Materials and methods |
The feedstocks used in this study came from various points within the city of Bangalore. Several individual types of fruits wastes were collected within 4–5 h of generation from nearby fresh fruit juice vendors who stored these fruit wastes separately for our study. Feedstocks were collected fresh once a week at the generation site and fed in the fresh state. For studies on biological methane potential (BMP) assay, individual types of fruit wastes were collected separately and subject to assay and physico-chemical analyses. Paddy straw and cane trash are major packing materials for fruits and furniture in urban areas. They are normally recycled as cattle feed but are also found in large quantities in OFMSW of areas manufacturing or trading with steel furniture, glass, etc. Paddy straw and cane trash were picked up from such locations. Tree leaf litter is an important component of avenues where fallen litter of the nearby gardens or roads is swept frequently. It occurs in large quantities during months of peak litter fall. Bangalore has a very large diversity of trees and litter collected is usually mixed (Jagadish et al. 1998; Sathishkumar et al. 2001). In this study we have used mature leaves of one tree species Brousenetia papyrifera (paper mulberry) that is found in large numbers in the vicinity of the laboratory – it was thus a convenient feedstock of the SSB as well as BMP studies.
Solid-state stratified bed (SSB) reactor (Fig. 1) made of opaque PVC pipes (Chanakya et al. 1998) were used to characterize decomposition of various feedstocks. These reactors consist of 0.6 m tall, 0.15 m dia PVC pipes placed in a pool of recycled digester liquid. The top of the pipes was modified to hold a water jacket such that the top cover could be opened out to put in the feed. Under this cover was placed a perforated plate that enabled the recycled liquid to be spread and sprinkled over the bed of decomposing biomass below. Once a week, the top hatch was lifted and a nylon mesh was placed over the existing bed of decomposing biomass and then a pre-determined quantity of waste was fed to them by placing it on the existing bed and closing the hatch. No preprocessing was adopted for any of the feedstocks. This operation lasted about 5 min in a week. We adopted a weekly fed operation to minimize interference by air inadvertently entering the reactor during the feeding operation. A total of six such reactors were operated for 70 d. The process was initiated by placing a 0.15 m depth of digested biomass in each digester. Digested biomass was taken from a 6 m3/d SSB reactor and it served as a methanogen rich bed to initiate the process. A total of 3 l of digester liquid from the large biomass reactor was used as the recycling liquid in each reactor. The liquid in lower reservoir (1.5 l) was manually recycled twice daily at 0900 and 1630 h. The gas produced in this reactor was led to a 0.6 m tall, 100 mm diameter PVC pipe and collected by downward displacement of air. When required, a sample of the gas was drawn with a micro litre syringe and analyzed for gas composition. The gas production was recorded twice daily. The feedstocks were initially fed at 0.5 g TS/l/d and feed rates were gradually raised. An attempt was made to reach pseudo steady state operation at 2 gTS/l/d (14 g TS fed once a week). In a few feedstocks however, this could not be achieved and hence were continued at 0.5 g TS/l/d feed rates.
After a digestion period of 70 d the fermenting mass was removed in the form of layers of decomposing mass subject to increasing solids retention time (SRT). Each layer was characterized for its residual TS/VS, moisture, mainly to determine the extent of decomposition it had suffered. Biomass such as fruit waste, paper mulberry, were extremely decomposed and/or pulpy and considered unsuitable for further re-use as biofilm support. Only fermented bagasse was recovered in adequate quantities and form after SSB digestion. We operated a reactor with bagasse+20% leaf biomass as biofilm support. Biofilm biomass was packed at 200 g fresh weight/l. This downflow reactor was operated by feeding synthetic feed (rice flour dispersed in hot water, cooled and diluted to required strength). The gas produced was connected to downward displacement type storage and measured on a daily basis. The COD of wastewater at inlet and outlet was determined and used to indicate the satisfactory functioning of the biofilm support. The composition of biogas collected was monitored at frequent intervals to determine the health of the reactor. This was tried to determine the possibility of co-fermentation of solid and liquid wastes in a single fermentor configuration (Chanakya et al. 1998).
The BMP assay was carried out in 133 ml serum vials with 3–6 replicates and fermented at 35±1°C in an incubator. Each vial contained 49.5 or 49 ml of methanogen enriched digester liquid and 0.5 or 1.0 g TS equivalent of specific feedstock. This gave a final concentration of 1 or 2% TS in the fermenting liquid. All vials containing 2%TS showed stalled fermentation and were subsequently not reported in this study. The gas production was measured by downward displacement of water in an inverted burette at progressively increasing intervals of time e.g. on days 1, 3, 6, 9, 12, 16, 20, 27, 35, 49, 63, 79, 100, etc. These intervals were standardized from previous experience with leaf biomass feedstocks. On all days when gas production volume was estimated, the gas composition was also measured using gas chromatograph connected to a thermal conductivity detector with H2 as carrier gas. From the volume and composition we estimated the total CH4 and CO2 produced. BMP assay was carried out as a reference to determine the potential of the SSB reactor to extract maximum biogas potential of the feedstock.
The total solids (TS) and volatile solids (VS) was determined for the feedstock mass residual in the SSB reactors which was removed after the 70 d fermentation period by simply lifting the reactor out of the digester liquid and gently pushing out the fermenting mass (bed) within. This mass was lifted out layer by layer (separated by nylon mesh) and individual layers representing various solid retention times (SRT) were weighed and sampled for TS/VS estimation. The ratio of the residual mass was expressed as per cent of original fed in order to overcome varying feed rates and attempts were made to fit an exponential decay pattern for each of the feedstock tried.
