Machine scheduling models in environmentally focused chemical manufacturing

A case problem involving the strategic repositioning of a chemical finishing operation from long cycle, make-to-stock production of high volume commodity products, to short cycle, make-to-order production of low volume speciality products is described. The focus of the analysis is on makespan feasible scheduling of parallel mixers over a short ordering cycle to minimize the resultant waste solvent stream. A model of the physical manufacturing process with respect to the scheduling criteria is described. This model is used in conjunction with a multi-start descent scheduling algorithm based on dispatching rules for clustering and sequencing followed by local improvement based on random numbers. A series of computational studies using the algorithm in conjunction with data from the case problem are described. Implementation issues associated with the case problem are emphasized.