By Brad Buecker, Contributing Editor
Steam turbine and surface condenser performance is just as important as ever, and is intimately tied in with chemistry control on both the steam side and cooling water side. In this article, we examine these issues, plus revisit a straightforward method to monitor condenser performance.
A Turbine Discussion First
Utility steam turbines typically consist of three stages, high-, intermediate-, and low-pressure. A question for the readers is, "Have you ever considered how much power is produced by each?" Let's look into this topic.
The figure on pg. 53 is from Babcock & Wilcox's book, Steam, now in its 41st Edition. The drawing outlines a realistic heat balance from a 2,400 psig drum unit with a 593 MW generator output.
|Figure 1 Drum Unit Heat Balance|
|Source: Babcock & Wilcox|
While a thermodynamic analysis of this system could seem quite daunting, in actuality the First Law of Thermodynamics can be utilized to reasonably determine the work output of each turbine section. For steady state operation, the turbine work (WT) can be calculated by mass flow rate (m) multiplied by the enthalpy drop (Dh) through the turbine.
WT = m(hin – hout)
Of course, in the steam generator outlined in the figure, a number of steam extractions are utilized for feedwater heating. However, the work done by the steam from all processes can be calculated by the straightforward application of the formula outlined above. An analysis of Figure 1 shows that the output of the three turbine sections (and in which author's calculations suggest a moisture content of the low-pressure exhaust at 6.5 percent) is as follows.
- HP turbine – 166 MW
- IP turbine – 138 MW
- LP turbine – 289 MW
Although each turbine generates a significant portion of the total power, the LP turbine has the largest capacity.
Each turbine section can be affected by operational and steam chemistry factors that reduce efficiency. We will consider several of the most important. Please note that these principles also apply to heat recovery steam generators (HRSGs), and thus are universal across the utility steam-generating industry.
First, in drum units some moisture always entrains with the steam as it leaves the water surface in the drum. Thus, drums are equipped with internal steam separators to remove most entrained moisture and return it to the boiler water. However, complete removal of droplets is impossible, and thus trace amounts of solids will carry over to the superheater and main steam. These impurities include chlorides, sulfates, silica, phosphates (in phosphate-treated units) and other compounds. A number of factors can cause carryover to be excessive. These include poor drum design, failed steam separators, high dissolved solids in the boiler water, and excessive ramp rates during startup, among others.
Carryover can also be vaporous, in which certain compounds will volatilize on their own and transport to steam. Silica and copper (in units with copper-alloy feedwater heater tubes) are two notorious agents for vaporous carryover.
Not to be forgotten is steam attemperation, where the common process is to control main and reheat steam temperatures by direct injection of feedwater. Impurities that enter the condensate via a condenser tube leak, makeup water treatment system upset, or other source will be directly introduced to the steam.
Both mechanical and vaporous carryover become more pronounced with increasing pressure, thus boiler water chemistry guidelines are increasingly stringent at higher pressures. But how do these impurities affect turbine operation? The list below outlines many of the most important issues.
- Silica solubility decreases as steam pressure decreases through the turbine. Thus, silica will deposit on turbine blades, particularly in the HP section.
- Chloride and sulfate salts will deposit, commonly in the LP turbine. These impurities, and chloride in particular, may cause pitting and stress corrosion cracking (SCC) of turbine blades and rotors. The most susceptible locations are the last stages, L-1 and L-0, where early condensate forms. During times of low load operation, early condensate formation may move backwards a bit in the LP turbine.
- Sodium hydroxide (NaOH) can also cause SCC of turbine blades.
- In units with copper alloy feedwater heaters that operate at or above 2,400 psig, copper carryover can be quite detrimental. For the most part, copper precipitates in the HP turbine, where even a few pounds of deposition will reduce capacity by perhaps 10 MW or more. This capacity loss can be critical during periods of high power demand.
The allowable concentration of contaminants in boiler water to prevent carryover is variable from unit to unit, and also obviously is a function of pressure. The Electric Power Research Institute (EPRI) has published many guidelines on this issue, but the details are much too expansive to be presented here, although I have included an example to highlight this issue. At 600 psia, the guideline for boiler water silica concentration is 6 ppm to ensure no greater than 10 ppb in the main steam from mechanical and vaporous carryover combined. Of course, this is a general guideline, and each steam generating unit needs to be evaluated separately. Regardless, the recommended boiler water silica concentration at 2400 psia is 0.2 ppm!  This example illustrates the large influence unit pressure has on carryover. Current limits for main/reheat steam sodium, chloride, and sulfate concentration are all 2 parts-per-billion (ppb), with, as we have seen, a silica guideline of 10 ppb. Saturated steam sampling can be a valuable tool to determine the level of mechanical carryover from drum units. I have seen excess mechanical carryover due to incorrect drum size, and also have observed contamination of boiler water and main steam by organics that produced foam. Excessive firing rates can also influence carryover.
The Condenser: Money Made or Lost Due to Heat Transfer
After the steam generator, the condenser is the largest heat exchanger in the power plant. Although air-cooled condensers are becoming more popular as a water conservation measure, steam surface condensers still dominate this technology. Condenser tube fouling or scaling, or air binding due to excess air in-leakage will reduce efficiency. Any of these mechanisms can cost a plant hundreds to thousands of dollars per day in lost performance. So, what are major issues of concern?
Often the most troublesome cooling water issue is microbiological fouling of condenser tubes, auxiliary heat exchangers, and, where cooling towers are utilized, cooling tower fill. Cooling systems provide an ideal environment, warm and wet, for microbes. Bacteria will grow in condensers and cooling tower fill, fungi on and in cooling tower wood, and algae on wetted cooling tower components exposed to sunlight. Biocide treatment is absolutely essential to maintain cooling system performance and integrity.
A problem with microbes, particularly bacteria, is that once they settle on a surface the organisms secrete a polysaccharide layer for protection. This film then collects silt from the water, thus growing even thicker and further reducing heat transfer. Even though the bacteria at the surface may be aerobic, the secretion layer allows anaerobic bacteria underneath to flourish. These bugs in turn can generate acids and other harmful compounds that directly attack the metal. Microbial deposits also establish oxygen concentration cells, where the lack of oxygen underneath the deposit causes the locations to become anodic to other areas of exposed metal. Pitting is often a result, which can cause tube failure well before the expected lifetime of the material.
Fungi will attack cooling tower wood in an irreversible manner, which can eventually lead to structural failure. Algae will foul cooling tower spray decks, potentially leading to reduced performance and unsafe working locations.
The core of the microbiological treatment program is feed of an oxidizing biocide to kill organisms before they can settle on condenser tube walls, cooling tower fill, and other locations. Chlorine was the workhorse for many years, where when gaseous chlorine is added to water the following reaction occurs.
Cl2 + H2O ⇔ HOCl + HCl
HOCl, hypochlorous acid, is the killing agent. The functionality and killing power of this compound are greatly affected by pH due to the equilibrium nature of HOCl in water.
HOCl ⇔ H+ + OCl-
OCl- is a much weaker biocide than HOCl, probably due to the fact that the charge on the OCl- ion does not allow it to penetrate cell walls. The killing efficiency of chlorine dramatically declines as the pH goes above 7.5. Thus, for the common alkaline scale/corrosion treatment programs, chlorine chemistry may not be efficient.
Due to safety concerns, liquid bleach (NaOCl) feed has replaced gaseous chlorine at many facilities. The major difficulty with bleach is that the product contains a low concentration of sodium hydroxide, thus when it is injected into the cooling water stream it raises the pH, if by only a small amount.
A popular alternative is bromine chemistry, where a chlorine oxidizer and a bromide salt, typically sodium bromide (NaBr), are blended in a makeup water stream and injected into the cooling water. The chemistry produces hypobromous acid (HOBr), which has similar killing powers to HOCl, but functions more effectively at alkaline pH.
The primary disadvantages are that an extra chemical is needed and feed systems are a bit more complex than for bleach alone.
Chlorine dioxide (ClO2) has found some application as an oxidant for two primary reasons. Its killing power is not affected by pH, and it does not form halogenated organic compounds. Also, chlorine dioxide is more effective in attacking established bio-deposits. New techniques for producing ClO2 are making this option more attractive.
An oxidizer that appears to offer a potentially effective alternative is monochloramine (NH2Cl). This chemical has been widely used in potable water systems, as it remains in solution much longer than free chlorine, and thus can be effective in large distribution systems. However, chloramines are less potent than free chlorine towards planktonic microbes in the bulk water. The potential benefit in cooling systems appears to stem from the fact that NH2Cl is much better at penetrating the slime layer of bacterial colonies to kill the sessile microbes underneath. A full-scale test at an anonymous utility indicated that the chemical removed previously formed microbial deposits and restored condenser performance during the summer of 2011.  The plant utilizes once-through cooling.
Space limitations prevent a full-blown discussion of scale prevention in cooling water systems, but the topic is of great importance, especially because restrictions on once-through cooling may force more plants to install cooling towers. Long ago, chemists and engineers discovered that without treatment, calcium and bicarbonate ions (Ca+2 and HCO3-, respectively), when subjected to heat in a condenser or when cycled up in concentration in a cooling tower would form a hard calcium carbonate (CaCO3) scale on condenser tube surfaces. The following equation outlines this process.
Ca+2 + 2HCO3- + heat → CaCO3↓ + CO2↑ + H2O
Unlike many salts, calcium carbonate is inversely soluble with temperature, thus as temperatures rise in a heat exchanger, the potential for CaCO3 scaling increases. For once-through systems, feed of simple polyacrylate is often sufficient to prevent scaling. Programs for systems with cooling towers are much more complicated. In the "Good Old Days" of water treatment, a common technique consisted of sulfuric acid feed to the cooling tower makeup to convert bicarbonate alkalinity to CO2 gas, with feed of chromate (CrO4-2) to the recirculating water to provide corrosion protection for metal surfaces. However, chromate use has essentially been banned due to the toxic characteristics of hexavalent chromium (Cr+6). Most towers now are operated on an alkaline regime to minimize corrosion, and where to this point in time inorganic and especially organic phosphates have been used for scale prevention and additional corrosion protection. These programs are not foolproof, as poor chemistry control can lead to scale formation from phosphate compounds. An important note is that phosphorous-based treatment programs are losing favor due to phosphate's nutrient effect in natural waters. Phosphate in cooling water discharge can lead to algae blooms in surface waters and difficulties thereby. All-polymer programs are gaining popularity as the technology improves. 
The Thermodynamics of Condenser Fouling
Let's go back to the example outlined at the beginning of this document and investigate the effects of condenser tube fouling or scaling. We will examine the LP turbine, and for simplicity's sake will ignore the extraction flows. The diagram shows that 2,744,971 lb/hr of saturated steam exits the LP turbine. At the design condition of 2.5" Hg pressure and a calculated 6.5 percent moisture level, the enthalpy of the exhaust is 1043.1 Btu/lbm. So the work done in the LP turbine, excluding that from the extraction streams, is,
2,744,971lb/hr * [(1386.7 – 1043.1) Btu/lbm] ÷ 3,413,654.6 Btu/MW = 271.4 MW
Now consider if condenser fouling causes the absolute backpressure to increase by 1" (Hg) to a level of 3.5" (Hg) absolute. Thermodynamic calculations show that the exhaust enthalpy increases to 1090.1 Btu/lbm. The work from the low-pressure turbine decreases to 238.5 MW. This is a loss of nearly 33 MW! This example illustrates the importance of maintaining condenser cleanliness.
Condenser performance can also be affected by excess air in-leakage. If the vacuum system is not capable of removing all of the air, the remainder will coat tubes and form pockets within condenser tubes. Air is a very good insulator, and air binding can cause efficiency losses as great as those induced by fouling or scaling.
Condenser Performance Monitoring
In 1992, I wrote about condenser performance monitoring for Power Engineering. In the 20 years since, many power plant engineers, chemists and operators have retired from utility life and have been replaced by new personnel.
The technique for condenser performance monitoring about which I wrote in 1992 is still just as valid today, and indeed I have discovered the calculations in sophisticated programs installed on plant distributed control systems (DCS). However, I have also learned that the data may be overlooked in the complexity of hundreds of other readouts, and that plant operators may not recognize condenser performance problems as they first arise. Prompt detection is often vital to prevent small difficulties from becoming large. Via the following straightforward process, plant engineers and chemists can monitor condenser performance and be prepared to react if problems are detected. Much of the following text is adapted from my 1992 article.
At that time, some of my former co-workers from City Water, Light & Power had attended the General Physics (www.gpworldwide.com) course, "Fundamentals of Power Plant Performance for Utility Engineers."
This excellent course included calculations for determining condenser cleanliness factors. With personal computers still being somewhat in their infancy, we first put the calculations in BASIC language and used them to monitor condenser performance. Of course, nowadays spreadsheet format is an excellent method to track condenser performance, and for any interested readers I have the calculations in Excel. The program utilizes the process variables of cooling water inlet temperature, outlet temperature, and hotwell temperature, along with fixed constants to calculate a condenser cleanliness factor. The fixed constants are,
- Circulating water flow rate (gpm)
- Number of condenser tubes
- Inside tube diameter
- Outside tube diameter
- Tube length
- Number of passes
Also necessary are two variables known as the circulating water correction factor and condenser tube correction factor, which exist in table format. These values, which were originally compiled by the Heat Exchange Institute (http://www.heatexchange.org">www.heatexchange.org) are easy to input into a spreadsheet.
A general rule-of-thumb regarding the program output, i.e., the cleanliness factor, is that a clean condenser will exhibit a reading of 85 percent. I found this to be true with some condensers, but for others came up with readings that exceeded 100 percent. The latter was due to design factors. However, the strength of this program is its ability to monitor trends. Using a clean condenser as a baseline, a factor above 100 percent is not an issue. The key is detecting a decline in condenser cleanliness factors, which the program will do without influence from seasonal changes in cooling water inlet and outlet temperatures. These changes can otherwise mask true problems.
I successfully used the program for over 15 years to track performance of a number of condensers and alert operations and maintenance personnel to problems. Both microbiological fouling and scaling typically manifest themselves as a gradual decline in cleanliness factor. Often, these problems arise due to a malfunction of chemical feed equipment. It is very important to catch such problems early.
In one instance, a summer drought had reduced the volume of the cooling water surface supply by a factor of four, and correspondingly increased the dissolved solids concentration by the same factor. Two once-through condensers that had never encountered fouling or scaling difficulties in the past began to exhibit gradually declining performance, as determined by the program. As it turned out, the drought-induced increase in dissolved solids coupled with summer temperatures initiated calcium carbonate scaling in the tubes. A mechanical contractor came in to scrape the tubes during autumn outages on both units.
The program is excellent for detecting excess air in-leakage, which occurs quite commonly. The strong vacuum in a condenser naturally will pull in air from any leaks in the condenser shell or the many lines that penetrate the condenser shell. In one case, the program showed a precipitous drop in condenser cleanliness factor from 75 percent to 45 percent in a single day on a 200 MW unit.
An inspection revealed that a crack had developed in the condenser shell where a heater drips line penetrated. The maintenance crew pad-welded the leak, and cleanliness factors returned to normal.
Approximately three months later, the readings once again dropped dramatically – the weld had failed. Maintenance personnel then welded a steel collar over the spot, which solved the problem permanently. On another unit, a malfunctioning trap on the condensate return line of a gland steam exhauster allowed outside air to enter the condenser. The problem was unique in that the air in-leakage was not evident at high loads, but certainly was at reduced load. Cleanliness factors sometimes fluctuated between 65 percent and 28 percent within the same day as load changed. These wild fluctuations ruled out tube fouling as the problem. After an extensive search, plant operators found the defective trap and it was repaired. The difficulty might have gone on for months without detection from this excellent program.
B. Buecker, "Computer Program Predicts Condenser Performance Monitoring"; Power Engineering, June 1992.
Kitto, J.B., and S.C. Stultz, Steam/its generation and use, 41st edition, 2005, the Babcock & Wilcox Company, Barberton, Ohio.
Cycle Chemistry Guidelines for Combined Cycle/Heat Recovery Steam Generators (HRSGs). EPRI, Palo Alto, CA: 2006, 1010438.
Personal conversation with Darrell Rose, Buckman Laboratories, Power-Gen 2011, and subsequent follow-up with personnel at the utility where the test was performed.
Buecker, B., Post, R., and R. Aull, "Chemical Treatment and Fill Selection Methods to Minimize Scaling/Fouling in Cooling Towers"; the 71st International Water Conference, Orlando, Florida, November 13-17, 2011.
Brad Buecker is a contributing editor for Power Engineering and also serves as a process specialist with Kiewit Power Engineers, in Lenexa, Kan. He has over 30 years of experience in, or affiliated with, the power industry, much of it in chemistry, water treatment, air quality control, and results engineering positions with City Water, Light & Power in Springfield, Ill., and Kansas City Power & Light Company's La Cygne, Kan., station. He has written many articles and three books for PennWell on steam generation topics. He is a member of the ACS, AIChE, ASME, and NACE. He is also a member of the ASME Research Committee on Power Plant & Environmental Chemistry, the program planning committee for the Electric Utility Chemistry Workshop, and the program planning committee for Coal-Gen.
Power Engineerng Issue Archives
View Power Generation Articles on PennEnergy.com