Viscosity of High-Iron Slags from Australian Coals - Energy & Fuels

---

Article pubs.acs.org/EF

Viscosity of High-Iron Slags from Australian Coals Alexander Y. Ilyushechkin* and San Shwe Hla Commonwealth Scientific and Industrial Research Organisation (CSIRO) Energy Technology, Post Office Box 883, Pullenvale, Queensland 4069, Australia ABSTRACT: There is a need to characterize slag flow properties for continuous operation of entrained flow gasifiers at the expected slag tapping temperature range. Slag flow characteristics include the slag phase composition, slag viscosity, and temperature of critical viscosity (TCV). Previous studies of viscosity behavior and viscosity modeling for slags from Australian coal ashes include compositions with up to 15 wt % FeO. This work investigates phase compositions and viscosity behavior of highiron slags (16.5−21 wt % FeO) from Australian coal ashes, their blends, and fluxed mixtures. The investigated slag compositions are selected from mullite, spinel, and anorthite (feldspar) primary phase fields. A series of viscosity measurements were made over the range of 1200−1600 °C using laboratory-produced slags, and TCV and viscosity contours are presented. The viscosity results are also modeled by a modified Urbain treatment, and two separate sets of polynomial coefficients are generated for evaluating viscosities as a function of the temperature and slag composition.

■

INTRODUCTION

content because it can be as rich as 30 wt % (as Fe2O3) in Australian coal ashes.4 Experimental viscosity data on high-iron coal slags are limited.3,8,9 In the present work, we extend experimental work on the Australian coal slag characteristics to slags with a high iron content (16.5−21 wt % FeO). Developing an accurate mathematical model for predicting slag viscosity as a function of the temperature and composition is in high demand, because producing viscosity data for high-temperature molten slag is experimentally difficult, expensive, and time-consuming. There are some theoretical models for estimating viscosity directly related to the composition of the slag mixture, but according to Vargas et al.,10 these models usually fall back on empirical testing of their performance. Hurst et al.5,6 applied a modified Urbain treatment to predict the viscosity for the quaternary SiO2−Al2O3−CaO−FeO system at 2.5−10 wt % FeO. They extended their work at 15 wt % FeO, and the same modeling approach was used to fit the viscosity measurements with ash composition.7 Because our study is a continuation of their work, an empirical Urbain style viscosity model based on slag compositions was used for the prediction of slag viscosities with high iron contents (16.5−21 wt % FeO).

Entrained flow gasifiers are slagging gasifiers, requiring the mineral matter in the feedstock to be melted. Smooth operation of such gasifiers strongly depends upon steady and reliable removal of slag. Viscosity is one of the important characteristics of slag flow behavior in terms of the applicability of coal for entrained flow gasification. To achieve steady slag tapping in entrained-flow gasifiers, it is generally considered that slag viscosity should be within the range of 5−25 Pa s at operating temperatures of 1200−1500 °C.1 Slag viscosity strongly depends upon slag composition, which reflects the parent coal mineral matter. It may differ from coal ashes because of partitioning of chemical components of the mineral matter in the coal between the gasification streams and reactions with gasifier wall slags.2,3 However, as a first approximation, coal ash compositions are typically used to assess slag flow characteristics, including slag viscosity and ash fusion temperature. The Australian coal ashes consist of four major components, such as silica, alumina, ferric oxide, and calcium oxide, which usually comprise about 90−95 wt % of the coal ashes. The concentration of these elements may vary significantly, as listed in Table 1.4 The suitability of Australian bituminous coals for entrained flow gasification has been previously studied for different coal ash compositions, for synthetic slag compositions, and gasifier slags containing 0−15 wt % FeO.5−7 There is a need to characterize viscosity of the slags with higher iron

■

Ash and Slag Samples. Several coal ashes, their blends, and representative synthetic oxide mixtures were used in this study. Compositions of the 23 samples used in the present work are given in Table 2. The investigated ash compositions represent coal with a high iron content, having ∼18−23 wt % as Fe2O3, which is equivalent to ∼16.5−21 wt % FeO in slag compositions. The listed compositions exclude some components that represent less than 1 wt % of the ash and were normalized to 100%. Artificial coal ashes were prepared by mixing laboratory- or analytical-grade Al2O3, CaO, Fe2O3, K2CO3, MgO, Na2CO3, S, SiO2, and TiO2 powders. Real coal ashes were prepared by ashing of the coal at 750 °C for 20−30 h.

Table 1. Range of Major Components of Australian Coal Ashes4 major oxides SiO2 Al2O3 CaO Fe2O3 MgO

wt % from from from from from

28 to >80 8 to 48 0 to 28 0 to 32 0 to 10

© 2013 American Chemical Society

EXPERIMENTAL SECTION

Received: April 4, 2013 Revised: June 20, 2013 Published: June 21, 2013 3736

dx.doi.org/10.1021/ef400593k | Energy Fuels 2013, 27, 3736−3742

Energy & Fuels

Article

Table 2. Slag Compositions Used in Present Study slag

SiO2

Al2O3

CaO

Fe2O3

MgO

K2O

Na2O

TiO2

P2O5

SO3

Mn3O4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

35.20 50.42 49.61 50.73 38.42 33.81 34.03 32.17 47.39 46.80 47.64 47.44 45.44 44.43 40.46 43.24 45.63 48.80 34.32 42.26 41.05 37.34 36.93

20.93 21.17 20.68 20.23 30.78 30.80 27.09 29.26 20.48 15.75 22.17 16.07 25.20 27.97 24.98 21.44 15.22 15.55 28.44 30.16 18.10 20.41 28.31

17.52 2.85 4.93 6.43 5.03 10.01 10.03 10.06 3.64 7.35 4.56 5.16 4.93 5.02 10.87 11.59 12.85 10.83 15.41 5.00 15.00 16.34 9.70

20.00 19.09 18.42 18.28 22.55 22.54 22.59 22.57 22.16 20.73 20.07 23.12 20.45 19.20 19.87 20.44 21.82 21.26 18.60 19.57 20.60 21.89 21.99

2.00 2.05 2.42 1.32 0.66 0.58 1.71 1.62 2.26 2.68 1.78 2.36 1.44 1.09 1.60 1.67 1.90 1.30 1.26 0.99 1.93 2.10 1.84

0.30 0.49 0.34 0.69 0.08 0.07 0.83 0.79 0.53 0.57 0.52 0.64 0.37 0.28 0.49 0.34 0.35 0.35 0.48 0.24 0.33 0.22 0.45

0.91 0.20 1.49 0.10 0.09 0.08 0.17 0.16 0.68 1.24 1.26 0.98 0.86 0.64 0.47 0.46 0.72 0.74 0.27 0.47 1.30 0.45 0.24

0.86 1.37 1.09 1.64 1.30 1.14 0.42 0.40 1.21 1.04 1.09 1.11 1.16 1.20 0.82 0.54 0.74 0.80 0.79 1.07 0.67 0.87 0.34

1.41 2.13 0.01 0.30 0.04 0.03 0.45 0.43 1.50 1.05 0.70 0.91 0.01 0.00 0.06 0.02 0.01 0.16 0.10 0.06 0.07 0.04 0.02

0.34 0.05 0.05 0.13 0.89 0.79 2.58 2.45 0.04 2.68 0.17 2.12 0.06 0.05 0.12 0.08 0.21 0.14 0.21 0.07 0.50 0.29 0.12

0.52 0.18 0.98 0.14 0.17 0.16 0.09 0.09 0.10 0.12 0.05 0.10 0.08 0.12 0.28 0.18 0.54 0.07 0.12 0.12 0.46 0.05 0.06

■

Slag samples were obtained after viscosity measurements by cold rod quenching at a fixed temperature for investigation of slag phase compositions. Cold rod quenching involves dipping a Mo rod in the molten slag, removing it, and then cooling it in water. Slag Viscosity. Slag viscosity measurements were performed by a Haake high-temperature viscometer using a rotational bob (molybdenum, 12.5 mm in diameter) supported by a 3 mm diameter molybdenum rod. Ashes or slags (approximately 100 g) were placed in a molybdenum crucible (32 mm in diameter and 100 mm in height) and heated up to 1600 °C in a nitrogen atmosphere. The crucible was placed in a sacrificial graphite sleeve to consume free oxygen in the constant nitrogen gas flow. Oxygen partial pressure (pO2) in the system has been measured separately using an oxygen probe (HT oxygen probe, Australian Oxytrol Systems). At temperatures between 1200 and 1450 °C, pO2 values near the slag surface were between 2.8 × 10−11 and 2.9 × 10−9 atm, indicating the reducing nature of the experimental conditions. The sample temperature was measured with a B-type thermocouple placed near the sample location. Viscosity measurements were conducted at incremental temperature reduction steps of 25−30 °C within the temperature range of 1200− 1600 °C. At least 30 min intervals were taken to equilibrate temperatures and compositions inside the crucible. At each step, shear stress is measured with varying shear rate (5−15 s−1) to detect non-Newtonian behavior, and the slag viscosity was measured several times to ensure the repeatability of measurements. Slag Analysis. Bulk compositions of samples were determined by X-ray fluorescence. Quenched slags were examined using scanning electron microscopy (SEM) in backscattering mode. Energy-dispersive X-ray spectroscopy (EDS) was undertaken to identify the elemental composition of selected regions of the slag. Electron probe microanalysis (EPMA) was used to identify the compositions of solid and liquid phases of quenched samples. The slag phase compositions corresponding to the temperature range of 1200−1600 °C were also calculated using models based on the FactSage thermodynamic package. The FactSage 6.2 equilibrium and phase diagram modules were used with the FACT53 and FToxid databases.

RESULTS AND DISCUSSION Slag compositions were selected to represent a specific range of Australian coals with high iron content. Viscosity behavior of slags was investigated above and below the liquidus temperatures and covers the temperature range of 1200−1600 °C. Slag Composition. The compositions of slag used in this work are normalized to four major components and plotted on a ternary-phase diagram in four-component system Al2O3− CaO−FeO−SiO2 at a normalized content of FeO (20 wt %), as illustrated in Figure 1.

Figure 1. Compositions of slags illustrated in four-component system SiO2−Al2O3−CaO−FeO at a normalized content of FeO (20 wt %).

Primary phase fields for selected slag compositions were determined by slag quenching experiments. Coal slag compositions used in the present work are found in three primary phase fields, anorthite, mullite, and spinel, and listed in Table 3 along with phases determined in quenching experiments. Univariant lines in Figure 1 are initially determined 3737

dx.doi.org/10.1021/ef400593k | Energy Fuels 2013, 27, 3736−3742

Energy & Fuels

Article

Table 3. Solid Phases in Coal Slags and Slag Primary Phase Fieldsa solid phases

a

coal slags

primary phase field

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

anorthite mullite mullite mullite mullite spinel spinel spinel mullite anorthite mullite anorthite mullite mullite anorthite anorthite anorthite anorthite anorthite mullite anorthite anorthite spinel

1400 °C

1350 °C

L+m

L L L L

+ + + +

m sp sp sp

1300 °C

L L L L

+ + + +

m sp + a sp + a sp + a

1250 °C

1200 °C

L+a

L+a L+m L+m

L L L L

L L L L L L L L L L L L L L L L L L L

+ + + +

m sp + a sp + a sp + a

L+m

L+m

L+m L+m L+a

L L L L

+ + + +

m m a a

L+m

L+a L+m

L + a + sp L+m

L + sp

L + sp

L L L L L

+ + + + +

a + sp m a a sp + a

+ + + + +

m sp + a sp + a sp + a m

+m + + + + +

m m a a a

+ + + + +

a + sp m a a sp + a

L, liquid phase; a, anorthite; m, mulite; and sp, spinel.

Table 4. Slag Tapping Temperatures, TCV, and Viscosity Parameters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 a

T at 5 Pa s (°C)

T at 25 Pa s (°C)

TCV (°C)

TCV precluded

1230 1540 1440 1470 1430 1400 1390 1400 1480 1410 1450 1480 1445 1480 1480 1440 1390 1350 1380 1470 1300 1240 1500

1220 1370 1290 1320

1230