Optimal design of flexible and reliable process systems

In the design and operation of chemical process systems, two categories of uncertainty are important: (1) realization of continuous process parameters (product demands, feedstock qualities, reaction kinetics); and (2) discrete states (typically related to equipment availability). This paper presents a general framework for incorporating both types of uncertainty in process system design. A combined flexibility-reliability index (FRI) is introduced to measure the capability of a process system to cope simultaneously with: (1) (stochastic) parameter variations; and (2) equipment failures, by explicitly accounting for process operational interactions and constraints. A design optimization procedure is then proposed for obtaining flexible, reliable process systems at minimum investment cost and maximum average profit; the potential inclusion of maintenance considerations is briefly discussed