ML20077K856
| ML20077K856 | |
| Person / Time | |
|---|---|
| Site: | Crystal River  |
| Issue date: | 06/13/1983 |
| From: | Ludington B, Overbeck G, Vosbury F FRANKLIN INSTITUTE |
| To: | Chow E NRC |
| Shared Package | |
| ML20077K859 | List: |
| References | |
| CON-NRC-03-81-130, CON-NRC-3-81-130, RTR-NUREG-0737, RTR-NUREG-737 TER-C5506-272, NUDOCS 8306160602 | |
| Download: ML20077K856 (18) | |
Text
-
l TECHNICAL EVALUATION REPORT
- ECCS REPORTS (F-47)
' TMI ACTION PLAN REQUIREMENTS FLORIDA POWER CORPORATION CRYSTAL RIVER UNIT 3 1
NRC DOCKET NO. 50-302 FRC PROJECT C55C6 FRC ASSIGNMENT 7 l
NRC CONTRACT NO. NRC 03-81 130 FRC TASK 272 hepared Q F, W. Vosbury Franklin Research Center Author: C. J. Overbeck 20th and Race Streets
- 3. W. Ludingten Philadelphia, PA 19103 FRC Group Leader: G. J. Overbeck Prepared for Nuclear Regulatory Commission Lead NRC Engineer:
E. Chew Washington, D.C. 20555 June 13, 1983 This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency tnereof, or any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or resocnsibility fcr any third party's use, cr the results of such use, of any infermation, appa-ratus, product er process disclosed in this rescrt, cr represents that its use by such third carty would nct inf ringe privately owned rights.
Prepared by:
Reviewed by:
Approved by:
')h M
Principal AutherQ Orcuo Leade Department Dir ctor]
Cate:
3 cate:
4 - es-a s cate:
r -i - a' 3 XAppy Has Been Sent.to PDR A
- b. J Franklin Research Center A Civisien cf Tne Franklin Insdtute ne Sen.arnn F arauin 7 mev. PMs., f a. :9103 (212> 4.18 ILv/J g3o6160 box Xg icie.
s TER-C5506-272 CONTENTS Section Title Page 1
INTRODUCTION 1
1.1 Purpose of Review.
1 1.2 Generic Background.
1 1.3 Plant-Specific Background.
2 2
REVIEW CRITERIA.
3 3
TECHNICAL EVALUATION 4
3.1 Review of Completeness of the Licensee's Report 4
3.2 Comparison of ECC System Outages with Those of Other Plants.
5 3.3 Review of Proposed Changes to Improve the Availability of ECC Equipment.
9 4
COtCLUSIONS.
12 5
REFERENCES.
13 l
l iii nidin Re$4310h Centef 4 Cresson of The Fraruen insange
,,en
--v w -
r
' ' ~ ' ~ " ' ' ' ' " " '
TER-C5506-272 FOREWORD This Technical Evaluation Report was prepared by Franklin Research Center under a contract with the U.S. Nuclear Regulatory Commission (Office of Nuclear Reactor Regulation, Division of Operating Reactors) for technical assistance in support of NRC operating reacter licensing actions. The technical evaluation was conducted in accordance with criteria established by the NRC.
Mr. G. J. Overbeck, Mr. F. W. Vosbury, and Mr. B. W. Ludington contributed to the technical preparation of this report through a subcontract with WESTE0 Services, Inc.
I 1
l I
V M
i l
SONnklin Research Center A h of N Fransen Insecute i
.?,
~
4 TER-C5506-272 1.
INTRODUCTION 1.1 PURPOSE OF REVIEid This technical evaluation report (TER) documents an independent review of the octages of the emergency core cooling (ECC) systems at Flordia Power Corporation's (FPC) Crystal River Unit 3.
The purpose of this evaluation is to determine if the Licensee has submitted a report that is complete and satisfies the requirements of DtI Action Item II.K.3.17, " Report on Outages of Emergency Core-Cooling Systems Licensee Report and Proposed Technical Specification Changes."
1.2 GENERIC BACKGROUh3
- Tollowing the Three Mile Island Unit 2 accident, the Bulletins and Orders Task Force reviewed nuclear steam supply system (NSSS) vendors' small break loss-of-coolant accident (LOCA) analyses to ensure that an adequate basis existed for developing guidelines for small break LOCA emergency procedures.
During these reviews, a concern developed about the assumption of the worst single failure. Typically, the small break LOCA analysis for boiling water reactors (BWRs) assumed a loss of the high pressure coolant injection (HPCI) system as the worst single failure. However, the technical specifications permitted plant operation for substantial periods with the HPCI system out of service with no limit on the accumulated outage time. There is concern not only about the HPCI system, but also about all ECC systems for which substantial outages might occur within the limits of the present technical specification. Therefore, to ensure that the small break LOCA analyses are consistent with the actual plant response, the Bulletin and Orders Task Force recommended in NUREG-0626 (1], " Generic Evaluation of Feedwater Transients and Small Break Loss-of-Coolant Accidents in GE-Designed Operating Plants and Near-Term Operating License Applications," that licensees of General Electric (GE)-designed NSSSs do the following:
" Submit a report detailing cutage dates and lengths of the outages for all ECC systems. The report should also include the cause of the outage (e.g., controller failure or spurious isolation). The outage data for nk!!n Research Center A Ome on d The Frenen insoeure s
eE >
TER-C5506-272 ECC components should include all ou'tages for the last five years of operation. The end result should be the quantification of historical unreliability due to test and maintenance outages..This will establish if a need exists for cumulative outage requirements in technical specifications."
Later, the recommendation was incorporated into NUREG-0660 (2], "NBC Action Plan Developed as a Result of the TMI-2 Accident," for all GE-designed NSSSs as TMI Action Item II.K.3.17.
In NUREG-0737 (3), " Clarification of TMI Action Plan Requirements," the NBC staff expanded the action item to include all light water reactor plants and added a requirement that licensees propose changes that will improve and control availability of ECC systems and components. In addition, the contents of the reports to be submitted by the licensees were further clarified as follows:
". The. report =ha"1d.contain.(1). outage datas.and. duration of outages; (2) cause of the outage; (3) ECC systems or components involved in the outage; and (4) corrective action taken."
1.3 PLANT-SPECIFIC BACKGROUND On March 31, 1981 (4], FPC submitted a report in response to NUREG-0737, Item II.K.3.17, " Report on Outages of Emergency Core-Cooling Systems Licensee Report and Propos.ed Technical Specification Changes." The report submitted by FPC covered the period from December 29, 1976 to March 16, 1981 for Crystal River Unit 3.
On August 13, 1982 (5), FPC submitted a second report in response to a request for additional information concerning diesel generator and surveillance testing outages. The second report covered the same period described in the first report.
FPC did not include any recommendations in its reports to improve the availability of ECC system components.
A b Franklin Research Center 4cm.aaw m r. - %.
~
5 TER-C5506-272 2.
REVIEW CRITERIA The Licensee's response to NUREG-0737, Item II.K.3.17, was evaluated tgainst criteria provided by the NRC in a letter dated July 21, 1981 (6) outlining Tentativ'e Work Assignment F.
Provided as review criteria in Reference 5, the NRC stated that the Licensee's response should contain the following information:
1.
A report detailing outage dates, causes of outages, and lengths of outages for all ECC systems for the last 5 years of operation. This report was to include the ECC systems or components involved and corrective actions taken. Test and maintenance outages were to be included.
2.
A quantification of the historical unavailability of the ECC systems i
.and. components due to test and maintenance outages.
l 3.
Proposed changes to improve the availability of ECC systems, if necessary.
The type of information required to satisfy the review criteria was clarified by the NRC on August 12, 1981 [7).
Auxiliary systems such as component cooling water and plant service water systems were not to be considered in determining the unavailability of ECC systems. Only the outages of the diesel generators were to be included along with the primary ECC system I
outages. Finally, the "last five years of operation" was to be loosely interpreted as a continuous 5-year period of recent operation.
On July 26, 1982 [8), the NRC further clarified that the purpose of the review was to identify those licensees that have experienced higner ECC system outages than other licensees with similar NSSSs. The need for improved reliability of diesel generators is under review by the NRC. A Diesel Generator Interim Reliability Program has been proposed to effect improved performance at operating plants. As a consequence, a comparison of diesel generator outage information within this review is not required. b) Frank!!n Research Center A Oromon of The Frannan Insende.
,- -,.,_. ___-, K
~
TER-C5506-272 3.
TECHNICAL EVALUATION 3.1 REVIEW OF COMPLETENESS OF THE LICENSEE'S REPORT The ECC systems at FPC's Crystal River Unit 3 consist of the following four separate systems:
o ccre flood tank o high pressure injection. (HPI) o low pressure injection (LPI) o borated water storage tank (BWST).
In Reference 4, FPC also included infctmation on diesel generator, closed
, cycle cooling water systam,M saa water. system outages. The two water systems support the ECC systems under certain accident conditions bat are not considered ECC systems. The results of FPC's re71ew were provided for the period from December 29, 1976 to March 16, 1981 for Crystal River Unit 3.
For each ECC system outage event, FFC provided the outage dates, the duration (when available), and the cause, plus sufficient description to discern the corrective action taken. Maintenance activities were included in the ECC system outage data. However, FPC did not provide any information on surveillance testing outages in References 4 or 5.
In Reference 5, referring to the NBC request for additional information on surveillance testing outages, FPC states:
"While surveillance testing of the ECC and diesel generator systems took place as required during the period of the initial report, the duration of these outages is unrecorded. Therefore, no summary of outages is forwarded as it would be meaningless withoct specific duratica information.
FPC cannot with good engineering judgment, estimate a duration for an equipment outage when such duration is not documented."
Based on the preceding discussion, it is concluded that FPC has submitted a report which partially fulfil'is the requirements of review Criterion 1, as no information concerning surveillance testing outages was provided.
nklin Research Center A Ormeson of The Frarwan insonne
TER-C5506-272 3.2 COMPARISON OF ECC SYSTEM OUTAGES WITH THOSE OF OTHER PLANTS The outages of ECC systems can be categorized as (1) unplanned outages due to equipment failure or (2) planned outages due to surveillance testing or preventive maintenance.
Unplanned outages are reportable as Licensee Event Reports (LERs) under the technical specifications.
Planned outages for periodic maintenance and testing are not reportable as LERs. The technical specifications identify the type end quantity of ECC equipment required as well as the maximum allowable outage times.
If an outage exceeds the maximum allowable time, then the plant operating mode is altered to a lower status consistent with the available ECC system components still operational. The purpose of the technical specification maximum allowable outage times is to prevent extended plant operation without sufficient ECC system protection.
'the maxi== allowable outage time, specified per event, tends to limit the unavailability of an ECC system. However, there is no cumulative. outage time limitation to prevent repeated planned and unplanned outages from accumulating extensive ECC system downtime.
Unavailability, as defined in general terms in WASH-1400 [9), is the probability of a system being in a failed state when required. However, for this review, a detailed unavailability analysis was not required. Instead, a preliminary estimate of the unavailability of an ECC system was made by calcu-lating the ratio of the ECC system downtime to the number of days that the plant was in operation during the last 5 years.
To simplify the tabulation of operating time, only the period when the plant was in operational Mode 1 was considered.
This simplifying assumption is reasonacle given that the period of time that a plant is starting up, shutting down, and cooling down is small compared to the time it is operating at power. In addition, an ECC system was considered down whenever an ECC system component was unavailable due to any cause.
It should be noted that the ratio calculated in this manner is not a true measure of the ECC system unavailability, since outage events are included that appear to compromise system performance when, in fact, partial or full function of the system would be expected. Full function of an ECC system would be expected if the design capability of the system exceeded the capacity required for the system to fulfill its safety function.
For example, if an ECC system ~
fJ2J Franklin Research Center s o~un e ne na-m enue
TER-C550 6-272 consisting of two loops with multiple pumps in each loop is designed so that only one pump in each loop is required to satisfy core cooling requirements, then asi outage of a single pump would not prevent the system from performing its safety function.
In addition, the actual ECC system unavailability is a function of planned and unplanned outages of essential support systems as well as of planned and unplanned outages of primary ECC system components. In accordance with the clarification discussed in Section 2, only the effects of outages associated with primary ECC system components and emergency diesel generators are considered in this review. The inclusion of all outage events assumed to be true ECC system outages tends to overestimate the unavailability, while the exclusion of support system outages tends to underestimate the unavailability, of ECC systems and components. Only a detailed analysis of each ECC system for each plant could improve the confidence in the calculated result. Such an analysis is beyond the intended scope of this report.
The planned and unplanned (forced) outage times for the four ECC systems (core flood tank, HPI, LPI, and BWST) and the diesel generators were identified from the outage information in References 4 and 5 and are shown in number of days and as percentage of plant operating time per year in Table 1 for Crystal River Unit 3.
Outages that occurred during nonoperational periods were eliminated, as were those caused by failures c: test and maintenance of support systems. Data on plant operating conditions were obtained from the annual reports, " Nuclear Power Plant Operating Experience" [10-13], and from monthly reports, " Licensed Operating Reactors Status Summary Reports" (14].
The remaining outages were segregated into planned and unplanned outages based on FPC's description of the causes. The outage periods for each category were calculated by summing the individual outage durations.
In the Final Safety Analysis Report (15] (FSAR) for Crystal River Unit 3, the following statement was made in Subsection 6.1.4, Tests and Inspections:
s
" Performance of active systems is tested by establishing flow and observing pressures and flows during scheduled shutdowns."
Thus, surveillance testing of the ECC systems that would normally render these systems unavailable for emergency service is ostensibly done when ECCS l
l
{
1
- bd Franklin Research Center 4 cm on er m r- %.
.a
. -. ~,.
Table 1.
Planned and Unplanned (Furced) Outage Times for Crystal River Unit 3 c;
aa
>m fh Core Flood HPI LPI BWST Diesel Generator gh Days of Plant Outage in Days Outage in Days Outage in Days Outage in Days Outage in Days 5
py Year Operation Forced Planned Fbrced Planned Forced Planned Forced Planned Forced Planned 1976 w
(n 1977 245.5 0.30 0.30 3.31 0.0 0.83 0.0 0.0 0.0 0.85 0.0
[6 (0.1%)
(0.1%)
(1.3%)
(0.3%)
(0.3%)
1978 151.2 0.0 0.0 0.75 0.0 0.38 0.0 0.0 0.0 0.49 0.0 (0.5%)
(0.3%)
(0.3%)
da 1979 215.0 0.15 0.0 0.25 0.50 1.00 0.0 0.0 0.0 1.00 0.0 (0.1%)
(0.1%)
(0.2%)
(0.5%)
(0.5%)
1980 197.3 0.0 0.0 0.0 0.0 10.00 0.0 0.0 0.0 1.25 1.40 (5.1%)
(0.6%)
(0.7%)
t Total 808.9 0.45 0.30 4.31 0.50 12.21 0.0 0.0 0.0 3.59 1.40 (0.1%)
(< 0.1 % )
(0.5%)
(0.1% )
(1.5%)
(0.4%)
(0.2%)
F3
- Nunbers in parentheses indicate system outage time as a percentage of total plant operating time.
- Commercial operation began March 13, 1977.
g I
tw I
TER-C5506-272 operation is not required. For this evaluation it was assumed that surveil-lance testing does not affect ECCS unavailability and that the observed outage times of various ECC systems at Crystal River Unit 3 are comparable with those of other PWRs.
Based on this comparison, it was concluded that the historical unavailability of the core flood tank, BWST, and the HPI systems has been consistent with the performance of those system throughout the industry. The observed unavailability was less than about one standard deviation above the industrial mean for these ECC components and system, assuming that the underlying unavailability is distributed lognormally. The outage times were also consistent with existing technical specifications. The outages of the diesel generators were not included in this comparison.
llistorical unavailability of the remaining ECC sy ' tem, IPI, did not compare favorably with the performance of similar systems throughout the ine' s try. The average unavailability for this system, shown in Table 1, is well above the industrial mean plus one standard deviation. Due to the higher unavailability of the LPI system, the outage times were reviewed in detail using the FPC submittal (4) and licensee event report (LER) information obtained for the Nuclear Safety Information Center (NSIC) of Oak Ridge National Laboratory (16]. Ihe purpose of this investigation was to ascertain the major contribution (s) to the high unavailability. This investigation revealed that most of the LPI system unavailability is attributable to failures of the decay heat pump discharge throttle valves (one in each of the two trains). The outages of these two valves contributed approximately 10 days out of a total of 12 days of LPI unavailability during the reporting period (8 2.8%).
In Reference 15, FPC states the "the decay heat pumps are providad with automatic flow control circuitry and throttle valves which maintain flow between 2800 gpm (with-a 2750 gpm low flow alarm) and 3100 gpm (with a 3150 gpm high flow alarm) to preclude pump runout during any plant upset conditions."
In addition to the automatic flow control mode, the valves have a manual mode to allow the operator to determine the flow rate. A listing of the outages of
_nklin Rese_ arch C_ enter J
TER-C5506-272 the two throttle valves, DHV-110 and DEV-lll, is provided in Table 2.
In an effort to identify failure mode trends, this listing includes all available failure information and is not limited to only those failures occurring dur'ing
~
Mode 1 operation.
From the data provided in Table 2, it can be concluded that valves DEV-110 and -111 have had continual flow control problems caused by the automatic control scheme. Review of outage information provided in References 4 and 5 and related LERs indicates that FPC had not identified a successful reoccurrence control measure as of late 1980. LER 80-036 described an event which occurred on August 27, 1980 and concludes that the cause was water in the sensing line. FPC indicated that the sensing line will be blown down af ter surveillance testing for 3 months to determine the extent of a condensatirm plam-. JPC. also ind4 e=*=d that. an. engineering evaluation of the control methodology for DEV-llo and -111 was in progress. On December 7, 1982 (17], it was confirmed that the engineering study is still in progress (over 2 years old) and that FPC is still experiencing automatic flow control problems with DHV-110 and -111.
The NRC Resident Inspector indicated that FPC has not been able to identify a successful reoccurrence control measure.
The automatic flow control problem has not affectes the ahility of the operator to take manual control; however, the system was designed to start automatically without operator action and to initiate on low reactor coolant system pressure or high reactor building pressure. In the system design, operator action was not anticipated until 25 minutes af ter a LOCA when the operator was expected to open suction valves from the reactor building emergency sump to permit recirculation of the spilled reactor coolant and injection water from the reactor building sump. Failure within the automatic control scheme would require operator action following a LOCA. During the initial licensing review of the facility, operator action to manually control decay heat removal system flow was not considered in the system design bases.
3.3 REVIEW OF PROPOSED CHANGES TO IMPROVE THE AVAILABILITY OF ECC EQUIPMENT In References 4 and 5, FPC did nct include any recommendations to improve the availability of ECC systems and components. However, it is recommended
-g_
EL Franklin Research Center 4 cm.an at wr.r a m.ou.
.-,-----T*
~
TER-C5506-272 Table 2.
Decay Heat Pump Discharge Throttle Valve Failures Power Duration valve Date Level thr)
Re ference Description DHV-llo 9/3/80 NA 48 4
failed valve operator DEV-lll 8/27/80 804 12 estimate 4, 16 failed to control flows water in sensing line DEV-110 8/27/80 804 12 4
failed to control flows air in sensing line DEV-llo 8/5/80 On 8
4 failed to control flows air in sensing line DEV-110 7/13/80 On 80 4
failed to control flow; no cause-related information provided DEV-lli 7/13/80 on 72.5 4
valve cyclings cause not determined DRV-llo 2/6/80 1004 168 4, 16 f ailed to control flows Ref.
4: caused by loose wires Ref. 16: cause unknown DEV-110 3/17/79 On NA 4
valve inoperabler no cause-related information provided DEV-110 3/11/79 04 MA 4
failed to cont ol flows cause not determined DEV-110 10/4/78 1004 9
4, 16 failed to control flows flow controller valved out DEV-111 6/21/78 0%
NA 4, 16 failed to opent cause not determined DEV-110 12/7/77 1004 0.8 4, 16 failed to control flow I
controller out of calibration DEV-llo 8/3/77 NA 12 estimate 4
f ailed to control flows fuses blown
_nklin Res,e_ arch C_ enter
TER-C550 6-272 that the cause of each DEV-110 or CHV-111 outage be examined in detail and a determination made as to whether modifications to the current control scheme can be accomplished to improve the reliability of the throttle ~ valves. Since previous attempts have not corrected the automatic flow control problem, it is further recommended that an evaluation be made of the viability of using passive flow control devices such as cavitating venturis as an alternative to active automatic flow centrol valves.
4 l
l l
I
(
l
, #8 Judd Franklin Research Center r c.n.an as m v n.u.n.au.
l
TER-C5506-272 4.
CCNCLUSIONS
- Flordia Power' Corporation (FPC) has submitted r'eports for Crystal River Unit 3 which contain (1) outage dates and duration of outages, (2) causes of the outages, (3 ) ECC systems or components involved in the outages, and (4) corrective actions taken. The reports did not include information en surveillance testing; however, the FSAR states that surveillance testing is performed during plant shutdown. Therefore, it is concluded that FPC has fulfilled the requirements of NUREG-0737, Item II.K.3.17.
In addition, the historical unavailability of the core flood tank, high pressure injection, and borated water storage tank systems has been consistent with the performance of those systems throughout the industry. The observed uDayaflability Was less than about one standard deviation above the industrial mean for the above ECCS systems. The observed unavailability of the low pressure injection system (LPI) did not compare favorably with those of similar systems throughout the industry. It is recommended that LPI outages subsequent to the reporting period be examined with particular attention given to outages of valves DEV-110 and -111.
If subsequent data identifies continued automatic flow control problems, then a determination shoulr'. be made as to whether modifications to the current control scheme can be accomplished to improve the availability of the throttle valves.
It is further recommended that an evaluation be made of the viability of using passive flow control devices such as cavitating venturis as an alternative to active automatic flow control valves.
, O bd Frank!!n Research Center A cum a at Th er.n.sa m
. -. _ ~. --.
~
TER-C5506-272 5.
REFERENCES 1.
" Generic Evaluation of Feedwater Transients and Small Break Loss-of-Coolant Accidents in GE-Designed Operating Plants and Near-Term Operating License Applications" NRC, January 1980 2.
NUREG-0660 "NBC Action Plan Developed as a Result of the TMI-2 Accident" NRC, March 1980 3.
NUREG-O'3 7 "Clariff s ition of TMI Action Plan Requirements" NRC, October 1980 4.
W. C. Jones (FPC) letter to Director of Licensing, D. G. Eisenhut
Subject:
Response to NUREG-0737 December 31, 1980 5.
W. C. Jones (FPC)
Letter to R. A. Clark (NBC)
Subject:
Additional Information in Response to NUREG-0737 August 19, 1982 6.
J. N. Donohew, Jr. (NRC)
Letter to Dr. S. P. Carfagno (FRC)
Subject:
Contract No. NBC-03-81-130, Tentative Assignment F July 21, 1981 7.
NBC Meeting between NRC and FRC
Subject:
C5506 Tentative Work Assignment F, Operating Reactor PORV l
and ECCS Outage Reports August 12, 1981 8.
NRC Meeting between NBC and FRC
Subject:
Resolution of Review Criteria and Scope of Work July 26, 1982 9
WASH-1400
" Reactor Safety Study" NRC, October 1975 10.
" Nuclear Power Plant Operating Experience 1976" NRC, December 1977
' nklin Research Center
l TER-C5506-272 11.
NUREG-0483 "Nuc1 car Power Plant Operating Experience 1977" NRC, February 1979
- 12. NUREG-0 618 "Itaclear Power Plant Operating Experience 1978" NBC, December 1979 13.
" Nuclear Power Plant Operating Experience 1979" NRC, May 1981 14.
" Licensed Operating Reactors Status Summary Report" Volume 4, ts.1 through 12, and Volume 5, No.1 NBC, December 1980 through January 1981 15.
Updated Final Safety Analysis Report, Crystal River Unit 3 Docket No. 50-302, Operating Licensee No. DPR-72 Flordia Power Corporation January 22, 1982 16.
Nuclear Safety Information Center, Oak Ridge National Laboratory Licensee Event Report Search Crystal River Unit 3, ECCS Failures October 22, 1982 17.
NRC Telephone Conversation Between T. Stetka (Resident Inspector, NRC) and G. J. Overbeck (FRC).
Subject:
Automatic Flow Control of Valves CHV-110 and -111 December 7, 1982
. O[dd Franklin Research Center IJ A Cwman of De Festwen manage