Good News And Bad About America's Leaking Methane

First the good news: most of the nation's natural gas leakage can be halted by plugging up a relatively small number of "super emitters," according to the Stanford researcher who first documented the government's underestimate of the methane problem.
"If you could solve the top 20 percent of the leaks, find them and fix them, you’re at least 90 percent of the way there," Adam Brandt said last Monday, briefing the public and colleagues at Stanford's Precourt Institute of Energy on a study yet to be published.
Now the bad news: the relatively small number of super emitters is still a vast number considering the size of the natural gas infrastructure and how little we know about it.
"There’s something like 80 to 100 million end-point connections to the gas grid, so this is an incredibly complex large distributed infrastructure," said Brandt, an assistant professor of Energy Resource Engineering. "Each one of these points can have many possible points of leakage."
If the leaks are not plugged, the escaping methane threatens to overwhelm any benefit the nation gains by switching from high-carbon coal to low-carbon gas. Unburned, methane is about 33 times more potent as a greenhouse gas than carbon dioxide over 100 years.
Sealing those leaks will be a challenge in part because of shoddy information, as evidenced by the uncertainty in that estimate of end-point connections to the gas grid: give or take 20 million. There are other uncertainties:
In 2012, the Environmental Protection Agency included methane emissions from hydraulic fracturing operations in its annual greenhouse gas inventory. The agency estimated the emissions based on 8,000 frack jobs it knew to be active that year, Brandt said, but the American Petroleum Institute corrected the agency: there were actually 15,000 that year.
Most methane comes not from fracking, Brandt has emphasized, but from other sources such as the livestock industry, garbage dumps and other facets of the oil and gas industry.
For example, methane escapes from release valves and faulty connections on oil-storage tanks. The government estimates there are two to four tanks per well, but no one knows for sure. When Brandt and his team contacted North Dakota regulators to get a more precise number in that state, they were told that information is kept on paper forms stored in six filing cabinets.
"If you generate an emissions factor—emissions per tank per year—you ought to know how many tanks there are," Brandt said, "and in many cases we don’t."
Brandt conducted his latest study, "Super-emitters in the Natural Gas Sector: Prevalence and Implications for Solving the Methane Leakage Problem," with Garvin Heath from the National Renewable Energy Laboratory and Dan Cooley, a statistician at Colorado State University. The study begins to narrow in on the discrepancy Brandt first revealed in 2014 between government estimates of methane emissions and the much larger volumes detectable from sensors on towers, aircraft, and satellites.
The study, which has just been accepted for peer review, incorporates data from 21 such surveys of emissions conducted over 15 years—including about 20,000 observations of leaking methane—to conclude that 20 percent of super emitters are responsible for nearly all of the leakage.
The top 5 percent are responsible for more than half.
But even this data may underestimate emissions, Brandt warned, because of sampling bias. Scientists sample emissions with the permission of drilling companies and delivery utilities. So sampling may have occurred more often at sites where leaks are less likely, because scientists may not have been welcomed to sample the worst.
The Aliso Canyon natural gas leak in Southern California, captured with an infrared camera in December of last year: