Ameliorated hydrogen production through integrated dark-photo fermentation in a flat plate photobioreactor: Mathematical modelling and optimization of energy efficiency

Abstract

Biohydrogen production through ‘one-pot’ integration of the dark fermentative and photofermentative routes has become one of the most researched versions of the bioprocess. Present investigation experimentally and mathematically assessed the performance of a single stage integrated dark-photo (SSIDP) biohydrogen production system focusing on amelioration of hydrogen yield and energy conversion efficiency (ECE) through optimization. Dark fermentative C. acetobutylicum and photofermentative Rhodopseudomonas sp. were combined in a flat plate photobioreactor (FPPBR) and operated adopting a ‘batch – repeated batch cycle’ strategy for biohydrogen production for the first time. Hydrogen yield and ECE were optimized with respect to the ratio of the photo- to dark- fermentative bacteria (P:D), the fraction of medium removed and refilled during operation (γ) and intensity of light (I) by using the statistical tool response surface methodology (RSM). Maximum cumulative hydrogen of 4.44 mol H2/mol glucose was produced from the operation of the SSIDP FPPBR under optimal condition (P:D (2:1), γ (0.1), I (90 W/m2)). The optimum yield was almost 2.24 times higher than the yield (1.98 mol H2/mol glucose) obtained from batch operation of the FPPBR using same quantity of glucose and optimum values of P:D and I. ECE of the optimally operated FPPBR in batch - repeated batch cycle mode (45.31%) was also enhanced by the same factor (2.24) in comparison to that (20.20%) obtained from the batch operation. Switch over to the batch – repeated batch cycle strategy from the batch mode operation led to attainment of 2.96 fold (196.75%) increment on overall energy conversion efficiency. A deterministic mathematical model has been developed for a priori prediction of the profiles of the substrate and products of the SSIDP FPPBR system. Reasonably high degree of agreement between the predicted and experimental data confirmed the validity and adequacy of the evaluative power of the developed model.