Science is about to make understanding the risks from the Kirtland Air Force Base fuel spill a whole lot more confusing.
Will the pollution from a decades-old leak, already contaminating groundwater beneath a southeast Albuquerque neighborhood, reach the nearest municipal drinking water well in 30 years, as an EPA scientist recently calculated?
Will it be 40 years, as an Air Force-funded consultant reported to the Albuquerque Bernalillo County Water Utility Authority Board Wednesday afternoon?
Or is it likely to be sooner than that, as some early estimates and at least three independent reviewers of the current round of models think?
“I’m confused,” Albuquerque City Councilor Rey Garduño, a member of the water utility authority board, said during the board’s most recent meeting. “We have gone from three years to five years to 22 years to 30 years to now it’s 40 years.”
With two computer modeling studies already out and two more to come, and many experts now weighing in, Garduño, his colleagues and the public are about to embark on an exercise in decision-making under genuine scientific uncertainty.
“The line is, ‘All the models are wrong. Some are useful,'” quipped University of New Mexico professor Gary Weissman, an expert on groundwater movement. “That’s because you’ve got so much uncertainty about what’s going on down there.”
Five hundred feet beneath the surface, the groundwater is moving slowly through a complex terrain that scientists can only glimpse narrowly through data collected from single boreholes. Even with a lot of holes drilled to gather data – 87 groundwater monitoring wells to date in the vicinity of the spill – there are significant uncertainties because of the complexity of the underground environment, according to Weissman.
Rich Shean, a groundwater hydrologist with the Albuquerque Bernalillo County Water Utility Authority, said he would prefer more data from the area between the known contamination site and the nearest drinking water wells. “I’d be more confident with the modeling efforts if the data gaps were filled in,” Shean said.
The area where the spill happened and where the nearest drinking water wells are located is a deep bed of old sediments that includes a complex mixture of clays, sands and gravels. Water moves at different speeds through different types of sediments and there will never be enough data to know exactly what is happening where, according to Weissman.
“The problem is, we don’t know how to map the variability down there,” Weissman said.
That means, according to Weissman, that there could be paths where water and contamination could move more quickly, paths the monitoring wells may have missed and which the models being used cannot simulate.
“I would always err on the side of, ‘It’s faster than you think,'” Weissman said.
Mike Amdurer, who heads up the team working on the project for Air Force contractor CB&I, said the public should expect the modeling efforts to come up with different answers. “They have different levels of detail, different starting conditions and different parameters,” Amdurer said. The thing to look for, Amdurer said, is whether they converge on answers that are similar.
More studies to come
The public and decision-makers need to realize that the uncertainties are real, involving questions to which science cannot give crisp answers, said Daniel Sarewitz, an Arizona State University researcher who studies the use of science in political and public policy decision-making. “What they have to ask themselves is how much risk they are willing to put up with,” Sarewitz said.
The stakes are high. Discovered in 1999, the fuel had been leaking undetected from an underground pipe at Kirtland for decades. In 2002, Air Force officials found traces of contamination in groundwater 480 feet beneath the site. In 2007, they realized that massive amounts of fuel had soaked through the ground and was now floating atop the underground aquifer, with some of the contamination dissolving into the aquifer and moving toward Albuquerque drinking water wells.
According to an independent study by the U.S. Centers for Disease Control and Prevention, the public health risk is low. Water agency managers can detect the contaminants in drinking water wells long before they pose a health risk, the CDC scientists found. The real risk is to the water supply itself, according to the CDC. The worst case scenario, according to the federal public health agency, is that municipal drinking water wells would have to be shut down in order to protect the public.
This has raised concerns at the Albuquerque Bernalillo County Water Utility Authority because the drinking water wells in the area are among its most productive.
Prodded by the water utility, the Air Force agreed to fund a contingency plan for the water agency with two goals: to help plan a network of “sentry” wells to provide early warning if the contamination nears the wells, and to provide water supply options if and when the wells are hit. That study is the one that yielded the 40-year estimate.
Separately, a scientist for the U.S. Environmental Protection Agency used similar computer modeling techniques with different assumptions to come up with a 30-year estimate.
The New Mexico Environment Department, doing a simpler analysis, in December estimated the contamination was five to seven years away from the nearest drinking water well.
CB&I, the contracting firm overseeing cleanup work for the Air Force, is working on a modeling study due out later this spring to guide cleanup decision-making. And the U.S. Geological Survey, a federal science agency with expertise in Albuquerque’s groundwater, is doing an analysis of its own due later this year.
An example of the scientific uncertainty was on display last Wednesday when Sharon Minchak of the consulting firm CH2MHill presented her findings to Garduño and the water utility board. Her team’s modeling work suggests that, at current pumping rates, contamination would be sucked into the nearest water utility drinking water well in 40 years.
But when the water utility hired a second consulting firm, Intera, to vet CH2MHill’s work, the Intera researchers concluded there was a risk that the contamination would arrive much sooner. The Intera scientists did not do their own model, but made a similar argument to Weissman’s – that the equations being used underestimate the speed with which the contaminated water can travel through the aquifer.
Sarewitz, of Arizona State, cautioned against spending too much time and energy arguing over uncertainties like these, because uncertainties can never be eliminated.
John Stomp, the water utility’s chief operating officer said that, in a sense, none of the models matter, because all the models are saying the contamination will eventually reach Albuquerque drinking water wells. He doesn’t want to wait to see which model is right. He wants the Air Force and state regulators to take action far sooner to clean up the mess.
“Is it 26 years or 40 years? If we let this thing go that long, we’ve failed,” Stomp said.