ML20150F322
| ML20150F322 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 07/08/1988 |
| From: | Walker R TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC) |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| NUDOCS 8807180234 | |
| Download: ML20150F322 (11) | |
Text
'
l N==
Log # TXX-88550 2
File # 903.1 clo C
C 903.6 7t/ELECTbC Ref.# 10CFR50.4(a)
July 8, 1988 Wi%re G. Counsil cea.me vue trans c--
V. S. fluclear Regulatory Commission Attent'on: Document Control Desk Washington, D. C. 20555
SUBJECT:
COMANCHE PEAK STEAM ELECTRIC STATION (CPSES)
DOCKET NO. 50-445 APPLICAT10fl 0F LEAK-BEFORE-BREAK METHODOLOGY TO BRANCH LINES Gentlemen:
On April 15, 1988, TV Electric submitted to you by our letter logged TXX-88382 ten (10) copies of the Leak-Before-Break (LBB) "CPSES-1 WHIPJET Program Report" for your review. As a reselt of your review, a t.elephone conference call was held on June 22, 1988, among representatives of the NRC staff, 10 Electric Licensing and Engineering, and Robert L. Cloud Associates (RLCA),
Inc. to discuss the results presented in the CPSES-1 WHIPJET Report.
The flRC Staff representatives requested that TV Electric perform the following six action items to enable the Staff to complete its review.
1.
The 14-inch pressurizer surge line high stress location occurs at the loop nozzle safe ?nd weld uhen torsion is included with the bending moments.
When torsion is excluded, the high stress location occurs at the opposite end of the surge line, i.e., the pressurizer nozzle safe end weld.
The CPSES-1 WHIPJET report provides LBB analysis results for the loop nozzle safe end weld, but the NRC staff requested the pressurizer r.ozzle safe end weld LBU analysis results. provides the requested information.
As shown in Attachment 1, the required LBB margins are satisfactory at this location.
2.
The 12-inch Residual Heat Removal lines have high stress locations in the Class 2 portion of the lines.
The Class I high stress location was analyzed and reported in the CPSES-1 WHIPJET report because the RHR Class 2 portion of the system experiences a lower pressure and temperature, and is isolatable.
As mentioned in the CFSES-1 WHIPJET report, the Class 2 highest stress location was also analyzed and found to have acceptable margins.
The NRC staff requested the results for the high-stress location in the Class 2 portion of the RHR system. provides the e
requested information on the Class 2 portion of the RHR system.
'PoM 8807180234 800708 1 \\
[DR ADOCK 05000445 PNU 400 North Ohve Street LB81 Dallas, Texas H201
TXX-88550 July 8, 1988 Page 2 of 3 3.
The 10 gpm crack length reported for the 14-inch pressurizer surge line in the CPSES-1 WHIPJET report, was 5.94 inches.
The NRC staff requested the 10 gpm crack length calculated for the "no torsion" case at the reported location.
RLCA reanalyzed this location (the loop nozzle safe end weld) for the case with no torsion and determined the 10 gpni crack length to be 8.02 inches. Attachment 3 provides further details of the pressurizer surge line loop nozzle safe end weld.
4.
The NRC staff identified a recommendation in Appendix B of the CPSES-1 WHIPJET report for using a 1% strain fit to tensile test data that was mistakenly referenced.
The correct reference is Peference 8 and not Reference 12. Attachment 4 provides further details.
5.
The NRC staff inquired about the material composition of the SIS accumulator nozzles based on the description of the CPSES-1 WHIPJET program scope presented by the CPSES-1 WHIPJET report.
The four accumulator nozzles are made of cast austenitic stainless steel.
The conservative approach taken by the CPSES-1 WHIPJET program in analyzing lower bound material properties bounds the LBB behavior of cast stainless steel.
The review determined that thermally aged cast stainless steel with an end-of-life condition, has toughness values in excess of the lower bound submerged arc weld metal analyzed by the CPSES-1 WHIPJET program.
In addition, Westinghouse considered and analyzed cast material and welds in the CPSES main loop LBB program. Attachment 5 and its Enclosure provide further details and references for the properties of cast material.
6.
The NRC staff asked whether the CPSES-1 WHIPJET program considers the RHR containment penetration as an anchor and analyzed systems on an anchor-to-anchor basis. As stated on page 2-2 of the CPSES-1 WHIPJET report, systems were analyzed on an anchor-to-anchor basis.
For the Staff's convenience, a copy of that page is provided as Attachment 6.
As discussed in the conference call, one anchor of the RHR system is at the containment penetration.
The above-discussion, six (6) attachments and two (2) enclosures completes the actions requested by the NRC Staff of TU Electric in the June 22, 1988 conference call concerning the "CPSES-1 WHIPJET Program Report."
.n w
TXX-88550 July 8, 1988 Page 3 of 3 We remain available to meet with the Staff to resolve any questions which may arise and to provide any additional information to complete its review of this report.
The FSAR will be amended to reflect the revised design bases as a result of implementing LBB methodology upon NRC acceptance.
Very truly yours, W. G. Counsil By:
EA c i
R. 0.Alalktr Manager, Nuclear Licensing HAM /mlh Attachments and Enclosures c - Mr. R. D, Martin, Region IV Resident Inspectors, CPSES (3)
Attechrent 1 TXX-88550,
July'8, 1988 Pego 1 of 1 l
DATA INPUT AND RESULTS FOR PRESSURIZER NOZZLE SAFE END. WELD LOCATION
("No Torsion" High Stress Location)
For node 800 (pressurizer nozzle), the 10 gpm crack length is 4.475 inches.
The resulting margins are:
~
Load margin Flaw Margin (2 1.0 required)
(1 2.0 required)
Base Metal 2.72 3.45 Submerged Arc Weld 2.25 3.08 Shielded Metal Arc Weld 2.50 3.29 l
l The LBB input for the pressurizer nozzle location is as follows:
Loading Forces (lbs)
Moments (in-lbs)
Line Conditions Fx My Mz 1
RCS 14-inch.
Deadweight 5170
-2230 5940 Fzr. nozzle Thermal
-7130
-100990
-2826530 SSE 7530 39850 110020 PICEP Inout Outside Diamet'er = 14.0 inches Pipe Thickness = 1.406 inches Circumferential crack orientation Diam 6ad Shaped Crack E = 25.5D+06 psi Yield stress = 24.5D+03 psi Yield strain = 0. (Default value)
Alpha = 3.75 FACN = 4.82 Flow stress = 46.2D+03 psi Pressure = 2249.7 psi Temperature = 653 F Absolute external pressure = 14.7 psi
Attschmnent 2 T%X-y8550, July 8, 1988 Page 1 of.1 DATA INPUT AND RESULTS FOR RESIDUAL HEAT REMOVAL (CLASS 2) HIGH STRESS. LOCATION For node 2680 (High stress class 2 node), the 10 gpm crack length is 5.763 inches.
The resulting margins are:
Load margin Flaw Margin (2 1.0 required)
(1 2.0 required)
Base Metal 3.11 3.49 Submerged Arc Weld 2.99 3.40 Shielded Metal Arc Weld 3.12 3.50 The LEB input for node 2680 is as follows:
Loading Forces (lbs)
Moments (in-lbs)
Line Conditions Fx My Mz RHR 12-inch.
Deadweight 30 1060
-44650 Thermal
-9880
-214360
-306870 SSE 6700 312660 468300 SAM 1090 113940
-100 SAM 1130 39550 150 PICEP Inout Outside Diameter = 12.75 inches Pipe Thickness = 0.375 inches Circugferential crack orientation Diamond Shaped crack E = 25.5D+06 psi Yield stress = 24.5D+03 psi Yield strain = 0. (Default value)
Alpha = 3.75 FACN = 4.82 Flow stress = 46.2D+03 psi Pressure = 414.7 psi Temperature = 350 F Absolute external pressure = 14.7 psi
I
}
. Attachment ~3 TXX-88550 July 8, 1988 l
Page.1 of'l l
l t
LEAK RATE DATA INPUT AND RESULTS FOR SURGE LINE LOOP NOZZLE SAFE END WELD LOCATION For node 1010 (pressurizer surge line nozzle at the main loop),
the 10 gpm crack length is 8.019 inches when torsion is not included.
Note that this node is not the high stress location when torsion is excluded (see Attachment 1).
The LBB input for tus loop nozzle location is as follows:
Loading Forces (lbs)
Moments (in-lbs)
Line Conditions Fx My Mz RCS 14-inch.
Deadweight 2370 49640 105310 Thermal
-4470
-896720 706880 SSE 5490 1272360 480810 PICEP Inout Outside Diameter = 14.0 inches Pipe Thickness = 1.406 inches Circumferential crack orientation Diamond Shaped crack E = 25.5D+06 psi Yield stress = 24.5D+03 psi Yield strain = 0. (Default value)
Alpha = 3.75 FACN = 4.82 Flow stress = 46.2D+03 psi Pressure = 2249.7 psi Temperature = 653 F Absolute external pressure = 14.7 psi
- 1 l
. Attachment 4 TXX-88550 '
July 8, 1988 Page-1 of 1 i
STRAIN RANGE FOR STRESS-STRAIN FIT With respect to the selection of the Ramberg-Osgood work hardening constants for the EPRI estimation scheme, a low strain-range fit is desirable.
Fitting the stress-strain data over a small strain range is consistent with the procedure used in EPRI report NP-4690-SR, Evaluation of Flaws in Austenitic Steel Pioina.
The discussion in Appendix B of the final CPSES-1 WHIPJET report concerning this approach (p. B-5) should have referred to Reference 8 vice Reference 12. To clarify this issue, the content of the attachment material contained in Reference 8 is summarized by a June 23, 1988 letter from D. M. Norris to D. Quinones and provided as an enclosure to this letter.
This same 1% strain fit approach was used for WHIPJET work accepted by the NRC on another docket.
The use of a strain range to 1% is a subset of the approach recommended in the EPRI PICEP tcanual (NP-3596-SR), e.g., "fitting up to strain of about 5%."
For these reasons, fitting the stress-strain data to a plastic strain of 1% is generally conservative.
- Attadhment 5' TXX-88550
- 9 July-8, 19,88
- Page' 1 of 1 CAST MATERIAL IN CPSES-1 ACCUMULATOR INJECTION LATERALS The four 10-inch accumulator injection 4E-degree laterals, at the 23-inch reactor coolant loop, are constructed from SA 351 Grade CF8A cast steel with E 308-16 manual walds. The CPSES-1 surge line and Residual Heat Removal lines have no cast components.
Table B-1-1 in the CPSES-1 WHIPJET final report should be modified with an additional row showing material SA 351 Grade CF8A for 45-degree laterals. A copy of the revised Table B-1-1 is contained as an enclosure to this Attachment.
LBB studies on the reactor coolant cast pipes and associated welds previously have been performed by Westinghouse.
The delta ferrite ranges from 20 to 21 percent for the four CPSES-1 accumulator laterals, which is typical for the type of nozzle. The fracture toughness of cast stainless steel manual welds for the thermally aged, end-of-life condition is no worse than the toughness of the worst-case submerged arc or shielded metal arc welds. Furthermore, the nozzle section is thicker than the adjacent pipe section.
For these reasons the cast stainless steel laterals have been conservatively bounded by the LBB analysis.
f
Attac) ment o
~g TKX-9f550 July 8, 1985 Pege,1 of 1 Selected CPSES-1 system lines were screened for abnormal crack growth mechanisms (such as corrosion or gross global fatigue) and other potential failure mechanisms (e.g., water hammer).
Additional screening considerations included unexpected flow stratification conditions, erosion /corrosi.on, and other adverse industry experience.
Applicable stress calculation analyses were examined anchor to anchor on a system line bacts to determine which locations (not
~
limited to normal postulated pipe break 1ocations) were most critical in terms of highest stress.
These locations were then coupled with all possible material properties (base metal and welds) in a deterministic evaluation that demonstrated sufficient margin against failure.
The pertinent system locations were analyzed to determine the stability of postulated through-wall flaws.
In flux welds, the loading conditions in the pipe (termed normal + SSE), were the combination of:
Safe Seismic deadweight + thermal + pressure +
Shutdown
+ Anchor Earthquake Motion or:
l For the consideration of base metal and non-flux welds, thermal loading (exparsion) was not included in the normal plus SSE.
Leak rate versus crack size was determined considering normal operating loads (DW + TH + P) and fluid conditions (1).
A limiting leak rate based upon a conservative detectable leakage limit'o# 1.0 gpu multiplied by a margin of 10 was used to determine the crack length corresponding to 10 gpm.
This laakage size crack length was analyzed for two margins against failure:
first, that a margin of 2 on crack size exists when compared to the (normal + SSE) loading condition critical crack i
l
Enclosure to Attachment 4 TXX-88550 Julj 8,'1988' 3 g @ g J N @,
Page 1.of I l
em e J.r' 2 4 MB w=
h Page 5 of 8 Jun? 23, 1988 Mr. David Quinones RLCA 125 University Avenu' Berkeley, CA 94710
Dear Hr. Quinones:
In response to your recent inquiry, this letter gives the recommended procedure for determining the Ramberg-Osgood constints for austenitic steels using the EPRI estimation scheme given in Reports NP-1931 and NP-3507.
Based on the attached work of Dr. Sumio Yukawa, the low strain range (typically to 14 plastic strain), should be used to fit experimental stress-strain curves.
The low strain-range fits produced the best match between the EPRI estimation schenie and experimental JR curve data.
Sincerely, M.
Dougl M. Norris Managur Structural Mechanics Program DMN:cmb/3406SM81*
l l
l l
l 3412 Hi# view Avenue. Post Offee Box 10412. Paho Ano. CA 94303 Teleonone (415) 855 2000 Washegton OYce: 1019 Noe:eenm Street. NW Swte 1000. Washegton. OC 20036 (202) 872 9222
Enclosure to Attachment 5 TXX 88650 1
July 8,1988 Page 1 of 1 -
TABLE B-1-1 TYPICAL PIPE FITTING STEELS IN CPSES-1 SIS, RCS AND RHR LINES F_ITTINGS MATERIALS Elbow SA403 WP316 SA182 c-RF316 or F304 SA240 Type 316 SA376 Type 304 Sockolet SA182-F304 Reducer Type 304 Type 316 Reducing Tee SA376 Type 316 Waldolet SA182-F304 SA182-F316 Special Ends SA403 WP304 Check Valve SA182 F316 Nozzles SA376 Type 316 SA312 Type 304 SA182 F16N Nozzle Safe Ends SA182 Type S16 or 316L SA376 Type 316 or 316L Coupling SA182 Type F304 or F316 45' Reducer SA351 Grade CF8A 4
_ ~.
-