ML20211D993
| ML20211D993 | |
| Person / Time | |
|---|---|
| Site: | Crystal River, Rancho Seco, 05000000 |
| Issue date: | 03/21/1986 |
| From: | Simmons G SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | |
| Shared Package | |
| ML19292G087 | List: |
| References | |
| 75, NUDOCS 8610220327 | |
| Download: ML20211D993 (160) | |
Text
T
(
_\\
et
(
TRANSIENT ASSESSMENT PROGRAM REPORT FOR RANCHO SECO TRIP LOSS OF ICS POWER ON DECEMBER 26, 1985 TAP NUMBER RS-86-01 8610220327 060926 I
PDR ADOCK 05000302 S
t 3
I TRIP REPORT #75 REACT 0P TRIP ON DECEMBER 26, 1985 LOSS OF ICS POWER Prepared By Grant Sirnmons 3/21/86 Revised -- 4/15/86 i
5
.b LOSS OF ICS POWER Trip Report #75 TABLE OF CONTENTS I.
Summary 1
.II.
Sequence of Events 2
III.
Pretrip Plant Status 15 IV.
Initiating Event and Reactor Trip 16 V.
Post Trip Transient Response 17 VI.
Key Operator Actions / Plant Procedures / Training 28 VII.
Investigation of Areas of Concern 39 VIII.
Root Causes 87 IX.
Corrective Actions 91 X.
Conclusions 93 XI.
Attachments / Graphs / Figures 95
a Lk ENCLOSURE 4.6 TRIP REPORT AMEN 0MENTS AND APPROVAL Prepared 8y Grant Simmons March 71. t o RA Engineering / Quality Control Oate PRC Amendments and Remarks:
PRC review of event on
, meeting number Minutes of the meeting (s) are attached PRC Chairman Date Plant Superintendent Remarks:
(
Approved Plant Superintendent Date END Rev. 8 AP.28-2f
a
~k LOSS OF ICS POWER TRIP REPORT NO. 75 I.
SUMMARY
On December 26, at 0414, the reactor tripped from 73% power on high RCS reessure due to a reduction in main feedwater flow to the steam generat' ors.
The reduction in flow was caused by a loss of Integrated Control System (ICS) power.
Following the trip an overcooling transient occurred and the Safety Features Actuation System (SFAS) was automatically initiated on low RCS pressure.
The unit was holding power at 75% when a total loss of ICS DC power occurred.
Feedpump speed decreased to minimum speed and feedwater control valves partially closed.
Atmospheric Dump Valves (ADVs), Turbine Bypass Valves (T8Vs), and AFW Flow Control Valves partially opened.
The reactor tripped on high RCS pressure.
The turbine and generator tripped from the reactor trip.
The Auxiliary Feedwater (AFW) pumps automatically started on low main feedpump dischargo pressure.
AFW flow to the steam generators started through the partially opened AFW flow control valves.
Operators were dispatched to manually close the ADVs, TBVs, and AFW control valves.
A rapid steam generator depressurization and RCS cooldown to 386*F was exparienced before operators isolated the valves and ICS power was restored.
RCS cooldown and depressurization led to SFAS initiation and emptying of the Pressurizer.
The steam generators were l cverfilled with feedwater.
The rapid cooldown with full HPI flow resulted in reactor vessel Pressurized Thermal Shock (PTS) operating limits being exceeded.
The primary system makeup pump's suction was inadvertently isolated during the transient.
The pump was severely damaged.
Makeup water spilled out of the pumps damaged seals and caused a minor offsite radiological release.
A Senior Reactor Operator (SRO) collapsed in the control room during the transient.
Following restoration of ICS power, the cooldown was halted.
The primary plant was depressurized below the Thermal Shock region in accordance with plant procedures.
RCS subcooling was adequate throughout the event.
The rapid cooldown was initiated by the opening of TBVs, ADVs,'and excessive AFW flow.
Following isolation of the ADVs and TBVs, the rapid cooldown is attributed to steam generator overfill with AFW, The makeup pump failure was the result of &
valving error committed when restoring from the SFAS system lineup.
Prior to restart, an extensive investigation was performad to determine the cause of the loss of ICS power and to identify and resolve the issues resulting from the transient.
Page 1 i
.h LOSS OF ICS POWER TRIP REPORT NO. 75 SEQUENCE OF EVENTS NEY TIMES 04:16:20 Clock time.
The IDADS computer time was used as the official clock.
All times have been normalized to the IDADS clock.
[07:23]
Time after initial event in minutes and seconds.
04:20:407 Time is estimated to the nearest second.
However, the associated event may have occurred a few seconds earlier or later.
04:16:7 Time is estimated to the nearest minute. The event has occurred sometime during this minute.
04: 587 This is the best estimate for the occurrence of this event.
It could have occurred a few minutes earlier or later.
SOURCE ABBREVIATIONS IDADS from the Interim Data Acquisition and Display System computer.
OPS from operator personal statement or control room logs.
FUI from operator follow up interviews.
CALC from an engineering calculation of pressures or flowrates.
SCRTY from security log.
ECL from Emergency Coordinator's log.
GACP from General Atomics Radiation Monitoring System Computer.
Page 2
e k
LOSS OF ICS POWER TRIP REPORT NO. 75 II.
CHRONOLOGICAL SEQUENCE OF EVENTS 12/26/95 INITIAL CONDITIONS Unit operating at steady state power of 75%, 710Pld(e)
Integrated Control System in full automatic Bailey Computer out of service (one of the Control Room's two main computer systems)
TRANSIENT INITIATOR TIME SOURCE EVENT / ACTION 04:13:47 IDADS Loss of ICS is caused by the simultaneous de-energizing of all redundant ICS DC power supplies.
All lights on the ICS stations go out.
All Bailey station demands go to 50%.
"lCS OR FAN POWER FAILURE" annunciator alarms.
SEQUENCE OF EVENTS 04: 13: 47 IDADS "ICS OR FAN POWER FAILURE" annunciator alarms.
[00:00]
Upon the loss of ICS DC power, all ICS demands went to midscale, corresponding to 0 volts.
The startup and main feedwater valves closed to 50% because of this decrease in demand signal.
The main feedwater pump speeds reduced to low speed stop, 2500 rpm.
De-energization of the ICS DC power supplies caused the main feedwater block valves to close.
Main feedwater flow to the steam generators decreased to zero.
The control rod drive system transferred to MANUAL upon loss of ICS power.
This prevented the rods from reducing power in response to decreasing feedwater flow.
The loss of ICS DC power also sent a c'emand to the Bailey AFW control valves, ADVs, and TBVs to open to 50%
demand.
The main turbine could not respond to changes in steem header pressure because signals to change governor valve pcsition originate in the ICS.
Total steam flow increased.
Steam generator pressures began to decrease.
The reduction in feedwater flow had a greater effect than the extra steam flow.
RCS pressure began to rise rapidly as the RCS heated up.
The main steam to auxiliary steam pressure reducing station also failed to 50% demand.
Aux'liary steam pressure began to increase.
Page 3
e A
LOSS OF ICS POWER TRIP REPORT NO. 75 04:13:557 OPS Operators notice MFW flow decreasing rapidly and RCS pressure increasing.
Operators manually open one of the pressurizer spray valves in an attempt to stop the RCS pressure increase.
04:14:01 IDADS Main feedpump discharge pressure decreases to a poirt which automatically starts the motor driven AFW pump, P-319.
The setpoint is 850 psig.
At this time, there was no main feed flow to the steam generators because steam pressure was higher than feedpump discharge pressure.
04:14:03 IDADS Reactor trip on high RCS pressure.
The turbine and generator
[00:16]
trips are also initiated by the reactor trip.
An operator closes the pressurizer spray valve.
OPS Immediately upon reactor trip, transformer yard fire alarm, seismic trouble alarm, and SFP high temperature alarm are received on the main annunciator panels.
A TSC fire system actuation alarm is received on the IDADS computer.
An auxiliary steam system relief valve lifts.
04:14:04 SCHRT Momentum of the RCS carries peak pressure to 2315 psig.
AFW IDADS flow begins to both OTSGs.
04:14:06 IDADS AFW dual drive pump, P-318, autostarts on low main feedpump discharge pressure (850 psig).
RCS hot leg temperature reaches a peak of 606.5 F.
04:14:077 OPS Operators perform the actions of Emergency Procedure Section E.01.
This included reducing letdown flow.
Operators then proceed with Emergency Procedures Section E.02.
04:14: 12 10 ADS Six OTSG code safety valves are lifting.
04: 14:25 IDADS Operators fully open "A" inject valve for more makeup addition to the RCS due to low pressurizer level in accordance with E.02, Vital System Status Verification.
04: 14:26 IDADS All OTSG code safeties have resented.
04: 14:48 OPS Makeup Tank level is decreasing rapidly due to high rate of makeup to RCS.
An operator opened 8WST suction valve on "A" side (SFV-25003).
04:15:04 IDADS Operators start "B" HPI pump to increase reactor coolant inventory from BWST I
04:15:18 10 ADS No level remains in OTSGs.
AFW is removing heat.
Page 4
o.
S LOSS OF ICS POWER TRIP REPORT NO. 75 04:15:30?
OPS Operators are sent to close AFW flow valves and place covers on MSR relief valves.
The MSRs go into a vacuum following a turbine trip. The relief valves have been a source of major vacuum leaks in the past. There were no condenser vacuum problems dur!.ng this event.
04:16:007 OPS Operators are sent to close TBVs and ADVs.
The TBVs and ADVs could have been shut from the Remote Shutdown Panel located two floors below the Control Room.
However, the operators failed to remember this fact.
Ranch Seco does not have Main Steam Isolation Valves (MSIVs).
04:16:02 IDADS An operator trips both main feedpumps.
Two operators verify 16:04 AFW flowrates to each OTSG are greater than 800 gpm.
Operators noted that both AFW pumps were running and that they had no control over main feedwater due to the loss of ICS.
04:16:14 IDADS OTSG levels begin to increase.
04:167 OPS Operator secures pegging steam from Control Room to ensure that it would not contribute to the cooldown.
A Control Room operator heard a steam relief valve blowing on the turbine deck.
Pegging steam had caused feedwater heater relief valves to lift in the past.
This had contributed to a recent overcooling event at Rancho Seco.
04: 16:40 IDADS RCS temperatures decrease below expected post trip value (550F.) RCS pressure is 1670 psig.
04: 16:57 IDADS RCS pressure has decreased to 1600 psig.
Pressurizer level is (03:20]
15 inches.
SFAS automatically initiates on low RCS pressure.
"A",
"B",
"C", "D" HPI injection valves travel to prothrottled position.
Selected SFAS equipment, including motor driven AFW pump P-319, trips and block loading of SFAS equipment begins.
AFW SFAS valves travel full open.
"A" and "B" DHR/LPI pumps autostart in their recirculation mode.
Diesel generators autostart but do not close onto vital buses.
There has beon no loss of power to the vital 4160 volt buses. SFAS also actuates containment building isolation.
The stripping of P-319 from its vital bus causes total AFW flow to decrease by half.
Page 5
LOSS OF ICS POWER TRIP REPORT NO. 75 04:16:59 IDADS "A" HPI pump autostarts from SFAS signal.
04:17:00 IDADS Pressurizer level goes offscale low.
Subcooling margin is 75'F and increasing.
04:177 OPS /
Operators unlock the ICS cabinets. They find all DC power FUI power supplies de-energized.
The dual switches (S1-S2), which provide power to the supplies, are both found in the full down position and are assumed to be closed. The automatic bus l
transfer device (ABT) which supplies ICS AC field power, is found powered from its normal supply.
An operator switches the ABT to its alternate supply, but ICS DC power does not return.
The A8T is switched back to its normal supply.
04: 17:107 OPS An operator closes AFW SFAS valves which were fully opened by the SFAS actuation.
These valves are in parallel with the already open ICS controlled AFW valves.
04: 17:15 IDA09 "A" and "B" CR/TSC Essential HVAC units start from the SFAS signal.
04: 17:27 IDADS Motor driven AFW pump P-319 is block loaded back onto its vital bus and immediately restarts.
The dual drive AFW pump has been running continuously since it started on low feedpump discharge pressure a few moments after the loss of ICS power.
AFW flow to both steam generators increases and is now approximately 1300 gpm to each steam generator.
04: 18:50 IDADS RCS temperature decreases below 500*F.
One RCP should have been stopped at this time to avoid core lift concerns.
An RCP was tripped at 04:28.
04:19:00 CALC Pressurizer surge line empties.
RCS pressure begins steeper (05:13]
decline.
Subcooling margin drops 8'F to 77'F and then begins to increase. Water at vessel head is flashing to steam.
i 04:19:15 IDADS Operators secure "A" train of CR/TSC Essential HVAC to reduce the ambient noise level in the Control Room.
04:20:01 IDADS Auxiliary steam relief valve reseats and does not lif t again.
Main steam pressure was 550 psig at this time.
There was enough flow across the pressure reducing valve to reduce the auxiliary steam pressure below the relief valve setpoint.
l Page 6
o LOSS OF ICS POWER TRIP REPORT NO. 75 04:20:20 IDADS Steam generator pressures have decreased to 500 psig.
Main feedwater flow begins.
At this pressure the running condensate pumps began to supply feedwater to the OTSGs through the idle Main Feedpumps.
This added approximately 1000 gpm to the feed rate of each OTSG for a little over two minutes.
04:207 OPS Operator sent computer technician to look at ICS power.
He confirmed all DC power supplies were de-energized and the A8T had not transferred.
04:21:25 IDADS RCS reaches minimum pressure of 1064 psig.
Operator is closing ADVs.
The co.nbination of water flashing in the head, the reduction in steam flow as the operator manually isolates the ADVs, and the increased HPI flow resulting from the RCS pressure decrease has stopped the pressure decrease.
HPI flow was now sufficient to keep up with the cooldown rate.
04:21:30 CALC /
Pressurizer surge line begins to refill.
RCS pressure and
[07:43]
OPS /
subcooling margin are increasing.
IDADS rormation of steam at the head has ceased.
HPI has refilled all steam voids.
04:22:00 IDADS Exceeded B&W recommended PTS curve for the reactor vessel.
04: 22: 40 OPS /
An operator is isolating TGVs.
RCS pressure is rapidly IDADS increasing.
04:22:50 IDADS Steam generator pressures have decreased.to 435 psig.
Main steamline failure logic closes the startup and main feed valves.
FW flow from the condensate pumps is stopped.
04:23:7 OPS Loca1' isolation of TBVs and ADVs is completed.
ADVs were the first valves closed, followed by the T8Vs.
04:23:10 OPS /
"B" AFW control valve partially closed using handwheel.
[09:23]
IDADS The operator thought he had completely closed the valve at this point.
Feed flow to the "B" OTSG, however, has decreased by about 60%.
This increased flow through the "A" AFW valve.
04:25:00 OPS /
Operator begins to close "A" AFW control valve with its IDADS handwhee l.
Page 7
)
LOSS OF ICS POWER TRIP REPORT NO. 75 04:25:30 IDADS Operator unisolates flPI pump SFAS recirc valves, opening the recire path to the Makeup Tank.
04:26:15 IDADS CR/TSC Essential HVAC train "8" is secured.
04:26:20 IDADS Pressurizer level returns on scale.
04:26:22 FUI/
"A" AFW valve closed. This stops AFW flow to the "A" OTSG.
IDADS operator believed the valve is only 80% closed, but could not close it any further by hand.
He left to locate a cheater.
04:26:47 IDADS Pressurizer level rising rapidly.
Subcooling margin is 170*F.
Operators start to throttle HPI injection valves to minimize RCS repressurization.
04:28:00 IDADS Operators stopped the "C" RCP per core lif t requirements.
RCS temperature is 410*F.
Flow from four RCPs at low RCS temperatures, may give excessive lifting force to the core components (fuel assemblies).
04: 28: 43 IDADS RCS letdown ficw is reestablished to help control pressurizer level.
Letdown flow is directed to the Makeup Tank.
Letdown flow can be directed to either the Makeup Tank or the Flash Tank.
It is normally lined up to the Makeup Tank.
04:28:45 CALC /
Makeup Tank level goes offscale high.
Makeup. Tank relief valve IDADS lifts and discharges to the Flash Tank.
Flash Tank level begins to increase.
The relief valve discharge is routed to the Flash Tank.
The Flash Tank pumps start on high level but do not have the capacity to handle the letdown flow and "B" HPI pump recirculation flow.
04:28:59 IDADS Operators stop "A" HPI pump.
The RCS pressure has peaked at 1616 psig.
RCS temperature at this time is 422*F.
04: 29:40 OPS /
Operator uses a cheater on "A" AFW valve.
The manual operator (15:53]
IDADS is damaged.
The valve reopens.
The operator calls the Control Room and is told to close downstream manual isolation valve FWS-063.
AFW flowrate to the "A" OTSG is approximately IDADS 1700 gpe, resuming the rapid cooldown.
Page 8
h LOSS OF ICS POWER TRIP REPORT NO. 75 04:29:45 IDADS "C" and "0" HPI injection valves are closed to reduce the rate of increasing RCS pressure.
06: 307 OPS "A" side BWST suction valve, SFV-25003, is closed in an actevat to decrease Makeup Tank level.
The operators assumed the running makeup pump would take a suction from the Makeup Tank if the BWST was isolated.
They forgot that the Makeup Tank outlet valve (SFV-23508) closes on SFAS actuation.
Their action isolated the suction of the Makeup pump, "A" HPI pump, "A" Decay Heat pump, and "A" R8 Spray Pump.
(Note that the "A" HPI pump and "A" Reactor Building Spray Pump were not running at this time.)
operators shift letdown flow to Flash Tank.
This action was taken to reduce the inflow to the Makeup Tank.
HPI recirculation flow continued to the Makeup
- Tank, l04: 30 OPS Shift Supervisor declares Unusual Fuent.
A Senior Control Room Operator begins notification of the state, counties, and the NRC.
04:30:40 IDADS Both OTSG operate levels are off scale high.
AFW flow continues to both OTSGs.
"A": > 1700 gpm, "0":.670 gpm.
04: 32:50 CALC Letdown flow is automatically diverted to a Coolant Waste Receiver Tank.
A 3-way valve diverted letdown flow when a high Flash Tank level was reached.
Flow to the Flash Tank was reduced but flow from the Makeup Tank relief valve continued to be slightly greater than the capacity of both Flash Tank pumps.
Flash Tank level continued to slowly increase.
04: 33:00 OPS /
Started depressurizing RCS to return to condition outside PTS IDADS region using normal pressurizer spray.
04:33:.!O OPS /
Operator arrives at "B" AFW valve and finds it partially open.
IDADS He closes it the rest of the way.
Feedwater to the "B" OTSG has been stopped.
This increases flow to the "A" OTSG to approximately 2000 gpm.
04: 33:40 IDADS "A" OTSG is full up to the top of the steam shroud and begins (19:53]
to spill water into the steam annulus.
Page 9
)
+
LOSS OF ICS POWER TRIP REPORT NO. 75 04:367 ops Operator is attempting to close FWS-063 but it will not move, even with a cheater.
04:39:00 IDADS RCS subcooling magin reaches peak of'201*F and begins to decline.
RCS temperatt're was 390*F.
RCS pressure was 1430 psig.
This is approximately 800 psi into the PTS region.
04:40:00 ops /
ICS power is restored.
An operator closed S1-S2 switches in
[26:13]
IDADS ICS cabinet 3.
Upon restoration of power, ADV demands went to 2005, T8V demands went to 01, "A" AFW valve demand went to 100%, 8 AFW valve demand went to 0%.
Operators reduced the 100% demand signals to 0%.
The "A" AFW valve responded and closed.
Operators regained control of ADVs, T8Vs, and AFW flow control valves.
The operators immediately reduced the demand signals to these valves to 0%.
Since the ADVs, were held closed by the manual handjacks, they could not respond to their demand signals and did not lif t.
The "A" AFW control valve was closed by thi operator action.
All feedwater flow to both OTSGs was then stopped and the RCS began to heat up.
The lowest RCS temperature of 306*F was reached and at this time RCS pressure (now 1413 psig) is being reduced to achieve conditions outside the PT3 region. The RCS had cooled down 196 degrees in 26 minutes.
04:40:10 IDADS Minimum OTSG pressures reached.
"A": 221 psig, "B":202 psig.
04:417 ops /
Operator calls control room and informs them that FWS-063 is IDADS stuck open.
He is told to disengage the handjack from the "B" AFW control valve.
Other operators are told to unisolate T8Vs.
04:41:10 IDADS "A" OTSG level goes below steam shroud.
An estimated 12,000 gallons of water spilled over the shroud and into the steamlines.
04:42:42 IDPDS A control room operator shuts down "B" HpI pump.
Makeup pump continuing to run.
04:42:56 IDADS A control room operator closes "A" and "8" injection valves.
Pressurizer level is 130 inches and increasing.
All sources of makeup water to the RCS are closed oncept RCP seal injection.
Page 10
e LOSS OF ICS POWER TRIP REPORT NO. 75 04:43:307 OPS The operators note a loss of RCP injection flow.
Seal flow slowly decreased as the "B" HPI pump coasted down.
The operators were puzzled by the fact that seal flow was being lost with the Makeup Pump continuing to run.
They were not yet aware of the Makeup Pump's isolated suction.
04:43:54 IDADS An operator restarts "8" HPI pump to supply RCP seals.
Seal flow returns to normal.
- 04:50:19 IDADS An operator stops "B" HPI pump again.
OPS The operators checked the RCP seal injection valves but could not find anything wrong with the valve lineup so the pump was shutdown again.
04: 50:30 IDADS The operators again notice loss of seal flow and restart "B" HPI pump.
04: 527 OPS A senior operator collapses in front of control panel.
He is moved to an office adjacent to the control room.
04: 57 OPS Control room operators start blowdown of both OTSGs to reduce lovel.
04:50 OPS Both Emergency Diesel Generators are shutdown.
They had run unloaded since the SFAS initiation.
04: 587 OPS The opera'. ors in the Control Room hear a loud noise.
The
[447]
" MAKEUP /HPI PUMP LUBE OIL LOW PRESSURE" annunciator alarms.
An operator notes that the Makee7 Pump ammeter is reading only a fraction of normal running current.
He realizes the pump has been damaged due to lack of suction.
The pump became separated from the motor when the pump seized ard broke the coupling between the speed changer and pump, 05:00:10 IDADS An operator trips the Makeup Pump.
I Page 11
LOSS OF ICS POWER TRIP REPORT NO. 75 05:01:22 OPS /
The operators open the Makeup Tank outlet valve (SFV-23508).
IDADS Water from the tank spills out of the damaged Makeup pump seals and onto the pump room floor, Approximately 1200 gallons is spilled.
The operators deduce why the sharp drop in Makeup Tank level had occurred and reclose the outlet valve.
The operators were aware that the known leak at the Makeup Pump to the Auxiliary Building was isolated from RCS pressure by a single check valve. They were concerned this might lead to a LOCA.
05:05 IDADS The RCS crossed out of B&W recommended PTS region and a 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> OPS soak is in progress.
An ambulance is called for SRO.
05:08 OPS The Auxiliary Building Gas Radiation Monitors (R15002B and R15045) go into alarm.
These instruments monitor air being exhausted from the Auxiliary Building basement.
One of these rooms is the Makeup Pump room.
The instrument is detecting gas from the mildly radioactive water spilled in this room.
05:09 IDADS Doth AFW pumps are stopped.
05:12:26 CALS/
Both Deca / Heat pumps are stopped.
- 12:40 IDADS The "A" DHR pump had been running without a suction from the BWST since the closing of the BWST outlet valve (SFV-25003).
The pump ran on its own recirculation and always had adequate NPSH.
There was no damage to the pump.
05:13 OPS doth Auxiliary Building Stack radiation monitor alarms clear.
05:19 SCyTY The ambulance arrives at the site.
05:21:45 CALC Flash Tank begins to overflow into Waste Gas System.
05:277 OPS Makeup Pump manual suction and discharge isolation valves are closed.
This isolates the pump from the Makeup Tank /BWST suction header.
The "A" HPI pump is now available for service, if needed.
05:29:04 IDADS Operators stop the "A" RCP per procedure.
Page 12
(
LOSS OF ICS POWER TRIP REPORT NO. 75 05:36:04 GACP Zone 20 fire alarm trips the Rad Waste Area Exhaust Fan
[82:17]
(A-542A er A-5428).
An interlock prevents restart of the fans until the fire alarm is manually reset.
The St AS initiation had closed the suction and discharge valves to the Reactor Building radiation monitor (R15001A&B). The compressor which normally pulls a sample from the Reactor Building overheated after running for over an hour with its flowpath isolated. The smoke from this compressor actuated the smoke detector in this room. The fans are located in an adjacent room within the same fire zone.
If a fire is detected in either room the fans are tripped. The fans draw air from the -20 an6
-47 foot areas of the Auxiliary Building, Spent Fuel Building and Chemistry labs. The Makeup Pump room is located in the -20 foot elevation of the Auxiliary Building.
05:387 OPS Control Room operators stop R15001A&B compressor.
05:39:37 GACP Zone 20 fire alarm is reset and one of the exhaust fans is restarted.
It runs for less than a minute and the fire alarm trips it again.
05: 40?
OPS Main Streamline Failure Logic is inhibited.
This action permits the normal feed flow pathway to the steam generators to be used.
05: 48 OPS TSC sprinkler system is isolated.
05:50 SCRTY The ambulance with the SRO departs for the hospital.
05:54 OPS Outside operator reports loss of security badge to Watch Commander.
The operator lost his security badge and radiation film badge around the time he was helping the licensed operator manually isolate the ADVs.
He was busy assisting in the stabilization of the plant and did not report the loss until this time.
He did not enter any radiologically controlled areas without his film badge.
05:59:07 GACP Zone 20 fire alarm is reset and one of the exhaust fans is 05:59:50 restarted.
It runs for a few seconds and the fire alarm trips 06:01:21 it again.
This sequence is repeated three times.
06:02:15 GACP Zone 20 fire alarm is reset and one of the exhaust fans is restarted. The smoke cleared from the area and the fan continues to run throughout the remainder of the shift.
Page 13
s LOSS OF ICS POWER TRIP REPORT NO. 75 06:02 OPS Health Physics samples showed 1.5 MPC's of Xenon 133 and 135 in -20 foot level of Auxiliary. Building.
06:04 OPS An operator bypasse s Safety Features signals.
06:11 OPS "ICS OR FAN POWER FAILURE" alarms on main annunciator panel but immediately resets.
There is no loss of power.
No
- equipment response is noted.
06:14 OPS "ICS OR FAN POWER FAILURE" alarms on main annunciator panel.
An operator immediately resets S1-S2 switches and ICS power is restored.
Many ICS demands go to 100%.
The operators reduce demands to 0% on ADVs, main and startup feedwater valves, and AFW control valves.
Operator response is quick enough to prevent another overcooling.
Feedwater flow from the condensate pumps through the open startup feedwater valves added approximately 750 gallons of water to each steam generator before operators can complete the closing of the valves.
06:15 SCRTY Security brings a spare visitor's security badge to the Control Room for the operator who lost his badge.
07:00 OPS The Senior operator released.from hospital.
07:02 OPS Site boundary release of.93 MPC from the Auxiliary Building stack is calculated.
State OES is informed.
07:057 OPS Makeup Tank outlet valve is opened.
07:15 OPS The "A" HPI pump is started. The "B" HPI pump is' stopped.
The "A" HPI pump took its suction from the Makeup Tank.
Makeup Tank level began to decrease. This terminated the Flash Tank overflow into the Waste Gas System.
08:41 ECL The Unusual Event is terminated.
[267:]
I Page 14
a LOSS OF ICS POWER TRIP REPORT NO. 75
'III.
PRETRIP PLANT STATUS (at 0412)
Reactor Power:
75%
RCS Tavg:
582*F RCS Pressure:
2160 psig Subcooling Margin:
40'F Pressurizer Level:
200 inches Generated MWe:
710 MWe (gross)
Condenser Vacuum:
28.7" Hg Steam Header Pressure:
882 psig A OTSG Level:
53%
8 OTSG Level:
58%
Delta Tc:
-1.6 F.*
All ICS Stations in AUTO 2 ADVs per OTSG Isolated **
Per procedure to reduce 8 OTSG level.
Rancho Seco has three (3) ADVs per OTSG.
Each valve has the steam relieving capability of three and a half percent full power.
Four of these valves are normally isolated to prevent a rapid cooldown should the ICS experience certain single failures.
These failures would have opened all the TBVs and ADVs.
Total' steaming capability with the present Rancho Seco alignment provides for 28% steam dump capacity.-
This brings the Rancho Seco plant in line with other B&W plants. The failure of ICS with the 28% steam dump capacity has been analyzed.
The r
j 40% capacity failure has not.
I l
l j
Page 15 l
l.
LOSS OF ICS POWER TRIP REPORT NO. 75 IV.
INITIATING EVENT AND REACTOR TRIP The unit had been running at a steady state power level of 75% for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> prior to the trip.
Power was being held at this level until one of the control room's two main computer systems was repaired.
The failed compute provided the operators with core thermal power and precise core imbalance and tilt calculations as well as many data points and alarms.
At 0413, a loss of all ICS DC power occurred.
The operators were immediately alerted that a major plant transient was beginning by the "ICS RUNBACK OR LIMIT" annunciator. The loss of ICS power was evidenced by an ICS power annunciator, loss
' of indicating lights on all the ICS stations on the main control console, and all ICS (dgewise meter indications going to midscale.
Many of the individual modules in the ICS are powered from a positive and a negative 24 VDC bus.
Each bus receives auctioneered power from either of two positive or negative supplies.
One set of positive and negative supplies are fed from the vital 120VAC "C" bus and the other set from the non-vital "J" bus.
A power monitor checks for low voltage on either bus.
It isolates all DC power supplies via a main switch (S1-S2) should it sense less than 22 volts on either bus.
Acceptable ICS performance cannot be assured at voltages less than this value so all DC power is removed from the ICS to fail controlled components to a predictable mode.
Most modules transform the dual power supply voltages to +/-10 VDC.
When all the DC power supplies are isolated, demand signals to plant components become zero volts.
This presents a 50% demand signal to all ICS air-operated valves, a low speed stop demand (2500 gpm).to main feedwater pumps, and a close signal to main feedwater block valves.
The main turbine EHC pulser also responds to signals from ICS. When ICS power goes to zero volts, no signal is generated to the pulser to tell it to open or close the turbine governor valves.
The turbine no longer acts as a pressure regulator.
The turbine governor valves remain in the positiun which just preceded the loss of ICS power.
Main feedwater flow to the steam generators was immediately affected by the loss of ICS power.
The flow rapidly decreased due to runback of feedpump speed and the partial closing of the startup and main feed regulating valves.
Feedpump discharge pressure decayed to the point where pump discharge pressure was insufficient to feed the steam generators.
Total steam flow increased as the TBVs and ADVs opened and the turbine governor valves remained fixed in position.
This additional heat removal from the RCS could not compensate for the loss of feedwater.
The rod control station transferred to MANUAL when ICS DC power was. lost and prevented a cross limit rod insertion response to the decreasing main feedwater flow.
RCS temperature and pressure increased.
An operator manually opened a pressurizer spray valve. The spray could not overcome the RCS pressure increase.
Fifteen seconds after the loss of ICS power, the Reactor Protection System tripped the reactor on high RCS pressure (2300 psig).
The turbine and generator tripped from the reactor trip. The RCS pressure continued to increase due to the momentum caused by loss of feedwater, reaching a maximum of 2315 psig before turning.
Within a few seconds RCS
(
pressure began to decrease and the operator closed the spray valve.
Two seconds prior to the trip, the motor driven AFW pump autostarted on low feedpump discharge pressure (850 psig).
Page 16
-O LOSS OF ICS POWER TRIP REPORT NO. 75 V.
POST TRIP TRANSIENT RESPONSE The momentum of the RCS pressure and temperature continued for a few seconds following the trip.
RCS pressure peaked at 2315 psig.
Six secondary side code safeties lifted for a brief period.
All resented properly. The con.rol room received a few false annunciator alarms when the reactor tripped. These are attributed to electrical noise or voltage transients generated when the site loads transferred from the Auxiliary Transformers to the Startup Transformers on the Main Generator trip.
The control room also received a Technical Support Center (TSC) fire alarm at the time of the trip.
The smoke detector for this zone is fed from the "F" 120 VAC bus. This bus is ultimately supplied by the Auxiliary or Startup Transformers.
When site loads transferred, the voltage transient was sufficient to actuate a system solenoid valve.
The protection system is a pre-action water system. The solenoid actuated only the first part of this two-part system.
There was no sprinkler discharge into the TSC.
A flow of water did pass through the open solenoid valve and ultimately onto the TSC floor.
The TSC is located next to the control room and is sealed as part of the Control Room protected boundary. The water accumulated on the floor as there are no drains in the TSC.
Cabling under the false floor was soaked and some water dripped into a vital 480 volt switchgear room directly below the TSC.
The flow of water continued for some time as operators attended to higher priority items.
The water did not cause any malfunction of equipment during the transient.
An auxiliary steam relief valve lifted coincident with the reactor trip.
It relieved excess pressure in the system caused by the fifty percent demand signal to the main steam to auxiliary steam pressure reducing valve.
The relief valve cycled open and closed for the next six minutes.
Main steam pressure was then low enough and the differential pressure across the reducing valve high enough to lower system pressure below the relief valve's setpoint.
There is no data on auxiliary steam pressure for this transient due the outage of the Bailey computer.
However, steam pressure must have been adequate.
The turbine gland sealing system and air ejector functioned throughout the event.
There were no condenser vacuum problems. The auxiliary boiler was placed in service soon after the trip.
The pressurizer level was decreasing due to the normal post trip contraction of reactor coolant.
An operator opened "A" HPI injection valve to increase makeup to the RCS.
The operators noted a rapidly decreasing level in the makeup tank.
This was due to the high rate of makeup (300 gpm).
The "A" BWST outlet valve ses opened.
This action stops the makeup tank decrease at the simu-lator.
At Rancho Seco, this will not stop the Makeup Pump from taking its suction from the Makeup Tank until either the Makeup Tank outlet valve is closed or the pressure in the Makeup Tank decreases to a low enough value where the height of the water in the BWST exerts more pressure than the combined height of the Makeup Tank level and its gas overpressure.
The Makeup Pump continued to take its suction from the Makeup Tank and the tank level continued its decrease.
(
Page 17
LOSS OF ICS POWER TRIP REPORT NO. 75 Three seconds af ter the trip, the dual driven AFW pump, P-318, autostarted on its steam turbine from low main feedpump discharge pressure (850) psig).
AFW flow to the steam generators through the half open AFW control valves began two seconds later. There was no main feedwater flow to the steam generators at this time.
Initially, the AFW flow rate could not keep up with the steam flow rate out the open TBVs, ADVs, and main steam relief valves.
OTSG levels decreased below low level limits for a short time. This did not jeopardize core cooling.
An actual level is not required for AFW to provide cooling.
Cool AFW is sprayed into the top of the steam generator.
The water droplets can be heated and changed to steam prior to reaching the bottom of the generator. Once AFW flow began to the OTSGs, th< RCS had more than adequate cooling.
As OTSG pressures decreased due to the partially open TBVs and ADVs and the spraying effect of the cool AFW into the top of the OTSG, the AFW flow rate increased.
Steam flow rate decreased as the main stem relief valves reseated. Two minutes after the trip, AFW flow overtook steam flow and OTSG levels began to increase.
A control room operator believed he heard a feedwater shell relief. valve lifting.
Pegging steam is automatically directed to the shell side of second and fourth point heaters following a turbine trip,taking the place of extraction steam.
Extraction steam flow is lost when the main turbine trips.
Pegging steam had lifted heater relief valves on recent Rancho Seco trips and contributed to an overcooling transient.
Two of the pegging steam pressure control valves were found leaking past their seats when fully closed.
This presents no problem when there is feedwater flow through the heaters.
The steam condenses at a greater rate than the rate of leakage.
Following this trip there was no main feedwater flow and no condensation of the pegging steam.
Leakage across the control valve seats built pressure up in the heater (s) until the relief valve (s) lifted.
pegging steam was isolated from the control room soon after the trip.
Core decay heat following.the reactor trip was relatively low due to the short power history.
The heat removal rate from excessive steam and AFW flow soon exceeded the j
heat input from the' core and RCS metal.
Within two and a half minutes, RCG average I
temperature had decreased below the expected post trip value. The RCS was cooling down at a rate in excess of one thousand degrees an hour.
The' operators noted that both AFW pumps had autostarted and that AFW flow was being fed to the steam generators at a high rate.
Non-licensed operators with walkie-talkies wert dispatched to close the AFW control valves.
The operators were planning to use the modulating SFAS valves to control feed flow rate.
An operator tripped both main feed pumps because he had no control over main feedwater.
The startup and main l
regulating valves were receiving fif ty percent open demands.
Pressurizer level was decreasing due to the cooldown.
Operators started the "B" dPI pump to increase the RCS inventory.
Recirculation flow from the makeup pump and HPI pumps is directed to the makeup tank. The additional recirculation flow from the "B" HPI pump caused the makeup tank level to decrease more slowly.
The additional l
i makeup flow from the second high pressure pump did little to slow the rate of pressurizer level decrease. The RCS was contracting at a rate of 200 cubic feet por l
oinute.
Total HPI, makeup, and RCP seal injection flows were 70 cubic feet per minute.
Page 18 l
1 l
l
i LOSS OF ICS POWER TRIP REPORT NO. 75 Operators unlocked the ICS cabinets. They found all DC power supplies deenergized.
The dual switches (S1-S2) which provide power to the four supplies were found in the full down position and were assumed to be closed.
The OFF indication was difficult to see.
Actually both switches had tripped. The operators believed these switches to be breakers. They expected them to be in a mid position if they were tripped.
Operators also checked the status of the Automatic Bus Transfer device (ABT) that lsuppliesACfieldpowertoICSinstrumentation.
It was found to be powered from its normal source, the "C" vital bus. The operators switched its power source to the "J" bus in an attempt to regain power to the DC supplies.
(The "J" inverter which supplies this bus had been removed from service several weeks earlier.
All loads normally powered by the "J" inverter were being fed from the "F" bus.). This action
- had no effect and the ABT was transferred back to the "C" bus.
An operator was sent to check the supply breakers to the ABT.
He reported that both breakers were closed. The operators decided to abandon attempts to regain ICS power and concentrated on arresting the transient under manual control.
Coincident with the inspections of the ICS. cabinets, non-licensed operators were told to manually isolate the TBVs using a handjack installed on each valve.
A control room operator left to manually close the ADVs.
A recent modification to the control circuit of the TBVs and ADVs could have been used in lieu of sending operators out to manually close the valves.
There are switches located in the remote shutdown panel that give the operator the option of closing the valves independent of the ICS signals to the valves.
This panel is located two floors below the control room in a switchgear room.
The modification was put in place for evacuation of the control room (Appendix R) and their use only appeared in-the evacuation casualty procedure.
These Switches were not remembered until the transient was over.
RCS pressure was decreasing rapidly because of the overcooling.
Within three minutes of the trip, Safety Features (SFAS) automatically actuated on low RCS
(
pressure (1600 psig).
Three HPI injection valves opened to their prethrottled l
position.
The "A" HPI injection valve had been fully opened previously.
This valve l
traveled in the closed direction until it reached its prethrottled position.
The I
AFW SFAS valves opened.
Thes9 valves are in parallel with the flow control valves.
Within a few seconds'after SFAS initiation, the operators closed these valves.
Through training they knew these valves can quickly lead to a severe overcooling a.,d steam generator overfill if not closed.
Both decay heat pumps started.
By design, selected SFAS equipment tripped and block loading of SFAS equipment began.
One of the components tripped was the motor driven AFW pump.
Thirty seconds later it block loaded back ' onto its vital bus.
This resulted in AFW flow rate decreasing for thirty seconds.
The dual driven AFW oump continued to run via its steam supply throughout this SFAS initiation.
Thetmakeup pump and "B" HPI pump continued to run and the "A" HPI pump was started by t$e SFAS signal.
The diesel generators also started from the SFAS signal but the output breakers did not close onto their respective dital buses as there was no loss of voltage to the vital buses.
f l
Page 19
i LOSS OF ICS POWER TRIP REPORT NO. 75 Less than five minutes after the trip, RCS temperatures were less than 500*F.
At this point, one RCP should have been tripped to avoid core lift concerns.
The operators ' forgot this step until ten dinutes later. when RCS temperature had decreased to 410*F.
The concern regards the lifting of fuel assemblies off the bottom core grid when denser water enters the bottom of the core as the RCS cools.
A B&W analysis of hydraulic core lift was performed. The results indicate that the fuel assemblies remained positioned in their lower grid supports.
There was no thermal hydraulic damage to any core components.
Control room operators shut down one. train of Control Room / Technical Support Center Essential HVAC soon after its autostart from the SFAS signal.
Running both systems-
- creates a high ambient noise level in the control room.
The systems are redundant so securing one train is not detrimental.
The additional HPI that SFAS provided was not sufficient to counteract the RCS contraction rate.
About six minutes after the trip, indicated pressurizer level went offscale low.
It remained offscale for approximately nine minutes.
A minute and a half after level went offscale, the surge line drained. The pressurizer steam exiting the surge line collapsed as it mixed with the turbulent flow of subcooled
~
water in the hot leg.
RCS pressure started a steeper decline.
Subcooling margin, which had been 85 F, started to decrease and reached a minimum of 77aF.
The RCS saturation temperature at this point was 580*F.
Just six minutes prior to this time, the reactor vessel head and upper internals were 601oF.
This area does not receive as turbulent a coolant mixing as does the other portions of the RCS.
The large mass of metal in this area would take longer to cool.
The RCS saturation temperature was low enough Jto start the formation of nucleate steam bubbles on the hottest, lowest flow regions in the head area.
The nucleate bubbles displaced the liquid coolant and RCS pressure decreased at a slightly slower rate.
Flow from the HPI pumps increased due to the decrease of RCS pressure.
The increased flow also played a part in the rate reduction of pressure decrease.
Subcooling margin began to increase once again.
The steam bubbles broke away from the metal surfaces and were swept into the cooler bulk coolant where they collapsed.
RCS pressure continued to decrease as the rate of contraction due to the addition of AFW and the steaming of the ADVs and TBVs was greater than the RCS inventory displaced by the steam bubbles.
Greater numbers of steam bubbles were produced as the saturation temperature decreased.
HPI flow increased as the RCS pressure decreased. The steep decline of RCS pressure started to taper off.
The nucleate bubbles collapsed within a few inches of their origin.
The bulk coolant remained subcooled.
There was no two-phase flow through the hot leg piping.
Page 20
9 LOSS OF ICS POWER TRIP REPORT NO. 75 Steam generator pressures had decreased below 500 psig by this time.
This pressure is low enough to allow the running condensate pumps to feed the OTSGs through the half open startup feed valves.
The loss of ICS power had caused the feedwater block valves to close and isolated the flow path through the main regulating valves. The flow continued for.about two minutes and added approximately two thousand gallons to the inventory of each steam generator.
Steam pressure' continued to decrease to the point where main steamline failure logic was actuated (435 psig).
Solenoid valves vented air from the bottom of the startup and main feed valve diaphragms and blocked their air supply. The valves closed. This stopped the feed from the condensate pumps.
Seven and a half minutes after the reactor trip, the RCS reached its lowest pressure of 1064 psig.
This corresponds to a saturation temperature of 554 degrees.
The pressure remained constant for approximately one minute.
Three factors caused the RCS pressure to stabilize.
1.
The saturation temperature was low enough to allow a great number of nucleate steam bubbles to form on the underside of the reactor vessel head.
The RCS pressure will tend to stay at the saturation pressure of these steam bubbles.
2.
HPI flow had increased due to the RCS pressure reduction.
3.
An operator was manually closing the ADVs.
Total steam flow and thus RCS cooldown and contraction rate were being reduced.
As the operator completed the isolation of the ADVs, the RCS contraction rate was reduced and the flow from the HPI pumps was sufficient to keep up with the cooldown.
RCS pressure began to increase and the nucleate boiling at the head
. ended.
The pressurizer surge line began to refill.
As the operator moved to the TBVs and started his isolation, the surge line filled and pressurizer level started to increase.
With the TBV isolation completed, the only remaining steam loads were main steam supplies to the dua'l driven AFW pump and the auxiliary steam loads. The reduction in steam flow caused RCS pressure and pressurizer level to increase rapidly, although level was not yet within the minimum indicating range.
The RCS was continuing to cool down at a rate of.650 degrees per hour.
Subcooling margin increased to over 100 degrees.
This exceeded the B&W recommended Pressurized Thermal Shock (PTS) guid11ines for the reactor vessel.
The steam release through the ADVs and main steam safety valves led to a minor radiological offsite release.
Residual radioactivity left in the steam generators and steamlines from previous OTSG tube leaks plus a suspected small, pre-existing primary-to-secondary leak led to some minor contamination of the OTSGs.
The rapid cooldown of the OTSG tubes relative to the OTSG shell placed the tubes in tensile stress. This condition may have aggravated the pre-existing leak or resulted in several new leaks.
A conservative estimate of the offsite release is 0.85 MPCs at the site boundary, the primary element being Cesium.
Page 21
LOSS OF ICS POWER TRIP REPORT NO. 75 Nine minutes af ter the trip, an operator was using the manual handwheel to close the "B" AFW control valve.
He believed he had closed the valve completely but flow had only been reduced by 60 percent.
He reported to the control room that the valve was closed. Total AFW flow did not decrease significantly. The flow reduction to the "8" OTSG resulted in a flow increase to the "A" side.
The operator then went to the "A" AFW valve.
He completely closed the valve but thought it was only 80 percent closed. He could not move the handwSwel any further by hand and left to locate a cheater.
The closing of the "A" AFW valve and partial closing of the "8" valve reduced total AFW flow from an estimated 2500 gpm tc 1000 gpm.
This and the OTSG pressure reduction decreased the RCS cooldown rate to 50 degrees per hour.
Subcooling margin was 160 degrees and increasing. Emergency Procedures permit HPI throttling as long as subcooling margin is adequate, regardless of pressurizer level. Operators first opened the HPI recirculation valves which had closed on the SFAS signal.
This resulted in a 100 gpm decrease in HPJ flow to the RCS.
Pressurizer level indication returned. Operators throttled HPI injection -in an attempt to minimize the RCS repressurization.
HPI recirculation flow is. directed to the makeup tank.
RCS letdown flow to the makeup tank was re-established to aid in the control of pressurizer level.
HPI pump recirculation flow increased as HPI flow to the RCS was throttled.
Total flow to the makeup tank was greater than 300 gpm.
The SFAS signal had closed the makeup tank outlet valve (SFV-23508).
Makeup tank level rose quickly.
Operators tripped the "C" RCP at this time.
RCS temperature was 410 degrees.
As HPI flow continued to increase pressurizer level and subcooling margin, operators secured the "A" HPI pump and closed two HPI injection valves.
The non-licensed operator returned to the "A" AFW valve with a cheater. While coplying force to close the valve, the manual operator was damaged and shifted.
i The l
valve reopened and allowed approximately 1700 gpm to be fed to the "A" generator.
The RCS began to cool down at 230 degrees per hour.
Both steam generator operate levels went offscale high.
The HPI injection balanced the contraction and letdown rates.
Pressurizer level and RCS pressure remained stable.
Subcooling margin continued to increase. The operator called the control room and informed them the control valve had. opened and he could not manually close it.
He was told to close the downstream manual isolation valve, FWS-063.
The valve could not be moved from its fully open position, even with a cheater.
Another operator who had also been directed to close the AFW control valves arrived
.it the "B" valve and found it partially open.
He closed it and stopped all feed to the OTSG.
The "B" OTSG reached a maximum level of 475 inches, or about 85 percent full.
Closing the "B" valve forced more AFW flow through the "A" valve, although the total feedwater flow had been reduced somewhat due'to the flow losses across the control valve at such high flow rates.
Page 22
LOSS OF ICS PGWER TRIP REPORT NO. 75 As the "A" OTSG continued to fill, there was less available tube length for the incoming AFW to be heated and vaporized to steam.
As water spilled over the top of the steam shroud, much of the AFW contacted only a few peripheral tubes and drained inte the steam annulus before it had a chance to heat to saturate conditions.
The dec. ease in total AFW flow and reduction of available tube length above the steam generator liquid level reduced the cooldown rate to 160 degrees an hour.
HPI injection rate was now greater than the contraction and letdown rates.
Pressurizer level began to increase once more. Operators opened one of the normal pressurizer spray valves to depressurize the RCS out of the Thermal Shock region.
Subcooling mar 3fn remained in the 190-200 degree range, however, as the cooldown rate kept pace with the pressure reduction.
HPI flow 'of 200 gpm was continued to restore pressurizer level to a " comfortable" margin above the minimum level indication.
Procedures direct the operator to maintain 100 inches following a trip.
A higher level also clears the pressurizer heater low level interlock.
These heaters will be required to establish saturation conditions in the pressurizer.
A disadvantage of allowing the level to increase was the compression of the steam bubble in the top of the pressurizer.
This action retards the efforts of pressurizer spray in reducing RCS pressure.
Operators noted makeup tank level and pressure offscale high.
The 100 psig relief valve lif ted and discharged to the flash tank.
Operators shifted letdown flow to the flash tank.
There was a concern that some of.the makeup tank water could find its way into the waste gas or vent gas systems.
Operators closed the "A" side BWST suction valve, assuming that the running makeup pump would take its suction from the makeup tank. This would have reduced the tank level and pressure.
They neglected to open the makeup tank outlet valve (SFV-23508) which had been previously closed by the SFAS signal.
This action would have been successful at the simulator.
The simulator logic will take makeup pump suction from the 8WST if the BWST suction valve is open and the i
l makeup tank if the valve is closed.
The motor-operated makeup tank outlet valve is not modeled.
(The simulator was changed to agree with Rancho Seco soon after this event.)
The makeup pump was running without a source of water.
The "A" HPI' pump had been tripped earlier.
The "B"'HPI pump continued to supply HPI injection and RCP seal injection since its supply from the BWST had not been interrupted. Without a suction, makeup pump recirculation flow to the makeup tank ceased.
The makeup tank remained full, its relief valve continuing to pass the "B" HPI pump recirculation flow to the flash tank.
The combined relief valve and letdown flows were greater than the capacity of the two flash tank pumps.
Flash tank level began to increase.
Letdown flow was automatically redirected to a coolant waste receiver tank when the flash tank reached a high level.
The rate of level increase declined but the "B" HPI pump recirculatien flow being passed through the makeup tank relief valve continued to be slightly greater than the combined capacity of the flash tank pumps.
Page 23
LOSS OF ICS POWER TRIP REPORT NO. 75 ICS power was restored thirty-six minutes after it had been lost.
A licensed operator noted that the SI-S2 switches were off.
He first tried to close just one of the switches.
It immediately tripped.
He then closed both switches simultaneourly and they remained closed.
Response of the ICS demand signals upon re-energizacion varies.
1.
100% demand:
ADVs, "A" AFW control valve, both startup feedwater valves, "B" main feedwater valve, both feedpumps.
1 2.
0% comand:
T8Vs, "B" AFW control valve, main steam to auxiliary steam reducer.
3.
25% demand:
"A" main feedwater valve.
All stations reinitialized in manual.
The operators quickly reduced the demand on all the ICS stations to zero percent.
The "A" AFW valve damaged manual operator did not restrict stem movement in the closed direction.
The AFW valve responded to the close demand and AFW flow to the OTSG stopped.
The ADVs did not lift because their manual handjacks were holding them closed.
There was no main feed flow to the OTSGs because the main steam failure logic signal was holding the valves closed as it has priority over ICS demand s.- Auxiliary steam was being supplied by the auxiliary boilers by this time so the closing of the main steam reducing valve had no effect on auxiliary steam pressure.
The rapid cooldown ended when the AFW flow to the "B" OTSG was stopped.
The minimum RCS temperature was 386 degrees.
The minimum steam generator pressures were slightly over 200 psig.
Subcooling margin also peaked about this time at 201 degrees.
The RCS began to heat up at a rate of 140 degrees per hour.
Operators were directed to unisolate the T8Vs and "B" AFW valve.
Control room operators slowly increased the steaming rate through the TBVs.
They were aware that the steam generators had been filled to a high level and assumed that there was water in the steam lines.
They did not want to cause water hammers with high steam flows.
A few water hammers were heard in the secondary plant at this time although their source could not be located.
A walkdown of the main steam lines did not reveal any damage to piping or piping supports.
Pressurizer level was 120 inchet and increasing due to the heatup and 200 gpm HPI injection. Oper ators shutdown the "B" HPI pump and closed the last two injection valves.
Operators noticed the RCP seal injection flowrates trending toward zero.
They restarted the "B" HPI pump' and seal injection flow returned to normal.
They checked the seal injection valve lineup but could find nothing wrong.
The "B" HPI pump was stopped again with the same results.
The "B" HPI pump was restarted.
l Page 24 l
l LOSS OF ICS POWER TRIP REPORT NO. 76 At this time one of the senior control room operators collapsed in front of one of the main control consoles.
He was the licensed operator who had left the control room to close the ADVs just after the SFAS initiation.
Since his return he had been complaining of storsch pains.
He was moved.to an office adjacent to the control room and was attent.ed to by a non-licensed operator.
It was hoped he would recover if left to rest.
The control room operators heard a loud noise and received the " Makeup /HPI Pump Lube Oil Low Pressure" annunciator a second later.
An operator noted the makeup pump ammeter was reading only a fraction of normal running current. The loud noise vas probably a result of the pump seizing and the breaking of the coupling between the
' pump and motor. The motor continued to run unloaded which accounts for the low reading on the ammeter.
The annunciator was actuated when the shaf t driven lube oil pump stopped when the pump seized. The operators realized ~the makeup pump had been severely damaged due to lack of suction and tripped the pump motor breaker.
The "A" HPI pump had not been run since the common suction had been isolated and was not damaged.
The makeup tank outlet valve was opened.
Makeup tank level decreased rapidly.
The outlet valve was 1.nmediataly closed.
Approximately 1200 gallons of mildly radioactive water spilled out of the makeup pump's damaged seals onto the pump room floor.
The water accumulated on the floor to a depth of one or two inches as the floor drain could not immediately handle all the water.
The drain is routed to a sump on a lower floor of the Auxiliary Building.
The sump is pumped to a tank in the Miscellaneous Radwaste System.
The pump room, sump, and tank are all ventilated by the Radwaste Exhaust System.
Radwaste ventilation fans pull air from these areas and exhaust it to atmosphere after it is passed through HEPA and charcoal filters.
The exhaust system radiation monitors alarmed two minutes after the spill.
The release lasted no more than ten minutes.
It was conservatively estimated at 0.93 NPC at the site boundary.
The primary isotope was Xenon.
The RCS continued to heat up but'at a slower rate as operators increased the steaming rate using TBVs.
RCS pressure was decreased with pressurizer spray.
The RCS pressure-temperature relationship passed out of the PTS region forty-three minutes after its entry.
The heatup was arrested and a three-hour soak period was begun per procedure.
RCS temperature was stabilized at 435'F and pressure at 770 psig.
Letdown slowly reduced pressurizer level.
OTSG blowdown to the condenser was
^
started to reduce steam generator level.
Both AFW pumps were stopped.
The normal RCS makeup path was unis31ated.
The "A" and "B" decay heat pumps were shut down.
The "A" decay heat pump shares the same BWST suction as the makeup and "A" HPI pumps cnd was isolated from tha BWST when the BWST outlet valve (SFV-25003) was closed.
The pump ran on its own recirculation.
It is located one floor below the HPI pumps cnd never lost its NPSH.
The pump was not damaged.
Page 25
LOSS OF ICS POWER TRIP REPORT No. 75 The senior reactor operator's condition had not made much improvement.
An ambulance was called. The operator was taken to a nearby hospital forty-five minutes later after examination by the paramedics.
He was released from the hospital an hour later, the diagnosis being c,xhaustion and hyperventilation.
The known leak from the Ma'<eup Pump to the Auxiliary Building was isolated from the RCS by a single check valve. Operators were concerned about the possibility of a LOCA. Two non-licensed operators were sent to inspect the damaged pump.
They were told they would probably encountered a great quantity of water on the floor.
A few minutes.later they were told to isolate the pump.
They could not locate plastic high-top shoecovers and did not wear any respirators on either entry into the room.
Some of their personal clothing was contaminated as they walked through the water on the floor. One operator sustained minor skin contamination.
The pump was l
successfully isolated.
The Reactor Building radiation monitor's supply and discharge valves were closed on the SFAS signals. The newer monitor (R-15044) senses loss of flow and shuts its compressor down automatically.
The older monitors' compressors did not have this feature and continued to run.
One monitor's compressor overheated (R-15001 A&B).
An hour and fif teen minuter. af ter the SFAS valves were closed, smoke from the compressor seals initiated the zone's smoke detector (Zone 20).
The Auxiliary Building radwaste exhaust fans are located ~in a room adjacent to the rad monitors, although in the same fire zone.
An actuation of any fire detector in Zone 20 trips the exhaust fans.
They cannot be restarted until the smoke clears and the zone alarm is manually reset.
The operators recognized the source of the fire alarm and shut the compressor off.
Numerous attempts were made to reset the fire alarm and restart the fan.
Twenty-six minutes after the fan's initial trip, it was successfully restarted and continued to run.
Nearly an hour af ter the flash tank had started receiving water from the makeup tank relief valve, the flash tank was full.
This tank is continuously vented to the Waste Gas System.
Water started to enter this piping.
The system is totally closed. No water or gas escaped to the atmosphere.
The system was not overpressurized.
Subsequent to the event, the system was drain,ed and inspected, and leak checked. No damage to the system resulted from the overfill and the system was returned to service.
Two hours after the initial loss of ICS power, the control room received the "ICS OR FAN POWER FAILURE" annunciator again.
It came in for an instant and then reset.
There was no actual loss of power.
Two minutes later the annunciator alarmed, the SI-S2 switches tripped, and ICS DC power was lost.
Within a few seconds, the switches were reclosed and power was restored.
Operators quickly reduced all demands to zero.
No plant transient was experienced.
The ADVs were still manually isolated and the T8Vs received a zero percent demand signal.
The condensate pumps did add warm feedwater to the steam generators through the open startup valves.
(MSFL had been bypassed earlier.) OTSG levels increased by approximately 750 gallons each.
(
Page 26
LOSS OF ICS POWER TRIP REPORT No. 75 The "B" HPI pump continued to take u.ater from the BWST and pump it into the RCS as makeup and RCP seal injection and into the makeup tank as pump recirculation.
Makeup tank level and pressure were still offscale high. The makeup tank outlet valve was opened and the "A" HPI pu'ap started.
The "B" HPI pump wai stopped.
As the "A" HPI pump took its suction from the makeup tank, level and pressure decreased. The makeup tank relief valve resented, flash tank level decreased, and water discharge into the waste gas header stopped.
The Technical Support Center was m nimally manned to determine the significance of i
any offsite releases and for commur:ication to offsite agencies.
The three-hour hold of RCS pressure and temperature was completed.
The Unusual Event was terminated.
The plant was taken to cold shutdown.
Page 27
LOSS OF ICS POWER Trip Report #7.1 VI. ~ Key Operator Actions / Plant Procedures / Training Operator Actions / Operating Procedures 28 Operator Actions / Operating Procedured - Conclusions 31 Training 32 Emergency Plan Implementation 35 Emergency Plan - Conclusions 37 9
i l
l
LOSS OF ICS POWER TPIF REPORT NO. 75 VI.
KEY OPERATOR ACTIONS / PLANT PROCEDURES / TRAINING OPERATOR ACTIONS / OPERATING PROCEDURES The loss of ICS power was immediately indicated to the operators by a main panel annunciator, loss of all AUTO / MANUAL indicating 1.ghts on the Bailey control stations, and all demand meters on the Bailey stations traveling to mid scale.
RCS pressure began to increase and an operator opened a pressurizer spray valve. This action had little effect and the Reactor Protection System tripped the reactor on high RCS pressure. The spray valve was reclosed when RCS pressure began to decrease following the trip.
- After the reactor / turbine trip, the operators automatically performed Emergency Procedure Section E.01, "Immediate Actions".
The reactor and turbine were manually tripped and letdown was reduced to 40 gpm.
The operators then proceeded with Vital System Verification Status, procedure E.02.
"A" HPI injection valve was fully opened in accordance with step 10 of E.02 to counteract the pressurizer level decrease due to the normal post trip cooldown.
Operators noted a rapid decrease in Makeup Tank level and opened the "A" BWST outlet valve.
The "B" HPI pump was started to increase RCS inventory as pressurizer level continued to decrease.
This is a collateral action in step 10 of E.02.
Operators recognized the automatic start of both AFW pumps, a high AFW flowrate through the failed AFW control valves, excessive steam flow through the half-open ADVs and TBVs, and excessive post-trip cooldown from primary plant parameters.
Excessive heat transfer was diagnosed and cperators procaeded with the initial actions of E.05.
The first step of E.05 instructs the operator to use HPI injection to maintain pressurizer level above 100 inches.
The operators had already taken the first three actions in this step when they performed E.02 steps. The fourth action instructs the operator to initiate full HPI.
This was not manually perfonned but SFAS was automatically initiated a short time later.
Neglecting this action did not have any bearing on the overall events of the transient.
The second step of E.05 is a branching step for overcooling events resulting from obvious problems concerning only one of the stear generators. As both OTSGs were contributing to the excessive heat transfer, the operators' advanced to step 3.
The third step calls for isolation of both steam generators.
The first set of actions under this step is performed 'when OTSG levels 11 crease to 95% on the operate range.
Levels were increasing, but still in the startuo range during this time in the transient.
Operators proceeded to the next set of actions.
These actions isolate both the feedwater inputs and steam outputs of the OTSGs.
Since the operators had no control over the TBVs, ADVs, or AFW control valves from the control room, operators were sent to manually isolate the valves locally.
As there was also no remote control of main feedwater control valves and AFW flow had been verified, the main feedpumps were tripped.
(
Page 28
LOSS OF ICS POWER TRIP REPORT NO. 75 The operators failed to remember a recent modification to the control circuitry of the ADVs and TBVs.
This modification allows for the closing of ADVs and TBVs from
'the Remote Shutdown Panel located two floors below the control room.
The switches on this panel will close the valves regardless of the status of ICS power.
Also forgotten was a remote station for operating the ADVs.
Tha remote switches and control station were installed to allow isolation or cont"ol of the valves in the event of an evacuation of the control room.
Their alternate use for loss of the ICS control was not recognized by the operators until after the transient.
Operators attempted to find the source of ICS power loss.
No specific Casualty Procedure existed for this condition, however, operators correctly recalted the
, system for ICS power distribution.
Initially, they misdiagnosed the positico of the S1-S2 feeder switches.
Having been unable to restore ICS power, their full attention was given to arresting the transient under local control.
Within the set of instructions for isolating both OTSGs was a statement:
"If feedwater flow continues, trip appropriate feed pumps to terminate flow". A decision was made by the operators early in the transient not to stop the AFW pumps.
This decision, in part, was based on two concerns.
1.
They did not feel confident the pumps could be restarted when needed.
With the main feedwater system disabled the AFW system was the only system available for the feeding the steam generators.
2.
They did not feel that stopping the pumps could be justified to station management or the NRC following the event.
They felt this action may be construed as the intentional defeat of an actuated safety system.
The excessive steam and AFW flow resulted in automatic SFAS initiation within three minutes of the trip.
Operators closed the AFW SFAS bypass valves.
These valves are included in E.05 for isolation of both OTSGs.
Pressurizer level continued to decrease and went offscale.
RCS temperatures decreased below 500 F.
-At this temperature one RCP should have been stopped to avoid core lift concerns per Normal Cooldown Procedure, B.4.
One pump was stopped ten minutes later.
Operators isolated the ADVs and TBVs within eight minutes of the trip. Total AFW flowrate was being slightly reduced when an operator partially closed the B AFW control valve. The pressurizer started to refill and subcooling margin was increasing.
As there was adequate subcooling margin and the SFOS showed the RCS pressure-temperature plot moving into the Pressurized Thermal Shock (PTS) region, operators began to throttle HPI in accordance with E.05 step 8 and EOP Rule 2, HPI Flow Control.
Rule 2 allows for HPI throttling when subcooling margin is adequate and termination of flow when a pressurizer level of greater than 100 inches has been regained. Their first action was to open HPI SFAS recirculation valves.
Subcooling margin continued to increase.
Operators further reduced HPI flow by first throttling all four injection valves, then by closing two valves. One HPI pump was stopped as pressurizer level continued to increase.
Letdown flow was also re-established in accordance with E.05, step 8.
Page 29
LOSS OF ICS POWER TRIP REPORT NO. 75 The opening of the HPI recirculation valves increased Makeup Tank.1.evel.
Operators closed the A BWST outlet valve to shift the Makeup Pump suction from the BWST to the Makeup Tank. They neglected to open the Makeup Tank outlet valve which had closed on the SFAS actuation.
The operator that partially closed the B AFW control valve believad it was fully closed. He proceeded to the A AFW valve.
He closed it but believed it was not fully closed and left the area to find a cheater. The steam generators continued to fill past 95% on the operate range.
E.05 step 3 states that the dual drive AFW pump should be stopped at this time.
The pump was not stopped for the two reasons given earlier.
Another operator found the B AFW valve partially open and closed it the rest of the way. The first operator returned to the A valve with a cheater, placed excessive force on the handwheel, and damaged the manual valve operator. The valve reopened and the A OTSG continued to fill. The operator informed the control room that the valve had reopened and he could not close it.
He was told to.close the downstrea.n manual isolation valve.
The valve could not be moved off its open seat.
The cooldown and increase in RCS inventory had left the RCS in the reactor vessel PTS region.
Normal pressurizer spray was started to reduce the RCS pressure temperature point below the PTS curve in accordance with E.05 step 12.
A licensed operator found the S1-S2 switches in the ICS cabinets open.
He closed them and r estored. ICS power.
Operators reduced the ICS demand signals to zero.
The open AFW valve responded and the overcooling was terminated. Operators were dispatched to unisolate the TBVs and the manually closed AFW valve.
The operators, unaware that the Makeup pump was running without a supply of water, stopped the B HPI pump.
They noted a decrease in RCP seal injection flow and the pump was restarted.
The seal injection system lineup was checked and found satisfactory.
The HPI pump was stopped again with the same results.
The pump was restarted.
Control room operators heard a loud " bang" and received a low lube oil pressure annunciator on the Makeup pump.
They noted low motor currents and correctly assumed the Makeup Pump had been destroyed by lack of suction. The Makeup pump motor was tripped.
Blowdown of both OTSGs to the condenser was initiated to reduce levels.
RCS pressure and temperature were stabilized outside the PTS curve and a three hour soak period was begun in accordance with E.05 step 12.
The dayshift licensed operators i
i were called in early to provide additional assistance.
l The Makeup Tank outlet valve was opened to establish a suction supply to the makeup Pump and A HpI pump.
Water spilled from the Makeup Pump's damaged seals onto the pump room floor. Operators noted the rapid decrease in Makeup Tank level and the outlet valve was closed.
Water accumulation amounted to a couple of inches as the room's floor drain could not handle all the water.
l l
\\
Page 30
LOSS OF ICS POWER TRIP REPORT NO. 75 i
Operators were first sent to survey Makeup Pump damage and later to isolate the pump.
They could not locate high-top shoe covers and did not wear any respiratory protection equipment.
They experienced contamination of their shoes and some parts of their clothing. One operator experienced minor skin contamination of one hand.
The AFW ~ pumps were stopped and their automatic start signals bypassed.
A1other RCP was stopped to establish one running RCP per OTSG. Core Flood Tank isolntion valve power supplies were racked in and the valves closed.
Condensate system cleanup flow to the condenser was started. These steps were taken in accordance with procedure B.4, Normal Cooldown.
Half an hour after the Makeup Pump room spill, a Zone 20 fire alarm tripped the Auxiliary Building and Spent Fuel Building supply and exhaust fans. The operators diagnosed the problem when low flow was noted on Reactor Building radiation monitor R-15001.
They realized the compressor had run without suction or discharge since the SFAS initiation.
The compressor was shutdown but the smokey atmosphere created by the overheated compressor seals took time to clear.
Restarting the fans was a high priority.
Without them, radioactive gasses and particulates from the spill would accumulate in the Auxiliary Building basement.
Operators reset the fire alarm and attempted to restart the fans numerous times but the fire alarm would actuate within a few seconds and retrip the fans.
The smoke eventually cleared and the fans were successfully restarted.
The Technical Support Center (TSC) sprinkler system, which had been venting water to the TSC floor since the trip, was isolated.
Operator = noted water dripping into the switchgear room below.
Poly sheets were placed over the switchgear for protection.
Two hours after the initial loss of ICS power, power was lost again.
Operators responded within a few seconds and closed the SI-S2 switches.
Many ICS demands again initialized at the 100% level.
These were reduced to zero.
Another overcooling did not occur.
At 0705 the Makeup Tank outlet valve was opened.
Operators started the A HPI pump and the B HPI pump was stopped.
Makeup Tank level decreased and came back on scale.
Flow to the Flash Tank and Waste Gas header stopped.
The Unusual Event was terminated following completion of the three hour soak.
Operators prepared to take the plant to Cold Shutdown.
Page 31 l
1
LOSS OF ICS POWER.
TRIP REPORT NO. 75 TRAINING The training department developed specific lessons for licensed and non-licensed operators from lessons learned from the December 26 transient.
The major issues and l
topics are described below.
1.
Sequence of Events Training The objective was to ensure that all plant operations personnel have the perspective of the cause of the event and the transient which ensued.
The lessons were given in a two part format.
The first was a detailed review of the actual sequence of events and the occurrences which took place.
The second lesson involved detailed discussions of the events and actions, with emphasis on the difficulties and problems which the operators faced while mitigating the event.
The training was completed by mid-February.
l Page 32 i
L
LOSS OF ICS POWER TRIP REPORT NO. 75 i
2.
Plant Modifications The modifications covered are those to the following components and systems:
ADVs, TBVs, AFW flow control valves, and ICS and NNI power supplies.
The basic objectives of the training are:
Describe the purpose of each modification.
Describe the controls, interlocks, and operation of the modified system or component.
Describe identified failure modes and conditions.
Understand related procedurs changes and requirements.
3.
Emergency Operating Procedure Changes The changes to include plant parameter limits-in the Excessive Heat Transfer section of the ECPs will be taught to all licensed operators and STAS.
The basic objectives of the classroom training will be the description and basis for the changes and the effects on plant control.
All licensed operators and STAS will receive simulator training on the dynamics of plant response to cooldowns using the revis'ed procedures.
4.
Loss of ICS Power Casualty Procedure f
This training will present the new Casualty Procedure and will include a review of the operation of the new ADV, T8V, and AFW control valve controls.
The basic objectives for the training will be a review of the ICS power supply eystem and a description of the purpose and basis for procedure actions.
All licensed operators and STA's will receive simulator training on the use of the newly installed controls and their effect on plant response.
5.
General Procedure Changes Various Operating Procedures, Casualty Procedures, and Administrative Procedures are being changed as a result of lessons learned from the transient.
The training conducted will describe the change, the purpose for the change, and the basis for the steps.
6.
Emergency Entry into Areas of Unknown Radioactive Conditions Evaluation of the operators entry into the Makeup Pump room during the event identified the need for several changes in procedure, in an addition to a restatement of management policy.
Training will address the changes made in the requirements for health physics technicians and operators, availability and use of protective equipment, and the procedures which govern work in areas of unknown radiological conditions.
i Page 33
l 0
0 LOSS OF ICS POWER TRIP REPORT NO. 75 7.
Manual Valve Operation This lecture'will specifically cover the manual operation of the ADVs, TBVs and AFW control valves.
It will focus on management's policy that valve wrenches or " cheaters' will not be used on devices which include " mechanical advantage", such as linkages, screw jacks, or levers.
The training will include:
Description of basic components of the valve controls.
Describe basic operation of the valves, both with air and manually.
Describe the proper method manually operating the valve.
Plant tour, operating the subject valves, as plant conditions allow.
8.
Command and Control Training Changes are being made to Administrative Procedures to revise the roles of the operators during normal and abnorr.al plant conditions, but this is only one part of the training.
The Operations Department has also provided specific areas of control, mainly dealing with the Emergency Plan, that are to be covered. The training will be presented based on the following objectives.
Define the line of command in normal and abnornal situations.
Ensure the operators understand the philosophy behind the changes to the Administrative Procedures concerning their specific roles and duties.
the specific topics idencified by the Operations Department:
Provide philosophy of change in control duties.
Provide specific guidance.
Discuss how to implement the guidelines.
The Shift Supervisor is to continue to maintain an overall perspective of the event, its mitigation, and compliance with administrative process.
The Senior Control Room Operator will assume the responsibility to directly oversee the activities of the Control Room Operators. This will be accomplished by interacting with the control panel operators while monitoring the EOPs and directing the activities both in and outside of the Control Room.
The individual assigned as " communicator" for the Emergency Plan, will not be diverted from that assignment.
This is to insure commitments to effectively implement the Emergency Plan are not delayed.
Page 34
LOSS OF ICS POWER TRIP REPORT NO. 75 9.
Simulator Training Additional Topics In addition to cooldowns allowing the operators to demonstrate their proficiency with the enhanced Emergency Operating Procedures, simulator training will include the following items:
Management of HPI recirculation flow, letdown, and Makeup Tank inventory when restoring from an SFAS actuation.
HPI ano AFW throttling and trip criteria.
Pressurized Thermal Shock recovery actions.
Cooldown rate interpretation and tracking.
Conversion from auxiliary feedwater to main feedwater flow.
Differences between the simulator and Rancho Seco which can result in improper operator responses or damaged equipment.
Watchstanding principles stressing command and control training, including implementation of Emergency Plan, when applicable.
EMERGENCY PLAN - IMPLEMENTATION The loss of ICS power lead to a rapid cooldown and depressurization of the Reactor Coolant System.
ECCS initiation (SFAS) occurred three minutes into the event.
On ECCS initiation the State Office of Emergency Services (OES) received an alarm at their communications center.
ECCS initiation is an entry condition for implementation of the Emergency Plan (EP).
The Shif t Supervisor declared an Unusual Event (UE) at 0430 in accordance with AP 501 Tab 8, Abnormal Coolant Temperature / Pressure.
Tab 6, ECCS Initiation, Tab 9, Secondary Depressurization, and Tab 10, Relief Valve Failure (in this case ADV and TBV failure) also applied to the plant conditions.
The Shift Supervisor assumed the role of Emergency Coord inator. Considering the Shift Supervisor had to base his estimate of plant status on steadily worsening parameters, the EP was entered before plant conditions were stabilized or stabilizing actions had been completed.
Some of the action items in the EP were carried out.
Some were not.
i Page 35
LOSS OF ICS POWER TRIP REPORT NO. 75 i
Notifications to the state, NRC and surrounding counties began.
A Senior Reactor Operator (SRO) made the initial notifications instead of the Auxiliary Operator who had been assigned the Communicator position at the beginning of the shift.
The AJxiliary Operator was occupied with duties in other areas of the plant.
It is roteworthy that an experienced Licensed Operator, Auxiliary Operator, and Equipment 1
Attendant were on vacation this shift, although shift manning requirements were met with regard to Technical Specifications and District operating requirements. The SRO split his duties between Communicator and mitigation of the transient. There was some question about the proper forms and completeness of information.
hotification forms were not used for initial notifications.
The classification tab number was not given to the state or counties.
The Communicator had difficulty explaining plant conditions to the NRC.
The Communicator assignment was clearly not a full time assignment.
An SRO who was working dayshift reported to work early at SAM.
He became the full time communicator with the NRC in their request for an "open line".
All agencies were notified within the required fifteen minutes of the declaration of the Unusual Event.
Followup notifications were not performed with the result that the counties expressed concern that there was "too much unconfirmed information to
- ' eel confortable, yet too little information to make a decision."
Other oversights in the initial phases of the event were as follows.
1.
The emergency alarm should be sounded for ten seconds and an announcement made over the plant's public address system informing site personnel of the Unusual Event (UE).
(AP 502) This was not done.
2.
A Technical Report Coordinator was not designated.
(AP 502) This person is taken "from available Control Room Staff personnel." At 0430, when the UE was declared, all personnel were engaged in stabilizing the plant or in communications.
There was no one available for this position.
3.
The Energency Coordinator's log was not started.
(AP 502) 4.
District management, Security Watch Commander, and public information representatives were not notified of the UE.
(AP 502 and 506) The Plant Superintendent was contacted but not by the EP Communicator.
An ambulance was called when it was determined the incapacitated SRO required outsida medical assis,tance.
The requirements of AP 514, Personnel Injury, were i
met.
Security was promptly notified and supplied security officer escorts to the ambulaace team.
l l
l Page 36
O LOSS OF ICS POWER TRIP REPORT NO. 75 The Makeup Pump room spill resulted in Auxiliary Building exhaust stack radiation monitor alarms.
The state OES also received this alarm.
This event is also characterized as an Unusual Event by AP 501.
A Control Room Dose Calculation is available. in the EP (AP 509) if the EC determines this information is needed by offsite agencies. The calculation was not performed.
Followup dose calculations proved the offsite dose was not significant.
When the state OES did not receive any notification from the plant within fifteen minutes of receiving their monitor alarm, they called Rancho Seco to inquire on the status of the plant.
The Plant Superintendent arrived on site about 0550.
Once appraised of the situation, he suggested the Technical Support Center (TSC) be minimally manned for performance of offsite dose calculations and communications with outside agencies.
The EC concurred.
Appropriate personnel were sent to the TSC as they arrived for their normal dayshif t work.
At 0710 the Plant Superintendent assumed the role of EC.
The three hour soak for reactor vessel PTS concerns was completed and the Unusual Event was terminated at 0841. Tnere was some confusion during the UE closeout notifications to the state and counties.
The TSC Communicator used an internal SMUD checklist instead of a notification c)oseout form.
l Page 37 i
l
e a
LOSS OF ICS POWER TRIP REPORT NO. 75
}
I l
1 Page 38 f
LOSS OF ICS POWER Trip Report #75 VII.
Investigation of Area' of Concern s
INTRODUCTION 39 A.'
Report of Smoke Prior to Trip 42 B.
Loss of ICS Power 42 C.
Control Room Instruments Which Fail on Loss 46 of ICS Power D.
Modifications to ICS Controlled Equipment 48 E.
Pressurized Thermal Shock 50 F.
Core Hydraulic Lift 52 G.
Fuel Pin in Compression Limits 53 H.
Comparison to USAR Design Basis 54 I.
Technical Specification Limits 55 J.
Comparison of Identical Initiating Event 56 With No Operator Action for Ten Minutes K.
Post Trip Shutdown Margin 56 L.
Primary to Secondary Leak Investigation 57 M.
"A" Auxiliary Feedwater Control Valve FV-20527 58 N.
"B" Auxiliary Feedwater Control Valve FV-20528 60 0.
"A" AFW Manual Isolation Valve FWS-063 61 P.
"B" AFW Manual Isolation Valve FWS-064 62 Q.
Makeup Pump Failure 6'3 R.
Other Effects of Closing the "A" BWST Outlet Valve SFV-25003 65 S.
Filling of the Waste Gas Header 66 T.
Minimum Pressurizer Lcvel Calculation 68 U.
Pressurizer Heater Operation 69 l
V.
Response of Main Steam Relief Valves 69 l
W.
Main Steam Line Analysis 71 X.
Loss of RCP Seal Injection 72 Y.
Pegging Steam Lift of Feedwater Heater Relief Valve g
73 Z.
"A" HPI Injection Valve Position Indication 74 AA. Actuation of the TSC Sprinkler System 75 BB. Main Steam Line Failure Logic 76 l'
CC.
SPDS Versus Steam Generstor Operate
-Level Recorders.
78 D0.
Damage to Reactor Buildirg Radiation Monitor 79 EE.
Radiological Assessment af Steam Releases 80 FF.
Radiological Assessment of Auxiliary Building Releases 80 GG. Radiological Assessment of Operators' Entries into Makeup Pump Room 81 HH.
Radiological Assessment of Operation of AFW Valves 83 II.
Timely Assessment of Offsite Radiological Releases 83 JJ.
Operations / Security Interface 85 KK.
Human Factors Review 86 t
6
(
LOSS OF ICS POWER TRIP REPORT NO. 75 VII.
INVESTIGATION OF AREAS OF CONCERN INTRODUCTION In addition to collecting the various data available following an event of this type, the Plant Manager immediately instituted a deliberate and detailed process necessary to govern the investigations and troubleshooting necessary to establish the root cause(s) of the event and associated equipment failures.
An Action List was generated to collect all developing issues and odersee their resolution. The issues were given priorities.
Those with a startup priority must be resolved prior
, to plant heatup above Cold Shutdown.
Refer to Attachment 7 for the Action List.
The implemented troubleshooting program identified four major milestones to resolve an issue.
These are explained below.
1.
Troubleshooting Action Plan Following a description of the question, issue, or problem being investigated, a summary of information supporting the probable cause is developed.
Included is a review of the item's maintenance, surveillance testing, and modification history.
From this body of information, the potential root causes are developed and an outline of the troubleshooting plans to prove or disprove each is presented.
2.
Engineering Report This is a report of the results of the troubleshooting efforts and provides the conclusion and justification of the identified root cause.
t 3.
Repair Action Plan i
(
Once concurrence is obtained that the root cause(s) have been identified, then those steps to be taken to repair or modify the component or procedure for return to service are developed.
This step is imposed to insure that repairs / modifications are properly coordinated and that troubleshooting is compl,ete and sufficient to ; allo w repair.
4.
Action Item Closure Report This report consolidates all of the developed information from the above efforts and completes the explanation and justification of the root causes(s) of the item.
It also provides for developing recommendations which will be useful in the development of lessons learned and programmatic improvements to guide in achieving excellence in operations and management.
(
Page 39
LOSS OF ICS POWER TRIP REPORT NO. 75 For those items placed on the NRC/IIT Quarantineo Equipment List, or if so designated by the Action List, the entire four phase program described above was implenanted.
Several items were added to the Action List for tracking and management attention which would not have benefitted from the entire program.
In those cases a closure Report Alone was required and used to document the scope of the analysis, or investigation, and to report the conclusions which resulted from analysis of the item.
ROLE OF OUTSIDE AGENCIES On December 26, 1985, the NRC Region V Administrator sent to the District a
- Confirmatory Action Letter requiring that root cause analysis of the trip be done and that the NRC would be briefed on this root cause and provide the justification as to why the facility is ready to resume power operation, This same letter established the hold of repair work on equipment which malfuntioned pending evaluation by SMUD and NRC inspection teams.
Within two days of the transient, the NRC had assembled an Augmented Region V inspection team to the site.
The purpose of this team was to determine the sequence of events, determine the major issues and review any troubleshooting plans.
They held interviews on the " record" with operators involved with the transient and with some of the members of the plant staff management.
After three days of interviews the NRC elected to upgrade their involvement in the event investigation by replacing the team with an Incident Inspection Team (IIT).
A quarantined equipment list was developed early in the NRC investigation effort.
A fee items on the original list were removed when it was determined these items did not contribute to the loss of ICS power or the overcooling transient.
The following items remained.
1.
Integrated Control System 2.
Power to the ICS DC Power Supplies and Power Monitor 3.
Main Steam Pressure Relief Valve, PSV-20544 4.
Auxiliary Feedwater Flow Control Valves, FV-20527 and FV-20528 5.
Auxiliary Feedwater Manual Isolation Valves, FWS-063 and FWS-064 Prior to the commencement of any troubleshooting of the above components, NRC IIT approval of the plan was required. Maintenance Instructions were provided to the NRC Resident Inspector prior to any physical work on the above components to allow for NRC monitoring and observation of the troubleshooting activities The effect of having an item on the Quarantined Equipment List was that its operation was restricted to safety needs or activities specified under the direction of the approved Troubleshooting Action Plan.
No servicing, clearances, configuration changes, surveillance or other work was permitted.
Work requests which might affect this equipment were held by scheduling.
The equipment itself was identified by appropriate barriers and signs.
l l
Page 40
LOSS OF ICS. POWER TRIP REPORT NO. 75 Due to the expected industry interest in an overcooling event of this magnitude, and due to the similarities to other facilities using the ICS, the District invited the Institute of Nuclear Power Operations (INPO), the Electric Power Research Institute (EPRI/NSAC), and the B&W Transient Assessment Team (TAP) to assist with investigating the event and reporting it to the utility industry.
The INPO team spent nearly a week gathering information and developirg a Significant Event Report (SER #3-86) which was issued to the industry on January 2, 1986. This SER was superceded by SER 6-86 issued January 17.
This report is included as Attachment 5.
EPRI/NSAC assisted in the Evaluation of Pressurized Theamal Shock (PTS) questions and analysis as they pertained to the Reactor Vessel and its integrity.
The B&W TAP team assisted in compilation of the Sequence of Events and in defining issues which needed further investigation. The team played an important role in developing the troubleshooting plans for the ICS.
They also assisted in the evaluation of operator. response and procedural adequacy.
B&W engineers in Lynchburg, VP, performed an independent analysis of minimum Pressurizer level.
They provided engineering judgements as to the status of the RCS when there was no indicated Pressurizer level.
B&W also performed an evaluation of the fuel and primary system components relative to the rapid cooldown and reactor vessel PTS concerns.
Power Safety instructors at the plant simulator in Lynchburg ran the 12/26 transient with no operator action.
Their results were compared to the Rancho Seco plots of primary system parameters.
All licensed operators and STAS received simulator training prior to plant restart.
The training stressed the mitigation of cooldown transients, use of the revised Emergency Procedures, implementation of the Emergency Plan, Command and Control training, and areas where simulator response and modeling differ from the plant.
Since the ICS is generic to the B&W designed NSSS, the NRC expressed its questions which were of potentially generic interest, to all B&W plant owners and operators.
The B&W Owners Group Regulatory Response Group (RRG) met with the NRC several times and discussed these issues.
Analysis of the event concluded that the "A" steam generator was filled and auxiliary feedwater was spilled over the steam annulus and into portions of the main steam lines.
Bechtel Corporation engineers performed a stress analysis of the affected main steam line piping for determination of any detrimental effects and usage factors.
Page 41 l
l 1
l
. LOSS OF ICS POWER TRIP REPORT NO. 75 AREAS OF CONCERN A.
Report of Smoke prior to Trip '
1.
Description of Concern
'Thie issue concerns Operator's reports of smoke in the controlled area.
shortly before the loss of ICS power.
An investigation was conducted to determine any relationship between the observations and the loss of ICS power.
2.
Description of Operator's Observations At approximately 0400, an SRO entering the ventilation equipment room on
-the mezzanine level of the controlled area believed he smelled smoke.
He informed a non-licensed operator in an adjacent room.
He too entered the room and smelled smoke.
The operators spent about five minutes looking for the source but were unsuccessful. They described the smell as the result of wood or paper burning, and not electrical equipment or burning insulation.
They noted the smell disappeared when they exited the room.
3.
Results of Investigation Although the investigation did not determine the source of the smell, it can be concluded that this was not a significant factor contributing to the loss of ICS.
The ventilation room is isolated from any significant electrical power equipment.
Two walls and a corridor separate the 480 VAC electrical equipment rooms from the location where the smell was noted.
Furthermore, the electrical equipment rooms have separate heating and ventilation systems from the radwaste areas.
Smoke sensed in the ventilation room would not have originated in the electrical equipment rooms.
Thus, the report of smelled smoke is unrelated to the failure of the ICS.
B.
Loss of ICS Power 1.
Description of Concern The extended loss of ICS power resulted in a reactor trio, excessive cooldown, and steam generator overfill.
A detailed, methodical investigation was launched to find the cause of the loss of power.
The ICS, its DC power supplies, and components which fed the supplies were on the list of NRC/IIT quarantined components, f
Page 42
LOSS OF ICS POWER TRIP REPORT NO. 75 This section of the report details descriptions of the physical causes of the loss of power and repairs to prevent the recurrence of a similar transient.
It does not include operator actions, procedural adequacy, overall effect on the plant, or human factors analysis.
Reviews of these subjects can be found in other sections of this report.
2.
Description of Associated Components The origin of ICS power is two battery fed 120 VAC inverters. One battery / inverter system, the "C" supply, is vital (i.e. class IE). The other battery / inverter system, the "J"
supply is non-vital.
The "J"
. inverter was out of service at the time of the transient.
All "J"
inverter loads were being fed from the "F" 120 VAC bus which is fed from one of two 480 VAC non-vital supplies.
Refer to Attachment 12 for a simplified schematic of ICS power distribution.
AC loads in the ICS system are fed from either the "C" or "J" supplies through an automatic bus transfer device (ABT).
The ABT normally feeds all ICS AC loads with power. from the "C" inverter.
It will switch its power source to the "J"
inverter (in this case, the "F" bus) if it detects a problem with "C" inverter voltage.
The ABT is a fast transfer _
device and allows uninterrupted power to be supplied to the ICS AC loads should a failure of the "C" battery or inverter occur.
The DC loads in the ICS system are fed by four identical DC power l
supplies.
Two power supplies provide +24 VDC and two provide -24 VDC.
l One positive and one negative supply is fed from the "C"
inverter through switch S2 and the other positive and negative power supply is fed from the "J"
inverter ("F" bus in this case) through switch S1.
The power supplies are diode auctioneered to provide for the loss of one supply, but under normal operation are adjusted so they share the load.
If the "C" or "J" power feed is lost, the remaining positive and negative supply are capable of handling the entire DC load of the ICS system.
Power monitors. continually check the voltage from each power supply and the voltages of the positive and negative busses.
If a low voltage is detected on any one of them, an annunciator alarms.
Additionally, the power monitor that checks bus voltages will send a signal to open the S1 and S2 switches if it senses low voltage on either bus.
The SI-S2 j
switches are designed to have a half-second time delay before tripping j
open.
If the power monitor senses that voltage has been restors d within this time period, the switches will not trip.
Opening both switches results in a total loss of ICS DC power.
This action is taken to avoid powering the ICS circuitry from only one side of a dual power supply l
with the resultant unknown output signals to ICS-controlled components.
page 43 i
(
l LOSS OF ICS POWER ~
TRIP REPORT NO. 75 The ICS sends a voltage demand signal to each valve it controls, +1C VOC being a full open demand'and -10 VOC being a full closed demand. When ICS DC power-fails, the valves receive a zero volt signal which corresponds to 50 percent demand and they travel toward a mid position.
ICS also controls main feedpump speed.
A zero volt demand causes the pumps to decrease their speed to the minimum speed setpoint of the ICS.
Main turbine governor valve positions are changed by varying voltage into a pulser circuit.
A zero voltage results in the valves remaining fixed in the position just prior to the loss of power.
Loss of power '
also causes the main feedwater block valve relays to de-energize and the valves are driven closed.
3.
Method of Investigation prior to any physical troubleshooting, detailed troubleshooting plans were reviewed and approved by District management and the NRC/IIT.
The program for the ICS equipment investigation included the following.
a.
Visual inspections and photographs of the ICS cabinets.
b.
Insitu verification of power supply setpoints.
c.
Load carrying capability of the individual DC power supplies.
~
d.
Insitu verification of power supply monitor bus voltage setpoints.
e.
Insitu verification of S1-S2 trip delay.
f.
Verification of system response to loss of DC power.
g.
Verification of system response to repowering.
h.
Bench test of bus power supply monitor.
i.
Remove and replace defective wiring for further troubleshooting.
j.
Verification of operability of 120 VAC power to ICS.
4.'
Results of Investigation Early studies of the potential causes of the ICS loss of power eliminated the independent, redundant 120 VAC power supplies.
Operator reported that the first annunciator to alarm was the loss of ICS power wi nd ow.
There were no breakers found tripped on the E, F, or J busses.
The ABT did not transfer so it can be concluded that the C bus power was not interrupted.
Further support to this conclusion is given in noting that the second loss of ICS power, at 0615, gave the same indications.
page 44
LOSS OF ICS POWER TRIP REPORT NO. 75 Troubleshooting plans then centered around the device that opens both S1 and S2, the bus power monitor. Early findings indicated the monitor was exhibiting drift. Understanding this phenomena was enhanced with the receipt of a new power monitor. This monitor was set up in a laboratory where a full set of performance data could be obtained and ccmpared to the installed monitor. It was found that small amounts of resistance, as little as one ohm, in series with the sensed voltage could cause less than optimum performance of the device and could result in the generation of a trip-signal.
This finding led to the troubleshooting of the voltages and resistances within the ICS DC distribution cabling. A resistance measurement in excess of two ohms was found in the auctioneering panel +24 VDC inter-cabinet wiring. The high resistance was traced to a connection on the cabinet #1 +24VDC bus bar. The crimp on this wire lug was so loose the lug fell off when the wire was disconnected from the bus.
This led to an inspection of the other crimped lugs in all ICS cabinets.
Several more factory cable lug crimps were found poorly made.
The size of the bolts securing the lugs to the bus bars were found too small for the bus bar holes.
The wiring layout from the power supplies to the power supply monitor was reviewed.
The power monitor module receives its supply power from cabinet #3 bus bars through a daisy chain of wire wraps stringing many ICS modules in the same row together.
These series connections were seen as a possible source of resistance.
A change of the wiring design was made.
Analysis of the trip timing of the SI-S2 switches revealed that the delay before tripping was.15 seconds instead of the acceptable range of i
.2 to.8 seconds.
The shorter delay made the trip slightly more sensitive.
5.
Root Cause l
The root cause of the loss of DC power has been identified as the loose l
power connection on the cabinet #1 bus supplying power to the bus power supply monitor.
6.
Corrective Actions i
l The factory bus to bus wiring has been replaced with current nuclear standard wiring.
Larger bolts which fit snuggly into the bus bar perforations were substituted.
Investigations into the power monitor's l
sensitivity to input resistance are continuing.
The power supply l
monitor and SI-S2 switches have been removed for subsequent analysis by an independent laboratory.
Replacement components meeting operational requirements have been installed in place of these removed items.
The Page 45
LOSS OF ICS POWER TRIP REPORT NO. 75 4
power monitor module has been wired directly to the auctioneering bus bars to eliminate the possibility of high resistance in the many daisy-chain connections.
A study showed that there has been 160 service years of successful operating history on this type of power supply monitor, and that this one is the only one to have tripped an ICS as a result of a failure.
In addition, this failure was external to the monitor itself.
Having replaced the power distribution wiring and connected the power monitor module directly to the auctioneering bus, the likelihood of a recurrence is remote.
On this basis, there is no need for a modification to the design of ICS bus power monitoring.
C.
Control Room Instruments which Fail on Loss of ICS Power 1.
Description of Concern Following the transient, testing was performed to determine the response of control room controls and instrumentation to a loss of ICS OC power and subsequent repowering.
This study was undertaken to verify actual response to that predicted.
Information gained from these tests were used to further understand the 12/26/85 transient and in the writing of the Loss of ICS Casualty Procedure.
2.
Method of Analysis ICS DC power was interrupted by reducing the positive or negative bus voltage, causing the power monitor to trip switches S1 and S2.
All ICS Bailey station demands were recorded before the loss of power, during the loss of power, and af ter the re-energization.
Personnel were stationed at ICS-controlled valves to observe their reaction to the loss j
of power and subsequent repowering.
During one test run, the l
de-energization period was matched to the actual time the ICS was i
without power on December 26.
A video camera was used to record the response of all control room instrumentation, controls, and alarms before, during, and following the loss of power.
This recording was given a detailed review to identify any anomalies which were not apparent from operator statements or the plant transient response.
Page 46
LOSS OF ICS POWER TRIP REPORT NO. 75 3.
Results of Analysis LOSS OF ICS DC POWER ACTUAL COMPONENT / INSTRUMENT INDICATION POSITION A&B FW Pump, Demands 50%
N/A A&B FW Main Feed Valves 49%
50%
A&B FW Startup Feed Valves 53%
50%
A&B Turbine Bypass Valves 49%
55%
A&B Atmospheric Dump Valves 49%
50%
A&B AFW Control Valves 48%
50%*
A&B FW Flowrate Recorders 3.1E06 lbm/hr**
N/A Auxiliary Steam Demand 50%
Unknown
- Valves under clearance.
Predicted actual position from input to E/P converter.
- Alternate Main and Startup Feedwater flowmeters are located on the same panel.
These are not affected by ICS.
RE-ENERGIZING ICS DC POWER A&B FW Pump Demand 100% decreasing to N/A 0% within 1 minute A&B FW Main Feed Valves A:24% B:100%
A:25% 8:100%
A&B FW Startup Feed Valves 100%
100%
A&B Turbine Bypass Valves A:90% B:0%
A:10% B:0%
A&B Atmospheric Dump Valves 100%
100%
A&B AFW Control Valves A:100% 8:0%
A:100% B:0%
A&B FW Flowrate Recorders Actual Flowrate N/A Auxiliary Steam Demand 0%
Unknown All Control Stations initialize in MANUAL mode.
- The A TBV had a 0% demand when the test was run a second time.
All other demands agreed with the initial test.
The de-powering and repowering test was run twice.
The ICS responded identically in both cases with the exception of the demand signal to the A TBV when repowering.
Based on operator statements and plant response, the results obtained by de-energizing ICS DC power agree with the transient sequence of events.
One control station unexpectantly responded to the loss of power.
The main steam to auxiliary steam reducing station failed to midscale.
After researching the electrical drawings, this station was found to be powered from the ICS cabinets.
Page 47
LOSS OF ICS POWER TRIP REPORT NO. 75 Based on operator observations during the repowering of ICS on 12/26/85, all ICS demands were expected to go to 100% during the test when power was restored.
Instead, some demands went to different values, and then proceeded to change during the first minute af ter power was applied.
It was concluded that a precise prediction of ICS response to repowering could not be determined without substantially increasing the scope of the test program.
As this issue is a generic concern to all B&W plants, the B&W Owners Group I&C Committee agreed to task B&W with development of a program to investigate the characteristics of repowering.
The repowering open item does not prevent the operator from effective control of the plant.
Controls independent of ICS power have been added to the ADVs, TBVs, and AFW control valves.
When ICS power is regained, the valves will be receiving demands from their alternate controllers.
Per procedure, operators will not switch demand signals back to the ICS until any transients in the ICS have settled out and ICS demands are aligned to the alternate controller demands.
This action provides for a smooth transfer back to ICS control.
Additional instruments and indications were seen to fail on loss of ICS power, although their failure had no direct effect on the 12/26/85 transient.
These are listed below.
Unit Load Demand Station Steam Generator / Reactor Master Station Reactor Demand Station Delta Tcold Station Turbine Header Pressure Setpoint Station Generator Frequency Meter (HISS)
All RUNBACK indicating lights (HIRI) are extinguished.
All LOAD LIMIT indicating lights (HIRI) are extirguished.
l D.
Modifications to ICS Controlled Equipment 1.
Description of Concern All ICS controlled velves' demand signals went to 50 percent when power was lost. Operators were sent to locally isolate eight of the valves contributing to the overcooling.
When ICS power was lost, the operators lost their only method for con. trolling these valves' positions from the l
Page 48 1
F 0
LOSS OF ICS POWER TRIP REPORT NO. 75 control room.
Automatic or backup methods of control in the event of a loss of ICS power were investigated. These modifications were not intended to prevent a plant trip.
They were intended to allow the operators easier, quicker methods of controlling valves critical to steam generator heat transfer independent of ICS power following the trip.
2.
Description of Modifications A great many different designs were considered for the Turbine Bypass Valves (T8Vs), Atmospheric Dump Valves (ADVs), and Auxiliary Feedwater (AFW) control valves.
Many of the alternatives were rejected based on engineering or human factors judgements.
Other B&W nuclear plants'were polled for the failure modes of their valves.
The final designs are summarized below.
a.
TBVs and ADVs. Override switches for these valves have been mounted on console H1RI, adjacent to the normal controls for the valves, where all relevant parameters are readily available.
The switches will have three positions.
The CLOSE position will close the valves regardless of the ICS demand signal being sent to them and the status of ICS power.
The NORMAL position will allow the valve to respond to the demand signals from the ICS control station unless there is a loss of ICS power.
If power is lost, an independent close signal will be sent to the valves.
This will be the normal switch position selected during power operation.
The AUTOCLOSE DISABLE position will allow the ICS station demand signal to be fed to the valve and will not close it on loss of ICS power.
~
This feature was included to allow the operator some gross control of steam generator steaming capabilities from the control room in i
the event of loss of ICS power, rather than relying solely on main steam relief valves.
For prolonged losses of ICS power, the ADV positions can be modulated with the use of the remote station in the turbine plant, grade level.
b.
AFW Control Valves.
Two hand / auto controllers have been added to control room vertical panel H2PS.
Normally, these controllers will be selected to the AUTO mode.
This mode allows control signals from the ICS controller to pass to the AFW control valve, unless a loss of ICS power is sensed.
On loss of power the new controller will position the valve to a preset opening.
The operator can modulate the control valve by placing the new controller in the MANUAL mode and varying the demand signal.
The power supply to the new controller ard its associated components is independent of the ICS power supplies.
Startup Level indications for A and B steam generators have also been added beside the new controllers on H2PS to allow the operator at the controllers to monitor the effect of AFW feed rate without leaving the H2PS area.
Page 49
LOSS OF ICS POWER TRIP REPORT NO. 75 On loss of ICS power the AFW control valves' positions will open to a preset 21%.
This setpoint yields a flowrate of 280 gpm per OTSG at an OTSG pressure of 1050 psig and both AFW pumps running.
According to a B&W calculation, a total flow of 560 gpm' is sufficient to cool the RCS from a trip at any power level without violating any minimum subcooling requirements.
As the decay heat output of the core decreases the operator will take manual control of AFW flowrate if ICS power is not regained to prevent an excessive cooldown.
At the opposite end of the spectrum, a loss of ICS power with the plant at hot shutdown conditions with no decay heat gives the operator 1.6 minutes to respond before RCS temperatures decrease below the Interim Brittle Fracture Limit of 500 F.
The analysis assumed the OTSGs are depressurized to 200 psig.
E.
Pressurized Thermal Shock 1.
Description of Concern The transient resulted in operation outside the optimum operating band in the Excessive Heat Transfer curves of the Rancho Seco Emergency Procedures.
Operation to the left of this curve (lower temperatures with the same, or higher pressures) places the RCS in the Pressurized Thermal Shock (PTS) region.
Operation in this region, by itself,.does not mean damage was done to the reactor vessel or coolant system but does require evaluation prior to continued operation.
2.
Method of Analysis The transient of 12/26/85 was initially compared to another rapid cooldown transient which occurred 3/20/78 at Rancho Seco.
The previous transient is sometimes referred to as the " Light Bulb Incident".
Babcox and Wilcox performed a fracture mechanics analysis to address the
'TS concern.
This computer program calculates the RCS pressure which would cause a reactor vessel flaw to propagate unacceptably, given the time / temperature history of the transient. The program analyzes crack propagation for cracks ranging in size from 1/40 to 1/4 the thickness of the vousel wall.
A fluence of 32 EFPY was used in the calculation.
This overestimates the current fluence.
The vessel material properties were teken from the B&W report BAW 1895 performed previously. The analysis was done based on the ASME Code,Section XI, Appendix A.
B&W also reviewed the transient data with regard to its effects on the structural integrity of the other RCS components.
This involved a comparison of the 3/20/78 temperatures, pressures, and cooldown rate with the 12/26/85 transient.
B&W had previously completed a detailed analysis of components following the first transient and found that the i
Page 50 r
i l
)
1 LOSS OF ICS POWER TRIP REPORT NO. 75 i
stresses and usage factors remained within the allowable limits of the ASME Code.
The Electric Power Research Institute (EPRI) assessed the impact of the,
transient on the reactor vessel to help resolve the PTS concern.
Their assessment was based on a draft appendix to the ASME Code,Section XI.
This method calculates a temperature margin which assures ductile behavior of reactor vesr el material. The draf t Appendix was designed to be applied if a plant violated its NOT Tech Spec curve established in accordance with ASME Code,Section III, Appendix G.
Rancho Seco did not violate the NOT curve during the 12/26/85 transient.
However, the 100 F/hr cooldown rate basis for the curve was violated.
The draft Appendix assumes a.400 F/hr Cooldown rate.
EPRI evaluated rates as high as 1200 F/hr.
Although the transient had cooldown rates approaching 1200 F/hr, this was for a short period.
A more reasonable rate is estimated at 300 F/hr below 500 F.
Actual vessel fluence was used for the calculation.
3.
Results of Analysis The comparison of the two rapid cooldown transients produced the results tabulated below.
3/20/78 12/26/85 Lowest RCS Temperature 280 F 386 F Corresponding RCS Pressure 1900 psig 1413 psig Obviously, vessel fluence and cooldown rate must be taken into account to fully analyze the stresses on the reactor vessel.
The B&W computer program found that the critical pressure for unacceptable crack propagaticn was above 2750 psig.
Since the pressure did not even approach this value during the 12/26/85 transient, there is no predicted unacceptable flaw propagation.
The EPRI ASME evaluation procedure allows demonstration of adequate structural integrity of the reactor vessel beltline without performing i
further integrity analyses as long as the reactor coolant pressure did I
not exceed design pressure (2500 psig) and the difference between RCS cold leg temperatures and the nil-ductility transition temperature had not been less than 55 degrees during the transient.
The transition temperature was calculated to be 217 F.
The minimum cold leg temperature was 386 F., leaving a difference of 169 degrees. The EPRI study concluded that the Ranc9o Seco reactor vessel beltline region has adequate structural integrit? for return to service without further evaluation.
Page 51 l
i
LOSS OF ICS POWER TRIP REPORT NO. 75 D&W also evaluated the effect of the pressure transient on the balance of RCS components, including the steam generator tubes.
A comparison of the 3/20/78 transient and the 12/26/85 transient showed that the two transients were not substantially different from one another with reference to component stress and fatigue.
It was concluded that the latest transient was bounded by the earlier transient.
Based on their analysis, B&W concluded that the structural adequacy of the RCS y
components has nit been impaired.
The Rancho Seco Reactor Coolant System was designed to accommodate 240 normal cooldowns at 100 degrees per hour.
There have been several~
transients during which this cooldown rate was exceeded and therefore B&W was asked to deterrine the cumulativ'w fatigue usage factor.
This evaluation concluded that the allowable nu'mber of remaining cooldowns, at 100*/hr should be reduced from 240 to 235.
Since only 31 cycles of this transient have been used to date, the reduction to 235 allowable cycles is expected to have no adverse impact on the current design life of forty years.
F.
Core Hydraulic Lift 1.
Description of Concern All four reactor coolant pumps were operated longer than is allowed by normal operating procedures.
One RCP should have been stopped prior to cooling down below 500 F.
The pump was not stopped until ten minutes later when RCS temperatures had decreased to 410 F.
Operation with four RCPs at colder temperatures could result in excessive core lift forces from the denser water pushing the bottom of the fuel assemblies in an upward direction.
This action could cause rubbing between adjacent fuel rods and spacer grids and lead to fuel assembly wear or damage.
This requires evaluation prior to continued operation.
i 2.
Method Of Analysis Babcox and Wilcox ran two analyses, one employing a "best estimate" model and the other a " bounding" model.
The first method uses a density compensated nominal RC flowrate using realistic pump head capacity curves and the core ir.let flow factor distribution obtained from Vessel Model Flow Tests (VMFT) at the Alliance Research Center (ARC).
The latter method assumes an upper limit value for RC flowrate and the application of the largest bundle inlet flow factor obtained from VMFT tests at ARC.
j 3.
Results of Analysis 1
Evaluation of core conditions during the ten minute period of four pump operation below 500 V using the "best estimate" model shows all fuel assemblies are predicted to have remained seated in their lower grid support s.
Therefore, no fuel assembly wear or damage is expected.
Page 52
LOSS OF ICS POWER TRIP REPORT NO. 75 Using the " bounding" model resulted in the prediction of 21% (37 assemblies) having a net upward lift force.
Upon further analysis, three conclusions were made.
1.
Fuel assembly holddown spring stresses were predicted to be within the design limits.
2.
The calculated maximum lift force resulted in a maximum lift of 0.437 inches for the worst case assembly.
The engagement of the lower end fitting of the fuel assembly to the lower grid pads on the lower grid plate is more than 1.4 inches.
It is concluded the assemblies did not disengage from the lower grid structure.
Assembly }ateral position and spacing were maintained.
3.
Flow induced vibration during the uplift mode is not expected to be greater than vibration during normal power operation with four RCps.
This was confirmed by previous tests run at ARC.
In summary, the " bounding" model analysis coupled with further analyses found that no fuel assembly wear or damage was caused by the thermal hydraulic lifting forces.
G.
Fuel Pin in Compression Limits 1.
Description of Concern l
l Due to the depressurization of the RCS during the first half hour of the t
transient, Fuel pin in Compression Limits may have been violated.
This may lead to weakened fuel cladding and, at worst, small fissures through the cladding leading to release of fission matarial into the reactor coolant.
2.
Basis of Limit During normal power operation, the internal fuel pin pressures are less than RCS pressure.
If the plant trips and is subsequently depressurized shortly thereafter without giving time for the fuel pellets and clad gap gasses to cool, internal pin pressures can exceed RCS pressure and result in fuel rod hoop stresses.
The free hydrogen which is liberated l
by small amourts of clad zircalloy reacting with water at high l
temperatures during power operation forms zire hydrides along I
circumferential grain boundaries if the rod is in compression.
If the I
rod is in tension the hydrogen migrates along radial grain boundaries.
l The hydrides oeaken the zircalloy in this area and fissures can form l
along the radial grain boundaries.
Keeping the RCS at higher pressures I
during the initial cooldown period ensures the rods remain in compression.
Page 53
LOSS OF ICS POWER TRIP REPORT NO. 75 3.
Results of Analysis B&W has concluded that Fuel Pin in Compression Limits were not violated during the transient.
Although RCS pressure was rapidly reduced, the coinciding overcooling reduced fuel pin pressures in step with the pressure reduction.
Fuel rod integrity was maintained and is acceptable for continued operation.
H.
Comparison to USAR Design Basis Transients:
1.
Description of Concern Althouch a loss of ICS power and its effects on the plant are not specifically discussed in the Updated Safety Analysis Report (USAR),
rapid cooldowns are.
Although the initiating events may differ, the results on the plant are the same.
The USAR was reviewed to ascertain if the 12/26/85 transient was bounded by transients evaluated in the USAR.
2.
Results of Analysis USAR Section 14.2.2.1 describes the accident analysis for a main steamline failure.
The accident postulated by the USAR is a guillotine failure of a 36 inch steamline.
The accident and 12/26/85 transient are appropriate to compare against one another because they both produce dramatic cooling effects on the RCS and results in offsite releases of steam to the atmosphere.
An unisolatable steam leak did occur for the initial minutes of the transient until an operator isolated the ADVs.
Page 54
I LOSS OF ICS POWER TRIP REPORT NO. 75 COMPARISON Main Steamline Failure Analysis Assumptions and Results 12/26/85 Event 1% Defective Fuel Rods 0% Defective fuel rods.
Complete severance of 36" pipe.
6 Code safety valves lifted.
Duration of release - 4.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />.
<1 min. 2'ADVs and 4 TBVs 50%
open for <10 min.
Break between Reactor Building Open valves located in same and turbine stop valve.
portion of main steam system, 490 F reached in 43 seconds.
490 F reached in 6-7 minutes.
10 gpm primary to secondary
<0.01 gpm primary leakage.
to secondary leakage.
Site area thyroid dose - 12.7 rem.
Site area thyroid dose - 0.0 Whole body dose - 46 mrem.
mrem.
Whole body dose - 0.02 mrem.
The consequences of the 12/26/85 event are bounded by the USAR analysis.
I.
Technical Specification Limits 1.
Description of Concern The December 26 event was reviewed for any Technical Specification limit violations.
2.
Results of Investigation The RCS rapid cooldown exceeded Technical Specification 3.1.2 limits.
i The curve associated with this section limits the RCS cooldown rate to i
100 F/ hour.
During the December 26 event the RCS temperature dropped from approximately 582 F to 386 F in 25 minutes.
This rate violated the limit.
The RCS Pressure-Temperature Cooldown Curve was not exceeded.
The District generated a Licensee Event Report (LER) summarizing the event.
Included in this LER was the notice of Technical Specification violation.
The report was forwarded to Region V Office of Inspection and Enforcement on January 24, 1986.
A copy of this report is included as Attachment 6.
A full discussion of the significance of having exceeded this limit, and j
a related phenomenon, pressurized Thermal Shock (PTS) is included in l
paragraph E of this section.
I page 55
LOSS OF ICS POWER TRIP REPORT NO. 75 J.
Comparison of Identical Initiating Event with No Operator Action for Ten Minutes 1.
Description of Concern Operators performed numerous actions within the first ten minutes of the transient.
These actions may have significantly changed the overall plant response.
The NRC/IIT requested this analysis.
2.
Method of Analysis The Sequence of Events was reviewed and all operator actions were deleted.
Engineering judgement was then used to predict the plant response.
A qualitative response was generated.
Power Safety International instructors simulated a loss of ICS power at the, plant simulator in Lynchburg, Virginia.
The transient was allowed to progress without operator action for ten minutes.
Plots of important primary and secondary parameters were printed on a multipoint recorder.
These values were compared against those experienced during the December 26 event.
A summary of the major events of the simulator run is included in Attachment 15, 3.
Results of Analysis The reactor trip, automatic SFAS actuation, and rapid cooldown, would have occurred with or without operator action.
SFAS initiation would have occurred slightly earlier due to less inventory makeup from the manually started "B" HPI pump and open "A" HpI injection valve and continued steam flow through the idling Main Feedwater Pumps.
The simulator run showed subcooling margin was lost for a short time until pressure decreased to the point where the Core Flood Tanks dumped some of their inventory into the RCS.
The steam generators were filled more l
quickly due to the open SFAS bypass valves.
The simulator cooled 80 degrees below the Rancho Seco event within the first ten minutes.
Operator action to increase HPI flow within the first few minutes following the trip maintained adequate subcooling margin.
The closing of the AFW SFAS bypass valves shortly after SFAS initiation and the closing of the ADVs and TBVs within ten minutes reduced the cooldown rate.
i K.
Post Trip Shutdown Margin l.
Description of Concern As the Reactor Coolant System rapidly cooled following the loss of ICS power, the negative moderator coef ticient added positive reactivity.
A minimum one percent shutdown reactivity must be maintained following a plant trip.
Page 56
LOSS OF ICS POWER TRIP REPORT NO. 75 2.
Method of Analysis Actual shutdown calculations were performed at various temperatures from hot shutdown conditions to 386 F.
The calculations took only power deficit, control rod position, and moderator temperature coefficient into account.
Negative reactivity effects from the addition of thousands of gallons of highly borated water from the BWST was not counted.
3.
Results of Analysis The reactor core was always at least two to five percent shutdown during the cooldown.
This is a conservative calculation, especially at lower temperatures when the RCS boron concentration had increased significantly.
L.
Primary to Secondary Leak Investigation 1.
Description of Concern The thick OTSG shell did not cool as quickly as the OTSG tubes during the rapid cooldown transient.
This resulted in tensile stresses on the OTSG tubes and could have led to a weakening of the tubes or primary to secondary leaks.
In addition, cool auxiliary feedwater being sprayed onto the peripheral tubes increased the stress on these outer tubes.
Prior to the event, the condenser air ejector radiation monitor was indicating approximately 400 cpm of activity.
This is a low value, but is above background and may have been indicative of a small pre-existing primary to secondary leak.
Following the event the radiation monitor indicated from 100 to 300 cpm.
Mein steam line radiation monitors showed no change throughout the event.
However, due to the OTSG thermal transient it was deemed prudent to monitor the OTSGs, while in wet l
layup, to evaluate and quantify identifiable leakage.
2.
Results of Evaluation Samples of secondary side OTSG water were analyzed.
These results were-inconclusive as some minute traces of primary radioisotopes were found, so a more sensitive technique was applied.
Helium leak detection did identify two very small leaks in the B OTSG.
Investigations employing eddy-current techniques were used to characterize the leaks and to provide confidence that any incipient leaks in either 01SG are found.
l At the time of this writing, eddy-current tests were completing on the B OTSG and beginning on the A OTSG.
A 21% random sample of 8 OTSG tubes was run.
Tubes of special interest was also tested in both steam generators.
In addition to the two leaking tubes found by the Helium leak check, 53 tubes were identified as having greater than 40% through wall imperfections.
These 55 tubes will be stabilized and plugged.
Page 57 l
~
LOSS OF ICS POWER TRIP REPORT NO. 75 Results of A OTSG testing gave indications that 12 tubes will require plugging.
l M.
"A" Auxiliary Feedwater Control Valve FV-25027 i
1.
Description of Concern This valve received a 50% demand signal when the ICS power failed.
It reopened while an operator was attempting to manually close it using a 4
valve wrench on its handwheel.
Evaluation of computer data following the transient showed that the valve was fully closed before the operator applied force with the valve -
wrench. The manual operator was damaged.
The investigation was to determine the root cause of the failure of the manual operator. This component was an NRC/IIT quarantined component.
l 2.
Description of Component I
The ICS controlled AFW control valves are four inch, air operated, modulating valves which employ balanced valve guiding through a linear cage to counteract large stem forces during high flowrates.
Stem travel from full open to full close is two inches.
The air actuator is an air-to-open, spring-to-close type.
The positioner is a reverse loading voltage-to-pneumatic (E/P) converter.
The E/P converter works on a
+/-10 VDC signal originating in the ICS A top connected diaphragm actuator positions the~ valve stem.
Increasing air pressure above the i
diaphragm will move the stem in the close direction.
A zero volt ICS signal results in a 9 psig air signal to the controller which positions the valve half open.
Attached to the side of the valve yoke is a manaal handjack.
The handwheel assembly is fixed in place to the yoke by two dowel pins.
Two J-bolts hold the bottom of the assembly against the yoke while one U-bolt holds the top of the assembly against the yoke.
During manual l
operation, the dowel pins take the majority of the force, that being in j
the shearing direction.
The handwheel assembly has a center, neutral l
position which allows full valve travel by the air operated positioner. -
l Clockwise rotation of the handwheel limits the amount of stem travel in l
the open direction and will eventually compress the actuator spring and l
begin to close the valve.
As the operator continues closing the valve, the position feedback signal to the positioner decreases the air supply to the top of the diaphragm in an attempt to allow the spring to force the valve to its demanded open position.
As the valve is closed the resistance of the handwheel would increase.
The handjack assembly has the proper gear ratio to compress the spring until the valve is fully closed with no assistance from air pressure on top of the diaphragm.
Counterclockwise rotation will open the valve against air pressure on the top of the diaphragm.
Page 58
LOSS OF ICS POWER TRIP REPORT NO. 75 Local position indication utilizes a fixed vertical scale with an arrow pointing in the open direction. The closed position is not marked on the scale.
A pointer is not utilized on this position indication method. Two horizontal disks attached to the valve stem are used to couple the remote positioner feedback arm to the valve stem.
These disks, with approximately 1/8" spacing between them, are used as the pointer for the local indication method.
This type of local position indication is used widely throughout the plant on air operated valves.
3.
Method of Analysis Prior to the commencement of any troubleshooting activity, a written work plan was submitted and approved by District management and the NRC/IIT.
First the "as-found" conditions of the manual operator were documented, including photographs, and measurements were taken.
A list of probable causes for the failure was generated from the initial investigation.
a.
Excessive force on the handwheel assembly, b.
Improper mounting bolt torque, c.
Improper positioning of the handwheel assembly on the operator yoke.
The handjack assembly was then removed, disassembled in the shop, and inspected.
A review of the maintenance, surveillance testing and modification history was also conducted.
4.
Results of Analysis A review of the "as found" condition of FV-20527 and handwheel assembly revealed evidence that excessive force was applied to the handwheel assembly.
Broken dowel pins, elongated dowel pin holes, bent lever pivot pin, pivot bushing cracks, and handwheel assembly casting cracks substantiate the use of excessive force.
The dowel pins first elongated, then the pins sheared.
The valve spring forced the handwheel assembly to move up the yoke 1-5/8 inches, allowing the valve to pop back open.
When ICS power was restored, the handwheel assembly moved 5/8 inches down the yoke.
An operator placed the assembly back in the neutral position following the restoration of ICS power.
This scenario
- agrees with the "as-found" condition of the handwheel assembly.
The force developed by the handwheel assembly was at least 7360 lbf to cause failure of both dowel pins.
Since the spring force opposed the handwheel assembly force, the resultant force on the valve stem and valve seat was 5490 lbf.
The stem and valve seat was not damaged as their maximum allowable load is 28,000 lbf per Fisher Controls Company.
i The J-bolts holding the handjack to the valve yoke were found to be loose.
The U-bolt nuts were found not fully seated against the actuator. This does not appear to have contributed to the failure of 1
l the handwheel assembly.
Since the dowel pins were sheared, the Page 59
O e
i LOS3 OF ICS POWER TRIP REPORT NO. 75 handwheel assembly was against the valve yoke at the time of assembly failure.
The dowel pins, locating the handwheel assembly to the operator yoke, ensure proper positioning of the handwheel assembly on the operator yoke.
Newly chipped flakes of paint on the FV-20527 operator showed the handwheel assembly had moved as a result of the shearing of the dowel pins. The handwheel was in the proper position prior to the 12/25/85 transient.
The design of the local position indication may have contributed to the overstressing of the handjack assembly.
The 1/8 inch between the disks and lack of an absolute closed indication may have led the operator to believe the valve was only eighty percent closed when, in fact, it was fully closed.
In addition, the area lighting slightly increases the difficulty in determining actual valve position at night.
5.
Root Cause The results of the evidence indicates the following root causes contributing to the failure of FV-20527:
a.
Excessive force on the handwheel assembly, b.
- Training, c.
Valve stem position indication method.
d.
Area lighting enabling the operator to see the indicator.
N.
"B" Auxiliary Feedwater Control Valve FV-20528 1.
00scription of Concern The "B" AFW control valve was manually closed during the transient.
The first operator arriving at the valve only partially closed it but believed it to be fully closed.
The second operator completed closing it.
The same concerns which applied to the "A" AFW valve were applied to this valve, i.e. possible overstress of the manual operator and lack of positive closed indication.
This valve was also on the NRC/IIT list of quarantined items.
2.
Description of Component The "B" AFW control valve is identical to the previously described "A" AFW control valve.
3.
Method of Analysis The method of analysis for the "B" valve was the same as that for the previously described "A" valve.
As there was no failure of the handjack on this valve, there is no list of probable root causes, i
page 60
e LOSS OF ICS POWER TRIP REPORT NO. 75 4
Results of Analysis Both J-bolts were found loose and the U-bolt had dropped down approximately 3/16 inch in the rear.
It is believed this condition existed prior to the transient. The handwheel assembly levers were found slightly bent where the levers attach to the stem.
This condition also occurred prior to the transient.
No valve or handwheel parts were overstressed by the handwheel assembly and the handwheel assembly remained functional.
O.
"A" AFW Manual Isolation Valve FWS-063 1.
Description of Conct<n Following the failuva of the "A" AFW control valve handjack, the operator was told to close the downstream isolation valve, FWS-063.
The operator could not move this valve from its fully open position, even with the aid of a valve wrench.
2.
Description of Component FWS-063 is normally locked open, manually operated, rising stem Velan pressure seal forged gate valve.
It is six inches in diameter and rated at 900 psig.
The valve is located in an outdoor environment.
This component was an NRC/IIT quarantined item.
3.
Method of Analysis A corrective maintenance history search was conducted for FWS-063 and similar Velan valves.
There was no preventive maintenance performed on these valves.
Although there had been no corrective maintenance i
performed on FWS-063 since its initial installation, two similar valves, FWS-064 and FWS053, have had failures of the yoke not bearings, causing the valves to stick in the open position.
Based on the maintenance history of similar valves and vendor recommendations a list of probable causes for the FWS-063 failure was generated.
1 a.
Failed yoke nut bearing, b.
Valve packing too tight.
c.
Lack of stem / bearing / packing lubrication.
d.
Failed valve internals.
Prior to commencement of any troubleshooting activities, District management and NRC/IIT approval was given to the troubleshootirg plan.
The yoke nut assembly was disassembled and inspected.
l
(
Page 61 1
i
LOSS OF ICS POWER TRIP REPORT NO. 75 4.
Results of Analysis The yoke nut assembly required considerable force to free during disassembly. The upper and lower yoke nut assemblies and all races were found rusted. There was no lubrication in the yoke nut housing. The valve stem was found free to move internally and there was no stem thread damage. The yoke nut assembly was cleaned, lubricated and reassembled. The valve then stroked smoothly.
5.
Root Cause FWS valve failure was due to lack of yoke nut assembly lubrication causing the assembly to seize.
Failure of the yoke nut Searing was a result of the lack of lubrication.
Investigations of five valves similar to FWS-063 showed all lacked recent lubrication.
Reviews.of the existing preventive naintenance'(PM) program showed that these valves were not on the PM schedule.
In recognition of the desirability of having certain manual valves' readily operable, the Nuclear Operations Manager has identified a list of approximately 100 valves which will be verified operable prior to resumption of power operations.
These manual isolation valves will be characterized by their purpose and need to allow isolation of important active equipment such as pumps, valves, and heat exchangers.
They will be selected to include both primary and secondary plant systems.
Functions, not class will be the criteria.
The program will involve actual stroking of the valve and necessary adjustments and servicing.
Statistics will be collected and evaluated to determine a summary status of valves in similar service.
P.
"B" AFW Manual Isolation Valve FWS-064 1.
Description of Concern Valve FWS-064 is identical in design and service to FWS-063.
Since FWS-063 could not be moved off its open seat during the transient, it was judged prudent to check FWS-063 for a common mode failure.
FWS-064 was not operated during the event.
This component was on the list of NRC/IIT quarantined items.
2.
Description of Component Valve FWS-064 is identical in design and service to FWS-063.
See the previous description for FWS-063.
3.
Method of Analysis Prior to any troubleshooting actions, a written troubleshooting plan was reviewed and approved by District management and the ARC /IIT.
The valve was stroked and inspected with measurements of breakaway and closing torques being recorded.
Page 62
o o
LOSS OF ICS POWER TRIP REPORT NO. 75 4.
Results of Analysis The valve stroked smoothly with a minimum of applied torque.
Four other similar Velan valves were stroked.
All stroked smoothly.
Q.
Makeup Pump Failure 1.
Description of Concern The Makeup Pump was severely damaged due to a lack of suction.
Both the physical root cause and mental error which resulted in the isolation of the suction supplies must be thoroughly understood so appropriate actions can be taken to prevent similar occurren:es.
The Makeup Pump was not included in the list of NRC/IIT quarantined components.
2.
Description of Component Function The Makeup Pump is one of three identical, high speed, high pressure pumps that can inject water into the primary system at normal RC3 operating pressures.
The Makeup pump is normally running and supplies normal inventory makeup to the RCS and seal injection to the RCPs.
In normal system alignment it takes a suction from the Makeup Tank.
During an SFAS initiation it is automatically aligned to the BWST and its suction from the Makeup Tank is isolated.
When all three high pressure pumps are available, the Makeup Pump shares its suction with the "A" HPI Pump. There are no low suction pressure trips or alarms associated with the pumps.
All three pumps are designed with recircelation lines which are piped to the Makeup Tank.
A constant flow of water is required to ensure cooling of the high speed, close tolerance pumps.
The lines ensure a minimum cooling flow of 105 gpm flow through each running pump.
The operating procedure states " Minimum Makeup Pump time at less.than 40 gpm flow is fif teen seconds".
The recirculation lines are isolated during SFAS to provide full flow capacity to the RCS, 3.
Method of Analysis l
The sequence of events, control room and shif t supervisor's logs, operator debriefing statements, and operating and emergency procedures l
we r.1 reviewed.
Maintenance history and open work requests were reviewed for any past work or open items which could have contributed to the l
failure. The configuration and operation of the Makeup System at the plant simulator in Lynchburg was discussed with simulator instructors.
The Makeup Pump was inspected following the transient to determine the extent of the damage.
l Page 63 l
t
LOSS OF ICS POWER TRIP REPORT NO. 75 4.
Results of Analysis The operators closed the BWST suction valse to the "A" train of SFAS ECCS pumps in an attempt to eliminate flow into the Makeup Tank whose level and pressure were increasing offscale high.
They assumed the running Makeup Pump would take its suction from the Makeup Tank and reduce the level and pressure but the earlier SFAS signal had closed the Makeup Tank outlet valve.
This cas. sed a complete loss of suction to the Makeup Pump and "A" HPI Pump.
The HPI Pump had been stopped two minutes prior to this action and was not affected.
The Makeup Pump continued was reached. The pump then began to cavitate and heat until it seized approximately twenty-five minutes later.
The coupling between the speed changer and pump was broken.
The motor ~tontinued to run until tripped by the operator a few minutes later.
The Lynchburg simulator plant and control room are loosely modeled after Rancho Seco.
However, many components, controls, and alarms are not simulated.
One example is the Makeup Tank outlet valve.
The operator does not have direct control of it nor does it close on an SFAS initiation.
The BWST outlet valve is modeled.
When the BWST outlet valve is open, the "A" HPI pump and Makeup Pump take a suction from the BWST. When it is closed, the pumps automatically take a suction from the Makeup Tank.
To stop a Makeup Tank level increase at the simulator, all the operator need do is close the BWST outlet valve.
This was exactly the same action performed by the operator during the transient, though with different results.
Neither operating procedures nor emergency procedures identify specifically what actions occur on an SFAS initiation.
As a result there are no recovery procedures to provide guidance for restoration following an SFAS initiation.
Initial operator licensing and follow-up requalification training include discussions of SFAS actuation.
This training is usually directed toward the response of individual systems rather than an integrated response.
Operators are" taught integrated response at the simulator, however, as stated above, this method does '
have shortcomings.
5.
Root Causes a.
The root cause of the failure of the Makeup Pump was isolation of both its sources of water.
b.
The root cause of the failure of the operator to ensure a suction to the makeup pump is threefold.
The simulator does not reflect the Makeup System as it exists at fancho Seco. There is no procedure for restoration from an SFAS initiation.
Sufficient training on integrated response of the plant to an SFAS initiation and, particularly, restoration does not exist.
Page 64
~
s f
p.
LOSS OF ICS POWER TRIP REPORT NO. 75 R.
Other Effects of Closing the "A" BWST Cutlet Valve SFV-25003 1.
Description of Concern Atapproximately0430,theBWSToutletvalveSFh-25003wasclosed.
The Makeup Pump, "A" HPI Pump, "A" Decay Heat Pump, and "A" Reactor Building Spray Pump were isolated from their suction source (s).
The Make'p Pump u
suffered severe damage as a result.
An evaluation was made to determine if any of the other pumps were damaged from this action.
Nehe of the subject pumps were on the NRC/IIT list of quarantined components.
2.
Method of Analysis The sequence of events, maintenance, surveillance testing and modification histories were reviewed.
Plant elevation drawings and Manufacturer's data were consulted in support of NPSH calculations.
Surveillance Procedures which demenstrate operability and check performance were performed for the "A" HPI Pump and "A" Decay Heat Pump.
3.
Results of Analysis The "A" HPI Pump was stopped about two minutes pr[or*to the closing of SFV-25003.
It remained shutdown until 0715.
Price to restarting the pump, the Makeup Tank outlet valve was opened.
The pump was not run without a suction.
The Reactor Building Spray Pump did not start during the event.
The pump does not automatically start from an SFAS signal until Reactor Building pressure exceeds 30 psig.
There are no other conditions which could have autostarted the pump.
l The Decay /LPI pump autostarted on the SFAS signal and was running in its recirculation mode.
The recirculated water was cooled by the decay heat cooler.
It operated for 57 minutes until stopped b'y a control room operator.
Approximately forty minutes of this time was with the suction i
valve closed.
Under the conditions that existed, the decay heat system discharge is isolated.
There was no path for water to leave the system on the discharge side of the pump.
A surveillance test is run to ensure there is minimal leakage in the discharge piping of the Spray and Decay Heat Pumps.
The high point in tae line from the Decay Heat Pump suction to the Makeup /HPI Pump suction is forty feet above the Decay Heat Pump.
Even if the running Makeup Pump drained all.the water out of its suction l
line between it and the BWST before its destruction, a mini' mum static l
head of forty feet would remain i.t the Decay Heat Pump suction.
The l
NPSH requirement is only eight fast so the pomo was not damaged.
This I
is supported by operation of the pump at full' flow for decay heat cooling since the trip, without apparent degradation.
Page 65
o LOSS OF ICS POWER TRJk REPORT NO. 75 The data obtained in the performance of the surveillances of the "A" HPI Pump and "A" Decay Heat Pump was compared with data from surveillances run prior to the transient.
The comparison indicated there had been no change in pump performance in either case.
S.
Filling of the Waste Gas Header 1.
Description of Concern The Waste Gas Surge Tank was partially filled with water during the latter stages of the transient.
This investigation identifies the source of water, the cause of the Makeup Tank overflow, and the resulting consequences to the Waste Gas System.
The Makeup Tank, Flash Tank, and Waste Gas Tanks along with their associated valves and equipment were not on the NRC/IIT list of quarantined-items.
2.
Description of Affected Systems The Makeup Tank is a 4000 gallon tank normally used as the receiver of RCS letdown, RCp seal return, and Makeup /HPI pump recirculation flow.
It supplies water to the Makeup and "A" HPI pumps in its normal system alignment.
When the plant is operating at power, it has hydrogen overpressure of about 25 psig.
A 100 psig relief valve protects the tank from overpressurization.
The relief valve discharge is routed to the 1600 gallon Flash Tank.
On SFAS initiation,.the Makeup Tank is isolated from all its inflows and outflows. The suction for the Makeup and "A" HpI pumps is automatically switched to the BWST.
Normally, there is no flow into the Flash Tank.
RCS letdown flow can be directed to the tank if selected by the control room operator.
Flash Tank level is controlled by two pumps.
The first pump starts when level reaches a predetermined point.
The second pump will start if the level l
continues to rise and tank level reaches a higher point.
The numos transfer their water to the 60,000 gallon tank in the Coolant Radwaste System. Together, both pumps have a capacity of 140 gpm.
If the tank level continues to rise, letdown flow is automatically redirected to the 60,000 gallon tank, bypassing the Flash Tank.
The Makeup Tank relief valve flow is not redirected.
The Flash Tank is contin;ously vented to the Waste Gas System. This system consists of a gas collection header, collection (surge) tank, two compressors, and four storage (decay) tanks.
The compressors take their suction from the collection tank and pressurize the in-service storage tank.
i page 66
7 LOSS OF ICS POWER TRIP REPORT NO. 75 3
3.
Analysis of Event The SFAS signal isolated the Makeup Tank.
As operators prepared to throttle HPI injection into the RCS, their first action was to restore HPI pump recirculation.
This action started a 300 gpm flow into the Makeup Tank. The Makeup Tank outlet valve was not opened.
Makeup Tank y
level began to increase rapidly.
Letdown flow was re-established, adding anot.her 80 gpa flow into the tank.
Level and pressure increased until both were offscale high. The relief valve.iifted and water started to flow to the Flash Tank.
Operators stopped the "A" HPI pump.
This also stopped its recirculation flow to the Makeup Tank..Two minutes later operators closed the "A" 8WST outlet valve and shifted letdown flow to the Flash Tank.
The Makeup Pump drained the remaining water from its suction line and then ran dry.
Its recirculation flow also stopped.
At thisipoint, the Flash Tank was receiving letdown flow and Makeup Tank relief valve flow which consisted of "B" HPI pump recirculation.
These flows totaled to 240 gpm, far in excess of the Flash Tank pumps' capacity.
Flash Tank level continued to rise until letdown flow was redirected to the 60,000 gallon tank.
Only the "B" HPI Pump recirculation flow was coming into the Flash Tank.
Its flow was slightly higher than the combined capacity of the Flash Tank Pumps.
Level continued to slowly rise.
The tank totally' filled and forced water into the Waste Gas collection header, and Waste Gas Surge Tank.
Approximately two and a half hours after the Flash Tank began to fill, the Nakeup Tank outlet valve was opened and the "A" HPI Pump started.
The pump took its suction from the Makeup Tank. The "B" HPI Pump was stopped.
Makeup Tank level and pressure decreased.
The relief valve -
reseated and flow to the flash tank stopped.
Flow from the Flash Tank to the Waste Gas collection header stopped and Flash Tank level decreased.
4.
Results of Overfill The Waste Gas Surge Tank and gas collection header"were drained.
The waste gas compressors were drained and inspected for damage.
No damage was found.
The compressors are designed for water induction so their ability to continue operation was not jeopardized.
The Waste Gas System was helium leak checked.
A single small leak, external to one of the compressors, was found and repaired.
The system was returned to' service.
5.
Root Cause The physical root cause of the overfill was the unisolation of the HPI l
pump recirculation valves without opening the Makeup Tank outlet valve.
The root cause sf the mental error is closely related to the root cause of the Makeup Pump failure discussed under paragraph Q. of this section and in Part VI of this report.
I Page 67
LOSS OF ICS POWER TRIP REPORT NO. 75 T.
Minimum Pressurizer Level Calculation 1.
Description of Concern During the rapid cooldown, Pressurizer level was offscale low for more than nine minutes.
This calculation was to determine if the Pressurizer emptied and, if so, whether or not a bubble formed elsewhere in the RCS.
2.
Method of Analysis Both the District and B&W performed independent calculations on the mass balance of the RCS and Pressurizer from the beginning of the cooldown to the time pressurizer level was recovered.
Data for RCS temperatures and pressures, letdown flow, Pressurizer level, makeup and HPI flow, and RCP seal injection were obtained from the IDADS computer.
Pressurizer level was adjusted for temperature compensation and instrument zero shift.
3.
Results of Analysis Both studies confirm that the Pressurizer and its surge line emptied.
At the time Pressurizer level went offscale low, the RCS contraction rate was 200 cubic feet per minute while inflow from HPI and seal injection flows totaled 130 cubic feet per minute.
The mass balance showed that the surge line emptied at approximately the same time a sharp drop in RCS pressure began to occur about three minutes after SFAS initiation.
The steam exiting the surge line readily collapsed in the cooler RCS hot leg water.
As RCS pressure decreased, HPI flow increased.
The RCS saturation temperature reached a point where a small amount of steam voiding was present in the reactor vessel head area.
Because the vessel design forces flow through the upper head, any steam that formed wou;J tend to be condensed af ter a short time and no large accumulacion would occur.
These two effects reduced the rate of RCS pressure decrease.
The bulk coolant remained subcooled.
As the operator closed the ADVs, the rate of primary to sdcondary heat transfer decreased.
RCS pressure began to increase and steam voiding in the head ended. The coolant flow in the head from the four running RCPs caused the steam bubbles to collapse.
The surge line filled and Pressurizer level began to increase. The B&W calculation concluded a small, less than one hundred cubic foot, steam volume war formed elsewhere in the RCS.
If at any time during this period of the transient the RCPs would have been lost or shut off, the steam voids in the system would not have been large ene. ugh to impact natural circulation cooling.
In addition, loss of forced flow would have greatly reduced the primary to secondary heat transfer rate and RCS contraction rate.
HPI flow would have quickly repressurized the RCS and compressed any steam voids in the head.
(
Page 68
LOSS OF ICS POWER TRIP REPORT No. 75 U.
Pressurizer Heater Operation 1.
De scription of Concern During the transient the Pressurizer level was offscale low for approximately nine minutes.
The heaters may have been' damaged if they had been energized during this period. The heaters were not on the NRC/IIT list of quarantined items.
2.
Description of Operation and Associated Interlocks During normal operation the heaters are operated in their AUTO mode.
In ttis mode they respond to RCS pressure arid are automatically energized on low RCS pressure.
The RCS pressure dropped below the heaters actuation setpoint a few seconds after the reactor tripped and remained below this value for the duration of the transient.
The heaters are located in the bottom of the Pressurizer, their uppermost portions corresponding to a Pressurizer level of 24 inches.
The heaters are grouped into two major divisions, normal and emergency heaters.
The emergency heaters function as normal heaters unless a panel switch places them in their emergency mode.
This switch remained in the normal position throughout the time Pressurizer level was offscale.
In this mode, all Pressurizer heaters are turned off by a common low level interlock which functions at forty inches from a temperature compensated Pressure level instrument.
3.
Method of Analysis A resistance bridge test was conducted on the heater coils to determine if any heater degradation had occurred.
The' data was compared to similar tests of pre /ious years.
4.
Results of Analysis Data showed that the heater resistances changed little over the past
(
five years and that the heaters were not damaged during the transient.
The low level interlock functioned properly.
It was also noted that following the event, the plant was stabilized and no problems were experienced controlling RCS pressure with the Pressurizer heaters.
V.
Response of Main Steam Relief Valves 1.
Description of Concern A review of the IDADS computer alarm typer indicated that one secondary
(
code safety valve (PSV-20544) lifted early and opened and closed repeatedly.
IDADS also showed an Auxiliary Steam code safety valve opened and closed repeatedly.
If true, these actions could have contributed to the cooldown rate.
page 69
e T
LOSS OF ICS POWER TRIP REPORT NO. 75 2.
Description of Associated Instrumentation The main steam and auxiliary steam relief valves and ADVs are monitored by an acoustic monitoring system.
Sensors are strapped to each valves' discharge piping.
When the system picks up sufficient noise and vibration it sends an alarm signal to the IDADS computer system.
The IDADS system responds with its own alarm on a monitor in the control room. The' alarm is also printed out on a high speed alare printer, also located in the control room. ~ When the valve reseats the alarm is reset and the typer prints that the valve is closed.
-3.
Description of Analysis The IDADS alarm typer printout was reviewed. The opening and closing i
times were compared against an OTSG pressure plot obtained from IDADS computer data.
Information from temperature sensitive stickers affixed to the discharge piping of each steam generator code safety was considered. The discharge piping of each relief valve and ADV was struck with a hammer to see the effect on the acoustic monitoring system.
4.
Results of Analysis The setpoint of PSV-20544 is 1050 psig.
The acoustic monitor indicated the valve lifted at 896 psig, opened and closed repaatedly, and finally reseated for the last time at 560 psig. The temperature sticker concurred that the valve lifted.
The hammer test of this acoustic monitor showed the instrument-to be very sensitive.
It would go into alarm when adjacent valves' discharge piping was struck.
It is concluded the valve did lift, but its lift and reseat pressures cannot be verified.
It is believed the auxiliary steam relief valve lifted.
A loss of ICS l
power. test was run following the transient.
When the power suoplies were de-energized, the Bailey station controlling the demand to the main steam to auxiliary steam reducing valve failed to mid scale. Normally the demand to the valve is fif teen to twenty percent.
The fifty percent -
demand signal would have overpressurized the auxiliary steam header, l
causing the relief valve to lift.
As steam generator pressure rapidly decreased, flow across the pressure reducing valve decreased.
I
~
Eventually steam pressure in the auxiliary steam header was reduced below the relief valve setpoint and the relief valve reseated.
A further review of IDADS data and temperature stickers revealed six secondary code safeties lifted.
The five remaining valves were compared to steam generator pressure data to determine their lift and reseat
)
points.
One valve, PSV-20549, was found to have lifted 23 psi early.
The results of the hammer test showed some acoustic sensors were too sensitive and some not sensitive enough.
Page 70
LOSS OF ICS POLER TRIP REPORT NO. 75 5.
Recommendations PSV-20549 should be tested to determine lift setpoint, and adjusted a.
as necessary.
b.
The gains on the amplifiers of the acoustic monitors should be adjusted as necessary.
W.
Main Steam Line Analysis 1.
Description of Concerns During the transient, OTSG A was overfilled with auxiliary feedwater.
Approximately 12,000 gallons spilled into the steam line.
Because of the large difference between the temperature of the steam line and auxiliary feedwater, there was a concern that high thermal stresses may have occurred.
There was also a concern that flooding may have caused water hammers.
Water hammer noises were heard in the turbine building sometime after the flooding had occurred.
Their source could not be located.
Another overfill-concern was the potential to inject water into the AFW pump steam turbine.
2.
Method of Analysis Bechtel performed a stress evaluation of the main steam line. The fatigue evaluation considered the loads imposed by thermal stratification caused by cold water in the bottom half of a hot pipe and thermal gradient stresses caused by the thermal fatigue.
Fatigue usage factors were calculated for two cycles of this event, plus operating l
basis earthquake (OBE)- stresses, design basis earthquake (08E) stresses, i
and 1000 cycles of pressure temperature loads.
To determine any damage that could have resulted from water hammers, walkdowns of the A main steam line and A main steam line bypass to the condenser were performed.
Supports were checked and measured using previous detailed information from I&E Bulletin 79-14 walkdown packages.
Steam line piping capacities and relative elevations were calculated to evaluate the possibility of water injection into the AFW pump turbine.
3.
Results of Analysis The results of the stress evaluation indicate that the fatigue usage factors for the most critical components in the steam line are all below 0.3, compared to a allowable maximum of 1.0 Page 71
i l
LOSS OF ICS POWER TRIP REPORT NO. 75 No visable evidence of water hammer was found on the A main steam line or bypass line to the condenser as a result of the configuration comparison.
One anomaly was noted where the bypass line penetrates the Auxiliary Building wall.
Sheet metal flashing covering the piping penetration was pulled away from the wall in this location.
The pulled flashing is thought to be caused by thermal expansion of the bypass line.
To ensure that water hammer did not occur in this area, insulation was removed and a detailed examination of the welds and piping was performed.
The suspect piping was subjected to. visual, ultrasonic and magnetic particle examinations. No indications of damage were found. Any postulated water hammer in this area would have been limited to this run of pipe.
Over 19,000 gallons of water would have to spill into the A main steam line before the steam supply line to the AFW pump turbine was reached.
Since less than 13,000 gallons was spilled, it was concluded that no water was injected into the turbine.
X.
Loss of RCP Seal Injection 1.
Description of Concern RCP seal injection flow was lost when operators stopped the "B" HPI pump while the running Makeup pump was isolated from its sources of suction.
An investigation was made to determine possible negative effects on the pumps and seals.
2.
Method of Analysis IDADS data for RCP seal injection flowrate was reviewed.
Control room charts and recorders for RCP pump, motor, and seal parameters were examined. Control room logs of pump performance prior to the following the transient were reviewed.
The RCP manufacturer was consulted.
3.
Results of Analysis IDADS data showed total seal flow decreasing to 4 gpm twice, corresponding to the two times the "B" HPI pump was stopped during the transient. Normal seal supply is 35 gpm.
The durations of the low flows were 13 and 9 seconds.
The RCPs are designed to operate with no seal inject flow provided the pump is running and 90 seconds if it is not.
Since the low seal injection flows occurred for such a'short time period, seal damage is not a concern.
The manufacturer concurred with this assessment.
f Page 72
LOSS OF ICS POWER TRIP REPORT NO. 75 Y.
Pegging Steam Lift of Feedwater Heater Relief Valve 1.
Description of Concern operators reported they heard a feedwater heater relief valve lifting shortly after the plant trip.
Pegging steam had caused these valves to lift in the past.
Pegging steam was isolated from the control room.
Two months previous, an overcooling event was initiated by pegging steam lifting feedwater heater relief valves. The October event was evaluated and modifications were made to the pegging steam system to prevent recurrence.
An investigation was made to determine the reason for the relief valve lift.
2.
Description of Associated Systems Pegging steam is directed to the second and fourth point feedwater heaters when turbine extraction steam can no longer provide enough flow for sufficient feedwater heating.
Pegging steam automatically starts when the turbine is at low power or is tripped.
High pressure steam supplies the pegging steam header.
Pressure reducing valves decrease this pressure to 185 psig to the second point heaters and 150 psig to the fourth point heaters.
The heater relief valves have lift pressures of 250 psig and 175 psig respectively.
3.
Method of Analysis The sequence of events, operator statements and operator followup interviews from the transient were reviewed.
Power supplies to the pegging steam components were reviewed to see if the loss of ICS power would have had any effect on the system.
The modifications and testing performed after the October incident were reviewed.
4 Results of Analysis The runback of Feedpump speed due to the loss of ICS power resulted in a cessation feedwater flow to the steam generators at the time of the l
trip.
There was no feedwater flow through the second point-heaters at this time and the only flow through the fourth point heaters was the feedpump mini-flow to the condenser.
When the main feedpumps were tripped, this mini-flow stopped.
The pegging steam condensation on the tubes of the heaters will stop as the feedwater inside the tubes heats to the saturation pressure of the steam in the shell.
At this point the pegging steam control valves will close.
However, both the pegging steam control valves to the fourth point heaters are known to leak through.
This causes a pressure increase on the shell side of the heaters and eventually lifts the relief valve.
The operators took the proper action to isolate the pegging steam header.
I Page 73
LOSS OF ICS POWER TRIP REPORT NO. 75 The 125 VDC "E" bus supplies power to the pegging steam solenoid valves. This bus has no affiliation with ICS power. The loss of ICS power had no direct affect on the pegging steam pressure reducing valves.
5.
Follow-up Actions Suspect pegging steam control valves were disassembled during this outage. One control valve was found seriously " steam cut" and requ. red reworking.
The valve was repaired and repeat relief valve opening is not expected.
Z.
"A" HPI Injection Valve Position Indication 1.
Description of Concern During the post trip transient, the "A" HPI injection valve, SFV-23811, was throttled to a low flow rate along with the other three injection valves.
The operators used indicated flow rates to balance the four valves at approximately 100 gpm each. When they performed this action, the position indication for SFV-23811 indicated " closed" although the flow meter suggests the valve was opened slightly.
This condition was observed only on this one valve.
The other three HPI injection valves performed as expected.
2.
Method of Investigation The maintenance and testing histories of the valve and flowmeter were checked.
The flow transmitter had been the subject of a detailed investigation following a trip in October, due to flow indication errors at low flowrates.
Based on the results of this investigation, flowmeter error was not expected to be the problem, since the required error is in the opposite direction of the expected error.
The setpoint of the Limitorque operator limit switch which gives the closed position indication was checked.
3.
Results of Investigation The limit switch setpoint was found to give a closed indication when the valve was two turns of the manual handwheel of f the seat.
The setpoint was adjusted and the valve stroke tested.
Page 74
LOSS OF ICS POWER TRIP REPORT NO. 75 AA.
Actuation of the TSC Sprinkler System 1.
Description of Concern The Technical Support Canter (TSC) fire detector actuated at the time of plant trip.
Although there was no sprinkler discharge into the rocm, water vented from the fire valve complex, quickly filled a bucket, and ran underneath the TSC false floor.
Water accumulated around computer cabling and eventually dripped onto switchgear and relay cabinets in the floor below.
Operators isolated the system an hour and a half after initial actuation, poly sheeting was placed on top of the electrical cabinets.
An investigation into the system design was made.
2.
Description of the Components The TSC sprinkler system is a pre-action system.
It takes a fire detector actuation plus a fusible link melt in a sprinkler _ head to start any sprinkler discharge into the room.
If only the electronic fire detector actuates, as was the case in this event, air pressure on an upstream check valve will prevent water from entering the s'prinkler header.
The fire detector de-energizes a solenoid valve which vents water pressure off a diaphragm valve.
Water flow continues through the solenoid valve by way of an orificed line which is meant to keep pressure on the diaphragm valve when the solenoid valve is closed.
The solenoid valve remains open until the electrical portion of the fire detection system is reset with a manual pushbutton located in the TSC.
Refer to Attachment 16 for a schematic of the pre-action sprinkler system.
i The Control Room and TSC are adjacent rooms.
Both rooms are served by a common Essential HVAC system and were designed to form a sealed area in' the event of a radiation or toxic gas release.
As a result, there were no floor drains designed into the rooms. The vent lines from the fire system were run to a five gallon bucket just below the fire valve complex.
The bucket will quickly fill when the system is actuated.
Actuation of the system is alarmed on the IDADs computer.
During normal operation, the alarm will result in the Control Room sending an opsrator to investigate within a short time, before the bucket overflows.
l Page 75 l
l
LOSS OF ICS POWER TRIP REPORT NO. 75 The electrical portion of the system is powered by 120 VAC from the "F"
non-vital bus. The "F" bus is fed from the "C" or "D" non-vital motor control centers through an ABT, and ultimately from a station auxiliary transformer when the plant is greater than 40 percent power.
The auxiliary transformer is fed from the plant's main generator.
When the reactor trips, it sends a signal to trip the main turbine and generator.
The site loads which were fed from the auxiliary transformers are transferred to startup transformers which are fed from the grid.
The transfer takes only a split second.
3.
Results of Investigation The TSC fire alarm was received at the same time as the reactor trip.
It is believed the loss of power for the few cycles it took the site loads to transfer to the startup transformers was responsible for actuating the fire system.
Once actuated, the flow of water through the orificed line continued until operators manually reset the solenoid valve.
The flow continued for quite a while as operators attended to higher priority items.
The water flow quickly filled the bucket and ran onto and through the false floor, eventually dripping into the switchgear room below.
There were no malfunctions of computer or electrical equipment due to the water.
I l
l 88.
Main Steam Line Failure Logic 1.
Description of Concern Operators tripped the main feedpumps shortly after the trip when they saw they had no control over main feedwater flow.
As steam pressures decreased below the discharge pressure of the running condensate pumps flow was initiated through the idle feedpumps and half-open startup feedwater valves. This flow was stopped by the automatic actuation of the main steam line failure logic (MSFL) a couple of minutes later.
There are three concerns with the MSFL actuation.
a.
An initial study of the IDADS computer data revealed that the logic actuated significantly lower than the 435 psig setpoint, thus prolonging the flow from the condensate pumps, page 76
LOSS OF ICS POWER TRIP REPORT NO. 75 b.
Consideration should be given to increasing the NSFL setpoint above the condensate pump shutoff head.
c.
Control Room operators were unaware that MSFL had actuated.
2.
Method of Analysis The failure logic was examined to determine 2xpected behavior, vendor Manuals were checked for setpoint inaccuracies and drift.
IDADS computer data was reviewed to determine the time of MSFL actuation and coincident steam pressures.
3.
Results of Analysis The pressure switch actuation setpoints have a tolerance of +/-5%.
Thus the lowest setpoint would be 364 psig.
The IDADS data for the pressure switch actuation is scanned on a one minute time basis.
The pressure in the OTSGs at the time of the IDADS indicated the pressure switches actuated may be significantly lower for rapid overcooling transients.
In this transient the pressure indicated by IDADS one minute prior to the indication of switch actuation was approximately 410 psig.
This is only slightly less than the expected setpoint and is :within the tolerances of the pressure switches.
It is concluded the failure logic operated as designed.
Operators were not aware of the MSFL actuation because there is no main panel annunciator to remind them this action has taken place.
There is an alarm on IDADS for each pressure switch, however, in a transient situation one operator would have to dedicate a major portion of his time reading the blarm typer to note the actuation.
Two hundred seventy-five alain messages were generated by IDADS during the first thirty minutes of the transient.
Also, the IDADS alarm for the pressure switch reads as follows:
MAIN STEAM PRESS PSL 20602 LOW ALARM ALARM STATE The message states that steam pressure is low, a fact probably already obvious to the operator.
It does not tell the operator that the pressure switches are part of MSFL and that isolation of the main feedwater path to the steam generators has been er. acted.
Although the logic actuation aided the, operators in this transient, there have been other times when main feedwater was lost during a cooldown and operators did not recognize the loss until the steam generators were almost dry.
,~,
Page 77
LOSS OF ICS POWER TRIP REPORT NO. 75 CC.
SPDS Versus Steam Generator Operate Level Recorders 1.
Description of Concern The STA and several control room operators reported that Safety parameter Display System (SPDS) OTSG operate levels were not observed to increase above 97% on the A OTSG and 93% on the 8 OTSG.
The control room strip chart recorders indicated that OTSG levels were of f scale high for a period of thirty to forty minutes.
The IDADS points which monitor OTSG startup and full range level indicate that levels in the steam generators were high enough that all operate range level instruments should have read offscale high.
2.
Description of Instrumentation The SPDS and IDADS computer receive each steam generator operate level indication from the same transmitter.
The IDADS system monitors this signal as-is, i.e. uncompensated for temperature.
The SPDS compensates this level using an input from an OTSG pressure transmitter.
This calculation assumes the OTSG is in a saturated condition.
The control room level recorder is driven from either af two selectable, temperature compensated instruments.
Uncompensated levels from the same instruments are fed to IDADS.
3.
Results of Analysis A review of data during the time of steam generator fill to 100% in the operate range showed the two IDADS uncompensated levels from the A OTSG differed by no more than 1.5%, and typically by 1%.or less.
Data from the 8 OTSG uncompensated level transmitters showed even closer agreement, typically within 0.7%.
The uncompensated level signals could not have been the cause of the discrepancy.
l 1
l Page 78 i
LOSS OF ICS POWER TRIP REPORT NO. 75 The temperature compensation system was seen as a possible cause of the difference. A previous calibration check on SPDS level verified the SPDS compensated level agreed with the control room level recorder. The previous check was run when the plant was at power and operate levels were near 90%.
The SPDS pressure compensating algorithm was checked
-against this data and was found to be valid. The algorithm was then used to calculate levels of 951, using calculated differential transmitter pressures and varjing steam generator pressures.
Compensated level changes varied by only 0.5 percent for pressures ranging from 200 to 1050 psig.
Uncompensated operate level data and OTSG pressure for the transient were then retrieved from the IDADS archive history files. The algorithm was applied to these values.
Calculation of compensated level yi9lded values of 98% to 99% for both steam generators.
Three different values of uncompensated level were input into SPDS as dummy signals.
This was done at three different pressures, also input as dummy signals.
The results showed the SPDS response within 0.6% of the expected values.
From this data it was concluded the SPDS performed correctly during the transient.
At most, SPDS level reads about 1% to 2% less than the operate range recorders.
The SPDS was probably indicating 98% to 99%
during the transient when operate levels were offscale high.
DD.
Damage to Reactor Building Radiation Monitor 1.
Description of Concern One of the Reactor Building radiation monitors (R-15001 A&B) was damaged when its suction valve was closed by SFAS initiation and its compressor continued to run.
Approximately eighty minutes after the suction valve was closed, the compressor seals overheated and tripped a fire alarm in the room. The fire alarm tripped the running Auxiliary Building Radwaste Exhaust fan.
An evaluation was made to determine the corrective action to prevent recurrence.
The radiation monitor was not on the list of NRC/IIT quarantined components.
2.
Results of Evaluation A contact from the existing low flow switch will be incorporated into the control circuitry of the compressor.
The compressor will stop when low flow is sensed.
page 79
LOSS OF ICS POWER TRIP REPORT NO. 75 EE.
Radiological Assessment of Steam Releases 1.
Description of Concern Because of residual activity in the steam generators from previous primary to secondary leaks and the possibility of a continuing small tubeleak, the secondary code safeties, ADvs, feedwater heater relief valve, and auxiliary steam relief valve discharged a small amount of radioactivity to the atmosphere.
An evaluation was made to. determine the fraction of maximum permissible concentration (MPC) at the site boundary.
2.
Method of Analysis Main Steamline samples were taken the morning of the event.
Each isotope from each oain steamline was conservatively added together. The dose calculation assumed that all valves were open for six minutes.
Actually, the highest volume steam flow source, the secondary code safeties, were open for only a few seconds.
An average atmospheric dispersion factor was used.
3.
Results of Analysis l
The NPC at the site boundary was 0.0085 MPC.
The largest contributors to this value were two Cesium isotopes, Cs-135 and Cs-137.
FF.
Radiological Assessment of Auxiliary Building Releases 1.
Description of Concern A spill from the damaged Makeup Pump seals resulted in an estimated 1200 I
gallons of mildly radioactive water on the pump room floor.
The floor drain is routed to a sump on a lower floor.
The sump's pump transfers excess water to a radwaste holding tank.
The room; sump, and tank air is drawn through HEPA and charcoal filters by the Radwaste exhaust fan and discharged to the environment.
Analysis was performed to determine the offsite dose as a result of the spill.
2.
Method of Analysis The two Auxiliary Building stack gaseous radiation monitor printouts were reviewed.
Calculations were based on the monitor indicating the highest reading. The calculation conservatively assumes that both radwaste exhaust fans are e unning.
Only one fan was running during the spill.
An average disperrion factor was used.
An air sample of the Page 80
LOSS OF ICS POWER TRIP REPORT NO. 75 Auxiliary Building basement was taken during the time the fans were tripped by the Zone 20 fire alarm.
Charcoal cartridges in service during the length of the event from the two radiation monitors were analyzed for radioiodine.
The calculated dose included the consideration that the exhaust fans were not running for about half an hour.
3.
Results of Analysis The total number of noble gases released was 33 curies.
The major isotopes were Xenon-133 and Xenon-135.
The analysis of the charcoal filters showed no radioiodine was released. Whole body dose to a person at the highest downwind sector at the site boundary for the entire length of the event was calculated at 0.02 mrem.
GG.
Radiological Assessment.of Operators' Entries into Makeup Pump Room 1.
Description of. Concern Following the destruction of the Makeup Pump and the spill into the room, the control room directed two operators to first perform a quick visual inspection of pump damage then later,-to isolate the Makeup Pump.
The operators experienced minor skin contamination and contamination of their personal clothing taking these actions.
An evaluation was performed to find the root cause and to prevent recurrences.
2.
Description of Event The control room di.ected the operators to survey the Makeup Pump for camage.
The operators were told there had been a loss of coolant (LOCA) from the pump into the pump room.
Neither the control room nor the operators notified the on-duty HP technicians of their impending entry.
Both operators dressed in cloth and paper coveralls, two pairs of rubber gloves, rubber booties, and flat PVC shoecovers.
The operators could not locate any high-topped PVC shoecovers, however, a later investigation showed there was an ample supply in the changeroom. One operator also wore a set of plastic raingear and taped his gloves to his l
cloth coveralls.
The operators took a portable instrument with them to assess radiation le/els.
The operators did not wear respirators.
No assessment was made of contamination levels or airborne activities prior to entry.
The operators entered the pump room and found a couple of inches of water on the floor plus a water and oil mist vapor in the air.
They assessed the damage and phoned the control room with their results.
They sport approximately one minute in the room.
Page 81
o LOSS OF ICS POWER TRIP REPORT NO. 75 A few minutes later the control room contacted the operators and told them to manually isolate the pump. The operators returned to the room and performed this action, which took about five minutes.
Upon exiting the controlled area, they both set off alarms on the Hand and Foot Monitor.
An HP technician happened to be in the area of the monitor.
He found one operator's shoes, socks, and pants were contaminated.
The other operator experienced minor skin contamination of one hand and contamination of his shoes.
One operator received 20 mrem external exposure and the other 10 mrem as read from their pocket dosimeters.
Before they lef t site that morning, both operators had body counts.
The results of one operator indicated one percent maximum permissible body burden of Silver-110M.
The other indicated 0.3 percent of the same isotope.
3.
Method of Evaluation An evaluation of plant policy in regards to the interface between the operations group and Health Physics group was undertaken.
The logistics of radiation protection clothing and equipment during emergency conditions was investigated.
4.
Results of Evaluation A lack of clearly defined separation of responsibilities between Operations and Health Physics was found.
Operators have been trained in the fundamentals of radiation exposure, radiation protection, and the use of the common hand-held doserate instruments.
They routinely monitored their own exposure in high radiation areas without the escort of an HP technician.
New policy has been formulated which requires an HP escort for all entries into high radiation areas.
It also requires an HP escort into any area where radiological conditions are unknown or may have recently changed.
An HP technician has been assigned to Operations around-the-clock.
His sole duty will be the support of operators in the radiologically controlled areas of the plant.
Respirators and SCBA's were not immediately available to the operators in the basement of the Auxiliary Building.
This equipment is located three floors above the Makeup Pump room.
Page 82
LOSS OF ICS POWER TRIP REPORT NO. 75 HH.
Radiological Assessment of Operation of AFW Control Valves 1.
Description of Concern Both AFW control valves are located outside, one between the Reacter and Auxiliary Buildings and one next to the Reactor Building. One valve is in the radiologically controlled area and one is not.
Due to the seriousness of the event and the high priority given to closing the valvas, operators " jumped the fence" crossing back and forth between the two areas.
This is a violation of radiological work practices and procedures.
2.
Description of Controlled Area The "A" AFW control valve is located within the controlled area.
A small portion of this area is a contaminated area.
This' portion is around the Spent Fuel Cooler.
Another portion of the area is a radiation area due to the conts.nts of the BWST.
The vicinity.around the valve is not contaminated and not within the radiation area.
The operators did not go through either area on their way to the valves.
Entry into the major part of the area requires dosimetry and sign-in sn a general RWp.
No anti-C's are required.
3.
Results of Evaluation The majority'of the subject area has been made uncontrolled'.
An area of the chain link fence which previously separated the controlled and uncontrolled areas has been removed.
Operators now have easy access to both AFW control valves from either area.
II.
Timely Assessment of Offsite Radiological Releases 1.
Description of Concern There was an inabililty to obtain meaningful data on secondary code safety and ADV valve lifts.
This prevented a timely assessment of releases from the secondary system.
Equipment design problems and procedure deficiencies prevented a timely assessment of releases from the Auxiliary Building.
l page 83
a LOSS OF ICS POWER TRIP REPORT NO. 75 2.
Description of Components and Instrumentation Lifts of each of the secondary code safeties, auxiliary steam relief valves, and ADVs are monitored by an acoustic monitoring system. 'These monitors provide inputs to the IDADS computer system.
IDADS generates an alarm message or reset message each time the monitor senses the valve is open or closed.
The alarms are printed on a typer in the control room. Temperature sensitive stickers are also affixed to the discharge of each main steam safety valve.
The effluent from the Auxiliary Building exhaust fans are sampled by two different radiation monitoring systems. The older system reads in counts per minute (cpm) and prints out on a multipoint recorder.
The newer system reads in microcuries per cubic centimeter (uci/cc) and displays both current and historical data on a television monitor in the control room.
Both systems provide inputs into the IDADS computer system and both have alarm capabilities.
The IDADS system contains programs (RAC1,2, and 3) which provide the operator in the control room or TSC monitoring capability of effluent activity leaving the Auxiliary Building stack.
3.
Results of Evaluation The varied sensitivities of the acoustic monitors flooded the IDADS system with alarms.
One hundred forty-six messages were printed on the IDADS alarm typer from nine different valves within six minutes of the loss of ICS power.
The assimilation of this data into a meaningful package to inform Health Physics which valves lifted and for how long takes, at a minimum, several hours.
problems with the acoustic monitoring system are detailed under paragraph V. of this section.
The temperature sensitive stickers gave a good indication of which valves lifted but do not yield any information on the duration of the lif t.
The alarm setpoints for the two different Auxilia'ry Building effluent radiation monitors were found to differ by a factor of about ten.
Although this does not affect a dose calculation, it can give the operator conflicting information.
The RAC computer programs could not be called from the IDADS consoles.
These programs were not included in the reboot procedure for the IDADS computer.
Cue to continuing work on the computer system, the reboot
-procedure is used frequently.
page 84
LOSS OF ICS POWER TRIP REPORT NO. 75 The multipoint recorder which gives a history trace of the older radiation monitors prints the number of the scanned point only a few times an hour.
It prints a dot for the scanned point between these times. Due to the number of scanned points, it is not possible to see the trend of an individual monitor between the prints of its associated point number.
The trend between numbered points becomes a job of judgement or " guesswork".
Refer to Attachment 17 for a sample of the multipoint recorder printout.
JJ.
Operations / Security Interface 1.
Description of Concerns During the event, one outside operator lost his security badge.
The operator followed his assigned helper into any security areas he required access until the plant was stable and he had time to inform Security.
This was a violation of security procedures (tailgating).
The Security Watch Commander was not informed that the plant was in an Unusual Event until one hour after the initial declaration.
The security area access card reading time at the centrol room door began to impact the timely dispatch and reporting' back of operators as they attempted to stabilize the plant under manual control. Eventually, a guard was posted at the door, allowing it to remain open for the remainder of the event.
Although the hinderances did not significantly affect the outcome of the event, it was apparent that improvements could be made to ensure Security is a part of event response and that effects of anticipated occurrences could be minimized by appropriate plans and procedures.
Page 85
o LOSS OF ICS POWER TRIP REPORT NO. 75 2.
Results of Evaluation The Emergency Plan, Notification Section, was revised to provide notification of the Watch Commander prior to outside agencies whenever the decision is made that the event requires Emergency Plan response.
Upon notification, a security officer will be dispatched to the control raos alcove area to communicate with Security any needs which involve security doors or badges.
Any operator who loses his badge will inform the control room.
Control room personnel will inform the security officer who will, in turn, notify security personnel at the main security building.
A temporary badge, cleared for all vital areas, will~ immediately be made available to the operator at the main security building.
KK.
Human Factors Review 1.
Overview of Program The District determined that a human factors evaluation should be made to assess the operational usefulness of the control room and operational concerns in the plant during the transient.
Many of the concerns mentioned by the operating crew had previously been documented by a Control Room Design Review (CRDR).
The human factors evaluation consisted of an initial interview with all the operators involved in the transient, a review of post trip documentation, follow up operator interviews when necessary, and walk-throughs of specific events.
2.
Results of Evaluation The process identified 26 items which were worthy of study.
Of these, 12 items were previously identified in the CRDR.
The 14 new items were l
beyond the scope of the original review.
A complpte listing of the items is provided as Attachment 19.
Disposition of the individual recommendations is being handled by assigning appropriate priority and resources to each.
Those which justify immediate resolution have been scheduled for resolution prior to startup.
l l
I l
l Page 86 i
]
I
LCSS OF ICS POWER TRIP REPORT NO. 75 VIII. ROOT CAUSES(S)
The Rancho Seco Incident Analysis Group (IAG) has responsibility to determine the
" root cause" of problems or events at the facility.
This involves addressing the programmatic.causes as well as the direct causes unique to specific occurrences.
The IAG charter causes this effort to be separate from other efforts to seek and resolve causes.
In the instance of the December 26 event, the IAG monitored the activities and findings of the Transient Analysis Organization, but used methods and resources to independently arrive at the " Root Cause(s)" of the event.
These root causes are presented below.
Although developed independently, their basis is supported by the material presented in this summary report.
The December 26 event was subdivided into five " event themes" for root cause determination.
They are:
Loss of ICS Power Rapid Cooldown Makeup Pump Health Physics Emergency Plan Analyses of these event themes were then studied for the following considerations:
Procedural Adequacy Unique Design Features Human Factor Consideration Related Issues -
Training Preventive Maintenance Personnel Access Vendor Technical Analysis The following discussion identifies the Root Causes and the contributory causes, as appropriate.
1.
Loss of ICS Power l
Root Cause: Manufacturing Error i
l A lug was improperly installed on a factory prepared wire.
The resulting connection exhibited variable resistance which was on the input to the Power Supply Monitor.
The resulting variable voltage lead to the PSM tripping when the ICS was still being supplied with nominal voltage and power.
Corrective action involved installing new wiring and lugs.
Page 87
LOSS OF ICS POWER TRIP REPORT NO. 7S Contributory Causes:
1.
The Power Supply Monitor is sensitive to resistance in series with its voltage input.
As little at one ohm was found to cause the trip point to increase.
Approximately 5 ohms at the failure point was sufficient to cause the PSM to trip at its nominal operating voltage, 24 VOC.
Corrective action involves wiring the PSM directly to the DC source bus l
rather than through a daisy chain of wire wrapped terminals.
2.
The Si and S2 source switches were found to have short built-in time delay characteristics, approximately 0.15 second while the specification l
is for 0.2 to 0.8 seconds.
This made them more sensitive to short term trip signals generated by the PSM.
Corrective action is to obtain new switches.
2.
Rapid Cooldown of NSS Root Cause:
Delay in Implementing Design Changes to Mitigate Effects of Loss of ICS Pcwer The susceptability to this event has been recognized for some time.
In response, a Class I design modification called EFIC has been developed.
Revisions in design criteria, delays in reaching scheduled refueling outages, equipment manufacturing croblems and delays have compounded to delay its implementation.
Interim corrective action is to install modifications which will provide the Control Room Operator with necessary controls powered independently of the ICS, while aggressive efforts are being made wnich will install EFIC at the next refueling.
Refer to attachment 20 for a discussion on the background and history of EFIC implementation.
Contributory Causes:
1.
Procedures The overcooling procedure did not clearly identify criteria which would cause the operator to take second level actions to terminate the condition when the initial effort proved ineffective.
Corrective action has been taken to incorporate this guidance into the EOPs.
i l
l 2.
Training could have compensated for the lack of the above features or procedures.
Training which had been given was not able to compensate for those deficiencies.
Corrective action; training developed to understand the event, revised policies, and new procedures will increase the awareness and knowledge of the operators.
Page 88
LOSS OF ICS POWER TRIP REPORT NO. 75 3.
Damaged Makeup Pump Root Cause:
Procedures A procedure specific to restoration of normal equipment lineups following SFAS initiation was not available.
Procedural references in the EOPs for insuring pump suction / discharge paths were not consistently included.
Corrective action is to provide the missing procedure and add the appropriate caution steps.
Contributory Cause: Training The operator was trained and aware of the consequence of operating th'.s pump without suction.
This event demonstrates that training alone may not always be sufficient to insure requirements will be remembered.
Training should not be expected to compensate for lack of appropriate procedures.
Corrective action; this event reinforces previous trainirg with an example of the ec, sequences.
Retraining and new procedures are sufficient to preclude r= occurrences 4.
Health Physics Procedure Implementation Root Cause:
Human Performance The perception of the individuals involved was that they were following appropriate guidelines for the conditions which prevailed and the directions they have been given.
Corrective action is to clearly restate Administrative Policy that procedures will be followed and provide training on implementing procedures.
Contributory Causes:
1.
Training These individuals did not clearly understand their obligation to follow established procedures and to utilize the protective equipment and other personnel available to them.
Corrective action is retraining of all operational personnel to assure that others may not harbor similar concepts and perceptions.
2.
Imprecise Definition of HP Responsibilities, Authority, Outies Administrative policy is being restated while policy is being changed to assign a HP Technician to Operations for the single purpose of supporting their activities.
I
\\
i i
l l
Page 89
a LOSS OF-ICS POWER TRIP REPORT NO. 75 1
5.
Emergency Plan Cornnuriicatio.ts Root Cause: Training Training was not sufficient to insure the Emergency Plan would be properly implemented during n' plant transient.
Priority of responsibility to that plan was not fully appreciated by some. Corrective action is to provide training on implementing command and control policy and obligation to assigned duties.
Contributory Causes:
1.
Human Performance The operators were faced with several independent " Emergencies." The overcooling, fire alarms, radiation alarms, damaged equipment, disabled operator, and failed equipment were the major ones.
These diverse events challenge the ability to be effective and prioritime.
Corrective action is to restate policy and process in the training program while developing training exercises which will better practice complex scenarios.
The relationship between the roles of the Shift Supervisor / Emergency Coordinator and the Senior Reactor Operator are being defined and practices in plant simulator training.
2.
procedures l
Complex branching within the Emergency Plan and its implementing l
procedures makes smooth implementation-difficult.
This event highlighted the situation and led to revisions to remove many of the obstacles.
Corrective actions, in addition to increased training, the necessary forms for an event are being packaged into a folder which will l
insure all are convenient and ready to use.
I I
l l
I
,(
Page 90
LOSS OF ICS POWER TRIP REPORT No. 25 l
}
}
l I
i l
Page 91
e LOSS OF ICS POWER TRIP REPO.tT NO. 75 i
i l
l Page 92
~
r LOSS OF ICS POWER TkIP REPORT NO. 75 X. CONCLUSIONS The loss of ICS power initiated events which led to a. Reactor / Turbine trip and an overcooling of the RCS.
The. Engineered Safeguards System was activated and functioned as designed.
Twenty-five minutes after the start of the transient, ICS power was restored and the cooldown and steam generator, overfill was terminated.
Complications arose when restoring from the SFAS condition.
A detailed study into the causes of the many subevents led to an in-depth review of pertinent areas of system design, plant procedures, operator performance, human factors, and management policy.
Several significant changes resulted from the analysis of this transient :
which will enhance plant reliability, safety, and staff performa'nce.
The loss of ICS power has been traced to a manufacturing error in the power distribution wiring.
This wiring has been replaced and wiring in other critical cabinets in the Reactor Protection System and Safety Features Actuation System is being investigated.
A new control system has b'een installed Chich will prevent the excessive steam and feedwater flow that initiated the rapid cooldown during this event.
A procedure addressing appropriate actions for a loss offICS power is now in-place.
The ' modifications performed to the ICS eliminate the likifnood of a similar occurrence.
The addition of the new control switches and Casualty Procedure will mitigate the consequences of a loss of ICS power and reduce the chances of a rapid l
cooldown.
3 Extensive studies and analysis have been performed to assure the operability of the major components subjected to the excessive cooldown.
The results on the reactor vessel, fuel cladding, NSS components, and main steam lines analyse _s proved these components were stressed within their design limits and are acceptable for continued operation.
Any detrimental effects to the steam generator tubes were found using eddy-current techniques.
Suspect tubes.will be plugged prior to resuming operation.
A study of the Updated Safety Analysis Report(USAR) was conducted to compare the December 26 event to previously analyzed events.
This latest event was found to be within the bounds of the USAR rapid cooldown ev6nt.
The December 26 event was also compared to the March, 1978 "Lightbulb Incident".
The 1978 event resulted in a much lower temperature at a much higher pewssure than the Loss of ICS transient.
These transients plus other recent cooldown events at Rancho Seco were taken into account in the calculation of fatigue usage factors of RCS' components.
A post trip analysis of the reactor coolant mass inventory revealed that the Pressurizer completely emptied and a small steam void formed elsewhere in the RCS, probably in the vessel head.
The voiding did not pose any problems with core cooling as there was adequate subcooling of the forced flow of coolant through the core at all times during the transient.
The overcooling was allowed to progress while operators focused on. closing the Auxiliary Feedwater control valves, rather than.i.nplementing the procedure step which would have stopped the associated feedwater pumps. The Emergency Procedures have been revised to include references to specific plant parameters where appropriate to prevent the recurrence of excessive cooldown from main or auxiliary feedwater.
Page 93
LOSS OF ICS POWER TRIP REPORT NO. 75 A Reactor Building radiation monitor was damaged due to a design oversite when the SFAS signal isolated its suction. The Makeup Pump was severely damaged and the Waste Gas System partially filled with water as the operators attempted to restore the Makeup and Purification system from the SFAS lineup.
A Casualty Procedure has been generated which outlines the appropriate steps to be taken when securing from an SFAS.
Emergency Procedures.have also been revised to prevent recurrence of these events.
The radiation monitor has been repaired and returned to service.
The "A" HPI pump will assume the function of the Makeup pump until repair or replacement can be affected.
Other components which experienced failure during the transient have been disassembled, inspected, repaired as necessary and returned to service.
The Training Department has given all licensed and non-licensed operators classroom and
" hands-on" training on the proper manual operation of the ADVs, TBVs, and AFW control valves.
The Emergency Plan was not effectively implemented during the transient.
The. Plan
.will be revised to provide more guidance in responding to multi-event emergencies, less branching, and a reduction in the number of forms and logs required.
The roles of the Shift Supervisor and Senior Reactor Operators in a transient situation have been redefined.
Command and control training will be given and practiced during transient situations at the plant simulator.
Use of the Newly installed control switches outlined above will be used to mitigate loss of ICS power transients at~the simulator.
Proficiency with the revised Emergency Procedures, Emergency Plan,.and new Casualty Procedures will be demonstrated for this and other rapid cooldown simulations.
The described modifications and training will provide the desir d and necessary capability to control events such as loss of ICS power from within the Control Room.
The transient resulting from the loss of power will be much less severe, even assuming no operator action.
New or revised procedures and operator classroom training backed by practice in simulated events will enhance operator response to.
transients and ensure adequate communication with plant management and outside
- agencies, i
Page 94
LOSS OF ICS POWER Trip Report #75 XI. Attachments / Graphs / Figures 1.
IDADS Computer Graphs 100 2.
Control Room Strip Charts 113 3.
Calculated AFW Flow 117 4.
NRC I.E. Information Notice 85-04 118 5.
INFO Sig91ficant Event Report 6 - 86 121 6.
Rancho Seco Licensee Event Report 85-25.
128 7.
December 26, 1985 Transient Action List 131 8.
Main Feedwater Simplified Schematic 150 9.
Auxiliary Feedwater Simplified Schematic 151
- 10. Main Steam Simplified Schematic 152
- 11. Letdown / Makeup /HPI/ Waste Gas Simplified Schematics 153 12.
ICS Power Distribution 159 13.
Modifications to ICS Controlled Components 161 14.
ADV/TBV/AFW Valve Operators 165 15.
Simulator Run With no Operator Action 168
- 16. TSC Sprinkler System 170
- 17. Radiation Monitor Multipoint Recorder
,171 18.
Emergency Operating Procedure Revisions 172
- 19. Human Factors Study Results 175
- 20. EFIC Implementation 178 l
l l
l t
s R EACTO R TR 3
DEC.
26, 1985 RCS PRESSURE / TEMPERATURE (Tove) 090 -
2.2 -
O
-Rew.uzee Seac,e twe Emp ress 080 -
j 070 -
2.1 -
- C u>-,w c, ADW 060 -
2-
- Co o., ac., AF# CoMTRot. VAWGs e 50 -
Reasee.estee Spa.4y duva.T o40 -
1.9 -
gg,,3, c 30 -
1.8 -
Rx Trip Q Id5 /b e4 #erareeb e 20 -
10 - 9 1.7 -
o 500 o
1.6 -
490-Uj@m 1.5 -
,4 480 L o 470 - h 1.4 _
v 460 -
13~
450 -
440 -
1.2 -
umane-A 430 -
L uaq in asAn t
ll-b 420 -
U
- 410 -
1-j i
Sv.u.e Lin s _
g'
-+
l400 -
R e cit.t w c, O.9 -
l390 -
gm gu,;
y Re.s dMA!
f380 -
0.8 -
b 04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 L
TIME 8 RCS PRESS B-LOOP
-O-RCS-Tave B-LOOP IDADS P9000 AVERAGE OF IDADS T9010 AND T9016
REAC-O R TR 3
D EC.
26, 1985 RCS PRESSURE /PZR LEVEL 2.4 i
2.2 -
m u
2-1.8 -
.m
@,y g
1.6 -
o m"
E
- a. u 1.4 -
w 9
w g
'O Wo
,o te a 1.2 -
3s O
(A(nf 1_
.J Wu Q'
L*"
a-O.8 -
Soaqe LenE T? GALLS ICS TLee Eas ronso Ql O.6 -
Rx TRIP O.4 -
STAR T Eb O.2 --
Avw c.c ygo c vgtoe hs Owd
-MPL TutoTTeen n"
?
O-u a
a a
a g*
04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 S
TIME
- B --
RCS PRESSURE
-O-PZR LEVEL (UNCOMPENSATED)
IDADS P9000 IDADS L9005
00 00 5
n 0
e y
5 z
A o
t N
r r
e E
8 se P
0 S
2 O
0 9
T n
e 0
.a s
1 s
i p
5 z
o n
s A
P V
os B
s.
a s
e T
u 6
c u
t I
P A
d 0
l V
1 w
2IN b
P A
0 i
PA 3
u R
G e
H 0
G o
T c
s R
V 4
c r
j q
b g
A A
T e
a.
A g
CM T
S c
e r
c S
EG w
r u,
V t
N d
r o
0 og DL I
A e
e Eh g 0
MTl i
t O
C 3
0 O
M I
3 C
3 TSOR UF ND l
B s
I E RU u
MTA au,**
L S
Ls i
s, e T
tU s
0 aC t
S siv I
0 sA o e 0
C
- b 5tE TC 2
0 RR tS 2
O T
)
(
C 0
P A
,0 I
R T
0 F
x 1
R R
00
- 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 04 1 0 9 8 76 5 43 21 0 9 8 7 6 5 4 30 221 1 1 1 1 1 1 1 1 1 u h $$tEg2$2p g
li 1);
Il ii
{
i
}
26, 1985 HPI & MAKE UP FLOW 300 280 -
Rx TRIP O A
.g. ggpg Spy
^
260 -
C' S'" II"" I 240 - "A" IFI TV flit OPEN---
."B" IIPI SFV 8" llPI Pige STOP TilROT1 LED 220 -
"B" llPI PUMP START--
"B" HPI SFV CL SF n
"A" IPI PUMP STOP "B" HPI Pimp S ART 200 -
M C
y 180 -
"C. IPI snf CLOSED MAKE UP PUMP TRIPPED-O_
\\
(f0LL)
AT 0500 g
160 -
D s
- n" nPi Srv g
140 -
CLOSED (futL)
-. B" HPl
- 3. HPI PUMP START-L 3
\\
PUMP i
d 120 -
A-B-C-D llPI SFV's STOP/ START 4
TilROTTLED POSITION g
A*
fw,AC 100 -
ir-y r
r
80 -
H 4
Wicates the C
C C
Minningof 60 -
3 40 -\\'
o_
o
'tual flow was n
kbablyzero. 20 -
[
M
_H__
g m n i-- -- n
.-...-O H
H p
^
d^
O-04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 5
TIME A
B C
!)
M/U ilPI HPI IIP I IIP 1 IHJECTION f9003 09001 F9004 f9007 09000
26, 1985 TOTAL HPl & MAKE UP INJECTION 900 HP[ Rec.ac VAtoss Opsg sn -
g 800 - Lavai. cerscats' Tig n o r y.m a lossevi.a VAtoEs 700 -
HM wP SToP PG.b Rx TRIP Q k
Tu Iw3acTieu JAt0ES (uSED s
600 - SFAS W mATiou,
'A' HPI. Ao c n2 TRAVEL.s To
~'
Q-500 - TheTuaorvi.eb \\
S UC.T to e4 LEVA t. hd h
g on w s.e.
' T "*I
'wP SToPi'E b,
b 400 -
y LAST Tuso losser oa a d
v n to.i s c t o s s a s
' s' a p r p u,w p
[ Res rAa.T r=_ b 300 -
- '# ""1 N "P
\\.
Sropeab A ab l
200 -
REsrA rcip; )
7
(
j 3
100 - "
~ --
1
-M
\\l
=-
.m
- = -
a m
2 04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 TIME A
B.
C D
M/U TOTAL r#M ML
<!!L'4 rW>
"'WIl "
NTW
D 9
s' C
9
_oO
! 5
- o to 00
(
3 0o o)
R 4%
Jj h j<
r u
p A I O
a 3n ai e
6 a
u e
t
+
.W M
ft F.
g (J aD N
mQ 3'
N a
a s
4 W
Wx g
k*
a a
o$
z w
y
. o.
2x l
- g Fz O_s sc
(?
w Ex 2;
l h
!q a es o
x~
6 5
I M
O
/\\
l r
s 0
<C 3
8 c 1a
! o u.
Im5 I
ry 4b I
w1%
o kd4 6
i i
i i
i
.c C
0 6
O O
ggQh
]
] $QO f
N (OZH NI) lE page 105
\\
R EAC-~O R T R P,
D EC E V 3 ER.
2 6, 1985 MAIN FEEDWATER FLOWRATES 4.5 W
4-l 3.5 -
1 l
3-I E
I Rx Trip
( )
l 2.5 -
foacean7g R,,,,p gego 5
v 2-gf
/ Haw Sreat lm
/9/aws l
1.5 -
44se deru4 nod I
)
.l 1-I 1
1 0.5 -
'l i
m g
i O
3 04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 TIME A MN FD WFR B MN FD WFR IDADS f9601 IDADS F9600
R EACTO R
~~ R 3
JECEV BER 26, 1985 AUX FEEDWATER FLOWRATES 1.4 l'
1.3 - A'1A*!l O "'G Of '8 ' AAV
~N
/
1.2 -
gvevi N
l heO El O
3,,
N
/
Ze5 ben 1-(
a f
&^S7026*D SFAS dyinAnod
(
O.9 - Afon.g - Dav.zw AAv e1w/ /s "
y]
~ g O8~
/niffra AND
~~
a.
8g 0.7 -
Lu
- !g g Ba u 6x n> Ifat t
Oc O.6 -
B</s b
G'b
'8 'AFW VALs's U
/
4'et/ 0"#
0.5 -
04~
l I
%'Afw 14rius l
0.3 - Cicsix4 'A '
'4 5 0 " ^'
AAV // Alvi O.2 i
(
- POSSIBLY INVALID DATA i
y 0.1 -
h PROBABLY ZERO FLOW
-- = -
g O-04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 TIME
-B-A AUX FEED FLOW
-O-B AUX FEED FLOW IDADS F1660 inADR F1661
R EACTO R TR 3
DECEV BER 26, 1985 OTSG PRESSURES 1.1 4
Six ('oae J u s n,c.s* OOra/ p' EES u r 1-O.9 -
O.8 -
Cuisis's M VM U
.G 9,y _
(n ^
b O
y A ri v V,n ai s Po s s O P m o,g _
Wo Ma 5o O.5 -
2'07.K, Swrs E JPru &'E2 enj'
.Sram - %'Abub w-E O.4 -
n-tes wwan Restoneb Y
O.3 -
U O.2 - Cto:,ian AD/s Aes 78 Vs 0.1 -
O-g 04:00:00 10:00 20:00 30:00 40:00 50:00 05:00:00 TIME A-OTSG B-OTSG
.IDADS P9301 IDADS P9300
O 9
O 9.
m D
@i
~
O G
b d
S W
C
$b h
3 o
Q fU U
9
~
O g
F kD k
64 O
J T
b '(f Qg W
(O eg' Pa T
t
\\,
O O
OE Ng
. 9.
p5 W
O o,
- h I E.
Qg QS WO 1
Ob 8 si W
dg C
p O.- $
55S ef WK
- e i
a e8
^
i J
G8 O
D
$"m
.9.
J
[k ab5 O
W l
W c. E N
(:
l O
s S m.
a_
O 05
=
O I
O 4
O Om to
]
'g I "d I 5 W
3
[a EN
.<c
~3 r
m.
$Q d
bD 5
k k
O I
- w A:d y
I i
i i
a i
i i
O O
O O
O O
O O
O O+
l O
O O
O O
O O
O O
l 00 6
(D 0
t
'M N
e (OEH NI) 13A31 Page 109
O 9.
g O
00 M
-O
/
~
g (O
(N e O
T2 WS m
m.
3 Q
25 E
Wx Os o
Uo 9.
Ze e
a Wm o
do Os
's m
g w
o v;
W 10 2M
-S FW Tm M
u j
g w
2 m
T<
q-1 o
i
'O kW 24
- O W
.O (Y
i i
i i
i i
i i
i i
i iiiii O
O O O O O O O O O O O O O O O O O OO 65d4ida65565d4ddado J
ocoo t/ L.(7W/on) A.uAU.OV Page 110
Rx TRIP CO V PARISON PRESS-TEMP to TECH SPEC COOLDOWN CURVE 2.4 B&W LIMIT-Rx TRIP 2.2 -
- CROSS B&W LIMIT AT =0422:00 INTERIM BRITTLE j
- RETtRN TO NORMAL RANGE AT =0505:00 FRACTtRE CURVE
/
2-h A
/[
1.8 -
/
2 1c-un f
g.g 1.4 -
o8 ya 1.2 -
til o
[{
1-b O'8 -
NDT LIMIT
(
120 MINUTES E
TECil. SPEC. CtRVE AFTER Rx TRIP O.6 -
\\
\\
SATURATION O.4 -
/
\\ CURVE f
g O.2 -
A o
f O
O 200 400 600 TEMPERATURE Tc DEGREES F
- G --
-A-B&W LIMIT
-O--
SATl.IRATinH snanc Tante Tree s e nts' rac 1 1 9 o Dima r in O tt!
1 rnnu acur cTrau Taperc
O O..
- O i
[
)
- Om m
M3 i
z
!g si F
Om El m8
- 19. T(O $
io c.
- =
e H N 1.
- O
-e xN 4
<O N
z.m 1C 5
Ei o
O 22
.0 og hi Os CO 5
5 m
0 c
-t m
1 ".
-e
- o
[N
- O w
Hd is 3 w
m o
X o
1 b@
Y 9:
c:
H O
O e
m Ng I
e N
- O Ne
- O N
n i
I I
I I
I I
i 1
l i
I I
I I
I d
~
O O O O O O O O O O O O O O O O O O O I
O C W t N O O @ + N O C e + N O O e eooooo+++++nnnnnN N O
i i
3 see.15ec 3Mn1VB3dPl31 Page 112
,,1 O I p
o o
o e,
o 6
c,l is ca 2;
.: e c2 O.' Attachment 2
-d===.
C
?!
O:
I t
c NI 5l. b [ [3 [ O1 Dj)
O-
)
d
?
e
-=
D C
<i-(
c 5
C i o
o o
o e,
s las c6 as i e e
e C
e 4
l 0l 1
m q
W l
l l
l N l
i i
o, I
I l
l l
'l.
ld) l l
l l
o!
g
.N l l
l l
I l
I7 ci f
l l
l M
I c.
. m 3 n.vN l
l 1
E o p!&..=w.ww:c w.;c lJ--
- p.,e$5iG.sg
-T Y. :.e.. j.,
Q;%...,d., ty
=
O s. G,:! #r.. %. %..... R 4
\\
9 v;,.\\
. 9.,,.,m,., :.. g m.
e n.
- e."....
g f
l 5
,l lo.IoI
'i o.1 l,, -j l,,1 l.. l l.. l
^
g s!
ul j ap.la;
- 2 lnl ej l-70; l
l kl l
O!
i s
I A_
O f
1 j!$
0 l
t
, %o l
l I
d l
l 0 5
[l l
j O
gi I
I d
I I
l 0;
gI I
eh l
1 e
1,
+ =u I
T a3 C
ll kNE O
l e.
l l
y O
e, g
jaj al l :t p
e j+
c
"~
g I
o page m sN l
c:
N,y________._____.__L_LL I
I O
i s,
o LR-2 3 So 2.
[%lA. % k levef
/o m a./ k m 10 0
~ ~ T '
r
~ --" 10 0 t-i 9g j.
9g Il
.t 8O.
d 80 h,[F
^
7 l
- ' ~. * *.j.
.... f;o..
- i
- ~
i 7,0....
L.
y t
6 60 60 Lo co cn 50 S0 c=
CQ O
...4 0....
4o La.1 Q
C 3 0--
30 Q
00 f
2 0--
20 C
l to.
to -
i i
g
......g.
.g g
-0
_. L..._ _. _
0 __
),
pla.7 oncy %tvE ma.7 Ourter CPCHED k EECLO3ED
$ASWOD-l
'A ' HPI Puid J7A276b i
I Page i14
.I r.__
O 4
I 1 a
LP. 20 So *5 A CDM -0NNE W
NL
. ima a.. __ s._. t
._....... _ t... l u. u.._.. __.... n._.....
.... i
.l
.-.-..__.a.
_... _ _ = _._,...._.._t_..__..___
_..._.r._.._..__-._.-.--n._._.._
. _. ~.... _
- 1... _.
4.
... +..
_,(
,....._.._.._.u._.-._.__
n... _ _ _
..,. _ - - - _ ~
~
_--.v......
_. _...s s
4
~..
..6_....
._u
... ~..
- t. -. -......
_ _e 44, W
.-.,.g..
t
._s
. L L-
'."..i -
..-.j._
...Ia
...,..-.....,.gr.
.e e
g
. _ ~. _.....
_r
.4.-
_ 4% m. wy.
._..{.-.
f
....e
.........._. ~.
J..
...,.. n.... 6.,
n
.-.,._..L t
-~..-
s._.
,v.
g........._
.....e...
.p a
m
,... _..i...............
.J g
-_.-i
-w
.e..
[.
p......
L p
s...
.g.
-...m....
n-.....
._w9-...._.....
v
- _.. _4 4.....
g......... -.. _..
..-....L...._...
g
. &...a
.~.
l-I I==
). 4 j
. - I. =-. b
. + -.. ---
.. =...
q
.... L
.4.
__g_...
.g t
-. ii
,-h....e.
p..I..
g
...p.
.b
- 4....
j,
......~1..
s r..
4.~.
1..
~
........n...................
(
.s
...g......... _....6..... = _. -.
t an
. g
. ~.
.Aw.
...........l..-........ g..
l J.
\\._,......._._........._....,.......f...._. _...
e
.. -.. - =.
p
..p.
q
..... _. _.o..
...i.
e...
W r
n s
\\
=.
- r.......L g
..-q
.... e..
o......-....
.. _. _ =....
nm_ _
. A.,
..4
_-e g _.
,i
... u
~, _ _..
- ._.a_.__-.__. g Mff loss Cr Id5 %M 6pwsATE SA fffD Ck w M 9
8 o
0'MME N", Wo...
Lp 2OSo4 8 07S4
- *
- f A f 00
......i t.
.u....
.s.
.. wo.!
i.
l
,.e_.._.
._.i.._.........
.e.
6
..g i
...a r
4
.. _g
. e. c. c t.
i
~
i j
s
... s s...g. m i.'
.l i
i
- ---?
.I, 1
- a4 o
I
...o u
.....T...
..+
.. Q...J
.. g.....g.
......4.
I'
?
.n._
5- --
.h..
.L i7 i. L.*
t i
e a.
g g
..._i-
,.g i
l g
i c--
...J...
.,.....g
. ~.. _.
i 4p
..l. _. _
{
t O
..i.
.w
.J. _ _... _ __ q -
l
' ~
1
.O 50 h
= _.:- 5 0 j
1
-6
= _ _..
.Q
._.s.,
-i g
1 g
4o
,y l
- a. _
\\.._.
c p
i A
i a
I l
I g.
....'..........,.'.aT 7
t-.
a1 3n..
v l
g
- t...
.i_
4 p.
2,U
,.._sv._..._...,.....
6
.l l
t.- _.-....,....s.,.
..g t.._w..
... _ -._e..........
1. - --..
,y IU.
l
. 66L...
.h.,_6_.
i I
.~f--.
..a..
...10 a_...
....T _-.
~..._...._t.
._.~_,__,.___.3 d,
cats bss Of IC.S
\\
Ame hgDSMSA7*/ PVMP /~fsb
(
Page f16
O ATTACH ENT 3 AFW FLOW - IDADS
(
INCLUDING CALCULATED CORRECTION (*)
9 TIE "A" OTSG "B" OTSG (Hour: Minutes:
(gpm)
(gpm) 04:14:00 (NONE)
(NONE) 04:15:00 587.0 437.7 04:16:00 1114.
1050.
04:17:00 662.2 617.9 04:18:00 1248.
1165.
04:19:00
- 1304 1213.
04:20:00
- 1310 1250.
04: 21:00
- 1368 1224.
04:22:00
- 1758 730.0 04:23:00
- 1860 694.9 04:24:00
- 1756 678.0 04:25:00
- 1044 865.2 04: 26:00 106.5 990.0 04: 27:00 22.70 1010.
04: 28:00 22.70 1013.
04:29:00 22.70 1011.
04:30:00
- 1744 675.3 04:31:00
- 1712 667. 2 04:32:00
- 1759 652.0 04:33:00
- 1750 650.8 l
04:34:00
- 1960 457.7 04:35:00
- 1970 101.6 04:36:00
- 1989 93.64 04:37:00
- 1930 93.64 04:3CiOO
- 1950 93.64 04:39:00
- 2035 32.11 04:40:00
- 122 32.11 04:41:00 39.33 32.11 i
I Page i17
l ATTACHENT 4 i
UNITED STATES NUCLEAR REGULATORY C0peqISSION OFFICE OF INSPECTION AND ENFORCEMENT WASHINGTON, DC 20555 January 31, 1986 6*'
IE INFORMATION NOTICE N0s 86-04:
TRANSIENT DUE To LOSS OF POWER TO INTEGRATED 1
CONTROL SYSTEM AT A PRESSURIZED WATER REACTOR DESIGNED BY BA8 COCK & WILC0X 1
Addressees:
All nuclear power facilities ho! ding an operating license (OL) or a construction permit (CP).
Purpose:
i This notice is to inform recipients of a recent event at an operating pressurized water reactor resulting from loss of power to the integrated control system.
Recipients are expected to review the information in this notice for applicability to their facilities and consider actions, if aporopriate, to preclude similar problems from occurring at their facilities.
However, suggestions contained in this notice do not constitute NRC requirements; therefore, no specific action or written response is required.
Description of Circumstances:
On December ~ 26, 1985, Rancho Seco was operating on automatic control at a t
constant power level of 710 MWe (76% of licensed power).
At 4:14 a.m., power to the integrated control system (ICS) was lost.
The annunciator alarm for l
" Loss of ICS or Fan Power" sounded.
As designed, ICS deman,d signals went to midscale.
The main feedwater valves closed to 50%, and the atmospheric dump valves, turbine bypass valves, and one set of auxiliary feedwater valves opened j
to 50%.
The main feedwater pump speed was reduced to minimum.
Low discharge pressure at the main feedwater pump caused the motor-driven auxiliary feedwater pump to start automatically.
The net decrease in feedwater flow caused the reactor to trip on high reactor coolant system (RCS) pressure.
After the reactor trip, the above ICS valves remained at 50% (i.e., could not be operated from the control room) causing excessive cooling of the RCS which was exacerbated by autostarting of the dual-drive auxiliary feedwater pump.
During the 26 minutes required to restore ICS power, operators acted to mini-mize the resulting transient.
However, difficulties were experienced with i
manipulation of valves, operation of pumps, and control of various' liquid levels, pressures, and temperatures.
RCS pressure decreased to a minimum of
(
1064 psig at 4:21 a.m.
At 4:40 a.m., the lowest RCS temperature (386'F) during the cooling transient was reached.
RCS pressure at that time was 1413 psig.
Eventually, a senior reactor operator discovered that switches which supplied
sg Attachttent 4 Page 2'of 3 January 31, 1986
,e power to the ICS de power supplies were in the off position and set them to 1
the on position. Although manual (i.e., hand) operation was now possible in the control roce, the valves initially received a 1005 demand signal.
Opera-tors quickly shut the valves. At 5:00 a.m., RCS pressure and temperature were stabilized at 716 psig and 433'F and saintained there for 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. This unusual event, which was declared at 4:30 a.m., was terminated at 8:41 a.m.
Before the event was terminated, a large number of problems were experienced, including:
The RCS was cooled 180*F in 24 minutes violating the technical i
specifications limit of 100*F in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.
Recommended pressure / temperature limits for pressurized thermal shock were exceeded; however, nil ductility temperature limit in the technical specifications was not violated.
Pressurizer level was low and off scale.
After loss of ICS power, ICS controlled valves could not be manually operated from the control room.
One auxiliary feedwater isolation valve could not be closed.
One auxiliary feedwater flow control valve was overtorqued using the manual handwheel, and the manual operator failed.
Operators had considerable difficulty determining (locally) the position of the auxiliary feedwater flow control valves.
One steam generator was overfilled.
A main feedwater flow recorder in the control room failed at sidscale because of the loss of ICS power although main fee.dwater flow was essentially zero.
An RCS makeup pump was run without water (i.e., suction valve shut) and severely damaged, specifically, seals for the makeup pump failed and approximately 450 gallons of water were spilled in the auxiliary building.
A containment radiation monitor was damaged because it continued to run after the suction valve had been shut by a Safety Features Actuation Signal.
Four senior reactor operators were present during the event.
At 5:01 a.m., one of them collapsed from exhaustion in front of a control panel.
He was trans-ported by ambulance to a loc'al hospital and subsequently released in satisfactory condition at 7:00 a.m.
ATTACH'ENT 4 Discussion:
The NRC sent an incident investigation team (IIT) to Rancho Seco shortly after the event.
The licensee has agreed to hold in abeyance any work in progress or planned (except as required by plant safety considerations) until the licensee and the NRC have had an opportunity to develop detailed trouble-shooting plans for failed equipment.
Further, the licensee has agreed to maintain the unit in a shutdown mode until NRC concurs with the licensee that the unit can be returned.to power safely.
Review by the IIT is continuing.
As additional information about the event is obtained, this notice will be supplemented, if appropriate.
No specific action or written response is required by this information notice.
If you have any questions about this matter, please contact the Regional Administrator of the appropriate NRC regional office or this office.
l N
M ard P. ordan, Director Divisi(n f Emergency Preparedness and gineering Response Office of Inspection and Enforcement Technical
Contact:
R. W. Woodruff, IE (301) 492-8597
Attachment:
Li'st of Recently Issued IE Information Notices 6
0 0
e
- -, + - _.
= __
e l
ATTACHENT 5 1
4 i
i INPO SIGNIFICANT EVENT REPORT 6/86 l
t i
10 581 FLYNN (INPO) 17-JAN-9s 13:09 PT i
IvcJacT INFO O!GNIFICANT EVENT PEPOPT (:EP):
4-is i
j JJECT:
LOOO OF PCMEP TO THE INTEGPATED CONT 8CL ;Y0 TEM PEOULTING IN OVEPC00 LING TPAN0!ENT UNIT (TYPE)
PANCHO OECO (PMP) i DOC N0/LEP NO:
50-312 NA i
EVENT DATE:
12'26/95 l
N000/AE BABCOCK & MILCOX SECHTEL 2
PEFERENCEOs 00EP $2-7
.0EP 81-2 NOTE:
THIO OEP OUPEPCEDEO AND PEPLACE OEP 3-96 i
l OUMMARY:
j WITH THE UNIT OPEPATING AT 76 PERCENT POMEP, ALL DC CONTPOL POMEP i
W! THIN THE ICO WAS LOOT TO THE INTEGPATED CONTPOL YOTEM (ICO).
THIO PEOULTED IN A PAPID PEDUCTION OF NAIN FEEDMATEP FLOM.
FOLLOWED BY A PEACTOP TPIP DN HIGH PEACTOP COOLANT ;YOTEM (PCO>
i POE00VRE, AND AUTOMATIC INITIATION OF THE AUXILIAPY FEEDMATEP (AFW) SYOTEM. ADDITIONALLY. WITHOUT ICO POWEP, AFM FLOW TO THE OTEAM SENEPATOPS AND OTEAM FLOW THPOUGH THE ATM00PHEPIC DUMP VALVE (MDYr) AND TURR!NE SYPMSO VALVE 0 (TBYr> COULD NOT BE i
CONTROLLED FROM THE CONTPOL POOM.
AFM FLOW. TOGETHER WITH TH',0 OTEAM FLOW. PPODUCED AN EXCE00!VE AND *APID PCO C00LDOWN AND
( '0E00UP!0ATION SUFFICIENT TO AUTOMATICALLY INITIATE THE OFFETY l
\\
eTUPE0 ACTUATION OY0 TEM (CFAT).
HIGH PPE00VPE INJECTION FLOM PEOULTED IN PCO PEPPE00VPIOATION MHILE TEMPEPATUPE MA: OT!LL.
DECPEA!!N6.
TH!O EVENT 10 O!GNIFICANT RECAU0E LOO:: OF DC CONTPOL POWEP WITHIN THE ICO PEOULTED IN A PAPID AND EXCE:0!VE COOLDOMN OF THE REACTOP COOLANT OYOTEM AND PEACTOP VE! EL.
THE i
POE0!UP!EP LEVEL WAT OFFICALE LOW AT ONE POINT IN THE TPAN0!ENT
'___AND LOOO_OF,PPE000P10EP_FUNCT!ON ALN007 8E"ULTED.
DE;CPIPTION:
ATTACHENT 5 THE PLANT WAS OPEPATING IN A OTEADY ? TATE CONDITION 47 76 *EPCENT POWER.
AT 0413, +<- 24-VOLT DC CONTROL POWE5 MITHIN THE ICO MAI l-T,PEOULTINGINTHEFOLLOWING8 o
MOST ICO DEMAND SIGNAL; WENT 70 MID-!CALE, COPPE!PONDING TO CEPO VOLT 7.
THE IMMEDI ATE EFFECT NA! THAT TWE :TARTUP AND MAIN FEEDWA1EP CONTROL VALVES CLOOED TO ABOUT 50 PEPCENT, THE MAIN FEEDWATER BLOCK VALVES CLOOED AND THE ICO-CDNTROLLED AUXILIARY FEEDWATER CCNTROL VALVEO, THE ADvs, RND THE T3Vs OPENED TO ABOUT MID-#00! TION.
o MAIN FEEDWATER PUMP SPEEDS DECREATED TO MINIMUM SPEED (APPROXIMATELY 2500 RPM) RESULTING IN INSUFFICIENT DISCHA#6E HEAD TO FEED THE STEAM GENERATOPS.
o THE RAPID PEDUCTION IN MAIN FEEDWATER FLOW PP3DUCED INCREASES IN PCO TEMPERATURE AND P#ESOUPE.
THE REACTOP l
TP!PPED ON HIGH PCO PPE00UPE ABOUT 15 OECDNDS AFTEP THE LOSO OF ICO DC CONTROL POWEP.
AT ABOUT THE !AME TIME, AUXILIARY FEEDWATER INITIATED DUE TO LOW MAIN FEEDhATEP PUMP DISCHAPGE PRE 000PE.
BOTH MAIN FEEDWATEP PUMP! MEPE MANUALLY TRIPPED.
a
!!NCE THE LOOO OF ICO DC CONTPOL P0uEP *EPOOITIONEI THE ICO-CONTPOLLED AUXILIAPY FEEDuATEP FLOW CONTROL VALVE:. THE ADVs. AND THE T3Vr TO PA*TIALLY CPEN POOITION!. THE UNIT SEGAN A PAPID COOLDOWN.
EQUIPMENT CFEPATOP! WEPE 4
DIOPATCHED TO MANUALLY CLOOE THE ADvs. TEvr. AND ICO-CONTROLLED AFW FLOW CONTPOL VALVE 0.
< RANCHO IECO DCE: NOT HAVE MAIN STEAM 100LATION VALVEO.)
i GC0 PRE 30VRE DECPEASED DUE TO THE PAPID CCCLDOWN. AND THE IAFETY FEATUPE ACTUATION SYOTEM (OFA!) INITATED ON LOW RCO PSE00VPE.
AS A PEOULT. THE IFAC-CONTROLLED AFM FLOM CONTPOL VALVE: CPENED FULLY.
THE OPEPATOPO 700V mRNUAL CONTDCL AND CLCOED THE !FAO-CCNTROLLED AFM FLOW CONTPOL VALVE FSCM THE CONT 50L ACOM 70 REDUCE THE EXCE00!VE FEEDWATEP FLOW TO THE : TEAM GENEFATCPO.
THE !NDICATED PRE 00VPICEP WATEP LEVEL MENT OFF ICALE LCM.
- TEAM i
GENEPATOP PPE000PE DECPEASED TO THE POINT WHEPE THE CONDENOATE CUMPS BEGAN FEEDING EACH OTEAM GENEPATOR.
FLOW TO THE TEAM GENEPATOP0 INCPEA0ED BY AN ADDITIONAL 1000 GPM FDP A OHORT TIME.
THIO CONTINUED UNTIL MAIN FEEDWATER 100 LATED AUTOMATICALLY WHEN MAIN OTEAM LINE PRE *0UPE DROPPED BELOW 415 PTIG.
THE EQUIPMENT OPERATOPO MANUALLY 100 LATED THE ADVr AND TEvr EUT ENCOUNTERED DIFFICULTY MHEN ATTEMPTING TO CLOOE THE AFM FLCM CONTROL VALVEO.
AN OPEPATOP PARTIALLY ONUT THE "B" AFM VALVE.
PELIEVING IT MA7 FULLY CLO3ED.
HE THEN OHUT THE "A" AFil VALVE.
O!NCE HIGH RE0!0TANCE MAO ENCOUNTERED PEFDPE THE "A" VALVE i
3 POO!! TON INDICATOP OHOWED THE VALVE TO BE FULLY CLCOED. HE i
CONTINUED TO TUPN THE VALVE OPEPATOP !N THE CLOOED DIPECTION u!TH A VALVE WPENCH.
THE VALVE OPERATOP FAILED, AND THE VAL /E OEDPENED.
HE THEN TP!ED TO OHUT THE MANUAL !!OLATION VALVE DOWNSTREAM OF THE "A" CONTROL VALVE BUT MA! UNAPLE TO MOVE THE i
'LATION VALVE FPCM ITO OPEN P00! TION.
A *ECOND CPEPATCP
.!VED AT THE "B" AFu VALVE AND l HUT THE VALVE THE PE;T CF THE WAY.
l WHILE THE AFM VALVE MEPE PEING MANUALLY CLOOED. OPEPA7080 IN TWE i
CONTROL POOM DETEPM!NED THAT THE PCMEP OUPPLY IPEA>EF
<:1 AND
' 02) FDP THE ICO WEPE T*!PPED AND PE TORED !CO PCMEP BY PECLO:ING wA* THEN CLO*ED FECM THE CONT 50L i
DUF1NG THE CCCLDOWN FEFIOD. TWE FC0 COLD LEG TEFFEFrTUFE DFCFSED RBOUT 190 DEGPEE3 FAHEENHEIT IN 24 MINUTE 0.
THE NIL-DUCTILITY
'EMPERATUEE LIMITO IN THE TECHNICAL OPECIFICA" ION! MEEE NOT
/IOLATED.
. CAUSE OF THE LOOO OF DC CONTPOL POWEG TO THE ICO II NOT YET k NOWN. -
COMMENTS:
- 1.. WHEN AUXILIARY FEEDWATER, ATMOSPHERIC DUMP. OR TURBINE BYPA!!
VALVES FAIL OPEN, PROMPTLY STOPPING THE AUXILIAPY FEEDWATER PUMPS COULD APPRECIABLY SLOW THE COOLDOWN TRANOIENT.
THIS !!
ESPECIALLY TRUE IN PLANTS LIKE RANCHO SECO THAT DO NOT HAVE MAIN STEAM ISCLATION VALVES.
2.
AT RANCHO SECD, A RECENT MODIFICATION HAD BEEN MADE TO ENABLE THE TURBINE BYPAST VALVE 0 AND THE ATM00PHERIC DUMP VALVE: 70 BE RAPIDLY CLGOED FROM THE REMOTE !HUTDOWN PANEL (CUTOIDE THE CONTROL ROOM), INDEPENDENT OF THE AVAILABILITY OF ICO POWEP.
THIS CHANGE WAS MADE TO ACCOMMODATE PGOTULATED CONTROL ROOM FIPE SCENAPIDO.
THE CONTPOL PCOM FIRE PPOCEDURES INCLUDED INSTRUCTIONS FOR OPEPATION OF THE VALVE!.
AND GPERATORO WERE TRAINED ON THE UOE OF THE MODIFICATION.
HOWEVER, THIO PAPID CLOOUPE PEVIOION MA! ONLY MADE TO THE CONTPOL POOM FIPE P90CEDUPE0 AND NOT INCCPPCPATED INTO OTHES 900CEDUPE0 THAT IPECIFIED OF PEGUIRED GPERATION OF THE E VALVE! FACM OUT0!DE THE CONTPOL POOM.
THE OPEPATOP0 DID NOT PEMEMBEP TH!! ALTEPNATE METHOD OF VALVE CONTGOL EEFDPE THE VALVE 0 HAD BEEN MANUALLY 100 LATED.
uwEN ADDITIONAL CAPABILITIE0 APE ADDED TO A PLANT. A VAG!ETY OF PPCCEDUEE0 GTHER THAN THE PRIMARY APPLICATION CAN OFTEN BENEFIT FPCM THEIP UOE.
THE PLANT PPOCEDUREO OHOULD THEREFDPE BE PEVIEWED TO IDENTIFY OTHEP PDTENTIAL APPLICATION!.
I.
TPAINING PPCGPAMO FOR OPEPATCPO lHOULD INCLUDE AND :TPE 0 THE FULLOWING:
a.
ACTION PEQU! PED WHEN POWEP IO LOOT 70 CPITICAL FUNCTICN*
CP CONTROL OYSTEMS 7UCH A! THE INTEGPATEr CCNTROL :YOTEM THE CONOEQUENCEO OF LOOING PPE00VP!OEP LEVEL AND'n4E s.
CDPRECTIVE ACTION: TO BE TAkEN TO RE070PE PRE 00VPIOEP LEVEL c.
THE CONCEPT THAT WHILE ADEQUATE CORE COOLING 10 CF PARAMOUNT IMPORTANCE, THE POTENTIAL CONOEQUENCEO OF AN EXTENDED. RAPID C00LDOWN MUOT BE UNDEROT00D.
IN THII EVENT. THE OPERATOPS WERE RELUCTANT 70 OTOP AFW PUNP; BECAUOE OF RECENT EMPHA7!! CN FPOBLEM! A000CIATED u!TH LOOO OF AFW FLOW.
c.
IN ADDITION TO FORMAL DIAGNO TIC TRA!N!NG, OHIFT j
SUPERVISION OHOULD ENCQUPAGE D10CUOO!ON DMONG OPEPATCR0 OF THE VARICUO COUP 0E0 AN EVENT OUCH At rH!! CNE CAN FOLLOW.
IT !! THROUGH THIO k!ND OF DIOCUt010N AND "WHAT IF" ANALYOIS THAT QPERATOP! APE ABLE TO RECALL THE t
AVAILABILITY OF ALTERNATE METHOD (OUCH At U!E OF THE CONTROLS AT THE RENGTE ONUTD0uN PANEL / AND APE ABLE 70 PECOGNICE THE DE0!RABILITY AND ACCEPTABILITY OF TCPPING AFW PUNPO IN CERTAIN O!TUATION!.
a.
THE IMPORTANCE OF THOROUGH TPAINING ON ADDITIONAL e
ATTACHENT 6 I
w a =wc6s...th..,c=. :.,
.sa o=
=.c me
.e oveo oue no t*es sim
"*"8'"
UCENSEE EVENT REPORT (LERI
..,uu...
..a i
u,,...
mANCHO SECO NUCLEAR GENERATING STATION UNIT NO. I o I5 t o io I o i ?!1 i2l 1iod c #3
.. 6g e.e REACTOR TRIP FOLLOWING LOSS OF ICS POWER ev.
n.
, u n,,,
o,....ciu riu... 6v e....
r.
ou
.u.6..--.-=
occ ur w - u +
r.
u.
.. l.... l n!.';p -
a,;;;;;
oisicioici e i 1l2 2l6 8l5 815 Ol2l5 Ol0 0l 1 2l0 8l6 NCNE o,s,oio,c.
~
m. =.
m.
v.v.........,c-..,~....
g m....
r
~,....
=
,2 u _,,
y e y ; g,..g,.
n,n4
~.,
.n.
y seisw.n.a l ] se,s.nsn nas j
m.s
.. s a nm e se,,unan
..e.
g y es.ts_asm n
s_,
=
s.
se.n w I se.tsmaanw l l se,si.nsn..
6.c.
.ecom.c,.o.,=s6s. nei s s.-c=es..e.
- .... u
- e.
Acn W. Colemco, Regulatory Comolfance Superintencen:
,e,.,4l4,g,7,,,.
7..,3 e
m u.....
.ci. ca
..,...u... o ne..n.............
l
- .atge
{ep:;,*ga
{,,,,,(,,,,,,j,,,,,,,,,
, jg,
., p,;
- 4.islsti.i es ou=,
l l.!
t a
Cl 3 i i ' A 9 !S,8 t 0 Y
l,I l,,,l I,
i,
.u u.
6..
o n e,..n.
i
.e...
bo l
Iweu.58ic9
!o1 I 1,7 Ie.6 7.u u,
,me,o...,.. u n, On Decemoer 26, 1985, at 0413 hours0.00478 days <br />0.115 hours <br />6.828704e-4 weeks <br />1.571465e-4 months <br />, the reactor triocec on nign reactor :: clan- (;CI:
- ressure initiated by a loss of electrical cower to the unit's integratec c
- ntrol system (lCS). Prior to the trip, the unit was ocerar.1ng in a steady-state c:nci:icn n '6 percent power.
The loss of power to the ICS resulted in all ICS demand signals assuming their mic-scale values corresponding to zero volts. This caused the main feedwater (MFW) valves to close to their mid-position, while the auxiliary feedwater control valves, atmoseneri:
dumo valves (ADVs), and the tureine bypass valves (TBPVs) opened to their mid-oositien.
The_ closing of the MFW valves reduced flew through the steam generators, thereoy in-creasing RCS pressure and temperature and producing the hign RCS :ressure trio.
!eca.se the ADVs and TBPVs were partially open and auxiliary feedwater initiateo, the clant underwent a rapid RCS cooldown exceeding the Technical Specification limits.
The safety features actuation system (SFAS) was actuated on low RCS pressure and cerfor ed as designed.
An " unusual event" was declared by the site at 0430 hours0.00498 days <br />0.119 hours <br />7.109788e-4 weeks <br />1.63615e-4 months <br />.
8 ewer was restoreo to :Me ICS at 0440 hours0.00509 days <br />0.122 hours <br />7.275132e-4 weeks <br />1.6742e-4 months <br />.
The reason for the power loss was a malfuncti:ning cower succly voltage monitor module in the ICS.
This module is designed to trio tne ICS OC cower r
supplies when they have ICW output voltage; however, it triDDed the powe sucolies anen voltage was normal.
During *Me event, a make-up pumo esceived extensive damage as s result of a valve lineuo error.
The ccrrective actions to Drevent a recurrence of *nis svent 41
e suc.~1:*t0 'n 1 suaalement to this recor*.
~*s.. m
ATTACHENT 6 u 8 *wuSee.gh..A*;a t CO*me.seS40's e.C 8e.e 3ama UCENSEE EVENT REPORT (LER) TEXT CONTINUATION (48'agg 4 3 a
- +o. : :ve c : s:.:m
.J. v
...e..,
y....o.
4.,:
.1CHO SECO NUCLEAR GENERATING
*
- j [ " }y j
i STATION UNIT NO. 1 o is lo lo ol3!1:2i315 O! 2f 5 H 0101012 'C' O 13 mn ~
eo man,m On December 26, 1985, at 0413 hours0.00478 days <br />0.115 hours <br />6.828704e-4 weeks <br />1.571465e-4 months <br />, the reactor trioned :n high reactor coolant system (RCS) pressure initiated by a loss of electrict.1 power to the unit's integratec c:ntr:1 system (ICS). Prior to the trip the unit was coerating in a steady-state concition at 76 percent power. The reason for the pow'er loss was a malfunctioning power succly voltage monitor mocule in the ICS. This module is designed to trio the ICS CC power supplies when they have low output voltage; however, it tripped tne power su:clies when voltage was normal.
The ICS provides the proper coordination )f the reactor, stea:n generater feecwater ::n.
trol, and turbine under various operating conoitions.
The c:ntrol mocules usec ::
accomolish this function utilize a -10 to +10 VOC signal ange.
Upon :ne Icss Of I'S cower, all ICS demands went to micscale, which corresconcs to zero volts. This causec :ne main feedwater (MFW) valves to close to their mid-cositien, wnile the auxiliary feecwater (AFW) c:ntrol salves, atmes:herie :uma valves (ADVs), anc :ne :urc'ne :y ass talses USPVs) openec to :neir mic-cost:1en.
The cicsing cf :ne VFW valves recuced ficw :necugn tne steam genert: Ors, :nere:y 'ncreas-ing RCS cressure anc tem:erature, anc crocucing :ne nign ;CS :ressure tric.
A: 3::r:x1-nately :ne same time AFW :umo P-318 was autcmatically startec on Icw feec:um: cisenarge pressure.
' cause the ADVs and T5PVs were cartially opened (as a result of ICS pcwer #ailure), anc
,e AFW flow initiated, ne unit uncerwent a rapid RCS coolcewn exceeding :ne Tecnnical Specifications Section 3.1.2 limits.
During the ecolcown period the RCS tem erature croopec frem accroximately 582*F :o 386'F in 25 minutes.
As a result of :ne acic :::1-ocwn, RCS pressure cecreased celew 1600 asig anc tne safety 'eatures ac:uati:n system (SFAS) was initiatec.
During tne event equipment opera:ces were dispatenee to reanually close the ACVs, ~SPVs, and AFW valves.
The operators experienced difficulty in etermining that :ne "A"
AFW valve was closed and acclied a valve wrench. As a result, the valve operator fa1*ec anc the valve recoened.
An attemot to snut the manual isolation valve downstream of :ne "a" AFW valve faileo.
At approximately the same time (0440 neurs), Operations personnel determined that the ICS power sucoly breakers were tripoec and restored power oy ecicsing the breakers. Technicians had previously checked the power sucolies at accreximately 0a20 hours, but had not ooservec :ne trioped breakers.
The "A" AFW valve was nen :1:sec from the Control Roem.
At 0430 hours0.00498 days <br />0.119 hours <br />7.109788e-4 weeks <br />1.63615e-4 months <br /> an " unusual event" was declared and the drecer state anc county agencies tere notified.
The NRC Outy Officer was notified at 0442 hours0.00512 days <br />0.123 hours <br />7.308201e-4 weeks <br />1.68181e-4 months <br />.
The " unusual event" nas terminated at 0841 hours0.00973 days <br />0.234 hours <br />0.00139 weeks <br />3.200005e-4 months <br />.
During the subsecuent plant recovery, trike-up pumo P-236 suffered extensive damage wnen coeration without suction resulted in the failure of the pumo seals.
Acoroxtrately 4C0 gallons of water from the make-uc tank scilled onto the pumo room floor price te :ne ' low Jeing valved off. Minor atrborne activity was released to the atmosonere througn :ne miliary building ventilation system.
An Auxiiiary Operat:r (A01 and !:ut:-en: Attendant (!;) t. ice made e e'gency en:r es '.::
the make-vo rocm, ' irs: :: isola:e re aiake-up flow, anc seconc to assess.ne :: :
camage. No respira :ry protection as acen during :rese entries wnien :::al'ec 4::r: -
g :......
ATTACHfENT 6 o
w g etwChla..ggwga;;e. CClwese6460=
.*f eere Jena UCENSEE EVENT REPORT (LERI TEXT CONTINUATION
- Eev es :. e. r w.e u uaer.
occa et asus.o.is ist gae seuseman ses ease y
-acats f v messe e,e i.. I i"h'. ' t' i ::'c.*.J::
..aCHO SECO NUC'. EAR GENERATING i
t 131112 At 5 iml 0!2 i 5 ll-; Oi 0l, O t 3 ios 0 i 3 STATION UNIT NO. 1 o ts io io Io rw..
-. - - e -, e m mately six minutes. Both the A0 and IA received whole body counts during the.'orning of December 26, 1985. The results indicated I". MPBS (maximum permissaole bocy burden) of Ag-110m received.,y the A0 and 0.3f. M.cSB of Ag-110m received by the EA.
The external exposures received by the AJ and EA were 20 and 10 mrem, respectively.
Also, the Senior Reactor Operator (SRO) ' collapsed in the Control Room at 0452 hours0.00523 days <br />0.126 hours <br />7.473545e-4 weeks <br />1.71986e-4 months <br /> after a : erica of strenuous activity curing the event. He was taken to :ne 'lospitai Oy amou_ lance, diagnosec as naving suffered hyperventilation, and releasec.
Both Saccock & Wilcox (B&W), the unit's nuciear steam sucoly system vendor, anc E?R* :er-formed analyses to assess the effect of the RCS transient on :ne structural integrity of the vessel. The results of both analyses demonstrate that the reactor vessel nas acecuate structural integrity 'or return to service witnout further evaluation.
- n accitien, S&W
'9as c0moleted its initial evalua::en Of the fuel anc primary system comconents anc nas conclucec that no camage to the rimary Oressure :ouncary or 'ue; assamoi:as, <<nicn
..cuic Orecluce a return :: Ocwer ::eration, resultec frcm the transient.
' he District is centinuing its invest gation into this event anc will ;revice :ne NRC's Region V :ersonnel with a triefing anc root cause analysis of :ne event orter ::
e: urn-e ing the plant to cower operation.
A succlement to this recor: cetailing :ne ::rrective actions taken to revent a recurrence of this event will be suomittec following :ne moletion of :ne triefing and root cause analysis.
I is being submitted in c:moliance witn paragracns 10 CR 50.73(a)(2;(i,
This recor:
50.73(a)(2)(iv), anc.50.73(a)(2)'v).
e I
O i
I l
1 l
a..
=
~
LEGE.
SU = Startup Required LT = Lcng Term STATUS DATE 02-10-86 PE = Power Escalation DECEMBER 26, 1985 TRANSIENT TIE 0800 NA = Not Applicable Page 1 of 19 ST = Short Ters
- ACTION LIST -
- Closure Report Requ'. red 1;o NRC
- Troubleshooting /Repa' r Gudelines Per GAC 85 1001 Apply DESCRIPTION RESPONSIBILITY SCHEDULE STATUS WR No./NCR/
C0t#ENTS ETC.
~
c 1
w i
4 o
@ild :/
1 A
d m
1
{
s e
o
~
n~
t t' 143 w
93 a5p 1
/
ey #
j,-
...m aJ w
A s~
x g
w q
- o Q
A '74 k
h
. p-- t,jd
&.--A
( n ww ex
- _Perssugjll, g w
t 1
K
.= =- A I
T g-
.P
\\
f sm
+
$'t 0
v A L A
8 HP1 h3 4
=
u1 -
papon
1
['gE
);
CK Oh j/
j Q
E J,
@4 g
4 r
O O
a ohoj ai!
s
}
Nsf
.3 p.
(
k)
!t+J 9'e
.: zi a
(
a 4
{ (A h b
q N'.
4 41 6
tu
[
b
~
...3 p Q
?
5).xms/
P4
..:]3
- x l
g
/
J O}Be i
~
gd g :k x
A-9 Y,-."g X
u 5
i h
r' k'
kh$
BYI j
HPZ y
e l
x N0f'ON HPI WFCPON
i 0 "y k(kh hhii-2
'xp f<g Je a
e o.
e o 5}
ils s
}
[
hA.
g
\\
1 Py
\\ kki) o I
N k
l l
g W
X_
N
_1
.[
I 8
-/
k JA45 J
J
~k l
k
-- ^Q, X m
sns Ag F3 T
~?
E
(
L S
q j
u m HPZ c'
Nf07' N NP Idif4 DOM
1 k
(
S e4 9
v Q
! k!? *E " !+----@
m 3
h, Cltf d
1j 3
w o.
t d
is' l fla e
o+m e
i e
0 j
i
(
4
' !! *3 n
is g
N
,{-Xh 1
od
-/
y n) 4 q
sns k-7-NMk b
3d l
T, e
nee.ssuaa a X
AfAS A&ft NAMNP
--A
[
l c*'
0" mue g
2-s
~a~M e
\\
ZwfdPCA HPI IWf4 ROM
&g v
qy
.y ;e e
- g h
I~
c o ag x
4) fla
]
3 4
kN q
d 4 k I")
'p s
s x
k
' I!
%)
l 43 h
\\
\\
( 0 b
- $4mum/ _.
(l l R
) ms
}f mh"R Yh' W D
r LIW sns t
ma x
k
%,"=
. [h[
(fi t
e z
u miN [
j NPIa x
m
...>M.
. pr.
r
.s.
- h; - '". ~. M... del.. ',,.... '
-.7
-....I ATTACHENT 8
,, ; r_.
> :r:. ;..,
k
.1 5 q p
.. % w n -"
,y:.Qf=g 1
.g'
. g(-
}
g(
)
y ~
(:;._'
a
_ g
$ g ".'..?,.$ Alt 4 >
t >
-Gtt::.QF.
\\\\,g
.c..,.
. - lp ~lc* g w.,u i
6' '
t 4
m W
'N.
?
uag'n l,
ji oKiig eK! e'
-o w
D e
4
=g
.C
.2
- -e
%g lE*---
w N
hM E
E g
ps=1 e
i gg.
l I
3 8
I'IE=
S W
M s:.l lllla-s
~
N E8 c-
'1.$
og' S,R 3="
l[l5 li G
r.
rn
=-
gypq A m cni n
... % y Q, w q,k
...3,-
g
~-
- - ~ -
>g. -
w.+
- m.,, _ _.
c.7('
)
- (
)5.>,3..
5 2N i
i a
4 E 4 I
, : k'._'-
g 4h
- ;j ';.'
- m e. amm
~
ammess. - ee amme amm am e m m en.
[gg e
.ggg;
=
..y.} :f.v
-. - -_ : :L+.
. w.
e
'k
~
@!lq!IIIGi(kl5,I e]
[o 03[li at II e
.m e
n
- k h.
[
@s u_ g.
5 5
5>-@W l
E E~
l e
m_
-s hd s
ti e
QM 3
1 s
g R
38 E8
{l Ig*
X e
-N f
, r5 r n A Il~-
I lls 4
s n..,-,---
,- --,--- --.-,,,,...-.. ~.---- -- - ----
n m
m u
u
~8 8 7g. M. $.
e 4 4 4._4 4.,_4....4 4
,6
=
g 9
x l
^
O O
lM
- e-p-
-le*
- i s >
g *-
s-L f
e o- -o o -
Ev7. Y Y
'8'
.E' l
c r$,,
^
UN g ECB $188T01
)
i gf
~d ITPICAL(4) l N1 s.,_.
/ ek-L eq :m si _j :
\\m..f s
J oi
<,=
s l[ \\
O j.
B-) :[IH :G-: :
i I
f f
s~
s i3
,b
. I ngl'[
f f*
34 2-gg l-6, gg I
b r
l
- -*4-~
- s - --
b b
b 5
D G-4 C I
3 S
cse.a,ime..t.
/*
.o 3 sus.v vuv.i see l (- 3 M i[
-y g
M I
E p;
o e
o o
i
" " '** g,%.::,'.v.=.' :.
g N
V Y
MAIN STEAM SYSTEM
- s. t e ca *
'lM x
M RANCil0 SECO MAIN STEAM R
'~
SIMPLIFIED SCHEMATIC
G VITAL tus ic
- peoog-
- t. Sus IJ (C07)
(J04) t ABJ ll$00ICS) 184 VAc
__N ass yac h
i
,.) s g ------
) a 3 --- -
l l
__n I
I l l l
\\
l l
V V
M M
+
t34
-24 t 34
- 24 e cAs e i Voc yac ypc Voc
_f_0WE.a_ a40N4,i,0M,
- "" " "I#8I M
t t -
t t
e cAs 4 3 g
6 Y
? CAR e 8
$7
- (
jg
- AuCil0NEEfilNG p60ers l couuoM SUS l
. - 24 VDc aus
[__ _._ _I_.
t 24 VDc sus POWEa AucT4044EEA cAtle4Ei ( 2 d?
x M
\\
\\
l
\\
\\
\\
\\
2 w
cAs e I cAs e 3 cAu e 6 I
l lNTEGRATED CONTROL SYSTEM l
POWEP DISTRIBUTION
ATTACHMENT 12 1F Sus (118. VAC)- c S4C e
n IJ Bus,(118 VAC)
J L
\\- _ \\ C33 J
J
.g i
I Trude i i
~
swi tc*.r i
I I
l P
I L_____l VNaiBus1C (118 VAC)
)Cas
~
l Id l (Not in Invertar servica)
~
l (s1
' sz ICS AC POWER SUPPLIES; y.
V To 2 OC To 2 OC power supplies power supplies t
y IC3 AC power (E/P convertars) s
~
n MoowicAT t ord I
8 e
i.
- e
^*
t 1
1 A
l centkusorsu f
ad uss or,
3 sesrowsA i
niecuA swaren selecroA sweitN t.acat. srtscred fA nu artAv/
\\1n ELS ou NIR1 eN N150 SWETtN(AW6 oNLY) 1.(LoJE FALVE 5 f.(tosE VALVES
- s. ctosa intvfi
- 2. A6410 ft0$E OM toss
- 2. YALVEG DN AtiTO 2.\\fALVES ltt Au fe er scs rowsn 3.DD$ABLE AseTo clos (
n 1
U l'
Atj[M 58tENotD VALVE 5 y
DEEME268tE.To IDAny ClosE ADV's 04 TSV\\
u
$4 s
T T
T,,
d A =
~
E
(
1 '
EEb$
ea <
w son"d x2 IT Cf 2
g m
F
=
--t OJ e
e
D ATTACHMENT 13-(
hlM 00rTLC MiR SorTLL ICS ICS T
T
!. '#(
fn T7 aa l
~
l l
A 3-l M obistic.AT I O N l
A 3-I I
I I
,uro-ct.are" ;l I;,
H250 l
l HS I
,=
Hs
-2 1
il l
l ll "Ct.osE 4
~g ALARM -
f lI
" Auto c t.o M 10A05
- p. _.
gjg Old l
Qorrzz I
ocfril A:.3-A:.S '
I
/,
O O
$3_g
~ -
~
E/P l
E/P 1
L ICS ICS
- Solexeids Normnlly CNeryIkCal
(.S I A$ I)
{
- .rhAbt Alnrm nsoulkes $r..co/ exes d p w er TSV CONTROL 1
s c s t- - - - - o - - -
{'
og
/P l
ALARM l
PY
^
J L4
- A.S.
g A s.
l I
o i
hN P y - 20S'il A o----
--.-_------._--.--.o E/P
- A.s.
MoDIFICATIO N
[
A.S l
u pN g---.--g ry-205Fla g-_______._____________.__p.)g i
ggg l
_ _ ys l
E/P l
%8 O
i ^~'
l n
J L j
I" roetose'I Q.S ;
PY r- -
A. h,.,
MANUAL l
t LOADING 4
u * -
==l pq f
STATION P V - 2057 s c Hs H2SD
P HS "Ik/TO - ClosE..
l PV-2Ob62 A
,t------------1 av1'o -clOsc
- ,j l
6 I
l 4
PV - 20b 62 ts f 6- - - - -- - - - ----4 I
I
[J Lf a
MA AL 4
PV-20b62C t--------------
N L OADING STAT BON J YPLC A.L.0LI tILal!OY6._.YALYL1 Qg ADy_M.A.Huai con [ligt_ SIAL 10H_
=i U
ALMOSEllBlG_ DUMP VALVE CONTROLS
ATTACHENT 13
~
'IMI-l
(
foyAcp._ + _____________ + _____.,______,
I H 2PS 1
[
lH2F I
I MAuuAL 24 V I
l CONTROL POWER LED St6uAL SU P Pt.Y I
i (4-20 M A)
I i
IcS O
i l
M AWu AL - RED I I
I cs 20 VAC F-,
1 AUTO-6REEMI I
I 8
I l
t I
i I
ccMTAcT l
+24 i r- - _ -
OPEuau i
1 1
I I
I
' LcSS OF
-24 I I
ICS Dd i
M ACTS M Au. d AUTO. MAME AUTC.
P ESE1 LD6s OF i
I H 7.FS ICS Ic$ POWER 9
I AUX. REL AY l
8 8
8 MAMUAL L_________o____a j
gg7ggg l S7ATiou Slui-l
,. - - - - _ J iso VAc f - - - -
,i i
I RED Ll6HT HtPS DM WHEM Q
AUX.
I 50LEuclD gggAy i EM ERE12P.D 2
_S I/
HISS O P G M j- - -
"5 pKE AFW p
- LOSE:
f 3
4 16.8 OPEM d
,l' E/p
,l'
,l' f AIR SUPPLY VEwT-r 0-AD*
r 3
V g
A AFW ccMTROL Po6 I4A
' AIN VALVE
> SUPPLf I
l l
AUXILIARY FEEDWATER CONTROL INDEPENDENT OF I.C.S.
ATTACHENT[14. --
nrc g.
- n..
'4 04APMRAGM CASE ASSEMBLY
)
o APHRAou l
1 31
' N
// d '
j.
_ n n
J YOKg ACTUATOR SPRING
=
l ACTUATOR STEM m""
E l
l
/
~
' hk 1
TRAvat inoicAroR
!Nl TRAVEL INDICATOR l
i
/
scALs
, w AssameLY
/
/
ATMOSPHERIC DUMP mp
/
I VALVE /TJR3 NE BVPASS m Nur # r"
(,js VALV E OPE MTOR
s-
.-----,ne a
__m
,_L 8
5 ATTACHENT 14, ' - -
I I
8
,=.
I Il /
I I
I i
IE lj{
E I
E E
E
- gRe
\\,
e r n c
m-o,yg gllg A
% ^/
i.,,,.
I
\\
CO3 I
RL=
\\
A E
/
l l
/
O l-V 2
I.k
,k AFW CONTROL d
VA'_VE MANUAL 4
ffANDJACK l
, qye$
ATTACHfENT s..
->:.w.LT,-
k-[f')M:
Ci-f.
- e.,
sHANownart.
uppen asAnimo YOKINuT g
onsasa mvine towan esARING
- i ab
[b
=
U at.A m o s u s HI N G f'IY f
1 j
LANo EYE Bot.T g
,g gp-, e -
. _.T 9
f' 1 f
i T
sACKSEAT kh
' YOKE sooy
/,
//
/.
J "
///in
. ~ ~ s
~ ~ +
//
(
AFW MANUAL ISOLATION VALVE FWS - 063
o ATTACHENT 15
(
OFR SIMJLATOR RUN OUTLINE RE.E 1-3-86 LOSE ICS PWR INITIAL POWER: 75%
EIEltI IDE DE1TA T START 3:10 0
AUX PLMPS 3:10:10 10 SEC RX TRIP 3:10:15 15 SEC LLL 3:11:00 1 MIN HPI START 3:12:35 2 MIN 35 SEC LOSE T.W MAR 3 : 13 :05 3 MIN 5 SEC CFT 3:14:40 4 MIN 40 SEC 0 TAVE 453 F 3:15:00 5 MIN REGAIN STMAR 3:15:54 5 MIN 54 SEC HLL B OTSG 3 :16:04 6 MIN 4 SEC OPEN SV160 3:16:50 6 MIN 50 SEC 0 TAVE 380 F 3:18:00 8 MIN 0 TAVE 354 F 3:20:00 10 MIN l
NO OPERATOR ACTION FOR 10 MINUTES l
[
i SV160 is steam supply valve fron Auxiliary Boiler 1
1
ATTACHIENT 15 w
m e
m cn t
ua cw ce ce m
e cM e
ir,
u3 c3 cws es>
Mr*
--e C9 m
b e
o M
E m
c>
c :n e-.
=n e-.
m a
=
=
=
e
'E M
1... 1.1 1.I.1.......1..!..l....l..l.
4 4 1..!. 4 4 4..
i i i '
..... 1.
1.1......-..... l.. l.. i.... l.. !.. A. A.1.1.1. A. A..
G S
...l. 1... 1. 1.1.1.1. 1...
..l..
..!..l..l...... 1.1.1. 1.1.1..
P =-
N
..I...i..
....l........i.... i..... I. 1..
. d. %.'.. I..... T,. T,. T. i...
i i e
-g
.....i...,...e...i...
..,..i..
N.
.g.,i...,...,... g...:..,1..
, g w..t-
-*---t-f----'
- t t-t--**t*-
E r
h...l.....
.... 1.y.....y.....
a a
m m
g..
..l..
........ 4 4... 4 4..
... g
%....l....l....l... 4 4 4.g.4. 4 4.
, q p:e:w. 4..ts,y 4..,..%
g..........l... 4 4.7,... f....,. 7,...
+.t...,,..,......g..g.
N.
e
...;... w.... iw..g..
... 4 4.t-t --
T T --
e
. 7.
7.p. 7..y.,.....+....g m
.....+.+..,p..;.....+..,...
a,
...g.,...+
...g.. g..:...,.. w +.....,.. c%.. s.. A....
- - w. J.... i..,,. l..K.
- .. !...i...i. 4. +y..,.. + - +..
m m
... 4 4..
e o: o - - :_ *
- - !. M... A. A. A. A. A.1. A..
ns r.na
..... A. A. A. A. A. A. A.1%
......t..... i.. i...i. 1. i.. i....i...i...:..
...I. i.. i.. i.. i...!. 1...
4.a.
g g g g g y g g g g g g g g g g g g g g g g e.a.
m.
- - r. 7 7 7 7 7 7..i. 7.. I,- r... r,.. r,.. i - r,..,.
T i. 7.i. 7. T - -
e i
i i i i e i i
- 7. 7.. + t.,. 7.. _ a 3
...#...,..3...:..t.+..i..+.+..i..i..
t.-
e r.n e
...;. 4.+.+.+.+.+.+..;...;..a..,!..i..;..g..!..
... i. 4 4..i.. 2. 4. +. 4 4.. i.. l.. i.. i.. l..,.. ;.. !. i.i...
. 4 4... 4..
u : m x.
2.4.+.++.-
i.. i.. i. 4
- i
... :.. A.1. A. A. A. A. 2. A.. :.. :.. :.. :.. :.. :.. :... 1,:4 4 A. A. A. A..
e es
..b. 4, !... I,. 1..i...i. +i. i.. i....e...i....i...i...i...i...i..
',.a. i.. i.. +1..
...a a
.cs cs:n i
i e M rm3 i i
.,i...,i...!,...,i.
,i.. e.. i..,;.. i..,.. i.. i...e.
i e w
... g...g...;.. 9...J..... +:..;......i...
g l
".+.t..
.g 3
...a...s.. +. +..p.. +. +..a.. +..!..;;..g..,..g..g....g...;...g....s...
r.J+.+..
...i. 4 4..i. 4..!...i. 4..;...i..i..(..l..g..;..a..q.. +. +. +.
..;..s..
ga B
a w,a i.. l.. l.. i. 4 4..i...,. 4
.a...i...
- s
... I. 1. A.1. A.1. A. A. A.. !.. !.. !.. l.. !.. l.. !.. !.. A. A. A.
'.'.A..
m
=
... l. 1. 5 i 5 i 1.1.1..l..l..l..l..l..l..l..I i 5 i i
T
...,I...i...i,.. !
.j...!... I.. y..,!.. l..,!..,I.. i.. i...,i...i...i.
_. 9..,.y..i...
! l !
i 3
, y
, g u.
l..
.p:.
3
[
...g...p..p..p..p..p. p.
..p..;..g..,y..g..g..g..g..g...y.+ +......
7.
...,.........,..,..,..,..,..,....,...,.....p..
z
....i..!.
4 4 4.,
A..!......l..
......i..
..#..i...i. 4 4 4 4
...i...
ie m
. 1. i 5 i i i i i i.. i.. !.. i.. i. 1.. i.. i. 1.1.1.1.1.
e
.... 4..i..e...e.. a.. a...a. i....i...i.. l...i...e.. I.. I...i.. i...i. i.. i..i. i i
i i i
...i.e i e a i i i i a a e i i
e i i i i F. _ est gas:
......y
.y y. }..l G. ii
.... 7.. i.. t.y T..t - ; t.- -.
. - u - - : - -
- - : - ?.- 4. :..T - - - T.
c e
4
..;. y..y..y }.
..l..
..l. 4.1...
.A
...l. 1.1 4 1 4 4 4 4..
..,.....:.........A.1.4.4 4 4
... I. 4.1. A. A. A. A. A. A..
A.A.A.A.
...l.
- 1. 1.1.1..!.
- 1. 1. i i..l..i..i..i... 1.1.1..f. 1.1.
m I
i e i i i I i M
... g...g'.. i
......,. 7,. 4,. 7. 7..g...;...g... y.. g.. ;.. l..
.. l.. l.. j...{.. 9,.. 9,. 7,..g...g...
- i i i i i w
3
,...g...,...,..,..,..,....,..,..,...y.........
.,...y...
C8'
- - t - t. t. - '- - t. - t. - -j - -!- - - { - - l - - l, - l - - l. - - l. - - l_ - - l. - - -l
-t. - -i- - -l.- - -! - t. - -
est e
c.e:n I
i I
cw e
4:8 c.:n 83B S
c:
83b e
- s Ep".B C
S tr3 S
EMB a
CM Q
=4
-G u3
.a
 s
c "4"I*
Q f""*
- t u
N M
C1 4
AmMMMDmW y
A
[
S R
W
~
.a w
ATTACHENT 16 1
as w4FcF2
--. o---
- - - -f sees os.7s.tT o RS e
1 4
8
~
manaan. 99u.
$TATt0K 7
a.7J I
.- _ p.....
~
~ 4
- To SFR\\MKtP.RS
....._J
'p "M
PJW6 i4a
.y g.
, g, %7 gg
_\\
l swsTEM
-i
'l
' 6" Low
.A (pusswasu.s sc. Athu.r.n pspiuch
.?
CR\\FtCh.
fru\\.
wea.
(ALAAMS c3 TM$S w*c M v
y;
~%
s*
t.aw ur mus
- n. -
(Lhat nN h91f*Gr/ SPRW4% LEAN f'U.? {
t.['APts.-Ac.TsonNALve.
I
( W - W dei)
PuamwG
'WE^"
9w -57o) i
.g sgg, g Agg _
4 Ywwi ( AM PSH CW T.MM md esttu AcTuarew)
Tsc PRt.- AcTicR SPM\\NKG.R SYSTEM
'(simeuP\\E.D SCwf.MATtC)
(
ATTACHMENT 17 I
4__
,hl
.j l
i I
9 i
i a
-i
---i
- -4 i
j i
VICT0 HEN MDIATION I
l MOB ITH
' l' _TEPOINT UCR]ER i
i i
l l
j l
I s
... a.
.,. 1
.,_..e i.
.. ',. y.!..
- 1
\\.
!=-
a..,..
l I
I 1
l
- l. '.
i g..l..
s t.
3
~
g..
l s
i
.wa.
^.
t.
m
..w
.g
- 1..,
- m..
.).
......g>
.'.t.s i..
4
,.....a..
7..l w
~.
... ~..
.....c
... t.,
- ;. +.. :,. ;..
- r.....a..
..:.,.n
,..a.
- w...
- x. >
... +.,
....q.
..c.
..a..
. l.
~,
- . m eu
....,>e. o.
,a
,s.q
.. 3 g.
- i.,..
i.e
- y..
.i.
i i
i i
i.
i
- i. i i
r
.i i
l-I
- i. -
I i
l l
l l
i i
!I IJ lil I
ATTACHMENT 18 E.05 EXCESSIVE HEAT TRANSFER
(
~
OVERCCOLING IS OCCURRING FROM AN UNIDENTIFIED OTSG
.............................**.....................,,..........,t CAUTION
=
Shutting mai'n steam,co aux steam HV-20560 and MV-20565 causes you to lose the air ejectors and sealing steam to the turbine."
Condenser' vacuum will decrease to 20" very quickly, causing the T3V to fail shut.
Therefore, close main steam to aux steam HV-20560 and HV-20565 lase and reopen fitst (as scon as possible).
~
3.
Isolate both OTSGs 3.
Isolate both OTSGs.
Use Control Room Use Control Room controls.
- ****** * * ""~"""
Controls.
f 1.
IF AT ANY T!ME DURING EXCESSIVE g
HEAT TRANSFER.
I Pressurizer level decreases to g Revisiou
<10 inches indicat 4----.----ed. ----
~
ei'ther OTSG level increases to 95% on the operating range.
. ---=.--- - -,,
E OR l
I RCS Teold decreases to 525'F.
g a= = = = =========
THEN Trip MFW and AFW pumps which i
are still supplying ficW.to the CTSGs, OTHERWISE CCNTINUE.
DO NOT WAIT for one of these co'nditions to occur EMERGENCY OPERATING PROCEDURE REVISIONS f
Rev. 6 E.05-2
ATTACHPINT 18 E.05 EXCESSIVE HEAT TRANSFER
(
7.
.2 IE MFW is not available, IEEE establish AFW to only the good OTSG (non-steaa leaking).
If.,AFW control valves FV-20527, A OTSG, and FV-20528 B OTSG, were closed at R1SS, the valve to be used, needs to be placed in automatic at R1SS to allow manual / auto control at ElRC.
.3 Adjust TBVs on OTSG without the steam leak to maintain RCS temp ar its present value.
.1 SPDS Post Trip Display OTSG Tsat Line (purple vertical Revisiou lin*) should be approximately equal to Tc.
.2 T
for OTSG pressure, at sat ElRI approximately equals wide range Tc at ElRI.
.3 IE TBVs fail to operate, use ADVs until TBVs are operable.
wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwdes
~
d+
CAUTION l
Prior to throttling HPI flow below 105 gym per pump, HP I
'miniflow valves SFV-23645,'STV-23646, and Makeup Tank l
Outlet SFV-23508 zust be open.
...wwwwwwww.wwwwwwwwwww.wwwwww.wwwwww.www wwwww.wwwww.wwwwwwwwwwww.
8.
Control HPI an 8.'
Control EPI as necessary.
(Rule 2) necessary. (Rule 2) following makeup and purification valves have closed.
1.
SFV-22025, Letdown Cooler Inlet Common Isolation.
2.
SFV-22005, Letdown Cooler E-220A.
3.
SFV-22006, Letdown Coolers E-220B and E-220C.
I 4.
SFV-22023, RC Letdown Coo'lers.
Rev. 6 E.05-6
ATTACHtENT 18
[
T RULE 8. LOSS OF ICS POWER - CONTROL OF AFW VALVES. TBVs AND A0Vs 1.0 AFW CONTROL VALVES FV-20527/FV-20528
.1
'AUT0" mode of L&N controllers for FV-20527 and FV-20528, panel H2PS, will position valves to a predetemined position providing
-280 GPM to each OTSG on loss of ICS power.
.2
' MANUAL
- mode of L&N controllers provides full range of valve control independent of the Bailey controllers and the ICS.
~
Operator should ensure that trie L&N controllers for FV-20527 and
.3 FV-20528 are in ' MANUAL' before attempting to re-establish power the ICS.
Refer to C.40 for restoration of ICS Power.
2.0 TBVs AND A0Vs
.1
' NORMAL" mode for the TBV/ADV ' AUTOCLOSE' switenes will cause the TBVs and. A0Vs to close on loss of ICS power.
.2
'AUTOCLOSE DISA8LE' mode will cause the TBVs and ADVs to remain under the control of the ICS regardless of the status of ICS power.
(Valves will fail 50*. open on loss of ICS power).
.3
'tLOSE' mode will force the TBVs and ADVs to close or remain clost regardless of the status of ICS power.
l 4
Operators should ensure that the TBV/A0V ' AUTOCLOSE,* switches art l
in the 'CLOSE' made before attempting to re-establish power to :ne; ICS.
Refer to C.40 for restoration of ICS Power.
/
V R eu s sso a E
.}.
1 O
LOSS OF ICS POWER TRIP REPORT NO. 75 f 9 HUMAN FACTORS STUDY RESULTS WORKSPACE 1.
Excessive noise in the control room when both CR/T3C Essential HVAC units started, made it difficult to communicate.
2.
There were difficulties or time delays for the operators who were trying to gain access to various control areas.
This is due to the need for some areas to be locked for security or under HP control.
The area where the "A" AFW control valve is located has been changed to an uncontrolled area.
An opening in the fence has been created to allow easier sccess to both AFW va.1ve s.
3.
Emergency equipment such as high top PVC shoe covers and respirators were not located where the operators thought they would be most useful to deal with emergency situations.
Radioligal response lockers with Anti-C's, respirators, and SC8As has been placed in the Auxiliary Building basement and l
grade level.
4 Time was wasted waiting for the control room access door to cycle between exits and entrances and there were difficulties in coordinating with Security so that the door could be guarded but left open.
A change was recently approved to make one set of control room doors entry only and another set exit only.
COMMUNICATIONS 5.
The operator who was communicating with the NRC on the " red phone" found it difficult to keep track of plant status from within the clearance room area where the phone is located.
The phone was previously located in the control room where it was found to contribute to noise and confusion during the transient.
The recommendation is to install a long cord on-the receiver so that the communicator can easily observe control room events, yet remain out of the way and on the phone continuously.
6.
A plant trip causes a voltage transient which oftentimes causes the NRC phone to ring.
7.
There were a few problems establishing and maintaining communications between the control room and personnel in the plant.
Walkie-talkies will not work in the Tank Farm area due to the large amounts of metal in the area and the area's location relative to the repeater station.
(
Page 175
4 LOSS OF ICS POWER
^
TRIP REPORT NO. 75
(
ANNUNCIATOR AND ALARM SYSTEMS 8.
There is no main annunciator indicating the status of the Main Steamline Failure Logic.
The individual pressure switches in this. system are alarmed on low steam pressure on IDADS but the alarm statement does not inform the operator that MSFL has actuated.
An actuation of a system that stops main feedwater to the steam generators should be annunciated on.the main annunciator panel.
9.
The IDADS screen display alarms were largely ignored because of the number of alarms seen by the computer during the transient (275 alarms and resets came in on IDADS during the first 30 minutes of the transient).
10.
There is a common. annunciator window in the control room for diesel generator trouble. This annunc.iator is actuated by numerous parameters but there is no reflash capability.
Currently, an operator is sent a few times an hour to check on the status of a running diesel generator.
' CONTROLS 11.
It was hard to tell if the AFW control valves were fully closed when locally operating them.
The label on these valves may be misleading because the pointer must travel past the label edge.
Many of the plant pneumatic valves have the same type of labeling.
These valves should be relabeled, giving the most important or most frequently manually operated valves priority.
12.
The ability to control the T8Vs, ADVs, and AFW control valves from the control room using redundant controllers powered front a separate power source was recommended. Modifications to the controls of these components have been approved and will be in place prior to startup.
VISUAL DISPLAYS 13.
It was difficult for the operators to track the status of the pressurizer relief tank (pRT) over time during the transient.
A two-pen trend recorder will be installed to track PRT level and pressure. This is a long term item.
14.
A difference between SPDS steam generator operate level and control room operate level recorders was identified by the STA and several operators. The calibration between SPDS, IDADS, and level recorders were checked.
No inconsistencies could be found.
15.
A few control room recorders were found untrustworthy during the transient, yet the operators tend to rely upon data pres 9nted by the recorder.
The continuous and impact trend recorders should be replaced with more reliable, more readable recorders.
A study is currently underway to investigate other recorders as potential replacements.
t Page 176
LOSS OF ICS POWER TRIP REPORT NO. 75
(
16.
The operators were given conflicting data on main feedwater flow when the recorders failed to mid scale and the edgewise meters showed no flow.
This occurred because the power supply for the recorders was ICS power and the supply for the meters was NNI power.
During the CRDR the recommendation to label the power sources of all control room instrumentation was made.
LABELS AND LOCATION AIDS 17.
When ICS switches S1 and S2 tripped, several operators and a computer technician checked the switches and believed them to be closed. The switches are located high in an ICS cabinet. The only position indication is the word "ON" painted on the bottom of the switch.
When the switch trips, the switch moves so the word disappears.
There is no indication that the switch is in a tripped position. The S1 and S2 switenes should be relabeled..
PANEL LAYOUT AND CONTROL / DISPLAY INTEGRATION 18.
The AFW indications and controls are spread throughout the control room.
As part of EFIC, scheduled to be functional by the start of cycle 8, a single operating area will be established which will contain all normally used controls and displays associated with the auxiliary feedwater system.
19.
The makeup system isolation valve SFV-23604 was put into MANUAL on one SFAS
. panel but not on the other.
Switches on both panels must be placed in MANUAL before the valve can be opened following an SFAS actuation.
This resulted in j
a ten minute delay in opening the isolation valve.
The BWST outlet valves, SFV-25003 and SFV-25004, also have multiple AUTO / MANUAL stations because the valves supply water to pumps in three different channels of SFAS equipment.
Following SFAS initiation, permissive signals from the MANUAL mode of all actuated channels must be given before the valve can be closed.
A recommendation has been made to change the MANUAL pushbutton lenses for these valves to key the operator's memory.
MISCELLANEOUS 20.
The Emergency Operating Procedures (EOPs) provide no guidance as to how long l
it would take for something to occur or whether one step should or could be occurring simultaneously.
There was no bres(po,nt to indicate when to move onto the next step (no timeline).
There @s a' t sufficient guidance for securing AFW.
A revision to the Excessies de--
Transfer EOP has been made to resolve this issue.
21.
There should be an investigation into the installation of low suction trips on vital pumps, such as the Makeup /HFI pumps.
i k
page 177
LOSS OF ICS POWER TRIP REPORT NO. 75 22.
The operating crews abilities were stretched to their limits trying to manually control the plant during the transient.
Analysis should be performed to determine the minimum number of personnel to handle different transients. There is a need for 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> support from maintenance, including engineering support.
Also, a procedure should be developed to call in additional operational help without tying up an operator on the phone.
23.
Sorse operators were unsu e of component response to the repowering of ICS.
Tests were conducted which revealed differing component response or responses which could not be easily explained.
B&W has been tasked with the program to predict and explain component response from the repowering of ICS.
24.
The operators believed they heard feedwater relief valves lifting.
There is I
no indication immediately available to the control room operator to tell him which relief, if any, is lifting.
25.
Some radiation monitoring equipment is in an area adjacent to the control room while other equipment is located within the control room.
It was difficult for operators to get a clear picture of the overall release as information on Auxiliary Building is available in both places.
Previous CRDR recommendations have suggested this information be centrally located.
l e
Page 177
asexLnrtni eu LOSS OF ICS POWER TRIP REPORT NO. 75
\\
EFIC IMPLEMENTATION 8y letter dated August 15, 1980, the NRC identified a situation where failure of power supplies to NNI or ICS could result in ADVs opening to 50% open position.
The District concurred with this scenario in its October 6, 1980 submittal.
The District proposed to correct this ADV response as part of its EFIC (Emergency Feedwater Initiation ar.d Control) AFW system upgrade.
The design concept was presented to the NRC at a September 4, 1980 meeting.
Equipment delivery for EFIC was originally estimated to be in early 1982. When actually signed, the contract specified equipment delivery for April 1983.
During the initial design review process, additional improvements to EFIC were identified.
As a result of these design changes, the deliver schedule was adjusted to May 1984.
NUREG 0737 required AFW automatic initiation and flow indication (II.E.1.2.1 and II.E.1.2.2).
The NRC issued safety Evaluation Reports in January and September 1982.
In October 1982, the District indicated that it would install interim safety grade AFW modifications and that EFIC was separate ~and beyond the AFW upgrade requirements of NUREG 0737.
The District also submitted a new schedule for EFIC implementation showing completion by Cycle 7.
This schedule was confirmed by the District in December 1982.
1 The District informed the NRC in April 1983 that the installation was tied to Control Room Design Review (CROR) and RG 1.97 modifications.
This was based on the need for an EFIC control panel in the Control Room that was compatible with the CROR effort.
Part of EFIC are the associated RG 1.97 instrumentation commitments for Rancho Seco; as a result it was necessary that EFIC be rescheduled for Cycle 8 (i.e., the next scheduled refueling).
In late 1983, the District implemented an Integrated Living Schedule to better control resources, scheduling of modifications, and enhanced operations at Rancho l
Seco. Since the AFW requirements of NUREG 0737 were previously completed EFIC was I
considered a plant betterment.
Using the Living Schedule, to prioritize the use of District resources, the District scheduled EFIC to be installed in two phases--Cycle 8 and Cycle 9.
The Living Schedule process determined that other NUREG 0737 modifications, 10CFR50.49 - Environmental Qualification of Electrical Equipment, Appendix R - Fire Protection, Generic Letter 83-28 ATWS, and NUREG 0737 Supplement 1, items receive high priority which resulted in heavy commitment of District resources during the Cycle 7 outage.
/
k Page 179
o ATTACHMENT 20 LOSS OF ICS POWER TRIP REPORT NO. 75
./
It became clear in meeting the requirements of NUREG 0737, that the number of modifications imposed in Rancho Seco would exceed the' electrical capacity of its existing emergency dierel generators.
The District decided in 1980-81 to purchase two additional diesel generators to augment the existing system. The District originally planned the installation of these new generators during the Cycle 7 refueling outage.
This schedule was compatible.with the installation of the majority of the TMI modifications, as well as the implementation of EFIC.
The diesels purchased were made by TOI and the District, as well as several other utilities, were forced into a major TOI generator requalification program as a result of design problems discovered on the Shoreham plant diesels.
This requalification program required both time-(several years) and resources to complete.
The' current schedule will have the diesels operational during the Cycle 8 refueling outage.
Since EFIC, and several other modifications, were tied to the installation of the diesels, the District.was forced to defer implementation of EFIC.
This delay also afforded the District time to take a closer look at EFIC as installed at CR-3 and ANO-1.
Because of some initial startup and operational difficulties at these installations, the District decided on installing the indication portions of EFIC during Cycle 8..This would allow the operators to gain familiarity with the system during an operating cycle.
Likewise, the District has been interfacing with the staff of ANO-1 to minimize any operational problems and benefit from the ANO experience, as the District's EFIC will closely resemble the ANO EFIC.
/
In October 1985, the District committed to accelerate implementation of EFIC.
The District outlined the specifics of the EFIC implementation in a letter to the NRC dated January 17, 1986.
This implementation will result in the majority of the EFIC actuation and control functions being operational at the completion of the Cycle 8 refueling outage.
t N
Page 180 1
1 -.
- --