Design and Operation of High Performance Chemical Processes

点击翻译按钮获取中文摘要

This project continues the principal investigator's experimentation with algorithmic methods for the design of complex processes that are difficult to operate and control; that is, processes that are often overdesigned because process engineers are reluctant to design near or within regimes of complex operation, where these processes are often economically optimal. To achieve better designs, work will continue to investigate development of design methodologies that involve the coordination of design, operations, and control optimizations. The prototype software system, PRODOC, will be extended to permit experimentation with new methodologies for eliminating overdesign in a broader class of chemical processes. In these new methodologies, models for design and control are refined in a coordinated fashion as the design evolves. A Model Predictive Controller (MPC) developed previously, enables the improved control needed to operate near or within complex nonlinear regimes. MPC optimization is coordinated with economic optimization to efficiently evaluate the controllability of the most promising designs as the economic optimization proceeds. The project is concerned with two typical chemical processes for experimentation with the new algorithms and design methodologies. It extends and perfects the optimization techniques for locating all solutions, including the global optimum, of complex nonlinear programs, such as arise when implementing the new design methodologies. The contributions, which can be expected to advance the state-of-the-art for all design engineers, are beginning to have an impact on chemical process engineers, who are gaining the ability to reliably design more highly-integrated chemical processes that are more energy- efficient and profitable.