ML20245K641
| ML20245K641 | |
| Person / Time | |
|---|---|
| Issue date: | 06/14/1985 |
| From: | Edison G Office of Nuclear Reactor Regulation |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20245A157 | List: |
| References | |
| NUDOCS 8905050111 | |
| Download: ML20245K641 (21) | |
Text
_
);74cA m l W 2 -~~
l i.
- /**%
UNITED STATES guygy gffg,c 4//y s
[,,,*,[#
NUCLEAR REGULATORY COMMISSION y jfg. gjyffe 7 j
g s
M 8
CASHINGTC N, C. C. 2M55
{
h f
1
\\>,,,,, #
June 14,1985 MEMORANDUM FOR:
Harold R. Denton, Director Office of Nuclear Reactor Regulation THRU:
Jesse L. Funches, Director Planning and Program Analysis Staff l
J FROM:
G. E. Edison, Chief Technical and Operations Support Branch Planning and Program Analysis Staff SUSJECT:
LEAK TEST REQUIREMENTS FOR REACTOR COOLANT SYSTEM PRESSURE ISOLATION VALVES (PIVs)
Purcose The purpose of this memo is to advise you on the various pressure isolation valve issues that have been raised by the EDO, DEDROGR and the NRR staff anc
.i l
provide a recommendation for resolution.
Backcround Since early 1981, the staff has been backfitting ors and placing new 1.
i requirements on NT0Ls (via tech specs) by requiring leak testing of all PIVs that connect the high pressure RCS to lower pressure systens.
TETs is a backfit for ors because current requirements call for leak testing only those PIVs involved in Event V (interfacing system LOCA).
For NT0Ls, the policy established and reviewed by upper management approved the forward fit.
2.
In February,1985 DEDROGR objected, claiming that the practice is causing excessive operator exposure to radiation, decreased plant availability, and diversion of utility resources.
DEDROGR also notes (correctly) that no clear safety rationale have been documented for PIV testing (to discriminate between PIVs or between systems on the basis of safety significance), other than for Event V valves.
DEDROGR further notes (probably correctly) that there are inconsistencies in the way the staff has imposed PIV testing require ents on different licensees.
About the same time (March 19, 1985) Till-1 filed a formal backfit clair with the staff in which about 775K in additional PIV testing impact is claimed, and Beaver Valley complained about the same subject.
The EDO note of 3/28/85 requested NRR to cease imposing the requirements 3.
for non-Event V valves on additional plants until completion of the CEB review. The ED0 noted he was assuming it would not take long to complete the review.
8905050111 870213 PDR REVGP NROCRGR MEETING 110 PDC l
i
. l.. ' '
\\
e.
4.
Many NRR staff are concerned because they believe that isolation of the high pressure and radioactive RCS is very' important to public safety.
The RCS PIVs are an especially important subset of valves because their failure can cause a LOCA plus equipment damage which then opens the door for operator error in dealing with an unexpected and rare situation. The TMI-2 accident was an example of such an instance.
r A large number of ECCS valve failures have occurred in actual operating practice over the years, some of them serious. A number of actual.PIV-failures and precursors to failure are documented in Enclosure 2.
It is this kind of operating experience that adds to the staff's concern.
Of course some PIVs and systems are more safety significant than others.
A conservative safety approach (the current staff approach) is to put stringent leek test requirements on all RCS PIVs, regardless of size of pipe (i.e., size of potential LOC'AT function of valve and its system, location of valve (i.e., LOCA inside vs. outside containment),
differential pressure at PIV interface, number and type of PIVs in series, etc.
Because the staff has never done a valve-by-valve safety significance evaluation, and because a wide spectrum of system and valve designs exist,
(
the staff is uncertain about the degree of safety impact of limiting PIV testing to only Event V valves.
It is clear that a reduction in safety margin from current staff practice would occur which will vary from plant to plant, but how large a reduction is uncertain until a valve-by-valve evaluation is made for each reactor unit.
5.
Because of upcoming licensing decisions (for example, Watts Bar) and IST 10-year reviews, an early decision is needed on the issue.
A more extensive background and sunnary of correspondence on RCS PIV testing is provided in Enc 1osure 1.
- The Issue The issue boils down to a few questions.
l 1.
Should the staff expend resources to refine the safety rationale for I
leak testing of PIVs? The objective would be to develop a range of leak test requirements to cover the spectrum of RCS PIVs and plants, improve consistency in applying our requirements, and increase our understanding j
of the safety significance of each RCS PIV.
l 2.
Regardless of the answer to (1), if PIV leak testing of non-Event V f
valves is suddenly stopped, is the magnitude of the reduction in safety margin so uncertain and variable that the staff should stand firm with its i
current approach? Or, in contrast, is the magnitude small enough to be acceptable?
l l
?
1 1
4 Options At present there are 3 principle options NRR can pursue in addressing these questions.
1.
Continue the current apprcach.
2.
Continue the current approach,.and initiate in parallel an evaluation of the safety significance of all RCS PIVs (the work for Event V valves is essentially complete). Develop a range of leak test requirements for all reactors and RCS PIVs based on safety significance.
Process through CRGR, and implement with c)nfirmatory orders.
'3.
Stop imposing non-Event V leak test requirements on (a) NT0Ls, and (b) those ors submitting IST programs after 10 years of operation, until the evaluation and. criteria in 2 are complete and implemented. A variatice, of this option would sirvly eliminate the study, and test only Event 7 l'
valves-from now on.
Evaluation of Options Option 1 provides the most assurance at this time that an RCS LOCA will r.ct occur as a result of PIV failures. By requiring leak testing of all RCS FIVs, the. staff is conservatively dealing with its uncertainty in the safety significance.of various.PIVs. The uncertainty exists because the staff bas not expended the resources to evaluate the safety significance, in a
' discriminating ' fashion, of the wide variety of safety systems and PIV designs and configurations in over a hundred reactor units.
Option 1 also has the advantage of being in place now with responsibilities and schedules assigned. The staff has been implementing it for over 4 years.
In addition, Option 1 requires minimum staff resources at a time when our available resources are being reduced.
. The negatives of Option 1 include operator exposure, plant unavailability and diversion of resources, possible additional cost of modifications in certain cases, and staff / industry understanding of the relative safety significan:e of the various 20 or so RCS PIVs in each reactor unit.
No estimates of the extent of availability decrease have been seen. Hcwever, the leak tests themselves do not appear to contribute much to plant un-availability.
It is only when a PIV fails the leak test and requires repair that availability is affected. The same is true of operator exposure.
I' a l
PIV successfully passes a leak. test, the leak testing does not normally result in much operator exposure.
It is only when a PIV fails a leak test that operator exposure occurs.
For example, one rough calculation made by DEP0GR staff indicated that about 95% of the operator exposure resulted frem de311ng with test failures (e.g., repairing a valve).
g
. In certain cases there may be a cost impact because a recirculation line might Such a cost have to be added in order to isolate a valve for leak testing.
increment has not been evaluated by the staff.
At this time, the NRC staff does not possess a detailed risk-essessment backed, cost-benefit analysis to support imposition of these existing requirements. However, within the literature there are a limited number of documented studies that support the current NRR staff position.
For exempie in 1980, the NRC published NUREG-0677, a study on the relative benefits of testing pressure itolation valves. The study documented in NUREG-0677 provided estinates of LOCA frequency reductions realized from testing pressure isolation valves.
For the case where testing was assumed the total estimated LOCA frequency decreased several orders of magnitude.
It is worth noting that in 4 years there has only been one formal backfit claim filed with the staff on this matter (Tfil I on 3/19/85).
Dotien 2 would continue the current approach while en evaluation is performed and more refined leak test requirements developed. Option 2 has all the advantages of Option 1 and, in addition, provides for removing all the negatives of Option 1 after a period of time. However, more NRC resources would be required.
It is estimated the evaluation of safety significance and development of leak test requirements for various RCS PIVs for sach reactor PRA unit would take a team of several engineers 1 to 2 years to conplete.
techniques shculd play an important part of the evaluation.
For the. purposes of CRGR, cost / benefit analysis would be needed, followed by the CRGR process Ancther and, finally implementation, which would probably take another year.
advantage of doing the evaluation study in Option 2 (and 3) is that generit issues #96 (RHR valves) and #105 (BWR PIVs) could be folded in and resolved at the same time. All staff I have talked to in NRR and DEDROGR support performing an evaluation of safety significance of all PIVs which j
would serve as a basis for setting leak test requirements.
j 1
i Option 3 has the advantage of saving utilities some operator exposure and
. resources, and increasing their plant availability by a small amount.
In addition, all the advantages that would come from the evaluation study in Option 2 would be enjoyed. A major disadvantage include reduced safety margin
- it is not clear to what extent. There are at least a dozen different PIV configurations, many different types of valves, four different RCS vendors and a number of different vintage reactor designs having different emergency water
}
Without analyzing the safety significance of the PIVs of each plant, systems.
one cannot know the safety significance of failing to leak test those PIVs.
}
Another disadvantage is that the staff would have to reopen generic issue B-63.
That issue was resolved by the staff's confirmatory orders of 1981 I
(
plus the policy of addressing all other PIVs as in the current approach of However issue B-63 could also be included in the evaluation study
)
Option 1.
above.
l l
w.
, ~..-
X
- 5'-
i-4 Recommendation-i recommend NRR adopt Option 2.- The basis.is the~ uncertainty in reduction of safety margin caused by the wide variety of PIV configurations and varying safety significance of their isolation function.
Better safety rationale.can be developed which should lead to reduced testing and less stringent criteria.
However, it will take years to complete the analysis and implement.
If Option 2 is_ adopted, the staff should formalize it some way. An early decisicn is needed, or Option 3 will begin to occur by default.
t<
A G. E. Edison, Chief Technical and Operations Support Branch Planning and Progran Analysis Staff
Enclosures:
As stated
. cc:
D. Eisenhut j
a I
i i
/
\\
BACKGROUND AND CORRESPONDENCE RELATED TO THE PRESSURE ISOLATION VALVE ISSUES
~
This. enclosure provides background and correspondence related to the issue raised by CRGR and the EDO on our current requirements for pressure isolation valves.
Correspondence Via memorandum dated February 14, 1985, from you to V. Stello, NRR transmitted a proposed generic Technical Specifications change package to CRGR for review.
The proposed change relaxed the current leak rate criteria for pressure isolation valves.
Via memorandum dated February 15, 1985, from J. P. Knight to you, the Division of Engineering requested guidance on Technical Specifications regarding pressureLisolation valves for Beaver Valley Unit 2.
Via nemorandun dated February-25,1985, from V. Stello to.you, CRGR requested that the staff. provide a safety rationale for current pressure isolation valve test requirements. CRGR stated that a relaxation in leak rate criteria appeared to be appropriate, however, they were concerned over the variation
(
between licensees on the scope (number of valves) and requirements for inservice testing'of pressure isolation valves.
.Via memorandum dated March 1, 1985, from V. Stello to you, CRGR reiterated its concerns regarding the wide variability in pressure isolation valve test requirements being applied to different plants, plant systems and licensees.
CRGR requested a review of the pressure isolation valve issue prior to NRR's imposing its current requirements on any additional plants.
Via memorandum dated March 7, 1985, from J. P. Knight to R. Bernero, the Division of Engineering requested Division of Systems Integration support in responding to the CRGR request. Specifically, Division of Systems Integration
'was requested to define the valve configurations requiring testing and the safety basis.
Via memorandum dated March 7, 1985, from C. J. Heltemes to you, AE00 reported
}
the results of a recently completed case study of BWR operating events involving actual or potential overpressurization of low pressure systems.
From the results of this study AEOD concludes that the likelihood of an interfacing loss-of-coolant accident is significantly greater than had been previously assessed.
Via note dated March 28, 1985, from W. D "ks to you, the EDO directed NRR to cease its practice of imposing additional requirements on pressure isolation valve testing until a CRGR review of the issues is completed, l
l l
l i
l l
s.
4
- Backoround In October 1975, the Reactor Safety Study (Appendix V, WASH 1400) identified the potential intersystem LOCA in a pressurized water reactor (Event V) as a significant contributor to risk resulting from a core melt.
In this scenario, check valves in the injection lines of the reactor heat removal or low-pressure injection systens could fail allowing the high pressure reactor coolant to communicate with the low-pressure piping outside of containment.
Rupture of the low-pressure system would result in loss of coolant outside of containment, at the same time disabling that part of the ECCS needed to mitigate the effects of the LOCA, and possible subsequent core meltdown.
Through operating experience evaluations in the 1979-1980 time frane (see ), a significant number of Event V precursers were identified.
Accordingly, the NRR staff initiated actions to enhance safety at all operating PWRs and three operating BWRs. By issuing confirmatory orders in April 1981, operating reactor Technical Specifications were modified to impose limiting conditions for operation and surveillance requirements on the Event V isolation valves.
For example, the order modifying the Beaver Valley Unit 1 Technical Specifications was issued on April 20, 1981.
Via this order, an enhanced surveillance program was prescribed for six valves that included leakace testing during each refueling outage and prior to returning the valve to service af ter maintenance, repair or replacement.
Although the scope of the confirmatory orders was limited to the Event V valves because of the C
perceived risk (LOCA frequency and consequence), the staff reccgnized the need to address, at a lower priority, the surveillance of other PIVs.
The policy adopted by the staff at that time was:
(1) For operating reactors, to address all other (non-Event V) PIV configurations during review of the inservice test programs at the 10 year program renewal, and to then nodify the tech specs to incluce specific PIV test requirements; (2) For NT0Ls, to address all PIV configurations during the licensing review, and to include tech specs for leak testing all PIVs in operating licenses issued after April,1981.
In 1982 this policy was re-examined as the staff prioritized the safety issues identified in NUREG 0471, " Task Problem Descriptions Category B, C and D Tasks".
Documentation of this re-examination is contained in NUREG 0933, "A Prioritization of Generic Safety Issues" under item B-63, " Isolation of Low Pressure Systems Connected to the Reactor Coolant Pressure Boundary." B-63 was classified as RESOLVED based on issuance of the Event V confirmatory orders in 1981 and the policy for addressing all other PIV configurations discuss above.
The Systenatic Backfit of New Requirements The flRR staff has systematically backfit surveillance requirements for all PIVs in ors and NT0Ls since 1981 via the review of inservice inspection programs. This backfit has been completed on all plants licensed subsequent to the accident at Three Mile Island and on approximately 15 plants licensed prior to that time.
Utilizing a deterministic approach, all reactor coolant systen interface PIVs in lines equal to or greater than 13 inch diameter are j
~
1
\\
l categorized by the staff as Category A* or Category AC* as defined in Article IWV-2000 of ASME Code Section XI, and therefore subject to leak rate tests at least once every two years per Article IWV-3422 requirements.
In addition, the NRR staff has imposed enhanced surveillance requirements via Technical Specifications that include the following provisions:
(1) leak rate testing at 18 month intervals; (2) leak rate testing following each cold shutdown lasting 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or more if such tests have not been performed in the previous 9 months; (3) leak rate test for valves prior to returning a valve to service following valve maintenance; and (4) within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following check valve actuation or flow through a check valve.
Relief has been granted by the staff from items (2), (3) and (4) above where plant features such as valve position indication, interlocks that prevent inadvertent valve opening and j
high-pressure alarm annunciation for low-pressure systems are provided.
l Most disagreements on the new requirements between the NRC and utility staffs l
have been over the testing required by item (4).
Because of this requirement i
several plant startups have been delayed in order to repair leaking valves.
To prevent startup delays, some licensees have developed test programs that j
include provisions for testing and repairing valves just prior to or during a j
shutdown followed by a quick retest during the return to power.
Through this scheme a licensee can ensure the check valve is seated and avoid a startup delay.
At this tine, the NRR staff does not possess a detailed risk-assessment backed, cost-benefit analysis to support imposition of these requirements.
However, within the literature there are several documented studies that supp' ort the current NRR staff position.
For example, in 1980, the NRC published NUREG 0677, "The Probability of Intersystem LOCA:
Impect Due to Leak Testing and Operational Changes." The study documented in NUREG 0677 provides an evaluation of six different pressure isolation valve configura-tions to determine the probability of an intersystem LOCA. Valve failure probabilities were those utilized in WASH 1400. Estimated LOCA frequencies were developed for the valve configurations for cases where no testing, and For those cases where no testing was assumed periodic testing were performed.
stimates for the various configurations ranged from 2.8 x 10" LOCA frequen
,to9.5x10"gyeandforthosecaseswherepeyfodictesting,,gasassumedLOCA frequency estimates were between 4.7 x 10 and 4.2 x 10 A total estimated frequency for intersystem LOCA per plant year was derived by summing the estimated frequencies for 18 interfaces. The total estimated LOCA frequency Article IWV-2200 of the ASME Code,Section XI defines Category A as valves for which seat leakage is limited to a specific maximum amount in the closed position for fulfillment of their function. Category C includes valves which are self actuating in response to some system characteristics, such as a check valve.
r
~
for the' case where no testing was assumed was estimated to be 1.8 x 10'3 For the case where testino was assumed the total estinated LOCA frequency decreased to 4.3 x 10"6 These estimates are, at this time, considered conservative due to the higher observed valve failure rates than those used in WASH 1400.
In addition to NUREG 0677, a recent staff review of the Indian Point and Zion PRA's found that the most risk significant interface systen LOCA events are through the RHR suction line isolation valves. This issue is currently being prioritized as Generic Issue 96, "RHR Suction Valve Testing."
It should be noted that the orders modifying the Indian Point and Zion Technical Specifications issued in April 1981 for Event V did not address these RHR systen valve configurations.
Technical Specifications Supplement 1 to NRR Office Letter 38 sets forth the policy for the preparation of technical specifications for second units at multi-unit sites.
Supplerert I specifies that the technical specifications for second unit to be licensec at a nulti-unit site will be virtually identical to those of the previously licensed unit.
The only changes that are permitted are those which are absciutely necessary and justified (e.g., actual design differences or new regulatory requirements applicable only to the unit to be licensed). Althe.gh not identical, the safety-related systems for Beaver Valley Unit I and Beaver Valley 2 are essentially the same design.
Construction of Beaver Valley Unit I was conpleted and the plant was licensed in 1976; the fourth plant C
licensed under the Standard Technical Specification Program. The NRC staff has estimated that construction of Beaver Valley Unit 2 will be completed in June.1986. Accordingly, in June 1985 the staff will begin the process of
]
finalizing the. Technical Specification for Beaver Valley Unit 2 utilizing the i
Beaver Valley Unit 1 Technical Specifications as a model. The Division of Engineering has proposed that the Unit 2 Technical Specifications differ frcm
)
the Unit 1 Technical Specifications to accommodate post-TMI regulatory requirements on pressure isolation valves.
Should the Beaver Valley Unit 2 Technical Specifications nelude All Pressere Isolation Valves?
- Although the technical specifications serve as a convenient and comprehensive instrument for specifying staff requirements on pressure isolation valves, there is another vehicle available to dccument these requirements, the inservice test program.
10 CFR Part 50.55 requires each applicant and licensee to establish an inservice test program in accordance with Section "I of the ASME Code.
The Code recuires that pressure isolation valves be tested every two years.
Part 50.55 (a)(g) authorizes the Commission to recuire a licensee to follow an augmented inservice inspection program where added l
assurance of reliability is necessary.
Rather than including the Beaver Valley Unit 2 PIV surveillance requirements in Technical Specifications, these l
Once augmented requirements could be included in the inservice test program.
an inservice test program is approved by the staff, the program cannot be changed without prior NRC approval.
It should be noted that we have not consistently achieved the came level of visibility and compliance utili:irg l
1
\\
4.,-
5-l methods other than the technical specifications to document augmented Therefore, the staff has in the past utilized technical requirements.
specifications ainost exclusively to delineate augmented inservice inspection requirements.
With regard to limiting conditions for operation, there do not appear to be existing methods to address these requirements other than via the technical specifications. However, limiting conditions for operation contained within the Beaver Valley Unit 1 Technical Specifications, (3.4.6.2 OPERATIONAL LEAKAGE) in conjunction with specific provisions on leakage rates and prerequisites for reactor startup and test sequencing in the inservice test program could serve to provide a sufficiently conservative set of requirements.
Notwithstanding the arguments that can be made to include the PIVs in Beaver Valley Unit 2 Technical Specifications either to ensure ccmpliance or as a rew regulatory requirement, a suitable alternative does exist for documenting staff requirements for PIVs in the inservice test program.
I l
l 1
)
l
SUMMARY
OF REPORTED OVERPRESSURIZATION PRECURSORS AND EVENTS A review of available operating experience, data and licensee event reports shows that numerous valve failures have occurred in systems important to safety in nuclear power plants.
The events addressed herein are thought to be most of the reported valve failures that were either precursors to an overpressurization or high-pressure / low pressure boundary failures.
Sources of information for this summary include the following:
NUREG-0090 Volume 7, No. 3, " Report to Congress on Abnormal Occurrences"; memo from C. J. Helteres to H. R. Denton dated March 7,1985, " Preliminary Case Study Report"; mero from P. J. Polk to D. G. Eisenhut dated October 20,1980, " Primary Coolant
. System Pressure Isolation Valve Failures"; LER 272-80-67, Reportable Occurrence at Salem Unit 1; Daily Highlight - Oconee Nuclear Station Unit I for February 24, 1981; LER 269-81-5, Reportable Occurrence at Oconee; LER 317-05-001, Reportable Occurrence at Calvert Cliffs.
Vermont Yankee December 12, 1975 ilhile performing monthly LPCI pump and motor-operated valve operability surveillance testing with the plant operating at 99% power, LPCI injection valve V-10-25A failed to respond to an open signal generated from its remote C
control switch.
In order to determine if the failure was caused by an excess differential pressure across the valve seat, or a specific mechamcal/
electrical malfunction, V-10-25A was manually cracked open. Then the valve was successfully cycled fully open and closed.
Immediately following the cycling of V-10-25A, a steam-water mixture was observed to discharge from three RHR system relief valves and the RHR heat exchanger tube sheet to shell flange area, indicating an overpressurized condition.
Janua ry 21, 1977 Cooper Station During performance of high pressure coolant injection (HPCI) - turbine trip and initiation logic functional test, with the plant operatinc at about 97',
power, the HPCI testable check valve failed to seat (i.e., tc be fully closec) which allowed feedwater to flow backwaros through the HPCI injection line into the lower pressure rated HPCI suction pipino.
September 14, 1983 La Salle Unit 1 The plant was in cold shutdown for an extended maintenance outage.
Plant operators were conducting a routine surveillance test of the RHR system relay logic. As part of the test, an RHR system lineup was established for the "B" loop which opened both containment spray valves and the suppression pool spray valve. The test procedure then called for opening the "B" loop RHR injection l
valve, which left only the injection check valve to isolate the RHR system from the reactor ccoling system.
1
\\
. At this point, a rapid decrease in the reactor water level was observed.
Water in the reactor vessel drained from the reactor vessel for 'about three minutes until it was teminated by a combination of operator action and an automat.ic containment isolation triggered by the low reactor vessel water
)
level.
However, the reactor core was covered at all times.
The water level dropped 50 inches during the event, but remained 161.5 inches above the top of the reactor core. Normal reactor water level was restored about 45 minutes after the event using the control rod drive pumps.
I Subsequent investigation detemined that the injection check valve was stuck in the open pcsition instead of being in its normally closec position. The water (between W and 10,000 gallons) drained through the injection check valve and into tnc suppression pool through the open suppression pool recircu-lation valve and into the containment through the open containment spray valves.
September 29, 1983 Pilarim With the plant operating at about 96% power, the licensee was performing functional testing of the HPCI system logic when a HPCI high suction pressure alam and HPCI area smoke detector alarm were received in the control room.
Investigation showed that a feedwater pressure transient occurred and involved the HPCI suction piping. The low pressure section of the piping was briefly
]
overpressurized due to backflow of feedwater into the piping.
Investigation showed that miscommunication between two station personnel, while conducting more than one surveillance test at the same time, resulted in both HPCI pump discharge valves being inadvertently opened.
This left only l
the testable injection check valve to isolate the HPCI from the RCS. However, the movable internals of the latter valve were bound by rust which apparently held the valve partially open during normal operation. This allowed feedwater pressure to overpressurize the low pressure piping through the open discharge valves.
October 28. 1983 Hatch Unit 2
.With the plant in cold shutdwon, operations personnel found that an isolation check valve was open, and would not close, during perfomance of a valve operability testing procedure. The valve is located in the "B" train of the RHR system and is equipped with an air actuator for periodic testing purposes.
The valve had been open for approximately four months while the plant operated at close to full power.
The valve is a swing-type testable check valve with an air actuator controlled
]
by a four-way solenoid pilot valve. The rotary air actuator is used to I
perform in-service testing of the valve when the plant is in a cold shutdown condition. The valve, and its actuator and solenoid valve, are installed on the 24-inch LpCI line inside the primary containment structure. The valve serves as one of the two isolations between the high pressure RCS and the low
^
pressure RHR system. The second isolation valve is located imediately outside containment and is a normally closed, motor-operated injection gate j
valve. This gate valve is designed to open automatically when the RCS pres-sure drops below the low pressure permissible setpoint. The gate valve is interlocked with a third valve in a manner which prevents both valves being opened if excessive RCS pressure is present.
Even though no overpressurization of low pressure piping actually occurred, the event is significant since the check valve had been held open for such a long period of time. During the period, a postulated failure or inadvertent opening of the gate valve could allow discharge of high-pressure reactor coolant into the low pressure RHR system.
Browns Ferry Unit 1 Acoust 14, 1984 With the plant operating at about 100% power, a core spray (CS) logic func-tional test was being performed.
For this test, the outboard injection valve remains in its normally open position, and the inboard injection valve is supposed to be closed.
Since the test would sinulate automatic core spray actuation (which would normally open the inboard valve), procedures specify that the valve breaker to this valve should be opened so that the valve remains closed during the test.
However, a licensed operator failed to open the breaker; therefore, the valve opened durino tne test.
With both the inboard and outboard injection valves open, isolation of the high pressure RCS to the CS systen is provided by a testable check valve.
However, as later investigation indicated, im3 roper maintenance oreviously perforned on the check vhlve caused it to be 1 eld open while indicatino closed.
This came about as a result of an incorrect plunger being installed in the solenoid pilot valve of the actuator leading to a pneumatic pressure reversal holding the check valve open.
The check valve position indicators were also reversed by personnel in the belief that the valve was not misposi-tiened. Therefore, the high pressure RCS (about 1050 psi) was open directly to the low pressure CS system (designed for 500 psi).
~
- During the test, the contro1 room operator did not notice a system pressure change, and the CS system high pressure annunciator, located outboard of the outboard isolation valve, did not alarm.
Several minutes into the test, a roving fire watch noticed smoke near the loop 1 CS piping and phoned in a fire alarm (the smoke was caused by the pipe paint overheating when hot reactor coolant backflowed into the CS piping).
The fire brigade entered the reactor building and correctly assessed the reason for the pipe paint sroking. The CS systen's one inch relief valve, set at approximately 400 psig, has lif ted.
The assistant shift engineer phoned the Unit operator and instructed him to close the inboard isolation valve to isolate the system.
This action terminated the overpressurization event which lasted approximately 13 minutes.
Steam and/or water was seen coming from the CS pump "A" seal and several fire brigade members were slightly contaminated due this water.
l
q..i s
a 4-1 October 5, 1982 l
La Salle Unit 1 l
With the plant operating at 20% power, quarterly surveillance testing on the high pressure core spray system (HPCS) was being conducted. The testable isolation check valve IE22-F005 and its associated bypass valve IE22-F354 failed to indicated completely closed after they were opened for the test.
Both the testable isolation check valves and its bypass valve were situated on the HPCS injection line inside primary containment.
The HPCS systems was declared inoperable. The motor-operated HPCS injection valve was closed and deactivated.
During the surveillance test, the check valve bypass valve 1E22-F354 was first openec to equalized the pressure on both sides of the testable check valve disk.
The testable check valve was then tested open by operating a remote handswitch. This handswitch energized a solenoid valve to allow instrument air to be supplied to one side of the piston cylinder of the air operator of I
the testable check valve, causing the piston cylinder to move a rack and gear assembly against spring tension. The rack and gear assembly movement rotated the actuator rod which lifted the valve disk off its seat. When the switch was returned to its closed position, the solenoid valve was de-energized, cutting off instrument air su This should have allowed the spring (tension) pply to the piston cylinder.
to return the rack and gear assembly to its normal position, allowing the valve disk to reclose by its own weight and differential pressure.
The failure of testable check valve 1E22-F005 to reclose was investigated by the ' licensee and was determined to have been caused by (1) dried lubricant on the actuator piston cylinder; (2) insufficient preload en the actuator spring assembly; and (3) the stuck open testable check valve bypass valve IE22-F354 Together, these causes prevented the piston cylinder of the check valve air operator fro" returning to its fully retracted position.
La Salle Unit 1 June 17, 1983 On June 17, 1983, with the plant at 48% power and quarterly operating surveil-lance of the HPCS system in. progress, HPCS testable isolation check valve
- 1E22-F005 and its associated bypass valve IE22-F354 failed to indicate closed
'after being tested open.
l The licensee determined that the failure of the testable isolation check valve to reclose was caused by (1) the stuck open bypass valve 1E22-F354 which prevented a pressure differential from developing across the valve disk of the testable check valve, and (2) possibly thermal binding of the check valve disk. With respect to the latter cause, the licensee indicated that the Anchor Darling check valve and bypass valve have a tendency if tested hot to remain partially open after being cycled.
The failure of the bypass valve to reclose was traced to insufficient return spring tension in the bypass valve.
While shutting down the plant, both the bypass valve and the testable check valve cicsed without any assistance as reactor pressure and temperature decreased.
D Calvert Cliffs Unit 1 -- January 16, 1985
~
At 1416 on January 16, 1985, a ten minute Safety Injection Tank (SIT) check valve inleakage test was completed.
Initial results indicated excessive inleakage into two SITS. The results were presented to the Plant Operatiers and Safety Review Committee (POSRC).
It was decided that the high pressure Safety Injection flow rate specified in the Technical Specifications coulo r.ot be assured. Additionally, under certain circumstances, the SIT inleakage could render the tanks inoperable. Based on this information the POSRC recommended that both Unit 1 High Pressure Safety Infection headers be de-clared inoperable.
Reactor shutdown was completed at 1845.
Two SIT outlet check valves were overhauled on January 17, 1985. Each valve's seating surface o-ring was found approximately one-third degraded.
The Ethylene Propylene o-rings had been upgraded previously oue to their inability Ecth to withstand the temperature environment in which these valves operate.
o-rings were replaced with a more heat resistant 0-ring.
Both check valves were subsequently satisfactorily leak tested.
March 7, 1981 Ocoree Unit 3 At approximately 0700 on March 7, 1981, Unit 3 was at 292'F and 750 PSIG,v. hen the check. valve 3CF-13 ("3B" Core Flood tank (CFT) discharge) was discoverec to be leaking excessively during performance of the "Intersystem LOCA Leak Test". 3CF-14, the check valve between the reactor vessel and the core ficed
(*
discharge, was tested shortly after 3CF-13 and revealed no leakage.
The "23" Core Flood tank. motor operated valve (MOV), 3CF-2, was also closed during tnis time.
The "Intersystem LOCA Leak Test", a test performed during refuelino shutdc.ns, i
uses a pressure increase vs. time method which is correlated to GPM leakage.
Each valve is tested with a driving pressure of 750 PSIG on one side of the valve, and the pressure increase starting at 0 PSIG is observed on the other side.
.The cause of the leakage by 3CF-13 was due to t e va ve disc not seating h
l properly. The disc has apparently become cocked due to a bu11 dup of deposits in the pivot tolerance on the side closest to the hinge pin.
February 22, 1981 Oconee Unit 1 At approximately 1800 hours0.0208 days <br />0.5 hours <br />0.00298 weeks <br />6.849e-4 months <br /> on February 22, 1981, a check valve in the Uni 2 Low Pressure Injection (LPI) System was discovered M be leaking excessively l
during the performance of an intersystem LOCA leak test. The leaking vahe i
was 1 CF-12, which is the final valve in LPI loop "B" before reaching the reactor vessel.
Examination of the valve disc-hinge assembly in I CF-12 found the dise had becone frozen at the pivot in a cocked position, Consequently, only about l
half of the cisc was seating.
The " freezing" of the disc at the pivot was I
I
[
.s i ' e.
e apparently caused by a buildup of deposits in the gap between the hinge and the disc " knob" on the side of the knob closest to the hinge pin'.
While there is flow through the valve, the disc is normally in a cocked position, and it is postulated that the flow could carry deposits into the pivot gap area, where they could accumulate. The accumulation of deposits could then cause the disc to remain slightly cocked when the flow was stopped.
December 14, 1980 Salem Unit 1 With the reactor in hot standby, safety injection system check valve 1155139 was found to be stuck open during the performance of the leakage rate test.
With 1155139 open, failure of one boundary valve, 11SJ156 could have resulted in an inter-system LOCA.
A radiographic examination of the check valve revealed that it was cocked open. The valve, a stainless steel two inch check valve, resented after being tapped several times with a mallet.
October 8,1980 Davis Besse On October 8,1980, Davis-Besse plant staff conducted leak tests on a check valve which isolated the primary coolant system (PCS) from the discharge line of the low pressure injection system (LPIS). Excessive leakage caused the licensee to further investigate.
Upon valve disassembly, the licensee found that the valve disk had come free from the valve body. The two 5/8" bolts (and locking mechanism) which hold tne disk and hinge arm to the valve body were found missing. Therefore, the valve was totally non-functional.
The manufacturer of the subject 14" valve is the Yellan Company.
Valve failure was attributed to flow induced vibrations since valve bonnett gaskets had been replaced in 1978 and plant personnel had sighted correct valve assembly during this maintenance.
(The valve had last been tested in 1977).
Evidence was not offered in support of the assumed failure mechanism. The i
valve manufacturer had not been contacted to ascertain possible design defi-
,ciencies.
i I
I i
1
~_
AMadmcad4-Hasca r.@ cA M
?
Af/) # no 2/2 W8 '?
v
~,
f t
BIBLIOGRAPHY OF NRC MEMORANDA AHD OTHER DOCUMENTS RELATED TO LEAK-RATE TESTING OF PRESSURE ISOLATION VALVES
REFERENCES:
1.
Memorandem from A. Cappucci to R. Bosnak dated 2/20/80, MEETING WITH DOE TO RESOLVE DIFFERING INSERVICE TESTING REQUIREMENTS.
2.
Memorandum from R. Vollmer and P. Gamel to A. Schwencer, D. Ziemann, T.
i Ippolito and R. Reid dated 2/25/80, PERIODRIC TESTING OF PRIMARY COOLANT SYSTEM PRESSURE ISOLATION CHECK VALVES (Inc1r tes draft letter to all LWR licensees signed by E. Eisenhut).
~
3.
Memorandum from J. Knight to R. Tedesco dated 10/15/80, LEAK TIGHT INTEGRITY OF PRIMARY COOLANT SYSTEM PRESSURE ISOLATION VALVES, 4.
Memorandum from R. Bosnak and T. Speis to R. Vollmer and D. Ross dated 10/29/80, STAFF RECOMMENDATIONS ON VERIFICATION OF PRESSURE BOUNDARY INTEGRITY AT OPERATING PLANTS.
j 5.
Memorandum from R. Vollmer to D. Eisenhut dated 11/17/80, STAFF RECOMMENDATIONS ON VERIFICATION OF PRESSURE BOUNDARY INTEGRITY AT PWR AND j
BWR OPERATING PLANTS.
1 6.
Memorandum from R. Vollmer to D. Eisenhut dated 2/12/81, ORDER FOR MODIFICATION OF LICENSE: PRIMARY PRESSURE ISOLATION VALVES.
~
7.
Memorandum from D. Eisenhut to R. Vollmer dated 2/25/81, LWR PRIMARY l
COOLANT SYSTEM PRESSURE ISOLATION VALVES (WASH. 1400, EVENT V) - GENERIC ACTIVITY B-45.
8.
Memorandum from R. Vollmer to D. Eisenhut dated 3/30/81, LWR REACTOR COOLANT SYSTEM PRESSURE ISOLATION VALVES.
9.
Memorandum from R. Bosnak to J. Knight dated 1/29/82, TECH SPECS.
- 10. Draft Regulatory Guide and Value/ Impact Statement dated October, 1982, Task MS801, IDENTIFICATION OF VALVES FOR INCLUSION IN INSERVICE TESTING PROGRAMS.
j
- 11. Ec&G Report EGG-NTAP-6175 dated 2/83, INSERVICE LEAK TESTING OF PRIMARY 4' iSSURE ISOLATION VALVES.
I
- 12. tUREG-0677, dated 5/80, THE PROBABILITY OF INTERSYSTEM LOCA:
IMPACT DUE 1
TO LEAK TESTING AND OPERATIONAL CHANGES.
l 1 of 5 j
1 1
l t
j
$l Ne,.
j
.}
- 13. Memorandum from Frank Hiraglia to Darrel Eisenhut dated 3/31/83 RECENT j
OPERATING REACTOR EVENTS.
j i
.14.
NUREG 0933, A PRIORITIZATION OF GENERIC SAFETY ISSUES, ITEM B-63:
j ISOLATION OF' LOW PRESSURE SYSTEMS CONNECTED TO THE REACTOR COOLANT
=
PRESSURE BOUNDARY, page 2.B.63-1 dated 11/30/83.
(Proposed Appendix A to SRP3.9.6 attached)
- 15. Memotandum from J. Knight to R. Yo11mer dated 5/11/83,' PRESSURE ISOLATION VALVE LEAK TEST ACCEPTANCE CRITERIA - NTOL's,
- 16. Memorandum from R. Bosnak to B. Sheron dated '7/19/83 LEAK TESTING OF i
PRESSURE ISOLATION VALVES - FARLEY 182 AND MEB CRGR PROPOSAL.
l f
- 17. ~ Memorandum from B. Sheron to R. Bosnak dated 8/4/83, LEAK TESTING OF.
' PRESSURE ISOLATION VALVES.
- 18. Portion of an enclosure to a letter to Duquesue Light dated 2/9/84.
Question 210.40 (POSITION ON TESTING PIV'S).
- 19. Memorandum from P. Lainas to K. Seyfrit dated 5/31/84, STUCK OPEN ISOLATION i
CHECK VALVE ON THE RESIDUAL HEAT REMOVAL SYSTEM AT HATCH UNIT 2 (AEOD/E414).
Memorandum from C. h1temes to H. Denton dated 6/1/84, STUCK OPEN 20.
ISOLATION CHECK VALVE ON THE RESIDUAL HEAT REMOVAL SYSTEM AT HATCH UNI (Copy contains marginal coment by G. Edison. This memo forwarded reference 19.)
- 21.. Memorandum from D. Eisenhut to T. Speis, R. Vollmer, R. Mattson and H.
. Thompson dated 6/18/84, AE0D ENGINEERING EVALUATION E414. " STUCK OPEN ISOLATION CHECK VALVE ON THE RESIDUAL HEAT REMOVAL SYSTEM AT HATCH UN 22.
IE Infonnation Notice 84-74:
ISOLATION OF REACTOR COOLANT SYSTEM FROM LOW PRESSURE: SYSTEMS OUTSIDE CONTAINMENT. date'9/28/84 Memorandum from H. Denton to V. Stello dated 2/14/85, PROPOSED TECHNICAL l
23.
SPECIFICATION CHANGE AND NOTICE REGARDING ACCEPTABLE PRESSURE ISOLAT i
VALVE (PIV) IN-SERVICE TEST LEAK RATES.
j Memorandum from V. Stello to H. Denton dated 2/25/85 GENERIC TECHNICAL
]
24.
SPECIFICATI0HS REGARDING ACCEPTABLE PRESSURE ISOLATION VALVE (PIV)
IN-SERVICE TEST LEAK RATES.
{
- 25. Memorandum from J. Knight to H. Denton dated 2/15/85. TECHNICAL SPECIFICATIONS FOR PRESSURE ISOLATION VALVES AT BEAVER VALLEY 2.
(refers to reference 55)
Memorandum from G. Edison to H. Denton dated 6/14/85, LEAK TEST 26.
REQUIREMENTS FOR REACTOR COOLANT SYSTEM PRESSURE ISOLATION VALVES.
Memorandum from H. Denton to V. Stello dated 6/20/85, GENERIC TECHNICAL 27.
SPECIFICATION REGARDING PRESSURE ISOLATION VALVE (PIV) IN-SERVICE TES LEAK RATES.
2 of 5
__- - _ - _ a
^
d 4.'~
e..
28.-
Memorandum from V. Stello to W. Dircks dated 8/21/85 MINUTES OF GRGR MEETING NUMBER 79 (Enclosure 4.- Briefing on Leak Test Practices for PIV's).
Memorandum from V. Stello to H. Denton dated 11/15/85 BEAVER VALLEY 29.
POWER STATION - UNIT NO. 2 (BVPS-2) PRESSURE ISOLATION VALVE (PIV) LEAK TESTING.
(Forwarded letter from Duquenese Light dated.10/31/85)
- 30. Henorandum from R. Bernero to H. Thompson, J. Knight, T. Speis. W.
Russell and C. Heltemes dated 11/18/85 GENERIC ISSUE 105 " INTERFACING
-SYSTEM LOCA AT BWR'S" - TASK ACTION PLAN (TAC 57979).
- 31. Letter from Duquesne Light Company to H. Denton dated I/9/86. BEAVER VALLEY POWER STATION - UNIT NO. 2.
PRESSURE ISOLATION VALVE LEAK-TESTING.
- 32. Memorandum from D. Muller to G. Lainas dated 8/29/84. FARLEY.152 SAFETY EVALUATION OF REQUEST FOR RELIEF FROM IST REQUIREMENTS FOR ACCUMULATOR' CHECK VALVES AND RWST CHECK VALVE.
- 33. ORDER FOR MODIFICATION OF LICENSE CONCERNING PRIMARY COOLANT SYSTEM PRESSURE ISOLATION VALVES, dated April 20,1981 (Coninonly referred to as the " Event V Order").
- 34. Informal Summary of the Event V Orders Issued on 4/20/81, dated 5/5/81.
- 35. Memorandum.from J. Knight to R. Vollmer dated 2/24/83, RECENT OPERATING EXPERIENCE INVOLVING SAFETY RELATED CHECK VALVE FAILURES.
36.
IE Bulletin No. 83-03 CHECK VALVE FAILURES IN RAW WATER COOLING SYSTEMS OF DIESEL GENERATORS dated 3/10/83.
- 37. Memorandum from R. Vollmer to J. Knight dated 3/29/83. RECENT OPERATING EXPERIENCE INVOLVING SAFETY RELATED CHECK VALVE' FAILURES.
- 38. Calvert Cliffs LER 85-001 dated 2/8/85 SAFETY INJECTION TANK CHECK VALVE FAILURE TO PASS LEAK TEST.
- 39. Memorandum from H. Denton to J. Sniezek dated 3/3/86, LEAK RATE TESTING -
PRESSURE ISOLATION VALVES.
40.
Paper entitled HISTORY AND
SUMMARY
OF ISSUES PRESSURE ISOLATION VALVES.
dated 2/1/86.
- 41. Draft Staff Position dated 2/18/$6 entitled STAFF CURRENT POSITION ON LEAK TESTING OF PIV'S.
- 42. Memorandum from J. Snierek to H' Denton dated May 1,1986 CRGR REVIEW 0F PROPOSEDCOURSEOFRESOLUTIONONPIVTESTING.
- 43. Memorandum from R. Mattson to R. Vollmer dated July 15, 1981. EVENT V -
COMPETIhG RISK OF CLOSING A HIGH PRESSURE MOV.
3 of 5
_e1~
9 w
.,y
,44.
AEOD Case Study Report AEOD/C507 dated September 1985 0'.'ERPRESSURIZATION
-0F EMERGENCY CORE COOLING SYSTEMS IN BWR'S.
-45.
Paper entitled "AN EVALUATION OF BWR OVERPRESSURE INCIDENTS IN LOW PRESSURE SYSTEt1S" By J. D.-Harris (ORNL) and J. W. Minarick (SAIC)-
Prehminary dated May 1985.
- 46. Memorandum from P. Polk to all Operating Reactor' Branch Chiefs dated April 23, 1981, CLOSE00T OF LWR PRIMARY COOLANT SYSTEM PRESSURE ISOLATION VALVES (WASH-1400 EVENT V).
=47.
Note to H. Denton from Bill Dircks regarding imposing PIV testing dated 3/28/86.
48.
IE Information Notice 86-40 " DEGRADED ABILITY TO ISOLATE THE RCS FROM LOWPRESSURE SYSTEMS IN BWR'S dated 6/5/86.
- 49. Memorandum from C. Paperiello to T. Speis and Y. Jordan dated 6/17/86, REQUEST FOR GUIDANCE TO POWER REACTOR LICENSEES ON HANDLING CLOSURE VERIFICATION OF NORMALLY CLOSED CHECK VALVES.
- 50. EG8G Report EGG-NTAP-6175 INSERVICE LEAK TESTING OF PRIMARY PRESSURE ISOLATION VALVES, dated February 1983.
- 51. Menorandum from T. Speis to H. Thompson, F. Miraglia and R. Bernero, LEAK RATE TESTING OF PRESSURE ISOLATION VALVES (PIV'S) TO ASSURE THE INTEGRITY 0F THE REACTOR COOLANT PRESSURE BOUNDARY (RCPB), AT ALL PLANTS NOT YET LICENSED,. dated July 3,1986.
- 52. Mbmorandum from P. Polk to S. Varga, et. al., CLOSE OUT OF LWR PRIMARY COOLANT SYSTEM PRESSURE ISOLATION VALVES (WASH-1400 EVENT V) dated.
April 23, 1981.
- 53. Memorandum from J. Zerbe to R. Bernero, et. al.,
SUMMARY
AND ISSUE IDENTIFICATION FOR CRGR MEETING NO. 101, dated October 23, 1986.
- 54. Memorandum from H. Denton to J. Sniezek, dated August 29, 1966, NRR PROPOSED ACTIONS REGARDING TESTING OF PRESSURE ISOLATION VALVES (PIV'S)
- 55. Memorandum from V. Stello to H. Denton, dated March 1,1985, STANDARD TECHNICAL SPECIFICATIONS FOR PRESSURE ISOLATION VALVES (PIV'S) - TESTING AND INSPECTION (refers to reference 25)
- 56. Memorandum from J. Sniezek to R. Bernere et. al., dated October 23,19E0, CRGR MEETING NO. 101
- 57. Memorandum from J. Sniezek to R. Bernero et. al., dated November 3, 1986.
Cancellation of CRGR Meeting #101.
- 58. Memorandum from H. Denton to J. Sniezek dated February 9, 1987, ADDITIONAL' INFORMATION REGARDING NRR PROPOSED 50.5 and (F) LETTER TO ALL LICENSEES ON PIV's, 4 of 5
g 4:
i'.,I a
o.
6.
- 59. Memorandum from R. Bosnak to K. Kniel dated February 18, 1987, REVISED DESCRIPTION IN SIMS FOR GENERIC ISSUE 105.
- 60. Menorandum from R. Capra to R. Bernero, et. al., dated Febuary 19, 1987,
SUMMARY
AND ISSUE IDENTIFICATION. FOR CRGR MEETING NO. 110.
~
e 5 of 5
_ _. _. _ _ _ _ _ _ _ _ _ _ _. _.