Results and discussion |
The potential to carry out decentralized processing of OFMSW using only door to door collection greatly averts the high cost and need for daily fossil fuel based MSW collection as is stipulated in national and state level MSW rules and guidelines in India. The OFMSW component in Bangalore’s MSW is reasonably high necessitating its rapid removal and processing. Here we consider that anaerobic digestion (AD) has the potential to provide two saleable byproducts biogas and anaerobic compost to make it more economic. Further using a simplified 2-stage system, as in SSB, will be a potential solution to decentralized AD where feedstocks vary widely both in mass and composition (Chanakya and Moletta 2005). It is thus important to establish and characterize that each of these subcomponents of MSW with respect to their amenability to be fermented under SSB conditions.
Decomposition under SSB conditions Mixed fruit wastes could be fed at high feed rates (2 g TS/l/d) without significantly affecting operation (Fig. 2). Fruit wastes rapidly underwent decomposition and only a small fraction was left behind after a fermentation period (Fig. 3). However, such a feed rate and >90% TS reduction did not commensurate with gas production. We noticed that pH levels in the digester were generally low and long-term operation gave rise to fungal growth in exposed digester liquid. This low pH (less than 5.5) implicates high VFA in digester liquid (VFA not measured) and thus incomplete conversion to gas. For this extent of TS loss in the reactor, the expected gas production rates need to be above 1.0 l/l/d. Over 75% decomposition occurred in 14 d. This high rate of decomposition and VFA accumulation suggests that SSB reactors for mixed fruit wastes need to be operated with a much deeper biomass bed to ensure complete conversion to biogas, efficient operation and high biogas recovery. It is clear that compost recovered from such reactors is going to extremely low and will be less than 10% of the original weight of mixed fruit waste added.
It appears possible to support high feed rates for paper mulberry (leaf litter) based SSB (2 g/l/d) but gas production pattern seen, like in the case of mixed fruit waste, did not commensurate with the feed rate (Fig. 2) or TS removed. The high TS removal suggests that just as in the case of fruit waste the generation of required methanogenic bed and its subsequent conversion to biogas is poor. SSB digesters designed for large fraction soft leaf feedstock (paper mulberry) require longer SRT such that the active methanogenic bed height is adequate to decompose the VFA during its downward passage. It is seen that 75% decomposition took 35 d. Here the rates of decomposition are as reported before (Chanakya et al. 1997, 1999) however, it is important to ensure that TS to gas conversion efficiency is improved. Just as in the case of mixed fruit waste the quantity of compost generated is expected to be low. Table 2 lists the decay pattern for five of the six feedstocks and the decay pattern is exponential (y=me-kt). In the case of bagasse, it was found that by not using the 7 d value of decomposition, the fit is much better and suggests that the delay in onset of decay if overcome, the decay process would be truly exponential.
Owing to the bulky nature of bagasse (BG), sugarcane trash (ST) and paddy straw (PS) the lab scale SSB bioreactors used here could not be fed beyond the 0.5 g TS/l/d. The conversion to gas is however good reaching values between 0.2–0.25 l/l/d volume efficiency in all these feedstocks. There is a gradual increase in gas production rates during the initial period but the gas production data is highly scattered and reasons for scatter are not clear at this stage. PS and ST decompose to the tune of 50% of TS and may be considered suitable feedstocks for this approach. Although exponential decay pattern is visible, the current fit does not adequately explain for the initial rapid decomposition stage in these two feedstocks. Similarly in the case of bagasse while there is a reasonably good fit with exponential decay pattern and yet it does not explain for the delay in the initiation of decomposition and TS loss. The causes for these deviations need further investigation and necessary corrections need to be made to use these decay rates in future. We expect that in larger reactors where the mass of feedstock will be higher, the compaction rates of feed would be high and it in turn would permit higher feed rates making SSB reactors acceptable for these feeds as well as economic and viable.
The case of photocopying paper is unusual. The gas production has gradually built up to reach a rate of 0.25 l/l/d after 50 d period. A feed rate of 1.5 gTS/ l/d could be sustained without running out of digester space as was the case with the dry feedstocks or high VFA build up as in case of FW mentioned above. From this gradual build up of biogas production rate it is clear that the decomposition rate under present conditions is low (Fig. 2). This has happened in spite of adding 10% digested leaf biomass to compensate for low nutrient status of paper. This suggested that the current level of nutrition supplement is either inadequate or is not the right approach to handle photocopier (PC) paper decomposition to biogas.
Biological methane potential and realizing BMP under SSB conditions The BMP of the six feedstocks along with the gas composition is presented in Fig. 4. All the BMP assay vials using 2% feedstock exhibited low biogas production with low CH4 content. It was clear that the methanogen populations were inadequate to handle the various inhibitory intermediates generated especially in the case of FW where the pH was low (5.2, data not shown). We therefore draw inferences only from BMP assay carried out at 1% substrate concentration. The BMP of fruit waste was similar to what was achieved under SSB conditions. Under SSB conditions it was possible to obtain a specific gas production of about 0.5 l/l/d at a 2 g/l/d feed rate. At nearly 90% VS/TS degradation as found in Fig. 3 it would result in =1.5 l/l/d at the above feed rate. While a large fraction of TS/VS was degraded it did not translate into gas, it was seen that the digester liquid contained a significant extent of fungal growth indicating among others, the presence of VFA that has flowed through the methanogenic bed without being converted to biogas. From this it is inferred that fruit waste based SSB biogas fermentors need to start with a deeper and more active methanogenic bed compared to what is used for other materials. The BMP was low, arrested by inadequate conversion of feedstock to methane and corresponding low methane content in the ensuing gas (Fig. 4).
Paper mulberry (leaf litter), bagasse, paddy straw and cane trash showed a BMP level of nearly 600 ml/gTS with gas production rates tapering off between 40 and 80 d of fermentation. The methane content of the gas reached =50% within a reasonable fermentation time (c.5–20 d). All these indicate that these feedstocks have the potential of being converted to biogas with reasonable gas yields and a stable process not easily afflicted by VFA overproduction is possible under SSB conditions. However, the efficiency of conversion under SSB conditions (Fig. 3) has not been commensurate with results of the BMP assay (Fig. 4). The average gas production at pseudo-steady state operation (Chanakya et al. 1999) works out to be 200, 380, 380 and 360 ml/g TS for paper mulberry, bagasse, paddy straw and cane trash respectively. This accounts for realization of 60% of BMP of dry feedstocks and only 30% of the BMP of paper mulberry. Dry feedstocks decompose slowly (Figs. 3 and 4) and therefore are not easily subject to VFA flux induced methanogen toxicity. Thus they can produce reasonable quantities of biogas with high methane content compared to paper mulberry (leaf litter) under the current set up of SSB reactors. This evidence suggests that as found in the case of fruit waste, the methanogenic bed at the lower part of the SSB reactor did not function adequately to convert VFA leaching downwards to biogas. Thus good VS destruction achieved did not result in good gas production. In solid waste management (SWM) terms when SSB reactors are operated with such feedstocks, it is necessary to ensure that the start-up methanogenic bed is acclimatized to high VFA flux before being operated at full loading rates. Photocopier paper when augmented with 10% digested biomass does not stabilize and reach pseudo-steady state (Fig. 2). The highest gas production is also poor (c.140 ml/g TS) nor did the decomposition follow a predictable pattern. In BMP assay the gas production tapered off at around 40 d and reached a little above 400 ml/g TS. The reasons for such low conversion and gas production rate for a feedstock that is expected to be largely cellulose is not clear from available data and needs to be investigated.
At longer SRT, greater methanization takes place in the hydrolyser reactors than in methanizer, because VFA are at a lower concentration, not inhibitory to methanogens and thus phase separation is not very effective. At low organic loads, VFA concentrations are poor, because they are immediately converted and the yields (l/g VS) are high. Thus in order to enhance yield in a two-phase system, many more investigations are needed with respect to the system set-up, the control of pH in both reactor, etc. Only then it is possible to optimize conditions in the hydrolyser and methanizer. Thus a single phase system would be the appropriate choice as it is simpler, can be applied successfully to the treatment of this type of waste without any type of process control while simultaneously using the concepts of co-digestion (Callaghan et al. 2002). It also permits mesophilic operation at 25°C (Castillo et al. 1995) to avail of steady yield and economic operation (low heating needs).
Reuse of digested feedstock as methanogen biofilm for wastewater treatment Chanakya et al. 1998, 2004, showed that digested herbaceous biomass forms a precolonized biofilm support for carrying out high rate wastewater treatment. Green leaf biomass however had a short half life of 120 d as support after which it would disintegrate and was lost in the effluent. This suggests that fibrous biomass feedstocks, such as bagasse, paddy straw, leaf litter, cane trash, paper, etc. can be first digested for their energy recovery through SSB mode of biomethanation in decentralized MSW biogas plants. Later the spent residue removed from these digesters could be sold as ready to use biofilm support for anaerobic digestion of wastewater. It was found that only bagasse fed SSB reactors had adequate digested biomass for further trials as biofilm support. Bagasse with 25% digested leaf biomass was used as biofilm support in a down flow reactor. Bagasse with biomass showed good performance measured as the difference between inlet and outlet COD. At feed rates up to 1,000 mg/l it has shown around 50% conversion (data not shown). This reuse potential leading to zero discharge needs to be examined further.
