What’s up with methane? [Gas In Transition]
Atmospheric levels of methane gas are rising at an alarming rate. In 2021, the methane concentration in the atmosphere exceeded 1,900 parts per billion (ppb) for the first time in human history, according to the records of the Global Monitoring Laboratory, Earth System Research Laboratories, National Oceanic and Atmospheric Administration (NOAA). That is triple the pre-industrial level of 700 ppb, and the increase has been accelerating over the past 13 years.
Local methane concentrations can vary widely depending on proximity to emission sources. The 1,900 ppb figure from the agencies that Hunziger cites is derived from thousands of direct measurements in air samples collected at high altitudes well away from local emission sources. It should be a reliable estimate for average concentration in the global atmosphere. Methane levels have been monitored for decades, and there’s little doubt that both the reported rise and the acceleration in the rate of rise are real. But what caused them, and what do they portend?
New measurement tools
Coincidentally, satellite instruments for indirect measurement of methane concentrations in the lower atmosphere have been developed. They’re not as accurate as direct measurements from air samples, but they cover the globe. They allow emission regions to be mapped out, and for point sources of anonymously high emissions to be flagged in near real time. If a major leak develops in a gas pipeline anywhere in the world, monitoring agencies can learn about it within 24 hours.
Also recently developed are mobile instruments that allow plumes from small leaks to be imaged and measured. The instruments can be carried by drones flying over old oil and gas fields to detect leakage from old wells that have not been properly plugged. They can also be mounted on ATVs that drive along pipeline routes. If a buried gas pipeline is leaking, methane seeping up through the soil above it will be detected.
Upon hearing about rising methane emissions, the first impulse of many is to assume that it must be due to oil and gas operations. Some of it is, but isotopic analysis of methane from air samples reveals that roughly two thirds of it is not. It’s of recent biogenic origin. (“Recent” meaning that samples still contain a measurable fraction of the radio-isotope 14C from when the biomass was formed.) Sources of biogenic methane include forest fires, rice paddies, wetlands, landfills, and sewage treatment plants. The biggest sources, however, are enteric fermentation of biomass in the digestive systems of termites and ruminant livestock.
Emissions from O&G operations
The fact that two thirds of the emissions are not from natural gas does not let oil and gas producers off the hook. It’s still important that they do whatever they reasonably can to cut fugitive methane emissions. Gram for gram while present in the atmosphere, methane is 120 times more potent as a greenhouse gas than CO2. Methane doesn’t persist in the atmosphere for much more than a decade on average, but even small amounts have a big effect on greenhouse warming while they’re present. So cutting fugitive emissions has a disproportionate positive impact.
There has been a lot of confusion in the popular press – and even in some academic literature – regarding the amount, the sources, and the implications of fugitive methane emissions. A recent study using the new tools for finding and measuring methane emissions found that methane emissions from oil and gas operations in New Mexico’s portion of the Permian Basin are three times higher than official EPA estimates. That finding has added fuel to a movement to ban natural gas connections to all new residential construction. There are valid reasons to favour electrical heat pumps over gas for residential space heating and hot water going forward, but the issues are not cut and dried. Misinterpretation of fugitive emissions data warps cost-benefit analyses and can lead to poor allocation of resources.
To put the cost-benefit analyses on a firmer footing and to provide guidance for what actions will deliver the most environmental benefits for resources invested, we need to go beyond sweeping net ratios between fugitive emissions and natural gas consumption. We need to understand the various types of emission sources and their relative impacts. Then we can consider appropriate options to reduce emissions from each.
The natural gas supply system
Figure 1 illustrates the general parts of a natural gas supply system. It’s from the pipeline & hazardous materials safety administration (PHMSA) of the US department of transportation. While each of the parts illustrated is a potential source of methane leakage, they differ greatly in terms of the magnitude and nature of risk they present.
The first and probably most important point to emphasise is that the greater part of methane emissions from oil and gas operations are not from any of the components represented in the figure. That is to say, they are not directly associated with the natural gas production and pipeline system at all. Rather, they are from wells not connected to that system.
Nearly all oil wells produce some amount of associated gas – especially if they’re tapping shale formations. But since oil typically delivers higher profit margins and since operating companies have been running tight on cash for a long time, it may not be worth it for them to lay gathering pipelines to deliver the associated gas into the gas supply system. In those cases, the gas should be flared. However, flares aren’t totally reliable. They can blow out, and even when functioning properly, are typically only 98% efficient. That means that at least 2% of the gas fed to them escapes into the atmosphere. In 2020, surveys sponsored by the Environmental Defense Fund (EDF) found that unlit or malfunctioning flares alone accounted for 300,000 metric tons of methane emissions from the Permian basin.
Fortunately, flaring intensity in US shale plays has fallen sharply from the highs reached in mid 2019. Those highs were an artefact of the rush to expand oil output from Permian and Bakken shale plays. Oil took priority, and associated gas was flared. According to data from Rystad Energy, flaring declined slowly but steadily from mid 2019 through the early months of 2020. Then from April to May of 2020 it fell sharply. The steady decline would have been from laying of gathering lines skipped in the rush to complete new oil wells. The sharp drop from April to May reflects well shut-ins and a hiatus on well completions following the onset of COVID-19.
From May of 2020 to Sept 2021 – the last month covered in Rystad’s press release – flaring intensity rose slightly, but then fell again. The rise likely reflected recovery of oil production as the pandemic waned and world economies recovered. But gas prices in Europe began climbing sharply in the second half of 2020. That gave oil producers in the Bakken and Permian basins incentive to continue laying gathering lines for associated gas from new and reopened oil wells. By Sept 2021, flaring had dropped below its May 2020 low. In the months since then – and especially since the imposition of sanctions against Russia – gas prices have exploded to unprecedented heights. That makes it a pretty sure bet that expansion of gathering pipeline networks has continued. Flaring of associated gas should now be approaching rock bottom.
The drop in flaring intensity and the associated drop in fugitive methane emissions are of course good news for climate change mitigation. However, it’s worth pointing out that the drop in fugitive emissions has been a consequence of higher, not lower demand for natural gas. That undercuts the case for fugitive methane emissions as a reason to push for lower consumption of natural gas. That’s not to say there aren’t other reasons, but so long as oil and gas production remain as “joined at the hip” as they are, pressure to cut gas consumption while demand for oil remains high risks increasing fugitive methane emissions.
Flaring – or sometimes outright venting – of associated gas are only one part of the emissions picture for upstream oil and gas operations. Another part is leakage from abandoned wells. But before we get to that part, let’s briefly jump to the other side of the natural gas supply system. Let’s look briefly at emissions from gas consumption by end users.
As the PHMSA figure earlier illustrates, consumption of natural gas downstream of the city gate is divided among small industrial, commercial, and residential users. (There’s also a small slice of the transport sector for vehicles powered by compressed natural gas, but it’s minor.)
In an effort to quantify fugitive emissions from end uses of natural gas, two studies bear looking at. One is Large Fugitive Methane Emissions From Urban Centers Along the US East Coast by Genevieve Plank et. al., published in Geophysical Research Letters. Dr. Plank’s team flew a NOAA research plane on multiple runs along the US eastern seaboard. It flew downwind of metropolitan areas from Washington, DC northward to Boston, MA. The instruments it carried measured not just methane concentrations, but also ethane, carbon monoxide, CO2, and water vapour. Ethane concentrations allowed the researchers to determine how much of the methane they measured was from leakage of natural gas vs. biogenic sources. The differences between CO2 concentrations in the plumes vs. known background levels allowed calculation of how much the plume had mixed with surrounding “virgin” air. That enabled gas concentration levels to be translated into overall emission rates. The bottom line for fugitive methane emissions from the sampled metropolitan areas was calculated as 0.8% of the volume of natural gas consumed.
The other study is much smaller and limited in scope. It’s Methane and NOx Emissions from Natural Gas Stoves, Cooktops, and Ovens in Residential Homes. In the study, a group of Stanford researchers measured leakage from gas cooking appliances in 53 California homes. Leakage from the appliances monitored was found to be 1.3% of the volume of gas consumed. Surprisingly, 80% of leakage was found to be from small leaks in the gas supply connections and control valves to the burners. That leakage was present whether or not the burner was on. The remaining 20% of leakage was from a combination of raw gas released after the burner was turned on but before it became lit, and gas from incomplete combustion. The burner flame also yielded traces of nitrous oxides (NOx) and carbon monoxide (CO).
It’s unlikely that the difference between the 1.3% leakage finding of the Stanford study and the 0.8% finding of the Plank study is due to any errors of measurement or methodology in either of the two studies. The error is in how we tend to think about fugitive gas emissions. We want to think of them as a percentage of gas consumed; i.e., if we consumed less gas then we would expect fugitive emissions to drop proportionately. But that’s not how it works.
Leakage depends mainly on the number and quality of imperfect valves and pipe connections; consumption depends on the average rate of gas flow through all those valves and pipes. So leakage as a percentage of consumption can vary widely. The 1.3% leakage that the Stanford researchers found was high, not because the leakage itself was particularly high, but because consumption was low. They only looked at consumption in cooking appliances. The greater part of consumption, in buildings with gas service, is for water and space heating.
Leakage and consumption aren’t totally independent, since one of the minor components of leakage is the small percentage of the gas fueling a burner that escapes combustion. In a well designed burner with a controlled ratio of fuel to air, however, that’s less than 0.1%. Moreover, it seems that even that small percentage can largely be eliminated in a manner that is reasonably economical and environmentally friendly.
If hydrogen is blended into natural gas at just 5% by volume, the volumetric heat content of the gas is reduced only slightly. The blended gas can be safely used in place of unblended natural gas without changes to the distribution system or to appliances burning the gas. But the H2 content increases flame propagation speed and raises combustion efficiency. That reduces emissions of unburned methane or incompletely burned methane (i.e., CO). If the burners are adjusted to provide a leaner fuel to air mixture – enabled by the presence of hydrogen – the cooler flame reduces NOx emissions. At the same time, greater oxygen availability further reduces emissions of CO and methane. In unvented cooking appliances, the result is a super clean flame that improves indoor air quality. That’s important in new well-insulated homes with low air exchange rates.
Let’s now move on to what is undoubtedly the thorniest problem regarding fugitive emissions: abandoned wells.
There’s an old war ballad from the time of the First World War. Its many improvised verses all end with the familiar refrain: “old soldiers never die, they just fade away”. Substitute “oil and gas wells” for “soldiers” and the refrain approaches literal truth. Old oil and gas wells never die, their production just fades until it’s no longer profitable to operate them. However, if abandoned, their bore holes will slowly fill with water and seeping oil, with methane and other gases bubbling up and into the atmosphere. Proper plugging requires filling the bore with a special sealing cement. It’s an expensive operation.
A lot has been written about the abandoned well problem. Good videos have been produced. A web search on “abandoned oil and gas wells” will return page upon page of relevant, high quality results. I won’t try to reprise all that here. Though methane leakage from abandoned wells is a significant problem that needs to be addressed, it’s primarily a legacy problem. It’s about remediating what we’ve done in the past. Cumulative emissions from the millions of abandoned wells that now exist are large enough to affect climate change, but they are not increasing. They’re slowly decreasing. Part of the decrease is natural, and part is from slow progress in identifying and plugging the highest emitters.
What matters, in terms of the future gas industry, is not the question of “how much methane is leaking from abandoned and unplugged oil and gas wells”, but a subtler one. We need to ask whether the pool of abandoned and unplugged wells is increasing. Do our efforts to maintain or increase natural gas production result in more emissions from a growing pool of abandoned wells?
There’s good news and bad news on that front. The bad news is that in the kind of shale plays now supplying the bulk of US gas production, decline rates for individual wells are ferocious. The “old soldiers” fade very quickly indeed. New drilling is constantly needed, and there’s a theoretical potential for the number of played out wells to expand rapidly. The good news is that a large and increasing fraction of new drilling is in laterals from existing boreholes. Or if not from existing boreholes, then from existing drilling pads hosting multiple wells.
The upshot is that the number of non-viable boreholes and pads is not rapidly increasing. More importantly, when further production from an existing pad does fade into non-viability, it’s likely that the well owners are still in business. So the boreholes get properly plugged and the pads decommissioned.
Misuse of emissions data
I pointed out earlier that the bulk of fugitive methane emissions commonly attributed to upstream oil and gas operations came from wells not connected to the natural gas supply and distribution system. Reducing consumption of natural gas will not reduce emissions from those sources. In fact, if natural gas prices were to fall due to reduced consumption, it would be more likely to increase fugitive methane emissions. There would be less incentive for the owners of new oil wells to expend the capital needed to deliver their associated gas to the natural gas supply system. It would also reduce the incentive for supply system operators to be vigilant in finding and fixing leakage from their gas treatment plants and pipeline systems.
Despite that, there appears to be a strong movement to demonise natural gas and discredit its use as an economical and comparatively low-carbon backup for renewable energy. Worse, the fugitive methane issue has been used to argue that blue hydrogen, made from natural gas with CCS, far from being a carbon-free resource for clean power generation, has a CO2 equivalency footprint that makes it “no better than coal”. That’s the assertion of a recent high profile article by professors Robert Howarth and Mark Z Jacobson attacking blue hydrogen.
The gist of that argument boils down to a patently illogical chain of reasoning:
● The net total of fugitive methane emissions is X;
● The net total of natural gas consumption is Y;
● Therefore, an application whose consumption of natural gas is Z causes fugitive methane emissions equal to Z × (X / Y).
There’s an implicit assumption of causality there that the authors don’t even feel the need to mention, much less justify.
Conclusion: a plea
I’d be the last person to minimise the danger of increasing methane emissions. I find the situation very alarming. I’d also be the last person to play down the potential for reducing emissions from leaky gas appliances and abandoned oil and gas wells. But I abhor the knee-jerk tendency of many in the RE advocacy sector to automatically blame rising methane emissions on the gas industry. That’s just not where the evidence points.
So let’s try to understand where those accelerating emissions actually are coming from, and figure out what we can do about it. More light, less heat. Please?
About the author
Roger Arnold is a former physicist, aerospace engineer, and systems architect. He has worked at IBM, Boeing Aerospace, AT&T, and a number of electronics companies and startups in San Diego and Silicon Valley. Now retired, he pursues independent research and writing. His writings focus on climate, clean energy, and sustainability. He is especially interested in exploring how technologies interact and create opportunities for synergistic solutions to critical problems.