ML20031D671
| ML20031D671 | |
| Person / Time | |
|---|---|
| Site: | Allens Creek File:Houston Lighting and Power Company icon.png |
| Issue date: | 10/09/1981 |
| From: | Fields M Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20031D662 | List: |
| References | |
| NUDOCS 8110140008 | |
| Download: ML20031D671 (8) | |
Text
.,
'r 10/Ch d1 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICE SING BOARD Ir. the Matter of
)
)
HOUSTON LIWiTING & POWER COMPANY
)
Docket No. 50-466
)
( Allens Creek Nuclear Generating
)
Station, Unit 1)
)
NRC STAFF TESTIMONY OF MEL B. FIELDS RELATIVE TO MANNINGS COEFFICIENT
[lexPirg Contention 6]
Q.
Please state your name and position with the NRC.
A.
My name is Mel B. Fields.
I am employed ai, the U. S. Nuclear Regulatory Commission as a Containment Systems Engineer in the Containment Systems Branch.
I have testified previously in this hearing on Board Question 4B, Compliance with GDC 50; Board Question 9, Bypass Leakage; and Board Question 4A, Combustible Gas Control.
l Q.
What dch TexPirg Contention 6 allege?
l i
A.
TexPirg Contention 6 states as fol.ows :
Petitioner contends that the drywell planned for Allens i
Creek Unit I will not withstand the pressure generated in l
a LOCA. The water within the weir wall will not clear the first row cf vents before the differential pressure exceeds 28 psi. This is due to failure to properly account for the Mannings roughness factor within the weir wall and the bent pipe. By delaying the time to clear the first row of vents i
i by only 0.5 second the drywell will be damaged allowing the escape of high pressure steam into the containment without being condensed. This will lead to the containment vessel l
pressure exceeding 15 psig so that it will crack allowing I
the escape of radioactive gases above the limits allowed by l
8110140000 811009 gDRADOCK 05000466 PDR.
. ~
Q.
What is the purpose of your testimony?
A.
The purpose of this testimony is to respond to the board comments contained in the September 1st order on ti:is contention.
I will address each of the board comments separately.
Board Comment #1 The Mark III containment is characterized as being designed to with-stand an internal pressure of 15 psig. The Board wishes to understand the margin of safety (expressed as an incremental pressure in excess of the 15 psig) between design pressure and that pressure at which the yield strength will be reached for the weakest components.
If containment leakage is not expected to occur when an overpressure corresponding to yield strength is attained, then it is important to document at what excess pressure beyond yield strength containment leakage will begin to occur and at what excess pressure significant conteinment failure will occur.
Response
Although the staff has not reviewed in detail the statement made by Miguel A. Lugo in hi testimony regarding the internal pressure capaci'-
of ACNGS, the value of 50 psig is typical of the yield strength for low pressure containments such as Mark III containments and Ice Condenser pl ants. The containment is designed to be a leak tir,ht structure; how-ever, there will be a small amount of leakage whenever the pressure l
inside the containment exceeds atmospheric.
From experience on leak l
(
testing operating nuclear power plants a good approximation of the leakage
, for a plant under design can be determined. However, the individual ch.aracteristics of plants _and.the changes that coul,d occur in the plant's capability to resist leakage over time make the exact quantification of this leakage impossible.
For this reason, preoperational and periodic leak tests are required using the peak calculated accident pressure as
" ' ^ '
~ - - - -
the test pressure (see Appendix J to 10 CFR 50 for further details). The leakage rate of the containment for pressures hisner than this test pressure cannot be determined exactly but since significant material deformation is not expected until the material reaches its yield strength, signifi-cant increases in leakage should not occur for pressures less than 50 psig.
Board Concern #2 The board wishes to understand the basis for confidence in the conclusion that data from the General Electric Company's test in their Pressure Suppression Test Facility are applicable to the ACNGS.
Figures A-12 and it ' 3 attached to Fields' affidavit offer no indication of reliability
( icertainty, accuracy or error band) associated with the experimental r..alts.
Response
The test data from the C' Press 9re Suppression Test Facility (PSTF) is applicable to ACNGS because the factors that govern vent clearing for Mark III containments; drywell pressure rist the cross-sectional areas normal to the flow direction, and the height of watar over the top vent, have been encompassed by the PSTF testing program. As T. R. McIntyre notes in his testimony regarding this contention, the relevent system paramaters for the PSTF are almost identical tc the ACNGS vent geometry. The dry-well pressure rise measured in the PSTF bounded the expected Mark III pressure rise and the top vent submergence for ACNGS (7,5 feet) was also bounded by the PSTF testr.
This PSTF test data is used to verify that GE's vent clearing model con-servatively predicts vent clearing times.
Figure 4-6 (attached) of NED0-20533, "Th'e Gener'l Electric Mark III Pressure Suppression Contain-ment System Analytical Model," shows that the vent clearing mode consistently overpredicts vent clearing times for full scale steam blowdown tests.
The experimentally measured drywell
4 and v:ntainment pressure was programmed into the vent clearing model as well as initial submergence and vent system and.gool geometry of the PSTF tests w best show the accuracy of the model. Because of the simularities Setween the PSTF and A.CNGS as far as vent clearing considerations are concerned, the staff has a great deal of confidence in the ability of the vent clearing cadel in GE's pressure suppre:sion containment analytical model to predict the vent clearing times in a Mark III containment, The potential inaccuracy in defining vent clearing as the time at which the air-water interface passes a prob (located 6 inches from the end of the vent) is equal t4 or less than the scan accuracy of the data acording system, which was 0.016 second for test series 5701 and 5702 (one vent and two vent fi i scale steam blowdown tests, respectively) and 0.18 seconds for test series 5703 (three vent full scale steam blow-down tests). At 1 and 1.5 seconds on Fig. 4-6, the average conservatism of the vent clearing model is approximately 15 and 10%, respectively, which does a good job of bounding the experimental error.
Board Comment #3 Fields' affidavit refers to NEDO-10320, and represents that Figure 4,4 therefrom is attached. The Board's copy of this affidavit provides Figure 4.1 from NEDO-10320 and Figure 4.2 from an unidentified source, l
there being no Figure 4.4.
Please explain, and again address uncertainty, accuracy or error band to be associated therewith,
Response
Attached is Figure 4.4 of NED0-10320 which is a comparison of Humbolt vent clearing data with predictions of vent clearing model, Figure 4,2 l
' e of my previous affidavit is also from NED0-10320 and shows the vent velocity and applied pressure differer.c: during vent clearing calculated for Limerick.
The purpose of Figures 4.1 and 4.2 is to show that friction losses between the moving water and the ver.t system are neg11gible when compared to the applied pressure difference. Notice that the peak fric-tion pressure drop in vent liquid flow as a % of applied differential pressure in Figure A.1 is never greater than 1.3%; this ratio is less than 1% for the majority of the transient. This t esult is applicable to the Mark III vent system because the differences in vent system geometry between Mark II and Mark III containments'is irrelevant from a fluid friction standpoint, as shown in T. R. McIntyre's testimery on this contention.
The comparison of vent clearing data with predictions of vent clearing model in Fig. 4.4 was provided to show that the basic approach used to derive the vent clearing model (applying the integral form of the con-tinuity and momentum equations to a series of control volumes) will give good results when applied to different vent geometries.
The basis for the staff's acceptance of the accuracy of GE's vent clearing model is based on the comparisons with the PSTF full scale steam blowdown tests (as discussed in the response to Board Comment #2).
Board Coment #4 l
Intervenor's response raises questions regarding, for example, smooth vent tubes versus rough concrete walls, drywell corner weakness, right-l angle turns in fluid flow paths, and the necessity to clear two rather l
than one set of vents. Without more infonnation than is currently before us, the Board cannot assess the importance of these :onsiderations, i
~-
3 6-i
Response
The entire suppression pool boundary, which includes the weir wall, dry-well wall, containment wall, floor, and the horizontal vents is steel lined so the intervenor's question regarding rough concrete walls is not applicable.
Irreversible losses are included in the vent clearing model for the turning losses the fluid experiences when flowing from the annulus into a horizontal vent. GE's analytical model also simulates i
the sequential clearing of all three rows of vents, f
It is important to include these considerations in the modeling of the l
vent clearing phenomena. The fact that this vent clearing model does accurately predict, with a 10% to 15% conservatism, the PSTF test data
\\
shows that these considerations have been correctly modeled.
l 1
l l
~.. _. _ _
I f
NEDO-10320 t *.
t 70 i l
l l
1 I
I so gi7 no E
o
-wuweEns naPER TO EXMn! MENTAL RUNS E
g 16 r.
1 o
i E
y 13.ts.ac.msJsJe y so 3
5 i
20 Ste cALcutATro l
i l
10 34 I
l 1
I I
I I
I O
0.1 0.2 R3 OA Es 0.7 WNT CLEARING YlE ised FIGURE 44. COWARISON Of HUMBOLDT VENT CLEARING DATA WITH PREDICTIONS OF VENT CLEF. RING MODEL 44/44
2.5 l
l (CONSERVATIVE)
\\
7.0 l
)
E A
a 6
b 1.5 g
b i
(NONCONSERVATIVE) i o
l E
1.0 j
p I
O
?
5 O
5 O TOP VENT 3
I l
A VE Y I
I 0
O 1
2 3
4 EXPLPIMENTALTIME OF VENT CLEARING lasc)
Figre 4 6 f.orpSite Stem Blondown Teits - 1,2, and3 Vena l
-