ML20004D354
| ML20004D354 | |
| Person / Time | |
|---|---|
| Site: | Browns Ferry  |
| Issue date: | 06/02/1981 |
| From: | Mills L TENNESSEE VALLEY AUTHORITY |
| To: | Ippolito T Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8106090282 | |
| Download: ML20004D354 (14) | |
Text
.e e
TENNESSEE VALLEY AUTHORITY CH ATTANOOGA. TENNESSEE 37401 400 Chestnut Street Tower II June 2, 1981 y/
g Director of Licensing s
Attention:
Mr. Thomas A. Ippolito, Chief j
-3 dll// g 198 7 g -- j Operating Reactors Branch No. 2 U.S. Nuclear Regulatory Commission g
u.s.
Washington, DC 20555-c%'grg y
Dear Mr. Ippolito:
D q$
N In the Matter of the
)
Docket No. 50-259 Tennessee Valley Authority
)
50-260 50-296 This letter is in response to D. G. ',isenhut's letter dated September 27, 1979 to All Light Water Reactors and your letter to H. G. Parris dated October 22, 1979, concerning containment purge valve operability and purge operations. TVA has previously provided information on this subject in our letters from J. E. G111 eland to you dated fune 12, 1979, from J. L. Cross to you dated December 10, 1979, and from me to you dated March 17, November 19, 1980, and April 22, 1981.
Enclosed is an operability analysis of the containment purge valves installed in the Browns Ferry Nuclear Plant. The enclosed analysis addresses containment purge valve operability under Design Basis Accident Loss-of-Coolant Accident (DBA LOCA) forces as required by the NRC letter dated Septem'oer 27, 1979. The analysis shows that the purge valves are adequate for closure against DBA LOCA forces.
Therefore, modification of the valves to meet the interim staff position of the October 22, 1979 letter is not necessary.
During the current refueting outage on Browns Ferry unit 1, TVA is performing modifications to satisfy the requirements of NRC Branch Technical Position CSB 6-4 regarding valve closure times and debris screens. The valve closure time for the large purge valves will be reduced from approximately 15 seconds to less than 2.5 seconds.
l This significantly reduces the analytical dose and also protects
(
againe.i. pressurization of appertenant duct work and secondary -
L containment as e.iscussed in item 5 of our March 17,-1980 submittal.
I
['
An Equal Opportunity Ernployer i
8106080 N
. Mr. Thomas A. Ippolito June 2, 1981 TVA considers submittal of the enclosed analysis and the planned modifications es resolving all applicable NRC concerns to date on this subject. Accordingly, following completion of these modifications on Browns Ferry unit 1 and later on units 2 and 3, we will resume purge operations as allowed by our present technical specifications. The present license restrints use of the main purge valves during operation to within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after placing the reactor in "run" mode and 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> before shutdown. These specifications allow purging to be conducted without adversely affecting unit output yet ensure that the overall time spent purging is low.
Very truly yours, TENNESSEE VALLEY AUTHORITY G/h w' L. M. Mills, fenager-Nuclear Regulation and Safety SubscribeA an'. sworn to ber re me this cv &
day of 1981.
Notary Public Hy Com:nission Expires Enclosure M
1 i
1 1
i 4
1 l
1 ENCLOSURE I
I CONTAINMENT PURGE VALVE OPERABILITY ANALYSIS l
}
.I i
j.
}
i i
l l
)
5 i
I l
1^
4 An analysis has been performed to address the question of containment purge valve operability at the Browns Ferry Nuclear Plant (BFNP) detailed by the Nuc1 car Regulatory Commission (NRC) in a letter dated September 27, 1979, to All Light Water Reactors. The purge valves in question are shown schematically in figure 1.
Table 1 gives pertinent information for all of the containment purge valves and associated actuators at the plant.
The Rockwell 18-and 20-inch butterfly valves, equipped with Bettis air-operated spring-r turned operators, are pressure-temperature rated 2
at150lbfin. At operating temperature, the valves are designed for 230 lb/in. The pneumatic actuator requires air to open and is spring loaded to close. There are two air pistons and one spring on a single shaf t (figure 2).
The Fisher 10-inch butterfly valve, provided with a Fisheg air-operated spring-loaded to close operator, is rated i
at 150 lb/in. All of the purge valves have symmetric dicks and steel bodies with elastic inserts to provide a low leakage seat.
The basic conditions of the analysis are:
4 1.
The accident chosen for determining valve operability was the design basis loss of coolant accident (LOCA). This accident produces the highest. *ywell pressures and therefore maximizes the loads on the valves.
2.
All vgives were analyzed using a constant upstream pressure of 44 lb/in gauge. This value bounds the peak drywell pressure given in the Mark I Containment Program Plant Uniqua Load Definitions j
report.
i 3
The fluid dynamic torque is defined by a relationship developed by Fisher Controls for compressible fluid flow through butterfly valves with symmetric disks by Fisher Controls (Floyd P. Harthun, l
"Effect of "luid Compressibility on Torque in Butterfly Valves,"
ISA Transa ; ions: 8 281-286, 1969).
4.
Input paran aters that were not attainable for the specific Rockwell butterfly vahes were obtained from Fisher Itenry-Pratt Valve Companies for valves that ware ' essentially identical to the Rockwell valves.
5.
Major conservatiems included in the analysis are that pressure losses due to ber.ds, elbows, tees, and debris screens were not considered. Pressure losses due to upstream valves were not l
I considered in the analysis either.
The various torque constituents considered and analyzed were:
1.
Operator Torque The operator torque has two components:
(a) the operator spring l
torque which causes the valve to close and (b) the operator air I
torque supplied by the decaying air pressure and pistc;; cylinder volume as the valve closes, which slows the closure of the valve.
i l
&'-'~
_,e.m, m.-,
w
Individual spring torques have been listed in table 1.
A typical spring pecvides a torque of 8,200 in-lbs when the valve is fully closed and 16,400 i: -lbs whpn fully open. This spring requires an air pressure of 80 lb/in gauge to opgrate the valve.
Nominal air pressure at BFNP is 90 lb/in gauge. The torque on the air side of the operator was calculated during closure. The analysis was performed, considering bleeding down the two air cylinders, based on flow calculated using the Bernoulli equation for incompressible flow through a square edge orifice, while accounting for tbc change in air cylinder volume with time. The pin and. slot coupling between the piston-spring assembly and the valve shaft was also modeled to incl.ude the nonlinear effects of this connection.
2.
Hydrodynamic Torque This is the torque that the fluid flowing through the purge lines I
exerts 2pon the valve disk. As,tated earlier the purge valven at BFNP have symmetrical disks In discussions with valve manufacturers such as Henry-Pratt, Fisher, and Mosser Industries all manufacturers stated that butterfly valves with symmetris disks always tend to close due to fluid flow over the disk. This car.clusion was based,on analytical and experimental work performed by the valve manufacturers. This conclusion is also supported by literature available on compressible flow through butterfly valves.
As has been stated earlier, the hydrodynamic torque on the disk was evaluated using a relationship developed by Fisher Valve Company for compressible fluid flow through a butterfly valve (this basic relationship is fairly standard based on contact with manufacturers and a literature search).
The relationship developed by I'isher Valve Company for the hydrodynamic torque (T 8 E **" "8' D
TD*
1
'e (1) wh?re K is a dimensionless torque coefficient determined by 3
l experiment, D is the inside diameter of the valve, and C
C 3
p 2
dip,= P3 59.64 Sin E)
(2) dip i the effective pressure differential contributing to the e
dynamic torque on butterfly valves considering compressible flow, where P is the upstream or inlet absolute pressure, C, is 3
determined from.the ratio of the gas sizing coefficient, C E
to the flow coefficient, C, C, is a correction factor for y
i variation in specific heat ratio and -() is given as:
O= 12 63 AP_
(3) 1 2 1
rad 4LP is the valve pressure differential.
where O
l i
(
r 1
-w'
- V*
wo',
-at-'ty 4r--
W-
'y
-g*?
F % #
T-
The dimensionless torque coefficient, K, was provided by the HenryPrattCompanyforvalvesofessenkiallythesamedesignand disc shape as the BFNP valves. C is determined from catalogue dataforFishervalveswithdiscdhapessimilartotheRockwell valve discs.
Figure 3 shows the hydrodynamic torque as a function of disk angle for the 20-inch purge valve. The maximum fluid torque for the 20-inch valve is less than 10,000 in-lbs and less than 8,000 in-lbs for the 18-inch valve. The peak hydrodynamic torque for the 10-inch valve is less than 1,400 in-lbs. The shape of the torque curve is similar for all valves, but differs in the magnitude of the load.
3 Frictional Torque Internal valve and operator friction resists any movement but is small in magnitude relative to the operator and hydrodynamic torque.
As has been stated, the Rockwell 18-and 20-inch valves and the Fisheg 10-inch valvo are pressure-temperature rated at 150 lb/in. Under any plant condition, the valves are structurally adequate and possess conservative margins of safety. The valve operators are designed for larger than specified torque values.
The design values provide adequate margin to withstand LOCA induced 2oads without failure or damage. The design of the operator has been verified for a scismic environnent which has a margin of safety approximately four times greater than the BFNP siesmic loads. Vital accessories have also been qualified for operability to seismic and other conditions which are greater than anticipated. Hence, the valves and accessories are structurally adequate for their intended service.
Having taken into account the various forces acting on the disk, shaft, and the actuator, the valve closure rate versus time has been determined and plotted for the 20, 18, and 10-inch valves as shown in figures 4, 5, and 6, respectively.
The flow directions chosen in the analysis are those that would be present during a LOCA; that is, the normal flow direction is chosen for the exhaust purge valves and opposite the normal flow direction for the supply valves.
As has been previously stated, a constant and therefore 2
conservative containment pressure of 44 lb/in gauge was used in
'the analysis.
The highest loads on the valve are produced for a single valve closure. Simultaneous closure of both valves in a line will reduce the pressure drop across each individual valve, thereby reducing the fo'ces on a given valve. All of the purge valves at
BFNP are located outside of the containment, so the operators vent to atnicspheric pressure. Therefore, the containment pressure redacing flow out of the operators is not a concern at our plant. The pneumatic operator on the purge valves are supplied air to open the valves and provided with springs to close the valve when air pressure is removed. Accumulators are not used on the valves and torque limiters are not required on the valve operators.
The effects of the piping system or. the fluid mixture as it relates to valve disc and shaft orientation has been studied. As can be seen from table 1, valve numbers 64-17, 64-29, 64-30, and 76-24 are relatively close to a change in flow direction (from an elbow, bend, tee, etc.). The above listed valves which are currently not in the correct position will be rotated at the next available outage. The rotation will be such that the valve shaft is in the same plane as the source ceasing the flow change (for example, an elbow) so that a desirable uniform flow distribution is obtained.
Conclusions 1.
The fluid torque always tends to close the valve.
2.
A nonuniform fluid force on the valve diso does not appear to be a major problem.
3 The valve and operator are' well within their structural limits.
4.
The purge valve operability is therefore not a safety concern.
E50297.06
Table 1 Containeent Purge Valves and Actuators at Browns Ferry Nuclear Plant l
Unit 1 Valve Operator Manufacturer Spring Torque (in-lb)
Pipe Diameter
$1va No, Manufacturer Size (in. )
Purpose and Model No.
Closed Open Nearest Flow
- C 64-17 Rockwell 20 main supply Bettis 732A - SR51 S,400 17,500 1.92 i 64-l'8 18 drywell supply 722A - SR63 1,000 13,500 11.33 64-19 20 torus supply 732A - SR51 8,400 17,500 10.2 64-29 18 drywell purge 722A - SR63 7,000 13,500 1.73
- 64-30 18 drywell purge 3.47 l64-32 18 torus purge 9.47 i64-33 18 torus purge 10.87 l76-24 Fisher 10 N2 supply Fisher 656 - 60 1,288 2,394 1.83 bit 2 (64-17 Rockwell 20 main supply Bettis 732A - SR51 8,400 17,500 1.92 l64-18 18 drywell supply 722A - SR63 7,000 13,500 11.33 64-19 20 torus supply 732A - SRS1 8,400 17,500 10.2 164-29 18 drywell purge 722C - SR80 6,300 12,600 1.73 I64-30 18 drywell purge 722C - SR80 6,300 12,600 3.47 l64-32 18 torus purge 722A - SR63 7,000 13,500 9.47 l64-33 18 torus purge 722A - SR63 7,000 13,500 10.87 l76-24 Fisher 10 N2 supply Fisher 656 - 60 1,288 2,394 1.83
[it3
!64-17 Rockwell 20 main supply Bettis 732C - SR80 8,200 16,400 1.92 l64-18 18 drywell supply 722C - SR80 6,300 12,600 11.33
@4-19 20 torus supply 732C - SR80 8,200
- 16,400 10.2
-@4-29 18 drywell purge 722A - SR63 7,000 13,500 1.73
- 64-30 18 drywell purge 722C - SR80 6,300 12,600 3.47
'64-32 18 torus purge 722C - SR80 6,300,'
12,600 9.47 l64-33 18 torus purge 722A - SR63 7,000 13,500 10.87 l76-24 Fisher 10 N2 supply Fisher 656 - 60 1,288 2,394 1.83 l
flow directions are those that are present during blowdown ts:
A R H X
O E
T CA G N
E I
RD L I
OU TG 3
2 3
3 4
4 6
6
_k_
k_
g 0
3N 9
4 2-6 4
6 1 k-
)
8 I
f
\\
R EB l
hA L
H M
L C
E E
W N
T Y
O S
R I
S Y
D S
S E
R E
P G
P I
U R
/
S EU s
RP U Y G R
[ (
IFR EF SN W
OR B
"8 1
8 1 -
4 6
_[_
t "0
2 s
YL
_[_
P 4
t P
2 U
0
_9_
S 6
2 1 -
7 2
4
/-
6 N
M
'o-ORF I'
7 1
4 6
Y R L O P T P C U A S E G R N I
.Ol I l
d". r " ~ ~ ~ ' ~, s, ALWAYS FURNISH SERIAL NUKiiR OF ACUATOR WHEN ORDER NG PARIS, IFART4 U5f'R7D]
SE.:
SIR 1, is 7,
Q
. gI-HEAVY DUTY bm: TIS ACTUATOR 732c sa.s I'V h h W*)
O bh T ' UiOlM'MND d "') N
'S b
d[)
.W
?M T A ~ U@TE TFfiTff D r C di?
FiI
' Pjdf 7
3
- - -- - - - i. E 'n-
-f-Q~3 - E Jg - - r 7;
,l a.
=.=.
Is M*
I
{,\\
taw 93d.-
q A
- cy b,'
=
A
..> b L
l Cass l Space l
732C 5R H Spare 732C-5R-H trem p.,, no, Description Material Quan.
p.,n item p.,, N,,
Descr:pilon Materiat
}
p.,,,
I 18328 HOU$tNG DUCT. IRON I
19 33581 CYtlNDER STEEL 1
1-2 11330 COVER DUCT. IRON 1
20 SEE BELOW CYLINDER, SPRING STEEL I
i 3
18331 PIN, YOKE STEEL I
21 SEE CELOW SPRING CARTRIDGE A55Y, 1
4 11415 ROLLER. YOKE PIN STEtt 2
22 8387 SEAL, PI5iCN NITRITE 3
3 5
16364 SCREW, STOP ADJUSTING STEEL 2
23 17439 5EAL, YOKE NiTK'(E i
2 1 2
6 18335 NUT STOP SCREW STEEL 2
24 30556 SEAL, PISTON l'OD NITRitt 2
2 7
doo4 SCR EW, COVER SIEEL 4
22 8383 SE AL, CYLINDER NffeltE 2
2 8
3294 TAG, SERIAL NUMBER (NS)
SIN. STEEL 1
26 11334 GA5KET, COVER ASBESTO5 1
1 SCREW, DRkTAG (NS)
STN. STEEL 4
27 9094 CA5KET, CYtlNDER ADAPTER A5BESTOS i
2 I 2 9
19956 to 11332 ROD, PISTON STEEL I
28 3272 GA5KET, ADAPTER SCREW COPPER 8 1 8
11 3780 PISTON GRAY 1RON 2
29 20980 GASKET, STOP SCREW COPPE R 2
2 12 21922 ADAPTER, CnfNDER DUCT, IRON 1
30 5316 SEAL, PISTON HEAD NtTR;tE 2
7 38 3272 GA5KETdOVER SCREW COPPER 4
4 13 9096 BUSHING, ROD BRONZE 2_
14 833S NUT, HEX LOCK PISTON ROD STEEL 2
32 11967 SNUEBER VALVE A55Y.
I 15 29134 SCREW, CYLINDER ADAPTER STFEL AttOY 8
33 29193 BREATHE R, AIR, ALUMINU *A 2
16 3727 PLUG, PIPE STEEL 5
34 1997b PLUG, P!PE 51EFL
'2
- (
- 17 30501 VALVE, BtEED BRAS 5 2
35 29188 KEY, YOKE alli t 1
3S
'206(3 e tN, YOKE KEY si((L 1
f.ki 37 2032 ADAeTER, CYtiNCER
' DUCTS 2ON !
l g
g 7**.
..-s.-*
d.) g ','.
SPRING 732C 5R-H '
',,J StIE CYLINCER P/N CAR T. P/ N 40 16'93 16195 60 16193 16196.
s0 36i22
- i6io, Figure 2.
Typical Bettis Actuator 100 30412 16216 125 16224 16225 e
150 16224 16220 (N5)-NOT 5HOWN g
f
=
e.
i Wn =i%
Z[ d z g 51 iM i mkg 0
00 i
TN
~
A L
0 S
P 2
I S
R 7
X A
IA E
N LC A
Y U
e
\\
v A
N l
R S
a N
Y V
O R
h D
R c
E E
T n
N F
I S
S 0
- A N
0 2
I V
W 4) r 5 s H
T O
o e
f RD e
e r
l E
g g e
n
\\
A L
d
(
G sV E
N L
ue A
G qr
/
- 0. N oT 6A E
3 diu U
lF Q
R 3
e O
ru T
g iF 0
81 0
0 o2 oi g
L ain3 abo 4oH 231_I 4
b7 c
(
c
,z z
8 e!w -
Zy d ZP o
1i l
mh o
8L
/
~
TN AL S
P I
5 S
R 1
Y A
LA E
N LC A
U A
N Y
SN Y
R e
R B
R v
l O
/
E E
a N
F 2
V T
1 S
h A
N c
S
/
n T
W I
)
O s
I R
d 20 H
B nr o o f
'E c
M
/
- 0. (s e em iT 0
/
I E
.s T
Mv
/
I e
T 1 6 E
r-L A
k G
s iD 6
N 0
A
!+
erug iF D
,O s
0 O
f s
8*
8.
oO O
}8 8 k M$a
W$%
e Zy d Z9 la T>g 0%
8L TN AL S
P I
5 S
R L
Y A
/
LA E
N L
C A
U A
N Y
SN Y
R R
B R
e O
E E
v l
/
N F
2 a
T 1
V S
S A
N h
c mW
) n I
O s I R
H B
d 8 n1
/
o ro c f E
e M
e
- 0. (s m
i 0
T
/
M I
E T
.s
/
T e V
I E
lgn L
A G
ks J
i N
0' D
/
A 5
e rug iF 3
0 j
0 0
~
98
- to
- +
9N o0 QO8 5 M$Q 1
Wc %:;3 mC Zg d zP 3
e E5 a4 Mi N>g om.
0 3
TNA
/
L S
P 4
I 2
S R
Y A
LA E
N LC A
?s A
Y S
~
R N-R i
B R
e O
E E
v N
F l
T a
S V
S A
N 8
h V
W L) c I
T O
n H
R s
I B
d 0
n1 o
r E
c o
e f
M s
e
(
m I
i E
T T
M
/
s I
V 2T E
e 1
lg L
nA G
ks N
iD A
6 erug i
6 F
O 0
0
~
a*
oYa
- o6e, 8"
oo N10z4 Mre-neA
. ;.... -----....-.. ; 7.. --__.--.;--~-;----- ----
ANGLE / TIME HISTORY 5N c
b%
a aa
/
42 3 *9 a
1 8
9*
a>b ra AOZo 4 es w
M
/
s o
mz
>M O
- TVA* NEIM(SA ANAIXSIS gp H
BROWNS FERRY NUCLEAR PLANT Z
oP N
o oP 0.0 0.6 12 1.8 at 3.0 TIME (seconds)
Figure 6 - Disk Angle Vs. Time for 10-Inch Valve