EPA methane measurement regulations explained

Last month, the US EPA released its final rule regarding methane emissions from upstream and midstream oil and gas facilities. The rule was groundbreaking, including rigorous requirements to measure and mitigate emissions, broad scope covering nearly all facilities, and novel enforcement procedures leveraging third-party measurements of large emission events. Here’s what this means for measurement tech.

Welcoming new tech into the fold

A major development involves the use of continuous monitors and other advanced technologies designed to detect methane leaks. While previous EPA rules included requirements for methane leak detection, the only approved leak detectors were handheld technologies—like optical gas imaging and Method 21 analyzers—that have long track records of success but are considered inefficient compared with more modern approaches.

So, operators who have been piloting new detection technologies have been using them in addition to the traditional methods, as required by the previous regulation. In an effort to encourage innovation and deployment of more efficient technologies, the new regulation now allows some advanced technologies to meet the regulatory requirement. Going forward, modern technologies can be used instead of, not in addition to, traditional handheld detectors.

Frequency vs. sensitivity: acknowledging a trade-off

However, before advanced technologies can be used, they have to pass performance tests. The effectiveness of a leak detection program depends on both the sensitivity of the detector (more effective programs use sensitive tools that can find smaller leaks) and the frequency with which inspections are performed (more effective programs run inspections more frequently to find leaks sooner).

To balance this trade-off between sensitivity and frequency, the EPA defined a matrix of requirements where compliance can be achieved either by using a more sensitive tool deployed less frequently or a less sensitive tool deployed more frequently. For example, a tool that can detect only emissions larger than 5 kg/h would have to be deployed 12 times per year, while a tool that can detect emissions as small as 1 kg/h would have the be deployed only 4 times per year. While this matrix was designed for periodic screening tools like airplanes, all advanced detectors—including continuous monitors—are eligible to be approved under these criteria.

Another way of assessing continuous monitors

In addition to the matrix, there is a second set of criteria that can be used to approve continuous monitors. The second set can be used instead of the matrix for continuous monitors, but it does not apply to airplanes or any other noncontinuous tools. This second evaluation is more detailed. It requires continuous monitors to make a measurement every 12 hours, to quantify the emission rate (the matrix requires only detecting, not quantifying, emissions), to meet certain requirements for system uptime and have the ability to measure emissions as small as 0.4 kg/hr.

That sensitivity requirement seems counterintuitive. Because of the trade-off between sensitivity and frequency, detectors with lower limits of detection (better sensitivities) normally can be deployed less frequently: in the matrix, the 1 kg/h detector needs to be deployed only 4 times per year while the 5 kg/h detector needs to be deployed 12 times per year. So why does a continuous monitor, which has an extremely high frequency (remember the requirement to measure at least every 12 hours), need to have a small limit of detection instead of a large limit of detection?

Can this be right?

The answer to that question lies in the way continuous monitors are to be used, if they are approved under this second evaluation method. Many oil and gas facilities contain equipment that is allowed to emit methane. These allowed sources might emit at small rates for short periods of time or from equipment where there is no ready technique to eliminate the emission. Examples of these sources include tanks that intermittently vent flashed gas, pneumatically controlled equipment that vents during operations, or even flares that emit imperfectly combusted gas.

The purpose of leak detection regulation is not to detect these allowed emissions sources. Instead, the purpose is to detect and repair unintended fugitive emissions that occur in addition to these allowed emissions. Unfortunately, many emissions detectors struggle to distinguish allowed sources from fugitive sources.

Separating the expected from the unexpected

Advanced emissions detectors can detect emissions, but they often cannot tell if the emission comes from an allowed source or a fugitive source—after detection, a separate investigation is required to understand the cause of the emission. These allowed emissions are often a nuisance for periodic screening tools. The screening detects the emission, triggering an investigative analysis to identify the cause of the emission, which then identifies that the emission is an allowed event that is not actionable. These nuisance alarms result in considerable time and effort spent investigating small emissions that cannot be addressed.

While the EPA allows continuous monitors to be used as high-frequency periodic screening tools, that path is impractical because of the nuisance alarms. The EPA included this second set of criteria—the one specifically for continuous monitors—to solve this problem. In this approach, a site is first cleared of fugitive emissions by standard leak detection and repair procedures, and then the continuous monitor is installed.

For the first month of operations, the continuous monitor quantifies the average rate of methane emissions from the site. Because any fugitive emissions have just been repaired, the emissions detected over this first month are deemed to be baseline allowed emissions. After the first month, if an emission is detected, rather than immediately proceeding with an investigation as is required for periodic screening tools, the continuous monitor simply records the emission rate. That measured emission rate is averaged for a period of time (both 7 days and 90 days) and an investigation is required only if the average emission rate, minus the baseline emission rate determined in the first month, is larger than a threshold.

Detection for action

While that procedure is complex, it has the advantage of eliminating the nuisance alarms that affect periodic screening tools. Small or temporary emission events, which require inspection whenever they are detected by periodic screening tools, will not need to be actioned if detected by a continuous monitor and found to be part of the baseline emission or too small when averaged over time. The only emissions that require a follow-up investigation, if detected by a continuous monitor, are ones that are large, persistent, and exceeding the site’s baseline emissions. The benefit of the continuous monitor designation in the EPA regulation is that operators will have to follow up only on the significant emissions that should be prioritized, and not on the nuisance, unactionable emissions that periodic screening tools sometimes detect.

The EPA allows continuous monitors to be approved under either of two pathways: as high frequency periodic screening tools described by the matrix, or as “true continuous monitors” that meet the second set of criteria (the set that applies only to continuous monitors). Seeking approval as a high frequency periodic screening tool may seem attractive because the rules are looser: the limit of detection requirement is not as strict, measurements do not need to be performed once every 12 hours, emission rate quantification is not necessary, etc. But in practice, nuisance alarms make that pathway unrealistic. For continuous monitors to be viable in the field, they need to be approved by the EPA as true continuous monitors.

Choosing continuous monitoring tech: regulatory cues

With many continuous monitors on the market today, some operators are wondering how to decide the one is right for them. My advice is to consider two things.

First, look for a continuous monitor that meets all the requirements to be approved as a true continuous monitor, not as a high-frequency periodic screening tool. These requirements include the ability to make a measurement every 12 hours, to quantify the emission rate, to have a limit of detection below 0.4 kg/h, and more. These specifications are provided by vendors and some of them are measured independently at third-party test sites.

Second, look for a continuous monitor that measures emission rate without bias. True continuous monitors must quantify emission rate, and leak repair is required if that average rate exceeds the baseline by a threshold. That threshold can be as small as 1.2 kg/h, averaged over 90 days. In a single measurement, no continuous monitor available today can guarantee an error bar smaller than 1.2 kg/h. However, the error bar on a single measurement is not relevant. The 90-day average rate is determined by averaging hundreds to thousands of individual measurements. Random errors—where the continuous monitor is equally likely to overestimate as to underestimate the emission rate—will largely cancel out over hundreds of measurements.

On the other hand, biased errors—where the continuous monitor is more likely to overestimate or underestimate the emission rate—will not cancel out and can prevent a continuous monitor from measuring with the accuracy required to identify fugitive emissions. Biased error can result from sensors that are improperly calibrated or from algorithms that do not properly model all the effects that dictate how methane molecules move in the atmosphere. Some random errors are acceptable, but true continuous monitors should have minimal biased errors.

Conclusion

In summary, the new EPA rules require substantially more methane measurement than previous regulations required. The EPA will approve the use of advanced technologies, if they meet certain performance requirements. Continuous monitors can be approved either as high frequency periodic screening tools or as true continuous monitors. While the screening tool approval offers less-demanding standards, that approach is impractical because it results in a large number of nuisance detections.

The true continuous monitor approval eliminates nuisance detection and ensures that only significant methane emission need to be actioned. Operators should look for continuous monitors that meet the limit of detection and other requirements necessary to be approved as true continuous monitors and that quantify emission rate without bias.

While the EPA requirements are complex, true continuous monitors with unbiased accuracy can satisfy the regulatory requirements while allowing operators to prioritize the most important emissions and avoid nuisance detections.

Find out more about our continuous monitoring technologies: methane point instrument and methane lidar camera.