Microalgal cultivation for biodiesel production: Process analysis, modeling, and energetic optimization

Traditionally, algal cultivation in sparged photobioreactors has been optimized to maximize biomass productivity without taking into consideration the energy input or the energy that could be harvested. In this study, an energy-based methodology is presented to maximize the net energy gain of the cultivation process by minimizing the energy input and by maximizing the energy output. Options for minimizing the energy input through optimal gas-culture volume ratio and CO2-air ratio and options for maximizing the energy output in terms of lipid production through optimal levels of nutrition and CO2 are presented and validated with experimental results from 900-mL bubble column reactors.;For this purpose, 15 different options were tested and analyzed in this study in a bubble column reactor with working volume of 900 mL under constant light, with a test species Nannochloropsis salina. Of the options tested in this study, the two-stage operation was found to result in high power yields in the range of 9 to 72 W m-3.;As part of this study, new correlation models were developed and validated for timely assessment of lipid content and FAME content of algal biomass that can be of significant benefit in day-to-day operation of large-scale cultivation systems. These models were developed from biochemical reasoning presented in this study in terms of routinely measured spectrophotometric optical density readings at 680 nm and 750 nm. These correlations were calibrated and validated using experimental data on Nannochloropsis salina obtained under a range of laboratory conditions in bubble columns and under outdoor conditions in a closed photobioreactor. Lipid predictions by this correlation agreed well with the measured ones (r2 = 0.88). The ratio of the absorbance at the two wavelengths correlated well with total FAME for the calibration data set (r2 = 0.87, n = 21); and predicted well for the validation dataset (r2 = 0.72, n = 74). Based on this level of prediction, this correlation seems to hold promise for low-cost, continuous, onsite estimation of lipid content and FAME content in algal cultivation systems.