It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Tuesday, May 25, 2010

Horizon - Backing Off Optimism

OK, so first of all, somebody I've heard of (a physical oceanographer whose opinion on such matters I'd trust more than my own) is worrying about long-range transport of oil:
Niiler is not speculating. He has studied the way ocean currents and winds moved hundreds of "drifter" buoys around the Gulf. In the drifters' 90-day lifespan, he has seen them scatter to all parts of the Gulf with the help of a tropical storm with 40-knot winds.

Some drifters were found as far west as Texas and others were caught in the Loop Current that carries Gulf water out and around to the East Coast of North America.

What's more, said Niiler, there's no evidence that oil will be diluted by the time it reaches the East Coast.

"We see Mississippi water in the Loop Current all the way to Cape Cod," said Niiler. "It's not mixed up."

And neither will any oil slicks that are sucked into the Loop, he said, unless there is something causing the water to mix, like a hurricane. A powerful hurricane can cause the ocean waters to mix down to 150 meters.
Is this right? I'm not sure. The Mississippi delivers more than 100 K barrels/day. So the fact that Mississippi water is detectable offshore in the east doesn't settle the question whether the oil will be noticeable, never mind troublesome.

Actually, this brings me to one of the things I've been thinking about. We really need to think about four sorts of oil in this situation:
  1. oil in colloidal suspension in the deep
  2. oil in two dimensional configuration on the surface
  3. oil in two dimensional configuration on the sea floor
  4. oil in linear configuration on a beach
Each of these will have different impacts. So far, it seems like we are getting a lot of #1, and I still think this is a lucky break. Suppose for the sake of argument we have 100,000 barrels divided into four equal parts of 25,000 barrels each. So each portion is about a million gallons (four million liters). Now suppose a gallon of oil is floating in a thin film, about a centimeter thick, it will cover about a third of a square meter. The whole mess will cover 33000 m^2 or an area of about 181 meters on a side. Of course, it doesn't do us the kindness of staying in a neat square. So suppose we consider a coverage of about 1/1000; that will probably be enough to provide a sheen on everything. The mess will cover a square about 5.7 km on a side and will be very depressing.

Now let's distribute that same amount of oil over the water column. This will be those "plumes" we are talking about. Presume the water column is 2000 meters thick. Here we have a gallon of oil over 1000 m^2 * 2000 m = 2,000,000 m^3. What is the density of oil in the column? Well a gallon is 4 liters is .004 m^3, so the concentration of oil in the water column is .004/2e6 = 2 parts per billion. Ho hum.

Of course, that is a bit generous. Suppose the plume is constrained to a 2 meter thickness instead of a 2 km thickness. 2 parts per million. Still hard to lose sleep over. Here, the dimensionality of the problem is on our side. So if these plumes get into the loop current, they will get sheared out and nobody will ever hear of them again.

The question is whether the colloidal suspension is stable on time scales comparable with the surface weathering of oil. I guess some of it is buoyant and gets to the surface. I get the impression that it is.

On the other hand, there is the sea floor. If some of the oil gets there, or some of the colloid settles down to the bottom, the lifetime is probably very long. And if a significant amount of colloid gets dragged past the coral formations off the Florida Keys, surely that won't help the corals which are already stressed by lots of things. Both of those things convert from 3 D back to 2 D.

But the dimensionality of the problem cuts the other way on the shoreline. If a gallon is essentially trivial in a 3D column, and unpleasant on a 2 D surface, it is catastrophic on a shoreline. There, 25,000 barrels of oil can make a huge mess. The entire Santa Barbara spill amounted to 100,000 gallons. If we are approaching such quantities on a daily basis, the amount of shoreline damage can be spectacular.

This is why it has been such good news that the oil has been kept offshore until the last couple of days, and such bad news that it is beginning to accumulate.

So I'm not as optimistic as I was a few days ago. But I'm sticking to my guns on this part: I think the main issue is the shoreline. The bottom is a secondary issue. Stuff that remains in suspension is not a big worry, and so the exact measure of the flux through the hole is not an immediate concern. Some aquatic birds and animals will get oiled at the surface, but not many.

We will be able to measure things better when it all settles down. Certainly it's important to do that, to gain experience if this ever happens again.

On the other hand, most of what I've seen indicates that this was entirely avoidable. Like Chernobyl, it tells us how bad things can get if people really screw up badly. That's pretty bad, so the best thing to do is to avoid screwing up. So I still think the measurement issue is a red herring.

This isn't to defend or attack BP's post-spill actions. I really don't have the information or skills to judge, much as I enjoyed the forthrightness of the f***ing booming rant.

I think it's important to understand that the engineers are not pleased. I think it's important to understand that an entity the size of BP is a lot of people and a lot of decision-makers. Inevitably with an organization of that size some of its members take their responsibilities more seriously than others do. The culpability and liability of the organization itself I leave to the lawyers and others who like to cast blame at inanimate objects.

Image: NASA via SkyTruth h/t Hank. Note that the darker area in the east is not oil, but rather an area of calm water where the waves are too small to reflect sunlight directly to the satellite.


Hank Roberts said...

Here's the situation report page for today
the inset shows the loop current; the larger map shows a long streamer off to the southeast.

That situation report map seems to match up with the main part of the spill labeled on this image.

The long stream to the southeast does seem to align with the 'loop current' on the situation report, but coincidence is surely likely; I don't know if the loop current is mostly on the surface or if surface winds dominate, and they had offshore wind for quite a while that could explain this.

I've got family in Texas and Florida, so I'm watching with personal concern here.

Of course I have no idea what 'ground truth' is being used to support the satellite photo interpretation. If your expert says one of the labeled areas is definitely not oil, that's good news (and will be a relief to the folks along the Florida gulf coast).

PS, I don't believe today's '' "expert" reports about evacuation plans for Tampa Bay; I think we're starting to see people making money off mouseclick stories.

Hank Roberts said...

This image extends much farther than the Situation Report map:

They do label patch to the far southeast as oil, as Skytruth does; to me, the patch to the east-northeast looks like the same color, but isn't labeled here.

The Situation Report map just says they use remote sensing to identify what's there -- someone should have a database from boat reports using GPS, I suppose, but that's a level of detail we aren't getting.

Hank Roberts said...

Oh, and NOAA doesn't show any oil over to the east at all:,subtopic_id,topic_id&entry_id%28entry_subtopic_topic%29=809&subtopic_id%28entry_subtopic_topic%29=2&topic_id%28entry_subtopic_topic%29=1

PS, more imagery and radar images at

They also say down the thread appropriately:
"... MODIS / Terra satellite image, May 22, 2010.

Clouds and haze obscure the southeastern Gulf, but a small patch of what might be oil entrained in the Loop Current is visible. As we've said before, it is possible the Loop Current has a distinct color even without the presence of oil, so this is a low-confidence analysis and therefore is shown with a dashed orange line. Sure wish they'd send a vessel out there to do some sampling transects."

Michael Tobis said...

The darker area in the Northeast is not oil, and I learned that from your link.

That narrow feature spreading southeast obviously has something to do with the blowout. It's peculiar, really; I'd like it understand it from a fluid dynamics point of view. You would expect it to get pinched off at some point.

I doubt it matters all that much on the current scale of things though. I'd call it a curiosity.

Hank Roberts said...

Offshore trajectory maps/models,subtopic_id,topic_id&entry_id%28entry_subtopic_topic%29=809&subtopic_id%28entry_subtopic_topic%29=2&topic_id%28entry_subtopic_topic%29=1#downloads

This one shows that long stream off to the southeast of the well:

apology for oversharing, I'll quit for a while and hope for better info to surface. More please from your expert on mixing and the Gulf Stream, I know parcels of water travel in it with amazing integrity over long distances (I've swum in it during 'deepsea' day fishing trips as a kid off the North Carolina coast -- we could dive off the boat as it stopped in the Gulf Stream water, swim as deep as our breath would let us dive in clear, warm water, with the sun casting our and the boat's shadow down into the deep deep blue water below -- and yet not far away to the side was the gray-green cold cloudy Atlantic water we'd traveled through to get there.

Then they called all us kids back onto the boat, and spent the rest of the day doing nothing but fishing. Bother!

Hank Roberts said...

Here's more detail showing that feature as oil being pulled into the loop current; don't know where they get their detail:

"strategic fisheries forecasts that are the result of the integration of satellite and other fisheries oceanographic data....."

Hank Roberts said...

The ground truth is in the PDF files, very specific attribution by name and source for the detail they're using in their maps.


"... Today’s RGB data shows that the surface oil (olive green) has reached the counter-clockwise eddy (centered roughly near 85°45’W & 27°30’N) west of Tampa, Florida and the tip of the oil was observed near 85°50’W & 27°07’N). We have received visual confirmation of this oil from on-site sampling by Dr. Jim Franks Southern Mississippi University who described a sheen and visible oil globules, 1/8 inch to 5 inches in diameter at 86°57’W & 27°45’N and 86°54’W & 27°46’N. Based on the motion of the water west of Tampa it appears that this surface oil is anticipated to travel the same path as the subsurface oil-water-dispersant mixture that we have depicted in a gray color....
... If you plan to use these reports including the graphics you must give ROFFSTM full credit for this work. ROFFSTM would be appreciative if you would copy this analysis to others who may be interested in our efforts...."

Michael Tobis said...

My mistake. The loop current is farther north than I had thought. That filament actually is marking the boundary of the current and looks like it will be carried along in the loop current. Let's hope the main blob doesn't wander over that way.

Hank Roberts said...

The long feature's been visible since May 7th, apparently. (same effect changing from lighter to darker as the angle of reflection changes)

I wish someone had made a 'flipbook' of all these images stretched and aligned to cover the same area.

Gallery here:
The oil's more obvious in the one nighttime infrared photo but that's a zoom only on the well area.

John Mashey said...

Part 1
1) people need to understand the workings of big companies. If the incentives visible to individuals encourage cutting corners and taking what look like small risks for short-term profits ... at the chance of huge, but rare problem that may well surface after they've moved on, and that if big enough someone else will pay for...
then that's what you get.

We've just seen how that works in the financial markets (and read Michael Lewis (of Liar's Poker fame) in his recent book, "The Big Short", about a handful of people who *bet* that the crazed subprime mortgage business would have a total meltdown, and it did... but a lot of people who caused the mess also walked away with plenty of $.

Wile he sometimes gets a bit over the top, Nicolas Nassim Taleb's "The Black Swan" has good discussion of the problem:
make a bunch of money over a while, taking more risk than you realize ... then whammo.

2) SO, given that we are simply not going to stop having deepsea wells in the Gulf, how do you fix this problem, i.e., how do you, in practice, strongly reduce the likelihood of something happening, and how do you assure, that if something does happen, that contingency plans and resources are available to deal with the problem?

My answers later in Part 2.
But I can say that I think either minimizing or maximizing the wording doesn't help much. "nationalizing the oil industry" isn't going to happen, and wouldn't help if it did.

Real solutions might look towards nuclear submarines, the old Bell System, computer design, sales&executive compensation, and insurance companies, of which at least 4 actually depend on choosing the right metrics (whatever they should be) and measuring them, to create the right incentives for the right people.

Hank Roberts said...

Well, shit. I've had the live feed in a corner window most of the day and it'd been consistent for a long time -- black oil pouring out, occasionally faster for a while.

Right around 17:56 it suddenly increased, changed to gray-and-black with black chunks, and grew and grew; for a while up til now at 18:17 (this is the video time stamp) it's been just a churning cloud. The camera panned up and down once, another light source came briefly into view, the lights are still illuminating the mess, but it's an enormous mixture of oil and black chunks of stuff billowing around; now at 18:18 back to black shooting out but still way way more than before.


Jim said...

I don't see any reason to think that this isn't a catastrophe on the shore and a catastrophe on the bottom:

Philippe Cousteau Jr. dives beneath Gulf’s oily water: ‘This is a nightmare … a nightmare’

Michael Tobis said...

The video Jim points to is, in my opinion, manipulative. It offers no sense of scale or context. It just goes "yuk".

OK, yuk then. Fair enough. I don't want to swim there either.

Hank Roberts said...

One top kill diagram here

Another at the NYT.

Some of the commenters in the blog thread along with the video feed are at least competent to guess what we're seeing.

BP really needs someone to do what NASA does, give information along with the pictures or type in a caption string explaining what we're seeing. The ROV operators would be the people I'd like to eavesdrop on. The ocean a mile down is far less accessible than near earth orbit for human beings.

Aaron said...

I see different sorts of oil. A lot of methane is coming out of the pipe and that will end up in the atmosphere. Some long compounds are dissolved in the crude and they will fall out to form low toxicity tar balls on the sea bed. In between are a lot different compounds of varying vapor pressures, toxicity, and solubility including some metals and sulfates. This stuff plus Corexit is more toxic than the crude alone, and remains toxic longer then crude alone. We do not have good numbers for this, we do acute tests for 96 hour time periods. This stuff is going to be out there and the critters exposed to it for longer than our 96 hours test periods. In some cases, it will get passed up the food chain.

Ultimately the effect of the oil is on life, including little critters that are the base of food webs and little critters that grow up to be big critters. For a 2cm long critter on the beach, then that “1-dimensional” line of oil modeled by MT is its entire world, and it’s death. It may even be death if the critter is several feet away from the line drawn by MT. That is, MT’s line is tar balls on the beach, and soluble toxic materials leached from the oil kill the little critter.

I would model the oil as clouds or plumes of oil at various concentrations. At 1 ppb (of middles) most critters will live. At 1 ppm many little critters will live more than 96 hours. At 10 ppm, many little critters will live less than 96 hours. That is how I would sort the oil.

One little tar ball can keep leaching toxics for a long time. Thus, a few tar balls on the sea bed can affect the biological productivity of the area for an extended period.

We think we have “saved” a bird when we have cleaned the oil off of it. However, we do not clean the oil off the little critters that the bird eats. Then, the bird dies later of starvation. That is "natural", and we feel good because the bird did not die because of the oil spill. However, it was the oil spill that killed its food.

Greg said...

"The culpability and liability of the organization itself I leave to the lawyers and others who like to cast blame at inanimate objects."

Hmm, I seem to recall that the US Supreme court has given a lot of equivalent-or-better-to-people rights to those inanimate objects. A recent decision about political spending ....

Paul said...

I was pleased to see John Mashey's recognition that some lessons might be learned from the nuclear submarine program. Indeed they could. ADM Rickover met extensively with executives of the civilian nuclear power industry following Three Mile Island. They listened politely and ignored most of what he said. One of his central points was that the decision to start up a reactor or continue its operation in face any safety question could not rest in the hands of those whose primary responsibility was to the share holders.

Adm. Rickover pioneered not only the Navy nuclear propulsion program but also the civilian power industry with the construction of the Shippingport plant in 1956.

He was enthusiastic about the thorium cycle breeder. Over the years his concern for the safety of civilian plants increased as the impact of reactor saftey studies and potential harm from core meltdown level accidents became better understood. This concern was not allayed by the knowlege that a large portion of civilian plant operators and managers had been trained in his program and served on nuclear powered ships.

He recognized from the beginning that safe operation depended on plant design, maintenance, continuous testing, highly selected and trained operators and rigorous operational inspections.

Most of all it depended on the integrity of the operators to follow the rules, believe their indications and either not start up or continue operation in the face of any questionable safety condition.

On a nuclear powered vessel this responsibility rests with the captain. Enormous importance is placed on operational readiness and meeting underway commitments on time. Great is the temptation to ignore a safety issue and start up anyway to meet an underway commitment. It didn't take too many reliefs for cause of CO's and chief engineers who let operational pressures overcome their reactor safety responsibilities to get everyone's attention.

Navy nuclear propulsion has a 50 plus year history of unprecedented safe operation as a result of Rickover's basic approach. It has not been diluted with subsequent generations. We've done some dumb things operationally over the years but compromising reactor safety has not been among them.

Rickover lost faith in the civilian industry to make the costly long term decisions that would put similar policies in place for their operation.

The BP oil disaster resulted from the violation of essentially every principle Rickover espoused.

Paul Middents
Captain USN Retired
Operated and maintained nuclear submarines 1964 -1991

Hank Roberts said...

good collection of videos and some facts here:
including the big change I noticed; also including a 10-minute video as one of the ROVs is pulled up to the surface to clean its camera. As it rises, particularly around 7:00 on, you start to see the billows of oil at different levels below the surface, giving a better idea of what's down there.

I also saw an AGU piece mentioned warning that a Gulf hurricane is likely to cause currents even that deep strong enough to move the oil equipment on the floor and stir up whatever's there, as has happened in the past.

Michael Tobis said...

Paul, that's far too good of a letter to be wasted on the comments section of a minor league blog (even a very good one such as this :D ).

I'm very honored to be the recipient, but I can't do it justice.

I suggest you send it to the NY Times or such.