Xiaobin Qi1, Weijian Song1, Guoliang Song1,2. 1. Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China. 2. University of Chinese Academy of Sciences, Beijing 100049, China.
Abstract
It is crucial to accurately evaluate the slagging characteristics of coal before industrial application. However, when evaluating the Xinjiang high-sodium low-rank coal (XJc) according to the properties of the ash prepared at 815 °C, the ashing temperature of the coal sample specified by Chinese standards, forecasting failures are frequent. To figure out the internal reason, the influence of ashing temperature on predicting slagging characteristics of XJc was investigated via an ashing test and thermodynamic equilibrium calculation. Experimental results show that when XJc is ashed at 815 °C, the ash yield apparently reduces compared to the case when ashed at 500 °C since numerous volatile components are released. The release of these components contributes to an obvious inhibition in the formation of the liquid phase in the residual ash, especially at temperatures below 1200 °C. Hence, reducing the ashing temperature is conducive to the acquirement of more real ash compositions of XJc and thereby to the accurate prediction of the slagging behaviors. By a comparison with reported experimental results, it is found that the liquid ratio-temperature curve calculated from the ash compositions of the ash prepared at 500 °C basically reflects the actual slagging tendency of XJc. According to the evolution of minerals with temperature, two slagging mechanisms, self-fusion of sodium-bearing salts and low-temperature eutectics, are confirmed. In addition, effects of antislagging measures of adding refractory oxides greatly differ among coal types due to the diversity in ash compositions.
It is crucial to accurately evaluate the slagging characteristics of coal before industrial application. However, when evaluating the Xinjiang high-sodium low-rank coal (XJc) according to the properties of the ash prepared at 815 °C, the ashing temperature of the coal sample specified by Chinese standards, forecasting failures are frequent. To figure out the internal reason, the influence of ashing temperature on predicting slagging characteristics of XJc was investigated via an ashing test and thermodynamic equilibrium calculation. Experimental results show that when XJc is ashed at 815 °C, the ash yield apparently reduces compared to the case when ashed at 500 °C since numerous volatile components are released. The release of these components contributes to an obvious inhibition in the formation of the liquid phase in the residual ash, especially at temperatures below 1200 °C. Hence, reducing the ashing temperature is conducive to the acquirement of more real ash compositions of XJc and thereby to the accurate prediction of the slagging behaviors. By a comparison with reported experimental results, it is found that the liquid ratio-temperature curve calculated from the ash compositions of the ash prepared at 500 °C basically reflects the actual slagging tendency of XJc. According to the evolution of minerals with temperature, two slagging mechanisms, self-fusion of sodium-bearing salts and low-temperature eutectics, are confirmed. In addition, effects of antislagging measures of adding refractory oxides greatly differ among coal types due to the diversity in ash compositions.
Slagging is a major issue
limiting coal utilization. With the exploitation
and utilization of Xinjiang high-sodium low-rank coal (XJc) in Northwest
China, the slagging issue has become more prominent in recent years.[1,2] Due to the lack of adequate knowledge on this type of coal and an
effective prediction of slagging, most XJc-firing boiler plants suffer
from a severe slagging problem at an early stage, resulting in frequent
safety incidents and unscheduled shutdown of boiler units. Research
has shown that a high content of alkali and alkaline earth metals
(AAEMs) in the coal ash was responsible for this issue.[3−5]For an early warning, accurately predicting slagging characteristics
of coal before industrial application is necessary. Most current predictions
are based on ash fusibility and ash compositions. Regarding the determination
of ash fusibility, the most common approach is to measure the ash
fusion temperatures (AFTs) via the ash cone fusion test, including
the deformation temperature (DT), the softening temperature (ST),
the hemispherical temperature (HT), and the fluid temperature (FT).
AFTs reflect the real melting state of coal ash; thus, the prediction
based on AFTs is relatively accurate in theory. To date, numerous
correlative indices have been proposed to evaluate the slagging tendency
of coal.[6−8] According to the concept of ionic potential, ash
compositions are classified into acidic oxides (SiO2, Al2O3, and TiO2) and alkaline oxides (Na2O, K2O, MgO, and CaO).[9] The former generally increase AFTs, whereas the latter decrease
AFTs. Following such relevance, many slagging indices have also been
proposed to predict the slagging characteristics of coal.[6] However, in some cases, impacts of certain important
components such as sulfur[10] and chlorine
are frequently ignored.When evaluating the slagging characteristics
of XJc using the above-mentioned
methods, the accuracy is often questionable. For instance, Zhou et
al.[11] pointed out that the measured AFTs
via the ash cone fusion test were too high to accurately evaluate
the actual slagging situations of industrial XJc-firing boilers. In
much of the reported literature,[12,13] there was
also a mismatch between measured ash compositions and actual slagging
characteristics. The poor prediction most likely resulted from the
enrichment of highly volatile AAEMs in low-rank coals.[14] It should be noted that according to Chinese
standards (GB/T 212-2008, GB/T 219-2008, GB/T 1574-2007, etc.), the
coal ash for the ash cone fusion test and the ash composition measurement
should be prepared at 815 °C. For XJc, the ash yield at 815 °C
almost halved, compared to the case at 400 °C.[15] Further analysis showed that the release of sodium, sulfur,
and chlorine was significantly responsible for the huge ash loss.
In actual combustion and gasification conditions, the species released
are probably recaptured by ash particles under the action of gas-solid
disturbance, promoting the slagging.[10,16−18] However, in the static ashing process and the ash cone fusion test,
this cannot be achieved. This viewpoint is supported by a reported
result that the ST of XJc ash prepared at 500 °C was about 100
°C lower than that prepared at 815 °C.[19] In the studies of Li et al.[20] and Yao et al.,[21] similar results were
also reported. Besides the inaccuracy of prediction, large discrepancies
in properties of XJc among different mines[22] also increase the difficulty in understanding the slagging mechanisms[23] and reduce the universality of an antislagging
method.To date, thermodynamic equilibrium calculation has been
widely
used to evaluate the slagging characteristics of coals.[24,25] For XJc, due to the aforementioned impacts of ashing temperature,
the accuracy of thermodynamic equilibrium calculation based on the
ash prepared at a certain ashing temperature may be questionable.
To the best of the authors’ knowledge, the influence of ashing
temperature on the accuracy of such a calculation has not been determined.Against this research gap, the variation of ash compositions of
XJc with ashing temperature was first investigated in this study.
Then, the impact of ashing temperature on the calculating accuracy
was evaluated by a comparison with the reported experimental results.
After that, the detailed slagging characteristics of XJc and antislagging
strategies via adding mineral additives were discussed by a thermodynamic
equilibrium calculation. Considering the diverse properties of XJc
mined from different coalfields, four kinds of XJc were used. Two
kinds of non-XJc were also studied for comparison.
Fuel and Method
Fuel
Four kinds
of XJc (TMc, TWc,
SHc, and SEc) and two kinds of non-XJc (SMc and DTc) were used. The
six kinds of coal were characterized to obtain the results of proximate
and ultimate analyses, ash compositions, and ash fusibility, according
to Chinese standards including GB/T 212-2008, GB/T 476-2001, GB/T
1574-2007, and GB/T 219-2008. As required by these standards, the
coals were ashed at 815 °C. The analysis results of coal are
given in Table .
Table 1
Properties of Coal Used in This Studya
coal
TMc
TWc
SHc
SEc
SMc
DTc
proximate analysis
(wt %, air-dry basis)
fixed carbon
55.48
49.94
45.27
43.86
55.03
47.13
volatile
27.02
26.12
34.06
30.46
32.51
29.75
ash
3.16
9.88
5.03
14.66
4.76
18.39
water
14.34
14.06
15.64
11.02
7.70
4.73
LHV (MJ/kg)
23.70
21.84
17.63
17.93
27.68
23.72
ultimate analysis
(wt %, air-dry basis)
C
64.54
59.26
54.41
51.54
71.24
61.62
H
3.02
2.82
1.70
2.36
4.11
3.64
O
13.97
12.30
22.13
19.73
10.86
10.17
N
0.52
0.68
0.69
0.58
0.87
0.98
S
0.45
1.02
0.40
0.11
0.36
0.47
Cl
0.056
0.080
0.104
1.138
not detected
not detected
ash
compositions
(wt %)
SiO2
3.73
20.81
17.24
41.98
28.40
49.54
Al2O3
6.16
10.54
11.90
17.59
12.95
34.88
Fe2O3
5.37
29.70
5.76
6.78
10.62
4.34
CaO
33.45
11.36
28.74
19.39
30.60
5.64
MgO
5.42
4.51
5.34
2.49
1.05
0.38
TiO2
0.41
0.60
0.60
0.98
0.57
1.32
SO3
29.34
14.91
19.58
1.82
10.66
2.78
P2O5
0.00
0.04
0.05
0.18
0.20
0.10
K2O
0.45
0.89
0.38
0.66
0.42
0.31
Na2O
7.28
2.71
3.92
4.38
1.40
0.10
others
8.39
3.93
6.49
3.75
3.13
0.61
ash
fusion temperatures
(AFTs, °C)
DT
1360
1110
1320
1120
1200
1270
ST
1370
1130
1320
1130
1210
1470
HT
1370
1150
1330
1140
1220
>1500
FT
1380
1170
1340
1150
1240
>1500
LHV, low heating value.
LHV, low heating value.Results show that all four kinds of XJc’s are rich in AAEMs.
Even so, the measured AFTs of TMc and SHc fail to accurately reflect
their actual slagging behaviors. Similar to XJc, SMc is also a low-rank
coal but contains less alkali metals. During the early use, SMc had
demonstrated a certain slagging tendency,[26,27] while DTc is widely considered to be free from ash-related problems.
Ashing Test and Sample Characterization
The ashing test was conducted at 500–1000 °C, at a
temperature interval of 100 °C. Each coal sample underwent a
slow heating process. For instance, the temperature program in the
case of 800 °C was as follows. In the air, the coal sample was
first heated from room temperature to 400 °C within 1 h and the
temperature of 400 °C lasted for 2 h; then, the temperature was
raised to 600 °C within 1 h and the ashing process was kept at
600 °C for 1 h; after that, the temperature continued to be raised
to 800 °C within 1 h and the residual was further ashed for 4
h.Ash samples were characterized using an X-ray fluorescence
spectrometer (XRF, XRF-1800, Japan) and an X-ray diffractometer (XRD,
Panalytical, The Netherlands) to determine the ash compositions and
crystalline phases, respectively.
Thermodynamic
Equilibrium Calculation
In this study, the software of Factsage
7.2 was employed. Since the
ashing process of 500 °C sufficiently burned the coal out and
effectively reduced the escape of alkali metals,[28] the data of ash compositions of the coal ashed at 500 °C
were input for calculations. The databases of FactPS, FToxid, FTsalt,
and FTmisc were selected, as well as the product species of ideal
gas, pure liquids, pure solids, and corresponding solution phases.
The calculation was conducted at 600–1500 °C with an interval
of 25 °C, at 0.1 MPa. Although the reaction atmosphere plays
an important role,[10] the calculation was
carried out in an inert atmosphere according to ref (24). Via the calculation,
the evolution of the mineral phase in the ash prepared at temperatures
of 500 and 800 °C was compared, and the influence of ash composition
was evaluated.
Results and Discussion
Influence of Ashing Temperature on the Release
of Ash
Figure shows the ash yield of experimental coal as a function of ashing
temperature. For the four kinds of XJc, the experimental ash yields
gradually reduce as the ashing temperature increases. The apparent
downward trend indicates that the ashes are highly reactive and unsteady,
even at the temperature below 815 °C. However, for non-XJc, the
changes in ash yield are significant only in the narrow temperature
range of 600–700 °C and very limited in the range of 700–1000
°C. When the temperature exceeds 1000 °C, the ash yields
recover the apparent downward trend. A similar trend was also observed
in the study of Yao et al.[29] They considered
that the weight loss was assigned to the oxidation of unburned and
encapsulated char because at such high temperatures the melting of
ash allowed oxygen diffusion. However, in this study, the huge weight
loss is only attributed to the ash escape. It should be noted that
even at a temperature as high as 1200 °C, the ash yield of DTc
is still up to 93%, indicating that DTc ash is thermally inert. Hence,
mechanisms of ash loss are highly different between XJc and non-XJc.
Figure 1
Ash yields
of experimental coals as a function of ashing temperature.
The ash yield at 500 °C is normalized as 100%.
Ash yields
of experimental coals as a function of ashing temperature.
The ash yield at 500 °C is normalized as 100%.Figure shows
the
ash composition of coal ash by XRF as a function of ashing temperature.
According to the variations of content of ash composition with temperature,
the thermal behaviors of calcium, sodium, sulfur, and chlorine in
the ashing process shall be focused to reveal the ash loss mechanisms.
As indicated by XRD patterns in Figure , these species mainly exist in the raw coal (referring
to the coal ash prepared at 500 °C) in forms of anhydrite (CaSO4), calcite (CaCO3), halite (NaCl), and sodium sulfate
(Na2SO4). As the ashing temperature increases
from 500 to 700 °C, the diffraction peak of calcite gradually
disappears, which is extremely consistent with the decomposition feature
of calcite.[2] It is inferred that the decomposition
of calcite chiefly contributes to the ash loss of experimental coals
excluding TWc at the corresponding temperature range. For XJc, with
an increase in ashing temperature, the lime (CaO) decomposed from
calcite rapidly reacts with other minerals, forming gehlenite (Ca2Al2SiO7), diopside (Ca2MgSi2O6), etc. However, for non-XJc, such an evolution
of lime seems not to occur or to be very weak since the apparent peak
of lime still exists.
Figure 2
Contents of some typical ash compositions as a function
of ashing
temperature: (a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and (f)
DTc. The contents presented have been normalized by the ash yield
shown in Figure ,
i.e., the values are the product of the ash yield and the measured
content by XRF; for better presentation, the inert (Al2O3 and SiO2) and low-content (below 1%, including
Cl, K2O, TiO2, etc.) components are not plotted.
Figure 3
XRD patterns of experimental coal ashes at different ashing
temperatures:
(a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and (f) DTc. A, anhydrite
_CaSO4; Ss, sodium sulfate_Na2SO4; Sd, sodium disilicate_Na2Si2O5; C, calcite_CaCO3; L, lime_CaO; He, hemalite_Fe2O3; Q, quartz_SiO2; D, diopside_Ca2MgSi2O6; G, gehlenite_Ca2Al2SiO7; Ha, halite_NaCl; M, merwinite_Ca2Mg(SiO4)2.
Contents of some typical ash compositions as a function
of ashing
temperature: (a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and (f)
DTc. The contents presented have been normalized by the ash yield
shown in Figure ,
i.e., the values are the product of the ash yield and the measured
content by XRF; for better presentation, the inert (Al2O3 and SiO2) and low-content (below 1%, including
Cl, K2O, TiO2, etc.) components are not plotted.XRD patterns of experimental coal ashes at different ashing
temperatures:
(a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and (f) DTc. A, anhydrite
_CaSO4; Ss, sodium sulfate_Na2SO4; Sd, sodium disilicate_Na2Si2O5; C, calcite_CaCO3; L, lime_CaO; He, hemalite_Fe2O3; Q, quartz_SiO2; D, diopside_Ca2MgSi2O6; G, gehlenite_Ca2Al2SiO7; Ha, halite_NaCl; M, merwinite_Ca2Mg(SiO4)2.Figure clearly
shows that contents of both sodium and chlorine gradually decrease
with temperature for TMc, SHc, and SEc. At 900 °C, there was
almost no chlorine in ash. To date, the occurrence modes and transformation
behaviors of sodium and chlorine in the low-rank coal have always
been of concern. In the coexistence of sodium and chlorine, the two
elements may exist and release in the form of NaCl.[15] XRD patterns in Figure d strongly confirm such a behavior of sodium and chlorine
in SEc.Figure shows the
molar ratio of Na/Cl released on the basis of the ash prepared at
500 °C. For facilitating analysis, molar ratios of Na/Cl presented
in the ash prepared at 500 °C are also presented. It is found
that both molar ratios of Na/Cl exceed the value of 1. In theory,
if there exist more sodium and chlorine in the form of NaCl, the ratio
will be closer to 1. For the three coals presented in Figure , the deviation of the ratio
away from 1 basically follows the descending order of TMc, SHc, and
SEc. According to the deviation, it is inferred that for TMc and SHc
there may still be a considerable amount of sodium and chlorine existing
in other forms besides NaCl, while for SEc, most of the sodium and
chlorine may be present in the form of NaCl. The observation of cubic-shaped
NaCl crystal particles on the ash deposition probe verified that the
two elements migrated in the form of NaCl.[30]Figure can be explained
by our previous findings[23] that the content
of NH4Ac-soluble sodium in XJc follows the descending order
of TMc (36.2%), SHc (12.4%), and SEc (7.9%). The NH4Ac-soluble
sodium in coal exists as ion-exchangeable cations associated with
the carboxyl groups, forming part of the organic coal substance.[31] At high temperatures, such sodium will be transformed
into atoms,[32] oxides, and carbonates[17] and subsequently released into the gas, resulting
in the ratio released away from the value of 1.
Figure 4
Molar ratio of Na/Cl
released on the basis of the ash prepared
at 500 °C as a function of temperature: (a) TMc, (b) SHc, and
(c) SEc. The contents of Cl in TWc ashes are too low, below 1%; thus,
the relevant results for TWc are not presented.
Molar ratio of Na/Cl
released on the basis of the ash prepared
at 500 °C as a function of temperature: (a) TMc, (b) SHc, and
(c) SEc. The contents of Cl in TWc ashes are too low, below 1%; thus,
the relevant results for TWc are not presented.At the temperature above 700 °C, the peak of sodium disilicate
(Na2Si2O5) emerges, as shown in Figure a,c. It indicates
that sodium disilicate is easily formed. Actually, sodium disilicate
can be generated via the reaction between silica (SiO2)
and NaCl in the presence of moisture,[17] releasing gaseous chlorine in the form of HCl. However, in such
an ashing process, moisture is absent, and thus this reaction is impossible.
A possible reaction generating sodium disilicate is the one between
silica and Na2SO4.[3] Sodium disilicate and sodium sulfate have a low melting point, 874
and 884 °C, respectively. These low-melting-point components
make XJc face a potential risk of slagging.Different from TMc,
SHc, and SEc, the ash loss of TWc is apparently
caused by the decrease of sulfur rather than sodium and chlorine. Figure b shows that as the
temperature increases from 500 to 1000 °C, the content of SO3 reduces by almost 74%. Thus, the sulfur in TWc is thermally
unsteady. Such sulfur-bearing components include sulfides, sulfur,
organic sulfur,[33,34] etc. However, these components
are not detected by XRD analysis in this study.The above results
imply that when the ashing temperature of 815
°C is employed, for XJc, a lot of ash, mainly composed of sodium,
chlorine, sulfur, and calcium, escapes. Hence, the test results based
on the ash prepared at 815 °C cannot truly characterize the XJc.
There is of course a certain gap between the prediction of slagging
using these data and the actual situation. However, during the calculation,
all phases are kept in an equilibrium state under the principle of
minimum Gibbs free energy, i.e., the interaction among minerals is
adequate. As shown in Figure , the calculated ash yields are apparently higher than the
experimental ones and basically remain constant at temperatures below
1000 °C. This confirms that during the calculation the low-temperature
release of minerals is considerably relieved, which is a prerequisite
for subsequent evaluation of slagging characteristics using the calculating
results. More detailed discussions will be made later.
Slagging Characteristics Based on the Liquid
Ratio
Here, we compared two cases of ashing temperatures
of 500 and 800 °C to illustrate the influence of ashing temperature.
The liquid ratio, which is defined as the mass percentage of the liquid
phase (including liquid slags and liquid pure compounds) in the residual
slags, is adopted to evaluate the slagging characteristics.[10,25] The relevant results calculated via Factsage 7.1 are shown in Figure .
Figure 5
Liquid ratio–temperature
curves of the ash prepared at 500
and 800 °C: (a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and
(f) DTc.
Liquid ratio–temperature
curves of the ash prepared at 500
and 800 °C: (a) TMc, (b) TWc, (c) SHc, (d) SEc, (e) SMc, and
(f) DTc.As illustrated in Figure , the liquid ratio in the case
of 500 °C is basically
higher than that in the case of 800 °C, especially for XJc. Hence,
when a higher ashing temperature is used, the predicting slagging
characteristics of XJc will be less severe than the actual situations.
The differences between the two liquid ratio–temperature curves
reflect the importance of released species to slagging. For the cases
of XJc, the two curves significantly differ, but for the cases of
non-XJc, the trends of the two curves are very similar. According
to this point, it is believed that the released species play an important
role in promoting slagging for XJc but not for non-XJc. Through further
observation, it is found that for XJc the main differences between
the two curves (500 and 800 °C) exist in the temperature ranges
of 800–1350 °C (TMc), 900–1150 °C (TWc), 1000–1450
°C (SHc), and 800–1275 °C (SEc). Therefore, when
using ash compositions of the ash prepared at 815 °C, a lot of
information important to understand the low-temperature slagging mechanism
of XJc will be missed. In other words, the calculating results based
on the ash preparation method specified by Chinese standards are questionable
before further confirmation.It is also seen in Figure that the diversity between
the cases of 500 and 800 °C
varies considerably among the four kinds of XJc’s. This may
be explained by the differences in the amount and species of ash composition
released. For instance, for the case of TWc, a small difference corresponds
to a low ash loss (see Figure ). The detailed influence of ash composition released will
be discussed in Section .Slagging characteristics of XJc will next be analyzed
based on
the liquid ratio–temperature curves of the case of 500 °C.
According to the curves, it is found that the liquid ratio of TMc
rapidly increases to about 50% at 900 °C. For TWc, the starting
temperature with a rapid growth in liquid phases is 1100 °C.
Within a temperature increase of only 50 °C, the liquid ratio
increases by almost 43%, from 7.2 to 51.2%. For SHc, the temperature
range corresponding to a fast increase in the liquid phase is 900–1000
°C. In this temperature range, the ratio increases from 4 to
31%. Unlike the other three kinds of XJc’s, SEc is not characterized
by a low-temperature zone with a fast increase in the liquid ratio.
However, at a temperature as low as 800 °C, the liquid phase
remains at a high level, accounting for up to 18%, and gradually increases
until a steep rise at 1275 °C. Based on these results mentioned
above, it is predicted that all four kinds of XJc’s are characterized
by severe low-temperature slagging characteristics.To verify
the rationality of aforementioned predictions, the liquid
ratio–temperature curves are further analyzed combining the
reported experimental results of TMc and SHc. Figure a shows the variation of bed temperature
with time during circulating fluidized-bed gasification (CFBG) of
TMc.[23] During the process of CFBG, the
melting substances will adhere to the surface of the bed material
or ash particles, resulting in a phenomenon of slagging, named agglomeration.[35] In this study, agglomeration is considered as
a phenomenon of slagging. To better explain our calculating results,
agglomeration is assumed to mainly result from the molten ash, and
the role of bed materials is not considered. One signal of agglomeration
is the drastic fluctuation of bed temperature.[23] If timely measures are not taken, the agglomeration will
further develop into defluidization, accompanied by a rapid rise of
the bed temperature,[18] like Test III plotted
in Figure a. Following
this rule, it can be inferred that the CFBG process of TMc is free
from agglomeration at ∼850 °C (Test I) but undergoes this
event at above 900 °C (Test II). At 950 °C (Test III), defluidization
occurs when the CFBG process operates for less than 2 h. Apparently,
the higher temperature accelerates the formation of melting substances.
Such a slagging situation of TMc during the CFBG process is consistent
with the variation of the liquid ratio with temperature, partly confirming
the feasibility of predicting slagging characteristics based on the
liquid ratio.
Figure 6
Operational situations during the circulating fluidized-bed
gasification
of XJc. (a) Variation of bed temperature with time for TMc and (b)
weight of slags deposited on the slagging probe for SHc. More detailed
information can be found in our previous studies.[23,36]
Operational situations during the circulating fluidized-bed
gasification
of XJc. (a) Variation of bed temperature with time for TMc and (b)
weight of slags deposited on the slagging probe for SHc. More detailed
information can be found in our previous studies.[23,36]Figure b shows
the weight of slags deposited on the slagging probe during the CFBG
of SHc.[36] The weight of slags deposited
directly reflects the slagging degree. According to the weight of
slags, it is inferred that the CFBG process of SHc is free from slagging
at 900 °C because the weight of slags is almost zero. At 950
°C, 1.3 g of slags indicates the presence of slight slagging.
As the operating temperature further increases to 1000 °C, the
weight of slags drastically rises to 5.6 g, indicating that more melting
substances are generated in the temperature range of 950–1000
°C. Such a slagging situation of SHc also agrees with the relevant
results of the liquid ratio.Through further analysis of Figure , it is deduced that
the key temperature causing a
sharp generation of the liquid phase for TMc and SHc is within 850–900
and 950–1000 °C, respectively. In Figure , the two temperature intervals separately
correspond to liquid ratio ranges of 27–50 and 18–32%.
Considering the inherent close relationship between the liquid ratio
and slagging, there seems to be a universal liquid ratio such as 30%,
below which the reactor can maintain a long-term steady operation.
Influence of Release of Ash on the Liquid
Ratio
As mentioned above, the release of certain ash compositions
is the main reason for the differences in slagging characteristics
between the cases of 500 and 800 °C. For TMc, SHc, and SEc, the
ash released primarily includes Na2O, CaO, and Cl, while
for TWc, SO3 contributes to the ash loss. To reveal the
separate effect of each volatile ash composition on slagging characteristics,
the liquid ratio–temperature curves as a function of their
contents are calculated, as shown in Figure . Note that the effect of CaO will be discussed
as a mineral additive in Section ; thus, the result of CaO is not presented in Figure .
Figure 7
Three-dimensional color
mapping surfaces of the liquid ratio for XJc, for which independent
variables are the temperature and content of volatile ash composition:
(a) Na2O_TMc, (b) Cl_TMc, (c) SO3_TWc, (d) Na2O_SHc, (e) Cl_SHc, (f) Na2O_SEc, and (g) Cl_SEc.
The two black lines outlined on the surface indicate the two liquid
ratio–temperature curves of the ash in which the content of
ash composition is equal to the corresponding value in the ash prepared
at 500 and 800 °C, respectively. The input content of other ash
compositions has been normalized based on the corresponding content
in the ash prepared at 500 °C.
Three-dimensional color
mapping surfaces of the liquid ratio for XJc, for which independent
variables are the temperature and content of volatile ash composition:
(a) Na2O_TMc, (b) Cl_TMc, (c) SO3_TWc, (d) Na2O_SHc, (e) Cl_SHc, (f) Na2O_SEc, and (g) Cl_SEc.
The two black lines outlined on the surface indicate the two liquid
ratio–temperature curves of the ash in which the content of
ash composition is equal to the corresponding value in the ash prepared
at 500 and 800 °C, respectively. The input content of other ash
compositions has been normalized based on the corresponding content
in the ash prepared at 500 °C.It is clearly seen from Figure a,d that as the content of Na2O increases,
for TMc and SHc, the liquid ratio is on the rise. By further comparing
the two black lines outlined on the surface, the rising trend is relatively
apparent at temperatures above about 900 °C. Such an influence
of Na2O matches the results shown in Figure a,c, indicating the important role of sodium
in promoting slagging.By contrast, for SEc, the influence of
Na2O is worth
attention. The biggest difference to cases of TMc and SHc is that
when the Na2O content decreases, the liquid ratio increases
at low temperatures below 800 °C; see Figure f. This special trend may be attributed to
the special coal property of SEc, the feature of a high content of
chlorine. In the two cases of 500 and 800 °C, the molar ratios
of Na/Cl in the ash are 1.35 and 0.41, respectively. According to
the two ratios, chlorine is considered to be more inclined to exist
in the form of NaCl in the case of 500 °C and in the form of
other metal chlorides, such as CaCl2, MgCl2,
etc., in the case of 800 °C. It is known that the melting points
of alkaline earth metal chlorides are generally lower than that of
NaCl (801 °C). For instance, the melting points of CaCl2 and MgCl2 are 772 and 714 °C, respectively, and
FeCl2 melts at 670–674 °C. Hence, when the
Na2O content in SEc decreases, there will be more alkaline
earth metal chlorides. The lower-temperature melting of metal chlorides
is responsible for the higher liquid ratio at a low temperature.For TWc, the decrease in the SO3 content results in
a decrease in the liquid ratio, as indicated by Figure c. The decreasing trend of the liquid ratio
mainly exists at 900–1100 °C, coinciding with the results
of Figure b.In Figure b,e,g,
it is seen that the liquid ratio increases overall with the decreasing
content of chlorine, especially at temperatures above 900 °C.
Generally, chlorides have strong volatility. As the content of chlorine
increases, more alkali and alkaline earth metals, important fluxing
materials, prefer to be released into the gas phase in the form of
chlorides. Results of Figure that more gaseous species are released with an increase in
both temperature and content of chlorine confirm this view. However,
such an influence of chlorine is contrary to that of Na2O and SO3.
Figure 8
Percent of gaseous species released as a function of temperature
at different input contents of chlorine: (a) TMc, (b) SHc, and (c)
SEc.
Percent of gaseous species released as a function of temperature
at different input contents of chlorine: (a) TMc, (b) SHc, and (c)
SEc.Besides, as indicated by Figure , with an increase
in ashing temperature (namely, with
a decrease in the chlorine content), the liquid ratio overall decreases.
There is a contradiction between Figures and 5. It is because
the influence of a single component is one-sided and the final slagging
characteristics are the integrated result of all components in the
ash.
Evolution of Mineral Phase with Temperature
Figure shows the
phase assemblage–temperature curves for XJc, which are calculated
using the ash compositions of the ash prepared at 500 °C as input
data. The input ash compositions will be converted into corresponding
solid, liquid, and gas phases, following the principle of minimum
Gibbs free energy. According to the evolution of phase assemblage,
the detailed slagging mechanisms of XJc can be revealed.
Figure 9
Phase assemblage–temperature
curves for XJc calculated from
the ash compositions of the ash prepared at 500 °C: (a) TMc,
(b) TWc, (c) SHc, and (d) SEc. The minerals given in these figures
exclude those with a maximum content below 5%. The phases of SLAG
and GAS in figures refer to the liquid slags and the gaseous species
released, respectively.
Phase assemblage–temperature
curves for XJc calculated from
the ash compositions of the ash prepared at 500 °C: (a) TMc,
(b) TWc, (c) SHc, and (d) SEc. The minerals given in these figures
exclude those with a maximum content below 5%. The phases of SLAG
and GAS in figures refer to the liquid slags and the gaseous species
released, respectively.As shown in Figure a, it is inferred
that the low-temperature slagging mechanism of
TMc includes two parts, fast growth of liquid slags at 800–825
°C and melting of Na2SO4 at 884 °C.
According to the evolution of minerals, the liquid slags (the phase
of SLAG presented in Figure ) mainly originate from low-temperature eutectics, composed
of Na2Ca8Al6O18, Ca2Fe2O5, CaSO4, and NaCl (the
maximum content of NaCl is too low, only 3.3%; thus, NaCl is not plotted
in Figure a). By further
analysis of the composition of liquid slags, with the calculating
result at 850 °C as an example, the compositions with a content
above 3% in liquid slags include CaO (24.3%), CaSO4 (21.3%),
Al2O3 (17.3%), NaCl (12.8%), Fe2O3 (7.9%), Fe2(SO4)3 (7.2%),
Na2O (3.7%), and Na2SO4 (3.0%). CaO
and Al2O3 are generally refractory oxides, not
conducive to slagging. However, under the action of alkali metal components
such as NaCl, Na2O, and Na2SO4, these
refractory oxides are dissolved. The second fast growth of liquid
slags occurs at above 1400 °C, and the fluxing materials chiefly
originate from CaSO4, Ca2SiO4, etc.
Meanwhile, the mass fraction of the GAS (the gaseous species released
from residual slags) starts to evidently increase. The released components
primarily contain chlorine, sodium, oxide, sulfur, and potassium,
which are commonly considered as crucial components causing the fouling.[37]It is seen from Figure b that for TWc Na2SO4 also plays an
important role in promoting slagging. The self-melting of Na2SO4 is inferred to be responsible for the high liquid
ratio at 900–1000 °C presented in Figure b. After that, the molten Na2SO4 may react with Mg2Al4Si5O18, generating Mg2SiO4 and NaAlSi3O8. NaAlSi3O8 has a melting
point of 1100 °C. At above 1100 °C, the molten NaAlSi3O8 gradually dissolves other solid minerals including
CaSO4, Mg2SiO4, and residual Mg2Al4Si5O18. At this moment,
the liquid slags start to form fast. At 1100–1200 °C,
the liquid slags increase from almost 0 to 48.6%. The calculating
results show that at 1200 °C, the main components in the liquid
slags consist of SiO2 (36.3%), Al2O3 (19.6%), CaO (10.9%), Na2O (8.1%), MgO (7.4%), CaSO4 (6.0%), MgSO4 (5.0%), Na2SO4 (4.2%), etc. Overall, Na2SO4 initiates the
whole low-temperature slagging process. The mass fraction of GAS begins
to rise at above 1000 °C. At 1500 °C, the GAS accounts for
23.7%. Actually, the GAS mainly consists of two elements, oxygen and
sulfur, accounting for above 97%. The molar ratio of O/S is 3.01,
indicating that the GAS may originate from the decomposition of sulfates
and the two elements most likely exist in the form of SO3. The calculating results confirm the strong volatility of sulfur
in TWc, agreeing with the variation of content of SO3 in
ash with ashing temperature shown in Figure b.As indicated by Figure c, the low-temperature slagging
mechanism of SHc is also divided
into two parts, the self-melting of Na2SO4 at
884 °C and the generation of liquid slags at above 1000 °C.
Such a slagging mechanism is similar to that of TMc. According to
the evolution of phases, the rapid growth of liquid slags from 3.8
to 17.7% at ∼1000 °C mainly results from the interaction
of Na2Ca3Al16O28, Ca2Fe2O5, etc. As the temperature further
increases, Na2Ca3Al16O28, Ca3MgSi2O8, Ca2Al2SiO7, and CaSO4 successively dissolve,
further promoting the growth of liquid slags. At 1375–1500
°C, the mass fraction of GAS gradually increases to 12.6%. In
the GAS, the main elements include oxide, sulfur, chlorine, and sodium.
Despite the content of iron being relatively high in TWc (see Table ), the iron always
exists in the form of Fe2O3, even at high temperatures,
which is consistent with the XRD patterns in Figure b and other reported results.[38]Due to the high content of NaCl, the slagging
characteristics of
SEc are certainly different from those of the other three kinds of
XJc. As shown in Figure d, NaCl contributes to the first fast growth of liquid slags at below
800 °C. In the liquid slags, the mass fraction of molten NaCl
accounts for 88.6%. In the temperature range of 925–950 °C,
mass percentages of NaAlSiO4 and Ca2Al2SiO7 start to increase. The two phases may be generated
by reactions between Ca3Fe2Si3O12 and Na2Ca3Al16O28. Although NaAlSiO4 and Ca2Al2SiO7 have high melting points, 1520 and 1593 °C, respectively,
they are dissolved into the liquid slags at 1275–1300 °C,
thereby increasing the liquid slags from 26.2 to 94.9%. This is a
typical low-temperature eutectic phenomenon of XJc. In addition, numerous
components in SEc start to release at ∼1300 °C. At 1500
°C, the mass fraction of GAS is up to 23.1%. Obviously, the gaseous
NaCl is dominant in the GAS, accounting for a volume fraction of 55.1%.
Other gas components mainly include SO3, (NaCl)2, KCl, etc.
Influence of Refractory
Oxide on Antislagging
Refractory oxides in coal usually play
a role in improving resistance
to ash fusibility, which provides a guide to solving the ash-related
problems during coal thermal conversion by blending coal or adding
mineral additives.[39,40] In this study, the liquid ratios
at varying temperatures and varying contents of CaO, Al2O3, and SiO2 are calculated. Relevant results
are shown in Figure .
Figure 10
Three-dimensional color mapping surfaces of the liquid ratio for
XJc, for which independent variables are temperature and content of
refractory oxide: (a.1) CaO_TMc, (a.2) Al2O3_TMc, (a.3) SiO2_TMc, (b.1) CaO_TWc, (b.2) Al2O3_TWc, (b.3) SiO2_TWc, (c.1) CaO_SHc, (c.2)
Al2O3_SHc, (c.3) SiO2_SHc, (d.1)
CaO_SEc, (d.2) Al2O3_SEc, and (d.3) SiO2_SEc. The black line outlined on the surface indicates the
liquid ratio–temperature curve of the ash prepared at 500 °C.
Three-dimensional color mapping surfaces of the liquid ratio for
XJc, for which independent variables are temperature and content of
refractory oxide: (a.1) CaO_TMc, (a.2) Al2O3_TMc, (a.3) SiO2_TMc, (b.1) CaO_TWc, (b.2) Al2O3_TWc, (b.3) SiO2_TWc, (c.1) CaO_SHc, (c.2)
Al2O3_SHc, (c.3) SiO2_SHc, (d.1)
CaO_SEc, (d.2) Al2O3_SEc, and (d.3) SiO2_SEc. The black line outlined on the surface indicates the
liquid ratio–temperature curve of the ash prepared at 500 °C.Figure a.1–a.3
shows the 3D color mapping surfaces of liquid ratios in TMcash. By
a comparison with the ash prepared at 500 °C, it is found that
varying the contents of all three oxides in TMc can regulate the slagging
characteristics of TMc. For instance, the decreasing trend of liquid
ratio with the increasing content of CaO indicates that adding CaO
(i.e., the content of CaO exceeds 32.8%) improves the resistance to
ash fusibility. The improving effect is significant, especially at
above 900 °C. Besides the high melting point (2572 °C) of
CaO, the improvement can also be achieved by prior reactions of CaO
with other coal ash (such as SiO2 and Al2O3) inhibiting the generation of sodium-based eutectics.[41] However, different situations occur for the
cases of Al2O3 and SiO2. As seen
in Figure a.2, the
proportion of the liquid phase in ash effectively reduces at temperatures
above 800 °C only when the content of Al2O3 exceeds ∼20%. Otherwise, the liquid ratio increases. The
low content of Al2O3 in the ash of TMc (only
4.7%; see Figure a)
may be responsible for the increasing liquid ratio. According to the
variation of the liquid ratio, Al2O3 plays a
role of a fluxing material within the content range of 8–20%.
As indicated by Figure a.3, at contents below ∼14%, SiO2 also performs
a fluxing function. It has been reported that the low-content addition
of silicon- and aluminum-bearing additives in the Xinjiang high-sodium
coal lowers the ash fusion temperatures.[42] For TMc, SiO2 performs the best antifluxing behavior
at the content of ∼20%. Under this circumstance, SiO2 may prefer to react with other ash compositions, forming higher-melting-point
components. However, at other contents of SiO2 (<55%),
the liquid ratio basically remains at a high level, regardless of
the content. It indicates that the antislagging measure by adding
SiO2 is invalid in most cases.As observed in Figure b.1, for TWc, largely
increasing the content of CaO is invalid.
When the content of CaO is higher than 27%, the ash fusibility is
significantly improved. It is quite contrary to the case of TMc. Figure b.2 shows that
in the temperature range of 900–1275 °C a higher content
of Al2O3 in TWc basically corresponds to a lower
liquid ratio. In this circumstance, the participation of Al2O3 contributes to the mitigation of slagging. An increase
in SiO2 is only conducive to the reduction of generating
a liquid phase at 900–1000 °C, as indicated by Figure b.3. At a higher
temperature above 1100 °C, however, adding SiO2 results
in more liquid phases.It is found that there are similarities
in variation trends of
3D color mapping surfaces between TMc and SHc, by comparing Figure a.1–a.3
and c.1–c.3. However, effects of adding refractory oxides in
antislagging are discrepant due to different contents of ash compositions
in the two ashes. As observed in Figure c.1, increasing the CaO content from 29
to 41% first largely accelerates the generation of the liquid phase,
especially at temperatures below 1300 °C. After that, although
increasing the CaO content results in a reducing liquid ratio, the
liquid ratio still remains at a high level. This indicates that CaO
plays a role of fluxing in the SHc case. Because of an insignificant
growth of the liquid phase, adding Al2O3 or
SiO2 does not cause severe slagging risk. This trend is
different from the low-temperature fluxing action of Al2O3 and SiO2 in the TMc case. Similar to the
case of TMc, however, the lowest liquid ratio also appears at the
SiO2 content of ∼20%.Adding CaO into SEc
is also unfavorable to inhibit the formation
of a liquid phase, as indicated by Figure d.1, because the participation of CaO promotes
the formation of low-temperature eutectics. Figure d.2 and d.3 indicates that Al2O3 can be used as a mineral additive for antislagging,
whereas adding SiO2 cannot achieve this purpose.
Conclusions
In the temperature range of 500–1200
°C, the yield
of coal ash decreases with an increase in the ashing temperature.
The weight loss of coal ash is mainly attributed to two parts, the
decomposition of calcite at 600–700 °C and the release
of volatile components including calcium, sodium, sulfur, chlorine,
etc. The former is largely responsible for the ash loss of high-rank
coal such as DTc, whereas the latter mainly results in the ash loss
of low-rank coal such as XJc and SMc. Compared to the case of 500
°C, at the ashing temperature of 815 °C, numerous volatile
components in XJc ash are released. Thermodynamic equilibrium calculation
shows that these released components play an important role in promoting
the formation of the liquid phase, especially at temperatures below
1200 °C. This may be a main reason for the inaccurate prediction
of slagging characteristics of XJc using the ash compositions of the
ash prepared at 815 °C. However, for non-XJc (SMc and DTc), such
an adverse influence caused by a higher ashing temperature is basically
negligible, regardless of the coal rank. By a comparison with reported
experimental results, it is found that the liquid ratio–temperature
curve calculated from the ash compositions of the ash prepared at
500 °C can reflect the actual slagging characteristics of XJc.
According to the curves, all four kinds of XJc’s are predicted
to suffer from a low-temperature slagging issue. This issue mainly
results from the self-fusion of sodium-bearing salts and the low-temperature
eutectics. The antislagging measures of adding refractory oxides greatly
differ among the four kinds of XJc’s due to the diversity of
ash compositions. In most cases, adding SiO2 (the content
of SiO2 is below 60%) seems invalid for antislagging, while
the addition of Al2O3 into XJc performs an excellent
antislagging function. In addition, adding CaO to inhibit the slagging
issue seems effective only for TMc.