Efficient modelling and control design for suppression of pressure oscillations in an industrial condensation process

Abstract

In this paper, a simple yet versatile modelling approach for sinusoidal pressure oscillations in an industrial multi-stage condensation system and different control schemes for suppression are presented. The investigated condensation system is embedded in an extensive extraction system and is used to separate a hazardous substance out of a gas mixture. The main problem is posed by unknown and unmeasured process disturbances which may ultimately lead to pressure peaks exceeding the ambient pressure. When an over-pressure occurs, the hazardous substance may escape, and a safety shut-down is triggered. A surrogate model, based on observations and the physical motivation that a gas flow contains kinetic and potential energy, is developed. The model is then based on the momentum and the ideal gas equation. The validation shows a good agreement with measurements in the time and frequency domain. To suppress the oscillations and guarantee an under-pressure in the system a linear output feedback controller, an adaptive algorithm for rejection of sinusoidal disturbances and an LQR-Controller with integration of the control error are compared. The simulation shows that the readily implementable linear output feedback control scheme sufficiently suppresses the oscillations and safely keeps the pressure in the under-pressure range.