Volume 13 - Corrosion Part 14 docx

Brass and austenitic stainless steel condenser tubes, in particular, are known to have failed by pitting and crevice corrosion.. 4 Example of pitting in AISI type 316 stainless steel in

Corrosion of Fossil Fuel Power Systems

R.I Jaffee, Electric Power Research Institute

THE ELECTRIC POWER INDUSTRY uses three types of fossil-fired plants The most common plant is the pulverized coal-fired steam power plant, which may be used either as a baseload plant (where the plant runs continuously at capacity except for scheduled outages for maintenance) or for intermediate loads between the steady baseload and higher loads needed daily Gas turbines are used for peak loads that occur for an hour or two each day Combined cycles using both gas and steam turbines are generally intended for baseload sustaining intermediate-load service

The fuels used in fossil-fired plants are natural gas, petroleum, and coal Natural gas is generally extremely pure and does not constitute a corrosion threat, unless firing is substoichiometric for reducing NOx emissions Petroleum fuels can be corrosive to boilers if they contain vanadium and alkali metals; these produce liquid vanadium oxides or alkali sulfates, both of which are highly corrosive to metals in the combustion chamber or hot-gas passages Coal can contain such impurities as sulfur, chlorides, and alkali metals, which are extremely corrosive Many coals are virtually unusable except under deaerated service conditions

Steam Power Plants

Figure 1 shows an illustration of a fossil-fired steam power plant Three fluid flow loops circulate through the system: fuel-air, water-steam, and condenser cooling In the fuel-air loop, the fossil fuel is burned in air, transfers its heat to a series of heat exchangers, is cleaned of particulate matter, is scrubbed of sulfur oxides, and exits through the stack In the water-steam loop, clean feedwater is converted into superheated steam in a boiler, which expands through a series of turbines, converting its heat into mechanical energy, and is condensed, conditioned, pumped, and heated as feedwater In the condenser-cooling loop, cold water is passed through the condenser and can be recirculated if a cooling tower is used

or an be exhausted back to the source of the cooling water Each fluid loop possesses its unique corrosion problems

Fig 1 Schematic of a coal-fired power plant

The fossil fuel is burned in a very large chamber constructed of water walls consisting of vertical or spiral steel tubes about 60 mm (2.4 in.) in diameter that are welded together in a web about 20 mm (0.8 in.) wide The feedwater ascends

the water walls and is heated by the combusted fuel In subcritical boilers, the generated steam is separated from the water such that the water is returned to the bottom of the water wall through downcomers, but the saturated steam is super-heated in tubular heat exchangers suspended in the gas stream In supercritical boilers, the pressure is above the critical point, and the liquid becomes superheated vapor without undergoing a phase change The feedwater is conditioned to be slightly alkaline, but the fluid in the boiler may become acidic or caustic, depending on the presence of corrosion deposits and flow interruptions Under acidic conditions, the steel boiler tubes may be hydrogen embrittled; under caustic conditions, the tubes may be caustic gouged In supercritical boilers, overheating and excessive internal scaling in water walls may occur if boiling undergoes departure from nucleate boiling conditions at the region near the critical pressure

The superheaters and reheaters are subject to steam oxidation on their inner surfaces and to hot corrosion on their outer surfaces Steam oxidation results from attack by superheated steam, which acts similarly to oxygen at the same temperatures For low-alloy steels, steam oxidation is of concern above 540 °C (100 °F) This is about the temperature at which creep strengths limit low-alloy steels, and it becomes necessary to use alloys with higher chromium contents for higher allowable stress and for better steam oxidation resistance

Fire-side corrosion in superheaters and reheaters is a typical problem In coal-fired boilers, it exhibits a maximum rate at

700 to 750 °C (1290 to 1380 °F), where the corrodent is liquid, and decreases to a minimum at higher temperatures, where the corrodent does not condense The liquid ash is generally an alkali sulfate or a complex alkali iron trisulfate At higher temperature, corrosion is predominantly the oxidation of uncooled parts, such as hangers

Another types of boiler, the fluidized-bed boiler, was once thought to be free of fire-side corrosion because it operated under dry conditions at about 850 °C (1500 °F) The coal is burned in a fluidized bed composed of limestone or dolomite The calcium sulfate (CaSO4) product of desulfurization is in equilibrium with the calcium oxide (CaO) absorbant At low oxygen potentials, such as those that occur in the condensed bed itself, the sulfur potential may rise high enough to sulfidize many otherwise very corrosion-resistant alloys

Following the coal-air loop in a conventional power plant further, the flue gas passes through heat exchangers for preheating combustion air It is important to maintain the flue and gas temperature above its dew point in order to avoid the deposition of sulfuric acid (H2SO4) After passage through the precipitator while above its dew point, the flue gas is scrubbed of its sulfur dioxide (SO2) content in a flue gas desulfurization scrubber There are many points of corrosion concern in the scrubber, but the most serious areas are the inlet duct, where the SO2-laden gas is hottest and a wet/dry interface exists, and the outlet duct, where the scrubbed gas, still containing sulfur trioxide (SO3), begins to condense on the walls of the duct

Once the steam has been compressed and superheated, it leaves the boiler through heavy-wall pipes and enters the pressure turbine There it expands and returns to the boiler for reheating before entering the intermediate-pressure turbine for a second expansion After one or two expansions and reheatings, the steam enters the low-pressure turbine

high-As long as the steam is dry, there is little corrosion in the high- or intermediate-pressure turbines except when condensation of solid sodium hydroxide (NaOH) occurs However, when expansion of the steam in the low-pressure turbine reaches the point of initial condensation (at the so-called Wilson Line), high-concentration chloride and sulfate salt solutions may deposit if the steam is contaminated The distribution of salts between steam vapor and steam condensate is such that the condensate may be 106 times more concentrated than the vapor Thus, the premissible impurity levels in the feedwater are measured in parts per billion in order to protect the low-pressure turbine at the Wilson Line, which generally occurs at the next to the last (L - 1, or last minus one) row of turbine blades The Wilson Line shifts to a highest temperature point in the low-pressure turbine at reduced load such that in load-cycling plants there is alternate wetting and drying of the salt deposit as the load increases and decreases during cyclic operation This is a serious condition for corrosion of blades and disks in the L - 1 row

The salt solution is often acidic as a result of evaporation of ammonia (NH3) from the water-conditioning process During shutdowns, oxygen and carbon dioxide (CO2) may dissolve in the acidic salt, aggravating the corrosive condition at the L

- 1 row of a low-pressure turbine The commonly used 12% Cr turbine blade alloy becomes pitted under these conditions and may lose up to 90% of its fatigue strength Corrective measures include using blades designed to be strong enough to operate with pitted surfaces, cleaning and maintaining the steam to avoid corrosive salt deposition, using more corrosion-resistant low-pressure turbine blade materials (such as titanium alloys), or protecting 12% Cr steel blades with corrosion-resistant coatings

Corrosion problems may occur in the condenser-cooling loop, especially if the cooling water is sulfide-contaminated seawater or brackish water acting on copper-base tubes or tubesheets Also, pitting or crevice corrosion may occur under deposits or barnacles or between tubes and tubesheets However, the primary concern with condensers is leakage of seawater or contaminated cooling water into the water-steam loop, which operates below atmospheric pressure; this leakage would result in drastic corrosion effects on boiler and turbine components

Gas Turbines

In a gas turbine, inlet air is compressed in a compressor, reacted with fuel in a combustion chamber, and directed at stationary airfoil vanes and through rotor blades or buckets constituting the turbine stage Thus, the entire fuel and air input passes through the gas turbine without an intermediate heat exchanger Any corrosive impurities present in either fuel or air will affect the high-temperature components, primarily the combustor, nozzle diaphragm, and turbine The principal threats are oxidation and hot corrosion These have been largely met by using alloys of increased chromium content, particularly Ni20Cr and Co-30Cr alloys However, as the turbine inlet temperature increased to achieve higher thermal efficiency, it became necessary to strengthen the alloys, which resulted in lower chromium contents and greater vulnerability to oxidation and hot corrosion The use of bypass air cooling of the hot-section parts essentially reduced the metal temperature to a point at which the high-temperature strength was sufficient, while the gas temperature increased progressively The high-temperature high-velocity gas stream causes evaporation of volatile chromium trioxide (CrO3) from otherwise protective chromic oxide (Cr2O3) scales The alloys subjected to the highest turbine temperatures are protected by coatings containing over 5% Al, which from protective aluminum oxide (Al2O3) scales

If the fuel or inlet air contains alkali metals, sulfur, or vanadium as impurities, hot corrosion may occur This is combated

by limiting these impurities in the fuel Additives such as magnesium oxide (MgO) also help Air filtration is used for reduce ingestion of airborn impurities Vanes and blades are washed periodically to remove accumulated salts The coatings that form Al2O3 protective scales are sometimes improved by platinum metal additions or sublayers Also, coatings with small additions of yttrium promote adherence of all Al2O3 scale, which otherwise might spall off

Combined Cycle Plants

In a combined cycle power plant, the gas turbine is used as a high-temperature topping cycle whose exhaust gas enters a waste heat boiler, which raises steam to operate the steam turbine and generator To a large extent, corrosion problems in the combined cycle are simply the sum of the corrosion problems in the gas turbine and the steam boiler and turbine Control of impurities in the inlet air and fuels is essential Corrosion from the use of gasified coal, which may have caused severe problems in the gas turbine and steam generator, has largely been eliminated by scrubbing the gasified coal in a water-quenching operation There are severe corrosion problems in the radiant cooler used to generate process steam from the raw gasified coal, which contains hydrogen sulfide (H2S) and other corrosive agents before the scrubbing operation These problems can be handled by limiting the temperature of the radiant cooler to be commensurate with the heat-exchanger material used, which is generally coated or clad steel

Corrosion of Condensers

Barry C Syrett, Electric Power Research Institute; Roland L Coit, Consultant

A steam surface condenser is a shell and tube heat exchanger that is positioned immediately downstream of the pressure steam turbine Heat is transferred from steam on the outside of the condenser tubes (the shell side or steam side)

low-to water on the inside (the tube side or water side) A schematic of a typical electric power plant condenser is shown in Fig 2

Fig 2 Schematic of a typical condenser in an electric power plant

The condenser is a particularly critical component in a power plant because its failure can affect many other components

in the steam-water cycle The root cause of many of the corrosion problems in fossil fuel boilers, nuclear steam generators, low-pressure steam turbines, and feedwater heaters has been traced to condensers that have leaked and allowed contamination of the steam condensate with raw cooling water and air Most tube leaks are caused by corrosion, but some failures are purely mechanical, such as those caused by steam impingement (erosion), tube-to-tubesheet joint leaks, mechanical ruptures from foreign object impact, and tube vibration resulting in fretting wear and fatigue Purely mechanical failures will not be discussed further

Corrosion mechanisms that have led to failure or serious problems in power plant condenser are summarized in Table 1 Table 1 reflects known service problems to date, rather than susceptibilities that might be inferred solely from laboratory tests; each form of failure occurs only under specific environmental and metallurgical conditions Information on tubesheet materials requires special explanation Because tubesheets are very thick (>25 mm, or 1 in.), corrosion rates can

be 15 to 50 times higher than in condenser tubes and still be considered acceptable in most cases Furthermore, even if corrosion is a serious problem in a tubesheet, inspections are usually scheduled frequently enough, and the tubesheet is thick enough, that suitable repair can made or corrosion protection procedures can be instituted along before the tubesheet

is penetrated Thus, although Table 1 indicates that copper alloy tubesheets have suffered significant (often severe) galvanic corrosion under certain conditions, leakage of cooling water through the tubesheet from the water side to the steam side has rarely occurred Each of the corrosion mechanisms responsible for failures in condensers will be reviewed below Some of the methods of preventing these failures will also be summarized

Table 1 Corrosion mechanisms that have caused problems in power plant condensers under certain conditions

Corrosion mechanism Alloy

Erosion- corrosion

Sulfide attack

Dealloying Crevice

corrosion/

pitting

Galvanic corrosion

Environmental cracking

Condensate corrosion Copper alloys

Muntz metal (tubesheets) N (W) (W) N W N N

Aluminum bronze (tubesheets) N (W) N N (W) N N

Aluminum bronze W W (W) (W)(a) (W) (S) (S)

Admiralty brass W W (W) W(a) W W/S S

(a) Perhaps a problem only when sulfide is present

(b) Can induce galvanic-corrosion of adjacent copper alloys, iron, and carbon steels when used in

seawater or other highly conductive waters

(c) A problem in heats containing only 25.5% Cr and 3% Mo

Erosion-Corrosion

Erosion-corrosion is a relatively common waterside phenomenon that is a problem only in copper alloy condenser tubes

It occurs in areas where the turbulence intensity at the metal surface is high enough to cause mechanical or electrochemical disruption of the protective oxide film In these turbulent regions, pitlike features develop Turbulence increases with increasing velocity and is greatly influenced by geometry For example, turbulence intensity is much higher at tube inlets than it is several feet down the tubes; this results in the phenomenon of inlet-end erosion-corrosion Tube inserts have been used to circumvent this problem A tube insert is a tightly fitting internal sleeve, typically 150 to

300 mm (6 to 12 in.) long, made from a material resistant to erosion-corrosion that shields the susceptible tube ends However, unless there is a smooth transition between the end of the insert and the tube, the insert can itself create turbulent conditions and promote erosion-corrosion further down the tube (Fig 3) Recent experiments have demonstrated that inlet-end erosion-corrosion can also be prevented by installing a cathodic protection system in the water box region

Fig 3 Erosion-corrosion occurring immediately downstream of a nylon insert in an aluminum brass condenser

tube cooled by seawater

Erosion-corrosion may also occur when marine life or debris in a tube creates a partial blockage, resulting in locally high velocities through the restricted opening In these cases, the best solution is to keep the tubes clean, using one or more of the following methods:

• Install or improve intake screens

• Install on-line sponge ball cleaning

• Periodically reverse flow (backwash)

• Manually clean with brushes, balls, scrapers, and so on (off line)

• Prevent biofouling by chlorination or thermal shock

Alternatively, some copper alloys benefit from periodic dosing of the water with ferrous ions (Fe2+), which are usually added as ferrous sulfate (FeSO4) solution The Fe2+ ions deposit as a protective lepidocrocite [FeO(OH)] layer on the copper alloy surface

Sulfide Attack

This form of attack affects only copper alloys and occurs when the cooling water, most often brackish water or seawater,

is polluted with sulfides, polysulfides, or elemental sulfur As little as 10 mg/m3 (10 ppb) of sulfide in the cooling water can have a detrimental effect, and concentrations far greater than this are often measured in polluted harbors and estuaries Sulfide attack manifests itself in many ways It can greatly increase general corrosion rates, and it can induce or accelerate dealloying, pitting, erosion-corrosion, intergranular attack, and galvanic corrosion Penetration rates in polluted waters can be extraordinarily high, sometimes as high as 20 mm/yr (80 mils/yr) No copper alloy is resistant to sulfide attack, and the relative performance of copper alloys in polluted or brackish waters seems to depend on the precise environmental conditions

If the incoming water contains sulfide and there is no obvious method of eliminating the source, the most successful method of reducing or preventing the problem is to dose the water periodically or continuously with FeSO4or some other source of Fe2+ ions However, sulfide attack can also occur in condensers cooled with nominally unpolluted water if marine organisms trapped within the condenser during downtime are allowed to die and putrefy to produce sulfides This can probably be prevented, or at least reduced, by turning on the pumps for an hour or two each day to flush out the condenser with fresh seawater In addition, sulfate-reducing bacteria can produce sulfides under debris and deposits where the oxygen content is low Thus, the risk of sulfide attack is greatly reduced if the copper alloy tubes are regularly cleaned

Dealloying

Another water-side problem in brass-tubed condensers is dealloying Dezincification is an example of dealloying that has been observed in utility condensers In dezincification, zinc is selectively removed from brass alloys to leave a copper-rich surface layer

Dealloying is rarely the cause of condenser tube failure, but when it does occur, it is normally restricted to localized areas, such as beneath deposits or at hot spots This results in plug-type dealloying Clearly, maintaining clean tubes will reduce the incidence of this type of failure A much less localized from of dealloying, termed layer-type dealloying, is rare in tubes, but has occasionally occurred in brass tubesheets, particularly in conjunction with galvanic corrosion in induce by titanium or stainless steel condenser tubes Under such circumstances, a cathodic protection system installed in the water box will control both galvanic-corrosion and dealloying

Crevice Corrosion and Pitting

Some stainless steels and cooper alloys are susceptible to water-side pitting and crevice corrosion Brass and austenitic stainless steel condenser tubes, in particular, are known to have failed by pitting and crevice corrosion There is limited evidence that copper alloys have adequate resistance to these forms of corrosion if the cooling water is completely free of sulfide Certainly, susceptibility seems to be greatly increased when sulfide is present

Pitting and crevice corrosion of stainless steels are more dependent on the chloride content of the cooling water than on the sulfide content, although laboratory data have demonstrated that the detrimental effects of chloride are accentuated in

the presence of sulfide Some alloys, such as AISI type 304 and 316 stainless steels, which generally perform well in freshwaters or slightly brackish waters, suffer rapid pitting and crevice corrosion in seawater (Fig 4) The newer, more highly alloyed stainless steels, including AL6X (UNS NO8366), AL29-4C (Fe-29Cr-4Mo-0.35Si-0.02C-0.02N-0.24Ti), and Sea-Cure (Fe-27.5Cr-3.4Mo-1.7Ni-0.4Mn-0.4Si-0.02C-0.5Ti+Nb), generally perform well even in seawater However, a few failures have been reported for AL6X and for some of the early heats of Sea-Cure

Fig 4 Example of pitting in AISI type 316 stainless steel in seawater service

Again, tube cleanliness is a critical issue because debris and deposits promote the formation of concentration cells (the precursor to crevice corrosion) and because they favor the production of sulfides The tube-cleaning techniques summarized earlier in this section are therefore equally useful in preventing crevice corrosion and pitting in copper alloys and stainless steels

Galvanic Corrosion

Galvanic corrosion is not a problem is poorly conducting waters, such as those normally found on the steam side of condensers However, it can be a water-side problem in condensers cooled with seawater or with medium- or high-conductivity fresh and brackish waters

In seawater, tube materials, such as copper-nickel alloys, stainless steels, and titanium, are more noble than tubesheet materials, such as Muntz metal and aluminum bronze Consequently, the tubesheet may suffer galvanic attack when fitted with these more noble tubes (Fig 5) Laboratory tests have demonstrated that the rate of galvanic corrosion of a Muntz metal tubesheet fitted with titanium or stainless steel tubes can exceed 5 mm/yr (200 mils/yr) in seawater Similarly, if stainless steel inserts are installed in copper alloy tubes to prevent inlet-end erosion-corrosion, rapid galvanic corrosion can be promoted in the tube close to the insert/tube interface

Fig 5 Galvanic corrosion of a Muntz metal tubesheet, fitted with AL6X stainless steel tubes, after 1 year of

service

Other galvanic couples can exist in a condenser, but in each case, the recommended method of alleviating the problem is

to install a cathodic protection system in the water box Cathodic protection current requirements can be reduced by coating the tubesheet and water box with a nonconducting material

Environmental Cracking

Stress-corrosion cracking (SCC) and hydrogen embrittlement cracking are forms of environmental cracking that can affect condensers Hydrogen embrittlement cracking was identified as a problem on the water side of ferritic stainless steel tubes in a couple of condensers fitted with cathodic protection systems It is believed that hydrogen was generated

on the surface of the tubes by the passage of too high a cathodic protection current and that this hydrogen promoted slow crack growth and failures at the ends of the tubes The tube ends were particularly susceptible to hydrogen embrittlement cracking because roller expansion during fabrication introduced higher-than-normal residual stresses in these zones

Entry of hydrogen into cathodically protected titanium tubes is also possible if too high a cathodic protection current is delivered In such cases, the absorbed hydrogen can react with the metal to form a brittle titanium hydride phase which could conceivably crack and lead to premature failure However, so far no titanium condenser tube failures have been reported One electric utility that grossly overprotected the waterbox region of a titanium-tubed condenser discovered that the ends of the tubes were severely hydrided, but even here the affected tubes did not leak

Apart from the rather unusual failures in ferritic stainless steels, environmental cracking is a problem only in copper alloys, specifically the brasses Here, SCC (not hydrogen embrittlement cracking) is the mechanism of failure Most SCC failures initiate on the steam side of the tubes, and all occur when the steam condensate contains high concentrations of

NH3 and oxygen The NH3 is derived from the chemicals added for boiler feedwater chemistry control, and oxygen originates from air that leaks into the system through imperfectly maintained turbine glands, expansion joints, valve packing glands, and so on The NH3 and oxygen concentrations are particularly high in the air removal section, and it is here that SCC occurs most frequently

Steam-side SCC can be controlled by ensuring that the condensate on the tubes has a low oxygen concentration The maintenance of an airtight system requires continuing attention to all seals, glands, and joints that are subjected to internal pressures less than atmospheric during start-up, normal operation, or shutdown Helium tracer and similar techniques allow leaks to be detected with moderate ease

Water-side SCC of brasses has occurred less frequently than steam-side attack, and in most cases, the species responsible for the failure were not positively identified However, NH3 and its derivatives (nitrates and nitrites) are often suspected

of promoting SCC Possible sources of these species are farm fertilizers (runoff) and decaying organisms in polluted water Water-side failures frequently initiate beneath surface deposits, probably because the deleterious species can concentrate beneath the deposit to levels that favor SCC Thus, once again, tube cleanliness is important, and cleaning techniques can be used to minimize water-side SCC

No matter which side of the tube is susceptible to environmental cracking, the incidence of cracking can be controlled by reducing or eliminating residual tensile stresses Roller expansion of the tubes during installation will always introduce some residual stresses, but care should be taken to avoid expansion beyond the back of the tubesheet, and event that can lead to particularly high residual stresses In addition, fully stress-relieved tubes should be used, and during installation, they should not be bent or mechanically abused

Condensate Corrosion

Copper alloy condenser tubes, particularly brass condenser tubes, are susceptible to condensate corrosion in steam condensate that contains high concentrations of HN3 and oxygen Consequently, condensate corrosion, like SCC, is most prevalent in the air removal section Condensate corrosion, also known as NH3 attack, is a form of corrosion that is localized not by microstructural features in the metal but by the localization of the corrosive environment (Fig 6) For example, the slight tilt routinely given to condenser tubes may promote flow of some of the steam condensate toward one side of a tube support plate There, the flow from a large number of tubes can collect and run down the plate surface Such localized flow can create deep circumferential grooves, termed condensate grooving, in the tubes immediately next to the support plate (Fig 6b) if the condensate contains high levels of HN3 and oxygen Condensate corrosion can be controlled

by reducing the oxygen concentration in the condensate, as discussed previously for steam-side SCC, or by selecting more resistant alloys, such as copper-nickel alloys, or completely resistant alloys, such as stainless steels and titanium

Fig 6 Examples of NH3 attack on admiralty brass (a) The unattacked tube end (left) was protected by the tubesheet (b) Condensate grooving that occurred at one side of a support plate

Corrosion Prevention

Many corrosion modes operating in condensers can be prevented if only two maintenance procedures are followed First,

if air is eliminated from the steam, condensate corrosion and SCC of copper alloy tubes can be prevented Second, if condenser tubes are kept clean and free of deposits, debris, and biofouling on the water side, sulfide attack, dealloying, erosion-corrosion, crevice corrosion, pitting, and SCC can be prevented or minimized

Corrosion of Deaerators and Feedwater Heaters

Robert J Bell, Heat Exchanger Systems, Inc

Deaerators (direct contact deaerating feedwater heaters are used in fossil and a few nuclear power plants primarily to remove dissolved gases (mostly oxygen and nitrogen) from condensate/feedwater and to raise the condensate temperature

by exchange with extraction steam by mechanical deaeration Another function is to provide deaerator storage capacity and proper suction conditions for the boiler feed pump Closed feedwater heaters (Fig 7) are used in power plants to increase the overall cycle efficiency of the plant by delivering to the boiler or steam generator water at higher temperatures, thereby reducing the heat required to produce steam This is accomplished by heating the feedwater (condensate) using extraction steam from the turbine

Fig 7 Schematic of a 3-zone feedwater heater Zones: desuperheating, condensing, and subcooling

Deaerators and feedwater heaters are susceptible to various forms of corrosion This section will discuss corrosion in these applications as well as the measures that can be taken to minimize it

Closed Feedwater Heaters

Corrosion problems in closed feedwater heaters are usually manifested as tube failures Although tube failures may be a symptom of another problem, such as a baffle failure, the predominant causes of tube failures are destructive vibration, impingement erosion, cavitation-type erosion, and corrosion Corrosion is a lesser cause than the others The tube alloys most frequently used in closed feedwater heaters and their associated corrosion concerns are summarized in Table 2 and will be described in the rest of this section

Table 2 Tube alloys used in closed feedwater heaters and associated corrosion mechanisms

Corrosion mechanism and alloy susceptibility(a) Alloy

SCC Exfoliation Erosion-

corrosion

Crevice corrosion

Pitting Condensate

corrosion

General corrosion

Snake skin

Austenitic stainless steels S I I S S I I I

Ferritic stainless steels (for example,

Stress-corrosion cracking occurs when an alloy under a tensile stress (applied and/or residual from manufacture or

welding) is exposed to a specific corrosive environment The chloride ion (Cl-) induced SCC of austenitic stainless steels, for example, AISI type 304, in power plant systems is well documented Failures due to SCC in feedwater heaters have not been significant in number, and gross deviation from normal feed-water chemistry is the universal cause The potential for concentration of corrosively aggressive chlorides occurs mainly on the shell side in the desuperheating zone

in the event of a small leak This factor, combined with controlled feedwater conductivity, pH, and dissolved oxygen content, possibly explains why there have been few reported failures of stainless steel tubes due to SCC

The copper alloys are susceptible to NH3-induced SCC When many plants switched to all-volatile water treatment (a source of NH3), the incidence of SCC in copper alloys correspondingly increased The failures typically occur in poorly vented areas where oxygen and NH3, can concentrate Stress-corrosion cracking can be minimized by reducing the residual stresses in the material, for example, by stress relieving tube U-bends Because of the susceptibility of copper alloys to other corrosion mechanisms and because of the overall need to reduce copper within the system, the application

of the copper alloys has been greatly diminished

There have been a few reported failures in Monel 400 tubes, primarily at the U-bend because of SCC Such failures can

be prevented by limiting residual stress

Exfoliation is a form of corrosion in which the corrosive penetration runs primarily on a plane that is parallel to the tube

surface (Fig 8) Generally, attack is along boundaries of grains that are elongated in the drawing direction The expensive force of insoluble corrosion products tends to force these grains apart The outside surfaces of exfoliated tubes exhibit heavy scaling with a leafy appearance; hence the term exfoliation Both 70Cu-30Ni and 80Cu-20Ni alloys are susceptible

to this form of attack; the degree of susceptibility increases with nickel content

In the power generation industry, exfoliation was first encountered in units that were converted from baseload operation to peaking/cycling service (Ref 1) The peaking/cycling service results in greater and more frequent exposure to oxygen This problem is reduced by steam or nitrogen blanketing when the unit is out of service

Erosion-Corrosion Many closed feedwater heater tubes

are protected from accelerated corrosion by the formation

of a corrosion product film, which in turn acts as a diffusion barrier that can limit the net diffusion rate of the corrodent (for example, dissolved oxygen) to the underlying base metal, By limiting the net diffusion rate of corrodent through the film, the corrosion rate is also limited Turbulence at tube entrances can produce excessive shear forces on the protective corrosion product film; these forces reduce the section thickness of the film and consequently accelerate corrodent flux and the corrosion process This process is erosion-corrosion

Although other tube alloys used in closed feedwater heaters are theoretically susceptible to erosion-corrosion, only carbon steels have exhibited widespread failures (Fig 9) The primary causes are excessive water velocities, either by design or

by abnormal operation of the feedwater heaters, and channel geometries that cause local zones of turbulence at the tubesheet Low feedwater pH, for example, less than 8.8, accelerates this mechanism because of the impact of pH on the carbon steel corrosion product film

Fig 8 Exfoliation of tubes in a closed feedwater heater

Fig 9 Erosion-corrosion of carbon steel tubes in a closed feedwater heater Courtesy of the Electric Power

Research Institute

Erosion-corrosion is also predominant at temperatures between 150 and 230 °C (300 and 450 °F) At these temperatures, the protective corrosion product film is soft and comparatively friable (Ref 2)

Pitting of feedwater heater tubes is generally limited to austenitic stainless steel alloys and carbon steels, Pitting of

austenitic stainless steel alloys is caused by the contamination of feedwater by chlorides Pitting typically occurs when chlorides are concentrated under deposits, within crevices, and through evaporation, for example, in the desuperheating zone The few reported incidents of chloride-induced pitting in austenitic stainless steel (type 304) are generally associated with excessive seawater intrusion permitted by condenser tube leaks (Ref 2) There are no reported similar occurrences with ferritic stainless steels, such as AISI type 439

Failure to control feedwater oxygen content properly (<7 ppb) has resulted in a few reported incidents of pitting in carbon steel tubing (Ref 3) One incident of carbon steel tube pitting was also attributed to excessive saltwater intrusion due to operating with a condenser tube leak (Ref 2)

General corrosion occurs in all closed feedwater heater tube alloys Corrosion rate, however, varies substantially

among the typical feedwater heater tube alloys Passive materials, such as the austenitic and ferritic stainless steels, demonstrate negligible uniform corrosion rates and are therefore considered essentially immune to this form of corrosive attack In general, susceptibility to general corrosion decreases as alloy selection changes from carbon steel to brasses, copper-nickels, Monel 400, and stainless steels (immune) The nature of the corrosion product film is the primary controlling variable

In terms of the corrosion product film, two characteristics are important; the corrodent flux through it and the uniformity

of diffusion resistance at all points General corrosion typically occurs where diffusion resistance is essentially the same

at all points

The variables affecting general corrosion, and therefore the corrosion product film, are pH, temperature, fluid velocity, and concentration of corrodent The magnitude of influence exerted by each variable significantly increases or diminishes

as a function of the exposed tube material

Two very critical parameters are introduced into the general corrosion reaction when cycling service begins First, greater amounts of oxygen and CO2 are present during low loads and during outages Because the corrosion rate of carbon steel is directly proportional to corrodent concentration, greater corrosion is likely with cycling service Second, cycling service creates thermal expansion-contraction problems that arise from the difference in the thermal expansion coefficients of the carbon steel tube and its corrosion product film Therefore, cycling service leads to fracturing of the protective corrosion product film The metal directly beneath these fracture sites has little, if any, remaining film to provide a protective diffusion barrier It follows, then, that fossil plant cycling can accelerate the corrosion of carbon steel tubes

Snake skin are a result of the redeposition of copper corrosion products on Monel 400 or 70Cu-30Ni high-pressure

feedwater heater tubes The deposition results in a thin flaky film that shrinks and subsequently sheds from the tubes when dried The film appearance is similar to a skin shed by a snake (Fig 10) It can cause a significant reduction in heat transfer Low-pressure closed feedwater heaters employing admiralty brass tubes are usually the source of the copper

Fig 10 Snake skin formed by the redeposition of copper corrosion products on copper alloy reheater tubes

Courtesy of the Electric Power Research Institute

Deaerators

Deaerators are subject to general corrosion, erosion-corrosion, SCC, and corrosion fatigue The corrodent for the latter mechanism is dissolved oxygen even at concentrations of 7 ppb or less This corrosion process usually begins as oxygen-induced pits, which in turn act as stress risers, that ultimately promote the initiation of (typically) transgranular cracks This process requires low-frequency stress cycles that produce greater contact time between the metal and corrodent

Corrosion of Steam/Water-side Boilers

R.B Dooley, Electric Power Research Institute

The cost to electric utilities from corrosion and deposition on steam- and water-side boilers is high approximately $3.5 billion per year One-half of the outages forced by plant failures have also been estimated to be attributable in some way

to water-side corrosion Contaminant deposition in most cases reduces equipment efficiency and induces corrosion by a variety of mechanisms and is implicated in a variety of boiler tube failure mechanisms that are most common in water walls and economizers

Water side deposits often begin as accumulations of corrosion products transported to the boiler from other parts of the system The corrosion product deposit is porous, in contrast to the protective magnetite (Fe3O4) film This porous deposit serves as a trap for corrosive impurities, such as caustic, chlorides, and acid sulfates

Boiler tube failures initiating in steam-containing tubing have also almost exclusively been the result of entrainment of contaminants within the steam Chlorides, sulfates, and caustic are the most common contaminants However, the growth

of Fe3O4 on the inside tube surface can also be a secondary contributor to tube failure If its growth rate is excessive, this will act as a thermal barrier and cause the tube wall temperature to rise, sometimes above the point at which excessive creep damage will result in an over-heating failure

Water Walls and Economizers

These components are usually manufactured of plain carbon steel (ASTM A210, ASME SA-192); in a few exceptions, the tubes are a low-chromium ferritic material (ASME SA-213, grade T-11) In most cases, the tube metal temperatures are less than 400 °C (750 °C) for subcritical boilers

Normal Protective Fe 3 O 4 Growth It has been pointed out many times that the only thing that protects a boiler is a

thin film of Fe3O4 on the water-side tube surface The ability to use inexpensive carbon or low-alloy steels in contact with water at high temperatures and pressures is due to the reaction between iron and oxygen-free, neutral, or slightly alkaline water:

by which a film of Fe3O4 is formed on the water-side surface of the tube; the kinetics of the reaction are parabolic, and the protective oxide consists of two layers described as the inner and outer In practice, the outer layer is seldom formed, because as the iron diffuses outward the Fe3O4 formed at the outer interface usually becomes entrained in the boiler water flow and then deposits, together with feedwater corrosion products in another region of the boiler, which may be of higher heat flux

The growth of Fe3O4 in economizer tubing occurs in a similar fashion to that in the water wall However, a common observation is that the water-side surfaces are more uneven and contain more pits than those in the water walls

Breakdown of the Normal Protection Mechanism The corrosion resistance of water wall/economizer tubing

depends on keeping the protective Fe3O4 in place, on the pH level of the water, and on the amount of contaminants Figure 11 shows the effect of pH on the rate of corrosion of steel by water Magnetite is unstable and soluble at pH values below 5 and above 12

Fig 11 Effect of pH on the corrosion rate of steel in water at 310 °C (590 °F) Upsets in water chemistry that

increase or decrease the pH of boiler water or of wall deposits can result in corrosion of the water wall tubes

Source: Ref 4

The significant categories of water-side corrosion failure mechanisms are caustic corrosion, hydrogen damage, and pitting (localized corrosion) A significant factor in these mechanisms is the amount of corrosion product deposited on the wall tube Caustic corrosion and hydrogen damage result from the breakdown of the protective Fe3O4 layer by the concentration of corrosive chemicals within a wall deposit As indicated below, different failure mechanisms will be experienced, depending on the contaminants present

Caustic corrosion is sometimes referred to as caustic attack, caustic gouging, or ductile gouging Caustic corrosion develops from the deposition of feedwater corrosion products in which NaOH can concentrate to high pH levels At high

pH levels, the protective Fe3O4 layer of the tube steel becomes soluble, and rapid corrosion occurs (Fig 11)

Caustic corrosion is caused by the selective deposition of corrosion products and NaOH at locations at high heat flux As porous deposits accumulate in high heat input areas, NaOH concentrates through a process known as wick boiling The caustic levels can concentrate from less than 100 ppm NaOH in the bulk water to over 200,000 ppm adjacent to the tube surface The electrochemical nature of the attack is shown in Fig 12 The hydroxide ions (OH-) are concentrated within the deposit layer so that the hydrogen ion (H+) concentration is highest in the boiler water

Fig 12 Schematic of the mechanism of caustic corrosion A caustic upset in boiler water conditions can result

in concentration of OH - ions in the deposit and generation of hydrogen gas into the boiler water Source: Ref 5

The susceptibility of high-pressure boilers to such corrosion damage may be reduced by minimizing the entry of forming materials and by performing periodic removal of the water-side deposits by chemical cleaning Rigorous monitoring and control of the water chemistry is necessary to prevent high caustic levels

deposit-Hydrogen damage develops from the generation of hydrogen during rapid corrosion of the internal surface of the tube The atomic hydrogen migrates through the tube steel, where it can react with iron carbide (Fe3C) to form methane (CH4) The larger CH4 molecules become trapped at the grain boundaries and cause a network of discontinuous internal cracks to

be produced These cracks grow, and some will link up to cause a throughwall fracture

Hydrogen damage is caused by operation of the boiler with low-pH water chemistry and the concentration of contaminants within the deposits on the internal tube wall The electrochemical nature of hydrogen damage is shown in Fig 13 Under acidic attack, H+ ions are concentrated within the deposit so that the cathode site is localized and adjacent

to the anode

Fig 13 Schematic of the corrosion mechanism of hydrogen damage Acid (low-pH) boiler water conditions can

result in concentration of H + ions in the deposit and generation of hydrogen gas atoms into the tube material Source: Ref 5

Monitoring and control of boiler water chemistry are important in preventing internal tube deposits and hydrogen damage The most common method of preventing increased corrosion and hydrogen damage is to develop operating guidelines for boiler water and particularly for the action to be taken when the boiler water is outside the guidelines (see the section

"Boiler Water and Steam Chemistry" in this article) For example, chemical cleaning should be immediately considered when the boiler water pH has been below 7 for more than 1 h

Pitting. Boiler tube failures caused by pitting or localized corrosion result from oxygen attack or acid conditions on the internal surfaces of the boiler tube The localized corrosion produces perforations of the tube wall when a small area on the tube becomes anodic to the rest of the tube surface and preferentially corrodes The anodic condition can develop from exposure of the tube to water with high acid or oxygen concentrations or at crevices

The oxygen in the boiler water reacts with and rapidly removes the hydrogen produced at the cathodes (Fig 13), thus accelerating the cathodic reaction The oxygen will also oxidize the Fe2+ ion Hematite iron oxide (Fe2O3) will form as the corrosion product and cover the craterlike perforation in the tube wall

Pitting failure can occur anywhere in the boiler, particularly in economizers, superheaters, reheaters, and the nonheated portions of water wall tubes For full protection against oxygen pitting during shutdown, it is necessary to keep the boiler full with hydrazine-treated water and blanketed or capped with nitrogen Oxygen pitting attack of economizer tubing can

be prevented by proper operation of deaerators and their heaters; by elimination of air in-leakage paths in low-pressure feedwater heaters, extraction piping, and condensate piping; and by injection of an oxygen scavenger chemical

Superheaters and Reheaters

Unlike water wall and economizer tubes, the tubes in the superheater and reheater are designed for a finite life, which is not based on failure but on a conservative creep criterion For steam temperatures of 540 °C (1000 °F), tube metal temperatures can exceed 600 °C (1110 °F), especially in the last stages of the superheat and reheat sections Tube materials can vary from carbon steels to low-chromium ferritic (ASME SA-213, grades T-11, T-22) to austenitic stainless steels (ASME SA-213, grades T304, T321, and T347)

Normal Protective Oxide Growth Chromium-containing steels exposed to high-pressure power station steam at

metal temperatures from 550 to 650 °C (1020 to 1200 °F) initially form a spinel-type oxide consisting of two layers whose relative thickness depend on the chromium content of the steel In superheaters and reheaters, however, the outer layer is essentially pure Fe3O4, and the inner layer is an iron-chromium spinel-type oxide containing the steel alloying elements

Breakdown of the normal protective oxide in superheaters and reheaters occurs similarly to that in water walls

and economizers If the normal protective growth continues as described above, no problems will exist There are two ways in which this protection can break down; the first is excessive growth and exfoliation, and the second is SCC

Excessive Growth and Exfoliation. Although excessive internal growth of oxide can elevate the tube temperature, a much more serious problem is solid-particle erosion of the turbine components when the oxide exfoliates and travels into the turbine For ferritic tube materials, corrosion initially occurs parabolically with time, but at a later stage, it can become linear, depending on the temperature

The duplex scales produced during the parabolic period are approximately of equal thickness and are parallel sided Any deviation of the corrosion rate from parabolic is associated with a multilayer scale Exfoliation occurs only in ferritic materials when this multilayer growth occurs Exfoliation of the scale is related to the stresses that are induced in the scale during temperature cycles by differences in thermal expansion between the scale and the tube

Austenitic stainless steels generally corrode more slowly than ferritic stainless steels under the same steam conditions; as

a result, the scales are somewhat thinner This is because of the higher chromium content in austenitic stainless steels

The initiation of scale exfoliation is not as easily defined for austenitic stainless steels, but may correspond either to the breakdown of the inner layer into a laminated structure or to some critical level of defects (voids) at the oxide/oxide interface The cause of exfoliation is again due to the difference in thermal expansion between the scale and the tube Replacement of existing superheater or reheater tubes with chromatetreated, chromized, or stainless steels (for the original ferritic tubes) is an effective method of reducing scale exfoliation

Stress-corrosion cracking failures in a boiler usually occur in the austenitic stainless steels used for superheater and reheater tubing However, SCC failures can occur in some ferritic reheater tubing when high levels of caustic are introduced from the desuperheating or attemperator spray water station

Condition for SCC initiation and propagation arise from the contamination of boiler water or steam, the introduction of high tensile stresses from service conditions, or the production of high residual tensile stresses during fabrication and assembly Common contaminants are chlorides and caustic, which result in transgranular cracks, and sulfur from chemical cleaning, which results in intergranular cracks

Boiler Water and Steam Chemistry

The cycle chemistry is of paramount performance for all of the corrosion mechanisms that occur in water walls, economizers, reheaters and superheaters, operation, and control Through proper understanding of these mechanisms, as well as the others around the cycle, it is possible to define water and steam quality limits in order to eliminate or lessen corrosion The reliability and availability of the equipment can be improved by following water and steam quality guidelines for all types of operation (baseload, cycling, and peaking), by adopting target and action levels, and by taking the appropriate action

Corrosion of Steam Turbines

Otakar Jonas, Consultant

The steam turbine is the simplest and most efficient engine for converting large amounts of heat energy into mechanical work As the steam is allowed to expand, it acquires high velocity and exerts force on the turbine blades Turbines range

in size from a few kilowatts for one-stage units to 1300 MW for multiple-stage multiple-component units comprising high-pressure, intermediate-pressure, and up to three low-pressure turbines For mechanical drives, single- and double-stage turbines are generally used Most larger modern turbines are multiple-stage axial-flow units Steam pressure and temperature conditions are governed by the boiler and range from less than 1.4 MPa (200 psi) saturated to more than 35 MPa (5000 psi) at 650 °C (1200 °F) superheated and supercritical (Ref 6)

Turbines are typically built for a 25-to-40-year life Recently, a small 8-MW turbine built in 1908 was inspected and found to be in excellent condition During this long life, corrosion and other material damage can accumulate and lead to premature failures Corrosion usually results from a combination of water chemistry, design, and material selection problems

From 1971 to 1980, steam turbines (excluding controls) contributed 6.7% of forced outages in United States utility units (Ref 7), which cost the utilities 192,575 GW·h in more than 9000 outages Outages due to corrosion were a major part of

this total A survey of about 500 utility turbines larger than 100 MW showed that corrosion failures occur in 4 to 5% of operating turbines each year (Ref 8, 9) Considering that it takes an average of 4 years for corrosion to result in a failure, about 20% of operational large turbines are under attack at any given time A survey of industrial turbines conducted by the American Society of Mechanical Engineers (ASME) found a similar degree of corrosion and deposit problems (Ref 10)

It has been estimated that corrosion losses in utility steam systems amounted to about $1.5 billion of the $70 billion annual cost of corrosion in the United States in 1978 (Ref 11); the losses today are about $3.5 billion The cost of corrosion of fossil turbine blades in the United States is about $300 million annually (Ref 12) Adding the estimated costs

of disk, bolt, bellows, and piping corrosion, the total annual cost of utility turbine corrosion in the United States is about

$600 million The cost of replacement power can be as much as two orders of magnitude higher than that of replacement parts

Major Corrosion Problems in Steam Turbines

Corrosion fatigue, SCC, pitting, and erosion-corrosion are the primary corrosion mechanisms in steam turbines Figure 14 and Table 3 show the distribution of corrosion within the turbine Pitting and corrosion fatigue of turbine blades and SCC

of disks are currently the two costliest problems Much research has been devoted to these two problems (Ref 12, 13, 14,

15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29), and progress is being made both in operation (better steam chemistry) and design (lower stresses, no crevices, limit on maximum strength to reduce susceptibility to SCC)

Table 3 Corrosion mechanisms in steam turbine components

mechanisms(a) Rotor Forged Cr-Mo-V or Ni-Cr-Mo-V low-alloy steel P, SCC, CF, E

Shell Cast carbon or Cr-Mo-V low-alloy steel SCC, E-C

Disks, bucket wheels Forged Cr-Mo-V, Ni-Cr-Mo-V or Ni-Cr-Mo low-alloy steel P, SCC, CF, E-C

Blades, buckets Stainless steels (12Cr or 17-4 PH), copper alloys P, CF, SCC, E

Bucket tie wires 12Cr stainless steels (ferritic and martensitic) SCC, P, CF

Shrouds, bucket covers Stainless steels (12Cr or 17-4 PH) P, SCC

Stationary blades AISI type 304 stainless steel SCC, SCC-LCF

Expansion bellows AISI types 321 or 304 stainless steels, Inconel 600 SCC, SCC-LCF

Erosion shields Weld-deposited or soldered Stellite type 6B; hardened blade materials (see

above)

SCC, E

Bolts Incoloy 901, Refractalloy 25, Pyromet 860 SCC, SCC-LCF

Valve bushings and

stems

(a) P, pitting: SCC, stress-corrosion cracking; CF, corrosion fatigue; E, erosion; E-C, erosion-corrosion; LCF,

low-cycle fatigue; OX, oxidation in steam General corrosion is experienced by all carbon and low-alloy steel components Solid-particle erosion is experienced in high-pressure and intermediate-pressure inlets (nozzle blocks, stationary and rotating blades, and valves) It is caused by exfoliation of steam-grown oxides

in superheater tubes and in steam pipes

Fig 14 Schematics showing locations of corrosion in steam turbine components P, pitting; CF, corrosion

fatigue; SCC, stress-corrosion cracking; C, crevice corrosion; G, galvanic corrosion; E, erosion; E-C, corrosion; SPE, solid-particle erosion

erosion-Statistics reveal interesting correlations Blade failures are most frequent (in fossil utility turbines) in the L - 1 row, which

is immediately before the saturation line (Ref 13, 14, 15, 16) In industrial turbines, blade failures are also most frequent

in the first row after the stage in which first moisture occurs (Ref 17) Blade pitting and cracking are more frequent in once-through boiler units (these units use all-volatile water treatment and condensate polishers) than in drum boiler units (Ref 18) Statistics from West Germany, the United Kingdom, and Japan indicate a very low incidence of blade failures

A recent survey of 494 utility units in the United States reported that 167 (34%) of the units experienced blade failures between 1970 and 1981 In 1980, 1981, and 1982 the failure rate was about 4.7% (Ref 12)

Stress corrosion of low-pressure turbine disks is specifically distributed for each type of turbine and each disk

location, indicating a correlation between cracking and surface temperature and steam condition This correlation is often related to the impurity concentration by evaporation of moisture The tendency toward cracking increases with yield strength, stress, and operating temperature (Ref 8, 19, 20, 21, 22, 23)

There are pronounced thermodynamic effects that cause the concentration of impurities at surfaces and a turbine component failure They are most frequent where the metal surface temperature is slightly above the saturation

temperature of steam For the low-pressure fossil utility turbines, this usually occurs on the L - 1 blades and on various surfaces of the last two disks It is one of the reasons the L - 1 blades in fossil utility turbines have higher failure rates than any other blade row Another reason for the high L - 1 blade failure rate could be the effects of transonic flow, vibratory stresses, and changing temperatures resulting from an interaction of the shock wave with the Wilson line (periodic destruction of the Wilson line) (Ref 30)

Erosion-corrosion of carbon steel wet-steam piping is of primary concern in industrial and nuclear turbine units It is

most pronounced in carbon steel pipes with high-velocity turbulent flow and low-pH moisture containing high concentrations of CO2 or other acid-forming anions (Ref 31, 32, 33, 34, 35, 36)

Corrosion of other turbine parts is due to one or more of the same causes described above (Table 3) There is a strong SCC low-cycle fatigue interaction in many turbine materials (well recognized in piping) that may be important in stationary blade, bolt, expansion bellows, and turbine cylinder failures The highest frequency of turbine bolting failures has been in high- and intermediate-pressure cylinders and nozzle blocks Stress-corrosion cracking of expansion bellows (intermediate- and low-pressure pipes) used to be a problem in fossil units with high concentrations of NaOH (Ref 37)

To find the true causes of corrosion, it is essential to analyze the local temperature, pressure, chemistry, moisture velocity, and stress and material conditions

Turbine Materials

There is little worldwide variation in materials for blades, disks, rotors, and turbine cylinders, and only a few major changes have been introduced in the last decade Titanium alloy blades are slowly being introduced for the last low-pressure stages Also, improved melting practices, control of inclusions and trace elements, and weld repair techniques are being used for disks and rotors In addition to the materials listed in Table 3, carbon steel is used for low-pressure casings (shells and cylinders), chromium-molybdenum steels and austenitic stainless steels for high-pressure piping, and miscellaneous other materials for other components

Considering the typical design life of 25 to 40 years and the relatively high stresses, turbine materials perform remarkably well Turbine steels are susceptible to SCC and corrosion fatigue in numerous environments, such as caustic, chlorides, hydrogen, carbonate-bicarbonate, carbonate-CO2, acids, and, at higher stresses and strength levels, pure water and steam

No turbine can permanently tolerate concentrated caustic, sodium chloride (NaCl), or acids Under such conditions, disk and rotor materials would crack by SCC in a few hundred hours at stresses as low as 10% of the yield strength Chromium steel (12% Cr) blades may tolerate NaOH, but would pit and crack by corrosion fatigue at very low vibratory stresses (as low as 6.9 MPa, or 1 ksi) in many other corrodents With high concentrations of impurities for long periods of time, there would always be a weak link in the turbine or within the power cycle that would fail prematurely Although titanium could be used for blades and would tolerate most of the impurities (except hydroxides), corrosion-resistant materials for the large forgings of rotors and disks are difficult to find because of the cost, forgeability, banding of alloying elements and impurities, and other problems Plating and surface coatings appear to be only marginal temporary solutions (Ref 38)

In addition to the susceptibility to localized corrosion in concentrated impurities, the low-alloy and carbon steels used extensively in steam turbines have rather narrow ranges of passivity (Ref 39) This makes these steels vulnerable to pH and oxygen excursions and consequent pitting and other forms of localized corrosion during operation and layup In chloride solutions, the pH region for passivation becomes even narrower The material properties needed for designing

against corrosion, failure analysis, and evaluation of residual life include fracture toughness (KIc), stress-corrosion threshold stress ( SCC), threshold stress intensity (KISCC), crack propagation rate ( SCC), corrosion fatigue limit, corrosion fatigue threshold stress intensity ( Kth), corrosion fatigue crack propagation rate (da/dN), pitting rate, and a pit depth

limit

Environment

Corrosive impurities are transported into the turbine steam from the preboiler cycle (in the feedwater) and the boiler as a total (mechanical plus vaporous) carryover (Ref 6, 8, 9) Their major sources include condenser leaks, air in-leakage, makeup water, and improperly operated condensate polishers The turbine environment is controlled through a control of impurity ingress and various feedwater, boiler water, and steam chemistry limits (Ref 6, 8, 40, 41, 42, 43, 44, 45) The corrosiveness of the steam turbine environment is caused by one or more of the following:

• Concentration of impurities from low part per billion levels in steam to percent levels on surfaces and the formation of concentrated aqueous solutions (concentration by deposition or evaporation of moisture)

• Insufficient pH control (in both acid and alkaline regions)

• High-velocity, high-turbulence, and low-pH moisture

The situation is illustrated in Fig 15, which shows a Mollier diagram with a typical turbine steam expansion line, the thermodynamic regions of impurity concentrations (NaOH and salts), and the resulting corrosion The conditions on hot turbine surfaces (in relation to the steam saturation temperature) can shift from the wet-steam region into the salt zone and above This is why SCC of disks often occurs in the wet-steam regions (Ref 19, 23, 29), and it emphasizes the need to consider local surface temperatures in design and failure analysis The surfaces may be hot because of heat transfer through the metal (Ref 9, 22, 29) or because of the stagnation temperature effect (zero flow velocity at the surface and conversion of kinetic energy of steam into heat) (Ref 46) Depending upon their vapor pressures, impurities can be present as a dry salt or as an aqueous solution In the wet-steam region, they are either diluted by moisture or could concentrate by evaporation on hot surfaces

Fig 15 Mollier diagram showing the low-pressure turbine steam expansion line, thermodynamic regions of

impurity concentration (with impurities given in parentheses), and corrosion mechanisms Source Ref 6 and 9)

The steam impurities that are of most concern include chlorides, sulfates, fluorides, carbonates, hydroxides, organic and inorganic acids, oxygen, and CO2 Their behavior in turbine steam and deposits is well documented (Ref 6, 9, 14, 30, 46,

47, 48, 49, 50, 51, 52, 53), there are strong synergistic effects and interactions with metal oxides

In addition to corrosion during operation, turbines can corrode during:

• Manufacture (machining fluids and lubricants)

• Storage (airborne impurities and preservatives)

• Erection (airborne impurities, preservatives, and cleaning fluids)

• Chemical cleaning (storage of acid in condenser hotwells)

• Nondestructive testing (cleaning and testing fluids)

• Layup (deposits plus wet air)

Many of the above substances may contain high concentrations of sulfur and chlorine, which could form acids upon decomposition Decomposition of typical organics for example, carbon tetrachloride (CCl4) occurs at about 150 °C (300

°F) Therefore, the composition of all of the above substances should be controlled (maximum of 50 to 100 ppm S and 50

to 100 Cl has been recommended), and most of them should be removed before operation Molybdenum disulfide (MoS2)

is being increasingly implicated as a corrodent in power system applications (Ref 54, 55, 56, 57, 58)

Layup corrosion increases rapidly when the relative humidity of the air reaches about 60% When deposits are present, layup corrosion can be in the form of pitting or SCC

Design

Because of the long design life, steam turbines undergo a limited prototype testing in which the long-term effects of material degradation, such as corrosion, creep, and low-cycle fatigue, cannot be accurately simulated When development was slow, relatively long-term experience was transferred into new products Because of the rapid development of new turbine types, larger sizes, and new power cycles and water treatment practices during the last 25 years, the experience was short and limited, and some problems were developed that still need to be corrected and considered in new designs and redesigns Design disciplines that affect turbine corrosion can be separated into four parts:

• Mechanical design (stresses, stress concentrations, and stress intensity, K1)

• Heat transfer (surface temperatures and heated crevices)

• Flow (moisture velocity, location of the salt zone, stagnation temperature, and interaction of shock wave with Wilson line)

• Physical shape (crevices, obstacles to flow, and surface finish)

The various aspects of design against environment-sensitive fracture are well documented (Ref 9, 12, 13, 19, 20, 21, 24,

29, 30, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68), but only a few of these references deal with the complexity of the problem (Ref 9, 13, 14, 20, 60, 65, 66, 67, 68) Probabilistic approaches are discussed in Ref 67 and 69 Problems with such approaches include the lack of a large number of statistical data on corrosion properties of materials, service stresses, and environments The mechanical design concepts for avoiding turbine corrosion should include evaluation of safety factors against SCC, KISCC, SCC, (da/dN), Kth, corrosion fatigue limit, pitting rate, and a pit depth limit

A more conservative approach should be taken for high multiaxial tensile stresses, strain (mechanical or thermal) fluctuations, other mechanical interactions, and large component sizes True residual stresses, both microscopic and macroscopic, should also be considered Implementation of the above concepts requires that corrosion testing generate quantitative data that are useful to the designer

Hot Corrosion in Coal- and Oil-Fired Boilers

Ian G Wright, Battele Columbus Division

Corrosion from the firing of coal or oil is essentially related to specific impurities in the fuels, which can lead to the formation of nonprotective scales or can disrupt normally protective oxide scales The relevant impurities in coal are sulfur (about 0.5 to 5.2% in United States coals), sodium (0.01 to 0.7%), potassium (0.2 to 0.7%), and chlorine (0.01 to 0.28%) In oil, the important impurities are sodium, which from United States refineries may range up to 300 ppm by weight; vanadium (up to 150 ppm by weight); and sulphur (0.6 to 3.6%) During combustion, these impurities can be melted or vaporized and will deposit upon contact with surfaces at temperatures lower than the condensation temperatures

of the specific species This provides a mechanism for the accumulation of deposits of fly ash on cooled surfaces downstream of the burners (Ref 70, 71)

Water wall fire-side corrosion is generally found in regions around the burners The thick, hard external scales formed may be quite smooth, but often exhibit cracks or grooves that can resemble an alligator hide Cracking or grooving is usually circumferential, and it is more common in supercritical than subcritical boilers It occurs in the areas of the water walls that receive the highest heat flux and is apparently a result of superimposed thermal stress Figure 16 illustrates the appearance of grooving Figure 17 shows the grooves to be sharp-pointed cracks filled with corrosion product (mostly

iron oxide) but with a central core of iron sulfide Figure 18 shows a cross section of an ASME SA-213, grade T-11 water wall tube that has formed a thick, smooth scale The outer scale is a mixture of iron sulfide and iron oxide, with the chromium from the alloy dipersed in the inner layer The major causes of corrosion of the water wall tubes are, first, the reducing (substoichiometric) conditions caused by impingement of incompletely combusted coal particles and flames and, second, molten salt or slag-related attack

Fig 16 Example of grooving in water wall tubes Note the thick, adherent scale remaining in some areas

Fig 17 Cross section of circumferential cracks on the fire-side surface of a water wall tube Source: Ref 72

Fig 18 Cross section of the fire-side face of water wall tube subjected to reducing (substoichiometric)

combustion conditions From left to right: optical micrograph of deposit and x-ray elemental dot maps for iron, chromium, and sulfur, respectively

Gas Phase Corrosion

Reducing atmosphere corrosion may be a result of direct reaction of the water wall tubes with a substoichiometric gaseous environment containing sulfur or with deposited, partially combusted char containing iron pyrites, which gives rise to a very localized corrosive environment Reducing conditions have two main effects on corrosion First, they tend to lower the melting temperature of any deposited slag, which increases its ability to dissolve the normal oxide scales on the tubes, and second, the stable gaseous sulfur compounds under these conditions include H2S (Ref 73, 74, 75, 76), which is considerably more corrosive than the SO2 that predominates under oxidizing conditions Under reducing conditions, iron sulfide is the expected corrosion product on iron, rather than the oxides (Fe3O4 and Fe2O3) Sulfide scales allow significantly higher rates of transport of iron cations than oxides do and so are less protective However, Cr2O3, which is a very protective oxide, would be expected to form on chromium under reducing conditions; therefore, alloys capable of forming Cr2O3 scales, such as AISI type 310 stainless steel, Inconel 671, or chromized coatings, would be expected to exhibit good corrosion resistance in this application

Molten Salt Corrosion

Slag-related attack takes several forms Local disruption of the normal oxide film on the wall tubes by intrusion of molten slag can lead to accelerated oxidation or (if sulfur species are present in the slag) to oxidation-sulfidation In coal-fired boilers, alkali sulfates deposited on the water walls may react with SO2 or SO3 to form pyrosulfates, such as potassium pyrosulfate (K2S2O7) and sodium pyrosulfate (Na2S2O7), or possibly complex alkali-iron trisulfates The latter compounds are formed in thicker deposits after long periods of time at about 480 °C (900 °F) (Ref 77) The (K2SO4-K2S2O7) system forms a molten salt mixture at 407 °C (764 °F) when the SO3 concentration is 150 ppm The corresponding sodium system can become liquid at 400 °C (750 °F), but it requires about 2500 ppm SO3 for this to occur; such levels of sulfur oxides are likely only under deposits (Ref 70) Thus, molten salt attack on the tube metal by K2S2O7 is more likely and occurs according to the reaction

K2S2O7 + 3Fe FeS + Fe2O3 + K2SO4 (Eq 2)

By such a mechanism, K2S2O7 can react aggressively with any protective iron oxide scales on the tubes and lead to accelerated wastage through fluxing of the oxides and attack of the substrate metal Differential scanning calorimetry of samples of deposits taken from water wall tubes typically indicates melting points in the range of 335 to 410 °C (635 to

770 °F) (Ref 78)

The role of chlorine in fire-side corrosion is difficult to define However, it seems likely that under reducing conditions HCl or NaCl can render the oxide scales less protective by causing them to blister or crack or by reactions with the oxides

or base metal to form volatile products (Ref 71)

Molten salt related corrosion of water walls is seldom encountered in oil-fired boilers This is presumably related to the absence of chlorine in the oil and to the lower ash content of oil compared to coal (about 0.5% in oil, and up to 35% in U.S coals); therefore, the deposits formed on the water walls of oil-fired boilers are very thin and do not provide the necessary conditions for the formation of pyrosulfates In addition, the tube metal temperature in the water wall region of

a furnace is usually below 400 °C (750 °F), which is lower than the melting points of any compounds possible between sodium and vanadium oxides Therefore, any compounds formed are unlikely to melt and pose a corrosion threat under normal operating conditions

Prevention of Corrosion

Solutions to fire-oxide corrosion of furnace water walls are available from changes in operating procedures and changes

in tube materials Where the corrosion results from the presence of reducing conditions near the water walls, operational actions include adjusting the air and fuel distribution to individual burners and among burners in order to promote better mixing and more uniform combustion conditions, as well as resetting to design specification the coal fineness delivered to the burners from the milling plant Flame impingement can be rectified by changing the characteristics of the offending burners through adjustment of secondary air registers to control air flow and degree of swirl The most accurate way of making such burner adjustments is by monitoring the composition of the gas close to the affected water walls to ensure that oxidizing conditions (or acceptably low carbon monoxide, CO, levels: <1%) are achieved

Another method of countering reducing conditions near the water walls is to introduce a flow of air along the walls through opening in the membrane between water walls tubes This is often referred to as air blanketing or curtain air, and

it should also be implemented and adjusted in conjunction with local monitoring of the gas composition

Furnace wall corrosion can also be lessened by reducing the levels of the chemical species in the coal that are responsible for corrosion Approximately one-half the sulfur and alkali metal content of coal can be removed by standard coal-washing procedures However, washing generally does not remove the chlorine-containing species from coal; therefore, the net effect is that the chlorine content of washed coal is increased An alternative strategy is to blend the coal to reduce the average content of corrosive species Both of these strategies may involve some increase in fuel costs, which must be weighed against the increased tube lifetimes The same approaches for control of reducing environments near the water walls described for coal-fired boilers can be applied to oil-fired boilers

Materials solutions to water wall fire-side corrosion problems involve either direct replacement with tubes of a more corrosion-resistant alloy or the application of a corrosion-resistant alloy as a cladding on the affected tube Replacement

of the tubes with the same material and the use of regular wall thickness monitoring where wastage rates are only slightly greater than allowable have been recommended to give the desired tube life (Ref 79) For more severe wastage, little difference in performance has been found for alloys containing less than 9% Cr In this case, the materials choice is probably between thicker-walled carbon steel tubes and tubes with a coextruded outer layer of a high-chromium alloy, such as AISI type 310 stainless steel or 50Cr-50Ni The technique of cladding a tube fabricated from the alloy typically chosen for the application (based on strength considerations) with an outer layer of a corrosion resistant alloy by coextrusion can provide a cost-effective solution to this problem The Central Electricity Generating Board has found coextruded tubes of type 310 on low-carbon steel to be economically viable for base-loaded units and has used such tubes

in boilers since 1974 (Ref 80)

Increased corrosion resistance can also be gained by enriching the surfaces to be protected with such elements as chromium or aluminum This can be achieved by diffusion treatments or by the spraying of metal overlay coatings Although pack aluminizing can be applied commercially to tubes up to about 6 m (20 ft) long (Ref 81), there are as yet no reports of utility boiler experience with such materials Flame or plasma spraying has been used to apply high-chromium, high-aluminum, iron-chromium-aluminum alloy compositions to water wall tubes Although such alloys have shown good corrosion resistance in laboratory tests, problems have been experienced in practice as a result of a lack of reproducibility of coating application techniques

Hot Corrosion in Boilers Burning Municipal Solid Waste

H.H Krause, Battelle Columbus Division

The corrosion problems experienced in boilers fueled with municipal refuse are different from those encountered with fossil fuels in that chlorine rather than sulfur is primarily responsibly for the attack The average chlorine content of municipal solid waste is 0.5%, of which about one-half is present as polyvinylchloride (PVC) plastic The other half if inorganic, principally NaCl The chlorine in the plastic is converted to hydrochloric acid (HCl) in the combustion process The inorganic chlorides are vaporized in the flame and ultimately condense in the boiler deposits or pass through the boiler with the flue gases Zinc, lead, and tin in the refuse also play a role in the corrosion process by reacting with the HCl to form metal chlorides and/or eutectic mixtures with melting points low enough to cause molten salt attack at wall tube metal temperatures

Gas Phase Corrosion

Hydrochloric acid alone has little effect on carbon steel at temperatures below 260 °C (500 °F) (Ref 82) However, in the presence of excess air in the boiler, the reaction to form ferrous chloride (FeCl2) on the steel surface occurs more readily This corrosion product is stable at water wall tube temperatures, and the curve for metal wastage versus time is parabolic However, at a metal temperature of about 400 °C (750 °F), which may occur in the superheater, the FeCl2 is further chlorinated to the readily volatile ferric chloride (FeCl3) (Ref 83) If the gas temperature in the area exceeds 815 °C (1500

°F), the FeCl3 will evaporate rapidly, and breakaway corrosion can then occur These effects are shown in Fig 19, which presents data from corrosion probe exposures conducted in a municipal incinerator

Fig 19 Corrosion rates at two temperatures of carbon steel in a boiler burning municipal refuse Source: Ref

84

The corrosion probe results also demonstrated that FeCl2 was formed as a corrosion product at temperatures of 150 to 260

°C (300 to 500 °F); in this range, HCl is not particularly corrosive (Ref 85) This result can be explained only by attack from elemental chlorine rather than HCl In one study, corrosion of carbon steel by chlorine reached the breakaway condition at 205 °C (400 °F) (Ref 82); therefore, this agent must be responsible for any low temperature attack

A reducing atmosphere can also contribute to wall tube corrosion This effect was first observed in European incinerators,

in which excessive corrosion of side walls and furnace corners was found to be associated with reducing conditions in

addition to chloride deposits A recent study investigated the combined effect of CO and HCl on corrosion (Ref 86) Corrosion rates were linear with time in a simulated flue gas containing 400 ppm HCl, 10% CO, 10% H2O, 0.5% SO2, and the balance nitrogen At 400 °C (750 °F), the slope of the metal loss versus time curve was 1.14 mm/yr (45 mils/yr), without HCl in the gas mixture, a parabolic rate curve was obtained that leveled off at a maximum metal loss of only 127

m (5 mils) The synergistic action of HCl and CO in promoting corrosion was attributed to disruption of the oxide layer

on the metal surface, followed by spalling of the oxide, thus exposing the metal to further oxidation

Molten Salt Corrosion

Incinerator boiler deposits usually contain several percent of zinc and lead, with a lesser amount of tin In combination with chlorine, these elements can form low-melting compounds or eutectic mixtures that would be molted at tube metal tmeperatures Thus, for example, stannous chloride (SnCl2) melts at 246 °C (475 °F) and zinc chloride (ZnCl2) at 283 °C (541 °F) The eutectic mixture of 50.2% lead chloride (PbCl2) and 49.8% FeCl3 melts at 175 °C (347 °F) Examination of the layer of scale at the corroded metal surface of boiler tubes an probe specimens by the electron microprobe and by energy-dispersive x-ray analysis has shown these metals to be associated with chlorine Consequently, molten salt attack

by these metal chlorides can occur and will be more severe than gas phase corrosion

A recent investigation of an incinerator wall tube that was corroding at a rate of 2 mm/yr (80 mils/yr) showed that zinc and sodium were both associated with chlorine in the deposit (Ref 87) The presence of NaCl was confirmed by x-ray diffraction However, the high corrosion rate could not be accounted for in terms of attack by NaCl or HCl Consequently, laboratory tests were conducted to demonstrate that the corrosion could be caused by the eutectic mixture of 84% ZnCl2and 16% NaCl, which has a melting point of 262 °C (504 °F) After a 336-h exposure to this mixture at 315 °C (600 °F), carbon steel had a corrosion rate of 23 mm/yr (910 mils/yr), indicating that such molten salt attack was the likely mechanism in the incinerator There is as yet no evidence for participation of SnCl2 in the incinerator corrosion reactions However, its low melting point and the possibility of forming a eutectic mixture with NaCl that melts at 199 °C (390 °F) make it a likely contributor to molten salt corrosion

Prevention of Corrosion

Most of the methods of preventing incinerator wall tube corrosion exact some penalty in boiler efficiency The practice of studding the tubes and covering them with silicon carbide refractory has been widely used in European incinerators, but this remedy reduces heat transfer Increasing overfire air or blanketing the walls with air to prevent reducing conditions in the flue gas has been effective, but either approach will reduce boiler efficiency Lowering tube metal temperatures by operating at lower steam pressure also has a cost in efficiency

However, upgrading the boiler tube material to a corrosion-resistant alloy does not involve an efficiency penalty Although capital costs will be greater, the extended tube life resulting from the use of more resistant alloys can offset the initial expense and can be a cost-effective solution to the problem Extensive corrosion prove studies in municipal incinerator showed that in the temperature range of 150 to 315 °C (300 to 600 °F) a number of alloys provided good performance in resisting high-temperature corrosion (Ref 88) In decreasing order, the better alloys, were Incoloy 825; AISI types 446, 310, 316L, 304, and 321 stainless steels; and Inconel alloys 600 to 601 However, when subjected to moist deposits, simulating boiler downtime conditions, all of the austenitic stainless steels underwent chloride SCC The type 446 stainless steel Inconel 600 and Inconel 601 suffered pitting Consequently, uniess the boilers were to be maintained at a temperature above the HCl dew point during downtime, only Incoloy 825 was recommended

The laboratory tests discussed in Ref 87 showed that Inconel alloys 600, 625, and 690, Incoloy 800, and type 304 and 309 stainless steels were resistant to the ZnCl2-NaCl mixture at 315 °C (600 °F) In addition, an overlay of Inconel 625 applied to the carbon steel tubes in the incinerator offers promise Recent field tests conducted at a municipal waste incineration boiler indicated that alloy 556 (iron-nickel-cobalt-chromium) has excellent resistance in such applications (Ref 89) The test rack was exposed to flue gases at 800 °C (1475 °F) for 950 h Maximum depths of attack, that is, metal loss and maximum internal penetration, were 0.21 mm (8.2 mils) for alloy 556, 0.34 mm (13.4 mils) for AISI type 309 stainless steel, 0.69 mm (27 mils) for Hastelloy X, and more than 1.1 mm (45 mils) for Inconel 690 The Inconel 690 coupon was perforated Sulfidation was reported to be the major made of attack Consequently, Inconel 625, Incoloy 825, alloy 556, and possibly Incoloy 800 are the preferred materials solutions to wall tube corrosion in municipal incinerators

Corrosion of Superheaters and High-Temperature Air Heaters

John Stringer, Electric Power Research Institute

Corrosion of superheaters and air heaters is inevitable in coal-fired boilers and in fluidized-bed combustors This section will discuss factors that cause corrosion as well as measures that can be taken to minimize corrosion in these applications

Superheater Corrosion in Coal-Fired Boilers

The fire-side corrosion of superheaters in pulverized coal-fired utility boilers is one of the principal problems that have limited main steam temperatures to 540 °C (1000 °F) for the last 30 years Allowing for the temperature decrease through the tube walls and the typical non-uniformities in the temperature distribution across a superheater bank, this corresponds

to an outer maximum metal surface temperature of approximately 620 to 650 °C (1150 to 1200 °F) The metals and alloys used for superheater tubing are primarily determined by the ASME Boiler Code, which is concerned only with the strength of the materials

For typical steam pressures and tube wall thicknesses, Table 4 shows the maximum-use temperature for several boiler alloys for which code standards exist These are the wall midsection temperatures, which are typically 25 °C (45 °F) lower than the outer surface temperature A further limitation is imposed by the oxidation resistance of the steels The ASME code does not mention oxidation, and some allowance can of course be made by increasing the tube wall thickness However, for typical conditions, Table 4 also lists the maximum metal surface temperature that can be tolerated from an oxidation wastage point of view

Table 4 Temperature limits of superheater tube materials covered in ASME Boiler Codes

Maximum-use temperature Oxidation graphitization criteria, metal surface (a)

Strength criteria, metal midsection Material

°C °F °C °F SA-106 carbon steel 400-500 750-930 425 795

Ferritic alloy steels

(a) In the fired section, tube surface temperatures are typically 20 to 30 °C (35 to 55 °F)

higher than the tube midwall temperature In a typical U.S utility boiler, the maximum metal surface temperature is approximately 625 °C (1155 °F)

The fire-side corrosion discussed in this section is the process by which the wastage rate is increased above these design rates The enhanced corrosion is generally associated with the formation of a deposit on the tube walls Although the presence of a deposit does not necessarily indicate accelerated corrosion, accelerated corrosion is never experienced in the absence of such a deposit

Nature of the Deposit Figure 20 shows an illustration of the deposit formed on superheater tubes The deposit may

be several inches thick, with the outer layers formed of loosely sintered fly ash The outer part of this deposit will be essentially at the gas temperature The white inner layer is rich in alkali sulfate The black layer is largely Fe3SO4 Sulfur prints indicate that sulfide is present at the metal surface

Fig 20 Deposit layers on a corroding superheater or reheater tube

It has been shown that deposition of alkali sulfate, as well as corrosion, is often a more or less linear function of the chlorine content of the coal It is now generally believed that chlorine plays no direct role in the corrosion process; instead, it acts as a release agent for the alkali metals in the coal

It is generally believed that the salt deposit must be molten for accelerated corrosion to occur The melting point of sodium sulfate (Na2SO4) is 884 °C (1623 °F); that of potassium sulfate (K2SO4) is 1069 °C (1956 °F) The maximum melting point of a mixture of the two is 823 °C (1513 °F) Because the molten phase must be close to the metal, there is

no possibility that temperatures as high as this could be attained In the deposit on tubes cooled to room temperature, a complex sulfate having the general formula (Na,K)3Fe(SO4)3 has often been identified This phase has a minimum melting point of 554 °C (1029 °F) at a sodium:potassium ratio of 2:3 However, it is relatively unstable and at 540 °C (1000 °F) will decompose to the solid constituent oxides unless the local SO3 partial pressure exceeds 25 Pa (0.004 psi) This is approximately five times the normal SO3 partial pressure within a boiler, but it has been shown that the local partial pressure may be significantly higher near the corroding surface

Corrosion beneath the deposit is at a maximum toward the edges of the deposit at the 5 and 7 o'clock positions The

metal temperature will be highest under the deposit at these locations because beneath the thick part of the deposit the metal surface will be insulated from the hot gas Generally, corrosion consists of pits with relatively smooth metal/scale interfaces There is little evidence of internal penetration of internal sulfidation The pits eventually overlap to produce an apparently general corrosion

It is difficult to obtain much information on the influence of different factors on the corrosion from in-service data because the corrosion is very specific to a particular boiler burning a particular fuel under a definite set of conditions In addition, different materials are usually not exposed simultaneously On the basis of laboratory tests, it appears that the accelerated attack is present only for metal temperatures in the range of 560 to 700 °C (1040 to 1290 °F), with a maximum at approximately 670 °C (1240 °F) It is believed that the minimum temperature corresponds to the solidus temperature of the salt deposit and that the maximum is the dissociation temperature of the low-melting complexes Figure 21 shows the relative rates of attack for a range of typical boiler tube materials, including some coating and cladding alloys

Fig 21 Corrosion rates of alloys in a laboratory test using synthetic ash (37.5 mol% Na2SO4, 37.5 mol% K2SO4and 25 mol% Fe2O3) in a synthetic flue gas (80% nitrogen, 15% CO2, 4% oxygen, and 1% SO2 saturated with water) Exposure time: 50 h

For a given metal temperature, the corrosion rate increases with gas temperature Corrosion rate increases by a factor of three at a metal temperature of 650 °C (1200 °F) as the gas temperature rises from 800 to 1400 °C (1470 to 2550 °F)

Corrosion Prevention Empirical procedures are available for predicting the probable corrosivity of a given coal A

known corrosive coal may be blended with another to produce a less corrosive ash, and it may also be possible to use additives, such as limestone If it is known at the design stage that a corrosive coal must be burned, the superheater can be positioned further back along the gas path where the gas temperature is lower Such positioning also reduces the amount

of corrosive material reaching the superheater, because some corrodent is deposited on upstream tubing

If control of coal chemistry or design changes are not feasible, corrosion-resistant materials must be used For a boiler that has exhibited corrosion, it may be sufficient to wrap strips of AISI type 310 stainless steel around the leading tubes in a platen; this is termed bandaging If bandaging is insufficient, it may be necessary to replace the finishing superheater tubing with a coextruded material, in which the inner layer is a strong coded alloy and the outer layer is a corrosion-resistant material, such as type 310 stainless steel or Incoloy 671 (Ni-50%Cr) Over the years, there have been many efforts to develop coating systems that could be applied in the field by such techniques as plasma spraying However, the results have generally been disappointing, although for a particular situation this may still be a cost-effective approach

Corrosion of Superheaters and High-Temperature Air Heaters in Fluidized-Bed Combustors

A recent development as an alternative to the pulverized coal-fired boiler is the fluidized-bed combustor In this unit, the coal is burned in a fluidized bed of particles; very good combustion efficiency is achieved at relatively low temperatures

(typically 850 to 900 °C, or 1560 to 1650 °F) This results in low NOx emissions In addition, sulfur can be captured in the bed by adding an absorbant, such as CaO

Because of the low combustion temperature, it was anticipated that there would be no ash fusion, little or no ash deposition on the in-bed heat-exchanger surface, and limited alkali release It was therefore assumed that in-bed fire-side corrosion of superheaters by a mechanism similar to that described in the preceding section would be unlikely This appears to be true However, a different mechanism of corrosion does present a potential problem; this will be described

in the section "Corrosion Mechanism" in this article In addition, the in-bed environment is potentially erosive, and the combination of erosion and corrosion can result in accelerated wastage

The In-Bed Environment The bed consists of particles with an average size typically close to 0.8 mm (30 mils) The

major constituent is CaSO4, and there are relatively minor amounts of coal ash, still smaller amounts of unreacted CaO, and typically less than 2% C At metal temperatures above approximately 500 °C (930 °F), reddish-brown deposits form that consist mostly of CaSO4, with about 15% coal as related material The deposit is extremely dense; its porosity is below the limits of detection of most conventional techniques It is very well bonded to the oxide on the metal, and when

it detaches, for example, upon thermal cycling, it generally takes the metal oxide with it

The oxygen potential in the bed varies from approximately 10-1 to 10-14 atm The oxygen potential fluctuates relatively rapidly (a frequency of the order of 0.5 to 2 Hz) In some regions, the oxygen potential is generally low, with peaks to high values; in other regions of the bed, the potential is generally high, with peaks to low values If local equilibrium were established, this low oxygen potential, combined with the CaO/CaSO4 equilibrium, would generate a sulfur partial pressure of the order to 10-5 atm

The Corrosion Mechanism Materials within the bed can suffer a form of corrosion in which the normally protective

oxide (usually Cr2O3) is disrupted; this allows the rapid growth of the less protective oxides of the base metal Within the metal itself, chromium-rich sulfides are present In extreme cases, with nickel-base alloys at temperatures above about

650 °C (1200 °F), liquid nickel sulfides may appear, and the degradation of the alloy becomes catastrophic It has been shown that corrosion is more severe in regions of the bed where the oxygen potential is generally low, with peaks to higher values; regions near the coal feed ports are particularly corrosive

Materials for In-Bed Superheaters and Air Heaters The conditions for in-bed superheaters are similar to those

in a conventional boiler in terms of the internal temperature and pressure and the lifetime generally required for commercial success The maximum metal temperature is of the order of 650 °C (1200 °F), and the available materials are those defined by the ASME Boiler Code The conditions for an air heater generating hot air to be expanded through a gas turbine are somewhat different The internal pressures are significantly lower than those in utility boiler tubes (of the order of 1 to 1.5 MPa, or 145 to 218 psi, as opposed to 17 to 25 MPa, or 2465 to 3625 psi), but the temperatures are considerably higher The maximum metal temperature is essentially equal to the bed temperature

Nickel-base alloys are generally very sensitive to this form of corrosion, and their use should be avoided within fluidized beds unless low local oxygen activities can absolutely be eliminated Stainless steels, such as type 304 and type 347, are relatively resistant Type 310 is very resistant to the corrosion, but its mechanical properties at elevated temperatures make it unsuitable for boiler tubes, although it could be used as a cladding Incoloy 800 has a higher nickel content and good high-temperature properties Its behavior in the bed is intermediate; that is, in locations where the oxygen activity is always high, it has excellent corrosion resistance, but it can suffer very rapid attack in regions of low oxygen activity

Therefore, the materials of choice for the superheater are the austenitic stainless steels The principal question is whether these steels may suffer breakdown corrosion in several thousand hours in this type of environment Long-term investigations of this question are in progress or have recently been completed If it is demonstrated that breakdown does take place or if the simultaneous action of erosion leads to accelerated attack, coextruded tubes may be necessary

Selection of materials for air heaters is considerably more difficult because many of the alloys with adequate strength at the higher temperatures have high nickel contents In some cases, the corrosion-sensitive alloy Incoloy 800H has been selected, with efforts to locate the hotter tubes in regions of the bed with high oxygen activities This presents a long-term risk because it is difficult to be sure that the conditions within the bed will remain constant, and fluctuations in the distribution of coal of air in the bed can profoundly affect local oxygen activities Again, coextruded tubes may be necessary

Corrosion of Combustion Turbines

R Viswanathan, Electric Power Research Institute

Combustion turbine components are susceptible to aqueous corrosion and to oxidation-sulfidation corrosion occurring at elevated temperatures Aqueous corrosion problems are encountered in the compressor section, but oxidation and sulfidation problems are encountered by blades and vanes in the hot-gas path of the turbine section

Corrosion of Compressor Blades and Disks

Compressor components are generally made of ferritic stainless steels or low-alloy steels At many utilities, the humidity

in the atmosphere is sufficiently high that water condenses on these components during operation In addition, water often collects on the compressor components of peaking units that are idle; some utilities heat their turbines when they are not used to avoid this problem General corrosion and pitting can occur as a result of the accumulation of water on compressor blades and disks The corrosion product fouling of compressor blades associated with corrosion reduces the efficiency of the compressor and therefore increases the heat rate of the turbine If corrosion is extensive, parts may need

to be replaced Extensive corrosion is often encountered in the low-pressure compressor section of turbines, in which the disks are made of low-alloy steel Corrosion problems are compounded when aggressive environments are present, such

as salt air or air containing contaminants from other units at the plant

Periodic cleaning of the compressor has been known to restore the efficiency lost because of corrosion Utilities use many methods of cleaning, including nutshelling (abrasive blasting using crushed nutshells) as well as washing with water, water plus detergent, or water plus organic solvents Cleaning often cannot be performed without interrupting the operation of the turbine; this presents problems for the operators of turbines used in intermediate- or baseload service Nutshelling may also lead to clogging of the fuel nozzle and does not provide any cleaning to the turbine section

Many coatings have been used with mixed results in an effort to reduce the extent of corrosion in compressors The most disappointing has been the nickel-cadmium family of coatings These coatings are subject to pitting in acidic environments They are cathodic to the blades and therefore can accelerate the rate of localized blade corrosion at coating defects In addition to pitting, bare spots in the coating are often present where the electrodes were placed during electrodeposition of the coating, and this further accelerates corrosion In some turbines, nickel-cadmium coatings on the blades of the latter stages have been lost, reportedly by vaporization of the cadmium

More recently, sacrificial coatings (that is, anodic to the blades) have been used in the compressor section These aluminum-base coatings have met with moderate success; many utilities have reported reduced corrosion In some cases, however, the rate of corrosion (or at least the rate of fouling) was essentially unchanged, presumably because of the porosity of the coatings This problem is reportedly being addressed by some coating suppliers through use of a subsequent sealer overcoating In the cases in which coated blades become fouled, the coatings still have been found to be beneficial because the coating facilities cleaning of the blades In these cases, the recovery of efficiency by cleaning was greater for coated blades than that for uncoated blades

Hot Corrosion of Turbine Blades and Vanes

Blades and vanes located in the hot-gas path in the turbine section are subject to a combined oxidation-sulfidation phenomenon that is commonly referred to as hot corrosion These components are generally made from nickel- or cobalt-base superalloys Three basic types of hot corrosion attack have been recognized (Fig 22) In the temperature range of

650 to 705 °C (1200 to 1300 °F), layer-type corrosion characterized by an uneven scale/metal interface and the absence of subscale sulfides is observed At temperatures above 760 °C (1400 °F), nonlayer-type corrosion (type I) is observed Type

I corrosion is characterized by a smooth scale/metal interface and a continuous, uniform precipitate-depleted zone containing discrete sulfide particles beneath the scale The transition from one type to the other, which occurs in the range

of 705 to 760 °C (1300 to 1400 °F), is characterized by an uneven scale/metal interface containing intermittent pockets of subscale precipitate- depleted zones and sulfides The layer-type and the transitional corrosion to gather are variously referred to as type II hot corrosion, low temperature hot corrosion, and low-power corrosion

Fig 22 Three forms of hot corrosion in Udimet 710 turbine blades (a) Layer type (b) Transition type (c)

Nonlayer type

Figure 23 shows the essential features of the corrosion products associated with the three types of corrosion In the nonlayer-type high-temperature form of hot corrosion (Fig 23c), discrete chromium and titanium sulfide particles are present in a region of the matrix depleted in these elements, adjacent to the base metal The surface scales consist of protective Cr2O3 with some titanium oxides With decreasing temperature, the chromium and titanium sulfides are increasingly agglomerated into large interconnecting sulfide networks, and the surface scales contain predominantly the oxides of nickel and cobalt (Fig 23b) Complete layer-type corrosion (Fig 23a) is characterized by the chromium and titanium sulfides forming a continuous layer The surface scale in this case contains only the unprotective oxides of nickel and cobalt

Fig 23 Schematic of corrosion products formed in the three types of hot corrosion of turbine blades (a) Layer

type (b) Transition type (c) Nonlayer type

In addition to the corrosion features described above, grain-boundary spikes (sharp-pointed cracks) are present in the zone

of transition-type corrosion The spikes usually contain sulfides alone or sulfides followed by oxide penetration They occur over a narrow region of the blade in a manner suggesting that spike formation is dependent on stress

The high temperature form of hot corrosion involves the formation on the hot-gas path parts of condensed salts that are often molten at the turbine operating temperature The major components of such salts are Na2SO4 and/or K2SO4, which are apparently formed in the combustion process from sulfur from the fuel and sodium from the fuel or the ingested air Because potassium salts act very similarly to sodium salts, specifications limiting alkali content in fuel or air are usually taken to be the sum total of sodium plus potassium

Very small amounts of sulfur and sodium or of potassium in the fuel and air can produce sufficient Na2SO4 in the turbine

to cause extensive corrosion problems because of the concentrating effect of the turbine pressure ratio For example, a threshold level has been suggested for sodium in air of 0.008 ppm by weight; hot corrosion will not occur below this level Therefore, nonlayer-type hot corrosion is possible even when premium fuels are used This has been especially true

in aircraft-derivative turbines, which have turbine blades made from B-1900 (UNS N13010) Alloy B-1900 has performed well with ultraclean aircraft fuels, but has experienced numerous corrosion problems in land-based service Other fuel (or air) impurities, such as vanadium, phosphorus, lead, and chlorides, may form with Na2SO4 mixed salts having reduced melting temperatures and thus broaden the range of conditions over which attack by molten salts can occur Agents such

as unburned carbon can also promote deleterious interactions in the salt deposits

Research over the past 15 years had led to better definition of the relationships among temperature, pressure, salt concentration, and salt vapor-liquid equilibria so that the location and rate of salt deposition in an engine can be predicted

In addition, it has been demonstrated that a high chromium content is required in an alloy for good resistance to type I hot corrosion The trend toward lower chromium levels with increasing alloy strength has therefore rendered most superalloys inherently susceptible to this type of corrosion The effects of other alloying additions, such as tungsten, molybdenum, and tantalum, have been documented; their effects on rendering an alloy more or less susceptible to hot corrosion are known The near standardization of such alloys as IN-738 and IN-939 for first-stage blades and buckets, as well as FSX-

414 (Co-0.25C-29.5Cr-10.5Ni-7W-2maxFe-1maxMn-1maxSi-0.012B) for first-stage vanes and nozzles, implies that these are the accepted best compromises between high-temperature strength and hot corrosion resistance It has also been

possible to devise coatings with alloying levels adjusted to resist type I hot corrosion The use of such coatings is essential for the protection of most modern superalloys intended for duty as first-stage blades or buckets

The low-temperature form of hot corrosion produces severe pitting and results from the formation of low-melting eutectic mixtures of essentially Na2SO4 and cobalt sulfate (CoSO4), a corrosion product resulting from the reaction of the blade surface with SO3 in the combustion gas The melting point of the Na2SO4-CoSO4 eutectic is 545 °C (1013 °F) Unlike type I hot corrosion, a partial pressure of SO3 in the gas is critical for the reactions to occur in low-temperature hot corrosion Knowledge of the relationships between SO3partial pressure and temperature inside a turbine allows some prediction of where layer-type hot corrosion can occur Because first-stage blade metal temperatures in heavy-duty engines range from about 650 to 855 °C (1200 to 1575 °F), all three types of hot corrosion can occur when sulfur and sodium are present in sufficient quantities

To avoid hot corrosion in stationary combustion turbines, fuel specifications for sulfur, sodium, potassium, and vanadium are typically set at approximately 1% S, 0.2 to 0.6 ppm Na + K, and 0.5 ppm V Impurity content limitations can be varied if blade coatings are used, and corrosion inhibitors, such as magnesium, can be added to the fuel Where the ambient air at the site is contaminated, as in industrial or coastal locations, air filtration is also often practiced

Problems have been experienced with occasional batches of fuel containing higher-than-specified levels of impurities A problem that has had to be addressed is the difficulty in accurately measuring low levels of elements such as sodium in fuel oil Compliance with stringent specifications requires careful supervision and the use of such techniques as centrifuging the oil, which result in increased costs Impurities from other sources, such as the plum stones (which contain sodium and potassium) used in a carboblast cleaning technique at low engine power, have led to cracking of aluminide coatings and corrosion of blades and vanes Entrapment of plum stone fragments in these components places the corrosive species in contact with the alloy surfaces

Air filtration is not a panacea It is expensive and requires proper maintenance and monitoring to prevent the periodic release into the engine of material captured on the filters For example, there are reported instances of collected contaminants being washed off of high-efficiency filters and into engines by sudden heavy storms

Overall, with stringent control of fuel specifications and good air filtration, essentially no unexpected corrosion-related problems are encountered The life limitation is then the creep strength or thermal fatigue strength of the first-stage blades

or vanes Where such controls cannot be exerted, alloys with some inherent corrosion resistance are used, together with a coating The alloy used and the type and thickness of the coating are generally the least costly options that correspond to the planned engine maintenance schedules

Historically, the development of corrosion-resistant coatings was aimed at combating high-temperature hot corrosion The earliest coatings were the diffusion aluminides It was found that chromium-modified aluminides offered little additional protection against high-temperature hot corrosion compared to the basic aluminides, but that the platinum-aluminides offered superior protection compared to the basic aluminides The chromium-modified have since been found to be particularly beneficial against low-temperature hot corrosion, giving results equivalent to those of the platinum-aluminides; both modified aluminides performed better than the basic aluminides

Although these diffusion aluminides have been successful in reducing hot corrosion, the chemistry of these coatings is not readily modified for further improvement in corrosion resistance Thus, increased attention has been given to the development of overlay coatings, which offer significant compositional flexibility

The actual compositions of these coatings depend on their intended use Because Al2O3 is used for protection against high-temperature hot corrosion, coatings that exhibit the greatest high-temperature protection are generally high in aluminum (11%) and low in chromium (<23%) Low-temperature hot corrosion, on the other hand, depends primarily on

Cr2O3 for protection; therefore, coatings exhibiting the greatest low-temperature corrosion protection are high in chromium (>30%) and low in aluminum Other elements, such as silicon, hafnium, tantalum, and platinum, are added to these coatings in an attempt to improve resistance to corrosion and spalling High-chromium MCrAlY coatings have been developed to offer superior low-temperature protection without sacrificing high-temperature protection because industrial gas turbines sometimes operate under varying load conditions that could result in exposures to both low- and high-temperature conditions Overlay coatings have been applied by such techniques as electron beam physical vapor deposition, plasma spray, and sputtering

Apart from overall material wastage due to hot corrosion, an additional concern has been the degradation of mechanical properties, particularly creep and fatigue resistance Figure 24, based on accurate laboratory tests, depicts for several alloys the variation of the ratio of time to rupture in a salt environment to the time to rupture in air with applied stress at

705 °C (1300 °F) The pronounced degradation of the stress rupture life (sometimes by a factor of as much as 105) due to the corrosive action of the salt mixture is evident The plots tend to converge at low stresses; this indicates that a threshold stress level may exist for each alloy, below which rupture life may become insensitive to corrosion Additional studies are needed to understand the mechanism of and implications of the environmentally induced mechanical property degradation

in the context of the field performance of components

Fig 24 Degradation in rupture life for various superalloys due to hot corrosion at 705 °C (1300 °F)

Components Susceptible to Dew-Point Corrosion

W.M Cox, D Gearey, and G.C Wood, Corrosion and Protection Centre, Industrial Services, University of Manchester Institute of Science and Technology

Dew-point corrosion is the attack in the low-temperature section of combustion equipment resulting from acidic flue gas vapors that condense and cause corrosion damage to the plant materials It occurs when gas is cooled below the saturation temperature pertinent to the concentration of condensable species contained by the gas Waste flue gas produced by the combustion of fossil fuels may contain several components, such as SO3, HCl, and H2O, and therefore may display several dew-point temperatures at which the various species begin to condense Corrosion rate peaks related to individual condensation processes may appear, but the formation of protective corrosion products and the deposition of soot ash can

moderate the corrosive effects of deposited acids In conventional boiler plants, the risk areas normally include the posteconomizer flue gas handling section, that is, air heaters, ducting and precipitators, induced-draft fans, and chimney stacks

Dew-point corrosion problems in nominally dry flue gas handling systems will be discussed in this section Problems in wet FGD systems are covered in the section "Corrosion of Flue Gas Desulfurization Systems" in this article This section will emphasize the large coal- and oil-fired power generation equipment, but similar problems are found in many other combustion systems The underlying mechanisms of attack are similar, but their detailed natures are affected by the operation variables Dew-point corrosion and the secondary factors affecting it are thoroughly discussed in Ref 90 and 91

Most fossil-fired power plants are constructed of carbon steel, and past work has largely been) conducted on this material

In specialized applications, low-alloy steels, stainless steels, stainless, nickel-base alloys, and organic and inorganic coatings are also use Much recent dew-point corrosion work is reported in Ref 90 The following discussion draws heavily on the information in Ref 90 and on a recent 6-years collaborative investigation sponsored by the United Kingdom Department of Trade and Industry (Ref 91)

Areas Susceptible to Attack

A conventional coal-fired power generation boiler is illustrated in Fig 25; the locations of items prone to dew-point corrosion are indicated The most susceptible areas are discussed below

Fig 25 Schematic of a fossil-fired power generation boiler showing areas susceptible to dew-point corrosion

(black areas)

Penthouse Casing and Hanger Bars The penthouse casing encloses the tube header pipework above the boiler

furnace roof Flue gas leakage from the furnace can cause corrosion of the casing and furnace support steelwork The hanger bars holding the furnace tube bundles are particularly susceptible to attack near the seals retaining the bars passing from the casing to the external environment (Fig 26) Bars can corrode such that insufficient cross section remains to support the load, and failure allows collapse of the furnace roof and superheater tube bundles

Fig 26 Schematic of a hanger bar packing box showing areas prone to corrosion Source: Ref 90

Air Heater Cold Ends Air heaters are commonly either of honeycomb matrix (Ljungström or Rothemühle) or shell

and tube construction For both, damage is usually most severe at the cold end

A Ljungström rotating matrix air heater is shown schematically in Fig 27 Rothemühle air heaters are similar except that the matrix is stationary and the flue gas/inlet air is supplied by an arrangement of rotating hoods In the Ljungström type, the ductwork is stationary, and the heat-exchanger matrix rotates in either a horizontal or a vertical plane

Fig 27 Schematic of a Ljungström air heater showing location of the air seal and cold end basket corrosion

sites

Corrosion damage is sustained by the thin steel heat-exchanger elements and by support steel-work and air seal materials

It is partly caused by cold air leaking from the inlet to the outlet duct (bypassing the air heater matrix), but damage is often worsened by mechanical interaction between the rotating and fixed components as well as by displaced heat-exchanger elements that have fallen from their baskets The highest dew-point corrosion rates are often associated with operation of a cold-end sootblower, which removes fouling deposits by steam or air jets Increased dew-point attack seems linked to moisture droplets entrained in the blowing medium impinging directly on the lower edges of the air heater elements, causing dissolution of aggressive salts, removal of protective bonded deposits, mechanical abrasion or erosion

of the element surface, and fatigue cracking of the element plates The local injection of considerable moisture may also tend to increase the acid dew-point temperature of the flue gas in a region where the metal temperature may already approach this dew point

In tube-type air heaters, corrosion is again normally found at the cold end The same temperature and sootblowing parameters described for matrix air heaters also apply to tube-type heaters, but bypass air leakage is not a problem until tubes have been perforated However, tube air heaters are often prone to poor gas distribution across the heat-exchanger surface This leads to localized cool spots where dew-point corrosion takes place

Ductwork, Expansion Joints, Inspection or Sampling Ports, and Access Doors Damage at the locations

shown in Fig 25 is often caused by constant low operating temperatures, which are normally related to air entry at fabrication faults, leaking expansion joints and door seals, or careless replacement of sampling port covers Such leaks can cause an appreciable reduction in duct metal temperature due to stratification effects within the gas stream Attack rates of 5 mm/yr (200 mils/yr) have been reported in precipitator outlet manifold ductwork Such damage substantially affects ductwork integrity and frequently leads to increased attack rates on plant components downstream, particularly on the electrostatic precipitator housing and fittings and on the chimney stack

Electrostatic Precipitators and Filter Bag Houses Fly ash precipitators are not normally used on oil- or gas-fired

equipment, but are usually present on coal-fired systems or furnace exhaust streams Filter bags are sometimes used instead; however, both methods of dust collection are located in large insulated housings, and their internal components are expensive to maintain and replace