Crystalline-SiO2-Component-Induced Cracking in Coal Ash Deposits Exposed to Temperature Changes.

Frequent load changes or daily shut down and restarting have been imposed on thermal power plants in many areas to fill the gap between the demand for electric power and the variable outputs from renewable energy sources. These operations result in temperature changes in boiler furnaces, and the temperature changes may induce cracks in problematic coal ash deposits, which would contribute to their spontaneous shedding from heat-transfer surfaces. In this work, we microscopically investigated the origin of temperature-change-induced cracks in ash deposits. Analyses of the coal ash clinker generated in a utility boiler indicated that many cracks were generated close to the crystalline SiO2 components, such as quartz or cristobalite. To elucidate the impact of crystalline SiO2 components on the generation of cracks, we conducted temperature cycle tests for coal ash clinkers in an electric furnace. The results suggested that cracking was promoted by temperature changes that cover the transition temperature of crystalline SiO2 components. Moreover, we successfully demonstrated that the addition of quartz additives could promote the generation of cracks significantly in coal ash clinkers. The findings in this work will advance the cracking behavior in ash deposits during temperature changes. Factors for the propagation of cracks or other factors for the shedding should be discussed in further studies, which will contribute to developing a method for promoting the shedding of ash deposits during temperature changes caused by load changes or daily shut down and restarting of thermal power plants.

Introduction

The use of variable renewable energy sources, such as photovoltaic power or wind power systems, is increasing worldwide.[1] The outputs of renewable energy sources fluctuate remarkably depending on the weather conditions.[2,3] By imposing frequent load changes or daily shut down and restarting, thermal power plants play a significant role in filling the gap between the demands for electric power and the outputs from renewable energy sources.[4−6] The load changes result in temperature changes in boiler furnaces. Herein, we focused on the impact of temperature changes on ash deposits that are generated by the combustion of coal or biofuels.[7−10] As thick ash deposits can inhibit heat transfer and lead to an unplanned temporary shutdown of power plants, the ash deposits should be removed spontaneously before their amount increases. The temperature changes induce thermal stress in the ash deposits, which may promote the generation of cracks and facilitate their shedding from the heat-transfer surfaces.[11] Therefore, understanding the impact of temperature changes on cracking is important for controlling the growth and shedding of ash deposits.

Macroscopic studies on thermal stress in ash deposits have been conducted previously.[12−16] In previous studies, the mechanical properties of ash deposits were considered to be uniform. However, the microscopic behavior of ash deposits, which is critical to the origin of cracks in other materials, is not well studied. Ash deposits are generally composed of multiple crystalline and amorphous components.[17−20] The microstructures with various components may be one of the dominating factors for cracking because local stress is generated close to the boundaries of components with different mechanical properties during temperature changes. This behavior is similar to that of glass ceramics.[21,22] Considering this, microscopic studies may provide insights into the origin of cracking in ash deposits, which is critical information to discuss the cracking behaviors during temperature changes caused by load changes or daily shut down and restarting of thermal power plants. Moreover, understanding the dominating factors for cracking from a microscopic viewpoint may contribute to developing additives that promote the generation of cracks by controlling the microstructure of ash deposits. Although various types of additives have been studied to inhibit the growth of ash deposits,[23−28] additives that induce the generation of cracks by utilizing temperature changes have not been investigated.

In this work, we focused on the crystalline SiO2 component as one of the factors for the generation of cracks, which can be naturally contained in some coal ash deposits depending on the ash chemical compositions or environments. Additionally, the crystalline SiO2 component is known to promote cracking in other materials due to its large volume change caused by the temperature-dependent transitions of the crystalline structures.[29] We investigated the microstructure of the coal ash clinker generated in a utility boiler, and this result indicates that the crystalline SiO2 components also induce cracking in ash deposits. To advance the understanding of the generation of cracks caused by the crystalline SiO2 components, we conducted temperature cycle tests for the coal ash clinkers in an electric furnace. The results suggested that the generation of cracks particularly increased when the temperature changed beyond the transition temperature of the crystalline SiO2 components. Furthermore, we demonstrated that the generation of cracks could be successfully promoted in three types of coal ash clinkers with different chemical compositions by employing quartz additives.

Experimental Section

Investigation of the Clinker Generated in the Utility Boiler

The image and chemical composition of the coal ash clinker generated on the superheater at the furnace exit in the pulverized coal-fired boiler (700 MW) are shown in Figure and Table , respectively. The clinker was physically hard and grew large. The clinker seemed to have melted in the boiler furnace, which implies that the clinker was generated by slagging.[8] The cross-section of the clinker was observed by scanning electron microscopy (SEM) to investigate the distribution of cracks. In addition, an electron probe microanalyzer was used to determine the elemental distributions around the cracks. X-ray diffraction (XRD) was performed to investigate the structures of the crystalline components. The analysis conditions for XRD and SEM/electron probe microanalysis (EPMA) are summarized in Tables S1 and S2 in the Supporting Information, respectively. Based on the combined results of all the analyses, the crucial factors for the origin of cracks were discussed.

Figure 1

Photograph of the coal ash clinker generated on the superheater at the furnace exit in the pulverized coal-fired boiler (700 MW).

Table 1

Chemical Composition (wt %) of the Coal Ash Clinker Generated on the Superheater at the Furnace Exit in the Pulverized Coal-Fired Boiler (700 MW)

clinker generated in the utility boiler
SiO2	58.26
Al2O3	20.99
Fe2O3	11.44
CaO	5.40
MgO	1.12
TiO2	1.52
P2O5	0.40
Na2O	0.11
K2O	0.66
SO3	0.09

Temperature Cycle Test

Coal ash clinker samples for the temperature cycle tests were prepared from fly ash particles generated in coal combustion tests conducted in an experimental furnace. Fly ash particles were collected from a bag filter. The chemical compositions of the fly ash samples are summarized in Table , and the results of the XRD analysis are shown in Figure S1 in the Supporting Information. Three types of fly ash particles, generated from different coals and with different chemical compositions (termed FA1, FA2, and FA3 herein), were chosen. FA1 and FA2 were generated by the coal combustion tests reported in a previous study.[17] The clinker samples were prepared by heating FA1, FA2, or FA3 on a thin alumina plate in an electric furnace (termed CL1, CL2, or CL3, respectively, herein). Additionally, clinker samples were also prepared from fly ash containing quartz additives. The X-ray diffractograms and particle size distributions of the quartz additives are shown in Figures S2 and S3, respectively, in the Supporting Information. The operating program of the electric furnace is shown in Figure . First, each fly ash sample was heated at a rate of +400 °C/h up to TM, the temperature at which the fraction of molten ash reached 70% according to thermodynamic equilibrium calculations shown in Figure S4 in the Supporting Information. Based on the results of the calculations, FA1, FA2, and FA3 were heated to 1360, 1245, and 1300 °C, respectively. The fly ash was maintained at these elevated temperatures for 1 min, following which each sample was slowly cooled at a rate of −40 °C/h to allow solidification and crystallization of the samples. Crystallization also occurs in ash deposits on the heat-transfer surfaces in actual combustion furnaces, depending on the thermal history.[17,20] The clinker samples were continuously cooled to 100 °C when the temperature cycles were not conducted [pattern (a) in Figure ]. In the other cases, the cooling process was suspended, and the temperature cycles were conducted, which consisted of five heating and cooling processes [pattern (b) in Figure ]. During these cycles, the clinker samples were heated and cooled at a rate of ±200 °C/h. After the temperature cycles were completed, the clinker samples were cooled to 100 °C at a rate of −40 °C/h. Images of the clinker samples after cooling are shown in Figure . XRD was performed, and the cross-sections were observed using SEM and EPMA, using the same procedure as that for the clinker generated in the utility boiler. Additionally, the lengths of cracks in these SEM images were measured, and cracks longer than 80 μm were counted using the ImageJ software to estimate the differences in the cracking behavior.[30] Moreover, thermodynamic equilibrium calculations were performed using the Factsage 7.1 software to estimate the stability of the quartz additives in FA1, FA2, and FA3 at each heating temperature (TM).[31]

Table 2

Chemical Compositions [wt %] of the Fly Ash Samples and Ash Fusion Temperatures [°C] of the Original Coal Asha

		FA1	FA2	FA3
chemical composition	SiO2	65.31	56.76	46.33
	Al2O3	22.56	23.06	21.59
	Fe2O3	5.82	6.76	11.44
	CaO	1.26	5.47	11.31
	MgO	0.91	2.31	3.99
	TiO2	0.92	0.94	1.02
	P2O5	0.21	0.92	0.42
	Na2O	0.68	1.01	1.54
	K2O	1.49	1.96	1.22
	SO3	0.83	0.82	1.13
ash fusion temperature of the original coal ash (oxidizing atmosphere)	IT	1370	1205	1190
	HT	1465	1280	1220
	FT	>1500	1375	1270

IT: initial deformation temperature, HT: hemispherical temperature, and FT: flow temperature. The ash fusion temperatures were measured based on the Japanese Industrial Standards (JIS M 8801).

Figure 2

Operating program of the electric furnace. TM: heating temperature for ash melting, TH: heating temperature during temperature cycles, TL: cooling temperature during temperature cycles, and TR: room temperature.

Figure 3

Photographs of the coal ash clinkers generated in the electric furnace (after cooling).

Results

Distribution of Cracks in the Clinker Generated in the Utility Boiler

Figure a shows the SEM image and elemental distribution of the cross-section of the clinker generated in the utility boiler. Cracks were obviously generated around the Si-rich component with a diameter of approximately 40 μm. Moreover, as shown in Figure b, many cracks were also generated in or around the fine Si-rich components. The Si-rich components were determined to be crystalline SiO2 components (quartz or cristobalite) based on the XRD analysis (Figure ). These results suggest that the crystalline SiO2 components in coal ash deposits have a significant impact on cracking, similar to other materials.[29]

Figure 4

(a) SEM image and elemental maps of Si, Al, Fe, Ca, Mg, Ti, K, and Na for a cross-section of the coal ash clinker generated in the utility boiler (700 MW). The main crack in the SEM image is shown by a red circle. (b) SEM images for views 1, 2, and 3 in (a).

Figure 5

X-ray diffractogram of the coal ash clinker generated in the utility boiler (700 MW). A: anorthite, C: cristobalite, H: hematite, M: mullite, and Q: quartz.

Distribution of Cracks in the Clinker after Temperature Cycles

The XRD analyses of FA1 and CL1 without temperature cycles (CL1-W/O) are shown in Figure . The results indicate that heating and slow cooling of FA1 could successfully crystallize the sample and that CL1-W/O contained abundant cristobalite (or traces of quartz). The SEM images and elemental distributions of the cross-sections of CL1 after temperature cycles between 500 and 700 °C (CL1-500/700) are shown in Figure . Almost all the cracks were generated close to the Si compounds in the observed areas, and these compounds could be quartz or cristobalite, based on the XRD analysis (Figure ). Furthermore, the cracking behavior of the clinker was investigated in the electric furnace. Figure , which shows the crack densities in CL1 following the loading under different temperature cycles, indicates that more cracks were generated owing to temperature cycles in the low-temperature regions.

Figure 6

X-ray diffractograms of FA1 and CL1-W/O. C: cristobalite, M: mullite, and Q: quartz.

Figure 7

SEM image and elemental maps of Si, Al, Fe, Ca, Mg, Ti, K, and Na for a cross-section of CL1-500/700. Cracks in the SEM image are shown by red circles.

Figure 8

Crack densities in CL1 after different temperature cycles.

Promotion of Cracking by Crystalline SiO2 Additives

Figure shows the results of the XRD analysis of the clinker samples prepared from fly ash with or without the quartz additives. For the clinker samples without the additives, crystalline SiO2 components (quartz or cristobalite) were detected only in CL1-W/O. By adding the quartz additives, the peak intensity of quartz in CL1-W/O and CL2-W/O increased, although some parts of the additives seemed to have transformed into cristobalite. For CL3-W/O, on the other hand, quartz was not detected in any case, and a small peak of cristobalite was detected when 20 wt % of the additives were added. The thermodynamic equilibrium calculations in Figure indicate that crystalline SiO2 components can exist in CL1 and CL2 but not in CL3 in the equilibrium states at each TM when 20 wt % of the quartz additives were added. This implies that the quartz additives in CL3-W/O were consumed through reactions with other elements, and a small amount of cristobalite remained in CL3-W/O with 20 wt % additives. Figure shows the SEM images and elemental maps of Si in CL1-500/700, CL2-500/700, and CL3-500/700; these samples contained 20 wt % of the quartz additives. In the case of CL1-500/700 and CL2-500/700, many additives remained in the clinkers, and cracks were observed close to these additives. Although fewer additives remained in CL3-500/700, cracks were observed close to the remaining additives. Figure shows the crack densities in CL1-500/700, CL2-500/700, and CL3-500/700. The crack densities in CL1-500/700 increased gradually with increasing amount of the quartz additive whereas those for CL2-500/700 and CL3-500/700 increased sharply when the amount of additive was increased to 10 and 20 wt %, respectively. Similar tendencies were observed for CL1-150/350, CL2-150/350, and CL3-150/350, as shown in Figures S5 and S6 in the Supporting Information.

Figure 9

X-ray diffractograms of CL1-W/O, CL2-W/O, and CL3-W/O with the added quartz additives. C: cristobalite, Q: quartz, and ×: alumina derived from the plates that could not be completely removed from the samples.

Figure 10

Temperature dependence of the fraction of crystalline SiO2 components in CL1, CL2, and CL3 with 20 wt % quartz additives. Throughout the heating process, crystalline SiO2 components could exist thermodynamically in CL1 (TM: 1360 °C) and CL2 (TM: 1245 °C) but not in CL3 (TM: 1300 °C) when 20 wt % of the quartz additives was added. The values were obtained from the thermodynamic equilibrium calculations.

Figure 11

SEM images and elemental maps of Si for the cross-sections of CL1-500/700, CL2-500/700, and CL3-500/700 with or without the quartz additives. Main cracks in the SEM images are shown by red circles. The SEM image and elemental map of Si for CL1-500/700 without the additives are shown in Figure .

Figure 12

Crack densities of CL1-500/700, CL2-500/700, and CL3-500/700 with or without the quartz additives.

Discussion

As shown in Figure , the generation of cracks in CL1 was promoted by the temperature cycles in the low-temperature regions. This implies that the transition of the crystalline structures of quartz or cristobalite may induce cracking. The crystalline quartz structure undergoes phase transition at 573 °C, while cristobalite undergoes phase transition in the temperature range of 200–300 °C.[29,32−35] These transitions involve large volume changes and can cause local stress. Since the temperature cycles between 500 and 700 °C and between 150 and 350 °C include the transition temperatures of quartz and cristobalite, respectively, it is likely that the transition of the crystalline structures of quartz and cristobalite in the clinker samples promotes the generation of cracks, as observed for other materials.[29] Even if the temperature cycles did not cover the transition temperatures of quartz and cristobalite, the clinker samples attained these temperatures during the cooling process, thereby inducing cracks to some extent (Figure ).

Additionally, the quartz additives successfully promoted the generation of cracks in the coal ash clinkers, as shown in Figures and 12. The drastic increase in cracks in CL2-500/700 upon the addition of the quartz additives indicates that the additives can induce the generation of cracks in ash deposits, wherein crystalline SiO2 components are not generated from their original chemical components. Moreover, a sharp increase in cracks was observed for CL3-500/700 when the amount of the quartz additives was increased from 10 to 20 wt %. This implies that the additives must remain as crystalline SiO2 components in the ash deposits to effectively promote the generation of cracks. The gradual increase in the cracks observed for CL1-500/700 may be attributed to its chemical composition, in which crystalline SiO2 components can exist without the quartz additives. The thermodynamic equilibrium calculations in Figure indicate that the quartz additives may decrease as the temperature increases, implying that some part of the quartz additives is melted during the heating process. Therefore, there may exist two kinds of crystalline SiO2 components in the coal ash clinkers to which the quartz additives were added. One is the additive that remained as the crystalline SiO2 component during the heating process without melting, and the other is the crystalline SiO2 component that was recrystallized from the molten phase during the cooling process. The ratio of the two kinds of crystalline SiO2 components may largely depend on the ash chemical composition and heating temperature, considering the results in Figure .

Figure shows the steps to extend this work to the shedding of ash deposits. As mentioned in the Introduction, the changes in operating conditions of boiler furnaces cause temperature changes, and ash deposits on boiler tubes are exposed to temperature cycles. The crystalline SiO2 components in ash deposits induce the generation of cracks, as shown in this work. Additionally, the generation of cracks can be promoted by utilizing quartz additives even if the crystalline SiO2 components cannot be generated from the original chemical compositions of ash deposits.

Figure 13

Schematic of the steps to extend this work to the shedding of ash deposits.

To extend this work to the shedding of ash deposits, the propagation of cracks should also be induced as shown in Figure . In addition to the origin of cracks, the dominant factors for the propagation of cracks should be discussed from a microscopic viewpoint in future studies. Moreover, other factors, such as gravity force and gas flow, should be considered to develop a method for promoting the shedding of ash deposits in the actual environments of utility boilers. Furthermore, we should carefully consider the effect of the crystalline SiO2 components on other problems such as the erosion of boiler tubes[36] to discuss a method for promoting the shedding of ash deposits based on this work.

Conclusions

In this work, we microscopically investigated cracks in coal ash clinkers to advance the understanding of the influential factors for the origin of temperature-change-induced cracks. Especially, we focused on the crystalline SiO2 components, which are known to induce the generation of cracks in other materials because of the temperature-dependent transitions of the crystalline structures involving a large volume change.

Many cracks were observed close to the crystalline SiO2 components in the coal ash clinker obtained in the utility boiler. The temperature cycle tests in the different temperature ranges with the electric furnace indicated that the generation of cracks was induced by the temperature cycles that covered the phase transition temperatures of the crystalline SiO2 components, such as quartz and cristobalite. These results imply that the phase transitions of the crystalline SiO2 components have significant effects on the generation of cracks in ash deposits during temperature changes. Moreover, we demonstrated that crystalline SiO2 additives could promote the generation of cracks even if crystalline SiO2 components are not generated in ash deposits from their original chemical compositions. These findings will advance the understanding of the origin of cracks in ash deposits during temperature cycles. Further studies on the propagation of cracks or other factors for the shedding should be conducted, which will contribute to developing a method for promoting the shedding of ash deposits during temperature changes caused by load changes or daily shut down and restarting of thermal power plants.