Are gas flares methane’s forgotten super-emitters?

25 September 2026

The EU Methane Regulation requires operators to repair leaks detected at rates as low as 1 gram of methane per hour. Yet many continuously operating gas flares may emit hundreds – even thousands – of kilograms per hour. Could tackling methane from flaring be one of the biggest opportunities hiding in plain sight?

A Capterio Perspectives Insight prepared at and for the Methane Regulation and Emission Mitigation conference in Berlin 24-25 September, 2026, where our CEO Dr Mark Davis was a speaker.

Flaring is big. And not just for CO₂.

Global gas flaring reached approximately 167 billion cubic metres (BCM) in 2025, rising for the third consecutive year. It is an enormous waste of energy and a significant source of carbon dioxide. But it is also a potentially major source of methane because no flare operates with a 100% combustion efficiency.

A “best practice” flare may achieve 98–99% combustion efficiency (meaning “methane slip” is some 1-2%). But real-world combustion efficiency can be materially lower and highly variable, depending on flare design, operating conditions, wind and gas composition. The IEA’s global estimates imply an average combustion efficiency of around 93%, which we believe is plausible. At this level, flaring accounts for 13 BCM of emitted methane (out of a total of 101 BCM from the upstream). See our paper on country performance of flaring and methane.

Data from a flare in the North Sea illustrate this point – courtesy of Accord Energy Solutions, who have developed a new model to compute flare combustion efficiency. Per-minute data for three flares on a single North Sea platform show just how variable combustion efficiency can be. Over the period shown, the three flares averaged 95.9%, 98.5% and just 72.4% combustion efficiency, respectively.

Figure: real-time estimates of combustion efficiency for 3 flares on a single oil platform in the North Sea. Combustion efficiencies are material less than the “best practice” 99%. Chart courtesy of Accord.

At these efficiencies, unburned methane emissions are 1.5× to 28× higher than under a 99% assumption. Even after accounting for the fact that lower combustion produces less CO₂, total CO₂-equivalent emissions can be around 1.1× to 7× higher. The UK is an OECD country with strong regulatory oversight. Similar standards may not apply in other countries.

Flares as continuous super-emitters of methane

The industry has been focussing not only on fixing small leaks (potentially as low as 1 gramme per hour), but also those “super emitters” – those above 100 kg per hour. Most methane super emitters are transient (lasting perhaps for a few days per year). But large flares are the super-emitters in their own right. A very large flare of 50 million scf/day could plausibly be emitting 4 tonnes of methane per hour – that’s 4 million times the minimum limit prescribed by the EU. Interestingly, one of the largest gas fields that supplies gas to Europe routinely flares close to 50 million scf/day.

Consider the flare below. It averages approximately 30 million scf/day and, apart from two shutdown periods, has operated almost continuously for years. This is not a transient methane event detected during an occasional satellite overpass. It is a persistent source hiding in plain sight.

Figure: example of a large “super emitting” gas flare in a market that supplies gas to Europe.

Even at an optimistic 99% combustion efficiency, we estimate methane emissions of approximately 242 kg/hour, which is well into the “super emitter” category. At 95%, this rises to 1,263 kg/hour; at 90%, 2,667 kg/hour. Several major European energy companies hold non-operated interests in the asset below, making this particularly relevant to commitments to extend zero-routine-flaring ambitions across non-operated portfolios. Table 1 has some details.

Observed flare (million scf/day)	CE	kg/hour of CH4	thousand tonne/year of CH4	Total CO2-e million tonnes/year	Revenue $/year at $10 per mmbtu
30	99.0%	242	2.1	0.8	100
30	95.0%	1263	11.1	1.5	105
30	90.5%	2519	22.1	2.4	110
30	90.0%	2667	23.4	2.5	110
30	80.0%	6000	52.6	4.9	124

Table: examples of the impact of this flare as a function of combustion efficiency. Table shows the emissions of methane per hour and per year, the total emissions in CO2-equivalent terms (using a GWP of 82.5), and the revenue upside.

Interestingly, US regulation already recognises the danger of simply assuming high flare combustion efficiency. Under the US EPA’s Greenhouse Gas Reporting Program (40 CFR Part 98, Subpart W), operators that do not undertake the specified flare monitoring must use a default combustion efficiency of just 90.5% (and 92% for a destruction efficiency*) for emissions calculations. Higher assumed efficiencies are permitted where appropriate monitoring and operating criteria are demonstrated. The principle is important: high combustion efficiency should be evidenced, rather than simply assumed.

The EU Methane Regulation – albeit mostly about methane emissions – also has obligation on gas flaring. With reference to Article 27(1), Article 28(1), (2) and (3), and Article 29(1), Annex IX (6) states that importers shall provide information on “measures to control and restrict venting events and flaring events, including a description of those measures, together with, where available, relevant reports from LDAR surveys and from venting events and flaring events with respect to the last available calendar year”. Article 30 also states that each country that exports to the EU should state whether it has “particular restrictions on venting and flaring” and is part of global initiatives such as Zero Routine Flaring Initiative.

These aren’t small residual errors. They are methane sources comparable in rate to what would ordinarily be regarded as very large emission events. But unlike many intermittent super-emitters, this super-emitter operates every day – and has done, for years. Rates at these levels are regularly captured by public and private methane-tracking satellites, but mostly attributed to intermittent sources.

Why we need better flare inventories

There are good reasons to question whether conventional inventories (such as company reported data, or data submitted to organisations such as OGMP 2.0) fully capture methane from flaring. Analysis of company reports often finds (through back-calculation of company-reported data) that implied combustion efficiencies are as precise as 99.0%, year after year.

We think three reasons suggest that conventional inventories might be worth of a deeper look, because:

Firstly, flare rates are often underestimated. In general, most flares are not metered (though they are often for *operated* assets of leading companies). Reported volumes and emissions often therefore rely on assumptions based perhaps on historical “gas oil ratios” or “mass balance” calculation. Even flares that are metered are not always accurate (especially if flare rates are outside design ranges, or the flare meters are not best-in-class technologies). On many occasions our work with clients (which have subsequently been confirmed by ultrasonic metering) has been able to establish the true level of flares are 2-10x larger than initial company estimates.
Secondly – in our experience, not all companies have comprehensive inventories of their flares. Without third-party validation, corporate HQ that rely on self-reported data can be missing flares within both the operated and non-operated portfolios.
Thirdly, the commonly-used default assumption of 99% combustion efficiency is likely not to represent the real-world situation. Wind and gas composition are major drivers of lower combustion efficiencies.

This triple whammy creates a substantial risk that methane emissions from flaring are understated—potentially by an order of magnitude in some cases.

By implication, the “methane intensity” of the gas (or oil) is likely understated – and importantly – money is being left on the table. In the example above, $100 million per year is a material opportunity. It’s likely that the capital to fix this would deliver strong pre-tax returns.

Satellites can help – and Europe now has a reason to care

Infrared satellite observations can identify flares, estimate flared volumes and track operational changes over time. Capterio’s FlareIntel platform (www.flareintel.com) tracks every flare – for every asset, for every company, in every country – every day. Our data helps companies to improve visibility (even for metered flares, and especially for non-operated assets), to drive operational excellence (especially as flares are a useful diagnostic of asset performance) and to identify and prioritise attractive investment opportunities.

Methane-detection satellites can provide complementary evidence of plumes when conditions permit – although they typically have 2-4 orders of magnitude less frequent data (vs our platform, which integrates up to 6 detections per asset per day). The difference in observational density can be dramatic. In Libya, for example, 255 plumes are reported by IMEO in 2025, while FlareIntel recorded around 150,000 flare detections (i.e. between 2 and 3 orders of magnitude more data points). These are different measurements of different phenomena, but they illustrate why persistent thermal monitoring can provide a much richer time series of flare behaviour.

Yet, together with ground measurements and operational records, these observations can help identify discrepancies, prioritise investigation and independently assess progress. Satellites can therefore provide an increasingly important independent layer of evidence against which reported methane emissions can be tested, reconciled and verified.

This is especially timely for European energy imports. The EU Methane Regulation (EU 2024/1787) introduces producer-level monitoring, reporting and verification (MRV) equivalence requirements for relevant imported oil and gas under Article 28. It requires equivalence for monitoring, reporting and verification to Article 12 or OGMP 2.0 Level 5, with independent third-party verification meeting the relevant requirements of Articles 8 and 9 (although it is unclear – at least to the authors – whether the verification is around process or, (as we think is necessary) around data.

Two recent papers examine how to make this work. The Energy Emissions Modeling and Data Lab’s August 2026 protocol sets out a detailed framework for independently verifying producer-level methane reporting for oil and gas placed on the EU market. Independent satellite observations can form a valuable part of the wider evidence base against which reported emissions are assessed. The Oxford Institute for Energy Studies’ September 2026 paper examines the wider compliance architecture, including verification, accreditation and the chain of custody needed to link producer-level information with imported energy.

The methane challenge is not just about finding the next short-lived super-emitter. Some potentially very large methane sources are already visible, operating every day, and have been for years. Gas flares may be the methane super-emitters hiding in plain sight. Better data can help us see them – and act.

Notes and references: