ML20024B597

Transcript of IE Interview of H Hartman on 790522

ML20024B597
Person / Time
Site:	Crane
Issue date:	05/22/1979
From:	Hartman H GENERAL PUBLIC UTILITIES CORP.
To:
References
TASK-01, TASK-04, TASK-06, TASK-08, TASK-1, TASK-10, TASK-4, TASK-6, TASK-8, TASK-GB B&W-1021, NUDOCS 8307090384
Download: ML20024B597 (61)

Text

-

','*A h'

i No Exhibits i

l l

I&E INTERVIEW of HAROLD HARTMAN

{

5/22/79 i

/

l

(

(This sheet was made for 8307090384 790522 I.D. purposes only - not PDR ADOCK 05000 9

part of original)

C5K@M 1902

+o NRC SICv,

e e

n s g

(

.att.,

~~

e m

I'-

gg q

hpy 1*J8 N W /

.t t

w s

9-

..o 3

1 Wi$ !! A W"IK*M **7'.*1:';7 !?.0 *dt :*t;***a*;;*. 'M !*C G :-

'*(

tt.:.,18v!D. rygt.;.g.:st.e z.1,.., 2 prg;s : a.::::q gg -g gg r-

! -:; 43. w..

...f......., Aa 7D...t.as.

.ts 6 A.".D A*dC1 f..FCG*..*?i 1

l I

{

I l

l I

(

I, I(

0 b

4 l

(

0-

s e

(_

wARSH: The data is May 22, 1979. The time is c: If p.m.

This is Ecb Marsa, MARSH, and I'm and investigater wit the 'J.S. Nuclear Regulat:ry C:==f ssion assigned := Regien III, Chicago, Illincis.

This evening we are Icestad in Reca 119, the 2,ed Reef Inn, in Swatara, Pennsylvania.

Sat's S*dATARA, and we art here t: cencuct an intarview cf Mr. Hal V.

Mar =an, who is an u-CR0 fer.w.e - H a: the Three.S.ile Island sita.

A'.

this -f u I'd lika the other indivicuals in :ne rce= := ican-ify them--

selves, := spell their las na:e, and to icentify their ::esitien.

CRET4 ELL: This is James S. Cres-ell, CREI4 ELL.

I'm a react:r inscac:c-icca:ac at Regica III.

(

FASANO:

I am Antaeny N. Fasano, FASANC.

• am an hspec-icn hecialis:

out of Region I.

1 CRETWELL:

I'd like to make a ref t tnca te Hal's fi s: name teing Marc 1d Hart =an.

.uARSM: Thank you. Hal ycu incicztad that you go by Hal even :ncugn ycur name is Harold, right.

HAR?AN:

Rignt.

(

es !s a sica. css *. r. E 20 *! 'T:". 'T*'3 "IIE*T 'I 'I n :,.. :;. 1;.1:U}**

re-e've'r,ti;2reU

(*.

~ ;0:.". :::c::t- ::*:0 I :.... **.!:.. *.: i : 4

• 8 'a

.aTg;g og a;;;2,171 13Gf. J.*T f

J.

(

TMI TAP: 254 2

i MARSH: Hal, befort we turn the ta:e en we had sa hart and discussac this do pagan memo and I just want : make a few ita:s in there a 8

= attar of de taped rec:rd. As I indica:ad de mano coes ::ver tne pur;ose and sc:;e of cur investiga:icn and gces : seme degree int: the rights of the individual being intarviewed.

Cn tr.e last page :r.ere's I

several questions wni:n I just woulc lika :s get your rescense to en s

i the taos and that is (1) do you understan: the a:cve wnich secressas the d o page same?

I HARWAN: Yes.

M.aRSH:

Tne sec:nd question reads, de we have ycur ;er::issica to u:e tais intarview?

MARWAN:

Yes.

MARSH:

And thirdly, it says cc you want a c:py of the tame 7 r

1u HARWAN:

Yes.

I

.ine.

On the tame I will get ycE a ::py of this ta:e ;r:ba:1y MARSH:

l' temorr:w and I will get a c:py of the transcript wnen its availa:1e

sailed cut to ycu so you have that also. There's a fourth questien

vered in the tecy of the text that does ac: :eruin s:ecifically :s' 1

you, but it accresses the incivi:uals rigr.:s.

We can a f:a :y it if I..

' i

_. ~ -

- - ~ ' -

~~ "'

.s 9

t r

TMI TAFE 254 3

~

1 you so want, and eat indicatas :t.a fer.ae Met-Ed em:1cyees in tha if they en cesirs, they c;uld have a unien re resanutive er a Me -Ed 1

representa.ive present.

I think you've i-dicatad Ost ycu did not, bu.

can I get your respense nowh i

HARTHAN:

I waive my right have any...

i

.WARSH:

Fine. Thank you. Hal, to ga: going win, we'd a :reciau 1:

.=

very =uch if you could giye us sc=e words regarding ycur back;rtund, you association witt./:na nuci3tr field, and Lycur ex;eriencas wie Met-Ed.

I'd also lika to inc1cce :te datas taa: ycu wen. is crt for Pet-Ed and the data Mat ycu sacaratad.

!(-

'ADIAN:

I suruc my nuclear career in ce Unitad Sta as Navy.

I went ta basic nuclear power sencol in laineriega, Maryland.

I graduatae

(

in the t:p quartar of my class there.

I wen:.: ' des: Mi1=n 52G pr:5:ype in West Milton, New York.

I was then transferred :: tr.e 'docct:w 'Wils:r..

I served aheard her 2 1/2 years.

I was transferrec :s the U.5.5.

3

~,.

I who was in overhaul in Pearl liar:' r Naval Shi; Yard.

Iwasi o

I separttad fr:a the nevy in Nove=cer '73.

I, starud ncrt win-Met-Ed,'

1 Januarj of 1574 I believe'it was the 252.

I was hired tr.ere as an auxiliarj cperator.

I,want u rzugn six r.catas cf.achnical. raining :n

\\-

ce operations, 'systa=s, and technical training, rc=a reac=r t.ecr/

t l

stout late:ck and Vil=x reec :e 1 ants.

I s en 21/2 years is an 3

auxiliar/, c: era =r and was ;r=ctac :s Ur.it 2 c:nti:1 r:cs :ceraur in t

l

i

(

l' TNI TAPE 254 4

3 September of 1975.

I attanded the, S week c:1c licansing pr:ges: a Lynchnurg, Virginia, the simulater training.

I graduated n ort No. 1

(

out of 5 and in Octscer of 1977 I passed ce requirements for a Neclear Regulatory Cemission operators license.

Fr:m Oc::cer of '77 until April 13 I served in ce ct:acity as a licansad c:ntrol r:es c;erater t

in Unit 2 and I resigned my e micyment as of April 13, 1572, 1979. Anc t

tha*'s aheut it.

I e

CRErdELL: Hal, I wonder if you could go back ts ce time of Mar:n 23, 1975 and tall us briefly when ycu get on shift and what went on as ycu 1

got on shift.

-(

l HAR7wAN: Tnis is genna he amusihg because I was in Lynenburg, Virginia at the one week react:r peration training c: ursa d:wn cert.

Sc I was in Lynchturg on the actning of the 2Sth.

I guess, well I woke up accu:

7:00 that scrning and y shift fortsan came ever and said :nat his girl said Unit 2's down. 7he safety's were blewing f:r a c:emie of hcurs.

I tacught'uat was rather odd sinca cey shculd never blew that leng.

And threugh the c:urse of the day we jus gatered bi.s and piecas ef; infonnation as it went Ly, as they became availacle :s us down thers. '

Mostly from B&*d pe:ple.

In fact, we had a shift superviser, Bernie

'4 Smith, he was thers, with us at the time my su;erviser of =cerations in Unit 2, Jim Floyd was thers, a: the time and I guess latar in the cay i

when they figured things weren': so hectic, they callec un and g:: a y

li--le bit ::re taennical inf rmati:n as s -na na:: ente. Anc =e way l

0

(

TMI TAPE 254 e

I undersued it at that time was that they had a esactor...a.1 css of feedwatar to both staam generat:rs...and for sc:a reason they didn't have auxiliary feed. Now we didn't kr.cw the real reasons wny they cidn't have auxiliary feed Ehen the pu=s failed to start, er we didn:

rsally knew.

So the rest of. the day was then just spent. siculating the accident, trying to see exactly what r.a:pened.

'de had several :aramatars that we knew happened, the pressure excursien in the pri=ary systam. 5 They assumed that they had maybe one tube, had ruptured, it se:arated.

in the tube sheet on the B staam generator, tacause that's where the activity ca=a frem, that they. knew was released at that time.

'de just, tasically then, well, Jim Ficyd and Earnie Smi-h, they left Tnursday in "a aftarncen. They t:ok a plane cut and came back ts ce site and t' '

May left myself and the two cuer coera=rs and the snift fereman dew..

care just ts c:ntinue with the training. And really tne only infer-mation we got back then was eitner f :a the e.aws pa:ers, the televisien, er what we esuld get f :s late:ck and Vile:x anc :ney cien't want =

ad: nit tEs much.

Especially ahcu. fuel damage and any design deficiencias that may have been present or have tacught say have been a cause of the I

accident.

I get tack from Lynchburg en Fricay anc I dica't really fi c un much cut then.

But I did hand in my resignatica that cay, it was' the 31st of March. Cr the 20th of Marta, excuse me, it was a F.-iday.

(

And I went in to work then the next Wednesday and basically the ac:icent They were st.ill in a state =f general emergency ac=rding.:

was over.

the raciatien, emergenet plans. And my cuties.ners wert, I was :nly t

f

.t

' ('

m.e.. s e..r.

e4 s

,t

.'.ers fer :..res :ays fr:m Oe time ce 1::icen: ham:enec ::ntil I resi;.e:.

D.e tr.ree cays ca: I s:en :nere were tasically a ca:a taker. a-1:q katter, and sucn as. hat.

I :i:n' really ge int: = a ::eratien :1

=a piant.

d CRET=?tL: Okay,iial. i&.: was your snift f:re.an ua:.as.i u you u:

hers on the...

MAR 9AN:

It was Dick H:yt.

" RET *' ELL:

Oicx Heyt?

i MAR M N:

Dick Hey; is =.v snift f.: reman,.ves.

• RErn' ELL: Oc ycu rec:llac who the c her CKGs.ere taere?

.4 MARMN:

Y e'ah.

Ray Ecyer anc he's a li:ensac 20, an: 'enn 51essing, he is a traines.

t i

,..-..,: Now you, i,.. rec llect ; :cer.:y,,.earnec cf ne even; l

or.::a: _

tar: ugh ycur shiit i: reman...

(

t

.9R7AN:

Yes, wne had learnec tr.r:ugn a ;nene :all fru.is ;.irif-iene.

(

ORET-?L* :

His giri'riend snert.as s.e :::stac?

(-

i e

f.

p-g r,

,v.---,

n.

.,.,.r

(

(

TdI. TAPE 254 7

HARTMAN: Well, I didn't really ge: that perscnal wita Dick, but :: na best of my rec:llection he was dating a guard.

It.as a Gregg quart.

Cr...

CRESW5LL: Gregg's f acurity?

s

. HARTMAN:

Grtgg Security. And I'm not sure if she was on that evening.

or what.

Her infor=ation may have been see:nc hanc, I c:n't knew.

4 MARSH:

I think that was the pcin as to wnether she jus; lived in the area or...

' ~

RARTMAN:

Yeah.

She lived in M:unt Gretna. ' As test as I can rent =:er she lived in Mount Gretna which is a small mount c:== unity accu 10 miles frem the. sita, I believe.

l

[

MARSH:

But also employed in and ar:unc ne sita, rign:.

(

~

HARTMAN:

Yes.

i

• I ORET45LL: Gkay. New basically at that point in time was, is it a fair i

enaractari:ation that the inf:r:ation was restrictad to the relief, tne safety relief valves blowing for a sucstantial perice of time?

(

(

i

-(

TMI TAPE 254 8

.I HARTMAN: Yeah. At that time when sne saic cat t::e safattes ::ac ble a fer two hours and it raised a cuestien in =y mind cat the safties, they c:uldn't have blevn for two hcurs unless sc=e:hing was really 4

wroiig. The reacu r would have had to stay at power in ceder for those things to blev.

i CRETWELL: Okay.

MARTHAN: So there's no two ways about it.

CREF4?LL: So you...

MARDAN:

So I assumed that Oey ere ut accspneric relief vaivts.

~

Now I, the atses;heric du=n valves, yeah.

I really cien't xn:$< the circ =stancas wnica cpened.

I knew hcw cey cculd cpen, but : really

(

dicn't knew the stecifics en wny ney c;ened en the =ct.ing cf the l

22:n.

(

l GEF4 ELL: Okay. What hacpens aftar, are you at breaxfast eating. hen you found this cut er..

(

HARWAN:

No.

• de just got, we were in the metal recm.

CREFJELL: Oh, ycu -ere in te sctal reca.

~f

(

( f.

,.m.,.,

g 4

y-

3

/

6 TMI TAPE 254 9

3ARTMAN: t!m um.

CREI4 ELL: Okay. So then una: do ycu do? Ycu gc on inn werk...

3 HARWAN: Yeah. We went i'n, we went rign: inu the training cantar Dere.

CREF4 ELL: On Old Forest Road. Does Jim Ficyd try t: get in ::ven win the plant er any.hing at that point in time?

MARWAN:

I'really can't rece= ar.

I knew, I cink he die. But ce infer:atien, yenn, as a mattar of fact he did. He called es and he g-(

get, he was really intarestad in finding out some cf the ;arame:ars, sema of te things, seme cf the events, ce sa:uencas of the event, scme of the readings that they were getting on the.-adiatien :eni.crs',

f-and seme ;finary and sec:ndary chemistry analyses sc that we :suld simulata it, so eat we csuid see if we =uld simula a it, see sc=e cf t e transian u that t:ck placa. They had'saic that, ce pressu.-i ar

(

we'nt solid. They said that ce primary systam ;ressure went v a:cvei the safety limit. At t*:e time we didn't knew tha: the pressuri ar l

relief valve, electr:matic, I guess it was, it was stuck.

I r.:n't even

(

knew at this time, to this data.

I jus: assu=e that it was the elec:r:-

matic relief valve. We didn't knew mat until I think it was Thursday acining we finally simula:ad t at ce valve ac:: ally failed c enee.

i

(

x E.

O I

TMI TAPE 254 10

- 4 FASANO: This is Fasano speaxing. Hal, you cid =entien =at, in :ne c:nversation and the infomation that came :: you en the first day, that you bew that the aux feed was not feeding. 'ders you 01d that 4

ever the phone?

+.AR W Ath Yeah.

eney nac saic that the auxiliary faec, they cidn'. get it inn the generatsrs at the initial ;cin:.

• ne dicn't at that time, we didn't bcw hcw the feed was st:; ped. We didn*: bow whether it was a malfunction of the pu==s, the piping, or valves.

~4e really bow wna:

tne cause was.

FASANO: Do you knew wnat valves they we e talking about?

~

HARWAN: Absclu.aly.

I checked these every ti=e I c:ce en shift.

. (

FASANO: 'nhy?

l l

MARTPAN: They' re.fm;:cr. ant. And 1 is they're always doing su:/ei!-

f-lanca.

They de a acnthly surveillance on each cne cf these ures ;u=fs wnen we're at power.

Ir fact, they have to do a once a scnth in acce a cr above ar.d I knew that these wo valves had :: he shu. in order to cc

(

the surveillance so that they dicn't feed :he.atar thr:ugn the air cperatad regulating valves int: the generators.

(

en

..--r-w-

9

_.y c.--._

1

't

/'

TAI TAPE 254 n

i

(

,CREF4 ELL:

This is Jim C.tswell again.

Have you ever found ecsa valves closed bef:rt?

i.

HARWAN: Yes.

CRE5 DELL: Heir =any titts?

1 HARWAN: Cnca, that I can ruember, f

'CREF4 ELL: '4hy was it, vny were they c!csec?

HARWAN:

I can't recall why t.ity were clcsad.

I askec y fertman, k

when I fcunc mem elesad, I asked =y fernman if this was a genersi precedurt that I felicw, I asked him i: you knew why the E3V 12 valves are shut. He said no, I d:n't.

I said skay, Dick, I'm genna c:en

~'

te.s, and then I c:ened nem anc every-hing lets icosa. O'

-t s

CREF4 ELL: Tne foreman's name is...

.l(

HARWAN: Dick Hoyt.

l

.' g CREF4 ELL: Dick Hoyt.

Do your reen11ect when tais happen 4d?

HARWAN:

Tnert wert so many things that, cert wert so =any nings ::

rte:11ect, I tally can't even put a ti e frame on it.

~ ta:1y ::n'-

t. ow.

-(

(

THI TAPE 254 12

,.(

't CRE!WELL: 'Would it be like weeks befers tte event?

HARTMAN: Yes.

It would have been weeks. Maybe, perna;s menchs.

(

CREEWELL: Okay.

i.

I i

HARTMAN: Several months.

CREFWELL: Okay.

Bu: as far as you knew it, there's not a practica cf

'I leaving these valves closed fer a specific reason?

Ycu knew cf nc rtason?

t '~

HARTMAN: No.

Ncne whatsoever exca;; ina; the surveillanca precadurt s

calls fer them tc te snu: wnen Ee actual tes is ::eing cencuc:ic.

CREFWELL: Do you know if that was rt;cr:ac te :na NRC na tacsa valves wers snut?

l HARTMAN:

Nc, I do nct.

j i

CREFWELL: Os you knew if :ne unit was c;erating at ;cwer when :na:

hac;ened?

(

HARTMAN:

I can't recall tha eitner.

t l

3

(-

TMI TAPE 254 13 a

CREI4 ELL: Okay.

MARWAN: But it dcasn't maka any differenca, in Mcde 4 Oney're required.

CREF4 ELL: Okay. Getting back to te saq::enca of ce event, tha time secuenca of that day, you went en ints the R&W facility on Old Forest :I 1

Read in Lynchburg and you started simulating =a event en the simulatdr and I guess all' of ycu were in the simulater at that...

i HARWAN: Well, it.was, they kind of pusned us aside, really. And every onca in a while if we get a break, we had seme classhes trainir$g g

wnen they were running uis and the e.ree CR0's, in fact, it was the u

urte CRO's, well myself, and tno cther guys' and Dick Hoyt, the fere=an.

We s,.ent most of the ::rning in class wnile they ran the tas s...

CRET4 ELL: When you say they...

HARW AN: Jim F1cyd, Bernie Smita, there was one er do other inst:.:::: s, I knew then latar there was some of de big wneels dcwn thers fr:m 5&$

i cat'we a on the tast sita, I den't knew their names. But I guess they c:nducted tests from if ke 10:00 in the =cening wnen ney got e.e infer-4 mation that cey needed until probably 2:00 in the aftarncen.

CREidELL: Okay. What basically was ne information taa-e.ey were i

f usir.g when uey sta tac at 10:00 in ta.icining?

Cic ycu un c:-n

(

ur:ugn it?

I n

(

?

TMI TAFE 254 1,.

(

HAR7uAN:

Loss of both feed pum:s causec by Icw suction pressure and that caused a high pressure react:r trip and they said tha: 8 =inutas it latar they got emergency feed.

So that's when we simulated turning en the emergency feed pu=ps.

It was a minutas ina the accident. We also simulatad, we also assumed, we didn't get this I can't believe we ge:

q this from any infor=ation that was sen. ::wn :nat morning, tu:

s we knew that the crassuri:er was on c:ntinuous s : ray t: ecuali:n bcr:n er to :

keep bar:n in the pressuri:er and the RC3 acuali:ed.

CRErdELL:

And that was the becausa of the leaking valves en the pres-suri:er?

'r HAR M N: The laaking ressuM :ar ::ce safaties.

CRErdELL: The safety valves.

MARSM:

Excuse me.

Have you nected in affr=a: fen :s

..a stat =ent, M gnt?

(

i l

MARTPAN: Yes, yes.

r

(

FASANO: Hal, you knew tha: the c:ce safeties ere leaking I mean...

HAR M N: Absolutaly. W.ey were leaking, I kn:w fcr a. leas-3 acntah

(

j be,f:re the ac:f: ant.

{.. '

f.

t

-(,

THI TAFE 254 15 t

FASANO: New, ry understanding the ele: r: motive, the electr=a.ic, was the main causa of leakage prict to the event. New tais is a little 1

different so...

MARSH: Well, this is his understanding.

I i

1 FASANO:

I understand...I just wondered,*nere he gets his inf:=a:fon.

f HARTMAN:

I can Icek at the c:=utar. They have an anaicg value of ide ta=eratures at the outlets of :nese valves.

J1CANC: These would be the themec:u:les?

'(

HA RT%'.N: Rignt. Tha ther. :cou:les dcwnstream. The ele

matic relief valve was the icwes ef :na three ar.d it had been for 3 men =s.

4 The otner do would kind of weep u: and c:r*n and they wculd scmetimes maybe ever/ cnce in a while you'd see taem ahove 200 degrees, but ses; of the time they stayed be heen 150 and maybe 120 wnica before tney

i startad leaking they were always dcwn at:und 100, 105.

I knew for a '

a fact a leak rata is regrired ever/ 3 days. That leak rata had '.s be fudged ever/ time we g:t, Just accu: ever/ time that we gs-it, we had to do-ssat,: Ling :s saka it rign:. Ve as_c:n. 1 r:cm c: era-drs on my shift, I knew, we kept asking what are you genna doJout these valves.

They're leaking. We,can't get a leak et.a ou :f :ne c:cr_u ar.

'at can

(

har:iy even cc a hanc calcula:icn and nave it c=e :ut rign:.

• de c:n :

('

have := may:e g: 1cok at sesetning.

I: was fus a :ac sf.uati:n.

' dica's lika it.

f.

o a'

i lli! TAPE 254

-=

l(

CREF4 ELL:

Let me ask you this.

'4hc cid you inform?

t MARWAN: This wculd be Dick Hoyt knew abcut it.

I knew !!arnie Smith knew about it,' an everf other shift superviser and shift fere=an,and ~

cent ci reca cperater i.iat cperatad the plan: in the previous 2 =enths

~

had to knew aheut it.

i CRET4 ELL: Ncw you said cat you falt tha: the figures nere inac= rata.

What cther evidenca did you have?

~I MARWAN:

I =entioned that wnen we simulatad ne ac::ident se si=ulatac v

it dcwn at the simulatsr with te spray valve c::en and Oe pressuri::er

!k heatars on. And the reasen that.I knew :na: this, relief valves were laaking was the fac. that if ycu turned the sprty valve eff and put ce spray systam back in its aut==atic mcde tha ycu'd have a =ntinucus '

I red acticn in whien indicatad tna the plant bas cacerating.

'4 hen you turn the spray back cn and recircuinted ue pressuri::er, f~r a tna:

beratad water cat was in :nere back into the primary syste the rces

(

wculd move cut.

CRErn?LL:

Ckay....

l f.

HARWAN: We fcught that for 3 men:r.s. I natad it.

Ever/ sinuta cf it.

ll 1

\\

ww y

r

---

o.-

- - - + -, -,

,,->-w

--->-w w-

5

/

TMI TAPE 254 17 CREF4 ELL:

Sc this was a substantial beren change that ycu were ga ting in systam.

KAR7AN: Absolutely.

.i CREI-TLL: Que ta distillation in the pressuri:ar.

• i

~

MARTF.AN: Right.

CREF TLL: Okay. So we should be able to lock like at the pcwer range charts or the :d posi-ions.

Pr:bably r:c ;csittens.

.e

!g MAR 7 AN: This was during the early days. You know there's a le: cf people um there, a le of, I consider, t=p net:h c; erat:rs.

Edtwnen semetimes it c:mes devn to the basics cf knewing what ha:: pens, ycu know, tney really, they look for the c:: licated picture a lo cf times. They don't go back to the basics.

I picked un hew :: ::erata tais systas rignt off the bat. Ycu just put it in aut:matic anc you leave it there.

If scaething happens,...or not in auto I mean ycu ;;i;.

it into manual...and yet just let it c:ntinuously recirc that way yeu*

have no heren change to worry about exce::: normal leakage, =ayce ycu t

have ta add some domin watar fer fuel turneu: just to bring the recs back in a little bit f:r control.

A let of the cperat:rs dicn't uncer-s-and that anc tney'd get themselves in tr:ccle wita all the r cs :ut' t

at !!% ;cwer =r they get them d:wn in ::a far s: taa ycu ge: :1:sa t:

(

na r:d incex curves and I d:n't know...

(

t

(

TMI TAPE 254 h

C.

, a me ask you this,.. l. Regarding One react:r c:olan s: n:.

na drain tank, ce leakage frem thesa valves c:ute go i..:o the reac::r 1

c:oTant train tank.

~

HARUAN: That's c ritet.

CRE? DELL: New if t' art was e.xcess leakage it wculd require frequen:

s. art::3 of on :ransier pu=ps.

C rnc:?

s HARTMAN: Tnat's c:r act.

CREF4 ELL: 'das cat an operatien that you cust:marily go.nreugn?

(_

% WAN:

Sinca the relief valves were leaking I can tmercer, and there of late we had t: pt.7 it at less: a ti=es a shift.

l' l

l l

CRET ?LL: Okay.

1 HARU aN: And that was, I can't even rtce :ar t.e num:er. 'da pu== it!

for about 5.minutas and'it was pr:bably 1C0 gallons ;te si..u.a.

So pretacly a t:tal cf 500 gallens each tir.e.

!(

CREF4 ELL: Or 2000 gallons per i hour shift?

l 1

i i

't.

l l

3

(

TMI TAPE 254 13

(

HAR M N:

Right.

In fact.- I :ck a back1cck at the legs ft:m the ti=e-I that we started 'ta have to acd watar in.: the makeup tank t: keep RC5 lt inven.=ry. And at ene time I can reme=ar they wculd pu=m 2000 gallcas of demin watar a day. New that is your infer =ation.

Ycu can get that right out of the c:nt 31 reem cperat:rs leg :: verify that.

I even, !

!?

ycu knew, they, the pecple that I had to reper..c dicn't even uncerstan:

i tha sericusness and I believe that dat was a saricus presies.

t CRE5WELL:

In what way?

\\

HAR M N:

Centrol wise, beczuse eve:j:::y hac a ciffertn way of :n-trolling. One time I would c:me in and the s:rzy wculd be en au :mati:

(

building up boren in the prer.suri:sr.- Pretty scen I'd enc up with =y recs at the incax limit. Ncw, wnert's =y bcr:n in the RC. ? I cen't 1

' knew.

Now, what de I have :: de t: get the rocs cut? Well, I can cnly

(

assume cat the bcr:n is in ce pressuri:ar and manually,scray.

How long it was thers, ce enly thing I can do is g: tack ts.he 1:g.

If the ree:rts were kect ac:urataly, then I c:ulo maka a pretty goed juegn

(

i of, ycu kncw, wnecer my rods wara genna go cut the t:p cr wnether I !

I was genna have to add seme demin watar to Keep them in.

CRE!WELL: Well, let ee ask you this, Ha1. C uld ask for a samole on the pressuri:tr and ask for a sample en letdcwn, would that help?

i

/

\\..

s E

s'

~

!c.

TMI TAFE 254 20 a

HARTMAN: Yeah.

I'm not sure hcw cftan those sa=cles wert aken.

I

' ow they wart taken 'at scoe intarval and I believe it was enca a week.

c

~(

And I can remember at cne ti,me a sam:le came back and,it was 100 and, well to the best of my knowledge, it was around 120. Differenca bar aen

rtssuri

er boren...

l i

1 CREFWELL: 120 ;;m differenca beween ce ;rtssuri:tr and ue react:r?'

e HARTMAN: Right, between ;rtssurizer and react: P.

And if I remem:er c: rely, that was with c:ntinuous scray. Aftar a wnile we hac gottan the point whers everybody was kind of ccarating te ';rtssuri:er i

l

(.-

spray systam in a, I c:n't want ts say it, a c:creinaad fashion, whert s

everybecy kind of die it the same utt'ycu could know whers you,wert at, how.t cperata it.

'e CREF4 ELL: What brought about this c:nsistancy :f :;eration?

MARWAN: Well, I did a let of 3C taming.

(

l l

l CEEF4 ELL:

Do you have any indicatica c at manage =ent beycnd c;eraticas was infer =ed er knew about this pr:blem?

(

1 i

MARSAN: Oh, they had to. My supervisce, Bernie Smith, would, he wculd make a remincer ts all ne c: erat:rs en our shift, the ccers::r' i

l lt tnat had tne panel, ce. console that cay, hey, ::n't f:r;st ::

i s: ty f'

I I

J 1

(

,(

TMI TAPE 254 h

-(

the pressure and ccn't ferge: :: recir: the pressuri:sr for a: leas: t c:uple of hcurs. Anc one of the c; erat:rs, Kay Scyer, ne aiYays jus:

<g liked t put it on pacir: for a couple of hours and then taka 1: cff.

And mayce ha'd do that two times a shift and ever/ time he did it he i

would and um with rods out. And, you knew, it wasn't really funny tut i

then again you kind of had to. laugn at the guy because he was igr.crant.

.i He cien't you know, I don't itka to ake any benes abcut the way a i

fellow operates but...

3 CESI4 ELL: Met me ask ycu this. '4hy wculdn't =anagemen have snu: ccwn.

and recaired tr.csa leaking valves?

i HADAN:

My i=cressica of Me:-id management was nu=:er cae, O ey ;u f tse react:r ints c:=mercial operatien :afort it was reacy.

It was se obvicus I could run dcun a ' ist, and may e I will la ar, I don': 'e.cw.'

l li I'll run d:wn a list of pr:blems, design deficiencias, : st really may should have never gene up with them.

They shcuid it even have,.ever have, attom::ad t: Up with them.

(

CREr*TLL: Let's go int:. that list, let's go d:wn.

(

MARTMAN:

Afsnt new? Okay.

CREFaTLL:

ust taxa ycu time and we'll give ycu plenty of tise ::

(

thinx.

o 0

y

,._,---.-._,_mw

---

,----p

v. c.. -,-v.-

.g.,

,,,,w--.e.,

-,--+_w-

---

e.

m

TMI TAPE 254 22 4

HARW AN: We sar.ad pcwer cceraticas back in, well I wcn't say ;cwer' cperations, maybe we cid tco, back accut a year, almes a year tafere, if hack in March, I believe we, maca initial criticali y.. We cid the 1cw l

pcwer physics tasting and I believe we escalated to 15 cr 2C: pcwer.

No, I take that back. We get c ::.tC% Enan we had tr.e' safety valve 1

prcbl e=s.

There was one of the biggest cesign ceficiencies that ces:-

them ' millions of dellars f$r that jch. And te me it -as fus sisdesii;f

~

i CEEFdELL: That was wnen they rapiccad the tenergan valves with the Dressar...

HA Q AN:

0 esser...

CREFw?LL: Oresser valves.

MA D AN: Dressar valves.

The ::ncensata ;clishing sys am.

It was a nign =are.

They didn't have an autc=atic typass.

If you les: ins: u-s ment air, all 2 discharge valves frem the polisher vessels wculd fail

(

closed.

If that ha;; ens the tecstar pu=cs 1csas sucticn pressure, :.7ay trip, they cause the feed pu=;s ta lose suction pressure and they tri:.

l l

Seven vessels is normally all we wers casigned te c:ersta vita. Okay,

(

ycu cculd ccerata with 8 but that's hac engineering ;ractica.

Seven vessels cculd hardly take the lead at 52% pcwer let along 100. The cendensata reject valve which.as locatac te: ween the cencansa a tecstar I

pume suc-ica and One polf sners, if it wculd cycle because of a.ign

~

(

l.

't

.(

TMI TAPE 254 23

.I hereell level, if we are...cr a 1cw hereell level...if it 'wcula cycle-c;en because cf. a Icw herne11 level, it wculd starve the be star pues cf watar that they desferately neeced :: pu= and a hecstar pe== woule trip en 1cw suctica pressure, taking a feed,pu=n alcng win it.

This really wasn't ::c evicent until we g:: c= ::, to higner pcwer levels,'

wnen tec feed pu=s were really required :: su: ply all of ue feed ficw. Whenever the turbine bypass valves, I can't reme _her the nu=:ers.. 22 A & 5, 24 A & S I believe they are...whenever these valves c;an en a '

transient, say we had it in a trip, anc c e bypass valves wculd c;en...:c==

sta m in.: ce c:ndenser...hcreell level indica:icn wculd fail 1:w.

It would aisc cause tne =nt iler na: c:nt :ii the ner:a1 and the I

emergency sakeu: valves t: see a 1:w level and tacsa valves would fail x

cpen, er they vculd go c;en thinking ua ' here was a 1cw level...New ce opera :r a-this time saw less can 10 incnes in the het ell and I don't knew hcw many of the other c:ers.::rs reali:ed tais, but wnen T see less than 10 inenes in the horsell and I've g:: 3 ;u==s satti g there sucking at 1,000 hersepower a:f act, ! 2: very c:ncarned ahcut that-equipment da= age.

So I wculd watch the hetwell level, it just i

~

wculd stay 1cw.

It was herrible. Anc I wculd wat:n the amps en the;:n I

the pu=p, and watch ths discharge pressure.

hat way I c uld tall if the level was actually 1cw then. Tnat was, yeah.

Other than I knew tne

)

incica :r said less than 10, I can only believe. y incica:icn, but I also reali:ed the necassity for c:ndensata flew curing a transien-liks' tnis. So I was a l'ittle hesitant t: c : =e ;ncs cff s-that particbsr J

poin~.

1 A

(.

W.I TAFE 254 24 i

CREraTLL: That was the incication you had of howell level was the absenca... that the c:ndansata pu.: won't run. Turn tne ;u== cff? If ycu 1est the condansata ficw, ycu would trip the =ain feed ;u=;7 i

MARWAN: Trip the bocstar pu=ps en low suctica pressurt trips te fee:

M s.

I i

i CREinTLL: So, you get a loss of fee: water event?

s HARWAN: Right.

FASANO: H:w many of thesa did you have?

s MARWAN:

I can rtmem:er for sure 2 times.

I c:n't know tne eEac:

dates.

It was curing one of the many tri:s they had there. Well, I I

wasn't really en them but I was a bystancer.

I was en cayshift.

I was pr:bably en the lead shift or training shift er seme ning when taey had

.a trip and I ran up to the panel a see what I c:uld cc. Then I saw t

it.

'u. than see, the next problem is with that, I mentienec that.r.e' nemal and emergency ma: aup valves saw that 1cw level. They woule ::en and they wculd dump tons of watar in the condensor. Ncw,thecc:bai

.(

level is going hign. Ncw, if it ge.s too hign, vacuum pum:s.

You ais use all that ' space that r.c= ally was vacuum, is n:w watar, and any staam that you've ge c: ming ints t e thing, it c: vers t :es, ycu can' 7

c:n=ense the s:=am as well.

So wna: a: pens? The vacuum, ;:w.

Y:=

k' losa vas:vm, a.mes:nerie :um= valves :;en!

Y-(

TMI TAPE 254 25 CRE!WELL: What a: cut them?

3 HARTMAN:

If you ge: OSTG tube leaks, you're in had sha:e.

o CRESVELL: Have they cparated properly, the at:es:heric.dum:s?

I 1

MARTF.aN:

The only time I ever rs=em:er =cse things c;erating, we had.

l just gotten off shift at 3:00. At 3:30, the enc: ming shift had a tri; and they lost, I dont knew hew they Ics: vacuum..Mayce they Ics: cih:

I O. ink it wa's ene of the sa=e type transients ! Just describe:

watar.

with the emergency makeu:. They just les: vacuum becausa' f a hign level in the horeell anc the a.mesc.heric cu.:. valves c:ened ar.d jus-( "

tr. art was anybo.y dcwn in that.r:cm, tr.ey would have been PAR be,ilec h a

they t: tally wiped cut the pressuri:tr heatar catinets with staam.

Staam was noad t: have escapec tr.reugn the area wnera they art 1ccatad.

I guess that was called ue M20 area.

Thr ugh the ;iping heles in taa c:ncreta s: veture cown over int: ce c:ntr:1 builcing area and it '.ent as hign as the c:nt :1 r:cm ficer, back inn the instrument shop. They (I

had stamm fr:m the bellows ru=turing on the cischarge c' that valve. !

i l

FASANO:

Seth bellows ru=tured or one teilcw?

(

HARTMAN: One that I can recail.

I knew that ene bellcws did rupture.

I can't recall, I con't think the ::her :ne cid, but they re;iacac it.

4 ni.h cr.e cf tr.e similar design :: :ne = = er one.

J

1

( +

fMITAFC254 25 4

.rASANO. ~0kay. Cther equipmen pr:blems?

.URW.AN: Main steam line su;;cr:s.

Two years age I can reameer semebecy c: ming up to me and saying I d:n't want t: be ar und when :ney trip the tuttine from 10C:: with the restrain; systam tnat they have en these pipes. Ycu'11havestam,pipesandlaggingeverynersiftha.!

tur:ine trips.

I don't really know, I 'c.ov we went uc, we neatad u= s:

we did have saturntad stamm in these pipes bef:rt the restraints vert.

put in. I believe $csa restraints wert put in curing tha relief valke cutage when we reclacad all the relief valve, they redest;ned the pipe l

hangers and snubber arrangement down :nere en.hese.

(1) 'I's ne :: -

i h:: en Burns and ?.:we because tay never desigt.e: a ;rtsstri:ac wa:d

(

g ratctor plant, they only ever designed boiling watar reac.:rs wnica is cbvicus becausa of the 5 f ct ::nc sta wall be -een tne turbine builci..;

and ue c:ntr:1 recm.

I can't c' ow.

Did you ever n:: ice tha.?

ASANO: Tners is a, ekay, ycu'rt talking a: ut wnere tne fire :: r is be,een the tursine building and the...

l(

1 l

l u.ARW AN: And be control builcing, yean, where they are.

i 1(

C.._._LL:

I've oftan wendered abcut that wall that's, n : :na wall 1

..::w-l beween these do buildings but if y u'11 notica.here's a wall that c:mes dcwn in Sat hallway wnert dat c:or :;ees v: and its se:aratad l

l t

fr:m :ne fle r by an inch to 2 inches.

r k

l

(.

l i

t

(

TMI TAPE 254 27 HARWAN:

Yeah, I've seen that.

I never really wencartc wny that was thers but I just kind of chalked it up to O.e res Of One cra:y :nings I saw.

e CRE5WELL:

Okay.

'4 hat ateu: the engineered safety features ecuf; men:

itself? The hign pressurt it.dtc.ica pu= s, wnert thern ever any :r=ics with.them?

l I czt really rt:all any pr: lass with'these.

Unit 2's HARTMAN: N:.

n kind of lucked cut. 'de hadn't burned any u:. Uni: 1 wen: -J.r:ugn a :f them, I guess, befers tney learned their lessen.

(

E CRErdELL:

I understand the suctica switenes, icw su::ico switches have been taken eff these pue s.

HARTPAN: D a,' s c : r t c. 7"ney used to have 1 3 ;cund'Icw suctica pressurt tri; on them, I believe.

I can't really reme=er ::o muen

nat far back, tu: I knew that I di:n't lika :ta particular taing i,

because a ict of times you wculd star-the ;== u: and it wculd tripj rign away because of *.he 1cw suction ;rtssura.

It also has a icw i

i edscharge ;rtssure. No, no it dica't.

It just had a 1:w discharge l

pressure alarm that I can reme=ar. No, these pu s ever/-ime I've operatad nem I never really had any :r:blems wita them.

(

d(.

I il

l 4

TMI TAPE 254 23 I

\\'.

FASANO:

Scuncs like acs: of your casign deficienies are c:ncan rt:ac, 1

en the balanca of plant on the sec:ndary sica at least, were thire any cn the, I mean about, are there any more you that you have in =ind?

t Can you c:ntinue en you liit? And if indeed en the H53 side?

i, MRTMAN-Well, the whole nuclear staam su:piy sys as, in fac. 04 wnole plan: was desig'ned fcr a placa dewn in New Jersey. So in ordety-t= ac::medata fuel handling buildings uey had t take and r:ta.a te, reacter building, tha guts of the react:r builcing, ce insice ; art, SC degrees. So that ycu c:uld go ints the react:r building and leck at the wall and yeu'd have ancther 150 f ct cf pipe running ir:une tne cutar ed;t of the walls tha should have never been thers.

It should

(

have gene'straignt. cut but I can'.t help but think that because s' trying ts, having a plant en the drawing board 15 years ago for Ferkac River, New Jersey, and then just taka and accifing these plans, bringing them to Thien Mile Island and c:nstructing a pcwer plant of ::a:

c:::lexity

~

that things aren't scing ts be wr:ng. Y=u knew, there are :afinitaly g:ing ts se design deficiencies.

Semebody tnat designed the rystam, i

scmebecy else is gonna c:me along and change it to fit TMI. One thin's that I have...that I never really had close c:n.act with, I knew that:

ne auxiliary perat:rs, just because :ney wert cut in the plant, taey t

had a close c:ntact with it, was the fact tnat they had ex ansica c:ntrois that went througn the wall and they had clu :hes that cperata:

the valve behind a c:ncreta wall f:r racia-ion. Th:sa each limit, li=iting type aing. And mes: ef :ne ti:e the valves wcui: never

(

1

3

(

W.I TAPE 2E4 29 T

c erata, so you'd have to go thr ugh tnesa, back into ce valve alley-to cperata the valve anyway and you had t: spend anywhere fr:m 2 4 3 times as much time in the valve alley, crawling over all '.hese ex:ansic=

. (

c:ntrols that were in your way that, ycu knew, the peor guy would and up getting three times as auch radia-icn than he shcuid.have.

l J

l i

CREra?LL:

Is this a wide spread problem er an isolatad preslam?

HARTMAN: This is wice spread. Especially at filtar r:c=s, the =akaus valve sue icn alley, the discharge alley and ne 305 valve alley.

CRErai!.L: Those are only hign radiatica areas during c;eration?

HARW1.N: Yes. New I'm net sure, I knew the levels were creeping up there in the latar days of power operation.

I knew they were cli:hing u: there and there wert still valves in there tnat had t: he cpera ad for surveillance pr:catures, valve l'ineups and the like. Filtar r: cms is tne same way cnly the filtar r:c=s, you can't ge in.o them.

(

CRErnitt: Do you recollec a trip tha: cc:urred back in around Nove=tr 3rd, Neverer 4th of 1878, a loss of fetena:ar type of tri;? This is I

where an instrument tachnician threw the wr:ng swite, and in c:ncensata polishing systas, all feedwater was icst.

Do you rs=ec er any of the!

details of that event?

(

(

s I

I L

(

'/

A.

TMI TAFE 25.L 30 b IH:

Y:u wert..:=ing in af'i-a:ic.., wert y =.c:7 t

H;R MN: Yes, I was.

I rema::ered.

I'm : ying = ::: ink.

I nave ::

rt:r: grim =yself, hers.

I have sc: :: get ta:k inn that..I've been l

thr:ugn two t::al less Of fetc'naters.

. g e_r,,

.u.g g t c _ _..

HA M AN: Yes. One was 1 :tr: ;cwer P very 1:w ;:wer and One : na.-

ens was at IT. Oc-er.

Anc : really :en't races:er =c.any cf ne

n ails.

an k.:w na: -J.e r.it:n tha ne ;;y urtw.as ::n:r:1

wer f

r all Of.he valves ir. =a ::.:ansata :elisnir.g systam tha:

r.

i,.

= ace them shut = :i.; off ti; ::= ansa a f*.:w :a=.

CREF4EL':

. Do.ycu races:er any c;: erat:rs :n snift cis :ssing =at win yeu, hearing any:.ing 1: ut it?

MA D AN: ' dell,

remac er, rey hac. =e '.ER =a: was :ir:uiati..g an ney had a syncesis of :ne event ca: we riac in.

'ae hac :: sign an:

initial.

I ORET4 ELL: Well, this par.fcular even;. =a: I': ::eaxic,; Of, I can't teiieve there was an LIR generatac.

t 2:. un:

see.

7 A.

.f e

,,a

,,.-e,--w

-.-,-,-e--,

,.-------s---

---

e.

i

(

TM* TAFE 254 31 5

CRE_r4 ELL: Thers was one Neve::er 7th that was a loss :f :ne feed pu=p, hu. that was with tne r.:n :acx.

Okay. At t..is ;cin; in ti e we're '

getting ver/ c1:sa ::~the end of tne ta:e, s: ~we'll break rign: hers i

and c:ntinue with a naw: tape.

x 4

MARSH:. Time is,7:32.

I'm genna brtak at this :cin. and turn.ne to:a e

9 over.-

M:

Resuming-1: nis time, the ti=e is 7:3~.

CREF4 ELL: Okay. % hat a::u. ycu training at Th e Mile Is'lanc L' nit 2, hew's t.'.a: heen?

- s e

i, s

F.A D AM:

I han t: say this but I'm genna have ::.

They train, y

- s m

training I theuqht was very well d:ne. They ; e. u tar.ugn a pre ty

'x t

~

c:=:rtnensive Jr:gra.m. th2 3 weeks at Lyncheur;. '/ us we hac la s cf s l

s 4

x time := cursel res just c:ns ra::ing 0.e plant wnan :nings '.ersn't.je y s

x t:usy.

'de c:uld get cu. ints'-he plant tracz :ysta:s anc alika. Jie\\

K

~

went thr:ugn a meck NRC tas: that was given by a Generti.8hysics, tre walk areund ar.d the tatt Aad then wt had the actual tast.

'eie had a 1c-of ;rslicansing training thati thougn: was a big hel:. Cns."dt: we ac' 0

several lectures. They waren't always d:ne as planned st wfl i:: get [

s in s me training ensnift.1 The guy would ske a systam and give every::cy I,

-s en.he shift a lec :.:rs abou; it.

Suih*.elica.hsadcetft::rs.hathavet '

s N

T s

h

\\

s N-

=* ~

h

\\

\\.

^

g g

\\

o

,l

(,.

TMI TA;E 254 32 i

.i v'

c:me en sinca :ne c:ld licensing gr ups, sinca ne initial gr u: Of cperat:rs tnat went u:, th:se c; erat:rs art trainec :: take an NRC

>f ext =ination. They art net trained t: c erata tne plant. They have c: pies of HRC tas.3, the questions, you knew.

They'rs...

f CEES'4 ELL: Hal, le: =a ask you nis. Y u sta:ac tha general physics' c:ncuc.ad a c:/ run of the licansees. Oc ycu fine that taa NRC exams i are predictible, the way the test will te c:ncuctac?

MARWAN: Yeah, I think they are fairly well ;tedic".ible. TMI has g::

an excalien rec:rc.

I c:n't knew the rec:r:s Of any of the c:ner

(

c; era: fag plants. Eu I do b :w that TMI's rec:rd is gecd.

I d:n't

=:1 think.we've had but ene failure and taat was in Unit 1.

'de ' ve had several senice c;erners tha. wen u: fer a sanier's licanse bb: did fail the senier par. but g: a react:r :; erat:rs licansa. And, ycu 4

k.:w, I can just say that I feel tha this operating rec:rd is indi-cative cf knowing wnat : ex;ec:.

They can tuild u: on it.

If they 9

know tha. a cartain examiner is going := c:=e, tney can dig cut all his [

6 z

r old tests, they can, we have old ir.tarviews that semeboty =ight have :

I i

snuck a little tape rec:.rder in their ;ccke: and tamed ne entire walk i

arcund and you can get the ta:e c:nversaticns of tacse.

I've seen I

tacse flying ar und.

/

/

s L

MARSH: Have you actually seen a ta:e er heard a :2:e bef:rt tais was'i

(

c:na, or just transcri::s thertof?

C 2,f,

i

~

t I

TNI TAPE 254 33 t

HARWAN:

I've just settn transcri;*J no ta:es.

In fact, I c n't even knew of anybecy at Three Mile Island making a tape.

f CREMLL: Do you knew wnat the scurca of the ta:e was?

HARTMAN:

I do knew this, I kncw that it was a Sa :::% ar.d Wilc:z 1

rea:: r.

u CRE.NELL: Okay.

MA D AN: And it was a spe:1fic.txaminer.

I can't even rimer:er his L.

No, I wouldn't even begin := atta=:t t: r e e:ter his

.a=e.

I nea.

d:n't rs= amber.

CRE!WELL: One in.ariiew, it was one intariiew that ycu had neard of?

(

MARBAN:

I think taere was one, one f:r surt, maybe twc.

l CRESWELL: Okay, i

t RaRSH: khart would I 1cok if I wan ad to find a 'sa of those? Whc l4 would be my best sr.ot inat I c:uld talk to?

HA D AN:

I tarew all my old stuff away, t

a

\\

\\

(

g.

s TMI TA.8E 234 24

'\\

i

\\

v MARSH: You had nad a sat of tne=?

HAF.TMAN:

I had a sat at one time.

CREF4 ELL: Did ycu recaive these through your e. c1cyment at Three Mile

~

, Island?

i HARTMAN: Yean.

I can't recall who I get them ficc.

I can't reser.cer if it was the training de::a-:=ent, unica I don't thing i: was.

I thinx it was one of the et.'.er opera:crs and he =ign: have ge::an i-fr.s training.

~

(

PAP.5H: Okay.

If ycu have any acettianal reccilections on :nesa, I'd appreciate you get in touch with me.

I'll give you a cz-: and a pnene nuncer and all that where you can reach se.

HARTNAN: Yes, i

CRErdELL: You mentioned before you'd cc:a en shift and fcunc these '

l tw ive valves shu. Befo.e. What about, have yeu cene your lineum on' ycur panel before and found c: hor valve mislineups?

(

'I

(

d t

,m

i i

[

i TMI TAPE 254 25 1

i i

I cue in one day, this is jus an\\.)exar.sie of seme of ce HARWAN:

things that I'*)e had to c:=e int:. Met-Ed was always focus f:r per-

?

forming an evolution 20 minutas hefere snift relief. Turning tne plan:

over in ::tal chaos.

I hated to turn the plan over that way myself and I more than hatad recaiving a plant like that.

I ca=e in ene cay' relieve the shift tha had had a trip and I can't, recall ut exact,i i

t:

wna: ha:pened, bu,I know that there was an c; era:Or was t ying to j

ntrol pressuri:er level wi.h MW 153. New that's a high pressure infection valve en the A locp. He was thre : ling uis valve, pressur'i:ar level wculd go ut, he wculd c1csa it, it wcule c:me back d:wn again, and he wculd jus keep c:ing tais. And I askac him, "'.nai are y u dcing?" He said, "I'm mainuining pressuri:er level." I said, "waa:

<(

ha::ened t: ne acimal?"

"I den't knew, it just c:esn't wort.". And ne went over to.the pneumatic csn: roller f:r MW 17 anc he shewec me, nothing ha;pened. I said "did ycu check MW 12?" Tha 's tne manual' l'

is:1ation to 17.

CRET4 ELL: You're indicating semetning there...

t 8

i MAR 7AN:

I'm indica it.1 that I pointad te MW 18 and Oa valve was in fact 3 hut.

And he "says, 'Ah," and :t..ar ex licitives. And you know, I

(

says ycu gotta wake up.

I said you can't opera.a a plan: -his way.

came in another time.

New, I wasn't really taking Over the shift but we were walking ou the passage way c: ming in f :n 'Jni: 1 whicn is j

nor ally =e way.e came in anc avery :nca in 1 nile e:: near a

(

-(

. (,

TMI TAP 5 254 36 t

safety gc. And it would g: f:r saybe 30 sec:nes then it would rtseat.

We'd walk a little further and it, ;cw, wen again. What art they c:ing up thern? I was with ancther CRO, we were just a: ut reacy :

.I take the shift.

I don't kn:w. They're net tasting them.

I th:ugt.: we f

were at pcwer. Y u knew, what wcuid they be testis.g nem n:w f:r? We 4

get up thers and wnat had ha:;ened was iney lest fstewatar, pressurt in

(

the stats generators went dcwn to the ;: int whers iney actua:ac in :ns' fenesatar latching systam which cuts eff all feed :: the genera :rs fr:m the normal fetcwatar pu=;s, and it also shuts MSV 1A, 5, 7A, an:

A, wnica art the =ain staam isclaticn valves. The sensing ;:in: fcr tur:ine header pressurt which c:ntr:ls the tur:ine hycass ' valves is dcwnstream of the MSV 4's and 7's.

They rest:rtd ner=al fenc, but :ney k-f:rget t= c;an the MSV 4's and 7's. So that the tur:ine ty; ass valves were seeing 750 pcunes pressurt and the turoine, tne reifef valves in the stamm generaters were seeing 1C50,1050. So everyti=a they tried

?.

c:ntrol pressurt with tne bypass valves, they had these in manual, they woulo close them d wn becausa they'd ses ;rtssure was starting ::

cacrease, in clesing dcwn the ;rtssur woule c:me tacx up again anc I

oof.

It wasn't two minutas in:= the shift and Ray Scyer, the guy t$a:

i.

was taking the panel. sa'd, "wnat art ycu trying to do?" He saic, i

"ycu'rt blewir.g safties cut thers." "But we can't c:n r:1 pressurt

(

enough. And lock at header pressure.

It's dcwn icw."

Right nex: :s it is OT5G pressure.

It was 1C50. He says, " Man, what'i :he differenca hert." And he Iceked up and th'a MSV 4's ar.d 7's wert snu. He says,'

(

[ (.

l.

f.

I O

e

,e

(

TMI TAPE 254 37 t

" crack those valves." They cracked the valves, ney cicsac :ne tur:ine by; ass valves, put them in aut:=atic and the thing cz:e righ: :n up 522 215. New thesa art trained :;erators.

I den't lika :: say :na: =y shift was tne best but, ycu knew, I think we wert.

f tiiT4?it: Las ce ask you tats. 'ahat art the difficulities na; ycu.'

enc:untar in cperating a plant of this casign?

i RARTMAN:

Fetcwatar is very sansitive.

It, I c n' t ka w.

I c:n't want to say tha fascwatar is sensitive. 'nhen you acve il millien po$nes o' watar an hour,.that, neecless to say, is g:ing :: he ::vch'y, ycu knew, whatever its just that a slight change in ficw is genna causa a big i

k-cnange in tne staam gsnera::r, but no :nly tha; :ne pri=arly syszam is very sensitive. The ;rtssuri:sr was ::: ally ::o s=ali. Any cecitasa in primary systam tar: era urt, which woulc result fr:m in inc$tase in e'

fee: flow, would cause the pressuri:er level t: g: cewn and the ;rtssur':er pressure to go ::wn.

It was really very hart :: ::ntrol in :nis respe::.

Ycu knew, I, in fact, wnen we, when I ::n rel the feecwater, the a:cun:

I s

of feedwater to knew hew much :s put into tne stats generat:rs to kes:

the reactor basically as s.able, as ' stable as I can keep it in a tra sier.t situation.

I look at react:r pressure.

If pressure g:es up I feed a

(

little scen. ' shen I star-to see it c::e ecwn again, I :ack'i: cff a 1

lit:1s bit. And that's how I knew wners to kee: fasewatar flew.

l

(

C f.

g k

TMI TAPE *254 28 1

CREF4 ELL: Hcw do yeu, after a rea::cr tri:, wnat are ce ic=ecta a actions cat y u taka? Could ycu walk us thr:ugn vna; a::: ally gees I

on?

i MARiFAN:

Well, the first thing that ycu do is you insure t.?.at, ycu i

ma. inly trip the reae::r. Okay, that just insures cat, well, I c:n't knew why it would, but the sec:nd thing says t at you lock u;:

t en =a.:

panel and verify that all ce red in-11=it lignis are on.

Ycu check t:

make sure tnat the turbine is tripped. That it's auxiliary oil ;u==s are cperating. Makasurethatt$etur.inebypasssysta.: is c:ntr:11ing stamm heater pressurs at 1010: You clcse A"/ 375, wnica is a le:dewn.

isciation valve.

If ;ressurizer level ge.s c:wn belew 1C0 inches, you're supposed to start a sec:nd makeup ptim: and just kete i tacy t:

go if you need it.

If pressuri:ar level gets c:wn telew, I taink its 20 inches, then yeu'rt supposed t: open MUV 153 to acmit mers wa:ar. '

If be saxeup tank is icw and the pressuri:er level is Icw, taen ycu shut, or you open the DHV 550, wnich is a11cws the F45 :: c:=e ccwn :

the suction of the =akeup pumps and then ycu snut E'V 12.

'{

t I

C.t!IdELL: Is that a va;ve ycu have t: go cut ar.d manipulata manually-or do you...?

(

MARTNAM:

Its ene you have walk around 20 feet of panel and tacx 20 feet to c:en.

i f

(

i J

~k TMI TAPE 254 39 i

CRESWELL: What abcut the feteestar c:n:rci...?

1 MARWAN: Vell, if you have any fenceatar sutions in hand, ycu should verify it, you shculd run these back censistant ts the parametar that they should be, in cuer words just uke it and take itiall me way i

i down.

I CRE5WELL: What if cey're in auto?

,I i

. HARWAN: Well, if they'rt in auts you just veri *y that femenatar flow is c=ing back at a rata censistan win ce hencer pressu're.

CRE5WELL:

Mw, they'rt sat, tr.e' fascwatar is set f:r 30 inches,. rign-!

?

't.

HARWAN: On ce low level limi.4, right, yeah.

CRE_r4 ELL: Has that always been tne czse? It's thays been sat at 30' inches?

1 1

?

HARWAN: Well, 30 inchos is the acminal, is it ncminal nu=cer.

It c:uld be plus or minus. What they do is they set the levels in the

(

staam generators at 532 cegrees so that they c:ulc get 532 cegrees ESS

sig.

If they need a little bit more heat transfer ta get that !E5, then they wculd raise the level slign:1y, ycu c' ew, variec =ay:e plus!

(

cr minus 3 incnes fr:m 30 en eitner genera::r.

I k.,ew :ne ua: was

(,

accut 32 and tne str.er's 22.

I

---

*--E

A

(

TMI TAPE 254 0

4 CAirdELL:

I guess you've had a caance to talk to Oe c;eraters since the event. Have they netad anything to you that was pe:uliar a: cut the 1

event.? Where they had t take special acticas anc...?

FA D AN:

I really didn't get that s ecific with taem..The :nly time !

I ever talked was in a bar reem. Ana I, you knew, I con': like to re:ea

~

wnat I hear in har r: cms.

I

.ASANO: You sentiened that y:u at ene ti=e f:und.*UV la shut. Whak' rassen wculd anyone have to have that valve in a closed ;csitten?

MA D AN: To the best cf my ka:wledge cat particular event came wnen

- k scmettecy tried to :.ange a ligt.t bulb in.he ICV i valve ::n: oller.

They rsmeved the lens covers and everything, They pulled the olc lig.:

bulb cut and.they went to stick the new one in and, you knew, t.'.ese A

were tnese PS3 120's that they're telepnene lignts. They're a::ut taa long and they have a c:nact making surfaca en either sica cf

.em abcut mayte a half inch. And as they slid this thing into tne sccket, b

P it made c:ntact with the het side and tne ground, it hiew the fuse fer i

that indicating circuit. It also t:ck away the incica icn fer all the ether valves and ha:pened te fail den MUV 18 saut, when tney re-energi:ac I

that.

l l

CRE!WELL: Was it, did maintanance repair it ;r:=:t'.y?

t C..

l

(

4.

.I TMI TAPE 254 41 HARTMAN:

I really don't kn:w the time fra.me. Wha Saw, I unters an; it, hcw it ha:yened.

I can' remamcar when tha trip ha;;ened.

I knew

ney ba'd a trip on that partic$1ar instan:a.

3 I can't rsmem:er what tne time frame was he: ween the ti=e that the fusa actually blew and the time they got it fixed.

I wasn't on shift at na time, Then I ce i

recall a sign hack en that par-icula-cperating panel that said c; era::rs w

ts enange lignt bulbs in this panel, call the electricians. he are no we have ts call the electricians and have a light bulb c anged. Anctuar ene of Surns and R:e's designs.

CRE!WEL' : Okay. Cne thing we haven't talka: :s y:u a: ui is wny you quit.

RARTMAN: Why I quit? Abcut a year age, well, it was even, no, it wasn't a year ago.

It was a year age.

Back in June I had wantad :: '

leave this racket for a unile.

I,think it can he a good carter.

In fact I enjoyed my wert um there.

'4 hat I did, :r what I had : g:

tnreugh :s do what I did, it was hec-ic. 'When I left Me:-Ed my bleed f

t pressure was 180 over 110, which fer a man of 30 years :14 is cuiragicus.

t My biced pressure had teen hign f:r over a year.

In fact, I had gene' jet hunting back aucu-this time last year and I had g: tan a jcb bu

[

it was a little less seney than I really wantad to take an: it was in St. Louis, and I didn't wan* to move.

So it wasn't really a spur of the moment type decisien, whica a let of pec;1e tacugnt it was becaus'e l

ef : e ac:icent.

I did know that enca :ne ac:icen na::enec, wcen

• C

(

o g

(

INI TAPE 254

.:2 t

reali:ed the severity cf the accicent, I bew that it was g:nna te a long time.

I hac just gene thrcugn 21/2 grueling years cf bu11 snit, l

whics is what it was, with GPU st1tr:tp, UELC star::c, and then Met-Ec.

And I didn't like being hassled by 3 cr 4 differsn: fereen anc 2 cr 3 8

different su:ervisers plus 3 cr 4 shift tas engineerings anc various.'

(

other mechanical and elec rical engineers, you knew, it was jus: co much I csuidn't taka it anymcre.

I thcugn: I was a gced...

There was shift supervisers that wculd actually stand over y' cur shcuicer anc tell you exactly what to do. Raise sta m watar, you ge: a ge fandwatar uc, new check this and check tnat.

It was, ycu knew, ridiculcus.

I knew hcw t: operata the plant.

He sncule have been haci inere taking care of his paper work, but instanc, he was right us there in c e fic.:

lines, and ::. ring to keep myself crientad 'in-my own henc ! always nac to lis an ts uis guy. And if I wculcn't do something he sid to do, well, he was righ cn me.

He says wny dicn't ycu cc that, wny dicn't l

de cat. Well I cidn't see any i=;ct anca :: cc that rign..a: ca:

particular time. And ses times I was rign:.

(

i CRE5WELL: Well, did they bew what they were doing? The snift sucer-visors?

1(

MAD.eN: Yeah, for the most par., but cey wcrried abou ce wrcng '

things.

Ycu know, they put their priorities a little different can I di d.

(

r a

-. _ ~

(

(

.TMI TAPE 254 43 t

CRET4 ELL: What shoule they've been worrying thout?

8 1

HARTMAN: The overall picture. They should have been back staying taking a big, a big 1cck at everything that.was going en around the:

l and no getting themselves involved in what the c:ntrol.r:cm emerat:n i

er the shift feraman is ceing.

The snift feresan c:uld direc.

Isa

?

team effort and I tried t: pr:mota tha y 21/2 years that I spen en t.at shift, I tried to pr meta team work. And it just dicn't verk. I Thert vers personnality c:nflicts, c:nflicts of intarest, period, you g

kr.cw, tne guy just c:esn't do it the way I tall'him, er the way I ask hid or tne way I shew him that maybe that's the tes way 5f :cing...

He woul:n't de it just. ;1um because I id him, and ne would do it fus

(

to defy me.

Now, this, ycu can": have taam work -hat way.

CREF4 ELL: Let me ask ycu this.

'4 hen :ne sucerviser was stancing ever

(

i ycu shculder, was this during a trip ::ncitien r during ner ai c;ers: ten?

[

l HARTWN:

No, this was curing, in fact, I rtmester this sta-tu: 7 tally I

well.

My pr: clam has been diage.csad as labile hypertansica wnica se as i

gces up and then I'L gracual : tring it back ccwn.

I dcn't have!a c:nstant high biced pressure pr:blem.

I keep things he:: led uc insica T

cf me. This par-icular nignt I came in, I had the canel, we were.

1 I

l schedulad to do a star.um f cm 1% snutdown to 15 pcwur. Vi.a all the pacer work and every ning that's involved in ::ce t: mece checxlists I i

anc -he surveillanca tas s that had : he cene :ri:r : ;:ing critica:.

l(

t

~

I Y

(

TMI TAPE 254 44 I,

getting the c; erat:rs sta:icned wnere they nestad, ;ust getting t e plant in a general s'able c:ncitien t: perf rm this thing, it takas a.

t let of c:ncantratien.

It takas a lot of eff:rt en the ;ar: cf the u ntrol roca operat:r to do that. At that particular ti=a there was, in our alarm systam is t:: ally ridiculcus, there were ::o many of them and the systam that they have is it fails all the time. An alarm car:

t c:uld go bad and it wculd just send an alarm.

It would just ket; flashing in ar.d out. Ycu would silenca it and it would just ket; c: ming in. And uat God awful horn, it would just, ycu knew, you're

• i

f ng te c=ncantrata on doing a startup and you get this 5t::er going eff and I'm ;articularly c:nscious of alams becausa if ye'u just igners inis stupid thing, what if you get ene that's impo ant.

If ycu ge:

T-one that's i=portant and it gees unre=gnited becausa y u're ig:: ring this one over here you might as well net be in the c:nt ci r:cm. My i

feb it is keep the plant safe.

If I, you knew, if I 'can' see wnat's C

g:ing on around me then I felt apprenensive ateu the wn:le thing.

~

didn't feel like I can kae; as close a surveillanca en me whole tning as I really wantad to.

Iaskedbissu:erviser,I'llevenmentionhis t

name, Brian Menler.

I askad, I said, " Brian, c=uld you have an ins:=.-

t i

l ment tach take a 1cok at that alam?" I said, "its driving me nuts.

\\

I said, "am I gonna have to listan to that all night th ugh this I

start =?"

"' dell, I guess you're genna have to."

Those were his werds pretty close. And with that I uld him, I said, "I won't do the sur:=

under those c:nditions " I said, "sicer get mat uing fixed =r fin i

se a relief."

Ar.d he said, "y:u k.cw if y:u tec:me, if jeu are relieve:

C

t G

^

TMI TAPE 254 45 you might as well get ycur lun : cx and take eff." Well, at that particular point I wasn't reacy to leave my employment there. May:e it I

was a prayer, I d:n't know what it was. if.e alarm se!.: led d:w'n and I didn't have ts listan t: it at that point.

It did c:me back latar but not until aftar the wnole thing was d:ne. They had an ECP there reach

~

l for me.

I 1ceked it ever and I can generally tall whe:ner the thing is gonna be c1csa by just looki,ng at the nur.:ers.

I've dene encugh f i

them, you knew. And by just locking at the nu :ers I can tall wae:na they' re genna be...

.ASANO:

ECP and c:ncantrator pointar...

a CRE5VELL: No, estimated criticai pcsitien.

.ARTFAN: Esti=ated critical pcsition, yeah.

And this ene Icekac g:cc.

W t

It was, I think, somewnere ar:und EC% on greu:s 5 anc 7.

It was wnere l

they wanted it no, no, I'm sorry.

It was a: cut GE%, its the elesas: I can remascer, en group 5.

No wait, no wait, I'm gatting screwed u:

1 l

here.

I've been away for a men:A anc a half and it's a little...

I i

l l

CRE5WELL: Vell, I den't think that ::is is critical infermation, t

MARTMAN: Anyway. Well, just the way they cperata.

Its, the estimatac critical pcsition was scmething. We have a gui:eline if you ;s cri:1a

,I hafers half a ser: ant less :an wnen yeu're anticipa:ac :s g:. ycu k

e

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

t TMI TAPE 254 46 I

should shut back dcwn. Ycu should put all the reds in until you get !

the safeties in and then investigata why.

Sc, I wasn't even I was just getting to the minus.5% position and all of a sudden I 1 coked up and I had an alarm, it was the startup rata red withdrawal inhibit circuit. The only thing that threws that into ecun: is 1 c;:s in the 1

it scur:s range. And I lockad dcwn and I cid have 3 c;m in the scur:a i

range.

It step;ed the red motion.

I put the r:d stick in and he says, "no, no, no. Just take it dcun one."

I said, "what c: you mean. We' jus went critical here at 2!: en gr:um 5."

I said, "taa EC? called for a half a per: ant above that er bet:ar." New that's alri;nt.

That's alrigns. We'll calculata a new EC? fer where we wen: critical.

Now that's wna they did. New that dcasn't shew en any Icg becks er

(

anything like that. Eut that is a fac... '

CRE!VELL: Was Mr. Mehler still a shift superviser s: this ;cint?

(

WAR 7 FAN: Mr. Mahler is still a shift superviscr at that plant.

I l

CRE3WELL: At that point.

I HARTYAN: Yes, yes.

3 MARSH: Menler was the one that was on duty at that time...

i 1

4

(

TMI TAPE 254 47 MARTMAN: Yes. He was the one Onat I kact talling you that he wculd icek over my sn'oulder.

So I t:ld him, I said, "I d:n't believe tha:

that's rignt.

I think that tnat is unsafe.

I think tna; there's sc=ething wrong here." Well, we'll recalcul,ata it and :aka it rignt. !

i Whica is wna they did. They redid the numcers and semanew :ney fuegite 1

(

them, I c:n't knew.

They...

i CREIWELL: Did they have a nuclear engineer c:me ini I

i MARTMAN: To the bes ef my knewledge, nc.

This was en a mics'hift.

This was pre::y late at nignt, which acr= ally that coesn't ha::er them fr:m calling anybecy cut bu I d=n't recall any nuclear engineer c:mir.;

> \\,'

l in.

MARSH:

Hal, you say taey. Who else besides Mehler wcule be involved in that calculatien?

i HARTMAN:

The snift foraman.

(

I i

MARSH: Which would be.ho?...

I l

!t HARTMAN: That was. I can't recall.

CR!i4 ELL: Okay.

I l

1<

.o 3

(

TMI TAPE 5 4 44

't MARWAN:

I d:n't want : name any names if I can't te sure.

Becausa' we were well, Menle~r is not my normai shift superviser so I'm net sure 1

that I was with my normal fersman at the time.

i CREr4 ELL:

Could you estimata when in ti=e, it culd help us : ;ursue I

(

this mattar if you c:uld give us the appr:ximata da:a. Vas it early,in th,e star.us tast pr: gram?

i l:

MARWAN: Yeah, I believe it.as right back before -he relief valves CEEF. ELL: And taat would have been lilie in Acrit er May in 1572.

(k Mt.R7AN: Rignt.. Yean, it.as about that ' time.

CRETWELL: Okay. Y:u senti:ned taam wert bef:re. Are tne panel assign-(

ments clearly indicatac to pecple when they're ensnift?

HARWAN: Generally on our shift wnat e dic was we hac a panel :: era::r.

(

He was in charge of taking the, filling cut the log bcok and just generally evenesing t;e operation of ce plant in a wide se:ee.

i Normally we were at stancy stata. We also had a serson that was assi; ec

(

to what we call the swit:hing and tagging des <, and he wculd taxe care j

of any safety tags tnat neeced to be hung.

He would.also perf:rm the daily logs and the snift and daily surveillar. as anc any c:m:utar I

4 l

infermation. hat r.ad to estainec ::at :ay.

Then.e usually.ac a :nir:

i t

\\

=

(-

TMI TAFE 254 49 coeratar tha was assigned t: surveillanca cask and he wcule eversee '

tae, take care cf c::rdinating the c:ntr:1 r:cm and ce auxiliary operators in perfer=ing m:n:hly cr weekly surveillane tas:s.

If anything went wr:ng genera.11y what would happen the c1cses: guy := a section of panel, you knew, like we used ts civide it un ints 3 sec-15 s.

'l We had the sec:ndary, we hac ce react:r plant and we had tne nuclear i

staa.2 su;;1y systam which was basically makau: pu::s, low pressure 3

infection and that scr: of thing. And can the fere:an, he woulc xinc cf r:ve around and =ayce if he was in the back panel and if there was scme valves back the e that he c:uld c;erata. This is generally ne way we werked.

If secrething ha::ened, y:u ka:w, ycu'd' yell cut and say...analy:e the pr:blems as best ycu c:uld...and say, ".e icst fee

(

water." And at that time the t.irse cperat:rs woule c::e up and cey l

would key in en a pcsitien tha: wasn't ec:::ied and then take ever.

1 C.tEF.E!.: What about shift turncvers? What scr. cf turnovers c'd

eeple crdinarily make to ycur kncwiecge?

1 lt l

l Mim8AN: Veil, May varied be::veen saf fts.

It cepenced wnc ycu turnec 1

t j

over to and I always tu ned cvar ta =ajcr stuff, any, I always let t'ies knew where they were in rods, what they were dcing,.hecer :ney ere 4

I c: ming in or going cut. I let them knew of any abncrmal conditions tha 1

existad in the seconcary plant, ycu knew, like they art dcing a ;clisner l

vessel, regenerating that. Or we ur :: led cis d:wn anc pu thisi$

lt au=at'c and seme:..ing, may:e a =a,':r :nange Or:ughcu: ne day. Any l(.

1 l(

l

4

(

TMI TAPE 254 50 x

jchs that were in pr:gress, I'd ge ever to the c:mputar and snew him the primary systam parame:ars, pressure, beron c:ncentra ica, reac::r

(

pewer, and such inings like,that.

'4e generally have a writtan sheet.

Sc=etimes guys,wculd draw it cur for 3 or 4 pages and taey wculd put the mes miniscule itams en taere that, ycu knew, generally : Icekac i

ever tacse.

But tae majer stuff, the turncvers, ! guess, were generally pretty geed anc I say tnat en a stancy sista basis.

Like, I cen't knew i

hew may times I've takan the plant over in a transient and it was total chaos. Y u knew, they'c leave, ycu knew...

CREEWEL': Ycu would have a turncver in the miccle cf a tr'ansient?

s MARTMAN: ' dell, =ayne ner necessarily in tte middle of a transi n:.

3 let me rephrase that. Let me say after the transient was ever, perna:s curing sne rec:very wnica semetimes ::ck days. But generally, you c:uid ge things straignianed cut in a snift to the pein: where, ycu knew, the major work is cene. The big things that have :s be d:ne and the small things. Mest of the small. things are cent..The big taings

.and they can c:me latar, you kn w.

But its really hart when you have' a i

trip. There's so many.hings that ycu might have seen :nat ycu t:ck I seme carrective actico fer, that mayc'e seme:ccy else signt net have t

done that.or maybe the precadure dica't call for but it was ekay is to e

it.

It was safe, it was c:nservative. Anc you ferget to sention that

ycur relief, and you go tais valve c;en and :ne.k you go anc try anc 1

ce semaining else and, you knew, it jus; ccesn't its:cnc rign; an: you

.(

O 4

r WI TAFE 254 51 t

Iceking ar und for a pr:blem wny taa: cummy left that valve c:en. Why-cid he de that? You get mac at tne guy because he didn't tall ycu

(

ab' cut it but, ycu knew, it was an hones; mistake.

I very seldem g::

mad at guys for pcci turr.cvers because I generally made a pretty scod.

ur cf the, rign; aftar he left I would go at:und the ;lant, tne

anels, and look for aancr=alities that I saw. If I hac any questiens i

I would ask the fersman.

FASANO:

Did ycu have a eneck snee: cr did ycu de this jus: by ka: wing the systam, its all in your head.

MARTHAN: Jus by kacwing the systas, knewing hcw the bcard was ::

1co k.

The pcsitions of valves, you knew, its like it al=cs: bec:=es instinctive after a while, t:

FASANC:

Aftar a wnile.

RARTMAN:

Yes.

t I

FASANO: You mentiened that you did have design c:ncarns en bcth the !

nuclear and the steam side of the plan:. Art thert any ways etner than I

c:mplaining ver: ally that you c:uld have reportad these t: ycur :anagement?

I mean, are you pec;1e...

(

HARTNaN:

Yenn.

'k 1

P t

714I TAPc 254 52 t

Fa!ANO:

A11cwed t: re;crt things en pa:ar?

I I

[

HARTMAN: Yeah. We can generally writa a little lettar to, ycu knew, ce c:gni: ant engineers of the pr:blem. And generally se=e of tne s

f cperating c:nveniencas, they used to take can f bu: mings like, an:

=

1 we'd writa these or even a vertal c:=ent, they wculd ncrsally fc: it ccwn themselves. They. era pre :y ge:c.

Sc etimes they never g::

d:ne, but at least they'd listan. Su: tne safer.hings, you knew, lika the relief valves, condensata het well thing.

Idon'tevenknewiff.e has yet t: day, wneuer that systam nas been ::cified. But, yeah, yee c:uld writa him a lettar and even a lit:ia ef agram snowing na you's i-like te see, ax:laining say:e even perna:s hcw :: go ac::..clishing the t:(

change.

FASANO: How ateut reter:ing :s say c:her agencies? I mean can you :

=ean talk to other pecple lika :urselves or...?

MARWAN: Well, that is, I forget the part nu=ter.

10 CFR 20?

I can:t 4

remember the regulation..,7 FASANO:

217 t

I HARTNAN: Yeah.

Its the one wnere if ycu see the ::mpany ::ing scmetning in violation of any rules that you can go u =e NRC win ne c:=;a}.r.:.

(

C.

l

(

1, t

l

.r

~

6 i '

(

TMI TAPE 254 53 CREWELL: Have you done that?

HAR MAN: No.

,=

CREWELL: 'mhy not?

t 4

MARTWAN: 3ecause I falt that if I had of gene uere, :: the NAC, snan uey wculd c:me c:wn on me.

I was a little afraid of that.

1 l

MARSH:

• hat makes you think that way? Do you knew of ouer instancas a

wnere tais has hap:enec? Have ycu been ::1c anyming f:r:aily cr inf:r. ally?

r.'

g MAR FAN: Nc. Tney never really ca=e rign: out ar.g t:1c us that, you know, tha you would be ";:r:sacutat." But I jus-felt ca: knewing un way they c; era:ad that it wculd he kinc of like being the black sheto now all of a succan an: it was ::ugn ancegn :: get along.

I founc is t

l was t=ugn enougn t: jing to get along with the c u er pacple up there.

I Just, you know, taam verk type thing. Ar.d trying :: develop tae ra;:or:

cf the other s. Tift sunevis rs as well as with =y cwn. That would jus sake.1: : tally siserable for myself.

i t

FASANO:

Ycu're mainly en Unit 2?

t i

'AR~ VAN: Unit 2.

(

-(

i 4

?

(

o k.

TMI TAFE 254 54 i

FASANO: Do ycu have any bewiedge that this is si ilar type of c: ara:i:n on t' nit 1 or art the'y different? Tc ycur best kncwledge.

Ycu ucw, I 1

mean if you dcn't knew, you d:n't knew.

If ycu de....

HADA2: The only thing I can mata is an asst:_:icn.

Tna only thing lI 4

i can ass 0=a is that the shift superviscrs ca: are a:.he plant :: day '

I wert either former c:ntrol oca cperat:rs in Unit 1 er they were for=ar shift supervisors in Unit 1.

New I can't. help but tnink that they.

t

(

gained seme experienca in Unit 1.

That's all I'm ;:nna say. There's nc hing really else I can say about it.

(.

C?.YF4 ELL: Hcw abcut ce perfer:anca of surveillan tas:s a-TM. Car you c:=ent on that?

HA D AN: Surseillanca precedurts, I guess, sc:stices I, well, I did a let of them..In fact, I did tne =r.e that caused tne fi s safety features actuation whern I ripped the 11:arna.a fetc su :Iy ts ue invertar and Ics: a DC sucaly.

I :er te:ad that ::c, by the way, whij e i

i e

overycocy else s:ced around, Iceking around like they cien't knew wna.

i was going on. Now generally, the surveillance prc:adurts, we did them and scmetimes they required a change, ycu knew, lika well thers was a'

(

small pr:cedural deficiency or say:e a valve nu=:er was wrong ce :ayce there was a battar way of doing it to ge: the and resul:. 'de <cule f

I l

change :ncse, the 71.Ns would take two ifcansac :: era rs Or two senier 1

licensac :; era. rs unless nuclear safety was involvec, ben i:.cuic C

B y----.----

-_y..

7

(

s 8

.f-TMI TAPE 254 55 t

take 70RC a:;reval. We wcule change :nese, make them rignt. Scmetimes in the perfermanca cf a test ycu c:ulcn't ge-the required resul s anc we'd go back cut with the shift fersman and he would get the pre;er i

results. Sometimes we'd...

l lI

• RETWELL:

Ex: sa me. 0:uld you elas:ra a :n -ha-?

s 9.ARTHAN: Well, an exa= le, the emergency fats pu=:s, running at st /'A!11anca it was a bear.

Every time that we did the surveillance taa ney called f:r a thrust tearing vibrati:n measurement anc it als: :211ec -

for a tampera urs rtacing en tar : earing and called f:r a :er ain 1

, r~

differenita' pressure, su::ica ar tssurs had :: :e :e: ween a car zin

\\,

amount.

'de've never t:ne that das: wnert i: :a=e :u: :ne same way twice.

So we.t:ssed ue :ur hancs and we say, ycu knew, wha-

.e cci We can't get the reference values, we can't ge: the ;rt;er cata. Okay,

(

well never mind.

I'll take tais precedure an: I'll -tr:w 1-d wn at the surveillanca... I assign :ec;1a, ne insarvice ins:ectica :y;e an:

they would evaluata the cata and then ney wculd ::me up with a new sa:

(

of reference data everytime. Andofc:urse:nesurveillancethatwe!

i did would fall rign in;s taat.

I never did uncerstand that.

CRE!'4 ELL: Any otner systams basicas :ne emergency fate ater system "

involved?

l(

(1 t

~

t o

t I

I I

i i

\\

(

TMI TAPE 254 35 l

I HA RTYAN:

I can' t rtcall.

I kn:w that t lo: Of the talanca Of ;lan: '

surveillanca. It wasn't taen specs er talan:a er any ning li<a nat, '

that would go by the wayside. Ycu knew a 1c: of ti=e it wculd cali f:r maybe the sec:ndary servica c:clers to be ta:kwashed and we c:n't havt time for that.

Then ycu just sign it, ycu knew, tar:w it in -he basre:

not c:=;1stad, and it -culd :::e ta:X nLx week t: de it, ycu knew, i

i never get dene.

Thers was a Ic of things like that en the sac:n:ary side. We ner: ally did all :ne surveillanca :na was recuired f r tac's l

specs, but thert was a lo: cf ti=es, ycu knew, I can't really name t y specific ins: Incas tut...fer specific precacurts...tu: I kn w that thers was excap icns and in t.9e excapti:ns they c:uld pa:er thesa away 1

semencv.

I naver did rtally understand.

I ::n't go in fer na-kinc

(.

cf thing.

I figurtd if it can' te c:ne by -he surveillanca pr::adurs, l

you change it so that it can be c ne ::rrsc:1y and witnin :nu se::e cf the surveillanca recuirement er you c:n' ce tnam, you knew, ycu get' the thing. ignt anc then c3 t.9tm.

r i

C?.ETWELL: Hal, art tners any e-her c: erat:rs like you :na art :ncarrec 1

't secut sc=e of these cccuritccas?

I i

i HARTMAN: Yeah.

I wculd imagine that -here's quita a faw that art

(

c:ccarned.

I'm not surt that their attituce is the same as =ine. I' knew that they like money.

I mean ebv,icusly I cidn't quita u: :nere I

fer a higner paying feb and ina: I knew that a ic: them stay ar:und

(

nert just becausa cf the =cr.ey, na :ney'rt afrai: : leavt :ecause,

(

p.

d I

i d

a ; 4'

(

g

(

T.MI TAPE 254 57

(

ycu knew', they like meney se mu:h. Sut e.c: :nly cat I guess cay're a little bit =cre, well, they are just en: cu cf a cifferent mel:.

(

Tney'rs net as, they d:n't say nings the way I say the=.

Theyhave-i be c:ncarr.ed.

I can't really see where taey c:ulca't be c:ncerr.ed.

I have respect fer everjene of the cceraars c: thert.

I have res;cc:

I

.g f:P the su:arvis:Ps ::e because 'aell Oey went tartugn har: Oises but se=euings that they c: I really lese respec f, r.

Tnern's a ::u:le just, ycu kr.cw, I wcuidn't werk win tem for anyming and that's one of ne reas:ns I left.

(

.uA R5H:

'de are ge::ing :: war:s the end of' :ne ta:e.

The di=a teing 3:1.*, so at tais time I'm genna break fer a :cment wnfle I put a new

{

ta:e e..

t

(

I t

i

(

1 g

I i.

4

(

...... : --- 1 n e

1..-

i

.=g**.

1

.t

.