Algae show great potential as an energy feedstock, but its production costs must be reduced to be competitive with traditional energy sources. To this end, CFD-based (computational fluid dynamics-based) optimization studies provide a robust method of determining the maximum algae production. This study presents a framework for maximizing algae production in a packed bioreactor using an artifical neural network as the reduced order model of a reactive computational fluid dynamics simulation of algae growth. The optimized reactor with the kinetic properties of Chlorella species is presented and discussed. It is concluded that convection vortices centered on the packed region, caused by low column aspect ratios (height-to-diameter) and high ratios of gas to liquid Reynolds numbers, produce higher algae growth rates.