| The worldwide energy demand has been continuously increasing, thus requesting more sustainable alternatives to the rapidly depleting fossil fuels. Therefore, biofuels such as hydrogen, bioethanol and biodiesel are gaining more importance as a renewable and pollution-free solution, which might give a significant contribution to the future energy mix. In recent years, the exponential growth of biodiesel production has led to a glycerol glut, however, according to some authors, crude glycerol might represent a suitable, abundant and low-priced feedstock for fermentation technologies.So far, most studies were focusing on the use of pure cultures(or uncharacterized mixed cultures) grown on pure glycerol, while the experiments on crude glycerol typically led to a decrease of the fermentation performance, due to the presence of inhibiting compounds. So an alternative approach was investigated in the present study, trying to select and acclimatize a mixed culture directly on the biodiesel-derived crude glycerol. Enrichment of activity sludge through repeated transfers in small batches with a minimal medium(without nutrient supplementation) allowed the selection of a suitable and stable microbial mixed culture, able to grow on crude glycerol as the only carbon source, and using a very simple synthetic medium without any extra nutrient supplements. Microbial characterization, based on 16 S r DNA cloning library, showed that about 58 % of the community was affiliated with the genus Klebsiella sp. and about 32% belonged to the genus Escherichia. Both taxonomic groups are well-known glycerol fermenting microorganisms, able to produce hydrogen and ethanol. The remaining 10 % of the microbial community was affiliated to the genus Cupriavidus, which is not able to use glycerol, but has the ability to convert methanol, a typical contaminant of crude glycerol, which can be inhibiting for the other microbial strains. The enrichment process significantly increased the glycerol conversion efficiency into hydrogen and ethanol and allowed to obtain a stable mixed culture, useful for the subsequent optimization step.Following to statistical optimization in 125 mL batch fermentation tests, the selected pool was able to effectively convert crude glycerol into H2 and ethanol, both at a nearstoichiometric yield. Maximum H2 yield of 0.96 mol H2/ mol glycerol consumed(maximum production rate was 2191 m L H2/L/d), was estimated at the temperature of 37.0 °C, initial p H of 7.9 and glycerol concentration of 15.0 g/L, with around 1 mol Et OH/mol glycerol consumed(corresponding to a concentration of approximately 8 g/L) and a highly efficient substrate conversion efficiency(around 98%). Total energy conversion efficiency reached 98.25%(around 82 % coming from the contribution of ethanol and 16.25% from H2), thus clearly demonstrating the advantage of targeting a joint production of hydrogen and ethanol. The results showed that it is possible to apply design of experiments to a stable mixed microbial culture, to effectively model and optimize glycerol fermentation process, maximizing production of both, hydrogen and ethanol. Besides, more than 50% of the glycerol was converted into ethanol, which represented the main fermentation product.Yields were as high as the best ones reported in literature with pure glycerol in batch tests,andpractically matching the theoretical ones. Moreover, this was probably the first time to maximize production of both products(H2 and ethanol) together. Therefore the joint use of highly selected and stable microbial mixed pool, together with experimental design, seems to be a useful tool to obtain highly efficient conversion of crude glycerol into biofuels, without the need of complex medium containing expensive vitamin- and trace element solution, tryptone or yeast extract.Subsequently, upscale tests were performed, using a 3L bioreactor equipped with controller(working volume of 1 L), to validate the statistical model and further enhance production of target metabolites. The validation of the predictable model provided good results: the difference between predicted values, calculated with the Box-Behnken Optimization Design, and experimental results were as low as 6.5% for H2 and 2.3 % for ethanol, thus showing that the statistical model was valid also in upscaled conditions.Comparison tests using different glycerol types(pure glycerol and two crude glycerol types, both in sterile and non-sterile conditions) showed no substantial differences, leading to a H2 production rate of 2.96 L/L/d ± 185 and a yield of 0.9 mol H2/mol glycerolconsumed. Average cumulative H2 production potential(Pmax) reached 3.62 L/L/d ± 199, with an H2 content in the biogas of around 54% ± 0.99. Kinetic characterization by means of the modified Gompertz equation(to fit experimental data) confirmed the importance of using a mixed culture on a complex substrate. In fact, differently from most papers published in literature, there was no loss of performance compared to pure glycerol, and maximum H2 production was achieved also in non-sterile conditions. Moreover, substrate degradation efficiency did not decrease in upscaled conditions, reaching on average 97.42%, ± 0.98, with a carbon balance of the fermentation products of around 88%.Fed-batch tests, in non-sterile conditions, led to a three-fold increase of ethanol concentration, thus reaching 26g/L, with an average yield of around 0.4 g Et OH/g glycerol consumed, while producing up to 9L of H2, corresponding to an average yield of 0.14 L H2/g glycerol consumed. These results were obtained, using a feed of around 20 g/L glycerol diluted with a minimal medium and without nutrient supplementation. Substrate consumption reached 65.3 g/L of glycerol, with a total of 74.7 g/L of glycerol fed(roughly corresponding to 95 g/L crude glycerol), with an overall substrate degradation efficiency of around 87.4 %, over the 380 h operation time. Ethanol was the clearly the dominant fermentation product, representing about 81% of the soluble metabolites.Based on these findings, I performed a techno-economic study to evaluate the possibility of reaching viability of this process on an industrial scale, with a plant able to produce 28 MWth. Results showed that with 26 g/L of ethanol and a theoretical retention time of 120 h, the calculated energy cost would be about 0.019 €/k Whth and 0.057 €/k Whel, considering the contribution of both, hydrogen and bioethanol. Moreover, bioethanol cost would be as low as 0.21 €/L, even without taking into account the possible hydrogen revenues. These results suggest that the process has reasonable chances to achieve economic viability, thus deserving further attention. The procedure followed in this work provided a realistic and concrete target to pursue in the future lab and scale-up experiments, in order to bring this technology closer to the market.Besides, the possibility to totally exclude the use of any synthetic medium for glycerol fermentation, through the co-fermentation with additional organic waste deriving from agro-zootechnical activities, was also investigated, using a previously selected mixed culture(F210), able to grow on such organic waste. Mixture Design was used to investigate the optimal mixing ratio of crude glycerol co-fermented with cheese whey and buffalo slurry, and to verify how the H2 production yields were affected by the variation of the substrate components(% of substrate composition). In optimized conditions, maximum hydrogen yield of 117 m L H2/g VSadded(corresponding to approximately 460 m L H2) was estimated at a substrate composition of 30% buffalo slurry and 70% cheese whey(R2=0.962; p-value=0.0001; lack of fit>0.05), in good agreement with the experimental data. Even though crude glycerol represented a less efficient substrate in comparison to CW(using F210 as inoculum), however it could still be used with a contribution of 47%(together with 49% CW and 4% BS) in suboptimal conditions, thus still obtaining a H2 production > 379 m L. |
PDF Full Text Request