Improved anaerobic digester stability to organic loading rate shocks with the use of an environmentally derived inoculum
Open Access
- Author:
- Steinberg, Lisa Marie
- Graduate Program:
- Environmental Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- February 12, 2009
- Committee Members:
- John Michael Regan, Dissertation Advisor/Co-Advisor
John Michael Regan, Committee Chair/Co-Chair
Christopher Howard House, Committee Member
Bruce Ernest Logan, Committee Member
Jenn Macalady, Committee Member - Keywords:
- methyl coenzyme M reductase
pH
shock loading
methanogen
anaerobic digestion - Abstract:
- Anaerobic digestion broadly describes technology that utilizes microorganisms to break down organic matter under anaerobic conditions through the coordinated efforts of several trophic groups of microorganisms. The last step is catalyzed by methanogens which produce primarily methane, carbon dioxide, and water as products of metabolism. Anaerobic digestion occurs naturally in a variety of water-saturated sediments, but is also used to treat waste in constructed reactors. There are a number of advantages to treating waste with anaerobic digestion, but perhaps the greatest is that waste treatment can be coupled to energy generation by the production of a methane-rich biogas. Despite the advantages, anaerobic digestion is severely under-utilized in waste treatment mainly due to the belief that anaerobic digesters are less stable than aerobic treatment processes. Anaerobic digesters are typically operated under warm temperatures and circumneutral pH, with operation outside of these conditions leading to instability and potential reactor failure. Reactor failure is expensive and time-consuming to recover, and has mainly been attributed to the sensitivity of the methanogens in the microbial consortium. The most often-cited reason for digester failure is acidic pH which may result from overfeeding of the reactor leading to the buildup of fermentation by-products. In contrast to constructed reactors, methanogenesis in acidic wetlands proceeds at pH values less than 4. Based on previous findings, there were two hypotheses driving this study. The first was that the methanogens present in an acidic wetland would be distinctly different than methanogens found in a constructed anaerobic reactor. The second hypothesis was that a reactor inoculated from an acidic peatland would be more tolerant of sudden increases in feeding rate, and the resultant drop in pH, than a reactor inoculated from a traditional anaerobic digester. To investigate the first hypothesis, the methanogens in a local oligotrophic, acidic fen known as Bear Meadows Bog and a mesophilic anaerobic digester treating municipal wastewater sludge at the Pennsylvania State University Wastewater Treatment Plant were analyzed with clone libraries. Two clone libraries were created for each environment, one for the 16S rDNA and a second for the methyl coenzyme M reductase alpha subunit gene (mcrA). The Mcr is unique to methanogens and shows similar phylogeny as the 16S rDNA so it may be used independently of, or in conjunction with, the 16S rDNA to determine methanogen taxonomy. A quantitative framework was developed to assess the differences between these two communities by calculating the average sequence similarity for 16S rDNA and mcrA within a genus and family using sequences of isolated and characterized methanogens and an established methanogen taxonomy. The average sequence similarities for 16S rDNA within a genus and family were 96.0 and 93.5%, respectively, and the average sequence similarities for mcrA within a genus and family were 88.9 and 79%, respectively. The clone libraries of the bog and digester environments showed no overlap at the species and genus levels, and almost no overlap at the family level. Both libraries were dominated by clones related to uncultured methanogen groups within the Methanomicrobiales, specifically the Fen Cluster in the bog and MCR-7 in the digester. Despite differences in the sequences detected, diversity was similar in both environments. Based on the results of the clone libraries, real-time quantitative PCR (qPCR) methods were developed to distinguish and quantify members of different methanogen clades. Biases inherent in PCR prevent inference of an organism’s environmental abundance based on end-point PCR product abundance, but qPCR overcomes some of these biases. A SYBR-Green I qPCR assay was developed to quantify total numbers of mcrA genes, and TaqMan probes were also designed to target several different phylogenetic groups. Groups which were targeted by TaqMan probes included Methanosaetaceae, Methanosarcina, Methanocorpusculaceae, Methanospirillaceae, Methanobacteriaceae, and the environmentally-defined clades Fen Cluster, MCR-7, and MCR-2 subgroups a and b. Many of these groups represent largely uncultured clades whose taxonomic standing has yet to be determined. Both SYBR-Green I and TaqMan qPCR assays showed good reproducibility and specificity for their methanogen targets. These methods were further tested by determining total mcrA and mcrA of different methanogen groups from six samples: four samples from anaerobic digesters treating either primarily cow or pig manure, and two aliquots from an acidic peat sample stored at 4°C or 20°C. The three samples obtained from cow manure digesters were dominated by members of the genus Methanosarcina, whereas the sample from the pig manure digester contained detectable levels of only members of the Methanobacteriaceae. The acidic peat samples were dominated by both Methanosarcina and members of the Fen cluster. In two of the manure digester samples only one methanogen group was detected, but in both of the acidic peat samples and two of the manure digester samples, multiple methanogen groups were detected. To investigate the second hypothesis of this study, lab-scale reactors were inoculated with either bog sediments (Bog reactors,) wasted sludge from the anaerobic digester (Digester reactors), or with a mixture of these two inocula (Hybrid reactors). Reactors were maintained at 30°C and operated in semi-batch mode with feeding and wasting every four days, and gentle mixing once daily. Digester performance was monitored chemically and microbiologically using the previously designed qPCR assays. After the establishment of active methanogenic cultures in all reactors, a set of test reactors were inoculated from the control reactors and used to examine the effects of periodic doses of glucose to the consortia derived from different inocula. A total of three sets of organic shocks were delivered. Glucose pulses were equal to 1-10 times the normal amount of volatile carbon in a feeding, and these resulted in the production of large amounts of volatile fatty acids which reduced reactor pH to less than 5. The first shock was 10 g of glucose, and this caused the Digester- and Hybrid-inoculated test reactors to fail, but the Bog-inoculated reactor recovered. After re-inoculating the Digester- and Hybrid-test reactors, three successive glucose pulses of 1, 5, and 10 g of glucose were delivered. The Hybrid-test reactor failed after receiving the 5 g glucose pulse, but the Bog- and Digester-test reactors only failed after receiving the 10 g glucose pulse. All reactors were recovered by the addition of NaOH to raise reactor pH to neutral levels. A second set of test reactors was established from the first set, and a third glucose pulse of 10 g was delivered. One set of reactors received mineral nutrients with the glucose pulse, but the second set did not. All reactors failed, but the Bog- and Digester-inoculated reactors recovered after pH was raised to only 4.8-5. The Hybrid-inoculated reactor did not show signs of recovery until the end of the testing period. Reactors with the same inoculum acted similarly, suggesting mineral nutrients did not prevent reactor failure. Originally the Bog-inoculated test reactor was dominated by members of the Fen Cluster, but this clade only persisted through the second set of glucose shocks. The Fen Cluster was detected in the Hybrid-test reactor after glucose pulses, but was never detected in the control reactor. Methanosarcina and Methanobacteriaceae came to quickly dominate all control and test reactors, and the presence of Methanosarcina correlated with reactor recovery after glucose pulses. The control reactors showed less diversity than test reactors, and all control reactors were dominated by Methanosarcina with fewer numbers of Methanobacteriaceae and MCR-7. Although distinct differences existed in the methanogen communities from the bog and digesters, these differences could not be maintained in constructed reactors. The methanogen communities in control reactors were less diverse than those in test reactors, and the acidophilic Fen Cluster was detected when reactor pH was acidic. The test reactors showed greater resilience to low pH conditions than has been previously reported in the literature, and reactors were able to recover activity after a glucose pulse from a pH as low as 4.8-5. Further research is needed to understand the role that different methanogen groups, specifically the Fen Cluster and Methanosarcina, play during periods of acidic operation, and how their presence may be maintained and their activity enhanced in order to provide greater stability to anaerobic digesters.