Update: Point-Counterpoint on Shale Gas Methane in Water Wells
Debate on water-well contamination from shale gas extraction.
Last May, Duke University post-doc Stephen Osborn et al published a paper in the Proceedings of the National Academy of Sciences (PNAS) showing evidence of drinking-water contamination by methane in the vicinity of shale gas drilling activity. (See my original post on the paper.)
Following on its heels came a media blitz and gobs of criticism, largely from industry sources, disputing the findings and dismissing the science. (See here, here and here.) The comments and queries were so numerous that the authors published answers online to the most common ones.
Discussions online and in the media are fine, but the real arena for scientific debate should be and is the peer-reviewed literature. The earnest start to that debate comes in the form of two peer-reviewed critiques on the paper along with a reply by the authors. They were published in last week’s online edition of PNAS here and here ($ub req’ed). Below are some of the salient items, each prefaced with summaries of the original paper and/or the new comments, and followed by comments from the original paper’s authors and yours truly.
Volley 1 From Saba and Orzechowski
By far, in my opinion, the more substantive comments came from Tarek Saba of Exponent, Inc., and Mark Orzechowski of Civil and Environmental Consultants, who based their comments on Osborn et al’s data, noting apparent inconsistencies.
1. Spatial correlation of methane contamination
A major finding of the Osborn et al paper: Drinking water wells close to active shale gas sites tended to show enhanced methane concentrations; water wells far from active sites did not.
Saba and Orzechowski comment: Water samples collected by Osborn et al from the Genesee formation aquifer (which overlies the Marcellus in upstate New York) showed the opposite correlation: 1.5 milligrams per liter in water far removed from active wells and 0.3 milligrams per liter near an active well.
Osborn et al reply: Both 0.3 and 1.5 milligrams per liter are essentially background concentrations, much lower than the average concentration of 19.2 parts per million they found in methane-contaminated drinking water near shale gas wells.
Saba and Orzechowski comment: Unpublished data from the Pennsylvania Department of Environmental Protection (DEP) show the existence of high concentrations of thermodgenic methane in Tioga County that is not related to gas extraction.
Osborn et al reply: The DEP data from Tioga County are of questionable applicability because the region has “been used for commercial has “been used for commercial underground gas storage, leading to documented leaks into well water.” With respect to there being other sources of thermogenic methane beyond shale gas extraction, Osborn et al agree. Additional data collected in June 2011 from a natural methane seep near the Catskill study area showed that this methane was also thermogenic, but was isotopically distinct from the methane found contaminating wells near active shale gas wells in the area.
2. Isotopic Issues
Osborn et al used isotopic labeling to connect the methane contamination in drinking wells with methane from shale gas.
Saba and Orzechowski comment: Osborn et al found a large degree of variation in the isotopic signatures of the methane in the contaminated wells. Such variation calls into question the conclusion that the methane contamination was from shale gas.
Osborn et al reply: Such isotopic variation is not inconsistent with their conclusions as it is within the published range for shale gas in the region.
Saba and Orzechowski comment: The lack of isotopic data on ethane in the drinking water wells is a serious deficiency. Without such data Osborn et al cannot “conclusively determine” that shale gas operations are source of methane in Catskill formation water wells.
Osborn et al reply: A point of disagreement.
Volley 2 From Schon
The comments of Samuel Schon of Brown University were far less germane in my opinion.
3. Not Coming From Deep Fractures
Osborn et al use their data to argue that methane from shale formations is contaminating drinking water wells in the vicinity of fracking operations. They do not claim to know the mechanism and in fact suggest that it is unlikely to be from fractures opened up in the deep source rock of the Marcellus during fracking. The most likely explanation they propose is leakage from the shale gas wells themselves.
Schon echo Osborn et al’s conclusions with a twist.
Schon comment: Methane contamination is highly unlikely to have come from the shale gas entirely because seepage from the deep formation is unlikely: “The data presented simply do not support the interpretation put forth that shale-gas development is leading to methane migration from the Marcellus into shallow groundwater.”
To me Schon’s comment is a bit disingenuous. Osborn et al’s conclusion does not depend on this one interpretation — that the methane contamination is caused by shale gas movement from the Marcellus via fractures opened during hydraulic fracturing.
Osborn et al reply: Schon “offers no alternative to explain the differences in chemistry and gas concentrations we observed.”
In fact, Schon does offer an alternate explanation and it is the same explanation Osborn et al favor. Schon states that “inadequate well construction [as suggested by Osborn et al (1)] could conceivably enable methane migration from shallower horizons.”
4. The Industry Has Got It Covered.
As noted above, Osborn et al speculate that the contamination might be coming from leaking well casings.
Schon comment: “Industry best practices recently codified in Pennsylvania drilling regulations (5) ensure that the region’s substantial shale gas resources
can be developed safely and environmentally responsibly.”
My reply: It is always interesting to hear arguments that natural gas extraction is not causing contamination that are then followed by an argument that new practices mitigate all problems — in this case practices that were codified in 2011. The fact that new practices were needed kinda suggests that there was some recognition that there were problems with the old practices. And it follows that it might be prudent to test the effectiveness of these new practices before concluding that the issue is moot.
Along these lines Osborn et al suggest that one might “want to test the effectiveness of the new regulations before concluding that they ‘ensure success.'”
5. Researchers Discussing Work in Popular Press
The Osborn et al paper got a lot of media attention, and some of those published words in the media were straight from the researchers themselves.
Schon took issue with that latter kind of coverage.
Schon comment: Here’s one example: “These data especially do not justify coauthors’ reports in the popular press ["Strong Evidence That Shale Drilling Is Risky," Philadelphia Inquirer, May 10, 2011 (4)] about the process of hydraulic fracturing.
Osborn et al do not reply to this one. I see this as being really off point, in the context of a comment on a scientific paper. The fact that the authors published a lay piece on their own work is hardly a valid criticism of the work itself. More appropriate here would be a letter to the editor.
Agreement at the End?
Osborn et al’s final point is an acknowledgement that they completely agree with Saba and Orzechowski and Schon that a larger study, complete with measurements of well water before and after the onset of shale gas drilling, is needed. More details of what they would propose to do can be accessed here: www.nicholas.duke.edu/hydrofracking/responses-about-gas-shale.
Who knows. With the types of data Osborn et al propose collecting, we might just reach consensus on this issue. Consensus — what an odd concept.