ML20046A804
| ML20046A804 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/23/1993 |
| From: | Broughton T GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| C311-93-2109, NUDOCS 9307300077 | |
| Download: ML20046A804 (5) | |
Text
.
GPU Nuclear Corporation l
Route 441 South p*-
tt mA M
UOEhh[
P.O. Box 480 Middletown, Pennsylvania 17057-0480 i
(717)944-7621 Writer's Direct Dial Number:
i (717) 948-8005 July 23, 1993 C311-93-2109 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D. C.
20555
Dear Sirs:
Subject:
Three Mile Island Nuclear Station, Unit 1, (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Technical Specification Change Request (TSCR) No. 207 Request for Additional Information (TAC No. M86085)
The referenced letter presented questions about 1) the calculation of the applied stress intensity for the postulated flaw, 2) the initial RT,,, (IRT) of reactor vessel welds, 3) the margin used in the adjusted RT,.y (ART) calculation, and 4) the selection of the limiting material. The following are responses to these questions, in the order in which they were presented in the reference:
1.
In accordance with the requirements of the Code, GPUN will assess the controlling weld on the basis of the major axis of the postulated defect to be oriented in a longitudinal direction regardless of the weld orientation.
Since welds WF-25 and SA-1526 are made of the same weld wire, are more limiting than the WF-70 weld, and weld WF-25 receives a higher fluence than weld SA-1526, the THI-1 vessel beltline circumferential weld WF-25 will become the limiting weld with a postulated defect oriented in the vertical plane.
2.
There appears to be some confusion with respect to the available information on initial RT values and the meaning of these values.
Enclosure No.1 of the referenced NRC letter contains some errors and redundancies in IRT values, and the text of-the question indicates a misunderstanding regarding the meaning of the mean value of -SaF with an associated standard deviation of 20aF which is the result of a statistical evaluation of Linde-80 welds, reported in BAW-1803 Rev. No.l.
Attachment No. I to this letter provides the correct values, as well as the source of information and comments as to their basis.
The statistical study, sponsored by the B&WOG, reported in BAW-1803, yielded -Saf and 20af values for the mean initial RT value and a e h'ty T \\ \\
9307300077 930723
-i, PDR ADOCK 0 % O289 N P
PDR f
\\
GPU Nuclear Corporation is a subsidiary of General Pubhc Utthties Corporation i
C311'-93-2109
- Page 2 of 4 standard deviation, respectively. The intended use of these values was to provide a basis for defining the IRT of a weld wire if there was no other data available by which the IRT could be defined. Thus, if Charpy data is available for a given weld wire, then the IRT can be established on the basis of such data. The IRT values for the Linde-80 weld wires have virtually all been established or controlled by the 50 Ft-Lb Charpy transition temperature.
Since the ASME Code Section NB-2330 was issued subsequent to the fabrication and testing of some of the B&W fabricated reactor vessel weld specimens, drop weight tests were not performed at the time Charpy tests of the surveillance capsules weld materials were performed.
In 1976 B&W utilized excess surveillance capsule weld material, nozzle belt dropout (NBD) weld material, and Linde-80 weld metal qualification material to perform additional Charpy testing and to perform drop weight tests. The resulting values of the nil-ductility-temperature (T,m) as well as the Charpy results for the 50 Ft-Lb transition temperature and 35 mils lateral expansion are tabulated in tables No.
3-1 and 3-2.of BAW-1803.
In all of these test results, the 50 Ft-Lb transition temperature controls the IRT.
For the TMI-I controlling weld metals (WF-25 and SA-1526), the 50 ft-Lb transition temperature is shown to be controlling the IRT. Table No. 3-2 in BAW-1803 contains the results of the aforementioned supplementary testing of excess TMI-1 surveillance capsule weld metal WF-25(5) which show a T m value of -20*F compared to the IRT value of o
-13*F established on the basis of the 50 Ft-Lb transition temperature.
This IRT value is only a few degrees higher than the -21*F IRT established by the initial Charpy testing.
Likewise, the results shown in table No. 3-2 for SA-1526 (CR-3 NBD) yielded an IRT of -20*F compared to the Surry Unit I surveillaace capsule IRT of -10*F which was established on the basis of the 50 ft-Lb transition temperature.
3.
GPUN is cognizant of the activities to study the IRT of weld metal WF-70. The issue regarding the WF-70 weld material is rooted in a finding that the IRT of two samples, each from different sources, had relatively high measured IRT values which could not be reconciled with the available reactor vessel surveillance capsule data from several vessels.
To date, the B&WOG had presented a plan to the NRC to investigate several potential causes for the differences in the IRT values. These investigations included microstructure characterization, evaluation of the effect of the post welding heat treatment, examination of the differences in chemical composition, Charpy impact testing, drop weight testing and fracture toughness testing of various specimens of WF-70 weld wire, including such material from the Midland usel. The results of these investigations were reported in BAW-2100 and submitted to the NRC. The results were insufficient to develop a firm conclusion regarding the cause of the differences in the IRT values.
The present plan, which is mentioned in the reference, entails fracture toughness testing of irradiated WF-70 weld material in anticipation of being able to demonstrate the conservatism contained in the indexing of the fracture toughness by IRT values which are based on the 50 Ft-Lb
i
~
C311-93-2109
~ Page 3 of 4 q
l
'transition temperature and the shift which is based on the shift in the 30 Ft-Lb transition temperature.
)
Until this issue is resolved. (expected resolution by early 1994), the NRC is requesting that GPUN utilize an IRT value of -5'F and an associated standard deviation value of 20*F to assess the embrittlement trend of the WF-70 weld in the TMI-1 reactor vessel.
Based on calculation of the chemistry factor (C.F.) per position No. 2 of NRC Regulatory Guide 1.99, Rev.2, using all WF-70 & WF-209 welds irradiated surveillance data, a C.F. of 164.45 is obtained. The resulting RT,, at 1/4T is 147aF when the limiting weld RT,, is 186*F at 15.2 EFPY. This result demonstrates that the WF-70 weld in the TMI-1 vessel is not the limiting weld when compared to the predicted RT., for the WF-25 weld at 15.2 EFPY.
4.
GPUN's evaluation of the B&WOG integrated reactor vessel surveillance program data indicated that the procedure prescribed by position No. 2 of NRC Regulatory Guide 1.99, Rev.2 is inadequate in evaluating the r
bounding RT., trend of Linde-80 weld metal irradiation embrittlement, when multiple sources of weld metal. irradiated capsules, all made of the same weld wire heat, are available.
The regulatory guide procedure is adequate for evaluating the ir. adiated trend of a specific section of a weld when all of the irradiated capsules are cut from that same weld section, and the margin is 28aF.
The IRT for each of the reported B&WOG integrated surveillance capsule program welds is an upper bound value established on the basis of _the Charpy impact tests, for that weld section, at which 50 Ft-Lbs of absorbed energy will be attained. While it is true that multiple welds made of the same weld wire, but from different vessels, exhibit different IRT values, the standard deviation of the range of these values does not represent the standard deviation of the precision of the test method as defined in the Regulatory Guide. Thus, the reported IRT values are upper bound values for which the standard deviation for purposes of calculating the margin term, as defined in the Regulatory Guide, should be (0) zero.
It can be argued that the range of the IRT values for welds made with a given weld wire simply represents the variability which may exist in a vessel weld. Thus, the mean of those values and the associated standard deviation should be used as the base from which the shift in embrittlement is to be measured. However, as indicated in the TSCR submittal, it is noted that for welds made of the same weld wire the trend in the shift is lower for welds with higher IRT values. Also the use of such an argument yields RT,, trends which become overly conservative in the lower fluence region as compared to the available irradiated surveillance capsule data.
I GPUN's TSCR No. 207 submittal was based on evaluation of various methods of predicting the bounding trend of irradiation embrittlement and concluded that use of the chemistry factors in the regulatory guide, based on the mean plus one standard deviation of the copper and nickel content of the all of the weld data of a given weld wire, when shifted by a small amount will yield a bounding trend of the irradiated RT., for all of the data. The values of the amounts by which the trend
a C311-93-2109
' Page 4 of 4
' formed by the chemistry factor needed to be shifted were 10af for WF-25
)
and SA-1526 and 13aF for the WF-70 welds. We believe this approach to yield an adequately bounding trend of embrittlement behavior of the TMI-l reactor vessel limiting welds as compared to use of the TMI-1 vessel surveillance capsules for WF-25 and use of the Surry Unit No. 1 SA-1526 surveillance capsules as a surrogate for the TMI-1 SA-1526 welds.
During our July 7,1993 discussion it was agreed that the current study by the NRC Staff and B&W of the IRT of weld WF-70 should be expanded to include weld WF-25.
Resolution of the issue is expected in early 1994.
The Staff also agreed to consider the methodology proposed by GPUN.
On July 14,1993 the Staff responded to GPUN's request. The Staff confirmed its position that all of the unirradiated IRT values (Attachment No.1) for the limiting TMI-1 weld wire must be used to establish a mean IRT value and that the standard deviation of those values represents the sigma I value to be utilized in the margin term. GPUN understands the Staff position but continues to believe that the alternate methodology is technically sound and plans to pursue the issue in the future. However, in order to resolve the near term need of the TSCR, GPUN has performed a conservative calculation of the embrittlement trends, per the NRC Staff position, of the TMI-I limiting WF-25 weld so that an interim extension of the present TMI-1 P/T limits could be obtained. The calculation utilizes an IRT of -7.29'F with a sigma I value of 20.84*F (margin term equal to 50.21*F) and chemistry factor of 221.25 which is calculated in accordance with regulatory guide position No. 2 from welds WF-25 and SA-1526 irradiated surveillance capsules. The results of this calculation are that weld WF-25 will reach an RT,,, at 1/4T of 186af at 15.2 EFPY. This is expected to occur near the end of fuel cycle No.12, approximately June, 1999 based on a 90% availability factor. The concurrent calculated RT,,, value at the 3/4T location is 135*F.
In order to satisfy the 10CFR50.61 PTS Rule, GPUN has also calculated RT,,, at the inside surface of the vessel to be 255*F at the end of 32EFPY.
Although an interim resolution would allow extension of the present operating lirits for another 5 EFPY, GPUN requests that the NRC address the issue of the various IRT values as expeditiously as possible.
Resolution of the issue will enable GPUN to address its reactor vessel embrittlement management plans for the present license period and consider the viability for potential plant life extension evaluations.
Sincerely, b
T.G.Broudton Vice President and Director, TMI-1 TGB/YN Attachment cc:
Administrator Region I TMI Senior Resident Inspector THI-I Senior Project Manager I
I
4 ATTACHMENT NO. 1 INITIAL RT DATA FOR WF-25 & SA-1526 WELDS ELANT WELD SOURCE IRT 'F BEFERENCES AND COMMENTS THI-1 WF-25 RVSP
-21
. BAW-1803 R-1, pg. A-20 3dh-60
. IRT = T t
.Tg=
. IRT = 39-60 = -21F
. No drop weight tests.
TMI-1 WF-25(5)
RVSP
-13
. BAW-1803 R-1, pg. 3-3
. IRT = T
- 60 fh
. Ikh== 4
.T 47-60 = -13F 20F; drop weight
.T gg = tests in 1976.
BAW-1803 R-1, pg. A-35 Crystal WF-25 RVSP
-27 IRT = T
- 60 River 3
.T dh.
.Ikh==333-60 = -27F
. No drop weight tests.
TMI-2 WF-25(6)
NBD 33
. BAW-1803 R-1, pg. 3-3
- IRT = T
- 60 i
m = 9Ih
.T
. IRT = 93-60 = 33F I
.Tgg = -10F; drop weight tests in 1976.
48 Hr. stress relief.
I Oconee 3 WF-25(9)
NBD 7
= BAW-1803 R-1, pg. 3-3
. Tm = 6 h - 60
. IRT = T IkT = 67-60 = 7F s
-40F; drop weight
.T gg = tests in 1976.
l 48 Hr. stress relief.
Surry 1 SA-1526 RVSP
-10
. BAW-1803 R-1, pg. A-72 5df-60
. IRT = T
.Tg=
+ IRT = 50-60 = -10F
. No drop weight tests.
Crystal SA-1526 NBD
-20
+ BAW-1803 R-1, pg. 3-3 River 3
. IRT = T
- 60 INb== 4dh
.T 40-60 = -20F
.T
= -40F; drop weight gg tests in 1976.
48 Hr. stress relief.
Note:
The parenthetical ( ) number in the weld designation column represents the B&W NSSS number for the reactor vessel from which the nozzle belt dropout (NBD) weld section was taken.
mean IRT = -7.29'F ;
Sigma I =-20.84'F Margin = 50.21'F ;
C.F. = 221.25 i
n-e
.