ML20209G881
| ML20209G881 | |
| Person / Time | |
|---|---|
| Issue date: | 04/30/1987 |
| From: | Ellen Brown NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20209F141 | List: |
| References | |
| NUDOCS 8705010089 | |
| Download: ML20209G881 (75) | |
Text
I i
'r '
i 4
FESOLUTION OF PEER REVIEW COMMENTS i REGARDING THE APRIL 1986 AE00 PRELIMINARY i CASE STUDY REPORT "A REVIEW OF MOTOR OPERATED VALVE PERFORMANCES" APRIL 1987 Prepared by:
Earl J. Brown Reactor Operations Analysis Branch Office for Analysis and Evaluation of Operational Data U.S. Nuclear Pegulatory Comission 0705010089 870408 PDR ORG NEXD PDR
TABLE OF CONTENTS fage Coments from INP0 and AE00 Response . . . . . . . . . . . . . . . 1 Comments from IE and AE00 Response . . . . . . . . . . . . . . . . 5 Comments from RES and AE00 Response ............... Il Comments from M0 VATS, INC. and AE00 Response . . . . . . . . . . . 21 Coments from NRR and AE0D Response ............... 45 Coments from Region I and AE0D Response . . . . . . . . . . . . . 70 Comments from Region II and AE00 Response ............ 71 Coments from Region III and AE00 Response . . . . . . . . . . . . 72 Coments from Toledo Edison and AE00 Response .......... 74 l
l l
l
C0pmENTS FROM INPO AND AE00 RESPONSE Letter from P. W. Lyon to C. J. Heltemes, Jr., dated May 20, 1986 i
coment 1 Recently a utility modified valves in the shutdown cooling system in response to IE Bulletin 85-03, and the modification contributed to an overcooling event. INPO issued SER 12-86 and subsequently the NRC issued IE Information Notice 86-29 describing this event. The modification changed the torque switch bypass setpoint for the opening cycle to approximately 16% open. A motor operator with a two-rotor limit switch assembly uses one rotor for both torque switch bypass on the opening cycle and closed indication. As a result, when the valve indicated " closed", it was actually partially open. This event demonstrates the importance of thoroughly understanding motor operator design prior to implementing modifications. It also illustrates the weakness of two-rotor limit switch assemblies particularly for valves that do not " seal-in" when operated. The report should address this important concern.
Response to Comment 1 This issue was added to the report. The event is discussed in Section 2.5 together with a suggestion for both caution and the need for a thorough investigation before implementing a simple change in the setpoint of the y bypass limit switch.
l
2-Comment 2 We support the expedited reevaluation of the current Regulatory Guide 1.106 recomendation to bypass thermal overload protective devices. A similar problem may be developing as some utilities either have or are considering bypassing the torque switch contact for the entire opening valve stroke.
While it is certainly important for a valve to operate in an emergency, it is much easier to recover form a tripped themal overload or torque switch than from a broken valve stem or burned-out motor.
Response to Coment 2 We agree with this coment and continue to recomend appropriate action.
___ _ _ _ x
. o Coment 3 We believe finding number 9 on page 61 is not supported by the data and we recomend a reanalysis of the NPRDS failure reports described in Section 2.4.
The attachment t.o this letter documents sone of our concerns with the results of the NPRDS failure report analysis. Since NPRDS is the dominant source of l POV failure information from 1984 and 1985, some of the report conclusions may l change subject to this reanalysis, l
Attachment:
Coments on Section 2.4 i Finding number 9 on page 61 is based upon the conclusions in Section 2.4 that approximately 60% of the NPRDS failure reports merely stated that the valve failed to open or close and did not attributed a cause to these failures to I operate as needed.
INPO staff have independently reviewed the data sample used in the report
! (633 MOV failures discovered in 1984 and 1985 and reported as of 9/20/85, the date of the NRC data retrieval). From our brief review to date, we have two observations that contradict the sumary presented in Table 2.4-1.
First, after a mt.nual review of the failure narrative fields in a randem
! sample (approximately one-fourth) of the reports, we feel that over 90% of the l reports do attribute the failure to a specific cause such as torque switch or limit switch problems.
Second, the table states that only 4.7% of the NPRDS reports identified a i rcuse related to torque switch / limit switch problems. However, a text search of the failure cause fields for the works TORQUE or LIMIT along with SWITCH produces 229 records,or36%(vs4.7%).
Resperse 'to Coment 3 The section on NPRDS data was completely reanalyzed. It was discovered that the initial data searches had inadvertenly omitted certain information such l
o . 4 Response to Comment 3 cont'd that the output did not provide the information that we had intended to obtain.
The new analysis resulted in new problem categories that suggests the NPRDS 4
data identified as "cause of failure" is actually a symptom of the failures such as an adjustment to a torque or limit switch. The analysis indicates nearly 80 percent of the identified causes of failure are most likely symptoms
- of inadequate procedures, inadequate maintenance or indicative of component i
degradation that should have been investigated and resolved prior to the failure. However, the information contained in NPRDS does not permit an indepth review or independent evaluation needed to determine the true root cause of failure.
i l
l 4
1 4
l I
COM ENTS FROM IE AND AE00 RESPONSE Merorandum through R. H. Vollmer to C. J. Heltemes, Jr., May 7,1986 l
j Comment 1 It is reconeended that:
- a. The paragraphs describing the AE00 case studies be expanded to includetheassociatedIEgenericcommunication(s),
I
- b. The paragraph containing the listing of associated IE generic communications be deleted, and
- c. The two subsequent paragraphs be rephrased to indicate that they are addressing IE generic connunications pertaining to subjects beyond those in AEOD case studies.
Response to Comment 1 The report was changed to incorporate these connents.
I
- 6-Connent 2 The first sentence of the second paragraph of Section 2.1 (page 17) refers to !
the recontendations in Section 1.2 of the subject case study. The proper :
reference should'be to Section 4.0 of the case study, j Response to Coment 2 The sentence with the reference to Section 2.1 was correct as written.
However, the sentence vas modified so the reader can recognize that Section 2.1 is the proper reference.
l a
s i
t 4
i l
l i !
1 1 1 i
____ i
l 7
Coment 3 1
The last sentence of the twelfth paragraph (third from the end) of Section 2.2 (page 32) states that the cause of the Davis-Besse event was the torque bypass switch being set. so that the torque switch was active during the hamerblow portion of the valve's close to open cycle. Had this been the case, then the l valve would not have operated whenever it was called upon to open. The fact 1
the valve operated properly when not subjected to differential pressure refutes this contention. Rather, the problem was caused by the lengthening of the portion of the valve stroke during which high torque was required because of the differential pressure increasing the valve disc to seat friction load.
While this distinction does not change the apparent cause of the valve failure (i.e., improper setting of the torque bypass switch), it is crucial in under-standing the root cause.
l Therefore, it is recommended that this paragraph be revised to reflect the true root cause of the Davis-Besse failure.
Response to Comment 3 The paragraph was modified to recognize the effect of differential pressure over a portion of the stroke.
l 1
, . Comment 4 The seventh sentence (second from the last) of the last paragraph of Section 2.3 (page 40) states that the cause of the recent Catawba valve failure was the same as the Davis-Besse event--improperly set torque bypass switch 4
settings. However, it is unclear how this can be the case when the valve stopped in mid-stroke, a position not normally covered by the torque bypass switch. Rather, the failure was caused by improperly set torque switches (they had been set at the low end of the recommended range).
Therefore, it is recommended that this paragraph be revised to reflect the proper root cause of the failure.
Response to Comment 4 This event was investigated further and the report was expanded to enhance understanding and illustrate the potential significant safety issues. The actual position of the valves when they failed was unknown even though it was reported to be at mid-stroke. Also, the valve operators did have a bypass on the torque switch and the differential pressure was instrumental in the failure.
Thus, the ir. properly set bypass was the more likely cause of failure rather than the suggested improperly set torque switch.
1 l
l
Comment 5 Pecommendation 3 (page 65) contains, in reality, at least three recommendations--namely: (1) developing a root cause diagnostic capability; (2) expanding the scope of IEB 85-03; and (3) developing a training program.
Therefore, it is recommended that this recommendation be broken into three separate recommendations.
Pesponse to Comment 5 The initial recommendation 3 was divided into three separate recommendations to accommodate the comment.
t 1
10 -
g Coment 6 ..
The last sentence of Recommendation 3 (page 65) it either a statement of fact or a conclusion, not a rettmmendation.
Therefore, it is recommended that either:
- a. It be moved to either the finding (Section 3.1) or conclusions (Section 3.*c, portion of the case study, or
.s ,
- b. It be expanded to present a true recommendation based on the ,
~
observatINnwhichismade.
Response to Recomrhndation 6 The sentence in quest.fon was included as part of a recommendation to cover a training program and similar words are included in the conclusions section (3.2).
% t y s
3 s
y
. COMMENTS FPCM PES AND AE00 RESPONSE Memorandum from Guy A. Arlotto to C. J. Heltemes, Jr., dated May 23, 1986.
Enclosure 1, Comment 1 Page 19, and 2.1 addresses the issue of the failure of a valve assembly to operate against a differential pressure. At times an assembly may be designed for operation against a specific differential pressure and tested to verify such capability. If the specific installation applies a greater differential pressure and the assembly fails to operate the failure mode would be valve failed to operate but the valve did not fail - the failure was in the design engineering process. Such a cause does not appear in the report.
Response to Enclosure 1, Comment 1 The issue raised about differential pressure is addressed by Finding 6 in Section 3.1.
Enclosure 1, Coment 2 ,
The coment raised a concern about the discussion on page 29 about spring pack gap and recomended additional details.
Response to Enclosure 1, Coment 2 This portion of the report was revised to discuss the physical changes that occur with the torque switch.
Enclosure 1, Coment 3 The final recomendation states that "the scope of the IE Bulletin 85-03 should be expanded to cover all safety-related motor-operated valves required to be tested for operational readiness in accordance with 10 CFR 50.55a(g)."
That section of title 10 involves various editions and addenda of the ASVE Boiler and Pressure Vessel Code in regard to valves. Since the earlier editions are less detailed, it would be helpful if the recomendation identified the earliest edition / addenda that provides assurance of valve assembly operability and addresses the issue of " unannounced" failures.
Response to Enclosure 1, Coment 3 The subject paragraph (g) addresses requirements to identify specific valves for an inservice inspection program. Each plant updates this program at prescribed intervals to accomodate ASME Code changes and we think it may be confusing to suggest other time frames or editions. A statement was added to the recomendation to identify the intent was to cover all those motor-operated valve assemblies addressed by the inservice test program.
i l _ _ - -
Enclosure 2, Coment 1 The term " valve assembly is defined in Section 1.1, but it appears that subsequent sections use different terms when " valve assembly" is appropriate.
Response to Enclosure 2, Coment 1 The document was reviewed and terms were changed to provide consistent usage in the report.
i
Enclosure 2, Comment 2 The terms operability and performance / reliability raise some questions concerning precise meaning.
i l
1 Response to Enclosure 2, Comment 2 The intent of the term operability was to cover concerns related to proper definition of accident conditions and resultant loads together with adequate demonstration (e.g., functional qualification testing and surveillance testing) that a specific valve assembly can operate under design basis conditions. The intent of the term performance / reliability was to cover concerns related to whether the valve assembly will operate when needed and involves evaluation of operating experience that may signify degradation or changes that could impact previously assumed loading conditions or equipment status.
Enclosure 2, Comment 3 The first paragraph in Section 1.2 addresses valve stem failure due to inter-granular stress gorrosion cracking. A review for adequacy "of ASME code requirements" concerning hardness of martensitic stainless is suggested.
Recommend the suggestion be revised since valve stems are not " code items" and, therefore, there are no code requirements; for example, specified mechanical properties, for valve stems. The requirements are left up to the valve manu-facturer.
Response to Enclosure 2, Comment 3 A request about valve stems was formally sent to RES on June 14, 1985 as a result of the AEOD study. That study concluded valve stems were not a " code item" but felt this was a deficiency in the Code because it had resulted in inoperable valves. We understand the Code has decided they will not take action on our request. However, their inaction does not mean the AE0D request was not a reasonable and prudent course of action. The statement documents the AE0D conclusion.
4
, , Enclosure 2, Comment 4 It has been a concern of some NRC staff that flow-induced vibrations may contribute more to valve assembly failure than what has been documented. It is clear that these kinds of excitations can occur particularly when the flow area through the' valve is reduced by a valve disc, etc. Although the report identifies the effects of these kinds of loads (p 19 (second bullet)) and p 58 (fifth bullet)) as " damage due to misues of a valve...." it is conceivable that the damage occurred during normal operation of closing and opening the valve.
Therefore, the word " misuse" may not be an accurate description of the cause for failure.
Response to Enclosure 2, Comment 4 There may be instances where vibration damage could result in a manner such that " misuse" may not be an accurate description. However, the report AE00/E315 clearly involved misuse of the valve that involved improper throttling of the valve with resultant vibration. The statement is, therefore, appropriate.
i i
I
- 18 _
Enclosure 2, Comment 5 In the evaluation of NPRDS event data, the difference between the " root cause of failure," for. example, fuse not in place, and what IEEE* calls the " failure mode," for example motor fails to run is discussed. Reconmend AE00 consider adopting the IEEE usage of failure mode.
Response to Enclosure 2, Conment 5 Although the IEEE standard represents an attempt to establish a consistent apprcach to reliability data, the terms used in this report follow those used by NPRDS in order to provide easy access to information and comparisons by people who wish to do similar studies, i
4
- IEEE 500-1984 failure rode is the effect by which a failure is observed to occur, ... for example, for a relay which fails to close on command, the failure mode is..." fails to close."
l . - --. _
l
o
- 19 Enclosure 2, Comment 6 On page 59, it does not appear that the phase " failure occurs immediately (underscore added) after" is appropriate.
Response to Enclosure 2, Comment 6 The word Immediately was replaced by "during or soon" to be consistent with other portions of the report. This change was made in other sections, but was overlooked on this page prior to release of the preliminary report.
d
- t. _. - - - _ - _ - . - . - _ .-- .-. . . .. - -
a
- Enclosure 2, Comment 7 In light of Comment 4, a question is raised about the possible interaction of the failure mechanisms described on p. 58. For example, is the evidence presented in the databases sufficient to clearly blame a failure on inaccurate limit switch settings? It is conceivable that excessive vibrations at the valve can propagate to the actuator and influence valve assembly behavior. Therefore, a failure that may appear to be due to limit settings may have been caused by vibrations.
Response to Enclosure 2, Comment 8 In general, the difference between an inaccurate limit switcli sattino and vibration influenced change on the valve assembly behavior is that the limit switch can or would be loosened by vibration. This could be detected and would be different from an inaccurate setting.
. COMMENTS FROM M0 VATS, IFC AND AE0D RESPONSE Letter from A. G. Charbonneau to C. J. Heltemes, Jr., dated April 30, 1986 Comment on Page 2, paragraph 1 ,
Throughout the draft report, motor burnout is addressed as a typical failure mode for motor-operated valves. We wish to point out that, in our opinion, motor burnout is an indirect failure resulting from a previously existing degradation.
Response to Comment on Page 2, Paragraph 1 AE00 believes that motor burnout may be related to several possible deficiencies that result in degradation and eventual burnout. The ideal solution would be to correct the deficiencies; however, the safety concern is that current practice with inadequate protection and detection methods may lead to increased potential for motor burnout when the valve assembly is needed to perform a safety function.
l
. s Conment on Page 3, Paragraph 1 This paragraph includes a statement that reversing the direction of a valve while it is being operated could exceed design requirements which could result in valve failures. M0 VATS, Inc., conducted testing on this valve with zero differential pressure and the test results did not show unusual loading conditions.
Response to Comment on Page 3, Paragraph 1 The information reported to NRC by both the licensee and valve operator manufacturer indicates that reversal could result in overload and failure. No information has been presented by the licensee concerning any M0 VATS, Inc.,
testing. With the information we have, the statement appears prudent as a cautionary procedure.
-l a
, . Comment on Page 8, Paragraph 1 There are a number of organizations within the industry who have expressed concern with age related failures of motor-operated valves. As we have stated in the past and will continue to do so, we find it extremely difficult to justify any degradation that we have seen to date with perhaps one exception (grease hardening and/or separation) that is age related. It is true and we recognize that an improperly set up POV could be inducing a wear condition which, over a period of time, could result in inoperability. It is our opinion, however, that this type of degradation and resultant failure should not be classified as " age related."
Response to Comment on Page 8, Paragraph 1 Our concern is that if degradation occurs with time in service it is age related even if the ultimate cause is improper procedures, methods, or settings. k'e therefore propose to retain the term " age related" with the understanding that proper set up could reduce the number of such events.
~l l
l
_ 24 Comment on Page 9, Paragraph 9 We would recommend, based on our previous courents, that the word aging be deleted from this particular paragraph.
. Pesponse to Comment on Page 9, Paragraph 3 The response is the same as that to comment on Page 8, paragraph 1. Keep the word " aging."
i i
1 1
I i
- - - . . - - _ _ . - ~ - . . - .
I l
, e Comment on Page 13, Last Paragraph This paragraph should be revised after due consideration of our previous comment regarding the impact of a change in direction during operation of an MOV.
f Response to Comment on Page 13, last Paragraph We retain the wording based on the reason provided in the response to the comment on Page 3, Paragraph 1, i.e., it is consistent with the information provided by the licensee to NRC.
J 4
i
.i 1
4
. Comment on Page 14, Paragraph 1 During our investigation of the PORY block valve event discussed earlier, it is not clear to us that the toroue switch contacts closed due to a reduction in the system pressure drop across the valve. In all likelihood, it was due to a relaxation of the springpack which, in most cases, may be independent of any changing pressure drop condition across a given valve.
Response to Comment on Page 14, Paragraph 1 Information presented to NRC only mentions the torque switch trip and does not reference other investigations. Therefore, we propose to keep the paragraph unchanged.
i f
- - . , . - - , _ . - . _ _ , . _ _ _ . , , , . , , . , - , _ , , _ , , , , . _,.,.__,,,.,_,_,.,.-.-m__ , y .2-.-.
- .= - -
Coment on Page 14, Paragraph 2, Sentence 2 We question the justification that "the failure mechanism can be comon mode failure for a given plant," since motor burnout, in our opinion again, is more a result of an incorrectly set up MOV.
Response to Coment on Page 14, Paragraph 2, Sentence 2 l There is a high probability that a single individual (or a few people) may set most of the valves. Thus, if they are setting up incorrectly, the result could be comon mode failure of several valves.
1 i
. , - --m,, - , , _ _ - - _ . - - . - _ . , - . - - . . . . . _
.._...-._...-,_,.----w, . , . ~ , , . . . . . . - ,
g . Coment on Page 14, Paragraph 2, Sentence 4 It is not apparent to us how a motor failure could prevent manual operation of the valve.
Response to Coment on Page 14, Paragraph 2, Sentence 4 The motor burnout resulted in cracking the valve operator so that operator components were Jamed which prohibited manual operation of the valve.
1
Comment on Page 19, Paragraph 1, last Sentence l
Again, our position is that motor burnout should not occur with a properly set l up motor-operated valve.
Response to Comment on Page 19, Paragraph 1 Last Sentence We propose to retain the wording based on the reasons provided in the response to the comment on Page 2, Paragraph 1.
l l
1 4
. . Comment on Figure 2.2-1 The springpack deflection (thrust) signature shown in this figure is for a close to open cycle rather than an open-to-close cycle. The thrust signature as shown is technically incorrect and should be revised so as to avoid confusion with those knowledgeable of signature analysis in the industry.
Response to Comment on Figure 2.2-1 This figure was reproduced from reference 30 of the preliminary case study.
The intent of the signature was to show representative behavior of maximum thrust and running thrust from a reference point without signifying which direction was positive or negative. Since discussing the issue with you, we have had communications with the author of the referenced document and have obtained permission to revise the signature curve in question. The change was incorporated into the report.
l 1
l ,- . . _ -
1
f Comment on Figure 2.2-2
, The springpack deflection (thrust) signature shown in this figure is for an open to close valve cycle rather than close to open.
Response to Comment on Figure 2.2-2 Based on discussions as indicated in the response to the comment on Figure 2.2-1, this curve was changed with the permission of the author of the referenced report.
I l
-,= . , . .. . . - - - . - - . _ , _ - - . - - - - - . . - - .
. - . _ - _ - = . . - . ._- -- .- - - _- . _ - - .. - . _ .
. . i Comment on Page 25, Paragraph 4 From a technical accuracy point of view, we would suggest that the words "in other words it shows decrease in thrust" be deleted.
4 Response to Comment on Page 25, Paragraph 4 Table 2.2-1 was reproduced from reference 30 in the preliminary code case.
The author of that report agreed to this change and the deletion change was incorporated in the final report.
1 q
1 i
i I
- . ~ - - , , , , - _ , . - - . . .
.n.. - - ,- - . - - - , ,. - - , - , - - . . - - - . -- , , - , , - . - - . . - ~ , , . , , - - - , _ , -- , .-
33 -
Comment on Page 26, Paragraph 11 We would suggest, again for clarification purposes, that the first sentence of this paragraph be revised as follows: "When properly set, the bypass switch opens just after hammerblow or valve unseating."
. Response to Comment on Page 26 Paragraph 11 We do not believe the qualifying phrase "when properly set" is necessary when paragraph 11 just explains the intent of the bypass switch to open.
I i
l i
l f
Comment on Page 27, Table 2.2-2 We would suggest that the word " excessive" be deleted in describing springpack gap and packing tightness since the criteria for this work has not been established in the draft report.
i Response to Comment on Page 27, Table 2.2-2 i This table was reproduced from reference 30 in the preliminary case study.
The author of that report indicated that the use of the adjective excessive was to illustrate valves greater than normal. We propose that the Table not be changed.
t i
f 1
l 4
4 i
l Comment on Page 27, Table 2.2-?
Grease hardening should not be reflected in this particular table since it is not normally detectable per se by a M0 VATS signature, rather by visual inspection.
We strongly disagree with the column shown in this table as " time dependent degradations" with the potential exception of grease hardening. From a technical standpoint, we have no justificaticn that any of those degradations occur other than due to improper setup of the M0V. The term " incorrect torque i
switch calibration" should probably be revised to " torque switch unbalanced."
Response to Comment on Page 27, Table 2.2-2 This table was reproduced from reference 30 in the preliminary case study. We recognize that some types of degradation and classification such as time dependency are subject to interpretation. We propose to leave these areas unchanged on the basis that it was the author's interpretation with the information available. The term incorrect torque switch calibration used in the reference was the sum of other degradations involving the terms "torgue switch setting excessive" and " torque switch setting inadequate." We propose to leave the Table unchanged because it was reproduced from reference 30.
d
. . Comment on Page 28, Paragraph 1 The statement " incorrect torque switch calibration" appears to be definition of an unbalanced torque switch condition.
Response to Comment on Page 28, Paragraph 1 We retain the wording based on the explanation given in the Response to Comment on Page 27, Table 2.2-2.
s i
1 f
j
_,[
37 -
Comment on Page 29, Paragraph 2 l
We suggest that in describing the consecuences of a springpack gap that the discussion be limited to the statement that a premature trip could result, since the torque switch senses the gap, however, it translates this gap motion as an incorrectly assumed compression of the springpack or a developing stem loading condition.
Response to Comment on Page 29, Paragraph 2 This portion of the report was revised to reflect these coments.
t
l j
38 -
i i
Coment on Page 30, Paragraph 2 i In reality, the primary indicator of a loose stem nut lock nut is a change in
!4 y;. the time at ONich hamerblow and disk unseating occur. \
~~. .
ResponsethCommentonPage30, Paragraph 2 This portion of the report was revised to reflect this comment.
'I .v ll
.p (
l; * .,
b
'1 1
l k
- ( *-
[
A
,' Comment on Page 31, Paragraph 1 Quoting an unbalanced ratio of 3 to I can be misleading and should be clarified. 3 to:1 does not necessarily mean that the load in one direction is three times the. load in the other direction for the same torque switch setting. We would be willing to discuss this further with you.
Response to Comment on Page 31, Paragraph 1 We recognizant that the ratio of 3 to 1 does not mean the load in one direction is three times that in the other direction. Since the explanation would require excessive detail, we have modified the report to delete reference to the ratio 3 to 1, but have retained the concept that an unbalanced torque switch can lead to possible potential excessive or low operator thrust loads which may result in damage or inoperability of the valve assembly.
n +
.\
\
i l
l l
Comment on Page 47, Paragraph 2 We suggest that you review the sentence dealing with the torque switch settings being changed from 1.5. It is not clear to us that as presently worded it is technically correct.
Response to Comment on Page 47, Paragraph 2 The sentence is correct. However, we have rephrased the sentence to clarify the settings and resolve the concerns that were raised.
I 1
. 41 _
Coment on Page 53, Paragraph 3
, Again, we suggest that the basis for the statement that an overload. trip was caused by reversing the operating direction of a valve in mid-stroke be confirmed.
Re_sponse to Coment on Page 53, Paragraph 3 i
We retain the wording based on the reason provided in the response to the
- coment on Page 3, Paragraph 1.
i i
i l
-, ,--r- - = .,s., ,-- - . - - - - - - - - . - -e-,--- - , - -
me- - . - - + - g -- --m-e--- w-- w-+-- - - - - w - , r--- --w +w no--- -
. i Comment on Page 54, Paragraph 1 We would suggest and only approve authorization of the " industry average" information which has been previously published can be made available upon request. The industry average information is available from recent ANS publications which we would be willing to provide.
Response to Comment on Page 54, Paragraph 1 This section of the report was changed to included the ANS publication that should be available in July 1986. Also, the " Industry Average" column in Table 2.6-1 was revised to be consistent with the new reference and a note was added to the Table to identify the reference.
Coment on Page 59, Paragraph 5 Again, we point out that reversing direction does not necessarily infer that a l high loading or unknown leading condition will exist.
Response to Coment on Page 59, Paragraph 5 I
We retain the wording based on the reason provided in the response to the Coment on Page 3, Paragraph 1.
0
, . Comment on page 64, Paragraph 2 We wish to point out at this time for your information that M0 VATS Incorporated has developed a system for use at the MCC, which will enable the user to determine that (1) whether normal operating loads, under the zero differential pressure conditions, have increased to a point indicating the possible development of a degradation which could affect operability and (2) more importantly, whether or not the operator has, in fact, sufficient thrust available between its present operating load under the no-load condition to accommodate the maximum pressure differential condition. This new system has only recently been introduced to the industry and should A.E.0.D. or other element of the N.R.C. be interested, we would be willing to demonstrate and technically explain the system to interested parties.
Pesponse to Comment on Page 64, Paragraph 2
, The nature of the comment does not require a change in the report. The information about a system for use at the FCC was forwarded to NRR for information in their programs concerning valves.
i
.- . ~ , - . - , _ . _ . , - --_e--
1
., , COMMENTS FROM NRR AND AE0D RESPONSE Memorandum from H. R. Denton to C. J. Heltemes, Jr., dated June 6, 1986 General Comment The discussion in Section 1.3 on page 16 of the report, with respect to the schedular problems associated with the resolution of Generic Issue II.E.6.1, is somewhat misleading. In addition, work is in progress within NRR, RES and I&E on the resolution of several topics which concern check valves and the schedule for results is not firm a this time. A new task action plan, which reflects recent events and findings, as well as the new NRR organization, is in preparation and will be issued shortly. Work is not currently being delayed, as the report implies, but is being actively pursued. However, as the report points our in some detail, this issue is complex and will require extensive effort in order to resolve.
Response to General Comment 1 The report was changed to indicate a new task action plan was being prepared.
General Comment 2 In section 4.0 (2) of the report, a concern is identified in which a valve assembly is tested and performs satisfactorily during the test but, unknown to the plant personnel, becomes inoperable either as a result of incorrect performance of the test or some incorrect action taken directly after the test. It is pointed out that this situation results in having an inoperable component which is relied upon to perform a safety function and, because the component has just been tested, its condition could go undetected for a long period. It is recognized that such a situation is related to inservice test methods themselves. The quality assurance aspects of performing tests and/or maintaining components which are being questioned by this finding of the report are not currently being investigated as part of the resolution of Generic Issue II.E.6.1. They are being addressed to some extent through broad NRC and industry programs aimed at improving maintenance and surveillance practices; however, this may be inadequate. Therefore, the recommendation made in the report that this situation should be investigated under Generic Issue II.E.6.1. "In-Situ Testing of Valves" may be appropriate and we are considering it.
I Response to General Comment 2
, We believe the AE0D recommer.dation is appropriate.
1 l
i
l
- 47 _
, General Coment 3 The report contains many recomendations regarding enhancement of current MOV testing criteria for switch settings. Consideration should also be given to adding a recomendation suggesting adequate record keeping regarding these items in order to establish trends for future evaluation. (i.e., main-tain all records of switch settings and signature tests etc.)
Response to General Coment 3 3
We anticipate the subject of record keeping in order to establish trends would be a part of the overall procedures, surveillance test procedures, or maintenance practice that was addressed in recomendation (3) even though the report does not specifically mention the item. Also, the use of signature tracing equipment should include record keeping for all current settings and readings. This was covered in recomendation (1)(b).
1 I
. . General Coment 4 The AE0D study makes extensive use of the finding of two AE00 reports.
AE00/C203, " Survey of Valve Operator Related Events occurring During 1978, 1979, and 1980,"~and AE0D/S502. " Evaluation of Recent Vale Operator Motor Burnout Events."
The report AE00/C203, issued in May 1982, disclosed various causes that resulted in valve motor burnout events and made specific recomendations for improvement. However, based on the more recent findings of report AE0D/SE03 issued in September 1985, the valve motor burnout events appeared to be occurring more frequently than before (180 events for the recent 4 years as compared to 19 events for the 3 years from 1978 to 1980). It should be made
- clear what causes contributed to the large increase in the frequency of the motor burnout events and what progress the utilities and equipment vendors have made or are making in coming to grips with the problems.
Response to General Coment 4 The report 5503 discusses the large increase in the frequency of motor burnout and identifies that no progress has been made concerning the problem. The report recommends that NRR assess this situation and it was included as part of Generic Issue II.E.6.1 i
i l
. . 49 -
General Comment 5 l We note that the AE00 study does not provide a comparison of M0V failure events among MOV assemblies of differe'nt designs or POV assemblies made by different vendors, e.g., the performance of motor-operated cate valves compared with that of motor-operated globe valves and comparisons as to which equipment vendor has more mechanical problems, such as loosened parts or improper assembly, than the others.
Response to General Comments 5 We believe the AEOD studies have identified generic problems of MOVs rather than globe valve or gate valve problems. Also, mechanical problems have been related to misuse of a valve in service. In addition, all valves use the same motor operator. Thus, the evidence does not suggest that specific valve designs are specific problems.
General Comment 6 I
The report provides an evaluation of industry maintenance practices on POVs based on past experience and, therefore, does not consider IE Safety System Functional Inspections in which MOV designs and maintenance activities were reviewed. These reviews include the verification of programs for valve testing, programs for preventive / diagnostic techniques and program interfaces within the maintenance orgainizations of the inspected licensee.
Response to General Comment 6 The reviews of operational data suggest that MOV problems have existed for at least eight years with no improvement and the same things continue to occur.
Thus, it is evident that existing programs do not address the root cause of the problem. We believe that implementation of the AE00 recommendations would correct this situation.
1 e
)
- Specific Comment 1 In the last paragraph on page 10 of the subject report, reference is made to AE00 Report AE0D/T410 which describes the omission of an electrical bypass circuit for a torque switch and subsequent discovery that a significant number of other valves at the subject plant and other plants also had the same problem. The final recommendations of the report do not appear to address this concern which involves the omission of equipment or components.
Response to Specific Comment 1 The omission of the bypass circuit was addressed by issuance of IEIN 84-13.
However, recommendation 3 was changed to cover the concern by' adding the phrase
" including operational effects from omission or improper installation of components or bypass circuits."
i e
-.--r , c - _ - . - - . --.
c- . _ -- ,m7_ _- _ - ,, _- _-. _ _ _ y _..,- ., ,,, ----- . - ,- - - - - - --
4 Specific Coment 2 The findings of several IE documents are discussed in the last paragraph on page 14 of the report and reference is made to " misapplication" of motor-operatedvalves(MOVs)atthetopofpage15. Do any of the findings indicate that some MOVs are dedicated to too many potential service requirements (e.g., open or close several times to fulfill a safety function in response to diverse logic such as the Davis-Besse auxiliary feedwater valves) making it difficult or impossible to properly adjust the valves for all required conditions? If so, a recomendation should be made regarding this problem.
Response to Specific Coment 2 This area about IE documents was revised to relate them to issues similar to the AE00 studies. In addition, the recomendations section was changed to provide more guidance with specific reference to specify all conditions of operation to ensure that necessary requirements are covered.
1 1
_ . - _ _ . . . _ _ . . . , , _ _ - _ . - _ . _ _ _ _ _ . _ _ - , . _ . _ , . _ . _ , , _ . - _ __ __.,-...___-,_.._.__,,m.e
Specific Comment 3 In the second paragraph on page 18 of the report, reference is made .to what is considered to be a human factors consideration in that the maintenance philosophy at several operating plants appears to be to correct problems in order to return the valve to service without making significant efforts to investigate root causes of failure. IE Notice 86-29 dated April 25, 1986, entitled " Effects of Changing Motor-Operated Valve Switch Settings" implicitly refers to such practices. Recommendatons should be made to develop guidelines for quality assurance and training which will perhaps encourage a more 3
thorough and aggressive investigative attitude on the part of licensees.
Response to Specific Comment 3 The recommendations were revised to include specific recommendations for development of training programs to address these concerns.
l l
I
(
54 -
Specific Comment 4 On page 20 of the report reference is made to test conditions that "could be less severe than design basis conditions." It is observed that most' currently invoked inservice test conditions for valves are universally less severe than design basis conditions or even normal operating conditions. This should be pointed out in the report.
Response to Specific Comment 4 The phrase "could be" was changed to "usually are."
e 1
l l
l l
e i
1 i
l
. , . Specific Coment 5 ,
In the second paragraph on page 30 of the report it is indicated that a
" loose" stem-nut locknut will result in a low running load on the MOV. This statement appears to be inaccurate in that a " disconnected" stem-nut would exhibit such behavior but not necessarily just a loose stem-nut lock nut.
, Response to Specific Coment 5 I
This paragraph was revised. The revision eliminates the issue raised in this
] comment.
O 1
i
, I l
l l
i I
3
, . Specific Coment 6 At the top of page 32 of the report, it is parenthetically stated that an MOV torque switch could still actuate while being bypassed. This statement is considered somewhat confusing since it would not necessarily be known that the switch actuated. Additionally, the report does not explain what significance the switch actuation is, even if it were known that the switch actuated.
Additionally, on line 3 of page 32, it is stated that the toroue switch would
" remain" in the control circuit. This wording should be changed to say
... torque switch would re-enter the control circuit too soon and a hamerbl ow" . . .
Response to Specific Coment 6 The significance of the torque switch actuation is to illustrate that the bypass circuit does not modify operation of the torque switch in any way.
Thus, if the bypass switch setting is incorrect or a high sustained load is encountered that exceeds the bypass setting, then a torque switch trip could occur and terminate valve assembly operation. Line 3 on page 32 was changed to accomodate other coments so that the concern raised by this coment no f
longer appears.
l i
_ , _ , _ _ , _ - _ _ . - _ . . . _ _ . _ . - . , . _ , _ _ ._ . _ _ _ . . . _ _ . , _ . . _ . _ _._._.-..-e , -.
. a Specific Comment 7 In the first paragraph at the top of page 37 of the report, there is a discussion of valve '" operator" failures and further on reference is made only to " valve" failu.res. This paragraph should be clarified with respect to which components are being discussed (i.e., motor-operated valves including their operators or just the valve operators, or just the valves).
Response to Specific Comment 7 The primary issue is valve assembly (valve, operator and accessories) inoperability. The report discusses this in Section 1.1 and indicates that various terms involving motor operator or electric valve operator are introduced to identify the specific search strategy used for the available data bases. The issue raised on page 37 is in this category. The phrase
" valve assembly" has been used in the report to refer to the combination of valve, operator, and accessories as a system. The problem arises because categories to report failures utilize the phrase " valve operator." We believe the report adequately addresses these concerns.
. . . . - , - - - _ --y - - -- .- -,---.,. -_ .. - . . -
Specific Comment 8 Page 1: 2nd paragraph - Signature tracing techniques are good diagnostic tools, however, the more conventional meggering tests for various environ-mental conditions (such as steam, humidity, and wet weather) should also be performed to compliment the signature tracings.
Response to Specific Comment 8 The report concentrates on overall conceptural problems and potential solutions. We believe meggering testing is useful and is used by some licensees. In addition, that testing would fit into a thorough diagnostic i
program discussed in recommendation 3 even though the ters is.not specifically referenced.
4
)
l
Specific Consent 9 Page 2: 2nd paragraph - It may be appropriate to include industry reports such as INP0's SEE-ins and other MOV-related evaluation reports in the AE00 review.
Response to Specific Convent 9 The data bases used were limited to those that are available to most people.
e a
S
. s Specific Comment 10 Pace 3: 2nd paragraph - It appears that the second new area of concern is attributable to operator error or inadequate operating procedures. The reversing of valve direction while the valve is in motion is contrary to vendor instructions in that some elapsed time must be provided between valve strokes.
Further, it does not seem that there is adequate data in the AE00 data base to support this as an observable trend, but only as isolated cases. l Response to Specific Coment 10 The concern raised was actually discussed on page 54 of the preliminary case study. That portion was kept in the final report to indicate.that the data was not adequate to assess the extent of this problem when reversing the disection of motion during mid stroke. However, since the failure mechanism could occur, it is important to keep the information in the report.
. -' Specific Comment 11 Page 5: Recommendation (3) - Data from a recent DNFT survey of Resident Inspectors regarding licensee maintenance programs (NUREG-1212) indicate that most plants. (70%) have some system in place for trending and analysis of reported failures, and almost all plants (93%) conduct post-trip and pre-start-up analyses on equipment that has failed. It seems that the root cause of the AE00-identified problem is not in the requirements for a establishing procedures, but rather in the lack of technical expertise and licensee ability to investigate and diagnose during maintenance and restoration of the failed equipment.
Response to Specific Comment 11 The conclusions'section of the preliminary case study identified the need to
! understand equipment operation in order to enable diagnosis of valve assembly failure to operate. The intent of the recommendation was to cite the need to develop the expertise (and equipment) for diagnosis and develop appropriate l procedures to accomplish the effort.
4 l
l
Specific Comment 12 page 18: 2nd paragraph, 3rd sentence - It appears from the previous discussion that (1) the lack of training on a specific type of equipment (2) the involve-ment of equipment vendors in the establishment of a more technically sound and morereliablere'solutiontothesettingoftorqueswitchproblem,and(3)the lack of performance of appropriate analyses of the interaction between the torque switches and other valve sub-actions or parts are the main issues.
Therefore, the report should be more specific on the recommendation for
" development of methods and procedures for setting torque switches." Such methods and procedures may have already been established with the adequacy and completeness being in question.
Response to Specific Connent I?
Evidence suggests that setting of torques switches involves more than just awareness of a certain value. The process involves a predetermined policy, a knowledge of system interactions to determine maximum loads, and knowledge of operator action and requirements. It is for these reasons that the reconnenda-tion was kept in broad terms to develop methods and procedures rather than just find the torque switch value.
l Specific Comment 13 Page 22-23: Technical details on these pages are rather lengthy and, where applicable, references should be made to NUREG/CR-4380. Relevant and essential information from the NUREG should be extracted to correlate the NRC signature
! tracing test and the AE00 database review. Section 2.2 and 2.5 should be organized in similar fonnats and be utilized to support the AE00 findings in a j similar manner.
i
! Response to Coment 13 I
I
! Section 2.2 of the report was intended to provide both a background understanding -
l of signature tracing test equipment and the NRC signature tracing test program -
! results. The body of the report identifies which figures and tables were i reproducedfromthereferenceddocument(NUREG/CR-4380). In addition, the tables also reference that document. The detailed descriptions of certain phenomena i
illustrate the physical significances and effect on valve assembly operability.
This was only provided for the six most predominant types of degradation observed during the test program. Section 2.5 does not follow the some format as Section 2.2 because it reviews specific valve assembly failures and does not
- attempt to provide numbers of failures for each category. However, a description l of the failure relationship to the major degradation categories discovered ,
i during the signature tracing testing is provided when appropriate. I l
l 1
)
f i
i t
)
.- I 64 -
Specific Comment 14 ,
] Page 41: The Table shown on this page should be broken down in more detail to ,
reflect those failures discusseo on pages 44 and 46. / l
\
i I
Response to Specific Comment 14 i I e The subject table was changed completely as a result of additional data and ,
I analysis suggested by other peer review comments. The failure depicted in the table in question generally can not be subdivided in greater detail because the database does not provide that information. In fact, the revised analysis for this table suggests that the information does not represent the root cause of valve assembly failure, but is more related to symptoms. Also, the failure identified on pages 44 and 40 were the result of detailed analysis of specific events where the root cause of failure was identified. Thus, there is no easily defined cross reference as suggested by this comment.
I I
l h
I l
65 -
I Specific Comment 15 Page 44: Failu're mechanisms as listed should have associated statistical distributions (in percent) to highlight the importance of each mechanism.
3 Resporse to Comment 15
//
Many of the failure mechanisms identified on page 44 were established through a detailed investigation of a few events. The general population of failures has not been reviewed in that detail so that it would not be feasible to provide a failure distribution to each mechanisms.
f
\
1 Specific Comment 16 Page 49: Make reference that Figure 2.5.1 is for the Davis-Besse plant.
Response Specific Comment 16 i
It was mentioned in the body of the report that Figure 2.5.1 was reproduced from Reference 2 for Davis-Besse. However, a note was also added to the figure to identify that the schematic applies to the Davis-Besse plant.
4 e
"~
\
l 4
l 1,
s 1
67 -
Specific Comment 17 Page 53: 3rd paragraph - Identify if the PORY block valve is motor-operated.
Response to Comment 17 l The operating data review was restricted to events involving motor-operated valves as depicteC by the title of the report and explained in the scope section of the report. Since all events are of this nature, we do not believe it is necessary to identify the block valve as an MOV when it was not done for the other events in the report.
/
.g=
I
= _ - .- _- -. . -
l, , I Specific Comment 18 !
I Page 54: 1st paragraph - The term " Industry Average" is misleading since the failure data is based on the failure rate of a limited number of valves tested.
l Response to Specific Coment 18 The report clearly defines the term " Industry Average" to be those valves throughout nuclear plants that have been tested with the same type of signature tracing equipment that was used during the NRC test program and at the Davis-Besse plant. Also, the test population includes several hundred valves (300-400) located at more than 27 nuclear plants. The term was introduced to indicate the broader spectrum of tested valves in contrast to l the limited NRC test program.
I 1
i l
Specific Comment 19 The title of the study indicates that all motor-operated valves were studied, however, all technical problems identified in the report are related to one specific vendor,' i.e., Limitorque Valve Operator. DHFT suggests that the title be changed accordingly.
Response to Specific Comment 19 i
The only restriction of the database search strategies was that the event involve electric motor-operated valves. There was no effort to search Limitorque valve operators. Additionally, there is no doubt that Limitorque valve operators represent possibly 90 to 95 percent of the motor operators used in the industry. However, the technical problems cited do not appear to be restricted to one manufacturer. In fact, we are aware of several LER reports involving valve operators from a different manufacturer. Therefore, we do not believe the title should be changed.
- y. - . ~ _ . .,,..,.---,.--_,e-- -
7- r, ,,m, , , y--
l
. 3 I
COMMENTS FROM REGION II AND AEOD RESPONSE l Memorandum from J. N. Grace to C. J. Heltemes, Jr., dated May 13, 1986 General Comment The report provided a comprehensive assessment of motor-operated valve performance which would provide a sound basis for further regulatory action regarding this important issue.
It appears to us that little improvement in performance or reliability has occurred in this area and that current programs do not assure that valves will !
be operable when needed. We suggest that a new bulletin or generic letter to direct licensee activities toward a resolution of these problems be prepared.
Response to General Comment The case study recommended independent industry action, including extending IE8 85-03 to all safety-related MOVs, rather than regulatory action to address MOV problems. The ED0 sent a letter to NUMARC to request that they take the lead for industry action on a timely basis. The case study recommended renewed regulatory action if the industry action does not provide effective licensee response within two years.
i l
1
l- -
72 -
COMMENTS FROM REGION III AND AE0D RESPONSE Memorandum from J. G. Keppler to C. J. Heltemes, Jr., dated May 15, 1986 General Coment Region III has completed its review of the subject preliminary case study as requested by your memorandum of April 17, 1986. We agree with the findings, conclusions, and recommendations in the subject case study report. Power plant field inspections in this area conducted by Region III reinforces the report's findings and indicate an urgent need to implement the recommendations.
l ,
l We offer one additional recommendation that a standard should. be developed that addresses all aspects of a quality valve maintenance program. It should be one that utility management could adopt with confidence, field personnel could implement, and NRC could enforce. Hence, we recoseend that findings and recommendations in your report be pointed in this direction. Such a ,
standard should have industry input, should be flexible and readily revised to I
- allow state of the art advancements, and should include the following
- (1) minimum valve technician qualifications; (2) formal classroom and on-the-job i training requirements; (3) guidelines for maintaining technician proficiency; '
(4) preventive maintenance and predictive maintenance guidelines; (5) inspection, surveillance, and post maintenance test requirements; and (6) quality control guidelines.
Fesponse to General Comment
)
AEOD considered the coment about development of a standard to address all aspects of MOVs in a very broad context. As a result, the case study recommended a concerted industry wide effort on KOVs. The EDO sent a letter to request that NUMARC take the lead for industry action. If effective industry action is not forthcoming within two years, the case study recomends renewed regulatory action.
t
- . , , . - _ , . . - _ ,m., _ _,__,-7.__,. ....y, ,..y,,_,,,,,,,m_,-_-,-.,.r,,y .,.,nw%,,. . - . -wy-v - , - -- , .
- ' Comment on Field Inspectica Observations Field inspection observations include:
Many valves were not tended until they totally failed to stroke in use or during a surveillance test.
Recent MOV experience and information publishing by NRC INPO, and EPRI had not been effective in achieving desired results in a number of licensee programs because poor or unacceptable practice continues.
Site management often supports no more than ASME Code,Section XI MOV stroke timing as evidence of valve operability. .
Industry has confused the meaning of maintenance and repair, with repair now referred to as " corrective maintenance."
Response to Comment on Field Inspection Observations
! We believe the case study addressed many of the issues observed during
- inspection. However, the final version of the case study included strengthened recommendations covering the need for site management support, development of strong training programs for the maintenance staff, and development of diagnostic capability to identify " root cause" of MOV failure to operate.
j '
i
.. . . . _ _ , . - - , . - . _ . _ - _ . - , ., - - , . , , .-,...,,..x. - . _ - _
, . .- , , ,--, , . .. . - - - - -m_.-
.~ COMMENTS FROM TOLED0 EDISON AND AE00 RESPONSE Letter from R. F. Peters, Jr., to C. J. Heltemes, Jr., dated June 20,19P6 Coment The one item of coment is that regarding page 27 Table 2.2-2. The category
" Bent Stem" should be in the " incorrect adjustments" category. Our experience shows this indication is possible, but not time dependent.
Response to Coment The Table referred to was reproduced from a reference document. The issue was reviewed with the author of the references report and it was decided not to change the Table. The issue is subjective in nature and open to interpretation.
Since the issue does not affect the conclusions of the report, it was decided that this report should not change the reference data.