U-600265, Forwards Response to SER Outstanding Issue 9 Re Suppression Pool Hydrodynamics.Existing Design Basis & Justification for Use of NEDE-13426P,Figure 6-9 Re Hydrodynamic Mass Values,Discussed
| ML20138R262 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 11/14/1985 |
| From: | Spangenberg F ILLINOIS POWER CO. |
| To: | Butler W Office of Nuclear Reactor Regulation |
| References | |
| U-600265, NUDOCS 8511180514 | |
| Download: ML20138R262 (15) | |
Text
,
e U-600265 L30- 85(11-14)-L 1A.120 ILLIN018 POWER COMPANY CLINToN POWER STATION. P.o. Box 678. CLINToN. ILLINolS 61727 November 14, 1985
-Docket No. 50-461 Director of Nuclear Reactor Regulation Attn:
Mr. W. R. Butler, Chief Licensing Branch No. 2 Division of Licensing U. S. Nuclear Regulatory Commission l
Washington, DC 20555
Subject:
Clinton Power Station Unit 1 SER Outstanding Issue #9 Suppression Pool Hydrodynamics l
Dear Mr. Butler:
l This letter is in respo.ise to concerns identified by the NRC Staff l
in its review of Illinois Power Company (IP) letter U-600020 date(.May 16, 1985. The concerns identified were discussed informally with your L
Staff on August 6, 1985, and September 11, 1985.
Attached for your review is the information requested. Attachment 1 provides responses to questions about the existing Clinton Power Station design basis. Attachment 2 provides the justification for the use of. Figure 6-9 (
Reference:
General Electric Report No. NEDE-13426P) hydrodynamic mass values for calculating impact loads on certain circumferential structures used by Burns and Roe, Inc., (BRI) in its evaluation of structures within 6 feet of the suppression pool surface.
IP believes that this supplemental information resolves your concerns on this issue. Please contact us should you have any further questions.
Sincerely yours, F. A. Sp gen rg Manager - Lic ising and Safety JLP/kaf i
Attachments (2) cc:
B. L. Siegel, NRC Clinton Licensing Project Manager p)
NRC Resident Office Regional Administrator, Region III, USNRC A
Illinois Department of Nuclear Safety
' i 1
8511180514 851114 PDR ADOCK 05000461 E
U-600265 L30- 85(11-14)-L 1A.120 NRC REQUESTS FOR INFORMATION REGARDING FOOL DYNAMIC LOADS FOR CLINTON POWER STATION Q.'
GESSAR-II specifies a 12.8 kip impact load to be applied to the weir wall over the projected area of each vent. Has this load been considered in the design of the Clinton Plant?
A.
Yes.
Q.
Reference Question #5 in letter U-600020 dated May 16, 1985. Are there any structures between the top vent and an elevation of 1 foot above the top of the weir wall? Discuss how design loads used
.for these structures compare with criterion 3.0 of Appendix C of NUREG-0978.
A.
Yes, there are pipe supports located in the weir annulus and radially oriented floor support beams that rest on the top of the weir wall. The structures are comprised of I beams and are subject to the '" flat pool" impact restrictions of Section 3 of Appendix C of NUREG-0978. The guidelines provided therein serve as the basis for the design of these structures for impact loads. Conserva-tively, the impact velocity of 8 feet per second determined for CPS is applied in place of the 4 feet per second impact velocity specified in GESSAR-II. The grating above the top of the weir annulus is located above the zone where " flat pool" impact is appropriate.
Q.
The use of Equation 3 of the attachment to the response to NRC Question 480.27-2 is a questionable approximation with consequences that are difficult to assess insofar as jet front velocity is concerned. Denonstrate that the results obtained for impact velocity above the top of the weir wall (Figure 480.27-2-5) are reasonable and conservative by comparing the results, the results that would be ottained using the staff approved method described in the attachment to the letter from H. C. Pfefferlen, GE, to W. R.
Butler, NRC, dated June 18, 1982.
A.
The General Electric weir swell backflow and jet methodologies of Reference 1 (see below) have been reviewed and compared with the CPS weir swell backflow and jet methodologies of Reference 2.
The results of this comparison are summarized below.
j 1.
The hydrostatic head due to the difference in the level of the water in the weir annulus and in the suppression pool is treated differently in the two methods. The GE method treats the resulting pressure differential as a constant and the CPS method represents the pressure term as a function of the difference in the two water levels.
l L
2.
The drywell to containment pressure differential is l
represented as a ramp function in the GE method and as a i
^
j Page 1 l
t
U-600265 L30- 85(ll-14)-L 1A.120 (cont.)
realistic time history in the CPS method. Moreover, the absolute level of the pressure differentials is considerably different as the GE value represents a bounding value for design of Mark III plants and the CPS value represents a realistically conservative CPS design value.
3.
The General Electric method is derived from conservation of momentum in the " lumped" suppression pool / weir annulus / vent system. This " lumped" system assumption is a crude approximation since the velocities in the suppression pool and the weir annulus / vent system are drastically different. The CPS method properly accounts for the velocities in the various parts of the system.
4.
The two methods use different initial conditions. The GE method begins the solution to the backflow and jet models with the water level in the weir annulus at rest at the top of the weir wall. The jet initial velocity profile is generated by a series of steady state solutions. The pressure differential from the drywell to the wetwell does not equate to the head loss due to differing water levels in the suppression pool weir annulus / vent system. The CPS method begins the backflow solution with the weir annulus water level at rest at the center line of the top vent. Pressure equilibrium is maintained. The results of the backflow solution at the top of the weir wall are the forcing function for the jet flow solution above the top of the weir wall.
5.
GE's solution methods are simplified by neglecting acceleration terms. Therefore, the solutions are valid only for steady state situations and the problem solutions are approximated by an envelope of a series of steady state solutions. The CPS methods are solved numerically to produce the transient results.
These differences combine to produce two colution methodologies which bear little resemblance to one another.
Further, a direct comparison without consideration of these differences is meaningless. The only meaningful comparison which can be made without modifying one or more of the equations is to use CPS solution for velocity at the tops of the weir wall as the initial condition in the GE jet methodology. This comparison is shown in Figure 1.
This figure indicates that the results for the two sets of initial conditions are similar in both the velocity and the affected height. To illustrate that the CPS results are conservative with respect to the GE results, the velocity as a function of height is cross plotted in Figure 2.
In Figure 2, the CPS impact velocity bounds the Mark III velocity for values of i
Page 2 i
o.
U-600265 L30- 85(11-14)-L 1A.120
~ (cont.)
non-dimensional height (y
- ) less than 2.15.
For values of y
- above 2.15, the flow fie18 is decelerating to zero velocity in a i
short distance (i.e.
y * = 2.4).
Therefore, the physical distance involved is (2.4-2.15) E 3.11 = 0.78 f t. and for y * = 2.15, V * =
0.29, the velocity range is 0.29 x 22 = 6.38 ft/se8 at y * = 2 15 and 0 ft/sec at y *-= 2.4.
Thus, the zone where the S&L results do not bound the GE 8esults is at the very top of the weir swell zone where.the water jet is decelerating to zero velocity. Since, all
~
other factors being equal, the impact velocity is the controlling factor, the CPS methodology is conservative with respect to the GE methodology.
The specific references are:
- 1..
General Electric Company letter MFN-084-02, H.~C. Pfefferlen to W. R. Butler dated June 18, 1982.
Subject:
Requested Information on Jet Front Velocity.
2.
Illinois Power Company letter U-0698, J. D. Geier to A.
.Schwencer dated February 17, 1984.
Subject:
SER Issues -
Pool Dynamics, Attachment to response to NRC Question 480.27-2.
Q.
As we interpret the discussion on drag loads during upward motion of the weir water, the method used is identical to that described for fallback in Tables 480.27-2-3 except for a sign reversal for the second term in the equation for F.
ecn a
n at D
this is'a correct interpretation.
A.-
The following equations are used to calculate acceleration drag loads durir.g weir swell and fallback..These equations are given in References 3 and 4.
-r For. weir swell FD" M
~
E sO W
s c
E sPE!c cs0 (2)
For fall back-FD" H r
where U is the acceleration of the water slug during weir swell (ft/sec)',
F is acceleration drag force (1bf)
D i
3 Vg is displaced volume of the structure (ft ),
p is the density of the fluid.(lbm/ft ),
2 gg s the gravitational constant (32.2 lbm-ft/lbf-sec ),
i g is the acceleration due to gravity (-32.2ft/sec ),
C is the virtual mass coefficient (dimensionless),
g Page 3
U-600265 L30- 85(11-14)-L 1A.120 (cont.)
CH=M is the hydrodynamic mass coefficient H
(dimensionless), and e is direction of the structure's axis from horizontal (degrees).
Note also that CM H + 1 and the coordinate system is positive
~
upward; therefore, the gravitational acceleration acts downward,
Thus the negative sign in Equation (1). The acceleration term U is dependent on the flow field and has the general form U = dU = - dP Ec+g dt dx p
For fallback, dP/dx is zero and Equation (1) reduces to Equation (2). The virtual and hydrodynamic mass coefficients C and C are M
H given in References 3 and 4.
Examples of these values are shown in Figures 3 and 4.
The specific references are:
3.
Patton, K.
T., " Tables of Hydrodynamic Mass Factors for Translational Motion", ASME Publication 65-WA/UNT-2.
4.
Fritz, R.
J., "The Effect of Liquids on the Dynamic Motions of Immersed Solids".
J. of Engineering for Industry - ASME Transactions (February 1972), pp 167-173.
Q.
The use of a drag coefficient of 2.0 for " flat surfaces" is not in accord with the staff approved GESSAR-II method. The requirement is that a universal value of C = 1.2 is used for all structures, butthatfornon-cylindricalskructures,and" equivalent" diameter equal to the diameter circle which circumscribes the cross section of the structure be used.
Show for all non-cylindrical structures located above the weir annulus, the use of a drag coefficient of 2.0 represents a bound of this GESSAR-II method.
A.
The C value cited is used for box beams which have the dimension D Thediameterofthecircumscribedcircleis-(2aandthe a x a.
l drag force per unit length is FD"N 2gc or F = 0.85 a pV D
8e l
l l
Page 4 I
r U-600265 L30- 85(11-14)-L 1A.120 (cont.)
With a drag coefficient of 2.0 the drag force per unit length is FD" 2 g? "
e or D"
"h F
Ec Therefore, the use of C = 2.0 results in a load 15% higher than D
that of the GESSAR-II method.
For closed beams with unequal sides, the experimental values were determined by Bostock, Mair, Bearman and Trueman. Their work was reported in Aeronautical Quarterly in 1972. The specific references are:
1.
Bostock, B. R. and Mair, W.
A., " Pressure Distribution Forces on Rectangular and D-Shaped Cylinders", Aeronautical Quarterly February 1972.
2.
Bearman, P. W. and Trueman, D. M., "An Investigation of Flow Around Rectangular Cylinders", Aeronautical Quarterly, August 1972.
A copy of the pertinent results is attached as Figure 5.
Q.
In the evaluation of the CPS containment design, is the pressure loading defined in GESSAR-II Figure 3B-57 and described in 3B-6.1.6 utilized? If not, provide a description of the loeding function specified to account for wetvell pressurization.
A.
The containment design pressure is 15 psig. The pressure specified in the reference figure is lower, 11 psig. The 11 psi differential pressure was also used in the design of the solid portion of the HCU floor. The pressure differential has been considered in the design of the CPS containment structures.
Q.
For small structures other than the ones considered in the Burns &
Roe, Inc. Technical Report attached to the May 16, 1985, transmittal, is the limitation imposed by NRC acceptance criterion of Section 1.3.1 of Appendix C of NUREG-0978 conformed to in the CPS design?
A.
Yes.
Q.
Are the drag load data provided in GESSAR-II Figures 3B-19, 3B-72 and 3B-75 utilized in determining the CPS bulk drag loads?
A.
Yes.
Page 5
U-600265 L30- 85(11-14)-L 1A.120 (cont.)
Q.
Are the bulk drag loads applied to all structures located up to an elevation of 18 feet above the initial pool surface?
A.
All essential structures'within 18 feet of the pool surface are designed for bulk pool swell drag loads.
Q.
When using the data given in Figure 3B-75, is the limitation imposed by the NRC Staff acceptance criterion 1.5 of Appendix C of-NUREG-0978 taken into account?
A.
Yes.
Q.
When using the data in Figure 3B-75, is the ratio a/b limited to values greater than or equal to unity?
A.
Yes.
- Q.
Are the drag load data in Figures 3B-19, 3B-72, and 3B-75 utilized to determine fallback loads?
A.
Yes.
Q.
When evaluating fallback loads, what value of density is used in
.the calculation?
A.
Full-water density, 62.4 lb,/ft3, is used in the bulk pool swell zone. In the transition zone,-linear interpolation between full water density and froth density is used.
Q.
Are all expansive structures at the HCU floor level designed to accommodate the impact load specification required by the acceptance criteria 1.4 of Appendix C of NUREG-0978?
A.
Yes, as committed in IPC letter U-0451 dated April 2, 1982.
i Q.
Is the HCU floor and other expansive structures designed to accommodate an 11 psi two phase or froth drag load as required by Section. 3B.11 of GESSAR-II?
A.
As previously stated, the 11 psi load is considered in the design.
The design controlling load is the impact load.
Q.
When applying the 4.6 psi &P across the grating at the HCU floor level, is the load applied to the total area and not just to the solid portion?
A.
Yes.
I i
i Page 6 e
U-600265 L30- 85(11-14)-L 1A.120 (cont.)
Q.
Does the CPS design include structures that are subject to the loading condition described in Section 3B.12 of GESSAR-II? If the.
answer is yes, have they been designed in accordance with the loads
.specified in Section 3B.12 of GESSAR-II as modified by the acceptance criteria 1.4 of Appendix C of NUREG-0978?
A.~
Yes and Yes. Typical components subject to this load are pipe, pipe supports, valves, conduit and conduit supports close to the containment wall.-
Page 7
3.0-Vo* = V /22 Yo* = Yo/3.11 to* = to/0.141 GE Method per Reference 1
~
/
GE Method with CPS Initial Conditions /
/
'Y*
o 2.0-
/
/
/
/
/
V 6,.Y o
- j
~
/
/
/
/
l.0-
/
/
/
/
/
/
/
N,_,
Vo*
/
p
/
/
\\
/
- es '
\\ s 0
i i
i i
i 0
1 2
3 4
5 6
7 8
9 10 11 12 to*
Figure 1:
Comparison of GE and CPS Initial Condition in GE Jet Methodology
a 5
0 2
\\
\\
\\
\\
\\
\\
. \\
g
\\
\\
N x'
\\
\\
\\
0 y
g
\\
2 o
\\
lo
\\
do
\\
h t
g e
M N
E G
e h
ts
.n mo 5
oi rt 1
Fi d
dn eo vC i
rl o
ea Y
Dit s
ti n
hn o'
gI i
i t
eS i
HP d
.C no p
0 sd C
/
0 vn g
a l
1 y
1 a
/
tE i
iG e
t c
c i
/
oh n
n lt e
I ei r
/
Vw e
S f
P
/
e C
R 2
h
/
r t
e e
i p
w
/
rug
/
d d
5 i
o o
F g
h h
/
0 t
t e
e M
M
/
E E
/
G G
//
/
/ / ! !
0 V
5 4
3 2
1 0
0 0
0 0
0
r Figure 3
~ Hydrodynamic Mass TWO.DIMENSIOllAl. ILODIES T R AHR AMHR HMOMWC MAM SECT 10H THROUGH BODY go.jpeg i
EilRECTI0tt PER UHlf !.ENGTH Ve rucal 8
m,=Inpa
- 14) :
.. y. -.
Ysitical
/
8 m,=1rpa (4) :
Vertical si
=1 epa N)t 8
g Vertical I
" " * ~ " ' ' '
,*, g, p,s (4) e, i
- l..y (6) e Vertical 8
s/b.
pi,.Iaps 141 :
s l
e/b 30 ei,, = 1.14 a p a
- g43, W
e/b = $
es, 1.21 = p s (4) e 8
/
s 26 e/b - 2 se, = 1.3/> e p a
,4 9,
' / ', 'd.
e/b = 1 is, l.11 e p. 8 141 6 l -7, 150 8
s/h = 1/2 m,,
pa (4)
./6 - v3
- e.,,. i. u.,...
m, 8
e/h a1/10 m,
= 2.7) e p a (4)
'%'7 Vertical A
g ar. =.os
,, - i.c i.,...
n,.
]
il/n 10 8
,1 m,,. I,72 e p e
- 14) e i
l Q-u. p af...n m,,. 2 3, p. '
me I
a l'
l
\\
l le a ke
Figure 4 Hydrodynamic Mass for Two - Dimensional Bodies a
n nt,-u-.c a,c,.4, aan, u..
ut.
u ei,.ti.
l a.r.
u.,
ma..
m.u
- 1. % Critu.,
tre.,on a
1
- e. i., e,ia,m.si u.,.
I?. / g,,,
ua..
u.
,2...
1n
.a 7
@J
- e. na..,i, m',
, d1-4.
~$
m i
, p y,,,
,,,,,.a 8 * *=r '< =
n s g,.;
- s. r,. e,ue,u.i n.a...,
rA.
=^
ua...
ut, u.
g a..
utal ri.
1 so 3
go g,gg Aumal Le%'ta h e<<a 5
1.rt "n1* arkt*8 41*eharse, antat:p a
1.g tu.sto.at.u.
1 1.11 L 8* -
1/2 1.70 1/s 1.Se
",.a Esb.
1/10
'2.33 10.
Psette langst Tras.wree to g
ni. usa.a ie I*O b +
Combtsattee er
- 1. b' Stees 9 med 9 t
- j. Pt a s.ett
.A E.-8-~
6 Dt o entat**te rar= 1* dea *
$8,,,, a vtute.o to 126 tas**r ::aa tre '
legia o,
.a
...ui.a.ua..,
2
.83 8'r.12 tot e/e. 9
'O 1
.76 l
- fg
,gy 11.
Land Tbta' Aasalat fresse.re. to y.-g (88 181**1 'P**#A
.l V)
.61 8**'*'*8 A
11 m 8e 1
=
Potating b e,
- Cy116te, e e saceles tat:4. sees a. usessrsn.s,e >> s t. aw. ave les4sa, & >> a
$. 3111ptisal Dtet ser.st te Otst '
b/s E.
k d.
ba?
12 Ctreule, Diet s
%,,4 ts a
1.m y
Ib.)
.991 s
ta 6.00
.96h 3
3.e0
.9m 1.M e' S 1.M
.537 1).
tag his strip
[3_L v.
a b.
- s. J.t.a.aa,,at..
,. "r..
1
. a t.,mt.
t/.
dQg
- 8 s
1 478 3
a.360 1884th L'e
\\
/
)
1.0 23 0>'
a ao IM
\\ /
e E 4/) EM as 4
ITa' M I T. 31113eete Yortigt og s
,s s
g I
I
/b es tat nst teal s
1 I
t.=
. s*
.3=
e.
a n
2.m
.m
.r=
Aa tat 1.00
.121
. Am
- de 1.00 059
.P)$
8.co
. cyp 9.g Cementale so 0
- 1. 0 23 sgnort.at Am.1.s 9
Figure 5
~*
l
. Standard-Drag Versus Aspect Ratio
(
For Rectangular Cross Sections o Nu A*/.' CHI ET At ! Re f 212-6 s Id 5
3, o PE!f ENT LAPIRIVEroTS, &'a ;O ol
$[\\
D' N C,
II
-o ur o
k n
b a
gl v
9ll 1'
s I
i i
I.LO "
Ms.k h
-] :
,a
$.- g '
i i
i 0
l 4
2 h
3 8
3-Y i
25
~\\
L k
Y*
ff 4
g y
U 13-O t=
15 T.
,h :
nse l}
g 0
Ff tsy.t fe%lts A, W Nara;uth., HASNmoto ; Mato 41%
- l 0 FA;e L Jchansm (1927) -
i
_i u
i
?
0 02 04 06 08 to 1.2 l
dih I'
t
{
U-600265 L30- 85(ll-14)-L 1A.120 JUSTIFICATION FOR THE USE OF FIGURE 6-9 0F GENERAL ELECTRIC REPORT No. NEDE-13426P HYDRODYNAMIC MASS FOR CIRCUMFERENTIAL STRUCTURES Mark III pool swell impact tests (Series 5805), reported in Reference 1, indicated that the hydrodynamic mass for circumferential structures was much less than that for the radially oriented structures. This can be seen in Figures 6-8 and 6-9 from Reference 1.
CE has recommended that the hydrodynamic mass values obtained for circumferential targets, which are shown in Figure 6-9, are applicable for circumferential1y-oriented structures, if the structure length is at least as long as one cell width. This is to provide justification for this recommendation.
The quality of the data used to obtain the results in Figure 6-9 is the same quality as the radially oriented beam data used for Figure 6-8.
The total impulse values used for Figure 6-9 were determined from measurements of the acceleration of the impacted structure and the forces in the columns supporting the impacted structure. The impact was also measured with pressure transducers located at the center and both ends of the circumferential targets. The targets extended across the width of the test section which was a 1/3 area scale representation of one cell of a Mark III suppression pool. The impulse data consistently indicated negligible impact at the ends of the circumferential targets as shown in Figure 6-7 from Reference 1.
Even though the impulse data indicates that the local impulse varies significantly over the length of the target, the total impulse values determined from the support column forces and target acceleration (basis for Figure 6-9) are shown in Reference 1 to be reasonably consistent with the local impulse values determined from pressure measurements on the impacted structure. An example of this is given in Figure 6-6 which shows the impulse on the beam from the column force and target acceleration (labeled " Test Data")
compared to the impulse from the impact pressure measurements (labeled
" Traveling Wave Model"). These comparisons, which are described in detail in Reference 1, confirm that the total impulse values which are the basis for Figure 6-9 are consistent with the local impulse values.
Therefore, the data are an accurate characterization of the conditions in the test.
The test facility was a 1/3 area scale model of a single cell of a Mark III suppression pool. The side walls of the test facility were representative of planes of symmetry between cells. There is a small boundary layer at the walls which is less than one inch (Reference 2),
so it can be considered to be insignificant in affecting the velocity and momentum distribution in the pool. Reference 2 discusses how the momentum distribution results in the low impact pressures at the ends of the circumferential targets. The partial velocities of the surface at the ends of the circumferential targets are apparently very low based on the local pressure measurements. These measurement locations on the ends of the targets correspond to the midpoint between vent rows in a Mark III containment, because boundary layer effects at the wall are Page 1
U-600265 L30-85(11-14)-L 1A.120 (cont.)
negligible. Therefore, if any circumferentially-oriented structure in a Mark III containment extends through the plane midway between two rows r
L of vents, it should experience insignificant impact loading at this location.
i i
In summary, the Test Series 5805 values for total impulse for circumferential structures are considered to be valid, because they can be determined.from two independent measurements, i.e., column forces and.
i structure acceleration or local pressure measurements. The correlation for hydrodynamic mass based on the total impulse measurements for circumferential structures (Figure 6-9 from Reference 1) is valid for application to Mark III circumferential structures which extend more than one cell width, because the pool momentum distribution in the test including the region near the side walls is representative of the plant.
Reference 1.
T. R. McIntyre, et al, " Mark III Confirmatory Test Program --
One-Third Scale Pool Swell Impact Test". NEDE-13426P, August 1975.
2.
GESSAR II, Section 3B0.3.2.33, Response 3B.33, Attachment B, Rev. 1 l
l l
l l
I I
I l
l l
i Page 2 L