A recent industry-backed study of diluted bitumen, the Canadian crude oil that would be shipped through the proposed Keystone XL pipeline, contradicts what environmentalists have said for years—that diluted bitumen, or dilbit, sinks in water and is much more difficult to clean up than conventional crude oil.
Instead, the study found that dilbit floats when it spills into water, a claim that contradicts what happened during a major dilbit spill in Michigan's Kalamazoo River in 2010. The cleanup of that spill already has cost more than $810 million, and the Environmental Protection Agency is still struggling to figure out how to remove the submerged oil.
The study is important because there is little scientific research on how dilbit reacts in water, and because the Keystone would cross thousands of U.S. waterways, including the critically important Ogallala aquifer in Nebraska.
But five scientists interviewed by InsideClimate News say the study was so narrowly constructed that it shouldn't be used to draw conclusions about dilbit. The experiment's laboratory conditions didn't reflect most real-life situations, they said, and the study ignored the consequences of the Michigan spill.
"One single study doesn't answer very much at all," said Merv Fingas, a retired scientist from Environment Canada, the Canadian equivalent of the EPA, who studied dilbit in the 90's. "You really need a much more detailed study ... to explain the science behind all of this."
Fingas said Environment Canada tested different kinds of dilbit in the 1990s and found that they "all sank in freshwater." But the research was abandoned before it could be broadened, he said.
Dilbit is produced by diluting bitumen—a thick, heavy crude mined from Canada's tar sands region—with a cocktail of light liquid chemicals that allows the bitumen to flow through pipelines. About two weeks after the Michigan spill, the EPA discovered that the liquid chemicals were evaporating, leaving the bitumen to sink slowly to the river bottom.
Ian Buist, an SL Ross engineer who was part of the study team, said they didn't take into account any of the findings from the Kalamazoo spill "because we weren't asked to. It wasn't part of the contract."
Buist declined to discuss how the experimental setup compared to a real life spill environment. He referred questions about the scope of the study to Enbridge.
In an email, Enbridge spokesman Todd Nogier said the study was done "to evaluate whether diluted bitumen would sink due to weathering processes alone. We attempted to simulate the conditions that exist in the environment that contribute to weathering of oil in a [water] setting such as UV lights, surface mixing and wind."
The SL Ross researchers poured 1.3 gallons of dilbit into a metal tank holding about 900 gallons of freshwater. They used a fan and UV light to simulate wind and sun, and circulated water from the bottom of the tank to create currents. After 11 days, they said no oil had sunk to the bottom of the tank. Most of the oil remained on the surface, with less than 5 percent suspended in the water.
No Sediment in the Tank
Several of the scientists InsideClimate News interviewed—none of whom worked on the study—said SL Ross has a reputation for conducting good research. But they also said there's a lot of room for improvement.
"What the experiment did is what I would say a good start, or maybe a place to consider the outcome and how you might be able to make it more sophisticated," said Chris Reddy, a geochemist at Woods Hole Oceanographic Institution who studies the fate and transport of crude oil in water. "I don't want to discount laboratory experiments. But you have to recognize that Mother Nature is a much more complex [environment] than what is found in a laboratory."
It's difficult to capture the variations in temperature, currents and organic matter found in a real river, he said. "To me, if you want a really good [experiment], you go to a site that's been impacted," like the Kalamazoo.
The scientists said the study didn't consider a crucial element—suspended material such as sediment and organic matter—found in virtually every body of water. Sand, clay and floating plant matter can drag down floating oil. Even conventional crude oil, which is almost always lighter than water, will sink if enough suspended material sticks to it. Heavier oils, like bitumen, can sink even faster.
"Once it starts to pick up the sediment, it begins to get heavier and heavier," said Rich Gaudiosi, president of the Delaware Bay and River Cooperation, a nonprofit coalition of oil response companies. The oil "may not sink all the way to the sea bottom or river bottom, but it will hang below the surface, which makes it obviously harder to clean up because you don't know where it is."
Gaudiosi, a retired Coast Guard officer, has never worked directly with dilbit but has experience with other heavy oils. In 2004, his group helped clean up the Delaware River after a tanker spilled 265,000 gallons of heavy Venezuelan crude. He said they had to improvise methods of tracking the submerged oil, just as the EPA is still doing on the Kalamazoo.
"There aren't tried and true techniques on the recovery of heavy oils once the lighter ends go away," Gaudiosi said.
The Delaware cleanup was complicated by the river's many currents and eddies—another variable that wasn't factored into the SL Ross study.
"Trying to reproduce a spill on a small scale, where the surfaces are flat, and you don't have the mixing crannies of a riverbed, which is like an English muffin—trying to capture that is not easy," said Reddy, the Woods Hole geochemist. "I would be incredibly prudent about transferring knowledge that was gained from a laboratory experiment to a broader interpretation."
In a real spill, the oil's behavior would depend on the type and amount of suspended material in the water, its grain size and how fast the water is moving. "All these factors are at play, which means you have to think about this on a case by case basis," he said.
Only One Type of Dilbit Studied
Fingas, the retired Environment Canada scientist, said another limitation of the SL Ross study was that it examined only one of the many varieties of dilbit on the market. Each one will react differently during a spill depending on the type of bitumen, the proprietary chemicals used to dilute the bitumen, the density of the water and other factors in the surrounding environment.
Using this study to make decisions about dilbit management would be like analyzing a random scoop of soil and concluding that it's identical to all the soils on earth, he said. "What you need to do is study the whole breadth of products."
The Cold Lake dilbit blend—the type of oil found in the Kalamazoo and also studied by SL Ross—contains one of the lighter bitumens mined in Canada. Bitumen from the Athabasca tar sands deposits, for example, is often heavier than Cold Lake bitumen.
According to its manufacturer, Cold Lake dilbit has a density of 0.9 to 1.2. Since freshwater has a density of 1.0, it appears that the SL Ross experiment was conducted with one of the lighter Cold Lake blends. If the scientists had chosen a heavier sample of Cold Lake dilbit, the oil would have sunk, regardless of the currents or sediment.
Enbridge Says It Can Deal with Sunken Oil
Enbridge made the SL Ross study public last month at a hearing in British Columbia about its proposed Northern Gateway pipeline, which would carry dilbit and other tar sands oils 731 miles from Alberta to the British Columbia coast. Once the oil reaches the coast, it would be loaded onto tankers bound for international ports.
The project is being fought by environmental groups and Canada's First Nations who fear a spill would irreparably damage the environment. The opposition is so strong that some industry analysts predict the pipeline will never be built.
Enbridge filed the SL Ross report with Canadian regulators as evidence to boost its claim that dilbit rarely sinks, but it wasn't discussed in detail during the hearing, which focused on oil cleanup in marine environments. Oil tends to float more easily in the ocean because saltwater is denser than freshwater.
Edward Owens, a member of Enbridge's emergency response team, assured the regulators that Enbridge has "the capability" to clean up submerged oil. "It's difficult, admittedly, but we know how to look for this type of oil that's sunken and we know how to deal with it."
He did not mention the 2010 spill in the Kalamazoo River, where submerged oil continues to accumulate.
The scientists interviewed by InsideClimate News say cleaning up sunken oil remains a challenge.
"In Canada, emergency responders are not as prepared for submerged oil as they should be," said Fingas, the former Environment Canada scientist.
The situation is similar in the U.S.
"All of our response capabilities assume oil floats," said Jacqueline Michel, president of Research Planning, Inc., a South Carolina-based consulting firm. She's been helping the federal government deal with oil spills since 1978. "We know very little about dilbit's long-term behavior."
Neither Michel nor the other scientists could name another publicly available study from the last decade that analyzed how dilbit behaves in water.
"Maybe someone's done studies, but I haven't seem them," Michel said. "I know a lot of my cohorts who've done work for the industry have probably done all kinds of work which I may not be aware of."
The lack of research is important because most U.S. pipeline regulations were written before dilbit began flowing into the country in 1999. Congress commissioned a study last year to determine whether dilbit is more corrosive to pipelines than conventional oil, but it won't cover what happens when dilbit spills into water.