Effect of Feed Loading on Biogas Methane Production in Batch Mesophilic Anaerobic Digesters Treating Food Waste

Transcription

1 February 214, Volume 5, No.1 International Journal of Chemical and Environmental Engineering Effect of Feed Loading on Biogas Methane Production in Batch Mesophilic Anaerobic Digesters Treating Food Waste Musa Idris Tanimu; Tinia Idaty Mohd Ghazi * ; Mohd Razif Harun; Azni Idris Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 434 UPM Serdang, Selangor, Malaysia * Corresponding author tinia@upm.edu.my Abstract Food waste mixture upgraded to a carbon to nitrogen ratio of 3 was co-digested at different feed loadings of.5, 1.5, 3.5 and 5.5gVS/L in batch and mesophilic conditions (37 o C). Results showed that the production of biogas methane increased with an increase of the feed loading to the digester. A maximum cumulative biogas methane yield of.535l/gvs was attained at feed loading of 3.5gVS/L. Generally, it was observed that higher feed loading to the digester led to ph reduction and a decrease in treatment efficiency from 96% to 75%. Key words: Batch digester; Biogas methane; Feed loading; Food waste; Methane; Anaerobic digesters 1. Introduction Recent statistics showed about 5% of the municipal solid waste generated in Malaysia (765 tonne/day) is food waste [1]. These food wastes (FW) are mainly from commercial restaurants, school cafeteria, residential homes etc. Anaerobic digestion (AD) has been considered as the most viable method of FW treatment because of its Manuscript accepted December 5, 213. This study was supported by the Research University Grant Scheme (RUGS) Universiti Putra Malaysia. Musa Idris Tanimu is a M.Sc research student in the Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, on study leave from the Chemical Engineering Department, Kaduna Polytechnic, PMB 221, Kaduna, Kaduna state Nigeria ( musandadrisu@gmail.com) * Tinia Idaty Mohd Ghazi is an Associate Professor, at the Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 434 UPM Serdang, Selangor, Malaysia ( tinia@upm.edu.my, tiniaghazi@gmail.com, Tel: ) Mohd Razif Harun is a Senior Lecturer at the Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 434 UPM Serdang, Selangor, Malaysia ( mh_razif@upm.edu.my) Azni Idris is a Professor at the Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 434 UPM Serdang, Selangor, Malaysia ( azni@upm.edu.my) high moisture content (above 8%). Besides this, the AD treatment options produce biogas methane that is a renewable source of energy. The concentration and composition of FW can be highly variable depending on their sources. A number of researchers who has studied the FW anaerobic digestion reported different feed loadings (FL) tolerable by the digesters treating food wastes (FW). In a mesophilic AD, [2] achieved only a feed loading of less than 3gVS/Ld in a stirred digester due to volatile fatty acid (VFA) accumulation using fruits and vegetable as substrates. In a similar digestion, [3] digested the domestic FW at OLR of 2gVS/Ld and resulted in digester accumulation problems from high release of ammonia concentrations. Reference [4] also studied the methane yield of meat waste and mixed FW at 37 o C and obtained.482 and.472l/gvs respectively after 28 days of AD. Increase in digester loading, measured in gram of volatile solids per litre of substrate fed (gvs/l) into the anaerobic digester (in the case of batch process) can result in digester accumulation affecting the activities of methane forming anaerobes and leading to digester failure. Basically, two challenges have been linked with increasing feed loading: If the rate at which volatile fatty acids (VFAs) produced especially acetic acid is greater than the rate at which it is used up by the aceticlastic methanogens during AD, then,

2 this could lead to gradual decrease in ph which is unfavourable for biogas methane production [5]. The second challenge of increasing feed loading (FL) especially for feed substrates rich in nitrogen content such as protein as in FW is the risk of digester failure through free ammonia accumulation arising from protein solubilization [6]. Attempts have been made to overcome these challenges by improving the ph and temperature conditions for optimum microbial activities as well as improving the carbon to nitrogen ratio (C/N) through codigestion [7]. Literature showed that an optimum microbial activity requires 3 times carbon as much as nitrogen in a feed substrate to an anaerobic digester. The advantages of co-digestion as carried out in this study, was among other reasons, to help achieve higher C/N ratio, reduce the nitrogen content of substrates by combining FW rich in carbon content with that rich in nitrogen sources such as protein in FW fractions. A high C/N ratio FW means low protein solubilization resulting in lower risk from free ammonia accumulation. This study was initiated to investigate and compare the methane yield and feed loading that could be attained at batch FW substrate loadings of.5-5.5gvs/l when the required C/N ratio of the mixed commercial restaurant FW (C/N=17) was further upgraded to the optimum value of 3 through co-digestion of meat waste, fruit/vegetable waste and the commercial restaurant FW. 2. Materials and Method 2.1 Source and nature of digested food waste Food wastes were collected from a food waste anaerobic digestion treatment plant in Taman Sri Serdang, Selangor, Malaysia. The anaerobic digestion (AD) plant which treats a mixture of source sorted food wastes obtained from the commercial restaurants, market and meat industries around Serdang area (population of 3, people) is part of an integrated waste management system for the city of Serdang. The main components of the restaurant food waste (FW) include: kitchen wastes such as rice and noodles (77%), leafy vegetables/ salad (7%), soup (6%), and cooked meat/fish (1%). The carbon to nitrogen ratio of the restaurant FW (C/N=17) was upgraded to 3 by co-mixing with fresh chicken meat waste (C/N=5), fruits waste (C/N=35) and vegetable wastes (C/N=11). The average composition of the vegetable wastes includes: baby corn (5%), lettuce (23.6%), carrot (4.47%), broccoli (18.2%) and green leafy vegetables (48.73%) all on wet weight (ww) basis. The average composition of the fruit wastes includes: papaya (26.84%), orange (18.63%), pineapple (38.78%), watermelon (1.55%) and berries (5.2%) on ww basis. Substrate and Inoculum preparation 4 The mixture of restaurant food waste, vegetable wastes, fruits waste and raw chicken meat waste was ground using a heavy-duty blender model 39BL11 (Waring commercial, USA) and sieved with a 1.mm sieve size. It was prepared in bulk and stored at -2 o C for later use. The composition of food waste, meat waste, fruit wastes and leafy vegetables added to achieve C/N ratio of 3 are presented in table 1. Cow dung was used as source of inoculum in this study. The proportion of feed mixture to inoculum was 1:1 by volume. The characteristic of the inoculum is presented in table 2. The ratio of the FW to water dilution in the various bottles were: 1:3, 1:9, 1:3 and 1:1.5 to give volatile solid (VS) contents in the 1ml glass bottles as.5, 1.5, 3.5 and 5.5gVS/L respectively. 8 bottles of 1ml capacity were used as digesters each with a working volume of.8litres. Nitrogen gas was bubbled for 5 minutes through each digester medium to eliminate oxygen. Each reactor prepared in duplicate with the feed loading mentioned above was placed in an oven set at 37 o C. The characteristics of the FW substrate used are presented in table Analytical methods Total solids (TS), volatile solids (VS), COD, alkalinity and NH 3-N were carried out every other day according to the standard methods [8]. Carbon and nitrogen tests were carried out using the CHNS machine model CHNS 932 (Leco corporation, USA). Portions of the digestion medium were collected and centrifuged every other day using a centrifuge (model 242, Kubota Corporation Japan) at 5 rpm for 15 minutes before carrying out COD test. Analysis on alkalinity was performed by titration of 1ml of the sample against 1N H 2SO 4 to ph 4.5. Ammonia-nitrogen (NH 3-N) test was performed by Nessler s method using Hach spectrophotometer (Odyssey DR 25). Volume of biogas collected in the gas bag was measured using water displacement method. The gas composition was obtained using a gas chromatography instrument (Agilent 689N network GC system) equipped with a thermal conductivity detector and carboxen 11 plot column 3m x 53µmx µm nominal (Supelco 25467, USA). The injector, detector and oven temperatures were: 2 o C, 23 o C and 4 o C respectively. Argon gas was used as the carrier gas at a flow rate of 3.mL/min. ph was determined using a combined ph and temperature meter (Trans Instruments, Singapore). 2.3 Data analysis A one-way ANOVA using the Microsoft Excel was carried out to know if there was any significant difference in biogas produced at the various feed loadings tested with level of significance (p) set at 5%.

3 Cummulative methane yield (L/gVS) Alkalinity (mg/l) Results and Discussion 3.1 Feed characteristics The characteristics of the food wastes treated are presented in table 1. The 95% moisture content (MC) of the FW showed that it is highly putricible and therefore well suited to anaerobic digestion. This level of moisture content is quite higher than those reported for FW in UK Details Table 1. Characteristics and composition of feedstock used Characteristics Parameter value Moisture content (%) VS (g/l) VS (%TS) TS (g/l) TS (% FW) NH 3-N (mg/l) 82 Alkalinity (mg/l) 953 ph 4.4 COD (mg/l) C/N 3.36 Na (%) 1.6 K (%) 3.33 P (%).66 Lignin (%) NDF (%) ADF (%) GE (cal/g) Feed Composition (weight %) a Restaurant food waste 6 Fruits waste 1 Vegetable waste 7 Meat waste 23 a percent fresh weight and USA (range 74-93% MC) [9]. This could be as a result of seasonal or weather conditions especially the higher humidity conditions in Malaysia (about 8%). A high volatile solids (VS) content of 98% means that the food substrate is high in organic solid content from which biogas methane can be obtained during AD. The C/N ratio of the restaurant FW (C/N=17) was upgraded to C/N ratio of 3 by co-mixing with common FW sources from fruits, vegetables and chicken meat wastes. This can help release trace and major elements (e.g. Na, K, P etc) essential for AD [1]. Apart from trace element addition, other advantages of co-digestion of the complementary wastes among other reasons is to help provide adequate nutrition, increase C/N ratio and dilution of potentially poisonous components that may be contained in the FW [5]. A high COD of up to 294,735mg/L in the FW means that it contains a high level of organic pollutants, which could be potentially converted to biogas. 3.2 Biogas methane production The profile of cumulative methane yield during the 3 days AD at feed loading ranging from.5-5.5gvs/l is represented in Fig. 1. The highest cumulative methane yield at FL of.5, 1.5, 3.5 and 5.5gVS/L were.3,.43,.54 and.49l/gvs, respectively. In each of the feed loading, the gas production continued up to day 3 except at FL of.5 and 5.5gVS/L when the biogas production stopped on day 26 and 28. The reason for this can be seen to be due to a decrease in ph from neutral to 4.5 in the case of 5.5gVS/L FL. Decrease in ph during AD has been identified to be the result of significant volatile fatty acids (VFAs) production [1]. Lowered ph as a result of VFAs accumulation can lead to inhibition of the activities of the methane producing methanogens during AD. The cessation of gas production for.5gvs/l FL on day 26 could be due to limited feed concentration..6 Table 2. Characteristics of inoculum Parameter Value Moisture content TS (mg/l) TS (%FW) 19.5 VS (mg/l) 36. VS (%FW) 69.5 NH 3-N(mg/L) 625. ph FL=.5gVS/L

4 VS destruction (%) Methane in biogas (%) Figure 1. Methane yield during the 3 days AD of food waste at different feed loadings. This is because the ph of the medium was still within neutral bounds of 6.8. This reason is further supported by the highest treatment efficiency of up to 95% obtained for the FL of.5gvs/l as compared to others (Fig. 2). The highest cumulative yield of methane obtained in this study was.535l/gvs at FL of 3.5gVS/L. It was observed that the biogas methane yield increased from.3l/gvs at the loading of.5gvs/l to.535l/gvs at the loading of 3.5gVS/L. At the FL of 5.5gVS/L, the methane yield reduced to.488l/gvs. The ph profile at FL of 5.5gVS/L was observed to reduce from 7.5 to 4.5 most likely due to gradual VFA accumulation. If the rate of VFAs produced as intermediate product during AD is higher than the rate used up by the aceticlastic methanogens, this can lead to gradual accumulation which is noticeable by a gradual reduction in ph of digestion medium [11]. Reduction in ph away from the neutral bounds ( ) affects the activities of the methane producing anaerobes. This phenomenon could largely be responsible for the decrease in methane yield at FL of 5.5gVS/L. suggest that the digestion FL of 3.5gVS/L was not just the highest yielding in terms of cumulative gas produced but also has the fastest gas production rate among the other FL tested. This suggests that the rate of methanogenesis was fastest at FL of 3.5gVS/L followed by 5.5gVS/L, 1.5gVS/L and.5gvs/l. The methane content of biogas produced during the first 1 days of AD ranged between 27.6% and 65.5% for all FL tested. However, the highest values of percentage methane recorded were 58%, 64%, 75% and 66% at FL of.5, 1.5, 3.5, and 5.5gVS/L respectively (Fig. 3). The percentage of methane in the biogas again showed that the methanogens were most active at feed loading of 3.5gVS/L followed by 5.5gVS/L, 1.5gVS/L and then.5gvs/l. The cumulative average yield of biogas obtained during the digestion was 7.82L, 1.88L, 13.74L and 15.89L at FL of.5, 1.5, 3.5, and 5.5gVS/L, respectively. The methane composition in biogas as obtained above, is comparable to 75% obtained by [12] in a similar FW digestion at mesophilic conditions FL=.5gVS/L Figure 2. Volatile solids (VS) destruction during the 3 days AD at different feed loadings. A single factor ANOVA test was carried out on the values of methane yield obtained in this study with significance level (p) set at 5%. The results showed that the methane yields obtained at all FL were quite significant with p level at.4. The methane yield obtained in this study at FL of 5.5gVS/L (i.e..488l/gvs) was higher than that obtained by [12] (i.e..435l/gvs) in a similar FW digestion for 28 days at FL of 6.8gVS/L. The difference in methane yield can be attributed to the variation in feed composition and the lower C/N ratio of FW used by the researchers (C/N=14.8). In all the FL studied, at least 6% of the methane yield was achieved within the first 1 days of digestion. The linear regression lines of Fig. 1 showed that the yield rate of methane per day at each FL studied above were.133,.174,.212 and.23 L/d. These values FL=.5gVS/L Figure 3. Methane composition in biogas produced at different feed loadings. Figure 4 showed the alkalinity profile of each medium during digestion. Alkalinity rose to about 3mg/L within the first 4 days of digestion at FL.5-3.5gVS/L. The profile generally showed an increase in medium alkalinity as the FL increased. While the digestion medium of.5gvs/l remained at 3mg/L up till day 3, that at 1.5gVS/L and 3.5gVS/L rose to a maximum of 38mg/L and 4mg/L respectively at day 3. The 5.5gVS/L digestion medium had the highest alkalinity of 58mg/ have been found to have a buffering effect on VFA production, which lowers the ph during AD [7]. The increase in alkalinity at higher organic loading is partly due to the increase in the level of ammonianitrogen released from protein solubilization. Ammonia-nitrogen profile (Fig. 5) showed the range of NH 3-N released during the digestion to be within the range of 12mg/L - 28mg/L at FL of 5.5gVS/L. These 42

5 Ammonia-nitrogen (mg/l) Alkalinity (mg/l) ph values of ammonia-nitrogen are below the reported inhibition value of 6mg/L observed by researchers [13], [14]. The relatively lower value of NH 3-N observed during this study could be attributed to the higher C/N ratio of FW used (C/N=3) Figure 4. Change in alkalinity during the 3 days AD at different feed loadings. This resulted in less solubilization of protein to ammonia as compared to a similar study reported by [13] that produced ammonia-nitrogen to an inhibiting level of 6mg/L due to the rapid solubilization of protein as a result of lower C/N ratio of 15.5 of the FW substrate digested. The counter balancing effect of increasing alkalinity on the VFA produced at the varying feed loadings must have also been responsible for the relatively stable ph profile observed (Fig. 6) FL=.5gVS/L FL=.5gVS/L Figure 5. Variation of ammonia-nitrogen concentration during the mesophilic digestion of food waste for 3 days at different feed loadings Figure 6. ph profile during the anaerobic digestion of food waste at feed loading of.5-5.5gvs/l and 37 o C. 3.3 Treatment Efficiency The COD removal and volatile solid (VS) destruction are two indicators of treatment efficiency of an anaerobic digester. COD removal achieved were 98%, 96%, 9% and 8% at FL of.5, 1.5, 3.5 and 5.5gVS/L. Corresponding maximum VS destruction achieved at the above mentioned FL were 96%, 95%, 9% and 75% (see Fig. 2). The COD and VS reduction stated above showed that the treatment efficiency of the FW decreased as the FL to the digesters increased. The highest treatment efficiencies were obtained at FL of.5gvs/l with 98% COD reduction and 96% VS destruction. This trend seems logical, assuming that the same amount of inoculum or media concentration has been supplied to each batch of digester at increasing organic loading. Therefore, the higher the FW substrate concentration, the less the capacity of the microbial population to convert the intermediate products formed during AD to biogas [15], [16]. The treatment efficiencies attained can be attributed to increased C/N ratio and a balanced feed nutrient due to co-digestion. This has also helped to overcome limitations posed by high ammonia-nitrogen concentrations associated with digesting FW alone [17]. 4. Conclusion FL=.5g VS/L FL= 1.5g VS/L FL= 3.5g VS/L FL= 5.5g VS/L Increase in C/N ratio of the food waste treated through co-digestion has led to an increase in biogas methane yield up to.535l/gvs when compared with similar digestion by other researchers at lower C/N ratio. There was significant difference in biogas yield (ranging L/gVS) at each feed loading tested (i.e gvs/l). Digestion at feed loading of 3.5gVS/L attained the highest biogas methane yield during the 3 days batch anaerobic digestion. The ph of reaction medium was observed to decrease rapidly with the increase in feed loading. The lowering ph affected the methanogenic activity contributing largely to the 43

6 reduction in methane yield in biogas. These together affected the digester treatment efficiency, which decreased from 96% at feed loading of.5gvs/l to 75% at feed loading of 5.5gVS/L. REFERENCES [1] Periathamby, A., Hamid, F.S., & Khidzir, K. Evolution of Solid Waste Management in Malaysia: Impacts And Implications of the Solid Waste Bill, 27. Journal of Material Cycles and Waste Management, (29), 11: [2] Mata-Alvarez, J., Llabres, P., Cecchi, F., & Pavan, P. Anaerobic Digestion of the Barcelona Central Food Market Organic Wastes: Experimental Study. Bioresource Technology, (1992), 39: [3] Zhang, Y., Banks, C.J., & Heaven, S. Co-Digestion of Source Segregated Domestic Food Waste to Improve Process Stability. Bioresource Technology, (212), 114: [4] Cho, J.K., & Park, S.C. Biochemical Methane Potential and Solid State Anaerobic Digestion of Korean Food Wastes. Bioresource Technology, (1995), 52: [5] Bouallagui, H., Rachdi, B., Gannoun, H., & Hamdi, M. Mesophilic and Thermophilic Anaerobic Co-Digestion Of Abattoir Wastewater and Fruit And Vegetable Waste in Anaerobic Sequencing Batch Reactors. Biodegradation, (29), 2: [6] Elbeshbishy, E., & Nakhla, G. Comparative Study of the Effect of Ultrasonication on the Anaerobic Biodegradability of Food Waste in Single and Two Stage Systems. Bioresource Technology, (211), 12: [7] Zeshan, O., Karthikeyan P., & Visvanathan C. Effect of C/N Ratio and Ammonia-N Accumulation in a Pilot Scale Thermophilic Dry Anaerobic Digester. Bioresource Technology, (212), 113: [8] APHA Standard Methods for the Examination of Water And Waste Water. (25). Washington: American Public Health Association, American Water Works Association, Water Environment Federation. [9] Tchobanoglous, G., Theisen, H., & Vigil S.A. Integrated Solid Waste Management (international edition). (1993). New York: McGraw-Hill Inc. [1] Jiang, Y., Heaven, S., & Banks, C.J. Strategies for Stable Anaerobic Digestion Of Vegetable Waste. Renewable energy, (212), 44: [11] Siegert, I., & Banks, C.J. (25). The Effect of Volatile Fatty Acid Additions on the Anaerobic Digestion of Cellulose and Glucose in Batch Reactors. Process Biochemistry, (25), 4: [12] Zhang, R., El-Mashad, H.M., Hartman, K., Wang, F., Liu, G., Choate, C., & Gamble, P. Characterization of Food Waste as Feedstock for Anaerobic Digestion. Bioresource Technology, (27), 98: [13] Hansen, K. H., Angelidaki, I., & Ahring, B.K. Anaerobic Digestion of Swine Manure: Inhibition by Ammonia. Water Research, (1998), 32: [14] Ganidi, N., Tyrrel, S., & Cartmell, E. The Effect of Organic Loading Rate on Foam Initiation During Mesophilic Anaerobic Digestion of Municipal Wastewater Sludge. Bioresource Technology, (211), 12: [15] Liu, G., Zhang, R., El-Mashad, H. M., & Dong, R. Effect of Feed to Inoculum Ratio on Biogas Yields of Food and Green Wastes. Bioresource Technology, (29), 1: [16] Forster-Carneiro, T., Perez, M., & Romero, L. Influence of Total Solid and Inoculum Contents on Performance of Anaerobic Reactors Treating Food Waste. Bioresource Technology, (28), 99: [17] Yabu, H., Sakai, C., Fujiwara, T., Nishio, N., & Nakashimada, Y. Thermophilic Two Stage Dry Anaerobic Digestion of Model Garbage with Ammonia Stripping. Bioscience and Bioengineering, (21) 111: