Study Finds Natural Gas Leaks in Transmission Pipelines That Exceed EPA Estimates

KEY TAKEAWAYS

• The leakage of natural gas from distribution pipelines appears to be larger than what the EPA accounts for in its greenhouse gas inventories, which is a problem for climate.
• New technologies to measure methane continue to improve and make this type of work possible.
• Science doesn't always have to be done in a lab. Sometimes we get to go out in nature, go for a walk, and do science at the same time. 

Johns Hopkins researchers have found that methane leaks from natural gas transmission pipelines around Baltimore and Pittsburgh are more frequent and emit about four times more methane than current national estimates suggest. From their walking survey, they detected 23 leaks over about 12 miles, raising concerns that pipeline leaks may be a far more significant source of atmospheric methane than previously thought. 

These findings add to growing concerns about natural gas, which overtook coal as the country’s largest source of power generation nearly a decade ago. While it’s often labeled a “clean” energy source, natural gas is still a fossil fuel—and its primary ingredient is methane, a potent greenhouse gas.

“Methane by itself is not a direct health risk other than it causes climate change, which has downstream effects,” said team leader Peter DeCarlo, PhD, associate professor in Environmental Health and Engineering. “But methane is a potent greenhouse gas, so underestimating leaks means an important source of climate-warming emissions may be missed.”

DeCarlo and Ellis Robinson, an assistant research engineer, undertook the study in the winter and early spring of 2021 after the COVID pandemic sidelined another project. The idea for this work came from previous research done by DeCarlo and PhD students who measured methane leakage from natural gas infrastructure in Pennsylvania, where there is a lot of hydraulic fracturing or fracking. When driving around in a mobile lab looking for leaks, they noticed higher concentrations of methane whenever the road crossed an area with a buried pipeline. 

“We thought that these pipelines may be leaking more than people think they are. That is what took us off-road, so we could follow these pipelines for some distance and see what leaks we detected,” said DeCarlo, noting that pipelines typically do not run parallel to roads, but cross under them.

Armed with maps available through the National Pipeline Mapping Service Public Viewer, Robinson donned a GasScouter backpack that measures gas coming up through the soil and went on the hunt for “fugitive methane” leaking from large transmission lines. Leaks are identified by spotting sudden jumps in methane levels and using ethane measurements to confirm that the gas was from fossil sources. Overall, the survey covered about 30 miles of pipeline.

“One of the challenges was to identify where we are allowed to walk,” said DeCarlo. “Many pipeline rights of way are private land. State forests, state parks, and federal forests where natural gas pipelines pass through are public domain, so we are able to walk along those rights of way and make these measurements.”

In addition to determining the number of leaks, Robinson measured how quickly methane was escaping from five of these leaks. Each leak released about 172 grams of methane per hour, ranging from 17 to 452 grams—well above the EPA’s estimate of 44 grams per hour for similar pipelines, DeCarlo said.

Why the discrepancy?

The EPA estimates the number and volume of natural gas leaks based on ideal conditions—assuming that pipelines are maintained regularly, used properly, and operated according to best practices. That doesn't always represent reality, DeCarlo said.

“A great example is cars. Most cars get their emissions tested and are fine, but you always see that one car on the road that’s putting out white smoke. And that one car may be responsible for emissions equaling 100 cars. So, we're always worried about that kind of super emitter,” he said.

Ten years ago, it would not have been possible for us to both pinpoint the location of a leak and measure how much methane was emitted. With emerging technologies, we have the ability to do a better job.

Peter DeCarlo, PhD

Although this study covered a relatively short distance and is not necessarily representative of transmission lines across the Mid-Atlantic or the country—the United States has roughly 2.5 million miles of pipeline—the researchers say their findings raise questions about how the natural gas industry detects and reports leaks. Transmission pipelines are typically inspected by companies once a year, but the Hopkins team’s results suggest current inspection methods may miss or underreport leaks. Factors such as terrain, pipeline depth, and how emissions spread through soil make measuring leaks difficult. Different detection thresholds and equipment sensitivity can also significantly impact whether a leak is even noticed. Still, DeCarlo says it’s feasible for companies and the EPA to adopt more accurate measurements techniques.

“Ten years ago, it would not have been possible for us to both pinpoint the location of a leak and measure how much methane was emitted," said DeCarlo. “With emerging technologies, we have the ability to do a better job.”

The study was published in ACS Air.

Environmental Health and Engineering is a cross-divisional department spanning the Bloomberg School of Public Health and the Whiting School of Engineering. This hybrid department is uniquely designed to lead pioneering research and prepare the next generation of scholars to solve critical and complex issues at the interface of public health and engineering. Learn more about our programs.