Moving Horizon Scheduling of an Air Separation Unit under Fast-Changing Energy Prices

Abstract

Maximizing the benefits of time-of-use pricing for industrial electricity consumers requires varying production rates, such that energy use is shifted from peak price periods to off-peak times during the day. Assuming that excess capacity and product storage are available, production of energy intensive processes can be increased at off-peak times beyond nominal rates, and the stored product can be used at peak times when the production rate is lowered. Under these rapidly changing circumstances, scheduling calculations must take into consideration explicitly the dynamic model of the process, often rendering the scheduling problem intractable in practical amounts of time. To address this challenge, we introduce a class of scheduling-relevant low-order process models, which capture the closed-loop input-output behavior of a plant. We use these models to close the scheduling loop, whereby the scheduling problem is formulated over a moving horizon with feedback. We apply the theoretical concepts to an industrial-scale air separation unit model, demonstrating that variable production rate operation with product storage has the potential for significant operating cost savings while abiding by product quality and safety constraints.