Temperature dependency of granule characteristics and kinetic behavior in UASB reactors

Journal of Chemical Technology and Biotechnology J Chem Technol Biotechnol 79:797–808 (online: 2004) DOI: 10.1002/jctb.999 Temperature dependency of granule characteristics and kinetic behavior in UASB reactors Hsin-Hsien Chou, Ju-Sheng Huang∗ and Wen-Feng Hong Department of Environmental Engineering, National Cheng Kung University, Tainan, Taiwan 701, People’s Republic of China Abstract: When an inhibitory substrate, phenol, was treated under mesophilic conditions (25, 30, 35, and 40 ◦C), the upflow anaerobic sludge bed (UASB) reactors at 30 ◦C resulted in the greatest amount of biomass and the largest granule size, while the UASB reactors at 25 ◦C resulted in the smallest granule size and the greatest amount of wash-out of sludge. The granule size tended to be negatively correlated with the amount of wash-out of sludge.With an increase in temperature, the kinetic constant k for anaerobic phenol degradation increased and the half saturation constant (Ks) decreased. The mass fraction of methanogens (f ) increased with increasing operational temperature in the UASB reactors and the activation energy (Ea) for acetate methanogenesis was larger than that for phenol acidogenesis in the batch reactors, indicating that the operational temperature imposes a more influential effect on methanogens than on acidogens. From the results of the activity of acidogens and methanogens (expressed in specific COD utilization rate), the rate-limiting step is phenol acidogenesis.  2004 Society of Chemical Industry Keywords: upflow anaerobic sludge bed; mesophilic; granule characteristics; phenol acidogenesis; acetate methanogenesis; rate-limiting; kinetics NOTATION Af Arrhenius frequency factor (day−1) COD Chemical oxygen demand (mg dm−3) dp Average size of granules based on surface area (mm) Ea Activation energy (cal mol−1) f Mass fraction of methanogens (dimen- sionless) H Sludge-bed height (cm) k Reaction-rate constant or maximum spe- cific phenol utilization rate constant of mixed culture (day−1) k1 Maximum specific phenol utilization rate constant for phenol acidogenesis (day−1) k2 Maximum specific acetate utilization rate constant for acetate methanogenesis (day−1) k′2 Maximum specific acetate utilization rate constant of mixed culture (day−1) Ki Inhibition constant (mg phenol dm−3) Ks Half-saturation constant for phenol degra- dation (mg phenol dm−3) Ks1 Half-saturation constant for phenol aci- dogenesis (mg phenol dm−3) Ks2 Half-saturation constant for acetate methanogenesis (mg acetate dm−3) Qi Inflow rate (dm3 day−1) r Specific phenol utilization rate (day−1) R Recycle ratio (dimensionless); universal constant (cal mol−1 K−1) SB,−1,SB,0, SB,+1 Acetate bulk concentrations in three consecutive samples (mg acetate dm−3) Se Residual concentration (mg dm−3) Si Phenol influent concentration (mg phenol dm−3) �t Time interval (h) T Absolute temperature (K) us Superficial flow velocity (m h−1) UASB Upflow anaerobic sludge bed VSS Volatile suspended solids (mg dm−3) VR Reactor volume (dm3) VFAs Volatile fatty acids (mg acetate dm−3) XB Average VSS concentration in batch reactor (mg VSS dm−3) Xf Microbial density (mg VSS dm−3) Xi Biomass concentration (mg VSS dm−3) θ Temperature coefficient (dimensionless) SUBSCRIPTS 1 Phenol 2 Acetate ∗ Correspondence to: Ju-Sheng Huang, Department of Environmental Engineering, National Cheng Kung University, Tainan, Taiwan 701, People’s Republic of China E-mail: [email protected] Contract/grant sponsor: National Science Council of the Republic of China; contract/grant number: NSC 89-2211-E-006-108 (Received 20 March 2003; revised version received 5 November 2003; accepted 6 December 2003) Published online 19 May 2004  2004 Society of Chemical Industry. J Chem Technol Biotechnol 0268–2575/2004/$30.00 797 H-H Chou, J-S Huang, W-F Hong 1 INTRODUCTION In the last two decades, upflow anaerobic sludge bed (UASB) reactors have been widely used for treating various kinds of industrial wastewaters. In UASB reactors, anaerobic bacteria are immobilized by a process of spontaneous aggregation of the bacteria, resulting in dense granules with high microbial activity and good settling characteristics.1 The great success of UASB reactors lies in their capability to retain a high concentration of granular sludge, which allows the operation of high organic loading rates and maintenance of a long retention time for biological solids. Successful operation of UASB reactors treating phenol wastewater at mesophilic temperatures (35–37 ◦C) and volumetric loading rates of 0.4–10.66 g COD dm−3 day−1 also have been claimed.2–7 The inhibitory effect of phenol on anaerobic bacteria in the UASB reactors can be reduced with recycling of the reactor-effluent.3–5,7 Nonetheless, phenol imposes a significant inhibitory effect on anaerobic bacteria in the UASB reactors if the phenol concentration in bulk liquid reaches approximately 500 mg dm−3.8 Depending on the temperature range, microor- ganisms are generally grouped into three categories: thermophilic, mesophilic, and psycrophilic. Of major concern in biological wastewater treatment processes are mesophilic microorganisms, which grow well over the temperature range of 20–40 ◦C. Even under mesophilic conditions, the maximum specific sub- strate utilization rate of methanogens increased with an increase in temperature;9,10 and the anaerobic reac- tors sustained a higher organic loading rate at a higher operational temperature.11,12 To adequately interpret the experimental results regarding the effect of tem- perature on reaction rate of biochemical reactions, the Arrhenius relationship13–15 is commonly used to estimate either activation energy (Ea) or temper- ature coefficient (θ). A positive Ea value or a θ value of greater than unity depicts that the reaction rate increases with increasing temperature and, the greater the Ea (or θ) is, the more influential the effect of reaction rate imposed by the temperature will be. Most researchers mainly focused on the treatment efficiency of the UASB reactors under mesophilic conditions.2–6,11,16–18 It has been noted that the average granule size measured from UASB reactors decreased from 1–2 mm to 0.5–1 mm when the operational temperature increased from mesophilic (37 ◦C) to thermophilic (55 ◦C).19 Under mesophilic conditions, the activity of methanogens (utilizing volatile fatty acids (VFAs) or H2/CO2) increases with an increase in temperature.9,10 However, few studies have investigated the temperature dependency of granule characteristics and the activity of acidogens and methanogens in UASB reactors treating complex organics (eg inhibitory substrate phenol) under mesophilic conditions. In this work, four laboratory-scale UASB reactors (with four different superficial flow velocities (us)) treating phenol wastewater were used. Four test runs were conducted at four different operational tempera- tures (25, 30, 35, and 40 ◦C) for each UASB reactor to generate experimental data. Granule size distribution in the UASB reactors was also determined. By carry- ing out independent batch experiments, the intrinsic Haldane biokinetic constants for phenol acidogenesis (k1,Ks1, and Ki) and the Monod-type biokinetic con- stants for acetate methanogenesis (k2 and Ks2) as well as the mass fractions of acidogens and methanogens (f ) at the four different operational temperatures were estimated. To comprehend the temperature depen- dency of reaction rate (in the batch reactors) and the specific phenol utilization rate (in the UASB reactors), the Arrhenius relationship13 was used to estimate the activation energy (Ea) and temperature coefficient (θ). Moreover in this paper, the variations in the activity of acidogens and methanogens (expressed in specific substrate utilization rates) with different operational temperatures in the UASB reactors are presented as well. 2 MATERIALS AND METHODS 2.1 Experimental apparatus Four Plexiglas UASB reactors were used. A schematic diagram of UASB reactor with six-equal distance sampling ports is shown in Fig 1. Each UASB reac- tor had dimensions of 6.0 (length) × 6.0 (width) × 105 cm (height) and a volume of 3.78 dm3 (ie reac- tor volume). The dimensions of the gas–liquid–solids separator of the UASB reactors were 6.0 (length) × 6.0 cm (width), and the liquid volume of the UASB reactor including gas–liquids–solids separator equalled 4.14 dm3. 2.2 Bioreactor operation The four UASB reactors used in the present study have been operated in our laboratory for the treatment of phenol wastewater for more than 1 year. The original anaerobic seed sludge was taken from a local piggery wastewater treatment plant. The chemical composition of synthetic wastewater (diluted with tap water) and the operating conditions of the four UASB reactors are shown in Tables 1 and 2, respectively. The superficial flow velocities (us) of the four bioreactors were maintained at 0.5, 1.0, 2.0, and 4.0 m h−1 by controlling the inflow rate of 10 dm3 day−1 and four different recycling flow rates of 33, 76, 163, and 336 dm3 day−1, respectively. By circulating hot water in the outer-water jacket, the operational temperatures of the four bioreactors in test runs 1, 2, 3, and 4 were respectively maintained at 25, 30, 35, and 40 ◦C. Steady state of the UASB reactor system is defined as follows. Each of the four bioreactors (in test runs 1–4) for the anaerobic degradation of phenol must be conducted and continuously operated for at least 3 months; the biogas (CH4 plus CO2) production rate and phenol and volatile fatty acids (VFAs) concentrations in the effluent are within 5% deviation 798 J Chem Technol Biotechnol 79:797–808 (online: 2004) Effects of temperature in UASB reactors Figure 1. Schematic diagram of the UASB reactor. for three consecutive samples (sampling twice weekly). Thus, the steady-state test results obtained in test runs 1–4 were used for thorough discussion. 2.3 Determination of biokinetic constants Accurate determination of granule kinetic constants for substrate utilization in the UASB reactor is difficult because the internal and external mass transfer resistances of the granule may mask true reaction kinetics. In addition, due to the fact that phenol imposes an inhibitory effect on anaerobic bacteria at high concentrations, the Haldane inhibition model was commonly used to describe anaerobic phenol degradation kinetics.7,20–22 Thus, the intrinsic Haldane biokinetic constants for anaerobic phenol degradation (k, Ks, Ki; mixed culture) and the intrinsic Monod-type biokinetic constants for acetate methanogenesis (k2, Ks2; enrichment culture) at 25, 30, 35, and 40 ◦C were determined using batch reactors with dispersed sludge. Dispersed sludge for anaerobic phenol degradation (ie to determine k, Ks, and Ki) was prepared by putting the sludge granules (removed from the steady-state UASB reactor) together with a few glass beads into a 1 dm3 serum vial, followed by placing the serum vial on a shaker (100 rpm) for over 1 h to break up sludge granules (ie to obtain dispersed sludge). The size distribution of the obtained dispersed sludge was determined using a laser scattering particle size distribution analyzer (Horiba, model LA-920, Japan) ranged from 0.02 to 2000 µm. More than 90% of the particles were found within the colloidal size range (0.259–101.46 µm). Thus, the resistances of both external mass transfer and diffusion within tiny microbial cells in batch culture were considered to be negligible. Dispersed sludge for acetate methanogenesis (ie to determine k2 and Ks2) was also prepared by load- ing the anaerobic sludge (collected from the effluent of the UASB reactors) into a 5 dm3 batch reactor (approximately 1500 mg VSS dm−3) installed with a mixing device (approximately 100 rpm) to break up sludge granules. Similarly, more than 84% of the particles were found within the colloidal size range (0.226–101.46 µm). Prior to determining the biokinetic constants k2 and Ks2, the enrichment of methanogenic culture was carried out for 4 months with the following procedures. The synthetic wastewa- ter that only contained acetate and nutrients was added into the well-agitated batch reactor (100 rpm), which was already loaded with dispersed sludge. An initial pH value of 6.8 was measured. At the end of methano- genesis (pH ∼= 7.2), the agitation was stopped, and the anaerobic sludge allowed to settle, followed by decant- ing of the supernatant. The synthetic wastewater that contained acetate and nutrients was again added to the batch reactor. To determine biokinetic constants k, Ks, and Ki (mixed culture) and to avoid a possible sudden initial decrease in phenol concentration resulting from adsorption of phenol on the biomass, samples from batch reactors were analyzed, 1 h after liquid mixing and phenol addition to the culture, for the biomass and phenol remaining in the solution. To determine biokinetic constants k2 and Ks2 (enrichment culture), samples from batch reactors were analyzed for the biomass and acetate remaining in the solution. Thereafter, the Levenberg–Marquardt algorithm method23 was applied to search for a set of biokinetic constants which would fit in with the experimental data. Table 1. Chemical composition of synthetic wastewater (diluted with tap water)a,b Test run Phenol (mg COD dm−3) Yeast extract (mg dm−3) K2HPO4 (mg dm−3) KH2PO4 (mg dm−3) NH4Cl (mg dm−3) Alkali NaHCO3 (mg CaCO3 dm−3) 1–4 3980 60 115 83 360 2000 a Trace metals added: Ni2+, 0.5; Fe3+, 0.5; Co2+, 0.3; Mo6+, 0.6; Zn2+, 0.5; Mn2+, 0.5 mg dm−3. b Influent pH: 7.8. J Chem Technol Biotechnol 79:797–808 (online: 2004) 799 H-H Chou, J-S Huang, W-F Hong Table 2. Performance of upflow anaerobic sludge bed reactorsa Xid (mg VSS dm−3) Effluent Test run Temp (◦C) Vol loadingb (g COD dm−3 day−1) us (m h−1) Hc (cm) Lower Middle Upper Weighted mean Biomass (g) See (mg COD dm−3) VFAsf (mg dm−3) VSS (mg dm−3) COD removal (%) A1 25 10.5 0.5 74 42 900 40 800 37 140 40 380 107.6 255 (72) 16.3 56 93.6 B1 25 10.5 1.0 78 42 210 41 660 36 460 40 120 112.7 210 (61) 9.8 52 94.7 C1 25 10.5 2.0 78 44 680 43 120 41 210 42 950 120.6 208 (58) 8.3 68 94.8 D1 25 10.5 4.0 71 39 650 38 760 37 320 38 660 98.8 294 (83) 21.8 74 92.6 Arithmetic mean 75.3 40 530 109.9 242 (69) 14.1 63 93.9 A2 30 10.5 0.5 71 46 780 46 040 41 920 45 200 115.5 105 (30) 5.6 36 97.4 B2 30 10.5 1.0 70 49 280 47 070 45 240 47 320 119.3 65 (4.9) 4.0 38 98.4 C2 30 10.5 2.0 70 50 740 52 340 51 110 51 520 129.8 62 (2.0) 2.6 38 98.4 D2 30 10.5 4.0 68 43 130 43 020 40 010 42 370 103.7 98 (19) 6.0 42 97.5 Arithmetic mean 69.8 46 600 117.1 83 (22) 4.6 39 97.9 A3 35 10.5 0.5 69 43 110 41 650 40 470 41 840 103.9 90 (28) 5.6 33 97.7 B3 35 10.5 1.0 72 42 010 41 120 37 050 40 300 104.4 58 (3.2) 4.2 42 98.5 C3 35 10.5 2.0 70 47 170 47 960 42 980 46 460 117.1 42 (0.8) 2.0 45 98.9 D3 35 10.5 4.0 72 40 670 39 220 36 190 38 850 100.7 92 (14) 3.1 47 97.7 Arithmetic mean 70.8 41 860 106.5 71 (12) 3.7 42 98.2 A4 40 10.5 0.5 69 42 270 42 460 39 110 41 600 103.3 98 (29) 3.8 40 97.5 B4 40 10.5 1.0 71 42 600 40 570 37 550 40 430 103.2 68 (6.3) 3.8 42 98.3 C4 40 10.5 2.0 70 46 940 46 120 44 670 46 020 116.0 56 (3.2) 2.1 44 98.6 D4 40 10.5 4.0 69 40 150 40 010 35 360 38 940 96.7 106 (20) 3.1 52 97.3 Arithmetic mean 69.8 41 750 104.8 82 (15) 3.2 45 97.9 a Qi = 10 dm day−1, Si = 1670 mg phenol dm−3 (3980 mg COD dm−3), VR = 3.78 dm3, R = 3.3, 7.6, 16.3, and 33.6. b Vol loading = QiSi/VR. c H = sludge-bed height. d Xi = biomass concentration. e The number indicated in parenthesis is based on phenol. f Expressed in mg acetate dm−3. 2.4 Determination of mass fraction of methanogens In test runs 1–4, an adequate amount of sludge granules was respectively removed from the lower-, middle-, and upper-part of sludge bed of each steady- state UASB reactor operated at a designated opera- tional temperature. Together with a few glass beads, the sludge granules were loaded into a 100 cm3 serum vial, which was immediately sealed. The serum vial was placed on a shaker (at a designated tempera- ture; 100 rpm) for 10–15 min to break up sludge granules. Then the synthetic wastewater that con- tained acetate and nutrients was added into the serum vial, which was immediately sealed. At a designated temperature and in an agitation state, the initial biomass concentration (XB) in the serum vial was measured, while the acetate concentra- tion remaining in the solution (SB,−1, SB,0, SB,+1) was measured hourly. Thereafter, the initial rate method7 was used to determine the mass fraction of methanogens (f ) as shown in eqns (1) and (2), as follows: k′2 = Ks2 + SB,0 XBSB,0 SB,−1 − SB,+1 2�t (1) f = k ′ 2 k2 (2) where SB,−1, SB,0, and SB,+1 represent the acetate bulk concentrations in the initial three consecutive samples: XB represents the biomass concentration in the serum vial; and �t represents the time interval. 2.5 Determination of granule size When the UASB reactor reached steady state, the granule diameter was determined using the method of image analysis. Approximately 5–10 cm3 of granular sludge were randomly taken from the lower-, middle-, and upper-part of the sludge bed in each bioreactor. Then sludge granules were poured into a flask and washed several times with deionized water. After that, an enlarged mouth dropper was used to take 1 cm3 of the washed granules from the flask, followed by placing them on a filter paper (11 cm in diameter) and overlapping granules were separated with a pair of tweezers. The filter paper with the granules on the top was photographed. The prints were analyzed for granule area with an image analyzer (Bischke, model CD-5024N, Canada). Granule size distribution (ranging from 0.1 to 4.0 mm) was determined using a commercial statistical software package (Matrox, Inspector version 2.2 for windows NT). 800 J Chem Technol Biotechnol 79:797–808 (online: 2004) Effects of temperature in UASB reactors 2.6 Determination of activation energy and temperature coefficient To adequately describe the effect of temperature (T) on reaction rate (k) for chemical or biochemical reactions, Arrhenius (in 1889)13 postulated the following relationship: k = Af e(−Ea/RT) (3) The logarithmic form of eqn (3) gives ln k = −Ea R 1 T + lnAf (4) By plotting ln k versus 1/T together with linear regression, the activation energy (Ea) and frequency factor (Af ) can be estimated. The expression that is frequently used in biological wastewater treatment to correct reaction-rate constant (k) for temperature variation is the modified Arrhenius relationship (ie to introduce a temperature coefficient term θ): kT = k25θ(T−298) (5) where θ = e Ea298RT (6) By plotting ln (kT/k25) versus (T − 298) together with linear regression, the temperature coefficient (θ) can be estimated. 2.7 Determination of phenol and VFAs The 4-aminoantipyrine colorimetric method24 was used to measure phenol. A gas chromatographic column (GC) system (Shimadzu, model GC-14A, Japan) was used to measure VFAs with a 15 m × 0.53 mm Supleco-Nukol fused-silica capillary column, FID detector, and nitrogen carrier gas. Injector and detector temperatures were 230 ◦C. The fluid sample was filtered through a 0.45 µm membrane filter and acidified to pH 3 with concentrated phosphoric acid prior to injecting into the column using the fast injection technique. 2.8 Statistical analysis To ascertain whether the granule size in the UASB reactors vary with different operational temperatures, the t-test for paired observations was performed by using eqn (7): t = average difference of paired observations standard error of average difference of paired observations (7) To test the null hypothesis (H0), H0: ParameterT1 = ParameterT2 (two-tailed test) H0: ParameterT1 ≥ ParameterT2 (one-tailed test) or H0: ParameterT1 ≤ ParameterT2 (one-tailed test) If the null hypothesis is rejected, the alternative hypothesis (H1) is true. 3 RESULTS AND DISCUSSION 3.1 Performance of UASB reactors The operating conditions and experimental results of the UASB reactors treating phenol wastewater are presented in Table 2. At the volumetric load- ing of 10.5 g COD dm−3 day−1 (us = 0.5, 1.0, 2.0, and 4.0 m h−1), the COD removal efficiency at 25 ◦C, (arithmetic mean = 93.9%) was relatively lower than those at 30, 35, and 40 ◦C (arithmetic means = 97.9–98.2%). In addition, excluding the opera- tional temperature of 25 ◦C, the biomass retained in the UASB reactors tended to decrease with increasing temperature (30–40 ◦C). According to the literature,9,14,25 the specific growth rate, the yield coef- ficient and the decay rate all increase with increasing temperature under mesophilic conditions. Thus, the results obtained in the present study implied that the increasing temperature imposed a more influential effect on the decay rate than on the other two factors. In addition, the reason why the biomass retained in the UASB reactor at 25 ◦C did not reach the expected greatest amount may be because the sludge-bed height at 25 ◦C (arithmetic mean = 75.3 cm) was higher than those at 30, 35, and 40 ◦C (arithmetic means = 69.8–70.8 cm); the biomass concentration (Xi) and microbial density (Xf ) at 25 ◦C (arithmetic mean = 40 530 mg VSS dm−3; 147 830 mg VSS dm−3) were lower than those at 30, 35, and 40 ◦C (arithme- tic means = 41 750–46 600 mg VSS dm−3; 151 920– 160 270 mg VSS dm−3). Consequently, the wash- out of sludge from the UASB reactor at 25 ◦C (arithmetic mean = 63 mg VSS dm−3) was greater than those at 30, 35, and 40 ◦C (arithmetic means = 39–45 mg VSS dm−3) (Table 2). These findings may explain why the COD removal efficiency at 25 ◦C was relatively lower than those at 30, 35, and 40 ◦C. Moreover, under mesophilic conditions (25–40 ◦C), the UASB reactor retained less biomass when operated at low and high superficial flow velocity (us; reactors A and D). This may be because a low recycle ratio (ie low us) would cause insufficient dilution of inflow wastewater, imposing an inhibitory effect of phenol on anaerobic bacteria in the UASB reactor. In contrast, a high recycle ratio (ie high us) would cause wash-out of small granules from the UASB reactor. Accordingly under mesophilic conditions, the us of 1.0–2.0 m h−1 is suggested when the UASB reactor is applied to treat high-strength inhibitory substrates. 3.2 Granule characteristics Table 3 shows that the specific gravities (sp gr) of the granule in the UASB reactors at 25, 30, 35, and 40 ◦C were 1.05, 1.05–1.06, 1.05, and 1.05, respectively. The sp gr of granules mainly depends on the occurrence of inorganic precipitation26 (ie wastewater composition). In the present study, the J Chem Technol Biotechnol 79:797–808 (online: 2004) 801 H-H Chou, J-S Huang, W-F Hong T ab le 3. G ra nu le ch ar ac te ris tic s of up flo w an ae ro b ic sl ud ge b ed re ac to rs a G ra nu le ’s sp gr d p (m m ) X i c (m g V S S dm −3 ) Te st ru n Te m p (◦ C ) V ol lo ad in gb (g C O D dm −3 da y− 1 ) u s (m h− 1 ) Lo w er M id dl e U pp er W ei gh te d m ea n Lo w er M id dl e U pp er W ei gh te d m ea n Lo w er M id dl e U pp er W ei gh te d m ea n A 1 25 10 .5 0. 5 1. 06 1. 05 1. 03 1. 05 1. 41 1. 03 0. 86 1. 10 15 4 10 0 15 1 70 0 14 6 10 0 15 0 87 0 B 1 25 10 .5 1. 0 1. 06 1. 05 1. 03 1. 05 1. 65 1. 60 1. 31 1. 53 15 6 20 0 15 2 10 0 14 4 00 0 15 0 90 0 C 1 25 10 .5 2. 0 1. 06 1. 05 1. 03 1. 05 1. 67 1. 67 1. 35 1. 58 14 6 20 0 14 1 30 0 14 0 60 0 14 2 61 0 D 1 25 10 .5 4. 0 1. 06 1. 05 1. 03 1. 05 1. 76 1. 75 1. 40 1. 67 15 1 40 0 14 5 20 0 13 7 00 0 14 5 14 0 A rit hm et ic m ea n 1. 47 14 7 38 0 A 2 30 10 .5 0. 5 1. 06 1. 05 1. 03 1. 05 1. 60 1. 47 0. 91 1. 37 15 8 70 0 15 5 30 0 15 1 30 0 15 5 34 0 B 2 30 10 .5 1. 0 1. 06 1. 05 1. 03 1. 05 1. 81 1. 73 1. 64 1. 73 15 4 10 0 14 9 20 0 14 6 90 0 15 0 20 0 C 2 30 10 .5 2. 0 1. 07 1. 06 1. 04 1. 06 2. 05 2. 08 1. 78 2. 00 15 6 90 0 15 7 00 0 14 1 40 0 15 3 07 0 D 2 30 10 .5 4. 0 1. 06 1. 05 1. 04 1. 05 2. 00 2. 03 1. 80 1. 97 15 6 20 0 14 8 30 0 14 0 80 0 14 9 08 0 A rit hm et ic m ea n 1. 77 15 1 92 0 A 3 35 10 .5 0. 5 1. 06 1. 05 1. 04 1. 05 1. 52 1. 35 0. 91 1. 30 16 6 20 0 17 1 20 0 17 1 90 0 16 9 74 0 B 3 35 10 .5 1. 0 1. 06 1. 06 1. 04 1. 05 1. 72 1. 68 1. 60 1. 67 16 1 30 0 16 0 00 0 14 7 00 0 15 6 89 0 C 3 35 10 .5 2. 0 1. 06 1. 05 1. 05 1. 05 2. 01 1. 71 1. 60 1. 78 15 9 20 0 16 0 00 0 14 5 70 0 15 6 17 0 D 3 35 10 .5 4. 0 1. 06 1. 06 1. 03 1. 05 1. 98 1. 78 1. 73 1. 83 16 4 30 0 15 9 20 0 15 0 00 0 15 8 30 0 A rit hm et ic m ea n 1. 65 16 0 27 0 A 4 40 10 .5 0. 5 1. 06 1. 06 1. 03 1. 05 1. 53 1. 25 0. 90 1. 26 16 4 00 0 15 5 30 0 13 9 60 0 15 4 38 0 B 4 40 10 .5 1. 0 1. 06 1. 05 1. 04 1. 05 1. 74 1. 72 1. 56 1. 68 16 2 80 0 15 6 70 0 14 9 30 0 15 6 70 0 C 4 40 10 .5 2. 0 1. 06 1. 05 1. 05 1. 05 2. 03 1. 70 1. 62 1. 79 15 9 70 0 15 1 90 0 14 1 80 0 15 1 88 0 D 4 40 10 .5 4. 0 1. 06 1. 05 1. 03 1. 05 1. 84 1. 82 1. 68 1. 79 15 8 70 0 15 6 40 0 15 8 80 0 15 7 72 0 A rit hm et ic m ea n 1. 63 15 5 17 0 a Q i = 10 d m 3 d ay −1 ,S i = 16 70 m g p he no ld m −3 (3 98 0 m g C O D d m −3 ), V R = 3. 78 d m 3 ,R = 3. 3, 7. 6, 16 .3 ,a nd 33 .6 . b V ol lo ad in g = Q iS i/ V R . c X f = m ic ro b ia ld en si ty . 802 J Chem Technol Biotechnol 79:797–808 (online: 2004) Effects of temperature in UASB reactors same wastewater composition was prepared for the four UASB reactors at the operational temperatures of 25–40 ◦C. Thus, the sp gr of granules at this operational temperature range varied slightly. Typical sp gr of granules in UASB reactors ranged from 1.03 to 1.08.3,26 Table 3 also shows that the average gran- ule sizes (dp) in the UASB reactors at 25, 30, 35, and 40 ◦C were 1.10–1.67 mm (arith- metic mean = 1.47 mm), 1.37–2.00 mm (arithmetic mean = 1.77 mm), 1.30–1.83 mm (arithmetic mean = 1.65 mm), and 1.26–1.79 mm (arithmetic mean = 1.63 mm), respectively. According to the t-test for paired observations, the dp at 30 ◦C was the largest, the dp values at 35 and 40 ◦C were the next largest, and the dp at 25 ◦C was the smallest at the 0.05 level of significance (Table 4). By observing the performance data of the wash-out of sludge (ie VSS concentration in the effluent; Table 2) and granule size (dp; Table 3), the granule size tended to be negatively correlated with the amount of wash-out of sludge. In the present study, the greatest amount of wash-out of sludge at 25 ◦C gave the smallest granule size, while the least amount of wash-out of sludge at 30 ◦C gave the largest granule size. 3.3 Evaluation of kinetic parameters 3.3.1 Parameters k2 and Ks2 for acetate methanogenesis The data points shown in Fig 2 imply that the Monod- type kinetics can be used to adequately describe the anaerobic degradation of acetate in batch reactors (enrichment culture) at the temperatures of 25, 30, 35, and 40 ◦C.7,9,27 Thus, the Levenberg–Marquardt algorithm method23 was applied to estimate the Monod-type kinetic constants (k2 and Ks2). When the temperature was increased from 25 to 30 and 35 ◦C, the k2 value gradually increased from 3.1 to 4.2 and 5.1 mg acetate mg VSS−1 day−1. With a further increase in temperature (40 ◦C), the k2 value only slightly increased (5.2 mg acetate mg VSS−1 day−1). This can be explained by the fact that, under mesophilic conditions, denaturation of the enzyme protein in acetate methanogenesis did not occur. Thus, k2 was clearly a rate coefficient, and as such, its value increases with increasing temperature. A similar result was also reported by Lawrence and McCarty9 at the Table 4. Statistical analyses of granule size H0 t valuea t(α,n−1) Results dp,25◦C ≥ dp,30◦C −6.48 t(0.05,3) = 2.35 Reject H0 dp,25◦C ≥ dp,35◦C −11.67 t(0.05,3) = 2.35 Reject H0 dp,25◦C ≥ dp,40◦C −8.55 t(0.05,3) = 2.35 Reject H0 dp,30◦C ≤ dp,35◦C 3.31 t(0.05,3) = 2.35 Reject H0 dp,30◦C ≤ dp,40◦C 3.83 t(0.05,3) = 2.35 Reject H0 dp,35◦C = dp,40◦C 1.04 t(0.025,3) = 3.18 Accept H0 a t = (D − 0)/Sd/(n)0.5, where D and Sd represent the mean and standard deviation of the differences of paired observations, respectively, n represents the number of paired observations; degrees of freedom = n − 1. Figure 2. Time course of acetate utilization in batch reactors (enrichment culture). temperature range of 25 to 35 ◦C. In contrast, when the temperature was increased from 25 to 30, 35, and 40 ◦C, the Ks2 value gradually decreased from 240 to 202, 133, and 122 mg acetate dm−3. This implied that acetate bulk concentration influenced the specific acetate utilization rate. Similar results were also reported by Lawrence and McCarty9 at the temperature range of 25 to 35 ◦C and Lin et al28 at the temperature range of 15 to 35 ◦C. Nonetheless, a contradictory result was reported by Westermann et al10 at the temperature range of 20 to 37 ◦C. Grady et al29 pointed out that other factors (including the kind of substrate and its concentration, sludge age, cultivation modes of pure or mixed culture, and batch or continuous-flow stirred tank reactor) would possibly affect the estimated Ks2 value. J Chem Technol Biotechnol 79:797–808 (online: 2004) 803 H-H Chou, J-S Huang, W-F Hong Figure 3. Time course of phenol degradation in batch reactors (mixed culture). 3.3.2 Parameters k, Ks, Ki for anaerobic degradation The data points shown in Fig 3 reveal that the Haldane kinetics can be used to appropriately describe the anaerobic degradation of phenol in batch reactors (mixed culture) at the temperatures of 25, 30, 35, and 40 ◦C.7,20 Thus, the Levenberg–Marquardt algorithm method23 was applied to estimate the Haldane kinetic constants (k, Ks, and Ki). When the temperature was increased from 25 to 30, 35, and 40 ◦C, the k value gradually increased from 0.6 to 0.65, 0.73, and 0.76 mg phenol mg VSS−1 day−1 (ie denaturation of the enzyme protein in anaerobic phenol degradation did not occur), while the Ks value decreased from 183 to 162, 92, and 73 mg phenol dm−3 (ie phenol bulk concentration influenced the specific phenol utilization rate). However, the trend towards variations of Ki with increasing temperature was not observed. The k value (at 35 ◦C) obtained from the present study falls in typical range.7,20 3.3.3 Parameters k, Ks1, and Ki for phenol acidogenesis To estimate the Haldane kinetic constants for phenol acidogenesis (k1, Ks1, and Ki) at the temperatures of 25, 30, 35, and 40 ◦C, the same dispersed sludge used for the determination of the Haldane kinetic constants k, Ks, and Ki, was also used to determine the mass fraction of methanogens. Similar to the results of the k values, the k1 value gradually increased from 0.78 to 0.88, 1.03, and 1.23 mg phenol mg VSS−1 day−1 when the temperature was increased from 25 to 30, 35, and 40 ◦C (Table 5). 3.3.4 Activation energy and temperature coefficient To comprehend the temperature dependency of reaction rates (k, k1, and k2) (in the batch reactors) and the specific phenol utilization rate (r) (in the UASB reactors), the Arrhenius relationship was used to estimate the activation energy (Ea) and temperature coefficient (θ). Figure 4 shows that the Ea values estimated for the anaerobic degradation of phenol (k), phenol acidogenesis (k1), and acetate methanogenesis (k2) in the batch reactors at the temperature range of 25 to 40 ◦C were 3063, 5640, and 6505 cal mol−1, respectively. The Ea value for acetate methanogenesis (k2) was larger than that for phenol acidogenesis (k1), indicating that the temperature imposed a more influential effect on methanogens than on acidogens. Moreover, Fig 4 shows that the Ea value estimated for the anaerobic degradation of phenol in the UASB reactors with the four respective us at the operational temperature range of 25 to 40 ◦C varied greatly from 778 to 1810 cal mol−1. The reason for great variations in Ea (in addition to the effect of the operational temperature) is that some other factors (eg superficial flow velocity, substrate bulk concentration, granule size, sludge age) also affect the performance of the UASB reactors. However, these Ea values were all smaller than that estimated for the anaerobic degradation of phenol in batch reactors (k) (3063 cal mol−1), disclosing that the operational temperature imposed a less influential effect on the UASB reactors than on the batch reactors. This is because mass transfer resistance in granular sludge Table 5. Parameters k1,Ks1, and Ki for phenol acidogenesis Temp (◦C) f k1a (mg phenol mg VSS−1 day−1) Ks1 (mg phenol dm−3) Ki (mg phenol dm−3) 25 0.23 0.78 183 60 30 0.26 0.88 162 82 35 0.29 1.03 92 69 40 0.38 1.23 73 68 a k1 = k/(1 − f). 804 J Chem Technol Biotechnol 79:797–808 (online: 2004) Effects of temperature in UASB reactors Figure 4. Activation energy (Ea) estimated from temperature dependency of reaction rates and anaerobic phenol degradation rate. grown in the UASB reactors is greater than that in dispersed sludge grown in the batch reactors. Figure 5 shows that the temperature coefficient (θ) estimated for the anaerobic degradation of phenol (k), phenol acidogenesis (k1 and Ks1), and acetate methanogenesis (k2 and Ks2) in the batch reactors at the temperature range of 25 to 40 ◦C were 1.017, 1.030, 0.941, 1.042, and 0.953, respectively. The θ value for acetate methanogenesis (k2) was larger than that for phenol acidogenesis (k1), implying that the temperature imposed a more influential effect on methanogens than on acidogens. In addition, both the θ values for phenol acidogenesis (Ks1) and acetate methanogenesis (Ks2) were smaller than unity, showing that both Ks1 and Ks2 decreased as the temperature was increased for phenol acidogenesis and acetate methanogenesis. The θ values for acetate methanogenesis (k2 and Ks2) in the present study were consistent with the results of Lin et al.28 Moreover, Fig 5 shows that the θ value estimated for the anaerobic degradation of phenol in the UASB reactors with the four respective us values at the operational temperature range of 25 to 40 ◦C ranged from 1.003 to 1.008. A θ value of close to unity implied that, under mesophilic conditions, the effect of the operational temperature on the specific phenol utilization rate in UASB reactors was somewhat insignificant. 3.4 Mass fractions and activity of acidogens and methanogens When the UASB reactor (under mesophilic condi- tions; 25, 30, 35, and 40 ◦C) reached steady state, the mass fractions of methanogens (f ) in the lower- , middle-, and upper-part of the sludge bed were determined. The biomass and initial acetate concen- trations measured in 48 batch reactors (serum vials) ranged from 1020 to 1450 mg VSS dm−3 and 92 to 172 mg acetate dm−3, respectively. The obtained f values (arithmetic mean) ranged from 0.23 to 0.40 (Table 6), and the f value tended to increase with increasing operational temperature. This is consistent with our earlier results (Section 3.3.4) that the oper- ational temperature imposed a more influential effect on methanogens than on acidogens. In addition, the rate-limiting step in the anaerobic degradation of the inhibitory substrate phenol is acidogenesis.30 Thus, under the adverse operating conditions (eg low opera- tional temperature), a high yield of acidogens relative to methanogens might occur, resulting in a low f value. To comprehend whether the activities of acidogens and methanogens vary with different operational temperatures in the UASB reactors, the specific COD utilization rates of acidogens plus methanogens, acidogens, and methanogens were calculated using the performance data (Table 2) together with the experimental f values. Table 6 shows that the specific COD utilization rates of acidogens plus methanogens and acidogens at 25 and 30 ◦C were the same; both increased with a further increase in operational temperature (35 and 40 ◦C). However, the specific COD utilization rate of methanogens decreased with an increase in operational temperature (from 25 to 40 ◦C) (Table 6). Moreover under mesophilic conditions, the specific COD utilization rates for phenol acidogenesis were significantly lower than those for acetate methanogenesis, implying that the rate-limiting step in the anaerobic degradation of the inhibitory substrate phenol is acidogenesis. 4 CONCLUSIONS Under mesophilic conditions (25–40 ◦C) and at the volumetric loading of 10.5 g COD dm−3 day−1, the UASB reactors at 30 ◦C retained the greatest amount of biomass, resulting in very efficient for COD removal (97.4–98.4%); the greatest wash-out J Chem Technol Biotechnol 79:797–808 (online: 2004) 805 H-H Chou, J-S Huang, W-F Hong Figure 5. Temperature coefficient (θ) estimated from temperature dependency of reaction rates and anaerobic phenol degradation rate. Table 6. Mass fractions of methanogens and specific substrate utilization rates of acidogens and methanogens in UASB reactorsa f Biomass (g) Specific substrate utilization rate (mg COD mg VSS−1 day−1) Test run Temp (◦C) us (mh−1) Lower Middle Upper Total Acidogens Methanogens Overall Acidogens Methanogens A1 25 0.5 0.22 0.21 0.19 107.6 85.0 22.6 0.35 0.45 1.68 B1 25 1.0 0.25 0.23 0.22 112.7 86.8 25.9 0.33 0.44 1.46 C1 25 2.0 0.28 0.26 0.25 120.6 89.2 31.4 0.31 0.43 1.22 D1 25 4.0 0.30 0.25 0.24 98.8 74.1 24.7 0.37 0.51 1.52 Arithmetic mean 0.26 0.24 0.23 109.9 83.8 26.2 0.34 0.46 1.47 A2 30 0.5 0.24 0.23 0.23 115.5 88.9 26.6 0.34 0.44 1.47 B2 30 1.0 0.26 0.26 0.25 119.3 88.3 31.0 0.33 0.45 1.28 C2 30 2.0 0.30 0.28 0.25 129.8 93.5 36.3 0.30 0.43 1.09 D2 30 4.0 0.31 0.27 0.23 103.7 75.7 28.0 0.37 0.52 1.40 Arithmetic mean 0.28 0.26 0.24 117.1 86.6 30.5 0.34 0.46 1.31 A3 35 0.5 0.30 0.28 0.26 103.9 74.8 29.1 0.37 0.52 1.34 B3 35 1.0 0.29 0.28 0.26 104.4 75.2 29.2 0.38 0.53 1.36 C3 35 2.0 0.34 0.32 0.25 117.1 79.6 37.5 0.34 0.50 1.06 D3 35 4.0 0.33 0.27 0.22 100.7 73.5 27.2 0.39 0.54 1.45 Arithmetic mean 0.32 0.29 0.25 106.5 75.8 30.8 0.37 0.52 1.30 A4 40 0.5 0.37 0.38 0.32 103.3 65.2 38.1 0.38 0.60 1.00 B4 40 1.0 0.41 0.41 0.33 103.2 60.9 42.3 0.38 0.65 0.94 C4 40 2.0 0.39 0.40 0.35 116.0 69.6 46.4 0.34 0.57 0.86 D4 40 4.0 0.40 0.39 0.37 96.7 59.0 37.7 0.40 0.67 1.04 Arithmetic mean 0.39 0.40 0.34 104.8 63.7 41.1 0.38 0.62 0.96 a Vol loading = 10.5 g COD dm−3 day−1. 806 J Chem Technol Biotechnol 79:797–808 (online: 2004) Effects of temperature in UASB reactors of sludge from the UASB reactors was at 25 ◦C, resulting in less efficient COD removal (92.6–94.8%). By observing the performance data, the granule size was negatively correlated with the amount of wash- out of sludge. The least amount of wash-out of sludge occurred in the UASB reactors at 30 ◦C, resulting in the largest granule size. To avoid the negative effect of inhibitory substrates on anaerobic bacteria and wash-out of small granules, the superficial flow velocity (us) of 1.0–2.0 m h−1 is suggested if the UASB reactor is applied to treat high-strength inhibitory substrates. According to the obtained kinetic constants for anaerobic phenol degradation (k and Ks) and acetate methanogenesis (k2 and Ks2), the k and k2 values gradually increased with an increase in temperature (ie denaturation of the enzyme protein in anaero- bic phenol degradation and acetate methanogenesis did not occur), while the Ks and Ks2 values grad- ually decreased with an increase in temperature (ie phenol and acetate bulk concentrations influ- enced specific substrate utilization rate). In addi- tion, the mass fraction of methanogens (f ) esti- mated from the UASB reactors under mesophilic conditions ranged from 0.23 to 0.40; and the f value increased with increasing operational tem- perature (ie the operational temperature imposed a more influential effect on methanogens than on acidogens). The activation energy (Ea) for acetate methanogen- esis in the batch reactors (6505 cal mol−1) was larger than that for phenol acidogenesis (5640 cal mol−1), indicating that the temperature imposed a more influ- ential effect on methanogens than on acidogens. Consequently, the mass fractions of methanogens (f) in the UASB reactors increased with increas- ing operational temperature. In addition, the Ea for the anaerobic degradation of phenol in the UASB reactors (778–1810 cal mol−1) was smaller than the Ea in the batch reactors (3063 cal mol1), disclos- ing that the operational temperature imposed a less influential effect on the UASB reactors than on the batch reactors. The estimated values of tempera- ture coefficient (θ) were close to unity, disclosing that the effect of the operational temperature on the specific phenol utilization rate in the UASB reactors was somewhat insignificant. Moreover, the specific COD utilization rates for phenol acidogen- esis under mesophilic conditions were significantly lower than those for acetate methanogenesis, imply- ing that the rate-limiting step is phenol acidogene- sis. ACKNOWLEDGMENT Financial support of this research by research grant NSC 89-2211-E-006-108 from the National Science Council of the Republic of China is greatly appreciated. REFERENCES 1 Lettinga G, Hobma SW, Hulshoff Pol LW, de Zeeuw W, de Jong P, Grin P and Roersma R, Design operation and economy of anaerobic treatment. Wat Sci Tech 15:177–195 (1984). 2 Duff SJB, Kenndy KJ and Brady AJ, Treatment of dilute phenol/PCP wastewaters using the upflow anaerobic sludge blanket (UASB) reactor. Wat Res 29:645–651 (1995). 3 Chang YJ, Nishio N and Nagai S, Characteristics of granular methanogenic sludge grown on phenol synthetic medium and methanogenic fermentation of phenolic wastewater in a UASB reactor. J Ferment Bioeng 79:348–353 (1995). 4 Fang HHP, Chen T, Li YY and Chui HK, Degradation of phenol in wastewater in an upflow anaerobic sludge blanket reactor. Wat Res 30:1353–1360 (1996). 5 Zhou GM and Fang HHP, Co-degradation of phenol and m- cresol in a UASB reactor. Biores Tech 61:47–52 (1997). 6 Tay JH, He YX and Yan YG, Improved anaerobic degradation of phenol with supplemental glucose. J Environ Eng (ASCE) 127:38–45 (2001). 7 Jih CG, Huang JS and Huang SY, Process kinetics of upflow anaerobic sludge bed reactors treating inhibitory substrate. Wat Environ Res 75:1–10 (2003). 8 Huang SY, Kinetic behavior of UASB reactors treating inhibitory substrate. MS Thesis, Department of Environmen- tal Engineering, National Cheng Kung University, Tainan city, Taiwan, ROC (1999). 9 Lawrence AW and McCarty PL, Kinetics of methane fermen- tation in anaerobic treatment. J Wat Pollut Control Fed 41:R1–R17 (1969). 10 Westermann P, Ahring BJ and Mah RA, Temperature compen- sation in Methanosarcina barkeri by modulation of hydrogen and acetate affinity. Appl Environ Microbiol 55:1262–1266 (1989). 11 Kalyuzhnyi SV, Martinez EP and Martinez JR, Anaerobic treatment of high-strength cheese-whey wastewaters in laboratory and pilot UASB-reactors. Biores Tech 60:59–65 (1997). 12 Joubert WA and Britz TJ, The effect of pH and temperature manipulation on metabolite composition during acidogenesis in a hybrid anaerobic digester. Appl Microbiol Biotechnol 24:253–258 (1986). 13 Arrhenius S, U¨ber die reaktionsgeschwindigkeit bei der inver- sion von rohrzucker durch sauren. Zeitschrift fu¨r Physikalische Chemie 4:226–248 (1889). 14 Hnze M and Harremoes P, Anaerobic treatment of wastewater in fixed film reactors. Wat Sci Tech 15:1–101 (1983). 15 Kettunen RH and Rintala JA, The effect of low temperature (5–29 ◦C) and adaptation on the methanogenic activity of biomass. Appl Microbiol Biotechnol 48:570–576 (1997). 16 Barbosa RA and Sant’anna JR GL, Treatment of raw domestic sewage in an UASB reactor. Wat Res 12:1483–1490 (1989). 17 Fang HHP, Li YY and Chui HK, UASB treatment of wastew- ater with concentrated mixed VFAs. J Environ Eng (ASCE) 121:153–160 (1995). 18 Agrawal LK, Harada H and Okui H, Treatment of dilute wastewater in a UASB reactor at moderate temperature: performance aspects. J Ferment Bioeng 83:179–184 (1997). 19 Fang HHP and Chung DWC, Anaerobic treatment of pro- teinaceous wastewater under mesophilic and thermophilic conditions. Wat Sci Tech 40:77–84 (1999). 20 Suidan MT, Najm IN, Pfeffer JT and Wang YT, Anaerobic biodegradation of phenol: inhibition kinetics and system stability. J Environ Eng (ASCE) 114:1359–1376 (1988). 21 Wen TC, Cheng SS and Lay JJ, A kinetic model of a recirculated upflow anaerobic sludge blanket treating phenolic wastewater. Wat Environ Res 66:794–799 (1994). 22 Lay JJ and Cheng SS, Influence of hydraulic loading rate on UASB reactor treating phenolic wastewater. J Environ Eng (ASCE) 124:829–837 (1998). J Chem Technol Biotechnol 79:797–808 (online: 2004) 807 H-H Chou, J-S Huang, W-F Hong 23 Press WH, Flannery BP, Tenkolsky SA and Vetterling WT, Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, London (1986). 24 APHA, AWWA and WPCF, Standard Methods for the Examina- tion of Water and Wastewater, 19th edn. APHA, Washington, DC (1995). 25 Speece RE, Anaerobic Biotechnology for Industrial Wastewaters. Archae Press, Nashville, Tennessee (1996). 26 Schmidt JE and Ahring BK, Granular sludge formation in upflow anaerobic sludge blanket (UASB) reactors. Biotechnol Bioeng 49:229–246 (1996). 27 Wu CS and Huang JS, Bioparticle characteristics of tapered anaerobic fluidized-bed bioreactors. Wat Res 30:233–241 (1996). 28 Lin CY, Noike T, Sato K and Matsumoto J, Temperature characteristics of the methanogenesis process in anaerobic digestion. Water Sci Technol 19:299–310 (1987). 29 Grady CPL, Daigger GT and Kim HC, Biological Wastewater Treatment, 2nd edn, revised and expanded. Marcel Dekker, Inc. New York (1999). 30 Jih CG and Huang JS, Effect of biofilm thickness distribution on substrate-inhibited kinetics. Wat Res 28:967–973 (1994). 808 J Chem Technol Biotechnol 79:797–808 (online: 2004)