In this study, we optimize the quality and economic performance of biodiesel production through a scenario-based adaptive state feedback control framework. Our goal is to maximize the operational efficiency while ensuring the production of on-spec biodiesel, despite uncertainties in reaction kinetics. A first-principle model is employed to describe the process dynamics, with kinetic parameters estimated online using a moving horizon estimator (MHE). A pool of state feedback controllers is created via off-line nonlinear optimization and metaheuristic algorithm based on sampled kinetic scenarios. Then, the uncertainty space is partitioned into several clusters, each linked to an optimal controller from the pool. During online operation, the appropriate controller is selected by matching the estimated kinetic parameters to the corresponding cluster. Simulation studies on a semi-batch reactor demonstrate that manipulating the methanol feed flow rate and heat duty with our control approach significantly improves operational efficiency, reduces online computational time, and enhances robustness to kinetic uncertainties compared to model predictive control (MPC).