ML20235E843
Text
_ _ _ _ _ _ _ _
e 5
Reg tory m Cy.
1 NSF NORTHERH STATE 5 POWER COMPANY M t N N E A PCLI S. M I N N E SOTA 55401 9
9 E
g 4
February 18, 1972 e
"M
\\p W[//Il/t )
j d:
FEB 221972* s d 83 s
Dr. Peter A Morris, Director REGULATORT fh,,[hgb T
' #'N:
Division of Beactor Licensing gQ gy*,. M United States Atomic Energy Commissi g
Washington, D C 20 % 5 g
co Or
Dear Dr. Morris:
MONTICELLO NUCLEAR GENE ATING PIANT Docket No. 50-263 License No. DPR-22 Follow-up Report on Main Steam Isolation Valve Pmblem This report is submitted as a follow-up mport to the M3IV items described in the " Report on Main Steam Isolation Valve Problems," dated November 26, 1971.
Subsequent to November 26, all of the Main Steam Line Isolation Valves were itak tested in accordance with approved procedures and maintenance was per-fomed or the valves where necessary. The following is a summary of the leak test results prior to performing maintenance on the valves:
A0 2-80A 0.2 SCFH A02-86A
__70 gpm H O 2
A0 2-80B 0.6 SCFH A02-86B 150.0 SCFH A0 2-80C 1.1 SCFH A02-86C 6.7 SCFH A0 2-80D 153 9 SCFH A02-86D 83 1 SCFH It is noted that four of the eight MSIV's were leaking in excess of the 115 SCFH Technical Specification leakage limit.
Maintenance work on the valves extended from November 23 until January 16.
Dur-ing the repair period, it was necessary to disassemble two of the four MSIV's three times and one of the four MSIV's twice to reduce the leakage rates to with-in the allowable limit. Problems of interest encountered during the repair per-iod included:
a.
Main Poppet Guide Wear As note'd in the November 26 report, an inspection of A0 2-864_ performed on November 23 revealed that one of the three poppet guides showed ex-l cessive wear on the lower 1 inch of guide material.
There was also in-l dication on the side cf the main poppet of wear between this guide and l
the poppet. Wear on the guide and poppet was caused either by foreign l
8709280293 870921 S 4dE 2 5 % }9 #
NENZ87-111 PDR O2.3
1 NOFh HERN STATES POWER CbAPANY Dr. P A Morris February 18, 1972 matter whi:t became wedged between the guide and the poppet or by for-eign matter which became lodged under the seat during flushing opera-tions alleving the poppet to vibrate freely back and forth 86ainst the guide ribs. The wear area of the poppet guide was overlayed with stel-lite. The ;cppet was rotated prior to reassembly to provide a new mat-ing surface between the poppet and the guides.
This or similar effects were not p:ssent on any other valves which were inspected.
b.
Spring Colls: Tack Weld Failures Each MSIV his a pilot spring seat collar which is screwed to the stem.
Each collar is tack velded to the stem to impede rotation.
One of the MSIV's which was disassembled had two pilot spring collar tack velds of which one vss found to be cracked.
The other three NSIV's which were disassembled had only one pilot spring collar tack weld and each was cracked.
C: A0 2-86B the collar had unscrewed 13/32 inch. An investi-gation of -lis problem revealed that the pilot spring collar could not completely iisengage from the stem because of the physical assembly of the MSIV.
It was concluded that the tack weld failure problem could not prevent an MSIV from closing, however, if a high differential pres-sure existed across the EIV, the MSIV could possibly not be opened if the tack we'.d had failed and the pilot spring collar had unscrewed to j
the point a which the pilot valve (which equalizes the pressure) would t
not open.
he Monticello Operations Committee and Safety Audit Commit-f tee concluded that it was not a safety problem if an EIV could not be opened.
The origina~ tack welds were made using 410 stainless steel which is a hardenable :sterial.
This material loses its ductility when hardened.
Since the c:'lar and stem cannot move after the tack veld is applied, the tack veli vould have a tendancy to build up internal stresses as it cools which =ay have resulted in a cracking of the velds.
'Ibe only me-chanical losiing applied to the veld would be that of a slight turning moment causei by the spring deflection.
The tack velis on the four EIV's were replaced with longer velds made with 309 stainless steel which is an austenitic material. An austenitic material vill retain its ductility when heat treated and therefore should not crack as readily as the 410 stainless steel previously used for the tack welds.
l Pilot Valve Scratches During Reassembly c.
As noted pre-iously, it was necessary to disassemble three of the MSIV's more than one time to correct the leakage problem.
One of the reasons j
for this was that the pilot valve was being scratched during the reas-sembly of the valve.
It was found that upper edges of the pilot valve guides were very sharp and unless extreme caution is taken during the reassembly, -le pilot valve would get scratched.
The upper edges of the guides nre chamferred ani the assembly procedures were modified to minimize the possibility of scratching the pilot valve.
I
J. -
- " ' ~
NOA. e4ERN STATES POWER CL APANY '
l Dr. P A Wrris
' l+ -
February 18, 1972 rates leads us. o believe that the Main Steam Isolation Valves can reliably perfom their fcetion.
It is our intent to" retest the MSIV's during the next scheduled ;1 ant outage.
The November 26 report also. discussed the slow closure of A0 2-80B.
It was noted that A0 2-30B would close slowly only if the valve was opened and im-
' mediately reclosed.' It is believed that this characteristic may have been-caused'by a slight contraction of the spool' sleeve due to a cooling effect of the air as i passes through the sleeve when the MSIV is opened. This contraction wou i prevent the spool piece from responding rapidly if the MSIV is immediately given a' closed signal after it has been opened.
A Significant 0;erating Event report has been written for this occurrence and will be available to' the Begion III Compliance Inspector for review during his next visit.
1
. Yours very truly, i
h*
\\
L 0 hyer, P.. -
Director-Nuclear Support Services LOM/DDA/bjr I
cc: B H Grier.
4 i
o Fue Cy.
y ry NSF
~
4 i
NOMTHERN STATES POWER COMPANY K)EICELLO NUCLEAR GENERATI fG PLAW
(
Monticello, Minnesota 55362
)
'6ili 6T j
4+\\
e i
m y
l S
A
A
{
February 21, 1972 k'..
{
l, N
DocmEO y
..,, g-i Uwc
=
Dr. Peter A. Morris, Director 1[
FEB 2019n e, 5,,,)
{ $
M Ib,:
i v
u Division of Reactor Licensing REGULATORY d'/
United States Atomic Energy Commissio 4 gEcDONDh' h
g j
Washington, D. C.
20545 cLtax e
g Lpg, 4
i a
j
Dear Dr. Morris:
co UDEICELLO f0 CLEAR GENERATifG PLAN j
Doeket No. 50263 License No. DPR-22 ADDITIONAL INFORMATION ON 1f6TRUMEEATl0N Pf0BLEMS This letter is to inform you of additional information obtained concerning i
several plant instrumentation problems which were the subject of previously J
submitted reports.
1.
Main Steam line Hioh Flow I instrumentation Problems (See report to Dr. Morris dated November 23, 1971)
Sixteen switches with a range of 0-70 psid were installed in the main steam line high flow isolation system on September 1,1971, to replace the original 0-200 psid switches. The lower range switches were installed because of lower than expected differential pressure measure-f ments from the main steam line flow nozzles.
During surveillance 1
testing on November 11 and 12,1971, three of the switches were found to exhibit signs'of high friction or binding in their mechanisms and another switch was found to exhibit non-repeatable trip settings. T he four defective swi tches were immediately replaced by 0-200 psid j
switches previously used for the protective function. The remaining j
twelve switches were replaced on November 13, 1971.'
One of the defective switches was returned to the factory for inspection.
The vendor's inspection revealed that the torque tube jewel bearing was contaminated by a lead paste compound that was used to seal the joint between the instrument case and the di fferential pressure uni t.
l The contamination which occurred during switch assembly af the factory, caused excessive hysteresis in the torque tube movement. The vendor indicated that the jewel bearing contamination was a random occurrence in switches of this type and model, and therefore there was no assurance I
that other similar instruments did not have the potential for similar failures.
A review of the instrumentation used in the plant protection f
system was conducted to identify all instruments with the potential for 1
t cs a n. m +
- r-p g,
3 1 -l 1 g <.g j
s a
h 3
2-a contaminated -jewel bearing.
A total. of 39 instruments.were so :
identi fied and inspected following _ the vendor's re:ommended proce-du re.
No jewel bearing contamination was found.
The switch which was' found to exhibit' a non-repeatable trip setting-was_' investigated at the site. Tests disclosed that the erratic operation was caused by'a loose pivot pin on the cam follower; assembly and was not due to the snap action switch, as previously reported.
The loose pivot pin. caused the trip point to vary between 109% steam 1
flow and 127% steam flow. This problem also appears to have resulted from a manufacturing error. All plant instruments'of this t inspected for excessive deadband (greater that 5% of range) ype were.
and were found to be satisfactory.-
2.
Failure of ECCS Pumo Start Permissive Switch and Relav, (See report.
to Dr. Morris dated December 20, 1971)
~
On November 20,1971, while performing a surveillance test, the trip setting of ECCS Pump Start Permissive Switch PS 2-3-53 A, switch #2, was found to be 12 psi-below the required setting of _>. 450 psig.
During this surveillance test it was also discovered that a relay in the ECCS pump start permissive logic failed to er.ergize when pump start permissive switch PS 2-3-53 A, switch #1 was closed.
An investigation of the setpoint problem revealed that the two switches contained in PS 2-3-53 A were not designed for the 125 V DC application for which they were being used, but were designed for 115-j V AC service. The wrong switches were initially specified for this application.
An analysis performed by the instrument manufacturer revealed-that the microswitch contacts.in the instrument were burned due to excessive current. The vendor believes the burned contacts may have caused the change in the instrument setpoint.
A check of plant instrumentation disclosed that 22 instruments with microswitches rated for 115 V AC were being used in 125 V DC ci rcui ts.
.These switches (listed below) have all been replaced with switches rated for 125 V DC.
I ns t rument Function PS 2-3-49 A & B LPCI Recire Loop Selection PS 2-3-50 A & B LPCI Reci rc Loop Selection PS 2-3-52 A ECCS Valve Open Permissive PS 2-3-53 A & B ECCS Pump Start Permissive PS 23-68 A through D HPCI Steamlire Low Pressure isolation PS 2-128 A & B RHR Shutdown Cooling isolation PS 14-47 A & B Core Spray Header High Pressure Alarm PS 14-44 A through D Core Spray "AC Interlock" PS 13-78 RCIC Turbine Exhaust Diaphram High Pressure Alarm
=*
r' Funct ion I ns t rument PS 13-72 A & B RCIC Turbine High Pressure Alarm
' The investigation of the relay which failed to erergize when the switch #1 of PS 2-3-53 A was closed, revealed tha: the pull-in. voltage was set too high. ;This resulted in an intermitte-t' failure of the,
relay to energize. - The pull-in voltage of the relay was found to be 110 V DC (the operating voltage available at the relay was measured to be-125 V.- DC).- The pull-in voltage was reset to 80% cf.the operating-voltage and the relay ai,r gap was reduced. All of the DC relays of this type.in the plant had been previously set and tested'at.80% pull-in
- voltage in accordance with a field engineering me o in August 1970.
Additional investigation revealed that the 80% se-ting. is applicable only to AC relays. All DC relays of this -type were reset-to the factory recommended 60% pull-in voltage. The relay which failed to ' operate during the surveillance test was the or.ly relay of this type to experience a failure since.the DC relays were set and tested in August 1970.-
3.
Failure of ECCS Valve Ooenino Permissive Switch (See report to Dr.
Morris dated December 21, 1971.)
On November 24, 1971, while performing a surveillance test, the trip setting of ECCS Valve Opening Permissive Switch PS 2-3-53 B, switch
- 1 was found to be 42.5 psi below the requi red setting of.>_. 450 7
psig.
An investigation revealed that the instrument lacked a setpoint
" locking" device, a modification recommended by. the instrument manu-facturer. The locking device was. immediately installed and the instrument was calibrated to trip at'460 psig.
Locking devices were also installed on all'similar instruments with snap action switches used in the plant protection systems.
The main steamline' high flow instruments have mercury type switches and are not designed for setpoint locking devices. Since past-experience has shown the setpoints of these mercury -switches to be quite stable, the switches have not been redesigned or modified to include a setpoint locking device.
Locking. devices were installed on the following instruments:
i l ns t rumen t Funetion 4
DPIS 23-76 A & B HPCI High Steam Flow Isolation l
DPIS 23-77 A & B DPIS 13-83 RCIC High Steam Flow Isolation DPIS 13-84 DPIS 14-43 A & B Core Spray Line Break Detection PS 2-52 B ECCS Valve Opening Permissive j
a r
$l
.r
I ns t rument Function DPIS 2-129 A through D LPCI Loop Se'ection DPIS 2-136 A & B
All the above instrutrentation problems were corrected prior to plant startup on January 24, 1972.
Detailed reports on these problems have been prepared at the plant and are available for -he Compliance Inspector.
Yours very truly,
[.
L. O. Mayer Director of Nuclear Support Services LOWLLWkik cc:
B H Grier 3
')
$a