ML20042F084
| ML20042F084 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 04/27/1990 |
| From: | Shelton D TOLEDO EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| 1793, NUDOCS 9005070210 | |
| Download: ML20042F084 (10) | |
Text
g7 t: y<
Wg
? ll":
?
f.i Q;;y e
- b c
s.
TOLEDO -
-EDISON:
' a crwenor Energy cowv j
DONALD C. SHELTONL
- w. pe-wucw Docket Number 50-346 License Number NPF-3 1,
Serial Number 1793 April'27,L1990
.a
-United States Nuclear Regulatory Commission.
p
-i
. Document Control Desk Vashington,-D. C.. 20555'
Subject:
~ Response'to Inspection Report 50-346/89-201 - Intertacing System Loss of Coolant Accident (ISLOCA)
Gentlemen:
On December 22, 1989 the NRC issued Inspection Report 50-346/89-201 for a'n:
Interfacing System Loss of Coolant Accident audit (Log Number 3130) performed during the period from October 30 through November 9,.1989, at Davis-Besse Nuclear Power Station Unit 1.
.Although the inspection-report did~not require a response, enclosed is Toledo Edison's response co those specific findings and open-items discussed during-the March,28, 1990 1SLOCA meeting with the NRC at Glen Ellyn, IL.
If you'have any questions concerning this matter, please contact-Mr. R. V. Schrauder, Manager - Nuclear Licensing, at (419)-249-2366.
Very truly yours, 71 f
l~
/
RTH/ssg
!l cc:
P. M. Byron, DB-1 NRC Senior Resident Inspector A. B. Davis, Regional Administrator J. A. Isom, NRC Operations Engineer T. V. Vambach, DB-1 NRC Senior' Project Manager r
?O050'70210 900427 9
{DR ADOCK 05000346 PDC
- 8-THE TOLEDO EDISON COMPANY EDISON PLAZA 300 MADISON AVENUE
- TOLETM, OHIO 43652 e
4
.n
D'ock'et Number 50-346 -
t
'Li* cense Number NPF-3
<*/ /
Seriel Number 1793
. Attachment Paga 1 ATTACHMENT RESPONSE TO INSPECTION REPORT 50-346/89-201 Human Reliability NRC Finding; Executive Summary; Page ES-1 "The team found that operators were not aware of potential ISLOCA scenarios ~or indications of ISLOCAs and ISLOCA precursors during conduct of normal plant evolutions.
Interviews with operators and: (nchnicians, review of operator' tra!.ning..and review of various plant procedures by the team found that the potential for ISLOCAs had not been addressed by the training department nor lacorporated into proceoures used by operations."
Toledo Edison Response:
Davis-Besse operators are aware of and understand the potential for Reactor Coolant System (RCS) leaks and Loss of Coolant Accidents (LOCAs) that are equivalent to the Interfacing System Loss of Coolant Accident (ISLOCA).
Periodic training addresses "ISLOCAs" without specifically using the terminology. Toledo Edison's training program has been audited and accredited by INP0.
Plant procedures provide sufficient instruction and guidance to prevent and or mitigate the consequences of an ISLOCA. The plant emergency procedures (e.g.,
DB-PF-02000 RPS, SFAS, SFRCS, or, SG T"'a Rupture) are symptom oriented and adequately address actions required tot aCS leaks and LOCAs. ' Abnormal procedures address small RCS lesks, including leaks to' areas outside containment. The small RCS leak procedure (DB-0P-02522) provides adequate direction to mitigate the event irrespective of leak location or source.
During the period of January 31 through February 9, 1990 the Wkc performed an audit of Davis-Besse's Emergency Operating Procedures (EOP). No significant findings were identified during this audit.
Open Item 50-346/89-201-03; Section 5.1.3; Page 18 "The team noted that the plant startup procedure did not specify retaoval of control power from the Decay Heat-Removal (DHR) suction valves. The team considered this omission from the startup procedure to be important because, in addition to the MOV breakers, which were opened in the procedute, removal of control power provided a second layer of protecticn against an ISLOCA by preventing inadvertent operation of the DuR suction valves."
Toledo Edison Response The plant startup procedure (DB-0P-06901) vill be revised prior to restart from the sixth refueling outage (6RF0). The revision vill addr.ess the removal of control power from the DilR suction valves (DH11 and DH12) to supplement the existing steps to open the Motor Operated Valve-(H0V) breakers.
l l
I 6 2 Mi W M M W 4 D Nf D 's
y Dock:t Nuihar 50-346 Lictnsa Numiar NPF-3 Serial Number 1793:
Attachment Page 2 Operator Task.
.i I
NRC Findingj,p ecut h Summary; Page ES-1 NRC Findings; Section.5.1.3; Page 18 j
9 "The Team found that, since~there vere no formal procedures for responding to
.j computer alarms which could be used-to detect'ISLOCA events, the computer i
alarms could be ignored for some time before an ISLOCA event was noticed."
l
\\
"The team found that valve HP 1556, which was one of three valves required to-j open in order to vent the HPI piping that had become pressurized during the surveillances, was omitted from the procedure. Consequently, the HPI j
discharge high-pressure alarm could be in the alarmed condition for a j
prolonged period of time resulting in reduced operator. sensitivity to an actual ISLOCA event."
Toledo Edison Response.
One computer CRT is normally maintained on display for computer alarms in the;
?
Control Room.
Operators are knowledgeable of the appropriate action to take upon receipt of a computer alarm. The High Pressure Injection (IIPI) discharge u
-high-pressure alarm would come to the attention of.the_ Operator by computer i
alarm 0479 "HP INJ VLV LEAKING TRBL/NORH".
This alarm is actuated from-computer point P465.
For the specific concern identified, Davis-Besse procedures have provisions which would preclude the HPI'high-pressure alarm being in alarmed condition for a prolonged time period.
The normal system operating procedure DB-OP-06011 does not provide a lineup to depressurize this piping. -However, DB-0P-06011 requires the operator to verify the valves are'not leaking. The l
monthly jog test procedure (DB-SP-03216) contains the proper valve lineup to clear the alarm. In addition, the quarterly test procedure (DD-SP-03218) establishes the correct lineup prior to running the pump. During performance of this procedure, when the pump ~is stopped the header vill depressurize 1
through the established flovpath to the Borated Water Storage Tank (BVST),
'l thus clearing the alarm. As stated above, when performing surveillance tests, proper valve lineup is required prior to running the pump per DB-SP-03216 and DB-SP-03218. Therefore, Toledo Edison does not agree that the HPI discharge high pressure alarm could be in the alarm condition for a prolonged period of time due to the above tests being conducted on this system, j
NRC Finding; Section 5.1.5; Page 19 "The team found inconsistent references to the position of HP 1556 which is one of the three valves used to vent the HPI piping.
Duting review of plant documentation and discussions with operations personnel, it was noted that the pesition of this valve, and the type of valve, was inconsistently described.
The team's repeated attempts to determine the position for the valve were unsuccessful.
Direct observation of the local handwheel and valve stee were i
a
n Ddektt Nurb;r 50-346
.Liccns] Numbrr NPF-3
-Serial Number 1793 Attachment-Page 3-necessary to determine the_ actual positfon of the valve..The team found that various licensee' documents ~ referred to this valve as either a normally closed air-operated valve or closed, but not locked; or: locked closed. The team eventually determined-that HP 1556 vas a normally closed, manually-operated valve with its air-operator ren.oved.
The team's interviews with the operators found that t' heir knowledge of the -
valve was just-as inconsistent. At:various-points in the inspection,
- operations and maintenance personnel reported'that the HP 1556 valve status was mechanically " gutted." vith internal parts removed (i.e., normally open);
i normally closed, air operated valve; failed open, with air' removed; locked l
closed =with air removed; or normally closed with air power removed (the l) as-found condition)..
The team also found that the local control station for this valve was intact, with no tags or labels to' indicate that it was inoperable."
.l Toledo Edison Response l
I Piping and Instrument Diagram (P&ID) M-033A shows HP-1556 closed with air j
power removed.
Procedures which reference or manipulate HP-1556 vill be.
reviewed to ensure'that they are consistent, and.have correct descriptions and directions for manipulation.
Information concerning the current status of
]
HP-1556 (i.e., normally closed with air power removed) vill be provided to-all g
licensed individuals. The local control station vill be labeled to identify 4
the valves deactivated status by August 15, 1990.
l 1
NRC Finding; Section'S.I.5; Page 19 l
\\
The team found.that the plant P& ids and valve lineups c'ontained in the plant-i startup procedure (DB-0P-06901, Revision 0, October 10, 1988)' indicate'that the reactor drain valve, RC'1773A, and reactor vent valve, RC'1719A, were normally open. The as-found position for these' valves was closed, and therefore did not appear to increase the probability of an ISLOCA event.
.}
Operators were unable to provide a rationale for these valves being in a position other than what was indicated in the P&ID."
i Toledo Edison Response 4
Toledo Edison's preliminary review of the above' concerns has shown that no 4
changes are required to procedures or drawings.
The as-found positions of the L
valves were correct for the status of the plant at the time of the inspection.
g Information on recognizing the correct status of the valves vill be provided
_{
to all licensed. individuals.
NRC Finding; Section 5.2.4.2; Page 23
{
" Ladders which were rcquired to onarate some manual valves were not kept near the immediate work area where they might be used."
i k
i d
i F
b
Docket Number 50-346 License Number NPF-3 Serial Number 1793' Attachment Page'4 Toledo Edison Response l
Toledo Edison has reviewed the above concern and determined'the current locations and number of. ladders, platforms or other access methods to. operate i
manual valves are adequate. Also, the time needed to obtain portable ladders to operate the valves is reasonable.
1 NRC Finding; Section 5.2.1; Page-20 "The team also identified several concerns with regard to monitoring for an.
ISLOCA event at the local stations. Accessibility for local operation of.
valves vas cumbersome and the lighting level vas low and'non-uniform with deep shadows even with normal lighting (lighting vith emergency lights was not inspected), and operators carried flashlight as normal equipment."
i i
Toledo Edison Response Toledo Edison has reviewed the concerns and determined that there are adequate means available'to access local stations.
This may require the use of a i
flashlight and other portable lighting sources (i.e., lanterns, drop lights)
. l to provide adequate lighting to accomplish the job.
Check Valve Preventive Haintenance NRC r.
'Ing; Executive Summary; Page ES-1 "The team's review of the licensee's maintenance history indicated thatLthe licensee had repeated problems in seating one of the four pressure isolation check valves in the D!!R system and that there was no preventive maintenance program for any check valves, which is contrary to recommended industry practice."
Toledo Edison Response The Toledo Edison Check Valve Reliability Program was established in 1986.
This program was developed prior to the issuance of INP0'sLSignificant Operating Experience Report (50ER 86-03) on check valves..The program initially consisted of a review of approximately 200 check valves in 15 ot 34 systems containing check valves from the System Review and Test Program.
Elements of the program include testing, inspection and design reviews.
A short description of each element is provided below:
- 1) Testing Toledo Edison reviewed check valve test procedures to ensure that the safety function (reverse flow or forward flow) of safety related check valves was being verified by an adequate test.
In many cases, Toledo Edison vent beyond the requirements of the ASME Section XI Valve program to ensure this was being accomplished.
J
Docket Numinr 50-346 Licensa Nu;btr NPF-3:
Serial Number 1793-Attachment.
Page 5-j
- 2) Design Reviev' Toledo Edison categorized check valves as being in service at all times, only during outages, or only during testing.
Design and operating experience data was gathered and sizing' calculations completed to-provide a basis for evaluating which check valves may be susceptiole to
" chattering" or inutability. Also, check valves were reviewed to determine if the valves had a history of failure.
Valves with potential.
~
instabilities, based upon this reviev,: vere selected for inspection.
Toledo Edison identified the safety function of each valve to see which valves could-be eliminated.or replaced with another type of valve. During the fifth refueling _ outage (SRF0), eleven Service Vater valves were
' removed, six Service. Vater check valves were replacedtwith either butterfly valves or a vafer' type check valve and two containment Hydrogen Dilution _ check valves vere removed.
During the sixth refueling outage c
(6RF9), Toledo Edison.has scheduled replacement of eight Component Cooling water stop check valves 1vith manual ~ valves.-and two Auxiliary-Feed Vater valves, and one Make-Up system valve with a valve of a different design.
Additionally, based on these design reviews Toledo Edison. modified Velan,,
Schutte and Koering anti-rotation pins, replaced some hard seat' check-valves with' soft seat check valves, and made. changes to interna 1' locking devices.
l
- 3) Inspections Toledo Edison inspected seven check valves during the fifth refueling outage. During the on-going sixth refueling outage, thirty-two check valves are scheduled for inspection.
Toledo Edison's goal is to have a comprehensive prioritized' Check Valve Preventive / Predictive Maintenance Program with a basis similar to that developed for motor operated valves (HOVs)'and air operated valves (A0Vs) at Davis-Besse..To accomplish this goal Toledo Edison has determined that the Check Valve Reliability Program would need some enhancements.
Experiences from the SRF0 and 6RF0 inspections and recent industry experience vill provide the basis for any changes-to the' existing program. _The check valve PH program is currently scheduled for completion by the end of September, 1990, In addition, Toledo Edison has assumed a very aggressive and proactive role in various industry activities relating to check valves. 'The highlights of'this industry involvement includes the following:
1.
Provided input to MPR Associates in developing EPRI application.
guidelines for check valves in Nuclear Power Plants (NP-5479).
2.
Toledo Edison is a member of the Nuclear Industry Chock Valve Group (NIC).
3.
Toledo Edison chairs the EPRI/INPO m '.deline eb..
tee.
4.
Toledo Edison presented a check valve ' silt ds p at the-1989 EPRI conference.
i 1-
Docket Number 50-346 License Number NPF-3 Serial Number.1793 Attachment Page 6
~i 5.
Assisted the' Babcock and.Vilcox Ovner's Group (B&V0G) valve working group on check valve specific tasks.
Check Valve Installation NRC Finding; Executive Summary; Page ES-2 "The~ team's review of maintenance work orders (MV0s) found that two of the four pressure isolation valve (PIV) check valves in the DHR-system could experience Mgher than-normal failure rates because they were installed vertically in the DHR piping, and because they were both subjected to upstream flow disturbances. The team noted that these check valves were recommended by:
o the vendor's manual for horizontal installation only."
Toledo Edison Response; g
Per.the original design of the plant the two PIV's (CF-30 and.CF-31) were.
intentionally installed in the vertical section of. piping so as to locate these valves as:close as possible to the-reactor vessel to reduce the probability of a LOCA due to a core flood tank discharge line break. The original design conditions were taken into account in evaluations performed for the check valve reliability program.
These conditions are as follows:
1.
The-valves are 14" Velan sving check valves.
)
2.
CF-30 has a 90 degree elbow one pipe diameter-upstream.
3.
CF-31 has a 90 degree elbow six pipe diameters upstream.
4.
Normal decay heat flow of 3000 gpm vill maintain CF-30:and CF-31 in the full open "asition, j
The following operating history was also taken into consideration ir the check valve reliability program.
1.
Normally, CF-30 and CF-31 experience flow only during outages when on decay heat removal.
2.
At normal decay heat flow rate or lover, with upstream flow disturbances nearby causing unstable flou given its vertical installation, CF-30 and CF-31 are susceptible to " chattering".
3.
From 1977 to June 5,-1985 CF-30 underwent approximately 663 days of service.
4.
During the outage follc ng the June 9, 1985 event at Davis-Besse CF-30 underwent approx 1.,staly 377 days of service and at times at lover decay heat flows (less than 3000 gpm) than normal, causing a more severe wear on the:e valves, i
5.
In December 1986 and March 1987 Toledo Edison had difficulty in seating valve CF-30.
~
Docket. Number'50'-346 License Number NPF-3 JSerial Number 1793 Attachment-Page 7 Based on the above, Toledo Edison ruodified both check valves (CF-30 and CF-31) to improve operation (new discs, hinge pins and larger anti-rotation lugs were
{
installed) and changed plant operating procedures _to minimize operation of CF-30.
Although the valves are not installed per the vendor's basic i
recommended configuration, the valves vill perform their intended function.
This is further supported by leak testing of valves CF-30 and CF-31 as required by the Technical Specifications.
Opening.the valves for periodic PMs i
vould cause excessively high radiation exposure.
Therefore, a frequent PH
- 1 program to-open and inspect the valves is not varranted.
Toledo Edison is evaluating what the appropriate inspection interval should be, j
Surveillance Testing
'NRC Finding; Executive Summary; Page ES-2 "The team found that the licensee's acceptance. criteria in their surveillance test for determining back-leakage rates for two of their DHR interfacing valves was nonconservative in that the acceptance criteria did not incorporate the effect of instrument inaccuracy.
Consequently, the team concluded that the failure to incorporate instrument inaccuracy into the test acceptance criteria could lead to. test results greater than the maximum leakage rate
~
allowed by the plant Technical Specifications."
Open Item 50-346/89-201-02; Section 4.2.2; Page-12
-I
" Leakage rates for the DHR sving check valves (CF-30 and CF-31) were calculated by measuring the-change in the core flood tank (CFT)~ level.. Using the maintenance test line and the RCS as the pressure medium,Hthe CPT-level increase was recorded off the computer in the control room. After the quantity of leakage was known, it was_ divided by the test duration of one-half hour to ob.tain the leakage rate. The inspection team requested that the licensee provide information regarding the accuracy of the-instrument used in the test because the CPT level displayed in the plant computer implied that instrument accuracy Las one one hundredth of a foot (.01 foot).
Ilovever,. the team's review of the licensee's calculation showed that the' accuracy was plus-or minus.0336 feet. The instrument accuracy of.0336-feetfimplied-that'the flow rate calculation was only accurate to plus or-minus.568 gpm for the test-duration.
Therefore, the team concluded that, in the event the licensee obtained a test result greater than.432 gpm, the actual value_could potentially be above the surveillance acceptance limit of 1 gpm.
Similarly, the team noted that the inaccuracy associated with this test could lead to test results greater than 5 gpm.
Five gpm is the maximum rate allowed-by the Technical Specifications (TS) before the licensee must enter a limiting condition for operation, (LCO) as specified by the NRC Order of April 20, 1981 (Event V Order)."
I
Dockht Nu;bar 50-346
- I
.Licinsa Numb 2r NPF "
-Serial Number 1793 Attachment-Page 8 Toledo Edison Response Leakage greater than 0.422 gpm'across CF-30 or CF-31 could have resulted in exceeding the Technical Specification alert limit of 1 gpm when accounting for instrument error.
However, a~ review of-past test results indicated that this 1 gpm alert limit when accounting for instrument error was not exceeded.
Therefore, Toledo Edison vill revise the RCS Isolation Check Valve Leak ~ Test procedure to incorpcrate instrument _ inaccuracy. This procedure change. vill be completed prior to the conduct or this test at-the end of:the 6RFO.
i' Valve Classification NRC Finding; Executive Summary; Page ES-2 "The team's review of the PIVs found that there vere two additional valves which should have been classified as PIVs, but vere not.
Also it was not clear whether two additional valves referenced in a note in the licensee's response to Generic Letter 87-06 were being considered as PIVs by the licensee."
Open Item 50-346/89-201-01; Section 4.2~.1; Page 12 "The team reviewed the licensee's response to Generic Letter (GL) 87-06 and the piping and instrumentation diagrams (P& ids) for RCS, DHR, HPI and makeup and purification systems to ensure that the licensee had identified two valves in each system as PIVs at all high-pressure to low-pressure interfaces. The valves which were identified by the licensee as PIVs are listed in Table 4.2.1.
Because PIVs should be individually tested, the team considered the back-to-back check valves in the HPI' injection lines, which were velded-together, to be only one PIV. Therefore, the team considered the
-high-pressure to low-prescure boundary to be. located further up the HPI injection piping. As a result, the team considered that;the HPI discharge / recirculation cross-connect valves, which isolated the HPI recirculation line from the borated water storage tank (BUST), and the HPI pump discharge check valves, to be PIVs.
In addition, the' team's review of' the licensee's response to GL 87-06 found the two HPI discharge check valves were not included with the other PIVs.
Although these valves were tested to determine their back-leakage and vere included in a note at the and of Table a
II in GL 87-06, since these valves were considered to be PIVs by-both the team i
and the licensee, the team concluded that they should be placed in the table and removed from the note."
{
Toledo Edison Response:
l Toledo Edison was unable to identify any regulatory requirement or guidance j
identifying that to be considered a PIV a valve must be individually leak tested. Pressure Isolation Valve is defined in Generic Letter 87-06 as those l
.)
I 1
l
Docket Number 50-346' License Number NPF-3
. Serial Number 1793 l
Attachment Pcge 9-j
. valves in "each interface coolar* ?ressure boundary which separates the high pressure reactor coolant system i (CS) from an attached Ivy pressure system.
These valves are normally closed during power operations."
In' Toledo. Edison's response to Generic Letter 87-06, the two valves that sepacate'the high-pressure RCS from an attached low pressure system (back-to-back check valves)L are listed as being the PIVs. Generic Letter 8?-06 alsa stated that " periodic tests or verifications of the 'eak. tight integrity of all pressure isolation valves is necessary in order asure the integrity of the reactor coolant a
+
pressure boundary in accordant with 10CFR50, Appendix A and-B."
Since check
-valves HP-22 and HP-23 are tested to ensure the integrity of the back-to-back check valves, Toledo Edison does not believe that it should be a requirement to include HP-22 and HP-23 on the list of PIVs.
However, Toledo Edison' vill re-review these two valves to confirm this conclusion by June 30,.1990.
- 1 Valve Testing-i NRC Finding; Executive Summary; Page ES-2 l
"The team found that none of the motor operated and manually-operated gate and globe valves which were classified.as PIVs were leak tested to verify.their
_1 pressure isolation capability. These types of valves constituted j
approximately 60 percent of all PIVs at Davis-Besse.
l Toledo Edison Response:
.i Generic Letter 89-04 on the ASME Section XI' Pump and Valve Program clearly stated that the only PIVs required to be leak tested'are those specified in the Technical Specifications.. For Davis-Besse this applies only to valves CF-30, CF-31, DH-76 and DH-77.
Generic Letter 87-06' stated that'" Periodic I
tests or verifications of the leak tight. integrity of all pressure isolation i
valves are necessary in order to assure the integrity of the reactor coolant j
pressure boundary in accordance with 10CFR50, Appendix A and B."
At Davis-Besse, other Pressure Isolation check valves are leak tested to ensure leak tight integrity since check valves have more of a generic leakage problem in the industry than gate or globe valves. These are. tested as an augmented part of our ASME Valve Test Program.
There are 32 motor operated and manually operated gate and globe valves that are considered PIVs at Davis-Besse.. Twenty-two-of the gate valves drain to a tank with overpressure protection.
These tanks either have high' level and pressure alarms in the control room, at means to relieve pressure and maintain level. Three PIVs are tested per the Appendix J program.
Seven other PIVs are located in one inch lines that vill be evaluated in the future for possible testing. Toledo Edison considers on-line monitoring of tank levels a more meaningful identification of leakage than testing valves unda a test program every two years.
i 1
_.. _.. _ _