NUREG-1511, Discusses NUREG-1511, Rv Status Rept, Re NRC Comprehensive Assessment of All PWR RPVs in Us Plants.Rept Identified Beaver Valley 1 & Palisades as Only Plants Projected to Potentially Exceed PTS Screening Criteria Before End
| ML18064A746 | |
| Person / Time | |
|---|---|
| Site: | Palisades, Beaver Valley  |
| Issue date: | 05/05/1995 |
| From: | Strosnider J NRC (Affiliation Not Assigned) |
| To: | Thadani A NRC (Affiliation Not Assigned) |
| References | |
| RTR-NUREG-1511 NUDOCS 9505100187 | |
| Download: ML18064A746 (8) | |
Text
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555 0001 0 5 '995 MEMORANDUM TO:
Ashok C. Thadani, Associate Director for Technical Assessment Office of Nuclear Reactor Regulation THRU:
Brian W. Sheron, Di rector (~ L.
Division of Engineering fJ T-FROM~*
~Jack R. Strosnider, Chief
SUBJECT:
Materials and Chemical Engineering Branch Division of Engineering ASSESSMENT. OF IMPACT OF INCREASED VARIABILITY IN CHEMISTRY ON.THE RT,rs VALUE OF PWR REACTOR VESSELS NUREG-1511, "Reactor Vessel Status Report* documents the staff's comprehensive assessment of all the PWR RPVs in U.S. plants. The report identified Beaver Valley 1 and Palisades as the only plants projected to potentially exceed the pressurized thermal shock (PTS) screening criteria before the end of license.
The NUREG also noted that.these results were based on the docketed information that was available at th~t-tinie and were subject to change.
Subsequent to the issuance of NUREG-1511, the licensee for Palisades submitted for staff review a revised PTS evaluation.
As part of the revised evaluation, the licensee submitted chemistry data from welds in its retired steam generators that were fabricated using the same procedure and weld wire heat number as the limiting welds in its reactor vessel. These data indicated large variability in the reported chemistries, i.e., copper and nickel contents, for welds fabricated from the same heat of weld wire.
The staff has assessed the ~mplications of the large variability observed in the Palisades chemistry data with regard to the current methodology for assuring protection against PTS events. This methodology is contained in 10 CFR 50.61, "Fracture Toughness Requirements for Protection Against Pressurized Thermal Shock Events.* Based on its evaluation, considering plant-specific data, the staff concluded that Palisades could operate in compliance with and consistent with the margin of safety intended by 10 CFR 50.61 through late 1999.
However, the staff is concerned that the large variability in chemistry data could 111Pact other plant RPV evaluations. This concern will be thoroughly revt..-cl as part of the staff's review of plant specific PTS evaluations MMt a reassessment of the PTS rule and its implementation.
- If a rule change ta necessary, it could take several years to complete.
CONTACTS: 8. Elliot, NRR 415-2709 A. Wilford, NRR 415-2735 J. Tsao, NRR 415-2702
Ashok To provide assurance that all plants will maintain adequate protection against PTS events while the above assessments are in progress, the staff has evaluated all the-PWR reactor vessels using generic values of chemistry and increased margin terms to account for potential variability in chemistries.
The increased margin terms were established using generic data for various classes of weldments.
The results of applying generic values of chemistry and increased margin terms is that plants would be predicted to reach the PTS screening criteria at an earlier date than that given by the PTS assessment methodology given in 10 CFR 50.61.
The staff's assessment using generic values of chemistry and increased margin terms based on generic data indicates that all PWR reactor vessels are predicted to remain below the PTS screening criteria except for six reactor vessels. Five of the six reactor vessels would be projected to reach the PTS screening criteria a few years prior to the end of their licenses. However, if beltline variability of plant-specific fluence and surveillance data are considered, it is likely that the five reactor vessels would not exceed the PTS screening criteria until after the end of their licenses. The only plant significantly affected by the staff's assessment is Ginna.
Ginna would be projected to reach the PTS screening criteria in about seven effective full power years {EFPY) from January L 1995, when their embrittlement is conservatively calculated using the generic values of chemistry and increased margin terms.
However, Ginna has surveillance data that indicate that the amount of embrittlement may be less than the amount projected in this assessment. A detailed review of these surveillance data will be performed.
It is important to note that, based on currently available information, the analyses and results reported here are considered conservative evaluations performed solely for the purpose of demonstrating that there is no inunediate cause for concern and that there is adequate time to perform a more rigorous.
assessment of this issue. Recognizing the significance of RPV integrity and
- the potential impact of any changes in RPV evaluations on the industry, the staff intends to pursue a timely resolution of this *issue. Close cooperation with the industry will be important in this effort. is an explanation of the methodology used in the staff's evaluation, and includes a discussion of each PWR plant.
Attachments:
As stated cc:
W. Russell F. Miraglia G. Lainas
ATTACHMENT I DESCRIPTION Of METHODOLOGY USED IN THE STAFF'S PTS EVALUATION The purpose of this evaluation was to assess the potential impact of increased variability in reactor vessel material chemistries, i.e. copper and nickel, on the projected pressurized thermal shock reference temperature, RTprs* for PWR's.
This evaluation was performed by grouping vessels into sets based on their vendor and methods of fabrication, specifically welding procedure, and using available chemistry data to establish a generic distribution of copper and nickel for each of these sets of reactor vessels. These generic distributions, which were intended to bound potential variability in plant specific chemistries, were then used to determine RTPrs values.
Plates were not assessed because the variability in copper and nickel observed in the Palisades PTS review affected welds and not plates. Reactor vessels with projected end of license RTPrs values less than 200°f were considered acceptable because the increased variability in copper and nickel should not result in their vessels exceeding the PTS screening criteria for a significant amount of time. Reactor vessels with copper less than 0.10 percent were considered acceptable because reactor vessels with this small amount of copper were probably fabricated with copper limits for these welds and the variability in percent copper should not be a concern.
Reactor vessels with nickel values *less than 0.25 percent were considered acceptable because vessels with this low value of nickel would require extremely high neutron fluence (7El9) to exceed the PTS screening criteria.
The majority of the reactor vessels for U.S. nuclear plants were fabricated by Combustion Engineering (CE) and Babcock and Wilcox (B&W).
The small population of vessels fabricated by o~her manufacturers (Chicago Bridge and Iron, Rotterdam Dockyards, IHI-Toshiba and framatome) were evaluated by the staff and found to have either low values of copper and nickel iQ their weld materials or low fluences or both.
As such, the projected RT rs values for the limiting welds in these plants were significantly below t~e screening criteria at end of license.
All B&W fabricated reactor vessels that were assessed were welded with copper coated primary electrodes and with nickel in the primary electrode. All CE fabricated reactor vessels that were assessed were welded with copper coated primary electrodes, but had two sources of nickel.
One welding procedure had the main source of nickel from a cold nickel wire (e.g. Palisades) and the other procedure did not.use cold nickel feed and therefore, had the main source of nickel from the primary electrode.
The large variability in copper observed in the Palisades PTS data is believed to be due to the variability in the copper coating between weld wires.
The large variability in the percent nickel observed in the Palisades PTS data is believed to be due to the variability in the feed rate of cold nickel wire.
Welds with nickel in the primary electrode would not have large variability in percent nickel because the amount of nickel in weld wire is relatively uniform.
Hence, welds fabricated with a cold nickel wire are expected to have larger variability in nickel than welds with the nickel in the primary
2 electrode.
Based on vessel vendor and weld procedure, the reactor pressure vessels (RPVs) were divided into three subsets: B&W fabricated vessels, CE vessels fabricated with cold nickel wire feed, and CE vessels fabricated with nickel in the prtmary electrode.
The staff used Monte Carlo analyses to determine the distribution of adjusted RTPrs values for the assumed distribution of copper and nickel for each of
- these subsets of reactor vessels.
The increased margin term used in the screening analysis was then determined by the following equation:
M = 2* [ (u, )2 + (u4)2] 112 where u1 = the standard deviation of the initial (unirradiated) RTpts u4 = the standard deviation from the simulated list of RTpts This margin term was used along with the mean of the generic copper and nickel distributions and. the mean value of unirradiated reference temperature for each vendor to determine the critical fluence at which the mean adjusted R1,ts value would equal.the PTS screening criteria. Then the projected fluence ror each plant in the subset.was compared to the critical fluence.
The mean and*
standard deviation values used to simulate the copper and nickel variability, the initial reference temperature and the neutron fluence are identified in Table 1.
The source of each set of data is described below.
The three subsets of RPVs were evaluated using margin terms derived from (a) generic data from surveillance welds, and (b) discrete data from a weld fabricated using a specific heat number of weld wire.
B&W Evaluation The mean and standard deviation of the copper distribution used for screening B&W fabricated welds were determined using 36 discrete copper values from B&W owners group, nozzle drop-outs and surveillance welds.
The variability of nickel was established using the mean value from (36) surveillance welds and the standard deviation of nickel from welds fabricated using heat number 72105, a weld wire containing nickel.
The best estimated mean value for percent nickel for each unit was the value reported by the licensee in response to Generic Letter 92-01, Revision 1.
The staff's assessment using.generic data to establish the percent copper and an increased margin term indicates that all B&W fabricated PWR reactor vessels are predicted to be below the PTS screening criteria at end of license except for Ginna, Point Beach 2, Turkey Point 3, and Turkey Point 4.
CE Evaluation Cold Nickel Wire Feed The mean and standard deviation of the copper distribution used for screening CE welds with cold nickel wire feed was determined using 30 discrete copper values from CE fabricated surveillance welds with copper greater than 0.10 percent.
As mentioned, welds with copper less than 0.10 percent were probably
3 fabricated w1 th copper 1 imi ts for the we 1 ds and the variability in percent copper should not be a concern.
The mean and standard deviation of nickel was established using data from a Palisades weld fabricated using cold nickel wire feed (heat number W-5214) and the data from three additional surveillance welds that were fabricated using cold nickel wire feed.
The staff's assessment using generic data to establish the percent copper and an increased margin term indicates that all CE PWR reactor vessels fabricated with cold nickel wire feed are predicted to be below the PTS screening criteria at end of license except for Robinson 2 and Salem 1.
No Cold Nickel Wjre Feed The mean and standard deviation of the copper distribution used for screening CE welds fabricated with the primary electrode as the source of nickel was determined using 30 discrete copper values from CE fabricated surveillance welds with copper greater than 0.10 percent.
The standard deviation of the nickel distribution was established using the mean value from (11) surveillance welds with nickel in the primary electrode and the standard deviation of nickel from a Palisades weld fabricated with nickel in the primary electrode and no cold nickel wire feed (heat number 27204).
The best estimate mean value for percent nickel for each unit was the value reported by the licensee in response to Generic Letter 92-01, Revision 1.
The staff's assessment using generic dat~ to establish the percent copper and an increased margin term indicates that all CE PWR reactor vessels fabricated without col~ nickel wire feed are predicted to be below the PTS screening criteria at end of license ex~ept for Calvert Cliffs 1. *However, the RTPrs value for Calvert Cliffs 1 was determined using plant specific surveillance data which indicates that the RPV would not reach the PTS screening criteria until after end of license. The Calvert Cliffs 1 surveillance data were previously reviewed by the staff in a letter to the Baltimore Gas and Electric Company (BG&E) dated July 29, 1994.
In this letter the staff approved the use of surveillance data from McGuire 1 to determine the effect of radiation on intermediate shell axial welds in the Calvert Cliffs 1 reactor vessel.
The intermediate shell axial welds in the Calvert Cliffs reactor vessel were fabricated using the same weld wire heats as the McGuire 1 surveillance welds.
The July 29, 1994 letter to BG&E reconunended that a chemistry factor of 180 be used to determine the Mlers value for the intermediate shell axial welds.
The MTprs 1*s0 defined in tne PTS rule as the mean value in the adjustment in reference temperature caused by radiation. A margin term is added to the MT Prs to account for uncertainties in the in it ia 1 reference temperature, copper and nickel contents, fluence, and calculational procedures.
In assessing Calvert Cliffs 1, the staff used a generic value of initial reference temperature, a chemistry factor of 180, and a margin value from the Monte Carlo simulation for CE welds fabricated with the primary electrode as the main source of nickel.
The Calvert Cliffs reactor vessel welds were fabricated by CE with a weld procedure in which the main source of nickel is the primary electrode.
4 A concern raised by the relatively large variability observed in the W-5214 (heat # for the limiting weld in the Palisades reactor vessel) chemistry data is that the specimens included in plant surveillance programs may not have chemistries as lfmiting as the weld material actually in the RPV.
The staff reviewed the appropriateness of the chemistry of the surveillance specimen being referenced by Calvert Cliffs and concluded that it represents the best estimate chemistry for that heat of reactor vessel material.
Conclusions The staff ~nalyzed discrete data from a CE weld fabricated using weld wire heat number W-5214 (Palisades) and from a B&W weld fabricated using weld wire heat number 72105.
The results of the analyses using discrete data were not limiting when compared to the analyses performed using the generic data described in this attachment.
The staff's assessment using generic data to establish the mean value of the copper and the increased margin term indicates that all B&W fabricated PWR reactor vessels are predicted to be below the PTS screening criteria at end of license except for Ginna, Point Beach 2, Turkey Point 3, and Turkey Point 4. _
CE fabricated PWR reactor vessels at Calvert Cliffs 1, Robinson 2 and Salem 1 would be projected to reach the PTS screening criteria before-end of license
.if generic copper and increased margin terms are used.
However, none of the above units, except for Calvert Cliffs 1 and Ginna, would be projected to reach the PTS screening criteria prior to a few years before the end of their current licenses.
The Calvert Cliffs 1 reactor vessel would be projected to reach the PTS screening criteria in less than 2 effective full power years (EFPY) if its RTPrs value were calculated using generic data.
However, the staff has concluded that the Calvert Cliffs 1 PTS assessment can be performed using the McGuire 1 surveillance data to determine the RTPTS' since the surveillance data represents the best-estimate value of the increase in reference temperature resulting from irradiation and the copper in the McGuire 1 surveillance weld is equivalent to the generic mean value of the CE surveillance welds.
The amount of copper in the McGuire 1 surveillance weld is 0.21 weight percent and the mean value of copper from the CE surveillance welds is 0.226 weight percent. Using the McGuire 1 surveillance data and the margin value from the Monte Carlo simulation for CE welds fabricated without a cold nickel wire feed, the Calvert Cliffs 1 reactor vessel is predicted to reach the PTS screening criteria after the expiration of its current license in 2014.
The next surveillance capsule will be removed from McGuire 1 in 1997.
The test results from that capsule could affect the results of this assessment (i.e., if the measured shift indicates a higher rate of embrittlement than the previous test, the screening criteria could be reached sooner).
The staff was recently informed that the licensee has acquired new data relevant to the Calvert Cliffs 1 reactor vessel that will also need to be considered in its evaluation.
Ginna would be projected to reach the PTS screening criteria in approximately 7 EFPY from January 1, 1995 'when their embrittlement is conservatively calculated using the generic values of chemistry and increased margin terms.
5 However, Ginna has surveillance data that indicate that the amount of embrittlement may be less than the amount projected in this assessment.
Robinson 2, Turkey Point 3, and Turkey Point 4 also have data that could affect the dates ~hat these vessels are predicted to reach the PTS screening criteria. However, the use of generic data is conservative because the Monte Carlo simulations for these units resulted in margin values greater than those used in their plant specific PTS assessments *when credit is given for credible surveillance data. A more detailed review of the plant specific surveillance data is warranted in order to determine if the data used represent the best estimates of the chemistry values.
Based on the above conclusions, no licensee should reach the PTS screening criteria in the near future.
This assessment was performed using data reported by licensee in their responses to generic Letter 92-01, Revision 1.
CE has recently completed the Phase I review of records from reactor vessels that it fabricated.
The records have been entered in a computer database that CE considers to be proprietary.
The staff has performed an initial review of the CE chemical composition database and is assessing its impact on licensee PTS evaluations.
It is also recognized that other.sources of data, e.g., other vendors, foreign plants, etc, may exist). If the staff assessment concludes that these additional CE data could significantly affect PTS evaluations, the staff will determine the appropriate actions to assure that all data are incorporated by licensees in their.PTS evaluations.
The staff on March 28, 1995 became aware of ABB-CE proprietary data that could affect the PTS assessment of the Kewaunee reactor vessel.
ABB-CE provided the licensee for Kewaunee a sunwary of the data for its evaluation in a letter
- dated April 6, 1995.
The staff met with the licensee for Kewaunee on April 13, 1995 to discuss the effect that the ABB-CE data and its plant specific surveillance data would have on their PTS assessment.
The licensee presented its plant specific surveillance program results and some new information related to the reactor vessel weld chemistry variability. Based upon this information, the licensee believes that the Kewaunee reactor vessel will not reach the PTS screening criteria before the end of their license. The staff has not completed its review of the new information on the Kewaunee reactor vessel.
The new chemistry data could significantly change the Kewaunee PTS evaluation. However, based on conservative evaluations, the staff has concluded that the Kewaunee reactor vessel will not reach the PTS screening criteria in the near future.
~I~
6 Table 1 Simulation Cases for CE l B&W Fabricated Plants CE Plants CE plants BlW Plants Generic IJ'ata Generic Data Generic Data wtth cold wtth nickel 1n ntckel wtre electrodes Cl C2 Bl Mean Cu, I 01scr1t1 CE 01scr1t1 CE 01scntt Surv1tllanc1 Surv11111nc1 BlV Data Data Data
(*an-0.287)
<*an-0.221>
<*an*0.226)
Std. Dev.
(0.064)
(0.014)
(0.062) for Cu Mean ft1. I 1.01 0.1&*
O.&o Std. dtv.
0.163 0.033 0.016 for Nt Std. Div.
20I of the 20I 20I for flu1nc1...
.. 51
.. y
.. 5 Std. Dev.
17 17 20 for IRT NOTE:
The use of discrete values precludes the use of *an anct standard devtatton values tn the st*lattons. Th* nUllbers 1n paret1thes11 are equivalent *an and standard devtatton values.
'.