Coordinator MPC with focus on maximizing throughput

In cases where disturbances of economic importance have a dynamic character and if they occur frequently compared to the controlled plant responses, dynamic optimization is more suitable than traditional RTO. However, in many cases the optimal operation is the same as maximum plant throughput. To realize maximum throughput, the bottleneck(s) must be identified and maximum flow at the bottleneck must be implemented. In this paper we suggest to use a coordinator MPC with experimental step response models to maximize throughput. The local MPCs exploits its models and constraints to estimate the remaining feed capacity in each unit at each sample. The coordinator has then information from the local MPC with distance to the bottleneck and can manipulate on feeds and crossovers to maximize the throughput. The coordinator MPC has been tested on a dynamic simulator for parts of a gas processing plant and performs well for the simulated challenges.