|Ph.D Student||Solovyev Boris|
|Subject||Integration of Process Design and Process Control|
|Department||Department of Chemical Engineering||Supervisor||Professor Daniel Lewin|
Integration of process design and process control is a necessary part of any modern design approach. The necessity of this integration arises from everyday process practice recognizing the design nature in many operational problems erroneously referred as control ones. One of the important process control characteristics influencing the quality of process design is resiliency that can be defined as the ease with which the process can meet its objectives despite disturbances and design uncertainties. Considering a process developed for steady, long-term operations, the main driving force in its design is steady-state economics. From the steady-state economics point of view, resiliency as stated previously cannot provide any useful information. However, it does indicate an infeasible design directly. Consequently, the resiliency of the design at steady state is a constraint that must be satisfied.
Incorporation of the resiliency into a design procedure involves two activities: (a) developing the appropriate resiliency index, (b) constructing the dependency of the index on economically-relevant design parameters. The present work deals with both.
The evaluation of the resiliency index is reduced to the solution of an appropriate optimization problem that is quite complex due to its dual nature. On one hand, it is desired to improve resiliency using the design degrees of freedom, and on the other, it is required to search for the possible process disturbances and design uncertainties leading to infeasible design. The proposed resiliency index is based on a mathematical process model, whose ease of preparation and solution is crucial for the evaluation of the index. It is clear that more precise models will lead to more accurate resiliency estimation avoiding unnecessary over-design and will ultimately reduce process costs. The use of a process simulator provides a commercially-available vehicle for rigorous, rapid, model development and its solution. However, for industrial scale processes, even state-of-the-art commercial simulation packages cannot provide the desirable ease of process solution. The present work proposes a systematic approach to the computation of a resiliency index and its implementation in the design procedure based on rigorous simulation models. The two industrial distillation examples are considered. In both examples, the required dependency of the resiliency index on the economically-relevant design parameters is successfully derived.