Energy 27 (2002) 457–469 www.elsevier.com/locate/energy Kinetic studies of dehydration, pyrolysis and combustion of paper sludge Yong Ho Yu a, Sang Done Kim b,*, Jong Min Lee b, Keun Hoo Lee a a R&D Center, Samsung Engineering Co., Ltd., 39-3 Sungbok-Ri, Suji-Eup, Yongin, Kyunggi-Do, South Korea 449- 844 b Department of Chemical Engineering and Energy and Environment Research Center, Korea Advanced Institute of Science and Technology, TaeJon, South Korea 305-701 Received 22 May 2001 Abstract The reaction kinetics of drying, pyrolysis and combustion of paper sludge have been determined in a thermogravimetric analyzer (TGA). The effects of heating rate (5–30 K min�1) and sample weight (10– 50 mg) on drying and pyrolysis of paper sludge have been determined. The kinetic parameters of char combustion are determined at the isothermal conditions (723–1173 K). For dehydration, pyrolysis and combustion of paper sludge, temperature can be divided into drying (�470 K), pyrolysis [low (470–660 K), medium (660–855 K)] and combustion (�855 K) ranges. From the kinetic parameters (frequency factor and activation energy) of water decomposition, two major degradable compounds are found and the experimental thermogravimetric curves predicted by those parameters. For char combustion, the reaction order is found to be unity. The char combustion is well expressed by the shrinking core model with chemical reaction controlling and the activation energy is changed from 24.3 to 10.14 kJ mol�1 K�1 at 873 K. 2002 Elsevier Science Ltd. All rights reserved. 1. Introduction Sludge production from sewage treatment and pulp and paper industries has been increasing due to the growth of such facilities. Therefore, sludge disposal is one of the most important tasks for environmental protection. Disposal through land filling is expected to decrease due to the limits of existing filling capacities. Therefore, incineration of sludge in fluidized bed combustors [1–6] is expected to expand as one of the alternatives to land filling. Wet sludge which contains water over 60 wt% can be incinerated in fluidized beds with some auxiliary fuel [6]. Therefore, * Corresponding author. Tel.: +82-42-869-3913; fax: +82-42-869-3910. E-mail address:
[email protected] (S.D. Kim). 0360-5442/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S0 360- 544 2(01 )000 97- 4 458 Y.H. Yu et al. / Energy 27 (2002) 457–469 Nomenclature Ai frequency factor of decomposition of compound i (s�1) Ei activation energy of decomposition of compound i (kJ mol�1K�1) Wa degradable weight of sludge at a time t (mg) Wc initial weight of dry ash free char (mg) Wf weight of sludge residue (mg) Wi weight of degradable compound i, at a time t (mg) Wi,0 Initial weight of compound i based on ash free content (mg) Ws weight of sludge at a time t (mg) kc frequency factor of char combustion (s�1) Ki kinetic constant of decomposition of compound i (s�1) mN2 nitrogen mass flow rate (mg s�1) PO2 partial pressure of oxygen (bar) rc conversion rate of char (mg s�1) T temperature (K) X conversion of char (-) b constant heaing rate defined by dT/dt, (K s�1) Subscripts h high temperature l low temperature d drying i compound kinetic parameters such as frequency factor and activation energy for drying, pyrolysis and com- bustion of sludge provide important information for design of fluidized bed sludge incinerators. It is necessary that the drying rate of wet sludge is included in the kinetic equation when wet sludge that contains water over 60 wt% is directly used as fuel. Incineration of sludge is a consecu- tive process comprising of drying, pyrolysis of volatile matters, and the combustion of char. Therefore it should be noted that the kinetic parameters of each step are needed to understand the whole incineration process. However, kinetic studies of drying, pyrolysis and combustion of paper sludge are relatively rare due to its complicated reactions. Urban and Antal [7] analyzed the sewage sludge pyrolysis as the nth order reaction with the derived kinetic parameters. They analyzed the pyrolysis reaction as the parallel two equations, which represent decomposition reac- tions of settled undigested organics and dead bacteria, respectively. However, their model pro- duces some errors to fit the different heating rates at the weight fractions less than 0.8 because of variability in the sludge composition. On the other hand, Dumpelmann et al. [1] studied pyrol- ysis kinetics of sewage sludge by using TGA and described pyrolysis of sludge as a competitive reaction to volatile matters and the intermediate solid that subsequently decomposed into volatile 459Y.H. Yu et al. / Energy 27 (2002) 457–469 matters and char. However, it is known to be very difficult to use a simple global kinetic model for sludge pyrolysis reaction over the wide temperature range. Whereas, Kim et al. [8] and Kim and Chun [9] developed the analytical technique, namely the subtraction method that enables to estimate the kinetic parameters for pyrolysis reaction of scrap tire, with an assumption that there may exist undisturbed temperature regions of each group. Therefore, in the present study, the pyrolysis process can be characterized by high, medium and low temperature regions based on pyrolysis pattern of the sludge in differential thermal- gravimetry (DTG) and TGA with an assumption of the wet sludge comprised of water, two major degradable compounds and char [8]. Thus, the kinetic parameters (frequency factor, activation energy and reaction order) of dehydration, pyrolysis and combustion reactions of paper sludge at three different temperature regions are determined in TGA and DTA. Moreover, it is found that the shrinking core model can be successfully applied to analyze the char combustion reaction. 2. Experimental Thermogravimetric experiments were carried out in Setaram TG/DTA92 with a sensitivity of 10�6 g. Samples of the sludge (10–50 mg) were loaded in a Ni–Cr wire mesh basket suspended from an electronic balance that located inside a 0.055 m i.d×1.0 m high stainless steel reactor. The heating rate was controlled with an electric heater (4 kW) and temperature inside the reactor was monitored by a thermocouple located at 5 mm under the basket. To determine the effect of heating rate on the kinetics of drying and pyrolysis, heating rate was varied from 5 to 30 K min�1. The operating condition was maintained at a desired temperature for 2000 s to maintain an isothermal condition with variation of temperature from ambient to 1073 K. The flow rate of nitrogen as a carrier gas into the bottom of the reactor was ranged from 20 to 50 cm3 min�1. In the present study, the undigested wet sludge after treatment by a filter press was sampled at the paper plant in Korea, where the recycled waste paper was used for 30–40% of raw material. The proximate and ultimate analyses of the paper sludge used in this study are shown in Table 1. Char produced from the paper sludge (0.7 g) was loaded into the Ni–Cr basket to study the kinetics of char combustion. In order to produce char, the paper sludge was heated up to 1173 K with a heating rate of 30 K min�1 in nitrogen atmosphere at 1 bar and temperature was main- tained at 1173 K for 30 min. Thereafter, char was cooled to room temperature with a cooling Table 1 Results of paper sludge analysis Ultimate anlysis Proximate analysis Element wt% Item wt%(raw) C 39.82 Water 59.80 H 4.53 Volatile matter 22.19 O 19.22 Fixed carbon 3.70 N 0.70 Ash 14.31 S 0.11 Cl 0.02 HHV(wet), 3816 kJ/kg 460 Y.H. Yu et al. / Energy 27 (2002) 457–469 rate of 30 K min�1. Oxygen partial pressure was controlled in the range of 0.01–0.05 bar by varying flow rates of nitrogen and air. The thermogravimetric experiment for combustion of char was also carried out at the isothermal conditions from 723 to 1173 K. 3. Kinetic equations The kinetic equation for the decomposition of sludge is described as dWa dt � � i � 1 n dWi dt � �� i � 1 n KiWi (1) where Wa � Ws�Wf, degradable weight of sludge at a time t, Ws=weight of sludge measured at a time t, Wf=weight of residue, Wi=weight of degradable compound i at a time t. Whereas, the rate constant, K, can be described by the following Arrhenius equation [7,9] with an assumption that the activation energy (Ei) is independent of temperature. Ki � Ai exp�� EiRT� (2) where Ai, Ei, R and T are frequency factor, activation energy, gas constant and temperature, respectively. If the sludge is comprised of three major compounds and there is no interactive reaction among the compounds, Eq. (1) can be expressed as Eq. (3), being normalized as [8–10]. ln�� bWa dWadT � � ln�A1 exp��E1RT��W1Wa� � A2exp��E2RT��W2Wa� (3) � A3exp��E3RT��W3Wa�� where, b � dT / dt as a constant heating rate. To determine the kinetic parameters of tire pyrolysis, Kim et al. [8] proposed three temperature regions (high, medium and low) of pyrolysis in TGA and DTG experiments. Whereas, in the present study, pyrolysis of the paper sludge can be divided into four different regions in the TGA and DTG curves as shown in Figs. 1 and 2 including drying region with an assumption that three compounds including water can be volatilized. Therefore, the DTG curve can be expressed as the follow corresponding equations for each region as: At high temperature region (W2 � W3 � 0 and Wa � W1), Eq. (3) can be reduced to ln�� bW1 dW1dT � � ln A1��E1R �1T (4) At medium temperature region (W3 � 0), Eq. (3) can be reduced to If W1,0, W2,0, W3,0 are the initial weight of compounds 1, 2 and 3 in the sludge respectively, ln�� bWa dWadT � � ln�A1exp�E1RT��W1Wa� � A2exp�E2RT��W2Wa�� (5) 461Y.H. Yu et al. / Energy 27 (2002) 457–469 Fig. 1. Thermogravimetry analysis of paper sludge at a heating rate of 20 K min�1, W0 � 40 mg, FN2 � 30 cm3 min�1. Fig. 2. DTG data of paper sludge at heating rate of 20 K min�1, W0 � 40 mg, FN2 � 30 cm 3 min�1. Eq. (3) can be expressed at each temperature region as follows. At low temperature region (W1 � W1,0), the initial weight of compound 1, W3 � 0 and dW1 /dT � 0, the above equation can be expressed as Eq. (6). ln�� bW2 dW2dT � � ln A2��E2R �1T (6) 462 Y.H. Yu et al. / Energy 27 (2002) 457–469 At the drying region (W1 � W1,0, W2 � W2,0, dW1 /dT � 0, dW2 /dT � 0), Eq. (3) can be expressed as: ln�� bW3 dW3dT � � ln A3��E3r �1T (7) As can be seen in Figs. 2 and 3, it is not necessary to divide the intermediate region between low temperature and drying regions since the Arrhenius curve exhibits discontinuity between those two regions. The kinetic parameters for decomposition of compound 1 can be obtained from the following equation from integration of Eq. (4) as: Wa � W1 � W1,0exp��A1RT2hb E1 �1�2RThE1 � 4�RThE1 � 2�exp�� E1RTh�� (8) where, Th is the temperature at the high temperature region. To determine the initial weight of compound 2 as in the case of compound 1, the following equation can be derived. Wa�W1 � W2,0exp��A2RT2lb E2 �1�2RTlE2 � 4�RTlE2 � 2�exp�� E2RTl�� (9) where Tl is the temperature at the low temperature region. As can be seen in Fig. 2, rapid decrease of the sample weight between the drying and low Fig. 3. Graphical expression to estimate the stepwise kinetic parameters of drying and pyrolysis process of the paper sludge at heating rate of 20 K min�1, W0 � 40 mg, FN2 � 30 cm 3 min�1. 463Y.H. Yu et al. / Energy 27 (2002) 457–469 temperature regions can be observed. Therefore, it can be assumed that compounds 1 and 2 would not be decomposed during the drying step, keeping the weights of compound 1 and 2 as W1,0 and W2,0, respectively. Weight variation of compound 3 (water) along temperature in the drying step can be expressed as: Wa�W1,0�W2,0 � W3,0 exp��A3RT2db E3 �1�2RTdE3 � 4�RTdE3 � 2�exp�� E3RTd�� (10) where, Td is the temperature at the drying region. The frequency factor and activation energies for the decomposition reaction of compounds 1, 2 and 3 can be obtained from the slopes and intersection of TGA curves in the each region as defined by Eqs. (4), (6) and (7). Furthermore, weight fractions of compounds 1, 2 and 3 in the sludge can be estimated from TGA by using Eqs. (8)–(10). 4. Results and discussion 4.1. Drying and pyrolysis The rates of weight loss in TGA and DTG of the wet paper sludge are shown in Figs. 1 and 2, respectively. As can be seen in Fig. 2, it can be considered that both TGA and DTG curves may be divided into four regions, namely drying (�470 K), low (470–660 K), medium (660–855 K) and high temperature (855 K) regions [8]. Although, complicated decompositions may occur simultaneously in each region, it can be assumed that the wet sludge may be comprised of three compounds including water. For sewage sludge, Dumpelmann et al. [1] assumed that the pyrolysis process is represented as the competitive reaction between volatile product and the intermediate solid that decomposes into char and volatiles with further heating. Though similar concept has been applied in this study, the regions are classified by temperature based on the assumptions of Kim et al. [8]. As can be seen in Figs. 1 and 2, the predicted TGA and DTG curves by Eqs. (1)– (10) exhibits an excellent fit to the experimental data of the present study. To determine the frequency factor and activation energy, an Arrhenius plot is shown in Fig. 3. The Arrhenius plot illustrates four regions in which three straight lines exhibit with different slopes. Therefore, it can be considered that the analysis of drying and pyrolysis characteristics with the assumption of four regions is a reasonable one. The effect of heating rate on weight loss of the sample is shown in Fig. 4. The weight loss curves in the drying region shifted to higher temperatures with increasing heating rate but there is no change in the other regions. Vaporization of water may be affected by heat transfer rate at low temperature because latent heat of water is larger than other volatile product. The frequency factor and activation energy derived from the Arrhenius plot in each region are shown in Table 2. The frequency factor related to the compound 1 increases with increasing heating rate and nitrogen flow rate, but decreases with an increase in the sample weight due to the thermal lag as described by Gronli et al. [17] who reported that the mass transfer resistance may shift the DTG curve at higher heating rates and the activation energy and frequency factor decrease due to the thermal lag. Throughout the present experiment, the mass transfer may affect 464 Y.H. Yu et al. / Energy 27 (2002) 457–469 Fig. 4. Weight loss of paper sludge with temperature at different heating rate, Wa0 � 40 mg, FN2 � 30 cm 3 min�1. the frequency factor because of the sample weight employed. Therefore, the intrinsic practical frequency factor can be expressed in terms of (mN2/W0). However, the effect of thermal lag on decomposition of compounds 2 and 3 is relatively small since the sample weight and the heating rate do not much affect the frequency factor and activation energy. By increasing heating rate and nitrogen flow rate, decomposition of compound 1 increases due to increase of both driving force of mass and heat transfers inside the particle. On the other hand, the frequency factors of A2 and A3 are less affected by the heating rate and sample weight. Since there is no significant effects of heating rate, sample weight and gas flow rate on the decomposition kinetics of compounds 2 and 3, the average values of frequency factors, activation energies and weight fractions of compounds 2 and 3 in the same region can be utilized. Decomposition of compound 1 produces the highest activation energy and frequency factor compared with those of compounds 2 and 3 and the activation energy of compound 2 is larger than compound 3 as observed with the sewage sludge used by Dumpelmann et al. [1]. They observed that frequency factor and activation energy of the final step, decomposition of the intermediate into volatile product and char, exhibits the higher values than the former decomposition of sludge. On the other hand, the water content estimated from Eq. (3) is in good agreement with the value in the proximate analysis as shown in Table 1. Throughout the graphical analysis of TGA, DTG and the Arrhenius plot, the frequency factors and activation energies of decomposition of each compound are estimated to be A1 � 8.35 × (mN2/W0)11.23s�1; (correlation coefficient � 0.84) E1 � 225.7 kJ mol K�1 A2 � 6.60 × 103s�1 E2 � 53.05 kJ mol K�1 A3 � 1.64 × 104s�1 E3 � 35.63 kJ mol K�1 465Y.H. Yu et al. / Energy 27 (2002) 457–469 Ta bl e 2 Fr eq ue nc y fa ct or s, ac tiv at io n en er gi es an d w ei gh tf ra ct io ns o f co m po sit io na lc o m po un ds o f pa pe r slu dg e F N 2a (m N 2)b dT /d tc W 0d ln A 1 E 1 W 1, 0/W a ln A 2 E 2 W 2, 0/W a ln A 3 E 3 W 3, 0/W a 30 (0. 57 45 ) 0. 16 7 40 23 .4 5 20 0. 32 0. 03 01 7. 54 50 .4 1 0. 16 7 9. 50 35 .7 0 0. 78 3 30 (0. 57 45 ) 0. 25 40 22 .1 0 18 9. 47 0. 02 55 8. 45 53 .5 5 0. 16 0 9. 45 35 .8 9 0. 80 1 30 (0. 57 45 ) 0. 33 4 40 27 .5 9 23 1. 63 0. 02 67 9. 15 56 .1 1 0. 15 8 8. 90 34 .5 6 0. 81 5 30 (0. 57 45 ) 0. 50 40 21 .9 9 18 7. 26 0. 02 76 8. 11 50 .1 0 0. 15 4 7. 76 31 .6 2 0. 81 3 30 (0. 57 45 ) 0. 33 4 10 37 .5 6 29 4. 94 0. 01 49 9. 09 53 .5 0 0. 21 7 12 .1 7 40 .2 9 0. 76 8 30 (0. 57 45 ) 0. 33 4 20 27 .3 7 22 0. 46 0. 01 85 10 .5 0 59 .9 8 0. 23 2 10 .4 3 36 .8 3 0. 73 7 30 (0. 57 45 ) 0. 33 4 30 26 .2 5 21 7. 52 0. 04 53 7. 66 46 .9 1 0. 23 4 9. 93 36 .0 5 0. 71 0 30 (0. 57 45 ) 0. 33 4 40 21 .4 1 18 0. 97 0. 03 67 8. 27 49 .4 3 0. 19 6 8. 48 32 .6 5 0. 75 9 30 (0. 57 45 ) 0. 33 4 50 21 .8 0 18 3. 35 0. 02 84 8. 53 51 .2 6 0. 20 4 8. 48 33 .4 9 0. 76 1 10 (0. 19 15 ) 0. 33 4 20 24 .4 7 19 8. 85 0. 01 62 9. 92 57 .5 2 0. 19 6 10 .3 8 37 .0 7 0. 79 8 20 (0. 38 3) 0. 33 4 20 33 .5 2 26 8. 34 0. 02 11 9. 01 53 .2 2 0. 20 1 10 .7 3 37 .2 4 0. 76 4 40 (0. 76 6) 0. 33 4 20 41 .9 0 33 5. 00 0. 02 24 9. 30 54 .6 4 0. 19 5 10 .2 3 36 .2 0 0. 76 9 A ve ra ge 27 .4 5 22 5. 68 0. 02 61 8. 79 53 .0 5 0. 19 3 9. 70 35 .6 3 0. 77 3 A ve ra ge 18 .7 16 .8 24 .3 8. 0 5. 4 11 .5 9. 7 4. 8 3. 1 de vi at io n (% ) a N itr og en v o lu m et ric flo w ra te (cm 3 m in � 1 ). b N itr og en m as s flo w ra te , m N 2 (m g s� 1 ). c H ea tin g ra te (K s� 1 ). d Sa m pl e am o u n t (m g). 466 Y.H. Yu et al. / Energy 27 (2002) 457–469 4.2. Char combustion According to previous studies [10–16], it is known that the shrinking core model describes well the reaction rate of char gasification or combustion. The reaction rate by the shrinking core model can be expressed as: dX dt � KcP n O2(1�X)2/3 � kcexp��EcRT�PnO2(1�X)2/3 (11) where dX/dt is the char conversion rate; PO2 is oxygen partial pressure (bar); Kc is the kinetic constant for char combustion. With integration of Eq. (11), the following relationship between time and conversion factor, 1�(1�X)1/3, can be derived. t t � 1�(1�X)1/3 (12) where t is the time for complete conversion. The relationship between the conversion factor 1�(1�X)1/3 and time is shown in Fig. 5. As can be seen, the shrinking core model with chemical reaction controlling predicts well the combus- tion rate of char from the paper sludge. On the other hand, the apparent reaction order (n), can be determined from logarithmic plot of the reactivity and oxygen partial pressure (Fig. 6). The reactivity is defined as the following equation and the reaction rate (dW/dt) is estimated as an average value of weight loss up to 10 wt.% [11]. Fig. 5. The conversion factor with time based on the shrinking core model with variation of reaction temperature, PO2 � 0.03 bar. 467Y.H. Yu et al. / Energy 27 (2002) 457–469 Fig. 6. The apparent reaction order with respect to oxgen partial pressure at 873 K. rc � � 1 Wc dW dt (13) As can be seen in Fig. 6, the apparent reaction order [Eq. (11)] to oxygen partial pressure in char combustion process at 873 K is found to be unity [12–15]. The Arrhenius plot based on the shrinking core model is shown in Fig. 7 where two straight lines having different slopes can be found. Even though the activation energy decreases in the high temperature region (�873 K), the reaction is still belongs to the chemical reaction rate control regime, since there is linear relationship between 1�(1�X)1/3 and time. Therefore, the shrinking core model with chemical reaction resistance can describe well the combustion mech- anism of the char. Therefore, the frequency factor and activation energy from Eq. (11) for each region with the graphical analysis of char combustion are found to be kc � 1.603 bar�1 s�1, Ec � 24.3 kJ mol�1 K�1 at T � 873 K and kc � 0.212 bar�1 s�1, Ec � 10.14 kJ mol�1 K�1 at 873 K �T for high temperature region. 5. Conclusions The pyrolysis process of the paper sludge can be divided into drying (�470 K), low (470–660 K), medium (660–855 K) and high (�855 K) temperature regions based on the data of TGA, DTG and Arrhenius plot. The kinetic parameters for decomposition of each compound are determ- ined and the kinetic equations corresponding to each compound predict well weight loss of the sample in thermogravimetric analyzer. For char combustion reaction, the reaction order respect to the oxygen partial pressure is found to be unity. The char combustion can be well expressed 468 Y.H. Yu et al. / Energy 27 (2002) 457–469 Fig. 7. Arrhenius plot for combustion of paper sludge based on the shrinking core model at PO2 � 0.03 bar. by the shrinking core model with chemical reaction controlling and the activation energy is changed from 24.3 to 10.14 kJ mol�1 K�1 at 873 K. Acknowledgements Authors would like to acknowledge a grant-in-aid of research from the Korea Science and Technology Policy Institute. References [1] Dumpelmann R, Richarz W, Stammbach MR. Kinetic studies of the pyrolysis of sewage sludge by TGA and comparison with fluidized beds. Can J Chem Eng 1991;69:953–63. [2] Lee JK, Lee KH, Jang JG, Lim JH, Lim JS, Chun HS. Combustion characteristics of digested sewage sludge in a two-stage swirl-flow fluidized bed incinerator. Hwahak Konghak 1992;30:499–508. [3] Lee JK, Lee KH, Jang JG, Lee NS, Lim JS, Chun HS. Fluidized bed incineration of digested sewage sludge, Part 1. Combustion characteristics of the single-stage fluidized bed incinerator. Hwahak Konghak 1992;30:228–36. [4] Ogada T, Werther J. Combustion characteristics of wet sludge in a fluidized bed: release and combustion of the volatiles. Fuel 1996;75:617–26. [5] Shin YS, Kim HC, Chun HS. Drying of water treatment process sludge in a fluidized bed dryer. Korean J Chem Eng 2000;17:22–6. [6] Douglas MA, Latva-Somppi J, Razbin VV, Friedrich FD, Tran HN. Combustion of paper deinking solids in a pilot-scale fluidized bed. Tappi J 1994;77:109–19. 469Y.H. Yu et al. / Energy 27 (2002) 457–469 [7] Urban DL, Antal Jr MJ. Study of the kinetics of sewage sludge pyrolysis using DSC and TGA. Fuel 1982;61:799–806. [8] Kim S, Park JK, Chun HD. Pyrolysis kinetics of scrap tire rubbers I: using DTG and TGA. J Envir Eng ASCF 1995;121:507–14. [9] Kim S, Chun HD. Analytical techniques estimating kinetic parameters for pyrolysis reaction of scrap tire rubbers. Korean J Chem Eng 1995;12:448–53. [10] Flynn JH. The effect of heating rate upon the coupling of complex reactions. I. Independent and competitive reactions. Thermochim Acta 1980;37:225–38. [11] Kwon TW, Kim SD, Fung DPC. Reaction kinetics of char-CO2 gasification. Fuel 1988;67:530–5. [12] Lee JM, Kim YJ, Lee WJ, Kim SD. Coal-gasification kinetics derived from pyrolysis in a fluidized-bed reactor. Energy 1998;23:475–88. [13] Park KY, Park YC, Son JI, Park WH. An experimental study on intrinsic combustion kinetics of Korean anthracite by thermogravimetric analysis. Hwahak Konghak 1987;25:345–54. [14] Dutta S, Wen CY. Reactivity of coal and char. 2. In oxygen–nitrogen atmosphere. Ind Eng Chem Pro Des Dev 1977;16:31–7. [15] Smith III FG, Pendarvis RW. Combustion of cellulosic char under laminar flow conditions. Combustion and Flame 1992;88:61–73. [16] Sorensen LH, Saastamoinen J, Hustad JE. Evaluation of char reactivity data by different shrinking-core models. Fuel 1996;75:1294–300. [17] Gronli M, Antal MJ, Varhegyi G. A round-robin study of cellulose pyrolysis kinetics by thermogravimetry. Ind Eng Chem Res 1999;38:2238–44. Kinetic studies of dehydration, pyrolysis and combustion of paper sludge Introduction Experimental Kinetic equations Results and discussion Drying and pyrolysis Char combustion Conclusions Acknowledgements References