Advanced process control of pantolactone synthesis using nonlinear model predictive control (NMPC)

In this paper, the discontinuous synthesis of racemic pantolactone has been described. The chemical steps of the synthesis take place in two stirred tank reactors, operated batchwise. The chemical reactions are highly exothermic. For a good quality of the product, the reactor temperature must be maintained between 12-14°C. The control of the reactor temperature was studied in two different situations, first using Proportional-Integral-Derivative (PID) controllers and second using a Nonlinear Model Predictive Control (NMPC). The aim of this process control study was to reduce the cooling agent consumption and to improve the quality of the product by better reactor temperature control. The pantolactone synthesis process was modeled and simulated using MATLAB/SIMULINK software package. It was demonstrated that using PID controllers the cooling agent consumption can be reduced with 10% (comparing to real plant operation), but the reactor temperature control could be improved. In case of using an advanced reactor temperature control (model predictive control), the cooling agent consumption can be further reduced with 8% and the temperature of the reactor is very well controlled. The applications developed for racemic pantolactone synthesis were validated by comparison with data collected from real plant operation and can be used to improve real plant operation.