Scheduling a Two-Stage Multiproduct Process with Limited Product Shelf Life in Intermediate Storage

Many commonly occurring multistage facilities in noncontinuous chemical industries can be modeled as a two-stage process with intermediate storage. A two-step mixed-integer linear programming (MILP) approach is developed for such a class of problems with real and complex features such as limited shelf-lives of intermediate products, batch splitting at the storage, a batch filling multiple orders, and general product specifications. Step 1 forms optimal lots and batches to meet the orders, while step 2 schedules them. Twelve new formulations based on several novel constraint sets to assign consecutive orders to a unit are proposed and evaluated using several examples involving up to 90 batches. These sets are useful even for multistage scheduling problems other than the present problem. For this problem, minimizing setups and tardiness together seems easier and better than minimizing tardiness alone.