The 98 Percent Nobody Measured

Most methane credit programs assume a flare destroys 98 percent of the gas it burns. In 2022, a research team flew aircraft over the largest oil and gas basins in the country and measured what flares actually do. The real average was 91 percent.

That seven point gap does not sound like much. It is the difference between a climate fix that works and a credit paid for combustion that never fully happened. The number on the certificate was an assumption. Almost nobody had checked it from the air until recently.

Ninety-Eight Percent Was Never a Field Result

The 98 percent figure has a history, and it is older than most of the policies that lean on it. It traces back to U.S. EPA flare research in the early 1980s, work that was mostly concerned with assisted flares and visible smoke, not with measuring how much methane survived combustion under real field conditions. Over time the number hardened into a convention. Two percent escapes, ninety-eight percent burns, and that became the default in emissions accounting and in many reporting rules.

The assumption is reasonable on paper. A well designed flare with enough energy in the gas stream can destroy nearly all of the methane it burns. EPA flare guidance puts combustion efficiency around 98 percent and higher and ties that performance to conditions like a minimum heating value in the combustion zone, near 270 Btu per cubic foot. Source: https://www.epa.gov/sites/default/files/2020-10/documents/13.5_industrial_flares.pdf

The problem is that a design target is not a measurement. A flare can be built to hit 98 percent and still fall short if it goes out, burns unsteadily in wind, or runs on gas that is too thin. The certificate says 98. The sky can say something else.

What the Aircraft Found

A 2022 study published in the journal Science put the assumption to the test. Researchers led by the University of Michigan sampled flares from the air across the Permian Basin, the Eagle Ford in Texas, and the Bakken in North Dakota and Montana, the three regions responsible for most U.S. flaring. They measured the methane left in the plume. Source: https://www.science.org/doi/10.1126/science.abq0385

The flares destroyed 91.1 percent of the methane, not 98. Two things drove the gap. Between three and five percent of flares were unlit, venting raw methane straight to the air while appearing to do their job. Others were lit but burning badly, with some operating closer to 60 percent efficiency. The study estimated this raised methane emissions from flaring roughly fivefold above what the standard assumption predicted.

The lesson is not that flares fail. It is that an unwatched flare is a guess. The same hardware can earn a passing grade or release a large volume of methane depending on whether it is lit, maintained, and burning hot enough, and none of that shows up in a spreadsheet that simply writes down 98 percent.

A Good Flare Still Reaches 99 Percent

This is where the farm story diverges from the oil field, and the difference matters. EPA landfill guidance states that properly operated flares can achieve methane destruction efficiencies greater than 99 percent. Source: https://www.epa.gov/lmop/continue-operate-gas-collection-system-and-flare

The gap between 91 and 99 is not luck. It is design, fuel quality, and supervision. Enclosed flares hold the flame inside an insulated chamber, which shields combustion from wind and holds temperature steady, the two failures that dragged the field flares down. Farm biogas also helps its own case. It is roughly 60 percent methane, which carries far more energy than the thin minimum that flare standards require, so the gas burns hot and clean when the equipment is working.

So the farm question is narrower and more answerable than the oil field question. The risk is not that manure gas cannot be destroyed. It is whether anyone confirmed the flare was running when the credit was issued.

Measure It, Do Not Assume It

Farm methane programs are set up to repeat the oil field's mistake if they credit destruction by assumption. A credit that pays out on a nameplate rating of 98 percent, with no requirement to show the flare was lit and burning, is paying for a number, not a result.

The fix is not exotic. It is a flow meter on the gas line, a methane concentration reading, an operating log that shows hours of runtime, and a downtime rule that stops the credit when the system stops. With those four records, a program is paying for measured destruction. Without them, it is paying for a brochure.

This standard protects farmers as much as the climate. A farmer running a flare correctly should not be lumped in with a site that lets one go dark for a month. Measurement is how the careful operator proves the work, and how the public learns the difference.

What Should Change

Every methane program that credits combustion should require destruction to be measured, not defaulted. That means regular monitoring of whether the flare is lit, what share of the gas is methane, and how long the unit ran. It means publishing those numbers so the people paying for the fix can see what they bought.

The 98 percent convention survived for forty years because almost no one checked it. When someone finally did, from the air, the real figure was seven points lower and the cause was ordinary neglect. Farm methane policy is young enough to skip that lesson. Credit the methane that is measured out of the air, fund the monitoring that proves it, and the gap between the promise and the result closes.

A flare is a good tool. It is only a climate fix when someone confirms it did the job.