Corrosive Effect of Wood Ash Produced by Biomass Combustion on Refractory Materials in a Binary Al–Si System

In terms of its chemical composition, biomass is a very complex type of fuel. Its combustion leads to the formation of materials such as alkaline ash and gases, and there is evidence of the corrosive effect this process has on refractory linings, thus shortening the service life of the combustion unit. This frequently encountered process is known as “alkaline oxidative bursting”. Corrosion is very complex, and it has not been completely described yet. Alkaline corrosion is the most common cause of furnace-lining degradation in aggregates that burn biomass. This article deals with an experiment investigating the corrosion resistance of 2 types of refractory materials in the Al2O3-SiO2 binary system, for the following compositions: I. (53 wt.% SiO2/42 wt.% Al2O3) and II. (28 wt.% SiO2/46 wt.% Al2O3/12 wt.% SiC). These were exposed to seven types of ash obtained from one biomass combustion company in the Czech Republic. The chemical composition of the ash is a good indicator of the problematic nature of a type of biomass. The ashes were analyzed by X-ray diffraction and X-ray fluorescence. Analysis confirmed that ash composition varies. The experiment also included the calculation of the so-called “slagging/fouling index” (I/C, TA, Sr, B/A, Fu, etc.), which can be used to estimate the probability of slag formation in combustion units. The corrosive effect on refractory materials was evaluated according to the norm ČSN P CEN/TS 15418, and a static corrosion test was used to investigate sample corrosion.


Introduction
Worldwide, 80% of electricity is produced using fossil fuels. According to the International Energy Agency (IEA), electricity production reached approximately 25.8 T-kWh in 2020, and an increase to 36.5 T-kWh is expected by 2040 [1]. As stated by the World Bioenergy Association, 59.2 NPP/year, i.e., 10.3% of the global energy supply, comes from biomass [2]. Biomass is becoming a popular source of energy which can be used in various ways. Electricity produced from biomass currently corresponds to 493 TWh, which is approximately 2% of the world's electricity production [2]. Using biomass as a raw material for powerplants is certainly interesting and useful; however, this technology also has certain disadvantages. The use of biomass in powerplants leads to the formation of residue, called biomass ash. It is estimated that around 480 million tons of ash are produced every year by biomass powerplants worldwide. This is similar to coal ash, with 780 million tons per year [2].
The most frequently burned material is wood (64%), followed by cereals and plant residue from agricultural production. In general, it can be said that the average percentage of ash produced by burning biomass ranges between 1 and 6%; for wood, it is 0.6-1.6%; for bark, it rarely exceeds 3%; straw produces an ash content of around 5%, while grass produces 7%. At the other end of the spectrum, the ash content produced by black coal is significantly higher, reaching 20-30%, and from brown coal, this amount can be even greater [3]. Ash represents a variable composition of mineral and inorganic components. significantly higher, reaching 20-30%, and from brown coal, this amount can be even greater [3]. Ash represents a variable composition of mineral and inorganic components.
During the combustion process, ash continuously changes its physical and chemical properties, the final product being a molten mixture of original minerals, various eutectics, and elements. Ash causes various problems, especially corrosion, erosion, stickers, etc. [4] If the melting temperature of ash during combustion is tash < tflame, then the grate of the hearth can become clogged. Ash layers on the walls of the furnace diffuse into the lining, which then peels off in thin layers. The combustion chamber of the boiler must therefore be structurally adjusted such that the flame temperature drops below the ash melting temperature, i.e., the temperature on the grate should be lower than the melting temperature of the biomass ash [5].
The major problem from a chemical point of view is corrosion, which comes from the interaction between a refractory and a corrosive medium: gas, molten metals, molten glass, molten salts, or slag. It results in a loss of mass and thickness and in the degradation of the material properties [6]. The corrosion of refractory materials is a combination of external and internal physical and chemical influences.
The process is basically a chemical reaction between the refractory material and the slag or metal. Reactants are transported to the interface of the refractory material, and, in turn, the product reacts and is transported to the liquid phase. The dissolution of refractory materials in the melt is controlled by diffusion. Three types of corrosion have been defined: surface, dimple, and undersurface corrosion [7].
Alkaline corrosion, or "alkaline oxidative bursting", is extremely common, effective, and particularly harmful to alumina-silicon (Al-Si) lining systems, and it is usually observed in the temperature range of aggregates of 800-1000 °C. During biomass combustion, damage to the refractory lining is observed (Figure 1) as the peeling of surface layers, cracking, the bending of individual parts of the lining, the bulging of entire walls, and eventually their collapse [8,9].  [8,9]: (1) the damage to refractory materials after 1 year of the combustion of wood chips; (2) the furnace vault after 1.5 years of the combustion of chipboard; (3) the corroded part of refractory samples after 2 years of combustion of plant biomass, and (4-9) the presentation of the defects of the refractory lining after the combustion of biomass for 10 years of operation.
The main difference between coal ash and biomass ash is that coal ash contains higher amounts of SiO2 and Al2O3, but it contains lower amounts of K2O and Na2O. The eutectic of Al-Si forming fly ash lies above 1200 °C, while the eutectic of plant fly ash is much lower. Eutectic temperatures for mixtures of alkali metals together with silica or phosphorus have a low melting point: Na2O.2SiO2 (874 °C), K2O.4SiO2 (770 °C), and 2CaO.3P2O5 (774 °C) [10].  [8,9]: (1) the damage to refractory materials after 1 year of the combustion of wood chips; (2) the furnace vault after 1.5 years of the combustion of chipboard; (3) the corroded part of refractory samples after 2 years of combustion of plant biomass, and (4-9) the presentation of the defects of the refractory lining after the combustion of biomass for 10 years of operation.
The main difference between coal ash and biomass ash is that coal ash contains higher amounts of SiO 2 and Al 2 O 3 , but it contains lower amounts of K 2 O and Na 2 O. The eutectic of Al-Si forming fly ash lies above 1200 • C, while the eutectic of plant fly ash is much lower. Eutectic temperatures for mixtures of alkali metals together with silica or phosphorus have a low melting point: Na 2 O.2SiO 2 (874 • C), K 2 O.4SiO 2 (770 • C), and 2CaO.3P 2 O 5 (774 • C) [10].
Aluminosilicate refractories are based on the SiO 2 -Al 2 O 3 system. The equilibrium diagram of this system is given in Figure 2, marking various refractories. The main phase in the Al-Si binary diagram is mullite (3Al 2 O 3 .2SiO 2 ) [11], which increases the resistance of the refractory material against the corrosive effects of ash [11].
Materials 2022, 15, x FOR PEER REVIEW 3 of Aluminosilicate refractories are based on the SiO2-Al2O3 system. The equilibrium agram of this system is given in Figure 2, marking various refractories. The main phase the Al-Si binary diagram is mullite (3Al2O3.2SiO2) [11], which increases the resistance the refractory material against the corrosive effects of ash. [11]. In the AL-SI system, new phases are often formed as a result of different chemi reactions, gradually degrading the system. The newly formed products have a larger v ume than the original material, with expansion being reported between 7 and 30%. T creates compounds in the lining or on its surface that have chemical compositions a physical parameters different from the lining itself [12]. Ternary diagrams of the typ Na2O-Al2O3-SiO2 and K2O-Al2O3-SiO2 also describe the formation of individual phases the given system (see Figure 3). . Ternary diagram of Na2O-SiO2-Al2O3 [13] and K2O-Al2O3-SiO2 [14], marking the individ phases formed during alkaline corrosion.
The corrosion mechanism in the Na-Al-Si system includes the formation of alb (NaAlSi3O8), nosean (Na8Al6Si6O28S) [15,16], and natrosilite (Na2Si2O5) by Equation ( which further reacts with mullite (Al6Si2O13) to form albite (NaAlSi3O8) and aluminu oxide according to Equation (2). Nepheline (NaAlSiO4) can also be formed according Equation (3). Nosean is rarely reported in the literature as a corrosion product. Howev In the AL-SI system, new phases are often formed as a result of different chemical reactions, gradually degrading the system. The newly formed products have a larger volume than the original material, with expansion being reported between 7 and 30%. This creates compounds in the lining or on its surface that have chemical compositions and physical parameters different from the lining itself [12]. Ternary diagrams of the types Na 2 O-Al 2 O 3 -SiO 2 and K 2 O-Al 2 O 3 -SiO 2 also describe the formation of individual phases in the given system (see Figure 3). Aluminosilicate refractories are based on the SiO2-Al2O3 system. The equilibrium diagram of this system is given in Figure 2, marking various refractories. The main phase in the Al-Si binary diagram is mullite (3Al2O3.2SiO2) [11], which increases the resistance of the refractory material against the corrosive effects of ash. [11]. In the AL-SI system, new phases are often formed as a result of different chemical reactions, gradually degrading the system. The newly formed products have a larger volume than the original material, with expansion being reported between 7 and 30%. This creates compounds in the lining or on its surface that have chemical compositions and physical parameters different from the lining itself [12]. Ternary diagrams of the types Na2O-Al2O3-SiO2 and K2O-Al2O3-SiO2 also describe the formation of individual phases in the given system (see Figure 3). The corrosion mechanism in the Na-Al-Si system includes the formation of albite (NaAlSi3O8), nosean (Na8Al6Si6O28S) [15,16], and natrosilite (Na2Si2O5) by Equation (1), which further reacts with mullite (Al6Si2O13) to form albite (NaAlSi3O8) and aluminum oxide according to Equation (2). Nepheline (NaAlSiO4) can also be formed according to Equation (3). Nosean is rarely reported in the literature as a corrosion product. However, The corrosion mechanism in the Na-Al-Si system includes the formation of albite (NaAlSi 3 O 8 ), nosean (Na 8 Al 6 Si 6 O 28 S) [15,16], and natrosilite (Na 2 Si 2 O 5 ) by Equation (1), which further reacts with mullite (Al 6 Si 2 O 13 ) to form albite (NaAlSi 3 O 8 ) and aluminum oxide according to Equation (2). Nepheline (NaAlSiO 4 ) can also be formed according to Equation (3). Nosean is rarely reported in the literature as a corrosion product. However, In the case of high-alumina refractories (>45% Al 2 O 3 ) containing mullite (A 3 S 2 ) and cristobalite (SiO 2 ), reaction with NaO 2 above 1000 • C forms nepheline (NaS 2 ) and α-Al 2 O 3 according to Equation (5). As can be deduced from the ternary diagram K 2 O-Al 2 O 3 -SiO 2 , at a lower content of Al 2 O 3 < 30%, orthoclase KAS 6 is formed, and at a content of Al 2 O 3 > 30%, new phases of leucite (KAS 4 ) are formed according to Equation (6): Since the composition of biomass ash encourages the formation of eutectic melts, it is advisable to use high-alumina refractory materials with an Al 2 O 3 content > 80% or to add silicon carbide for these linings. The compound, aluminosilicate-based materials mainly include products containing oxide-less constituents-graphite and silicon carbide Materials in Al 2 O 3 -SiO 2 -SiC systems combine the high thermal conductivity and chemical inertness of silicon carbide with the chemical and thermal stability of aluminosilicate and corundum. The products are therefore highly resistant to corrosion by liquid metals, as well as to sudden changes in temperature.
SiC oxidizes according to Equations (7) and (8) and creates an amorphous SiO 2 film on the surface [18,19]: SiC + 1.5O 2 → SiO 2 + CO SiC + 2O 2 → SiO 2 + CO 2 (8) To prevent graphite oxidation, firing is carried out without any contact between the fired product and oxygen. The firing temperature is chosen to create a ceramic bond in the products. At present, the process of quick firing is used, ensuring a reducing atmosphere in the kilns at higher temperatures and while cooling the products. K 2 O and Na 2 O, in the form of alkaline vapors, are capable of diffusing into the refractory matrix, and then they react with Al 2 O 3 and SiO 2 components to form K-aluminosilicate and Na-aluminosilicate phases [20]. In the AL-SI binary system, potassium pairs react according to Equations (9)- (11). The most harmful is the presence of free SiO 2 and Na 2 O, which increase the reaction rate at high temperatures and support the formation of reactive glassy phases, according to Equation (12) [20]: The so-called slagging/fouling index can be used to estimate the probability of slag formation in combustion units during biomass combustion. Slagging/fouling means the formation of layers (sticky, melted, or soft) of ash particles on heat exchange surfaces. A summary of slagging and fouling indices and their calculation are presented in Table 1.
The SiO 2 index is often the predominant element in biomass samples and causes the formation of melt, or "stickers", therefore giving it the characteristic of being slag-forming.
The chlorine index Cl acts as an accelerator of the reaction between K and SiO 2 , which leads to the formation of fused glass deposits and the formation of slag at boiler operating temperatures of 800-900 • C [23].
Ash-deposition potential may be evaluated in terms of base-to-acid (B/A). The basicity index B/A (base/acid ratio) is based on the general rule that basic oxide compounds lower the melting point, and acidic compounds raise it. The B/A ratio is an indication of the fusion and slagging potential of ash. I/C (iron/calcium ratio) stands for Fe 2 O 3 /CaO, e.g., ash with a ratio of Fe 2 O 3 /CaO = 0.3/3.0 containing eutectics that increase slag formation.
The Fouling index Fu (fouling index) is the B/A ratio, also taking into account the alkali content (Na 2 O + K 2 O). Fouling refers to the dry deposition of ash particles or the condensation of volatile inorganic components on heat transfer surfaces. The normal percentage of alkali in biomass ash is between 25 and 35%, and it forms a eutectic in combination with silica.
Ash has a high viscosity (Sr) value, slag viscosity index Sr [24], so it will have a low tendency to slag. The TA (total alkali) index assesses the fuel's ability to form ash layers. Values of individual ash samples, defined based on the above-mentioned indices, are summarized in Section 3.2.
The chemical composition of ash is a good indicator of the problematic nature of biomass. For biomass fuels, massive slagging of heat exchange surfaces of boilers occurs during combustion. Ash composition and atmosphere in a combustion chamber influence the ash-melting temperature [10]. Indicators tell us of the characteristics of ash in terms of their influence on the formation of the glassy phase, and thus their tendency to slag and clog linings, heat exchange surfaces, and gas flow routes. These indices are based on chemical composition of biomass and its combustion. The equations are mainly based on fuel evaluation. However, since there is no specific index for biomass, it is possible to apply these indices to this type of fuel as well.

Ashes from Wood Biomass Combustion
Seven different types of ash from different types of wood biomass were used for the experimental portion of our study. All of these were obtained from the Czech Republic, mainly from the Moravian-Silesian Region, but one was from the Central Bohemian Region. Ashes utilized during the experimental portion were used in the original form for the crucible test. the granulometry was not adjusted. More information about the ash samples is presented in Table 2. the crucible test. the granulometry was not adjusted. More information about the ash samples is presented in Table 2.

Refractory Materials
Tested refractory materials were manufactured by one of the largest producers and suppliers of refractory products and raw materials in the Czech Republic. Two types of shaped refractory materials, belonging to the silica-aluminum group, were selected for the corrosion experiment.
The first type was quality labeled as STV. It is a shaped refractory material classified as standard fire clay. The second type was quality labeled as ARS60N and is classified high alumina. The parameters of the mentioned tested materials with their properties are shown in Table 3.

Corrosion Crucible Test and Evaluation Method
The crucible test gives only approximate results. The refractory cube was filled with corrodent and heated to the testing temperature for a specified period. The testing conditions (temperature and corrodent composition) may reflect the expected service conditions, but in some situations, a more aggressive corrodent and/or high temperature may be used to speed up the attack to determine the resistance of the refractory to the corrosive liquid in a relatively short time. The crucible test is described step by step in Figure 4. The refractory cuboid sample with a cylindrical hole in the central portion was filled with cor-

Refractory Materials
Tested refractory materials were manufactured by one of the largest producers and suppliers of refractory products and raw materials in the Czech Republic. Two types of shaped refractory materials, belonging to the silica-aluminum group, were selected for the corrosion experiment.
The first type was quality labeled as STV. It is a shaped refractory material classified as standard fire clay. The second type was quality labeled as ARS60N and is classified high alumina. The parameters of the mentioned tested materials with their properties are shown in Table 3.

Corrosion Crucible Test and Evaluation Method
The crucible test gives only approximate results. The refractory cube was filled with corrodent and heated to the testing temperature for a specified period. The testing conditions (temperature and corrodent composition) may reflect the expected service conditions, but in some situations, a more aggressive corrodent and/or high temperature may be used to speed up the attack to determine the resistance of the refractory to the corrosive liquid in a relatively short time. The crucible test is described step by step in Figure 4. The refractory cuboid sample with a cylindrical hole in the central portion was filled with corrosive, medium/powdered ash with a heating temperature of 1200 • C for 2 h. After cooling, the tested sample was cut through along the vertical axis, and the corroded portion was measured. rosive, medium/powdered ash with a heating temperature of 1200 °C for 2 h. After cooling, the tested sample was cut through along the vertical axis, and the corroded portion was measured. After the corrosion test, samples were visually checked for compactness, potential cracks, and holes in the sample and walls. The ČSN P CEN/TS 15418 method [25] and the internal regulation method of P-D Refractories CZ a.s. [26] were used for test evaluation.
The classification used for reporting the condition of the crucible with defined parameters [25] U: unaffected/no visible attack; LA: lightly attacked/minor attack; A: attacked/clearly attacked and C: corroded/completely corroded. In addition to the abovementioned evaluation regulations, another internal regulation method of P-D Refractories CZ was also used [26]. Table 4 shows the parameters of the classification after the corrosion test. Two evaluation methods may sometimes be requested by a customer or company testing laboratory, and the parameters can be used for comparison.

Characterization Methods
The chemical composition (XRF) of the ash was determined by energy-dispersive X-Ray fluorescence spectroscopy (ED-XRF) on the SPECTRO XEPOS (Spectro Analytical Instruments, Kleve, Germany). Powdered samples were shaped/pressed into tablets for XRD measurement.
The mineralogical composition (XRPD) of the samples was evaluated using X-Ray diffraction analysis on the X-Ray diffractometer MiniFlex 600 (Rigaku, Tokyo, Japan) equipped with a 0Co tube and a D/teX Ultra 250 detector. XRD patterns were recorded in a 5-90° 2θ range with a scanning rate of 5° min −1 .

Ash Characterization
Chemical analysis is a good indicator for determining the problematic nature of biomass. The chemical composition of all of the ash types is presented in Figure 5. Biomass After the corrosion test, samples were visually checked for compactness, potential cracks, and holes in the sample and walls. TheČSN P CEN/TS 15418 method [25] and the internal regulation method of P-D Refractories CZ a.s. [26] were used for test evaluation.
The classification used for reporting the condition of the crucible with defined parameters [25] U: unaffected/no visible attack; LA: lightly attacked/minor attack; A: attacked/clearly attacked and C: corroded/completely corroded. In addition to the abovementioned evaluation regulations, another internal regulation method of P-D Refractories CZ was also used [26]. Table 4 shows the parameters of the classification after the corrosion test. Two evaluation methods may sometimes be requested by a customer or company testing laboratory, and the parameters can be used for comparison.

Characterization Methods
The chemical composition (XRF) of the ash was determined by energy-dispersive X-ray fluorescence spectroscopy (ED-XRF) on the SPECTRO XEPOS (Spectro Analytical Instruments, Kleve, Germany). Powdered samples were shaped/pressed into tablets for XRD measurement.
The mineralogical composition (XRPD) of the samples was evaluated using X-ray diffraction analysis on the X-ray diffractometer MiniFlex 600 (Rigaku, Tokyo, Japan) equipped with a 0Co tube and a D/teX Ultra 250 detector. XRD patterns were recorded in a 5-90 • 2θ range with a scanning rate of 5 • min −1 .

Ash Characterization
Chemical analysis is a good indicator for determining the problematic nature of biomass. The chemical composition of all of the ash types is presented in Figure 5. Biomass ash almost always contains carbonates, especially calcite, and very often portlandite, as well as a proportion of organic carbon.  In terms of chemical composition, ash resembles low-melting glass. The vari of chemical composition complicates accurate representation in a ternary diagra approximate composition based on the largest content of wt.% of oxides is shown diagram. Four ash types, labeled P020, P033, P059 and P060, are marked in the CaO-Al2O ternary diagram, and two types, labeled P031 and P019 are marked in the K2O-SiO system, as presented in Figure 6. In terms of chemical composition, ash resembles low-melting glass. The variability of chemical composition complicates accurate representation in a ternary diagram. An approximate composition based on the largest content of wt.% of oxides is shown in the diagram. Four ash types, labeled P 020 , P 033 , P 059 and P 060 , are marked in the CaO-Al 2 O 3 -SiO 2 ternary diagram, and two types, labeled P 031 and P 019 are marked in the K 2 O-SiO 2 -CaO system, as presented in Figure 6. The next method of ash characterization was X-ray powder diffraction phase analysis (XRPD). The samples were compared to the reference diffractogram database published by ICDD (PDF-2) in the range of 5-90° 2theta. The results for the analyzed samples are Figure 6. Approximate position of ash types P 020 , P 033 , P 059 , P 060 , in ternary diagram CaO-Al 2 O 3 -SiO 2 [27] and P 031 and P 019 in ternary diagram K 2 O-Al 2 O 3 -SiO 2 [28]. The next method of ash characterization was X-ray powder diffraction phase analysis (XRPD). The samples were compared to the reference diffractogram database published by ICDD (PDF-2) in the range of 5-90 • 2theta. The results for the analyzed samples are presented in Table 5, where there is an overview of the phases in the samples. Table 5. Phase composition of analyzed biomass ash samples.

Phase Composition
Labeled of Sample P 019 P 020 P 031 P 032 P 033 P 059 P 060 As confirmed by the analysis, the most frequently recurring phases are quartz, anorthite, calcium silicate, hematite, anhydrite, and microcline. In ash samples P 059 and P 060 , there were seven phases identified as portlandite; microcline, leucite, and portlandite occur in both. Samples P 019 and P 032 , were especially rich in the glass phase.

Calculation of the Slagging and Fouling Indices
To predict slagging/fouling in a combustion furnace, it is possible to use indices for the SiO 2 , basic/acid ratio, silica/alumina ratio, fouling, iron/calcium ratio, and total alkalis, as summarized in Table 6. A special index only for biomass does not exist, but many authors have calculated these indices with regard to the probability of slag forming in combustion units. In the case of SiO 2 content in P 020 , P 033 , P 059 and P 060, they have a high inclination towards slagging. The high levels of silica in wood biomass ashes may have been caused by contamination with different elements (clay, sand, etc.); also, each part of the wood plant may contain different amounts of oxides. According to chloride content, extremely high fouling inclinations were observed in samples P 032 = 1.74 and P 020 = 0.47, while a low fouling inclination with a value > 0.2 was calculated for P 033 = 0.1.

Index B/A confirmed that all wood biomass ash samples included in this study
showed high to extremely high values of slagging (>1), as shown in Table 6 and Figure 7. The highest value B/A = 8.1 was for ash P 059 , which will have a tendency toward slag formation. The low B/A index for P 033 ash can be attributed to its high SiO 2 content, which implies an increased presence of acidic compounds. No ash samples had a B/A value lower than 0.5. Samples which had the highest B/A index and a high Na 2 O content showed an inclination towards fouling.
In the case of SiO2 content in P020, P033, P059 and P060, they have a high inclinat wards slagging. The high levels of silica in wood biomass ashes may have been cau contamination with different elements (clay, sand, etc.); also, each part of the wood may contain different amounts of oxides. According to chloride content, extremel fouling inclinations were observed in samples P032 = 1.74 and P020 = 0.47, while a low ing inclination with a value > 0.2 was calculated for P033 = 0.1.
Index B/A confirmed that all wood biomass ash samples included in this showed high to extremely high values of slagging (>1), as shown in Table 6 and Fi The highest value B/A = 8.1 was for ash P059, which will have a tendency toward sl mation. The low B/A index for P033 ash can be attributed to its high SiO2 content, implies an increased presence of acidic compounds. No ash samples had a B/A lower than 0.5. Samples which had the highest B/A index and a high Na2O content sh an inclination towards fouling. The I/C value for the tested samples was >0.3 and represents a small risk for th radation of refractory materials. S/A index values were higher for all tested ash sa the slagging tendency based on the silica/alumina ratio (S/A) was high for all the as ples because no value was lower than 3 ( Figure 7). The I/C value for the tested samples was >0.3 and represents a small risk for the degradation of refractory materials. S/A index values were higher for all tested ash samples. the slagging tendency based on the silica/alumina ratio (S/A) was high for all the ash samples because no value was lower than 3 ( Figure 7).
As a final result, higher Fu values correspond to higher fouling tendencies. The extreme value of the fouling index (Fu) was observed in sample P 019 , with Fu = 240.1, and P 031 Fu = 72.1, as presented in Figure 8. Fouling index values over 40 indicate a high tendency to fouling. The fouling tendency based on total alkalis was high for the types of ash mentioned below. As a final result, higher Fu values correspond to higher fouling tendencies. The extreme value of the fouling index (Fu) was observed in sample P019, with Fu = 240.1, and P031 Fu = 72.1, as presented in Figure 8. Fouling index values over 40 indicate a high tendency to fouling. The fouling tendency based on total alkalis was high for the types of ash mentioned below.
The value of total alkali content (TA) was never less than 0.4. (Figure 8). The lowest TA index was observed in P060 TA = 5.3, and it was 6 times higher in P019 = 29.7. The maximum value was 0.4, and there was a higher tendency towards the fouling of refractory materials during biomass combustion. Slagging probability was high for ash samples based on their Sr ratios, which were lower than 65. In this study, the calculated index for Sr corresponded to five ash types. Samples P020 and P033 had the middle value between 65 and 72. Based on calculations, the ash will tend to form slag and deposits.

Corrosion of Refractory Materials
Two types of refractory materials were selected for this portion of the corrosion experiment, STV fire clay (containing 53 wt.% SiO2, 42 wt.% Al2O3), shown in Figure 9, and The value of total alkali content (TA) was never less than 0.4. (Figure 8). The lowest TA index was observed in P 060 TA = 5.3, and it was 6 times higher in P 019 = 29.7. The maximum value was 0.4, and there was a higher tendency towards the fouling of refractory materials during biomass combustion. Slagging probability was high for ash samples based on their Sr ratios, which were lower than 65. In this study, the calculated index for Sr corresponded to five ash types. Samples P 020 and P 033 had the middle value between 65 and 72. Based on calculations, the ash will tend to form slag and deposits.

Corrosion of Refractory Materials
Two types of refractory materials were selected for this portion of the corrosion experiment, STV fire clay (containing 53 wt.% SiO 2 , 42 wt.% Al 2 O 3 ), shown in Figure 9, and ARS60N high alumina (containing 28 wt.% SiO 2 , 46 wt.% Al 2 O 3 , and 13 wt.% SiC), shown in Figure 10. Refractory materials underwent a crucible test under the following conditions: 2.9 g of ash as corrosion agent/2 h on maximum temperature at 1200 • C. All ash types melted at the suggested testing temperature.  Resistance to eutectic melts was determined by the refractory quality of STV at 1200 °C. As shown in Figure 9, infiltration was relatively low in samples P019 and P060 with refractory materials, whereas the attack of ashes P032 a P031 achieved a high intensity with STV refractory materials.
The evaluation of the corrosion attack by two methods is presented in Table 7. The Figure 9. Corrosion of STV refractory materials using different types of wood ash and defined parameters of crucible test: 2.9 g ash/2 h/1200 • C.
Materials 2022, 15, x FOR PEER REVIEW 12 of Figure 9. Corrosion of STV refractory materials using different types of wood ash and defined p rameters of crucible test: 2.9 g ash/2 h/1200 °C. Figure 10. Corrosion of ARS60N quality refractory material with different types of wood ash wi defined parameters of crucible test; 2.9 g ash/2 h/1200 °C.
Resistance to eutectic melts was determined by the refractory quality of STV at 120 °C. As shown in Figure 9, infiltration was relatively low in samples P019 and P060 with r fractory materials, whereas the attack of ashes P032 a P031 achieved a high intensity wi STV refractory materials. Resistance to eutectic melts was determined by the refractory quality of STV at 1200 • C. As shown in Figure 9, infiltration was relatively low in samples P 019 and P 060 with refractory materials, whereas the attack of ashes P 032 a P 031 achieved a high intensity with STV refractory materials.
The evaluation of the corrosion attack by two methods is presented in Table 7. The corroding interaction was most intensive between P 032 and STV refractory surface. P 032 had the highest proportion, 1.77 wt.% Cl, a high content of alkalis (Na 2 O + K 2 O), namely 16 wt.%, a high CaO content of 26 wt.%, and 19 wt.% SiO 2 . B/A = 2.3 ex was extreme, and the ash can be defined as basic. The slag viscosity index was Sr = 35.2 h and index Fu = 46.6 h was high, too. The ash showed a tendency towards slagging, and CaO and K 2 O reduced the viscosity of the slag. Basic oxides lowered the ash melting point. Cl acts as an accelerator of reactions between K 2 O and SiO 2 , which leads to the formation of fused glass deposits and the formation of slag at the boiler operating temperature of 800-900 • C [23]. According to Table 7 and normative regulation [25], the reaction between the STV quality refractory and ash P 032 was evaluated as attacked/clearly attacked, and according to [26], the distinctive attack is >7 mm corrosion and/or infiltration/slight cracks. P 031 attack had an effect in contact with refractory materials that was very similar to that of P 032 . All index values were categorized as high, but the amount of Cl was only 0.16 wt.%, unlike that of P 031 . The crucible test of STV corrosion by P 031 was evaluated according to [25] as attacked/clearly attacked and [26] slight attack/<6 mm corrosion and/or infiltration/no cracks.
From the visual point of view (see Figure 9), the sample tested by P 033 formed melt, which represents a low-melt eutectic compound, because K 2 SiO 3 and Na 2 SiO 3 present low melting temperatures. P 033 contained 55.1 wt.% SiO 2 , alkali (Na 2 O + K 2 O) 7 wt.%, and only 0.1 wt.% Cl. Index B/A was only 0.5, but a high index Sr = 71.7, which had a negative effect to refractory materials. A similar effect was observed on samples attacked by P 020 . STV refractory materials, after the application of P 019 and P 060 , may appear to be inert to ash, but a little infiltration was noticeable upon closer examination. Nevertheless, the evaluation according to the methodology was uniform.
An increase in temperature can be a significant indicator of changes in refractory materials. A problem may occur if temperature fluctuations occur-in this case, the combustion temperature could be 200 • C higher than the normal operating temperature [29], and the corrosion of refractory materials can be very intensive.
The formation of slag in the furnace depends on the chemical and mineralogical composition of the ash and on the conditions in the furnace (temperature, reduction, or oxidation zone, etc.). When fuel is burnt, combustion produces ash. In this case, residue will always form on the walls of the hearth and the heat exchange surfaces [18]. Slag deposits can be further divided into so-called "sintered deposits". (They are formed by dust particles trapped in a liquid or plastic state and stick to the wall.) The second type is hard deposits having a layered structure. The viscosity of the slag depends on the composition of the ash [13,14,30].
The second tested sample was high-alumina refractory ARS60N with a content of 46 wt.% Al 2 O 3 , 28 wt.% SiO 2 , and 13 wt.% SiC (Figure 10). From the visual point of view, the refractory material was resistant to ash attack at 1200 • C. None of the samples showed total corrosion leading to disintegration. P 032 showed negative effects on the refractory material ARS60N, just as it did with the STV quality. P 032 caused slight changes on the surface structure with very fine cracks. The infiltration of ash was 6.4 mm, and in the structure of the tested refractory materials, it formed spaces without grains, which changed to compact slag in the middle of the sample. However, the integrity of the samples was preserved. A detailed photo of the changing structure of P 032 is presented under the main pictures. As already mentioned, this ash has a tendency towards slagging. A very similar effect was observed in the middle of the refractory material, which formed after the attack of ash P 033 . According to Table 8 and the literature [25], the corrosion effect between refractory quality ARSN60 and P 032 described as attacked/clearly attacked, and according to Plibrico [26], it presented as a distinctive attack />7 mm corrosion and/or infiltration/slight cracks. The viscosity index of P 032 is Sr = 35.2 h , and that of P 031 Sr = 25.3 h . According to [24], ash with a value (<65) viscosity Sr will have a high tendency towards slagging. On the other hand, P 020 , with index Sr = 67 m and P 033 Sr = 71.7 m , will display the opposite trend. The effect of P 020 is visually comparable for both STV and ARS60N refractory materials. The interaction between P 020 and ARS60N samples formed a very slight layer of glassy character with an infiltration of about 6-8 mm. In terms of chemical composition, P 020 contained 46 wt.% SiO 2 , 16 wt.% CaO, and 11 wt.% alkali oxides (Na 2 O + K 2 O). This ash is acidic, as the ratio is B/A = 0.7 m . TA = 11.5 m and Fu = 7.7 m [24], so this ash should have a low tendency towards slagging. Other tested ash types showed invisible or minimum attack with refractory materials. ARS60N presented a higher level of corrosion resistance against the tested ash types than to STV. This portion of the experiment showed that the suggested temperature of 1200 • C during the corrosion crucible test should not be exceeded for tested STV and ARS60N refractory materials with the above-mentioned tested wood ash types.

Conclusions
STV quality refractory materials contain 53 wt.% SiO 2 and 42 wt.% Al 2 O 3 ), while in ARS60N, the quantity of oxides is 46 wt.% Al 2 O 3 , 28 wt.% SiO 2 , and 13 wt.% SiC). These materials belong to the alumina-silica system and were tested for their corrosion resistance. The results obtained within this research can be summarized as follow: • Seven types of wood ash were used as corrosive media in their original forms. All types of ash were obtained by biomass combustion. The testing method was a crucible test at 1200 • C.

•
The intensity of the corrosive effect was determined by two regulations. Neither a corrosion test nor visual evaluation revealed a greater penetration or destruction of the samples and that was not recorded, but only light/middle corrosion. From all tested ash samples, the one labeled P 032 achieved the highest corrosive effect on both tested alumina-silica refractory materials.
• A less-intensive corrosion attack was observed in P 031 with STV and ARS60N with P 020 . Visible signs of corrosion included melts, fine cracks, infiltration, and in some cases, structural changes. The above-mentioned results suggest that the poorer the material is in silicon dioxide, the better it resists corrosion.

•
The chemical stability of the phases increases as follows: amorphous silica (SiO 2 ) < cristobalite (SiO 2 ) < quartz (SiO 2 ) < andalusite (Al 2 SiO 5 ) < mullite (Al 6 Si 2 O 13 ) < corundum (Al 2 O 3 ). • Upon a visual check, fire clay materials with SiO 2 > Al 2 O 3 content presented a higher infiltration of the corrosion medium into the sample with an apparent porosity similar to that of sillimanite refractory materials.

•
This may lead us to the conclusion that a higher level of Al 2 O 3 content can increase corrosion resistance, as is often reported. An effective solution is also a partial replacement of the oxide with silicon carbide SiC, which has shown a favorable effect on the resistance of refractory materials. A high refractory-material-testing temperature can be a good indicator of a possible corrosive effect on the tested refractory material if there is undesirable temperature fluctuation during combustion.

•
The nature and behavior of biomass ash can be defined by calculating slagging/fouling indices (B/A, Sr, Fu, TA, etc.), which are applied to the calculations of conventional fuels.

•
Based on the chemical composition and calculated indices for wood biomass ash samples, it was confirmed that many of the values are above the limits and should tend to form slag and fouling. However, not all samples confirmed this prediction. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.