Developing a statistical model to predict hydrogen production by a mixed anaerobic mesophilic culture

Optimizing the hydrogen (H2) yield at several initial pH conditions in a mixed batch anaerobic mesophilic culture fed with glucose and linoleic acid (LA) was performed using a three factor three level Box–Benkhen design (BBD). Based on the BBD approach, a statistical model was developed to predict the H2 yield. The variables considered for the experimental design were the LA concentration, the initial pH and the number of times glucose was added to the culture. The D-optimality method predicted a maximum H2 yield of 3.49 mol H2 mol glucose−1 for cultures fed 1.9 g l−1 LA, maintained at an initial pH of 5.15 and received 1.79 glucose additions. The response outcome (H2 yield of 3.38 ± 0.22 mol mol glucose−1) at the nearest setting of the experimental factors (2.0 g l−1 LA, an initial pH of 5.0 and two glucose additions) was 3.3% less than the predicted maximum value. The model provides a useful approach for predicting H2 production when H2 consumers are inhibited in mixed batch anaerobic cultures.