Microbial community analysis, influence of reactor hybridation and effect of the proportion of glycol ethers/ethanol mixtures in egsb reactors

Resumen

Volatile organic compound (VOC) emissions can cause different problems in the public health and in the environment, acting as a primary pollutant and allowing the formation of secondary pollutants as tropospheric ozone. Due to these problems VOC emissions are regulated in many countries, such as USA and the European Union, in this case by the Council Directive 2010/75/EU. Because of the use of solvents in its productive process, flexographic industry is one of the major contributors to the emissions of these compounds, and biological techniques have been considered as one of the best available technologies for the treatment of VOC emissions in this industrial sector. Among these processes, a new technology the anaerobic bioscrubber is emerging as a feasible technology (patent number WO2015114436A1). In this process, VOC are transferred from the gas phase (air emission) to the liquid phase (water) and then transformed into biogas in an anaerobic expanded granular sludge bed (EGSB) reactor. So, the VOC emissions can be converted into bioenergy. However, the use of EGSB reactors for this process has some inherent barriers as the lack of information about the anaerobic degradation of some of the compounds typically used in the flexographic industry or the loss of biomass in the effluent due to the use of a high up flow liquid velocity. For this last aspect, an alternative configuration to the EGSB reactors should be studied in order to avoid the biomass leakage. In this regard, an anaerobic hybrid reactor configuration, which consists in the installation of a filter of polypropylene rings inside the gas-liquid-solid separator in the upper zone of the reactor, seems to be a good alternative to improve the biomass retention capacity of the EGSB reactors. Furthermore, the flexographic industry uses synthetic organic solvents as glycol ethers, such as 1-ethoxy-2-propanol and/or 1-methoxy-2-propanol, but the anaerobic biodegradation of mixtures of these compounds remain unknown yet. In addition, in the literature there is also barely information available about the possible negative or toxic effect of these glycol ethers on the microbial population responsible of the biotransformation of VOC emission into biogas. In this regard, different molecular techniques, as denaturing gradient gel electrophoresis (DGGE), quantitative polymerase chain reaction (qPCR) and high throughput sequencing technologies, are available to characterize the microbial community and to analyse the microbial evolution in biological systems, such as anaerobic reactors, which can be helpful to check these negative or toxic effects by the presence of some of these solvents. In this context, the main objectives of this PhD thesis are: i) to evaluate an alternative reactor to the EGSB reactor to improve the biomass retention capacity and the performance of a reactor treating glycol ethers and ethanol mixtures; ii) to study the pathways for the anaerobic degradation of glycol ethers as 1-ethoxy-2-propanol (E2P) and 1-methoxy-2-propanol (M2P) used in the flexographic industry and the possible impact on the microbial community; iii) to evaluate the effect of the ethanol / glycol ethers ratio in an EGSB reactor treating mixtures of these compounds; and iv) to compare the performance and microbial communities from laboratory scale reactors and an industrial prototype reactor. An alternative configuration based on the modification of the EGSB with a filter of polypropylene rings, called anaerobic hybrid reactor (AHR), has been compared with the conventional EGSB (control reactor). Both reactors were operated at the same conditions and the experiment was divided in seven stages (from S-I to S-VII). First (S-I), the organic loading rate (OLR) was increased step by step up to 45 kg chemical oxygen demand (COD) m-3 d-1 using a readily biodegradable substrate such as ethanol, then E2P was introduced (S-II and S-III), resulting in a binary mixture of ethanol and E2P, and the total OLR was maintained around 45 kg COD m-3d-1. After that, M2P was also introduced as a new substrate in the reactor feed (ternary mixture of ethanol, E2P and M2P) maintaining the same total OLR (S-IV). Later on, the feeding was switched off to simulate a long-term shutdown of production process, which typically occurs in the printing facilities, and so to check the influence of a long starvation period (S-V). Then, reactors were restarted again by increasing the OLR, step by step with the ternary mixture, and finally, in the last stage the proportion of M2P was slightly increased (S-VI and S-VII). Proportion of glycol ethers in this experiment was always lower than 30% in weight. Results showed a high performance of both reactors with global removal efficiencies (RE) higher than 92% even treating OLR of 54 kg COD m-3 d-1 and RE only decreased below 90% in both reactors during the first days after the feeding of E2P, indicating that biomass was not adapted to this solvent and an adaptation period was needed to be able to metabolize it. The adaptation period was also observed in the evolution of RE of E2P, as only around 20% RE was achieved in both reactors when this compound was introduced, but after 40 - 50 days the RE of E2P increased to 80% and maintained during all the experimental period. Regarding M2P, RE was almost complete (100%) immediately after its introduction in the reactor feed and no adaptation period was needed, which would suggest that both glycol ethers have the same mechanism of degradation. Furthermore, the removal of M2P was practically complete along all the experimental period, even after the increased in its proportion in the feed (S-VII). In addition, during the first days of exposure to the glycol ethers, some intermediate products of their degradation (methanol, acetone and isopropanol) were detected and identified, allowing the clarification of the anaerobic degradation pathways of these compounds. Regarding the biomass retention capacity, the accumulated solids in the effluent (563.2 g in the EGSB reactor and only 293.7 g in the AHR) showed that the AHR had a higher biomass retention capacity than the EGSB reactor. However, the filter installed in the AHR was finally clogged and the rings of the filter had to be replaced by new polypropylene rings to maintain its higher biomass retention capacity. Despite the higher biomass concentration in the AHR, both reactors performed similarly. These results, included in the Chapter 4 of this PhD thesis, have been published in the Journal of Environmental Management (Ferrero, P., San-Valero, P., Gabaldón, C., Martínez-Soria, V., Penya-roja, J.M., 2018. Anaerobic degradation of glycol ether-ethanol mixtures using EGSB and hybrid reactors: Performance comparison and ether cleavage pathway. J. Environ. Manage. 213, 159–167). The dynamics of the microbial community of both reactors was also analysed using different molecular tools, such as denaturing gradient gel electrophoresis (DGGE), quantitative polymerase chain reaction (qPCR) and high throughput sequencing technologies. These analyses revealed an important impact in the microbial populations caused by the introduction of both glycol ethers (E2P and M2P). DGGE technique showed an evolution in the bacterial community from the beginning to the end of the experiment. So, some initially predominant bands (inoculum and/or early stages) decreased their intensity in later stages, and the opposite happened with other bands, which appeared or increased their intensity even becoming predominant with the evolution of the experiment. In addition, DGGE also evidenced that sludge bed of both reactors behaved as a mixed reactor, as the same microbial community structure was found in the bottom and in the top ports of both reactors. qPCR results indicated a toxic effect of the glycol ethers over the bacterial and archaeal populations, as both population concentrations decreased after the introduction of E2P. Later, the population concentration values were recovered after 30 days of exposure to this solvent, which also indicated the necessity of an adaptation period of the microorganisms to degrade anaerobically E2P. Besides, when M2P was afterwards introduced into the feed, no negative influence in the concentration of the bacterial population was found, demonstrating that M2P was anaerobically degraded in the same pathway that E2P, so the bacterial population was already adapted to this substrate (M2P). In contrast, archaeal populations as Methanobacteriales and Methanomicrobiales showed a decreased in their population concentrations after the introduction of M2P and also after the increased in the OLR of this substrate, which suggest a toxic effect of M2P over these archaeal populations. Additionally, the results of high throughput sequencing technique emphasized the importance of the type of substrate over the predominance evolution of the microbial communities along the experimental period. Through this technique, Proteobacteria phylum was observed as the predominant phylum when reactors were fed only with ethanol as carbon source, and the predominance of this phylum was replaced by Firmicutes when glycol ethers, especially M2P, were introduced as substrates. This study, included in Chapter 5, was developed in collaboration with the ‘Laboratoire du Génie de l’Environment Industriel, IMT Mines Alès, Université de Montpellier’ during a research stay carried out under the supervision of Professor Luc Malhautier. Recently, it has been sent for its publication to the journal Bioresource Technology (Ferrero et al., 2018. Link between the anaerobic degradation of glycol ether-ethanol mixtures using EGSB and hybrid reactors and the dynamics of the microbial community structure, Bioresour. Technol. Submitted for decision). To complete the previous study an experiment with high proportion of glycol ethers was carried out using an EGSB reactor. This experiment was divided in 8 stages. First, from stage I to stage III, the start-up of the reactor was carried out using ethanol as only carbon source, increasing the OLR in these stages up to a value of 45 kg COD m-3 d-1. Then, in the following stages (IV-VII), ethanol was progressively replaced by E2P and M2P maintaining the OLR constant at 45 kg COD m-3 d-1, until the last stage (VIII) where only the glycol ethers were fed into the reactor. Reactor had a high performance (RE >95%) during the start-up period, and when glycol ethers were introduced global RE dropped initially to 80% and then it was recovered to 90%, showing again the corresponding adaptation period. Later, when ethanol was progressively replaced by glycol ethers in the feed the global RE slightly decreased, and in the last stage (VIII) where only E2P and M2P (without ethanol) were fed into the reactor, global RE diminished to 80% indicating that the removal of glycol ethers could not be complete. Regarding the individual RE of each compound, a low RE was registered during the first days of exposure to each compound (≈20%), but thereafter their maximum RE were achieved. The maximum RE of E2P was around 70% and it was achieved after 15 days of exposure, while the maximum RE of M2P was 100% and it was achieved after 24 days of exposure. Furthermore, in this last stage (VIII) without ethanol in the feed, a partial degranulation of the sludge bed was observed, which could suggest, among other things, that ethanol is a necessary substrate to maintain the granular structure of the sludge. Extracellular polymeric substances (EPS) content and the microbial community of both kind of final sludges (granulated and degranulated) were studied, revealing that the microbial community of both sludges was almost identical, which indicate that fast changes in the physico-chemical properties of the granular sludge did not cause any change in the structure of the microbial populations. However, the EPS content analysis showed a lower concentration of protein and polysaccharide in the degranulated biomass than in the granulated one, which indicates that the loss of these compounds should be related to the degranulation process. Additionally, specific methanogenic activity (SMA) of each sludge was determined using as substrate both only ethanol or a mixture of ethanol and glycol ethers. Results of SMA analysis demonstrated that: 1) an adaptation period to degrade glycol ethers is needed; and 2) the anaerobic degradation of ethanol carried out by the granular sludge was faster than obtained with the degranulated sludge. This study, included in Chapter 6, has been prepared to be sent for its publication to the journal Applied Microbiology and Biotechnology (Ferrero et al. 2018. Behaviour, stability, and microbial community analysis of EGSB reactor at hihg content of glicol ether solvents in mixtures with etanol. Appl. Microbiol. Biotechnol, ready to be submitted). In the last part of this document, a comparison of the microbial community structure of laboratory reactors and a prototype scale reactor has been shown. For this purpose, samples of biomass of an industrial prototype reactor installed and operated in a printing facility were collected, analysed and compared with those obtained in a laboratory control reactor. The on-site prototype reactor treated mainly ethanol, ethyl acetate and E2P. Typical fluctuations in the OLR derived from changes in the production process of the facility were observed. In spite of these fluctuations stable performance of the reactor was achieved (average RE of 93%) and volatile fatty acids were only accumulated when the organic load was higher than 3 kg COD h-1, which suggested a limit OLR that can be used for assuring the stable operation of the reactor. The microbial community structure was analysed through the DGGE technique and it showed an evolution during the first months of operation in the domains Archaea and Bacteria. Taking into account that the initial inoculum was obtained from a brewery treatment plant, the limitation of the carbon source to only a few organic solvents inherent to the flexographic emissions was suggested to be the cause of this shift. Archaeal populations were the most affected by this change in the carbon source, resulting in a diminution of the biodiversity as the Shannon index showed (from 1.07 to 0.41 in the first 123 days of experiment). Methanosaeta was the dominant microorganism in this domain, and its dominance persist during all the experimental period. The proportion of the archaea Methanospirillum and Methanobacterium increased along the experimental period, which was related to variations in the temperature and in the load, respectively. Regarding the domain of Bacteria, species from Geobacter and Pelobacter genera, which are microorganisms specialised in the ethanol degradation, were the predominant microorganisms. Besides, Methanosaeta and Geobacter are syntrophic partners able to use direct interspecific electron transference for methane production, so the predominance of these species indicated that this phenomenon occurred in the prototype reactor. Comparing laboratory and industrial prototype reactors, both performed similarly with high RE (>90%), showed a similar adaptation period for E2P degradation, and the only difference was the limit value for the organic load associated to the volatile fatty acid accumulation. The microbial community structure showed similar trends in both reactors, with an evolution from the initial sludge during the beginning of the operation in Archaea and in Bacteria. A stable community in the domain of Archaea was finally obtained with the predominance of Methanosaeta. In the domain of Bacteria the predominance of the microorganisms showed a dependence on the used substrate. These results have been published in the Journal of Environmental Management (Bravo, D., Ferrero, P., Penya-roja, J.M., Álvarez-Hornos, F.J., Gabaldón, C., 2017. Control of VOCs from printing press air emissions by anaerobic bioscrubber : Performance and microbial community of an on-site pilot unit. J. Environ. Manage. 197, 287-295)