ML19305A662
| ML19305A662 | |
| Person / Time | |
|---|---|
| Site: | Crane, Davis Besse  |
| Issue date: | 12/16/1977 |
| From: | Domeck C TOLEDO EDISON CO. |
| To: | Novak E TOLEDO EDISON CO. |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001160722 | |
| Download: ML19305A662 (6) | |
Text
"M
/M.14/'/9U.5Y4-}
0k$YQ
)'
f'l~f INTRA' COMPANY MEMORANDUM s
dan E 06214 -i To
~,1 777 l
Dece t
l Fild W % 2 h enord 32-?7-20 Eugene C. Novak
' ME 1
Tite.T Charles R. Domeck
}
suancT Davis-Besse Nuclear Power Station Unit No. 1 Summary Of A November 29, 1977 Incident TS SF f.
u Nsu
SUMMARY
OF EVENT On~ November 29,E1977, an incident occurred at Davis-Bt en h lae r Pnt er Station Unit No. I which caused a reactor trip and a t E ' - d
'11 AC power. As a result, the emergency diesel generators were both auto-matically started and one of the emergency dAesel generators at the unit tripped on overspeed.
The following is a brief suamary of what happened. 5A' wiring' error in the patch panel for_the Reactimeter caused the unit load demand signal from the ICS to be shorted to a control rod position indication.
This resulted in the indication showing that the rod had dropped and at the same time it caused the unit load demand in the Integrated Control System (ICS) to go from 40% power to slightly greater than 62b% power.
When this patch panel was plugged into the Reactimeter, these error signals were.sent to the control. rod drive system and to the ICS, During the time the reactor is tested at 40%
power, the high flux trip set points on the Reactor Protection System (RPS) will be set at 50% power.
panel was plugged into the Reactimeter, Thirty-five seconds after the patch the reactor tripped on high flux.
This caused the main turbine to be tripped at the same time.
When the operator observed that the main turbine generator and the reactor had been tripped, he manually tripped the two 345 KV Air Circuit Breakers (ACB) on the high side of the main transformer.
He tripped these breakers nine seconds after the reactor had tripped.
This re-sulted in the unit's housepower being tied to the main turbine generator with no steam to the turbine, and it was no longer synchronized to the transmission grid.
The generator has been designed to trip 30 seconds after the turbine has been tripped.
This generator trip causes the following events to happen:
the 345 KV ACB's are opened, the 13.8 KV breakers from transformer 11 to buses A and B are opened, the generator field is tripped, and the breakers from transformer 01 to bus A and from transformer 02 to bus B are automatically closed if the 345 KV generator ACB's have been closed.
Th'is causes a fast transfer of buses A and B to transformers 01 and 02.
5' "
L. E. Roe, w/a 2 -P;- D. Murray, w/a 8 001160 7M 1
A l
s q..
On November 29, the fast transfer did not occur because the generator had coasted down to a reduced frequency, and when the 30 second timer had timed out the housepower buses A and B were out of phase with the As a result, when the 30 secondary side of the 01 and 02 transformers.
second timer timed out, the generator was tripped, no fast transfer occurred and we had a total loss of housepower at the unit.
Upon the
. loss of housepower we lost all four reactor coolant pumps which sent a
! signal to the Steam and Feedwater Rupture Control System (SFRCS) to
' start both auxiliary feedpumps.
At the same time, the emergency diesel generators starting logic was initiated due to the loss of voltage on the two 4160 volt essential buses.
The two emergency diesel generators The received start signals 3 seconds af ter the loss of all housepower.
1 emergency diesel generator tripped on overspeed three seconds No.
after it was started. Housepower was restored to buse,s D1 and F1 from the emergency diesel generator No. 2 seven seconds after all housepower was lost at the station.
This allowed auxiliary feedpump No. 2 valves to start opening. At the time of loss of all housepower, the two main feedpumps tripped due to the loss of the AC motor driven' lube oil pumps.
This resulted in a main feedwater AP trip of the SFRCS 15 seconds after the loss of all housepower at the unit.
Housepower was restored from 01 start-up transformer to C1 and El buses 1 minute and 19 seconds after the loss of voltage'to these buses. This allowed auxiliary feedpump The balance of the housepower was restored to the unit No. I to start.
in 5 minutes and 39 seconds.
EVENT DESCRIPTION The following is a detailed description of everything of significance that happened during this event.
22:42:50, a control rod drive asymmetric rod alarm was received by the This alarm was created when the patch panel with the wiring computer.
error was plugged into the Reactimeter.
22:43:25, the reactor tripped on high flux.
22:43:26, the main turbine tripped.
22:43:34, the 345 KV air circuit breakers on the high side of the main transformer were manually tripped by the operator.
22:43:57, the reactor coolant pressurizer low level alarm came on.
" 22:43:54, the station housepower went dead at the time the 30 second j timer timed out to cause the generator trip and the fast transfer of
- buses A and B to transformers 01 and 02.
22:43:54, the SFRCS tripped on the loss of four reactor coolant pumps.
22:43:57, emergency diesel generstors ' and 2 were started.
22:44:00, emergency diesel generator 1 tripped or. overspeed.
22:44:01, buses D1 and F2 were auto transferred to emergency diesel generator 2.
22:44:02, the auxiliary feedpump 2 main steam 2 inlet isolation valve started to open.
i 22:44:05, bus B voltage was restored manually to normal by the operator closing in the 13.8 KV breaker to 02 transformer.
22:44:09, a' full SFRCS trip occurred on high main feedwater AP on steam generator No. 2.
22:44:14, main steam isolation valve 2 closed.
22:44:15, main steam isolation valve I closed.
22:44:19, the 13.8 KV breaker to 01 transformer.voltag was restored to bus A by the operato 22:44:47, bus D2 voltage was restored to normal by the operator closing in the feed from bus B.
22:44:57, bus C2 voltage was restored to normal by the operator closing into the feed from bus A.
22:45:13, bus C1 and El voltages were restored to normal by the operator closing the 4 KV breaker from bus C2.
22:45:13, auxiliary feedpump No. I started through the restoration of voltage on El bus which allowed the motor operated valves for auxiliary feedpump No. I turbine to open.
22:46:37, the reactor coolant pressurizer low level heater trip occurred due to the shrinkage of the reactor coolant fF drop in the level in the pressurizer.
and the subsequent 22:46:52, reac~cr coolant makeup pump No. 2 was manually started by the operator.
22:47:16, essential bus D1 was transferred from the emergency diesel generator 2 back to bus D2.
22:47:51, bus F3 voltage was restored to normal.
22:47:51, bus E2 voltage was restored to normal.
~ J 22:48:23, bus F2 voltage was restored to normal.
I 22:48:55, bus E6 voltage was restored to normal.
22:48:56, bus E4 voltage was restored to normal.
g..
22:49:04, reactor coolant makeup tank level vent low.
-22:49:05,' bus F6 voltage was restored to normal.
22:51:51, reactor coolant makeup pump 1 was started manually by the operator.
reactor. coolant makeup tank level was restored to norm' l by a
22:53:07, the operator adding fluid from boric acid addition system.
2f tS8:20, reactor coolant pump 1-2 was started by the operator.
reactor coolant pressurizer low level heater trip was removed 23:00:06, by restoring the level in the pressurizer.
23:00:46, reactor coolant pump 2-2 was started by the operator.
From this time on, the system conditions gradually were restored to normal.
SYSTEM-EQUIPMENT MALFUNCTIONS I was a major The overspeed trip of the emergency diesel generator No.
equipment malfunction. Station staff has found that the governor stops were not adjusted properly, and this has been corrected by properly l
adjusting these stops.
Miscellaneous alarms on the computer and the annunciator were not operat-All of these alarms that have been noted ing properly during the event.
during the review of the incident have been turned over to the Davis-Besse staff for correction.
At 22:48:24, a discontinuity occurred in much of the data that had been Discontinuities were observed on the follow-recorded on the Reactimeter.
A) Steam generator No. I wide range ing points on the Reactimeter:
1 level; B) Steam generator No. 2 outlet pressure; C) Steam generator No.
outlet pressure; D) Megawatt electric generation; E) Main feedwater temperature; F) Group VI rod positions; G) NI power range to ICS; F) Low A check of pressure condenser pressure; G) Reactor coolant pressure.
the station log indicated that F2 voltage was restored to normal at the
. When bus F2 was restored, this same time this discontinuity occurred.
At re-energized the static voltage regulator that supplies bus YBR.
this (see instant that voltage was restored to bus YBR, bus YBU was automatically transf erred to YBR through the static switch on the non-We are still trying to find out if all of the essential inverter YVB.
. troubled instriments or the Reactimeter was powered from bus YBU and
-this valtage transient on YBU caused this discontinuity in the data
%ecorded by the Reactimeter.
I The reactor coolant pressure signal that was recorded on the.Reactimeter was a wide-range signal that came out of a buf fered output frem the
~
'SFAS.
This signal indicated svings of about 100 psis throughout the A similar phenomenon occurred during our September 24 transient.
and nobody has been able to expizin why these pressure 1977 transient, Nobody swings are occurring on the input signals for the React 1 meters.
to the SFAS:
has observed these pressure signale to occur on the inpu:
therefort, we can only conclude that it is in the heacrimeter or in the output signalt troc the SFAS.
m
~
?
- - ~ - -...
a I
i,a,
.e,
b Batteries IN and IP did not have battery chargers on them until voltage l
was restored to bus El.
However, low voltage alarms occurred on these
']
batteries 26 seconds after the battery chargers went dead, and the voltage was not restored to normal until after bus El voltage had been restored to normal, which picked up the battery chargers.
I do not feel that low voltage alarms should come in this quickly on this battery.
!r Bechtel(and the statiodl staff)has been asked to investigate why these low voltage alarms came in so early.
Either something is wrong with the battery or the low voltage alarms are set at too high a voltage.
li PROCEDURAL DEFICIENCIES i
Tre first procedural deflu ey noted as a result of this incident was the manual opening of the 345 KV generator ACB's.
The station procedures have been revised to tell the operators not to touch the 345 KN breakers, so that the automatic transfer occurs after the 30 second timer is timed l-out.
li Reactimeter procedu
..s need.to be updated. First of all, all patch panels should be thoroughly checked before they are used.
The secend thing that should be done is a careful record kept of the exact source of all signals going into the Reactimeter. This record ghould include the positions of transfer switches where the signal can come from two i!
transmitters through a hand-operated transfer switch.
At the present time, we cannot determine the exact transmitter that was used to supply signals to all the inputs to the Reactimeter during the November 29th event.
SYSTEM TRANSIENTS l
The major system transient, as a result of this es.ut, was the rapid cool-down of the reactor coolant system when auxiliary feedwater was introduced into the steam generators.
Since we had no reactor coolant pump flow as a result of loss of all AC power, the reactor coolant pumps were all off. This resulted in natural circulation flow in the reactor coolant system. During the initial portion of the event on the secondary j
side, the secondary pressure was relieved to the condensor through the turbine bypass valves. However, after the full trip of the SFRCS,the main steam line isolation valves were closed and the turbine bypass valves were no longer available.
.owever, the Reactimeter. data indicates (thatYaIno31Eiduring$ughitofliftithc[chdel;s'afetyvalves.the eventAd'th6 '
~
~
generat'5rsYge' [hfish.Len In fact, ~ he' 5 t
t pressure steadily decreased on the secondary side of the steam generators and this caused the saturation temperature on the secondary side to be reduced which allowed the primary system temperature _to be_ reduced.:, As the
[ pres, primary, system reactor coolant shrunk, weilost the levelito theiThis eventually suriser; l
heaters which made it more difficult to maintain the primary system pressure. The attached curves show the pressure and temperature transient on the primary and secondary sides of the reactor coolant systems. These curves also show steam generator levels and pressurizer j
levels.
L.
wi
~ 6-t l, ,
k
., xt We haveksskedIB&W: to investigate why we had this rapid decrease in tempera *,ure on the reactor coolant system.
Specifically, we have asked the following:
1.
Is there enough decay heat at this time, considering the small burn-up of the core and the initial power level of 40% to maintain the reactor coolant system tercerature and pressure under the i
condition of the incident.
FArewemaintaininNoohighislevel'oflrdativ'elycoidfwaterinthe 2.
2
~.
'ssi$nd$ry"si'de ~of the steam generatoss.~
We introduced approximately 90" of 70 F water to each steam generator through the automatic starting of the auxiliary feedpumps.
UNRESOLVED DESIGN PROBLEMS 1
The first thing that we noticed during the incident was: why not fast j
transfer the 13.8 KV buses to the start-up transformers immediately after a reactor trip? This fast transfer would only occur if the start-up transformers were energized and available to carry loads.
Bechtel is investigating this at the present time.
The second item noticed was: why do we trip.all the 13.8 KV breakers serving the 480 volt unit substations and tha 4160 volt buses C2 and D2 upon the loss of voltage on buses A and B? Ar *he present time, when the voltage is lost on buses A and B, all 13.8 KV breakers on the buses are automatically tripped. As a result, when voltage was restored to these buses, the operator did not restore voltage to any other loads in the station.
Bechtel is reviewing this to determine why we have tripped all these 13.8 KV breakers on buses A and B.
We are reviewing the possibility of not tripping the makeup pumps on the loss of power on the buses C1 and DI, and not closing the reactor coolant makeup and the reactor coolant pump seal injection lines, with SA signals.
We find that during any transient on the secondary ImakAup" flow thaF we can~to miintainl bvellinit'hsireside,iwe need,al1Jthe;
~
~
actor coolant pressur-
}ir.erijThi;s)phin$ mets (occurredlonourl September 24theventandfagain lpccurred on;the; November 29th; event. We feel that if it is possible, we should preserve the makeup flow to the reactor coolant system.
If such a change were made we would probably have to trip the makeup pump on SFAS incident level 3 when the decay heat pumps are started in order to prevent the overloading of the emergency diesel generators.
db c/1-6 e
e 3
T