Almost all of the methane leakage calculated from the Texas research “could be affected by this measurement failure,” according to the paper; “their study appears to have systematically underestimated emissions.”
Original post | In 2011, a Cornell research team led by the environmental scientist Robert Howarth published “
Methane and the greenhouse-gas footprint of natural
gas from shale formations,” a widely discussed paper positing that gas escaping from drilling operations using hydraulic fracturing, widely known as fracking, made natural gas a bigger climate threat than the most infamous fossil fuel, coal. (Natural gas
is mostly methane, and methane is a potent heat-trapping gas.)
There’s been a lot of backing and forthing in the literature on natural gas and the greenhouse ever since. In the meantime, the climate concerns of Howarth and his Cornell colleague Anthony Ingraffea have been pushed hard by foes of expanded fracking, most notably in the updated version of the Josh Fox film “ Gasland.”
As I and Cliff Krauss wrote in 2009, methane leakage is a big problem — and an opportunity (among other things, stanching leaks results in more gas to sell).
But what’s been missing is a comprehensive, independent and open set of measurements taken around drilling operations. That’s produced a pattern I’ve seen on a host of similar environmental issues: A paucity of data leads to a overabundance of assertion.
That’s all starting to change as the first of a planned batch of 16 papers probing emissions data produced by a comprehensive on-the-ground gas measurement initiative has been published today in the Proceedings of the National Academy of Sciences.
The analysis, led by David T. Allen and other energy and environment researchers at the University of Texas, finds that estimates of methane escaping from gas drilling made by the Environmental Protection Agency are fairly accurate, over all, while those from industry critics and some indirect studies of leakage (from aerial measurements, for example) appear far too high. A comprehensive package of background on the research has been posted by the university. [Updates | Seth Borenstein and Kevin Begos filed a helpful Associated Press piece on the study. The New York Times story is here.]
While the researchers found that emissions from a critical stage of well construction — “completions” — are far lower than regulators had estimated, they pinpointed an important under-appreciated source of escaping gas — pneumatic devices powered by the pressure of the extracted natural gas. Authors said this should help regulators and industry close in on ways to further reduce emissions. Here’s Allen describing the work:
Attacks on the credibility of the new work are already beginning, focusing on how it was funded in large part by a batch of gas and oil companies that also provided the research team with access to drilling sites. Leading the charge is Physicians Scientists & Engineers for Healthy Energy, which has posted a five-page critique calling the study “fatally flawed” and noting the small sample size and potential sources of bias.
Other money for this research effort came from the Environmental Defense Fund, which deserves a lot of praise for organizing this effort. Keep in mind that the only way to make such measurements is to be at the fields and facilities where gas is flowing. Aerial studies and other indirect methods of tracking methane are vital, as I’ve stressed. But direct measurement is the key to tightening operations.
Here, John Seinfeld from the California Institute of Technology, a member of an independent advisory panel, describes how integrity was maintained [Update, Sept. 20, 11 a.m.| Oddly, Seinfeld and other panel members are listed as paper authors; it’s hard to see how they can be both, as I explored on Twitter]:
The protocols that were established to maintain independence, sustain transparency and avoid any bias in choosing inspection sites are impressive. The best place to review them is the University of Texas Web site on the project. (It’s important to note, as well, that the researchers say this project has no connection to the University of Texas energy institute that faced appropriate criticism over unstated industry ties.)
Read on for the prime conclusions of the new study, along with reactions from Howarth and responses to his reaction from Lawrence Cathles, another Cornell researcher focused on fracking, and Steve Hamburg, the chief scientist of the Environmental Defense Fund.
Here’s an excerpt from the Environmental Defense Fund news release on the study and the broader research effort:
The UT study, which only deals with the extraction phase of the natural gas supply chain, is the opening chapter in this broader scientific effort designed to advance the current understanding of the climate implications of methane emissions resulting from the U.S. natural gas boom. Methane, the primary component of natural gas, is a powerful greenhouse gas – 72 times more potent than carbon dioxide over a 20-year time frame. The nation’s largest single source of methane emissions is the vast network of infrastructure, including wells, pipelines and storage facilities, that supplies U.S. natural gas. Experts agree that methane leaked or vented from natural gas operations is a real concern, yet estimated emission rates vary greatly ¾ from 1 to 8 percent of total production.
“We know that immediate methane reductions are critical to slow climate change,”said Fred Krupp, president of EDF. “But we don’t yet have a handle on how much is being emitted. We need better data, and that’s what this series of studies will deliver. As we understand the scope of what’s happening across the natural gas system, we will be able to address it. We already know enough to get started reducing emissions, and thanks to the first study, we know that new EPA regulations to reduce wellhead emissions are effective. EPA got it right.”
Launched last year, the overall research effort is designed to collect methane emissions data associated with natural gas production, gathering lines and processing facilities, long-distance pipelines and storage, local distribution, and commercial trucks and refueling stations. A variety of scientific methods are being used across the various studies, including approaches that measure emissions directly at the source and those that use airplanes or towers equipped with sensors to measure total emissions in a given area. In some cases, these methods are paired to provide greater insight and certainty. EDF anticipates all of the projects will be submitted for publication in peer-reviewed journals.
“The scientific talent leading these studies, the partnership with industry and access to their facilities, and the diverse research methods used, gives us the confidence that when the project concludes in late 2014, we’ll be able to greatly increase our understanding of the climate impacts of switching to natural gas from other fossil fuels, through this unprecedented collective research effort,” said EDF Chief Scientist Steve Hamburg.
UT’s peer-reviewed study, the first work published in this overall series, reports data from emission sources from natural gas production – the first part of the supply chain. Study results show that total emissions are in line with EPA estimates from the production of natural gas, yet the distribution of those emissions among activities differ. Methane emissions are lower than estimated by EPA for well completions and higher for valves and equipment used to control routine operations at the well site. All of the data generated in this study are available for public scrutiny.
According to [Steve] Hamburg, UT’s low well emissions finding indicates an early phase-in of EPA’s New Source Performance Standards (NSPS), which requires all new fractured natural gas wells to either burn-off or use “green completions” (an emissions control method that routes excess gas to sales), is working. Results also suggest that these new regulations, which will be fully implemented in 2015, are having the desired effects. No national survey of how many operators currently use green completions is available, but the data suggest that once this practice is required, emissions from this phase of the production process will decline….
Robert Howarth wrote a response a few days ago under the journal embargo. I just posted the full comment on Slideshare. Here’s an excerpt and link, followed by some reactions from Hamburg and from Lawrence Cathles, an earth and atmospheric scientist at Cornell who has published work challenging Cathles and Ingraffea’s conclusions on the climate impact of gas drilling.
Howarth’s reaction:
A major call from our April 2011 paper was for direct measurements of methane emissions by independent scientists. Amazingly, few such studies had been conducted over the preceding decades, and most of the extremely-limited available information was coming from industry sources, usually unverified and undocumented. I am delighted to report that many scientists have taken up the challenge of measuring methane emissions in the short time since our paper came out 29 months ago.
The EDF and industry-funded study by Allen and colleagues is one of these new studies. So far, I have only seen the preprint of their paper to be published soon, and I have not had access to any of the supporting on-line documents. But with that qualifier, I believe Allen et al. have done a fine job of characterizing emissions in the sites they have studied. Their conclusion is that upstream emissions are low, 0.42% of natural gas production (lower than we estimated for shale gas back in our April 2011 paper, and towards the low end of what we estimated for conventional natural gas).
This is good news. It suggests that the oil and gas industry – when sufficiently motivated – can produce natural gas with modestly low emissions. There are a couple of caveats, however.
First, this study is based only on evaluation of sites and times chosen by industry. The Environmental Defense Fund over the past year has repeatedly stated that only by working with industry could they and the Allen et al. team have access necessary to make their measurements. So this study must be viewed as a best-case scenario. Perhaps when industry is motivated, methane emissions can be kept to this relatively low level.
Second, many other scientists have proven over the past 2 years that you can measure methane emissions from gas development without industry cooperation, for instance by using aircraft to fly over operations. Many studies have now been published, and many more presented at national scientific meetings, on methane emissions using techniques which capture the emissions at regional scales and do not require industry permission to sample. The National Oceanic and Atmospheric Administration (NOAA) laboratory in Boulder, CO, has been a very important player in this work, but other labs including many academic institutions have also been pursuing this line of research. All of these studies are reporting upstream emission estimates that are 10- to 20-fold higher than those reported in this new paper by the Allen and colleagues. Most of these other estimates from NOAA and independent non-industry funded academics are in fact at the upper end of what we estimated in our April 2011 paper, or far higher. This should be truly alarming to anyone concerned about global climate change.
How can we explain this huge discrepancy? My take at the moment – again without access to the data sources and analysis behind the Allen et al. paper – is that the gas industry can produce gas with relatively low emissions, but they very often do not do so. They do better when they know they are being carefully watched. When measurements are made at sites the industry chooses and at times the industry allows, emissions are lower than the norm. But the norm may well be 10-fold or more higher, based on the other accumulating research by NOAA and other independent researchers. [Please read the full comment here.]
I asked Steve Hamburg from the Environmental Defense Fund to react to Howarth’s assertions. Hamburg is hardly someone who can be painted as an industry apologist. He was an author of a paper in the same journal last year with this title: “Greater focus needed on methane leakage from natural gas infrastructure.”
But he said the work is trustworthy, noting that it would be very difficult for gas companies to speedily alter practices at fracking sites to clean up their act ahead of the researchers’ visits, given very short lead time on choosing well sites and that procedures and equipment are developed well ahead of time. He added that this work complemented, rather than clashed with, the valuable aerial surveys, some of which the Environmental Defense Fund is also paying for.
Lawrence Cathles at Cornell said he had confidence that more observation, from direct measurements to global tracking of methane concentrations, would allow gas development to proceed while limiting climate impacts. Here’s a link to his full comment.
In the meantime, the Public Accountability Initiative, which has been doing valuable watchdog service probing for conflicts of interest in university research on natural gas, found at least one issue with the new study. Here’s the key line in a note sent to me just now by Kevin Connor at that group:
The authors of the study declared no conflicts of interest, according to the article, but one author,Jennifer Miskimins, is currently the employee of a petroleum engineering firm, Barree & Associates, and has been since 2012. The firm offers a range of consulting services related to fracking. The paper lists her affiliation as Colorado School of Mines.
It’ll be interesting to see if this was the unintended result of a recent job change or the like or is a legitimate departure from the conflict of interest reporting standards of the journal.
|Update, Sept. 20, 10:30 a.m. |
There are some significant issues being raised about the procedures used to guarantee the scientific integrity of this research despite the unavoidable involvement of industries with an interest in the outcome. Here’s one Twitter thread in which I note that members of what had been described as an independent scientific advisory panel are also paper authors:
|Update, 9:55 p.m. | Steven Hamburg, the chief scientist at E.D.F., sent this comment after scanning the initial batch of comments tonight:
I appreciate the sincere concerns being expressed here. Regrettably the conversation seems to be focused more on impressions than on the data reported today in the Proceedings of the National Academy of Sciences. All of that data and all of the methods used to generate the data are available for others to review, analyze and interpret (see UT web site) – that is as transparent as it is possible to be.
Some have claimed that the companies picked the timing and locations of sampling. That assertion is false and based on nothing more than conjecture. The paper clearly lays out the basis for how the sites were selected and the companies were given no input or control over that process. The scientific team gave its criteria to the companies and all sites meeting those criteria were made available. There was no cherry picking and no sanitizing. Natural gas wells are not restaurant kitchens that can be ‘cleaned up’ before the inspectors come — emissions levels from activities like well completions are determined by geology and the equipment on the site, the companies couldn’t have influenced them to any significant degree if they tried. And we saw zero evidence that they tried.
The wells were in areas of more recent drilling because measurement of well completions was the top priority of the study — you can’t measure emissions from well completions of old wells. So there is a skewing toward newer wells, but that is transparent in the study.
We do not claim that this study demonstrates everything is fine when it comes to methane emissions. We never would. The study does indicate that EPA got it right when it required green completions for shale gas wells. That is why completion emissions in this study are lower — because the regulation was beginning to have the desired effect – remember the data reported were collected in 2012. Now those rules should be extended to shale oil wells and hybrid oil-and-gas wells. And there’s plenty to be done reducing emissions from valves and other equipment. Those emissions were higher than EPA estimates and must be addressed.
A word about flyovers. EDF agrees that top down studies are important — that is why we directly funded NOAA to make top-down measurements when the larger study began (more than a year ago). We are as eager as anyone to see those results published and available to the public. We will not speculate about the results until they are published. We need to wait until the results and data are made transparent as has been done with the UT results.
While the flyover results from one study completed by NOAA suggest higher emissions, there is additional work that has yet to be published. Results from the study that has been published are from one day over a relatively small production area, we need more data before we can draw robust conclusions. While flyover data will provide important insight it brings its own challenges, as ensuring that you can differentiate among methane sources can be difficult.
Bottom line, the data collected and reported in the UT paper released today in PNAS reflect direct measurements of methane leakage from activities for which there had been no or little data. The paper was subjected to peer-review and made all methods and data transparent. The paper was accepted for publication in one of the most prestigious science journals. A conversation about the results — what they tell us and what they do not — will be far more productive than one that simply restates preconceived notions of good and bad. Science requires open and honest discussion of data and ideas. I hope that can begin.
| Update, 4:30 p.m. | Gary Susswein, a spokesman for the University of Texas, sent this reaction to the concerns about this author:
Transparency has been our top priority throughout this process. UT Austin required all authors to disclose any potential conflicts of interest and to affirmatively approve the language about them that was included in the PNAS article. We are obviously anxious to look into the issue raised in this memo.