05000302/LER-1992-001, :on 920327,inverter Output Connected to Dc Sys Input Isolated,Causing Two Relays to Actuate & Opening Feeder Breakers.Caused by Relay Design & Failed Components in Relays.Relays Disabled & Dc Bus Monitored

:on 920327,inverter Output Connected to Dc Sys Input Isolated,Causing Two Relays to Actuate & Opening Feeder Breakers.Caused by Relay Design & Failed Components in Relays.Relays Disabled & Dc Bus Monitored

ML20095J402
Person / Time
Site:	Crystal River
Issue date:	04/27/1992
From:	Boldt G, Stephenson W FLORIDA POWER CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
3F0492-12, 3F492-12, LER-92-001, LER-92-1, NUDOCS 9205040001
Download: ML20095J402 (6)

LER-1992-001, on 920327,inverter Output Connected to Dc Sys Input Isolated,Causing Two Relays to Actuate & Opening Feeder Breakers.Caused by Relay Design & Failed Components in Relays.Relays Disabled & Dc Bus Monitored

Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2)(iv), System Actuation
3021992001R00 - NRC Website

text

....

.. l.

i W

t Florida Power C 0 R P O R AT t O N.

Crystal River Unit 3

- Docket No. 50-302 April 27, 1992 3F0492-12

- U. S. Nuclei. Regulatory Commission

. Attention:- ']cument Control Desk Washington, D.'C.

20555

Subject:

Licensee Event Report (LER)'92-01

Dear Sir:

Enclosed is Licensee Event Report (LER) 92-01 which is submitted in

- accordance with 10 CFR-50.73.

. Sincerely,

'G. L. Boldt Vice President Nuclear Production

~EEF: mag:

Enclosure

<xc:.-Regional-Administrator, Region,Il ProjectLManager,:NRR-Senior Resident-Inspector.

POST OFFICE BOX 219 + CRYSTAL RIVER, FLORIDA 326234219 a (904) 7954486 g

O100% t

popam O S NUCLFAR GOULATofW COMMISSION

- I APOROVED OMM 31604104 W)

Euws mo/s2 LICiENSEE EVENT REPORT (Lt_R) gig ig g g gs g tgen gugig TEXT CONTINUATION g g g gao ge 1,ogg f f M ilt 1"We% "4"w'* "oa^Wedrt'oM oc aun or uAnAccutut Ano suooct,wAs FACtLITY NAuE (1)

MU"EA N

N novum amse.

CRYSTAL river UNIT 3 (CR-3) o\\ slo lol o} 3l ol 2 ol2 olol1 ojo ol3hflol5 text y

-.-.au wmer.-=.m The reactor was stabilized in MODE 3, Hot Standby. The OP1 was checked, re-powered, and the ES busses returned to normal alignrnent. The Diesel Generators were secured.

The plant ns cooled down to MODE 5 for maintenance work on control rod position indication prior to post-trip restart.

A four hour report was made as required by 10CFR50.72(b)(2)(ii). This event is reportable under 10 CFR 50.73(a)(2)(iv).

CAUSE

The root cause of this event was relay design combined with the specific off-normal alignment of equipment utilized in the troubleshooting effort. Failed components in the inverter may have additionally contributed to the event.

The emergency power scheme for Crystal River 3 incorporates several levels of power sousces, one of these being four uninterruptible power supplies (UPS) [EE,UJX] called the Vital Busses.

Each 120 volt AC bus is powered from two sources, the preferred being the Vital Bus Static Inverter [EE.INVT] (a dual input inverter), end the alternate source being the 480 volt ES bus via a dedicated voltage regulating transformer which bypasses the inverter.

The inverter normally rectifies 480 volt ES AC power to DC power. The inverter then invertc the DC power back to 120 volt AC power, through a constant voltage transformer (EE,XFMR] within the inverter, which supplies the load, if the AC power input is lost, the inverter will instantly draw power from banks of lead-acid batteries [EE,BTRY] providing DC power and invert that to 120 volt AC power for the Vital busses.

At the beginning of this event, the 'C' Vital bus was being supplied from the alternate source, the 480 volt ES bus and voltaga regulating transforreer. because the normal ~ source, the inverter, was out

>..ervice for maintenance.

Under the troubleshooting package, several test configurations were to be established in the inverter to locate the root cause of the problem. A test corfiguration required that the Constant Voltage Transfcrmer [EE,XFMR], withir the inverter, be isolated by lifting the transformer leads. The inverter was then to be connected to ihe DC power input by closing the DC inpui. breaker. In the process of isplating the transformer, the electricians had only lifted one lead.

While this did take it out of the circuit, it did not isolate the transformer. When the DC input breaker was closed, y

tne partf ally isolated transformer induced an AC voltage (350 volts peak-to-peak)

L cnto the DC bus. The only apparent effect was the tripping of the icterposing relays used for normal 0PT feeder breaker control.

later testing showed a unique sensitivity in these relays, not shared generically threaghout the DC power system. The relay act Wa isolated the transformer and de-powered both ES 4160 volt busses. The loss c u bus power caused the EDGs to start.

NHC Form WeA (6-49)

=

[gqc m pena U.S. MuttENtIEGULATOfW COWulsecoN CPPfCDVED OMB Peo. 3150-4104

<m minE.mo.

LICENSEE EVENT REPORT (LER) gt,E,ogggnE,sgrgegougig g g g g o gEsguo j

TEXT CONTINUATION s?A*4h","E=JMTer$gT,Ogg

1. 1 W #PJ OF MANAGEMENT Areo BUDGET, VeASPENGTON DC 20603.

FACUTY NAME (O DOET NUMBER p)

LER NUWGER $8)

PAGECD SEQUENTIAL R4MFIN CRYSTAL RIVER UNIT 3 (LR-3) ol sl ol Jl ol 3l ol 2 ol2 o l ol 1l-o l 0 o l 4 lCFl o f 5 rw,, _. n m m.on M NT EVALUATION:

There were four main consequences of the loss of ES busses: first, both EDGs started en a valid undervoltage signal on the ES 4160 volt busses. Second, the motor driven Emergency Feedwater Pump (EFWP) auto started on ES bus undervoltage and EDG output breaker closure.

Third, there vas no power to the

'C' Vital Bus because it was normally aligned to the "A" " 4160 volt bus /"A";ES 480 volt bus.

Fourth, due to electrical alignment and effected busses, there was no power to the CRD motors and all control = rods inserted on loss of power. Each of these consequences is discussed below.

j The loss of power to the ES 4160 volt bu*ses is a'ccounted for'in the~ design of the plant. Should ~ power be lost to the busses, the EDGs auto start, come to synchronous speed and automatically power the ES bus loads. This action occurred as expected.

The EDGs carried the ES 4160 bus loads until 1538 for the "B" EDG and until 1918 for the "A" EDG. The only anomaly in EDG performance was a leak in the Jacket cooling

- system for the 'B' diesel.

The EFWP auto started, though there was no Emu jency Feedwater Initiation and Control (EFIC) system actuation. This is as designed.

Terever there is an undervoltage on the ES 4160 volt busses followed by an EDG output breaker closure, the motor driven EFP (EFP-1) auto starts as the bus is block loaded by the EDG. The EFIC system and

'EFP-1 both worked as designed and expected.

The 'C' vital: bus was deenergized because it was being fed by the ES 4160 volt bus /ES

- 480 volt hus.

This bu: powers the 'C' Chaisnel of the Reactor Protection System (RPS). When the power to the syste r fails, the channel trips.

In addition to the RPS, channel "C",

the "C" Vital Bus also powers the Recall system, a passive data recording system, and an annunciator events recorder.

The loss of power caused a less of some transiant information normally used to analyze an event.

The CP,D motors lost power and all the control rods inserted into the core. The CRD motors are designed so that a sectioned roller nut engages a lead screw on the control rod. ~ The roller nut sections are designed to be disengaged from tht lead screw by springs. The roller nut is held in the engaged position by electromagnetic form. The roller nut is turned by progressing the electromagnetic field around the contul rod (moving in discrete steps), turning the roller nuts around the lead screw, raising and lowering the rod in the core. When power was lost to the CRD motors, the roller nuts disengaged and the rods inserted.

"NRCFORM3eM.

U.S. NUCLEAR HEGLAATORY COMutseeON APP 50VED OMJ N'l 31M104 EXPIRES 4/30/92 l

LICENSEE EVENT REPORT (LER) gs,r g r3D 4nDc g g sP goutgeouPgtg TEXT CONTINtJATION gg7g,y,ogguATEto;7g T5NERwoYrdtEdEE5c7 eioYoEc'[

oF MANAGEMENT AND BUDOFT, WAsem TON DC 20603.

F ACILITY IWeE (11 DOOf' NUMBER CO LER NUM90R (19 PAGE (3) stouturw Reviseon YEAR mim wem

~ C8YS !.

RIVER UNIT 3 (CR-3)

]o ( 5 l 0) o l 0 l 3 l oj 2 el2 ol0l1 ol0 o l 5 lOFl 0 l 5

,cxt_,,

_. -. -, -.on CORRECTIVE ACT11NSt There were several corrective actions taken to preclude recurrence of this event.

Prior to plant restart, the relays that send the remote open signal to the feeder breakers for the OPT were disabled.

These relays provided no protective relaying

- functions so-there. is no loss of equipment protection. The feeder breakers can now

.be manually opened with control switches installed in the 230 KV switchyard prior to startep.

A second action to monitor the DC bus for noise prior to reduced RCS inventory ' operations will be implement J during the upcoming ' refueling outage.

Lastly, a numan performance review will be conducted on the inverter troubleshooti.ng evolution to determine if the risks should have been known or anticipated. The first action _is already completed, the others are schedu}ed for completion _by July 1,1992.

PREVIOUS SIMILAR OCCURRENCESi

- A 'similar actuation of these relays occurred during the mid-cycle 8 mai. tenance outage.

See LER 91-10 for details of that event.

J NHC Form 3a.A (6-49)