After Six Months` Operation, the Steam Turbine Condenser Tubes at a Cogeneration Plant Near Las Vegas, Nev., developed multiple leaks which resulted in an extended forced outage. Visual inspection of the condenser tubes initially revealed that the tube ends and inlet rolled areas were experiencing deep pitting and erosion-corrosion attack. The original tubes were replaced and 100 percent of the new tubes were eddy current (EC) tested for baseline data. A follow-up EC examination after six months showed no indications of metal loss in the tubes inspected. There was also no measurable change in the thermal efficiency of the condenser. The equipment has now operated for more than five years with no further problems.
The Nevada Cogeneration Associates #1 (NCA #1) facility is located 16 miles north of Las Vegas, Nev., adjacent to a gypsum wallboard plant. NCA #1 is owned by a partnership of Texaco Inc. and the Bonneville Pacific Corp. The facility produces 85,000 kW of electricity for Nevada Power Co. The 210 MMBtu/hr condenser is a vertically split, two-pass design with 3,558 tubes. The original tubing material was 7/8 in, 18BWG, 90:10 copper-nickel (Cu-Ni). This is a zero discharge plant, with the cooling water cycled 30 times from the initial make-up (well) water using a side-stream clarifier.
Visual inspection identified tube damage in the form of deep pitting and erosion attack in the inlets of both passes and within the tube roll area. The second pass showed the worst damage to the tube inlets (Figure 1). This area exhibited a bright metal surface with little evidence of the protective cupric oxide barrier required for corrosion protection.
There was additional cutting (erosion) of the aluminum-bronze tubesheet in the ligament between adjacent tubes (Figure 2). This resulted from the high inlet velocities and turbulence due to the condenser design. Pitting of the tube wall stick-out at the ligament, which provided a flow path between adjacent tubes, preceded the cutting of the tubesheet in all cases.
Five tube ends were removed and sent to a laboratory for metallurgical evaluation. The laboratory identified erosion and underdeposit corrosion as the primary causes of tube degradation. Examination of the tube end in Figure 3 showed that, at the tube inlets, high water velocity and aggressive water combined to cause erosion-corrosion of the susceptible tube material.
The nature of the protective corrosion film that is formed, its continuity and adherence, determine the performance of copper-based alloys. For these alloys, the susceptibility to erosion-corrosion increases slowly with increasing flow velocity until a critical point, the breakaway velocity, is reached. For 90:10 Cu-Ni, the breakaway velocity is approximately 8 ft/sec for salt water and 10 ft/sec for fresh water.
The surface condenser was designed for a nominal 8 ft/sec flow rate. At 8 ft/sec and above, erosion continually removes the protective oxide film that forms. The rate of attack increases rapidly because of the combined effect of erosion and galvanic action between the large area of metal oxide (cathode) and small areas of metal surface (anode) exposed by erosion. Once a pit develops, the turbulence associated with the roughened surface may cause the attack to continue (Figures 4 and 5).
Local conditions can create high turbulence that exceeds the velocity resistance of the material, causing areas of severe damage. Based on an analysis of the return and water box design, flow velocities exceeded 10 ft/sec in localized areas at the tube inlets. The characteristic erosion-corrosion attack results in the formation of clean, bright metallic colored, horseshoe-shaped pits, which are often undercut on the downstream side.
EC examination identified pits beneath the tube scale deposits in the 90:10 Cu-Ni tubes downstream from the tube inlets. Several tubes were retrieved for further examination, and actual pit locations and depths corresponded well with the EC test results. Pitting was randomly distributed along the length and around the circumference of the tubes. The average depth of attack was slightly less than 2 mils. However, the maximum pit depth was 8 mils, or 16.3 percent of the tube wall thickness. This metal loss represents severe corrosion for tubes that had been in service for such a short time.
Underdeposit corrosion occurs under the deposit or scale on the tube surface, which shields the alloy surface from the bulk of the water. The water under the deposit becomes depleted in oxygen, and a differential aeration cell is set up, promoting metal dissolution. In this instance the deposit was siliceous material with phosphates of iron and calcium. The metallurgical examination revealed light scale. Removal of the corrosion products from pitted regions showed the underlying metal had been attacked.
There was a secondary form of corrosion attack, both in the deposit regions and at the tube inlets. In both regions dealloying (denickelification) of the metal had occurred, leaving behind a spongy layer of copper (Figures 6 and 7). In this case dealloying occurred when the more active element, nickel, was selectively removed from the alloy.
A water analysis indicated high concentrations of chloride, sodium and sulfate ions, together with high concentrations of calcium and magnesium. The relatively high concentrations of bicarbonate and calcium make this water likely to scale. Light calcium and silica deposits on the surface of the inner diameter supported this assumption.
The excessive wastage of the zinc waterbox anodes provided additional evidence of the aggressiveness of the water. The average life of the six, 25 pound water box anodes at each end of the surface condenser was approximately six months. The lack of electrical isolation between the tubesheet and the waterboxes augmented this excessive anode wastage. The initially uncoated tubesheet and tube ends provided a significant surface area which the anodes were attempting to protect in addition to the coated waterboxes. Even after the tubesheet was coated, excessive anode consumption continued until isolation kits were installed on the waterbox flanges.
Inspectors conducted EC examinations on random tubes, both first pass and second pass. A machined, notched section of spare 90:10 Cu-Ni condenser tubing was used to calibrate the EC instrumentation. Only indications greater than 20 percent of the wall thickness were reported. EC examination showed that the number of indications had increased in both intervals of February to March, and March to April.
In order to extend the life of the 90:10 Cu-Ni tubes, technicians applied a special tube end (amine-cured) epoxy coating to all of the inlet tubes and to the tubesheets as a short-term fix. This short-term solution provided protection of the tube inlets and tubesheets until the tubes could be replaced.
The plant retubed the condenser with 6Mo-superaustenitic stainless steel tubes. The engineers chose a thinner walled 22 BWG to maintain heat transfer near the original design. The 6Mo stainless provides excellent resistance to the high chloride environment and to underdeposit corrosion, as well as high velocity tolerance.
After the 6Mo tubes were installed, EC inspection on 100 percent of the tubes established baseline tube data for subsequent inspections. A follow-up EC examination on 5 percent of the condenser tubes during the next scheduled outage found no indications of erosion-corrosion, pitting or cracking. Visual examination of the tube ends every six months to a year since the retubing has found no evidence of degradation. As a result of this upgrading of the surface condenser metallurgy, NCA #1 has avoided lost revenues in excess of $128,000 per day. p
Century Brass Products Inc., Heat Exchanger Manual, p. 30.
Permutit Corporation, Permutit Water Conditioning Data Book, 1961, p. 106.
Drew Corporation, Principles of Industrial Water Treatment, New Jersey, 1961, p.66.
Tuthill, Arthur H. “The Right Metal for Heat Exchanger Tubes,” Chemical Engineering, Volume 97, No.1, January 1990.
“Chemtreat Corrosion Test Data Report,” September 23, 1993.
“Guidelines of Selection of Marine Materials,” International Nickel USA. Bulletin 10M2-66 4385.
Franson, Ivan A., “Selection of Stainless Steel for Steam Surface Condenser Applications,” ASME 85-JPGC-Pwr-15, October 20-24, 1985.
“AL-6XN Alloy,” Allegheny Ludlum Corporation, Bulletin 1023 2/91 5M., 1991.
Steven R. Johnson is operations supervisor for Texaco Global Gas and Power domestic operations in Las Vegas, Nev.
Jess C. Hassell is a project corrosion engineer in Texaco`s materials and corrosion group.
Michael Fahrion has been employed by Texaco Inc. for 10 years, and has 31 years` experience as a welding and metallurgical engineer.