Apart from the type of microbial inoculum and feedstock, which is used for fermentative hydrogen production, many other factors such as pH, temperature, hydraulic retention time (HRT), nutrients concentration, hydrogen partial pressure, the presence of inhibitors and hydrogen-consuming microorganisms and the reactor configuration, influence the process. Although the role of each parameter in fermentative hydrogen production is well defined, the optimum conditions of a given factor are not clear, so far. For example, it is well known that the pH influences the activities of hydrogen producing microorganisms, since it directly affects the hydrogenase activity (Dabrock et al., 1992) as well as the metabolic pathway followed. However, there is a wide range of pH values, which have been proposed as optimum for fermentative hydrogen production from different feedstocks. The pH range of 5-7.5 (Fang et al., 2002a; Calli et al., 2008) is usually reported as optimum, even though lower or higher pH values such as pH of 4.5 (Ren et al., 1997) and 9.0 (Lee et al., 2002) have also been proposed that are supposed to give the maximum hydrogen yield.

The operational temperature is another important factor affecting the metabolic pathways involved and influencing the whole process. Up to now, most studies on hydrogen production have been carried out under mesophilic conditions, even though it is well known that hydrogen fermentation at high temperatures (thermophilic conditions) has higher hydrogen yield than the mesophilic equivalent, owing to higher suppression of hydrogen-consuming bacteria. Nevertheless, mesophilic biohydrogen production is preferred for preventing the need for external heating, improving the economics of the process.

Regarding the HRT, for pure substrates such as glucose and sucrose, the widely used values are in the range of 3-8 hours, with the lowest being 1 hour (Chang et al., 2002) and the highest 13.7 hours (Fang and Liu, 2004), while for more complex substrates such as starch, an HRT of 15 or 17 hours is suggested to be necessary due to the slow initial step of hydrolysis (Hussy et al., 2003; Lay, 2000).

From this discussion, it becomes clear that the optimum value for each aforementioned key factor depends on the feedstock, the inoculum used and the prevailing conditions under which the experiments are carried out. Thus, predictions of the reactor performance in terms of hydrogen yields and rates as well as carbohydrate conversion efficiency under different conditions are not accurate. So, the selection of the operational conditions of a real scale hydrogen — producing bioreactor at this stage may be safely predicted only on the basis of lab-scale and pilot-scale experiments.