Our lab has experience with treatment of swine waste with ASBR technology, and therefore we will use this technology as an example of a high-rate anaerobic digester system on the farm (Garcia and Angenent 2009). The ASBR is a single-vessel system that requires no feed — distribution and gas-solids-separation systems, which simplify its design (Figure 4.2). However, intermittent mixing is needed to provide sufficient contact between substrate and biomass, distribute heat, and prevent scaling. A settling step before decanting effluent is instrumental to facilitate high biomass levels and long sludge retention times (internal settling; Sung and Dague 1995). Developing a well-settling biomass, while not disturbing bioflocculation, was accomplished by mixing ASBRs intermittently (e. g., 2 minutes of mixing every hour for a laboratory ~ scale bioreactor; Sung and Dague 1995; Zhang et al. 1997; . Development of a well-settling biomass from poorly settling inocula, however, requires long startup periods. Several laboratory; scale studies and one full ; scale study have been per­formed to investigate swine waste treatment with ASBRs (Dague and Pidaparti 1992; Zhang et al. 1997; Angenent et al. 2002b, 2008). Indeed, for efficient swine waste treatment at the design volumetric loading rates, the start-up period was longer than 6 months.

The choice of inoculum influences the start-up period of ASBRs. This became apparent in a study with two 5-L ASBRs that treated diluted swine waste with a VS level of 20 gVS/L (Angenent et al. 2002a). When swine lagoon sludge with a very well-settling ability (sludge volume index [SVI] < 10mL/g total solids [TS]) was used as an inoculum, the period to achieve a design VSLR of 4 g/L/day was 250 days due to the fact that biomass levels in the bioreactor remained around 20gVS/L during the initial 50 days of the operating period (Figure 4.3a, b). The operating period to achieve this design loading rate was 100 days longer when poorly settling anaerobic digester sludge from a CSTR treating waste activated sludge was used as an inoculum (SVI was ~56mL/gTS). Biomass washout in the ASBR with this sludge decreased the biomass levels in the reactor from 20 to 10gVS/L during the first 50 days of the operating period (Figure 4.3b). At the end of the study after a 1-year operating period, the settleability of the biomass in both reactors was identical (17mL/gTS), showing that ASBRs select for a well-settling biomass (Angenent et al. 2002a).

Besides developing a well-settling biomass, the concentration of methanogens in the reactor biomass may also need to be increased to obtain high volumetric loading rates. Timur and Ozturk (1999) reported that an ASBR showed favorable performance during the treatment of landfill leachate. In addition, these researchers reported that the volumetric methane pro­duction rate (VMPR) increased linearly with the volumetric loading rate to achieve a constant methane yield over the operating period during which a stable performance was reported (Timur and Ozturk 1999), which we also observed for animal wastes (Angenent et al. 2002b; Angenent et al. 2008; Hoffmann et al. 2008; Garcia and Angenent 2009). Timur and Ozturk (1999) observed an enrichment of methanogens over the operating period, with an increasing specific methane production rate (in L of methane per g volatile suspended solids [VSS] per day) to keep up with the increase in the food to microorganism (F/M) ratio. This showed that the biomass in the ASBR became more active in terms of methane production per amount of biomass present over the operating time and that a considerable start-up period was required before the ASBR could manage volumetric loading rates of up to 10 gCOD/L/day (Timur and Ozturk 1997) . We have shown higher concentrations of methanogenic small — subunit ribosomal RNA (rRNA) in ASBRs treating swine waste at the higher VMPR under stable reactor performance compared with the lower methane production rate (Figure 4.4). Therefore, anaerobic digesters select for a biomass that is more active in methane production to sustain higher loading rates (Angenent et al. 2002a).

In summary, both the development of a well-settling biomass and an increase in the con­centration of methanogens necessitates a long start-up period for the ASBR. In our work with ASBR treatment of diluted swine waste at 25-35°C, the investment of a relatively long start­up period, however, ensured an operation at a much shorter HRT of 5 days compared with a 15-day HRT for low-rate anaerobic digesters (Angenent et al. 2008; Garcia and Angenent 2009). This shorter HRT directly results in a 66% reduction in reactor volume. At the rela­tively low HRT of 5 days for a 5-L ASBR treating swine waste at a concentration of 2 gVS/L and 25°C, the methane yield was 0.47L methane per g VS fed with an ~52% VS removal efficiency at a VSLR of 4 g/L/day (Angenent et al. 2008). This yield was close to the ultimate methane yield (i. e., methane yield at an infinite HRT), which was estimated to be between 0.44 and 0.52L CH4/g VS fed for waste from swine fed a corn-based diet (Chen 1983), and shows that the ASBR performance is suitable even under relatively low HRTs.

In a subsequent study, we operated the same ASBRs under similar environmental condi­tions, but with swine waste from a different farm. We achieved a lower methane yield of 0.31 L methane/g VS fed (Garcia and Angenent 2009), and this was not caused by an inferior ASBR performance compared with the study described earlier (a 51% and 52% VS removal efficiency at a VSLR of 4 g/L/day for the Angenent et al. [2008] and Garcia and Angenent [2009] studies, respectively). Instead, the large difference in methane yields for both studies was probably due to variations in the composition of swine waste. This shows that it is very difficult to predict the methane yields due to the large variations in conditions at the

farm, including farm practices, such as animal feed composition, antibiotic use, manure collection, manure storage period before transfer into the digester, water consumption, and cleaning habits.