Shale-gas extraction has transformed the US energy landscape, but its environmental effects are unclear. Credit: Les Stone/Reuters/Corbis

As shale-gas operations expand across the United States, industry officials and environmentalists are at loggerheads over whether or not shale-gas extraction can contaminate groundwater. Now researchers have traced low levels of methane and other contaminants to a source of shale gas: the sprawling Marcellus Formation, which lies beneath much of New York state, Pennsylvania, West Virginia and Ohio (see ‘On tap’) .

The study, led by researchers at Duke University in Durham, North Carolina, expands on an earlier analysis of drinking water in northeastern Pennsylvania, where energy companies have used hydraulic fracturing (fracking) to crack the Marcellus Formation and release gas. In that work, the researchers found that contamination rates increased with proximity to wells (S. G. Osborn et al. Proc. Natl Acad. Sci. USA 108, 8172–8176; 2011). Their latest analysis, published on 24 June, goes a step further, by tying the chemical fingerprint of the ground­water contaminants to the gas being siphoned out of the ground some 2,000–3,000 metres below (R.B.Jacksonet al.Proc.NatlAcad.Sci.USAhttp://doi.org/m3j;2013).

“The problems we’ve seen are probably more common than people realize,” says Rob Jackson, director of Duke’s Center on Global Change and lead author of the paper. Jackson stresses that the contamination is probably due to poor well construction, rather than hydraulic fracturing itself. But he says that the results are another “wake-up call” for the industry to improve its drilling operations.

The study is the latest salvo in an ongoing debate about the environmental impacts of shale-gas extraction, which has transformed the US energy landscape and is now moving abroad. Such operations are also causing concerns about air pollution and methane emission, which could offset some of the climatic benefits of replacing coal with natural gas. But fears about the potential impact on ground­water resources have taken centre stage in a number of high-profile disputes between scientists, regulators and industry.

Last week, for example, the US Environmental Protection Agency (EPA) cancelled its probe of groundwater contamination in Pavillion, Wyoming, more than two years after issuing a controversial draft report pinpointing fracking as the cause. The decision leaves the investigation in the hands of state officials, who had criticized the EPA’s assessment. And in Dimock, Pennsylvania, which lies in the area covered by the Duke research, allegations of ground­water contamination led state regulators to halt some drilling operations temporarily in 2009 and to fine the primary operator, Cabot Oil and Gas Corporation of Houston, Texas, which settled a lawsuit out of court but did not admit fault.

The Duke team’s data provide further evidence that shale development could harm ground­water quality in some areas. Trace amounts of methane are often found in drinking water, and the levels identified by Jackson’s team do not necessarily represent a health threat. But in extreme cases, methane contamination can be dangerous: homeowners can sometimes ignite the gas in water coming out of their taps.

Jackson’s team found methane in 115 of the 141 shallow drinking-water wells that it sampled. To trace its source, the team examined carbon-isotope ratios of the methane molecules — these ratios differ depending on whether the gas was produced by microorganisms in relatively shallow water or by heat and pressure deep in the Earth. The team also looked at the relative concentrations of ethane, propane and helium — other common by-products of gas extract­ion that are rare in shallow groundwater sources.

The results suggest that home­owners living up to 1 kilometre from shale-gas wells probably have groundwater contaminated by gas from the Marcellus Formation. But the team did not find evidence that chemicals used in fracking migrated from depth to contaminate aquifers. Jackson says that the methane is probably from leaks in the well casing itself, which would allow direct contamination. The data also suggest that some gas migrated up from geological layers between the Marcellus Formation and the groundwater table.

The data “raise important issues”, says Mark Zoback, a geophysicist at Stanford University in California, who served on a high-level panel that in 2011 recommended improving industry practices to minimize environmental impacts. “There’s a real need in the future to have this kind of sampling before, during and after shale-gas development.”

Officials from Cabot, which operates most of the wells in this area, declined to comment on the study. Steve Everley, a spokesman for Energy In Depth, a research and advocacy arm of the Independent Petroleum Association of America, based in Washington DC, says the Duke study is inconclusive. He notes that recent baseline testing by the US Geological Survey found non-microbial methane in two groundwater wells in northeastern Pennsylvania even before drilling.

Jackson says that his results do not necessarily mean that all drilling operations will have problems. Other research by his team, now in the press in Applied Geochemistry, found no evidence of contamination in a shale formation in Arkansas. More importantly, he says, the results suggest that the problem is relatively simple to fix.

“This is about well integrity,” he says. “The industry knows how to do this, and they work hard to maintain well integrity.”