By Ravi Jethra, Endress+Hauser
Liquid analysis is one of the most important measurements in any type of power plant (fossil fuel or nuclear). Generating electricity is a fundamentally inefficient process. The efficiency of current generating capacity is approximately 35 percent (globally about 30 percent).This implies that about two-thirds of the potential energy in all fossil fuels is wasted. With fuel accounting for roughly three quarters of a coal-fired power station’s operating cost, the need to ensure optimum energy efficiency is critical.
This article examines how carrying out monitoring throughout the plant using analytical instrumentation can help ensure optimum efficiency throughout the water and steam loop.
The primary objectives of liquid analysis are to :
- Minimize local corrosion of boiler materials, which can occur in regions of restricted flow, particularly under deposits and within crevices
- Minimize transport of corrosion products into the boilers, condensate and feedwater section
- Minimize erosion in steam , turbine, feedwater and condensate areas.
Boiler water chemistry has a direct effect on boiler efficiency and fuel use. Improper water treatment will allow scale formation and corrosion. Scale insulates boiler tubes and requires increased fuel use and is a major culprit for boiler failures.
It is impractical to completely eliminate the presence of potential contaminants in boiler feedwater. Elevated temperatures and pressures inherent in power generation applications greatly increase the speed of the chemical reactions taking place in the boiler. The result is an aggressive environment that can dramatically reduce boiler life if not properly controlled.
By measuring and monitoring not just the boiler chemistry, but also other areas around a power plant, it is possible to obtain a better overview of current conditions. When incorporated into a planned preventive maintenance program, this information can substantially help reduce the risk of unplanned outages. For example, in a 500 MW plant, 1,500 tons of water are boiled off each hour; that equates to one million tons a month. Consider that most of the resulting contaminants present in the water will remain in the boiler and the need for close monitoring and control becomes apparent.
The most common source of corrosion in boiler systems is dissolved gas: this can include oxygen, carbon dioxide and ammonia. Of these, oxygen is the most aggressive as it is a source of pitting and iron deposition. Even small concentrations of this gas can cause serious corrosion problems. Pitting can produce failures even though only a relatively small amount of metal has been lost and the overall corrosion rate is relatively low. The degree of oxygen attack depends on the dissolved oxygen concentration, the pH and the water temperature.
The influence of temperature on the corrosivity of dissolved oxygen is particularly important in closed heaters and economizers where the water temperature increases rapidly. Elevated temperature by itself does not cause corrosion. The temperature rise provides the driving force that accelerates the reaction. As a result, even small quantities of dissolved oxygen can cause serious corrosion.
Makeup water introduces appreciable amounts of oxygen into the system. Oxygen can also enter the feed water system from the condensate return system. Possible return line sources are direct air-leakage on the suction side of pumps, systems under vacuum, the breathing action of closed condensate receiving tanks, open condensate receiving tanks and leakage of non-deaerated water used for condensate pump seal and/or quench water.
What to Measure and Where?
To operate the steam generator, feedwater with special properties is necessary. The feedwater must be free of so-called hardness (amount of alkaline earths), that leads to boiler scale and boiler sludge formation. Scale on the boiler body and on the wall of the steam pipes increases the temperature and reduces the strength of the boiler material. Feedwater should not contain dissolved gases like oxygen and carbon dioxide and also should be free of salts, as these contaminants speed up corrosion of the boiler wall and pipes in the water/steam system. Feedwater should also be free of mechanical impurities and colloidal dissolved bodies like oils and fats. A special value of dissolved salt content is also important to avoid foam of the feedwater within the boiler.
Figure 1 Boiler feedwater circuit in power stations
Important characteristics are hardness, conductivity and concentration of hydrogen ions (pH).
In a range of pH 9 to 10.5, materials such as steel pipes will be less vulnerable to attack than at lower pH values. If the alkalinity is not satisfied (in other words, for pH > 7) chemicals are used to condition the feedwater. These chemicals are typically ammonia or caustic soda. An overuse of alkaline protects the boiler and pipes against corrosion and boiler scale. The aggressivity is nearly zero at pH = 9.6. Therefore, pH measurement is used to control the quality of feedwater, for protecting the water/steam cycle of corrosion and boiler scale. At the same time, the pH-value provides information on the quantity of chemicals (ammonia and so on) that have to be dosed into the cycle to realize a stable pH value.
Condensed steam is pure water, which is normally fed back into the boiler instead of more expensive treated water. The condensate return system is an informative sampling point used to monitor overall boiler water conditioning system performance. The quantity and nature of the contaminants often point to problems such as entrainment and corrosion, which require corrective measures. A frequently occurring cause for contaminations in the condensate return is a leak in the coolant line from the turbine condenser. Dissolved gases and lube oils can enter through return pumps and dissolved ions. Suspended metal particles can be absorbed as corrosion by-products. Condensate polishing is often used to prevent these contaminants from entering the boiler. Detecting the cause of contamination and checking the conditioning systems depend on careful monitoring of key parameters. Extreme hardness, conductivity and turbidity can be a sign of an internal leak in heat exchangers and condensers.
Iron, copper and dissolved oxygen also point to corrosion problems. Sodium and silica in the condensate can be an indication of entrainment problems or condenser leakage. Analyzers for silica and turbidity can be used to highlight the effectiveness and regeneration point of a polishing system.
The control parameters that are to be monitored are silica, hydrazine and other oxygen-scavengers as well as phosphate, pH value and conductivity. The condenser and the feedwater system are the most likely areas for system contamination. In the condenser, steam from the turbines is condensed using cooling water. Although this water is pretreated to remove mud, silt and any organic matter, problems still can occur if the water becomes mixed with turbine steam condensate. Over time, condenser leaks are almost inevitable, allowing contaminated cooling water to enter the condensate compartment.
Figure 2 Steam water quality system in power stations
With the feedwater system, de-ionised water is preheated and chemically treated before entering the boiler. Some chemicalssuch as sodium hydroxide or sodium phosphatecan speed up boiler corrosion if misapplied.
To keep the boiler/steam section of a power plant running at peak efficiency, the following parameters should be monitored constantly:
Dissolved Oxygen: When dissolved oxygen enters the steaming boiler, corrosion manifests itself in the form of severe deep pits, almost exclusively at the water level in the steam drum. Even small parts-per-billion concentrations of oxygen dissolved in the feedwater stream can cause boiler pitting, drastically affecting its operating life. The concentration of dissolved oxygen therefore needs to be checked throughout the feedwater loop, from the extraction pump through to the deaerator and the boiler inlet.
As an oxygen scavenger, hydrazine is widely used to remove trace levels of dissolved oxygen in the boiler feedwater. Hydrazine is a colorless liquid, which is highly soluble in water. It is a powerful reducing agent that reduces oxygen to form nitrogen and water. At high temperatures and pressure, it will also form ammonia, which increases the feedwater pH level, reducing the risk of acidic corrosion.
Placing a hydrazine monitor at the feedwater inlet will help check that feedwater is being dosed with the correct amount of hydrazine. Too much hydrazine is both wasteful and costly. Too little will inadequately control dissolved oxygen levels, preventing the formation of magnetite layer.
Measuring conductivity in the feedwater and steam loops provides an indication of water and steam purity. By measuring the feedwater electrolytic conductivity (that is, the ability of the feedwater to pass an electrical current) it is possible to ascertain the level of contamination, which can then be used to dictate the level or duration of treatment required. For example, where boiler blowdown is used, conductivity will be one of the main parameters used to control the frequency of the blowdown process.
Silica: Although not a source of corrosion, silica can form extremely hard and dense scales in the boiler and turbines, lowering heat transfer efficiency and increasing the risk of mechanical failure such as turbine blade malfunction. Dissolved silica gets weakly ionized and is difficult to detect by conductivity measurement. Thus, dedicated silica analyzers are required to obtain accurate measurement.
Sodium: Sodium is a critical parameter to monitor in a power plant. Although conductivity measurement is typically used to indicate total dissolved solids or chemical conductivity, it lacks adequate sensitivity. As sodium is present in the critical dissolved compounds, it can be detected with on-line sodium monitors, which are very sensitive.
Other parameters that operators also tend to monitor include phosphate, ammonia and chloride using sensors that offer quick response times, are temperature tolerant and require minimal maintenance.
To maximize online monitoring system efficiency it is always advisable to use instruments with fast response times and self-diagnostic capabilities. The location of monitoring equipment is critical. Ideally, monitoring equipment should be in an environment offering less potential for damage while providing easy access for maintenance and superior accuracy.
pH and Conductivity: pH is an extremely important parameter to measure, as it gives an indication of the degree of acidity or alkalinity of the feedwater.
With the reduction in cost, the user has also attained greater reliability and accuracy in the measurement due to improved sensor designs. Smart, two-wire transmitters are less expensive than they were 10 years ago and also offer advanced digital bus and sensor diagnostic capabilities.
In addition, advances in automation capabilities now enable users to reduce associated maintenance and calibration costs. Within the pH sensor itself, several performance improvements have evolved; in particular, the traditional glass electrode has dramatically evolved. Glass formulations have been refined to improve the selectivity of the glass toward hydrogen, reducing the effects of sodium ion error.
Thermal stress has long been a problem. Fusion of the glass-measuring bulb to the supporting stem glass enables thermal stress fractures to occur at the glass weld. Due to slight variations in the composition and the corresponding differences in elasticity or thermal expansion coefficients, repeated changes in temperature will induce cracking.
To overcome this problem, optimized thermally matched glass is often used. Close comparisons of raw glass materials and slight changes in glass formulations also reduce the effects of repeated temperature changes. For certain applications, ion-selective field-effect transistor sensors can serve in lieu of traditional glass electrodes.
Many of today’s transmitters offer the ability to assess the condition of the measuring and reference electrodes, temperature compensation element and current calibration. By measuring the impedance of the measuring glass, the pH meter can detect aging or cracking caused by abrasion, thermal stress, or simple glass fatigue. Monitoring the resistance between the reference junction and the reference electrode can assess material buildup or junction fouling. Monitoring the continuity of three-wire resistance temperature detectors allows detection of temperature compensation element failure.
Lastly, evaluating the sensor slope, asymmetry and zero-point alerts the user to potential failures and provides the user with insight into the sensor’s approximate remaining useful life. Each of these diagnostic features allows the user to perform predictive maintenance rather than rely exclusively on schedule maintenance. This translates to savings in unneeded labor.
Overall, today’s pH measurement solutions can provide their users with a competitive advantage. Anyone who has worked in a laboratory knows the problem. A simple pH measurement can turn into a nightmare simply because the workbench is not grounded. Simply stated, it is not that easy to get an undistorted signal to the sensor. There can be a lot of problems hidden along the path to the sensor. Moisture and contamination can distort the signal. Cable length and cable characteristics can affect signal quality. Measurement in the lab is anything but easy and online analysis while a plant is running is even more challenging.
Sensor technologies are evolving to provide unique ways of achieving undistorted signal transmission. This would solve the electrical and mechanical problems as well as issues that occur with high impedance measurements in moist, contaminated environments.
A silicon chip inside the sensor housing converts the raw measurement value into digital signals. The inductive, non-contact connector guarantees an undistorted signal path to the transmitter. The electrode is a defined, closed system which can be calibrated as a separate entity. The cable is only used for power and signal transmission and has no influence on measurement accuracy. A warning is raised if there is a problem with the connection between the sensor and the transducer; for example, if the cable is disconnected. By contrast, a conventional sensor would provide false data.
The transducer’s electrical isolation simplifies wiring requirements and enhances electromagnetic compatibility. The cable is attached to the sensor via a bayonet connector. This prevents accidental removal of the process connector when the cable is detached.
However, built-in “intelligence” enables the smart digital pH sensor to do much more. The ability to store calibration data is a huge advantage. Sensors can be pre-calibrated in the lab under optimal conditions, and then can be quickly installed on the system in the field. Generally, digital sensors are available for monitoring pH, conductivity and turbidity.
Apart from advances in sensor technology leading to digital sensors, the platforms to which these are connected have undergone transformation as well. The smart platform allows users to read sensor information such as serial numbers and calibration data, as well as process information including sensor operating time, number of calibration operations, maximum operating temperature and operating time under extreme conditions; for example, a defined temperature limit. Diagnostic messages provide timely warning of changes and errors. Visual indication is provided for alarm limits and fault conditions. Modular design gives the smart platform great user flexibility in terms of different digital communications protocols that may be needed on plant-site. Different types of sensors such as ISFET, glass electrodes for pH and conductive and sensors can be attached to the smart platform.
Ravi Jethra is an industry manager – power at Endress+Hauser in Greenwood, Ind. He received his Bachelor’s degree from Bombay University in Instrumentation Engineering and an MBA from Arizona State University. He is a Senior member of ISA, and a member of ASME and IEEE. The author would like to thank his colleagues Andreas Schmidt and Tracy Doane-Weidemann for their expert guidance and assistance in putting together this article.