Fuel 216 (2018) 697–706

Contents lists available at ScienceDirect

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

Full Length Article

Understanding ash deposition for Zhundong coal combustion in 330 MW T

utility boiler: Focusing on surface temperature effects
⁎
Yongzhen Wang, Jing Jin , Dunyu Liu, Haoran Yang, Xuesen Kou
School of Energy and Power Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China

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

Keywords: Zhundong coal as a high quality steam coal has not been widely utilized due to severe fouling and slagging
Zhundong coal problems during combustion. It is crucial to reveal the slagging and fouling mechanisms in a boiler burning
Ash deposits Zhundong coal. In this paper, the properties of ash deposits collected from various heat transfer surfaces in a
Sulfates 330 MW boiler were characterized by ICP-OES, SEM−EDX, and XRD. These results were also verified by an
Silicates
experiment burning Zhundong coal in a drop-tube furnace and thermodynamic calculation by the software
Thermodynamic calculation
FactSage. The results indicated the ash deposited on different heat transfer surfaces in 330 MW boiler burning
Zhundong coal could be classified into silicates and sulfates based on their elemental and mineral analysis.
Further, the ash deposited on the division panel superheater is obviously divided into three layers. In the first
layer, wollastonite and sodium silicate are the main minerals with the contents of quartz and feldspar being
relatively small. The types of mineral in the second layer are the same with the first layer, but their contents are
different from the first layer. Meanwhile, the amount of crystalline phase in the third layer is lower than that of
the first and second layer. In addition, the temperature of the flue gas or the tube played an important role in the
mechanism of ash deposition. Thermodynamics calculation also shows CaSO4 can be formed at all temperatures
and there only exists CaSiO3 formed in the range 900–1200 °C. The ash deposits along the flue gas flow attributes
to sulfates, silicates and sulfates, respectively. The result of the thermodynamic calculation is in accordance with
the experiment results obtained in the drop-tube furnace.

1. Introduction
Zhundong coal field in Xinjiang province is currently the largest
coalfield in China and the reserve for this coal is estimated to be 390
billion tons [1–3]. Zhundong coal is considered as a high quality steam
coal due to its low-to-medium ash yields, low phosphorus and sulphur
contents, high volatile matters, and high calorific value [4]. However,
severe slagging and fouling occur during the combustion of Zhundong
coal. The mechanisms for slagging and fouling are different. Slagging
mainly takes place in the radiant heat transfer section in the boiler,
while fouling occurs in the convective heat transfer section of the
boiler. In addition, slagging is caused by the molten ash particles while
fouling results from the behaviour of various components in the fly ash
as the flue gas cools down. These components include silicate liquids,
and sulfates formed from the reaction between sulphur oxides and al-
kali components [5].The ash deposit layer on heat transfer surfaces
cannot be easily removed by traditional soot blowing techniques [6–8].
It has been widely accepted that temperature, coal ash composition
and particle size are three main factors that have strong influences on
the formation of ash deposits. The ash deposition is affected by wall
temperature and furnace gas exit temperature. Slagging occurs at wall
temperature above 882 °C while fouling occurs during the condensation
of gaseous species such as NaCl or CaSO4 to form a sticky coating layer
on cooled surfaces [9]. Typically, glassy and easily molten deposits
form on high temperature surfaces while porous and sintered (not
molten) deposits form on low temperature surfaces [10]. Additionally,
higher furnace gas exit temperatures can make some minerals in fly ash
reach a molten state, causing severe slagging problems in the boiler
[11].
While previous studies show the importance of temperature on ash
deposition, presumably through the chemical reactions of different ash
components. Wu et al. [12] demonstrated that the formation of NaCl, or
Na2SO4 can cause a high fouling propensity. According to Tang et al.
[13], Ca and S are the main components of combustion residues and
play an important role in the ash deposition process on convective heat
transfer sections through the formation of anhydrite. Song et al. [14]
also indicated that a compound salt CaSO4-Na2SO4 formed burning coal
can be considered as an important cause of ash deposition. Wei et al.

⁎
Corresponding author.
E-mail address: alicejin001@163.com (J. Jin).

https://doi.org/10.1016/j.fuel.2017.08.112
Received 17 March 2017; Received in revised form 24 July 2017; Accepted 30 August 2017
Available online 04 January 2018
0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
Y. Wang et al. Fuel 216 (2018) 697–706

Nomenclature R-WW rear water wall

D-PS division panel superheater
Tflue gas temperatures of flue gas, °C R-PS rear panel superheater
Twall temperatures of the tube, °C R-PR rear panel reheater
wt the weight percentages of elements, % FR final reheater
B/A the base/acid ratio FS final superheater
f fouling index, % PS primary superheater

Fig. 1. The schematic for the 330 MW boiler (a) the sam-
pling positions in the boiler, (b) temperature distributions
of flue-gas and heat transfer surfaces.

698
Y. Wang et al.
[15] investigated the effect of SiO2 on ash fusion during Zhundong coal
combustion. Results revealed that with the increase of SiO2, free CaO
generated by the decomposition of CaSO4 will react with other com-
ponents (SiO2 and Al2O3) in the ash to form calcium aluminosilicate,
leading to the reduced ash fusion temperatures. On the other hand, if
the amount of SiO2 is over 9%, the ash fusion temperature increases due
to SiO2 itself having a high fusion temperature. Furthermore, these free
CaO may react with SiO2, MgO, and Al2O3 to form calcium silicates,
such as gehlenite or akermanite, causing serious ash deposition [16,17].
Hence, understanding the migration of the major elements(Ca, S, Si,
etc) is important for revealing the mechanism for ash deposition.
In addition, particle size has also been found to be an important
factor in the formation of ash deposition [18]. Xu et al. [19]found that
some small particles (< 10 μm) are abundant in the ash deposits, and
believed that small particles results in higher ash deposition propen-
sities. Zheng et al. [20] demonstrated that the initial layer is mainly
composed of fine particles (∼2 μm) and thickness for the outer layer
formed by many particles(> 10 μm) is higher. Furthermore, ash de-
posits have corresponding stratification phenomenon during Zhundong
coal combustion. Sato et al. [21] revealed that initial deposits are made
up a powder layer with a low distribution of the particle packing
fraction due to adhesion of fine particles. According to Zhou et al. [22],
mineral composition is similar for different layers while the proportions
of minerals are different. Moreover, the degree of sintering becomes
more serious from inner layer to outer layer.
Overall, the studies on ash deposition mechanisms for Zhundong
coal combustion are mainly conducted in bench scale apparatus.
699
Fuel 216 (2018) 697–706
Combustion conditions in a lab scale furnace are different from a large
scale power plant. Thus, the aim of this paper is to address ash de-
position mechanisms in a large furnace burning Zhundong coal with
5 wt% kaolin additive. Therefore, characterization of ash deposits
formed from the combustion of Zhundong coal provides new insights
into ash deposition. This understanding can be beneficial for developing
new techniques to relieve ash deposition from Zhundong coal com-
bustion.
Along the flue gas flow, ash deposits on various heat transfer sur-
faces were collected; the mineral compositions of ash deposits were
analyzed in the laboratory. Based on these results, the migration of
major elements including Ca, Si and S is addressed. Additionally, sev-
eral representative ash deposits from the division panel superheater
were also investigated to uncover the stratification phenomenon and
explore the formation mechanism of ash deposits. Finally, the effects of
surface temperature on ash deposition were also verified by the using a
drop-tube furnace and equilibrium simulation based on the FactSage
software.
2. Experimental
2.1. Experimental system
Ash deposits were sampled in a 330 MW utility boiler. They were
located on various convective heat transfer surfaces along the gas flow
direction: rear water wall(R-WW), division panel superheater(D-PS),
rear panel superheater(R-PS), rear panel reheater(R-PR), final reheater
Fig. 2. Schematic diagram of a 5000 W drop-tube furnace
system.
Y. Wang et al.
(FR), final superheater(FS), and primary superheater(PS). The sampling
positions in the boiler are marked in Fig. 1a. Further, the temperature
distributions of flue gas and heat transfer surfaces are also provided in
Fig. 1b. The gas temperature data came from the measurements in the
boiler, and the points for measurements were located on the left/right
side of heat transfer surfaces. The wall temperature was measured by a
thermocouple to the tube in the steam outlet sections.
Ash deposition tests were also conducted in a Drop Tube Furnace.
This system was consisted of a main tube reactor (2400 mm of total
height × 100 mm of inner diameter), a deposition tube and an ash
collecting unit as shown in Fig. 2. There were two heating zones for the
furnace, and temperature for each can be controlled separately. Coal
particles (70 μm of average diameter) were introduced with a speed of
8 g/min in the experiments. The excess air ratio was 1.2 and the
duration time for coal feed was 5 h. The resulted burn-out rates were
higher than 97%. The deposition tube was made up of two inner con-
nected steel tubes with sealed ends. Two side tubes were connected
with each main tube to provide water inlet and outlet. The temperature
of the tube was controlled by water flow rate to simulate different heat
transfer surfaces in boilers. Five experiments with different tempera-
tures of flue gas and the tube marked with Tflue gas/Twall (1200 °C/
450 °C, 900 °C/450 °C, 900 °C/550 °C, 900 °C/650 °C, 600 °C/450 °C)
were conducted.
2.2. Coal properties
One typical Zhundong coal, namely Yihua coal with high volatile
and low ash content was selected in this study, this being a good coal
for pf combustion. However, serious ash deposition occurred when this
coal was used in the boilers due to the abundance of alkali-alkaline
earth metal (AAEM) elements in ash. Table 1 gives the proximate and
ultimate analysis for coal and it also provides ash fusion temperatures
for standard ash samples. Table 2 shows elemental analysis for coal ash
and ash with 5 wt% kaolin addition. In coal ash, the amounts of Na2O
(up to 7.12%) and CaO(up to 32.9%) are high, while those of
SiO2(5.63%) and Al2O3(4.77%) are low. In particular, the amount of
SO3 in coal ash reaches 34.00%.
2.3. Experimental and simulation method
Coal samples were pulverized and sieved to a size fraction of
0.1 ∼ 10 μm to meet instrument measurement requirement. The test
was conducted three times and averaged for the elemental content.
Elemental content in ash deposits can be quantified with Inductively
Coupled Plasma Optical Emission Spectrometer (ICP-OES, Prodidy).
Samples were digested in a solution of acid or alkali in a 100 ml vo-
lumetric flask before the analysis to make sure the elements in ash
deposits were completely dissolved. Microstructure and distribution of
the major elements for ash deposits were also analyzed by SEM-EDX
from Phenom Prox. XRD-6100 from Shimadzu was used to analyze the
mineral composition of slag samples from various locations in the
boilers. The samples were run on XRD operated at 45 kV, 30 mA, di-
vergence slits of 1°, receiving slits of 0.3 mm and the scan speed by 6°/
min respectively. Furthermore, the traces were run in the step mode at
0.02° step intervals over an angular range of 10° to 90°. The curves for
XRD were further processed using Jade.6.
Based on analyses of elemental composition obtained from ICP-OES,
an equilibrium calculation was also conducted using FactSage 5.2. The
thermodynamics module was used to obtain the mineral forms in the
ash samples. Proportions of the liquid and solid phases for a specified
condition for the multi-component system can also be obtained
[23–26]. Calculation conditions are given as follows. The air atmo-
sphere with 21% O2 and 79% N2 was employed; The pressure was
chosen as 0.1 Mpa; The amount of Si, S, Ca, Na, Al and O were given on
the basis of 1 kg of ash deposits; The temperature region of
400–1400 °C was calculated with an interval of 100 °C.
700
Fuel 216 (2018) 697–706
3. Results and discussion
3.1. Characteristics of ash deposits
Elemental analysis of ash deposits from various heat transfer sur-
faces is shown in Fig. 3 and Table 3. By comparing the amount of ele-
ments in ash deposits from different locations of boiler, the result shows
that the weight percentages of Na and K in the ash deposits are greater
than these contained in fly ash collected from the dust remover. This
shows the enrichment of Na and K in ash deposits. The weight per-
centages of Ca and Fe in ash deposits do not differ significantly from
these contained in the fly ash, while the contents of Al and Mg in the fly
ash are higher than those in ash deposits. Additionally, it is worth to
compare the ash deposits and the chemistry of the coal ash. Ash de-
posits with 5 wt% kaolin additive were sampled in a 330 MW utility
boiler. As is done in the power plant, the chemistry of the coal ash with
5 wt% kaolin was prepared according to GB/T 1574-2007. Table 2
shows that the weight percentages of Na, K and Fe in ash deposits are
higher than those contained in the standard ash with 5 wt% kaolin.
Moreover, there is a clear enrichment of Si in the deposits obtained at
D-PS and R-PS. However, the weight percentages of Ca and Mg in ash
deposits are lower than these in standard ash with kaolin. Overall, SiO2,
CaO and SO3 are the three main species in most ash deposits burning
Yihua coal. Therefore, the deposition of these substances during Yihua
coal combustion is substantially higher than the deposition of other
substances.
Viscosity is an important parameter which indicates the tendency of
ash deposition. The viscosity of ash deposits increases with the increase
of SiO2 and Al2O3 in deposits, as Si and Al in the ash deposits are
considered as the network-forming elements. In silicate melts,
Al3+occupies the position by 4 and 6 coordination in the deposits, and
in particular, the existence of [AlO4] can increase the viscosity of ash
deposits [27]. Furthermore, if SiO2 and other oxides such as CaO with a
strong bond emerge at the same time, [AlO6] may be easily converted to
[AlO4], leading to an increase in the viscosity of ash deposits [28].
Therefore, the viscosity can be raised due to an increase of Al2O3. Si-
milarly, SiO2 as a network former is a prime factor affecting the visc-
osity of ash deposits, and the viscosity in ash melts is also determined
by the degree of connections for the silicon-oxygen tetrahedron struc-
ture [29]. With an increase of SiO2 in coal ash, the tetrahedral network
grows and internal friction force resulted from internal motion of a
particle with fluid flow increases. In a certain range, the viscosity of ash
can be raised with an increase of SiO2 [30]. Fig. 3 shows that the weight
percentages of Si and Al in the ash deposits from D-PS and R-PS are
higher than those obtained from other locations. This may indicate that
Table 1
The basic analysis for coal and ash.
Name of Analysis Symbol Value
Proximate analysis/% Mad 12.27
Aad 3.29
Vad 27.22
FCad 57.22
Ultimate analysis/% Cd 73.44
Hd 3.23
Od 17.78
Nd 1.31
Sd 0.49
Qnet.ad/kJ/kg 25109
Ash fusion temperatures/°C DT 1372
ST 1375
HT 1436
FT 1444
M, Moisture; A, Ash; V, Volatile; FC, Fixed Carbon; ad, air dry basis; d, dry basis; DT,
deformation temperatures; ST, soft temperatures; HT, hemispherical temperatures; FT,
flow temperatures.
Y. Wang et al. Fuel 216 (2018) 697–706

Table 2 Fe2 O3 + CaO + MgO + Na2 O+ K2O

B/A =
Elemental analysis of coal ash and ash with 5 wt% kaolin. SiO2 + Al2O3 (1)
Chemical components (wt.%) f = B / A × Na2O (2)
Na2O K2O CaO Fe2O3 MgO Al2O3 SiO2 SO3 Fig. 5 shows XRD patterns of ash deposits from various heat transfer
surfaces. Strong peaks identified in the R-WW deposit are those of
Coal ash 7.12 1.17 32.90 3.72 4.64 4.77 5.63 34.00
CaSO4 (Anhydrite), and minor amount of CaSiO3 (Wollastonite), Na
Ash with 5% kaolin 5.27 0.85 24.24 2.72 3.39 11.41 26.74 25.12
(AlSi3O8)(Albite), SiO2(Quartzs), (Ca,Na)(Si,Al)4O8(Anorthite) and
Al2SiO5(Mullite) are also found. CaSiO3 and (Ca,Na)(Si,Al)4O8 are the
primary phase present in the samples from D-PS and R-PS, while small
amount of CaSO4, SiO2 and Na(AlSi3O8) are also observed. The dif-
fraction peaks for CaSO4 is significant from R-PR to PS, and small
amount of CaSiO3, Na(AlSi3O8), SiO2 and (Ca,Na)(Si,Al)4O8 are ob-
served. Consequently, the ash deposits can be classified into silicates
and sulfates based on the elemental and mineral analysis as shown in
Table 4.
As CaSO4, CaSiO3 and (Ca,Na)(Si,Al)4O8 are dominant minerals for
ash deposits at different locations in the boiler, it would be interesting
to discuss how these minerals are formed. In terms of CaSO4, CaSO4
may be formed by the reaction between CaO produced by the decom-
position of calcite during coal combustion and SO3 in the gas phase.
Some of the formed CaSO4 will attach to R-WW, giving rise to ash de-
posits. The CaSO4 present in the high temperature flue gas will de-
compose to free CaO; which will then react with silica and other mi-
Fig. 3. Elemental analysis of ash deposits from various heat transfer surfaces. nerals in ash, to form CaSiO3 and (Ca,Na)(Si,Al)4O8 etc. [33,34] based
on formulas (3) to (6). Consequently, ash deposits composed of massive
Table 3 CaSiO3 and (Ca,Na)(Si,Al)4O8 occurs in D-PS and R-PS. In addition, gas
Elemental analysis of ash deposits from various heat transfer surfaces. phase calcium based species in the flue gas quickly reacts with sulphur
oxides to form calcium sulfates, and the aerosols of calcium sulfates
Location of heating Content (wt%) formed may deposit on heat transfer surfaces with decreasing flue gas
surface
Na2O K2O CaO Fe2O3 MgO Al2O3 SiO2 SO3
temperature [1]. It can be seen from Fig. 5 that a significant amount of
sulfates aerosol deposit on heat transfer surfaces under the temperature
R-WW 5.78 1.41 17.26 8.00 2.93 9.27 35.37 16.03 of 300 ∼ 650 °C in flue gas to aggravate the adhesion of ash particles
D-PS 6.01 1.29 18.89 8.33 2.80 12.56 43.17 2.70 on heat transfer surfaces.
R-PS 5.07 1.54 16.23 9.83 2.93 12.62 46.93 1.85
R-PR 7.88 1.19 19.75 9.93 2.57 7.20 23.33 25.58 CaO + SiO2 → CaSiO3 (Wollastonite) (3)
FR 8.90 1.18 17.60 5.71 2.65 7.75 24.08 25.6
FS 6.13 1.21 19.12 8.47 2.78 8.41 25.45 24.58 CaO + Al2 O3 ·2SiO2 (Metakaolin) → CaO·Al2O3 ·2SiO2 (Anorthit) (4)
PS 7.52 1.68 16.98 6.77 2.93 8.14 28.46 22.25
FA 3.18 0.58 15.92 6.83 4.60 17.23 44.14 2.25 2Al2 O3 + Al2 O3 ·2SiO2 → 3Al2 O3 ·2SiO2 (Mullite) (5)

CaO + 3Al2O3 ·2SiO2 → CaO·Al2O3 ·2SiO2 (6)

the viscosity of ash deposits from D-PS and R-PS may be higher than
others. 3.2. Formation mechanism of silicate ash deposits
The tendency of ash deposition can be estimated by the slagging/
fouling indicies. Especially, the base/acid ratio(B/A) is common used to In D-PS and R-PS, a significant amount of wollastonite is present,
evaluate the degree of ash deposition in various heat transfer surfaces and those ash deposits are harder compared with other locations.
[31]. Furthermore, due to the high content of Na2O in the ash deposits,
the fouling index(f) is used to estimate the tendency for particles to 1.4 14
adhere on surfaces [32]. The base/acid ratio(B/A) and fouling index(f) B/A
based on formulas (1) and (2) are used to estimate the tendency of ash 1.3 f
deposition in Fig. 4. The result shows that B/A of the ash deposits at D- 12
1.2
PS and R-PS is lower than those at other locations. Furthermore, B/A of
the ash deposits at R-PR and FS is higher than those at other locations. 1.1 10
Previous research indicates that, as the value of B/A increases, the
tendency for slagging and high temperature fouling will increase. 1.0
8
f
B/A

Nevertheless, under normal circumstances, increasing SiO2 and Al2O3 0.9

contents can cause an increase in the viscosity of the ash. Therefore, if
B/A increases, the viscosity of ash deposits can decrease, which is 0.8 6
consistent with the behavior shown in Fig. 4.
0.7
Furthermore, the tendency of f is the same with B/A and f can be 4
used to estimate the tendency for particles to adhere on surfaces. From 0.6
Fig. 4, f of the ash deposits at D-PS and R-PS is relatively low. This
indicates that fouling at these locations can be lighter than that at other 0.5 2
R-WW D-PS R-PS R-PR FR FS PS
locations. However, the viscosity of D-PS and R-PS is higher, indicating
that more serious ash deposition may occur on these surfaces. the location of heat transfer surfaces
Fig. 4. Ratio of B/A and f of ash deposits from various heat transfer surfaces.

701
Y. Wang et al. Fuel 216 (2018) 697–706

Fig. 5. XRD patterns of ash deposits from various heat transfer surfaces.

Table 4
The classification of ash deposits from various locations.

Location of heating surface The temperature of flue gas /°C mineral The classification of ash deposits

R-WW 1000–1400 CaSO4 mainly; CaSiO3, Na(AlSi3O8), SiO2, (Ca,Na)(Si,Al)4O8 and Al2SiO5 sulfates
slightly
D-PS
R-PS 650–1000 CaSiO3 and (Ca,Na)(Si,Al)4O8 mainly;CaSO4, SiO2 and Na(AlSi3O8) slightly silicates
R-PR 300–650 CaSO4 mainly;CaSiO3, Na(AlSi3O8), SiO2 and (Ca,Na)(Si,Al)4O8 slightly sulfates
FR
FS
PS

Therefore, it is important to study the formation mechanism of silicate
ash deposits at these two locations. In particular, the samples from D-PS
were taken for further analysis.
The ash deposits (D-PS) can be visually divided into three layers in
Fig. 6, which shows that in the first layer (Layer A), microstructure of
ash deposits exists in the forms of beads and clinkers. The layer A is
composed of condensing submicron particles, fine silicates and sticky
large particles, and inertial impaction of sticky particles may result in
rapid deposition of layer A [18]. A large number of submicron particles
are composed of alkali metal and alkaline earth metals such as sodium
and calcium. In the second layer (Layer B), the microstructure is com-
posed of euhedral crystals and ash particles. These may be silicate
minerals. In the outer layer (Layer C), the ash deposits are formed by
aggregation of long columnar and acicular with no obvious gaps. Due to
the layer C located on the radiant zone of the boiler, the temperature is
very high and the ash particles are melted. Rather than melting, the
image of layer C seems to indicate adherence of molten ash which is
then cooling and crystallizing. This results in the conversion from
crystal into non-crystal phase on the outer layer of the ash deposits.
The spot analyses of the crystals are shown in Fig. 6. EDX data
shows that silicon, oxygen, sodium, calcium and aluminum are identi-
fied in three layers, while the proportions for them are different. Sul-
phur does not exist in the layer A, but there is sulphur in the layer B and
layer C. The proportion of calcium and sulphur in the layer C are
highest than other layers. That indicates that its related substances may
be generated in the layer C. The content of silicon is high in all three
layers.
Fig. 7 shows XRD patterns of ash deposits at D-PS. In the layer A,
wollastonite and sodium silicate are the main minerals, while quartz
and feldspar contents are relatively low. This is because volatile species
including CaO, NaCl or NaOH may react with SiO2 to form wollastonite
and sodium silicate, causing formation of a sticky layer, which can
enhance adherent efficiency. In the layer B, the mineralogy is similar
but the content of the minerals is different. The decrease in the contents
of wollastonite and sodium silicate are concomitant with the sharp in-
crease in the contents of anorthite, albite and quartzs. Calcium or so-
dium aluminosilicate between ash particles adhering to the sticky layer
by inertial impaction can agglomerate. Additionally, due to cooling and
crystallization for the molten ash adhering to the ash deposits, the
diffracted intensity of the layer C sample is lower than that of the first
and second layer. Mullite and anhydrite can be found in layer C, while
the shape of anhydrite and mullite is mostly long columnar and acicular
[35,36]. This agrees very well with the analysis of SEM. And the ex-
istence of anhydrite in the layer C can cause the high content of Ca and
S. Combined with the result of EDX, the proportion of calcium and
sulphur in the layer C are highest than other layers. This is consistent

702
Y. Wang et al. Fuel 216 (2018) 697–706

Fig. 6. morphology and elemental contents of ash deposits

at D-PS.

703
Y. Wang et al. Fuel 216 (2018) 697–706

Due to different temperatures of the tube, main minerals are not the
same under the temperature of 900 °C in flue gas. In Fig. 8, the strong
peaks of CaSiO3 and (Ca0.94Na0.06)(Al1.94Si2.06O8) are found in the
condition of 900 °C/550 °C, while CaSO4 is the main mineral in the
condition of 900 °C/450 °C. Furthermore, the thermal conductivity of
CaSiO3 is smaller than that of CaSO4 [37,38]. Combined with Fig. 7, the
sticky layer of the ash deposits is made of CaSiO3 and Na2Si2O5 at D-PS.
Due to the low thermal conductivity of CaSiO3, heat transfer is reduced.
Additionally, the high surface temperature for 550 °C and the low
thermal conductivity of CaSiO3, causing higher temperature of the de-
posit layer in 900 °C/550 °C. This promotes the formation of
(Ca0.94Na0.06)(Al1.94Si2.06O8).
Thermodynamic calculations were performed to obtain the forma-
tion of shown in Fig. 9. Here, only calcium is considered as the positive
charge species. From 1400 °C to 1200 °C, some CaSO4 is formed and a
large amount of free CaO is also observed. However, CaSiO3 is not
found. From 1200 °C to 900 °C, the content of CaSiO3 increases with the
decrease in temperature, and it reaches a maximum at 1000 °C and then
decreases. From 900 °C to 400 °C, CaSiO3 disappears, and CaSO4 is the
main mineral formed. However, CaSO4 can be found at all tempera-
tures. The ash deposition sample taken from the power plant had been
cooled before the XRD analysis. The compounds formed at higher
temperature may transform when cooled. So Na2Ca3Si6O6 can exist by
thermodynamics calculation in 400 °C–1000 °C due to the crystal con-
version of (Ca,Na)(Si,Al)4O8. Thermodynamic calculation also agrees
well with results in Table 4 in that CaSiO3 and (Ca,Na)(Si,Al)4O8 are the
primary phases in the temperature ranges 650 °C–1000 °C.
Combined with Fig. 7, CaSiO3 and Na2Si2O5 cause formation of a
sticky layer, and the sticky layer can enhance the adherent efficiency of
ash deposition. In addition, the melting point of CaSiO3 is also lower
than that of CaSO4. With an increase in the temperature of the ash
deposits layer, melting points are reached for silicates and this results in
Fig. 7. XRD patterns of ash deposits at D-PS.

with the result of EDX.

3.3. Results for the drop-tube furnace and the software FactSage

The parameters (temperature of flue gas, temperature of the tube,

and atmosphere) are easily controlled in bench scale apparatus. In order
to further reveal the mechanism for the migration of the major elements
(Ca, Si and S) and understand the effect of the surface temperature on
ash deposition, experiments in a drop-tube furnace were conducted in
order to verify the observations made on the 330 MW Boiler samples.
Furthermore, thermal equilibrium calculations were also performed to
verify the boiler sample observations using the thermodynamic calcu-
lation software package FactSage.
Fig. 8 shows XRD patterns of ash deposits from the DTF. With the
surface temperature of ash collection tube kept at 450 °C, the tem-
peratures of the simulated flue gas were controlled at 1200 °C, 900 °C
and 600 °C, respectively. The strong peaks are of CaSO4 for the condi-
tions of 1200 °C/450 °C, 900 °C/450 °C, 600 °C/450 °C. Meanwhile, a
minor amount of SiO2, Na(AlSi3O8), CaSiO3, (Ca,Na)(Si,Al)4O8 and
Al2SiO5 were also found. Thus, the ash deposits can be classified into
sulfates. However, this does not agree with the results from Table 4.
Therefore, the temperature of the flue gas is not the only factor.
The temperature of the tube can play an important role in the for-
mation of ash deposits. When the surface temperature of the tube was
increased to 550 °C and 650 °C respectively, the composition of the ash
deposits changed significantly. Specifically, the strong peaks can be
identified as CaSiO3, (Ca,Na)(Si,Al)4O8, (Ca0.94Na0.06)(Al1.94Si2.06O8)
and a certain amount of CaSO4, SiO2 and Na(AlSi3O8) may also be
observed for the condition of 900 °C/550 °C and 900 °C/650 °C. As a
result, the formation mechanism of ash deposits can be affected by both
Fig. 8. XRD patterns of ash deposits from the drop-tube furnace.
the temperature of flue gas and the tube.

704
Y. Wang et al.
Fig. 9. Speciation of Ca-based mixed oxide formed under different temperatures calcu-
lated by thermodynamics calculation.
the formation of rigid clinkers. Silicates are easily adhered to the tube
and deposit layers than sulfates. Thus, silicates are the main minerals in
650 °C–1000 °C and therefore, the ash deposits can be classified into
sulfates, silicates and sulfates along the flue gas flow.
4. Conclusions
The properties of ash deposits collected from various locations in a
330 MW boiler were investigated by ICP-OES, SEM-EDX and XRD.
Conclusions drawn from analysis of the boiler samples were also ver-
ified by drop tube furnace experiments and thermodynamic calcula-
tions using FactSage. Four conclusions can be drawn as follows.
(1) The ash deposits can be classified into silicates and sulfates: a) from
1000 to 1400 °C, the ash deposits at R-WW belongs to sulfates; b)
from 650 to 1000 °C, the ash deposits at D-PS and R-PS are silicates;
c) from 300 to 650 °C, the ash deposits (R-PR, FR, FS and PS) are
classified as sulfates.
(2) The ash deposits (D-PS) is obviously stratified: a) in the first layer,
wollastonite and sodium silicate are main minerals, while quartz
and feldspar contents are relatively low. As volatile CaO, NaCl or
NaOH may react with SiO2 to form wollastonite and sodium silicate,
a sticky initial layer may be formed; b) in the second layer, the
mineralogy does not change, but the content of minerals does; c)
the diffracted intensity of the sample of the third layer is lower than
that of the first and second layer because of presence of a glassy
melt phase.
(3) Drop tube furnace experiments can be used to identify the ash
formation mechanisms. These can be affected by the temperature of
flue gas and the tube. CaSO4 is the main mineral foemed in the
temperature range 900 °C/450 °C while CaSiO3 and (Ca,Na)
(Si,Al)4O8 are the main minerals occurring in the range 900 °C/
650 °C.
(4) FactSage calculations agree with the experiments in that ash de-
posits may also be classified into silicates and sulfates. From
1400 °C to 1200 °C, some CaSO4 is found and a large amount of free
CaO is also observed. However, CaSiO3 is not found. From 1200 °C
to 900 °C, the content of CaSiO3 increases with the decrease in
temperature, reaching a maximum at 1000 °C and then decreasing.
From 900 °C to 400 °C, CaSiO3 disappears, and CaSO4 is the main
mineral formed. However, CaSO4 is present at all temperatures.
Acknowledgment
The authors highly appreciate the financial support from National
“Twelfth Five-Year” Plan for Science & Technology Support of China
705
Fuel 216 (2018) 697–706
(No. 2015BAA04B03).
References
[1] Wang X, Xu Z, Wei B, Zhang L, Tan H, Yang T, et al. The ash deposition mechanism
in boilers burning Zhundong coal with high contents of sodium and calcium: a study
from ash evaporating to condensing. Appl Therm Eng 2015;80:150–9.
[2] Li J, Zhu M, Zhang Z, Zhang K, Shen G, Zhang D. Characterisation of ash deposits on
a probe at different temperatures during combustion of a Zhundong lignite in a drop
tube furnace. Fuel Process Technol 2016;144:155–63.
[3] Li J, Zhu M, Zhang Z, Zhang K, Shen G, Zhang D. Effect of coal blending and ashing
temperature on ash sintering and fusion characteristics during combustion of
Zhundong lignite. Fuel 2017;195:131–42.
[4] Jin H, Chen Y, Ge Z, Guo L. Hydrogen production by Zhundong coal gasification in
supercritical water. Int J Hydrogen Energy 2015;40:16096–103.
[5] Couch G. Understanding Slagging and Fouling in PF Combustion. London: lEA Coal
Research; 1994.
[6] Xu L, Liu H, Fang H, Gao J, Wu S. Effects of various inorganic sodium salts present in
Zhundong coal on the char characteristics. Fuel 2017;203:120–7.
[7] Chen C, Zhang S, Shi D, Lv J, Liu D, Guo X, et al. Research on sodium removal from
Zhundong coal. Adv. Mater. Res. 2013;734–737:614–8.
[8] Li G, Li S, Qian H, Yao Q. Fine particulate formation and ash deposition during
pulverized coal combustion of high-sodium lignite in a down-fired furnace. Fuel
2015;143:430–7.
[9] Qi X, Song G, Song W, Yang S, Lu Q. Effects of wall temperature on slagging and ash
deposition of Zhundong coal during circulating fluidized bed gasification. Appl
Therm Eng 2016;106:1127–35.
[10] Tomasz K, Krzysztof Z, Roman W. Effect of fuel type and deposition surface tem-
perature on the growth and structure of an ash deposit collected during co-firing of
coal with sewage sludge and sawdust. Energy Fuels 2009;23:3429–36.
[11] Bilirgen H. Slagging in pc boilers and developing mitigation strategies. Fuel
2014;115:618–24.
[12] Wu X, Zhang X, Dai B, Xu X, Zhang J, Zhang L. Ash deposition behaviours upon the
combustion of low-rank coal blends in a 3 MW th, pilot-scale pulverized coal-fired
furnace. Fuel Process Technol 2016;152:176–82.
[13] Tang Z, Chen X, Liu D, Zhuang Y, Ye M, Sheng H, et al. Experimental investigation
of ash deposits on convection heating surfaces of a circulating fluidized bed mu-
nicipal solid waste incinerator. J. Environ. Sci. 2016;48:169–78.
[14] Song G, Yang S, Song W, Qi X. Release and transformation behaviors of sodium
during combustion of high alkali residual carbon. Appl Therm Eng
2017;122:285–96.
[15] Wei B, Wang X, Tan H, Zhang L, Wang Y, Wang Z. Effect of silicon–aluminum
additives on ash fusion and ash mineral conversion of Xinjiang high-sodium coal.
Fuel 2016;181:1224–9.
[16] Zhang L, Dong X, Liu K, Tan H, Wang X, Wei B. Effect of kaolin on ash slagging and
mineral conversion of Zhundong coal. J. Fuel Chem. Technol. 2015;43:1176–81.
[17] Yao Y, Jin J, Liu D, Wang Y, Kou X, Lin Y. Evaluation of vermiculite in reducing ash
deposition during the combustion of high-calcium and high-sodium zhundong coal
in a drop-tube furnace. Energy Fuels 2016;30:3488–94.
[18] Wall TF. Mineral matter transformations and ash deposition in pulverised coal
combustion. In: Proceedings of the Symposium (International) on Combustion;
1992.
[19] Xu J, Yu D, Fan B, Zeng X, Lv W, Chen J. Characterization of ash particles from co-
combustion with a Zhundong coal for understanding ash deposition behavior.
Energy Fuels 2013;28:678–84.
[20] Zheng S, Zeng X, Qi C, Zhou H. Mathematical modeling and experimental validation
of ash deposition in a pulverized-coal boiler. Appl Therm Eng 2017;110:720–9.
[21] Sato N, Priyanto DE, Ueno S, Matsuzawa Y, Ueki Y, Yoshiie R, et al. Growth and
gravity shedding of ash deposition layer in pulverized coal combustors. Fuel Process
Technol 2015;134:1–10.
[22] Zhou B, Zhou H, Wang J, Cen K. Effect of temperature on the sintering behavior of
Zhundong coal ash in oxy-fuel combustion atmosphere. Fuel 2015;150:526–37.
[23] Liu X, Yu G, Xu J, Liang Q, Liu H. Viscosity fluctuation behaviors of coal ash slags
with high content of calcium and low content of silicon. Fuel Process Technol
2017;158:115–22.
[24] Christoph W, Benjamin K, Gundula B, Hartmut S. Evaluation, comparison and va-
lidation of deposition criteria for numerical simulation of slagging. Appl Energy
2012;93:184–92.
[25] Yan T, Kong L, Bai J, Bai Z, Li W. Thermomechanical analysis of coal ash fusion
behavior. Chem Eng Sci 2016;147:74–82.
[26] Yan T, Bai J, Kong L, Bai Z, Li W, Xu J. Effect of SiO2/Al2O3, on fusion behavior of
coal ash at high temperature. Fuel 2017;193:275–83.
[27] Xiao Z, Cheng J, Wu H. Effect of Al2O3/SiO2 ratio on the viscosity and workability
of high-alumina soda-lime-silicate glasses. J Chinese Ceramic Soc 2012;40:1000–5.
[28] Kim WH, Sohn I, Min DJ. A study on the viscous behaviour with K2O additions in
the CaO–SiO2–Al2O3–MgO–K2O quinary slag system. Steel Res Int 2010;81:735–41.
[29] Hurst HJ, Novak F, Patterson JH. Viscosity measurements and empirical predictions
for some model gasifier slags. Fuel 1999;78:439–44.
[30] Dexia LI, Zhou Z, Guo Q, Guangsuo YU. Ash fusion and viscosity-temperature
characteristics of yulin coal. CIESC J 2012;63:9–17.
[31] Danišovič L, Varga I, Polák S, Bajčíková B, Adamkov M, Vojtaššak J. Inhibition of
lignite ash slagging and fouling upon the use of a silica-based additive in an in-
dustrial pulverised coal-fired boiler. part 1. Changes on the properties of ash de-
posits along the furnace. Fuel 2015;139:720–32.
[32] Paula T, Helena L, Ibrahim G, Nuno L, Pedro A. Evaluation of slagging and fouling
Y. Wang et al.
tendency during biomass co-firing with coal in a fluidized bed. Biomass Bioenergy
2012;39:192–203.
[33] Ozturk ZB. Microstructural characterization of mullite and anorthite-based
Porcelain tile using regional clay. J Ceramic Process Res 2016;17:555–9.
[34] Wu X, Zhang X, Yan K, Chen N, Zhang J, Xu X, et al. Ash deposition and slagging
behavior of Chinese Xinjiang high-alkali coal in 3 MWth pilot-scale combustion test.
Fuel 2016;181:1191–202.
[35] Yoon W, Sarin P, Kriven WM. Growth of textured mullite fibers using a quadrupole
706
Fuel 216 (2018) 697–706
lamp furnace. J Eur Ceram Soc 2008;28:455–63.
[36] Pengfei L, Fei D, Dang Y. Influence of fe∼(3+) on calcium sulfate whisker by
hydrothermal synthesis. Chem Ind Eng Progress 2014;33:165–8.
[37] Müller-Steinhagen H. Heat exchanger fouling – Mitigation and cleaning technolo-
gies. Germany: DLR; 2000.
[38] Radhakrishnan S, Sonawane PS. Role of heat transfer and thermal conductivity in
the crystallization behavior of polypropylene-containing additives: a phenomen-
ological model. J Appl Polym Sci 2010;89:2994–9.