Supply-demand pinch based methodology for multi-period planning under uncertainty in components qualities with application to gasoline blend planning

Uncertainty in component quality in gasoline blending due to measurement errors and variation in operation leads to planned blends which may not meet quality specifications and re-blending is required. Formulating gasoline blending as chance constrained programming enables a decision maker to decide what percentage of blends will be guaranteed to meet the specifications and balance the increased cost of blends vs. the cost of having to re-blend the off-spec blends. Chance constrained formulation makes the gasoline blend problem nonlinear and nonconvex. In this work, we employ a supply-demand pinch based algorithm to optimize gasoline blend planning with uncertainty in components qualities and examine its performance vs. full-space model. The supply-demand pinch algorithm decomposes the problem into two sub-problems, top-level (NLP) computes optimal blend recipes and the bottom-level (MILP) computes an optimal production plan using the recipes computed at the top-level. Computational efficiency of the algorithm is verified by case studies.