Environ. Sci. Technol. 2007, 41, 1413-1419
Operation of a Two-Stage Fermentation Process Producing Hydrogen and Methane from Organic Waste YOSHIYUKI UENO,* HISATOMO FUKUI,† AND MASAFUMI GOTO Environmental and Bioengineering Group, Kajima Technical Research Institute, 2-19-1, Tobitakyu, Chofu-shi, Tokyo 182-0036, Japan
A pilot-scale experimental plant for the production of hydrogen and methane by a two-stage fermentation process was constructed and operated using a mixture of pulverized garbage and shredded paper wastes. Thermophilic hydrogen fermentation was established at 60 °C in the first bioreactor by inoculating with seed microflora. Following the hydrogenogenic process, methanogenesis in the second bioreactor was conducted at 55 °C using an internal recirculation packed-bed reactor (IRPR). After conducting steady-state operations under a few selected conditions, the overall hydraulic retention time was optimized at 8 d (hydrogenogenesis, 1.2 d; methanogenesis, 6.8 d), producing 5.4 m3/m3/d of hydrogen and 6.1 m3/m3/d of methane with chemical oxygen demand and volatile suspended solid removal efficiencies of 79.3% and 87.8%, respectively. Maximum hydrogen production yield was calculated to be 2.4 mol/ mol hexose and 56 L/kg COD loaded. The methanogenic performance of the IRPR was stable, although the organic loading rate and the composition of the effluent from the hydrogenogenic process fluctuated substantially. A clone library analysis of the microflora in the hydrogenogenic reactor indicated that hydrogen-producing Thermoanaerobacterium-related organisms in the inoculum were active in the hydrogen fermentation of garbage and paper wastes, although no aseptic operations were applied. We speculate that the operation at high temperature and the inoculation of thermophiles enabled the selective growth of the introduced microorganisms and gave hydrogen fermentation efficiencies comparable to laboratory experiments. This is the first report on fermentative production of hydrogen and methane from organic waste at an actual level.
Introduction In Japan, among various organic wastes, solid organic wastes, such as garbage, wastes from the food industry, and waste paper, are mostly treated by incineration and disposed of in ash landfills. Some of these organic wastes are high in water content; therefore, supplemental fuel is required for treatment by direct incineration. The shortage of landfill space for waste disposal is also a problem in Japan. Thus, innovative * Corresponding author phone: +81-42-489-7524; fax: +81-42489-2896; e-mail:
[email protected]. † Present address: Environmental Engineering Division, Kajima Corporation, Japan. 10.1021/es062127f CCC: $37.00 Published on Web 01/13/2007
2007 American Chemical Society
technology for the reuse and recycling of organic wastes have been required to develop. Anaerobic digestion is one of the effective technologies used to recover energy resources from organic wastes in addition to being a simple and effective biotechnological means of reducing and stabilizing organic wastes (1). Various anaerobic processes including the up-flow anaerobic sludge blanket (UASB) process have been widely used for the treatment of wastewater in breweries and food factories (2). However, an efficient anaerobic process that decomposes wastes containing high concentrations of solid materials, such as garbage, has not been developed. Research and development on a rapid and effective anaerobic digestion system, which is a highly desirable organic waste treatment and energy recovery system suited to Japanese social conditions, have been promoted and conducted (3). The degradation of solid substances is the rate-limiting step in anaerobic digestion (4). Thus, enhancement of solubilization is expected to improve the overall process performance. There have been a considerable number of research publications on the two-stage anaerobic process; these have mainly focused on the combination of solubilizing acidogenic processes for solid materials and a methanogenic process (5-7). In the anaerobic digestion process, surplus electrons that can be recovered as hydrogen gas are formed through fermentation metabolism. In order to recover hydrogen gas from organic waste, the hydrogen fermentation of organic wastes has been studied and optimized at fundamental levels (8-11). Continuous hydrogen production was observed from industrial wastewater (12) and artificial garbage slurry (13) using natural populations of microorganisms without sterilization. These reports suggest that the selection of suitable microflora and their cultivation under appropriate conditions would enable stable hydrogen fermentation if the process is operated at an actual level. The two-stage system essentially comprises acidogenic and methanogenic processes. In the first, acidogenic process, organic polymers, carbohydrates, proteins, and lipids are degraded to volatile fatty acids (VFAs), which are metabolized to methane in the subsequent methanogenic step. In order to realize a two-stage fermentation process comprising both hydrogen and methane fermentation, a high-performance methanogenic reactor is required subsequent to the hydrogenogenic process. Because the retention time of the hydrogenogenic operation is relatively shorter than that in conventional methanogenic reactors (11), the required reactor dimensions should be substantially different between the two processes if a conventional slow-rate methanogenic process is connected as the second stage. Fixed-bed or packed-bed anaerobic reactors are capable of efficiently converting organic wastes to biogas even if these contain a substantial amount of solid matter. A thermophilic downflow anaerobic packed-bed reactor (TDAPR) has been studied and reported to be one of the reactor designs that enable efficient digestion of wastewater containing a high concentration of VFAs (14). An nternal recirculation packed-bed reactor (IRPR) has also been developed as a high-performance reactor for the anaerobic digestion of solid matter (15). We have previously conducted and reported on laboratory-scale experiments of two-stage anaerobic processes using artificial organic solid waste. It was confirmed that the maximum allowable organic loading rate (OLR) achieved in the two-stage process was higher than that in the single methanogenic process. Furthermore, the hydrogenogenic operation was more suitable for combination with the VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1413
TABLE 1. Properties of Garbage Slurry Amended with Shredded Paper (GSP)a average total chemical oxygen demand (COD) dissolved COD suspended solids volatile suspended solids lactate formate acetate propionate i-butyrate n-butyrate i-valeriate n-valeriate NH4-N total N total P total carbohydrates dissolved carbohydrates lipids pH
range
140000 (18800) 111100-171300 57500 (8200) 68300 (7300) 66600 (7400) 6800 (1700) 530 (990) 2000 (380) 190 (140) 0 (0) 5600 (1700) 0 (0.7) 34 (37) 41 (15) 2700 (700) 200 (40) 45000 (6600) 21500 (5000) 20700 (6300) 4.3 (0.3)
45400-70700 54700-81800 53100-81700 3170-9870 0-2840 1090-2600 0-440 0 3030-8330 0-2 0-137 12-59 1880-4010 153-285 31000-56000 12000-30000 12000-37000 3.8-4.7
a Values (except for pH) are shown as concentrations (mg/L). Average values are shown with standard deviations in parentheses.
methanogenic process than the solubilizing operation, since the required retention time of the hydrogenogenic operation is much shorter than that of the solubilizing operation; it achieves an almost identical level of methanogenic efficiency in terms of both chemical oxygen demand (COD) and volatile suspended solid (VSS) removals (16). The operation of a two-stage fermentation process producing hydrogen and methane has been reported for a model system at the laboratory scale (17). In the present study, a pilot-scale experimental apparatus was designed and constructed to examine the process performance of the two-stage fermentation process using real organic solid waste. Garbage from restaurants and shredded paper from offices were mixed and treated by the two-stage fermentation process. We report here on the results of the implementation of hydrogen fermentation using a pilot-scale plant. In addition, a microbial community analysis of the hydrogenproducing microflora was performed using samples obtained at the steady state of hydrogen fermentation. The overall process performance of the two-stage biogas production process comprising hydrogenogenic and methanogenic processes was also evaluated from the perspectives of total energy recovery and total retention time.
Materials and Methods Preparation of the Garbage Slurry. Garbage slurry containing shredded office papers (GSP) was used as the feedstock. The garbage and shredded office papers were collected daily from the restaurants and the offices of the National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan. Foreign materials that are not biologically degradable were removed manually. The garbage was roughly crushed by a compact chopper (MKBC-42; Masuko Sangyo Co., Ltd., Saitama, Japan) and then delivered to a slurry tank equipped with a cutter pump unit (KD80MS; Komatsu Zenoah Co., Saitama Japan) for intermittent pulverization and mixing. The GSP was prepared by adding an equal volume of tap water and a 1% volume of shredded office papers to the garbage slurry in the tank. The slurry tank was maintained at 4 °C. The properties of the GSP are shown in Table 1. Pilot-Scale Experimental Plant. The pilot plant for the two-stage anaerobic digestion process was designed and constructed in Tsukuba, Japan. The plant comprised a slurry 1414
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
tank for feedstock storage, a hydrogenogenic reactor (HR), a buffer tank, a methanogenic reactor (MR), and two gas holders. A schematic diagram of the pilot plant is shown in Figure 1. The HR was a continuous-flow, stirred tank reactor with a 200 L working volume. The MR, an IRPR with a 500 L working volume, was designed by numerical analysis of fluid dynamics. The IRPR, which consisted of an impeller at the bottom and a packed bed in the upper part, was designed to generate a vertical circulation pattern that enabled optimal contact between the liquid phase (slurry) and the microorganisms attached to the packed bed (15). The packed bed was of unwoven carbon fiber textile (approximately 2-mm thickness) that wrapped around the support structure (outer diameter, 60 mm; thickness, 5 mm; and height, 600 mm), and was installed in the direction of flow in the reactor. The reactor body was a double-walled structure designed to maintain an operational temperature of 55 °C by recirculating hot water in the outer jacket. The effluents from each reactor were discharged through the overflow line. The effective volumes of the hydrogenogenic and methanogenic reactors were 110 and 340 L, respectively. The contents of both reactors were continuously mixed by the impeller at 15 rpm. The gas-flow lines were sealed with water to maintain anaerobic conditions. The biogas produced from the reactors was measured and stored in the gas tanks. The biogas from the HR was periodically released into the atmosphere and the biogas from the MR was burnt in an open gas incinerator at appropriate intervals. Operation of the Two-Stage Process. The thermophilic microflora that had been enriched from excess activated sludge compost were used for hydrogen fermentation as described in our previous studies (8, 12, 13). Prior to the pilot plant experiment, the fermentation broth was collected from the reactor in a laboratory, in which stable hydrogen fermentation had been established in a 5-L jar fermentor using the GSP as substrate and was used as the seed miroflora for the pilot plant experiment. One hundred liters of the seed microflora was transferred to the HR and purged with nitrogen gas to remove dissolved oxygen. The seed microflora was then incubated at 60 °C to activate the microorganisms while being agitated at 15 rpm. After the decrease in the oxidation-reduction potential (ORP) due to the decomposition of residual organic substances in the seed microflora was confirmed, the chemostat operation was initiated. The supply of the GSP to the reactor was conducted at different flow rates to test the effect of different hydraulic retention times (HRTs). This was accompanied by the concomitant removal of an equal amount of effluent from the reactor via the overflow. The cultivation was conducted at 60 °C and the pH was regulated at 5.8-6.0 by automatic titration of a 6.25 N NaOH solution. The effluent from the reactor was collected and stored in a buffer tank at 4 °C until fed to the MR. Prior to the two-stage operation, the MR was filled with an anaerobic sludge collected from a commercial thermophilic methanogenic reactor for garbage (METAKLES System, Kajima Co., Tokyo, Japan). The acclimation and proliferation of the microbes in the reactor were then accomplished by gradually increasing the feed rate of the GSP in a single-stage operation. After the start-up period, the effluent from the HR was supplied intermittently to the reactor at the predetermined OLR and HRT. During the experimental period, the pH within the MR was not adjusted or regulated. Upon changing the HRT set point, it was assumed that a new steady state was attained after three HRTs had passed. Analyses. The volumes of biogas produced from the reactors were measured by wet gas meters (WS-1A; Shinagawa Co., Tokyo, Japan). The composition of the biogas collected in a Teflon bag (Ohmi Odor Air Service Co., Shiga, Japan) was analyzed by gas chromatography as described previously
FIGURE 1. Schematic diagram of the two-stage fermentation process. Arrows in schematic diagram indicate the flow directions of raw material (solid line) and biogas produced (dashed line). Arrows in the methanogenic reactor indicate the flow directions of liquid recirculation. Packed bed was installed in shaded part of the methanogenic reactor. (8). Prior to the chemical analysis, the fermentation broth was sampled and filtered using a 1.0-µm pore size membrane filter. To determine the concentration of suspended solids (SS) in the broth, the membrane filter was dried in an oven at 105 °C for 3 h and then weighed after cooling to room temperature in a desiccator. To determine the concentration of VSS in the broth, the SS on the membrane filter was heated to 600 °C for 3 h in an electric furnace and then weighed after cooling to room temperature in a desiccator. The COD was determined using the dichromate method according to the Japanese Industrial Standard (JIS) K-1012 with a COD analyzer (DR-3000; HACH Co., Loveland, CO). The filtrate of the sampled broth was used for the subsequent analyses of the soluble components. The volatile fatty acids (C2-C5) and lactate were determined by using a liquid chromatograph fitted with an organic acid analysis system (LC-10A; Shimadzu Co., Kyoto, Japan). The concentration of carbohydrates was determined by the phenol-sulfuric acid method (18). Clone Library Analysis. A sample of the fermentation broth was taken from the HR and centrifuged at 20 400g at 4 °C for 5 min. The pellet containing the cells was washed twice with 1 mL of extraction buffer (100 mM Tris-HCl [pH 9.0], 40 mM EDTA [pH 8.0]), and then the cells were resuspended in 1 mL of the same buffer in a 2.2-mL screwcap polypropylene vial, which contained 1 mL of baked glass beads and 50 µL of sodium dodecyl sulfate. The cell suspension was subjected to mechanical disruption for 2 min using glass beads (diameter, 100 µm) on a reciprocal shaker (Minibeadbeater; Tomy, Tokyo, Japan). The disrupted suspension was incubated at 50 °C for 5 min and then subjected to a second 2-min mechanical disruption. Three hundred microliters of 3 M sodium acetate (NaOAc) was then added to the solution and the resulting mixture was kept on ice for 15 min to remove the cell debris (19). The crude DNA obtained by this method was further purified by phenol extraction and ethanol precipitation. The amplification of 16S rRNA genes from the purified DNA mixture was performed by a polymerase chain reaction (PCR) using Taq polymerase (Takara, Kyoto, Japan) in accordance with the manufacturer’s instructions. The PCR primers used in the amplification were the universal primer
530F (5′-GTGCCAGCMGCCGCGG-3′; M represents A or C, 514-529 Escherichia coli position) and the prokaryotespecific primer 1490R (5′-GGTTACCTTGTTACGACTT-3′; 1491-1509 E. coli position) (20, 21). The reaction conditions were as follows: initial denaturation at 95 °C for 9 min, followed by 15 cycles of 95 °C for 1 min, 50 °C for 1 min, and 72 °C for 2 min. To reduce the possible bias caused by PCR amplification, the rDNA was amplified in triplicate and the individual PCR products were combined for further analyses. The combined PCR products corresponding to the expected size of the amplified rDNA (1.0 kb) were purified with a Rapid PCR Purification System (Marligen Biosciences Inc., Ijamsville, MD). The DNA fragments were cloned into E. coli TOP10 using the pCR4 TOPO vector from the TOPO TA cloning kit (Invitrogen, Carlsbad, CA) in accordance with the manufacturer’s instructions. The clonal rDNAs were prepared from randomly selected recombinants and were used as templates for sequencing. The sequencing was conducted with the M13 primer and a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) and an automated sequence analyzer (Model 3100, Applied Biosystems, Foster City, CA). The sequences obtained were compared with similar sequences of reference organisms by using a BLAST search (22). The nucleotide sequences reported in this paper have been deposited in the GSDB, DDBJ, EMBL, and NCBI nucleotide sequence databases under accession nos. AB271021-AB271043.
Results Process Performance. The time courses of operation and reactor performances are shown in Figure 2. The overall performance of the two-stage fermentation process at steadystate operation is summarized in Table 2. During the 60 days of continuous operation, the concentration of the GSP and the HRT were adjusted to optimize reactor performance and maximize biogas recovery. Continuous hydrogen fermentation was started at the HRT of 1.1 d (36.2 kg COD/m3/d OLR) using a 3-fold water-diluted GSP after the batch incubation. The OLR in the hydrogenogenic process was then increased by using a 2-fold waterVOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1415
FIGURE 2. Time course of operation parameters and reactor performance for the two-stage fermentation process. (A) Hydraulic retention time (HRT) and organic loading rate (OLR) in the hydrogenogenic reactor (HR). (B) HRT and OLR in the methanogenic reactor (MR). (C) pH and oxidation-reduction potential (ORP) in each reactor. (D) Decomposition of organic substances. (E) Biogas production. Number with arrow indicates selected period of the steady state. Symbols: -, HRT; -, OLR; ], pH in HR; [, ORP in HR; 0, pH in MR; 9, ORP in MR; 4, volatile suspended solids (VSS) in HR; 2, VSS in MR; ×, chemical oxygen demand (COD) in MR; O, hydrogen; b, methane. diluted GSP and the GSP without dilution. The OLR was calculated to be approximately 40-160 kg COD/m3/d throughout the continuous experiment. The biogas production was stable although the concentration and content of the GSP applied fluctuated with a little variation, due to the changes of garbage components on a daily basis. The composition of the biogas produced was 50%-60% hydrogen 1416
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
and 40%-50% carbon dioxide. No methane was produced in the HR. On the 15th day of operation, the effluent from the HR was fed into the MR, in which stable methanogenesis had been established using the GSP at the HRT of 4.3 d (11.3 kg COD/m3/d OLR). The HRT applied in the methanogenic process was then increased stepwise to 6.8 d as the concentration of the effluent from the hydrogenogenic process increased. Biogas production with concurrent stable organic decomposition was observed at each of the OLR and HRT applied in the methanogenic process. The typical composition of the biogas was 63% methane and 37% carbon dioxide. The decomposition of VSS was over 80% in the methanogenic process although it did not exceed 30% in the hydrogenogenic process throughout the experiment (Figure 2D). The COD removal efficiency during hydrogen fermentation was not expected to be high since the organic substances produced (VFAs and alcohols) remained in the process and methane gas production, which results from the decomposition of organic substances, had not occurred. Significant difference was not observed in the values of the COD measured before and after the hydrogenogenic process (data not shown). However, a COD removal efficiency greater than 75% was maintained in the methanogenic process during the two-stage operation. The pH remained approximately stable at less than 8 in the methanogenic process, even though a slight increase in the value was observed when the OLR was increased. This could have been due to an increase in the amount of NaOH solution added when the OLR in the hydrogenogenic process was raised. The ORP values varied between -300 and -420 mV in the hydrogenogenic process, whereas in the methanogenic process the value was approximately stable at -600 mV (Figure 2C). As for the overall performance, the most efficient COD removal (79.3%) was realized at the HRT of 6.8 d with the initial OLR of 15.7 kg COD/m3/d in the methanogenic process that was combined with the hydrogenogenic process operated at the HRT of 1.2 d with the OLR of 97.0 kg COD/m3/d. The rates of hydrogen and methane gas production were 5.4 and 6.1 m3/m3/d, respectively, with a total retention time of 8 d (operation 4 in Table 2). On the other hand, the shortest retention time was achieved at the HRT of 4.3 d with the initial OLR of 12.4 kg COD/m3/d in the methanogenic process that was combined with the hydrogenogenic process operated at the HRT of 0.6 d with the OLR of 95.5 kg COD/m3/d. The rates of hydrogen and methane gas production were 2.5 and 4.2 m3/m3/d, respectively, with a total retention time of 4.9 d (operation 2 in Table 2). The dilution of the GSP with water brought about a reduction in the allowable HRT in both hydrogenogenic (0.6 d) and methanogenic processes (4.3 d), although the COD removal efficiency was slightly decreased. Optimization of Hydrogen Fermentation. The hydrogen production yield and formation of VFAs at each steady state established in the hydrogen fermentation process are summarized in Table 3. The VFAs detected during the hydrogenogenic operation consisted mainly of lactate, acetate, and butyrate. This implied that homoacetogenic fermentation and acetate/butyrate fermentation were predominant in this process since lactate had already accumulated to high concentrations in the GSP. The hydrogen production yield, which is calculated from the amounts of carbohydrates decomposed and hydrogen produced, was approximately 1.5-2.4 mol/mol hexose throughout the experiment. When the maximum production of hydrogen gas (5.4 m3/ m3/d) was obtained at the HRT of 1.2 d (97 kg COD/m3/d OLR) (Table 2), the hydrogen gas production was accompanied by the accumulation of significant amounts of
TABLE 2. Overall Performance of the Two-Stage Fermentation Process by Operationa 1
operation condition organic loading rate (kg COD/m3/d) retention time (d) organic decomposition (%) total chemical oxygen demand volatile suspended solids carbohydrates biogas production (m3/m3/d) hydrogen methane
2
3
4
HR
MR
total
HR
MR
total
HR
MR
total
HR
MR
total
53.8 1.1
12.6 4.3
5.4
95.5 0.6
12.4 4.3
4.9
138.1 0.6
16.6 5.7
6.3
97.0 1.2
15.7 6.8
8
33.9 65.6
76.9 82.2 89.6
76.9 88.2 96.4
15.4 40.8
75.5 83.4 88.3
75.5 86.0 93.1
17.7 19.3
69.9 81.4 56.4
69.9 84.7 64.8
26.4 58.1
79.3 83.4 83.5
79.3 87.8 93.0
2.4 NP
NP 4.4
2.4 4.4
2.5 NP
NP 4.2
2.5 4.2
3.3 NP
NP 5.5
3.3 5.5
5.4 NP
NP 6.1
5.4 6.1
a HR: Hydrogenogenic reactor; MR: Methanogenic reactor; NP: not produced. Data are the averages of the values obtained during selected periods of the steady state.
TABLE 3. Hydrogen Production Yield and Concentration of VFAs in the Hydrogenogenic Processa operational condition OLR (kg COD HRT (d) /m3/d) 0.6 0.6 0.6 1.1 1.2 a
89.2 95.5 138.1 53.8 97.0
ORP (mV)
pH
-338.3 -363.9 -313.3 -308.8 -415.8
5.92 5.92 5.95 5.96 6.02
hydrogen production yield
concentration of VFAs (mgL)
(mol/ m3/kg mol hex.b) (COD) lactate formate acetate propionate i-butyrate n-butyrate i-valeriate n-valeriate 1.7 1.6 1.7 1.5 2.4
3.7 2.9 2.6 4.2 6.6
3782 3995 4708 2760 6945
0 0 0 0 0
2248 1745 2286 2020 3330
257 160 155 223 163
0 0 0 7 0
2437 1305 3728 3410 14275
11 191 32 5 52
32 23 12 28 0
Data are the averages of the values obtained during selected periods of the steady state. b Hexose was calculated as C6(H10O5)n.
acetate and butyrate. The observed production rate of hydrogen (5.4 m3/m3/d) was in good agreement with the theoretical production rate (5.2 m3/m3/d). The latter value is based on the calculation of the amounts of acetate and butyrate formed according to the stoichiometry of hydrogen production via acetate/butyrate fermentation from carbohydrates (8). Hence, among the various organic substances contained in the GSP, carbohydrates were exclusively metabolized at the particular culture conditions. At a low HRT of less than 1 d, carbohydrates were demonstrated to be the main substrate for hydrogen production from sugary wastewater (12). This predominance of acetate/butyrate fermentation was observed at a relatively low ORP value (Figure 2 and Table 3). This decrease in the ORP value did not correspond with the tendencies of HRT and OLR, whereas lowering of the ORP value was observed at the OLR of nearly 100 kg COD/ m3/d. The desirable operational conditions for hydrogen fermentation from the GSP could be optimized by strict control of both the HRT and concentration of raw materials. The pH value for hydrogen fermentation was optimized in conjunction with both the HRT (10, 13) and the substrate concentration (23). Since no reducing agents were supplied in this experiments and the GSP contained a relatively high amount of dissolved oxygen, it is also suggested that an appropriate amount of organic substances is required in order to generate the reducing power required to maintain a low ORP value. Clone Library Analysis of the Microbial Community in the HR. In order to identify the microorganisms functioning in the HR during hydrogen fermentation, a clone library was constructed from a sample obtained at the steady state of hydrogen fermentation during which acetate/butyrate fermentation was predominant (Operation 4 in Table 2). A total of 50 clones were obtained and classified into 4 groups. The most frequently observed clones (48%) were closely related to Thermoanaerobacterium thermosaccharolyticum with a
similarity of over 98%. Others were related to Lactobacillus keferi, Saccharomyces barnettii, and Saccharomyces cariocanus with similarities of over 99%, 100%, and 98%, respectively. There were some genotypes among the identified groups in which one or two sequence mismatches were detected. These mismatches could reflect either differences at the strain level or different PCR products generated by slightly different copies of the 16S rDNA gene within a single cell. Among the clones related to T. thermosaccharolyticum, two operational taxonomic units (OTUs) (OTU-1 and -2), comprising 12 clones, were identified in the seed microflora (AB207930 and AB207931) (13). T. thermosaccharolyticum is a hydrogen-producing microorganism that was found in the microflora that had been enriched from sludge compost (24, 25). This implies that the hydrogen-producing microflora was successfully inoculated into the GSP without aseptic operation and functioned adequately in the hydrogen fermentation of real organic waste. Three additional species of microorganisms also detected could have been derived from the GSP itself or have been coincidental contaminants acquired during the preparation process at ambient temperature; these were mesophilic microorganisms that had not been identified in the seed microflora (13). Lactate accumulation in the GSP before feeding into the HR supports this speculation since Lactobacillus keferi is a lactic microorganism. Although they were predominant in the GSP, hydrogen-producing T. thermosaccharolyticum could have proliferated in the HR since the HR was operated at thermophilic conditions (60 °C).
Discussion In the present study, a two-stage process for the fermentation of organic solid wastes was examined at the pilot scale. This is the first demonstration of a two-stage fermentation process that is combined with hydrogen fermentation from garbage and waste paper. VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1417
In the hydrogenogenic process, the VSS decomposition efficiency was less than approximately 30% or lower under all applied experimental conditions while the carbohydrate decomposition efficiency was approximately 40%-65% except for operation condition set 3 where the OLR was set excessively high (Table 2). This implies that most of the gaseous (molecular) hydrogen was derived from the fermentation of soluble carbohydrates. Almost 70% COD removal and more than 80% VSS decomposition were obtained under all the conditions studied. The total retention time realized in the present study (4.98.0 d) is the shortest so far reported for the two-stage fermentation of organic solid wastes (26, 27), although a digestion rate greater than the present experiment has been reported for the treatment of wastewater that contained less solid materials (28-30). A shortening of the retention time can contribute to an increase in the total recovery of energy since long retention times directly affect the energy consumption required for plant operation. This short retention time could be achieved by combination with a highperformance methanogenic reactor. Effluent from hydrogenogenic process still contains a relatively high amount of organic solid materials if solubilization is imperfect. The methanogenesis facilitated by the IRPR applied as the second reactor was stable, although the OLR and the composition of the effluent from hydrogenogenic process fluctuated substantially. Stable and efficient hydrogen fermentation was established by the inoculation of seed microflora and the regulation of culture conditions under continuous operation. It can be concluded that the hydrogen fermentation of real organic waste is more readily established with thermophilic cultivation than with mesophilic cultivation. Mesophilic culture conditions can easily result in unintended contamination that inhibits the proliferation of hydrogen-producing microorganisms in the process. There have been several reports that describe the enrichment of mesophilic hydrogenproducing microorganisms in mixed populations of microflora (31, 32). Most of these studies used raw materials that were not contaminated with microorganisms. In contrast, waste materials, such as garbage, are generally contaminated with mesophilic microorganisms. In the present study, operation at high temperature and the inoculation of thermophiles promoted the selective growth of the introduced microorganisms and this facilitated hydrogen fermentation with efficiencies comparable to those obtained under laboratory conditions. Sakai et al. also reported open fermentation initiated by the inoculation of thermophilic lactic bacteria (33). In the present study, positive results were obtained from a two-stage operation. It should be noted, however, that the two-stage process ideally requires evaluation as a total system; that is, an evaluation that is based not only on the criterion of energy recovery but also on process operability and economical feasibility. Recycling of the effluent from the MR could mitigate the alkalinity requirements for pH control in the hydrogenogenic process, and this reduced operational costs. As reported previously, a massive decline in hydrogen productivity has also been observed (16). Further important factors that should be taken into consideration when planning for operational use include the collection of garbage and wastewater treatment after the two-stage process. As an integrated system, the utilization of a mixture of hydrogen and methane gas would be one of the possible options for a cogeneration process to be combined with the two-stage fermentation system.
Acknowledgments This work was carried out as a part of a research program in the High Efficiency Bioenergy Conversion Project, which 1418
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
was supported by New Energy and Industrial Technology Development Organization (NEDO) in collaboration with the National Institute of Advanced Industrial Science and Technology, Ebara Corporation, Nishihara Environmental Technology Inc., and Japan Bio-industry Association (JBA). We wish to thank Mr. Y. Suzuki of Marubishi Bioengineering Co. Ltd for his technical support in the collection of engineering data and the maintenance of the experimental facilities. We also extend thanks to Miss Yoko Abe for her analytical support.
Literature Cited (1) Archer, D. B.; Kirsop, B. H. The microbiology and control of anaerobic digestion. In Anaerobic Digestion: A Waste Treatment Technology; Wheatley, A., Ed.; Elsevier: London, 1990; pp 4389. (2) Lettinga, G.; van Velsen, A. F. M.; Hobma, S. W.; de Zeeuw, W.; Klapwijk, A. Use of upflow sludge blanket (USB) reactor concept for biological wastewater treatment, especially for anaerobic treatment. Biotechnol. Bioeng. 1980, 22, 699-734. (3) Biomass Nippon Strategy: Government of Japan 2002. http:// www.maff.go.jp/biomass/eng/biomass_honbun.htm. (4) Eastman, J. A.; Ferguson, J. F. Solubilization and particulate organic carbon during the acid phase of anaerobic digestion. J. Water Pollut. Control Fed. 1981, 53, 352-366. (5) Ghaly, A. E. A comparative study of anaerobic digestion of acid cheese whey and dairy manure in a two-stage reactor. Bioresour. Technol. 1996, 58, 61-72. (6) Ng, W. J.; Wong, K. K.; Chin, K. K. Two-phase anaerobic treatment kinetics of palm oil wastes. Water Res. 1985, 19, 667-669. (7) Mata-Alvarez, J.; Mace´, S.; Llabre´s, P. Anaerobic digestion of organic solid wastes. An overview of research achievements and perspectives. Bioresour. Technol. 2000, 74, 3-16. (8) Ueno, Y.; Kawai, T.; Sato, S.; Otsuka, S.; Morimoto, M. Biological production of hydrogen from cellulose by natural anaerobic microflora. J. Ferment. Bioeng. 1995, 79, 395-397. (9) Kataoka, N.; Miya, A.; Kiriyama, K. Studies on hydrogen production by continuous culture system of hydrogen-producing anaerobic bacteria. Water Sci. Technol. 1997, 36, 41-47. (10) Lay, J. J. Modeling and optimization of anaerobic digested sludge converting starch to hydrogen. Biotechnol. Bioeng. 2000, 68, 269-278. (11) Ueno, Y.; Haruta, S.; Ishii, M.; Igarashi, Y. Changes in product formation and bacterial community by dilution rate on carbohydrate fermentation by methanogenic microflora in continuous flow stirred tank reactor. Appl. Microbiol. Biotechnol. 2001, 57, 65-73. (12) Ueno, Y.; Otsuka, S.; Morimoto, M. Continuous production of hydrogen from industrial wastewater by anaerobic microflora in chemostat culture. J. Ferment. Bioeng. 1996, 82, 194-197. (13) Ueno, Y.; Sasaki, D.; Fukui, H.; Haruta, S.; Ishii, M.; Igarashi, Y. Changes in bacterial community during hydrogen and acid production from organic waste by thermophilic anaerobic microflora. J. Appl. Microbiol. 2005, 101, 331-343. (14) Tatara, M.; Yamazawa, A.; Ueno, Y.; Fukui, H.; Goto, M.; Sode, K. High-Rate Thermophilic Methane Fermentation on ShortChain Fatty Acids in a Down-Flow Anaerobic Packed-Bed Reactor. Bioprocess Biosyst. Eng. 2005, 27, 105-113. (15) Fukui, H.; Ueno, Y.; Tatara, M.; Goto, M. Highly Efficient Digestion of Organic Solid Wastes by Internal Recirculation Packed-bed Reactor. In Proceedings of the 4th International Symposium on Anaerobic Digestion of Solid Waste; Copenhagen, Denmark, August 31- September 2, 2005; pp 271-276. (16) Ueno, Y.; Tatara, M.; Fukui, H.; Makiuchi, T.; Goto, M.; Sode, K. Production of hydrogen and methane from organic solid wastes by phase separation of anaerobic process. Bioresour. Technol. in press. (17) Kraemer, J. T.; Bagley, D. M. Continuous fermentative hydrogen production using a two-phase reactor with recycle. Environ. Sci. Technol. 2005, 39, 3819-3825. (18) Saha, S. K.; Brewer, C. F. Determination of the concentrations of oligosaccharides, complex-type carbohydrates, and glycoproteins using the phenol-sulfuric acid method. Carbohydr. Res. 1994, 17, 157-167. (19) Zhu, H.; Qu, F.; Zhu, L. H. Isolation of genomic DNAs from plants, fungi and bacteria using benzyl chloride. Nucleic Acids Res. 1993, 21, 5279-5280. (20) Lane D. J. 16S/23S rRNA sequencing. In Nucleic Acid Techniques in Bacterial Systematics; Stackebrandt, E.; Goodfellow, M., Eds.; Wiley & Sons: New York, 1991; pp 115-175.
(21) Weisburg, W. G.; Barns, S. M.; Pelletier, D. A.; Lane, D. J. 16S ribosomal DNA amplification for phylogenetic study. J. Bacteriol. 1991, 173, 697-703. (22) Altschul, S. F.; Gish, W.; Miller, W.; Meyers, E. W.; Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 1990, 215, 403410. (23) van Ginkel, S.; Sung, S.; Lay, J. J. Biohydrogen production as a function of pH and substrate concentration. Environ. Sci. Technol. 2001, 35, 4726-4730. (24) Ueno, Y.; Haruta, S.; Ishii, M.; Igarashi, Y. Microbial community in anaerobic hydrogen-producing microflora enriched from sludge compost. Appl. Microbiol. Biotechnol. 2001, 57, 555562. (25) Ueno, Y.; Haruta, S.; Ishii, M.; Igarashi, Y. Characterization of a Microorganism Isolated from the Effluent of Hydrogen Fermentation by Microflora. J. Biosci. Bioeng. 2001, 92, 397400. (26) Nielsen, H. B.; Mladenovska, Z.; Westermann, P.; Ahring, B. K. Comparison of two-stage thermophilic (68 °C/55 °C) anaerobic digestion with one-stage thermophilic (55 °C) digestion of cattle manure. Biotechnol. Bioeng. 2004, 86, 291-300. (27) Lee, J. P.; Lee, J. S.; Park, S. C. Two-phase methanization of food wastes in pilot scale. Appl. Biochem. Biotechnol. 1999, 77-79, 585-593.
(28) Ince, O. Performance of a 2-phase anaerobic digestion system when treating dairy wastewater. Water Res. 1998, 32, 27072713. (29) van Lier, J. B.; Rebac, S.; Lettinga, G. High-rate anaerobic wastewater treatment under psychrophilic and thermophilic conditions. Water Sci. Technol. 1997, 35, 199-206. (30) van den Berg, L. High Rate Reactors for Methane Production. In Biotechnology and Renewable Energy; Moo-Young, M., Ed.; Elsevier Applied Science: New York, 1986; pp 268-275. (31) Chen, C. C.; Lin, C. Y.; Lin, M. C. Acid-base enrichment enhances anaerobic hydrogen production process. Appl. Microbiol. Biotechnol. 2001, 58, 224-228. (32) Logan, B. E.; Oh, S.-E.; Kim, I. S.; van Ginkel, S. Biological hydrogen production measured in batch anaerobic respirometers. Environ. Sci. Technol. 2002, 36, 2530-2535. (33) Sakai, K.; Ezaki, E. Open L-lactic acid fermentation of food refuse using thermophilic Bacillus coagulans and fluorescence in situ hybridization analysis of microflora. J. Biosci. Bioeng. 2006, 101, 457-463.
Received for review September 6, 2006. Revised manuscript received December 8, 2006. Accepted December 8, 2006. ES062127F
VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1419