ML003779604
| ML003779604 | |
| Person / Time | |
|---|---|
| Site: | Dresden  |
| Issue date: | 12/06/2000 |
| From: | Rossbach L NRC/NRR/DLPM |
| To: | |
| rossbach L, nrr/dlpm, 415-2863 | |
| References | |
| TAC MA8346, TAC MA8347 | |
| Download: ML003779604 (9) | |
Text
UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 December 6, 2000 MEMORANDUM TO: File FROM:
Lawrence W. Rossbach, Project Manager, Section 2 Project Directorate III Division of Licensing Project Management Office of Nuclear Reactor Regulation
SUBJECT:
DRESDEN, UNITS 2 AND 3 - RECORD OF TELEPHONE CONVERSATIONS, NRC QUESTIONS ON FLUENCE, AND RECEIPT OF INFORMATION REGARDING FLUENCE FOR P-T CURVE AMENDMENT REQUEST (TAC NOS. MA8346 AND MA8347)
On February 23, 2000, Commonwealth Edison Company (ComEd) submitted an amendment request to revise the reactor vessel pressure-temperature (P-T) limit curves for Dresden, Units 2 and 3. On June 26, 2000, Nuclear Regulatory Commission (NRC) telecopied several questions (attached) to ComEd concerning the methods they used to derive the neutron fluence values used in developing the proposed P-T curves. These fluence issues were discussed with ComEd during a telephone call on June 28, 2000. Attending for NRC were: L. Lois, J. Medoff, and L. Rossbach. Attending for ComEd were: A. Haeger, G. DaBoo, T. Heisterman, and R.
Geier. No conclusions were reached during the phone call, however, CornEd stated that General Electric Company (GE) had provided a justification to ComEd for the fluence values used. On July 5, 2000, ComEd telecopied this information to L. Rossbach (attached memorandum from GE to ComEd dated December 3, 1999.)
On July 6, 2000, another telephone call was held with CoinEd to discuss the fluence issues.
The July 6 phone call attendees were the same as the June 28 call. No conclusions were reached during the phone call. However, in order to proceed with the P-T amendment, ComEd suggested that approving use of the new P-T curves for a limited time would be conservative.
The NRC participants agreed to review such a proposal. ComEd subsequently submitted this new proposal to NRC by letter dated July 17, 2000.
Docket Nos. 50-237 and 50-249 Attachments: As stated
December 6, 2000 MEMORANDUM TO: File FROM:
Lawrence W. Rossbach, Project Manager, Section 2 /RA/
Project Directorate III Division of Licensing Project Management Office of Nuclear Reactor Regulation
SUBJECT:
DRESDEN, UNITS 2 AND 3 - RECORD OF TELEPHONE CONVERSATIONS, NRC QUESTIONS ON FLUENCE, AND RECEIPT OF INFORMATION REGARDING FLUENCE FOR P-T CURVE AMENDMENT REQUEST (TAC NOS. MA8346 AND MA8347)
On February 23, 2000, Commonwealth Edison Company (ComEd) submitted an amendment request to revise the reactor vessel pressure-temperature (P-T) limit curves for Dresden, Units 2 and 3. On June 26, 2000, Nuclear Regulatory Commission (NRC) telecopied several questions (attached) to ComEd concerning the methods they used to derive the neutron fluence values used in developing the proposed P-T curves. These fluence issues were discussed with ComEd during a telephone call on June 28, 2000. Attending for NRC were: L. Lois, J. Medoff, and L. Rossbach. Attending for ComEd were: A. Haeger, G. DaBoo, T. Heisterman, and R.
Geier. No conclusions were reached during the phone call, however, CornEd stated that General Electric Company (GE) had provided a justification to ComEd for the fluence values used. On July 5, 2000, ComEd telecopied this information to L. Rossbach (attached memorandum from GE to ComEd dated December 3, 1999.)
On July 6, 2000, another telephone call was held with ComEd to discuss the fluence issues.
The July 6 phone call attendees were the same as the June 28 call. No conclusions were reached during the phone call. However, in order to proceed with the P-T amendment, ComEd suggested that approving use of the new P-T curves for a limited time would be conservative.
The NRC participants agreed to review such a proposal. CornEd subsequently submitted this new proposal to NRC by letter dated July 17, 2000.
Docket Nos. 50-237 and 50-249 Attachments: As stated DISTRIBUTION:
PD3-2 r/f M. Ring, RIII DOCUMENT NAME: G:\\PDIII-2\\dresden\\FluenceMTFa8346.wpd To receive a copy of this document, Indicate in the box: "C" = Copy without attachment/enclosure "E" = Copy with attachment/enclosure "N" = No copy OFFICE PM:LPD3 E LA:LPD3 SC:
L NAME LRossbacla;,-z-P,\\ CMoore W n (9,a DATE 1-1/J-/00 12/ 1/00 0
Officia ecord Copy
-0 0
SwRI-6901-002 "Dresden Nuclear Power Station Unit 2 Reactor Vessel Irradiation Surveillance_ Program Analysis Capsule No. 8N March 1983 and 0
SwRI-7484-003/1 "Dresden Nuclear Power Station Unit 3 Reactor Vessel Irradiation Surveillance Program Analysis of Capsule No. 18" dated February 1984.
Both reports have several non-conservative practices:
The cross sections used (DLC-23/CASK) are obsolete and no longer recommended for capsule analysis.
0 The number of groups used is inadequate, particularly in the computation of the spectrum.
The P1 approximation used for the scattering cross section is inadequate for the representation of the forward flux.
The iron transport scattering cross has been corrected (in ENDF/B-VI), because the older value was non conservative.
The capsules removed from the Dresden units were behind (in the radial direction) the jet pump. The combination of the P, approximation with the CASK cross section is very likely that resulted into a significant underestimate of the fluence.
0 The number of mesh points used seem to be inadequate to represent the capsule as shaded by the jet pump.
0 The (r, 6) and (r, z) values in the two capsule reports seem to be identical, yet the value of the fluences are different.
The staff recognizes that the report were state of the art at the time they were published. Also the staff is aware that all units have adopted low leakage loading in order to obtain longer fuel cycles. However, because of the reasons stated above we feel that the values used in the latest submittal are not defensible. The staff would appreciate the licensee's comments on the proposed fluence values of the Dresden and Quad Cities Units.
JU I -UU-WU Ur.1tu4ni r[WHI-%OWNI &W TrUW6A1Wn1 OLRYIU64 Genera Electric Company 175 CurMer A4venue. &*M Jose. CA 9512S GENE-B 13-02057-00-TCI December 3, 1999 To:
Guy DeBoo, CornEd cc: D Bouchie, GE BI Branlund, GE Fa-IX Moser, CornEd From:
TA Caine, GE
Subject:
Justification for Vessel Fluence Based on Dosimeuty Results Ajustification for using dosimetry results to determine the vessel fluence is provided in this letter. Supporting calculations and evidence of verification are in DRP B13-02057-00. Please call me at (408) 925-4047 if you have questions or need additional information regarding this evaluation.
Ritckground:
The amount of vessel embrittlement caused by irradiation in the belline region is important to establishing pressure-temperature (P-T) curves with the appropriate conservatism to assure ductile behavior of the vessel materials. The amount of embrittlement is established in accordance with Regulatory Guide 1.99, Revision 2 [1].
The amount of embrittlemenr consists of a calculated ARTNDT plus a Margin term.
The Margin term was added to [I] to account for uncertainties in vessel material chemistries, vessel fluence and surveillance test results that form the database from which
[1] was derived. When typical beltline embrittlement is calculated, the Margin term added is 347F for plate or 567 for weld material. These Margin terms can be anywhere from 30% to 100% of the ART~vr values calculated for a given BWR beltline material. In contrast, for PWRs, the Margin term may be only 15% of the ARTNT value. Given the large size of the Margin term relative to the AX.TNm-value, GE's practice for vessel fluence determination has been to use the best estimate dosimeter fluence with the peak ID lead factor from the neutron transport calculation. UQWWWQI&4 I _W9.0 F - W&F W I F-F W19
Jiji-U5-UU U 4manl Pr-rom-L.uM tW. Kt.ULAIUKT RKVIC.
Ju3IU i-U r U~FUI r-Iuo GTM-B 13-02057-00-TC 1 Obiecive:
The purpose of this evaluation is to demonstrate that the use of best estimate dosimeter results with associated lead factors to determine vessel fluences is an acceptable approach given the size of the Margin term prescribed in [I].
InDuts:
An evaluation of the Margin term in the overall BWR embrinlement should look at both plate and weld materials. Further, there are some general groups of BWRs which should be evaluated. Specifically, they are I) vessels with low fluence and high chemistries,
- 2) vessels with moderate fluence and high chemistries and 3) vessels with high fluence and low chemistries (it turns out that BWR/6s which have the highest BWR fluences also had vessel material specifications for low copper). Plate and weld materials for a typical plant from each of these three groups are evaluated here. The plants, limiting plate and weld chemistries and 32 effective full power year (EFPY) tluences are listed below.
PlantlMaterial
% Cu
% Ni Fluence (n/era)
Ouad Cities 2 Plate 0.14 0.54
- 5. lxlO' Weld 0.24 0.37 5.14xl0' Plate 0.14 0.60 1.46x10' Weld 0.22 1.00 1.37x101' Clinton Plate 0.10 0.64 6.5x10C M
Weld 0.10 1.08 6.9x101 Other inputs to this evaluation are referenced as they are used.
Assumptions:
- 1. Comparison of dosimeter flux results among similar plants provides a conservatve assessment of the uncertainty of dosimeter measurements. This assumption is valid because the variation in factors that affect dosimeter results, such as flux wire location, fuel design and plant operation, would be greater between different plants than they would be if numerous sets of flux wires were in a single plant.
- 2. The uncertanry in the lead factor derived from neutron transport calculations is smaller than the uncernainzy in the fluence derivedfrom the calculatons. This assumption is valid because the factors that create uncertainty in the neutron transport calculations, such as dimensions, modeling of the core, cross sections used, etc., tend to cancel out when ratioing the flux at one location relative to another location.
.GENE-B 13-02057-00-TC I Methods:
The approach that has been used by GE for most plants to establish peak beltline fluence has been to take the dosimeter best estimate fluence, extrapolated to 32 effective ful power years (EFPY), and divide it by the dosimeter-to-vessel peak lead factor to calculate the peak fluence on the vessel inside surface (ID). Reg. Guide 1.99, Rev. 2 is used to establish the I/4T fluence from the ID fluence.
Two points are demonstrated in this evaluation:
- 1. The overall uncertainty associated with a dosimeter flux measurement is relatively small, and
- 2. The amount of increase in ART= associated with the uncertainty in the fluence, calculated from the dosimcter flux and lead factor, is relatively small, and therefore covered in the Margin term.
A conservative assessment of the uncertainty associated with dosimeter measurements can be made by comparing flux wire results from similar BWRs. Two groups of similar BWRs are evaluated here: the BWR/4, 251 group and the BWRJ4, 218 group. These are the two largest groups of BWRs that could be labeled "similar". The similarities considered are vessel diameter, shroud diameter, number of fuel bundles, annulus water temperature, core design and plant operation. For example, the Cooper plant is not included in the BWR*4, 218 group because some time ago, four of the core "corner" bundles were replaced with stainless steel "'dummy bundles". These dummy bundles are near the surveillance capsules and decrease the fluence at the dosimeters by about 25%.
Obviously, there are as-built differences in features which are nominally the same between plants. These plant-to-plant differences introduce variation in the dosimeter results, which makes the variation in the fluxes for the group conservative compared to the variation due to uncertainty in the flux determination method used.
For lead factor calculation, the relative fluence at two locations is the critical determination. Many of the sources of uncertainty in determining an absolute fluence value cancel out of the lead factor determination, so the overall lead factor uncertainty is less than that for a fluence value. In one evaluation [2], the la uncertainty was estimated to be 15% for the lead factor.
The combined uncertainty for fluence would be the SRSS combination of the dosimeter uncertainty and the lead factor uncenainty:
Oacl =
+ al]acw]
(1)
The impact on ART= of a total fluence uncertainty of 2a is evaluated relative to the Margin term from [I]. The Margin term is 347F for plate or 567F for weld, and is *i*llUZIU1. Uf;qcdm rrom-6wm ri; ArGU6AIUT ar.KYI69a DOOU99IL411 I-capg
- r. uqiu f r-ruj
GENE-B 13-02057-00-TC 1 intended to cover uncertainties in the shift correlation in [I] related to material chemistry, to fluence and to Charpy test measurements. Most of the data that forms the basis of the shift correlation is high fluence PWR data with relatively large shifts. Thus the Margin terms for plate and weld can be viewed as having "excess margin" relative to smaller BWR shifts. This "excess margin" is not quantified, but in cases where fluence uncertainty causes a small potential increase in ARTND,, the excess margin can be qualitatively applied as justification for using best estimate fluence as described previously.
ResulTs:
Table I shows the dosimeter flux values for the BWR/4's in the 218 inch and 251 inch diameter category. The la uncertainties, conservatively treating the values as samples from the same population, are 6% and 9o/a of the mean value. When a 90/a dosimeter uncertainty is combined with a 15% lead factor Icr uncertainty, the overall uncertainty per equation (1) is 17.5%, so the 2a uncertainty would be 35%.
Table 2 has ARTNDT values for the three representative plants' limiting materials, based on nominal fluence and based on upper bound fluence, 1.35 times nominal. The differences in ARTNDT between nominal and upper bound range from 40F to 67F for the plates and from 6*F to 15*F for the welds. It is quite reasonable to conclude that the Margin terms of 34°F and 560F for plate and weld, respectively, add sufficient margin to the nominal AR.TT values to cover the potential increases associated with the upper bound fluences.
==
Conclusions:==
The Margin term in (1] is large enough that use of a nominal fluence, based on dosimeter results and a calculated lead factor, results in an acceptably conservative determination of vessel embrittlement for use in P-T curves. This conclusion is applicable as long as a plant does not make a core change that would significantly change the relative fluence distribution and lead factor, as might occur when changing fuel types. In such cases, a neutron transport calculation should be performed and supplemental dosimetry should be considered.
References:
[1]
Regulatory Guide 1.99, Revision 2, -Radiation Embrittlement of Reactor Vessel Materials," May 1989.
[2)
Rogersm DR, -Kernkraftwerk Leibstadt AG (KKL) Reactor Pressure Vessel Neutron Fluence Analysis," GE Report GENE-B 1100705-01, February 1997.
jUI-U2-UU ur:4*am jrom-LuM tv KiULAIUNT UKVII.3 GENE-B]3-o2057-O0-TCI Table 1. Evaluation of Dosimeter flux Uncertainties Among Similar I3WRs Measured Flux BWR4 -251" Dia. Plant (nCM"A2/S)
Peacm Bottom 2 7.50E+08 Peach Bottom 3 6.80E+08 Browns Ferry 2 5.90E+08 Susquehanna I 6.60E÷08 Hope Creek 7.49E+08 Susquehanna 2 6.70E+08 Mean 6.83E+08 Sigma 6.03E÷07 Sigma %
8.83%
Measured Flux BWR/4 - 218" Dia. Plant (nfrcA2/s)
Brunswick 1 1.1 BE+09 Brunswick 2 1.18E÷09 FitzPatrick 1.1OE,09 Hatch 1 1.02E+09 Hatch 2 1.12E÷09 Mean 1.14EO09 Sigma 7.16E-&-07 Sigma %
6.29% T-515 P.O5/UT F-TOH
I GENE-B 1 3-02037-00-TC I Table 2. Comparison of ARTjr Values for Nominal and Upper Bound (Nominal + 2o) Fuences Upper i
Nominal Bound 32 EFPY 32 ]PPY Difference Material Thickness 32 EFPY ID Fluence,
%Cu
%Ni CF FF FF IUB/Nom Delta RTnd Delta R'hdI (deg F)
(in)
Nominal (n/cm"2)
(0)
(I.35x))
Nominal UpperBound Quad *it* 2 Plate 6.13 5.1EE417 0.14 0.54 97.3 0.241 0.285 1.182 23.5 27.8 4.3 Weld 6.13 5.1E+17 0.24 0.37 140.6 0.241 0.285 1.182 33.9 40C 6.2 Plvim Plate 3.63 1.46E4 18 0.14 0.6 100 0.425 0.488 1.149 42.5 48.8 6.3 Weld 3.65 1.37E4 18 0.22 1
232 0.412 0.474 1.151 93.6 110.0 14.5 Clinton Plate 5.59 6.5E418 0.1 0.64 65.4 0.787 0.869 1.105 31.4 56.9 5.4 Weld 5.59 6.91418 0
1 1.08 133 0.803 0.886 1.104 108.4 119.6 11.2 l
..-