Live Feed from the ROVs working on the “Top Kill” procedure on Deepwater Horizon.
I don’t really know what I’m seeing but I find the motion oddly calming.
Other sites with feed or commentary include:
There are plenty more to view - see here for listings in the latter portion of the first ‘article’ (one has 7 different views at once)
Kinda neat to see changes occurring in the riser/BOP output as things happen.
I hope the top kill works, and it appears to have temporarily slowed or stopped the release of oil and gas, but right now I can think of a few reasons it may not succeed in the long run.
What follows is somewhat speculative, and is not based on any non-public knowledge of what BP is attempting here.
I’ve had a look at the BP live feed. The feed seems to show mainly weighted drilling fluid (based on the fluid’s color) escaping from several points. The live caption superimposed on the image is not clear enough for me to read, but I can read the acronym “LMRP” so it’s likely that the view is of the top of the Lower Marine Riser Package (including the BOP stack) of the sunken Deepwater Horizon. I’m pretty sure the view is not of the riser end, which is where the previous live feeds have mostly been located.
If this is the case, what’s happening probably is this: drilling mud is being pumped at a high rate into the wellhead, via high-pressure choke and kill lines attached to the BOP stack. A large amount of the injected fluid is escaping from existing or newly created leak points in the riser just above the injection point, as this is the path of least resistance.
The combination of dense fluid and high circulating pressures is pushing back against the oil and gas flowing from lower down in the well, but at this point it’s unclear whether enough back pressure can be generated, given the rather large leaks apparent in the video, to ensure that the injected fluid can push the oil and gas all the way back to the formation it is flowing from.
Over time, the leak points are going to become larger as the solids in the drilling fluid erode the metal bordering the holes where the fluid is escaping. This may be a matter of hours, or a few days, but it will happen. In addition, the riser may fail completely if the back pressure within it exceeds the pressure rating of the connecting flanges of the riser sections above the wellhead, which, IIRC, may be around 500 psi, i.e. not giving much leeway for this application.
It is not necessarily required, however, to push the wellbore fluids all the down the hole at this time. It’s possible that the engineers believe that if the top of the wellbore fluids can be lowered as little as a few hundred feet down from the wellhead, they could take a shot at injecting cement via the choke and/or kill lines to set a plug at the wellhead level. Even if this works, though, it would not be particularly secure, so in my estimation it will still be necessary to complete at least one of the relief wells so that cement can be injected at a lower point in the well to provide a permanent seal.
The above, I believe, is a reasonable basis for why BP is saying that definitive results may not be known for 48 hours or so.
Some discussion of that here:
http://www.theoildrum.com/node/6513#more
I think you are right, they are wanting to put their plug a good bit lower than they actually need to, and I don’t blame them in this situation. It’s certainly a fascinating thing to watch unfold.
OK, as of 06:45 central the live feed has changed to a different view, and the general setup is much clearer. Now I’m looking at a large ROV viewed from above, with the top of the blowout stack and part of the riser curving off to the right. The ROV in the image is where the feed was coming from earlier in the day. Looks like they are still pumping mud; they may have reduced the flow rate somewhat. There are large leaks originating at the top of the stack; could be mainly mud, could be mud and oil; bit hard to tell. If it’s mainly mud, good news. If mud and oil, bad news.
Things are looking up.
From here.
Yes, the holes in the stack don’t seem to have gotten much larger overnight. It is beginning to look as if this might work.
I am seeing reports they were successful!
Top kill’ plugs gulf oil leak, official says.
I sure hope so.
It looks like it worked from the feed…
I am cautiously optimistic. I don’t think we will really know until the cement plug has had time to set, probably another 24 to 36 hours.
Halliburton could do that job in twelve hours! 
What are the downsides to this method? i.e. why was it not the first choice and acted upon immediately? Does it close the well forever making the remaining oil unpbtainable?
First off, nobody (outside of BP) cares whether or not this particular well is usable in the future or not. In fact, BP is taking steps to ensure that this well is permanently sealed forever, efforts that will continue even if this “top kill” method seems to be working.
The downside of the “top kill” effort is that there is a chance of making the spill worse. The high pressure drilling fluid could actually increase the size of the leaks. Also, this method was thought to have a relatively low chance of working. The most optimistic estimates of success have only been something like 60-70%.
BP’s efforts to date have been focused on attempting to manually shut the blowout preventer, and putting two different collection devices on top of the well head. All of these methods were thought to have a higher chance of success, but ended up failing. BP has also been drilling relief wells in an effort to intercept and sever the original wellbore, but this takes months.
From here:
Also, from here:
To obtain the remaining oil in the formation only takes the installation of another well, hopefully one with a functioning blowout preventer this time.
I’m confused. The live feed still shows oil spewing from a broken pipe. Why are we saying the top-kill has worked, again?
Post 5 addresses this. If you’re still confused after re-reading it, let me know and I’ll try to explain further.
Aaaah. I see. Thanks.
Damn, not working - too much mud coming back out. 
Seems they’ve temporarily stopped the operation.