Nonlinear model predictive control of biodiesel production via transesterification of used vegetable oils

• A first principle model is used to describe a batch transesterification process.
• Nonlinear MPC control of a biodiesel fed-batch reactor is demonstrated by simulation.
• The role of nonlinear MPC is the economic optimization of the fed-batch biodiesel process.

The economic performance of an industrial scale semi-batch reactor for biodiesel production via transesterification of used vegetable oils is investigated by simulation using nonlinear model predictive control (NMPC) technology. The objective is to produce biodiesel compliant to the biodiesel standards at the minimum costs. A first-principle model is formulated to describe the dynamics of the reactor mixture temperature and composition. The feed oil and mixture composition are characterized using a pseudo-component approach, and the thermodynamic properties are estimated from group contribution methods. The dynamic model is used by the NMPC framework to predict the optimal control profiles, where a multiple shooting based dynamic optimization problem is solved at every sampling time. Simulation results with the economic performance of an industrial scale semi-batch reactor are presented for control configurations manipulating the methanol feed flow rate and the heat duty.