SUMMARY OF THE LOW-ALLOY STEELS SESSION – Chaired by Hans-Peter Seifert

This session on Tuesday late afternoon consisted of three presentations on chloride and -dK/da effects on SCC and one talk about high-temperature water effects on the rapid fracture behavior of low-alloy RPV steels. Two further contributions, one in the special session on stress and strain effects and one in the summary presentation session are also dealing with EAC in LAS and are included in this session summary.

In the first talk, Martin E. Bjurman (Studsvik, Sweden) reported about a study on the effect of chloride (3 to 90 ppb) on the SCC growth in a RPV steel (Barsebäck archive material) with medium sulfur content (0.011 wt.% S) in BWR/NWC environment (500 to 2000 ppb DO) at high KI values (50-70 MPa√m) under constant K and with periodical partial un-/re-loading (PPU) in 1.5T C(T) specimens. The resulting SCC CGRs were in the expected regime and below the corresponding BWRVIP-233R2 disposition lines. Initiating and sustaining crack growth was more difficult than in previous studies at PSI and GE-GRC (with mainly high-sulfur steels) and SCC crack growth usually consisted of phases of slower and faster transient growth.

The effect of chloride (high-purity water with 2.5 to 50 ppb chloride) on the SCC crack growth behavior in a very low-sulfur RPV steel (0.001 wt.% S) steel was evaluated by constant K contour CDCB specimens in BWR/NWC environment by Yasufumi Miura (CRIEPI, Japan). Most tests started with a high chloride concentration, which was then stepwise reduced. In contrast to high-purity water, fast SCC was observed and the crack growth usually consisted on phases with slower and faster growth. In most cases, the SCC growth rates decreased with decreasing chloride content. The CGR in S-L orientation tended to be faster than that in L-S orientation. The CGRs of the specimens that started from high levels of chloride contents of 50 ppb remained high after returning to low levels of chloride or high-purity water. The SCC rates in the very low-sulfur steel are remarkable.

Both studies were in line with previous work of PSI, GEGRC, TEPCO, AREVA/Framatome GmbH or MPA and the recently developed BWRVIP-233R2 disposition lines. It is expected that the chloride effects are synergistic with steel sulfur content and other material parameters like DSA or yield stress and that the chloride-K threshold combinations with fast SCC are shifted to lower values with increasing steel sulfur content. This was also confirmed by tests in BWR/HWC environment at GE-GRC and PSI, where low-sulfur crack crevice chemistry prevailed and where the addition of 1000 ppb of chloride (as HCl) did not result in fast SCC although the chloride level in the crack-tip environment is expected to be higher than in BWR/NWC tests with 5 ppb of chloride (chloride enrichment of 30x, i.e., 150 ppb), where high-sulfur crack crevice chemistry and fast SCC prevailed.

In a related presentation, Katsuhiko Kumagai (TEPCO, Japan) discussed the effect of -dK/da on SCC rate in BWR/NWC environment with chloride (5 to 30 ppb). In core shroud support RPV attachment welds, the compressive residual stress in the RPV part results in slightly negative -dK/da in the range of 2 MPa√m/mm for SCC crack growth in the RPV. The applied -dK/da were a factor of 10 to 50 higher than in the plants and only moderately reduced the SCC growth rates. In several cases the growth rates remained very high in spite of a large -dK/da. This behavior shall be further confirmed by tests with constant displacement control. The results are in line with corresponding investigations with SS or Ni-alloys that showed strong effects of +dK/da, but little beneficial effects of -dK/da. Old tests with bolt- or wedge-loaded DCB, 1T and 2T C(T) specimens in chloride contaminated, oxygenated high-temperature water suggest that fast SCC can be also sustained under such conditions.

In the last talk of the LAS session, Zaiqing Que (PSI, Switzerland) summarized the main results of his PhD thesis on the effect of PWR and BWR environments on the fracture behavior of RPV steels. The fracture behavior of low-alloy RPV steels with different microstructures and DSA, EAC and tempering embrittlement (TE) susceptibilities in simulated LWR environments was evaluated by elastic-plastic fracture mechanics (EPFM) tests (with different temperatures and strain rates) in the upper shelf region. Only very moderate reductions in fracture initiation and tearing resistance were observed in both PWR and BWR environments, which is mainly due to the low hydrogen availability in high-temperature water. A high yield strength, DSA and TE susceptibility or an aggressive occluded crevice chemistry may cause a moderate reduction in upper shelf fracture resistance in high-temperature water environments. The materials failed by stable ductile TG tearing by microvoid coalition in air and water with additional various and varying but small amounts (a few %) of secondary cracking, macrovoids, quasi-cleavage and IG cracking in high-temperature water. No unstable cracking and no change in the dominant fracture mechanism was observed in high-temperature water. HELP and HESIV are the most probable mechanisms (minor role of HEDE). Open issues are mainly related to irradiated RPV steels and the behavior at <150°C.

Please note that two further low-alloy steel related presentations are included in the minutes (one was presented in the summary presentation session and one in the special session):

In an extended contribution for the minutes, Hans-Peter Seifert (PSI, Switzerland) discusses the relevance of homogeneous LAS specimens for the assessment of cladding penetrating RPV cracks. A recent, very sophisticated crevice chemistry model by VTT/UCTM suggests a very strong enrichment of chloride in homogenous LAS specimens due to the high Fe cation concentration, both being much higher than in a cladding penetrating crack that reaches the fusion boundary towards the LAS. It was thus argued that the homogenous LAS specimens are not representative and unduly conservative for these important crack situations and several of the conclusions in BWRVIP233R2 were thus indirectly challenged. Direct evidence from Alloy 182-RPV steel specimens is given that clearly discounts this argument. The identical SCC behavior of homogenous RPV and bimetallic Alloy 182/RPV steel specimens (GE-GRC, PSI, TEPCO) with semi-elliptical surface SCC cracks in Alloy 182 with regard to SCC growth rates and KI and chloride thresholds is the most striking argument. There is no evident reason why the behavior of the reference stainless steel cladding flaw should be fundamentally different, although SCC is more difficult in the stainless steel cladding than in the Alloy 182 weld meal. Microsampling tests show orders of magnitude lower enrichment than predicted by the model and a comparable enrichment in stainless steel and LAS. The model in its present form does not consider the dominant effect of MnS inclusions, crack growth and the interrelation between crack crevice chemistry and crack growth (and vice versa) in LAS. Furthermore, the model predicts only a small chloride enrichment, if chloride and sulfate are simultaneously present in the bulk environment. This prediction was used to check the model by tests with sulfate, chloride and sulfate/chloride combinations. The observed behavior is in contradiction to the model predictions in its present form. The observed effects of chloride on SCC is thus not an artefact of homogeneous RPV steel specimens and the results and BWRVIP-233R2 are relevant for cladding cracks also.

In the special session on stress and strain effects on EAC, Hans-Peter Seifert (PSI, Switzerland) discussed the SCC behavior of LAS in BWR environment at high KI values and gave a mechanistic explanation for this transition. The transition of slow to fast SCC at high KI values is an inherent characteristics of EAC in LAS. Violation of SSY limits is not the root cause for this transition, although it may affect SCC (e.g., due to dK/da and low-temperature creep effects). This transition is primarily related to the difficulties to sustain SCC and maintain a crack-tip strain rate at KI < ~50 MPa√m and the interrelation between da/dt ↔ dε/dt and crack crevice chemistry and vice versa. It is affected by S, DSA, yield stress, Cl–, ECP, T, dK/da, specimen size, constraints, etc. The exact location is not yet adequately characterized, but has a strong impact in flaw tolerance and safety evaluations. There is thus a clear practical need for SCC tests in the high KI range from 50 to 100 MPa√m with sufficiently large specimens and high sulfur steels to evaluate the role (and synergism) of dK/da, yield stress (weld HAZ, irradiation hardening, etc.) and DSA, sulfur content and very low chloride concentration (and other halides?) on the SCC transition behavior in this region to assess the real safety concerns with SCC.

Concluding remarks from the session chair:

High-temperature water effects on fracture behavior of RPV steels are usually moderate and do not represent a serious safety concern, as shown by the comprehensive work of Zaiqing Que, but the chloride effects on SCC were further confirmed with very low sulfur steels and persisted under -dK/da conditions also. Potential long-term effects after severe chloride transients as shown by Yasufumi Miura and previously reported by PSI further increase our concern. An even bigger concern represent the cases of fast SCC of RPV steel in high-purity water (e.g., weld HAZ with high hardness, steels with high DSA susceptibilities). The possibility of fast SCC of many cm’s per year in RPV steels at low KI values in BWR/NWC represents a serious potential safety concern, but is usually overseen or even ignored and RPV steels are often (erroneously) regarded as inherently resistant to SCC. The good service record is primarily based on the stainless steel cladding and the high EAC initiation resistance from smooth surfaces, particularly under turbulent flow conditions. In fact, very fast SCC may occur in RPV steels in case of incipient cracks under conditions that are not very far away from typical operating and material conditions. The exact threshold conditions are not sufficiently known. Although service record is excellent and SCC appears as unlikely, we currently do not know enough to qualify SCC as a neglectable residual risk, in particular, in the context of the high SCC rates and LTO and SCC DMW incidents.