05000289/LER-1992-001, :on 920122,inadvertent Emergency Feedwater Initiation Occurred During Planned Maint.Caused by Installation Error.Wire Errors Affecting Sys Operation Corrected

:on 920122,inadvertent Emergency Feedwater Initiation Occurred During Planned Maint.Caused by Installation Error.Wire Errors Affecting Sys Operation Corrected

ML20097D720
Person / Time
Site:	Three Mile Island
Issue date:	06/04/1992
From:	Broughton T, Knight M GENERAL PUBLIC UTILITIES CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
C311-92-2074, LER-92-001, LER-92-1, NUDOCS 9206110094
Download: ML20097D720 (7)

LER-1992-001, on 920122,inadvertent Emergency Feedwater Initiation Occurred During Planned Maint.Caused by Installation Error.Wire Errors Affecting Sys Operation Corrected

Event date:
Report date:
Reporting criterion:	10 CFR 50.73(a)(2) 10 CFR 50.73(b)(2)(ii)
2891992001R00 - NRC Website

text

l e

j GPU Nuclear Corporation i

,[

Q Mf Post Office Box 460 Route 441 South Wddletown, Pennsylvania 1M67 0191 717 944 7621 i

TEl.EX 84 2386

" riter's Direct Dial Number:

(717) 948-8005 June 4, 1992 C311-92-2074 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555

Dear Sir:

Three Mile Island Nuclear Station, Unit 1 (TMI-1)

Operating Licensing No. DPR-50 Docket No. 50-289 LER 92-001-01 This letter transmits Licensee Event Report (LER) No 92-001-01 regarding.in inadvertent Emergency Feedwater System (EFW) actuation which occurred on January 22, 1992 during planned maintenance due to an installation error.

Public health and safety were not affected. This revision is being submitted to clarify the root cause to include inadequate startup testing as a contributor to the event.

This LER is being submitted pursuant to 10 CFR 50.73.

The abstract provides a brief description of the event.

For a complete understanding of the event, refer to the text of the report.

Sincerely, 2Ki-u lh'-

T. G. Br ghton Vice President and Director, TMI-1 MriX Attachment cc:

Region.I Administrator THI-l Senior Project Manager TMI Senior Resident Inspector 9206110094' 9'20604

~

k PDR ADOCK 05000289 7

g S

PDR GPU Nucleaf Corporation is a subscary of General Pubirc Ubhbes Corporabon

e m..

u i wu i.... om. w., w

.n.c su a

o. e e o.,, w,w n LICENSEE EVENT REPORT (LER)

THREEMILEISLAND, UNIT 1 0.1.mluoj218 :9 FLl0 6

" ' " ' " ~ " * *

^

INADVERTMT~ ERG 5ti!CTTEEDVATtf ITiT11A110N DURING PLAIIRED MAINTENANCE

"'" +

DUE 10 INSTALLATION ERROR

....a......-

u..v..-

.i~. e. c o, i...c n n.u...os. i e -.

u.

.u c ~4i

" d.;;t-eft,;

n.

.n.

o! 610 ! %.8 1 ! !

O11 2 l2 92 9 12 0 ! 0 11

~~

0 11 O l6 0 14 912 oisioicici i L

~

,i....o.,... - n...u..v..,v.,

..u......no i.c..i,e

..,m,

===+

N 1

.y

.1.,

....n.,,,

y n.m.,

... in n,.

= = 4. n i

. n u 4.,

i n,,,,

g1 00

, n =,. ~ i o n.ns,...,

] g.g 4,,.;g

, i q n.

..n..

. n.. ann, n...nium.,

...e u.

M.Oli.*H H.I

.,$i.it$lle I 04, $ 6.hl b s t t ie. u. co.a.c,...,m. u. n e, 1.. m-i

=

n a,-2.......

4 M. R. Knight, TMl-1 Licensing Engineer 7l7 948 i 8554 i 1 i i i i co a n...........c co o..,

.itv.

...c

.... n.,.....o.,

n, C.vst 5 v 8,14 CDu.0%tst

' ', f M,'

,,, h

c.ust svPtw C0w.0%IN, v.)..C 98,8

,, i' g

I

_.L t t

' I i 1

i i !

! I I l l l 1

! ! I I l ?

- o..., i t...,.... cm o.,

o..

4 th in, p.e one. th.gpgO Sv8M@ON 04 70'

%Q l

l

.,. ic, a., o _.,..

~.., n.,

Inadvertent Emergency Feedwater (EFW) Actuation During Planned Maintenance Due to Installation Error TMl-1 was operating at 100% power. During a planned maintenance activity.

Emergency Feedwater (EFW) was inadvertently initiated for a short time.

This event is reportable in accordance with 10 CFR 50.73(a)(2)l,iv). The actuation occurred due to a construction wiring error resulting from n.odifications during a previous outage. This event was caused by lack of drawing clarity, inadequate supervisory oversight and failure of

.e test procedure to verify separation between HSPS channels and trains as required in the modification test program.

All equipment functioned as expected considering the wiring errors There was no adverse impact on nuclear safety.

Wiring errors which could affect system operation have now been corrected. A detailed walkdown will be performed and the applicable drawings will be revised to reflect the as built configuration. TMl-1 has in place sufficient procedural controls to preclude or identify wiring errors and as such this event is considered to be an isolated case.

The potential for single failures to cause an inadvertent EFW actuation had been previously evaluated and with the NRC's concurrence this wat deemed acceptable.

No additional action is considered necessary.

.,c,..

ase c,,or:at zesa u s wuct eam atovtatom V corwsio-

...,m.,

LICENSEE EVENT REPORT (LEW

',5,'s",','.',"f,,Tf,"o*'J2!!,' 4*J.,'l,*J," '/;;j TEXT LONTINUATlON C",,',W,';!%'."nti 'rO',t !','J,"",'l# '",' 0'RTf!

74,1. ','#,M..a?"*23.u*i,!.2. Hnn10 'l3a o

.. a.

w

~w x mi

, acit t1 T 8e.W6 6 0 D8[R I 1.fUw, $ R I D gg a e,pygg a,g, p.g j 3

" c !.T,*-

l'N',9 o-THREE MILE ISLAND, UNIT 1 0l 6 o p ;o o o l2;Oj 9 92 0lOj1 0;I 0; 2 or rum -.

,. uc %,u.. n n Inadvertent Emergency Feedwater (EFW) Actuation During Planned Maintenance Due to Installation Error I.

Plant Operating Conditions before Event:

TMI-l was operating at 100% rated power.

The ICS was in full automatic control. Work was planned to be performed on the "lB" Inverter.

11.

Status of Structures, Components, or Systems that were Inoperable at the Start of the Event and that Contributed to the Event

None III.

Event Description

This event occurred at approximately 1300 hours0.015 days <br />0.361 hours <br />0.00215 weeks <br />4.9465e-4 months <br /> on January 22, 1992. The operating crew had performed the prerequisites for transferring the "B" 120 vac vital bus (VBB) (ED/BU]* from its normal source, the "lB" inverter (ED/lNV]*, to its alternate source (TPA), in accordance with Operating Procedure (0P) 1107-2, " Emergency Electrical System." During the transfer, it was necessary to temporarily deenergih vital bus kBB.

OP 1107-2 includes actions and precautions which identify the loads on the bus, the effect of loss of VBB and what precautionary actions are to be taken.

When VBB was deenergized in a aordance with the procedure, the follcwing events occurred in addition to those described in the procedure:

1.

Main Steam supply valve, MS-V13A (SB/V)', opened and the turbine-driven EFW Pump (EF-P1) developed discharge pressure sufficient to produce flow into the Once Through Steam Generators (OTSGs).

2.

Valve controllers asscciated with EFW control valves, EF-V30B/D [BA/V]*, switched to 50% operating level setpoints, an ope, ating range level input, and received a level signal less than the 50% sotpoint.

Both control valves opened and EFW flow was initiated te both OTSGs (AB/SG]*.

3.

Annunciator J-1-3/J-1-4 (IB/ ANN]* "EFW ACTVA'ED OTSG A/B" went into alarm.

The operator identified the improper system response, placed EF-V308/D in

" manual" and closed both valves.

This terminated flow to the OTSGs very shortly after it had been initiated.

l l

ac r.,v sess ues t

l

v. ~uu..... uut. o., u-,o-

.j. g on. m

,,,, m.,,,, _,

...,.n.~

LICENSEE EVENT REPORT (LEfD

"'3^4'f, ',,q'fJ',"OQi',/ W$l.Ta'J;2 TEXT CONTINUATION D7?,'f ';'/.,'."O]!rO',M" I',','J*l lM"f 5 o, o.s a isi.w.m, C. r# %C WQt

" W,l C 2 7 8 M.'

smem

,&CILif,h&W840 DCeC D G 1 NJarag e @

(A hvMS$A +

P&Gt 0 n ':,;:-

?);;:

THRE: MILE ISLAND, UNIT 1 opo o ;o 2 l8 l9 9j2

__ 0l0l1 0;1 0; 3 0l6 os v u r -

.-y,- -, w, The Control Room Operator (CRO) sent an Auxiliary Operator (AO) to EF-P1 (BA/P)*.

The CR0 attempted to close MS-V13A from console center.

MS-V13A did not close.

Within the first minute following the EFW actuation the overhead Annunciator J-1-3/J-1-4 cleared and the CR0 was able to close MS-V13A from the Control Room. These events occurred without any EFW functions being placed in " defeat."

EF-V30B/D continued to indicate that an actuation signal was present after the overhead annunciator was clear (the AUTO Light was "on' above each valva controller). The operators attempted to clear this signal by placing the Train "B" Loss of Reactor Coolant Pumps (RCPs) Defeat / Enable Switch to " defeat." This action had no apparent effect.

An operator verified that the Heat Sink Protection System (HSPS) Train "B" cabiret had swapped to the backup power source as designed and that there were no unexpected indications locally at the HSPS cabinets. With EfW flow terminated and without any cause of t'l actuation determined, the shiff supervisor and operations management decided to reduce the present vulnArability by restoring power to VBB from TRA.

When VBB was reenergized from TRA, EF-V30B/D transferred (without any operator action) into " Auto," with 0% setpoint and a good startup level input. This is the normal state for these controllers.

With power restored to VBB all indications for HSPS both locally and in the Control Room appeared normal.

a The actuation occurred due to a modification construction error during TMI-l's Cycle 8 Refueling Outage (8R) which had gone undetected until this time.

8R occurred between January and March, 1990.

Two train "B" nests had been powered from the Channel 11 DC distribution bus (located in HSPS Section A2, Rack 3) instead of the "B" Train DC distribution bus (located in Section A2, Rack 4) as required. The net effect was that deenergizing VBB caused a loss of power to a portion, but not all, of the Train "B"

nests.

The causes of this event were identified as follows:

1.

The 8R construction drawings were not clear and could easily be misinterpreted.

it was believed that all of the problems associated with drawing presentation and wiring errors had been identified and corrected prior to turnover. The subject installation error went undetected.

Problems with drawing clarity were addressed in the Post 8R critique of the HSPS modification to prevent a recurrence in future modifications, p._ _.,

I'

~

.g oa.

u. ~un....

.ut.,o.,co - w o~

.....n.w,,

LICENSEE EVENT REPORT (LERI M,'g;l#lf,,T,l',"c "'!;ti's,'%fl"".!l.*1'.M TEXT CONTINUATION 20".'S"",!\\'."$1'C',MC',','/ F/ OR."1 Uh.t.','.WJf,7. 3%.oc'i,O..s ;P E M2 Mi ';

n o s.ciois o

o~~~

,wu,, m. m

=.., e. m

,,,,,vo....

,.o,,

" e,;e-

'v,p THREE MILE ISLAND, UNIT 1 o s lo lo lo 2l8l9 92 0 l 011 01 O! 4 or 0l 6 1

1 rw w, -.

,u

- m - a..n m 2.

The Startup and Test (SV/T) group was not involved during tha init.ial phase of construction as has besa a practice in other outages including the most recent 9R outage which occurred between September and November, 1991. As a result, no THI-1 cognizant SU/T engineer was present during SR at the work site to provide technical guidance to the construction workers during the initial construction phase of the modification.

Inadequate supervisory oversight was identified as a probable contributor to construction errors described above during the BR critique and recommendations from the critique were implemented to increase SU/T involvement during construction of complex modifications with safety significance.

3.

Th0 generic electrical testing that verifies all construction wiring shous 3 have corrected this problem early in the test program, Functional testing requirements to verify power separation between HSPS channels and trains should have identified and resulted in correction of the wiring problems.

Inadequate implementation of the SU/T modification test program requirements allowed the power supply wiring error to go undetected during both generic an ' functional tests.

IV. Component Failure Data

There were no component failures associated with this event.

V.

Automatic or Manually Initiated Safety System Responses:

(

All equipment associated with a Train "B" EFW actuation functioned as expected considering the improper wiring configuration. MS-V13A opened after receiving an actuation signal caused by loss of power to modules (JB/lH0D]' within the actuation control logic, t

Based upon obsetlations during the event and subsequent testing it was

oncluded that the actuation signal to MS-V13A and MS-V130 cleareu before MS-V13B received an "open" demand.

By design, MS-V13B is opened approximately 40 seconds after MS-V13A.

EF-P1 came up to speed and provided sufficient discharge pressure to deliver water to each OTSG. The motor-driven EFW Pump, EF-P2B, did not start because the HSPS actuation signal has an interposing relay powered from VBB which starts the pump.

With VBB deenergized, the relay did not energize to start EF-P28.

EF-V30B/D received invalid actuation signals and level inputs due to deenergized modules in the logic which provides the inputs to the controllers.

Given the faulty inputs, the valves functionad as expected.

se:C pom 2n.A t.491 l

i

l jig =.

u. ~oc o.... o vu, o., m

..o~

i aP4s,t $ 4 '30 9J LlCENSEE EVENT REPORT (LER)

' T*',7,%t"C',"J ' $T's / %lf'51,'.* '.".' M

~

TEXT CONTINUATION EC'",,'M',llPo'."ini';"l'.11"O,'G '"i MTU O, M. ~*,,'a".'4,*,",,U. v,'om' :.::O ' NMT.' i : ? M a

.-~caa oc na o

a oi s.w

,.c io,, ~.. n,

e n -.. a.

o.....,,,

..a+

yr,;P n.*,p THREE MILE ISLAND, UNIT 1 jo siojojo 2 l 8 l 9 9l 2 0l 0l 1 0{0l 5 or 0l 6 i m,, -.,

o.c %.m.

• VI. Assessment of the Safety Consequences and Implications of the Event:

A.

There is no adverse impv.t on nuclear safety from Train "B" modules being powered from Channel 11.

The Train "B" modules incorrectly powered from Channel 11 should receive power from Train "B" power supplies which are also powered from VBB. This condition would not resalt in a single failure which wculd have consequences worse than a failure of VBB which has been evaluated.

If power is lost to VBB, the HSPS System is still capable of performing its design function. With a loss of power to VBB, Channel 11 is actuated and Train "B" is incapable of performing its design function, but Train "A" is unaffected by loss of VBB.

B.

The design basis for EFW/HSPS considered system failures in which the EF-V30 valves would fail closed or open, t.oss of air or signal causes valve closure.

Under conditions of partial train power failure, the EF-V30 valves on nne train may fail open.

This does not impact nuclear safety because a reasonable time period (at least 15 minutes) is available for operator action to locally isolate the failed valve and the ability to terminate flow to one or both OTSGs is still available from the Control Room. This can be accomplished with EF-P2A and EF-P2B control switches and EFW discharge header crossconnect isolation valves, EF-V2A/B [BA/V]*, pushbuttons.

VII. Previous Events of a Similar Nature:

None.

Three previous EFW actuations nave occurred at TH1-1 since the HSPS was installed.

In the case of each previous EFW initiation, the system responded as designed when a valid actuation signal was sensed at the input to the system. The cause of previous events was external to the HSPS.

The cause of this event was an internal wiring error.

Therefore, the previous events were not similar.

Vill. C(rrective Actions Taken:

1.

Through entries in the night order book, the operating crews were advised of the events which h:d taken place and given instructions on

/

the actions to be taken if power were lost to VBB.

2.

The power cables from Train nests "A2-6-6" ana "A2-6-7" have been rewired to provide proper termination from the Train "B" power supply.

Functional testing, performed in accordance with a Special Test Procedure (STP-92-004), verified power separation between the HSPS channels and the "B" Train.

3.

A walkdown of Channels 1 & 11 and Trains *A"

& "B" was performed.

It was verified from the walkdown that the channel and train nests are povered from the proper train and channel powar supplies.

No l

GCc Penn,asA 16Jh91

1

'ee Q,Pu 366A y$veVCg,AhAf0VL4TD4'CDM" N O" apenh(D OUS ND 318D444 g neim,, 4 30'92 LICENSEE EVENT REPORT (LEP.'

'"'"'"**"D'S""""' '"'"'"'"S

TEXT CONTINUATION

' ' U'N O "'.',S".? "'J'ila".K.'Es'?iMAf fJM su.'ZO, "&n'J?,'"OGZ"iM"'"

..c m..... n,

noc...

v.... i,

" M;', :~

pyuz w-THREE MILE ISLAND, UNIT 1 o plo ojo1 2l8l9 9l 2 0l 0l 1 0j1 0l 6 or 0l 6 w.

..e,- =...

additional testing of the "A" Train power supply was required.

Other discrepanciesbetweenthedrawingsandtheexistingconfigurationwerel noted during the walkdown. However, it was determinid that these discrepancies would have no effect on system operatic,1.

IX.

Corrective Actions Planned:

Wiring errors that could affect system operation have been corrected.

In order to correct the other noted discrepancies, a more detailed walkdown of the HSP5 povar supply wiring is required and will be performed at the next outage of opportunity.

The affected HSPS drawings will be revised to reflect the as built configuration.

X.

Conclusion The potential for single tallures, as occurred during this event, to cause an inadvertent EFW actuation were identified during THI-l's Cycle 8 Refueling Outage (BR).

Prior to startup from the 8R outage, modifications were performed to eliminate the potential for inadvertent MFW isolations resulting from a pari.ial or total loss of power to the HSPS.

However, evaluations during 8R confirmed that the consequences of an EFW actuation are acceptable. The prevention of inadvertent EFW initiation resulting from a single failure is not a regulatory requirement.

This was reported to the NRC in a letter dated July 5,1990

_ during THI-l's Cycle 8 Refueling Outage (8R). With NRC concurrence it was concluded-at that time that modifications to prevent EFW actuations on partial loss of power would not be required. Although, partial loss of power to the HSPS r" 4 not result in loss of safety function and does not result in a safety nard, this event was undesirable.

Having corrected the wiring errors that resulted in this event, GPU Nuclear continues to believe further modifications are not required.

The contributors to this event, which caused the wiring error, have been identified as a problem with drawing clarity and inadequate supervisory oversight by individuals knowledgeable of the HSPS design and the modification be 49 performed.

In addition, inadequate implementation of the SU/T modification test program allowcd the wiring errors to remain undetected. The involvement of the SU/T group during-the initial phase of construction, as in nther outages including the most recent 9R Outage, and. proper implementation of the SU/T modification test program requirements would have avoided the wiring error or identified its existence. Therefore, this ; vent fs considered to be an isolated case.

No additional ~ action is considered to be necessary.

• The Energy Industry Identification System (Ells), System Identification (SI) and Component Function Identification (CFI) Codes are included in brackets,

"[SI/CFI)", where applicable, as required by 10 CFR 50.73(b)(2)(ii)(F).

3 BAC 7een 3e6416491 s