Acta Polytechnica doi:10.14311/AP.2018.58.0057 Acta Polytechnica 58(1):57–68, 2018 © Czech Technical University in Prague, 2018 available online at http://ojs.cvut.cz/ojs/index.php/ap ANAEROBIC DIGESTION OF LANDFILL LEACHATE WITH NATURAL ZEOLITE AND SUGARCANE BAGASSE FLY ASH AS THE MICROBIAL IMMOBILIZATION MEDIA IN PACKED BED REACTOR Hanifrahmawan Sudibyoa, b, Zata Lini Shabrinaa, Hartika Rafih Wondaha, Retno Tri Hastutia, Lenny Halima, Chandra Wahyu Purnomoa, Wiratni Budhijantoa, b, ∗ a Chemical Engineering Department, Faculty of Engineering, Universitas Gadjah Mada, Jalan Grafika No. 2, Yogyakarta 55281, Indonesia b Center for Energy Studies, Universitas Gadjah Mada, Jalan Bhinneka Tunggal Ika Sekip UGM K-1A, Yogyakarta 55281, Indonesia ∗ corresponding author: wiratni@ugm.ac.id Abstract. To enhance the digestion rate of a landfill leachate in the anaerobic packed bed reactor, a natural zeolite and sugarcane bagasse fly ash (BFA) were tested as the immobilization media. In order to scale up this process and systematically optimize the reactor performance, a kinetics model was needed. The suitability of the Contois and Haldane growth kinetic models were tested on the experiment data. It turned out that Contois gave the best fit for both acidogenic and methanogenic steps. A statistical analysis on Contois kinetic parameters using the Pearson correlation coefficient indicated that, in comparison with the BFA, the zeolite, as an immobilization media, showed more positive effects on the performance of the anaerobic digestion of a leachate. Keywords: landfill leachate, natural zeolite, sugarcane bagasse fly ash, growth kinetics, Contois, Haldane. 1. Introduction 1.1. Landfill leachate problem in developing country The increase of the municipal waste accumulation in conventional landfill sites has caused severe envi- ronmental impacts. In Indonesia, the high organic fraction in the municipal solid waste leads to an ex- cessive leachate release. The leachate generated from landfill sites usually contains a high amount of or- ganic and inorganic contaminants [1]. The organic and inorganic contaminants in the leachate were com- monly characterized by the high values of chemical oxygen demand (COD), pH, ammonia nitrogen and heavy metals and strong colour and bad odour. The removal of an organic material represented by the COD, biochemical oxygen demand (BOD), and am- monium from the leachate is the mandatory prereq- uisite for discharging the leachates into water bod- ies [2]. A biological treatment of the leachate was quite complicated due to the excessive amount of it, possi- ble contents toxic to the digesting microbial and the uncertainty of its composition. The leachate composi- tion can vary depending on several factors, including the degree of compaction, waste composition, moisture content in the waste, composition and volume and the age of the landfill [3, 4].Several methods are cur- rently available to treat the landfill leachate. Most of them are adapted for wastewater treatment processing and can be divided into two main categories: biologi- cal treatments and physical/chemical treatments [5]. The latter was often chosen over the biological treat- ment because it was easier and faster. However, it was usually a costlier and energy intensive operation and created other environmental problems due to the chemical release to the water bodies [5]. In a densely populated country like Indonesia, the large landfill site input could be as high as 5000– 6000 ton/day of municipal solid waste (MSW), with a 60–70 % organic fraction. This MSW characteristic produced high amount of a leachate with a high or- ganic content. An anaerobic digestion of this organic- rich landfill leachate would potentially produce a sig- nificant amount of biogas. Energy generation from the leachate in the form of a biogas makes anaerobic digestion even more attractive to Indonesia, because this country is now heavily relying on imported fossil fuel [6]. The biogas produced in the landfill site could be converted into electricity that could be used for further treatment of the effluent aerobically, using more energy-efficient aeration system [7, 8]. However, the common hindrance to run the anaer- obic digestion for the leachate treatment was the slow growth of microorganisms, so the usage of a conven- tional anaerobic digester required a huge volume of the digester [9]. Besides, a wash out often happened at the conventional anaerobic digester, especially for 57 http://dx.doi.org/10.14311/AP.2018.58.0057 http://ojs.cvut.cz/ojs/index.php/ap H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica a high flow rate feed. The solution for stabilizing and maximizing the microbial growth is immobilizing the cell on solid media [9, 10]. The cell immobiliza- tion on solid media can be defined as a localization of an intact cell via a physical adsorption between the carrier and cell membrane [11]. The microbial immobilization can be applied in various designs of reactors, such as the fixed bed, fluidized bed, or the membrane reactor to minimize the possibility of the cell to be washed out. 1.2. Material of the immobilization media The natural zeolite and sugarcane bagasse fly ash (BFA), which are both widely and abundantly avail- able in Indonesia, have a potential to be used as an immobilization media. As an inorganic mate- rial, the natural zeolite is able to immobilize the biological species through offering interesting char- acteristics, such as mechanical and chemical resis- tance and a high surface area. It has the advan- tage of the mineral content, such as silica and alu- mina (the major component) [12] calcium, magne- sium, etc., which could enhance the cell growth. Ze- olites are also known to be stable both in wet and dry conditions and well-tolerated by microorganisms, and therefore, normally compatible with a bioprocess application. The sugarcane bagasse, a residue obtained after crushing the sugarcane to extract the broth, was the most abundant lignocellulosic residue, with 1–1.2 ton of bagasse produced for 10 tons of sugarcane con- sumed for a sugar production [13]. Although most of the bagasse had been used in the sugarcane industry itself to generate energy, there was a surplus of this agro-industrial residue and several alternatives for its utilization had been evaluated, among which was the production of vanillin [14] and xylitol [15, 16]. In this respect, the sugarcane bagasse has already been used with promising results as a cell support in different bioprocesses [17]. The performance of both the BFA and the natu- ral zeolite as the immobilization media in the landfill leachate anaerobic digestion was evaluated in this work. The comparison was conducted quantitatively by means of a mathematical model to compare the ki- netics parameter between the process using a natural zeolite and the process using the BFA. The appro- priate growth kinetics model was chosen based on the best fit of the experimental data indicated by the minimum sum of squares of errors (SSE). The Pearson correlation coefficient was selected as the statistical tool to verify the correlation between the additions of the immobilization media to the digester performance represented by the kinetics parameters. Furthermore, the mathematical model suggested in this paper could be very useful in the future for scaling up and opti- mizing the design of the full scale reactors. 1.3. Kinetic model of anaerobic digestion with immobilized microbes In digesting the leachate, there are two processes that must be carried out - acidogenesis and methanogen- esis. Currently, there are two suitable models for describing the leachate or wastewater digestion - Con- tois kinetics [18] and Haldane kinetics [19]. On the one hand, the Contois kinetics has been proven to describe the acidogenesis step well, especially with the substrate type like wastewater and leachate [18]. On the other hand, for the methanogenesis step, there are still some discussions about the most suitable model due to some existing phenomenon. After the acidogenesis begins and starts producing a volatile fatty acid (VFA), this VFA usually inhibits the mi- croorganism growth, this is known as the substrate inhibition. That is why using Contois kinetics for the methanogenesis is not sometimes suitable to describe the phenomenon well. Fortunately, the Brigg-Haldane kinetics accommodates the substrate inhibition phe- nomenon. Therefore, it was necessary to compare which model suited better to describe the mechanism of the anaerobic leachate digestion, especially for the methanogenesis step. To do so, an anaerobic batch digestion inside the packed-bed reactor would be con- ducted to obtain the necessary data for the kinetics study. This kinetics study would be beneficial for op- timizing the performance of the continuous anaerobic packed-bed reactor. According to the aforementioned explanation, there were two scenarios that could be set up to find the mechanism of the anaerobic leachate digestion. The first scenario consisted of the acidogenesis step, which was described by the Contois kinetics, and followed by the methanogenesis step, which was described by the Haldane kinetics. The other scenario consisted of aci- dogenesis and methanogenesis steps, which were only described by the Contois kinetics. A set of differential equations was derived from both scenarios. The first scenario’s set of differential equations consists of (1), (2), (4), (5), and (6), whereas the second scenario’s set of differential equations consists of (1) (3) (4) (5), and (6), where dX1 dt = µm1sCODX1 KSX1X1 + sCOD − kd1Xc11 , (1) dX2 dt = µm2CVFAX2 KSX1X2 + CVFA + C2VFA/KI − kd2Xc22 , (2) dX2 dt = µm2CVFAX2 KSX1X2 + CVFA − kd2Xc22 , (3) dsCOD dt = 1 YX1/COD dX1 dt , (4) dCVFA dt = YVFA/X1 dX1 dt − 1 YX2/VFA dX2 dt , (5) dCCH4 dt = YCH4/X2 dX2 dt . (6) These differential equations are solved numerically 58 vol. 58 no. 1/2018 Anaerobic Digestion of Landfill Leachate and the corresponding kinetics constants could be determined by minimizing the Sum of Square of Error (SSE) between the calculated and experimental data of the organic matter concentration as the acidogenic cell (X1), methanogenic cell (X2), substrate (sCOD), volatile fatty acid (CVFA), and methane (CCH4). Afterwards, the change of kinetics constants related to the additions of more immobilization media would be verified through the Pearson correlation coefficient: r = n ∑ xy − ∑ x ∑ y√( n ∑ x2 − ( ∑ x)2 )( n ∑ y2 − ( ∑ y)2 ). (7) The calculated correlation coefficient was transformed into absolute value and then compared with the criti- cal value of the Pearson correlation coefficient. The absolute value of the calculated correlation coefficient must be greater than the critical value to show that there is a correlation. After being verified that there was a correlation, the integer value of the correlation coefficient (either positive or negative) shows what the correlation was. The interpretation would be linearly correlated, inversely correlated, or not correlated at all. 2. Materials and Methods 2.1. Materials A fresh leachate was obtained from Piyungan San- itary Landfill, Yogyakarta, Indonesia. Starter, in the form of an active digester effluent, was supplied by the cow-manure-based biogas mini-plant located at Gadjah Mada University’s PIAT (Pusat Inovasi Agroteknologi) at Berbah, Sleman. The immobiliza- tion media was produced from the Lampung natural zeolite and PT. Madukismo sugarcane bagasse fly ash, which were supported by bentonite as the adhe- sive agent. High purity chemicals were used in this work for analytical routines, which included H2SO4 98 % (Merck), HCl 37 % (Merck), NaOH (Merck), C8H5KO4 p.a. (EMSURE), HgSO4 p.a. (EMSURE), AgSO4 p.a. (Merck), K2Cr2O7 p.a. (EMSURE), Na2B4O7 · 10 H2O 99.55 % (Merck), and CH3COOH 96 % p.a. (Merck). 2.2. Production of Immobilization Media from Natural Zeolite The BFA and the raw natural zeolite powder (under- size 100 mesh) were each mixed with bentonite with the mass ratio of 1 : 1. Afterwards, the mixture was moulded to form the Raschig ring with the size of 1 cm inside diameter, 5 mm thickness, and 2 cm long by the extruder. Lastly, the moulded mixture was kept heated at 110 °for 12 hours by using the furnace Thermolyne Tube Heater Type F21100. 2.3. Anaerobic Digestion of Leachate The anaerobic digestion was operated in a batch sys- tem using a vertical-cylinder-formed digester made of *Corresponding author. Tel.: +6281328183160 E-mail address: wiratni@ugm.ac.id Figure 1 Figure 1. Experimental set-up. Month Parameter sCOD, mg/L VFA, mg/L 1 3660 1132 2 5895 1220 3 2620 970 Table 1. Leachate characteristics over three different months in 2016. acrylic and equipped with a vertical tube gasometer (Figure 1). In this work, the digester volume was 3 L with the L/D ratio of 4.6. The fresh leachate was fed to the digester without any dilution so that the initial concentration of the leachate (in the form of sCOD) would be different for each experiment batch. The sCOD and VFA concentration analysis, identifying the initial leachate characteristics, resulted in the data shown in Table 1. The difference was caused mostly by rainfall. To identify the effect of the immobilization media addition to the digester performance, the number of the immobilization media varied. Since both im- mobilization media had different bulk density, the number of the immobilization media was calculated based on the volume fraction inside the digester. Because of the same basic area, the immobiliza- tion media would be 1/4, 1/2, and 3/4 of the di- gester height. Through this way, the zeolite ratios were 150 g/g sCOD, 240 g/g sCOD, and 350 g/g sCOD, while the BFA ratios were 43 g/g sCOD, 110 g/g sCOD, and 164 g/g sCOD. The anaerobic digestion with- out immobilization media was set as the control (0 g zeolite/g sCOD) and was executed for each im- mobilization media. 59 H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica 2.4. Analytical method of sCOD, VFA, CH4, and microbial concentration The quantification of the microbes in the digester was conducted using the heterotrophic plate counts method explained by Boothe et al. [20] and by APHA [21]. Obtained number of the cell (in cell/L) was then converted into a mass concentration (g/mL) by multiplying with the mass of a cell (1.15 pg/cell) [22]. The mass concentration of the aci- dogenic and the methanogenic microbes was estimated by using the ratio of 9 : 1 (acidogenic : methanogenic microbes) [23]. The calculation is as follows: X1 [mg L ] = 0.9 ( # of cells [cell mL ] 1.15 pgcell 10 6 mg pg mL L ) , (8) X2 [mg L ] = 0.1 ( # of cells [cell mL ] 1.15 pgcell 10 6 mg pg mL L ) . (9) In this work, the variable used to represent the sub- strate concentration was the soluble COD (sCOD). The analysis of the sCOD, VFA, and ammonia during the experiment would follow the standard procedure by the APHA [24]. The sCOD analysis was conducted as the closed reflux colorimetric method. The VFA analysis used the titrimetric method and the ammonia measurement used the ion selective electrode (ISE) measurement. The gas volume was measured using the gasometer method outlined by Walker [25] while the methane content was analysed by using the Gas Chro- matography (GC) Shimadzu GC 8A. The data of the methane production was presented using mL/g sCOD (removed) unit [26] and in a volume percentage to describe its purity. 3. Results and Discussion 3.1. Acidogenic and methanogenic cell growth behavior According to the experimental data (Figure 2), both for the acidogenic cell (X1) and the methanogenic cell (X2), it could be seen that each growth phase ran for the same period of time when the natural zeolite and BFA were used as the immobilization media. For instance, the lag phase ran for the same period of time before and after the addition of the immobilization media (from day 0 to day 7). From day 7 to day 13, the slope was increasing sharply, so it could be con- sidered as the beginning of the log phase. Between the day 13 and 35, the slope started declining, so it could be assumed that the stationary phase began at this period. However, the maximum concentration of the cell was the differentiator. For instance, with the increase of the number of the immobilization media, the maximum concentration of the cell reached during the lag phase was greater than the one without using immobilization media. The maximum concentration increased largely when the immobilization media filled 3/4 of the digester volume. 3.2. sCOD and CVFA profile As the consequence of the cell growth, four parameters experienced changes – the sCOD, CVFA, cumulative (a) (b) (c) (d) Figure 2 0 2 000 4 000 6 000 8 000 10 000 0 7 14 21 28 35 X 1 , m g /L Days 100 1 100 2 100 3 100 4 100 5 100 6 100 7 100 8 100 0 7 14 21 28 35 X 1 , m g /L Days 0 200 400 600 800 1 000 1 200 0 7 14 21 28 35 X 2 , m g /L Days 0 200 400 600 800 1 000 0 7 14 21 28 35 X 2 , m g /L Days Figure 2. Concentration of acidogenic cell (X1) and methanogenic cell (X2) in leachate during anaerobic digestion: (a) X1-zeolite; (b) X1-BFA; (c) X2-zeolite; (d) X2-BFA (diamonds — no media; triangles — 1/4 height of digester; squares — 1/2 height of digester; crosses — 3/4 height of digester). methane purity (% CH4), and cumulative methane production. In the digester immobilized with the ze- olite, the sCOD and CVFA decreased consistently in time (Figure 3ac). This consistent decrease stood in line with the growth characteristic of the acidogenic cell (Figure 2a), which increased in time. However, the fastest significant sCOD decrease occurred in the digesters, which were immobilized by the natural zeo- lite by as many as 1/2 and 3/4 of the digester volume (240 g/g sCOD and 350 g/g sCOD). In those digesters, the sCOD decreased significantly from the day 7 to 13 60 vol. 58 no. 1/2018 Anaerobic Digestion of Landfill Leachate (a) (b) (c) (d) Figure 3 1 000 1 500 2 000 2 500 3 000 0 7 14 21 28 35 sC O D , m g /L Days 1 000 2 000 3 000 4 000 5 000 6 000 0 7 14 21 28 35 sC O D , m g /L Days 400 500 600 700 800 900 1 000 0 7 14 21 28 35 V F A , m g /L Days 700 900 1 100 1 300 1 500 1 700 1 900 0 7 14 21 28 35 V F A , m g /L Days Figure 3. sCOD and CVFA profile during anaero- bic digestion: (a) sCOD-zeolite; (b) sCOD-BFA; (c) VFA-zeolite; (d) VFA-BFA (diamonds — no media; triangles — 1/4 height of digester; squares — 1/2 height of digester; crosses — 3/4 height of digester). (shown by a steeper slope). Meanwhile, the digester without any immobilization media and the digester with the immobilization by the natural zeolite by as much as 1/4 of digester volume had a significant but late sCOD decrease from the day 21 to 35. Therefore, the addition of the zeolite seemed to be influential to the rate of the sCOD consumption by the cell after the zeolite ratio was greater than 110 g/g sCOD. However, the digester immobilized by the BFA had both the sCOD and the CVFA increasing from the day 0 to 21 (Figure 3cd). This phenomenon could possibly be explained as the degradation (through hydrolysis) of the insoluble compound, such as com- plex carbohydrates and proteins in form of particulate, into the simple ones, which were soluble. Thus, the sCOD could increase, because the sCOD increase (a) (b) (c) (d) Figure 4 0 5 10 15 20 25 30 35 0 7 14 21 28 35 C H 4 , m L/ g s C O D Days 0 5 10 15 20 25 0 7 14 21 28 35 C H 4 , m L/ g s C O D Days 0% 10% 20% 30% 40% 0 7 14 21 28 35 C H 4 , % v /v Days 0% 5% 10% 15% 20% 25% 30% 35% 40% 0 7 14 21 28 35 C H 4 , % v /v Days Figure 4. Cumulative methane production and methane percentage ( % CH4) profile during anaer- obic digestion: (a) CH4-zeolite; (b) CH4-BFA; (c) % CH4-zeolite; (d) % CH4-BFA (diamonds — no me- dia; triangles — 1/4 height of digester; squares — 1/2 height of digester; crosses — 3/4 height of digester). caused by the hydrolysis was greater than the sCOD decrease caused by the acidogenic cell consumption. After the day 21, the sCOD started decreasing, which means that the insoluble compounds had completely degraded into the simple and soluble compounds. 3.3. Cumulative volume of biogas and methane content The cumulative volume of methane produced from the digester immobilized either by the natural zeolite or by the BFA increased along the time and tended to be stable (reach the asymptote point) at the stationary phase (see Figure 4ab). However, the digesters immo- 61 H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica Variable No media 1/4 1/2 3/4 Contois-Haldane X1 1 168 700 126 620 141 820 837 480 X2 677 380 100 750 226 880 1 719 100 sCOD 5 084 200 2 987 200 752 470 951 060 CVFA 52 551 187 090 37 764 15 562 CCH4 110.24 43.62 12.8 0.01 Contois-Contois X1 8138 101 610 72 256 280 810 X2 7117 62 042 60 491 161 210 sCOD 3 240 500 2 604 100 561 560 1 321 900 CVFA 60 428 89 010 29 270 59 133 CCH4 0.47 45.95 12.15 0.01 Table 2. Comparison of SSE results of two proposed models. Variable No media 1/4 1/2 3/4 Contois-Haldane X1 42 534 252 370 143 390 1 745 400 X2 13 440 63 460 351 560 1 140 600 sCOD 411 590 17 974 208 410 232 730 CVFA 78 732 53 479 68 476 28 177 CCH4 12.07 129.41 10.57 280.21 Contois-Contois X1 12 788 28 870 11 343 481 880 X2 4344 25 903 8244 265 090 sCOD 230 980 174 150 105 620 257 410 CVFA 19 792 17 427 26 633 27 526 CCH4 9.54 79.21 2.67 113.48 Table 3. Comparison of SSE results of anaerobic leachate digestion using sugarcane bagasse fly ash as the immobilization media. bilized by the BFA by more than 1/2 of the digester volume were unfortunately unable to produce more methane. Thus, in this work, the optimum number of the BFA to have a large cumulative volume of methane was 1/2 of digester volume (101 g/g sCOD). Different from the BFA, the more natural zeolite could increase the cumulative volume of the methane produced. With respect to the methane content, the digester immobilized by the natural zeolite and the one immobi- lized by the BFA had different methane content profile. The methane content of the zeolite-immobilized di- gester had specific trend line in which a lower volume of a higher-purity methane was produced during the lag phase and the log phase. Afterward, during the stationary phase, the methane purity decreased and tended to be stable (see Figure 4c). The highest stable methane content was reached when the natural zeolite filled 3/4 of the digester volume. The lower number of the natural zeolite only produced methane purity in the range of 12–15 % at a stable conditions. When the digesters were immobilized by the BFA, the maximum methane content reached at a stable conditions was about 14 % (see Figure 4d). The in- crease of the methane content tended to be similar among the growth period (lag phase, log phase, and stationary phase). When the number of the BFA in- side the digester was increased to 3/4 of the digester volume (164 g/g sCOD), the digester became unpro- ductive in terms of the biogas production rate and methane content. To understand this phenomenon, kinetic study was conducted on the aforementioned data. 3.4. Kinetic study Generally, the SSE results of sCOD, CVFA, and CCH4 depicted the same performance of both scenarios to fit the experimental data (see Tables 2 and 3). For each number of the immobilization media both for the natural zeolite and the BFA, the SSE result stayed in the same order of magnitude. For instance, when the digester was immobilized by the natural zeolite by as many as 1/2 of the digester volume, the SSE result of the sCOD was at a hundred thousand order of magnitude. Although the SSE had a gap about two 62 vol. 58 no. 1/2018 Anaerobic Digestion of Landfill Leachate Constants No med. 1/4 1/2 3/4 Natural zeolite amount µm1 1.1989 1.2028 0.8779 2.5125 µm2 1.1782 0.9671 0.8564 1.2192 KSX1 69.5063 18.3186 11.8596 8.9268 KSX2 25.0034 8.2281 8.0195 2.824 YX1/COD 1.1205 1.1066 1.5885 2.9432 YX2/VFA 0.464 2.4561 0.5124 5.9081 YCH4/X2 15.7046 18.2752 16.8141 16.476 YVFA/X1 0.9845 0.0134 1.1001 0.0283 kd1 4 · 10−6 3.1014 2.6587 10.5556 c1 0.3601 0.3599 0.3613 0.3589 kd2 0.0281 0.103 0.0386 0.1899 c2 0.8321 0.8301 0.8311 0.8278 Table 4. Kinetics constants of Contois-Contois sce- nario (second scenario) for natural zeolite as the im- mobilization media. Constants No med. 1/4 1/2 3/4 BFA amount µm1 0.1709 0.2847 0.3766 0.6775 µm2 0.1222 0.2746 0.3655 0.6695 KSX1 14.6413 5.6303 2.7611 4.1738 KSX2 2.7716 9.0647 0.4556 2.8435 YX1/COD 0.7595 0.8443 1.1089 2.7326 YX2/VFA 0.0663 0.1116 0.1856 0.1587 YCH4/X2 16.4151 50.1092 23.6147 0.144 YVFA/X1 9.9998 6.0225 3.552 4.1674 kd1 3 · 10−6 0.2414 0.6659 0.5249 c1 0.0092 0.7588 0.7588 0.7588 kd2 0.1182 3 · 10−8 0.2997 0.0735 c2 0.5788 0.9571 0.9521 0.9571 Table 5. Kinetics constants of Contois-Contois sce- nario (second scenario) for BFA as the immobilization media. Constants In this study Previous studies µm1 0.17–0.25 0.315 [27], 0.156 [28] µm2 0.12–1.22 0.271 [29], 1.2 [28] KSX1 2.76–69.51 126.32 [27], 0.983 [29], 20–50 [28] KSX2 0.46–25.00 151.32 [29], 0.4 [30], 20–50 [28] YX1/COD 0.76–2.94 0.82 [30] YX2/VFA 0.07–5.91 0.983 [30] YCH4/X2 0.14–50.11 0.27 [30], 11–25 [31], 74 [32] YVFA/X1 0.01–10.00 0.4 [33] kd1 4 · 10−6–10.56 0.48 [30] c1 0.01–0.36 — kd2 3 · 10−8–0.30 0.48 [30] c2 0.58–0.96 — Table 6. Comparison of the obtained parameters with previous studies. hundred thousand, it was considered as a small gap due to the SSE concept. When there were six data for a one experiment condition, the SSE would be about 33000 in average. According to the SSE concept, this average meant that there was a difference of only 180 between each experimental data and each calculation result. Differently, the SSE result of X1 and X2 depicted that the anaerobic leachate digestion in the zeolite- immobilized and the BFA-immobilized digester were well described using the Contois model for both the acidogenesis and methanogenesis steps. The compar- isons between the X1’s and X2’s SSE of each scenario showed that there was a huge difference because of the different order of magnitude (see Tables 2 and 3). For instance, when the digester was immobilized by the BFA by as many as 3/4 of the digester volume, the second scenario had the SSE at a hundred thousand order of magnitude while the first scenario had the SSE at a million order of magnitude (see Tables 2 and 3). Visually observed, the huge difference of the SSE was caused by the inability of the first scenario (Contois-Haldane models) to fit the experimental data. It clearly revealed that by using the Contois model for both steps, each growth phase, such as lag phase, log phase and the beginning of the stationary phase, was well depicted (see Figures 5 and 6). The kinetics constants value for Contois-Contois scenario, resulted from the numerical calculation using MATLAB, was shown on Tables 4 and 5. Compared with previous studies focusing on finding the kinetics of an anaerobic digestion of organic or food waste, the obtained parameters only showed a slight difference. In Table 6, there were two parameters, which cannot be compared with the previous studies since in this work, the death rate equations opened the possibilities of the non-elementary kinetics model (not on the first order) together with proving if it was 63 H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica (a) Zeolite - ¼ (b) Zeolite - ½ (c) Zeolite - ¾ (d) BFA- ¼ (e) BFA- ½ (f) BFA- ¾ Figure 5 Figure 5. MATLAB calculation result for acidogenic cell (X1) inside the digester using natural zeolite and BFA as immobilization media. true that the first order was applicable in this study. To identify the effect of the increase of the immo- bilization media number, a statistical approach was used. According to the correlation coefficient result for the experimental data of the zeolite-immobilized digester, the most absolute values of the correlation coefficient were not greater than its critical value. The critical value set in this work was 0.951, gained from a level of significance of 0.05 and a degree of freedom of 2. Therefore, statistically, it revealed that the ad- dition of more natural zeolite to the digester didn’t have any correlation to the digester performance (see Tables 7 and 8). However, the correlation coefficients, which were greater than zero, still showed that, actu- ally, there was still a correlation though it was a weak correlation. Both the zeolite-immobilized and the BFA-immobi- lized digesters had the KSX1 and KSX2 value decreas- ing with the increase of the immobilization media number (see Tables 4 and 5). This decrease meant that cells/microbe could easily attach because KS is the parameter representing the affinity of microbes to the solid substrate [34]. A large KS usually indicates a low affinity and vice versa for a small KS [34]. Because of the lower value of KSX1 and KSX2, the cell/microbe preferred the BFA to the natural zeolite for the at- tachment. However, the cell/microbe attached to the solid substrate had been found to grow at a much 64 vol. 58 no. 1/2018 Anaerobic Digestion of Landfill Leachate (a) Zeolite – ¼ (b) Zeolite – ½ (c) Zeolite – ¾ (d) BFA – ¼ (e) BFA – ½ (f) BFA – ¾ Figure 6 Figure 6. MATLAB calculation result for methanogenic cell (X2) inside the digester using natural zeolite and BFA as the immobilization media. slower rate than the cell/microbe that was unattached due to its lack of a direct access to food [35]. When more cell/microbes attach to the solid substrate, the overall growth rate of the cell is slower. Thus, the growth rate of the acidogenic and methanogenic cell was better when the digester used the natural zeolite as the immobilization media, since the number of the cells/microbes attached to the zeolite was quite low. It was revealed by the value of µm1 and µm2, which was greater when using the natural zeolite inside the digester (see Tables 4 and 5). The values of YX1/COD and YX2/VFA were also greater when using the natural zeolite inside the digester due to the ability of the cell/microbe to gain the food (in form of the sCOD and VFA). Other kinetic constants dealing with the aforemen- tioned explanation were kd1 and kd2 (see Tables 4 and 5). The more immobilization media added, the greater the value of both variables. This meant that the addition of more immobilization media caused more cells/microbes to attach on it, leading to more cells/microbes dying afterwards due to a lack of food. This result stood in line with the previous work by Wang et al. [35]. As the consequence of a better growth of the cell/microbe by using the zeolite, the rate of methane 65 H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica Constants Correlation Indication coefficient (r) Nat. zeolite as the immobilization med. µm1 0.617 No correlation µm2 −0.046 No correlation KSX1 −0.899 No correlation KSX2 −0.935 No correlation YX1/COD 0.851 No correlation YX2/VFA 0.732 No correlation YCH4/X2 0.266 No correlation YVFA/X1 −0.447 No correlation kd1 0.885 No correlation c1 0 No correlation kd2 0.746 No correlation c2 0 No correlation Table 7. Pearson correlation coefficient of kinetics constants of Contois-Contois scenario (second scenario) for natural zeolite as the immobilization media. Constants Correlation Indication coefficient (r) BFA as the immobilization media µm1 0.944 No correlation µm2 0.96 Correlated KSX1 −0.868 No correlation KSX2 −0.205 No correlation YX1/COD 0.831 No correlation YX2/VFA 0.865 No correlation YCH4/X2 −0.375 No correlation YVFA/X1 −0.913 No correlation kd1 0.871 No correlation c1 0.833 No correlation kd2 0.119 No correlation c2 0.834 No correlation Table 8. Pearson correlation coefficient of kinetics constants of Contois-Contois scenario (second scenario) for BFA as the immobilization media. production was more stable in the zeolite-immobilized digester. The yield of methane produced per the methanogenic cell increase (YCH4/X2) was still greater though the methanogenic cell growth rate was also greater in the zeolite-immobilized digester. For the order of the death-rate constant, according to Shuler and Kargi [9], the order for the cell concen- tration of the death rate equation was first order. This work tried to prove the validity of this order. The numerical calculation using MATLAB showed that the death rate of the acidogenic and methanogenic cell had order’s value almost identical to first order. Thus, its validity was verified. 4. Conclusions The addition of the zeolite and the BFA as the immo- bilization media in the anaerobic digester showed dif- ferent digester’s behaviour. In the BFA-immobilized digester, the hydrolysis seemed to take part at the beginning of the operation, therefore, causing the sCOD and CVFA to increase at first. In contrast, the zeolite-immobilized digester had only acidogen- esis and methanogenesis as the limiting processes, causing the the sCOD and CVFA to decrease from the start. Besides, the methane purity and the cumulative methane volume of the zeolite-immobilized digester were greater than those of the BFA-immobilized di- gester. Kinetically, the anaerobic leachate digestion using the natural zeolite and the BFA as the immo- bilization media followed the Contois model both for the acidogenesis and for the methanogenesis step. A statistical analysis showed that a higher ratio of the immobilization media did positively affect some kinet- ics parameters. It indicated that the natural zeolite was plausible to be further studied as the potential im- mobilization media for anaerobic digestion purposes. Acknowledgements The study was conducted under the CLEAN Project fi- nancially supported by USAID PEER-Science Research Grant [NAS Sub-Grant Award Letter Agreement Number 2000004934 and Sponsor Grant Award Number AID-OAA- A-11-00012]. The authors also expressed the highest ap- preciation to the Office of Civil Work and Energy/Mineral Resources of Yogyakarta and the bureau of Waste Treat- ment, Infrastructure, and Municipal Water Supply of the Government of D.I. Yogyakarta Province. List of symbols µm1 maximum specific growth rate of acidogenic cell (day−1) µm2 maximum specific growth rate of methanogenic cell (day−1) KSX1 half-saturation constant associated with sCOD (mg sCOD/mg acidogenic cell) KSX2 half-saturation constant associated with VFA (mg VFA/mg methanogenic cell) YX1/COD yield of cell formation per mg sCOD reduction (mg acidogenic cell/mg sCOD) YX2/VFA yield of cell formation per mg VFA reduction (mg methanogenic cell/mg VFA) YCH4/X2 yield of CH4 formation per mg methanogenic cell/L increase ((mg CH4/L)/(mg methanogenic cell/L)) YVFA/X1 yield of VFA formation per mg acidogenic cell (mg VFA/mg acidogenic cell) KI inhibition constant associated with VFA (mg VFA/L) kd1 death rate constant of acidogenic cell kd2 death rate constant of methanogenic cell c1 order of acidogenic cell death rate equation c2 order of methanogenic cell death rate equation References [1] El-Salam, M.M.A., Abu-Zuid, G.I., 2015. Impact of landfill leachate on the groundwater quality: A case 66 vol. 58 no. 1/2018 Anaerobic Digestion of Landfill Leachate study in Egypt. Journal of Advanced Research, 6 (4), 579-586. doi:10.1016/j.jare.2014.02.003 [2] Kettunen, R.H., Hoilijoki, T.H., Rintala, J.A., 2009. Anaerobic and sequential anaerobic–aerobic treatments of municipal landfill leachate at low temperatures, Bioresource Technol. 58, 40–41. [3] Silva, A.C., Dezotti, M., Sant’Anna Jr, G.L., 2004. Treatment and detoxification of a sanitary landfill leachate. Chemosphere, 55 (2), pp. 207-214. doi:10.1016/j.chemosphere.2003.10.013 [4] Im, J.H., Woo, H.J. Choi, M.W., Han, K.B., Kim, C.W., 2001. Simultaneous organic and nitrogen removal from municipal landfill leachate using an anaerobic–aerobic system. Water Res. 35, 2403–2410. doi:10.1016/S0043-1354(00)00519-4 [5] Wiratni, W. and Subandiyono, 2009. Enhancement of Methane Formation in Biogas Production by Addition of Landfill Leachate. Proceeding of Regional Conference on Chemical Engineering, De La Salle University, Manila. [6] Santosa, N.B., 2014. Pemanfaatan LNG sebagai Sumber Energi di Indonesia. Jurnal Rekayasa Proses. 8(1), 33-39. [7] Deendarlianto, D., Wiratni, W., Tontowi, A., Indarto, I., & Iriawan, A. 2015. The Implementation of a Developed Microbubble Generator on the Aerobic Wastewater Treatment. International Journal Of Technology. 6(6), 924-930. doi:10.14716/ijtech.v6i6.1696 [8] Budhijanto, W., Deendarlianto, D., Kristiyani, H., & Satriawan, D., 2015. Enhancement of Aerobic Wastewater Treatment by the Application of Attached Growth Microorganisms and Microbubble Generator. International Journal of Technology. 6(7), 1101-1109. doi:10.14716/ijtech.v6i7.1240 [9] Shuler, M., Kargi, F., 2002. Bioprocess Engineering Basic Concepts, second ed. Prentice Hall, New Jersey. [10] Mshandete, A.M., Björnsson, L., Kivais, A.K., Rubindamayugi, M.S.T., Mattiasson, B., 2008. Performance of biofilm carriers in anaerobic digestion of sisal leaf waste leachate,”Electronic Journal of Biotechnology 2008; 11 (1), pp.1-9. doi:10.2225/vol11-issue1-fulltext-7 [11] Kourkoutas, Y., Xolias, V., Kallis, M. Bezirtzoglou, E., and Kanellaki, M., 2005. Lactobacillus casei cell immobilization on fruit pieces for probiotic additive, fermented milk and lactic acid production. Process Biochem. 40, 411–416. [12] Wirawan, S.K., Sudibyo, H., Setiaji, M.F., Warmada, I.W., Wahyuni, E.T., 2015. Development of natural zeolites adsorbent: chemical analysis and preliminary TPD adsorption study. Journal of Engineering Science and Technology. Special Issue 4 on SOMCHE 2014 & RSCE 2014 Conference, 87-95. [13] Rainey, T.J., 2009. A study of the permeability and compressibility properties of bagasse pulp. Brisbane, Australia: Queensland University of Technology. [14] Mathew, S., Abraham, T.E., 2005. Studies on the production of feruloyl esterase from cereal brans and sugar cane bagasse by microbial fermentation. Enzyme Microb. Tech. 36 (4), 565–570. doi:10.1016/j.enzmictec.2004.12.003 [15] Carvalho, W., Santos, J.C., Canilha, L., Silva, S.S., Perego, P., Converti, A., 2005. Xylitol production from sugarcane bagasse hydrolysate: metabolic behaviour of Candida guilliermondii cells entrapped in Caalginate. Biochem. Eng. J. 25 (1), 25–31. doi:10.1016/j.bej.2005.03.006 [16] Santos, J.C., Carvalho, W., Silva, S.S., Converti, A., 2003. Xylitol production from sugarcane bagasse hydrolyzate in fluidized bed reactor. Effect of air flowrate. Biotechnology Progress. 19 (4), 1210–1215. doi:10.1021/bp034042d [17] Sene, L., Converti, A., Felipe, M.G.A., Zilli, M., 2002. Sugarcane bagasse as alternative packing material for biofiltration of benzene polluted gaseous streams: A preliminary study. Bioresource Technol. 83 (2), 153–157. doi:10.1016/S0960-8524(01)00192-4 [18] Nelson, M., Sidhu, H. Reducing the emission of pollutants in food processing wastewaters., 2007. Chem. Eng. Proc. Process Intensification. 46 (5), 429-436. doi:10.1016/j.cep.2006.04.012 [19] Hussain, A., Dubey, S. K., Kumar, V., 2015. Kinetic study for aerobic treatment of phenolic wastewater. Water Resources and Industry. 11, 81-90. doi:10.1016/j.wri.2015.05.002 [20] Boothe, D.D.H., Smith, M.C., Gattie, D.K., Das, K.C., 2001. Characterization of microbial populations in landfill leachate and bulk samples during aerobic bioreduction. Adv. Environ. Res. 5, 285-294. doi:10.1016/S1093-0191(00)00063-0 [21] American Public Health Association (APHA), 2005. Standard Methods for the Examination of Water and Wastewater, twentieth ed. American Public Health Association, New York. [22] Fabregas, J., Herrero, C., Cabezas, B., Abalde, J., 1986. Biomass production and biochemical composition in mass cultures of the marine microalga Isochrysis galbana Parke at varying nutrient concentrations Aquaculture. 53(2), 101-113. doi:10.1016/0044-8486(86)90280-2 [23] Wirth, R., Kovács, E., Maróti, G., Bagi, Z., Rákhely, G., Kovács, K.L., 2012. Characterization of a biogas-producing microbial community by short-read next generation DNA sequencing. Biotechnol. Biofuels. 5, 41. doi:10.1186/1754-6834-5-41 [24] American Public Health Association (APHA), 1984. Compendium of Methods for the Microbiological Examination of Foods, second ed. American Public Health Association, Washington, D.C. [25] Walker, M. Zhang, Y., Heaven, S., and Banks, C., 2009. Potential Errors in the Quantitative Evaluation of Biogas Production in Anaerobic Digestion Processes. Bioresource Technol. 100, 6339-6346. doi:10.1016/j.biortech.2009.07.018 [26] Budiyono, Syaichurrozi, I., Sumardiono, S., 2014. Effect of Total Solid Content to Biogas Production Rate from Vinasse. IJE Transactions B: Applications, 27 (2), 177 – 184. 67 http://dx.doi.org/10.1016/j.jare.2014.02.003 http://dx.doi.org/10.1016/j.chemosphere.2003.10.013 http://dx.doi.org/10.1016/S0043-1354(00)00519-4 http://dx.doi.org/10.14716/ijtech.v6i6.1696 http://dx.doi.org/10.14716/ijtech.v6i7.1240 http://dx.doi.org/10.2225/vol11-issue1-fulltext-7 http://dx.doi.org/10.1016/j.enzmictec.2004.12.003 http://dx.doi.org/10.1016/j.bej.2005.03.006 http://dx.doi.org/10.1021/bp034042d http://dx.doi.org/10.1016/S0960-8524(01)00192-4 http://dx.doi.org/10.1016/j.cep.2006.04.012 http://dx.doi.org/10.1016/j.wri.2015.05.002 http://dx.doi.org/10.1016/S1093-0191(00)00063-0 http://dx.doi.org/10.1016/0044-8486(86)90280-2 http://dx.doi.org/10.1186/1754-6834-5-41 http://dx.doi.org/10.1016/j.biortech.2009.07.018 H. Sudibyo, Z. L. Shabrina, H. R. Wondah et al. Acta Polytechnica [27] Tomei, L., Altamura, S., Bartholomew, L., Bisbocci, M., Bailey, C., Bosserman, M., Cellucci, A., Forte, E., Incitti, I., Orsatti, L., Koch, U., 2004. Characterization of the inhibition of hepatitis C virus RNA replication by nonnucleosides. J. Virol. 78(2), 938-946. doi:10.1128/JVI.78.2.938-946.2004 [28] Grady Jr., C.P.L., Daigger, G.T., Love, N.G., Filipe, C.D.M., 2011. Biological Wastewater Treatment. CRC Press Taylor & Francis Group: Boca Raton, Florida, p. 296. [29] Geed, S.R., Kureel, M.K., Giri, B.S., Singh, R.S., Rai, B.N., 2017. Performance evaluation of Malathion biodegradation in batch and continuous packed bed bioreactor (PBBR). Bioresour. Technol. 227, 56-65. doi:10.1016/j.biortech.2016.12.020 [30] Fedailaine, M., Moussi, K., Khitous, M., Abada, S., Saber, M., Tirichine, N., 2015. Modeling of the anaerobic digestion of organic waste for biogas production. Procedia Comput. Sci. 52, 730 – 737. doi:10.1016/j.procs.2015.05.086 [31] Stucki M., Jungbluth N., Leuenberger M., Life cycle assessment of biogas production from different substrates, Final report, Federal Department of Environment, Bern (2011 Dec), Transport, Energy and Communications, Federal Office of Energy. [32] Achinas, S., Achinas, V., Euverink, G.J.W., 2017. Technological Overview of Biogas Production from Biowaste. Engineering. 3 (3), 299-307. doi:10.1016/J.ENG.2017.03.002 [33] Chiu, S.F., Chiu, J.Y., Kuo, W.C., 2013. Biological stoichiometric analysis of nutrition and ammonia toxicity in thermophilic anaerobic codigestion of organic substrates under different organic loading rates. Renew. Energ. 57, 323-329. doi:10.1016/j.renene.2013.01.054 [34] Liu, Y., 2006. A simple thermodynamic approach for derivation of a general Monod equation for microbial growth. Biochem. Eng. J. 31, 102–105. doi:10.1016/j.bej.2006.05.022 [35] Wang, Z.W., Hamilton-Brehm, S.D., Lochner, A., Elkins, J.G., Morrell-Falvey, J.L., 2011. Mathematical modeling of hydrolysate diffusion and utilization in cellulolytic biofilms of the extreme thermophile Caldicellulosiruptor obsidiansis. Bioresour.Technol. 102, 3155–3162. doi:10.1016/j.biortech.2010.10.104 68 http://dx.doi.org/10.1128/JVI.78.2.938-946.2004 http://dx.doi.org/10.1016/j.biortech.2016.12.020 http://dx.doi.org/10.1016/j.procs.2015.05.086 http://dx.doi.org/10.1016/J.ENG.2017.03.002 http://dx.doi.org/10.1016/j.renene.2013.01.054 http://dx.doi.org/10.1016/j.bej.2006.05.022 http://dx.doi.org/10.1016/j.biortech.2010.10.104 Acta Polytechnica 58(1):57–68, 2018 1 Introduction 1.1 Landfill leachate problem in developing country 1.2 Material of the immobilization media 1.3 Kinetic model of anaerobic digestion with immobilized microbes 2 Materials and Methods 2.1 Materials 2.2 Production of Immobilization Media from Natural Zeolite 2.3 Anaerobic Digestion of Leachate 2.4 Analytical method of sCOD, VFA, CH4, and microbial concentration 3 Results and Discussion 3.1 Acidogenic and methanogenic cell growth behavior 3.2 sCOD and CVFA profile 3.3 Cumulative volume of biogas and methane content 3.4 Kinetic study 4 Conclusions Acknowledgements List of symbols References