ML20212M604
| ML20212M604 | |
| Person / Time | |
|---|---|
| Issue date: | 03/09/1987 |
| From: | Grimsley D NRC OFFICE OF ADMINISTRATION (ADM) |
| To: | Aftergood S COMMITTEE TO BRIDGE THE GAP |
| References | |
| FOIA-87-33 NUDOCS 8703120093 | |
| Download: ML20212M604 (4) | |
Text
U.S. NUCLEAR CE!Ull. TORY COMMISSION Nnc*F0lMbuusEn<si-33 a necytsT f.e ~.s,,,
s RESPONSE TYPE 3
N""^'
I l'*""^'
fY I
RESPONSE TO FREEDOM OF v
~
^
- 5 INFORMATION ACT (FOIA) REQUEST g
00CAET NUM8'ERisi W aspecacei h$
b(
0.
~('
'I PART 1.-RECORDS RELEASED 01: NOT LOCATED (See checked boxes)
No agency records subsect to the request have been located.
No additional agency records subsect to the request have been located.
Agency records subject to the request that are identshed in Appendia are already available for public inspection and copying in the NRC Public Document Paom, 1717 H Street, N.W., Washington, DC.
Agency records subject to the request that are identifed in Appendia are being made available for public inspection and copying in the NRC Public Doc. rent Room,1717 H Street, N.W., Washington, DC, in a folder under thes FOIA number and requester name.
The nonproprietary verson of the proposalts) that you ag'eed to accept in a telephor.e conversation with a member of my staff is now being made avaiable for public 6nspection and coying at the NRC Public Document Room,1717 H Street, N W. Washington, DC, in a folder under thus FOIA number and requester name.
Enclosed is information on how you rnay obtain eccess to and the charges for copying records placed in the NRC Public Document Room,1717 H Street, N.W., Washingtor. OC.
Agency records subject to the request are enclosed. Any applicable charge for copies of the records provided and payment procedures are noted in the comments sectior Records subsect to the request have been referred to another Federal agencyties) for review and direct response to you, in view of NRC's response to this request, no further action is being taken on appeal letter dated PART ll.A-lNFORMATION WITHHELD FROM PUBLIC DISCLOSURE Certaia information in the requested records is being weheld from public disclosure pursuant to the FOtA enemptons described in and for the reasons stated in Part II, sec-tione 8. C, and D. Any released portone of the documents for which only part of the record is being withheld are be ng made available for public inspection and copying c the NRC Public Document Room,1717 H St set, N W., Washington, DC, in a folder under this FOIA number and roquester name.
Romments 0703120093 070309 PDR FOIA AFTEHCOOB7-33 PDR A
. 2 SIGN iMECTOR, Divt$ ION F AIS A DS A2x-
\\
NRC FORM 464 iPart t) 9 asi
FREEDOM OF INFORMATION ACT RESPONSE FOIA NUMBE3St DATE M 9 m
PART II 8-A:wLICABLE OlA EXEMPTIONS Records subject to the request that are desenbod in the enclosed Appendices i are being withheld in their entirety or in part under FOIA Enemptions and for the reasons set forth below pursuant to 5 U.S.C. SS2(b) and 10 CFR 9.5(a) of NRC Regulations.
- 1. The wthheld informaten is property classded pursuant to Executive Order 12356 (EXEMPTION 1)
- 2. The wthheld information relates solely to the intomat personnel rules and procedures of NRC. (EXEMPTION 2)
- 3. The wthheld information is specif cally enempted from pubrc desclosure by statute indcated: (EXEMPTION 3)
Secten 141146 of the Atome Energy Act which prohibits the disclosure of Restrcted Data or Formerly Restncted Data (42 U.S.C. 21612166L Section 147 of the Atomic Energy Act which prohibits the disclosure of Unclassded Safeguards information 142 U.S.C. 2167).
- 4. The withheld information is a trade secret or commercial or fmancial information that is being withheld for the reason (s) indicred: (EXEMPTION di The information is considered to be confdential busmess fpropnetary) information.
The information is considered to be propnetary information pursuant to 10 CFR 2.790(d)(1).
The information was submitted and rece=ed in confidence from a foreign source pursuant to 10 CFR 2.790(dH2).
- 6. The withheld informaten consists of intecagency or intraagency records that are not available through discovery during litigation. Declosure of predecisional information would tend to inhibit the open and fiank enchange of ideas essential to the deliberative process. Where records are withheld in their entirety. the facts are inextricably intertened with the credeceeonelinformation. There sino are no reasonabtv se inerect mqury into the prodoceenal process of the egency. (EXEMPTION $ gregable factual portions because the roisase of the facts would permet an 4
- 8. The withheld information a enempted from publec declosure bor.ause its esclosure would result in a clearty unwarranted invasion of personal privacy. (EXEMPTION 6)
- 7. The withheld information consists of investigatory records compied for law enforcement purposes and is bemg withheid for the reason (s) indicated. (EXEMPTION 7)
Declosure would interfere mth en enforcement proceedmg because it could reveal the scope, directen, and focus of enforcement efforts. and thus could possabty allow them to take acten to sheid potential wrongdoeng or a violation of NRC requirements from investigators. (EXEMPTION 7(All Disclosure would constitute an unwarrented invason of personal privacy (EXEMPTION 7(Cl) s The information consists of names of indmutuals and other ir. formation the disclosure of which would reveal identities of confidential sources. (EXEMPTION 7(D))
PART ll.C-DENYING OFFICIALS Pursuant to 10 CFR 9 9 and/or 9.15 of the U.S. Nucteer Regulatory Commissen regulatons, it has been determined that the informaton mthheld is enempt from producten or disclosure, and that its producten or disclosure is cor*trary to the pubic interest. The persons responsible for the denial are those offlCials identifed below as denying officials and the Director, Drvisen of Rules and Recorde. Offica of Admerustraten, for any denials that may be appealed to the Executive Director for Ooorstens IEDot DENYING OFFICIAL TITLE / OFFICE RECORDS DENIED APPELLATF OFFICIAL SECRETARY EDO hs B.FitzmaW hskh&wed Cael X-9 r
PART II D-APPEAL RIGHTS The denist by each denying official identified in Part ll.C may be appealed to the Appellate Official identified in that section. Any such appeal must be in writing and must be made within 30 days of receiot of this response. Appeals must be addressed as appropriate to the Executive Director for Operatens or to the Secretary of the Commission, U.S. Nudeer Regulatory Commission, Washington, DC 20566, and should clearly state on the envelope and in the letter that it is an " Appeal from an Initial FOIA Decesson mac Foau ase iret a U.S. NUCLEAR REGULATORY COMMISSION
- asi FOlA RESPONSE CONTINUATION
-h Re:
FOIA-q APPENDIX l
^ RECORDS MAINTAINED IN THE PDR UNDER THE ABOVE REQUEST NUMBER NUMBER DATE' DESCRIPTION
,Oli@
h _{ L-1(,c v9 An 'c k d(t55 l
hecatTer m 'ec hn b ' g gg g l
9reas
L o es) l l
J Re:
F0IA-
-b RECORDS PARTIALLY WITHHELD NUMBER DATE DESCRIPTION & EXEMPTION A 10 %
horaaltun }(em ws 5. StqafakI
-lbVbd 6fe)0,obb 8.SUl b N 7
4 TnacpaWt Evprts of 1serti bs Eatta e scura. Tarn ua2rhnEs.
Evathnf l
pe %
December 12, 1986 POLICY ISSUE SECv-86-369 (Information)
For:
The Comissioners From:
Victor Stello, Jr.
Executive Director for Operations
Subject:
plan TO ADDRESS SOURCE TERM TECHNICAL UNCERTAINTY AREAS purpose:
To inform the Commission of plans to address the areas of major uncertainty described in the staff's report, " Reassessment of the Technical Bases for Estimating Source Terms," NUREG-0956
Background:
On July 9,1986 the staff briefed the Comission on the severe accident source term assessrnent that was soon to be published as HUREG-0956 (the report was published later that same month). At the meeting the Comission requested the staff to submit a plan to review each of the problem areas listed in Table 6.1 of l
NUREG-0956. The Comission further requested the staff to recomend whether an independent group of experts should be constituted to review each of the major areas of uncertainty and to recommend a research program for resolution, if needed. The Commission's requests are described in a staff Requirements Memorandum (H860709 Chilk to Stello, July 18,1986),anda brief memorandum describing our progress on these requests was sent to the Comission froni my office on October 27, 1986.
Discussion:
Table 6.1 of NUREG-0956 is reproduced in Enclosure la and lists the eight areas of major uncertainty identified during the staff's source term reassessment. A detailed plan to reduce uncertainties in these areas has been developed and is presented in Enclosure 2.
This plan includes timetables, milestones, and budgetary reouirements needed to address these issues adequately.
In addition to research programs specifically aimed at addressing these issues, other closely related work is being performed in support of these programs.
Specifically, the
Contact:
P. Silberberg, RES i
443-7997 or G. Parino, RES A
I 443-7622
'l)'
y )_, m m /> m P L
]O
~r t " 7 " N '
/
.=. -
.=
The Commissioners 2
supportworkconsistsof(a)codedevelopmentandvalidation, (b) experiments to provide an improved data base for code validation (c)(phenomenological analysis of specific accident sequences, and d)uncertaintyanalyses. This supporting work is also described in Enclosure 2.
A sunnary of budgetary requirements for the programs addressing the major issues and for the support work is shown in Enclosure Ib.
In response to the Connission's second request, the staff has recomended that the NRC obtain the independent expert reviews of the major areas of uncertainty and has asked Brookhaven National Laboratory to provide for such reviews, coordinate the reviews, and summarize their technical findings. Dr. Herbert Kouts, Chairman, Department of Nuclear Energy, Brookhaven National Laboratory (BNL), will lead the review activity. Several issues have been combined so that it is necessary to conduct only four independent reviews as shown in Enclosure Ic. The reviewers will include experts from other countries, as suggested by Chairman Zech.
BNL will utilize the information from the independent reviews and prepare a report containing technical conclusions on the ability of our current research programs to achieve their stated goal.
This report will be submitted to the NRC in March 1987. A revised final plan for resolution of the eight areas of uncertainty will then be developed and submitted to the Comission in April 1987, in time for any potential revision to the President's FY 88 budget and planning for the FY 89 budget.
Additional details on the BNL review effort are given in Section i
'i IV of Enclosure 2.
Enclosures Id through Ik are sumary charts of the research time line milestones for each uncertainty area. Also included are timelines for an accelerated schedule, pfven the additional funds shown, and for the implementation p an, which includes revisions to source tem regulations and Individual Plant Examinations (IPE).
The staff is now completing work on NUREG-1150 scheduled for publication in January 1987. This study will provide quantification of the risk significance of the major areas of uncertainty, not previously available. This infonnation will
(
be taken into account in future revisions of the research
- program, i
Finally, there will be some impact from the Chernobyl accident on the scurce term programs. The staff is completing its implications assessment and has tentatively concluded that i
several mechanisms for the release of fission products from core debris need further study. These mechanisms played an important role in the Chernobyl accident but they are not included in i
i
The Comissioners 3
our current source term assessments.
It has not yet been deter-mined whether these mechanisms would be important in accidents in LWRs.
Further discussion of this subject will be included in the staff's report on the Chernobyl implications assessment.
In addition, the review groups discussed above will also be asked to consider any relevance of the Chernobyl accident to the eight areas of major uncertainty identified in NUREG-0956.
Adjustments to the staff's plan for resolution of source term issues related to the Chernobyl accident will be made in the final version of the plan that will be submitted to the Comission next April.
The unfunded needs of the programs addressing source terms areas of major uncertainty and related support exceed $9.5 million. The unfunded budgetary requirements of this magnitude should be met by requesting additional budget authority.
,A D teflo.
c Executive Directof for Operations
Enclosures:
As stated cc:
SECY OGC i
,---,--.-_,-v,
..-_,_,-.ry.--,_.,,.._..e
.v
-..,,--x---,,
,.,-,__-_,___y_.,__,_o.--.._
--,.-_.,____~_,---,,__
y
Enclosure la Table 6.1 from NUREG-0956, " Reassessment of the Technical Bases for Estimating Source Tems."
Area of Uncertainty 1.
Natural circulation in reactor coolant system 2.
Core melt progression and hydrogen generation 3.
Steam explosions 4.
High-pressure melt ejection 5.
Core-concrete interactions 6.
Hydrogen combustion 7.
Iodine chemical fo m 8.
Fission product revaporization D
,--,,y-----
..,m-e.,.-,.4.,-,
-,<--.,--,s-.
_,.-~,-w=_.----w+~-
-~_-,.,-~.
Enclosure Ib Summary cf budgetary requirements for programs addressing source tems areas of major ur. certainty and related support.*
(DollarsinThousands)
FY 87 FY 87 FT 88 FT 58 FT 89 Area of Uncertainty Funded Unfunded Funded Unfunded Funded Natural Circulation 1.000 400 1.000 900 1.000 Core Melt Progression 3.934 1.350 4.810 2.000 4.915 Steam Explosions 567 260 425 375 425 i
High Pressure Melt Ejection 1.740 200 1.800 250 1.500 Core-Concrete Interactions 1.665 850 2.000 0
2.200 Hydrogen Combustion 579 530 600 200 0
fodine Chemical form 1.693 400 2.050 300 1.700 Fission Product Revaporization 190 500 200 700 100 f
Subtotal 11.368 4.490 12.885 4.725 11.840 t
Program Support M
300 4.350 0
4.350 Total 15.726 4.790 17.235 4.725 16.190 i
- Does not include new work related to Chernobyl.
?
4 i
i t
h i
j s
i
Enclosure Ic BNL Plan for implementation of the review of NRC research plans for addressing source ters areas of ma,ior uncertainty using independent experts. Dr. Herbert Kouts, Chairman, Department of Nuclear Energy, BNL will lead the review activity.
Areas of Uncertainty
- Natural Circulation and Core Melt Progression Energetic Events High Pressure Melt Ejection and Steam Explosions Core-Concrete Interactions Idoine Chemical Form and Fission Product Revaporization
- A recent National Academy of Science report on hydrogen combustion will be utilized in lieu of a special review on this topic.
F MP N." {
.d N
.~
g co n n oc u g.~
B m ee w w a
- o. -* a tu m t ~ ta g er o s o o
4 met *34 3 3jjj uraab25Q **p2 S f = Em # nv5m E
QEE S$ T1*
Emon E
x a w a" r e % ? a !.
, r a. 0 5 8.b * -3 8 A 4 j.2ss A
- cer:apmana#'
cag..a
,craa s
-La mtg ene= e; pgg.-ga-aa e
3 $ " % *,$ u
- E n "S
Dm SFw5"
~1I g %
s aama x?.
r,. g.* s ? 9 a
"FW 2
- t*S4*88~2 a8.
s Ba2 g
o
.x:
s e1:s 3:es a,s s
a e
mann-
.a ar 5,
e a
e psegg g
g m
e aa n
O E
o C
E' O
3
.g g
2, M
O 3:
a ri d
2 COMMIX Calculation for 0
E 7 z
3 Heat Loss to S.G.
<3 issue IPE Guidelines 9l l Cooperative plant with c) and Hot leg Piping 5
2 EPRI for addt'l W tests
(
3 g
MELPROG/ TRAC
' E O
m
-l q [ Calculation for Surry E dissue first source F, 4 SCDAP/RELAPS Calculation term regulation change 4 P for Surry
(
Analysis na omparison 3,
C z
O m
l l Ad itional High M,ELPROG/ TRAC VaHdation p
O with W Data Pressure Tests 9$
p u
d First IPE (Utility)
@W Facility gg g
d PRA for future plant g
> a O
C
> Calculation of Pump seal LOCA with MELPROG/ TRAC r
Q St'aff review of first IPE l l Assess Need for Mose Teste d
O
( ) Start additional ar.alyses g
M of PWR sequences O:
l l MELPROG/ TRAC v i tion z
with W data C
$p
( ) MELPROG/ TRAC SENSITIVITY O
m m analysis d La61 IPE (Utility) m m 4.
-4 m
- ['.
d issue last source term D$
>= SFT Analysis ge g regulation change for Surge Line. Hot Leg '
Weld, and Penetrations E
>"' Break Sire Analysis E
E and Uncertainty Estimate 0,
O b
C L
ca
'C Staff review of last IPE j
pt aansotou3 7
f
]
F f
?
F."
arss esrgs tr
= -mess cess ar a gnraeggen:sr s~
=
GT~~1";#ta;
- MG47'ag82.."d*2 a 22:15 3 "dg ra a 3 % 2 e
oaE5"aI=1%2AT[S M
5 E5o" O
a ' a y a -l g e. E. % :r " Y 1 w - *,as :t egn e
=
1 52TPS$o%o$
ascasaz$nat
-,r 5See C
l d $*ES$1Eo*%Ed"O M U 25 n 2 E"E* s* $ ES I E E E5m a %E*-
9 E
72*Eu$
250%
1 % ". " l " E * "e N o
","d*o' n
w-M 2 " 2 * * ' a.-
- avas? ara" Mar
- SI r4 8
"E"R COT"**8A=O2 j,gg las-at"E4"s 400
"*N 42R%
m, n i: g *.4aa%;; e i*.a
- dT S f
3022.7
=
a
- an= s.~s,
- are R,2::::
m r _ a :" 4 m
- 8 3 8 ?"? r n r 2 SE18
" C
- S *r = " "
2 a#W 8.
"Cr
& Ca8g4231 lit B 0 887*
a 4
- 5 B*
5 a:ma n'
x
. = *
=
-e og gi g
O 0 () Report on PSF test SFD 11 E
O C
g
( ) DF-4 test (SWR)in ACRR o
5 m
m 5
i m
l l Start PIE of NRU tests IIh'(ppg 2 Malysis we T
E 2
g d issue IPE Guidelines a
g g
j p
5 O
o( ) Start small-scale tests
,n Z
for MELPROG validation g
2 O
d
( ) Report on NRU test FLHT 2 g
g
% Q lssue first source 6
m term regulation change l
Preliminary TMI 2 4 > Complete PIE of PBF tests 2
x analysis with MELPROG
( > NRU FLHT-5 test Q
I hReport on hydrogen generation O
M
(
in P8F tests ST and 11 2
O M
oport on ACRR tests g
j DF 2 ft OF-3 Z
g 7
4 4 > Report on DF-4 O
t-x
-4 m m
Q First IPE (Utility) 4 gg
(
ep rt on FLHT 4 g
( > Report on tests for E
2 4 PRA for future plant
(
MELPROG validation 2
I
( > Report on P8F test U
g
.SFD 14 d Staff review of first IPE
> LMF 1 test 4
c.
(
U 5-y
( ) NRU FLHT-4 test 2
O
> ACRR MP 1 test rn
=--.-
g Q
m 4 bReport from OECD-NEA c
on TM1-2 standard problem E
2
-(
Report on LMF-1 O
4 Last IPE (Utility)
$rn t-m g$
(
Report on P8F test SFD 13 g
Qlssue last source term q
regulation change Report on FLHT-5 M
3
> LMF-2 test
)
3 9
- g S. %
-f m
> >** v o
..y y
\\
z;
\\
> ACRR MP-2 test i
\\
.\\
dStaff review of last IPE !
_,_, y y g _ gegrtgLg2 '
~.
ca e
at a.msotou3
i i
.N r
iP; F
FN
?
E
?
F 2
?
C3-n
-wewwsm
--snenzu e V,wunemm -
8 wms,wMnzes's-4 m
-1 ne 3yOxeH x e -% N (D -+ C O - fD f3
- 2 g V
-a m 3 N O vi x N -% r+ -* 3 -* 3 x 3 fD ft a?Og"a.2**G*";*"VRagt
+ 3 k m a;;.
an e
12 2 a ** "**
? 2:
mCM3 mom-2 V
- On n o m e s w1 C mW 2
52 woo 33 on v 9 (D m CL U ehMm9 c+ #9 woe m
& 7 m ne -1 e f
m 3 r* r, =1 n1 Va
-1 m
c+ V O c+ -*V -*= (D 1 m (D
=* m c+(D r* 3"Os C4 -a.
m
=* m
- w m
- Q.
t e y fB ft 3 p 3 1 ("1 2* Ge fvt O
--* m
-* O Qe r*
M C
1 CL x e -* ot OO r+
'O fD fD. r+ N m
-*-*CLfB O O M M (D n r* n O CL O t c * -4 M
=* V CL O 3 r*
f9 3OO V 3 O m 40 --* m C
-4 O O fD 3C Of V W 0s O 3== m c* -1 O #D 3 08 *=
2
% C r* w
- 3 m -1 M 3 -1 3*
3 m==* O 3 r*
am -*9 3
9 Of n m CL O MT
- r+ Q. 2 O C (D =* < (D r* p Op 3 3 3. =a-M O -% c*
3
-4 y toOOn m C e+ ft
==
-% < O 4 fD m Q. m -1 x (D x ft M --* m -%
O r+
n n -1 fD & re M
e 3 or e op -1 O m =* -1 e 2
- =
3 fD a m we.* (D ft =** -1 'O O fD (D w Os fD 3"
(D Op 3 Os 3" Os 'D os os fD a
-8 0s -% e O M r
-1 e c*
O x 3 m it m m w -* -% O m =* Or r* ne vs 9
--4 Q. c* n L O r*
===c*
n T
m fD E e 3. Os-* 3 o V 10 =*= l O m La I --* -% -% N m V c+
-a. < f5 3 c)
C O fD 3 ft *. 3 o ft 9O 3m
-'" c+ x (D
-a C (D O
- ta fD C (D e
3" -* O
-a.
fD
%-13O 3 O ro 3 c. O R S
2 3
MMn%m Z
"O 9 O 08 3 fB =*= O 3
-13 Q. 5
- e c+
3 th H
- =* C ft *Q 7 O
m fv1 J3 m3 C 4 to -% 9 08 e+
(D =an 3 C
O9 Os e r*
- T1 Co V -1 % (D no O. f O
a
-4 m -a. $
fD CL 2
(D su m
(
a
- 2 a', 2. g "
88
.a ?
S g
C ". '
SF T
O* O e
fD Q* n fD 3 O
-1 C
3 0
~
O E
O C
z
?
o g
"t m
O E
o 2
m z
-4 z
y u
H r
C 2l l Start natural circulation z
o z
analysis with g
2 Qz dissue IPE Guidelines o
MELPROG/ TRAC 2
E m
3 0
6 c
l l Complete stratified 5 dissue first source tests
+
term regulation change (f) l l Calculation of head y
temperature with natural circulation
( ) Complete jet mixing tests IT1 rA
---d P
O z
l l Start hi pressure h g
cutoff tests 2
m O
Q y.
o 6
n!
d First IPE (Utility) 3l l Head failure h$
I E
"c d PRA for future plant h
model natural Q
circulation U
i $ Assess let mixing models g
i D Complete jet O
E energetics tests 2
d Staff review of first IPE l l Evaluation need for (f) large-scale tests D Complete IFCI model l l Natural circulation alpha mode assessment 6
q Jet energetica report O
2 l l Mecharlistic IFCI.' Q model i z'
O n>
mm dLast IPE (Utility) 3m
=
u
-4 m m
ll Integrate RFCI R$
g y d issue last source term model intoi Z
regulation change M El. PROG' I
Report on hi pressure cutoff tests g
u Q Staff review of last IPE; go J[ 8JflSO[3u3 4,
iIIIIj;l1
)l
)
lll 4)Ill 1"CI " 3IM@C:m E Id mtmOj_OZ U T D
I fcOc5Ess,BEQ9r l
n e
d o,
o
)
l F ys$$4*.,S3C Ca.
M it e
y t
yr
- EE o* s%
d s
v
)
fie l
a y
t n e
y C
EtEsg= ?S n1~
it
)
y y
ne c
a t
r r
n s
o o
aG
)
r
(
v n
la2Er= aR*
n p N.
e) t t
u a
a e S B
i Tlo a a
Qn
)
C
)
d r
v r
d s
g m
F Qi0"e%.qMEa" r e
e e lap i f
o m m
og
)
y h
o e ir
)
t-u t
e A v
r r
g r
t a
a IT4 xn% "01C K
y r
o n
i t o t
r r
l d a l
a e
W f
u r
ir sr s
al P
h it H
o e
itde a
y 0
p Fe n n
td W C s
E aEl* e*3U" t
8 S s
(
(
t o
o sy p
(
(
m
(
(
tr ino w
T-e t
t a C C
eH s
n C7$23 F EgE o* s.13S=
s i eW l
t t
t t
i t
s T
t t
u s )e iM t o sl s o sms s
s
)
)
e l
s e
e Sn e
e eF em e e
e e e
T e*T aD aD el sac S*.
r t e ei T
Tt Ta T T t.
w nB nB t
t T
T da Tu t s e*1 o
d e
le.
0 o
es l
t s
6A le EREC8.n3E* g=
H.i t Hu mla Hi Ht m
Ci
- H m H m*Ha it(
it(
pe iT lp"1loiT 2
3 eP 4c lpy 5e p
S p
l s 7t 8 9 1
H
- s
- n H
H d
d mV l
t t
v s r
r C-s s
C C ee Ct on Ce Ce o C C o
C o Ch de de o x-f r
$,eE?G5E
- 5M "
D D RT D(
n S
D C^D( AT AT CE
) I
)
)
)
p
)
)
>)
I
)
)
)
$F
(
(
)(
". 01* E% 7 eS8' M
h4
(-
(
( I
(
(
gmE2H 2:go 1#" %2-7SOg m4.DEgG QZ; nC$mm$EpV
%um' mm mm 9% o m;$Ca o* s*
Ea* 8* " 2" E9' F 7* E7%* $G5R,dM r
$tD7.aR%3 i" EE2~ a.aE% o" Ex S5* 2lr' 2C5* a E* $* " aE2 e8RMw S5 n2Sar* # %
13E2s1"E$C8
(
4 t
(
q
$.l 2"
> o5Oz? a2O 2o E;--
g I
.k1-2 p
EaE* 5Er E
l m
P E
e P
g I
s n
t I
r e
a n
t F 3% v2" s* 4#
i ec
)
lp f
e t
i s
n h
a s
r t e a
d%CE.:O" l
c y
y cn f
)
eg l
e d
r n i e f
E%71 m.4C4 <3 iu oi it t
t r a o
r o
=
i uh uo l
u
=
lit o c w
w
=
E a%*3: aa1" G
sa U(
f ie U
s t
t tl u
so ie n
E e%S ~..C*. >
u
=
(
t E
s v
v r g E
r E
P i e P o e
ai e
2n0E,$:
I f
I f
P t
r l
r r
a I
e e
t A
f el f
3 F Ei 7E5
- g" "
- u um t
u u f
f s
a s
sg a
u s
sr i R r
Er E.Ee aa1*
is se F P S
L lse o
t a
t it r
S d
d 4d Q
d d G* o t
R*. s aH$*
TaES ~
s eR*
F TtD1* ;:E@*
n g ? mE g y e @ " @ m E r"'
i.
%a a
3a3$* ra ET*
Q,w
.<3 n
m t
$t 7*
mE 3 E -
6
> M OZ 0umE >
- u">g O z C c.
, p" O ZO' ' E pE e.
,lI
5
" Y E
N S
$ noemregm j
xme-o%moz u n-e n-n o nv e n o m lEEoEES$SS$E55
$ SS U5 S " -5
$ 2 8'S E ' O = $
E 15I E
m ZEME&Sei&2*&l*
8&a**22*
C 1R"S1
=
- !871 4
20 cada%3**017
- n.E a"g k
15%='E 5
a 3"
7
- Fr
- T 3 3 8 **ta"28=
.ECS@*.8"3
"$sN IEEE972E 5
"2$$$c 3
o i ~" ""#25Ta%x E < s.
- x cr.
2 av en '
Se S<oNgo@%
$*##S?$#
d 1 E7 ESES E
oo=
%1 Re**n*
a:Wi" W 1
=",4*
g o
Re *a "Q%: *
$;"G*8" C
a%
- Q m
=g a
2*";
- r E'S
.m 2*
M a
8*28
.a.
=
0 O
-E o
e 0
3 2
3 E4 > SURC tests 3 Er 3a m
I I Report on bubble-induced Zl l Start compilation of 2( ) interlayer heat transfer 6
y>
concrete properties g
n C
O.( g Start debris spreading tests E,,
5 z
d Issue IPE Guidelines
@(
2O Start debris quenching studies g
p g
z oll Start measurement of 4 > Complete WITCH / GHOST tests Q
- E m
activity coefficients O
33 u
c dissue first source term regulation change E.
( > Report on Kfk BETA tests m
0 O
rn
> SURC tests 1 & 2 g
( ) Complete integration of O
o CORCON Er VANESA g
ji z
n>
m M
~.I
[hl O
M I > Report,on debris u
Complate spreadmg phenomena n
Q First IPE (Utility) l l measurement of l
E activity.
Complete debris spreadingj H
EE 4 PRA for future plant
( coefficients
( ) tests and modeling m
Implement improved CORCON Er E
) VANESA models E
l l Complete compilationi 4 Staff review of first IPE, q of concrete propertiesi- ( ) SURC TESTS 4 Er St O
6
) Assessment report on O
CORCON fr VANESA y
> Complete debris quenching studies 2
go 0:
( ) SURC test 8' 6
( ) Report SURC resultsi Complete hot solid / concrete tests Q Last IPE (Utility) b
~
A m-41ssue last source term I$j E2 regulation change I
3 V
> Complete CORCON fr VANESA p
uncertainity reassessment -
E c!
bi 4 Staff review of last IPE l I
4L aansotou3
F F
2 F
F M. " "
F."
E
-4 n
c em
- Cmm SOWOm 904 2
m S wc
- E
"*E"gia 2&25208EE8Ea%
?
Og%3*"*"
e mm%30@ 74w 2
a c e C wmV 9 m e mWO
=
C3mm e30 2
Ut C# LD -1 3 Os m
De m e mO O
M C) = *tI3 M
e ce =* to
-* C to n
OZam Wane Ma k nD 3 Oma 3
8~4&a822 8E8""8F s"&n T
88Can C
T C) M m -1
--a 1
M e
-% T 3
w-*
-4
-si Ow1
-1
=
O 'C.C.)n o e3" O* 'n -a. O 3
S<
1 Os e S e 3
M m 3 m
=a.
C#
Z to e-a
- w3M "5 m (A e 3" 1M 5
C% M V Or C3
-1 M 3
kY b
we O
M&a 7
4 O
D Ea AW%*s" Wa 15 cs:
0 a
- 2
- an "S
9"2 2*
a*4 5 4
A n -*
-a. n De Os =.*
-4 3
40 n m Os N m
O Qe Os
=4 r3 m 3 C#. w 3
(M NO 3wN Qe
==
3, 3
m n.
a 3 c.
e cn 0
a m
p a
O 3
O 3
n 5
0 c
2 O
2 4
2 m
0 I
m o
2 2
2
- 4 ( >-Report on HDT test (series 1) 8-y m
m p
r C
-.e n
2 E
o d Issue IPE Guidelines 5
E n-< 2 2
> ce m oq HReport on HECTR code n
2 u
E 2
O m
- t A c.
Oc dissue first source hi
,,,,, Assess % scale tests c
m term regulation change E
= Improve detoM code 1 p. Report on HDT tests (series til y IbFlame accelerat.
I HReport on DDT tests 4
ion 4 (Document HMS-3 URN code Q
[*("CONCHAS SPRAY m
~""""-----l modeling with
- ~
3 O
Reanalysis of local 2
detonations for large dry O
subatmosperie ff1 2
L PWR oy g
containments m a 4First IPE (Utility)
$g J hUpdate HECTR models m-g E Q GPRA for future plant h$
Q 3
1
>I V
E E
W QStaff review of first IPE C
c_
8-l l Document large d
z y
scale Q>
{ VUpdate HMS 8 URN Models hydrogen tests a M M
-e $ _~~~
Z i--
- - _ _ =
O Em O
4Last IPE (Utility) 8-m dissue last source term E
regulation change 3
m r
L m
Q Staff review of last IPEi l
[t aansotou3
I IODINE CHEMICAL FORM 5
S Ig 4
5 PAJOR UNCERTAIN PARAMETFRS:
_j j.
ggf
.g l
l u
.i E
- 1. Chemical form of released E"
- g.3i, j
o
- ]E 3s.l iodine u 'E U
.E*:
o IE'"s E
1 j
- 2. Effects of reactor coolant 1; $
- j @ q:5 us s's {
E8 Kf
.5 %
E g o g.,. ' g ; 1 8 l 6 2 g g g P, ec system environment on
[
l
.O E 2 3*.
, 5 = s ].
chemical fom e
m o s am m E E=l :3 m ::
a s
yr ig jsy((
[ gj gigg
[
}-
-l y g j ((y []0m4UU>U>U E.g =]w ACCOMPLISHMENTS TO DATE:
y
.o0 -0 0
o aa a o. o. a a
o o
U >>
- >uoum g
hh F t st s ow ixed
=
CURRENT FUND NG l [
l[
[
results on chertical form
- 2. Parametric study of the
~
l
.~4SSISS impact of volatile iodine l
ASSESS indicates lower effect UNCERTAINTY s UNCERTAINTY l
g on the source term than l
previously thought l
l FUTURE NEEDS:
l
- 1. Experiments to detemine l
the cause of cesium iodide g
(Cs!) dissociation in the I
I N
$s nt f ect of
l E
l E
E 3
0 stability E
E
- 3. Complete high temperature j
.5 l
i G
l
~&
5 release tests (VI series) 2
$5 E2 E $5 o
o 5
35 at ORNL g]
j j g
3 3 ;g j!
l
- 4. Complete volatile iodine g
gg l
5 o
E Si 2
study for other plants 5g 9:
l 0
I'3 m
and accident sequences I
gE
% 4 i
S. Validation of the TRENDS E
jj d: E j jy j
{
n l
q q q
o code y
y yq q
IMPLEMENTATION SCHEDULE l
l l
r I
I l$
0NDJ.F MAMJ J AS ONDJ F MAMJ J AS O NDJ F MAMJ J AS
1 1
N
.N
." I 3
4 n
a S&EIEEEN199EE E TEREN
@ S257 E
2 a:E* a%G#4 =4 E.el E
? 3:*j c
%^2J38k=ETPt E *: 73 a G ;EL M
3
~ Cbt"PP%%%E G F3*3.
E 4",7 9
EESstI*$$m*-
T $*S$w 9 *7*d U
i 73%. a 8 "4 02R;s d
RE" E
i n 3
- m '1
- Hea 7:E ?
Rae L'S ;g*.
aa
=
g
=mm
=
=
a w.
=a em 1rC; e
a n
Q1 er 1"
Pa 2"*
M p
- 1 9
C 1
?
9 a
E 8
r o
E n
o C
Z
?
O m
e O
E 5
z m
z H
c y
r C
m 2
'C w
1 Q
j D() Complete RELAP5 Improvements E
Q issue IPE Guidelines z
> 0 m
E o
o I
E
( ) Complete TRAP-MELT improvements 3 b Qlssue first source
( ) Complete scoping calculations c
m term regulation change I
with RELAP5 O-Complete scoping calculations m
( ) with TRAP MELT 2
O
(
T O
2=
z O'
3 l Start experiment Qg U-
_fll oI to conlitm.'
- n m
~
g y y;;
thermodynamic C:
dFirst IPE (Utility) data m
gm g;
2.
4 PRA for future plant jl
) Complete integral analysis y
E Q
for PWR
> 13 2:
E QStaff review of first IPE u
l l Start multi-effect
),
validation tests I Complete integral analysis
[for BWR
'g !
u O,
>l
[ Complete verification of l
l g!
(
thermodynamic data base 2
--~_p-g zi 5>
ma y
o!
Qn -
>m O*
'l 4Last IPE (Utility) g Qlssue last source term
.n lC regulation change l l Complete thermodynamic
> 45 data confirmation E
i t
i l
l f4 I Q Staff review of last IPE I l l Complete multi-effectl validation tests i
=
ML aansotau3
=
'0FFICE OF RESEARCH PLAN TO RESOLVE SOURCE TERM TECHNICAL UNCERTAINTY AREAS I. INTRODUCTION During the Staff briefing to the Comission on the " Reassessment of the Technical Bases for Estimating Source Terms (NUREG-0956)" on July 9.1986 -
Consnissioner Bernthal requested that the Staff prepare a program plan to resolve eight areas of technical uncertainty illustrated by Table 6.1 of that report. The Conunission further requested the Staff to recommend whether an independent group of experts should be constituted to review each of the major areas and.to reconsnend a research program for resolution, if needed. 'The groups of experts would have a charter similar to that of the review group on steam explosions.
At the time of this request, the Accident Evaluation Br'anch (AEB) of the-Office of Research had already begun work on a similar plan which required cognizant staff members and contractor personnel to submit to the AEB Branch Chief research reports on each of eleven areas of technical uncertainty in draf t form (See-Attachment 1 - dated June 20,1986). These eleven areas encompassed each of the eight areas in Table'6.1 as well as an additional three areas. Attachment 1 also requested on page four (item IV) that each uncertainty area leader establish a " Technical Uncertainty Working Group" whose charter was to have been to review the state of knowledge in each area, the adequacy of the current NRC research program.for achieving resolution of i
the uncertainty to the extent needed for NRC licensing and severe-accident policy decisions, and to deterinine the additional research that might be needed to achieve this resolution. As a result of the above and the later Commission request, the Accident Evaluation Branch has revised the original list of uncertainty areas to correspond to Table 6.1 of NUREG-0956 and revised the overall plan and the expert working groups to correspond more closely to the Commission request for " independent" reviewers.
Section II of this research plan discussion gives background information on 4
and a history of the determination of the areas of uncertainty.
Section III gives a brief description of each area of uncertainty, accom-i plishments to date, expected FY 87 milestones and resources, unfunded needs, and expected milestones for FY 88 and beyond. Each of the eight areas listed in the beginning of Section II is discussed individually.
In addition, there is a section, General Support, which gives the same information for code development and uncertainty estimation programs which encompass all of the areas.
For these areas, it is impossible to apportion the costs among the l
other eight, and, therefore, they are discussed separately.
Section IV gives the background and objectives of the independent review groups that are being established to review the eight areas of uncertainty.
Information on the schedule for the review group meetings is also given. This schedule is designed to support a Comission request to Congress for additional resources for FY 88, should that action prove to be necessary.
1
II. BACKGROUND INFORMATION The eight technical areas of uncertainty of Table 6.1 of NUREG-0956 are as follows:-
- 1. NATURAL CIRCULATION IN THE REACTOR COOLANT SYSTEM 2.-CORE MELT PROGRESSION AND HYDROGEN. GENERATION
- 3. STEAM EXPLOSIONS
- 4. HIGH PRESSURE MELT EJECTION
- 5. CORE CONCRETE INTERACTIONS
- 6. HYDROGEN IGNITION AND BURNING IN CONTAINMENT
- 7. IODINE CHEMICAL FORM
- 8. REVAPORIZATION OF PREVIOUSLY DEPOSITED FISSION PRODUCTS HISTORY OF THE DETERMINATION OF THE AREAS OF TECHNICAL UNCERTAINTY J
One of the main objectives.of the Office of Research's Severe Accident Research Program (SARP) has been the identification and resolution of technical issues arising from the complex phenomena encountered during postulated LWR meltdown events. The important phenomena of such events encompass the sciences of physics, chemistry, therinodynamics, and mathematical modeling to an extent rarely encountered in practical applications of scientific principles. For example, models are needed to account for temperature ranges from 300C to over 2800C; pressure ranges from 10 psia to 2200 'sia; simultaneously-present phases of liquids, solids, and gases; p
hundreds of chemical species, compounds and eutectic mixtures; and complex geometric and structural environments. The methodology used to attack such a complex problem requires that it be divided into several phenomenological areas covering the time period of the accident sequence. After that process has been completed and the accident sequence evaluated, certain phenomena can be identified which have large uncertainties or different modeling descriptions (due to the lack of physical and chemical data) associated with them. When this occurs, a " technical issue" or " technical area of uncertainty" is identified.
In the latter part of 1983, the RES staff identified approximately 60 such technical. issues, because of the paucity of data present at that time. These l
issues ranged from the "T/H behavior of the RCS", and " Containment Leakage" from various causes, to " Changes to Emergency Response Capabilities". At the same time, interactions between the NRC and IDCOR (Industry Degraded Core Rulemaking) staffs began with the purpose of identifying technical areas of agreement and disagreement between the methodologies used by both parties in assessing severe accidents in LWR's.
Initially, the NRC staff presented, in general terms, the above 60 issues (later combined into 29) to the IDCOR staff for discussion purposes. The NRC staff also produced and presented documentation outlining their concerns in these areas.
2
~
The general phenomenological technical exchanges were completed in early 1985.
.At that t me, t e NRC/RES staff, using the discussions and results of inose i
h meetings, compiled a list of the remaining unresolved issuas, met with and received concurrence from consulting contractor personnel, and presented the IDCOR management with the revised set of eighteen issues in February,1985.
Also, at this time, control of the issue resolution procedures with IDCOR was transferred to the Office of Nuclear Reactor Regulation (NRR). This " final set of issues" was 'tenned the NRC/IDCOR TECHNICAL ISSUES.
In July of 1985 the American Physical Society (APS) published its report on radionuclide release from severe accidents at nuclear power plants (Reviews of Modern Physics, Volume 57, Number 3 Part II July 1985). They 11sted eighteen recommendations for'Tuture research which, because of their similarity to the above NRC/IDCOR issues, became known as the 4
APS TECHNICAL ISSUES.
In fact, all of these recomendations could be considered as restatements of the NRC/IDCOR issues either in more general terms or, for some cases, in more specific terms. This similarity and restatement of importance by an independent study group confirmed the significance of the effort by NRR towards resolving the NRC/IDCOR issues.
In July of 1985, the draft report NUREG-0956 was published for review and comment. This extensive study was carried out by RES over a two year period with support from the ongoing RES-sponsored Severe Accident Research Program (SARP).
In the report, eight major areas of uncertainty were identified which were essentially generalized statements of most of the NRC/IDCOR and APS issues. These areas of uncertainty were reported on and discussed with NRC staff and Contractors and have gradually assumed the role of the eight main 4
areas of technical uncertainty related to source term calculations as reported to the Commission on July 9, 1986.
1 The major areas of uncertainty in NUREG-0956 and the NRC/IDCOR issues are i
approximately the same as those areas which are driving the uncertainty in NUREG-ll50.
It is the characterization and the reduction of the uncertainties associated with these issues that will provide the principal backup for the implementa-tion of the Commission's Severe Accident Policy Statement (SAPS) over the next few years. RES staff members have been working closely with NRR staff members to provide the needed data and analytical techniques for state of knowledge application of the policy statement.
The main goals of the previous, current, and future research in this area are to:
O Conduct experimental programs and analyze the results to provide a base of data covering the range of conditions predicted to be found during
+
4 severe reactor accidents. Those experiments have in the past provided the basis for model development and will in the future provide data for i
further model development and for model validation. The experiments include in-pile tests as well as out-of-pile tests and separate effects j
tests as well as integral tests.
3 1
..,.... _ ~.
0 Develop, maintain, and apply simplified, integrated computer codes that.
describe the time history of severe accident phenomena..These simplified codes, together with estimates of the uncertainty, can provide information covering many plants and accident sequences at reasonable costs.
O Develop, maintain, and apply an integrated, best-estimate, mechanistic computer analysis package which includes all the complex source term phenomena noted above. These codes will be used to provide specific information on aspects of severe accident phenomena as well as to benchmark the simplified, integrated codes.
4
.)
III. PLAN OF ACTION NATURAL CIRCULATION IN THE REACTOR COOLANT SYSTEM.
Description Natural circulation in the reactor coolant system (RCS) of a PWR is defined as the buoyancy-driven flow circulation of steam (and hydrogen and fission gas as the core is being damaged) between the uncovered core and the upper plenum with possible counter-current flow in the hot leg piping between the vessel and steam _ generators. This kind of " multi-dimensional" gas circulation serves as a means of transferring the decay heat from the core to upper plenum structures, hot leg piping, and possibly steam generator tubes. As a result, the RCS pressure boundary may be heated to a temperature at which the structural integrity is challenged.
Natural circulation in the,RCS is likely to occur during the core uncovery and core melt period of high-pressure severe accidents in a PWR such as the TMLB' accident (station blackout with the loss of auxiliary feedwater). The Source Term Code Package (STCP) is not capable of analyzing reactor coolant system natural circulation because the flow is multi-dimensional and it is coupled with core melt progression.
This issue addresses the effects of RCS natural circulation during high-pressure severe accidents in a PWR on:
O Location, size, and timing of the res pressure boundary failure Will the failure be large enough and occur long before vessel lower head failure to preclude high pressure melt ejection and consequent direct heating that may challenge containment integrity?
Will steam generator tube rupture occur and lead to fission product bypass of the containment?
O Fission product retention and revaporization in the res before and after the failure of the pressure boundary either in the vessel or elsewhere.
O Hydrogen generation due to longer predicted times for vessel failure.
O Timing of core uncovery and core damage.
O Recommended operator actions to mitigate and recover the accident.
5
. ~.
f Sumary of Program Strategy Three key parameters determine the structural integrity of the res pressure boundary:. structure temperature, system pressure, and the. failure temperature beyond which the structure will likely fail in a few minutes (because of creep rupture). The first two parameters are calculated by the best-estimate MELPROG/ TRAC code. However, the uncertainty in the calculated temperature is much greater than that in the calculated pressure. The failure temperature of a structure such as piping or a steam generator tub'e is obtained by the creep rupture analysis, and its uncertainty is believed to be smaller than that in the calculated structure temperature. Therefore, our effort is primarily focused on the uncertainty reduction for the structure temperature history during a high-pressure accident sequence in a PWR such as TMLB'.
To quantify and reduce the uncertainty in MELPROG/ TRAC calculations on structure temperature, the code must be validated by comparing calculations with appropriate experimental results which are currently scarce. The only data on the multi-dimensional natural circulation flow is from the Westing-house 1/7-scale facility using water at ambient pressure or SF,d program and gas at pressures up to 600 psia.
(Note that this is an EPRI -sponsorE the results will be provided to NRC contractors for code validation.) Data from this. facility may not be sufficient for simulating transient high-i pressure accidents in a PWR. Therefore, additional experiments may be needed.
The staff would like to reduce the uncertainty regarding the impact of this issue in a year or so. To achieve this goal comparative code calculations using SCDAP/RELAP5 are required to repeat the same calculations being per-formed by MELPROG/ TRAC. The SCDAP/RELAPS code is more mature than MELPROG/
TRAC and has been validated against severe core damage experiments from LOFT l
FP 2 test, PBF tests, and other experiments, although the code can only analyze core melt progression prior to core slump. SCDAP/RELAPS will also be used for sensitivity analysis on the time-temperature history of the structures. Should the SCDAP/RELAPS results reasonably agree with the MELPROG/ TRAC results on the time history of structure temperatures and system prestures, the staff will have more confidence in the conclusions. On the other hand, if there is no agreement between these two codes without good reasons, the staff will have to rely more on MELPROG/ TRAC validation and this will take more time to reduce the uncertainty.
In parallel with the code calculations, additional rupture analyses are needed to determine the failure temperatures for res piping and components other than the hot leg nozzle, piping, and steam generator tubes. The size of the structural rupture also needs to be estimated.
6
Research Accomplishments to Date O
Preliminary MELPROG/ TRAC calculations indicate that natural circulation may lead to failure of the res pressure boundary long before vessel lower head failure. However, surface heating due to fission product deposition and ex-vessel flow were not modeled in the calculations.
O Creep rupture analyses have determined the failure temperatures as a function of pressure for hot leg nozzle (carbon steel), hot leg piping (stainlesssteel),andsteamgeneratortubes(Inconel). However, the rupture size is yet to be estimated.
O TRAC-PF1 and COBRA-NC in-vessel calculations have been performed to compare with MELPROG/ TRAC results for intact core geometry. Preliminary COMMIX calculations have been performed to provide support to MELPROG/ TRAC.
FY 87 Milestones at Curren't Budget Level O
Complete and document a MELPROG/ TRAC (2-D) calculation analyzing natural circulation during the high-pressure TMLB' accident in the res of the Surry Plant. The calculation will include those physical phenomena not modeled in the existing MELPROG/ TRAC calculations, namely, decay heating due to fission product deposition on structure surfaces, pilot operated relief valve (PORV) cycling open and closed, and ex-vessel flow in the hot leg piping and steam generators. Another calculation, similar to the one above but with a stuck-open PORY will also be performed and reported.
These two calculations will aid in estimating the uncertainty of the structure temperatures calculated by the MELPROG/ TRAC code under TMLB' conditions.
O Complete and document a SCDAP/RELAP5 calculation and sensitivity study for the TMLB' accident in Surry (with cycled PORV). Results will be used to compare with the MELPROG/ TRAC calculation to help in estimating the temperature uncertainty calculated by these codes.
If both codes predict a hot leg nozzle temperature high enough to cause an induced LOCA before lower head failure or should reasonable agreement exist between these two codes, this will give us some confidence in the results.
FY 87 Resources Allocated - $1000K FY 87 Unfunded Needs - $400K 0
Perform creep rupture analysis for the surge line, the hot leg instrumen-tation penetrations, and the weld connecting the hot leg piping to the 7
f
'n vessel; perform simple tests to obta'in rupture data of the weld material
~
for the use'in the analysis. -This information will provide a basis for evaluating other potential RCS failure locations.
O Analyze the break size due to creep rupture in reactor coolant system piping and components.
O Perform MELPROG/ TRAC calculations for the TMLB' sequence with the maximum pump seal LOCA expected in Surry to evaluate impact of pump leakage on-natural circulation parameters.
- 0 Analyze cooling of the vessel and system depressurization by accumulator water injection into the cold legs.
O Perform COMMIX. calculation in support of MELPROG/ TRAC calculations.
O Review adequacy of current test matrix for the EPRI-sponsored program and determine the basis'for an expanded program and cooperative resource plan.
FY 88 and Beyond Milestones O
Complete expanded natural circulation experimental program under PWR high pressureconditions(Dec.87).
s i
0 Evaluate need for and cost-benefit of more advanced experiments (FY 88).
i 0
Validate and improve MELPROG/ TRAC with the above data (FY 89).
O MELPROG/ TRAC sensitivity analyses (FY 89 and 90).
0
' Additional MELPROG/ TRAC full-size PWR calculations for risk-dominant high-pressure accident sequences (FY 90).
F I
8
+
CORE MELT PROGRESSION AND HYDROGEN GENERATION Description In-vessel core-melt progression is concerned with the state of the reactor core from core uncovery to reactor-vessel failure, including in-vessel hydrogen generation and the thermal and mechanical attack upon the reactor structure and the reactor vessel. The state of the core during in-vessel core-melt progression is the primary oeteminant of in-vessel hydrogen generation, fission-product and aerosol release, the mode of vessel failure, and the state of the core debris at vessel failure. The state of the core debris includes the mass of debris expelled into the reactor cavity, and the distributions of debris temperature, composition, and melt fraction. These distributions provide the. initial conditions for assessing the fraction of the core participating in high pressure melt ejection and direct containment heating, core-concrete interactions and ex-vessel fission product release, and other threats to containme.nt integrity. The current substantial uncertainties regarding in-vessel hydrogen release and transport and the threat from core-debris to containment integrity are in large part driven by the substantial uncertainties in the state of the core during in-vessel core-melt progression.
These uncertainties are in turn reflected in the parametric inputs on the 4
state of the core currently used in the Source Tenn Code Package (STCP) and other accident-analysis codes.
The current IDCOR and NRC (STCP) analytic ~al models of in-vessel core-melt progression are greatly simplified parametric representations of very complex processes. In these models, the three significant separate and distinct material relocation processes in core-melt progression are treated as a single i
" core slump" that occurs at an assumed " core-slump" temperature. These separate relocation processes are:
- 1) relocation of the molten unoxidized Zircaloy cladding that also includes some dissolved fuel (sometimes called liquified fuel) to form a possibly permeable metallo-ceramic hard-pan layer across the lower core; 2) collapse of the remaining free-standing ceramic columns of cracked fuel pellets fand oxidized Zircaloy cladding) onto the hard pan to form a rubble bed; and 3? failure of the lower molten-corium-pool crust by thermal, chemical, and mechanical attack to release the molten core-material mixture into the lower plenum.
Substantial confidence exists in the modeling of the steam /Zircaloy reaction in the reactor core as long as the original geometry is maintained.
Substantial uncertainties arise, however, when the unoxidized Zircaloy melts and starts to relocate. These uncertainties include the threshold for relocation, the extent to which flow channels will begin to block, whether cladding material will run out of the hot zone, how much oxidation and hydrogen generation will occur as it moves, and whether the relocated cladding will subsequently be reheated and exposed to steam. Because of these uncer-tainties, results obtained by the industry's IDCOR and NRC analysts have of ten differed substantially. For example, for a PWR small-break LOCA sequence without ECCS, IDCOR models yield 200 kg of hydrogen produced, whereas, the the STCP calculation yields 450 kg of hydrogen; a significant difference when translated into peak containment loading pressures from burned hydrogen.
I 9
l l
l Sumary of Program Strategy i
The information currently available on in-vessel core-melt progression from past research is almost entirely for the early phase of the process, through metallic (primarily Zircaloy) but not ceramic (UO,, Zr0,) material relocation and including hydrogen generation. Almost the onTy data available on the later phase of the in-vessel core-melt progression process, which involves ceramic debris relocation, molten corium pool formation and crust melt-out,.
and melt attack on the reactor structure and the reactor vessel, come from the
.TMI-2 core examination. THI-2 has furnished valuable information, but this information relates directly to an unusual set of conditions for accident recovery by core reflooding that are not well characterized, and not for the conditions of the unrecovered accidents currently analyzed and modeled for risk assessment.
The detailed mechanistic SCDAP (Severe Core Damage Analysis Package) code, for recovered accidents and for detailed fuel-damage experiment analysis, and the MELPROG (Melt Progression) code, for core-melt progression through vessel failure in unrecovered accidents, were developed on the basis of existing experimental data. The later stages of in-vessel core-melt progression treated in MELPROG are based on analytical modeling of the key processes and the limited relevant available data, primarily from the TMI-2 core examina-tion. Experimental priority in resolving the uncertainties in core melt i
progression is based upon obtaining data on the governing processes in the later phases of core-melt progression, although a few significant uncer-tainties regarding the early phase still need further research. Much of this needed new data will be acquired in a new program consisting of out-of-pile separate-effects experiments and analysis for MELPROG assessment and validation.
The major areas of phenomenological uncertainty regarding in-vessel core-melt progression and the primary research programs to resolve these uncertainties are as follows:
0-Molten Zircaloy relocation Threshold and mechanisms of relocation 3
AddressedinnewCORAfueldamageexperimentsatKfK(Germany).NRU tests, ACRR Damaged Fuel (DF) tests, PBF data, some MELPROG l
Validation out-of-pile experiments. SCDAP and MELPROG modeling 0
Effects of PWR Ag-In-Cd and BWR B C Control-Rod Materials 4
Lower temperature alloying, liquifaction, and relocation f
Potential for opening pathways and also forming blockages 1
10 i
e
-..-.,-n..
a
-___n _._ _,_ _ _ _ _ _ _
Addressed in KfX experiments, PBF SFD 1-4 data, ACRR DF-3 and DF-4 data, SCDAP and MELPROG modeling 0
Longer-term hydrogen generation with Zircaloy relocation.
Addressed in NRU full-length bundle tests, KfK CORA fuel-damage experiments, SCDAP and MELPROG modeling.
~
O.
Formation of across'-core metallo-ceramic hardpan in fuel-rod stubs by relocated saaterial.
Permeability of blockage, effect on steam flow and hydrogen generation, depth and strength of blockage Addressed in data from PBF, ACRR DF, NRU, KfX CORA, and SCDAP and MELPROG modeling 0
Collapse of free-stan' ding columns (rods) of cracked-pellets onto metallo-ceramic hard-pan to form debris bed.
Collapse did occur at TMI, not in PBF, ACRR, or NRU tests Addressed in new out-of-pile MELPROG Validation experiments,' new ACRR Melt Progression (MP) tests, and in MELPROG modeling.
O Characteristics of growing molten corium pool, insulating crusts, and surrounding debris.
Compositions, temperature distributions, melt fraction Addressed in new out-of-pile MELPROG Validation experiments in Large j
Melt Facility (LMF), new ACRR MP tests, and in MELPROG modeling.
Collapse of the rods strongly accelerates the growth of the central, molten fuel pool.
O Melt-out failure of pool crust and relocation of melt.
Mode of melt out Characteristics of Corium pool and debris at melt out Addressed in new out-of-pile MELPROG Validation experiments in 'LMF, new ACRR MP tests, and in MELPROG modeling.
0 Explosive and Non-Explosive Steam Generation from melt interaction with lower-plenum water.
Steam and hydrogen generation Effect on core-melt progression 11
Potential for direct steam-explosion failure of the reactor-vessel lower head Addressed in steam-explosion research program and in MELPROG modeling 0
Failure of reactor structure and the reactor vessel under core-melt thermal and mechanical attack.
Mode of failure of reactor structure and vessel Failure of vessel penetrations Particular failure of BWR di.stributed core-support structure Characteristics of-the core debris and melt at vessel failure
^
Characteristics of the ejected stream or jet (pressurized ejection) of melt and debris Addressed in MELPROG modeling Research Accomplishments To Date Results of four PBF tests, three-ACRR experiments, three full-length NRU tests, corollary analysis, and results of the German fuel-damage experiments at KfK have provided the following information on in-vessel core-melt-progression:
0-A rapid autocatalytic-oxidation temperature transient starts at about 1700K, with local fuel temperatures rising above 2500K in tens of seconds L
unless limited by steam starvation.
l 0-Molten unoxidized metallic Zircaloy cladding with dissolved U02 (liquified fuel) relocates downward and refreezes to form a metallo-ceramic hardpan or crust that has blockage formation potential.
0 0xidation heating decreases with the relocation of the molten cladding downward and away from the high-temperature region.
0 In a recent coolant bo11down test in NRU with a full length (twelve-foot long) test fuel bundle, hydrogen generation continued for a thirty-minute high temperature hold without being cut off by steam' blockage from Zircaloy relocation.
12
m
~-
l O
Examination ~ of the TMI-2 core by DOE has demonstrated the three separate and distinct major material-relocation processes in core-melt progression: 1) relocation of the molten unoxidized metallic clad and dissolved UO., to form a metallo-ceramic hardpan; 2) collapse of the free-standing columns (rods) of cracked ceramic pellets onto the hardpan; and 3) melt-through of the crust (hardpan) by the newly-molten corium mixture.
0 In-vessel natural circulation can significantly affect core-melt progression by producing more uniform core temperatures and by delaying the rapid oxidation transient, by melting upper-plenue structure, and possibly by failing the pressure boundary.
l 0
The oxidation models in current codes are adequate for oxidation and hydrogen generation prior to molten Zircaloy relocation.
0 With PWR silver-indium-cadmium control rods, the silver alloy melts and runs down the rods rather than imediately forming aerosols as previously assumed.
0 The cadnium vaporizes upon control-rod failure, and it later forms an aerosol upon leaving the core, for later deposition in the upper plenum.
O With BWR boron-carbide control blades, the' boron carbide alloys with and liquifies the stainless steel about 200*C below the stainless-steel melting point. This prevents significant steam oxidation of the boron carbide. The liquid alloy attacks and dissolves Zircaloy, and has the potential for blockage of the channels between the BWR can walls.
O The three ACRR tests, using visual diagnostics, showed that a dense tin aerosol is formed upon Zircaloy melting that can transport fission products.
O The information from the above experimental programs has provided strong support for our understanding and current modelling of the initial meltdown phase of core-melt progression, as embodied in the SCDAP code, p
0 Also, a more mechanistic core-melt progression model has been developed for the MELPROG code based on experimental information from the PBF, the l
ACRR, the NRU tests and the German KfK experimental programs.
l FY 87 Milestones at Current Budget Level PBF 0
Issue final Experiment Report on the SFD 1-1 test (trace-irradiated fuel).
l 13
c 0
Complete Post-Irradiation Examination (PIE) of the SFD 1-3 (high-burnup fuel) and the SFD 1-4 (high-burnup fuel, PWR Ag-In-Cd Control Rods) test-fuel bundles, and issue report on Non-destructive Examination.of the SFD 1-3 Fuel Bundle.
O Issue an Integrated Sumary Report on hydrogen generation in the PBF
-Severe Fuel Damage tests.
NRU O'
Issue Experiment Report on test FLHT-2, and Preliminary' Report on test FLHT-4 on fuel relocation and hydrogen generation.
O Perform test FLHT-5 in NRU to study fuel relocation and hydrogen generation with sustained heating to higher temperatures.
n ACRR O
Perfom Experiment DF-4 (BWR B C Control Blade) and Issue DF-4 Experiment 4
Report.
0' Issue final Experiment Reports on Experiments DF-2 (Moderate Oxidation Rate), DF-3 (PWR Ag-In-Cd Control Rod), and DF-4 (BWR B C Control' Blade),
4 andissuereportonthe(PIE)ofExperimentDF-3.
MELPROG Validation Analysis and Experiments 0
Issue report on MELPROG comparison with the ACRR DF-1, 2, and 3 experiments.
I O
IssuereportoncomparisonofMELPROGfission-product (FP) behavior
' ~
module VICTORIA with results of the early and recent ORNL FP experiments, the ~new in-pile ACRR ST experiments, and other FP data.
i 0
Issue interim report on MELPROG analysis of the initial German CORA experiments at KfK on Zircaloy relocation and hydrogen generation.
l 0
Perfom intermediate-scale experiments on the attack of molten Zircaloy on thin zirconia shells and on Inconel grid spacers, and report on the results.
O Perform intemediate-scale (50kg) crust-failure experiment with molten coriu.t and report on the results.
O Develop and improve MELPROG models on molten Zircaloy relocation and crust failure.
O Begin MELPROG analysis of the TMI-2 accident as part of the OECD-NEA TMI-2 standard problem exercise.
FY 87 Resources Allocated - $3,934K 14 i
FY 87 Unfunded Needs - $1350K 0
PIE of NRU FLHT-2 and FLHT-4 test fuel bundles to obtain
$250K physical data on fuel relocation and blockage to support thermal hydraulic and hydrogen data.
O Accelerate and complete and report on initial MELPROG
$100K analysis of the TMI-2 accident for earlier use in MELPROG validation.
O PIE of PBF SFD 1-3 and SFD 1-4 lower plenum debris for
$100K information on metallo-ceramic hard pan fonnation 0
Perform large-scale (200kg) LMF test on crust failure
$400K under molten corium attack to expedite late melt-progression information for MELPROG validation.
O Prepare to perform ACRR Melt Progression (MP-1)
$500K Experiment in early FY 88 to expedite late melt-progression j
information for MELPROG validation.
FY 88 to FY 90 Program 0
ACRR Perform two Melt Progression (MP) tests per year on governing phenomena in late-phase core-melt progression to vessel failure (ceramic pellet collapse, pool growth and melt out of insulating i
l crusts,etc.)
0 NRU Perform one test per year to obtain data using 12-foot long bundles on continuing hydrogen generation after Zircaloy relocation.
O PBF Complete analysis of results including PIE of the last two Severe Fuel Damage (SFD) tests with high-burnup fuel and issue final experiment reports.
Issue sumary reports on the application of the results of the four PBF SFD tests to the analysis of severe accidents.
O KfK (German results in joint international program)
Complete and issue reports on the series of twelve CORA tests on molten Zircaloy relocation and hydrogen generation.
O HELPROG Validation Analysis and Experiments i
15 l
Continue comparison of MELPROG with data on in-vessel core-melt progression as they become available.
Complete and issue MELPROG analysis of the TMI-2 accident as part of the OECD/ DOE TMI-2 Standard Problem exercise. '
Perform out-of-pile experiments on crust failure and corium-pool meltout with up to 200kg melts in the Large Melt Facility (LMF) and in smaller. facilities.
Perform out-of-pile experiments on other key phenomena in the latter phase of core-melt progression. including the collapse of free-standing columns (rods) of cracked ceramic pellets to form a debris bed.
l 4
i l
16 l
lc
STEAM EXPL0SIONS Description The contacting of a mass of hot liquid with a low-boiling-point cold liquid may produce a purely thermal explosion which is characterized by very-rapid fragmentation, mixing, and heat transfer between the liquids and by the rapid generation of vapor and formation of a shock wave. When the cold liquid is water, the explosion is commonly called a steam explosion. The principal
. reactor-safety significance of steam explosions is the probability that a sufficiently energetic steam explosion will occur when the molten core slumps into the lower-plenum water to eject the vessel upper-head through the containment as a large-mass missile, producing high-consequence early con-tainment failure. Such an occurrence, comonly called alpha-mode containment failure, requires acceleration of a mass of material or. slug overlying the steam explosion to impact upon and detach the upper-vessel head. The mechanism of slug accelera, tion and impact was demonstrated in the SL-1 accident, where the slug impact severely strained but did not fail the upper-vessel head. There is general but not complete consensus among experts that the alpha-mode failure has a conditional probability of less than about 0.01.
This was the conclusion of the Steam Explosion Review Group (SERG), a panel of experts convened in August,1984 by NRC to review this question.*
The report of this review group was issued as NUREG-1116. Further research primarily by T. Theofanous at UCSB and by W. Bohl at LANL has supported this conclusion. Recent analysis with potential contrary implications has indicated that in-vessel natural circulation can result in significant heating of the upper vessel head with resultant loss of strength of the head and the head bolts. The effect of this upon the probability of the direct alpha-mode containment failure by steam explosions has not yet been evaluated. Analysis also indicates that the probability of failure of the vessel lower head by direct shock over-pressure from an energetic steam explosion is much greater than the probability that of alpha-mode failure of the containment by steam-explosion-induced slug acceleration and impact. This direct failure of I
the lower head by a lower-plenum steam explosion strongly reduces the probability of alpha-mode containment failure by venting the slug-accelerating steam pressure, and leads to a lower-consequence core slump into the concrete reactor cavity.
It should be noted that there are other generally less serious but still significant consequences of steam explosions upon the progression of accident sequences besides alpha mode containment failure and direct shock-failure of the vessel lower head. The steam-explosion produced fragmentation and dispersal of the melt can produce additional fission-product release and steam-oxidation of the metallic component of the melt to generate additional hydrogen. The debris dispersal and steam generation in the steam explosion i
l
- 51nce the meeting, W. Bohl has lowered his upper-limit probability from 0.1 to 0.01 on the basis of his further research.
l l
e i
4 affect core-melt progression. Recent analysis has shown a possibility that an ex-vessel steam explosion in the water-flooded reactor cavity following vessel melt through in a MARK-III BWR can fail the drywell, so that the suppression pool is bypassed.
Extensive research over the last twenty-five years has failed to produce a mechanistic model of the steam explosion process capable of accurate quantitative prediction. A non-quantitative understanding of this complex process has been developed, however, in tenns of thermal detonation in analogy to chemical detonation theory. A data base on steam-explosion behavior has also been developed from drops of thermitically-generated fuel melts into water or sodium and with non-reactor materials for melt masses up to about 30 kg.
In energetic steam explosions the conversion ratios of thermal energy in the melt into mechanical work are typically a few percent, with a few cases of conversion ratios up to about a quarter of the thirty percent thermodynamically possible. Recent research has been concerned with potential limitations on conversion ratios with very-large melt masses from limitations 4
on the pre-detonation breakup of the melt mass and mixing with the water and steam. The most recent work has concentrated upon the break up and mixing of jets of melt coming through the holes in the intact core-support plate, as occurred at TMI-2.
It is thought that there was no steam explosion at TMI-2 i
because the vessel pressure of about 1800 psi at the time of melt ejection into the lower-plenum water was too high for a steam-explosion to be triggered.
Summary of Program Strategy In the absence of a truly predictive mechanistic model of steam-explosion behavior and of the prospects for achieving such a model in the NRC steam-explosion research program, evaluation of the probability of alpha-mode i
containment failure has been performed by the use of semi-empirical modeling i
of part of the steam-explosion process, the direct use of experimental data for other parts, and by analysis of the steam-explosion work potential (slug kinetic energy) required of an in-vessel steam explosion to produce alpha-mode containment failure. There is considerable engineering judgement and uncertainty involved in this process. This is reflected in the wide range covered by estimates of the upper and lower uncertainty bounds for this probabilitygivenbythedifferentmembersofthe1984SERGreviewgroup(0to O.1,sincereducedtoOto0.01).
The strategy currently used in assessing the alpha containment failure mode from a lower-plenum steam explosion is to evaluate the separate steps in the process and to treat them as independent factors. These separate factors are:
[
- 1) The mass, temperature, and composition of the core melt dropped into the lower-plenum water are assumed.
In the future these quantities can be based on analysis with the new mechanistic MELPROG core-melt-progression code. 2)
The mass of melt that breaks up and mixes with the water and steam to form a thennaly-detonable mixture is taken from current models which are based on a very-limited small-scale data base.
It is thought that a scale-dependent mixing process limits the masses of core melt and water that can be mixed l
i 18
(
i
i 2
before detonation, and therefore limits the energetics of the resultant steam explosion. This mixing process is the principal area of current steam-explosion research. 3)Ahigh-pressurecut-offtosteam-explosion l
triggering is used that is based upon very limited experimental data. This pressure cut off is thought to be the reason why an energetic steam explosion did not occur at TMI-2 (as opposed to the observed non-explosive rapid steam generation or " steam spike"). 4) The conversion ratio of thermal energy in the melt into mechanical work in the steam explosion is taken in the range from about half that thermodynamicly possible (40%) to about an order of magnitude less than half, primarily on the basis of experimental results.
- 5) The steam-explosion work potential actually required to produce alpha-mode l
containment failure is evaluated. This process includes acceleration of the overlying core debris as a slug to impact the upper vessel head, to fail the head bolts, and to eject the separated upper head through the containment as a large-mass missile.
The purpose of the current-research on steam explosions is to reduce the j
largest of the uncertainties in the factors in evaluating the probabilities of alpha-mode containment failure listed above.
In response to recomendations of the SERG panel, the research has concentrated upon limitations on the masses of the interacting liquids in the pre-explosion mixing process that would occur at full reactor scale. More recently, the THI-2 core examination showed that the melt poured through holes in the undamaged massive core-support forging and flow distributor plates as individual jets without significant damage to the plates. Most of the current NRC steam-explosion
~
research consists of non-explosive mixing experiments with single and multiple jets of corium in water. This work is providing a data base for developing models of the pre-detonation mixture for analysis of limitation on the work potential (energetics) of a subsequent detonation because of mixture composition.
For some time there has been debate on whether large-scale (about 1,000 kg) mixing and explosion tests are needed for application of steam-explosion mixing models and conversion-ratio results to full reactor scale (about 50,000 kg melts).
In 1984 the SERG panel recomended that such large-scale tests were premature before more small-scale experiments and model development were t
completed. With the recent experiments upon the mixing of arrays of melt jets in water, the problem may be solvable by a one-dimensional treatment, and large-scale testing may not be required.
4 Further small-scale experiments on the detonability limit and the pressure threshold for suppressing spontaneous triggering of the steam explosion are currently planned. Experiments on steam-explosion conversion ratios (energetics) for multiple corium jets injected into water are also currently planned. Analysis of past experiments is continuing and future experiments for specific conditions of mixing may be required.
If a steam explosion to disperse the melt does not occur during the fall of the melt through the lower-plenum water, then the more dense melt agglomerates in the lower head beneath a vapor blanket and the lower plenum water.
If such li 19 4
5._.,__,_.___._..
a metastable stratified system occurs at low pressure, it has been demonstrated that an energetic steam explosien can occur. It is known that mixing and detonation propagate together along the interface surface, but the
- ~
mechanics of the process is not known. Small experiments have been started to provide basic data on the stratified contacting mode, and these experiments should be completed and the results analyzed.
t i
Recent analysis and experiments have indicated that in-vessel natural circulation during core uncovery and core-melt progression can produce substantial heating and weakening of the head bolts and the upper plenum head itself. Weakening of the head bolts will lower the threshold steam explosion i
energy necessary to produce alpha-mode containment failure. Weakening of the upper head may also produce head failure under slug impact in a " petaling back" mode, preventing ejection of the head as a large-mass missile, and possibly preventing alpha-mode containment failure altogether. A high i.
priority in future steam-explosion research is analysis of this in-vessel natural-convection heating-of the upper-head and head bolts using the new mechanistic MELPROG/ TRAC 2-D core-melt progression code. The effects of this heating upon the mechanics of slug impact and upon the consequences relevant j
to alpha-mode containment failure then need to be examined, i
I T. Theofanous of UCSB has been performing an integrated study of all the processes contributing to the alpha containment-failure mode for NRR. His 1
l preliminary results which are now undergoing outside review, are consistent 34 with his earlier 10 es'timate for the upper limit for the alpha-mode
~
containment failure probability made in the SERG review, the low values reflecting particularly Theofaneous' analysis of the pre-detonation mixing process.
In any case, such integrated analysis of the entire process is very valuable and needs to be continued along with future research.
i The need for, the priority of, and the content of further research on steam explosions is to be reviewed in FY87. This review will include the needs for further research to reduce the uncertainties in assessing the probability of the alpha containment failure mode. Also to be considered is possible research to better define steam-explosion pressure-time histories for evaluating direct shock failure of the lower vessel head; ex-vessel steam-explosion consequences, especially for MARK-III containments; the effects of explosive and non-explosive rapid steam generation on core-melt progression; and steam explosion effects upon hydrogen generation and fission-product release.
It does not appear that research to achieve truly predictive mechanistic modeling of steam explosions is achievable within potentially available NRC resources, nor does this scientificaly difficult task appear to be programatically justified.
Research Accomplishments to Date r
Current knowledge of steam explosion behavior is based on extensive international research on thermal explosions in both LWR systems (steam explosions) and LMFBR systems over the last twenty-five years. The NRC steam 20 e-y
~ar -w---9
---,----,,,,,,,,,+-.m.w--.---_...-m-,,.,-,-,,,
f
+
explosion research is a major part of this research, and has included the following:
4 0
Twenty five tests with pours of up to 30 kg of themitic m$lt into7 water -
under varying conditions. Some of these tests had transparent walls and cinematography to observe the mixing and detonation process. Energetic steam explosions occurred in about half of these tests, and prcvided
~
measurements of the thermal-to-mechanical energy conversion ratio.
0
-Experiments have begun with themite melts and with simulant materials en the process of break up of melt jets in water and the mixing of the' fragmented melt with water and steam.
O Analysis of the alpha containment failure mode process of steam-explo'sion acceleration of an overlying slug of material into impact with the vessel upper head and ejection of the head through the containment, including analysis of the probability of this occurrence as reported in
'i NUREG/CR-3369 and NUREG/CR-4500.
1.
O Steam-explosion model development including the TEXAS and IFCI codes and the development of theoretical models of the pre-dstonation mixing process for application at large scale. No large-scale data exist for testing these mixing models.
l FY 87 Milestones at Current Budoet Level O
Perfom intermediate-scale (up-to-50kg) experiments with cinematography on mixing and steam explosions from the injection of single and multiple jets of iron-alumina thermite melt into water under sca' led prototypic,
geometry and lower-plenum conditions. Analyze data, compare with mixing models, and report results.
O Complete small-scale (up-to-10kg) one and four jet mixing tests with urania-zirconia thermite melts in water to quantify celt mixing, quer ch rate, and debris characteristics. Analyze data, compare with' mixing models, and report results.
O Complete and report on analysis of previous steam explosion experiments in the FITS test facility with themite melts.
O Continue development of the IFCI 2-D, 4-field explosive and non-explosive i
molten-fuel-coolant-interaction model in conjunction wtth the MELPROG core-melt-progression code development and assessment program.
FY 87 Resources Allocated - $567K FY 87 Unfunded Needs - $260K 0
Analysis of the effects of in-vessel natural circulation heating upon the temperature and strength of the head bolts and the upper head itself 21
usingthe2-D(r,z)MELPROG/TRACcode. Head-bolt heating can lower the slug-energy threshold for alpha mode containment failure, but weakening of-the upper head may eliminate alpha-mode containment failure completely-by producing a petal-like failure of the upper head itself under slug impact rather than ejection of'the detached upper head through the containment as a missile. Also needed is analysis of the natural-circulation heating effect upon the alpha-failure-mode probability.
O Completion of small-sca'le experiments on the stratified mode of thermal explosions and analysis and reporting of the results, to provide a small basic data base for treating this contact mode.
FY 88 to FY 90 Program The needs~and priorities for further research on steam explosions is to be reviewed in FY87. - The siz~ and content of future research programs will be e
determined by this review. Here it is assumed that the current programs to reduce the uncertainties in assessing the probability of alpha-mode containment failure will be carried to fruition, but that major new initiatives will not be undertaken in such areas as large-scale (1,000kg) mixing and explosion testing, the stratified contacting) mode, specialex-ves
- hydrogen generation and fission-product release. -The planned continuation of current programs is:
0 Complete analysis and reports on expartments with visual diagnostics on mixing mechanisms of multiple jets or corium in water.
O Complete experiments on steam-explosion energy-conversion ratios with
- multiple jets of corium in water and analyze and report results.
l 0
Develop a model for the mixing of multiple-jets of corium in water for use in steam-explosion energetics assessments.
tests.in the FITS (Fully Instrumented Tests) previous steam-explosion Complete and report on detailed analysis of i
0 facility.
i 1
0 Perform small-scale experiments on a pressure cut off in steam-explosion triggering.
O Complete development of the 2-D, 4-field IFCI (Integrated Fuel Coolant Interactions) model for both explosive and non-explosive rapid steam generation. IFCI is to be a module in the mechanistic MELPROG core-melt progression code and is also to be used in stand alone form.
l i
0 On the basis of current data, report on the effect of steam-explosions i
upon hydrogen generation and fission-product release.
22 l
~>
s.,
O Perform and report on integrating analysis on the effect of new 4 #..
steam-explosion research results on the uncertainties regarding the probability of alpha-node containment failure and on other
.y_
J risk-signficant reactor safety problems.
')
. ;t.
.\\ ' p.
a 1
L i -
e 1
J
,8 s
e e
N k'
f h.
l i
)';
. r 4
a i
l 23 i
i HIGH PRESSURE MELT EJECTION
. Description This issue is critical to containment failure timing for high-pressure se-quences, such as TMLB' in large-dry PWR containments..It involves
. high-pressure ejection of molten core material from the reactor vessel into the region beneath the vessel and into various subcompartments of the containment building. The melt consists of oxides and unreacted zirconium and iron. The stored energy in the melt consists of the sensible energy of the melt and the chemical reaction energy of the various components assuming that they can react chemically with either the oxygen or the steam within containment. The melt can also transfer energy to water in containment, either to water droplets suspended in the atmosphere or to water collected in liquid pools within containment. Calculations have indicated that for melt at about 2500K, assuming reaction with oxygen, that 2/3 of the energy is chemical, the remaining thermal. The melts may first react with steam, if local conditions favor this, and thereby produce hydrogen. However, in a t
TMLB' sequence, at this point current information indicates that only about 4
20% or less of the metal phase will remain unoxidized, leading to lesser amounts of hydrogen production via this mechanism. The hydrogen may then burn at some later. time at a different location in containment if conditions permit.
In order to predict the containment response, one must follow the motion of the melt through the various subcompartments of the containment, while computing the integrated release of energy from the melt to the con-tainment atmosphere, the quantity of hydrogen produced during the time period that the melt-is suspended and the energy transfer to water in containment.
These various exchange processes must be integrated into prediction of the pressure and temperature response of the containment building from direct containment heating (DCH) phenomena.
The potential for " direct heating" of the containment atmosphere (DCH) is being investigated by the NRC in experimental and analytical programs which are underway at SNL and at BNL.
The issue of direct heating is closely related to the issue of in-vessel natural circulation.
If recirculating flow patterns within the reactor coolant system result in a failure of a sufficient size at alternate locations such as the hot leg or the steam generator tubes, then the RCS will depressurize prior to melt through of the lower head and the core debris will not be rapidly dispersed from the reactor cavity. However, investigation into phenomena which would lead to pressure boundary failure and system depressurization prior to core melt may not eliminate the residual probability of containment failure by high pressure melt ejection.
Summary of Program Strategy Direct containment heating is a complex physical and chemical process.
Its outcome depends on a host of variables such as:
RCS pressure, melt temperature, amount and composition of molten core materials, size of debris particles, presence of water, plant configuration, and the composition of the 1
24
containment atmosphere. This issue is relatively new and no existing calculational tools are capable of modeling the transport of fuel debris in the various compartments between the vessel cavity and relatively unobstructed, upper part of the containment' building. Similarly sufficient test data from separate effects experiments and scaled-down test facilities are not-available for evaluating existing codes or to provide the basis for developing new codes or improvements to existing codes.
SNL has conducted 1:20 scaled System Pressure Injection Tests (SPIT) and 1:10 scaled High Pressure Streaming Tests (HIPS) to confirm the potential for containment pressurization and aerosol generation. However, these investigations could not quantify the effects because the tests were conducted without a pressure vessel to simulate the containment volume. During FY86, a large ( 100m3) steel vessel, named Surtsey, was installed at SNL for this t
purpose. The shakedown test for the facility and the first matrix test have been completed. Eleven tests are currently planned in the Surtsey facility.
These tests will provide a data base for understanding the phenomena and evaluating existing computer codes or developing new codes for calculating molten fuel transport from the reactor vessel into the reactor cavity followed by the various pathways in the subcompartments leading to the containment.
Specific features covered in the currently planned tests are:
1)
Rates of exothermic chemical reactions and heat transfer 2
Effect of water in the reactor cavity 3
Effect of structures on debris dispersal and transport 4
Aerosol generation and transport 5
Burning, promoted by dispersing hot debris, of hydrogen that may preexist in the containment atmosphere.
A program at BNL will investigate subissues that are important to high pressure melt ejection but are either difficult or too expensive to be studied l
in the Surtsey facility. An example of such a subissue is the core debris t
dispersal. The high temperature ( 2500*K) melt simulant used in Surtsey tests e
i makes it difficult to observe trajectories of dispersing debris. Further, many tests will be needed to study sensitivities of the debris dispersal to various conditions predicted for different accident scenarios and the relatively high cost of Surtsey tests makes it impractical for such studies.
Therefore, small scale tests in the BNL program are designed to narrow uncertainties regarding core debris transport and mixing in the containment atmosphere.
t The analytical strategy involves the use of existing methods for calculating l
transport processes which will be compared with test data. Based upon these initial results, if appropriate, more advanced methods will be developed to reduce uncertainties. At the end of each stage of model improvements, best estimate analyses will be made of direct heating loads in selected containments for assessment of uncertainty.
25
Research Accomplishments To Date O
High pressure ejection tests were conducted to investigate the melt-dispersal hypothesis predicted in the Zion Probabilistic Safety Study.
0
' Preliminary results not only confinned that core debris would be swept from the reactor cavity, but also will cause direct heating of the containment atmosphere that could' threaten containment integrity.
O Preliminary runs of the MELPROG/ TRAC code indicate that natural circulation may indeed lead to early failure of the reactor coolant ~
system boundary thus precluding high pressure ejection of the molten core. However, this result is very preliminary and extensive code validation and further analysis will be needed to resolve the uncertainty in the results.
FY 87 Milestones at Curren't Budget Level O
Complete documentation of the Systems Pressure Injection Tests and High Pressure Streaming tests.
O Conduct three high pressure melt ejection tests in the Surtsey vessel to obtain data on major uncertainty parameters in direct containment heating for transport model assessment and improvement, including the effects of structures in the reactor cavity keyway and define flow paths.
l 0
Conduct small scale tests at BNL to investigate important parameters in core debris transport and mixing phenomena such as dispersal and deentrainment for use in developing and testing advanced transport models.
i 0
Complete initial phase of analytical model evaluation including com-parisons with Surtsey test data. Determine need for and select advanced analysis method.
O Document all of the above via test reports, topical reports, and the l
regular SNL and BNL bi-monthly reports.
FY 87 Resources Allocated - $1740K FY 87 Unfunded Needs - $200K to accelerate test program l
FY 88 and Beyond Milestones O
Complete 11 tests in the current high pressure melt ejection test matrix using the Surtsey facility (Sept. '88)
I i
26 l
O Conduct tests to quantify ex-vessel hydrogen generaticn due to metal (Zr, Fe) - steam reactions in and around the reactor cavity.(Sept. '89) 0 Document test results (thru Sept. '90)
O Complete separate-effect tests and document the results (Sept.' '88)
~ 0 Complete development of analytical models for high pressure melt ejection analyses and validate these models against test results (thru FY 90).-
0 Apply experimental and analytical results for assessment of high pressure melt ejection uncertainties to determine revised impact on contribution to risk for specific plant types.
(FY 88 FY 89 and FY 90).
O Determine extent to which specific plants cannot be categorized by a reduced number of " standard" plants for assessment and evaluate need for additional plant analysis and confirmatory tests.
(FY 88 and FY 8,9).
I 27
CORE CONCRETE INTERACTIONS Description This issue is associated with (a) the magnitude and mechanism of energy
-transfer fron: molten debris to the concrete, to the containment atmosphere, and to overlying water pools (if present), and (b) ex-vessel fission product and aerosol release during the core / concrete interaction process. The issue can potentially impact the mode and timing of containment failure and the ex-vessel release of fission products..In the case of the BWR Mark I. molten core debris could spread on the reactor cavity floor and melt through the dry well liner wall, leading to containment by-pass and higher source terms.
In the' case of PWRs, gases generated from concrete decomposition including hydrogen can lead to containment failure. This is an area in which important differences exist in the modeling assumptions used in the IDCOR and NRC analyses because of the current limited understanding of the phenomena.
Complex themal-hydraulic ind chemical phenomena occur when molten core debris reacts with concrete; these processes are interactively modeled in the CORCON N002 code to produce containment-loading parameters such as the composition, rate and temperature of gases generated. Concrete decomposition gases sparge through the molten debris and generate aerosols when the bubbles break at the surface. The VANESA code models these aerosol-generation processes to predict the aerosolized fission-product release to containment from which the ultimate radiological source term is computed. There are many gaps in.(1) our understanding of these processes at high temperatures, (2) the data base of material and chemical. properties, and (3) modeling uncertainties accompanying the codes. This program is aimed at identifying and closing these gaps so that uncertainties associated with the prediction of containment-loading and source-term predictions can be reduced.
Core-Concrete tests in the PETA facility in _the FRG and at Sandia (involving sustained urania/ concrete heating tests) in FY 85 and FY 86 are progiding an expanded data base for improving and valfdating models.
The results of calculations or core contrete interactions are also highly dependent on the initial conditions of the core debris, that is, the conditions at vessel failure. That aspect of the problem is being addressed under the aegis of core melt progression.
Summary of Program Strategy The strategy on which this program is based is to identify and prioritize the sources of uncertainty and then carry out research which will pemit reduction of these uncertainties. Although the containment-loadirg and radioactive aerosol release parameters have an interrelated impact on the ultimate risk, f
the two aspects of core concrete interactions are considered separately. The 4
following research program is intended to address and resolve these issues:
T 28 j_
i
=
Thermal-Hydraulic Uncertainties 1.
The experimental programs have demonstrated that concrete composition affects its behavior under thermal ablation. A research program is planned which will result in construction of a comprehensive catalog of the important properties of different kinds of concrete as a function of composition: strength, heat of ablation, ablation temperature, viscosity, surface tension, etc.
2.
The CORCON and VANESA code predictions depend critically upon the geometric configuration of the debris pool. Little is known about the fluid-dynamic properties of core debris as it exits the reactor vessel i
and spreads out over the cavity floor. A program is planned to study and develop models for this debris spreading process.
3.
Considerable uncertainty remains in the prediction of radial concrete erosion relative to downward ercsion. This information impacts the risk associated with possible basemat penetration and the potential degradation of supporting concrete structures such as the reactor 2
pedestal. These predictions are controlled primarily by the heat-transfer models in CORCON. The research programs in place and-planned at SNL and BNL are aimed at reducing this uncertainty by developing refined models and conducting the SURC experiments with both metallic and oxidic melts.
4.
Radiological source-tem predictions depend critically on a knowledge of debris temperature. The temperature, in turn, depends upon the interfacial heat transfer between pool layers and/or the onset of interlayer mass entrainment, if the pool is stratified. The research at BNL is addressing this problem.
i 5.
The fission-product release rates predicted by VANESA depend upon melt temperature and gas sparging rates. Both of these variables are strongly influenced by the amount of unoxidized (metallic) zirconium present in the debris. Better understanding of, and improved models for, this
" coking" process will be derived from the SURC program at SNL.
6.
If the core-debris temperature can be lowered soon after the material exits the reactor vessel, the radiological source tem will be drastically reduced. The opinions of technical experts differ on whether or not the presence of overlying water will accomplish this temperature reduction. The SNL SWISS experiments indicated that the overlying water goes into film boiling and that the melt temperature remains high.
The IDCOR assumption is that the water will cause a surface crust to form which will then fragment and ultimately render the debris coolable. To resolve this uncertainty, a program is planned to study this phenomena and to develop models which can be extrapolated to reactor-scale conditions with an acceptable level of confidence.
29
7.
To date, most of the research has addressed the early phase-(first-few hours)propertiesofcore-concreteinteractions. To extend code-predictive: capability into the longer-term phase of reactor accident analysis, further study is needed on crust formation and freezing phenomena:and the nature of hot solid debris attack on concrete. The program in place at SNL is aimed at this problem and will be extended.
Aerosol and Radionuclide_ Release Aemsols are generated at the pool surface by vapor condensation and-bubble-ftim fragmentation. The VANESA code models these processes; model improvements and an augmented experimental data base for validation are needed.
i 1.
A series of special-effects experiments will be established at BCL to measure chemical-activity coefficients of prototypical chemical species to reduce the uncertainty in VANESA fission-product release predictions.
2.
The WITCH and GHOST experiments under way at SNL are designed to independently study the vaporization and mechanical-fomation processes.
Both metallic and oxidic melts are used in these tests and the material is doped with aporopriate fission-product simulants so that the data.can be used for VANESA. validation and guidance in model improvements.
3.
It has been established experimentally that when surface crusts fom, they are usually porous so that the sparging gases continue to leave the pool surface. The degree and manner of aerosol decontamination that may be accomplished by this process is being studied at BNL.
Research Accomplishments To Date 4
l 0
Experiments are in progress at SNL in which simulant fission products are included in the core-concrete attack tests.
These data are providing a basis for testing C0RCON and VANESA modeling assumptions.
i O
A series of experiments has been performed at Sandia National Laboratories (SNL) to study the behavior of molten-core materials released from the reactor vessel under high or low pressure, with or j
without the presence of water. The earlier tests used molten steel to simulate molten core materials. The more recent tests used inductively heated molten corium.
O The more recent tests show drastically different heat transfer characteristics because of early crust formation. Quantities of the test melts ranged from 10-80 kg for the high pressure tests, and up to 200 kg for the low pressure ones as compared with several tons expected for reactor accidents.
O Some wet tests were conducted with a water filled cavity and some with the addition of water after the melt was released from the reactor vessel.
30 l
. Expected FY 87 Milestones at Current Budget Level O
Program planned to study spreading behavior of high-temperature core
' debris for resolution of'BWR Mark I drywell liner failure issue.
'0_
Program to study surface crust formation and breakup caused by overlying water; determine influence on debris bed coolability; initiate program with feasibility study.
O Incorporate new information into CORCON-M002.
O Conduct tests to determine the effects of boron carbide during molten core interactions with limestone and siliceous concretes..
O Compile a summary review of data on core debris / concrete interactions-available for model validation.
O Complete the WITCH tests of aerosol generation by mechanical processes during core / concrete interaction.
Issue report and reconsnendation of changes to VANESA.
i 0
Complete the GHOST tests on aerosol generation by vapor-condensation processes during the core /conrete interaction.
Issue report and recommend changes, if any, to VANESA.
i FY 87 Resources Allocated - $1665K FY 87 Unfunded Needs - $850K 0
Study boiling and non-boiling heat transfer from submerged porous
~
surfaces. Investigate the effects of gas bubbling on heat transfer.
O Complete film boiling experiments with freon R-il and liquid metal melts.
O Program te study the material properties and high-temperature erosion characteristics of different types of concrete as a function of material composition.
l O
Establish special-effects program to measure fission-product chemical properties and activity coefficients needed for VANESA model improvement.
O Complete experiments on boiling of water overlying imiscible liquid pools.
O Conduct tests to determine the effects of boron carbide during molten core interactions with limestone and siliceous concretes (SURC).
O Compile a sumary review of data on core debris / concrete interactions available for model validation.
l 31
.. ~.
~
Expected FY 88 and Beyond Milestones O
Complete integration of the CORCON MOD 2 and VANESA codes into a single consolidated code. The CORCON/VANESA code package will perform the following analyses: -(l) rate and direction of attack on structural concrete. (2) release rates and composition of vapor species from the core debris, i.e., water vapor, noncondensable.and flamable gases. (3) release of radiant and thermal energy from the debris pool to the containment,(4)characterizetheradioactiveandnonradioactiveaerosol release to the containment atmosphere (5) characterize the scrubbing -
(decontamination factor) effects of an overlying water layer, and -(6) continue code predictions associated with the foregoing phenomena into the long-term accident phases due to continued concrete erosion by solid, or partially-frozen core debris.
0
. Conduct experimental program on aerosol decontamination by porous / solid crusts.
O As the experimental and theoretical data bases improve, refine the phenomenological models so that uncertainty can be reduced.
O Increased emphasis will be placed on the analysis of experimental data for the purpose of code validation (this effort is ongoing).
O The SURC series of experiments in which the interaction between prototypical core melt materials and concrete are tested under conditions of sustained induction heating will be completed in FY88.
4 O
The experimental program on heat-transfer phenomena at Brookhaven National Lab (BNL) together with model development, will continue in close coordination with the SNL CORCON/VANESA development and validation work.
O In connection with the SURC experiments, an augmented effort to understand and model the zirconium-oxidation issue will continue. This is particularly important with respect to BWR safety because of the high Zr inventory in the core.
O A research program will be completed on the study of debris spreading phenomena and the influence on the BWR Mark I dry-well liner failure issue and on release fraction for refractory fission-products; issue draft report on the research and its implications for severe accident analysis.
O A research program will be designed and. completed for the purpose of resolving the question of whether or not surface-crust formation and break-up rey or may not provide mechanism by which debris beds may be rendered coolable.
32 l
0 Continue to prepare and release periodic code up-date sets-for CORCON/VANESA and maintain ongoing user support as needed.
O Continue to maintain close liaison and cooperation with Japan, the Federal Republic of Germany and the United Kingdom where cooperative research and information-exchange agreements have been established.
. 0 Conduct experiments on boiling of water overlying immersible liquid pools.
33
b HYDROGEN IGNITION AND BURNING IN CONTAINMENT Description Hydrogen Program The production of hydrogen in severe accidents could lead to one or more combustion events inside of containment. To evaluate the threat to the containment given a specific quantity of hydrogen, the hydrogen distribution, the possibility and strength of an ignition source, the likely mode (s) of combustion -as well as the containment volume and design pressure must be considered. The objective of the NRC Hydrogen Behavior Program is directed at investigating and quantifying combustion phenomena including deflagration, accelerated flames.and transition from deflagration to detonation (DDT),
~ detonation and diffusion flames. The threat to containment and safety related equipment from hydrogen burns, is strongly dependent on the combustion mode.
Ordinary deflagration is a form of combustion in which the flame moves at subsonic speed relative to the' unburned gas. Detonation is a form of combustion is which the flames move at supersonic speed relative to the unburned gas. Accelerated flames is a form of combustion intermediate between deflagration and detonation and for highly accelerated flames, a transition from deflagration to detonation can occur.
In addition to the research sponsored by the NRC, the nuclear power
~
. industry also had research programs to investigate hydrogen behavior.
There are substantial differences between the IDCOR and NRC treatment of l
ignition, burning, and flame propagation in mixtures of air, hydrogen, and steam.
IDCOR analyses base ignition on a calculated flame temperature criterion which is a function of the composition of the atmosphere within the j
compartment. NRC analyses, in addition to considering hydrogen and oxygen concentrations within a compartment, give explicit consideration to steam l
inerting and the availability of ignition sources. The IDCOR models appear to predict continuous burning in essentially all cases, whereas the NRC treatment l
tends to predict a number of discrete burns. The NRC's approach tends to allow the buildup of higher hydrogen concentrations and hence can lead to the l
prediction of higher containment pressures. The differences between the IDCOR i
and NRC treatment of hydrogen ignition and burning are particularly pronounced I
in multi-compartment systems, such as the ice condenser containment, and in the absence of deliberate ignition.
There is experimental evidence that hydrogen combustion can alter the chemical form of airborne fission products, e.g., release of molecular todine from cesium iodide aerosols due to hydrogen flames.
It is not clear whether such changes in chemical form increase or decrease the consequences of severe accidents. Neither IDCOR nor the NRC analytical models at present consider such changes in fission product chemistry.
I Resolution of this issue will come, in part, from continued comparisons be-L tween experiments and analyses.
It must be recognized, however, that experimental data may not be available to address issues related to burning in ccmplex geometries.
34
Sumary of Program Strategy-The_ uncertainties associated with hydrogen loading lie in the state of understanding of various phenomena associated with hydrogen generation and combustion. These phenomena are as follows:
l 1.
hydrogen transport and mixing 2.
ordinary deflagration 3.
accelerated flames and transition from deflagration to detonation (DDT) 4.
detonations 5.
diffusion flames Different tests with varie'd geometries have been perfomed to investigate deflagration as a form of combustion and the understanding in this area is good but the' uncertainty associated with extrapolating these tests to reactor size and geometrics still needs to be reduced. Assessment of data from tests at the Nevada Test Site to improve existing flame speed, flame propagation, containment spray. characteristics, and burn completion correlation coupled with improved turbulence modeling could make the scaling of deflagration data a
more reliable.
1 Uncertainty associated with hydrogen transport and mixing is significant in that it will influence the initial and boundary conditions for combustion.
The uncertainty in transport and mixing is closely coupled to the remaining uncertainties in that the degree of mixing and rate of transport can detemine which mode of combustion will dominate.
Existing codes give good results for i
hydrogen transport and mixing for a unifomly mixed system. However, there is significant uncertainty associated with prediction for nonuniform mixtures.
i These uncertainties can be reduced by comparison (model refinement) to a limited number of lar Comparison to ongoing HC06 (Hydrogen Control Owners Group)ge scale tests.1/4 scale tests and to proposed HDR tests wo i
useful but are currently unfunded. However, HCOG test data will be used to reduce uncertainty in existing diffusion flame modeling. This experimental data will be used to assess the model as predictive tools for flame length and I
temperature / velocity fields in the vicinity of safety related equipment.
Pesults from HECTR and HMS-BURN calculations show that for certain accident scenarios hydrogen could be present in locally high concentrations and that flame acceleration and transition from deflagration to detonation (DDT) and i,
i detonation could occur. Because of the likelihood that these locally high hydrogen concentrations could occur sometimes during the course of an L
accident, the potential for flame acceleration DDT, and detonation, as well as the consequences of these combustion modes on containment integrity were investigated. Experimentshavebeenperformedatvariousscalestoassess(1) i the potential for flame acceleration and transition from deflagration to 35 i
I w.-c.,-.--,--~,,.-,em
_m.,~,,wm.e._
_p.,,,,.
%%w--.e.we-,,,w
,,r-,--ym
detonation (DDT) and (2) the potential to propagate a detonation under accident conditions (mixture composition, pressure, temperature) and geometries. Correlations have been developed and scaling trends identified to-quantify the uncertainty in these areas and to further analyze existing data to further reduce the uncertainty associated with flame acceleration, transition from deflagration to detonation, and detonations to ensure that interpolation and extrapolation to reactor scale / conditions will be reliable.
Review ff Hydrogen Combustion Research by the National Academy of Sciences The committee on Hydrogen Combustion was formed in 1985 and was composed of seven experts in the area of nuclear reactor safety, physics, gas dynamics, combustion, chemical kinetics and hazard analysis. These experts were asked to perform an independent assessment of the technical issues related to the behavior and control of hydrogen generated in severe accidents to determine the degree to which current knowledge may support regulatory decision making.
In recent years a great deal of safety related hydrogen combustion research has been carried out at different scales and various geometries and computer codes have been developed based on these experimental results. Because of the importance of this work, the comittee was asked to assess 1) the ability to scale up, and to interpolate and extrapolate data from experimental scale to i
actual plant conditions and 2) to assess whether all important areas of research have been adequately addressed.
The comittee concluded that the current research program on hydrogen combustion had properly covered most technical aspects in this area and that an extensive large-scale experimental program is not needed. However, the comittee did recomend that some further work is required. Among its recommendations are the following:
further reduce uncertainties in hydrogen release, transport, mixing and combustion models by comparison with experimental results and incorporation of some zone and field modeling efforts.
(unfunded i
FY87/88need) l develop methods for improving the reliability of igniter systems during station blackout scenarios and continue to develop catalytic igniter systems.
(unfunded/ unplanned) reduce uncertainties in diffusion flame and subsonic premixed flame j
phenomena through planned experiments.
(unfunded/ unplanned) 1 reduce uncertainties associated with the likelihood of failure of subatomspheric containment under detonation loads through more detailed analysis. (unfunded/ unplanned) reduce uncertainties associated with the possibility of detonation for large dry containments having fan coolers through more detailed analysis.
(unfunded/ unplanned) 36 i
i w
- - -,, -r--e-m- -,,
v-m,,wwem,-.,,-,,.m,--,-,---_,-nv,,--wws,-w,v.wn,,,-,_,,v,,,,
.n.,,,,-m,,.-,m,,a,_,m,c--m..
,,,-.--,,.:n----,-
---v
t Finally this committee of experts reconsnended that a new review panel be formed to provide a critical assessment of modeling techniques currently being developed or used.
Research Accomplishments To Date O
A hydrogen compendium (NUREG/CR-1561) was published to provide a resource describing hydrogen behavior during accidents in LWRs.
O A LWR hydrogen manual was published (NUREG/CR-2726) which provides general guidance in developing plant-specific procedures for handling hydrogen during normal and off-nonnal conditions.
i 0
Analytical methods have been developed for analyzing hydrogen and/or steam transport in containment, and these methods have been successfully applied to experiments conducted at EPRI, Battelle Frankfurt, SNL, and the large-scale facility at the Nevada Test Site.
l 0
The HECTR code was developed (NUREG/CR-3913) for the analysis of hydrogen i
combustion events in various reactor containment types. The models in the code serve as the basis for the hydrogen burn models in the containment systems code CONTAIN and the simplified integrated codes, STCP and MELCOR.
0 The Hydrogen Control Owners Group (HC0G) has completed experiments at 1/20 scale and are currently performing tests in a 1/4-scale facility representative of a Mark III containment.
It is anticipated that these i
results will enhance our understanding of combustion behavior in a Mark III containment and may provide a sufficient experimental basis for diffusion flame modeling in reactor accidents.
i 0
Recent experiments in the FLAME facility at Sandia, experiments at Battelle-Frankfurt and at the Fraunhofer Gesellschaft indicate a much l
higher potential for the transition to detonation than previous I
experiments had indicated. Additionally they indicate that scaling is l
possible based on theoretical considerations.
l 0
Experimental work sponsored by the NRC has resulted in establishing lower values for the propagation limits for a hydrogen detonation in air is as low as 13 percent hydrogen under ambient conditions.
O In the area of hydrogen burn mitigation studies, several options have been reviewed (NUREG/CR-1762, 2767, and 2865) and, aside from a delib-erate ignition system and a passive catalytic igniter, most have been i
i eliminated because of engineering considerations, cost, or other negative j
effects.
I l
Expected FY 87 Milestones at Current Budget Level 37
0 Incorporate, test, and validate a spray model in HMS-BURN with Nevada Test Site (NTS) and 1/4 scale data.
O Issue topical report on NTS experiments.
0
-Initiate documentation of the HMS-BURN code.
O Complete HECTR code assessment and sensitivity studies using data from VGES, FITS, NTS, and the HCOG 1/4 scale tests.
Issue report.
O Issue a final report on all flame acceleration and deflagration to detonation studies performed at the FLAME facility.
O Issue a final report on the detonation hot tube experiments.
O Provide support for incorporation of the improved hydrogen combustion models into the MELCOR and CONTAIN codes.
O Provide experimental design support for large scale hydrogen mixing I
experiments in the HDR facility.
J FY 87 Resources Allocated - $579K FY 87 Unfunded Needs - $530K 0
Continue support of CONCHAS-SPRAY code calculations in support of the flame acceleration (F/A) and transition from deflagration to detonation (DDT) experiments for the FLAME, mini-FLAME and McGill experiments to provide the analytical capability ~to predict F/A, DDT, or detonation for reactor scale / accident conditions.
O Continue further developmental work on HECTR to provide increased capability to scale up experimental data and to improve modeling of accident conditions such as sprays and flame propagation between com-partments. (FY87 tasks will be limited to maintenance and code application.)
O Provide an assessment of HCOG 1/4 scale experiments for improved diffusion flame models for HMS-BURN and HECTR and to investigate flame lift-off and attachment to improve confidence in the results and to make more reliable assessment of equipment survival from hydrogen burns.
O Continue further development of the HMS-BURN code to extend capability to all plant types by developing an ice condenser model.
(codetasklimited to documentation and application of current version).
O Provide support for incorporation of recent experimental data into detonation hot tube report and for improved detonation codes to provide the capability to calculate lean detonation limits as a function of steam content for those mixtures that might be expected to occur during some accidents.
38
0 Complete documentation of the-HMS-BURN code to improve confidence in the results for use in benchmarking HECTR.
0
' Continue assessment of likelihood and consequences of local detonations for subatmospheric and large dry PWR containments to determine if an adequate safety margin exist.
Expected FY 88 and Beyond Milestones 0
' Improve transport and mixing models in hydrogen combustion codes.
O Improve diffusion flame models.
O Release HECTR version 2.0.
FY 88 Resources Allocated - $600K FY 88 Unfunded Needs - $20DK 0
Continue support of large scale hydrogen mixing studies at HDR to further reduce uncertainties in the transport and mixing models of both HECTR and HMS-BURN.
0 Extend. studies at SNL on flame acceleration, transition to detonation, and detonations to allow for gaps in data that may become apparent when the final reports in these areas are available in late FY 87 and to respond to NAS report.
All work in this area should be completed by the end of FY 88, provided the unfunded needs are met.
39 l
i CHEMICAL FORM OF IODINE i
Description After the accident at Three Mile Island, it was cosmonly assumed that almost all of the fission product iodine was transported as aerosol particles of Csl in the reactor coolant system and in the containment. Analyses reported in NUREG-0772 showed that this was the preferred fem in the reactor coolant system based on thermodynamic considerations. Current source term evaluations discussed in NUREG-0956 are based on that assumption. There has, however, been a chronic concern as to whether or not this assumption is valid.
Furthermore, recent experimental results on this subject do not give a clear picture. Results from the PBF program showed the presence of one Csl particle on a deposition rod in the 1-3 test. The EPRI in-pile experiments at TREAT-and the NRC out-of-pile experiments at the ORNL Hot Cell Facility found that most of the iodine detected was accompanied by cesium. Data from Battelle's high pressure release experiments indicated that cesium and iodine were released from the fuel separately. Separate-effect experiments conducted at Sandia indicated that Cs! may be unstable in the reactor coolant system.
Laboratory experiments performed to study the chemistry of iodine in the i
containment showed that volatile iodides can be formed in the containment even if iodine enters the containment as Cs!. Finally, filters installed in the LOFT containment for the FP-2 test detected the presence of volatile iodides, although the sample was a very small fraction of that released from the bundle J
or transported in-the LOFT primary system and may not be conclusive.
At first glance, these experimental results seem to be in conflict, but they are merely pieces of information that need to be tied together.
It is obvious that the chemical form of iodine depends on the environmental conditions which are functions of the tirie into the accident and the location in the reactor i
coolant system or containment. Additional experiments are planned and data analysis of completed experiments will continue to fill in the data gaps.
l Currently the uncertainty in the iodine chemical form is high and the fraction of iodine considered to be volatile ranges from zero to 100 percent of the iodine released to the reactor coolant system; however, the more likely range is believed to be narrower.
It is expected that this uncertainty range will be narrowed as additional results become available.
l Susuary of Program Strategy Details of the experimental plan are as follows. The ORNL release experiments and the Battelle high pressure release experiments will be extended to temperatures beyond 2400K to determine the chemical form of iodine at high temperatu.es. As for the Sandia program, it is believed that, at the low radiation level used, the instability of Cs! observed in previous radiation experiments was probably caused by effects other than radiation.
Experiments being conducted in the EPRI-sponsored program at the Atomic Energy of Canada Limited facility at Whiteshell will provide the confirmatory data.
For FY 87, Sandia will focus their effort to detemine the cause of Cs!
decomposition in the absence of radiation. The need for studies on the effect 40
n.-
.1
.syu +
u w
-a u
.n w.
t of high level radiation on fission product stability will be assessed following the White Shell experiments and sensitivity studies noted below.
4 In the containment area, ORNL will continue to improve their models for.
transporting different iodine chemical forms in the TRENDS code, to use the 1
improved code to reevaluate. existing sensitivity calculations, and to conduct
. new sensitivity calculations for other plants and accident sequences, as recommended by the EPRI-NRC Steering Consiittee that reviewed this work. These f
sensitivity studies will place the-importance of the volatile iodine in proper perspective. Further experiments will also be performed to support code improvement activities and to determine the fate of the iodine-132 formed by the decay of tellurium-132. Research on the latter also applies to the reactor coolant system when the deposited tellurium decays into iodine. The i
in-pile experiments at NRU conducted to study hydrogen generation will also provide information on the chemical fom of iodine.
Research Accomplishments To Date 0
Deposition profiles in PBF-SFD Tests indicated the co-location of Cs and I.
Post-test-analysis of deposition samples showed the presence of at
'least one CsI particle.
O Analysis of the thermogradient tube samples from the Oak Ridge fission l
product release experiments also showed the co-deposition of Cs and I.
3 O
High pressure fission product release experiments at Battelle Columbus indicated separate releases of Cs and I from the fuel.
1 O
Separite-effect experiments conducted at Sandia (and later at Battelle Colum>us) iqicated that Cs! may be unstable in the reactor coolant system.
\\
i s
O A study of iodine behavior at ORNL in containment and in suppression pools showed tisat volatile iodides can be generated, i
0 Filter samples in the LOFT containment during the FP-2 test detected the i
presence of volatile iodides.
0 A parametric study to assess the impact of volatile lodide formation on source tem was performed for two accident sequences for Surry and one accident sequence for Peach Bottom. The results were incorporated into the final version of NUREG-0956. Preliminary results indicate a lower importance for iodine chemical form than previously expected.
i
- FY 87 Milestones at Current Budget Level l
0 Extend current experimental matrix to determine fission product chemical forms at more severe core damage conditions.
That is, extend bulk core temperatures in future experiments to beyond 2400K.
t 4
l 41 r
Conduct three ORNL high temperature fission product release tests (Tests VI-3.through VI-5).
Issue final reports for Tests VI-1
.through VI-3.
. Issue final fission product behavior reports for PBF Tests SFD 1-3 and SFD 1-4.
~ 0 Conduct experiments to determine the cause of Cs! dissociation in the absence of radiation for the SNL experiments.
- 0 Conduct further experiments to determine the functional dependence of Cs!
decomposition on factors identified in the above experiments.
O Continue experimental study at ORNL to provide data to improve the TRENDS code.
O Redo existing calcula,tions with the improved TRENDS code and perform analyses for other accident sequences and plants to assess the impact of volatile iodide formation on source terms.
O Perform analysis to study boron compound effect on Cs! stability.
0 Evaluate data from FLHT-4 test at NRU to determine the chemical form of iodine at in-pile experimental conditions. The results should be able to assess the EPRI finding in the STEP test in TREAT.
FY 87 R_esoJrces Allocated - $1693K FY 87 Unfunded Needs - $400K FY 88 and Beycnd Milestones O
Continue and complete the high temperature investigation of fission product chemical forms.
Conduct the remaining CRNL high temperature fission produri release tests (Tests.VI-6throughVI-10),
f Conduct melt-progression source-term experiments at ACRR.
0 If necessary, conduct experiments to determine the effect of high-level
(
radiation on Cs! stability.
O Develop a model to predict the chemical forrs of iodine in the reactor ccolant system under varicus severe accident conditions.
O Conduct multi-effects experiments to validate the TRENDS code.
l l
0 Determine the fate of I-132 generated from the decay of Te-132.
42 l
l v -~ ~---- - - -, -, - -, -,, - -
,-.-.gwnn,.v,
,_,,,,s.
e-w
_-_,,c,,,,--a-n,-w _,_ _-w
_,m,,,, _ nw w n~,
em-
k REVAPORIZATION OF PREVIOUSLY DEPOSITED FISSION PRODUCTS Description Current models for in-vessel transport and deposition of fission products predict that for some sequences a large fraction of fission products released i
from the core can be deposited on structures in the reactor coolant system.
Associated with these fission products is a significant portion of the decay heat with the potential for heating structures and revaporizing the deposited fission products.
In some analyses, when revaporization is taken into account it results in enhancement of the environmental source term for some reactor types and accident sequences. A number of major uncertainties remain, however, including: a very sparse data base related to the mechanisms of reaction between fission product species and surfaces which would affect the potential for revaporization; the modeling of natural-convection-driven flow patterns in the reactor coolant system prior to, and subsequent to, lower head i
failure; and the behavior of reactor coolant system insulation material in the i
accident environment.
i The chemical forms of the deposited radionuclides detemine the vapor pressures of the radionuclides and the potential for and timing of fission product revaporization. At present, the chemical forms of most fission i
products adopted for analytical calculations are based on assumptions.
l Iodine, cesium, and tellurium are assumed to be in the forms of Cs!, Cs0H, and j
elemental tellurium, respectively, regardless of the changes in the environmental conditions with time and location. Recent results from various 4
experimental programs indicated that iodine, cesium, and tellurium can exist in other low-or high-volatility foms. The vapor pressures of other foms are, of course, different from those of Cs!, Cs0H, or Te. The potential for and the timing of revaporization of these foms will also be different. High vapor pressure species revaporize early. Early revaporization would cause little additional consequence since there are natural processes such as 4
settling and washout in containment to reduce this component of the airborne aerosol. Additionally, if the containment fails early, the source tem is i
large and any augmentation due to fission product revaporization would not be l
significant. Low vapor pressure species may not revaporize or they may revaporize late. The consequences of delayed revaporization may be larger i
since the source terms for delayed containment failure accidents are currently l
1 assessed to be low, and the containment mitigation processes at late times are i
less potent. Therefore, accurate predictions of the chemical forms of deposited radionuclides are necessary and models are being developed and incorporated into TRAP-MELT.
IDCOR claims that, in large dry PWRs, heat loss through deteriorated insulation on reactor coolant system components will prevent heat up of l
structural surfaces and consequently preclude fission product revaporization.
An evaluation of the IDCOR claim is needed.
Finally, studies based mainly on j
an empirical approach to natural circulation indicates a need for a mechanistic analysis to confim the flow and mass transport conditions in the vessel after vessel breach.
i j
43 l
The current estimate from NUREG-1150 of the uncertainty range for this issue is zero to 0.9 fraction of the deposited radionuclides revaporized Sunnary of Program Strategy The uncertainties in post-vessel-failure thermal-hydraulics and fission product chemistry contribute to the overall uncertainty in the issue of fission product revaporization. To resolve this issue, research on both areas is conducted in parallel. Specific details of both areas of research and the L
integrated analysis are given below. Plans for code validation are also i
discussed.
l A. THERMAL-HYDRAULICS (FY 1987)
EG&G will search the literature for heat loss data at high temperatures on the typical insulation materials used in nuclear power plants and perform sensitivity calculations to assess the effect of heat loss through insulation on structural surfaces. The task is scheduled to be completed by December 1986.
Scoping calculations will also be performed to determine magnitudes of flow rates, flow directions, and surface temperatures.
It is scheduled to be j
completed by June 1987.
]
B. REVAPORIZATION CHEMISTRY (FY 1987) j Sandia will develop gas phase and surface chemistry models, based on t
thermodynamic data, to predict the chemical forms of deposited iodine, cesium, i
and tellurium. All models will be delivered to EG8G by December 1986.
Subsequently, they will be coded and added to the EG4G version of the TRAP-MELT code. The improved TRAP-MELT will subsequently be used in a separate-effect analysis to assess the effect of chemistry on the timing of fission product revaporization. Conditions for the calculations will include the flow data obtained in the separate-effect thennel-hydraulic analysis.
l TRAP-MELT improvement activities should be completed by June 1987 and the revaporization calculations should be completed by September 1987.
4 i
C.INTEGRATEDANALYSIS(FY1988)
F j
The separate analyses conducted for the thennal-hydraulic and chemistry study
(
do not consider feedback between thermal-hydraulic results and chemistry results. One example of feedback that needs to be considered is surface temperature. As fission products vaporize from the surface into the gas phase, the heat source associated with the decay of deposited fission products becomes smaller, and the surface heat-up decreases. Tiis information needs to be fed back into the thermal-hydraulic analysis so that a new surface temperature can be estimated for subsequent revaporization calculations. The result will be a more accurate estimate of the timing of revaporization.
Additionally, an estimate is needed on the effect of revaporization on severe accident source terms for delayed containment failure accidents. The CONTAIN 44
code will be used for this calculation. The calculation will be conducted for a PWR and a BWR plant.
For each plant selected, a transient and a small-break LOCA sequence will be used as conditions for the study. Task completion is expected by September 1988.
D. EXPERIMENTAL VALIDATION There are large uncertainties associated with the themodynamic data base used
-for-the development of the chemistry models. The data are based mainly on extrapolations or estimates, some from published data, and need to be experimentally verified. Work is planned at Battelle Columbus where gas phase interactions and surface interactions including nonideal interactions in liquid mixtures will be considered. The experiments will provide data on the chemical forms of radionuclides and their ciaracteristic vapor pressures.
Activity coefficients in the liquid mixtures can then be calculated using the experimentally obtained vapor pressure data.
In the Battelle program, multi-effect experiments are also being considered to verify tte integrated analyses of fission product revaporization. Some data are expected on this subject from experimental programs sponsored by other organizations. For example, the EPRI revaporization program at Argonne will provide data on the revaporization of simple mixtures. These data, although helpful, will not be able to verify the complex chemistry envisioned in the reactor coolant system during severe accidents. Nevertheless, the Argonne experiments (and others being planned) will be integrated into the Battelle test plan. The separate effects experiments are scheduled to be completed by September 1988. The multi-effect experiments are expected to be finished by September 1990. One difficulty associated with the multi-effect experiments is the simulation of realistic specimens for the revaporization study whereby the expected complex interactions are simulated. One option is to use the piping from the PBF Severe Fuel Damage Tests. Another option is to simulate surface conditions according to analytical predictions in NUREG/CR-4624.
If the integrated analysis shows that fission product revaporization is important, modifications will be made to the mechanistic MELPROG/ TRAC code to consider this phenomenon.
Subsequently, calculations will be conducted with the modified MELPROG/ TRAC code to confim the previous results predicted using the RELAP/1 RAP-MELT code.
Research Accomplishments To Date 0
Coupling of the TRAP-MELT and MERGE codes has been achieved to permit analysis of the effects of revaporization during the period prior to vessel melt through and, in a parametric manner, after melt through. A more rigorous treatment will be provided by the coupled RELAPS/SCDAP/ TRAP-MELT and MELPROG/ TRAC codes.
O A revaporization chemistry model was partly incorporated into the RELAPS/SCDAP/ TRAP-MELT Code package.
0 A parametric study with the partially completed RELAPS code package was conducted for NUREG-1150. The effect of the Cs0H-Csl interaction on the timing of revaporization was determined.
45
C A model for the heating of structural surfaces due to decay of deposited fission products was incorporated into the RELAPS code package.
FY 87 Milestones at Current Budget level O
Complete development of a revaporization chemistry model and incorporate into the RELAPS code package.
C Assess IDCOR's heat transfer models for insulated and uninsulated surfaces in the reactor coolant system.
O Conduct separate-effect analyses to determine the effects of: 1) revaporizationchemistry,and2) postvessel failure thermo-hydraulics en fission product revaporization.
FY 87 Resources Allocated - $190K FY 87 Unfunded Needs - $500K FY 88 and Beyond Milestones O
Conduct separate-effect experiments to confirm thermodynamic data used in the revaporization chonistry model.
0 Validate the chemistry model with results from existing multi-effect experiments.
If not adequate, conduct additional experiments.
O Conduct integral analyses for selected sequences to estimate the impact of revaporization on source terms. The analyses will include containment calculations to assess the effect of natural mitigation processes on the revaporized fission products, t
t 46 l
a s
GENERAL SUPPORT FOR ALL AREAS In addition to.the above, general support for all areas of uncertainty is required which includes code development and validation contracts, code' application contracts, uncertainty analyses, peer review groups, and general phenomenological analyses of plant-specific issues, i
Accomplishments to date for these support programs include-0 Development of the Source Tem Code Package ($TCP).
l 0
Development of the best-estimate mechanistic codes such as MELPROG, SCDAP. MELPROG/ TRAC linking. SCDAP/RELAPS linking, CONTAIN, CORCON, VANESA, HECTR, HMS BURN TRAP / MELT, and VICTORIA.
O Peer review and publication of the BMI-2104 source tem methodology.
0 Peer review and publication of the reassessment of source term l
technology. NUREG-0956.
l 0
Severeaccidentsequenceanalyses(SASA)forNRRandriskstudies.
0 Out-of-pile fission product release experiments conducted at the BCL.
j.
CORSOR-M gives good agreement with these data for temperatures up to 2400K.
l 0
Separate effects experiments at Battelle Columbus have shown that two Te retaining mechanisms are operative for the range of severe accident conditions of interest. At low temperatures, the formation of zirconium telluride was found to be important whereas, at high temperatures, the i
j formation of tin telluride was favored. Tin is a component in the i
Zircaloy cladding. The PBF tests also confirmed the Te retention of Te.
The effect is now mechanistically modeled in VICTORIA, the new mechanistic fission product release module in MELPROG.
Expected FY 87 Milestones for these Support Programs CODE DEVELOPMENT AND VALIDATION:
0 MELPR06/ TRAC /M001 will be completed and documented.
O A status report on SCDAP/RELAPS validation will be issued.
l O
A report on the EWR version of MELPROG will be published.
j 0
Benchmarking of the STCP will be continued.
O Benchmarking of MELCOR will be continued.
I i
f~
47
0 A MELPROG assessment report using ACRR and PBF data will be issued.
O Improved and additional models will be added to the CONTAIN code including melt ejection under high pressure, aerosol behavior with the Kelvin effect and the effect of hygroscopic aerosols, and physical and numerical treatment of heat deposition en containment surfaces.
Alternative models will be considered to the present lumped parameter approach in order to assess effects such as natural circulation.
stratification in open volumes, and buoyant plume mixing.
PHENOMEN0 LOGICAL ANALYSES OF PLANT-SPECIFIC ISSUES 0
Complete the station blackout reevaluation using CONTAIN.
O Conduct an accident sequence analysis for MARK II containments.
0 Evaluate structural support of MARK III pedestals under attack by core meit/ concrete mixtures.
4 0
Benchmark Mark I MELCOR analysis.
UNCERTAINTY ANALYSES PROGRAMS 0
Complete detailed uncertainty source term analyses for Peach Bottom TC sequence with the STCP.
O Issue a sumary report on the important phenomenological uncertainties.
O Begin sensitivity and importance analyses.
0 Use the CONTAIN code to detennine the uncertainty associated with the 4
high pressure melt ejection phenomenon for specific plants and sequences.
EXPERIMENTAL DATA FOR CODE VALIDATIONS:
O Continue to support the LACE program with pre-and post-test analyses.
O Conduct LACE tests LA-5 and LA-6.
O Issue topical reports on the following deposition phenomena: Cs0H reaction with 304 stainless steel, Te interaction with structural materials in steam, competitive Te reactions, Cd behavior in steam, and the effect of borated coolant on fission product chemistry.
O Complete ORNL experiments on aerosol resuspension and issue sumary reports.
48 i
s i
0 Complete development of a mechanistic aerosol resuspension model for incorporation in the TRAP / MELT module of the STCP.
0 Issue report on Sn/Te chemical interactions with emphasis on tht."
roteidi n of Te in the melt by such a mechanism.
FY 87 Resources Allocated - $4358K s
FY 87 Unfunded Needs - $300K Expected FY 88 and Beyond Milestones CODE DEVELOPMENT AND VALIDATION O'
Release MELPROG/ TRAC / MODI and SCDAP/RELAP5/M002.
O Continue code validation for MELPROG/ TRAC and SCDAP/RELAPS.
O Benchmark of BWR MELPROG.
0 Continue benchmarking of MELCOR.
O Complete verification and benchmarking of the STCP.-
4 0
Develop methods to analyze aerosol behavior that are compatible'with multidimensional calculational methods used to assess material l
circulation, stratification in open volumes, and buoyant plume mixing.
O Validate the HECTR and HMS-BURN hydrogen models that have been i
incorporated into CONTAIN.
i O
Continue the assessment of CORCON/VANESA as new data become available.
PHENOMEN0 LOGICAL ANALYSES OF PLANT-SPECIFIC ISSUES 0
Conduct dominant risk accident analyses for Mark II Containment.
0 Based on the completed Mark III pedestal analysis, assess potential i
modifications to severe accident scenarios and benchmark source terms.
l 0
Conduct an evaluation of plant specific effects for Mark I, II and III i
containments.
UNCERTAINTY ANALYSES PROGRAMS 0
Complete detailed uncertainty source term analyses Sequoyah S HF and 2
l l
Sequoyah V sequence with the STCP.
0 Extend the source tem uncertainty analysis for the Peach Bottom TC sequence analysis utilizing best estimate mechanistic codes.
(MELPROG, CONTAIN,etc.)
i i
i i
49
t-
.IV. BACKGROUND AND OBJECTIVES OF INDEPENDENT REVIEW 0F RESEARCH PLANS
-The technology used to estimate radiological source terms for postulated severe accidents has been reported in NUREG-0956, where the technical basis for the scientific tools that the U.S. Nuclear Regulatory Commission (NRC) has
' developed for future use in its regulatory considerations of postulated severe accidents are documented.
However, as discussed in previous sections the understanding of severe accident phenomena:is far from complete, and many uncertainties still exist.-
While recent severe accident studies have pemitted advances in the analysis of source tems, this improved understanding has also allowed a more clear identification of major uncertainties. The phenomenological issues summarized in Section II of this report are believed by NRC staff and its contractors to contribute most significantly to uncertainties in current predictions of the source term.
In response to the Connission's request, as part of the issue resolution plans, the NRC staff has recommended obtaining the independent review by experts of the research plans for the major areas of uncertainty. Brookhaven National Laboratory (ENL) has been selected by the NRC staff to provide for the review. BNL will conduct and administer t..e review process, and Dr. Herbert Kouts, Chairman, Department of Nuclear Energy. BNL, will lead the review activity. The objective of the BNL review effort is to provide:
1.
A review of:
(a) the state of the art in each technical area identified.
(b) the staff's description of the extent of uncertainty in each
- area, (c)currentresearchplansforresolvingtheseuncertainties,and (d) any aspects of the Chernobyl accident related to each area.
2.
Technical connents on the staff's description of the extent of i
uncertainty in eLeh area.
3.
Technical conclusions on the adequacy of current research plans for resolving these une.wt:inties.
4.
Identification of any implications of the Chernobyl accident for the technical area in question.
Definition of Review
~:
In order to avoid duplication of effort, the BNL review process will use the technical conclusions of two recent review groups namely, the Steam Explosion Review Group and the National Academy of Sciences study on hydrogen combustion. Furthermore, the remaining issues will be combined into i
{
50
four groups to consider the interdependency of the various physical processes and comon features of some of the issues. The grouping of issees is as follows:
GROUP I NATURAL CIRCULATION IN THE RCS CORE MELT PROGRESSION AND HYDROGEN GENERATION GROUP II HIGH PRESSURE MELT EJECTION STEAM EXPLOSIONS HYDROGENCOMBUSTION(ASNEEDED)
GROUP III CORE-CONCRETE INTERACTIONS GROUP IV IODINE CHEMICAL FORM REVAPORIZATION OF PREVIOUSLY DEPOSITED FISSION PRODUCTS Schedule 1.
BNL has infomed us that they have scheduled their review, including the independent expert reviews of the research plan, for resolution of the major areas of uncertainty as follows:
o Review Group Meetings January 1987 o
AdditionalReviewGroupMeetings(asneeded)
February 1988 o
BNL report on technical conclusions from reviews to NRC staff March 1987 2.
The NRC Staff will consider the conclusions from the BNL report in order to submit a revised research plan to the Comission in April 1987.
51
Schedule The expert review groups will meet in-the Washington, D.C. area at locations to be announced.
The plan to address the major areas of uncertainty is to be sufficiently detailed in order to support a Commission request to Congress for additional resources, if necessary. Therefore, reports from the individual comprising the review groups containing any recommendations for expanded research programs must be available to BNL for subsequent publication in'a BNL formal report by the end of March 1987. To meet this target, the following timetable has been established:
Meeting Purpose
.Date Overview Group Discussion of objectives and November 14, 1986 Meeting I guidelines Panel Group To obtain relevant information Early January 1987 Meetings I by group members from principal technical contractors Panel Group For group members to formulate Late January 1987 Meetings II opinions and recomendations Overview Group To sumarize individual Mid February 1987 Meeting II findings and recommenda-tions to BNL draft report
(
Overview Group Review of BNL report which Mid March 1987 Meeting III summarizes and correlates the findings and recomendations t
of the individual members 52
ATTACHMENT 1
_ j -s ae g 9,,
UNITED STATES f
g NUCLEAR REGULATORY COMMISSION
.{
[.
wAsNewcToN, D. C. 20555 e...*
..A 2 0 m MEMORANDUM TO: Those on attached list FROM:
M.
Silberberg, Chief -
Fuel Systems Research Branch R. T. Curtis, Chief Containmer.t Systems Research Branch
SUBJECT:
S VERE ACCIDENT TECHNICAL UNCERTAINTY RESEARCH REPORTS In view of the impending conpletion of NUREG-1150 and the recent reviews of the Source Term Reassessment document (NUREG-0956), I consider it necessary that the Accident Evaluation Branch have in place definitive and extensive research documentation which support our positions on areas of technical uncertainty. These areas have been defined and stated many times vis-a-vis our interactions with IDCOR, the APS review of BMI-2104, the NUREG-1150 study group, and our own NUREG-0956. These research reports will supplement the NUREG-1150 issue papers, and will be a basis for discussion of these issues for the 1150 review in the Fall. These reports will also be useful to NRR for their October 1986 milestone, "Research Update," in the SAPS Implementation Plan.
It is also important to establish a framework by which the research pro-grams which ultimately quantify and reduca the uncertainty for each area can be implemented into analysis packages such as the STCP, MELCOR, CONTAIN, HECTR, MELPROG, and SCDAP.
In this manner, we can be assured that state of knowledge infonnation is available and usable by members of the staff who must make decisions regarding the Comission's Severe Accident Policy Statement.
Accordingly, we have developed a list of research reports and suggested authors for those technical areas under the cognizance of the AEB (see attachment).
We have also included a suggested outline and fonnat f6r the papers. We expect a completed draft of each paper to be available for review by G. P. Marino, J. Mitchell, R. Meyer, and M. Silberberg no later than October 1,1986, and G
f the final report published (either as a NUREG/CR or a NUREG) by November 1,1986.
It is expected that the named AEB staff person (s) contact and arrange for contractor personnel participation for their reports.
In view of the AEB staff persons' ultimate responsibility for completing this work, we will entertain requests for modifying selected contractor personnel and/or technical content.
Please see Dr. Marino if you have any suggestions.
Sk.
7 M. Silberberg, Chief Fuel Research B
- h R. T. Curtis, Chief Containment Systems Research Branch t
i
7_
SEVERE ACCIDENT TECHNICAL UNCERTAINTY RESEARCH REPORT-(ACCIDENT EVALUATION BRANCH)
It is expected that detailed research papers be developed for the following technical uncertainty areas by the Accident Evaluation Branch:
- PAPER TITLE AND AUTHORS-1.
MELT PROGRESSION & HYDROGEN GENERAT20N by R. Wright & W. Camp (SNL) 2.
CORE / CONC. INTERACTIONS & HEAT TRANSFER by B. Burson & D. Powers (SNL) 3.
REVAPORIZATION by'L. Chan, D. Powers (SNL) & G. Berna (INEL) 4.
NATURAL CIRCULATION by J. Han & W. Camp (SNL) 5.
RCS F.P. & AER0f0L DEP. by L. Chan & J. Gieseke (BCL) 6.- RCS F.P.. RELEASE & CHEM. FORM by R. Meyer & R. Lorenz (ORNL) 7.
EX-VESSEL F.P. & AEROSOL RELEASE by B. Burson & D. Powers (SNL)
- 8. ;iX-VESSEL F.P. & AEROSOL DEP. by P. Worthington & T. Kress (ORNL)
- 9. ! DIRECT HEATING by T. Lee & D. Powers (SNL)
- 10. HYDROGEN IGNITION & BURNING by P. Worthington & M. Berman (SNL)
- 11. ALPHA MODE CONT. FAILURE by T. Lee & M. Berman (SNL)
I i
i e
,w
-r,,, - -,, - - - - -, - - - -, - -. - -,,,,.
.c
.--v-
UNCERTAINTY PAPER DUTLINE The format and content of the technical uncertainty research papers should follow the suggestions given below:
I. INTRODUCTION State what the uncertainties are, why they are considered important, and estimate their effect on source term, containment loading, and containment failure times from current knowledge. Compare current modeling (i.e., the STCP) with current IDCOR models. Please reference any existing issue papers.
II. DESCRIPTION OF PAST, PRESENT, AND FUTURE RESEARCH Discuss past, present, and future research (through FY 1988) and describe how and why it will help quantify and reduce the uncertain-ties.
Include all accomplishments achieved to date. Give a schedule of experimental milestones and documentation deliverables. Discuss and propose research beyond FY 1988 if its considered necessary.
III. TECHNICAL UNCERTAINTY EVALUATION Discuss how the uncertainties will be reduced by the end of the current program (through FY1988). -Discuss potential future (beyond FY1988) reduction in terms of program needed to accomplish expected results and the benefits of better estimate.
IV. IMPLEMENTATION OF RESEARCH RESULTS Recommend the membership of an NRC/ Contractor " Technical Uncertainty i
l Working Group" whose charter will be to recomend and implement changes and improvements to the STCP, MELCOR, MELPROG CONTAIN, and SCDAP codes resulting from the new knowledge gained.
V. SUPMARY Sumarize all of the above in no more than two pages being sure to highlight the importance of the technical area and the projected minimum uncertainty expe.cted (i.e'.Section III above)...
l l
G w
-r-m--
s+.,wr-,--,->w--e a-w-,+,e-e m---
w-e--,--m--n.-
- mse m--
-eme--
e w--
-,e--a,w-w-ir wee.
e--w---v--.-
+v-, - - -
-e
---w-
-e-
--wT- - +- - -
'-irw--w,wwr--r m---w w-e----