(4b) MINLP Optimization Approach for Managing Integrated Petrochemical Business Planning

In this presentation, we describe an optimization approach for managing polyolefin (PE) planning problem with its upstream ethylene plant as the source of its main raw materials.

The main challenge it this integrated business planning is to optimize profits based on the allocation of ethylene to be used in various PE product grades. We consider the inventory of PE grades, monthly demands in the production planning and all possible product sequences that can be produced each month. Instead of solving the planning problem as two separate businesses (ethylene and PE), here we solve the entire supply chain from ethane as the ethylene main feedstock. The advantages for solving as integrated business planning, we do not need to use ‘transfer price’ for the ethylene as its PE main raw material. The profit optimization is based on true market value for the raw materials and products, with considering all process constraints.

The modeling approach used in the planning model is based on mathematical formulation of mass, energy balances, together with nonlinear cracker yield model and PE production sequences. The model described the entire production material flow from the ethane feedstock to all by products of Ethylene, utility and three PE plants. The entire ethylene product with annual capacity of roughly one million ton annually is used for PE production. The main challenge in this model approach is that the upstream unit requires nonlinear model due to component mass balances, yield model and other non-constraints; and the downstream requires binary variables to manage production sequence selection for feasible production operation. The resulting model is in the form of Mixed Integer Nonlinear Programming (MINLP) model. Here, we will discuss the practical usage of optimization of MINLP planning model for a world class ethylene producer with its polymer businesses.