Coal, Gas

Steam Purity Debate Rages On

Issue 2 and Volume 106.

By Brad Buecker,
Contributing Editor

In recent years I have reported on papers presented at the International Water Conference held every October in Pittsburgh, Pa. I have often provided an overview of important topics related to steam generation. This time, however, I wish to focus on an issue that dominated much of the steam generation chemistry sessions at the 2001 conference. This is the topic of steam purity, its effects upon turbine operation, and the criteria for monitoring steam purity.

As boiler technology improved during the last century and steam pressures reached higher and higher levels, steam turbine manufacturers tightened steam purity limits to protect turbine components. From about 1970 onwards, EPRI began contributing additional valuable data on chemistry influences in turbines. Evidence clearly indicates that carryover products or impurities injected through attemperator systems can initiate and propagate serious turbine corrosion and deposition. Chlorides, and to a lesser extent sulfates, are notorious corrosive agents, while the deposition of copper on high-pressure turbine blades has received much attention in recent years. Many other corrosion/deposition mechanisms are also well known. What is still being hotly debated is the influence of organic oxygen scavengers and pH conditioners, and especially their decomposition products, on steam system corrosion. This is an issue that will not be resolved soon.

Research and chemical analyses have shown that the high temperatures in a boiler and superheater will cause organic feedwater treatment chemicals to decompose into short-chain acids and the final breakdown product, carbon dioxide. These reduce the pH in the turbine, especially at the low-pressure end where condensate begins to form. That acids can initiate corrosion is obviously a well-established fact, but what still fires the debate is “Do the often very small acid concentrations produced by organic treatment chemical breakdown generate enough product to cause corrosion?” I have seen compelling data from both sides of the issue, and we will look at the topic in more detail shortly.

Tied in with this issue is that of steam chemistry monitoring. All steam turbine manufacturers provide limits on the concentration of contaminants that should be allowed into the turbine. Impurities of most concern include chloride, sulfate, sodium, silica and copper. Another item on most lists is cation conductivity. This measurement increasingly seems to be emerging as kind of a catchall for steam purity monitoring and evaluation. Of course, the conductivity of water increases with increasing dissolved solids concentration, especially if the contaminants are elements or compounds that ionize. So, conductivity can be a useful tool in monitoring contaminant excursions into a water or steam process. It is particularly valuable in high-purity systems when even a small impurity increase can be detected quickly.

The problem is that some people may be trying to simplify steam chemistry monitoring to that of cation conductivity only. A limit that has emerged for proper steam quality is 0.2 µS/cm. Some guidelines suggest that steam with a cation conductivity below this limit is considered to be in good condition, while if the conductivity is higher than 0.2, the steam is suspect. Perhaps this guideline results from the “new” plant environment where facilities are operated with minimal staff and little chemistry expertise. When operators only have to monitor basic data, chemistry decisions are greatly simplified. Unfortunately, such simple measurements do not take into account the much more complex chemistry that may be present.

Luis Carvallho of BetzDearborn Canada presented a very interesting survey of Canadian steam-generating plants, many of which have been unable to meet a 0.2 µS/cm limit for years, but yet have turbines in excellent condition. Many of these units have also operated with organic oxygen scavengers and pH-conditioning chemicals, so the two issues that strike me upfront are that cation conductivity is too simple to use as a sole steam purity monitoring criterion and that organic treatment chemical breakdown and its effects on turbine corrosion need much more study. Oxygen scavenger concentrations are typically in the low part-per-billion range, and carryover of decomposition products to the turbine may be so slight that corrosion is minimal to non-existent.

An issue that continues to influence interest in organic oxygen scavengers is the concern over the safety of hydrazine. I have reported before that some plants have installed systems to minimize employee exposure to hydrazine vapors and certainly any contact with liquid hydrazine. Yet, with any chemical the fact that it is on-site means that some accidental exposure is possible. It is understandable that plant managers and environmental personnel might wish to avoid hydrazine-related safety concerns.

It was interesting to note that Tom Pike of Western Farmers Electric Cooperative discussed both organic oxygen scavenger feed and oxidation-reduction potential (ORP) monitoring in the same system. I believe that ORP monitoring is becoming a very valuable tool for chemistry control in mixed-metallurgy feedwater systems, where plant personnel must minimize copper corrosion while at the same time preventing potentially catastrophic flow-assisted-corrosion (FAC) failures of carbon steel. I have attended presentations before where the authors discussed ORP monitoring in systems on hydrazine, but ORP should also be an excellent monitoring technique in units on organic oxygen scavengers.