ML20030B463

Forwards Revised Model Study of Reactor Containment Sump Flow Characteristics

ML20030B463
Person / Time
Site:	Summer
Issue date:	08/10/1981
From:	Nichols T SOUTH CAROLINA ELECTRIC & GAS CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
NUDOCS 8108180145
Download: ML20030B463 (15)

Text

{{#Wiki_filter:SOUTH CAROLINA ELECTRIC a GAS COMPANY D eost omer son re. CotuMe:A, SOUTH CAROLINA 29218 y T. C. NicHOL S, J n. Agust M, im ,h ~+=;;;,;;o;;,- I .6(LL Lu Aucf171981* ~2 Mr. Harold R. Denton, Director u,h g Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission k Washington, D. C. 20555 g us

Subject:

Virgil C. Summer Nuclear Station Docket No. 50/395 Model Study of Reactor Building Sump

Dear Mr. Denton:

In a previous submittal on the model study of the reactor building sump dated July 29, 1981, South Carolina Electric and Gas Company committed to re-issue the study after several minor errors had been corrected. The report has now been re-printed and we hereby provide twenty-five (25) copies. As described previoitsly, no significant vortex phenomena occurred in the model test program. Therefore, South Car:lin Electric and Gas considers the existing sisap design acceptable. If you have any questions, please let us know. Very truly yours, / T. C. Nichols, Jr NEC:TCN:lkb Attachments i I cc: Page Two l \\ 40 9 'V q' @)

• f</

l f 'l 8108180145 810810 PDR ADOCK 05000395 A pop _..~ _ _

Mr. H. R. Denton August 10, 1981 Page 2 cc: Mr. H. R. Denton (25) Y. C. Sumer w/o attach. G. H. Fischer w/o attach. T. C. Nichols, Jr.w/o attach. H. N. Cyrus w/o attach. J. C. Ruoff D. A. Nauman W. A. Williams Jr. w/o attach. R. B. Clary O. S. Bradham A. R. Koon M. N. Browne

3. A. Bursey J. L. Skolds J. B. Knotts, Jr.

H. E. Yocom w/o attach. J. B. Cookinham w/o attach. C. A. Price NPCF w/o attach. File w/o attach. t i ,.._._,. ~.-.._....-...-.-..._.-.__ ~._...

~ ALDEN RESEARCH LABORATORY WORCESTER POLYTECHNIC INSTITUTE May 6, 1981 Mr. Gary Moffatt South Carolina Electric & Gas Company Post Office Box 764 Columbia, SC 29202 VCS/ SUMP-ARL-260E

Dear Mr. Moffatt:

With respect to the question of conducting high temperature tests as part of the ongoing reduced scale model study at the ARL for the Virgil C. Dimmer Station, we would lik2 to provide a genert_ review of model similitude con-siderations, a summary of tests on the scale model to date, and a review of why high temperature tests will be conducted as part of the full scale para-metric sump study through Shndia for NRC-DOE. This will be followed by wr recorrendation regarding the possible need to conduct high temperature tests using the V.C. Summer model. Review of Model Scaling Considerations General Models involving a free surface are conetructed and operated using Froudo similarity since the flow process is controlled by grarity and inertia forces. The Froude number, representing the ratio of inertia to gravita-tional force, F = u/ h (1) where u = average velocity in the pipe g = gravitaticnal acceleration s = submergence is, therefore, made equal in model and prototype F = F /F =1 (2) r m p where m, p, and r denote model, prototype, and ratio between model and proto-type, respectively. i l HOLDEN MASSACHUSETTS 01520 + TELEPHONE 617-829-4323 \\ /

r ~s Page 2 Mr. Gary Moffatt May 6, 1981 In modeling an intake sump, it is important to select a reasonably large geom tric scale to achieve large Reynolds numbers and to reproduce the o c..rved flow pattern in the vicinity of the intake (4). An asymptotic be-havior of energy loss coefficients with Reynolds number is usually observed in similar flows (2). Hence, with Fr = 1, the basic Froudian scaling cri-terion, the Euler numbers, E, will be equal in model and prototype. This implies that the flow patterns and loss coefficients are equal in model and prototype. Similarity of Vortex Motion Fluid motions involving vortex formation in sumps have been studied by several investigators (1, 4, 5, 6). In addition to the primary forces of sravity and inertia, viscous and surface tension forces could influence the formation and strength of vortices (1, 5). The relative magnitude of these forces to the fluid inertia force is reflect-ed in the Reynolds and Weber numbers, respectively which are defined as: u d/v (3) R = (4) W = (o/pr) ! where d = intake diameter r = characteristic radius of vortex o = surface tension force per unit length v = kinematic viscosity, a function of water temperature-p = mass density per unit volume l It is important for a reduced scale model study to ascertain any deviations l in similitude (scale effects) attributable to viscous and surface tension forces in the interpretation of model results. For large R and W, the ef-fects of viscous and surface tension are minimal, i.e., inertial forces pre-dominate. Surface tension effects are negligible when r is large, which will be true for weak vortices where the free surface is essentially flat. Con-versely, only strong air core vortices are subject to surface tension scale effects. Moreover, an investigation using liquids of the same viscosity but 3 3 different surface tension coefficients (a = 4.9 x 10 lb/ft to 1.6 x 10 lb/ft) showed practically no effect of surface tensic.a forces on vortices (1). The vortex severity, S, is therefore mainly a function of the Froude number, but could also be influenced by the Reynolds number. S=S (F, R) (5) /

] Page 3 Mr. Gary Moffatt May 6, 1981 Anwar (4) has shown by principles of dimensional analysis that the dynami : similarity of fluid motion at an intake is governed by the following dimen - sionless parameters 40__ u Q d_ g vs 2s ud /2gs g where Q = discharge through the outlet u = tangential velocity at a radius equal g to that of outlet pipe The influence of viscous effects was defined by the parameter Q/(v s), known as a radial Reynolds number, BR. For similarity between the dimensions of a vertex of strengths up to and including a narrow air-core type, it was shown by experiments that the influence of RR becomes negligible if Q/(v s) was 4 greater than 3 x 10 (4). For the prototype of this study, the value of R for the operating temperature range of 70* and above, and using the submer R 5 gence to the floor grating, was greater than 1.1 x 10. In the scale model, 4 the value of R for the RHR sumps was 2.6 x 10 for Froude velocity and 4.4 R 4 x 10 for prototype velocity, both for water temperatures of 50'F. Thus, vis-cous forces would have a negligible role in this model study. Dynamic simi-2 larity is obtained by equalizing the parameters 4Q/u0d, u//2gs, and d/2s in model and prototype. A geometrically similar Froudian model satisfies this condition. To compensate for any possible excessive viscous energy dissipation (Reynolds number scale effect) and consequently less intense model vortex, various in-vestigators have proposed increasing the model flow and, therefore, the' intake and approach velocity, since the submergence is maintained constant. Operating the model at the prototype inlet velocity (pipe velocity) is believed by some researchers to achieve the desired results (1). This is often referred to as the Equal Velocity Rule. The test procedure for the present study incorporat-ed testing the model at prototype pipe velocities to achieve conservative pre-dictions. ARL Vortex Activity Projection Technique ARL has conducted independent research to assure that no scale effect on vor-tex activity due to Reynolds number exists in models with weak vortices. A technique was developed (9) to extrapolate model vortex activity to prototype Reynolds numbers by using elevated model water temperatures and varying model flow velocity (Froude ratio), and this has been applied to several studies (7, 8, 10, 11, 12). Figure 1 illustrates the method. The ordinate, F, is the j

Page 4 Mr. Gary Moffatt May 6, 1981 ratio of nodel to prototype Froude number, while the abscissa is the inlet pipe Peynolds number, R. The objective is to determine flow conditions at Fr = 1 at prototype R from tests at lower than prototype R. Assume the model to operate at flow less than Froude scaling (Fr < 1) at point al. By increasing the discharge in the model while keeping the same submergence and water temperature, F and R are increased in increments, corresponding r to a point, aN, where a vortex of type N is observed. The model Reynolds number can also be changed by varying the kinematic viscosity with tempera-ture changes, and similar tests performed to locate b, another point on N the locus of type N vortices. Extrapolation of the line of constant vortex strength of type N can be made to a prototype Reynolds number at the proper Froude number (Fr = 1), point P. The locus could represent any expedient N measure of vortex severity, including inlet loss coefficient and $nlet swirl angle. Any scale effects due to viscous forces would be evaluated and taken into account by such a projection procedure. Figures 2, 3, and 4, and Table 1, show the effect of increasing Reynolds num-ber and increasing Froude ratio for the final designs of fcur containment sump model scudies conducted at ARL having scale ratios of 1 to 2.5 to 1 to 4. The vortex types designated on these figures are shown in Figure 5. In all cases, the data show no measurable changes in vortex strength, even for types 3 and 4, with Reynolds number at constant Froude ratio. This is reasonable since the Reynolds numbers are all above the limiting value (1, 4), a previously des-cribed similitude requirement. Minor increases in vortex strength occur when the Froude ratio is increased. Other measurements, such as swirl in the inlet pipe, have also shown no measurable dependence on Reynolds number. This indi -~~7 cates that reduced scale model tests are a direct indication of prototype per-formance for weak vortices, particularly if vortex suppressors are part of the design, even at Froude scaled flow (i.e., F = 1). Tests at higaer than Froude r scaled flow are seen to give conservative results, i.e., somewhat stronger vor-tices than expected in the prototype. V.C. Summer - Test Results Approximately 50 conditions of varying approach flow distribution, and vari-ous combinations of blocking the bar racks and screens have been tested in the reduced scale model. Measurements include classification of vortex type, swirl in the inlet pipe, and inlet loss coefficient. These tests show that no vortices stronger than type 1 (incoherent surface swirl) occurred for any combination of test variables, including operating the model at prototype inlet velocities as well as at Froude scaled velocity. The only swirl ob-served is generated by obstructions in the approach flow, and these swirls are weak (type 1) and transient. Pipe Reynolds numbers are greater than 1.2 x 305 for the minimum flowrate modeled, the RB spray inlet. This Rey-nolds number is comparable to the minimum for the previous studies which indicated no increase in vortex activity for increasing Reynolds numbers at constant F' oude ratio. 1 The horizontal floor grating and the relatively deep submergence of the in-let pipes apparently act to suppress the formation of any coherent vortices above the reactor floor.

Page 5 Mr. Gary Moffatt May 6, 1981 Full Scale Parametric Sump Study The generic study of full scale containment sumps being conducted at the ARL via Sandia Laboratories for the NRC-DOE addresses an unresolved safety issue (Task A-43) associated with ECCS systems. As part of this general study, testing will be conducted using water heated to about 160*F to approach pro-totype operating conditions. Two sump configurations will be selected for those tests, both configurations having been previously testing with normal water temperature, and showing evidence of intermittent air core vortices and air ingestion into the inlet pipes. The basic reason for the high temperature tests are to determine the effects of decreased water viscosity and surface tension on two phase flow phenomena involving air and water. The effects of increased vapor pressure per se' are expected to be negligible since the amount of air and water vapor released from the liquid phasa is minimal compared to the ingestion of air by vorti-ces. Also, the effects of increased Reynolds number on single phase flow parameters such as inlet loss coefficient and swirl angle (except due to change in vortex strength) are expected to be minimal since Reynolds numbers with normal water temperatures are already above any transitional values for changes in flow patterns. Basically, the high temperature tests are to deter-mine if a coherent but weak vortices (type 3 or 4) becomes an air core vortex, and if a weak air core vortex becomes stronger and ingests more air (i.e., will the void fraction increase). Coherent vortices could become stronger with higher water temperature since vortex cores are regions of high shear (velocity gradients) and a decrease in visccsity could increase tangential velocities. Based on work by others, the effects are expected to be small since the Reynolds number for all test conditions are well above suggested minimum values relative to scale effects (1, 4). All testing will be con-ducted without any vortex suppressor such that the vortex can become stronger should that be the tendency. Evaluation and Conclusions Based on general similitude considerations, the results from ae V.C. Summer sump reduced scale model testing program, and the intent of the high tempera-ture tests in the full scale sump facility, the following conclusions may be made regarding high temperature tests for the V.C. Summer model. The V.C. Summer model was designed for minimal scale effects at Froude scale flow, but was also tested for conservatism at prototype velocities. No co-herent vortices were observed for any combination of test parameters. Past work at ARL with high temperature testing of scale models has shown that swirls and weak vortices are not subject to scale effects when Reynolds num-bers are above recommended minimum values. The horizontal floor bar rack of the V.C. Summer sump acts as a vortex suppressor, which together with the n. r.,-- - - - - - - -..,.

_ _ _ _ _ ~ _. _.. i Page 6 q .Mr. Gary Moffatt May 6, 1981 lack of coherent vortices and scale effects, makes the question or how 1 higher Reynolds may effect vortices a moot point. There is no technical reason that the observed weak swirls (type 1 vortices) will become ob-jectionable vortices for the V.C. Summer sump design, and tests at ele-vated water temperatures are, therefore, not warranted in this case. Please contact us should you have any questions or if you would like any additional material. Sincerely, l ?['/ >*. - f4

s "

-*L George E. Hec er Professor and Director GEH/nnb Enclo sures i 1 l ( i e a f i i h d

r REFERENCES 1.

Daggett, L.L.,

and Keulegan, G.H., " Similitude Conditions in Free Surface Vortex Formations," Journal of Hydraulics Division, ASCE, Vol.100, pp. 1565-1581, November 19.4. 2.

Daily, J.W.,

and Harleman, D.R.F., Fluid Dynamics, Addison-Wesley Pub-lishing Company, 1965. 3. Rouse, H., Handbook of Hydraulics, John Wiley & Sor.s,1950. 4.

Anwar, H.O., Weller, J.A.,

and Amphlett, M.B., " Similarity of Free-Vortex at Horizontal It. (e," Journal of Hydraulic Research, IAHR 16, No. 2, 197f. 5. Hattersley, R.T., " Hydraulic Design of Pump Intakes," Journal of the Hydraulics Division, ASCE, pp. 233-249, March 1965. 6.

Reddy, Y.R.,

and Pickford, J., " Vortex Suppression in Stilling Pond Over-flow," Journal of Hydraulics Division, ASCE, pp. 1685-1697, November 1974. 7. Durgin, W.W., Nea3 e, L.C., and Churchill, H.L., " Hydrodynamics of Vortex Suppression in the ?aactor Building Sump Decay Heat Removal System," ARL Report No. 46-77/M202FF, February 1977. 8. Padmanabhan, M., " Hydraulic Model Studies of the Reactor Containment Building Sump, North Anna Nuclear Power Station, Unit 1," ARL Report No. 123-77/M250CF, July 1977. 9. Durgin, W.W., and Hecker, G.E., "The Modeling of Vortices at Intake I structures," Joint Symposium of Design and Operation of Fluid Macninery, Colorado State University, June 1978. 10. Padmanabhan, M., " Hydraulic Model Investigation of Vortexing and Swirl Within a Reactor Containment Recirculation Sump," D.C. Cook Nuclear Power Station, ARL Report No. 108-78/M178FF. 11. Padmanabhan, M., " Assessment of Flow Characteristics Within a Reactor Containment Recirculation Sump Using a Scale Model," McGuire Nuclear i Power Station, ARL Report No. 29-78/M208JF. i 12. Padmanabhan, M., "'.nvestigation of Vortexing' and Swirl Fithin a Contain-ment Recirculation Sump Using a Hydraulic Model," Seabrook Nuclear Power Station, ARL Report No. 25-81/M296HF, 1981. a / j

t g E GNAR E P N Y O TO Y T T S O R I R RE P EB VM x EU S N F s X S 3 E D T[ T L R O \\ O N / S" V Y E F O R = N E D R C N i / 3 NA c E 1 4 D E a N D u b U E b PO { I 4 ER 8 DF X /. E T 3 1 R 0 O E V R U N r/ 2a G E I P F YT ta SU CO L I 2 m P F I = F I \\

111 b S E P TY F NT EP 5 E

• 3 Y 6 SX 1

NEE 2 T I RT 3 O - 7 HPR T G TEO L O F N I R W V E R I 2 P 0 W 7 S M E ST E C O REU S A H EKM A F S B CI C R MA X R UR A U S E NBM BP R E MM E T UU T A ) O N S N W E 3, S F D H 2 0 ( b LT 5 l 1 ) O N 3, l NI l2( Y F 5 5 S E v R 2f ) '/ P / 1 d Y 3, / u SU T l2 F ( / S R X E R 0 ) f B E I l[( b M T I 0 3 E 1 ' / V U R N O O r il S I V D T ( L AD O RE N V Y R R ld 2 E E E R B S )3 F 2 MB '8 UO ( 4 N M I E U DM UI ) O X 1( R A F M } ) / ); 2 1 2( ( E R U GI F 0 / ) 0 1 1( 0 5 1 5Qc c fl 1 lll

-l U l l l Il 1: I 4 4 4 I MINIMUM WATEH LEVEL MAXIMUM FLOW BOTH PlPES OPEF;ATING (CASE 8) BLOCKAGE SCHEME NO.12 = NUMBERS WITHIN BRACKETS DENOTE MAXIMUM VORTEX TYPES OBSERVED T"F = 56' 82 121 155 22 ^ I ^ g (2,3) (2,3) (2,3f2,3) Y E d x +> = x 1.8 = g (1,2) (1,2) (1,2) (1,2) Ez o / 3 1.4 = + 1,2) (1) ' 1,2 ) '(1,2) u. PROTOTYPE 1.0 b1) [(1) (1) (1) RANGE 5iku ~ 6 "*

• m *m I

i, I n n l l u l u u 1 n:l i! I i u l-ll ! n E 9f 2 3 4 5 6 789/ g 10 4 5 89 7 5 10 10 1 I REYNOLDS NUMBER, 1 l FIGURE 3 FROUDE NUMBER RATIO VERSUS REYNOLDS NUMBER SHOWING MAXIMUM OBSERVED VORTEX TYPES IN THE SUMP N J

CASE 1 PIPE 2 AT 9,500 GPM W.S. EL 602 FT 10 INCHES NOTE: NUMBERS WITHIN BRACKETS WITH GRAT!NG BLOCKAGE SCHEME 5 D6 NOTE VORTEX TYPES 12 ~~ ) i I l l i i i I I l I I i i 2.0 (3,4) (3,4) (3,4) (3,4) t 1.8 (2,3) (2,3) (2,3) (2,3) 9' E a: 1.6 (7,3) (2,3) (2,3) (2,3) i 1.4 (2,3) (2,3) (2,3) (2,3) ew 1.2 (1,2) (1,2) (1,2) (1,2) 1.0 + PROTOTYPE RANGE-*' (1,2) (1,2) (1,2) (1,21 l i 1 I I I I I I I I I I I I l 6 7 2 '3 4 5 6 7 8910 2 3 4 5 6 7 8 9 10 REYNOLDS NUMBER ud/v FIGURE 4 VORTEX TYPES OBSERVED DURING HIGH TEMPERATURE-HIGH VELOCITY TESTS (MODIFIED SUMP) M /

VORTEX TYPE 1 INCOHERENT SURFACE TWIRL 2 SURFACE DIMPLE: COHERENT SWIRL AT SURFACE 3 DYE CORE TO INTAKE; COHERENT SWlHL THROUGHOUT WATER COLUMN 4 VORTEX PULLING FLOATING 1 TRASH, BUT NOT AIR 4k l 5 VORTEX PULLING AIR ($ BUBBLES TO INTAKE 1 g......... FULL AIR CORE 6 b TO INTAKE FIGURE 5 VORTEX STRENGTH SCALE FOR INTAKE STUDY M

1 TABLE 1 EXPERIMENTAL OBSERVATIONS OF VORTEX ACTIVITY IN THE SUMP PHASE 3 TEST SERIES; SCHEME 4 WITH SCREEN BLOCKAGE WATER MODEL TEMPECATURE fl x 10-5 VORTIMETER READING k

• F INLET 2 REV/100 SEC.

REMARKS TEST NUMBER DATE 6-1 S/22/77 1.0 64 1.50 34, 35, 35 Surface swirls and dimples observed (Types 1-2) 6-2 6/22/77 1.4 65 2.10 49,49 Surface swirls and dimples observed { Types 1-2) 6-3 6/22/77 1.8 65 2.70 51, 51, 49 Surface swirls and dii.iples observed (Types 1-2) C-4 6/23/77 1.8 104 4.37 46,46 Surface swirls and dimples observed (Types 1-2) 6-5 6/23/77 1.4 107 3.40 45,45,46 Surface swirls and dimples observed (Types 1-2) 6-6 6/23/77 1.0 103 2.43 15,15,15 Surface swirls and dimples observed (Types 1-2) s \\ 6-7 6/24/77 1.0 140 3.38 25, 25, 24 Surface swirls and dimples observed (Tynes 1-2) 6-8 6/24/77 1.4 143 4.73 47,48,45 Surface swirls and dimples obsemed (Types 1-2) 6-9 6/24/77 1.8 139 5.74 49, 50, 48 Surface swirls and dimples observed (Types 1-2) .1 m

i MODEL STUDY OF REACTOR CONTAINMENT SUMP FLOW CHARACTERISTICS VIRGIL C. SUMMER NUCLEAR GENERATING ' STATION by James B. Nystrom Researcli Sponsored by Soutli Carolina Electric and Gas Company ALDEN RESEARCH LABORATORY 7 WORCESTER POLYTECHNIC INSTITUTE ~ n,4 July 1981 Y $\\ Q &O3 ObfC

MODEL STUDY OF REACTOR CONTAINMENT SUMP FLOW CHARACTERISTICS a VIRGIL C. SUMMER NUCLEAR GENERATING STATION by James B. Nystrom Research Sponsored by South Carolina Electric and Gas Company George E. Hecker, Director ALDEN RESEARCH LABORATORY WORCESTER POLYTECHNIC INSTITJTE h0LDEN, MASSACHUSETTS June 1981

nBSTRACT l A hydraulic model of the containment building sump for the Virgil C. Summer Nuclear Generating Station was constructed at a scale of 1:2.93. Residual heat removal pumps and reactor building spray pumps withdraw water from the sump after a postulated loss of coolant accident for re-injection into the core and building. To assure acceptable operation of the pumps, the model was tested for a wide range of possible approach flow distributions, floor grating blockage schemes, and screen blockage schemes. The tests were de-signed to assure that no air entraining vortices were formed, head losses across the screens and in the inlet were acceptable, and swirl in the pump suction pipes was acceptable. The maximum vortex activity noted was a surface dimple, which originated as a vortex shed from obstructions in the approach flow and was carried across the sump by the approach flow. Test results indicated that the maximum swirl angle was 9.5 degrees, while average swirl angle was about 3 degrees. For an RHR flowrate of 4500 gpm per line, loss measurements indicated an average pipe inlet loss of 0.37 ft and screen losses ranging from 0.05 ft for clean screens to 0.19 ft for the worst case of 50 percent blockage. The reactor building spray lines had flowraten of 3000 gpm, which reduced pipe inlet losses to 0.30 ft and screen losses to 0.12 ft maximum. O O i l \\

ii 2ABLE OF CONTENTS Page No. ABSTRACT i TABLE OF CONTENTS ii INTRODUCTION 1 PROTOTYPE DESCRIPTION 2 SIMILITUDE 4 Froude Scaling 6 Similarity of Vortex Motion 8 ARL Vortex Activity Projection Technique 10 Dynamic Similarity of Flow Through Screens 11 MODEL DESCRIPTION 13 INSTRUMENTATION AND OBSERVATION TECHNIQUES 15 Flow Measurement 15 Pressure Gradelines 15 Pipe Swirl 15 Vortex Activity 16 Observation of Flow Patterns lo TEST PROCEDURE 16 TEST RESULTS 17 Vortex Activity 17 Swirl Angle Measurements 20 Screen Head Loss 21 Inlet Losses 23 Flow from Mezzanine Floor Level 23

SUMMARY

24 REFERENCES 26 FIGURES PHOTOGRAPHS APPENDIX A APPENDIX B

INTRODUCTION l The reactor containment building of the Virgil C. Summer Generating Station is provided with both a residual heat removal (RHR) system designed to covl the shutdown reactor core and a reactor building (RB) spray system to cool a the containment building, both systems to operate only in the event of a Loss of Coolant Accident (LOCA). nitially, water for these systems is drawn from the refueling water storage tank. When the water level in this tank reaches a predetermined level, the residual heat removal system is switched from th_ njection mode to the.ccirculation mode. At this point, water is drawn fram the containment surp, which then contains water drain-i ed from the break and from the containment spray system. Flow approaching the sump is affected by the geometry of the flow path including various ap-purtenant structures and equipment. Water level, pump discharge, and water g temperature could vary durinj the recirculation mode, which lasts for an ex-tended period to provide sufficient heat removal. 44 The Alden Research Laboratory ( ARL) of Worcester Polytechnic Institute (WPI) was authorized by South Carolina Electric and Gas Company to construct and test a model of the Virgil C. Summer Nuclear Generating Station containment sump with the object of investigating free surface vortex formation, swirl in the inlet piping, inlet losses, or any otber flow conditions that could adversely affect the performance of the residual heat removal pumps and the reactor building spray pumps in the recirculation mode. Operating conditions involving a wide range of possible approach flow distributions, floor grat-ing blnckages, screen blockages (due to debris), and combinations tharcof were tested in th. model. I This report presents the findings of the study and includes a description ~ E of the prototype and the model, and summarizes conditions investigated, similitude considerations, test procedures, instrumentation, and interpre-tation of results.

_m i 2 J PROTOTYPE DESCRIETION 4 Both the RHR and the RB spray systems have a pair of pumps and sumps to main-tain independent redundant systems. An RHR sump and an RB spray sump are Jo-cated in each of two containment sumps which are located in the reactor build' c ing floor at elevation 412 ft between the bioshield wall and the containment I wall, as shown in Figure 1. The bioshield wall protects the sumps from di-E rect impingement of possible breakflow jets. Each containment sump contains j two sets of fine screens and two pump sumps from which the pump suction line exits. The shallow containment sumps are approximately trapezoidal in shape, I about 22 ft by 10 ft in plan, and are 4 ft deep. A 4 ft wide wall extending 3 ft high from the containment sump floor separates the RHR and RB spray pump sumps. A 6 inch high curb surrounds the containment sumps and stand-i j ard floor grates cover the sump area at floor elevation. Within the shallow containment sumps, two 4 ft square pump sumps descend i 8 ft to elevation 400 ft. The cross-section of the pump sumps, Figure 2, shows the pump suction lines exiting the sumps spproximately horizontally with initial center 2:r, elevation 402 ft. The RHR and RB spray pump suc-f tien lines have diameters of 14 and 12 itsches, respectively. Quasi-bell-f mouths consisting of standard reducers and flanges are used on both inlets. l The inlet piping to the pumps extends at a shallow slope about 56 ft to an isolation valve prior to the pump. I i i Two sets of vertical screens protect each pump sump from ingestion of debris into the pump systems. An outer screen, shown in Figure 3, is 6 ft square in l plan and extends from the bottom of the containment sump, elevation 408 ft, about two feet to elevation 410 ft. A solid cover extends from elevation 410 ft to elevation 412 ft, the grating level, and a horizontal solid plate i covers the outer screen. The outer screen has 1/2 inch mesh. The inner i screen, l'igure 2, has the dimensions of the pump sump, 4 f t square, and -has 1/4 inch mesh. The inner screen extends from elevation 410 ft to elevation I 911 ft 11 inches. A solid 1 slate is located from elevation 408 ft to eleva-tion 410 ft. The horizontal cover has access doors, and a ladder provides l access tc the pump sump. L

l 3 In the recirculation mode, after a postulated LOCA, water approaches the sumps laterally through the annulus created by the bioshield wall and the contain-ment wall. A secondary flow path is from the next level above the sump. The RB spray flow may collect on the upper floor, which has a 6 inch high curb surrounding all openings except a stairwell near the southwest sump. Assum-ing the floor drains are completely blocked, the stairwell provides the only flow path for the RB spray collected at that level. Minimum water level for recirculation mode is elevation 417 ft. Runout flow-rates for the RHR and RB spray pumps are 4500 gpm per line and 3000 gpm per line, respectively. A site visit was conducted to assure the interpretation and completeness of drawings in regard to the primary approach flow paths, possible secondary approach flow naths, and various equipment obstructing the flow paths. Various equipment, located at elevation 412 ft, with diameters greater than 3 inches were considered relevant in influencing flow conditions and these are shown in Figure 6. The main pieces of relevant equipment are the accumu-lator and its pipeline, an RHR pipe loop and valve, auxiliary piping over the southwest sump, a fan, lubrication lines, support columns, and instrument cabinets. Photographic documentation during the site visit allowed details ) to be chccked as model design and construction proceeded. Photographs 1 and 2 show the areas in the prototype surrounding the West and Southwest 7 sumps during construction, when temporary scaffolding was in place.

o 4 SIMILITUDE The study of dynamically similar fluid motions forms the basis for the de-sign of models and the interpretation of experimental data. The basic con-cept of dynamic similarity may be stated as the requirement that two systems with geometrically similar boundaries have geometrically similar flow patterns at corresponding instants of time (3). Thus, all individual forces acting on i corresponding fluid elements of mass must have the same ratios in the two sys-tems. The condition required for complete similitude may be developed from Newton's second law of motion: F. F +F +F +F (1) = i p g v t where inertia force, defined as mass, M, times the F. = 1 acceleration, a F pressure force connected with or resulting from = P the motion F gravitational force = viscous force F = force due to surface tension F = Additional forces may be relevant under spe-ial circumstances, such as fluid compression, magnetic or Corriolis forces, but these had no influence on this study and were, therefore, not considered in the following development. l Two systems which are geometrically similar are dynamically similar if both satisfy the dimensionless form of the equation of motion. Equation (1) can be mado dimensionless by dividing all the terms by F.. Rewriting each of 1 the forces of Equation (1) as:

1 5 P = net pressure x area = o Ap L g F = specific weight x volume = a2 Y L i P p Au/Ay x area = a puL F,, = shear stress x area = a3 3 F = surface tension x length = a oL 4 3 2 2 2 F = density x volume x acceleration = a P ""5 P" 5 where a ' "2, tc. = proportionality factors l 4 6 L = representative linear dimension p = net pressure y = specific weight p = dynamic viscosity o = surface tension p = density u = representative velocity a substituting the above terms in Equat'on (1) and making it dimensionless by dividing by the inertial force, F, we obtain "1 ~ 4 -- F~ + 5'- R- + W" =1 (2) E a5 "5 "5 "5 D

6 where = -u Inertia Force Euler number; E = Pressure Force /Ap/O u Inertia Force Froude number; F = = Gravity Force /gL uL Inertia Force Reynolds number; R = = j ,scous Force u Inertia Force = Weber number; W = Surface Tension Force 7p Since the proportionality factors, af, are the same in model and prototype, complete dynamic similarity is achieved if all the dimensionless groups, E, F, R, and W, have the same values in modcl and prototype. In practice, this is difficult to achieve. For example, to have the values of F and R the same requires either a 1:1 "model" or a fluid of very low kinematic viscosity in the reduced scale model. Hence, the accepted approach is to select the predominant forces and design the model according to the appropriate dimen-sionless group. The influence of the other forces would be secondary and are called scale effects (2, 3). Froude Scaling Models involving a free surface are constructed and operated using Froude similarity since the flow process is controlled by gravity and inertia forces. The Froude number, representing the ratio of inertia to gravita-tional force, u//gs (3) F = 4

l t 7 where u = average velocity in the pipe g = gravitational acceleration s = submergence, the representative linear dimension was. therefore, made equal in model and prototype. r F /F 1 (4) F = = r m p where m, p, and r denote model, prototype, and ratio between model and pro-totype, respectively. In modeling of an intake sump to study the formation of vortices, it is im-portant to select a reasonably large geometric scale to achieve large Rey-nolds numbers and to reproduce the curved flow pattern in the vicinity of the intake (4). At sufficiently high Reynolds number, an asymptotic beha-vior of energy loss coefficients with Reynolds number is usually observed (2). IIence, with F = 1, the basic Froudian scaling criterion, the Euler numbers, E, will be equal in model and prototype. This implies that flow patterns and loss coefficients are equal in model and prototype at suffi-ciently high Reynolds numbers. A geometric scale of L = L /L = 1/2.93 r m p was chosen for the model, where L refers to length. From Equations (3) and (4), using s =L, the velocity, discharge, and time scales were: = 1//2. 93 = 1. 71 (5) u =L r r Q =L u =L = 1/(2.93) (6) = r r r r 14.66 L = 1/ /2. 93 = 1. 71 (7) t =L r r

I 8 Similarity of Vortex Motion Fluid motions involving vortex formation in sumps of low head pump intakes have been studied by several investigators (1, 4, 5, 6). Viscous and surface tencion forces could influence the formation and strength of vortices (1, 5). The relative magnitude of these forces on the fluid in-ertia force is reflected in the Reynolds and Weber numbers, respectively, which are defined as: u d/V (8) R = I' (c/pr)1/2 where r = characteristic radius of vortex and d = intake diameter. It was important for this study to ascertain any deviations in similitude attribu-table to viscous and surface tension forces in the interpretation of model results. For large R and W, the effects of viscous and surface tension are

minimal, i.e.,

inertial forces predominate. Surface tension effects are negligible when r is large, which will be true for weak vortices where the free surface is essentially flat. Conversely, only strong air core vor-tices are subject to surface tension scale effects. Moreover, an investi-gation using liquids of the same viscosity but different surface tension coefficients (o = 4.9 x 10' lb/ft to 1.6 x 10~ lb/ft) showed practically no effect of surface tension forces on the vortex flow (1). The vortex severity, S, is therefore mainly a function of the Froude number, but could also be influenced by the Reynolds number. e S = S (F, R) (10) i l Anwar (4) has shown by principles of dinensional analysis that the dynamic similarity of fluid motion in an intake is governed by the dimensionless parameters given by 1 l 4Q u Q d and ue

9 where Q = discharge through the outlet U = tangential vclocity at a radius equal to 0 that of outlet pipe d = diameter of the outlet pipe Surface tension effects were neglectad in his analysis, being negligibic for weak vortices. The influence of viscoas effects was defined by the parameter Q/(v s), known as a radial Reynolds number, R p. For similarity between the dimensions of a vortex of strengths up to and in-cluding a narrow air-core type, it was shown that the influence of R becomes R 4 negligible if Q/(v s) was greater than 3 x 10 (4). As strong air-core type vortices, if p' resent in the model, would have to be eliminated by modified sump design, the nain concern for interpretation of prototype performance based on the model performance would be on the similarity of weaker vortices, such as surface dimples and dye-cores. For the prototype of the present study, the values of R for the operating temperature ranges of 70* and above, and p using the submergence to the floor grating, was greater than 1.1 x 10. In 4 the model, the value of R for the RHR sumps was 2.6 x 10 for Froude velo-p 4 city and 4.4 x 10 for prototype velocity both for water temperatures of 50*F. Thus, viscous forces would have only a secondary role in the present study. Dynamic similarity is obtained by equalizing the parameters 4Q/ , u/ gs, 0 and d/2s in model and prototype. A Froudian model would satisfy this condi-tion. To compensate for any possible excessive viscous energy dissipation and con-sequently less intense model vortex, various investigators have proposed in-creasing the model flow and, therefore, the approach and intake velocity, 3 since the submergence is maintained constant. Operating the model at the prototype inlet velocity (pipe velocity) is believed by some researchers to achieve the desired results (1). This is often referred to as Equal Velo-nity Rule, and is considered to give conservative predictions of prototype performance. The test procedure for the present study incorporated testing the model at prototype pipe velocities to achieve conservative predictions.

ls ,ARL Vortex Activity Projecti(n Technique ARL has conducted an extensive research program to assure that the conclu-sions regarding the effect of Reynolds number on vortex activity in the mo-del are valid for the prototype. A technique of extrapolating model vortex I activity to prototype Raynolds numbers (17) by using elevated model water temperatures and varying model flow vele ity (Froude ratio) has been appl 1-ed to several studies (7, 12, 18, 19, 29). Figure 4 illestrates the method used to investigate scale effecta e 4 predict vortex types in the prototype based on nodel results (7). The Ordinate, F, is the ratio of model to pro-totype Froude number, while the abscissa is the inlet pipe Reynolds number, R. The object 2Ve is to determine flow conditions at F 1 at prototype R = g from tests at lower than prototype R. Assume tue model to operate at flow less than Froude scaling (F < 1) at point a. By increasing the discharge in the model while keeping the same submergence and temperature. F and R are increased corresponding to a point, a, where a vortex of type N was g first observed. The model Reynolds numbee can also be changed by varying the kinematic viscosity with temperature changes, and similar tests per-formed to 1ccate b, an ther point on the locus of type N vortices. Extra-N polation of the line of constant vortex strength of type N can be made to a prototype Reynolds number at the proper Froude number (F = 1), point p

• H The locus could represent any expedient measure of vortex severity.

Any scale effects due to viscous forces would be evaluated and taken into ac-count by such a projection procedure. The high temperature-high flow tests were used in the similar fashion for projecting the inlet loss coefficients (from the pressure gradient measurementa) and the swirl severities (from vortimeter readings) over a wide ranga of Reynolds and Froude numbers. i Experience has shown that incoherent swirling flow is even Icss dependent on Reynolds number than a coherent vortex core. Eliminating the tendency for coherent vortices axiomatically removes possible sea.e effects. In reactor sumps, the design criteria eliminate the possibilicy 'f coherent vortex cores in an acceptable design. 1 l

11 Figure 5 shows the results of one recirculation sump model (19) which are typical of the other four studies conducted. As can be seen from the duca, waich are for the final design with vortex suppressor grids, there are no meastrable changes in vortex strength with Reynolds number. This is rea-sonable since the Reynolds numbers are all above the limiting value (1, 4), a previously described similitude requirement. Minor increases in vortex strength occur when the Froude ratio is increased. Other measurements, such as swirl in the inlet pipe, have also shown no measurable dependence on Rey-nolds number. This indicates that reduced scale model tests are a direct indication of prototype performance for weak vortices, particularly if ver-tex suppressors are part of tt; design, even at Froude scaled flow (i.e., F = 1). Tests at higher than Froude scaled flow are seen to give conser-vative results, i.e., somewhat stronger vortices than expected in the pro-totype. Since for this study the minimum Reynolds number is comparable to the minimum for the previous studies which indicated no increase in vortex activity for increasing Reynolds numbers at cc:~ tant Froude ratio, it is concluded that no scale effects will be present in the final design. Dynamic Similarity of Flow Through Screens addition to providing protection from debris, screens tend to suppress 7 non-uniformities of the approach flow. The aspects of flow through screens of concern in a model study are: (1) energy loss of fluid pass'ng through the screen; (2) modification of velocity profile and the def1cc Jn of streamlines at the screen; and (3) production of turbulence. As all these factors could affect vortex formation in a sump with approach flow directed y through screens, a proper modeling of screen parameters is important. i The loss of energy across the screen occurs at a rate proportional to the drop in pressure, and this loss dictates the effectiveness of the screen in altering velocity profiles. The pressure drop across the screen is analogous to the drag induced by a row of cylinders in a flow field and could be expressed in terms of a pressure-drop coefficient K (or alternately a drag coefficient), defined as (8), L

12 (ll) K = 2 1/2 p U U /29 where Ap = drop in pressure across the screen U = mean velocity of approach flow p = density of the fluid all = head loss across the screen g = acceleration due to gravity From the available literature on the tonic (8, 9, 10), it may be seen that f(R, S', Pattern) (12) K = g where screen Reynolds number, U d /v, d being the R = g wire diameter of the screen s' = solio.cy ratio, equal to the ratio of closed l area to total area of screen Pattern = geometry of the wire screen l If the solidity ratio and the wire mesh pattern are the same ia the model cnd i prrtotype screens, the corresponding values of K would only be a function of the screen Reynolds number. This is analogous to the coefficient of drag in the case of the circular cylinder. It is known that K becomes practically in- } dcpendent of R at values of R greater than about 1000 (8, 11). However, for g model; with low approach flow velocity and with fine wire screens, it is nec-essary to c3certain the influence of R on K for both the model and prototype s i l screens before selecting screens for the model which are to scale changes in volteity distribution.

l l I 13 Velocity modification equations relating the upstream velocity profile and downstream velocity profile have been derived based on different theories (8). Most of these indicate a linear relationship between upstream velocity profile and downstream velocity profile, shape and solidity ratio of screen, 4 and value cf F. If the wire shape and solidity ratios are the same in the model and prototype screc.ts, it is possible to select a suitabic wire dia-meter to keep the values of K approximately the same for the model and pro-totype screens at the corresponding Reynolds number ranges. Identical velo-city modifications'would be produced by the respective screens if the loss coefficients were identical. The pressure loss coefTicient to Reynolds number relationship of fine screens have been investigated at ARL (12). Based on the similarity of pressure loss and velocity modifications, an appropriate model screen was chosen, which had a loss coefficient within 10% of the predicted prototype loss coefficient. This was considered sufficient since actual losses and, therefore, velocity profile modifications, were small (about 0.05 f t) and screen blockages cause changes in velocity distributions far outweighing changes aue to screen. MODEL DESCRIPTION The model was constructed to a geometric scale of 1:2.93 with boundaries, as indicated in Figure 1. Model boundaries were chosen at locations where flow l pattern control in the prototype would be sufficiently removed frcm the sump areas to avoid boundary effects, especially once screen blockage is censider-ed. Screen blockage has consistently generated the most severe vortices and swirl in the numerous past ECCS sump studies at ARL. The model was located in ? an existing elevated tank to provide access to observe flow patterns in the pump sumps. Photograph 3 is an overall view of the completed model. Inflow war provided from a sump beneath the model by a vertical pump, and the water level in the model was controlled by an adjustable weir. Flow straighteners at the model boundaries provided a uniform initial velocity distribution with relatively low turbulence levels. Portions of the prototype structure with outside dimensions greater than 3 inches, such as pipes, columns, conduit supports, and a stairway, in the immediate vicinity of the sump and below the water surface were modeled to the geometric scale, as shown in Figure 6.

14 The model was constructed using a combination of wood, steel, and clear acrylic, which allowed observation of flow patterns. One clear acrylic sump is shown in Photograph 4 with the RHR spray suction line. Horizontal section pipes were modeled for about 16 pipe diameters, had access ports for vortimeter installation, and had five sets of piezometers for pressure gradeline measurement. ASME standard orifice flowmeters were provided to 4 measure flow in each suction line. The two containment sumps and nearby details are shown in detail in Photo-graphs 5 and 6. These photographs may be compared to the similar perspec-tives of the prototype shown in Photographs 1 and 2. Clear PMMA plastic was used for sump covers to allow observation of flow patterns between the screens. Narrow slots were provided in the cover plates to allow screen blockages to be changed without model disassembly. The model screens were chosen on the basis of percent open area. The model outer screens were 3/16 inch mesh with 0.063 inch wire diameter and the inner screens were 1/8 inch mesh with 0.041 inch wire diameter. Model screen Reynolds num-bers were greater than 100, which resulted in loss coefficients a few per-cent greater than the predicted values for the prototype screens. The floor grating ased in the model was prototype dimensions. The flow from the RB cpray from the above floor was modeled by a tank with an opening simulating the stairway. Flow was supplied by a 4 inch pipe with orifice meter for flow control. t

15 INSTRUMENTATION AND OBSERVATION TECHNIQUES Flow Measurement Flowrates were measured by ASME standard orifice meters and coefficients us-ing air-water manometers for differential pressure measurement. 3 Pressure Gradelines Each pressure gradeline in the suction line was measur?d by a pair of piezo-meters at five locations in each pipe using air-water manometers with the sump water level as reference pressure. The pressure gradeline was extra-polated to the entrance by a linear least equares (linear regression) curve fit of the pressure measurements. The area average velocity was used to calculate the pipe velocity head, which was added to the extrapolated pres-sure gradeline. The total head within the sump was determined from a pres-sure measurement and the velocity head at that location. The pipe total head was subtracted from the sump total head to determine the inlet loss. An entrance loss coefficient was calculated by: 1 bH$ K= (13) 2 vmean 2g where K = loss coefficient AH = inlet head loss, ft g Pipe Swirl Average swirl in the suction pipes was measured by cross vane vortimeters. Studies at ARL (22) have shown that a vortimeter with vane diameter 75% that of the pipe diameter best approximates the solid body rotation of the flow. The rate of rotation of the vortimeter was determined by counting the number of blades passing a fixed point in one minute.

a ) 16 J An average swirl angle was defined as the arctangent of the maximum Langen-tial velocity divided by the axial velocity. The maximum tangential velo-city of the vortimeter is the circumferential path travelled by blade tip per unit time, w D N, and the average swirl angle is defined by: 0 = arctan (y D N) (14) W mean where N = revolutions per second D = rotameter diameter, ft V = mean axial velocity Vortex Activity Vortex activity was recorded by observing vortex strength on a scale from 1 l l to 6 (see Figure 7). Vortex strength was identified by using dye injection ar.d addition of " trash" consisting of a slightly buoyant ball of paper. Observation of Flow Patterns Visual aids, such as dye, were used to observe flow patterns. Photographic documentation was taken whenever appropriate. TEST PROCEDURE Tests were conducted at the normal laboratory water temperature. The model was filled to an appropriate level, and all piezoneter and manometer lines were purged of air and zero flow differentials checked. The required flow-4 rates were then set and the water level allowed to stabilize. The water i icvel was checked and adjustments made if required and flourates were re-cP.ccked and re-adjusted, if necessary. A 15 minute minimum settling time was allowed prior t, initiation of the data recording. Fifteen minutes of vortex observations were recorded and the seguired physical parameters, such as depth, manometer deflections, and vortimeter readings, were recorded. 1 Entrance losses were determined for the suction lines without vortimeters.

17 TEST RESULTS Six floor grating blockages, see Figures 8 through 10, and 8 screen block-ages, see Figures 11 through 14, were used in the test program. The ap-proach flow distribution was varied by blocking 50% of the flow straightener area en one side at a time. Various combinations of floor grating blockage, screen blockage, and approach flow distribution were, tested. The floor grat-ing was removed to determine whether it had an effect on the vortex activity. Vortimeters were located in the west containment sump inlet lines and the inlet pipe pressure gradelines were measured in the southwest containment sump lines. Screen losses were measured for all four inlet lines. Vortex Activity Table 1 summarizes vortex activity for the 76 tests conducted. Maximum ac-tivity was a surface dimple, type 2, indicating some swirl at the surface. Dye injection indicated the dimple to be a surface phenomenon and no coher-ent core was detectable, even at the surface. In only one case of 33 using Froude scale velocity was a surface dimple noted. For prototype velocity, this increased to 23 cases of a total of 43. In all cases, the surface dimple was unstable and was carried across the sump by the general circu-lation patterns and quickly dissipated. Dimples over the west sump were caused by vortices shed from the support columns, which translated across the sump. Over the southwest sump, the vortices were generated by the col'imns and piping in the area. Due ta the approach path and local geo-metry, a general clockwise circulation developed over the southwest con-tainment sump. This circulation became somewhat more pronounced when the south flow straightener was 50% blocked. Removal of the floor grating did not increase vortex acti?ity. Combined floor grating and screen blockages caused the greater vortex activity. Approach flow distribution had little effect on vortex activity.

18 l TABLE 1 Vortex Activity I Velocity Scale Test Number Froude Prototype Ph Sueu F P Grating Blockage E West Southwest West Southwest 1 6 0 1 1 1 1 2 1 1 1 3 1 1 1 4 2 1 1 5 2 1 1 7 1 0 1 1 8 2 O 1 1 9 3 0 1 2 10 4 0 1 2 11 5 0 1 1 12 6 0 1 1 13 1 0 1 2 14 2 0 1 1 15 3 0 1 2 16 4 0 1 1 17 5 0 1 1 52 18 6 0 1 1 1 1 19 7 0 1 1 20 8 0 1 1 3E 21 1 8 0 1 1 1 1 37 22 1 8 1 1 1 1 2 J 38 23 1 8 2 1 1 1 1 33 24 4 8 0 1 1 2 2 l 34 25 4 8 1 1 1 2 2 35 26 4 8 2 1 1 2 2 32 27 6 8 0 1 1 1 1 31 28 6 8 1 1 1 1 2 30 29 6 8 2 1 1 1 1 56 62 1 2 0 1 1 1 1 57 63 4 2 0 1 1 2 2 58 64 6 2 0 1 1 1 1 1 59 65 1 4 0 1 1 2 1 40 1 4 2 1 1 41 1 4 1 1 2 j 42 66 4 4 0 1 1 2 2 43 4 4 2 1 1 44 4 4 1 1 1 45 67 6 4 0 1 1 1 2 46 6 1 1 1 47 6 4 2 1 1 l 53 68 1 6 0 1 1 1 2

19 TAELE 1 (continued) Velocity Scale Test Number Floor Screen F P Grating Blockage F West Southwest West Southwest 69 2 5 0 1 2 70 3 6 0 2 2 54 71 4 6 0 1 1 2 2 72 5 6 0 1 2 55 73 6 6 0 1 1 2 1 48 59 None 4 0 1 1 2 49 61, None 4 1 1 1 1 1 50 60 None 4 2 1 i l 1 51 77 None 6 0 1 1 1 2 78 None 6 0 1 2 79 None 2 O 1 1 See Figures 8 through 14 for floor grating and screen blockage configurations. For Approach Flow Distribution, (FF) indicates no flow straightener blockage, 1 indicates west blocked 50%, and 2 indicates south blocked 50%. i

20 Swirl Angle Measurements Rotameter rotation rates were used to calculate swirl angles by Equation (14). Rotameters were located in the RHR and RBS lines in the west containment sump for all tests. Rotameter rotation rates were unsteady for sevcral tests with reversing direction during the one minute observation period. In these cases, the greater rotation rate was used to calculate swirl angle and, therefore, in some tests opposite rotation directions appear for Frorde and prototype velo-city scale tests. Appen/.ix B lists the calculated swirl angles, blockage configurations, and ap-proach flow distributions for all tests conducted. Approach flow distribution had little effect on swirl angle when floor grating and screen blockage were combined. Floor grating blockage had little effect when screen blockage were in place, with the exception of the horizontal screen blockage tnich imparted no swirl. Table 3 summarizes the swirl angles averaged for each screen block-age, since screen blockage was the dominant factor in most cases. TABLE 3 Average Swirl Angles, Degrees Screen Blockage,* RHR Inlets RB Spray Inlet None 3.2 2.4 None, all floor grating blockages 1.8 4.3 All screen only 4.7 3.6 2 4.8 3.5 4 4.8 5.5 6 1.9 1.8 8 2.8 4.1 4 without floor grating 5.7 5.0

• See Figures 8 through 14 for screen blockage configurations 4

l

21 r The swirl angles for the RHR and RBS inlets were 3.2 and 2.4 degrees, re-spectively, for cican screens. Screen blockage configuration 4 created the greatest average swirl with values of 4.8 and 5.5 degrees for the RHR and RBS inlets, respectively. With the floor grating removed, the swirl angles were similar, 5.7 and 5.0 degrees. The maximum swirl angle measured was 9.5 degrees for acreen blockage confi-guration 3. Average swirl angles for all tests were 3.6 degrees for the RilR inlet and 3.9 degrees for the RBS inlet. Since about 48 diameters of straight pipe exists prior to any fittings in the inlet lines, the swirl angle will decay considerably. Using a conser-vative estimate for the swirl decay parameter, S = 0.02, from available literature (27, 28), the swirl remaining at the end of the straight pipe vill be about 40 percent of the initial swirl. This results in a maxi-mum swirl angle of 3.8 degrees and average swirl angles of less than 2 degrees. S,irl angles of similar magnitudes may result from single bends (24) and swirl angles resulting from combined bends could be about three times greater (25, 26). Therefore, the measured swirl angles are not con-sidered excessive. Screen Head Loss The head losses due to the floor grating and two sets of screens were mua-sured for all four pump inlets for all tests. The velocity head of the ap-proach flow was neglected such that the measured water level outside the screens was assumed to be the initial total head. Static head was measured in each sump with two piezometcrs at elevation 404 ft. The velocity head in the samp, calculated using the area average velocity, was added to the sta-tic head to determine che total head. Screen. Loss was determined by subtract-ing the sump total head from the measured water level. The measured head loss was corrected to the runout flowrate for each pump and converted to prototype dimensions.

22 As a check, the RI:R screen losses were calculated (8). For clean screens, the calculated loss was less than 0.01 ft and with the screens 50% blocked, the calculated loss was less than 0.03 ft. These values are basad on ap-proach flow r.ormal to the screen with a relatively uniform approach velo-city distribution. Actual losses will be censiderably greater due to the complicated approach flow path which has several changes in direction due to the orientation of the floor grating and the vertical offset of the screens. With blockage, horizontal screen offsets could also be includcd. The magnitudes of the screen losses are small in relation to experimental uncertainty and, therefore, averages will be used to illustrate losses. Appendix A lists the measured losses for all tests for the four pump inlets. Table 2 summarizes the average loss measurements. Floor grating blockage and non-uniform approach flow distribution did not cause losses to vary greatly. Screen blockage could cause losses to vary due to the flow path variations. Therefore, losses for a given screen blockage are averaged over floor grating blockage and approach flow distribution configurations. TABLE 2 Screen Loss Summary Loss - Feet Screen Blockage

• RHR RB Spray None 0.06 0.05 None, all floor grating blockages 0.06 0.04 2

0.10 0.06 d 4 0.09 0.05 6 0.08 0.05 8 0.10 0.06 4 without floor grating 0.09 0.05

• See Figures 11 through 14 for screen blockage configurations.

23 l Clean screen losses averaged 0.06 f t for the RHR inlet and 0.05 ft for the Rb spray inlet. Tests with floor grating blockages resulted in losses of 0.06 f t and 0.04 f t for the RHR inlet and the RB spray inlet, showing lit-tle change. The four screen blockage configurations with sufficient data to average resulted in an increased loss of about 0.03 ft for the RHR in-let and about 0.1 ft for the RBS inlet. For the RHR inlet, the increase is about what was calculated for the blocked screen losses. The losses due to the flow path are significantly higher than the screen losses. In the case of the RBS inlet, the losses should be about 45 percent of the RHR in-let losses, due to the decreased flow. The measured losses were somewhat high in comparitan to the RHR inlet loss, but the increase due to screen blockage was near what would be calculated. Inlet Losses Inlet losses were measured and an inlet loss coefficient was calculated by Equation (13). The inlet loss c 3 efficient was essentially equal for the RHR and RBS inlets and had a value of 0.27. This compares well to published data (23) and data from previous studies (29). The head losses were 0.37 f t and 0.30 f t for the RHR inlet and the RBS inlet, respectively. Flow from Mez:.anine Floor Level Tests were co d_rted with flow from the mezzarine floor stairwell opening to determine whether any adverse conditions, such as bubble formation and subsequent air entrainment, might exist. A series of flowrates were used with the maximum flowrate of 1300 gpm corresponding to a depth of about six inches on the mezzanine floor. Since the mezzanine floor has a six inch curb around all openings, significantly greater depth would be impos-sible, The flow pattern from the stairwell is shown in Photograph 7 for the maximum flowrate. The first flip a of stairs was modeled and they de-flected the majority of the flow vertically downward. The stairwell en-trance is between the fan duct wall and the bioshield wall, therefore, the majority of the flow fell into the fan duct. The flow that was able to travel horizontally enough to impact in the front of the fan duct wall was

24 distributed over a large area. The remainder of the stairway, not modeled in these tests, would further dissipate the energy of the falling water and further spread the impact area. No air bubbles penetrated the water surface sufficiently to be detectable in the sump area. As the flowrate and depth on the mezzanine floor dei reased to 990 gpm, the initial horizontal velocity decreased and less flow impacted outside the fan duct. At the 4 inch depth corresponding to a flowrate of 680 gpm, only a small amount of flow impacted outside the fan duct, as shown in Figure 9. Flow from the mezzanine floor level did not cause any adverse effects.

SUMMARY

l A 1:2.9 scale model of the containment building sump for the Virgil C. Sum-me. Nuclear Station was constructed and tested. In the recirculation mode, residual heat removal and reactor building spray pumps withdraw water from two containment sumps after a postulated loss of coolant accident. A hori-zontal floor grating, 1/2 and 1/4 inch mesh vertical screens, surround each pump sump to assure no debris is entrained into the pumping systems. The debris could block both the floor grating and screens, thereby producing ad-verse flow patterns in the sump. A wide range of possible approach flow dis-tributions, floor grating blockages, and screen blockages, and combinations thoreof were tested to simulate possible undesirable flow patterns which could result in poor pump performance during the recirculation mode. The model was operated with both Froude scale velocity and prototype velocity. Vortex a ti uty was observed and recorded. Head losses due to the floor grating, screens, and pump inlet and the flow rotation in the suction pipe were also measured. A surface dimple was the greatest vortex activity observed. For Froude scale velocity, only one test in 33 h. a surface dimple. Increasing the velocity to prototype velocity increased vortex activity such that in about one-half of the 43 cases a surface dimple formed. The surface dimples noted were pro-I duced froa vortices shed from obstructions in the flow, such as support col-umns and, therefore, traveled with the general flow patterns. No coherent dye core formed in conjunction with the surface dimple. Tests without the floor grating showed no increase in vortex activity.

25 Averagc swirl angle in the suction pipe =, was less than 4 degrees and maximum value measured was 9.5 degrees. Screen losses varied from about 0.05 ft for a clean screen to 0.19 ft for the worst case of 50 percent screen blockage. The pipe inlet head loss averaged about 0.3 times the inlet pipe velocity head. 9 L 4 =

26 REFERENCES 1.

Daggett, L.L.,

and Keulegan, G.H., " Similitude Conditions in Free Sur-face Vortex Formations," Journal of Hydraulics Division, ASCE, Vol.100, pp. 1565-1581, November 1974. 2.

Daily, J.W.,

and Harleman, D.R.F., Fluid Dynamics, Addison-Wesley Publishing Company, 1965. 3.

Rouse, H., Handbook of Hydraulics, John Wiley & Sons, 1950.

4. Anwar, H.O.,.Weller, J.A., and Amphlett, M.B., " Similarity of Free-Vortex at Horizontal Intake," Journal of Hydraulic Research, IAHR 36, No. 2, 1978. 5. Hattersley, R.T., " Hydraulic Design of Pump Intakes," Journal of the Hydraulics Division,'ASCE, pp. 233-249, March 1965. 6.

Reddy, Y.R., and Pickford, J.,

" Vortex Suppression in Stilling Pond overflow," Journal of Hydraulics Division, ASCE, pp. 1685-1697, November 1974. i 7. Durgin, W.W.,

Neale, L.C.,

and Churchill, R.L., " Hydrodynamics of Vortex Suppression in the Reactor Building Sump Decay Heat Removal System," ARL Report No. 46-77/M202FF, February'1977. 8.

Baines, W.D.,

and Peterson, E.G., "An Investigation of Flow Through Screens," Trans. ASME, pp. 467-477, July 1951, 9.

Papworth, M.,

"The Effect of Screens on Flow Characteristics," British Hydromechanics Research Association, Report TN1198, November 1972. 10. Weighardt, K.E.G., "On the Resistance of Screens," The Aeronautical Quarterly, Vol. IV, February 1953.

27 11. Tennessee Valley Authority, " Flow Through Screens," Report No. 87-8, May 1976. 12. Padmanabhan, M., " Hydraulic Model Studies of the Reactor Containment Building Sump, North Anna Nuclear Power Station - Unit 1," ARL Report No. 123-77/M250CF, July 1977. i 13.

Govier, G.W.,

and Aziz, K., "The Flow of Complex Mixtures in Pipes," Van Nostrand Reinhold, 1972. 14. Chainshvili, A.G., " Air Entrainment and Vertical Downward Motion of Acrated Flows," IAHR, 8th Congress, Montreal, Canada. 15.

Muakami, M.,

Suehiro, H.,

Isaji, T.,

and Kajita, J., " Flow Entrained Air in Centrifugal Pumps," 13th Congress, IAHR, Japan, August 31 - September 5, 1969. ( 16. Final Safety Analysis Report, J.M. Farley Nuclear Plant, Appendix 60, Nuclear Regulatory Commission, 1977. 17. Durgin, W.W., and Hecker, G.E., "The Modeling of Vorticos at Intake Structures,".cir.t Symposium of Design and Operation of Fluid Machinery, Colorado State University, June 1978. 1 l 18. Padmanabhan, M., " Hydraulic Model Investigation of Vortexing and Swirl Within a Reactor Containment Recirculation Samp," Donald C. Cook Nuclear Power Station, ARL Report No. 108-78/M178FF. t 19. Paimanabhan, M., " Assessment of Flow Characteristics Within a Reactor Containment Recirculation Sump Using a Scale Model," McGuire Nuclear Power Station,. ARL Report No. 29 78/M208JF. 20. Padmanabhan, M., " Selection and Scaling of HorJzontal Gratinas for Vortex Suppression," ARL Report No. 68-78, July 1978.

28 21. Padmanabhan, M., and Vigander, S., " Pressure Drop Due to Flow Through Fine Mesh Screens," Journal of the Hydraulics Division, ASCE, HY8, August 1978. 22. Du rgin, W.W., and Lee, H. L., "The Performance of Cross-Vane Swirl Meters," ASME Winter Annual Meeting, 1980. 23. Miller, D.S., Internal Flow Systems, BHRA Fluid Engineering, 1978. 24. Unpublished ARL Experimental Results. 25. Padmanabhan, M., " Investigation of Flow Distribution and Swirl Due to a Combined Pipe Bend," McGuire Nuclear Power Station, ARL Report No. 12-79/M208MF, December 1978. 26.

Nystrom, J.B., "The Effects of Combined Bends on the Velocity Dis-tribution and Swirl at the Inlet to a Pump," ARL Report No.

122-80/M105AF. August 1980. 27.

Baker, D.W.,

and Sayre, C.L., " Decay of Swirling Turbulent Flow of Incompressib2 e Fluids an Long Pipes," Flo,i, Its Measurement and Con-trol in Science and Industry, 1974. 28.

Janik, C.R.,

and Padmanabhan, M., "The Effect of Swirling Flow on Pipe Friction Losses," ARL Report No. 26-81/M296KF, February 1981, t 29. Padmanabhan, M., " Investigation of Vortexing and Swirl Within a Con- [ tainment Recirculation Sump Using a Hydraulic Model," hBL Report No. 25 81/M296HF, February 1981. I i f 1

N FIGURES

MODE L BOUNDARY ey / \\ l /,g g f s. / s,s / 'q' / / l rassas (+,,3 / Ret,E,1ANx / I J F m-g,, / I L I 3, = I e + i LF ~\\ \\ s {f"~~~~ ' ! '\\ f % i f ~ l 270" - T'- 7! I b ~ g n ) H REACTOR l ~ 63' s 's r s \\ 'N h \\ ' %, h ' \\ $__.. \\ N i \\ s \\ 'N \\ N 180' MODEL BOUNDARY

• = = = =, _ _ _ _, -

FIGURE 1 REACTOR BUILDING LAYOUT AT EL 412 FEET N

l 1/4" OPENING SCREEN (4 SIDES) OUTER SCREEN OOR GRAMG 0.120" WIRE Di AMETER SEE FIGURE 3 13/4 X 3/16 1 3/16 O.C. 4' 0" L 3 x 3 x 3/8 TYP. E L 412'-0" p - E ". D C,*, .4.,7 ~* ' " i " s i

Kg

,TT.(Wisiiiiaisi gjg. fo:..j [tg -e l l g e id 3EI gj: j l l '- 1 1 " f, p '. V i i E l l l l l [ 2 '-0 " o 'A l k al .p-EL 408'-0", o i 4>r .4' i u .h,. . :..,.;j ;.j - .i.,-,/ * ;p -, l. :,. -a a I o.,.- ACCESS l : i '.j T .. g, LAD D E R --+ j i l u-l l 1/2" LINER PLATE l i I :.. I V _ :.* :. i I. :,q., i* 3 :...- '. -[ l l:{o'. NOTE: , ' f. M,'l.; l :7. SUCTIO.J PwE OUTER SCREEN NOT 14" RHR L!NES p. EL l ;,g. SHOWN FOR CLARITY. i 12" RB SPRAY LINES 402'-O " :4 SEE FIGURE 3 FOR j OUTER SJREEN DETAIL ,} SLOPE VARIABLE l.e a, . p,! U l.D.. i i . o. D l l E L 401'-1" _ g.: E L 400*-0" k h k h S.- ' Y. 6 p

• O

. ~0' FIGURE 2 SECTION OF SUCTION PIPE SUMPS AND INNER SCREENS \\i l I

6 '-0 " ~ 5 ~ ~5: x i 5 -[_--Z_~ i ~z=;'5r '~ ^ t '

z_z__z_ z_.r_ _Z__. __ _ __ Z_Z_ Z Ji,

I 7 _ it i l.;l:0 [I Il I! ! I g I i il w_ _- l l l l i I i 11 "pg N-II l i a r $1 i 't I> I I I I l i i + i _i i _> gi l f k/ I 6 '-0 " il l fl> l l ACCESS < )l l ii l111 l

ll l

IHATCH

_____.4 Nl l l j ( j

lllllj, i

-'11 l i l' l j 4 II I A I A l ilp=l y _ =_=_=_=_=_=_ _=dpr:4 =_ _ =al:ll I lii 3 i ; -t_Ac;;=::::::::= _ _:=_ n! 1 PLAN VIEW 3/8" CHECKED PLATE T EL 412'-0" ~ l m3 sa ' ' Es um 1 2'-0-3/8" i 1/4" PLATE i (4 SIDES) I 4::tg[:t a N. l'-11 5/8" 1/2" OPENING 'b tq-SCREEN 0.162" EEtt ~E' A_I

EE i

EL 408'-0" WIRE DIAMETER y . ;v,..: y. 4 ...- y - l...g ? - L 3 x 3 x 3/8 TYP. I U ' *: ' SECTION A-A O*I . :?. o FIGURE 3 OUTER SCREEN DETAILS S l

3 I i l i i" l LOCUS. TYPE N VCRTEX j F, l T I T F j 2 \\~ / l 2 3 Fl j T i m a 4 [p s !! p l l i U d4 N l l r.__.__.. l I l Cu I U4 \\ l a3 4 i C b !' \\ ' 3 N t 1 ~ P a2j q N I al PROTOTYPE RANGE i IR I FIGURE 4 ARL VORTEX ACTIVITY EXTt:APOLATION TECHNIQUE M =

l l' jll l N 9 1 S E E C P N TY F t NT E E R P 5 SX 3 Y 6 E N F NEE 2 T O 1 I RT 3 E O - R I HPR 7 T T TE O L O F M LA WIR VI E R 0' O 2 P O S S T M E 7 R P E C F REU S A A EKM A F T R B CI C R MA X E UR A U T X NBM S O E R N E Y E T T T O A I F 3, N W V ) I 2 T 0 6 C ( 5 f A 1 j ) 3, X Fj2 i ( / E 5 / v RT 2 3j/ ) 1 2, / O ) d 3, a V ( l2 F / ( R 0 ) E R I 0 ) 2 J 3, ( M A B 1 ((2 2 U ( R N O / J S F ( D / L S O T N L Y U ) 2 E )3, R SE F 2 R ( 8 / 4 L E J ) 2 D O ( E M ) U j j LQ AIN C H IPC J ( / ) y 2 j YE g TT 5 E R UG I F / 5b ) j l ( 5 5 0 1 m6 dP Su l

FLOW ONLET STRAIGHTENER FLOW STRAIGHTENER i ACCUMULATOR r-VENT!LATlON LINE 9 FAN g I 4, Nff D3: COLUMN (TYP.) ) N-k \\ ~"] / LUBRICATION (- LINES sy WEST-3 SOUTHWEST SUMP ~- SUMP FIGURE 6 PLAN VIEW MODEL WITH ' DETAILS I M I

VORTEX TYPE 1 INCOHERENT SURFACE SWIRL 2 SURFACE DIMPLE: COHERENT SWlRL AT SURFACE 3 S DYE CORE TO INTAKE; ( COHERENT SWlRL THROUGHOUT WATER COLUMN 4 VORTEX PULLING FLOATING TRASH, BUT NOT AIR U 5 VORTEX PULLING AIR h BUBBLES TO INTAKE ( g......., 6 FULL AIR CORE 5 TO INTAKE FIGURE 7 CLASSIFICATION OF FREE SURFACE VORTICES M

l f '4 3vg \\ 'N f; g *i % A " igf,;,b,' [ N-g,-% j Un3 r,i' WEST S'JMP SW SUMP FLOOR GRATING BLOCKAGE 1 i N 6 \\. h ( p i xh}'d, g~

c. n E

WEST SUMP SW SUMP FLOOR GRATING BLOCKAGE 2 FIGURE 8 FLOOR GRATING BLOCKAGE SCHEMES M

t ~ .g. A'\\ z \\ 4 / / L e l s) f;& w% i w"a , $ 4;,9 w};FQ>e ,R Q T e ..s WEST SUMP E SW SUMP FLOOR GRATING BLOCKAGE 3 $i4 %s / d a / y g, WEST SUMP E SW SUMP FLOOR GRATING BLOCKAGE 4 FIGURE 9 FLOOR GRATING BLOCKAGE SCHEMES M

s ; v, #= k ' /9 A f J 7 / Q 'N [ ,I p $4 /. CT I WEST SUMP E SW SUMP j-FLOOR GRATING BLOCKAGE 5 1 d\\ x 'k / h xxx /f\\ M '%b j g,; - g // ~ 4($ s k ((

(

'\\._ ~ WEST SUMP rg SW SUMP l FLOOR GRATING BLOCKAGE 6 l FIGURE 10 FLOOR GRATING BLOCKAGE SCHEMES M

a -}$5/}'#;'fN'f~#EI2/rfdOIN'/1MflC(//.*Ihm fI-197dE.M?Ed*Mk A $mw HFr2%;,Q-We;r r-mt?nwA Wi5?!M?bWJRfiaf3MUt $!!96W I SCREEN PLOCKAGE 1 NOTE: COMPLETELY BLOCKED VERTICALLY SCREEN ORIEN1 ATION AS IN clGURE y.W~NewmMr.im.2cm.,MwO.vimWF T.!t2rrs.l J. T 21 9 i C 1 8 n s $I, k, ,i r p 3 t 2 ? 6 u W y 0

C d

9 k 5 S bi R 't

7 a$

1 >WGNMautav1M&iMauw1l ',> y; i.I h 1 i a SCREEN BLOCKAGE 2 FIGURE 11 INNER AND OUTER SCREEN BLOCKAGE SCHEMES M

Wi%*b.4?JCit4_I.MM fin'Al i s n 2 s N, ,e ?- N 0 N P h h h.1 +! kWJMMtWi2/W i nxiv.cusremswc.eA BLOCKAGE 4 l p kawsa'EW'5fp5*.eles.T/1 ek ty .c ?% C ? >s s \\ u Yy 'f

• i 9

4 r

2 5'r Kh'MJ. iiiB #sT.h"1 n u w m.a.w w w,,n BLOCKAGE 3 FIGURE 12 INNER AND OUTER SCREEN BLOCKAGE SCHEMES M

zur.evawne.v.mmemwmva.- N, 9 r.1 y t a ej s M4 y y e T -[4 ?.e ? -1h f; u i ti 4Y@ 4 BLOCKAGE 5 W4acesswFMWMn a k ljYweALU rssN y e A

.w 4

4 a C o (* g b i P f;' ?$.

s i8,

1

'h / d' a N EMEdNMM*I4. 34 4r II IndtTa.1ffT4' 6 9 9 2 n :6 F N:il BLOCKAGE 6 FIGURE 13 INNER AND OUTER SCREEN BLOCKAGE SCHEMES b3 l l

yxw+:+xw<6xiww+:wwiwwwi+:+:mx+x x 3-F J d ( g m h BLOCKAGE 7 NOTE: COMPLETELY BLOCKED HORIZONTALLY. 5%N%5%X%XnnMXnNM%%% M

//,

f 6 E h l t / I l 1 l BLOCKAGE 8 l FIGURE 14 INNER AND OUTER SCREEN BLOCKAGE SCHEMES M

1 PHOTOGRAPHS 2

'TM >- -_ JP:>ed%e e!:;jf7W \\ [' W A' [1.;. \\' . _ N D, s' j q F2 I l-j y 4 I + 3 y '; 3 (f of' l ..w ~ _l I f g ' 5._ lt)}l} } @J, I N ' $, ) . nib; 1 a ~q p; a  !? . _f, ~ u tu y. flill[3 ogf g ga gyyggy ' N~, !} E h ' ' ", ~j iky: 3 N I r 3_,.

4. U n

',. J m "2 g.7,gj.;i ? :. ji: [,. ~;gggg. t . '9.55.Et .[# y, Photograph I West Sump - Prototype [0

li i 5Rjfy; p s,- MN} b ll r. ~!- i .g ? .a G ey-y .: v;. l l ~',- "O

==4 m m d h u. 7- - st7"MMR x "[ -; ;. e-o . kik m,, Q; g . m.. m. %":f r mag =K_k??,"'fll_' !;Q,i mme. x C Q@.,a q n - {d t

I e

'p;. \\j 'I D ;;.*'[ ,d + +j ..a ;;,, l l liiolograph 2 Southwest Sump - Prototype l N

l l ) \\ l g -S . i r-{ N. ' \\ '.* ;.%h. - j- '-[.N.'.fg. - N$ -

• 'e 2

yw -.. ' p% .j i 4 vi - - .,*y .. ' k.g ' q.. M 2 y e'q- ~'*..+.pk' 1 ?gW-+ 7' A+f r' '9 kg.[;.'s- $. s' " ' - ~ 1"' s., 55 $. a

• '2

( =. s: -e k a-Aeh 3 -n.-,6 ~ a. r _ #P %4 y*... ' - e c.- a - e ".', '.. 4(k, ; g. M, ,.> 4 '- j f. - ' 4-(,[ h 1.fQ 7 % .l.W.k '.}g ' (h f ' j.v'O 4 M s.k.-

f. [.,; I j. g T-.S1..Q $

M E ? '~ k,..$ # ' l.,-l, y, ( *3. #. -.,, %( 4 _ W Tt.- J .. g,y ? ,3 , w.h.

' -- ;1 44..e,W*y_

a.*,- y - 3. r ...a... _, .s ,. r. ~ . #F*** = 4 ,h g.lI E,.. t f.. ' l f,w.N. ;4 '. M.k i(, Q a 4.. f ~ ? g',L '. Q v f m.-- Y L. V. \\. n.*s 4. W* Y *&as. -' ','",e, .w ' ' ~ U Q P._ e ~.,.i.s *fl 6 9-u'.,= ^ As s .A'Ni. ~-, - Q sp ? .n .F. 4 k,,.. ' Q . = 4 ew '- ~_ - yr _l, :. /'. W_ f,g,T, f ?..Y( h.. l '. I 5 k. f. n 3 j k $ l y- & y. i ~ . r-?a n^.l f i ' 1. a m;..ne. + s%,h. f..;, I e l( },~. 4 - L.y ' - Y. [ [' (. " _. J Q. ny~. Fr. ;.A,e. :7,.o. w. c). n,,, ' 1 cy. .; '.y J3 ef.- j . -, ' sf .- a 39-v,=, .. / p mnu , 4.> E. g } 7',. :

• v J

e. v 7.

.'-kh.I = - W. . )a. ,a r. If . y. 7 p.;.f.c:. .y 7,, % +. C. b..$ Ylj.N{ f. ff. R.htyy4 Y S. rl 5. r g _. f j ' Y?. 3 'Ihki

3

g..

Photo;.raph 3 Oserall View ot \\1odel I 1 i t l I l i E1

l s ~- nt.g e., _s 9t ~. L,3 y, . y;* *' y.e,. r Q.:.j p 4 7..;.. +,. $ s.,.u, , yh.i. % ' ' qp+s. .i r- ?g u. n , 1.. -,, .p. ,$9 p.d[hy< k'ih i:? ,f " .'). f _.Y$ly-k- '('.'.fh Yf ,,.) _7 ' Q@e. ?Q :#.... Ui,p ' k, ],', . I' . N-r y .n -m k5 h {,. '&u- ?.f. T;g_.d.bg..

y

.'l '. i*p.. gv;k - o e 4 4 g- ;- - e ;s. .$ r k .. f. . i.. c N-h )..,.. h'hf. E - \\. / ',[ ,\\j ' .. m;[ q..-4. .r a - g. < g: ~~ ? h.& +, ~ ~ , ;; '. 9,%g. i :^j. 3 4 ,y +_ws t

47;..

's . 3s ,,k ', >.i. 4.,..f , s t e-q.._ r _ : .../l l + /.. / ) }' y. ' ' < } Q. L x,.. _.w. h l e .. q p + ,.4_y .,,.~_.. r; x ^ ..s ~ sk lj j Photograph 4 Model Containment Sump and Pump Smnp N

ym. ,nr,-- v ' m r~ y - ~ m p,.. z -.4 ~ ..,.; l M i- . s.u j r_m >1 {' _g W Q g%a% i 1 .j 3 'ew

:a.

1 k' .:U.:'. j ~' .Q ..p ;., A --. w \\ g a ~ ll 7 b } 1 3 Photograph 5 West Sump - Model 4

.r - A t.r ' ~ 'g f ,,.c-t $.1 f s' j 3

n. nn.7'

.- A J,,-5 s 3 a

?

9 ':3 ~~ c.. '?. ' '.!. 4.. d '!l ? $ ;_ *: '. , ' ' ~, Q 'y*.", ,-l 'p.{_-l. - 'e. y y. g,; :;g,.[ ; m. [ _ $, .c .1 v .v, (.... - r T .i &;g - ) .,,,3 + s .q

gg....

e' l_..:. f- ~ ' .? 3. g'.,, n...~. _. p_c 4 --* g

• Q _. _,.

-g a,1 w. ts. .. - ' ' '.^ R

• f..

Q{ g ee k.-. 8 g, Wh~~ I J .'1> ~ h~ ~ ' b

4.,

,s 8 .. > 4, fr *,.,,,,,I. .y,

• {. i;s,

'Me' \\ l e i, ,? je, i

,,,,.p= ~

Photograph 8 Flow from Mezzanine Floor Depth = 5 inches, Flowrate ' 990 gpm t i

4;] g. g ] e e 4 i 1 1 LA f e j9 . ~. ~e r 4; -0 t 4 s w ~ I 1.1 te. .h ~ 4 g [ q s ' I .n ~ ~ ,l

g.. -

'g'.. i', Photograph 9 Fle v from Slezzanine Floor Depth = 4 inches. Flowrate = 680 gpm M

b i APPENDIX A \\

I APPENDIX A TLSI Num6LM nLOCKAGE RNR SCREEN LOSS RB SCREEN LOSS F P F. t.. SCHEEN FF WEST SW WEST Sh F P F P F P F P A 6 n 0 0 0.068 0.072 0 091 0.038 0.058 0.049 0.106 0.061 e 0 0 0 1 0.043 0.000 0.081 0.000 0.049 0.000 0.042 0.000 3 0 o 0 1 0.043 0.000 0.064 0.000 0.035 0.000 0.044 0.000 0 t-0 2 0.026 0.000 0.091 0.000 0.006 0.000 0.048 0.000 a u 0 2 0.040 0.000 0.094 0.000 0.024 0.000 0.059 0.000 u 7 1 0 0 0.000 0.040 0.000 0.098 0.000 0.014 0.000 0.064 w e 2 0 0 0.000 0.044 0.000 0.0m7 0.000 0.034 0.000 0.072 9 9 = 0 0 0.000 0.035 0.000 0.060 0.000 0.019 0.000 0.062 v 10 4 0 0 0.000 0.022 0.000 0.076 0.000 0.020 0.000 0.062 = 0 0 0.000 0.060 0.000 0.061 0.000 0.024 b.000 0.014 s 11 6 0 0 0.000 0.029 0.000 0.083 0.000 0.034 0.000 0.070 12 6 1 G 0.000 0.064 0.000 0.092 0.000 0.056 0.000 0.048 13 a v s 14 i 2 0 0.000 0.097 0.000 0.186 0.000 0.061 0.000 0.062 s 15 3 0 0.000 0.112 0.000 0.163 0.000 .0.067 0.000 0.070 4 0 0.u00 0.067 0.000 0.104 0.000 0.039 0.000 0.036 v 16 o u 17 o 5 0 0.000 0.085 0.000 0.123 0.000 0.055 0.000 0.039 5e 18 a 6 0 0.079 0.080 0 175 0.116 0.035 0.044 0.060 0.055 19 4 7 0 0.000 0.070 c.000 0.076 0.000 0.030 0.000 0.014 a w 20 a 8 0 0.0uu 0.086 C.000 0.107 0.000 0.051 0.000 0.u56 3J 21 1 6 0 0.061 0.078 0.165 0.145 0.038 0.029 0.063 0.117 31 22 i 8 1 0.07b 0.110 9.167 0.073 0.003 0.065 0.110 0.041 Jo 23 1 6 2 0.007 0.071 C.185 0.091 0.001 0.036 0.118 0.049 33 4 4 8 0 0.089 0.056 c.104 0.098 0.041 0.03? 0.027 0.000 o 1 0.102 0.087 0.114 0.002 0.068 0.021 0.057 0.034 S. 25 o 2 0.J00 0.096 0.127 0.116 0.000 0.082 0.054 0.085 3s 26 34 27 6 0 0.066 0.032 c. 912 0.11u 0.055 0.031 0.038 0.041 34 26 6 8 1 0.097 0.075 s.105 0.101 0.044 0.051 0.077 0.045 q Ja 29 8 2 0.090 0.081 0.086 0.096 0.053 0.066 0.031 0.038 ba 62 1 2 0 0.084 . 135 0.079 0.118 0.044 0.07A 0.053 0.075 51 63 s 2 u 0.0'< 0.117 0.095 0.127 0.035 0.085 0.047 0.059 bu u4 4 2 0 0 sd9 0.133 0.085 0.117 0.056 0.082 0.027 0.069 3> 65 1 4 0 0,0o5 0.066 0.071 0.070 0.031 0.070 0.015 0.034 4J u i 4 2 6.0 f 3 0.000 0.097 0.000 0.028 0.000 0.065 0.000 41 0 4 1 0.064 0.000 0.100 0.000 0.072 0.000 0.009 0.000 44 66 '+ 4 0 0.062 0.102 0.096 0.095 0.044 0.062 0.059 0.007 43 0 4 4 2 0.101 0.000 0.091 n.000 0.052 0.000 0.059 0.000 44 0 4 1 0.075 0.000 0.089 0.000 0.019 0.000 0.039 0.000 4J 67 8 4 0 0.039 0.086 c.132 0.115 0.006 0.058 0.035 0.063 +b U A 4 1 0.09u 0.000 0.119 0.000 a.030 0.000 0.048 0.000 47 0 s 4 2 0.093 0.000 0.105 0.000 v.053 0.000 0.074 0.000 33 68 1 0 0 0.079 0.093 0.071 0.075 0.038 0.040 0.036 0.055 w 69 9 6 0 0.000 0.r88 0.000 0.uS9 0.000 0.070 0.000 0.046 70 t 6 0 0.000 0.086 0.000 0.810 0.000 0.051 0.000 0.047 e 5* 71 4 6 0 0.073 0.080 0.053 0.074 0.040 0.055 0.024 0.053 J 72 a 6 0 0.000 0.105 n.000 0.103 0.000 0.062 0.000 0.055 33 73 6 0 0.077 0.059 c.042 0.076 0.000 0.010 0.043 0.039 43 59 7 4 0 0.081 0.088 o.106 0.087 0.068 0.070 0.050 0.046 49 61 7 4 1 0.097 0.071 0.080 0.101 0.050 0.057 0.036 0.045 bJ 60 1 4 2 0.070 0.074 0.100 0.100 0.056 0.035 0.024 0.053 bl 77 7 o 0 0.075 0.084 0.165 0.109 0.045 0.055 0.106 0.045 a 78 7 6 0 0.000 0.082 0.000 0.076 0.000 0.024 0.000 0.035 s '9 7 2 0 0.000 0.088 0.000 0.143 0.000 0.043 0.000 0.080 NOTr. FLOOR GRATING (FG; AND SCREEN BLOCKAGE CONFIGURATIONS ARE SHOWN IN FIGURES 8 THROUGH 14. APPROACH FLOW DISTRIBUTION (FF), O INDICATES NO FLOW STRAIGHTENER BLOCKAGE,1 INDICATES WEST 50% BLOCKED, AND 2 INDICATES SOUT. 50% BLOCKED. )

1 APPENDIX B e

APPENDlX B Ts.bf r.UhLLet dLOCNAGL SWIRL Ai4GLE F P r.G. SChlEm FF F 4 0VI.L PROTOTYPE RHR %d NHR RB 1 0 0 0 -1.9 2.7 -4.3 2.6 2 0 0 0 1 5.0 3.0 00 0.0 3 0 u C 1 -2.6 1.6 0.0 0.0 4 w 0 0 2 4.5 3.3 0.0 0.0 3 0 0 0 d 3.1 1.1 0.0 0.0 o i A C L 0.0 0.0 -1 3 -2.2 0 6 4 0 0 0.0 0.0 3.2 3.m b 9 3 0 0 0.0 0.0 1.6 -6.0 e i te 4 0 L u.0 0.0 -1.1 3.5 ( 11 5 0 0 0.0 0.0 -2.6 7.4 0 12 o 0 0 0.0 0.0 -1.4 3.3 0 la u 1 C U.0 U.0 7.4 -3.5 6 14 u 2 0 0.0 0.0 4.2 2.4 0 la u o 0 0.0 0.0 -9.5 -9.5 L lb 0 4 0 0.0 0.0 2.4 2.3 e 17 u b u 0.0 0.0 4.3 2.4 si it 0 6 s 1.4 1.4 -3 1 -0.4 9 19 0 7 0 0.0 0.0 -2.9 3.2 9 d( o 6 0 0.0 0.u -4.1 5.4 at 21 1 6 0 3.1 -1.7 -1 6 4.2 a7 2d 1 e 1 1.1 0.0 1.0 6.1 et 23 1 8 e 1.1 -1.7 -2.5 4.7 2 P= 4 6 u 0.7 -d.1 3.5 5.0 a4 it 6 1 -u.9 -1.1 31 5.1 ab 2E 4 6 2 -1.3 -6.0 2.9 4.7 u 27 b 6 0 b.1 -2.7 4.6 6.0 .A et 6 e 1 5.4 -2.o 4.3 6.8 at 21 6 8 2 4.8 5.5 3.9 7.1 su es 1 2 0 5.7 3.3 5.0 4.1

7 63 4

2 0 3.0 4.1 3.2 3.5 ab e4 6 2 u 4.9 2.8 S.O 3.1 s9 bL 1 4 0 .4.5 -5.5 -2 3 6.1 wb b A 4 2 -5.0 -6.o 00 0.0 = 1 L 1 4 1 -5.1 -6.3 0.0 0.0 9, ob 4 4 0 -5.6 -5.s -3.1 -6.6

• 3 b

4 4 2 -b.4 -5.0 0.0 0.0 .4 L 4 1 -b.1 -5.3 0.0 0.0 .5 of a 4 0 -4.8 -4.1 5.8 5.5 40 L 6 4 1 -5.1 -4.3 P.0 0.0 ./ L u 4 2 -b.5 -4.b u.0 0.0 .3 66 1 b 0 1.4 1." 1.1 1.3 L c9 4 6 b 0.0 0.0 1.7 1.6 t-70 3 6 6 0.0 0.0 1.8 1.5 34 71 4 6 2.4 1.7 2.6 1.6 0 72 6 L 0.0 0.0 1.5 2.5 's b 73 b 6 0 2.4 1.9 2.0 2.7 .e 59 7 4 0 -4.9 -5.8 -5 7 -6.0 49 61 7 4 1 -b.7 -3.9 -6 2 -4.9 JU bb 7 4 i -6.0 -4.4 -5.7 -4.9 31 77 7 6 0 4.7 2.7 P.6 2.2 U 76 7 6 0 0.0 0.0 -1.5 5.4 C 79 7 2 0 0.0 0.0 5.3 4.7 4WEMAGE bWIRL M4uLLS 3.9 3.4 3.4 4.2 NOTE: FLOOR GRATING (FG) AND SCREEN BLOCKAGE CONFIGURATIONS ARE SHOWN IN FIGURES 8 THROUGH 14. APPROACH FLOW DISTRIBUTICN (FF), O INDICATES NO FLOW STRAIGHTENER BLOCKAGE,1 INDICATES WEST 50% BLOCKED, AND 2 INDICATES SOUTH 50% BLOCKEO. b

N L M M M l B,!!h,!h.;!h,;$, !BMiM m. - ,, M. i t li!! n.- yNd,NN$$$f!!N..,E sk!E!E!!hh , k. r.h( :., kh. ;. <.u.: :.. s. se.:.. .:-:-y,...y.t Q,.y-p p,:;yp;p;;

,y.

?., r.yg <.. '.y ~ 7 _. - ,gr

y, ; f. y9.,;c}.,,, p.,7. s_.Q,g

.y e ~.,y. ~,; ~., r, .?, v sc, z. g ,r. r . ~.x_ z.__r.z,,j....,. p

c.,.. A. e e

c c.- 4 - _ 4;,,- , 3. ' 4_ . _.,_.,, g. o; se,,,, ,;,; y,. ,e p_ ' r f,C' rf. '. '. f.1 i e

r., :.r

. f,

r. y e

_y_ ~ @,.Ac) 5.'D- 't47T F - (M Ak e W! ~ -%p E Ao ?^'c c % % 1C ),r M. +,-4 e i' ,.9, s , f &',1%,,y.,,.4;,s,e.;. - e, -.. ': ;.,.,. :,..,,p~ *;:,.',. _. '4r

-; y..;-;,%~ r,:-*_:,:-y,.y;.vy,.,;y.,',,,;

.. 5 ; y. - o., - :-. - : c _ %;:: f - 4,, p- <- y. ,? ;; :-.' sy*, ;.-...;,s*- f,,. A - l ' jfr.} r. - r.; ;- /..,._Q.. ;,. ',' v

.'.' ;t

'. i ) . ; l- [.. _'-[ 9.... (s,.,o,. ,...a- .,,;:, -..-..,,.'*j g.,' ; s'y*/ ^;;,'!' '3 ,g,f, [o. 4 =

v..

, ~. ? ,n rn do .h ur r a m - ur s s -

• 4...sr

.4. .., T 'n % b wVR T *p u p# ",b,,,9er'- 1 '[,(,' Q $ W. ; M '.' ' * #,e ,, e3 4 vQ' h:,[:[., '*'ed k [,,,* I $ *= ya w w, c < u." Q J r mn.s.s. ; ;.. W,. , j gjjp e c.., y,7 f, .[! !'? / b C A d' [ %< N./., k.: M: N".*/,o Je 34 '# .N N, f i w M..

• O.-

Q% O h' 'O O' f') ' 'S, 9) 9 -) y , Tt - aC

• k:,,.,

,,,ke

'i,p??}..- >

~{ 4 e- ..4 A ,. T.,,..g,;,5 t,,s y. sj,q: :.. a,.a .w:.n' i.? ? : :. :... - *se e 4-f $9. T3 T* '. i T E' WT> Q E gt,AQ,s,Q.w,'T a he .,.c.. h aj,,. m....f.:', k 4 t.. 3 9.o u+ 'P - n. . _ _ _? _..? ,,c,o. a y, ~ g-.9.p'<.,..,...;, y. .,s 9. k'. .'l tl, ' '. 'l , Er'*W,' '8 - # ~ t.I. d

fj.$s, h;<L. h, h..,

h....*.

> N '.. U .)b Ik y h d......,,,w'*~f I~... ,p.*.. . -.,,. w m :...,,,,,. o,, n...... ,.'c .r ...v, .m....... J O. (7 @M M9M-M )(37,*e'S')e av N g v.'4 * ( ' 's) .. 'O' M gr,g '"" 'j*A } ;,p?q;We;j q.y, f ' ';: y.. g. n M.. v: f : j, .y,

,..,; j

.:-::3 fr n - ) 3. u.,,, ) .p :,. ?';,. ..... s.. e -r r ..,.. ' '.....,, gr;r te,ne -y.. 3 .r e..g '/. 2/.. s. ;. '. J f., t ..sf,..,/ p... . ' ei.,: 4. s ( .7..-;. ;. L, ;.-7.f..'r.-; *i'.. :. m. = a ' ?,;.,.4.,,,. @ %.M.N y ;.-. %.-X.Qf ?c '. ;.J(, '..r, p.;. ';. t%) v.% hv - :,... r..........,,...:. ... s..., g + <. ~ s.- .c.. v.. i.. MM eW8*e9 dran } ww e'aserS w, 94

I ~ [ I l WORCESTER POLYTECHNIC J INSTITUTE a l I ALDEN RESEARCH LABORATORY g HOLDEN, MASSACHUSETTS ' 01520 i mm-ww=_--

m...}}