Influence of reference temperature on the thermal stress of slag-layer cooling in an atmospheric entrained-flow gasifier with high-speed circulating gasification agent

Applied Thermal Engineering 131 (2018) 446–454

Contents lists available at ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Influence of reference temperature on the thermal stress of slag-layer cooling in an atmospheric entrained-flow gasifier with high-speed circulating gasification agent Haopeng Wang, Zhichao Chen ⇑, Xiaoyan Zhang, Zhengqi Li, Neng Fang, Xiaoying Liu School of Energy Science and Engineering, Harbin Institute of Technology, 92, West Dazhi Street, Harbin 150001, PR China

h i g h l i g h t s
The thermal stress of the slag layer in an atmospheric entrained-flow gasifier is studied.
Creep relaxation is taken into consideration in the numerical model.
A continuous reference temperature is set to acquire a stress-free state at cooling onset.
A more reasonable outcome is obtained with a continuous reference temperature.

a r t i c l e

i n f o

Article history: Received 11 July 2017 Revised 20 November 2017 Accepted 2 December 2017 Available online 5 December 2017 Keywords: Atmospheric entrained-flow gasifier High-speed circulating gasification agent Membrane wall Slag layer Thermal stress

a b s t r a c t A slag slayer can protect the membrane wall in an entrained-flow gasifier. Maintaining a certain thickness of slag layer is particularly important for allowing an entrained-flow gasifier to operate steadily. Thermal stress is a major cause for a slag layer to break. This study aims to provide a numerical model that accurately represents the variation of thermal stress in the cooling slag layer in a novel entrained-flow gasifier. Based on experimental data from an industrial-sized atmospheric entrained-flow gasifier with high-speed circulating gasification agent, the thermal stresses owing to cooling are simulated numerically using transient thermal analysis. Creep relaxation is taken into consideration in the model. A continuous distribution of reference temperature based on that at the onset of cooling is applied to the numerical model. The results indicate that the thermal stresses in the slag layer are tensile during cooling, and that the von Mises stress increases. The von Mises stress appears to peak (at 72 MPa) near the initial liquid–solid interface at the end of the cooling process, which is therefore where the slag layer is most likely to crack or even be shed. The numerical model is also calculated with a fixed distribution of reference temperature by way of comparison. In that case, the von Mises stress in the initial solid slag layer decreases during cooling, which is contrary to common sense. The results show that the reference temperature used in the numerical model is crucial to the calculated thermal stress. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Entrained-flow gasification technology forms part of clean coal technology. Because of its high carbon conversion rate, tar-free raw gas production, and low sensitivity to coal type, it is widely applicable in industrial technology [1–4]. Because an entrained-flow gasifier operates at high temperature (1500 °C), the coal ash melts into liquid slag, most of which is deposited at the membrane wall to form a slag layer that can be divided into two parts: solid slag in contact with the colder refractory, and liquid slag facing ⇑ Corresponding author. E-mail address: [email protected] (Z. Chen). https://doi.org/10.1016/j.applthermaleng.2017.12.013 1359-4311/Ó 2017 Elsevier Ltd. All rights reserved.

the hot syngas [5]. For the slag layer to protect the membrane wall from syngas radiation and liquid-slag erosion, and thereby to allow the gasifier to operate steadily, a certain thickness of slag layer must be maintained [6]. However, many factors can cause the slag layer to crack or even be shed, such as erosion, gravity shedding, or thermal shock [7–9]. Of these, thermal shock is a major source of slag-layer damage in an entrained-flow coal gasifier. Thermal stress occurs because of an uneven thermal expansion between the deposit and the tube metal, or between different layers [7]. Sudden heating or cooling may cause thermal shock of the slag layer, and sudden shrinking caused by the latter can decrease the adhesion force between the slag layer and a tube, causing the slag layer to crack. This is most

447

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

2.4 m

Top view Coal & N2

O2 & H2O

O2 & H2O Sect. A-A

9m A

A

Syngas

Slag layer Stud SiC lining Water tube Fin

Fig. 1. Schematic of atmospheric entrained-flow gasifier with high-speed circulating gasification agent. (The membrane wall contains 160 water tubes, here only for illustration purpose).

Table 1 Fusion temperature of Saimengteer Coal Ash. Fusion temperature (°C) DT

ST

HT

FT

1130

1140

1150

1180

DT, deformation temperature; ST, softening temperature; HT, hemispherical temperature; FT, flow temperature.

likely to happen during cooling, and so thermal stress analysis of this stage is a major research topic [6,10,11]. Much research has been conducted on the thermal and mechanical properties of the slag layer [6,7,10–12]. Because of the high temperatures and pressures in an actual gasifier and the associated limitations of test instruments, numerical simulation is a common way to gain a better understanding of the thermal stress in the slag layer. Using a laboratory-scale gasifier, Lin et al. [6,10,12] investigated the influence of the thickness and porosity of the slag layer on its thermal stress while cooling. Zhou et al. [11] developed a two-dimensional model for simulating the thermal stress in boiler slag under different temperature conditions with static analysis. Because the gasifier geometry and operating conditions dramatically affect the flow and deposition of slag or ash particles [13], different gasifiers results in different thermal stresses in the slag layer. As the temperature of the material when it is free of stress, the reference temperature (Tref) is one of the most important parameters in any numerical simulation to calculate thermal stress. In many reported numerical simulations of slag-layer thermal stress, the value of Tref for the membrane wall and the initial solid slag layer was fixed, for example to 25 °C [6,10]. However, an entrained-flow gasifier is typically run for 6–18 months [6] at high temperature (1500 °C) [14,15], which means that high-temperature creep relaxation of the material becomes an important factor in determin-

ing the stress state. Creep during the high-temperature phase can relax the stress as the material layers grow, resulting in a stressfree state at the end. This phenomenon is usually seen in thermal barrier coatings [16–18] that, like gasifier slag layers, are multilayer structures. Hence, during typical operation of an entrained-flow gasifier, the thermal stress in the membrane goes through three stages: (1) During heating, the thermal stress in the membrane wall increases from zero. To simulate the thermal stress during this process, the whole model can have a fixed value of Tref (i.e., the temperature at the onset of heating, which could be ambient temperature, e.g., 25 °C). (2) During stead operation, creep relaxation of the material gradually dissipates the thermal stress of the membrane wall. After a long time, the membrane wall returns to a stress-free state. (3) During cooling, the thermal stress in the membrane wall again increases from zero. To simulate the thermal stress during this process, the model must somehow acquire a stress-free state at the beginning of the cooling process. Given the physical meaning of Tref, it is necessary to set the temperature distribution of the membrane wall and solid slag layer at the beginning of cooling as a continuous Tref on the model. Hence, a fixed Tref (e.g., ambient temperature, say 25 °C) is unlikely to be suitable for the thermal stress in the membrane wall during cooling after the gasifier has been in operation for a long time. Such an approach would lead to unreasonably exaggerated thermal stress because of the substantial difference between the imposed Tref and the actual temperature. Although entrained-flow gasifiers tend to operate at higher pressures [19,20], atmospheric gasification technology (such as

448

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

Table 2 Composition of Saimengteer Coal Ash. SiO2

Al2O3

Fe2O3

TiO2

CaO

MgO

K2O

NA2O

SO2

P2O5

MnO

30.92

11.91

15.85

0.48

20.82

1.65

0.5

1.98

14.56

0.61

0.18

fixed-bed or fluidized-bed gasifiers) still has a share of the market. An atmospheric entrained-flow gasifier would be an ideal choice if reestablishing a small or medium-sized coal chemical plant because it requires only small investment and is environmentally friendly. Based on this societal need, our research group has developed a novel atmospheric entrained-flow gasifier with high-speed circulating gasification agent [22,23]. To extend the lifetime of the membrane wall, it is important to study how the slag layers breaks, focusing in particular on the thermal stress of the slag layer during cooling. Herein, we study the thermal stress of the slag layer in an atmospheric entrained-flow gasifier with high-speed circulating gasification agent (where the gas production is 35,000 N m3/h). The temperature and thermal stress are analyzed using a threedimensional finite-element model with boundary conditions based on industrially measured temperatures. We impose on this numerical model a continuous Tref distribution based on the temperature distribution at the onset of cooling. To investigate the contribution of Tref, we also analyze the model with fixed Tref.

2. Experimental setup 2.1. Configuration of gasifier Fig. 1 shows a schematic of the atmospheric entrained-flow gasifier with high-speed circulating gasification agent. The operating principle is given in [22,23]. The membrane wall is formed from 160 water tubes distributed evenly at the same height. Adjacent water tubes are connected by welded round steel with a diameter of 8 mm. We selected Saimengteer coal for the industrial test of the gasifier. The fusion temperature and composition of Saimengteer coal ash are given in Tables 1 and 2, respectively. After gasifier shutdown, a smooth dense surface can be seen, as shown in Fig. 2, meaning that a stabilized slag layer was formed, the thickness of which was roughly 10 mm. In this study, the membrane wall is simplified by the local symmetry of the geometric structure. Because the diameter of the gasifier is far greater than the thickness of slag layer, we can ignore the curvature of the surface slag. As shown in Fig. 3, the numerical model consists of a cooling-water tube (carbon steel, Q235), a fin (carbon steel, Q235), a stud (carbon steel, Q235), the SiC lining, and the slag layer. The thermal and mechanical properties of the slag layer are listed in Table 3 [7,10,11,21]. In Fig. 3, two key lines (AB, CD) are marked to show how the membrane wall changes in the depth direction.

2.2. Mathematical model In this study, we used ANSYS Mechanical APDL (ANSYS Parametric Design Language) for the finite-element analysis. The static temperature distribution was calculated first and then used as an initial condition to calculate the thermal stress. Because heat transfer is complicated in an atmospheric entrained-flow gasifier with high-speed circulating gasification agent, we made the following assumptions [6,10]:

Fig. 2. Surface of slag layer.

(1) All materials properties are assumed to be isotropic, and the contact thermal resistance between any two materials is neglected. (2) Slag hotter than the fusion temperature of Saimengteer coal is regarded as a liquid with zero thermal stress. For the linings and the solid slag layer, only elastic deformation is considered, with no cracks generated during cooling. (3) The temperature drops sharply after the gasifier is shut down. This causes the viscosity to increase rapidly once the temperature drops below that of the critical viscosity, ending the flow of the liquid slag. Consequently, any change in the thickness of the slag layer is neglected. The boundary conditions for calculating the temperature distribution are shown in Fig. 3. In the gasifier, heat is transferred between the slag-layer surface and the syngas mainly by means of radiation and convection. However, such heat transfer is usually difficult to calculate because of the complicated conditions inside the gasifier. Therefore, for the thermal load, we used temperature data about the slag surface as determined by a thermocouple in the membrane wall passing through the SiC lining by 10 mm. The variation of the temperature of the slag layer during cooling is shown in Fig. 4. The membrane wall is cooled by the natural circulation of highpressure water. The coefficient h of convective heat transfer in the membrane wall [10] is given by

k h ¼ 0:023 Re0:8 Pr 0:4 f ; d f

ð1Þ

where k is the thermal conductivity of the fluid, d is the inner diameter of the tube, and Ref and Prf are the Reynolds and Prandtl numbers, respectively, of the fluid. The coefficients of heat transfer in the membrane wall and steel surface [10] are given by

8 h ¼ hc þ hr > h
 < t
T 4 i T 4 a hr ¼ eC 0 100  100 =ðT  T a Þ; > : hc ¼ 1:35ðT  T a Þ

ð2Þ

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

where hc is the convection heat-transfer coefficient, ht is the total heat-transfer coefficient, hr is the radiation heat-transfer coefficient, C0 is the blackbody radiation coefficient, T is the absolute temperature of the wall surface, and Ta is the absolute temperature of the air. During the calculation, the liquid region is determined by the temperature distribution above the fusion temperature, and the thermal stress of the liquid zone is estimated as free. The function of element birth and death is used [10], which sets the matrix multiplier to 106, thereby killing the liquid element. When the liquid slag solidifies, these elements are then activated with a value of Tref of 1180 °C (the fusion temperature of Saimengteer coal). For the initial solid slag, the stub, and the water tube, we set a Tref distribution based on that at the onset of cooling.

2.3. Validation The geometric model was meshed in ANSYS Mechanical APDL with an adaptive mesh. Because the present study is aimed mainly

449

at the distribution of thermal stress in the slag, we refined the mesh in the slag layer to achieve results of a better accuracy. The variation of temperature with distance L along the centerline of the geometric model is shown in Fig. 5 for different mesh sizes. The position L = 10 is at the interface between the slag and the refractory material. The result shows the model containing 34,653 nodes has similar results with the others. Because the version of the models with 74,511 and 40,423 nodes would impose an excessive computational load, we chose to work with the version with 34,653 nodes instead. To verify the model of thermal and thermal stress used in the present study, we show in a comparison of the volume fraction of the liquid slag between our simulation results and those of Lin et al. [10] (Fig. 6(a)) and a comparison of the von Mises stress between our simulation results and those of Lin et al. [6] (Fig. 6 (b)). Given the good agreements, we reason that the present model can be used to simulate the temperature and thermal stress of the slag layer.

Fig. 3. Three-dimensional model of membrane wall and slag layer.

450

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

Table 3 Thermal and mechanical properties of slag. Properties

Temperature T (°C)

Conductivity k (Wm1K1) Young’s modulus E (GPa) Poisson ratio l Expansion coefficient a (K1106) Specific heat c (Jkg1K1) Density q (kgm3)

300

600

900

1100

1300

1.03 9.7 0.25 6.0 879 2700

1.71 9.4

2.40 9.1

2.85 8.9

3.31

934

971

996

1047

1600

3.2. Von Mises stress distribution in slag layer

1400 1200

o

T/ C

1000 800 600 400 200

0

1000

2000

t/s

3000

4000

5000

Fig. 4. Variation of slag-layer temperature during cooling.

34653 40423 74511

1400

1000

0

T/ C

1200

800 600 400 0

2

4

6 L / mm

8

10

Fig. 5. Temperature along centerline of geometric model for different mesh sizes.

3. Results and discussion 3.1. Temperature distribution Fig. 7 shows the temperature distribution in the slag layer and membrane wall at t = 0 s. The temperature is highest at the firefacing side. There is a strong temperature difference (600 °C) across the slag layer because it has a lower thermal conductivity than others in the membrane wall. The volume fraction of liquid slag is estimated according to the temperature distribution, and the simulation results are plotted in Fig. 8. There is approximately 40% liquid slag by volume at the onset of cooling; this molten slag then solidifies, causing the liquid volume fraction to decrease with decreasing temperature and therefore with increasing time. At t = 240 s, the temperature of the slag surface drops below 1180 °C, whereupon the molten slag solidifies completely.

The boundary conditions used in calculating the thermal stress are shown in Fig. 3(b). In this study, we compare continuous and fixed (i.e., 25 °C) distributions of Tref along lines AB and CD in Fig. 3. As shown in Fig. 9, the continuous Tref distribution approximates to a certain extent the temperature distribution at the onset of during cooling, whereas the fixed Tref distribution comes nowhere near. We use the von Mises stress to analyze the thermal stress of the slag layer. In Fig. 10, we show planar (Z = 20 mm) distributions of von Mises stress at t = 0, 120, 1200, and 4800 s. At each of those times, we find completely different distributions and variations of von Mises stress according to whether continuous or fixed Tref is used. In the case of continuous Tref, there is no distribution of thermal stress at the onset of cooling. As the temperature decreases, so gradually does the thermal stress increase and the liquid slag solidify. With this decrease in temperature, the thermal stresses increase in both the initial solid slag and the solidified slag, with the latter being higher than the former. For the model with fixed Tref, differences can be seen between the initial solid slag and the solidified liquid slag, with the von Mises stress being discontinuous at the initial interface. For the initial liquid slag layer, the von Mises stress shows an upward trend due to the tensile stress generated as the solidified slag volume shrinks during cooling. However, for the initial solid slag layer, the von Mises stress shows a downward trend with decreasing temperature, which is contrary to common sense. The strength of the slag layer under actual conditions cannot be acquired easily through experiments. Lin et al. [12] found the maximum von Mises stress is 165 MPa at the crack tip of the slag layer during cooling. Therefore, the slag layer cannot crack under a von Mises stress lower than 165 MPa. Because Lin et al. used similar gasifier operating conditions during cooling as those used in the present study, these data provide a reference value for this work. As shown in Fig. 10(d), at t = 4800 s the maximum von Mises stress of the slag layer is nearly 80 MPa. This is far lower than 165 MPa; therefore, in the condition calculated, the slag layer appears to have no possibility of cracking. However, the slag layer is a complex, heterogeneous and porous material [6,7], and some areas of relative low-strength exist. When the thermal stress becomes relatively large in these low-strength positions, the slag layer can crack or even be shed. Without the protection of the slag layer from syngas radiation and liquid-slag erosion, the lifetime of the membrane wall will be reduced. As shown in Fig. 10, the thermal stress of the total slag layer continues to increase due to the increasing difference between the temperature at the onset of cooling and the actual temperature during cooling. Increasing the final cooling temperature helps to keep the slag layer intact, so as to extend the lifetime of the membrane wall. Slowing the rate of temperature decrease or adopting stepped cooling can decrease or even eliminate the thermal stress

451

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

180

0.3

160 Von Mises stress / MPa

Fraction of liquid slag

Calculated Data Data from [10] 0.2

0.1

0.0

140

0 s, 2D, data from [6] 0 s, 3D, calculated data 3480 s, 2D, data from [6] 3480 s, 3D, calculated data

60 40 20 0

0

30

60 Time / s

90

120

0

1

2 3 Thickness / mm

(a)

4

5

(b)

Fig. 6. Comparisons of the volume fraction of the liquid slag and von Mises stress results. (a) Comparison of the volume fraction of the liquid slag. (b) Comparison of the von Mises stress of the slag layer.

Temp/K

Fraction of liquid slag

0.5 0.4 0.3 0.2 0.1 0.0

0

Tcv=1180 C

250

C-D

1000

O

1000

200

liquid slag o Tcv=1180 C Solid slag

1200

o

Temperature/ C

O

Temperature/ C

Solid slag

150

1400

A-B

1200

t/s

In Fig. 11(a), the first (r1 ) and third (r3 ) principal stresses are positive for the model with continuous Tref, which means that the thermal stress is mainly tensile. This accords with the principle that most solid material shrink as they cool, generating tensile stress. The stress r1 is divided into two sections by a peak (72 MPa) near the initial liquid–solid interface. This indicates that the largest thermal stress appears near the initial liquid–solid interface, which is the place most likely to crack.

in the slag layer by providing enough time to creep. This also protects the slag layer and extends the lifetime of the membrane wall. The time point t = 4800 s is selected as the end of the cooling process; the temperature of the slag surface is then only around 300 °C. The thermal stress distributions for continuous and fixed Tref at that time are shown in Fig. 11, where distinct differences can be seen especially for the initial solid slag.

Liquid slag

100

Fig. 8. Volume fraction of liquid slag.

Fig. 7. Temperature distribution inside atmospheric entrained-flow gasifier with high-speed circulating gasification agent at t = 0 s.

1400

50

800 600

TEM at t=0 s Continous Tref Fixed Tref

400 200 0

0

10

20

X/mm

30

40

800 600 400

TEM at t=0 s Continous Tref Fixed Tref

200 0

0

10

20

Fig. 9. Distributions of Tref along lines AB and CD in Fig. 3.

30

X/mm

40

50

60

452

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

100

Von Mises Stress(MPa)

Von Mises Stress(MPa)

100 80 60 40 20 0 25

80 60 40 20 0 25

20

20

15 Y( 10 m m 5 )

0

6

4

2

-5 0

0

X(mm )

-5

Continuous Tref

4

2

0

10

8

6 X(mm)

Fixed Tref

(a) t = 0 s 100

Von Mises Stress(MPa)

100

Von Mises Stress(MPa)

15 Y( 10 m m 5 )

10

8

80 60 40 20 0 25

80 60 40 20 0 25

20

20

15 Y( 10 m m 5 ) 0

8

6

4

2

-5 0

15 Y( 10 m 5 m )

10

X(mm)

0 -5

4

2

0

Continuous Tref

10

8

6 X(mm)

Fixed Tref (b) t = 120 s 100

Von Mises Stress(MPa)

Von Mises Stress(MPa)

100 80 60 40 20 0 25 20 15 Y( m 10 m ) 5 0 -5

4

2

0

8

6

80 60 40 20 0 25 20 15 Y( 10 m m 5 )

10

0 -5

X(mm)

4

2

0

8

6

10

X(mm)

Fixed Tref

Continuous Tref (c) t = 1200 s 100

Von Mises Stress(MPa)

Von Mises Stress(MPa)

100 80 60 40 20 0 25

80 60 40 20 0 25

20

20

15 Y( 10 m m 5 )

0 -5

0

2

4

X(mm)

6

8

10

15 Y( 10 m m 5 ) 0 -5

Continuous Tref

0

2

4

6 X(mm)

Fixed Tref (d) t = 4800 s

Fig. 10. von Mises stress distributions on plane (Z = 20 mm) at t = 0, 120, 1200, and 4800 s.

8

10

453

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

80 70

80

Continuous Tref σ1 σ3

a

60

Fixed Tref σ1 σ3

60 50

Stress/MPa

50

Stress/MPa

b

70

40 Initial liquid slag Initial solid slag

30 20

40 initial liquid slag

30

initial solid slag

20 10

10 0

0

-10

-10

0

2

4

6

8

10

0

2

4

Fig. 11. Thermal stress distributions for (a) continuous and (b) fixed T

In Fig. 11(b), for the model with fixed Tref, the thermal stress is again mainly tensile in the solidified slag layer. However, it acts as a compressive stress in the initial solid slag layer (r1  0, r3 < 0), which is contrary to common sense. 4. Conclusions The thermal stress of the slag layer in an atmospheric entrained-flow gasifier with high-speed circulating gasification agent (with a gas production of 35,000 N m3/h) was simulated. A continuous Tref distribution based on the temperature distribution at the onset of cooling was set for the numerical model to consider creep relaxation. The numerical model was also calculated with fixed Tref by way of comparison. The temperature and thermal stress of the slag layer were investigated. The conclusions are as follows: (1) The distributions and variations of von Mises stress of the models with continuous and fixed Tref were fundamentally different at fixed times. For continuous Tref, there was no thermal-stress distribution at the onset of cooling. As the temperature decreased, the thermal stress increased in both the initial solid slag and the solidified slag. However, for the model with fixed Tref, an unreasonably large discontinuity in von Mises stress was obtained in the initial solid slag layer at the onset of cooling. The thermal stresses in the initial solid slag and the solidified slag showed different variation mechanisms. The von Mises stress increased in the solidified slag layer during cooling but decreased in the initial solid slag layer, the latter being contrary to common sense. (2) The stress state of the slag layer as indicated by the r1 and r3 stresses differed fundamentally according to whether continuous or fixed Tref was applied. For the model with continuous Tref, the thermal stress of the whole model was tensile during cooling. This accorded with the principle that most solid materials shrink when cooled, thereby generating tensile stress. The r1 stress was divided into two sections by a peak (72 MPa) near the initial liquid–solid interface, indicating where was most likely to crack. However, for the model with fixed Tref, the thermal stress of the whole model was different on either side of the initial liquid–solid interface. As in the model with continuous Tref, the initial liquid slag layer showed a tensile stress distribution. However, the initial solid slag layer showed a compressive stress distribution, which was contrary to common sense. Overall, the results indicated that the model with fixed Tref was not a reasonable model of the cooling process.

6

8

10

X/mm

X/mm ref

at t = 4800 s.

Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant Nos. 51406043 and 51576055), the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 51421063), and the National Quality Inspection of Public Welfare Scientific Research Project (Grant No. 201510067). References [1] H. Watanabe, M. Otaka, Numerical simulation of coal gasification in entrained flow coal gasifier, Fuel 85 (2006) 1935–1943, https://doi.org/10.1016/ j.fuel.2006.02.002. [2] H.J. Jeong, D.K. Seo, J. Hwang, CFD modeling for coal size effect on coal gasification in a two-stage commercial entrained-bed gasifier with an improved char gasification model, Appl. Energy 123 (2014) 29–36, https:// doi.org/10.1016/j.apenergy.2014.02.026. [3] L. Kong, J. Bai, W. Li, X. Wen, X. Li, Z. Bai, Z. Guo, H. Li, The internal and external factor on coal ash slag viscosity at high temperatures, Part 1: Effect of cooling rate on slag viscosity, measured continuously, Fuel 158 (2014) 968–975, https://doi.org/10.1016/j.fuel.2015.02.055. [4] L. Wang, Y.J. Jia, S. Kumar, R. Li, R.B. Mahar, M. Ali, I.N. Unar, U. Sultan, K. Memon, Numerical analysis on the influential factors of coal gasification performance in two-stage entrained flow gasifier, Appl. Therm. Eng. 112 (2017) 1601–1611, https://doi.org/10.1016/j.applthermaleng.2016.10.122. [5] I. Ye, C. Ryu, J.H. Koo, Influence of critical viscosity and its temperature on the slag behavior on the wall of an entrained coal gasifier, Appl. Therm. Eng. 87 (2015) 175–184, https://doi.org/10.1016/j.applthermaleng.2015.05.027. [6] W. Lin, Q. Liang, G. Yu, H. Liu, X. Gong, Numerical modeling for non-steady thermal stress analysis of slag layer in a membrane wall entrained-flow gasifier, Fuel 90 (2011) 2396–2403, https://doi.org/10.1016/j.fuel.2011.02.040. [7] A. Zbogar, F. Frandsen, P.A. Jensen, P. Glarborg, Shedding of ash deposits, Prog. Energy Combust. Sci. 35 (2009) 31–56, https://doi.org/10.1016/j. pecs.2008.07.001. [8] H. Zhou, H. Zhang, L. Li, B. Zhou, Ash deposit shedding during Co-combustion of coal and rice hull using a digital image technique in a pilot-scale furnace, Energy Fuels 27 (2013) 7126–7137, https://doi.org/10.1021/ef401814y. [9] N. Sato, S. Ueno, D.E. Priyanto, E. Ohno, Y. Matsuzawa, Y. Ueki, R. Yoshiie, I. Naruse, Effects of coal type on growth and shedding of ash deposit in pulverized coal combustors, Energy Fuels 30 (2016) 6059–6069, https://doi. org/10.1021/acs.energyfuels.5b02836. [10] W. Lin, Q. Liang, H. Liu, X. Gong, G. Yu, Study on the temperature and thermal stress of the slag layer cooling process in a membrane wall entrained-flow gasifier, Energy Fuels 25 (2011) 2579–2586, https://doi.org/10.1021/ ef200245c. [11] J. Zhou et al., Chem. Ind. Eng. 54 (2003) 1678–1682 (in Chinese). [12] W. Lin, Q. Liang, H. Liu, X. Gong, G. Yu, Study on the thermal stress of slag layer with cracks and the crack extension in a membrane wall entrained-flow gasifier, Energy Fuels 26 (2012) 518–523, https://doi.org/10.1021/ef2013024. [13] J. Ni, Q. Liang, Z. Zhou, Z. Dai, G. Yu, Numerical and experimental investigations on gas-particle flow behaviors of the Opposed Multi-Burner Gasifier, Energy Convers. Manage. 50 (2009) 3035–3044, https://doi.org/10.1016/j. enconman.2009.07.023. [14] Y. Wei, H. Li, N. Yamada, A. Sato, Y. Ninomiya, K. Honma, T. Tanosaki, A microscopic study of the precipitation of metallic iron in slag from iron-rich coal during high temperature gasification, Fuel 103 (2013) 101–110, https:// doi.org/10.1016/j.fuel.2011.09.024.

454

H. Wang et al. / Applied Thermal Engineering 131 (2018) 446–454

[15] I. Ye, C. Ryu, Numerical modeling of slag flow and heat transfer on the wall of an entrained coal gasifier, Fuel 150 (2015) 64–74, https://doi.org/10.1016/ j.fuel.2015.01.111. [16] J. Rösler, M. Bäker, K. Aufzug, A parametric study of the stress state of thermal barrier coatings, Acta Materialia 52 (2004) 4809–4817, https://doi.org/ 10.1016/j.actamat.2004.06.046. [17] M. Bäker, J. Rösler, G. Heinze, A parametric study of the stress state of thermal barrier coatings. Part II: Cooling stresses, Acta Materialia 53 (2005) 469–476, https://doi.org/10.1016/j.actamat.2004.10.004. [18] L. Ajdelsztajn, D. Hulbert, A. Mukherjee, J.M. Schoenung, Creep deformation mechanism of cryomilled NiCrAlY bond coat material, Surface Coat. Technol. 201 (2007) 9462–9467, https://doi.org/10.1016/j.surfcoat.2007.03.054. [19] T.F. Wall, G.S. Liu, H.W. Wu, D.G. Roberts, K.E. Benfell, S. Gupta, J.A. Lucas, D.J. Harris, The effects of pressure on coal reactions during pulverised coal combustion and gasification, Prog. Energy Combust. Sci. 28 (2002) 405–433, https://doi.org/10.1016/S0360-1285(02)00007-2.

[20] D.G. Roberts, D.J. Harris, Char gasification with O2, CO2, and H2O: effects of pressure on intrinsic reaction kinetics, Energy Fuels 14 (2000) 483–489, https://doi.org/10.1021/ef9901894. [21] A. Zbogar, F.J. Frandsen, P.A. Jensen, P. Glarborg, Heat transfer in ash deposits: a modelling tool-box, Prog. Energy Combust. Sci. 31 (2005) 371–421, https://doi. org/10.1016/j.pecs.2005.08.002. [22] Z. Li, et al., Vergasungsvorrichtung für Kohlepulver mit mehrstufiger Zuführung von Vergasungsmittel und starker Verwirbelung sowie entsprechendes Vergasungsverfahren, Germany, DE112015000112T5 [P], 2016-16-26. [23] Z. Li, et al., Intensified rotating coal dust gasification apparatus having multistage gasification agent provision and gasification method: WO Patent, PCT/ CN2015/091286. https://patentscope.wipo.int/search/en/detail.jsf?docId= WO2017041338&recNum=1&maxRec=&office=&prevFilter=&sortOption=& queryString=&tab=PCT+Biblio#atapta0.