Two-stage stochastic optimization of sustainable bioethanol supply chain based on demand uncertainty and seasonal availability of biomass

Bioethanol-blended gasoline fuel is one of the alternatives for reducing CO2 emissions in the transportation sector. Since its production is still dependent on demand fluctuations and feedstock availability, an optimal supply chain design needs to be evaluated. A two-stage stochastic optimization model with mixed-integer linear programming (MILP) was developed for designing the optimal bioethanol supply chain under the constraints of multi-feedstock second-generation biomass based on seasonal availability and demand uncertainty. Two optimization models were implemented. The goal of the first model was to maximize the expected profits of the supply chain based on the optimal combination of biomass type, plant location and biomass-bioethanol pathways. The second extension model focused on maximizing expected profits of the supply chain with the least amount of impact on the environment via the reduction of net emission to result in an optimal supply chain. The validation process was then performed by comparing the stochastic and deterministic models in terms of plant location and operating decisions. The first model was applied to actual data of a Thailand case study involving 26 existing plants and agricultural residue availability. The optimization result illustrates that the stochastic model is 16% more profitable than the deterministic model for all random data sets. Furthermore, the extension model shows that carbon credit can contribute to around 5% of the total profit earned. Thus, the proposed models can address demand uncertainty in supply chain design and can be implemented by policymakers to achieve sustainable bioethanol production.