There is another shale-gas-in-water story doing the rounds. Robert Jackson, a researcher from Duke University in North Carolina, says that he has found higher concentrations of alkanes in water samples near shale wells than in equivalent sites that have not been subject to hydraulic fracturing.
Interestingly the study appears to have been conducted in Pennsylvania, where in towns like the now notorious Dimock, water supplies have long been contaminated by gas naturally leaking from the ground. In fact, according to the Supplementary Information (Fig S1) several of the sampling sites came from around Dimock itself. I don't quite follow why, if you were trying to learn something about the effects of hydraulic fracturing on groundwater, you would choose to study somewhere where the water was already contaminated. Would it not make more sense to go to, say, the Bakken or Utica shales?
I'm not the only one thinking along these lines. The Telegraph, discussing the same story, notes that no methane contamination has been found elsewhere:
Studies in different shale formations have also shown no signs of escaped gas, suggesting that leakages could only be a problem in certain areas.
A paper worthy of closer examination I would say.