Insider View on Causes of BP Deepwater Horizon Blowout – Spill

Posted: June 10th, 2010 by: h2

Here’s a reasonably new anonymous sourced insider view of what probably happened to cause the BP Deepwater Horizon blowout/spill. This was posted in today’s The BP Deepwater Oil Spill – the Hurricane Season – and Open Thread comment thread.

I will try to find a more original source on this story, but for now, here’s the raw report. [Update 1: source may be forums – SEDCO445 – May 31st, 2010]

Peter B. on June 10, 2010 – 12:39pm Permalink | Subthread | Comments top

The following analysis was first (?) brought to light by tvhawaii on June 4th.

The original source is from elsewhere on the web.

The original author felt the need to preserve his anonymity…. and judging by his familiarity with the event, he may have had very good reason to wish for same.

(I am NOT the author, and do not have any related information. I just thought bringing the article back to light would be of use.)

Peter B.

(Anonymous Source)

The following is my theory on what happened on April 20th. I have listed factual information to the best of my knowledge, and base this theory on 33 years of experience working on these rigs, with 16 years working as a consultant worldwide. The contractor (Transocean in this case) typically does not do anything without direction and approval from the operator (BP in this case). I believe that there was nothing wrong with the BOP, or the conduct of the crews prior to the catastrophic failure. If any operator drills a similar well using the same flawed casing and cement program, the same results will be very possible.

The well was drilled to 18,360 ft and final mud weight was 14.0 ppg. The last casing long string was 16 inch and there were 3 drilling liners (13 5/8″, 11 7/8″ and 9 7/8″) with 3 liner tops. A 9-7/8″ X 7″ tapered casing long string was run to TD. The bottom section of casing was cemented with only 51 barrels of light weight cement containing nitrogen, a tricky procedure, especially in these conditions.

The casing seal assembly was set in wellhead and pressure tested from above to 10,000 psi. Reportedly, a lock down ring was not run on the casing hanger. The casing string was pressure tested against the Shear rams, only 16.5 hours after primary cement job. A negative test on the wellhead packoff was performed.

The rig crew was likely lead to believe that the well was successfully cemented, capped and secured. Normally a responsible operator will not remove the primary source of well control (14.0 ppg drilling mud) until such conditions were met. However, the crews were given the order to displace heavy mud from riser with seawater, prior to setting the final cement plugs. They were pumping seawater down the drill string and sending returns overboard to workboat, so there was limited ability to directly detect influx via pit level. This is the fastest way to perform the displacement operation, and the method was likely directed and certainly approved by operator. There was a sudden casing failure during this displacement procedure that allowed the well to unload, with ignition of gas and oil. Evidently, the crew was able to get the diverter closed based on initial photographs, showing flames coming out of diverter lines.

It is likely that pressure built up between the 9 7/8″ and 16″ casing under the casing hanger, due to gas migration from the pay zone. Based on reported mud weight, the reservoir formation pressure is in excess of 13,000 psi. The pressure building in the cross sectional area below the casing hanger would have increased casing tension and caused casing to collapse and part (rapidly separate) at a connection, probably a joint or two (50′ or 90′) below wellhead. The collapse pressure for 62.8 ppf 9-7/8″ casing is +/- 10,300 psi. However, the collapse resistance of casing is considerably reduced in presence of axial stress (i.e. tension). Engineers – see formula from API bulletin 5C3, section 2.1.5 and run the math. The well then came in violently through parted casing and caused the blowout. Without lockdown ring on hanger, the casing hanger and joint(s) were slingshot up into BOP. That would explain why all components of the BOP are unable to seal or shear. The parted casing section remains across all BOP ram cavities and probably all the way up into the riser.

Shortcut #1: Running a tapered long string rather than a liner with 9-7/8″ liner top packer, followed by tieback string and pumping heavy cement all the way to seabed. Perhaps the original permits for this casing program were based on a planned appraisal well, and changed midstream to a producer well, then hastily approved by the complacent or under-staffed MMS. This tragic shortcut may have saved about 1.5 rig days.

Shortcut #2: Insufficient time was used to cure the mud losses prior to cementing the open hole reservoir section, depending instead on using lightweight cement to prevent losses to the formation.

Shortcut #3: The nitrified primary cement job. This is difficult to pull off, even under ideal conditions.

Shortcut #4: Hanger without lock ring may have used due to the previously unplanned long string, and to avoid waiting for hanger with lock ring to be fabricated or prepared.

Shortcut #5: No cement evaluation logs were performed after a job with known high calculated risk (mud losses to formation). This shortcut may have saved 8 hours of rig time.

Shortcut #6: Pressure testing casing less than 24 hours after cement in place can expand the casing before the cement is fully set. This shortcut can “crack” the cement and create a micro annulus which will allow gas migration.

Shortcut #7: Displacing 14 ppg mud from 8000 ft MDRT with 8.7 ppg seawater, less than 20 hours after primary cement is in place. How many tested and proven barriers can you count? I count zero satisfactory barriers. Industry standards dictate that at least two tested (to maximum anticipated pressure) barriers are in place prior to removing the primary source of well control (weighted mud or brine).

There was a response to this report in the same thread from aliilaali, which answers some of the above questions. He goes down the points, which is a nice counter-view, cool stuff if you are trying to really understand what is happening.

aliilaali on June 10, 2010 – 2:15pm Permalink | Subthread | Parent | Parent subthread | Comments top

it is a good analysis but then again hindsight is 20-20 …..but most of these points can be argued both ways…..

shortcut#1 – there is really nothing wrong with running a tapered string long as cement is pumped all the way down….similar approaches have been used all over the GOM …a cheaper approach YES …but not wrong from an engineering perspective

SC # 2 — this is correct

SC # 3 – there is nothing wrong with using nitrified cement…it is used all over again ….what has not come to light is what excess was used…cmt calculations need to account for nitrified cmt delivery ….simply put 30-40% excess is used….but this is a tested approach and not cause for concern in itself..infact in highly deviated holes it is the preferred option…why it was used in a vertical hole is a pause for thought but then nothing that is wrong here..

SC # 5 — again this has been misunderstood… has focused on the CBL log….now a CBL log only proves zonal isolation for a completion job….NOT to check for cmt integrity and if used the results are always ambigious …..because here a nirtified cmt delivery was used…the problem with this process is a higher than normal chance of micro-annuls forming…a CBL log DOES NOT check for micro-annulus ….so how a CBL log would have helped here I dont understand….a leak-off test is the only correct option here….would have been a good practice to run a CBL here but not a requirement…

SC # 6 — this is correct and the major cause for concern….nitrifed cmt delivery requires a longer WOC time (waiting on cmt) than the normal cmt approach ….the small WOC time is what started a hain reaction of problems IMHO

SC # 7 – correct but again ties into WOC times…the displacement of the mud without good WOC times again was not the best idea but the returns should have been monitored …especially since the company man knew he was cutting it close …the company man should have been HIMSELF watching returns …anything bubbling downhole could have been countered if needed

most of the things are bad judgment calls with nothing to do with bad processes being employed …just my 2 cents ….once a well blows and leaks like this one…looking back every decision seems like the worst possible decision …
aliilaali on June 10, 2010 – 3:23pm Permalink | Subthread | Parent | Parent subthread | Comments top

good centering is important because it minimizes axial loading on the casing ….this was a vertical hole and I am not surprised BP did not agree to 30 ….10-15 is what most operators would agree to in vertical holes …..also it makes sense since nitrified cmt delivery takes care of tighter clearances …..the way nitrified cmt delivery works is …nitrogen is added to fluff out the slurry ….this make the slurry more mobile and can move well through tight spots and find its way along ziggy zag flow paths ……

I would bet my 5 bucks …if the caliper log is released ….it will show a crappy hole …combine that with the 7 centralizers ….nitrified cmt would really make sense here then …
aliilaali on June 10, 2010 – 3:57pm Permalink | Subthread | Parent | Parent subthread | Comments top

just ran a quick calculation …..out of 1,192′ of open hole drilled with @ 8 3/4″, running a 7″ csg, volumetircs come to 0.0268 bbls/ft so running up a total of 1,192′ the volume of cmt needed is or a total of 31.9456 bbls (32 bbl) of cement required to fill “newest hole dug” annulus.

so 32 bbl of cmt required and 51 bbl of cmt were pumped in ….equates of excess of 19 bbl or 59% excess……now this is plenty of excess cmt by any standards ….atleast on paper things seem right……
in real life drilling —- the caliper log is needed here to see what the actual avg dia of the open hole was ….this calculation assumes a good consistent caliper log which is hard to imagine given the well history in march and april ……so if the caliper log is not available more than likely this whole calculation is not worth putting any comfort in ……same caulcuations need to be run with using the avg hole dia from the caliper log to work out if enough cmt with enough excess was pumped in …..
tollertwins on June 10, 2010 – 4:06pm Permalink | Subthread | Parent | Parent subthread | Comments top

Halliburton apparently documented their disagreement w/ BP on this point saying that ‘they could (would?) have SEVERE gas flow issues’ w/ the fewer number of centralizers.

Media reports of the hearings said that:

a) yes – very small clearance between hole and casing (something like 3/8 – 3/4″).

b) HBT recommended 21 centralizers. There were only 6 of the right configuration on the ship – so they went w/ 6. So HBT covered it’s hind assets w/ the above documented statement.
aliilaali on June 10, 2010 – 4:40pm Permalink | Subthread | Parent | Parent subthread | Comments top

TTwin —

3/8 – 3/4″ seems like a little exaggerated …and i dont really see how anyone can come up with this number or even this tight range without looking at the caliper log ……

given halliburton advised 21 centralizers …..again points to a high chance that the caliper log showed a hole in bad shape ….now the only way for a csg to run in hole is a hole that is wider than what is expected and not narrower 🙂 … IMHO i can say with a high degree of confidence that 3/8″ – 3/4 sounds unreasonable and probably something CNN would report given their outstanding job of pulling information out of their ass interms of this whole incident ….

a disagreement on centralizers between operators and contractors is very common really…….if I was the drill engg and halliburton wanted 21 on a vertical hole….i would need some convincing too…. not to say its not happened…I ran 38 centralizers as recent as 4 months ago on a less deeper hole but i needed to be convinced by schlumberger …took not their field techs but a senior engineer to convince me …point being such instances are common and in itself nothing thats not part of routine work…thats why there are techs , engineers and senior engineers to provide many heads and levels of knowledge to counter such problems at different levels just like any other industry

2 Responses to “Insider View on Causes of BP Deepwater Horizon Blowout – Spill”

  1. Ann says:

    were you able to identify the source of this information? I would love to know “the rest of the story”.

  2. h-1 says:

    No, I wasn’t, this event is moving too quickly to spend a lot of time on the various stories that pop up daily.

    See today’s revised estimates of 35-60,000 barrels per day spilling into Gulf, for example.

    There’s ongoing discussion of what was really the main problems among some oil guys at, but one thing most seem to not disagree about is that the mud return flows were not being watched properly, probably because they were in a hurry to get things closed up. But as has been noted, it takes no time to watch for things like that. That’s not the only problem, there are a lot of others.

    I was a bit skeptical when people started waving around that ‘worst disaster in US history’, but now it’s starting to look like that might well be the case, especially if the new oil lines fail to capture all the oil from the BOP unit before it spills into the Gulf waters.

    In general most oil field guys I read are pretty concerned about this, and are providing a lot of insider type background information. Unfortunately, the massive size of the liability of BP is probably putting a stop to any further leaks from BP people who were on the rig, unless BP tries to throw them under the bus, then they might start talking.

    But there’s been a series of lesser revelations that are really making BP look very bad, no denying that.

    I’m paying attention to what the peers of such insiders as the one above say far more than to what bloggers like myself say, and I try to follow the most promising stories, but it’s literally a full time job, plus some, to even keep up with the basics daily, let alone all the documents that are now available at