Analysis of oil/gas behavior pre and post Deepwater Horizon blowout BOP
Posted: May 24th, 2010 by: h2
Here’s another little gem of a comment posting, again at theoildrum.com, Deep Water Horizon Oil Spill Multiple Plumes. This is for those of you who are wondering how all this oil/water/pressure stuff actually works. If you’re not interested in the technical parts of the Deepwater Horizon spill, just move along, this is all meat.
roger_rethinker on May 24, 2010 – 9:56pm Permalink | Subthread | Comments top
Deep Water Horizon Oil Spill Multiple Plumes
By Roger Faulkner
Re-posted with edits May 24, 2010
(originally posted http://www.theoildrum.com/node/6499#comment-628572)I have consulted with several experts, and I have modified this blog post somewhat from previous posts, but the essential ideas are intact. The Deepwater Horizon oil spill is different from all previous blow-outs because of four separate unusual or unique aspects of this particular blowout:
1. The gas: oil ratio (GOR) in this well is reported to be about 3000, which means about 150 pounds of gas per 285 pounds of oil (34% gas by weight, more than 70% by mole ratio methane + ethane). This well is between a typical gas well and a typical oil well. The high amount of gas at the high pressure of the reservoir means that the properties of the reservoir must be understood as a supercritical solution which I here term petrogas. It is possible that there is no fluid phase boundary within the reservoir, but the expert I spoke to (Dr. Robert M. Enick, Bayer Professor of Chemical and Petroleum Engineering, Swanson School of Engineering, University of Pittsburgh) thinks that is unlikely. On the other hand if two phases do coexist within the reservoir, it is very likely that more than 50% of the weight of the petroleum is in the supercritical phase, since at 12,000 psi Methane is a very strong solvent. We bet a beer on this; I still think the petrogas is a single supercritical phase in the reservoir. We both agree that by the time the petrogas rises to the wellhead, it is probably a two-phase flow.
a. According to information given to the team that is tasked to estimate the flow, the pressure in the reservoir is about 12,000 psi, but only 180 Fahrenheit, which surprised both of us (TOD bloggers: is this credible?). If this pressure is correct, and the 8500 psi estimated pressure behind the BOP is correct, then the average density of the petrogas in the drill pipe is 0.62 g/cc, which is reasonable for a supercritical solution of gas + oil.
b. The supercritical nature of at least part of the petrogas persists all the way up the drill pipe to the seabed (13,000 feet). The 13,000 foot rise of the supercritical petrogas is expected to be a nearly adiabatic expansion against gravity. In essence, the work to lift the petrogas 13,000 feet is performed by pressure-volume work done as the petrogas expands and cools coming up the drill pipe.c. The expansion and cooling of any supercritical solution reduces the solvent power. It is possible that the least soluble components of the crude oil (highest molecular weight and/or most polar components) will precipitate out of solution during the 13,000 foot nearly adiabatic rise of petrogas from the reservoir to the BOP. This implies that the material entering the BOP is likely to be phase separated into a supercritical methane-based solution, and liquid droplets containing the least soluble components. (Dr. Enick was skeptical about whether this oil contains much asphaltene, but he has not yet seen samples. The anomalously low temperature of this reservoir makes it more likely that it does contain some asphaltenes.) I continue to think that most of the oil is still contained in the supercritical methane-based phase (at 8000-9000 psi) when it enters the blow out preventer (BOP). The phase makeup just before the BOP can be determined experimentally by recreating these conditions in a lab. I’ve been talking to various scientists & engineers who have the right kind of equipment to do these experiments. (These experiments are pretty vital; any influence that TOD bloggers can bring to bear on getting these measurements funded will be helpful)
d. Dr. Enick points out that some of the observed tar balls may be from partially burned oil.
2. Most of the pressure drop going from the reservoir to the environment occurs very fast, probably in milliseconds as the supercritical methane-based petrogas, and possibly a viscous liquid phase as well, passes through a severe flow restriction at the BOP which is partially closed. In the Horizon Spill, the pressure of the oil goes from 8000-9000 psi before the orifice to 2650 psi right after the BOP according to Admiral Thad Allen of the Coast Guard on May 15 (http://blog.al.com/live/2010/05/national_incident_commander_oi_1.html).
a. The sudden reduction of pressure at the BOP must produce a phase separation just downstream of the orifice, with most of the heavier molecules condensing out to form one or more liquid phases, and most of the methane and a portion of lighter fractions staying in the gas phase. Given that the temperature is apparently much lower that I had estimated earlier, not much oil is likely to remain dissolved in the gas downstream of the BOP. As the pressure goes from well into the supercritical region (~8500 psi versus methane’s critical pressure of 6600 psi) to subcritical conditions, it is possible that several precipitations occur. It seems likely that most of the initially formed liquid phase droplets are quite small and mutually miscible and will form a single liquid phase given enough time.
b. The expansion through the BOP is sort of a Joule-Thompson expansion, but because it results in formation of a liquid phase, it is expected to produce a temperature increase due to the heat of vaporization that is released. It is quite possible that the temperature downstream of the BOP could be higher than the reservoir temperature because of the condensation of this oily phase.
3. The environmental pressure at the ocean bottom is around 2225 psi. Expansion to this pressure rather than to atmospheric pressure has an effect on the resultant phase separation. Although the gas phase formed downstream of the BOP orifice is subcritical, it is still fairly dense and has solvent properties. Right after the BOP, it is likely that a major portion of C6-C12 molecules will remain in the hot gas phase. The pressure right after the BOP is still about 400 psi above the local sea water pressure, and the flow is trapped inside a damaged and twisted riser pipe. There are two escape routes from the damaged riser pipe.
a. A small leak is just above the wellhead where the kinked riser pipe lays over onto the seabed. On April 15, it was estimated that 15% of the effluent from the blowout was exiting this hole (this fraction of the total flow has been increasing since then). The material blowing out of this hole has had very little time to cool. Insofar as there is very little time between the BOP orifice and the first leak into the ocean just above the riser, the phase structure and partitioning of components between the phases at the first leak from the riser pipe is expected to be very similar to the properties immediately downstream of the orifice. If a heavy oil phase separated from the petrogas on the 13,000 foot rise through the drill pipe, they are likely to survive as a third distinct phase at the first leak.
b. The second leak from the collapsed riser pipe is about a mile away from the BOP. The two phases formed at the BOP orifice will cool significantly during passage through the collapsed riser pipe, and they will remain in contact for a goodly while. I think it is very likely that if a heavy oil phase did separate from the supercritical methane-based petrogas while it rose the 13,000 feet to the BOP, they will re-dissolve into the hot liquid phase as the liquid phase moves along the mile long riser pipe.
4. The majority of the total methane entering the ocean will be in the high pressure gas phase, though some will be dissolved in the oil phase too. Unlike spills at low depth, the pressure at the Deep Horizon spill is well above the pressure required to form methane hydrate [46(H2O)•8(CH4)]. The spill itself may heat the water too much for methane hydrate to form near the leaks. As the plume carrying the methane mixes with more cold sea water, it will become cold enough for some methane hydrate crystals to form. Crystallization of the methane hydrate will release more heat. One aspect of the plume is that it contains warmed water; it is not as if petrogas bubbles need only rise within a vertically stationary water column; there is a plume of warm water that also is rising, at least for a while. I expect a lot of the methane to eventually precipitate out as methane hydrate “snow.” This snow will probably rise, and dissolve into the water rather than make it to the surface as bubbles.