ML19316A491
| ML19316A491 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 05/27/1970 |
| From: | Cady K US ATOMIC ENERGY COMMISSION (AEC) |
| To: | Schwencer A US ATOMIC ENERGY COMMISSION (AEC) |
| References | |
| NUDOCS 7912200745 | |
| Download: ML19316A491 (7) | |
Text
__
a.J
( g,,- j
?-
4 FIT.o.
. -T
~3 s UNITED STATES
((;CMi h ATOMIC ENERGY COMMISSION l
WASHINGTO N. D.C. :'0545
- \\D fg g
May 27, 1970 p.
A. Schwencer, Senior Project Leader, PWR Project Branch 2 Division of Reactor Licensing (DOCKET No. 50 A)
C OCONEE MODEL FOR RELAP-3, 269
/
As an appendix to Paul Norian's memo t utes, March 27, 1970, Trip Report to I"C, I offer the follouing additions.
Initi.al Pressures i
I have taken what I think are reasonable friction drops between control l
volumes for the Oconce model to be used in RELAP-3.
They agree closely uith B&W's friction drops used in their "LASl! II nodelins; of the Oconee plant.
Using the appro::imate f riction drops appearing in the attached table and adding the approximate elevation aP's I calculate the the rmo-dynamic iP's appearing in the t ab le.
~ have then chosen initial prescures consis tent uith these pressure drops and they are listed in the table.
I suggest we use these initial pressures in our Cconec analyses and that frictica f actors, k, be put in as zero so that PILAP will concute them consistent with the initial flows and pressures.
Included in the table m
H.a a p g, r n. Or.- t u e t1..u f f.c :_.. a r
9 psi-sec~
3 (ft -lbm y uhich RELAP will calculate and use.
Two Phase Friction Drops I have been unchle to discover any mention of tuo-phase friction factors in the RELAP-3 writeup or in the source deck.
I am afraid they are missing, and that junction friction drcps are simply given by
=b W!Wl, SP friction a
where AP is in psi, a in ibm /f t3, W in Ibm /sec, and k in psi-sec /lbm-ft3, 2
It anpears that k is assumed ccnstant, independent of quality.
If so, I recommend that we incorporate a table of two-phase friction factors as a function of quality and pressure.
Denny Ross has some ideas on this and I would be glad to help incorporate them into our version of "2 LAP-3.
7912200145
-- -- - - - - - - - - - - ~ - - ~ - - - - - - ' - ' ~ " ~ ~ ~
^ ~
- TABLI: I INITIAL FLOWS AND PRESSURI: DROPS FOR OCONEE MODEL
~ ~ ~ ~ ~
Friction l'ri c-Dais i 4, Factor K Initial' Junc-Pump Elev.
tion Accel. Flow E
~ psi-sec g
10 Vol.
Pressure tion From/
No.
(ps i. )
No.
To (psi)
(psl)
(esi)
(psi)
(psi)
(#/sec) ft ft -lbm l
1.
2231.0 1.
2/14 11.5 3.2 P,. 3 0
33,580 45.7 0.34 I
- 2 2255.5 2
14/1 13.0 4.7 8.3 0
33,580 43.7 0.32 3
2204.3 3
'2/15
~11.6 3.3 8.3 0
2,890 46.7-46.00 4
2188.1' 4
1/3 26.7 5.3 21.4 0
18,240 42.7 2.7 15 2180.0 5
'3/4 16.2
-0.5 16.7 0
18,240 42.7 2.1 6
2276.6 6
4/5 8.1
-10.6 18.7 0
18,240 44.7 2.5 7
2259.0.
7 5/6
-96.6 117.2
-0.2 20.8 0
18,240 46.7 2.9 8
2204.3 8
6/7 17.6 2.5 15.1 0
18,240 46.7 2.1 9
2188.1 9
15/1 12.9 4.8 8.1 0
2,990 44.7 43.00 10 2180.0 10 1/8 26.7
- 5. 3 21.4 0
18,240 42.7 2.7 11 2276.6 11 8/9 16.2
-0.5 16.7 0
18,240 42.7 2.1 12 2276.6 12 9/10 8.1
-10.6 18.7 0
18,240 44.7 2.5 13 2200.02 13' 10/12
-96.6 117.2
-0.2 20.8 0
9,120 46.7 11.6-14 2244.0" 14 12/7 17.6 2.5 15.1 0
9,120 46.7 8.4 15' 2243.9 15 10/11
-96.6 117.2
-0.2 20.8 0
9,120 46.7 11.6 16 600.0 16 11/7 17.6 2.5 15.1 0
9,120 46.7 8.4 17 7/2 3.5
-4.8 8.3 0
36,470 46.7 0.29 18 8/0 0
0 0
19 13/3
-4.28 4.28 0
0 0
?
20 16/7
-1659.0 0
0
?
C osed
t,
. 3
'Pumo Caviation Model-
~ I have. gone thru the function subprogram PUMP and have some criticisms
- of the. pump model used in RELAP-3.
The basic problem'is when the pump cavitates the head.goes to zero instead of giving a frictional head.
I suggest we rewrite PUMP and recompile it.
p As you know, RELAP requires as input:
1 l
1a) a he'ad verus capacity curve, H(q) l b) a frictional factor for a zero-speed pump, k c) a pump coastdown curve of fractional pump head as a function t
of. time after pump trip signal, C a-cavitation constant CAVCON.(psi p(AT) i d)
).
l c) a required NPSH(ft).
' RELAP-2 separates the pump head into two parts, an H ump an'd a frictional p
}
- hecd, H
=H
-kQlQl' 3
T0T pump 1
~
RELAP-3 determines a constdcwn table as a function of time af ter pump trip.
It then sets the f rictional f actor equal to k = (1-C )
D o'
where k 'is' the input, zero-velocity friction factor.
RELAP-3 then determines the dif ference between the actual NPSH and the 1
required NPSH, 40 = NPSH-(actual) - NPSH (required),
where the NPSH (required) is an input constant of 150 feet for Oconee.
1 If AP>o, it sets the-cavitation constaat equal to one (CAV.= 1.0)~.
If aP<o, it forms a cavitation constant
. CAV. = :L - CAVCON x (AP) 2,
^
where CAVCON is an ' input cons tant for Oconee of about 0.002_ (psi-2),
77 1CAV from the above Eformula ~is less.than acro it sets CAV-equali to zero.
If. CAV is between zero and one, 'it -uses CAV as the cavitation constant (dimensionless).. CAV is.the effective fractional pump head. for a par-
~
tially,cavitating pump.
4 i
I t
I e
+
+
y
=
Q-g
-y F
e-p y
, =.
9 y
i-w er y
=
4
.i
~ The physica1' meaning of CAVCON. (psi-2) is that /1/CAVCON is a differential j.
pressure (about 22 psi for Oconee), and if the actual NPSH is below the required NPSH by more than this differential pressure, then the pump is fully cavitating.
RELAP-3 calculates H as pump i
H
=C x CAV x H(Q),
j Pump D
}
where H(Q) is the input head-capacity table for the pump at full RPM.-
Adding the frictional head to the pump head gives a total head of
^ * "(9} ~ o( ~ D}9 0
- H
=C yy D*
The problem with this model in RELAP-3 is that for a full speed. pump l
(C = 1.0) which is fully cavitating (CAV = o) the model.gives no pressure D
i drop across the pump.
It should actually have a large friction drop, not unlike the zero speed case.
A simple way to fix this up would be to change the coding in RELAP-3 so that it gives a total pump head of u
-C x CAV x H(Q) - k,(1-C xCAV)QlQ!.
total g
D I
which states that the total head is a linear combination of the full speed, i
non-cavitating head and a frictional head, where the combining coef ficient is CD*
A*
I recommend that we change the PUMP function subprogran in RELAP to make~
this change.
4 t
Of course, ultimately, we should put in pump. characteristic curves of f
head and torque versus RPM and capacity and solve the pump constdown differential equations as a subroutine in RELAP.
In other words, we would omit the coastdoun table and let RELAP compute pump coastdown taking into account the torque.on the shaf t caused by the accelerating flows, 4
. For the current' version' of RELAP-3, I recommend a zero speed frictional' coefficient' for the Oconee pumps, of k = 0.732 x 10 '
~
- 8
~8*"
,1bm-f t (
which;gives the-following head table:
b 4
h e
-a,--_--,aa__--.---
a..n
.s s
w
i 5
~
~ TABLE II-HEADLVS CAPACITY q
100% rpm 0 rpm.
' Head Head O(GPM)
H(f t)
H(f t)
AH(ft)
-396,000 8,750 8,204 546
-299,200 5,250 4,684 566
-140,800 1,750 1,037 713 0
550 0
550.
63,000 460
-208 668 88,000 360
-405 765
^
140,000 0
-1,025 1,025 264,000
-1,750
-3,646' 1,896 400,400.
-5,250
-6,386 3,138 I arrived at this by assuming that the difference in head. going from zero to 100% rpm is approximately. equal to. the shutoff head of 550 feet for.
I large, backward flows thru. the pump,- and assuming. that the pressure drop
-thru the~ pump is the same for forward and bac irdiflows.
TF 100%' rpm curve comes'from B&U's FLASH II input and.I-have checked the-zero rpm curve against ~ Bingham Pump Company zone map (curve ~ No. 24927,.
1 i
test 176102-for single-suction pump.vithLa specific speed of N
=.4165).
g The-Bfngh i test data;is compatible with a k-:of from.
e g
-5 psi-sec 0.5 to '1.iD ' x -10
,1bm-f t j-av
^
s s
a
.i
+'
4 n.
6
- Pump Coastdown Curve 4
j The pump coastdown curve, CD versus time af ter trip used by B&W in FLASH-II in simply the square of the relative speed versus time curve given in Figure 14-16 in the Oconee FSAR.
To a good approximation this is
'2 y
C
=
,1 +.t/T, i
where t(sec) is tin:e af ter pump trip and.T is_a time constant of 12 seconds.
1 B&W has made' an error and the Oconee pumps will coastdown with a time cons tant of approximately 9.'4 seconds.
A slight refinement of our Oconee i
model for RELAP-2 would be to correct this time constant.
i
/
- e
/
K. B. Cady
_ PWR-Project Branch 2 Divis' ion of Reactor Licensing cc:
.C.~G. Long P. E. Norian T. M. Novak D. F. Ross Docket Files PUR-2 Reading Orig: KBCady I
Pos ts crip t: ' June 1, 1970'
^
.I notice that L the pressuriner level. pressure in the table is 2200.02 psi.
which makes it 4.28 ps1 below the hot leg, Volume 3.
-Actually it should be 4.28 psi above the hot leg pressure in order to balance 'the elevation thead and. provide for no initial acceleration of the flow in' the pressuriser leg.-
This-is probably a trivial point,.but could be corrected.by changing
^
. the pressurizer :(Volume 13). initial pressure to 2208.58 psi. Also, check-allainitial pressures against' the table for typographic errors.
There is
- one, I believe.
11have discovered a coding error,in the h'ect e:: changer subroutine HTXQ.
' Page -46 of the RELAP-3 manual: says that the heat ' exchanger index for ivolume-regions'say be repeated.
However, the subroutine.is improperly coded"and uses {the 'same : input and output junctions forf each volume.ragion..
indexing the heat exchanger. data card.
~
. This is nonsense, 'of course. 'iThe
~
M 1
~...,
7 simplest way to correct the situation is to avoid repeating a heat exchanger index.
That is, give each heat exchanger volume its own heat exchanger data card.
This is easier than trying to correct the RELAP coding.
I notice another bug in our input data.
The heat exchange volumes (four steam generator volumes and the core volume) are not compatibic with our initial flows and temperature rises.
I recommend a careful look at this problem so that we do not have an initial readjustment transient super-imposed on the blewdown transient.
.