A multiobjective programming approach to waste minimization in the utility systems of chemical processes

In the held of process synthesis, heat integration methodologies have been matured considerably during the past two decades. Today, these techniques are widely accepted as effective tools for improving chemical processes in terms of capital investment and energy consumption. However, as the problems of environmental pollution have become more and more serious in recent years, the development of process integration methods for waste reduction is now recognized as an area of urgent research. Tn this paper, mathematical programming models, which take into account both economical incentives and environmental penalties, are formulated for the design of ''best'' utility systems. The pollution problem associated with a utility system can be mainly attributed to gas emissions (e.g. COx, NOx and SOx()) caused by burning fuels for generating power and/or heating utilities. In some cases, in order to satisfy the additional demand for power, electricity is imported from a central power plant which may also consume fuels. This demand for external electricity should therefore be considered as a hidden source of emission indirectly caused by running the utility plant. To address these environmental concerns, an improved version of the traditional MILP model for utility network design is proposed in this work. Not only the problem of cost minimization can be handled efficiently with an elaborate heat recovery scheme embedded in the modified superstructure, but also the concept of global emission can be incorporated in the model formulation. By making use of the goal programming techniques, appropriate designs of the the utility networks can be obtained according to the decision maker's priority. From the experiences we have gathered so far in solving the improved MILP model, it can be concluded that the proposed techniques are applicable for a wide variety of processes having extremely different utility demands and, also, it is a sensible design approach for establishing a compromise among conflicting evaluation criteria. Copyright (C) 1996 Elsevier Science Ltd