While formal methods to account for uncertainty in process optimization have been developed in the literature, little use of this research work has been reported for industrial applications. An attempt is made in this paper to integrate techniques for modelling and optimization under uncertainty in order to explore flexible operating scenaria and energy management schemes of real industrial utility systems. Multiperiod optimization principles are employed to account for variable demands and/or uncertain operating conditions, while discrete operating decisions (e.g. switching on/off a piece of equipment) are explicitly considered, enhancing, thus, the concept of operating cost for flexible operation. (C) 1998 Elsevier Science Ltd. All rights reserved.