[Pops]

Saw the news about a GOM blowout and wondered if our resident experts could explain to me exactly how drilling a relief well is accomplished. Seems pretty miraculous to be able to hit a 1-2 foot wide target in solid rock several thousand feet down.

[Oily Stuff]

I am not qualified to answer this except to say that most relief wells don't need to intersect that small a target. They simply need to get in the same formation the blowout flow is occurring from, and fairly close, for instance using typical steering tools, to be able to pump massive amounts of kill fluids (mud). In the case of the Macondo well John Wright, the Dally Lama of relief wells, was able to drill right to the liner shoe of the blowout well using something called magnetic ranging. The down hole tool on the relief well and the steel in the liner of the blowout well emitted opposing magnetic fields and allowed precise steering to the steel target. In the unlikely case a relief well is required in this current event magnetic ranging might be used also for intersection as the well was being sidetracked from existing casing.

Magnetic ranging is cooler than snot. It is used for avoidance purposes in drilling multiple wells from the same platform, for instance. I have seen subsurface maps of well bores drilled off the beach into S. Bay in S. California where many well bores are drilled off a lateral track and deviated out into the bay. Casing of individual wells is intertwined with each other, up, down and around other well bores like a big bowl of noodles. Magnetic ranging in this case allows them to avoid drilling into other casing strings.

[dinopello]

I can grasp (or guess) as to how the operators know where the drill head is in 3-D space. Probably some sort of telemtry system similar to DGPS which gives something like 2 inch precision. I'd like to know more about how the drill head actually manuevers (the mechanical design).

[rockdoc123]

Pops I believe you are thinking of the Macondo incident which was a bit of a special case where to do the remediation they needed to intersect the annulus of the original borehole. This is different than the example oily stuff mentions where it is sufficient to just bleed off the reservoir pressure.
The BHA is steerable and as mentioned above uses GPS in order to understand where it is. There is a magnetic locator device that can be used that can help steer them towards metal such as casing, but in general they have a very accurate deviation survey for the original well bore and given the accuracy of modern GPS technology it could be sufficient to just plan the coordinates and let it steer itself.

[dinopello]

The geo-pilot steers by "flexing its shaft"...

I guess this one got a little off course !

[rockdoc123]

Basically not using MWD (monitoring while drilling) or they would have known where the BHA was.
MY guess is the driller got run off, the superintendent probably as well.

Rockman or Oily can correct my observation here but that looks like an 8 inch bit....which means this was meant to be a pretty shallow hole.

[Oily Stuff]

The internal flex geo-steer system I understand is nifty; I love the animation. Some well bore profiles off platforms look exactly like that. Cool.

The other video is old and I have forgotten the story behind that. Something reminds me it was kind of a hoax (note the dirt levees and that the hands were waiting for the bit to breach the surface) but I am not sure. As Doc eluded to earlier steering is pretty precise with real time plotting of the BHA location, etc.; I could not in my wildest dreams imagine how that would even be possible, unless intentional, but...I noticed the hands were speaking what sounds like Russian. Anything is possible over there.

Hoax or pilot error of immense proportions lets hope the folks in the Marcellus back East don't see this...they'd have a cow, yes?

[ROCKMAN]

Geosteering capabilities: Just last week I geosteered my well in Texas and hit my target just right. But I had a very big target compared to hitting an existing well bore. At 4,600’ I was able to keep the drill bit within a 5’ vertical window and at about 1 mile away I was left about 20’ from the TD target. Accurate enough for this EOR effort.

But had it been necessary I could have been much more accurate if I had spent the time/money. With the geosteering GPS-like systems doc describes I could put a drill bit thru a car’s front window 2 miles down. The trick is that I need a very accurate location of that windshield. Which is why they use magnetic locating systems as OS described. They will never have the location of the well bore they’re trying to intersect defined much better than 10’ +/-. But by knowing were the casing is (left or right of the relief well [projected path) with the sounding system they can make the final adjustments. But it’s a slow process and if they are too far off course they might have to plug back the RW and sidetrack to get close enough on a second attempt.

From the radio report this morning sounds like there is still some NG leaking but isn’t burning. If the report that the production platform collapsed is correct then the leak is probably come from close to the seafloor. If so it may be nearly impossible to cap it and the RW will have to be drilled.

Also heard a report this morning that Halliburton pleaded guilty to destroying evidence regarding the Macondo blowout. The data was two computer simulations of their cementing process. Typical bone head move we’ve seen so often in politics: the cover up effort is worse than the original incident. As I’ve mentioned before Halliburton didn’t decide what cement to pump or how to pump it on the BP well. They make recommendations but BP makes the final decision…not Halliburton. If H. pumped the correct cmt chemistry and in the manner that BP approved they aren’t liable for any failure. Which is the exact wording in every cmt contract written by every cmt company: if they pump the cmt as determined by the operator in the manner the operator dictates they are not responsible for a cmt failure.

Cmt failure is a very common incident. Happens hundreds of times every year in the Gulf Coast. And often the effort to fix a failed cmt (a “squeeze job”) doesn’t work the first time. I once had a well that was squeezed 23 times to fix one failed cmt job. Operator finally gave up and plugged that $48 million worthless hole in the ground. What’s strange about the H. story is that while the may have erased the two computer simulations it would have been a simple matter to run them again. These are rather standard computer programs and not only could the govt have H. run those two models over again they could have done so dozens of times by varying the inputs. I suspect H. either just paid up to make the continuing bad PR go away. Or perhaps they actually didn’t follow BP’s directions for the cmt job. If that were the case and BP could prove it Halliburton could lose a great deal more.

But what will always be difficult to understand: now everyone here knows just how common cmt failure is and now you know that everyone in the oil patch is acutely aware of the possibility. Then why, knowing they were intentionally putting the well in an unbalanced state and that a bad cmt job could lead to a catastrophic event, weren’t they monitoring the well for a kick which would have indicated a potential blowout was coming up the hole. That’s SOP even when you have a balanced well and no concern about a bad cmt job. There was an argument amongst the drilling managers on the rig whether the test info on the cmt was correct or not. And they still didn’t monitor for a kick.

[Pops]

Thanks all.

GPS or DGPS signals are line of sight, how does the bit assembly down the hole receive signals?

[dinopello]

Thanks all.

GPS or DGPS signals are line of sight, how does the bit assembly down the hole receive signals?

At the time I guessed that I was assuming there was some lower frequency version of DGPS that could penetrate - but that probably isn't feasible at these depths and for other reasons. Accelerometers/IMU could probably get pretty good accuracy over short term. I've been involved with robotic systems that can navigate for quite a while just altering its position perception based on optical flow from cameras it carries (e.g. navigate turns for a mile and know where it is to within a few feet). I doubt cameras are in play here, but IMU may be.

Just found this wiki...

[Pops]

Very good Dino, thanks, I looked but couldn't find it.

[ROCKMAN]

Pops - that's why I said "GPS-like". Think of something like an internal inertial guidance system. For more details you can search " directional drilling assembly".

[John_A]

Thanks all.

GPS or DGPS signals are line of sight, how does the bit assembly down the hole receive signals?

MWD tools do not use GPS. They use magnetometers and accelerometers to determine direction and angle off of vertical. Rockman has probably been around long enough to seen holes drilled using the original single shot technology, which when combined with whipstocks and particular configurations of the BHA were how directional holes were drilled. Steering tools, a wired device stabbed into a receiver on the backside of the BHA was another older technology and the signals were sent back to the surface via wire. Problem with that is that you had to keep stabbing and unstabbing as drill pipe connections are made.

The modern world in MWD arrived with the MWD tool actually sending mud pulse telemetry back through the mud column which could be picked up by transducers in the standpipe on the drill floor. It generated its own power downhole using the mud flow the same way a dam makes electricity, and it constricted the flow of mud to create pressure pulses. These pulses, sort of like morse code, were then decoded by some electronics and delivered an inclination and direction. By stringing these together at the same distance as a single pipe joint, all of those changes in direction and inclination could be assembled into a path.

Modern MWD can send continuously while drilling, the old analog gear has been replaced by digital stuff, the signals (which to the layman just look like noise from the mud pumps) are much shorter so there are more of them packed into every second of time. Some tools might still be battery operated (Sperry Sun used to have quite a few of these which were junk) and with modern mud motors and bits, decreases in drilling time are pretty phenomenal. Once upon a time you couldn't rotate the mud motor, once upon a time you certainly couldn't change the orientation of a bent sub without tripping the pipe, there is just all kinds of new stuff allowing drilling ahead where once upon a time you would be tripping pipe. Biggest increase in "technology" has really been decreased drilling time, and the things which contribute to that.

[ROCKMAN]

Great details John. On my phone so difficult to post long answers.
Pops - what's really amazing is that some of the steering system have become so sophisticated that it reads the telemetry directly and correction it's course itself. Think of an autopilot that will automatically change it's course if the wind nudges it off course. And all this done with tools that are typically less than 5" in diameter and sometimes as much as 3 miles from the drill floor and human intervention.

How much has drilling control changed in my 38 years? Today we can hit a coffee can horizontally 2 miles down. In 1975 I would drill a straight hole and hope I would hit within 500' of my target. And sometimes I wouldn't. And this was just trying to drill straight down.

[Pops]

Rock I was confused because "GPS-like" seems to infer that the BHA knows where it is somehow (GPS) and reports its location back to the surface. Dino's link explains that the BHA doesn't report where it is but where it's going. Those reports stitched together tells where it went and so where it is.

MWD tools are generally capable of taking directional surveys in real time. The tool uses accelerometers and magnetometers to measure the inclination and azimuth of the wellbore at that location, and they then transmit that information to the surface. With a series of surveys; measurements of inclination, azimuth, and tool face, at appropriate intervals (anywhere from every 30 ft (i.e. 10m) to every 500 ft), the location of the wellbore can be calculated.

Gotta go slow for my little brain, lol

[ROCKMAN]

Pops - actually the bha does report where it is. Ever time I drill 30' and stop to add another jnt of drilling pipe the bha does another "survey" showing how it has changed location from the last survey. It sends back the hole angle and the azimuth of the bha. With that info we can calculate the change in position. Imagine you are standing at Location A and blinded folded you walk 30'. I tell you that you walked 5 degrees to the left and walked up hill at a 2 degree angle. With a little trig you can calculate your new position. But for you to be able to know where you are in absolute terms you need to know your exact location before you started your walk.

So from the vertical hole before we start the direction drilling we run a gyro survey so we know the exact location of that point. Even though in theory that begin point of the direction curve should be the same as the surface coordinates it never is: all well "walk" a bit from that vertical projection. And sometimes a lot.

One more complication. The survey while drilling is taken about 40' feet behind where the drill bit is so we have to project that last footage. But we won't know how accurate that projection is until we drill about 40' beyond that point. I know tha does't seem very far at the end of a 5,000' to 10,000' hole but it is especially when the hole able is at 86 degrees or more. At those angles to change the depth of the hole just 1' might take drilling 100' of hole. Geologists that haven't worked in this world have a very difficult time. Most don't have a goog grasp of trig at such high angles. Long ago I made a small poster to hang on my wall to remind myself: "The world get weird at 87 degrees...deal with it". LOL

[Pops]

lol, OK. Thanks.