About the SA Blog Network



Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

Dispersed oil proves less toxic in EPA tests

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint


The U.S. Environmental Protection Agency released results of its second round of toxicity testing on dispersants—chemicals used to break up the oil that spewed for nearly 90 days into the Gulf of Mexico from BP’s Macondo well. This new round focused on the specific oil in question—Louisiana sweet crude—alone and in conjunction with eight dispersants, including COREXIT, the chemical employed in bulk by BP in the Gulf.

"The eight have similar toxicities to one another when mixed with crude oil," EPA Assistant Administrator and chemist Paul Anastas said during an August 2 press briefing to announce the results. "The dispersant and oil mixtures were no more toxic to test species than the oil alone," except for Nokomis 3-AA, which BP has not reported using during the Gulf disaster.

Specifically, the oil alone is moderately toxic to juvenile silverside fish (Menidia beryllina) and mysid shrimp (Americamysis bahia), both natives of the Gulf of Mexico, according to tests conducted at a single laboratory under EPA contract. Similarly, the oil mixed with seven of the eight dispersants is also moderately toxic.

The EPA found that only COREXIT 9500A (the primary dispersant used in the Gulf disaster), Nokomis 3-4F and ZI-400 made the dispersed oil less toxic than oil alone to the shrimp, though all the dispersants registered as "moderately toxic." The least toxic dispersant, per the EPA tests, is ZI-400.The levels at which the eight tested dispersants killed half the tested animals ranged widely, from as little as 0.39 milligrams-per-liter for the shrimp to 13.1 mg/L for fish. Oil alone killed half the shrimp at 2.7 mg/L and roughly 7 percent of fish at the highest concentration tested: 2.9 mg/L. In general, the EPA data matched that provided by industry. "While there are differences in the data we are presenting today with the data submitted to [government by industry], there are not significant differences," Anastas said.

One mystery: COREXIT 9500A showed a "non-petroleum hydrocarbon chemical peak," meaning it was breaking down into unspecified hydrocarbon compounds on its own, though this was not included in the final results. It remains unclear what this result might mean for both toxicity testing and in terms of the dispersant’s long-term effect. The EPA also did not test COREXIT 9527, which BP reported using in relatively small quantities at the beginning of the spill.

Anastas called the decision to employ roughly 7 million liters of COREXIT a "wise choice," especially given that oil "degrades 50 percent faster when dispersant is used." He added: "It was interesting to see that the dispersant-oil mixture was about the same toxicity as the oil alone."

According to Anastas, roughly 100 times more oil is present in the Gulf than dispersant. And dispersant use ceased on July 19 when the well was capped.

Of course, the new data does not determine where the dispersed oil has ended up; the U.S. National Oceanic and Atmospheric Administration has detected large plumes of it throughout the water column and some scientists have found dispersed oil in crab larvae, among other animals. Nor do the tests mimic the actual conditions faced by the dispersant-oil mixture in the cold, dark, deep waters of the Gulf of Mexico. And it does not dispel fears that the dispersed oil may have long-term toxic effects as it breaks down, both in the water and in Gulf sealife. "In toxicology, it’s quite often not the original compound that’s the toxic entity," says toxicologist Carys Mitchelmore of the University of Maryland, who co-authored a 2005 National Research Council report on dispersants.

"A very important longer-term research question is to characterize the various metabolites and see if there are any metabolites of concern in the breakdown products," Anastas said in response to a question from Scientific American.

Image: U.S. Air Force Master Sgt Paul Tatar, a C-130 Hercules aircraft aerial spray aircraft maintainer, refills a C-130 with oil-dispersing chemicals on the runway of the Stennis International Airport, Miss, on May 9, 2010. U.S. Air Force photo by Tech. Sgt. Prentice Colter.

Rights & Permissions

Comments 4 Comments

Add Comment
  1. 1. kgriffin 8:52 pm 08/2/2010

    did you read the study? the quote from epa here was "The dispersant and oil mixtures were no more toxic to test species than the oil alone," Unless you are just looking at the corexit 9500 results, but then you have to assume either they never used 9526 or that its use was insignificant even though it was used before the EPA told BP to decrease dispersant use. And the 9500 results was for one species only.

    Link to this
  2. 2. gttaylor 9:50 pm 08/2/2010

    You missed the point! "the oil degrades 50% faster when the dispersant is used." So if there is no increase in toxicity and the oil goes away in half the time, the right answer is pretty obvious.

    Link to this
  3. 3. kgriffin 11:36 pm 08/2/2010

    The tests were "acute toxicity tests". Acute. that is toxicity in 24-96 hrs. How fast the oil degrades and how that degradation impacts long term effects was not looked at. You are making assumptions – this is about a research article. The likelihood of the oil reaching x or y has no relevance to acute toxicity. And again they did not say it was less toxic.

    Link to this
  4. 4. dbiello 10:04 am 08/3/2010

    If you look at the actual numbers in Table 3 of the EPA results, COREXIT 9500A dispersed oil is less toxic than Louisiana sweet crude alone. Specifically, LSC alone kills 50% of shrimp at 2.7 ppm and 7% of fish at 2.9 ppm while COREXIT-LSC mix kills 50% of shrimp at 5.4 ppm and 50% of fish at 7.6 ppm. That means the dispersant-oil mix is less toxic since, as EPA forgot in its directive to BP, the lower the concentration required to kill, the MORE toxic a given chemical is.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article