Experimental evaluation of extremum seeking control in semi-industrial microalgae raceway ponds

This work presents the implementation and experimental evaluation of an extremum seeking control (ESC) strategy for pH regulation in semi-industrial microalgae raceway photobioreactors, where pH is one of the most critical variables to control for ensuring stable and productive operation. A model-free optimization approach is proposed based on the classical modulation-demodulation ESC scheme, particularly suited to bioreactors whose dynamics evolve slowly and vary over time due to irradiance, temperature, mixing and biomass growth. To enable within-day convergence, the dither frequency was deliberately placed outside the dominant bandwidth of the process, ensuring sufficient excitation cycles during daylight operation. The high-pass filter (HPF) cutoff was then tuned to reject slow diurnal drifts while providing the phase lead required by the out-of-band design, thus preserving an accurate gradient estimate. In addition, a static data-driven feedforward of solar irradiance was incorporated to cancel predictable disturbances associated with photosynthetic activity, improving transient response and reducing corrective effort from the ESC loop. The approach was deployed on full-scale 80 m2 raceways under realistic outdoor and semi-industrial conditions. Results demonstrate robust pH regulation and fast daytime convergence toward the optimal operating region, with minimal modelling effort, highlighting ESC as a practical and effective tool for real-time pH control in microalgal bioprocesses.