Abstract The microbial ecology of the sequential, leach-bed, mesophilic anaerobic digestion of unsorted, coarse municipal solid waste (MSW) was examined ...
ws/41(3)/A2
13/4/00
1:42 pm
Page 9
P. Silvey,1 P.C. Pullammanappallil,1 L. Blackall2 and P. Nichols3 CRC for Waste Management and Pollution Control Ltd. 1Department of Chemical Engineering, The University of Queensland, Brisbane QLD 4072, Australia 2Department of Microbiology, The University of Queensland, Brisbane, QLD, 4072, Australia 3CSIRO Marine Research, Hobart, TAS, 7000, Australia
Abstract The microbial ecology of the sequential, leach-bed, mesophilic anaerobic digestion of unsorted, coarse municipal solid waste (MSW) was examined over 80 days. The methane yield was approximately 75% of the ultimate biochemical methane potential (BMP) of the waste loaded into the digesters. The operational strategy involved a sequence of two digesters containing fresh and anaerobically stabilised MSW respectively. Cell wall phospholipid fatty acid (PLFA) and ether lipid (PLEL) analysis was used to monitor changes in microbial biomass. Both Bacterial and Archaeal biomass were heavily influenced by pH during the two-week start up period. Archaeal biomass peaked just before the methane production rate reached a maximum whereas Bacterial biomass peaked at a later stage. Changes in the phylogenetic diversity of the population were monitored by denaturing gradient gel electrophoresis (DGGE). An analysis of the changes in DGGE banding patterns suggested that rapid start-up of a new reactor was effected by inoculation as well as the provision of buffering capacity from the mature reactor leachate. Keywords Anaerobic digestion, microbial ecology, PLFA, DGGE, methanogenesis, leach-bed
Introduction
The putrescible component of municipal solid waste (MSW) can be rapidly treated using an anaerobic leach-bed process in which leachate is recycled between new and mature (anaerobically treated) waste beds to provide moisture, nutrients and inocula for rapid start-up of the new bed (Hall et al., 1985; Chynoweth et al., 1992). The leach-bed digestion process requires two reactors: a “new” reactor containing fresh waste and a “mature” reactor containing anaerobically stabilised waste (Figure 1). The fresh waste is moistened and water in excess of bed saturation requirements is added. The excess water appears at the bottom of the reactor as leachate. This leachate is recirculated to the top of the bed after flushing it through the mature bed. This recirculation process is called “sequencing” and is repeated daily until the pH of the leachate and the gas composition from the new reactor is 6.5 and 30% respectively. When these conditions are met the new reactor is disconnected from the mature reactor and the leachate from the new reactor is recirculated directly to the top of the bed. This process is continued until the waste has degraded. This bed of anaerobically stabilised waste is used to start up another fresh waste bed and the process is repeated. The sequencing lasts for approximately two weeks and is referred to as the “start-up” period. The mature reactor is usually “shut-down” and emptied at this stage. Preliminary experiments in our laboratory over the previous year showed that when a single stage reactor system was employed with a similar recirculation protocol the reactor eventually acidified. A major challenge in operating this process is to achieve a short start-up period while ensuring balanced microbial population development. The acidification of the waste bed and its consequent inhibitory effect on the methanogenic population is a common problem. An appropriate microbial community structure and continued stability of that community under varying loads and environmental conditions are key factors in ensuring efficient and
Water Science and Technology Vol 41 No 3 pp 9–16 © 2000 IWA Publishing and the authors
Microbial ecology of the leach bed anaerobic digestion of unsorted municipal solid waste
9
ws/41(3)/A2
13/4/00
1:42 pm
Page 10
P. Silvey et al.
reliable biological treatment processes. However, the microbial ecology of MSW treatment has not been studied extensively. Analysis of fatty acids of cell membrane lipids, particularly phospholipid (PLFA), has been used to study the biomass, community structure and metabolic status of natural microbial communities. The use of phospholipids to estimate biomass is well documented (White, 1983; Balkwill et al., 1988; Tunlid and White, 1990). Denaturing gradient gel electrophoresis (DGGE) is a molecular method which separates polymerase chain reaction (PCR) amplified regions of genes encoding for 16S rRNA, based solely upon differences in the nucleotide sequence (Myers et al., 1987). DGGE was first used to assess the genetic diversity of complex microbial populations by Muyzer et al., (1993). With the DGGE method, a difference in base sequence will be indicated by a difference in the final position of bands in the denaturing gel. The banding patterns provide profiles of microbial communities that can be used to identify temporal or spatial differences in community structure or to monitor shifts in structure that occur as a result of changes in environmental conditions. Each band in the profile is assumed to have been generated from a distinct phylogenetic population and therefore the number of bands provides an estimate of the species diversity. In this paper both the PLFA and DGGE approaches were used to examine changes in microbial biomass and community structure during leach-bed digestion of unsorted MSW at mesophilic temperature.
Materials and Methods
Details of the experimental equipment and the design of reactors used in this study can be found elsewhere (Chugh et al., 1999). The experiment involved monitoring a set of new and mature reactors in duplicate. Both sets of reactors showed a similar performance profile (Silvey, 1998). The experiment was then repeated with minor modifications in the method of loading. Only one set of reactors from each experiment was subjected to microbial analyses. This paper presents the findings from the second experiment which essentially yielded similar microbial population profiles to the first experiment (Silvey, 1998). Each reactor contained approximately 70 kg (wet wt.) or 37 kg VS (volatile solids) of unsorted coarsely shredded waste. At the start of our study the mature reactor had been operating for 3 months. During the two week start-up, leachate (amounting to 10±2% of the waste bed volume) was transferred between the beds daily. The leachate pH, and the gas volume and composition were monitored daily. Leachate was sampled at the mid-point of the sequencing process to ensure a homogenous, representative sample. The liquid was removed via a tap inserted into the sequencing line. Standard anaerobic techniques were employed for sampling and storage. Duplicate samples were prepared for each sampling event and these were stored immediately at –20ºC. Lipids and DNA were extracted from both samples. Samples of MSW (100 g) in pre-placed mesh bags attached to a length of twine, were retrieved from a reactor when required. Samples were stored immediately at –20ºC under an anaerobic environment prior to lipid extraction and analysis. Samples of headspace gas were obtained using a gas tight syringe (10 mL luer lock). Methane concentration was measured with a Perkin Elmer Autosystem Gas Chromatograph equipped with a thermal conductivity detector (GC-TCD). PLFA and PLEL analysis
10
Fatty acid extraction, fractionation and methylation was performed according to standard methods (White et al., 1979). Duplicates of each sample were extracted and analysed. Methods used for the derivatisation and simultaneous estimation of microbial PLFA and PLEL by capillary gas chromatography are reported elsewhere (Virtue et al., 1996). PLFA
ws/41(3)/A2
13/4/00
1:42 pm
Page 11
P. Silvey et al.
Figure 1 Schematic of the SEBAC process
and PLEL concentration data were converted into microbial abundances on the basis that Bacteria and Archaea contain on average 100 µmol PLFA g–1 dry wt., and 2.5 µmol PLEL g–1 dry wt respectively; cell mass is equal to 5.93 012 cells g–1 dry wt. (White et al., 1979); and the average molecular weights of polar lipid derived fatty acids and diether lipid are 270 and 740 Da, respectively. This conversion assumes that bacteria contain a constant proportion of their biomass as phospholipids under natural conditions, and the approach has been verified by Balkwill et al. (1988). The detection limit for Archaea under the procedures employed in this study was estimated to be 13 108 cells ml–1. DGGE analysis
DNA was extracted from duplicate samples of leachate. PCR products were generated from the DNA using broad specificity (universal) primers, 530f and 907r. The products were applied to DGGE gels. Methods for the DGGE analysis (including PCR and DNA extraction methodology) are reported elsewhere (Silvey, 1998). Results and discussion Process performance
The total methane produced during operation of the new reactor was 5.3 m3. Given that the starting weight of the waste was 37 kg VS (volatile solids), the total methane yield was
Figure 2 Methane production and pH fluctuations during treatment of the MSW
11
ws/41(3)/A2
13/4/00
1:42 pm
Page 12
P. Silvey et al.
approximately 0.15 m3 kg–1 VS. The BMP calculated for the unsorted MSW used in the current study was 0.19 m3 kg–1 VS. Therefore we estimated that close to 75% of the anaerobically degradable organic matter in the reactors was converted to methane. Sequencing leachate between new and mature reactors enabled rapid start up and balanced degradation in the new reactor. In this reactor, daily methane yield was
