When it comes to water used in boilers, it is important to remove potentially problematic impurities. This offers a potentially huge return on investment, even with the prevention of just one major upset, let alone recurring failures
Several decades ago, a popular TV commercial showed a cantankerous man driving down a two-lane road. The vehicle’s tailpipe was spewing smoke and nearly blinding the drivers behind. One of his passengers suggested he consider changing his brand of motor oil, to which the man growled, “Motor oil is motor oil.” The announcer went on to say that “Motor oil definitely is not motor oil,” followed by a pitch for a particular product.
Unfortunately, suggesting that poor water quality may cause problems in plant boilers can yield a similar response from plant personnel: “water is water.” Nothing could be further from the truth.
This article provides straightforward arguments against the “water is water” philosophy that appears much too often at industrial plants. Increased emphasis on improved water-system operation and staff training offers a potentially huge return on investment, even with the prevention of just one major upset, let alone recurring failures.
It starts with makeup
As a basis for discussion, we will use Figure 1, which shows a representative diagram of an industrial steam-generating system.
FIGURE 1. Common schematic of an industrial steam generating system. The blowdown heat exchanger and feedwater heater are auxiliary items that may or may not be present
In power-generation boilers, the high temperatures and pressures require high-purity makeup water, with contaminants limited to low part-per-billion (ppb) concentrations. However, lower-pressure industrial drum units can tolerate some impurities, as the milder temperatures reduce the potential for scaling and corrosion. The American Society of Mechanical Engineers (ASME; www.asme.org) recently updated its long-standing industrial boiler-water-chemistry guidelines [1]. Table 1 of this article is an extract from Table 1 of these guidelines for watertube drum boilers with superheaters, and is representative of the steam generator type shown in Figure 1. We will refer to these data several times in the following discussion.
Even in low-pressure steam generators, impurity ingress control is important. Observe the low limits for feedwater hardness. For decades, a popular method of makeup-water hardness removal has been sodium softening by ion exchange. However, the authors and colleagues have seen numerous instances of hardness excursions caused by inadequate softener operation and maintenance. Even more alarming are cases where operators were directed to bypass the system and feed raw water directly to the boilers when the softener malfunctioned. The typical result of hardness excursions is shown in Figure 2.
FIGURE 2. This layered boiler-tube scale is caused by multiple hardness incursions
Calcium carbonate (CaCO3) is the most common scale, but other deposits are possible, including calcium sulfate and magnesium and calcium silicates. All deposits are strong thermal insulators and can cause tube overheating and failure (Figure 3).
FIGURE 3. A boiler tube that has bulged and blistered by overheating
Calcium carbonate forms from the inverse solubility with temperature of dissolved calcium and bicarbonate alkalinity (HCO3–).
As added protection against this scaling mechanism, some plant makeup systems include an ion-exchange dealkalizer or split-stream dealkalizer or downstream forced-draft decarbonator. With acid feed ahead of the decarbonator, most alkalinity converts to carbon dioxide, which then vents from the unit.
It is not unheard of to simultaneously see poor performance from both the softener and the decarbonator, which greatly increases the potential for CaCO3 scale formation in the boilers.
Another benefit of makeup dealkalization is that alkalinity, upon reaching the boiler, is in large measure converted to CO2 via the following reactions:
Carbon dioxide flashes off with steam and can increase the acidity of the heat exchanger condensate and condensate-return piping when the CO2 redissolves.
Although the pH generated by Reaction (5) has a relatively mild lower limit, the acidity is more than enough to cause significant carbon-steel corrosion in condensate-return systems (Figure 4).
FIGURE 4. Shown here is carbonic acid corrosion of a carbon steel condensate return line. Besides direct attack, carbonic acid and other corrosion mechanisms release iron oxide particulate matter that can settle in boiler tubes to reduce heat transfer and initiate underdeposit corrosion (UDC)
Even with a properly operating softener and decarbonator, many other dissolved ions pass directly to the steam generators. These include chloride, sulfate and silica. While lower-pressure boilers can tolerate some dissolved solids, adherence to the conductivity guidelines shown in Table 1 often requires frequent or steady blowdown, which wastes water and energy. Furthermore, conditions may arise where the ions induce direct corrosion. Some of these mechanisms are explored in the next section.
A makeup-treatment method being adopted at some facilities is reverse osmosis (RO) installation or retrofit. RO lowers the concentration of virtually all dissolved ions. Reliable RO design and operation requires detailed feedwater analyses. A typical design includes upstream particulate filters, or perhaps even micro- or ultrafiltration, to minimize mechanical fouling of RO membranes [2].
Condensate impurities
For many boilers like the type shown in Figure 1, as well as other units, ASME feedwater guidelines are very important. Apart from makeup, feedwater conditions are often significantly influenced by condensate return chemistry. Because of the enormous variety of materials produced at industrial plants, many compounds may potentially enter the condensate return from heat exchanger leaks or other sources. Unfortunately, as with makeup-water-system operation, condensate-return chemistry is frequently neglected until a major upset occurs.
A classic example comes from the power industry, but can easily apply to many heavy-duty industrial boilers, including co-generation units. This power plant’s generating fleet included a relatively small, low-pressure (1,250 psig) unit with a condensing steam turbine. The condenser was on once-through cooling from an artificial lake. After a pre-planned maintenance outage, plant chemists discovered the feedwater total dissolved solids (TDS) concentration had increased from normally low parts-per-billion levels to values approaching 1 part per million (ppm). The laboratory staff traced the leak back to the condenser and requested an immediate unit shutdown to find and repair the leak. Plant management refused because the utility was selling power (another example of the “water is water” mindset). This condition continued for three weeks. Plant chemists worked diligently to maintain boiler water chemistry as close to normal as possible, but within two to three months, boiler tubes began failing with regularity. Eventually, the entire boiler had to be re-tubed at a cost many times higher than the company made from the power sales. Metallurgical analyses revealed underdeposit corrosion that led to hydrogen damage throughout the boiler (Figure 5).
FIGURE 5. Shown here is a sudden failure of a boiler tube from hydrogen damage. Notice the thick-lipped fracture, which indicates failure with little metal loss
The equation below represents a well-known reaction that can occur under boiler-tube deposits.
One product of this reaction is hydrochloric acid (HCl). HCl can definitely cause corrosion on its own, but the compound can also concentrate under deposits where the reaction of the acid with iron generates atomic hydrogen that penetrates into the metal wall and reacts with carbon atoms in the steel to generate methane (CH4):
Gaseous methane formation causes cracking, greatly weakening the steel’s strength. Hydrogen damage is very troublesome because it cannot be easily detected. The phenomenon continues to plague many steam generators around the globe.
While underdeposit corrosion (UDC) can occur underneath a variety of precipitates, porous iron oxide deposits usually have the greatest influence. This explains the low iron limits shown in Table 1. Iron particulates are generated by condensate return and feedwater system corrosion, sometimes by the carbonic acid attack as outlined earlier, but also by oxygen attack. This is a clear example of how one form of corrosion has potential manifestations for other corrosion mechanisms.
The following example illustrates the effect of organic impurities on steam chemistry. Author Buecker once evaluated a fouling issue at an organic-chemicals plant that produced a primary product and four derivatives thereof. Steam was generated by several 550-psig package boilers in which solids deposition in the superheaters (and associated tube overheating) required tube bundle replacements every 1.5 to 2 years. An extracted superheater-bundle inspection revealed internal deposits up to 1/4 in. in depth. The inspection team also discovered foam issuing from every boiler saturated-steam sample line. The plant had minimal analytical chemistry capabilities with no trained water chemists on staff, but a condensate return data review provided by the plant’s water treatment vendor showed total organic carbon (TOC) concentrations as high as 200 ppm. Compare this to the recommended upper limit of 0.5 ppm, as shown in Table 1. The excessive organic contamination in the boilers generated the foam that carried over to the superheaters.
Beyond a “water is water” philosophy, some plant managers are reluctant to install condensate polishers because of cost and the extra personnel needed to operate the equipment. However, such an investment can pay large dividends by minimizing or preventing boiler tube failures and corresponding equipment shutdowns. Ion-exchange polishing is a potential treatment option for condensate that contains inorganic ions, such as hardness, chloride, sulfate and similar elements and compounds. Temperature is an important factor in treatment selection for these systems. Activated-carbon filtration may be effective for organics removal. Several are possible for particulate control, including direct filtration. The aforementioned boiler-makeup alkalinity reduction can reduce CO2 transfer to steam and the potential for condensate carbonic-acid corrosion. Alternatively, carbon-steel corrosion is minimized by operating within a mildly basic pH range. Alkalizing amine feed for condensate and feedwater pH conditioning is common. These are small-chain organic molecules with an ammonia group attached to, or embedded within the compound (Table 2).
Selecting the best product or product blend is sometimes difficult, as each compound has a different basicity and distribution ratio (the tendency for the product to depart with steam or remain dissolved in the boiler water). It is desirable to have comprehensive pH control throughout the network, but a single compound may not be sufficient to achieve broad results. Recently commercialized blended amine products can provide wide-ranging coverage. A thorough analysis of system design, metallurgy, current chemistry, operating temperatures, and other parameters is a prerequisite for proper program selection, as it is for any other condensate polishing or treatment choice. Filming amines that directly protect metal surfaces have also been successfully applied in some applications, but space limitations prevent a discussion at this time.
Also note the low feedwater dissolved oxygen (D.O.) guidelines in Table 1. Depending on the application, D.O. control can be a thorny issue. For power-generating units with high-purity makeup and feedwater, research has shown that a small amount (5–30 ppb) of dissolved oxygen at the economizer inlet is necessary to minimize flow-accelerated corrosion (FAC). FAC has been an extremely concerning issue since the first observed FAC failure killed four power plant personnel in 1986 [3].
However, in industrial units with lower purity makeup and condensate return, D.O. must be maintained at near-zero concentrations to minimize iron and copper corrosion. A key piece of equipment is a mechanical deaerator, which, when functioning properly, will reduce D.O. to approximately 7 ppb. Additional oxygen reduction is achieved with reducing agent/oxygen scavenger feed. Over time, deaerator trays and spray nozzles can become misaligned or damaged, which may be overlooked by plant personnel. On-line dissolved oxygen monitoring is just one part of a comprehensive sampling and analysis program [4].
Boiler-water treatment
The boiler has the harshest conditions within a steam-generating system, and chemistry excursions can easily lead to rapid corrosion or other problems. In power boilers, which typically receive consistent high-purity makeup, the primary issue is pH control to minimize carbon-steel corrosion. Table 1 outlines the typical mildly basic ranges that are optimal for many boilers.
Tri-sodium phosphate (Na3PO4) is still commonly employed for pH control in both industrial and many power boilers.
Knowledge beyond this basic concept can be lacking at plants without trained water chemists. A prime example is in systems where hardness and silica may periodically or regularly enter the boiler via contaminated condensate. Without proper chemistry control, these constituents can form very tenacious calcium and magnesium silicate scales that are extremely difficult to remove. In large measure, it was this chemistry that influenced the development of phosphate treatment with accompanying hydroxide alkalinity. Phosphate and hydroxide alkalinity react directly with calcium to produce calcium hydroxyapatite:
Magnesium and silica react with hydroxide alkalinity to form the non-adherent sludge serpentine:
These two compounds exist as soft sludges that are much easier to remove than the hard scale that would otherwise form. The materials typically settle in the mud drum or lower headers and are removed by blowdown. Careful analyses and system chemistry evaluation are necessary to design and operate the chemical feed systems to maintain proper conditions. Various synthetic polymers have been developed to assist with hardness and iron oxide deposition control. These compounds cannot be introduced arbitrarily to boilers, but must be selected and dosed based on rigorous chemistry evaluations. Some polymers are better at sequestering hardness ions, while others are more effective in controlling particulates. No “one size fits all” exists.
How plant personnel handle boiler chemistry upsets may generate additional problems if incorrect procedures are followed. For instance, the reaction shown in Equation (6) illustrates that some chemistry excursions can lower boiler pH significantly and quickly. As pH drops below 8.0, steel corrosion rapidly increases. The best procedure in such cases is to kill the boiler fires, increase blowdown, drain the boiler when safe, and eventually refill it with fresh, treated water. Before restart, it is imperative to find and correct the contamination source. However, these procedures may curtail plant production, which can cause management angst. A “knee-jerk” alternative is to keep the boiler in operation and inject caustic to raise the pH. Unfortunately, the rapid pH elevation induces precipitation of the recently-generated iron corrosion products, which settle on the boiler tubes and add many more locations for future underdeposit corrosion.
A final note
All systems are different, and, as with all other technologies, due diligence is necessary to determine the feasibility of utilizing these methods. Always consult your equipment manuals and guides. Also, reputable chemical vendors can offer the products to achieve your goals.
References
1. “Consensus on Operating Practices for the Control of Feedwater and Boiler Water Chemistry in Industrial and Institutional Boilers,” The American Society of Mechanical Engineers, 2 Park Avenue, New York, N.Y. 10016, 2021.
2. Buecker, B., and K. Perryman, High-purity water treatment: Using membranes to protect membranes, Water Technology, May/June 2022.
3. Guidelines for Control of Flow-Accelerated Corrosion in Fossil and Combined Cycle Power Plants, EPRI Technical Report 3002011569, the Electric Power Research Institute, Palo Alto, California, 2017.
4. Buecker, B., Monitoring of Water and Steam Chemistry for Steam Generators, Chem. Eng., September 2019. pp. 33–42.
Acknowledgment
Special thanks to Richard Tribble of ChemTreat for supplying the photos in this article.
Authors
Brad Buecker is president of Buecker & Associates (Lawrence, Kan.; beakertoo@aol.com), specializing in technical writing and consulting. He most recently served as senior technical publicist with ChemTreat, Inc. He has over four decades of experience in or affiliated with the power industry, much of it in steam generation chemistry, water treatment, air quality control, and results engineering positions with City Water, Light & Power (Springfield, Illinois) and Kansas City Power & Light Company’s (now Evergy) La Cygne, Kansas station. He also spent two years as acting water/wastewater supervisor at a chemical plant, and eleven years with two engineering firms, Burns & McDonnell and Kiewit. Buecker has a B.S. in chemistry from Iowa State University with additional course work in fluid mechanics, energy and materials balances, and advanced inorganic chemistry. He has authored or co-authored over 250 articles for various technical trade magazines, and has written three books on power plant chemistry and air pollution control. He is a member of the ACS, AIChE, AIST, ASME, NACE (now AMPP), and the Electric Utility (and Co-Generation) Chemistry Workshop planning committee.
Wendy McKinnon is vice president, Applied Technologies for ChemTreat, Inc. (Glen Allen, Virginia 23060; Phone: 804-935-2000; Fax: 804-965-0154; Email: wendym@chemtreat.com). She has been with ChemTreat for 20 years. Prior to becoming vice president, she was in a technical consultant role specializing in water and wastewater clarification, reverse osmosis and deionization. She has a B.S. in chemistry from the University of Mary Washington.