ML20040A404
| ML20040A404 | |
| Person / Time | |
|---|---|
| Site: | 05000561 |
| Issue date: | 03/30/1976 |
| From: | Israel S Office of Nuclear Reactor Regulation |
| To: | Rosztoczy Z Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML111090060 | List:
|
| References | |
| FOIA-80-515, FOIA-80-555 NUDOCS 8201210029 | |
| Download: ML20040A404 (16) | |
Text
i
~
e
)
c 4p UNITE 3 STATES
- a NUCLEAR REGULATORY COMMISSION g
j WASHINGTON, D. C. 20555
+
(
%*****j*
ggg 3 g Et6 Zoltan R. Rosztoczy, Chief, Analysis Branch, DSS THRU: Thomas M. Novak, Chief, Reactor Systems Branch, DSS B&W CODE " TRAP", TRANSIENT REACTOR ANALYSIS PROGPAM FOR OTEAM AND FEEDWATER LINE BREAKS Enclosed is a description of a B&W code which was included in the B-SAR-205 PSAR.
Since new methods are involved in this case, it would appear appropriate (as guided by Standard Review Plan 15.0) to initiate a review covering the model description, data correla-tions and empirical relationships, solution techniques, sample problems, experimental verification, and comparative calculations.
It is requested that such a review be conducted by the Analysis Branch.
In the interest of maintaining the current B-SAR schedule, it is suggested that your initial set of inquiries be submitted at the Round One milestone (April 26, 1976).
6fC 1
S. 1srael, Section Leader Reactor Systems Branch Division of Systems Safety
Enclosure:
Code Description ec:
D. Ross T. Novak T. Coxe G. Mazetis J. Watt 8201210029 810403 PDR FOIA MADDEN 80-515 PDR
B-SAR-205 2/1/76 Rev 0 O
,..i APPENDIX 153 TRAP - Transient Reactor Analysis Program f or 0
Steam and Feedwater Line Breaks 9
5-SAR-205 2/1/76 Rev 0
~
,3.1.
Method of Solution b
15B.1.1.
Introduction Th2 transient reactor analysis program CIRAP) was developed to permit analysis thermal-hydraulic behavior during an accidental depressurization of transient The TRAP computer code is cf the secondary side of a PWR steam supply system.
primary system loss-of-coolant. code through the l
cn extension of the CRAFTi cddition of a detailed steam generator model and by provision for substantial representation of secondary steam and feedwater piping. TRAP requires a user input control volume-fl,ow path network model of both primary and secondary system elements identical in type to that utilized in CRAFT primary coolant system studies as illustrated in Figure 15B.1-1.
The method and logic of CRAFT cre applied to solve the equations of conservation of mass, energy, and momen-2 tum in both primary and secondary networks.
TRAP relies on CRAFT 2 pressure the end property scarch subroutines to determine the fluid state throughout system.
In TRAP, each steam generator is represented as N primary, secondary, and tube metal volumes. These control volunes are determined by dividing the steam gen-erator height into segments as shown in Figure 15B.1-2.
15B.2.
Steam Generator "5B.2.1.
Initialization
..a initial steady-state energy balance is performed on primary and secondary Based steam generator control volumes to establish precise node enthalpfes.
on user input values of the heat rates at each control volume, and taking the path enthalpy to be that of the associated upstream node, the energy equations
=0 WhHL + Qi - W hP P,1 P
=0 P Pel + Q2 - W hP P,2 Wh Wh
+Q
-Wh
=0 P P,n-1 n
p p,n Wh
+Q
-Wh
=0 p p,N-1 N
p p,N cre solved for the primary enthalpics b
,n.
The same balance is applied to thefeebwaterinlet.
the secondary side, starting at h
=0 MhsW+QN - s s,N
=0 s s,N + O
- s s,N-1 N-1 t
-Wh
=0 a s,n+1 + Q Wh n
s 8,n
B-SAR-205~ ' - ~ ~ -
Rav 0 2/1/76
=0 Wh
+ Qt - W hs s,1 de-a s.2 Control volune fluid temperatures, energies, and saturation properties are termined as functions of pressure and enthalpy.
Two options are offered for calculation of initial primary side heat transfer coefficients and tube metal temperatures:
User input values of tube metal' temperatures are used to calculate the initial primary heat transfer coefficients when the user has input H 1.
p
= 0.0.
'NT ZT (T"'" - T#'") - A 2
H
=Ai Q
P,n p,,
User input values of the primary side heat transfer coefficients are used to calculate initial tube metal temperatures when the user has input 2.
T
= 0.0.
g
'A I
/NT ZT T
=T
+0
+A2 tn p,n 7,n (H
p,n where ggy-A1 = 1/nD,
i O
2 = in(D /D )/2nK,
A 2
3 4p.
Di = tube ID, D2 = h(D3+D)*
1 D3 = tube OD, K = tube conductivity, NT = number of tubes, ZT = control volume height.
transfer coefficients are determined from the tube Initial secondary side heat metal and secondary fluid temperatures.
'NT ZT (T
-T
)
H
=A/
- Au 3
8'"
s,n D)
In(D3 2
1 Ag 2nK
=
A3
- 39 3 t
Initial and steady-state control volume mixture heights are assigned _according to the node enthalpics:
h < h, ZM = ZT r
n f
n ost h 5h 5 h, ZM = ZT f
n g
n Babcock & \\Milcox
.,=,
I
~.-. -.-
B-SAR-205 2/1/76 hv 0 h > h, ZM = 0.0
{ ;..
n g
O I
The switch to transient calculations occurs at a time value input by the user.
Prior to that time, all heat transfer coefficients are carried at initial cteady-state values.
Phase stratification is precluded during the initialized cteady state, and phase distributions are taken to be homogeneous within each l
ctntrol volume.
15B.2.2.
Transient Analysis Transient primary side heat transfer coefficients are determined for each nodal segment based on the associated tube metal temperature, the primary vol-l une fluid temperature and state, and the average mass flow.
l i
1.
Subcooled fluid, all Tt,n, Tp,n forced convection l
Saturated fluid, T T
p tm Values of the convective heat transfer coefficient are based on the ini-tial steady-state value and the average mass flow through the control volume.
i
' 0. 8 go H
=H0 P,n P,n p,n go P, n.
The average mass flow is taken to be the arithmetic mean of the mass ad-r dition and extraction rates during the_ time 1.tep._ _
2.
Saturated fluid, T
>T
= TSAT
+ nucleate boiling s
t,n p,n p,n Nucleate boiling heat transfer coefficients are calculated using the Roh-senow correlation.
The Rohsenow factor is assumed to be given by l
3 Cu
'C u'
~ '
l 3
R = (2.5 x 1010)
(p _ p )
- 1. 5 p
fg so that the boiling heat transfer coefficient is expressed I
p,n)2, H
= R(TWALL
- TSAT P,n p,n A pressure fit to the Rohsenow factor is used in TRAL' and is represented in coded form:
ROHSE = ((((((-4. 6165E-19*PBARR+2.754E-15)*PBARR-l 6.667E-12)*PBARR+8.331E-9)*PBARR-5.58E-6)
- PBARR+ 001754)*PBARR+.008247)*PBARR+9.5.
l The tube wall temperature, TWALLP'", is calculated at each time step from the heat balance:
Babcock t.Wilcox u n _,.
~
B-SAK-ZU)~' ~
/
- 2/1 ?6 Rev 0
)-
AT" R*A+(6T )3 =
1 A*R+(AT,)2 + Rm where R = ROHSE, A = NT ZT,/Ag,
= A /NT ZT,
R 2
a n
AT = TWALL
- TSAT W
p,n p,n, AT
=T
- TSAT m
t,n p,n This equation is solved directly for AT, allowing the tube wall temperature to be calculated TWALL
= AT + TSAT p,n w
p,n 3.
Superheated fluid, all t
,T
+ superheat forced convection p
transfer coefficient applied in superheated regions is calculated The heat from the relationship recommended by Dittus and Boelter :
h = 0.23 - (rcd )0 8 (Pr)0.4, h
h
)'
The Reynolds and Prandt1 numbers are calculated using fluid properties and 0
the average mass flow, flow area, and hydraulic diameter of the associated control volume.
The primary side heat transfer in all regimes is calculated by p,n)NT ZT (T
-T t,n n
q p,n A
3
^2 11p,n llent transfer coefficients applied to the secondary side during the transient are determined as follows:
1.
Subcooled fluiu, all T '", T"'"
forced convection Saturated fluid, T
>T sn t,n The convective heat transfer coefficient is based on user input values specified for the initial mass flow rate:
90 ' O,8 0'
s,n
};
11 s,n s,n w
where the everage mass flows are calculated in the same manner as those of the primary side.
Babcock & \\Vilcox 15B-5
~
(O 2.
Saturatsd fluid Tt,n, T TSAT
+ nucl=t3 bsiling a
o,n 0,n Nucleate boiling coefficients are calculated using the pressure fit to the Rohsenow correlation previously described.- Again, tube wall temperatures are directly solved at each time step and used to calculate the surface
[
conductance.
i 3.
Superheated fluid, all Tg,,, T,,,+ superheat forced convection The Dittus-Boelter relationship is used to determine all surface conduct-ances to superheated control volumes.
In stratified secondary control volumes, where ZM, < ZT, the rate of heat transfer is given by ZH (ZT ZM )
+
s,n " NB ZT FC ZT n
n H
= nucleate boiling coefficient, NB H
=~ convective surface coefficient.
g To prevent minor discontinuities in heat transfer coefficients when the switch from initial steady state to transi~ent analysis occurs, heat transfer coeffi-cients are normalized in each regime against the values calculated at steady state. This normalization is performed in each control volume through a con-stant multiplier applied to the surface conductance.
The.value of the coeffi-g cient multiplier is established one time step prior to transient calculations.
The noronlization procedure is applied only to those control volumes in which heat transfer occurs in the initial regime.-~
~
.iecondary side heat transfer for all regimes is calculated by (T
-Ts,n)NT ZT t,n n
q
,s,n A 3
+ Aq Hs,n Tub 6 metal temperatures are evaluated at each time step through finite dif-ference solution to the lumped capacity equation:
dT
'" = B (T
-T t, n) + B (T t,n) 2
-T t
dt p,n s,n where NT ZT" B1=
,g C
+A2 Hp,n s
NT ZTn B2" rA F
3 C
+A g
q
,s,n 15B-6 babcock & WiiC0X
.,.,.o
~ ' ~~'
~
Rev 0 C = total tube heat capacity, input.
(~Mpresenting the time derivative by a forward difference leads to the solution:
- I = T"t,n + [B g (T"
- T*t,n) + B (T"
- T*t,n))6t.
T 2
tn pn sn buring the initial steady state, it is assumed that slip flow between vapor cnd liquid phases is accounted for by the initial net mass, energy, and mo-mentum balance.
During the transient, however, the relative vapor velocity may be the primary means of phase separation. The rate of phase separation in l
ccch steam generator control volume is taken to be d(GL )
VBUB l
dt
" " n ~ ZM GL = stratified vapor mass, MB = mixture bubble mass, n
f ZM = mixture height.
a To preclude the existence of liquid-vapor strata in the lower control volumes, separated vapor is assumed to displace liquid in upper generator control vol-The vapor mass transferred upward from node n is calculated:
i umes.
VBUB GL = MB
" 6t 0
ZM, 1
n n
where VBUB = bubble rise velocity.
The change in total mass due to phase separation is determined:
TM = TM - GL - LG
+ GL
+ LG n
n n
n-1 n-1 n
I TM = total mass, GL = vapor mass transferred to volume n+1, LG = liquid mass displaced to volume n, GL _3 = vapor mass transferred to volume n, t
LG _3 = liquid mass displaced from volume n.
9 The vapor transferred cannot exceed the liquid volume available for displace-ment from the adjacent upper volume.
0b n+1( f}n+1 g)n' n
The liquid mass displacement downward is given by LG = GL (V ) /(V )
,urr
~
Rev 0 The energy balanc'e maintained for secondary volumes is calculated:
w h, - [ W,,g out h
+
Un"O+
in g n-1 g n-1 n
t
+ LG,(h )g - LG, (h ), - GL,(h ),.
g f
g Two bubble rise models are available by user input:
1.
Redfield variable bubble rise velocity, input VBUB = -1.0.
j 2.
Constant bubble rise velocity, input by user VBUB 2 0.0 ft/s.
15B.2.3.
Feedwater Path Model Capability is provided for feedwater delivery by either one or two feedwater psths to either one or two specified secondary system control volumes.
At sich time step, the feedwater mass flow rate is determined by solution of the feedwater path momentum equation:
b NU
[Ag dt S
D + oP P
-P pump where
[f=f
, evaluated'for each feedwater path and input by the user, ft-1, W = feedwater path mass flow rate, lbm/s, f
Pg = feedwater pump suction pressure, psf, P = delivery node pressure, psf, D
= feedwater pump head, psf.
At each time step, the pressure PD and mass flow rate W are carried from the calculations of the previous time step. The pump suction pressure is obtained by interpolation in a user input table of suction pressure and enthalpy versus time (and/or time af ter scram).
Feedwater pump head is calculated using the volume flow rate based on the mass flow from the previous time step and the pump speed obtained from the ap-propriate user input speed versus time tables.. The volume flow rate is based on the suction side density calculated from input values of pressure and en-
'thalpy.
Assuming the feedwater pump behavior to occur as a sequence of dynamically similar states, the pump head 11 may be determined from the flow Q and speed N by applying; the homologous curve relationships.
2 2
(h/a )g = (h/a )2 when (v/a)g = (v/a)2 and 2
2 (h/a ) g = (h/a ) g when (a/v) g = (c/v)2
(
15B-8 Babcock & Wilcox
._.' 'B-SAR-205~ ~~~~ -
Rev 0 a = N/N, the normalized speed, R
b v = Q/Q '
I R
l h = H/H, the normalized head.
R i
Tha subscript R denotes values at the rated or best efficiency point of opera-tion.
At each time step, the quantities v/a and a/v vill.be calculated and cpplied with the following user input pump tables to calculate the nomalized hrd:
h Vs v at a = +1.0 for 0 s v s 1.0 when v/a < l.0
. h Vs a at v = +1.0 for 0 s a s 1.0 when a/v < l.0.
Far v/a = (v) h/a2= (h) table cr, when a/v = (a) table, h/v2= (h) table
- The pump AP is calculated from the normalized head and the suction side density:
6P
=h H
p pump R
S.
The feedwater path momentum equation is then solved by direct integration to determine the change in feedwater mass flow rat.e.
The rate of energy delivery 0
- o the selected secondary node is given by the product of the input suction de enthalpy and the calculated feedwater mass flow.
15B.3.
Conduction From Vessel Wall Associated with each control volume is a mass of metal repre.senting the metal adjacent to the water in the control volume.
This metal does not include the fuel pins since the calculation is handled independently.
Tht. heat capacity of the metal (me), and the heat transfer coefficient (U.\\)w from the average metal temperature to the average water temperature are input and remain con-stant during a run.
However, an option is provided that makes the heat trans-fer coefficient proportional to the mixture height in the control volume:
C = (UA) (T
-T
}
er or C = (UA) (T
-T
)(Z /Z )
- dl T
The metal temperature is calculated for the next time step:
T
=T
- (C)(6t)/(me) y 11 s
Babcock t.Wilcox 15B-9
^
B-SAR-205 2/1/76 Rev 0
'iB. 4.
References 1
W. L. Jensen, B. H. Hopper, and P. W. Daggett, CRAFT - Description of Model for Equilibrium LOCA Analysi.s Program, BAW-10030, Babcock & Wilcox, October 1971.
2 R. A. Hedrick, et al., CRAFI2 - Fortran Program for Dip, ital Simulation of a O
Multinode Reactor Plant During Loss-of-Coolant, BAW-10092, Babcock & Wil-cox, March 1974.
3 J. P. Holman, Heat Transfer, McGraw Hill Book Co., New York, New York (1968).
G 0
v
V V
W
,y
-j e,
s Figure 15B.1-1.
Example of Consolidated Base Case 8
0 car 0,
_i,.
i.'~
O i!
i se e
r-m :
,O@"
p,i gI y
Q g
- tad 3
'6 ai 2
3 s
89 e
m(m)'-Q
@ l1 as as
=
2.
as 1
g r--2 2
g i -@
g a
u a
a= ;;e -
2-e
~
l" e
7f g
Ot es 7
- 1 a
e e -$ M,c.,3
>-- -@@S
~I
~
sr y
U U,g
_g ss e
e 3
e Ku f6
-9
-w888
_8 O
n r- - - 1 O
l M
l
-~
O s
CS..)
e
-v4 (rwtv)
T' 8O O L _ _J 6
h,o e
~@
i,, i e
i,, i e
L,J:'"
g"d e.. "s g c.,o
-j so
@ l si
}--N--j st l l
'5
.r... ~e.19,,,. 3 4
76 tu m
cr 8
y o
x ta P
Sunnary:
76 nodes (includes one " containment". node and one " turbine" node).
(
107 paths (includes 24 leak paths, 13 variable area vs time; 5 check valves; 4 fill paths; yh 2 pump paths; 1 core path; 1 surge line; 57 regular paths) 1 feedwater path.
<g y
o vi x
O
l
,2/1/76
.B-SAR-205
{
l Rev 0 O
r I
e
(
1 e
Figure 15B.1-2.
Steam Generator Representation Tube Hot Leg Nodes
[
Q 9
1 1
1 1
r-U o
Q Q
2 2
2 2
r Y
h Secondary
(
Primary Side l
Side 3
k 3
3 V
s o
I i
i I
i li v
E E
N N
N N
t t
j s.-
l 15B-12 Babcock & Wilcox
,,mv-w-,
e-
u BSAR-205:
REQUIRED DATA STILL TO BE SUPPLIED BY B&W AS OF 02/16/77 -
7 y
o I.
Containment Design Basis LOCA Calculations 1.
Increased break spectrum 3
2.
Sensitivity study on effect of containment pressure 3.
24 Hr. data Containment Subcompartment Calculations l
1.
Provide mass / energy release data using appropriate correlation l
such as modified Zaloudek, for subcooled flow 2.
Noding study in primary system l
Steam Line Break Calculations 1.
Demonstrate that TRAP 2 is adequate for small break analyses:
r f
a) by appropriate use of a condensing heat transfer coefficient b) by using finite bubble rise velocity that correlates with at least one set of existing data c) by adjusting small break flow to assume no entrainment d) by completing several sample cases (for different break sizes) using TRAP 2 and a) amW c) above fhru.
r l
l l
i"E
})ockEr No. STW s'o-561 ; BSAR-Zof STEAM LINE BREAK ANALYSES May 14 '76 In request number 222.9, staff asked for steam separation assumptions used in TRAP and why they were conservative and asked for heat transfer assumptions used in calculations of reverse heat flow.
July 20, '76 Amendment 2 response by B&W did not address conservatism of separation assumptions or the assumptions on reverse heat flow.
August 27, '76 In 222.11 staff noted that answers to 222.9 were r
inadequate and asked specifically for 1) a justified heat transfer coefficient from secondary liquid to tube walls; 2) staff also asked separately for a comparison to an appropriate condensing heat transfer coefficient.
Staff asked for response by October 18 (7 1/2 weeks).
6 September 16, '76 B&W letter stated response to 222.11 would be in by November 15.
November 18, '76 Staff received letter response to 222.11 as advance copy, to be identical to Amendment 7.
7 December 10, '76 Relative to B&W November 18 submittal, staff telecopied requests for infonnation to B&W; 1) stating that forced convection coefficient was inadequate for secondary side heat transfer in intact generator 2) asking for the
" detailed SG performance study referenced in response to 222.11(1) and a justified minimum value for
" reverse" heat flow; 3) stating that zero bubble rise velocity may not be conservative for small breaks and asking for justification thereof.
December 17, '76 B&W submitted Amendment 7, documenting 11/18 response to 222.11 and presenting for the first time a response to 022.7.
December 27, '76 Staff formally issues requests telecopied on 12/10, as request numbers 222.18,19, 20 and 21.
January 15, '77 B&W submits Argendment 8, states intent to respond to 222.18 through 21 on January 31.
(
l i
\\
S6L e
F.
s January 28, '77 B&W telecopies draft responses to 222.18 through 21.
February 04, '77 B&W documents responses to 222.18 through 21 in Amendment 8.
On this date staff informs B&W (telephone) that 222.18 is OK but responses to 19, 20 and 21 are not responsive.
February 08, '77 Cox (NRC) and Happell (B&W) discuss information needs and NRC position on 222.19-21.
NRC position is that staff concerns raised in this series of requests must be resolved prior to PDA degision.
B&W cannitment to resolve to satisfaction of staff at some time after PDA issuance is not acceptable.
-