Process design in chemical industry has to satisfy a lot of requirements. On the one hand, a process has to fulfill the chemical and technical requirements like, for instance, a demanded amount of product of certain chemical purities. On the other hand, this process has to be as cost efficient as possible due to the increasing competition. In the traditional approach, the process design is based on experience, heuristics and other tools like, for example, flowsheet simulators. Although these processes satisfy the chemical and technical requirements very well, no guarantee on optimality with respect to its cost can be given. Due to this the development of innovative tools and methods like the multi-objective optimization of chemical production apparatuses and processes gain in importance. The obtained Pareto-optimal solutions, visualized as the Pareto-front, represent the best trade-offs between the multiple and often opponent objectives. This approach provides an important advantage to the process designer because he is now in a position to choose between the various best trade-offs.
For this purpose, an innovative software tool has been developed at the Institute of Process and Plant Engineering at the Hamburg University of Technology. This tool is capable of solving single- and multi-objective optimization problems by using specially tailored evolutionary algorithms. Various processes taken from literature and practical industrial processes like the styrene process and the dividing-wall column have been already successfully optimized.
However, every Pareto-front determined by an algorithm or method is always an approximation of the true Pareto-front. As a consequence, the goal of every multi-objective optimization is to approximate the true Pareto-front as precise as possible. The complex problem definition in the case of the multi-objective optimization of whole chemical production processes turns this into a challenging task. Multimodal and rugged functions, limited computing time, just to name a few, can hinder the convergence of the determined Pareto-front to the true Pareto-front as well as the distribution of solutions on the determined Pareto-front. Both result in a loss of important trade-offs for the process designer.
In order to overcome this problem a new method was developed, implemented and tested in detail. The new method, the Two-Step-Optimization, provides an innovative approach of performing the optimization and was successfully applied for the multi-objective optimization of various apparatuses and processes. The Two-Step-Optimization suggests, as indicated by the name, a stepwise approach, by combining single- and multi-objective optimization, to the true Pareto-front. The new method as well as the obtained results for the optimization of a complex MINLP-problem will be the subject of the presentation.
See more of this Group/Topical: Sustainable Engineering Forum