Catalyst N° 033 of 125 29 Aug 2024
Why are we still flaring gas?
with Tomás de Oliveira Bredariol, energy and environmental policy analyst focused on methane, International Energy Agency (IEA)
In this note
The question
Natural gas is one of the most heavily traded commodities on earth, so why do oil producers still burn about 150 billion cubic meters of it a year at the wellhead instead of selling it?
The answer
Because for the operator it is a capital allocation decision more than a technical one. Mostly there is no easy way to get the gas to a market, and where a fix exists it has to compete for money against drilling another well. Bredariol’s view is that several of the fixes do pay, just not well enough to win that competition, which is why he ranks regulation as the most important lever, ahead of voluntary industry action, with infrastructure, financing and simple awareness of the available options alongside it.
03The argument
Start with why flaring is a defensible practice at all, because the episode only makes sense once that is clear. Natural gas is roughly 90% methane, and methane is about 80 times more damaging to the climate than carbon dioxide over 20 years and about 30 times over 100. Burning it converts the methane to carbon dioxide and cuts that impact sharply. The alternative to flaring is not capturing the gas, it is venting it, which is worse for the climate and also an explosion and health hazard for workers and nearby communities. So flaring is the better of two bad options, and that is why it survives as an accepted industry practice rather than being treated as an accident.
The scale is what makes it indefensible anyway. Roughly 150 billion cubic meters were flared in the most recent year of World Bank data, more than Norway produced, and Norway is the eighth-largest gas producer in the world. The line has been essentially flat since 2010, through the World Bank’s Zero Routine Flaring initiative, the Global Methane Pledge and the Oil and Gas Decarbonization Charter, all of which point at zero routine flaring by 2030. Flaring intensity, meaning gas flared per unit of oil produced, has improved, but only marginally. And the emissions figure of roughly 500 million tons of carbon dioxide equivalent is not purely carbon dioxide, because no flare achieves complete combustion. The IEA puts global flare combustion efficiency near 92%, well below the 98% the industry tends to cite, because flares operate in high winds, go unlit after upsets, and are often poorly monitored. A meaningful slice of flaring emissions is therefore raw methane, which is precisely the thing flaring exists to prevent.
Kann spends the middle of the episode running through the obvious fixes and expecting one of them to work, and the interesting part is that Bredariol agrees the economics are often favorable and still explains why nothing happens. Pipelines are the leading answer, and most of the 8,000-plus flaring sites worldwide could probably connect to one, often by adding 10 or 20 kilometers of line or striking a deal with whoever owns the existing trunk. But Permian wells co-produce small volumes of gas that decline fast, so the operator is being asked to build durable infrastructure for a short-lived stream. Using the gas on site to displace diesel generation should be free money, except the gas often needs cleaning first to strip carbon dioxide, hydrogen sulfide and liquids. Reinjection is better than it sounds because it can lift oil recovery later, but only in reservoirs whose pressure and maturity allow it. Small-scale compressed or liquefied natural gas needs roads or waterways, gas-to-liquids and small-scale methanol cost far more, and going offshore takes nearly all of it off the table. The honest answer Bredariol gives is that operators are generally aware flaring is harmful and simply have better places to put capital: using the gas may be economical, but not to that extent. In some cases, particularly the long tail of small flares, it is genuinely not economical at all.
That reframing changes what the United States’ presence on the list of worst offenders means. Kann’s instinct is that flaring concentrates where regulation is weak, and the nine countries that dominate it do include Russia, Iran, Iraq, Venezuela, Algeria, Libya, Nigeria and Mexico. But the United States is on that list too, and Bredariol separates volume from intensity to explain it: on intensity the US is a relatively good performer, though not among the best, and it appears on the list because it is the largest oil and gas producer. His second correction is that flaring is not reliably a small-operator, low-infrastructure phenomenon. US flaring rose about 20% in a year driven mostly by the Permian, a mature basin with plenty of infrastructure, where operations are nonetheless dispersed enough that an individual company without relationships with its neighbors may have no practical route to a pipeline. Which is what makes his closing argument about regulation follow rather than being a reflex. Norway has prohibited routine flaring since about 1971 and its industry functions normally, and Nigeria cut flaring from about 20 billion cubic meters a year in 2000 to about six through a combination of regulation, fiscal incentives, infrastructure investment and LNG export capacity. The shove works where it has been tried.
04What you need to know first
- Associated gas
- Natural gas that comes up alongside oil from a well drilled for the oil. Nobody set out to produce it, which is the root of the whole problem.
- Routine versus non-routine flaring
- Routine flaring is the continuous burn-off of associated gas with nowhere to go, which Bredariol puts at roughly two-thirds to 70% of volumes. Non-routine flaring handles upsets, shutdowns and maintenance, and is a management and monitoring problem rather than an infrastructure one.
- Flaring intensity
- Gas flared per unit of oil produced, as opposed to total volume flared. The two rank countries very differently, and the gap between them is the whole explanation for why the US sits on a list it otherwise looks out of place on.
- Combustion efficiency
- The share of methane a flare actually burns. Even optimally run flares reach roughly 99.8% rather than 100%, and the global average is far lower, so flaring emits methane as well as carbon dioxide.
05Details worth keeping
- The venting alternative is what makes flaring rational at the well. It is also why “stop flaring” is not by itself a climate instruction; the useful instruction is “capture it instead.”
- The non-routine share has its own solution class, distinct from pipelines and conversion: better monitoring of production conditions, better planning, and portable equipment brought on site for short interventions.
- A newer category of technology does continuous monitoring of combustion efficiency itself, tuning flows, speeds and temperatures to cut the unburned fraction. That lowers the emissions from flaring without lowering flaring.
- Kann raises using stranded gas to power data centers, and notes that in practice this mostly means bitcoin mines, which can be sited anywhere and are extremely sensitive to power cost. Bredariol confirms it is seeing some use in the United States and says he is unsure how applicable it is elsewhere.
- The cost ladder among conversion options runs from small-scale compressed natural gas, to small-scale liquefied natural gas, to gas-to-liquids, which is why there are few small methanol or gas-to-liquids plants. Compressed gas suits low volumes and local vehicle fleets; liquefied suits larger volumes and longer hauls.
- Operator size matters through capital access and obligations rather than competence. Larger operators reach capital and infrastructure more easily and carry clearer targets, reporting and financing conditions; small independents often have none of that.
- By volume, most flared gas comes from mid-sized flares, with a large population of small flares where the economics get hardest.
06Claims worth citing
All figures as stated on 2024-08-29 and attributed to the speaker. Annual flaring volumes and country totals get revised every year, so treat every number here as a snapshot.
- About 150 billion cubic meters of gas flared in the most recent year, from World Bank data. He later restates the same figure as 148 billion cubic meters, so the precise number is loose. World Bank, cited by Bredariol
- Norway’s gas production is about 120 billion cubic meters, making it the eighth-largest producer, so more gas was flared globally than Norway produced. Bredariol
- Flared volumes have not meaningfully declined since 2010; flaring intensity has improved marginally. Bredariol
- Roughly 500 million tons of carbon dioxide equivalent, using a 100-year global warming potential of 30 for methane and counting both the methane and the carbon dioxide. More than all international flights in a year, and a little below the carbon dioxide emissions of Korea’s energy sector. Bredariol
- That works out to roughly 1% of global emissions against a rough 50 gigaton annual total. The arithmetic is Kann’s own and the guest neither confirms nor disputes it. Kann
- Natural gas is about 90% methane; methane is roughly 80 times more damaging than carbon dioxide over 20 years and 30 times over 100 years. Bredariol
- Global flare combustion efficiency around 92%, against an industry figure usually given as 98% and an optimal-condition ceiling near 99.8%. The 92% estimate is the IEA’s and is described as consistent with the scientific literature. IEA, cited by Bredariol
- Nine countries account for about 75% of flared volumes and about 45% of global oil production: Russia, Iran, Iraq, the United States, Venezuela, Algeria, Libya, Nigeria and Mexico. Bredariol
- US flaring rose about 20% in the most recent year, with most of the increase from the Permian Basin. Bredariol
- More than 8,000 flaring sites worldwide; the majority could probably be connected to a pipeline, many needing only 10 to 20 kilometers of new line. Bredariol
- Roughly two-thirds to 70% of flared volumes are continuous routine flares and the rest are periodic. Bredariol
- Nigeria cut flaring from about 20 billion cubic meters a year in 2000 to about six, a reduction he estimates at roughly 70%. Bredariol
- Norway has banned routine flaring since roughly 1971, a date he flags he may be misremembering. Bredariol
07Where it’s contested
- Combustion efficiency is a live factual dispute. The IEA’s 92% and the industry’s customary 98% are far apart, and the entire gap is methane. This is the one place in the episode where the guest names an opposing position and rejects it.
- The host’s premise and the guest’s answer do not fully match. Kann opens by calling flaring the weirdest market inefficiency he knows and returns repeatedly to the idea that operators are leaving money on the table. Bredariol agrees they need a shove but keeps resisting the framing: operators are largely aware, they have competing investment opportunities, and in some cases the economics genuinely do not work. The divergence worth holding is between “irrational” and “rationally deprioritized.”
- Kann’s structural guess gets corrected. He assumes flaring concentrates among small operators in low-infrastructure regions; Bredariol says that is “not perfectly correlated,” and points to the Permian as the counterexample.
- Nothing is claimed for the hardest geographies. Kann asks explicitly about Iran and Russia, and neither speaker offers a route to reducing flaring there. The regulatory optimism is scoped to countries where regulators can act.
- The guest hedges many of his own numbers. The routine share is “maybe two-thirds or 70%,” the Norwegian ban is dated “if I remember correctly,” the pipeline finding is that most sites “could probably” connect, and he is explicitly unsure whether flare gas powering data centers travels outside the United States.