A risk-averse two-stage stochastic programming for biomass supply chain planning problem

Abstract

This study addresses the problem of designing a sustainable biomass supply chain (BSC) network under uncertainty. The main challenge lies in determining how to optimally locate biomass processing facilities and manage the flow of materials, such as biomass, biogas, fertilizer, and water, while accounting for uncertain factors. A mixed-integer linear programming model is proposed. The model identifies optimal plant locations, determines the quantities of biomass to be delivered and processed for biogas production, and manages the distribution of outputs to agricultural fields. The objective is to minimize transportation and production costs across a two-echelon BSC network. A risk-neutral two-stage stochastic programming (SP) model is presented to incorporate uncertainties associated with electricity demand and transportation costs. In addition, conditional value-at-risk is used as a risk measure in the modeling and robust solutions are obtained by applying a risk-averse two-stage SP model. Sensitivity analysis is performed to support decision-making processes in BSC management. The proposed BSC models are tested in a sustainable BSC network involving two-echelon biomass supply and biorefinery sites in the municipal area of Izmir in Türkiye. The empirical study on BSC models confirms that the risk parameters influence the objective function value. The experimental findings prove that BSC risk models provide optimal results with lower costs from a cost minimization perspective.