Codigestion of pig slurry and organic wastes from food industry

PROCEEDINGS OF THE II INTERNATIONAL SYMPOSIUM ON ANAEROBIC DIGESTION OF SOLID WASTE. BARCELONA. JUNIO 1999. PAG. 192-195. CODIGESTION OF PIG SLURRY AND ORGANIC WASTES FROM FOOD INDUSTRY Elena Campos, Jordi Palatsi and Xavier Flotats Department of Environment and Soil Science, University of Lleida. Rovira Roure, 177. 25198 Lleida, Spain. ABSTRACT Anaerobic digestion of pig slurry can be a good option to get a higher value for these wastes. Although thermophilic anaerobic digestion is more efficient than mesophilic anaerobic digestion, it presents some limitations, like less stability and an increased effect of some inhibitors. The main inhibitor for pig slurries is free ammonia. In order to improve the methane production, some mixtures of slurry and organic wastes from food industry were tested, like wastes from fruit and olive oil refinery industries (pear waste and oil bleaching earth). A batch experiment was carried out, in order to determine the maximum potential of methane production and biodegradability of these wastes. The experiment was developed in mesophilic and thermophilic ranges (35º and 55º). The mesophilic results were better than the thermophilic ones, showing a large inhibition by ammonia in the thermophilic range. In both temperature ranges, the methane production was improved by addition of a co-substrate. The higher methane production was obtained from the codigestion of slurry and oil bleaching earth (95% and 5% respectively). The methane yield was 344 mL CH4/g VSinitial, which is 2.4 times the methane yield for slurry (144.0 mL CH4/g VSinitial). KEYWORDS: Anaerobic digestion; codigestion; ammonia inhibition; organic industrial wastes; pig slurry; temperature. INTRODUCTION In Catalonia there are more than five million pigs, producing more than 30.000 m3/day of wastes. The anaerobic digestion can be a good option to get a higher value for these wastes. Thermophilic anaerobic digestion has some theoretical advantages over mesophilic, like a better sanitation of effluents, faster process rates and bigger methane yield. However, the slurry presents a very high concentration of ammonia, which has been described as a very important inhibitor for anaerobic process. Free ammonia is the active component causing ammonia inhibition (Angelidaki et al., 1993), and its concentration depends on the total ammonia, pH and temperature. The inhibitory concentration depends on the micro-organisms adaptation phenomena (Van Velsen et al., 1979; Koster, 1986; Hansen et al., 1998). So, a wide variety of inhibitory values are found in the literature, i.e. 80 mg N-NH3/L (Koster, 1986), 700 mg N-NH3/L (Angelidaki et al., 1993) or 1100 mg N-NH3/L (Hansen et al.1998). By codigestion of different types of wastes, such as manure, household solid waste and organic industrial waste, a higher gas yield can be obtained from biogas reactors. Industrial wastes are more easily degradable and have a higher gas potential than manure, varying from 30 to 500 m3/ton (Ahring et al., 1992; Angelidaki et al., 1997). Slurry can be an excellent basic substrate for the codigestion, because it has a high water contents, a high buffering capacity and a wide variety of needed nutrients for anaerobic bacteria. (Angelidaki et al., 1997). Previous experience of codigestion of slurry and sewage sludge, with satisfactory results, has been reported (Flotats et al., 1999). Codigestion of manure and oil bleaching earth (OBE) has provided very good results in Denmark (Ahring et al., 1992), due to the high lipids content of this kind of wastes, that provides a high potential for biogas production (Ahring et al., 1992). However, an excess of lipids can cause inhibition (Hanaki et al., 1981; Ahring et al.,1992). Other interesting industrial wastes are those produced in fruit juice factories. The main problem associated with digestion of ligno-cellulosic wastes is the maintenance of stable pH, which is associated with its poor buffering capacity and volatile fatty acids accumulation. Combination with another waste with high buffering capacity can be a good way to digest it (Banks et al., 1998). Both types of wastes (OBE and pear waste) are produced in large quantity in the area of Lleida (Catalonia, Spain), and the objective of the present work is to analyse the performance of its codigestion with pig slurry, using batch digestion tests. MATERIAL AND METHODS The batch reactors were 120 mL glass vials, filled with 50 g. of sample and 5 g. of inoculum, initially bubbled with N2/CO2 gas mixture for air displacement, tightly closed with rubber taps, placed in incubators held at 35ºC (mesophilic test) and 55ºC (thermophilic test), and followed for 75 days. Five repetitions were involved for every treatment and every temperature. Duncan’s multiple statistical test was applied for gas production comparison among treatments and temperatures. The inoculum proceeded from an anaerobic digester treating pig slurry since 1983, in the mesophilic range. The pig slurry used came from a fattening pig farm near Lleida. The pear waste came from a fruit juice factory, close to Lleida. The OBE waste came from an olive oil bleaching and filtering factory. The experimental design consisted of eleven treatments and a blank run, performed at 35ºC and 55ºC temperatures. Table 1 shows proportions used for each co-substrate and substrate composition for each treatment. TABLE 1. SUBSTRATE COMPOSITION FOR EACH TREATMENT Treat ment Slurry % Pear Waste % OBE % Water % TS (1) VS (1) COD (2) TKN (3) NH4 + (3) pH Total Alk (4) VFA Alk (5) Alk. Rate T1 100 - - 0 7.92 5.87 82.50 5.06 3.43 8.06 10.55 3.66 0.29 T2 95 - - 5 7.52 5.58 78.37 4.82 3.28 7.96 10.05 3.30 0.27 T3 87.5 - - 12.5 6.93 5.14 72.18 4.44 2.97 8.03 8.98 3.09 0.29 T4 80 - - 20 6.33 4.69 66.00 4.05 2.70 7.95 8.20 2.88 0.29 T5 95 5 - - 9.29 7.29 102.4 4.91 3.21 7.55 9.75 3.66 0.31 T6 87.5 12.5 - - 11.35 9.45 132.3 4.98 3.01 7.25 8.73 3.54 0.34 T7 80 20 - - 13.41 11.63 162.1 4.66 2.71 6.88 7.58 3.39 0.37 T8 - 20 - 80 7.08 6.98 96.14 0.89 0.08 3.45 0.00 0.00 - T9 95 - 5 - 12.40 7.54 136.9 4.76 3.20 7.69 9.50 3.75 0.33 T10 87.5 - 12.5 - 19.13 10.09 218.5 4.48 3.04 7.33 8.10 3.66 0.38 T11 - - 12.5 87.5 12.2 4.94 58.54 0.38 0.04 6.93 0.95 0.90 0.79 (1)Total and volatile solids are expressed as % (w/w); (2) Chemical Oxygen Demand in g O2/Kg; (3) Total Kjeldahl Nitrogen and total ammonia are expressed as g N/Kg; (4) Total alkalinity is expressed as g CaCO3/L; (5) Volatile fatty acids alkalinity is expressed as g AcH/L. Analytical procedures. Gas analysis (N2, CH4, CO2) were carried out by gas chromatography using a Thermoquest 8000 chromatograph with a TCD and a packed steel column Porapak N. Total and volatile solids, COD, pH, alkalinity and Kjeldahl and ammonia nitrogen were determined according to Standard Methods (APHA, 1995). Free ammonia concentration was calculated from the equilibrium relationship. RESULTS AND DISCUSSION Accumulated methane production and yields obtained at 75 days batch test are presented in Table 2, with the results of the Duncan test. Results in mesophilic experiment. Non statistically significant differences were obtained among the different slurry dilutions tested (treatments T1-T4), showing non-inhibition effect by ammonia for this temperature. Although the high total ammonia concentration (2.7-3.4 g N-NH4/L), the free ammonia concentration presents a low value at the beginning of the digestion. The effect of slurry/pear mixture was no significant for methane (treatments T5-T8). For T7 treatment (20% of pear waste), the process failed, probably due to a less buffering capacity and a higher fibre contents. Digestion of pear waste (T8) couldn’t develop due to a low pH and buffering capacity. Codigestion of slurry and OBE (treatments T9-T10) showed a very good behaviour at 35ºC, presenting a statistically significant higher methane production than those obtained for the other treatments. The T10 process showed a lag phase, probably due to a slight inhibition by long chain fatty acids. The lipid contents in this mixture was 35.7 g/L, higher than values reported by other authors (Ahring et al., 1992; Hanaki et al., 1981). TABLE 2. METHANE PRODUCTION AND YIELD BY TREATMENT AND TEMPERATURE, AT 75 DAYS BATCH TEST Methane Production (mL CH4) Methane Yield (mL CH4/gSVini) Methane Yield (mL CH4/gSVini) 55ºC 35ºC 55ºC 35ºC Differences due to temperature T1 192.6 de 439.8 c 63.1 c 144.0 b 80.9* T2 199.1 de 425.1 c 70.0 cd 144.0 b 74.0* T3 246.2 ef 365.8 c 90.8 def 134.8 b 44.0* T4 293.4 f 367.3 c 114.6 f 143.6 b 29.0* T5 419.1 g 534.0 c 110.7 ef 141.0 b 30.3* T6 518.0 h 828.4 d 107.4 ef 171.6 b 64.2* T7 535.3 h 188.4 b 88.1 de 31.0 a -57.1* T8 0.0 a 0.0 a 0.0 a 0.0 a - T9 6.8 ab 1289.0 f 1.8 a 343.9 d 342.1* T10 136.5 cd 1050.1 e 26.9 b 206.6 c 179.7* T11 77.3 bc 9.8 a 29.7 b 3.8 a - Duncan’s multiple test with a 95% confidence level: Different letter means significant differences among the means, by columns (for each temperature); (*) means significant differences between temperatures for each treatment. Results in thermophilic experiment. For slurry dilution treatments, the highest methane production was obtained in the T4 treatment (20% of water). In these experiments, methane yield and initial free ammonia concentration are very well correlated, as can be shown in Figure 1, indicating a significant inhibitory ammonia effect. For slurry/pear mixtures, the highest yield value was 110.7 mL CH4/g VS for T5 treatment, which is smaller than T4 treatment, but not statistically different. Taking into account that the T5 treatment has higher initial ammonia concentration than the T4, it can be said that the main effect of pear waste is the addition of organic matter. Co-digestion of slurry and OBE presented low gas production, indicating inhibition by ammonia and LCFA. All treatments present significant differences, at 95% confidence levels, between mesophilic and thermophilic ranges. Inhibition by free ammonia concentration at thermophilic range, increasing to 1000 mg N-NH3/L during the experiment, could be the main effect. Taking into account some previous good experiences with higher levels of ammonia contents (Hansen et al., 1998), the importance of the adaptation of biomass used as inoculum must be remarked. Figure 2. Correlation between methane yield and free ammonia concentration, at 55ºC CONCLUSIONS Codigestion of slurry with other substrates could be a very good option to increase the methane production. Codigestion with OBE showed a very good behaviour at mesophilic range, with a methane yield 2.4 times bigger than for slurry digestion. The main effect of pear waste addition was increase of organic matter. In general the mesophilic range showed a better behaviour than the thermophilic one, showing a clear inhibition by free ammonia. The previous adaptation of inoculum is basic for process performance. ACKNOWLEDGMENTS This research has been supported by the Lleida Council (Paeria), and by a grant from CIRIT (Generalitat of Catalunya). REFERENCES Ahring, B., Angelidaki, I., Johansen, K.(1992).Anaerobic treatment of manure together with industrial waste. Wat. Sci. Tech. 25 (7), 311-318 American Public Health Association (1995). Standard Methods for the Examination of Water and Wastewater. 19th ed., APHA-AWWA-WEF, Washington DC, USA. Angelidaki, I., Ahring, B. (1993). Thermophilic anaerobic digestion of livestock waste: the effect of ammonia. App. Microbol Biotechnol., 38, 560-564. Angelidaki, I., Ahring, B. (1997 ). Anaerobic digestion in Denmark. Past, present and future. In: III Curs d’Enginyeria Ambiental. Lleida, 336-342. Banks, C.J., Humphreys, P.N.(1998). The Anaerobic Treatment of a Ligno-cellulosic substrate offering little natural pH buffering capacity. Wat. Sci. Tech. 38 (4-5), 29-35. Flotats, X., Bonmatí, A., Campos, E., Antúnez, M. (1999). Ensayos en discontinuo de codigestión anaerobia termofílica de purines de cerdo y lodos residuales. Efecto del amonio. Inf. Tecn., 10 (1) 79-85. Hanaki, K., Matsuo, T., Nagase, M. (1981). Mechanism of Inhibition Caused by Long-Chain Fatty Acids in Anaerobic Digestion Process. Biotech. Bioengin., 23, 1591-1610. Hansen, K., Angelidaki, I., Ahring, B.(1998). Anaerobic digestion of swine manure: Inhibition by ammonia. Water Research, 32 (1), 5-12. Koster, I.W. (1986). Characteristics of the pH-influenced adaptation of methanogenic sludge to ammonium toxicity. JChem. Tech. Biotech 36, 445-455. Van Velsen (1979). Adaptation of Methanogenic sludge to high ammonia-nitrogen concentrations. Water Research, 13, 995-999 CH4/VSi = -0.0636*[N-NH3] + 72.962 R2 = 0.7777 CH4/VSi = -0.2889*[N-NH3] + 341.56 R2 = 0.8573 0 20 40 60 80 100 120 140 700 750 800 850 900 950 1000 N-NH3 (mg N/L) mL CH4/g VSi Day 15 Day 75