ML20249A641

Responds to NRC Re Violations Noted in Insp Rept 50-002/98-202.Corrective Actions:Nrc Issued Amend 44 on 980416 Which Puts Licensee in Full Compliance

ML20249A641
Person / Time
Site:	University of Michigan
Issue date:	06/13/1998
From:	Fleming R MICHIGAN, UNIV. OF, ANN ARBOR, MI
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
50-002-98-202, 50-2-98-202, NUDOCS 9806180034
Download: ML20249A641 (2)

Text

{{#Wiki_filter:_ _ _ _..____. _ _ _ - _ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ eJ M a l l MICHIGAN MEMORIAL -PHOENIX PROJECT PHOENIX MEMORIAL LABORATORY FORD NUCLEAR REACTOR g.g ANN ARBOR, MICHIGAN 48109-2100 own, at e, puu ww June 13,1998 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington D.C. 20555 Docket 50-2 License R-28 Re: REPLY TO A NOTICE OF VIOLATION dated May 13,1998. Gentlemen: The Notice of Violation refers to two distinct instances in which the requirements of our License were not followed:

1) The first violation involved failure to adequately perform a required 10 CFR 50.59 evaluation of Modification Request No.120 which installed a new primary cooling pump and motor and removed the pump discharge

[~ check valve internals in April 1996. l /

2) The second violation involved failure to notify the NRC in writing within 30 days as required by Technical Specifications 6.6.2.b.2 following the discovery of the condition described previously on October 8,1996.

/ /'g o Reasons for the Violations The reasons for the violations have been provided as follows:

1) Inspection Report No. 50-002/98202(DRPM) by Tom Burdick.

2) Presentation to the Enforcement Conference of April 22,1998, with attachments by Ronald F. Fleming.

9906180034 990613 9 PDR ADOCK 05000002 PDR Phone:(734) 7644213 Fam: (734) 763-7863 www.umich.edu/~mmpp/

t Cor'rective Steps Taken to Avoid Further Violations The Arrective actions taken fall into two categories, those involving technical aspects of our license and those involving the adequacy of our communication with the NRC. The technical issues with the license have been resolved by License Amendment No. 44, submitted on April 2,1998, and approved by the NRC on April 16,1998. These issues are discussed in detail in my Presentation to the Enforcement Conference of April 22,1998. The communication issues were also addressed at the April 22,1998 meeting and included dual approval lines as well as additional specific review by our Safety Review Committee. Perhaps the best indication that these Corrective Steps have borne fruit is our

• Report of Non-routine Occurrence - Hot Demineralized Leak Causes 75 Gallon Loss of Pool Water to Radioactive Liquid Retention Tank System" dated June 3, 1998. I have included a copy of this Report with this Reply.

Date When Full Compliance Will Be Achieved We believe that we were in full compliance as of the approval date of Amendment 44, April 16,1998. We will continue to improve our license in terms of Safety Limits for Operation, and we will continue to improve out,Jrocess of review and communication. However, we are now in full compliance with the issues leading to these Violations. Sincerely, ) N-Ronald F. Fleming Director

he Lmversity if Michigan

.icnigan Memorial Phoenix Project

hice

:..e Director

201 Ben: steel Boulevarc A.nn Aroor..'.lienigan 48109-200 June 3.1998 Docxet 50-2 License R-23 United States Nuclear Reculatorv C:mm:ssion Document Control Desx Washington. D.C. 20555 Re:

Report of Non routine Occurrence. Hot Demineralized Leak Causes 75 Gallon Loss of Pool Water to Radioactive Liquid Retention Tank System. Summarv of Events On May 10.1998 at 0:50 reactor operators at the University of Michigan's Ford Nuclear Reactor discovered a leak in one of the two resin columns comprising the "B" Hot Demmeralizer (Hot DD system. Tne leak resuited in the loss of approximately 75 gallons of reactor pool water to tne factiity s radioactive liquid waste retention tank system (total textor pool water voiume is approximately 48.000 gallonsi. Hot DI "B" was isolated and

Mw rextor operauon conunuca. Safe overanon of the reactor was not (Mr.w.eed at any time during this occurrence. No radioacuvity was released from the facility as a result of this occurrence. This report is bemg mace cecause the leak resulted from a failure in a component of the pnmary coolant boundarv.

The leak was discovered while the reactor eperators were investigating the case of an unusually large ooserved decrease m pooi water level amounung to one-half O/2) inch dunng the pnor two nours. The normal pooi water loss rate is about one and one-half i1-1/2)incnes per cay. An automauc shutorf rais e tnat was installed with the most recent Hot DI system upgrade was not m service cecause of past erranc cenavior in its tiow sensor. Hot DISystem The FNR Hot DI system consists to two sets of fiberglass reinforced resin tanks containing mixed canon and amon ion excnange resin ocads. The purpose of the Hot DI system is to mamtam the pH and concucuvity tne pnmary coolant water. The tanks are physically located in the basement of the reactor t'uilding. The Hot DI system taps off of the pnmary coolant retum piping civemng a smail fracuan t 14 gallons / minute) of pnmary coolant water to the Hot DI's. Nommal pnmary coolant : low rate is 1050 gallons / minute. Valves are installed that allow isciation of the Hot Di system from the pnmary coolant system. A check valve crevents eacx : low from ine reoi throuch the Hot Di rerum line. Most of the Hot DI syste'm has eeen switenec to Seneauie 50 chionnated polyvinyl chloride (CPVC) pipe fittings and valves. See tne two enciesed tigures. ciW5%7TCt28 &fP

, Dunng nonnai operanon ene set et DI coiumns will be put on-line wnen concucuvity of the

ooi water reacnes 2.5 w.no/cm. Arter tne conductivity has been brought down to 1.5

'amnoicm. tne DI's a tii ~r e cypasse:: unui :nauctivity a' gam reaches 2.! umno/cm. Then tne on-line set ::::.umns :mer 3. ~ B"' is no ioneer effecuve tne otner set is ~

ut

nto semce. 2.e cut-et-sem:e set ci ::.umns are restorea after a suitaoie cecay

enoa ov reunne tne utstream cc:umn.. v:ng tne cownstream column to tne uostream position.' ana replacmg tne.:ownstream cc:umn with a tanx containmg fresn ton exchange resms.

Safety Anairsis Testing conducted on Stay 13 dunng tne snutcown mamtenance period immediately following the event showea tnat tne maximum onmar; coolant loss rate throuen a cuillotine breax in the Hot DI piping is 15.9 gallons /mmtite. This loss rate would cause'a decrease in

he water height above tne core at tne rate ci 6.4 incnes per hour (one inch of pool height is eaual to 150 gallons of wateri.

The water would flow into pits and floor crams in the reactor basement that drain to a pair cisumps. Sump pumps transfer tne water to tne 3000 gallon radioactive liquid retention tanks in the Phoerux 51emonal Lacoratorv rutiding. 29 gallons is pumped each time the sump pumps actuate. An alarm also annunc:ates in the reactor control room when the reactor control circtuts are energizea ano e:tner sump pump actuates. The normal frequency at which the sumps pump out ano alarm is acout once per eight hour shift. Reactor Operatmg A worst case leak in tne Hot DI system wouid cause repeated sump alarms and would quickly be invesugated. ciscovereo and isoiatea by the reactor operators when the reactor is coeraung. This operator accon would take p:xe well before the low pool level alarm at nunus five inches or tne automanc reactor r:ndown at mmus 12 inches would occur. (Note: the mmus 12 inch level corresoonas to tne reol water level LSSS of 19 feet above the core. one foot aoove me safety iinut of 18 feet.) ' Reactor Secured Two poei level alarms are acuve wnen tne reactor is secured. A local alarm is actuated at a rooi level of about minus seven menes. The second alann, actuated by the same sensor inat gives the pool levei reactor rundown signal at minus 12 inches. is 'sent direcdy to the alarm panel at the Of Depanment of Public Safety which is statfed 24 hours per day. The worst case Hot DI leak situanon u nen tne reactor is secured would occur if the local alarm actuates just after tne secunty guara nas passed through on his rounds. In this situation, knowledge that a proolem exists will probably not be known unul the minus 12 inch alarm actuates. Given a one nour response time between Public Safety contxting the On-Call Supemsor iOCS) and the OCS isciatme the Hot Di system. an additional 6.4 nenes wouid be iost frem tne cool. The pect iev'el wouid still be at least IS feet 5 inches

cove tne core wnen tne

e:K is isolatec.

l J l L__________.

Raciation ievels at me sud=e ci me pool tould not be sigm6cantly aitered by a drop in pooi levei from 20 feet acove me cere to 15 feet 5 inenes. The intensity oidirect gamma ediation from the core would increase ey a factor of approximately 60 at the average f:ssion gamma enerzy oi 00 keE However the contnbution direct core gammas make to ne total dose rate at me coot sur:re :s msienincant as comoared to tne cose cue to nuvateo imounnes i:nmaruv 'U _4. in me rooi water. Die reactor nas ocerated with the rooi level lo'wereo chs mucn 5 :.ve :nenes'uncer special c:rcumstances.'No ciscemible I ~ ncrease m surface =atauon. '.e: t as cetec:ec at tne ume. Normal pooi surface radiation .:veis are 10 to 15 mRihr. Reactor pool water ieaxmg from tne Hot DI system would be captured by the facility's radioacuve licuid waste system. 2ere are no direct oatnways between the basement of the TNR building', where tne' dor Drs are located. and tinrestn'eted areas. Therefore. no cdioacuvity would be releasea to unrestnctea areas. De conclusion is tnat a leak from tne Hot DI system will have no reactor safety signi5Cance. ?ossible Root Cause In the course of tesung the Hot DI system following this event, it was observed that a pamal vacuum could be drawn en tne resm columns depending on how the system's valves were operated. This causea an inward flexmg of the resm tank. Such flexing may have initiated the crack in the failed resm column. lmmealate Actions Taicen Immediate xtions taken were: D :o isolate at,d secure Hot DI "B", and 2) to isolate and secure the entire Hot DI system cunng penoas when the reactor was not operating until d:rtner testing could be completec. '.ane Term Acnons to Prevent a kcurrence The vendor who supplied the resm columns was contacted. The burst pressure was quoted

o oe 250 psi. The maximum recommenced working pressure was said to be 100 psi. A

'00 psi pressure test will be performea on each DI resm column before it is put into servlCe. A number of tests were performea on tne entire Hot DI system on May 18. 26 and June 2. l ~he entire Hot DI system was isolated anc men pressurized and hydrostaucally leak tested

t 85 psi for more inan one hour en Slay is. No leaks were found. The system was hydrostatically tested a secono ame on June 2. One 5tting containing a conductivity cell

=cked at a pressure cetween 100 and 120 esi. The fitting was replaced by a CPVC pipe n:pple and the system was testea again at 1:0 psi. No further leaks were detected. The

.aximum operatmg pressure m me Hot DI system occurs in the piping immediately after l ne return pump. The Hot DI retum pump is normally operated with a back pressure of 50 - l

0 psi to prevent cr. vim: ton. Alax: mum pressure will occur when the pumps are running 1

l 1 L.

emme_. ma :ne Hot DI re:um :soianon.aise :s snut. Under tnis conoition a pressure of 80 psi was ecserved on May. 3. i ne ;Lne..eax test snowed the Hot DI system piping is capaole of withstanding the nignest anucipatec coeraung pressure without failure. A new auto-snutorf vaive =a now +ensor using a different cesien have oeen crocured and ~ nsta21ed. but are not yet wirec elec:..::uiy. This will maxe tne Hot DI system self-isolating

n :se at a major ::2K.s nen :ne :ns:ananon is complete.

Operaune Procecure No. ;;;. :!ec Oreranne the Hot Demineralized Svstem has been ~ m'ooined to secure tne iniet anc outlet rumos enor to enancing valve line-uns in the system. This should eliminate unneceisaritlexing of the resiin tanks. ~ ~

Sigerely, f(' A f.

== Ronald F. Flenung, Director ~ Michigan Memonai-Phoemx Project ene.: as stated

c:

Theodore Michaels. USNRC Proiect Manager Thomas Burdick. USNRC Region III ~

LTER e 3YPass

{ liTCl -

LTERS POOL CHECK 200L CONOUCTMTY

. BYPASS VALVE

!LL 1.

l i -p = _._. 1 i .NLET i 4 - l ~! AUTO SHUTOFF i VALVE i i B B ' lNLET ! . PUMP A i i I i i ) I r I _d* m

1. A" ana "B"

, y g CONDUCEVITY CHECK VALVE ~ DI _0 CAL ~ 0" Low CCNDUCTMTY 1 ~ N OLD ~ ~. mROWW

RESSURE

~NDICATOR VALVE p;p A A 0-100# i i 9 RETURN A I TO - -\\ l DOOL \\v y

g OUTLET PUMP 01 TANK

!LL d.

HOT DJ e ~~ P]P!NG $'f-- s.RRANGEMENT h

...em r .e w.ver.uvua;.wr u s wa

.gure...

4ot Jemineraliser System m

m e RX m

N I

^ 1 LJ 1 g g ,a n u.a. 3 r a L w r J k 4 k HK Ys Al 5* *" Y -%n s..... il (J (T l /7 i kJ AJ f LA pr g rT r_ T L Q,,2 w p r,- ?d.;m*an '" ~ ' dik db h ah T*.* m., 3m e

• nn nn nn 3,rg
• nk 5P
• 5?n

.o. =L

• PA JPn

'#;2 J{.A{2 { -nOrna 12 2 'An wa um e enn nn nn 9:2 s =es. -.t r e -3

.Orzmat issue:. 0997 Revision i: :50398 OPERATING PROCEDL*RE NO. 211 U!C %R OEACTOR LtBORATORY

Nuclear Reactor

P.xna.\\!e=cnal Lacoratorv 2e il=versary of.\\licmgan Ann Aroor. Miciugan HII2 OPERATING TIE HOT DE.\\IINERM.I7FR SYSTEM (HOT DD l.

?URPOSE The purpose of this procedure is to provice a means to ensure the punty of the pnmary

ooiant Dy removal ofimpunnes in tne form of nunerais, dissolved solids. and conosion prooucts.

3. EOLUPMENTCOVERFD The hot DI pumps. cemmeralizer columns. tilters. as weil as valves and pipeg required to

rculate pool water tmougn tne not DI system.

4 RESPONSIBILITIES 2.1 Shift Suoemsor The Shift Suoervisor will be a licensed senior reactor operator. S/he will supervise

.nd authonze all acuvines relatec to the performance of this proceaure.

A; Shift Ooerator Se Shift Ocerator wiii be a licensed senior rextor operator. a licensed reactor ~ coerator, cr a person cualitied for on-shift dunes. Tlie Shift Operator will assist in me performance of tms crocecure uncer the supervision of the Shift Supervisor. 5. PROCEDL*RE 5.1 Censole ocerator review Oceraung Procedure 211. 5.2 Normal cooi water concucuvity ts mamtained between 1.5 umho and 2.5 4mho. Denuneralizer column "A" ("B'"i is piaced on line when conoucuviry is 22.5 mho. When the concucuvity ei the pooi water is 51.5 pmho, the column that is on line is bypassed. f.; ? LACING A HOT DEMINERAI 77CR ON LINE. f.:.1 When pooi water condue::vity is 22.5 pmho. checic the Hot DI L~tilizatior.

nput recera. ;cstec in== controi room. to determine which column is to be piacea on ime.

evi::n. 30398 O

..; - r -- M:: :::::5 f..3

he:s cpen me :sotauen.aives for tne "A * '"B"i DI coiumns.

f.2.4. pen me vaive :aceiec "A ' "B"1 DI on me not DI mamfold. l ^ f.2.5 ihut me.alve =e:ec "3. ass :n me not DI mamiold. m f.:.5 z: ca Het CI;um:5. f..7 2e x me pressure cut :: ne outlet pump, adjust tne clue oudet valve as eecca to mamtmn a 50-c0 psi back pressure. f.2.3 ' form me Censoie 0;emtor mat "A" "B") DI is on line. f.2.9 Conscie Orerator. upc:.te me Hot DI Utilizaucn Input recort. r.nd record i: ms enange m reactor.cg coor. 5.3 3YPASSING A HOTDEMDGpAU7FR. 3.3.1 When cooi water concucuvity is $1.5 umho, Check the Hot DI Unl 7arion l

nput recora. posted in me co'ntrol room. to determme which coiumn is on

.:ne. 5.3.2 T;m off Hot DI pumps. l f.3.3 Cren tne valve :abeled "3vpass cn the hot DI mamfold. l 5.3.4 Shut me vaive iabelec "A * "B") DI on the hot DI manifold. l \\ l f.3.5 Tum on Hot DI pumps. l f.3.6 Check de pressure out c:ie outlet pump, adjust the blue outiet valve as l .eecca to mamtam a f 0-ed psi back pressure. 5.3.7 '-iorm tne Censoie 0:emtor that "A" ("B") DI is bvpassea. l 5.3.S Consoie Operator. upcate tne Hot DI Utilization Input report. and record l status change :n reacter:og book. en

I. .w j J l 2- ? <CO "W '/'7m ,;o /.M '[/C : n ov

/.l = W-

~ A~ A^~ ewIp-sEVIEW ANC 'URNOVNN~ a_ ), ~. /[-{.,G

I

_ yat..;'~ $ Y. s.H ~ 2 Oate l i? _ -gm__ rna g vii L, d.- c -- A 4 . N. W ' (; _w ~ ,30PV. 4FR _ z. _ j'. j'[ i.__ ... W. i"** /_4, .,a......n, u.. ~ m ~j; 4_ ,yj) :(f. g, y " * ' ~ " " j ^ ~ ff."jW ~.morary CperannnQ 'nMM (- -y guyxn 7x.m enaiiawomem M [y,^ Q; 3 > 7- - .a sca ang p demets 5vi MN <v ,p;.., y _:r: mo;_ y.._ t,yf Ma - N f*- y & L/l W -_ O ~' f"- yoo 49_da Aso ), '~ , *fy fgw hel5 O'- ., ' ~ -.- y, .s h ra~ t E"5 b,a bs/i CK-

• ~

g,99 pygg fp f 3 & _i N K "~ '~ 2300 bu 0 05- /dre.) 'w:.. 7: n'~ ?CA ~ ^MMW WL v " ~~ ' '~ ~ ewIpr einfi9N AND TUPNOVFA "~ 4 l W. 9_ e 4,, a "*M.. gy .5 % / f ?vrt*- NbK i d Vf. Cr in; SM pe:mirq #dR v-w . v- /w. e E/2 3.

V.

CS 'UW. OPE. o ( )

• f _

'f . :Occx intr H M/ O

• t-. v_

f* y3 tuns no. ^2*' St ,/ ) q b, A" A i $$ } ~ ^^., - gj i,s.,s m,,,,r. 7y,~ -. :...- m. -* ~ ~ c s,,. _ ~.. _,, ,)) 4-x1 w

ae n. n c m.....-c w a s zu

'e.c*cr.tm C. . W = = W ' N-A*- ,,7 .. o. p_ (p p) n.'cr, io u -,S . g....;..__ { g. ~ ' h - C/ ..:.-~~' ,.n v h 4' e-=*=--

• ~~d

';Q c?% - M- . u.( f. [ y u, p 9 & y,.....

.. ffp ~'J ^ ^^ ^l _.&s- /* a ~ r J

• >~ M +et A li / 4 f

/[/kfD I. d%) ' ' Dd]_ GM 2Ak _=

m da2&d-PtL.)vf J A M -,,t. & sadsldsksb kN RIE

' hou uk / 3 al cJn LAnsaib EE kg>~Ti

1. 1 3 & J.adhkl Cu'i

'r & n. p.) c'.75', - mS [ C'^2 G & f W d ?N S k 16 m /2 + Q i Y" ^ ?*d W '.sw r! If % 1.. W n.< y ( ) Y .. " sfp f bw \\ Ohi j&

, I-

/* i 1 Y &,, = r

=

lt t' V ~~~~~ A ? l?) i, U ere d / w a/ C6 r-- sw ~- - z-s - -~ ,p f., f ~ r) ! 8" Sh ft o S ?JPV. CIS. A .,m m .: 5 ~ 6, ODersennelCEsomalet cui .a g, 4 4 .( 4 ~ ) 9 Joaclef and Co 008entaleSeteouse I! 4 1, d (CSM 69% riu-Mi J tt e.; c ./ N ilCC :.,4 g - ISys f en s,wnd 1 ' '/JR ) (( s' A

Di

~t.

• r.*

1 Ia es og.. ;.,1, i > C hw 9/ b 9.b. ';54 inG ~,:=2 w na icrailY 0 % /*r %) .Y l, $.l a CC0 aa w -,. .;.e..,,,,,. p,..,, (e g) .}}