16.2 Coal‐Fired Boilers

The principle of a thermal power plant is to produce electricity by burning a fuel generally consisting of coal, biomass, or waste. Electricity is produced by the process of heating water in a boiler to yield steam. The steam under pressure, heated to high temperature in the superheater part, flows into a turbine, which drives a generator to produce electricity (Figure 16.1).

In today’s conventional coal‐fired power plants, the steam is brought to above 22 MPa and above 540 °C in a supercritical state.

Since the energy crisis in the 1970s, there has been a constant need for increasing the efficiency of thermal power plants. In other words, the development aimed at increasing the amount of electricity produced for the same fuel consumption. The general route that has been followed consists of increasing the steam parameters, temperature, and pressure in the plant. As an example, the total efficiency of a plant increases by nearly 6% when changing the steam parameters from 538 °C/18.5 MPa to 593 °C/30 MPa. It could even reach an 8% increase at 650 °C (Viswanathan and Bakker 2000). One of the most modern coal‐fired power plants commissioned in 2002 in Niederaussem in Germany demonstrates quite well the current state of the development with a fuel conversion efficiency of 43% and steam parameters up to 580 °C and 28 MPa (Heitmüller and Kather 1999). Furthermore, the Kyoto Protocol ratified in 1997 by many countries, including the European Union, plans a significant reduction of greenhouse effect gas emissions. These include first of all CO₂ but also NO_x (Kyoto Protocol 1997).

At the same time, predictions indicate that the global electricity demand will continue to increase at an average of 2% per year until 2020 (Parker 2002). To cover these new capacity demands, the global coal consumption is planned to increase by 60% between 2000 and 2020 (Figure 16.2). Coal is a cost‐effective fuel, which will have a significant role at least until the efficiency of renewable energy products is improved in the next 20 years.

The electricity production from waste or biomass is also planned to increase. Waste firing offers a solution for reducing the increasing amount of waste, especially municipal waste. Biomass, like wood or straw, is a renewable fuel that can be considered as a long‐term alternative to the decreasing amount of fossil sources.

Meeting both the Kyoto requirement and the increase of the electricity production leads to the need of improving the efficiency of the fuel to power conversion processes, which by extension means an increase of the steam parameters. For coal‐fired systems, temperatures up to 720 °C or even 760 °C with a pressure of 35 MPa are currently under consideration in Europe and in the United States, respectively. The aim is to achieve a thermal efficiency of 60% by the year of 2020 (Ruth 2003).

The superheater tube temperatures often exceed the steam temperature by as much as 30 °C (Viswanathan 2004). This application thus requires a material with high temperature creep resistance. During the last decades, high temperature steels or more recently nickel‐based alloys have demonstrated their potential with regard to high temperature creep strength. Nevertheless, metallic materials are well known for expanding when increasing the temperatures. The use of materials with low coefficient of thermal expansion is thus required in order to minimize the generated mechanical stresses. Moreover, the thermal conductivity of the superheater components has to be kept as high as possible in order to maximize the efficiency of the energy conversion. For building a superheater, several tubes must be welded together. Therefore, the weldability of the materials involved has also to be optimized.

Depending on the type of fuel, the combustion produces hot gases of various concentrations but is usually composed of O₂, CO_x, SO_x, HCl, H₂O, and NO_x. Furthermore, depending on the tube temperature, solid or liquid ashes and salts are deposited on the superheater tube surfaces. Table 16.1 gives the typical species found in various deposits. For coal‐fired plants, the deposits contain usually sodium or potassium sulfates, whereas for waste‐fired plants, the deposits are rather composed of alkali or heavy metal chlorides, due to the unceasing increase of the polyvinyl chloride (PVC) plastic content in refuses. Concerning biomass firing, the deposits are usually composed of potassium chlorides or potassium sulfates.

Molten sulfates and chlorides, as well as the gases mentioned earlier, are known for provoking corrosion damage at high temperature (Kofstad 1988). Moreover, on the inner side of the tube, the high temperature steam can lead to rapid oxidation. As a consequence, in addition to high creep strength, low thermal expansion, high thermal conductivity, and good weldability, the superheater tubes have to resist high temperature corrosion.

The mechanisms describing this form of corrosion are various and complex with strong dependence on the type of material and environment. They involve reactions such as oxidation, chlorination, carburization, sulfidation, and chemical/electrochemical in molten or semi‐molten salt deposits. The reader can find detailed explanations on these reactions in Chapters (oxidation), 8 (sulfidation), 9 (carburization), 11 (chlorination), and 13 (molten salts).

A large utility boiler consists of a large firebox, within which the combustion takes place. A number of burners are installed in the walls, through which the pulverized coal is blown. The walls of the firebox are manufactured with vertical tubes, typically between 70 and 100 mm of outside diameter, joined by webs of the order of 30 mm width. This is “membrane wall” construction. Preheated water enters the bottom of these walls, which are usually called the waterwalls, and boils at some location usually a little above the top of the burner array. The effective temperature in the combustion zone is in the range of 1200–1800 °C. As the combustion gases rise in the furnace, they cool; at the top of the furnace, the temperature is of the order of 1000–1200 °C, and the gas turns into a horizontal section. In this section, and sometimes at the top of the furnace itself, there are tubular platen heat exchangers hanging down into the combustion gas stream: these are commonly superheaters, and the principal corrosion problems are in the superheaters and the waterwalls (Meadowcroft 1987).

The majority of the utility boilers are subcritical recirculating boilers with a main steam pressure of approximately 16 MPa. The boiling point at this pressure is about 350 °C, and the outer surface of the waterwall tubes will be about 50 °C above this. The superheat temperature is 538 °C, and the maximum outer metal temperature in the superheater may be as much as 100 °C above this because of nonuniformities in the temperature distribution across the superheater bank.

The materials for the pressure parts are selected on the basis of their mechanical properties. For the lower temperature parts, the tensile strength is limiting; for the higher temperature parts, the creep strength is critical. The majority of the tubes, pipes, and drums are manufactured with plain carbon steel or low‐alloy ferritic steels; the highest ferritic alloy normally used is T (or P) 22, which is essentially 2Fe–1/4Cr–1Mo. The very last sections of the superheater and reheater exchangers may be manufactured with austenitic steel (Meadowcroft and Manning 1983).

There are two main high temperature corrosion problems that may be identified in utility boilers:

1. The fireside corrosion of the superheaters and reheaters.
   
2. The fireside corrosion of waterwalls.

Figure 16.3 indicates the typically affected regions in a fuel‐fired boiler.

There are several methods for burning coal. Firing with pulverized coal has been the most dominant method for utility boilers. Coal is burned as fine powders suspended in the furnace and almost all types of coal from anthracite to lignite can be burned by pulverized firing. Pulverized coal in particle sizes of 50 μm diameter or smaller can be completely combusted in a matter of one to two seconds.

Stultz and Kitto (1992) developed the cyclone furnace for firing coal grades that have a low fusion temperature and are not suitable for pulverized coal firing because of potentially forming a molten slag, thus developing a severe slagging problem in superheaters. In cyclone boilers, the cyclone barrels burn coal in such a way that most of the coal ash is captured to form a molten slag that coats the inside surface of the cyclone barrels. The combustion flue gas from the cyclone barrels then enters the main furnace to generate steam (Stultz and Kitto 1992).

There are four different types of coal. The youngest, or lowest‐rank coal, is lignite, which is then followed by an older, or higher‐rank subbituminous coal, then bituminous coal, and then anthracite coal. Heating values, moisture content, volatile matter content, ash content, and sulfur content can be different among these different types of coal. Bituminous coal is the most commonly used coal for utility boilers in the United States. Subbituminous coal in the United States generally contains very low sulfur, with many deposits containing less than 1%. Table 16.2 shows the properties of some US coal. There is a great variation in moisture, ash content, ash softening temperatures, and sulfur content from various grades of coals. The material factors associated with coal can directly or indirectly affect the boiler tube material performance at different locations in the boiler.

Sulfur is the most important impurity in coal for causing high temperature corrosion in the boiler. Sulfur is present in coal in the form of organic sulfur, pyritic sulfur (i.e. pyrite), and iron sulfate. High sulfur coals also cause SO_x emission problems and require expensive air pollution control equipment. As a result of the US Federal Clean Air Act emissions issues, the low sulfur Powder River Basin (PRB) coal (a subbituminous coal) has become extremely popular in the past years. The PRB coal, which is from mines in southern Montana through northern Wyoming, contains less than 0.6 kg of sulfur per million Btu, making it compliant with the Clean Air Act emissions limits without air pollution control equipment.

One of the important characteristics of ash is its fusion temperature. Table 16.2 lists the ash fusion temperatures of several coals under reducing conditions. These fusion temperatures can affect the nature of the ash deposits, whether in the form of “dust” or a tenacious slag. If ash reaches the heat‐absorbing surface at a temperature near its softening temperature, the resulting deposits are likely to be porous and can be easily removed by sootblowing. Also, if such a deposit is subjected to high gas temperature, the ash deposit can reach its melting point (due to the thermal insulating properties of the ash) and run down the furnace wall surface. This solidified slag is tightly bonded and is difficult to remove. This slag may require water lances or water cannons to create thermal shock for the removal of this slag deposit. The furnace walls that are subject to radiant heat are likely locations for developing this slagging problem.

The analyses of ash for several types of US coal are shown in Table 16.2. The compositions shown in Table 16.2 are presented as oxides. However, most ash constituents in coal are minerals. Typical minerals found in coal are shown in Table 16.3. When these minerals are exposed to oxidizing environments at appropriate high temperatures, oxides and/or complex salts are formed as stable phases. This is illustrated in Table 16.4.

If the ash constituents are present as oxides, the ratio of basic oxides (e.g. Fe₂O₃, CaO, MgO, Na₂O, and K₂O) to acidic oxides (e.g. SiO₂, Al₂O₃, and TiO₂) may determine the fusion (fusibility) temperature of the reaction product. It was reported that the ash may exhibit low fusibility temperature with higher slagging potential when its base–acid ratio is in a range of 0.4–0.7. Many other parameters, such as SiO₂/Al₂O₃ ratio, Fe₂O₃/CaO ratio, Fe₂O₃/(CaO + MgO) ratio, (Na₂O + K₂O), and so forth, are also used for predicting the fusibility temperature of the coal ash. It has been suggested that SiO₂ is more likely than Al₂O₃ to form lower‐melting species. The fusibility temperature of coal ash will be lowered when the Fe₂O₃/CaO ratios are in a range of 0.2–10. Alkalis are important in affecting the fusibility of coal ash and the furnace slagging potential. Many sodium compounds melt at temperatures below 900 °C, playing an aggressive role in fireside corrosion.

Chlorine content is also an important indication for fouling potential. When chlorine in coal is greater than 0.3%, fouling potential becomes high. Alkali metals and chlorine can also play a significant role in high temperature corrosion in boilers.

We have been discussing complex corrosion phenomena and other issues in typical boilers being evident that to help the reader’s understanding, the important high temperature corrosion aspects in coal‐fired power plants should be rapidly summarized at this stage.

In coal‐fired power plants, besides corrosion damage due to molten salts, corrosion can occur by oxidizing flue gas and other reactions as already reported. Depending on the way the combustion is carried out, the gases emanating from the combustion can be either reducing or oxidizing. The typical gaseous species encountered in oxidizing conditions are O₂, NO_x, HCl, SO_x, H₂O, and CO_x.

With respect to the high sulfur content in some types of coals, particular attention has to be given to sulfur‐containing gases. Indeed, it is well known that in most cases, sulfidation kinetics are much faster than those of oxidation and that sulfides are usually non‐protective other than oxides.

The second corrosive element is chlorine, which is usually found as HCl in combustion products. Chlorine is particularly detrimental because of the formation of low temperature eutectic melts and high volatile metallic chlorides that can form at high temperature and lead to rapid metal loss. Furthermore, Rahmel and Tobolski (1965) proposed a mechanism explaining the acceleration of the oxidation of pure iron in the presence of water vapor or CO₂. This mechanism is based on the formation of cracks within the oxide scale. After some time, these cracks are filled with H₂O/H₂ or CO₂/CO, respectively. As shown in Figure 16.4 for H₂O‐affected oxidation, a repeated reduction of H₂O at the metal surface and oxidation of H₂ at the gas–oxide interface is put in place within the crack. As a consequence, the presence of H₂O or CO₂ enables further oxidation of the metal without substantial inhibition. This mechanism is still accepted nowadays. Many studies focused on the corrosion mechanisms of one or two types of corrosive species, for example, SO_x or SO_x combined with O₂ as reviewed by Kofstad (1988). Others focused on the effect of chlorine, in particular HCl combined with O₂ (Zahs et al. 1998). On the other hand, in‐plant studies are generally governed by molten salt corrosion and are hardly comparable due to differences in plant operation and the difficulty for precisely analyzing the combustion gases. In fact, there seems to be only a limited number of studies addressing the corrosion resistance of 9–12% Cr steels in simulated coal firing flue gas, containing three or more oxidizing species, as will be presented later.

Fireside corrosion of the steam‐containing superheater and reheater tubes due to the deposition of fuel impurities from the flue gas at its highest temperature is a significant problem, and considerable research has been carried out concerning this corrosion process. Since it has been shown that accelerated wastage of superheater and reheater tubes occurred under an ash deposit, much research work has involved exposing metals and alloys to molten salt baths, building up a thin sulfate film of around 1–2 mm thickness, to replicate typical deposits found on the tubes (Figure 16.5).

The black inner layer is iron oxide, essentially Fe₃O₄, and is typically 2.20 μm thick. The white layer contains majorly sulfates of Na, K, and Al, with some Fe. This is the molten sulfate layer and is usually around 1 mm thick. The “red slag” layer is high in Fe, low in alkali, and with a sulfate content intermediate between that of the white deposit and the ash. The adhering fly ash is rich in Fe₂O₃ and remains porous, even if sintered. Dissolution of both white and red layers in water produces acidic solutions.

Figure 16.6 shows the temperature distribution through the system for a typical set of conditions. This diagram is however oversimplified. The overall heat flux to a superheater tube is of the order of 0.2 MW m⁻². It is however evident from Figure 16.5 that the thickness of the insulating ash layer is greater at the front, leading, position of the tube than the rear, trailing, area. This results in a variation in the heat flux around the tube. Wastage takes place at the “2 o’clock” and “10 o’clock” positions relative to the flow of the flue gases, the “12 o’clock” position. The heavy buildup of ash on the front surface reduces the temperature flux within the deposit below that required for the salt layer to be molten. The thinner ash levels at the side of the tube allow for a higher heat flux, and a molten layer beneath, at the deposit–metal interface.

The origin and role of the molten salt layer, and its associated fly ash, may be summarized as follows. The cations arise largely from the clay component of the coal, and the sulfate ions from the oxidation of the sulfur in the fuel to SO₂. The air entering the combustion chamber contains about 10% more oxygen than necessary for stoichiometric combustion, and Fe₂O₃ in the fly ash acts as a catalyst to convert the SO₂ to SO₃. Fuels containing vanadium, such as fuel oil, often used in the start‐up process in boilers, provide an even more effective catalyst for SO₃ formation. SO₃ readily dissolves in the molten sulfate, forming essentially pyrosulfate ions [S₂O₇]²⁻: SO₃ and O₂ are the main oxidants and SO₂ is insoluble in sulfate melts.

The SO₃ present in the flue gas has the major effect on the corrosion of metal tubes, since this determines the oxidizing potential of the sulfate deposit, in terms of the redox equilibria:

16.4

The corrosion process is thus two separate reactions. First, an anodic oxidation, by

16.5

to form metal ions in the corrodant (n may be 2 or 3 but here will be assumed as 2). Thus

16.6

Second, the cathodic reaction is transfer of electrons to the oxidizing agent

16.7

More specifically, for the reduction of and/or SO₃, we can write

16.8

16.9

These equations are oversimplified and the reaction mechanism is better understood by considering a steady state in which SO₂ and O₂ from the flue gas enter the porous fly ash. The catalytic action of Fe₂O₃ then produces equilibrium concentrations of SO₃. SO₃ dissolves in the melt and migrates to the oxide–melt interface. Then, it diffuses through the porous oxide to the oxide–metal interface.

16.10

This is regarded as the overall result of the electrode Reactions 16.6 and 16.7. Solid metal oxides and sulfates result from the subsequent oxidation of the metal together with metal ions in the sulfate melt. Dissolution of the metal oxides at the oxide–melt interface takes place, viz.

16.11

with some of the metal sulfide being converted to the oxide, introducing imperfections in the oxide scale as it forms.

The solubility and stability of the dissolved metal ions in the melt is dependent upon both SO₃ partial pressure and temperature, increasing with increase in SO₃ content, but decreasing with increasing temperature. A mutual opposition is thus seen for the effects of SO₃ partial pressure and temperature on traversing from the oxide–melt to the melt–fly ash interface such that a temperature is reached where the dissolved metal ion is precipitated as the oxide. A resulting concentration gradient of dissolved metal ions across the melt was thus envisioned and has been shown (Griffiths et al. 1982).

The corrosion process is therefore considered as the continuous dissolution of metal and metal oxide, reprocessing to form a non‐protective layer at the melt–fly ash interface. Considering all the factors involved, the rate‐determining step for the dissolution of the protective layer that the steel tubes generate, the overall rate of corrosion process, is that of the diffusion of metal ions away from the metal–metal oxide interface.

The transition metal ions from the steel tubes in the sulfate melt do not behave like the alkali metal ions Na⁺ and K⁺ – they form complexes with the sulfate ions:

16.12

The evidence for this comes from the electronic absorption spectra of such melt systems. The spectra obtained all correspond to the presence in solution of the metal ions surrounded octahedrally by six oxygen atoms. It is therefore concluded that each of the metal ions has three bidentate sulfate ions coordinating to it.

In 2002, Masuyama presented a review of the last 50 years of alloy development for power plant applications. Four generations of the evolution of ferritic steels since the 1950s are distinguished on the basis of their creep properties. The alloy development has indeed been motivated by the enhancement of the creep strength. The 10⁵ hour creep‐rupture strength at 600 °C is the parameter typically used for alloy classification (Figure 16.7). It reaches 140 MPa in the current state of the art. But among the commercialized 9% Cr steels, the P91 has been extensively used for heaters and steam pipes in plants operating at temperatures up to 593 °C.

From a metallurgical point of view, the steels presented in Figure 16.7 are either called ferritic or martensitic steels. In fact, both designations are correct depending on the heat treatment. For instance, P91 results from a specific heat treatment that consists of an austenitization step followed by rapid cooling for the martensitic transformation. A subsequent tempering is then required to transform part of the martensite into ferrite and allow carbides to precipitate homogeneously within the martensite.

The steel development shown in Figure 16.7 has mainly consisted of adding alloying elements with strengthening effects (Viswanathan and Bakker 2000). W, Cr, Mo, and Co are used for solid solution strengthening, and the latest tendency was to replace Mo by the more beneficial W. W, Cr, and Mo also contribute to precipitation strengthening by formation of carbides, whereas V and Nb used in junction with N contribute to precipitation strengthening by forming carbonitrides. Although carbon is required for the fine carbide precipitation, the carbon content remains limited, below 0.12 wt%, in order to obtain a good weldability. Nickel additions improve the toughness but are detrimental to creep strength. Therefore, some developments, such as for HCM12A, consisted of replacing part of the Ni by Cu. Finally, Cr additions have been rather motivated by the enhancement of the corrosion resistance, which will be discussed.

The oxidation resistance of 9–12% Cr steels, especially the P91, has been extensively studied in various environments including air in order to simulate fireside corrosion.

It was demonstrated that changes of the surface conditions of the 9% Cr steels influence the formation of a protective chromia layer (Grabke et al. 1997), in particular during the very early stage of oxidation. First of all, the early formation of chromia is strongly influenced by the Cr diffusivity toward the surface. This is enhanced when the dislocation density is increased by previous deformation of the subsurface zone, using sandblasting, for instance (Ostwald and Grabke 2004). Moreover, in the first 10–15 minutes of oxidation, local effects are of great importance. Fast Cr diffusion along grain boundaries lead to the nucleation and growth of chromia. However, in other areas, formation of less protective Fe‐rich oxides was reported for 9% Cr, whereas a Cr content higher than 11% is more favorable for the formation of Cr‐rich oxides without local differences (Grabke et al. 2004).

By considering exposure times of several thousand hours, the minimum Cr content required for the formation of a protective chromia scale was estimated to about 7.2 wt% (Vossen et al. 1997).

Obviously, a little drop in the Cr content completely changes the nature of the oxide scale. As a consequence, the evolution with time of the Cr content in the subsurface zone has to be considered with care. Indeed, the oxidation is consuming Cr; thus, after some time, the Cr concentration falls under the level required for forming the protective oxide layer. This can be particularly detrimental in the case of scale cracking, because the Cr‐depleted metal is then directly in contact with the corrosive environment. The resulting oxide scale is thus not protective anymore. This phenomenon explains the high corrosion rates observed for cyclic oxidation (Pillai et al. 2004). In this case, scale cracking is accelerated due to the repeated cooling process and leads to accelerated Cr depletion.

Grünling et al. (1979) reviewed the corrosion mechanisms that can lead to a decrease in creep resistance. A few examples of these mechanisms are pore formation near the surface due to the growth of an oxide scale by cationic diffusion, depletion of the matrix in solid solution‐hardening elements as a result of selective oxidation, decomposition of hardening precipitates, and enhanced grain boundary slip by diffusion of noxious elements into the grain boundaries. More recently, it has also been suggested that injection of vacancies deep into the metal due to cationic oxidation may also play a role in the reduction of creep resistance due to oxidation (Dryepondt et al. 2005).

As mentioned before, the heat exchanger tubes are submitted to several constraints, resulting in creep, e.g. due to the high pressures inside the tube or other operational parameters. The formation of thick oxide scales has therefore several consequences for the service behavior of components:

1. The loss of material reduces the availability load‐bearing cross section, so that as oxidation proceeds, the stresses acting on the component increase. The magnitude of the effect will, of course, depend on the initial wall thickness and the oxidation rate. By considering creep data measured in air as well as the oxidation rates mentioned earlier for steam and simulated coal firing, Ennis et al. (1993) and Nickel et al. (1998) evaluated the influence of oxidation on the creep resistance. For example, for a thin‐walled heat exchanger tube of about 5 mm wall thickness, the reduction of the load‐bearing cross section due to corrosion would have severe consequences for service life.
   
2. The thermal insulation effect of a thick oxide scale, which reduces the heat transfer across the component wall, can lead to overheating. A small increase in component operating temperature can reduce the stress‐rupture life considerably.
   
3. The matrix becomes depleted in those elements that are selectively oxidized to form the thick oxide scales, leading to microstructural changes and associated potential effects on the long‐term strength. This effect has been investigated in particular by comparing the creep‐rupture time in air of specimens preoxidized in steam with heat‐treated specimens (Ennis et al. 1993).

By the same method of testing the creep resistance in air of specimens preoxidized in steam, Schütze et al. (2004) investigated the effect of load on the oxidation behavior. This method particularly allowed comparing the healing properties of the oxide scale. Indeed, the imposed load brings the oxide scale to crack. The ability of the material to reform a protective scale will thus be crucial for its lifetime. For 9% Cr steels, it was shown that silicon offers some advantages for healing the protective oxide scale, but this effect is spoiled in the presence of Mo in dry environment.

The preceding description of the properties of 9–12% Cr steels shows that further increasing the heat exchanger temperatures will require higher corrosion resistance and creep strength. If this has to be achieved with chromia‐forming alloys, higher Cr contents are required (Staubli et al. 2002; Viswanathan et al. 2002). Table 16.5 shows the benefit of higher Cr contents with regard to corrosion resistance.

Furthermore, steam oxidation tests showed that depending on the grain structure, a minimum of 20–25 wt% is recommended for forming protective Cr₂O₃ oxide scale at 650 °C (Otsuka et al. 1989). This amount lies clearly above the limit of martensitic steels, and since body‐centered cubic ferritic steels have poor mechanical properties, the ferritic–martensitic steels are preferentially replaced by austenitic steels or Ni‐based alloys. However, at higher temperature, the low Cr diffusivities in austenitic steels limit the Cr supply toward the metal–oxide interface and therefore the ability of the alloys for healing a cracked oxide scale. In austenitic steels, increasing the Ni content seems to have a beneficial effect on the resistance to water vapor oxidation (Peraldi and Pint 2004). However, when comparing other properties of these materials, not only the higher costs are a drawback for the replacement of ferritic–martensitic steels. Firstly, the lower thermal conductivities of austenitics and nickel‐based materials limit the heat transfer (Table 16.5). Secondly, a higher thermal expansion has to be taken into account for plant design. This also means that the thermal expansion mismatch between the alloy and the oxide scale is higher, resulting in a higher tendency for oxide spallation. Thus, even though some austenitic steels form a thin protective Cr₂O₃ layer, with a thermal expansion of about 8 × 10⁻⁶ K⁻¹ (Robertson and Manning 1990), this might spall off due to temperature changes. For comparison, up to 400 °C, the thermal coefficient for magnesite lies between 11 and 14 × 10⁻⁶ K⁻¹ (Armilt et al. 1978), which is close to the expansion of P91. The tendency for oxide scale spallation is thus higher on austenitic steels compared with ferritic–martensitic ones. The concern is that spalled flakes appearing inside the exchanger tube lead to erosion and blocking of the steam turbine components. Yet, Peraldi and Pint (2004) observed that increasing the Ni content in austenitic steels reduces spallation. A minimum of 20 wt% Ni is nonetheless recommended for observing a significant effect at 650 °C.

Eventually, from a mechanical point of view, ferritic–martensitic steels offer the best creep properties at low temperature. The most recent modifications could be used up to 650 °C. Furthermore, austenitic steels can be used in a small range up to 670 °C. Higher temperatures require the use of Ni‐based alloys (Smith and Shoemaker 2004; Viswanathan et al. 2002).

16.3 Coal‐ash Corrosion

Corrosion in coal‐fired boilers is frequently referred to as coal‐ash corrosion particularly when the aggressive phenomena are mainly due to the nature of the ash that deposits on the heat‐absorbing surfaces of metallic components (waterwalls, superheater, and reheater tubes of boilers and furnaces).

The waterwall tubes can undergo a very rapid wastage even though the metal temperature may have been over 400 °C and sometimes would have been cooler than this. In the vicinity of the combustion zone, much of the ash will still be molten, and some of this is deposited on the waterwalls as a slag layer. The combustion zone is designed to be located in the center of the furnace enclosure, at a distance from the walls. Poor adjustment of the burners, wear of the burner nozzles, a change in the combustion characteristics of the coal, or several other factors can result in the combustion zone being displaced, or simply being larger than had been expected. In this case, the slag deposit in the region where the combustion zone approaches the walls contains unburnt carbon and unoxidized pyrite. It is under these circumstances that rapid wastage can result.

Early investigators thought that this wastage was also associated with alkali sulfate species. In fact, alkali sulfates deposited on the waterwalls may react with SO₂ or SO₃ to form pyrosulfates, such as potassium pyrosulfate (K₂S₂O₇) and sodium pyrosulfate (Na₂S₂O₇), or possibly complex alkali iron trisulfate. The latter compounds are formed in thicker deposits after long periods of time at about 480 °C. The K₂SO₄–K₂S₂O₇ system forms a molten salt mixture at 407 °C when SO₃ concentration is 150 ppm. The corresponding sodium system can become liquid at about 400 °C, but it requires about 2000 ppm SO₃ for this to occur; such levels of sulfur oxides are likely only under deposits. Thus, molten salt attack on the tube metal by K₂S₂O₇ is more likely and occurs according to the reaction

16.13