A 2017 study that claimed shale gas development is a “climate detriment,” based on findings of high leakage rates in the Marcellus Shale region, was retracted this week due to an “error in wind measurements” that led to an overstated emissions estimate.

Led by researchers from the University of Maryland and published in the American Geological Union’s (AGU) Journal of Geophysical Research Atmospheres last yearthe peer-reviewed study claimed to find a methane leakage rate of 3.9 percent based on three flight measurements in September and August 2015. “Finding” such a high leakage rate led the authors to reach some bold conclusions when the report was released last year. From the original study:

“Hence, the production of energy from CH4 extracted from our surveyed area with current technologies is a climate detriment, over the next two decades, if our measured leak rate is representative of typical conditions for extraction in the Marcellus Shale…

“At our measured leak rate of 3.9 ± 0.4% for the Marcellus Shale in southwestern Pennsylvania and northern West Virginia, the use of natural gas rather than coal for combustion will result in a relatively greater climate impact over the next few decades.” (emphasis added)

But upon discovering their wind measurement gaffes in October, the researchers have changed their tune, concluding that the leakage rate was roughly half of the level reported in the study. From the retraction notice:

“[T]he original wind measurements led to an overestimate of methane emissions from oil and natural gas operations. A reanalysis with corrected winds reduced the total estimated emissions by about a factor of 1.7, with a correspondingly larger reduction in emissions of methane attributed to oil and natural gas in the southwestern Marcellus Shale area. This is expected to reverse a conclusion of the paper, which had asserted that leakage from oil and natural gas extraction in this region results in a climate penalty compared to the use of coal.” (emphasis added)

Putting it another way, rather than supporting the original conclusion that shale development is a “climate detriment,” this study’s corrected data show a leakage rate far below the 3.2 percent threshold for natural gas to maintain its climate benefits, adding to a long list of studies that have reached the accurate conclusion that natural gas is a climate winner.

With updated data, this is now the second study released in the past three months concluding that leaks from Marcellus shale gas development are well below the 3.2 percent threshold. As EID reported last November, a Department of Energy-funded study conducted in the northeast Marcellus Shale region by researchers from Penn State University found that methane leakage rates from natural gas wells and other infrastructure are roughly 0.4 percent of production. The comprehensive study — which used a combination of “top down” and “bottom up” methodology — included measurements taken from 12 aerial flights.

Considering the complete reversal of the study’s conclusion, the following March 2017 tweet from the AGU trumpeting the Maryland study’s original conclusion hasn’t aged well.

Though the University of Maryland study received virtually no media coverage last year and hasn’t been cited in research, it’s clear that the authors were hoping their original findings would be used to influence public policy; more specifically, to justify greater methane regulations. From the study,

“Given the current low cost of natural gas, a purely economic driver to reduce the atmospheric release of CH4 by leakage does not exist. Thus, regulatory approaches are needed in order for energy derived by the combustion of natural gas to be a net climate benefit, relative to energy derived from coal combustion.

“Although new regulations on the completion venting step of the hydraulic fracturing appear to have improved relative leak rates, further actions are needed in order to reduce natural gas losses.” (emphasis added)

Though the authors should be commended for admitting their error and trying to correct the record, erroneous wind measurements are not the only questionable elements of the study, bolstering the argument that it deserves further scrutiny.

For instance, the paper cites several thoroughly debunked top-down methane studies to support its claim that the original finding is “broadly consistent with the results from several other recent studies based on atmospheric obsrvations.”

The most notable of these debunked studies cited by the authors is Caulton et al. (2014). The report, led by Princeton University professor Dana R. Caulton, was conducted in conjunction with noted anti-fracking Cornell University professors Tony Ingraffea and Robert Howarth — the Lennon and McCartney of hyperbolic methane studies based on eronious data. Interestingly, the Maryland researchers thanked Caulton for “helpful comments” in the acknowledgements section of their study.

Similar to the University of Maryland study, Caulton et al. relied on data from a handful of flights over southwestern Pennsylvania and ran modeling exercises on the data to determine the source of the detected emissions.  That study’s highest methane reading was over Greene County, an area that also happens to be the state’s largest coal producing region. Finding the highest methane readings over this area is noteworthy, as several state laws require coal mines to have methane ventilation systems to prevent explosions and to protect workers.  However, the researchers just briefly noted the presence of coal mines – stuffing the mention away in the supplemental information section.

As for the Maryland study, it includes a map showing eight coal mines in its study area, and the researchers acknowledge that coal mines are a major source of methane emissions. With this in mind, the authors also admit that their estimates of methane emissions attributable to coal, based primarily on EPA Greenhouse Gas Reporting Program (GHGRP) data, are uncertain. Translation: they could be higher than the researchers estimate they are. From the study:

“However, a substantial source of CH4 was found to contain little ethane (C2H6), possibly due to coalbed CH4 emitted either directly from coalmines or from wells drilled through coalbed layers.”

“The GHGRP (EPA Greenhouse Gas Reporting Program) only reports underground coal mines and does not report emissions from surface or abandoned coal mines.”

“Uncertainties in emissions of CH4 from coal mines and other non-O&NG sources contribute an additional uncertainty to the range of possible CH4 leak rates inferred from active O&NG operations in our study area.” (emphasis added)

Therefore, even though the researchers corrected a major flaw in their study, resulting in a lower — and presumably more accurate — emissions estimate, the report certainly deserves greater scrutiny based on data uncertainty and questionable associations with academics tied to anti-fracking researchers. This is especially true considering the study was completely funded by taxpayer supported agencies. From the study’s acknowledgements section:

“This work was supported by National Science Foundation (NSF grant CBET-1438400), Maryland Department of Environment (MDE contract U00P4401029), National Institute and Science and Technology (NIST Cooperative Agreement 70NANB14H333), National Oceanic and Atmospheric Administration (NOAA), and the National Aeronautics and Space Administration (NASA) AQAST program.”

Fortunately, the retraction notice states that, “The authors are in the process of submitting a new manuscript based on an updated analysis that will describe the process to correct the erroneous wind measurements used in the original manuscript, provide a more accurate estimate of the methane emissions, and assess the implications of the fossil fuel production from the Marcellus Shale.”

EID anxiously awaits the release of the revised study.