Assessment of the scale of potential indirect emissions due to higher oil use
Petroleum life-cycle emissions—did we forget something?
Some biofuels proponents have recently been at odds with regulators over the extent to which some fuels policies could unfairly burden biofuels against fossil fuels. The question at the heart of the debate is that if regulations factor indirect land-use change (ILUC) into estimates of the carbon intensity of biofuels, should we not also estimate indirect carbon emissions from petroleum in addition to its direct emissions? If we don’t, say these biofuels advocates, we allow fossil fuels to look better, and thereby cause biofuels to look worse relative to oil than they really are.
One place this argument has been going on is in Oregon. The State of Oregon recently introduced a Clean Fuels Program, similar to California’s Low-Carbon Fuel Standard (LCFS) and the European Union’s Fuel Quality Directive (FQD), aiming to reduce the average carbon intensity of fuels consumed in the state over time. Because biofuels are often blended into gasoline and diesel in order to lower the carbon intensity of the finished fuels, Oregon’s Department of Environmental Quality (DEQ) proposed estimating the ILUC emissions of biofuels in order to more accurately estimate how much they’d lower the overall carbon intensity of transport fuels. But not all biofuels proponents support the inclusion of ILUC accounting in regulations, and the Renewable Fuels Association (RFA) called the Oregon DEQ proposal “inconsistent and unfair.”
The RFA expressed two primary concerns: First, that indirect land-use change remains purely theoretical, and estimates of its magnitude rely on highly disparate model results instead of measurements of direct emissions. Second, that fossil fuels also generate indirect emissions, and therefore including only ILUC for biofuels creates an unlevel playing field and unjustly penalizes biofuels.
There is by this point a lot of solid work that refutes the first of these concerns. If arable land—a limited resource—is diverted to meet the demand for biofuels, that shift must be accounted for somewhere else in the world. While the magnitude of emissions and where they occur differs between various models that researchers use to analyze that shift, nearly all modeling exercises agree that these emissions occur and that they are big enough to significantly affect the life-cycle emissions of biofuels. For example, the modeling underlying California’s LCFS estimated that corn ethanol made from corn grown in the Midwestern U.S. has an ILUC emissions value of 19.8 grams of CO2-equivalents per megajoule—about a quarter of the net life-cycle emissions of the finished fuel that ethanol would be blended into. Excluding ILUC would fundamentally change the scope and effect of the LCFS.
As for the RFA’s second objection, it’s not unreasonable on the face of it to ask whether policy-driven shifts in demand for petroleum could also trigger behavioral or market shifts that, in turn, impact GHG emissions. In fact, it’s such a reasonable question that we recently assessed six potential sources of indirect emissions for petroleum and compared their certainty and magnitude to ILUC. The emissions sources we examined included:
- Induced Land development
- Military involvement
- Carbon intensity of marginal oil
- Accidents (including oil spills and fires)
- Co-products and the carbon intensity of refining
- Price effects (i.e., the rebound effect)
How did the indirect emissions from petroleum production compare to ILUC? The two effects that had the biggest potential impact on the carbon intensity of petroleum were the rebound effect and shifts in marginal petroleum production. The rebound effect covers the extent to which reduced petroleum demand could trigger a decline in prices and a partial bump in demand (hence the term “rebound”). And on the margins of production, the petroleum produced or avoided in response to policies could have a very different carbon intensity than the average petroleum mix. The impact of these two effects ranged from -5 to 10 gCO2e per megajoule, indicating a wide variation in the magnitude and even sign of the emissions. Further research on these emissions sources would be necessary before they are incorporated into regulations.
For the remaining four emissions sources, such as military involvement in the protection of petroleum supply routes, emissions from accidental spills, and induced land development, the potential emissions were an order of magnitude smaller than ILUC emissions. Across the board, they were far closer to zero than to the 19.8 gCO2e/MJ that the LCFS estimates for corn ethanol ILUC. It’s also worth noting that the production of biofuels would have some of the same indirect effects. If similar treatment is a key goal, we should also assess those emissions in the biofuel case. For example, development of oil fields in the rainforest can generate infrastructure development (e.g. road construction) that results in increased rates of local deforestation. Similarly, biofuel production may also drive infrastructure development in forested regions, e.g., in the palm oil supply chain.
As demonstrated by the high variability and lack of consensus on impact of the indirect emissions sources above, the body of research on indirect effects of petroleum production simply hasn’t drawn similarly strong conclusions as on ILUC emissions. None of the indirect effects for petroleum mentioned above is significant enough to fundamentally invalidate policymakers’ understanding of “average” fossil fuel carbon intensity. Whatever else happens in the oil supply chain, fossil fuel combustion emissions remain the dominant term in the lifecycle. In contrast, ILUC emissions are likely an order of magnitude above most indirect emissions sources for fossil fuels—and more importantly, may counteract the relative benefits for some types of biofuels. RFA’s criticisms of Oregon’s Clean Fuels Program simply don’t hold up to closer analysis: the absence of indirect effects for petroleum-derived fuels has a far smaller impact on their carbon intensity than the corresponding exclusion of ILUC would impact biofuels’ carbon intensity. While there’s a case to continue research into indirect effects of fossil fuel production and improve our understanding of market-mediated GHG emissions, those efforts shouldn’t preclude the inclusion of ILUC in policy design.