Crystals and the Gulf of Mexico Blowout

The situation in the Gulf, in regard to the Deepwater Horizon fire and oil spill continues to evolve. Since my last post on this, the size of the operation has continued to grow, with some thought now being given to dredging barriers along the coastline. What I thought I cover a little today, among other things, is the problems that occur when oil leaves a hot reservoir and suddenly enters a pipe on the floor of a very cold sea, causing some of the constituent chemicals to crystallize out. To illustrate this, I took a picture of one of the pipes that used to sit in my office, since it helped visualize a problem we had been given, and is along the lines of that now facing BP.

Barium Sulfate Crystals growing in a 3-inch I.D. pipe that carried oil from the deposits under the North Sea. (Scale is in cm) (Growths of this size can occur in less than 24 hours).

You will note that the crystals have reduced the effective diameter of the pipe down to about 1.5 inches, cutting the flow the pipe can carry by 75% (roughly). These are but one example of a number of different chemicals that can precipitate out in the pipes that carry oil from the reservoir, where it hot, up through colder zones as it rises through the different pipes on its way to land. There are two different aspects to the problem that I want to cover since it relates to the formation of methane hydrates in the capture box that BP had fielded at the end of last week.


The first of these is on the growth of the crystal hydrates within the confinement box that BP had lowered over the middle leak in the riser. However, rather than dig back into my notes, I am going to suggest that those who want to understand the chemistry/physics go to The Obligate Scientist. He uses a plot from the USGS to illustrate the conditions that transition methane gas to methane hydrate as a function of temperature and pressure (depth in the ocean).

(Source USGS )

To which, to give it current relevance, the USGS have added what happens at the sea bed, exemplifying this with a seabed at 2 km deep, while in this case it is only about 1.6 km. However the temperature is still sufficiently low to cause the hydrates to form.

(Source USGS )

The temperature of the fluid is critical, as can be seen, but for the temperature to reach that level requires a certain amount of residence time, for heat transfer to take place. Unfortunately in constructing the box the way that BP did, while providing stabilizing plates that would stop the box sinking into the sediment, these also act as heat radiators, ensuring a very rapid heat loss to the surrounding sea and a faster drop in temperature for the gas, accelerating the formation of the hydrates.

As I mentioned above there are two problems that then arise. The first of these is the immediate constriction of the flow path, as the first figure shows, the second is that if the crystals were to break off, or grow freely to that size and then enter the flow channel, then unless they are all smaller than about 30-40% of the flow channel diameter, then they will jam together and block the flow channel. (I have photographs somewhere of a pair of particles, both nominally less than half a tube diameter wedged together in the middle of said tube and effectively blocking it – hence the rule of thumb).

So as the crystals grow they both reduce the flow channel itself, and also increase the risk of total blockage. Now it might be suggested that if the speed of the fluid were increased that this would not give the crystals time enough to form and cause the above problems. These are three, originally similar sized nozzles through which water was fed at 40,000 psi. (It is used in industrial cutting) The top orifice was totally blocked after 48 hours of operation using untreated tap water, the center orifice is largely closed after 104 hours using softened water, and the bottom orifice is still clear after 200 hours using de-ionized water. The water velocity was around 2,400 ft/sec and, as you may note deposits still grew.

After Hashish (Flow Industries)

Keeping deposits from clogging feed lines is thus a tad more complicated than it may first appear. And while I note that BP is now apparently planning on lowering a second smaller cap into position, there may need to be other modifications (such as an inner liner of Teflon to insulate the flow from the radiative metal of the box) that have been added (perhaps an internal heating system) to help alleviate the problem.

The Unified Command Center has noted the following update to the resources mobilized to help with the spill:

Total response vessels: more than 290

Boom deployed: more than 1.1 million feet (regular plus sorbent boom)

Boom available: more than 1.4 million feet (regular plus sorbent boom)

Oily water recovered: approximately 3.6 million gallons

Dispersant used: approximately 372,000 gallons

Dispersant available: approximately 180,000 gallons

Overall personnel responding: approximately 10,000

There was a “real time” transcript of the press conference that BP held at noon (6 pm in the UK) in which they commented that the second dome that they will try to cap the well with is going to be much smaller (about the size of a barrel) and that this should be tried within three days , and if that doesn’t work they are thinking of tapping into the pipe to drain the oil without seawater contamination. One of the problems, apparently has been the high gas content of the oil. Much higher than had been anticipated.

The box is being made smaller so that warm water (from the surface) can be poured over it to keep the temperature above that at which the hydrates form.

Now they've gone from "one extreme to the other". The "top hat" dome is 5ft tall, 4ft wide. Warm sea water will be pumped down. As they lower the dome it will already be operational - hooked up to a ship on the surface. It will be more difficult to place but they will succeed.

It should be in place by the end of the week.

The Gulf of Mexico oil spill, a review and some progress

The last two days has seen what appears to be some progress in the capping of the Gulf oil leaks. BP is currently lowering the 100 ton dome that is intended to cap and capture oil from the middle leak of the three. The open end of the riser was also closed on May 4th by fitting a valve to the end of the pipe and closing it. There are however, two different results from the different methods of treating the problem.

Plan for sealing the second leak in the riser.

The oil is flowing out of the well, through the Blow-Out Preventer stack (BOP) which is mounted on the sea bed and into the pipe that would have initially carried the oil to the surface and the drilling rig. When the blow-out occurred, natural gas flowed up that riser, and was ignited as it left the pipe, on the platform. To help with understanding I am going to include a couple of items that I have been sent to try and give a better picture of the overall situation as well as where we are today.

UPDATE: There is a new website that is now co-ordinating information for the Unified Command, and, among other things they have posted this picture from late last night.
Cap for the second leak being lowered below the surface.

The site also has pictures from one of the initial test burns.


Firstly, to help with some of the discussion on the structure of the rig there are a couple of pictures, first of a sister rig while aboard a ship, so that you can see the pontoons, and then with the Deepwater Horizon in the water.


When the rig is in the water the lower parts are submerged, and only the legs and platform are above the water.


When the fire began it was fueled by the initial flow of oil and gas that came up the riser from the wellhead on the sea bed, through the rig floor and engulfed the drilling mast. As I noted earlier, there is some indication that the ignition came after the power to the rig was cut, and then the emergency generators kicked in. However, initially, as can be seen from the form of the fire, the fuel was coming up the pipe from the sea bed.


Over time, the vessel began to develop a list, as the fire continued to be fought



And by this time, on the second day, the heat had burned out parts of the aluminum structure (see the helipad for example)


The rig, as you know, sank shortly thereafter.

I am going to add some illustrations to help those of us who don’t do this every day understand some of the terms and conditions that are referred to either in discussion or in the media. (Thanks to Roland V).

The first of these is the casing shoe, which is the piece of pipe that is threaded onto the bottom of the casing to help guide it into place.

Some of the components used for the casing and cementing of a well (Schlumberger)

The other area that I want to address is where the well starts on the sea bed. Remember that when we are drilling a well mud is used to cool the bit, clear the cuttings and stabilize the walls of the well. Well this normally flows up around the outside of the drill pipe, but when it gets to the sea bed it is captured between the casing that was initially set into the sea bed to stop if falling into the hole, and the high pressure drilling pipes. Thus when one looks at the area around the wellhead there are actually several pipes going into the ground.

The first of these is a relatively low pressure outer casing, that holds the surrounding ground in place as the well penetrates through it and down towards the rock with the oil in it. Within that there is a second, higher pressure tubing that will carry the fluids down to drill the well, which may be an additional 20,000 ft deeper (at 1 psi per foot rock pressure perhaps) and which must contain the oil and gas flow that will then come from that site. As an example, you can see, from this view of a Cameron assembly, through a window in the lower pressure casing to the higher pressure fittings inside.


It can also be seen in this drawing of a section of a competing design from Vetco

Vetco Subsea well assembly

Notice that there have to be seals between the high pressure drilling/production line and the lower pressure confinement casing/housing. Should high pressure fluid get into the space behind the outer casing, then it could be sufficient to crack it, and there is not always monitoring equipment in place, subsea that can check to see that this has not occurred.

There is also a passage so that the drilling fluid which flows down the central bore to the bit, can also has to come back out of the hole, and be confined as it is carried out of the well and back up to the rig where it is cleaned of cuttings and re-circulated.

It should be born in mind that the BOP sits on this assembly, and that the lowest leak is about 50 ft above the BOP, so that there are apparently no leaks in this section of the structure, as far as I can determine, at this time. The current condition of the riser above the BOP has been badly distorted by the collapse of the pipe to the sea floor. And the difficulty in fitting a cap over the assembly can be estimated from the condition of that area.

Photo of the top of the BOP (I have color corrected the image)

When the production casing is floated into the well the high-pressure drill pipe is removed, and the production casing floated into the well. It will have the assembly shown in the earlier picture, and will be sealed at the joints along its length so that the cement flows down to the shoe, and then fills the bottom of the well, and then flows back through the gap between the production casing and the rock walls, displacing the mud and filling that gap. The intent is to ensure that all the oil and gas will be forced to flow through access holes that will later be inserted through the production casing and the cement liner, using shaped charges.

There are a couple of considerations when the wells get this deep, the first being that the rock is very hot at that depth, and so the cement chemistry has to be carefully controlled to ensure that it stays liquid long enough to totally fill the gap that it is being injected into. The second area of concern lies with the pressure at which the cement is injected. Because the rock is under a lot of pressure initially, and can also be quite weak, the cement must be injected at a pressure that will be enough to drive it up along the gap, but not at too high a pressure. If the pressure is too high, then the rock around the well can be cracked. In that case some of the cement can be pumped into the crack, and the full volume fill may not be achieved, and the circulation path for the cement back to the well head may be lost.

The shape of the wiper plugs are slightly different to those that I showed in the earlier post.



The plugs are activated by first causing the plug releasing ball to be locked into the plug, sealing the flow passage, and then the pressure of the cement/driving mud will push the plug down the casing, wiping the walls as it goes. When the plug gets to the bottom of the well, the ball in the bottom plug is pushed out, and this allows the cement to flow into the gap (annulus) around the casing.

Some of this additional explanation, as I have said, is more to help fill out the background to the ongoing story.

At the present the large cap is being lowered to the second leak, and we may not know if this works until Monday. It has one advantage over the first step in the process. When the initial seal was put over the end of the riser, it closed the open end of the pipe. However the riser was split in two additional places. It is the intermediate leak that is the current target. When the pipe was capped the pressure driving the oil up out of the ground did not change. As a result, since the other two leaks are still open, all the oil that was going out of the three holes is now flowing out of two.

However when the intermediate cap is placed over the riser it will capture some of the oil, and depending on how the pressure in that cap is controlled, by drawing the oil to the surface, it might be possible not only to capture that portion of the spill, but also to reduce the pressure at that point a little, relative to the third split, so that the flow from it is also reduced a little.

It has been suggested that this idea is relatively novel, and the patent on the idea is relatively recent about 2000) though it appears to have been quite successful at shallower depths. The technique is a variation on the Riserless Mud Recovery System (RMR) that has been successful in a number of earlier wells, though at shallower depths. We will wait to see how it works this time, though I can't at the moment, see why it should not.