Fuel 220 (2018) 233–239

Contents lists available at ScienceDirect

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

Full Length Article

A new prediction method for the viscosity of the molten coal slag. Part 2: T
The viscosity model of crystalline slag
⁎
Jie Zhoua,b, Zhongjie Shena,b, Qinfeng Lianga,b,c, Jianliang Xua,b, Haifeng Liua,b,
a
Key Laboratory of Coal Gasification and Energy Chemical Engineering of Ministry of Education, East China University of Science and Technology, P.O. Box 272,
Shanghai 200237, PR China
b
Shanghai Engineering Research Center of Coal Gasification, East China University of Science and Technology, P.O. Box 272, Shanghai 200237, PR China
c
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: The complex chemical composition of coal slag separated out different types and shapes of crystals during the
Coal cooling process. The morphology of the crystal in the molten slag influenced the viscosity of the coal slag. The
Crystalline slag aim of this paper was to obtain the viscosity model of crystalline slag using a new calculation method. The model
Crystal morphology of the suspension viscosity obtained by the previous study was used to predict the viscosity of molten slag. The
Viscosity model
correction factor (β) was introduced in the model of the suspension viscosity. The solid phase volume fraction
(φ) in the molten slag was calculated by FactSage. The liquid phase viscosity (η0) was fitted by the viscosities of
molten slag in high temperature section. Ten kinds of crystalline slags were applied in this study to prove the
accuracy of the modified viscosity model. The viscosity of molten slag predicted by the model (CSM) agreed with
experiment data.

1. Introduction Xuan et al. [7–10] studied the influences of CaO, Fe2O3, SiO2/Al2O3 on
crystallization characteristics of synthetic coal slags. With the increase
Entrained flow gasification technology is a type of clean coal con- of CaO, crystallization of the slag became significant, especially in those
version technology, which plays an important role in reducing carbon with a calcium range between 15% and 35%. The crystallization tem-
dioxide emissions and improving energy efficiency. It allows for various perature increased but only slightly. With a higher ratio of Fe2O3, more
combinations of electricity, liquid fuels, hydrogen, chemicals and heat crystallization heat was released and the crystallization shifted to a
with the characters of high efficiency and fuel flexibility [1]. Entrained higher temperature, potentially leading to a higher Tcv in viscosity. The
flow gasifier usually operates at high temperature (above the ash flow kinetics under isothermal 1100 °C showed that the growth rate of
temperature) to ensure a suitable slagging condition for the stable op- crystals increased with the addition of iron oxide. As the S/A ratio in-
eration [2,3]. The viscosity of molten slag is the key factor in de- creased in the range from 1.5 to 3.5, energy barrier was significantly
termining whether the slagging condition is smooth and stable. The lowered which increased the crystallization ratio. Shen et al. [11] stu-
viscosity of molten slag exhibited a rapid increase because of the in- died the effect of cooling process on the generation and growth of
crease of crystals in the molten slag when temperature was lowered crystals in coal slag. The variation of the cooling process obviously
below a certain temperature which was referred as the temperature of affected the crystallization behavior of molten slag. Low cooling rate
critical viscosity (Tcv) [4,5]. benefited the generation of the crystals and long residence time below
Many factors affect the viscosity of molten slag, including compo- the initial crystallization temperature promoted the generation and
sition, cooling rate, residence time and so on. And these factors influ- growth of crystal. Louhich et al. [12] reproduced the experimental T
ence the viscosity of molten slag by changing the volume fraction of dependence of the crystallization temperature with numerical calcula-
crystal phase in the molten slag. A number of scholars have studied the tions based on standard models for the nucleation and growth of hard-
factors that affect the crystallization of slag. Fredericci et al. [6] con- sphere crystals, classical nucleation theory and the Johnson-Mehl-Av-
sidered the crystallization mechanism of an unaltered blast-furnace slag rami-Kolmogorov theory. These results suggested that deep analogies
composition. They suggested that most crystallization was on surface existed between hard-sphere colloidal crystals and pluronics micellar
and this suggestion was confirmed by the study of nucleation kinetics. crystals, in spite of the difference in particle softness. Studies about

⁎
Corresponding author at: Key Laboratory of Coal Gasification and Energy Chemical Engineering of Ministry of Education, East China University of Science and Technology, P.O. Box
272, Shanghai 200237, PR China.
E-mail address: hfl[email protected] (H. Liu).

https://doi.org/10.1016/j.fuel.2018.01.056
Received 9 September 2017; Received in revised form 30 November 2017; Accepted 16 January 2018
0016-2361/ © 2018 Elsevier Ltd. All rights reserved.
J. Zhou et al. Fuel 220 (2018) 233–239

glassy slag are relatively rich in the glass industry. Karamanov et al. Table 1
[13] summarized results of the crystallization of iron-rich glasses. The Chemical composition of coals in the experiment.
results indicated that magnetite and pyroxene were the main crystal
Composition/wt% SiO2 CaO Al2O3 Fe2O3 MgO Na2O TiO2
phases and that the kinetics of pyroxene formation could be explained
as growth on a fixed number of magnetite nuclei. Pacurariu et al. [14] SF 37.65 26.16 19.69 11.89 0.96 2.62 1.03
studied non-isothermal crystallization kinetics of some glass-ceramics YL 27.20 28.12 25.44 16.65 1.07 0.77 0.48
JJT 32.19 25.91 19.85 16.12 3.12 2.31 0.49
with pyroxene structure. Many other scholars [15–19] have also studied
YH 26.66 22.27 20.31 24.98 3.01 2.37 0.41
the effect of other factors on the slag crystallization process including BD 45.23 36.09 9.51 4.55 1.40 1.10 2.11
residual carbon, trace elements and so on. NM 48.85 16.22 13.28 15.88 3.42 1.64 0.71
About the viscosity of molten slag, Kondratiev et al. [20], Ilyush- TX 50.26 30.42 1.72 11.82 0.56 2.17 3.05
echkin et al. [21] and Zhang et al. [22] studied the characters of XJW 50.10 30.67 1.53 11.95 0.56 2.16 3.04
BS 39.31 23.89 9.65 21.20 3.23 1.29 1.45
flowability and Tcv of coal ash slag. Kong et al. [23–25] proposed the
ZX 55.23 21.97 0.20 19.24 0.55 2.04 0.78
internal and external factors influencing the slag viscosity at high
temperature, including CaO content, residual carbon and cooling rate.
It was found that the viscosities of both the glassy and the crystalline
slag were declining when the cooling rate increased. However, when
the temperature was lower than the flow temperature, the impact of the
cooling rate on crystal slag was more obvious. At the same time, Tcv also
decreased with increasing cooling rate. Residual carbon in the slag was
the internal factor affecting the slag viscosity. With the content of re-
sidual carbon increasing, the slag viscosity also increased. When the
content of residual carbon was more than 5%, a significant impact on
the slag viscosity appeared. They also studied the effect of CaO content
on the viscosity curve. When the CaO content increased, the slag visc-
osity increased, which exhibited the same properties as the viscosity of
crystal slag versus S/A ratio. Other documents also mentioned the in-
fluence of silicon, aluminum and other trace elements on the slag
viscosity. Zhang et al. [26] found that the viscosity of CaO-SiO2-Al2O3-
CaF2 system increased monotonously with gradually increasing sub-
stitution content of Al2O3 for SiO2. Feng et al. [27] produced that the
viscosity of slag decreased with the increase of TiO2 content. Wang
et al. [28] found that the viscosity of liquid slag decreased with the
addition of V2O5 increasing at high temperature. Ilyushechkin et al.
[29] studied the viscosity of high-iron slags from Australian coal.
Folkedahl et al. [30] considered the effect of atmosphere on the visc-
osity of selected bituminous and low-rank coal ash slags. Under the air
atmosphere, the slag viscosity was related to the mole ratio of basic to
acidic oxides. However, the changes in viscosity observed in using
hydrogen instead of air atmosphere appeared to be related to the
amount of iron which was originally in the slag.
Kondratiev et al. [31], Hosseini et al. [32], and Ramacciotti et al.
[33] have proposed models to predict the slag viscosity. These models
were based on the method of simplifying the synthesis of oxides, so
there were great limitations in the accuracy of prediction. Browing
et al. [34] proposed a classical prediction method for the viscosity of
molten slag which exhibited Newtonian fluid properties. Fulcher et al.
[35] proposed the viscosity prediction model based on temperature Fig. 1. Schematic diagram of the calculation method.
correction factor by using Arrhenius equation:
B Reid [41] suggested that the viscosity of the slag at a given tem-
logη = A + . perature can be expressed as an expression of SiO2 content. Watt et al.
T −T0 (1)
[42] studied the viscosity of homogeneous liquid slag in relation to slag
where A is a constant, B is the activation energy, and T0 is the tem- composition. The equation was given by
perature correction factor.
107m
Urbain et al. [36] established the semi-empirical model to predict logη = + c,
the viscosity of Al2O3-CaO-FeO-SiO2 system according to Weymann- (T −150)2 (4)
Frenkel equation [37], which was the most widely used in recent years. where η is in poise, and T is in °C. Here
η = aTe1000b / T , (2) m = 0.00835SiO2 + 0.00601Al2 O3−0.109, (5)
and and
ln(a) = −0.2963b−11.6725. (3) c = 0.0415SiO2 + 0.0192Al2 O3 + 0.0276(equiv. Fe2 O3) + 0.0160CaO−3.92,

In the equation above, T is the slag temperature, the unit is K. a and (6)
b are the empirical constants, b is related to the slag composition, as a and slag components are expressed on weight percentages.
function of the slag system components. Many other researchers, such It can be seen from the literatures that bulk of irregular crystals
as Kalmanovitch and Frank [38], Kondratiev and Jak [39] and Hurst generated in the crystalline slag during the crystallization process
et al. [40] had also modified the Urbain model. which improved the difficulty in predicting the viscosity of the

234
J. Zhou et al. Fuel 220 (2018) 233–239

Fig. 2. Variation of solid phase volume fraction of different coal slags with temperature.

crystalline slag, especially the viscosity near Tcv. Therefore, a new consists of the following parts: Brookfield DVIII rheometer, high tem-
model (CSM) for calculating the viscosity of crystalline slag was ob- perature heating furnace, gas circulation system, water cooling system,
tained by using the relationship between the suspension viscosity and temperature control system and data acquisition system. The DilaSoft32
the solid phase volume fraction in this paper. Correction parameter (β) software on the computer controls and records the temperature data of
was introduced to modify the equation. The liquid phase viscosity (η0) the furnace. The high temperature heating furnace of the viscometer
was fitted by the viscosity of molten slag in high temperature section. was heated by the silicon aluminum rod, whose maximum temperature
Ten kinds of crystalline coal samples were used to verify the accuracy of could reach to 1700 °C. The height and diameter of the cylindrical
the model. corundum crucible were 120 and 30 mm. The height and diameter of
the spindle were 16 and 12 mm. Before the measurement, the rhe-
2. Experiment ometer should be calibrated through the standard silicone oil, and the
measurement error was controlled within 1%. During the viscosity test,
2.1. Experiment sample approximately 45 g of coal ash was placed in a ceramic corundum
crucible and fixed by a corundum bracket of the high temperature
Ten kinds of crystalline coal samples were used to prepare the coal furnace. The entire heating zone and reaction zone were sealed by
ash in the N17/HR-K muffle furnace (Nabertherm Company, Lilienthal, metal flanges and vacuum evacuated for 5 min using a vacuum pump.
Germany), according to the Chinese standard GB/T1574-2007. In the The gas CO/CO2 (molar ratio 60/40) was then fed from the bottom of
experiment of coal ash preparation, the temperature first rose from the furnace and the gas flow rate was 100 ml/min to create a reducing
room temperature to 500 °C within 30 min and then was held for an- atmosphere. The gas was removed from the top of the furnace in order
other 30 min to remove the water of crystallization. Then the tem- that the sample was in a reducing atmosphere. The ash sample was
perature was heated to 815 °C within 30 min and held for about heated from room temperature to a temperature higher than the ash
120 min for the complete reaction of coal and air. The aim of the pro- melting temperature of 200 °C at a rate of 5 °C/min in the heating
cess was to obtain coal ash samples with inorganic minerals. The coal furnace. The entire rotor was then slowly put down below the slag level
ash sample was analyzed by Advant'X Intellipower 3600 X-ray fluor- and the temperature should be kept at a constant value for 30 min to
escence (Thermo Fisher Scientific, America) [11]. The ash compositions measure the slag viscosity smoothly.
of the ten coal samples are shown in Table 1.
3. Results and discussion
2.2. Experiment equipment and method
3.1. Viscosity model
2.2.1. High temperature viscosity measurement
Viscosity measurement instrument is RV DVIII type high tempera- It can be seen from the previous studies that the morphology of the
ture rotary viscometer produced by Theta in America [17]. It mainly crystals in the molten slag was mainly acicular and occupied a large

235
J. Zhou et al. Fuel 220 (2018) 233–239

Fig. 3. Variation of viscosity with high temperature range of different coal slags.

Table 2 viscosities at different temperatures. And the viscosity of the high

Viscosity equations of high temperature section of different coals. temperature section obtained by the two methods were fitted as the
liquid phase viscosity (η0) of the molten slag. The liquid phase viscosity
Sample η0c η0e
obtained by the experiment data was denoted as η0e, and the liquid
SF 4.261 × 108e−0.0142T 3.831 × 105e−0.0089T phase viscosity calculated by FactSage was denoted as η0c. 3. FactSage
YL 4.302 × 10 4e−0.0076T 3.436 × 107e−0.0115T was used to calculate the solid phase volume fraction (φ) under the
JJT 7.115 × 103e−0.0064T 1.556 × 103e−0.0055T composition mentioned above, where the volume fraction was instead
YH 1.192 × 103e−0.0059T 1.068 × 106e−0.0091T of the mass fraction. 4. According to the conclusion given by Shen et al.
[11] on the changes of crystal shape and aspect ratio during the crys-
tallization process. It can be concluded that the average crystal size was
volume fraction during the slag crystallization process. A number of 50 μm, the average crystal aspect ratio was 10 and the constant C took
crystals had great influence on the viscosity of molten slag during 1.235. The Eq. (8) was used to obtain the correction factor β with the
crystallization process. Through the previous study, a modified model value of 1.19. 5. The corresponding temperature and the solid phase
of the suspension viscosity respecting to the solid phase volume fraction volume fraction were substituted into Eq. (7) to calculate the viscosity
was obtained as follows: at different temperatures and finally get the viscosity-temperature
curves.
η = η0 (1−βφ)−2.5, (7)

and
3.2. Model verification
β = 0.9672Ce−0.0022de 0.0126(θ − 1), (8)
3.2.1. Changes of the solid phase volume fraction with the temperature
where η0 was Newtonian viscosity of the suspending liquid, β was the In this paper, the viscosities of ten typical crystalline coal slags were
correction factor, d was the particle size, and θ was the particle aspect calculated by Eq. (7). The ash compositions of the ten coal samples are
ratio. C was a constant related to the shape of the particle. When the shown in Table 1.
particles were spherical, C was 1. When the particles were non-sphe- Fig. 2 shows the changes of solid phase volume fraction with tem-
rical, C was 1.235. Therefore, Eq. (7) was used to calculate the slag perature of different coal slags (SF, YL, JJT and YH). It can be seen from
viscosity. The schematic diagram of the calculation method is shown in the Fig. 2 that with the decrease of temperature, the solid phase volume
Fig. 1. fraction in the molten slag increased, and when the temperature
The calculation steps are as follows: 1. X-ray Fluorescence (XRF) reached a certain temperature, the solid phase volume fraction in-
was used to analyze the different coal ashes, including the main com- creased rapidly. This phenomenon indicated that the increment of
ponents of the oxide composition and content. 2. High temperature crystal was violent at this time. For SF, YL, JJT and YH, the tempera-
rotary viscometer and FactSage were used to calculate the slag tures were about 1230 °C, 1280 °C, 1310 °C and 1315 °C respectively.

236
J. Zhou et al. Fuel 220 (2018) 233–239

Fig. 4. Viscosity-temperature curves of different coal samples.

Table 3 was reasonable to use the viscosity data of the high temperature range
Comparison of characteristic temperature of different coals. to fit the equation of the liquid phase viscosity. The fitting equations
obtained from different coal samples were shown in Table 2. Among
Sample T25/°C △T25-c/°C △T25-e/°C Tcv/°C △Tcv-c/°C △Tcv-e/°C
them, η0c was the viscosity of the slag in the high temperature section
SF 1220 25 23 1228 15 13 calculated by FactSage, η0e was the viscosity of the slag in the high
YL 1307 −34 −12 1326 −35 −15 temperature section obtained by the experimental data. And T was the
JJT 1163 38 29 1158 50 46
slag temperature in units of K. Although the viscosity of slag in the high
YH 1282 34 6 1285 34 19
BD 1372 5 71 1380 10 66
temperature section obtained by the two methods had a certain dif-
NM 1286 −14 −60 1290 5 10 ference, the difference of absolute viscosity was small. The viscosity
TX 1375 60 35 1395 20 10 temperature curves of different coals were obtained by the liquid phase
XJW 1455 −29 −31 1470 −33 −35 viscosity in the high temperature section of two methods.
BS 1295 15 16 1298 16 17
ZX 1364 41 16 1380 35 30
Average 14.1 9.3 11.7 16.7 3.2.3. Viscosity prediction
error The comparison of the viscosity predicted by model with the ex-
periment data is shown in Fig. 4. The black curve represented the
changes of viscosity measured by experiment, and the pink curve re-
The abrupt change point of the solid phase volume fraction had great presented the changes of viscosity calculated by FactSage. The red
influence on Tcv. When the temperature was further decreased, the in- curve (FCE) showed the changes of viscosity whose liquid phase visc-
crease rate of the solid phase volume fraction decreased, indicating that osity was obtained from the experiment data, and the blue curve (FCF)
the growth of crystal in the slag was basically complete. Shen et al. [11] indicated that the changes of viscosity whose liquid phase viscosity was
found that the number of crystals increased with the increase of the obtained by the FactSage. It can be seen from the Fig. 4 that the visc-
residence time. The influence of residence time was not considered by osity calculated by the model was consistent with the trend of viscosity
FactSage. Therefore, the solid phase volume fraction calculated by measured by the actual measurement and the difference in the value
FactSage had a little difference with the experimental value at the same was small. In the actual project, the requirement of the slag viscosity
temperature. was below 25 Pa.s. So 25 Pa.s was chosen for the critical point for
analysis in this paper. The comparison of the characteristic tempera-
3.2.2. The viscosity of molten slags at high temperature tures of the different coals samples is shown in Table 3. When the
The changes of liquid phase viscosity with temperature in different viscosity was 25 Pa.s, the temperature difference between the actual
high temperature range of molten slag were shown in Fig. 3. In Fig. 3, value and the model whose liquid phase viscosity was calculated from
slag viscosity increased with the temperature decreasing, but the rise of FactSage (△T25-c) were 25, −34, 16 and 34 °C respectively. The average
the viscosity was small. This phenomenon indicated that no crystal was error of △T25-c was 14.1 °C. The temperatures difference between the
formed in slag and the slag was still in the molten state. Therefore, it actual value and the model whose liquid phase viscosity was calculated

237
J. Zhou et al. Fuel 220 (2018) 233–239

Fig. 5. Comparison of the measured viscosities predicted by the proposed methods for different coal samples.

from experiment data (△T25-e) were 23, −12, 14 and 6 °C respectively. authors’ model agreed with the experiment data. The Urbain model and
The average error of △T25-e was 9.3 °C. For the critical viscosity point the Watt and Fereday correlation expressed poor coincidence with the
(Tcv), the temperature difference between the experimental data and experiment data. Comparing the results predicted by the three methods,
the model whose liquid phase viscosity was calculated from the Fact- the authors’ model in this paper and the Watt and Fereday correlation
Sage (△Tcv-c) were 15, −35, 15 and 34 °C respectively. The average agreed with the experiment data in the high temperature section. The
error of △Tcv-c was 11.7 °C. The temperature difference between the Urbain model was different with the experiment data in this range.
actual value and the model whose liquid phase viscosity was obtained When the temperature was further decreased, the prediction of the Watt
by the experimental data (△Tcv-e) were 13, −15, 13 and 19 °C respec- and Fereday correlation appeared obvious difference with the experi-
tively. The average error of △Tcv-e was 16.1 °C. In conclusion, the ment data. The Urbain model and the authors’ model were in line with
average error of characteristic temperatures of different coal samples the experiment data. Tcv of the two classical models also had a certain
obtained by the model was below 20 °C. The results illustrated that the gap with the experiment data. With the point of 25 Pa.s, the prediction
model had advantages in predicting the viscosity of crystalline slags. temperatures of the two classical models had a larger difference with
However, it can also be seen from the Fig. 4 that the viscosity value the actual temperature. According to the comparison, it can be con-
calculated by FactSage differed from the viscosity value measured by cluded that the model obtained by authors had advantages in predicting
experiment. The difference was mainly due to the fact that the calcu- the viscosity of crystalline slag.
lation of FactSage was based on the phase equilibrium principle. This
principle was different from the actual situation. Thus little difference 4. Conclusion
in viscosity was existed.
In the process of slag crystallization, the crystal morphology
3.3. Model comparison changed greatly. Most of the crystals showed acicular shape, and the
size of the crystals was relatively large. A small part of the crystals had
To verify the accuracy of the model in this paper, two empirical the performance of square. The morphology of the crystals in the slag
models were chosen to compare with the viscosity data measured from had a great influence on the slag viscosity. The viscosity of crystalline
the experiment. The first method was produced by Urbain [36] which slag containing a large amount of crystals was difficult to predict.
was probably the most commonly used method for predicting the slag Therefore, to accurately predict the viscosity change during the crys-
viscosity. The second method was commonly known as the Watt and tallization process of the molten slag, a new method for predicting the
Fereday correlation [42]. viscosity of the crystalline slag (CSM) was obtained by using the model
Comparison of the measured viscosity predicted by Urbain model, of the suspension viscosity given in the previous study. The solid phase
Watt and Fereday correlation and the authors’ model for different kinds volume fraction was calculated on the basis of FactSage. The model was
of coals in this study is shown in Fig. 5. On the whole, it can be seen verified by ten typical crystalline slags. It was found that the viscosity
from the Fig. 5, with the decrease of the temperature, the viscosity of calculated by the model agreed with the experiment data, and the ab-
the coal slag increased sharply. The increasing trend of viscosity of the solute error was small which proved the model was available for

238
J. Zhou et al. Fuel 220 (2018) 233–239

predicting the viscosity of crystalline slags. However, a little difference crystallization process and crystal compositions of iron-rich coal slag. Energy Fuels
between the experiment data and the FactSage calculation results in the 2015;29:3640–8.
[18] Trasi NS, Baird JA, Kestur US, Taylor LS. Factors influencing crystal growth rates
fitting of the viscosity at high temperature section was existed. from undercooled liquids of pharmaceutical compounds. J Phys Chem B
Therefore, the problem of determining the liquid phase viscosity and 2014;118:9974–82.
solid phase volume fraction should be solved in the future work. [19] Burkhard DJM. Nucleation and growth rates of pyroxene, plagioclase, and Fe-Ti
oxides in basalt under atmospheric conditions. Eur J Mineral 2005;17:675–86.
[20] Kondratiev A, Jak E. Predicting coal ash slag flow characteristics (viscosity model
Acknowledgments for the Al2O3–CaO–‘FeO’–SiO2 system). Fuel 2001;80:1989–2000.
[21] Ilyushechkin AY, Hla SS, Roberts DG, Kinaev NN. The effect of solids and phase
compositions on viscosity behaviour and TCV of slags from Australian bituminous
This study was supported by the National Natural Science coals. J Non-Cryst Solids 2011;357:893–902.
Foundation of China (U1402272), the Foundation of Shanghai Science [22] Zhang GH, Zhen YL, Chou KC. Influence of TiC on the viscosity of
and Technology Committee (14dz1200100), the National Nature CaO–MgO–Al2O3–SiO2–TiC suspension system. ISIJ Int 2015;55:922–7.
[23] Kong LX, Bai J, Li W, Wen XD, Li XM, Bai ZQ, et al. The internal and external factor
Science Foundation of China (21376082), and the Foundation of State
on coal ash slag viscosity at high temperatures, Part 1: effect of cooling rate on slag
Key Laboratory of Coal Conversion (Grant No. J16-17-301). viscosity, measured continuously. Fuel 2015;158:968–75.
[24] Kong LX, Bai J, Li W, Wen XD, Liu XC, Li XM, et al. The internal and external factor
References on coal ash slag viscosity at high temperatures, Part 2: effect of residual carbon on
slag viscosity. Fuel 2015;158:976–82.
[25] Kong LX, Bai J, Li W, Wen XD, Li XM, Bai Z, et al. The internal and external factor
[1] Yamashita K, Barreto L. Energyplexes for the 21st century: coal gasification for co- on coal ash slag viscosity at high temperatures, Part 3: effect of CaO on the pattern
producing hydrogen, electricity and liquid fuels. Energy 2005;30:2453–73. of viscosity–temperature curves of slag. Fuel 2016;179:10–6.
[2] Gong X, Lu WX, Guo XL, Dai ZH, Liang QF, Liu HF, et al. Pilot-scale comparison [26] Zhang Gh, Zhen Yl, Chou KC. Viscosity and structure changes of CaO-SiO2-Al2O3-
investigation of different entrained-flow gasification technologies and prediction on CaF2 Melts with substituting Al2O3 for SiO2. J Iron Steel Res Int 2016;23:633–7.
industrial-scale gasification performance. Fuel 2014;129:37–44. [27] Feng C, Chu MS, Tang J, Qin J, Li F, Liu ZG. Effects of MgO and TiO2 on the viscous
[3] Krishnamoorthy V, Pisupati S. A critical review of mineral matter related issues behaviors and phase compositions of titanium-bearing slag. Int J Miner Metall
during gasification of coal in fixed, fluidized, and entrained flow gasifiers. Energies Mater 2016;23:868–80.
2015;8:10430–63. [28] Wang ZG, Bai J, Kong LX, Wen XD, Li XM, Bai ZQ, et al. Viscosity of coal ash slag
[4] Yuan HF, Liang QF, Gong X. Crystallization of coal ash slags at high temperatures containing vanadium and nickel. Fuel Process Technol 2015;136:25–33.
and effects on the viscosity. Energy Fuels 2012;26:3717–22. [29] Ilyushechkin AY, Hla SS. Viscosity of high-iron slags from Australian coals. Energy
[5] Oh MS, Brooker DD, De Paz EF, Brady JJ, Decker TR. Effect of crystalline phase Fuels 2013;27:3736–42.
formation on coal slag viscosity. Fuel Process Technol 1995;44:191–9. [30] Folkedahl BC, Schobert HH. Effects of atmosphere on viscosity of selected bitumi-
[6] Fredericci C, Zanotto ED, Ziemath EC. Crystallization mechanism and properties of nous and low-rank coal ash slags. Energy Fuels 2005;19:208–15.
a blast furnace slag glass. J Non-Cryst Solids 2000;273:64–75. [31] Kondratiev A, Jak E. Modeling of viscosities of the partly crystallized slags in the
[7] Xuan WW, Whitty KJ, Guan QL, Bi DP, Zhan ZH, Zhang JS. Influence of CaO on Al2O3-CaO-“FeO”-SiO2 system. Metall Mater Trans B 2001;32:1027–32.
crystallization characteristics of synthetic coal slags. Energy Fuels [32] Hosseini M, Ghader S. A model for temperature and particle volume fraction effect
2014;28:6627–34. on nanofluid viscosity. J Mol Liq 2010;153:139–45.
[8] Xuan WW, Whitty KJ, Guan QL, Bi DP, Zhan ZH, Zhang JS. Influence of Fe2O3 and [33] Ramacciotti M, Journeau C, Sudreau F, Cognet G. Viscosity models for corium
atmosphere on crystallization characteristics of synthetic coal slags. Energy Fuels melts. Nucl Eng Des 2001;204:377–89.
2015;29:405–12. [34] Browning GJ, Bryant GW, Hurst HJ, Lucas JA, Wall TF. An empirical method for the
[9] Xuan WW, Whitty KJ, Guan QL, Bi DP, Zhan ZH, Zhang JS. Influence of SiO2/Al2O3 prediction of coal ash slag viscosity. Energy Fuels 2003;17:731–7.
on crystallization characteristics of synthetic coal slags. Fuel 2015;144:103–10. [35] Fulcher GS. Analysis of recent measurements of the viscosity of glasses. J Am Ceram
[10] Xuan WW, Zhang JS, Xia DH. Crystallization characteristics of a coal slag and in- Soc 1925;8:339–55.
fluence of crystals on the sharp increase of viscosity. Fuel 2016;176:102–9. [36] Urbain G. Viscosity of silicate melts. Trans J BrCeram Soc 1981;80:139.
[11] Shen ZJ, Li RX, Liang QF, Xu JL, Liu HF. Effect of cooling process on the generation [37] Weymann HD. On the hole theory of viscosity, compressibility, and expansivity of
and growth of crystals in coal slag. Energy Fuels 2016;30:5167–73. liquids. Colloid Polym Sci 1962;181:131–7.
[12] Louhichi A, Tamborini E, Ghofraniha N, Caton F, Roux D, Oberdisse J, et al. [38] Kalmanovitch DP, Frank M. An effective model of viscosity for ash deposition
Nucleation and growth of micellar polycrystals under time-dependent volume phenomena. Miner Matter Ash Deposition Coal 1988:89–101.
fraction conditions. Phys Rev E 2013:87. [39] Kondratiev A, Jak E. Review of experimental data and modeling of the viscosities of
[13] Karamanov A, Pelino M. Crystallization phenomena in iron-rich glasses. J Non- fully liquid slags in the Al2O3-CaO-‘FeO’-SiO2 system. Metall Mater Trans B
Cryst Solids 2001;281:139–51. 2001;32:1015–25.
[14] Păcurariu C, Lazău I. Non-isothermal crystallization kinetics of some glass-ceramics [40] Hurst H, Novak F, Patterson J. Viscosity measurements and empirical predictions
with pyroxene structure. J Non-Cryst Solids 2012;358:3332–7. for fluxed Australian bituminous coal ashes. Fuel 1999;78:1831–40.
[15] Xu J, Zhao F, Guo QH, Yu GS, Liu X, Wang FC. Characterization of the melting [41] Reid WT. External Corrosion and Deposits: Boilers and Gas Turbines. American
behavior of high-temperature and low-temperature ashes. Fuel Process Technol Elsevier Publishing; 1971.
2015;134:441–8. [42] Watt JD, Fereday F. Flow properties of slags formed from ashes of british coals. 1.
[16] Gan L, Zhang CX, Zhou JC, Shangguan FQ. Continuous cooling crystallization ki- Viscosity of homogeneous liquid slags in relation to slag composition. Journal of the
netics of a molten blast furnace slag. J Non-Cryst Solids 2012;358:20–4. Institute of. Fuel 1969;42:99.
[17] Shen ZJ, Liang QF, Zhang BB, Xu JL, Liu HF. Effect of continuous cooling on the

239