Differences between Mesophilic and Thermophilic Anaerobic Syntrophilic Digestion of Volatile Fatty Acids in a Continuous Reactor

Transcription

1 Differences between Mesophilic and Thermophilic Anaerobic Syntrophilic Digestion of Volatile Fatty Acids in a Continuous Reactor T. Amani, M. Nosrati, and S.M. Mousavi Abstract Volatile fatty acids (VFAs) are key intermediates in anaerobic digestion. Enriched acetogenic and methanogenic cultures were used for syntrophic digestion of VFAs in a continuous reactor at thermophic and mesophilic conditions. Interactive effects of main microbiological and operating parameters (propionate (HPr), butyrate (HBu), acetate (HAc), hydraulic retention time (HRT) and methanogen to acetogen population ratio (M/A)) were analyzed at both temperatures and compared. The results showed that performances of the mesophilic and thermophilic processes were similar and sensible differences were not observed. Also, optimization of the processes was showed that optimum conditions (corresponding to maximum VFA removal and biogas production rate (BPR)) for mesophilic (HPr= mg/l, HBu= mg/l, HAc= mg/l, HRT=21 h and M/A=2.5) and thermophilic (HPr= mg/l, HBu= mg/l, HAc= mg/l, HRT=21 h and M/A=2.5) were only little different. Results of verification experiments and predicted values from fitted correlations were in close agreement at 95% confidence interval. Keywords Anaerobic syntrophic digestion, Acetogens, Methanogens, Thermophilic digestion, Mesophilic digestion A I. INTRODUCTION NAEROBIC digestion follows four major steps: hydrolysis, acidogenesis, acetogenesis and methanogenesis [1], and latter two are the most critical. Acetogens (HPr- and HBudegrading bacteria) and methanogens (hydrogenotrophs and aceticlastics) interact syntrophically [2, 3]. HPr and HBu are the most important syntrophic intermediates, and their oxidation is regarded as the rate-limiting step because of thermodynamic restrictions (ΔG Propionate =+76.1 and ΔG Butyrate 48.1 kj/mol at 25 C) [4]. The anaerobic degradation of HPr and HBu can be accomplished through the syntrophic cooperation of HPr- and HBu-oxidizing bacteria and H 2 /formatescavenging partners, which maintain a low H 2 partial pressure [2, 3]. Temperature affects thermodynamics of syntrophic HPrand HBu-oxidation. At elevated temperatures acetogenic conversions (ΔG Propionate =+62.3 kj/mol and ΔG Butyrate=+37.9 T.Amani1 Chemical Engineering Department, Faculty of Engineering, University of Kurdistan, Sanandaj, Iran (corresponding author to provide phone: ; fax: ; t.amani@uok.ac.ir). M. Nosrati2, Biotechnology Group, Chemical Engineering Department, Tarbiat Modares University, Tehran, Iran ( mnosrati20@modares.ac.ir). S.M. Mousavi3 Biotechnology Group, Chemical Engineering Department, Tarbiat Modares University, Tehran, Iran ( mousavi_m@modares.ac.ir). kj/mol at 25 C) are somewhat more favorable [4]. Nevertheless, thermophilic anaerobic digestion often manifests chronically higher VFA concentration (especially HPr) than that for mesophilic [5], which the reasons of this phenomenon has not known yet. VFA conversion studies have been mainly focused on the syntrophic association of acetogens with methanogens undergoing co-cultivation [6 8] and on the effects of all VFAs, especially HPr, on the activities of these microorganisms at mesophilic conditions. Recently, in our previous works [9 12] interactive parameters, their toxic effects and extent of these inhibitions in the anaerobic syntrophilic digestion of VFAs were studied in different reactors (different growth) at mesophilic conditions. Thermophilic anaerobic syntrophilic digestion processes has been rarely studied in the literature, therefore, the main differences between mesophilic and thermophilic proceeses have not reported yet. Response surface methodology (RSM) was chosen because it useful for the modeling and analysis of problems in which a response of interest is influenced by several variables and the objective is to optimize the response [13]. In the last few years, RSM has been applied to analyze, optimize and evaluate the interactive effects of independent factors in numerous chemical, biochemical and bioenvironmental processes, but its application to the analysis and modeling of anaerobic digestion processes, particularly syntrophic anaerobic degradation of VFAs only has been reported in our previous studies. The main objective of this research was to investigate, analyze and model the suspended sludge process of VFA syntrophic anaerobic degradations and comparison of the performances of the enriched acetogenic and methanogenic microorganisms under mesophilic (37 C) and thermophilic (55 C) conditions. The RSM was used to analyze and model the process with respect to the simultaneous effects of five microbiological and operating variables (HPr, HBu, HAc, M/A and HRT), and four parameters were assessed. II. MATERIALS AND METHODS A. Inoculums Enriched acetogens (HPr- and HBu-degrading bacteria) and methanogens (hydrogenotrophs and aceticlastics) were used as inoculums in this research. They were enriched from granular sludge (ph, 7.4; volatile suspended solid (VSS), 67.2 g/l; total suspended solid (TSS), 92.4 g/l) from an UASB reactor treating dairy wastewater. The enrichment processes were 277

2 described in our previous works [9 12]. The concentrations of the HPr-degrading bacteria in the enriched cultures were VSS= 68.4 g/l and TSS=78.2 g/l. For HBu-degrading bacteria, the concentrations were VSS=71.2 g/l and TSS=81.3 g/l; and for methanogens, they were VSS=74.3 g/l and TSS=84.5 g/l. B. Synthetic Wastewater HPr, HBu and HAc (99%, Merck Inc., Germany) were diluted in tap water to achieve synthetic wastewater with desired chemical oxygen demand (COD) levels. The COD:N:P ratio was maintained at 100:4:1 by adding NaNO 3 and KH 2 PO 4 as nitrogen and phosphorus sources, respectively. Oxygen was removed by N 2 sparging for 10 min before feeding to the bioreactor and a balloon containing the N 2 gas was placed on the feed reservoir to prevent oxygen entering into the feed vessel. The ph of the feed and within the reactor was not regulated throughout the experiment and was , corresponding to high and low loads, respectively. To increase the alkalinity, 4 g of NaHCO 3 was added per 1 L of feed. C. Continuous Reactor and Operating Conditions The bioreactor was a glass cylinder with a diameter of 125 mm, a height of 165 mm, and 2-L working volume (Fig. 1). The reactor was seeded by adding 400 ml of the enriched methanogen and acetogen cultures (with defined M/A ratio; 1 to 3 g VSS of enriched methanogenic sludge to g VSS of enriched acetogenic sludge). At start up, the ratio of M/A was taken as the relative amount of their (enriched acetogenic and methanogenic sludge) VSS concentrations. The temperature was monitored with a probe connected to a transmitter and was maintained at mesophilic (37±1 C) and thermophilic (55±1 C) with an electrical heating tape attached to the outside surface of the reactor. Mixing of the cultures and release of gas bubbles trapped in the medium was performed by magnetic stirring for 5 min per every 10 h. The produced biogas was collected by the water-displacement method. The concentrations of HPr, HBu and HAc, HRT and M/A were selected as the primary factors in order to investigate the syntrophic anaerobic digestion of VFAs and central composite design (CCD) was used to design the experiments. The levels of the factors are shown in Table 1. Each factor was varied at five levels, while the other parameters were kept constant. Consequently, 47 experiments were conducted; 32 of them organized in a full factorial design and 10 experiments were related to axial points. The remaining five involved repetition of the central design to obtain a good estimate of the experimental error. Experiments were designed by Design Expert Software (State-Ease Inc., version 7.0.0). D. Analytical Methods Methane content in biogas was determined with a Figaro TGS 2611 methane sensor (FIGARO Inc., USA). Analyses of liquid reactor samples were conducted after centrifugation at 12,000 g for 15 min and for acidification of the supernatant 500 µl of 1.0 N HCl was added to the samples. HPr, HAc and HBu were quantified using an Agilent gas chromatograph (model 7890) equipped with an auto-injector (7683 B series), a flame ionization detector (FID) and a Chrompack Cp-Wax 52 CB fused-silica column. VSS, TSS and COD were determined according to the standard methods [14]. The ph was measured using the Metrohm 620 ph meter (Metrohm Inc., Germany). Fig. 1 Schematic representation of the hydrogen generator set-up: (1) rectifier, (2) positive electrode, (3) negative electrode, (4) electrolysis vessel, (5) enrichment culture vessel TABLE 1 The levels of factors in the experiments based on CCD Factor Low Axial Low Factorial Center High Factorial High Axial ( α) ( 1) (0) (+1) (+α) A: HRT (h) B: M/A (g methanogens/g acetogens) C: HPr concentration (mg/l) D: HBu concentration (mg/l) E: HAc concentration (mg/l) III. RESULTS AND DISCUSSIONS A. Statistical Analysis The experimental conditions and their responses for mesophilic and thermophilic anaerobic digestion processes (data table not shown) were obtained and the responses were fitted to quadratic correlations, and then adequate correlations were found to predict the response variables. The analysis of variance (ANOVA) results for the responses have been summarized in Table 2. The relatively high R 2 values indicated that the quadratic equations for the effluent HPr, HBu, HAc and BPR were very capable of representing the system under the given experimental domain. As seen in Table 2, the fitted correlations were significant at the 95% confidence interval for both of temperatures. The correlation adequacy was tested by the F-test for lack of fit [13]. The lack of fit F-statistics were not statistically significant because the p-values were greater than Adequate precision is a measure of the range of the predicted response relative to its associated error or, in other words, a signal-to-noise ratio. Its desired value is 4 or more [15] and this was the case for the four correlations (Table 2). Simultaneously, low response values for the coefficients of variation (CVs) indicated good accuracy and dependability of the mesophilic and thermophilic experiments. ANOVA analysis results showed that the performances of mesophilic and thermophilic processes were similar and only a little differences between corresponding coefficients were 278

3 observed in the fitted correlations. TABLE 2 ANOVA results for the correlations from DX-7 for the studied responses Response Correlations with significant terms P-value R 2 Adj. R 2 Adequate CV Precision HPr Mesophile A 49.7B+470.8C+102.3D+132.0E +18.4AB 25.1AC 42.5BC+30.0BD+34.7BE+13.7CD < CE 12.1DE+70.6A B C D E 2 Thermophile A 46.9B+458.3C+108.8D+137.2E +18.3AB 11.7AC 31.6BC+30.7BD+26.5BE+47.3CE < A B C D E 2 HBu Mesophile A 77.0B+96.8C+602.3D+153.5E +24.1AB 24.0AD+38.2AE+14.1BC 25.1BD+34.3CD < CE+84.5DE+46.1A B C D E 2 Thermophile A 58.6B+97.2C+590.2D+154.2E +24.9AB 25.8AD+34.7AE+17.0BC 16.0BD+35.2CD < CE+82.8DE+50.3A C D E 2 HAc Mesophile A 80.7B 11.0C+135.9D+805.4E +24.4AB+18.6AD 14.1AE+10.8BC 44.3BE 16.7CD < CE+161.2DE+99.6A B C D E 2 Thermophile A 96.6B 6.1C+133.8D+816.1E +13.5AB 46.1BE+51.5CE+148.4DE +98.6A 2 < B C D E 2 BPR Mesophile B+0.6C+2.3D 3.1E 3.3AB 1.6CD 1.0CE 4.9DE 3.6A 2 1.6B 2 2.9C 2 5.2D 2 6.2E 2 < Thermophile A+4.3B+0.6C+2.6D 3.9E 3.4AB 1.3CD 0.9CE 5.8DE 4.6A 2 2.0B 2 3.6C 2 6.0E 2 < R 2 : determination coefficient, Adj. R 2 : adjusted R 2, CV: coefficient variation. B. Effects of VFAs on HPr Conversion Since acetogenic degradation of HPr and HBu is highly endergonic its anaerobic degradation is repressed thermodynamically at both mesophilic and thermophilic temperatures. Also, generally ph drops due to presence of VFAs leads to inhibit the anaerobic digestion of VFAs. The effects of HAc and HBu on HPr anaerobic degradation is shown in Fig. 2. As seen from the figure, the pattern of the effects of HAc and HBu on HPr conversion was similar at mesophilic and thermophilic temperatures. The fitted correlations in Table 2, shows that coefficient of HPr term is greater at thermophilic condition (470) than that of mesophilic (458) temperature. This means self-inhibitory effect of HPr conversion was steeper at mesophilic process. Therefore, it could be expected that HPr anaerobic oxidation at thermophilic temperature must be a little more favorable. The effects of HAc, HBu, HRT and M/A on HPr oxidation were similar at both temperatures. A. Effects of VFAs on HBu Conversion Like HPr conversion, the patterns of VFAs effects on HBu conversion were similar at mesophilic and thermophilic temperatures. Fig. 3 shows the effects of HAc and HPr on HBu degradation. The fitted correlations in Table 2 show that the positive effect of M/A on HBu oxidation was a little greater at mesophilic (77) temperature than thermophlic (59) condition. In other hand, self-inhibitory effect of HBu was greater at mesophilic condition, slightly. The effects of other variables, HRT, HAc and HPr on HBu degradation were very similar at both temperatures. Consequently, syntrophic anaerobic digestion of HBu was almost the same at mesophilic and thermophilic temperatures. B. Effects of VFAs on HAc Conversion Both HPr and HBu inhibited methanogenic conversion of HAc and obstructed the growth of the aceticlasts at mesophilic and thermophilic temperatures [3, 4, 16, 17]. Fig. 4 shows the effects of HPr and HBu on the anaerobic oxidation of HAc conversion in contour plot. As seen from Table 2, HPr in some ranges had positive effects (negative coefficients) on HAc conversion at both temperatures. The self-inhibitory effect of HAc was a little steeper at thermophilic temperature. Effects of other variables on HAc degradation were similar for both processes. Fig. 2 Effects of HAc and HBu on HPr conversion at mesophilic and thermophilic processes A. Effects of VFAs on BPR Since HPr, HBu and HAc affect growth of all syntrophic microorganisms, especially methanogens, BPR was influenced by the VFAs levels, strongly. The results showed that the patterns of effects of VFAs on BPR were very similar at mesophilic and thermophilic processes. Fig. 5 shows the effects of HAc and HBu on BPR. The fitted correlations in Table 2 show that the effect of HPr on BPR was smaller in compared with the effects of HBu and HAc and also, in some ranges had a positive effect on BPR. The effects of HAc and HBu in all ranges were negative and inhibited the biogas production. The M/A (high coefficient) had strong positive effect on BPR at both temperatures. IV. MAXIMUM VFA REMOVAL AND BPR According to the main target (maximum VFA removals and BPR), the optimum conditions obtained were: HPr= mg/l, HBu= mg/l, HAc= mg/l, HRT=21 h and M/A=2.5 for mesophilic temperature and HPr= mg/l, HBu= mg/l, HAc= mg/l, HRT=21 h and M/A=2.5 for thermophilic process. The results showed that optimum 279

4 conditions for mesophilic and thermophilic processes were similar, therefore we could conclude that syntrophic reactions at mesophilic and thermophilic temperatures are little different, and reported sensible differences between two temperatures could be related to the acidogenic and methanogenic steps in the anaerobic digestion process. To check the accuracy of the fitted correlations at the 95% confidence interval of the optimum conditions, the continuous bioreactors was operated accordingly to compare the actual and predicted responses. The results of the experiments conducted at the optimum conditions showed that verification experiments and predicted values from fitted correlations were in close agreement at a 95% confidence interval for both temperatures (data not shown). and thermophilic processes V. CONCLUSIONS All variables affect the performance of mesophilic and thermophilic processes, similarly. However, acetogenic degradation of HPr and HBu was better at thermophilic temperature, slightly. In other hand, methanogenic conversion of HAc was more favorable at mesophilic condition. Also, biogas production rate at thermophilic temperature was a little greater than that of mesophilic process. Finally, great dissimilarities between anaerobic mesophilic and thermophilic digestion are related to the first two steps of the anaerobic digestion process not due to performance of the syntrophic (acetogenic and methanogenic) stages. Fig. 4 Effects of HPr and HBu on HAc conversion at mesophilic and thermophilic processes BPR (ml/h.l) BPR (ml/h.l) Fig. 3 Effects of HAc and HPr on HBu conversion at mesophilic Fig. 5 Effects of HAc and HBu on BPR at mesophilic and 280

5 thermophilic processes REFERENCES [1] G. Bitton, Wastewater microbiology. A John Wiley & Sons Inc. Publication, 3 rd Edition, [2] B. Schink, A.J.M. Stams, Syntrophism among prokaryotes. In: Dworkin M. (Ed.), The prokaryotes: an evolving electronic resource for the microbiological community, 3 rd Edition, 2005, Springer, New York. [3] T. Amani, M. Nosrati, T.R. Sreekrishnan, Anaerobic digestion from the viewpoint of microbiological, chemical, and operational aspects a review. Environ. Rev., 2010, 18: [4] F.A.M. de Bok, C. Plugge, A.J.M. Stams, Interspecies electron transfer in methanogenic propionate degrading consortia. Water Res., 2004, 38: [5] M. Kim, Y.H. Ahn, R.E. Speece, Comparative process stability and efficiently of anaerobic digestion: mesophilic vs. thermophilic. Water Research, 2002, 36, [6] C. Plugge, M. Balk, A.J.M. Stams, Desulfotomaculum thermobenzoicum subsp. Thermosyntrphicum subsp. nov., a thermophilic, syntrophic, propionate-oxidizing, spore-forming bacterium. Int. J. Syst. Evol. Mic. 2002, 52, [7] T. Kosaka, T. Uchiyama, S. Ishii, M. Enoki, H. Imachi, Y. Kamagata, A. Ohashi, H. Harada, H. Ikenaga, K. Watanabe, Reconstruction and regulation of the central catabolic pathway in the thermophilic propionate-oxidizing syntroph Pelotomaculum thermopropionicum. J. Bact. 2006, 188, [8] M. Tatara, T. Makiuchi, Y. Ueno, M. Goto, K. Sode, Methanogenesis from acetate and propionate by thermophilic down-flow anaerobic packed-bed reactor. Biores. Tech., 2008, 99, [9] T. Amani, M. Nosrati, S.M. Mousavi, R.K. Kermanshahi, Study of syntrophic anaerobic digestion of volatile fatty acids using enriched cultures at mesophilic conditions. Int. J. Environ. Scie. Tech., 2011, 8(1), [10] T. Amani, M. Nosrati, S.M. Mousavi, Using enriched cultures for elevation of anaerobic syntrophic interactions between acetogens and methanogens in a high-load continuous digester. Biores. Tech., 2011, 102, [11] T. Amani, M. Nosrati, S.M. Mousavi, Analysis of the syntrophic anaerobic digestion of volatile fatty acids using enriched cultures in a fixed-bed reactor. Water Environ. Res., Accepted for Publication. [12] T. Amani, M. Nosrati, S.M. Mousavi, RSM analysis of anaerobic syntrophic degradation of volatile fatty acids in a UASB reactor inoculated by the enriched cultures. Biotech. Bioeng. Eng., Accepted for Publication. [13] D.C. Montgomery, Design and analysis of experiments, John Wiley & Sons, New York, [14] APHA Standard methods for the examination of water and wastewater. 18 th ed. Washington, DC: American Public Health Association, [15] R.I. Mason, R.F. Gunst, J.L. Hess, Statistical design and analysis of experiments, eigth applications to engineering and science. 2 nd ed. John Wiley & Sons, New York, [16] C. Gallert, J. Winter, Propionic acid accumulation and degradation during restart of a full-scale anaerobic biowaste digester. Biores. Tech., 2008, 99, [17] A.J.M. Stams, K.C.F. Grolle, C.T.M.J. Frijters, J.B. van Lier, Enrichment of thermophilic propionate-oxidizing bacteria in syntrophy with Methanobacterium thermoautotrophicum or Methanobacterium thermoformicicum. Appl. Environ. Micro., 1992, 58(1),