Waste solvents from chemical process industries not only reduce material economy but also deteriorate environmental quality. Solvent recycling is a major endeavor in batch as well as continuous chemical process industries as determining optimal separation sequences is a difficult process synthesis problem. This dissertation presents a coupled solvent selection (chemical synthesis) and solvent recycling (process synthesis) approach to pollution prevention. The simultaneous integration of chemical synthesis and process synthesis provides better economic throughput and superior environmental quality. However, this integration poses a challenging problem of multiple conflicting objectives, combinatorial explosion of alternatives, and uncertainties. This dissertation focuses on the development of a new and efficient multiobjective optimization programming (MOP) framework under uncertainty for this simultaneous integration.; The main contributions of this dissertation include: (a) Derivation of general heuristics for optimal batch column configurations by defining, optimizing, and analyzing four performance indices; (b) Hierarchical improvements in discrete optimization algorithms under uncertainty by using a new sampling technique; (c) Development of a stochastic solvent selection model for systematic design of separating agents by using a new efficient discrete optimization algorithm under uncertainty, resulting in totally different but promising solvents; (d) Development of a new and efficient MOP framework under uncertainty for discrete & continuous decisions; (e) Two industrial case studies for this coupled solvent selection and solvent recycling approach in continuous and batch processes, namely (1) acetic acid extraction from water (Eastman Chemicals, Kingsport, TN) and (2) acetonitrile separation from water (Mallinckrodt Chemicals, St. Louis, MO).
展开▼
