ML20141G708

To Rept on Vortexing in Condensate Storage Pool

ML20141G708
Person / Time
Site:	Waterford
Issue date:	07/05/1997
From:	Gilmore R, Welcks R ENTERGY OPERATIONS, INC.
To:
Shared Package
ML20141G700	List: ML20141G695 ML20141G703 ML20141G708 ML20141G713
References
NUDOCS 9707100197
Download: ML20141G708 (24)

Text

... _.

i l

Report on Vortexing in the Condensate Storage Pool Revision 1 i

l R.6ttnogy Prepared by:

A /I [&

7/v/92

~< wa Prepared by:

& l& d,s 7/4/f/

s Reviewed by: d%Mk 7477 Approved by:

/[/b 72r//7

/-

a 9707100197 970707 DR ADOCK 0500 2

4

}

Table of Contents i

1.0 Background

1 2.0 Test Program 1

l 3.0 Scaling Analysis Description 2

l 4.0 Test Procedure 5

l 5.0 Results Summary 6

l 6.0 Results Interpretation 11 7.0 Conclusion 13 l

l 8.0 References 14 i

a

].

Figures j

Figure 1 Vortex Submergence vs. Froude Number with Vortex Breakers i

Figure 2 Vortex Types I

Figure C1 CSP Model Flow Diagram Figure C2 CSP Model Plan View

{

Figure C3 CSP Model Side View l

Figure C4 CSP Model Screen & Vortex Breaker Appendix A - CSP Test Notes 3 pages l

1

]

{t Pego 1 of 14 l-l 1.0

' BACKGROUND f

j

' Waterford 3 submitted Technical Specification Change Request NPF-38-179 on July 17,1996 and supplemented on June 3,1997. NPF-38-179 requests the minimum required level in the Condensate Storage Pool (CSP) be increased from 82 percent to 91 percent to account for vortexing and instrument j

uncertainties.

l In support of this Technical Specification change, scaled testing was performed i

to investigate whether vortexing, and thus air ingestion, is a valid concern.

Based on the predicted submergence without a vortex breaker, a vortex breaker 4

is designed and fabricated. This report documents the methodology.

i Revision 1 of this report is generated to include recirculation flow rate (170 gpm) in the discharge flow rate. Also, required flow to the steam generators is revised, l-for a total discharge flow rate of 474 gpm. Testing at the increased flow rate was performed (Tests 12 through 17) to further demonstrate that the analytical approach is applicable to CSP conditions. Note that use of this increased discharge flow rate does not invalidate model testing at lower flow rates. Model testing is performed simply to validate the use of an accepted analytical method j

of predicting vortex formation.

i 2.0 TEST PROGRAM l

Scaled testing was performed on the CSP to determine the height at which type

- 2 vortexing occurs.(See Figure 2). The testing was based on the Froude number, with certain other considerations (i.e. distance of nozzle from pool wall)

I to ensure that scaling can be accurately used to predict CSP behavior. Test rigs j

were designed and built, with a pump and hnse used to draw water from the

{

tank. The required flow rates were measured using a stopwatch and level l

measurement. The flow rate was checked and adjusted each time a test was l

performed, and was set conservatively high. The observation of the onset of a noticeable dimple was used as evidence of a type 2 vortex. Although such a l

vortex would not entrain air into the pump, it was decided that it would be difficult

^

to scale the type of vortex up to the full scale tanks, and so the onset of a i

sustained type 2 vortex was used as the criterion for critical submergence. This i.

was very conservative because it was observed that the onset of actual air L

entrainment, with the vortex breaker, occurred at a water level below the onset of the type 2 vortex, and that sometimes no air entrainment occurred when a i

vortex breaker was used.

a

}

Different size nozzles were used to determine scale effects, and the tests were performed multiple times to demonstrate repeatability. Different flow rates were i

used to vary the Froude and other relevant numbers to determine how the data J

l

Pags 2 of 14

. compared to the correlations and methods for predicting vortex onset, for proper scale-up to the actual full-size pool.

In all tests, appropriate items in the tanks were scaled to the test rig, within a relatively small distance of actual locations. Figures within this report demonstrate the scaling used.

3.0 SCALING ANALYSIS DESCRIPTION Froude Number The Froude number is a ratio of inertial to gravitational forces, and is of paramount importance in vortexing. The Froude number is:

V Fr = /(g)(D) in which Fr = Froude number V = velocity g = gravitational constant D = characteristic length The relevant literature indicates that the velocity is typically taken to be the pipe velocity, and either the submergence (the water level above the pipe) itself is

- taken to be the characteristic length or the nozzle inside diameter is taken to be the characteristic length. When the pipe diameter is taken to be the characteristic length, another dimensionless group is used, which is the required submergence divided by the nozzle diameter, H/D. In that case, a correlation between H/D and Fr is used to determine the required subnergence.

Both of these methods were explored to determine which provides a better tit to the actual test data.

Reynolds Number The Reynolds number is a ratio of inertial to viscous forces, and is defined as:

Re = #

P in which:

Re = Reynolds number p = fluid density V = velocity

i Pcgs 3 of 14 L = characteristic length

= fluid viscosity i

The Reynolds number using outlet velocity and the nozzle diameter for the characteristic length is a critical parameter, in that above a threshold value there is no scaling effect. Above a Re of 100,000, again using the nozzle diameter as the characteristic length, there is no scale effect on head loss, per Reference 5.

In this case, the head loss is not a test parameter because of the very large NPSH margin (27 feet minimum) and the negligible head loss across a vortex breaker.

Therefore, this criterion was not considered to be a test parameter.

Another Re defined as l

Re = Q/( vH) 4 in which Re = Reynolds number O = Flow rate v = Kinematic viscosity H = Submergence l

This version of the Reynolds number is used to ensure that there are no scaling i

effects for vortexing. Because the tests were run to determine the height at which a sustained type 2 vortex appeared, the submergence was not a controlled parameter. The desired minimum value of this version of the Reynolds number is 30000, and this was about what was found in the tests as is shown in Table 3. The values varied over a range, and H/D was plotted as a function of Froude number in Figure 1. It was found that the bounding values of H/D were close to the correlation recommended in Reference 7. Test 9 indicated no vortex at all, which is atypical behavior and reflects the fact that there was no recirculation flow in that test.

Weber Number The Weber number is an indication of surface tension effects, defined as:

We = p*L*V' / cr in which We = Weber number p = fluid density L = characteristic length V = fluid velocity

P 1

Pegs 4 of 14 o = surface tension I

and reference 6 indicates that above a Weber number of 120, with the 4

characteristic length taken as the nozzle diameter, there is no scale effect for L

surface tension. The Weber number is not expected to be as critical for

- determination of when a Type 2 vortex occurs, because the surface is not broken in that case. These criteria were met for all tests in which there was i

recirculation, which is the actual case modeled.

)

Area Ratio of Tank to Nozzle 1.

There are no scale effects if the ratio of tank diameter to nozzle diameter exceeds 16.5, and in this case, the tank diameter is taken to be the hydraulic j

' diameter because the tanks are rectangular. This criterion was met in that the hydraulic diameter of the tank is 48", and the largest test nozzle used was 2" schedule 80 pipe, with an intomal diameter of 1.939". The ratio of these two i

lengths is 24.8, which exceeds 16.5.

l Recirculation Flow I

Recirculation flow into the tanks occurs during the drawdown of the tanks. The predominant feature of the recirculation flow with respect to vortex formation is

[

that it induces circulation in the pool. Flow has dimensions of cubic feet per j

second, in order to model the pool circulation induced by recirculation flow, this 3

flow was divided by the cube of the scale factor of the model, which is the ratio of-outlet nozzle diameters.

3 The recirculation flow is a maximum of 170 GPM. The scale factor for the 1-1/2" i

nozzle is 0.2206, and the cube of the scale factor is 0.0107. The recirculation flow corresponding to this scale factor is 170*.0107, which equals 1.64 GPM. A l

recirculation flow of 3 GPM was used for conservatism for the 1-1/2" nozzle. The scale factor for the 2" pipe is 0.3197. The cube of this scale factor is 0.0327.

l j

The corresponding recirculation flow is 5.56 GPM.. A recirculation flow of 8 i

GPM was used for conservatism for the 2" nozzle. The flow was introduced from i

a horizontal hose,6" from a side wall, allowing the water to fall onto the pool l

surface, similar to the CSP.

Details of Scaling for Condensate Storage Pool (CSP)- All Flow Assumed j

Through One Nozzle

]

In the full scale CSP the three nozzles are 6" schedule 40, and the test for the i

CSP was performed at two scale factors, using a 1-1/2" schedule 160 pipe and a j

2" schedule 80 pipe (see Table 1). The Fr number based on the nozzle diameter L

was used with the correlation between H/D and Fr based on D published in

[

Reference 7. Note the pipe velocity was conservatively preserved (increasing j

the Froude number in the models) in order to maintain the required Re.

j-The mesh around the CSP nozzles is 1/16" wire on 1/4" centers. The test rig j

employed both this size mesh, and window screen which scaled down i

reasonably closely to determine whether mesh size was a significant parameter.

4 I

e

... ~ _

e

l.

Pago 5 of 14 i~

i 2-

' It was found that the mesh delayed the onset of vortexing slightly, but not significantly enough to matter.

Table 1 4

Scale Parameters l

i Full Re '-

Se=ta 1 Seala 2 l

l Nozzle 6" Sch 40 2" Sch 80 1-1/2" Sch q

Ne**ta1.D.

6.065" 1.939" 1.338" Scale 1:1 3.13:1 4.53:1 l

Nozzle Velocity 5.26 5.27 5.27 l

f Fr.VNaD)"

1.30 2.31 2.78 j

Recirc Flow. GPM 170 8

3 i

j Flow. GPM 474 48.5 23.1-l j

4.0

. TEST PROCEDURE General Apparatus An apparatus was utilized which models the CSP when supplying water to the Emergency Feedwater Pumps. The model was designed to allow control and

{

measurement of both the tank discha'ge flow rate and recirculation flow rate.

j (See Figure C1)

' Model Equipment Description I

A model tank was selected with the general shape of the CSP. The model tank

- was sufficiently large to ensure that a ratio of tank size to nozzle size exceeded 16.5:1, so that the tank size would not be a factor in scaling. Two nozzles were l

used, to provide two separate scales. These are described in Table 1. The l

nozzles were cut to the inside diameter of the pipe, reflecting the actual detail in i

the CSP. The cylindrical" top hat" screen over the nozzles was scaled for each L

nozzle in the test rig. The screen model was constructed of the same mesh size as the actual CSP screen (1/4" holes). A sensitivity study was performed, and determined that a smaller screen size (scaled) had little e%ct. Tank surface area and water level drain rate were measured in order to calculais +he model tank discharge flow rate. The recirculation flow rate was measured by discharging into a separate known volurne.

j The CSP model tank was placed indosra in order to eliminate wind effects.

Downstream of a 90' elbow (located at least two pipe diameters from the hole),

flexible hose was connected to route the discharge to the pump. The pump was l

j a Dayton brand,1-1/2",50 gpm centrifugal pump for tests 1 through 11. A 1-1/2 i

j inch globe valve was connected to the pump discharge to allow control of the j

tank discharge flow rate.

(

I y-c

,e.

m,-

..,.,-m.s a4,,

4

. -. ~ - -. - - - - -

_.. - - -., ~. - - -.

Pcgn 6 of 14

{

A tank with a discharge throttle valve was placed above the model tank.

Recirculation was simulated by throttling discharge from this tank through a flexible hose to the tank model. Recirculation injection was in the horizontal plane above the tank model, located at a lateral distance from the model discharge hole which was scaled to the CSP configuration.

The apparatus was refined for tests 12 through 17. The pump was a Yanmar brand,150 gpm diesel driven centrifugst pump. A 0-250 gpm flow indicator and i

3" globe valve were connected to the pump discharge to allow control of the tank discharge flow rate. This improved configuration eliminated the need to measure tank discharge rate based on tank level changes.

Materials & Testing Equipment (M&TE)

As stated above, the tank discharge and recirculation flow rates were calculated

{

- and controlled by measuring the time required to fill / drain a known volume; j

discharge flow was calculated by timing the tank level change rate, and j

recirculation flow was calculated by determining the time required to fill a known volume. The stopwatch (MlFT357.012) used for these measurements was a calibrated (M&TE) device. Tank level was measured with a measuring tape and

- ruler for tests 1 through 11, which were keer calibrated for accuracy by the plant meteorological lab standard length calibration device. The flow indicator utilized for tests 12 through 17 is a certified calibrated device.

j Procedure i

(

The tank was allowed to settle prior to each test. Tests commenced only after all l

surface rotation ceased. The tank was also leveled prior to testing.

j Each test commenced by setting the discharge flow rate, and introducing the required recirculation flow rate. The tank was then observed as water level dropped, and the tank height at which a type 2 vortex occurred was documented.

I Each test was recorded on video tape, and later reviewed to ensure accurate documentation of the results.

I s

5.0 RESULTS

SUMMARY

Correlations reported by Hecker of Alden Research Laboratories in Reference 7 were applied to compare to the empirical results. The correlation reported by Hecker in Reference 7, attributed to Reddy and Pickford, for required submergence with a vortex suppresser was found to provide reliable, repeatable predictions of vortex formation for the CSP. The correlation recommended for use in general in Reference 7 was found to conservatively bound the data without a vortex suppresser.

The flow velocity at the nozzle was preserved at full scale based on the conservatism assumption that all of the flow is through one EFW suction header.

e

Pcg3 7 of 14 This is 5.26 feet per second. Tests were performed with various configurations.

Tests 1 through 3 were performed to study vortex formation without a vortex breaker and to deterrr..ne the effect of different size mesh configurations. Tests 12 through _17 scale to actual expected CSP flow rates. Figure 1 and Table 4 4

compare the correlation reported by Hecker (H/D = Fr Mth the vortex suppresser) to the test results. Within the limits of experimental uncertainties, d

i the empirical results provide very good agreement with this correlation. The i

correlation reported by Hecker with no vortex supprescer (H/D = 1+2.3*Fr) is also compared to test 1.

The highest, or bounding, test results were about 10% higher than the 4

correlation, so additional conservatism was used in selection of a required submergence for the CSP. The required submergence for the CSP was chosen i

for the case in which all flow is conservatively assumed to pass through only one EFW suction header. The flow and hence the Froude number would be only half of the assumed flow, and the actual level at which vortexing occurs would be l

about half of the required submergence for the CSP. This represents a conservatism in the minimum level required from this testing.

t 1

i i

2 Y

Pcgo 8 of 14 Table 2 i

Results for Condensate Storage Pool Suction Recirc.

Nominal Screen Type Date Test Flow Flow Pipe Config.

2 Comments No.

(gpm)

(gpm)

Size Vortex (in.)

Height (in) i 6/22/97 1

20 3

1.5 No VB 6.5 Air ingestion at 3.5" 2

20 3

1.5 Fine 4.25 Air ingestion Below Screen mesh (approximately 1")

3 20 3

1.5 Fine 2.25 Air ingestion at 2.25" mesh 6/23/97 4

20 3

1.5 VB 2.5 Air Ingestion at 1.75" 5

20 3

1.5 VB w/

2.0 Air ingestion at 1.5" tail 6/24/97 6

40 8

2.0 VB 4.0 Air ingestion at 2.0" 7

40 8

2.0 VB 4.0 Air Ingestion at 2.0" 8

20 8

2.0 VB 2.0 No Airingestion 9

20 None 2.0 VB None No Air ingestion i

10 20 8

2.0 VB 2.0 No AirIngestion 11 40 8

2.0 VB 3.0 Air ingestion at 2.0" 7/3/97 12 48.5 8

2.0 VB 4.8 None i

13 48.5 8

2.0 VB 4.7 None i

14 48.5 8

2.0 VB 4.6 None 15 23.1 3

1.5 VB 1.7 Air ingestion at 1.2

• 16 23.1 3

1.5 VB 1.3 Air ingestion at 1.2 "

17 23.1 3

1.5 VB 1.3 Air ingestion at 1.2

• Notes VB = Vortex Breaker, with mesh screen made of 1/16" wire on 1/4" centers, as in the full scale.

Notes:

1 The fine mesh was window screen, which approximately scaled to the test rig.

2 In tests 6 through 11, no fine mesh on top of the strainer was used, because there were no appreciable differences among results with a full-scale screen, a window screen, or no screen. This facilitated observation of vortexing below the top of the strainer.

3 The test rig was constructed to reflect a situation in which all of the flow passes through one of the two nozzles in the actual CSP. The actual case is one in which the flow passes through both nozzles. Because of this conservative approach, the radial Reynolds number was not as high as desired in tests 8 through 10, which were run at flow rates more appropriate (although slightly less than) to the actual case. To account for uncertainty in scaling, the setpoint is taken as the CSP level required as if all flow were through only one EFW suction header.

Page 9 of 14 Table 3 Critical Non-Dimensional Numbers for the CSP and its Scale Model onset of Sustained l

CSP l

l Scale blodel J

Type 2 Test Nozzle Flow Velocity Reynolds Weber Nozzle Flow Velocity. Reynolds Froude Weber Vortex (h)

Number Dia (In.)

(gpm)

(ft/sec)

No., D No Die (in.)

(gpm)

(fttsec)

No.

• No. a No. 3 (In.)

1 6.065

,381 4.234 229853 3501 1.338 20 4.566 8838 2.410 899 6.50 2

6.065 381 4.234 229853 3501 1.338 20 4.566 13517 2.4.0 899 4.25 3

6.065 381 4.234 229853 3501 1.338 20 4.566 25531 2.410 899 2.25 4

6.065 381 4.234 229853 3501 1.338 20 4.566 22978 2.410 899 2.50 5

6.065 381 4.234 229853 3501 1.338 20 4.566 28723 2.410 899 2.00 6

6.065 381 4.234 229853 3501 1.939 40 4.349 28723 1.906 1131 4.00

_7 6.065 381 4.234 229853 3501 1.939 40 4.349 28723 1.906 1181 4.00 8

6.065 381 4.234 229853 3501 1.939 20 2.174 28723 0.953 295 2.00 9

6.065 381 4.234 229853 3501 1.939 20 2.174 N/A 0.953 295 0.00 10 6.065 381 4.234 229853 3501 1.939 20 2.174 28723 0.953 295 2.00 11 6.065 381 4.234 229853 3501 1.939 40 4.349 38297 1.906 1131 3.00 12 6.065 474 5.267 285959 5419 1.939 48.5 5.273 29022 2.312 1736 4.80 13 6.065 474 5.267 285959 5419 1.939 48.5 5.273 29639 2.312 1736 4.70 14 6.065 474 5.267 285959 5419 1.939 48.5 5.273 30284 2.312 1736 4.60 15 6.065 474 5.267 285959 5419 1.338 23.1 5.274 39029 2.783 1199 1.70 16 6.065 474 5.267 285959 5419 1.338 23.1 5.274 51038 2.783 1199 1.30 17 l

6.065 474 5.267 285959 5419 1.338 23.1 5.274 51038 2.783 1199 1.30

' - Reynolds No. = O/vh -

Should be around 30,000 or more per Reference 7. However, the submergence at which a sustained Type 2 vortex appeared was not a controlled parameter. Therefore, the test was run over a range of Froude numbers and Reync4ds numbers to determme if the scahng relationship desenbed in Reference 7 held. In tests 4 through 11, which were for the vortex breaker model, this Reynolds number is approximately 30,000 2 - Froude No. = V/(gD)u2 8 - Weber No. = pDV /Segma - Should be greater than 120 per Reference 6 Temperature

(*F) 80 (Lb - sec/ft')

1.80E-05 Viscossly (u) f Viscosity (v)

(ft /sec) 9.31E-06 Density (p)

(Lbm/ft')

62.23 Sigma (Lb ttt) 0.005 f

Page 10 of 14 Figure 1 Vortex Submergence vs. Froude Number with Vortex Breakers

3. 5 ---- - ---

Test rs 4. 6 through 17 - Cruciform inside Screen O Onset of Air ingestion Test #5 - Cruciform Modified witti a Tad Extending into Pipe Two Diameters e Onset of a Sustained Type 2 Vorter Test #9 - No Recirculaton Added into Scale Model 3.0

.~

j.**

Test rs 12,13.,1,

[

.. ~~

e Test rs 6 & 7,. *

• 8 E 2.0 H/D = 1.1*V/(gD)112.

_cn Required Submergence

,.~ '.e Test # 4 5 e

for CSP (S~)

I

~.-

Test # 11

\\

g e 1.5

-xr Test # 5 a -

~

t h

,... ~

Test # 4 o Test #15 m l

Test rs B & 10,.: ~ ~

Test rs 6,7 & 11 Test # 5 0 Test rs 15,1 5& t7 1.0 -

- - - - H/D = V/(gD)w O

L--

8 Test r31s,17 0.5 Ref.: Swithng Flow Probledis at intakes, edited by Knauss, contributon by. G Hecker, pubirshed by A.A. Abalkema,1987.

Test rs 8,9 & 10 0.0 a'

Test # 9 0.5 1.0 1.5 2.0 2.5 3.0 Froude Number (V/(gD))

I Pags 11 of 14 i.

l 6.0 -

RESULTS INTERPRETATION j

Comparison with the correlations reported by G. Hecker in Reference 7 are shown in Figure 1 and Table 4. Tests 12 through 14 used the 2" pipe, with flow l

scaled to expected CSP flows. These tests utilized the vortex breaker. - The l

average height at which a type 2 vortex was observed for these tests was 4.7";

l

. the correlation predicts 4.5". This is the most representative series of tests for L

the CSP, and the correlation is very good. All other vortex breaker tests (tests 4-11 and 15-17) also correlated well with Hecker.

The use of a " tail" on the vortex breaker (test 5) provided better results than no tail (test 4), but it was decided that there was not a sufficient improvement to warrant the extra construction burden. Tests 2 and 3 demonstrated that a finer mesh had little effect on vortex formation. Therefore, all further testing conservatively used larger mesh size (1/4").

The correlation used in the right-most column (Table 4) to predict the onset of vortices is appropriate whether there is a vortex breaker or not. With a vortex breaker, the recommended correlation is H/D = Fr. With no vortex breaker, the recommended correlation is H/D = 1 + 2.3

• Fr.

t n

i I

i

\\

Pgga32og34 Table 4 Actual and Predicted Measurements Test Number Fr,D Vortex Measured Height of Type 2 Predicted Height (In.)

V/(gD)1i2 Breaker Vortex (In.) from Tests From Ref. 7 1

2.41 None 4.5 8.75 2

2.41 None 4.25 8.75 3

2.41 None 2.25 8.75 4

2.41 X

2.5 3.22 5

2.41 X with 2.0 3.22 Tail 6

1.906 X

4.0 3.7 7

1.906 X

4.0 3.7 8

.953 X

2.0 1.85 9

.953 X

None 1.85 10

.953 X

2.0 1.85 11 1.906 X

3.0 3.7 12 2.31 X

4.8 4.5 13 2.31 X

4.7 4.5 14 2.31 X

4.6 4.5 15 2.78 X

1.7 3.7 16 2.78 X

1.3 3.7 17 2.78 X

1.3 3.7 X = Cruciform vortex breaker above the nozzle to fit into existing screen

a

'4 Paga 13 of 14 i

l 7.0

. CONCLUSION l

The Condensate Storage Pool submergence required to prevent vortexing, j

based on the correlation reported by Hecker in Reference 6, is 7.9" (Froude x D

= 6.065" x 1.30). The testing program described here supports this correlation as a bounding equation, except for the results of Tests 6, 7,12-14. The results j

of these tests indicated a need for about 10% more submergence to prevent Type 2 vortexing than this correlation predicted. This would yield a required j

submergence in the CSP of 8.7". Although the testing was done conservatively, j

in that the onset of a sustained type 2 vortex was considered the criterion for critical submergence instead of air ingestion as the criterion, and the actual outlet I

velocity and Froude number would be half as high in the CSP compared to that

]

used for choice of the level _setpoint, additional conservatism was added to the results to account for possible test uncertainties.

i The submergence value is taken as 9" in the Condensate Storage Pool with vortex breakers installed.

l i

)

e l

i y

Pcgs 14 of 14

8.0 REFERENCES

1. NUREG/CR-2758,"A Parametric Study of Containment Emergency Sump Performance"

2. NUREG/CR-2759, "A Parametric Study of Containment Emergency Sump Performance: Results of Vertical Outlet Sump Tests"

3. Crane Technical Paper No. 410,1988

4. Waterford 3 Calculation EC-M35-012 " Minimum Pipe Submergence to Prevent Vortexing"

5. NUREG/CR62-2760, " Assessment of Scale Effects on Vortexing, Swirl, and Inlet Losses in Large Scale Sump Models"

6. " Vortex Formation at Vertical Pipe intakes, by Jain, Raju and Garde, Journal of the Hydraulics Division, ASCE, HY10, October,1978, p.1429ff

7. Swirling Flow Problems at intakes, edited by Knauss, contribution by G.

Hecker, published by A. A. Abalkema,1987

8. Waterford 3 Calculation MN(Q)-10-9,"NPSH Available to Emergency Feedwater Pumps"

9. Waterford 3 Design Basis Document W3-DBD-003," Emergency Feedwater System"

10. Mark's Handbook for Mechanical Engineers, by Avallone and Burmeister McGraw-Hill,9'th edition 11.ASME Steam Tables j

1 12.Waterford 3 SES isometric Diagram 4305-6636 Rev. 9 13.Waterford 3 SES isometric Diagram 4305-6634 Rev. 8 14.Waterford 3 SES Drawing 1564-G-905 Rev. 5 i

15.Waterford 3 SES Drawing 1564-G-906 Rev. 3 16.Waterford 3 SES Drawing 1564-G-907 Rev. 9 17 Mechanical Engineering Reference Manual,8'th Edition, by Lindeburg, Professional Publications i

l

VORTEX TYPE 1

V, INCOHERENT SURFACE SWIRL 2

SURFACE DIMPLE:

COHERENT SWIRL AT SURFACE i

3 DYE CORE TO INTAKE; COHERENT SWlRL THROUGHOUT WATER COLUMN i

i 4

VORTEX PULLING FLOATING y

TRASH, BUT NOT AIR 2

5 VORTEX PULLING AIR h

BUB 8LES TO INTAKE g........

b 6

FULL AIR CORE q

5 TO INTAKE f

a d

Figure 2

i 1

i RECIPC.

TANK i

1

)

1 CSP w/

VN

\\

THROTTLE i

PUMP CSP MODEL FLOW DIAGRAM FIGURE C1

i 4

i l-1 j

4 i

i A

i i

a I

6%'

9%g' i

=

r lb*o SCH 160 i

m i

e 4

i 2*eSCH 80 i

1 i

I i

i t

RECIRC.

i e

a

.-,i -

i a

i 48-

=

CSP MODEL PLAN VIEW FIGURE C2

(

1

1 I

i

[

t i

48-I~

=I u

cc v

IV2'eSCH isa 2'OSCH 80 CSP NOJEL Sf0E VIEW FllGURE C3 1

1 i-..

-r-

t 5

VCRIEX BREA<ER

~

~

i E 1 i

4 I

1

'x w-17e*

4 4

N L

a i

i 1%*e PIPE 1

i VORTEX BREAKER m,

='

l l

1 m

s-

?

2%

j h

~*

\\

o i

l 2'e PIPE

)

i 7

I

)

4 1

4 4

1 CSP MODEL SCREEN & VORTEX BREAKER FIGURE C4 1

t 4

i t

n

Appendix A l

CSP Test Notes 1)

Wire Mesh Recirc. line to the left (looking at flow drain). Recirc coming into the tank

=1' from the edge of tank Type 2 vortex at 6.5" l

Air ingestion at 3.5 2)

Recirc:

3 gpm i

6" from side l

1-1/2' from Hole 1/2" Screen with fine mesh (1/16" dia.)

Pump flow: 20 gpm 4.25 Type 2 Air ingestion below screen (approximately 1")

3)

Swirl at 5'25" 3

Swirl at 4.5" l

Disappeared at 4.0" Swirl at 3.0" Type 2 Vortex & Air at 2.25" 4)

Measured Pump flow at 20 gpm Recirc. 3 gpm 6" off the side 1-1/2" pipe 1" cruciform without cage Swirl / Vortex at 2.5" Air at 1.75" 5)

Same as (4) y r

Vortex at 2.0" Air at 1.5" 6) 2" drain 8 gpm recirc Standard VB

_ _.. _ ~.. -.

APP 1.A P C.

2 of 3 i.

Swirl at 5"(Type 1) i Vodox at 4" Air at 2" i

7)

Recire. 8 l

Outlet 40 i

Swirl / Vortex et 4" Air @ 2" l.

8)

Outlet gpm: 20 Swirl at 4" 3

Vortex at 2"(No air) l 9) 20 gpm no recirc j

no Vodex no air I i

j 10)

Same as (8) j Swiri at 2.5"(Type 1)

Vortex at 2.0"(No air) j 11) 8 gpm recire 40 gpm same as (6)

Vortex at 3.0" Air at 2"

7. ca w, y Wa &,m [2w, h ses fr /rs}s 9.jj )

,,, /-/- r Performed By: f?,La & L.

(/V L/

G 30 MJ;L

[for; i-5)

"/ [/

Witnessed By:

s u

/

i1 J

l APPY. 4 PC. 3 or 3 THURSDAY JULY 3RD,1997 l

CSP Retest for 2" Dia. Drain

' Pump Flow @ 48-49 GPM l

Recirc Flow @ 8 GPM l

Cruiciform Vortex Breaker for the 2" Drain with 1/4" Mesh Screen Installed Test Results:

TEsY $

Test 1) Type 2 vortex @ 0.40 feet, Vortex with funnel @ 0:38 feet j

i v.

!3 Test 2) Type 1 vortex @ ~ 0.5 feet, Type 2 @ ~ 0.40-0.38 feet, Vortex with funnel @ 0.38 feet i4 Test 3) Vortex with funnel @ 0.38-0.36 feet er vsso Test 4) No VB Installed - @ 0.44 feet vortex with funnel CSP Retest for 1 1/2" Dia. Drain Pump Flow @ 23 GPM Recirc Flow @ 3 GPM Cruiciform Vortex Breaker for the 1 1/2" Drain with 1/4" Mesh Screen Test Results:

15 Test 1) Type 1 vortex @ 0.14 feet, Vortex @ 0.12 feet, Air ingestion @

0.12-0.10 feet 4

Test 2) Type 1 vortex @ 0.16-0.12 feet, Type 2 vortex with air ingestion

@ 0.11-0.10 feet l

17-Test 3) Type 1 vortex @ 0.14 feet, Type 2 vortex with air ingestion @

O.11-0.10 feet per v sto Test 4) No VB Installed - Type 1 @ 0.16 feet, Vortex with air ingestion @

0.12 feet i

i Performed By:

.7.9. Pere 2 k 7 f7 Y"

Witnessed By:

F.

NA 6As m N

hA

~7-3 -9 7 1

i

.