Industrial utility plants are usually comprised of many interconnected units that must constitute a flexible and reliable system capable of meeting process energy requirements under different circumstances (e.g. varying prices, demands, or equipment shutdowns). Also, in order to avoid large economic penalties, the design and operation of a utility plant should consider that the equipment is not fully reliable and that each item needs to receive preventive and corrective maintenance. Conventionally, these issues are handled by installing additional units according to rules of thumb or heuristics, which usually imply excessive capital costs and might even result in designs that cannot satisfy the specified demands for certain situations. In contrast, during the present work a systematic methodology has been developed to address the design and operation of flexible utility plants incorporating reliability and availability considerations. The suggested method is based on a novel modelling and optimisation framework that can address grassroots; design, retrofit, or (pure) operation problems in which design and operational parameters are optimised simultaneously throughout several scenarios. Thereafter, it is possible to define maintenance and failure situations in different operating periods to ensure that the plant will be able to cope with them, while meeting process requirements at minimum cost. Hence, for design cases, the most cost-effective elements of redundancy can be determined without pre-specifying any structural options in the final configuration (e.g. equipment sizes, types, and number of units). Furthermore, the proposed (multiperiod) MILP formulation is robust enough to tackle problems of the size and complexity commonly found in industry, and has the potential of yielding significant economic savings. (c) 2008 Elsevier Ltd. All rights reserved.