Description
This project explored the anaerobic co-digestion of food waste with two different co-substrates, thickened waste activated sludge (in the wet system) and decomposing municipal solid waste (in the dry system), simulating co-digestion in a water resource recovery facility (WRRF) and landfill or landfill bioreactor. Three types of food waste – protein, carbohydrate, and lipids – were prepared based on USDA consumption data. The objective of this study was to better understand the changes in methane yields as loading rate and substrate type are changed over time. The wet system was studied using fed-batch (semi-continuous) reactors, while the dry system was studied using batch reactors. In the wet system high protein reactors were able to withstand ammonia levels of 6000 mg/L and were the only food waste category that did not experience failure throughout the experiment. High carbohydrate and high lipids reactors were inhibited at lower organic loading rates than the protein reactors but were able to recover, and successfully converted the substrates at the previous organic loading rate. In the dry system, lipid substrates yielded the highest methane production while carbohydrates resulted in the fastest production of methane. Overall, the lipids reactors achieved the highest methane yield in both wet and dry system, 475 and 615 mL CH4/g-VS, respectively.
A second experiment identified mixtures of food wastes that resulted in maximum methane production in mixtures of food waste types. Twenty-one dry reactors were loaded with various food wastes in a unique simplex centroid mixture design. Three reactors served as controls, and 18 reactors were fed either carbohydrate, protein, or fat-rich (lipid) waste (6 reactors each type). Once methane generation passed peak production, one reactor from each category was destructively sampled for future microbial community analysis. The remaining reactors were fed a combination of 50% of their original food waste type and 50% of a new one. Again, once methane generation passed peak production, one reactor from each food waste combination was destructively sampled. The remaining reactors were fed an equal part mixture of the three food waste types. The protein-fed reactors exhibited a large reduction in peak production after being introduced to either a lipid or carbohydrate substrate. Lipid-fed reactors experienced a decrease in peak production rate but an increase in cumulative methane yield when introduced to carbohydrate waste. Carbohydrate-fed reactors increased cumulative yield when introduced to either protein or lipid waste. The statistical analysis of the cumulative methane yields found the lipid and protein mixture to be the most optimal for methane production. This study provides insight into adaptation to substrate changes and informs the development of operational procedures to optimize the AcD of food waste.
$0.00
