The goal of this study is to develop a process for energy recovery from the treatment of food wastes which constitute about 30% of the total municipal solid waste (MSW) in Korea. In order to develop an anaerobic digestion process for Korean food wastes containing 15~30% total solids (TS) the biochemical methane potentials (BMP) of food wastes and their components (cooked meat, boiled rice, fresh cabbage) were measured. The methane yields of cooked meat, cellulose (as control), boiled rice, fresh cabbage, bibimbab (Korean food based on boiled rice mixed with vegetables, meat and eggs) and 10 food waste samples were 482, 356, 294, 277, 472 and 340-440 mL $CH_4$/gVS added and anaerobic biodegradabilities based on the stoichiometric methane yield were 0.82, 0.92, 0.72, 0.73, 0.86 and above 0.88, respectively.
To elucidate the digestion mechanism for solid state anaerobic digestion of food wastes discharged at 15~30% TS, digestion tests with various initial organic loading in a single reactor were carried out. Since the food wastes contain much soluble organics at a high organic loading (>5%TS), they were converted rapidly to volatile fatty acids (VFAs) at an early stage of digestion, and VFAs produced were accumulated so excessively to inhibit subsequent methane fermentation. Therefore, in order to digest the solid state food wastes properly, VFAs produced rapidly at an early stage should be fed to a separate reactor which maintains good conditions of methane fermentation using two-phase anaerobic digestion system, and the effluent of the methane reactor was recirculated to provide buffer capacity in the solubilization/acidification (S/A) reactor.
Bibimbab with 20 % TS was digested using a lab scale two-phase anaerobic digestion systems of 1 L and 8 L methane reactors in a batch mode. In the 8L methane reactor system also the effect of recirculation was studied. No nutrients, buffer and seed inoculum were added in the 5 L S/A reactor. In all cases, 87~90% of TS in S/A reactor was reduced and approximately 90% of the initial bed volatile solid (VS) was converted to biogas. The treatment periods for the complete digestion of waste depended mostly upon the performances of the methane fermenter, especially on hydraulic retention time (HRT) and loading rates. The methane yield was around 405~415mL/gVS reduced during entire periods. This value is about 86~88% of the BMP test data (the ultimate methane yield) with the same sample.
And also, the actual food wastes discharged at a restaurant were treated in a batch mode, fed batch mode and continuous mode using a mini pilot scale two-phase anaerobic digestion system which consists of a 70 L S/A reactor and 200 L methane reactor. In a 30 days of batch experiment for the treatment of 50kg food wastes, the methane yield was around 367mL/gVS added, and this value was about 86% of the BMP test data with the same sample. In 35 days of fed batch experiment for the treatment of 128kg wastes, the methane yield was lower than that of batch experiment, but the treatment rate was about two times. In continuous treatment of food wastes at a loading of 5, 7.5, 10, 15kg/d, the methane production rate on the basis of total reactor (S/A reactor + methane reactor) volume was around 1.6-1.7, 2.1-2.4, 2.7-3.0, 3.2-3.5 L/L.d (vvd), and the average efficiencies of VS reduction were 91, 87, 84, 69%, respectively. Clear phase separation was obtained in continuous anaerobic digestion because the total VFAs ($C_2$ - $C_6$) levels exceeded 7g/L and the methane contents was maintained below 25% in S/A reactor during entire periods.