How the long shadow of model inputs could dilute the ambition of the Biden Administration’s SAF Grand Challenge

Last Wednesday, leading U.S. airlines (including United, Delta, and American Airlines) co-signed a letter to the U.S. Treasury Department alongside major corn ethanol producers. Their letter seems, at first glance, like a minor technical suggestion. But that seemingly small suggestion could profoundly weaken the effectiveness of the Biden Administration’s efforts to curb greenhouse gas (GHG) emissions from the aviation sector.

Both the Biden Administration’s sustainable aviation fuel (SAF) Grand Challenge and the Inflation Reduction Act (IRA) notably defined an “SAF” as an alternative aviation fuel that generates at least 50% less GHG emissions across its entire life-cycle than its fossil jet counterpart. On its face, this is much more ambitious than the EPA’s GHG reduction threshold for fuels to qualify under the Renewable Fuel Standard (20%) or the International Civil Aviation Organization’s (ICAO) threshold for aviation fuels to qualify for the international CORSIA policy. However, high life-cycle thresholds are only as meaningful as the underlying methodology to calculate those life-cycle GHG emissions. If the methodology to estimate GHG savings is diluted or weakened, it challenges the credibility of that claimed 50% GHG savings.

At its core, the industry’s letter to the Administration’s is a request is for an easier path to 50% GHG savings. Their case is that the ambiguity within the IRA can be used to ensure that their life-cycle assessment (LCA) model of choice, the Department of Energy’s Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET), is used to calculate which fuels can exceed the 50% GHG savings threshold to qualify for valuable IRA tax credits. LCA models are of high interest to research organizations like the ICCT, but not usually to airlines. So why the interest now?

It all comes down to the substantial differences in emissions estimates that can be calculated for certain fuels between GREET and the IRA’s proposed methodology, CORSIA. Though these two approaches are largely similar, small differences in the inputs and assumptions in these models may make a big difference in the path to tax credits for fuel producers.

Granted, modeling results are only as good as the data and assumptions that go into them. GREET is a versatile and flexible analytical tool that allows users to input their own data and estimate project-specific emissions. For most fuels and for most parts of a fuel’s life-cycle, the expected results from using a GREET approach and a CORSIA approach would be nearly identical. The International Civil Aviation Organization (ICAO) even used GREET to develop its own default LCA emissions for aviation fuels for the CORSIA process.

But the primary issues with using GREET lie outside the model itself, and instead with the choices of which models and parameters feed into it. A recent ICCT fact sheet illustrates how policymakers’ choices of external indirect land-use change (ILUC) modeling and soil organic carbon (SOC) modeling alone can more than halve the emissions attributable to corn ethanol-to-jet. The large changes in emissions from these choices thus warrant further scrutiny—how are they calculated, and why are they so different from previous regulatory assessments?

However, the ILUC modeling cited in some configurations of the default GREET model estimates land-use emissions for corn ethanol and soy that are only 25% – 33% of the total emissions estimated through public regulatory assessments by the U.S. Environmental Protection Agency and the California Air Resources Board. Compared to those previous public regulatory assessments, the lower scores in some GREET configurations are often driven by optimistic modeling assumptions and soil carbon shifts that remain controversial and may not have passed muster in a transparent regulatory assessment. All else being equal, using these ILUC factors could mean substantially lower estimated emissions for a fuel without any process improvements by fuel producers.

Similarly, the soil organic carbon modeling in the CCLUB model referenced in GREET forecasts 30 years of consistent land management practices to calculate credits for fuels, making these credits similar to expected behavioral changes like those calculated as part of an offset program. However, the model on its own it lacks any safeguards to determine the land-use history of that land, the permanence of sequestration, or its additionality. Given the uncertainty associated with estimating soil organic carbon shifts, including these credits in the IRA goes far beyond the scope of existing carbon offset programs, and with even less scrutiny.

Many U.S. airlines have tried to burnish their credentials on climate, but regretfully this latest letter is consistent with a recent trend. Airlines have exerted considerable influence to dilute sustainability criteria for sustainable aviation fuels (SAFs) and exclude the industry from binding policies that would regulate its emissions.

To ensure that IRA money drives meaningful, additional GHG reductions rather than incentivizing business-as-usual fuels, the Administration still has substantial flexibility in interpreting the IRA and deciding how to model emissions. Using ILUC factors consistent with CORSIA or existing U.S. regulatory assessments, and excluding the use of SOC offsets would create a higher bar for what qualifies as a SAF, while retaining flexibility for fuel producers to demonstrate meaningful, measurable GHG reductions in their supply chain by using carbon capture and sequestration or renewable electricity. Researchers like me are usually glad when there’s LCA modeling up for discussion. But it’s important that those discussions empower policymakers to make the right choices in how those models are used and what assumptions feed into them.