ML19308B740
| ML19308B740 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco, Crane |
| Issue date: | 03/31/1978 |
| From: | Mattimoe J SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001160870 | |
| Download: ML19308B740 (5) | |
Text
D h
fBDJW3 SAtRAMENTO MUNICIPAf. UTILiiY DISTRICT D E201 S Street. Scx 1TO, Sacrarnent:i. California P'313; (SIE) O M 211 March 31, 1978 Director of Rct,ulatory Operations ATTN:
Mr. R. H. Engelken
.M OM NRC Operations office, Region V F1Af2C)4 oDi 1990 N. California Boulevard
'n'aln u t C:cek Plar.a, Suite'202 Winut Creck, California 94596 l
Re: Operating License DPR-54 Docket No. 50-312 l
Reportable Occurrencu 78-1
Dear !!r. Engelken:
In accordance v.ith Technical Specifications foy3uNhDek[ Nuclear Generating Station, Section 6.9.4.1.b, the Sacramento !!unicipal Utility District is hereby subt:titting a fourteen day follovup report to Reportable Occurrence 78-1, which was initially reported to your office on(March 20,g978),-
Cn the date of that initial report, a ground short in the plant non--
nuclear instrumentatibn rt sulted in a reactor trip and subsequent RCS cobldown which exceeded limits set forth in Technical Specifications Figurc 3.1.2-2.
Tnis followup report will describe the sequence of events surrounding the incident, detail how the instrumentation loss precipitated the transient, discuss the analysis perforced on the transient and its effects on the RCS, and set forth the corrective action that has and vill be taken to prevent.
recurrence and to' insure the integrity of plant systees.
<g' L'/.r Scouence of Events w,
., f.q a.,4
,v Prior to the cooldown transient, the plant was operating at a steady sta,te power level of 70 percent, with all four reactor coolant pumps operating and-an average reactor coolant system temperature of._582*F._ Shortly before 0425, a control roca operator begad'riplacing;a burned-out Llight, bulblin a back-
~
a li$ted pushbutton switch on "ond lif the control consoles.
The 'DC power for
.this evitch is provided from the "Y" portion of the Non-Nuclear Ins trumentation' (N L '!).*
To change out the light bblb, the light assembly was pulled out from the panel and flipped down, exposing the bulbs.
During the change, a bulb was dropped into 'he open light assembly cavity, creating a short to ground.
Tne et.rrent-licitiis and undervoltage protec' tion for the NSI-Y DC pov5r supplies
Preliminary investigations have shown that approxit.ately two-thirds of the NNI signals (pressure, temperature, 3evel, flow, etc.) were af fected 8001160 [7O d.~
o.s o m n u m s -,( - - m. m,, m - u w
.~
\\
.'t
,a Director of Regulatory Operations March 31, 1978 by' the pcwer loss.
We erroneous signals provided both the Control Room and the Integrated control System (ICS) with faulty infort:ation.
The ICS, atter.:pting to tutch equiptent output with plant requirceqnts, reduced main feedwater flow to tero in response to faulty signals. Thei reduction 'in'fe'e"dwater fflow; htiisd RCS ~ pressure ~to11EEreasEl with a[rc'adtor, trip;occurrids 6n hi'gh" pressure
~~~~
at'0425~. "
^
In the period following the reactor, trip, the operators were stim [hlepered$
by'the lack"of 'instruffentaion"avdilable ind by~eq"uipit.ent;redonding to in.icc~u-
{i rQto.signalsj. ' ' Tor" api roxiihately nine' minutes "following ' the" trip, pic'ss'ure '
i slowly decayed in the RCS, remaining at approximately 2000 psig.
It has been postulated that pressure reinained fairly constant during this period due to the cooling provided by :akeup flow into, the RCS and to the lif ting' of a'Pressuiizer'
[ Code safetyjalM Nih71ts"setpoint, of 2500 psig.
An a{tiiliaryf feedGater2 "
~
~
' jump had starredTda lhs " loss 'of feedkater. flow, howeverl the auxiliary feedwater Ivalves e:$ained 'closedLin response to erroneouslonce Tarough Steam Cencrator (OTSG Lstartup level.sigua3s.
nose two sisnals were rendered inoperable by the Pil-Y DC power failure, the "A" steam generator level signal drifting to zero indication over a nine-ninute period while the "S" stca::2 generator IcVel drifted full scale.
The actual plant conditions showed thatiboth;OTSGfs 6eni 7
fdql during this period.
When the startup icvel for.thc "A" OTSG drifted belev
~
, the Ice level setpoint, the 91CS. opened the auxiliani feedwater valve {admittini -
Lvater' to the shell side of the "A" OTSG.
This inflow of water created a heat sink for the RCS resulting in a.. rapid pressure drop.
n e operatorslaiso man
,i have (increased the main feed ptntp flovfdt.this ' time, providing another source of feed flov t'o the "A" steam generator.
The rapid drop in pressure took the RCS through the safety Features Actuation System (SFAS) setpoint (1600 'psig).
Ca SFAS signal, both auxiliary feedwater bypass valves opened which began the filling of both steam generators with water.
UntiiL power waslestored"SojNIyI~ap~prhx'ihahely odihotirlaiidLtcQtinudes IAf tar ~ thFrcActor tripl, 'thEoperato'rs continudd'Ihc inJ& tion"6f's&kiliaff~"
'feeddaier 'th'at'vas started on SFAS signal.
It did not appear that any RCS
~
tc:perature indication was reliabic, so the operators raintained RCS pressure as well as possibic utilizing the pressurizer level indication and the RCS pressure indication that was available.
Control was obtained by adjusting high pressure injection flov.
Tbc pressurizer heaters ucre not available due to the SNI pcVer loss, preventing their use for pressure control..The cont. ;ous injection of auxiliary feedwater resulted in cocplete filling of.
l bo t).
tehn generators, af ter which water began to enter the steam lines. This l
S arge heat sink continued to cool off the RCS, although the operators were not avarc of system tecporatt re.
Wen the power to the NNI-Y DC power supplies was finally restored, the eperators realized the RCs tenperature had dropped to around 285'F, which placed the plant in the restricted region of Technical Specifications Figure 3.1.2-2.
I=nediate action was taken to return to the pernissible.
l 1
i e
Director of Regulatory operations March 31, 1978 operating region, including spraying the pressuriser to reduce pressure, kcuping three RCP's operating (pu p combinations were changed) to increase temperature, shutting off auxiliary feedwater flow, and draining the OTSC's.
Non-Nuclear Ins trucentation Power Loss _
The short caused by. the light bulb drev excessive current through the 24-volt DC power supplies which ervice cocponents in NNI cabinets 5, 6, and 7.
T"no pouar for choco cabino,co Ic Neignated UNI-Y, vtsh the pouce foe cabincts 1, 2, 3 and 4 being designated :,NI-X.
The four power supplies for NNI-Y are operated current-limited with a setpoint of 7.5 amps.
The subsequent reduction in voltage caused an undervoltage monitor to operate, opening the two shunt breakers through.which AC power from inverter D and inverter J is supplied to the DC ' ower supplies.
Loss of these power supplies = cant that every co=ponent p
in ci,inets 5, 6, or 7 operating on DC power was not functioning properly.
An NNI signal could have been affected two ways between its source and the receiving component.
The sip,nal could be interrupted completely due to a centact opening on being deenergized.
Because mos t of the signais are
~10 volts to +10 volts, this vould have resulted in a mid-scale reading or in some cases a reading anywhere between -10 volts and +10 volts being transmitted to the indicator or sent to the ICS as an actual plant para =eter.
If a signal conditioning component (buffer amplifier, square root extractor) was a f fected, this vould have meant that the desired conditiening would not hhve been performed on the signal or that the component tight not pass the true, signal, resu1ging in erroneous values being cent to the indicator or to the ICS, Since signal paths in the NNI are not restricted to either the X or the Y cabinets, about two-thirds of the signals passed through at least one I
component in cabinet 5, 6 or 7 and..wcre thus rendered invalid.
Tausw spurious, readings,had -severalr ef fects.~It was Idifficult' for ^ thal
~ _ -,
r tpcrators. toLascertainiwhichiuff theiriindicators;were v lid, given W1hii[ging plant conditions and the vide variety of possible errors that were intro'duced, Only a select few parame ters were known to be valid readings, and the operarcrs had to, control the p} ant, based on that information.
T"ne second effect was the p
Lspurious signals; vere _ fed,intolthe ICSi so equipment was operated automatically, without regar'd~t3'3ct'ual' condi~ti6fid.' "The first evidence of this was the runback of the enin feed pumps to zero, which caused the reactor tri~p.
- Later, the automatic actions -involved with adding feedwater to the dry steam generatora hindered operator actions nnd precipitated the rapid depressurization leading to S* - initiation.
Fower was finally restored to the NNI-Y vhen operators realized that the shuat breakers were open.
Restoring power returned the non-nucicar instru-nentation to operation, permitting proper operator response to the plant condition, which at this time was outside the pernissible operating region.
Director of Regulatory Operations March 31, 1978 Transient Analysis
'Ihcre was very little permanent record of the plant,pargoeters during 7
- e. transient. A.r.ajor_ source,of informatica was the!Pont 'Irip Transienth th g RevieV.;which s
tQr_eactor; trip [ prints ou't ! selec ted l computer points l periodically; following ;aIE data ovailable (recorder outputs, hourly logger typer, etc.) during the transient.
Over a period of several days following the incident, engineers were able to trace which signals were valid, determine what equipeent operated at which tires, cad then interpolate to arrive at a te=perature trace for the RCS.
Irr>idiately following the trip, temperaturcs slowly increased while pressure decayed away, twintaining approximately 2000 psig. At approximately nine minutes after the trip, fcedvarer began to enter the "A" sceau generator, resulting in an RCS depressurization leading to SFAS initiation.
Wich the full auxiliary feedvater flow initiated by SFAS, RCS temperature fell from the high of about 595*F to 185*F in the span of slightly r. ore than one hour.
This cooldown rate of approxir2tely 300*F per hour is well above the permitted rate of 100'F per ho.ur stated on Tigure 3.1.2-2.
When Nh'I power was restored with the RCS at 285'F, aeasures were taken to bring the plant back within limits by increasing tenperature and reducing pressure.
To assure that all components of the plant were not damaged by the transien't, available inforcat '.on on the RCS pressure and te=perature, OTSG presourc, temperature and lev,el,' feedvater flev (eain and auxiliary), pressurizer level and other relevant parameters were transmitted to the reactor vendor, D.bNck and Wilcox, for c'nalysis.
On March 23, the District received a respons'c from B&W indicating that their analysis was cocpicte. Af ter evaluating the ef fects
'of the tz'ansient on the reactor vessel, the reactor coolant piping, the pressurizer, the OTSG's, t.he fuel asse blics, the RCP's and seals, and the control rod drive t>achanisms, it was recommended that Rancho seco be pertitted to return to power, under certain specified conditions.
.~,n----,
-mn~.~
- mw g,
_..~.Theianalysis wa,s subsitted; tolthe=0ffice ofJNuclear Reactor Regulation,1 knd on March;24.; they_agreedithat> Rancho,Seco could; return; to operation; providing the conditions v.cntioned in the 35W 1etter were fulfilled. When the District receives the complete analysis from B&W, a copy will be forwarded to the Kac Regional of fice.
Corr + 'ive Action
~
Tne ir:ediate corrective action taken following the rapid cooldevn was to return the plant to the permissibic operating region of the Technical Specif f ention figure, rne unit was then kept shutdown while data was ' gathered and sent to B&h' for analysis, and further investigations into the incident Vere cade at the plant.
Tne Management safety arview Correittee held a recting on the subject on March 21, and is' sued several directives.
A cor=itte'e of
- hree engineers from the cicetrical, techanical and nuclear d1sciplines v'as e
1
l
?
I March 31, 1978 Director of Regulatory Operations.
appointed to determine the circumstances Icading to the shutdown, to detercine if the energency shutdown systers are adequate or whether som3 redesign is indicated, to evaluate the significant parameters involved in the cooldown to determine if any component damage had occurred, and to recoczend corrective actions.
The findings of this committee will be relayed to the NRC Regionni of fice vbca.they_are..available The PRC was directed to review CEd~ciYcIstances s
connected with th'e incident, and to review the 36W recommendations which were forthcoming. The PRC was to reco=nend a return to power if they agreed that the recocrendations were acceptable.
If any disagreecents developed, no return to power was permitted until the 11SRC reviewed the catter and determined appropriate action.
Upon receipt of the 36W recommendations, the PRC reviewed them and requested that a special test procedure and a casualty procedure be written to assure co=pliance.
PRC concerns about the following itets also had to be satisfied prior to startup:
1.
A question was raised concerning possible datage to steam.
lines from the injection of water, so the lines vere checked for any deforeations.
2.
A 2255.psig leak test was performed on the acs to insure integrity.
3.
The overvoltage trip setpoints on the NNI DC power supplies were increased from 27 volts to 29 volts to pravnnt spuri.ous trips.
Thn special test procedure addressed the conditions imposed by B&W, such as reactor caneuvering limits for the first startup, increased surveillance of the loose parts monitors for a ucek, an operability check of on-line and redundant NNI instrumentation, and daily surveillance of the pricary and secondary radiochemistry for a veck to check for leaking components.
The casualty procedure was written to provide required operator actions for' restoration of NNI power following a trip similar to that experienced.
The PRC also committed to having a procedure written by April 7 giving operator instructions if FMI power cannot be restored.
Current' S tatus_
r llowing NRC's review of the B&W analysis of the transient, they agreed o
'that,. return to power was acceptable if the B&W recommendations were followed.
On 7: day, March 24, the reactor was taken critical, and the initial power ascension was begun.
The one-ucek surveillance programs required by 36W are still progressing, and the STP will be reviewed upon completion.
The startup was poreal, and no unusual circumstances developed that would indicat'e damage to reactor systems as a result of the transient.
Respectfully.
V t. fniGi
/*
&'J.'J. Mattinoe Assistant General Manager I
_;M:rrg:sc and Chief Engineer i
s