Methane, the major component of natural gas, punches way above its weight when it comes to damaging the climate. Humans send less than half a billion tons of methane into the air every year, only one percent of the amount of carbon dioxide we spew from our cars, homes and factories. Yet methane is responsible for about 20 percent of global warming. This fact should motivate humanity to reduce our reliance on natural gas as a fuel as quickly as possible and, in the meantime, plug even small leaks in the existing natural gas system from wells to gas-burning appliances, says Marc Fischer, a scientist at Lawrence Berkeley National Laboratory.

Researchers in Boston have been studying methane leakage in urban areas for eight years and the pandemic gave them a unique natural experiment to compare methane levels during the lockdown with those from normal activity. Publication of their findings this month give ammunition to scientists who have long contended that much more gas is leaking from city sources than has been widely believed. The findings suggest also that far more natural gas is seeping out from pipes and appliances in buildings than previously thought, and that slowing global warming requires fixing these leaks as soon as possible or, possibly, phasing the fuel out faster than policy makers are planning. 

One strategy, said Maryann Sargent, a climate scientist at Harvard University and lead author of the new paper, might be to stop connecting new buildings to gas lines. “If you stop consuming, you stop emitting,” she said. “You could have more of an impact than you might think in reducing methane emissions now.”

Indoor stoves, heaters, ovens and gas-powered appliances 

For thousands of years methane generated by human endeavors has wafted into the air. Rice paddies and cattle produce it. Farming obviously still emits methane today. But the largest source now is production and distribution of natural gas (which is 90% methane) and, to a lesser extent, coal mining and oil. Natural gas usage is growing fast both in the U.S. and abroad, adding urgency to efforts to prevent leakage. Climate researchers have warned about methane leaks from gas wells, pipelines, and other fossil fuel production and transportation infrastructure. But methane emissions from urban sources in buildings – for instance, stoves, ovens, heaters, and other gas-fired appliances inside homes and business – have not been on most experts’ radar.  

Sargent’s latest research could change that. In 2020, soon after the first reported Covid-19 deaths in the United States, Boston University sent home its 34,000 students and the vast majority of its 10,000 staff members. Along with BU’s several hundred buildings, BU’s Tsai Performance Center, near the center of the urban campus, locked its doors, but an electric pump kept drawing exterior air into the Tsai Center through a thin hose that snaked from the roof to a device called a gas concentration analyzer measuring the amount of methane in Boston’s air. The setup was part of Sargent’s study of how much natural gas escapes into the air around the city from appliances and the fuel’s labyrinthine supply network of indoor plumbing and underground pipelines. 

Sargent’s research team included colleagues from BU, the National Oceanic and Atmospheric Administration (NOAA), the Environmental Defense Fund, and other Harvard scientists. They analyzed methane measurements taken at BU and three other Boston-area sites. Sargent said the research record, much longer than that of any similar study, gave the team a unique view of emissions trends. “It kind of tells a story,” she says, offering researchers new insights into the location of leaks in Boston.

COVID pandemic opened way for new analysis, understanding

Before the pandemic year, the long-term experiment was already on course to support a dawning realization that cities are an important source of natural gas releases, and to give some clues about where leaks occur. But the unprecedented global health crisis supplied an additional vein of data they hadn’t expected. The Covid quarantines enabled a clever natural experiment that probed how much gas escapes inside buildings such as those around the BU campus. When the pandemic closed the university, less natural gas leaked into the air. That result would not be expected if emissions leakage of methane inside buildings were negligible, as regulators have generally assumed.

The study could suggest new ideas for reducing leakage and avenues for further investigation, and it shows that many familiar objects around us are more threatening to the planet than has been imagined.

“That’s a marvelous result,” said Fischer, the Lawrence Berkeley National Laboratory scientist, when he heard about the Covid findings. Fisher has spent more than a decade researching natural gas leakage.

He gave “three cheers” for the additional evidence that a large share of the unwanted, or “fugitive” emissions in Boston – and, by extension, in other cities – arise not only from leaky pipelines under streets, the usual suspect, but also from a combination of utility distribution equipment previously considered blameless and sources inside buildings. 

Science at work: How the study was conducted

One day back in 2012, Fischer explains, he propped open the oak Arts and Crafts front door of his house, a couple of miles from downtown Berkeley. With help from Toby Haass Walpert, an undergraduate student at the University of California, Berkeley, Fischer fitted a heavy-duty nylon panel known as a blower door into the frame, completely blocking the passage. 

The scientist and student switched on a fan secured to the cloth barrier like a window-mounted air conditioner. It blew air out of the house, creating negative pressure and preventing gasses inside the house from seeping out from anywhere other than through the blower door. For the next hour the scientist and student measured methane in the fan’s slipstream. They turned the kitchen stove on and off and, for calibration purposes, released a standardized puff of methane from a tank. 

Fischer and Walpert’s investigations were the first serious attempt to determine how much natural gas leaks out of residential homes. The results surprised them.

Most of the effort to stanch natural gas leakage in the United States focuses on wells, processing facilities, and the 300,000 miles of high-capacity pipeline. According to one well-regarded assessment of methane emissions from the U.S. oil and gas supply chain, 12 million tons of natural gas, or about 2.5 percent of production, escape each year from this infrastructure before arriving in urban areas.

Utility interests and regulators long have insisted that, compared to releases in the production and long-distance transport of natural gas, urban leakage is small. Their argument relies on the inventory method for estimating leaks, a procedure only an accountant could love. Leakage from each component in a city’s gas distribution network, including every mile of pipeline and every valve and gas meter, is estimated and tallied up. Regulators usually don’t even bother to include emissions occurring inside residences and commercial buildings, which they have considered very low.

A calculation of leaks using the inventory method with data published by the Massachusetts Department of Environmental Protection determines that 0.4 percent of the natural gas supplied to the state’s urban distribution network escapes: It’s a small figure compared to losses in long-distance production and transport of natural gas. But Sargent says that her research shows that the inventory method low-balls fugitive urban emissions. It also suggests that a large share of the uncounted urban emissions might emerge from inside buildings. To come to that conclusion, she and her colleagues used a distinctly different technique. 

For more than a decade, some researchers have studied urban leaks by making direct measurements of the amount of methane in urban air. Their results have raised doubts about inventory studies and have caused concern that large leaks in urban areas have been overlooked. For instance, one paper that made direct air measurements found that 1 percent to 2 percent of the natural gas entering the Baltimore-Washington, D.C., region escapes into the air: That’s about 10 times as much as the inventory method suggests. 

Sargent says calculating emissions by measuring methane in ambient air is complicated. She and her colleagues set up a network of ambient detectors, built a computer model of the air flow around Boston, and collected a huge database of known methane sources, including the natural (such as marshes) and human-made (such as landfills). 

The investigators established background levels of methane – levels unaffected by local gas leaks – with tower-mounted detectors set up at two rural sites just outside the metropolis. Using fine-grained meteorological records, the scientists modeled step-by-step paths of hypothetical air parcels transiting from the region’s outskirts to the detector at BU, and another on top of a downtown skyscraper. At increments in the simulated trips, the researchers boosted the methane in the made-up parcels by the amount that would be gained from the actual gas sources at those geographic locations. When simulated portions of air hypothetically arrived at a detector, they should have contained as much methane as actual gas measured there, assuming all methane sources had been properly inventoried. 

Efforts to reduce leaks under streets ‘didn’t make difference in atmosphere’ … Why? 

But the simulated methane figures didn’t match up with actual measurements. There must be unknown sources somewhere. To make the numbers work, the scientists determined that 2.5% of natural gas delivered to Boston must escape into the air, six times the amount suggested by figures used by Massachusetts regulators.

“We know this is coming from the natural gas system,” Sargent says. Yet “we don’t know what part of the system” is leaking. However, in addition to the results from the Covid shutdown, the researchers’ paper does provide some clues.

Over the eight years examined, Boston’s annual average emissions remained level, despite substantial efforts by utilities to plug distribution leaks under streets. Between 2012 and 2019, utilities replaced 1,000 miles of antiquated cast and wrought iron mains with modern plastic pipe. “And guess what,” says Sargent, “it didn’t make any difference in the atmosphere. We can’t even see the impact of it.” This conclusion suggests that leaky pipes and valves buried underground might be only a part, perhaps a small part, of the problem, she says. 

Seasonal trends tell the same story. The authors of the new research found that leakage increases in winter, when natural gas usage, for heating buildings, is highest, and declines in summer. “That’s kind of a sign that maybe we shouldn’t actually be looking at the pipeline system in terms of the biggest sources of natural gas losses,” says Sargent. Distribution mains are pressurized year-round, she explains. If distribution pipes were the only big source of fugitive emissions, leakage would not change much between seasons.

The Boston research points to the probability of considerable leakage from residential and commercial buildings, though the researchers say there are probably also uncharted leaks in the urban distribution network. 

Evidence leads to greater investigation

Fischer says he has been expecting results like this for years. When he studied his house in 2012, he noted that nearly 1 percent of the natural gas supplied to it escaped. “Whoa, we should look at some other houses,” he recalls thinking. A few years later, he surveyed natural gas escaping from a sampling of 75 California houses. 

He and colleagues at Lawrence Berkeley reported in a 2018 paper that on average, California homes lose 0.5 percent of their natural gas, a figure far higher than generally assumed at that time. That’s one-sixth of California’s leaked natural gas, or 35,000 tons of methane.

Natalie Pekney, an environmental engineer at the National Energy Technology Laboratory, was intrigued when she read these findings. “Maybe this does merit further investigation,” she thought. “Maybe it could be something that’s significant.” She and a coauthor later published their own paper highlighting evidence that sizable amounts of natural gas escapes from houses.

Fischer’s own house leaked more natural gas than almost any he ever found: Not enough to be a health or safety risk, but not so good for the planet. He blames the high losses in his house on an outdated stove with inefficient pilot lights and a tankless water heater that expelled a lot of gas before the flame takes when turned on. He has since shut off his pilot lights.

Not as much attention has been given yet to commercial establishments, such as restaurants, but Sargent says she thinks they’re also part of the problem. The pandemic decline in methane leaks around BU, a neighborhood crammed with offices, restaurants, and laboratories, suggests that leakages come from operation of appliances. If the worst offenders could be identified, perhaps they could be repaired or redesigned.  

Or, maybe we have to “just move away from natural gas altogether,” Sargent says. That’s what Fischer is planning. He’s installing solar panels that, along with electricity sourced from non-fossil fuel generation, will halt reliance on natural gas.

Daniel Grossman

Daniel Grossman, Ph.D., is an award-winning freelance print journalist and radio and web producer with more than 20 years of experience. He earned his B.S. in physics and his Ph.D. in political science,...