Advancing life cycle assessment of bioenergy crops with global land use models

Bioenergy crops can cut greenhouse gas (GHG) emissions, yet often bring hard-to-quantify environmental impacts. We present an approach for integrating global land use modeling into life cycle assessment (LCA) to estimate effects of bioenergy crops. The approach involves methodological choices connected to time horizons, scenarios of GHG prices and socioeconomic pathways, and flexible data transfer between models. Land-use change emissions are treated as totals, avoiding uncertain separation into direct and indirect emissions. The land use model MAgPIE is used to generate scenarios up to 2070 of land use, GHG emissions, irrigation and fertilizer use with different scales of perennial grass bioenergy crop deployment. We find that land use-related CO2 emission for bioenergy range from 2 to 35 tonne TJ-1, depending on bioenergy demand, policy context, year and accounting method. GHG emissions per unit of bioenergy do not increase with bioenergy demand in presence of an emission tax. With a GHG price of 40 or 200 $ tonne-1 CO2, GHG per bioenergy remain similar if the demand is doubled. A carbon tax thus has a stronger effect on emissions than bioenergy demand. These findings suggest that even a relatively moderate GHG price (40 $ tonne-1 CO2) can prevent significant emissions, highlighting the critical role governance plays in securing the climate benefits of bioenergy. However, realizing these benefits in practice will depend on a coherent policy framework for pricing CO2 emissions from land-use change, which is currently absent. Overall, our approach addresses direct and indirect effects associated with irrigation, machinery fuel and fertilizer use as well as emissions. Thanks to a global spatial coverage and temporal dimension, it facilitates a systematic and consistent inclusion of indirect effects in a global analysis framework. Future research can build on our open-source data/software to study different regions, bioenergy products or impacts.