ML20195E323
| ML20195E323 | |
| Person / Time | |
|---|---|
| Issue date: | 11/13/1998 |
| From: | Tim Reed NRC (Affiliation Not Assigned) |
| To: | Sullivan E NRC (Affiliation Not Assigned) |
| References | |
| NUDOCS 9811180314 | |
| Download: ML20195E323 (74) | |
Text
-
(
{' November 13, 1998 g MEMORANDUM TO: Edmund Sullivan, Acting Chief Materials and Chemical Engineering Branch
@[b Division of Engineering Office of Nuclear Reactor Regulation THRU: EmmeNur y, k Ckef Section B j Materials and Chemical Engineering Branch Division of Engineering Office of Nuclear Reactor Regulation original signed by FROM: Timothy A. Reed, Senior Project Manager Materials and Chemical Engineering Branch Division of Engineering Office of Nuclear Reactor Regulation
SUBJECT:
SUMMARY
OF THE OCTOBER 28 AND 29,1998 TECHNICAL I MEETING WITH NEl/EPRl/ INDUSTRY TO DISCUSS RISK GUIDANCE I FOR INCORPORATING SG TUBE DEGRADATION INTO PRAS On October 28 and 29,1998, the NRC staff met with representatives of Nuclear Energy Institute (NEI), Electric Power Research Institute (EPRI) and industry to discuss different approaches for estimating the risk acsociated with steam generator (SG) tube degradation. Meeting attendees are identified in Attachments 1 and 2.1he meeting agenda and slides are provided as Attachment 3.
The two day meeting began with the staff discussing in detail its work to assess SG tube degradation risk as documented in NUREG-1570. The industry followed with a detailed description of the various aspects of its effort to perform pilot risk assessments that incorporate tube degradation. Additionally, the status of ongoing PRA work for Calvert Cliffs was discussed.
Throughout the two day meeting, industry representatives and the staff exchanged the insights they had gained from the different approaches utilized for considering SG tube risk. The staff was favorably impressed with the significant progress made by industry in developing a methodology for addressing SG tube degradation risk and believes there is a significant probability that the effort will produce a successful approach.
Attachments: As stated CONTACT: T. Reed, EMCB/DE 415-1462 DISTRIBUTION: PUBLIC PDR SJCollins FJMiraglia EMCB RF j SMagruder JStrosnider GClainas OGC Document Name: G:\ REED \OCT98RSK. MIN {
INDICATE IN BOX: "C"= COPY W/O ATTACHMENT / ENCLOSURE,"E"= COPY W/ATT/ ENCL, *N"sNO COPY OFFICE E M CJ k' E 1 b EMCB:DE N E y
NAME Tree SLong ur hy //
DATE iI/fh98 l I/ $98 / G/98 _
,[ y;.
,, o OFFICIAL RECORD COPY -
K 9811180314 9'1113 8 PDR REVGP ERGN t C y L
g EEN
7
(
& CE%ts M 1 UNITED STATES
[
j 4'
NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20556-0001 g,,* November 13, 1998 MEMORANDUM TO: Edmund Sullivan, Acting Chief Materials and Chemical Engineering Branch Division of Engineering Office of Nuclear Reactor Regulation THRU: Emmett Murphy, Acting Chief //v Section B c Materials and Chemical Engineering Branch Division of Engineering Office of Nuclear Reactor Regulation , ;
7 i FROM: Timothy A. Reed, Senior Project Manager l Materials and Chemical Engineering Bran i Division of Engineering )
Office of Nuclear Reactor Regulation
SUBJECT:
SUMMARY
OF THE OCTOBER 28 AND 29,1998 TECHNICAL MEETING WITH NEl/EPRl/lNDUSTRY TO DISCUSS RISK GUIDANCE FOR INCORPORATING SG TUBE DEGRADATION INTO PRAS On October 28 and 29,1998, the NRC staff met with representatives of Nuclear Energy Institute !
(NEI), Electric Power Research Institute (EPRI) and industry to discuss different approaches for estimating the risk associated with steam generator (SG) tube degradation. Meeting attendees are identified in Attachments 1 and 2. The meeting agenda and slides are provided as Attachment 3.
The two day meeting began with the staff discussing in detailits work to assess SG tube degradation risk as documented in NUREG-1570. The industry followed with a detailed description of the various aspects of its effort to perform pilot risk assessments that incorporate tube degradation. Additionally, the status of ongoing PRA work for Calvert Cliffs was discussed.
Throughout the two day meeting, industry representatives and the staff exchanged the insights they had gained from the different approaches utilized for considering SG tube risk. The staff was favorably impressed with the significant progress made by industry in developing a methodology for addressir,g SG tube degradation risk and believes there is a significant probability that the effort will produce a successful approach.
Attachments: As stated l
CONTACT: T. Reed, EMCB/DE 415-1462 I
i
'(
NEl/EPRl/lNDUSTRY TECHNICAL MEETING SG TUBE INTEGRITY RISK GUIDANCE .
OCTOBER 28,1998 J.lST OF ATTENDEES NAME ORG/ POSITION l
- 1. Tim Reed - NRC/NRR/DE/EMCB
- 2. _Clive Callaway NEl
- 3.' Mati Merilo EPRI l
- 4. Ron Gamble Sartrex Corp
- 5. Dave Modeen NEl i
- 6. Bruce Mrowca BGE '
_.7. Bill Hannaman SAIC
- 8. David Finnicum ABB
< 9. David Stellfox McGraw-Hill
~ 10. Emmett Murphy NRC/NRR/DE/EMCB
- 11. Steve Long NRC/NRR/DSSA/SPSB
- 12. Joe Muscara NRC/RES 13.' Richard Lee NRC/RES -
- 14. Ted Sullivan NRC/NRR/DE/EMCB
- 15. Jack Hayes NRC/NRR/PERB
- 16. Mike Schoppman FPL
.17. Gary Elder Westinghouse
- 18. Ed Fuller Polestar Applied Technology
- 19. Marc Kenton Creare
'20 Joe Donoghue i
. NRR/DSSA/SRXB 21 Scott Barber NRC/OEDO I
- 22. Anthony Saccavino BGE
- 23. Fred Anderson - Tetra Engineering l
(. ATTACHMENT 1
!~
J 4
, - - c.----, .- -
y . _ _.
1 NEl/EPRl/ INDUSTRY TECHNICAL MEETING SG TUBE INTEGRITY RISK GUIDANCE OCTOBER 29,1998 LIST OF ATTENDEES NAME ORG/ POSITION
- 1. Tim Reed NRC/NRR/DE/EMCB
- 2. Clive Callaway NEl
- 3. Mati Merito EPRI
'4. Ron Gamble Sartrex Corp
- 5. Dave Modeen NEl
- 6. Robert Cavedo BGE
-7. Bill Hannaman SAIC
- 8. Charlie Tinkler NRC/RES
- 9. Noel Dudley ACRS
- 10. Emmett Murphy NRC/NRR/DE/EMCB
- 11. Steve Long NRC/NRR/DSSA/SPSB
- 12. Joe Muscara NRC/RES
- 13. Richard Lee NRC/RES
- 14. Ted Sullivan NRC/NRR/DE/EMCB
- 15. Don Streinz ABB CENO
- 16. David Ayres ABB CENO
- 17. Ed Fuller Polestar Applied Technology
- 18. Marc Kenton Creare
- 19. Anthony Saccavino BGE
- 20. Fred Anderson Tetra Engineering
- 21. Joe Donoghue NRC/NRR/DSSA/SRXB ATTACHMENT 2 i
' l NRC/NEl/ INDUSTRY MEETING RISK ASSESSMENT RE:SG TUBE INTEGRITY OCTOBER 28 AND 29,1998 l MEETING OBJECTIVES:
- 1. NRC and industry gain a better understanding of each other's risk assessment approaches
- 2. Identify areas of agreement, areas of disagreement, and areas where further work is needed to better understar.d differences or to resolve differences AGENDA: OCTOBER 28,1998 (OWFN CONFERENCE ROOM 384):
9:00 a.m. INTRODUCTION / PURPOSE NRC//ndustry 9:15 a.m. OVERVIEW OF PREVIOUS NRC RISK WORK
~ NUREG-0844 E. Murphy (NRC)
- NUREG-1570 S.Long (NRC) 11:00 a.m. OVERVIEW OF EPRI SGTI RISK ASSESSMENT METHODOLOGY E. Fuller (Polestar) 11:30 a.m .
SELECTION OF ACCIDENT SEQUENCES E. Fuller (Polestar)
NOON LUNCH 1:00 p.m. .
HUMAN RELIABILITY ANALYSIS AND EXPANDED OPERATOR ACTION TREES G.W. Hannaman (SAIC) 1:45 p.m.
THERMAL HYDRAULIC ANALYSIS M. A. Kenton(Creare) 2:30 p.m. ACCIDENT PROGRESSION EVENT TREES E. Fuller (Polestar) 3:00 p.m. PROBABILITY OF TUBE RUPTURE M.A. Kenton (Creare) 3:30 p.m. DISCUSSION NRC/ industry 4:00 p.m. ADJOURN ATTACHMENT 3
- )
- ~
NRC/NEl/ INDUSTRY MEETING j RISK ASSESSMENT RE:SG TUBE INTEGRITY j OCTOBER 28 AND 29,1998 J
! AGENDA: OCTOBER 29,1998 (TWFN CONFERENCE ROOM 10A1)i l t
i j f 9:00 a.m. CALVERT CLIFFS RESULTS R. Cavedo(BGAE) 4 11:00 a.m. WRAP UP NRC//ndustry !
-AREAS OF AGREEMENT 1
-AREAS OF DISAGREEMENT ;
-AREAS FOR FURTHER WORK (TO RESOLVE OR BETTER l UNDERSTAND DIFFERENCES)
NOON ADJOURN 11necessary - Conference Room OWFN 1684 is reserved for attemoon if meeting runs over the above schedule.
- ** W es e e o eere d ,, ,
's l NUREG-1570 Overview ,
(S. Long) l
Purpose:
To estimate the risk associated with allowing SG tube degradation to go to the point contemplated by the proposed rule (i.e., the conditional probability of tube rupture under " steam line break conditions" becomes 0.05)
Relevant seauences in PRA-level 1: spontaneous tube rupture tube ruptures induced by secondary system depressurization events tube ruptures induced by primary system over-pressurization events .l level 2: tube ruptures induced by secondary depressurizations at normal temperatures tube ruptures induced by secondary depressurizations at elevated temperatures (found to depend on secondary depressurization, also)
Assumntions:
- 1. The spontaneous rupture frequency was assumed to remain unchanged.
- 2. Conditional tube rupture probabilities estimated for end of cycle were applied to entire fuel cycle.
- 3. Flaws > 0.25" long that propagated through wall under high temperature conditions were treated as ruptures due to concem about the effects of high temperature steam erosion / cutting.
- 4. SG mixing based on 1/7 scale model data Methods:
l Spontaneous SGTR methodology and results were taken from Surry NUREG-1150 model. 1 ISGTR from secondary system depressurization events was evaluated in INEL 95/0641.
. )
Successful mitigation assumed to require completion of cooldown to cold shutdown 1 conditions before RWST exhausted.
. RELAP analyses used to establish success criteria, required operator actions and tumng.
- HEPs calculated in a manner that coupled time available for all actions, assuming surplus l RWST inventory consumption at various rates at the different times for the various actions.
(
1$GTR due to ATWS events was bounded in Section 2.1.3 of NUREG-1570.
\
T
, r 2
- Based on NUREG-1150 ATWS model.
- Concern about changes in MTC since NUREG-1150 analyses were completed.
ISGTR during core damage sequences based on frequency of challenges extracted from NUREG-1150 level 2 results
- Frequency of"high-dry" plant-damage states extracted from Surry and Sequoyah analyses.
= Frequency of secondary depressurization during accident progression extracted.
+ 0.05 conditional probability of tube rupture added for each SG that depressurizes while primary remains pressurized and RCS remains near operational temperatures.
. Probability of thermally-induced SGTR reevaluated, adding effects of flaws.
- Dir.inction made in release categories for ISGTRs with open secondary system (or primary pressure still above secondary relief valve setpoints) and ISGTRs with intact (although leaky) secondary.
Methodology for estimating thermally-induced SGTR probabilities:
- RELAP/SCDAP T-H calculations for each sequence in accident progression event tree
- RELAP/SCDAP files for time histories of RCS pressure and tube temperature used to recalculate creep damage index for tubes
- ANL's modified Larson-Miller equation used for computing creep damage as a function of flaw size (length & depth)
=
distribution of failure times for other RCS components (surge line & hot legs) computed from RELAP/SCDAP time histories of pressure and temperature and uncertainty of Larson-Miller parameters for surge line & hot leg)
Probability of specific tube flaw failing first was based on predicted time to failure of flawed tube compared to probability of other component failing earlier.
Probability that no flawed tube fails first was done in two parts:
- 20 flaw length / depth bins were each assessed with Monte Carlo treatment of l (intra-bin) length and depth plus tube thickness and diameter and L-M parameter
! - estimates of flaw size distribution were binned combined with probabilities of i failure for each bin
- Note that flaws that were assumed to fail due to AP conditions prior to heat-up l were not included in the thennally-induced probability estimates, l
l Sensitivity studies:
l Sensitivity studies were performed on tube temperature (as a function of time), flaw
- distribution estimates, RCS pressure, probability of SG depressurization, flaw stress concentration factor and predicted surge line failure time.
Some sensitivities to variations in pairs of parameters were addressed Most sensitive factors wen flaw distribution and tube temperature Other designs also considered (Sequoyah for lower secondary depressurization probability and ANO-2 for different T-H response of RCS to core oxidation)
m i
3 Findings:
l =
Tube failure probabilities for elevated temperature conditions appear to be low except for conditions where secondan is depressurized Nan'-e ofsensitivity to one variable can depend strongly on assumptions about other sens:tive variables Thermally-induced failures of flawed tubes may not be as important as are pressure-i induced failures at normal temperatures, but this is highly dependent on many factors, l including flaw distributions, and human error probabilities. Also influenced by RCS j , design (surge line materials and orientation, RCP seal type, loop seal stability) and risk profile ofplant.
Difficulties in further applications:
l Currently available ECT technologies do not produce results in terms of crack lengths and depths a
Concems about cutting rates for high temperature / high pressure steam passing through ;
tube cracks '
Differences between MAAP and RELAP in predicting tube temperatures, especially for partially depressurized RCS conditions associated with accumulator injection during core oxidation, including:
- loop seal stability l
- mixing model 1
l 1
l
. ;i- - t!i i!e !;iiI' l! >!;
0 0
0
_ ~
0
- 6 1
e ,
i m -
T W s -
v -
s .
0 e
0 c a ' 0 0
e 5 d 1 I
n -
_ g e -
a m
a -
D -
p -
e - 0 e
r .
0 0
0 C - 4 1
l e se -
p en i
bl g m tuee a gl -
x Guo Ssh rt E '
0 0
0 0 8 6 4 0 O
2 0.3 1 0 0 0 O1 m L xe? eceEeV Cees.o
!\! ,;( l,l!l lli1tl ,l ,
.- J RCS Pressure vs Time Case RES-9 20 . .
1 15 -
(1. -
y I
t i
e_
s _
U m
m 10 -
8 -
o.
t m _
O OC -
5 -
b k
0.0 - ' - '
0.0 10000.0 20000.0 time (seconds) i
!i: '.; !! .. !, .::iil!lli !l{il; e
i T
m P 0 s - 0 v .
0
' 0 0
s * < 0 2
e r \
t u
a r
e p
m k
PL d n d e g ne d
)s e9 -
nge d n
TS E e el
/
)
l gt d o eoo l c b
eR 1
( l t h c e s
u esa h o- -
(
e TC - ss ee 0 0i m
- sbb t C .
b euut t
' 0 0
r uww 0 t
o t wnnoo 1 a
r l
o mm e f-t uu t n uteer G
e / '
i t
or eee vv v cccie t -
eee m www a r r r ooo t
e ccc
)))
S 1
( (2 (3
- Q- 0 0 0 0.0 0 0 0 0 0 0 5 1
2 B. G ES O S
.s RCS Pressure Boundary Failure Probability vs Time Case RES-9 1.0 -
I l ,'
I i'
. :s .
l lI '
/
I RCS PB intact .h' /
I ,,
0.8 l/
l hot leg B tailuse i i I I j.j O.6 -
surge line fature Il g
/
f'
'I
/
d hot leg A failure ; ,
2a .
\1 y :l 'i
,./
g hot leg C failure -
& I
,3 ll ', s' -
O.4 -
I i '
I i ,,' .i I io/
i
I !If I it .
I i' 0.2 -
i j'l -
I g-I q f
/
O.0 - ' / ' '
O.0 10000.0 20000.0 time (seconds)
RCS Pressure Boundary Failure Probability vs Time Case RES-9m StaWess Steel Hot tags 1.0 .
RcS m- r ,.-
,i 0.8 -
hotlog B taBure I!
t; - -
i surge EnefaBuse l
'I 4 3 t i 4 0.6 - : ' -
5.# l 2 -
e :
a 0.4 -
hotleg A faBure ,' -
I
, t i
~
t t
t t
0.2 -
l t
l t
. l #
hotleg C faRure -
I u -
h ,
0.0 - ' - = -
O.0 10000.0 20000.0 time (seconds)
. . J Comparison'of NRR " Average" and RES " Moderate" Distributions 2.0 . . . -
c c c '
c o RES amind cracks > 1" c o NRR auds = 1" 1.5 as x
0 W
w o
o u
1.0 -
e
.O -
E o
C i
0.5 -
i
~
. - x.O a 0.0 -
20.0 40.0 60.0 80.0 100.0 ;
b* , f 5
- h .
4 * * .* I 4+ (IM ' OU** 'tT*'*" #
k' *'#
o e i i i Q
4 C -
O
- bo 5 -
.m_
Q ,
.c D
C "
($
5 m "s 6 a m .
[
CD W
T o e
d
, , o o
J e e w m od d 6 6 6 6 X 416uel pesoxe ley) uollosa;
. c; Temperature Ssnsitivity of TI-SGTR 1.00 . . . . . . .
o -o case 1R - RES flaw distr.
0.90 -
e o case 3R - RES flew distr. -
e o case 3R - NRR flaw distr.
0.80 - 7 7 case 9R - RES flaw distr. -
0.70 -
5 0.60 -
.8 ~
e
" 0.50 0~
V-0 y 0.40 -
p -
0.30 - -
0.20 -3 -
0.10 - -
0.00 0 '
' -- ' ^
- ^ ^ ^ ^
-80.0 -40.0 -20.0 0.0 20.0 40.0 e0.0 80.0 100.0 SG tube temperature offset (K) l
_J
e o
- ^
E -
e . .
2
'E , .
k .
D .
o 4C
- ~
R ^
3 '
O (D
E -
3
, e l
e - -
o l @@ d O. E l CO j o . d o,
CC e
.s
~~
C X1
' u> x mm o mz (n on LLI
$ oa a3 b
i 5
.g ci e O C
g . . . . -
'~
CO o 9 o,
, o o
! bl10S-1110 4!I!qegoJd l
l I
l- \
l l
l C.
A "
l A _
%A _
eo s$
'h.
Op _
L1 -
m o-dg hb '
- 3*
o w _
t3 88 en 1 -
mm o.
l EZ oo
%A _
00 i .
q)AA _
. o
" ao o 6
! eJni!s; meg eqni es }o 4)ll!qeq0)d
(
l
[
i.
e l
Selection of Accident Sequences Edward L. Fuller Polestar Applied Technology One First Street, Suite 4 Los Altos, CA 94022 !
650-948-8242 .
edfuller@ polestar.com l l
l Presentation to Nuclear Regulatory Commission l Rockville,MD October 28,1998 i
Categorization of Accident Sequences
, With Tube Rupture Events Spontaneous SGTR-initiated sequences.
Secondary side depressurization sequences with induced tube ruptures.
Severe accidents with pressure- or temperature-induced tube ruptures.
- Primary system at high pressure and steam generators dried out (high/ dry sequences),
l 1
Practical Aspects of Candidate Sequence Selection
!
- Need to consider both internal and external events
- For PRAs with small event trees, can select the candidate sequences directly from the event trees.
- For sequences with large event trees, must select candidate sequences from examination of key plant damage states.
- Some candidate sequences are design- or procedure-specific. For example,
- Some CE plants don't have PORVr; consequently RCS pressure can remain high even when power still avaliable.
- Some B&W plants still allow pump bumping, potentially allowing high-temperature steam to reach SG tubes .
Typical Times to Core Damage for High/ Dry Sequences
- Early Core Damage sequences
- about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> due to loss of primary and secondary cooling.
- Mid Core Damage sequences
- up to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> due to AFW turbine driven pump failure upon depletion of batteries supplying instrumentation power
- Late Core Damage sequences
- about 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> due to failure to establish RHR cooling 2
I
< . 1 l
End State Parameters for Selecting CanrF ! ate Accident Sequences for Plant A @ Plara)
(PRA Has Large Event Trees)
Position
- End State Parameter Values 1 RCS pressure S = set point, H = high, I = intermediate, L= low 2 Steam generator cooling A = available, X = not available, N = NA 3 Water from RWST and Y = yes, N = no backup sources injected j
- Position of parameter value in end state label from left to right I
1 Description of Key End States (Plant A)
SXN; RCS pressure at set point, no SG cooling, no injection from RWST.
l Representative sequence: loss of switchgear ventilation, AFW
! system fails. (early CD)
SXY: RCS pressure at set point, no SG cooling, injection from RWST.
Representative sequence: reactor trip, AFW system fails, feed and bleed cooling fails. (late CD)
HAN: RCS pressure high, SG cooling until dry out or overfill, no injection from RWST.
Representative sequence: loss of off-site power, deisel generators fail, RCP seal integrity lost, AFW system initially available, but fails after level control lost and operators fail to l isolate SGs. (mid CD) l INN: Spontaneous SGTR, RCS at intermediate pressure, AFW j terminated to ruptured SG, no makeup to RWST.
Representative sequence: Non-isolated SGTR, secondary side leak to atmosphere, makeup to RWST fails. (late CD) 4 3
High/ Dry Sequence End States for Plant A (IPE Has Large Event Trees)
High & Dry Frequency (per reactor year)
Total Timing End States intamal Seismic 3 SXN 3.1E-06 5.19E 061.8E-07 8.47E 06 Earty CD SXY 3.74E-06 4.04E-071.18E-06 5.32E-06 Late CD HXN 3.72E-071.16E 06 2.2E-08 1.55E-06 Earty CD HXY 1.13E461.61E-07 4.2E 09 1.29E-06 Late CD g 3.59E461.95E-05 2.1E-06 2.52E-05 Mid CD alt. 1.19E45 2.64E 05 3.49E46 4.18E46 i
Supplementary Analysis of Plant A
- For Mid Core Damage sequences the AFW system functions as long as batteries operate.
- After batteries are depleted AFW system continues to function, but level indication is lost.
- Level control is difficult, but still possible. The turbine-driven pump could still be operated and ADVs could be manually operated.
- Backup water supplies are also available, which were not credited in originalIPE.
- Significant credit can thus be taken for averting a thermal challenge situation.
4 1
I
l <
Some Candidate Sequences for Plant B l
(CE Plant; IPE Has Small Event Trees)
= Spontaneous SGTR with failure of both RCS and secondary heat sink removal l
Main steam line break with failure to isolate the faulted i
Loss of power conversion system (loss of feed water),
l AFW unavailable, pressurizer safety valves re-close.
l Loss of off-site power, AFW not available.
l Station blackout, de power initially available and l operators maintain AFW flow by manually operating ADVs until SGs either dry out or overfill.
i I
l Table 2.4 Candidate Accident Sequences for Plant B Sequence Accident Event Tree Number Clan Free /ItY Timinefl)
Small LOCA 12 1A 1.lE47 early Small.SmallIDCA 10 IA cl .0E48 _._ early
. SGTR 10.20 IA 2.lE47 early i
Main Steam time Break 7.11.13 1A 6.l E47 early l Less of CCW 3.6 IIA 5.1E-07 late less of125V DClus 5 1A l.7E47 early loss oll25v_pC_B_us 33 IIA 1.6E-08 early__
Less of Offsite Ppwer 11 IA J.0E46 earty Station Blackout $ IC J.1E-07 mid Station plackout 20 IC <l .0E48 _early I Less of PCS _ 5,19 _1A_ l.6E46 _early_
Transient with PCS Initally Available 5 1A 1.3E-06 early
[ 9,11,13,15, ATWS 17.25.27.29 IV 1.3E-06 early l Total 8.9E-06 Motes: (l) Core damage ummg: early = 1 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />; mid = about 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> due to loss of battery power; late > 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> due to RCP seal failure l
5 i
1 Candidate Accident Sequences for Plant C (B&W Plant)
- Spontaneous SGTRs with failure to isolate the SGe
- MSLB events leading to tube leakage or rupturee
- Severe accident high/ dry sequences with pressure-induced tube ruptures.
- Severe accident high/ dry sequences where operators start reactor coolant pumps during the accident and l
! clear one or more RCS loop seals.
l l
Table 2.5 Candidate Astident Sequences for Plant C F 4e nt Type' Timing e M6gh & Dry Feedwater ECCS Falls High & Dry &
Damage F regl RY Systems Fall g,8 RCP Por S ta te To FreemV ED4 Emall LOG A sedy 2.1 E 07 sted inje cten 0 sD5 EmesLOcA e ersy s.7E so start tarsstea o 505 EmesLOCA e arly 3 7 E-07 etert tapesten 3.7E 07-are EmesLOcA seri, t .O E -O s sis rt enfe sten t .s E.O s itDi EmsaLOcA no te i.eE Os sta n E wnshower 1.9E 06 1105 EmesLOCA la te 3 1 E.05 start 5 washo wet a tE 09 1201 EmeELOCA late 4 5E-07 rua sw tchever 4 sE 07
. 4 Di Trans ents e arly 1,2 E.O s start imposten 0 1401 T ransen te early 4 DE 07 htart lapesboa 1.stE 07 1405 T ra nsents medy a 7E 10 sts,t ange stien 0 tari Transents earty s .2 E-os siert anpo sten 0 14P5 Treas ente e arty a sE-Os start tarschen 0 isP6 Transe nte med 1.3 E 0 7 rua laseshen s 03E-05 1sPE Trenoients m MI 2 SE-10 rua inse chen 0 1sO6 T ran sents iste 2 4 E-05 etert S witche ver 1.4 E -05 iso 5 Tre memats note 4 3 E-O s start 5 enshe ve r 1. l t -De ISPI T ra n s.e n ts tale 3 EE 07 sta rt Ewitchever 3 st.07 7F5 T reno ma ts is te t .s E-0 9 start G oncheve r i 3E 09 1701 T ra nsents hie 2 2 E-07 rwa 5 edshever 22E07 17P6 T reassate tote 4 s E.O s run G odsho wer 5.s E -O s To teis 3.3 E.e s s .sE 45 Notes.1. Tressients insiede cycling RC5 relief valve
- 2. Core uncovering time: [arjy. Ebout 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. LE. abost 20 boers
- 3. Switchover = Failars to make switchover from injection to recirculation l
6 i
l l
,+ - _ _ . . - - -_ _. - . _ _ _ . - . .. . . - - . - -
O I
I l
Overview of EPRI's Steam Generator Tube Integrity (SGTI) Risk Assessment Methodology Edward L. Fuller 1 Polestar Applied Technology One First Street, Suite 4 Los Altos, CA 94022 650-948-8242 edfuller@ polestar.com Presentation to
! Nuclear Regulatory Commission l Rockville, MD j October 28,1998 l
\
l l Purposes of EPRI Project i
Develop a methodology to support requests to use Alternate
!- Repair Criteria (ARC) for assessing risks from loss of steam l generator tube integrity.
- Build upon previous EPRI work.
- Be consistent with approaches developed by the NRC.
, - Treat systems, operating procedures, and thermal hydraulic behavior as realistically as possible.
- Demonstrate methodology on a number of pilot plants.
- Model actual systems and procedures of the pilot plants.
- Use current probabilistic risk assessments for pilot plants as starting points.
- Evaluate risks and compare against criteria derived by NRC.
e 1
l
NRC Documents Pertaining to SGTI Risk Assessment i 1
- NUREG-0844 and NUREG-1570 contain risk analyses.
{
DG-1074 states that proposed changes to licensing basis must be supported by a risk assessment, appropriately considering defense-in-depth.
- Applies to changing probabilistic structural and/or accident leakage performance criteria, since these would be changes to the licensing basis.
Risk implications generally defect type- and SGDSM-specific.
Guidance for submitting licensing basis changes is provided in Regulatory Guide 1.174 (now issued).
- Risk impacts are changes in core damage frequency (6-CDF) and in large early release frequency (5- LERF).
Hierarchy of SGTI Risk Measures Risk Measure Where ADDlled Off site consequences NUREG-0844 NUREG-1150 EPRI TR-106194 6-LERF DG-1074 & RG 1.174 Bypass frequency NUREG-1570 Tube failure probability 11/96 Draft RG Thermal challenge frequency 11/96 Draft RG EPRI TR-107623, V2 2
l l
l l
l Risks Are From Accidents Leading to High l RCS Pressure and a Dry Steam Generator
- Design basis accident- or ATWS-Induced tube ruptures (e.g. MSLB with failure to isolate faulted SG).
( - Pressure-induced tube ruptures possible.
!
- Severe accident-induced tube rupture events (e.g.
station blackout).
- These are called High/ Dry Sequences.
- Pressure-or thermally-induced ruptures possible.
l
(
I I
l , Thermal Challenge to SG Tubes i
During High/ Dry Sequences
" reasonable" likelihood of thermally-induced SG tube creep rupture prior to either hot leg or vessel failure.
,
- Thermal challenge frequency (TCF) is an imprecise, l and conservative, measure of the risk from thermally-L induced SG tube creep rupture. It was originally proposed by NRC as a risk measure.
- Original pilot project focused only on severe accident-induced sequences, and stopped at TCF.
i 3
i Containment Bypass and l Large Early Release
- Large early release frequency (LERF): "the frequency of those accidents leading to significant, unmitigated releases from containment in a time frame prior to effective evacuation of the close-in population such that there is a potential for early health effects."(as defined in Regulatory Guide 1.174)
Containment bypass frequency is a main contributor to LERF.
Tube rupture probability is a primary contributor to containment bypass frequency, and is necessary to compute LERF.
I I
l Possible Ways of Calculating S-LERF l for Operational Assessment l 6-LERF = LERF(EOC,.i)- LERF(BOC,.i)
Estimated change in LERF over the upcoming cycle. Assumes that, as part of the operational assessment, tubes have been repaired at the end of the n acycle to enable successful operation over the next cycle, according to NRC/NEI guidance.
S-LERF = LERF(EOC,.i)- LERF(MOC,.i)
Estimated risk accrued from not having a mid-cycle inspection.
i S-LERF = LERF(EOC,.i)- LERF(XOC,.i)
The time during the cycle that can be justified deterministically, such as from a 3*(delta-P) criterion.
S-LERF = LERF(EOC,.i,,,,,,,,,,4) - LERF(EOC,.i, pi,,,,,)
For ARC, to determine the effects of not plugging some defected tubes.
t 4
Possible Ways of Calculating 6-LERF for Condition Monitoring 6-LERF = LERF(EOC,) - LERF(EOC,.i)
- Measures risk accrued from one cycle to the next from progressive degradation.
S-LERF = LERF(EOC,)- LERF(BOC,)
- Measures risk accrued during the n* operating cycle.
- Could be part of the condition monitoring analysis.
4 Risk Assessment Methodology:
One of Two Methods Used
- Review existing risk assessment and identify candidate sequences.
Refine existing results to more realistically treat AFW system, MSSVs, ADVs, and RCP seal LOCAs.
Account for recent plant upgrades and EOP changes.
Carry out a refined human reliability analysis (HRA).
Construct and quantify Expanded Operator Action Trees (EOATs) to tie HRA to the EOPs.
5
l 1
l Risk Assessment Methodology (cont.)
Do thermal hydraulic analyses with MAAP 4 to aid in quantifying Accident Progression Event Trees (APETs) and provide input for thermally-induced tube rupture probability analysis.
l Quantify APETs, link EOATs with APETs, and determine thermal challenge frequency.
Quantify tube rupture probabilities, add to APETs, l and determine change in LERF over an operating cycle, after ARC applied (6-LERF).
Identify potential improvements; do sensitivity studies.
Examples of Top Events on EOATs Recover AFW pump locally Depressurize using ADVs Isolate steam generators SG level indication extended beyond four hours Local control of pressure Level control of SGs out to 1000 min.
Control ADVs to 1040 psi or above
- Cool down re-initiated and RHR established 6
l 1
Examples of Top Events on APETs ADVs remain available for limiting high pressure
=
No stuck-open MSSVs from liquid challenge l No stuck-open MSSVs from steam challenge
- No large RCP seal LOCAs No concurrent loop seal and core barrel clearing High SG pressure in loop where loop seal clears No pressure-induced tube ruptures No thermally-induced tube ruptures I l
Thermal Hydraulle Calculations With MAAP 4 Used for Three Purposes Provide insights into thermal hydraulle behavior and accident timing.
Aid in assigning split fractions to various top events on EOATs and APETs.
Provide thermal hydraulle inputs to tube rupture probability calculation.
I i
7 l
l l
1 Pressure-Induced Tube Rupture Considerations Type and degree of tube degradation.
- from run-time analysis Effects of SGDSM.
- Operator actions.
- i. - maintaining or re-establishing secondary side integrity MSSV reliability during multiple challenge events in a drying-out secondary side.
l l
Thermally-Induced Tube Rupture Considerations Principal thermal-hydraulic factors
- natural convection in RPV, RCS, and SGs (type and strength).
l - RCP seal LOCA potential and size.
- likelihood and consequences ofloop seal clearing.
- reactor vessel vent valves (B&W plants).
- likelihood and consequences of pump bumping (B&W plants).
Principal structural integrity factors
- heat transfer to RCS components (such as hot legs and surge line) and to SG tubes.
l - secondary side integrity, particularly MSSV reliability.
f u
4 8
l
. )
1 i
Secondary Side Integrity Considerations Potential failure to close ADVs when appropriate Inability to isolate a loop when appropriate Potentialleakage through MSIVs l MSSV reliability when challenged either by liquid (in I case of SG overfill) or by steam i
Determination of SG Tube Creep Rupture Probability Extends analyses that already need to be rigorously performed with the MSLB design basis accident, using the same characterization of the condition of the SG tubes.
Recent ANL experiments have demonstrated that creep rupture behavior of damaged tubes at h_Igh l temperatur' es can be quantitatively related to tube burst pressure behavior at nominal temperatures.
- Repeating the current burst pressure probability calculations at various differential pressures provides a way to input plant-specific state of tube damage into creep rupture calculations.
EPRI and BG&E have produced a methodology to do this.
9
Use of SG Tube Rupture Calculations
= Pressure- and temperature-induced tube ruptures will be top events in the APETs.
- Pressure-induced probabilities of burst determined from the "run time" analyses for condition monitoring and operational assessment.
- Thermally-induced creep rupture probabilities determined for each sequence using the EPRl/BG&E methodology.
- Approach similar to NUREG-1570 in that one set of tube rupture probability calculations can apply to a number of sequences judged to behave in a similar fashion.
l l
i I
10 T
. . . .= _ - .- -. . - . . - - . - - - . - . - ~-.
l e 1 1
i 1
Presentation Agenda
- Relo of MRA m 80Tleweiuston Human Reliability Assessment Methodology and Use of . Technical Elements of HRA Process EOATs to support integrated SG Tube Integrity . Ti hemme.e Asseem.m min. - n t-se . e. - .s.i so v.
.no e.
Pepset.eMti SQosM N . v. - .en e, -
Coneeny. eAIC
. TS W
"- es,,me e.e.enno
. exi ease e
- ex 2 Pune twiipene m enes twoush SGs
- Summrypombs 4
HRA Role in SGTI evaluations T1 HRA Queltouw Aamunwnt Ekment
- Rovesis SGD na moonent scess by eened spectos min 0 e Dernie Situation Contend T1e 8'*
- Identify Potengel Human actions to model Tib
. Systemshesy renne emehne MWHRA to e.iste WeeCi cf i l
ces,etor echons on CDF, LERF, and TCF to suppen SGTI
- Apply Qualitatin Screening Tic ousnetone nok - _
Detailed Qualitative AssessmentTid I
. m-.e, e v ase,s . w e ser i
son
. os e na,eductone non. an ein e-.
ennes asetamen e swees waamnea 4 Define Situation Context Tia Ex 1 Define Situation Context Ann. Orgentes siformaton needed to ested and eierefy himon . . airvrm aos arver pr@ * '88** ****' Tepw in.se a se a swa ahmuss
. Rever Cw.e4 Pe4 Assedet gegense er e4 Bel Geodeness . % n.
- e nt eusse sur hwien essons andener medes to 34AT>GANA.
- A posess . er asennes e auseed,
.e> .e . - e e
. om.n vn ogs .emmc e e
, ec ease ein, ann : =, ~,,,,-*g,,,- :;; :,---~~~
. men.9.se ese ,=y
.
- L8* lB .-- . eem esseg atek ee,dy ...,,s.,.,,..,,,,,,,,,,.,,.m.,,,,,,,,,,,
. tease e m ammenenenes a.a.emase. ems %ummeer
- W M had.se W m heen eiese esse WW em W M
- ones masviness aguasy asusseses e enseg iPeMit esisnes arm sine.e s mMoss e,esamed 4 1
, w -. + ,- ,--- , . , - - .- - -
v Ex 2 Defining Ac6on in Accident Context Select key Human Actions Task ib
- Ass.epsons er 96w ones sceneh tes
. e e,s. . - w e=s 4,,i..essus no.si. e.n. nay n n m
. Tee e.,s ..e se w.see ..es -ca***=
- so -
..es--
.s.
.a ., e - ,.e see.-u tesi es.
.s,e=*= emwe me -
e
- Delormme f e.stig stJdy has ineryn for Operater ottens
. esusese8 e coses.a... oe= e -
- Tehey ses stee.n toorm.seesses n,sid eme ise esse
--een) een pensely seen denue.s
. me..e. me.see. . . .wasa r ner.y -
e en.
to ne.ss. e.n.es. . = = - =
_ e,. sm e _ -- e
- EX 1 Existing Analysis poes.esamer mese.ory assens seemed m IPes .e ese e inital Coro dems0s frequency
! = Loss orenese poser (Seense inesang . Canaderstesterest sowenes
! Heren e and nonsenenes towns or tunes, tegenos, chsgers.
- . Does nei maiude pesonnel ads.mel reemeses ese.g esseng .ei sweners. and Geest sene olors mwentry see.ned es,n pwnp tiespag l
- AshsenenlaumeWdeedgenero."
- Thermal challenge
- Co.n.n mes <- emi e=eni. , ,,,,,,,,,,n,,,,,,, ,,,,n,,,,,n ,y. ,,
l so,*. emes ==m m yaeses.e*=aene.y nasy. e ,n,,,,
. Esesen.sy,.w d t, e.,mese .es ytne -g",,*,,',,*,,*,""*******"'""Y""'
. Co. woe.,imeneswme.esi s a Potonnel Margin in suidiiery feedwater recovery
.e esesem r
. an.s e,ar .e =nas sie.e. esnea.
I .e.se .e e mne - .
L l
Apply Qualitative Screening Rules and Ex 1 Potenhal Recovery Actions in PWR 880 Assumptions Task 1C (high dry events)
- Asm.Swreidy Assensment a restore aussary feedneter (turtime pung) .neh .sier supply.
reher on ADVs er Safety welwes
. Renee ses.no ses.npenas Her.1 A screas som PmA (t esoneeze le.ss men 1EJ . reshan a e emner eessis nor ensam ennemier mycoon
. Pnenese accens Dy sese.e g sees . Ramass esscene seds )
CDP.LORP.memsf
- restert eseel pen.shre(eteestcod bened cunes)
. emer ne seen. hen,sesen a eenes:
- '"'"' P*"w motsecal band ames)
. t., es.e,.nore - .e,,o, s e,s -
- - - o nec) -
. Denne dnesind _ _ ;iesues b'**d cenhnetconsid roosai) e feed hgh prenewe inpocharftenbel reem)
.~ siend edih human ecsons streedy Desied ti IPEe
_ sm s_ .ese.s - .
2
1 l
l Detailed Qualitative Assessment Task id 1 m A h h Aux h y M m W j system failures i Ann owocwee esamed tusen actens M estas
,,,,,,._.,,.,, wa-. e-* e e
- =l; =:===:=
ll::ll: == l,.y_.=: .
- . . -~
g.':;;,_" ;;; _
. For each eesswnst todwo enose seneider esses of
- _ ==-y = ,- _=. g llllllll l::::-~'--'
. ,0 -
- f- ll=". llll::=ll;.".,;"* ==t.".:llll*.=l"
. llllll:: :::--- 1l: :::. ,
.e: e,, o m
.== ==.m,*- . mm Date on Performance Shaping Factors Example Performance Shaping Factors b Error forcing situations
. Denne PSF categories (e. g., ATHEANA, Swain *"-* ny =e.*
- ==
(NUREGICR-1278, or 4772)) , ; p_.=.- .-===
. Training and operemons inteMews , p---*===
- Plant visit to essess PSF in6uence l lll':"*"l .,, ..
- can me me rnaar : :=.- .
- = prac=*a =*c=at '.~ . ._
. con twee uw.w.no res==d..e wt :=
- e . - e, - .
i i
Development of EOATs T2 Aen . Construct EOATs tredne operator accens (and two. ore) to twndge bot.een esehn0 PE CoF and LERF to tierrred I chosenge pequeicy
. Shome lope of hoy schens required M moenartes
. Pionens quetteuwe brems of EOPs and seier pocedures to acomort esquences
- Drining of esquence _ __
. Dipuis in APETs, Event Wees, or systern Fault toes
.in.e er.mme n se 1
/
i l 3 o
I
e Ex 1 EOAT mid recovery l Ex 1 EOAT Early Recovery
= - ;- e=a?g,*s--T+ +
~
j w it?? _-P- 5--- was -
M ; _ l 1
_ .1 _ l _ _ . ,
_ _ - ~y
. mamar _..
4__ _ . .. ,
, asr= . --
. . -m -. .
-- ~ - -
- H_- :::: _._ _
g: = _
i
-. m_._. . _ i n_ ._ __.. -
- - . e.e , .
Ex 2 EOAT for Pump bumping Ex 2 Use EOPs to build EOATs l w -
w - - g ,
- - ._ _. _ in =-- .
-.- .:::::::r.L..a.. ,.- -. .,
I _.a..s. _
-~y,.r._,,_,,,., .- :: . =_ =_ _
- i EEE
... =_ -
- 5" - ---
.:= =
s'E
- Mi.._
E_ E_i_
-4.'.=. . --
=rm - ::r. - r -c:: i E_ . . E. 5_ 7 m :-
- == =
Quantitative Assessment T3 Genen.c HEP quanbfication a-.P e a P. r_ . som .,,,,,,,,,s,.,,,,,,,,,,,,,, ~ ' , ,
- h Scenes (Gelect tenuteng er emesses esosiing -mamanammme ensem% shmeen essansen ogsammus,t, e. esenee,a,m. papeggeasemean,gW gaan geen tir HEPs) 8I
- Queensseen HE
- Pi*
- P1"
- PI ' P8
. Dummaseel ammesn p. g, uns esas siese sie
- Pr W
- Moment poner aman,ay, tas, sop shey, minamarmamanaq 3R06 sak)
- Reaswery sudyes. Asen twismu sulyuln esupsus age 1H
- Omensme per tyna tuhas Deutsume tes Omenho P1, pg, a ps uses **""
- bum,emse resuas too tiennai enen,,,en,,wis, -**==.a==am=ea.= -am. .a 4 985.esGenftte .**94
- m - - --.- -
- arm.one.,m. m 4
- . . . . - _ _- . - . - - ~ _ . . . - - . .- . . - .
t 1 1
1 .
l.
l Timeline elements interaction with :- .- .. . .
l MAAP T I i.I \ -adm j
w is s w
'm -
v
- i I
I' - _-
t ._-- ,
- I
.. . l l
. == . . . . . . .
1
. .,-. a i .ma.sm, .
l i
i l
i l
1 1
EOAT Ex 1 Case 1 EOAT Ex 1 Case 2 l
~ ~ =.=, =,,*,,=;ll=,:=, = =
., - .m .===-*==_
l __-= t _ _ .n== ==.- - -
=.-.=.g . .
n,. -. mas l
- ,L, .- . .
.,--- =
.,.e.. ,.
~~"" ~ ===e ,. ~
, :::, =_=~ ,. *"""
t t
- -, =_== _. =- a==
---.h na gM
~
i l :=: ! I iE!
x..
- : m : :
A M.. : :
e
. :=.:.
m
!s: . .::::
- === ,_ _ .
1 I
1 Ex 2 change in sequence timing Ex 2 EOAT Cases
- 9. 9 tel .e.e esse e.e i
es.es.80 e M
.e
_.- - _- :=r =
. - .e .e
. . . .. , ... ., m .
32 888 eee
- e .ese ..
.se 4.39 ..34, . =
ese e he e .M ,,, ,,
I.9W .m.GR has .B a
5 l'
I
Sensitivity & Uncertainty T4 Ex 2 Sensitivity of Cases Aim - Evoluete sensitMty of pseult to variations -
In ensumptons and values used for HEPs & l3lElll' l".*
9enerate insights m e, "..".3%ll'" ,T.e.
+ SensitMty seeeeements look at ellect of " " " ' ~ ~ ' "
different assumphons on overal results , ; , , = ,;-=
+ Each HEP input has uncertainty eseccieted =====a'= **=
_ **=^
wth it ,,,,,,,,,, _ , . ,, ,
lllll '*'."""e'""
, T."* O '"* 2".
~ , - ~~ =
I Ex 2 UncertainW Ex 2 HRA Uncertainty Intemal events Quantify basiclogic models
- Denne reigne . essensens ter M P,eassesse
- Typioni assuns te feu.e to ,sesvar ensemanor saar Weel tot.e
.%,.,,, e. .
- c. ,
. o.s,
.u .
---e .. e
. e - . es e e t. es I .. e
~
i Document HRA T5 Insights Ex 1 e Case 1 representation of current "Q",e "w * " * * *d situation indicates TC exceeds 104Nr.
. os ,,.e,
- D'"**"**
e eunenenas Aeneste oogenene D Mnd bry life by load shedding
. s =nesse emans.o. emen one se PRA teoAn ser esom en a Permit temporary hookup of level indicaton
.egnav eemenn esonnese a
Pedes =mses ser senereis ens esamens _ ,
e e.nmeias W.ieres sanos was e.mne m 6
l l
\
l l
l
l l
-e ;
i I
I I Insights Ex 2 l TCF can be a ston0 funcien of procedure oestegy
- wei no essen owe a ne some ehesense, but omre l denses heter
- we na p. nase tunoed euas paar e ese dennes) poseen
- het one pom to 90 eessed
. w tesesses.Essesand who tesen toevyr,
= Hoheycorshausseehmese%
- we eso of too pumpe bungled no msnmei ehenerge unsees I ener by operaser
- W send emkBMmeer
- NWiey CDF desseen abad 17%
(
some pusencame se i
l l
l t
I i
f r
l O
l l
t i
d 7
6 t
Key points i
- Identify plant specific key actions that change TC, CD and LERF values
- Clarify emergency procedure and training issues for protection of SGTI
- Identify effect of operator actions on SGTI uncertainties and sensitivities
- Illustrate SGTI margins available on risk measures from operator actions
- Identify effect of vendor specific issues (B&W pump bumping) 10/28/98 EPRI.SGTI.gwh.saic 1
. .l
E Hi THERMAL HYDRAULIC ANALYSES TO SUPPORT DEVELOPMENT OF EPRI SGTI METHODOLOGY i i
Presentation to NRC
. October 1998 i
Marc Kenton i Creare Inc.
l Work Performed on Behalf of EPRI
- I MTG-98-10-958 4
t
Goals of T/H Analyses '
.. HI
! E Qualitative .
(1) Obtain phenomenologicalinsights (2) Elucidate sequence timing to support HRA and APET ,
(3) Assess effect of design differences on phenomena and results E Quantitative and Semi-Quantitative (1) Assess whether sequences contribute to
" thermal challenge" i
(2) Supply temperatures for explicit tube rupture probability calculation C MTG-98-10-958
Analysis of Tube Rupture Probability EIDi i
E All but our most recent analyses focussed on ,
assessing whether tubes are " challenged" in a given sequence.
N Most recent (Calvert Cliffs) analyses include a quantitative calculation of probability of tube rupture.
4
{Q MTG-98-10-958
In this Program, l Four Sets of Calculations Performed
~
llll E Methodology Development (" thermal challenge'?)<
(1) Westinghouse-comprehensive (2) B & W - assess phenomenological differences due to geometry (3) CE - assess phenomenological differences due to ,
geometry E Plant - specific evaluation of tube rupture probability for Calvert Cliffs.
eeare .,e... 1,.,,,
Three Types of Sequences Make up Vast Majority of High/ Dry Frequency in the Westinghouse Plant HHI ;
, E Fast SBOs. .
i E SBOs with "Mid" Core Damage Timing: Turbine-driven AFW available initially.
E Transient loss of feedwater with failure of feed and bleed.
E Aim: What fraction of these result in thermal challenge?
MTG-98-10-958
t
- Focus of MAAP Calculations for Westinghouse ;
Plant was on "New" Types of Sequences EIEl
~
~
E " Usual" SBO sequences with early core melt givie results highly consistent with previous industry .- l' analyses and with NRC calculations. ;
E Loop seal clearing.
E Mid SBOs. -
1~
a
@ % re y10 9810 938
I l
Key Results for Early High/ Dry Sequences m Westinghouse Plant without Loop Seal C1 aring
. lll o
E High/ Dry High-negligible thermal challenge (Cases la,1b). .
E High/ Dry / Low - modest thermal challenge based on At ,, ~500 secs (cases 1c, Id).
E Cases with stuck MSSV and PZR/PSV depend on timing and size of PSV failure - concluded conservative and best not to credit stuck PSVs in the SGTI methodology (Cases Ifx).
(
% {Q MTG-98-10-958 l
l Sununary of Westinghouse Plant Calculations with Stuck PSVs and MSSVs 5 lll Fraction of full Sequence of Thermal flow (percent) .
Events Challenge?
100 accumulator discharge, No repressurization, relocation, hotleg rupture 75 relocation and Yes accumulator discharge i simultaneous 50 relocacon No then surge line rupture 25 similar to no-stuck Yes PSV base case, ruptures occur first '
MTG-98-10-958 t
. - - - -- e v. > v ------
Loop Seal Clearing IBI
- M If large cold leg seal LOCAs develop, detailed code results (SCDAP/RELAPS) indicate one crossover leg can clear.
E Clearing one crossover leg eliminates differential pressure between core and cold leg and presumably prevents other loop seals from clearing.
E Can't model the clearing process with MAAP, but can force it to occur and see what happens.
h%S MTG-98-10-958 9 9
Do Loop Seals Really Clear in these Sequenpes?
. HI i
E Experimental evidence compromised by scaling !
considerations and not modeling sequences of interest to us here. ;
E Available detailed code results compromised by numerical problems, perhaps bad models, and seal LOCAs too large.
i i
E If a sufficiently large bypass flow area exists through core barrel ("T-cold" plants (?), B&W) shouldn't occur. i E Non-SBO: If operators " bump" RCPs, will occur (will they if steam generators dry?).
E Conclusion - For SBOs we don't know, so have to treat. I C re MTG-98-10-958
4 i
k l
l
' Collapsed Ilquidlevel -
Collapsed liquid levet Elevation (m) distribution just prior distribution just after to intact loop seat clearing -
intact loop seat clearing 14 -
m '
l 13 -
I 12 -
7
- ./ O
!. 11 -
\ / s
- g /
{ 10 - -
4 l 9 -
l 8 - '
l 7 -
l t
1 6 -
a l 5 -
'. 4 - O .
. m
{ 2 -
Bypass
" L $
F 4 4
- g-g
- gypgs, -
, line
~ line a I. E =
$ l\k
, hf '} \
0 - {
N i ,-
.t -
Break Break
')
2 -
3 4
Broken loop p intact loop Broken loop D
Intact loop w
.g -
6- -
as4:e Figure 12. Ittustratloa of the system liquid mass dis.
F1gure13. Illustracion of the system tiquid mass dis.
tribution just prior to intact loop seat tribdtton just arter intact ioop scal clearing -
clearing (5% break S-LT8).
,(55 tneak S-UT4).
O
If One Loop Seal Clears, Does it Matter?
w--wwmewawa ameww a -
as w e ., y ; lll EIf the base of the core barrel also uncovers, then and only then will a complete path 'for unidirectional flow develop in one loop.
I i
E This is bad for two reasons: ,
(1) Eliminates inlet plenum mixing phenomenon which wpuld otherwise mitigate tube temperatures relative to hot leg.
(2) Flow rates, energy and convection from upper plenuni to affected S/G much larger.
MTG-98-10-958
I l Case 1E: Fast SBO,180 GPM SL, Clear LS, Stuck MSSV ,
eme#%wenewe6*vwM;/4;#9tM6r;#,55 iw 4WW Sep --
$t l a t
~lw$l 25 ,- - - -
'N --
B Interm. Leg 20 - -
h 2-P h
- '\.)
- i. i i
2 15 - ! -
i t. .
E I :;
a .
e !
lii i 3: 10 -
j -
ds
\\
5 - - --
i'.-
V
~ .
1
' ' ' ' ' ' ' ' ' ' *^
0 - - - - " " - ^
4 0 0.5x104 1.0x10 1.5x104 2.0x104 t i
Time !
I t MTG-98-10-958 i
._- . . . _ . . - . . - . . . - . . _ - . . . - - . - . _ . . . - . ~ . - - . . . . - - - . . -
15.0 ---
l .
a F
1 i
10.0 -
d l
Top of fuet -
~
5.0 -
Bottom of fust i
- . .. .. ,j
\
l .
0.0 ' ' ' ' - '
O.0 5000.0 10000.0 15000.0 20000.0
.- _,. Time (s) .
Figure 3.13 Co' lapsed liquid level in the reactor vessel: Case 8R I
- 20.0 . .
' i 15.0 d
I l
- 1 10.0 8-I -
e i j -
5.0 L accumulator -
1 Ir$Ntkxl Qne -
0.0 ' -
1 0.0 5000.0 10000.0 15000.0 20000.0 Time (s)
Figure 3.14'fressurein'hereactor'vessellowerhead:
t Case 8R -
Insights on Loop Seal Clearing from j MAAP Calculations 1
- - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~* -
- E *
- Iil E Core barrel uncovering seems very unlikely (caveat: -
contradicts SCDAP/RELAP5).
E If assume it does and if loop seal clearing happens in a loop not containing a depressurized S/G, results actually better:
hot leg temperatures more uniform and so it fails long before i steam generator tubes.
B If loop seal clearing happens in a loop with a depressurized secondary, even pristine tubes calculated to fail. ,
MTG-98-10-958 G e
Case 1F: Fast SBO,180 GPM.SL, Clear 2 LSS
- me*utca +a,r=$www$$te* +4;We rryt.+46 !%9fster, ou gga ey i
ein. i@ u $
t .. hl i
80 . - - - -
.k . '-
! l *w !
j\ --- B-loop hot leg j i
- U Cntr-curflow(perloop) !
i
\N -
i
~
g 15 l % '
$' .f .
s <
l .b s
-i 3
0#
(N t
'b. ::--l'
!$ 1 K[v?T"/,y./J-1 k,.1 r
i i
=
4 - . . . ' . . . .
4 4 1x10 2x10 Time SECOND CGreare MTG-98-10-958 i
( -
. ! ;!! Il!; !itI\ !l:lt- )1l ;[ ,i t !!l:!!l
- 8 S 93 l
5 9-L Sg 0
1 S 8
- L -~
4 0
1 x
9-G T
M 2 /
- g. - - .
0 M
- - 9 1
Prg
- a Ge d
- ~
t e
c l .
de 0
Cg
. 8 _
r 0 8 P _
. 1 e . x _
. 1 - - t r
u 5
8
, -- p . 1
.. u .
R O
. g .
e
~ -
B
~
~ L .
t o D N
S H
. 4 0 O
_ C
.
- 1 t *
. x E 0 S s -
- 8 a
- . 1 e
[
. m
_ F 321 i eee . T ddd mp ooo NNN
/
- w F
3 - .
4
+ -
0
. p . - 1 1 :.
5,,
5 7
x
_ e o r.
. 1 s s a
r e
C
/
e -
- e .
h .
4 0
1 x
%0 0 3
0 0
2 0
0 0
0 0
0 0
7 w1 1 1 0 9 1 1 1 i
eYr T8ao t
Mc O
t a
w M
r G
- O
- T 5
1 8
5 9-L 0 1
8
- S lj
(
4 0
1 9-G
- . - x T
$' 0 M t ._
4 M:
- 1
/,
., ~
G g. g -
}/ 'y/e
^
1 254
/'j / / e r
u 4
0 1
x
/.f t
/ . p '
5 u 3
, p - ,
R de f'f' /.'//
' 1 O 'p gt
/' ec Lde t
or B -
<e** . ,/
/'s/ '/ ,
, P '
SQ t :
. ee do dode
.s '
4 0
1 0
x D
N O
C E
S NNoN 3 s4- -
dr dns 's /'s e a t . ' 1 m
- s,.
321 i T
F
~
['-
q -
~ .
- g - - -
/ /
4 0
A%
1
~
5 x
2 1 44 -
' 1 ems.
.f.'
s -
a% ~.-
. d'/'
C4g
. 4 m . 0
- - I
. 1 0
0 x
0 0 0 g 0 3
1 0
2 1
0 1
0 0 M2 B 1 e 1 1 g g@*E ?4x m N=
w n
1 v
e w
~ _
n+
1 Effect of Assumptions on Likelihood of S/G Tube Rupture in Westinghouse Plant
~ ~ ~ . - w. - == . ., e w w
- 41 w ll Seauence Difference between hotleg and ;
S/G rupture time (sec)
RCS high pressure, '
N/A S/G high RCS high, S/ Glow 490 RCS high, S/G high, 3000 clear 2 loop seals RCS high, S/G low, 2300 clear 2 loop seals in unaffected loop RCS high, S/G low, -2300 (-200*)
clear 2 loop seals in affected loop
- pristine tubes .
I g MTG-98-10-958
e Mid Core Damage Sequences
.-. - m - .. ,, - 2 .* '
EIBI EHigher CDF than early SBOs.
E Used MAAP mainly to get insights into operator actions rather than physical phenomena. j E
Conclusion:
operator actions drive the problem.!
t r-
. j m
I
, . - . - - - - - . . - - - - - . . . - .- . .-.. - . ._. .--.---.-- ._n T
e
- Early Core Damage Sequence Cases Case Descr;sailon j la Base case with no sealLOCAs
- lb Base case with 4xl80 GPM sealLOCAs i Ic- No sealLOCAs with stuck MSSV l Id SealLOCAswith stuckMSSV j lo Seal LOCAs, stuck MSSV, and loop seal cleadng .
, 10 Seal LOCAs, loop seat clearing, forced core bansl uncovering j 10 Seal LOCAs, loop seal clearing, forced core barret uncovering, stuck MSSV 4
inunaffectedloops i If Seal LOCAs, loop seal clearing, forced core barret uncovering, stuck MSSV
j afrectedloop, pristine tubes til Original MAAP treatnent of axial '+.i-w in hot leg
! 113 Original MAAP treatment of axial temperatures in hot leg, no rupture of '
carbon steelnozzle i 114 25 percent tManad tubes
- 115 Pristine tubes ,
i lh MSIV leaks at 2 kg/sec at nominal pressure j lj . 100 spm nominallenkage through steam generators
! Ik Noleakage through steam generators l
. Il1 Stuck MSSV and fully stuck-open PSV l 1l2 Stuck MSSV and 75 percent stuck-open PSV 1l1 Stuck MSSV and 50 percent stuck open PSV l
! Il3 Stuck MSSV and 25 percent stuck-open PSV
< Mid Core Damage Sequence Cases
] Case Descr5 tion 4 2a Base case with no sealLOCAs i 2b Base case with sealLOCAs i 2c No seal LOCAs, no manual ' control of ADVs 2d SealLOCAs,no manualcontrolofADVs i 2el No seal LOCAs, ADVs manually controlled,125% nominal AFW flow i 2e2 No seal LOCAs, ADVs manually controlled,75% nominal AFW flow
- ~
2f No seal LOCAs, no manual control of ADVs, MSSVs don't stick open 23 No seal LOCAs, no manual control of ADVs, AFW falls early on overfil!
- ' 2h ADVs controlled manually, AFW falls to run at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> 2hl ADVs controlled manually, AFW ikils to run at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, stuck MSSV
! 2g AFW falls early, ADVs manually controlled
] 2m Seal LOCAs, AFW fa!!s early, ADVs manually controlled 4 2n Seal LOCAs, AFW falls early, ADVs controlled, loop seals cleared l 3a No cooldown base case
{ 3b No cooldown, lose AFW and MSSV sticks open at 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> Table B-1 List of MAAP calculations of Early and Mid Core Damage sequences 4 .
I k
b~
4 1
i
!- B-2 i
1 l Mid SBO Base Case: Assumed
! Sequence of Events e Iil l
l E Lose all power at time zero except for DC, instrument AC.
I E Turbine-driven AFW starts, steam supplied by 2 S/Gs.
l E Operators cooldown using ADVs to ~260 psia.
t
! E Seal LOCAs may occur (P=15%).
l E When lose battery, feed reg. valves fail as-is; operators re-l open fail-closed ADVs per procedure to keep S/G pressure
! constant.
4
- E S/Gs may fill as decay heat drops or if flow mismatch.
1 i j E Ingest water into Terry turbine?
i l MTG-98-10-958 i
Mid SBO: Issues i '
HI E If operators do subsequently control S/G pressure constant and AFW is reasonably well matched when batteries lost, sequence is simple and benign through mission time (24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />).
E Otherwise, behavior depends on details of:
(1) What operators do or fail to do.
(2) How AFW flow varies with S/G pressure.
i (3) What happens to AFW and MSSVs if S/Gs overfill.
i re MTG-98-10-958
Case 2C: Slow SBO, No SL, CD/AFW .
Const at 4 Advs Close MMMbEE Hi 20~ ,
i
/l . - ~ -- ~~~~'*\
U S/G
/, ...._ g g,9 I i
. /. - . :
./* \. I ,
2' " \
.- .r*~~~ \
bs \.%. }.
t f
u
@ 10
- \. i -
8- \. i c \. : ,
i 3 )- :
o O
\ i
'g : '
- I 5 -
i i
l.
. \ i i
\.
i
\
i I
\' i O
0 0.4x105 0.8x10s 1.2x10 5 i MTG-98-10-958 .
l Case 2C: Slow SBO, No SL, CD/AFW
- Const at 4, Advs Close Hill t 2.0x1'07 - -
i
- N i 1.5x10 7 - -
! Is -
i i !
! i i i
1 ,
\ ;
e : I
- a. : : :
I 7 I f -
U S/G y ,1.0x10 i ! ---
B S/G y ; j . ....
acs i
! 4 !
- 1 r
t 8
il p .
l O.5x107 /
J;' 4 f:
\.\ !
I -
A,. i
.\ . ..y
! \~~
e_ _ - _ _s
/ \
t ~~..... , \
%.:: ,1._N--
. i* -
0 -
8 0.8x10 8 0 0.4x10 1.2x10s Time !
l
@ MTG-98-10-958 l
l t
4
Case 2B: Slow SBO, SL, CD/AFW and Press. Constat 4 t
~
BI i
1.8x16 . - -
l U S/G ;
B S/G !
RCS 1.2x107 t ca : !
- n. 1 !
E i 8-
'i . t 2 i f
- n. :
i !
0.6x107 I -
. i !
l'
\.
. k '., .
6 l
\ ' .N---e- .-~r--- __ . *%,
s'; 4...._
i
-- _---~ . . .. ._. ..m.m m .* _ __ ___
. f t
0 O 1.5x10 4 3.0x10 4 4'.5x10 4 6.0x104 7.5x104
{ '
Time MTG-98-10-958 i
i l
Case 2A: Slow SBO, No SL, CD/AFW and Press. Const at 4
( ]l 20 ' - . .
i I
U S/G l B S/G
\ !
15 -
x ,
'\.
2 d \ i
- r-N. [
E l \. !
8 \ !
a N.
h, 10 - -
g g~ \ t 3
O
\+
f l
I o \
\. :
5 -
\ ~
'\
'\
. \.
)
\. t
. , , . \- . .
0 5 0 0.5x10s 1.0x10 5 1.5x10 !
r Tima !
I r MTG-98-10-958 i
l t
6
Conclusions:
Sequences in the Westinghouse-Plant that Contribute to Thermal Challenge
- HI E Early SBO:
(1) If no loop seal clearing, need one or more stuck MSSVs to challenge tubes.
(2) Cases with seal LOCAs, loop seal clearing and core barrel .,
uncovertig require a depressurized steam generator in affected loop (b* t then even pristine tubes may fail). l' E Mid SBO:
If operators don't re-open ADVs:
V (1) Cases where lose S/G inventory prior to core melt with a depressurized secondary; (2) Predicting the likelihood and timing of such sequences is difficult, especially lacking a very detailed AFW model.
E If operators do re-open ADVs, should maintain AFW long enough to recover.
e@ TOGO MTG-98-10-958
i l 6 B & W Analyses: Background
. : Ill E In Westinghouse plants, MAAP calculations were performed primarily to elucidate sequence -
specific and plant - specific details; the overall~
behavior was already understood.
E In B & W plants, much smaller plant-to-plant differences exist and little work done previously, so MAAP work primarily focused on understanding what general features were necessary for thermal challenge to occur.
- eeoe m e.. .,e. ,.
Effect of Hot Leg Geometry on Natural Circulation When Cold Legs Blocked-ggll E Unlike Westinghouse and CE plants, steam generators are not the highest point in the RCS. >
E Based on natural convection experiments run in!
vertical tubes for other issues, expect on general [
grounds to see:
(1) Fairly vigorous, stratified flow in horizontal part of hot leg (2) Unorganized, weak, turbulent flow in vertical l portion of hot leg (3) No flow into tubes from top of candy cane .
E Westinghouse one-ninth scale model results are ,
completely consistent with these expectations. l C re u ro.gs.1o.gss i
l 1
I B & W NSSS i
? .
E H!
I
)
I
{ - Vent line Relief valve nozzles !
we co,e ,
D G g [ bay h D, 3 T !
p Inject nozzle Pressurizer Temperature !
sensor %. M8 f) g line i 1 I
~
s &
I .
g g
,g _.- YMh cx ' Pump #
P18 8 a
/
["'^
i
% ===5V ', -
Temperature !
[ -
sensor :
(
Decay heat . .
gggg - " nouie
\ einiectionnouse H i
. I ;
M Q MY W :li ne.ctor vessel Q~ v\ .etoown t.oop A t outlet nouie i gp <U, nP etion nonie {Q MTG-98-10-958
i
- Average Temperatures (oC) and Heat Flux Densities on the Pipe Walf of Hot Leg Model With 750 Watts Heating Ill t
I O 1 t
.. ., x .
N g ....
,,, ., x .
/ !
{
/ ;
.... .a s # :
. .... i e.m
- N/
, i e.m ud --.
l l I g ( :e::, t :: /
) !
F MTG-98-10-958
.. - _ _ _ _ _ _ _ _ - _ - _ _ _ _ - - - _ - _ - - . - - - - - - - - - - __ 1
Implications of B & W Geometry on ' Thermal Challenge
~ ;- . Elli E Vigorous natural circulation in horizontal hot leg should lead to induced hot leg rupture, as in Westinghouse and CE plants.
E No counter-current natural circulation flow into tubes implies no thermal challenge in any sequence where whole-loop circulation does not occur (cold
- legs remain blocked).
i M For cases with seal LOCAs, flow goes from core directly to cold legs through RVVVs; could lead to cold leg rupture if hot legs don't fail first.
% I6 MTG-98-10-958
[ Conclusions on High Pressure Seyere 'l Accident Sequences with Blocked Cold Legs WElli E No thermal challenge possible. - E Pressure induced rupture relatively more likely in the plant studied because of absence of MSIVs and , depressurized steam generators. E Hot legs, surge line, and perhaps even cold legs will rupture, depressurizing the RCS. MTG-98-10-958
Possibility of Cleared Loop Seals in B & W Plants EI81 E B & W plants have either " raised-loop" or
" lowered loop" configuration.
E RVVVs open at <2,000 Pa, smaller than the static head of water in the cold leg of a raised loop plant (~9,000 Pa). E An open RVVV will relieve the pressure drop across the cold leg water seal, so we don't expect loop seal clearing due to cold leg LOCAs to occur as has been predicted in Westinghouse plants. E However, the operators can elect to " bump" an , RCP, which may empty the associated cold legs. MTG-98-10-958
$I ,
l Effect of " Bumping" R, cps M EINII l i E If operators bump an RCP in one loop, and if this empties the cold legs in that loop (not clear in the ! case of lowered loop plants), nothing much should i happen: other loop and base of core still blocked. i E Above situation is analogous to case of one spontaneous loop seal clearing in a Westinghouse plant. E If an RCP in the other loop is bumped and that loop also empties, expect whole-loop circulation. MTG-98-10-958
What Happens if Both Loops Cleared by Bumping RCPs? llll E Flow goes out from upper plenum into one loop, returns to upper plenum from the other. E Note that flow does not directly enter core initially; instead:
- a. Core - upper plenum natural convection cools
- core by heating upper plenum structures.
- b. Whole-loop flow transfers heat from upper plenum to loops E Very strong whole-loop flows give rise to:
- a. Delays in vessel failure because of indirect l' core cooling to structure.
- b. Nearly equal temperatures in steam generator
- @ % re ura.gs.1o.938
Conclusions on B & W Plants KEHI E Expect no thermal challenge unless whole-loop F circulation begins. k E Unless RCPs bumped, whole-loop circulation seems to be nearly impossible in a lowered loop i plant and quite unlikely in a raised loop plant. E If RCPs are bumped and both loops clear, will have thermal challenge, even of pristine tubes, unless the PORVs are used to depressurize the RCS below the SG pressure fast enough. E Whether thermal challenge leads to tube rupture is not yet known-B & W and CE hot leg steel is less. thermally resistant than that modeled in MAAP. C re x10 981o.958
e d Effect of Hot Leg Materials llll E Westinghouse plants use stainless steel for hot legs, ASTM A533 carbon-moly steel for nozzles. E Most MAAP analyses of creep rupture in Westinghouse plants use Larson-Miller correlations for nozzle material. E Correlations for A508 Class 2 pressure vessel steel give results similar to those obtained for nozzle material and are often used in SCDAP/RELAP 5 calculations. E B & W and CE plants use A516 Gr 70 carbon steel ' which is less resistant to high temperatures. MTG-98-10-958
,ll. l s!l i j !i>
8 _ 5 _ 9-rl; g 0 o e 8 1 f tg e- 9-G T rS ! M ino 5 vo ab
.~ ' -
2 1 1 hr ea l t e a n at s BC P i t sd r l e eo 0 N ;~, e eBMW-l
. ~ 0 t
e e .
~-
~
1 S t ste ee 1 ed r n o Vlaaa PPPP t l t l
N q n br '
R SSS t u a C t a o P TTT A UUU Dhhh
~
~
Ciiiggg 5 pl n o . SHHH * ' 7 uee 0 K i . t l a 1 e r - e r RtS C r o t u a r e
~
P p p A m eo-eM r D C S M n a e N . N ' 0 5 1 0 e T C-C f o n o s - f i a r 5 o
' 2 p 0 1
m o
.N n .
C N N o . i s x~ r 0 0 r 4 0 3 0 1 0 8 0 1 a 1 1 6 0 0' 1 1 1 p 8 i5$ m o - C a
CE Plant Phenomenological Analyses EII J E Prunarily directed toward understanding the cause of significantly higher calculated tube temperatures than is predicted for Westinghouse 1 designs. I i i re MTG-98-10-958 i f
t What Causes Higher SG Temperatures at Time of Hot Leg Ruptyre? M HI i E Much higher system heat-up rates (hot leg rupture follows Zr-oxidation). r E Smaller steam generator to hot leg flow ratio (less difference between hot leg and steam generator temperature, all else being equal). E Less inlet plenum mixing. i
%S MTG-98-10-958 i
I - 2 Ratio of Steam Generator to Hot Leg Flow is 1 Lower in CE Compared to Westinghouse EEll i E In Westinghouse 4-loop MAAP calculations, ratio drops from 4.5 to'2.2 in time frame ofinterest. E In CE, see a drop from 2.8 to 1.7, i.e. ratio is about 70% of value in Westinghouse plant.
- E This behavior is consistent with a hand calculation
- that suggests that the ratio should vary roughly as
, .66 I D 3 1.66
- F=N o r
! g Da,
- E For two plants we analyzed
F '" ~ 0.75 F,o ! lV MTG-98-10-958 . t
Ratio of Steam Generator to Hot Leg Flow is Lower in CE Compared to Westinghouse (cont.) REll
~
~
l E This formula also provides a way to numerically alter a CE steam generator to get the Westinghouse-like flow ratio. E If this is done, indeed see same flow ratio and no thermal challenge. E Conclusions ,
- a. MAAP predictions of flow ratio are consistenE with a first-principles hand calculation.
- b. The differences in geometry that lead to reductions in steam generator / hot leg flow ratio appear to explain why thermal challenge is more likely in CE plants.
% re mfg-98-10-958
i ; i Conclusions
.- lll
, E In Westinghouse plants, MAAP-guided assessment of thermal challenge is consistent with published NRC work with ; exception ofloop seal clearing issue. E MAAP analyses suggest that assuming a cold leg loop seal clears, nothing is likely to happen because base of core barrel ; remains covered until after hot leg fails. i E On fundamental grounds, B & W plants appear to be immune from thermal challenge unless RCPs are bumped-and this empties both loops. E CE plants are qualitatively similar to Westinghouse plants, but geometry differences lead to higher tube temperatures. E However, CE hot leg material is less creep-resistant than that ; used in Westinghouse plants, so need a more detailed analysis i to assess threat to tubes.
@rOOre MTG-98-10-958 i
)
i i Accident Progression Event Trees (APETs) Edward L. Fuller Polestar Applied Technology One First Street, Suite 4 Los Altos, CA 94022 650-948-8242 edfuller@ polestar.com Presentation to Nuclear Regulatory Commission Rockville, MD October 28,1998 1 l 1 APETs Primarily Treat Valve and Pump i Failures, and Thermal Hydraulic Effects a The EPRI approach to constructing APETs is similar to that in NUREG-1570, but the details differ. EOATs and tube rupture probabilities are also combined with APETs,in order to compute LERF.
- In EPRI TR-107623, V2, e APET was constructed for each of the three high/ dry accident classes.
- The APETs were quantificd and the thermal challenge frequency (TCF) was determined.
- A sensitivity study was carried out to determine which top event uncertainties had the greatest impacts on the TCF.
I
e
.;'i;;. .";.".'.i
".. ".';.".'i;T; "" " :." . ;;' ' , ' .. " . .;. i;; .
.C
;;;. T; .
.. . . = e l .. e e ee. e W.
e
-La__e-- .
- = :
h .. e .. ..
. .. ..u .
.- .e .m. .m.
u i .. . e me. C .. b .. m l& . ..e et
.m e
. . 4 4. .
. . e. 44.
- 9. .4. e 4.
. .4
. a. . 0 4. . e4.
- a. . e. .4.
Figere 4 3 E arly C ore D en ese A seideos P rogreseles E veet f ree (A P E T).lasledle s the E erst C ore D om ese E speeded O perator A ssies f ree (E 0 AT) l l l
..............u., . .. 0 6 W......
D..... .. tee ee . u...... W
.u...
q ..g
..e .e.mg.e. .t..E . .s.
b.
.e ...q.a.. ...a ..F. .u. . au. u..u . .
e s. e s e dt e
- e. e ne4e e
. 4 3ee de e 9.y og. e e. e i 3 2A at e .. e ese 4. e des de e.4 a te et f easee e h 1 e.e 4. ee 4 S.hL * **e 'e *** 4 *** 'e
.a. e e.e 4e e l
ee e ses 4. e l
. e ese de e i
i ee I a .. e e.e e. . e ese 4, i h ee e e.e ae e
. . . e. . .. .e e e,e se iu _L , g _._ e e.e.0 e
,u__ . .e .e ,ee a a. i e.e de . see 4e ee l e. e e.e 4 g f
M, g _._ e.. e
.u._ _ . e .e e e.e i e e. e e e 4.
O e. 4 .. . e e e 44
. 0. S 4*. SS r e a a. w. .e.e se e see 4# e e e a e
. e ses 4e e see 4e te.at e ate de Fagste 4 3 M id Core D ece ase A stident Fregreselos E vent f ree (A P ET) i 1
2
.. . ~ . -
e LaleQue APWpwm Gesidann Nghprusswo Thumul Tturmal W feceweed mMumind bl80s Questge? Chdiago True leanNy malmet pumwus emto proupancy army Pres tai Lys1 ses e mes 0 95 1 4.8508 fe 0 N 2 14507 ro 0 SA548 016 3 134548 to O 005 4 11508 ps 410548 10fAL. 110548 Hgm 84 Onbined IJde On Damage Expanded Opender Acdon Tm (EONI) and Accident Prognusica Emd Tm (APEI)
- l 1
I Top Events in APETs for Plant A ! (Excluding Tube Rupture Top Events) ADVs remain available for limiting high pressure when SGs are isolated. (Early CD sequences) l No stuck-open MSSVs (liquid challenge). (Mid CD) , No stuck-open MSSVs (steam challenge). (Early/Mid CD) No large RCP seal LOCAs. (Early/Mid CD) No concurrent loop seal and core barrel clearing. (Early/Mid CD) High SG pressure in loop where loop seal clears. (Early/Mid CD) High pressure in SGs (failures due to operator error).
! (Late CD)
M 3
Thermal Challenge Potential Assessed by Estimating Times of Hot Leg, Surge Line, and SG Tube Failure Using MAAP 4
- Various cases run for Early and Mid Core Damage sequences
- Larson-Miller correlations used for predicting creep rupture failure
- Degraded tubes simulated by reducing wall thickness by 50% (probably a conservative criterion)
* " Pristine" tubes maintain normal thickness
- Thermal challenge frequency reflects the likelihood of SG tube failure before either hot leg or surge line failure ADVs Remain Available for Limiting High Pressure When SGs Are Isolated Operation of ADVs can be restored by local manual actions, after they have failed closed due to loss of control power or air supply.
- Success means that secondary side pressure remains high enough to prevent a large pressure differential across the SG tubes, but the MSSVs are not challenged.
- Failure means either operator errors resulting in secondary side depressurization or hardware failures that lead to MSSV cycling.
- Success split fraction is 0.9 from HRA analysis.
4
4 0 No Stuck-Open MSSVs (Liquid Challenge)
=
Applies to Mid Core Damage sequences. MSSVs could undergo liquid challenge if SGs are at high pressure, and are being over-rdled. MSSVs are not designed to withstand such challenges. No data exist to evaluate the extent to which the valves would stick open, so the " total ignorance" split fraction i of 0.5 is assigned for success. l l No Stuck-Open MSSVs (Steam Challenge)
= Failure upon first demand due to maintenance errors
- Operating history shows one failure in 218 verified lifts.
- Failure probability per demand is 4.5E-03. For a four-loop plant, this yleids a failure probability of about 2%.
Failure rate from steam challenge during subsequent demands is established by tests by valve manufacturers
- No failures observed in over 1400 tests by Crosby and Dresser
- A conservative estimate of the failure rate per demand is:
P,,, = 0.55/1400 = 3.93E-04
- MAAP 4 calculates about 340 demands in a SBO accident in a four-loop plant for a dead band (blowdown) of 5%. This yields a failure probability of 13% during subsequent demands.
MSSV failure probability from steam challenge is thus 15% for this plant. 5
*\
l No Large RCP Seal LOCAs
- Large seal LOCAs (about 170 gpm per pump) could l develop subsequent to loss of component cooling water. l
- Seal LOCAs smaller than this would not lead to i significant depressurization of the RCS. l
- Expert clicitation for NUREG-1150 is the source for selecting the success split fraction. i
- Success split fraction is 0.84 for Plant A, which has new O-rings in its pump seals.
i l 1 No Concurrent Loop Seal and Core Barrel Clearing
- If both a loop seal and the bottom of the core barrel are cleared, once-through natural circulation flow to the SG tubes could result.
- If the secondary side is also depressurized, then high temperature tube rupture is likely.
- This is judged to be unlikely, based on MAAP results and hand calculations.
- A value of 0.9 is assigned to success of this event.
l l 6 l l
High SG Pressure in Loop Where Loop Seal Clears
- Thermal challenge only results if the RCS loop where the loop seal clears is connected to a depressurized SG.
- For sequences where a MSSV has stuck open and a loop seal and the core barrel have been cleared, the success split fraction is 0.75 for a four-loop plant.
- For sequences where the secondary side has been depressurized through operator actions, and a loop seal and the core barrel have been cleared, the success split fraction is 0.0.
EOATs and APETs Were Combined and Quantified In Order to Determine the Thermal Challenge Frequency Most Early Core Damage sequences, as identified in the IPE/IPEEE, have the AFW pump recovered locally, and are transferred to the Mid Core Damage EOAT
- TCF for Early Core Damage sequences results only when secondary side integrity is lost '
- Prime contributors to TCF for Mid Core Damage sequences are sequences where secondary side integrity is lost via failure to isolate (MAAP Case 2L), or when MSSVs fall from liquid challenge (MAAP Case 2C)
Late Core Damage sequences are small contributors to the TCF for this Westinghouse plant 7
e e e Results for Plant A Accident HigWDryo Highorys TC Freq. For TC Freq., Class No Recovery With Plant as it is Improved Recovery Today L:vellnd. Early Core 1.0E-05 3.0E 06 1.5E-07 1.5E 07 Damage Mid Core 2.5E 05 8.8E-06 5.3E-06 6.4E 07 Damage Late Core 6.5E 06 1.6E-06 8.lE-08 8.lE-08 Damage Tom! 4.2E-05 1.3E 05 5.5E-06 8.7E-07 Toble 101 Results of Sensitivity Steeles: Cases levolving Changes la EO AT Split Frsetiens Thereal Challesse F reguesey (SV) I' Case Case descripdee Chasses to est evente Mid Teest-(seesees values) segeasses ' I Base case Base case S 35 4 9 SE4 !- 3 Combined lead shedding and O A 4
- 9.M 6.45 7 5.75 7 leveliedieetles 3 Temely lead sheddies esly D Al
- O 575 i OE4 1 35-6 4 Backup bettery supply esly O A 1
- O 495 1 PE4 3 IE4 I 5 Behassed abihty es seawel D A3
- 9 99 3554 4 OE-4
' pressure leeally I e Lessi operesars' abihty se D A3 3
- O 94 4GE4 4954 l operses emetime anchest De Figure A.88) indicssiee impeeved l 7 Lesal sentrol is abandeced DA3 3
- 5 43 6 4E-4 6 65-6 (to Figure A le)
[ g4g4 5 Operators fasi se desece stae C A3 3 e g pg 3.j g4 l I level sentrol neede es be (la Figure A.10) l estabitsbed (1PE assumpties) l 9 5troisaic iselshen D A 5
- O 80 4 3E4 4 UE-6 le Fressure eestrel and eeeletiee 0 A 3
- 9 99 3.3E4 3.4E4 performed as needed o A$ = 0 to i- la Casse 3 and 40 sembined D Al . 0 94 5.55 7 5157
! oat =00 oA5
- 0 00 (SLI) oAl
- 8 80 (NSLI) 8 t
l l l L. -
----~ -- _ _ _ _ .~ . . . - .. - - . - . , _ . --. - - - . - - - - - . - - . . - _ . . -
Q-1 l I 1
- 8. af a 8 Y I T .61. I e,.3 aa c..,. i.s. .ta8.c. a1:1 a.u.nin .. 8.s.,.8 i. .e. d.I,..8.C
. c .. ... ..l.4.,3 I. .. 8C,6,..
ni .. s..... l
. . .. . .g;.. . . . .
. . . w.. ..... . .. .
l . ........ ........ ..... >u llMii'.'i.'".'.l1.*:7."ll * " ;;;; :i :
". ".'.!*.t"."t'.:.*::7'"' * " i;:ll:I;l fl.'."l'??.1" *Jt'.*;"."/,. 1;;1::::
l1. ;;:1::;I t
, , , ;;l2.11.'.*!.*,'.l n
;,.;,.g.,.; o,.,. ... . . . . .... ;..; . .. . n .. .o
!!.it.::r.'.t.'r.lir4' :::
l
.. .............. . . . . . . . ..4 ...
,r- .........n.......
.......6o.
........u i
1 Key Findings from Plant A Analysis
=
Thermal challenge frequency is fairly high, given current plant system configurations, procedures, and levels of training
=
Minimizing TCF depends strongly upon maintaining extended operation of the AFW system
=
A review of the AFW system and its support indicates that there is available margin for its extended operation
=
With improved training, intelligent load shedding, and extended SG level control, the TCF could be significantly reduced l i I t I 9
i w w - * % w e=,. w w w.-a. w s w .w .- .. e., e 4m x, . ..
. 'EH t i CALCELATION OF TEBE RUPTURE PROBABILITY
^ Marc Kenton Creare Inc. Presentation to NRC October 1998 Work Performed on Behalf of EPRI l MTG-98-10-959
Need for Methodology _m ,_.< _ . . . , . . - EEEH E Consider-High/ Dry / Low sequence in a Westinghouse plant. E We consider this " thermal challenge" because 50 percent thinned tubes calculated (by MAAP) to rupture ~500 secs after hot leg. E However, by the time of tube rupture, hot leg is very hot and creep damage very large. E The 'l thermal challenge" assessment nee ~ds to more completely treat relative heat-up rates, uncertainties, and plant-specific status of tubes. MTG-98-10-959
e -t New Approach
. - e mme. - e - we w .-
E T Bil
~
l EFunded by BG & E and EPRL . E Leverages a key finding of ANL work to conveniently interface severe accident evaluation to.
- . run-time analysis.
E Explicitly addresses various uncertainties including temperature. I lL MTG-98-10-959
Previous Attempts
# r.-+ g m e . m .. . e w + - .
t ll[ E Both NRC and INEL have performed tube rupture ~ probability analyses based on explicitly constructing defect distributions. E NRC concluded in draft NUREG-1570 this would t be difficult to use on a plant-specific basis. E Also,little explicit consideration of effects of uncertainties in temperature. MTG-98-10-959 ,
Coupling the Severe Accident Evaluation to the Run-Time Calculation 4~ , - em 4- 4., . gg[l E A series of STEIN, RTLIFE (or equivalent) calculations are performed at various assumed tube AP's. E Obtain probability of tube rupture vs. AP. E These provide probability distribution for M,: if one or more tubes fail at AP=5000 psid with ! probability 0.1, and nominal burst is at 10,500 psid, then prob (M, > 1050 00)=0.1. l 5 l i l MTG-98-10-959
Assessing Impact of Tube Damage on Creep Rupture
- - - . , . . ,. 3HI
- E ANL experiments
- effect of tube degradation on high temperature creep rupture is accurately estimated by increasing Stress used in Larson-Miller correlation by the same factor My necessary l to calculate low temperature burst pressure.
E The probability distribution for M ycan be obtained by exercising the run-time code (e.g. STEIN, RTLIFE).
Analysis Methodology (cont.)
.- . m %. w. e , . e u g(([
E Previous result is multiplied by the probability that My lies between assumed My and M + dm- (from run-time analysis) and process is repeatedfor otfier M, intervals. E Preceding result is multiplied by the likelihood assigned to the assumed temperature profiles and the
- process repeated for each profile treated.
E Sum of preceding represents the tube rupture probability for the sequence. i MTG-98-10-959
Analysis Methodology
- . . . . r- .,,, < aggi E Perform a MAAP calculation to obtain hot leg and SG temperatures as a function of time for one set of phenomenological assumptions.
E Calculate best-case, best-estimate, and worst-case rupture times for hot leg using Larson-Miller correlations. E Do same for tubes for a given M,. E As in NUREG-1570, assume rupture times are distributed normally and fit distributions to these rupture times. E Probability of tube failure prior to hot leg failure is then calculated by integrating over these distributions.
% re ura.ga.io.g3, 1
t Comparison of Predicted and Measured Failure Temperature using NRC Worst-Case Creep Rupture Correlation n n .u m e n n.wa n .;
-a w ,. . - .
t lll
- 900 880 -- ;
860 -- t 840 -- 2 820 -- ** ** , M 800 --
=
*+* +
+. ;
E 780 -- *%
++ + :
760 --
- g ;
740 -- + + ;
+ +
720 -- , ! 700 : : : : : ; ; ; ; 700 720 740 760 780 800 820 840 860 880 900 Exp.Tf ( K) i i i, - l
& MTG-98-10-959 l;
Validation of Methodology
- , ---,=w xe w = . + -
lll ; E Input ANL's "EPRI Ramp" and "INEL Ramp" into PROBFAIL code. t E Calculated M, using ANL correlation with measured crack length and depth. E Based rupture on maximum, not average hoop , stress (as recommended by earlier work on creep in thick-wall cylinders). ; i i l : i i MTG-98-10-959 l
Comparison of Predicted and Measured Failure Temperature i using NRC Best-Case Creep Rupture Correlation
==. - ==ee w - = - s E -IDi 900 880 -- '
860 --
+g 840 .
I [x820 - 800 -- v *
' E 780 --
760 - - 740 - - 720 -- 700 : : : : : : : : 700 720 740 760 780 800 820 840 860 880 900 Exp.Tf( K) i i I MTG-98-10-959 i
r I Comparison of Predicted and Measured Failure Temperature Using NRC Nominal Creep Rupture Correlation
.n me m .a n - m n -m: . e e n M. s Ql 900 ~-
i 880-- . 860 --
+% +, :
840 --
,+,
@ 820 -- *
- ku 800 -- +
E 780 - - 760 -- 740 -- 720 -- + 7 0 0 ,,- : : 700 720 740 760 780 800 820 840 860 880 900 l Exp.Tf ( K) I b- .. --id MTG-98-10-959 i
t Assessment of Severe Accident Uncertainties i 1~... ___ ..m . . .
+ IBi E Based assessment on calculated difference in temperature between hot leg and hottest SG tube around the time when the first rupture occurs.
E Believed a priori that in-vessel uncertainties l relatively unimportant because would tend to affect both hot leg and tubes. E By contrast, variations in hot leg phenomena would affect differences between hot leg and steam generator tube temperatures. E Sensitivity calculations support this view. b MTG-98-10-959
Consideration of Uncertainties
+ mnm noen a av e -
H }} E Methodology explicitly incorporates the uncertainties in burst pressure, NDT, etc. via the run-time calculations. E Uncertainties in base material properties are included using 3 Larson-Miller correlations for Alloy 600 and hot leg steel. E Include temperature uncertainties by performing a series of MAAP calculations in which phenomenological parameters have been varied.
@reare _ .,s . ,s,
\
?
l CE: Effect ofIn-vessel Uncertainties i
*t w wre%4ttAws<ngayn @ Mahep;%9sn.snga n '*: -
%n sw n ., i; h
25 - -
/ / *
. /:P, : -
'..- / '.,e / s .
~~~
t M 0 - - a.
.s
/ / s s-H E
e . t f
/*,r>4::'..
/
/./.'.
1 M
,.' r ..c.'.
/
,s
/
s' ,.'i a -25 - : j/ - i en O
~ /*,// . -
/
./
< * . 4. ./. ,/
/s -. . /.r.s-m -
e a ..- e
$ -50
,/ ' C' Slow Relocation .
, -- --- -- Effective Core Collapse :
J "Eutectic" Model High Core-UP NC Flow Low Core-UP NC Flow Base Case u
-75 O 100 ---
400 P Time since Damage Significant, Secs I g MTG-98-10-959 ;
t h I Westinghouse: Effect ofIn-vessel Uncertainties , e. a
- g,* . ' '. j ' -.
- yWe 44% ' k:k'W9 's 4Q$ $:q %Q+'wc:o Q ' Q.. ; 'v' f-i e ; . l ,'r :: 4, );
. + :. ;q . ; y.
I 225
~ -
/
**,.. 7 210 -
/ ,
p- /* i M ,/ ',/ ,
/7.
- / / .
< / .. '
,/ - +
i ' s-. g 195 - j j
,. **s -
y 6
/ /*,/ , ,/ /'
j f
./
us 6 . /
. W*
d 180 -
.. i
. - I -
en j .....- i
~.. ** /
. /
$ .~ /
/ i
/ -
Slow Relocation
~~~
* * * **5*
165 -
*** I t' -
.- .- / s ,e "Eutectic" Model -
- s High Core-UP NC Flow
.f ,. /* ,s/ ,
yf., Low Core-UP NC Flow l i Base Case . . l 150 i 0 50 100 iou av 250 l Time Since Damage Significant, Secs i re MTG-98-10-959 1
CE: Effect of Hot Leg Uncertainties
- 1. M e
- w d 4. W e % 5t n gle r p r e m 4 Accg m m k g s n o o ne- .:y; , -?
50 . . High Mixing .
~ --- ~.
*W Afixing .
SGA1L finW Ratio ,. . H19h Rag, M ~ ~ ~ ~ ~ ** Ramatin,
~
[ ~--- g 0 case 2 o
~ ~~ ~~~ ~~~~~~~ s
,. ,.f f * -
O
.=
'- , , , , ~, - > . _ . <
, ~ .
r.. d ,,. . u
-~ ~ . ..,... # ~~~
3 -so .
,.~.-
~
E s
'~ ~ ~ ' - ~ .
- z. ' ,.,_
3.,.
-100 0 100 200 300 ~ 400 Time Since Damage Significant, Secs Greare MTG-98-10-959
_m_ _ _ _ _ _.- - _ _ _ - _ _ - - - _ _ - -- m.
.)
Westinghouse: Effect of Hot Leg Uncertainties I
. g -tse g 7 w atr u e e w g ew s w : 9.rt' s p p ., e y t y. 1 a 6. t4 9;_ ;e .w. 3 .
v M . g{ 250 . M 225
/
/
/- - -
ct / *... :
/ .-
E / [ y / g 200 / /,. ,,/,/ ,,..' - l 1 / e . v2 6,
. / .. ,..
> . /
s,/ i
-/ ,.. -;/"**""**... f S. .. . - ,. . . ,. . -
l R 175 .g:,..
,/ -
3 /
./,. --- '.p'.
',/,
-E **,,. **
j
.p.
High Mixing - 150-7,
- f * . '.. .' - --- ---
Low M. .nxnng : Low SG/HL Flow Ratio ,
>.I.<' -----
High Radiation ; L w Radiation m 125 ' 0 50 100 Base case '250 MTG-98-10-959
ouemwry . Sensitivity Calculations
< E 81 Developed by assigning a likelihood to having high, nominal, or low ~ values of the important phenomena.
For each accident sequence, about 8 MAAP runs l are made by combining those factors. MTG-98-10-959
Effectiveness of Thermal Radiation is a Particularly Important Phenomena
, . - . gggj E Effective radiation between gas and hot leg surface j favors hot leg rupture relative to tube rupture and conversely.
E To supplement (very) simple MAAP model, additional analyses are performed using a detailed radiation model (Balakrishnan and Edwards, 1979). E This auxiliary model also treats conduction in hot leg in more detail, t E The temperature calculated by the detailed model are actually used in the rupture calculations. i MTG-98-10-959
N Benchmark #1
+ $wete4ua < 'iN vf M dwM- F*FF.v4fMW.:": 7' '- -T'ht
"+'.
4 W # - 1E+0 1E-1
/
/
? }
)
fIE-2 f .f li y . m *
-*-P1 CM 1 or more
-e-P1 OA 1 er more
-P1BOCn+1,1 or more 1E-4 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Pressure Difference KSt
@eare re.,s.1e.es ,
Phenomena Investigated to Assess Uncertainties in Hot Leg Heat Transfer Phenomena
. - . _ , - - . . . . . - - ~ a., _ ,_ . gg ll
~
Phenomena Variation Relative Likelihood
- Assigned Hot Leg 150% of nominal 0.2 Radiation Nominal 0.6
- 50% of nominal 0.2 SG Plenum High 0.2 Mixing Nominal 0.7 Low 0.1 SGRIL Nominal 0.8 Circulation Reduced "Out" 0.2 Strength Fraction l l (Q MTG-98-10-959
Results of Tube Rupture Calculations w m e O W>.s w m m ,w w w m m w w :. e n ., ;- lQl Probability of Rupture Case #*
. Thermal
- 1. Stein Benchmark 1 EOCn .26 .05 BOC.13 .22 .M EOCnu .35 .11
~
- 2. Stein Benchmark 2 Condition Monitoring .002 ~3.E-5 Operational Assessment .006 ~4.E-5 MTG-98-10-959
Burst Probability of STEIN Benchmark Problems .
'M %' q .Y - f %% -
4 J y . 1' , e. i 10 0 -
. ',.,.* 7 , / *'*s.n ,
/
*** /
10-1 -
/
/*. '.' / ,
/
/ / '
/ !
/
D
$ 10-2 .
E
**** /
./ .
~
o '
/ -
/
.D Z-
<a -
/ i S- 10 4 - / , !
' / '
/ '
l i 10 4 - / ,
/ .
/ --- STEIN Bench 2 Op. Assess. '
----- STEIN Bench 1-EOC ,
10-s 2000 1000 3000 4000 5000 6000 1 q Differential Pressure (psid) MTG-98-10-959 ! 4 3'
Conclusions
- p +e e . -
Mgl E A methodology and associated software package have been developed to facilitate the calculation of tube rupture probability. E The predictions of the software agree well with ANL rupture data on tubes with machined defects. E The key difference from previous work is the method used to represent tube defects using the results of the run-time analysis. l MTG-98-10-959
1 Method m .. ,.. . - - v . . , ,. . . .. . . , ., . EIHI E Ran a series of MAAP calculations for following
~'
sequences: (1) High/DryAligh (early SBO) (2) High/ Dry / Low (early SBO) (3) Transient with loop seal clearing and forced uncovering of core barrel l (4) Same as (3) with depressurized SG in uncleared loop (5) Same as (3) with depressurized SG in cleared loop B Utilized as input, calculated burst probability vs pressure provided by Aptech at BOC, mid-cycle; and end of cycle.
@ % re y10 981o.96o
- a G
$ I
I
- , . . ~ , - - . . - .: . . . ,3; g3gl RESELTS OF TLBE REPTERE PROBABILITY CALCELATIONS FOR CALVERT CLIFFS t
Marc Kenton ! Creare Inc. !
, Presentation to NRC Staff i October 1998 :
i I !
, Work Performed on Behalf of EPRI I
--{g MTG-98-10-960 ;
I i
8 4 4 Phenomenological Uncertainties Considered t III E Cases with counter-current flow t (1) Base . (2) High radiation (3) Low radiation (4) Low SG/HL flow ratio (5) Low inlet plenum mixing , (6) High inlet plenum mixing (7) Best combination of above
~ (8) Worst combination of above E Cases with loop seal clearing and uni-dimensional flow (1) High radiation (2) Low radiation (3) Low flow (effect of high resistance, e.g. pump impeller)
I MTG-98-10-960
Probability. of Burst vec cw.**now4 heur4Ammw+:v.*g.uw a te ww ow .. .. u g ., -- -
]
10"- - - - s.,
'"*, s
. ,/ ,,. - ,- . ..- ,
/ -
10- r /',, *-
/ .. ,. / /
s
/ ',' /
n ,/ /
$ 10'2 -
/ ,,- j ,
z% :
*,/,
- i M l g- ,,- / ,
4 *' / Su 1.0 a-
' /
. ,s
/
./ :
I 104 - - -- '
/ - STEIN Bench 1 EOC +1 n ,
! / -- STEIN Bench 2 OA
. ...- Aptech EOC 10-s 1000 2000 3000 4000 5000 6000 Differential Pressure (psid)
MTO-98-10' 960
1 Overall Results
- e.ww.u w ,4+wmawe
- u 49 EEm 1 3 Ql BOC Mid EOC High/ Dry /High nil .001 .01 High/ Dry / Low .09 .33 0.69 Clear loop, all SGs pressurized nil .004 .02 Clear loop, with depressurized SG .08 .35 0.76 Clear loop with pressurized SG .001 .02 .08 .
1 DRAFT l U & MTG-98-10-960 l 48 8
Conclusions Based on Preliminary Results
. - * - - ~ = ++ * *
- E F I31 E High/ Dry / Low cases with no loop seal clearmg appear to present a significant risk of tube rupture, caused by relatively high calculated temperatures.
E High/ Dry /High cases pose little risk of tube
- rupture.
E Forcing loop seal clearing and the uncovering of the core barrel does not significantly alter the results. i p MTG-98-10-960
=.e l
a w Overview
. First, this is an inprogress evaluation, and noformal results are available.
Backcround NUREG analyses indicate that induced SGTR may be a significant contributor to LERF. As a result, this issue must be incorporated into the CCPRA to evaluate the significance ofinduced SGTR. Obiective Incorporate induced SGTR Events into the Calvert Cliffs PRA (CCPRA) using the latest SG inspection results. Scope ofInduced SGTR Modeling Areas are evaluated based on predicted SG degradation:
. Increase in spontaneous SGTR (considered negligible impact) e Increased risk ofpressure induced SGTR e Increased risk of temperature induced SGTR Each of these areas is evaluated as a function of time in cycle to allow subsequent evaluation of mid-cycle SG inspections.
Mid-Cycle SG Inspections Use the model to evaluate the increase in LERF resulting from not doing mid-cycle SG inspections. The comparison will be based on the amount of - time justified deterministically vs. a full cycle run.
v( 10/26/1998 Pressure induced SGTR REU/SGTR-PLN.XLS Challenges i
# item -
1 Steam Line Break (Unisolable) 2 Steam Line Break (Isolable) 3 Excessive Feed Event (not isolated) 4 Stuck Open TBV(Isolated) ' 5 Stuck Open TBV(Unisolated) 6 Stuck Open Safety 7 Stuck Open ADV 8 8 Turbine Trip Failure (Unisolated)
- 9 ATWS
?s Answer --
if decay heat removal is available can this impact be No, based on NUREG, and draft this should be reasonably screened. . considered. Can all of the isolated events (i.e. MSIVs close) be ,
'No, must be evaluated screened Can all of the unisolated events be reasonably binned to! .No, preliminary binning is contained in sheet a single frequency of SGTR "PM Case Layout" If not whatis a reasonable binning? 4 10/21/98 binning is reasonable Given the SSRV are cycling with AFW available, but no i l(Currently using existing SSRV frq: 2.87E-3 ADVs and TBVs what is the likelihood that the SSRVs 'per demand,10% max; better information will stick open over 24 hrs? (Answer should include :may be available )
demands over 24 hrs, and failure likelihood for each ! i cycle after the 1st) i Does LOOP Seal Clearing impact Pressure induced i ;No, LOOP Seal Clearing affects the SGTR? ' ! temperature induced phenomenon only. Can ATWS be screened due to low frq and S/G tubes l !SGDSM indicates Yes. NUREG hints yes, but strongest when ATWS mostlikely l defaults to we should check. Ed Fuller states that this should be considered. Currently
. modeling.
-- Ex.
vv%SEEFL c mWsdG=fLihMmn =&caEwrM"PiML%irum*7B &TK mitwnw m2 .'W (pg. 6-10 SGDSM/PSA: ATWS) l 'Other Cycle Affects BOC 0% l 'MTC (Impacts PORV & ATWS recovery) MOC 21.30 % i EOC 46.20 % ! i. E i + Page1
10/27/1998 High-Dry induced SGTR REUISGTR-PLN.XLS Challenges
! Issues 1 LOOP Seal Clearing 2 2ndry Pressure 3 RCS Pressure All cuestions assume a complete loss of decay heat removal
]~ +
- Can all 2ndry Pressure losses be lurnped together or does the way we de-pressurize affect the SGTR conditionallikelihood?
No, cases are sepsisied as in *PM Case Layout". Does LOOP Seal Clearing significantly affect the likelihood of SGTR? NUREG indicates yes (pg. 2-50 pristine tube could fail). Ed also states yes. is the only way this can occur by means of a RCP Seal LOCA. Others possible, but RCP should dommate; RCP Bumpmg is also an issue, but RCPs bumping is not done or rare. It is brought up in the drills, but always dismissed. How much should the RCP Seal LOCA likelihood be increased during a CDF event? Ed Fuller suggested ask Marc Kenton, Marc felt an increase is not warranted as the water in the loop itself prevents the seals from seeing much higher temps than normal. What frachon of the time wit an RCP Seal LOCA cause LOOP Seal Clearing (Diablo Canyon used 10%) (Right now we are using 100%) Given the SSRV are cycling wdhout AFW available, what is the likelihood that the SSRVs will stick open~ prior to aN inventory bemg lost? (Answer should include number of demands until S/G is depleted). (Cs......^4, using existing SSRV failure probabilities) El High/ Low SGTRs pg. 6-29 SGDSM/PSA BOC 0% MOC 10.7 % EOC 33.8 % Page 1 d
n$ vervu~ncevuswu Simulation Results
- l Time in Cycle - EFPYs j Delt kP 9E 1.21 1.83 Mi 0.00 % 0.00 % 0.04 %
Lil 0.12 % 0.14 % 0.41 % Mi 0.20% 1.21 % 8.86 % ! 14 1.65 % 8.69% 41.38 % i Simulation Results with Zeros Removed J.oaarithmicInterpolation Results l Time in Cycle - EFPYs Time in Cycle - EFPYs Delt kP 9E 1.21 1,83 Delt kP 0.01 1.21 1.83 1 1.45 0.00 % 0.00 % 0.04% 1.45 4.20E-05 4.90E-05 4.00E-04 Lig 0.12 % 0.14 % 0.41 % Lgi 5.61E-04 i 6.54E-04 2.41 E-03 M1 0.20% 1.21 % 8.86 % M 8.20E-04 9.57E-04 i 3.14 E-03 14 1.55% 8.69% 41.38 % Li 1.04E-03 i 1.22E-03 3.71E-03 2.56 1.20E-03 i 1.40E-03 4.09f.-03 I I I i 1.45 to 2.56 Case 0.01 1.21 1.83 Slope 5.898 1.000 0.693 y intercept -12.26914387 10.15415068 -0.942140588 i l I l Page1
NEARLY FINAL TUBE RUPTURE CALCULATIONS FOR TRANSIENTS THERMAL CHALLENGE ONLY; NOTE USING LOGARITHMIC INTERPOLATION BETWEEN APTECH-SUPPLIED POINTS 10/22/1998 Results with an e-mail add on 10r24/98 Probability of tube rupture due to thermal challenge ' Probability of Burst on a Single StGs Case . Early Mid EOC
- 1. Transents with no RCPs and no seal LOCA (so no loop seal cleanng)
Fully pressurized pnmary and w, vie.if (high/ dry /high) 0.0% 0.1% 1.0% Fully pressunzed pnmary and 1 depressurized sec. (high/dryllow) 8.0% 33.0 % 69.0 %
- 2. Transeis with 1 loop sealclearing '
Loop seal clears, both S/Gs pressunzed 0.0% 0.3% 2.0% Loop seal clears in loop with depressunzed S/G, other loop pressunzed 8.0% 35.0 % 73.0 % Loop seal clears in loop with pressunzed S/G, other loop depressunzed 0.1% 2.0% 8.0% I ! i l O b , i l . l l
uo q 10f28/1996 PM Case Layout REUISGTR-PLN.XLS A B C D E F G H- 1 J L.OOP Decay RAmmircum BOC ISOC EOC Seal Neat Nature of Delta 1 Case Value_ Value_ Value_ Description Cleared _ Removal? _ChaIIange_ Pressure _ Notes 2 1 0% 0% 0% 2ndry Pressure is Normal n/a Yes None 0 2 0.11 % 0.13 % 0.48% 2ndry De-Pressunted with MSNs successful n/a Yes Pressure 2250 (single SM3 de-pressunred)(e.g. Stuck Open ' ADV, Stuck Open TBV, MSLB, etc.) 3 3 0.22 % 0.26% 0.96% 2ndry De-Pressurized with MSIVs faBed [heo nfa Yes Pressure 2250 SM3s :': r i_
- 4 (e.g. Stuc te Open ADV, e
Stuck Open TBV, MSLB, TT fenure.TBV 4 fadure, etc.) 5 5 0.16 % 0.19 % 0.63% ATWS n/a Yes Pressure 2400 MTC worst at BOC 6 8% 33% 69% ATWS No No Pressure or 2400 Values assumed to be~ 6 Temperature same ac Case 8 7 8% 35% 73 % ATWS Yes No Pressure or 2400 Values assumed to tA 7._ _ Temperature sarne as Case _8 _8 0% 0.1 % 1.0% 2ndry Pressure is Normal No No T. . e/_-. _1850_ _ , Based on MAAP 4 8 celcs. 9 8% 33 % 89% Stuck Open Safety and other Segle Header No No Pressure or 2500 Based on MAAP 4 De-pressunred Scenenos (e.g.1 ADV Stuck Temperature celcs. 9 Open, MSLB with MSlVs closed) 10 15% 55% 90 % As other two header low pressure scenanos No No Pressure or 2500 Values assumed to ti (any studt open 2ndry equipment with MSN Temperature same as Case 9 to faaure [eg. TT faRure, SSRV isBure, etc.D 11 0.0% 0.3% 2% 2ndry Pressure is Normal Yes No Termerature 1850 Values assumed to be 11 same as Case 8 12 4% 19% 41 % Stuck Open Safety and other Smgle Header Yes No Pressure or 2500 Values assumed to be Dwak.d Scenarios (e.g.1 ADV Stuck Temperature same as Case 9 12 Open MSLB with MSIVs closed) 13 15% 58 % 93% AR other two heeder low pressure scenanos Yes No Pressure or 2566 - Values aIsinined to be (any stuck open 2ndry equipment with MSN Temperature same as Case 9 13 faRure [eg. TT feBure, SSRV fa5ure, etc.D 14
~ ~- ~
15 Note: If a scenano includes both a pressure and temperature phenomenon, then the IBteEhood of failure noted above~
" ~
shoukimdude__both,HkM j ~l
~
l ,, , _ l ,
. 17, For example, scenario 6 indudes a short durahon ATWS pressure chaNenge, then a long duration temperature chagenge ,_ _
i 3 If the quecess of the short durabon chatenge reduces the likeWhood of a temperature failure 19 cheNenge in the kmg tenn_(ie. denenstrated tube strength)3 N, en this should be indudal in the 20 single faikne probab5my provided for a case. l l Page1
W $7 i,/f- 10/26/1998 PM SFs Rules REU/SGTR-PLN.XLS w $15 SF or Macros PM # of S/Gs Case ATWS De-prsami SSHR ' Seal Rule LOOP Value , OC_ . . , OCS 0% n/a n/a n/a n/a n/a n/a (-(INIT=LLOCA+1 NIT =MLOCA+1 NIT =SLOCA)) OC1 50 % n/a n/a n/a n/a n/a n/a 1 _ SSHR:= Secondary Heat Removal Avadable F1=S*F3=S+SHSD+SLPF__ _ _. SSPH:= Secondary Pressure High (BS=S*TT=S*(-(INIT=L.SLBD))+MS=S)*DV=S*SW=S*(-(INIT=LSLBU+1 NIT =SLBI)) _ WRS 0% n/a 1 No O n/a n/a SSPH*RQ=S+PT=F WR1 0.12 % S 2 No 1 n/a n/a RQ=S*MS=S*OC=S f WR2 0.31 % F 2 No 1 n/a' n/a RQ=S*MS=S WR3 0.24 % S 3 No 2 n/a n/a RQ=S*OC=S , _ _ _ _ WR4 0.61 % F 3 No 2 n/a n/a RQ=S WR7 0.18% S 5 Yes O n/a n/a SSPH*OC=S WR8 0.41 % F 5 Yes O n/a n/a SSPH _ ___ _ WRF 100 % n/a n/a Any Any n/a n/a 1 TRS 0% n/a 1 No 0 Yes n/a WR=S*(SSHR+0TCC)+PT=F TR1 20.5% S 6 Yes O Lost Intact WR=S*SSPH*SL=S*OC=S*RQ=F TR2 51.0 % F 6 Yes 0 Lost intact WR=S*SSPH*SL=SPQ=F ~ TR3 21.5 % S 7 Yes 0 1.ost Failed WR=S*SSPH*OC=S*RQ=F >
~ ~ ~ -
TR4 '~54'.0% F 7 - e' s 0 Lost Failed WR=S*SSPAPQ-F _ TRS 0.1% S 8 No 0 Lost intact WR=S*SSPH*SL=S*OC=SPQ=S___ ' ~ ~ '
-TR6 TR7 0.6%
21 % F S 8 9 No No 0 1 Lost intact WR=S*SSPH*5i_'=SW55-Lost intact WR=S*RQ=S*MS=S*OC=S*SL=S
~~[~~ , ,_,
TR8 51 % F 9 No 1 Lost . Intact .WR=S*RQ=S*MS=S*SL=S l TR9 35 % S 10 No 2 Lost intact WR=S*SL=S*RQ=S*OC=S TRA 73 % F 10 No 2 Lost intact WR=S*SL=S*RQ=S I TRB. 0.2% S 11 No 0 Lost Failed WR=S*SSPH*OC=S*RQ=S TRC 1.2% F 11 No 0 Lost Failed WR=S*SSPH*RQ=S TRD 11 % S 12 No 1 Lost Failed WR=S*RQ=S*MS=S*OC=S ,
- ~ ~ -- ~ ~ ' ~ ~ ~ ~ ~ ~ ~ ~ ~ "
I _ __TRE __30% F_ _12_ __No_ _ _ _ 1_ _ , Lost Failed WR=S*RQ=S*MS=S TRG 37% _S 13 No 2 Lost. Failed _WR=S*RQ=S*OC=S TRH 75% F 13 No 2 Lost Failed WR=S*RQ=S TRF 100 % n/a n/a Any Any Lost Failed 1 Notes. OC=S is a MOC trip, OC=F is an EOC trip When the MSIVs close (i.e. MS=s) only a single S/G is de-pressunzed
. ___ _ _ _ ___- - __ _ _}}