A systematic approach for scheduling production in continuous processing systems

This thesis addresses the problem of scheduling production in continuous processing systems for a specified production horizon. It is concerned with the development of a strategy for the systematic synthesis and analysis of short term production schedules for cases considering constant and time-varying conditions.; Initially, the major reasons preventing optimal operational performance are identified. It is shown that the absence of appropriate computer-based technology at the production scheduling level restrains the integration of production objectives and process operations. It is also shown that the development of suitable technology is being delayed due to insufficient understanding of the continuous processing system scheduling problem. As a first step towards an inclusive solution of this problem, a comprehensive definition is presented in this thesis.; Next, the synthesis and analysis of production schedules are considered. A multistage multiperiod model for cases considering constant conditions is derived from the nonlinear programming formulation of the continuous processing system scheduling problem. The model offers an overall solution for the production horizon under consideration as well as a solution for each production period and provides the means for the integration of production objectives and process operations. Since the model presents severe computational problems, a decomposition strategy which retains the attributes of the multistage multiperiod model but reduces computational costs and difficulties is proposed. Based on hierarchical and relaxation concepts, the strategy decomposes the overall problem into a production operations planning problem and a process operations planning problem.; Finally, a new approach for scheduling production in continuous processing systems is presented. The approach consists of a strategy for the systematic synthesis and analysis of short term production schedules for cases considering constant and time-varying conditions. The strategy incorporates inventory management policies into the solution of these type of problems. Since the strategy makes use of the proposed decomposition scheme, it also enables the calculation of an overall solution for the production horizon under consideration as well as a solution for each production period. To demonstrate the utility of the approach, the strategy is tested on a benchmark problem, the scheduling of production in a petroleum refinery.