O&M, Plant Optimization

Protecting Your Steam Turbine from Corrosion

Issue 10 and Volume 121.

By Brad Buecker

Many power plant personnel are aware that chemistry upsets in a steam generator may cause severe corrosion and failure of boiler waterwall tubes and other components.

These failures place the staff at risk, and also lead to severe economic hardships for the plant.

However, often overlooked is that even trace levels of some impurities in steam can induce severe to catastrophic corrosion of turbine blades and rotors under certain conditions.

These issues are often magnified by the high cycling duty of most plants, including formerly base-loaded units.

With the aid of some excellent information presented at the recent 37th Annual Electric Utility Chemistry Workshop, the author will outline several of the most important issues with regard to protecting this most valuable piece of equipment.

How Do Impurities Enter Steam?

Because steam generating power plants, both conventional units and heat recovery steam generators (HRSGs), operate at high temperatures and pressures, very pure makeup water is a requirement to prevent corrosion and scaling. The core treatment process at many plants for high-purity water production is, following suitable pretreatment methods, two-pass reverse osmosis with either mixed-bed ion exchange or electrodeionization polishing of the RO effluent. Regardless of the exact makeup system design, it must produce water with very low part-per-billion (ppb) impurity concentrations. So, in the absence of condenser tube leaks or contaminated boiler water treatment chemicals, the water entering the steam generator is typically quite pure, with the normal sodium concentration less than 2 ppb and cation conductivity less than 0.2 μS/cm. More on these measurements appears shortly.

Many power plant personnel are aware that chemistry upsets in a steam generator may cause severe corrosion and failure of boiler waterwall tubes and other components. This article outlines the most important issues related to protecting this valuable piece of equipment.

The issue with drum units is that impurities “cycle up” in concentration due to production of steam, which leaves dissolved solids (typically analyzed and reported as total dissolved solids [TDS]) behind in the boiler. Dissolved and suspended solids are controlled via boiler blowdown, but the fact remains that some impurities can enter steam via any of three methods, mechanical carryover, vaporous carryover, or direct introduction from attemperator sprays. With regard to the former, virtually all high-pressure steam drums are equipped with moisture separators to remove entrained water droplets from the steam. But no separator is 100 percent efficient, and some moisture escapes the drum. The density difference between water and steam is the primary driving factor in moisture separation efficiency, and because this density difference decreases with increasing pressure, mechanical carryover becomes more pronounced in high-pressure steam generators. At a pressure of say 2,500 psia, moisture carryover may reach 0.2 percent of the steam flow. Mechanical carryover is the typical method by which most impurities, and particularly those which we will most focus upon in this article, chloride and sulfate, enter steam.

Some elements and compounds, most notably silica (SiO2), will vaporize on their own and enter steam. As with mechanical carryover, vaporous carryover becomes more pronounced as pressure increases. Consider for example the recommended boiler water silica concentrations designed to limit the steam concentration to 10 ppb. At 1,000 psia the general boiler water limiting concentration is around 2.3 ppm, but at 2,500 psi the limit drops to about 0.18 ppm. Silica does not cause turbine corrosion, but as the steam pressure decreases during passage through the intermediate- and low-pressure turbines, the compound will precipitate and modestly degrade turbine aerodynamic efficiency.

An element that once was quite problematic with regard to vaporous carryover is copper, especially in many of the coal-fired units from decades past. A large number of these steam generators were originally equipped with copper-alloy feedwater heater tubes, where corrosion introduced copper and copper oxides to the boiler. At pressures above 2,000 psi and especially at 2,400 psi and above, vaporous carryover of copper becomes quite troublesome. Deposition typically occurs in the HP turbine, and just a few pounds of copper deposits can cause the loss of several megawatts. Very few if any HRSGs have feedwater heaters, so this is a non-issue in modern combined cycle units. And, of course, other materials besides copper alloys are possible for any units that still have feedwater heaters.

Impurities may also be introduced to steam in a direct manner, via attemperator sprays. This is normally not an issue unless the feedwater has been contaminated from a condenser tube leak, or, less likely, a makeup water system upset. Then, harmful compounds will enter the entire system. This is yet another example why comprehensive on-line sampling, including condensate pump discharge and feedwater, is critical for protecting steam generators.

Current normal steam purity limits are outlined in the table below. These values may seem very minute, but as we shall see, even slight impurity ingress, especially of chloride, can still present serious difficulties.

Stress Corrosion Cracking.

How Impurities, Even at Trace Levels, Can Damage the Turbine

Power-generating steam turbines are typically divided into three sections, high-pressure (HP), intermediate pressure (IP), and low-pressure (LP). In virtually all modern units, HP exhaust steam returns to the steam generator for reheating, and then is introduced to the IP turbine, whose exhaust “crosses over” to the LP turbine(s). (Often, the configuration may have two or perhaps occasionally even three LP turbines per overall system.) Consider as an example of common operating conditions the following approximate values from a planned combined cycle project. The data is based on an average summer day at the site, with no supplemental duct firing to the HRSGs.

  • HP Inlet Steam: ~2,000 psia, ~1,050o F
  • Cold Reheat Steam: ~570 psia, ~700o F
  • Hot Reheat Steam to IP Turbine: ~510 psia, ~1,050o F
  • LP Inlet Steam: ~60 psia, ~610o F
  • LP Exhaust to Condenser: ~1.2 psia, ~110o F

This data clearly illustrates the drop in steam pressure and temperature across the turbine sections as the energy is converted to mechanical work and electrical output. We have already noted that silica will precipitate from the steam as pressure drops in the IP and LP turbines. But what about the really bad actors, chloride and, to a lesser extent sulfate? Most difficulties with these salts occur in the LP turbine. The entering steam still has a reasonable amount of superheat, but as the steam reaches the last few rows of turbine blades, some of it begins to condense. This location is known as the phase transition zone (PTZ), and it is in the PTZ that chloride and sulfate salts precipitate on the turbine blades and rotor. During steady-state operation, the precipitated salts are neutral, but in units that cycle on and off regularly (many plants nowadays) the LP turbine may be frequently exposed to humid, outside air. When the salt deposits become moist, they can initiate pitting of blades and rotors. Pitting in itself is a very serious issue, but other factors exacerbate the problem. Rotating turbine blades, and particularly the long blades in LP turbines, develop stress points during operation. Pitting is often the precursor to stress corrosion cracking (SCC), in which the combination of a corrosive environment and metal stress can induce severe localized corrosion.

Another problematic mechanism is corrosion fatigue (CF), which, as the name implies, is influenced by repeated cycling. A simple example of basic fatigue is to bend a paperclip at one spot back and forth several times until it fails. Cycling duty in a plant initiates fatigue points at many locations, including rotating turbine blades and attachments. If a corrosive environment exists, the time to fatigue failure is shortened. Weakening of turbine blades and attachments from SCC and CF can lead to blade failure while the turbine is in operation. The only word that applies to this situation is “catastrophe.”

Information from the Electric Utility Chemistry Workshop

The corrosion issues outlined in this article were the subject of an excellent paper at the most recent Electric Utility Chemistry Workshop (EUCW). The lead author outlined in the first half of the paper that during scheduled outages on two power generating units at his company, non-destructive testing (NDT) revealed stress corrosion cracking in blade attachments within the last three stages of some of the LP turbines. Prompt blade and blade attachment repair prevented the problem from becoming a serious issue. This work came in conjunction with upgrades to the plant’s on-line chemistry monitoring system. Accurate and reliable monitoring are aspects that seem to often be overlooked by the plant staff, even though the cost for instrumentation and training for plant personnel can be recovered many times over by prevention of chemistry upsets. Critical sample points include:

  • Makeup water system effluent
  • Condensate pump discharge
  • Feedwater and economizer inlet (economizer outlet is also a good location)
  • Boiler / HRSG evaporator water
  • Superheat, reheat, and saturated steam

With regard to steam chemistry, Table 1 previously illustrated current steam purity guidelines as established by the Electric Power Research Institute (EPRI) and other top organizations such as the International Association for the Properties of Water and Steam (IAPWS). But, the author is aware from these sources that the values are considered to be too high, particularly with regard to chloride and sodium, if the latter occurs in the form of caustic (NaOH) carryover. (Caustic can also induce stress corrosion cracking.)

For years, the primary power plant steam measurements have included some or all of sodium, silica, and cation conductivity (more properly known as conductivity after cation exchange [CACE]). CACE is essentially the electrical measurement of any anions, generally chloride and sulfate, after the cations (in steam generator water primarily ammonium and sodium) have been stripped from the sample. Because carbon dioxide (CO2) influences CACE, now being recommended is degassed CACE, where the sample is routed through a reboiler or perhaps a nitrogen-sparged compartment to remove CO2. For a long time, plant owners, startup personnel, and equipment manufacturers focused on a cation conductivity limit of 0.2 μS/cm as a good guideline for steam purity. But, CACE is only a surrogate for chloride and sulfate, and it is now known that 0.2 μS/cm corresponds to a chloride or sulfate concentration greater than 10 ppb. This is at a time when many consider 2 ppb of these ions to be excessive. The question has naturally arisen, “What about measuring trace chloride and sulfate directly?”

Heretofore, such measurements have been possible with ion chromatography (IC), but from direct experience the author will attest that although the technique can provide accurate results, it is expensive and requires trained chemistry personnel (something combined cycle plants unfortunately don’t always have) to keep IC units in proper working order. However, changes in the ability to monitor trace Cl and SO4 are imminent. Another paper at the EUCW outlined results from several field tests of a new trace chloride/sulfate analyzer.

A key aspect of this technology is that the instrument uses a process with the ponderous name of microfluidic capillary electrophoresis to separate chloride and sulfate, which may then be detected at concentrations down to 0.1 ppb. The electrophoresis module is calibrated at the factory, thus the instrument can be started up in the field without additional calibration. The field tests so far have provided excellent results. The technology could be expanded to detect other anions that may be in the steam samples.

A notable example is phosphate, which is often a mechanical carryover product in those plants that utilize tri-sodium phosphate (TSP) for boiler water chemistry control. Although TSP is not a corrosive agent in the steam system, carryover and deposition in superheater and reheater U-bends have been known to cause overheating failures.

Don’t Forget About Shutdowns and Layups

Often, plant personnel tend to focus on issues that may occur during normal operation. But, off-line corrosion is a very serious issue that must be addressed to ensure good unit reliability.


Brad Buecker is a senior process specialist in the Water Technologies group of Kiewit Engineering Group Inc.