ML20063E048

From kanterella
Jump to navigation Jump to search
Core Spray Crack Analysis at Brunswick Steam Electric Plant,Unit 2
ML20063E048
Person / Time
Site: Brunswick Duke Energy icon.png
Issue date: 07/31/1982
From: Brandon R, Gridley R, Kiss E
GENERAL ELECTRIC CO.
To:
Shared Package
ML20063E042 List:
References
82NEDO73, NEDO-22171, NUDOCS 8207130282
Download: ML20063E048 (119)


Text

- - . - - - _ . .-

NEDO 82171 82NED073 CLASSI JULY 1982

C 4

CORE SPRAY SPARGER l! '

CRACK ANALYSIS AT

BRUNSWICK STEAM ELECTRIC  ;

PLANT - UNIT 2

. C.

(\

7

~

t

!!PA!8!Pogggggg, GENER AL h ELECTR PDR

' (~

NEDO-22171 DRF E21-0067 82NED073 Class I July 1982 CORE SPRAY SPARGER CRACK ANALYSIS AT BRUNSWICK STEAM ELECTRIC PLANT UNIT 2 e

Approved- __ _f__

Approved:( 9 s

~

. . ranJ5n', Manag'er E. Kiss, 'ana er clear Services Engineering Plant Materials and

Operation Mechanics Section

' Nuclear Power Systems Engineering Department Approved-

. . Gridley, Manager Q

Fuel and Services Licensing Safety and Licensing Operation i

NEDO-22171

p. IMPORTANT NOTICE REGARDING CONTENTS OF THIS REPORT PLEASE READ CAREFULLY This report was prepared by General Electric solely for Carolina Power & Light (CP&L) for CP&L's use with the U.S. Nuclear Regulatory Commission (USNRC) for the Brunswick Steam Electric Plant Unit 2. The information contained in this report is believed by General Electric to be an accurate and true representa-tion of the facts known, obtained or provided to General Electric at the time this report was prepared.

The only undertakings of the General Electric Company respecting information in this document are contained in the General Electric Company Proposal No.

424-TY706-KB0, Rev 1. (" Core Spray Sparger Crack Contingency Analysis for Carolina Power & Light Co. , Brunswick Steam Electric Plant Unit 2", dated May 1982), and CP&L P.O. No. 769128. The use of this information except as defined by said contract or for any purpose other than that for which it is intended, is not authorized; and with respect to any such unauthorized use, neither General Electric Company nor any of the contributors to this document makes any representation or warranty (express or implied) as to the complete-ness, accuracy or usefulness of the information contained in this document or that such use of such information may not infringe privately owned rights; nor do they assume any responsibility for liability of damage of any kind which may result from such use of such information.

V

(

ii

NEDO-22171 CONTENTS P_ age,

1. INTRODUCTION AND

SUMMARY

1-1 1.1 Structural 1-1 1.2 Lost Parts 1-2 1.3 Effect on LOCA Analysis 1-2 1.4 Conclusion 1-2 1.5 Reference 1-2

2. CORE SPRAY SPARGER STRUCTURAL INTEGRITY 2-1 2.1 Sparger Configuration 2-1 2.2 Fabrication Sequence 2-2 2.3 Installation Sequence 2-2 2.4 Performance History 2-3 2.5 Potential Sources of Stress 2-3 2.5.1 Fabrication Stresses 2-3 2.5.2 Installation Stresses 2-6 2.5.3 Stresses During Normal Operation 2-8 2.5.4 Stresses During Core Spray injection 2-9 2.6 Materials Aspects of Cracking 2-12 2.6.1 Potential Causes of Cracking 2-12 2.6.2 Conclusions of Sparger Cracking 2-14 2.7 Crack Arrest Assessment -

2-14 2.8 Structural Integrity With 360* Throughwall Crack 2-14

{~>t, 2-15 2.8.1 Normal Operation 2.6.2 Core Spray Injection 2-15 2.8.3 Flow Induced Vibration 2-15 2.8.4 Conclusions 2-15 2.9 References 2-16

3. LOST PARTS ANALYSIS 3-1 3.1 Introduction 3-1 3.2 Loose Piece Description 3-1
3. 3 Safety Concern 3-1 3.4 Safety Evaluation 3-1 3.4.1 General Description 3-2 3.4.2 Postulated Loose Pieces 3-2 3.5 Discussion of Probability 3-6 3.5.1 Fuel Bundle 3-6 3.5.2 Control Rod Mechanism 3-11 3.6 Conclusion 3-11 3.7 Reference 3-11

(

iii

NEDO-22171

(~ CONTENTS (Continued)

P_ age

4. IASS-OF-COOLANT ACCIDENT ANALYSIS WITH A CRACK IN ONE CORE SPRAY SPARGER 4-1 4.1 Introduction 4-1 4.2 Limiting Break Size and Single Feilure Analysis 4-1 4.3 Phenomena Involved in the Analysis of Sparger Performance 4-3 4.4 Analysis Results 4-5 4.5 Bounding Sensitivity calculation 4-6 4.6 Conclusions 4-6 4.7 References 4 -7 APPENDICES A. STRUCTURAL ANALYSIS OF THE BRUNSWICK 2 CORE SPRAY SPARGER A-1 B. FLOW VELOCITY CALCULATIONS B-1 iv

NEDO-22171 TABLES f'

Table Title Pg 3-1 Probability of Fuel Bundle Blockage 3-10 3-2 Summary of Results of the Probability Study 3-10 4-1 Key Phenomena Related to Core Spray Cooling Performance 4-4 1

l l

l l

l r

k I

l

}

t v/vi

NEbO-22171 ILLUSTRATIONS -

g.

t Figure Title P_ag 1-la Plan View of Cracked Core Spray Sparger 1-3 1-lb Elevated View of Cracked Core Spray Sparger 1-3 2-1 Core Sprey Sparger - Elevated View 2-17 2-2 Core Spray Sparger - Plan View 2-18 2-3 Sparger.to Shroud Attachment Method 2-19 2-4 Sparger. Nozzles 2'-20 2-5 Sparger Support Method 2-21 2-6 Pipe Bending Method 2-22 2-7 Sequence of Events Leading to the Residual Stress Distribution 2-23 2-8 Bilinear Stress-Strain Curves for Type-304 Stainless Steel 2-24 2-9 Stress and Strain Distribution in the Pipe Under

(- Applied Moment 2 25 2-10 Moment Versus Outer Fiber Strain 2-26 2-11 Resultant Residual Stress Distribution After Fabrication 2-27 2-12 Postulated Installation Stresses 2-28 3-1 Reac-or Vessel 3-12 3-2 Steam Separator 3-13 3-3 Largest Piece That Can Fit Through the Turning Vane (End View) 3-14 3-4 Orificed Fuel Support 3-15 3-5 Fuel Assemblies and Control Rod Module 3-16 3-6 Loose Piece Potential Upward Flow Path 3-17 4-1 Brunswick 2 DBA (Limiting LOCA) Analysis 4-8 v11/viii

NEDO-22171

1. INTRODUCTION AND

SUMMARY

One of the scheduled tasks during the Reload 4 refueling and maintenance outage in May, 1982 at the Brunswick Steam Electric Plant Unit 2 was the performance of a visual inspection of the Core Spray Spargers using underwater television cameras. This inspection was conducted as required by and in accordance with IE-Bulletin 80-13 (Reference 1-1).

During the inspection of the core spray spargers, a crack indication was observed on the upper s'parger in the heat-af fected-zone of the sparger to T-box weld. It is GE's understanding that this crack is approximately 120 in length, 5-10 mils in width. The crack location is shown in Figure 1-1.

The discovery of the crack was reported to the NRC by CP&L.

On June 14, 1982 CP&L and GE met with members of the staff to present tech-nical justification to support plant start-up with the core spray sparger in its current condition. This presentation included the technical basis to establish continued structural integrity of the upper core spray sparger for all normal and injection conditions. A discussion of the possible consequences of potential loose pieces from a cracked sparger was also presented. Finally, the effect of a postulated Loss Of Coolant Accident (LOCA) with a cracked core spray sparger was discussed. This information is documented in this report.

Supplemental information provided by GE to CP&L on June 15, 1982 for use in continued discussions with the staff is also documented here.

l 1.1 STRUCTURAL i

A structural analysis is presented in Section 2, which describes the potential sources of stress in the spargers resulting from fabrication, installation, l

normal operation, and operation during postulated LOCA. It is concluded that the structural integrity of the sparger will be maintained for all conditions of operation. In addition, potential causes of cracking are discussed, and it i

is concluded that the most likely cause is Intergranular Stress Corrosion l

Cracking (IGSCC).

i l

l-1 l

I l

NEDO-22171

( l.2 LOST PARTS Because continued sparger structural integrity has been demonstrated, lost parts (loose pieces) are not expected. However, a lost parts analysis has been performed and is presented in Section 3. It is concluded that the prob-ability of unacceptable flow blockage of a fuel assembly or for unacceptable control rod interference is essentially zero. The potential for corrosion or other chemical reaction with reactor materials is essentially zero because the sparger material is designeu for in-vessel use. It is also shown that loose pieces are not expected to cause damage to the other reactor pressure vessel internals.

1.3 EFFECT ON LOCA ANALYSIS Section 4 presents the results of LOCA analyses assuming one cracked core spray sparger. GE believes that the current reload analysis calculations are still valid because coolant injection to the upper plenum is maintained.

( It is concluded that there is no basis to impose a Maximum Average Planar Linear Heat Generation Rate (MAPLHGR) penalty on Brunswick 2.

1.4 CONCLUT10N 1

l A detailed evaluation of the Brunswick 2 core spray sparger cracks has been J

performed. This evaluation included structural, lost parts and LOCA analyses 1

to determine the impact on plant operation of the cracked sparger. It is concluded that Brunswick 2 can safely operate in this condition and that no operational changes or restrictions are required.

1.5 REFERENCE l-1 USNRC IE Bulletin No. 80-13, Cracking in Core Spray Spargers, May 12, 1980.

1-2

NEDO-22171 o'

s ,

p .

\ On M afn ,

a i

\t /

7

/

N s.

/

< /

N // .-

CORE SPRAY SPARGER I-w" d

_ _- . I._, . ..___. _ ..

_g i .

f C. \ '

/

\\

\ s

'/'

?Q If0! CATION Y k

/ --

  • / ': @

I fL 'l ,

i $. t 180*

Figure 1-la. Plan View of Cracked Core Spray Sparger O

( INDICATION v 170* T 40X Figure 1-lb. Elevation View of Cracked Core Spray Sparger 1-3/1-4

NEDO-22171

2. CORE SPRAY SPARGER STRUCTURAL INTEGRITY 2.1 SPARGER CONFIGURATION The core spray sparger configuration is shown in Figure 2-1 through 2-5.

The spargers are mounted in the upper shroud, as shown in Figure 2-1. Ver-tical spacing is 10-1/4 inches between header pipe centerlines. The upper sparger has bottom-mounted nozzles and the lower sparger has top-mounted elbows. The plan view (Figure 2-2) shows that the spargers are asymmetric.

The shorter header pipe has an arc length of 80*, and the longer header pipe has an arc length of 100*. The T-boxes for the spargers are located 110*

from the vessel 0* and 180* azimuths.

Figure 2-3 shows the attachment of the T-box to the shroud. The T-box is a 5-in. Schedule 40 section of pipe with an end plate toward the vessel center-l line. The 5-in. pipe extends through the shroud wall and is butt-welded to external piping. The T-box pipe is attached to the shroud by the seal ring

(~

with the attachment welds to the 5-in. pipe and the exterior surface of the shroud wall. This arrangement (as opposed to direct attachment) eliminate's high thermally induced stresses during core spray injection of cold water into the reactor at operating temperature.

The sparger flow nozzles are depicted in Figure 2-4. The Brunswick 2 upper and lower core spray sparger headers use 1-in. shielded VNC nozzles alternat-ing with SPRACO 3101 nozzles.

The 100* header pipe and the 80* header pipe are each supported at th 2e locations. Figure 2-5 shows the support arrangement at locations between the T-box and end supports. The brackets are 1/2-in. thick and are welded to the shroud. The pipe-to-bracket mating surfaces are not welded to allow circum-ferential relative motion between the header pipe and the shroud during a core spray injection of cold vter into a system at reactor operating tempera-ture. The header pipe is 3-1/2 in. Schedule 40 Type-304 stainless steel.

(.

2-1

NEDO-22171

'~ The street elbows, 90' elbows, orifices, tees and the close nipples (used to connect the elbows and orifice the elbows) are all Type-304 stainless steel.

2.2 FABRICATION SEQUENCE Fabrication records show that the Brunswick 2 spargers were fabricated as follows:

1. The pipe was bent using a four-roll bending process as shown in Figure 2-6. The rollers have 2 in, radius grooves, and rollers 3 and 4 are adjustable to accommodate the pipe size and to bend the pipe to the required radius. In this case, the design radius is R = 91.25 inches. The maximum strain in the pipe is calculated to be 2.2%.
2. After the pipe is bent to the proper radius, it is placed in the shroud to verify that the pipe fits the shroud as-built conditions.

During this fit-up process, the T-box 5-in. pipe is marked for

(

drilling the header pipe holes.

3. After removing the pipe from the shroud, the headers are welded to the T-box.
4. The holes for each nozzle are drilled in the header pipes.
5. Stainless steel orifices are bevel welded at each nozzle opening.
6. The elbows are screwed into the assembly and roughly aimed.

2.3 INSTALLATION SEQUENCE The sparger is installed in the shroud in the following manner:

1. The brackets are welded to the shroud, thereby positioning and I holding the spargers. The T-box is attached to the shroud by 2-2

NEDO-22171 welding the seal ring to the T-box and the shroud. It is assumed

{

that, because of interference between sparger ends, one or more of the spargers would be cold sprung during installation. This opera-tion was not addressed in the fabrication records.

2. The next operation was to aim the nozzle as required by the sparger drawing.
3. The elbows were then tack welded to assure that the threaded connec-tions remain intact.

2.4 PERFORMANCE HISTORY Brunswick 2 first went critical in March 1975. There have bee no inadvertent core spray injections. Brunswick 2 does flush the core spray spargers during refueling outages. Water is pumped from condensate storage at a temperature of approximately 70*F into the vessel at a temperature of approximately 200*F.

The resulting AT is sufficiently low that fatigue is not a concern.

2.5 POTENTIAL SOURCES OF STRESF The potential sources of stress in the core spray sparger which could result from f abrication, installation, normal plant operation, and operation of the core spray system during postulated loss-of-coolant accidents are presented l in this section.

l 2.5.1 Fabrication Stresses Residual stresses are developed when an initi '.ly straight pipe is subjected l

to a moment sufficient to cause yielding and later unloaded, as would occur during the fabrication of the core spray spargers. The fabrication operation j is idealized in Figure 2-7. The steps involved in the calculation of the residual stresses are:

' l. Determine the moment-curvature curve for the pipe assuming simple l beam theory.

l 2-3 l

l __

NEDO-22171

2. Cr.lculate the applied moment, M , corresponding to the final un-

{ g loaded radius of curvature. Determine the . stress distribution associated with this moment.

3. Calculate the elastic stress distribution corresponding to the applied moment (-M g ) to describe the unloading.
4. Determine the residual stress in the pipe which is the algebraic sum of the elastic-plastic stresses due to Mg and the elastic stresses due to (-M ).

In calculating the movement-curvature curve for the pipe, thin shell theory was applied and a representative bilinear stress-strain curve (Figure 2-8) was used.

As shown in Figure 2-9, the strain varies linearly through the section, while the stress follows the bilinear curve for angles greater than 0.

The applied moment (Mg ) is given by:

h M =2 (Ec Sin 4) (a sin 4) (2atdc) p,/2

+2 {(c Sin 4 - c ) E + Ec } (a Sin $) (2atdt) (2-1) where l

c, = a/R = outside' strain a = radius of pipe R = radius of curvature

{

i' 2-4 I

NEDO-22171

(.

ic c,o = yield strain y y E. E = elastic and plastic modulus, respectively g

The first term in Equation 2-1 is the contribution from the elastic part of the stress distribution, and the second term corresponds to the plastic por-tion of the stress distribution.

After integration and rearrangement, Equation 2-1 becomes:

t (1 - Eg/E) 00 - S M O)

E

" ~

+ cose t o w sine + E sine c R Y

  • and Sine = c /c 9 ,

M = moment corresponding to the first onset of yielding on the outside surface = a w a t.

Clearly, for fully elastic behavior, 0 = r/2, and Mg =M.

Figure 2-10 shows the variation of the applied moment with the outside fiber strain and also the bend radius R. As shown in the figure, in order to get a final radius of 91.25 inches, the outer fiber strain during bending is 2.37%.

2 The corresponding moment is 1.43 o na t, The residual stress distribution can now be determined by combining the elastic stress corresponding to (-M ) and the elastic-plastic stress during t

bending. Figure 2-11 shoes the resulting stress distribution.

Figure 2-11 shows that the pipe is subjected to high residual stresses (approaching the yield stress), and that the stress distribution varies around the circumference of the pipe. In particular, it shows tensile stresses on the surface facing the centerline of the vessel. It should be noted that the actual stresses could be higher due to local yielding at locations where 2-5

NEDO-22171 Hertzian contact stresses (between the roller and the pipe) occur during

(

bending. Since this would be most likely to occur on the surface of the sparger facing the center of curvature, higher stresses could be expected at this location.

The residual stresses shown here were calculated for room temperature condi-tions. However, for reactor operating temperatures =550*F, the residual stresses are expected to relax to the yield value at that temperature (18.8 ksi at ' 550*F) .

The conclusions from the evaluation of fabrication stresses presented in this section are summarized below:

1. Stresses due to f abrication could be significant and would exist throughout plant operation.
2. A possible synergistic combination of adverse metallurgical condi-tions (e.g. , sensitization, cold work) and high residual stresses

(

may explain the observed cracking.

3. Since the stresses change sign (become compressive) around the circumference, a crack that initiates in the tensile region can be expected to arrest in the compressive regions.

I 2.5.2 Installation Stresses Stresses sufficient and necessary to cause initiation and propagation of l

l cracks by intergranular stress corrosion cracking (IGSCC) can be identified l by postulating certain installation variables. Figure 2-12 shows two cases which might be postulated.

i l

In Case 1, it is postulated that differential weld shrinkage occurred during welding of the header pipes to the T-box. The outer bracket would provide a l

force to cause the header to contact the shroud wall. For simplicity, the i (,

l arm is assumed to have an are length of 90*. A l/8-in. differential weld 1

2-6

NEDO-22171

- shrinkage is assumed. The deflection resulting at the header end would be approximately:

(8 =,1;2,,s=2.85 inch.

Then, from Reference 2-1. Table 13.4, Case 1:

A = WR - sin 24) , where $ = 90*F .

4EI

Solving W = 647 lb, assuming:

R = 91.25 in.

6 E = 28.3 x 10 ksi I = 4.79 in.

Since M = WR y , WRC , 647 x 91.25 x 2.0 1 4.79 o = 24700 psi e = 25000 psi (elastic)

For Case 2, it is assumed that R is incorrectly fabricated to a radius of 90.75 inches. It is further assumed that the vessel brackets cause a uniform moment on the pipe, thus increasing the radius to 91.25 inches.

I, The initial inner length is w/2 x 88.25 = 138.62. Af ter forming, the inne?

length is w/2 x 89.25 = 140.19:

i e

I""*# 140.19 - 138.62 = 0.011 f S rain = c =

t

=

138.62 inner v.

= 1.1%

2-7 f

NEDO-22171

( Using a stress strain curve for Type-304 stainless steel, the resulting secondary stress is found to be 37,000 psi for 1.1% strain.

For the postulated conditions, these two examples show that high deflection limited tensile stresses can occur during installation. These stresses have not been confirmed. In addition, the welding process produces residual stresses in the pipe near the weld. The magnitude and sign of the stresses vary with distance from the weld and depend on pipe size and welding speed. These stresses are likely to vary circumferentially. Maximum tensile residual stresses in the range of 18 ksi to 40 ksi have been measured in weld pipe tests (Reference 2-2).

Installation stresses considered in conjunction with the material considera-tion discussed later (Section 2.6) may explain the cracks that have been observed. It should be emphasized that the installation stresses postulated above are all deflection-limited secondary stresses that will relax to the elevated temperature yield strength of the material during normal plant operation.

{,

2.5.3 Stresses During Normal Operation All identified stresses during normal operation were found to be negligible.

Loadings that were considered include impingement loads (i.e., flow past the spargers), seismic, pressure, thermal mismatch, stagnant line top-to-bottom i

temperature gradients, stagnant line throughwall temperature grtdients and weight. Stress calculations are given in Appendix A.

l It should be noted that, during normal plant operation, there is no core spray l flow. The sparger AP = 0 and AT = 0. Impingement loads are 1.0 lbf/in. on the header arm, resulting in negligible stresses. Weight of the spargers and water is only 1.37 lbf/in., again resulting in negligible stresses. Stagnant line temperature gradient ca.iculations are not provided since the maximum AT for top-to-bottom gradients and for through wall gradients were found to be less than 8'F, which would result in insignificant stresses. It should be noted, however, that the AT for core spray injection is addressed in Section 2.5.4.

2-8

NEDO-22171

~

It is concluded that the normal operating loadings do not result in stresses that could explain the crack observed in the Brunswick 2 core spray sparger.

2.5.4 Stresses During core Sprav Injection Stresses during core spray injection are the design stresses for the spargers.

Design loadings include all those discussed in Section 2.5.3 plus those that occur because the system is no longer a passive system. The pressure differ-ential in the sparger at rated flow is approximately 28 psid. The hoop stress Impingement 1oad stresses are less during spray

~

in the pipe is about 230 psi.

injection than during normal operation. Thermal stresses due to the through-wall temperature gradient are high and are given by:

o E a . AT T" 2(1 - p)

These stresses are not a concern for one or a few cycles. The radius of the sparger shrinks when the sparger is cooled, resulting in secondary bending

(-

-' stresses of approximately 3100 psi. The axial stress in the pipe due to AP and bracket friction is low--less than 220 psi. Flow through the nozzles results in stresses in the nozzle-to-pipe weld which are low--less than 500 psi.

Weight stresses are negligible. Water hammer is not expected because the pipe is essentially an open pipe, and the nozzle opening areas are approximately equal to the pipe internal area, even for the short leg. However, water hammer is addressed in the following section.

2.5.4.1 Water Hammer Loads Water hammer loads as discussed herein are those hydraulic loads associated with injection of core spray water into a core spray system, where the system piping downstream of the check valve in primary containment is assumed empty (or filled with steam) because of the draining of water from the spargers and/or the flashing of water to steam during depressurization prior to core spray injection.

2-9

NEDO-22171 For the purpose of maximizing injection loads on the core spray spargers, it is assumed that reactor pressure is essentially atmospheric (as for a large LOCA), enabling system flow to increase to runout controlled only by the injection valve opening characteristic. Upon valve opening, the head (H) is available to accelerate the flow, but as the velocity increases, the acceleration head is reduced by friction and local losses. If L, is the equivalent length of the pipe system, the final velocity V isf given by application of the energy equation:

2 L V

"' 2g The maximum velocity attainable is limited to that at system runout flow (6000 gpm), which produces a velocity of 54 ft/sec in the sparger (at the entrance to the long sparger arm to be more concise; the velocity at the ends is zero).

t This is conservative because the velocity of the water first entering the sparger will be less than runout velocity because of the relatively slow opening chracteristics of the injection valve. The injected water fills the pipe line between the injection valve and the sparger at a time prior to full valve opening and therefore, before the final runout velocity is attained. ,

Assuming the maximum velocity attainable, the resulting momentum load in the sparger is:

2 V (54) = 39.3 psi m " 144 gv " (144(32.2) (0.0160) or m A = 39.3 (9.89) = 389 lb.

F =P m p u

2-10

NEDO-22171

~

where P, = momentum pressure (psi);

F, = momentum load (lb);

V = velocity (ft/sec);

g = gravitational acceleration (32.2 ft/sec );

v = specific volume (0.0160 ft 3/lb) (%80*F water); and A = pipe flow area (9.89 in.2) (3-1/2-in. Schedule 40 pipe).

p If the end plates at the ends of the spargers were removed, it is obvious there would be no impact load. Now cap the ends and also plug the sparger nozzles. Again, there would be no water impact load because the trapped gas

{ in the line acts as a surge tank.

The actual end condition of the spargers is somewhere in between these two extremes. It is much closer to the open end condition, except that there are several " ends" instead of one end, and they are located along the length of the sparger arms.

The exit flow area of the sparger nozzles is computed as follows:

Area Total Area Number (in.2) (in.2) 27.5 1-in. VNC Nozzle 27 1.018 3101 Nozzle 25 0.307 7.7 Total Open Flow Area Per Sparger = 35.2 The exit flow area of the nozzles and elbows is actually 78% greater than the flow area of the two sparger arms (2 x 9.89 = 19.78 in.2),

2-11

NEDO-22171 An estimate of pressures induced in the sparger at the end of the filling time i of the spargers and piping can be made by considering a sparger with only one cpen elbow located at the end of each arm. Steam would be pushed ahead of the oncoming front of water, exiting the sparger through the assumed single nozzle.

The developed differential pressure to expel the steam would be approximately 7 psid. Adding all sparger elbows and nozzles to this logic clearly demon-strates that the sparger indeed behaves like an open-ended pipe, and conven-tional water hammer loads of any significant magnitude would not be present.

Injection conditions at higher reactor pressure would clearly be bounded by the runout case presented here.

2.6 MATERIALS ASPECTS OF CRACKING The potential causes of Brunswick 2 core spray sparger cracking are discussed in this section.

2.6.1 Potential Causes of Cracking

( The potential metallurgical causes of core spray sparger cracking can be divided into those relevant to cracking in the heat-affected zone (HAZ) of the T-box to sparger arm weld, as well as those related to cracking in the sparger arm remote from the weld. Cracking in both regions is discussed here even though the crack in the Brunswick 2 core spray sparger .is located in the HAZ of the weld.

Near the T-box, three possible causes of sparger cracking have been identified.

First, sensitization by welding the sparger arms to the tee box is the most likely causc. It is known that ICSCC can result in the presence of this metal-lurgical condition with sufficient sustained stresses. In piping systems, the weld residual stresses plus sustained primary or secondary stresses equal to the materials yield stress can result in cracking. Moreover, the small amount of cold work inherent in sparger arm fabrication prior to welding could serve to increase cracking susceptibility. When the material is cold worked and subsequently sensitized, the susceptibility to IGSCC is markedly increased for J

cold work levels less than 20% (Reference 2-3). Carbide precipitation kinetics 2-12

NEDO-22171 are markedly increased and some martensite can be formed. It has been shown that the presence of martensite accelerates kinetics of crack initiation and propagation.

Second, fatigue induced by thermal variations in the environment may be the cause of the sparger indications. However, the variations in temperature dur-ing operation of the reactor (8'F, see Section 2.5.3) are expected to be small.

No evidence of a driving force for thermal fatigue has been identified.

Finally, fatigue resulting from flow-induced vibrations could be hypothesized.

However, the natural frequencies of the sparger are high relative to any flow-induced excitation sources (Appendix A.2).

In the arms remote from the T-box by distances greater than 2 inches, welding cannot be considered a major influence on cracking. Sensitization may still be present if the original solution heat treatment was inadequate, either in temperature or quench rates. No direct evidence exists of this condition.

Secondly, if cold work from arm bending were followed by local heating, a sousceptible condition would more readily exist. Again, no direct evidence exists. Thirdly, surface cold work resulting from arm bending or straightening could hasten crack initiation and subsequent growth could occur from redisual or installation stresses. Finally, fatigue by either of the sources cited above for the T-box area could induce cracking, although there is no confirmed source of fatigue loading.

It is postulated that cold work, both local and overall, associated with the bending could produce cracking remote from the T-box. It is known that high levels of cold work without subsequent sensitization can result in initiation of transgranular stress corrosion cracks which can subsequently propagate in an intergranular fashion (Reference 2-3). For the sparger, the overall cold work could be due to the sparger arm formation alluded to above. The local cold work can result from local smearing of metal during the bending operation.

While the overall cold work level is small - about 2%, the local cold work level can be quite high.

(-

2-13

NEDO-22171 A sparger with fabrication history very similar to that at Brunswick 2 was examined and found to have these locally smeared areas. Constant Load and Constant Extension Rate stress corrosion test specimens were fabricated from sparger segments with and without the smears. In addition, specimens contain-ing the cold work smears were subjected to solution heat treatment and also tested. Ferrite scope measurements taken on the smeared areas indicated the presence of about 5% ferrite indicating the presence of a magnetic phase such as martensite. Hardness tests on the smeared areas also revealed high local cold work. Normal hardness and no ferrite were found in the non-smear type specimens. The constant extension rate tests were conducted in 550*F, 8 ppm oxygenated water. Transgranular cracking was noted in specimens containing the cold work smears along with some expected intergranular cracking of the underlying material resulting from the crevice formed by the transgranular crack. No such cracking was found in sparger material without smears or in solution annealed material. Constant load tests resulted in similar results with failure times of the smear specimens 69 to 325 hours0.00376 days <br />0.0903 hours <br />5.373677e-4 weeks <br />1.236625e-4 months <br /> as opposed to about 4000 hours0.0463 days <br />1.111 hours <br />0.00661 weeks <br />0.00152 months <br /> on test for the other specimens without failure.

2.6.2 Conclusions of Sparger Cracking

(

In summary, the most probable cause of core spray sparger cracking at Brunswich 2 is cold work and subsequent sensitization.

2.7 CRACK ARREST ASSESSMENT An assessment of crack propagation has been prepared for CP&L by a third party for CP&L's use with the USNRC for the Brunswick Steam Electric Plant Unit 2.

l 2.8 STRUCTURAL INTEGRITY WITH 360* THROUGHWALL CRACK l

Even though GE believes that a 360* throughwall crack is improbable, a struc-tural analysis was performed (See Appendix A) which conservatively assumed that the existing crack propagated 360* throughwall. Loads which were considered l included all loads applicable to the intact sparger (See Section 2.5.3). The analysis ignored the ef fect of clamp (or assumed a clamp was not installed).

t 2-14

NEDO-22171 2.8.1 Normal Operation Bending stress in the broken aparger arm due to impingement is low -- 868 lb/in . During the postulated seismic event, bending stress is calculated to be 2380 lb/in . All other operating loads result in negligible stresses.

2.8.2 Core Sprav Injection Stresses on the broken sparger during core spray injection are bounded by the stresses given for the intact sparger (Section 2.5.4). Normal stress will therefore be less than 3320 lb/in in the sparger pipe. In the nozzle-to-pipe weld, the maximum normal and shear stresses are 4270 lb/in and 4400 lb/in ,

respectively. In the welded brackets, bounding normal and shear stresses are 2 2 5980 lb/in and 752 lb/in , respectively.

2.8.3 Flow Induced Vibration Flow induced vibration is not a concern for the broken sparger case. The ratio of natural frequency of the broken sparger arm to the vortex shedding

{ frequency is greater than 6, which exceeds the GE design basis by more than a factor of 2.

2.8.4 Conclusions Stresses during normal operation and during core spray injection were found to be well below allowables. The natural frequency of the assumed broken sparger remains high enough so that flow induced vibration is not a concern.

It is concluded that the sparger will lose no pieces and will remain attached to the shroud wall under the conservative assumption that the existing crack propagates 360' throughwall.

2-15

NEDO-22171

2.9 REFERENCES

2-1. Hopkins, Design Analysis of Shafts and Beams, McGraw-Hill Book Company.

2-2. H. H. Klepfer, et al. " Investigation of Cause of Cracking in Austenitic Stainless Steel Piping," NEDO-21000-1, July 1978.

2-3. R. L. Cowan and G. M. Gordon, "Intergranular Stress Corrosion Cracking and Grain Boundary Precipitation of Fe-Ni-Cr Alloys," NEDO-12399, September 1973.

l

(

2-16

zmhOGV O U T-I I

_ f S

E w L e Z i V

Z 0 n N o i

> Y t a

. A v R e

\ P l

\\ S E N .

\ -

r

\ t e wr LO\

g r

a Ci[pI s '

S p

r L

F I y L ." 'iJ}. a r

A I H oo S p

sn R e E r G o C

R A

P 1 S -

_ . 2

/ e

. r

_ 1

/ L u g

_ / i F

.I I

p O ,_3

[ 'a Y

,&n

NEDO-22171 o

mA i la 4fa

'o .

/

g\

_ 7- ,

7 s

  • /

~N // .-

CORE SPRAY SPARGER M

b..)._, t __

., _t 3, g _

8 i

k /

\ -

~

/ \ .

N  !

g

~~

-j ,.

'\

l P \

1 ii l80*

l Figure 2-2. Core Spray Sparger - Plan View 2-18

NEDO-22171 h,

. _.sumoup comE M AI m

i I

'O i l

g i 7~... _-m it'O LOWE R_ SP AR G_E R

\ l i TOP guide V

Figure 2-3. Sparger to Shroud Attachment Method 2-19

NEDO-22171 SMAou6 1

F;;t t m af I \

Figure 2-4. Sparger Nozzles 2-20 l __

NEDO-22171 swRouc

[ LDE TO SMROUD

//

1 L

f%=)

[ ONE E I Hon

.Lowsm

-- sR ACKET_

Figure 2-5. Sparger Support Method 2-21 i . --

NEDO-22171

'(

91.25 in.

T

( -

k.

= 07_

l l

( Figure 2-6. Pipe Bending Method 2-22

NEDO-22171

\

INITIAL CONPtGum AT10N M ..

l' *M,

~

l LOAD AMLICATION DumlNG F ASalCATION l

\

/ / g

/

A g N

/ s ' \

N

/ g ,e s N N N *%, " ' - f bMlg \.N  %

,/

/

/ .. _

(- M

....t'

===. '

, FINAL DEFOmMED SHAPE AFTER LOAD REMOVAL Figure 2-7. Sequence of Events Leading to the Residual Stress Distribution 2-23 l

l l

NEDO-22171 l

l l

l moou TEMP (74* pi I l

~'

~

'y m 0

~

1 'y - 34.3 (,,

'y

  • 0.0012 E* 28 3 a s03 y,,

3 g,,

ET* 0.26m10

'Y E

(a)

(-

$50* 8 0 oy =%

  • 18 8 ksi v

'y

  • 0 0007 3

E

  • 25.8 = so g,,

3 g,,

ET* 0 26 n 10

'v E (b)

Figure 2-8. Bilinear Stress-Strain Curves for Type-304 Stainless Steel 2-24

NEDO-22171 l l

l l

I i

L a 3., .e - ,

g- - - -

t. ... E,,,,_ ,,, T i M A

8 8 ELASTIC.PLAg77C

. u

,s ' ~ ~ _ -_____

h...

/ ..

[i @ 57!C 2

N'y

. - - 9- j

( ' " 'y

=

1

! . __ # 5TM E'53 l _- 'DISTMISUTsON

E-( STRAIN

!DisTMisuvios F.'.gure 2-9. Stress and Strain Distribution in the Pipe Under Applied Moment 2-25

NEDO-22171

o. .-
a. -

r S o n

R I "

l f

.T U 4g N m

W.

es N

a m

_. w_______ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ . .

e O

h m m = C

o. a._ m p W W

.O e

W E -

g

~

o W u.

9 5 6 i  ; e r > =

' V.

Y .

5 >

a .

. . I u C

0 E

C E

o O e h -

7 n

U W

D C.2 EA.

P=

w I I I I I I e w f'1 'A 4 . efi ,O

, 4 == .= 8 0 .G -

-- - - O.. 10. . s. e_

. . ~

2-26

NEDO-22171 g -

3-I'o 4

1 ,

g. -

l~

i;i / si

.E E

.a- im e

[g , .

's* 8 i.T_

I$ 5 0 n

asy

.8

._: ,e - e In.e C

3

'L~ ,t 4-3 o

8

.a - a k d b m

- .-g S e e

1 3

gC .-e D

saan E

'c3 =

E

!E B i2 -

'E E i$

!60 l l2

  • c 5

o iL y i

o ._

2-27

NEDO-22171 INSTA LLATION A ADIAL MISM ATCM DIP F ER ENfi AL WELD sumiNatAGE AT T 80x E

~

R1 = h

[ [u 5

I

' FORCE  :

- FOR SHRINK = 1/8 in.

O AT T-BOX = 25 ksi(ELASTIC)

ASSUME R1 = 90.25 R2 = 91.25 UNIFORM FORMING e

  • 1.1%

o = 38 ksi(FROM o--e CURVE)

ETsU T +?' < R2 R1 R2 s

Figure 2-12. Postulated Installation Stresses 2-28

NEDO-32171

3. LOST PARTS ANALYSIS i

3.1 INTRODUCTION

Based on the structural analysis given in Section 2, it is expected that the Brunswick 2 core spray sparger will not break and result in loose pieces in the reactor. However, an evaluation of the possible consequences of a potential loose piece is presented in this section.

3.2 LOOSE PIECE DESCRIPTION Since a piece has not been lost, it cannot be uniquely described. Three different types of loose pieces are postulated in Section 3.4.2: (1) a section of sparger pipe; (1) an outlet nozzle; and (3) a small piece of the sparger. The entire sparger is fabricated of Type-304 stainless steel.

3.3 SAFETY CONCERN

'~

The following safety concerns are addressed in this safety analysis:

1. Potential for corrosion or other chemical reaction with reactor I

materials.

2. Potential for fuel bundle flow blockage and subsequent fuel damage.
3. Potential for interference with control rod operation.

l

4. Potential for damage of other reactor internals.

l 3.4 SAFETY EVALUATION The above safety concerns for the postulated loose pieces are addressed in this section. The effect of these concerns on safe reactor operation is also addressed.

3-1 1

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

NEDO-23171 I 3.4.1 General Description The core spray spargers are attached to the inside of the core shroud (Figure 3-1) in the upper plenum. For a piece of the sparger to reach and potentially block the inlet of a fuel assembly, it would have to be carried out of the upper plenum and pass down into the lower plenum. To accomplish this, it would have to be carried by the fluid flow in the upper plenum up through the steam separators then outward to the downcomer annulus, the through the jet pump nozzle into the lower plenum, then make t. 180 F turn and be carried up-ward to the fuel assembly inlet orifices. A part of the sparger cannot reach the fuel assembly inlet orifices by falling down inside the core shroud as the core support plate and the loaded core will prevent this. For a part of the core spray sparger to reach a control rod, it must first traverse the upper plenum from the outer region of the shroud toward the center, then either fall through the restrictive passage between two fuel channels; or fall through an opening between the outside of the peripheral bundles and the core shroud, both of which are unlikely.

b?

Since all parts of the core spray sparger are designed for in-reactor service, there is no possibility that any loose part will cause any corrosion or other chemical reaction with any reactor material.

3.4.2 Postulated Loose Pieces 3.4.2.1 Sparger Pipe The sparger pipe is 3-1/2 in. Schedule 40 pipe and is attached to the core shroud at seven locations (T-box plus six brackets). The maximum span be-tween the crack and the support is approximately 61.5 inches and between

[

supports is about 65 inches. In order to generate a loose piece of pipe, a minimum of two throughwall cracks would have to propagate 360* around the t

sparger. The weight of the largest pipe segment would be approximately 84 lb.

Because of the slow rate of potential crack propagation and based on previous experience with cracks in core spray spargers, it is judged that pieces will I not break off and become loose. However, for purposes of evaluating the 3-2

NEDO-23171 probability for loose pieces, which might potentially cause core damage, a conservative estimate of this probability has been chosen at 10~3 (see Section 3.5).

A pipe segment could come to rest in any of three locations: (1) the top surface of the top guide outboard of the fuel assemblies; (2) the top surface of the fuel assembly handles; or (3) in an unlikely event, the top surface of the core plate. In all three of these locations, the flow velocity is low and insufficient to lift a segment of the pipe (see Appendix B). Therefore, it will remain at one of these locations and is not expected to lift or rattle around. An 84 lb piece of pipe which falls from the core spray sparger will not harm the core plate, top guide or fuel assembly handles, since these components are designed for much larger loads.

Since the pipe cannot be lifted by the flow and since the pipe cannot fit through either the steam separator or the jet pump, it will not cause any flow blockage at the fuel inlet orifice. Since the pipe is too large to fit r- between fuel channels, it will not cause any interference with control rod operations.

3.4.2.2 Spray Nozzle Each spray nozzle consists of two 1-in. elbows fabricated of Type-304 stainless steel, which are welded to the sparger. In order to generate a loose nozzle, a throughwall crack would have to propagate 360 around the nozzle. The weight of each nozzle assembly is approximately 1-3/4 lb. A loose nozzle would most likely come to rest on the top surface of the core plate or on the top surface of the top guide. The flow velocities in these regions are insufficient to lift the nozzle, and it will remain at one of the above mentioned locations (Appendix B).

Since the nozzle cannot be lifted by the flow and since the nozzle cannot fit through the steam separator, it will not cause any flow blockage at the fuel assembly inlet orifices. The nozzle is too large to fit between two fuel channels; thus, it cannot cause any control rod interferences.

3-3

- _ = .

NEDO-22171 3.4.2.3 Small Pieces

('

In order to generate small pieces of the sparger, both longitudinal and circumferential throughwall cracking must occur. A small piece could be lifted by the flow if it maintained an orientation with its maximum projected area perpendicular to the flow. Due to flow turbulence and assymmetry of the loose part, the part would tend to rotate so that the minimum projected area will be perpendicular to the flow. With this orientation, and based on the velocities in the upper plenum (Appendix B), all parts with a length of greater than approximately 0.4 in. would sink. Thus, most pieces would not be carried by the flow toward the steam separator. However, in the unlikely event that a piece reached the steam separator, it would have to pass through the steam separator turning vane (Figure 3-2). There are eight curved vanes with the outlet of each vane overlapping the inlet of the adjacent vane. The longest straight piece that can fit through the turning vanes is approximately 6 in.

long and it must be oriented with the long dimension in the vertical direction.

As shown in Figure 3-3, the maximum dimension of such a piece is approximately 6 x 2 x 2 inches. Thus, a piece 0.4 in. can pass through the separator.

i/

,C The fluid momentum is reduced as the water is removed in passing through the separator. At the separator exit, the fluid momentum is further reduced as the water is removed. At the separator exit, the fluid is almost entirely steam. A typical water content is 1 weight percent. Thus, it is very unlikely that any piece could be carried out of the separator by the steam. If any piece were carried through the separator by the steam, then it could be car-l l ried into the downcomer annulus, through the jet pump and enter the lower plenum. A piece that entered the lower plenum would probably be driven by the jet pump flow to the bottom of the reactor pressure vessel where it would be expected to remain. However, a small piece <0.4 in. could be carried.by the flow to the fuel bundle flow inlet orifices. There are three different sized orifices: 1.225, 1.433 and 2.09 inches.

j It is extremely unlikely for a piece larger than the 1.225 in. orifice and essentially impossible for a piece larger than the 2.090 in. orifice to be i carried through the steam separator. The outside diameter of the sparger is 3-4

NEDO-22171 4.0 in., while the fuel inlet orifices are slightly recessed relative to the surface of the control rod guide tubes (Figure 3-4), which have an outside diameter of 10-3/4 inches. Due to the different radii of curvature, flow would be able to enter the fuel assemblies. Thus, unacceptable flow blockage as defined by Reference 3-1 would require that more than one loose piece be carried to the same inlet orifice. This is based on the size of the piece (s) that, in a highly unlikely circumstance, have the potential of reaching the vessel lower plenum. The probability of unacceptable flow blockage of any fuel orifice is insignificant. This would require multiple pieces at the same orifice 'at the same time and the probability of several pieces blocking a significant portion of the bundle inlet to cause significant fuel damage is essentially zero (see Section 3.5.1).

Flow velocities near the sparger are lower than those above the fuel assemblies. Thus, it is unlikely that a small piece would be carried over the fuel assemblies (see Appendix B). If the piece were carried over the fuel assemblies and then rotated so that the flow could no longer carry it, the piece would fall on top of a fuel assembly or between fuel assemblies.

,(

Figure 3-5 shows a typical unit cell of four fuel assemblies and one control rod. The control rod moves in the gap between the fuel channels. The gap between fuel channels is 0.75 inch. The length of the gap between the channel spacer and the channel fastener is 2.3 inches. Thus, any piece larger than 2.3 in. by 0.75 in, cannot cause control rod interference. The control rod thickness is 0.312 in, and the diameter of the control rod rollers is 0.520 inches. Thus, pieces smaller than 0.334 in, will fall past the control rod without causing any interference. A piece of precisely the right size could be in contact with the control rod and one or two fuel channels. Such a piece might be detected during the normal control rod exercising. The rods are inserted one notch and withdrawn one notch each day. It is also possible, though unlikely, that a piece might wedge between two fuel channels above the control rod and thus not be detected by normal control rod operations. If the rod were to be inserted, the control rod mechanism has enough force to lift one fuel assembly with the reactor at normal operating pressura. If the fuel assembly were lifted 1 or 2 inches, it would be able to move horizontally 3-5

NEDO-22171 at both the bottom and the top, thus almost certainly relieving any interference.

The rod would then insert and the fuel assembly would fall back into place.

It is not credible that any control rod will fail to insert.

One of the licensing bases of the reactor is that with the highest worth control rod fully withdrawn, the reactor can be safely shut down. Thus unacceptable control rod interference will require multiple precisely-sized pieces inter-fering simultaneously with control rods that are in close proximity to each other. The probability of this is judged to be insignificant (see Section 3.5.2).

3.5 DISCUSSION OF PROBABILITY This section provides the basis which supports the conclusions discussed above that the probability of loose parts from the core spray sparger causing a safety concern are negligible.

Based on operating history of BWRs with cracks in the core spray sparger in

( the last few years, structural evaluation of the core spray sparger cracks, and the fact that no loose parts have been found, the generation of a loose part is believed to be incredible, i.e. , the probability to have a loose part of any size as the result of a crack is believed to be very nearly zero.

However, for the purpose of this study, a bounding probability for having a

~

loose piece break off of the sparger is assumed to be 10 .

3.5.1 Fuel Bundle For a core spray sparger loose piece to reach a fuel bundle and potentially cause some safety concern it would have to be carried out of the upper plenum and pass through the downcomer, jet pumps, down into the lower plenum and then into the core region. However, the probability of a piece being carried out of the upper plenum, through the steam separator, and outward to the downcomer annulus is limited by the minimum projected area (perpendicular to the flow) that can be lifted by the fluid flow and the size of piece that can physically fit through the turning vanes of the separator (see Section 3.4.2.3). Likewise, 3-6

NEDO-22171 if a piece were to be carried to the lower plenum, the probability of the piece potentially being carried to the core is also limited by the minimum projected area that can be lifted by the fluid flow in the lower plenum.

As discussed in Section 3.4.2.3, it is physically possible for a piece approxi-mately 6 x 2 x 2 inches to fit through a separator; however, the fluid veloci-ties in the upper plenum are not sufficient to carry this size piece out and hence it would remain there in the upper plenum. The maximum size piece that can be carried out of the upper plenum is limited to approximately 0.4 x 0.4 x 0.4 inches as discussed in Section 3.4.2.3. If this piece were to be carried to the lower plenum, which is unlikely, it could be lif ted toward the core because the vertical fluid velocity in the lower plenum is high enough to lift this sized piece. Therefore pieces of this size are used to evaluate the potential of them reaching the core region from the lower plenum (the thickness of the core spray sparger pipe is only 0.226 in.).

Figure 3-6 is a diagram of the path that a potential loose part from the sparger would follow if it were carried out of the upper plenum into the lower

( plenum and core. For each flow path, a probability is assigned and the cumulative probability is also shown (in brackets). As discussed above, the probability for a part to break off in the upper plenum is conservatively

-3 estimated to be Py = 10 . Since it is postulated that a piece of this size is small enough to be carried by the upper plenum fluid velocity it is assumed that it will leave the upper plenum, i.e., P = 1.0. Once out of the upper 2

plenum it is assumed that the piece travels with the liquid flow out of the separator and is carried out down the downcomer, i.e., P = 1.0. There, the 3

piece is likely to come to rest at the jet pump support plate at the bottom of time shroud and remain there (P6 ). However, because of the potential for the piece to be ingested into the jet pump flow (P ), or to be sucked into the 4

recirculation line (P ) and driven into the jet pump (P ), it is conservatively 7 8 estimated that the probability is the sum of the probability for these flow paths, hence P = 0.75. The probabilities, P and P , are based on 9 =P8+ 4 4 8 the projected flow areas and biased by the flow velocity ratios in the jet pumps and annulus downcomer for P and by the recirculation suction line and down-4 comer for P . Once in the jet pump it is further assumed that the part will 7

C be carried by the flow stream toward the core, hence P10 " *

  • 3-7

NEDO-22171 To enter the core region from the lower plenum, the part must first pass

(

, through the fuel bundle inlet orifices (Pyy) (there are three different sized bundle orifices). Once past the orifice, the part must pass through the lower tie plate (P12) and into the fuel bundle (Py3).

The path from the lower plenum through a fuel bundle to the upper plenum is restricted by the following flow areas:

1. Inlet orifices; three sizes 1.225, 1.433 and 2.09 inches; flow area between 1.18 and 3.4 sq. inches;
2. Lower fuel bundle tie plate; made up of 49 irregular shaped holes with a total flow area approximately 11 sq. inches;
3. Fuel bundle with 64 rods inside a square channel with a free flow area of approximately 15 sq. inches;
4. Seven grid spacers along the bundle with a free flow area of approxi-(. mately 13 sq. inches; and
5. Upper tie plate similar to the lower tie plate.

For a part to enter a fuel bundle it would physically have to pass through the lower tie plate which would further limit its size. A part approximately 1/8 x 1/8 inches would either pass completely through the bundle, beginning the cycle over again, or be trapped within the fuel bundle possibly at one of the fuel rod spacers, if it remained trapped at a spacer for a sufficiently long period of time, there is a potential for fretting wear. Extensive fretting may ultimately lead to local fuel rod perforation and possibly some small release of :ission products. This is not a safety problem because offgas radiation monitors are designed to detect fission product release in order to limit offsite dose to within the 10CFR100 limit.

l For BWR fuel, the smallest and most restrictive flow path is at the fuel bundle inlet flow orifice. The percentage of blocked orifice area must be greater

, (_ than 75% before a boiling transition condition would be approached for the l

l 3-8

NED0-22171 most limiting, peak power fuel bundle (Reference 3-1). These bundles have the largest orifices. However, for a conservative estimate the potential for flow blockage for the smallest sized orifice is evaluated. To block 75% of the area of the smallest orifice, at least five of the loose pieces must somehow migrate to the same fuel bundle inlet orifice. Five pie: es that pass through the inlet orifices and remain trapped within the bundle would block con-siderably less flow area and hence would be less restric;ive.

It is extremely unlikely to have more than one loose piece, and the probability of more than one piece migrating to the same fuel bundle is also negligible.

However, for evaluating this potential it is assumed that the probability for one loose piece to enter a particular fuel bundle is the cumulative probability for a piece from the upper plenum to reach lower plenum. This is called P and later it is set equal to the cumulative probability at P There are 560 10 fuel bundles in the Brunswick-2 core, each with an inlet orifice. The proba-bility of having n loose pieces is P". The probability of having n loose pieces from the lower plenum simultaneously, partially blocking one given fuel bundle is substantially lower. We estimate this on the number of fuel bundles, h Hence, the probability of n pieces at a given fuel bundle is:

P = (1/560)n-1, where n > 1.

Therefore, the probability for blockage of a given bundle is:

0

- (1/560)n-1 To block 75% of the minimum inlet orifices (which control the lowest powered fuel bundles) would require at least 5 small pieces somehow arriving and becoming wedged at the inlet orifice simultaneously. To estimate this prob-ability we take n = 5 in the above expression to block 75% of the smallest inlet orifice. As can be seen the probability of one piece entering the lower plenum is 7.5 x 10 '. When this is used in the above expression for P and for 75% blockage n = 5, the resulting probabilities are incredibly 10 small.

b The probability for n pieces being carried from the upper plenum to the lower plenum and going to the same orifice is shown in Table 3-1 below.

3-9

NEDO-22171 Table 3-2 summarizes the estimated probabilities and the potential consequences related to fuel damage associated with a piece or number of pieces either entering a fuel bundle or potentially blocking a bundle. As shown in this table the potential consequences do not pose any safety concerns and are of extremely low (incredible) probabilities.

Table 3-1 1

PROBABILITY OF FUEL BUNDLE BLOCKAGE n (1/560)"~1 Pn

~9 2 2x10~ 1x10 3 3x10

-6 1x10

-15

-8

<<10

-15

>3 <10 n = Total number of pieces generated P = Probability that there is at least one orifice which

, is blocked by two or more pieces up to a total of n.

(_

Table 3-2

SUMMARY

OF RESULTS OF THE PROBABILITY STUDY No. Pieces or Percent Flow Blockage Probability Consequences

~

1 piece 7.5 x 10 ' May become trapped within a bundle and after extended period could lead to local fuel rod perforation and subsequent fission product release in RPV

-15

<5 pieces <10 Same as above

-15 5 pieces <<10 Same as above with possibility of 75% blockage local boiling transition and several perforated rods.

f-3-10

NEDO-22171 3.5.2 control Rod Mechanism The probability of forming a small loose piece (small enough to fall down

~

from the sparger to the bypass region) is assumed to be 10 as tareviously' discussed. Small pieces could be lifted by the flow velocities in the upper plenum and during a hot shutdown condition of the reactor could drop or fall back and possibly pass through the flow area from the upper plenum to the bypass (approximately 18.5 2f t total). The total flow area in the upper plenum above the fuel bundles is approximately 182 ft . Therefore the prob-ability of such a piece (af ter being formed) to fall into the bypass region is 18.5/182 = 0.1. The total probability of the piece forming and reaching the bypass is 0.1 x 10~ = 1 x 10 -0 .

The probability of multiple pieces (i.e., n pieces) reaching a specific con-trol rod guide tube is:

P(n) = (10~ ) / (no guide tubes)"~ .

h For two pieces (n = 2), P(2) = (10~ )2/137 ::: 7x10 -11 ,

Therefore the probability of multiple pieces reaching the same guide tube (control rod drive) is negligible. The probability of multiple pieces inter-fering with more than one (or adjacent) control rods is substantially less.

3.6 CONCLUSION

There is no possibility for unacceptable corrosion or other chemical reaction due to a loose piece. The probability of unacceptable flow blockage of a fuel assembly or unacceptable control rod interference is essentially zero.

Therefore, it is concluded that there is no safety concern from a loose part perspective.

3.7 REFERENCES

3-1. " Consequences of a Postulated Flow Blockage Incident in a Boiling Water Reactor," October 1977 (NED0-10174, Rev. 1).

3-11

NEDO-22171 itI ase Oa, e a ..s e. esc ,4 7 ,

wer.se =as p.a. # ' -

f e* __

d ~~.

. hc ,;;.7 ;

. ."ma T.I",'"

= ~ ~'s "

. :x ~..'q; .

l

} Plass se paastoa l'i i jy'h[yi

n 8tt9meatta ya ga

= e. .,

u.yhgggl_g), - -a. -

Was paar paagga ". _M - - _ _- - g G .

l IG - - - ,; ,l;}. -

if

. k- s r . ,,, TS* Cuee 4 l} l- -

,' 11

~ ' " " * ' '

k  !

[ .l { 'E!l , =i~

j T 7 ;

r$ N - --

x .

artswe. nee l 'I i

Q4! p .. ! , y =t =

f 7 .e -

b -

'.l:

a.wi a.ni # k 'I,h'I E -

S!' ,

[h

-/  :.J :

- /h3lt sb 8 .

'N . ==-

une.atetvaammuvon Figure 3-1. Reactor Vessel 3-12

NEDO-22171

. , (, .

  • .* v.
  • tT**gau 5 at wn,isc warta b y' [

N '

ci _\h\

ml6 '

T y

  • ATE m LE v t L

)ff s S~t Av watta vix?:.;8 E I T.  ;

/ [

' agTammi~g wayg, ~

L j y .

v! \. .

f B

Y m . ~ , ~ . .~ . . . .. ., , m. . ,

1

( n' u

I a

7

/

UANOPtPE TO DOWN COutm Anga #

, Aho etciscutat C%

AND atT PvwP svCTIONs h 7 6 Figure 3-2. Steam Separator 3-13

NEDO-22171

~~

- og'.n; 2.23 a.

f.t ag ,,

3.032 - 1 ,

l Figure 3-3. Largest Piece That Can Fit Through the Turning Vane (End View) 3-14

NEDO-22171

_ ___ C

/ f>&,

s

- \

f e;OM@i y _

_~ _-

i l

cet%:sc sea es% .et .e: .g,o,

.no, g __< _

f Y (~

]

4  !

aos cone GulDE Tust 's h?

t !il

[

h

,F " =.....

j p=

,h}

s/ s =:n Omaric Figure 3-4. Orificed Fuel Support 3-15

NEDo-22171

- o.'- ..

p=. tr .

. . ,+ j. r ;t, {

'.. o. 3. , . ,... -l;g s :

. ,  ?.r

.... Q> s.s. q , . t ,,1. .. .*

f.v.,-

p f..
  • i
  • *:n We .'l

}s . 'sf. .. ? kr{ .  ?

.. .- u ~ .

I.. L. N.kj..v- ...tsi. . . [...7... Nlf.3.y.h.. i. ;,. ,$..[g I< .

y.3I m .! %.J.,: 5 .>$m. ,1 m 3 . .. e #cgg e .s.:

.g"*

  • r s 1 .; * *w

.v,

, ; vg.l;,,:.., f* ;s ,

, : .(b y.. .c..::.

. s. -

-:.- ~, :

. .;g FUEL .. a. , .

-' ,#a mm 'M ~

.a /

ASSEMBLIES

!L' tr .., ..

m - . .

g t

& CONTROL ..; y . . jy2 q .

'.l ROD MODULE '.5.13

! +

'3e ,

. . . ;j}s :i WW.:.

. . . . ,s {. .'.

i a ,<~

, . L

m. - > ,.:

6.s. ,

1. TOP FUEL SUlOE .

3 j ,*, ,*

T - l

' ]~g 4' J, y4... 70 , .

}s ,y*. u'sSj "T'i "(' sj;'.Cy 't

,;' N .t-4 2.CMANNEL . .ls .'/

FASTENER f ,. **

, . . P . ~dy . , '., ' j.;6 g.f y ,' ., .f',.$ *. '; . . , ft

3. UPPER TIE ,.
  • ff.,.# J,) ,

f , ,,

-' j h lb  %

  • M. i f /, s,.'I t PLATE .' i. 1, : .  ; .1 I 1 T. ' *
4. EXPANSION 3 .s .

ji <

%} i

. . r - t-SPRING

'[,' 5,*g.. ,,; . . . $', [ . 'f, i ,),' 'i

. 1  ; ,

j 'd, . p

'g j *.>:'755; il cf . .s#l~

5. LOCKING TAS 't *4;' e : . ' . * *
  • 4. re* i l .

J 1 J rdi -

( 7. CONTROL RCD e.CmANNEu

' ?:li:0.?' . ;

  • f ' * ,* \. '. L't,'.f

,s . *.'

l 7

&)l '

a

.~

, .',*Q. , 1.' . l'. .I .

h Y g ; [,'. :'l'. ha. p fl ). ~*

l 4 g %

i S. FUEL MOD .-[i,T.,d,',:/!,. .'y *!

W ' 7*'d '.+,?

I* ', f,t - -

9.SP ACER t' g g 5', .

. ~'* N I'e .F * '. Cll ,

~

10. CORE PLATE -

. P; . . , e . . . ', 'l , N. ~.'i, , [  ;

  1. N' ASSEMBLY
  • W- E E ' * * . .

M fe,[;

11. LOW ER 3 **.-fs  :[p.n .{I 'I'.1 V. . *fi 4dL". 53 *
    • !. $' y;.~ d. iQ,./

i

.. I Q,/y;'3.'i.i.7, ";.d Tie PLATE '
12. FUEL SUPPORT

... * . 'i .?c, **

J#'f. [. k,I*'

. )2 4. g .i'.'

c' i.i

  • PIECE i'.*' '*;.)4';; (~'1.." ' ., 1
    • b'
13. FUEL PELLETS  ; . -l., '..i . . '.' *; , # *..s.1 [hij gg .. . g - .' . [ . f. , . .
  • 14 END PLUG

. .,f,1 ;. ', g {. , *. ( * [ 4, jQ. T. ,- 4 ,g j, .

. . j, ..' . *y ;

15.C MANNEL {Le_g g ., 4 e ,s . , , h g - ?. [. . -

SPACER if , g ., I - .

f . , 1

16. PLENUM {  : . ,, '. .' g 3[7 ;
  • *Q ' -%g LJ N.v 9  ; u

./.

, ** /.

l - , .f. t% -

1 SPRING l!:

a '

E.

  • s '

t

~

R.y,Af e -q, ). .-Q&;  ;

GENER AL $ ELECTRICI. y *.. "'

[

.. N o .' '-

/ ) '

Q , . .

Figure 3-5.

&& @ en . GElll" Fuel Assemblies and Control Rod Module 3-16

NEDO-22171 b

P3 = 1 A H n 10 i SEPARATORS I I 2= A (1 a 0-3)

P CORE MAY WARGER UPPER PLENUM Pg = 1 x 10-3 f

b ANNULUS CORE P

4 j i y3 = 14 P12 P5= 0.5 P 3 P4 = E5 P3 "

R< 10~I E)

P ,= 0.5 Pg q[ NR TIE ME

'6 = 5P 5 jgp 1.0 P, If k< 10-153

' U lI S40E ENTRY ORIFICE P8 =P 7 RECIRC m KT p.P+P PUMP (2.5 a 10")J PUMP 8 4 8 (7.5 x 10 ) j i Pgg= (1/900)not P 10 k< 10'I0 *l

  • Pg= 1.0 (7.5 x 10 4)*

LOWER PLENUM

' ACCUMULATED PROSABILITY OF ONE PIECE

    • PROS AstLITY TO BLOCK 75% OF THE AREA (

OF ONE ORIFICE y Figure 3-6. Loose Piece Potential Upward Flow Path 3-17/3-18

NEDO-32171 l

4. LOSS-OF-COOLANT ACCIDENT ANALYSIS WITH A CRACK  !

g' IN ONE CORE SPRAY SPARGER 1

i

4.1 INTRODUCTION

A crack in the core spray sparger at the Brunswick 2 plant, located in the upper sparger near the T-Box to sparger veld, has been clamped. The structural integrity of the sparger and the intended cooling function of the spray system are not adversely affected by the presence of the crack. Therefore, no change in Emergency Core Cooling System (ECCS) performance analysis or Maximum Average -

Planar Linear Heat Generation Rate (MAPLHGR) limits is required. However, at the request of Carolina Power & Light (CP&L), conservative ECCS sensitivity analyses have been performed and are presented here.

This section describes the methods used to evaluate the MAPLHGR requirements to meet 10CFR50 Appendix K for the current Brunswick 2 operating cycle with a cracked core spray sparger. The potential effect of spray distribution on the Peak Cladding Temperature (PCT) of the limiting break size and a single failure r is discussed in Section 4.2. The phenomena involved and the inputs to the approved 10CFR50 Appendix K computer codes are discussed in Section 4.3, the results of analyses performed are given in Section 4.4 and 4.5, and the conclusions are presented in Section 4.6.

4.2 LIMITING BREAK SIZE AND SINGLE FAILURE ANALYSIS For the Brunswick 2 plant, there are no single failures for any break location (other than a core spray line break) that can result in less than one core spray system injecting water into the upper plenum above the reactor core.

I For a core spray line break, there are always at least three low pressure ECCS l

pumps infecting water into the reactor vessel, thereby ensuring that this break i

l is not a limiting event. For medium and large break sizes (which depressurize l relatively quickly), the most limiting failures are those that result in the least number of ECCS pumps remaining operable (i.e., injecting water into the l

reactor vessel).

l 4-1 I

NEDO-22171 The only two single failure candidates that are potentially limiting for s medium to large break sizes are:

A. Diesel Generator Failure - 1 core spray (LPCS) + 1 Low Pressure Coolant Injection (LPCI) + HPCI + the ADS operable; B. LPCI Injection Valve Failure - 2 core spray (LPCS) + HPCI + the ADS operable.

Since the High Pressure Coolant Injection (HPCI) is steam turbine powered, it is not a significant contributor to mitigating medium to large breaks which depressurize rapidly. Also, since the function of the Automatic Depressuriza-tion System (ADS) is to depressurize the reactor as a backup to the HPCI, it contributes little toward mitigating medium and large break LOCAs. Therefore, failure candidates A and B are limiting and each result in a dependence on only two ECCS pumps.

Per the plant specific LOCA analysis (Reference 4-4), failure candidate B (LPCI Injection valve f ailure) is limiting by a large margin because of the conservative modeling of counter current flow limiting (CCFL) at the fuel assembly upper tie plates. The calculation limits the coolant delivery or downflow from the core spray systems to the fuel bundles and further prolongs core reflooding by neglecting the water held back in the upper plenum.

Both single failure candidates (A and B) were re-examined for large breaks acknowledging the crack in one sparger. The limiting single failure, break size, and location were found not to change. This is because the calculated I

core uncovery and recovery times and the reactor depressurization rates are insensitive to changes in spray cooling heat transfer, and even with more realistic treatment of CCFL the failure candidate B yields (marginally) the limiting result.

! For smaller break sizes, the limiting single failure is the high pressure ECCS (HPCI) since the transient is a relatively slow depressurization event that is 4-2 I

NEDO-22171 dominated by the time required to either reflood the reactor with the high pressure system or the time to depressurize the reactor so that the low pressure systems become effective. Furthermore, the effects of CCFL in limiting coolant delivery to the core are not as large at higher reactor pressures. The small break LOCA transient is, therefore, insensitive to spray distribution because reflooding occurs very rapidly once any one or two of the six low pressure ECCS pumps begin injecting coolant into the reactor vessel.

Therefore, only medium and large break LOCA calculations have any potential for dependence on spray distribution, and detailed LOCA calculations need only be performe for large limiting break sizes with the current limiting single failure.

4.3 PHENOMENA INVOLVED IN THE ANALYSIS OF SPARGER PERFORMANCE The key phenomena involved in evaluating core cooling performance resulting from the injection of spray through the core spray sparger in the BWR are listed in Table 4-1. The analytical assumptions regarding these phenomena

('. which are important to understanding system performance and the predicted core cooling are also tabulated.

The approved Appendix K models include these phenomena, but the input assumptions used in the standard reload analysis are overly conservative.

The extent of this conservatism is evident from Table 4-1 in light of the realistic phenomena observed and tabulated. The bases for the first three of these realistic inputs are derived from a recently completed, jointly sponsored, large scale BWR safety research program between NRC, EPRI and GE (Reference 4-1).

The relevant phenomena do not depend on the distribution of the injected spray through the nozzles but on the injection of coolant into the upper plenum (Reference 4-1). Recently, the NRC staff has evaluated the issues related to the adequacy of the core spray systems in the BWR and their ability to distribute spray water to the core (Reference 4-2). This evaluation was 4-3

NEDO-22171

/ Table 4-1 KEY PHENOMENA RELATED TO CORE SPRAY COOLING PERFORMANCE Analytical Assumptions Used in Phenomena the Current Reload Analysis Realistic Assumptions Upper Plenum Conservatively assumed to not Pool of water present Inventory interact or contribute to core throughout transient reflood during LOCA transient assures coolant delivery to all fuel bundles (supported by Large Scale Tests)

Counter Current Saturated water in upper plenum Some subcooling and less Flow Limiting above core CCFL occurs. A residual pool of water remains during and af ter core reflooding. (supnorted by Large Scale Tests)

No CCFL breakdown Breakdown of CCFL shortly after spray initiation causes rapid reflooding (supported by Large Scale t' Tests)

~

Core heat Limited spray cooling after Steam cooling contribution transfer blowdown (Appendix K credit as much as 10 times greater only) than Appendix K spray cooling Decay Heat 1971 ANS + 20% specified by 1979 ANS (GE has submitted Appendix K a technical basis as a part of the Standard Plant docket which is based on the 1979 ANS decay heat correlation) s 4-4

NEDO-22171 in response to concerns that the core spray systems may not distribute any spray to certain regions of the core when injected into an upper plenum steam environment. The staff testimony in Reference 4-2 concluded that the spray distribution adequacy is not a safety concern because the coolant injected into the upper plenum will either disperse uniformly in a pool of water above the core or will flow to the lower plenum producing rapid reflooding.

Therefore, the current reload calculation using the plant specific LOCA analysis basis is applicable and conservative despite the presence of any crack (s) in the core spray sparger.

4.4 ANALYSIS RESULTS The current reload analysis for the limiting LOCA with the most limiting fuel type and exposure combination results in a calculated PCT of 2200*F. This is for 8x8R fuel at an exposure of 20,000 mwd /t and a MAPLHGR of 11.8 KW/ft.

Figure 4-1 shows the heat transfer assumed as a function of time (Curve 1) compared with the realistic heat transfer (Curve 2). A bounding calculation k (Curve 3) of the limiting LOCA with approved Appendix K models but with CCFL breakdown input based on observed large scale tests, and no convective core cooling prior to reflooding results in a maximum PCT of less than 1260*F at a MAPLHGR of 11.8 KW/ft. This result demonstrates that the current reload calculation is conservative by more than 940*F. No credit for steam cooling or the improved decay heat correlation are included in this calculation which would further reduce the PCT.

A comparison of the current reload analysis with the conservatively calculated PCI using CCFL breakdown is shown in Curve 4 at the bottom of Figure 4-1. It is clear from this figure that the overly conservative treat-ment of CCFL results in the unrealistically slow core reflooding time and high calculated PCT in the reload analysis.

4-5

NEDO-22171 4.5 BOUNDING SENSITIVITY CALCULATION At the request of CP&L,a simplified and conservative sensitivity calculation was also performed with the following assumptions: 1) no credit for upper plenum inventory, 2) no CCFL breakdown, and 3) no cooling contribution from the sparger with the crack before core reflooding. This analysis results in a bounding MAPLHGR reduction of 8.5 percent at 20,000 mwd /T exposure to meet the 2200*F Appendix K limits for all fuel types and exposures.

To perform this bounding calculation, the cooling contribution from the cracked sparger was ignored prior to reflooding. With only the core spray system with the cracked spray sparger operating, the spray heat transfer coefficient was set to zero. For the core spray system with the uncracked sparger operating, the spray heat transfer was set to one half the value used in the reload calculation (Reference 4-3). This is the approved assumption when one spray system is totally inoperable.

GE considers this calculation to be unrealistic and overly conservative but has performed this calculatior. at the requer.i of CP&L to quantify the sensi-tivities of the core spray performance to various assumptions.

4.6 CONCLUSION

S An analysis of one cracked core spray sparger in the Brunswick 2 BWR was performed utilizing the approved Appendix K evaluation models. The results of this analysis demonstrate that with CCFL breakdown (derived from a conserva-tive interpretation of recent large scale tests) the calculated PCT is at least 940*F less than the current reload circulation. Without CCFL breakdown the upper plenum inventory (pool of water) ensures adequate ecolant delivery to the core. Therefore, the current reload calculation is applicable and conservative and there is no basis to impose a MAPLHGR penalty on Brunswick 2 for the next or succeeding cycles.

l l

I l 4-6 I

NEDO-22171

4.7 REFERENCES

4-1. Refill-Reflood Program Task 4.4-CCFL/ Refill System Effects Test (30 F Sector) SSTF System Response Test Results, D. G. Schumacher, T. Eckert, J. A. Findlay, General Electric Co., NUREG/CR-2568, EPRI NP-1525, GEAP-24893-2, March 1982.

4-2. Transcript of Testimony before the Atomic Safety and Licensing Board For Long Isand Lighting Co., May 28, 1982 - Docket No. 50-322-OL. NRC Staff Testimony of Summer B. Sun on ECCS Core Spray.

4-3. SER, O. D. Parr (NRC) to G. G. Sherwood (GE), Review of General Electric Topical Report NEDO-20566, Amendment 3, June 13, 1978.

4-4. " Loss-of-Coolant Accident Analysis Report for Brunswick Steam Electric Plant Unit No. 2", NED0-24053. September 1977.

w.

4-7 I

NEDO-22171

= -

i RELOAD CALCULATION 2.5 l GCURVEil g

,_ .. I

= ..

2

.l ,,

1 -

i

! I i 1 i i e i i i

~100 -

STEA[ -

' ~1" l + SPRAY COOLING f e

r- -

i j j REALIETIC 2 -

l R:U'lVE 2}

l 1

l I il I .

25 - llI Illl l l I 2.

s u l #'

E 2 - lll l DOUNDING CALCULATION OF CCFL DREAKDOWN

=v5

i lIl1

~l __ f 0l L

s 8 I l

' T I

l uvv"F LIMIT

= -

l I l i l BOUNDING CALCULATION

,E l l l, l l OF CCFL SREAKDOWN

, RELOAD AN ALYSIS l lll l (CURVE 4)

' I l

l ,,,, s fi l i i 2m i

2w i

0 12 12 h d600 0 h TIME (seconds) [

- ACTUAL REFLOOO D3 suc) REFLOGD (EMI FOOL sueCOOLS (=tio ese) F D R ATED CORC 2*R AY 152A sec) (248. 3 se c.)

POOL FORMS (=45 rec) l CORE UNCDVERY D4.2 anc) l '

l' -Figure 4-1. Brunswick 2 DBA (Liriting LOCA) Analysis 4-8

NEDO-22171 APPENDIX A

( STRUCTUPAL ANALYSIS OF THE BRUNSWICK 2 CORE SPRAY SPARGER TABLE OF CONTENTS Page,

SUMMARY

A-1 A.1 DESIGN LOADS A-3 A.1.1 Weight of Sparger A-3 A.1.2 Impingement Loads (90* Deflection of Flow) A-4 A.1.3 Impingement and Seismic Load A-4 A.l.4 Pressure / Flow Loads A-5 A l.5 Nozzle Thrust Loads A-6 A.l.6 Thermal Expansion Mismatch Loads - Intact Sparger A-8 A.l.7 Thermal Expansion Mismatch Loads - Broken Sparger A-ll A.2 FLOW INDUCED VIBRATION - NATURAL FREQUENCY A-13 A.2.1 Flow Induced Vibration A-13 A.2.2 Natural Frequency A-13 A.2.3 Conclusions A-21 A.3 STRESSES FOR UNBROKEN SPARGER A-22 A.3.1 Pipe Stresses A-22

/ A.3.1.1 Seismic and Impingement A-22 A.3.1.2 Differential Pressure A-23 A.3.1.3 Mismatch Due to Thermal Expansion A-24 A.3.2 Nozzle Stress A-26 A.3.2.1 Nozzle Thrust A-26 A.3.3 Bracket Stresses A-28 A.3.3.1 Lower Bracket Stresses Due to Thermal Expansion A-28 A.3.3.2 Middle Bracket Stresses Due to Thermal Expansion A-31 L A.4 STRESSES FOR BROKEN SP/RGER A-33 l

A.4.1 Pipe Stress Due to Seismic Load A-34 A.4.1.1 Mismatch Due to Thermal Expansion A-37 A.4.2 Nozzle Stresses Due to Pressure Load A-38 A.4.3 Bracket Stresses A-39 A.4.3.1 Lower Bracket Stress Due to Seismic Loading A-39 A.4.3.2 Lower Bracket Stress Due to Thermal Expansion A-40 A.4.3.3 Middle Bracket Stress Due to Pressure Load A-42 A.5 CORE SPRAY SPARGER TEMPERATURE CALCULATIONS A-43 A.S.1 Temperature Difference - Bracket to Pipe A-44 A.S.2 Heat Transfer Coefficients A-46 A.5.3 Pump Head / Runout A-48 A.6 REFERENCES A-49 l A-i/A-li

NEDO-22171 p-

- Appendix A STRUCTURAL ANALYSIS OF THE BRUNSWICK 2 CORE SPRAY SPARGER

SUMMARY

This appendix contains structural analyses that support Sections 2.5.3, 2.5.4, 2.6 and 2.8 of this report and, also, the presentation to the USNRC on June 14 and 15, 1982 reported in Section 1.

Section A.1 contains the calculation of loads during normal plant operation and during the core spray injection event for input into Section A.2, A.3 and A.4 of this Appendix.

Section A.2 contains flow-induced vibration and natural frequency calculations that show that flow-induced vibration is not a problem for the intact sparger condition and for an assumed broken sparger condition.

O

\ Section A.3 contains the structural analysis of the core spray sparger in an intact condition. The stresses were found to be low during all identified loading conditions.

Section A.4 contains the structural analysis of the core spray sparger which conservatively assumes a 360* throughwall crack in the larger sparger arm at the T-box. The analysis ignores the effect of a clamp (or assumes no clamp is installed).

The stresses were found to be low during al1 identified loading conditions.

Section A.5 contains heat transfer calculations that determine the maximum (bounding) temperature differential between the sparger pipe arms and the shroud wall.

I l

A-1

NEDO-22171

- Because the calculated stresses were well below the allowables, it was deemed

- . unnecessary to calculate principle stresses and stress intensities. The material properties are given below for comparison purposes.

e Material - 304 SS e Temperature - 550*F e Material Properties (ASME Section III)

S, = 16.9 ksi (upset allowable)

S = 18.8 ksi y

S = 57.3 ksi 6

E = 25.75 x 10 lb/in.2 The structural analysis results of the intact sparger and the broken sparger

_ are summarized below.

SUMMARY

OF STRESSES FOR INTACT SPARGER Seismic Impingement Thermal Mismatch Bending Bending Normal Stress Stress Stress -

(1b/inr2) (lb/in.2) (1b /in.2)

Sparger Pipe 476 174 3319 Normal Shear Stress Stress (1b/in.2) (Ib/in.2) i Nozzle 369 101

( (Weld)

Lower Bracket 3845 954 (Plate)

Lower Bracket 3067 617 (Weld)

Middle Bracket 1422 31 (avg)

(Plate) l Middle Bracket 654 44 (avg) i (Weld)

A-2

NEDO-22171

SUMMARY

OF STRESSES FOR BROKEN SPARGER r.

Seismic Impingement Thermal Mismatch Bending Bending Normal Stress Stress Stress (1b /in. 2) (lb /in.2) (lb/in.2)

Sparger Pipe 2380 868 3316 Normal Shear Stress Stress (lb /in.2) (1b /in.2)

Nozzle 4267 4396 (Weld)

Lower Bracket 3031 752 (Plate)

Lower Bracket 5974 534 (Weld)

Middle Bracket 5210 158 (avg)

(Plate)

Middle Bracket 1800 224 (avg)

(, (Weld)

U A.1 DESIGN LOADS This section contains the calculation of loads on the core spray sparger during normal plant operation and during the core spray injection event. The loads are used in Section A.2 for natural frequency calculations and in Sections A.3 and A.4 for calculating stresses in the intact sparger condition and the assumed broken sparger condition respectively.

A.1.1 Weight of Sparger 3 1/2" sch. 40 pipe

~

o

-d f p, p = 488 lb/ft PiPe 0 3. 8

( =f _

(488) = 9.1 lb/ft A-3

NEDO-32171 5'

([ W water

( 12 62.2 = 4.3 lb/ft 1 !I' Wnozzles" W = 9.1 + 4.3 + 3 = 16.4 lb/ft

=

6[2 = L_32 lb/in.

A.l.2 Impingement Loads (90* Deflection of Flow)

F = PA = #

31/2 SCH 40 PIPE F oV D a

t= ,

p = 45.87 lb/ft @ 550*F ,

V = 5 ft/sec D= y2 ft

(' (conservative value -

more realistic value =

p c

is 43.5 f t/sec) c F_ , 45.87 (5)2 (4.0/12)

L 32.2 h d n f=11.87lb/ft=Mlb/in. v l

! A.1.3 Impingement and Seismic Load l

Impingement Only W = -1.0 lb/in. (upward) f (Sect A.1.1) j Seismic Only - Assume lg (conservative) l W,=W2 (1.0) W W = 1. 37 lb/in. (Sect A.1.2)

W, = 1. 37 - 1 (1. 37 ) = O_ lb /in . (upward)

W, = 1.37 + 1(1. 37) = 2.74 lb/in. (downward)

A-4 L

NEDO-22171 Impingement & Seismic W = -1.0 - 0.0 = -1.0 lb/in. (upward)

T WT = -1.0 + 2.24 = 1.74 lb/in. (downward)

Seismic loading (absolute value) is more severe than seismic plus impingement loading.

A.l.4 Pressure / Flow Loads Maximum Flow = 6000 gpm (Rated Flow = 4625 gpm) (Section A.5.3)

Q = 6000 gal / min x 60

  • 7.4 gal.

= 13.37 ft 3/sec e Maximum pressure in sparger arm AP = 28 psig @ 4491 gpm measured I6000h AP, = 28 l = 1Qd psig

\ggy }

e Pressure load on sparger segment F = APA A=fd g =f(3.548)2 = 9.89 in.2 F

max

= 50.0 (9.89) = 495 lb e Maximum nozzle flow The one inch VNC nozzle has the highest flow and will produce the greatest nozzle thrust.

t A-5

NEDO-22171

-- >- + s/is n 2 2 A1 = 7 (1.181 - 0.313 ) = 1.018 in.2 (min.)

h w 2 a A 2 " I (1.75 - 0.875 ) = 1.804 in.2 p = 62.2 lb/ft3 0 80*F

\

g = 64 gpm @ 4491 gpm

  • I- 7/8 = 641 = 85.5 gpm N l

< > 1.181 MIN W

max " 6 7 = 11.85 lb/sec

, - 13/4 62.2(1.018) = 2b_d1 ft/sec @ exit from nozzle max "

A.1.5 Nozzle Thrust Loads HEADER Y I PIPE I f

~ ~

-X - ,

e U J g

m h F z a - \ _

N Jf Fy 1 IN VNC NOZZLE A-6

NEDO-22171 F = APA + P

^

y 8 AP = 28 @ 4491 gpm test flow AP =

28 (6000/4491)2 = 50.0 psi @ mount A =

{d where d = 1.181 in (minor diameter of 1" strai,ht t internal threads) 2 A = = 1.095 in

{(1.181)

  • 8 (

V = =

=

25.05 ft/see @ exit from HEADER F

y

=

50(1.095) + 62.2 32.2(25.05)2 (144) (1.095)

= 63.93 lb F

z

^ + jd V

26.95 ft/sec @ EXIT FROM N0ZZLE

62.2 (26.95) (1.018)

, F, 32.2 (144) 9.92 lb i

4 A-7

h M -22171 A.l.6 Thermal Expansion Mismatch Loads - Intact Sparger iso

  • 113*

\ I 210-1/2 0 149-1/2'

\ /

t i CENTERLINE g gINLET /

CENTERLINE TEE CENTERLINE BRACKET BRACKET 24Q' 1J9 R

CEN'TE R LIN E CENTER LINE CENTERIINE BRACKET SHROUD BRACKET C

CENTERLINE CS PIPE

  1. [- , CENTERLINE 92;1/2

~-

BRACKET I

= ,1.5 = 94 in.

R, 2 R = = 91.25 in.

c 2 - 2 r

SHROUD TEMPERATURE = 550*F See Section A.5-1 CS PIPE TEMPERATURE = 200*F ,

. . AT = 350*F AR = oRAT For 304 STAINLESS STEEL, a = 9.6 x 10

-6 in/in. *F 186.5 R = = 93.25 in, at shroud-to-pipe interface 2

A-8

NEDO-22171 R

90

=

9.6 x 10-6 (93.25) (350') = 0.313 in.

(For 90' are)

FOR SE2iENT:

Assume AR =

AR90. ( -cose)

=

0.313 (1-cose)

O' l

1/2*

/

40-1/2* g

/

-51*

\ l CENTERLINE IN LET TEE 71 % (FIXED) k '

-77 1/2*

T 97 1/2*

AR 40

=

0.313 (1 - cos 40 *) = 0.075 in.

AR 71

=

0.313 (1 - cos 71') = 0.211 in.

AR 979

=

0.313 (1 - cos 97h') = 0.354 in.

AR = 0.0198 in.

-20 . - 0.313 (1 - cos 20 *)

AR

-51 0.313 (1 - cos 51') = 0.166 in.

i AR_77q. = 0.313 (1 - cos 77 *) = 0.245 in.

A-9

NEDO-22171

/

, /

' /

/

w /

/

R 8

sw Assume the AR is resisted only by each bracket support in turn:

AR =

(20 - sin 20) - (cos 20 - 4 cos e + 3) y ,

4EIaR R (2 0 - sin 20 - p cos 20 + 4u cos 0 - 3u) 6 E = 28 x 10 lb/in. R=R = 93.25 - 2.0 = 91.25 I

= [ (4.0 - 3.548') = 4.79 in.

( u = 0.2 (coefficient of friction) l 0

p , 4 (28 x 10 ) (4.79) LR (91.25)3 (20 - sin 20 -0.2 cos 20 + 0.8 cos 0 - 0.6) 1 y ,

706 AR (20 - sin 20 - 0.2 cos 26 + 0.8 cos 0 - 0.6)

W , ,, , 706 (0.075) = 131 lb

  • n 0.403 y , 706 (0.211) = 89 lb 71, 1.68 y , 706 (0.354) , 79 3 97h, 3.15 y , 706 (0.0198) = 2M lb

, 20

  • 0.0579 i

A-10

NEDO-22171

)

l W-51, 706 (0.116) = 110 lb l f

0.747 y-77h, , 706 2.04(0.245) = 85 lb A.l.7 Thermal Expansion Mismatch Loads - Broken Sparger o*

1. Assume break in sparger at T-Box during core spray injection. 80-1NPN a
2. Assume AT at sparger-to-

\

shroud = 350*F (maximum -

see Section A.I.6). yi',,

AR = = R AT

~- =

-6

= 9.6 x 10 in./in. *F

' ~~

(304-SS) ,,,,o,,

R = 93.25 in. at shroud-to-pipe interface AR 90* arc = 9.6 x 10-6 (93.25) (350*) = 0.313 in.

Assume AR O = AR90* arc ( - c s 0)

Take 70* bracket as 0 = 0*

Then: 40

  • bracket is at 0 = 40 - 71 = -30 *.

97

  • bracket is at 0 = 97 - 71 = 26 *.

AR

-30 . = 0.313 (1 - cos 30 *) = 0.0433 in.

AR 26

= 0.313 (1 - cos 26 *) = 0.0329 in.

A-ll

NEDO-22171

~

h-

/

/

w e /

nh ow Assume that the 71* bracket is fixed and that the corresponding AR is resisted by the brackets to each side of it:

D AR = El ( - sin 20) 4E c s 20 - 4 cos 0 + 3) y ,

4EIAR R (2e - sin 29 - u cos 20 + 4 u cos 0 - 3p) 0 E = 28 x 10 lb/in. R=R = 93.25 - 2.0 - 91.25 c

h I

=h(4.0 - 3.548 ) = 4.79 in.

p = 0.2 (coefficient of friction) 6 y ,

4 (28 x 10 ) (4.79) AR (91.25)3 (20 - sin 20 - 0.2 cos 20 + 0.8 cos 0 - 0.6) 706 AR

" (26 - sin 20 - 0.2 cos 20 + 0.8 cos 0 - 0.6) 706 x 0.0433

-30\* " (2n x 30.5/180 - sin 61* - 0.2 cos 61* + 0.8 cos 30.5* - 0.6)

= 168 lb 706 x 0.0329 26 * " (2n x 26.5/180 - sin 53* - 0.2 cos 53* + 0.8 cos 26.5* - 0.6)

= 190 lb (maximum)

~.

A-12

NEDO-22171 A.2 FLOW INDUCED VIBRATION - NATURAL FREQUENCY GE Design Basis requires that the natural frequency f is equal to or greater than three (3) times the vortex shedding frequency.

A.2.1 Flow Induced Vibration The vortex shedding frequency, f ' I'"*" I n

fy .D y = 0.21 V = velocity past shroud wall = 5 ft/sec (Conservative value - more realistic value is 43.5 ft/sec) 4.0 D = sparger pipe diameter = g ft 0.2 f

y

= ) = 3.15 Hz A.2.2 Natural Frequency

.. The sparger natural frequency is now calculated for four different cases. The first two cases are bounding for the intact sparger. The last two cases are for the broken sparger.

Case 1 - Intact Sparger Calculate the natural frequency of the sparger by examining the longest segment between support brackets. Assume this section has a uniform load w per unit length, both ends simply supported.

W K hhhh hhhhk f n"h wL d H K = 9.87 n

I 6

E = 25.75 x 10 lb/in. I = 4.79 in.'

L = 40.5 180 x v x 91.25 = 65.4 in. (distance from T-box centerline to the first bracket)

W = 1.37 lb/in.

A-13

l NEDO-22171 f

,9.87 25.75 x 106 (4.79) (32.2) (12) = 70.4 Hz n 2r 1.37 (64.5)'

Ratio = =

3 5>3 Case 2 - Intact Sparger (Missing Bracket)

Calculate the natural frequency of the sparger by examining the longest segment between support brackets ignoring an intermediate support. Values other than L are same as previous case.

L =

0 f = 70.4 = 22.9 p, Ratio = =

3

>3 v v Case 3 - Broken Sparger Calculate the natural frequency of the unsupported sparger segment. Assume the segment acts as a cantilever and has a uniform load v (force / unit length).

NOTE: Length of unsupported segment is taken 3" from T-box centerline to account for location of crack.

ASSUMED W k seco CRACK f

n" 2i wL 4 I c L k, = 3.52 6

E = 25.75 x 10 lb/in.

A-14

NEDO-22171 l I = 4.79 in.'

L = [4015 (w/180) x 91.25] - 3 = 64.5 - 3 = 61.5 in.

(M istance from crack to bracket) w - 1.37 lb/in. (Section A.l.1) f

,3.52 25.75 x 106 (4.79) (32.2) (12) = 27.6 Hz n 2r 1.37 (61.5)0 Ratio = = >3 3

v Case 4 - Broken Sparger (Include Effect of Two Near Brackets)

Rayleigh's Method:

(

(.

Tne total kinetic energy of the system is zero at the maximum displacement but is maximum at the static equilibrium point. On the other hand, the total potential energy is maximum at the maximum displacement but is zero at the static equilibrium point. From conservation of energy:

(K.E.) = (P.E.) max 2 2 W v Ey ,

= E W, where W = weight, lb y = deflection, in.

A-15 l .

NEDO-22171

( g = gravitational constant = 386 in./see w = natural frequency, radians /sec The deflection of the member is obtained by successive graphical integrations starting with the shear diagram.

The loading diagram is shown below. The reaction loads are calculated to develop the shear diagram. The slope at the left end (R y) is calculated to initialize the slope (dy/dx) diagram, c 40 IN.

Z 81 IN. r W = 1.27 LEAN.

R9 (BR ACKET) 11 if If 11 Il ll If I

[ ll il il il il il R2 (BR ACKET) (ASSUMED 300 CRACK)

LOADING 1* 37 (49)

IM y = 0 49 R2 = 1.37 (61) (49 + 61/2) 2 R = 102.0 lb 2

+ .

IM 2

49 R y

f61[

Ry = 85.6 lb 85.6 - 1.37 (49) - 102.0 + 1.37 (61) = 0 (Check)

Slope at R y:

w1 M E 2

+ = 2 M in.-R e

1 =h EI EI 2" 2 3

vi M i 3

+ 2549 (61) = 32630 lb-in.

i OEI = + =

24 6 24 6 A-16

NEDO-22171 The shear diagram is now constructed. It is then graphically integrated to construct the moment diagram, so -

so -

4o -

N 30 -

r 767 631 403 367 219 82 y 715 064 527 ago 263 M

38 4

-30 -

-40 -

(_ -ao - /

-so -

[85.6 f ~

368 to47 1643 21o1 2423 2164 1406 904 476 191

'301 i

' 658

% 716 b,

5 -tono -

5 1151 2

1379 1782 1906

-3000 -

2296 2549 A-17 l - -. . - _ _. _ _ - _ _ _ . .-

l 1

NEDO-22171 The slope (dy/dx) diagram is now constructed by graphical integration of the moment diagram.

l 1

ao.neo -

32A30 30,000 - 29,410 20.000 -

is.eso 10.000 -

c g I $

d N og 2sto

- o -

h $. $. $. $. $. $. $. $.

< s R H E 5 $ 5 i

-10000 - -

1s,soo

-20,000 -

(, -30,000 -

  1. D00 -

42,730

-e0,000 -

-e0,000 -

e4,370

-70,000 -

79.030

-e0.000 -

es o

-co.oco -

, q

    • ,Y70 ssj20 1

-100,000 -

l Graphical integration of the slope diagram yields deflections. The deflections must be adjusted so that the deflection at both brackets is zero.

l A-18

NEDO-22171

.. (1b-in. ) (in.)

Length Adjusted yEI (X100) y (10-3) (in.) y (10-2)

Left End Bracket 0 0 0 0 2790 2.26 9 1.91 5210 4.23 19 3.49 6260 5.08 30 3.91 5420 4.40 39 2.88 Bracket 2360 1.91 49 0 (Must be zero)

-3000 -2.43 59 -4.73

-10170 -8.25 69 -10.94

-18530 -15.03 79 -18.11

-27580 -22.37 89 -25.84-

-36970 -19.98 99 -33.84 Right End -47420 -38.46 110 -42.75 6

El = 123. 3 (10 ) lb-in.

Adjustment = 0.001 in/in. length 9

The deflection diagram is now constructed.

Y A-19 I . ._ - - -- - -- _ .

NEDO-22171 to 3.40 - 3El _12.3 lb I#' 13.7 tb 13.7 lb 12.3 lb o

I 13.7 lb 4.73 13.7 lb

~' ~

10.94 13.7 ab E

^

'7 te.11 o

_3o -

13.7 lb a

2524 13.7 lb 2324 15.1 tb

-40 -

42.75 The weight of the member is now distributed as shown above. The natural frequency can now be determined.

A-20

4 NEDO-22171

( Wy2 = [12.3 (0.95) + 13.7 (2.7)2 + 15.1 (3.8)2 ,

12.3 (3.5)2 + 13.7 (1.8)2 + 13.7 (2.1)2 ,

13.7 (7.5)2 + 13.7 (14.0)2 + 13.7 (21.7) +

13.7 (29.9) + 15.1 (38.0) ] (10-6)

= 0.04454 lb-in.

Wy = [12.3 (0.95) + 13.7 (2.7) + 15.1 (3.8) +

12.3 (3.5) + 13.7 (1.8) + 13.7 (2.1) +

13.7 (7.5) + 13.7 (14.0) + 13.7 (21.7) +

13.7 (29.9) + 15.1 (38.0)] (10-3)

= 1.778 lb-in.

"n = (386) 0b22 = 15425-9 v

n

= 124.2 rad /sec f = = = 19.8 Hz Ratio = = = 6.3 > 3 3

v A.2.3 Conclusions

1. The natural frequency of the intact sparger, even when a bracket is assumed missing, is greater than three (3) times the vortex shedding frequency. Therefore, fatigue resulting from flow-induced vibrations cannot be hypothesized as a cause of cracking. Also, the loads used to calculate stresses for the intact sparger condition (see Section A.3) do not require amplification.

A-2' l-

NEDO-22171 7 2. The natural frequency of a broken sparger (assumed 360* throughwall crack near the tee-box) is greater than three (3) times the vortex shedding frequency. Therefore, flow-induced vibration under this assumed condition will not be a problem and the loads used to cal-culate stresses for the assumed broken sparger condition (see Section A.4) do not require amplification.

A.3 STRESSES FOR UNBROKEN SPARGER This section contains the calculation of stresses in the intact sparger con-dition during normal plant operation and during the core spray injection event. During normal plant operation, there is no core spray flow. The sparger AP = 0 and AT = 0. Impingement loads and weight loads are low and are bounded by the postulated seismic event.

Design loads during core spray injection are the design loads for the sparger.

Thermal mismatch between the cold sparger (due to injection) and the hot shroud produce significant loads on the sparger pipe arms and on the brackets.

Pressure and thrust loads produce stresses in the pipe arm and nozzles.

A.3.1 Pipe Stresses l

A.3.1.1 Seismic and Impingement For simplicity assume continuous beam - three equal spans.

If1P If ififif1f U ifIf1f If if1f1f 1P lfifif JL JL JL JL 4 e  ; ; e ;  ; e y R = 0 40 wt 4 RB= 1.10 wt RC= 1.10 wt RD = 0 4C 0.40 wt 0.50 wt 0.60 wt SHEAR (Ib) 0.60 wt 0.50 wt 0.40 wt A-22

NEDO-22171

-0.10 wt 2 -0.10 wf 2 MOMENT (enAb)

> \ J \

+0.025 we N% /

%s-_ '/  %.

+0.Os wc 2 ., ,, ,;9 40.5 (Assume maximum span length 1 =

180 x v x 91.25 = 64.5 in.

for conservatism)

M = 0.1 (2.74) (64.5) = 1140 in.-lb o = I = 4.79 c= 'j 0

= 2.0 in.

o =

1140 (2.0) = 476 lb/in. (Seismic) max 4.79 o

(Impingement)

g. max =476(2.74)=173.7lb/in.

1 A.3.1.2 Differential Pressure LP = 50.0 psi R = 2.0 in.

Rg = 3.548/2 = 1.774 t = 0.226 nom.

t,g = 0.226 - 2 [(0.003) corrosion allowance] = 0.220 in.

e Hoop Stress PR g , i , 50.0 (1.774) = 404 lb/in.

r t 0.220 s.

A-23

NEDO-22171 ,

o Axial Stress a= = = 202 lb/in.

A.3.1.3 Mismatch Due to Thermal Expansion M

/ j

/ t m = WR sin 0 - p WR (1 - cos 0) /.

R

= WR { sin 0- p (1 - ces 0)] w Assume u = 0.2 R = 94.25 __

(See Section A.1.6) uw e 40

  • Bracket M

40 . = 131 (91.25) (sin 40.5* - 0.2 (1 - cos 40.5*)]

= 7191 in.-lb e 71* Bracket M71. = 89 (91.25) (sin 71* - 0.2 (1 - cos 71*)]

= 6583 in.-lb e 97b* Bracket M 97q. = 79 (91.25) [ sin 97 * - 0.2 (1 - cos 97h*)]

= 5517 in.-lb e -20

  • Bracket

-20 . = 241 (91.25) [ sin 20b' - 0.2 (1 cos 20b*)]

= 7423 in.-lb (maximum)

A-24

NEDO-22171 e -51' Bracket f

M

-51. = 110 (91.25) [ sin 51' - 0.2 (1 - cos 51*)]

= 7056 in.-lb o -77

  • Bracket M_7

. = 85 (91.25) (sin 77 * - 0.2 (1 - cos 77h*)]

= 6357 in.-lb Sum of Stresses Attributable to Core Spray Injection:

max " " bending + bracket friction + 4P

-f++f

(.

c= = 2 in., I = 4.79 in. , uW = 0.2 (241)

Ap = 2.68 in. , = 202 lb/in. (Section A.3.1.2) g , 7423 (2.0) 0.2 (241) + 202 max 4.79 2.68 l

f = 3099 + 18 + 202

= g lb/in.

l l

l A-25

NEDO-22171 A.3.2 Nozzle Stress A.3.2.1 Nozzle Thrust Y' y Y

l R

l 3 l 2

X - = ,

" Z l 1.76 + + l 1.96 r ,

L E'

Z O.12 V ARIES -

Fy I I ASSUME 15*

, 3.29 m  % 15' Weld Properties:

I = h (1.76 1.52') = 0.209 in.

( K = h (1.76' - 1.52 ) = 0.418 in.

2 (1.76 -

1.52 ) = 0.618 in.

A=f C = 1 76 = 0.88 in.

t = 0.12 in, r= = 0.88 in.

F = 64 lb y

F , = 9.9 lb (Section A.1.5)

P = 50 psi (Section A.1.4) u.

A-26

NEDO-22171 ,

7.U l ..

/ _

- The resulting loads at the veld are: -

t

~

F =F = 64 lb axial y .

T = F , = 9.9 lb shear z T = 3.29 F , = 3.29 (9.9) L32.6 in.-lb ,

torsion z -

s.

P M = 1.96 F , = 1.96 (9'.'9) =.19.5 in.-lb -

moment z . .

The stresses are censervat,1vely calculated 'as: -

Mm c F a

+. A- + _Pr o = 1 y I 2t .

64

= -+ 19.4 0.209 (0.8'81 + 0.618 +

50 (0.88).,

2 (0.12) .,, .

-: ~

/

~._ '

/

(- =

2 81.7 + 103.6 { 18L . s s,

, a  : - ,- . ,

N - . -

o = 369 lb/in.2 y . 205 lb/in.

,. e

- s f ,

- t' s Tc F- . ./~

T s - ~

t = +a a = 2 (thin vall cylinder) xy K A '

It, f '

> 'i- ,.:?

' w ~,.

' -- ' ~

=. (32.57) 0.418 (0.88) + 20.618 9.9 ,~,.

,s j '?

M -

= 60 + 32 ~ -._

,'~. .

n -r . / ., ,

/ W j

^

t = 101 lb/in. '!* '

e' '

s 4 j

  • 7 gr ,, ,,

w ,'.-

-i - - .,

h'.i 4 , ,.

s j ,

o w

! ],,

, , ,i%

n . , l>

t ** p *o

,'- , ,  % s ;n y

/

e*

j ,

's-

"l *

/

/

,.e* ,%

,s m .,

, , - , f ,

  • ~

'q /m .w= . J~ '

y ,

  • /*t j 'w
  • pi e

i be  %, .

, e *-

/r ,

f e 7' a, e

l- A-27 ,) *

~' C .,

,. .. ,r '

, ,p . l' ' .

s;

~~ r f,, 'p ,

-- p

, '/ ,  ; # e y*

4 4 ' N

NEDO-22171

,' A.3.3 Bracket Stresses s.

A.3.3.1 Lower Bracket Stresses Due to Thermal Expansion

[ (1 1

"'Jft n

T

,,  ;. F H L ()

r h U \ _ e ...

\ _ U ,

[ ,

LOWE R OR ACKET

('

t = 4.0 - 2.0 = 2.0 in.

L = 2.0 (1 + cos 30') = 3.73 in.

A'= j + (1 - sin 30') (2.0) = 2.0 in, b = 0.5 in.

! h = 0.25 in.

1 R = 241 lb Rz =uR = 0.2 (241) = 48.2 lb R =0 y

['

I' l

r (Section A.l.6)

(-

A-28

NEDO-22171 Maximum shear stress in the fillet weld is i

6 R, + R, +gT (b + h) g

    • I h1 (t
  • h) h where M = 1' R z 6 (2.0) 48.2
  • "T6 '(241 0.25 +(2.0) 48.2)
  • I (0.5 + 0.25) (2.0 - 0.25) (0.25)

T = 409 + 208 = 617 lb/in.

Maximum normal stress in weld is:

R d x z 2 , (b+h) ,36M max 2 ht ht (b+h) 2 h1 2

s.

where M = L' R x d (241) (241) max " I 0.25 (2.0)

  • 0.25 3 6 (2.0)2 (2.0) 48.2

~

I* )2 + (0.5 + 0.

2 25) 0.25 (2.0) (0.5 + 0.25) o = 341 + 2045 + 681.4 - 3067 lb/in.2 i

A-29

.j

NEDO-22171

- Maximum normal stress in plate is ...

R t' R C LR zC2x g . x, , x_.,,,,g ,

max A I I, A = 0.5 (2.0) = 1.0 in.2 I = bt } 0.333 in.'

xy " =

2 1

=

.0 = 1.0 in.

C xy 2 I, =

If2 1

= 0.02083 in.'

5 =

C ,

0.25 in.

zx 241 + (2.0) (241) (1.0) 3.73 (48.2) (0.25) o,,x = 1.0 0.3333 0.02083 C'

s.

- 241 + 1446 + 2158 o,, = 3845 lb/in.

Maximum shear stress in plate R

x + R, + l' R, (3t + 1.8b) bt g2b 2

, 241.0 + 48.2 ,(2.0) (48. 2 ) (3 x 2.0 + 1.8 x 0.5) 0.5 (2.0) (2.0)2 (0.5)2 t = 289.2 + 665 + 954 lb/in.

%e A-30

NEDO-22171 A.3.3.2 Middle Bracket Stresses Due to Thermal Expansion

/ f 30' 2.0

\A  ? O R., h

^ (; j h s h 8'

s

/

F y 4 F (PRESSURE LOAD -

SEE SECT. A.1.4) h@

~~ - . e

\/,-

/ /

$ Y Ryg

/  :

\

O ~ ~" )

L e) )

/

4 t '

L L = 10.25 - 2 (2.0) = 6.25 in. b = 0.5 in.

h = 0.25 in.

f t' = 6.25 + (1 - sin 30*) (2.0) = 4.13 in.

L = (1 + cos 30*) (2.0) = 3.73 in.

I F

=

(2.0 + 1.96) cos 15* + (1.50 + 1.18) sin 15' - 2.0 A-31

NEDO-22171 E

p

= 2.52 L

F

= 2.0 - (2.0 + 1.96) sin 15' + (1.50 + 1.18/2) cos 15' L = 2.99 in.

F (NOTE: 17Lp used in Section A.4.3.3)

Assume R xy =

R, = 241 lb (Section A.I.5)

R =

R, =

0.2 (241) = 48.2 zl Shear Stress R +y i

/2" \/ *1 ~'

E 96.4 T

AVG 2 ht 2 (0.25) (6.25) 1,yc = 44 1b/in.2 c ,1,,

(R +R )

1 *2 "

96.4 A

AVG bt 0.5 (6.25) i i

gyg

= 31 lb/in.2 (Bracket)

Normal Stress R +R R +R 6

I

  • 1 ht
  • 2 *1 h1 (b+h)
  • 2 2 + (b+h)2

't A-32

NEDD-22171

" /T 241 (2) 2 (48.2) *

( "2' O.25 (6.25)

  • 0.25 (6.25) (0.5 + 0.25)

~

$ 2 (3.73) + (0.5 + 0.25)2

= 218.1 1 436 o = 654 lb/in. , -218 lb/in. (WELD)

R +R (L - h) (R +R )C

  • 1 *2 *1 z2 bE I e = 0.25 in. I = 0.0651 in.'

, , 2 (241) . (3.73 - 0.25) (96.4) (0.25)

.g, 0.5 (6.25) - 0.0651

\_.

= 154 1 1288

( ACKET) o = 1442 lb/in. , -1134 lb/in.

i A.4 STRESSES FOR BROKEN SPARGER t

l This section contains the calculation of stresses during normal plant operation end during the core spray injection event. The analysis conservatively assumes a 360* throughwall crack in the longer pipe arm at the T-box and ignores the effect of a clamp (or assumes no clamp is installed). The analysis includes all the loading conditions identified in Section A.3. In addition, the stresses in the middle bracket due to the pressure load are calculated. (A nozzle will contact the middle bracket in order to axially restrain the broken sparger arm when it is pressurized). Stresses in both the middle bracket and the lower bracket (assumes lower sparger cracks) are calculated in order to bound the l

( condition.

A-33 l

l

NEDO-22171 A.4.1 Pipe Stress Due to Seismic Load Assume uniformly loaded beam with three supports, end moment M 3 and force P 3 i

and the 3rd support.

P 3

ll M'\ 2 #"2 I

= ig --

2 2

r3 7 El 2 A

3 using theorem of three moments...

3 3 IEl Eh2 ME 32 "1 E

l "2 E

2 M)1 11+ M 2 1 41 41 1 (T1 + 7)1 2 + 2 1 2

= 0 1 " "

My 1 I

2 "1 "2 "I +

(1 2

+

H 3 2 1 2) 2M ( 7j + R \

2 1 I 41 (t 1 +12)-M3 2_ _ _ A g .

2 (ty+1) 2 M3 is caused by the contilevered section of pipe between the support bracket and the break 2

wt M

3" 2 Likewise, P is caused by the cantilevered section 3

P 3 **3 For Seismic w = 2.74 lb/in. R y

=

2 615x 180 x 91.25 - 26.5 x 1.5926 = 42.2 2 013 x 1.59 = 48.6 in.

A-34

NEDO-33171

_ 1 = 40 x 1.59 = 64.5 in.

3 P

3

=

2.74 (64.5) = 177 lb M = 2.74(64.5)2 = 5700 in-lb 4 (42.2 - 48.6 ) - 5700 (48.6) 2 2 (42.2 + 48.6) i

= -809 in-lb 4.0 o = 1 - 4.79 in. c = = 2.0 in,

.0) 2380 lb/in.

o ,,x =

4,f9 =

T max

=

2f=2 8

= 150 lb/in.2 For impingement i

w f

l-1.0l = 1.0 lb/in.

P 3

=

177 (1.0/2.74) =

M lb.

l l M 3

=

5700 (1.0/2.74) = 2080 in-lb.

M 3

=

-809 (1.0/2.74) = -295 in.-lb.

o, = 2080 (2.0/4.79) = 868 lb/in.

! e Determine reaction loads for seismic only 3

"S IM 2

= 0 For segment 2: uj i M 3

l = ,, /

a2 a3 l

A-35

NEDO-22171 R = P "s 2+Ms "~

2

~.

3 3+ 2 2 2

, 177 , 2.74 (48.6) , 5700 , (-809) 2 48.6 48.6

= 377.5 lb "s

IM2 = 0 For Segment 1: 0 t t t 't I t i 2

'= t, =d Ry R,

"s 1 2 1 2 - I-1

, 2.74'(42.2) ,

I-809) 2 42.2 y

= 77.0 lb t

IMy = 0 For Segment 1, IM = 0 For Segment 2:

3 "s 2 "s 2+M1 2 3 2

2 1+T+ 1 2

2 E

2

, (2.74) (42.2) , (-809) , (2.74) (48.6) , (-809) _ 5700 2 42.2 2 48.6 48.6

- -28.7 lb Ry+R2+R3 = 426

> CHECK

= 6 (2.74) (11+E2+E) 3 a

I A-36

NEDO-22171

,-- A.4.1.1 Mismatch Due to Thermal Expansion s_

M = WR sin 0 - p'n'R (1 - cos 0) V u

=

WR (sin 0 - u (1 - cos 0)) m

9) \

h e Assume p = 0.2 R = 91.25 (See Section A.1.7)

V = coef, of friction M

-30.5

=

168 (91.25) (sin 30.5* - 0.2 (1 - cos 30.5'))

= 7356 in.-lb.

M 26.5

=

190 (91.25) (sin 26.5* - 0.2 (1 - cos 26.5'))

=

7372 in.-lb (maximum) max BENDING + # BRACKET FRICTION +

= + +

c = 4.0/2 = 2.0 in., I = 4.79 in.2 uW = 0.2 (190), Ap = 2.68 in.2 h = 202 lb/in. (Section A.3.1.2 - BOUNDING VALUE) g ,x , 7372 (2.0) ,0.2 (190) + 202 4.79 2.68

= 3316 lb/in.

A-37

NEDO-22171 A.4.2 Nozzle Stresses Due to Pressure Load Assume 360* break, nozzle loaded by bracket, f, 1 F = 495 L8 (SECTION A.1.4)

~

I p \

L tJ 19s

\ _ JL -

o.12 j+

/

/ 2.68 r

The resulting loads at the weld are.

F " "

SHEAR T

TORSION

2.68 F

2.68 (495) = 1327 in.-lb.

bOMENT = 1.96 F =

1.96 (495) = 970 in.-lb The stresses are conservatively calculated as...

o = +

7 P = 50 psig (see Section A.l.4)

, - 970 (0.88) ,50 (0.88) For Weld Properties, see 0.209 2 (0.12) Section A.3.2.1

= 1 4084 + 183 J

A-38 L

NEDO-22171 1

<- o = - 3901 lb/in.2 , 4267 lb/in.2 l

TT c F T = +, o= 2.0 k

, 1327 (0.88) + 2 0.618495 0.418

= 2794 + 1602 t

xy

= 4396 lb/in.2 A.4.3 Bracket Stresses A.4.3.1 Lower Bracket Stress Due to Seismic Loading n 1 2.0 "Y f av f "I 3, t M o 3

r b U k ,, e ...

U

\ .

h b L __

LOWER 9R ACKET 1= 4.0 - 2.0 = 2.0 in. L =

2.0 (1 + cos 30*) = 3.73 in.

E'= j + (1 - sin 30') (2.0) = 2.0 in, r~ b= 0.5 in. h = 0.25 in.

C A-39

- - . . _ _ _ _ _ _ _ _ _ _ _ __ __. J

NEDO-22171 R

y

= 377.5 lb R, = 0 R, = 0 (Section A.I.3)

(conservatism - uses highest bracket load at weakest bracket)

The stress in the fillet weld is ...

T AVG 0 25 ( 0)

= 534 lb/in.2 3dM 3d LR y BENDING 2 2 h1 h1

, 3 6 (3.73) (377.5) = 5974 lb/in.2 (0.25) (2) b.

A.4.3.2 Lover Bracket Stress Due to Thermal Expansion R

x

= 190 lb R, = uR x

= 0.2 (190) = 38 lb (Section A.I.7) R y

= 0 Maximum shear stress in the fillet veld is...

6 Rx + d M I h z*T (b+h) (R-h) (h) where M = t' R z

6 190 + 38 d (2.0) 38

(. T 0.25 (2.0)

  • T (0.5 + 0.25) (2.0 - 0.25) (0.25)

's . '

A-40

NEDO-22171 e t - 322.4 + 163.8 = 486 lb/in.

k Maximum normal stress in veld is ...

_d_

z 2L 2 , (b+h)2 36m 2 ht b ,ht (b+h) max 2 2 where M = t' R (190) max

, _d 0.25 2 (2.0) ,3 6 (2.0) (190) 0.25 (2.0)2 52.8

,0.25 (2.0) (0.5 + 0.25) 2 (3.73) + (0.5 +2 0.25)2

(:

o,,, = 268.7 + 1612.2 + 746.5 = 3360 lb/in.2 Maximum normal stress in plate is...

R l ' R, C, LRg C,x max A+ I xy I zx A =

0.5 (2.0) = 1.0 in.

= ,

0.5 2.0)3 I 0.333 in.

xy 2 C = = 1.0 in.

s.

A-41 L.

NEDO-22171

<2.o <0.5) c 1,, =

Rf = 1

= 0.02083 C

zx

=

j = 0.25 190 (190) (1.3) , 3.73 (38) (0.25)

, max 1.0 , (2.0) 0.3333 ,

0.02083

190.0 + 1140.0 + 1701.1 o

3031 lb/in.

Maximum Shear Stress in Plate R +R 1 R, (31 + 1.8b) x z T *

  • bi g2b 2

, 190 + 38 ,(2.0) (38) (3x 2.0 + 1.8 x 0.5)

(2.0)2 (0.5)2

= 228 + 524 = 752 lb/in.

A.4.3.3 Middle Bracket Stress Due to Pressure Load l (Refer to Figure. Section A.3.3.2)

F = 495 lb (Section A.1.4)

Shear Stress (Neglect Torsion - Small)

  • AVG 0.25 .25)

= 224 lb/in. (kTl,D) j- -

A-42 L

NEDO-22171

~ ~ =

T 158 lb/in. (BRACKET)

AVG 0.5 6.25)

Stress Due to Bending F

b 2 (b+h)2

' max h1 (b+h) F 2 495 0.25 (6.25) (0.5 + 0.25) 2 (2.99)2 + (0.5 + 20.25) o = 1800 lb/in. (WELD) 0.5 e ,x = c = = 0.25 in.

I = =

6.25 0.5)3 = 0.0651 in.'

r M =

(LF - h) F l

y ,

(2.99 - 0.25) (495) (0.25) = 5210 lb/in.2 (BRACKET) max 0.0651 -

A.5 CORE SPRAY SPARGER TEMPERATURE CALCULATIONS This section contains heat transfer calculations that determine the maximum (bounding) temperature dif ferential between the pipe arm and the shroud.

I q.-

A-43

NEDO-22171 A.5.1 Temperature Difference - Bracket to Pipe

,.s 2 ' D* ~

h 1

= 5145 Btu /hr-ft *F T*

h, = 390 Btu /hr-ft -

  • F (Section A.S.2) ,',

i l sf l l h," Tg k = 10 Btu /hr-ft *F t 9

Dg = 4.0 in, t = 0.226 in. D = 3.548 in.

1 Tg = vater in sparger = 80*F T, = vater outside = 550*F

.S' Ag = w i(1) = 0.9289 ft (1 ft long section)

\ y2 /

4.8 A, = r y2 (1) = 1.0472 ft (1 ft long section)

( A = n g) I (1) = 0.9880 ft 2 (1 f t ,long section) p Tne thermal resistance, R is 1 1 R l 1 t Ah 4 Ah ,Ak i i oo p T, - Tg 9 1 1 1 Ah +Ah 0 0 Akp

, (.

A-44

NEDO-22171

~

(TO -T)

A h, f AT ~

0 Ah

  • FIDI OUTSIDE 1 1 t Ah g Ag h, Ak p

~

A g1 0 1) 0 ~

FILM INSIDE 1 1 t Ahgf

  • A,h, + A k Ahgg +Ah +Ak 1 1 0.226/12

! (0.9289) 5145 + (1.0472) 390 + 0.9880 (10) x = 0.000209 + 0.002449 + 0.001906 = 0.004564 AT = 0.002449 (550-80) = 252*F FIU! OUTSIDE 0.004564 Outside metal temp = 550 - 252* = 298'F LT = 0.000209 (550-80) = 21.5'F FIU! INSIDE 0.004564 Inside metal temp = 80 + 21.5 = 101.5'F Average sparger (pipe) temperature =

298 + 101.5 = 200*F AT " - "

SHROUD TO PIPE

'w A-45

NEDO-22171

~

In practice, the core spray pumping system cannot inject into the reactor until the pressure reaches about 300 psia where T SAT = 417*F. In this case, the AT BRACKET TO PIPE is less than 337'F (417-80)*. Thus the above calcula-tion bounds the inadvertent injection case. It also bounds the case of core spray operation during LOCA for the same reason.

Conservatisms in the calculation are:

1. Bounding for reason described above
2. Assumes steady state conditions (Q, = pQ =Q) g
3. Neglects heat conduction to clamp from pipe.
4. Assumes runout flow.

A.S.2 Heat Transfer Coefficients A. Inside sparger arm (near tee box)

Assume AVG film temperature = 90*F (Check: (101.5+80)/2 = 91*)

D, = 3.548/12 ft + AFLOW =

(3.548/12)2 = 0.0687 ft p = 62.1 lb/ft Q R0 = 6000 gpm (Section A.5.3) v = 0.833 (10-5) fg2 /sec 3

w 797g = 6000 gpm (1 min /60 see) (ft /7.48 gal) (62.1 lb/ft3)

= 830 lb/sec FILM INSIDE "" PIPE

"#* 'E" #*

  • A-46

NEDO-22171

. w Long Side Arm

=

830 (100/360) = 231 lb/sec

(

l V =

g = 0.068 62.1)

=

54.1 ft/see Nu = 0.023 Re

  • Pr !

Re =

3.548 I 54.1 = 1.92 x 10 6

12 5 (0.833 (10 )j Pr = 5.20 0 90*F 6

Nu =

0.023 (1.92 x 10 )0.8 (5.20)1/3 - 4237

= Nu b where:

3.548 D = f 12 k = 0.359 Btu /hr-ft *F @ 90*F

= 59) = 5145 Btu /hr-ft *F

..h 3 5 12 B. Outside of Sparger Arm Assume Average Velocity is 2 ft/sec Assume Average Film Coef. Temperature = 425'F (Check: (550+298)/2 = 424*F) i:

(,- .

A-47

NEDO-22171

. Assume Heat Transfer is like a cylinder in Cross Flow s.

0.166 x 10 -5 2

v = gg /sec @ 425'F k = 0.374 Btu /ft-hr *F Pr = 0.927 hD 0

=

[0.35 + 0.56 (Re)0.5) Pr .31

= ft D,

I R

= = 4.02 x 10 5 12 (0.166) (10-5)

' 0 h, =

2)

[0.35 + 0.56 (4.02 x 10 )0.5] 0.927 31

/

^

. . h, = 390 Btu /hr-ft *F A.S.3 Pump Head / Runout Rated Flow = QR = 4625 gpm @ 128 psia Shut-off Head = 380 psia P = P SH - CQ 128 = 380 - C (4625) 1.178 x 10 -5

~

C = =

(4625)

@P = 14.7, i u..

A-48 L

NEDO-22171

- P -P SH Q R0 " = 5570 gpm 1.178 (10-5)

For conservatism, assume runout flow to be 6000 gpm A.6 REFERENCES Roark, R. J., " Formulas for Stress and Strain", McGraw Hill, Fourth Edition, 1965.

Hopkins, R. B. , " Design Analysis of Shaf ts and Beams", McGraw Hill.

" Machinery's Handbook" The Industrial Press, 16th Edition, 1959.

Blevins, R.D., " Flow-Induced Vibration" Van Nostrand Reinhold Co., 1977.

Kreith, " Principles of Heat Transfer", International,1969.

(~

Welty, et. al., " Fundamentals of Momentum, Heat and Mass Transfer" John Wiley, 1969.

Shields, C. M., Wade, G. E., " Core Spray Distribution Report No. 16, 218 Standard Plant", NEDE 13006-3.

GE Drawing 730E854, " Shroud".

C A-49/A-50

i NEDO-22171

- Appendix B FLOW VELOCITY CALCULATIONS l

This appendix describes the calculations for the flow velocities supporting statements in Section 3.4.2.1 of the text.

B.1 FLOW VELOCITY IN BYPASS REGION Assumptions:

6

1. The plant is operating at rated power (2513 MWt) and flow 77.0 x 10 lb/hr.
2. The flow in the bypass regions is homogeneous.
3. The bypass flow fraction is 12% (9.24 x 106lb/hr).
4. The water in the bypass regions is saturated.

u

5. There is no down flow in the bypass region. This assumption is discussed later.

There are two parallel flow paths in the bypass region--one is between the fuel channels, and the other is between the core shroud and the outermost fuel assemblies. The flow areas for these paths are shown schematically in Figure B-1. The simple analysis that follows will give an estimate of the relative flow velocity in the neighborhood of the spray sparger.

1 I

Path 1 is between the core shroud and the outermost fuel channels. The flow area along path 1 changes from Ay , between the bottom and top of the active fuel, to A at the t P guide to A immediately above the top guide:

5 6 Ay - 3758 in.2 ,

A 5

= 1172 in.2 ,

m B-1 e

NEDO-22171

^

> A 6

= 7157 in.2 s.,, ,

Path 2 is between the fuel channels. The flow area along path 2 changes from A2* A at 3 the top guide to A above 4 the fuel channels:

1 A

2

= 3118 in.2 ,

A = 1486 in. ,

3 A

4

= 20160 in.

From the geometry and the flow areas, the loss coefficient, K, for path 1 is approximately 1.0 and for path 2 is approximately 1.1. Therefore, for equal pressure drop:

(K Wyy) Kw 22 A A 5 3 r

(1.0)Wy (1.1)W2

~

1172 1486 Wy = 0.93 W 2

= 9.24 x 10 6, But, Wy+W2 Therefore, 6

Wy = 4.5 x 10 lb/hr The velocity is the bypass region between the core spray sparger and the fuel assemblies is then:

6 V =

Wy /pA5 + 4.5 x 10 /[3600 x 45.8 x (1172/144)] = 3.4 ft/sec.

The fluid in this region is primarily saturated liquid.

B-2 l

NEDO-22171

_ The upward fluid velocity in the periphery of the core bypass region is therefore conservatively estimated to be 3.4 ft/sec. The maximum sized piece that can be lifted from this vicinity in the upper plenum is approximately 0.4 inches in length (Reference 3-1).

B.2 FLOW VELOCITY AT BOTTOM OF BYPASS (TOP SURFACE OF CORE PLATE)

( Total Bypass) PA Since W

Total Bypass

=

9.24 x 106 lb/hr 9

A =

w/r (D' - Nd)2 D = inside diameter of shroud - 174.5 in.

N = number of control rod guide tubes = 137 f

d = outside diameter of control rod guide tube = 10.75 in.

o = density = 45.8 lb/ft Then V = (0.24 x 100 /(3600 x 45.8 x e (174.52 - 137 (10.75)2)jg4 x 344))

V =

0.70 ft/sec l

From Reference 3-1, a core spray nozzle cannot be lifted with this velocity and therefore the nozzle will remain at the bottom of the bypass.

i i N -

B-3

NEDO-22171 B.3 FLOW VELOCITY AT THE TOP OF THE FUEL ASSEMBLY HANDLES p-6 6 6 W

Total

= 77 x 10 - 9.24 x 10 = 67.76 x 10 lb/hr ,-

l A = na n = number of fuel assemblies = 560 *

/ '<

a = area associated with each fuel assembly a '(6)2 = 36 in.2 .

~

i The nominal single phase velocity is:

,' -J' '

V =

(WTM)/pA --

Therefore, 6

V =

(67.76 x 10 )/(3600 x 45.8 (560 x 36/144)) = 2.93 ft/sec.

I f' At this location, the fluid is a mixture of steam and water. Therefore, to estimate the lif ting force due to the mixture, a two-phase friction multiplier is used:

t m = 1+x(p /p - 1) - ' /

f g ,

-  ; )

where:

/

mass fl w rate of steam x = quality =

total mass flow rate 6 6 x = (10.5 x 10 )/(77 x 10 ) = 0.136 with, 3

p =

45.8 lb/ft f

l' -

' =

3 o 2.35 lb/ft g

B-4 t '.-

l NEDO-22171 4 = 1 + 0.136 (45.8/2.35 - 1) = 3.51 The total lifting force on a section of core spray pipe per unit length is:

F CD A of 4 M/(2g)

=

with C = drag coefHcient = 1.2 D

2 A = area (4.0 in. x 1 (ft/ft))/12 (in/ft) = 0.333 ft jgt Then the maximum weight than can be lifted is:

F = 1.2 x 0.333 x 45.8 x 3.51 x (2.93)2/ (2 x 32.2) = 8.6 lb/ft The buoyant weight of the sparger segment is about 11 lb/ft. Therefore, the f sparger pipe segment will not be lifted by the upward velocity lifting force.

w B-5 l

i

- - ~ - .

.4"

(

~ '

' ~

NEDO-22171 -

F .-

,f , <

s e .

y y- .., <

n P ,

I .- .

\~ , j r" p .n*

, / .

r

/ a '

_s ~

, i a

s .,

e.:

f

  • ,T

/ '

s' ~

J , .

  • * * -s s /

.. s A -

E' we

-- a l s  %

r*

f

  • 5 9 x

w -

m A - - A -

6 --

1r A -

, /

w- .

< - a j

$ s.

W

/ ,

ar

/

  • e .f' e/

, p , 1 -

M A 5

~

N T .e > >

m g) &

, , -. - p 1 >

,r m' ~

y

/

l f  %

.a . -

. /;  ? # _~.

r '* P ATH 1 <- PATH 2 m

- - am, . /

, -f ~ - ., ,. sm. .

, u /=

~~ ~

/,

/ /s y ...

f .,a r ,

4

. ~, ,

% s ~~ . .*/, ,

m ,, ,

,i - ~,

's -w ,

e eA_3 + *

< a%

  • _ ,

6 a  ! .* ,f A ,m-k s P l c.

.~ _; -

, e . ,

, s .

ti

~ - .

-,, u s .-

,. .- / , +

~

>,.v . j,~

v 1

i

, / '

s . # \

f a

J.m m i

./ , , -<

"i.; <

.r

,/ , f .

. ., . ..,J

/ <. - *

, e

. ,s l

~

_. a l

~ a 71 Sat r.' I: ~1.i Flow i.Pa th.s ,

, 1 -d

/ '*

  • ne ,,

.s m. ,

  • y Aw J p*ms3  %

",?m W* p n  %

l, r a

/

, .a-

- '" -- . ^___h_.__ _ _ _ __ _. ._ _ _ _ m_