MULTIOBJECTIVE ANALYSIS AND PLANNING OF THE PETROCHEMICAL INDUSTRY AND THE SYNTHESIS OF DISTILLATION SEQUENCES WITH ENERGY INTEGRATION SCHEMES

Energy related problems in chemical processing systems are addressed in the present dissertation. Particularly the petrochemical industry complex which is a large scale chemical processing system that consumes large amounts of energy and feedstocks to achieve its goal, is studied with respect to the efficiency of energy and feedstock utilization. For that purpose the thermodynamic availability, lost work and carbon content of the feedstocks have been selected as our criteria.; Thermodynamic availability is briefly reviewed, emphasizing its application to steady state flow systems undergoing physical or chemical changes. While thermodynamic availability provides an upper fixed bound for the energy performance of an industry, the lost work (creation of entropy) provides a lower evolutionary one. The structuring then of the petrochemical industry has been considered both as a single objective and as a multiobjective problem. Three objective functions have been considered: The maximization of the thermodynamic availability change, the minimization of the lost work and the minimization of the feedback consumption. The first two objectives aim at structuring the industry for optimum energy utilization, while the third aims at the optimum utilization of raw materials. The single objective analysis provides three optimum structures that constitute bounds of the actual performance of the petrochemical industry. The multiobjective analysis on the other hand provides the process designer a set of alternative solutions and subjective criteria have to be used in order to select the "best" structure.; The problem of the synthesis of distillation sequences with energy integration scheme is addressed. Detailed analysis of the properties and the structure of the problem reveals that a weak decomposition of the sequencing and heat integration subproblems can be accomplished. The overall temperature difference between the condensing and reboiling streams and the overall heat load of a sequence, are identified as the two most important properties for this decomposition. It is then shown that the optimum and near optimum non-integrated sequences will provide the best heat integrated distillation sequences. The synthesis strategy is then applied to a number of example problems.; Finally some preliminary work of the synthesis of sloppy distillation separations is presented.