Comparative Study on the Modeling of Char-CO 2 Reaction Kinetics under Pressurized Conditions

Transcription

1 Topic: Gasification / Co-Gasification - # Comparative Study on the Modeling of Char-CO Reaction Kinetics under Pressurized Conditions Martyna Tomaszewicz, Grzegorz Tomaszewicz, Marek Sciazko, Tomasz Chmielniak IChPW, Institute for Chemical Processing of Coal, Zamkowa St., Zabrze, 483, Poland mtomaszewicz@ichpw.zabrze.pl Abstract Comprehensive kinetic studies were performed to establish the validity of numerous kinetic expressions that describe the behavior of char samples during gasification toward CO. The gasification experiments were carried out isothermally in a pressurized thermogravimetric analyzer (TG-HP5s, TA Instruments, USA). Two char samples derived from Turów lignite B and Janina sub-bituminous C were used for the studies. The models that were employed are regarded as the most popular: the volumetric (VM); the modified volumetric (MVM); the grain (GM); the random pore (RPM); the unification theory (UTM); Johnson s, Dutta and Wen s; and the Langmuir-Hinshelwood model. The models were validated at 85, 95 and,5 C, and at CO pressures of.4,, and bar. The effects of temperature and pressure were significant and increased the reaction rate. The calculated values of activation energy were similar for the various models employed and ranged from 65 to kj/mol for the Turów sample and 86 to 38 kj/mol for the Janina coal char sample. The UTM, GM, RPM, and MVM fit the experimental data well. The LH model was also satisfactory. Overall, the UTM was distinguished as the most applicable over a wide range of process conditions.

2 Topic: Gasification / Co-Gasification - # Introduction The gasification of coal is considered to be an effective and clean method to produce gas that can be used to generate power and heat and/or be used as a chemical feedstock. Two major steps can be distinguished during the gasification process: the initial rapid pyrolysis of coal to produce char, gases and tars and the subsequent gasification of the nascent char. The rate of the second step is relatively low due to the poor reactivity of char. This rate significantly affects the gasification efficiency. To improve this efficiency, a novel gasification method has been developed at the Institute for Chemical Processing of Coal (IChPW) in which additional carbon dioxide is added to the process. As indicated in the thermodynamic model, the CO plays a dual role in the process as both a carbon and oxygen carrier []. The beneficial effect of the CO addition is associated with a decrease in the demand for both carbon and oxygen while maintaining the syngas quality as in conventional air-blown systems by increasing the influence of the char-co reaction []. The process is designed to operate in a 6-kW PDU-scale pressurized circulating fluidized bed reactor rig, which is part of the Clean Coal Technologies Center of the IChPW. A thorough understanding of the char reaction kinetics is essential to optimize the process conditions, especially those of the char-co reaction, which is one of the slowest independent gasification reactions, and simultaneously plays a critical role in the proposed process conception. Knowledge of char reactivity and understanding the char-co reactions kinetics is fundamental to both the design of gasification reactors and the optimization of process conditions. The reactivity of coal chars is influenced by many factors, such as the conditions during devolatilization, parent coal rank, pore structure, and mineral matter content and its composition [-3]. Additionally, the pressure affects the char gasification rate, especially for chars derived from lower rank coals [3-5]. Numerous models can sufficiently describe the CO -gasification rate variation, irrespective of whether the models consider the structural changes in the coal char as the reaction proceeds. The models regarded as the most popular include the volumetric; the modified volumetric; the grain (shrinking core); the random pore; the unification theory; Johnson s, Dutta and Wen s; and the Langmuir-Hinshelwood models. The present work primarily aimed to characterize the kinetics of the pressurized CO -gasification of chars prepared from two different rank coals from Polish mines at different temperatures (85-,5 C) and CO pressures (.4,, and bar). The second objective was to compare the results of fitting the aforementioned kinetic models to the experimental data. Experimental Two coals, sub-bituminous C from the Janina coalmine and the lignite B from the Turów open-pit mine, were selected for this study. A thorough characterization of the studied coals in terms of proximate and ultimate analyses is given in Table with the ash and maceral composition. Chars were prepared by devolatilizating the parent coals in a fixed-bed reactor heated by an electric furnace under inert gas flow (nitrogen). The samples were pyrolyzed at a heating rate of 5 K/min to a final temperature of C; the samples were maintained at this temperature for 3 min. After cooling the char samples to room temperature under flowing nitrogen, the samples were ground and sieved, and particles less than µm were used for thermogravimetric studies.

3 Topic: Gasification / Co-Gasification - # Table. Characterization of parent coals by the means of proximate and ultimate analyses, chemical composition of ash, maceral composition and vitrinite reflectance Proximate analysis Ultimate analysis Chemical compositi on of ash Maceral compositi on Reflectan ce Parameter Turów LigB Janina SubC M ar t, wt. % M ad, wt. % 3.9 A d, wt. % 8.. V daf, wt. % C ad T, wt. % H ad T, wt. % N ad, wt. %.6.94 S ad T, wt. %.. S ad A, wt. %.8.5 S ad C, wt. %.74.7 O ad diff, wt. % SiO, wt. % Al O 3, wt. % Fe O 3, wt. % CaO, wt. % MgO, wt. % TiO, wt. %.86. Na O, wt. % K O, wt. % NDA.46 V, vol. % L, vol. % 7 I, vol. % 6 8 R, %.3.44 To characterize the char pore structure, a 3Flex, produced by Micromeritics, USA, was employed to determine the pore volume and surface area by means of N and CO adsorption isotherms at 77 and 73 K, respectively. Prior to the measurements, the samples were degassed under a vacuum at 35 C for 3 h. The surface area was calculated by the BET method. The Dubinin-Astakhov theory (D-A) was employed to calculate the volume and surface area of micropores inaccessible for nitrogen adsorption at 77 K. The results of the pore characteristics are summarized in Table. Table. Pore characteristics of the coal chars used for experiments Parameter Turów char Janina char Nitrogen adsorption at 77 K: S BET, m /g V T, cm 3 /g V mi, cm 3 /g.9. V me, cm 3 /g Carbon dioxide adsorption at 73 K: S D-A, m /g V D-A, cm 3 /g.55.3 The kinetic measurements were conducted in a TG-HP5s high-pressure thermogravimetric analyzer (TA Instruments, USA), which is shown in Fig.. The gasification experiments were performed 3

4 Topic: Gasification / Co-Gasification - # isothermally between 85 and,5 C at CO pressures of, and bar. Moreover, experiments were also conducted at a CO partial pressure of.4 bar using a mixture of 4% CO in helium at a total pressure of bar. A weighed (5 mg) char sample was placed inside the alumina crucible. The flow of the gasifying agent (CO or CO /He mixture) was set to ml/min for all experiments. The carbon conversion degree was determined by means of Eq.: w w X () w w ash where w, w and w ash are the initial weight of char, the instantaneous weight of char and the weight of ash remaining after complete carbon conversion was reached, respectively. Figure. The TA Instruments TG-HP5s high-pressure thermogravimetric analyzer used for studies on gasification kinetics Kinetic models Numerous models describe changes in the gasification rate of carbonaceous materials, which are derived by modifying a global kinetic expression that is defined as follows: dx kt, PCO f X dt () where dx/dt is the reaction rate, k is the apparent gasification reaction rate that considers the effect of temperature (T) and gasifying agent pressure (P CO ), and f(x) denotes a term expressing the physical changes of the solid occurring as the reaction proceeds. The rate constant depends on the temperature by means of an Arrhenius-type relationship with the reaction order (m), which represents the effect of pressure of CO. This expression is given by Eq.3: m E k APCO exp (3) RT where A is the pre-exponential factor, E is the activation energy, R is the universal gas constant, and T is the reaction temperature. Several models are employed in the present study to describe the changes in the gasification rate. These models generally differ with respect to their formulation of f(x). 4

5 Topic: Gasification / Co-Gasification - # The Volumetric Model (known also as the I-order kinetic model) assumes a homogeneous reaction throughout the whole particle and a linearly decreasing reaction surface area with conversion [6]. The overall reaction rate is expressed by: dx kvm X dt (4) Separating the variables yields a linearized expression that can be used to estimate the kinetic constant: ln X k VM t (5) The Grain Model or Shrinking Core Model, introduced by Székely and Evans [7], assumes that a porous particle consists of an assembly of uniform, individual nonporous grains and that the reaction takes place on the surface of these grains. The porous network is formed by interparticle voids between these grains. The shrinking core behavior applies to each of these grains during the reaction. Assuming kinetic control and that the grains are spherical, the overall reaction rate in these model is expressed as follows: dx kgm X 3 (6) dt This model predicts a monotonically decreasing reaction rate and surface area because the surface area of each grain is receding during the reaction. The linearized expression is given by Eq. 7: X 3 3 k GM t (7) The Random Pore Model developed by Bhatia and Perlmutter [8] considers the overlapping of pore surfaces, which results in significant variability of the area available for reaction as the degree of conversion increases [8]. The basic equation for this model is: dx krpm X ln X (8) dt This model can predict a maximum in the reactivity as the reaction proceeds because it considers the competing effects of pore growth during the initial stages of gasification and the destruction of the pores due to the coalescence of neighboring pores for higher conversions of carbon. The RPM model defines a structural parameter, ψ, which is directly related to the pore structure of the initial sample by means of Eq. 9: 4L (9) S where S, L and ε represent the pore surface area, pore length, and solid porosity, respectively. Another method to calculate ψ is based on experimental data and uses the experimental conversion value at which the reaction rate is maximized, X max. By differentiating Eq.8, ψ can be estimated in terms of X max as follows: () ln X max The following expression is used to calculate the reaction rate constant: ln X krpm X () The Modified Volumetric Model (MVM) was first introduced by Kasaoka et al. [9] as an empirical modification of the VM. However, the MVM assumes that the rate constant is changing with conversion of the solid (X) as the reaction proceeds [9]. The reaction rate can be expressed by means of Eq. and the change in conversion degree with Eq. 3: 5

6 Topic: Gasification / Co-Gasification - # dx M VM dt X at X X k () b ln (3) where k MVM (X) is the model-corresponding kinetic constant, and a and b are the empirical parameters. The constant b is considered related to the physical structure of chars, while the constant a is regarded to be more closely related to the intrinsic reactivity of chars [9]. The rate constant k MVM can be calculated by using Eq. 4: k M V M b X a b ln X b b (4) The above expression can be integrated to obtain a mean value of the rate constant, which can be useful to obtain the kinetic parameters by means of Eq.3 and comparing other constants derived from the analysis of kinetic expressions. Eq.5 gives the formulation for the integration: k MVM k X dx (5) MVM The Dutta and Wen model introduces a new parameter f into the rate expression []. Notable differences in the reactivity behavior are observed when the change in the relative pore surface with conversion is considered. The term f was introduced to express this phenomenon, which is defined as the ratio between the available pore surface area per unit weight at any stage of conversion and the initial pore surface area per unit weight: SX f (6) S The value of f varies with conversion and temperature. When neglecting the effect of temperature, the change in f with increasing conversion can be fitted to a function valid for X<.9: f X exp X (7) The ν and β parameters are characteristic of a given sample. The value of ν denotes the conversion at which the relative available surface area is maximized. Eq.8 therefore gives the rate of the reaction: dx fk DW X (8) dt which is the same equation as that for the VM (4), except the new factor f has been introduced. An dx Arrhenius-type dependence is also assumed for the constant k D-W. Originally, the values of dt X for a fractional conversion degree of. are plotted against the reciprocal of absolute temperature to derive the kinetic parameters (A D-W and E) from the Arrhenius relationship. Therefore, the values of A D-W and E are calculated from this plot. The values of the pre-exponential factor A D-W are thus considered to represent the intrinsic reactivity of samples. Dutta and Wen`s paper also stated that the value of the activation energy was found to be approximately 55 kj/mol for all studied coal and char samples []. The Unification Theory Model is often used to describe the rate variations during the gasification processes [-3]. This model is based on an approach that approximately reduces the char gasification curves (X vs. time) for different temperatures, pressures, gasifying agents and chars to a single curve when X is plotted against the dimensionless time τ, where τ = t/t /, and the t / is the time at which.5 conversion has been reached. The t / time is often called the half-life of the reaction. The unification approach is often used for different models, especially the GM and the RPM. In the present study, it will be used for the GM only. In this case, integrating Eq.6 and setting X=.5 and t = t /, along with defining the dimensionless time, derives the expression for the unification curve: 6

7 Topic: Gasification / Co-Gasification - # 3 3 X.5 (9) Due to the unification approach, the normalized reaction rate, dx/dτ, is only a function of the conversion degree. However, after averaging the reaction rate over the entire range of conversion degrees, this dependence is also eliminated. Hence, if a quantity R u is defined as follows: dx R u dx d dx d dx dx dx d where R u is a constant for the unification curve. Correspondingly, the average reactivity for a particular gasification experiment is defined as follows: dx dx RC dx dt () dt The value of RC is constant for each gasification experiment. Furthermore, the following expression is obtained from Eqs. and and from the definition of dimensionless time: t /RC R u () The relationship in Eq. states that when the gasification profile (X vs. t) data from different experiments are unified with the half-life, the average reactivity, R C, for each experiment is inversely proportional to the half-life, t /, of that gasification run with R u as the proportionality constant. Moreover, R u is unique for each gasification run. Raghunathan and Yang [] found that using Eq. for data from experiments performed at different laboratories, coals, temperatures and gasifying agents yielded values of R u that ranged from.366 to.393. They assumed that R u.38; hence:.38 RC (3) t/ Eq.3 can determine the kinetic constant R C from only the half-life of each experiment. The average reactivity constant is a function of temperature by means of an Arrhenius relationship: E RC U exp (4) RT where U is a constant. After combining Eq.4 with Eq., the Arrhenius relationship can be expressed as follows: Ru E t / exp (5) U RT Because R u is constant and equals.38, a plot of ln t / versus /T yields a straight line with a slope of E/R, and the intercept denoted here as S, which equals R u /U. The mechanism by which carbon reacts with CO is not yet fully understood. The reaction is widely accepted to proceed by the sorption of the reactant, the reaction on solid surface and the desorption of the product. The simplest mechanism of the reaction is given by two steps, which are expressed by Eqs.6-7: C C f k O CO C CO (6) O k k COC (7) f () 7

8 Topic: Gasification / Co-Gasification - # where C f represents an active site and C(O) an occupied site. Reaction (6) proceeds via CO adsorption. When significant amounts of CO are present during the reaction, carbon monoxide acts as an inhibitor by reducing the steady-state concentration of the C(O) complexes in an inverse reaction. Reaction (7) describes the CO desorption where CO is removed from the occupied site on the carbon surface. The desorption step (Eq.6) is considered much slower than the sorption step, thus it limits the global reaction rate. Ergun [4] proposed a rate equation for carbon-co gasification by using a Langmuir Hinshelwood rate expression, given by Eq.8, which involves the adsorption and desorption steps together with the effect of reactant pressures. kpco r (8) k k PCO PCO k k where k, k - and k are the kinetic constants, which depend on the temperature according to the Arrhenius law, and r represents the apparent reactivity. In the absence of CO, Eq.8 simplifies to the following form: kpco r (9) k PCO k The reaction rate profiles are different for different carbonaceous material, i.e., they increase or decrease as the reaction proceeds and can also exhibit a maximum or minimum. Thus, the apparent reactivity should be chosen while considering this behavior. Normally, reactivity refers to a specific value of the conversion degree, such as.5 [5] or.5 [6]. However, using the representative value of the reactivity calculated by averaging the two conversion degrees is also acceptable [5]. The most commonly used equation to calculate the reactivity is defined as follows: dx r dt (3) X Johnson [7-8] originally studied the kinetics of coal char gasification in atmospheres containing hydrogen and steam. He proposed a modification of the GM (Eq.6) by introducing the relative reactivity factor and adding a term that represents the effect of surface area, which diminishes as the reaction occurs. The Johnson model is given as follows: dx f 3 L k T X exp X (3) dt where f L is the relative reactivity factor that depends on the particular coal char and the pretreatment conditions used during pyrolysis, k T is the kinetic constant defined as a sum of kinetic constants of individual reactions and is influenced by partial pressures of reacting gases, and α is the kinetic parameter defined as a function of pressure and gas composition. The relative reactivity factor can be calculated by means of Eq.3: f L f exp 8467 (3) Tp T where f represents the reactivity index that depends only on the inherent nature of the coal char, T p is the pretreatment temperature and T is the gasification temperature (both temperatures are given in R). The equation is valid only when the T p T ; if T p T then f L = f. After integration of Eq.3 we obtain: 8

9 Topic: Gasification / Co-Gasification - # X ex p X X M dx flktt (33) X 3 The plot of M(X) versus time should yield a straight line with a slope that equals f L k T. The values of α should be predefined prior to analysis. Generally, α is assumed to equal when purge gases are used, but dx against X. The slope of this linearized relationship yields value of α. For pure gases, the relationship given by Eq.33 is the same as for GM (Eq.7), except for the term f L k T, which is equal to k GM. this variable can also be estimated by plotting the values of ln dt X Results and discussion As mentioned before, the rate of coal char gasification towards CO can by influenced by many process variables. The effect of temperature on carbon conversion for all studied temperatures and pressures is shown in Fig. a, c, e, g and b, d, f, h for the Janina and Turów char, respectively. The carbon conversion curves show that the reaction rate increases linearly during the initial stages of conversion; the reaction slows and after the conversion reaches approximately.9 until the conversion is complete. The temperature significantly increases the gasification rate, and this trend is observed for both the Janina and Turów chars. Therefore, the time to complete the conversion is shortened by a factor of several dozens when comparing the processes at 85 and,5 C. The pressure also noticeably affects the gasification rate. A small increase in gasification reactivity was observed when increasing the pressure from.4 bar to bar. Further increases in the CO partial pressure resulted in significant increases of the reaction rate, especially at a pressure of bar. The effect of the CO partial pressure observed at.4 bar was likely obscured by the enhancing effect of the balance gas in the mixture (helium). As stated by Walker, Jr. et al. [9] and Zhang and Calo [], the effect of the inert diluent gas on the measured kinetics cannot be neglected because the gas affects the active sites on the carbon matrix by changing their rate of rehybridization, thus changing the reactive lifetime of the active site. The parent coal rank also noticeably affected the reactivity. When comparing the experimental data for same temperature and pressure conditions, char from less metamorphized parent coal (Turów) showed an approximately two times higher reactivity than a higher rank parent coal (Janina). Notably, the content of samples and the composition of mineral matter also differ among the coal samples. To determine the best model to analyze the Arrhenius equation parameters, all aforementioned models were employed to describe the experimental data from this study. First, the VM, GM, RPM and MVM were compared together in one figure due to their simplicity. The results of the data prediction are illustrated in Fig. 3. The experimental data used for the kinetic analyses covered conversion degrees ranging from to.99. The determination coefficient R was used to validate models and indicate the quality of the model fit. The VM yielded the poorest fit with the experimental data because the model failed to correctly predict higher conversion values, where the reaction rate is notably slower. The GM yielded a significantly improved fit, which indicates that reaction mechanism is connected with the structural effects that occur while reaction proceeds. Eq. was used to calculate the structural parameter, ψ, in the RPM. This equation yielded ψ values between. and.6 for most of the Turów coal chars, which indicates that the reaction rate was maximized at similar conversion values. Furthermore, it directly indicates that the pore surface changes due to pore creation and coalescence of neighboring pores is maximized at similar conversion degrees. The same range of ψ values was determined for Janina coal char gasified at CO pressures of bar. The values of the structural parameter were notably higher (up to 6.3) for other temperature and pressure conditions employed to gasify Janina coal char, but in most cases this parameter yielded values between and 5. The structrural parameter ψ did not noticeably depend on the reaction temperature or pressure in either sample. The accurate fit of the MVM can also be explained in terms of the change in the pore surface. Compared to other models the MVM better predicts high conversion degrees, at which the rate is relatively small due to the collapse of the pore surface. It also describes the initial stage of gasification reaction very well. The RPM together with the MVM yield 3 9

10 Topic: Gasification / Co-Gasification - # similarly satisfactory fits, which agrees with previous studies where different thermobalance systems were employed to study the CO -gasification kinetics []. Both the RPM and MVM are based on empirical outcomes but are still simple to apply to experimental data. Because the VM, GM and RPM equations are linear, the slope of the obtained lines equals the kinetic coefficients. The parameters a and b of the MVM were calculated by means of a non-linear curve fit with the least-squares method (Eq. 3). The average value of the kinetic constant was then calculated by employing Eq. 5. The obtained parameters and kinetic constants are summarized in Table 3 and 4 for the Janina and Turów coal char, respectively. Both tables show the data obtained at.4,, and bar and 85, 95 and,5 C.

11 Topic: Gasification / Co-Gasification - # a) Janina coal char b) Turów coal char bar X, X, c) d).8.8 bar X, X, e) f) bar X, X, g) h).8.8 bar X, X, Figure. Changes in the carbon conversion degree over time for different temperatures of the gasification of Janina (left-hand side) and Turow (right-hand side) coal chars, observed at CO pressures of.4 (a, b), (c, d), (e, f), and bar (g, h).

12 Topic: Gasification / Co-Gasification - #.4 bar -ln(-x) y =.893x R² =.93 Janina coal char Turów coal char y =.358x R² =.94 VM fitting y =.69x R² = ln(-x) y =.3599x R² =.865 y =.85x R² =.98 VM fitting y =.3x R² = y =.x R² =.9947 GM fitting y =.76x R² =.976 GM fitting bar 3(-(-X) /3 ) y =.35x R² =.9948 y =.46x R² =.984 3(-(-X) /3 ) y =.543x R² =.9977 y =.89x R² = bar (/Ψ)((-Ψln(-X)) / -) y =.78x R² =.9949 y =.44x R² =.9955 RPM fitting y =.8x R² = (/Ψ)((-Ψln(-X)) / -) y =.73x R² =.998 y =.46x R² =.9947 RPM fitting y =.74x R² = y =.36x.68 R =.99 MVM fitting 5 4 y =.438x. R =.9964 MVM fitting bar -ln(-x) 3 y =.x.66 R =.9894 y =.65x.79 R = ln(-x) 3 y =.955x.63 R =.998 y =.7x. R = Figure 3. Plots of the VM, GM, RPM and MVM for the Janina (left-hand side) and Turów (right-hand side) coal char gasified at temperatures of 85, 95 and,5 C and CO pressures of.4,, and bar.

13 Topic: Gasification / Co-Gasification - # Fractional conversion degree, Dimensionless time, τ, - Model 85 C.4 bar 95 C.4 bar 5 C.4 bar 85 C bar 95 C bar 5 C bar 85 C bar 95 C bar 5 C bar 85 C bar 95 C bar 5 C bar Figure 4. Unification curves for X vs. time relationships for all Janina coal char gasification experiments performed at 85, 95 and,5 C and.4,, and bar. The master curve of the unified GM is given by the thick dotted line Fractional conversion degree, Model 85 C.4 bar 95 C.4 bar 5 C.4 bar 85 C bar 95 C bar 5 C bar 85 C bar 95 C bar 5 C bar 85 C bar 95 C bar 5 C bar Dimensionless time, τ, - Figure 5. Unification curves for X vs. time relationships for all Turów coal char gasification experiments performed at 85, 95 and,5 C and.4,, and bar. The master curve of the unified GM is given by the thick dotted line 3

14 Topic: Gasification / Co-Gasification - # Table 5. Values of the half-life of reaction and the R c and R u parameters of the unification model for Janina coal char gasified at temperatures of 85, 95 and,5 C and pressures of.4,, and bar Pressure, bar R Temperature, Half-life of u, calculated R C C reaction, min., /min acc. to Eq., , , , , Table 6. Values of the half-life of reaction and the R c and R u parameters of the unification model for Turów coal char gasified at temperatures of 85, 95 and,5 C and pressures of.4,, and bar Pressure, bar Temperature, Half-life of R R u, calculated C C reaction, min., /min acc. to Eq., , , , , The Langmuir-Hinshelwood expression is often used to predict the gasification rate observed at various pressures. It presents a significant advantage because it is applicable over a wider range of conditions than other models, e.g., the VM, GM and RPM. However, the Langmuir-Hinshelwood kinetic model yields uncertain results at pressures above bar. In these cases, extra steps are added to the reaction mechanism given by Eqs. 6 and 7. First, the representative reactivity was calculated according to Eq.3 to confirm the validity of the LH equation to predict the kinetic behavior of the studied samples, and the average conversion degree ranging from. to.8 was chosen. The relationships between these representative reactivities and the CO pressure are illustrated in Fig. 6 for temperatures of 85, 95 and,5 C for both studied samples. To estimate the model parameters, the reaction was assumed to be prevented by P CO =. 4

15 Topic: Gasification / Co-Gasification - #. Janina coal char Turów coal char.6 85 C r, /min r, /min P CO, bar 5 5 P CO, bar C r, /min.4.3. r, /min P CO, bar 5 5 P CO, bar 5 C r, /min P CO, bar Figure 6. Fitting of the LH expression for relationships between the representative reactivity and CO pressures for the gasification of Janina (left-hand side) and Turów (right-hand side) coal chars at temperatures of 85, 95 and,5 C r, /min P CO, bar Fig. 6 shows that an apparent shift in the reaction order can be distinguished as the CO pressure is increased. The LH model appears applicable, particularly for the relationships observed at,5 C for the Janina and Turów coal chars. The shifts of the reaction order are still present at lower temperatures, but the significant inconsistencies are observed at and bar. The LH model expression explains the saturation effect of the reaction surface, which desensitizes the rate to pressure increases. This desensitization occurs because the surface is maximally concentrated with C(O) complexes, and further increases in the CO pressure do not impact their concentration or the reaction rate. Saturation was observed for both samples gasified at,5 C. Nevertheless, the reactivity notably increased when the pressure was increased from to bar at 85 and 95 C. This effect is not fully understood because the apparent reaction order for these pressure range was observed to be nearly zero []. The values of kinetic constants k and k were calculated directly by non-linear estimation using the Levenberg- Marquardt method. The resultant values corresponding to both samples and temperatures of 85, 95, and,5 C are given in Table 7. 5

16 Topic: Gasification / Co-Gasification - # Table 7. Kinetic coefficients calculated for Langmuir-Hinshelwood kinetics for Janina and Turów coal chars gasified at temperatures of 95, 95 and,5 C k Sample Temp., C, k, (bar min) - bar - R, - Janina coal char Turów coal char We also analyzed the model first introduced by Dutta and Wen [], which is thoroughly explained above. The model considers the effect of changes in the pore surface area as the reaction proceeds. Therefore, additional experiments with the Turów coal char were at a temperature of 95 C and a CO pressure of bar. After reaching specific values of the carbon conversion degree, the reaction was halted, replacing CO with inert gas. After cooling under an inert atmosphere, the pore characteristics of the partially reacted samples were analyzed by nitrogen and carbon dioxide adsorption in the same manner as the initial char samples. Four values of the carbon conversion were selected for this study:.,.4,.6 and.8. The micropore surface areas were measured with CO adsorption at 73 K (with Dubinin-Astakhov theory). The BET surface area was measured by N adsorption at 77 K, and the relative surface areas corresponding to micropores and mesopores are summarized in Table 8. Table 8. Pore surface area and relative pore surface area calculated by means of the Dubinin-Astakhov and BET equation for the adsorption of CO and N isotherms of the partially gasified samples of Turów coal char X, - S D-A, m /g S BET, m /g f D-A, - f BET, The values presented in Table 8 indicate how the surface area changes over the course of the reaction. The sample of Turów coal char can be considered microporous. During the reaction, the surface area of these pores increases to a maximum value observed at a carbon conversion of.6, after which the surface area decreases. Incomparably higher increases in the surface area are observed for mesopores given by the BET equation, but a maximum is still observed at a conversion of.6. This observation is not consistent with the literature [3], where maxima in the surface area are observed for conversion degrees lower than.3 when calculated from the CO adsorption isotherm and for conversion degrees higher than.5 for the N adsorption isotherm. The gasification reaction is generally accepted to occur primarily at the char surfaces located within the micropores, and this conclusion was also supported by the present work. Thus, the values of the relative surface area calculated by the CO adsorption were used for further calculations with the Dutta and Wen model. The relationship between the relative available surface area and the conversion degree is presented in Fig. 7. According to Eq.7, the change in f can be expressed with a relationship that incorporate two empirical constants: ν, which represents the value of the conversion at which f is maximized; and β, which is the fitting constant and does not reflect a specific meaning. Knowing that ν equals.6, the β parameter can be determined by a non-linear estimation with the leastsquares method. The estimated value of β was 5.8, which falls in the same range reported by Dutta and Wen []. Nevertheless, the relationship given by Eq.7 for the relative surface area fails to predict 6

17 Topic: Gasification / Co-Gasification - # conversion degrees of.8 (Fig. 7). To calculate the kinetic coefficients of the Dutta-Wen model for the Turów char, the values of ν and β were assumed to also be applicable to temperatures of 85 and,5 C..6.5 f D-A, ν =.6 β = 5.8 Figure 7. Change in the relative pore surface area of micropores with the increasing conversion degree calculated by the Dutta and Wen relationship The last model analyzed in this study was developed by Johnson [7-8]. This model modifies the GM equation by adding a term that is related to the decrease in available surface area. This term equals when pure gases are used as gasifying agents, as stated by Johnson. Additionally, a plot of dt X 3 ln dx against X was proven to yield a constant value for all experiments, so α =. Therefore, Eq.33 results in a modification of the GM as follows: dx flkt X dt 3 (34) Thus, the constants from the GM and Johnson model are related to each other as follows: k f k (35) GM L T.. Exp. data Conversion degree, X, - In this case, k T can be expressed in several ways. Johnson proposed an LH-type expression for representative reactivity as being equivalent to k T, and this approach has been accepted in the present work. Because only one reaction is analyzed, only the parameters for the char-co reaction are needed to calculate k T. Table 9 summarizes the parameters of Johnson equation; the values of the relative reactivity factor are similar to those reported by Johnson [7-8] for hydrogen gasification. The values of f L are independent of the coal char rank, but the effect of the CO pressure was less pronounced at higher temperatures, which is indicated by a narrower range of calculated f L values. The rank, temperature or pressure did not affect the f values. The Arrhenius equation parameters (activation energy and pre-exponential factor) were calculated for all analyzed models, except for the Johnson model, for which both f L and f do not hold a kinetic meaning. The GM-, RPM-, MVM and L-H-based kinetic constants can be related to the absolute temperature according to the relationship given by Eq. 3. For the unification theory and the Dutta and Wen model, expressions based on the reaction rate for the specific conversion degree (here.) and the half-life of the reaction are used instead of the kinetic constants. Because the VM poorly fit the kinetic data, the kinetic constants of this model have been excluded from the calculation of the kinetic parameters. Fig. 8 shows the Arrhenius plots for all of the kinetic coefficients or their representatives for the GM, RPM, MVM, unification theory model (a-d), Dutta and Wen model (e), and the Langmuir-Hinshelwood model (i). The Arrhenius relationship for the LH-derived kinetic constants are gathered together for both analyzed samples; the relationships for the other models are divided to the left- and right-hand side for the Turów and Janina coal char, respectively. The vast majority of Arrhenius plots create parallel lines, indicating similar values of the activation energy obtained from different kinetic expressions. Interestingly, the Model 7

18 Topic: Gasification / Co-Gasification - # relationships for the MVM and the unification theory model (UTM) were so close that they can be considered overlapping for both the Turów and Janina coal chars and all pressures. The GM and the RPM yield lines parallel to the MVM and UTM. However, the RPM shows a significant discrepancy for the Turów char gasified at bar (Fig. 8e). The coefficients derived for the Dutta and Wen model of the Turów coal char gasification at bar (Fig. 8e) exhibit parallelism and closely approximated the GMderived constants. The Arrhenius relationships obtained for the LH-type kinetic expression distinctly shows the effect of parent coal rank on the value of the activation energy of the k constant. However, the relationships obtained for k for the Janina and Turów coal char yield parallel lines. This relationship indicates that the rate of the desorption step in the LH mechanism (represented by the kinetic constant k ) is not connected with the coal rank but more likely with the applied process conditions. In summary, the calculated values of the activation energy and pre-exponential factor (or fk D-W for the Dutta and Wen model and R u /U for the UTM) for both samples and all models and pressures are gathered in Table. Table 9. Parameters of the Johnson equation for the Janina and Turów coal chars at the studied temperature and pressure conditions Janina coal char Turów coal char Temp., C P CO, bar k T, /min f L, - f, - k T, /min f L, - f, For all studied kinetic expressions and applied pressure conditions, the values of the activation energy ranged from 65 to kj/mol for the Turów coal char and 86 to 38 kj/mol for the Janina coal char. This agrees with the data reported in the literature. However, the activation energies for Janina coal char seem to be too small. The slightly higher values for the Janina coal char do not represent the real diversity in the reactivity between studied samples. The activation energies calculated for the same pressure conditions but with different models are similar, and the differences are less than kj/mol, indicating that the real activation energy likely falls in this range. For the L-H type expression, the differences in the rank of the parent coal are manifested in the k constant. Thus, the activation energies were 8 and 5 kj/mol for the Turów and the Janina coal char, respectively. As mentioned above, the activation energies corresponding to the kinetic coefficient k are similar and vary from 86 to 9 kj/mol. 8

19 Topic: Gasification / Co-Gasification - #.4 bar ln ki; -ln t / ln k GM = -64 T R² =.9883 Turów coal char Janina coal char /T, K a) ln k MVM = -344 T R² =.998 ln k RPM = -94 T R² = ln t / = -4 T R² =.985 ln kgm ln krpm ln kmvm ln t/ ln k i ; -ln t / ln k MVM = T b) -3 R² = ln t / = -498 T R² = ln k GM = -534 T R² =.994 /T, K - ln k RPM = -643 T R² =.9886 ln kgm ln krpm ln kmvm - ln t/ bar bar bar ln k i ; -ln t / ln k i ; ln (dx/dt/(-x)); -ln t / ln k i ; -ln t / c) ln k GM = -57 T R² =.983 /T, K - ln k MVM = T R² =.9877 ln k RPM = -5 T R² = ln t / = -444 T R² =.978 ln kgm ln krpm ln kmvm - ln t/ /T, K ln k MVM = T e) R² =.9999 ln k GM = T R² =.999 ln k RPM = -984 T R² = ln t / = -359T R² =.994 ln (dx/dt/(-x)) = -46 T R² =.984 ln kgm ln krpm ln kmvm ln (dx/dt/(-x)) for X=. -ln t/ ln k MVM = -334 T g) - R² = ln t / = -8T R² = ln k GM = - T R² =.9756 /T, K - ln k RPM = -669 T R² =.974 ln kgm ln krpm ln kmvm -ln t/ ln k i ; -ln k / ln k i ; -ln t / ln k i; -ln k / ln k MVM = -866 T d) R² = /T, K - ln k GM = T R² =.9956 ln k RPM = -857 T R² = ln t / = T R² =.997 ln kgm ln krpm ln kmvm - ln t/ /T, K f) ln k GM = -464 T R² =.9979 ln k GM = -39 T R² =.9973 ln k RPM = -449 T R² = ln k / = -4759T R² =.999 ln kgm ln krpm ln kmvm -ln k/ ln k MVM = -89 T h) R² = ln k GM = -44 T R² =.9939 /T, K - ln k RPM = -565 T R² = ln k / = -386T R² =.986 ln kgm ln krpm ln kmvm - ln k/ 9

20 Topic: Gasification / Co-Gasification - # ln k i /T, K i) Janina char: ln k = -363 T R² =.995 ln k = -369 T R² =.9998 Turów char: ln k = -6 T R² =.987 ln k = -39 T R² =.9968 ln k Janina char ln k Janina char ln k Turów char ln k Turów char Figure 8. Arrhenius plots for the kinetic constants or their representatives corresponding to the GM (a-h), RPM (ah), MVM (a-h), UTM (a-h), Dutta and Wen (e) and Langmuir-Hinshelwood model (i) for Turów (left-hand side) and Janina (right-hand side) coal chars gasified at CO pressures of.4,, and bar Table. Summary of the Arrhenius law parameters obtained for various kinetic expressions for the Turów and Janina coal chars at pressures of.4,, and bar Model GM RPM MVM UTM Press., bar Turów coal char Janina coal char E, kj/mol A -6, /min E, kj/mol A -6, /min E, kj/mol A -6, /min E, kj/mol Ru/U 7, /min U -6, min E, kj/mol L-H k k A -8, E, /min kj/mol A -6, /min E, kj/mol Dutta and Wen`s model fk D-W -7, /min Not applicable Not applicable k D-W -7, /min Conclusions The CO gasification rates of two coal chars derived from Polish sub-bituminous coal and lignite were studied experimentally. Twenty-four tests were performed in a pressurized thermogravimetric analyzer between 85 and,5 C and pressures of.4,, and bar. The char derived from the lower rank lignite was approximately two times more reactive (in terms of the time needed to complete reaction) than the sample derived from sub-bituminous coal. Increases in temperature and pressure significantly affected the rate of gasification observed for both samples. However, increases in the temperature affected the reaction rate much more dramatically than increases in pressure. Extending the reaction for Turów coal char gasified at 95 C and bar significantly increased the mesopore surface area. Conversely, smaller changes were observed for the micropore surface area. To perform a comprehensive kinetic analysis, eight models were used to describe the kinetic behavior of studied the samples: the VM, GM, RPM, MVM, unification theory model, Langmuir-Hinshelwood, Dutta and Wen and the Johnson model. The validity of these models was examined after analyzing their fit with the experimental data, and the values of the R determination coefficients were found to be adequate indicators of model applicability. The results obtained for the VM were considered unsatisfactory, but the results from the GM, RPM and MVM were significantly better, especially for the last two models. Very good results were also obtained using the UTM, and the unification curves showed that all performed 3

21 Topic: Gasification / Co-Gasification - # experiments could be described by one single relationship up to conversion values of.6. The Dutta and Wen model accurately predicted the change in pore surface area of micropores and therefore was successfully adopted to predict the intrinsic reaction rate. The Johnson model was applicable to the same cases at the GM because the term corresponding to the decrease in the available surface area yielded for the studied process conditions. The most general Langmuir-Hinshelwood model was applicable to all pressures, even though it is not recommended for pressures higher than bar. The results were found to be satisfactory, especially at 5 C, when both samples showed a saturation effect at and bar. The calculated values of the activation energy were similar for the various models used, and ranged from 65 to kj/mol for the Turów char and 86 to 38 kj/mol for the Janina coal char. For the Langmuir- Hinshelwood kinetics, the activation energy for the formation of active complexes was a function of the rank of the parent coal and was higher for more altered coal char (8 and 5 kj/mol for Turów and Janina char, respectively). The activation energy of the second step was found to be lower than that of the first step, and the values were similar for both coal chars and ranged from86 to 9 kj/mol. The comprehensive kinetic analyses suggest the most suitable kinetic expression, but most models yielded accurate results, which complicates the selection of the most appropriate model. However, the unification theory model still stands out among these models because of its simplicity. This model only requires raw data (X and dx/dt) to perform kinetic analyses and to calculate the activation energy and pre-exponential factor, which does not require model fitting. Acknowledgements The research results presented herein were obtained during the course of the Development of coal gasification technology for high-efficiency production of fuels and energy project, Task No. 3 of the Strategic Program for Research and Development: "Advanced energy generation technologies" funded by the Polish National Centre for Research and Development. Martyna Tomaszewicz has received a scholarship under the project DoktoRIS Scholarship Programme for the Innovative Silesia, which was co-financed by the European Union in the frame of the European Social Fund. References [] Chmielniak, T; Sciazko, M.; Tomaszewicz, G.; Tomaszewicz, M., Pressurized CO -gasification of coal. 5 th International Freiberg Conference on IGCC & XtL Technologies, -4 May, Leipzig, Germany [] Molina, A.; Mendragon, F., Reactivity of coal gasification with steam and CO. Fuel 998, 77, [3] Irfan, M.F. ; Usman, M.R.; Kusukabe, K., Coal gasification in CO atmosphere and its kinetics since 948: A brief review. Energy, 36, -4 [4] Roberts, D.G.; Harris, D.J., A kinetic analysis of coal char gasification reactions at high pressures. Energy & Fuels 6,, 34-3 [5] Fermoso, J.; Stevanov, C.; Moghtaderi, B.; Arias, B.; Pevida, C.; Plaza, M.G.; Rubiera, F; Pis, J.J., High-pressure gasification reactivity of biomass chars produced at different temperatures. Journal of Analytical and Applied Pyrolysis 9, 85, [6] Ishida, M.; Wen C.Y., Comparison of zone-reaction model and unreacted-core shrinking model in solid gas reactions I isothermal analysis. Chemical Engineering Science 97, [7] Szekely, J.; Evans J.W., A structural model for gas-solid reactions with moving boundary. Chemical Engineering Science 97, 5, 9-7 [8] Bhatia S.K.; Perlmutter, D.D., A random pore model for fluid-solid reactions: I. Isothermal, kinetic control. AIChE Journal 986, 7,

22 Topic: Gasification / Co-Gasification - # [9] Kasaoka S, Sakata Y, Tong C. Kinetic evaluation of the reactivity of various coal chars for gasification with carbon dioxide in comparison with steam. International Chemical Engineering 985, 5, 6-75 [] Dutta, S.; Wen, C. Y; Belt, R., J., Reactivity of Coal and Char.. In Carbon Dioxide Atmosphere. Industrial & Engineering Chemistry Process Design and Development 977, 6, -3. [] Raghunathan, K.; Yang R.Y.K., Unification of Coal Gasification Data and Its Applications. Industrial & Engineering Chemistry Research 989, 8, [] Ochoa, J.; Cassanello, M.C.; Bonelli, P.R.; Cukierman A.L., CO gasification of Argentinean coal chars: a kinetic characterization. Fuel Processing Technology, 74, 6-76 [3] Liu, H.; Luo Ch.; Kaneko M.; Kato Sh.; Kojima T., Unification of gasification kinetics of char in CO at elevated temperatures with a modified random pore model. Energy & Fuels 3, 7, [4] Ergun, S. Kinetics of the Reaction of Carbon Dioxide With Carbon. Journal of Physical Chemistry 956, 6, [5] Chen, G.; Yu, Q.; Sjostrom, K., Reactivity of char from pyrolysis of birch wood. Journal of Analytical and Applied Pyrolysis 997, 4-4, [6] Ollero, P; Serrera, A.; Arjona, R; Alcantarilla, S, The CO gasification kinetics of olive residue. Biomass and Bioenergy 3, 4, 5-6 [7] Johnson, J.L., Relationship between the gasification reactivities of coal char and the physical and chemical properties of coal and coal char. American Chemical Society, Division of Fuel Chemistry Preprints 975,, 85-. [8] Johnson, J.L., Kinetics of coal gasification 987 John Wiley & Sons, New York. [9] Walker, P.L. Jr; Pentz, L.; Biederman, D.L.; Vastola, F.L., The influence of inert diluent gases on the rate of carbon gasification. Carbon 977, 5, [] Zhang, L.; Calo J.M., The inert gas effect on carbon reactivity. Proc. Carbon '95,. nd Biennial Conf. Carbon, University of California, San Diego 995, [] Tomaszewicz, M.; Łabojko, G.; Tomaszewicz, G.; Kotyczka-Morańska, M., The kinetics of CO gasification of coal chars. Journal of Thermal Analysis and Calorimetry 3, DOI:.7/s [] Roberts, D. G.; Harris, D.J., Char Gasification with O, CO, and HO: Effects of Pressure on Intrinsic Reaction Kinetics, Energy & Fuels, 4, [3] Hurt, R.H.; Sarofim A.F.; Longwell, J.P., The role of microporous surface area in the gasification of chars from a sub-bituminous coal. Fuel 99, 7,