Many of you may have seen this kind of video, showing the effects of methane in drinking water near some shale gas extraction wells:
Before now, I've never known what to make of this kind of thing. Is this a very rare, if spectacular occurence, or is it common where shale gas drilling goes on?
Now, there is a paper in this week's PNAS, Methane contamination of drinking water
accompanying gas-well drilling and hydraulic fracturing by Osborn et al, researchers at Duke University. It appears to answer the question, and the answer is not good.
Basically, they sampled well water from 68 private wells above the Marcellus and Utica shales in Pennsylvania and New York. About a third were in an "active extraction area", defined as within 1km of a gas well, while the rest were further away. They then measured the concentration of methane in the water.
The results are as follows:
Technical Measures for the Investigation and Mitigation of Fugitive Methane Hazards in Areas of Coal Mining. That report establishes the following guidelines for action:
You can see that a good number of the samples above fall into the above 28mg/L category that is potentially explosive and calls for immediate mitigation. Meanwhile, a considerable majority are above at least the 10mg/L threshold for concern and warning.
There are a number of other interesting points in the paper that I'll briefly mention:
- They do isotopic analysis to establish that the source of the excess methane near gas wells is thermogenic methane from sediments, not biological methane from near the surface.
- They also tested the water for various possible toxic chemicals from the fracking fluid, but didn't find anything.
- They also didn't find evidence of high levels of salt or radioactivity suggesting of deep saline aquifer water intruding into the shallow aquifers, at least not in the areas they sampled.
I have just one potential caveat with the paper. They don't say anything about exactly how they selected private wells to sample. They report probability numbers suggesting that they treated it as a random sample. That could be reasonably implemented, for example, by selecting houses in the study area at random and approaching the homeowners to see if they'd agree to testing. However, something like advertising in the local paper for people who'd like to have their water tested risks a self-selected population that already suspected problems with their water. The results would still be interesting, but much less statistically meaningful. In the absence of any description in the paper, I'm giving them the benefit of the doubt, but I have an email in to the authors querying this point.
At any rate, if this paper holds up, it appears that drilling within 1km of household wells presents an unacceptable risk to the residents.