New IFPRI MIRAGE iLUC study released by European Commission

Alternative fuels Life-cycle analyses

The European Commission has just released the long awaited new report on the modelling of indirect land use change using IFPRI’s MIRAGE economic model ( The paper adopts a new scenario based on the National Renewable Energy Action Plans and applies several model improvements, to update both the overall conclusions of the modelling published last summer and the marginal iLUC factors for individual feedstocks.

Graph of carbon emissions including direct and indirect

Carbon intensity of biofuel feedstocks in the EU: Direct carbon emissions from the Renewable Energy Directive defaults; indirect emissions from the new IFPRI study (revised to include the ICCT’s higher estimate of peat emissions).

Here are a few key points from the new study:

  •  European biofuels mandates are likely to cause significant indirect land use change emissions. IFPRI calculates an average for the policy of about 38 g CO2e/MJ. Using improved estimates for the 20-year impact of peat degradation, we suggest that this figure could rise to about 50 g CO2/MJ.  
  • Biodiesel has much higher iLUC emissions than ethanol. Based on IFPRI’s results, biodiesel from rapeseed, soy, sunflower or palm is more emissions intensive than fossil diesel. Sugar and cereal based ethanol production, however, could potentially deliver 50% net carbon reductions.
  • A large fraction of carbon emissions, especially for biodiesel, come from peat degradation in Indonesia and Malaysia. Effective peat protection in these countries would substantially improve the emissions profile of biodiesel.
  • The model includes effects from price led yield increase, displacement of animal feed with co-products (DDGS and oil meals) and reduction in food demand.
  •  If food demand is held constant in the model, the emissions increase by about 20%.
  •  IFPRI notes that there are limitations to the ability of CGE models to reflect some of the detailed dynamics of agricultural systems. They also note that the results are inevitably subject to substantial uncertainty.