Nobody likes a methane leak.
Now, now, don’t go there. We’re talking about leaks from natural gas production — and more specifically from horizontal drilling and hydrofracturing (or fracking) to extract natural gas locked inside shale.
The Seesaw Saga of the So-called Bridge Fuel to a Bright, Clean, Renewable-fueled Future
- First came the natural gas “gold rush” because it burns cleaner than coal and emits less carbon dioxide (CO2) per unit of energy.
- But there was the problem of cost: Estimates of recoverable natural gas reserves were not extensive, and natural gas prices were relatively high.
- But then along came hydrofracking which made it possible to extract natural gas from shale. This uppped recoverable gas resources and lowered prices, making the coal-to-natural-gas switch more viable.
- But then prospects turned when it was realized that hydrofracking carries the potential for serious environmental degradation (see news articles here and here, and this post).
- And now the debate has heated up even more, with a recent paper by Cornell University’s Robert Howarth, which argues that the fugitive losses (or leaks) of methane (a very potent greenhouse gas) during hydrofracking are so large that the use of hydrofracked natural gas may be worse from a climate perspective than sticking with coal.
A Game Changer on the Game Changer?
The scientific community’s consensus seems to be that the paper’s conclusions are based on thin data but are not beyond the realm of possibility and thus raise issues that deserve further study.
Folks from the natural gas industry on the other hand are crying foul: the paper is a put-up job against fracking.
Which is it? You be the judge. Let’s take a look at the authors’ numbers for methane vented during flowback. These are emissions that accompany the fluids used in hydrofracking when they return to the surface in the days following the fracturing process but before well completion.
An Example of Thin Data: Flowback
The authors estimate that 1.6 percent of all the gas produced during a well’s lifetime is lost during flowback. This number is important because the authors estimate that an unconventional (i.e., fracked) well loses between 3.6 and 7.9 percent of its total production to leakage, while conventional wells lose anywhere from 1.7 to 6 percent. In other words almost the entire difference in methane losses between conventional and unconventional wells comes from flowback.
So where does this 1.6 percent number come from? To derive the percentage, Howarth et al could have used numbers from a study [pdf] by the Environmental Protection Agency on this very subject. In fact they used that study for estimates of the methane emissions from conventional gas- and coal-mining. But for unconventional gas wells, they chose instead to use data on the losses from wells in four basins:
The authors then calculated an arithmetic average of 1.6 percent. But this average is based on just four numbers that vary by a factor of five. In such cases a geometric mean (a method that places less weight on extreme numbers) is generally preferable and that calculation would yield 1.3 percent instead of 1.6. A small point perhaps but one worth noting.
The next question to ask is where those estimated percentages come from. And here’s where things get murky. As duly noted by the authors, the sources they used for the four wells are pretty obscure — trade publications and Powerpoints — a far cry from the rigor of an EPA report. Nevertheless, I was able to pick up a number here and a number there, and basically reproduce the percentages obtained by Howarth et al. But to do so, I had to assume that all of the gas generated before well completion is uncontrolled and vented to the atmosphere. This is certainly not always the case.
Howarth et al do mention toward the end of their paper that “emissions during the flow-back period in theory can be reduced by up to 90 percent through Reduced Emission Completions technologies… [but] are currently not in wide use.” However, they make no effort to assess how the use of such technologies would impact their results, nor do they mention that according to their own source more than 90 percent of the gas generated before completion of the Piceance wells was captured and some of the remaining gas was flared rather than vented.
All in All
So what do you think? Thin data or a setup? I think I know how the folks from the natural gas industry would answer. But there’s a way for them to make this whole thing moot. Capture or flare that gas.