ML20136H477
| ML20136H477 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 11/20/1984 |
| From: | GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| References | |
| SP-I-TMIA-015, SP-I-TMIA-15, NUDOCS 8508200353 | |
| Download: ML20136H477 (36) | |
Text
m' i
~~
~
~
my _ , (
! l9')
s,e ,
1 me DOCKETED -
umc t
t
'85 AUG ~7 P2 36' 4 0Evl't >~
-.4 .,f
. w.in,L.i, k .
y -
, c w;;.c=:
uva; .s . w . wn.,s <
- -.- .i ; * * * *. *. *.!, *- t j an o
- /. 9 /A #
N 1
} p, .s. j o Q * )
h'. ', '
- 4 .$ e .' e- u s $~
i .#
.a=
(
l l
l l79 T/W M l l
\
9 f
., /- r s n} #
4 ar -
gM i .
D:24e - p....
i&/m/n 7 RUCLLAR ElGULR10t1 CONMIS$\0R 0,enRo 399.sf_ona e,s. n fMA ~ /s-
- n. .e fN/ Af/w'y /, w <#
\
- n. s, -
IDENTIFIED StaN_
a,,sicant - RECtivto latervener V REKCTED .
((* 7 W O
Cent's Ofra
- 0471 Ceebactu Othe Witan O/ d 8#N[
nere$ _ $YW
& % Qv %sI Ae /
i.,..,,,..
Inter-Offico Mcmsrandum September 17, 1980 EEL-2794 Date:
Q
% dP =
!a INTERNAL WORK RELATED TO GPU'S h Sutect KNOWLEDGE OF CORE DAMAGE FOLLOWING l ,
THE TMI-2 ACCIDENT File 2359.
< 6..p;J -
~
t Ta Vice President - Generation Locanort HeadqN rthrs2 O I CT > ~~
- a. c. Arnold ,7 3,p,p,g g
.. //
IhaveattachedforyourinformationthreepapersY..h'ichla[
the only items that I am aware of r 1.ated to our understanding of -
the core damage following the TMI-2 acoident. The draft TDR-115 was the initial effort undertaken by John Hilbish. This TDR was never approved or c mpleted. The interview memorandum of Brian Mahler was done in connection with investigations by R. W. Keaten's
- Task Force. I have not provided a copy of the Keaten Task Force Report, which I believe you have. The untitled piece reflects the complete efforts of Bill Behrle, Scott Guilbord and Don Reppert that was undertaken at your request of about December of last year.
This dochment was never formally transmitted within the company, but was prepared in anticipation of other investigations, which did not. materialize at the time prepared.
I have talked with Bill Behrle about the effort he was in-volved in to see if any other material was available. He noted that, in preparing his work, he and the other members filled out a matrix, which reflected the interview records they reviewed and the subjects of interest contained in that interview. Bill
- in searching his records and will forward a copy of the matrix, if found, to you separately.
I know of no other work complete, or incomplete, that would-have any further bearing on this issue.
. . Ac w E. G. Wallace Licensing Manager EGW:bjo Attachments cc: E. Blake, Esq. - w/ attach.
B. Behrle - w/o attach.
k I
i 1
_ GPU Servce Corporation is a subsdary of General Pude UMt es Corporaton, ,
j A*d //00% i// 7 3
~
T_w =__2. _ .
__ m n. ,., s w , n- .
sha_vr '
(' .
4- As\ 'Te}u y $ % jg bgIyppyL 3yc & b &
--<w.w.
- v. .:.
. . , w . . o.
.... - -.... .- -..-...-i....----
...a -... g w
v... -
v a :.s, %s ?A*l.n.a . %;J_,.fa .
- i. .
, r _-
., '%> Pu.- t.e., * '.ru k Y ,_. #g...**?h hW .+
_ [' l ' iy.*x-4y 's, -l-% R k-Sorewh-V:h,..?8Y - -
_F .? .'.
2/ -k M r,.th su a*,. Q.c. -
y Qk.Ia Ekuk<a t 5'.d ~YY-i 1
- . . F'.= L aO_c Pn4a ... . . _
.M9sr -
l p , es ... .
r j _
Ra w - ens w t < s ,.u .
Vn P_) .
p , m m --- r . a .,ss,. 9-s -- - - - --- -- - - - - ---
l w % -, - e a r.c g,.
[ ('
. u .g.D 1.. L, k & P ,..n s.Q -r su ac 0. .
l
%' A L ca. s -
. - ft - A . J .3 ym -
O't. h l erwh O s ** Cr** >dral wr ed [ar/. -
- . L..
i /21 ~ SS'o*f tJ w -
)
l 64% eov-1 8,,4.;,.yy.-l: . t y.5r L,a.,< c -
-. ( A. *ft sm o T_kef L^.".2. ' .' .. _d M _ _ _ . . . _
i i, .
~ A s.4, ,
- u. t.j., .. _... utg.4_c,.p_o_.s ,J ,'8 0 W.a . _. _ __ _. ,
8 C f. - b4 L Un b.8 rr.c_(_f./.*,"L.
~
- f- .
L 6*.ga.n 'cM_i._2l c_p,w. St..l.m:ft.h. '
~
(.. .. __.
- 0 (8 e d .
' bw S % irn~' ' ~ 3. . .,*ij etr4 ND' ' f' i.c..
......,....s,..
.......,..........,,,,,,..,...,.,,,.,..,....,......,,,_,,,,,i I
. c ds/;'i _
l,( _ __ . M . i. . . '. .
,w . .. .L .: _ -_ GD .
4_.c_, ; . ,, ,, . . ., . w, 4 6. , 7,
-+, p. ,. t . , c.~. ... ' . _
- 4 %,n (6s 9_. ' s a qMLse ? '
._4_o . .
1 ....
.. m,...c..
v =wc _m: :a.s:. /_w..,n . .
. f ._.,
f i D w do.b_ M p a_a.a b_ %_t_ M l
. . . T_h_.u. .. _3..4. e_e_k_ _)._._ l
_ _. ___ f, I
A y,g.(
$ F!.A) A ti11*i em- 'g,g m _.,_g_um
. , ,a d JShig l t Q Q Q .
1 kN M.M$1 O Ii ABO i
( M I
i
=
O ir---..---_... ..
~
[ .
Q ~_ _ _ - .. ..
M_..._e .m gg
C v.' ...
. yasA1 worn n
Q. ,_n.
e -- e = f 4 f.r y / M _ y/ w - -. -
==,
. r . . . . -
ak_ . -_- _-.L k e._cu. _ % . Am _%or A. a w w c '.!.1.. *
._ak.mg
- k . Q % i r s u a :t .'
_q ala M W.m _+r&.,. Awes _+k_Re. Em .
.a.naa w e 3.kr.gg<.t..? u )
Dde .O.. , .M -
z s.6.. _ GD.. T+ * + (bh. . .
_6# Gb_Ar A4 NJ 'A ,
. m v- .
. (f u n,.e ,. e, 4 . g 4 a .? ' .
G.s..._ @ ...luso C,re m ?-
4( buaJ Awy - 6 Ay A.tre_).' .
3Pm (3L _.% G& ~ tA +%'- tuk+.a Lk.&
"'" Stfwee .h .h_EFWm
-* %... e L A.. . _ _ _ _ _
_ cpm _ @ _. ft fLa - ,4__ _
gID
_tdatu,.%. W l_ _-
l % M_&_m p j.stu.+I
_ _.-..... - p _. w _,.-au -
_]Mm %..._.tcht_.V9g ', . _ - .. \
l
.-____.. W . w . W .? .. _._ _ __. _ _._...
-_. _ L4 .4 e.A . M A ?
i
~ .;._-
i
. JPm . _.@ __ RC fc.4.hd . da bJ . . . . ...__._ ,
_ Y m . p .6,v 3 _ __
h w_4v s _Lb-k. . . . . _ _ _ . . . _ . ._
3
W u n. .
(.. /afet" M/~5
. ... g ___ . . . _ _ _ _ _ . _ _ _ _ __ _. .._ __ _ _ __.. _.
..._ mom - S --
_. aa.maen Pla.~+ GDir LMak' wW'.! . . . .
- w..y_n . -
g-
? ,
j
_.g .. ._
.. .res
_ _ _ S kA) 21_dM . _ _ .
M CR-N )
- om ..%kt._E. L_. -
_off_.LA. _ f:L c _ _ .- - - _-
O"_
fra wn, L 6 4 um..JrM.. .
O" -
4 Dameup. '.
1( .
. . - . _ _. __. . - . _ . _ _ _ _ _ . - = . ._.
/ _ __ _ . ___. _____ . . . _ _ _ _ _ . . . _ . _ . _ _ _ . _ . . . . .
. 1 __ _.. _ _ _______. __ _ _ _ _ _ _._ _ _ . . . _ _ . . . . _ . . _ _ _ _.. . _..
e 6 _M M_
- N _ .s__ gM _ mgg apg . ...._ _ ..g g m..g. .._.6- _. _.
(
O 9*
f ,e .
e =.*
l l
l 1 e
l I
f . - -.. . . - - - - . , , , _ . .
d
.5
) . . . - -
~
.. . h Q*^. ra_A $_ d d. M ..-. s .. O. p. r ._Omd_. )_. .
_ _. ..___.__e.,_,. ___ -_ __
e
.- C gg gm.eg . .m * @@ m & S m e t ese e em e W.D 99 e
i
. )
)
- e. .us em
.e m.m GNS 6 MM GW WN# M g eem e e e.see.m e e e e W e m 6 i
O9 O
l l
(' \
- s. ..
l l
M %
m em #66 M9M W O I
-e e e e e m e.ee --o m m e. . G GM4.
4 O e me e G
. mm g MGM- MSM MN9 m M. m
. O.
SS*W W9
- O .e + 9eSe eGem o m se e e. e e e em mm 8 MN N..e MM NN8 M6 9 9 W M*S .
G #
6 M M M e ..m.m mem_ ..+m*@ e.m-M*W M M W# 6 489 *.#6 N# .6 6M
.m O e
/ N .mo e.g.m .m og m N 9.-.We. m
- WM 6 6e m e e eme em o m o e ae We W9 M -- S S S M MOD 9 6"M # #. MMW4 W # "#***#
l y.,
. sn f 21 f - - -
- #P e .
CC.
-_ _ ._% & %c_.be I ~...a & m A. e M
- __ _......_ u.rLa e %W._h avx ._% n._4bd . ff.s 1 l
______ Pm .._t o'_?_ . ._
y +. ~. g.. ,. ..e-._
u ra w e-i
- s. - e w f m p _ b ..C.s w _ = _ o a b . h s = _ _ p . m .g _ _
.. __,. % a G ..- +g _t b __
_ _ . - . _:iaw. . .L w 1 A sc., _ __ _ .____
2 i
g-e .
X P_ _%. S 4 _ % --- = = s .. Q & . y + & .4., u _ + h._
ap_e_m.
b J=I 6 S .4 Q _e.n M _ - . \
_ _ _ ______ p ..:s L
- q .e.c 9=- w=_ h w m .g. W g c.4_u.ts.B
~
bTt>%* .Lsh : % 1^
.W..A .M.%
_._ . ..._& cw ' TO __Yw _y i r w. F.. *
.._ _____._ N, cert.cuA .L. > W ........_._ --
) __
_ L4, H;h.r.__-tr_y~ _ E h _.. --. ~ h- ..
. /h _
M.vg . p. 44 e ..%.k.. . _
( (. __ _._ .___. _ W s 4 cr u A ..c e u ; . w r. ,_ v
- h. W _ _ _
1 I
i
. QNM # ~ MM] a- Asr
- 3zrpj
_( .
A -
,b d.a:Embes :sW 4..'
. _ A t- 5 e. h . M w"E. .hr.% . . . .___.'.
_. m.g p. _ ..
kp ..o t.__Lo1.e - ._111on h _
m
__. _. - M _DV*_7_,.2A4v_h i h - _ _ _ -
M _MCedz% 3..C.,A h_..W. .!&_ A p.TSt---
. .__M W .9% . ._ . _. ._
. .\.# _% W ^_ h- $' ._. g ] &_._s % k M .
~
.$ -(5.%..._L.ao-7ed..p..____9.e_
S %.. ph atw._
.('y? p s e- p :- % -_ w .._c-- += f > _ w . % % ._-
p.. ._._.G. f ._\ Set Q m _ _..
.g-_ _ ..
_ . _ _ _ _ hon _1._ b A q & 7 W ..h l
-hs M_Wh s_ . .h l - . _ - . _ _ -
L- -.___.- _. . - - ..._.._ _ ._ . _ . . . . . . . - . . . . . _ . . .
\
(
\ _
. . o.. .
j gpy
., . __ __ . __% L asm. . . . . . _ _ _ _ . _ 53 .
l Q M 9P-
'- ~
%._w_IsN _f M. ._
__ .tacaw w .a__tw =- + M. q '? % ..W_ fa . k. e*~- '
,_ h 6 .w = e _ g ;_ __ % _so,T- , !
_.tw ._T . ; g_: f u q - p e _G e_1 _a..-
_ % "C .y_
He a s._ . _b-. . ! _tw. _
~G .2n~: ,
__{mm .bw.d. _ 4=w _.$ _.k _W= =-4. - S m ?
- de.n..
G% <L a pd)
_.____RPI 4 _-.. ~ p .L a_w a. p w.
l C g*_.5ke _V %h__:._d,,._L-A. A y 4 qc= S.
1 e r - ...< w .. g e . m _ t - . g ... -
j __ ~ _ __ _ _. . c:gw
- p. 6_% _h .h %. .g._:ty _.LU_ h . -
l L( _
E v -s, sac. % ..:hn .% ax =.-:' _G_-lo nm -
b _s u p h . -
Rc_ Q_A-.r, _
.. _._@.~c.2.g'Is_= g , _ %. g y wra. +.s,_k M. d
. _ . . . . T g _ % f.a _9.e. q ; - & k le a_ w y = = - z.
%4 .tL._A.it _ hi.4 _Q_ h_% _ lop. ..
afan
%.+ .k. .r. =s_ sne ..
! .__.__ .__ __ __ .fr EM. >
%N r ..s+w w..e m 'k
% h p .4.-e_%b.w.
. _ _ _ _ . . . . . _ A.hy %.-s2.v.oV . ...
l
. k y , a w .. ______ _ .....___._.__ f 4 Vy _
F._ __
l___*pr_.._G.h(Lz p C . y 14= 4.
- z 1 A b.e A . h . / c . , L i-)
. p* ._
.. . _ b W., p +_. N. % .. % g h L=r_l24_(b.p. _
, /..__..__________....____.._... . _ . _ _ _ .____.- _ __.._
l _ (. _ . _ . - %,-+ .Am.-><*p the r>, :=p u.m m. ..._ _ _ _ . .
o
y -
- 'cg .- . .
_ _ _ _ _ . _ . _ . . _ _ _ _ _ . _ . 3 :tE./.11 _
._ __ ./s ~6 ._ M .. b . b ._ W . b.ak.C .T W ..W.**. -.
1
..__....... 6 4a.s A -- . _tt
_ _ l
\
l 1
I
\
. _2:5". 0. p. .b2.3._.4s. _.o..RA. .A. _ __.
_kMiM.y .b_
, ..1 2oav.p - !.L p: ~ is3 .ha .Ast.a h n _RA.W. '. l
_ ._- s,. e p _ _. . . . .__ . , .
g.*. ..
h.. ..g. 9_e = p a. u._ ... .
l L..
., . a . r,M.._ . ).
___fotA k 4c - p ~2.a. - % set P
_ . _ _ ._. b .2.m==. p .. w w h=ths._ % = p 3_ts a f" h...= M . _ .
__- ___ %ch.A m . h.k4_ M = '. . ^ = M W . f .- - r ~ : _ -- - ._
. *": , & YM M %
- a
)
.C I
. p. _.,*1..-.c
. Q._p g ...J. q-
...._.y'.. ~0.- =
. - ~~
I .. u . a_% .g._4.s;.o.g**. . W... _2._.3.4 . . __. ._ _.
._ t _99.3 btthb_sek.._e _q . _= . .de te A. {f .
- c. w ,
, , w <i.e g i
. _ _ _ ._ W _4 ew.,,,fL. - w.w .2+e_w. _%_ . ..
p ._ _ _ __._. . _ . _ ..
. _- __ ._\O. \s a ju<~.t .% W.s . .
k .is l= - % Y-- _ . . . _
._.. - . . _ . . m__ _.. .. a... n_.w. . . . _
w_ . . .* v ... . . . _e_ . _ . . . _ _
--a --
. - - . _ .. . . . - y a w s n,_4 m g a_, . . . -- -- . -
. ._ _. c e f e .. W G .2.c.. o w . 6 % _ _.._ _ ._ _ ..... _ _._
i _.
- s. . _ ..
1 - .... .w wt 9 .w t.B 4/4. p' .... __..._...........___.__._
. , - _ - - , - , . . . - - _ . . , - . - - _ _ , ,- - --- - - - - . - + , - - - - --
e .. . I p h o cT54 A % 4euaikdL-@^:-'- M
. h % Q % M w *. W h B_..Sd _h W..% - A ,: M ._t.< "
)
. . . . _ . . _ ~4ad. EN. b K . e '= pa A .Ec 6 p_.W - _ .
~
. . . _ _ .. . .G * .6 .c.T.SC . be.<A d _ _ . _ ___ _
n m..._. .'__.r._.
.R ~. ..ma_s_exsc,_A_9. ... .e.
_ maA2 M._
M _a_os. sci _w.% eew_%'- hd_A.S_wq ; .. ,
7r y" .....__.....__ w . n._ m .c- _ \
. _ . . . . . - .% .~. .__, 00 .F___ __
- p. ,. ,.g _ _
l..A s..oed.@._47.q y _ ", , :
- W _s _g,& . _w . -
-~' ' **
- S r w u _ % e -L._ _ __
4_h %_, p L.
_-. . h r '.. .e_.o.N. .
%- -2 X_m a ~O
. . u_.v ww t-M .w.L.a mu.6.:. .
, g- --
.w-p u , ,r u,. p w y w.-. - _ _ _ .
I '
,/ . . .
I
,, h ,' '. ' C JY<o/
I *
,I. Corn Dam gc/Fu,1 Uniovnring From a reading of the depositions and interviews, it is obvious that both the licensee and the NRC knew on Wednesday, March 28, that core dam-age and fuel uncovering had occurred. However, the severity of the dam-age was not realized until the results of the primary coolant sample were received on Thursday, 3/29. e 1
l A. Licensee's Knowledge- !
On Wednesday, Chwastyk (Shift Supervisor) was aware of high incore Thermocouple readings, assumed there was some core damage, realised magnitude of problem when it took 50,000 gallons of HPI to fill the 88,000 gallon RCS, and knew there was core damage ;
from the explosion in the building. Frederick (Control Room Operator) was concerned about possible core uncoverage and twice suggested full HPI. Frederick and Fandt (Control Room Operator) later initiated HPI after PORY isolation, but were told by some-one out of the Shift Supervisor's Office to secure HPI. Flint (B&W Physics Test Coordinator) believed core had been uncovered or had voiding due to increase in leakage flux. Flint told 1togers' this around 10:00 a.m. (Rogers does not remember this, but Kunder does,),
. t Flint thought extent'of damage was a role,ase of gap activ-ity. Hits (Shift Supervisor) realized failed fuel at 6:50 a.m.
when'someone yelled 600R/hr. In sample line (really was 600er/
hr.). Kunder (Unit Superintendent - Technical Support) felt l
the core had been uncovered due to the impression he got from I
Porter that some of the core contained superheated steam.
Kunder also said that when the reactor coolant pump pumped ,
1
. . l l
(l
l 1 steam, everyone was concerned that the core might be uncovered. .
i Logan (Unit 2 Superintendent) knew the core was damaged at '
l 6:40 a.m. Mehler (Shift Supervisor) felt fdel failure occurred - l at 6:45 a.m. when radiation alarms came on, and, prior to this, l
wasn't sure whether core was completely covered. Miller (Station l e
Superintendent) realized we had failed fuel (gap activity) at 7:05 a.m. when he entered the control room. However, he has -
made several conflicting statements about core coverage. Porter (Instrument gngineer) learned of failed fuel when the site emer-gency phone calls were being made. Forter recalled no conversa-tion with anyone regarding core coverage, which conflicts with Yeager who said he discussed this with Porter around 8:0d a.m.
I Rogers (B&W Site Manager) did not believe core had been uncov-ered since a large volume of water did not come out of core flood tanka when they were floated on the core. Ross (Opera-
- tions Supervisor - Unit 1) knew arount 7:00 a.m. that we had failed fuel, but didn't consider whether core had been uncow-ered. Seelinger (Unit-1 Superintendent) believes that it en-tered his mind on Wednesday that the core had been uncovered.
l i Yeager (Instrument Tech.) cold Porter around 8:00 a.m. that
! '. he thought the core was uncovered based on high thermocouple L ,
- readings. Fortar.does not recall thie. Neve (Shift Super-
' visor) did not think (on Wednesday) that the core was uncovered.
- . Arnold (Vice President - CPUSC Ceneration) first become con- , j
. 1 cerned of cladding / fuel failure around 8:00 a.m. when he re-
< ceived high re'ading on the done monitor, and was continuously 1
- . j concerned on Wednesday as to whether or not the core was
=
covered. He was first concerned wich 9.ignificant fuel damage j l
a 1.--_-_._.- . . _ . _ - _ _ _ __ _- , _ _ _ _ _ _ _ _ _ _ _ , _ _ _ ___ _ __
a _
on Thursday morning. Creits ,(President - Met-Ed) cold Dieckamp .
(President - CPUSC) at 9:30 a.m. of Arnold's concern of possible Herbein (Vice President - Met-Ed 'q desage to the fuel assembly.
- Generation) talked with Arnold early Wednesday morning about l
possible fuel damage, and concluded after talking to Miller He believed later in the morning that we had some failed fuel. .
e
=
the core had been covered when he received the results of the core flood tank level change upon floating. Herbein said that l
up' until Wednesday evening, the plant staff believed the core was covered (conflicts with statements of numerous plant staff members).
He received the high reading from the primary coolant sample on Thursday, and briefed the press that we had failed fuel. ~
r
- 5. NRC's Enowledae .
4 Callina (site) knew on Wednesday that there was core dam-4 i
age, and believed on Thursday that there was gross fuel damage He based on the 1,000R reading of the primary coolant sample.
said that ,whatever the licensee or NRC on-site team know, it I
was ismediately available to Region I and to NRC headquarters.
Baunack (site) said neither the licensee nor the NRC realised l
the severity of the problem until the primary coolant sample
' was taken Thursday night. He said that th a NRC was totally i .
aware of and infor=*4 of *werythina soing on. Mosely felt on Wednesday that there was core damage, but didn't know it was extensive.until the primary coolant samples were obtained .
\ Stello (Bethesda) become aware of extensive
. on Thursday.
core damage on Wednesday, and felt it should have been aware to everyone.
Both Vollmer and Stallo realized on Wednesday i
that the core had been uncovered. .
l
s .
Maggie Reilly (PaBRH) to'1d the NRC Regi,onn I at 10:45 a.m. .
I that B&W representatives believed that fuel assembly gap activ-
- Ity had been released, but that theE had been no fue'l melting. 2 !
Hits (Shift Supervisor) vaguely remembers a conversation with Stello on Wednesday about the core being uncovered or in a super- ,
, . e
),wk *,.'. . L M. u%'. so.* *
- A p
- 4 * ~*
heated condition, and about the floating of the core floodjtold - -
Hits that this would not guarantee core coverage and Hits no-3 tified Unit 2 between 4:00 and 4:00 p.m. of Ste11o's concern.
Hits does not remember who in Unit 2 he spoke with, and Miller was never aware of Ste11o's concern. Runder (Unit Superinten-dent - Technological Support) notified Haverkamp at NRC Region I office of failed fuel and curtailed HPI at 9:35 a.m. on s -
Wedneday.
\
. O G .
l l
)
i .
l . <
II . , Pressure Spike / Hydrogen Several of the licensee's employees were aware on Wednesday of a real pressure spike, while others were not. A few of the licensee's employees were aware of an instrument indication of a pressure spike, ,
but they attributed it to an instrument or electrical malfunction. i 1
Other employee's who heard loud noises (thuds) attributed them to the
. 1 cycling of ventilation dampers. Those who were not aware on Wednesday -
of an actual pressure spike learned about it over subsequent days.
Two of the licensees employees (Chwastyk and Mehler) who were aware 6f an actual pressure spike may have believed on Wednesday that it was due to a hydrogen explosion. If they did then believe this, it does not 1
appear that they communicated their belief to anyone. The remaining asployees who were aware of an actual pressure spike did not associate it with hydrogen on Wednesday.
T'he earliest that tho' NRC appears to have become aware of both the pressure spike and the hydrogen explosion was Thursday, although one of the licensee's employee's (Chvestyk) says he mentioned the possibility of an explosion to an NRC man in the control room on Wednesday, and i, 4
another employee (Mehler) says he pointed out the spike to an NRC man in the control room on Wednesday.
L A. Licensee's Knowledge
- ?
(.
Chwastyk saw pressure spike occur and thought it was an gi f
instrument malfunction until the spray pumps came on. Chwastyk i l
i<
said he put together cycling of block valve and pressure spike. - -
!l and told Miller that spark from valve caused explosion and i' asked Miller for permission to draw a bubble and repressurine l
, (Miller does not remember this). He assumed the explosion ,
was caussed by hydrogen, but 'may not have related this to Miller
l -
i .
Jie assumed the explosion was ,
-(Miller does not remember this).
caused by hydrogen, but may not have related this to Miller t
(Miller does not remember this). Frederick saw pressure spike He saw pressure come occur but did not know what caused it. i back down to normal, perceived no consequence, and continued On Thursday he learned that the pres- a on with his functions.
amane sure spike was due to a hydrogen explosion. Flint heard a double thud and thought it was the ventilation dampers cycling.
The high pressure instrument indication was believed to be an He had no reason to electrical or instrumentation aalfunction.
' believe there was hydrogen present. Kunder was concerned (he's not sure if it was Wednesday) about energising electrical equip-He I ment due to long-term generation of hydrogen from aluminum.
uma not aware of the pressure spike until Friday. 14 san heard a doise and was told (thinks Miller told him) that it was the ventilation system cycling. Marshall knew of pressure spike 4
indication and said the consansus of opinion was that it was due ,
~
to instrument or electrical fault. Mahler said he saw pressure
~
! spike and that he told Miller and Ross in Shif C Supervisor's l Of fice of spike (Miller and Ross don't remember this; Miller and I Ross were in the Control Room at the time). Mehler said in his 4 ,
August statements that he discussed the possibility of a chasi-cal reaction with Chwastyk and that hydrogen could have been Later, mentioned (Chwestyk does not remember this conversation). -
ia October, Mehler said that hydrogen was definitely not con-sidered on Wednesday, Chwestyk did not mention hydroden on Wed-nesday, and that he (Mehler) did not hear of hydrogen until it I
1
(
_ _ . _ _ _ _ _ _ . . _ ._ _ .- _ _ _ _ _ _ . _ . _... __ __ .I
came cut in tha press.
- Mahlor -
said ha' thought M111cr gave crd:e on Wendesday not to start actors (Chwestyk, Miller, Rose5 and
$7 ewe all told him in discussions that it was on a later day).
~
liiller heard a thud'and was told it was the ventilation dampers cycling and did not know on Wednesday that it was a real pressure spike. He learned from Bill Love on Friday of hydrogen detona- ;
tion. Porter did not know about pressure spike until Friday '
i l morning when it became " general knowledge " and later that morn-j ing was asked by Miller to Icak at recorder charts (Miller s
remembers this). Boss said he and Miller had their attention called to the pressure spike by Newe (M111 der doesn't rer:.sber this, and kewe doesn'c remember discussing it with Miller).
He~
said they figured i: was an electrical malfunction or else they ,
didn't understand it, but ,it was not considered important because pressure returned to normal. Seelinger heard about pressure spike
+
"I l some time after Wednesday. .Ne,.ws saw presure spike, discussed it with Chwestyk and Ross (Ross remembers this), and they concluded
.,s l
i it was some sort of electrical transient. Newe did not percieve on Wednesday that there was hydrogen in the building, and did not learn unt,11 Thursday that a hydrogen burn caused the spike.
Arnold did not learn 'of pressure spike until Friday morning, and d as not r3 member shether he thought the then-existing bubble
-
- e.s v -
contained hydrogen. Herbein became of pressure, spike sometime betweenThursdayandSaturday,andthec'o'n[urfdIanalysisby Bill Lowe indicated to Herbein that we had had a hydrogen burn. ,,
B. NRC'S Knowledae Gallina in his May 7 interview said that, on Thursday, the NRC believed the pressure spike was hydrogen but that Met-Ed did I
- . . . .i
o not. Later, in his May 31 interview, he said he first became '
aware of Wednesday's hydrogen explosion on Friday, and that it
- was late Thursday or early Friday that the NRC inspectors in ~
the Unit 2 control room were told of the pressure spike and a possibie hydrogen burn. Baunack said that the primary coolant sample received on Thursday indicated serious core damage, and '
- 4. .
prior to then Noone had realized that the existing bubble could -
i.
be a hydrogen bubble. Higgans said that the control roca una so busy that neither he not plant management realized until Friday, when they went over the charts, that a pressure spike had occured on Wednesday. The NRC Incident Message Forms in Region I show that at 3:12 a.m. on Friday, the NRC completed their experiments and estimated the size of the bubble, at 5:55 a.m. they concluded that it was a hydrogen bubble, and at 9:00 a.a. Raymond (NRC-site) cold Region I that hydrogen was what had caused ths pressure spike (detonation) on Wednesday.
Chwestyk says he told an NRC man on Wednesday that he thought we had an explosion. Hahler recalls telling an NRC man in :he control room on Wednesday of the pressure spike.
- g O O O e
- g
\
- 4-
= - - - _ .. . -. __. .. ._ _ - _
e i,
III. In-Core Thermocouples A. Licensee's Knowledge The in-core termocouple readings print out on sheets of the ,
' control room computer. On Wednesday morning, many of the thermo-4 couple readings were off the high and of the computer's scale t
I (over 700*F). In an attempt to get actual temperature readings, g several maintenance and instrument people went to the cable room where the thermocouple lines fed into the back of the computer. Of the four thermocouple readings taken with a fluke thermometer, two
' were unusually high (above 2300*F) and two were unusually low I
(slightly above 200*F). The Zeergency Director (Miller) was in-formed of these readings. Most employees who b re aware of these .
four readings found them confusing and inconsistent, and therefore ,
unreliable. One technician (Yeager) said he told an instrument andcontrolenginbr(Porter)that,basedonthehighreadings, he thought the core we's uncovered. A second set of readings ;
1 i
taken with a digital voltmeter at about 8:50 a.m. confirmed the 1 I
i
' first four readings - some were extremely high and some were en-tremely low. While the presence of low readings was pussling.
l the fact that the second set also contained extremely high !
! I I readings convinced the other technicians of probable core uncovery. l .
However, the instrument and control engineer (Porter), who was serving as the liaison between the technicians and the Emergency
. Director (Miller), forgot that the second set of readings had been taken. Therefore, the Emergency Direcott may never have
~
been informed of the second set of readings Mn the engineer #s!
so conclusions,.' ,Whereby he might have determined that the core had been uncovered.
9
_ _ _ _ . . _ . _ . _ . . . . _ _ . _ . _ . _ _ _ _ _ _ - ~ _ _ _
' - A similar breakdown in communications occur d Wednesday ,
morning involving the B&W site engineer (Flint) and his liaison .
(Rogers) with the Emergency Director (Miller). The B&W site
- I engineer reviewed the 'in-core thermocouple computer printouts j.,?
(most were off scale) along with other information concerning ei the course of events, and between 10:00 and 10:30 a.m. con-
- cluded that the core had been uncovered but was no longer un- ,
a covered. The B&W site engineer recalls informing his liaison of his conclusion, but the liaison doefnot remember such a conversation. It therefore again appears that the Emergency Director was not informed of an engineers evaluation of the
- b-in-core thermocouples which ley to a determination that the core had been uncovered.
In any event, the Emergency Director (Miller) knew of ,
the first set of readings where two reading were off-scale high. His liaison (Porter) with the technicians told him that if the thermocouples had gotten really hot, they might have melted and formed other junctions whereby the readings would be unreliable. As a result, the Emergency Director has* stated that in in-core thermocouples were so hot that they scared you and from that he knew we were super heated. .
. to However, he had also stated that due to the wide disparity in the few thermocouple readings he received, he did not .
believe that any of the readings were reliable and he did 4 '*
, not pause to consider their significance.
- 3. NRC's Knowledge 5
The director of the NRC's TMI Support Task Group and acting assistanh director of NRC's Systematic Evaluation ,
Fregram (Vallscr) knew cs Wedn:sdcy that tho tempercturo.' in tho i .
hot leg above the core were higher than the saturation tempera- l tures of the liquid (and concluded the core wa's superheated and
- therefore had been uncovered). The director of NRC's Office of Inspection and Enforcement (Stello) has testified that at about l l
4:00 p.m. on Wednesday, when he was at the Incident Response ;
Center (IRC) in Bethesda, he was informed that the computer ,
in-core thermocouple data was mainly questions marks, from -
which het 1) correctly inferred that the temperatures were off-scale high, 2) felt it would be prudent to believe that the thermocouple readings indicated superheated steam, 3) felt the licensee should believe that the core was uncovered, despite pressuriser level indications and 4) called the TMI-1 control
. room and told them that the operator should understand clearl'y .
that the core is uncovered because the temperatures were too high.
However, the NRC tapes clearly show that it was a T-hot temperature (rather than thermocouple readings) of 550 degrees with a pressure of 450 pounds which led ,stel'io to this con-clusion, sad that he did not know at this time that the in-core thermocouple readings were off-scale high.
The NRC tapes also.show that at 10:15 a.m., Haverkamp
- . (NRC-Region I) asked and was told that the T-hot was off-l scale high (over 620'), and at 10:30 a.m. he wa's told that T-cold was extremely low (about 220*). At 2:15 p.m.,
Capthon (NRC - Region I) was told that the T-hot was 600'. ,
- p. .
The whole center of attention by the NRC with regard to $~ -[ aM** **
was T-hot and T-cold and the licensee was continuously re-quested by the NRC to provide, and di)1 provide, this l Information.
i 3-
,, - , - - - , , - - -,_n - - - - - - - , - -,,,w ,-e, -
.r---- - - , - - - - - - - - - - , - , - . -
Although Wilbur (NRC - Bethesda) indicates that it was I
Wednesday morning, our reading of the documents show that it ,
i was not until around 4:05 p.m. on Wednesday that Wilbur (or ,
anyone else from the NRC) requested the licensee (Hits) to get in-core thermocouple readings. At 4:10 p.m. , Wilbur was 4
told by the licensee (Hits) that the computer in-core thermo- ,
a couple data was printing question marks. ,
The licensee (Bits), unaware of ths readings taken Wednesday morning in the cable room then told Wilbur that .
the question marks meant that either the computer point was Nessed up" or that the sensor line was broken. The licensee (Hits) also stated that, we were trying all of the in-cores
- l to see if any of them would print, and that this would take some time.
At 4:30 p.m., Riggins (NRC - site) commmunicated a T-hot temperature of 580* to Wilbur. At 5:15 p.m. . tad 5:65 p.m. ,
"i
'l?
Riggins again communicated this temperature to Wikt (NRC -
Bethesda).* Higgins coussunicated T-hot and T-cold temperatures to Witt and Haverkamp (NRC - Bethesda) at 6:00 p.m., 6:12 p.m.,
~
j 6:45 p.m.', 7:00 p.m. (T-hot 560') 7:10 p.m. (T-hot 557'),
j ., 7:20 p.m. (T-hot 560*) and 7:36 p.m. (T-hot below 520'). '
At 7:56 p.m., Smith (NRC - Unit 2 Control, Room) told Cagliardo (NRC - Bethesda) that they were still unable to read any of the in-cores and that they would still try to
{ . get some to print out. A reactor coolant pump had been l started at 7:46 p.m. and at 8:07 p.m. Baunack (NRC - Unit 2 l
Control Room) reported to Cagliardo that the in-core thermo-couples were then reading with a high of 611' and a low of 1
l
1 254*, but that these were random readings and probably useless.
Around 8:45 p.m., Baunack told Wilb.e,r that T-cold was 320* and l
. l pressuriser temperature was 520*. At 9:15 p.m. Beunack told headquarters that the in-core thermocouples were not realiy reliable.
In summary, the NRC concentrated on T-hot and T-cold temperatures throughout Wednesday, and even *when they got I
actual in-core thermocouple readings they do'ubted their re- ,
liability and usefulness and did not utilise (them for detern-
\
ining system or core conditions. It was not until 1:30 a.m.
on Thursday that, the NRC began to give credibility to some of the in-core thermocouple readings and plot them on a core grid. However, Raymond (NRC - Unit 2 Control Room) in re-laying this data in Bethesda noted that the high numbers (617*, etc.) were questionable since the engineers had told i
the operators, who in turn told Raymond, that some o( the thermocoup,les were probably damaged through the transient.
At 7:30 a.m. on Thursday, Raymond relayed 38 of the 5.7 pos-sible in-core thermocouple readings to Bethesda (the re-maining 14 were unavailable). At 11:00 a.m. on Thursday. .
Raymond notified Bethesda that he had notified the licensee
. . that the NRC had )een tracking the in-core thermocouple
~
readings and that Met-Ed should do the same and either verify or discredit the in-core thermocouple readings. . .
It therefore appears that even by this point in time the NRC was uncertain of the validity of the in-core therscouple readings which were actually printing out, and was not
' relying on. %tuo to assess plant conditions.
5- .
, , 4 ---- --,,- -.- --- - -.-. w e,y- - - - - - , - - y , y-- -
g --w1- - - -- e-- --
IV. Calculated Dose Rate of 10-40 R/hr in Coldsboro -
The highest dose rate which was calculated for Goldsboro was 20R/hr. ,
This dose rate was a predicted dose rate, based on a hypothetical worst- ,
l case situation where one assumes a containment building pressure of 55 j pounds and a leak rate of .2% per day of the containment building volume (2.3 million cubic feet). At 7:44 a.m. on March 28, the done monitor ,
(HP-R-214) was properly read and the predicted worst-case dose rate was ,
calculated to be 20R/hr in Goldsboro. Thereafter, the wind speed in-a creased from 2 to.,3 miles per hour, resulting in a correction of the predicted calculated worst-case dose rate from 20R/hr to 10R/hr for Goldsboro. The corrected calculated dose rate of 10R/hr was then com-17 municated to Dick.Rubiel (Supervisor - Chemistry and Radiation Protection) at 7:46 a.m. However, over the next few minutes, actual field data was received enabling the engineers to calculate real-case dose rates as e:.u. L,s.-
opposed to the hypothetically predicted worst-does. cates for Goldsboro.
et By 7:50 a.m., the real-case dose rate for Goldsboro was calculated to be t- 1 millires/hr. As a result, the 10R/hr predicated calculated
/10R/hh worst-case dose rate was high by a factor of 10,000 Var /ht;/and was not reported since it was negated and made meaningless by the 1 millires/hr calculated rea'l-case' dose rate. The 1 millires/hr calculated reat-case
, ' i c.I.
e i dose rate was reported by Dulld to the State around 8:00 and to the NRC Region I office sometime between 8 and 9:00 a.m.
i There has been some belief that a 40R/hr predicted worst-case dose rate was calculated. This belief resulted from an error made by an
.p i- . .
engineer.in transforming the predicted calculated worst-case dose rate from his Off-Site Dose Calculation Sheet to his Radiation Correction Summary Sheet. As noted.above, the initial predicted worst-case dose rate for Coldsboro was calculated to be 20R/hr. When the wind speed changed from 2 to 4 miles per hour, the engineer corrected'the 20R/hr 6
e-=w
, ._ --. - - . - - . . , , - , , _ _ , -- _ - - - , . - . - , y
~
t .:- .
dose rate to 10R/hr by placing a 1 over the 2 (IO). In transferring this number to his Radiation Correction Sussaary Sheet, the engineer misread the 10R/hr entry as 40R/hr, and wrote it as such on, the Radiation Correction Susumary Sheet. In June of 1979, the NRC Inspection and Enforcement group interviewed the two engineers involved in calculating the dose rates on a
March 28. These engineers stated that a predicted dose rate of 40R/he had ,
1 been calculated for Goldsboro. Their statements were based on the above error made in transferring the 10R/hr figure from one sheet to another.
In July of 1979 Tom [asig (NRC) was notified by one of the two engineers i
interviewed in June (Mike Benson) of the above error. However, Mr. dalg
! apparently did not relay this information to the appropriate persons, for
- RJRRG 0600 was published witt. the 40R/hr dose rate.
In any event, it is clear thser the engineer correctly read the done monitor, correctly calculated the predicted worst-case dose rate for l
l Goldsboro at 20R/hr, correctly adjusted this figure to 10R/hr due to change in wind speed, and correctly calculated the real-case rate dose .
rate for Goldsboro to be less than 1 millires/hr based on actual reported field data. The 1 millires/hr dose rate negated the predicted worst-case dose rate, and was comununicated to the NRC as soon as they returned i our call. There were never a calculated dose rate of 40R/hr for Goldsboro.
4 i
e e e
9 e
4 4
9
- - - , , , - - - , . . . , , . . _ . ~ ,,. -_.--.,--..w.m~ ,g s._ - ~ _ - _ , , . - . . - . , . . - . , - , - - - . ,. , 4__ , ..m.....-e. -
, . . . ,.- u4 tu6.t
- Int r-Offico M mtrandum -
.i
.ciece.
January 29, 1980 Interview with Erian Nehler k e Service i
o
- l
'
- I MD10 TO FILE I
. i M
As part of the TMI-2 Accident Investigation I met with Brian Mehler, .
Shif t Supervisor, to discuss two items 1) his perception of the ex:ent of the core damage on March 28 and 29 and 2) the stat ts of the plant at the time he reported to the control room on
! March 28 sad his subsequent actions. . The following is a sussary
- of these discussions.
- Extent of Core Damese Brian stated that he realized the core had been damaged as soon as he saw the widespread indications of increased radiation levels throughout the plant. He associated this, however, with release
- of the fission gas from some fraction of the fuel pins. He stated that to the best of his recollection, he 3id not recognize the true
- extent of core damage during March 28 or 29, and in fact, had little time for such considerations because of his other duties.
He also was not svare of anyone else who recognized the true extent of core damage during that period. He suggested that we have a i
similar discussion with Jim Floyd.
Control Room Actions .
4
) Brian reported to the Unit 2 control room at approximately 6:00 a.m.
' on March 28. He received some briefing from the night shif t which I apparently include'd cons'iderable emphasis on the fact that the ETV - 12 valves had been erroneously closed. He then inspected i various plant parameters and recognized the fact that coolant system
. pressure was abnormally low. He inquired whether the status of the l pressurizer heaters had been che:had a:.d although he was told that it was, decided to dispatch a foreman to double check their operability.
He then called up from the computer the temperature readings on the
- tailpipes downstream of the pressuriser relief and safety valves.
4 These readings showed that the temperature downstream f rom the FORY was significantly higher then the temperature downstream ,
of either of ,the safety valves. Brian correctly deduced that this - -
indicated that the PORY was still open and took action to close the PORY block valve.. Subsequently he was involved in various support -
activities which we did not discuss in detail.
]L w .
- n. u. use.n CC: R. L. g g. L. Williams / T. L. VanWitback/ E. G. Wallace/ T. C. Broughton e
. - . ~ , - - , , -, .--_.,. --- . -- - -, .g.-.- . - -_ , - -n.-.--,
Nf:,C'/3 .
o
!4.h .* D .3i. '
'/W 7FI-//S"
.. ,- , .. i ,
- DrvttoPMDrf 0F UNDERSTANDING t
l) /Ap**f s ** s
- I The object of this task is to re-create, as best as possible, the l chronology and an on-going status understanding of the TMI-2 accident '
as it developed.
- It is obvious from the sequence of events and the resposse actioss
, taken by the operators that perception and understanding of the transient were changing, especially in the early hours. '
- I,evels of complication in as'sessing this developing understanding are introduced by the various chcation interfaces taich were established about the data source. These interfaces provided data (as l well as on-going assessment) to an increasingly growing number of people I
who int. urn contributed to what may be regarded as a cosmoon understanding of the accideas. .
As time wore on the number of coma.asicatios interfaces grew s
geometrically,and became so intertwined as to give rise to an apparent cosmos pool of understanding. Consequently the task of unraveling the growth of un'dersta'ading becomes increasingly more difficult as more i
sammunication interfaces were established (later in time following the accident).
Necessarily then, the subject assessment wast be attacked starting with the TMI-2 operator's understanding and procede through that of the I * '
site ma,nagement, Met-Ed management, GPU management, the GPU response team, and the? industry advisory group. Further the peripherical
, t m 1
I-I
}
a.- , --- , - - . - - - . = - . . . - . , , . - . - , , . - - . . ~ . . , , .- . . . _ . . . . . - . , ,. , ~ . - , . - , , . - - _ , , - - . . _ - . .-,.-- -. ,,,,,,,.-- ..,_.
\
i branches of BGT and NRC understanding could also be examined. Figure !
l 1 attempts to illustrate the challenge and the scope of the above.
[gct( An analysis of the perception ofsignificant problems and growth ,
i of knowledge during the event can be divided into three areas A) Efforts to maintain control of the plant during the first fear e
hours, 3) Assessment of the radiatios emergency, and C) Perception .
M of the non-condensabia gas bubble in the reactor vessel. .
A. Efforts to Naintain Control of tihe Plant , . ,
Durine the first Four Hours of the Event. the main concern of the operators was to bring the primary and secondary systems to a stable condition. Several key factors should be discussed to focus on the basis for operator actiosa. Details used to formulata each of ,
these hay factors were aerived from interviews with the shift supsrvisor, i
shift foreman and two control room operators.
- 1. Pressuriser Level Indicottos - From very early into the -
transient, operations personnel were very concerned with pressuriser level indication. .
Within five seconds 'after the reactor trip the operator had started a second aske-up ptzsp in anticipation of the expseted rapid decrease in '
s .' l . .
level never occurred, and within six minutes the.pressuriser level was off s. ale high. The operators felt they had caught the expected level i
decrease with increased high pressure injection. The major concern of the operators at this point was to not take the R.C. systes solid.
Based on high level indication and concern of taking the system solid, the operator bypassed Safety Injection, stopped NTPIC, and throttled l
. 1 l
( -
Lihahy EvenY M T u s k o M eau;' h, .
n i
1r
- pa6s e n; g4 hiih g,
Seleandlad. a '
a
! Aqeuss 2 5 a U .
- if
( 1}I SiehM ,c : Gy,6C l
, Fet appt .
! NRf if 3 h h -
5
. 7 p y gg,
. . h.le b - Ed.
i 7 7,,%
Eaaagewer, y
. /
694 Faq uec rquee i.. b e ica b %!u ues t wha ca4te ne belopued o. L*esh%,
,y--..---.--,r - ------+--w-- - - - - - - , - + - - - - - - ------r-----m = -
t- - - -- - + - - - -i-*- --
i
- . 1 l I-( the W V16's. At this point RC pressure was still decreasing, and *
! judgments were made primarily on pressuriser level based on es,ther past -
I experience during feedwater transients.or reactor ' trips, training, or procedural guidan'ce. -
1
- 2. Failure of the Pressuriser Electromatic Relief Valve to C1cse = #
ring the initial reactor coolant system pressure increase due to the .
i ine trip, the electromatic relief valve on the pressuriser opened, I l designed, at 2255 psi. After the reactor trip, the valve failed toislose as pressure decreased through 2205 poi, although the operator did verify that the valve indication did not signal en open valve.
nis valve remained open for the first two hours and tueaty-two i M aue== of the event. Mas, the control room operators had failed to i resognise a constant loss of coolant e rough the opss relief valve for
( that period. Ca at least three different ossasions, the operators 1
checked the computer output for the theauecouple bands on the relief i valve discliarse piping to determine whether the valve had properly
- seated as indicated; however, the computer data was misinterpreted and the blosk valve was not shut for two hours and twenty-two minutes.
Readings from the the'unocouple were in the range of 2308-2808 na '
operator judged these valves to be quite low compared to pressuriser temperature (approximately 603 8 ) and therefore concluded that the electreestic must be closed.. D e operator did not realise that.the
! temperature indication was from .a thermocouple strapped to the outside
- of the discharge pipe and based os heat losses, readings in the range *'
! of 250' were sa indication of art open relief valve.
! I i I
~
. l l
i ,
j
.I
.. + .
\
During the tims that this valve was open, many other indications ,
of a loss of costant accident were present in the Control Room, such ,I as rapidly falling RC pressure, rapidly increasing RC drain tank pressure and temparatu're, increas'ing RB sump level while both s up 7 pumps running, increasing EB temperature and pressure. Althosch '
. . ises A these indications of a,10CA existed, the operators did not associate ,
mum l thus with the stuck open relief valve. The operators continued to focus on the initiating event (loss of feeduster) and attempted to deal with the consequences of that event on the condensate system and the turbine heat sink.
- 3. Esactor Costant Pass Operation - At one boer and thirteen 4 sinates into the event, two Reactor Coolant pumps were tripped due to observed " flow fluctuations" and allouable NFSR requirements of oper.
ating four RCP's. Approminately thirty minutes later the remaining ;
two Reestor Costant pumps were t' ripped based os siellar concerns. .
Approximately two minutes after the teactor Coolant pumps were
' . tripped, the operator bases to raise steam generator level from thirty percent on the start-up range to fifty parcent on the i operating range to further induce natural circulation. Within the -
i next thirty minutes RC hot les temperatures were increasing to off 4
scale (greater than $208 ). It was realized that natural circulation j
1 was not occurring probably due to a steam bubble formation in the f
A loop (the 3 loop was isolated). Attempts were made to again start a RC pump to force circulation through the core; however, setempts -
were aborted due to pump motor low running current. (The 23 BC ptemp
\ i i was run for a period of about 19 minutes some 75 minutes after the l
last pany was tripped.). The 1A Rasetor Coolant pamp was started
{-
f 2
_..,.._..,._____,____..._.,_.___-.---.,__.._,._~.,_.._,,..._...-,_....-,.__.__,,_..m._,_.._,.,,,
. ~, 3 I
approximately fourteen hours after all pumps had been stopped.
- It appears that the actions of securing the Reactor Cosiaat ,
p g s were taken based on minimizing equipment damage (i.e. to prevent psamp cavitation and protection of pamp sesis) and not based on an understanding of what was occurring in the primary system
- e i
B. Assessment of the Raitation Emernency - The growth of .
knowledge in this area developed rapidly apprawinetely two hours and forty-five minutes into the event, as the first radiatios .
monitoring alarms were received thr~ f -a the plant. This know-ledge was accuentated quickly, and used effectively to determina acties levels according to the radiatios emergency plan. Although the extent of fuel failure was not realized this early in the event, the perception of the significance of the radiation monitoring
(- system readings use accurate.
C. Percention of the Non-Condensable Gas ubble in the Reactor f,ggg31 - Early in the evening of March 29th, a group of engineers met to di'scuss present plant status. Two of these engineers reported to the control roon to back up the operating staff. gy approxi-mately 2103, it wa's apparent to this group that a non condensable bubble existed in the reactor coolant system. Prior to 2300, a formia was derived to calculate the size of the ses space in the system. Gas
- bubble volumes were routinely calculated throughout the 30th of March and calculated voltanes began to decrease late in the evening of the 30th. Jased on analysis of the 3/28 reactor building pressure spike and containasse air saigle analysis which began at approximately 0400 .
on 3/31, it was determined that the gas bubble in the reactor coolant A
e
.d. .
. i
.p Volume of the gas space decreased system was primarily hydrogen. -
This was confirmed ts che increase in hydrogen s steadily through April 2. -
1
, coscentration of contatament atmosphere leveled off as the bubble in the reactor coolant system diffused. ,
1
'k4ar3 The first management communications concerning the THI-2 accident were -
drafted in a telephone conversation between Mr. Berbein and Mr. yabian At this time, they mucus 117 .
approximately 7:15 a.m. Wednesday, March 28.
drafted a statement for respassa to press inquirse that related that the TMI-2 reactor was shutdown due to a malfunction in a feedwater system.
The entire unit systematically shutdown and was espected to be out of
)
servise for about a week while equipment is checked and repairs were
~
mode.
At approximately 9:30 'Jednesday mornins, Gary Miller called Mr. Troffer I h
( to relate his conversations with t,t. Governor scrantaa 4escerning t e l
During these conversations, Gary indicated that thera unit status.
was some fuel pin leakage, however, he noted that he didn't have any '
The prepared seatement to the press was updated indication of fuel melt. ,
by noon, March 28. .This statement revealed radiatiou l'evels were' bei monitored in and aroJud the plant and that there had been no recordings !
f of any significant levels of radiation and none were expected outside i the plant. No svacuation of tha lo:s1 population was indicated at that time and that the reactor was being cooled according to design by the reactor costant system and should be cooled down by the end of the..day, "
of a melt down.
March 28. It added there was no danger During the Met-Ed press conference in Hershey, on March 29, Mr. Herbein l
said it was too early to tell the extent of the fuel desage'at TMI-2.
L However, he notti that fuel failure had been experienced during the accident.
He related this fuel failure to the point of turning off He updated the plant the reactor coolant p.rsps during the transient. * ,
. . . , _ . _ ..~,,...._.,-...~,_.m,.- . . - - = - _ _ _ . . . , , - - _ _ - . , _ _ _ , _ . . - _ . _ , , . , _ , , , . . , . - , _ _ . . _ _ - . , - . _ . - , , . ...-.-
~7*
status to say that a reactor coolant pump was running and cosidown was
(
proceeding and that he aspected to be on the decay heat system in .
apprestimately 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. In response to questions from the press.
Mr. Eerbein related perhaps one half to one percent of the rois any have experienced some melting and that the fuel had primarily failed due to a the reactor coolant system depressurization and the need to shutdown ,
the reactor cosiaat punys. Es noted that it was possible for sons steeming in the upper core region at that time that lead to the fuel failure.
Early in the evening of Thursday, March 29, Mr. William Iowa, Mr. J. p. Meere had gone to the Unit 2 control gesa to assist the
! operating staff. Based on observed indications this group assessed that there was e aos-condensable gas bubble above the core. Later'that -
evening, salaulations begna to determine the volume of the gas bubble.
C Duri.g ae.,-se seafe-e gives . ner. 3., t. .e. eta -1 4 I
the evidense of the gas bubble s'beve the sore. .owever, he noted i that it appeared that the fuel assemblies were covered at that time
- and that decay heat renoval was progressing. Es suggested at this time that the fuel failure was caused by a momentary uncovering of the fuel during the transient.
i .
During the press conference on Marek 31, Mr. Eerbein revealed that efforts were underway to reduce the size of the bubble over the top of the fuel. Initial indications indicated that the venting process was successfut and that the bubble has reduced in size. He did usatios -
- at this time, however, a soncern that the venting process has lead to a build up of hydrogen in the reactor building. During the evening a
. _ _ . , . . . - . . ,