A Bi-objective, Risk-Aversion Optimization Model and Its Application in a Biofuel Supply Chain

Abstract

AbstractThis chapter discusses approaches to incorporate risk aversion in supply chain network design. The design of supply chain networks involves multiple uncertainty sources. Most of the previous works took a risk-neutral approach by modeling the problem as two-stage stochastic formulation. However, most decision-makers are not risk neutral, and a better understanding of the risk involved is germane. The contributions of this chapter are threefold: methodological, algorithmic, and application. From the methodological perspective, we propose a novel mathematical formulation of a bi-objective two-stage stochastic programming model that measures the trade-off between the expected cost and the conditional value at risk (CVaR). From the algorithmic point of view, the augmented ε-constraint method is used for solving the model and getting the Pareto solutions set. From the application side, a real-life data-driven case study at a state level is solved to optimality to obtain pragmatic and managerial insights that enable the investigation of solutions for different levels of risk aversion; this, in turn, helps to increase the production of reliable and cost-effective biofuel. The mathematical model can be transferable to other applications that seek to provide risk-averse solutions to decision-makers.