White Papers & Presentations Extending the Run Time of Dirty Gas Fans with Advanced Wear Protection Technologies Download complete paper here. (436 KB PDF) Download presentation here. (877 KB PDF)  Abstract The desire to extend run times between scheduled outages and dwindling maintenance budgets has increased the power generation industry’s reliance on advanced wear protection technologies to lengthen equipment life while maintaining clean, quality power production. The use of boiler heat transfer improvement systems, such as sootblowers, can have a damaging impact on downstream equipment, including induced draft (ID) fans and other erosion prone components. By protecting equipment with erosion resistant materials, plants can reliably extend planned outage cycles, while reducing overall costs and decreasing the risk of unscheduled outages. This paper will summarize field erosion experiments conducted by the Electric Power Research Institute (EPRI) and the Tennessee Valley Authority (TVA) Kingston Power Plant, on one of the plant’s power boiler ID fans, in an attempt to reduce severe wear and extend intervals between planned outage cycles. Several wear protection materials were tested to determine which could endure the severe erosion experienced by the plant’s fans. Test results showed that brazed tungsten carbide cladding is superior to other protective materials in this extreme environment. The plant successfully increased the run time of their dirty gas fans from 5-8 months to over 30 months by cladding the fan blades with infiltration brazed tungsten carbide cladding. Independent reviews of severe wear protection methods on low NOx burners and superheater boiler tubes will also be presented, with similar results. Introduction In today’s power generation industry, extending scheduled maintenance shutdowns is becoming increasingly important. At the same time, plant managers, process engineers, and maintenance staff have become accountable for increasing plant efficiency and reducing downtime and operational costs, while adhering to strict environmental standards. The erosion and corrosion of steel components continues to plague plant personnel by triggering expensive equipment replacements, costly and inconvenient downtime, and reductions in plant productivity. The use of advanced wear protection on equipment, such as dirty gas fans, low NOx burner components, boiler tubes, shaft seal sleeves, and material conveyance components, can enable power plants to reliably extend scheduled outage cycles while reducing maintenance budgets and diminishing the risk of unexpected downtime. Practical field experiences and independent tests have demonstrated that infiltration brazed tungsten carbide cladding, as an alternative to chrome carbide weld overlays, thermal sprays, and other wear resistant coatings, provides increased erosion, corrosion, and abrasion protection. Infiltration brazed tungsten carbide cladding has been proven to be superior in fly ash conveyance fans, and in other applications, where both erosion and corrosion, or a combination of the two, are significant mechanisms of failure. The reliable severe wear protection offered by infiltration brazed tungsten carbide cladding can reduce plant operating costs, improve productivity, and lessen the risk of unplanned outages. Tennessee Valley Authority Kingston Power Plant The TVA’s Kingston Power Plant began operations in 1955 and is one of eleven coal-fired power plants owned and operated by the TVA. The plant is located west of Knoxville, Tennessee at the junction of the Emory and Clinch Rivers. For more than a decade after its completion, Kingston was the largest coal-fired power plant in the world. The Kingston plant operates nine Combustion Engineering coal-fired boilers and produces approximately ten billion kilowatt-hours of electricity each year, supplying more than 700,000 homes. In order to meet demand, the plant burns about 14,000 tons of coal daily. All nine boilers use a blend of low-sulfur coal to decrease SO2 emissions. In order to reduce NOx, Units 1-4 and Unit 9 use combustion controls and boiler optimization, while Units 5-8 utilize low NOx burners. In 2004, Selective Catalytic Reduction (SCR) systems were added to Units 1-4 and Units 7-8 to further reduce NOx emissions during the summer ozone season. SCR systems will be installed in Units 5-6 in the summer of 2005. Each boiler uses two Model 16MVID ID fans supplied by the Sturtevant Division of Westinghouse. Kingston’s Induced Draft Fan Erosion In 1977, electrostatic precipitators were added to all nine boiler units. Due to space and financial limitations, the ID fans for Units 5-9 were left upstream of the electrostatic precipitators and mechanical fly ash collector, while the ID fans for Units 1-4 were moved downstream of the precipitators. Within six months after the precipitators were installed, maintenance personnel discovered that the ID fans for Units 5-9 were being severely eroded by fly ash. The fans were experiencing high dust loads of approximately 3.6 grams/acfm. At an average run time of 12-14 months, the steel fan blades, supporting hardware, and center hub had to be repaired or replaced. (See Figure 1.) Erosion progressed to the point where several fan blades wore completely through, resulting in a fan failure and causing the fan to be lifted off of its concrete foundation. In 1999, the plant installed new sootblowers and Units 5-9 ID fan life shortened to 5-8 months. This suggests that erosive wear increased due to the pumping of more fly ash through the system.    Figure 1. Kingston’s Worn ID Fan Blades, Supporting Hardware, & Center Hub ID fan erosion was reported as one of Kingston’s highest maintenance cost items, costing the plant over $500,000 annually in parts and labor. The plant evaluated moving the ID fans in Units 5-9 downstream to reduce fly ash erosion rates. However, calculations revealed that it would take the plant approximately 20 years to recoup the related renovation costs, and the project would involve a major fan redesign and a large increase in fan motor horsepower to maintain required performance levels. Due to the increasing costs associated with shorter fan run times and the need to operate at full capacity during periods of peak demand, Kingston plant personnel enlisted TVA’s Energy Research & Technology Applications (ER&TA) group for assistance. ER&TA supports TVA’s plant and transmission system operators with the research and development of new technologies. EPRI/TVA ID Fan Erosion Testing EPRI, a nonprofit organization that provides science and technology based solutions to global energy customers, conducted testing in cooperation with ER&TA, to determine which protective materials could endure the severe erosion experienced by Kingston’s Units 5-9 ID fans (1). The project was initiated in January 2001. An EPRI project manager conducted the search for suppliers of wear resistant materials and coordinated daily activities with TVA’s staff. Description of Materials Tested The study examined sixteen wear protected fan blades from six commercial suppliers. Table 1 lists the protective materials tested and their attributes. Tested materials included chrome carbide weld overlay, tungsten carbide manual flame spray with a furnace fuse, two types of tungsten carbide HVOF, tungsten carbide plasma spray, and infiltration brazed tungsten carbide cladding.   Protective Material Process Coating Thickness Estimated Particle Velocity Bond Type Tungsten Carbide Infiltration Brazed 80 mil pad, 40 mil blade N/A Metallurgical Tungsten Carbide Thermal Spray with Furnace Fuse 80 mil pad, 40 mil blade 40-200 Metallurgical Tungsten Carbide HVOF 10 mil 600-800 Mechanical Chrome Carbide Bulk Weld N/A N/A Metallurgical Tungsten Carbide Plasma Spray 20 mil 1133 Mechanical Chrome Carbide Weld Overlay. Fan blades were provided coated by chrome carbide weld overlay plates, composed of a mild steel base with a chrome weld layer (see Figure 2). A thick coating is produced from the welding process, generating a heavy fan blade. The thickness of an overlay can be increased by applying multiple weld layers. However, multiple weld layers create a fragile surface, leading to check cracking. Thicker applications are not practical on ID fans due to weight concerns, and the instability and extra horse power required.   Figure 2. Typical Weld Overlay (Dark spots are chrome carbide particles) A metallurgical bond, with bond strengths from 30,000 to 50,000 psi, is produced from bulk welding, however, the process also creates inherent surface flaws. Extreme localized heating, combined with the difficulty in controlling cooling rates, typically results in material check cracking and component heat affected zones. Channeling will occur as surface check cracks create a path for erosive materials to undermine the base material, jeopardizing structural integrity and possibly leading to a catastrophic failure.   Tungsten Carbide Thermal Spray with Furnace Fuse.  Blades protected with a process utilizing a manual tungsten carbide flame spray followed by a high temperature furnace fusing were also supplied. The process begins when oxygen and a fuel gas, such as acetylene, are fed into a torch and ignited, creating a flame. Next, powder is injected into the flame, where it is melted and sprayed onto the substrate (2). Manual flame spraying typically creates a particle velocity of 40 m/s (130 ft/s). Post-spray fusion temperatures are high enough to create a metallurgical bond by sintering and diffusing, or by brazing, which wets and diffuses elements of the sprayed coating into the substrate. This fusion process can be performed in a furnace, typically with atmosphere control, or with a flame, laser, tungsten arc, or other suitable heat source. As with weld overlays, manual application of sprayed coatings for subsequent fuse are subject to quality difficulties associated with the manual process. Uneven application and variable tip-to-work distances contribute to unpredictable quality. This can be corrected with the automated application of the spray coating. The post-spray fusion process often requires a furnace with a controlled atmosphere to limit oxidation. With the introduction of solid state lasers however, the field application of spray and fuse may become more practical. Tungsten Carbide Plasma Spray. The experiment also included blades coated with a plasma spray, applied with the Gator-Gard process. The application process utilizes a high temperature, high velocity ionized gas stream to propel metal particles onto substrate materials. Higher particle velocities, and the uniformity of temperature and velocity profiles, differentiate the process from conventional plasma spray processes. Resulting coatings have a higher density, hardness, and bond strength, and porosity and oxide contents are reduced (3). The plasma spray process creates estimated particle velocities of 1133 m/s (4000 ft/s). When applied properly, the process is capable of producing a metallurgical bond. However, when applied incorrectly, a mechanical bond with poor adhesion and interparticle strength is created. The effectiveness of the plasma spray process is unpredictable due to variations in substrate geometry and application conditions. The preferred spray position is flat, but flat-vertical applications have shown recent success. The application to small areas such as bars, shafts, and corners is difficult because a smaller percent of powder affixes to the surface. The process is difficult to perform manually and requires automated application. Finally, the thickness of plasma spray is limited in the field to less than 0.020", although there may be some exceptions. Tungsten Carbide HVOF. Blades protected by two similar types of tungsten carbide HVOF were tested. The HVOF method injects process gases, such as hydrogen and oxygen, into the torch’s combustion chamber at high pressure. The gases are ignited in the chamber and gas velocities that reach supersonic speeds are achieved. The tungsten carbide powder is heated by the flame and accelerated at very high speeds, resulting in a dense thermal spray coating (2). The HVOF process creates a mechanical sticky bond between the tungsten carbide particles and the coating layer. Bond strengths of more than 10,000 psi are achieved. Typically, multiple coats are applied on top of one another to increase the material thickness. This process can cause oxidization and other contaminates between layers (see Figures 3 and 4), limiting the thickness, and, in some cases, threatening coating adhesion. This can result in spalling and chipping.   Figure 3. Typical Flame Spray at 100 µm  (Maximum coating thickness approximately 15 mils)  Figure 4. Typical Flame Spray at 10 µm   Infiltration Brazed Tungsten Carbide Cladding. The infiltration brazing method constitutes filling, by capillary action, a porous coating or structure with molten filler metal. While there are many methods for applying the carbide and braze in preparation for infiltration brazing, the principle technique used to manufacture the brazed tungsten carbide cladding discussed in this paper involves a non-woven preformed cloth. The particles used in this process are sized and mixed to provide a homogeneous, stable, dense cladding.     Figure 5. Photomicrograph of Infiltration Brazed Tungsten Carbide Cladding The increased wear protection provided by infiltration brazed tungsten carbide cladding can be attributed to the benefits of the brazing process, which metallurgically bonds the hard particles and matrix metal to the substrate. The cladding has a bond strength greater than 70,000 psi and is extremely crack resistant due to the controlled application and cooling during the brazing process. Hard particle densities of more than 70%, by volume, can also be achieved during the brazing process (see photomicrograph in Figure 5). High particle densities are reached because the infiltration process does not require the transfer of hard particles in a fluid matrix, a requirement that limits hard particle densities for most coatings and overlay methods. The method does not generate significant carbon dilution into the protective layer, ensuring a uniform wear rate. The vacuum furnace process places limitations on the size of components that can be directly clad. However, to enable field application on large equipment, such as large ID fans, liners with brazed tungsten carbide cladding applied to a thin substrate are fabricated and then weld-attached on site. Infiltration brazed tungsten carbide cladding may have a higher initial installation cost than traditional protection methods. However, financial profitability analysis demonstrates that this wear protection can generate overall cash savings and higher Internal Rates of Return (IRR) by extending capital equipment life. Testing Procedures   Figure 6. Wall Loss After 69 Days in Operation (not to scale) The tested fan blades were in operation on Kingston’s Unit 9 ID fan for 69 days, from January 11, 2001 to April 16, 2001 (1). The fan was a double inlet, single exhaust, 400,000 CFM Westinghouse model 16MVID, with forward curve fan blades. The fan was comprised of 120 blades with a shaft speed of 593 RPM. The original fan blades weighed 34 pounds each. Because Conforma Clad’s infiltration brazed tungsten carbide material added five pounds to each fan blade, a new blade was designed. The wear pad was removed and a full penetration weld was used. Wear protected blades were distributed throughout the fan. In order to facilitate balancing, the heavier infiltration brazed tungsten carbide clad blades were located 180° apart. Because material wear rates were unknown, test organizers carefully distributed the test blades so that erosion-induced weight change would not require fan rebalancing. Testing Results At the end of the 69-day test, all but the four blades protected with infiltration brazed tungsten carbide cladding were removed. Test results are shown in Figure 6. The blades protected with chrome carbide weld overlay and tungsten carbide HVOF were removed due to complete coating wear-through. The blades protected with chrome carbide weld overlay experienced a material loss of 0.150" (4 mm), suffered from a crack at the center junction plate and experienced extreme wear at the leading edge. Measurements taken from the infiltration brazed tungsten carbide clad blades showed a material loss of less than 0.010" (.25 mm) at the leading edge. Figure 7 shows the degree of wear experienced by blades protected with thermal spray with post-spray furnace fusing, HVOF, weld overlay, and infiltration brazed tungsten carbide cladding.   Thermal Spray with Post-Spray Fuse   HVOF   Weld Overlay Brazed Tungsten Carbide Cladding   Figure 7. Wear Experienced by ID Fan Blades During EPRI/TVA Kingston Testing Based on the test results, Kingston retrofitted Units 5-9 ID fans with blades protected with infiltration brazed tungsten carbide cladding (see Figure 8). The first blades were installed in October 2002, on fan Unit 8B. After seven months of run time, the infiltration brazed tungsten carbide clad blades showed a material loss of 0.014" (.35 mm) or less, primarily at the leading edge. Based on the applied infiltration brazed tungsten carbide cladding thickness, the blades were expected to last more than 30 months. The test data shows that the densely packed tungsten carbide cladding wears at a uniform and predictable rate. High bond strengths result in a protective barrier that is highly resistant to chipping, cracking, and flaking. The limited amount of material wear on Kingston’s fan Unit 8B after seven months of run time further substantiated the test results. EPRI indicated that the final report is in its draft stage.     Figure 8. ID Fan Blades Clad with Infiltration Brazed Tungsten Carbide The EPRI/TVA rainbow fan wheel experiment demonstrates the real world wear protection success of infiltration brazed tungsten carbide cladding in a severely erosive environment. The TVA Kingston Power Plant improved its power generation productivity and has plans to enhance other production processes, including material conveyance, through the use of this cladding. Similar results have been realized by independent wear protection. Supporting Field Performance Experiments have been performed on various power generation applications, including Low NOx burners and waterwall and superheater boiler tubes, to determine the best wear protection materials. Results are similar to those produced by the EPRI/TVA collaboration. Major OEM Burner Supplier Tests Erosion Protection Technology on Low NOx Burners A major OEM burner supplier interested in protecting their low swirl coal spreaders against abrasion degradation and maintaining long-term performance conducted laboratory testing on industry accepted wear protection materials (4). As coal is fired through the burner, abrasion severely wears the vanes of the coal spreader. Wear on the spreader’s leading edge alters the swirl pattern of the coal, reducing the spreader’s effectiveness. The spreader’s useful life is 1-2 years, depending on the operating conditions and the type of coal being burned. The supplier would like to extend the spreader’s life to 3-4 years.     Figure 9. ASTM G73 Test Fixture The following abrasion resistant materials were tested: Stellite Alloy 6 Weld Rod Stellite Alloy 31 Weld Rod WC219 Infiltration Brazed Tungsten Carbide Cladding Stoody 101HC Weld Overlay Sure Alloy SA1750CR Fusion Clad Weld Overlay Chromium Carbide Coating Silicon Carbide SiC Ceramic Tile Casting A532, Class I, Type A Casting A532, Class II, Type B Casting The experiment followed ASTM standard G73 methods (see test fixture in Figure 9) and analyzed highly abrasive fine grit black beauty coal slag as the erodent materi­al. Testing was conducted at a 90° impingement angle, with particle velocities of 68 m/s (240 ft/s) for 30 minutes. The infiltration brazed tungsten carbide cladding performed best, with a material loss of 0.21g. All other materials showed a material loss from 1.15g to 2.36g. Figure 10 demonstrates the percentage of lost mass for the tested materials. Figure 10. Mass Loss as a Percentage (%) As a result of this testing, burner designers selected infiltration brazed tungsten carbide cladding as the best available control technology for protecting burner compo­nents against abrasion deterioration. They began testing the cladding in the field on low swirl coal spreaders.     Figure 11. Unprotected Burner Component after 22 Months of Service Typical coal/primary airflow velocities through the burners at full load were approximately 24.6 m/s (87 ft/s). The pulverized coal fired at the sta­tion is blended with approximately 9% petroleum coke. The erosive environment was considered to be moderately high because of the high nozzle velocity, and the high silica and alumina content of the coal. Figure 11 demonstrates wear on unprotected burner components. Low swirl coal spreaders were installed into existing low NOx burners in February 2003. Three low swirl coal spreaders were supplied for instal­lation, including an alloy (50Cr/50Ni with Cb) clad with 0.040" (1 mm) of infiltration brazed tungsten carbide cladding. One spreader cast with silicon carbide, and one with stellite weld overlays on the leading edges, was also installed.   In October 2003, the spreaders were inspected. The stellite protected coal spreader, shown in Figure 12, was missing approximately 1 1/2" (38 mm) of the coal spreader vane after 22 months of service. The silicon carbide spreader showed 0.070" (1.8 mm) of wear, was severely cracked, and broke during disassembly. The spreader protected with infiltration brazed tungsten carbide cladding showed no visible signs of erosion (see Figure 13). Measurements showed a material loss of 0.007" (.18 mm), or less than 20% of the total protective layer thickness. The predicted life, before the base material begins to erode, of the spreader protected by infiltration brazed tungsten carbide cladding is estimated to be five years.   Figure 12. Spreader Coated with Stellite Weld Overlay      Figure 13. Spreader Clad with Brazed Tungsten Carbide   Burners protected with infiltration brazed tungsten carbide cladding produce NOx levels that test low at initial startup and remain low throughout the majority of burner life between major outages. This extended performance is achieved by the reduction in abrasion driven deterioration of component geometry. Results include prolonged compliance with NOx emissions, reduced risk of both planned and unplanned downtime, and an increase in overall productivity and reliability. Infiltration brazed tungsten carbide cladding performed similarly well at protecting superheater inlet boiler tubes. EPRI Hot Erosion Testing on Waterwall and Superheater Boiler Tubes Over time, waterwall and superheater tubes can develop soot buildup. A layer of low thermal conductivity material is created, leading to poor heat transfer. Sootblowers can be used to clean tube surfaces and restore optimum heat transfer capacity. Sootblowing can damage tubing and cause tube leaks, reducing boiler efficiency. In December 2003, EPRI completed a hot erosion comparison utilizing ASTM G73 test methodology, to determine the best protective material for waterwall and superheater boiler tubing (5). The test, which used fluidized boiler bed ash as the erodent material, compared commonly accepted corrosion and erosion resistant materials. Testing measured elevated-temperature solid-particle erosion under a normal oxidized state. Tests were conducted at both 30° and 90° impingement angles, with particle velocities of 40 m/s (141.2 ft/s) for three hours. Testing occurred at temperatures of 900°F (482°C) and 1100°F (593°C). Four conditions were utilized to simulate the erosion experienced by waterwall and superheater tubing in a boiler environment. 900°F (482°C), 30° impact angle - waterwall environment, ductile-erosion failure 1100°F (593°C), 30° impact angle - SH environment, ductile-erosion failure 900°F (482°C), 90° impact angle - waterwall environment, brittle-erosion failure 1100°F (593°C), 90° impact angle - SH environment, brittle-erosion failure Table 2 lists the protective materials tested, along with the application processes and testing results, reported as thickness loss. Table 2. Results from EPRI Hot Erosion Boiler Tube Testing No. Target Material Application Process Thickness Loss (µm)   At 900°F (482°C) At 1100°F (593°C)   a="30° a="90° a="30° a="90° 1 Cr3C2-NiCr coating HVOF 5 19 11 38 2 WC200 cladding Infiltration brazing 6 23 13 56 3 LMC-M+WC coating HVOF 20 28 25 99 4 Inc. 625 weld overlay GMAW 54 51 74 90 5 Inc. 622 weld overlay GMAW 56 54 71 104 6 602CA weld overlay GMAW 63 72 75 84 7 Inc. 52 weld overlay GMAW 65 62 68 83 8 Inc. 72 weld overlay GTAW 66 58 73 94 9 312 SS weld overlay GMAW 67 64 70 74 10 309L weld overlay GTAW 71 65 74 85 11 SA387 steel — base metal No application 76 65 90 97 12 Duocor coating TWAS 187 752* 226 82* *coating worn through Testing results indicated that the CR3C2-NiCr HVOF coating and the WC200 infiltration brazed tungsten carbide cladding had the highest erosion resistant levels. These materials were followed by the LMC-M plus tungsten carbide HVOF. All other materials tested proved to have comparable erosion resistance to the base metal, except for the Duocor coating, which performed worse that the plain base metal. CR3C2-NiCr HVOF coatings require automated application in order to produce coatings of adequate quality, limiting on site applications. All spray processes are limited by line-of-site application. The optimum impingement angle for a spray process is from 90° to 60°. Anything outside this range could result in over-spray, jeopardizing bond strength. Typically, the coating has a limited thickness of less than 0.040"s (1 mm). The coating is also prone to spalling, chipping, and flaking as a result of its limited mechanical bond. Conclusion Advanced wear protection applied to equipment components can reduce operating and maintenance costs, reliably extend planned outages, and reduce the risk of unscheduled downtime. The EPRI/TVA Kingston Power Plant field experience, along with other independent testing, has determined that infiltration brazed tungsten carbide cladding can outperform other protection methods in highly erosive power boiler environments. TVA Kingston Fossil Plant - Induced Draft Fan Erosion (Draft Report), EPRI, Palo Alto, CA: 2002. ER&TA Number 021807. Dobler, Klaus. (2003). The Advantages of Thermal Spray. PF Online. Retrieved July 7, 2004 from http://pfonline.com/articles/040301.html. Gator-Gard vs Conventional Plasma. (2002). Sermatech Power Solutions. Retrieved July 10, 2004 from http://www.sermatechpowersolutions.com/PremiumContent/Gas%20Turbine%20Protective%20Coatings/GG-1.pdf Trunkett, K. Scott, and Douglas Goebel, and Michael Saari. (2004). Advanced Erosion Protection Technology Provides Sustained Low NOx Burner Performance. New Albany, IN: Conforma Clad, Inc. Tube Repair and Protection for Damage Caused by Sootblower Erosion, EPRI, Palo Alto, CA: 2004. 1008037 Written by: Jennifer Broadwater, Marketing Communications Specialist, Conforma Clad Inc. K. Scott Trunkett, Power Generation Market Manager, Conforma Clad Inc. Deming Gray, Tennessee Valley Authority