ML19321A531
| ML19321A531 | |
| Person / Time | |
|---|---|
| Site: | Crystal River  |
| Issue date: | 07/16/1980 |
| From: | Baynard P FLORIDA POWER CORP. |
| To: | James O'Reilly NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II) |
| Shared Package | |
| ML19321A532 | List: |
| References | |
| 3-0-3-A-3, IEB-79-02, IEB-79-14, IEB-79-2, NUDOCS 8007230562 | |
| Download: ML19321A531 (7) | |
Text
'
Flodda Power July 16, 1980 File: 3 3-a-3 Mr. J. P. O'Reilly Director Office of Inspection and Enforcement U.S. Nuclear Regulatory Commission Suite 3100 101 Marietta Street, N.W.
Atlanta, GA 30303
Subject:
Crystal River Unit 3 Docket No. 50-302 Operating License No. DPR-72 IE Bulletins 79-02 and 79-14, Including Revision 1 and Supplements 182
Dear Mr. O'Reilly:
The attached report, prepared by Gilbert Associates, Inc., for Florida Power Corporation, is in response to the open items of IE Bulletin 79-14, identified in the NRC Inspection Report, 50-302/80-2, dated February 21, 1980.
It is the conclusion of the report that nothing has been found to indicate that the Crystal River Unit 3 is unsafe to operate.
The reanalysis and modifications of the piping supports is com-plete within the scope of Bulletins 79-02 and 79-14, and the modifications were accomplished completely during the current outage.
\\
THis DOCUMENT CONTAINS 8W&#@ s..QL POOR QUAUTY PAGES General Of fice 320i inirtf tourin streei souin. P O Box 14o42. St Petersburg F+cnca 33733 813-666 St51
Mr. J. P. O'Reilly Page Two July 16, 1980 If you have any questions regarding this report, please contact this office.
Very truly yours, FLORIDA POWER CORPORATION Patsy Y. Baynard Manager Nuclear. Support Services Simpson(IEBu179-14)DN-75-3 Attachment cc: Director Office of Inspection & Enforcement U.S. Nuclear Regulatory Commission Washington, DC 20555 Director Division of Operating Reactors Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555 i
FPC PESPONSE TO OPEN ITEMS OF IE BULLETIN 79-14 IDENTIFIED IN INSPECTION REPORT 50-320/80-02 SECTION 1 - GENERAL INFORMATION This report is in response to the open items identified in the NRC audit held on January ;4-25,1980, of NRC IE Bulletin No. 79-14, dated July 2, 1979, including Rev.1, dated July 18, 1979, Supplement 1, dated Aug-ust 15, 1979, and Supplement 2, dated September 7, 1979, for the Crystal River Unit 3 power plant.
It is the conclusion of this report that, having completed the review of the open items of the bulletin, the seismic analyses as originally done apply to the actual installed safety-related piping systems. Exceptions to this conclusion were of a type which do not compromise the safe oper-ation of the plant or for which remedial actions have been completed.
SECTION 2 - AUDIT OPEN ITEMS The four open items identified in the January 24-25, 1980, audit of NRC l
IE Bulletin 79-14 are as follows:
A.
Verification of material properties.
B.
Verification of valve type.
C.
New support loadings will be calculated for reanalyzed lines.
D.
Verification of the use of response spectrum curves.
SECTION 3 - OPEN ITEMS FINDINGS A.
Verification of material properties.
P rocedure A sample of the pipe fabrication spool sheets was selected to include one spool sheet fran each seismic analysis.
The ma-terial report code was obtained for one piece of pipe for each of these spool drawings.
The material report code provides traceability of the fabric & tion spool sheet to the material report sheet, which contains material specification data, in-cluding size, schedule, chemical and physical properties, heat treatment data, etc.
The material report sheet was then com-pared to the manufacturer's spool sheet for agreement with the piping material, outside diameter, and wall thickness.
The 1
manufacturer's spool sheets had previously been compared
.against the data used in the seismic analyses.,
SECTION 3 - OPEN ITEMS FINDINGS (Continued)
A.
Verification of material properties. (Continued)
Results j
Two hundred and one (201) analyses were checked.
Spool draw-ings or material codes could not be found for twelve (12) analyses. These twel <e analyses had low stress levels and are still considered satisfactory.
One hundred and eighty-nine (189) spool drawings checked showed complete agreement for outside diameter and wall thickness data.
In all cases, the type of material, carbon or stainless steel, was in agree-ment.
Twenty (20) spools listed different grades within the same type of stainless steel, which means the analysis is sat-isfactory, because the modulus of elasticity and allowable stresses are the same for the same type of stainless steel.
Three (3) spool sheets disagreed on the type of stainless steel used.
However, these analyses are satisfactory, since the analysis stress levels are below the allowable stress lev-els for the installed material.
B.
Verification of valve type P rocedure A sampling of valves was checked in the field to determine agreement between the as-installed valve versus the valve used in the analyses.
This sample included all five (5) valves that were dual-ordered due to late delivery of the original valves, and the four (4) valves that were listed in the Master Valve Listing as being replaced. The other valves were picked at random to give a total of sixty-two (62), which represents ten percent of the total number of valves incl uded in the seismic analyses.
Results The five (5) valves that were dual-ordered and the four (4) valves that were listed as changed were not the same as the valves included in the seismic analyses.
The actual weights and centers of gravity were close enough to consider the orig-inal seismic analyses still valid.
The remaining fifty-three (53) showed agreement between the analysis and the field-installed valve.
C.
New support loadings will be calculated for rea.ialyzed lines Procedure The existing hanger designs were checked with the new loadings generated by the sixteen (16) reanalyses found necessary by the IE Bulletin 79-14 review.
l SECTION 3 - OpEN ITEMS FINDINGS (Continued) 1 C.
New Support loadings will be calculated for reanalyzed lines (Con-tinued)
Results Of the two hundred and twenty (220) supports checked, fifteen (15) required adjustments and eleven required modifications.
The adjustments and modifications have been completed.
D.
Verification of the use of response spectrun curves in piping anal-yses Procedure The response spectra input decks were reviewed for being an accurate representation of the response spectra curves. Also, all two hundred and one (201) analyses were reviewed as to el-evation of piping and building locaticn to detennine if the correct response spectrum was used for each analysis.
Results Three (3) errors were found in the response spectra input deck.
These errors were determined to be minor by comparing the correct response spectra to the incorrect response spectra used and by examining the natural frequencies of the analyses involved.
)
Approximately half of the seismic analyses would have differ-4 ent response curves selected if today's conservative criteria of using the response curve corresponding to the maximum height of the piping configuration or enveloping two (2) building's response curves even if a line had only one (1) at-tachment in one (1) of the buildings.
The selection of the j
Floor Response Spectra Curves (FRSC) was recognized in the original perfonnance of analyses (1973-1975) as a critical in-put to the seismic dynamic modal analyses.
As such, it was included in the Design Review Checklist.
The philosophy in the selection of the FRSC for Crystal River Unit 3 was to se-lect tha one (1) that was most representative of the piping being analyzed.
The FRSC was viewed as a enaservative repre-sentation.of the response of the structures to an earthquake.
While attempting to be conservative, judgment was used in the selection of the FRSC in piping analysis so as not to intro-duce unnecessary conservatism of the FRSC in piping analysis so as not to introduce unnec assary conservatism arbitrarily.
The location of piping anchors (equipment nozzle or a fabrica-ted support) were considered to be the greatest contributors of the earthquake input from the structure into the piping 3 - - - -- - - - - -... -.
.m.
.m-
m system.
Linear supports, especially rigid rods, were consid-ered to have a lesser input.
Piping systems located entire-ly in one (1) structure, but having a singular attachment to a second structure (such as a building penetration), were sub-jected in most cases, to the FRSC applicable for the main structure, although differential building movements were eval-uated.
In reflecting on the CR-3 seismic analyses, it is felt i
that a more conservative FRSC could have been used, but it is also believed that the FRSC were chosen adequately.
Sample analyses (CR-22
-26, -52, -53, -90, -107, - 114, -127, and
-142) were re-run to evaluate these beliefs.
It was shown that the results for FRSC CRA1 (Auxiliary Building elevation 119') are almost identical to those for FRSC CROS (grcund re-sponse).
(Many times FRSC CROS was used, when, by today's standards, FRSC CRA1 would have been chosen.)
Other sample runs were made using a FRSC one (1) elevation higher than the one originally selected.
These runs also showed a sma (10%
to 18%) increasein stress levels and support loads.
The reason for these negligible or small increases appears to be:
(1) for the first case, the CR05 FRSC is handled in a differ-ent manner than the other FRSCs.
Its vertical response ic set equal to the horizontal responses, whereas, for all other FRSCs the vertical response is set equal to two-thirds of
>e horizontal responses; (2) for the second case, the responses for different levels within the s me building vary mainly at the peaks, for only a small range of frequency values.
A third sample run was made where only one (1) FRSC was used, 1
while present-day criteria indicates that an envelope of two 1
(2) FRSCs for two (2) different buildings should have been 1
used.
These runs were made using the FRSC from a structure that wasn't originally included and showed the stress levels and support loads to be lower, indicating that the original curve selected was the more severe of the two.
In conclusion, it is felt that all the seismic class piping will function as designed during a safe shutdown earthquake event for the following reasons. Although a more conservative criteria is. used today to select FRSCs, the met'iod utilized for CR-3 was deemed satisfactory as demonstrated by the sample reanalyses performed.
Also, many conservatisms were included in the CR-3 seismic design which are not used today.
One of these is that the FRSCs were constructed using 1/2% equipment
~
damping, while today,' 1, 2, or 3% equipment. damping is used, depending on pipe size, and whether an OBE or an SSE analysis
-'c being performed. Conservatism also exists in that present-l day response spectra do not use SSE response as twice ~the mag-nitude of OBE response.
This, in reality, means that the SSE
.is approximately 1.6 OBE. Moreover, as shown in Attachment A, CR-3 used -double the 0BE stresses and pipe t rdport loads to compare to 1.2 S, whereas, today's methodology compares the h
OBE stress levels to 1.2 Sg and the SSE stress level to 1.8 S -
h 4-
e SECTION 4 - CONCLUSION Nothing was found in the verification of material properties, valve types, or response spectra used, to indicate that the CR-3 plant is not safe to operate.
The pipe support modifications due to reanalyses is complete within the scope of Bulletins 79-02 and 79-14.
l j
.