Deactivation kinetics for lignite gasification in a fluidized bed reactor

Fuel 236 (2019) 1050–1056

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Deactivation kinetics for lignite gasification in a fluidized bed reactor Tansu Uyar, Yahya Suyadal

⁎

T

Ankara University, Faculty of Engineering, Department of Chemical Engineering, 06100 – Tandoğan, Ankara, Turkey

A R T I C LE I N FO

A B S T R A C T

Keywords: Deactivation kinetics Gasification Afşin–Elbistan lignite Fluidized bed reactor

Lignite gasification may be characterized by the active site restriction result from deactivation. As was aimed here, a deactivation model (DM) can be developed by “pseudo-steady-state” mass balance, i.e. describing variation of available surface with time. For this purpose, bench-scale fluidized bed reactor (FBR) was operated for obtaining gasification product profiles. Gas samples were taken from the reactor effluent stream, and fed continuously to the gas analyzers for on-line simultaneous measurements of H2, CO2, CO, CH4 and O2. Data were used for extracting model parameters (k// and kD) from the linear form of DM. Experiments at temperature range of 973–1173 K were conducted. Fluidizing mixture (air and steam) entered the bed through a distributor of a 200-mesh stainless steel sieve and fluidized the single charge of lignite with a mean particle diameter approximately 250 µm. The agreement between the experimental and predicted carbon concentrations was confirmed for DM. The latter may be successfully used to design the fluidized bed combustors or gasifiers.

1. Introduction

grinding of the raw material to fine or ultra-fine particles. With increasing quantities of coal fines occurrence of sintering seems more likely as it was observed in some of our studies [4]. Agglomeration and sintering may cause unforeseen consequences in the operation. Silica sand is the most typical bed material in fluidized beds because of its good mechanical properties and abundant reserves [5–9]. However, agglomeration of silica sand bed material is widely reported [10]. Inorganic alkali material in the fuel, mainly potassium and sodium cause agglomeration by the formation of silicates with the silica from the sand which has low melting points. The content of these inorganic components can vary between fuels; especially in the case of some biomass types as well as various low-rank coals. As a result, sand particles get coated with an adhesive layer. With the collision of solid structure sticky sand particles create larger agglomerates [11]. Some precautions for reducing the agglomeration risk, is to use mineral based bed materials, such as Al2O3, olivine and dolomite which are resistant toward agglomeration or to add some materials like kaolin, calcium oxide, calcium carbonate and bauxite which have the ability to trap or react with components causing agglomeration. Minerals are considered to be more or less brittle and sensitive toward attrition when used as bed material in fluidized bed gasification [10]. Most of the previous reviews focused on the agglomeration characteristics of fluidized beds based on chemical properties of fuel. Physics based parameters like particle size, velocity and collision frequency also affect the agglomeration. It is important to conceive these effects along with the particle chemistry. For example, bed additives not only transform the

Coal is generally used as an energy source by direct combustion. But combustion of coal with an unprocessed form causes various environmental problems. The amount of harmful emissions depends on the efficiency and utilization method of combustion process and the properties of the fuel source used. In this regard gasification seems to be an appropriate technique for coal utilization [1]. But no matter the process, some obstacles like agglomeration and sintering still appear as the common operation problems. Fluidized Beds have fuel flexibility concerning the particle size and moisture content but suffer from two major setbacks: a generic problem associated with increased tar content in the gas products that inhibits its efficient utilization and a fuel-specific problem, caused by the low melting temperatures of ashes that paves the way for particle sintering, agglomeration, leading to the inevitable defluidization of the bed [2]. Agglomeration problems are related to the transformations of mineral matter in solid fuels which depend on the type and composition of the fuel. The melting behavior of mineral matter in fuel is considered to be an important parameter. Where a sorbent is used for emissions control, such as the use of limestone in FB combustion, mineral species derived from the sorbent can also have a role. A model for prediction of agglomeration problems in specific operating conditions seems crucial for the utilization of solid fuels [3]. As for the sintering, present mining processes produce large quantities of coal fines, and also some beneficiation techniques require

⁎

Corresponding author. E-mail address: [email protected] (Y. Suyadal).

https://doi.org/10.1016/j.fuel.2018.09.028 Received 25 August 2017; Received in revised form 15 May 2018; Accepted 6 September 2018 0016-2361/ Published by Elsevier Ltd.

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal

Nomenclature k// kD C F V Q y r S t ks EA ED

D L m P T dp xC

reaction rate constant, s−1 deactivation rate constant, s−1 concentration, kmole.m−3 molar FLOW rate, kmole. s−1 reactor volume, m3 volumetric flow rate, m3.s−1 mole fraction, kmole.kmole−1 reaction rate, kmole. m−3.s−1 particle surface area, m2 Time, s surface reaction rate constant, m.s−1 activation energy for surface reaction, kJ.kmole −1.K−1 activation energy for deactivation, kJ.kmole−1.K−1

reactor diameter, m reactor height, m mass, kg pressure, kPa temperature, K particle size, m carbon conversion ratio, kmole.kmole−1

Greek Letters α β γ ρ

stoichiometric coefficient of steam, (–) stoichiometric coefficient of air, (–) total amount of converted carbon, kmole.m−3 density, kg.m−3

model (ASIM), in terms of a simplified calcium-catalyzed mechanism of char gasification to characterize a conversion-dependent maximum in reaction rate [15]. Dahlin et al. evaluated different additives to assess the ability to prevent ash agglomeration during the high-sodium lignite gasification. They conducted some series of muffle furnace tests selected meta-kaolin for a following work with a pilot-scale coal gasifier. Agglomeration and deposition problems during gasification of high-sodium lignite successfully prevented at a maximum operating temperature of 1200 K and a meta-kaolin (mean size of 920 μm) feed rate with roughly equivalent to the ash content of the lignite (approximately 10 wt%) [16]. Khadilkar et al. conducted a research with particle classes of the composite fuels, based on differences in density and size, in order to understand the physics and chemistry at particle level. They determined slag-liquid formation tendencies under fluidized bed operating temperatures both computationally and experimentally. They used a thermodynamic simulation software and proposed an integrated ash agglomeration model that accounts for particle hydrodynamics as well as particle class level ash chemistry to predict agglomeration kinetics [17]. As a continuation of previous studies outlined above, the aim of this study is to examine deactivation kinetics for lignite gasification in a

chemical structure of the bed materials but also change the mean particle size and density. This would affect the hydrodynamics in bed and physics-based parameters, which may affect agglomeration rate [12]. Namkung et al. investigated agglomeration tendency using different operating conditions consisting particle size, coal/sand ratio, temperature and fuel type. They used various materials like kaolin, alumina and additives to prevent bed agglomeration and found out that agglomeration tendency increased with smaller particle sizes and higher temperatures [13]. Lin et al. performed thermodynamic equilibrium calculations to identify the stable silica, potassium, chlorine and sulfur species. Their results showed that potassium silicates were the main form present in the bed. They also developed a simple model to describe the defluidization time including parameters like function temperature, fluidization velocity and particle size. The model was based on a competition between the breaking force induced by bubbles in the bed and the adhesive force caused by ash coating and sintering [14]. Tang et al. studied two mineralogically different lignite chars and reported that commonly known kinetic models like VM (Volumetric Model), SCM (Shrinking Core Model) and RPM (Random Pore Model) are unsatisfactory, also in some cases invalid to express the kinetics of char gasification. Therefore proposed a new active site/intermediate Table 1 Lignite characterization tests. Tests

Results

Pore size analysis: AUTOPORE II 9220

Total pore volume: 0.116 mL/ g; Mean pore radius: 0.003 μm; Porosity: 17.417%

Tests

Results

Proximate analysis: (ASTM-D-7582)/(ASTM-D-5865)

Weight (%):

Volatile matter

Moisture

Fixed carbon

Ash

Calorific value (kJ/kg)

35.65

6.28

11.96

46.12

9 990

Tests

Results

Ultimate analysis: (ASTM-D-5373)/(ASTM-D-7582)/(ASTM-D-5016)/(ASTM-D-121) Tests

Complete chemical analysis: (ASTM-D-4326-04)

Dry basis (%): Ash-free dry basis (%):

C

H

O

N

S

Ash

30.97 60.98

2.69 5.29

10.62 20.90

4.82 9.50

1.69 3.33

49.21

Results

Ash (%): Slag (%):

MgO

Al2O3

SiO2

CaO

Fe2O3

SO3

K2 O

Na2O

P2O5

H. Loss

2.49 2.11

13.27 14.69

43.00 46.40

21.05 11.95

7.93 6.97

6.56 2.12

1.08 1.18

1.32 0.61

0.48 0.35

1.66 12.47

1051

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal

fluidized bed reactor. In fact, there is no information in the open literature about an attempt to use deactivation model itself to obtain the gasification kinetics [18]. Therefore, a novel approach was presented here to determine the kinetic parameters for this model. Although it may seem reasonable to model the gasification process based on the various models, namely, Shrinking Core Model (SCM), Grain Model (GM), Random Pore model (RPM), Volume reaction model (VRM) and the other models encountered in the corresponding literature reported [19–37]. The application of these models to the lignite gasification have been described by various investigators, and in most cases good agreement between theory and experiment has been found. A disadvantage of these approaches, however, is that the model equations are generally non-linear and it is difficult to incorporate them into a hydrodynamic model of a fluidized bed gasifier or combustor without having to resort to prohibitively lengthy computer calculations. As a result, a number of semi-empirical models have been developed by investigators [38–40]. The latter investigators used the analogy between lignite gasification and the deactivation of catalyst particles by coke formation, both of which can be described in terms of an exponential decrease in available surface with time. Similarly, a good description of carbon conversion ratio by the deactivation model (DM) makes it more appropriate for lignite gasification by analogy [41]. This work concerns here, on one hand, to explore the variations of DM parameters (k// and kD) with operating temperature (T), volumetric ratio of gas feedstock (H2O/O2) and particle diameter (dP), on the other hand, to test the DM for the first time for the description of performance curves (CC/CC0) for lignite gasification in addition to carbon conversion ratios (XC) obtained from an integral FBR. Eventually, DM was used to successfully describe experimental carbon conversion ratio for the lignite gasification in a fluidized bed reactor.

was suitable for a bubbling-bed regime. Gas samples were continuously taken from the reactor effluent stream and it went through a conditioning unit which included two water traps and an active carbon column. Mole fractions of H2, CO, CO2, CH4 and O2 in H2O-free clean gas were determined by a biogas analyzer (GASBOARD-3100P) via on-line measurements. Firstly, air was fed to the bubble column and air-steam mixture entered the reactor through the preheating region. When the steadystate gas flow was provided, charred lignite particles were loaded into the bed instantly by a screw feeder with a batch-wise manner. Each run was terminated when the steady state was reached, i.e. when the exit concentration of gasification feedstock and products approached to their initial values. Table 2 shows the FBR operating conditions used for the experiments. 3. Model development Throughout the experiments, four product components (CO, CO2, CH4 and H2) were detected by the analyzer from the exit gas stream. In order to explain the variety of products, following reactions (Eqs. (1)–(5)) were considered to be dominant in the process. However, these five reactions were not the main focus of this work. Since DM solely relies on the solid phase data, a simplified global reaction was presented here for obtaining the kinetic parameters.

Complete combustion O2 + C⇔ CO2

(1)

1

2. Experimental

Partial combustion 2 O2 + C⇔ CO

(2)

Water−−gas reaction H2 O+ C⇔ H2 + CO

(3)

Shift reaction H2 O+ CO ⇔ H2 + CO2

(4)

1

1

1

Methanation H2 + 3 CO ⇔ 3 CH 4 + 3 H2 O

(5)

2.1. 2.1 materials and pretreatment Lignite was provided from Afşin–Elbistan coal basin of Turkey and its characterization test results were shown in Table 1. First, the lignite samples were charred with the inert gas (N2: 99.9%) as a fluidizing medium at 1173 K in the same reactor which was also used for gasification experiments. It was expected that increment of the available surface and carbon content of lignite should increase the rate of gasification at temperature range of 973–1173 K investigated, due to the absence of calcinations and impurities which were eliminated by this preliminary process. Then, according to Tyler standard [42], a sieve analysis was also applied for determination of mean particle diameter of charred lignite. 2.2. Apparatus and procedure In the course of lignite gasification experiments, the set-up given in Fig. 1 was used. This set-up involves mainly a fluidized bed reactor (FBR) and auxiliary equipments for gas preparation, sampling/analysis, and on-line data logging system. Stainless steel FBR had a height of 750 mm and 50 mm diameter. It was fitted with a 200 mesh gas distributor. A silicon carbide heater was used to heat the reactor. Bed temperature was monitored with a Ni–Cr/ Ni type K thermocouple and adjusted by a PID loop. The reactor was also connected with a preheating region to heat the incoming gas stream. In order to drag steam from bubble column to the reactor, an air compressor was harnessed. Bubble column was heated with a PID controlled circulator, so the gas feed was arranged with desired ratios. The volumetric flow rate of fluidizing air (it was 65 × 10−6 m3/s for the experiments conducted to determine the DM parameters) and steam were set to render a space velocity for the chosen U0/Umf ratio which

A AC A/D BC C C/P CW DM EH F

Analyzer Activated Carbon Analog/Digital Bubble Column Cyclone Compressor/Pump Condenser Digital Measurement Electrical Heater Furnace

FB GF GM P PD RV SF T V V3

Fluidized Bed Gas Filter Gas Mixer Pressure Pressure Dropper Regulator Solid Feeder Temperature Flow Rate Three way valve

Fig. 1. The experimental set-up. 1052

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal 3

Table 2 The operating conditions used.

⎡ t ⎤ C(H2O)0 ⎢ 1 ∞ i=Σ1Fi ⎥ CC dt = 1− ⎢ (CC )0 τ 0 F(H2O)0 ⎥ (CC )0 ⎢ ⎥ ⎣  ⎦  

∫

Properties

Values

Bed material [43] FBR: ϕ (D × L) Operation period (t∞) Particle charge (m Lignite) Particle density (ρ Lignite) Pressure (P) Air flow rate Temperature range The ratio of H2O to O2 range Mean particle diameter range

Lignite particles in Group B 50 × 750 mm 300 s 6.000 × 10−3 kg lignite 1.500 × 103 kg/m3 101.3 kPa (Atmospheric) 65 × 10−6 ≤ QAir, m3/s ≤ 150 × 10−6 973 ≤ T, K ≤ 1173 25 ≤ H2O/ O2, v/v ≤ 100 250 ≤ dP, µm ≤ 450

On the other hand, carbon mass balance or the variation of solid carbon concentration (from (CC)0 to (CC)∞) in the FBR with time (t) was calculated from the sum of molar flow rates of carbon containing gases (i = 1–3; CO, CO2, CH4) only, and as well known H2-free and O2-free, by substituting Eq. (10) into Eq. (11). In order to obtain a kinetic model for representation of the process, the following assumptions may be imposed: a. FBR is isothermal, and it is reasonable for reactants to use physical properties at its mean temperature. b. Pseudo steady-state assumption is valid in the FBR, so it can be represented as a batch-solid reactor. c. Deactivation of lignite from result agglomeration is the first order with respect to the particle surface area, and it can be also described in terms of an exponential decrease with time in its available surface as follows (Eq. (12)):

Lignite gasification process in the presence of steam together with the air can be generally expressed by the following chemical reaction (Eq. (6)) between the carbon (C) and the steam-air mixture (H2O + O2 + N2): According to this reaction process, if α, β and γ are greater than zero, there are H2O, O2 and C in the reaction medium. Because of this reason; gasification (water-gas, shift and methanation reactions) and combustion (complete and partial oxidation reactions) simultaneously carry out in the solid and the gas phase, the H2, CO, CO2 and CH4 gases form. The stoichiometric coefficients of the feedstock can be easily determined by means of the atomic mass balances (Eqs. (7)–(9)).

α H2 O(Steam) +

FH2 FCO 1 H2 + CO βO2(Air) + γ C(Lignite) → ⋯→ F(H2O) 0 F(H2O) 0 2 FCH4 FCO2 CH 4 CO2 + + F(H2O) 0 F(H2O) 0

dS = −kD S⇒ S= S0 exp(−kDt) dt

(−rC) // = −

C:

CC0

(7)

(FH2O)0

=

dCC kS = −⎛ S 0 ⎞ CC ⎝ V ⎠ 

t

∫ exp(−kDt)dt 0

(14)

= k //

CC k // [1−exp(−kDt)] ⎫ = exp ⎧− ⎨ ⎬ (CC )0 ⎩ kD ⎭

(8) t∞

∫ 0

(FCO + FCO2 + FCH4 ) ⎤ dt⎥ F(H2O)0 ⎦

yi (FN2 )0 ⎡ ⎤ ⎥ (FH2O)0 ⎢ (100 Σ y ) − i 1 5 = − ⎦ ⎣

(15)

Substituting of t → ∞ into the Eq. (15) (i.e. CC(∞) = (CC)∞) and the rationing of Eq. (15) to Eq. (16), these can be rearranged to obtain Eq. (17) and Eq. (18), respectively.

(9)

Where, γ (=(CC)0xC) stands for total amount of converted carbon which can be calculated by summing the mole amounts of carbon containing gases (nC0 xC = nC0 − nC = nCO + nCO2 + nCH4), for the operation period (t∞). Also, τ (=V/Q0), V and Q0 are the space time, FBR volume and volumetric flow rate of the inlet stream, respectively. When Fig. 2 is examined, a residence time distribution effect can be easily noticed. This effect was eliminated using the component mass balances individually and also by harnessing curve fitting techniques (Fi = Fi(t)). So the data obtained from analyzer were processed properly and used for the determination of molar flow rates, along with their initial values as was illustrated in Fig. 3. Fig. 3 shows a typical variation of dimensionless molar flow rates of gasification products in the reactor effluent stream with time. A reliable method based on dry basis via inert (N2) was used for the calculation of molar flow rates and Eq. (10) was formed. As was given here, the dimensionless molar flow rate of products based on gasification agent (H2O vapor) for several measurable components (i = 1–5; CO, CO2, CH4, H2, O2) and complementary inert gas evolution (yN2 = 100–Σyi=1–5) was determined by the mole fraction data (yi) for each set of experimental operating conditions given in Table 2.

Fi

(13)

with result

FCO + 2FCO2 −α F(H2O)0

C(H2O)0 ⎡ 1 γ = xC = (CC )0 ⎢ τ (CC )0 ⎣

1 d (VCC) = kS CC S dt

or in an integral form

(6)

CC

O:β =

(12)

Mass balance for carbon (C) at any time in the solid phase can be given by Eq. (13). Substitution of Eq. (12) into Eq. (13) yields Eq. (14):

∫

FH2 + 2FCH4 H:α = F(H2O)0

(11)

= xC

(CC )∞ k // ⎞ = exp ⎛− (CC )0 ⎝ kD ⎠ ⎜

⎟

22

(16)

T=1123 K H2O/O2=50 (v/v) (dp)mean=250 μm

20 18 16 14

CO CO2

10

CH4

yi , %

12

H2

8

O2

6 4 2 0

(10)

0

50

100

150

t,s

200

250

Fig. 2. Variation of experimental mole fractions. 1053

300

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal

0,050 0,045 0,040

Model

0,035

CO CO2

0,030

CH4

0

2

Fi /F(H O)

to deactivation.

T=1123 K H2O/O2=50 (v/v) (dp)mean=250 μm

4.3. Influence of feed ratio and particle size on rate constants In our previous studies [43,44] the effects of H2O/O2 ratio and particle size on gasification were also investigated. So the model parameters were determined by making use of these studies for other different operating conditions which were shown by means of Table 4. As it is easily seen in this table, rate constants increase with H2O/O2 ratio increment. In this regard the feed ratio of inlet gas stream seems to have effect on the rate parameters but actually this effect is really about the mole amount of gas flow which also involves the H2O/C ratio. Because in these set of experiments O2 and C amounts were constant but the H2O feed rate was changed so with more H2O, carbon conversion increased reasonably. As it was expected, with smaller particles agglomeration occurred more intensely and it caused an enhancing effect on the deactivation rate [13].

H2

0,025

O2

0,020 0,015 0,010 0,005 0,000

0

50

100

150

200

250

300

t,s 5. Conclusions

Fig. 3. Variations of dimensionless molar flow rates.

CC (CC )0 k // exp(−kDt) ⎤ = exp ⎡ ⎢ ⎥ (CC )∞ (CC )0 ⎣ kD ⎦

The novel deactivation model developed was easily applied and verified for fluidized bed gasifiers in this investigation provides a simple way to determine the surface reaction rate constant with deactivation rate constant needed for equipment design. These type of reactors can be employed e.g. as gasifier or combustors. It may be concluded that no other analysis is more suitable for fluidized bed gasifiers to use than to obtain relevant model parameters from the concentration profiles of a carbon containing materials. The experimental carbon concentration profiles can be described in terms of a simple deactivation model (DM) containing only two constants, one of which, k// describes specific reaction rate constant for any reactor system while the other, kD describes the terminated reaction or agglomeration. Therefore, two model parameters can be incorporated into a fluidized bed models that assumes the bed to behave essentially as a well-mixed flow reactor. Obtained activation energies seem close to each other and it can be interpreted that deactivation of solid fuel and surface reactions are in a competitive state. According to the magnitude of activation energies, deactivation carries out slightly easier compared to surface reactions. Because of this, lignite gasification process substantially depends on deactivation which may be a result of several reasons (such as agglomeration, defluidization, deposition etc.) so determining its kinetics is extremely important. Generally two rate constants obtained from the DM, follow the same

(17)

//

C k ⎞ ln ⎡ln ⎛ C ⎞ ⎤ = ln ⎛ −kDt ⎢ ⎝ (CC )∞ ⎠ ⎥ kD ⎠ ⎝ ⎣ ⎦ ⎜

⎟

⎜

⎟

(18)

Thus, in order to determine relevant model parameters, a simple way was developed, using Eq. (18). 4. Results and discussion 4.1. Model parameters and confirmation Fig. 4 depicts the variation of dimensionless carbon concentration with time (Eq. (15)) in the FBR. As was clearly illustrated in this figure, carbon conversion ratio increased with higher temperatures. In order to test the proposed DM, left hand side (ln[ln(CC/(CC)∞)]) of Eq. (18) was plotted as a function of time (t) for each temperature. Model parameters were calculated from the straight line with a slope equal to −kD and intercept giving ln(k///kD), from which k// can be obtained. The linearity of data points can be seen in Fig. 5. As was given in Fig. 6, the effect of temperature on the rate constants (k// and kD) can be given by Arrhenius relationships. Here, the units of frequency factors are the same with k// and kD, the magnitudes of EA and ED are 39.0 kJ/mole and 37.4 kJ/mole, respectively. Thus, the relevant rate constants based on the consumption of carbon feedstock and their Arrhenius parameters for lignite gasification process can be obtained. For confirmation of the results, first the model parameters (k// and kD) were inserted into Eq. (15). Then both experimental and estimated values of carbon concentrations were used for calculating the carbon conversion ratio which was defined by means of Eq. (19) obtained from Eq. (11) and the agreement between them was shown in Fig. 7.

CC (CC )0

H2O/O2=50 (v/v) (dp)mean=250 μm

T, K 973 1023 1073 1123 1173

0,9

Cc / (Cc)0

x C = 1−

1,0

0,8

(19) 0,7

4.2. Influence of temperature on rate constants When Table 3 is examined, it can be seen that both rate constants increase with the rise of temperature. Increasing temperature basically works in favor of reaction rates so that k// increases and in parallel deactivation rate also shows the same trend. This could be related to negative change on porous structure of solid and eventually leading up

0,6

0

50

100

150

200

t,s Fig. 4. Variations of CC/CC0 values with time. 1054

250

300

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal

Table 3 Effect of temperature on k// and kD. H2O/O2 = 100 (v/v) (dP)mean = 250 µm T, K 2

//

−1

10 × k , (s ) 102 × kD, (s−1)

973

1023

1073

1123

1173

0.24 0.77

0.29 0.68

0.36 0.84

0.39 1.04

0.54 1.18

Table 4 Effect of H2O/O2 and dp on k// and kD. T = 1173 K (dP)mean = 250 µm H2O/O2, v/v 2

//

−1

10 × k , (s ) 102 × kD, (s−1)

25

50

75

100

0.46 0.74

0.54 1.18

0.68 1.25

0.73 1.77

250

350

450

0.73 1.77

1.30 1.48

0.72 1.18

T = 1173 K H2O/O2 = 100 (v/v)

Fig. 5. Linear form of DM (Eq. (18)) for k// and kD.

(dp)mean, µm 2

//

−1

10 × k , (s ) 102 × kD, (s−1)

agglomeration inflicted by adhesiveness of melted ash. It can be concluded that the real reason for the deactivation in this work was the increment of density which caused defluidization and deposition. In addition pore plugging caused by components in the reaction medium can be considered as a deactivating factor [18]. Nevertheless additives, different bed materials and using coal-biomass blends may be harnessed to prevent such problems or any other operation obstacles, so in this regard further investigation is needed. Acknowledgments The authors gratefully acknowledge the laboratory support from Ankara University as well as the instrumental support from the Scientific and Technical Research Council of Turkey (TÜBİTAK) under grant no: 114M017/3001.

xC

Fig. 6. Arrhenius plots of k// and kD. 0,4 0,3 T=973 K 0,2 0,1 0,0 0,4 0,3 T=1023 K 0,2 0,1 0,0 0,4 0,3 T=1073 K 0,2 0,1 0,0 0,4 0,3 T=1123 K 0,2 0,1 0,0 0,4 0,3 T=1173 K 0,2 0,1 0,0 0 50

References 2

r =0,999

[1] Canel M, Mısırlıoğlu Z, Canel E, Bozkurt PA. Distribution and comparing of volatile products during slow pyrolysis and hydropyrolysis of Turkish lignites. Fuel 2016;186:504–17. [2] Fryda LE, Paopoulos KD, Kakaras E. Agglomeration in fluidised bed gasification of biomass. Powder Technol 2008;181:307–20. [3] Khadilkar A, Rozelle PL, Pisupati SV. Models of agglomerate growth in fluidized bed reactors: Critical review, status and applications. Powder Technol 2014;264:216–28. [4] Gürses A, Doymuş K, Doğar Ç, Yalçin M. Investigation of agglomeration rates of two Turkish lignites. Energy Convers Manage 2003;44:1247–57. [5] Karimipour S, Gerspacher R, Gupta R, Spiteri RJ. Study of factors affecting syngas quality and their interactions in fluidized bed gasification of lignite coal. Fuel 2013;103:308–20. [6] Azargohar R, Gerspacher R, Dalai AK, Peng D. Co-gasification of petroleum coke with lignite coal using fluidized bed gasifier. Fuel Process Technol 2015;134:301–16. [7] Herguido J, Corella J, Gonzállez-Saiz J. Steam gasification of lignocellulosic residues in a fluidized bed at a small pilot scale. Effect of the type of feedstock. Ind Chem Eng Chem Res 1992;31:1274–82. [8] Delgado J, Aznar MP, Corella J. Biomass gasification with steam in fluidized bed: effectiveness of CaO, MgO, and CaO-MgO for hot raw gas cleaning. Ind Chem Eng Chem Res 1997;36:1535–43. [9] Atimtay AT, Kayahan U, Unlu A, Engin B, Varol M, Olgun H, et al. Co-firing of pine chips with Turkish lignites in 750 kWth circulating fluidized bed combustion system. Bioresour Technol 2017;224:601–10. [10] Zhou J, Rosén C, Engvall K. Biomass oxygen/steam gasification in a pressurized bubbling fluidized bed: Agglomeration behavior. Appl Energy 2016;172:230–50. [11] Bartels M, Lin W, Nijenhuis J, Kapteijn F, Ommen JR. Agglomeration in fluidized beds at high temperatures: Mechanisms, detection and prevention. Prog Energy Combust Sci 2008;34:633–66. [12] Khadilkar AB, Rozelle PL, Pisupati SV. Review of Particle physics and chemistry in

r2=0,999

r2=0,996

r2=0,995

r2=0,996

100

150

200

250

300

t,s Fig. 7. Comparison of experimental and predicted.

trend with the variations of different operating conditions. This may be attributed to the change of pore structure with the reactions occurring in the solid phase. In our experimental studies operating temperatures were under 1273 K which can safely be considered as a threshold for ash melting. So it seems reasonable to exclude the probability of 1055

Fuel 236 (2019) 1050–1056

T. Uyar, Y. Suyadal

[13] [14] [15] [16]

[17] [18]

[19] [20] [21] [22] [23] [24]

[25] [26]

[27]

[28]

fluidized beds for development of comprehensive ash agglomeration prediction models. Energy Fuels 2016;30:3714–34. Namkung H, Kim C, Kim D, Yuan X, Kang T, Kim H. Effect of bed agglomeration by mineral component with different coal types. J Energy Inst 2016;89:172–81. Lin W, Dam-Johansen K, Frandsen F. Agglomeration in bio-fuel fired fluidized bed combustors. Chem Eng J 2003;96:171–85. Tang J, Wu X, Wang J. Kinetic study of steam gasification of two mineralogically different lignite chars: An active site/intermediate model. Fuel 2015;141:46–55. Dahlin RS, Dorminey JR, Peng W, Leonard RF, Vimalchand P. Preventing Ash Agglomeration during Gasification of High-Sodium Lignite. Energy Fuels 2009;23:785–93. Khadilkar A, Rozelle PL, Pisupati SV. A study on initiation of ash agglomeration in fluidized bed gasification systems. Fuel 2015;152:48–57. Suyadal Y, Erol M, Oğuz H. Deactivation model for dry desulphurization of simulated flue gas with calcined limestone in a fluidized-bed reactor. Fuel 2005;84:1705–12. Hannula I, Kurkela E. A semi-empirical model for pressurised air-blown fluidisedbed gasification of biomass. Bioresour Technol 2010;101:4608–15. Engin B, Atakül H. Air and oxy-fuel combustion kinetics of low rank lignites. J Energy Inst 2018;91:311–22. Loha C, Chattopadhyay H, Chatterjee PK. Three dimensional kinetic modeling of fluidized bed biomass gasification. Chem Eng Sci 2014;109:53–64. Jin H, Guo L, Ge Z, Cao C, Lu Y. Study on gasification kinetics of hydrogen production from lignite in supercritical water. Int J Hydrogen Energy 2015;40:7523–9. Doherty W, Reynolds A, Kennedy D. The effect of air preheating in a biomass CFB gasifier using ASPEN Plus simulation. Biomass Bioenergy 2009;22:1158–67. Kaushal P, Proell T, Hofbauer H. Application of a detailed mathematical model to the gasifier unit of the dual fluidized bed gasification plant. Biomass Bioenergy 2011;35:2491–8. Gungor A. Modeling the effects of the operational parameters on H2 composition in a biomass fluidized bed gasifier. Int J Hydrogen Energy 2011;36:-6592-6600. Loha C, Chatterjee PK, Chattopadhyay H. Performance of fluidized bed steam gasification of biomass – Modeling and experiment. Energy Convers Manage 2011;52:1583–8. Hannula I, Kurkela E. A parametric modelling study for pressurised steam/O2blown fluidised-bed gasification of wood with catalytic reforming. Biomass Bioenergy 2012;38:58–67. Xie J, ZhongW Jin B, Shao Y, Liu H. Simulation on gasification of forestry residues in fluidized beds by Eulerian-Lagrangian approach. Bioresour Technol

2012;121:36–46. [29] Nguyen TDB, Ngo SI, Lim Y, Lee JW, Lee U, Song B. Three-stage steady-state model for biomass gasification in a dual circulating fluidized-bed. Energy Convers Manage 2012;54:100–12. [30] Dai P, Dennis JS, Scott SA. Using an experimentally-determined model of the evolution of pore structure for the gasification of chars by CO2. Fuel 2016;171:29–43. [31] Xu J, Hu S, Xiang J, Yang H, Sun L, Su S, et al. Kinetic models comparison for steam gasification of coal/biomass blend chars. Bioresour Technol 2014;171:253–9. [32] Schulze S, Richter A, Vascellari M, Gupta A, Meyer B, Nikrityuk PA. Novel intrinsicbased submodel for char particle gasification in entrained-flow gasifiers: Model development, validation and illustration. Appl Energy 2016;164:805–14. [33] Keller F, Küster F, Meyer B. Determination of coal gasification kinetics from integral drop tube furnace experiments with steam and CO2. Fuel 2018;218:425–38. [34] Gonzalez V, Rußig S, Schurz M, Krzack S, Kleeberg J, Guhl S, et al. Experimental investigations on lignite char gasification kinetics using a pressurized drop tube reactor. Fuel 2018;224:348–56. [35] Jayaraman K, Gokalp I. Effect of char generation method on steam, CO2 and blended mixture gasification of high ash Turkish coals. Fuel 2015;153:320–7. [36] Mandapati RN, Daggupati S, Mahajani SM, Aghalayam P, Sapru RK, Sharma RK, et al. Experiments and Kinetic Modeling for CO2 Gasification of Indian Coal Chars in the Context of Underground Coal Gasification. Ind Eng Chem Res 2012;51:15041–52. [37] Tanner J, Bhattacharya S. Kinetics of CO2 and steam gasification of Victorian brown coal chars. Chem Eng J 2016;285:331–40. [38] Svishchev DA, Kozlov AN, Donskoy IG, Ryzhkov AF. A semi-empirical approach to the thermodynamic analysis of downdraft gasification. Fuel 2016;168:91–106. [39] Zhang Y, Ashizawa M, Kajitani S, Miura K. Proposal of a semi-empirical kinetic model to reconcile with gasification reactivity profiles of biomass chars. Fuel 2008;87:475–81. [40] Aydin ES, Yucel O, Sadikoglu H. Development of a semi-empirical equilibrium model for downdraft gasification systems. Energy 2017;130:86–98. [41] Levenspiel O. Chemical reaction engineering. New York: John Wiley & Sons; 1999. [42] Kunii D, Levenspiel O. Fluidization engineering. New York: John Wiley & Sons Inc.; 1969. [43] Geldart D. Gas fluidization technology. New York: John Wiley & Sons; 1986. [44] Uyar T, Gasification of Afşin–Elbistan Lignite with Air in a Fluidized Bed Reactor, Master Thesis, Ankara University, Ankara, Turkey, 2017: p. 145 (In Turkish).

1056