Using Many-objective Optimization to Design Low Carbon and Weather Resilient Biofuel Supply Chains

As a potential low-carbon transportation energy source, biofuels may offer certain advantages over electrification, but also a number of challenges, including water quality and quantity impacts, high cost, carbon footprint, and land usage. There are also well-founded concerns that biofuel supply chains could be disrupted if extreme weather negatively impacts feedstock yield. In this work, we use a multi-objective optimization framework to explore the competing relationships among cost, environmental footprint, and reliability objectives in the design of national-scale biofuel production pathways. A many-objective evolutionary algorithm (BORG) is used to optimize biofuel production cultivation and processing decisions for the 107 USDA agricultural districts in the U.S. Corn Belt under weather uncertainty using de-trended historical yield data from 1998-2013. Detailed geospatial datasets on land costs and availability, energy prices, and yield for corn, soy, switchgrass, and algae serve as inputs to a techno-economic analysis and life cycle assessment model of feedstock cultivation and fuel processing. BORG is then used to find approximate Pareto optimal biofuel production pathways, measured in terms of four objectives: 1) minimize per unit ($/MJ) costs; 2) minimize greenhouse gas intensity (CO2/MJ); and 3) minimize the maximum annual energy supply shortfall relative to a pre-specified quota 4) minimize water quantity, quality impacts and minimize water inequity. Results show that the overall mix of biofuel feedstocks used strongly controls objective performance across non-dominated biofuel production pathways. Results quantify the potential role of spatial diversification in reducing exposure to supply chain risks in the form of weather-caused drops in crop yield.