Energy Savings Through Application of Model Predictive Control to an Air Separation Facility

Abstract

Air separation process plants have the interesting characteristic that energy is considered the raw material that we pay for. The true chemical feedstock, air, is widely available, has the attribute of uniform composition, and is free. Thus, we adopt the viewpoint that energy is the feedstock to be parsimoniously managed, and focus on the application of technologies to achieve the highest product yields per unit of energy. Careful attention to thermodynamic efficiency has always been inherent in the process designs for cryogenic air separation plants. Equally important is the adherence of operating conditions to their optimal values, a task assigned to the plant's control system. This paper addresses the application of Model Predictive Control (MPC), an advanced process control technology common in the process industries, to air separation facilities. The effective utilization of energy is emphasized, with examples of results from several implementations at operating plants.