ML20091C891
| ML20091C891 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 07/20/1983 |
| From: | Mertens J AFFILIATION NOT ASSIGNED |
| To: | Martin J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| Shared Package | |
| ML20091C867 | List: |
| References | |
| TAC-52491, TAC-52492, NUDOCS 8405300680 | |
| Download: ML20091C891 (63) | |
Text
EXHIBIT A
- s. .
. JOHN MERTENS Mr. J. B. Martin Regional Administrator U. S. Nuclear Regulatory Commission Office of Inspection and Enforcement Region V 1450 Maria Lane, Suite 210 Walnut Creek, California 94596-5368 July 20, 1983
Dear Mr. Martin:
The purpose of this letter is to bring to your attention technical inadequacies in safety related piping systems and other recent incidents at the San Onofre Nuclear Power Station.
They are itemized below. For a detailed discussion please see
.the remainder of this letter and the attached copy of an engineering evaluation I recently completed.
- 1. Evidence indicates that Unit 2 operated for 21 months with inoperable snubbers on the safety related main feed water line FW 189- (from March 1981 through February 1982 with six, and from March 1981 through December 1982 with five internally damaged snubbers).
- 2. Inspections of these snubbers carried out over those periods failed to detect their inoperability.
3 An attempt was made to obstruct the engineering evaluation I conducted.
- 4. The method presently used for inspecting snubbers is in-adequate and misleading.
=
8405300680 840427 PDR ADOCK 05000361 '
0 PDR 3109 S. El Camino Real San Clemente, CA 92672. (714) 498-6311 y
5 No effective QA program exists for monitoringsnubber performance as per. Technical Specification page B3/4 7-6, Snubber Basis.
- 6. A reconmandation to install minimum instrumentation for monitoring operability of snubbers in safety related piping was dismissed.
7 The engineering evaluation performed by Bechtel Power Corporation to determine the cause of failure of these snubbers is inadequate and misleading.
- 8. Snubbers on the main feed water lines inside the con- >
tainment were not designed for dynamic loads, they are undersized and may not be able to ensure structural inte-graty of the main feed water lines.
9 An attempt. was made to intimidate me (transferred or fired) in response to my questioning BPC's performance at Songs 2 & 3
- 10. The original and all copies of NCR S01-P-1308, Rev. 1, were destroyed.
As to my qualifications for reporting such matters, I am a station engineer here at Songs., I am 58 years old and have extensive experience in design and~ development work, and failure analysis.
One of my.recent assignments was analysing why five snubbers on the main feed water line, FW 189 Unit 2, inside the containment at elevation 63', hanger locations H 010, H013, H017 had become inoperable. Suspected inoperability was first reported on December 8,1982.
I)
,. . . . . _ .- . =. -- .
_3_
I had the five snubbers shipped to the manufacturer, Pacific Scientific, for disassembly and inspection. Their internal
. parts were found severely damaged. The sixth snubber, which was the twin of one of the damaged ones, had been stroked at the time the others were removed and found operational.
This, however, seemed irreconcilable in the light of the severely
, damaged five snubbers. There were two explanations: one, the sixth snubber was not it: place when the others were damaged, and second, the sixth snubber had somehow survived the destruc-tive event.
It became necessary, in my judgement (T/S 3/4.7.6 (g) ist pgr.), l to remove the sixth snubber for two reasons: first, to ascertain !
operability through functional testing and internal inspection l.
by the manufacturer, thereby ensuring Unit 2 was not operating 1 with a damaged snubber on the main feed water line, and two, if the sixth snubber was found operational, it would be a clue
, as to when the damaging event had occured.
My immediate supervisor agreed with this reasoning. I submitted then a work order to Startup Maintenance Support requesting removal of the sixth snubber. It was cancelled. And so l were two further work orders. As to the fourth one, some one I called Startup Maintenance Support in my name and instructed
- them to cancel it. I then submitted a new NCR to accomplish I removal of that snubber. Yet, even that was recommended for I cancellation by Project personnel. Only on my insistance was that sixth snubber removed. It was taken to the manufacturer l for functional testing and internal inspection. It was found f' to be operational. This was convincing evidence that the sixth anub'oer had been installed after the damaging event.
Researching records suggested that the original sixth snubber (SN 2609) was replaced by SN 4322 on February 1, 1982. I say t
6 4
, .__.._m__ __._ _ - . , . _ . - _ . , _ _ _
, , . _ _ - . _ _ _ , _ _ _ , _ s
suggested- because NCR S023-F-463 required rejection of the damaged snubber, but the respective work order ( 2113) reports only that a new tag was attached. No record of SN 4322 appears until 12-23-82 on work order 19773 On 4-26-83 snubber SN 2609, the original sixth snubber, was retrieved from the Bechtel Warehouse. An examination showed that it was as badly damaged as the other five. This con-firmed that the damaging transient had occured before February of 1982, and that the causing event was the Waterhammer Test 2HA-201-01 of January 30, 1981.
A final test conducted at the Pacific Scientific Laboratory revealed that the PSA-10 snubber fails under shock loading at 37,000 lbs. But more important was the revelation that the destruct tested snubbers showed no visible evidence of failure, and further, they could be stroked without in'dication that the internals were damaged.
The disturbing element in these conslusions is the implication that a) Unit 2 operated for 21 months with damaged snubbers while going through modes 4 - 1, b) visual in'spections of ,
snubbers are unreliable, c) QA and maintenance work as related to snubbers is inadequate, d) no effective means exist to detect damaged snubbers.
. The need for a better method of curveillance for snubbers became obvious. Therefore, I submitted a written recommendation to install minimum instrumentation on the main-feed water lines.
for monit'oring snubber conditons, transients, and for alerting operating personnel when a transient had occured. With such instrumentation operability of snubbers and respective systems can be checked quickly. As of today no such instrumen-tation has been considered as neccessary.
b I
e
Another matter, which is of personnel concern, arose when I questioned Bechtel's performance here at Songs. I discussed this with other engineers who advised me that previous
~
questioning of Bechtel's performance had been countered by SCE management with a message that had quenched further
. inquiries. I decided,therefore, to pursue.the matter as a stockholder. After an exchange of fruitless correspondence with the General Office the technical station manager called me to his office and told me on instructions from top management that if I continued my inquiries into thes,e, matters I would either be i transferred or fired. A' note that I had been counseled in i certain matters was put into my performance report ',se attached copy).
Because of these incidents I request that you take appropriate action. I also recommend for your consideration the exclusion of BPC from performing any engineering failure evaluations for Units 1, 2 and 3 at San Onofre Nuclear Power Generating Station because of conflict of interest. All investigative avenues that were pursued for the attached engineering evaluation were open to BPC. They chose to ignore them.
Please inform me as to what action you will thke. .
Sincerely John Merten k,,e 4 c & b (.' n -*
/.
Enclosure's 6
- T fl
- EXHIBIT B Southern California Edison Company -
~.A.~. :
4 RO.90x8CO 2244 WALNUT G RCVC AVENU C OAvlO .J. FOGARTY T E L E ** *C N C i n ce ~, u .. *" "' ' ' ' ' " ' * *
- July 27, 1983 Mr. John Mertens 3109 South El Camino Real ,
San Clemente, California 92672 Re: San Onofre Nuclear Generating Station, Units 1, 2 and 3
Dear Mr. Mertens:
Your letter to Mr. Gould concerning alleged technical ,
inadequacies in safety-related piping systems, dated July 20, 1983, was received on July 22 and referred to me for evaluation l and response. You requested a response by July 27, 1983, concerning whether or not a letter from you to the Nuclear !
Regulatory Commission (NRC) contained false statements. {
The time available to respond to you is very short, and ;
I request that you give us until August 5, 1983, to respond so that our response can be complete and the details can be i checked. In addition, I urge you to follow the procedure described in Mr. Gould's letter dated July 11, 1983 (copy !
attached), for pursuing concerns about nuclear safety. (This !
letter updated Mr. Gould's letter of September 13, 1982, which l described this procedure to all personnel.) I have directed the :
Onsite Review Committee to proceed with review of your nuclear i safety _ concerns in any event. Please let me know if we may have until August 5 to respond and if you will agree to pursue your l nuclear safety concerns in accordance with Mr. Gould's letter. {
Nevertheless, a response to your request has been !
developed in the time available. Attached hereto are our comments on the content of your proposed letter to the NRC. A best effort has been made to investigate the facts and circumstances involved; however, the response must be considered preliminary.
In summary, as described further in the attachments, the technical matters related to your nuclear safety concerns are adequately identified and documented within our design, testing and quality assurance programs at San Onofre.
Sincerely, rm -23 -
.N,hsM, D. J Fogakty Executive Vice President DJF:CRK:dkg i Attachments
- i i
Southern California Edison Company wd5 R O. Box 3 00 2244 WALNUT GRCVC AVCNUC witLIAM se, C OULO AD, CAW a tA WO
........ .......... tttce=o=c July 11, 1983 ' ~ ' ' "
TO ALL PERSONNEL ADMINISTRATION ADVANCED ENGINEERING ENGINEERING & CONSTRUCTION '
FUEL SUPPLY-NUCLEAR ENGINEERING AND OPERATIONS POWER SUPPLY SYSTEM OEVELOPMENT SUSJECT: Review Process for Nuclear Safety Concerns Purcose ,
i To remind personnel of the existence of a review process which is available to address nuclear safety questions and concerns. l t
Discussion 5 i
Since 1957 when the Company received an operating license for San Onofre Unit 1, safety review organizations have been in effect whose functions ,
include review and initiation of action to solve problems related to nuclear i safety at San Onofre. These organizations now include the "On-Site Review Committee" located at San Onofre and, at the General Office, the Nuclear Audit and Review Committee" for Unit 1 and the " Nuclear Safety Group" for Units 2&3.
These organizations are available to consider questions or concerns related to .
nuclear safety from employees who become aware of existing or potentially serious problems. :
I l
In keeping with long-standing practice, should you become aware of a nuclear l safety question related to a Company facility, the matter should be brought to the attention of your supervisori He or she will, in consultation with the !
proper so request.
personnel, resolve the question and advise you of the resolution if you !
If you do not feel the question or concern was satisfactorily 'j resolved,
. organization.
you should notify the head of the appropriate safety review i For operations and maintenance personnel working at San Onofre i this is the Chairman of the On-Site Review Committee and for other personnel -
this is the Chairman of the Nuclear Audit and Review Committee for Unit 1 matters, or the Manager, Nuclear Engineering and Safety for Units 2&3 matters. ,
~
. i f, ,-._..-,..c.-,,--,,--y . , - - , , - - , e, , , --,-_,--m e,-,em . . , - - - . , r-,-y , - - ~ - - - . - , - - - ,
In ordsr to insure your concern is properly understood and that you are informed of the outcome, this notification should consist of correspondence from you describing your concern.
Following review of your concern you will be notified of the outcome. You should also be aware that the records of these actions are subject to audit by the Nuclear Regulatory Commission, and the disposition of your concern will be made a part of those records.
As you know, we are relying on nuclear generating stations for a significant share of our future generating capacity and are dedicated to maintaining our exemplary nuclear safety record. To accomplish this, it is important that potential nuclear safety questions be identified and promptly resolved. Each employee involved in the Company's nuclear program should consider nuclear safety and compliance with NRC regulations as the first priority in executing their duties.
This procedure is in addition to and not in derogation of any rights or obligations provided under-the regula~tions of the Nuclear Regulatory Commission.
_I b - -
e
COMMENTS ON PROPCSED LETTER DATED JULY 20, 1983, FROM JCHN MERTENS TO J. B. MA RTIN , USNRC The proposed letter refers to ". . . technical inadequacies in safety related piping systems and other recent incidents at the . . . " San Onofre Nuclear Generating Station (SONGS) . Comments, based on a preliminary review of information immediately available, are provided below for the items listed in the proposed letter to Mr. Martin. ,
Item 1: Evidence indicates that Unit 2 operated with several mechanical snubbers not detected as inoperable in main feed water line FW189.
Comment: As developed in Mr. Mertens' report to ,
Mr. Katz dated May 23, 1983, this certainly did occur !
for some period, and it is . likely that the mechanical !
shock arresters (snubbers) discussed were initially damaged in the March 21, 1981, startup test indicated. ;
It should be noted that operability of main feed water
{
line FW 189 was first required for Unit 2 when it , l entered Mode 3 on May 18, 1982. .
The startup testing program for Unit 2 includeo -
monitoring this line for thermal expansion, and it noted an anomaly which resulted in snubber inspection at the 50% power plateau in December, 1982. This inspection l did identify snubber damage and led to the replacement !
of five snubbers at that time. Replacement of the first l
.two of these was reported to the NRC in LER 82-165. I
'This LER is being revised to include the three additional snubbers replaced during that same time.
Item 2: Inspections of mechanical snubbers in main feed water line FW189 between when the damage probably occurred in {
March, 1981, and when it was discovered in December, j 1982, did not detect the inoperability.
Comment: Pre-fuel loading snubber inspections in January, 1982, did identify two damaged snubbers on this line, which were replaced.
Damage to other snubbers was ,
not identified until December, 1982. During the period ;
from January to December, 1982, a number of plant i transients occurred which could have caused, or ;
increased, damage to these snubbers. j
I Item"3: An attempt was made to obstruct the engineering evaluation assigned to Mr. Mertens.
Comment: This appears to relate to difficulty experienced in March, 1983, in removing the snubber installed in January, 1982. As indicated, Mr. Mertens'
. supervisor supported this effort, which was successful when the appropriate documentation (an NCR) was prepared. This snubber was found to be undamaged and operable.
Item 4: Mechanical snubber inspection methods are inadequate and misleading.
Comment: . Snubber inspection methods include periodic 100% visual inspection and stroke testing of selected snubbers. In addition, a sample of the snubbers is periodically removed for force testing by machine. The sample is enlarged when problems are noted. Each of these inspection methods is capable of identifying deficiencies, as demonstrated by extensive experience.
Item 5: No effective quality assurance program exists for monitoring snubber performance in accordance with the basis for snubber operability described on page B 3/4 7-6 of the Technical Specifications.
Comment: Page B 3/4 7-6 of the Technical Specifications discusses the requirements for monitoring the service life of snubbers subject to environmental degradation (i.e., hydraulic snubbers).
I Presently, an effective program, implemented by ,
procedures, is in place to monitor both hydraulic and mechanical snubbers. This program consists of both a '
visual and functional test to ensure proper snubber l performance. A computer-based system which records the maintenance history of all snubbers is utilized to input ,
information into the snubber surveillance program. Any failures that are identified during these inspections are subjected to a detailed engineering analysis and, if determined necessary, an increase in the frequency of inspection or the number of snubbers inspected is in.plemented.
Item 6: A recommendation from Mr. Mertens to install minimum instrumentation for monitoring the operability of snubbers in safety-related piping was dismissed.
[ Comment: Mr. Mertens participated in a meeting between representatives of NUS Corporation, the Electric Power 9
L
Research Institute (EPRI) and SCNGS in early June, 1983, which included discussion of this recommendation.
Pursuant to that meeting, SONGS expects a proposal from NUS in early August, 1983, which will include the possibility of use of monitoring instrumentation. Also, on July 12, 1983, Mr. Mertens' report of May 23, 1983, which includes this recommendation, was forwarded for evaluation by Bechtel.
Item 7: The engineering evaluation performed by Bechtel to determine the cause of failure of snubbers in the main feed water line FW189 was inadequate and misleading.
Comment: The evaluation performed initially was primarily concerned with verifying that the line was not overstressed as a result of the event which damaged the snubbers. This was established in an adequate and straightforward evaluation. The initial Bechtel evaluation did not definitely establish the cause of the i event, and Bechtel was requested to perform further evaluations in letters from the SCE Project Engineer {;
dated February 3, 1983 and April 7, 1983. Funding for this further evaluation was approved by SCE on May 18, 1983, and the work is ongoing. As the work is not complete, it is incorrect to characterize it as inadequate and misleading.
Item 8: Snubbers on the main feed water lines inside the containment were not designed for dynamic loads, they are undersized and may not be able to ensure structural integrity of the main feed water lines.
Comment: Main feed water lines inside the containment, including their supports, are designed in accordance with the ASME Boiler and pressure vessel Code,Section III, Class 2. The design has been verified as correct.
As indicated in the design basis (FSAR paragraph 10.4.7.1.d), it does not include dynamic loading resulting from severe water hanmer transients, as these are to be avoided by operational and design measures.
The startup test on March 21, 1981, which is hypothesized to have caused the damage to the five snubbers which was revealed by the inspections of ,
December, 1982, and to the two snubbers which were replaced in January, 1982, was substantially more severe than is included within the design basis. Further, following steam generator feed ring modifications in I mid-1981, a revised startup test was successfully l i
conducted in June 1982 which confirmed the adequacy of ;
the main feed system design basis. In summary, the -
i snubber damage probably occurred during startup testing, !
4 t
L 0
which included a severe water hammer transient. Should such a transient beyond the design basis reoccur during plant operation, careful inspection and testing of affected snubbers would be performed including consideration of this startup testing experience. Even so, analysis shows that feed water line FW189 was not overstressed by the event which damaged the snubbers.
Item 9: An attempt was made to intimidate Mr. Mertens (transfer or dismissal)*in response to questoning Bechtel's performance at SONGS 2 and 3.
Comment: No attempt was made to intimidate Mr. Mertens. In an exchange of correspondence between Mr. Mertens, SCE and others commencing in November, 1982, he requested:
o Total expenditure data concerning SONGS 2 and 3 o Payments to Bechtel for engineering work on SONGS 2 and 3 o A copy of the contract between SCE and Bechtel for !
SONGS 2 and 3 i t,
o The same information as above for SONGS 1 !
Mr. Mertens pursued these documents and data as a f stcckholder with the Secretary's office, rather than as an employee. In his letters, he included the comment {
that the reason for his request " . . . is to determine
~
whether or not the [ contract] contains provisions for redress in the event cf unsatisfactory performance, and f
'to study the terms of such provisions." At no time" i prior to receipt of his letter to Mr. Gould dated j July 20, 1983, were questions of nuclear safety raised :
in this correspondence. !
l Mr. Mertens was counseled in February, 1983, to stop his repeated attempts to obtain the information above, as i SCE has declined to provide it. This counseling did not i adversely affect his overall performance rating in his March, 1983, performance appraisal, and he was not t threatened with transfer or dismissal
- i l
Item 10: The original and all copies of NCR sol-P-1308, Rev. 1, ;
were destroyed. ,
I Comment: Revision 0 of this NCR for SCNGS 1 was opened on December 8, 1982. The condition described relates to corrosion of a heat exchanger foundation and is also l
i
__ . _ _ -- . _ ._ __= ___ _
described in an NRC inspection report transmitted to SCE by NRC letter dated January 7, 1983. The work required to' repair the foundation is ongoing and is about 80%
complete. Revision 1 of the NCR was drafted to i readdress the original evaluation of heat exchanger operability which had been made for Revision O. Some difficulty was experienced in deciding how to treat operability during the repair. It was concluded prior to validating the NCR revision that readdressing operability was not necessary, and the repair could be made with the heat exchanger in service, so the revision was not completed. Revision 0 of the NCR remains in effect, and it governs the work which is now nearing completion.
In addition to the ten items identified and commented on above, the letter discusses Mr. Mertens' assignment to investigate the cause of the failure of snubbers in main feed water line FW189. This assignment was made December 23, 1982, as i
a result of a station incident report written to cover the l matter. The assignment resulted in his report to Mr. Katz dated l" May 23, 1983, which was forwarded for evaluation by Bechtel via a startup problem report dated July 12, 1983. At the same time, as the result of a series of exchanges between SCE project i engineering and Bechtel, further evaluations of the cause of the !
event resulting in the snubber damage was approved on May 18, l' 1983, and is ongoing.
! i l
i t
l i
I i
I i
u 1
-es o- eo s eo e -e m em _
- _ , , ,
I e
i
, EXHIBIT C
\
Southern California Edison Company L., -
S A N ONOF R E N U C L E A R G E NE R /. TIN G S T A TION P.O. B OK 128 5 AN C LEMEN T E. C A LIF CR NI A 92672 H. O. R A Y iELEpwCNE S r.i ON u.N.G' a August 5, 1983 <" " m a Mr. John Mertens 3109 South El Camino Real San Clemente, California 92672 Re: San Onofre Nuclear Generating Station, Units 1, 2 and 3
Dear Mr. Mertens:
Mr. Fogarty's letter to you dated July 27, 1983, provided our initial response to your letter to Mr. Gould dated July 20, 1983. Attached hereto for your information is our more complete response, as described in Mr. Fogarty's letter.
Also, the Onsite Review Committee (OSRC) at San Onofre has met to review those of your concerns which relate to nuclear safety. We have had the benefit of your participation in clari-fying those concerns.- OSRC has provided input to the attached response and will review the conclusions of actions which are not yet completed.
In summary, we conclude that adequate means exist to ensure that your concerns which relate to nuclear safety are identified, documented, evaluated and appropriately resolved.
The OSRC will continue its oversight, and you are encouraged to bring additional information to the committee and to participate in its further review.
Sincerely,
/ '- -
g r
Attachment .
9 jf Ah 33
O I.
SUMMARY
In a letter to Mr. W. R. Gould, dated July 20, 1983 (Attachment 1), an employee of Southern California Edison Corapany described certain concerns regarding safety-related piping systeras at-San Onofre Nuclear Generating Station. Southern California Edison provided a' preliminary response to those s
concerns in a letter, dated July 27, 1983 (Attachment 2), from Mr. D. J. Fogarty to Mr. J. Mertens, the involved etaployee. That letter ,
indicated a complete response would be provided to Mr. Mertens by August 5, 1983. This report provides that complete response.
In order to provide the background necessary to understand the concerns raised by Mr. Mertens, the operational history of the raain feedwater piping in question is discussed in Section II. Following the background inforiaation, ,
each of his concerns is repeated verbatim in Section III. SCE's responses are provided after each itera. -
On July 28, 1983 and August 4,1983, special Onsite Review Comraittee meetings, t
,A attended by the NRC, were held so that Mr. Mertens could present his concerns i as they relate to nuclear safety. Section,IV of this report repeats the. l
, i l concerns as they were described in these meetings (References 1 and 2), and I provides the results of the corrnaittee's review.Section V suiaraarizes those
- activities still in progress at the time of this report. ;
1 4
I
II. OPERATIONAL HISTORY OF MAIN FEEDWATER PIPING i
1 The operational history of'the main feedwater piping begins with the NRC's review of the piping design criteria. As stated in the Final Safety Analysis, paragraph 10.4.7.1.D, the system is not designed for aonormal hydradlic loads since design and operating considerations preclude their occurrence. The NRC requested that a special test be added to the startup prograta to verify the F5A design criteria in that unacceptable waterhammer would not occur as a result of uncovering and draining the feedring (refer to Fi 3ure 1 for a sketch of the steam generator internals). In March 1981, When this test was performed on steam generator E088 and feedline FW189, a thud was heard and attributed to a check valve slamming open. No damage to piping or supports was noted during a subsequent visual inspection of the piping. Later, in July l'981, during a scheduled generator internals inspection, the feedring was e found to be damaged. The cause of failure was attributed to the dynamic differential. pressure forces applied to the feedring due to rapid injection of .
I auxiliary.feedwater into the drained feedring. The design was uodified to {
increase the time required to drain the feedring, to reduce the magnitude of the differential pressure dynamic loading, and improve.the capability to withstand compressive loading. A supplemental feedring integrity test with less severe conditions was scheduled for later in the startup program to verify these modifications. This test was conducted in June 1982. No indications of waterhanaer, such as noise or vibration, were observed during l the test. A subsequent' visual inspection of the feedring, tne feedwater piping, and piping supports showed no damage. , !
I i
s
& 9 s
s
=
I FIGURE 1
?
LOCATION OF FEEDRING IN STEAM GENERATOR -SONGS 2& 3 STEAM NOZZLE i I
- , O STEAM DOME STEAM DRYER ^
(TYPICAL) O -
-< UPPER LEVEL MOISTURE SEPARATOR -
n .
TAP (100%)
(TYPICAL)
J FEEDRING RISER TOP OF "J" TUBE TON FEEDRING FEEDWATER NOZZLE
\ \.r e r- *3 _'p / 46.8%
(_
,*3 EL. 71' 2'3/4" /
+
FEEDWATER LINE #
7 BOTTOM OF FEED-TOP OF 1 RING 36.51 EL. 66' 11" / TUBES p JUNCTION 23%
-36.7%
< LOWER LEVEL TAP ( O tt )
DOWNCOMER EVAPORATOR REGION WRAPPER -" TUBES erator Feedring i \ .
i 6
\-
J-Tubes
. , SECONDARY SIDE l
, PRIMARY SIDE
-m Thermal N Sleeve TOP VIEW OF FEEDRING HOT l tee
/ 's j
, ".-~-
i I In J.anuary.1982, a pre-fuel load visual inspection was conducted of all safety-related snubbers. Two mechanical snubbers on FW1J9 failed this inspection when they would not rotate in place. The failure was attributed to an installation error.and the affected snubbers were replaced. (Figure 2 is a
/ .
diagram of FW189 inside containment, and Table 1 lists the history of the raechanical snubbers'for FW189.)
In May 1982, during post-core hot functional, testing, excessive vibration was L
noted in the auxiliary'feedwater giping inside ind outside containment on threeseparateoccasjons. A special test program was conducted in June 1982 toconfirmthecauseofdh[evibrations. Vibration'raagnitudes of the auxiliar/
1
- feedwater piping inside cont.ainment; were coaitored and found acceptable. The
- y i, A e f.
. program (see Attachment 3) confirmed the cause of, the vibrations was not steara
. generator waterhara er, but das caused b hydraulic instabilities induced by
~
backflow through'a 2-inch Y-type globe valve located outside contairuaent. '
Q fi e On November 9, 1982, the reactor tripped from 20% power due td low steau j .
Steara generator lebel caused. by;, loss of feedwater' ccafrol 4 Reference 3).
~
generator E089 levei dop;ied below the feedring'for approxiuately 10 uinutes.
i Level was restored utiilisg auxiliary feedwatcr pumps. These transient
' ~
conditions were bounded by,,the conditions of the supplemental feedring integritytest/describedabdve,thus,
, i .;
fur,therigvestigationintothiseventis
-not required. '
C ',
- 7. .
, .Y s
- i On November 13, 1982 and sovember 26,1932,Jtne reactor tripped froia high J -
steam generator level'(References 4 and'5). Steara genefator feedline waterharaaer was not e doncern in either of these events since the feedring was not uncovered. . '
e
/,
J
/,,
3-
FIGURE 2
- l o r= .s s..e. ,
4 80eet my M oss k" H
$7 t ang
- ,f N$"*"' ,
" o* J 1 -
O w
- G I
M ai"it C4 *= - se,. n *
-oe
~
e4 + oe p y r
w -... w s '
r1 - . 1 Jg o . .. ... g
/%
Ws '
g a e- .. o JW't.. .
o '/
(>
O*i1 No 503 018
%E E T 2 Of g y
- .'q-f.P . ,
/
c
.,.I APPROVED O REFERENCE DESIGN B
-....) . . . ~ -
og w .
v ,i . .av..
2
=. G m sntg.us == n.s;.r,r:c.
v.==u --.;n4im m yg3 u ..... .i.s r. ,,,,6...
n- ea.a 1 l ,' - ___piu -. :
p =.w..,,, . .= C~m s_._,..
J. m__~~.:-
. ., p, q' 'z!#.-
r...= m .
-e..
d .?A..J ee .o s
, <n. a on rA c= .
-(-
v., a
.. umt z- ret w stow wc ar t etencva.ve.u<w m Iu' ~. M==Ae- a'or coaien wm 'ato n's A
" " ' " ~
8 6 7 3 - ,i Or..ri -l~ f=-e o 9976 fos 098 od 6 5 5 L 8
3 i 2 L g at< is
. . . _ _ _ _ . . _ _ _ _ _ _ _ . _ _ , . . . _ _ _ _ _ . ~ _ _
TABLE 1 MAIN FEEDWATER LINE S2FW189 - MECilANICAL SNUBBER IIISTORY Initial Snubber January 1982 December 1982 April 1983 Location Serial Number ' Inspection Inspection Inspection ,
S2-FW189-IIO10 2609 2609 replaced 4322 4322 replaced .
(Dual Snubbers) with 4322 with 15371 .
2603 2603 2603 replaced 11077 with 11077 S2-FW189-012 390 390 replaced 1145 1145 with 1145 S2-FW189-11013 2113 2113 2113 replaced 3472 (Dual Snubbers) with 3472 2107 2107 2107 replaced 268 with 268 S2-FW189-il014 3098 3098 3098 3098 S2-FW189-II017 4357 4357 4357 replaced 11086 with 11086 4352 '4352 4352 repinced 11085 with 11085 Notc(*): Snubbers found to be damaged' L- - - - - - -
J0n December 9, 1982, while performing the feedwater piping tneraal expansion test'at the 50% power plateau, a mechanical snubber at location S2-189-H013 (see Figure 2) was found damaged. The initial engineering investigation (Attachment 4) concluded that the remaining mechanical snubbers, five additional for a total of six in question, at locations S2-189-H013, S2-189-H010, and 52-189-H017 were prob ~ ably also damaged. Eventually, the original damaged snubber and four others were replaced. The sixth, on support 52-189-H010, was manually strcked with satisfactory results and was not replaced.
.s \
Table i summarizes the status of each snubber at these supports. At the time of the event, a preliminary investigation into the failures postulated the cause to have been the November 9, 1982 transient discussed above.
As part of the followup review of this occurrence (Reference 6), Mr. Mertens was requested to evaluate the cooldown in light of the snubber damage report, confirm the postulated cause, and make recommendations for plant or procedural modifications to preclude recurrence. During this investi 3ation, Hr. Hertens had the five failed snubbers, found in December 1982, shipped to the manufacturer for failure analysis. The manufacturer reported the cause of l
failure as gross overload (report included in Attachment 1). As a result of l
this report, Mr. Mertens reconaended removal and inspection of the single I snubber not replaced at location S2-189-H010. That snubber was replaced, sent to the manufacturer and passed a functional test and internals inspection.
i 1
1 l i 1
l
~
Mr. Mertens reported the results of his investigation, the test results discussed above, and his conclusions in a memorandura to the Station Technical Manager, dated May 23, 1983 (included in Attachment 1). This memorandura was forwarded to Bechtel Power Corporation for further evaluation on July 12, 1983 (Reference 7). The evaluation of his memorandum is contained in Section III of this report.
f l
l l
l r
< i
'III. RESPONSE TO CONCERNS STATED IH PROPOSED LETTER DATED JULY 20, 1933, FR0tt JOHN M2RTENS TO J.-B. MARTIN, USNRC The proposed letter refers to ". . . technical inadequacies in safety related piping systems and other recent incidents at the . . ." San Onofre Huclear Generating Station (SONGS). The responses are provided below for the items listed in the proposed letter to Mr. Martin. '
s Item 1: " Evidence indicates that Unit 2 operated for 21 months with inoperable snubbers on the safety-related main feedwater line FW 189 (from March 1981 througn February 1982 with six, and from March 1981 through December 19d2 with five internally dauaged snubbers)."
Reply: The initial evaluation of the inoperable snubbers found in December 1982 postulated the cause to be a transient occurring about one month earlier.
A more extensive evaluation of this event has been undertaken. Based on a review of the following data and as stated in our LER 82-165, Revision 2, it can now be concluded that initial damage to the snubbers on feedwater pipe FW189 occurred prior to the November 1982 cooldown events and probably as a result of the steam generator feedring integrity test conducted in March 1981:
Item 1:' (Continued)
- A. Dates of snubber inspections and replaceraent, and dates of linear main feedwater piping hydraulic transients discussed in Section II of this report.
B. Piping therraal expansion measureuents, recorded per piping Thermal Expansion Test 2PA-102-01
. raatch analytical results more closely than those taken at 20% power prior'to snubber replace;aent in December 1982.
C. The other raain feedwater pipe (RJ-190) connected to the steam generator that did not undergo the feedring integrity test has exhibited acceptable thermal movements throughout the test program.
Both raain feedwater lines are moving as predicted by original design calculation since replacement of the damaged snubbers on line FW-189.
Additionally, since the calculation stresses as result of the event were found to be acceptable as discussed later in this report, no unacceptable loads were imposed on the piping system.
l
q
, I. tem li .(Continued) i Discovery of an inoperable snuocer on uain
- feedwater piping inside containiaent is reportaole 3 to the NRC. Therefore, the two priraary
< methodologies for determination of reportability y and operability, the Non-Conforiaance Report (NCR)'
r and the Station Incident Report (SIR), were 7 reviewed for proper application to feedwater pipe 1 support probleras. For the December 1982 snubber 9 failure, three NCRs were issued and correctly v classified as reportable and not operaole. The
.O engineering evaluation required by Technical
? Specification 4.7.6 9 was specified in these NCRs r) and perforiaed by dechtel Power Corporation. A ty Station Incident Report was also prepared to
/ T docusaent the snubber failures. This report also fu correctly determined the failure to oe p reportable. Therefore, it is concluded the li reportability and operability assessiaents were K made correctly.
To The failure was reported to the NRC via LER 71 82-165 for two of the five snubbers. However, as r1 a result of an administrative oversight, this LER 7 2, had not been revised to reflect all of the v[ failures. Revision 2 to LER 82-165, reflecting y the additional damaged snuobers and conclusions n
with respect to the cause of damage has been subraitted. Since SCE's April 1983 evaluation of the reactor trip breaker failures at Units 2 and 3 (Reference 9) resulted in iraprovements in the identification and impleinentation of the reportability requirements of the Technical Specifications, a repeat of this type of adrainistrative oversight is not likely to recur
, and no further corrective measures are planned.
N.
Item 2: " Inspection of these snubbers carried out over those periods
[between when the daraage probably occurred in March 1981, and when it was discovere'd in Decembar 1982] failed to detect their
' inoperability." [The discussion in brackets has been added for cicrity.]
Reply: Detween March 1981 and December 1982, a auuder of snubber inspections of Feedwater Line FW-109 were conducted and are listed in Taule 2. Since it is uost likely that the feedwater snubbers were damaged in ;
March 1931. Table 2 indicates that the visual and I in-place rotation checks identified only 2 of the seven ;
i failed snuboers out of a total of eight snubbers on line FW189.
However, the theriaal expansion test program and uanual stroking did identify the rest of the failed snuboers.
One snubber in support 52-F.l-189-11010 was originally
[
Table 2 i Inspections of Snubbers on Main Feed Line FW-189 i
Date . Reason for Inspection Type Findings March - April 1981 Following Special Visual No indications of damage to S/U Test of S/G pipind or sdpports Feedwater integrity July 1931 Walkdown following Visual No indications of damage to discovery that piping or supports March 1931 Feedring integrity test had collapsed the feedring
! December 1981 - January 1982 QA/QC Walkdown Visual, o UCR F-462, snubber at S2-FW-189 i!-012 to verify snubber Hands On (1) could not be rotated, snuober #390 integrity prior was replaced with J1145 to initial fuel load o NCR F-453, snubber at S2-FW-ld9-H-01J could not be rotated, snubber #2609 was replaced with 14322 o No other indications of snubuer I damage were noted November 1982 Walkdown following Visual No indications of dauage excessive feedwater to piping or supports.
addition to steau generator resulting
! in overcooling of the i RCS. (SIR S02-82-260 Reference 12)
NOTES: ,
- .(1) Ha.nds On means an in-place rotation of mechanical snuubers aoout their axis looking for hesitation or bindinj.
4 -. A -
F La. .,,A L-1 2 _a--
Table 2 Inspections of Snubbers on Main Feed Line FW-189 Date Reason for Inspection Type Findings -
- December 1982 50% Power, Thermal o Visual o NCR 2-032 S2-FW-189-H013 Top-extension l Expansion Test o Manual arm buckled Stroking 52-FW-189-HJ13 Bottom-loose ,
i extension nut '
1 o NCR 2-053 S2-FW-189-H010 both-jamued
! o NCR 2-069 S2-FW-189-H017 Doth-jammed All of the above snuboers, exept serial number 4322 on sup were replaced (2) port S2-FW-ld9-11010 NOTES:
! 2) Serial number 4322 was manually stroked with satisfactory results and reinstalled.
.i I
- t i
) i i +
i 4
l
?
_ Item 2: (Continued) thought to have failed. However, after successful manual stroking it was reinstalled. In his review of the December 1982 failures, Mr. Mertens requested that this snubber be removed and sent to the manufacturer for a functional test under load and an inspection of the internals. This snubber was found to have no internaldamageandiobeoperational,aconfirmation ofthemanualstroketest1)Ithefield. Note, that this snubber was one of two previously replaced in
. January 1982 and hence should not have damaged at the time of the functional load testing. Thus, with the appropriate inspection technijue, snubber inoperability has been detected in the field without the use of functional load testing.
Since the manufacturer's test program conducted by Mr.
Mertens has shown that all failure modes cannot be detected by manual stroking, additional test measures will be required. Further discussion of this is provided in response to item 4.
F l
l L
'Ltem 3d ,
"An attempt was made to obstruct the engineering evaluation I conducted."
Reply: This appears to relate to difficulty experienced in March, 1983, in removing the snubber (S/N 4322) installed in January, 1982. As indicated, Mr. Mertens' supervisor supported this effort, which was successful when the appropriate docuiaentation (a ilCR) was prepared. As discussed above, this snuboer was found to be undaraaged and operaole. '-
5 e
S s
{
l
. 'Iteu 4f , The method presently used for inspecting snuubers is inadequate and misleading."
Reply: The inspection pro 3rau for mechanical snubbers is defined in the S023 Technical Specifications Section 4.7.6 and the ASME Jailer and Pressure Vessel Code,Section XI. The Technical Specifications require periodic visual and functional inspections. The visual inspection acceptance criteria verify that:
- 1) there.are no visiiile indications of damage or impaired operability, and;
- 2) attachments to the foundation or supporting structure are secure.
The functional test acceptance criteria verify that:
- 1) activation (restraining action) is achieved within the specified ran se in ooth tension and compression, except tnat inertia dependent, acceleration limiting uechanical snuboers uay be tested to verify only that activation takes place in both directions of travel.
- 2) Snubber bleed, or release rate where
! required, is present in both tension and compression, within the specified range.
i 1
l 1
l Ltem 4f- (Continued)
- 3) llhere required, the force required to initiate or raaintain motion of the snubber is riithin the specified range in both directions of travel.
- 4) For snubbers s,secifically required not to displaceyndercontinuousload,theability
. of the snubber to withstand load tithout N '-
displaceraent.
- 5) Fasteners for attachiaent of the snubber to the component and to the snubber anchorage are secure.
Additionally, tne Technical Specifications require that a special inspection shall be perforiaed each refueling oittage of all snubbers attached to sections of safety related piping systems that have experienced unexpected, potentially damaging transients as deterrained from a review of operational data and visual inspection of the systeras. This inspection shall satisfy the requireuents of the visual acceptance criteria and confirm freedou of taction of the snubber.
e C'
r 1 I. tem 4: (Continued)
An additional inspection uethod used for uechanical type snubbers in the field is to stroke the snubbers through their range of travel, looking for any binding of the unit. This method is consistent with the recoi.vaendations of the snuocer uanufacturer and iaeets the requirements of the special Technical Specification inspection aantioned aoove.
Inservice inspection requirements for snuboers at Units 2 and 3 are contained in Section XI, ASME B&PVC, 1977 Edition and addenda through Suaaer 1979. Visual !
and functional testing acceptance criteria contained therein are consistent with those specified in the )
Technical Specifications.
f As was noted earlier in responding to Item I, the visual raathod will not disclose all types of snubber failure modes. It is for this reason that iaanual stroking was performed in December 1982. During that inspection, the manual stroking uethod was an ,
effective, discriminating uethod for determining snubber operability. Since the testing performed under Mr. tiertens' cognizance at Pacific Scientific demonstrated that visual inspection and manual stroking cannot detect all failure modes, the choice of inspection method should consider functional. load
[
)
7
', I.teu 4: (Continued) testing. An onsite bench testing uachine has been on order for some time and delivery is currently forecast for October 1983. Should testing be required oefore
. October, portable onsite or offsite testing services are available in a reasonable time period (approximately two days).
Thus,thereareseveralleveofmechanicalsnubber inspection available. Should operating conditions develop which are known to produce significant dynamic transients, a Station Incident Report would be prepared and an engineering evaluation performed. The scope of this evaluation would include testing requirements for uechanical snubbers and selection of the most appropriate inspection technique. l' r i f I I i N f l Item Sr "No effective QA program exists for uonitoring snubber perforaance as per Technical Specification page B 3/4 7-6, Snubber Basis." Reply: Page B 3/4 7-G of the Technical Specifications discusses the requirements for monitoring the service life of snubbers subject to environaental degradation (i.e., hydraulic snubbers). Presently, an effective program, implemented by ; procedures, is in place to mo'nitor both hydraulic and i mechanical snubbers. This program consists of both a visual and functional test to ensure proper snubber performance. A computer-based system which records the maintenance history of all snubbers is utilized to input information into the snubber surveillance program. Any failures that are identified during these inspections are suojected to a detailed engineering analysis and. if determined necessary, an increase in the frequency of inspection or the number of snuobers inspected is iuplemented. 1 ! I
. Item 6:' "A recommendation to install rainimum instruiaentation for iaonitoring operability of snubbers in safety-related piping was dismissed."
Reply: Mr. Mertens participated in a iaeeting oetween representatives of NUS Corporation, the Electric Poder Research Institute (EPRI) and SCE in early June 1983, which included discussion of this recommendation. Purs ant to that meeting, SCE expects a proposal from NUS in early August, 1983, which will include the posssibility of use of raonitoring instrumentation. t
. i o
t i I i i I l. i 1 i I i f.
'I. tem 7i . "The engineering evaluation performed by Bechtel Poser Corporation to determine the cause of failure is inadequate and misleading."
Reply: The primary purpose of the engineering evaluation performed by Bechtel was to determine if allo able piping stress limits had been exceeded and the extent of inspection and repair required to return the line to service. The analysis which is sumarized later in this section, clearly indicates acceptable stress levels and N provides appropriate directionfor inspection. Since the engineering failure analysis showed that stress levels were acceptable, the replaceuent of all
' i failed snubbers on the main feedwater line FW-189 and a review of the subsequent thermal expansion data .
i indicated that the line was moving as predicted by original design calculations. After replacement, , I snubbers at support 52-FW-189-li-013 were ueasured to 1 I extend 1.90 inches in the north direction at 50% [ power. This compares almost exactly with analytically , i calculated uovement of 1.85 inches in the north j direction. Il f f SCE requested BPC on April 7, 1983, to conduct a f further evaluation as to the specific causes of the j snubber failure and to determine any long term i corrective action. This evaluation is scheduled to be couplete by Aujust 19, 1933. l
' Item 8i . " Snubbers on the iaain fecdwater lines inside the Containraent were not designed for dynamic loads, they are undersized and .aay not be dble to ensure structural integrity of the uain feedwater lines."
Reply: Main feedwater lines inside the Containuent, including their supports, are designed in accordance with the ASME Boiler and Pressure Vessel Code, Section III, Class 2. The design has been verified as correct. As indicated in the design bas,is (FSAR paragraph x 10.4.7.1.d),' it does not include dynamic loading resulting from severe waterharaner transients, as these are to be avoided by operational and design ueasures. The startup test on March 21, 1981, which is hypothesized to have caused the damage to the five j snubbers which was revealed by tha inspections of ! Decesaber,19J2, and to the two snubbers which were i replaced in January, 1982, was substantially more severe than is included within the design basis. Further, following steam generator feedring modifications in mid-1931, a revised startup test was successfully conducted in June 1982 which confirued the adequacy of the main feed system design basis. This inforination was provided to the NRC at their request in Responses to Questions 010.18 and 010.54 (Attachments 5 and 6). Further evaluation of the potential for damaging waterharsaers in the feedwater piping was l
, Item 81 (Continued)
- performed in 1932 in response to requests from Station and Startup Engineering. Attachinents 7, 8, 9 and 10 provide copies of these requests and the evaluations performed. In suanary, the snubber dainage probably occurred during startup testing, which included a severe waterhaiaiaer transient. Extensive modifications, testing and evaluations have been perforined to reduce the potential
- for dauaging waterhanaer events and to show that the feedwater piping will not be overstressed. Should such a transient beyond tne design basis recur during plant operation, careful inspection and testing of affected snubbers would be performed in accordance with existing Station procedures including consideration of this startup testing experience. As discussed in Section IV of this report, criteria for determining when such a transient uight have occurred will be developed.
G
^ J
l . . Item 9: , "An attempt was made to intilaidate me [lir. Ilertens] (transferred or fired) in response to my questioning BPC's performance at S0t1GS 2 and 3." [The comments within brackets have been added for clarification.] Reply: lio atterapt was made to intimidate !!r. liertens. In an exchange of correspondence between 14r. ilertens, SCE and others commencing in ilovember, 1982, ne requested: o Total expenditure dat concerning S0ilGS 2 and 3 o Payments to Bechtel for' engineering work on S0!iGS 2 and 3
- o A copy of the contract between SCE and Bechtel for SONGS 2 and 3 o The same information as above for S0tlGS 1 Mr. tiertens pursued these documents and data as a stockholder with the Secretary's office, rather than as an employee. In his letters, he included the couraent that the reason for his request ". . . is to deter $ine whether or not the [ contract] contains provisions for redress in the event of unsatisfactory perforiaance, and to study the terms of such provisions." At no time prior to receipt of his letter to Mr. Godld dated July 20, 1983, were questions of nuclear safety raised in this correspondence.
6
y
- I. tem 9: (Continued)
. Mr. Mertens was counseled in February, 1983, to stop his repeated attempts to obtain the information above, as SCE has declined to provide it. This counseling did not adversely affect his overall perfonaance rating in his March, 1983, performance appraisal, and he was not' ;
threatened with transfer or dismissal. x N .. f l 9 I 1 4 1 I. tem 10: . "The original and all copies of NCR S01-P-1303, Rev. 1, were destroyed." Reply: Revision 0 of this NCR for SONGS 1 was opened on Deceiaber d,1982. The condition described relates to corrosion of a heat exchanger foundation and is also described in an NRC inspection report transiaitted to SCE by HRC letter dated January 7, 1983. The work required to repair the foundation is ongoing and is about 80% complete. Revision'l of the NCR was drafted to readdress the original evaluation of heat exchanger operability which had been saade for devision U. Sotae ; difficulty was experienced in deciding how to treat f operability during the repair. It was concluded in May 1983, prior to validating the NCR revision that readdressing operability was not necessary, and the repair could be raade with the heat exchanger in service, so the revision was not completed. Revision 0 of the NCR remains in effect, and it governs the work which is now nearing coiapletion. 4 j i
l
; 'In addition to the ten concerns identified above, additional iteus ware presented in Mr. Mertens' memoranduia to Mr. B. Katz dated May 23, 1983, (Attachment 1). Each is listed verbatim below along with a reply.
Item 1: "The pressure differential of 500 to 800 psi (see Attachment [Aj, Itera 6, Page 2) responsiole for feedring i collapsing the steam generator 4 nduced a pressure wave back through the water-filled raain feedwater line producing, at the instant of reflection from the 20" check valve flapper, the dynamic force that damaged the snubbers at locations H010, H013 and H017." [See Attachraent 1, Paragraph 8, Conclusion 2.] : b Itera 2i_ "G-Forces produced by toe respective transient uay have exceeded design limits. (An inspection of main and auxiliary j feedwater lines for signs of overstressing is advised)." [See Attachment 1, Paragraph B, Conclusions 2 and 4.] Reply to The postulated differential pressure and resulting G-forces j Items 1 & 2: produced by the transient did not result in piping stresses that exceeded design limits. This conclusion was reached through inspection and analysis performed immediately following the discovery of damaged snubbers. 1 and 2: Reply (Continued) Inspection s
~
Visual inspection of the snubbers and associated snuober support structures identified only snubber coraponent daiaage,
'and verified that no daiaage w6s evident on associated support structures, i ,
4 i y Analysis -
's ,
x Four analyses were perforraed to evaluate stress levels in the s . riging. A. Lirait Load Analysis c A limit load is the calculated load at which the Weakest part of tha snuober support structure would fail.This failure woul' be d evidenced by physical deforination of structural iaembers or cracked welds. A i , lisait load o,f approxiiaately 56,000 los, was calculated
. . , - 2
,, at support H015. This liiait load was then applied to
, the feedwater piping systera to detertaine resulting stress in the piping.
s l % i t ,
^
1 and 2 , Reply , (Continued)
~ . B. Check Valve Slara Forces up to 50,000 los. were calculated to occur due to a pressure transient. These forces were then applied along the axis of the pipe to determine resulting stress' levels.
C. Steam Generator Nozzle Move e'nt The steam generator nozzle was analytically moved to simulate tneraal growth of the llSSS and steara t generator. At the same time snubbers at support , e location ;1013 were modeled in a locked condition in . t order to simulate r sistance to feedwater pipe moveuent. i D. Cold Pull i I i The piping was analytically " cold pulled" to simulate ; the effects of thermal growth against locked snubbers. ; This additional analysis was performed after the ! remaining snubbers at support locations H010 and H017 , were found to be damaged late in December 1932. The calculated results of these analyses indicated low stress levels in the piping:
.1 and 2: -Reply ,'(Continued)
- l. Local stresses at support attachment locations were a;) proximately 15,000 psi compared to an allowable field strebs of 32,000 psi at 400*F.
- 2. Piping component primary stresses were approximately
~
18,000 psi. This is also below allowaole stress of 21,000 psi.
'N
- 3. Piping component secondary plus primary stresses were maximum of 34,000 psi due to consideration of steau l generator nozzle thermal growth and piping thermal growth against locked snubbers at H010, H013 and H017.
This compares with an allowable secondary stress of l 43,750 psi. I 4 I l i. i l
~
l f l l I 1' ( -
, , . It.ea 3 : , ,"The engineering evaluation of the damaged snubbers on FW-189 Unit 2, performed by BPC (Attachment A, Iteis 4) raade ]
. \
no attempt to examine the snubbers internally to deterraine the true magnitude of the shockforce; the recommen'dation for i larger snubbers should not have been cancelled without further study, and it should not have advised SCE in a follow-up letter (Attachment A, Item 5) that no further investigation of the snubber failure was recommended." [See Attachment 1, Paragraph C, Finding 5.] Reply: An inspection of the snubbers on feedwater line FW-189 was directed by Bechtel Engineering in December 1982, as a result of monitoring piping movements as required by Test Procedure ! 2PA-102-01. The subsequent inspections at supprt locations H010, H012, H013, H014, and H017 resulted in the replacement of five damaged snubbers. { The engineering evaluation recomuended that the damaged t I snubbers be inspected and replaced if necessary. SCE g Engineering raade arrangements to send these snubbers to { t Pacific Scientific for failure analysis as part of the f continuing evaluation program. , t i l l l . . . . -
. - Itera 3 : , (Continued) . The recoramendation that no further evaluation of the snubber failure was required was based upon the determination that acceptable piping stress limits were not exceeded and future events that could cause similar snubber damage would require a detailed inspection per the plant Technical Specifications. Further evaluation as to the mechanism that imposed abnorual loads on the snubbers is scheduled to be couplete by August 19, 1933. The mignitude of shock force was deterrained as discussed in Iteu 4 below.
e e L i l l l l i l l- . . . , .
n
. . Item 4: .
" Magnitude of Dynamic Forces "
. The magnitude of the shock forces that acted on the snubbers can be estimated from the severity of the damage their internal parts suffered.
An examination of the five damaged snubbers at the manufacturer's test fac'ility revealed that all failed under
. compression. This fact tells us the direction of the force
~s s
and its point of origin. With the did of Figure (a) on Drawing No. 100, this is explained as follows: Figure (a) shows vector diagrams corresponding to compressive forces It can be seen that the force F1 I acting on the snubbers. must have been the result of a check valve slaia and that component forces F2, F3 and F4 caused the snubbers to fail. l t i From the extent to which the snuboer internals were damaged, one can establish the approximate value of the destructive force. On all of the five snuboers the threads oetween the ball screw shaft and the inner thrust bearing race (Figure 1) were sheared, and further, all shaf t ends had hit bottoia and suffered plastic deformation. Shearing the threads requires a static force of approximately 30,000 lbs. as confirmed by a test conducted at the manufacturer's test lab (see test report Attachment A, Item 2). Item 4: ,(Continued) I Exataination of Figure 1 reveals that, after having sheared the threads, the shaft was free to travel another 3/16" before hitting bottoia and becoraing deforraed. Therefore, shear and deformation forces are additive. The transient force causing the daformation is estiraated as follows: The shaft laaterial is ASTM A322, having a yield stress of 75,000 psi. The shaft end area is '0'2 inches squared. From these two quantities, we obtain 15,000 los. as the force of deforiaation. Adding the two forces fields 45,000 los. as the approxiraate minimura force acting on one snuocer. Since the , snubbers are arranded in pairs, we obtain, for the staallest coraponent force, F2 = 90,000 lbs. (Figure (a). For establishing a minirauia value for the resultant F1, the following reasoning was applied: Vector diagraias in Figures (a) and (b) on Drawing No.100 show coraponent forces F2 and j I F4 to be the smaller ones. If 90,000 lbs. is assigned to j both F2 and F4, ',he priraary force F1, acting on the check f valve plate, becomes F1 = 170,000 lbs. and coraponent [ i' force F3 = 140,000 lbs. These are rainitaum forces because l the extent of daraage inflicted upon the feedwater pipe snubbers, when corapared to the degree of damage seen in the snubber subjected to the dynamic destruct test, leaves no doubt that greater forces were at play." [See Attach;aent 1, ! Paragraph E, f.nalysis 1.] l l l
. 3,-
Reply: The damaged snubbers at support locations H010, H013 and H017 arc PSA-10's which are capable of reacting approxiuately 32,000 lbs. before failure. These loads were developed by Pacific Scientific thru testing. A review of the structural members that transmit the snubber loads into the building structure reveals that the highest loads that could have occurred at support locations H0ld, H013 and H017 are approximately 65,000, 56,000, and 62,000 lbs., respectively. Since there was no visible physical evidence of damage to these pipe support structures, thes'e'-loads are considered to be upper bound. Application of these loads to the piping system has been performed to confirm the acceptability of stress levels on the piping as detailed in the reply to Items 1 and 2. Detailed inspections of the piping system and associated pipe supports including snubbers indicated no failures of the structural steel ueabers or connecting welds. Therefore, i lo. ids on the order of 90,000 lbs. and 140,000 lbs. could not have developed since structural steel aemoer failures would have occurred at significantly lower levels (65,000, 56,000 and 62,000 los.). e f
~ *' ~
- Shock Arrestor Size
_ Item 5_: . It has been shown that each snuboer was exposed to a shock force of at least 45,000 lbs. They are designed for 15,000 lbs. Since conditions exist that could cause another waterhamaer in FW-189, it is advisable to increase the size of snubbers on the main feedwater lines of Units 2 and 3. This may require redes'i'gn of the components to which the snubbars are attached." [See Attachment 1, Paragraph E, Analysis 2.] - Reply: Our evaluation'does not concur tnat shear and deformation . forces are additive and that each snubber was exposed to a
~
shock force of at least 45,000 11's. We oelieve that a load of approxiiaately 32,000 lbs. to 31,000 lbs. could have been exerted on each snubber. The destructive tests conducted by Pacific Scientific on a ] number 10 snubber indicated a failure load of 31,600 lbs. and , i dynamic loads up to 37,000 lbs. This load resulted in I sheared threads between the ball screw shaft and the inner I thrust bearing race. Subsequent to shearing of the threads the shaft must travel 3/16" inches before coming into contact with the snubber housing. Therefore, the loads generated to s cause shearing of the threads did not occur at the same time as the load causing shaft end defonnation. Therefore, the shear and deforuation loads are not additive. The uaxiuuta load therefore, is the larger of the two loads, i.e. 31,600 lbs. to 37,000 los.
~
l -. . - - , . , .- . - _ . _ _ - _ . -
'. 5. Reply (Continued)
+ The design of the backup structure for the paired snubber arrangement is such that it will sustain visible damage for loads of approximately 60,000 lb:. Since there was no visible physical damage on the backup structure, it is not credible to postulate a failure load of 90,000 lbs. (a combined load of 45,000'lbs. on each of two snubbers).
', N Iteni 6: " Inconsistency between Existing Snubber Design Loading and Technical Specification.
An inconsistency exists between the intent of tne T/S and the , design basis for snubbers on the safety-related main feedwater system, as the following explains. i T/S 3/4.7.6 Snubbers, Plant System Bases reads as follows: I PLANT SYSTEMS l BASES j i 3/4.7.6 SNUBBERS r
}
}; . .
Reply: There is no inconsisic,,;y between the design basis for the feedwater piping system and Technical Specification 3/4.7.6. FSAR Paragraph 10.4.7.1.D states " Design considerations of the feedwater piping and steara generator feedring preclude the occurrence of hyraulic instabilities." Accordingly, the main feedwater line snubbers were not designid for abnormal hydraulic dynamic loads, but were designed 'or Seismic Category I loads. Unusual vents such as the feedring integrity test of March 1931 or.the cooldown of November 1982 were analyzed in detail. Procedures and design changes have minimized the possibility of recurrence. Additionally, analysis showed that these events did not result in excessive pipe stress. Other scenarios will be evaluated as part of the continuing snubber failure analysis which is scheduled to be complete by , August 19, 1983. h I , 4 i 9 4 4
. . ' . Item 6: ,
(Continued) All snubbers are required OPERABLE to ensure that the structural integrity of the Reactor Coolant System and all other safety-related systems is maintained during and following a seismic or_ other event initiating dynalaic loads. Snuobers excluded from this inspection prograra are those installed on nonsafety-related s
. systems and then only if their f ailure, or failure of N
the system on which they are iiistalled, would have no adverse effect on any safety-related system. f The main feedwater line snubbers, however, were never i
. 1 designated for dynamic loads (see Attachment A, Item 10, DYN. LDS column). j i
i Since conditions that can cause a waterhanaer in the main feedwater line still exist, (1) the event can repeat itself, causing the snubbers to fail again. Therefore, the snubbers, as they are, cannot ensure the f i structural integrity of the systeia during and following a dynaraic load event. This requires resolutions in regard to . operability determinations for the feedwater system." [ [See Attachment 1, paragraph E, Analysis 4.] I i i i l
Iteu _7:
,"QA as related to snubbers is inadequate" (see Attachiaent 1, Proposed letter to Mr. J. B. Martin, page 4).
Reply: In the July 20, 1983 proposed letter to Martin, Attachment A, Item 7 is a copy of SCE Nonconformance Report (NCR) S023 F-463, dated 1/25/82, and identifies that snuober #2609 cannot be rotated about its axis and that cold setting inforraation on ID tags of both snubbers #2609 and #2603 do not match the design drawings.
's In the July 20, 1983 proposed letter to Martin, Attachment A, Item 8 is a copy of work order 2113 which identifies a snubber as being incorrectly tagged and indicates that a new tag was affixed.
The inference of inadequate Quality Assurance, regarding Attachment A, Items 7 and 8, appears to be based upon a comparison of the identification of an inoperable snubber (#2609), including the disposition requiring its rejection, quarantine and subsequent replacement, with work order 2113, requiring only replacement of the ID tag. This apparent inadequacy is further alluded to on page 4, first paragraph, of the letter proposed to be submitted to Mr. J. B. Martin, HRC Regional Administrator, Region V. b
o
' Itta 7: (Continued)
~
': s .i
~ <
It must be noted that verification of the removal, quarantine and replacement of Snubber #2609 is evidenced in work order 2114, not 2113 as specified. This work order (2114) is additionally referenced on NCR S023 F-463 in Block ~ 24 (Reference 8). Based upon the proper research of tiCR S023 F-463 and verification documents utilizing both work orders, there is no validity to tne comment inferri.ng inadequate Quality t Assurance. 6 6 i 1 i i i I 3 l
r .
}
- l. ,
IV. PRESENTATION OF CONCERNS TO ONSITE REVIEW COMMITTEE In ac:ordance with the SCE procedure for the review of nuclear safety concerns (see Attachment 2), a special Units 1, 2 and 3 Onsite Review Cor.wittee meeting was held on July 28, 1983. Mr. Mertens presented four concerns related to nuclear safety. These concerns (taken from the minutes, Reference 1) are listed below: N s (1) Visual inspection methods for mechanical snubbers do not tell us anything about operability of the snubbers. (2) There are no means of detecting transients that ct 'J damage snubbers. , i 1 (3) There are no means of detecting damage to snubbers except for gross I l' deformations. (4) Piping (specifically, FW189) may be overstressed as a result of the damaging transient. (Concern with respect to power operation in-the next week.) i Item 4 was declared a restraint to Mode 2 entry for Unit 2 until a review of the feedwater line FW189 strest analysis was completed. The results of that analysis are contained in Section III of this report. The analysis concludes ! I the line was not overstressed. l 1 l (
' . A second special Unsite Revied Corcnittee meeting was held August 4,1983 (Reference 2) to complete the review of items 1, 2 and 3. The Comraittee
' concluded that potentially damaging dynamic transients can be detected by observation of operational parameters. Additionally, the Coraaittee concluded that visual inspection is not capable of detecting all modes of mechanical snubber failure. Thus, the Coraaittee directed that criteria shall be established, in adiainistrative procedures, for performing functional mechanical tests of snubbers. This criteria will be developed based on the conclusion that transients which will induce mechanical snubber failure will be events that are known and that current procedures' lack a definition for events that should trigger functional testing of mechanical snuubers in selected areas based on the possibility that a potentially damaging waterhammer could have occurred.
V.
SUMMARY
OF ACTIONS _ NOT COMPLETE AT TIME OF THIS REPORT
- l. From Section III, further evaluation as to the mechanism that imposed ]
abnormal loads on the snubbers is scheduled to be complete by August 19, 1983. x
- 2. from Section III, other scenarios will be evaluated as part of the
.y continuing snubber failure analysis which is scheduled to be complete by August 19, 1983.
- 3. From Section III, further evaluation as to the specific causes of the i s6ubber failure and to determine any long-term corrective action.
- 4. From Section IV, the Onsite Review Committee has directed that criteria i for determining when potentially daaaging waterhammers might have i
occurred will be developed by August 30, 1933. i i I 1 i
7-ATTACHMEitT 1 t l O
\
i N N, 9 l e l 9 I.
- i i
4 l I I I I I I i l, a 4 i l l r
~_ 0 l- .
Mr. W. R. Gould Chairman of the Board , pg'CEIVED Southern California Edison Company P.O . ' Box 400, 2244 Walnut Grove Avenue Rosemead, California 91770 M22E
.WM. R. GOULD July 20, 1983 -
Dear Mr. Gould:
I am writing to you because I must assume that you are not i-being fully informed of certain matters concerning nuclear op'erations. _ I am a station engineer at Songs and I have enclosed for your review two letters that I have written. one to Mr. J. B. Martin, HRC,.and the other to Mr. L. Grimes, Chairman, CPUC. I believe Company policy reqires employees to inform management of such correspondence and submit the same for review. The explanation for writing them can be found in their text. ; It is my intention to mail the letters on July 27, 1983 I would appreciate receiving your reply as to whether or not they contain false statements so that I may enclose it. ), l Sincerely J hn Mert e
/
Enclosures . , _ - g #,,1:~' I l 3109 S. El Camino Real San Clemente, CA 92672 (714) 498-6311
---- - - - - . - .}}