A nuclear steam-generator vessel constructed of 100-mm thick SA302, grade B, steel was found to have a small leak. The leak originated in the circumferential closure weld joining the transition cone to the upper shell. The welds had been fabricated from the outside by the submerged arc process with a backing strip. The backing was back gouged off, and the weld was completed from the inside with E8018-C3 electrodes by the shielded metal arc process. Striations of the type normally associated with progressive or fatigue-type failures including beach marks that allowed tracing the origin of the fracture to the pits on the inner surface of the vessel were revealed. Copper deposits with zinc were revealed by EDS examination of discolorations. Pitting was revealed to have been caused by poor oxygen control in the steam generators and release of chloride into the steam generators. It was concluded by series of controlled crack-propagation-rate stress-corrosion tests that A302, grade B, steel was susceptible to transgranular stress-corrosion attack in constant extension rate testing with as low as 1 ppm chloride present. It was recommended to maintain the coolant environment low in oxygen and chloride. Copper ions in solution should be eliminated or minimized.

Argonne National Laboratory has conducted analyses of failed components from nuclear power-generating stations since 1974. The considerations involved in working with and analyzing radioactive components are reviewed here, and the decontamination of these components is discussed. Analyses of four failed components from nuclear plants are then described to illustrate the kinds of failures seen in service. The failures discussed are (1) intergranular stress-corrosion cracking of core spray injection piping in a boiling water reactor, (2) failure of canopy seal welds in adapter tube assemblies in the control rod drive head of a pressurized water reactor, (3) thermal fatigue of a recirculation pump shaft in a boiling water reactor, and (4) failure of pump seal wear rings by nickel leaching in a boiling water reactor.

Cracking occurred in an ASME SB166 Inconel 600 safe-end forging on a nuclear reactor coolant water recirculation nozzle while it was in service. The safe-end was welded to a stainless-steel-clad carbon steel nozzle and a type 316 stainless steel transition metal pipe segment. An Inconel 600 thermal sleeve was welded to the safe-end, and a repair weld had obviously been made on the outside surface of the safe-end to correct a machining error. Initial visual examination of the safe-end disclosed that the cracking extended over approximately 85 deg of the circular circumference of the piece. Investigation (visual inspection, on-site radiographic inspection, limited ultrasonic inspection, chemical analysis, 53x metallographic cross sections and SEM images etched in 8:1 phosphoric acid) supported the conclusion that the cracking mechanism was intergranular SCC. No recommendations were made.

The secondary cooling water system pressure boundary of Savannah River Site reactors includes expansion joints utilizing a thin-wall bellows. While successfully used for over thirty years, an occasional replacement has been required because of the development of small, circumferential fatigue cracks in a bellows convolute. One such crack was recently shown to have initiated from a weld heat-affected zone liquation microcrack. The crack, initially open to the outer surface of the rolled and seam welded cylindrical bellows section, was closed when cold forming of the convolutes placed the outer surface in residual compression. However, the bellows was placed in tension when installed, and the tensile stresses reopened the microcrack. This five to eight grain diameter microcrack was extended by ductile fatigue processes. Initial extension was by relatively rapid propagation through the large-grained weld metal, followed by slower extension through the fine-grained base metal. A significant through-wall crack was not developed until the crack extended into the base metal on both sides of the weld. Leakage of cooling water was subsequently detected and the bellows removed and a replacement installed.

Jet pumps, which have no moving parts, provide a continuous circulation path for a major portion of the core coolant flow in a boiling water reactor. Part of the pump is held in place by a beam-and-bolt assembly, wherein the beam is preloaded by the bolt. The Alloy X-750 beams had been heat treated by heating at 885 deg C (1625 deg F) for 24 h and aging at 705 deg C (1300 deg F) for 20 h. Jet pump beams were found to have failed in two nuclear reactors, and other beams were found to be cracked. Investigation (visual inspection, metallurgical examination, tension testing, and simulated service testing in oxygenated water) supported the conclusion that intergranular SCC under sustained bending loading was responsible for the failure. The location of the cracking was consistent with the results of stress analysis of the part. Recommendations included either replacing the beams, reheat treatment, or preload reduction.

An intergranular stress-corrosion cracking failure of 304 stainless steel pipe in 2000 ppm B as H3BO3 + H2O at 100 deg C was investigated. Constant extension rate testing produced an intergranular type failure in material in air. Chemical analysis was performed on both the base metal and weld material, in addition to fractography, EPR testing and optical microscopy in discerning the mode of failure. Various effects of Cl-, O2 and MnS are discussed. Results indicated that the cause of failure was the severe sensitization coupled with probable contamination by S and possibly by Cl ions.

A metallurgical failure analysis was performed on pieces of the cracked vent header pipe from the Edwin I. Hatch Unit 2 Nuclear power plant. The analysis consisted of optical microscopy, chemical analysis, mechanical Charpy impact testing, and fractography. It was found that the material of the vent header met the mechanical and chemical properties of ASTM A516 Grade 70 carbon-manganese steel material and microstructures were consistent with this material. Fracture faces of the cracked pipe were predominantly brittle in appearance with no evidence of fatigue contribution. The NDTT (Nil ductility Transition Temperature) for this material was approximately -51 deg C (-60 deg F). The fact that the material's NDTT was significantly out of the normal operating range of the pipe suggested an impingement of low temperature nitrogen (caused by a faulty torus inerting system) induced a thermal shock in the pipe which, when cooled below its NDTT, cracked in a brittle manner.

A rupture of a thirty-year-old U-tube on a steam generator for a closed-cycle pressurized-water nuclear power plant occurred, resulting in limited release of reactor water. A typical tube bundle can be over 9 m (30 ft) tall and 3 m (10 ft) in diam with over 3,000 22-mm (7/8-in.) diam Inconel Alloy 600 tubes. Tube support plates (TSP) separate the tubes and allow flow of the heating water/steam. Inconel Alloy 600 is susceptible to intergranular stress-corrosion cracking over time, so investigation included review of operational records, maintenance history, and procedures. It also included FEA (thermal gradients, nonlinear material behavior, residual stress, changes in wall thickness during the formation of U-bends, and TSP distortions near the ruptured tube) of three-dimensional solid models of the U-tubes. The conclusion was that distortion of the TSPs and resulting “pinching” of the U-tubes, combined with the operational stresses, caused high stresses at the location where the tube cracked. The stresses were consistent with those required to initiate and propagate a longitudinal crack.

This paper describes the analysis of the failure of a Zr-2.5Nb pressure tube in a CANDU reactor. The failure sequence was established as: (1) the existence of an undetected manufacturing flaw in the form of a lamination, (2) in-service development of the flaw by oxidation of the lamination, (3) delayed hydride cracking, which extended the flaw through the wall of the tube, resulting in leakage, and (4) rupture of the tube by cold pressurization while the reactor was shut down. The comprehensive failure analysis led to a remedial action plan that permitted the reactor to be returned to full-power operation and ensured a low probability of a similar occurrence for all CANDU reactors.

The accident at Three Mile Island Unit No. 2 on 28 March 1979 was the worst nuclear accident in US history. By Jan 1990, it was possible to electrochemically machine coupons from the lower head using a specially designed tool. The specimens contained the ER308L stainless steel cladding and the A533 Grade B plate material to a depth of about mid-wall. The microstructures of these specimens were compared to that of specimens cut from the Midland, Michigan reactor vessel, made from the same grade and thickness but never placed in service. These specimens were subjected to known thermal treatments between 800 and 1100 deg C for periods of 1 to 100 min. Microstructural parameters in the control specimens and in those from TMI-2 were quantified. Selective etchants were used to better discriminate desired microstructural features, particularly in the cladding. This report is a progress report on the quantification of changes in both the degree of carbide precipitation and delta ferrite content and shape in the cladding as a function of temperature and time to refine the estimates of the maximum temperatures experienced.

Nuclear power plants typically experience two or three high-cycle fatigue failures of stainless steel socket-welded connections in small bore piping during each plant-year of operation. This paper discusses fatigue-induced failure in socket-welded joints and the strategy Texas Utilities Electric Company (TU Electric) has implemented in response to these failures. High-cycle fatigue is invisible to proven commercial nondestructive evaluation (NDE) methods during crack initiation and the initial phases of crack growth. Under a constant applied stress, cracks grow at accelerating rates, which means cracks extend from a detectable size to a through-wall crack in a relatively short time. When fatigue cracks grow large enough to be visible to NDE, it is likely that the component is near the end of its useful life. TU Electric has determined that an inspection program designed to detect a crack prior to the component leaking would involve frequent inspections at a given location and that the cost of the inspection program would far exceed the benefits of avoiding a leak. Instead, TU Electric locates these cracks by visually monitoring for leaks. Field experience with fatigue-induced cracks in socket-welded joints has confirmed that visual monitoring does detect cracks in a timely manner, that these cracks do not result in catastrophic failures, and that the plant can be safely shut down in spite of a leaking socket-welded joint in a small bore pipe. Historical data from TU Electric and Southwest Research Institute are presented regarding the frequency of failures, failure locations, and the potential causes. The topics addressed include 1) metallurgical and fractographic features of fatigue cracks at the weld toe and weld root; 2) factors that are associated with fatigue, such as mechanical vibration, internal pulsation, joint design, and welding workmanship; and 3) implications of a leaking crack on plant safety. TU Electric has implemented the use of modified welding techniques for the fabrication of socket-welded joints that are expected to improve their ability to tolerate fatigue.

Several mercury diffusion pump stages in the Tritium Purification process at the Savannah River Site (SRS) have been removed from service for scheduled preventive maintenance. These stages have been examined to determine if failure has occurred. Evidence of fatigue around the flange portion of the pump has been seen. In addition, erosion and cavitation inside the throat of the venturi tube and corrosion on the other surface of the venturi tube has been observed. Several measures are being examined in an attempt to improve the performance of these pumps. These measures, as well as the noted observations, are described. Six stages [two machined (MP) and four electron beam (EB) welded] from the mercury diffusion pumps operating in the Tritium Purification process at SRS have been analyzed to determine their condition after nine months of usage. Several cracks were found around the necked region of the two MP stages. The EB welded stages, however, seemed to perform better in service; only two of four stages showed cracking. The cracking is caused by fatigue that has been enhanced by high stresses and tritium in the flange area. The EB welded stage appears to be a step in the right direction. Since the EB weld is a shrink fit, the surface is in compression, thereby eliminating crack propagation. In addition, shot peening has been employed to produce a compressive material surface since fatigue usually originates at the surface. Pitting was observed down the throat of the venturi. This pitting was caused by cavitation and erosion along the length of the venturi tube. Corrosion and pitting was seen on the exterior walls of the diffuser tubes. Stress-corrosion cracks were observed emanating from these corrosion pits. The corrosion likely occurred from the chloride ions present in the process cooling water. Shot peening is now being used in an attempt to place the outside of the diffuser tube in compression to eliminate the stress-corrosion cracking.

Two AISI type 316 stainless steel components intended for use in a reducer section for sodium piping in a fast breeder test reactor were found to be severely corroded—the first soon after pickling, and the second after passivation treatments. Metallographic examination revealed that one of the components was in a highly sensitized condition and that the pickling and passivation had resulted in severe intergranular corrosion. The other component was fabricated from thick plate and, after machining, the outer surface represented the transverse section of the original plate. Pickling and passivation resulted in severe pitting because of end-grain effect. Strict control of heat treatment parameters to prevent sensitization and modification of pickling and passivating conditions for machined components were recommended.

A service water pump in a nuclear reactor failed when its shaft gave way. The fracture originated in the threaded portion of the sleeve nut on the drive-end side of the shaft. Results of the failure analysis showed that the cracking initiated at the thread root as a result of corrosion fatigue. Crack propagation occurred either by corrosion or mechanical fatigue. Evidence was found indicating high rotary bending stresses on the shaft during operation. The nonstandard composition of the En 8 steel used in the shaft and irregular maintenance reduced the life of the shaft. Recommendations included use of a case-hardened En 8 steel with the correct composition and regular maintenance of the pump.

Nickel anodes failed in several electrolysis cells in a heavy-water upgrading plant. Dismantling of a cell revealed gouging and the presence of loosely attached black porous masses on the anode. The carbon steel top, plate was severely corroded. An appreciable quantity of black powder was also present on the bottom or the cell. SEM/EDX studies of the outer and inner surfaces of the gouged anode showed the presence of iron globules at the interface between the gouged and the unattacked anode. The chemical composition of the black powder was determined to be primarily iron. Cell malfunction was attributed to the accelerated dissolution of the carbon steel anode top, dislodgment of grains from the material, and subsequent closing of the small annular space between the anode and the cathode by debris from the anode top. Cladding of the carbon steel top with a corrosion-resistant material, such as nickel, nickel-base alloy, or stainless steel, was recommended.

Cracking occurred in type 304L stainless steel sheaths on nichrome wire heaters that had been in storage for about 5 years in a coastal atmosphere. The cracks were discovered when the heater coils were removed from storage in their original polyethylene packing materials and straightened for use. Fractography established that fracture occurred by stress-corrosion cracking. The cracks originated at rusted areas on the cladding that occurred under iron particles left on the surface during manufacture. High hardness values indicated that solution annealing following cold working had not been carried out as specified. It was recommended that the sheathing material be fully annealed and that the outer surface be pickled and passivated.

Pressure testing of a batch of AISI type 316L stainless steel thermowells intended for use in a nuclear power-plant resulted in the identification of leakage at the tips in 20% of the parts. Radiography at the tip region of representative thermowells showed linear indications along the axes. SEM examination revealed the presence of longitudinally oriented nonmetallic inclusions that were partly retained and partly dislodged. Electron-dispersive x-ray analysis indicated that the inclusions were composed of CaO. Based on the overall chemistry of the inclusion sites, the source of the CaO was determined to be slag entrapment during the steel making process. It was recommended that the thermowell blanks be ultrasonically tested prior to machining and that the design be modified to make internal pressurization possible.

Three ASME SA106 grade B carbon steel feed water piping reducers from a pressurized water reactor showed indications of flaws near welds during ultrasonic testing. Further examination and testing indicated that the cracks resulted from a low-cycle corrosion fatigue phenomenon.

Various aluminum bronze valves and fittings on the essential cooling water system at a nuclear plant were found to be leaking. The leakage was limited to small-bore socket-welded components. Four specimens were examined: three castings (an ASME SB-148 CA 952 elbow from a small-bore fitting and two ASME SB-148 CA 954 valve bodies) and an entire valve assembly. The leaks were found to be in the socket-weld crevice area and had resulted from dealloying. It was recommended that the weld joint geometry be modified.

Leaks developed at random locations in aluminum brass condenser tubes within the first year of operation of a steam condenser in a nuclear power plant. One failed tube underwent scanning electron microscopy surface examination and optical microscope metallography. It was determined that the tube failed from crevice corrosion under seawater deposits that had formed on the inner surface. Mechanical cleaning of the condenser tubes every 6 months and installation of intake screens of smaller mesh size were recommended.