全球智库信息集成服务系统

In a new study published in the journal Science, a group of authors estimate that roughly 2.3 percent of natural gas produced in the United States escapes as methane, roughly 60 percent higher than US EPA estimates. This important study synthesizes results from previous research, and offers the most comprehensive view of methane emissions to date.

Spurred by advances in horizontal drilling, hydraulic fracturing (aka “fracking”), and other technologies, the US oil and natural gas industry has grown rapidly in the past decade, leading to economic and energy security benefits alongside concerns over water contamination, human health impacts, air quality, induced seismicity, and more (see some of RFF’s recent work on these topics here and here).

At or near the top of this list of concerns is the impact of “unconventional” oil and gas development on climate change. Methane (CH₄), the primary component of natural gas, is a shorter-lived but more powerful greenhouse gas than carbon dioxide (CO₂). When methane escapes from oil or gas wells, pipelines, compressor stations, or other infrastructure, its climate impact is potent. If more than about 4 percent of natural gas escapes into the atmosphere before it is burned and converted into CO₂, the climate benefits of switching from coal to natural gas in the electric power sector vanish over a 20-year time frame (roughly 8 percent would need to escape for those benefits to disappear over a 100-year period).

Less discussed is the fact that roughly two-thirds of US natural gas consumption occurs outside of the power sector, heating homes and businesses, providing heat for industrial processes, and serving as a feedstock for manufacturing plastics, fertilizer, and more. In 2016, US natural gas consumption reached an all-time high, and while consumption declined slightly in 2017, most projections forecast its continued growth.

To put it simply, methane leaking from natural gas systems contributes substantially to climate change, but until recently, we haven’t had a good understanding of how much of the stuff is escaping.

Recent Research

In recent years, hundreds of researchers have taken on this topic and published dozens of studies, many of which have been coordinated by the Environmental Defense Fund (EDF), which deserves praise for focusing attention and gathering evidence on the topic of methane emissions. These studies have yielded an enormously wide range of results, with estimates of methane emissions in some regions as low as 0.1 percent, and others as high as 10 percent or more.

Figure 1 demonstrates this spectrum, and includes all recent studies that examine either (1) the full natural gas supply chain, or (2) individual oil and gas producing regions, where most emissions appear to occur (many, but not all of these studies have been coordinated by EDF). At the left side of the figure, results from three of the most comprehensive studies (each meta-analyses in their own right) appear alongside two recent EPA estimates, while the right side of the figure demonstrates the range of studies from specific regions. “Alvarez et al 2018” is the study released last week, and its results are broadly similar to the other recent meta-analyses.

Figure 1. Methane Emissions Rates from Recent Studies

Researchers generally take two approaches to measuring methane emissions from the oil and gas sector: top-down and bottom-up. The top-down process estimates emissions by attaching methane sensing equipment to tower networks, aircraft, or satellites. Bottom-up studies measure emissions at or near the source, such as wellheads or compressor stations, then extrapolate from those measurements to produce broader estimates. The new study incorporates bottom-up measurements from nine producing regions (the Arkoma, Bakken, Barnett, Denver-Julesburg, Fayetteville, Haynesville, Marcellus, San Juan, and Uintah regions) then validates those results using top-down data from the same areas. Notably, the study omits several major producing regions, including the Permian basin and Eagle Ford shale, highlighting the ever-present need for more data collection.

While the headline from this study is that EPA undercounts methane emissions by about 60 percent, Alvarez et al. help readers understand the discrepancy by examining each stage of the natural gas lifecycle. They find that the bulk of the disparity is due to differing estimates of emissions during the production stage.

Figure 2 shows these differences at each stage of the natural gas lifecycle. While EPA estimates methane emissions of 3.4 teragrams per year (Tg/y) during production, Alvarez et al. find that emissions are 116 percent higher, or 7.6 Tg/y. If those results are accurate, it would mean that about 1.3 percent of US natural gas production is emitted as methane during the production phase alone.

The shaded areas on the Figure represent the differences between the two sources. While the authors do not attempt to make new estimates for emissions during local gas distribution or oil refining and transportation, they note that some studies have suggested that methane emissions during these stages may also be well above EPA’s estimates. Practically speaking, however, these emissions are likely to be a relatively small portion of the overall whole.

Figure 2. Methane Emissions by Stage

Alvarez et al. argue that a key reason for these differences is that EPA does not fully account for low-frequency, high-impact events, such as equipment malfunctions. These malfunctions, which can be as simple as a valve not properly closing, can contribute disproportionately to the problem, and a small number of “superemitters” appear to be responsible for a large share of methane emissions (for examples, see here and here). The challenge is that these events are often unpredictable, and given the hundreds of thousands of natural gas wells (not to mention millions of miles of pipelines) across the United States, keeping tabs on each potential leak has proved a major challenge.

The Bottom Line

The results from Alvarez et al. provide the best evidence to date on the subject of methane emissions from oil and gas systems, and its results are broadly consistent with previous meta-analyses highlighted in Figure 1. They reinforce two key points which, despite the heated nature of the rhetoric in the debate over “fracking,” are not mutually exclusive.

First, methane emissions contribute substantially to climate change and are likely well above existing EPA estimates. Alvarez et al. estimate that methane emissions from natural gas systems have about the same warming effect on the climate over a 20-year timeframe as all of the CO₂ emissions from US coal plants during 2015. In other words: this is an important problem.

At the same time, the study also provides the strongest evidence to date that switching from coal to natural gas in the power sector produces climate benefits over all time frames. In recent years, natural gas’s displacement of coal in the power sector has been the single largest driver of the decline in US CO₂ emissions, which in 2017 were down to levels not seen since the early 1990s.

So where do we go from here?

First, there is ample opportunity to reduce methane emissions at low cost. A recent report from the International Energy Agency finds that 40 to 50 percent of methane emissions can be avoided at zero net cost, as operators earn revenue from the gas that is no longer wasted. New technologies, such as EDF’s effort to launch a methane sensing satellite, could help pinpoint leaky wells and pipes, reducing costs further. New technologies are also being deployed in cities to reduce emissions from old distribution pipelines.

Second, state and federal leaders can encourage methane abatement through smart policy. While the Trump administration is seeking to roll back Obama-era methane emissions rules, states like Colorado have demonstrated that methane emissions regulations can be implemented effectively, and even garner buy-in from industry. And where policy falls short, industry and non-governmental organizations can help by developing voluntary initiatives to reduce emissions across the oil and gas system.

Finally, it’s important to note that while methane emissions abatement can help reduce some of the most acute near-term impacts of climate change, it is no substitute for a broader policy that covers all greenhouse gases, particularly CO₂, the most important contributor to the long-term risks of a changing climate. To achieve cost-effective solutions to such an important challenge, decisionmakers will need to craft policies that encourage the private sector to identify and act on the lowest-cost mitigation options, whether they are from CO₂, CH₄, or other contributors to climate change.

Sources for Figure 1

(Alvarez et al. 2012, 2018; Brandt et al. 2014; Zavala-Araiza et al. 2015; Caulton et al. 2014; Schneising et al. 2014; Karion et al. 2013; Peischl et al. 2015; Howarth, Santoro, and Ingraffea 2011; Jeong, Millstein, and Fischer 2014; Pétron et al. 2014; Pétron Gabrielle et al. 2012; Robertson et al. 2017; Schwietzke et al. 2014; Heath et al. 2014; Karion et al. 2015; Lyon et al. 2015; Barkley et al. 2017; Mark Omara et al. 2016)