BP Deepwater Horizon Relief Wells
Posted: May 31st, 2010 by: h-2
Before you say to yourself, oh, this will all be fine, consider the complexity of achieving success with relief wells. As this author has already noted, it can take many weeks to mill through the original well casing once the relief well intersects it at around 18k feet. And that itself is hardly guaranteed, these wells are hard to drill, and frequently fail.
This guy considers, if the first relief well succeeds, a target date of around September to be realistic, and August to be too best case to be considered as a real goal. And that’s IF and ONLY IF the first well A: intersects the existing well, and B: doesn’t fail in the process.
In the Ixtac blowout, it took 10 months and numerous relief well attempts before they succeeded.
Here we have another oildrummer drilling guy, aliilaal, explaining a bit about why Relief Wells (RW) are so complicated. Keep in mind that the relief wells are what will actually fix the problem, everything else is just a stop-gap measure.
aliilaali on June 1, 2010 – 12:36am Permalink | Subthread | Comments top
idle thoughts on problems in hitting the target with RW ….had said my 2 cents on wellbore surveys and now form the RW standpoint there are some things to consider with regards to RW target
1- like i’d said ….current technology’s theoretical limits can hit a 10ft radius ball with a confidence interval of 90%…this limit applies on RW with with a grain of salt since expected interception is 18000 rkb ….but really depth can be +- 50 ft on depth …the problem here is azimuth of RW (think 3d geographical grid) ….so essentially the target for the RW is not a circle but a rectangle (in cross sectional view of leaking well when looked at from right or left) of approx 75′ (length) x 2′ (width)
2- now there are two options to establish pressure communication b/w RW and LW (leaking well)…(1) mill into the LK csg or run a hot tap (pull along LK and run a perf gun)….high pressures will likely preclude a hot tap and most likely it will be the milling option
3- for milling we have to hit the LK with our milling bit and hit in a way so that atleast 25-30% of the cross-sectional on the LK makes a connection with the milling bit …to be able to mill and expose sufficient portion of the LK to pump the kill pill..this means the target is really a rectangle (cross-sectional view) of approx 75′ X 1.5′4- the Mississippi delta is a known have a high iron content in the formation…this will screw up the gyroscopes and the accelerometers a bit….small errors on the instantaneous will add up over 18000’….
5- accuracy will be increased in the last 3/4 joints of the DP using wellspot RGR (radial gradient ranging) in combination with PMR (passive magnetic ranging) ….this will get a good lock on the target zone and time drilling will be used ….(the driller will need sufficient supplies of Gatorade, Copenhagen/skoal/husky or whatever dip he prefers but fine cut and complete concentration here)
git er’ dun
And here’s an earlier comment he wrote explaining this a bit more (but man it would be nice if he could spend a few minutes writing it out legibly)
idle thoughts on wellbore surveys and how to work out where the original well is and how to guide the RW to the original well
a wellbore survey is the location of the wellbore in reference to its geographic grid (essentially a Cartesian 3D space where the exact location of the hole is determined with relation to the wellhead using certain sensors)…..there are two parts to this ….first is the inclination (which refers to its inclination form a true vertical) and the second is the azimuth (direction with respect to the grid) ….these measurements at taken repeatedly at the instantaneous to collect information of where the bit is at anytime…..this can be used to generate a 3D plot of the wellbore….most common sensors used here are multi-axis gyroscopes where the where the primary axis of reference is the spin-axis aligned with the BHA axis…two further axis are aligned in an orthogonal setup in relation to the spin-axis and to to each other…now this gyroscope can be rotated physically in the BHA and knowing the original spin-axis alignment multiple offset axis can be created essential to accuracy when the dog-leg severity of the wellbore is any higher than 4 degrees/100ft.
the other important sensor here is the accelerometer (each BHA has 2 of thse) and each has two axis …the primary aligned to the BHA-axis while the other is orthogonal….each of these sensors again has the capability to be physically rotated in high dog-legs and off-set axis generated as required….now that you know where the well is in relation to the wellhead at any point you can attempt a relief well using directional drilling…now all this has to be corrected for depth as the earth’s magnetic field interferes with all instruments and the deeper you go the more it interferes…..while it is relatively easy to correct ….very small errors at the instantaneous add up when you are talking 20000 odd feet of weelbore…..at present accuracy is such that a 10ft radius ball can theoretically be hit with a confidence interval of 90%…..but in pratice its another story…and the accuracy is a function of multiple parameters…..amount of trace iron or other magnetizable elements in the wellbore like cobalt , nickel…the compaction of the formation and many more….so actual accuracy is a little less than the theoretical….now the Mississippi delta has above average iron in the formation…something that doesn’t help
Some updates on this discussion
And so it goes, but it’s good to have real understanding and not the standard corporate mainstream media garbage stew of sound bites and 10 second video clips, don’;t you agree?
Is there in fact any published or unpublished (such as MMS-filed data required of BP) giving the TVD of the blown well, and its accurate deviation (inclination and azimuth) at specific depths below the seabed–such as 5K feet/7.5K feet/etc.?
It’s seems quite strange that in all the screaming and shouting and nonsensical media hyperbole, there’s not been published any detailed technical data concerning the bore itself. In past major disasters, such as the Challenger and Columbia episodes, the New York Times and other major media had extensive recitations of the failure modes??
Completely agree with Robert Mandell’s last point. As to the technical aspects of hitting the target well, I wish to make the following point: It may be better to consider a strategy whereby a first relief well hits the formation at some distance (of the order of a few tenths of feet), and after completing this well, to inject water into the formation at a rate that will ensure the encroachment of the original well by water and eventual cut-off of HC production. At that point water injection has to continue until a second relief well has successfully intersected the original well, allowing to proceed with cement injection. Such a strategy might well save a lot of time and limit the spill substantially. The idea is not new and has been applied successfully in the field. See: F. Lehner & A.S. Williamson, SPE Jrl., Aug. 1974, pp. 321-329.