ML20003C996
| ML20003C996 | |
| Person / Time | |
|---|---|
| Issue date: | 06/27/1980 |
| From: | Fabic S NRC |
| To: | |
| Shared Package | |
| ML20003C956 | List: |
| References | |
| NUDOCS 8103181030 | |
| Download: ML20003C996 (47) | |
Text
._
4 O
l CODE ASSESSENT STRATEGY Atm CODE ACCEDTANCE CRITERIA (TENTATIVE)
PRESENTED BY S. FABIC, NRC AT JOINT MEETING OF THE CODE ASSESSENT REVIEi GROUP AtlD THE ADVANCED CODE REVIEW GROUP SILVER SPRING, JUNE 26-27, 1980 l
l l-I e
8108181030
/
l MISSION OF BEST ESTIMATE CODE TO PROVIDE INFORMATION ON THE EFFECTS, CAUSED BY A VARIET(
OF POSTULATED ACCIDENTS AND TRANSIENTS, O!! CERTAlti KEY PARAMETERS IN LWRs.
AUXILIARY MISSION:
TO HELP IN DESIGN OF TEST FACILITIES AND INTERPRETATION OF THEIR RESULTS, WICH Ef PARAMTERS7 DEGREE OF FUEL CLAD OXIDATION (PAXIMUM LOCAU AMOUNT OF HYDROGEN GENERATIO!!
GELATED TO CORE-WIDE DEGRE.0F CLAD OXIDATION)
CORE GEOMETRY GHETHER OR NOT DISTORTED, DAMAGED AND IF MELTING HAS OCCURFID) i ALL ARE RELATED TO CL@ TEMPERATURE DISTRI3UTION, IN TIME AllD SPACE.
c
1
.i SAFE 7Y 4~/SN/f/tWNCG AS bErERM/Nfb 7NR/W4N 7NE RSd GF VEA/r/fa.
3Esr atr/NATE l'obd fir 4M47'sF key Afskers A N/ OCN7 ZCGNAA/0 m Passer /t.77 of OceMJ4Nca m A.
X lAA oria4 7*"
'norstitir1 or (Lo*M. Mex )
44/M4 dxC204 Gb
- (86)
J '$~
.0.
PA94Ad/4/!? #F C.BN4AA794 dt/NG dxtf6dGo (8 6)
O $~
-N 6
mm -
6 m 6
l l
\\
.ym-m-sv--
,,r.-,r e
-w--,,--v-,--
l
?
Il i
l yy i
I I
I st j
i l
I E
5 I'
[
l I
(m
(
l I
I tt i
I i
f I
I
\\
l 2
4 k
e 2
,a i
i U$
I
,4 s
x h,?
b I
k 4
I I
I 3
w 5%
kl 4
A Rg Sk k
y*3x0 s
u y
e 4
2 n
=+*h j 4478 M D (u E
e' n
c, s
s k9 9
I9 4
g" w
- DN{ I U <9 vJ x
t e
w%Nv s
(5 E
4
% h b, NT%e 0k%
g n<<<
t "s
e' w w w&
R4 G
4 1
ACCURACY c-0ALS I-OR THE BEST ESTIMAE CODE THE OVERALL GOAL IS TO PREDICT L'4R BEHAVIOR WITH FEaSONABL:.
RATHER THAN EXTREN ACCUPACY.
THE REASONABLE ACCURACY IS CONTROLLED BY
- THE LICBSING FEED
- STATE-OF-THE ART IN tEASURE'ENTS, PHYSICAL UNDERSTANDING AMD MODELING, !!UMERICAL SOLUTIGH TECHNIQUES RES RESOURCES - COST /EFFECTI'/ENESS i
i i
1 l
l L
s-TO SATISFY THE LICENSING NEED OUR AIM HAS BEEN TO RELATE, AS MUCH AS POSSIBLE, THE B.E. CODE ACCURACY REQUIREMENTS TOTHEEMCODE(RESULTSIACCEPTANCECRITERIA,BEARINGIN MIND THE PRIMARY MISSION OF THE B.E. CODE.
THE OVERALL PR0rFRS REQUIRES:
1.
SELECTION OF THOSE COMPUTED RESULTS WHOSE ACCURACY NEEDS TO BE DETERMINED AND WHICH PROVIDE SUFFICIENT MEANS FOR EVALUATION THE CODE CAPABILITY TO PERFORM ITS MISSION 2.
SPECIFICATION OF THE BOUNDS OF ACCEPTABILITY OF ACCURACY OF THE SELECTED RESULTS 3.
SPECIFICATION OF THE CODE ASSESSNENT PROCESS THROUGH WHICH THE ACCURACY DETERMINATION CAN BE ACHIEVED.
l t
t s
~
4 SF1FCTION OF THE COMPUTED RESULTS GROUNDRULES:
1.
TAKE CARE OF THE RESULTS THAT MATTER IN THE LICENSIt!G PROCESS, AS RELATED TO CLAD OXIDATI0il, H -GENERATION, AND CORE DAMAGE, INCLUDIllG THE PEAK CLAD TEMPERATURE 2.
DEFINE THOSE RESULTS WHOSE ACCURACY REQUIREMBIT MAY BE QUANTIFIABLE AND WHICH DEPICT THE CODE ABILITY TO CALCULATE IMPORTANT EFFECTS OCCURRING IN THE REACTOR COOLANT SYSTEM.
TH$5EWILLBEHAPED"KEYRESULTS."
i l
l
7 3.
DEFINE THOSE RESULTS THAT ARE USEFUL IN DETERMINING WHETHER (A) THE CODE MODELS THE PHYSICAL PROCESSES WITH SUFFICIENT REALITY S0 THAT THE AGREEMENT CIF 0BTAINED) BETWE91 THE CALCULATED AND THE MEASURED " KEY RESULTS" IS NOT FORTUITOUS; AND G1 WHETHER THE CODE MODELS HAVE A SCALE-UP CAPABILITY.
IT IS ANTICIPATED THAT IT MAY BE IMOSSIBLE TO DEFINE THE ACCEFTABIE ACCURACY LIMITS FOR THIS TYPE OF RESULTS Ai!D, THEREFORE, THEY WILL BE NAMED THE." DIAGNOSTIC RESULTS."
5 i
m 6~..,.
\\
8 KEY RESVL TS - FOR &usNnr4ityg A.ues.1ggyy C#NCff 7, CQN C MA.N/N6 sfRAtred t**RG i
i l
l AN!
iffAT-MP AMD AddNCM CycLdC Add CANZde ty to4LT PLew M
/ Luth beNft??' in Anyt/ G N /~Z Tu, i
I v[
!k
= n-,
g,em.
U l$ NEN l0YMS Y9 ANY ffMPN7A7/fMel CALL (Hogg)
W/TH/N FMd RAstint and, #A 172 Z/MMAT*d Q M/17947, /Z Atsbicrge A N 6 Alge N g,agen ge, b=b$
lZ iM PGAM 7*RMP6 Advt 4Mn /N THAT* ddLL, Atsgpassg 9Vd4 TN4 CRLL L s*Ct
/2 TN, fenex 78MManistAt /N Th6 4*ALL, fMtowf
~
g, AND M44ZUA64 FM TNG M97PfN WM646 TH4 LocnL h w 4.4 #444Anr/oM /t M/6M447 l
AAT5 ff $ X in d E 6 M & f W A l*?* N /Nig (An6&b/Np Avr Lf, :
e d&,,
A 83l*/24 s/R N},u
~
4 M
A. -^
2 =laW1/ %,g n u >, e a,.. >
WM4N /HrGMAree :
n b) ul; +t3 u%
s,-a -
o
-w.
an,--,s
,..--vm.
.-w,
,n.
,ne.-,n_,us-
--e
.+,-- - - -
a
9 Dox. - -[E ~k -
e r,,.
W HAA T H 6 S J4 g sc 4 / P r 4,' bgfA%65 A PART/ Cat.AR C4Lt.
{ s CAAb 7Wldtha153 fat. /*e n /O' Aen,)
i N g. A&/Ad$f4/T 6edkL PAAd7/0W W C4444/N6 7"N.4 7*
/$
$Xfbl120, // Nf4CLLAA Fit Rf.
Cl Abd/N& WdAd 7>
ER?CA/AN4:*
78a sam runseAsrue H/srasy.
FMf44Z/k&Z An>'t./7tehet spMe 4pAA7"/nWs ac 73Wyn. Skxg, 7Ni GAA4tfA ANA 44AGGA PGAM: f f/N'- 6/VGJ Meets te&d.yr.
s Do x = Hsx (as>;., a. /,,,.... - )
AG/AGZ6N7Z H/FM4Z7 LetAl QX/ed FAAC?"/0N s Pcr = nn O res.,,.,, n.... d AGPREZGN71 det2 -W/dd PGAult C4A6 74/t,*d9A.M &
i H1 = k f [ % (A-A-) at].
i bv'H545
}/
- NWngdA en PWit. AJdf /N C GLt. i 41m M&/4M7 08 C414 s'
K u p R o p e a r to m t,nny & w v 4 $t w n )
cowtrAnt ML ANPAftGNr1 7s' 744 PrwNr* #F MyAAo164 kEM6A.arsh /M 7WG A 6AtFIA CeR4.
l Gboy, $7c7 ANh N2 NAVG 7NG/A CJWMr?A?nArt /N 7W4 AM&Mb/k X L/rgas/NA CA/rst/n.
l i
a
.3 i
. cat snAt.t. assax 4ocs s.wsiyte r twa rou.swaa NKy AdsWL7C AAd.fM*601794, /N A4e/rrsmy 72 7 word g/greo hieht :
T' MARK a W
0 m
0 Y
GBA% s fWk.)
9 I
J6NX=
(Nd.g ~ % 4,,)df t f.nwes svoopy su sture i
- (sr****
- ~*"!
su.s.s sm, si
. e-F L4s/6 n w Z. A.f4 4 esen t a y 4/4 e
=a WHQt&
TA 7"fMet AT W//M 696As M piggr
/NP04VO40$ 79 Md6//y F4W/4 tswa/PrM GM Z.C J"3mMeeAy 4/eb;
'C" Is stas ix ?ds-rser /Asekisms.
F**
Ancr-rs.cr m y:4,t s,ersan Acrws.
M4AAron Acrrows AA4 xanN, g *asser p
/C in tf M:44 /NsPdAe of t.
2l,, = 7"/Ms er 4Asr fissacH, Auv wvgag
,~ -. <,,..
/p i* x 4 E" MCd W m
W
= insinwnNeones Pears nntz sess 7WAntfM 7W6 aftAx l
k,y a RNPHMM W M usa.;ineti upCPA cA n so rwe 44&nx
.f s /N.ll7AW7ANd ent! MAff Al#W #M Z73AM isswn, Tw< ar-sn n-an<
- GM7HAL AY At CNAM 14A WNe S. J.
If I
s /MZrAMr/n99Gs MAC / L eser GM L/dW/C JMrfA/N4 7Nd fr9AAf 64H02Ar*4 ZCinuedRT ZJed ENTHALM ff /Nf*M/N 6 L//W/b 3
$enW/td h'44K AN4 ARAM A26 A 6 LArnb 75 7He hints ANb P M M A M A l. Ah6A6Y ef f 4 M/e & Ad v/N r. 724G P4/MAAy dedLAM7 C)* LP9tt AMA/N4 79ws A4Asa y
$$ N X AGPR6KBHrs TW4 AM*dNr JE NBA7 AJdHANede 4671weFON TH6 Zr M er 40NdAArrt PA/Ms*4)r AMe 194 I
Z8toueAAy t/b es.
TN4 M A 7/ ext.s 4 roan was CHors-s l
ddinasa Ir cdxrnins MestaAHt e tw*Nrr 1-1as.
rue sinnce andaer su 78 4 Zdtowe4Ar Sibe enn 44 P9WM6 AAsN TH4 /JRAZWAdNON71 ff ' TMd AX/A4. P4SZ4Antt
&J7Alstarrf 9N 4NO PAdM THd AdZ34.M78 E4dZfhtd4 sortrw/x 7WE C. S. 444MAA r 4/ed. 4x/AL ra*rA*derude
.)w~astwrten W/ THIN THe tudesesse section en t.C.
14t**enAr s/s c /s 441/44s444 enr war asswrist Z/Med Metr of twe 2NdA*7 Ks:/est /x 7H4 2~9 3me AMO /H 7N4 474'M 3 **.
i M/NFGAbf a M/N/MWM N6/$NY #f Pddn 4 6 vfl M644M60
/N 7Nd AdA47dM V21:df. Oue/Np 7Nd 70g7*
FOAM 44YdL /3 THG W/PGA R*WNBAAY sF THd 4/dHfe = 4nWJr/NWows s?dNon/ pp 72een-/NMg Pf=4 W WHGA.5 40/44 kW/4 PdA crteN /$
j A R W e d. To et 4Mn44 4A VNAN d.t" (A M warsn m y),
..~r.,-
.- a
2.
PAQPOSEO UNK BETWEEN CGod ACCEP7ANCE CR/ rER/A, C006 WNCGATA/NTY AND RE/WLA70A y 4/H/TC d$SWMP7/oM: /*Af4A&/4/7Y b /STR/4Wr/GN FWNlr/GN /f A MEY AffMA7
/t WELL A&/AssGN710 8y &
NORM /4 O/frA/ fur /t:m/ fG4u.cr/ Ass).
bE7 h AA/ARZENT A )(gy AstyL7 4No lwn A* AGA/'t/Zf/fl& PADSA E/4/7Y s'NAr 95 &XfdeFA f I
bN / N P C.f 5 O A/N /7, a per Q, =.5rAHb4AO b fV/A7/ON l
,aa v.,rsi unamnn L
1
-F.
4, A
f L/M ~ hg iar i
rusu
?
- Vb)
Mn y'O) vs. f ts rseuzarab /N.srsNsss.s rex 7:
i Y bY,
/h kN$NN fNb em e,,,.*cus-4,
i1 AS fMM?779N Af9644/MS WNC547s+/N7y GF F&ol d&NAwoA,
Mobd1/N6 t/NtfA7AsN77 pp ZY171/1 THann44 Nybts. ruer, ANb (4NC647A/N/45 Attat/s74b tw/7N TH4 PWP
/N/7/Al AN4 29&NJAAt fpNb/7/oNZ 4ed /Nbsperve my 45 UAtN orHet. Hexes L
s
$*L4W7 Y
f J
4L Y*H l
{q,'-s'
-e*
G
=
/h*4H i*
kfL AYJ1* 7*. H WHfAE Q,.
/s F0WNo P9A A22M b //< Alt &d sA,< ass T *L
/t 19&Nd FAGM THE FWEL po
/ENAM*A l*d)6 Af 4C: MON 7 AAJ<4ff hjagy /C 79&M) /AJ/t ZYZ7?/7f d*444
.ew,N:e suAss<a n +,y
/N tvH/CH GN4y /4AN7 PAAAhe7&qf AAG VAA /db.
P29Av.Sfd
- f. pad ActG/7AN(G 4*A/74AloN :
47" b
ANb ft K n eie w.
//Arwree
/rmessoa F9A AN A Cfd?7'sE44 (90 & / 7" /t A OAU/AN NA s' bA lN YT$ ##
~
l'AS *I Mrs reme y,y, y4 misAsa, aresAusse or 7sse 7s<wer.res4e.
7*NG ARAf9N FoA Ade?7/Nd
/f 44/A4/4
,,, y MA.79.g/77 JP 7CJr C9HO/7/*N: ffA7kAd JL4/7A//A L
. ptWLAron: #5 fwd 4 A *J:.
dH4y JY!79MJ 76??? A A G 7'* f f C9N /Jdatta.
l
a PAST EXPERIENCE OBTAINED FROM THE LOCA CODE UNCERTAINTY STUDIES (IN THIS COUNTRY AND ABROAD) HAS INDICATED THAT THE GPCT OBEYS A NORMAL DISTRIBUTION.
~
HENCE, LIM 6) = LIM (GPCT) COULD BE TAKEN EQUAL TO THE CURRENT LICENSING LIMIT (1475*K) AND P >uM COULD BE o
CHOSEN, FOR EXAMPLE, SUCH THAT LIM 0) - BE D) = 2 BOT' GIVING P*>uM : 2.5%
THE VALUE OF BED) s GPCT COULD PERTAIN TO 3EST ESTIMATE EITHER TO A DESIGN BASIS LOCA (200: CL BREAK) OR TO A " SINGLE FAILURE CRITERION" SMALL BREAK LOCA THAT GIVES THE LARGEST VALUE OF GPCTaE, 'dHICHEVER IS THE l
LARGEST.
l l
l I
I
w-m
_.s l
+
b 2
3*1 9s s.
i w
bs g
m N
l g
3 0
s tty % y "I#
g
~
ag,n<a
=,=, 4
...a s
3*
tt' x go W
Q l.
ka 4 d 1
4442 RRe x
wL w 4
s.
gygg g
.U___
. w w
e 4
y
(
k l
y 4,.
g.
l s
y 4
L l
3 g
y>
u s
r s
t u
y z
=
w
\\s q.
q o
~
4 *...k N
g
,w g
4 7
%Q
. g;. #
5 4
O
~*
R l
I O'
i
/i DUE TO THE EXPONENTIAL NATURE OF DOX:GENCE OF GD0X AND OF H2) WITH RESPECT TO CLAD TEMPERATURE IT CAN BE SHOWN THAT GD0X AND H2 DO NOT OBEY NORMAL DISTRIBUTION AND, THEREFORE, COULD NOT BE DIRECTLY RELAits, THROUGH STANDARD DEVIATIONS, TO THEIR CURRENT REGULATORY LIMITS.
PhF I
e pf)g
(
/
j l L/
n.
L t
4 I
4 h ox L
i.,
TO SIMPLIFY THE MATTER IT IS PROPOSED THAT g Q) _
L/H (4)- 8E (<J )
nt (@)
WHERE * = GDOX OR H2, Lu (*) = CURRENT LICENSING LIMIT FOR +
BE D) = BEST ESTIMATE VALUE OF *, COMPUTED FOR THE DESIGN BASIS LOCA OR FOR THE SMALL BRE4K LOCA (WHICHEVER GIVES THE LARGEST VALUE OF BE D)
N D) = INTEGER C.>1) SPECIFIED BY NRC ON THE BASIS OF THE CODE UNCERTAINTY STUDY 2 E D) WOULD THEN SERVE AS THE ACCEPTABLE LIMIT ON THE ABSOLU VALUE OF TRE DIFFERENCE BETWEEl THE MEASURED AND THE CALCul.AT MAGNITUDE OF *, FOR 95% OF ALL EITRIES.
w-
- ~- - -.
w
i s
/f DUE TO THE EXPECTED LARGE VARIATION OF GDOX AND H2 WITH TEMPERATURE, IT IS RECOMMENDED THAT THE COMPARISONS BETWEEN THEIR CALCULATED AND THE " MEASURED" RESULTS BE DISPLAYED AS FOL'.0WS:
.L p' H
-,,a
&??
6 1em)
/
Y
(
l
.t
! '.Actdd b
m 4
i l
I l
l
tr IT APPEARS THAT ACCURACY OF THE :El RESULTS PERTAI? LING TO THE BREAK FLOW (MASS AND ENERGYL STEAM G 8ERATOR HEAT REMOVAL, AND THE MINIMUM FOAM LEVEL, CANNOT BE RELATED TO THE CURRENT LICENSING CRITERIA.
HENCE, IT IS PROPOSED THAT THE COMPARISONS OF CODE PREDICTI0tlS WITH TEST DATA, BE DISFLAYED AS SHOWN BELOW AND THE UPPER BOUND ON THE REU:TIVE ERROR BE SPECIFIED; FOR EXA?.PLE, 30%.
O
_m.-
1/1 MM. 1 4 4.:i]
6 l
?
i
, y p a.
V
Jo 1
THE MAIN PURPOSE OF DISPLAYING THE COMPARISONS OF THE CALCULATED VS. MEASURED VALUES OF LPCT AND D$ IS TO ESTABLISH WHETHER, AND TO WHAT DEGREE, DOES THE MULTIDIMENSIONAL OR DETAILED DISCRETIZATION CF REACTOR CORE ENHANCE PREDICTIVE CAPABILITY.
KNOWLEDGE OF THIS INFORMATION BECOMES MORE IMPORTANT WHEN PREDICTIONS OF THE CORE DAMAGE ARE SOUGHT.
WHILE BOTH THE OPDINATES AND THE ACCURACY ACCEPTANCE LE COULD BE THE SAME AS FOR GPCT AND GDOX, RESPECTIVELY, THE ABSCISSA SHOULD PERTAIN TO THE AXIAL LOCATION IN THE CGTC (7/Leagg
2/
k*
3 lb 3
5 T
l 1
g l
1 s
3 g3 s i.
as st se.
,w
,w
(
0 0/
0*L 05
- E
- 7. o
'S S
$. n,
.ts, i t-
't 4****
ou au l
7 T
T 7
y 7
f T
3 f
t' 3
1 BNTA/6f /N TW/f /AA/H AA6 R&#Rf144714 fy NWM&'AAls tv M/rH /H b /c A 7'd How HANY &sist)
AX/AL Nc6di 44Mt) WH445 MC44 /N THS dAttpurys,,y h. FfA &XAMPLd, /M Rve AX/AL 44V4Ls weAe of MMMsyth /N A CALCWLA7/*N QM A $/V4N
\\
7 4 Z 7 4*A 4 4,
7M8Z5 W/44 y/44 M/v6 RN7d/4$ /N IW4 41sV4' $.4AAy, AMM MNPAy 4A 44446 W/?W A N U M 4 A A 4 " f.*
js-LA444%
TNG AJ h/44 4 0cA 7/f y PdA ALL RN7A/41 /N TW6 4AA?M.
N /t TNG To744 N&n444 of A43/A L A /N4f (oA LA>PGAA L A /v/f/* H f ) M Z S 4 /N PMG cAlck ts 7/9N WH/ t.d TW& NWH9AAP*A, %, 443/Het,W g PAA7/CMLA X A/A/C foA AA 76A44 EsNG) /*A LeM/c4 YH4 RHrA y /s MA44.
MGNty, 7MGAi Av/44, 44 Z/X 24tW /*4A /N 2, t: 4 As l//, //, 2/s, //, L/j, 3/3 W/7W AN7Ard! F04 kaatoes
.a,xs w o oars are n oraNrsa.
Q. g S
b AdfX/4/AAY SAAM F#R. KKy A&Z WL 73 Le g g sus to barsome ansers os nexxxonsvr, unaarniner on.staanb m rxse.wcoune AeAbs8+t.
I
,/
,7
/
s aa
,,an,,.,
/
/
4/Mir
/ /
,m f
l' /
A
/
7
/
/
/
1
/?
)
y
'f
>7 t
l
- e
=
t I
a e
13 SELECTED RESULTS TO BE USED AS DIAGNOSTIC INDICATORS OF THE CODE'S CAPABILITY.
ACCEPTANCE CRITERIA NOT SPECIFIED. SUBJECTIVE JUDGMENT WILL BE USED TO DETERMINE WHETHER THE PHYSICAL PROCESSES ARE MODELED SUFFICIENTLY WELL S0 THAT AGREETITS IN KEY RESULTS ARE NOT FORTUITOUS.
l THE DIAGNOSTIC INDICATORS ARE THE MAIN VEHICLE OF CODE t
ASSESSMENT FOR'THE CASES INVOLVING THE SE?ARATE EFFECTS AND THE BASIC TESTS.
1 l
l l
l l
l
-4m-4
- + - -
a a a A-g A
b LOFT msasarsmanfs 1
O.
Ne
<-os, A..,om,w&
w
=
m et J
<<l x^
/*
Y ik'
%e
~
w c
N TM C
3 J
s g
w e
a Ah Y
J A
h Anw lap fluid Tem e new eda. prakurt.
a dans1.13 t
e---,+,--
---.---,-.,,,-,-_-.~r-n,,--,,-.,-,.a
. - -.. -. ~ -.. - - - - -.
u
/0/D FRACTION IN CELLS OF j
UWRAPPED DOWNCOMER se swarwa aarne 6
qg
~^)f V
1 l',
i 90 re se e
m m
b
.I
~
.p i
i i
re
'I9 4* *
?*
i Serrore cau.
}
r ma n
t u
Row conrM NIN6 Row conrAINIMi ZNTAcr war BAoKEt4 Loot ZHLar No33LE ZMLir M*1LLA
l M
RA dnortrive sasa, pan, 1
f*Lu/D MALs IN M Pf2R PLEN HH w
M tk s
p Q WITH Hot LES gg' t
i
\\
/H3 &CrisN R
' \\
/
i f
\\
/
-CoLb LEG Ett /Hyggytm s
s s
pHt,,y 1
I N
Au'-
5 1
TIM E 4,
ww t
w%
FLM/h M,4SS a
IN LOWGA PLEN uM N
4, e,---
~,,
.-.,..,c,
...,,,.-.,,wyy.,,.
i i
n 4
3 4
3 A
sa N
4 0
I T
a r
n-m l
'}
R 4
N-
'\\\\
\\\\
~
/
/
4
/
/
Q
/
/
/
~/
/
Q
/5
/
/
/
/
/
/
8
?
1 4,%
u
- E; k
~=*
A set Q
c w
6 s
i' l
Y
]
]
Y J
J-J J
J o,
=
6 4,.A s s,s. m.,,
e i
0
_'d.
- ?
j s
+
'$9 l
l JL 44 d,f s
f
4 1/
O 1t
='T A
l s _ _ _ z_.
- T
,i l
MAQ'll MEbl l
i *a ef CRITicA L FLod 5
V.
mrs
%I dL
\\
- Y l,
.-(
'; j.
I I
V
\\,I k
_,e s
e L.
MEA s tAREMENTS v
FoR door A S E =_SS M F MT~
n.
CODE ASSESSMENT PROCPl!RE A.
DEVF10PMENTAL ASSESSMENT PHASE PERFORMED DURING CODE DEVELOPMENT, T0 m DETERMINE CONSEQUENCE OF DEPARTURES FROM RECOMMENDED SYSTEM NODALIZATION y EVALUATE MODELS OF BASIC T-H PROCESSES AND DETERMINE UNCERTAINTY RANGE AND PROBABILITY DISTRIBUTION OF INPUT COEFFICIENTS y EVALUATE CODES' SCALE-UP CAPABILITY OF PREDICTED RESULTS, FROM THOSE IN TEST FACILITIES TO THOSE IN LWRs.
CODE MODIFICATI0fiS ARE ALLOWED DURI!!G THIS PHASE.
i SOME CONTRACTORS RE-RUN ALL COMPARISONS, WITHOUT CODE CHANGES, JUST PRIOR TO CODE RELEASE TO PUBLIC.
THOSE l
RESULTS SUPPLEPENT THE FINAL ASSESSMENT OF THE CODE.
l i
11 fB12AV47/rN$ spNdGAN/N4 41046.T ' fCAL/Ng 4fs9"/W 4 / L / 7 Y
/*
N64 t*
AVA/LAS/L/77 ff CGJ7A/N FW/4 f tJ94 4 JNPendMAf*/PN l
l l
$ U $4A42 7257 FA C/L/7/GC FWEL Ash!
NMtL&42:
LJF7~, NRM, /386 l
0.A R P/7fH i Ndtg1 D
l 24&t7A/c :
$4M/.tM42, PLACHQ T H M 74 729 4.#4.Zj 3
Fxx cc rF, serf i
fU54 $O8 NNd46A A :
NNN 4&N $7"H
&44c72/e :
FL MCH7*, 747A, 7H76 4 adC, P M 4, cc'7 6 ( C 7 F
.[3M/ZM 44/M95 f
$759M $4N.
Abfl, CC 7"F, p*LitWr-f2A.cA7 7"[] yygg b w d'or.:/),
Loprd) botvNcsMGg t4,07's l
l l
A&&d t-4f>*TF OGC/GN CHANF-GZ l
ff/PGA R &A/A gyyy d!!aer:, FoA. b / P F o n 4 N i' pwa vesent, couts se J.WR47'/0475d /N "MALLd<
t
. cAla 75.trl l
\\
if/9 J/AW/f/t.9N7" P4544dMS EW:7We c.evceAAI/N&
/-f.a 2-$ Flote /M /*/P/M?
AL.go Fon veAy CNM4 JA4Axt
14 k
m m
a a w' s,
4 s a w%
N
=
N g
^
Yk k
g 4
Es4 o
g C
Ng g(
N Q ksk b
\\
)
h k1
\\
k 4(n
,t e
2 m g
i e
Y y
=
{
}
' i..
g
.c
- . c =2 m
at
{ [y-4 h
1 P
r.-@a!
J s
1
. w by 2
N x
k 3
g q
o i
l a p gr n
Q%
~
u s 2
- e A
w s
I g
n$g c
4 F
... 1, x
w s
D k
l '.,i.,
L}
g 4
.mmm a '.
-6
\\J b
Ni k 4%
s
/
?
A.%
T k
$'I
/
k 4D I
3
( Si e
i n
3 m d' u
(
e
\\
y
=
5 i
CODE SENSITIVITY STUDIES (DONE BY CODE DEVELOPERS)
CODE SENSITIVITY DUE TO e MODELINGOPTIONS OF PHYSICAL PHENOMENA e GEOMETRICAL REPRESENTATION OF THE SYSTEM (NODAllZAT
- TIME-STEP CONTROL RESULTS:
~
e RECOMMENDED BEST ESTIMATE MODELS AND COEFFICIENTS e BEST ESTIMATE NODALIZATION OF LWR SYSTEMS AND OF TEST CONFIGURATIONS (LOFT, ETC)
= FROZEN TIME-STEP CONTROL
= DETERMINATION OF THOSE PARAMETERS WHICH HAVE A DOMINANT EFFECT ON Tile PEAK CLAD TEMPERATURE (PCT), AND WHICH CONTAIN UNCERTAINTIES TilESE PARAMETERS MAY BE PART OF CODE INPUT, OR MAY BE EMBEDDED IN VARIOUS CORRELATIONS USED BY Tile CODE.
(
E
A e
li B,
INDEPENDE?lT ASSESSidENT
> NOT PERFORMED BY THE CODE DEVELOPERS.
1.
MRC DEFINES CODE ASSESSMENT MATRIX FEATURING TEST
' CONDITIONS THAT SPAN A VARIETY OF LWR SYSTEM SIMULATO GE0ETRIC SCALES POSTULATED ACCIDENTS ECC INJECTION SCHEMES SYSTEM COMP 0MENTS BASIC T-H PROCESSES WITHIM COMPONENTS ALL IMPORTANT. INTEGRAL TESTS TO BE INCLUDED; OTHERS WITH ADEQUATE MEASURE NiiTS 2.
NRC DEFINES KEI RESULTS AND THE DIAGNOSTIC I'1DICATORS, l
AND SPECIFIES THE DISPLAY FORFf.TS 3,
NRC DE?INES CODE ACCE3TABILITY BOUNDS.
i
.r,
i 4.
CONTRACTORS PERFORM BLIND PREDICTIONS, WHEREVER POSSIBLE AND FORWARD RESULTS TO NRC BY AGREED
.UPON DEADLINE SINGLE-BLIND PREDICTIONS ARE MORE VALUABLE SINCE THEY EMFLOY ACTUAL INITIAL AND BOUNDARY CONDITIONS.
5.
CONTRACTORS DISPLAY CODE RESULTS (OF STEP 4) AGAINST TEST DATA, PER NRC SPECIFIED FORMATS 6.
CONTRACTORS PERFORM POST-TEST. ANALYSES TO M DETERMINE REASOf!S FOR DISAGREE?TJIT WITH TEST DATA y DtutRMINE CODE ERROR w DETERMINE MODELING SE!!SITIVITIES l
y Dt:.:tRMINE EFFECTS OF CODE USER'S INPUT OPTIONS 7.
CONTRACTOR PERFORM PREDICTIONS 0? LWR ACCIDEili CONSEQUENCES TO ASCERTAI!! REASONABLENESS OF CALCULATED TRENDS.
l-
~
.ta At/44&M ff NRW t*/bd VSAS/oNS
/.
SCNdbHLeb
/$ 240$
of NdW Cedd V222/sNf
/ PD2 - Ph 3 MC-PM sasur / yes.e sisar
~ PF.i
- PF.2 2-k N CC N eb HL&d Cod 2 V d a s/G N !
h/ftoVeAY #F /MPcArnNr dab /NS BAAoAf YHA7 MA Y PREV 5Nr dJ3E Nf6 /N CONG AGQW/ Add s/79Ar/*Nf y
b/stav44y #F /MPoATANr toJ # beA=/cseNoef P.-fA7 M4y PAGVEN7 CCb& Nf3 /N d&C/GN
/F AX)&A/NdN7C ISCWes fr New doAn Vaes/oNs teau/ Ads c2 ara /N AN/WNr $P A2= //lC2447/MV QF PR m V/rts.t4y PGAFsAndb CsM/AA/ZoNs :
y COMMA / son W/ro sN2 re:r,=AsM EncH /Nrt$,444 PA' d/4/77 PRit V/okt4 y C#//Z/bdadd.
.it A2 -CAL e2LA7/cN CF 74056 cAZGt WH/cN SAV2 P90A doMPAA/SoM w/7W 44rA
m m
-,m4 a
--2 A_2.
_m.
d a
_a
_-.-_am,
_a
-.+
19 9
]
4 TX'/ C ASSESSMENr (2Wwn)
.ospesur. emnsrion 3r g f> $
1TX
- Anef, 9
I I
4 I
b I [s 8 E S
E N
s s
4 s
s g*
h 4
D N
y A
e s
s v
f 4
4 k.
M
= K*
+
e.
~
A e
s s
s N
g U
kf k
,k,h k, 4
l 4
N
/
\\
M M
4 qk g
%g 4
4 g
( A 1
t, k/ 5 A
D k
A l
s sg s
y s
s s
IR Ui k ;i ti
ga C.
CODE UNCERTAINTY STUDIES CODE UNCERTAINTY STUDIES ARE USEFUL FOR
- 1. 'PRIORITIZING EFFECTS OF VARIOUS MODELS, COEFFICIENTS AND OTHER CODE INPUT ON THE KEY RESULTS, TO POINT OUT WHERE THE FUTURE CODE IMPROVEMENT WOULD BE MOST COST EFFECTIVE.
2.
DETERMINING THE EFFECTS OF THE INPUT UNCERTAINTIES THAT ARE NOT RELATED TO SYSTEM THERMAL-HYDRAULICS (E.G. PLa.NT CONDITION AND FUEL BEHAVIOR) 3.
OBTAINING A ROUGH IDEA 0F THE PROBABILITY DISTRIBUTION FUNCTION FOR THE COMPUTED KEY RESULTS.
4 CASE & PARAMETER SELECTION, BY NRC, FOR CODE UNCERTAINTY STUDY TO BE SELECTED AND/OR DEFINED:
- LWR PLANT, ACCIDENT CONDITION, DEGREE OF AVAILABILITY OF ENGINEERED SAFEGUARDS, AVAILABILITY OF OFF-SITE POWER, TIME OF LIFE, ETC.
- PARAMETERS TO BE VARIED (AMONG CODE INPUT PARAMETERS AND EMBEDDED COEFFICIENTS)
- RANGE OF PARAMETER VARIATIONS
- PROBABILITY DISTRIBUTION FOR EACH VARIABLE PARAMETER i
SOURCE OF INFORMATION: ' TEST DATA, CODE SENSITIVITY STUDIES, ENGINEERING JUDGEMENT l
t L..
i UNCERTAINTY STUDY FOR CALCULATED PEAK CLAD TEMPERATURE
- 2. ILLUSTRATIONS plX )
L F (X,X,... X,,I 2
3 2
sX 2
pix,,)
(
X,,
- L I'lPC11 CGHHECTED foA l'eht 5444 i
=
oe to ecT
1 9
CODE ERROR COMBINED EFFECTS DUE TO:
e MODELING SIMPLIFICATIONS AND TIME-SPACE AVERAGING e NUMElllCAL SOLUTION (DELETION OF lilGilER ORDER TERMS, TRUNCATION ERROftS, LOOSE CONVERGENCE CRITERI A, SPATI AL RESOLUTION, LINEAHlZATION) e SIMPLIFICATIONS IN DESCRIPTIGN OF MATERIALS PROPERTIES AND EOUATIONS OF STATE e SIMPLIFICATIONS IN CONSTITUTIVE EOUATIONS CONCERNING EXCHANGE OF MASS, MOMENTUM AND ENEllGY ACROSS LIOUID/ VAPOR INTERFACES:
ALSO FOR WALL FRICTION AND llEAT TilANSFER
- INABILITY TO CONSIDER PilENOMENA OF STOCilASTIC NATURE l
- /. '.
.~. / * :
~
n
-10%,./,*.*:*.',/
- :..~.. *E.k*,
- II 1
EXAMPLE S
t L
- /..
/*.
l PEAK CLAD o
l EN
- 7'. ' "
~
l TEMPERATURE j
g
- f.,.,/ *
- a0%
/> /
m
- E
/
l/
\\
BEST ESTIM ATE PREDICTION (*C) i
s4 OUR CURRENT PLANS DO NOT INCLUDE THE EFFORT IN SEPAR THE CODE ERROR EFFECTS FROM THE CJDE INPUT UNCERTAINTIE EFFECTS, MAINLY BECAUSE THIS fl0W APPEARS TO BE EXTREMELY DIFFICULT, IF NOT IMPOSSIBLE.
CONSEQUENTLY, OUR CURRENT APPROACH IS TO DEDUCE THE TOTAL CODE RESULTS UNCERTAINTY DISTRIBUTIONS FROM THE PLOTS OF' (*p - o ) VS TEST FACILITY SCALE.
g THIS REQUIRES SUFFICIENT NUMBER OF DATA ENTRIES TO ALLOW FOR THE STATISTICALLY SIGNIFICANT SAMPLING.
sr THE BASIC ASSUMPTION IS THAT THE PROBABILITY DISTRIBUTION FUNCTION, OBTAINED IN THAT MANNER, FOR THE DEVIATION OF THE KEY RESULTS FROM THE COMPUTED BEST ESTIMATE VALUES, REi'AINS THE SAME FOR ALL ACCIDENT SCENARIOS INVOLVING LOCA.
CODE RESULTS UllCERTAINTIES FOR NON-LOCA ACCIDENTS AND TRANSIENTS OUGHT TO BE SMALLER SINCE THE CODE WOULD BE LES SEVERELY CHAIIclGED.
l l
l l
l
.n SUffARY 1.
CODE ASSESSMENT STRATEGY IS DEFINED TO STRESS THE IMPORTAtlT RESULTS THAT ARE USERil TO THE REGULATORY PROCESS LINKAGE WITH THE CURRENT LICENSING CRITERIA SUFFICIENT BACK-UP IflFORMATION TO ASCERTAlff THAT THE CODE ADEQUATELY MODELS THE IMPORTANT PHYSICAL PROCESSES AND, THERE3Y, ASSURES NRAPOLGILITY TO LARGE SCALE LWRs.
2.
WORLD-WIDE TEST DATA BASE IS EMPLOYED 3.
E(PERT STAFF FROM A MUMBER OF NATIONAL LABORATORIES ARE UTILI7m1 TO ASSESS THE IMPACT OF THE CODE USER ON THE CALCULATED RESULTS.
THE CODE IS ALSO BEING USED BY FOREIGN E(PERTS, TdROUGH THE INTERNATIONAL STANDARD PROBLEM E(ERCISE TO FURTHER HELo IN THIS Dt.itRMINAT!0N.
!