Modelling and control of an integrated high purity methanol distillation configuration

Process integration in distillation unit operations promises improved process efficiencies. However, to realize these benefits fit for purpose control structures are needed to ensure stable and on-specification operations. In practice, this requires the development of regulatory and advanced process controls accounting for nuances and intricacies of a given design and application. This manuscript focuses on developing a control structure for a novel industrial mass-integrated twin-column distillation configuration, where a multi-component crude methanol feed is refined into AA grade (ultra-high purity) methanol, while meeting many technical and economic requirements. To this end, a Digital Model (a dynamic process model) of the column configuration was developed and used to carry out a thorough dynamic analysis. Based on the understanding gained, a regulatory control layer was introduced to achieve stable operations. An advanced supervisory level control structure and a Model Predictive Control (MPC) structure were then developed and compared on the Digital model where both alternatives achieved on-specification (technical and economic) operations. It was illustrated that a well-designed control structure could achieve stable and on-specification control for this specific application. Further disturbance testing showed the MPC is marginal efficient than a supervisory control but has similar robustness.

Automated summary

This manuscript focuses on developing control structures for a novel industrial mass-integrated twin-column distillation configuration to ensure stable operations and achieve on-specification (technical and economic) operations. The study shows that both a supervisory control and a Model Predictive Control (MPC) structure can achieve stable and on-specification control, with MPC being marginally more efficient but similar in robustness to supervisory control.