Evaluation of Multiple Feedstocks for Codigestion
In the U.S., a move toward the use of anaerobic digestion (AD) technologies for solid waste management has been slow. However, recently, a number of factors have coalesced to renew interest in AD because of its potential to produce renewable energy from the wastes. It is well known among AD practitioners that every waste material is unique, and each must be evaluated on a case-by-case basis for a particular digester project. A review of existing literature revealed that the guidance provided by it was not of a quality that would promote and assist with rapid U.S. AD development and advances. The aim of this research was to address some of these deficits and provide some new information to inform feedstock testing reliability and reproducibility. A database of codigestion articles published between 2000 and 2014 was compiled to examine the nature and quality of existing literature and also new researchers to filter research articles based on a variety of criteria that include feedstocks, operational parameters, and the types of information reported. The database is expandable and available for hosting on a website. Database analysis revealed that 32% of the authors measured biogas production but not methane (CH4) output specifically. Only 27% of the studies used food to microorganism (F:M) ratio as an operational parameter, and 8% reported both F:M and the nutrients expressed as carbon-to-nitrogen (C:N) ratio. The batch codigestion experiments revealed that even in the absence of acclimated sludge or alkalinity supplementation, poultry litter (PL) and DAF in equal weight percent loadings proved to be stable cosubstrates. Further, mixes in ratios of PL60:FW15 (poultry litter: food waste) with the remainder brown grease (BG) or DAF were also successful despite the fact that FW and DAF failed in the biochemical methane potential (BMP) and anaerobic toxicity assays (ATA) tests due to acidification. The BG was only mildly inhibitory in ATA testing and performed well in the BMP test with alkalinity supplementation. All batch, BMP and ATA experiments showed glycerin (GLY) and/or canola seed hull cake inhibited CH4 production. Although acidification was implicated in the canola ATA, it was not the cause of failure in the BMP test or in either GLY test; propionate accumulation or toxins may have been responsible. Methane yields in the BMP tests showed BG to be the most productive (371±76 mL CH4/gVS), and paper, PL, and cattle manure (CM) had yields in the 120-150 mL CH4/gVS range. PL and CM were stimulatory in the ATAs, with all other feedstocks showing varying degrees of inhibition. This finding is interesting in light of the fact that feedstock cell counts showed that PL (as well as CM and DAF) contain more live cells per gram volatile solids than the seed used as inoculum. The semi-continuous reactors demonstrated that a mix with up to eight feedstocks could be managed in a stable digestion; however, mixes with lipid-laden feedstocks, high organic loading rates or short solids retention time led to foaming and fouling of two of the reactors. DAF exposed to thermal pretreatment produced 170±22 mL CH4/gVS, while untreated DAF yielded no CH4. Similarly pretreated CM showed a two-fold increase in CH4 yield, but the same was not true for sewage sludge, where pretreatment inhibited CH4 production. Thermal pretreatment of PL had no effect on CH4 production, but, along with CM and DAF, it had a positive net energy balance. The energy analysis based on pretreatment studies reported in the literature revealed that chemical and biological pretreatment were the only methods that reliably yielded a net energy gain. Mechanical, thermal, and thermochemical pretreatment were less successful at yielding net positive energy values. Taken together, this body of work offers a roadmap for codigestion research. Ready access to recent literature is provided along with guidance about the important procedural, operational and reporting features required for sound study. A reproducible batch protocol is described that includes attention to nutrient and inoculant balance, to gas collection and analysis, and to the use of controls. If replicated, it will allow for better comparisons among laboratories. The importance of replication of semi-continuous or continuous flow studies is highlighted, due to the inherent variability between reactors. A discussion of the relevance and application of the BMP and ATA tests is offered, and data from some novel feedstocks is reported. Feedstock cell counts suggest that F:M ratios for some substrates need to be adjusted to account for live cells entering with the feed.