The amount of methane leaking from the nation’s oil and gas fields may be 60 percent higher than the official estimates of the Environmental Protection Agency, according to a new study in the journal Science.

The study, led by a group of scientists from the Environmental Defense Fund (EDF), presents some of the most compelling evidence to date that switching to gas from dirtier fuels like coal might not be as effective a climate strategy as its proponents suggest unless the gas industry improves how it controls leaks.

“It starts to have a material effect on just how clean a fuel natural gas really is,” said Ramon Alvarez of EDF, one of the authors of the study.

The authors estimated, conservatively, that methane equivalent to 2.3 percent of all the natural gas produced in the nation is leaking during the production, processing and transportation of oil and gas every year. That doesn’t count leaks from local delivery lines, another widespread problem.

This much leaked methane would have roughly the same climate impact in the short-term as emissions from all U.S. coal-fired power plants, the authors found.

Another way to put it: This rate of leaking methane is just as bad for the climate in the short term as the carbon dioxide that results from burning natural gas for fuel.

Infrared Cameras + Years of Spot Checks

Methane is a potent short-lived climate pollutant that doesn’t linger in the atmosphere nearly as long as carbon dioxide, but has a more powerful climate impact in the short term. With oil and gas production rising rapidly, it’s especially urgent to bring these emissions down.

The Trump administration has been attempting to roll back various federal regulations on emissions of methane. Its approach is tangled up in several court cases, some involving EDF.

The study in Science is the culmination of years of work by the team at EDF and other research scientists. In 2011, EDF launched a project with researchers from over 100 universities and with joint funding from foundations and the natural gas industry. The goal was to look at a wide swath of issues related to methane leaks and ascertain just how much methane was getting into the atmosphere.

The study released today builds on that earlier work, as well as research by scientists outside the project. The authors analyzed measurements from more than 400 well pads in six basins, from various facilities and components used in oil and gas production, and from aerial surveys across regions with oil and gas infrastructure. The aerial surveys confirmed the spot check findings, making the results more robust, Alvarez said.

It resulted in a comprehensive estimate for methane emissions.

Biggest Source: Leaking Tanks

One notable finding was that acute episodes of leaking due to sudden equipment failure or operator errors—not chronic conditions—accounted for a large amount of the deviation from official estimates of leakage.

Using helicopter surveys with infrared cameras, Alvarez said, they were able to find a likely culprit for these large leaks. “Ninety percent was coming from tanks—the vents and hatches,” he said. “These tank vents are designed to release pressure because otherwise they might burst. But why are they venting so frequently?”

Robert Jackson, who studies methane leaks at Stanford and was not an author on the study, said that the failure of the companies to report this kind of leak might help explain why the EPA has missed them in its emissions data.

“A company that finds such a leak might view it as an exception rather than as normal for their operations, so perhaps they don’t include that in what they report.” he said. “These large emissions are unusual but they’re real.”

Alvarez’s hope is that a combination of research and reporting from industry can help fix the problems and stop the leaks. 

The study’s authors said there is an “urgent need” for methodical surveys and measurements of these leaks, which must be followed by corrective measures