Environmentally assisted cracking (EAC) susceptibilities of Alloys 600, 690, and their weld metals with and without weld defects were benchmarked in pure and doped hydrogenated steam at 400°C and pressurized water reactor (PWR) water at 360°C on four-point bend (0.35% to 1.1% strain) and U-bend specimens (5% strain). Contrary to the expectations, no EAC initiation from existing weld defects was observed. As expected, Alloy 600 and its weld metals were much more prone to EAC initiation than Alloy 690 and its weld metals. Oxide structures remain the same in doped and pure steam and PWR water, although the environment has an impact on the EAC initiation times and oxidation rates. Because of the similar oxide structures, the EAC mechanism is assumed to be the same in all studied environments.

Alloys 182 and 82 (UNS W86182(1) and N06082, respectively) are used extensively in light water reactors (LWRs) in dissimilar metal welds, typically when ferritic steel and austenitic stainless steel or nickel-based alloy are joined. Typical applications in pressurized water reactors (PWRs) are Alloy 600 (UNS N06600) vessel head penetrations (VHPs), pressurizer penetrations, heater sleeves, instrument nozzles and reactor pressure vessel (RPV), and steam generator (SG) inlet and outlet nozzles. The same alloys are widely used in boiling water reactors (BWRs). In general, the performance of Alloys 182 and 82 in PWRs has been good. However, several cases of leaks caused by primary water stress corrosion cracking (PWSCC) of Alloys 182/82 have been reported at the locations mentioned above.1 

An example of PWSCC of nickel-based metals is Bugey 3. In 1991, a leakage was discovered on one head penetration from Bugey 3 pressure vessel. Investigations confirmed the presence of a through-wall crack resulting from PWSCC of Alloy 600 base metal and Alloy 182 weld metal. Between November 2000 and March 2001, leaks were also discovered from reactor vessel penetrations at Oconee-1 and -3 and ANO-1 plants. The cracking at Oconee-1 appeared to have initiated in the J-groove Alloy 182 weld. In most cases, the initiation has taken place in Alloy 600. Since that time, non-destructive examinations (NDEs) have revealed that several VHPs from various vessel heads have also suffered SCC. NDEs performed in a number of plants indicated (in 2002) that approximately 1.25% of the penetrations inspected in the United States and 6.5% of the penetrations inspected in France were affected by PWSCC.2-3 

The most severe case of RPV head penetration cracking in PWRs observed so far is the Davis-Besse case. In February 2002, severe corrosion of the reactor vessel head steel was observed at Davis-Besse (commercial start of operation in 1978). Investigations revealed axial indications in these nozzles. Crack initiation had occurred in the Alloy 600 penetration tube, and the crack continued into the Alloy 182 weld metal, resulting in a leak in one nozzle. The leaking nozzle caused corrosion of the RPV steel down to the stainless steel cladding within an area of ∼106 cm2. Cracking continued into the stainless steel cladding to an extent depending on the local thickness of the cladding, and seemed to stop at a level leaving a ligament with a thickness of ∼3.5 mm.4-6  The incident has resulted in intense investigations concerning the failure as well as extensive inspections of RPV heads.

Several cracking cases have been observed also in PWR pressure vessel to piping nozzles. Cracking and leakage took place in V.C. Summer hot leg nozzle weld in 2000.7  The reactor vessel nozzle was welded to Type 316 stainless steel (SS; UNS S31600) piping by Alloy 82 weld metal and Alloy 182 buttering. Investigation of the leakage location resulted in the conclusion that the cracking in Alloy 182 cladding/buttering and Alloy 82 weld metal took place by PWSCC mechanism. High residual stresses resulting from multiple weld repairs, hot cracking, and heavy grinding at the ID surface are believed to have contributed to the crack initiation. Similar nozzle to safe end cracking was observed in Ringhals 3 and 4 PWRs in 2000. In the Ringhals case, no leakage took place, but the cracks were detected by NDE. Defects produced during fabrication (e.g., hot cracks) and increased stress levels resulting from repair welding (Unit 4) may have been important factors for crack initiation.8-9 

Many other cracking incidents have been reported since then, also in other locations where Alloy 182 and/or 82 have been used in PWRs, and both weld metals have been affected, although cases of Alloy 82 cracking are scarce.10-12 

The main reasons why Alloy 600 and corresponding weld metals were chosen for application in nuclear power plants were extensive tests that showed that Alloy 600 was less susceptible to chloride-induced SCC than austenitic stainless steels. The tests had shown that the alloy was also generally SCC resistant.13 

Many of the reactor pressure vessel heads that had Alloy 600/182 penetrations have been replaced with heads having Alloy 690/52 (UNS N06690/N14052) (or 152 [UNS W86152]) penetrations and many other dissimilar metal welds (DMWs) are repaired by Alloy 52 or 152 or replaced.10,12,14 

The first case of intergranular cracking in BWRs in a safe-end made of Alloy 600 was observed in 1978 in Duane Arnold (commercial start in 1974). The crack had initiated in the heat affected zone of Alloy 600 and penetrated into Alloy 182 weld metal causing a leak.15  Subsequently, several cracks in DMWs have been found in BWRs in several different locations.16-17 

The crack initiation susceptibilities of Alloys 600, 690, 82, 182, 52, and 152 have been studied extensively in simulated PWR water at plant relevant temperatures, in simulated PWR water at elevated temperatures, and in hydrogenated steam at 380°C to 500°C in order to reduce testing time.1,18-24 

According to Economy and Jacko,25  the initiation time of low potential stress corrosion cracking (LPSCC) follows monotonic 1/T dependence from hydrogenated steam to water. This suggests that the mechanistic processes in water and steam are the same, at least for pure water. It has to be taken into account that the steam pressure has an influence on the corrosion mechanism and possibly also on the crack initiation time in hydrogenated steam. According to Guan and Macdonald,26  the corrosion process at lower steam densities is dominated by direct chemical reaction between the metal and the reactive species and at higher densities by electrochemical mechanisms involving partial cathodic and anodic charge transfer processes.

As in actual PWR temperatures, the crack initiation may take years or decades, and accelerated laboratory tests have often been conducted in high-temperature steam. Jacko, et al.,24  performed accelerated tests in hydrogenated 400°C steam doped with fluoride, chloride, and sulfate anions on Alloy 52M (UNS N06054) welds prepared to simulate Ringhals 4 field repairs of the reactor pressure vessel outlet nozzle. Comparison tests were performed on Alloy 182 weld metal and Alloy 600 base metal. The tests were done using bolt loaded 4-point bend specimens. Jacko, et al.,24  did not observe any cracking in Alloy 52M specimens within the total exposure time of 2,051 h which represents 45 effective full-power years (EFPYs) of the Ringhals 4 outlet nozzle. On the other hand, SCC initiated in Alloy 182 weld and Alloy 600 base metal specimens in less than one fifth of the total time that the Alloy 52M specimens were exposed. Analysis indicated that the degradation observed after 89,000 effective full-power hours (EFPHs) at 317°C in Ringhals 4 was duplicated with 289 h exposure at 400°C to sulfate, chloride, and fluoride doped hydrogenated steam. This translates to an acceleration factor of 308. For other temperatures Jacko, et al.,24  adjusted the acceleration factor using an activation energy of Q = 55 kcal/mol (230 kJ/mol). The activation energy was estimated by assuming Arrhenius type dependence for cracking probability between Alloy 600 in Ringhals 4 and Alloy 600 in laboratory tests at 400°C in doped steam.

The approach of Jacko, et al.,24  with some modifications, was followed in EAC tests performed within three DMW research programs in Finland. In addition, similar tests were performed in pure steam and PWR water. The aims of the three studies were to fully characterize the microstructures, mechanical properties, tendency to hot cracking, and susceptibility to EAC of different DMWs and all-weld metal samples. The experimental work was realized at VTT Technical Research Centre of Finland and Helsinki University of Technology (presently part of Aalto University) in 2003 to 2014. This article concentrates on the results of the EAC initiation tests. Characterization of mechanical and microstructural properties and hot-cracking tests and some of the EAC tests of the test materials are reported in project reports and conference papers elsewhere.27-32 

The aims of the tests were:

  • to benchmark the EAC susceptibilities of different Ni-based weld metals,

  • to study EAC initiation from weld defects, and

  • to study the influence of the environment on the crack initiation time.

Test Environment

The crack initiation tests were performed in three different environments:

  • I.

    Doped steam at 400°C temperature and 150 bar (15,000 kPa) pressure. The steam was generated from deionized water doped with 30 ppm SO42– (added as sulfuric acid), 30 ppm F (added as NaF), and 30 ppm Cl (added as NaCl). Target value for H2 partial pressure was 0.56 bar (56 kPa; Ni/NiO equilibrium).

  • II.

    Pure steam at 400°C temperature and 150 bar pressure. The steam was generated from deionized water. Target value for H2 partial pressure was 0.56 bar.

  • III.

    PWR water at 360°C with 1,200 ppm B (added as H3BO3), 2.1 ppm Li (added as LiOH), and 30 cc/kg H2.

The H2 partial pressure of 0.56 bar according to the Ellingham diagram is slightly above the Ni-NiO equilibrium which at the applied temperature and steam pressure is 0.42 bar(2) (42 kPa). The steam pressure was maintained using a high performance liquid chromatography (HPLC) pump feeding replacement water to the autoclave. The H2 partial pressure was measured using a Pd-membrane sensor inside the autoclave. The H2 partial pressure was controlled by a West 6100 PID controller connected to the pressure sensor and a high-pressure solenoid valve connected to the tube between the autoclave and the hydrogen bottle. The autoclave temperature was controlled using a controller and a commercial K-type thermocouple. The autoclave with the volume of 3 L was constructed from austenitic stainless steel and its internal surfaces were coated with a titanium alloy. A schematic picture of the test configuration is shown in Figure 1.

FIGURE 1.

A schematic picture of the test configuration.

FIGURE 1.

A schematic picture of the test configuration.

Close modal

The test implementation in steam was performed as follows:

  • The specimens were washed in ethanol and demineralized water.

  • The specimens were installed into the autoclave.

  • Autoclave lid was closed.

  • Air was replaced by N2 (pressure was raised to 3 bar [300 kPa], autoclave was emptied through a valve, and this cycle was repeated at least 10 times).

  • The autoclave was heated to the test temperature in approximately 5 h.

  • The autoclave was depressurized and doped or pure water was injected until the pressure was 150 bar. This took approximately 1 h.

  • Hydrogen was injected until the partial pressure was 0.56 bar (absolute). This took also approximately 1 h.

  • 400°C/150 bar steam pressure/0.56 bar H2 pressure was maintained until the exposure time was complete.

  • The autoclave was depressurized.

  • The temperature was decreased to room temperature.

  • At the room temperature, the autoclave gases were replaced by N2, the autoclave was opened, and specimens were taken out and washed in demineralized water and ethanol.

  • Liquid penetrant test was performed for the specimens and the indications were inspected with an optical microscope.

  • The process described above was repeated on specimens without indications.

The tests in PWR water were performed in an austenitic stainless steel autoclave with a volume of 10 L. The autoclave was connected to a water re-circulation loop. The test implementation was performed as follows:

  • The specimens were washed in ethanol and demineralized water.

  • The specimens were installed into the autoclave.

  • The autoclave was filled with demineralized water.

  • The autoclave was closed.

  • Demineralized water was circulated from the water loop to the autoclave and back through a mixed bed deionizer until the conductivity was below 0.1 μS/cm.

  • Boric acid and lithium hydroxide were added to the solution (at this point the demineralizer was disconnected).

  • Oxygen was replaced with nitrogen by bubbling through the water storage container until the oxygen content was below 10 ppb.

  • Nitrogen was replaced with hydrogen the same way as oxygen until the hydrogen content at 0.75 bar (75 kPa) overpressure (1.75 bar absolute [175 kPa]) was stabilized to ∼30 cc/kg (27°C).

  • The autoclave was heated to 360°C within ∼24 h.

  • The conditions were maintained until the exposure time of 500 h was completed. During the exposure the corrosion potentials were monitored using palladium and platinum electrodes.

  • The temperature was decreased to room temperature.

  • At room temperature, the hydrogen was replaced by N2, the autoclave was opened, and the specimens were taken out and washed in demineralized water and ethanol.

  • Liquid penetrant test was performed for the specimens and the indications were inspected with an optical microscope.

  • The process described above was repeated.

Test Materials

The studied alloys were Alloy 600, its weld metals 82 and 182, and Alloy 690 with its weld metals 52 and 152. The studied weld metals were in the form of nuclear power plant DMW safe-end mock-ups or they were all-weld metal samples welded on a steel plate. Several different mock-ups were prepared within the three DMW research programs:

  • Mock-up 1: SA-508 (UNS K13502) with Type 309L and 308L overlays (UNS S30983 and S30883), buttering 182, weld 182, Alloy 600.

  • Mock-up 2: SA-508 with Type 309L and 308L overlays, buttering 182, weld 182 with 82 root weld, Type 304 SS (UNS S30400).

  • Mock-up 3: SA-508 with Type 309L and 308L overlays, weld 52, Type 316NG SS.

  • Mock-up 4: SA-508 with Type 309L and 308L overlays, weld 152, Type 316NG SS.

  • Mock-up 5: SA-508, buttering 52, buttering 152, weld 152, Alloy 690.

  • Mock-up 6: SA-508 with Type 309L and 308L overlays, narrow gap weld 52, Type 304 SS.

The cross sections of the mock-ups are shown in Figures 2(a) through (f) and the compositions of the alloys in Table 1.

TABLE 1

Chemical Compositions of the Test Materials (wt%)

Chemical Compositions of the Test Materials (wt%)
Chemical Compositions of the Test Materials (wt%)
FIGURE 2.

Cross sections of dissimilar metal weld mock-ups: (a) Mock-up 1, (b) Mock-up 2, (c) Mock-up 3, (d) Mock-up 4, (e) Mock-up 5, and (f) Mock-up 6.

FIGURE 2.

Cross sections of dissimilar metal weld mock-ups: (a) Mock-up 1, (b) Mock-up 2, (c) Mock-up 3, (d) Mock-up 4, (e) Mock-up 5, and (f) Mock-up 6.

Close modal

Specimens

In most of the cases plate-shaped specimens were cut from the root side surface of the mock-ups. From the narrow gap mock-up, specimens were cut from several depths of the weld starting from the root side of the weld. The specimens were 80 × 15 × 3 mm3 and 80 × 15 × 2.5 mm3 plates. The specimen dimensions and cutting locations are schematically shown in Figure 3. The specimens were either cut in transverse (T) or longitudinal orientation (L) with respect to the weld beads. Some of the specimens were thermally aged at 400°C for 2,000 h in order to simulate long-time operation in typical PWR conditions (weld samples of aged specimens are indicated by “(B)” in Table 2). A smooth mechanical grinding by 600 grit emery paper was performed after plane-milling (and aging, where applicable) in order to remove the roughness and to reduce the degree of cold work of the specimen surfaces. The specimens were initially bolt-loaded in 4-point bend loading. Strain was calculated from the bending radius of the 3 mm or 3.5 mm thick plate. The target total strains for the specimens were 0.35%, 1.0%, or 1.1% depending on the specimen. Later some of the specimens were further loaded to ∼5% total strain and bolt loaded to U-shape if the lower strain did not result in crack initiation regardless of a reasonably long exposure time. The U-shape was produced by pressing the specimens between a 35 mm diameter cylinder and a rubber covered r ∼ 35 mm mold. In the final phase, the specimens were bolt loaded until they yielded slightly. The loading geometries are also shown in Figure 3.

TABLE 2

Summary of the Test Results on the Four-Point Bend and U-Bend Specimens in Doped Steam, Pure Steam, and PWR Water(A)

Summary of the Test Results on the Four-Point Bend and U-Bend Specimens in Doped Steam, Pure Steam, and PWR Water(A)
Summary of the Test Results on the Four-Point Bend and U-Bend Specimens in Doped Steam, Pure Steam, and PWR Water(A)
FIGURE 3.

(a) Schematic specimen cutting location and dimensions of the most often used specimen. (b) In most of the tests, 4-point bend geometry was used. (c) In some tests, U-bend geometry was used. In 4-point bend specimens the strains varied between 0.35% and 1.1% and in U-bend specimens it was 5%. Most of the specimens had a thickness of 3.5 mm (some 3.0 mm) and specimens with hot cracks had a thickness of ∼4.5 mm.

FIGURE 3.

(a) Schematic specimen cutting location and dimensions of the most often used specimen. (b) In most of the tests, 4-point bend geometry was used. (c) In some tests, U-bend geometry was used. In 4-point bend specimens the strains varied between 0.35% and 1.1% and in U-bend specimens it was 5%. Most of the specimens had a thickness of 3.5 mm (some 3.0 mm) and specimens with hot cracks had a thickness of ∼4.5 mm.

Close modal

Some of the specimens had hot cracks on their surface. The specimens with hot cracks were produced by Varestraint testing with an augmented strain of 4% (more detailed description of the Varestraint testing can be found in Hänninen, et al.27 ). The samples were cut so that the dissimilar metal weld was transverse to the loading axis. The specimens containing hot cracks were then loaded under an optical stereomicroscope until the crack mouth opening of at least some of the cracks was in the range of 0.02 mm. Higher loading would have resulted in strain localization and kinking of the specimens. The test surface containing the hot cracks produced by Varestraint testing was not machined at any point after welding. These specimens were also thicker than the mock-up and all-weld metal specimens, i.e., in the range of 4.5 mm at the top of the weld bead.

Alloy 600 and Its Weld Metals

In doped steam tests, the first mock-up samples containing Alloy 182 and Alloy 600 strained to 1.0% to 1.1% were cracked already after the first 220 h exposure (results are summarized in Table 2). All of the Alloy 182 mock-up specimens with that strain had cracked after 750 h exposure.

Alloy 182 specimens with 0.35% strain took somewhat more time before the first cracks were observed, i.e., 743 h. After 1,460 h exposure, all of the Alloy 182 mock-up specimens with that strain had cracked. The only observed difference between the L and T orientations was that the maximum crack lengths in L orientation specimens were shorter than the maximum crack lengths in T orientation specimens. Cracks were present both in the actual weld metal and in the Alloy 182 buttering.

In addition to Alloy 182 in the specimens prepared from Mock-up 1, cracks were observed also in Alloy 600 base metal with both 0.35% and 1.0% to 1.1% strains, although the cracks were considerably smaller.

All-weld metal Alloy 82 and Alloy 182 specimens with 1.1% strain showed similar exposure-time dependence as the specimens prepared from Mock-up 1. The first cracks were observed after 240 h exposure and all of the specimens had cracks after 750 h exposure.

An example of the evolution of the cracking in all-weld metal Alloy 82 specimen in the doped steam is shown in Figure 4. In the initial stage, the weld contains weld defects, of which one is large. After 240 h exposure, some EAC cracks are already present. After 480 h exposure, the main crack has grown through the width of the specimen. Figure 5 shows an example of cracking in one of the Mock-up 1 specimens after the exposure to the doped steam. In Figure 5, a liquid penetrant test indication is seen in the middle of the Alloy 182 weld, an optical photograph showing the crack in both of the specimen halves after the crack was opened, and a scanning electron microscope (SEM) image of the crack showing its interdendritic or intergranular nature are shown.

FIGURE 4.

Liquid penetrant indications on an Alloy 82 all-weld metal specimen before exposure, after 240 h exposure, and after 480 h exposure to doped steam. Note that the large opened weld defect does not grow further and final fracture takes place in the middle of the specimen.

FIGURE 4.

Liquid penetrant indications on an Alloy 82 all-weld metal specimen before exposure, after 240 h exposure, and after 480 h exposure to doped steam. Note that the large opened weld defect does not grow further and final fracture takes place in the middle of the specimen.

Close modal
FIGURE 5.

(a) Post-test liquid penetrant indications in one of the specimens with SA-508/Alloy 182/Alloy 600 (Mock-up 1). The indication is in the middle of the Alloy 182 weld. (b) A photograph of the opened fracture surface (both halves are shown). The crack formed during the test series is seen as dark brown area on the fracture surface. (c) An SEM image from the fracture surface showing the interdendritic nature of the cracking.

FIGURE 5.

(a) Post-test liquid penetrant indications in one of the specimens with SA-508/Alloy 182/Alloy 600 (Mock-up 1). The indication is in the middle of the Alloy 182 weld. (b) A photograph of the opened fracture surface (both halves are shown). The crack formed during the test series is seen as dark brown area on the fracture surface. (c) An SEM image from the fracture surface showing the interdendritic nature of the cracking.

Close modal

The films forming inside the cracks and on the outer surfaces of the specimens were analyzed from cross sections. Figure 6 shows details of the cross section of a crack in an Alloy 182 all-weld specimen tested in the doped steam. Energy-dispersive x-ray spectroscopy (EDS) analysis indicates that there is metallic Ni film in the middle of the crack extending almost down to the crack tip. The metallic Ni deposit in the middle of the crack is surrounded on both sides by a Cr-rich oxide layer which grows into the metal matrix on the adjacent walls and extends to the crack tip. Very close to the crack tip metallic Ni was not observed; only a thin Cr-rich oxide layer was present.

FIGURE 6.

An SEM image and EDS maps of a cross section of a crack in Alloy 182 all-weld metal specimen after the exposure to doped steam.

FIGURE 6.

An SEM image and EDS maps of a cross section of a crack in Alloy 182 all-weld metal specimen after the exposure to doped steam.

Close modal

Metallic Ni deposits are also forming in the fine scale inside the Cr-rich oxide films on the outer surface of the specimens (Figure 7). Additionally, on the outer surface larger metallic Ni deposit particles are present. Inside the Cr-rich oxide film metallic Ni deposits form spike-like structures in areas of deeper penetration of the oxide film into the metal matrix. The oxide structures were similar in Alloy 82 to those of Alloy 182 after the exposure to the doped steam.

FIGURE 7.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 82 weld metal after 2,580 h exposure to pure steam and subsequent 998 h exposure to doped steam (1.1% strain).

FIGURE 7.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 82 weld metal after 2,580 h exposure to pure steam and subsequent 998 h exposure to doped steam (1.1% strain).

Close modal

Exposure of specimens with 1.1% strains in pure steam for 1,500 h (Alloy 600) or 2,580 h (Mock-up 2 as well as all-weld metal 182 and 82 specimens) did not result in cracking. Further exposure of specimens strained to 5.0% resulted in small cracks in Alloy 600, Mock-up 2, and all-weld metal 182 specimens during 1,000 h exposure.

Pure steam exposure resulted in similar oxide structures in both alloys as the doped steam exposure except the oxidation and Ni deposits are less pronounced (Figure 8) and the EAC cracks are much narrower. The compositions of the oxide films within the cracks and on the outer surface of Alloy 600 are quite similar to those found from the Alloy 82/182 samples.

FIGURE 8.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 82 weld metal after 2,580 h exposure to pure steam (1.1% strain) and subsequent 1,000 h exposure to pure steam (5% strain).

FIGURE 8.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 82 weld metal after 2,580 h exposure to pure steam (1.1% strain) and subsequent 1,000 h exposure to pure steam (5% strain).

Close modal

Alloy 600 base metal and all-weld metal 182 and 82 specimens were exposed to PWR water with the strain of 1.1% for 1,000 h (Alloy 600 specimens) or 2,500 h (all-weld metal 182 and 82 specimens) and further with the strains of 5.0% for 1,000 h. No cracks were observed after the exposures. The oxide films on the specimens were very thin and had, locally, thick magnetite deposits originating from the surfaces of the water recirculating loop and autoclave. A more detailed analysis of the oxide films is given in an earlier publication.29 

Alloy 690 and Its Weld Metals

Mock-up specimens containing Alloy 52 or Alloy 152 (Mock-ups 3 through 6) and all-weld metal 52 or 152 specimens did not crack within the maximum exposure time of 2,178 h in the doped steam tests. No cracking was observed in Alloy 690 base metal, either.

Exposure to pure steam of Mock-up 6 specimens with the strain of 1.1% (for 2,580 h) and with 5.0% (for 1,000 h) did not result in cracking. Mock-up 3 through 5 or all-weld metal 152 and 52 specimens were not exposed to pure steam and none of the Alloy 690 or its weld metal specimens were exposed to PWR water because no EAC was observed in the doped steam tests in a reasonable time.

Alloys 52 and 152 had thinner oxide layers than Alloy 600 or its weld metals after the exposure to the doped and pure steam. Figures 9 and 10 show EDS maps of the oxide layers on the outer surface of Alloy 52 weld metal samples after exposure to both environments. The oxide scale is somewhat thicker after the exposure to doped steam. However, in both cases, the oxide scale composition is very similar and the oxide scales consist of thin, external layer of Cr-rich oxides with small, metallic Ni deposits inside the Cr-oxide layer.

FIGURE 9.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 52 weld metal after 2,580 h exposure to pure steam and subsequent 998 h exposure to doped steam (1.1% strain).

FIGURE 9.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 52 weld metal after 2,580 h exposure to pure steam and subsequent 998 h exposure to doped steam (1.1% strain).

Close modal
FIGURE 10.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 52 weld metal after 2,580 h exposure to pure steam (1.1% strain) and subsequent 1,000 h exposure to pure steam (5% strain).

FIGURE 10.

An SEM image and EDS maps of a cross section of the oxide scale on the surface of Alloy 52 weld metal after 2,580 h exposure to pure steam (1.1% strain) and subsequent 1,000 h exposure to pure steam (5% strain).

Close modal

Environmentally Assisted Cracking Initiation from Weld Defects

All weld combinations except Mock-up 4 were exposed to the doped steam after hot cracks were produced by the Varestraint testing. No crack extension was observed on the outer surfaces of any of the hot cracked specimens after the 2,178 h total exposure time.

After the test campaign, the largest hot cracks and other weld defects were opened for fractographic investigation. Also, cross sections were prepared for metallographic investigations for determination of possible environmentally assisted crack extension under the specimen surfaces initiated from the weld defects or hot cracks.

Porous metal debris that had dropped mainly to the crack tip area was observed inside the cracks in all hot cracked specimens. The debris was apparently related to the final melt that solidified inside the opening hot cracks. The chemical composition of the debris was almost the same as the composition of the weld metal. Only a thin Cr-rich oxide was present on the debris and the hot crack surfaces. In addition to hot cracking, clear liquation cracking had formed under the weld beads where the hot cracking had taken place. Post-test fractography of the opened hot cracks of all specimens revealed that no EAC initiation had taken place in the hot crack tips. Some of the hot cracks in Alloy 152 ended in large Laves phase formation. No EAC was observed in those locations, either. The characterization of the hot cracks is described more closely in previous publications.28,32 

Liquid penetrant indications caused by weld defects were observed during the test campaign in all of the weld combinations, and also in specimens with no hot cracks. However, no indications that the weld defects had extended, or acted as initiation sites for EAC, during the exposure in the doped steam were observed on the specimen surfaces. A weld defect seen on the surface of one of the Mock-up 3 specimens is shown in Figure 11, along with a cross section of it. The defect is in Alloy 52. The specimen was exposed to doped steam for 2,178 h with 1.1% strain.

FIGURE 11.

(a) An SEM image of a weld defect in Alloy 52 of the mock-up 3 specimen and (b) a photograph of the cross section of the defect.

FIGURE 11.

(a) An SEM image of a weld defect in Alloy 52 of the mock-up 3 specimen and (b) a photograph of the cross section of the defect.

Close modal

In the earlier studies27,30  in the doped steam conditions, Alloy 182 was apparently the most susceptible of the studied welds. However, when the number of tested specimens increased in the subsequent projects, the all-weld metal specimens of Alloy 182 and 82 seemed to be equally susceptible. No difference can be seen between the Alloy 182 in Mock-up 1 and the all-weld metal 182 regarding the crack initiation time, either. Alloy 600 base metal in Mock-up 1 seems to be less susceptible based on the smaller cracks. However, this may result from more homogeneous microstructure of Alloy 600 compared to the heterogeneous structure formed during solidification of the welds. Aging at 420°C for 2,000 h does not seem to affect the cracking susceptibility in the doped steam test.

Fractography of Alloy 182 and 82 all-weld metal specimens after crack initiation tests in doped steam shows that the outer surfaces and fracture surfaces are covered with a Cr-rich oxide layer. Metallic Ni deposits or a NiO layer formed on the top of the Cr-rich oxide layer, and covered the fracture surface close to the outer surface in the specimens that took a shorter time to crack. Metallic Ni deposits also form on the outer surface of the specimens, and on a finer scale, inside the outer surface oxide film. Close to the crack tip, NiO and Fe-rich spinel oxide particles are observed when metallic Ni is present, and the fracture surface is covered by a thin Cr-rich oxide layer up to the crack tip. This result is backed up by the earlier studies (see Hänninen, et al.27-28 ).

In recent studies by Persaud, et al.,33-36  and Bertali, et al.,37-38  the EAC mechanism in Alloy 600 is attributed to internal oxidation, embrittlement caused by it, related Ni expulsion, and/or local stress caused by grain boundary migration. Internal oxidation and Ni expulsion have been observed in tests in hydrogenated steam. In PWR water, internal oxidation has been observed but no Ni expulsion to the alloy surface. Ni expulsion is supposed to take place as a result of compressive stresses caused by the internal oxidation. Lattice diffusion is too slow and the Ni diffusion outward is supposed to take place through short-circuit pathways like oxide/metal interfaces, dislocation pipe, and grain boundaries. At or below 360°C in PWR water, these mechanisms are slower than between 400°C and 500°C and that may be the reason why it has not been observed in PWR tests.33 

The mechanism of cracking of Ni-based alloys in doped steam test has been explained also by the selective dissolution–vacancy creep (SDVC) model for EAC.31,39  In the SDVC model, cracking takes place by a selective oxidation mechanism where Cr and Fe are hydrolyzed within the crack producing excess hydrogen. Ni is transferred through the oxide film and deposited on the film in metallic form. This can take place even when the oxygen partial pressure in the ambient environment is slightly above the Ni/NiO equilibrium level. Vacancy injection into the base material is expected to take place and to enhance the creep rate at the crack tip. The mechanistic processes in steam and high-temperature water are thought to be the same.

In this work, Ni expulsion was observed both on the outer surfaces of the specimens and within the larger EAC cracks. There are no reasons why the above-mentioned mechanisms would not operate also within the cracks in addition to the outer surfaces.

The specimen surfaces and cracks in the Mock-up 1 and Alloy 82 and 182 specimens that took a longer time to crack did not contain metallic Ni, but NiO was observed instead. This indicates that those specimens were exposed to an environment where the oxygen partial pressure was above the Ni/NiO equilibrium. On the other hand, the oxygen partial pressure may have been high only in the end of the exposure(s) and the formed metallic Ni may have oxidized at that point.

The oxidation rates of Alloys 690, 52, and 152 are slower than that of Alloy 600 or its weld metals likely because the higher Cr content enhances the Cr flux to the surface. Because of the higher flux, Cr forms a protective film quickly on the grains and above the grain boundaries, thus preventing the oxygen inward flux and Ni expulsion. However, some metallic Ni particles were observed on the surfaces, which indicate that at least some internal oxidation has taken place. This means that the three alloys can be susceptible to EAC in long-term exposure.

Crack extension was observed neither in all-weld metal specimens nor in mock-up samples containing hot cracks or other types of weld defects after the 2,178 h exposure to doped steam. Only a little oxidation was observed on the hot crack surfaces. Similar selective oxidation was not observed as on the EAC crack surfaces. Based on these results, the local microchemistry and microstructures on the dendrite boundaries ahead of the hot crack tips may not be inherently EAC susceptible. In hot cracks in Nb-containing Ni-based alloys, Nb-rich phases are typically present. The adjacent austenitic matrix is also enriched in Nb up to 10 wt%. No Nb is present in Alloy 52 in which the hot cracking is associated with TiN phase.27,30  In addition, segregation, second phase particles, and high dislocation densities are present at the dendrite boundaries. However, it is not known if specimen loading after the hot cracks or other defects have formed resulted in excessive blunting and, thus, prevented the initiation of SCC from the defects.

The Ni expulsion and Cr/Fe oxidation are somehow influenced by the impurities in the doped steam tests, as can be seen when Figures 7 and 8 are compared. The mechanism is not exactly clear at this point and should be studied more closely. One possible explanation is that the impurities reduce the protectiveness of the oxide layer. In that case, a question arises: is the doped hydrogenated steam the correct environment to study PWSCC resistance of the Ni-based alloys or should pure hydrogenated steam be used instead? The difference between the crack initiation times in pure steam and steam with impurity ions was noted already by Tsuruta, et al.,40  in 1994.

In general, in the hydrogenated steam test the oxides forming a two-layer oxide structure on the crack walls and outer surface of the specimens are similar to those forming in normal reactor conditions.41-44  On the other hand, in the hydrogenated steam test the phases are larger, thicker, and more clearly separated.

  • Alloy 690 and its weld metals Alloys 52 and 152 are clearly more resistant to stress corrosion cracking in hydrogenated steam than Alloy 600 and its weld metals Alloys 82 and 182.

  • SCC and oxidation mechanisms seem to be similar in doped hydrogenated steam, pure hydrogenated steam, and based on literature, also in PWR primary water.

  • Impurities in doped steam test have an impact on the SCC times and oxidation rates. It is questionable whether doped steam or pure steam should be used in order to evaluate activation energy of Ni-based alloys for prediction of their SCC behavior in PWR primary water.

  • Hot cracks and other typical weld defects seem not to be preferable SCC initiation sites. On the other hand, crack tip blunting may have influenced the crack initiation in the experiments of this work.

(1)

UNS numbers are listed in Metals and Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International.

(2)

Source: HSC Chemistry software developed by Outokumpu Ltd.

Trade name.

The three projects within which the experimental work was performed were funded by Tekes, Teollisuuden Voima Oyj (TVO), Fortum Power and Heat Oy, Vattenfall Aktiebolag, OKG Aktiebolag, Fennovoima Oy, VTT Technical Research Centre of Finland, and Aalto University School of Engineering, which is gratefully acknowledged. Additional acknowledgments belong to STUK – Radiation and Nuclear Safety Authority of Finland for participating to the projects as a monitoring party.

1.
R.S.
Pathania
,
A.R.
McIlree
,
J.
Hickling
,
“Overview of Primary Water Cracking of Alloys 182/82 in PWRs,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
2.
F.
Vaillant
,
P.
Moulart
,
J.M.
Boursier
,
C.
Amzallag
,
J.
Daret
,
“Crack Growth Rates in Thick Materials of Alloy 600 and Weld Metals of Alloy 182 in Laboratory Primary Water Comparison with Field Experience,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
3.
G.V.
Rao
,
R.J.
Jacko
,
A.R.
McIlree
,
“An Assessment of the CRDM Alloy 600 Reactor Vessel Head Penetration PWSCC Remedial Techniques,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
4.
H.
Xu
,
S.
Fyfitch
,
J.
Hyres
,
“Laboratory Investigations of the Stainless Steel Cladding of the Davis-Besse Reactor Vessel Head,”
International PWSCC of Alloy 600 Conference and Exhibit Show
,
held
March
7-10
,
2005
(
Palo Alto, CA
:
EPRI
,
2005
).
5.
H.
Xu
,
S.
Fyfitch
,
J.W.
Hyres
,
“Laboratory Investigation of PWSCC of CRDM Nozzle 3 and Its J-Groove Weld on the Davis-Besse Reactor Vessel Head,”
Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Systems—Water Reactors
,
ed.
T.R.
Allen
,
P.J.
King
,
L.
Nelson
(
Pittsburgh, PA
:
TMS
,
2005
),
p
.
833
842
.
6.
“Davis-Besse Reactor Pressure Vessel Head Degradation—Overview, Lessons Learned, and Actions Based on Lessons Learned,”
U.S. Nuclear Regulatory Commission, NUREG/BR-0353, Revision 1
,
2008
,
p
.
11
.
7.
G.
Rao
,
G.
Moffatt
,
A.
McIlree
,
“Metallurgical Investigation of Cracking in the Reactor Vessel Alpha Loop Hot Leg Nozzle to Pipe Weld at the V.C. Summer Station,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
8.
A.
Jenssen
,
K.
Norrgård
,
C.
Jansson
,
J.
Lagerström
,
G.
Embring
,
P.
Efsing
,
“Structural Assessment of Defected Nozzle to Safe-End Welds in Ringhals 3 and 4,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
9.
A.
Jenssen
,
K.
Norrgård
,
J.
Lagerström
,
G.
Embring
,
D.
Tice
,
“Assessment of Cracking in Dissimilar Metal Welds,”
Proceedings of the 10th International Conference on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors
,
ed.
G.S.
Was
(
Houston, TX
:
NACE International
,
2001
).
10.
S.S.
Hwang
,
J. Nucl. Mater.
443
(
2013
):
p
.
321
330
.
11.
S.S.
Hwang
,
Y.S
Lim
,
S.W.
Kim
,
Corrosion
69
(
2013
):
p
.
1051
1059
.
12.
“U.S. Plant Experience with Alloy 600 Cracking and Boric Acid Corrosion of Light-Water Reactor Pressure Vessel Materials,”
U.S. Nuclear Regulatory Commission, NUREG-1823
,
2005
,
p.
79
.
13.
J.A.
Gorman
,
Corrosion
71
(
2015
):
p
.
1414
1433
.
14.
P.L.
Andresen
,
Corrosion
69
(
2013
):
p
.
1024
1038
.
15.
V.
Pasupathi
,
J.S.
Perrin
,
D.R.
Farmelo
,
G.P.
Smith
,
S.
Weis
,
Trans. Am. Nucl. Soc.
34
(
1980
):
p
.
232
233
.
16.
“IGSCC in BWR Plants,”
U.S. Nuclear Regulatory Commission, U.S.NRC Information Notice No. 84-41
,
1984
,
p.
2
.
17.
“Reactor Coolant Boundary Leakage Attributable to Propagation of Cracking in Reactor Vessel Nozzle Welds,”
U.S. Nuclear Regulatory Commission, U.S.NRC Information Notice 2004-08
,
2004
,
p.
5
.
18.
Y.S.
Lim
,
S.S.
Hwang
,
S.W.
Kim
,
H.P.
Kim
,
Corros. Sci.
100
(
2015
):
p
.
12
22
.
19.
C.
Amzallag
,
J.
Boursier
,
C.
Pages
,
C.
Gimond
,
“Stress Corrosion Life Experience of 182 and 82 Welds in French PWRs,”
Proceedings of the Fontevraud 5 Conference, Contribution of Material Investigation to the Resolution of Problems Encountered in Pressurized Water Reactors
(
Paris, France
:
SFEN
,
2002
).
20.
J.
Boursier
,
T.
Couvant
,
F.
Vaillant
,
B.
Yrieix
,
“A Review of PWSCC, Weldability and Thermal Ageing of Nickel Weld Metals in PWR Primary Water,”
International PWSCC of Alloy 600 Conference and Exhibit Show
,
held
March 7–10, 2005
(
Palo Alto, CA
:
EPRI
,
2005
).
21.
D.
Bousine
,
F.
Vaillant
,
P.
Vidal
,
C.
Gimond
,
“PWSCC Resistance of Nickel-Base Weld Metals with Various Chromium Contents,”
EPRI Workshop on PWSCC of Alloy 600 in PWRs, November 15-17
,
1994
,
p.
14
.
22.
M.
Psaila-Dombrowski
,
C.
Wade
,
J.
Sarver
,
W.
Van der Sluys
,
P.
Doherty
,
“Evaluation of Weld Metals 82, 152, 52 and Alloy 690 Stress Corrosion Cracking and Fatigue Susceptibility,”
Proceedings of the 8th International Symposium on Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
La Grange Park, IL
:
ANS
,
1997
),
p
.
412
421
.
23.
G.
Sui
,
J.M.
Titchmarsh
,
G.B.
Heys
,
J.
Congleton
,
Corros. Sci.
39
(
1997
):
p
.
565
587
.
24.
R.J.
Jacko
,
R.E.
Gold
,
A.
Kroes
,
“Accelerated Corrosion Testing of Alloy 52M and Alloy 182 Weldments,”
Proceedings of the 11th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
La Grange Park, IL
:
ANS
,
2003
).
25.
G.
Economy
,
R.J.
Jacko
,
Corrosion
43
(
1987
):
p
.
727
734
.
26.
X.
Guan
,
D.D.
Macdonald
,
“Corrosion of Austenitic Stainless Steel in Supercritical Aqueous Solutions,”
Proceedings of the 14th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
La Grange Park, IL
:
ANS
,
2009
).
27.
H.
Hänninen
,
A.
Brederholm
,
T.
Saukkonen
,
H.
Gripenberg
,
A.
Toivonen
,
U.
Ehrnsten
,
P.
Aaltonen
,
Hot Cracking and Environment-Assisted Cracking Susceptibility of Dissimilar Metal Welds
,
VTT Tiedotteita– Research Notes 2399
(
Espoo, Finland
:
VTT
,
2007
).
28.
H.
Hänninen
,
A.
Brederholm
,
T.
Saukkonen
,
M.
Ivanchenko
,
A.
Toivonen
,
W.
Karlsen
,
U.
Ehrnsten
,
P.
Aaltonen
,
Environment-Assisted Cracking and Hot Cracking Susceptibility of Nickel-Base Alloy Weld Metals
,
VTT Tiedotteita– Research Notes 2582
(
Espoo, Finland
:
VTT
,
2011
).
29.
H.
Hänninen
,
A.
Brederholm
,
T.
Sarikka
,
R.
Mouginot
,
P.
Holmström
,
A.
Saukkonen
,
P.
Toivonen
,
P.
Karjalainen-Roikonen
,
P.
Nevasmaa
,
H.
Keinänen
,
E.
Leskelä
,
M.
Ahonen
,
U.
Ehrnsten
,
P.
Aaltonen
,
Structural Integrity of Ni-Base Alloy Welds
,
VTT Technology 175
(
Espoo, Finland
:
VTT
,
2014
).
30.
H.
Hänninen
,
A.
Toivonen
,
A.
Brederholm
,
T.
Saukkonen
,
U.
Ehrnsten
,
P.
Aaltonen
,
“Environment-Assisted Cracking and Hot Cracking of Ni-Base Alloy Dissimilar Metal Welds,”
Proceedings of the 13th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
Toronto, Canada
:
CNS
,
2007
).
31.
H.
Hänninen
,
A.
Toivonen
,
T.
Saukkonen
,
A.
Brederholm
,
P.
Aaltonen
,
U.
Ehrnsten
,
“EAC Crack Initiation in Nickel-Based Dissimilar Metal Welds Using Doped Steam Test,”
Proceedings of the 14th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
La Grange Park, IL
:
ANS
,
2009
).
32.
H.
Hänninen
,
A.
Toivonen
,
A.
Brederholm
,
T.
Saukkonen
,
W.
Karlsen
,
U.
Ehrnsten
,
P.
Aaltonen
,
“Effect of Hot Cracks on EAC Crack Initiation and Growth in Nickel-Base Alloy Weld Metals,”
Proceedings of the 15th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
,
eds.
J.T.
Busby
,
G.
Ilevbare
,
P.L.
Andresen
(
Pittsburgh, PA
:
TMS
,
2011
),
p
.
197
222
.
33.
S.Y.
Persaud
,
R.C.
Newman
,
Corrosion
72
(
2016
):
p
.
881
896
.
34.
S.Y.
Persaud
,
A.
Korinek
,
J.
Huang
,
G.A.
Botton
,
R.C.
Newman
,
Corros. Sci.
86
(
2014
):
p
.
108
122
.
35.
S.Y.
Persaud
,
S.
Ramamurthy
,
R.C.
Newman
,
Corros. Sci.
90
(
2015
):
p
.
606
613
.
36.
S.Y.
Persaud
,
S.
Ramamurthy
,
A.
Korinek
,
G.A.
Botton
,
R.C.
Newman
,
Corros. Sci.
106
(
2016
):
p
.
117
126
.
37.
G.
Bertali
,
F.
Scenini
,
M.G.
Burke
,
Corros. Sci.
100
(
2015
):
p
.
474
483
.
38.
G.
Bertali
,
F.
Scenini
,
M.G.
Burke
,
Corros. Sci.
111
(
2016
):
p
.
494
507
.
39.
P.
Aaltonen
,
T.
Saario
,
P.
Karjalainen-Roikonen
,
J.
Piippo
,
S.
Tähtinen
,
M.
Itäaho
,
H.
Hänninen
,
“Vacancy-Creep Model for EAC of Metallic Materials in High Temperature Water,”
CORROSION 1996
(
Houston, TX
:
NACE
,
1996
).
40.
T.
Tsuruta
,
S.
Okamoto
,
T.
Kitera
,
Y.
Katsura
,
H.
Takamatsu
,
K.
Matsueda
,
T.
Matsunaga
,
“Effect of Impurity Ions on IGSCC Behaviour of Mill Annealed Alloy 600 Steam Generator Tubing under 400°C Steam Tests,”
Proceedings of a Conference on Steam Generators and Heat Exchangers
(
Toronto, Canada
:
CNS
,
1994
),
p
.
6.33
6.45
.
41.
S.E.
Ziemniac
,
M.
Hanson
,
Corros. Sci.
48
(
2006
):
p
.
498
521
.
42.
P.M.
Scott
,
“An Overview of Internal Oxidation as a Possible Explanation of Intergranular Stress Corrosion Cracking of Alloy 600 in PWRs,”
Proceedings of the 9th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
Pittsburgh, PA
:
TMS
,
1999
),
p
.
3
14
.
43.
L.E.
Thomas
,
S.M.
Bruemmer
,
“Insights into Environmental Degradation Mechanisms from Analytical Transmission Electron Microscopy of SCC Cracks,”
Proceedings of the 9th International Conference of Environmental Degradation of Materials in Nuclear Systems—Water Reactors
(
Pittsburgh, PA
:
TMS
,
1999
),
p
.
41
47
.
44.
F.
Scenini
,
R.C.
Newman
,
R.A.
Cottis
,
R.J.
Jacko
,
Corrosion
64
(
2008
):
p
.
824
835
.