Methane production from secondary paper and pulp sludge: Effect of natural zeolite and modeling

Chemical Engineering Journal 257 (2014) 131–137

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Methane production from secondary paper and pulp sludge: Effect of natural zeolite and modeling C. Huiliñir a,⇑, Alejandra Quintriqueo a, Christian Antileo b, Silvio Montalvo a a b

Department of Chemical Engineering, Universidad de Santiago de Chile, Santiago, Chile Department of Chemical Engineering, Universidad de La Frontera, Temuco, Chile

h i g h l i g h t s  The effect of zeolite on methane production from paper and pulp sludge was studied.  Modeling of methane production by three simplified models was evaluated.  Doses between 0.2 and 1 g/L of zeolite increased the accumulated methane.  A dose of 20 g/L of zeolite decreased methane production.  The modified Gompertz and logistic function models fit the experimental data.

a r t i c l e

i n f o

Article history: Received 21 March 2014 Received in revised form 10 July 2014 Accepted 12 July 2014 Available online 19 July 2014 Keywords: Paper and pulp sludge Methane production Zeolite Modified Gompertz model Logistic function model Modeling

a b s t r a c t The effect of natural zeolite on methane production from secondary paper and pulp sludge and its modeling are evaluated. Five tests with zeolite concentrations of 0, 0.2, 0.5, 1 and 20 g/L were evaluated at 30 °C. The modified Gompertz equation, the logistic function and the transfer function were evaluated. The results show that doses between 0.2 and 1 g/L of zeolite produce a statistically significant increase of accumulated methane, giving values greater than 183 mL for CH4/g of VSadded. On the contrary, the 20 g/L dose of zeolite reduced the methane production. The modified Gompertz and the logistic function models concur with the experimental data, the former having the best fit. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction The paper and cellulose manufacturing industry is one of Chile’s main industrial sectors. The country’s production of paper and pulp had increased to 4.94 million tons per year by 2008 [1]. The process produces wastewater that is treated by a conventional primary– secondary treatment. The secondary treatment, namely aerobic activated sludge, generates secondary sludge with high moisture content composed of microbial biomass, cellulose, cell decay products, and non-biodegradable lignin precipitates [2]. One way of treating this pulp and paper sludge (PPS) is anaerobic digestion (AD) to generate methane. This treatment helps to reduce sludge volume, degrading potentially toxic compounds, recovering energy and nutrients that can be applied as fertilizers or added to degraded forest and agricultural soil. Recently, several ⇑ Corresponding author. Tel.: +56 (2) 27181814. E-mail address: [email protected] (C. Huiliñir). http://dx.doi.org/10.1016/j.cej.2014.07.058 1385-8947/Ó 2014 Elsevier B.V. All rights reserved.

authors have explored the use of anaerobic digestion in order to recover or reuse secondary sewage sludge from paper mills [2–5]. However, the main problem regarding PPS is the low biogas production compared to other sewage sludge. According to Astals et al. [6], methane yield from sewage sludge coming from municipal wastewater plants varied between 324.5 and 379.7 mL CH4/g volatile solid added (VSadded), while methane yields from PPS ranged from a very low 50 mL CH4/g VSadded for bleached Kraft [7] to 199 mL CH4/g VSadded for chemi-thermo mechanical and Kraft pulp [4]. These values are almost 50% lower than the values reported for methane production from municipal wastewater plants. In order to improve this yield, several pretreatments have been studied, including thermal, ultrasound, ozone oxidation, alkaline, enzymatic and mechanical [8]; however, this process may increase the cost and cause operational problems regarding AD. A cheaper way to increase methane production is by using zeolite as amendment of anaerobic digestion. Zeolites are crystalline minerals that have cavities forming channels in their structure.

132

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

Because of this, some molecules can go into the channels or cavities, filling the available space. Several researchers [9–12] have found that natural zeolite is a good supporting material in the anaerobic digestion process of different substrates, because it has a high capacity for immobilizing microorganisms or trapping organic materials, making their interaction more accessible in the whole reactor. Several reports have shown that the use of zeolite in the anaerobic digestion of different kinds of wastewater (municipal, piggery industry, synthetic) and solid waste (grass silage) improves methane yield [13]. The effect of natural zeolites on the anaerobic degradation of synthetic substrates such as acetate and methanol was evaluated by Milan et al. [14], showing that the addition of zeolites determined an increase in the apparent kinetic constant of the process and achieved values twice as high as those seen in a control reactor. Montalvo et al. [9] studied the influence of particle size in the range of 0.07–1 mm and zeolite doses in the range from 0.05 to 0.40 g zeolite/g of inoculum (volatile suspended solids, VSS) on the anaerobic digestion of synthetic wastewater, showing that the anaerobic process was favored by the addition of zeolite at doses between 0.05 and 0.30 g/g VSS, with an optimum value of 0.10 g/g VSS. The effect of different natural zeolite concentrations on the anaerobic digestion of piggery waste was studied in batch mode by Milan et al. [10], who found that the anaerobic process was favored by the addition of natural zeolite in doses between 2 and 4 g/L and increasingly inhibited in doses beyond 6 g/L. The effect of adding natural zeolite to the batch thermophilic anaerobic decomposition of pig wastes was also studied at zeolite doses of 0, 4, 8 and 12 g of zeolite/L of waste at 55 °C [12]. In this case, methane production was up to 65% higher in treatments with natural zeolite at doses of 8 and 12 g/L of waste, compared to those without zeolite. Furthermore, in treatments with natural zeolite, the reduction of volatile solids and biological oxygen demand (BOD5) was statistically significant. Even though the use of zeolite as amendment has shown positive results, there are no studies that evaluate the use of zeolite as amendment in the anaerobic digestion of PPS, and this is one of the goals of this work. Studying the kinetics of methane production from feedstock(s) is important when designing and evaluating anaerobic digesters. First-order models are common models for describing methane production from solid waste materials compared with soluble substrates [15]. Few studies have applied mathematical models to anaerobic degradation using zeolite as amendment, focusing mainly on the removal of organic matter. Montalvo et al. [9], Milan et al. [10] and Montalvo et al. [11] applied a Contois kinetic model in order to determine the effect of zeolite on the anaerobic digestion of PW, focusing their efforts on reproducing chemical oxygen demand (COD) removal. Recently, Montalvo et al. [16] used the Monod kinetic equation to explain the effect of zeolite on anaerobic digestion of synthetic wastewater. None of these studies applied a model for methane production, so the study of the effect of zeolite on the parameters of three practical mathematical models was another aim of the present work, using natural zeolite in the production of methane from residual sludge derived from the paper industry. 2. Materials and methods 2.1. Experimental setup Every experimental run considered the installation of twenty 280-mL anaerobic mini-digesters with an effective operating volume of 250 mL. Thirteen of them were meant for measuring parameters in the liquid phase (chemical oxygen demand (COD), volatile solids (VS), pH); five were used to measure methane by liquid displacement using a system shown schematically in Fig. 1; and two were used as controls, one without adding inoculum and one with-

H2O T=30°C NaOH Fig. 1. Schematic of the experimental setup of the anaerobic digestion process.

out adding secondary sludge. The mini-digesters operated in discontinuous mode for 33 days, at which time gas accumulation remained constant. Manual stirring was performed once or twice per day before reading the volume displaced by methane. The volume of the digesters was completed with distilled water, they were stoppered with rubber stoppers and sealed with white silicone to ensure anaerobiosis, and they were covered with aluminum foil to prevent the growth of photosynthetic organisms. All the runs were done in duplicate. 2.2. Inocula, substrate and experimental design The inoculum was obtained from an anaerobic reactor of the La Farfana Water Treatment Plant of Aguas Andinas in Santiago, Chile. The substrate was secondary sludge from the Liquid Industrial Residues Treatment Plant of Papeles Cordillera, Santiago, Chile. Table 1 shows their characteristics. Zeolite was obtained from a company that commercializes natural zeolite and is located in Quinamávida, Linares, VII Region, Maule, Chile. In each treatment 25 mL of inoculum and 25 g of sludge were added, completing the reactor’s volume with distilled water up to 250 mL. With these quantities, the inoculum-substrate ratio was 0.25 g VS inoculum/g VS substrate in all the assays. The low value of inoculum-substrate ratio (high F/M ratio) was in order to reduce the lag-phase, because of the inoculum coming from an anaerobic digester of municipal wastewater. The zeolite masses were 0, 0.05, 0.125, 0.25 and 5 g for 0, 0.2, 0.5, 1 and 20 g/L concentrations, respectively. For each assay, three weekly measurements were made in duplicate of soluble and total COD, suspended VS, and pH. The temperature in the container was kept at 30 °C using three automatically controlled aquarium heaters. 2.3. Chemical analyses The following parameters were determined: total COD (tCOD), soluble COD (sCOD), total and suspended solids, volatile suspended solids, and pH. COD (total and soluble) was measured by colorimetric method according to APHA [17]. Total and suspended solids, Table 1 Characteristics of pulp and paper sludge and inoculum used in the study ((a) n = 6, (b) n = 8). Parameter

Inoculum

Pulp and paper sludge

TSS (mg/L) VSS (mg/L) TS (mg/L) VS (mg/L) pH %C, dry weight %N, dry weight C/N ratio

10630.667 ± 1307.934(a) 16696.667 ± 936.412(a) 28525 ± 2501.629(a) 17212 ± 654.413(a) 7.69 51.3 ± 1.5 1.8 ± 0.2 28.5 ± 7.5

– – 0.118 ± 0.014 g/g(b) sludge 0.107 ± 0.013 g/g(b) sludge – 51.1 ± 2.4 1.5 ± 0.3 34.1 ± 8

133

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

volatile suspended solids and pH were measured according to APHA [17]. Methane production was measured by volumetric displacement, connecting an inverted falcon tube containing a 3% w/w

(A)

NaOH solution in order to eliminate the CO2 and H2S as main impurities from the biogas, displacing only the methane volume (Fig. 1). The resulting biogas travels through the flexible tubing to the Falcon tube, where it comes in contact with the NaOH

16000

[z] [z] [z] [z] [z]

VSS concentration, mg/L

14000

= = = = =

0 g/L 0.2 g/L 0.5 g/L 1 g/L 20 g/L

12000

10000

8000

6000

4000 0

5

10

15

20

25

30

35

time, d

(B)

20000 2000 1800

tCOD

tCOD concentration, mg/L

sCOD 1600 16000 1400 14000

1200 1000

12000

sCOD concentration, mg/L

18000

800 10000 600 8000

400 0

5

10

15

20

25

30

35

20

25

30

35

time, d

(C)

7.5

pH value

7.0

6.5

6.0

5.5

5.0 0

5

10

15

time, d Fig. 2. Variation of the concentration of VSS, COD and pH in the process. (A) Variation of the VSS at different zeolite concentrations. (B) Variation of sCOD and tCOD for z = 0.5 g/ L. (C) Variation of pH for z = 0.5 g/L.

134

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

solution, forming sodium carbonate and only bubbling methane. The methane collected in the Falcon tube is sucked with a syringe and released in a flame to observe the presence of methane. 2.4. Modeling of methane production Three models were used to estimate the production of methane in batch tests. The modified Gompertz model and the logistic function model correspond to sigmoidal functions that relate methane production with the growth of methanogenic archaea in the biodigester [5,18,19]. Another model that has been used is the transfer function [18,19], which suggests that the production of methane resembles a first order curve with maximum rate at the beginning, decreasing practically to zero at the end of the experimental period [20]. The models are: Modified Gompertz model

n hl  e io y ¼ A  exp exp m  ðk  tÞ þ 1 A

ð1:1Þ

Logistic function model

y¼

A   1 þ exp 4lmAðktÞ þ 2

ð1:2Þ

Transfer function model

   l  ðt  kÞ y ¼ A  1  exp  m A

ð1:3Þ

where A represents potential methane production (mL CH4 g1), lm is the maximum rate of methane production (mL CH4 g1 h1), k is the lag time phase (h), y is the methane accumulated at time t, t is the measured time (h), and e is the base of the natural logarithms [2.718282]. 2.5. Statistical analyses To determine the differences between each treatment a onefactor variance analysis (ANOVA) was applied, using Excel 2010. The constants of the modified Gompertz kinetic model, the logistics and the transfer function were determined using the MATLAB 7.6.0 (R2008a) lsqcurvefit function, which solves nonlinear curve fitting problems using the least squares method, that consists in decreasing the sum of the squares of the deviations of the points from the fitted function. The degree of fit of the experimental data to the models was established through the estimated error variance, s2, determined from i X ^ i Þ2 ðyi  y

s2 ¼

1

ð1:4Þ

NK

^i is the value calculated by the where yi is the experimental value, y model, N is the number of samples, and K is the number of model parameters.

3. Results and discussion 3.1. Effect of zeolite concentration on VSS degradation, tCOD, sCOD and pH Fig. 2 shows the effect of zeolite on VSS, COD and pH. VSS concentration (Fig. 2A), decreases in all the tests, with the highest VSS reduction occurring at [z] = 0 (43%). At zeolite concentrations of 0.2, 0.5 and 1 g/L, VSS removal was almost 7% lower than in the process without zeolite. This lower VSS degradation can be explained by the higher microbial activity in the presence of zeolite, with greater biomass growth that coincides with higher methane production (see Table 2). Greater biomass growth produces lower VSS removal in the presence of zeolite at these concentrations. Furthermore, the addition of zeolite increases the probability of biofilm formation on the zeolite surface [21]. However, at a zeolite concentration of 20 g/L there is lower VSS removal and lower methane production. Therefore, at zeolite concentrations as high as 20 g/L, methanogenesis was not promoted. The VSS reduction obtained in this work agrees with values found in the literature. Lin et al. [3] got a VSS reduction of 52% using paper and pulp sludge and monosodium glutamate, while Hagelqvist et al. [22] got between 38% and 58% of VS reduction using secondary pulp and paper sludge and municipal sewage sludge as substrates. tCOD and sCOD are shown in Fig. 2B for [z] = 0.5 g/L. For the rest of the tests, the trend was similar. tCOD decreases in time, with its reduction agreeing with methane production and increasing pH (Fig. 2C). The variation was almost constant after t = 600 h, showing that after the 600 h the bioreaction had ended [11]. tCOD removal efficiency varied between 32.6% ([z] = 0.2 g/L) and 36.7% ([z] = 0.5 g/L). These values agree with tCOD removal reported in the literature [10,23]. Regarding sCOD concentration, it increased at first and then decreased, following the same trend as other reports [3,24]. This behavior can be attributed to the solubilization of particulate organic matter and the accumulation of volatile fatty acids (VFA) [3]. This can also explain the experimental drop of pH at 50–200 h (Fig 2C). The accumulation of sCOD ended at 250 h, coinciding with increasing methane production (Fig. 3). 3.2. Effect of zeolite on methane yield and methane production Table 2 shows the yield obtained in each test. The yield is expressed as mL CH4/g of VSadded (c) and as mL CH4/g of VSremoved (c0 ) A positive effect is seen when zeolite is added, resulting also in a small decrease at a zeolite concentration of 20 g/L. Since the data of the tests were found to be quite similar, they were subjected to an analysis of variance (ANOVA). The results are shown in Table 3, where it is seen that the methane yield varied significantly at the 5% level for zeolite additions of 0.2, 0.5 and 1 g/L. At a zeolite concentration of 20 g/L there is no significant difference in methane yield. Thus the methane yield using zeolite concentrations between 0.2 and 1 g/L was significantly different from that of all the other tests. The percentage increase or decrease of accumulated methane is also shown in Table 2, with the highest increase seen at a zeolite

Table 2 Yield of secondary paper sludge and percent removal of VS. Experimental test

c [mL CH4/gVSadded]

% increase from z = 0 g/L

% VS removal

c [mL CH4/gVSremoval]

% increase from z = 0 g/L

[z] = 0 [z] = 0.2 g/L [z] = 0.5 g/L [z] = 1 g/L [z] = 20 g/L

173.60 ± 5.87 191.46 ± 4.69 183.71 ± 4.99 186.19 ± 9.53 169.64 ± 6.55

– 10.29 5.82 7.25 2.28

43.33 40.41 42.27 40.19 37.32

400.92 ± 13.54 473.79 ± 11.61 434.61 ± 11.81 463.27 ± 23.71 454.55 ± 17.55

– 18.18 8.40 15.55 13.37

135

methane production, mL/g VSadded

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

Table 3 Results of the ANOVA analysis at the 5% level for different concentrations of zeolite compared to the test without zeolite.

250

A) 200 150 100 exp CH4

50

-50 0

5

10

15

20

25

30

35

methane production, mL/g VSadded

time, d 250

B) 200 150 100 exp CH4 GM model LF model TF model

50 0 -50 0

5

10

15

20

25

30

35

time, d methane production, mL/g VSadded

P

F

Fcrit

[z] = 0.2 [z] = 0.5 [z] = 1 [z] = 20

7.18  104 0.019 0.036 0.34

28.22 8.58 6.33 1.02

5.32 5.32 5.32 5.32

GM model LF model TF model

0

250 200

C)

150 100 exp CH4

50

GM model LF model TF model

0 -50 0

5

10

15

20

25

30

35

time, d methane production, mL/g VSadded

Experimental run

250

D)

200 150 100

exp CH4 GM model LF model TF model

50 0 -50 0

5

10

15

20

25

30

35

time, d Fig. 3. Models’ fit with methane production of anaerobic test using secondary sewage sludge as substrate. (A) Without zeolite; (B) with [z] = 0.2 g/L; (C) with [z] = 0.5 g/L; (D) with [z] = 1 g/L.

concentration of 0.2 g/L for both parameters c and c0 . The increase (10%) was similar to the maximum increase reported by Montalvo et al. [9] using PW and a zeolite concentration of 0.1 g z/g VSS.

The positive effect of zeolite has also been reported in the literature. For comparison purposes, the values used in this work were transformed from g/L of zeolite to g z/g of VSS, and the 0.2, 0.5, 1 and 20 g/L concentrations are equivalent to 0.1, 0.3, 0.6 and 12 g z/g of VSS. Montalvo et al. [9], working with pig and synthetic wastes, found that the anaerobic process was favored by the addition of natural zeolite at doses between 0.05 and 0.30 g/g of VSS, with an optimum value of 0.10 g/g of VSS. This last value agrees with the highest methane yield found in the present work. Therefore, concentrations lower than 0.1 g/g of VSS (0.2 g/L zeolite) will not significantly improve the methane production. Furthermore, Milan et al. [10] studied the effect of zeolite on the anaerobic digestion of PW, showing that zeolite doses between 2 and 4 g/L improve methane yield up to 35%, but the use of high doses (6 g/ L) decreases methane yield, a low value compared to the 20 g/L used in this work. Table 2 also shows that a zeolite concentration of 20 g/L decreases methane production by almost 3%, although c0 was 13% higher than the assay without zeolite. This decrease of methane yield concurs with the lower methane production and may be attributed to increased media viscosity with increasing amounts of solids, causing a more difficult mass transfer between the liquid phase and the microorganisms [9,11]. However, it has been determined that to guarantee an adequate flow regime the total solids concentration in the anaerobic processes must be between 4% and 8%, which corresponds to 1–12% and 2–99% for treatment at zeolite concentrations of 0 and 20 g/L, respectively, indicating that the solutions are diluted. These low percentages are due to the high moisture content of the sludge, which is 87.7%, so when 25 g of sludge are added, only 1.48 g of total solids are added. It should be mentioned that a decrease of 2.29% of methane production is obtained compared with negative effects at concentrations of 8 and 10 g/L [10], which cause reductions of approximately 10% and 30%, respectively, or as in the case of Montalvo et al. [9], where a decrease of 5% is obtained working at 0.4 g z/g of VSS. Working at higher zeolite concentrations, when a greater negative effect would be expected, this is not observed. In the cases in which PW is used, the effect of zeolite is related with the reduction of the ammonium ion concentration, which is produced during the anaerobic degradation of proteins, amino acids and urea [9,12]. In the case of paper industry sludge, the C/N ratio is high, so the nitrogen content is low [3], therefore there should not be an increase of free ammonium concentration. Taking this into account, the positive effect seen when adding zeolite (higher methane production and higher biomass growth) can be related to the immobilization of microorganisms [21,25]. This immobilization generated a biofilm where the microorganism can developed in a better environment, increasing its metabolism and improving the yield of the system. However, this biofilm also can affect the rate of process, specifically the methane production rate, due to the mass transfer problems related with this kind of systems [26]. In order to clarify this situation, a model that can calculate the methane production rate could help to visualize the apparition of this problem. In general, the yields obtained in this work are similar to other results reported in the literature. Puhakka et al. [27] determined an

136

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

average production of 220 mL/g of VSadded (570 mL/g of VSremoved) using sewage sludge from Kraft pulp-mill wastewater treatment, values slightly higher (18%) than those obtained in this work. Karlsson et al. [4] got values between 100 and 220 mL/g of VSadded for 20-day batch tests, with four out of six experiments with values similar to those obtained in this work. Bayr and Rintala [2] got values three to four times lower than those of this work, with yields between 50 and 100 mL/g of VSadded, even operating in the thermophilic range. Lin et al. [3], working with a co-digestion of pulp and paper sludge and monosodium glutamate waste liquor obtained a maximum value of 200 mL/g of VSadded, 10% higher than the values obtained in this work. The differences in these values are related to the different compositions and wastewater process used by the different plants [4]. With regard to the removal of VS achieved in this work (36–43%), these values are similar to those obtained in other studies. Bayr and Rintala [2] achieved a removal of 25–40%, Lin et al. [3] got up to 35% removal, while Puhakka [27] reported a 40% reduction of VS. This relatively low biodegradability of secondary sludge is related to its lignin content, since the anaerobic archaea and bacteria have a low degradation rate of this component. Furthermore, the composition of secondary sludge coming from paper and pulp mills can be very heterogeneous, with important amounts of anaerobically non-biodegradable material [2].

3.3. Application of simplified models Fig. 3 shows the model’s fit with the experimental data. The values of the parameters obtained are given in Table 4. Fig. 3 shows that of the proposed models, the modified Gompertz is the one that best fits the experimental data, with the lowest s2 value for all the runs. The logistic function can also predict the behavior of the experimental data, but with a lower fit (s2 between 15 and 27) than the modified Gompertz function (Table 4). In contrast, the transfer function does not properly reproduce the experimental data for any of the zeolite concentrations studied, with extremely high values for parameter A (Table 4). Therefore, the modified Gompertz function and the logistic function describe methane production adequately working with secondary sludge from the paper industry. This fitting order (Gompertz–logistic) is similar to that obtained by Zwietering et al. [28] when describing growth curves of Lactobacillus plantarum and by Mu et al. [29] recording hydrogen production. The modified Gompertz model and the logistic function assume that, under batch conditions, the methane production rate is proportional to the specific growth rate of methanogenic archaea in the reactor, with a sigmoidal production trend as usually seen in growth curves [30], with very slow growth at the beginning and

Table 4 Values of the parameters calculated for the different tests. s2

Experimental run Model A [mL g1]

lm [mL g1 h1] k [h]

[z] = 0 g/L

GM FL FT

184.294 171.893 444.373

0.415 0.446 0.348

134.512 6.04 156.327 15.45 72.781 104.06

[z] = 0.2 g/L

GM FL FT

219.828 196.440 1893.867

0.395 0.429 0.3119

139.487 5.01 165.488 15.07 70.6965 76.54

[z] = 0.5 g/L

GM FL FT

235.333 0.347 197.210 0.381 3956927.630 0.276

164.596 9.71 193.901 25.77 92.274 60.44

[z] = 1 g/L

GM FL FT

220.817 0.354 188.350 0.386 3741294.438 0.275

170.801 5.89 197.323 26.55 95.485 64.23

at the end of the study period, with an exponential increase between them. The modified Gompertz model and the logistic function were modified so that they would contain biological parameters [28]. As for the influence of the substrate used in this study, the origin of the growth models like those of Gompertz and the logistic function only describes the number of organisms and does not consider the consumption of substrate as is done with Monod’s equation, for example. Therefore, the substrate should not affect the application of those models [28]. In all the treatments the transfer function showed values that differed from those observed experimentally. The values of s2 were always higher compared to the Gompertz and logistic equations. As seen in Fig. 3, the model does not fit the points of the curve where there is no methane production (phase lag), so this model would be adequate only for describing the production of biogas instead of methane production. In the transfer function it is seen that parameter A (methane production potential) has high values that do not fit the range of methane produced in the runs. This behavior may be related to the shape of the curve, because in other studies in which this model was used [18,19] the curves consider only the exponential and stationary stages in the production of gas. Therefore, the transfer model can be used only if the lag phase of the culture is close to zero. Considering the modified Gompertz model, Table 4 shows that parameters A and k increase in all the cases in which zeolite was added, while parameter lm decreases, although this decrease is less than 17%. The greater value of A (potential methane production) is explained by the lower removal of VS and the greater generation of methane in the presence of zeolite, while the increase of the lag phase is explained by the greater presence of solids, making the hydrolytic stage difficult and increasing the acclimation of the microorganisms. The decreasing of lm (maximum production rate of methane) could be attributed to the presence of biofilm on the zeolite surface. According to the literature, the decreasing of maximum production rate of methane is related to mass transfer problems due to the increase of total solids (TS) content in the system [31,32]. In fact, Raposo et al. [31] indicated that the content of solids should not exceed 10% if an adequate mass transfer is to be assured. In our case, the TS was always lower that 10% (3.05% at Z = 0.2 g/L; 3.125% at Z = 0.5 g/L; 3.25% at Z = 1 g/L; 8% at Z = 20 g/ L), showing that, at least theoretically, the system is not under mass transfer problems by TS content. However, the value of lm decreases in presence of zeolite, showing an effect of mass transfer in the process not related to the TS content, but related to the mass transfer problems in the biofilm generated on zeolite’ surface [26]. According to the fit shown in Table 4, the highest production of methane (higher value of A) would be for a zeolite concentration of 0.5 g/L, the smallest variation of the maximum production rate of methane (lm) with regard to the process without zeolite would correspond to a concentration of 0.2 g/L, while the smallest increase of k with regard to the culture without zeolite would also be 0.2 g/L. From the above it can pointed out that the parameters of the Gompertz model agree in indicating that the best methane generation performance would occur with a dose of 0.2 g/L, a situation that was found experimentally. The modified Gompertz parameters were also determined by other studies. Donoso-Bravo et al. [18] showed that for secondary sludge, A had a value of 243.5 mL/g of VS, similar to that obtained in the present work. With respect to lm, it had a value of 1.48 mL/g of VS h, 3.5 times higher that the value obtained in this work. Using paper sludge, Parameswaran and Rittmann [5] obtained a value of A = 35.49 mL CH4/g of TS, 6 times lower that the value obtained in this work (213.94 mL CH4/g of TS). Regarding lm, Parameswaran and Rittmann [5] got a value of 0.64 mL/h, 1.85 times lower than the value obtained in this work (1.19 mL/h). Thus, the values obtained are in the range presented in the literature.

C. Huiliñir et al. / Chemical Engineering Journal 257 (2014) 131–137

4. Conclusions The addition of natural zeolite in the process at doses between 0.2 and 1 g/L causes a statistically significant increase of accumulated methane after 33 days of operation. Methane generation from residual sludge of the paper industry water treatment can be described by the modified Gompertz model, with better fit for all the treatments, compared with the logistic function and the transfer models. Acknowledgements This work was supported by FONDECYT (Chile) under Project No. 11121160 and FONDECYT (Chile) under Project No. 1130315. References [1] D. Rios, C. Perez, M. Sandoval, Phytotoxic effect of paper pulp sludge on Alfisol soil, J. Soil Sci. Plant Nutr. 12 (2012) 315–327. [2] S. Bayr, J. Rintala, Thermophilic anaerobic digestion of pulp and paper mill primary sludge and co-digestion of primary and secondary sludge, Water Res. 46 (2012) 4713–4720. [3] Y.Q. Lin, D.H. Wang, Q. Li, M.Q. Xiao, Mesophilic batch anaerobic co-digestion of pulp and paper sludge and monosodium glutamate waste liquor for methane production in a bench-scale digester, Bioresour. Technol. 102 (2011) 3673–3678. [4] A. Karlsson, X.B. Truong, J. Gustavsson, B.H. Svensson, F. Nilsson, J. Ejlertsson, Anaerobic treatment of activated sludge from Swedish pulp and paper mills – biogas production potential and limitations, Environ. Technol. 32 (2011) 1559– 1571. [5] P. Parameswaran, B.E. Rittmann, Feasibility of anaerobic co-digestion of pig waste and paper sludge, Bioresour. Technol. 124 (2012) 163–168. [6] S. Astals, M. Esteban-Gutiérrez, T. Fernández-Arévalo, E. Aymerich, J.L. GarcíaHeras, J. Mata-Alvarez, Anaerobic digestion of seven different sewage sludges: a biodegradability and modelling study, Water Res. 47 (2013) 6033–6043. [7] S. Bayr, P. Kaparaju, J. Rintala, Screening pretreatment methods to enhance thermophilic anaerobic digestion of pulp and paper mill wastewater treatment secondary sludge, Chem. Eng. J. 223 (2013) 479–486. [8] A. Elliott, T. Mahmood, Pretreatment technologies for advancing anaerobic digestion of pulp and paper biotreatment residues, Water Res. 41 (2007) 4273– 4286. [9] S. Montalvo, F. Diaz, L. Guerrero, E. Sanchez, R. Borja, Effect of particle size and doses of zeolite addition on anaerobic digestion processes of synthetic and piggery wastes, Process Biochem. 40 (2005) 1475–1481. [10] Z. Milan, E. Sanchez, P. Weiland, R. Borja, A. Martin, K. Ilangovan, Influence of different natural zeolite concentrations on the anaerobic digestion of piggery waste, Bioresour. Technol. 80 (2001) 37–43. [11] S. Montalvo, L. Guerrero, R. Borja, L. Travieso, E. Sanchez, F. Diaz, Use of natural zeolite at different doses and dosage procedures in batch and continuous anaerobic digestion of synthetic and swine wastes, Resour. Conserv. Recycl. 47 (2006) 26–41. [12] T.A. Kotsopoulos, X. Karamanlis, D. Dotas, G.G. Martzopoulos, The impact of different natural zeolite concentrations on the methane production in thermophilic anaerobic digestion of pig waste, Biosyst. Eng. 99 (2008) 105– 111.

137

[13] S. Montalvo, L. Guerrero, R. Borja, E. Sánchez, Z. Milán, I. Cortés, M. Angeles de la la Rubia, Application of natural zeolites in anaerobic digestion processes: a review, Appl. Clay Sci. 58 (2012) 125–133. [14] Z. Milan, P. Villa, E. Sanchez, S. Montalvo, R. Borja, K. Ilangovan, R. Briones, Effect of natural and modified zeolite addition on anaerobic digestion of piggery waste, Water Sci. Technol. 48 (2003) 263–269. [15] H.M. El-Mashad, Kinetics of methane production from the codigestion of switchgrass and Spirulina platensis algae, Bioresour. Technol. 132 (2013) 305– 312. [16] S. Montalvo, P. Gonzalez, C. Mena, L. Guerrero, R. Borja, Influence of the food to microorganisms (F/M) ratio and temperature on batch anaerobic digestion processes with and without zeolite addition, J. Environ. Sci. Health Part AToxic/Hazard. Subst. Environ. Eng. 47 (2012) 1785–1794. [17] APHA, AWWA, WPCF, Standard Methods for the Examination of Water and Wastewater, 18th ed., APHA, Washington, D.C., 1992. [18] A. Donoso-Bravo, S.I. Perez-Elvira, F. Fdz-Polanco, Application of simplified models for anaerobic biodegradability tests. Evaluation of pre-treatment processes, Chem. Eng. J. 160 (2010) 607–614. [19] L.H. Li, X.Y. Kong, F.Y. Yang, D. Li, Z.H. Yuan, Y.M. Sun, Biogas production potential and kinetics of microwave and conventional thermal pretreatment of grass, Appl. Biochem. Biotechnol. 166 (2012) 1183–1191. [20] G. Redzwan, C. Banks, The use of a specific function to estimate maximum methane production in a batch-fed anaerobic reactor, J. Chem. Technol. Biotechnol. 79 (2004) 1174–1178. [21] N. Fernandez, S. Montalvo, F. Fernandez-Polanco, L. Guerrero, I. Cortes, R. Borja, E. Sanchez, L. Travieso, Real evidence about zeolite as microorganisms immobilizer in anaerobic fluidized bed reactors, Process Biochem. 42 (2007) 721–728. [22] A. Hagelqvist, Batchwise mesophilic anaerobic co-digestion of secondary sludge from pulp and paper industry and municipal sewage sludge, Waste Manage. 33 (2013) 820–824. [23] M.C. Tomei, C.M. Braguglia, G. Mininni, Anaerobic degradation kinetics of particulate organic matter in untreated and sonicated sewage sludge: role of the inoculum, Bioresour. Technol. 99 (2008) 6119–6126. [24] Y.Q. Lin, D.H. Wang, S.Q. Wu, C.M. Wang, Alkali pretreatment enhances biogas production in the anaerobic digestion of pulp and paper sludge, J. Hazard. Mater. 170 (2009) 366–373. [25] S. Weiss, A. Zankel, M. Lebuhn, S. Petrak, W. Somitsch, G.M. Guebitz, Investigation of mircroorganisms colonising activated zeolites during anaerobic biogas production from grass silage, Bioresour. Technol. 102 (2011) 4353–4359. [26] P. Buffiere, J.P. Steyer, C. Fonade, R. Moletta, Modeling and experiments on the influence of biofilm size and mass transfer in a fluidized bed reactor for anaerobic digestion, Water Res. 32 (1998) 657–668. [27] J.A. Puhakka, M. Alavakeri, W.K. Shieh, Anerobic treatment of Kraft pulp-mill waste activated sludge: gas production and solids reduction, Bioresour. Technol. 39 (1992) 61–68. [28] M.H. Zwietering, I. Jongenburger, F.M. Rombouts, K. Vantriet, Modeling of bacterial growth curve, Appl. Environ. Microbiol. 56 (1990) 1875–1881. [29] Y. Mu, H.Q. Yu, G. Wang, A kinetic approach to anaerobic hydrogen-producing process, Water Res. 41 (2007) 1152–1160. [30] L. Altas, Inhibitory effect of heavy metals on methane-producing anaerobic granular sludge, J. Hazard. Mater. 162 (2009) 1551–1556. [31] F. Raposo, M.A. De la Rubia, V. Fernández-Cegrí, R. Borja, Anaerobic digestion of solid organic substrates in batch mode: an overview relating to methane yields and experimental procedures, Renew. Sustain. Energy Rev. 16 (2012) 861–877. [32] F. Xu, Z.-W. Wang, L. Tang, Y. Li, A mass diffusion-based interpretation of the effect of total solids content on solid-state anaerobic digestion of cellulosic biomass, Bioresour. Technol. 167 (2014) 178–185.