ML19248D384
| ML19248D384 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/17/1979 |
| From: | Minners W Office of Nuclear Reactor Regulation |
| To: | Mattson R NRC - TMI-2 LESSONS LEARNED TASK FORCE |
| References | |
| NUDOCS 7908150716 | |
| Download: ML19248D384 (4) | |
Text
a f * *
- C0r o
UNITED STATES
[ \\(V."', q [',h NUCLEAR REGULATORY COMMISSION
/, C W ASHir.G TON, D. C. 20555 t*
'.....'a JUL 17 379 MEMORA:DUM FOR:
Roger J. Mattson, Director TMI-2 Lessons Learned Task Force FROM:
Warren Minners
SUBJECT:
TRIP REPORT - TMI SYSTEM THETIAL HYDPAULIC NiALYSIS MEETING JULY 11 & 12,1979, PALO ALTO, CA The meeting by EPRI with the objectives of 1) gaining additional technical insights on the TMI-2 system thermal-hydraulic response, 2) to assess the capabilities of current computer programs and 3) to identify the improvements in the cc: putcr programs.
The reeting.,as attended by representatives cf all of the organizations (EPRI, GPU, ITI, INEL, LASL and NRC) analyzing the TMI accident data. An agenda listing of the participants is attachad.
Copies of the slides presented at the meeting are available in my file.
Two general approaches for evaluating the WI-2 system resp:1se, data analy-sis and cc:puter analysis,..ere preser' ed.
Data Analysis 20th EPRI and GPU presented their analysis of the data.
Tha majority of the discussion c;ntered on the first 100 minutes of the accident, that is until shortly af ter the reactor coolant pumps were shut off.
EPRI cor.cludes, based on aa energy balance on the relief valve drain tank,that the re'.ief valve failed in the full cpen position.
Mcwever, the m ss ficw rate though the valve is &.nsitive to the steam quality and has not yet kan fetermined.
Eased on their computer analysis, 23'.-l calculates twice the mass flew rate estimated by EPRI.
20th Ruf and EPRI estirate the fica cate in the letdc',,n line to be eaut 140 Spa bued on 1 heat ba!arce across the lat&. n line cooler.
Therefore the letdcun line fic is c. parable to the ficw Tecm the relief valve.
lcuever, 3?.N estiTates that the net letdan/ makeup ficw averaged a positive 35 gpm ahich sets the makeup ficw at 175 gpm.
EPRI 5cli. 'es such a high a' :up fl:w :hculd be cricalad by the response of yst a para...cters
.h ic h they do at sae.
EP'll estima tas the '"keup ficw to 'rve been 20 g,m,
- .ich result in a aegit'. te net letdan/ makeup ficw of 100 ;pm.
There is a c: 1cen-sus that the ater level in the RCS after the RCP..ere + urn:d off es not
'5sve Se ?renure vess 21 hot 129 noules.
790815 M m '
o
/
/
Roger J. Mattson EPRI speculates that since the system appears to have ';egun to repressurize before the rellef valve block was first closed, hydrogen may have been generated soon af ter the RCP were stopped.
EPRI also spectlated that a rapid 140 F increase in the cold leg terperature and similar 300 psi increase in system pressure at 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and 45 minutes may indicate the time when the core collapsed.
However, these increases can he also explained by increases in liPI ficw, outflow from the pressurizer and re-welting of the core. After generation of hydrog :n little or not heat transfer through the steam generators occurred and cooling was due to flow of HPI injected water through the core.
During the period between 11 and almost 16 hcors there was no significant heat removal from the core.
CcEputer Analysis Ccr..puter analysis of the accident are being nade to 1) e.aluate the capability of the codes, 2) fill in gaps in knowledge of the accident and 3) investigate the affect of possible alternatives.
The NRC representatives urged that the primary sphasis be placed on the last purpose, that is answer come questicns of ".,ha t i f".
The codes being used are RETRAN (CPU, ITI and EPRI), RELAP (INEL), TRAC (LASL) and CRAFT (PS.4).
The RETRAN codes used a detailed system model (about 100 volume ( including nodeling much of the secondary system. The RELAP and TRAC aedels used half as many.olumes (about 50) and.r.uch simpler secondary side mod el s.
Except for TRAC which had 3-D rcodeling of the vessel, the models
'ece one dir..ensional.
The computer analyses generally predicted the trends of the parameters better than the quantitative valves.
The ccdes that atter:pted to model the seccndary systcm in the rost d2 tail (RETRAN) nde poarer predictions, than codes that used secondary conditions as boundary conditions to a reactor coolant syste:n (GAFT).
The tire span calculated to date is short as shown by the following table.
CODE I:, VEST IOTOR TI"E SPAN REihN GPU S min.
RETRAN ITI 3 min.
RETRAN EFRI 1/2 min.
RELAP INEL 100 min.
TRAC IASL 100 min.
GAFT 3',W 100,ain.
The calculations show that the relief valve ass flow is insitive to tha
- team quality.
All analyses sverpredict the initial decense in pr_
arizer level.
The RCS flew rate decay &e to /;id ~cr nian is ;niar-;cy.ic ed by ha So c &s (RELAP and UC) for.hich t:a, c -tar as. iilable.
sO '
I
(_
~-
/
Roger J. attson In order to facilitate completion of the analyses and comparison of the results a " standard" set of initial and boundary conditions was developed based cn the best information available to date.
s J
, /' r e c c - // d m.~
'darren Minners cc: LLTF Dist.
POR Z. Ecsztoczy D. Ross
'J. Johnston S. Fabic f
w L.
t /
.)
l
~c
.. om b-,..r 1
f g
AGENDA for "TMI-2 s'ISTEM T!!EE"AL li'iDRAULIC ANAillSIS MEETING" LOCATION:
"m. 3-254, Bldg.
3, EPRI, Palo Alto, CA I
lI J u. l. A -.11, 1979 v
I 3:30 Introduction (T. Fernandez) 3:35-12:30
_ Plant Data Analys_es.
.a 3 : 3 5 - '; ; 4 5 TMI-2 Plant Data Analyces ;;PRI (J. -:t.~ia i l) a; t
9:45-10:00 BE'K I
10:00-11: 15 TMI-2 P' ant Data Analyses-GPU (G. 3 roughton) 11:1S-11: 3J Early Tima
,ur. ped Pa r a.ma t er I'
- '?
'nal.vsis-EPRI (J.
P. Sur30ck)
I, 11.'0-1?:30 DISCUSSICN (All Pirticiants) i i -
1? '0-1:30 LLNCH i C, k
t 1:30-5:30 cc.Luter Cc h._ _.__'_,_.a l v._o e s.
1.30 -2:.30
. c i'EJ. N 'nalysen-GPC/EI (N.Tri' cures, J.":rriren) 2:30-3:30
- E.T.W ir.a ly se s--
TI t-(K. :Iouse, C. S l ai er, C.Mi ller) 3:30-3:45 5.u AK j
i
'l 3:25--4:30 RT ;J N
.'.n a ly ; c c - E P R I (J. ';a s e r )
t t. 3 0-- 3 : 3 0 DISCUSSION i
( All P icticipant s) e i
.e i
7 ~10TM 0 T: ;;.R 1
t, t
j t
l \\,
_'_u l v.
1 2, 1979 i
6 3.30-12:30 C o.,n u t e r C : J e. ------. __ A_n a l v s e c ( c c a t. )
t
-.s_
l 3:30- 3:45
..C
':I-2 '.n a 1;, s
.',ct ; vi ti. s (S. F
.:)
f 3-
9:45
.W.'J 4. r3D 7 inalynac-: EL (M.lu i.
in, S.
Edl' ;)
I 3.<
_-10:00 2.. - i.x t
.B
'M:00-11rG0
- ' 1.C 1. 21. - - ; - ' ': L (J.
.1. vi) f
)
.. n, _,' 30
._".'.a-.
/
s
\\
I.
->J i
- ' J0-12:30 Di se c;i 'a 7
(All"._
.: ; n '. ; )
12.20-1-30 TJC:1 6
m
( O.,
-, a v.,. : 0
...-...n
,a)
. t
/
=.
r.-1=
',.3_
.ng
- .-1 g
f
.x s..
9
....,,,,..)
=
. ~ >
3 0 -
,3 :,. S
.s t sp, y.--
- ..)