LAUSR

Abstract Twenty-six new data envelopment analysis (DEA) models with 55 biohydrogen production experiments categorized into three groups including dark fermentation (DF), photo fermentation (PF), and dark-photo sequential fermentation (DF-PF) technologies, are used to evaluate their biohydrogen yield efficiency. The results reveal the average yield efficiencies of DF, PF and DF-PF are 0.2844, 0.3460 and 0.7040, respectively. The most efficient overall combination of biohydrogen inputs is PhBR1/Rhodobacter capsulatus B10/Rhodobacter capsulatus in DF-PF. Statistical tests demonstrate DF-PF has statistically double the efficiency of PF and DF, and the efficiency of PF significantly exceeds that of DF, supporting some of the literature findings. A flexible DEA model must be carefully chosen when evaluating biohydrogen production. All inputs and outputs of biohydrogen statistically influenced yield efficiency to a significant level. India and Japan are the top two economies benefitting from improved biohydrogen yield efficiency. Improving biohydrogen yield efficiency can improve macroeconomic growth and develop the renewable hydrogen and biohydrogen industry.