Battling boiler corrosion
Winning the war against boiler corrosion takes the
right combination of chemistry-related tactics and overall operational strategy
By Steven E. Kuehn, Senior Editor
To boiler operators everywhere the enemy is well known. It is corrosion, the creeping malaise that if not held in check will eventually eat away the availability and useful life of a fossil-fueled power plant. According to the Electric Power Research Institute (EPRI), it has been estimated that corrosion-related problems in utility steam generating plants are responsible for 50 percent of forced outages and account for the majority of all tube failures. Another EPRI study pointed out that failures typically occur in the boiler water wall, superheater and reheater tubes, and that corrosion problems such as stress corrosion cracking (SCC), pitting, etc., generally stem from the ingress of chemical contaminants throughout the water cycle.
Pacific Gas & Electric
In an effort to reduce cycle chemistry-related problems and to gauge the impact of the company`s participation in EPRI`s Cycle Chemistry Improvement Program, Pacific Gas & Electric Co. (PG&E) surveyed the chemistry control practices and operating environments at its various fossil-fueled plants. The results of this survey were documented in the paper “Cycle Chemistry Related Issues in Fossil Power Plants,” recently presented at the NACE International CORROSION/94 conference.
The survey`s scope, said authors Kenneth L. James and Ravindra M. Chhatre, involved reviewing the nature and causes of past failures, characterizing each plant`s water chemistry programs and chemistry management practices, and analyzing the interaction between chemistry, maintenance and operations departments.
To accomplish the survey`s goals, PG&E formed a task force team to collect and compile data gleaned from several areas. The task force included a senior operator with water quality assessment experience, a senior level chemist with water chemistry experience and a project manager with the same sort of credentials. Cycle chemistry parameter monitoring, instrumentation and maintenance, staff training, equipment reliability, chemistry-related costs, and inter-department and inter-plant communications were among the specific study topics.
Data was collected via a number of methods including questionnaires, group and individual interviews, equipment and process walkdowns, and historical data reviews. At the end of the survey, each task force team member compiled and submitted his results to the program manager. This information was organized and then sent to a working group represented by technical staff persons from each power plant and the Technical and Ecological Services departments.
According to the authors, the survey revealed that for a typical PG&E supercritical plant, unit chemistry parameters consistently exceeded EPRI guidelines. For example, excursions in the amount of dissolved oxygen in the condensate totaled as much as 128 hours.
In a five-year period, PG&E tallied almost $80 million in indirect corrosion-related costs (Table 1) and thousands of lost megawatt hours from forced outages and forced curtailments (Table 2). The data, said James and Chhatre, clearly showed why maintaining proper cycle chemistry is important.
Survey results pointed out that many deposit- and corrosion-related problems are the direct result of cycle chemistry excursions. Waterside deposits affect the heat transfer process and are generally responsible for boiler tube failures resulting from local overheating. The authors said that, generally, local concentrations of aggressive corrosion-forming species are significantly higher at these deposit sites and typically result in localized corrosion damage. Furthermore, the ingress of oxygen will cause general pitting corrosion of carbon steel components and that chlorides present in water at low temperatures and temperatures above 125 F to 150 F can cause stress corrosion cracking.
The paper listed a litany of chemistry-related failure mechanisms for a variety of systems and equipment. For example, waterwall tube failures via caustic gouging brought units at its Morrow Bay and Humboldt Bay plants down in the mid-`80s, and transgranular stress corrosion cracking caused feedwater heater failures and other problems for several of PG&E`s Pittsburg plant units. Quite a few of the utility`s units suffered through turbine blade and disk cracking problems as well.
A PG&E review of the equipment damage history and reported chemistry excursions showed that most of the corrosion-related failures could be reduced with improved cycle chemistry control.
By controlling dissolved oxygen pitting and deposit-related failure mechanisms (caustic gouging, hydrogen damage, etc.) boiler tube damage could be reduced significantly. It became obvious to PG&E managers that it is important to monitor and control pH, dissolved oxygen and chloride levels at any given time.
Six critical areas
James and Chhatre said that a workable water chemistry control program requires reliable sampling systems and effective, low-maintenance analyzers, monitors and recorders to provide reliable data.
Additionally, thorough personnel training in cycle chemistry issues will yield positive results and get everybody focused on correctly managing this critical operational aspect. The survey pointed out that personnel will benefit from a better understanding of cycle chemistry and the utility identified six critical areas to improve cycle chemistry management and help eliminate systems damage and unit downtime:
1. -Improve the exchange of cycle chemistry information between plants and departments,
2. -Develop and adopt operating procedures with operator action levels to mitigate cycle chemistry excursions,
3. -Provide adequate monitoring equipment to measure parameters that affect critical equipment,
4. -Train personnel to control cycle chemistry parameters,
5. -Restructure staff to make plant operations more efficient, and
6. -Document all failures and cycle chemistry-related problems involving component corrosion, solids generation, transportation and deposition, and all chemical parameter excursions.
Deposit weight a guide
PG&E`s survey established a clear link between cycle chemistry excursions, deposits, corrosion and tube failure. In another paper presented at CORROSION/94, Babcock & Wilcox Co.`s R. H. Weick agreed and noted in his paper, “Internal Boiler Tube Deposit Weights as a Basis for Chemical Cleaning,” that adhering to appropriate water chemistry limits is one way to limit deposits, but whenever water chemistry alone cannot maintain clean steam generating surfaces, chemical cleaning is one of the most effective options.
The simplest way to avoid deposit-related tube failures, said Weick, is to schedule chemical cleanings at safe, fixed-time intervals. But in this day and age, chemical cleaning costs and associated waste and downtime issues all point to extending this interval as long as possible. Ideally, operators would want to establish the longest and safest interval by letting deposits accumulate to a point just before tube overheating could cause damage. Provided boiler chemistry was maintained to prevent underdeposit corrosion, this would be the most cost-efficient time to schedule a cleaning.
Babcock & Wilcox developed a guide for chemical cleaning based on waterside deposit weight and boiler operating pressure (Table 3). Deposit weight is defined as the weight of deposits removed from the waterside surface of a sample boiler tube divided by the surface area from which it was removed. Deposit weight units are usually expressed in grams per square foot or milligrams per square centimeter. One gram per square foot is equal to 1.076 milligrams per square centimeter.
Determining deposit weight begins by taking tube samples. Although the process is expensive, the yield in information should cost-justify the procedure. Tube samples should be removed from areas with the heaviest accumulation of deposits. In the furnace, said Weick, these are normally areas with high heat flux or heat absorption rates. However, steam quality, tube orientation and other factors are also very important. Heat flux rates vary widely in the furnace with the highest heat absorption profile along the height of the furnace in a utility boiler. The furnace heat flux or absorption profile has roughly the same pattern for coal, gas or oil, said the authors.
Boilers with good water chemistry and no damage problems should be sampled in the following areas:
-In the sidewalls, one-third to one-half the distance from the front or rear wall near or below the top burner elevation on units with opposed-wall firing.
-In the center of the target wall near the top burner elevation on units with single-wall firing.
-In the center of the burner walls near the top burner elevation.
-The leading edge tube from division walls near the top burner elevation.
-Near the center of the sidewalls, at the primary air port area, in chemical recovery boilers.
-In sidewalls, one-fourth to one-third the distance from the burner wall at the center line of the burner in package boilers.
-The first pass tube from once-through boilers with multipass circuitry near the top burner elevation.
Weick noted that although it is normal to find the heaviest deposits in high heat absorption areas, if a boiler is operated in low-load cycle mode, heavier deposits are likely to be found in areas of low heat absorption.
The deposit weight values in Table 3 are based on deposits removed from the hot furnace side of the tube using a mechanical scraping method, one of three established methods. The other two are known as the solvent technique and mechanical glass bead blasting. The solvent method employs inhibited hydrochloric acid to remove deposits and weighing the sample prior to and after solvent cleaning. The mechanical methods are useful, but may produce ambiguity in the results.
According to Weick, although deposit weight is recognized as a most useful parameter, other information and data should be factored into the cleaning decision. Deposit characteristics such as density, porosity, layering, adherence and chemical composition must be evaluated as well.
Waterside deposits consist of two distinct layers. The first is an inner uniform layer of thin, hard, dense magnetite adjacent to the entire circumference of the tube`s inner surface.
This layer exhibits high thermal conductance and develops rapidly and stabilizes within several days of operation. In most cases the thickness of this layer reaches approximately 0.01 millimeters and can be up to 0.03 millimeters.
The second, outer layer forms on top of the protective magnetite layer and is predominantly iron oxide transported from the pre-boiler system by the feed water. This layer is soft, porous, nonadherent and has low bulk density. Because it is softer and more porous it plays a significant role in all overheating and corrosion-related boiler tube failures.
Figure 1 shows temperature conditions in a clean tube and one that contains internal deposits under nucleate boiling conditions. With nucleate boiling and a clean tube, the temperature differential across the fluid is small, approximately 20 F, and the overall temperature difference of 110 F is primarily due to the resistance of the tube metal. The tube with the internal deposit, however, requires approximately 60 F to drive the 315,000 W/M2 through the deposit; therefore, both inside and outside tube temperatures must rise accordingly. So, when internal deposits are formed, Weick explained, the material can act as a thermal insulator and a diffusion barrier deposit that produces hot spots and the potential to locally concentrate boiler water chemicals to corrosive levels.
In porous deposits a phenomenon known as wick boiling has been observed. Wick boiling occurs within a porous deposit when water in the tube flows through channels in the deposit to a point at the inner wall where boiling takes place. The resulting steam bubbles then flow back through the deposit to the center of the tube through a chimney-like passage (Figure 2). The effect of wick boiling is the formation of a concentrating mechanism. Chemical diffusion and mechanical mixing are significantly retarded within the deposit. Thermal conductivity will increase with respect to water originally filling the pores, but the pores also tend to concentrate it and any corrosive constituents within it and promote attack under the deposit.
Tube sampling alternative
Using deposit weight analysis as a means to characterize the extent of internal waterside deposit buildup is effective but it remains a time-consuming destructive examination technique that has some inherent drawbacks. In another paper presented at CORROSION/94, Nalco Chemical Co.`s Peter Hicks and Anton Banweg and Babcock & Wilcox`s Mike Parker discussed an alternative, non-destructive testing method based on ultrasonic wave generation.
In their paper, “The Use of Ultrasonic Testing in Determining Waterside Deposit Buildup in Boiler Systems,” the authors noted that there are a number of ways to gather tube deposit data. Chordal thermocouples have been forwarded as one method, but problems associated with where to locate them, data interpretation, and this technology`s inability to provide data on underdeposit corrosion keep them from being universally accepted. Direct borescope examination is another method, but one that does not provide enough information to get a complete picture of exactly what is happening with the deposits in a given tube.
Tube sampling, as mentioned, is a popular and effective way to gain a good perspective on deposit buildup. Hours of downtime, however, are required to extract the tube sections, and that comes only after spending many hours of analysis to decide which tubes to extract in the first place. Often, utility maintenance managers will use borescope examination to help determine which tubes to section but that only adds to the time, cost and complexity of the task and the results can still be imprecise. There must be a better way, and for now, ultrasonic examination holds great promise as both a stand-alone analysis technique and a way to supplement and enhance the effectiveness of the tube sampling method.
Waterside deposit formation
To determine the deposit accumulation in a given boiler waterwall tube, an ultrasonic scope and a hand-held computer are required to acquire data. Another computer, said the authors, can be used to download the data from the hand-held field unit in order to analyze and report data. At the heart of this process is a Babcock & Wilcox technique that analyzes the captured data.
Figure 3 shows schematically the ultrasonic output from a transducer coupled to the outside wall of a tube. The hardware includes an ultrasound generator and receiver, data acquisition unit and a probe coupled to a stepped ultrasonic standard. The display is manipulated to include first, second, third and fourth back echoes. The ultrasonic display is divided into 200 sectors along the X axis.
The first back echo is manipulated (using the ultrasonic scope controls) so that it lies within the first 40 sectors. The fourth back echo is manipulated so that the fourth peak lies in sectors 161 to 200. At this stage, according to Hicks, Banweg and Parker, the height of the first back echo is adjusted so that the amplitude is approximatley 80 percent of full screen height. Only then is the computer activated to analyze the signal.
The program is designed to obtain data by looking in the two aforementioned sectors. The software then curve-fits the data points in these sectors and calculates R1 and R2 values. R1 is the power (integrated area under the curve) for the first back echo, divided by the power for the fourth back echo. R1 numbers typically range between 8 and 25. R2 is the amplitude (height of the peak) for the first back echo, divided by the amplitude for the fourth back echo, and have numbers that typically range between 8 and 25 as well.
Changes in these R1 values are then correlated to deposit loading. R2 values are generally close to R1 values in magnitude, and if the difference between R1and R2 values is greater than 3, then that point is not used in any further data analysis because of poor peak symmetry. The authors said that each reading may take several seconds once the ultrasonic couplant and transducer have been applied to a specific location.
The A-gate shown in Figure 3 is positioned so that it includes the top sections of the first and second back reflections. This gate is set purposely to determine the wall thickness. Essentially, the ultrasonic scope distinguishes information above the A-gate, measuring the distance between the two peaks above the gate. This corresponds to twice the wall thickness of the tube. Wall thickness can be precisely measured (within 0.12 millimeters) in the boiler with ultrasonic. This kind of information is also available when performing the test for deposit accumulation. This wall thickness information can be an effective way to determine tube wastage, said the authors.
The ultrasonic technique was developed for the evaluation of internal deposits on boiler water wall tubes, observed from the fireside surface. This, however, does not preclude using it on other tubes or in serving the boiler and tubes where deposit buildup is a concern. According to Hicks, Banweg and Parker, in order to perform this procedure an operator should be familiar with both the ultrasonic apparatus and the computer and software. Many hours of experience are needed to become confident not only in performing the task, but also in the interpretation of the results. As with most NDT techniques, there is no substitute for experience.
According to Hicks, Banweg and Parker, ultrasonic testing is an ideal way to provide an adequate description of waterside deposits in situ. This method is especially well suited for outages when routine furnace wall thickness testing is being performed and it has the potential to make routine tube sectioning far more meaningful. Data can be gathered from a large number of tubes in a relatively short amount of time and thus it becomes a more economical way to determine the loading profile over large areas of the boiler without excessive tube sectioning.
Tube section selection becomes more accurate as a result, and permits utility and other power plant managers the luxury of reducing the number of tubes sampled and removing tube sections from only the most critical or most fouled regions. Once deposit mapping has been completed, said the authors, periodic inspection at the same locations would reveal how fast deposits are forming and help determine cleaning intervals. The ultrasonic technique has the potential to reduce forced outages, and trim much of the cutting, rewelding, quality assurance and the million other details associated with tube sectioning. The method may also be a very useful way to gauge the effectiveness of various chemicals used for cleaning and help analyze the impact of changes to cycle chemistry and other operational parameters.
Discipline and teamwork
To win the war against corrosion, a plan analogous to a military campaign may be most appropriate. One has to know the enemy, figure out where it`s going to attack, select the best weapons to fight it and then take steps to ensure it won`t regain any hard-won territory after the initial skirmish. The soldiers in this conflict, the operations and maintenance personnel, the chemists and technicians, need to maintain cycle chemistry discipline and continuously examine the tactics they use to achieve a strategic advantage over boiler tube corrosion. In the end it will be the right combination of leadership and teamwork that will defeat this mighty foe. END
Editor`s note: This article contains material excerpted from papers given at the NACE CORROSION/94 Conference and published with permission from NACE International. For more information or copies of the papers (Numbers 205, 208 and 210) contact NACE International at (713) 492-0535.