Worrisome levels of crude oil persist in a cove of Lake Conway eight months after ExxonMobil's Pegasus pipeline ruptured in Mayflower. But state authorities charged with monitoring the cleanup say the lingering oil is more of a threat to the area's ecology than it is to human health.
Those same state officials also say that they are confident the main body of Lake Conway was spared direct damage from the Pegasus pipeline disaster on March 29. To confirm that conclusion, further water sampling of the main body of the lake will continue.
Authorities reached those conclusions after reviewing data released Oct. 11 by Arcadis, a contractor hired by Exxon to gather and analyze soil and underwater sediment samples from dozens of sites the spill affected. An independent expert consulted by the Arkansas Times and InsideClimate News reviewed the data from Arcadis and agreed with the broad conclusions drawn by the Arkansas Department of Environmental Quality (ADEQ), the Arkansas Game and Fish Commission and the Arkansas Department of Health.
Soon after the pipeline break spewed at least 210,000 gallons of a heavy type of Canadian crude called diluted bitumen (or dilbit) into a neighborhood, as well as adjacent wetlands and Lake Conway's Dawson Cove, tanker trucks vacuumed up over a million gallons of oil and contaminated water. In addition, crews scraped over 8,000 tons of crude-saturated plants, soil and debris from the land around the cove.
A remediation plan for contaminated sections of Dawson Cove will likely be designed early next year after Exxon makes a follow-up set of soil and sediment tests available. The cove is a nearly 12-acre section of Lake Conway that drains an 11,000-acre watershed. It is separated from the main body of lake by a man-made dike, built to carry traffic on Highway 89.
When news of the spill broke on March 29, city and county workers sprang into action to contain the oil, creating a makeshift dam to block Dawson Cove from the rest of the lake. Using gravel, dirt and plywood, they plugged the culverts under Highway 89 that allow water to flow from the cove to the lake. (The culverts are no longer blocked.) Soon afterwards, a response team deployed booms and weirs, floating structures that contain the spread of oil in water.
It appears they were successful: Six samples drawn from underwater sediment in Lake Conway just beyond Highway 89 show no evidence of major contamination. The culverts are roughly a mile from where oil poured into the Northwoods subdivision.
Several days after the spill, rainfall caused the water level in the cove to rise. Cleanup organizers were forced to pump water from the cove into the lake to prevent it from flooding a nearby neighborhood. Ryan Benefield, deputy director of ADEQ, said this was done only after determining that the oil had been successfully contained farther back in the cove.
"We had stopped the progression of oil far from that point and we were testing the water on either side of Highway 89," Benefield said. "We had repetitive, dozens of booms and surface-to-bottom weirs throughout the cove, and free oil didn't penetrate past the first couple of booms."
Still, he said, chemicals in the oil may have been carried into the main lake as water was pumped over Highway 89, adding that "water comes in the top of the cove and it goes out into the lake. So, water that had been in the cove when we had free oil, we were pumping into the main body of the lake. But we weren't seeing levels of concern from our testing." So it seems likely that some amount of the Pegasus oil did flow past Highway 89 and into the lake — the question is just how much.
Arkansas Times and its news partner InsideClimate News asked Merv Fingas, a Canadian scientist who has researched and written extensively about oil spill cleanups, to analyze Exxon's data.
"It does not appear that major contamination of the lake occurred," said Fingas. "We may never know about minor contamination."
Fingas said the main body of the lake was already "fairly heavily contaminated" with hydrocarbons before the Pegasus spill.
The status of Dawson Cove is clearer. Oil sheens continue to appear in surface water, as evidenced by ongoing monitoring reports published by Arcadis, and the sampling data show varying levels of polycyclic aromatic hydrocarbons present in the soil and sediment of the cove and its surrounding wetlands. These organic compounds, abbreviated as PAHs, are found in all fossil fuels. They are also formed whenever a flame meets organic matter, so humans create PAHs when they burn wood, smoke tobacco or barbecue meat.
"PAHs are the source of the toxicity in oil," said Jacqueline Michel, an environmental consultant hired by the state Game and Fish Commission with years of experience working on oil spills. "That's what gets inside gills of fish, inside cells of organisms, and causes disorientation — and if levels are high enough, death."
State regulators say levels of PAHs in the cove are above normal, but they're no longer severe enough to cause acute harm to humans or large species such as fish or turtles. However, PAH levels might be high enough to harm small lakebed creatures such as worms and crustaceans that live in sediment and are a crucial link in the food chain of Lake Conway's freshwater ecosystem.
"It's those little communities that build the very foundation for an ecological recovery," said Ricky Chastain, deputy director of Game and Fish. "It's the effects on those organisms that we're trying to assess and see if there's a hot spot where we need to do more remediation to get those levels down."
Chastain said his agency is particularly concerned about PAH levels in a two- to three-acre section of the cove area. The testing sites that show the highest levels are concentrated along the drainage path of the oil as it flowed towards the water and in the spot where the first boom was placed to contain the spill. The dilbit pooled along the length of this boom, explained Michel.
Matt Moran, a professor of biology at Hendrix College, said he worries that the initial round of testing may have missed a significant amount of oil. The sample sites were systematically spaced along the path followed by the oil, but there are some places — such as along that first boom — where the substance may have penetrated more deeply into the ground.
"Because they've chosen these sites that are systematically spaced, they might be missing a lot of the oil that's there," Moran said. "I think what we've found is that a lot of it has been buried in sediment now, and it gets exposed every time it rains hard — something gets stirred up and released back into the environment. I think it's very patchy. But, you want to find those patches, because even if it's only a couple of places, those are places that can provide contamination, which can spread every time we have more rainfall.
"It does appear to be in decline," he continued. "There's less there now than there was two or three months ago. At the same time, it's still fairly easy to find petroleum product down there."
Michel said that it is no surprise oil is still surfacing months after the initial cleanup response ended. There is a trade-off, she said, between allowing oil to remain in the environment and damaging it by removing earth and vegetation.
"We leave a lot of oil behind," she said, "because we know from studies that if you get too aggressive you remove a critical ability of the habitat to recover, especially in sensitive areas like wetlands. Our goal is not to remove all the oil. Our goal is to remove as much of the oil as possible without causing more harm than good."
How PAHs affect aquatic creatures depends on the level of exposure, Michel said. With some toxins, such as mercury or lead, a principal concern is bioaccumulation. That is, over time, repeated exposure to a substance can allow it to build up in the body of an organism, which can slowly add up to a fatal dose.
However, PAHs don't bioaccumulate in larger animals such as fish or humans because their bodies break down the organic toxins.
"It's just like alcohol. PAHs get metabolized in your body," she said. "Vertebrates [such as fish] have enzymes in their gut they use to metabolize these hydrocarbons. A worm or dragonfly doesn't have those complicated body functions."
PAHs weren't the only contaminants of concern in the Mayflower spill. Testing was also done for heavy metals such as mercury and arsenic, and volatile organic compounds (VOCs). Metals and VOCs were of particular concern because dilbit is unusually rich in these toxins. Bitumen, which comes from Canadian tar sands, is too viscous to move through a pipeline in its unrefined form. It must be amply diluted with light hydrocarbons to give it a consistency similar to conventional crude.
After the breach of the Pegasus, those thinning agents evaporated into Mayflower's air, which fueled worries that local residents could have inhaled dangerously high levels of VOCs. Officials with the Department of Health said air quality tests taken since the day of the spill show that exposure levels never topped a threshold of concern. Many citizens of Mayflower insist their symptoms since March 29 prove otherwise. In any case, the soil and sediment tests show that levels of both metals and VOCs currently in the environment are at acceptable levels, and an independent expert consulted by the Times agrees with that conclusion.
Another scientific route to determining the level of toxicity in the environment is to examine organisms that have been exposed to the environmental conditions in question. Jennifer Bouldin, an associate professor at Arkansas State University's Ecotoxicology Research Facility, has been doing just that.
Last summer, the Log Cabin Democrat provided a grant to Bouldin and a graduate student to collect water and sediment samples from six sites in Mayflower, the lake, and the cove. The researchers then immersed three types of tiny organisms — fathead minnows, bloodworms and water fleas — in the water and sediment, watching for indications of harm. Some samples collected from cove water and sediment caused decreased growth and reproduction (see sidebar). Overall, Bouldin's results mesh with the Arcadis report's findings that the level of pollutants now likely constitutes more of an ecological concern than a public health hazard.
Lake already polluted
Environmental scientists testing for pollutants after a disaster must collect samples from non-impacted "background" sites to provide a baseline so they can compare the data. In Mayflower, scientists selected 12 background sites in and around portions of Lake Conway miles to the north of Dawson Cove, spots that could not have been touched by the Pegasus oil. Tests revealed significant levels of PAHs in these locations as well.
Michel said she's not surprised by the background data. The sites that are most heavily contaminated are closest to Interstate 40, indicating that petroleum runoff from the highway is a major culprit. Other likely sources are small boat traffic on the lake and heavy equipment used in nearby construction.
"After oil spills, people go out and take background samples and find a lot more [contamination] than they ever thought was there 'naturally,' " Michel said. "It's always kind of surprising to people to realize there are a lot of contaminants in the environment from human activity. PAHs are ubiquitous. They run off roads, they come out of power plant stacks when you burn coal. They're everywhere. They're in the Arctic, they're in the Antarctic. But that's why we have things like the Clean Water Act, to improve water quality."
Long before the spill, runoff from surrounding lands and I-40 carried pollutants and sediment into Dawson Cove and other wetland areas surrounding the lake, Michel explained. Because wetlands naturally absorb contaminants, they are both essential for preserving clean surface water and susceptible to harm by environmental disasters.
"The whole cove area is really a giant filter for Lake Conway, for all the water coming through that area," Game and Fish's Chastain said. "And all of that material has been removed out of [the cove]. When we're talking about recovery and getting that cove back to where its ecological function has been restored, those very foundational building blocks are important."
If the cove is unhealthy, then the overall health of the lake is diminished. After the spill, cleanup workers cleared about 11 acres of contaminated vegetation in the cove, much of it willow and buck brush. Game and Fish has planted a temporary cover crop of Japanese millet in the mud flats of the cove to hold sediment down and prevent erosion until a remediation plan is designed.
"We want the recovery of the perennial vegetation as quickly as we can," Chastain said. "That's part of the remediation process."
Some skeptics question why Exxon would be in charge of assessing the impact of a disaster it created. But in the wake of an environmental disaster, it is standard procedure for government regulators to identify the responsible party and make sure that party fixes the mess it has made, according to ADEQ spokesperson Katherine Benenati. That includes performing tests on environmental conditions post-cleanup according to established scientific guidelines.
"The goal is to make sure responsible parties live up to their responsibilities," Benenati explained. "Cleanups are handled by many parties, and as the lead agency/state on-scene coordinator we monitor all the cleanup and remediation activities and require use of an ADEQ-certified laboratory."
Following the release of Exxon's report in October, ADEQ requested a second round of soil and sediment testing, which should begin in December. Michel said the two state regulatory agencies are still discussing the details with Exxon. Those results will likely be ready by January. ADEQ officials will then require Exxon to design a remediation plan for the polluted area.
Tammie Hynum, chief of ADEQ's Hazardous Waste Division, said it's unlikely the second round of tests will reveal any major new developments. Requesting follow-up sampling is routine and not a sign that anything was amiss with the first round of data.
"It will allow us to compare it to when the original data set was pulled and let us know whether things are staying the same, or whether we're having an increase or decrease [in contamination]," Hynum said. "We also like to collect data at different seasons to see if temperature and environmental factors affect the results."
As the steward of Lake Conway, Game and Fish also has pushed Exxon to provide additional information beyond the report it submitted in October. The agency wants the company to provide "fingerprinting" of its PAH data that would indicate what contamination is derived from the Pegasus oil and what is from other sources. Because PAHs come in many varieties — the lab tested for almost three-dozen distinct compounds — different petroleum products leave different ratios of chemicals in their wake. Despite its attention to detail, the final report from Exxon made no attempt to use this forensic method to identify contamination sources.
After Game and Fish asked Exxon to fingerprint the data, though, the company balked, saying that that step was not included in its initial sampling plan. Chastain said he expects the company to cooperate with its request.
Chastain said Game and Fish also doesn't expect to find any big surprises — it just wants as much information as possible.
"There's nothing out there that indicates oil got that far, but that fingerprinting is a definitive tool that could tell us what's out there in the lake is not from the oil," Chastain said. That analysis, he said, "would clearly let the public know that contamination from the oil is not in Lake Conway proper. We want to be able to affirm and assure our constituents that we've done everything we can, we've looked at every piece of data, and we're sure that lake is safe and the fish are safe to eat. That's important to us."
Elizabeth McGowan and Lisa Song contributed to this report.
This story is part of a joint investigative project by Arkansas Times and Pulitzer Prize winning InsideClimate News. Funding for the project comes from readers who donated to an ioby.org crowdfunding campaign that raised nearly $27,000 and from the Fund for Investigative Journalism.