Results and discussion


In this study we soughtmicro-scale anaerobic or aerobic composting as a solution to irregular removal of decomposable solid wastes that are generated at specialized locations such as street side fruit and fruit juice vendors, food vendors using decomposable paper and agro-residues in packing, as also large gardens and parks, etc. Rapid decomposition by composting or biomethanation processes for such specialized wastes is a potential decentralized solution to maintaining good sanitation with cost recovery. In addition, it is important that the process is amenable for operation at micro-scale of 130 kg VS/day. We characterize feedstocks for this potential.

The changes in temperature of digesting mass and moisture held as well as the mass of decomposable components are presented in Fig. 1. Fruit waste, paddy straw and bagasse showed an initial rise in temperature that may be attributed to the initiation of rapid aerobic decomposing bacteria as well as the presence of required components to achieve this. The temperature rise is rapid reaching between levels between 10 and 22). There is thus a potential to achieve high temperatures needed to sanitize the waste by arranging the shape of microcomposters to have low heat loss. This strategy is not possible with leaf litter, copier paper and newsprint. While leaf litter and copier paper waste do not attract insect vectors, newsprint associated with fruit wastes or food wastes (often practiced by street vendors) will require a different strategy. The moisture held or that needs to be maintained by visual methods showed interesting results. Visual methods for moisture regulation appeared difficult for non-paper feeds. All these feedstocks lost moisture content rapidly during the early stages of composting. Fruit wastes lost moisture in the form of leachates while leaf litter, paddy straw and bagasse appear to lose moisture by evaporation only. For these three feedstocks moisture regulation by visual methods did not permit precise control as seen by highly fluctuating values for moisture content. In microcomposting, alternative strategies may be required to regulate moisture. Newsprint and copier paper did not show too much variation in moisture content. The pattern of degradation of TS, VS and a potential fit for the degradation of VS is presented in Fig. 2. In all feedstocks other than copier paper, the TS and VS degradation pattern followed each other. This suggests that bulk of the TS degraded

is a result of VS degradation.

Only paddy straw provided a perfect fit for exponential decay (y=e-kt. R2=0.93). Fruit wastes and leaf litter could reach decomposition levels >75%. In comparison to these bagasse, copier paper and newsprint decomposed to a lower extent. The decay among the all feedstocks followed a general pattern of exponential decay (Lopes et al. 2004). However, most deviations from a typical exponential fit arose at the point of initiation. Either the initiation of the decay was delayed or the initial decomposition was very rapid compared to the rest of the material. Thereby the fit did not begin at the point of origin. In the case of fruit waste, there is clearly a delay in the initiation of rapid decomposition. It is expected that much of initial decomposition leads to the generation of volatile fatty acids VFA) and

concomitant reduction in the pH levels of the fermenting mass. A similar pattern is also seen in the case of leaf litter decomposition. A rapid initial decomposition in the case of bagasse (VS loss) accompanied by a rapid rise in temperature at initial stages show that bagasse contains a significant fraction of easy to decompose substances. This accounts for the rapid initial decomposition and is possibly caused by inefficient extraction of sugars during the juice expulsion process. Juice extraction from fruit wastes is also expected to be less efficient leaving behind a significant portion of simple to decompose components in the waste. While bagasse is reasonably porous, decomposition of sugar not extracted occurs without concomitant accumulation of volatile fatty acids (VFA). In the case of paper residues the initiation of decomposition is slow and proceeds only up to an extent of around 50% or less. Thus it may be seen that recovery of compost or residues is the least in the case of fruit waste while newsprint and copier paper are expected to provide a larger fraction of compost like material.

The changes in the carbon and nitrogen proportion before and after the composting process are presented in Fig. 3. Only fruit waste and paddy straw showed significant reduction in the carbon content. This observation suggests that along with decomposition there is a translocation of mineral content which is lost from the composting mass. It is believed that this is either lost in leachate as recorded in the case of fruit waste or gone unrecorded in paddy straw and needs further investigations. Bagasse showed a small reduction in C content in the residual matter. In the other three feedstocks newsprint, copier paper and leaf litter there was only a very small change in the organic carbon fraction. Similarly the change in nitrogen content of the feedstock is presented in Fig. 3. All feedstocks other than bagasse showed a doubling of the N content. This phenomenon is attributable to firstly the break down of carbon fractions without concomitant volatilization of N as ammonia. Fruit wastes, straw and leaf litter made compost that is rich in nitrogen and is thus saleable. Bagasse produced compost that is only marginally useful. Supplementing newsprint and copier paper with N in the form of digested leaf material did not appreciably raise the measurable nitrogen content. The resulting compost also had a low N content. From this we suggest that the proportion of digested compost that needs to be mixed with paper wastes needs to be much higher than used in this study in case the decomposition needs to be hastened and the compost produced is to be sold (Table 2).

Only fruit wastes produced significant quantities of leachate such that it would be considered necessary to provide leachate treatment along with composting (Fig. 4). These values indicate that a total of 1,749 ml leachate in 20 days (c.150 ml/kg), 31.5 g COD (2.6 g COD/kg feed), 54 g TS=4.5 g TS/kg feed) for wet waste needs to be managed. The total solids measured in the leachate are much higher than the COD measured. This suggests the presence a large mineral content in the feedstock as visible from the ash content that accounts for nearly half the total solids measured in the leachate. The COD and BOD in the leachate rose to high levels and gradually fell after a 10-day period. However, this period is much smaller than recorded for landfills (Fan et al. 2006). There is thus a significant loss of COD/BOD lost in leachate. This explains to some extent the large change in organic C and nitrogen in digested fruit waste. It is clear from the above that in the case of fruit wastes the conventional windrow composting needs to be accompanied by treatment options for the leachate. A significant proportion of the leachate was released within 10 days after the initiation of composting process.

It was observed that only fruit wastes and leaf litter were found in proportions and quantities that suit its use on at-site for biogas generation. The BMP of fruit and vegetable wastes have been studied extensively and hence only BMP of leaf litter was determined and reported. The biological methane potential of collected leaf litter has not been determined extensively.

The fallen leaves undergo decomposition almost immediately. The biogas potential residual is therefore quite uncertain. Understanding the BMP of such material is tackled in the following way. Firstly the BMP of the major tree leaves contributing to the leaf litter has been determined in the laboratory for leaves harvested fresh and dried. Second the leaf litter collected was fed to a 6 m3 total volume plug flow biomass biogas plant (Jagadish et al. 1998) and the biogas production rates from this was determined by measuring the gas collected in an inverted drum. The BMP of leaf material is presented in Fig. 5. All freshly harvested leaves (except Mimusops) showed a BMP level >400 ml/g TS fed. Leaves of Ficus species and mixed leaf biomass showed a rapid conversion to gas and had high methane content. This suggested good decomposition pattern that is usually not inhibited by VFA accumulation within. The other four species showed a short period of reduced gas production between 1540 days and subsequently the gas production picked up along with increased methane content. These two suggest the potential for VFA flux induced inhibition of methanogenesis at higher feed rates and the need to ensure appropriate solutions to overcome these fluxes. The daily biogas production rate from an average daily feed rate of 6.27 kg of leaf litter is presented in Fig. 6. This showed that such operations required a start up time of about 30 days after which steady state was achieved. However, in place of about 1.8 m3 gas/day about 1.05 m3/day was recovered resulting in an average gas production of about 170 l/kg TS fed. From this it is observed that allowing leaf litter to accumulate on the ground for long and then subjecting them to anaerobic digestion loses about 50% of the gas production potential and nearly 60% of the BMP is not recoverable. To ensure higher recovery alternative management techniques need to be developed. Thus biogas from dried leaf litter collected and stored in open spaces could enable operation of anaerobic digesters for biogas and compost and is an attractive idea for managing garden wastes in many urban areas.