❮ Return to Our Work

Op-Ed | The American Prospect | September 23, 2022

The Problem With Emission Reduction Models

ClimateExecutive Branch
The Problem With Emission Reduction Models

This article was originally published in The American Prospect.

President Biden and Vice President Harris celebrated the Inflation Reduction Act’s potential to confront the global climate crisis at an invite-only event on the South Lawn of the White House last week. But how much will this new law actually reduce greenhouse gas emissions?

There are several answers to that question, and not all of them are factually correct. Mainstream media sourceshave largely repeated one statistic: that the IRA will reduce U.S. emissions by 40 percent from 2005 levels by 2030. This number is agreed upon by a small handful of outspoken research shops, namely the Rhodium Group, Energy Innovation, and a Princeton University group known as the REPEAT Project. It’s an encouraging number, considering that President Biden’s own goal was to cut emissions in half by 2030. The IRA would get us about 80 percent of the way there, according to these figures.

The problem, unfortunately, is that all three of these modelers rely on methane-related data and metrics used by the U.S. Environmental Protection Agency, which have been found over and over again to be inaccurate. All three models therefore may well underestimate not only the United States’ historical emissions, but also emissions over the next seven years.

That’s particularly worrisome considering that lawmakers may soon be voting on a dirty pipeline “permitting reform” bill, which Sen. Joe Manchin (D-WV) claims his IRA vote was contingent upon passing. The bill aims to approve the fracked-gas Mountain Valley Pipeline, and mandate expedited federal review of at least five more fossil fuel infrastructure projects, which will undeniably increase emissions and compound the burden of toxic, carcinogenic oil and gas infrastructure on frontline communities. Will the IRA offset those effects? That’s the threshold question.

To understand why these mainstream models’ optimistic projections must not be taken as guarantee, we have to get a bit wonky.

First of all, the EPA’s greenhouse gas emissions inventory is not based on actual measurements of planet-warming gases released by fossil fuel infrastructure. It uses a “bottom-up” computer modeling approach that estimates the average yearly release at almost 8,000 fossil fuel facilities. This is already incomplete, as plenty of emissions sources are not included, like the estimated 3.2 million abandoned oil and gas wells across the U.S.

Many smaller polluters are also exempt from reporting emissions to the inventory. Earthworks notes that “thousands of small polluters add up to create a big problem—a large but unknown volume of pollution is released every year without being tracked.” And as a terrifying new investigation from Inside Climate News and NBC reminds us, oil and gas infrastructure releases other extremely potent, “immortal” greenhouse gases that are not regulated by the EPA at all.

In contrast to the EPA inventory’s method, researchers who take a “top-down” approach of using atmospheric measurements or satellite imaging to track methane releases say the EPA’s inventory “underestimates the amount of methane leaking from U.S. oil and gas operations by as much as half.” A 2018 study in Science found that oil and gas supply chain methane emissions in 2015 were approximately 60 percent higher than the EPA inventory’s estimate, and a 2022 study found through aerial surveys that methane emissions in New Mexico’s Permian Basin were six times greater than the EPA’s estimate.

The EPA model’s assumed leakage rate doesn’t take into account the realities of human error and abnormal operating conditions increasing gas leaks. Much of the belief that natural gas is a “greener” fossil fuel than coal is based on a predicted methane leakage rate across its life cycle that is far lower than the real rate of leakage.

The EPA’s model also relies on inputs provided by oil and gas companies. This system is easy to game. Bloomberg reported last month that “an unorthodox reading of a single word in the EPA regulations allowed [U.S.-based natural gas producer] Range to slash its reported emissions from energy production by 93 percent in 2020 compared with the approach used by most oil and gas companies.”

The EPA’s inventory also employs an outdated estimate of methane’s carbon dioxide equivalency, dating back to the IPCC’s 2007 report. The IPCC has updated its calculation of methane’s global warming potential twice since then.

Not only that: The EPA also chooses to consider methane’s global warming potential over a 100-year time frame, which dilutes our understanding of methane’s short-term impact, as the gas is far more potent in its first two decades in the atmosphere. In 2021, the IPCC estimated that methane had a 100-year global warming potential (GWP) 29.8 times that of CO2, and a 20-year global warming potential 82.5 times that of CO2. The EPA model continues to use the IPCC’s 2007 estimate of methane’s 100-year GWP: 25 times that of CO2.

All three major modelers employ the 2007 estimate of methane’s 100-year GWP, and rely on the same incomplete EPA inventory. We must consider the likelihood that some of these models’ consistency with each other stems from relying on the same flawed data inputs.

Energy Innovation actually addressed this issue directly in a late-August research note about their IRA modeling. “Our methane values are tied to the EPA’s inventory,” they acknowledge. “Many studies have shown the EPA’s inventory undercounts methane leakage and associated methane emissions … In this instance, we chose to use the same factors as the EPA to ensure comparability to the inventory, to other modeling studies, and to international convention around emissions reporting and tracking. We nevertheless acknowledge that using more updated values or assessing impacts using GWP20 values would change the estimated reductions of emissions in CO2e.” But Energy Innovation is the only modeler to publicly wrestle with the problem.

Separately, each of the main modelers assumes that carbon capture and storage (CCS) technology will succeed at scale. The Princeton analysis predicts a 13-fold increase in the use of the technology by 2030, and both Rhodium and Energy Innovation incorporate CCS as a major part of the potential emissions reductions.

The problem here is simple: Even the largest under-construction CCS facility is only predicted to remove 0.0001 percent of the CO2 emitted globally every year—and much of the planned carbon capture isn’t even intended to be sequestered (i.e., removed) at all, but to be used in further drilling. CCS technology serves to extend the life of the fossil fuel industry; industry documents released last week by the House Committee on Oversight and Reform showed oil and gas companies admit as much internally.

Inaccuracies in climate modeling are unavoidable, as the global climate system is hyper-complex, and under a level of strain our species has never before witnessed. But it is worrisome when these projections are treated as all but guaranteed, particularly when the narrative they offer is conciliatory, and undercuts the perceived need for further action.

Rhodium Group has been especially outspoken in recent weeks on NPRE&E News, and elsewhere, downplaying the impact of the law’s mandated oil and gas leasing as “negligible.” Rhodium has also produced reports for fossil fuel industry clients, including the American Petroleum Institute. Its top partners are no foe to the industry either: Trevor Houser advocated for the repeal of the crude oil export ban in 2015, worked to keep a fracking ban out of the 2016 Democratic Party platform as Hillary Clinton’s top energy adviser, and even wrote a book using funding from BP, Cheniere, Chevron, and Dow Chemical that came out in favor of natural gas expansion. John Larsen, another frequently quoted Rhodium partner, has been a longtimeadvocate for U.S. natural gas expansion. A Rhodium representative could not be reached for comment.

Jesse Jenkins’s ZERO Lab at Princeton, which modeled the IRA’s impacts, has not received fossil fuel funding. Jenkins has separately received funding from the Princeton Carbon Mitigation Initiative, which is funded by BP, and as a researcher for Princeton’s Net-Zero America study, which is funded by BP and ExxonMobil. He joined the Net-Zero America study after its funding was negotiated.

Going forward, the chronically understaffed, underfunded Environmental Protection Agency desperately needs to be held to a higher standard of oil and gas industry oversight, which is impossible without a generous increase in its resources. This could be solved expediently if congressional Democrats chose to make it a nonnegotiable sticking point in omnibus negotiations later this fall.

Section 136 of the Inflation Reduction Act also opens up $850 million in appropriations to the EPA, the majority of which will be distributed to owners and operators of polluting facilities to help them submit greenhouse gas reports. But some portion of that funding could be used by the agency itself to “cover all direct and indirect costs required to administer this section, prepare inventories, gather empirical data, and track emissions.” Without any public pressure over the EPA’s flawed emissions tracking system, it may not take this opportunity to improve its inventory.

Big-picture, the media must ask hard questions about the reliability of the models they use to substantiate claims about the future condition of our planet. There are literally billions of lives at stake in the next decade’s climate policies; we need rigorous reporting now more than ever.

Note: This article has been update to more accurately reflect the funding relationships regarding Jesse Jenkins, the ZERO Lab, and various initiatives at Princeton.

ClimateExecutive Branch

More articles by Hannah Story Brown More articles by Dorothy Slater

❮ Return to Our Work