Bi-objective stochastic model for the design of large-scale carbon footprint conscious co-firing biomass supply chains

Abstract

Biomass is a sustainable alternative to fossil fuels, as it may be used in biofuel production and coal–biomass co-firing. Hub-and-Spoke networks have been proposed in literature as a modeling approach to design large-scale biomass supply chains (BSCs) to minimize investment, however, environmental aspects should be taking into account in the decision making as greenhouse gas (GHG) emissions are generated from their management due to farming and transportation. To reduce GHGs and investment in BSCs for co-firing, we propose a Bi-objective Two-stage Stochastic Hub-and-Spoke optimization model. Since Hub-and-Spoke networks are NP-hard. The proposed BSC is solved by using a tailored-made solution procedure based on a ϵ-constraint method in combination with a metaheuristic to provide Pareto frontier approximations with a lower computational burden than state-of-the-art exact methods, such as Bender’s Decomposition. The proposed metaheuristics are the Particle Swarm Optimization (PSO) and Simulated Annealing (SA) which find the binary first-stage variables (depot location). Numerical experimentation is conducted to identify the most promising solution procedure by considering a large-scale case study in the U.S. North East Region. The BSC shows a 17% higher cost when aiming to reduce GHG emissions, and the PSO outperformed the SA with respect to solution quality and computational time.