ML071020202

Memo from J. A. Olshinski to D. G. Eisenhut Re Loss of High Head Injection Capability at McGuire Unit 1 and Reconsideration of Technical Specifications 3.0.3 and 3.5.2

ML071020202
Person / Time
Site:	McGuire
Issue date:	04/12/1982
From:	Olshinski J NRC/RGN-II
To:	Eisenhut D Office of Nuclear Reactor Regulation
References
FOIA/PA-2007-0140
Download: ML071020202 (7)

Text

L.AJLA V To" .NUCLEAR UNITED STATES 0 ULERREGULATORY COMMISSIO c C ~REGION 11/ lC 101 MARlIETTA ST., N.W., SUITE 3100 ATLANTA, GEORGIA 30303 SAPR 1 2 1982 SSINS: 50-369 MEMORANDUM FOR: 0. G. Elsenhut, Director, Division of Licensing Office of Nuclear Reactor Regulation FROM: J. A. Olshinski, Director Division of Engineering and Technical Programs

SUBJECT:

LOSS OF HIGH HEAD INJECTION CAPABILITY AT MCGUIRE UNIT I AND RECONSIDERATION OF TECHNICAL SPECIFICATIONS 3.0.3 and 3.5.2 On February 12, 1982, while operating at 50% power, McGuire Unit I experienced a loss of both high head safety injection centrifugal charging pumps (CCP) due to entrainment of hydrogen gas in the suction line. The related LER that was sub-mitted describing the event is attached as Enclosure

1. Region II's preliminary evaluation of this event indicated both design and operational concerns.The design concern Is that a single failure in the non-safety related hydrogen dampener for the positive displacement pump (PDP) can disable both trains of the safety-related centrifugal charging pumps. Region II issued a Confirmation of Action letter on March 25, 1982 (Enclosure

2) to confirm isolation of the hydrogen dampener system from the high head safety injection system. Although our preliminary discussions with Westinghouse indicated that this particular dampener system may be a unique design, Region II- forwarded a proposed Infor-mation Notice (Enclosure

3) on the event to' the Office of Inspection and Enforcement because of the possibility of existence of similar, if not identi-cal, dampening systems. A proposed Abnormal Occurrence report (Enclosure 4)has been prepared and forwarded to AEOD.The operational concern (Enclosure

5) raised by the McGuire Senior Resident Inspector involved the action statement required by the applicable technical specification in this event. We view the concern on the action statement as a valid concern, but recognize that selecting the "safe direction" for the action statement is dependent on a number of assumptions including the probability of correction of the problem prior to loss of pressure control. We therefore request that NRR review the. adequacy of the Action Statement for this Technical Specification as described in Enclosure 5.My staff is available to discuss any questions or concerns you may have regarding this event.S A. 01 shinski

Enclosures:

• (See Page 2)*. CONTACT: Richard L. Fiedler (242-5550)

-88O~2p~~~G1 BeF AO~ 5C3~

DuxE POwEn COXtP.A.Y Pow1a flu-tD:dO 422 Sourn Cuiucu Strzz?, CIIARLOTE, N. C. woa.Laýrl cloi u fe I WSMLLe.AH

0. .&IPOCm4.4.

V',Ct 009%t $rtte March 24, 1982 273.4'c 3 Hr. James P. O'Reilly, Regional Ad=inistrator U. S. Nuclear Regulatory Commission Region 11 101 Yarietta Street, Suite 3100 Atlanta, Georgia 30303 Re: McGuire Nuclear Station Unit 1 Docket No. 50-369

Dear Hr. O'Reilly:

Please find attached Reportable Occurren:e Report RO-369/82-15.

This report concerns T.S.3.1.2.4, "At least two charging pumps shall be operable...";

T.S.3.1.2.2, "At least two of the following three boron injection flow paths shall beoperable...";

and T.S.3.5.2, "Two independent ECCS subsystems shall be operable with each subsystem comprise' of: a. one operable centrifugal charging pump." This incident was considered to be of no significance with respect to the health and safety of the public.An update to this report will be provided when the final corrective action has been determined.

Very truly yours,"O William 0. Parker, Jr./PBN/J fJ Attachmenr cc: Director Office of MýLnagement and U. S. Nuclear Regulatory Washington, D. C. 20555 Program Analysis Commission Records Center Institute of Nuclear Power Operacioas 1820 Water Place Atlanta, Georgia 30339 Mr. P. R. Bemis Senior Resident Inspect or-NtRC McGuire Nuclear Station DLKE POdER COMPAiN McGUIRE NUCLEAR STATION REPORTABLE OCCURRE2ZCE REPORT NO. 82-15 REPORT DATE: March 24, 1982 FACILITY:

McGuire Unit 1, Cornelius, NC IDENTIFICATION:

Loss of Both Centrifugal Charging Pumps (CC?) 'hen Hydrogen (H 2) from the Reciprocating Charging (PD)Pump Suction Dampeners Entered the Suction of the Centrifugal Charging Pumps DISCUSSION:

On February 12, 1982, during an attempt to fill and vent the PD pump suction piping in preparation for returning the pump to service, opening valve INV-217 (suction isolation to the PD pump) resulted in air and 112 in the PD pump suction piping flowing into the coron suction of the two CCP's. Con-trol room personnel noticed that CCP IA motor current and charging flow had begun to oscillate, indicating cavitation, and thus subsequently swapped to CCP IB and tripped CCP IA. Approximately 30 seconds later, CCP 1B began to cavitate and was tripped. This resulted in charging and letdo-n being secured.Both CCPs were declared inoperable at 2058 while unit I was in mode 1, 50%power operation.

This incident is reportable pursuant to Technical Specifi-cations 3.1.2.4, 3.1.2.2, and 3.5.2.When the Nuclear Equipment Operator (YEO) who had opened the valve heard a page announcement that charging had been terminated he immediately called the control room. As a result.of his call, the control operators then suspected that gas from the PD pump suction was entering the CC? suction, and instructed him to reclose the valve.The Shift Supervisor and three NEO's vented CCP IB and 1A thr6ugh the overflow piping. However, when CC? IB was restarted it i.n-ediately began cavitating, and was again tripped..

After donning anti-C's and shoe covers, the Shift Supervisor entered the pipe chase znd vented the suction pipes for both pumps, one at a time, for a total of approximately five minutes. During.this time, the NEO's revented CC? IB. Pump IB wad restarted and verified to operate properly after which charging and letdown were rc-establinhed.

EVWIUATION:

Due to the pulsating suction flow, characteristic of recipro-catift, pumps, the PD pump is equipped with a suction dampener consisting of a vertical section of twelve inch pipe with 112 gas overpressure.

The water level is controlled by two solenoid valves which supply gas when the water level is too high and vent off gas when the water level is too low. These valves can be controlled automatically by level switches, or manually by svitches mounted on a local panel.After the event, a check of the dampener level control system found the refer-.ence pot and leg empty. An empty reference leg would indicate high water level to the level switches which would result in a continuous supply of 11' to the Report No. 82-15 Page 2 dampener.

The glass cover and meter movement of the level switch for valve 1%V-838 were also found damaged, but the effect of the d.mage on switch opera-tion could not be determined.

The means by whIch the reference leg was drained could not be determined.

No leaks were found when the reference leg and pot were refilled.The PD pump was isolated and drained in order to install instrument taps and pressure sensing devices for a station modification.

After draining, the vent and drain valves were closed. Level in the reference leg was apparently lost while the piping was isolated since it had worked properly before but not after this period. When the NEO's attempted to fill the system, a slight hiss was heard. The NEO was not concerned because he knew that H7 gas was involved in the dampener operation.-

Actually, this indicated that the level controls might not have been working properly (supplying H2 when it should have been vented).The MZ0 was likewise not concerned by the sound of water flow when he opened the PD pump suction valve because he ex.pected water to flow into the PD pump suction.Hydrogen header pressure supplying the dampener was approximately 110 psi& which could eaiily displace water in the charging pump suction piping. Judging by the volume ccntrol tank level changes during the event, it is estimated that about fifty cubic fedt of water was displaced by the gas. VCT overpressure generally ranges from 50 -30 psig.The H 2 released to atmosphere did not fill any of the areas to concentrations sufficient to cause combustion as a result cf a spark. Hydrogen and air in the piping was in no danger of combustion due to the absence of an ignition source.SAFETY A-NALYSIS:

When the CCP's were inoperable, no emergency core cooling was available above 1500 psig (safety injection pump shutoff head). Emergency boration and reactor coolant makeup were also not available durinn this period.Any significant decrease tz Tave would have resulted in a corresponding drop in pressurizer level but system pressure could have been maintained by the pressur-izer heaters until level dropped below 17%. It was essential that the unit be maintained in a steady state condition until charging and letdown could b3 restorzd.The unit remained in a steady state condition during the incident end the health and safety of the public were not affected.If the restoration of makeup coolant flow had been delayed until after the pressurizer inventory was lost, or had a transient occurred forcing a loss of pressurizer inventory, a reactor trip would have ensued. With the loss of pressurizer water, loss of pressure control would occur. The lower limit of the pressure excursion would be determined by system hot spot saturation pressure.

Forced core cooling would continue unless system pressure loss or system voiding required stopping the pumps at which time natural circulation (Report No. 82-15 Page 3 would provide the =,ans for core heat removal. Reactor coolant umperatuce trends would be dependent on stecm generator heat transfer rates ard core decay heat generation.

When reactor coolant system pressure decreased to below 1Q00 psig, the safety injection pumps would begin to refill the system. Safet7 injection would con-tinue until pressurizer pressure control was regained.Steam generator water level, steam flow, and feed flow affect steam generator heat transfer and the reactor coolant natural circulation rates. The steam generator parameters are relatively simple to control; therefore, the recov,'y from the postulated loss of makeup flow incident is considered to be within the capabilities of the station.CORRECTIVE ACTION: The irmediate corrective action was to secure the CCP's, isolate the PD pump suction, vent the CCP's and suction piping, and to return CC? 1B to service. CCP IA was tested and returned to operable status later the same day.Due to several incidents Including this one, a memorandum to standardize the praciceteof isolating and draining equipment has been distributed.

Effective February 25, it is required that at least one drain and/or vent be red tagged open when any co=ponent is isolated for maintenance.

Although such action may not have prevented this Incident, it is good operating practice which was not iolloved in this instance.The PD pump and suction dampener have been isolated from the charging system.Duke Power Co=pany is evaluating the system design to determine what temporary and/or permanent changes are necessary to prevent recurrence of this type event.The PD pump will not be returned to service until such a temporary or permanent change is =ade. A followup report describing any temporary changes made to the system and permanent changes planned will be submitted when theze evaluations are complete.

M et F %A U. Pý N¶%qL5AA ReCJLATOeMY CCW4I.Z~jC-4 UC/NSE EVENT RfEP0f1T EXHIBIT A=h~rQ%. LOCC 1MU.A49 PONINTCR "YPt ALL o~c~wRCO ip~foRwAflcod-T%:-lc ix s I S t I _0! 0 o101 00 oi o-I 010 I C

---.JQ S Lhl.Z't&81 Czj 1b6C4 lowwi~klh J*b.A a %. p4z4 riftA SIC a(iJ I LI$ 015J 10 101DM 61jtC 1 01 21 11 1 R1 20a 0 1 1 2j 41 8[20=52 hile in mode 1, durinz an attemptl to fill and vent the rcciprocating charging_]

o t m (PM) suction nioine in prepar3ation for returning, the pump to service, both 11 -o,-s (CCp) were A-e1ared innnernble when hvdrogen fron thel E'D punr suction der-peners entered the suction of the CCP's ciusing cavita:tion.

06 Th's _ .3.1.., 3.1.,.2. and 3.5.2 which is rportiable 2er T.S.6.9..;

a, 112(c). Since the unit remaired in a steady state condlition until charging and og~ lletdown were restored, the health and safety of the public were unaffected.

uvrw CA= CLM co. VLI 5CCWit C.r WC.&Ica vALftVrTC -j c I SU f 0 ~ I?I Cr jjJ J XJ0 IX lX I X iX IX (E U ( D 3 S I6 II I3 I13 6I SI=lb 1Rs 2 to se M.I 2 ' L 1 5:U La -01P.f. ( VC-.N i laL~t~,.I

@ ID~w4 it l I 31 I _W L)O tW:4&CJ.Lt ! o N .-r w, ,,&#f =up. e SU A A. 7I 1'0 ftosd3 IEr~ 1. 01 .0!2 0~[i;o Th~s in.cident resultod from the failure of the V 4 ydrogen Control Svzte--n on the ?'D pun? sucti~dn darn-ener (the reference pot and leg were found empty, cause uiikno-.n'.

1, jbzallate corrective action was to secure the CC', isolate the ?D pump suec-ic~n j vent the CCP's and suction pipin-, and return the ^C?'s toj service. U'se z~f the I-?D svsrem -wil be discontinued until corrective action to prevent rIcurrence can, 1 a *a c and~n~ cr ai.-..fl P dcIJ*- -V (TI.O Op 1T 1 p~w" 4ýC4A STATUSLAr-

'9. 5R Of I A Y ~zCpcg t!SL.i(G I C1 5 C:2~ [AI ~~ Personnel CObSCr-,-ALAJo

^41IXAICZ ZO *d),UINO FCIVr 1.Mt'i4 P LLA48 fl .,;h ,/A , L I I I 10 II I4 1.pt"C I I- L0 373V- Ex 71 ~ K~i 12 27L1 _ 1 N/A AA3 a5 If OR O12 ý CE Ij~j 1 21 OkIP _N/A& a II ill I= CS 0A C"A41[~ tO FACUTY--- TW' O -/,A .'I .4'a to 4,4 do...IldA~iOFr4CPIRE?$A 14C Phillin B.. Nardoci (704) 373-7432 UNITtD STATES NUCLEAR REGULATORY COMMIISSION fEGION II 10 ATi LII A ST., L 0100 ATLANTA, GE'0IWIA 301303.Eceloa'e ý_".192 Duke Power Company ATThi: Mir. 11. 0. Parker, Jr.Vice President.

Steam Production P.O. Dox 2178 Charlotte, NC 28242 Shjct: Confirmation of Action -V)ockeot 11o. 50-369 Tlisi letter rcfers to the telephone conversitlo, on f1Mrch 24, 1982, I-etOlen:Y. J. 01shinski of this officL- and Mr. Ih. Imthrf'urd of yuur uffice conce"-i,;i o, mrct~ve actions on your repn-rtable occtrrunr.,;

repurt [~o. a 2-s. Thi e ;.'rtaI'ii:

tt) 91: i..:;s of bothi ctnntr r',ina1"cnar(iq

!1( i ;p-; .hi' to hyJrulr 'L Ijr..d;om Ftrai,':.,e;It

r, .,r, o .;j the pusitive ,lisplacentient (P)) t,:wpI) to sc.vice at y1!11lc,;t 1 01i" .i "y.:..::;.J t'p.un the r.att:r di,.Cussed, it is m'ir un.!ersutndini trM t thof: t,,,i.,i Iry at:' , *n': .;er' or wiil Iv initi,tLeJ:

rt. , PD~ I) Tr,, , ' '. v I- vt, a i- -b.ul , " i,,,..; t;hc2' 01-.1,'*jI, sYJ#.:. 1."o--.., n.pt% coi;trol .v, suct~ion datip';;,..] e to :.,f*%!V;!)

t' i; m -0,} toili. t/: ,.f 3. Foluwi ny any d~sion chati;es and prior -et,,nirgig h ' pu..tp ,,,,L -',ti.)n la*;pleer syste..; to service, sutxiit your evaluatiun to this office for cur re.v i edi.Viv.dz': note that the rýiiarting requir.Lunm, dt. identifit.-d ibovu arc, not t. i ,t).,i .z.x of I'mnagement j:d Oudget clearance as rcquired by P. L.90-511.!f y..ur* u,'oer, Ltad i:g of our discussio', is diff:renL fro:n that stat,ý'; ed2v,:-'vast inform this vffice pr(x..ptly.

Sincerel~y, 0 Rei i IlX A'i. imi s t r.\r i: .-!c ntnsh, Plant :'.ndg ," ( -*J. T. fluore. PrJL~ct ftIluagtr

.,