ML20210B260
| ML20210B260 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 07/19/1983 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML16340C148 | List:
|
| References | |
| FOIA-86-151 NUDOCS 8609180016 | |
| Download: ML20210B260 (6) | |
Text
. _. -. -
Appendix 4.5A STAFF REPORT EVALUATION l
Report
Title:
Independent Design Verification Program Shake Table Testing ITR #4 Revision 0 IDVP Designation:
P 105-4-839-004 Originator:
Robert L. Cloud and Associates, Inc.
INTRODUCTION The Interim Technical Report 4 (ITR 4) prepared by Robert L. Cloud and Associates (RLCA) for the Diablo Canyon Nuclear Power Plant Independent Design Verification Program (DCNPP-IDVP) has been reviewed by the staff.
ITR 4 summarizes the IDVP review of the seismic shake table testing of the Class IE electrical equipment and instrumentation at the DCNPP. The purpose of the review was to determine if the seismic testing procedure of PG&E and the service-related contractors conformed to the licensing criteria.
According to the Hosgri Report, shake table testing at DCNPP-1 was limited to items of electrical equipment and instrumentation.
Robert L. Cloud and Associates (RLCA) selected a sample of the items previously tested by PG&E and service-related contractors. This sample comprised seven groups of items tested by Wyle Labs.
RLCA reviewed the WYle grouping and testing sequence for electrical equipment and instrumentation. The review included both a field verification and a comparison of test results to the licesning criteria. As a result of the review, RLCA identified a concern and recommended additional reviews to address this concern.
SUMMARY
0F REPORT To obtain a representative sample, RLCA reviewed the list of Class IE electrical equipment and instrumentation qualified by shake table testing.
07/19/83 9
DIABLO CANYON SSER 1 8609180016 860908 PDR FOIA HOLMES86-151 PDR
-~
Seven groups of items tested at Wyle Labs were chosen as the sample.
- First, RLCA reviewed the test procedure Wyle used to test the seven groups of Class -
IE electrical equipment and instrumentation.
Each group contained items from a common plant location.
Second, RLCA verified the location of the electrical equipment and instrumentation included in all seven groups.
Third, RLCA developed " worst case" response spectra for each group. These spectra provide the highest seismic accelerations associated with the location of the group.
Last, RLCA made two response spectra comparisons.
The RLCA worst case response spectra (worst case spectra) was compared to the Wyle test response spectra (test spectra). The Wyle target test response spectra (target spectra) was compared to the test spectra.
The seven groups of items tested by Wyle Labs are located in the turbine and auxiliary buildings.
In cases where items appeared in more than one location (i.e., snap lock 1imit switch), Wyle placed them in the group with the highest Hosgri seismic spectra.
Because the equipment was grouped by location, a single target spectra representing the highest Hosgri seismic spectra associated with that location could be developed for each group.
The actual testing of these groups was done by mounting the items on the test machine table. A control accelerometer was mounted to the table to monitor the machine inputs. The equipment items were tested with simultaneous random biaxial vibration in two orientations:
side to side and vertical, and front to back and vertical.
The Wyle testing sequence included a minimum of five simulated operating basis earthquakes (OBEs) and two simulated safe shutdown earthquakes (SSEs). The
[
OBE levels were set at 60% of the SSE levels. The duration for each of the l
tests was 30 seconds.
I RLCA developed the worst case response spectra for each of the seven groups by enveloping the applicable Hosgri response spectra.
RLCA then compared the l
worst cases response spectra for each group with the Wyle test response spectra.
I According to the staff licensing criteria, the test response spectra must envelop the required response spectra by at least 10%.
07/19/83 10 DIABLO CANYON SSER 1
-,,.-r----n-,.-
-n.
,n.---,-,-m--,,
.-e------------------n.-
-.. - ~
l For each of the seven groups of items, RLCA plotted the horizontal (East-West and North-South) and vertical test spectra against the worst case spectra.
This was done by smoothing the Wyle test spectra through the minima.
An alternate smoothing method was used in cases where the test spectra that had been smoothed through the minima did not envelop the worst case by 10%.
This method consisted of visually examining the Wyle control accelerometer plots to determine if the curve representing an average of the minima and maxima would envelop the worst case response spectra by 10%.
The results of the comparison indicate that for Groups II, V, and VII, the Wyle test spectra for both the horizontal and vertical directions that RLCA smoothed through the minima enveloped the worst case spectra by 10%. Similar envelopment could not be established for Groups I, III, and IV.
However, using the alternate smoothing method, satisfactory envelopments were demonstrated.
For Group VI, neither of the above two enveloping procedures could be estab-lished. Specifically. the horizaontal test spectra does not envelop the worst case spectra for the 12.0 to 14.1 Hertz frequency range. TheIerticaltest spectra does not envelop the worst case spectra for the 7 to 10 Hertz frequency range.
RLCA further compared the W91e test ' response spectra for each group with the Wyle target response spectra, using the same method described previously for j
the comparison of worst case spectra, i
l The results of the comparison indicates that for Groups II through VII, the Wyle test. spectra for both the horizontal and vertical directions that RLCA smoothed through the minima enveloped the target spectra by 10%.
For Group I, althogh the same envelopment could not be established, it was found to be satisfactory by using the alternate smoothing method.
Four error and open item (EDI) reports, 2005, 1007, 1013, and 1049 were issued I
by RLCA as a result of the IDVP studies.
E0I 1005 and EDI 1007 both address a lack of documentation regarding formal transmittals of Hosgri spectra from i
PG&E to Wyle Labs. These open items were later closed because the Phase I 07/19/83 11 DIABLO CANYON SSER 1
. _ _ _ _,.,,. _ _. _ _ _ _. ~.. _. _ _ _ _. _,. _ _. _ _ _. _ _. _ _ _., _ _
1 plan specified that samples of the test response spectra were to be checked rather than auditing the transmittals of spectra.
a E0I 1013 reports test response spectra for Group VI that do not completely envelop the worst case response spectra, as stated previously. This item has been classified as a Class B Error. The RLCA comparison showed that the test response spectra completely envelopes the worst case spectra at all frequencies above 15 Hertz. The WYle sine sweep resonance search shows that the equipment has all resonant frequencies above 29 Hertz.
Since the test spectra envelopes the worst case spectra at all equipment resonant frequencies, this test was judged by the Independent Design Verification Program to qualify the equipment.
EDI 1049 was issued as the result of RLCA field verification.
RLCA found that the main annunciator typewriter was located on the control room slab above a column.
Except for the main annunciator typewriter, items located on the control room slab were placed in Group VII. The URS/Blume Auxiliary Report states that the lumped-mass model is applicable at the columns. The lumped-mass model is used for Group III. Therefore, the placement of'the main annunciator typewriter in Group III is acceptable. This item is therefore closed.
EVALUATION RLCA evaluated Wyle shake table testing by reviewing the W91e test procedures, verifying field location and developing a worst case spectra.
RLCA conipated the worst case spectra to the test spectra, and the target spectra to the test spectra.
Results given in the above form the basis for a concern with the target test response spectra.
Specifically, the target test response spectra for Group VI were incorrectly specified.
To assure that the target test response spectra were correctly specified for all items shake table tested by PG&E'and service-related contractors, RLCA has made the following recommendations:
1.
Confirm field locations of all equipment.
07/19/83 12 DIABLO CANYON SSER 1
-,,m
,--,p--m.,--
2.
Select the applicable Hosgri response spectra.
3.
Develop the worst case response spectra.
~
4.
Compare the worst case response spectra to the target test response spectra specified in the testing procedure.
CONCLUSION The staff finds the RLCA's independent review verified that the Wyle shake table testing of electrical equipment and instrumentation in the seven groups met the DCNPP Hosgri licensing criteria. The staff agrees with the IDVP findings and concurs in the recommendations as presented in the above. The staff further agrees that RLCA cannot yet evaluate the adequacy of all shake table testing until the recommendations are implemented by Diablo Canyon Project (DCP).
l 0
4 0
i 07/19/83 13 DIABLO CANYON SSER 1
-.... -.