Varun Sivaram

Energy, Security, and Climate

CFR experts examine the science and foreign policy surrounding climate change, energy, and nuclear security.

Print Print Cite Cite
Style: MLA APA Chicago Close


An Enlightening Study on Shale Gas and Water Quality

by Michael Levi
March 13, 2013


A team at Resources For the Future (RFF) led by Sheila Olmstead has a neat new paper in the Proceedings of the National Academy of Sciences (PNAS) that takes a rigorous look at water pollution due to shale gas development in Pennsylvania. (Hat tip: John Quigley.) The team collected thousands of data points measuring shale gas activity and water quality across a wide geographic area and more-than-ten-year span, and then used careful statistical analysis to test a series of hypotheses about how shale gas development might have affected water quality. What’s particularly interesting about this study is that it doesn’t require physical assumptions. It can also shed light on the cumulative impacts of large-scale shale gas development, going beyond analysis at the level of single pads and wells.

The team’s conclusions are fairly straightforward. They find enhanced chlorine concentrations downstream of waste water treatment facilities but not downstream of drilling sites. Chlorine is a good marker of contamination from well flowback. What the RFF analysis suggests is that leaks or spills aren’t statistically detectable, at least at the watershed level, but that impacts of poorly processed wastewater are. That points to the value of focusing on wastewater treatment facilities if one wants to reduce the impact of chemical contamination resulting from shale gas development. The authors are clear to point out that Pennsylvania has made significant changes in the last couple years in how it handles wastewater; whether those are sufficient given the costs and benefits of additional controls remains to be seen.

The RFF team also looks at “total suspended solids” (TSS) – essentially a fancy word for dirt. Here they find the reverse: no statistically detectable increase downstream from treatment facilities but now a non-trivial increase downstream from shale wells. The intuition makes sense: solids are readily removed by treatment facilities; solids displaced directly by shale gas activities, however, are not. But the paper also raises a puzzle. The most obvious way that shale development could increase TSS is through runoff from well pad construction. That should increase when it rains. But when the authors test for such a dynamic they can’t find it. They suggest other possibilities, such as road construction, as culprits. Further insight, though, will require more work.

The study, likely any of this sort, has important acknowledged limitations. It doesn’t say anything about the prospect of future outlier events that might have ugly consequences. And it doesn’t really say anything about damages that are concentrated more narrowly than at the watershed level. Moreover, like any statistical study, it’s always possible that non-detection of damage reflects limitations on the dataset rather than a hard and fast reality. I’d also love to see the authors do some more estimates to put their results in the context of other influences on water quality, and in the context of the growing scale of shale development.

Like I said, though, limits like these are inevitable. The RFF study is exactly the kind of work we need more of when it comes to the environmental impacts of shale development (and for that matter other activities too). One can only hope that regulators learn from the RFF results and that others continue to dig deeper into the issues they’ve raised.

Post a Comment 1 Comment

  • Posted by Mike Constitution

    No, it did not.

    Next question.

Post a Comment

CFR seeks to foster civil and informed discussion of foreign policy issues. Opinions expressed on CFR blogs are solely those of the author or commenter, not of CFR, which takes no institutional positions. All comments must abide by CFR's guidelines and will be moderated prior to posting.

* Required