Compufers them. Engng, Vol. 17, No. IO, pp. 957-970, 1993 0098-I 354/93 $6.00 + 0.00 Printed in Great Britain. All rights reserved Copyright 0 1993 Pergamon Press Ltd OPTIMAL DESIGN OF PERVAPORATION SYSTEMS FOR WASTE REDUCTION B. K. SRINIVAS and M. M. EL-HALWAGI Chemical Engineering Department, Auburn University, Auburn, AL 36849, USA (Received 30 October 1992;final revision received 28 January 1993, recerved_for publication 24 March 1993) Abstract-The purpose of this paper is to introduce a systematic approach to the design of pervaporation networks for waste minimization applications. The objective of this design strategy is to synthesize a pervaporation network consisting of pervaporation modules, booster pumps, vacuum pumps, heaters, coolers and condensers that can separate a waste stream containing some undesirable component into a lean (retentate) stream which is essentially free from the undesirable compound, and a rich (permeate stream) in which the undesirable compound is concentrated. A structural representation of the problem is first introduced that is rich enough to embed all potential configurations of the system. The problem is then formulated as an optimization program to minimize the total annualized cost of the entire network subject to environmental, technical and economic objectives. The solution of this program yields the optimal network configuration, the number and sizes of pervaporation modules, booster pumps, vacuum pumps, heaters, coolers and condensers that are required to perform the specified separation task. Finally, the applicability of the design procedure is demonstrated via two case studies on the separation of chlorinated hydrocarbons from wastewater streams INTRODUCTION The worldwide concern of environmental issues has stimulated government agencies to develop more stringent environmental regulations. In addition, in- dustry has invariably adopted waste-management techniques to address the health and environmental risks of improper waste disposal. These techniques can be classified into three options: source reduction via in-plant modification, waste recovery/recycle and waste treatment by detoxifying, neutralizing or de- stroying the undesirable compounds. The first two options, commonly referred to as pollution preven- tion, represent the most promising waste-manage- ment strategies. Indeed, waste recovery is a particularly attractive option. Significant environ- mental and economic benefits can accrue from separ- ating industrial wastes with the objective of recycling/reusing these valuables chemicals and/or the bulk of the water. The identification of an economically-viable waste- recovery system is often a challenging task. It requires the screening of a number of separation systems. These processes include traditional mass-transfer sys- tems such as mass-exchange operations (e.g. absorp- tion, adsorption, extraction, leaching, etc.), and/or emerging separation technologies. In the context of employing emerging technologies, membrane separ- ation processes offer significant advantages over other technologies. In addition to their high selectiv- ities, low energy consumption and moderate cost, they are compact and modular. Therefore, membrane units can be readily added to existing plants to recover wastes from effluent streams. While recent research work has lead to the development of system- atic techniques for the synthesis of waste-recovery systems involving mass-exchange operations (El- Halwagi and Srinivas, 1992; El-Halwagi and Manousiouthakis, 1989, 1990a,b), much less work has been directed towards synthesizing membrane separation networks (e.g. El-Halwagi, 1992). Pervaporation is an emerging technology in which a membrane is used to selectively separate certain species from a liquid feed based upon their preferen- tial permeability through the membrane. The feed flows along one side of the membrane and certain species (typically the undesirable compounds) are stripped as a vapor permeate at the other side of the membrane. The remaining feed is collected as liquid retentate. The driving force for transport is a chemi- cal potential gradient across the membrane resulting from a trans-membrane pressure difference. This pressure gradient can be achieved by employing a vacuum pump on the permeate-side of the membrane and/or cooling the permeate, thereby condensing the vapor. A more detailed description of the process may be found elsewhere (Rautenbach and Albrecht, 1989; Huang, 1991; Ho and Sirkar, 1992). Modeling equations that characterize a pervapora- tion system depend on several factors such as the kind of pervaporation module, properties of the mem- brane material, nature of the separation and the 957