Is the shale gas boom good or bad for climate? It largely depends on methane.
Methane, the primary constituent of natural gas, is a double-edged sword in climate terms. It burns much cleaner than coal—about half of the CO2 emissions and far less of most other pollutants for the same energy output. But released directly into the atmosphere it’s a much more potent GHG than CO2, at least over the short term. This leads to a pretty simple prescription: burn all the methane you pull out of the ground. But it is probably impossible to prevent all leaks, so the advantage of natural gas over coal, and its prospects as a “bridge” fuel, depend on exactly how much “fugitive” methane emissions occur. Estimating real world emissions is very hard, and has been surrounded by controversy. At something a bit over 3% emissions of total methane production, natural gas no longer beats coal in GHG terms (though this, too, is controversial, and depends on assumptions and policy goals).
EPA’s “bottom-up” estimates of fugitive methane emissions are pretty low—less than 2% of gas produced for all stages of gas production, processing and distribution, though these numbers are much lower than EPA’s previous estimates. Other “top-down” studies have found much greater emissions, as high as 6-12% in a Utah field. All of these studies have come under considerable criticism, with the result that fugitive methane emissions are widely regarded as the most important unanswered question on the environmental impact of the gas boom. A new study, funded in part by the Environmental Defense Fund and long-awaited by researchers, aims to fill part of that that gap. Results released this week look at gas production by measuring methane emissions from almost 200 natural gas production “sites” including production wells, newly drilled wells, and related facilities. Results for other parts of the gas lifecyle will be released later.
The results are good news for natural gas, including that produced by fracking. The study estimates emissions from the production phase at 0.42% of production, less than (but very close to) EPA’s bottom-up estimate of 0.47%. The study calculates sampling error, and even if the 95th percentile estimate is chosen instead of the mean, its estimated emissions are still less than EPA’s. If future estimates from the study for other parts of the gas lifecyle are similarly close to EPA’s estimates, we could then conclude that natural gas development contributes less (relative to energy output) to global warming potential than coal.
The similarity in final estimates doesn’t mean that the two studies were in agreement on the details. The EDF study found much lower methane emissions than EPA did for completion flowbacks from fracked wells, possibly (the authors speculate) because surveyed wells used “green completion” techniques. These will soon be required by EPA performance standards, but whether this small sample reflects average industry practice is not known. On the other hand, EDF’s estimates for methane emissions from chemical pumps, pneumatic controllers, and equipment leaks are larger than EPA’s, a result that the team plans to subject to further sampling and analysis. Despite these differences, the close correspondence in total emissions helps validate both approaches.
While the EDF study is to be commended for its presentation of uncertainty bounds, they have been created under the assumption of independence of the sources of uncertainty. This assumption may be questionable if given operators or given locations of wells create more favorable conditions for leaks in a variety of systems. If emissions rates across activities are correlated, the uncertainty bounds would be much larger. Therefore, as always, more research is needed.
This study, however—pending further results—is important new evidence in favor of gas as a climate-friendly fuel. That does not mean that more cannot or should not be done by companies and regulators to address fugitive methane emissions (and intentional venting). But it does make claims of pervasive major methane leakage harder to support.