Abstract

The effect of fish waste (FW), abattoir wastewater (AW) and waste activated sludge (WAS) addition as co-substrates on the fruit and vegetable waste (FVW) anaerobic digestion performance was investigated under mesophilic conditions using four anaerobic sequencing batch reactors (ASBR) with the aim of finding the better co-substrate for the enhanced performance of co-digestion. The reactors were operated at an organic loading rate of 2.46–2.51 g volatile solids (VS) l⁻¹ d⁻¹, of which approximately 90% were from FVW, and a hydraulic retention time of 10 days. It was observed that AW and WAS additions with a ratio of 10% VS enhanced biogas yield by 51.5% and 43.8% and total volatile solids removal by 10% and 11.7%, respectively. However FW addition led to improvement of the process stability, as indicated by the low VFAs/Alkalinity ratio of 0.28, and permitted anaerobic digestion of FVW without chemical alkali addition. Despite a considerable decrease in the C/N ratio from 34.2 to 27.6, the addition of FW slightly improved the gas production yield (8.1%) compared to anaerobic digestion of FVW alone. A C/N ratio between 22 and 25 seemed to be better for anaerobic co-digestion of FVW with its co-substrates. The most significant factor for enhanced FVW digestion performance was the improved organic nitrogen content provided by the additional wastes. Consequently, the occurrence of an imbalance between the different groups of anaerobic bacteria which may take place in unstable anaerobic digestion of FVW could be prevented.

1. Introduction

The easy biodegradable organic matter content of FVW with high moisture facilitates their biological treatment and shows the trend of these wastes for anaerobic digestion (Bouallagui et al., 2003). In general, hydrolysis is the rate limiting step if the substrate is in particulate form (Veeken and Hamelers, 1999). However, the anaerobic degradation of cellulose-poor wastes like FVW is limited by methanogenesis rather than by the hydrolysis. A major limitation of anaerobic digestion of FVW is a rapid acidification of these wastes decreasing the pH in the reactor, and a larger volatile fatty acids production, which stress and inhibit the activity of methanogenic bacteria ([Misi and Forster, 2001] and [Bouallagui et al., 2005]).

The addition of co-substrates with high nitrogen content is a solution to adjust nutrient content of FVW. The unbalanced nutrients of fish waste (FW), abattoir wastewater (AW) and waste activated sludge (WAS) characterised by a low C/N ratio were also regarded as an important limitation factor to anaerobic digestion of these organic wastes ([Mshandete et al., 2004], [Gomez et al., 2006] and [Gannoun et al., 2007]). Adding AW, FW and WAS in FVW feedstock to have a balanced C/N ratio was undertaken in this study. Their greatest advantage lies in the buffering of the organic loading rate, and anaerobic ammonia production from organic nitrogen, which reduce the FVW anaerobic digestion limitations.

Co-digestion is a technology that is increasingly being applied for simultaneous treatment of several solid and liquid organic wastes ([Poggi-Varaldo et al., 1997], [Callaghan et al., 1999], [Alatriste et al., 2006] and [Perez et al., 2006]). It combines different organic substrates to generate a homogeneous mixture as input to the anaerobic reactor in order to increase process performance ([Hamzawi et al., 1998], [Viotti et al., 2004] and [Zhang and Banks, 2008]). It permits the exploitation of complementarity in waste characteristics e.g. avoidance of nutrients (N, P) addition when a co-digested waste contains nutrients in excess ([Gavala et al., 1996], [Pavan et al., 2005] and [Neves et al., 2008]). Several studies have shown that multi-component mixtures of agro-wastes, rural wastes and industrial wastes can be digested successfully, although with some mixtures a degree of both synergism and antagonism occurred ([Misi and Forster, 2001], [Misi and Forster, 2002] and [Cavinato et al., 2008]).

The aim of this work was to examine the effect of AW, WAS and FW addition as co-substrates on the FVW anaerobic digestion performance in mesophilic condition using an ASBR. The criteria for judging the success of a co-digestion were process stability, VS reduction, biogas production rate, and methane yield.

2. Material and methods

2.1. Reactors design and operational conditions

Four laboratory-scale anaerobic sequencing batch reactors (R1, R2, R3 and R4) of 2 l effective volume were used (Fig. 1). The temperature was controlled at 35 °C by a thermostatically regulated water bath. Peristaltic pumps were used to fill the reactors and to draw off the effluents after settling. Mixing in the reactors was done by a system of magnetic stirring. Each digester was initially inoculated with anaerobic sludge obtained from an active mesophilic digester of FVWs treatment plant (Bouallagui et al., 2007).

The ASBR was operated with cycles including the following four discrete steps: (i) fill (15 min): 200 ml of different mixtures of wastes were added to the reactors at the beginning of a cycle, (ii) react (21 h): during this phase, the reactors were stirred and organic matter was converted to energy and new cells, (iii) settle (2 h and 30 min): settling started when the react phase was finished and (iv) draw off (15 min): at the end of the settling period, the volume of liquid added at the beginning of the cycle was drawn off from the reactors.

2.2. Wastes sources and characteristics

The fruit, vegetable and fish wastes used in this study were collected from the group market of Tunis. After shredding to small particles and homogenizing, they were stored at 4 °C. WAS was collected from the activated sludge plant (Cherguia, Tunis) treating domestic and industrial wastewaters. It is composed of settled suspended biomass. The AW was collected from an abattoir factory (El Ouardia City, Tunis). Analysis of the raw FVW, FW, AW and WAS were carried out several times and the average compositions are shown in Table 1. The FVW consisted of homogenised courgettes, lettuce, tomatoes, apple, orange, pear, potatoes and carrot to give 8.3% TS with VS content of 93%. The feedstock was made up by adding a percentage by volume of water, AW, WAS and FW to FVW. Four feedstocks (W1–W4) (Table 2) which were prepared with average TS contents of 2.7%, 2.74%, 2.9% and 2.8%, were used to load R1–R4. Approximately, 90% of VS in the different feedstocks were given from FVWs.

2.3. Technical analysis

The biogas produced was measured daily by gas meter (Ritter – Bochum Langendreer, Germany) and its composition was estimated using an ORSAT apparatus (Bouallagui et al., 2003). Total solids (TS), total volatile solids (TVS), total suspended solids (TSS), pH, alkalinity and total volatile fatty acids (VFAs) were determined according to the APHA Standard Methods (1995). Total organic carbon (TOC) was measured by catalytic oxidation on a TOC Euroglace analyser. Total nitrogen (TN) was estimated by the Kjeldahl method.

2.4. Statistical analysis

In order to determine whether the observed differences between digesters performances were significantly different, data were subjected to the ANOVA tests (StatSoft Inc, 1997). Differences between co-substrates' addition effects (p and p1) were compared with 0.05.

3. Results and discussion

3.1. Effect of co-substrates addition on the fermentation efficiency

3.1.1. Biogas production

The biogas production by the digestions of FVW alone and the co-digested wastes is shown in Fig. 2. Analysis of biogas production profiles for the substrates combinations showed that there were significant differences among the combinations tested. The results for co-digested substrates are better than those obtained from digestion of FVW. The average biogas production rate varied between 1.53 l d⁻¹ and 2.53 l d⁻¹, with value being highest for both mixtures W2 and W3 and lowest for 100% FVW. The specific biogas productions for the four digestion processes (R1–R4) were 0.403, 0.611, 0.580 and 0.436 l g⁻¹ removal VS.

The methane yields from FVW, which have been reported previously, are variable depending on the waste composition and the used reactor design. The reported range was from 0.16 to 0.4 m³ kg⁻¹ VS added (Bouallagui et al., 2005). The results presented in this paper for FVW digestion are therefore comparable with these earlier results.

The data for the co-digestion may be also compared with earlier works. Callaghan et al. (2002) examined the co-digestion of FVW with cattle slurry and chicken manure. The methane yields they obtained of 0.35–0.4 l g⁻¹ VS added were very similar to those in the current study for FVW–AW co-digestion. Gomez et al. (2006) have also reported similar results for the anaerobic co-digestion of FVW and primary sludge. However the methane yield obtained from FVW and WAS anaerobic co-digestion in two stages tubular digesters of 0.25 l g⁻¹ VS added (Dinsdale et al., 2000) was lower than that is presented in this work.

The addition of AW, WAS and FW enhanced the biogas yield by 51.5%, 43.8% and 8.1%, respectively. Biogas yields for FVW:AW and FVW:WAS co-digestions are much greater thanks to the better C/N ratio of these feedstocks (Fig. 3). The ANOVA of the data indicated that digesters (R2 and R3) performance enhancement was statistically significant (p < 0.05) (Table 3). Despite a considerable decrease of C/N ratio from 34.2 to 27.6, the addition of FW slightly improves the gas production rate and biogas yield compared to anaerobic digestion of FVW alone (p1 > 0.05). The C/N ratios of the co-digested FVW:AW and FVW:WAS which ranged between 22 and 25 were within the C/N ratio (20–25) required for stable and better biological conversions reported by others on anaerobic digestion of organic wastes ([Parkin and Owen, 1986], [Mshandete et al., 2004] and [Yen and Brune, 2007]). Kayhanian and Hardy (1994) reported C/N ratios between 25 and 30 as being optimal. However, Kivaisi and Mtila (1998) argue that the C/N of approximately between 16 and 19 is optimal for methanogenic performance.

3.1.2. Organic matter removal

The total volatile solids destruction for the various co-digested substrates combinations is given in Table 3. The higher degradation efficiencies were obtained for the digesters treating FVW:AW and FVW:WAS and operated at an organic loading rate of 2.46 gTVS l⁻¹ d⁻¹ and 2.51 gTVS l⁻¹ d⁻¹, respectively. They were associated with the higher specific biogas production and a lower content of volatile solids in the digested effluent, which represents a lesser amount of output stabilised effluent with a better dewatering properties. The data of Table 3 showed that about 84–85.4% of TVS were degraded to methane and carbon dioxide with the co-digestion of organic wastes. These results are in agreement with those obtained by Fernández et al. (2005) and better than those obtained by Callaghan et al. (1999) and Dinsdale et al. (2000). It is very likely that the high degradation efficiency in the co-fermentation was due to an improved ratio of nutrients and better availability of the organic substances, which facilitate their assimilation by anaerobic flora and increases the degree of degradation (Krupp and Schubert, 2005). Furthermore, the AW addition improves the proteins availability that were used by anaerobic bacteria to produce new cells and enzymes.

3.1.3. Alkalinity, total VFAs and pH variation

In a well balanced anaerobic digestion process, total VFAs levels are low ([Fernández et al., 2005] and [Chen et al., 2007]). In this study all the combinations examined, except FVW–FW showed lower levels of total VFAs in their digested effluent at steady-state (Fig. 4). During the period days 1–25 high levels of total VFAs of up to 2800 mg l^−l, 2000 mg l⁻¹, 1500 mg l⁻¹ and 2300 mg l⁻¹ experienced in digesters treating W1, W2, W3 and W4, respectively, indicating that the reactors were not operating at their optimum. An average of 750 mg l⁻¹, 300 mg l⁻¹, 520 mg l⁻¹ and 1900 mg l⁻¹ total VFAs were found, respectively for the different digesters at steady-state. Total VFAs concentration remained at high level for the digester treating FVW–FW indicating a digestion limitation accompanied by the lowest biogas yield and volatile solid removal. Although, the pH and the partial alkalinity of this reactor were high, indicating good process stability.

An average value of total VFAs between 1330 and 1800 mg l^−l was also found in the effluent of a successful methanogenic reactor treating FVW with WAS as a co-substrate (Dinsdale et al., 2000). In contrast, levels of 55–505 mg l^−l total VFAs were found in the reactors treating multi-component agro-wastes (Misi and Forster, 2001) indicating that higher and lower levels of total VFA are possible for organic wastes co-digestion.

The initial partial alkalinity ranged between 600 mg l⁻¹ and 2400 mg l⁻¹ while the final range was 1300 mg l⁻¹ and 7000 mg l⁻¹ (Fig. 4). The latter, demonstrated an increased partial alkalinity in the digesters compared to the initial values before anaerobic digestion stability. This provided further evidence that the co-digestions of FVW and co-substrates studied were successful. Previously, laboratory studies on mesophilic and thermophilic anaerobic organic wastes digestion reported a range of 2000–4000 mg l⁻¹ partial alkalinity as being typical for properly operating digesters ([Chen et al., 2007] and [Sharma et al., 2000]). The initial values reported in this study fall within this range. However, the final values are higher than the reported values, especially for FVW–FW co-digestion. This increase could be due to generation of NH₄⁺ during the digestion of protein in fish waste which resulted in an increased digester buffering capacity and hence stability of the digesters. This is an interesting cost effective approach since no external buffer sources were added.

The pH was monitored continuously in the digesters. The evolution of the pH values obtained under different conditions is presented in Fig. 4. Despite the low pH of the feed substrate (4.3–5.4), the pH increased to its neutral value (between 6.9 and 7.57) due to the process stability and the activity of methanogenic bacteria. The outlet pH value increased with the addition of high nitrogen content co-substrate on the FVW. The highest values were obtained for the digester treating FVW–FW co-substrates due to a high partial alkalinity of 7000 mg l⁻¹ and ammonia concentration of 2200 mg l⁻¹.

One of the criteria for judging digester stability is the VFAs:Alkalinity ratio. There are three critical values for this ([Switzenbaum et al., 1990] and [Callaghan et al., 2002]). If this ratio was lower than 0.4, the digester should be stable. While, when the ratio ranged 0.4–0.8, some instability will occur on the digester performances. However, the ratio higher than 0.8, indicates a significantly instability. When FVW being digested alone (Fig. 5), the VFAs:Alkalinity ratio (0.57) did not rise above the criteria value of 0.4, implying that despite the results for the biogas yield and VS reduction, there was the potential for instability. Generally, FVW is thought of as being highly degradable, but it is essential that there is an adequate alkalinity (Gunaseelan, 1997). Lane (1984) suggested that, for a balanced digestion of FVW, the alkalinity should not be less than 1500 mg l⁻¹ and that the VFAs:Alkalinity ratio should be less than 0.7. The addition of AW, WAS and FW in the feedstock of FVW produced a decrease of VFAs:Alkalinity ratio in the digesters to be around 0.07, 0.11 and 0.27, respectively, showing better processes stabilities and buffer capacities. Consequently, the occurrence of an imbalance between the different groups of anaerobic bacteria which may take place in an unstable anaerobic digestion of FVW process could be prevented.

3.2. Digesters performances

The results of the digesters performances are shown in Table 3. Compared to methane yield for the pure FVW, co-digestions of the FVW–AW, FVW–WAS and FVW–FW enhanced the biogas yield by 51.5%, 43.8% and 8.1%, respectively. The better biogas yield (0.61 l g⁻¹ removal VS) and VS removal (85.4%) were obtained by W2 and W3 co-digestions, respectively. This could be due to positive synergism in the digestion medium, especially for FVW–AW, FVW–WAS combinations, supplying missing nutrients and reducing of inhibitory materials in feedstock by the co-substrates (Mshandete et al., 2004). The average CH₄ content of the biogas produced from different treated wastes range between 64% and 66%. This range of methane content is closer to the range of 55–65% which is normally obtained from conventional anaerobic digestion of organic wastes conducted in single stage digesters.

The results of FVW–FW digestion showed a decrease of biogas production rate due to the high amount of ammonia (2200 mg l⁻¹) and total VFAs (1900 mg l⁻¹). In fact, the total ammonia nitrogen and VFAs both are important intermediates and potential inhibitors in the anaerobic digestion process. High concentration of ammonia and VFAs in the digester would decrease the methanogens activity and further accumulation could inhibit the anaerobic digestion (Chen et al., 2007).

The estimated free ammonia (FA) concentrations based on pH and total ammonia for the four digestion processes (R1–R4) were 3.2, 63.4, 34.4 and 264.5 mg l⁻¹. It is generally believed that ammonia concentrations below 200 l⁻¹ are beneficial to anaerobic process since nitrogen is an essential nutrient for anaerobic microorganisms (Liu and Sung, 2002). Gallert and Winter (1997) studied the anaerobic digestion of organic wastes and reported that methane production was inhibited 50% by 220 l⁻¹ FA at 37 °C and by 690 l⁻¹ FA at 55 °C, indicating that thermophilic flora tolerated at least twice as much FA as compared to mesophilic flora. Several mechanisms for ammonia inhibition have been proposed, such as a change in the intracellular pH and the inhibition of specific methane synthesising enzyme reaction (Calli et al., 2005). FA has been suggested to be the main cause of inhibition since it is freely membrane-permeable. The hydrophobic ammonia molecule may diffuse passively into the cell, causing proton imbalance, and/or potassium deficiency (Gallert et al., 1998).

4. Conclusion

An interesting option for improving yields of anaerobic digestion of solid wastes is co-digestion. Its benefits include improved balance of nutrients, synergistic effect of microorganisms, increased load of biodegradable organic matter and better biogas yield. Combination of FVW with other substrate like AW and WAS can significantly improve the waste treatment efficiency. This resulted in a highly buffered system as the high nitrogen content co-substrate contributed to high amount of ammonia. Fish waste was not as successful as a co-substrate for FVW digestion. As a consequence to the FW addition, the VS reduction deteriorated and the methane yield increased slightly. This appeared to be due to the concentration of ammonia. Results indicate that the ratio of C/N is a determining parameter which influenced the methane production and the organic matter bio-degradation. The biogas production yield was enhanced by 51.5% and 43.8% by the addition of AW and WAS, respectively to FVW feedstock. It was verified that these combinations could be a promising and practical alternative for the simultaneous recycling of different types of organic wastes with high stability. It seemed that carbohydrate rich substrates are good producers of VFAs and that protein rich substrate are yielding good buffering capacity. The high values for the methane yield and the VS reduction were indicatives for a high content of biodegradable organic matter in the co-substrate due to an improved ratio of nutrients and better availability of the organic substances.

Acknowledgements

The authors wish to acknowledge the Ministry of Superior Education and Scientific Research and Technology, which has facilitated the carried work.