Interaction Analysis for Process System Networks
Chemical processing plants are complex networks of interconnected process units. Traditionally, the control approach applied to process networks has been to control each unit individually with little consideration about the effects of the interaction between units. Clearly, this strategy can be effective only when the interactions are relatively insignificant. However, modern process networks are increasingly complex and usually include multiple recycle and bypass streams as well as energy integration configurations. Therefore more effective control strategies are required that can guarantee the safe operation of the network as well as the required performance objectives.
The dynamic behaviour of many process units obeys the laws of thermodynamics. These type of processes are called process systems. They are found to be dissipative with respect to a certain energy storage function related to entropy. In this article we study the properties of a process network formed by the interconnection of dissipative process units. In particular, we clarify how the material and energy flow among the units in the network influence the dissipative properties of the controlled variables. We provide an analysis method that can be used to assess the dissipativity of such process networks depending on the topology of interconnections between the process units. This approach is illustrated using a case study of a process system network that highlights the controllability issues caused by process interactions.