ML20129D523
| ML20129D523 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 06/22/1985 |
| From: | Fred Miller, Stalter L, Yarger K TOLEDO EDISON CO. |
| To: | |
| References | |
| NUDOCS 8507300076 | |
| Download: ML20129D523 (13) | |
Text
'
ACTION PLAN # 5.6&7 TITLE: LOW STEAM CENERATOR LEVEL TRIP OF SFRCS
' ~
V W o #ym'""
' W' b
CHAIRMAN REV DATE REASON FOR REVISION FORCE h,'$jffter O 6/22/85 Initial Issue K. Yarger [
8507300076 050622 ADOCK0500f"y6 gDR
TITLE: Low Steam Generator Level Trip of SFRCS REPORT BY: L. Stalter, F. Miller, K. Yarger PLAN N0.: 5,6 & 7 DATE PREPARED: June 22, 1985 PAGE 1 of 4 INTRODUCTION:
This report has been prepared in accordance with the " Guidelines to Follow When Troubleshooting for Performing Investigative Actions into the Root Causes Surrounding the June 9, 1985 Reactor Trip", Rev. 4 On June 9, 1985 a low Steam Generator (SG) level full trip of the Steam and Feedwater Rupture Control System (SFRCS) occurred when the main
/Df" M /)'
turbine tripped. This information was observed from the alarm log of the f. 8%
event. Additional SG low level half trips have recently occurred on June o 2, 1985 and on April 24, 1985. At the time of these trips, the time- evr y
averaged SG levels were at least 70 inches above the low level trip setpoint. No previous spurious low SG level trips have occurred at Davis-Besse (DB) prior to the 1984 Refueling Outage.
SUMMARY
OF DATA A detailed review of all of the available data from the June 9, 1985 event indicated that a full SFRCS trip occurred at 1:35:31 at the same time the main turbine stop valves closed. This full SFRCS trip occurred for a very short duration and it resulted in only the two Main Steam Isolation Valves (MSIVs) closing. At 1:35:31 the computer recorded a Channel 2 SG level trip. This alarm could have been caused by a full or a half trip of the SG level strings. No positive data is available to indicate which SG caused this alarm. Alarms were also recorded on the computer at 1:35:31 that indicated that SFRCS had sent trip signals to the main turbine and to the Anticipatory Reactor Trip System (ARTS) to trip the reactor. However, both the reactor and main turbine had been previously tripped after Reactor Protection System (RPS) actuation.
Review of past experience indicates the following:
- 1. Prior to the 1984 refueling outage, many turbine trips occurred that did not cause spurious SFRCS trips.
- 2. On April 24, 1985, a half trip of Channel 2 low SG level occurred when the main turbine tripped. No full trip was recorded during this event.
p 743
- 3. On June 2, 1985, a half trip of Channel 2 low SG level also occurred //
when the main turbine tripped. An SFRCS full trip slarm also was recorded on the computer at this time. No devices .ere actuated by this event and the SFRCS did not attempt to trip the main turbine.
Subsequent to the June 9, 1985 trip transient, it was determined from '
discussion with B&W that earlier testing at TMI-1 shows that SG startup range level transmitter output oscillates widely for about 10 seconds
after a main turbine trip from 100%. This information was taken using /
high speed recorders, and shows oscillations of 1.2 - 1.3 cycles /second. du No similar data exist for DB, but the SGs are essentially identical and w the same phenomenon could be expected to occur.
MAINTENANCE AND SURVEILLANCE / TESTING HISTORY The following is a listing of maintenance and surveillance test activities performed since the 1984 refueling outage:
- 1. Surveillance Test ST 5031.14 was performed on all channels for each input parameter once each month. No abnormalities were found, no drifting of components was noted since the 1984 refueling outage.
- 2. The apparent spurious trip of the SCLIC Channel 2 was checked out by WO 1-85-1444-00 dated 4/24/85. This WO called for running the
//
o, monthly surveillance test while checking for anomolies. No anomolies g were found.
1L
- 3. A relay driver board was replaced in the SFRCS logic Channel 1 for relay K201A (WO 1-85-1552-00 dated 5/4/85) . g/
- 4. A power supply was replaced in SFRCS logic Channel 2 (WO 1-85-1843-00 dated 5/27/85).
- 5. The status light socket for K602A relay in the SFRCS logic channel was replaced (WO 1-85-1860-00 dated 5/29/85). 7b.4, f
- 6. Troubleshooting was performed on SFRCS alarm logic Q963 (WO 1-85-1877-00 dated 6/2/85). This WO called for testing the alarm logic to determine why a full SFRCS alarm occurred on a half trip. In the M g process of checking, a connection was reterminated and the problem (M, leared. Testing on both SFRCS channel 2 and channel 4 was completed .
g esl,, per section 6.4 of ST 5031.14. No drifting of setpoints or other
. g#M anomalies were found.
A review of the maintenance and surveillance / testing activities indicates that they did not affect the functional operation of the SFRCS. M CHANGE ANALYSIS: % 2x The following changes were completed during or after the 1984 refueling outage.
- 1. All SFRCS/s':eam generator level transmitter amplifier and calibration boards were replaced to continue to meet the environmental qualifi-cation rnquirements (WO 1-84-1269-00).
All trans & ters were calibrated and response time tested by surveil-lance tests RT-5031.16). All acceptance criteria were met.
- 2. The level transmitters providing startup level indication on the control panel connected to the same sensing line as the SFRCS level
%gf' sensing tre.witter were changed from BY to Rosemount transmitters.
/
/ %c
p 3. The Sceam Generator Level Indication and Control (SGLIC) circuits g/
were modified by FCR 80-110 which installed new analog - bistable units to add the steam generator high level trip capability. This required that the original bistables used for low level trip also be /p A replaced to gain cabinet space. These units are now a composite unit -
in that the analog and bistable functions are contained in one unit as compared to separate analog and bistable units previously used for fc the low level trip function.
All modules were calibrated and response time tested by surveillance tests (ST-5031.15) and all acceptance criteria was met. 1
- 4. The SFRCS actuation channels 1 and 2 were modified to provide for W
g,-c.,
blocking of the steam generator high level trip by FCR 80-110. This g change was tested by surveillance test (ST-5031.14 Section 6.1) and Section 6.14. This verified that the installation was correct and that the level set points for both the high trip and the low trip were set properly.
The SFRCS actuation Channels No. 1 & 2 were modified by FCR 81-178 to provide for the additional opening of steam supply valves (MS106A & HN
[ MS107A) to the auxiliary feedwater pump turbines on steam generator high or low level, high steam generator feedwater differential 4 Y MY
.f pressure and loss of all four reactor coolant pumps. k QM 7 6 The addition of the valve actuation logic by FCR 81-178 was defeated by FCR 81-178 Rev. A as it may have caused pipe hanger damage to f
TV M ( occur in the steam lines to the auxiliary feedwater pump turbines.
j {( FCR 81-178A was tested by TP 580.00 and all acceptance criteria were grA met.
- 7. The SFRCS actuation channels No. 1 & 2 were modified to allow opera-tion of the steam generator blowdown valves HV 603 and HV 611 for modes 4, 5, and 6 operation. This change was completed and tested (TP 520.81 and ST 5030.18) during the 1984 refueling outage.
The only changes listed above that could affect the response time of the *st% ~
steam generator level instrument strings were changes 1, 2 and 3. None of the others had any effect on the response time of the SFRCS.
HYPOTHESES J-r4 The cause of the half trips of SFRCS appears to be the SG startup range k (fy y level transmitter output oscillations due to pressure oscillations from +J the main turbine trip. The modifications that were performed in the SG ' fj level instrument string used by SFRCS during the last refueling outage may have made the string's response time faster so that steam generator 8 FM ,
pressure oscillations are now detectable. This may account for the spurious SG level trips that have occurred since the 1984 refueling outage.
It is postulated that on June 9, 1965 </2 SG low level trips occurred l within each half of an SFRCS actuation channel with the following results. -
4 I
- 1. They overlapped long enough to allow the following alarms to occur:
- a. Q963 SFRCS FULL TRIP; this alarm requires that both halves of an SFRCS be tripped simultaneously,
- b. Q777 ARTS TRIP (Anticipatory Reactor Trip System); this alarm indicates that the ARTS was tripped. Since no other ARTS input alarms occurred, the only way that this alarm could have actuat-ed would be due to an SFRCS trip of ARTS.
- c. X044 T-G MN STM (Turbine-Generator Main Steam) and FW TURB TRIP (Feedwater Turbine Trip); this trip indicates that the SFRCS sent a trip signal to the main turbine. However, the main turbine had already been tripped by the reactor trip.
- 2. Each half trip occurred long enough to allow the electrically operat-ed solenoid valves and the air pilot valves in the MSIV air system to trip. These half trips overlapped long enough to allow the " seal-in" circuits to drop out, and thus, the MSIVs stayed closed after the half trips automatically reset.
- 3. These half trips did not overlap long enough to pick up and scal-in the starters on any SFRCS actuated motor operated valves (MOV). This would explain why we did not start the auxiliary feedwater pumps during this spurious full trip of SFRCS. In the past, when we were experiencing spurious high reverse delta pressure main feedwater trips, on some occasions only the MSIVs closed and MOVs did not operate.
4 Each half trip occurred long enough to allow the electrically operat- N ed solenoid valves in the Main Feedwater Startup Control Valve (SP7A, L1f ghs gg ,
SP7B) air system to trip. However, these half trips did not overlap long enough to allow the " seal-in" circuits to drop out, and there-fore, the Main Feedwater Startup Control Valves remained open.
Special tests are needed to determine the trip and trip reset response times for the components in the MSIV and the Main Feedwater Startup Control Valve control circuits.
Other less likely hypotheses might be possible. The action plan has been designed to preserve the equipment status should the above hypothesis prove false.
ap 5 & 7-1 O
O O O ..
ACTION Pt.AN PLAN NUMBEH PAGE *
'" ___.._ Rey,_0_ 5&7 I a' 4 TITLE ..DATE PREPARE D PREPARED BY Low Steam Generator Level Trip of SFRCS 6/20/85 L.C. Stalter SPE CIF eC OH sE CTIVE
\/ /
STEP
^" ' /
X \ PRIME ASSIGNED START TAHGET DATE
\
NUM8ER / ArSPONSIBILITY TO DATE DATE COMPLETED All steps of this Action Plan are to be performed in accordance with the latest revision of " Guidelines to Follow When Trouble- - - - . . . _ - . - .
shooting or Performing Investigative Actions into the Root Causes Surrounding the June 9, 1985 Reactor Trip".
The following steps may be performed in any sequence. Substeps of any step should be performed in the order listed.
i
~1 1 Review available data on the effects on the steam generator L. Stalter l
~
startup level from a turbine trip. Use available information from previous transient at Davis-Besse, and data from TMI-l supplied by B&W for this review. l A.' Use the data from 1 to verify the hypothesis. F. Miller i
l B.- Use data from 1 and work with Bechtel and CCC (SFRCS L. Stalter vendors) to establish an appropriate response time for the U___
P O ^
ACTN3N PLAN ' PLAN NUMBE M PAGE
- e m ee Rev,_0__ 5&7 2 a' - 4 TITLE DATE PREPARED PREPARED SY Low Steam Generator Level Trip of SFRCS 6/20/85 L.C. Stalter SPE CIFSC OBJECTIVE
/ T PRIME ASSIGNED START TARGET DATE ACTION STEPS ER ,
RFSPONSI8tLITY TO DATE DATE COMPLETED -
Steam Generator Level Indicator and Control (SGLIC) bistables to prevent inadvertant actuation of SFRCS following a turbine __ _ ,
trip, and still meet the required overall response time for SFRCS actuation on staan generator level.
2 Determine why the Main Steam Isolation Valve (MSIV) closed and F. Miller other SFRCS actuated equipment.did not change status.
A. Perform tests to determine the specific component trip K. Yarger l response times and seal in response times of MSIV and Main Feedwater Startup Control Valves.
B. Analyze the data from step 2A to verify that only the MSIV F. Miller would close upon a full SFRCS trip of short time duration, and other equipment would not actuate.
an5A7 L--
C' O 7 ACTION PLAN PLAN NUMBE H PAGE
- oo e Rev. 0 5&7 3d 4 TITLE DATE PREPARED PREPARED By Low Steam Generator Level Trip of SFRCS 6/20/85 L.C. Stalter sPeciric oaxcTivE p
/'
STEP PRIME ASSIGNED START TARGET DATE NUMsER k HFSPONSIBILITY TO s
DATE DATE COMPLET ED 3 Test the time delay relay (TED shown on drawing E-428, sheet 54) K. Yarger ,
in the SFRCS " full trip" reset logic to verify the total time the SFRCS full trip was in offect.
A. Determine the length of time the full SFRCS trip was in the F. Miller tripped condition by combining this time with the sequence of events data B. Analyze this data for verification of the hypothesis. F. Miller 4 A. Visually inspect each steam generator starting level K. Yarger transmitter for loose connections, cleanliness.
B. Verify the response time of each Steam Generator startup K. Yarger level transmitter.
an 5 & 7
r
- a t
D E
4 l T vo ct EE TL AP
- DS E
DM O
4 R . C A
E P C.
G E L T 3 A P
R P
D T _
R E E R B A 5 P
M E 8 U 7 R/ T E N P 0 RT _
N & E 2 TA A A 5 T / S D L
P A 6 D
D E
. S v S A
e R
V T r -
u r e Ee s
e g
t -
Mi l I s r a Rm a t Po Y S e
s . .
r R K L
/ p O k1 u
E V
I T
C
\
X t r
a t
s -
A r S
O o
t C a e r .
F e s I
C n e E
P e
g i
l S o m_ m S a o C e n R t a F s_
S s r P h o f E c f _
o T s
a p e k n c i
o f e r s o h T i c c A
n l o d e i n v t a e a .
L r r a b e t r i t a o l t d t a i a c m s r s i e e n h n h a t e t r G y t e z
m f l y a i e l e r v a C' $
f t
S V e
l e
A n -.
r A e 7 L e . .
P L C D &
s e R 5 O e. E PE I
T. L Ees p C s. T 4
Te Su a Ae T m
O ^
- ACTION PLAN PL AN NuaAtst H PAGE
- alcas Rev. 0 5, 6 & 7 1 a' 4 T4TLE DATE PHEPARED PREPARED BY Low Steam Cenerator Level Trip of SFRCS _ 6/20/85 L.C. Stalter spa caic oexcTivE STEP PHIME AS$tGNED START TARGET DATE NUMBE R
^ '
HF SPONSIBILITY TO DATE DATE COMPL ET ED All steps of this Action Plan are to be performed in accordance with the latest revision of " Guidelines to Follow When Tr_oubl_e __
, shooting or Performing Investigative Actions into the Root l
Causes Surrounding the June 9,1985 Reactor Trip".
The following steps may be performed in any sequence. Substeps of any step should be performed in the order listed.
1 Review available data on the effects on the steam generator L. Stalter j g" - ,
startup level from a turbine trip. Use available information duo,T sfrom previous transient at Davis-Besse, and data from TM1-1 I
supplied by B&W for this review.
\
. L2a(nw A. Use the data from 1 to verify the hypothesis. F. Miller B. Use data from I and work with Bechtel and CCC (SFRCS L. Stalter vendors) to establish an appropriate response time for the
ACTION PLAN PL AN NUUtst e4 PAGE s o .**
- Rev 0 5, 6 & 7 2 a' 4 TITLE DATE PREPARED PREPARED SY Low Steam Generator Level Trip of SFRCS 6/20/85 L.C. Stalter sPEceric oencTivt STEP PRIME ASSIGNE D START TARGET DATE NUM8ER R E SPONSl88 LIT Y TO DATE DATE COMPL ET E D Steam Cenerator Level Indicator and Control (SCLIC) bistables to prevent inadvertant actuation of SFRCS following a turbine trip, and still meet the required overall response time for SFRCS actuatics on steam generator level. .
j'-.
2 Determine why the Main Steam Isolation Valve (MSIV) closed and F. Miller other SFRCS actuated equipment did not change status.
A. Perform tests to determine the specific component trip __K . Yarger____
response times and seal in response times of MSIV and Main Feedvater Startup Control Valves.
B. Analyze the data from step 2A to verify thtt only the MSIV F. Miller .
would close upon a full SFRCS trip of short time duration, and other equipment would not actuate. f[ [/g
/d I_b Y Wl 4 'J4d Sb>d .
u 4 A & ] d d h w!< w$ f r QA nmm c> 5 & 7
- k,j8" u _
o -
ETm PLM PS N E EH P AG E.
ee s.ee Rev. 0 14TLE 5.6&7 3 a' 4 DATE PHEPARED PHEPARED BY Low Steam Generator Level Trip of SFRCS 6/20/85 L.C. Stalter sescuec osacTevE STEP PRIME AS$1GNED START TARGET DATE seuessER, ACTm STEM
, RFSPON$talLITY TO DATE DATE COMPLETED 3 Test the time delay relay (TED shown on drawing E-423. sheet 54) K. Yarger
)intheSFRCS"fulltrip"resetlogictoverifythetotaltime _ _ _ _ _. _
the SFRCS full trip was in effect.
o v -
A. Determine the length of time the full SFRCS trip was in the F. Miller tripped condition by combining this time with the sequence of events data.
B. Analyze this data for verification of the hypothesis. F. Miller . _ .
4 A. Visually inspect each steam generator starting level K. Yarger transmitter for loose connections, cleanliness.
B. Verify the response time of each Steam Generator startup K. Yarger l l
level transmitter. j i,
ca5a7 l
"T .*
I
~
ACTION PLAN
- L a r4 Nuunt u Pact
. _ - _ _ _ _ _ __ _ _ _ _ _. _ ._ ..__ _ ._____ ____ Rev._0_ _ _ 5,6&7 4 a' 4 TITLE DATE PREPAREO PREPARED Sy Low Steam Generator Level Trip of SFRCS __ _ _ 6/20/85 L.C. Stalter sPacietc osacreve STEP PRIME AhSIGNE D START TARGET DATE meutsSEH.f
- 3 HFSPONSIBILITY TO DATE DATE COMPL E T E D C. Verify the calibration of each steam generator startup K. Yarger ,
level transmitter. .__- _ _.. .. --- - - . - . .. - - _-_ ..- - _ __ __. _ _ _ _
D. Analyze this data and check for anomolies. L. Stalter www m l _ __ _
Pe. .meewh % %
rw- -* mee-Mn e-1.-NN--4Lem==
w-%-we -pw e-summumumn ww.1w-
_,_._m_---_-------- - - - - - ~~4- .<w ao5&7
~