Natural gas leaks are a significant source of methane emissions in cities, but it can be hard to pinpoint exactly where and why they happen. And methane, a potent greenhouse gas that is the main chemical component of natural gas, has been historically undercounted or omitted in emissions estimates, according to a team of atmospheric and climate science researchers in the Penn State College of Earth and Mineral Sciences.
To help address this, the team analyzed methane emissions readings from seven measuring towers in and around Baltimore and compared them to aggregated 2022 neighborhood gas usage data provided by a local gas distribution company. They found a daily cycle of methane emissions in urban and suburban neighborhoods, while urban sites also showed a seasonal trend in emissions. To explain this, the researchers found evidence that Baltimore's daily emissions trends could be correlated with consumer use of natural gas, while seasonal leaks could be caused by legacy cast iron pipes and urban repair schedules. They published their results in Environmental Science & Technology.
"This study shows progress toward a process-based understanding of methane leakage from urban natural gas distribution systems," said Ken Davis, professor of atmospheric and climate science and co-author of the paper. "Our method also provides a conceptual foundation for inferring patterns in local-scale methane emissions in Baltimore, which could be applied to existing atmospheric methane measurements in other cities."
When measuring methane emissions using atmospheric observations, emissions are usually estimated for the whole city but not parsed out spatially by neighborhoods, explained corresponding author Helen Kenion, a postdoctoral researcher at the University of Cincinnati who led this study as part of her doctoral dissertation in meteorology and atmospheric science from Penn State, completed in 2025 under Davis.
"By analyzing natural gas usage data in combination with these localized, high resolution emissions monitoring data, we were able to make some interesting connections and find new clues about where methane emissions come from," Kenion said.
Seven measuring stations, which were then overseen by the U.S. Department of Commerce's National Institute of Standards and Technology, took a reading of the amount of methane in the air 328 feet above the ground every two seconds. Kenion then used a method she developed in 2024 to estimate methane emissions from the neighborhoods surrounding the towers based on the readings.
At the two measuring stations where both methane measurements and gas usage data were available, researchers found that daily gas usage rates showed a daily cycle that was very similar to daily methane emissions rates, and that both usage rates and emissions were lowest at night and highest during the day.
Since gas lines in cities and neighborhoods are pressurized 24 hours a day, the daily usage and emissions cycle could not be attributed to gas pipeline leaks due to their clear pattern of daily fluctuations, Kenion explained.
"Since these daily fluctuations were not tied to seasonal trends, but rather a daily cycle present in both summer and winter, it's unlikely that postmeter leaks - leaks that occur after the pipeline has offloaded gas into smaller consumer lines and gas usage readings are noted - are entirely due to leaking heat furnaces," Kenion said. "It would have to be leaks originating by something used year-round, like water heaters or kitchen appliances."
Though the gas companies routinely monitor leaks in the main pipelines, the small lines into homes and buildings and hookups to appliances are not checked frequently and could leak small amounts when appliances are in use. Some small leaks could be attributed to lighting a gas stove, for example, where some gas must vent into the air, researchers said.
In addition to daily fluctuations, urban neighborhoods in Baltimore - accounting for two of the seven monitoring sites - saw a seasonal fluctuation in methane emissions. The emissions peaked in February, then gradually improved until reaching a low in August. When compared to the seasonal repair schedule provided by the local gas distribution company, the researchers found a clear correlation.
Since the suburban and rural locations had a flat seasonal pattern, the researchers hypothesized that it was due to the materials used in the gas pipelines: urban Baltimore neighborhoods use inflexible, legacy cast iron pipes. When the ground freezes and thaws, the pipes can shift and separate, causing leaks.
Suburban and rural neighborhoods surrounding Baltimore use newer PVC pipes, which can bend without breaking during freeze and thaw cycles.
"Seeing that emissions are higher in areas where there are cast iron pipes, the difference in material makes a difference," Kenion said. "It takes time for the gas company to identify leak locations and repair them. Workers continue to repair pipes even when the weather warms up, which explains why leaks and methane emissions hit a low in August."
In addition to Davis and Kenion, Natasha Miles, a former Penn State research professor of meteorology and atmospheric science, was a co-author of the paper.
This work was funded by the former Greenhouse Gas Measurement Program of the National Institute of Standards and Technology of the U.S. Department of Commerce under awards 70NANB19H128 and 70NANB23H188. This content is solely the responsibility of the authors and does not necessarily reflect the views of the funder.
