- Email: [email protected]
Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste
Accepted Manuscript Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste Jong-Hun Park, Gopalakrishnan Kumar, Yeo-Myeong Yun, Joong-Chun Kwon, Sang-Hyoun Kim PII: DOI: Reference:
S0960-8524(17)31112-4 http://dx.doi.org/10.1016/j.biortech.2017.07.025 BITE 18445
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
30 April 2017 3 July 2017 6 July 2017
Please cite this article as: Park, J-H., Kumar, G., Yun, Y-M., Kwon, J-C., Kim, S-H., Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.07.025
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste Jong-Hun Park1,2, Gopalakrishnan Kumar1,3, Yeo-Myeong Yun4, Joong-Chun Kwon5, Sang-Hyoun Kim1,3 *
1
Sustainable Environmental Process Research Institute, Daegu University, Gyeongsan,
Gyeongbuk 38453, Republic of Korea 2
Civil, Environmental and Architectural Engineering, Korea University, Anam-Dong,
Seongbuk-gu, Seoul 02841, Republic of Korea 3
Department of Environmental Engineering, Daegu University, Gyeongsan, Gyeongbuk
38453, Republic of Korea 4
Department of Civil and Environmental Engineering, Korea Advanced Institute of Science
and Technology, 373-1 Guseong-dong, Yuseong-gu, Daejeon 34141, Republic of Korea 5
Ecodigm Co., Ltd., 10-6 Expo-ro 339 beon-gil, Yuseong-gu, Daejeon 34124, Republic of
Korea
*Corresponding author; Prof. Sang-Hyoun Kim, Department of Environmental Engineering, Daegu University, Gyeongsan, Gyeongbuk 38453, Republic of Korea Tel.: +82 53 8506691; Fax: +82 53 8506699; E-mail address: [email protected]
Abstract The effect of feeding mode and dilution was studied in anaerobic digestion of food waste. An upflow anaerobic digester with a settler was fed at six different organic loading rates (OLRs) from 4.6 to 8.6 kg COD/m3/d for 200 days. The highest methane productivity of 2.78 L CH4/L/d was achieved at 8.6 kg COD/m3/d during continuous feeding of diluted FW. Continuous feeding of diluted food waste showed more stable and efficient performance than stepwise feeding of undiluted food waste. Sharp increase in propionate concentration attributed towards deterioration of the digester performances in stepwise feeding of undiluted food waste. Microbial communities at various OLRs divulged that the microbial distribution in the continuous feeding of diluted food waste was not significantly perturbed despite the increase of OLR up to 8.6 kg COD/m3/d, which was contrast to the unstable distribution in stepwise feeding of undiluted food waste at 6.1 kg COD/m3/d.
Keywords: Food waste, Methane, Organic loading rate, Upflow anaerobic digester, Continuity in feeding, Dilution
1. Introduction Recycle of organic waste has been growing continuously over the last few years due to rapid development of industries for economic growth and the generation of enormous amount of waste. Direct landfill of organic waste results in serious water/soil/air pollution, odor/nuisance issues, and global warming [Fadel et al, 1997]. The South Korean government has prohibited the direct landfill of food waste, which accounts for the greatest proportion of organic waste in the country (408,000 ton/year in 2016). Instead, food waste is separately collected and recycled as compost or animal feed. However, the market value of the recycling products is low due to quality issues and high operation cost. Furthermore, composting and animal feed production discharge process wastewater containing pollutants equivalent to more than 60% of the food waste [Kim et al, 2012]. Anaerobic digestion (AD) can provide propitious solutions for both organic waste treatment and energy recovery by producing methane gas (CH4) [Lin et al, 2013, Lee et al, 2009]. Anaerobic digestion is well-suited to the non-sterile, ever changing, complex environment of food waste. Upflow anaerobic digester provides advantages, such as low structural complexity and area requirements [Shin et al, 2001]. The reactor configuration is extensively studied for the treatment of domestic sewage, industrial wastewater and toxic waste, and, more recently, for the treatment of wastewater from agricultural and agroindustrial practices [Duda & de Oliveira, 2009]. However, successful operation is hardly reported for waste containing particulates at organic loading rates (OLRs) over 5 kg COD/m3/d [Chen et al, 2008]. In earlier studies, AD process has been carried out with continuous feeding, batch mode and intermittent batch systems for the effects of various parameters [Zaiat & Foresti, 1996, de Barros et al, 2016, Rollon, 1999, Lee et al, 2010]. High contents of particulate matters and lipids of food waste make the continuous feeding quite tedious in case of real
time applications. For example, Lee et al. (2010) reported the semi continuous feeding of food waste for methane generation due to the high solid nature of the waste when using it directly to feed the digester. Either pretreatment of the food waste or the dilution of the food waste has been applied as an efficient strategy for continuous feeding into the AD reactor. Up to author’s knowledge, feeding method of the organic material in AD has not been studied earlier although it is very important to scale up the process. Another issue in food waste digestion is dilution. Total solids (TS) concentrations of food waste in Korea is normally in the range of 155 to 230 g/L [Kim et al, 2012, Lee et al, 2015, Lee et al, 2017, Kumar et al, 2016], which is not suitable either for wet (TS < 150 g/L) or dry digestion (TS > 250 g/L) [BSWA, 2004]. Dilution with liquid waste stream or water can enable the application of wet digestion, however, too much dilution would increase operational cost in digester heating and the amount of effluent, which may arise another treatment issue. Therefore, by considering the major issues mentioned above, this study aimed to investigate the effect of the feeding mode and dilution on methane production from food waste. An upflow anaerobic reactor was fed either continuously or stepwise at various OLR for 200 days. Not only anaerobic digestion performance, but also microbial community structure and diversity of these systems were analyzed to investigate the relationship between the metabolic profiles of the microbiome and the process performances.
2. Materials and methods
2.1 Seed sludge and food waste Seed sludge was obtained from a mesophilic sludge digester in a local domestic wastewater treatment plant in Gyeongsan, South Korea. Food waste was collected from the cafeteria of Daegu University, South Korea. Before fermentation, the food waste was shredded using an electrical blender and filtered through a stainless-steel sieve (US Mesh No. 10 with corresponding sieve opening of 2.00 mm).
2.2 Upflow anaerobic reactor The configuration of the food waste digestion system is presented in Fig. 1. An upflow anaerobic reactor with a working volume of 75.0 L (34.4 cm in diameter and 104.4 cm in length) was used in this study. Mechanical stirrer with rubber scrapers was installed in the upflow anaerobic reactor and operated at 1 rpm for scraping the solids sent to the settling tank located in weir and discharged as excess sludge in the bottom. The effluent was collected into a settler with a working volume of 5.0 L (16 cm in diameter and 42 cm in length) which discharged supernatant through a weir. The settled biomass was recirculated into the substrate feeding line of the upflow anaerobic reactor using a peristaltic pump. Two different feeding modes were adopted, 1) stepwise feeding with syringe pump and 2) continuous feeding with peristaltic pump. In stepwise feeding mode, food waste filled in a syringe barrel was fed into the reactor to simulate which daily required amount is injected at a short period of time [Satoto et al, 2009]. At 4.6 kg COD/m3/d, the average daily dose of 1.88 L/d of food waste was injected into the reactor, taking 15.6 mL for 5 minutes at 1 hour intervals. In case of 6.1 kg COD/m3/d, an average of 2.5 L/d of food waste was injected per
day, which was taking 20.83 mL for 5 minutes at 1 hour intervals. As it was too viscous to be added using peristaltic pump, food waste was diluted to around 111 g COD/L using tap water in continuous feeding mode. Characteristics of food waste according to the feeding modes are summarized in Table 1.
2.3 Bioreactor operation Methane production in a continuous system fed with diluted food waste has been reported earlier at mesophilic and thermophilic conditions [Han et al, 2012]. In the paper, similar production performances have been noted in both mesophilic and thermophilic regimes. In this research, mesophilic digestion at 35±1°C was chosen due to its lower operation cost. The reactor was inoculated with 50 L of seed sludge and the rest of working volume was filled with food waste. Subsequently, the headspace of the reactor was flushed with nitrogen gas for 5 min, pH was maintained between 7.3-7.5 by adding NaHCO3 to 2 g/L in the feed to avoid the acidification of the reactor. The OLR was varied from 4.6 to 8.6 kg COD/m3/d. Biogas generated from the upflow anaerobic reactor and settler were connected to a wet-gas meter (Jinghao Int., JH-LMF-2-P, China). In each condition, the operation period was longer than 17 days, and pseudo-steady-state performance with stable methane production (±10%) was maintained for at least 5 consecutive days to obtain performance data.
2.4 Analytical procedure The measured biogas production was adjusted to standard temperature and pressure condition. Biogas composition including methane and carbon dioxide was analyzed using a gas chromatography (SRI Instrument, SRI 310, USA) equipped with thermal conductivity detector (TCD) as described in our earlier report [Kim et al, 2004]. Samples for microbial analysis were stored at -80°C. For analysis of the soluble metabolic products (SMPs), high-
performance liquid chromatograph (HPLC, Waters) equipped with ultraviolet and refraction index detectors were employed to detect acetic acid, iso-butyric acid, propionic acid, butyric acid, formic acid and lactic acid [Park et al, 2013]. The standard methods of APHA (American Public Health Association) were used for the analysis of total solids (TS), chemical oxygen demands (COD), ammonia, total nitrogen (TN), alkalinity, and sodium [APHA, 1998]. Elemental analyzer (EA 1108, Fisons) and Inductively Coupled Plasma Spectrometer (Optima 7300DV, Perkin Elmer) were used for elemental analysis. The methane production rate (MPR) and methane yield (MY) from the fermentation liquor using the Buswell formula (1952) as shown in Eq. (1) and (2).
273 760 − PT LCH 4 MPR + = M CH4 ( T ) ⋅VCH 4 ( T ) ⋅ 273 + T 760 L⋅d
(1)
LCH 4 SCOD ⋅ QSubstrate MY = MPR ⋅ VReactor kgCOD
(2)
Where MCH4 = CH4 concentration (%), VCH4 = total gas volume (L) measured by gas meter, T = temperature, PT = Saturated water vapor pressure at T°C (mmHg), SCOD = Substrate COD concentration (kg/L), QSubstrate = The amount of substrate injected per day (L/d) and VReactor = reactor working volume.
2.5 Microbial community analysis
To analyze the microbial diversity, DNA in the mixed samples from the reactor was extracted using an Ultraclean Soil DNA Kit (Cat #12800-50; Mo Bio Laboratory Inc., USA). The extracted DNA was subsequently purified with an UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, CA, USA). A 20 ng aliquot of each sample DNA was used for a 50 ul PCR reaction. Partial sequences of the 16S rRNA gene including the variable V1–V3 region (corresponding to the 9–536 regions in Escherichia coli) were amplified from the
obtained DNA using primers: Arc8f (5’-TTCCGGTTGATCCYGCCGGA-3’) and Arc519r (5’-TTACCGCGGCKGCTG-3’) for methanogen, 16S universal primers 27F (5’GAGTTTGATCMTGGCTCAG-3’) and 800R (5’-TACCAGGGTATCTAATCC-3’) were used for amplifying 16s rRNA genes [DeLong, 1992, Barns et al, 1994]. A Fast Start High Fidelity PCR System (Roche) was used for PCR as previous described. After the PCR reaction, products were purified using AMPure beads (Beckman coulter). Then sequencing was performed using a 454 pyrosequencing Genome Sequencer FLX Titanium (Life Sciences, CT, USA), according to the manufacturer’s instructions, by a commercial sequencing facility (Macrogen, Seoul, Korea). The sequences generated from pyrosequencing were mainly analyzed with the software MOTHUR for pre-processing (quality-adjustment, barcode split), identification of operational taxonomic units (OTUs), taxonomic assignment, community comparison, and statistical analysis. The method of sequences filtration and trimming were performed as previous described [Nam et al, 2012].
3. Results and Discussion 3.1 Methane production performance
Fig. 2 depicts daily variation of methane production performances according to feeding/dilution modes and OLRs. The summary of the production performance is provided in Table 2. Undiluted food waste contained carbon (525.9 g/kg TS), oxygen (326.1 g/kg TS), hydrogen (71.8 g/kg TS), nitrogen (33.8 g/kg TS), potassium (15,612 mg/kg TS), sodium (16,153 mg/kg TS), phosphorus (6,558 mg/kg TS), iron (77.46 mg/kg TS), sulfur (1,698 mg/kg TS), magnesium (1,556 mg/kg TS), zinc (353 mg/kg TS), boron (203 mg/kg TS) and manganese (21 mg/kg TS). When feeding undiluted food waste stepwise at 4.6 kg COD/m3/d, the pseudo-steady-state CH4 yield and CH4 production rate were attained as 232 L CH4/kg COD, and 0.98 L CH4/L/d, respectively. At 6.1 kg COD/m3/d of OLR, CH4 yield and CH4 production rate reached up to 320 L/kg COD and 1.8 L/L/d, respectively, once, but gradually declined. Furthermore, a performance deterioration in methane production happened from day 95 to day 99. It would result from the blockage of recirculation line caused by excessively accumulated mineralized residues. The recirculation of 20 Q was intended to prevent clogging of the viscous feedstock inside the process. The hindered recirculation caused build-up of the sticky feedstock, which consequently resulted in the local overloading and insufficient mixing between feedstock and anaerobic microflora. In day 99, organic acid concentration increased to 1,265 mg COD/L, which is discussed in detail in 3.2, and pH decrease to 6.2. To solve this problem, three actions were conducted on day 99. i) Recirculation line was substituted by a new one, ii) 20 L of mixed liquor in the process was replaced by 20 g/L of NaHCO3 solution, and iii) OLR was reduced to 4.6 kg COD/m3/d. The measures successfully recovered the process performance, therefore, OLR was returned to 6.1 kg COD/m3/d on day 104. The pseudo-state methanogenic performance, 266 L CH4/kg COD
and 1.51 L CH4/L/d, did not reach to the temporary maximum values. The low and unstable digestion performance would result from the high solid nature of food waste and the incontinuity of feeding. The viscous substrate and stepwise feeding would result in insufficient contact between microorganisms and substrate, decrease of effective volume due to settling and scum formation, and high-pulse dosage. After switching to continuous feeding of diluted food waste, performance stability was remarkably enhanced as shown in Fig. 1. Recirculation ratio gradually decreased to 10 Q as the rheology of the mixed liquor in bioreactor was improved by feedstock dilution and increased flow rate. Diluted food waste contained carbon (411.5 g/kg TS), oxygen (392.8 g/kg TS), hydrogen (71.2 g/kg TS), nitrogen (42.7 g/kg TS), calcium (26,267 mg/kg TS), potassium (22,946 mg/kg TS), sodium (22,194 mg/kg TS), phosphorus (9,163 mg/kg TS), iron (1,137 mg/kg TS), sulfur (100 mg/kg TS), zinc (72.3 mg/kg TS), nickel (5 mg/kg TS) and tungsten (5 mg/kg TS). Although ammonia concentration increased up to 4,110 mg N/L, there was no consequence of ammonia inhibition. The uprising of OLR from 6.2 kg COD/m3/d to 8.6 kg COD/m3/d resulted in the average CH4 production rate and CH4 yield ascend from 1.43 to 2.78 L CH4/L/d and 247 to 325 L CH4/g COD, respectively. Continuous feeding with dilution enabled stable operation of upflow reactor configuration with settler, which resulted in the stable and efficient performance. The increased CH4 production rate at higher OLR directly related to the enhanced revenue and saved process volume, which would compensate the disadvantage of dilution, increase of effluent and heating demands in dilution.
3.2 Effluent organic acids
Total concentration and distribution of organic acids are a vital signal in digesters. Fig. 3 presents the data in this study. At the pseudo-state condition of each OLR, total organic acids concentration was lower than 450 mg COD/L, which meant methanogenesis and
acidogenesis were well balanced. The most abundant organic acid in the stable methanogenic performance was acetic acid, which was 45 to 55% of total organic acids. However, at the deterioration period from day 95 to 99, the total organic acid drastically elevated up to 1,265 mg COD/L. The sharp decrease implied the broken balance between acidogenesis and methanogenesis. Organic acid composition was also different from that in stable methane production. Propionic acid concentration was as high as 687 mg COD/L. In pseudo-state conditions, propionic acid was less than 10 mg/L. It is known that accumulation of propionic acid is the sign for the damaged balance between hydrogen producing acidogenesis and hydrogen consuming methanogenesis [Shin et al, 2011]. This study confirmed that monitoring total organic acid and propionic acid would be useful to determine the deterioration of a digester.
3.3 Microbial community
Archeal community distribution is shown in Fig. 4. Methanosaeta spp. was identified in a very small in fraction. It is known that they are inhibited at acetic acid over 150 mg/L [Conklin et al, 2006], same as in this case. Methanosarcina spp., known resistant to acetic acid, figured to be relatively dominant [Barredo & Evison, 1991]. According to Cho et al. (2013), Methanosarcina spp. existing in cluster form allows it to be high in volume/surface, while for Methanosaeta spp. in filamentous form, it is low in volume/surface, making it vulnerable to toxins from outside. The high OLR in this study resulted in relatively high acetic acid concentration, and consequently Methanosarcina spp. dominant methanogenic community. Methanobacterium spp. is an alkaliphilic and halotolerant group, which is consistent with the neutral to weak alkali pH in this experiment [Galiano et al, 2017]. Fig 5. compares the bacterial community at genus level. At the deterioration period, the dominance of Proteiniphilum spp., was noticed. Proteiniphilum spp. are known to be
responsible for production of propionic acid [Chen & Dong, 2005, Kim et al, 2015], which is consistent to the accumulation of propionic acid at the condition. [Shin et al, 2011]. When food waste was diluted and fed continuously, the distribution of bacteria and archea showed little change according to OLR. The stable methane production performance of the continuous feeding with dilution would be attributed to stable distribution of microorganisms.
3.4 Prominence of this study
Korean food waste has marginal solid content, which is difficult to be applied to either dry digestion or wet digestion. In this research, stepwise and continuous feeding methods with diluted and undiluted food waste have been proposed an efficient strategy towards stable and improved methane production. The results are quite promising and also compared with other studies to evaluate the importance of this report. Recently thermophilic process is highly getting attention due to the sanitation and also high rate of hydrolysis efficiency. For example, a study by Kumar et al. (2016), conducted the AD process with food waste at mesophilic and thermophilic regimes and reported that thermophilic temperature resulted in better production performances, in spite of the consideration towards the cost of the process operation, when the process is at thermophilic and authors reported that 264 L CH4/kg TS as methane yield. Another thermophilic study by Lee et al. (2010) narrated that recirculation of digester sludge could efficiently increase the production performance and yielded about 287 L CH4/kg TS of food waste. However, the thermophilic condition requires energy for process heating. This study was conducted at mesophilic condition, but showed higher methane yield, 374 L CH4/kg TS, than the reported values. It implied that the reactor configuration and operating conditions in this study provided criteria for efficient anaerobic digestion of Korean food waste.
While comparing the studies with mesophilic conditions, many authors reported that co-digestion could improve the synergistic relationship between the substrates and ultimately result in the higher methane production. A study by Neves et al. (2009) found out that codigestion with oil at continuous mode ended up with the methane productivity of 0.51 L CH4/L/d and 96 L CH4/kg TS, respectively. Meanwhile co-digestion with piggery wastewater conducted by Zhang et al. (2011) mentioned that methane yield of 396 L CH4/kg TS, and this result is quite close to the results obtained in this study about 395 L CH4/kg TS as methane yield and methane productivity of 2.78 L CH4/L/d. Very recently, researches moved to the direction of 2 stage fermentation, which could efficiently enhance the production rates and yield due to the efficient hydrolysis at first stage and also the production of VFAs, which are the direct substrate for methanogens to utilize towards biomethanation. Zhua et al. (2011) performed 2-stage experiments with the hydrogen fermentation effluent demonstrated that methane yield of 522 L CH4/kg TS could be achieved, additionally, Siddiqui et al. (2011) also proved that 617.6 L CH4/kg TS could be attained while using 2 stage fermentation of food waste. These values are quite higher compared to our results, however, in terms of production rate, our value is the highest among the mostly studied report, and this might be attributed to the high OLR enabled by the continuous feeding of diluted food waste to the upflow digester configuration. Generally, microbial community information has been rare for anaerobic digestion, although it is essential towards the metabolic background of the process. This study showed that Methanosarcina spp. and Methanobacterium spp. rather than Methanosaeta spp. were dominant in the digester with high OLR, and Proteiniphilum spp. was flourished at unstable condition. Further investigation on scaled up operation would bestow novel insights for the industrial scale applications.
4. Conclusions
Two feeding strategies were examined for an upflow digester with a settler. Continuous feeding with dilution showed superior performance and stability compared to stepwise feeding without dilution. The peak methane yield and methane productivity were 325 L/kg COD and 2.78 L CH4/L/d, respectively on 8.6 kg COD/m3/d. Stepwise feeding without dilution showed lower performance and also instability. Increase of Proteiniphilum spp. and concurrent propionic acid accumulation were found in deterioration period. Dilution of food waste would result in increased biogas production and digestion reliability, which would compensate the increased amount of effluent.
Acknowledgements
This work was supported by National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2017R1A2A2A07000900).
References
1.
American Public Health Association, A.W.W.A., Water Environment Federation, 1998. Standard methods for the examination of water and wastewater, 20th ed. APHA, AWWA, and WEF, Washington, D.C., USA.
2.
Barns S. M., Fundyga R. E., Jeffries M. W., Pace N. R., 1994. Remarkable archaeal diversity detected in a Yellowstone National Park hot spring environment. P Natl. Acad. Sci. USA. 91 (5), 1609–1613.
3.
Barredo M. S., Evison L. M., 1991. Effect of Propionate Toxicity on MethanogenEnriched Sludge, Methanobrevibacter smithii, and Methanospirillum hungatii at Different pH Values. Appl. Environ. Microb. 57 (6), 1764-1769.
4.
Bluestem Solid Waste Agency, Iowa Department of Natural Resources, 2004. Anaerobic
Digestion Feasibility Study. R.W. Beck Inc., Washington, D.C., USA. 5.
Buswell, A., Mueller, H., 1952. Mechanism of methane fermentation. Ind. Eng. Chem. 44, 550–562.
6.
Chen S., Dong X., 2005. Proteiniphilum acetatigenes gen. nov., sp. nov., from a UASB reactor treating brewery wastewater. Int. J. Syst. Evol. Micr. 55 (6), 2257-2261.
7.
Chen, Y., Cheng, J.-J., Creamer, K. S., 2008. Inhibition of anaerobic digestion process: A review. Bioresource Technol. 99 (10), 4044-4064.
8.
Cho S.-K., Im W.-T., Kim D.-H., Kim M.-H., Shin H.-S., Oh S.-E., 2013. Dry anaerobic digestion of food waste under mesophilic conditions: performance and methanogenic community analysis. Bioresource Technol. 131, 210–217.
9.
Conklin A., Stensel H. D., Ferguson J., 2006. Growth kinetics and competition between Methanosarcina and Methanosaeta in mesophilic anaerobic digestion. Water Environ. Res. 78 (5), 486-496.
10. de Barros V. G., Duda R. M., de Oliveira R. A., 2016. Biomethane production from vinasse in upflow anaerobic sludge blanket reactors inoculated with granular sludge. Braz. J. Microbiol. 47 (3), 628-639. 11. DeLong E. F., 1992. Archaea in coastal marine environments. P Natl. Acad. Sci. USA. 89 (12), 5685-5689. 12. Duda R. M., de Oliveira R. A., 2009. Tratamento de águas residuárias de suinocultura em reator UASB e filtro anaeróbio em série seguidos de filtro biológico percolador. Eng. Sanitaria E Ambiental. 16 (1), 91–100. 13. Fadel E., Findikakis M., Angelos N., Leckie J. O., 1997. Environmental Impacts of Solid Waste Landfilling. J. Environ. Manage. 50 (1), 1-25. 14. Galiano, M. C., Ismail S. B., Stams A. J. M., Plugge C. M., Temmink H., Van Lier J. B., 2017. Biofilm formation and granule properties in anaerobic digestion at high salinity. Water Res. 121, 61-71. 15. Han M.-J., Behera S.-K., Park H.-S., 2012. Anaerobic co-digestion of food waste leachate and piggery wastewater for methane production: statistical optimization of key process parameters. J. Chem. Technol. Biot. 87 (11), 1541–1550.
16. Kim S.-H., Cheon H.-C., Lee C.-Y., 2012. Enhancement of hydrogen production by recycling of methanogenic effluent in two-phase fermentation of food waste. Int. J. Hydrogen Energ. 37 (18), 13777-13782. 17. Kim W., Shin S.-G., Han G., Cho K., Hwang S., 2015. Structures of microbial communities found in anaerobic batch runs that produce methane from propionic acid— Seeded from full-scale anaerobic digesters above a certain threshold. J. Biotechnol. 214. 192-198. 18. Kim, S.-H., Han, S.-K., Shin, H.-S., 2004. Feasibility of biohydrogen production by anaerobic co-digestion of food waste and sewage sludge. Int. J. Hydrogen Energ. 29 (15), 1607-1616. 19. Kumar G., Sivagurunathan P., Park J.-H., Kim S.-H., 2016. Anaerobic digestion of food waste to methane at various organic loading rates (OLRs) and hydraulic retention times (HRTs): Thermophilic vs. mesophilic regimes. Environ. Eng. Res. 21 (1), 69-73. 20. Lee D.-H., Behera S. K., Kim J.-W., Park H.-S., 2009. Methane production potential of leachate generated from Korean food waste recycling facilities: a lab-scale study. Waste Manage. 29 (2), 876–882. 21. Lee D.-J, Lee S.-Y., Bae J.-S., Kang J.-G., Kim K.-H., Rhee S.-S., Park J.-H., Cho J.-S., Chung J., Seo D.-C., 2015. Effect of Volatile Fatty Acid Concentration on Anaerobic Degradation Rate from Field Anaerobic Digestion Facilities Treating Food Waste Leachate in South Korea. J. Chem. 1-9. 22. Lee D.-J., Yoon Y.-M., Choi I.-W., Bae J.-S., Seo D.-C., 2017. Effect of seasonal variations of organic loading rate and acid phase on methane yield of food waste leachate in South Korea. Appl. Biol. Chem. 60 (1), 87-93. 23. Lee D.-Y., Ebie Y., Xu K.-Q., Li Y.-Y., Inamori Y, 2010. Continuous H2 and CH4 production from high-solid food waste in the two-stage thermophilic fermentation process with the recirculation of digester sludge. Bioresource Technol. 101 (1), S42-S47. 24. Lin C. S. K., Pfaltzgraff L. A., Herrero-Davila L., Mubofu E. B., Abderrahim S., Clark J. H., Koutinas A. A., Kopsahelis N., Stamatelatou K., Dickson F., Thankappan S., Mohamed Z., Brocklesby R., Luque R., 2013. Food waste as a valuable resource for the production of chemicals, materials and fuels. Current situation and global perspective.
Energy Environ. Sci. 6 (2), 426-464. 25. Lin J., Gan J. Z. L, Li P., Wang F. L. K, Chen L., Gan H., 2011. Effects of mixture ratio on anaerobic co-digestion with fruit and vegetable waste and food waste of China. J. Environ. sci. 23 (8), 1403-1408. 26. Nam Y.-D., Lee S.-Y., Lim S.-I., 2012. Microbial community analysis of Korean soybean pastes by next-generation sequencing, Int. J. Food Microbiol. 155 (1-2), 36–42. 27. Neves L., Oliveira R., Alves M.M., 2009. Co-digestion of cow manure, food waste and intermittent input of fat. Bioresource Technol. 100 (6),1957-1962. 28. Park, J.-H., Kim, S.-H., Park, H.-D., Lim, D.-J., Yoon, J.-J., 2013. Feasibility of anaerobic digestion from bioethanol fermentation residue. Bioresource Technol. 141, 177-183. 29. Rollon A. P., 1999. Anaerobic digestion of fish processing wastewater with special emphasis on hydrolysis of suspended solids. 1st ed. A. A. Balkema, P. O. Box 1675, 3000 BR Rotterdam, Netherlands. 30. Satoto E. N., Claudia G., Josef W., 2009. Foodwaste as a co-substrate in a fed-batch anaerobic biowaste digester for constant biogas supply. Water Sci Technol. 59 (6), 11691178. 31. Shin H.-S., Han S.-K., Song Y.-C., Lee C.-Y., 2001. Performance of UASB reactor treating leachate from acidogenic fermenter in the two-phase anaerobic digestion of food waste. Water Res. 35 (14), 3441–3447. 32. Shin, S.-G., Yoo, S., Hwang, K., Song, M., Kim, W., Han, G., Hwang, S., 2011. Dynamics of transitional acidogenic community along with methanogenic population during anaerobic digestion of swine wastewater. Process Biochem. 46 (8), 1607–1613. 33. Siddiqui Z., Horan N. J., Salter M., 2011. Energy optimization from co-digested waste using a two-phase process to generate hydrogen and methane. Int. J. Hydrogen Energ. 36 (8), 4792-4799. 34. Veluchamy C., Kalamdhad A. S., 2017. Enhanced methane production and its kinetics model of thermally pretreated lignocellulose waste material. Bioresource Technol. 241, 1-9.
35. Zaiat M., Foresti E., 1996. Cell wash-out and external mass transfer resistance in horizontal-flow anaerobic immobilized sludge reactor. Water Res. 30 (10), 2435–2439. 36. Zhang L, Lee Y.-W., Jahng D., 2011. Anaerobic co-digestion of food waste and piggery wastewater: Focusing on the role of trace elements. Bioresource Technol. 102 (8), 50485059. 37. Zhen G., Lu X., Kobayashi T., Kumar G., Xua K., 2016. Anaerobic co-digestion on improving methane production from mixed microalgae (Scenedesmus sp., Chlorella sp.) and food waste: Kinetic modeling and synergistic impact evaluation. Chem Eng J. 299, 332-341. 38. Zhua H., Parker W., Conidi D., Basnar P., Seto P., 2011. Eliminating methanogenic activity in hydrogen reactor to improve biogas production in a two-stage anaerobic digestion process co-digesting municipal food waste and sewage sludge. Bioresource Technol. 102 (14), 7086-7092.
Figure Captions Figure 1: Schematic diagram of the anaerobic digester system. Figure 2: Daily variation of methane production performance. Figure 3: Total concentration and distribution of organic acids according to OLR. All data
were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99). Figure 4: 454 pyrosequencing analysis of archaeal population in the pilot-scale anaerobic
digester. All data were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99). Figure 5: 454 pyrosequencing analysis of bacterial population in the pilot-scale anaerobic
digester. All data were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99).
Figure 1
Figure 2
Figure 3 1400
Distribution of SMPs (%)
1200 80 1000 60
800 600
40
400 20 200 0
0 4.6
6.1
6.1*
6.2
7.4 3
OLRs (kg COD/m /d)
7.9
8.6
Total organic acids (mg COD/L)
100
Acetic acid Propionic acid Butyric acid iso-butyric acid Formic acid Lactic acid Total organic acids
Figure 4
Figure 5
Table Captions Table 1: Characteristics of food waste Table 2: Digestion performances at pseudo-state conditions of each OLR
Table 1 Mode Undiluted Diluted
TCOD
SCOD
TSS
184,330 ±2,656 111,240 ±1,063
135,490 ±2,350 84,740 ±2,050
132,430 ±3,640 80,480 ±1,566
VSS (mg/L) 97,155 ±2,075 61,480 ±1,324
TS
VS
TN
170,230 ±4,870 96,990 ±2,542
158,270 ±2,371 88,480 ±2,042
5,423 ±213 4,280 ±159
Table 2 Item Average OLR (kg COD/m3/d) HRT (d) SRT (d) Average feed TCOD (g/L) Feeding mode Data collection Recirculation (Q) CH4 yield (L/kg COD) CH4 productivity (L CH4/L/d) pH Effluent Ammonia (mg N/L) Effluent TSS (mg/L) Effluent VSS (mg/L) Effluent TCOD (mg/L) Effluent SCOD (mg/L) Effluent organic acids (mg COD/L) TCOD removal efficiency (%) Alkalinity (mg CaCO3/L)
Parameter/Performance 6.2 7.4
4.6
6.1
7.9
8.6
40 40.87 184
30 30.83 184
18 18.11 111
15 15.39 111
13 13.30 111
12 111
Stepwise day 12 to 31 20 232 0.98
Stepwise day 63 to 88 20 266 1.51
Continuous day 127 to 142 20 247 1.43
Continuous day 143 to 164 15 299 2.01
Continuous day 165 to 186 10 326 2.52
Continuous day 187 to 203 10 325 2.78
8.0 975
7.5 1,310
7.7 2,120
7.7 2,470
7.5 3,060
7.5 4,110
24,400 19,770 32,250 2,500 157
23,140 12,970 35,150 2,180 390
24,070 19,230 33,810 2,050 407
23,140 12,970 27,930 1,720 449
17,550 14,330 31,970 1,500 413
24,770 18,650 29,910 1,820 402
82.5
80.9
69.5
74.8
71.2
73.1
4,720
5,120
2,740
2,620
3,061
2,360
26
Highlights
Feeding mode and dilution of food waste were examined.
Continuous feeding of diluted food waste showed stable performance.
The maximum productivity of 2.78 L CH4/L/d was achieved at 8.6 kg COD/m3/d.
454 pyrosequencing analysis divulged microbial distribution and diversity.
27
S0960-8524(17)31112-4 http://dx.doi.org/10.1016/j.biortech.2017.07.025 BITE 18445
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
30 April 2017 3 July 2017 6 July 2017
Please cite this article as: Park, J-H., Kumar, G., Yun, Y-M., Kwon, J-C., Kim, S-H., Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.07.025
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Effect of feeding mode and dilution on the performance and microbial community population in anaerobic digestion of food waste Jong-Hun Park1,2, Gopalakrishnan Kumar1,3, Yeo-Myeong Yun4, Joong-Chun Kwon5, Sang-Hyoun Kim1,3 *
1
Sustainable Environmental Process Research Institute, Daegu University, Gyeongsan,
Gyeongbuk 38453, Republic of Korea 2
Civil, Environmental and Architectural Engineering, Korea University, Anam-Dong,
Seongbuk-gu, Seoul 02841, Republic of Korea 3
Department of Environmental Engineering, Daegu University, Gyeongsan, Gyeongbuk
38453, Republic of Korea 4
Department of Civil and Environmental Engineering, Korea Advanced Institute of Science
and Technology, 373-1 Guseong-dong, Yuseong-gu, Daejeon 34141, Republic of Korea 5
Ecodigm Co., Ltd., 10-6 Expo-ro 339 beon-gil, Yuseong-gu, Daejeon 34124, Republic of
Korea
*Corresponding author; Prof. Sang-Hyoun Kim, Department of Environmental Engineering, Daegu University, Gyeongsan, Gyeongbuk 38453, Republic of Korea Tel.: +82 53 8506691; Fax: +82 53 8506699; E-mail address: [email protected]
Abstract The effect of feeding mode and dilution was studied in anaerobic digestion of food waste. An upflow anaerobic digester with a settler was fed at six different organic loading rates (OLRs) from 4.6 to 8.6 kg COD/m3/d for 200 days. The highest methane productivity of 2.78 L CH4/L/d was achieved at 8.6 kg COD/m3/d during continuous feeding of diluted FW. Continuous feeding of diluted food waste showed more stable and efficient performance than stepwise feeding of undiluted food waste. Sharp increase in propionate concentration attributed towards deterioration of the digester performances in stepwise feeding of undiluted food waste. Microbial communities at various OLRs divulged that the microbial distribution in the continuous feeding of diluted food waste was not significantly perturbed despite the increase of OLR up to 8.6 kg COD/m3/d, which was contrast to the unstable distribution in stepwise feeding of undiluted food waste at 6.1 kg COD/m3/d.
Keywords: Food waste, Methane, Organic loading rate, Upflow anaerobic digester, Continuity in feeding, Dilution
1. Introduction Recycle of organic waste has been growing continuously over the last few years due to rapid development of industries for economic growth and the generation of enormous amount of waste. Direct landfill of organic waste results in serious water/soil/air pollution, odor/nuisance issues, and global warming [Fadel et al, 1997]. The South Korean government has prohibited the direct landfill of food waste, which accounts for the greatest proportion of organic waste in the country (408,000 ton/year in 2016). Instead, food waste is separately collected and recycled as compost or animal feed. However, the market value of the recycling products is low due to quality issues and high operation cost. Furthermore, composting and animal feed production discharge process wastewater containing pollutants equivalent to more than 60% of the food waste [Kim et al, 2012]. Anaerobic digestion (AD) can provide propitious solutions for both organic waste treatment and energy recovery by producing methane gas (CH4) [Lin et al, 2013, Lee et al, 2009]. Anaerobic digestion is well-suited to the non-sterile, ever changing, complex environment of food waste. Upflow anaerobic digester provides advantages, such as low structural complexity and area requirements [Shin et al, 2001]. The reactor configuration is extensively studied for the treatment of domestic sewage, industrial wastewater and toxic waste, and, more recently, for the treatment of wastewater from agricultural and agroindustrial practices [Duda & de Oliveira, 2009]. However, successful operation is hardly reported for waste containing particulates at organic loading rates (OLRs) over 5 kg COD/m3/d [Chen et al, 2008]. In earlier studies, AD process has been carried out with continuous feeding, batch mode and intermittent batch systems for the effects of various parameters [Zaiat & Foresti, 1996, de Barros et al, 2016, Rollon, 1999, Lee et al, 2010]. High contents of particulate matters and lipids of food waste make the continuous feeding quite tedious in case of real
time applications. For example, Lee et al. (2010) reported the semi continuous feeding of food waste for methane generation due to the high solid nature of the waste when using it directly to feed the digester. Either pretreatment of the food waste or the dilution of the food waste has been applied as an efficient strategy for continuous feeding into the AD reactor. Up to author’s knowledge, feeding method of the organic material in AD has not been studied earlier although it is very important to scale up the process. Another issue in food waste digestion is dilution. Total solids (TS) concentrations of food waste in Korea is normally in the range of 155 to 230 g/L [Kim et al, 2012, Lee et al, 2015, Lee et al, 2017, Kumar et al, 2016], which is not suitable either for wet (TS < 150 g/L) or dry digestion (TS > 250 g/L) [BSWA, 2004]. Dilution with liquid waste stream or water can enable the application of wet digestion, however, too much dilution would increase operational cost in digester heating and the amount of effluent, which may arise another treatment issue. Therefore, by considering the major issues mentioned above, this study aimed to investigate the effect of the feeding mode and dilution on methane production from food waste. An upflow anaerobic reactor was fed either continuously or stepwise at various OLR for 200 days. Not only anaerobic digestion performance, but also microbial community structure and diversity of these systems were analyzed to investigate the relationship between the metabolic profiles of the microbiome and the process performances.
2. Materials and methods
2.1 Seed sludge and food waste Seed sludge was obtained from a mesophilic sludge digester in a local domestic wastewater treatment plant in Gyeongsan, South Korea. Food waste was collected from the cafeteria of Daegu University, South Korea. Before fermentation, the food waste was shredded using an electrical blender and filtered through a stainless-steel sieve (US Mesh No. 10 with corresponding sieve opening of 2.00 mm).
2.2 Upflow anaerobic reactor The configuration of the food waste digestion system is presented in Fig. 1. An upflow anaerobic reactor with a working volume of 75.0 L (34.4 cm in diameter and 104.4 cm in length) was used in this study. Mechanical stirrer with rubber scrapers was installed in the upflow anaerobic reactor and operated at 1 rpm for scraping the solids sent to the settling tank located in weir and discharged as excess sludge in the bottom. The effluent was collected into a settler with a working volume of 5.0 L (16 cm in diameter and 42 cm in length) which discharged supernatant through a weir. The settled biomass was recirculated into the substrate feeding line of the upflow anaerobic reactor using a peristaltic pump. Two different feeding modes were adopted, 1) stepwise feeding with syringe pump and 2) continuous feeding with peristaltic pump. In stepwise feeding mode, food waste filled in a syringe barrel was fed into the reactor to simulate which daily required amount is injected at a short period of time [Satoto et al, 2009]. At 4.6 kg COD/m3/d, the average daily dose of 1.88 L/d of food waste was injected into the reactor, taking 15.6 mL for 5 minutes at 1 hour intervals. In case of 6.1 kg COD/m3/d, an average of 2.5 L/d of food waste was injected per
day, which was taking 20.83 mL for 5 minutes at 1 hour intervals. As it was too viscous to be added using peristaltic pump, food waste was diluted to around 111 g COD/L using tap water in continuous feeding mode. Characteristics of food waste according to the feeding modes are summarized in Table 1.
2.3 Bioreactor operation Methane production in a continuous system fed with diluted food waste has been reported earlier at mesophilic and thermophilic conditions [Han et al, 2012]. In the paper, similar production performances have been noted in both mesophilic and thermophilic regimes. In this research, mesophilic digestion at 35±1°C was chosen due to its lower operation cost. The reactor was inoculated with 50 L of seed sludge and the rest of working volume was filled with food waste. Subsequently, the headspace of the reactor was flushed with nitrogen gas for 5 min, pH was maintained between 7.3-7.5 by adding NaHCO3 to 2 g/L in the feed to avoid the acidification of the reactor. The OLR was varied from 4.6 to 8.6 kg COD/m3/d. Biogas generated from the upflow anaerobic reactor and settler were connected to a wet-gas meter (Jinghao Int., JH-LMF-2-P, China). In each condition, the operation period was longer than 17 days, and pseudo-steady-state performance with stable methane production (±10%) was maintained for at least 5 consecutive days to obtain performance data.
2.4 Analytical procedure The measured biogas production was adjusted to standard temperature and pressure condition. Biogas composition including methane and carbon dioxide was analyzed using a gas chromatography (SRI Instrument, SRI 310, USA) equipped with thermal conductivity detector (TCD) as described in our earlier report [Kim et al, 2004]. Samples for microbial analysis were stored at -80°C. For analysis of the soluble metabolic products (SMPs), high-
performance liquid chromatograph (HPLC, Waters) equipped with ultraviolet and refraction index detectors were employed to detect acetic acid, iso-butyric acid, propionic acid, butyric acid, formic acid and lactic acid [Park et al, 2013]. The standard methods of APHA (American Public Health Association) were used for the analysis of total solids (TS), chemical oxygen demands (COD), ammonia, total nitrogen (TN), alkalinity, and sodium [APHA, 1998]. Elemental analyzer (EA 1108, Fisons) and Inductively Coupled Plasma Spectrometer (Optima 7300DV, Perkin Elmer) were used for elemental analysis. The methane production rate (MPR) and methane yield (MY) from the fermentation liquor using the Buswell formula (1952) as shown in Eq. (1) and (2).
273 760 − PT LCH 4 MPR + = M CH4 ( T ) ⋅VCH 4 ( T ) ⋅ 273 + T 760 L⋅d
(1)
LCH 4 SCOD ⋅ QSubstrate MY = MPR ⋅ VReactor kgCOD
(2)
Where MCH4 = CH4 concentration (%), VCH4 = total gas volume (L) measured by gas meter, T = temperature, PT = Saturated water vapor pressure at T°C (mmHg), SCOD = Substrate COD concentration (kg/L), QSubstrate = The amount of substrate injected per day (L/d) and VReactor = reactor working volume.
2.5 Microbial community analysis
To analyze the microbial diversity, DNA in the mixed samples from the reactor was extracted using an Ultraclean Soil DNA Kit (Cat #12800-50; Mo Bio Laboratory Inc., USA). The extracted DNA was subsequently purified with an UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, CA, USA). A 20 ng aliquot of each sample DNA was used for a 50 ul PCR reaction. Partial sequences of the 16S rRNA gene including the variable V1–V3 region (corresponding to the 9–536 regions in Escherichia coli) were amplified from the
obtained DNA using primers: Arc8f (5’-TTCCGGTTGATCCYGCCGGA-3’) and Arc519r (5’-TTACCGCGGCKGCTG-3’) for methanogen, 16S universal primers 27F (5’GAGTTTGATCMTGGCTCAG-3’) and 800R (5’-TACCAGGGTATCTAATCC-3’) were used for amplifying 16s rRNA genes [DeLong, 1992, Barns et al, 1994]. A Fast Start High Fidelity PCR System (Roche) was used for PCR as previous described. After the PCR reaction, products were purified using AMPure beads (Beckman coulter). Then sequencing was performed using a 454 pyrosequencing Genome Sequencer FLX Titanium (Life Sciences, CT, USA), according to the manufacturer’s instructions, by a commercial sequencing facility (Macrogen, Seoul, Korea). The sequences generated from pyrosequencing were mainly analyzed with the software MOTHUR for pre-processing (quality-adjustment, barcode split), identification of operational taxonomic units (OTUs), taxonomic assignment, community comparison, and statistical analysis. The method of sequences filtration and trimming were performed as previous described [Nam et al, 2012].
3. Results and Discussion 3.1 Methane production performance
Fig. 2 depicts daily variation of methane production performances according to feeding/dilution modes and OLRs. The summary of the production performance is provided in Table 2. Undiluted food waste contained carbon (525.9 g/kg TS), oxygen (326.1 g/kg TS), hydrogen (71.8 g/kg TS), nitrogen (33.8 g/kg TS), potassium (15,612 mg/kg TS), sodium (16,153 mg/kg TS), phosphorus (6,558 mg/kg TS), iron (77.46 mg/kg TS), sulfur (1,698 mg/kg TS), magnesium (1,556 mg/kg TS), zinc (353 mg/kg TS), boron (203 mg/kg TS) and manganese (21 mg/kg TS). When feeding undiluted food waste stepwise at 4.6 kg COD/m3/d, the pseudo-steady-state CH4 yield and CH4 production rate were attained as 232 L CH4/kg COD, and 0.98 L CH4/L/d, respectively. At 6.1 kg COD/m3/d of OLR, CH4 yield and CH4 production rate reached up to 320 L/kg COD and 1.8 L/L/d, respectively, once, but gradually declined. Furthermore, a performance deterioration in methane production happened from day 95 to day 99. It would result from the blockage of recirculation line caused by excessively accumulated mineralized residues. The recirculation of 20 Q was intended to prevent clogging of the viscous feedstock inside the process. The hindered recirculation caused build-up of the sticky feedstock, which consequently resulted in the local overloading and insufficient mixing between feedstock and anaerobic microflora. In day 99, organic acid concentration increased to 1,265 mg COD/L, which is discussed in detail in 3.2, and pH decrease to 6.2. To solve this problem, three actions were conducted on day 99. i) Recirculation line was substituted by a new one, ii) 20 L of mixed liquor in the process was replaced by 20 g/L of NaHCO3 solution, and iii) OLR was reduced to 4.6 kg COD/m3/d. The measures successfully recovered the process performance, therefore, OLR was returned to 6.1 kg COD/m3/d on day 104. The pseudo-state methanogenic performance, 266 L CH4/kg COD
and 1.51 L CH4/L/d, did not reach to the temporary maximum values. The low and unstable digestion performance would result from the high solid nature of food waste and the incontinuity of feeding. The viscous substrate and stepwise feeding would result in insufficient contact between microorganisms and substrate, decrease of effective volume due to settling and scum formation, and high-pulse dosage. After switching to continuous feeding of diluted food waste, performance stability was remarkably enhanced as shown in Fig. 1. Recirculation ratio gradually decreased to 10 Q as the rheology of the mixed liquor in bioreactor was improved by feedstock dilution and increased flow rate. Diluted food waste contained carbon (411.5 g/kg TS), oxygen (392.8 g/kg TS), hydrogen (71.2 g/kg TS), nitrogen (42.7 g/kg TS), calcium (26,267 mg/kg TS), potassium (22,946 mg/kg TS), sodium (22,194 mg/kg TS), phosphorus (9,163 mg/kg TS), iron (1,137 mg/kg TS), sulfur (100 mg/kg TS), zinc (72.3 mg/kg TS), nickel (5 mg/kg TS) and tungsten (5 mg/kg TS). Although ammonia concentration increased up to 4,110 mg N/L, there was no consequence of ammonia inhibition. The uprising of OLR from 6.2 kg COD/m3/d to 8.6 kg COD/m3/d resulted in the average CH4 production rate and CH4 yield ascend from 1.43 to 2.78 L CH4/L/d and 247 to 325 L CH4/g COD, respectively. Continuous feeding with dilution enabled stable operation of upflow reactor configuration with settler, which resulted in the stable and efficient performance. The increased CH4 production rate at higher OLR directly related to the enhanced revenue and saved process volume, which would compensate the disadvantage of dilution, increase of effluent and heating demands in dilution.
3.2 Effluent organic acids
Total concentration and distribution of organic acids are a vital signal in digesters. Fig. 3 presents the data in this study. At the pseudo-state condition of each OLR, total organic acids concentration was lower than 450 mg COD/L, which meant methanogenesis and
acidogenesis were well balanced. The most abundant organic acid in the stable methanogenic performance was acetic acid, which was 45 to 55% of total organic acids. However, at the deterioration period from day 95 to 99, the total organic acid drastically elevated up to 1,265 mg COD/L. The sharp decrease implied the broken balance between acidogenesis and methanogenesis. Organic acid composition was also different from that in stable methane production. Propionic acid concentration was as high as 687 mg COD/L. In pseudo-state conditions, propionic acid was less than 10 mg/L. It is known that accumulation of propionic acid is the sign for the damaged balance between hydrogen producing acidogenesis and hydrogen consuming methanogenesis [Shin et al, 2011]. This study confirmed that monitoring total organic acid and propionic acid would be useful to determine the deterioration of a digester.
3.3 Microbial community
Archeal community distribution is shown in Fig. 4. Methanosaeta spp. was identified in a very small in fraction. It is known that they are inhibited at acetic acid over 150 mg/L [Conklin et al, 2006], same as in this case. Methanosarcina spp., known resistant to acetic acid, figured to be relatively dominant [Barredo & Evison, 1991]. According to Cho et al. (2013), Methanosarcina spp. existing in cluster form allows it to be high in volume/surface, while for Methanosaeta spp. in filamentous form, it is low in volume/surface, making it vulnerable to toxins from outside. The high OLR in this study resulted in relatively high acetic acid concentration, and consequently Methanosarcina spp. dominant methanogenic community. Methanobacterium spp. is an alkaliphilic and halotolerant group, which is consistent with the neutral to weak alkali pH in this experiment [Galiano et al, 2017]. Fig 5. compares the bacterial community at genus level. At the deterioration period, the dominance of Proteiniphilum spp., was noticed. Proteiniphilum spp. are known to be
responsible for production of propionic acid [Chen & Dong, 2005, Kim et al, 2015], which is consistent to the accumulation of propionic acid at the condition. [Shin et al, 2011]. When food waste was diluted and fed continuously, the distribution of bacteria and archea showed little change according to OLR. The stable methane production performance of the continuous feeding with dilution would be attributed to stable distribution of microorganisms.
3.4 Prominence of this study
Korean food waste has marginal solid content, which is difficult to be applied to either dry digestion or wet digestion. In this research, stepwise and continuous feeding methods with diluted and undiluted food waste have been proposed an efficient strategy towards stable and improved methane production. The results are quite promising and also compared with other studies to evaluate the importance of this report. Recently thermophilic process is highly getting attention due to the sanitation and also high rate of hydrolysis efficiency. For example, a study by Kumar et al. (2016), conducted the AD process with food waste at mesophilic and thermophilic regimes and reported that thermophilic temperature resulted in better production performances, in spite of the consideration towards the cost of the process operation, when the process is at thermophilic and authors reported that 264 L CH4/kg TS as methane yield. Another thermophilic study by Lee et al. (2010) narrated that recirculation of digester sludge could efficiently increase the production performance and yielded about 287 L CH4/kg TS of food waste. However, the thermophilic condition requires energy for process heating. This study was conducted at mesophilic condition, but showed higher methane yield, 374 L CH4/kg TS, than the reported values. It implied that the reactor configuration and operating conditions in this study provided criteria for efficient anaerobic digestion of Korean food waste.
While comparing the studies with mesophilic conditions, many authors reported that co-digestion could improve the synergistic relationship between the substrates and ultimately result in the higher methane production. A study by Neves et al. (2009) found out that codigestion with oil at continuous mode ended up with the methane productivity of 0.51 L CH4/L/d and 96 L CH4/kg TS, respectively. Meanwhile co-digestion with piggery wastewater conducted by Zhang et al. (2011) mentioned that methane yield of 396 L CH4/kg TS, and this result is quite close to the results obtained in this study about 395 L CH4/kg TS as methane yield and methane productivity of 2.78 L CH4/L/d. Very recently, researches moved to the direction of 2 stage fermentation, which could efficiently enhance the production rates and yield due to the efficient hydrolysis at first stage and also the production of VFAs, which are the direct substrate for methanogens to utilize towards biomethanation. Zhua et al. (2011) performed 2-stage experiments with the hydrogen fermentation effluent demonstrated that methane yield of 522 L CH4/kg TS could be achieved, additionally, Siddiqui et al. (2011) also proved that 617.6 L CH4/kg TS could be attained while using 2 stage fermentation of food waste. These values are quite higher compared to our results, however, in terms of production rate, our value is the highest among the mostly studied report, and this might be attributed to the high OLR enabled by the continuous feeding of diluted food waste to the upflow digester configuration. Generally, microbial community information has been rare for anaerobic digestion, although it is essential towards the metabolic background of the process. This study showed that Methanosarcina spp. and Methanobacterium spp. rather than Methanosaeta spp. were dominant in the digester with high OLR, and Proteiniphilum spp. was flourished at unstable condition. Further investigation on scaled up operation would bestow novel insights for the industrial scale applications.
4. Conclusions
Two feeding strategies were examined for an upflow digester with a settler. Continuous feeding with dilution showed superior performance and stability compared to stepwise feeding without dilution. The peak methane yield and methane productivity were 325 L/kg COD and 2.78 L CH4/L/d, respectively on 8.6 kg COD/m3/d. Stepwise feeding without dilution showed lower performance and also instability. Increase of Proteiniphilum spp. and concurrent propionic acid accumulation were found in deterioration period. Dilution of food waste would result in increased biogas production and digestion reliability, which would compensate the increased amount of effluent.
Acknowledgements
This work was supported by National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2017R1A2A2A07000900).
References
1.
American Public Health Association, A.W.W.A., Water Environment Federation, 1998. Standard methods for the examination of water and wastewater, 20th ed. APHA, AWWA, and WEF, Washington, D.C., USA.
2.
Barns S. M., Fundyga R. E., Jeffries M. W., Pace N. R., 1994. Remarkable archaeal diversity detected in a Yellowstone National Park hot spring environment. P Natl. Acad. Sci. USA. 91 (5), 1609–1613.
3.
Barredo M. S., Evison L. M., 1991. Effect of Propionate Toxicity on MethanogenEnriched Sludge, Methanobrevibacter smithii, and Methanospirillum hungatii at Different pH Values. Appl. Environ. Microb. 57 (6), 1764-1769.
4.
Bluestem Solid Waste Agency, Iowa Department of Natural Resources, 2004. Anaerobic
Digestion Feasibility Study. R.W. Beck Inc., Washington, D.C., USA. 5.
Buswell, A., Mueller, H., 1952. Mechanism of methane fermentation. Ind. Eng. Chem. 44, 550–562.
6.
Chen S., Dong X., 2005. Proteiniphilum acetatigenes gen. nov., sp. nov., from a UASB reactor treating brewery wastewater. Int. J. Syst. Evol. Micr. 55 (6), 2257-2261.
7.
Chen, Y., Cheng, J.-J., Creamer, K. S., 2008. Inhibition of anaerobic digestion process: A review. Bioresource Technol. 99 (10), 4044-4064.
8.
Cho S.-K., Im W.-T., Kim D.-H., Kim M.-H., Shin H.-S., Oh S.-E., 2013. Dry anaerobic digestion of food waste under mesophilic conditions: performance and methanogenic community analysis. Bioresource Technol. 131, 210–217.
9.
Conklin A., Stensel H. D., Ferguson J., 2006. Growth kinetics and competition between Methanosarcina and Methanosaeta in mesophilic anaerobic digestion. Water Environ. Res. 78 (5), 486-496.
10. de Barros V. G., Duda R. M., de Oliveira R. A., 2016. Biomethane production from vinasse in upflow anaerobic sludge blanket reactors inoculated with granular sludge. Braz. J. Microbiol. 47 (3), 628-639. 11. DeLong E. F., 1992. Archaea in coastal marine environments. P Natl. Acad. Sci. USA. 89 (12), 5685-5689. 12. Duda R. M., de Oliveira R. A., 2009. Tratamento de águas residuárias de suinocultura em reator UASB e filtro anaeróbio em série seguidos de filtro biológico percolador. Eng. Sanitaria E Ambiental. 16 (1), 91–100. 13. Fadel E., Findikakis M., Angelos N., Leckie J. O., 1997. Environmental Impacts of Solid Waste Landfilling. J. Environ. Manage. 50 (1), 1-25. 14. Galiano, M. C., Ismail S. B., Stams A. J. M., Plugge C. M., Temmink H., Van Lier J. B., 2017. Biofilm formation and granule properties in anaerobic digestion at high salinity. Water Res. 121, 61-71. 15. Han M.-J., Behera S.-K., Park H.-S., 2012. Anaerobic co-digestion of food waste leachate and piggery wastewater for methane production: statistical optimization of key process parameters. J. Chem. Technol. Biot. 87 (11), 1541–1550.
16. Kim S.-H., Cheon H.-C., Lee C.-Y., 2012. Enhancement of hydrogen production by recycling of methanogenic effluent in two-phase fermentation of food waste. Int. J. Hydrogen Energ. 37 (18), 13777-13782. 17. Kim W., Shin S.-G., Han G., Cho K., Hwang S., 2015. Structures of microbial communities found in anaerobic batch runs that produce methane from propionic acid— Seeded from full-scale anaerobic digesters above a certain threshold. J. Biotechnol. 214. 192-198. 18. Kim, S.-H., Han, S.-K., Shin, H.-S., 2004. Feasibility of biohydrogen production by anaerobic co-digestion of food waste and sewage sludge. Int. J. Hydrogen Energ. 29 (15), 1607-1616. 19. Kumar G., Sivagurunathan P., Park J.-H., Kim S.-H., 2016. Anaerobic digestion of food waste to methane at various organic loading rates (OLRs) and hydraulic retention times (HRTs): Thermophilic vs. mesophilic regimes. Environ. Eng. Res. 21 (1), 69-73. 20. Lee D.-H., Behera S. K., Kim J.-W., Park H.-S., 2009. Methane production potential of leachate generated from Korean food waste recycling facilities: a lab-scale study. Waste Manage. 29 (2), 876–882. 21. Lee D.-J, Lee S.-Y., Bae J.-S., Kang J.-G., Kim K.-H., Rhee S.-S., Park J.-H., Cho J.-S., Chung J., Seo D.-C., 2015. Effect of Volatile Fatty Acid Concentration on Anaerobic Degradation Rate from Field Anaerobic Digestion Facilities Treating Food Waste Leachate in South Korea. J. Chem. 1-9. 22. Lee D.-J., Yoon Y.-M., Choi I.-W., Bae J.-S., Seo D.-C., 2017. Effect of seasonal variations of organic loading rate and acid phase on methane yield of food waste leachate in South Korea. Appl. Biol. Chem. 60 (1), 87-93. 23. Lee D.-Y., Ebie Y., Xu K.-Q., Li Y.-Y., Inamori Y, 2010. Continuous H2 and CH4 production from high-solid food waste in the two-stage thermophilic fermentation process with the recirculation of digester sludge. Bioresource Technol. 101 (1), S42-S47. 24. Lin C. S. K., Pfaltzgraff L. A., Herrero-Davila L., Mubofu E. B., Abderrahim S., Clark J. H., Koutinas A. A., Kopsahelis N., Stamatelatou K., Dickson F., Thankappan S., Mohamed Z., Brocklesby R., Luque R., 2013. Food waste as a valuable resource for the production of chemicals, materials and fuels. Current situation and global perspective.
Energy Environ. Sci. 6 (2), 426-464. 25. Lin J., Gan J. Z. L, Li P., Wang F. L. K, Chen L., Gan H., 2011. Effects of mixture ratio on anaerobic co-digestion with fruit and vegetable waste and food waste of China. J. Environ. sci. 23 (8), 1403-1408. 26. Nam Y.-D., Lee S.-Y., Lim S.-I., 2012. Microbial community analysis of Korean soybean pastes by next-generation sequencing, Int. J. Food Microbiol. 155 (1-2), 36–42. 27. Neves L., Oliveira R., Alves M.M., 2009. Co-digestion of cow manure, food waste and intermittent input of fat. Bioresource Technol. 100 (6),1957-1962. 28. Park, J.-H., Kim, S.-H., Park, H.-D., Lim, D.-J., Yoon, J.-J., 2013. Feasibility of anaerobic digestion from bioethanol fermentation residue. Bioresource Technol. 141, 177-183. 29. Rollon A. P., 1999. Anaerobic digestion of fish processing wastewater with special emphasis on hydrolysis of suspended solids. 1st ed. A. A. Balkema, P. O. Box 1675, 3000 BR Rotterdam, Netherlands. 30. Satoto E. N., Claudia G., Josef W., 2009. Foodwaste as a co-substrate in a fed-batch anaerobic biowaste digester for constant biogas supply. Water Sci Technol. 59 (6), 11691178. 31. Shin H.-S., Han S.-K., Song Y.-C., Lee C.-Y., 2001. Performance of UASB reactor treating leachate from acidogenic fermenter in the two-phase anaerobic digestion of food waste. Water Res. 35 (14), 3441–3447. 32. Shin, S.-G., Yoo, S., Hwang, K., Song, M., Kim, W., Han, G., Hwang, S., 2011. Dynamics of transitional acidogenic community along with methanogenic population during anaerobic digestion of swine wastewater. Process Biochem. 46 (8), 1607–1613. 33. Siddiqui Z., Horan N. J., Salter M., 2011. Energy optimization from co-digested waste using a two-phase process to generate hydrogen and methane. Int. J. Hydrogen Energ. 36 (8), 4792-4799. 34. Veluchamy C., Kalamdhad A. S., 2017. Enhanced methane production and its kinetics model of thermally pretreated lignocellulose waste material. Bioresource Technol. 241, 1-9.
35. Zaiat M., Foresti E., 1996. Cell wash-out and external mass transfer resistance in horizontal-flow anaerobic immobilized sludge reactor. Water Res. 30 (10), 2435–2439. 36. Zhang L, Lee Y.-W., Jahng D., 2011. Anaerobic co-digestion of food waste and piggery wastewater: Focusing on the role of trace elements. Bioresource Technol. 102 (8), 50485059. 37. Zhen G., Lu X., Kobayashi T., Kumar G., Xua K., 2016. Anaerobic co-digestion on improving methane production from mixed microalgae (Scenedesmus sp., Chlorella sp.) and food waste: Kinetic modeling and synergistic impact evaluation. Chem Eng J. 299, 332-341. 38. Zhua H., Parker W., Conidi D., Basnar P., Seto P., 2011. Eliminating methanogenic activity in hydrogen reactor to improve biogas production in a two-stage anaerobic digestion process co-digesting municipal food waste and sewage sludge. Bioresource Technol. 102 (14), 7086-7092.
Figure Captions Figure 1: Schematic diagram of the anaerobic digester system. Figure 2: Daily variation of methane production performance. Figure 3: Total concentration and distribution of organic acids according to OLR. All data
were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99). Figure 4: 454 pyrosequencing analysis of archaeal population in the pilot-scale anaerobic
digester. All data were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99). Figure 5: 454 pyrosequencing analysis of bacterial population in the pilot-scale anaerobic
digester. All data were taken during pseudo-state conditions except 6.1* which was obtained during the deterioration period (day 95 to 99).
Figure 1
Figure 2
Figure 3 1400
Distribution of SMPs (%)
1200 80 1000 60
800 600
40
400 20 200 0
0 4.6
6.1
6.1*
6.2
7.4 3
OLRs (kg COD/m /d)
7.9
8.6
Total organic acids (mg COD/L)
100
Acetic acid Propionic acid Butyric acid iso-butyric acid Formic acid Lactic acid Total organic acids
Figure 4
Figure 5
Table Captions Table 1: Characteristics of food waste Table 2: Digestion performances at pseudo-state conditions of each OLR
Table 1 Mode Undiluted Diluted
TCOD
SCOD
TSS
184,330 ±2,656 111,240 ±1,063
135,490 ±2,350 84,740 ±2,050
132,430 ±3,640 80,480 ±1,566
VSS (mg/L) 97,155 ±2,075 61,480 ±1,324
TS
VS
TN
170,230 ±4,870 96,990 ±2,542
158,270 ±2,371 88,480 ±2,042
5,423 ±213 4,280 ±159
Table 2 Item Average OLR (kg COD/m3/d) HRT (d) SRT (d) Average feed TCOD (g/L) Feeding mode Data collection Recirculation (Q) CH4 yield (L/kg COD) CH4 productivity (L CH4/L/d) pH Effluent Ammonia (mg N/L) Effluent TSS (mg/L) Effluent VSS (mg/L) Effluent TCOD (mg/L) Effluent SCOD (mg/L) Effluent organic acids (mg COD/L) TCOD removal efficiency (%) Alkalinity (mg CaCO3/L)
Parameter/Performance 6.2 7.4
4.6
6.1
7.9
8.6
40 40.87 184
30 30.83 184
18 18.11 111
15 15.39 111
13 13.30 111
12 111
Stepwise day 12 to 31 20 232 0.98
Stepwise day 63 to 88 20 266 1.51
Continuous day 127 to 142 20 247 1.43
Continuous day 143 to 164 15 299 2.01
Continuous day 165 to 186 10 326 2.52
Continuous day 187 to 203 10 325 2.78
8.0 975
7.5 1,310
7.7 2,120
7.7 2,470
7.5 3,060
7.5 4,110
24,400 19,770 32,250 2,500 157
23,140 12,970 35,150 2,180 390
24,070 19,230 33,810 2,050 407
23,140 12,970 27,930 1,720 449
17,550 14,330 31,970 1,500 413
24,770 18,650 29,910 1,820 402
82.5
80.9
69.5
74.8
71.2
73.1
4,720
5,120
2,740
2,620
3,061
2,360
26
Highlights
Feeding mode and dilution of food waste were examined.
Continuous feeding of diluted food waste showed stable performance.
The maximum productivity of 2.78 L CH4/L/d was achieved at 8.6 kg COD/m3/d.
454 pyrosequencing analysis divulged microbial distribution and diversity.
27