U-600588, Forwards Info Re Safety Relief Valve Discharge Line Sleeve Annulus Condensation Oscillation Load & Conservatism in Spatial Attenuation Per Request for Addl Info on SER (NUREG-0854) Confirmatory Issue 71 - Humphrey Concerns
| ML20198L208 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 05/30/1986 |
| From: | Spangenberg F ILLINOIS POWER CO. |
| To: | Butler W Office of Nuclear Reactor Regulation |
| References | |
| RTR-NUREG-0854, RTR-NUREG-854 U-600588, NUDOCS 8606040235 | |
| Download: ML20198L208 (10) | |
Text
r a
U-600588 L30-86 (05-30)-L 1A.120
//LINDIS POWER COMPANY CLINTON POWER STATION P.o. box 678. CLINTON ILLINOIS 61727 May 30, 1986 Docket No. 50-461 Director of Nuclear Reactor Regulation Attention:
Dr. W. R. Butler, Chief Licensing Branch No. 2 Division of Licensing U. S. Nuclear Regulatory Commission Washington, DC 20555
Subject:
Clinton Power Station Safety Evaluation Report Confirmatory Issue No. 71 - Humphrey Concerns
Dear Dr. Butler:
The purpose of this letter is to respond to a request for additional information concerning the Clinton Power Station (CPS) Safety Evaluation Report (NUREG-0854), Confirmatory Issue No. 71 - Humphrey Concerns. This request resulted from a telephone conversation on May 8, 1986, with the NRC staff and consultants and representatives from General Electric Company, Mississippi Power and Light, Gulf States Utilities and Illinois Power Company (IP) discussing IP's submittal to the NRC, U-600319 dated November 25, 1985, on the generic issue of Humphrey Concern 2.1 - Safety Relief Valve Discharge Line (SRVDL)
Resonance. Attachment 1 is the additional information that was requested for NRC Staff's review concerning the analysis of the SRVDL condensation oscillation load. Attachment 2 is a description of the conservatism present in the analysis. We believe that this response will resolve this concern for CPS. This response is not a fuel restraint item.
If you have any questions concerning this information, please contact me.
Sincerely yours, F. A. Spahgenb eg Manager - Lice
- ing and Safety LRH/kaf Attachments B. L. Siegel, NRC Clinton Licensing Project Manager cc
NRC Resident Office Regional Administrator, Region III, USNRC Illinois Department of Nuclear Safety 1(
8606040235 860530 PDR ADOCK 05000461 E
U-600588 L30-86(05-30)-L 1A.120 SAFETY RELIEF VALVE DISCHARGE LINE SLEEVE ANNULUS CONDENSATION OSCILLATION LOAD Introduction The NRC has expressed a concern that the methodology used to address Humphrey Concern 2.1 (Reference 1) did not account for a possible resonance between the sleeve annulus Condensation Oscillation (CO) frequency and the sleeve acoustic frequency. To address this concern, an alternative approach for estimating the Safety Relief Valve Discharge Line (SRVDL) sleeve C0 load has been used, which conservatively utilizes Mark I Full Scale Test Facility (FSTF) CO data in which significant excitation of the vent acoustic modes was observed (References 2 and 3).
This alternative approach shows that C0 occurring in the SRVDL sleeve combined with the main vent C0 gives pressures on the containment wall and drywell wall which are bounded by the pool swell and chugging loads already considered in design.
l Analysis There is substantial large scale C0 data available from tests of the Mark I and Mark II vertical vents (downcomers) and the Mark III hori-zontal vents. Of these tests, the Mark I FSTF data (Reference 2) has the most evident excitation of vent acoustic modes. Therefore, the FSTF i
data was used to address the NRC concern regarding resonant amplifica-tion of the CO loads in the SRVDL sleeve.
l The Mark I Load Definition Report (LDR) (Reference 3) includes a conserv-I ative definition of harmonic amplitudes for pressure oscillations in the Mark I downcomer during condensation oscillation based on the FSTF data.
These pressure oscillations were conservatively assumed to occur at the same amplitude in the Mark III SRVDL sleeve. No amplitude reduction was performed to account for the differences in the exit geometry of the Mark I downcomer and the Mark III SRVDL sleeve. The Mark I downcomer discharge is a two foot diameter pipe while the Mark III SRVDL is approximately one foot in diameter surrounded by a one-inch wide annulus. This small gap is expected to result in smaller amplitude oscillations of the steam-water interface so that the SRVDL sleeve should have much lower amplitude pressure oscillations than the Mark I downcomer. Also, no amplitude adjustment has been made to account for differences in flow conditions between the FSTF tests and the SRVDL sleeve. The Mark I tests showed that the C0 pressure amplitude increased with the vent enthalpy flow (Figure 6.2.2-56 of Reference 2).
The maximum enthalpy flux for the SRVDL sleeve is approximately the same as the maximum value in the FSTF tests.
The frequency range of the SRVDL sleeve CO load was determined by multiplying the Mark I LDR specified frequencies by the ratio of FSTF vent length to the SRVDL sleeve length. This adjustment is based on the assumption that the frequencies are controlled by the acoustic response Page 1 of 6
p,-
U-600588 L30-86 ( 05-30)-L 1A.120 in the sleeve. The resulting ranges are shown in Table 1-1 for Clinton Power Station (CPS) with the pressure oscillation harmonic amplitudes in the sleeve and on the drywell and containment walls.
The pressure amplitudes in the sleeve given in Table 1-1 are equal to the Mark I LDR values as discussed above. The amplitudes on the drywell wall and containment wall were determined by using a spatial attenuation equal to one over the distance from the end of the sleeve. The radius of the bubble at the sleeve exit was conservatively assumed to be equal to the radius of the sleeve. Since the annulus gap will act to limit the bubble size, the actual spatial attenuation of the pressures would result in pressures on the walls which are much smaller than the values given in Table 1-1.
The SRVDL Sleeve CO loads given in Table 1-1 were compared to the existing design loads on the containment and drywell wall using amplified response spectra (ARS). The ARS values for estimated CO loads g
for the SRVDL sleeve and the main vent are compared to ARS values for the pool swell and chugging wall loads in Figures 1-1 and 1-2.
The comparisons are based on the lowest frequency component of the SRVDL sleeve C0 load. The CO loads for the two higher frequency components are less significant relative to the pool swell and chugging loads so they are not included in the comparison. This shows that the addition of the estimated SRVDL sleeve CO load to the main vent C0 load results in a total CO load which is less than the chugging load on the drywell wall (Figure 1-1) and the pool swell load on the containment wall (Figure 1-2).
Summary An estimate of the CO load in the SRVDL sleeve for Clinton has been made using Mark I FSTF data which includes signific2nt excitation of the acoustic modes in the vent upstream of the discharge. This was done to address NRC concerns regarding resonant amplification resulting from coupling of the sleeve acoustic frequency and the C0 frequency. The FSTF data has been conservatively applied wittout any amplitude reduction. The ARS of the resulting SRVDL siteve CO load combined with the main vent CO load is lower than other des!.gn basis loads (chugging and pool swell). Therefore, it is not necessary to consider this load in design evaluations.
1 The main vent CO ARS values are based on calculations for Grand Gulf Nuclear Station which are representative for CPS.
Page 2 of 6
U-600588 L30-86 (05 -30)-L 1A.120 References 1.
Mississippi Power and Light (MP&L) letter No. AECM 82/574 dated December 3, 1982 from L.F. Dale (MP&L) to H. R. Denton (NRC).
2.
" Mark I Containment Program Full-Scale Test Program Final Report",
NEDE-24539-P, April 1979.
3.
" Mark I Containment Program Load Definition Report", NEDO-21888, Revision 2, November 1981.
Page 3 of 6
U-600588 L30-86 ( 05-30)-L 1A.120 Table 1-1 SRVDL SLEEVE CO IDADS FOR CLINTON BASED ON MARK I LDR Pressure Harmonic Lap 11tude (psi)
Amplified Response (psi)
Frequency Frequency SRVDL Sleeve Drywell Containment Drywell Containment Component Range (Hz)
Annulus Wall Wall Wall Wall 1
24 to 48 3.6 1.73 0.22 43.2 5.4 2
48 to 96 1.3 0.62 0.08 15.6 2.0 3
72 to 144 0.3 0.14 0.02 3.6 0.5 l
l 1
i l
1 i
l l
l l
l l
l l
l Page 4 of 6 i
i ;
?$$
~
[,? $ GY sYe gL$
=
0 l
O 1
Cl l
t nI e'
Vl n
i l a
M O
C e
v
}
e S
e o
a l
C i
R S
0 l
A e
1 v
)
L
+
l e
L
(
A i
V e
Z W
g M
H l
l M
n s
LL U
i
(
E R
g W
T g
~
1 Y
C u
Y R
i E
h 1
D P
C /l CN e
N
/
f S
E r
O
/
~
I U
ug N
E f f Q
i O
S E
F S
/
N I
R R
O
) /
\\
F A
P
/
P S
I M
j E
O R
/,
C S
8 D
R A
E 0
I F
'(
1
~
I L
M P
R e
_ 0 o
0 0
3 o1 e
6 4
a
^ Me -
&d B t b E
a i. om
- l i
i 4
l i
l
RMPLIFIED RESPONSE SPECTRUM
( ARS )
i I
12
^
~
Pool s
Swell y
9-f
_ _y a.
m
(
/
MV + Sleeve CO 6
o,,
~
~
~
- _ ~
- ~
~
~
~
~
~
~ '
a
)
l
,Slacyc O
L-e
-u.
3 Main Vent CO
~
0-II\\\\\\
G
'10 e 10 10 10 sp g ;e e
o FREQUENCY
( HZ )
g*S Figure 1-2 O*
ARS COMPARISON ON CONTAINMENT WALL s
U-600588 L30-86 (05 -30)-L 1A.120 CONSERVATISM IN THE SPATIAL ATTENUATION USED TO DETERMINE SAFETY RELIEF VALVE DISCHARGE LINE SLEEVE CONDENSATION OSCILLATION WALL PRESSURE The Safety Relief Valve Discharge Line (SRVDL) Sleeve Condensation Oscillation (CO) pressure amplitudes on the pool boundary were calculated by attenuating the bubble pressures at the sleeve exit by the ratio of the sleeve inner radius to the distance from the sleeve exit to the walls. This attenuation is based on the conservative assumption that the bubble radius at the sleeve exit is equal to the sleeve inner radius and does not account for the presence of the SRVDL in the sleeve.
To quantify the conservatism in the SRVDL Sleeve CO load resulting from this assumption, the pressure attenuation to the wall was redeveloped giving consideration to the presence of the SRVDL in the sleeve.
From potential flow theory, the pressure attenuation from a spherical bubble is proportional to the ratio of bubble radius to linear distance.
Therefore, an effective radius was developed which better descriaed the volumetric displacement of the annular steam bubble postulated for the SRVDL sleeve exit. Figure 2-1 gives a cross-sectional view of the assumed bubble shape. The annular bubble encloses a volume of 0.31 times the volume of the bubble volume assumed for the SRVDL Sleeve C0 wall load. The effective radius was therefore, the cube root of 0.31 or 0.68 times the bubble radius used previously in determining the attenuation to the pool boundary.
Since the sleeve exit to wall distances are not affected by this analysis, the attenuation factor with the effective radius was equal to 0.68 times the attenuation factor used for the previous wall load.
Table 2-1 compares the previous wall pressure amplitudes with values obtained using this more realistic spatial attenuation.
Page 1 of 3
U-600588 L30-86 (05 -30 )-L 1A.120 Table 2-1 Clinton SRVDL Sleeve CO Wall Load Comparison - Realistic vs Conservative Spatial Attenuation Pressure Harmonic Amplitude (psi)
Amplified Response (psi)
Frequency Component Drywell Wall Containment Wall Drywell Wall Containment Wall 1
1.73 (1.17) 0.22 (0.15) 43.2 (29.2) 5.4 (3.7) l 2
0.62 (0.42) 0.08 (0.05) 15.6 (10.6) 2.0 (1.4) l 3
0.14 (0.09) 0.02 (0.01) 3.6 (2.4) 0.5 (0.3)
Realistic values given in parenthesis.
4 4
Page 2 of 3
-~
r_
U-600588 L30-86(05-30)-L 1A.120
/
DRYWELL WALL
///
/
SRVDL
///
SRVDL SLEEVE
/
/
//
SRVDL ANNULUS
///////
ANNULAR BUBBLE SLEEVE BUBBLE I
/
Figure 2-1 SRVDL SLEEVE AND BUBBLE VOLUME CEOMETRY Page 3 of 3 l
l l
l t