ML18087A787

Submits Updated Info Re Investigation of Reactor Trip Switchgear Malfunctions.Technical Bulletin Recommending Independent Testing of Undervoltage & Shunt Trip Attachments for Manual Reactor Trip Expected by 830325

ML18087A787
Person / Time
Site:	Salem
Issue date:	03/22/1983
From:	Rahe E WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
NS-EPR-2737, NUDOCS 8303290144
Download: ML18087A787 (18)

Text

• .a*,.,., _

Westinghouse Watar Reat1or Got~

P~ ~~1$)'i'tae 15'1~

ElettTit <:on:ittratioo 01Yi$ions

. NS-EPR-Z7J7 March 22" 19&3 Mr. tt. Denton, Of~tor Office of Uuclear React.o~ Reguletion U.S. Nuc1ecr P~gu1atory ~ission Phf11ips 6uildiJ't9 7920 ~rfolk Avemie Bethesda, MD 20014 The purpose of this let"ter"{s to p'l"'ovide you t:rith the 1~test ittfor1r1ation on the Westingh\>use investigation into the ma1furreticn$ of the.Sal~m P1ant rea~tor trip Sli{itengear. Our invest.iiaticr., along wit.ti analyses.

perfo~d for the Sale-w: J>lantt demonstrates trnxt the W-estinghous~ plants W'ith trds. e'tUipment cari continue to o~r-&te ~thout und~ dsk* to public

~a1 th and safety.

Te~t and lnsp;e-s;tion Results On Maren 20-21, 1983. Westinghouse perforiPed a detailed, proeedurally contrci1ed inspection of th(! under-vo1tage trip (UV) attachment that was provided to Westi ngilouse and was represented by ?SE&G as the UV attachment that malfunctioned on Reacwr Trip Break.er 8 at S&1eJ:t Unit l on February 25, 1983. Photographs arn:f an a.udi o tape recording of the Karch 20-21 inspection are av~nilb1e at \iestinghousa for your review *. A

~tailed written 1r:$pecti on report is be1 ng developed from this eva1uat1on. To our knowledge, thi~ is the only such detailed inspection

• • .conduct~a to date of a UV attachment repre~nted ~s o.ne of tile two that

• it.al functioned on Februar.t 26. 1983, at Salem lktit 1 ~ . In prepgration fl)r this in$p.....ACtion, Westinghouse devaloped a list of postulat~d malfunction SCE:nari os for this device ($ee Table 1) *. This i n$;>ec;tion was conducted 1n order to establish ~hich of these potential scenario~ might have .,

produced the malfunct.ion of the Salem Unit 1 devices.. The fo11crwing is a s~ary of the key findings in this e~am1nation.

1. AS received. fros:: PSUG, the UV cieYice would not latch. Tr.ere was.

also a history of breaker closure problems on Reactor Trip Breaker B as reported in th& NRC Task Force Report NURE.G 0977. As a result,.

the test circuit breaker on which this device was installed could not be clo$ed either electrically or manually. This inability to

'~80329*0144 030322 PDR ADOCK 05000272 P PDR

latch the lN device. . s traced to a bent and defo. phosphor~

bronze leaf spring 'K'hich would not maintain the proper force Bgainst the traveling 1atch mechanism. Tru: defonriation observed on tt~ leaf sprl ng could not have been caused by nornoal operation or -wear of this device. Had this UV device been installed on Salem Unit l Reactor Trip Breaker B in the condition.as re<:eived by Westinghouse.

it lfrould have been difficult to close the associated circuit breaker. It shou1 d be noted that. h&d this particr.t\ar leaf sprl ng been excessively deformed against the traveling 1atch 1 the UV ~vice*

could have been prevented from unlatching automatically thus preventing the breaker from opening. A Westinghouse representative sent to the ~1em site on February 27, 1983.*noted that a 1e~f sprl rig sitas defonHed on at least one IJ¥ device sho~ to him at the site by PS£&G personnel. This device was described to him by PSE&G as one of the UY attachments that ma1 functi oOQd on February 25, 1983 1 at 5a1em ~it L

• 2. The exam1 nation at ~esti nghouse of the UV device disclosed. a missing loc~ washer on the drop-out voltage edjustment screw mechanism .. The adjustment screw was excessively "turned-inN. a condition which reduces the unlatching force available when the UY device is.  : .~

~energized.

3. ln the as-re~eived condition, a visual examinat1cn parfcrmed by Westinghouse revealed that the device was lubricated. PSE&G r~s adv1 sed i\'i!stinghouse that a 11.Jl)ri cant was added to thj s tie\'ice after the event of Februal)' 25, 1983. Westi ng00use is currently analyzing . .:,;

this lrJbMcant in order to determine 1ts type.  : .

4. Wear on the latch and i.,tch interface was not excessive and there was no ev1dence of burrs. Ho\ofever. excessi't'i: frictional_ force cannot be ruled out as a po~nti al ma 1function scenario si nee post intident handling {manualiy exercising the device and lubrication) prior to ~ceipt by Westingtwuse could have masked a friction force malfunction scenar1c. *

5. No v1s1ble evidence was found of *.corrosion or bro~en or missing partsi other than tne previously ~nt1oned lock washer. ~re were no obviou$ signs of 1mpro~r manufacture.. A check of each part against specified dimensions and tolerances 1s being ~de.

Funct1 ona i tests. ~-oonstrate\1 that the dev1 ce was cap ab le of perfom1ng 1ts electri'.cal function.

6. P.rt.ificia1 restraint of the UV dvice res~t era was required to measure the tr1p lei.'tH" forces generated by the uv device. Toe t~st breaker was trfpped nonaal ly by the UV attachment on several att...~ts witll no~1 tr1p t>ar load Qf 1.5 poi.mas and lrtth an increa.sea 1oad to Z.. 3 pounds. A. further fr.crease in trip bar load to 3.3 pQUnds res.ui"red in e~at1c breaker t.ripping by 'L"ie UV device. Tt.e maximum ex~teeC1f1cat1ons for the Westinghouse-supp1 ied 06-50 reactor tnp switchge~r.

west1nghou~ analys1s of the Salem events, transmitted tQ you on March

14. lS83 and 1ncluded as Attachment 1, cc:mtluded that the pub1fc health and safecy ~Ul d not trnvt: been affecteoth the undervoltage and shunt trip attacmrents of the DS-50 reactor trip switchaear. West1ngnouse has conducted. recent tests on a shunt *trip attachment and has detemi ned the dev1ce generates a force at least 30()'1; greater than tc"le force necessary to tr1p the breaker._ In response w NRC IE Circular 81-12, Westinghouse 1s prepaMng a Technical Bullet1n giving recDi~ndat1ons for independent-test'f ng of the unaervo1 ta_ge and shunt trip attachments for manuctl reactor tMps Th1s Technical 6u1letin will be issued to all Westinghouse plants, includ1n9 5al6i, by Maren 25, 1983.

UY Tr;f e Attac~nt Design:

At the COimt1$s1oraersJ sneeting of March 15~ 1983 the NRC Staff expressed uncertainty over the UY tM p ~ttacnment design 1i feti~ and tne inherent margin bet~n the trip force generated by the undarvoltage trip

<<ttachment and the force required to lift the brea~er trip l>ar. We believe thes.e uncertainties have b-een ~solved by further information.

Tests conducted by PSE&G~ FrankHn Research CenteY-, and westi nghouse fndicau a normal tr1p farce marg1n of 100-2'.,>Q percent. Also, in 19i2

~n undenolta.ge trip attacnment, mod1f1ed as a result Of 3 reported UV malfunctions at Robinson Unite, was*_successfuliY tested for more than 8000 operations without malfunction. fobd1fied undervolta.ge trip attg<:~nts were subseQt.tentlY sent es replacements to-an operating plants "ith OS-50 reactor trip ootchgear at that time: All $Ubsequent

~.estinghous-e ~nufacture of undervoltage trtp attach!ients has incorporated the modifications m~de in 1972 ..

A revi-el'( of cv4il*ble LERs on West1 ngh<>u5e UV tr1 p attachment

nalfunction:s si nee the l97Z rnodif1caton ind'fcate that a.pprox1matel,y tifO-thirds of the malfunctions appear ~o be mGintenance re1ated.

The o~-erall dat1t for mt1fum:tions ~r demand on Westinghouse DB-50 circuit breeke~ has ~n applied in plant ?RA studies and has not shown iln um.1ue cor.trlbutic,n ti.) total publi r;__r'( sk.

In view of ttre above, the i;;urrent design of the DB-SO reactor trip s\!!itchgecr is ~ pn:'lper ~pplication for nuclear power plant protect1on systems when properly installed, tested. and mainta1ned. _

4 4037Q.

Nobtithstanding the foregoing, rey letter to you of March 1, 1983 c0i-::nitted to a t:ru>rough evaluation and testing program of t~ UV trip attac~nt to be completed by June 2, 1983. Test objectives and pro9r~

definition "11"e expected to be c~pleted by March 25, 1983. Furthermo~.

~intend to.do an ;n-depth analysis of OB-50 react.or trip switchgear ma1fvnctions which have {;(;curred on operating nuc1e4r plants to provide additional date to further demonstrate its reliability *. This revig.w will be based on historical LER d*ta and any definitiv~ data which \!te can obtain from our operating plant- customers reiative to total nu:nber of reactor tn p marui' and ma1 functi Qn:& rlH:ordad. We wi 11 advi ~ you of tne eGtfmat&d completion date of this review.

Please contact ine if you wou1d like to discuss this matter further.

Very truly yours.,

~41~~.

E. P. Rahe, ~U~r Nuclear Safety Department Attac~nt(s)

Table l Postulated ~~1function Scenario~

l. Corrosion

z. Missing Parts

3. Broten Parts

4. Ele-ct.r1ca1 Failure*

a.. O'..rt of TQ1er~tKe Part~

b.. Mi sasse.-~ly

6. Insufficient iMp Force

i. frict1ona1 Are& Ar~li~s

a. ~'r

b. Burrs, ~fl9-aps c .. LubricatioR

8. Di rt/Cont.ami nati ort

.. S. Mi sadjust..~nt 1o.. Bent or- Def o~d P~rt$

6 4Q37Q"

I

~TTACHMENT

"/*

• I .* I Anc:1y5es of the Salem >>uc1ear P1ant for Postulat~d Feed~ater Malfunction ~fthout Aut(l)>>ati c Reactor Tri~

WESTINGHOUSE ELECTRIC CORPORATIOH

./

J .--

.3999Q

~

~--~

SCOPE ~2:.-

ln light of the recent failures of the reactor trip br-ea~er~ to ,)f]

automatica11y function.at the Salem pi~nt, thE purpose of th~s study is ~/

to real i ~ti~ally predict the consf:!quenc@~ of a failure t.o t.ri? for .

1imiting plant transients whi1e the p1ant is at full reactor p~wer. The trans1ents analyled. specifically for the Sa1em p1~nt~ ~r£ ~ pari;:.i~l loss of steam genet"ator main feeawater f1o~"due to the trfp of 3 single main feedwater pump and also a complete loss of m?in tee.dwate-r flow due to the loss of both main feedwater pumps. The 1atter~ 1ess pr~bsb1e, event is that pres~nted in the Salem plant fS~R. As st~ted previously, th~ purpose of this study is to reaiistica77y pr~d1ct tht Y"tspon~@ ~f the plant to these events and~ a$ $Uch, the p~ant sy$tems are assurrred to function normally ~;th the so1e ex~epticn being the cQrrman mode fai1ure of th<! reactor bteak~rs to automatically function as was ~-,,-+i~r*1~nced an rebruary 22 and 25, 1983. It shouid be noted that ~e sµur1ous steam generator level trip generated on 2/25/63 was ~s a ~e~~1t or nprmal expected f eedwater contro 1 sy&trtm di fficu1ties experi enct:!d at low n l\)

power 1evels. lt also should be noted that the los~ of a fe~dwater pum~

on 2/22/83 was due to a nonna1 manuevering of an e1e-=.t.rk.al bu$ while.

configuring the plant in preparation for~ po~r e$~a1ati~n. Both of these events are not normally expec;ted at full power and thus one *snou1d .

~onsider more credib1e ~vents *such a~ a f~edwater heater drapaut rath~r than the more 1imiting and much less frequent feeowater purn9 malfunctions.

The study considers a thirty second.operator re~ponse t1m~ for a manual.

reactor trip following the automatic protection system demand ~ignal, a simulation of the actual response time of the February Z5a 1983 evQnt,

• The study also considers a more con~ervative operator respon$e of five m1nutes in order to d@termine the $en~itivity Qf the plant response to operator a~tion.

OCSC~lP"T ION OF TRANSIEN'i £FF!CTS

~n~ric studies (We.AP 8330 Westinghouse Anticipated 1ransi~nt$ With~ut 7rip Analy~is} of failure to trip events previously s~bmitt~~ to the NRC have identified the limiting full power events to be malf~n~tions affecting st@am generator main feedwater flow. The reduction in main feedwater fiow affects the overall heat removal Ci§p&!bf Hey of the steam generators and, as a re.Sult of the mBmat.ch between th-a prfoUH.Y side he~t generation and the ~~condary side h~at r-emov~t ~rodute$ ~ he~tup Qf the primary system coolant+ If the rea,tor is trip?ed PJ".i;ptly~ the auxiliary feedwater system prov~des sufficient heat removal l;*apability to remove decay heat. fbwever lf feedwater f1ow to the steam generato~s i$ r~duted or tenninated without $ubs~quent reactor trip the $~condary sy~ t~m \ff ll be unable to remove a11 of th~ he& t that i ~ ~2-Mr& t-ed in t~

core. This heat buildup in the prima~t system is a f~nct~Qn of the amount of the feedwater reduction and is indicated by rising Nactor coolant syst~" temperat~re and pressuret and by incr-e4stn9 pres~urfter water level due to t.ie insur-ge of the expanding rea~tor 'oel~t. Wa~r 1eve1 in th~ ~team generators drops as the remoining ;nv~"toijt' 1n th~

steam generators 1s boiled off due to inadequate spp1y ~f fee~ater. .

When the steam generator water level falls to the pof nt whe'f"e the st~~

gene.rater tube hun<;il e i $ uncover.ed nnd pr1mll ry to s.e.co.rid'1 ry heat

• tr~nsfer is reduc~d, re~ctor eool~nt syst~ pr~$sure and temp~~atui"e 2*

3999Q

. tncrease at a greater r~- 7hi:s greater rate of te(:;;µ~~re ~1i-d

~ressure increase is main~ined a~ the pressurize~ fil1~~p1etely ano w~ter is aischarged thr6ugh the pressurizer reiief a~d ~&f~~Y ~alv~s.

P.e.a,ti vi ty feedback~ due to th~ M gh primary system tem;::er4.ture,. rec:tuces cor~ power. As a result the system presure begins to detre~se ~no a steam space is ~9a~n fornr::d in the pressurizer.

1he limiting criteria for the oostulated transient~ is that retctor coolnnt pressure be mainta1ned* sufficient1y below the -pre~~~re corre~ponding to the ft.Stt Code Service Leve1 C (£.mergenc~) ~tr-ess Hmits. for the reac:torcoolant system. the correspondin*a p.~ssure 1s 3200 p~ia.. "'

COtHRlJ_ ROOM lHMCATIONS ANO MIT!GATIN_G ACTIO~S Although the reactor ~s pr~varrted from tr-ipping ~utomatit~11y ~Y t~

c~n tr.tide f.anure of the re"actor trip b~akers, therv are ~art.Y 'cntr-ol room indic~tion$ end aiann$ which ar~ g~n~rat@d durin~ 't.be ~ansient which would serve to alert the operator tna.t the event has t~ken place.

These f ndications in addition to emergen.cy proteoures*f ~f¢h f'"f!~Uir~ th~

verification of a successful rea~tor trip before ali other dCt~ons, would ~upport the mitigation of the consequences of th~ transient.~

for ~ lo$$ of normal feedwater event, in ~dditicn to normal prr;cess contro1 ai~rms {pump trip~ temperature, pressure, 1eve1 and f1~w deviation alanns for both primary ~nd se~ond9ry sy$tems}~ t.n.e,, fo1)owrn~

audible alarms would be gen$rated:

l.. Steam/fe&dwater flow mismatch and low level (e~i;:h ste:Mi__~fferator}

2& OVertemperature O@lta~T turbine rynback

3. CNert~perature Oelta-T r~actor trip demand

• -4. ov*~rpowt!r- Del ta- T turbine r\inback 5~ overpower Oelta-T reactor trip demand

6. High pressurizer pressure reactor trip demand

7. High pressurizer 1~ve1 r~actor trip 6QJ\'i&nd S. Steam generator low-low level reactor trip d~and

9. Low ste~~ pressure s~fety injection {in co~ncide~e with high flow}

10. low reactor coolant loop fiow reactor trip d~nd Tables 1 and'2 ~how the time sequences for these aliH'msp As part of the procedures th!f operator is Nquir~d to ~):;~~ise following any reactor trip demand, the operator is required to first v~rify the successful accomplishment of the reactor trip by observing rpd position indicators. rod bottom 1i9hts? neutron flux~ or rea~~ trip breaker position indications, The following actions are av&ilable to the oper~i:.Qr in the main contro1 room if an unsuccessful l"'t'a,tor trip otc\!rs~

1. ~.anu~l reactor trip (with subsequent autcm~t1c turbine trip)

2. Manu~1 turbine trip

• 3~ Manual turbfne runback (200i/min.)

4. Manual ~afety {njection 5,. Manua1 control rod insertionw 3

3999Q

Outside the obvious benlt of an immediatf reactor tr** th-E trip o~ tur~ine runb~c~ action is the most important, if & re~ctor trjp turb~~

cannot be obtained rMriua1ly,. to termi'nate the stei'!m f1<?" de.1'~!14 from th~

steam gener~tors to preserve steam 9enerator f nventory. St~~m pre$~~r-e dnd h~n'e primary syst~ temperatur@ wi1i be CQntro11ed by means of th~

st~am aLimp controi system, steam generator relief and/or saf~ty valves.

Other means outside the main control room are av~1lab1e:. )

1. Local manuaJ trip of any ~&~tor trip breaket

2. local rn&nual trip of the rod control systf:il'I motol"'99ener~tor $ets

3. Local manual trfp of the turbine TRANSl(NT SIMULATION Ana1y$e~ were performed to sfmu1ate both a partial a<<~ ~omp1et~ 1o~s of m~in feedwater. TI>ese analyses are based upon previous mQaels con~istent with previous subm1ttals to tn~ HRC by Westinghouse on ATIIS (NS-~A-Z182~ T. M. And@~on to Dr. S~ Hanauer> 12/30/79} ~ut ~lso a~

modif1eQ to more accu~aU!-1y fnc)del the Sa1em Plant.

i..* Initial normal full power operati{ln at beginning ~'f core Hf~. "rh\s c~rrespcnds to the current condition of the ~1em ?la~t an~ is a1so

~'"re l imit1ng condition sirn::e the mol'.ierator temp~r-ature c~rfici~nt is at it~ 1.ea~t ne9at~ve vaiue. A value of -8 pcnv'~f,. wicich is va1id fQr 95% of core life, was ~$s~~dt s~th the pres$urli~r relief ~nd safety valves ere ass~"f'd t~

fur.cHon. There ~re two relief and three safEty va1~tes ..

?ressu~izer heaters and spr~y also fun~tion automati~~11y.

3. The automatic: turbine runba'k on either Overtemperatur~ \)T Overpower Dei ta-T si gna1 s is operable.. The r-unback setpoi nt is 3~ below the ,

t~ip setpoin.t. The turbine runbatk operates on a 30 ~econd cy~le.

Turbine load is first reducr?d 5i '.in l. 5 .seconds. Jf l1't th~ end of the 30 seconds the runback signal $till exists. tha load is further reduced another. 5t and so on. The load reduction nas $. miti9at1n9 effect on the trQnsient and help~ redue~ peak prfrn4ry syst~m pressure. ~

4. The rod 'Qntroi system is assumed to be 1n the manual moda consistent 'W~th actual prbctice. Al.itomatic C!ctioo of the rod control syster~ would eause rod in~ertion when primary tempar~tu.r:e f ncreases and would be iess conservat1ve.  ;

~. The ste&~ dump contro1 system is ~vai1.1hle. The capacity ~f the steam dump is soi of nominal steam fiow at full power.

Auxiliary feedwater f1ow {1760 gpm} begins at 10 se~on~s fvJlowing

~ceipt of the low-low st~am generator level signal~ Tn1s ~esponse time is based upon actual test data f'ff.>rn the Salem .Plfint ..

7. OperatQr a~tion is assumed to 1~itiate a successful m~~.:Ua1 trip.

Turbin~ ~rip is inftf atf!d v1~ the reactor trip breaker Qpenir~.

- ~.

a*~ For the c:omplete lo9of feedwater trans,ient, th~. f.ee.ct~~ter puiilps are assumed to coastdown to zerO fiow in five se-;~nd~. for

• the ioss of a single pump. on~ pump is assumed to coastdown to zerv flow ~n five seconds; however 7 the remaining pump has rated f1ow capac1ty of 7r:tb of nominal .fu1 l power feed~ater flo\ii. Therefore,.

the second pump (the Sa1~m Plant has two pumps) will increase its fi ow to 7~ flow. The re!>ponse time for the second pump is 20 seconds.

. I 9r Nomina1 <;:ontrol iand piotaction syst~ setpoints were. assumed ..

TRANS!LNT RESULTS

l. Loss of a Main Feedwater Pu-mp The sequence Qf events for both a SO ~~ond and 300 s?Cond de i ay* o*f

~anu~l reactQr trip ar~ shown in.Table l. The tran~i~nt primar.Y pressure t~1~u1Qtions are s~wn in figure L "rhe low-low -steam .*

gener~t~r level ~etpoint is*r@ached at 99 second$; auxiliary feedwater is automatici!iHy initisted~ Ten seconds lrsUft, *auxiH~ry f~ecfwater begins to be delivered to the steam generators.

• 30 Second Delay For the case where there is only ~ 30 second delay. the.re are no subsequent reactor trip signa1s gener~t~d. There *is no large he~tup of the reactor coolant because th@ $temn generator tube hundle dµe~

not uncover. Thus there is always ad#lquate ~@condary side Mat*

removaly The peak pressu~e of 2Z86 psia which occ~r> a~ JO second~~

is only slightly above the pre$s1,1re ~t whfch the presurizer sprays-.

are actuatt!d. *.

• For this transif!nt. the reactor coolant system int~gr-itY is not challenged.

5 J~.1 nute Oe 1ay For the cas~ where operator action is delayed 300 seconds (5 *.

min~tes), toe rea,tor ~oolant system tttmperature 1ncr-.eas~s~ reaching the Overpower Delta-T setpoint for turbine runback *at 190 second$.**

This signal is mai nt~i ned and thus turtsi ne power continues to redv.ci!

5\ every 30 seconds until th@ turbine load is at 75t. At thi$

pointJ the sum of the main feedwater fiow from one pump plus the auxi1 i acy feed-water flow is e_qual to the turbine steam fl-mil.

TI!erefores steam generator level does not continue**decreas:fng and stQbililesy The op@rator initiat@d ~eactor and turbine trJp at 399 seconds QCcurs ~fter the steam and feedwater flow have matched. The peak primary system pressure of Z330 psia at 267 $econds occurs before the steam and feed flow are matched. Thi$ pl"'ssur-e is below*

the relief valve setpoint {Z350 psia). The pressurizer sprays.

combined with the effect of reduced turoin~ load prevent ams significant overpre;;suri~ati~ny Again. reactor.coolant pressura stays be1ow servica Level c limits of 3200 p$ia. *

. *. 5 3999Q

2. Loss of Ail ~ain Fee.at~r The sequence of events for this t~~nsient are presented in Table 2.

The trans1ent pressure caic~lations are depict.ed in* F~9ur~ 2.

The low-)og' steam ge*nerator level setpoint is re~cnea at 33 $econcls; iO second~ later, auxi1i~ry feedwater is deliverea ta the 5t~~m generators.. .,

30 *~tond De i ~x An automatic turbine runback dua to an Overpo~r De1ta-T is initiated at 43 seconds a~d turbin~ 1oad f s reduced ~. The pressuriier relief valves op~n and maintain pressure at the ~et~ifit valve (23501 until the operator trips the p1 ant at 63 sees:mds.. .

Steam dvmp 1s initiated and ;educes the primary temperatD.re to t~

no .load value of 547°F. For this transient the react-Or ~oolant system pressure is l'iell below 3ZOO psiaF 5 Minute Delay As in the previous case.*the heatyrp of the primary c~nlant caused a turbine runhac~ initiate6 by an Overpower Oe1ta-i si~na1. The turbine load is reduced twice in S~ irn::rement~ until the 1~~d is 9~

of ~in&l ioad~ Steam pr@ssure starts to drop oue to t~a h~il ~ff of water in the steai.T. generators. generat1n9 a io~ ste~ pre$~are r?larn. At 'thi:? time primary pn;$si,:r~ stert~ to inc:r-e?s~ ~~4 t;:...F.!rt is an insr;rge intQ the press.1.lriler-, c:au~ing both pres~ur-i4~-,. high

• 1evel &na pressure trip alarms.to be actu~teo. The $team ~-ene~~t~r tube bundle begin$ tq uncover. '9~sing a l~rger rate of increase in ptim~ry pressu~ and temperature~ Thf. pressurizer- fin~ .a~d th~

peak pressurf! reached ts 3491 p~ia. Huclear p-o~r hos dee~asad at tfiis point to about 3~ of nth~inal due to the negative moderator tem-,.,eratur-e reijctivity feedb~,k. /..$ the relief rate of w~ter through the relief and safety valves ini;reases, the prim~ry iJ$"k~

pressurt starts to decreas~ and the ~afety and relif valves clos~

aoout.3S ~e<:onds eftel"' the time of peak pre~sure.. The operator trip~ the reactor manually at 333 seconds.

CONCLUSIOHS Los~ of a M!lfo f~e<ht~ter Pume The r?s~1ts presented here deµmnstrate that for t~ loss ~f o~ ~ain fe-erl~ater pump~ there are at least six major alarm~ in additf~n to others g£nsrat~d to ~1ett the oper~tor t-0 the f~tt that ~ m~1furn:t;on ha~ occured.. Furt..heraore, even fct the ~vent ~th a five m~nut~ dela,y 1n reactor trip. auwmati;; turbine r1,mb~ck reduces steam f1ow ~- match the capabf iity of the auxiliary f~dwater. fo~ this event ~r-e is n~

threat of rrverpre$surizat'ion in that t~ p-re~su ..i.ter relief v~1it.e setpoint is not ~ven reached.

6

• .. *' ..~* .. '** ~

. Complete Loss. of Main.Fe-ater For* the complet~ 1oss of f~edwater, operator action consistent ~ith the action tim~ ta~en at the pi~nt on the Febr~~ry 25, J983 ev~nt is

~ufficient to pr~v~nt overp~e~suri~ation of the reactor coolant system.

Peak primary system pressure results only ~n pressuri?er reiief valve actuation 'Hithoot the actuation of pressurizer saf~ty valves.

Furth@rmore, ther@ lit@ 3 major ai~m$ which are ii1ctu.ate<;i ,fn ad<£ttion t~

the steam generated 1ow-1ow level alan11 to alert. the operato~ t~ ta~e ili;ti on ..

As di~~US$eQ earlier>> 1t 1s a major r@duction in pr1mbT')' t~ se<;ond~r.f heat t-ransfer capabf Hty which cau~~s the primary system hE-.atup and pressure increase. A turbine trip reduces the amount of ~u~ flow ~nd the rate at which the levtrl in the steam generator drop$. 11 't.h~

turb~n~ i$ trfppe4 before there is a si9nificant loss of st@am g@nerato~

fnventory. the tubes will not uncov@r and the primary system will not overpressurize. Based upon th! results discus~~d in the previ"1J's s@ctfon. operator action to trip tha turbine ! t or be.fore one to one ana a half mfnut@s following the 1ow-low level trip -&nd a1~m wuld prevent overpressurization of the reactor coolant systtm beyond 320(} p$ia.

It shQuld be noted th~t the core nuclear characteristics (& ~oderator reactivity coof'ficient of .....S pemt*F) used ~re l'lOt r-apresentat"i'4'~ of the

~~tu~l core de5ign fo~ th.e Salem Plant. Prev1ous AnlS &nalys~s nave shown the peak pres~ure to be a strong function of the coe:ffiei-E!nt aoo

.thei"e i$*~ 1GO psi reduction for every*l pcm oecr~ase in* the co@ffieient. Th@ Salem co~@ i~ d~$igned tr> operate sui::h th~t. by the time the plant ~ached full po'tie~ it ~ould hav@ a coeffici~nt ~f -lO~S pcmr*r or 2.5 pcm les5 than the c0fficient fn the study .. This co~fficient would be reduced even further by approximately Z p<:m/aF per month of op~ration {s~e Figur@ 3). ih~ l0.5 pcm eoeffic1ent ~s~lts in a peak pressure for the limiting ease of five minute op@rator -~etfon cf 3241 p$ia (JJ 250 psia reduction *from 3491 psia) which i~ within the ca1culational band of the AStiE Stre$$ L~ve1 C 11mit. Therefore~ tne case ~PY'&!i@nted i n Ff gure 2 would not exceed tha acc~pt&ru;e Oi teri ~ ..

Sumnar;y In eone lu~iont thi$ $tudy has demonstr~ted the abl.Hty of the Salem Mlclear p1~nt to wi th$tanli th~ effett$ of postv1~ted grosi fee<Nt.ter malfunctions without r@actor trip at full powar with an artff1ef~1ly long- delay for operator action. The results $how acc~pUb1e re$pOn~~

which i :s within c;a1c:ul iationai uncertainties of the ASt£ Stress level C Hmit:s. Th~se r-esu1ts are furtnet" aff~cted by the low probab1Hty Qf these events occu.ring at full power in addition to the expected incrf}~singly beneficial nuclear characteristies Gf the plant o~~r eore

. life.

7 3999Q

i-~12-

~-

e::;_;

~ r-<":*.: ~

-__..,,, TABLE l '

Sequence of Events LOss of One Feecfweter Pump Event

• ~

loss of one pump {alarm) 0

~emaining pump delivers maxfmurn f1ow 20 20 Lbw-low SG 1~v~l setpoint {alarm);

auxiliary feedwater signal (alann) ~9 99 Auxiliary feedwater b~gins 109 100 Oper.ator trips reactor and turbine i29 ---

OP AT runb~r;:~ $etpoint (alarm)

. turb1 ne load reduced St ........ 190 iurbine load N!'duced si Z33 OPbT trip setpoint (a1am) ---

220

?50 Turbine 1oad r~duced 51 Peak Pressure Occuri 267 {~~SQ psia)

Turbine load reduced St

• 2SO Turbine load reduced 5~ 310 High. pressurizer level setpoint. (alarm) ...... 311 Operator trips rea~tor and turbine ..... 399 3 alarms - 6 alarm~

prior to prior t~

trip trip (1) so $econd delay before manual trip (2) 300 second delay before. manual trip 3999Q _.. _

. r

-~ TABLE 2 Sequence of Events Complete Loss of Main Feedwater

~vent ..

Loss of main feed.water pw.;ips (alarm) 0 Low-low SG level setpoint (a1~rm);

aux,iliacy feedw;iter $ignal generation 33 3l OP 6 T runback setpoint (alann) turbine load reduced 51 34 34 OP tJ T trip setpoi nt hil arm) 43 43 Auxiliary feedwater begins 43 43 Pressurizer reH~f valves open 55 55 Ope.rator' trips ~actor/turbine 63 Turbine 1oad redµc;:~d m --*

High pressu~zer level trip $etpoint {aiarm} ~-

Low steam pressure SI (alanu)

High pressurizer pressure setpoint (alarm) *--

SG tube.s begin to uncover;

~team f1Qw drops pressur-i.ter safety valves open ........ 92 Pressurizer fill$ .95 Pe~k pressure lJ4 ( 3491 p~ia)

Pressurizer :safety valves close --- . 14-t Pressurizer relief valves close Low RC flow s~tpoint (alann) 155 *.

159 .

Operator trips reactor/tyrbine 3_33:

4 alarms 7.alarms prior to** *p}'\f cir to trip ..trlp' (1) 30 second delay before manual trip (2} 300 second delay before manual trip 9

3999Q

- ...A*~..,:

. -**llW'

. :ne

.; i,.'

"'\ .

Sec:nC!i ..,.e: :.;1 ~~f'Jre

=.. -.......

• --9" ...

~

--"\Al

~soo

..:n.ru

--r.

...... ~--0 ~~' ~~c

~

~:-o t?O .IN

- * !!It ac.

300 Se,ono~ Oei~y ;efor~

,. * :Ill I

I I

--- \

~

\

~~

~

~

R-

\ ~

~

__ *, - .  :.cc ~

~  :~o ~ U}~ ~w I.

=..

• ~me i~C~

I

~

~

~

~':

?ressure*i"*s~errt FT*GUP.E 2

- Loss or All ~ain j .....

,-..i;,..~.....

e l, ..... '

30 Seconds Oe i ay 6efo r-~ Tr~?

~*

~ .li}!j

. l l

'II 1

\$00

....:II

' ~ tSOO f,.o_____.1.o_o_:::=:;:;;:;,..ZIJ:.,.,_Q_ __.;._~~00~-........~!(}Q~--~~~O~D---*-~~QO t 1nt! ( a.i:;: )

- -~-~- .

--~' . *'

j200 I

l, i!~ ~ I. '

~ ~ _,J 2-00 i

~000 r

r r

~00

- ~ *

, -~

e e

/-.'f~

FTGURE 3 MQDERAiOR TEMPERATIJRE CQEFF1CttNT DURING CYCLE 5 AT HFP, ARO, t:QUILI'BRIUM XENON CONOtiIONS"'

.5

\.Im C!

=

u 0...

J..j c: ..:10

- r:u u

u-a.I 0

~

.. 15 CD lO..

s tO i-

\ a1 So s:;.r

"""'t-0

-20

~

o'S q,*

"O

~

~. --

..-30 0 2 4 .6 8 10 .12 Cyc1e Surnup (~~0/MTU) 11

• WC.!\P 10242 2' The N11clear Design of Saiem Unit One Power Pldnt. Cyc:l-e S"