Anatomy of an Oil Plume May 24, 2010Posted by Jamie Friedland in Offshore Drilling, Politics.
Tags: BP, Deepwater Horizon, Methane Hydrates, Offshore Drilling, Oil, Oil Plumes, Oil Spill, Roger Faulkner, Subsurface Oil
Last week’s discovery of massive underwater plumes of oil, drifting beneath the surface, was widely reported. But how does that happen – doesn’t oil rise in water? Turns out it’s not quite that simple. I found a few sources to help explain this seemingly improbable phenomenon and theorize what is happening now.
Big bodies of water are not homogenous, they are layered, or “stratified.” Even in the same location, there may be great differences in water temperature and salinity depending on the depth. Those factors, along with the temperature, phase, and density of the gas and oil itself, have an impact on how water and hydrocarbons interact.
The Phases of an Underwater Oil Spill
There are three distinct stages in oil’s rise to the surface from a pressurized, underwater spill:
1) Jet Phase: In the initial phase of a high-pressure underwater leak, oil and gas are forcefully expelled from their pressurized confinement in a narrow, expanding cone. Within about a meter of the breach, the oil forms into tiny droplets, barely millimeters across, and natural gas forms into tiny bubbles. The higher the exit velocity, the smaller the droplets and bubbles are. This droplet size impacts buoyancy – larger droplets rise faster and more easily than smaller ones (discussed below).
2) Plume Phase: The momentum from all those little droplets shooting upward from the jet together entrains large amounts of nearby seawater, dragging it upwards as well. The rate at which plumes rise can vary; if large amounts of gas are leaking, the plume rises more quickly. A plume can also stop rising altogether – plumes can become suspended in the water column at “terminal layers” (explained below).
3) Post-terminal Phase: Above the final terminal layer, hydrocarbons rise to all the way to the surface solely under the power of each droplet’s own buoyancy.
Hydrate Formation: You may have noticed that Gas Hydrates appear on the chart but not in the plume. At depths of at least 300 meters, methane can form hydrates, a crystalline slurry of natural gas and water that is similar to ice. We know hydrates are forming at this site because they are what caused the containment dome to fail.
Active hydrate formation could help explain why so much oil is still suspended in the water column: natural gas is responsible for much of a plume’s buoyancy. When hydrates form, they settle out of the plume and sink, depriving the plume of much of its upward momentum.
Terminal Layers: At depths of over 200 meters, the ambient seawater is relatively dense. In a deepwater blowout, enough of this dense water eventually gets entrained in the plume that the suspension loses its relative buoyancy and begins to “mushroom” at a given depth, called a “terminal layer”. If heavier components sink out of the suspension, the plume may reform and begin to rise again past that terminal layer. This process, known as “peeling,” can happen numerous times en route to the surface.
Under certain circumstances, rising oil can “mushroom” and end up suspended at depth for an extended period instead of just rising.
As much as I hate to credit UNC for anything, the video above is an excellent demonstration/visualization of how a high-pressure jet can result in mushrooming. The first :50 shows how fine droplets sprayed from a jet can get trapped at depth instead of rising to the surface. The final :40 shows how bigger droplets rise straight to the top, just as you’d expect.
Dispersants Increase Oil Suspension
The high pressure of this spill is creating smaller droplets that can get suspended at depth instead of rising right to the surface. On top of that, though, the dispersants that BP is injecting into the well work by decreasing oil droplet size. That’s causing even more suspension.
BP’s flow estimate is so much lower than everyone else’s because they refuse to measure the flow at the seabed. Instead, BP is calculating the flow rate solely by the size of the surface slick. This is particularly odious, because it overtly and intentionally omits all the oil “successfully” dispersed by the hundreds of thousands of gallons of toxic chemicals BP is also injecting into the Gulf.
Rising Plumes can be Delayed and Disrupted
There are circumstances in which the plume dynamic can be disrupted. As pictured above, currents can bend or even transport the plume long distances, and especially strong current action can prevent the formation of a plume altogether. Emulsification poses a more pertinent scenario.
Oil and water normally don’t mix. However, under the right circumstances, they can emulsify into a stable mixture. Emulsified oil is much more dense and less buoyant than oil on its own, so it rises through the water more slowly. When oil is broken up into tiny droplets, this emulsification is far more likely to occur.
Oil Can Sink in Water
Like any other substance, oil will sink in water if it becomes more dense than the water around it. One way this can happen is if oil incorporates enough suspended sediment from the water column. This happens through adsorption, a process by which a liquid or gas forms a film around the surface of some other substance. So in this case, I assume that oil forms a coat around tiny clumps of floating sediment, and the whole new oily clump sinks.
So What’s Happening with Deepwater Horizon?
A polymer scientist and technology entrepreneur named Roger Faulkner has a very technical explanation of what is happening underwater.
His theory is as follows: in this ultra deep well with high methane content, conditions (~15,000 psi) are such that the oil and gas can actually mix in a “supercritical solution” until they exit the pipe. They cannot separate until the pressure is reduced. According to this hypothesis, there should be four subsurface separations (3 of which will rise as plumes):
1) As the oil begins to rise, the first separation should occur between the heaviest oil elements and the rest of the mixture. It should begin somewhere between the reservoir (18,000 ft beneath the sea floor) and the sea floor, but the pressure in the pipe is probably high enough to sweep the separated elements up and out of the pipe. This first separation should be complete within 40 feet of the sea floor. Faulkner did not explicitly say this, but I think this plume will be the slowest in rising of the three.
2) There is a very large pressure drop passing through the partially sealed BOP (~9000 psi to ~2250 psi). Within the first second after exiting the pipe, the natural gas should begin to separate. At first, it will still be hot enough to mix well with lighter oil components. As it cools it will decompose in two ways. Some of it will stay mixed with some heavier oil liquids saturated with methane. The rest will end up in the third plume.
3) The remaining mixture will become a dense gas solution with most of the gasoline and light oil as well as some heavy oil. This is where most of the methane will be, so it will be the lightest and fastest rising of the plumes.
4) As natural gas separates from the light oil in Plume 3 to cool and expand, it will form methane hydrate. These hydrates are slightly denser than water at this depth, and will slowly sink to the floor.
Faulkner describes this as “sort of a doomsday scenario” that can only happen because of the unique, gradual and stepwise pressure reduction as the oil first leaves the reservoir into the pipe, then goes through the BOP, then goes into the downed riser pipe, then enters the pressure gradient that is the deep ocean.
According to these predictions, MOST OF THE OIL IS STILL UNDERWATER BUT WILL SURFACE EVENTUALLY. Of the technical pieces of writing I’ve read before, none other has ever included the phrase “God help us.”
Tests Predicted Underwater Plumes
It is also worth mentioning that the possibility of underwater plumes was not unexpected. It has long been theorized that a blowout at deepwater depths would produce some unusual fluid dynamics behavior. Indeed, tests have been conducted to confirm and model the phenomenon.
In 2001, the now-embattled MMS conducted an experiment with 23 oil companies to try to see what would happen in the event of a deepwater blowout (note: the mere existence of this test provides FURTHER proof that such an occurrence was neither “inconceivable” nor “unprecedented”).
To test this, different mixtures of natural gas and oil were released half a mile underwater off the coast of Norway to simulate a blowout. Additional, related research was conducted in a pressure chamber at the University of Hawaii.
These projects confirmed that plumes of oil that do not immediately rise to the surface can in fact form in the event of a deepwater blowout.
The Deepwater Horizon spill has flowed for over a month now. And we may only just be getting started.