Has someone an answer for this problem?

I do not remember where I read this so I cannot cite the article but the permanent fix works this way:

The secondary drill path intersects the run-away well hundreds of feet below the soil bottom. A concrete slurry gets pumped into the failing well. The ground oil pressure and secondary well pressure push a stone slurry up into failing well head. The aggregate gets stopped by the restrictions at the surface transitions. After the flow rate gets reduced indicating that the stones are in place a concrete slurry now gets pumped in to plug the top of the failed well head.

The secondary well line now serves as the new well head for extraction.

I want to ask you to look at a few orders of magnitude so that the complexity will be more understandable:

- water depth ≈ 1500 m

- drilling depth ≈ let say at least 500 m since the drilling has to be directed on a radius and the drilling tubes cannot be bend as spaghetti

- diameter of the drilling column let say 0.300 m (about)

The ration L/d= 2000/0.3 ≈ 6666,7 ! ! ! ! As equivalent consider a wire 0.1" with a length of 55.5 ft.

At the end of this "item" there is a tool which has a limited life because of wear and has to be changed so that the whole "item" is pulled up in sections of about 20m and stocked till the tool can be reached and changed, then the whole process is made in the inverted order and a new drilling cycle begins.

The target is only 0.3 m wide, or if a partial intersection is accepted may be 0.5m and this is all. So that one has to pinpoint a 0.5 m target at 2 km distance with a "wire", do you think that it is a simple operation one can afford to do twice at each well attempt?

Do not forget that when a drilling is done not always the oil is present at the end, only part of drillings are productive and economic.

You imagine the difficulty now?

How many offshore wells work WITHOUT problems? How many are in the gulf still working without troubles? How much am I willing to pay for a gasoline gallon?

Those are questions one has to answer in the analysis of such a situation. Of course more safety has to be provided from the start but is a second hole the solution ? or any other approach which is safe "enough" but costs less?

According to the chief in charge of the coast guards there are THOUSANDS of wells performing correctly.

I completely agree that this is a daunting task that is cost prohibitive to be done for every well. It should not be considered until everything else has gone wrong and it will not happen quickly when it is needed. But I was merely trying to present a layman's understanding (I most certainly qualify as a layman on this) of the mechanics involved to plug the original well using a parallel well. The alignment complications are certainly staggering. However, from my understanding from the article I read this demonstration of precision drilling can and has been done.

GA for a neat diagram showing clearly the problem(s) as known at this time.....

I suggest we give you a visual 'izer qualification? Thanks

thanks.. but I didn't draw this one you know. I do like pictures though.

I hope you do not mind that my answer covers as well your question as the previous one.

In this case statistics have to play a role ONLY as failure probability but this probability is established under functional conditions considered as "normal" even if this covers several classes of normality. In this case it is a totally abnormal working condition which I consider as "accidental". Due to the extremely low probability it is impossible to match design with what happened. The load of a full platform falling on the spot is so unprobable that it was not considered as possible especially because the platform will under "normal" conditions float!

Between the drilled hole and the tubing there is a cement poured in the gap so that the seal is over a long depth. Depending on the drilling depth there can be several tubes one in the other so that the possibility to have a leak at this level is low probability but of course not nil. I do not know all details and also as mentioned previously I am not up to date so that it is only an assumption. It is not possible with the informations we have to discard any possibility. It would not be fair.

With respect to the other problem. Let us assume that it was not enough electrical energy available. This can be the result of a protection destruction due to the accident. If a short appeared some where even if an other source is connected it can not help since the energy will be wasted via the short. If the high pressure was lost then the pumps have to be reactivated to fill back the accumulators. Again the problem of circuit continuity arises as a possible limitation.

I apologise but my knowledge has limits and I am not ashamed to say and accept it.

As I mentioned in a previous comment: "the more you know the more you know that you will NEVER know enough". Who realises this has the humility to accept it.

I LIKE IT:-

As I mentioned in a previous comment: "the more you know the more you know that you will NEVER know enough". Who realises this has the humility to accept it.

Well put and VERY true....

Thank you, it is my philosophy since many many years it was and is my guide line!

Yes, very nicely done. A GA from me also.

And I thank you for your support.

When compared to an Apollo space craft or a nuclear submarine, I would like to think a lot of money, time and effort was spent on developing the Bop.

Yes clearly a lot of effort has already been spent on developing any BOP. On another thread, it was found that the drill sting still lies inside the pipe section that is this BOP. Now certainly this would not be an unexpected condition to have the drill string in place when a BOP was ordered to close. But possibly a very durable part of the string was exactly at the point that the BOP was supposed to close. This is of course all speculation.

BOP technology has been under development since the advent of drilling into volatile subterranean liquids and gases. Having worked with a professional to model a design for a small ram bop, I can say there is a great deal of detailed knowledge of hydraulics and seals, in a specialized function, that goes into them. Every one is different and has different properties.

The higher the Working Pressure, and the larger the bore, then the cost goes up exponentially, due to material casting costs, magnetic and radiographic testing, large bore seals, etc. it gets very complicated.

Chris

A blow out prevention valve is essential to all flowing wells whether it is an oil well or a water well. That being said, I have no experience in oil well drilling but have lots in water well drilling. I am assuming that some of the technology applied in water well drilling will also apply to oil well drilling. Many of the techniques used in water well drilling come from the oil well industry.

In the water well industry, when a flowing well is encountered it is mandated that the flow be directed so that it all flows through the well casing.You might think that all the water in flowing wells will flow through the casing. Not always true. When a well is drilled, the cased portion of the well (this is the section of the well that contains the well casing usually steel but sometimes plastic) is drilled as an open hole with a diameter slightly larger than the casing. The space between the casing and the formation is referred to as the annular space. Sealing of the annular space is the most critical part of drilling the well. If it is not sealed properly in the water well construction, surface contamination can readily enter the well. In the case of a flowing well, water can be force upwards under hydrostatic pressure through the annular space. This flow will cause headaches for the owner and the well driller. If there are exists other water bearing zones or potential seasonal water bearing zones, then the flow will enter these other zones and possible mix or cause contamination of the desired zone (think salt water for example). The lower water bearing zone can then travel to other sites and find its way into neighboring wells or to the surface through geo-hydrualic connections (a network of fractures in the bedrock formation). After spending several years inspecting water wells with a government environment agency and later my own business, I can say that the failure to properly seal the well casing (read annular space) in water well construction can account for the majority of well failures. The only fortunate part is that the cased portion of the water well in bedrock formation is generally less than 30 feet with the balance of the well being open hole to the water bearing zone. Insufficient casing can present its own problems.

If I can relate this information to the oil well industry. I would think blow out can mean that the annular space has been compromised and oil is now flowing up the annular space. This type of blow presents the worst case scenario in the recent Gulf incident. That would mean the oil could now find different passages to the ocean floor. I can only hope the annular space is intact.

The cure for the problem is now beyond a failed BOP. Instead I think designs should center on the abandonment of the well. This involves the removal of all drill string equipment. A tremie line will now have to be forced down into a cleaned out well (not an easy task) and the well be filled with a proper grout from the bottom (near the oil bearing zone) to the top of well at the ocean floor. The grout must be sufficiently heavy to prevent the flow of oil. Pulling the old well casing is generally not possible.

I do hope the engineers working on the problem get it right and soon. I am sure they are aware of the issues at hand. This post is based on potential issues and not real issues that may have occurred on the Gulf well. I wish this team of experts well and good luck.

All,

Out of curiosity, what would the pressure in the BOP have risen to had it closed on the Deepwater Horizon well head?