FVY-85-122, Forwards Responses to NRC 850103 Questions Re Unique Feature of RELAP5YA for BWR LOCA Analyses
| ML20141E622 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee, Yankee Rowe, Maine Yankee, 05000000 |
| Issue date: | 12/31/1985 |
| From: | Capstick R YANKEE ATOMIC ELECTRIC CO. |
| To: | Muller D Office of Nuclear Reactor Regulation |
| References | |
| FVY-85-122, FVY-85-139, FYR-85-139, NMY-85-206, NUDOCS 8601080229 | |
| Download: ML20141E622 (112) | |
Text
{{#Wiki_filter:epho"c (6 ") *' oo
-YANKEE ATOMIC ELECTRIC COMPANY TWX i 10-380-7619 j ' ,4 1671 Worcester Road, Framingham, Massachusetts 01701
.//iJKEip'
~'
FVY 85-122 NMY 85-206 December 31, 1985 FYR 85-139 United States Nuclear Regulatory Commission Washington, DC 20555 Attention: Mr. Daniel R. Muller, Director BWR Project Directorate No. 2 Division of BWR Licensing
References:
(a) License No. DPR-3 (Docket No. 50-29) (b) License No. DPR-28 (Docket No. 50-271) (c) License No. DPR-36 (Docket No. 50-309) (d) Letter, YAEC to NRC, FVY 83-4, dated January 14, 1985 (e) Letter, USNRC to MYAPCO, NMY 84-91, dated May 11, 1984 (f) Letter, USNRC to YAEC, dated January 3,1985 , (g) Letter, MYAPCO to USNRC, dated January 31, 1985 (h) Letter, YAEC to USNRC, FVY 85-18, dated March 1, 1985 (i) Letter, YAEC to.USNRC, FYR 85-48, dated April 30, 1985 (j) Letter, YAEC to USNRC, FYR 85-72, dated July 1, 1985 (k) Letter, YAEC to USNRC, FYR 85-87, dated August 15, 1985 (1) Letter, YAEC to USNRC, FYR 85-121, dated November 1, 1985 (m) Letter, VYNPC to USNRC, FVY 85-98, dated October 22, 1985
Subject:
Response to Additional NRC Questions on the RELAP5YA Computer Code
Dear Sir:
By letter, dated January 14, 1983 [ Reference (d)], Yankee Atomic Electric Company (YAEC) submitted RELAP5YA, a computer program for Light Water Reactor System Thermal-llydraulic Analysis, to NRC for review and licensing approval. A set of 197 questions concerning the use of RELAPSYA for LWR LOCA analyses was received from the United States Nuclear Regulatory Commission (USNRC) in May 1984 [ Reference (e)]. YAEC provided responses to these 197 questions in a series of submittals [ References (g) though (1)]. An additional set of 39 questions concerning unique features for BWR LOCA analyses was received from the USNRC in January 1985 [ Reference (f)]. The attachment to this letter responds to each of these additional 39 BWR LOCA-related questions. Sections II and III respond to questions on the development and assessment of the jet pump model in RELAPSYA. Section IV addresses questions on the RELAPSYA assessment against the TLTA tests. Section V responds to additional NRC questions concerning the use of RELAP5YA for BWR LOCA Evaluation Model Analyses. AD J. AN turr (! t r un ty . gn pl AW) LB (MALLARD) P58 (L. HULMAN) KIC+;n (RUSA) E!C55 (SNINIVASAN) PM (GAMMILL) uss ( ACTING) RSH (RFRLINCtH) FOR (VA%SALLO)
- u. WT ritLD (ltr ly) "
g$ La (W. JOHNSTON) ass ( m =As' 8601080229 851231
,i stess (r m an PDR ADOCK 05000029 PDR ros (n. ntc u ) p L i
1 i United States Nuclear Regulatory Commission December 31, 1985 Atten, tion: Mr. Daniel R. Muller, Director Page 2 As stated in Vermont Yankee's (VY) letter of October 22, 1985 [ Reference (m)], the development and licensing of the RELAPSYA LOCA Code is one of high importance and one in which NRC's review schedule has direct impact upon the planned application of this method for VY's Cycle 14 high energy bundle reload. We trust the attached information is satisfactory; however, should you have additional questions or require further information, please contact us. Very truly yours, i YANKEE ATOMIC ELECTRIC COMPANY [ Ro'oert W. Capstick Licensing Engineer Vermont Yankee Project RWC/dps Attachment i
YAEC RESPONSE TO ADDITIONAL NRC OUESTIONS ON RELAP5YA APPLICATION TO BWR LOCA ANALYSES I. INTRODUCTION The RELAP5YA computer program was developed and assessed by Yankee Atomic Electric Company for analyses of PWR small break LOCAs and the entire break spectrum of BWR LOCAs (References I-1, I-2 and I-3). Documentation for the RELAP5YA computer program was submitted to the USNRC for review on January 14, 1983 (Reference I-4). A set of 197 questions concerning the use of RELAPSYA for LWR LOCA analyses was received from the USNRC in May 1984 (Reference I-5). YAEC has provided responses to these 197 questions in a series of five submittals (References I-6 through I-10). An additional set of 39 questions concerning unique features for BWR LOCA analyses was received from the USNRC in January 1985 (Reference I-ll). This submittal responds to these additional 39 BWR_LOCA-related questions. Sections II and III respond to questions on the deve'lopment and assessment of the jet pump model in RELAP5YA. Section IV addresses questions on the RELAPSYA assessment against the TLTA tests. Section V responds to additional NRC questions concerning the use of RELAP5YA for BWR LOCA evaluation model analyses. In all sections that follow, each NRC question is identified by the NRC number. To facilitate review, each YAEC response is preceded by the same identification number. Likewise, the References Tables and Figure numbers retain the identification number for the associated question. References (I-1) R. T. Fernandez, R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, 1" L. Schor, R. C. Harvey and'R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume I: Code Description," Yankee Atomic Electric Company Report YAEC-1300P, Volume I (October 1982). (proprietary) w-. }
(I-2) R. T. Fernandez, R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, L. Schor, R. C. Harvey and R. Habert "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume II: User's Manual," Yankee Atomic Electric Company Report YAEC-1300P, Volume II (October 1982). (proprietary) (I-3) R. T. Fernandez, R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, L. Schor, R. C. Harvey and R. Habert, "RELAP5YA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume III: Code Assessment," Yankee Atomic Electric Company Report YAEC-1300P, Volume III (October 1982). (proprietary) (I-4) Letter, YAEC to USNRC, FVY 83-4, dated January 14, 1983. (I-5) Letter, USNRC to MYAPCo, NMY 84-91, dated May 11, 1984. (I-6) Letter, YAEC to USNRC, 2.C.2.1, FYR 85-22, FVY 85-18, MN 85-46, dated March 1, 1985. (I-7) Letter, YAEC to USNRC, 2.C.2.1, FYR 85-48, dated April 30, 1985. (I-8) Letter, YAEC to USNRC, 2.C.2.1, FYR 85-72, dated July 1, 1985. (I-9) Letter, YAEC to USNRC, 2.C.2.1, FYR 85-87, dated August 15, 1985. (I-10) Letter, YAEC to USNRC, 2.C.2.1, FYR 85-121, dated November 1, 1985. (I-11) Letter USNRC to YAEC, dated January 3, 1985.
, , _ . - , _ _ - _ . . - . - - . . _ _ , . , - _ _ - - . , , - . . . _ _ - _ . - . , . . _ . . - ,. _ _ _ _ ..x.,
II. JET PUMP MODEL Page 127 of Reference 10 states: "For this model it is assumed that at planes N and S the velocity of the drive stream is equal to that at the drive nozzle exit, and the suction steam velocity is based upon the horizontal projection of the area of the frustrum of a cone indicated as S on Figure 3.3-2." This would appear to give a suction stream mass flow rate of ' fggg V A cos d because V is based on the horizontal projection of 3 the area, A , as described above. g 0.II.1 Please clarify why all the terms using Ag in Table 3.3-4 of Reference 10 should not be replaced with A cos g d because the table (whose terms are the basis for the jet pump model development) was derived assuming m=fgVgAs rather than fgV 3Ascos d*/ A.II.1 In the jet pump model development, the flow area of the suction junction was assumed to be the frustrum of a cone as discussed in Reference II.1.1. However, during model implementation and assessment, this ares was taken as the horizontal projection of the area of the frustrum of a cone. Due to this variation, cos d was input as unity and no change was needed to Table 3.3-4. During the review of the jet pump model, it was determined that the area of the suction junction is better represented by the area of the frustrum of a cone, as discussed in Reference II.1-1. For the 1/6-scale jet pump geometry, this area is 0.008239 ft . The loss coefficients are detemined by the methods outlined in answer to Question II.9. Foe the 1/6-scale jet pump, the geometric parameters used in calculating the fom loss coefficients are: D = 1.143 inch D2 = 1.154 inch
a D3 = 0.566 inch t D4 = 1.196 inch
,DS= 0.126 inch D = 1.3004 sinch '
(d = . nch \,,
/
L3 ,= 2.337 inch - These parameters can be obtained from Figure 1 of Reference II.1.2. The numerical values of the loss cy,efficients are listed in Thule 11.1.1 and Table; II.1.2. It should be noted that the parameter FDK6 is the cosine of angle d shown in Figure II.9.1 for the 1/6-scale model jet pump. '
,The modified jet pump model was reassessed against the 1/6-scale etecdy-state tests and Blowdown Test 1. The ascessment results for the steady-state tests are shown in Figures II.1-1 through II.1-1 and show good agreement with the test data. The RELAP5Y4 calculations for Blowdown Test $ were very similar l'o those shown ~in Figures 2.3-16 through 2.'3-20 of Reference II.1-3. Therefore, these results are not included in this document.
x 5
- s.
- e )
mye
*b.
TABLE II.1.1 Forward Drive Loss Coefficients for Jet Pump Model Form Loss Coefficient Variable Value Comments KAS FDK1 0.040 Suction junction forward loss coefficient. Kpg FDK2 0.080 Drive junction forward loss coefficient. , (1-KSA) FDK3 0.92 (1-FDK4). KSA FDK4 0.08 Suction junction reverse loss coefficient. KSH FDK5 0.64 Empirical coefficient. Cos e FDK6 0.9075 Cosine of angle d shown in Figure II.1.1.
.- FDK7 0.0 Not used.
[' ' /, s ' TABLE II.1.2 j ' i ;i , ReverseDriveLbsgCoefficientsforJetPumpModel Form Loss f Coefficien,t Variable Value + ) Comments KTA RDK1 0.7964 . Suction junction reverso
' loss doefficie.nt.
.g.
KND RDK2r' O.1795 Drivs junctiom reverse loss coefficient. KTN RDK3 1.0 Empirical coefficient. (. . KAT RDK4 0.494 Suction junction forward loss coefficient. KAN .t ) RDK5 1.1 Er.pirical c'oefficient.
~
\
-~~
RDK6 0.0 Not used in calculations. (
- t RTK 7 0.0 Not used in calculations.
I'y
\
f, j \'
;l
?
.j?
l l
)
i fd a
STERDY STRTE MODEL ASSESMENT v: tilllllla 164ll16lll11111lllll4lllll111 litilllll 1111111ll lilllillllIllililli lilli II ] - o _ RELAP5YA - m _ g- - . g - o __
.r - -
N h-1 _
~
p o _ _
'i' -
- + -
g< - y u - m __ _ q- --
+ - -
+ -
._ = -
I -
+ + + -
T + T _ + ++ -
+y -
)
111lll1 111111111 lIl111111 1111lIl11 lillllitt 111111111 11111llll lllll11ll lifilllII
-5.0 -1.5 -1.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 M
Figure II. 1.1 Positive Drive; Normalized I'ressure Difference versus M (Suction Plenum to Throat)
STERDY STATE MODEL RSSESMENT o , . . . Jilllilli 11111ll11 611111111 111111lll 11i111111 1416i1111 11111111: 11 11111 l11111i1 _
~
A e ~ a- - RELAPSYA p - _ +, -- - a - - n ;- - 2: __ 3 _ o - t # o
~
~
m % ;- - 4
- g -
~
Io e ~
+
d- + - g _ *
+
+
d-_- 4 - _ + _ d- - f- -
+ -
o-
+ -
4 d 4 f i l i l l t 111111111 tillitill l' ':llil!Illi lilllllil !!!!! fill tillitill lilllll 4
-I.5 -3.0 -I. 5 -2.0
-5.0 -4'. 5 -4.0 1.0 0.0
~ -I.5 -N. 5 d.5 1.0 1'. 5 2.0 2.5 3.0 I.5 4.0 M Figure II. 1.2 Positive Drive; Normalized Pressure Difference versus M (Drive Line.'Lo Throat)
8 11 ' l ' I ' l ' I ' l ' I ' l ' l ' I ' l i l ' l i l ' I ' l i l i l ' I ' I ' I ' I 'l I I ' l i l ' 181 ' l i l i l ' I '!!1 l i l 'l4i l i _ + _ __ o, _ a e _ 25 - _ gg _
=
_ 8, ; _ u
+ -
s I.m
- 2 .:
9: om c. a _ 8, e g _
- _ ~
1 _ _ g
+ -
- o
! Tlilil i i i ilililifililililililililililililililililihiililililifi ' 8 8 8 8 8 8 8 8* 8 \ 5 a a l l i N ( .
l I I l I $ $ I I I j 1 j I I il44i11il
=
Iiiiiieiil46ii4444ljeil46464ijiie446Iill64ei11 ligiil1644ij ntn
- J
= 2
" -"~
o 7 r: _1
= d
= 2-2 .o wg b
- -, 5 E
I -
- 3
~
E
- : ,.7 s
- ,f e u
- 4 o @
7 b-l~d -e u en M wa iti : sn __ _- o C - - cu.c
- : o 4 d" r 7 ri'
*H 5 : I so g I 2
*"d M oe u=
< + -
. occ u o C
- w s =c~
m - o, I oc 0 c-w N
- 4 _
u-c
).e I a gg O
% _ o "n w
- gu
,. y. uu 6
- : o3
- - :z; e.n M I
- v
- a. >
- : P t
- : c 2 E g
- : o* -
T --" "
~y
- : oc
_5 E E o
- : T 2 : o
[ Y I o Tilllllll liilfillrlIlliffill i l l ' ,I l l t ! lilllllll I!!!!!!!! IliffillT $ i . i . , .
. 1 , n 0*/ S*9 0*S S*S 0 *S S4 04 S *C 0*C S*C 0 *C S*! 0*I S*O 0*D N
g Act h / (T.I.d Yd ) l _lo_
i i i e i i i i o di+' irti sicilisis seliitels i i i li s t e r n
- + 'iiissasil: liisiii I b
- o
~
r
}. .
T ~a I.
~_'
g - o
-a
~ ~-
T
- - + 1 I ~_
+ --
o
-f
+ $ h m : e
- Q o w
+ a -
.: =
A sa i
- 4 e -
o H : y
-k.
- ++ o o w
._. + _-
N i w U wn
,a oec o Ww Ai e,$15 d
- ,m g eo
~_ ~
p GM
- cG W -
o o 9 _
, u .o .c C : - '
5 % ~3 ww e t.a. - > e >* I
- o. Ew wc
-._ y g Ow g.) Y ..
- 1 e
_ o. 3u o
- 3
- . o u
_ f E, 2 -' z o-T
~
2 I o, 2 7
- +
I g 7 etvissileitsisitt iesis! !tlt ivisest i 'ti tt !I !tttl l tit o.!
. . i ,
i i i i i i , s 0 *I S*0 0*0 S *0- 0
- t- S *I- OT- S *C- 0*E- S*E- 0 *t- S *t- 0*S-N Ad 31/ (TJ,d Od )
Z
. . . , , , o d iiiii6i11li: 1i iill tillistil listilisi giti+ ie ri
- + =
=
o E
+ k 4 T -d
+% -
I I o I
=
_ o 7
- . o*
I < - in _= _.o n, _ i 2 _ _- 35u- - w g
--- _ o y'
.g
= = Z
= -
m
- I a 2 o* a
~
2 ~7 c$> b b
- H 2
-g ]
C o M e 8 0 "'"8 _ ww so
-- _ .o.co o* b >
2 ** u
.i e
- ec o
=_.o.
.i
=
o*
~n i = = = - z o. =
- .T
; o f
I E
= = o 9 flil!!!!!
I!lill!!f IllifIfIl
. I f l t i, l f f I !!f flI f f F d, 0*O S*0- 0*I- S *I- 0*E- S*E- 0*C- S*C- 0*t- S *)- 0*S-N
_ ;$171'I'I'I'I'I'I'I'I'I'I'I'I'I'I'I'I'I" -
'I TP P P F F FF F FFFF I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I ' I' I ' I ' I ' I '
-2.00 E--
3,00 ji-O Positive Drive Data
+ Negative Drive Data
-4.00 E t
RELAPSYA ]
-S.00 [-- ~-
-6.00 [+- -
=
-7.00 5- :3
-8.00 4 T
~
=
-9.00 b
-10.00 5 +
O i
- 3 T
-11.00 h O 5 E. -12.00 [--
-13.00 :-
'l
~
'a3
-14.00 :- "d T
hI,0
-15.00
-16.00 g-h ' 'E ,
-17.00 g- + 0 ]= -
& -18.00 E-- # * *
$ -19.00 5- =
g -20.00 5- ]r3 - a -21.00 E- + E j n.
-22.00 g-
+ ]E -
-23.00. -
-24.00 b ~
-26.00 E+ O : -5
-2ti . 00 z- ~
=
-27.00
-23.00 :-
T OT
, llllIlIll llllllll1ll I!!!!!!!IllI!!!!!!!!!!ll111!!!!!Il}lbdll!!!!!!!!!!!!!!!!!lI13IlllllIlllIIllllllllllIl'T , ,
-22.00 -16.00 -10.00 -4.00 2.00 8.00 13.00 19 00 25.00 31.00 37.00 IsovfVTf T (E*i } '
Figure II. 1.7 Dif fuser Pressure Drop versus Throat Dynamic Ilead
References (11.1.1) Fernandez, R. T., R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, L. Schor, R. C. Harvey and R. Habert, "RELAP5YA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume I: Code Description" Yankee Atomic Electric
-Company Report YAEC-1300P Volume I (October 1982). (proprietary)
(1I.1.2) Wilson, G. E., "INEL One-Sixth Scale Jet Pump Data Analysis," EGG-CAAD-5357; February 1981. (II.1.3) Fernandez, R. T., R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, L. Schor, R. C. Harvey and R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume III: Code Assessment" Yankee Atomic Electric Company Report YAEC-1300P Volume III (October 1982). (proprietary) 0.11.2 What is the basis for determining K " 9"* " *~ SH Reference 10? What is the magnitude? A.II.2 The parameter K " * "* ** " 8" " " *
- SH method for calculating this parameter is outlined in the answer to
~ Question II.9.
Q.II.3 How is the magnitude of K d termined for use in Equation 3.3-53 of TA Reference 107 A.II.3 The parameter K is defined as RDK1 in subroutine JETPMP. The method for calculating this parameter is outlined in the answer to Question II.9. 0 II.4 How is the magnitude of K determined for use in Equation 3.3-55 of TN Reference 10? A.II.4 The parameter K TN s Mned as RDK3 in subroudne MM. De method of calculating this parameter is outlined in the answer to Question II.9. Q.II.5 Please clarify the following concerning the use of the two-phase density from Page 142 of Reference 10: "The density in the relative velocity head term ( (V,43 [/2] is Mned such nat normalization with the drive or suction junction velocity (H, or Hg ) yields the results shown in Table 3.3-4 by (HR "N' "" (H R S* A.II.5 To clarify the meaning of f in the relative velocity head terms, equations for momentum source terms and the form loss coefficients were derived without making any assumptions about the densities. These relations are listed in Table 11.5.1, where the nomenclature of Reference 1I.5.1 is maintained. In implementing the jet pump model, it was assumed that all density ratios are unity, hence the need for defining p did not arise. This is a reasonabic assumption, as demonstrated by the use of this jet pump model in RELAP5YA simulations of large and small break LOCAs in the TLTA tests. s e s s o h L m r o 7 2 T T F P g- p ) d n M P T 1 , h a E g / J + S K f g E + ( n N I i T K 1 x
, i U N i O y 2 A K, M R /
B - 3 1 U A m S 2 ( - u N s S p S ) t I C ) n S 1
) M p e /
y 2 + / E M 1 m I T y ( p o T I C ) ( M T
- A A + M N p y )
r A U + o M C o Q o T C y A c, f A 4 / N + M A A s - . T" C y ( n . t p O
'M- '
T C o 2,- , ( + ) M i - 9 , J + t T n C, / t a g ,g K A T T ( ( u A K - " ( K - + q = A. =
) )
=
E =
, = =
) ) ) ) , ) )
U U J U , 1 p p T M 9 K K 0 J M 0 T T T + C T T 9
. D S D S + 1 - D S J J J J (
J J S. ( P ( P ( ( 1 g ( f I J J ( 2 C J J T E T E r F F F I M M = = = M M U 0 R R R . e e 9 O O C, y 2 O
- f 9 F F C C F T f
l b a T 0 M I < l M -
- T E C Vt t E I
OR RI I DFD P O 5 I V R I V E 1-O'
m
.l.
v mv D
?.
c m -
~ 4 a
c n n
<l< .
sp.. m U <l< v g U. v
.h.
n* "
.l.
."I .q U
M. d E
"i v
.= l ."
t V i v " E ~
- g
' m m
- , n fle
.5.
- <l<* .- v r - D c
ol v
- l v nE ~b ~
nE o a n . E s e w w w . E e -
. . v .
v w
= 1 =
m u s = = u u B E E E 9 E E a n a : - a n
;a t
; x : 5 ;
t t . E t E E E E ! E E ei E U3E w
?,
E
; I' OI w w 0 w E . $ Ei. Q h O ti M re M >W EWD<HM>W G 8C M >W l
l 1 1
References (I1.5.1) Fernandez, R. T. , R. K. Sundaram, J. Ghaus, A. Husain, J. N. Loomis. L. Schor R. C. Harvey and R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume I: Code Description," Yankee Atomic Electric Company Report YAEC-1300P, Volume I (October 1982). (Proprietary) O.II.6 Should Equation 3.3-93 of Reference 10 have an M rather than a 2 in the denominator of the second term on the right side? A.II.6 In Equation 3.3-93 of the cited reference, 2 should be replaced by M. This was a typographical error. Subroutine JETPMP of RELAP5YA contains the correct form of this equation. O.II.7 What is the meaning of the f term used in Equation 3.3-84 of Reference 10 in the discussion on Page 148 of Reference 10, and in Table 3.3-4 of Reference 107 A.II.7 See the answer to Question II.S. O.II.8 In Equation C-2 in Table 3.3-2 of Reference 10, should the minus sign on the Q P term be a positive sign instead? TT
A.II.8 The term Q P in Equation C-2 of Table 3.3-2 of Reference II.8.1 T should have a positive rather than a negative sign. This was a typographical error. Subroutine JETPNP in RELAP5YA has the correct positive sign for this term. References (II.8.1) Fernandez, R. T., R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis, L. Schor, R. C. Harvey and R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume I: Code Description," Yankee Atomic Electric Company Report YAEC-1300P, Volume I (October 1982). (Proprietary) 0.II.9 How are the magnitudes determined for the Ks used for the throat region when flows merge or split (sometimes with turning) for the six cases presented in Table 3.3-3 of Reference 10? A.II.9 The procedures for calculating the form loss coefficients for use in the RELAPSYA jet pump model are outlined in the following sections. The geometric parameters used in calculating these coefficients are shown in Figure II.9.1. (a) FDK1: Form loss coefficient for suction junction used for calculating mechanical energy losses in the suction junction during normal operating conditions. The magnitude of this parameter is determined from Diagram 3-3 of Reference II.9.1, where: y = ( D . 3,) j/g (see Figure II.9.1) k* 4
- DS >
4 w s, r i 4 L L
', G D 7F M KD 3 1 d p6 * ~ 'r-d N if I 1 7 FIGURE II. 9.1
-w
Given r and D , h e ss coe W eients based on G roat flow area ($) can be determined from table in Diagram 3-3 of Reference II.9.1. From $ , FDK1 is determined as: 2 FDkt = 3 Ar ) Where A g = Flow area of suction junction A = Flow area of throat T 7ID /4 (b) FDK2: Form loss coefficient for drive junction used for calculating the mechanical energy losses in the drive nozzle and the 180 bend under positive drive flow conditions. This loss coefficient is obtained as follows: (1) calculate the loss coefficient for the nozzle using Formula I given on page A-26 of Reference II.9.2, where : g -i T"" r ( b - B3 )/2 4 "I (from Figure II.9.2) 2 L Lg g
#= h/%
H e.n c. e. j
-8 ( bg - b3 )/2. ~l 3 X8 = 0. 8 Se.n Ta n I I- 124 1
4 J L J
-> bsstol on no n tit. sult vtictiVy (2) calculate the loss coefficient for the 180 bend fron Diagram 6-4 of Reference II.9.1.
(3) For the 1/6-scale jet pump, the loss coefficient for the drive junction is estimated as 80% of the sum of the coefficients calculated in Steps (1) and (2). This is an acceptable approximation because each of the above two coefficients apply to fully developed flow, while for the Relative Elevation (inches) 34.37 - s[:
/
"c Drive 32.36 - :
Suction DPs Bhffles
' D[E 29.76 _
g (typ. 4) DPR 3 DPg DPg I Cie e * '
- 18.74 DPE 4
0.0 - ........ ....... _ _ O
- 0.512 8.819 - -
D PE Discharge Region Diameter Drive Nozzle 0.566 in Drive line 1.260 " Throat 1.433 " Discharge 2.833 " Figure II. 9.2 Jet Pump Test Vessel Schematic i e
-?2-1
jet pump, the short distance (<one diameter) between the 180* bend and the drive nozzle does not allow the flow to develop before entering the nozzle. Hence, the total mechanical energy loss coefficient should be less than the sum of the two coefficients calculated in Steps (1) and (2). The factor of 0.8 (80%) was chosen by trial and error based on the 1/6-scale jet pump data. (c) FDK3: This parameter is defined as: FDK4 = 1 - FDK3 It was introduced in the model only for algebraic simplicity. (d) FDK4: Form loss coefficient used for calculating mechanical energy losses in the suction junction under forward drive and reverse suction flow condition. The magnitude of this parameter is determined from Diagram 5-7 of Reference 11.9.1 where: a F = Tf D, /4 1 F4 = Tih/h fg i3 fv.,n R$un. E 1. f (e) FDKS: This coefficient is used for calculating mechanical energy losses due to = hear between the drive and suction streams. The magnitude of this parameter was determined from the 1/6-scale jet pump data as follows. For positive drive and negative suction flow (M<-1), the pressure difference between Station A and T, in Figure II.9.2 can be written as: P-P. 4 r 1 f 834 + ksu (f/et)( As/A=) - i 3 'I 4H3 (H*i 0 i J Where:
= Total pressure at Station i Hg = Dynamic head at Station i A and A, = SucMon and drive juncuon areas g
K = FDK4 3 Using the definitions given in Table 3.3-1 in Reference 11.9.3, the above equation can be written in terms of static pressure difference as: P-P 3 r, *- L N34
- kJ ('/")(^2/#')-',
H-H+N. 5 A r (Heiel) Normalization by H 8 N I* Pa - Pr. _ _ , ( (,,gy(p,,a ny _ ,- H _ _s, _ H_a + H r. g, , g, u \ _ UN NN NN H Q (rt.9 0 Using the definitions given in Table 3.3-4 in Reference II.9.3, Equation II.9.3 can be written as: a A Pa - Pr* = - <f k34 3u r(#/m)(A
+ R 3 /Ay)-Ji -D, M (A,,/A 3 )
N .
- J A L
+ ( t + n)A (A~/Ar)A h ( An /Ap ')
(r* 1 4)
~
Figure II.9.3 indicates that for the 1/6-scale jet pump: 4 ( P - Pr* ) = 1.002. (E 9 O Pf= -1 From 1/6-scale jet pump geometry: Ag = 0.008239 ft A, = 0.001747 ft A = 0.007126 ft A A=0.M M K " SA *
- Substituting these numerical values in Equation II.9.4 yleids:
K gg
= 0.64 The basis for this parameter is the 1/6-scale jet pump data.
However, it is not expected to vary significantly with jet pump geometry. (f) RDK1: Form loss coefficient used for calculating mechanical energy losses in the suction junction under reverse drive and reverse suction flow conditions. The magnitude of this parameter is determined from the table in Diagram 5-2 of Reference II.9.1, where: F, = flow area based on D6 shown in Figure II.9-1 Fg a suction plenum flow area 4/2 = Angle d shown in Figure II.9-1
1
, , , O lIlIlIlI IlIlIlIlI I l I l I l I l 0 l I l I l I l I l I l I l I l I l l l I l I l I l I l 4 l_4
~ ~d m
m
?
T ' m M
~
- ~d m
O T 'N E
- e b .:
= - - ,.
u b - - C
- - e
- _ k m
-d 20 w c:
~
-
- O
=
w m c: c- -
+ + +
o es
-d d6oo
- - ew
- - ' (C
" b
- 4 - ,? w=g H : # Z ETE C .
- O U N H
(n _- i [*Oc p w u u
- + -
- .* C O
+ y zm~
p __
- es Q
~
- + -
<at 3 k
C
~ ~ ~ >
_ - d, -
- m M
- _ are em O
* - e C
-o _ ,7 b m 6
~ ~ -
-f
- i m
- -Ji 2 ~ ~f
;lililililililililililil,lililililil lililil,lililil.ililil;
. . . . , 3, S*1 0*1 S *O 0*O S*D- 0*I- S*l-e g "Ao : / ('IdY) 8 d
(s) RDK2: Form loss coefficient used for calculating mechanical energy losses in the drive nozzle and the 180 bend under reverse drive flow co'nditions. The magnitude of this coefficient is calculated as follows: (1) Calculate the loss coefficient for the drive nozzle from the table in Diagram 5-2 of Reference II.9.1, where: ol = 2 Tan I ( A h) /r ' L L, , (f. /F, ) = ( D3/ 3 ,')A (2) obtain the loss coefficient for the 180 bend from Diagram 6-4 of Reference II.9.1. (3) RDK2 is obtained by adding the above two loss coefficients. (h) RDK3: This parameter is used in calculating mechanical energy losses due to shear between the drive and suction streams under reverse drive and reverse suction flow conditions. A numerical value of 1.0 has produced satisfactory results for 1/6-scale jet pump and will not vary with the jet pump geometry. (i) RDK4: Form loss coefficient used for calculating mechanical energy losses in the suction stream under reverse drive and forward suction flow conditions. The magnitude of this parameter is determined as: RDK4 = 0.5(1 - (D1 /D y ) ) (j) RDKS: This parameter is used in calculating mechanical energy losses due to shear between the drive and suction streams under reverse drive and forward suction flow conditions. A numerical value of 1.1 is assigned to this parameter and is not expected to change for different jet pumps. i References (11.9.1) Idel' Chik, I. E., " Handbook of Hydraulic Resistances," 1966; U.S. Department of Commerce, Springfield, Virginia. (II.9.2) Crane Technical Paper No. 410. " Flow of Fluids Through Valve Fittings and Pipe," Crane Company, New York, NY; 1976. (II.9.3) Fernandez, R. T. . Sundaram, R. K. , Chaus, J. , Husain, A. , Loomis, J. N. , Schor, L. , Harvey, R. C. and Habert, R. ,
"RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analypls, Volume I: Code Description," Yankee Atomic Electric Company Report YAEC-1300P, Volume I (October 1982) . (proprietary)
III. JET PUMP MODEL ASSESSMENT 0.TII.1 How were the loss coefficient values obtained in Tables 2.3-2 and 2.3-3 of Reference 127 A.TII.1 See the answer to Questions II.1 and II.9. Q.III.2 In Figure 2.3-12 of Reference 12, why does the RELAP5YA curve not decrease as rapidly as the data for M values greater than about 1.97 A.III.2 The discrepancy between the RELAPSYA prediction and the dets for M >1.9 is probably due to vena contracta effects in the talipipe of the jet pump. For large reverse flows into the jet pump, this phenomenon affects the measurement more than at small reverse flows. Additionally, the uncertainty analysis presented in Reference III.2.1 Indicates that the uncertalnties in the data increase significantly beyond M >1.9. However, no major efforts were rpent to resolve this issue because, in our opinion, the performance of a jet pump model is better judged by the comparison of the calculated pressure differencer to data. Figures II.1-4 and II.1-5 compare these calculated pressure differences to the test data. The good agreement shows the adequacy of our model for M > 1.9. .0.III.3 Do any of the steady-state tests given in Section 2.3-2 of Reference 12 include two-phase flow? 29-
A.III.3 All of the steady-state tests were performed at single-phase subcooled conditions. O.III.4. 0.III.5. 0.III.6. 0.III.7 and 0.III.8 0.III.4 Why are the calculated jet pump pressures below the data af ter 14 s for the remainder of the Test 1 transient in Figure 2.3-16 of Reference 127 Pages 51 and 52 of Reference 12 state: "The volumetric flow rates through the drive (QDR ' "" " SU
"" ""
- I" 9 DC ""
are plotted in Figure 2.3-17. The RELAPSYA calculations for these flow rates are satisfactory." 0.III.5 Why are the calculated drive and suction flows as much as 50% below the data during the first 10 sf 0.III.6 Why does the calculated discharge flow show a rapid increase between 8 and 12 s and a rapid decrease between 24 and 26 s and yet the data for the discharge flow only changes very slowly during the entire transient? Page 52 of Reference 12 states: "In order to assess the performance of the jet pump model under transient conditions, the pressure drops across the jet pump calculated by the code were compared to data. In Figure 2.3-19, the RELAP5YA calculation of the static pressure difference between the suction and discharge plenums is compared to octa (DPE-6). The code calculated values match the data remarkably well throughout the transient." Figure 2.3-19 shows that RELAPSYA calculates about twice as large a negative value as the data between 16 and 32 seconds, the data is about twice as large a negative value as the RELAP5YA result at 36 s, and the data is about three times the magnitude of the RELAPSYA value and of opposite sign at 40 s. 0.III.7 Please explain what causes these differences. Page 52 of Reference 12 states: "The pressure difference between the drive line and discharge plenum (DPE-7) is plotted in Figure 2.3-20. A zero off-set in the DP cell is indicated in this figure. If the data is adjusted for this off-set, then the code calculations follow the data trends reasonably well." 0.III.8 Why does the RELAPSYA reault for the differential pressure decrease between 20 and 35 s while the data increases about 50% during this period? - A.III.4. A.III.S. A.III.6. A.III.7 and A.III.8 Questions III.4 through III.8 pertain to transient Tost 1. This test was simulated with the modified jet pump model as discussed in the answer to Question II.1. The results of this simulation were almost identical to those presented in Reference III 4.1. Hence Questions III.4 to III.8 are answered here in reference tc Figures 2.3-16 through 2.3-20 of Reference III.4.1 In transient Test 1, the jet pump operates in the normal flow configuration. The small pressure changes in the jet pump are overwhelmed by the resistances in the piping outside the jet pump test vessel. The flow splits and the total flows are determined by the resistances and thermodynamic conditions in the test assembly piping rather than in the jet pump. The discrepancy between the RELAPSYA results and the data for this test are mainly due to three reasons: q.
\
- 1) ' Heat structures were not modeled. Due to this reason, the total energy of.the syste as underestimated. The underprediction of N s the pressure and volumetric flow rates can partially be attributed to this deficiency.
- w
- 2) The pressdre vessel and the nitrogen accumulators shown in Figure 2.3-13 of Reference III.4.1 were modeled as a Time Dependent Volume. The pressure and temperature histories from Figure 8 and Figure 9ofReferenceIII.4.2wereinput[asboundaryconditions.
#. -' The errors introduced in reading this data from the graphs also may
+
have contributed to the discrepancies between the RELAP5YA r s-predictions and the ' data.
- 3) The sensitivity of the system 'b'ehavior to the model of the piping near the blowdawn valve. Through a sensitivity study, it was detedhinhd that the systerbte:havine is sensitive to the direction of the' junctions connecting Volume 50 to Volume 72, Volume 108 to
- , Volume 50, Volume 50 to Volume 105 and Volume 91 to Volume 105. The model shown in Figure 2.3-15 of,Rvference III.4.1 is an approximation of the systeS from fluid momentum considerations. r'\ However, during the tranvient, the Gominating momentum flow changes from one pipe to another, and the ilping model may not be the best representation of the system. Figur's 2.3-4 in Reference III.4-1 shows the complexity of the piping network near the blowdown valve where four small pipes converge. The transient hydrodynamics in this region were difficult to model. > The modeling deficiencies discussed above may also impact the results of Blowdown Test 2. However, in that test, the jet pump operates as a resistance in the flow path. The pressure drops in the jet pump are compardt hely larger, therefore, they influence the system conditions to a greater degree. In summary, the RELAPSYA predictions for Blowdown Test 1 and Test 2 reflect the performance of the piping model rather than the jet pump model. The behavior of the jet pump model under translenL and two-phase conditions is better assessed by TLTA Test 6425/2. ,This test has been re examined recently and is discussed in s Appendixd.'IV.1ofthisdocument. The results of this simulation show good agreement between the predicted jet pump behavior and the data. 5
- s. ,
References (III.4.1) Fernandez, R. T., R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis. L. Schor. R. C. Harvey and R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor Systen Thermal-Hydraulic Analysis, Volume III: Code Assessment" Yankee Atomic Electric Company Report YAEC-1300P Volume III (October 1982). (proprietary) (III.4.2) Crapo H. S., "BWR Jet Pump Report," EG4G Report, page 394. EG*G Idaho Inc., November 29, 1979. Q.III 9 Which of the 6 possible jet pump flow configurations have been assessed for two-phase flow and over what range of conditions have they been assessed
- A.III 9 The jet pemp model was assessed for all the six possible flow configurations under two-phase conditions. The experimental tests used l for the assessment are listed in Table III.9.1.
TLTA Test 6425/2 provided the most suitable conditions for tho jet pump model assessment. During this test, the drive flow rate in the broken loop jet pump varied between 8.2 lb/see and -4.6 lb/sec. In the intact loop jet pump, the drive flow rate varied between 8.6 lb/sec and -1.0 lbm/sec. The void fraction in the two jet pumps varied from 0.0 to 0.98. The results of this assessment are provided in Appendix A.IV.1 of this document, where total flow rates in each of the two jet pumps are compared to data in Figures A.IV.1-7 and A.IV.1-8. The results of the assessment against the Jet Pump Blowdown Test 2 are presented in Reference 111.9.1. r t TABLE III.9.1 h J. 1 Jet Pump Model Assessment i, 5 t
<---------------------POSITIVE DhIVE--------------------->
-m< M 1 -1 -1 1M1 0 0<M s
(1) TLTA Test 6425/2 (1) TLTA Test 6425/2 (1) TLTA Test 6425/2 5
-6
, t i
L
<---------------------NEGATIVE DRIVE--------------------->
-so < } M i -1 -1 iMi 0 0<M (1) TLTA Test 6425/2 (1) TLTA Test 6425/2 (1) TLTA Test 6425/2 (2) Jet Pump Blowdown Test 2 O
h m 0 r 1
peterences (III 9.1) Fernandez, R. T., R. K. Sundaram, J. Chaus, A. Husain, J. N. Loomis. L. Schor, R. C. Harvey and R. Habert, "RELAP5YA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume III: Code Assessment" Yankee Atomic Electric Company Report YAEC-1300P Volume III (October 1982). (proprietary) IV. TWO LOOP TEST APPARATUS EXPERIMENTS 0.IV.1. 0.IV.2. 0.IV.3. 0.IV.4. 0.IV.S. 0.IV.6 TLTA Large Break Test 6425/2 Page 204 of Reference 12 states: " Heat losses from the vessel wall to the environment have been represented in the RELAPSYA model by a heat flux boundary condition. These heat fluxes were estimated from the heat loss calibration tests performed on the TLTA facility...." 0.IV.1 How does the rate of energy removed from the system by these heat losses compare with the energy removed by the breaks for the various break sizes modeled? 0.IV.2 Is a constant rate used or is the rate based on fluid conditions? 0.IV.3 In Figure 5.2-3 of Reference 12, what causes the step in calculated pressure at about 160 sf Figure 5.2-8 of Reference 12 shows the broken loop jet pump flow. The RELAPSYA result has about the same shape as the data but underpredicts the reverse flow by up to 50% during the period shown from 0 to 18 s. 0.IV.4 What causes the RELAP5YA result to be smaller in magnitude than the data? Figure 5.2-9 of Reference 12 shows the data for the intact loop jet pump flow decreasing almost linearly during the first 7 s. In contrast, the RELAp5YA result increases to a peak almost immediately, decreases with approximately a constant slope until 4 s, and then decreases more rapidly between 4 and 7 s. 0.IV.5 If the first surge is caused by the excess liquid inventory, why does the flow rate decrease so much faster than the data from 4 to 7 s? page 209 of Reference 12 states: " Figure 5.2-13 shows the lower plenum mass inventory. In the test, the lower plenum level stabilized at the jet pump tailpipe exit. In the calculation, for the 70 to 140 second period, the lower plenum mass was lower than in the data primarily due to extra flashing at lower pressures. Beyond about 150 seconds, the lower plenum mass was calculated to continuously increase. This was due to the fact that ECC liquid, after having filled the bypass regions, tended to drain into the lower plenum rather than refill the bundle." 0.IV.6 Clarify why this occurred for the calculation and not in the experiment. A.IV.I. A.IV.2. A.IV.3. A.IV.4. A.IV.5 and A.IV.6 In examining the questions related to Large Break Test 6425/2, a reanalysis of the test was performed primarily to obtain more appropriate output parameters which could be used to explain the results of the calculation and comparisons to the data. In the reanalysis, some changes were made to the nodalization of the TLTA faellity for Test 6425/2 shown in Figure 5.2-2 of Reference IV.1-2. The reasons for the modifications and the results of the reanalysis are described in Appendix A.IV.1, attached. Answers to Questions IV.1 through IV.6 are provided below, utilizing the results of the reanalysis. A.IV.1 The rate of energy removed from the system by heat losses are generally small compared to the rate at which energy is removed from the breaks for both the large break and the small break tests. Figure IV.1-1 compares the heat losses to the break energy loss for Large Break Test 6425/2. Figure IV.1-2 shows a similar comparison for the Small Break Test 6432/1. References (IV.1-1) R. T. Fernandez, R. K. Sundaram, J. Ghaus, A. Husain, J. N. Loomis. L. Schor. R. C. Harvey and R. Habert, "RELAP5YA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume II: User's Manual," Yankee Atomic Electric Company Report YAEC-1300P, Volume II (October 1982). (proprietary) (IV.1-2) R. T. Fernandez, R. K. Sundaram, J. Ghaus, A. Husain, J. N. Loomis, L. Schor R. C. Harvey and R. Habert, "RELAPSYA - A Computer Program for Light-Water Reactor System Thermal-Hydraulic Analysis, Volume III: User's Manual," Yankee Atomic Electric Company Report YAEC-1300P, Volume III (October 1982). (proprietary) A.IV.2 The heat loss rate is based on the time varying fluid conditions in the downcomer regions of the TLTA vessel. As noted in Page 204 of i Reference IV.1-2, heat losses from the TLTA vessel are modeled in RELAPSYA as a heat flux boundary condition. RELAP5YA allows the specification of heat flux at a heat structure boundary only as a function of time. The total heat loss from the TLTA vessel has been estimated at about 50 kW for the steady-state initial conditions of the Small Break Test 6432/1. This corresponds to a fluid temperature of about 550 to 560 F in the downcomer region and provides one reference point for estimating the heat loss as a function of fluid conditions. 2
/
5 2 . 4 s6 s ot . K l s AL 0 e tT EL t 0 a RR '0 eA hT BW { 3 dT L o . n J af d o s
" 0 kn c ao ei 0 rt
.' '5 ba 2 l r
jI 1 s= . s os hu gm ui I o r f n hA
- o tY
. 0 c
G O z sP 5 N s sA oL
, 2 , lE E yR M gn ri
-* I e nl
,I
- 0. T el a
O fw 1
'5 o l
c 1 g . ne os J ss
- i e rv a
0 pm mo 0 or
. Cf d i
. - 1 1
V I 0
- e b r
. 5 u g
i F 0 E : ~. .
?: * - _E: 6~- -d .-
0 O~ O~ n O~ O..OX I*
@OE'$g .N& m8a >.yW* '
O*
I I I I I I i BRERKS - u<, c ang M oWRLL H (D _ W ~ l g - _ CE _ ; l _
& \
i
~
o
' D ~
m I .M,, l o__.)~ O _.
~_ r I
J, j h . _ b I e) _ m I' U _' " " " Z g _ CD ~ , r, ,, 0.0 INO.0 2NO.0 3NO.0 4NO.0 SNO.0 6N0.0 7N0.0 8bb.0 T I ME , sEconos Figure IV.1-2: Comparison of energy loss through breaks with heat loss from vessel wall in RELAP5YA simulation of TLTA Test 6432/1.
It is further assumed that the heat loss is directly proportional to the difference between the downcomer fluid temperature and the ambient temperature (about 70 F). The downcomer fluid temperature variation is estimated from the test data for each of the TLTA tests modeled. Thus, the heat losses from the vessel are known as a function of time for each test. This information is used to specify the boundary heat flux as a function of time for each heat structure used in RELAP5YA to model the vessel wall for each TLTA test. A.IV.3 Beyond about 70 seconds into the transient, the system pressure is primarily determined by the balance between the energy removal rate at the break and the heat transfer to the fluid in the vessel. Figure 5.2-11 of Reference IV.1-2 shows that the mid-plane of the fuel bundle quenches at about 160 seconds. This would cause a rapid increase in the heat transferred to the fluid and, consequently, a rapid rise ir 6.he system pressure. An analogous but less pronounced effect is observed in the reanalysis. There, the top and bottom of the bundle quench at about 130 to 140 seconds. Hence, at this time there is also a slight step latrease in pressure. This can be seen in Figures A.IV.1-22 and A.IV.1-24 siven in Appendix A.IV.1. This small pressure step is not expected to significantly affect the system behavior. A.IV.4 The magnitude of the reverse flow through the broken loop jet pump is partly affected by the jet pump nodalization. The more realistic nodalization used in the recalculation of Test 6425/2 shows much better agreement between calculated and measured flow rates, especially at earlier times. This is shown in Figure A.IV.1-7 in Appendix A.IV.1.
A.IV.5 The reason for the steeper slope between 4 and 7 seconds is because the jet pump suction line uncovers at 4 seconds in the original calculation. In the test, this occurred at 6.7 seconds (see Table 5.2-2 on page 231 of Reference IV.1-2). In the reanalysis, the jet pump uncovered at 9 seconds due to the slightly higher initial water level (Figures A.IV.1-8 of Appendix A.IV.1). However, the initial peak in flow is not observed. This is believed to be partly due to the more realistic initial water level and partly due to the more realistic jet pump nodalization. Figure A.IV.1-8 shows that the coastdown flow in the intact loop jet pump is predicted well. A.IV.6 In the test, as the ECC liquid filled up the bypass region, CCFL breakdown was observed at the bypass-to-bundle leakage path that led to water accumulation in the fuel bundle inlet piece. Steam in the lower plenum prevented this liquid from draining into the lower plenum through the side-entry orifice. Thus, the water was able to move
. upward into the bundle and quench the lower bundle regions. This phenomenon, combined with ECC water draining directly into the bundle from the upper plenum, led to the final bundle quench.
In the calculation discussed in Reference IV.1-2, ECC water filled the bypass region and started draining into the fuel bundle inlet piece at about the same time as observed in the test. However, this water was not able to move upward into the bundle. Instead, it drained into the lower plenum through the side-entry orifier. This can be seen in Figure 5.2-13, which shows the lower. plenum mass to continuously increase beyond about 150 seconds. Further, the ECC water in the upper plenum was also unable to drain into the bundle. This can be seen in Figure 5.2-15, which shows the upper plenum mass to be higher than that observed in the data beyond about 70 seconds. We believe the inability of ECC water to penetrate the fuel bundle as early as seen in the test is mainly due to high vapor generation rates in the bundle, calculated by RELAP5YA. This would cause steam binding in the bundle that further prevents ECC liquid penetration. This is generally consistent with other RELAP5YA calculations shown in Reference IV.1-2. An additional possibility is that the CCFL characteristics at the side-entry orifice are incorrectly represented in the RELAP5YA model. If this is the case, it is considered acceptable because early breakdown of CCFL at the side-entry orifice would result in.less liquid inventory in the bundle and thus provide a conservative calculation of the transient thermal response of the bundle. The reanalysis of Test 6425/2 described in Appendix A.IV.1 shows similar results. 0.IV.7. 0.IV.8 and 0.IV.9 TLTA Large Break Test 6426/1 Pages 213-214 of Reference 12 state: "The drive line break flow was lower than the data initially (during subcooled blowdown), but was compensated by the slightly higher flow rates calculated at later times (during saturated blowdown). The drive line break flow was only 10-15% of the total break flow. Hence, this discrepancy in the calculation is not expected to significantly affect the system pressure response." 0.IV.7 Clarify that the drive line break flow was only 10-15% of the total break flow because Figures 5.2-26 and 5.2-27 of Reference 12 seem to indicate that at least part of the time it was about 25% of the total break flow. Figure 5.2-27 of Reference 12 indicates the drive line flow rate is as much as twice the RELAPSYA calculated value before about 30 s and the RELAPSYA calculated value is more than twice as much as the data between about 40 and 70 s. 0.IV.8 Please clarify what causes these discrepancies, be ause the break flow differences may indicate modeling problems that may be significant for other cases. 0.IV.9 Why is the RELAP5YA lower plenum mass so much less than the data after 70 s in Figure 5.2-33 of Reference 12? A.IV.7. A.IV.8 and A.IV.9 A.IV.7 The measured break flows on the suction and drive side of the pump can be seen in Figures 5.2-26 and 5.2-27 of Reference IV.1-2. From these figures, it is confirmed that beyond about 30 seconds, the drive side break flow is indeed between 10% to 15% of the total break flow. In the initial 30 seconds, however, the drive side contribution to the break flow does appear to be about 25% to 35% of the total break flow. This is further discussed in the response to Question IV.8. A.IV.8 As noted from Figure 5.2-27 of Reference IV.1-2, the calculated drive side break flow is lower than the data by about 50% during the first 20 seconds and is higher than the data between 30 and 70 seconds. The reason for these discrepancies is not clear, especially considering the more reasonable predictions of break flow rate for Test 6425/2 (Figures 5.2-5 and 5.2-6 of Reference IV.1-2). The TLTA Large Break Tests 6425/2 and 64?6/1 were conducted with nearly identical reported initial conditions (Tables 5.2-1 and 5.2-3 of Reference IV.1-2). The timing of key events was also very similar until about 60 to 70 seconds when ECC injection in Test 6425/2 began to have an impact (Tables 5.2-2 and 5.2-4 of Reference IV.1-2).
Consequently, the break flows should also be very similar in the 0-70 second period. Indeed, as seen from Figures 5.2-5 and 5.2-26 of Reference IV.1-2, the measured suction side break flows are virtually identical. However, the measured drive side break flows are quite different (Figures 5.2-6 and 5.2-27 of Reference IV.1-2). This discrepancy cannot be explained by the small differences in the reported initial conditions. Perhaps the initial temperature in the stagnant fluid upstream of the drive side blowdown valve was different between these two tests. This could explain the different early critical flow histories between these two tests. However, the initial temperatures in this pipe are not reported, and therefore, this hypothesis cannot be confirmed. The RELAP5YA calculations of TLTA Tests 6425/2 and 6426/1 were conducted with similar initial conditions. Hence, the calculated break flows for the two tests are also very similar. Agreement with data appears to be reasonable for Test 6425/2 and also for the suction side break flow for Test 6426/1. The poor agreement between calculated and measured drive side break flow rates for Test 6426/1 cannot be explained because of the apparent data inconsistencies discussed above. O In general, RELAPSYA calculations of critical flow rates under subcooled and saturated conditions have been very reasonable (see Marviken test compacisons in Section 2.2 of Reference IV.1-2). Hence, we do not believe there are any major modeling problems in RELAPSYA calculations of critical flows. We expect to perform break spectrum sensitivity studies in licensing submittals and believe that uncertainties in break flow predictions will be addressed through these studies. A.IV.9 In Test 6426/1, the lower plenum mass is primarily determined by the nature of the flow leaving the lower plenum via the jet pump tallpipes. Considering the test data in Figure 5.2-33 of Reference IV.1-2, it can be seen that the lower plenum mass showed a very rapid initial decrease up to about 15 seconds. This corresponds to predominantly single-phase liquid flow up through the jet pump tailpipes. From 15 seconds to about 30 seconds, the lower plenum mass decreased less rapidly. This corresponds to two-phase flow through the jet pumps. Beyond 30 seconds, the two-phase level in the lower plenum fell below the jet pump tailpipe elevation and the flow through the jet pumps was mostly single phase vapor. This led to a very slow decrease in the lower plenum mass beyond 30 seconds.
- The dif ferences between the RELAPSYA calculation of lower plenum mass and the data are believed to be primarily due to the nodalization of the lower plenum region shown in Figure 5.2-2 of Reference IV.1-2.
RELAPSYA uses a donoring scheme to specify junction void fractions which are used in determining junction flow rates. Referring to the nodalization diagram, the mass flow through the junctions connecting the lower plenum to the jet pump tailpipes will be influenced by the void fraction in Volume 102. Figure IV.9-1 shows the calculated void fraction histories in all the lower plenum nodes. It can be seen that Volume 102 Ysgins to void at about 15 seconds and completes voiding at about 90 seconds. Thus, RELAPSYA calculates two-phase flow through the je; pumps over this 75-second period instead of 15 seconds as seen in the data. Figure 5.2-33 of Reference IV.1-2 shows that the rate of decrease of the lower plenum mass calculated by RELAp5YA approaches the rate seen in the data 1 as the void fraction in Volume 102 approaches unity. Since Volume 102 has to be completely voided before the jet pump flow becomes mostly single-phase supor, the lower plenum mass calculated by RELAPSYA is lower than tha t seen in the data.
~
TLTR LARGE BREAK 6426-1 RELHPSYR CALCULRT10N
- a. LOWER PLENUM VOID FRACTIONS e- m i . i . i i .
-a p -
5
;; 7 N-
-J .
O Volume 104 m : I volume 103 c5 - k Volume 102 - Volume 101 Z l O ~R_ go \. O T 0 5 o_ _ O 9 y a- -
\\ $[ .] ,
fj 1 i I i i i 0.0 50.0 100.0 150.0 200.0 250.0 300.0 350.0 T IME , sEconos Figure IV.9-1: Lower plenum void fraction history in RELAPSYA simulation of TLTA Test 6426/1 d
~
0.IV.10. 0.IV.ll and 0.IV.12 TLTA Small Break Test 6432/1 Page 222 of Reference 12 states: "The voiding of the lower plenum region is calculated reasonably well by the code. The fact that the lower plenum was calculated to be partially voided up to about 600 seconds indicates that the code was able to calculate the CCFL occurrence at the SEO. Also, the lower plenum is calculated to refill at about 600 seconds. This compares favorably with the experimental observation (Reference 5.2-9) that CCFL broke down at the SEO at 610 seconds. However, the lower plenum did not refill completely until beyond 750 seconds in the test." 0.IV.10 If CCFL broke down at about the same time for both the calculation and the test, what causes the 150 s difference in lower plenum refill time? Does Figure 5.2-44 indicate that the CCFL breakdown occurred earlier than 610 s for RELAP5YA? Page 223 of Reference 12 states: "The only discrepancy (although small) between the calculation and test data was in the underprediction of system pressure after ECC flow was initiated. Since the effect of bundle heat transfer is expected to be minimal for this test, it is suspected that the code may have overpredicted the condensation mass transfer rates associated with ECL injection. However, it appears that this did not seriously affect the overall calculation for this test." The results presented indicate other discrepancies, for example, the lower plenum refill time given in the previous quotation. 0.IV.ll How accurately has the code modeled condensation mass transfer rates associated with ECC injection for other tests where there is partial system voiding?
0.IV.12 Justify that the overprediction of condensation mass transfer rates would not give nonconservative results for analyses where ECC injection occurs into partially voided systems. A.IV.10. A.IV.11 and A.IV.12 A.IV.10 Figure 5.2-44 of Reference IV.1-2 does indicate that CCFL breakdown at the SEO was calculated by RELAp5YA to occur between 500 and 525 seconds. This is further illustrated in Figure IV.10-1 (attached) which shows the void fraction history in Volume 103 in the lower plenum (see nodalization diagram, Figure 5.2-35 of Reference IV.1-2). In the test, CCFL breakdown at the SEO appears to have occurred around 610 seconds. Thus, RELAp5YA calculated CCFL breakdown at the SEO about 100 seconds earlier than in the test. There are other locations in the TLTA facility where CCFL breakdown appears to have been predicted reasonably well by RELAp5YA. These include the bundle upper tie plate, bypass exit and the bundle-to-bypass leakage path. All these flow paths, including the SEO, are modeled similarly in RELAp5YA. The input parameters used to characterize these flow paths (or junctions) are the flow area and form loss coefficients. It is believed that this representation was adequate for the flow paths at the bundle upper tie plate, bypass exit and the leakage path. However, for the more complicated flow path at the CEO, the flow area and form loss coefficients may not have provided sufficient representation. This may have contributed to the early pre 61ction of CCFL breakdown at the SEO. The early prediction of CCFL breakdown at the SEO leads to earlier refill of the lower plenuro and less fluid inventory in the bundle. This can be seen in Figure 5.2-41 of Reference IV.1-2 which shows bundle mass to be underpredicted around 500 seconds. The voiding of the bundle caused a mild heatup of the bundle in the calculation which
TLTR SMALL BRERK 6432-1 . . CALCULRTED VOID FRRCTION oe VOLUME NODE 103 . l I I I I I I i
~
j. t VOLUME 103 9 - o- -
~Z +
0 =
.- R
- -. o - - .
O CC ' s E o k CCFL c T, BREAKDOWN ,
.-. O AT SEO O l
> d i ads OPENING i)f l q l o_ _
Y a . .
^
I i i i i i i 0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 T I ME, SECONDS Figure IV.10-1: Lower plenum void fraction history in RELAP5YA simulation of TLTA Test 6432/1.
was not seen in the data. This is shown in Figure IV.10-2 (attached) which compares calculated and measured inner clad temperatures in the mid-bundle region. The same thenomena were observed in the RELAPSYA calculation of TLTA Large Break Test 6425/2 (see Figures 5.2-13 and 5.2-16 in Reference IV.1-2) that contributed to a delayed bundle quench (Figures 5.2-18 through 5.2-20 of Reference IV.1-2). Hence, it appears that earlier prediction of CCFL breakdown at the SEO leads to conservative predictions of bundle fluid inventory and peak clad temperatures. For this reason, we propose to use the sarue modeling principles in reactor applications. If alternative techniques are employed to provide a better representation of the SEO flow path, they will be justified on a case-by-case bcsis. A.IV.11 In general, when water injection into a steam-water mixture is encountered, RELAPSYA will calculate condensation mass transfer rates very close to the equilibrium rates. In cases where the injected water is subcooled, this can result in an overprediction of condensation mass transfer and, hence, an underprediction of system pressure. The magnitude of the underprediction of the system pressure will depend mainly on system characteristics such as the initial thermodynamic state of the system and the system boundary conditions. The impact of overpredicting condensation rates in BWR Systems is further discussed in the response to Question IV.12. A.IV.12 In the analysis of TLTA Small Break Test 6432/1, RELAPSYA calculated the system pressure to be lower than .be measured value after injection of subcooled (90 F) ECC water (Figure 5.2-36 of Referr ice IV.1-2). This is believed to be due to the overprediction of condensation mass transfer rates by RELAPSYA. One consequerce of the underprediction of system pressure is that ECC flows are overestimated. Figures IV.12-1
e w ,
, 0
, 4
, M 4 0
_ - 0
- 0 1
E R U T A 2 C 2 2 1 R 1 2 6 5 2 E 0 0 9 9 0 1 1 0 0 1 P M , F F P F P F P P F P E T _ D _ T 7 e C S W S E S W S E N N W AS 6 s 4 _ l A d n
.w T M L u r Md T 1 7
1 7 1 7 1 1 U eb 7 7 N r L L L L L E r - E E E E E L - ud 1 P si N R 6 am e U 7 8 ; 5 4 E R
/
0 9 1 S 2 S 3 S 6 S W ^- me 2 O O P O OP O PO h 3 P P L A dt . 4 6 Y n 1 T 5 an/ S 1 2 3 4 5 6 P i2 E A d 3 T L es4 E t e6 R ar l ut ut s cae i c l rT r a ae l cpA 0 E mT l 0 5 f eL MI oTT T nd r S oao sl f iC
*e e e r aro pei n
ee e mng one ee e CI r ee e 2 ee E - e 0 1 ee e 0 l 5 V 2 I ee e e r
*e e e k i F
u g N 0 - 0 0 0 0 - 4 6 1 - E w
and IV.12-2 (attached) show the calculated and measured LPCS and LPCI-flows. It can be seen that the calculated flows are slightly higher than the data. . In order to estimate other consequences 'of the overprediction of condensation rates, an additional calculation of Test 6432/1 has been carried out with the ECC water temperature raised to 200 F (from 90 F used in the test). All other features of this calculation were identical to the previous calculation. The resulting system pressure response is compared to the data in Figure IV.12-3. It can be seen i that the calculated pressure is higher and closer to the data than with the colder.ECC water (F'.gure 5.2-36 of Reference IV.1-2). The resulting LPCS and LPCI flows are lower, as seen in Figures IV.12-4 and IV.12-5, respectively. The higher ECC water temperature had little impact on other system parameters. Figure IV.12-6 shows a comparison between the calculated and measured mass history in the lower plenum. Comparing this to Figure 5.2-44 of Reference IV.1-2, it can be seen that CCFL breakdown at the SEO occurred at about the same time as before, but the lower plenum took a little longer to fill completely. The bundle mass history was similar to that shown in Figure 5.2-41 of Reference IV.1-2, and a mild heatup of the bundle occurred at about 500 seconds. Figure IV.12-7 shows a comparison of the calculated and measured inner clad temperatures in the mid-bundle region. The results are similar to 4 those shown in Figure IV.10-2. A similar sensitivity study was also carried out for TLTA Large Break Test 6425/2. The base case for this study was the RELAPSYA calculation presented in Appendix A.IV.I. For the base case calculation, the HPCS water was at 131 F, and the LPCS and LPCI water were at 118 F as in the test. In the sensitivity study, ECCS temperatures of 160 F and 200 F were used. Figure IV.12-8 shows the system pressure response for the base case, the case with 160 F ECC water and the case with the 200 F ECC water. The base case pressure response was compared to data in Figure A.IV.1-9 in Appendix A.IV.1. It can be seen by ? 4 , __.m , , - _ _ . - - , , _ , . _ _ , _ _ . , , _ . _ . . _ _ . , _ _ _ , , - - _ _ _ _ - _ . - , __ _ _ - , _ . _ . ~ - _ _ _ _ - _ . . , _ _ _ _ - - .
- fl~ $'
1 0 0
- 0 1
d 0 5 e l 7 r D u E s T a . A e1 N m/ 2 MI d3 R n4 E a6 T C T S dt S A E es T T t e L aT T 6 l 6 P uA D cT 1 l l L N W t c aT U O L 0 e s c M F l 0 ( r 2 5 E f o 3 S M of 4 C I 6 P T L ns T S ow E A so T Y il 5 rf P a S A pS C L mC P E oP L R CL 1 Y, 0 2 1 V 7 S,, if ! 5 2 I e I - e A k'
=
r u g 9 i F r
~ -
3 0 2 1 o 3ioiOd 1$l
l ll _ y_ be 00 0
- 0
_ - 1 I 0 d
' 5 7
e r u s a . e1 m/ 2 d3 n4 a6 dt es C t e S aT A l T uA L cT T 3 t c l L 6 a aT e P D 0 t c 1 I 0 5 E r N M f o U W I T of 1 8 O
/
2 t A ns 3 F Y ow 4 6 I C S so P P i l T L rf S A a E L pI T 1 E R mC oP CL I C : P L 2 2 1 0 . 5
$ TJ 1
2 V I N f 4 I5 e r c L b - u I o lyl i F g T 9
@ F 0
3 2 , 0 3 gi a't j', !! l
i
)
1 IM 8570 TEST 6437/RUN 1 TLT A SC STE AMLINE VALVE CLOSE D ADS OPENS q, PI
! w_ j_ ~
L I 1 P1 STE AM DOME 8'.C - 5600 E ARLY RESPONSE GOVERNE D BY _k j RESPONSE OF RELAP5YA % a PHESSURE CON- " TROL V ALVE i d E l $ T E
- i i .
E a 400 - _ g
- -LPcs o
- LPCI _ F o
F T-/ k T i l [ ' TEST i TEHMINATED RELAPSYA O' I I ' O 7% O 53 ) O 750 1000 TIME (sec) Figure IV.12-3: Comparison of calculated and measured system pressure response for TLTA Test 6432/1
- with ECC water at 200 F.
j
j3-1 O0 0 0
- 1 d
0 eh 5 rt 7 ui D E sw a T e1 A m/ N 2 MI d3 R n4 E a6 T C T dt S S es A E t e T T aT L l T 6 6 uA . P cTF 1 D I l L W
)
c aT0 N U O es c 0 L 0 ( r2 R
/ F ! 0 E f o 2 5 of t 3 S M a C I 4
6 P L N T ns owr T soe S E ilt T A rf a Y a w S pS S P mCC C A oPC P L CLE L E R 4 2
;- 1 1 0 V 3 .-
ka, _: - 5 2 I c
,9 y
- e A r fI !
l ly u l g bt 1 _- i F _ f
~ -
O 3 2 1 0 ij8 m
_1g8e 1 O0 0
- 0 1
d e 0 r 5 7 u s a e1 m/ 2 d3 n4 a6 dt . esF t e aT0 C l 0 S uA2 A cT T l Lt L aT a T 3 i c 6 c P o s rr 1 D 0 ( f oe 0 E of t N 5 M a U W I nsw R
/
O L A T ow 2 3 F Y soc 4 I S i lC S C P rf E T P L A a S L pIh E E mCt T 1 R oPi CLw I 5 C - P L 2 1 0 V I 5 2 e r N f 4 .: u g
- i
?
[ 9l ,y g F T -
@ r
- O 3 .
O v agi ai
- .2 0
8 *
- _ 0 2 O 0 0
- 0
~ - ~ 1 F
0 i I 5 F 0 d 0 e 2 i r u t q s a a e r m e i rt d oa
- nf w a
yC C S > d rC eoE A t t T ash L li t T ) c e uhi 0 ( s c w 1 I 0 l s 5 E as1 N M ca/ U A m2 I H Y T
/ S f 3 2
3 P om4 4 A u6 6 L nn T E oet S R sl s E i pe T r T ar pea mwI S ooL _ N ClT E P
\ I O l :
S 6 D 0 - A I 5 2 2 1, V _ I T L- e _ r - _ u g i - _ F - Q-C 0 0 0 0 - 8 2 6 1 i -
.2 -
4@
, i
'- r ixo .enive . 1 I -
a ? i i i a r . k o e=0 aaE 8 t 5: : : s t t t t g f\ I
~
y
+p s
anar!! -
.8
-u t
- : n n
=
n ; Ts 12 eeeee: ; "e 1" 8
; < = =- 83" i s!EEar a t :"5 as s
eeeees 3 ." j ? m - g -~n . .. a .c 3y
?.0
%8*
'a ,
g3y i SSE g ;E 8 &<
- u "$
O E4 b
' = 8?8 LO e4 ~
x
"@@@@@@@@* 3g
@@@@@@@@ e a t,
@@'@@@@@@ o5
* @@@@@@@@ , a
@@@@@@@@ 6
? ;
@@@@@@@e :
e eeseeeeet s
\ Z T
s
, e I 1.
s i , s I _A l
$ 1 E ,
$ k i
(dd 3en1VW3dN31 s k i s ' i
I
. DOME PRESSURE o
k' E v ta3 '
- a i
[ ECC TEMPEPATURE
. 200 F 1 160 F z 118/131 F z__ _ ;E 3
00 e'0.0 1e'0.0 250D 3204 TIME (SEC)
. Figure IV.12-8: Variation of system pressure response with ECC water temperature in RELAP5YA simulation of TLTA Test 6425/2.
s
, .s
,i comparing Figures IV.12-8 and A.IV.1-9, that,the hotter ECC water
. temperaturemaintainsahighersystempressuheandisclosertothe data.
The impact of the ECC water temperature (and, hence, condensation rates) on'the bundle thermal response is shown in Figures IV.12-9(a), IV.12-9(b) and IV.12-9(c). These figures show a comparison of the inner clad temperatures for the three cases at the top, middle and bottom regions of the bundle. From these figures it can be seen that
.the calculated peak clad temperature at any one axial location is not very sensitive to the ECC water temperature. However, the hotter ECC water results in a more delayed quench. We believe this is due to the lower water density in the upper plenum and bypass regions that reduces the hydraulic force pushing water into the bundle. The order in which the bundle quenches (top, bottom and finally the middle region) is not affected by the ECC water temperature.
The sensitivity studies carried out for the TLTA Small Break Test
+
(432/1 and TLTA Large Break Test 6425/2 have shown that the ECC water temperature (and, hence, condensation rate) does not have a major impact'. on the phenomena calculated by RELAPSYA for these tests. Use of reallstic ECC water temperatures resulted in system pressures lower than those observed in the tests, but the thermal and hydraulic response of,the bundle was calculated conservativel[. A similar result is expectei in BWR applications. With the simultaneous;rse of EM
, t
' features, we believe that RELAPSYA will yield conservative predictions of peak clad temperatures in our BWR LOCA analyses.
^ .
4 r O.IV.13 and 0.IV.14
!; i TLTA Boil-Off Test 6441/6 Page 223 of Reference 12 states: "The boil-off tests attempted to simulate system conditions which would occur during a small break LOCA if none of the Emergency Core Cooling Systems including the ADS were available. The tests were conducted in the TLTA-5A configuration (see Figure 5.2-1). In these tests, the recirculation loops were blocked off and the liquid inventory
=
I NODE 3 ECC at 200 F (a)
- TOP REGION v l60 F OF BUNDLE 118/131 F J .
^b 3 5 5 i NODE 5 ECC at 2000F u (b) g 160 F MIDDLE REGION $ f. OF BUNDLE 118/131 F E
k
# 8 . . .
E NODE 7 ECC at 200 F (c) BOTTOM REGION g, 160 F OF BUNDLE
/ .-
118/131 F l 1 i I as aIma mino em ama TIWE (SEC) Figure IV.12-9: Variation of inner clad temperature response with ECC water temperature at three axial locations in RELAP5YA simulation of TLTA Test 6425/2. l i l was slowly boiled off at a constant pressure and constant bundle power. The power level was representative of decay heat in a BWR. The main objective of these tests was to evaluate heat transfer in a partially covered bundle at I 1 decay power levels and low (natural circulation) flows." l 0.IV.13 Why are the RELAP5YA calculated void fractions so much higher than the data at about 325 s in Figure 5.2-53 and at about 375 s in Figure 5.2-547 Pages 227-228 of Reference 12 state: " Figure 5.2-55 shows the clad temperature history at the highest measurement location (Elevation 143 inches) in the bundle. Alao shown in the figure is the RELAPSYA calculation. It can be seen that the clad temperatures in the post-dryout region are significantly underpredicted at this location. Some of this underprediction is due to the calculation being at a slightly higher elevation in the bundle than the measurement. However, the major portion of this underprediction is due to the inability of the code to calculate significant vapor superheat until the quality reaches unity. This can be seen in Figures 5.2-55 through 5.2-60 which show similar comparisons of clad temperature history at lower bundle elevations. It can generally be seen that the code is able to calculate reasonably correct temperatures closer to the two-phase level. At locations nearer the two-phase level, the vapor temperature would be close to the saturation temperature, as calculated by the code. Hence, the wall temperatures are reasonably predicted. However, further above the two-phase level, the vapor temperature would be considerably higher than the liquid , saturation temperature. This would cause the wall temperature to be correspondingly higher. RELAPSYA was not able to calculate this aspect of the thermal response of the bundle. However, the Peak Clad Temperature (PCT) calculated by RELAPSYA is comparable to the data. This is shown in Figure 5.2-63. This figure shows the locus of peak clad temperature history calculated and measured in the bundle, and indicates a favorable comparison. It is the PCT location that was miscalculated by the code."
0.IV.14 Why do the RELAP5YA calculated clad temperatures have maximums that decrease rather than increase in the upward direction? The data shows clad temperature maximums increasing in the upward direction. A.IV.13 and A.IV.14 A.IV.13 The main difference between the calculated and measured void fraction histories for TLTA Boil-Off Test 6441/6 is believed due to differences in initial conditions. As noted in Section 5.2.6.3 of Reference IV.1-2, the calculation and the experiment ut*lized different procedures to arrive at a set of initial conditions corresponding to zero time in Figures 5.2-47 through 5.2-63. This caused some initial perturbations in the calculation and also produced a slightly lower
, initial liquid inventory in the bundle. This can be seen in Figure 5.2-50 which shows the calculated void fraction in the uppermost bundle region at 0.0 seconds to be higher than that seen in the data.
This difference is equivalent to about 50 seconds on the time scale. Thus, the calculated boil-off history occurs about 50 seconds earlier than reported in the test data. This shift in the time scale can be seen consistently in Figures 5.2-50 through 5.2-54. In Test 6441/6, when the differential pressure measured by DP9 in the downcomer (see Figure 5.2-46 of Reference IV.1-2) decreased to about 0.65 psia, feedwater was injected in the downcomer to recover the bundle inventory. In the RELAPSYA calculation, because of nodalization differences, a control variable was used to estimate the differential pressure that would be measured by DP9. When the value of this variable decreased to 0.65, feedwater injection was initiated in the calculation. Figure 5.2-47 shows this to occur at 350 seconds at about the same time as in the experiment. If the time scale in the calculation was shifted to the right by about 50 seconds and the same method was used to initiate feedwater injection, then the injection would be delayed to about 400 seconds. However, at about 350 seconds, ) the calculated differential pressure would be about 0.07 psi higher
'than measured by DP9. This is within the uncertainty associated with the DP9 measurement and the representation of this measurement in RELAP5YA by the control variable.
The above discussion supports the following conclusions: (a) The higher void fractions in the calculation seen at 325 seconds in Figure 5.2-53 and at 375 seconds in Figure 5.2-54 are due to differences in initial conditions. A shift of 50 seconds on the time scale would resolve these differences. (b) Along with the shift in time scale, feedwater injection can be calculated to occur at about the same time in the test without compromising the test simulation. This would resolve the differences in the hydraulic response of the bundle between the test data and the RELAPSYA simulation of Test 6441/6. A.IV.14 The measured peak clad temperatut es do not increase monontonically with bundle elevation. Table IV.14.1 shows the measured PCT at various bundle elevations, as estimated from Figures 5.2-44 through 5.2-62 of Reference IV.1-2. Ir general, at a location above the two-phase mixture level, the clad temperature is primarily determined by the rod power at that location and the local vapor temperature. The rod power varies along the length according to a chopped cosine profile and is peaked near the bundle mid-plane. The vapor temperature increases monotonically above the mixture level. These effects combine to provide a measured PCT that is highest at the 130-inch elevation. Table IV.14.1 also shows the PCT calculated by RELAPSYA at similar elevations as the data. It can be seen that the PCT calculated by RELAPSYA is highest nearer the bundle mid-plane. This is because RELAPSYA calculates the vapor temperature to be near saturation at all the bundle elevations. Hence, the clad temperature response is more influenced by rod power variation than the vapor temperature. This was discussed in more detail in Section 5.2.6.4 of Reference IV.1-2. It should be noted that the TLTA Boll-Off Tests were designed to produce fuel rod heatup without any means of recovering the bundle fluid inventory. As such, they are not typical of conditions encountered in BWR LOCA analyses. The TLTA Large Break Test 6425/2 and the Small Break Test 6432/1 are more representative of BWR LOCA transients. For these cases, RELAPSYA generally provided a conservative calculation. In reactor applications, if system conditions occur that are similar to those observed in the TLTA Boil-Off Tests, then the RELAPSYA calculations will be reassessed and justification provided on a case-by-case basis. s TABLE IV.14.1 Comparison of Neasured and Calculated PCT at Various Bundle Elevations Elevation Above BHL PCT Measured PCT Calculated (inches) F F 107 600 1 20 760 115 770 1 10 770 120 815 1 25 720 123 740 1 10 710 130 835 i 15 680 135 820 620 140 815 i 5 570 143 730 1 20 530 4
APPENDIX A.IV.1 Recalculation of TLTA Test 6425/2
- 1. INTRODUCTION Some of the discrepancies between RELAPSYA calculation and the data have been found to be input-related. Hence, the TLTA Large Break Test 6425/2 was reanalyzed. Two types of changes were made to the code input prior to recalculation. These were:
o Changes in nodalization o changes in initial conditions. A. Nodalization The nodalization changes are as follows:
- 1. The jet pump nodalization in the original calculation had been established (Reference IV.1-2, Page 200) to enable the code to run with larger time steps. The mixing section had been lengthened and the diffuser and tailpipe sections shortened in order to mitigate the Courant limitation. Although the overall heights and volumes had been preserved and that jet pump nodalization greatly reduced the time to obtain an initial steady-state by allowing Isrger time steps, it has been found that the original jet pump nodalization had an impact on the transient jet pump behavior. The impact was observed in the first 18 seconds of the subcooled blowdown, and is discussed in detail later in this appendix with the results of the new calculation. The modifications in the jet pump input for the new calculation are:
- a. The lengths of the mixing section, diffuser and'tallpipe sections were reset to reflect actual geometry. The Courant limitations were endured rather than circumvented.
f
- b. Improved los't coefficients were used. The previous loss coefficients were based on an intermediate jet pump model.
The new loss coefficients reflect the recommendations of the jet pump model described in Section 3.3 of Reference IV.1-1. The method used in determining these loss coefficients is further described in the response to Questions II.1 through II.4 and Question II.9. In addition, surface roughness values were changed to reflect the actual surface finish of these components.
- 2. The second nodalization change was to relocate the valve on the suction side of the broken loop recirculation pump closer to the pump itself. This was done to more accurately reflect the mass isolated in that loop during the experiment. This change had no impact on the calculation.
B. Initial Conditions TLTA Large Break Test 6425/2 was initiated from transient conditions. Specifically, the facility could not operate in steady-state at the power level s' which the test was initiated (5.06 MW). Therefore, in the test procedure, an initial steady-state was reachec at approximately 2 MW. Then, the power was increased to 5.06 MW and the break was opened. While the reported initial conditions at the time of the break are known (within an experimental uncertainty band), there is considerable uncertainty in the history of how these initial conditions were obtained. It is only known that the power was ramped from about 2 MW to 5.06 MW and held constant at 5.06 MW before the break was opened. A few trial calculations were carried out by varying the time over which the power was ramped and the time over which the power was held constant. From these trials, the power history that led to initial conditions closest to the data was chosen. In the new initialization procedure, steady-state conditions were first established at a bundle power of 2.0 MW by using a feedwater controller to match feedwater flow to steam flow.
When steady-state conditions were reached, the feedwater flow was decreased and held constant at 1.4 lbm/s. This corresponds to a time of -10.0 seconds in the calculation. At -2.0 seconds, the power was ramped to 5.06 MW over 2 seconds. At 0.0 seconds, when the power reached 5.06 MW, the feedwater was cut off and the break was initiated. The motivation for attempting this initialization procedure was to follow the test procedure more closely. Table A.IV.1-1 presents selected calculated parameters at the 2.0 MW steady-state condition, as well as calculated and measured parameters at the 5.06 MW transient initial condition. The initial conditions are within the experimental uncertainty reported for the data. C. Results Figures A.IV.1-1 to A.IV.1-8 compare the short-term response of calculated and measured parameters. The side-entry orifice flow, bypass leakage path flow, bundle ficw, lower plenum to guide tube flow and steam flow, corresponding to Figures A.1V.1-1 to A.IV.1-5, respectively, show trends that are consistent with the data following the break opening. The coastdown flow through the two recirculation
, pumps also agrees very well with the data, as indicated in Figure A.IV.1-6.
The reverse flow through the broken loop jet pump (Figure A.IV.1-7) is in good agreement now, confirming that the previous overprediction was resolved by the improved jet pump nodalization and loss factors discussed in the introduction. For the intact loop, the early rise in the jet pump coastdr'. flow, which existed in the previous calculation, is not seen in the new calculation (Figure A.IV.1-8). This could be due to the initial distribution of the liquid in the new jet pump nodalization which is now more representat8're of the test facility.
Figure A.IV.1-8 shows that.the intact loop jet pump suction uncovery time is now 9 seconds rather than the previously calculated value of 4 seconds. In the previous calculation, the initial annulus inventory was too high and depleted too quickly when a subcooled discharge coefficient of 0.9 was used. The high C * * " D calculated subcooled suction-side discharge flow higher than the data. This led to an early uncovery time. In the current calculation, the coefficient was reduced to C = 0.7 and the D calculated suction-side break flow agrees well with the data. The present calculation was initiated with the water level (77 inches) within the data uncertainty band (73 16 inches), as opposed to the higher level (95 inches) in the previous calculation. However, the jet pump uncovery is now 2 seconds late. We believe the improved subcooled break flow calculation reflects the early annulus depletion history more accurately. Therefore, it can be concluded that, although within the experimental uncertainty band, the initiel water level in the annulus is still slightly high for the new calculation. This has no impact on the long-term response since the trends observed in the recalculation are identical with those of the original analysis reported in Reference IV.1-2. The long-term response of calculated and measured parameters are presented in Figures A.IV.1-9 to A.IV.1-20. The steam dome pressure (Figure A.IV.1-9) follows the experimental data well prior to the time rf LPCS initiation (Figure A.IV.1-20). However, there is a small (approximately 2 seconds) lag in the calculated timing for the initial pressure drop. We believe this is another indication of a high initial annulus inventory. Figure A.IV.1-10 reveals that the cubeccled bicwdown through the suction line was calculated to last about 2 seconds longer than in the experiment. This fact, a consequence of the higher annulus inventory, led to an extension of subcooled blowdown (as opposed to saturated blowdown) for a longer period. The lower energy depletion rate associated with subcooled blowdown allowed the pressure control valve (modeled in RELAPSYA) to maintain the pressure for the additional 2 seconds in question. This is confirmed by Figure A.IV.1-5, which shows that the steam flow was isolated about 2 seconds later in the calculation than in the experiment. Following LPCS initiation near the time of maximum HPCS flow (Figures A.IV.1-19 and A.IV.1-20), the calculated pressures become lower than the measured values. This is attributed to an overestimation of condensation rates by RELAp5YA when cold water is injected into partially voided volumes. This issue is discussed in 1 more detail in the response to Q.IV.11 and Q.IV.12. Figure A.IV.1-10 presents the suction-side break flow. Discharge coefficients of 0.7 (subcooled) and 0.6 (saturated) are used respectively. Agreement with the data is good, barr 'ng the slightly longer (2 seconds) subcooled discharge period caused by the higher initial annulus inventory. Figure IV.1-11 presents the drive-side break flow where good agreement is seen as well. Figures A.IV.1-12 and A.IV.1-19 show various regional inventory histories, which exhibit qualitative trends that are consistent with the previous calculation of Reference IV.1-2. Hence, the differences between calculated and measured values previously discussed in Section 5 of Reference ' IV.1-2 for these parameters cannot be explained in terms of code input (nodalization or initial conditions). These figures reconfirm the analysis presented in Section 5 of Reference IV.1-2 that after ECC injection, the ECC liquid tends to accumulate in the upper plenum and drains down through the bypass and guide tube into the lower plenum rather than penetrate the bundle. This effect leads to conservatively high bundle void fractions. The effect could result from an overprediction of vapor generation rate during the period when the bundle was mostly voided,
~
i which has been known to occur in other calculations with RELAp5YA. Figures A.IV.1-19 to A.IV.1-21 present the ECC flow rates. The differences between the measured and the calculated values are due to differences between the calculated and measured pressures. Figures A.IV.1-22 to A.IV.1-24 present the inner clad temperatures at the lower, mid-plane and upper bundle elevations, respectively. The qualitative trends are analogous to those observed in the previous calculation (Reference IV.1-2).
.b TABLE A.IV.1-1 Comparison of Calculated and Measured Initial Conditions for TLTA Test 6425/2 Initial Parameter Data SS Initiation at 2 MW Conditions Bundle Power (NW) 5.05 10.03 2.0 5.06 Steam Dome Pressure (psi) 1044 15 1043.8 1044.2 Lower Plenum Enthalpy 528 is 521 526.1 (Btu /lbm) Initial Water Level (inches) 73 16 76 77.6 Feedwater Enthalpy 41 12 42.9 42.9 (Btu /lbm) Bundle Pressure Drop (psid) 17 12 11.0 16.3 Steam Flow (1bm/s) 6 il 1.88 5.7 Feedwater Flow (lbm/s) 1.4 10.3 1.84 1.4 Intact Loop Pump Flow (lbm) 9.1 11 8.73 8.8 Broken Loop Pump Flow (1bm) 8.4 il 8.3 8.4 Intact Loop Jet Pump (1bm) 22 12 18.1 20.0 Broken Loop Jet Pump (1bm) 20 12 16.4 18.4 Bundle Inlet Flow (Ibm) 39 15 31.1 34.1 Lower Plenum Pressure (psi) 1071 15 1061.5 1070.1 i s
]} EO*
5 5 3 5 1 4 4 l - 5
~ 1
)
0 4 P D ( O ' L F m a t e T a E N $ R I L
,l l1' 3 w N m 1 l
o I F 2 s
,
- s O a M
E N S e U i c R f i i r O y r t 5 l e n s e E 2 t i t ,' t,! ; , ,! I e1!.f.l' -
!lli !'I 8 E e M d G' - I T
i S I t S e l E n I l- I e l d n u B A 1
;i! ,t t Y l' -l- .tl1 1!I- 3 -
S 1 P A. I
'w V I
E R ' A
/
',1
- e
( r u
, g E 0 i F
Ii l. .' ) * $ . * ;I 2 0 0 0 - 3 1 1 . j}=bI i 4v3
o eg# s e . i s e .. 1." , i
)
- 2 .
l
- i P
D ( O L F / 3
. !d T
I L / ' 3 1 2 S ' A P Y N U B R i 5 2 4 t 6 (
- t l
E B M r e e E 8 T l' T 5 S 5 E 8 T B W A Y : 5 2 P - A 1 L . E V R I.
- \ m '
A e r u g i F 1 r l i a . e *_ j}_Ii
. Sh='-"
e d e e e h
- e e e a
- =. .
8 t5 o 5*
*5 3 4 5
a 84
- =
"$ - n
~ I1 so z e o
E 5 u. m M
* - t I
o
=
ja d 5 so W e. M M w , 1 w e
$ s
- h b
D 4 00
-=- ___
_ . . . _ _ . _ _ ) . ., m D.4 # w - QG 8
>...- .ll 2 2*'
i (ses/wel 201d s
d68t
.Ili .
\" Ls i
)
8 6 P . D ( O L n
-j F e /
g E e B L T a s R t a w 2 E g , l o D QD _ F I . " U s s N G U M a R , e i b
- u T
e d i 5 1 m u 2 I s G
!it .l l' ! i, t o 6
r ! l: 1 : e sa l t T i T m m E T S 1 m l P e r e A w Y o L 5 P : A 4 L - E 1 R . V 4a I. A m . i F e r u g i o i-4 .
, a_
il=5i i
)i ' ! - I ,\
k
s . _ _ _ _ _ _ ICSI64h5 RUN 2 1 STEAM LI NE FLOW DPG2
=
?
f
'""'"" I'"1 l.f sk;I
,, lWs j-iuv\i : le
__ a _i 23
#e i l ' RELAPSYA t t l 3 l l r' i
}
, i l N
~
? ; .
y - j t' a. U
=
l I I .
! ,3 !
i \ I L l l . J -________
- i i ! l i i
- i. - - -os i e ,3 ,, 9 f tME lestl !
t t Figure A.IV.1-5: Steam Line Mass Flow Rate J l
} ,
-$ ga' p - -
l 2 g i I - ll . ' , , p 68 55 p-PP - DD WW ~V OO LL FF g ' t y i2 lla s p - j;.I :.II i! i I-PP _ HH - s 2 PP e
- V-t a
VV R
~
RR V w N DD U . V l o F R . l 2 s 2 s a M
- p m
5 u i P
;2' l; !i s c
a n
- i ' A
,I il : i ! i - ' ' f G .E o i
- MI t
- r T a
- J- n l
u E c I s r
% i c
e R I, N 6 1 A . Y V 5
, d ie P A ,
I. L
'l.
.1 , ;.!Y I 5
4' A E R . [- P A e r L u E g R i fl F i
\ Y - -
*\
I'
.,I e' " -
J s g
,I ,; .e ; ei t t !. '
w e
., 6.4 ll .!. 2 t1il:.I 2
i} 5I AO
9 . , - - +- e.. rss 6 h Id a s . 4 [ s - O g " a
~Wm 8 t
z, a $ a D l $
, X l 60 1 ._ 7
~
k 8 j
~
L y ..
< ~
.d >
n -
- 7 c m e o
\ %
60 en CAe _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ ___ q , ['
.3 .
O
...__w_,__- *
,g3 g Il. . I t a ll, s. Is* {III .7
- l ,1P 1 FLOW j
i l ! I I l i ' l s I
, 4 liS i
i
; ,,; , _ J .. i s__:' i i i.,
b i, V .y. j ; i l i
- 2 : ' l'%,% l 1
0 !
'it i
ii i
'i' !
! h I RELAPSYA
-- s/ -N l t,6 j i i
I- I '
!45 1
~"\ l l
g I i i
' l 1
\
. e
! # FLOsv CALCULATIO88 e peO LOf0GER RELIASLE l
I b'l \ lg ,;\ q. . I I
, i.%
gilf '9 '5. v, 1 h. - il
- _l I l f '. i _
., 12 0 l ! I'
- 3 5 g 13 IS Tea #E lesel Figure A.IV.1--8: Intact Loop Jet Pump Mass Flow Rate
3 sEIy3r n e o e e s a 2 o i a ENN MEE OLL DPP ~ j MRR AEE EWP T OP SLU n
! il 1 ,I m a
1 24 2 PPP
- 4 s f
e U r R u s 1 23 . s __ e r P m e 5 t c t d e s s y 4 - wi E i S b d n i T : I U 4 _ 3 9 1 L ' . i I
- V Tj' l I
NV
} .
5 -
. A f
_ e D T f r u g i liI I Il e2 j - 0 3 F A Y S A P Y A 5 L P E A R L E R w y d c c O n '- a s a
] 3E853*
i$'
i
^
l i I L :_. I bil'db l>Ull P
! UUC i l ut! FLU LBH/ CEC
'1 I
= I
.1 -
j e- RELAP5YA i 1 i
} }
g ~ l 5
% Gl\ .If n lf 0 o\v5'V f' '
Py h
~
y' )M , i
- l a
~ -10 70 150 230 319 Tia#E lasci i
j Figure A.IV.1-10: Suction Line Break Mass Flow Rate t i i I l
- l\
* ~IAe hg 3
s 3 2
- 2 o 0 1
3
=
f - C g E S y ~-
- / 4 M 4 B
L g - p O M L 2 - F t e 2 E Q - R a V w o W I R l F N D U s R s a 1 M
\S^' -
k a e
- r t
c B S h e 2 - 0 5 E e n I 1 M I+' ^ i G' _ I T L e I b A ^^:. _ i v r L l l w" D g A_ 4 :
% W_ l l
A EbwM 1
% V Y
5
' A 0 I.
7 A A L e E R _ r u g f i F V . I p
" - l l }L ' .N O -
'6 1
- ot t 9
- a. -
kgm- g
,ov1 oe j
!!ll! ' i {;'
(twehseg ggyn S S R .
,o n
I D Z LLI _J Q. E g m N$ % . x T 2 5
& C
$w s 5
- E m _- . . - - . . ._ g 5 s.
_m -- . - . . . - ... .
~
w u :s h s O 0 o
~
9
=>
< .CC
>* e iA lLe W
i e c a I o ~; ' a
*~
- 8 3
- s. -
N m o - 0 7 1 3
- ~
E L D N U , B y
;i,. !
4 S S 2 A M s N U d M s a R I k d l
,U y
e 5 !2I l 4 t I' !.
.p i t E
M I c s a l d n 6 I S L T E T ~ ;
- V I
A e Y 5" l1 A Y 5 p f s i E R n . o *-
~
s ik5 h
< i il
\
- I 4 ,
s x I
\
,\
g Sh N I I 2 i i ..
-5 M . .s
. J 4t 3 % .
' b 4
2 2 ' C *
\ ..
k L,
- - - - = . . . . . . -
@ ,R, is
- M I
<J C ~
E 2 6 D M 3:
.r.
E LS E I' LS i 2 k, '! g }
. Iw l .s
[. n .i lZ j I
.-e I
e
~
\ 4 e .
\ . 4
\ @
. ~ . - . . . _ . . 6*
- s g
,} .ec
*1
>= '
sn De 5 w
- j \
d , I I I "' 9 (***4:1 OsvM i
_iI,i= e s . g 3
,!I
- . el :
3 E B ' U T E g D I _ U M C 6v
^
I l l 'I 0 I 3 2 i
'l h 2 M --,
s s
-, a N M U d P
1 m N i u l F e b.t 5 _ t T c
,'1 J ; '
e s e ll ,!:i - i ' !!.- 0 t d 8, M, 5 1 f e e i u
-i I
/
l T G U / L 5 I i 1 1
< ^
V I.
, A e
r u
;i
* !l e I il g
l- i!: i lt .: ' i A
! I M F Y
5 P s 4 L E R e i
e s - e' .
i - s _k I*
i =*w ssvn 8 9 R ..
.-_y... , , ,
41 C r __.... . ___= R u, N .,
=
C_
>- e z x 3
2
.- \
=
i = C
- a m j .
^4 . .. _ _ . .. . . . .. . _ __.. p .s... . .. .
3 [ _:r .3 ..7
.e-
- g"
. 5. ..
h* [-. & b !- .1 _ _.-~.___ - . l
==___ -
3
- J v
u a ec a re p w sn 3 w a:
./
n.. O S S !! ** ~ . _- l l t=*ve ssvn S 9 R .. E > 5 2 Q. e w Q. Q. 3
-_ _ g
~2
$2 I 2
= -
3= 2 2 e __ E
- E b b A = b JL. . . _ _ _ __ . . . . _ . -. - g - g 2 , j &
.)
y e % ..
+ .....
/. ,
,s ..
>2 >
m k 5 g 5 w lee
>=
m D. 5 sa - M 8 o S !! si
- .I.l Omn
.I55I=
w 0 4 t 8 m ge
- 3 L
A T O T
! i1 0 3
2 S S 2 O M s s a N M U d R i 1 u l F l s e 5 t s c e e 2 i 0 f e V 4 t i; l 5 E 1 l 6 MI a T t T o T S E I : 8 1 1 V I. A e r Ii i M u A g Y i 5 F P - A - L . E
- R
,. 0
' 0 a-8' 0 3 1
=
_g. 93 , sY
_Ml $a s e0 1
- 3 5
6 P O _ W l . ; 0 i O tiI .l ,- l il:I 3 2 L F 2 _ S C 4 P e f H t U R a R 1
- w o
l F
~ s s
5 t c a 2 e a M
,f
- Il1 I liIi .; . 0 f t
; iiII i' 6 E S U
1 M C I P T H I U 9 E 1
- I
. 1 V
I. A
, e r
A u Y g I. i gi S i l* liI -i ,. l 4 Iii i'l 0 7 F _ A L E R i' 3 0 3 i 1 O-
.Iligm ie,
. i1 1 0 0 1
- 3 6
6 P D ~ W I l! I , 0
! itII ! i l' 3 O
~
2 L F J 5 C e P t U N L R a l w l l o F s s a
, - ) M
c e
e l , ,li . li
- 0 f S C
i ;I 6 E 1 P ib MI L T I U
- L :
I 0
- 2 1
V I. A A Y 5 e P r
!I.j Ij A j i l 0 u L 7 g E i R F 0
s 2 , 1 O-i) f 8
- h:
' O0 1
I
- 3 4
m1 3 - G P D 0 W ;.l ! ; u!1 :I II J 2 4t L F 2 I C e P t a N U L R R w
- o uI l
- l F
s s i a c s M 5 - 0 t a I I lg !lI - ! 5 1 E M C P 6' I T L I : [' 1 U 2 L - I 1 V I. A e r u g i t 0 F
,l8
,. 7 0
1
- 2 -
1 ,O I 1
- il: : l ' l '
_5E$5j"
. 0 e o s s e m 0 1
- - _ 3 8
11 893 0000M 1 1
...t 1l 1 l T P
e FFFF PPPPT A S WEWW SSSN y N 5555E . 3333L 0 n 3 i P 2 LLLL 5
) 3 EEEER 3 2 E n 2799W E i o
1 1 23O D t SSSSL O a N v N OOOO ( _ e U PPPP R _ l E A " - 1 2345 Y e 5 r P u A t a L,: r E R t e 5 c e p 2 ' 0 t e m 4 5 E e 1 M T 6 s T d a T - l C S E - -
~ e T . d i
s n
"y I
2 2
= 1 a
00egeO9O, Q, w 0 V 00egeO9O 7 I. A 00egeO9O , e 60egeO9O r u w 00GgGO8O i g 60OgGO9O F 00OgOO8O e 06OgGG0G # ' O I 0 0 e II!1g! " d
(My 3Wr11.. m ,1 h - k o
. a N(O N M e
- N O .% e 1 1 OOOOC E
....'w o p j
(A L t.L L ta. "' O. C. CL O CL 6-- G U) 3 W .2 U 2 U) Cn Cn Z Z j, z "{ w 55666 J G.
> g
- d JJJJJ s es m /
N U W U W W Ck' en - N w a bQ-07 2 W O~N2DO e $ enenenenen a o y z OOOOO v . 0 11111 -
-~m=ww . =
A. ss. 5 w i 3 m = -
-y' A
L S b t a - a
~
-{
o
.\ 1 w
4 c
- s ,.
.5
, 5 F@@@@@@@Gf
@@@@@@@@ n f ,
C
@@@@@@@@ 'd 4
. @@@@@@@@ , / g
@@@@@@@@ /
@@@@@@@@ a
- g
@@@@@@@G r e e e e e e e,,
e
. 2 I n' -
.......1 TEST 64 25 RUN 2
, I POS13 Ell 20 SW PF1.006 er 9 2 POS11 ELi 20 SE PF1.030 00000000 3 POS40 EL1 20 NW PFO.951 OO 4 POS62 Ell 20 NW PFl.024 OOOOOOOO '
gggggggg
- 5 LOWER PLEN SAT TEMP - see 09999999
'" OOOOOOOe OOOOOOOO
#00099999 o E k=
~.
5 5 4 4 5 5 g -mo t
# RELAP5fA (NODE 7) '
f a n ,s
, 7 I 4 n n
- - e e
-so ,, 23o 380 TIME fase)
Figure A.IV.1-24: Inside Clad Temperature-Elevation 120 in.
V. ADDITIONAL QUESTIONS CONCERNED WITH RELAPSYA BWR APPLICATIONS Page 50 of Reference 10 states: "The separator component, SEPARATR, provides a preliminary model to simulate the steam separation process in steam separators or dryers. This component is similar to the BRANCH component except for the following. At least one junction must be described in the input data. The first junction described is assumed to be the exit from the top of the separator. This junction is forced to have one hundred percent static quality for positive outflows. If reverse flow occurs through this junction or if the separator fills with liquid, then the convected quantitles revert to those of the upstream volume. The separator model is described in Section 3.2.8 of Appendix B and and Section 7.9 of Volume II.
"An alternative method was used for modeling the steam separator in the Two-Loop Test Apparatus to obtain steady-state initialization. This method is described in Section 5.2.3.3 of Volume III, and provided reliable results."
0.V.1 How will steam separators and drfers be modeled for Vermont Yankee and how are the models to be used a,sessed? A.V.1 The 129 steam separators in Vermont Yankee are modeled as one equivalent steam separator in the RELAp5YA System Model using a BRANCH component. Likewise, the 6 steam dryer panels are modeled as one equivalent steam dryer using a BRANCH component. Input parameters for each are based upon plant drawings and standard references for loss coefficients. For steady-state initialization runs, time dependent junctions and time dependent volumes are used downstream of the steam separator volume (BRANCH component) to separate the two-phase mixture from the core and upper plenum regions. Saturated steam is directed to the intermediate steam plenum below the dryer. Saturated liquid is directed to the downcomer region near the feedwater spargers. This modeling method has been assessed against plant data obtained at normal operating conditions and yields good results. Specifically, the liquid carryover fraction to the steam lines, vapor carryunder fraction and temperatures in the downcomer, and the regional mass flow rates calculated by the code agree with plant heat balance data. When a LOCA is initiated, the time dependent junctions and time dependent volumes are isolated. The two-phase flow passes from the separator volume to the intermediate steam plenum or the separator skirt through separate junctions at their respective elevations. This modeling method was used in the TLTA assessment cases and yielded good results during those accident simulations. O.V.2 How are core sprays modeled and assessed and how much penetration is assumed? A.V.2 The Low pressure Core Spray (LpCS) Systems are modeled using a TMDpJUN component. This component draws suction from a source (torus liquid) volume and injects into the lower region of the upper plenum. This is the same configuration that exists in the plant. The TMDpJUN component contains a table of LpCS mass flow rate as a function of the computed differential pressure between the upper plenum and the torus. This table accounts for the following:
- 1. LpCS pump performance characteristics.
- 2. Wall friction and form losses within the pipe system.
- 3. Hydrostatic pressure drop from the torus to the injection location.
-100-
The spray distribution pattern in a dry steam environment in the upper plenum is not explicitly calculated. This is based upon the experimental results from the BWR Refill-Reflood Program (References V.2-1 and V.2-2):
"The residual upper plenum pool effectively accomplishes the distribution of ECC liquid to the top of all channels.
Hence, the previously assumed requirement for appropriate ECC nozzle design to distribute liquid in a dry steam 7 environment is not necessary." (Page 8. Reference V.2-1)." References (V.2-1) Dix, G. E., "BWR Loss of Coolant Technology Review , " Volume I , Proceedings of the Second International Topical Meeting on Nuclear Reactor Thermal-Hydraulics, Santa Barbara, California, January 11-14, 1983 (published by American Nuclear Society, LaGrange, Illinois). (V.2-2) Sutherland, W. A., " Condensation Effects in Large Scale BWR Safety Experiments," Volume I, Proceedings of the Second International Topical Meeting on Nuclear Reactor Thermal-Hydraulics, Santa Barbara, California, January 11-14, 1983 (published by American Nuclear Society, LaGrange, Illinois). 0.V.3 How have single- and two-phase flow modeling for BWR valves been assessed? i: A.V.3 The modeling techniques and assessments for the four safety / relief valves < and the two safety valves in the VYNPS have been addressed in the answers l' to Qu*stions Q.IX.2, Q.IX.3 and Q.IX.4 of the 197 general RELAPSYA questions from the NRC (Reference V.3-1). 9
-101- ,,- , - . . - - ,--,,.y - -.- - - - - - - - , - - . , ~ . ,-- - - , , , -e - - . . - --
The MSIVs, located downstream of the flow limiters in each main st eam line, are modeled as a MTRVLV component. The trip logic provides signals to initiate MSIV closure. The MSIV closure time is physically adjustable between 3 to 10 seconds. A conservative closure time is used for licensing analyses. The valve throat area is essentially the same as the main steam line flow area; therefore, these valves do not control the steam line flow rate except during a very brief period prior to closure. Our main interest is to assure that MSIV closure occurs at the proper time. This is done by checking the computed results. The motor-operated discharge valves in each recirculation loop are also modeled as MTRVLVs. Plant data is used to set the characteristics in a manner similar to that described for the MSIVs. Again, our main interest is to assure that these valves close at the proper time. Reference (V.3.1) Letter, G. Papanic (YAEC) to J. A. Zwolinski (USNRC), " Response to NRC Questions on the RELAP51A Computer Program," 2.C.2.1, FYR 85-121; November 1, 1985. O.V.4 . How have the recirculation pump models been assessed for single- and two-phase flows? A.V.4 This question has been addressed in the answers to Questions Q.I.7 and Q.V.10 through Q.V.16 of the 197 general RELAPSYA questions from the NRC (Reference V.4-1). Additional assessment information on the TLTA recirculation pump behavior is contained in Appendix A.IV.1 of this document. Specifically, the intact and broken loop jet pump drive line flow rates presented in that section are a direct consequence, in part, of the recirculation pump histories.
-102-
Reference (V.4.1) Let.ter, G. Papanic (YAEC) to J. A. Zwolinski (USNRC), " Response to the NNC Questions on RELAP5YA," 2.C.2.1, FYR 85-72, July 1, 1985. Page F-17 of Reference 13 states: "GE should provide an evaluation of the uncertainties in their ability to predict system pressures, mixture levels, clad temperatures (including lower plenum flashing) and reactor vessel inventory distribution." O.V.5 Clarify how YAEC has evaluated the same uncertainties for RELAP5YA BWR applications? A.V.5 YAEC proposes to use a BWR LOCA Evaluation Model that complies with Appendix K to 10CFR 50.46. Modeling uncertainties have been extensively addressed in the RELAP5YA code assessment work documented in Volume III of our Topical Report YAEC-1300P and in our responses to the 197 general RELAP5YA questions and the 39 BWR related RELAPSYA questions from the NRC. This complies with Paragraph II.4 of Appendix K which states: "To the extent practicable, predictions of the evaluation model or portions thereof, shall be compared with applicable experimental information." Pages 271-272 of Reference 10 discuss the Appendix K requirement I.D.6 for convective heat transfer coefficients for boiling water reactor fuel rods under spray cooling compared with the RELAP5YA approach:
"Following the blowdown period, convective heat transfer shall be calculated using coefficients based on appropriate experimental data. For reactors with jet pumps and having fuel rods in a 7 x 7 fuel assembly array, the following convective coefficients are acceptable:
l
-103-
- a. During the period following lower plenum flashing but prior to the core spray reaching rated ficw, a convective heat transfer coefficient of zero shall be applied to all fuel rods.
- b. During the period after core spray reaches rated flow but prior to reflooding, convective heat transfer coefficients of 3.0, 3.5, 1.5, 1.5 BTU-be~ -it~ F~ shall be applied to the fuel rods in the outer corners, outer row, next to outer row and to those remaining in the interior, respectively, of the assembly.
- c. After the two-phase reflooding fluid reaches the level under consideration, a convective heat transfer coefficient of 25 BTU-hr~ -ft~ -
F~ shall be applied to all fuel rods. The acceptable model described above will not be used in RCLAP5YA calculations. Instead, the RELAPSYA heat transfer algorithm will be used.** 0.V.6 Please provide the data that justifies that the RELAPSYA heat transfer coef ficients are accurate for boiling water reactor fuel rods under spray cooling conditions so that the Appendix K recommen'etions do not have to be used. A.V.6 To justify the RELAP5YA heat transfer coefficiente for BWR fuel rods under spray cooling condition, data from the TLTA large break tests are used. Reference V.6-1 presents heat transfer coefficients derived from the data of TLTA Large Break Test 6422/3, which used average bundle power and average ECC injection. Similar data have not been reported for Test 6425/2 which was used for RELAPSYA assessment. However Test 6422/3 and Test 6425/2 were conducted with similar initial conditions and showed similar results. Hence, heat transfer coefficients derived from Test 6422/3 are compared to those calculated by RELAP5YA for Test 6425/2.
-104-l l
l
Figure V.6-1 shows the comparison in the mid-bundle region. Also shown in the figure are the heat transfer coefficients recommended by Appendix K. Figure V.6-1 shows that the RELAPSYA heat transfer coefficients, being more mechanistic than the Appendix K recommended values, follow the data trends more closely. However, they remain conservative, for conditions typical of BWR fuel rods under spray cooling. Figures V.6-2 and V.6-3 compare the heat transfer coefficients derived from the data of TLTA Test 6422/3 to those calculated by RELAP5YA for TLTA Test 6425/2 at two other bundle elevations. Again, the comparison shows that the calculated heat transfer coefficients are conservative. References (V.6-1) Lee, L. S., Sozzi, G. L. and Allison, S. A., "BWR Large Break Simulation Tests, Volume 1: Experimental Results and Analysis," EPRINP-1783, July 1982. Pages 272-273 of Reference 10 discuss the Appendix K requirement I.D.7 for boiling water reactor channel box heat transfer under spray cooling compared with the RELAPSYA approach:
"Following the blowdown period, heat transfer from, and wetting of, the channel box shall be based on appropriate experimental data. For reactors with jet pumps and fuel rods in a 7 x 7 fuel assembly array, the following heat transfer coefficients and wetting time correlation are acceptable.
- a. During the period af ter lower plenum flashing, but prior to core spray reaching rated flow, a convective coefficient of zero shall be applied to the fuel assembly channel box.
- b. During the period af ter core spray reaches rated flow, but prior to wetting of the channel, a convective heat transfer coefficient of 5 BTU-hr~ ft~ - F' shall be applied to both sides of the channel box.
-105-
l 1 lllll1I lll 0
- 7 7 6 6 3 5 5 A/
Y2 5 2 P4 _ - _ - 0 A6 L 0 E t 3 R s e nT i A s T t L 0 nT e I 5 if
. ) )
2 i c o
.L6L 1 f a LE E y- f t E e g6 e a d".lo od "9 00 No1 0 o V 7 1 1 d .
c h A of 0 rt ei t t Y oh e a aStP t Re d s d sl 0 f w s aaA" M o o 2 n2 R R a/ t t L1 E r5 aaE8M r r DDR( E (G e e ) t 2 4 n t t 6 r n a
- o e g
- C C 0 (
mhet se i 5 1 Eld eT
'I* lI1 A
- MuL nT b T 5 1 l
- lf oof A sl .i
'0 n n
' I 0 o o si A 1 i t r a al p u s.i mm oi C s d 1 i i tM A t
A
\I~ *l:
1 0 5 1 6 V i }I t p - t .\ - e r k-k ! l .I fiu
.! u g
- - ~ 0 i
F 0 0 0 0 0 0 0 0 0 0 1 0, 0, 0 1 1 1
- 0 0 0 1 1
- N "tsa * .5gm8smE@ ~2 l1'
j
- 100,000 "
Data at 71" EL
- - - - Da ta a t 79" EL g RELAP5YA Node 5 10,000 (61" to si" EL) _
": ,\
- 5 \ .
"i. \
- (
e I R 1'000 - k ~ 5670 t I'\ l 't Is 5 I l i 100 I
, 2 m
A( -
' 8 l I
{ w
'g _ja i
5
= 10 -
e t A - 56.7 li= A l 1.0 ; 0, 50 100 150 200 250 300 TIM E (soc) Figure V.6-2: Comparison of bundle heat transfer coefficients in RELAP5YA simulation of TLTA Test 6425/2 with data of TLTA Test 64-2/3
t 1 4 100,000
- - - - Data a t 100" EL k A RELAP5YA Node 7 j
10,000 g ~ cioi" to 122" eL) j ! . NA j 8 \ i o N l y 1000 -g - 5670 m
- I /
E
/
= l 4 /
38 '
!i I l
g 100 - sit i - 8 C i t 3 $ 8 5 ,' A 5 10 ~ A A i 5 s" 56.7 l H d x I 1.0 l A A l 0 50 100 150 M 250 300 Figure V.6-3: TIME (m) Comparison of bundle heat transfer coeffients in RELAPSYA simulation of TLTA Test 6425/2 with data of TLTA Test 6422/3.
- c. Wetting of the channel box shall be assumed to occur 60 seconds after the time determined using the correlation based on the Yamanouchi analysis (' Loss-of-Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors,' General Electric Company Report NEDO-10329, April 1971).
The acceptable correlations described above will not be used in RELAPSYA calculations. Instead, the procedure described for Requirement I.D.6 will be followed." 0.V.7 Please provide the data that justifies that the RELAPSYA heat transfer coefficients are accurate for boiling water reactor channel box heat transfer under spray cooling conditions so that the Appendix K recommendations do not have to be used. A.V.7 Data on heat transfer coefficients for channel boxes under spray conditions have not been reported in the TLTA test documentation. Hence, a direct comparison of calculated and measured heat transfer coefficients for the fuel channel box has not been carried out for the TLTA tests. In the RELAPSYA simulations of the TLTA tests, the fluid inventory in the bundle region was calculated conservatively (Figure 5.2-16 of Reference V.7-1). A similar result was also obtained for the annulus region. Hence, for the TLTA tests, RELAP5YA calculated the fluid inventory conservatively on both sides of the fuel channel box. The heat transfer coefficients used in RELAP5YA are based on best-estimate models. In EM calculations, the heat transfer lockouts are applied, as per Appendix K regulations. Overall, this would lead to a conservative calculation of the heat transfer coefficients for the fuel channel box.
-109-
References
-(V.7-1)~ Fernandez,' R. ' T. et al. , "RRLAP5YA - A Computer Program for LWR System Thermal Hydraulic Analysis - Volume III," YAEC-1300P, January 1983.
y
-110-II}}