Energy supply chains planning: Risk-based optimization

The planning of an Energy Supply Chain (ESC) aims at maximizing the benefits of the ESC agents, while satisfying the demands of the customers (Stadtler (2005); Shiyu et al. (2020)). Demand variability and supply disruption, originating from the connectivity between supply and demand, can disturb the agents interactions and impair the agents management (Govindan et al. (2017)). In this study, we propose a risk-based optimization approach for the management of ESC. We introduce a Conditional Value at Risk (CVaR) measure with the purpose of measuring and controlling the risk to the ESC management. A multi-objective optimization based by the Non-dominated Sorting Genetic Algorithm (NSGA-II) is performed to search for the solution optimal with respect to the maximization of the ESC total profit and the minimization of the risk under uncertainties. For demonstration, an application is carried out considering a specific oil&gas ESC model with five layers, including crude oil producers, storages, refineries, terminal storages and retailers. Results show that the optimization approach enables the trade-off between the ESC optimal planning and the source of risk that it is subjected to.