Optimal control for the flexible operation of zeolite crystallization in tubular reactors

The utilization of electric power from renewable sources is a major factor in the reduction of the CO2 footprint of the process industries. One of the main challenges is the fluctuating nature of the power production from renewable sources. This work investigates the application of a dynamic demand-side management scheme which shifts the production and the power consumption to periods of low prices. The scheme consists of an upper-level dynamic scheduling algorithm and a lower-level model-based controller that implements the schedule. The approach is applied to the production of zeolites in an intensified continuous tubular reactor at pilot and industrial scale and validated at the pilot scale process. First, the rigorous dynamic model of the process is developed and validated using experimental data. Then, the optimal operation and the transition between different operating points are investigated. To be able to adapt to frequent changes in the price of electric energy, a moving horizon optimization framework is developed that uses the dynamic model for the calculation of a price-driven schedule. This leads to reductions of the cost of energy by ~ 10 % in comparison to steady state operation. To ensure an accurate prediction of the response of the plant, precise state and parameter estimates are necessary. Therefore, the observability and optimal performance of estimators are evaluated. Based on this, an estimation scheme is designed and validated using experimental data. Finally, a robust economic model predictive controller is applied to realize a dynamic operation according to the demand-side management schedule under uncertainties. This work demonstrates the flexible usage of electrical power for the operation of the zeolite production process under uncertainties.