An experimental comparison of capacity rationing models

In this paper, we evaluate two approaches to capacity rationing to manage demand in make-to-order manufacturing firms that face seasonal demand in excess of installed capacity. We assume that the firms produce two classes of products, one having a higher profit contribution per unit of capacity than the other. In the first approach, capacity is reserved for the higher profit class in a dynamic fashion using an expected value analysis each time an order is received for the lower profit class. In the second approach, the amount of capacity reserved for the higher profit class is set at the beginning of the planning horizon and never changed. The two approaches are evaluated relative to the case with no capacity rationing (i.e. orders are accepted as they arrive as long as capacity is available) for a wide array of experimental situations by varying the ratio of unit profit contributions for the two product classes, the ratio of total available capacity to total expected demand, the ratio of order arrival rates and order sizes for the two classes, and variability in demand for each class. The results show that both rationing models produce significantly higher levels of profit than the model with no capacity rationing.