Production Scheduling of an Air Separation Plant

Abstract Cryogenic air separation plants consume a large amount of electricity to produce various gaseous and liquid products; they can reduce their operational cost by proper exploitation of energy contracts and intelligent utilization of liquid products. In this paper, we propose a State Task Network(STN) based model which replicates a representative air separation plant. This STN based representation owing to its significantly more granular modelling can reap higher benefits in terms of improved decision making than the approaches proposed in literature. Production Scheduling based on this model computes the optimal operating conditions for the plant under variable power pricing options and demand scenarios. The modelling framework is very rigorous and includes almost all the real world limitations and constraints in air separation plant operation. This unit wise scheduling approach provides more flexibility and also pave the way for integration of the scheduling layer with the lower layered (hierarchically) control system. The proposed framework has been evaluated on the representative air separation plant and the results demonstrate the benefits of optimal plant operation.