- Email: [email protected]
Regeneration Kinetics of Spent FCC Catalyst via Coke Gasification in a Micro Fluidized Bed
Available online at www.sciencedirect.com
ScienceDirect Procedia Engineering 102 (2015) 1758 – 1765
The 7th World Congress on Particle Technology (WCPT7)
Regeneration Kinetics of Spent FCC Catalyst via Coke Gasification in a Micro Fluidized Bed Yuming Zhanga,b,*, Guogang Suna, Shiqiu Gaob, Guangwen Xub a State Key Laboratory of Heavy Oil Processing, China University of Petroleum-Beijing, Beijing 102249, China State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China
b
Abstract In the present study, the spent fluid catalytic cracking (FCC) catalyst is regenerated via coke gasification instead of combustion in a micro fluidized bed reactor to investigate its reaction characteristics and kinetic parameters. The reaction rate first increased with carbon conversion ratio and then slowly decreased when reaching the peak. H2 and CO was found to be over 70 vol.% in the gasification gas. Two reaction models, homogenous model (HM) and shrinking core model (SCM) were used to calculate the kinetic parameters of catalyst regeneration, finding that HM had better fitting relevance for the data than SCM. The activation energy from these two models was close to each other, that is, about 150 kJ·mol-1 for the coke gasification over FCC catalyst. © 2014Published The Authors. Published by isElsevier Ltd. article under the CC BY-NC-ND license © 2015 by Elsevier Ltd. This an open access (http://creativecommons.org/licenses/by-nc-nd/4.0/). Selection and peer-review under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Selection and under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Academy Academy ofpeer-review Sciences (CAS). of Sciences (CAS)
Keywords: FCC catalyst; Steam gasification; Micro-fluidized bed; Regeneration kinetics.
1. Introduction Fluid catalytic cracking (FCC) process plays a key role in refinery in terms of converting heavier feedstocks (i.e., vacuum gas oil, atmospheric residue and vacuum residue) into transportation fuels, such as diesel and gasoline[1]. Conventionally, the spent FCC catalyst is reactivated via coke combustion in the fluidized bed regenerator. It will generate excessive heat in the system because of its high coke yield when treating heavy oil. External catalyst cooler
* Corresponding author. Tel./fax.: +86-10-89734820. E-mail address: [email protected] (Y. Zhang)
1877-7058 © 2015 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Selection and peer-review under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Academy of Sciences (CAS)
doi:10.1016/j.proeng.2015.01.312
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1759
or boiler should be applied to extract the heat and maintain the heat balance of the operation, which leads to great waste of carbon resources and high SOx and NOx emissions during coke combustion. On the other hand, hydrogen is usually of great deficiency in the refinery, especially processing heavy oil into light oil products [2]. As a result, the oil cracking combined coke gasification process is proposed, that is, to gasify the deposited coke on the spent catalyst using steam to produce syngas. Previous studies on catalyst regeneration are mainly involved in removing coke via air combustion[3,4], and little publications on FCC catalyst regenerating via coke gasification are found. However, reaction characteristics and kinetics on carbonaceous materials[5,6], such as coal and biomass, have been widely studied using TG analyzer. The samples are confined in a fixed bed and its reaction kinetics is obtained from the weight loss data under the specified reaction atmosphere and temperature program. As a result, TG analysis could hardly reflect the real reaction behavior of the spent FCC catalyst regeneration process in the fluidized bed. The micro fluidized bed reactor analyzer (MFBRA) developed by IPE, CAS, is highly suitable for investigating steam-involved gas-solid reactions [7] and has been used in studying biomass pyrolysis, char gasification and graphite combustion etc.. The fluidization operation could enhance heat and mass transfer and suppress the diffusion effect. Gas products are quickly entrained out of the reactor and measured with an on-line mass spectrometer (MS), further using the data processing software for the reaction kinetics. This study is devoted to investigating the regeneration characteristics and kinetics of FCC catalyst in MFBRA by simulating the possible catalyst regeneration conditions. 2. Experimental section 2.1. Apparatus and operation The schematic diagram of micro fluidized bed reaction system was shown in Figure1, mainly consisting of the gas-supply and steam-generation unit, the fluidized bed reactor, the pulse feeding and the product analysis part. The mixture of argon and steam was used as the fluidizing gas. Steam also served as the gasification reagent during the reaction process. Argon was the purging gas during the interval of each experiment and used as the calibrating gas during the gasification reaction. Partial oxygen would be introduced into the system as the gasification reagent together with steam. The micro fluidized bed reactor was made of quartz tube and had an inner diameter of 20 mm and a total length of about 160 mm. The reactor was divided into three sections by two gas distributors, that is, a preheating part filling with inert Al2O3 balls, a reaction area with fluidizing medium (inert silica sand with particle diameter of 100-150 μm) and a purification part for diminishing fine particles. Silica sand was acid-washed, filtered and calcined to remove the impurities before using as the fluidizing medium. Presssure sensor
Evacuation
Mass spectrometer
Gas cooler Gas filter-drier Water collector
Computer
Fluidizing medium
Sample container
Thermocouple
Eletromagnetic valve
Furnace Mass flow meter
Reactor Mass flow meter Three-way valve
Steam generator Pump
Gas valve
Gas cylinder
Water tank
Fig. 1. Schematic diagram of micro-fluidized bed reactor analyzer.
1760
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
Silica sand of 4 g was put in the reactor as bed materials and about 150 mg coke containing catalyst was set in the pulse-feeding container. When the temperature was reached, the sample was instantly injected into the middle of high-temperature silica sand. The gasification products were quickly stripped out of the reactor by the upward fluidizing gas, then purified and detected by the on-line MS. The composition of the produced gas was monitored until the response lines of MS become stable. Gas composition and concentration was determined using the internal standard method with the calibrating gas. Each experiment was repeated three times to ensure the relative error less than 3% and the average value of three experiments were used to calculate the kinetic data. 2.2. Materials and analysis The coke-containing FCC catalyst was prepared by cracking heavy oil with the fresh catalyst in a fluidized bed, as detailed in our previous publications[8,9]. The coke content on FCC catalyst was about 2.75 wt.%. The main composition and properties of FCC catalyst are given in Table 1. Table 1. Composition and properties of FCC catalyst. XRF analysis of catalysts (wt.%) Components
Al2O3
SiO2
Na2O
Re2O3
Concentration
54.15
37.71
0.25
5.37
Bulk density (kg·m-3)
Sauter Mean diameter (um)
Surface area (m2·g-1)
Pore volume (cm3·g-1)
Average pore diameter (Å)
824.3
62
235.2
0.13
48.7
The carbon content of the spent catalyst was measured with a coke analyzer (CS-344, LECO). The composition of FCC catalyst was determined using the X-ray fluorescence (XRF) spectrometry (AXIOS), and their particle size distribution was determined with the laser particle size analyzer (Malvern Mastersizer 2000). An automatic BET analyzer (Autosorb-1, Quantachrome) was used to measure the specific surface area and pore structure of the catalyst. The MS (PROLINE AMETEK) was adopted to monitor the real-time gas variation during gasification process. 2.3. Data processing The coke on the catalyst was mainly converted into H2, CO, CO2 and CH4 in the gasification reaction, and their corresponding concentration could be calibrated according to the response value of the MS. So the carbon conversion value of coke gasification was defined by calculating the carbon-containing gas species (i.e., CO, CO 2 and CH4) in the syngas, as shown in Eq. (1).
X
³
t t
t 0
³
t tg
t 0
(
FAr u (C CO C CH4 C CO2 )
(
22.4C Ar FAr u (C CO C CH4 C CO2 ) 22.4C Ar
u 12)dt u 100%
(1)
u 12)dt
where X (%) is the conversion ratio, t (min) and tg (min) represent the instant and the end of the reaction time, respectively. FAr (ml·min-1) is the flow rate of argon. CAr, CCO, CCH4 and CCO2 (vol.%) stands for the volume fractions of Ar, CO, CH4 and CO2, respectively. Gasification rate R (min-1) is defined as differential of conversion X to gasification time t,
R
dX dt
(2)
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1761
3. Results and discussion 3.1. Regeneration characteristics of FCC catalyst The effect of external diffusion on gasification reaction was determined at different gas flow rate in the preexperiments. The results showed that the external diffusion effect of gasification could be negligible in the temperature range of 800-950 ć when the flow rate of argon and steam was 200 ml/min and 0.3 g/min, respectively. Figure 2 shows the gas concentration varied with time in the regeneration reaction of coked FCC catalyst at 900 ć. The main gas components for the coke steam gasification were H2, CO, CO2 and CH4, and their corresponding concentration first increased and then slowly decreased to zero at the end of the experiments. However, the time needed for different gas species reaching the peak differed from each other and their release sequence could be proximately summarized as: CH4
Concentration (%)
40
30
H2 CO CO2 CH4
20
10
0 0
2
4
6
18
20
Time (min) Fig. 2. Gas concentration versus time for steam regeneration of spent FCC catalyst at 900ć.
Literature study[10] showed that the properties of coke deposited on the catalyst were different from the condensed coke of coal char or petroleum coke. The coke component on the catalyst usually has the H/C ratio of 0.3-1.0. Cerqueira et al. [11] indicated that there are four main types of coke identified in catalytic cracking of residue oil, that is, reaction or catalytic coke, dehydrogenation or metal coke, Conradson carbon coke and soft coke (i.e., incomplete stripping and entrained hydrocarbons). Methane is not easy to be produced by the carbon steam gasification at low temperature (< 1000 ć), so the relatively high content of CH4 in the syngas is probably from the breakage of condensed nucleus of aromatics due to the coke composition. The cleavage of aromatic compounds is easier to be conducted than the heterogeneous reaction of carbon-steam gasification, thus resulting that CH4 reaches the maximum value shorter than that of other gas species. The main reactions for coke steam gasification are carbon-steam gasification (Eq. 3) and water gas shift (Eq. 4) reaction. Higher temperature will facilitate the steam gasification of coke and meanwhile generate more H2 and CO. After the heterogeneous carbon-steam reaction, CO will further react with steam to produce CO2 via the homogenous reaction (Eq. 4). The reaction procedures and characteristics determine the time sequence of gas emission. The volume percentage of each gas component could be obtained via integrating the area of the MS curve. The results showed that H 2 took about 52 vol.% in the syngas, and the sum of H2 and CO was up to 75 vol.%. This kind of syngas could be potentially used as the hydrogen source for hydroprocessing the liquid oil in the refinery. C+H2OCO+H2 CO+H2O
CO2+H2
ᇞH0=118.9 kJ·mol-1 0
ᇞH =-45.2 kJ·mol
(3) -1
(4)
Figure 3 shows the carbon conversion with time (X-t) and gasification rate with carbon conversion (R-X) in the regeneration of coked FCC catalyst. High gasification temperature facilitated coke-steam gasification and thus shortened the completion time needed for coke gasification. Figure 3a shows that about 45 min was necessary for
1762
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
the coke gasification over FCC catalyst at 800 ć, while the completion time decreased to less than 10 min at 950 ć. For each specified gasification temperature, the gasification rate first increased with carbon conversion ratio and then slowly decreased, and reached the maximum at the carbon conversion ratio of 5-20% (Fig. 3b). Sahimi et al. [12] indicated that the difference of the initial pore volume was the main reason for the occurrence of the maximum gasification rate. In the cracking reaction of heavy oil, coke molecules can be heterogeneously distributed over the surface or diffused into the inner pores of the catalyst particles. At the beginning of gasification, steam first contacts with the coke on the surface of catalyst and the outside of pores. With the on-going of coke steam gasification, most pores would be open via removing coke at the external surface (pore mouth), thus greatly enhanced the gasification rate via increasing the contacting surface between coke and steam. Coke content of the catalyst will gradually decreased with the proceeding of coke-steam gasification, thus resulting in lower gasification rate.
(b)
70
800ć 850ć 900ć 950ć
60
80
50
60
-1
800ć 850ć 900ć 950ć FCC catalyst
40 20 0
R (min )
Conversion X ( %)
(a) 100
40 30 20 10 0
0
10
20
30
40
0
50
Time (min)
20
40
60
80
100
Conversion X (%)
Fig. 3. (a) Carbon conversion with time and (b) gasification rate with conversion of FCC catalyst regeneration at different temperatures.
3.2. Model description of gas solid reaction There is hardly any reaction function model reported specifically on the catalyst regeneration via coke-steam gasification. However, the reaction kinetics of coal char and petroleum coke gasification has been fully studied [13, 14]. The typical reaction models for coke gasification are mainly involved in the shrinking core model and homogenous model, respectively. Shrinking core model (SCM) assumes that the gasification reaction occurs only on the surface of spherical reactant particles, and the un-reacted core would shrink gradually in the reaction process. The reaction order of SCM is 2/3. When controlled by the chemical reaction, SCM could be expressed as, dX dt
k (1 X )2/3
(5)
The integral equation form is, 3[1-(1-X)1/3] = kt
(6)
Homogenous model (HM) makes the hypothesis that the active sites evenly distribute inside the solid particles, and the size of the particles remain to be constant in the reaction process, while only the particle density uniformly changes. The reaction order of HM is 1. When the chemical reaction is the control step, HM could be written as, dX k (1 X ) dt (7) The integral equation form is, ln(1-X) = kt
(8)
where X, t and k are the reaction conversion ratio, reaction time and reaction rate constant, respectively. The reaction rate constant k can be expressed via Arrhenius equation in Eq. (9). Applying logarithm to the Eq. (9) and
1763
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
expanding k leads to Eq. (10)
k = A exp(ln k
E ) RT
ln( A)
(9)
E RT
(10)
where A, E, R and T are the pre-exponential factor, activation energy, gas constant and temperature in K, respectively. The value of lnk is linear with 1/T at a fixed temperature, allowing the determination of activation energy E from the slope of the correlation line and also the pre-exponential factor A. The kinetic parameters of coke gasification on the FCC catalyst were calculated using the SCM and HM, as shown in Fig. 4. The calculated reaction rate constant k and the linear correlation coefficient R2 was presented in Table 2. For all the curves the value of R2 reached 0.90, while the HM had much higher fitting relevance (R2>0.95) than that of SCM. As mentioned above, SCM suggests that the particle size of reactants uniformly decreases during the reaction process, while HM assumes that the particle size is constant in the course of reaction. Coke distributed as a thin layer on the catalyst. The size of catalyst particles hardly changed even removing all the coke during the gasification reaction, which was more close to the assumption of HM. During the kinetic study of petroleum coke gasification in the micro fluidized bed, we observed that the R2 value of SCM was higher than that of HM, which was different from the results for coke gasification over FCC catalyst. The fine char particles pyrolyzed from coal, biomass or heavy oil (i.e., petroleum coke) usually have similar properties (i.e., composition) in the outer surface and inside cores, and will react with gasification reagent on the surface and then shrink uniformly as the reaction proceeds. Apparently, the reaction process of coal char or petroleum coke is more close to the assumption of SCM, thus resulting in higher R2 value of SCM than that of HM. This justified from another perspective that the characteristics of coke gasification could be distinguished from each other in the different reaction systems. (a) 2.0
(b) 2.5 2.0
1.2
800ć 850ć 900ć 950ć Model: SCM FCC catalyst
0.8 0.4 0.0
0
5
10
15
20
25
30
-ln(1-x)
1/3
3[1-(1-x)^ ]
1.6
800ć 850ć 900ć 950ć Model: HM FCC catalyst
1.5 1.0 0.5 0.0
35
Time (min)
0
5
10
15
20
25
Time (min)
30
35
Fig. 4. (a) Shrinking core model and (b) homogeneous model for FCC catalyst regeneration at different temperatures. Table 2. Reaction rate constant k and R2 for FCC catalyst regeneration with SCM and HM. Temperature (ć)
k (min-1)
SCM
HM
800
0.0487
0.9858
0.0706
0.9939
850
0.1072
0.9384
0.1579
0.9796
900
0.2212
0.9137
0.3268
0.9639
950
0.3333
0.9015
0.4966
0.9561
R2
k (min-1)
R2
1764
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
3.3. Regeneration kinetics of FCC catalyst The kinetic parameters were obtained from the correlation of lnk and 1/T according to the Arrhenius equation (Eq. 9), i.e., for the activation energy E and pre-exponential factor A. Figure 5 demonstrates that the reaction data are subjected to a good linear fitting of lnk with 1/T for coke gasification over FCC catalyst using HM, meaning that coke gasification in the examined conditions is mainly controlled by the chemical reaction. The calculated value of E and A was listed in Table 3, and the activation energy from SCM and HM had good repeatability, that is, about 150 kJ·mol-1 for the FCC catalyst regeneration. The differences of pre-exponential factor (A) between SCM and HM are mainly caused by the reaction assumption of these two models. The reaction characteristics for spent FCC catalyst regeneration could be investigated using MFBRA and meanwhile obtained its kinetic parameters, thus to provide some basic data of catalyst regenerating via coke gasification. -0.5 -1.0
lnk
-1.5 -2.0 -2.5 -3.0 0.00081 0.00084 0.00087 0.00090 0.00093 -1
1/T (K ) Fig. 5. Arrhenius equation plot for the kinetic parameters of FCC catalyst regeneration using homogeneous model. Table 3. Kinetic parameters of FCC catalyst regeneration with SCM and HM. Kinetic parameter Reaction model Value
E (kJ·mol-1) SCM 149.41
A (min-1) HM 151.37
SCM 4.31×10
HM 5
7.73×105
4. Conclusions The regeneration characteristics and kinetics of spent FCC catalyst were investigated in a micro-fluidized bed reactor analyzer (MFBRA). Coke gasification with steam over the FCC catalyst could produce high-quality syngas, with the contents of H2+CO up to 70 vol.% in the syngas, and simultaneously for catalyst regeneration. Coke gasification rate R on the catalyst was enhanced at high reaction temperature, and the gasification rate first increased to a peak and then slowly decreased with the increasing of carbon conversion ratio. Shrinking core model (SCM) and homogenous model (HM) were used to describe the coke gasification reaction kinetics of FCC catalyst. It was found that HM had better fitting relevance for the reaction data than that of SCM, and the activation energy obtained from these two models was close to each other. The activation energy of coke gasification for FCC catalyst was about 150 kJ·mol-1. MFBRA could be a useful tool to investigate the reaction kinetics of gas solid reaction, and fundamentally justify the feasibility of spent FCC catalyst regeneration via coke gasification. Acknowledgements The study was conducted with the research programs supported by Science Foundation of China University of Petroleum, Beijing (No. 2462013YJRC021), National Instrumentation Grant Program (2011YQ120039), National Basic Research Program of China (973 Program, NO. 2012CB224801). Acknowledgements are also extended to Mr. Deping Yu for his help on the experiments.
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1765
References [1] R. Sadeghbeigi, Fluid catalytic cracking handbook (3rd edition): An expert guide to the practical operation, design, and optimization of FCC units. Butterworth-Heinemann, Elsevier: UK, 2012. [2] N. Hallale, F. Liu, Refinery hydrogen management for clean fuels Production. Adv. Environ. Res. 6 (2001) 81-98. [3] C. Royo, M. Menendez, J. Santamaria, Kinetics and catalyst regeneration by coke combustion. I. Increased reaction rate due to the presence of chromium. React. Kinet. Catal. Lett. 44 (1991) 445-450. [4] P.B. Weisz, R.D. Goodwin, Combustion of carbonaceous deposits within porous catalyst particles I. Diffusion-controlled kinetics. J. Catal. 5 (1963) 397-404. [5] A. Karimi, N. Semagina, M.R. Gray, Kinetics of catalytic steam gasification of bitumen coke. Fuel 90 (2011) 1285-1291. [6] J.O. Jaber, S.D. Probert, Pyrolysis and gasification kinetics of Jordanian oil-shales. Appl. Energy 63 (1999) 269-286. [7] J. Yu, J. Yue, Z. Liu, L. Dong, G. Xu, J. Zhu, Z. Duan, L. Sun, Kinetics and mechanism of solid reactions in a micro fluidized bed. AIChE J. 11 (2010) 2905-2912. [8] Y. Zhang, D. Yu, W. Li, Y. Wang, S. Gao, G. Xu, Fundamentals of petroleum residue cracking gasification for coproduction of oil and syngas. Ind. Eng. Chem. Res. 51 (2012) 15032-15040. [9] Y. Zhang, D. Yu, W. Li, Y. Wang, S. Gao, G. Xu, H. Zhou, J. Chen, Fundamental study of cracking gasification process for comprehensive utilization of vacuum residue. Appl. Energy 112 (2013) 1318-1325. [10] E. Furimsky, Characterization of cokes from fluid/flexi-coking of heavy feeds. Fuel Process. Technol. 67 (2000) 205-230. [11] H.S. Cerqueira, G. Caeiro, L. Costa, F.R. Ribeiro, Deactivation of FCC catalysts. J. Mol. Catal. A, Chem. 292 (2008) 1-13. [12] M. Sahimi, T.T.Tsotsis, Statistical modeling of gas-solid reaction with pore volume growth: Kinetics regime. Chem. Eng. Sci. 43 (1998) 113-121. [13] L. Zhang, J. Huang, Y. Fang, Y. Wang, Gasification reactivity and kinetics of typical Chinese anthracite chars with steam and CO 2. Energy Fuels 20 (2006) 1201-1210. [14] J. Zou, Z. Zhou, F. Wang, W. Zhang, Z. Dai, H. Liu, Z. Yu, Modeling reaction kinetics of petroleum coke gasification with CO2. Chem. Eng. Process., Process. Intens. 46 (2007) 630-636.
ScienceDirect Procedia Engineering 102 (2015) 1758 – 1765
The 7th World Congress on Particle Technology (WCPT7)
Regeneration Kinetics of Spent FCC Catalyst via Coke Gasification in a Micro Fluidized Bed Yuming Zhanga,b,*, Guogang Suna, Shiqiu Gaob, Guangwen Xub a State Key Laboratory of Heavy Oil Processing, China University of Petroleum-Beijing, Beijing 102249, China State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China
b
Abstract In the present study, the spent fluid catalytic cracking (FCC) catalyst is regenerated via coke gasification instead of combustion in a micro fluidized bed reactor to investigate its reaction characteristics and kinetic parameters. The reaction rate first increased with carbon conversion ratio and then slowly decreased when reaching the peak. H2 and CO was found to be over 70 vol.% in the gasification gas. Two reaction models, homogenous model (HM) and shrinking core model (SCM) were used to calculate the kinetic parameters of catalyst regeneration, finding that HM had better fitting relevance for the data than SCM. The activation energy from these two models was close to each other, that is, about 150 kJ·mol-1 for the coke gasification over FCC catalyst. © 2014Published The Authors. Published by isElsevier Ltd. article under the CC BY-NC-ND license © 2015 by Elsevier Ltd. This an open access (http://creativecommons.org/licenses/by-nc-nd/4.0/). Selection and peer-review under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Selection and under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Academy Academy ofpeer-review Sciences (CAS). of Sciences (CAS)
Keywords: FCC catalyst; Steam gasification; Micro-fluidized bed; Regeneration kinetics.
1. Introduction Fluid catalytic cracking (FCC) process plays a key role in refinery in terms of converting heavier feedstocks (i.e., vacuum gas oil, atmospheric residue and vacuum residue) into transportation fuels, such as diesel and gasoline[1]. Conventionally, the spent FCC catalyst is reactivated via coke combustion in the fluidized bed regenerator. It will generate excessive heat in the system because of its high coke yield when treating heavy oil. External catalyst cooler
* Corresponding author. Tel./fax.: +86-10-89734820. E-mail address: [email protected] (Y. Zhang)
1877-7058 © 2015 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Selection and peer-review under responsibility of Chinese Society of Particuology, Institute of Process Engineering, Chinese Academy of Sciences (CAS)
doi:10.1016/j.proeng.2015.01.312
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1759
or boiler should be applied to extract the heat and maintain the heat balance of the operation, which leads to great waste of carbon resources and high SOx and NOx emissions during coke combustion. On the other hand, hydrogen is usually of great deficiency in the refinery, especially processing heavy oil into light oil products [2]. As a result, the oil cracking combined coke gasification process is proposed, that is, to gasify the deposited coke on the spent catalyst using steam to produce syngas. Previous studies on catalyst regeneration are mainly involved in removing coke via air combustion[3,4], and little publications on FCC catalyst regenerating via coke gasification are found. However, reaction characteristics and kinetics on carbonaceous materials[5,6], such as coal and biomass, have been widely studied using TG analyzer. The samples are confined in a fixed bed and its reaction kinetics is obtained from the weight loss data under the specified reaction atmosphere and temperature program. As a result, TG analysis could hardly reflect the real reaction behavior of the spent FCC catalyst regeneration process in the fluidized bed. The micro fluidized bed reactor analyzer (MFBRA) developed by IPE, CAS, is highly suitable for investigating steam-involved gas-solid reactions [7] and has been used in studying biomass pyrolysis, char gasification and graphite combustion etc.. The fluidization operation could enhance heat and mass transfer and suppress the diffusion effect. Gas products are quickly entrained out of the reactor and measured with an on-line mass spectrometer (MS), further using the data processing software for the reaction kinetics. This study is devoted to investigating the regeneration characteristics and kinetics of FCC catalyst in MFBRA by simulating the possible catalyst regeneration conditions. 2. Experimental section 2.1. Apparatus and operation The schematic diagram of micro fluidized bed reaction system was shown in Figure1, mainly consisting of the gas-supply and steam-generation unit, the fluidized bed reactor, the pulse feeding and the product analysis part. The mixture of argon and steam was used as the fluidizing gas. Steam also served as the gasification reagent during the reaction process. Argon was the purging gas during the interval of each experiment and used as the calibrating gas during the gasification reaction. Partial oxygen would be introduced into the system as the gasification reagent together with steam. The micro fluidized bed reactor was made of quartz tube and had an inner diameter of 20 mm and a total length of about 160 mm. The reactor was divided into three sections by two gas distributors, that is, a preheating part filling with inert Al2O3 balls, a reaction area with fluidizing medium (inert silica sand with particle diameter of 100-150 μm) and a purification part for diminishing fine particles. Silica sand was acid-washed, filtered and calcined to remove the impurities before using as the fluidizing medium. Presssure sensor
Evacuation
Mass spectrometer
Gas cooler Gas filter-drier Water collector
Computer
Fluidizing medium
Sample container
Thermocouple
Eletromagnetic valve
Furnace Mass flow meter
Reactor Mass flow meter Three-way valve
Steam generator Pump
Gas valve
Gas cylinder
Water tank
Fig. 1. Schematic diagram of micro-fluidized bed reactor analyzer.
1760
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
Silica sand of 4 g was put in the reactor as bed materials and about 150 mg coke containing catalyst was set in the pulse-feeding container. When the temperature was reached, the sample was instantly injected into the middle of high-temperature silica sand. The gasification products were quickly stripped out of the reactor by the upward fluidizing gas, then purified and detected by the on-line MS. The composition of the produced gas was monitored until the response lines of MS become stable. Gas composition and concentration was determined using the internal standard method with the calibrating gas. Each experiment was repeated three times to ensure the relative error less than 3% and the average value of three experiments were used to calculate the kinetic data. 2.2. Materials and analysis The coke-containing FCC catalyst was prepared by cracking heavy oil with the fresh catalyst in a fluidized bed, as detailed in our previous publications[8,9]. The coke content on FCC catalyst was about 2.75 wt.%. The main composition and properties of FCC catalyst are given in Table 1. Table 1. Composition and properties of FCC catalyst. XRF analysis of catalysts (wt.%) Components
Al2O3
SiO2
Na2O
Re2O3
Concentration
54.15
37.71
0.25
5.37
Bulk density (kg·m-3)
Sauter Mean diameter (um)
Surface area (m2·g-1)
Pore volume (cm3·g-1)
Average pore diameter (Å)
824.3
62
235.2
0.13
48.7
The carbon content of the spent catalyst was measured with a coke analyzer (CS-344, LECO). The composition of FCC catalyst was determined using the X-ray fluorescence (XRF) spectrometry (AXIOS), and their particle size distribution was determined with the laser particle size analyzer (Malvern Mastersizer 2000). An automatic BET analyzer (Autosorb-1, Quantachrome) was used to measure the specific surface area and pore structure of the catalyst. The MS (PROLINE AMETEK) was adopted to monitor the real-time gas variation during gasification process. 2.3. Data processing The coke on the catalyst was mainly converted into H2, CO, CO2 and CH4 in the gasification reaction, and their corresponding concentration could be calibrated according to the response value of the MS. So the carbon conversion value of coke gasification was defined by calculating the carbon-containing gas species (i.e., CO, CO 2 and CH4) in the syngas, as shown in Eq. (1).
X
³
t t
t 0
³
t tg
t 0
(
FAr u (C CO C CH4 C CO2 )
(
22.4C Ar FAr u (C CO C CH4 C CO2 ) 22.4C Ar
u 12)dt u 100%
(1)
u 12)dt
where X (%) is the conversion ratio, t (min) and tg (min) represent the instant and the end of the reaction time, respectively. FAr (ml·min-1) is the flow rate of argon. CAr, CCO, CCH4 and CCO2 (vol.%) stands for the volume fractions of Ar, CO, CH4 and CO2, respectively. Gasification rate R (min-1) is defined as differential of conversion X to gasification time t,
R
dX dt
(2)
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1761
3. Results and discussion 3.1. Regeneration characteristics of FCC catalyst The effect of external diffusion on gasification reaction was determined at different gas flow rate in the preexperiments. The results showed that the external diffusion effect of gasification could be negligible in the temperature range of 800-950 ć when the flow rate of argon and steam was 200 ml/min and 0.3 g/min, respectively. Figure 2 shows the gas concentration varied with time in the regeneration reaction of coked FCC catalyst at 900 ć. The main gas components for the coke steam gasification were H2, CO, CO2 and CH4, and their corresponding concentration first increased and then slowly decreased to zero at the end of the experiments. However, the time needed for different gas species reaching the peak differed from each other and their release sequence could be proximately summarized as: CH4
Concentration (%)
40
30
H2 CO CO2 CH4
20
10
0 0
2
4
6
18
20
Time (min) Fig. 2. Gas concentration versus time for steam regeneration of spent FCC catalyst at 900ć.
Literature study[10] showed that the properties of coke deposited on the catalyst were different from the condensed coke of coal char or petroleum coke. The coke component on the catalyst usually has the H/C ratio of 0.3-1.0. Cerqueira et al. [11] indicated that there are four main types of coke identified in catalytic cracking of residue oil, that is, reaction or catalytic coke, dehydrogenation or metal coke, Conradson carbon coke and soft coke (i.e., incomplete stripping and entrained hydrocarbons). Methane is not easy to be produced by the carbon steam gasification at low temperature (< 1000 ć), so the relatively high content of CH4 in the syngas is probably from the breakage of condensed nucleus of aromatics due to the coke composition. The cleavage of aromatic compounds is easier to be conducted than the heterogeneous reaction of carbon-steam gasification, thus resulting that CH4 reaches the maximum value shorter than that of other gas species. The main reactions for coke steam gasification are carbon-steam gasification (Eq. 3) and water gas shift (Eq. 4) reaction. Higher temperature will facilitate the steam gasification of coke and meanwhile generate more H2 and CO. After the heterogeneous carbon-steam reaction, CO will further react with steam to produce CO2 via the homogenous reaction (Eq. 4). The reaction procedures and characteristics determine the time sequence of gas emission. The volume percentage of each gas component could be obtained via integrating the area of the MS curve. The results showed that H 2 took about 52 vol.% in the syngas, and the sum of H2 and CO was up to 75 vol.%. This kind of syngas could be potentially used as the hydrogen source for hydroprocessing the liquid oil in the refinery. C+H2OCO+H2 CO+H2O
CO2+H2
ᇞH0=118.9 kJ·mol-1 0
ᇞH =-45.2 kJ·mol
(3) -1
(4)
Figure 3 shows the carbon conversion with time (X-t) and gasification rate with carbon conversion (R-X) in the regeneration of coked FCC catalyst. High gasification temperature facilitated coke-steam gasification and thus shortened the completion time needed for coke gasification. Figure 3a shows that about 45 min was necessary for
1762
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
the coke gasification over FCC catalyst at 800 ć, while the completion time decreased to less than 10 min at 950 ć. For each specified gasification temperature, the gasification rate first increased with carbon conversion ratio and then slowly decreased, and reached the maximum at the carbon conversion ratio of 5-20% (Fig. 3b). Sahimi et al. [12] indicated that the difference of the initial pore volume was the main reason for the occurrence of the maximum gasification rate. In the cracking reaction of heavy oil, coke molecules can be heterogeneously distributed over the surface or diffused into the inner pores of the catalyst particles. At the beginning of gasification, steam first contacts with the coke on the surface of catalyst and the outside of pores. With the on-going of coke steam gasification, most pores would be open via removing coke at the external surface (pore mouth), thus greatly enhanced the gasification rate via increasing the contacting surface between coke and steam. Coke content of the catalyst will gradually decreased with the proceeding of coke-steam gasification, thus resulting in lower gasification rate.
(b)
70
800ć 850ć 900ć 950ć
60
80
50
60
-1
800ć 850ć 900ć 950ć FCC catalyst
40 20 0
R (min )
Conversion X ( %)
(a) 100
40 30 20 10 0
0
10
20
30
40
0
50
Time (min)
20
40
60
80
100
Conversion X (%)
Fig. 3. (a) Carbon conversion with time and (b) gasification rate with conversion of FCC catalyst regeneration at different temperatures.
3.2. Model description of gas solid reaction There is hardly any reaction function model reported specifically on the catalyst regeneration via coke-steam gasification. However, the reaction kinetics of coal char and petroleum coke gasification has been fully studied [13, 14]. The typical reaction models for coke gasification are mainly involved in the shrinking core model and homogenous model, respectively. Shrinking core model (SCM) assumes that the gasification reaction occurs only on the surface of spherical reactant particles, and the un-reacted core would shrink gradually in the reaction process. The reaction order of SCM is 2/3. When controlled by the chemical reaction, SCM could be expressed as, dX dt
k (1 X )2/3
(5)
The integral equation form is, 3[1-(1-X)1/3] = kt
(6)
Homogenous model (HM) makes the hypothesis that the active sites evenly distribute inside the solid particles, and the size of the particles remain to be constant in the reaction process, while only the particle density uniformly changes. The reaction order of HM is 1. When the chemical reaction is the control step, HM could be written as, dX k (1 X ) dt (7) The integral equation form is, ln(1-X) = kt
(8)
where X, t and k are the reaction conversion ratio, reaction time and reaction rate constant, respectively. The reaction rate constant k can be expressed via Arrhenius equation in Eq. (9). Applying logarithm to the Eq. (9) and
1763
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
expanding k leads to Eq. (10)
k = A exp(ln k
E ) RT
ln( A)
(9)
E RT
(10)
where A, E, R and T are the pre-exponential factor, activation energy, gas constant and temperature in K, respectively. The value of lnk is linear with 1/T at a fixed temperature, allowing the determination of activation energy E from the slope of the correlation line and also the pre-exponential factor A. The kinetic parameters of coke gasification on the FCC catalyst were calculated using the SCM and HM, as shown in Fig. 4. The calculated reaction rate constant k and the linear correlation coefficient R2 was presented in Table 2. For all the curves the value of R2 reached 0.90, while the HM had much higher fitting relevance (R2>0.95) than that of SCM. As mentioned above, SCM suggests that the particle size of reactants uniformly decreases during the reaction process, while HM assumes that the particle size is constant in the course of reaction. Coke distributed as a thin layer on the catalyst. The size of catalyst particles hardly changed even removing all the coke during the gasification reaction, which was more close to the assumption of HM. During the kinetic study of petroleum coke gasification in the micro fluidized bed, we observed that the R2 value of SCM was higher than that of HM, which was different from the results for coke gasification over FCC catalyst. The fine char particles pyrolyzed from coal, biomass or heavy oil (i.e., petroleum coke) usually have similar properties (i.e., composition) in the outer surface and inside cores, and will react with gasification reagent on the surface and then shrink uniformly as the reaction proceeds. Apparently, the reaction process of coal char or petroleum coke is more close to the assumption of SCM, thus resulting in higher R2 value of SCM than that of HM. This justified from another perspective that the characteristics of coke gasification could be distinguished from each other in the different reaction systems. (a) 2.0
(b) 2.5 2.0
1.2
800ć 850ć 900ć 950ć Model: SCM FCC catalyst
0.8 0.4 0.0
0
5
10
15
20
25
30
-ln(1-x)
1/3
3[1-(1-x)^ ]
1.6
800ć 850ć 900ć 950ć Model: HM FCC catalyst
1.5 1.0 0.5 0.0
35
Time (min)
0
5
10
15
20
25
Time (min)
30
35
Fig. 4. (a) Shrinking core model and (b) homogeneous model for FCC catalyst regeneration at different temperatures. Table 2. Reaction rate constant k and R2 for FCC catalyst regeneration with SCM and HM. Temperature (ć)
k (min-1)
SCM
HM
800
0.0487
0.9858
0.0706
0.9939
850
0.1072
0.9384
0.1579
0.9796
900
0.2212
0.9137
0.3268
0.9639
950
0.3333
0.9015
0.4966
0.9561
R2
k (min-1)
R2
1764
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
3.3. Regeneration kinetics of FCC catalyst The kinetic parameters were obtained from the correlation of lnk and 1/T according to the Arrhenius equation (Eq. 9), i.e., for the activation energy E and pre-exponential factor A. Figure 5 demonstrates that the reaction data are subjected to a good linear fitting of lnk with 1/T for coke gasification over FCC catalyst using HM, meaning that coke gasification in the examined conditions is mainly controlled by the chemical reaction. The calculated value of E and A was listed in Table 3, and the activation energy from SCM and HM had good repeatability, that is, about 150 kJ·mol-1 for the FCC catalyst regeneration. The differences of pre-exponential factor (A) between SCM and HM are mainly caused by the reaction assumption of these two models. The reaction characteristics for spent FCC catalyst regeneration could be investigated using MFBRA and meanwhile obtained its kinetic parameters, thus to provide some basic data of catalyst regenerating via coke gasification. -0.5 -1.0
lnk
-1.5 -2.0 -2.5 -3.0 0.00081 0.00084 0.00087 0.00090 0.00093 -1
1/T (K ) Fig. 5. Arrhenius equation plot for the kinetic parameters of FCC catalyst regeneration using homogeneous model. Table 3. Kinetic parameters of FCC catalyst regeneration with SCM and HM. Kinetic parameter Reaction model Value
E (kJ·mol-1) SCM 149.41
A (min-1) HM 151.37
SCM 4.31×10
HM 5
7.73×105
4. Conclusions The regeneration characteristics and kinetics of spent FCC catalyst were investigated in a micro-fluidized bed reactor analyzer (MFBRA). Coke gasification with steam over the FCC catalyst could produce high-quality syngas, with the contents of H2+CO up to 70 vol.% in the syngas, and simultaneously for catalyst regeneration. Coke gasification rate R on the catalyst was enhanced at high reaction temperature, and the gasification rate first increased to a peak and then slowly decreased with the increasing of carbon conversion ratio. Shrinking core model (SCM) and homogenous model (HM) were used to describe the coke gasification reaction kinetics of FCC catalyst. It was found that HM had better fitting relevance for the reaction data than that of SCM, and the activation energy obtained from these two models was close to each other. The activation energy of coke gasification for FCC catalyst was about 150 kJ·mol-1. MFBRA could be a useful tool to investigate the reaction kinetics of gas solid reaction, and fundamentally justify the feasibility of spent FCC catalyst regeneration via coke gasification. Acknowledgements The study was conducted with the research programs supported by Science Foundation of China University of Petroleum, Beijing (No. 2462013YJRC021), National Instrumentation Grant Program (2011YQ120039), National Basic Research Program of China (973 Program, NO. 2012CB224801). Acknowledgements are also extended to Mr. Deping Yu for his help on the experiments.
Yuming Zhang et al. / Procedia Engineering 102 (2015) 1758 – 1765
1765
References [1] R. Sadeghbeigi, Fluid catalytic cracking handbook (3rd edition): An expert guide to the practical operation, design, and optimization of FCC units. Butterworth-Heinemann, Elsevier: UK, 2012. [2] N. Hallale, F. Liu, Refinery hydrogen management for clean fuels Production. Adv. Environ. Res. 6 (2001) 81-98. [3] C. Royo, M. Menendez, J. Santamaria, Kinetics and catalyst regeneration by coke combustion. I. Increased reaction rate due to the presence of chromium. React. Kinet. Catal. Lett. 44 (1991) 445-450. [4] P.B. Weisz, R.D. Goodwin, Combustion of carbonaceous deposits within porous catalyst particles I. Diffusion-controlled kinetics. J. Catal. 5 (1963) 397-404. [5] A. Karimi, N. Semagina, M.R. Gray, Kinetics of catalytic steam gasification of bitumen coke. Fuel 90 (2011) 1285-1291. [6] J.O. Jaber, S.D. Probert, Pyrolysis and gasification kinetics of Jordanian oil-shales. Appl. Energy 63 (1999) 269-286. [7] J. Yu, J. Yue, Z. Liu, L. Dong, G. Xu, J. Zhu, Z. Duan, L. Sun, Kinetics and mechanism of solid reactions in a micro fluidized bed. AIChE J. 11 (2010) 2905-2912. [8] Y. Zhang, D. Yu, W. Li, Y. Wang, S. Gao, G. Xu, Fundamentals of petroleum residue cracking gasification for coproduction of oil and syngas. Ind. Eng. Chem. Res. 51 (2012) 15032-15040. [9] Y. Zhang, D. Yu, W. Li, Y. Wang, S. Gao, G. Xu, H. Zhou, J. Chen, Fundamental study of cracking gasification process for comprehensive utilization of vacuum residue. Appl. Energy 112 (2013) 1318-1325. [10] E. Furimsky, Characterization of cokes from fluid/flexi-coking of heavy feeds. Fuel Process. Technol. 67 (2000) 205-230. [11] H.S. Cerqueira, G. Caeiro, L. Costa, F.R. Ribeiro, Deactivation of FCC catalysts. J. Mol. Catal. A, Chem. 292 (2008) 1-13. [12] M. Sahimi, T.T.Tsotsis, Statistical modeling of gas-solid reaction with pore volume growth: Kinetics regime. Chem. Eng. Sci. 43 (1998) 113-121. [13] L. Zhang, J. Huang, Y. Fang, Y. Wang, Gasification reactivity and kinetics of typical Chinese anthracite chars with steam and CO 2. Energy Fuels 20 (2006) 1201-1210. [14] J. Zou, Z. Zhou, F. Wang, W. Zhang, Z. Dai, H. Liu, Z. Yu, Modeling reaction kinetics of petroleum coke gasification with CO2. Chem. Eng. Process., Process. Intens. 46 (2007) 630-636.