(740a) Dynamics and Control of Energy Integrated Distillation Column Networks

Distillation is one of the most energy consuming units in a chemical plant, accounting for about 40% of the total energy consumption. Motivated by this, the design of energy integrated distillation has been as area of rich research activity, resulting in configurations such as vapor recompression distillation, multi-effect distillation, thermally coupled (divided wall) columns, etc. It has been well documented that such configurations lead to significant energy, and thus, cost savings. Energy integration, however, results in severe dynamic interactions between process units, slowly evolving network dynamics, reduction in control degrees of freedom and feedback interactions. Yet, contributions addressing dynamic analysis and control of such networks are rather scarce and are typically system specific.

In this talk, a generic framework for the dynamic analysis and control of such networks is presented. Numerous networks of energy integrated distillation columns are considered, focusing on the underlying energy flow structure. For the visualization of the energy flow structure, energy flow diagrams are used. The energy flow diagrams for these networks reveal the presence of specific patterns such as recycle loops and throughput paths. Each of these patterns has been analyzed previously in a different context and is shown to exhibit hierarchies of time scales in their dynamic response.

An example of double effect distillation (DED) is considered for a detailed dynamic analysis and the results are correlated with the key features of the patterns associated with the corresponding energy flow diagram. A hierarchical control strategy stemming from these patterns is proposed. For a simulation case study, a DED configuration separating a mixture of benzene-toluene-m xylene is modeled with gPROMS (PML v3.1 and IPPFO property package). The presence of multiple time scales in the dynamic response is demonstrated through open loop simulations of the system. A hierarchical control scheme, with controllers designed on the basis of reduced order models of the process, is applied. The effectiveness of the control scheme is demonstrated through closed loop simulations of the system under various operating scenarios.