Integration of Modeling, Design and Control for Efficient Operation of Chemical Processes

The purpose of this study is to develop a model-based methodology for integration of process design and control (IPDC) problem. The new methodology is organized in four hierarchical stages based on a decomposition of the general optimization problem. The objective of each stage is to define the search space and enumerate a set of promising candidates. In each subsequent stage, the search space is reduced until in the final stage only a small number of candidates need to be evaluated. Therefore, while the problem complexity increases with every subsequent stage, the dimension and size of the problem is reduced. The proposed methodology does not have difficulties in handling complex problem formulations with larger number of variables and constraints, and its applicability is highlighted in solving simple optimization problem. Introduction Traditionally, chemical process design and process control are two engineering problems that performed independently, with little or no feedback between each other. Figure 1 shows a schematic representation of the two problems. That is, first the process is designed to achieve an optimum objective based on a fully specified nominal case. Only after the process has been designed the operability aspects are taken into account. These might include the control system design and the safety, reliability and the flexibility of the design. Chemical processes therefore tend to be highly constrained with few degrees of freedom left for process control purposes. This conventional sequential-forward approach has some inherent limitations such as dynamic constraint violations, process overdesign or under performance and does not guarantee robust performance [1]. In practice, process design is often tackled by chemical and process engineers, while process control is often done by control and instrumentation engineers. To overcome the limitations encompassed by the conventional approach, a simultaneous approach for exploiting interactions between process design and process control is needed that will include the process design variables as optimization variables whilst, at the same time, optimizing the controller tuning parameters. The potential economic benefits of such a simultaneous approach are also investigated. Figure 1: Conventional solution approach for process design and control problems. Mohd. Kamaruddin bin Abd. Hamid Phone: +45 4525 2912