ML20078H521

Rev 0 to Plant-Unique Analysis Rept of Torus Suppression Chamber, Mark I Containment Program Technical Rept for James a Fitzpatrick Nuclear Power Plant

ML20078H521
Person / Time
Site:	FitzPatrick
Issue date:	08/11/1983
From:	TELEDYNE ENGINEERING SERVICES
To:
Shared Package
ML20078H509	List: ML20078H507 ML20078H521
References
RTR-NUREG-0661, RTR-NUREG-661 TR-5321-1, TR-5321-1-R, TR-5321-1-R00, NUDOCS 8310140202
Download: ML20078H521 (144)

Text

'

NEW YORK POWER AUTHORITY 123 MAIN STREET WHITE PLAINS, NY 10601 1

TECHNICAL REPORT TR-5321-1 REVISION 0 MARK 1 CONTAINMENT PROGRAM PLANT-UNIQUE ANALYSIS REPORT l OF THE TORUS SUPPRESSION CHAMBER FOR JAMES A. FITZPATRICK NUCLEAR POWER PLANT JULY 27, 1983

)

hDR DO OO 33 P PDR WTELEDYNE ENGINEERING SERVICES 130 SECOND AVENUE

.WALTHAM, MASSACHUSETTS 02254 617 M 3350

Technical Report TN TR-5321-1 -ii-MM ABSTRACT The work summarized in this report was undertaken as a part of the Mark 1 Containment Long-Term Program. It includes a summary of the analysis that was performed, the results of the analysis and a description of 19 significant modifications that were made to the structure and internals to increase safety margins.

In all cases, the stresses reported in this document meet the allowable levels as defined in the structural acceptance criteria (Reference 3). The methods and assumptions used in this analysis are in strict accordance with USNRC MUREG 0661 (Reference 2), except as noted in the text. The modifi-cations described in this report are also in compliance with NUREG 0661, unless otherwise noted.

l l

Technical Re ort TR-5321-1 -111_ ENGtEERNG SERVICES TABLE OF CONTENTS Page ABSTRACT ii

1.0 INTRODUCTION

AND GENERAL INFORMATION 1 2.0 PLANT DESCRIPTION 4 2.1 General Description 4 2.2 Recent Modifications 4 2.2.1 Modifications to Reduce Hydrodynamic Loads 5 2.2.2 Modifications to Strengthen the Structure 8 3.0 CONTAINMENT STRUCTURE ANALYSIS - SHELL & EXTERNAL SUPPORT SYSTEM 33 3.1 Computer Models 33 3.2 Load Analysis 34 3.2.1 Pool Swell Loads 34 3.2.2 Condensation Oscillation - DBA 35 3.2.3 Chugging 36 3.2.3.1 Pre-Chugging & IBA C0 35 3.2.3.2 Post Chugging 36 3.2.4 SRV Discharge 37 3.2.5 Deadweight, Thermal & Pressure 38 3.2.6 Seismic 38 3.2.7 Fatigue Analysis 39 3.3 Results and Evaluation 40 i 3.3.1 Torus Shell 41 3.3.2 Support Col'umns & Weld to Torus Shell 42 3.3.3 Support Saddles & Shell Weld 44 3.3.4 Earthquake Restraints & Attachments 45 3.3.5 Anchor (Tie-Down) System 45

Technical. Report TE WE TR-5321-1 ~ -iv- NBtNG N,ES TABLE OF CONTENTS (CONTINUED)

Page 4.0 VENT HEADER SYSTEM 57 4.1 . Structural Elements Considered 57 4.2 Computer Models 57 4.3 Loads Analysis 59 4.3.1 Pool Swell Loads 59 4.3.1.1 Pool Swell Water Impact 59 4.3.1.2 Pool Swell Thrust 60 4.3.1.3 Pool Swell Drag Loads 60 4.3.2 Chugging Loads 61 4.3.2.1 Downcomer Lateral Loads 61 4.3.2.2 Synchronized Lateral Loads 61 4.3.2.3 Internal Pressure 61 4.3.2.4 Submerged Structure Drag 62 4.3.3 Condensation Oscillation - DBA 63 4.3.3.1 Downcomer Dynamic Load 63 4.3.3.2 Vent System Loads 64 4.3.3.3 Drag Forces on Support Columns 64 4.3.3.4 Submerged Drag (Support Columns) 64 4.3.4 Condensation Oscillation - IBA 64 4.3.5 SRV Loads 65 4.3.5.1 SRV Drag Loads 65 4.3.6 Other Loads - Weight, Seismic, & Thermal 65 4.4 Results and Evaluation 65 4.4.1 Vent Header - Downcomer Intersection 66 4.4.2 Vent Header - Vent Pipe Intersection 66 4.4.3 Vent Header Support Columns & Attachments 67 I

4.4.4 Downcomer Tie Bars & Attachments 67 4.4.5 Vent Header Deflector & Attachments 68 4.4.6 Main Vent /Drywell Intersection 69 4.4.7 Vent Header, Main Vent, & Downcomers - 70 Free Shell Stresses 4.4.8 Vent Header - Mitre Joint 70 4.4.9 Fatigue Evaluation 70

l I

cn a Report 1eT54 W E

-- NERING SERVCES TABLE OF CONTENTS (CONTINUED)

Page 5.0 RING GIRDER ANALYSIS 78 5.1 Structural Elements Considered 78 5.2 Computer Models 78 5.3 Loads Analysis 79 5.3.1 Loads Applied to Shell 79 5.3.2 Drag Loads 80 5.3.3 Loads Due to Attached Structures 81 5.4 Results & Evaluation 82 5.4.1 Ring Girder Web & Flange 82 5.4.2 Weld to Torus Shell 82 6.0 TEE-QUENCHER & SUPPORT 87 6.1 Structural Elements Considered 87 6.2 Computer Models 87 6.3 Loads Analysis 88 6.3.1 SRV - Load 88 6.3.2 Pool Swell Loads 88 6.3.3 Chugging Loads 89 6.3.4 Condensation Oscillation Loads 89 6.3.5 Other-Loads 89 6.4 Results & Evaluation 90 6.4.1 Tee-Quencher Structure 90 6.4.2 Submerged SRV Line 90 6.4.3 Tee-Quencher Support 91 7.0 OTHER STR'JCTURES 93 7.1 Catwalk 93 7.1.1 Computer Models 93 7.1.2 Loads Analysis 93

Tcchnical Report TR-5321-1 -vi-WTF1 WE ENGNEERING SERVCES TABLE OF CONTENTS (CONTINUED)

Page 7.1.2.1 Pool Swell Drag 94 7.1.2.2 Pool Swell Fallback 94 7.1.2.3 Froth Loads 94 7.1.2.4 Drag Loads (Support Columns) 95 7.1.2.5 Weight & Seismic Loads 95 7.1.3 Results & Evaluation 95 7.1.3.1 Main Frame 96 7.1.3.2 Support Columns & End Joints 96 7.1.3.3 Welds to Ring Girder 97 7.2 Internal Spray Header 97 7.2.1 Computer Model 97 7.2.2 Loads Analysis 97 7.2.2.1 Froth Load 98 7.2.2.2 Weight, Seismic & Ring Girder Displacement 98 7.2.3 Results & Evaluation 98 7.3 Vent Pipe Bellows 99 7.3.1 Analysis Method 99 7.3.2 Loads Analysis 100 7.3.3 Results & Evaluation 100 7.4 Monorail 101

'7.4.1 Computer Model 101 7.4.2 Loads Analysis 102 7.4.2.1 Froth Loads 102 7.4.2.2 Weight & Seismic 102 7.4.3 Results-& Evaluation 102 8.0 SUPPRESSION P0OL TEMPERATURE EVALUATION 109 8.1 Maximum Pool Temperature Evaluation 109 8.2 Pool Temperature Monitoring 110

Tcchnical Report TR-5321-1 -vii-1PTF1 WE ENGNEERNG SERVICES TABLE OF CONTENTS (CONTINUED)

Page REFERENCES 112 APPENDIX 1 - USE OF SRV TEST DATA IN ANALYSIS Al-1 APPENDIX 2 - USE OF 32 HZ CUT 0FF A2-1 FOR C0 & POST CHUG ANALYSIS

, APPENDIX 3'- DRAG VOLUMES FOR SUBMERGED STRUCTURE ANALYSIS A3-1 t

l l

l 1

i

{

Tcchnical Report _yjjj. TN TR-5321-1 N NES FIGURES AND TABLES Figures: Page 2-1 Torus Plan View 11 2-2 Torus Composite Cross Section 12 2-3 Torus Modifications - Cross Section at Ring Girder 14 2-4 Torus Modifications - Cross Section at Mid-Bay 15 2-5 AP Pressurization System 16 2-6 Vent Header Deflector 17 2-7 Vent Header Deflector Attachment 18 2-8 SRV Tee-Quencher and Support 19 2-9 Pool Temperature Monitoring System 20 2-10 RHR Return Line Elbow and Support 21 2-11 RCIC Line Modification 22 2-12 Condensate Drain Modification 23 2-13 Torus Support Saddles and Saddle Anchors 24 2-14 Thermowell Detail 25 2-15 Downcomer Tie Rod and Gusset Modification 26 2-16 Catwalk Modification 27 2-17 Catwalk Modification- 28 2-18 Monorail 29 2-19 Torus Spray Header Support Modifications 30 2-20 HPCI Elbow Modification 31 2-21 Thermowell Locations 32 3-1 Detailed Torus Shell Model 47 3-2 Detailed Torus Shell Model 48 3-3 Detailed Torus Shell Model 49 3-4 Torus Beam Model (360 ) 50

Technical Report #N TR-5321-1 -ix-FIGURES AND TABLES (CONTINUED)

P_ age 3-5 Pool Swell - Net Vertical Load , 51 3-6 Pool Swell - Average Submerged Pressure 52 3-7 Pool Swell - Torus Air Pressure 53 3-8 SRV Shell Pressure - Typical 54 3-9 ' Location of Maximum Shell Stress 55 l 3-10 . Earthquake Restraint System 56 4-1 Detailed Vent Header Model 72 4-2 Detailed Vent Header Model 73 4-3 Detailed Vent Header Model 74 4-4 Vent Header Beam Model 75 4-5 Vent Header Deflector Analysis 76 4-6 Chugging Cases - Synchronized Lateral Loads 77 5-1 Ring Girder 83 5-2 Detailed Shell - Ring Girder Model 84 l 5-3 Detailed Shell - Ring Girder Model 85 5-4 Detailed Shell - Ring Girder Model 86 6-1 SRV Line Analytical Model 92 7-1 Catwalk Computer Model (Unmodified) 104 7-2 Catwalk Computer Model (Modified) 105 7-3 Spray Header Computer Model 106 7-4 Vent Pipe Bellows Motion 107 7-5 Monorail Computer Model 108

Technical Report WTF1 A'E TR-5321-1 -x- ENGNEERNG SERVICES FIGURES AND TABLES (CONTINUED)

Page Al-1 SRV Test Instrumentation - Shell Al-6 Al-2 SRV Test Instrumentation - Tee Quencher & Support Al-7 Al-3 SRV Test Instrumentation - Vent Header Support Al-8 Al-4 SRV Test Instrumentation - Downcomers Al-9 Al-5 SRV Drag Pressures Al-10 Tables:

1. Structural Acceptance Criteria for Class MC 114 and Internal Structure

2. Plant. Physical Dimensions 115

3. Plant Analysis Information 116

4. SRV Load Case / Internal Conditions 117 8-1 Results of Pool Temperature Responses 111 I

l l

l

Technical Rep;rt W TF1 W NE TR-5321-1 MEN SBt\/ ICES 1.0 GENERAL INFORMATION The purpose of the Mark 1 Torus Program is to evaluate the effects of hydrodynamic loads resulting from a loss of coolant accident and/or an SRV discharge on the torus structure. This report summarizes the results of extensive analysis on the Fitzpatrick torus structure and reports safety margins against established criteria. The content of this report deals with the torus 'shell, external support system, vent header system and internal structures. Analysis and results for piping attached to the torus (including shell penetrations and internal piping), for the SRV line (except for the submerge.d portion and tee-quencher), and for the SRV line vent pipe penetra-tion will be presented in a separate piping report, TR-5321-2.

The criteria used to evaluate the torus structure is the ASME Boiler &

Pressure Vessel Code,Section III, Division 1, with addenda through Sumer 1977 (Reference 11) and Code Case N-197. Modifications were done under Section XI of the ASME Code and meet the Summer 1978 Edition of Section III for design, materials, fabrication, installation and inspection.

A great many technical reports have been written and issued as a part of this program. These reports provide detailed descriptions of the nhonomena.

the physics controlling the phenomena, calculational methods and detailed procedures for plant unique load calculations. Several of thcse documents are listed as references in this report. The approach of this report will be to reference these documents, wherever possible, and to avoid a restatement of the same information.

A major part of this program has dealt with providing plant-unique load calculation procedures (Reference 1 is an example of this). In most cases, the loads used to support the analysis were calculated in strict accordance with those procedures, as amended by NUREG 0661 (Reference 2). In some cases, optional methods have been used; these methods are specifically referenced in

WTwi prf(NE N$$'"""' ENGNEERNG SERVICES Program documentation. Examples of these are the use of plant-unique SRV test data to calibrate SRV analysis, and use of plant unique quarter scale pool swell movies to refine certain water impact and froth loads. In a few cases, analysis assumptions have been made that do not appear in Program documen-tation; these are identified in the text.

Extensive structural analysis was performed as a part of this evalua-tion. The major analysis was for dynamic response to time-varying loads.

Analysis for static and thermal conditions also form a part of this work. The computer code used to perform almost all of this analysis was the STARDYNE code, as marketed by Control Data Corporation (CDC). STARDYNE is a fully qualified and accepted code in this industry; details of the code are avail-able through CDC. Cases where a computer code other than STARDYNE is used will be identified in the text. All dynamic analysis used damping equal to 2%

of critical, unless stated otherwise.

As an aid in processing the large amounts of calculated data, post-processors for the STARDYNE program were written and used. These programs were limited in function to data format manipulations and simple combinations j of load or stress data; no difficult computational methods were included.

l The loads and load combinations considered in this program required l special, consideration to determine the appropriate levels of ASME Code appli-cation. Reference 3 was developed to provide this standard. Table 5-1 of Reference 3 is the basis for all the evaluation work in this report; it is reproduced in this report as Table 1. This table shows 27 load combinations that must be considered for each structure. The number actually becomes several times that when we consider the many different values associated with various SRV discharge conditions. The approach used in the final evaluation of structures is to reduce this large number to a relatively small number of cases by conservative bounding. For example, load combinations including SSE seismic, have a higher allowable than the same combination with 0BE seismic.

l

Technical Report W F W NE TR-5321-1 ENGNEBUNG SERVICES For these cases, our first evaluation attempt is to consider the SSE combi-nation against the OBE allowables. If this produces an acceptable result, those numbers are reported as final. This procedure results in many cases where safety margins are understated; this is the case for most of the results.

As an aid in correlating discussion of particular load analyses to detailed program documentation, most analysis described in this report has been referenced directly to a paragraph in the Load Definition Report (Refer-ence 1). This has been done by identifying the applicable LDR paragraph in-parenthesis immediately following the title of the load. This referencing directs the : eader to a more detailed description of the load than can be included in this report.

I

r Technical Report YE E TR-5321-1 ENGNEERNG SERVICES 2.0 PLANT DESCRIPTION 2.1 General Description The configuration of the Fitzpatrick torus structure is shown in Figures 2-1 and 2-2.

Figure 2-1 shows a plan view of the torus. It is made up of the sixteen (16) mitred sections, connected to the drywell by eight (8) equally spaced vent pipes. It is supported by two external columns and an inter-mediate saddle at each of sixteen places, as shown. The columns and saddles are connected to the basemat floor with anchor bolts. Four earthquake restraints, spaced equally around the torus, connect the belly of the torus to the basemat (Figure 3-10).

Figure 2-2 shows some of the inside arrangement. Ring girders reinforce the outer shell at each of the sixteen planes defined by the exter-nal support system. The vent header system is supported off of the ring j girders and is directly connected to the drywell via the vent pipes. The j opening where the vent pipe penetrates the torus shell is sealed by a bellows ~.

The ring girder also supports the spray header, SRV tee-quencher and a partial catwalk. Figures 2-3 to 2-21 show several details of the torus structure.

Table 2.0 lists several of the plant specific dimensions.

2.2 Recent Modifications Over the period of the past few years, many modifications have been made to the Fitzpatrick torus, both to increase its strength and also to mitigate the hydrodynamic loads. The modifications are illustrated and listed in the composite sections of Figures 2-3 and 2-4, along with their installa-tion dates. A description and illustration of each individual modification follows:

Technical Report TN TR-5321-1 N SBh/ ICES 2.2.1 Modifications to Reduce Hydrodynamic Loads Drywell Pressurization System ( AP System _),

Installation of a system to maintain a pressure differential between torus and drywell was the first modification of this Program. The system is illustrated in Figure 2-5. It is designed to maintain a minimum positive pressure difference of 1.7 psi between the vent system (drywell) and the airspace inside the torus. The result of this pressure difference ( AP) is to depress the water leg in the downcomers and reduce the water slug that must be cleared, if rapid pressurization of the drywell occurs. Early generic test'ing in the Program demonstrated that this was an effective means to reduce shell pressures related to DBA pool swell. The 1.7 psi pressure difference was selected as the basis for the Fitzpatrick plant unique quarter-scale pool swell tests and is intended to be the normal operating condition of the plant.

As illustrated in Figure 2-5, pressure differential is maintained by using the nitrogen inerting system to pressurize the drywell to 1.7 psi; the torus remains at ambient pressure. No pumping or additional piping is involved.

Vent Header Deflector The vent header deflector at Fitzpatrick is illustrated in Figures 2-6 and 2-7. It is a 30-inch pipe with a one-inch wall.

The deflector extends under the belly of the vent header to protect the vent header from direct water impact during pool swell. It does this by shadowing the most sensitive part of the vent header and by separating and diverting the rising pool before it can reach the vent header. This deflector was included in the plant unique pool swell tests for Fitzpatrick to provide accurate vent header loading for detailed analysis.

Technical Report TE WE TR-5321-1 DJGPEstNG SSWICES SRV Tee-Quencher A tee-quencher has been installed at the ischarge end of each main steam relief lino to replace the existing ramsheads. The quencher and its support is illustrated in Figure 2-8. The quencher serves to divide the SRV discharge bubble into hundreds of smaller bubbles and to distribute them over an entire bay. This division and distribution of SRV discharge has been shown in generic testing to reduce torus shell pressure by factors of two or more when compared to ramshead pressures. The plant-unique character-istics of these devices at Fitzpatrick were determined by in-plant testing after their. installation.

The quencher support is also illustrated in Figure 2-8. It is a 20-inch schedule 120 pipe welded to the ring girder, as shown.

Temperature Monitoring System & RHR Return Lines, The addition of a pool temperature monitoring system and an elbow to the discharge end of the RHR return lines are both intended to assure proper operation of the SRV quenchers. These modifications are illustrated in Figure 2-9, 2-10, 2-14 and 2-21.

l The temperature monitoring system se.nses pool temperature through a set of 16 thermowells set around the inside wall of the torus (Figures 2-14 and 2-21). They are averaged to provide pool bulk temperature.

The system is hard wired to a console in the control room.

The elbows on the RHR return lines were added to provide pool circulation during periods of extended SRV blowdown. Circulation of the pool with these lines assures that local-to-bulk temperature differences will be minimized and. that SRV quencher performance will be maintained ~ during extended discharge. These two RHR return lines were further modified by re-

Technical Report WP WE TR-5321-1 ENGREERNG SERVICES routing them to the ring girders. The ring girders react drag loads on these lines and also provide for reactions due to elbow discharge loads (Figure 2-10).

Additional SRV Vacuum Breakers Each of the 11 SRV discharge lines at Fitzpatrick has been fitted with one, ten-inch vacuum breaker, in addition to the original small vacuum breaker. This modification minimizes the temporary formation of the high water leg in.the SRV line which could occur after an initial actuation; and thereby limits the high clearing loads which could occur if a second actuation occurred at that time. The location of these devices is different on each SRV line due to space constraints and is not illustrated.

Modification of Submerged Piping Early analysis showed that some of the submerged piping inside the torus might experience high drag loads which could. produce unaccept-able stress levels. Three such lines were identified; they are l 1. RCIC return lines (eight-inch diameter).

2. Condensate return lines (ten-inch diameter).

3. RHR return lines (described previously - two lines).

l Two cases required rerouting; all required additional sup-port to the ring girder. The modifications are illustrated in Figures 2-10, 2-11, and 2-12.

Early analysis also showed that the mitre joint on the HPCI return line might reach unacceptable stress levels. This was a direct conse-quence of the high stress intensification associated with the fabricated mitre joint. The line was modified by the addition of an elbow in place of the mitre. The modification is illustrated in Figure 2-20.

Technical Report YE NE TR-5321-1 ENGeEERNG SERVICES In addition, the eight vent (one-inch diameter) drain lines were cut off and capped above the pool.

2.2.2 Modifications to Strengthen the Structure Torus Support Saddles and Anchor Bolts Support saddles were added under each ring girder as shown in Figure 2-13. The saddles, support columns and ring girder all lie in the same plane and react til vertical loads on the torus - most of the load is reacted by the saddle, The , addle is constructed of 1 -inch SA 516 GR 70 steel plates, welded to the torus shell and resting on the concrete basemat. It is restrained from upwaro motion by six pairs of two-inch Williams rock bolts, set 24-inches into the basemat. The anchor bolt restraints are set with a small clearance to allow for normal radial growth of the torus due to tempera-ture changes. Sliding of the saddle is accommodated by lubrite plates between the saddle and basemat.

i Column to Anchor Bolts In addition to the six pairs of anchor bolts that attach to the saddles, there are four anchor bolts attached to each column base plate.

These bolts are identical to the saddle anchors and are set with the same 24-inch embedment.

1 Downcomer Tie-Rods and Vent Header Gussets _

The downcomer tie-rods and vent header gussets are illus-trated in Figure 2-15.

i

Technical Report TN TR-5321-1 6M The tie-rods are constructed from ? -inch schedule 40 pipe and provide greatly increased capacity to downcor.er lateral loads than the original tie bars. They are attached to the downcomers with specially fabricated 24-inch pipe clamps, constructed from 3/4-inch steel. The clamps are prevented from sliding on the downcomers by wilded steel clips both above and below the clamp.

The . gussets between the downcomers and vent header are necessary to reduce local intersection stresses due to chugging lateral loads on the downcomers. They are constructed from -inch thick 516 GR 70 steel plate, and are welded to the vent header and downcomers.

Catwalk Modification The catwalk at Fitzpatrick would have required major modifi-cations to avoid overstress during pool swell impact and fallback; the extent of the required modifications was considered unacceptable. Accordingly, the entire catwalk was removed except for three bays, centered under the equipment access hatch, as illustrated in Figure 2-16. The center section of the remaining catwalk is in a non-vent bay and includes a small platform that extends toward the center of the torus. Modifications to this part of the catwalk include:

e Addition of four new diagonal supports back to the ring girder (three-inch schedule 80 pipes).

e Replacement of the support for the extension plat-form.

e Replacement of the handrails with wire rope to reduce drag loads.

Figures 2-17, 7-1 and 7-2 also illustrate the catwalk struc-ture.

Technical Report TN TR-5321-1 MM Internal Spray Hedder Modification The internal spray header at Fitzpatrick is located above the vent header as shown in Figure 2-3. It was originally supported at every second ring girder (eight places around the torus) with full anchors. The system was modified by the addition of an additional restraint at each of the intermediate ring girders. These new supports provide lateral and vertical restraint but allow rotation and axial motion.

The original installation also included supports on the spray header branch lines near the penetrations; these were not modified.

The modification is illustrated in Figure 2-19.

Monorail A high capacity monorail was installed inside the' Fitz-patrick torus for use in handling materials. The monorail extends completely around the torus except for two bays where piping systems are installed. It l 1s illustrated in Figures 2-18 and 7-5.

SPTELEME Tcchnical Report TR-5321-1 _11 NM EXTERN AL SUPPORT COLUMNS

- l

\~'

29 '- 6 "

/O s

"^-" - 0\

l

/ &,

I VENT PIPE I

\Q.%

, PENETRATION e

g l a I

's,

/

l FIG. 2- 1 TORUS PLAN VIEW

WTELEDGE Technieal Report ENCNEBWiG SERVICES TR-5321-1 VACUUM BREAKER LINE l l_

s-fe'

/

/ /

l h

VENT HEADE

,,SRV

/ ,,

N LINE 1

__ __4.

p_y _ _ ' VENT gl

\,

a s. u 1 g DOW NCOM ER 1 BELLOWS

( [% \ '

'lj hCATWALK

/ (PARTIAL) i

! DEFLIEC TOR

/ f l

1 l - - i ,

tsc= l

\ / /

! _ t' l

_QUENCHE 4

~(x \- -

4 -

e SUPPORT COLUMN ffy ., e .

W 6  % ~4'_ J dg l 3jf f h5 k hhhNI 0$db

!$$(p  %

,0,yn;g#4@h did)g!hEG[h I fhi h

SUPPORT SADDLE 3 e .

e u2: + w +aa w e e)+

aw, e e :s .; - y 9 .

, .D -

' ^

ib,*: .

\ '..Q l....-.

( a-[.

E = MODIFIC ATIONS FI G . 2- 2 TORUS COMPOSITE CROSS SECTION FITZPATRICK

Tcchnica.1 Report WM TR-5321-1 M6 KEY FOR FIGURES 2-3 AND 2-4 Modification Completion Date

1. Mitre Joint Saddles 8/80

2. Saddle Anchor Bolts 8/80

3. Downcomer Tie Straps 8/80

4. Vent Header Deflector 8/80

5. Vent Header /Downcomer Stiffening 2/82

6. Monorail 6/80

7. Catwalk Strengthen & Partial Removal 2/82

8. Ring Girder, Drain Hole Complete

9. Drywell/Wetwell AP Control 11/77

10. Temperature Monitoring System Complete

11. Safety Relief Line, Vacuum Breakers & Supports 12/81

12. SRV Tee-Quencher 12/81 [

13. Tee-Quencher Support 12/81

14. RHR Return Line Elbow 12/81

15. Modify HPCI Line 12/81

16. Relocate RCIC Line 8" 12/81

17. Vent Header Drain Complete

18. Relocation of 10" Condensate Line Complete

19. Spray' Header Support Complete

WTE. EDGE l Technical Report BGEggeQSSMCES TR-5321-1 lg l

l W

l f er _

r .

s yf)

_, _ _._ -- l

\-

\ ,/ /

c

)M "

X 3

Ml  :

i3 L p g m .

P & 8 9

. h I$if?{E@ggg- .

7.

] j)f g *

[ .

3

'.?5 5 *,'. i f N[,,'a, i'?!: $. *$ ??.E $f-

. : . ' .' .t. m6 E = MODIFIC ATIONS g FIG . 2-3 FITZPATRICK TORUS MODIFICATIONS -

_ __ ._ CROSS-SECT:0N AT RING GIRDER

l WTELEDGE Tcchnical Report NN TR-5321-1 '

9 I l 15 16 p

@ l

1. 3 >

5 t

. j Q

~

C/ -

3 _ o _= __

/

I 17 wj 3 = MODIFICATIONS FI G. 2-4 FITZPATRICK TORUS MODIFICATIONS CROSS-SECTION AT MID BAY

WTELEDGE EN(NEERNGSERVICES Technical Report TR-5321-1 I l.7 PSIG- PRESSURE MAINTAINED BY N I N ERTIN G SYSTEM l l

A l

l

_4_

O PSIG l s I

^

\

l /%

l

\

= F -

=

WATER LEVEL I

J au. M. %. . J. . . . ,,4, ,_

k".Y! l l l l l_ 0 *

,]-[,'f FIG.2 - 5

WT sa

?X

,L a_'.

30" DI A. PIPE x X I" WALL 4 VENT HE ADER > "

d RE-INFORCING COLL AR <

h ( EXI STI N G ) s -

'3" /8

> 3,7 N s N

~

f s'-ob \

_ f - h REF.

f 6'- O z REE M' - . - . ,

( 7

/ _ _=._ b e 20'-9 r REE FI G . 2 - 6 ll VENT HEADER DEFLECTOR

WTERME Technical Report NM TR-5321-1 .

VENT HEADER RE-INFORCING COLLAR (EXISTING )

_ . _ _/

N i /

s'

~ - -

vT

's f

"3 '

/sV A NTI - ROTATION i e 30" SCH 120 PIP E l LUGS g! ,

j t --

j

=- -

-TORUS WATER LEVEL

. . - _ - ~ ~ ~ - ' ~ ' -- ~^ '- . .~

FIG.2-7 VENT HE ADER DEFLECTOR ATTACHMENT

~

WTERME Technical Report ENWWW8G5BMCES TR-5321-1 4

~

hN 8

15 / k 0  : ~.l - + < .:

._ 8 l I'

.  ? .

l 's

- ,f l \

_q _ , _ _ _ y l

% l F

l

• 12" TYPE 316L ARMS

~

j 1: I!!!!

f i:

~

\

li/ !!!!

IV

.I I I \

V

i. ii!!" 'il

\

T it i

• I:  ::::  : I  :  :::: *i :  :::: :t

Il.

1 l i::::

: I: 1  :::: , y% i  ::::< ' 11 :  :::: , :I  ::

! I a-3gh 7 7 _

kf- [l' j![]

f_.

1 ----- ~ T_==.a m.

/

if ,

20" SCH 120 PIPE

- . 1T/  : TO NEXT RING GIRDER y-RING GIRDER i - ____ . _ ._..._ _ _ . _ ._

( )

TORUS l

FIG.2-8 SRV TEE QUENCHER 8 SUPPORT

7 WTELEDGE Technical Report NSERWCES TR-5321-1 DATALOGGER CONTROL DISPLAYSAVEPKE DUPLICATE R00h (BULK) TEMPERATURE DISPLAY h Th URE PNEL 25-09

.a 80

'RTD e DUAL ELEMENT RTD 16 TOTAL, ONE PER BAY L

l FIGURE 2-9: POOL TEMPERATURE MONITORING SYSTEM

WTELEME Technical Report TR-5321-1 N S il M C E S I

EX I STE D j

c6 '

GII- N C s /s 75 N

N s NEW PIPE N

A RING GIR DER OII- NC

/pg.'  :

lU s\ WATERLINE m - g,

's Q. % -

IRTl l / III- N C U A ,5

,e Scs e eieE g 75 l

l l Y i

FIG.2-lO t

RHR RETURN LINE ELBOW 8 SUPPORT

MM Technical Report esGMggeSEMCES TR-5321-1 LBlf m -NC)

) /\

C 75*

EXISTED o

n' TORUS SHELL-'s f

, NEW PIPE RING GIRDER ,

g(Ill- NC s e

%J h

,/ . ',

WATER LINE i

...-n-~**'~ ~_., .

{

10 SC H 80 PI PE

\. .

FI G. 2 -i l RCIC RETURN LINE MODIFICATION _

WM Technical Report ENCNEBW4GSBMCES TR-5321-1 TORUS SHELL

'7

/l' s E XISTI N G PA D

IO"SCH 80 PIPE

- W6 X 15 BEAMS W ATER LIN E I / CUT TO SUIT y=

/x '

l _ = p=.======.======

I l

"U" B O LT l

l EXISTING L PAD 1

RING GIRDER FIG. 2- 12 10" CONDENS ATE DRAIN LINE

p WTBEME

, MM Technical Report ,

5/8 TR-5321-1 g LUBRITE 5.

c j SURFACE (SLIDES)

_f? ) N j - C1

[  ; /I/ ! ~ ' II~

I%l

~

g s --

fr W W bk ,

.: T ., .

,q a; r

l [ '

% CLEARANCE i 9 'Y.'

~-  !: .

TO ALLOW

. y ,. - l Jv SLIDING e w  ;'

'..~ -

p.:

,- 1 .<.. .- e '. S. '

- JL: E. .

/L V -

SECTION A- A RING GIRDER ss

/

1 1 PLATE '

s 2 t i

,N GUSSET TORUS ;

\ STIFFEN E RS

/

l i.

l ~.. _.'

A-+-

! r  !

I

___._4__..g_,.___.__..________ ._a, . . _ _ _ _ . . _ _ . .

, gy

~'

=I

_E3 t _ __ _ . ; i-- - - P

,. i ~~_J .

. : .' \

lY, , *.;, .

\

., i f , r >- r"i 1

l 3 f 12 -24" LONG 2" DI A. ANCHOR A-BOLTS ~ PER SADDLE FIG . 2 -13

WTELEDGE N!M!RVICES Technical Report TR-5321-1 13 REF.

\ s, s

\ 'N

'(

TORUS SHELL 3\

,5 8 l REF.[ c'. \

N.

CONAX B ARE WIRE THERMOCOUPLE GLANDS MODEL TYPE C P

FIG. 2- 14 THERMOWELL DETAll

Rep rt TR- 3 1-1

-VENT HE ADER w

MT

/

i e '/4 k /

/

I/4 V N l 4 ' 9" l. D .

,u

- PL ATE TYP.

2 1

4 h DOWNCOMERS

! t V4 A l

1

/ I/4V i

WATER LEVEL \ -

2"SCH 80 CLAMPS SCH 40 4" X {" TH K.

/~2 /

w 4 n 3

[ V SUPPORT BLOCKS (8 PER DOWNCOMER )

FIG. 2-15 DOWNCOMER TIE- ROD a GUSSET MODIFICATION

WTELEDGE ENWEERNGSEMCES Technical Report TR-5321-1 3/4" Wire Rope Nl %!

'D '

N 1

Grating Nh'N k Removed &

Stored During Plant

, Operation 1

\

\

-~

/

1

\\ 1 L

) '

y New 3" ,

Sch. 80 Pipe

\ p ruts (4)

/ k 1

i Modified Supports -==e -'-

ing Girder i

,a Pad On t t

Torus Shell V

i FIG 2-16 CATWALK & HANDRAIL MODIFICATION I

- -v,,----r- - - ,.,

WTRE30E ENWEBM3SSMCES Technical Report TR-5321-1 =a 5 13 " REF

- - _ _ _ =_ - _- __ _ _ _ _ _- ,__ _,

T}i i l

l 6a I- >

E XlST ED (m- N F- MC

\

d 1//

4 M r m - N F - MCH

\

'I/4 V

/

l

, EXISTING RING GIRDER

& FLANGE N x l

l TORUS SHELL 1

4 3" SCH 80 PIPE l III-N xF-MC)EE pq

'I//

4 l l

l1l >

l I!Id EU 4 -NF-MC

/

4; I/4 V l

FIG. 2-17 easesnuu>e_maaunaaswan _

WTELEDN!

Technical Report TR-5321-1 .

l l

4 il d n n RING GIRDER l P 4'-

L W N I

9 V '

l, /s i'j! ,

CONTINUOUS 315*

1  : AROUND TORUS i ,

sis x 54.7 1

FI G. 2- 18 MONORAIL

l WTREME Technical Report NMRVICES TR-5321-1 TORUS SHELL 1

/ TOP T

RING GIRDER

/

M RING GIRDER h ,

96 V k. ,

\

l i

N -

E,k 1 / f Y 5/sV N /

\

( -

_ It f3 U

+- "! ~

% ,* h!r y .

In _

) SUPPORT

\ PLATE

'5/16

(

NEW SUPPORT l

RIN G GIRDERS l

j f'

1 EXISTING SUPPORT -

I NEW SUPPORT FIG. 2- 19 TORUS SPRAY HEADER SUPPORT MODIFICATIONS

WTELEDGE SSMCES Technical Report TR-5321-1 4x N

N

' .(

\

2'-0" 2'- 0" .

NEW ELBOW REFI REk \ rEXISTING PENETRATION 24"SCH 80 PIPE a

g [ (24"O.D.XlI/ 4" WA LL)

QT f I!! #

sK;_ ~j 'u '

i

>" / /~

I l  :

/ Av >

~ -

RT l ,

l .

7 FORMER HPCI. LOC ATION + 12 g

'CU T LI N E EXISTI NG TORUS SHE LL 1

FIG. 2- 20 HPCI MITRE JOINT MODIFICATION

"E "

TR- 3 1 1 O

\ -

- .. 7?'

Thermowells (16) u 270 - -

t a - s" 90' 3 - o"- o +f j ,

l /

I -

ELEVEN (ll)"T QUENCHERS' CENTERED ON THE RING GIRDERS 180 TOTAL THERMOWELLS = 16 FIG. 2- 21 THERMOWELL LOCATIONS-FITZ PATRICK

Technical R:p;rt YM TR-5321-1 6M 3.0 CONTAINMENT STRUCTURE ANALYSIS - OUTER SHELL & EXTERNAL SUPPORT SYSTEM (INCLUDING ANCHORS)

The containment structure section of this report includes the analysis and evaluation of the following structures:

Torus Shell Support Columns Column-To-Torus Weld Support Saddles Saddle-To-Torus Weld Earthquake Restraints & Attachment

-Anchor (tie-down) System 3.1 Computer Models Analysis of the containment structures was accomplished using the computer models shown in Figure 3-1 to Figure 3-4. The detailed shell model shown in Figure 3-1 was used to calculate the effects of all loads on shell stress, as well as all symmetric loads on the support and anchor system. The beam model shown in Figure 3-4 was used to determine the effects of asymmetric loads on the support system. Asymmetric loads on the torus structure are horizontal earthquake, SRV and chugging. Evaluation of the, support system considered the combined effect of symmetric and asymmetric loads in accord-ance with the load combination table (Table 1).

The detailed finite element model shown in Figure 3-1 simulates one-half of the non-vent bay. It is bounded by the ring girder on one end and the mid-bay point on the other. The vent header system is assumed to be dynamically uncoupled from the shell by the support saddles and is not included in this model. This model was constructed with the assumption that the small offset that exists between the ring girder and mitre joint will not affect results; accordingly, the offset is not included in the model.

1

Tech.iical Report TF WE TR-5321-1 BIGNEstNG SENCES This model includes 641 structural nodes, 736 plate elements and 362 dynamic degrees of freedom. Symmetric boundary conditions were used at both ends of the model.

The model was modified for various load calculations to account for differences in the percent of the water mass that is effective for that load event. In all cases, modeling of the water mass was accomplished using a 3-D virtual mass simulation as an integral part of the structural analysis. The percent of water mass used is identified in the discussion of each load calculation that follows.

The 360 beam model of the torus is shown in Figure 3-4. This model was used to evaluate the effects of lateral loads on. the support system and earthquake restraint system. The beam element properties were selected to simulate combined bending and shear stiffness of the sections. Water mass was lumped with the structure weight on the wetted nodes.

3.2 Loads Analysis 3.2.1 Pool Swell Loads (4.3.1'& 4.3.2)

Analysis for pool swell loads was done using the finite element model shown in Figure 3-1. This was a dynamic analysis performed in the time domain by applying a force time history, to simulate the pressure-time histories of the pool swell event to each node on the computer model.

Input pressure-time histories were varied in both the longitudinal and radial directions in accordance with the information in References 1,10 and 2.

Typical pressure-time history curves are shown in Figures 3-5 through 3-7.

(These pressure-time histories are taken directly from Reference 10, before adjustment, as required by Reference 2. Therefore, the amplitudes shown are slightly different than the loads used in the analysis).

The computer analysis was run for two different pool swell conditions, full A P and zero A P. Figures 3-5 through 3-7 show comparative

Technical R: port TE WE TR-5321-1 ENGNEHtNG SERVICES values and time histories for the two cases. The only difference between the analyses was the input loads; the models were identical. Details of the full load distr.ibution can be found in References 1 and 10.

Plant-unique quarter scale pool sweil tests showed that the effective water mass was less than 100% af ter bubble breakthrough and was slightly different for both zero and full A P conditions (Reference 4). .The water mass used in the computer simulation was constant throughout the ar.aly-sis and was set at the average of the two reduced masses identified in the quarter scale tests. The reduced and average mass values are given in Table

3. This simplification in water mass analysis is consistent with the rela-tively slow (pseudo-static) nature of the pool swell load. This simplifi-cation only affects the inertial (frequency) calculation; the effects of weight are accurately calculated for each load and time in the deadweight analysis.

3.2.2 Condensation Oscillation - DBA (4.4.1)

Analysis for condensation oscillation (CO) was also done with the structual model shown in Figure 3-1.

The condensation oscillation shell load is specified as a spectrum of pressures in 1 Hz bands (Reference 1). The analysis for this load was performed by considering the effects of unit loads at each load frequency (harmonic analysis) and then scaling and combining the individual frequency effects to determine total stress at selected elements. The three variations in the C0 spectrum (Reference 1) were evaluated by re-scaling the results of the unit load analysis.- 100% of the water mass was used for all CO analysis.

A plant-unique factor was applied to the nominal condensation oscillation pressures as discussed in Reference 1; the factor is listed in Table 3.

The combination of individual harmonic stresses into total element stress was done by considering frequency contributions at 31 Hz and

Technical Report TM TR-5321-1 6M below. The actual combination was done by adding the absolute value of the four highest harmonic contributors to the SRSS combination of the others for shell stress. Loads on the support and anchor system were determined by

~

adding the absolute value of the three highest harmonic contributors to the SRSS of the others. These combination methods and use of the 31 Hz cutoff are the result of extensive numerical evaluation of full scale test data, which is reported and discussed in References 6 and 14, and in Appendix 2 of this reoort.

3.2.3 Chugging 3.2.3.1 Pre-Chugging & IBA/C0 (4.5.1.2 & 4.4.1)

The pre-chug load was evaluated for both the sym-metric and asymmetric distribution described in Reference 1. Results for the symmetric pre-chug analysis were also used for IBA/C0 as described in para-graph 4.4 of Reference 1.

Results for symmetric pre-chug were developed directly from the unit-load harmonic analysis done for CO. The results of that analysis were scaled to 2 psi (the pre-chug pressure) and all frequencies in the pre-chug range were scanned to determine the highest possible stresses.

Analysis for asymmetric pre-chug was performed using the beam model in Figure 3-4 by applying the unbalanced lateral -load through the prescribed frequency range.

3.2.3.2 Post Chugging (4.5.1.2)

Post chugging is defined as a spectral load across a wide band, similar in nature to the CO, but much lower in amplitude.

Analysis done on one of the TES plants produced very low stresses and loads that were bounded by pre-chug values. The analyses for pre- and post chug produced these results for maximum shell stress:

Technical Report 'A'TFI W NE TR-5321-1, ENGNEERNG SERVICES Maximum Shell Stress Shell Membrane Stress Pre-Chug 1284 psi l

Post Chug 774 psi

1. Based on frequencies to 30 Hz - sum of 4 maximum +

SRSS of others.

Additional work published in Reference 12 showed that pre-chug bounded post chug (to 50 Hz) for column and saddle loads (Table 5-1, Ref. 12). It also showed that PL+Pb stress due to post chug exceeded pre-chug by 53%.

TES analysis for post chug used the pre-chug stress values. The pre-chug stress may be increased by 53% to account for the 30 to 50 Hz contribution and they will still meet allowable stress.

No further post chug analysis was done for the shell. This position was also influenced by the fact that post chug stresses were very small.

3.2.4 SRV Discharge Calculat. ion of stresses, due tr SRV line discharge pressures, were also done using the finite element model in Figure 3-1. The loading function used for this analysis was based on data collected from in-plant SRV testing in this facility. Testing was done in general accordance with the guidelines given in Reference 2. In these tests, pressure amplitude and frequency were recorded and compared to calculated values for the test condi-tions. Factors were developed that related test to calculated values for both

Technical R: port TE WE TR-5321-1, ENGNEERNG SEB/ ICES amplitude and frequency (See Appendix 1). These factors were then applied to calculated load values for other SRV conditions; the structural analyses were performed using these adjusted values. Appendix 1 discusses the in-plant test and analysis in more detail. A typical set of SRV shell pressures is shown in Figure 3-8.

The method of modeling the water mass in the SRV computer model was the subject of extensive study in this program. Initial attempts to reproduce measured stresses by applying measured pressures to the computer models failed. After considerable study of the nature of the SRV phenomena itself, and the differences between it and the pool swell related loads, it appeared that a dry structure analysis should produce acceptable correlation.

The method was tested and correlation of calculated-to-measured shell stress was excellent. The dry structure analysis method was subsequently used as a basis for all SRV analysis.

3.2.5 Deadweight, Thermal & Internal Pressure Deadweight, thermal and internal pressure analyses were done using the computer model shown in Figure 3-1. Resulting stresses were calcu-lated and considered for all elements on the model.

For the thermal analysis, conduction into the columns and saddles from the torus was considered. Convection from the columns and saddles to ambient produced a calculated temperature gradient in these ele-ments. The torus water, internals and shell were all assumed at the same temperature.

3.2.6 Seismic Seismic analysis for shell stress was done by applying stat-ic G 1evels to the model in Figure 3-1. Load orientation and values were adjusted for vertical and horizontal earthquakes in accordance with Table 3.

Technical Report TN TR-5321-1 6M The effects of lateral seismic loads on the suppc,, t system were determined using the model in Figure 3-4. The effective water mass used in this (lateral) analysis was adjusted in line with test results which showed that net dynamic reaction loads due to the water mass were substantially less than those obtained from static application of the seismic acceleration. A discussion of this fact can be found in Reference 7; the effective water mass used can be found in Table 3 of this report.

3.2.7 Fatigue Analysis Fatigue analysis of the torus shell and external support system is described here. Analysis of the shell at piping penetrations will be described in TES report TR-5321-2, when the piping analysis is complete.

The f atigue analysis of the shell and support system was a conservative one, which was based on applying a stress concentration factor of 4.0 on the most highly stressed elements for each load case. In the case of the support system only, the column-to-torus and saddle web-to-torus welds were considered, since they have higher stresses than the rest of the support system. The process is conservative because:

e It applies the maximum stress concentration (4.0),

l recognized by Section III of the ASME Code to all elements (Reference 11).

and e

l It adds the maximum stress for each load case even though they do not usually occur at the same element.

l The procedure used in this ainlysis consists of the follow-ing steps.

Technical Report TF WE TR-5321-1 ENGNEN SERVICES

1. For a given load, locate the maxirnum component-level stresses (S x , S y , S,,) for the free shell, local shell, and the supports.

2. For these locations, establish the stress intensity ranges and the approximate number of cycles.

3. Repeat the process for all other loads in the load combination.

4. Add the stress ranges for all loads, independent of sign.

5. Multiply these total stress ranges by 4.0 (the SIF).

6. Calculate the alternating stress intensity and com-plete the fatigue analysis in compliance with Ref-erence 11.

Fatigue analysis resulting from chugging af ter an SBA was done assuming that the operator would use a procedure to end chugging af ter 15 minutes. Plant procedures are presently under study to provide for this action.

3.3 Results and Evaluation Re'sults are reported for each structural element of the containment system for the controlling load combination. Controlling load combinations are the.ones that produce the smallest margins against the allowable stress -

not necessarily the highest stress.

All load combinations listed in Table 1 have been considered. As stated previously, most results include some level of bounding analysis and, therefore, understate the margins which actually exist.

TGehnical Report TM TR-5321-1 MM 3.3.1 Torus Shell Results of shell stress due to individually applied loads were calculated and maintained on a component stress level until all the load combinations were formed. Stress intensities were then calculated from these total component-level values.

The controlling load combinations for the shell at Fitz-patrick are Cases 14 and 20 in Table 1, these are:

Case 14 IBA.C0 + SRV + Seismic (SSE) + Pressure + Weight Case 20 D3A.C0 + Pressure + Weight + Seismic'(SSE)

These load combinations control all categories of shell stress, although the location'of the elements is different for the different types of stress. The following table summarizes the controlling stresses.

Approximate locations of the controlling stresses are shown in Figure 3-9.

CONTROLLING SHELL-STRESSES - FITZPATRICK ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS LOCATION (psi) (psi)

Membrane (Pm) Free Shell 13,776 19,300 (Case 20) Element 17 Local (Pl) Local Shell 8,807 28,950 (Case 14) Element 160 Membrane + Free Shell 14,146 28,950 Bending Element 19 (Case 20)

Stress Range Local Shell 27,895 69,900 (Case 14) Element 148

Technical Rtport YE NE TR-5321-1 ENGNEERNG SERiv1CES Compressive Buckling - Acceptable (see below)

Compressive Buckling Reference 13 discusses the results of analy-tical studies and tests on Mark 1 torus structures to determine their compres-sive buckling capabilities. The report concludes that SRV is the dynamic load which presents the maximum chance of compressive buckling f ailure; but, that a safety factor of seven still exists for an applied SRV pressure of +29.3/-22.6 psi. The maximum worst-case SRV shell pressures for Fitzpatrick are +7.8 psi and -5.0 psi, which are lower than those used in the referenced study. Based on this, compressive buckling stresses are considered to be acceptable for the Fitzpatrick torus.

FATIGUE EVALUATION - FITZPATRICK CUMULATIVE USAGE FACTOR

SUMMARY

(STRESS INTENSIFICATION FACTOR 4.0)

NORMAL - EVENT TYPE ELEMENT OPERATING SBA/IBA DBA 19 .000 .001 .001 148 .001 .006 .038 3.3.2 Support Columns & Weld to Torus Shell The controlling load case for the support columns and the column base ,ioint is case 16 of Table 1:

Pool Swell (0 A P) + Weight and case 25 for the attachment weld to the shell.

Pool Swell (full A P) + Seismic (SSE) + SRV +

Weight

Technical Report TN TR-5321-1 MN For these load cases, the following controlling conditions were identified:

SUPPORT COLUMN - CONTROLLING AXIAL CONDITION LOAD CONTROLLING ACTUAL ALLOWABLE COLUMN DIRECTION CONDITION FACTOR FACTOR l

Inner Down Axial & Bending .55 1.0 Outer Down Axial & Bending .65 1.0 COLUMN BASE JOINT LOAD CONTROLLING ACTUAL ALLOWABLE LOCATION DIRECTION STRESS STRESS STRESS Tiedown Up Bending 18.8 ksi 28.5 ksi Clamping Plate COLUMN-TO-SHELL WELD LOAD CONTROLLING ACTUAL ALLOWABLE LOCATION DIRECTION STRESS STRESS STRESS Inner Down Shear 15.82 K/in 24.13 K/in Outer Down Shear 16.96 K/in 24.13 K/in

Technical Report TR-5321-1 ENGNEERNG SERVICES 3.3.3 Support Saddles & Shell Weld Stress levels in the saddle structure are very low for down-ward loads. The controlling stress occurs in the saddle clamping plate (which connects the saddle to the anchor bolts) during the upward loads associated with load case 21 in Table 1. Controlling stresses in the saddle-to-shell weld occur during downward loads associated with case 25. These cases include:

Case 21 DBA.C0 + SRV + Weight Case 25 Pool Swell (full AP) + SRV + Weight + Seismic (SSE)

Controlling stresses and loads for these cases are:

SADDLE STRESSES TYPE ACTUAL ALLOWABLE LOCATION STRESS LOAD LOAD Clamping Plate Bending 51.7 kips 75 kips SADDLE-TO-SHELL WELD TYPE ACTUAL ALLOWABLE LOCATION STRESS STRESS STRESS Outside End Shear 12.04 K/in 13.65 K/in i

Technical Report TN TR-5321-1 M $8{\/)CES 3.3.4 Earthquake Restraints & Attachments The earthquake restraint system is illustrated in Figure 3-

10. The controlling load case for this system is the one that produces the largest lateral load. This is case 15 which includes:

Chugging + SRV + SSE All three of these loads have been selected to produce the highest lateral load on one earthquake restraint; contributions from the individual loads were added directly.

The controlling stress results follow:

EARTHQUAKE RESTRAINT STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Attachment to Concrete 10 psi 65 psi Concrete Shear ATTACHMENT WELD STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Base of Shear 1,443 psi 21,000 psi tie plates 3.3.5 Anchor (Tie-Down) System The load combination which produces the highest upload and minimum margin on the anchor bolts is case 21 for the saddle anchors:

DBA.C0 + SRV + Weight

l I

Technical ' Report YM TR-5321-1 6M t

and load case 16'for the column anchors:

Pool Swell (zerotsP) + Weight For these cases, the anchor bolts with the snallest margin of safety (accounting for as-built conditions) are:

SADDLE ANCHOR BOLT ACTUAL FACTOR MAXIMUM MAXIMUM 0F LOAD CAPACITY,, SAFETY 51.7 K/ bolt 264 K/ bolt 5.1 COL"MN ANCHOR BOLTS ACTUAL FACTOR MAXIMUM MAXIMUM OF LOCATION LOAD CAPACI1f SAFETY ;

Inside Column 44.2 K 258 K 3.96 Outside Column 65 K 258 K 3.96

D CD 0 o m :7 LJ ::3 N+

WO

1. a' s' ,

OD i ', ,

i :o rp e i  !

o ,

/

/ D O

/

9 "'*o, / , / l

~ o's, , , 's e '

f 'F.:; s -? '

c 1 3 [

,. / ,/

m &a , '

o

, i j j

~;l '  !

u

's s' ,! /

h. n ', ' , .

'l n -

R I s ', 's' ' / ,

/

/ . l / - / ,

(* I s

' '..; ~j; ,7 / 's' y N h ;

,'s j si Q

• I / , ,, I,y ',o l t ',  ;, .. 'r, , si ,

/, l Q , $

/ '~

' o; s

( d'g s 2 \ -

't f M g(h ;t '%, ' ','-

'/

I :::,  ?'n, Y *%'% / S' e'

,' /' VIEW ROTATED AXES ,

! ' '"*$!$fg!!Ig .?.

4 p%3zi ,

J. A. FITZPATRICK NUCLEAR PLA T I

WTERME BGEWW43WMCES Technical Report TR-5321-1 be O

O en inside E.

s- $

sg ,# g i

,i g Y , _ _ _. . , ,- - _ ._ , .

st, s, 3

,/

9 ' ,f l s

=*-e_

.,i. ..em d.e e. --,. .= e -ii

/ 8 j s % ss,y s i g

%I 0 c #

f

, f t lt l1 l 1 6

/ ,,

8 9 e 4 110

, e p f -P- f= -r-=9- P- ~

\l / ,-

/ ,s ', ' t i,f. ,-A,- ,, ff*,-f,-t-t -

ii

,s s

o

< -)--y-

,,i* 4,,-*- .,

,-- , . is. .i,,, ,e.

s' s

r_p,;

?;/ ,

s

- Ts-si 'ii s.

.s

. w.r--

i :

s s

W. g

' ~.

-e ..

/ .... .-

q. . ;_a

'. _-'-)  :

4, 1, a

e O

4 l

c

's.

-9 _.

r.-e

! ,' r ,y;;

>+ ..

.., .J

~,

s, ,' .- ,

~. --t, ;-

l

-- s ,' l *,

,, ss .:  !

, s, ,+ _'s i-y t -

1. y 0/ I ,

s 4isi ,,9

$ _,_ I s 's _s4_ !/ ,i. ! .

.' ,- N' . , ," i ,y - ' : ! s

, , s

. N : 3-y -y - t, ; -

s ss s s , 4, ,_.r,. a.i_.

i

. , -

• g \ -i ,-, ,: . . . .

je <s g i i ,e r' s , , .i i i g i ,

l \ ' ' 'i #, , s y . ~ g

= -

outside n.

o 0.e c

5 l

l l

I 1

1 l 'F90URE 3.2 TORUS SHELL MODEL 1

l l

l l

M a,9-M E.

79 1/32nd 5'

_ TORUS MODEL.WITH ' 3 SADDLE E X,

C l ring girder a

• N'*x .- 4 *I b

n.m -

ASysG[R,04

' # ring girder m rr.rso.

c 4. 4.078 h i 15.269 124.932 5.6798 2.8469* E 41.nl -.

32.2215 i j 73.425 fs 0.0000* . u X2 177.0fT _.

Ig --

e 4 J. A. FITZPATRICK NUCLEAR PLANT

! I

--2 OUTSIDE EARTHQUAKE TIE COLUMN \

\ -

2 4

.J CL E

4 b

-INSIDE COLUMN o

3

=

= ,SADDtE ] ?] ]

TOnuS o

=

, 1 1 E m

k

=i l

l I

FIGURE 3.4 360%ORUS BEAM MODEL I l

l l

i I

1

6

^

gg ae M E.

4_ 7g OAP 5' 2- i" h_

O- N ^N

~

4 A P = 1.7 p s i '

'l 0 w

i E

D

! y g w

x o-f p

f 9

-l4_

O 20 400 6b0 800 10'0 0 IIOO 1400 16'00 1800 TIME (msec)

FI G . 3- 5

POOL SWELL- NET TORUS VERTIC AL LOAD -FITZPATRICK

, - yu' NN bi MX m

40-1

=

g 30-o OAP

$ A P = 1.7 psi a w '?

m O.

20--

10-k 0-O

),

200 400 600 800 1000 1200 i

1400

~

i 1600 1800

) TIME (msec) i FIG.3-6 POOL SWELL- AVERAGE SUBMERGED PRESSU RE-FITZPATRICK

WTREME ENGOEIWG!EfMCES Technical Report TR-5321-1 O O

~E O

.O 9

x 0

o E O

~w F-

<t Ct.

N 1 O b 4 O t O '0! I W

E

- D O o u)

.o si W m 9 e

,,,,, r0 m LtJ 1

2 $%

' ~

O i~

=O 00 m O

._ m a O O

s

O g n W g l

1 Lij d 3 m

8 a w O O

O.

O

-O N

O d b m m b O o e -

(!8 d ) 3 BASS 3Bd

E!#

8- .

32,o 7- BASED ON TEST 4, GAGE PZA- ~"

SCALED TO WORST CASE {

O 6- A 5-d I 4

$ 3-o i $

w 2- g W

n. l-I O-i

-l-7 J

j \

N d d.l d.2 O'.3 d.4 d.5 TI M E -+-

FIG. 3- 8 5 TYPIC AL SRV SHELL PRESSURE - FITZPATRBCK

1 1

W TE RhtE Technical Report

~

TR-5321-1 ~

5 Ev "T-~

2, m in cz  % s s.

b

%i-

.5 m

ld;c. 5 270* b l ~

outside W:' ~'l. I

.-:'F

.ii ..it---

260 154__.  ;TC'r T , .. ~l 3 an .-- .-- -

250 i Mox P L hs :,! I -

2<o 148 -

230 (h_'..33h.._.h.

-y

'. EM '  ?!

220 i.2 P4' 9t ,,,it-- .

210 MAX .

STRESS ' ' - , ., ' , ' ' --

Max P m aP L 4p t- -

190g

.;d 180* i I i e EBTI'om J."p'i ,  ; 180J - Ma x P m no ,: g .,1. ..

170 m

,f MODEL LOWER Y-;I'3{-j -

1so 32 -"'

. i.~

H ALF OF SHELL -5w' ' -

150

'i

.h'f.  !

9,;-

140

'j'% ,

iis -

130 T .i p;;7-- 120 uz - Q q___y ,ma - 110 -

t . .

pp'f.f--f- - 100 i- 90* <!  ;

> 90 o,

a  :

insideB< ls's 9: 3 9 3 l= s - ~

sz9 68 Tn CD [!

c s

FI G. 3-9 M AXI M U M SHELL STRESS LOCATIONS- FITZ PATRICK l

l..

YD ENGNgW4GERWCES Technical Report TR-5321-1 .

< A+ , TORUS SHELL u l4'.8" '

- F ' / .

,-e~s If 's, y

, - to},"

$)' ~

y [ U /U U;U U U 'fi,)

x x A

'S,,v l

1 i

TOTAL 4 RESTRAINTS \\'\h'N %'x'x 90* AROUND TORUS 1 i h /

16 8 ,,

q )" - - . -

y i s,16 ijk; '

. 9, (s *f

• r'

.  :;q

,' f

• l

., *: h l SECTION A- A FI G. 3- 10 E ARTHOUAKE_ RESTRAINT SYSTEM _

Technical Report TME TR-5321-1 ENGREERNG SERVICES 4.0 VENT HEADER SYSTEM 4.1 Structural Elements Considered The vent header system, as defined in this section, includes the following structural components:

a. Vent Header (V.H.)

b. Main Vent Pipe (V.P.)

c. Downcomers (D.C.)

d. Downcomer Tie Bars

e. Deflector

f. Vent Header Support Columns & Attachments

g. VH/DC Intersection

h. VH/VP Intersection

1. VP/Drywell Intersection J. Vent Header Mitre Joint The main vent bellows are considered in Section 7.0.

4.2 Computer Models Two computer models provided the means to analyze the vent header system; they are shown in Figures 4-1 through 4-_4.

The first of these is a detailed shell model (Figures 4-1 to 4-3),

which includes a highly detailed representation of one-half of the header in a non vent bay, complete with four downcomers.

The model also includes an approximate representation of one-half of the vent bay; this was intended to provide the proper boundary conditions and stiffness transition near the non-vent bay. The vent bay half of the

Technical Report TN TR-5321-1 MM model was not used for stress determination. This large finite element model was used primarily to determine shell stresses in the non-vent bay; some other uses are discussed in the following text. It was used for both static and dynamic analysis and provided detailed stress gradient inf.ormation in the downcomer/ vent header intersection region.

The second vent header model is the beam model shown in Figure 4-4.

This model represents a full vent bay, complete with vent pipe and downcomers; as well as a half non-vent bay on either side. It was used to determine boundary loads on the vent system components to support a more detailed stress analysis of those components. This model was used to define loads on the following elements:

e Vent Header Support Columns e Vent Pipe / Vent Header Intersection e Vent Pipe /Drywell Intersection e Vent Header Mitre Joint e Main Vent Pipe The loads and moments taken from the beam model were used in further analysis to determine stresses. The calculation methods used for these stres-ses are:

e VH support columns - hand analysis e VP/VH intersection - applied stress multipliers (stress intensification factors) from Reference 7 e VP/drywell intersection - used stress multipliers from -

Reference 16 (Bijlaard) e Mitre joint - used stress multipliers from detailed shell model (Figure 4-1) e Main vent pipe - hand analysis

Tcchnical Rep;rt Y TR-5321-1 - 5 9-- ENGSEERNG SERVICES The beam model used a stiffness representation of the VP/VH inter-section taken from Reference 7. Attachment stiffness between the vent pipe and drywell was calculated using References 17 and 18.

Pool swell water impact on the vent header deflector was calculated with a hand analysis. The impact forces were applied statically to a beam model and a dynamic load factor was applied (see Figure 4-5).

4.3 Loads Analysis 4.3.1 Pool Swell Loads 4.3.1.1 Pool Swell Water Impact Analysis for stresses due to pool impact and drag was done using both computer models.

Determination of shell stresses was done with the detailed model in Figure 4-1. For this analysis, force time histories based on QSTF test data was used (References 4 and 10). These time histories were applied at 100 nodal points on the shell model and the dynamic response of the structure was calculated. Relativetimingbetweenloadings(Reference 1)was maintained to preserve accurate representation of longitudinal and circum-

~

ferential wave sweep. Stresses in the vent header /downcomer intersect 1on, as -

well as in the free shell areas, were taken directly from this model. Stres-ses in the downcomer +.ie bars were also taken from this model. Analysis was done for both full and zero AP impacts.

l

! The beam model (Figure 4-4) was also used to deter-mine stress from pool swell impact and drag. This was done with a time history dynamic analysis Lsing loads developed by integrating the impact pressures over small areas and reducing them to nodal forces. Approximately

Technical Report TR-5321-1 EN SERVICES 95 nodes along the length of the beam model were dynamically loaded in this analysis, including lo ads on the VP/VH intersection and vent pipe. The results of this analysis were used to define boundary loads on VP/VH inter-section, mitre joint and other elements as listed in Section 4.2. Stress analysis for these elements was performed using the methods indicated in Section 4.2.

4.3.1.2 Pool Swell Thrust (4.2)

Pool swell thrust forces are defined as dynamic forces at each bend or mitre in the vent system, and are a consequence of the flowing internal fluids. Analysis for these loads was done using the beam model and applying the loads statically.

This is consistent with the slow nature of the applied pressure forces.

The calculation was performed with the maximum value of all thrust forces applied simultaneously; this is a conservative condition. ,

4.3.1.3 Pool Swell Drag Loads (4.3.7 & 4.3;8)

The vent header support columns are loaded by for-ces from LOCA-jet and LOCA bubble drag. By inspection, it was concluded that LOCA-jet loads would not combine with water impact on the vent system due to differences in timing and, therefore, would not contribute to the maximum stress calculations - LOCA jet forces were not considered further.

LOCA bubble forces were calculated and the maximum normal components (radial and longitudinal) were applied simultaneously to conservatively bound the bending moments on the support column. These peak values were applied statically at the midpoint of the column. Stress calcu-lations were done by hand.

• Technical Report TE WE TR-5321-1 ENGmEstNG SERVICES

_ 4.3.2 Chugging Loads 4.3.2.1 Downcomer Lateral Loads (4.5.3)

Reference 1 identifies downcomer lateral loads as static equivalents with random orientation in the horizontal plane. The major consequence of this loading is to produce high local stress in the VH/

downcomer intersection. Tiie detailed shell model (Figure 4-1) was used to identify stresses in.the downcomer intersection due to static loads applied at the base of the downcomer. Frequencies of the first downcomer response mode were taken from a dynamic analysis on the same model (Figure 4-1) with the downcomers full of water to the operating level. This frequency was necessary to determine the proper dynamic scale factor to apply to the static load.

The stress results from the statically applied load were used as a basis for a fatigue evaluation of the intersection in accord-ance with Reference 1.

4.3.2.2 Chugging - Synchronized Lateral Loads The random nature of the downcomer lateral chugging load provides for all combinations of alternate force orientations on adja-cent pairs of downcomers. Various load combinations were examined to deter-mine stress levels in the vent header and mitre joint as a result of these loads. The cases considered are shown in Figure 4-6.

s These cases were considered by applying static loads to the beam model (Figure 4-4) and determining final stresses as described in Section 4.2.

4.3.2.3 Internal Pressure (4.5.4)

Three vent system internal pressures exist during chugging. They are:

Technical Report TN TR-5321-1 MM e Gross vent system pressure - a .7 Hz oscillat-ing pressure with a maximum value of 5.0 psi.

This pressure acts on the entire vent system.

e Acoustic vent system pressure - a sinusoidal pressure varying from 6.9 to 9.5 Hz at a maxi-mum value of 3.5 psi. .This pressure affects the entire vent system.

e Acoustic downcomer pressure oscillation - a 40-50 Hz pressure at 13 psi that produces only hoop stress in the downcomers.

Responses to these pressures were estimated using hand analysis and were determined to be substantially less than those result-ing from internal vent system pressures during poo,1 swell. The values associ-ated with pool swell pressures were used in all combined load cases involving chugging pressures; this produces conservative results.

l 4.3.2.4 Submerged Structure Drag (Support Columns only) l Examination of the load combinations that include l chugging makes it clear that these cannot control maximum stress level in the '

support columns; combinations that include vent heade.r water impact will produce much higher stresses. For this reason, stresses in the vent header support columns were not calculated for chugging drag.

l Drag forces on the downcomers and downcomer tie

bars are already included in the Downcomer Lateral Loads, which were based directly on test data.

I i

Technical Report WW WE TR-5321-1 ENG#EERNG SERVICES 4.3.3 Condensation Oscillation - DBA 4.3.3.1 Downcomer Dynamic Load (4.4.3.2)

The downcomer dynamic load, due to condensation oscillation, is a sinusoidal pressure variation that can be equal or unequal in the two downcomers forming a pair.

The unequal pressure case produces a net lateral load on the downcomer much like chugging. The major considerations for this load are stresses in the downcomer intersection due to a net lateral load on one pair of downcomers and a more general stress case where loads on adjacent downcomer pairs are phased to produce gross lateral loads on the vent system or torsion in the vent header.

The horizontal component of the C0 downcomer load produces the same type of loading on the vent system as the CH lateral load, in terms of the stress analysis. A comparison of the magnitudes and frequen-cies of these two loads shows that stresses resulting from CH horizontal loads will bound C0 horizontal loads.

The C0 -downcomer load also produces a vertical component of load, which is not present during CH. The effects of this load were evaluated by static analysis of the detailed vent header model (Figure 4-

1) and consideration of dynamic amplification effects, using the beam model (Figure 4-4). This evaluation showed that the combined effects of the C0 downcomer load (horizontal and vertical components) would still be bounded by CH lateral loads.

Based on this, CH lateral load results were con-servatively used in all load cases in place of C0 downcomer loads.

Tcchnical R: port TN TR-5321-1 ENGNEERING SERVICES 4.3.3.2 Vent System Loads (4.4.4)

Vent system loads consist of a sinusoidal pressure in the vent header and downcomers superimposed on a static pressure. The dynamic pressure in the downcomers is used to calculate hoop stress only.

Stresses for all pressure loads were based on hand analysis using static analysis. The static analysis assumption is consistent with the low frequency of the applied pressure and the fact that the ring modes of the structure are very high.

4.3.3.3 Thrust Forces (4.2)

Stresses resulting from C0 thrust forces were conservatively taken from the pool swell thrust calculations and applied to all C.0. load cases (para. 4.3.1.2).

4.3.3.4 Drag Forces on Support Columns Inspection of approximate total loads on support columns due to CO, CH, and pool swell showed that condensation oscillation would not contribute to the maximum column load, due to differences in timing.

No detailed analysis was performed.

4.3.4 Condensation Oscillation - IBA Stresses and loads resulting from IBA condensation oscil-lation are bounded in all cases by either DBA condensation oscillation or chugging. No detailed analysis was performed for IBA condensation oscil-lation. -

Technical Report YE NE TR-5321-1 ENGREBtlNG SERVICES 4.3.5 SRV Loads 4.3.5.1 SRV Drag Loads An SRV discharge produces drag loads which act on the vent header support columns, downcomers, and downcomer tie bars. Analysis for drag loads on these structures was based on data collected during in-plant SRV tests.

Data collected during these tests was scaled to correct it for the appropriate SRV conditions and then applied to the struc-tural model to determine the resul. ting stress. A more detailed discussion of this procedure is provided in Appendix 1.

4.3.6 Other Loads Deadweight and seismic stresses in the vent system were cal-culated using the beam model of Figure 4-4.

Seismic stresses were calculated by statically applying the acceleration values in Table 3.

l Thermal stresses were determined for the steady state appli-cation of maximum vent system temperature, using hand analysis.

4.4 Results and Evaluation Results are reported for each structural element of the vent system for the controlling load combination. Controlling load combinations are thc.

ones that produce the smallest margins against the allowable stress, not necessarily the highest stress. All load combinations listed in Table 1 have been considered.

4 Technical R; port TF WE TR-5321-1 ENGmEERING SERVICES As stated previously, most results include some level of bounding analysis and, therefore, understate the margins which actually exist.

4.4.1 Vent Header - Downcomer Intersection ,

1 The controlling load on the vent header-downcomer inter- l section, both for maximum stress and f atigue, is the downcomer lateral load '

due to chugging. The worst load combination in which this load appears is j case 15 of Table 1. This cases consists of:

Chugging (IBA)+ Seismic (SSE)+ Weight +Presssure+ Thrust

+ SRV For this case, the following stress occurs at a point 90 from the plane of a downcomer pair. It is primarily the result of a longi-tudinal chugging force on the downcomer.

ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS ,

f Combined Maximum Stress 35,303 psi 37,635 psi 4.4.2 Vent Header - Vent Pipe Intersection The controlling load on the vent header / vent pipe inter-section occurs as a result of pool swell water impact. The controlling load condition is case 25 in Table 1 which includes:

PoolSwell(full /iP)+ Thrust + Seismic (SSE)+ Weight +SRV Pressure ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS Combined Maximum Stress 26,748 psi 28,950 psi

Technical RGport TM TR-5321-1 mM This load case was formed using a 0AP load, and was eval-uated to a level A allowable. This conservative evaluation was performed to eliminate the need to evaluate several other vent header load cases.

4.4.3 Vent Header Support Columns and Attachments The controlling load combination for the vent header support columns and the clevis joints at each end is case 25, Table 1. This case includes:

Pool Swell (full AP) + Seismic (SSE) + Weight + Thrust + SRV As before, the evaluation was conservatively performed using 0 A P loads and a level A allowable.

Controlling stress in the support column is:

ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS Axial in Column (tension) 17,420 psi 18,000 psi Controlling stress in the clevis joint at the and of the support column is:

ACTUAL ALLOWABLE LOCATION STRESS TYPE STRESS STRESS Clevis Plate Shear 13,015 psi 15,200 psi 4.4.4 Downcomer Tie-Bars and Attachments The controlling load combination for stresses in the down-comer tie bar and attachments is case 25, in Table 1. The major load is associated with pool swell impact on the crotch region of the downcomers which produces tensile loads in the tie bar.

Technical RCpsrt YME TR-5321-1 N S N ES The controlling case includes:

Pool Swell Impact (full /S P) + SSE Seismic + SRV + Weight +

Pressure'+ Thrust Zero 4sP pool swell loads and service level allowables were conservatively used in this analysis.

The controlling stress is:

ACTUAL ALLOWABLE LOCATION STRESS TYPE STRESS STRESS Tie Bar Clamp Bending 10,614 psi 22,240 psi 4.4.5 Vent Header Deflector and Attachments The major load on the vent header deflector occurs as a result of pool swell water impact. The controlling load condition is case 25 in Table 1 which includes:

Pool Swell (full tsp) + SSE Seismic + Weight + SRV The controlling stress in the deflector is:

l ACTUAL ALLOWABLE l LOCATION STRESS TYPE VALUE VALUE Center of Bending 6,236 psi 16,500 psi i the Long Span The controlling stress for the attachments is:

ACTUAL ALLOWABLE LOCATION STRESS TYPE VALUE VALUE l

Fillet Weld Shear 10,662 psi 18,000 psi 9

Technical Rep 1rt TM TR-5321-1 MM 4.4.6 Main Vent /Drywell Intersection The major load on the drywell penetration occurs as a result of chugging. The controlling load condition is case 21 in Table 1 which includes:-

Cnugging + Seismic (SSE) + Weight + Pressure (Drywell)

The controlling stress is:

ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS Primary and Secondary 23,664 psi 69,900 psi The effects of all loads from the vent system, and the pres-sure load were considered using Reference 16. Information regarding stresses due to seismic and thermal response of the drywell was taken from the Fitz-patrick FSAR update (Reference 19).

The controlling stress is:

ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS Primary and Secondary 17,724 psi 69,900 psi i

The total vent-to-drywell intersection stress is:

ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS l

Primary and Secondary 41,388 psi 69,900 psi

r Technical Report TR-5321-1 N MES 4.4.7 Vent Header, Main Vent & Downcomers - Free Shell Stresses It was established by inspection of the computer results that large safety margins existed in free shell regions and that minimum safety margins would be controlled by local shell stresses. No further work was done for free shell stress in these structures.

4.4.8 Vent Header - Mitre Joint The controlling load on the vent header mitre joint occurs as a result of pool swell water impact. The controlling load condition is case 25 in Table 1 which included:

Pool Swell (full AP) + Thrust + Seismic (SSE) + Weight + SRV

+ Pressure ACTUAL ALLOWABLE TYPE OF STRESS STRESS STRESS Combined Maximum Stress 18,935 psi 28,950 psi 4.4.9 Fatigue Evaluation The f atigue analysis of the vent system is a conservative one which assumes that all maximum stresses occur simultaneously, and that all cycles reach these maximum values. The duration of the major loads in this analysis is 900 seconds, the length of chugging associated with an SBA/IBA event.

5 The procedure used in this analysis consists of the follow-ing steps:

e For a given load and component, locate the highest stress.

Technical R': port YE WE TR-5321-1 ENG#EERNG SERVICES e For this location, establish the stress range, o Repeat this process for all other loads in the load combination.

e Add the stress ranges for all loads.

e Multiply this total stress range by the appropriate stress intensification factor, e Calculate stress intensity and determine the allow-able number of stress cycles.

e Determine the usage factor, using the methods of Reference 11.

The fatigue evaluation was performed for all high stress areas in the vent system. The major load, contributing to the fatigue evalua-tion is chugging following a DBA. The controlling load case is case 21 in Table 1, which includes: l IBA.CH + Seismic (SSE) + SRV + Weight The controlling usage factor for the vent system is:

VENT SYSTEM FATIGUE RESULTS ACTUAL ALLOWABLE USAGE USAGE LOCATION FACTOR FACTOR 7

Vent Header Support .98 1.0

Prisuanm Technical Report ENNEEIW4GSSMCES TR-5321-1 END NODES FIXED

\

\ -

s s SUPPORT COLUMNS fl 3

6 V .

+ ,

I).

y 7

"J f

SYMMETRY BOUNDARY CONDITIONS THis END

, TIE BARS FI G. 4 - 1 VENT HEADER MODEL

WN ENGNEIW4GSEMCES Technical Report TR-5321-1 d N i I 7G% k

/ x- ~ x y auxm

/ / / V>W i \% ,M/ /rM \ \ \

\\\M ($'L_: $[\ W4 / / /

DOWNCOMERS yM{ pc-j

~}&&

/ //X/#\ \% E7#2 \ \ \

h"bL-

\w:\\MV/('l _$]h\ WX / / /

h s w>c>c/

) \ /  ?

SUPPORT COLLAR [' N

~

~~

/.  !

/  !

1 THIS SECTION FOR BOUNDARY SI MUL ATION /

ON LY /

i FI G. 4- 2 DETAILED VENT HEADER MODFI

WTERME ENGWEEWGSERVICES Technical Report TR-5321-1 .

L:::::g GUSSE T

[---l

, _ _ _p.

3 -7 RE- E N FOR CEMENT y.

h,* / Y isgs s

e l <

s / ['%r,p[to,N,'

," t y

, Nst ,s'

/ h/ y 'h' b ,--i-+h:f. pt "' e

, s \(

, .s

/

, /

f 'yf.

/

/

/, s, N'

/-

/ , , r! ! ! ! , ',',y'[i

'mA' \

\ '

"s

- \

/ / 'Nj'N ' ///

\ g

'g \ \g, gg

/ / /

', \

/ / Ys ! *. ,'<, e # \ \ \

f j'/

/ /

/ /

/

/

//

/7

// T \

.'\\

\

\ \['

\* # r" g\ g\

\

//

/

/ /

/ 's' Y's

/

/

/Y

//

/

• */ -\

\

Y\

\

k" p#g\

\

g \

.g \

f \

g \ \.

/

// N 's s i'+~_'7 i  ; .4 / '

! I I I i  !!

i t

\,4l . -7, li_ k ~,

! g I I

-', I_ w_ _.1 I e f I t f i i

e I !

i I j

._ _ _ .s s _ _ _+ __ _u t ! t i !

I l i 1  ! !  ! I f I t t I t l ! I 1 l l l t i I t 4 i t t SECTION THROUGH ll l ll I t i

? I

-+ -- 4t --- +-4 DOWNCOMER PAIR- s - --+ --+-4 I t I l i I I I I ll t t l

t ll t ?

TIE BAR NOT SHOWN ll l ll

! I I I ! I f 8 I !

I f f I t  ! !  ! I t H___ __;.___4_a  ! I i  ! !

' i 1 1 1 s _ _ _4__ _4_;

! t I t i 1 1 I t I I t I 4 I I i 1  ! !

1 1

• t a f f i  !

1 i  ! I e t I i 1 t f  ! t t I t t t I, , .

i e

?, , 3 t t

. , ,I FIG. 4-3 DETAILED VENT HEADER MODEL

i WTELEDGE ENGOEERNGSEWICES Technical Report TR-5321-1 s -J s >

llNODES/ # [ - RING HEADER SPAN (TY PI o

,o DOWNCOMER VENT HEADER

( ,

SUPPORT COLUMN l

VENT PIPE i

i

, Ls tA B ot o i

\ '

=

! \  !

I VENT HEADER INTERSECTION

\ /

C g

/

/ (STIFFNESS MATRIX INPUT) g s ._ -

I >

i O O l

t ,

4> >

\

d '

O od SYMMETRY BOUNDARY CONDITIONS FIG.4-4 VENT SYSTEM BEAM MODEL

WTE.EME NM Technical Report TR-5321-1 STRUCTUR A L FREQ.= 59.3 hz LOAD FREQ. = 2 bz DLF=1.0 6.3' r 2 0.8 '

6. 3 '

ANALYSIS BE AM MODEL X

d b '

<N

[ 3.6 (K/g ) UNIFORM i

ca n n pc m- >

APPLIED STATIC LOADING 1

i FIG. 4- 5

, VENT HEADER DEFLECTOR AN A LYSIS - FITZ PATRICK

SE5MCES Technical Report TR-5321-1 _77_

4 9176 - T Y P.

3142 - TY P. ]

.au CASEI N

/

s LA, s--

g

/

f -o v

v s\

aw CASE 2

\. \

l x _o j\ o.

/ -o o g\

p u l

l l ANALYSIS CASES FOR SYNCHRONIZED LATERAL CHUGGING FIG 4-6 i

Tcchnical Report TN WE TR-5321-1 ENG4EERNG SERVICES 5.0 RING GIRDER ANALYSIS The ring girder for Fitzpatrick is shown in Figure 5-1. It is mounted in a vertical plane'that passes through the support saddles and the support col-umns. Because all major internal structures are supported by the ring gir-ders, the ring girders react to the largest number of individual loads.

5.1 Structural Elements Considered Elements considered in this section are:

(a) The ring girder web and flange.

(b) The attachment weld to the shell.

Local stresses at attachments have also been considered and added; i.e., vent header support columns, catwalk, etc.

5.2 Computer Models '

Two cortputer models were used as a part of the ring girder analyses; both are detailed 'models which also include the shell and external supports.

l l The first model is shown in Figure 5-2. This is a detailed model, which represents one-sixteenth of the torus structure; one half bay on each side of the mitre joint. It accurately simulates the ring girder offset (four-inches from the mitre joint), as well as structural differences between the vent and non-vent bays. Because the ring girder is not at the boundary of this model, out-of-plane motion of the ring girder can be accurately deter-mined. This model was used to evaluate all direct loads on the ring girder; these include loads from attached structures such as the tee-quencher sup-ports and vent header system, as well as all drag loads, i

TGehnical Report TN TR-5321-1 MM The one-sixteenth model used for the Fitzpatrick ring girder analy-

~

sis wss one that had been constructed for one of the other Mark 1 plants analyzed by TES. The dimensions of this other plant are very similar to Fitpatrick; the diameter of the torus, shell thickness and distance between the ring. girder and mitre joint are all similar. The ring girder flange in this model is slightly smaller than Fitzpatrick and, therefore, produces conservative results since lateral loads control ring girder stresses. The comparison is:

. Ring Girder Flange Dimensions (inches)

Fitzpatrick: 1.5 x 8 Model Used: 1.5 x 6 The second model used to determine ring girder loads is the Fitz-patrick 1/32 finite element model shown in Figure 3-1. This model was used previously to evaluate shell stresses of all symmetric loads that act on the shell. These same computer analyses produce information on ring girder stress for symmetric loads. Loads evaluated with this model include weight, internal pressure and all shell dynamic loads. The boundary conditions on this model restrict the ring girder to in-plane motion.

5.3 Loads Analysis 5.3.1. Loads Applied to Shell As stated, the ring girder stresses for all symmetric loads applied to the shell were taken from the appropriate analyses described in Section 3.0; these include:

(a) Pool Swell Shell Load (Paragraph 3.2.1)

(b) Condensation Oscillation (3.2.2)

(c) Chugging (3.2.3)

Tc,chnical Report WTF1 Fry (E TR-5321-1 ENGNEERNG SERVICES (d) SRV Discharge *

(e) Seismic (f) Deadweight, Thermal and Pressure

• SRV discharge is conservatively assumed to be a symmetri-cally applied load for shell analysis.

5.3.2 Drag Loads The ring girder is subject to drag loads from each of the dynamic shell loads as well as Fluid Structure Interaction (FSI) effects from C0 and CH. All these loads were evaluated by using the 1/16 model and applying static loads out-of-plane on the wetted nodes of the ring girder.

The use of static analysis was based on the assumption that the stiffening effect of the saddle, columns, and column gussets make the ring girder very stiff and would prevent frequency interaction with the dynamic loads. Because of this, no dynamic load factors were applied to the static analysis results (DLF = 1.0). Drag loads considered were:

(a) Pool Swell Bubble (b) Pool Swell Jet (bounded by a)

(c) SRV Jet (d) SRV Bubble (e)' C0 including FSI (bounded by g)

(f) Pre-chug including FSI (bounded by g)

(g) Post Chug including FSI The effects of SRV jet (c) and SRV drag (d) were evaluated based on data collected from in-plant tests. A discussion of the in-plant tests and the use of drag data from these tests is given in Appendix 1.

Calculation of ring girder drag loads due to condensation oscillation and post chug FSI was not in accordance with NUREG 0661 (Reference 2). An alternate method of calculating drag volume was used in this load

WTELBANE SERVICES calculation. It produced drag volumes for the ring girder of about half of those that the NUREG 0661 procedure would have produced. A discussion of this is included in Appendix 3. The FSI drag calculation was based on local pool accelerations at the ring girder resulting from the response of the entire shell. The post chug and FSI analysis considered frequencies to 31 Hz, which were combined by adding the values of the five maximum components to the SRSS sum of the others.

5.3.3 Loads Due to Attached Structure Loads applied to the ring girder by structures attached to it were evaluated by' equivalent static analysis, using the 1/16 model (Figure 5-2). The important loads are applied in the area of the support saddle and columns which make the ring girder very stiff and minimizes dynamic inter-action. Because of this, dynamic amplification of the static ring girder stresses was not done (DLF = 1.0). The load input to the ring girder was a result of a dynamic analysis of the attached system (or had an appropriate DLF applied) and, therefore, included the effects of dynamic amplification on '

load.

The following loads are applied to the ring girder and were considered:

e Tee-quencher support beam thrust due to SRV dis-charge.

e Tee-quencher and support drag loads.

e Vent header support column reaction loads during pool swell.

e Vent header support column drag loads.

As stated in Section 5.1, stresses resulting from attached structure have been included in the following results.

Technical Report TN TR-5321-1 6M 5.4 Results and Evaluations 5.4.1 Ring Girder Web and Flan,ge The controlling load combination for the rir a girder web and flange is load case 16 of Table 1; this includes:

Pool Swell (zeroAP) + Weight The controlling stresses are:

STRESS STRESS ACTUAL ALLOWABLE LOCATION ~ TYPE STRESS STRESS Web Membrane 14.9 ksi 19.3 ksi Flange Membrane 16.9 ksi 19.3 ksi 5.4.2 Weld to Torus Shell The controlling load combination for the shell weld is load ,

case 21 of Table 1. The controlling stresses are:

DBA.C0 + Seismic (SSE) + Weight STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Column Region Shear 7.64 K/in 8.53 K/in Inside Column Region Shear 8.27 K/in 8.53 K/in Outside Saddle Region Shear 7.29 K/in 8.53 K/in

WTREME BIGNEBW4G5iRVICES Technical Report TR-5321-1 u

- 26,2 5

- - ~h N j- ELLI PTIC AL

/ p

/ ,, i \

\

CIRCULAR '\

/

l i

I \4

's - -

t - -- -.

2 9 , ..

(', j 4

\

\

.\\

\

\ ,,

\ PLATE A+

N '

'sl N i in

. rg ,8-8 d ,

l

.._/

5 t A+

' 3" io

, l'- i g 1

51 ,_ m s y x x ,

r 3 s'4 as'-oj; j 1 ,,

L.- .

. . _-I2 TORUS ECTION A-A SHELL FI G. 5 -1 smU5UBUM6MhJM5WRTEMM

~

WTELEDGE Technical Report ENCMEEf54GSERVICES TR-5321-1 .

\

\

RING GIRDER Av _

tu l .

s g/

h)~ \ -

/ \,

\ s l

\/ \, ,

1 i

A~ A ,/ -

' ,7

, \

2

/

\f '

i

\

)

l { SUPPORT ' 'y \

COLUMyg \

l FIG. 5 -2 DETAILED gggyG GIRDeg ,SHELL MODEL

WTELJiDGE NSERVICES Technical Report TR-5321-1 MITRE JOINT OFFSET ~ ~

RING GIRDER e-J'  ?

iP f-g x--

> /

f/p Tk~~~

-Y,&i _

I y f. I  !

/gi' l

ggx .

4de%x  :

l i

• h

~

p;

T 4 xl 4

cx s

1 FI G. 5-3 DETAILED RING GIRDER- SHELL MODEL

Technical Report TR-5321-1 s 4 444 C[ '[ * '*0,9 g e\40 ,1's4 : r4,, v4 . s ,,, ,,, en$4

, 4

\,,..,

4 re y, 3., . g g,

) ,3 940 1 ,,8, , g

,,sl I ,,'4 3 t i, m' _.2s

_ *l . ,

s'3% ,'g ,' ,'[14 4

4 C's 4 2 ,,'b c 4'sbs s

,[ s 4't h

. 420 433 /

~

eC; eg 3 3 A- *

/ 4)Pd ** b-l)$

j f',,911 ' 434

~._ ,

4 ,01, g ,'452 4 IS',", ~

\ .~," 4. & P 418 433 ',a 400, ' ,

,#[11

- g- . ."- g_ .

y,- - - - g- .

25' ' 'Ca -

sG '.a -

' tid ' 4'O '

' Sci ' 4I,1 -

-YE 'I5-4  !

-(CI- ' 41,s - >

' ' ' '- QUENCHER & VENT HEADER

,"fC 3, ,

-4*C 1 ", '

,4cz,

, $0',

ATTACHMENT AREA ,sh-s 4YD,

, *ICO,% s ,4 N ,

,439f s

,499, ei n '

1. s 474 '

,49}, s,414 r eSy ' 415

,49 1 s 411 '

/**,T,y . fff28spQ:pzr.aczP25fp s .? ,'s"s

. t 7 7- .. r-e- gg s

,esy../' 7 'u

,, 4, eg.,j.,,6,epe)<sb8kj The model has additional elements, (not shown),

which are for more detail in the lower ring girder section, the torus-to-column weld joint and the saddle.

FIG. 5 -4 DETAILED RING GIRDER - SH ELL MODEL RING GIRDER ELEMENTS

Technical Report #@F WE TR-5321-1 ENGNEERIPG SERVICES 6.0 TEE-QUENCHER AND SUPPORT The following results for the tee-quencher and supports are conservative due to the combined effect of several factors, three of which are; e The calculational methods to determine applied loads improved af ter this analysis was complete, and would provide reduced stresses.

e Some loads were intentionally bounded by conservative values from other plants so a single calculation could be used for more than one plant.

e For submerged drag loads, individual frequency components were added to produce maximum stress without regard to load direction.

The effect of these conservatisms vary among stresses, but can be significant in some cases.

6.1 Structural Elements Considered The configuration of the quencher and support is shown in Figure 2-l 8. Fitzpatrick has eleven discharge lines, each enters the pool vertically.

The structural elements considered in this section include:

I e The quencher.

! e The submerged portion of the SRV line.

j e The quencher support beam and attachments.

Technical R: port WTF1 WE TR-5321-1 g 6.2 Computer Models The computer model used in this analysis is shown in Figure 6-1.

This is a STARDYNE beam model which represents all piping and struc-ture beh:;en the drywell jet deflector and the ring girder. For these analy-ses, the ring girder was assumed rigid and the vent pipe penetration was represented by a stiffness matrix which was developed from a finite element model of the penetration. Releases were modeled between the quencher and support plates to allow for free rotation of the quencher arms in the sup-ports.

This model was used for both static and dynamic analysis.

6.3 Loads Analysis 6.3.1 SRV - Load The calculation of stress due to SRV blowdown was done by applying the dynamic loads to the computer model and calculating the time-history response of the system. The applied loads included both the blowdown forces on the piping and the water clearing forces at the quencher. The l controlling condition was for a second, multiple valve actuation af ter an SBA/IBA break (SRV case C3.3). This case produces a high reflood level at the time of the second actuation and produces maximum load on the support system.

Loads for this analysis were developed using G.E. computer program RVFOR-04 and RVRIZ. The second actuation was assumed to occur at the point of maximum line reflood.

6.3.2 Pool Swell Loads l

The effects of pool swell jet and bubble loads on the quen-l cher and support system were conservatively estimated by static analysis and a dynamic load factor of 2. It was clear from this' analysis that combined pool l swell events would not control stresses - no further analysis was done.

l

Technical Report WM TR-5321 g 6.3.3 Chugging Loads Dynamic analysis of the quencher and support system was done for drag loads due to pre-chug, post chug and chugging FSI. All these analyses were based on a set of harmonic analysis which provided results for all steady-state frequency excitations from 1-31 Hz. Results for individual load conditions were determined by scaling individual frequency results of the computer analysis by the appropriate pressure amplitude.

The mass of the structure used in the computer analysis was adjusted to account for the "added mass" effect of the surrounding water. For FSI and post chugging analyses, individual frequency components were combined by adding the five maximum frequency contributors to the SRSS sum of the others(seeReference12fordiscussion). The maximum value of each frequency component was used in the combination, regardless of vector direction or time of instantaneous response. FSI loads were calculated by considering the calculated local accelerations in the pool due to response of the entire shell.

6.3.4 Condensation Oscillation Loads I The quencher and support system are subjected to conden-sation oscillation drag and CO-FSI drag. Analysis for these loads was based

. on the same harmonic analysis discussed in paragraph 6.3.3, scaled to the CO l

amplitudes. Each of the three C0 spectra shown in Figure 4.4.1-1 of Reference 1 were considered.

i All other discussion from paragraph 6.3.3 for chugging applies to the condensation oscillation analysis, except that the final load was. determined by adding the four maximum frequency contributors to the SRSS sum of the others.

6.3.5 Other Loads Calculations of stress due to weight, thermal and seismic loads was done by using the computer model in Figure 6-1 and static analysis.

Pressure stresses for the piping and quencher were calculated by hand.

l t

Technical Report YE WE TR-5321-1 ENGNEERING SERVICES 6.4 Results and Evaluation The results reported in this section are conservative depending on the effect of factors discussed in Sections 1.0 and 6.0 of this report.

6.4.1 Tee-Quencher _

The controlling stress in the tee-quencher, itself, occurs in the ramshead between the quencher arms. It is the result of load case 25 of Table 1. It includes:

Design Basis Accident + Seismic + SRV The controlling stress is:

STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Bifurcated Bending 26,292 psi 37,440 psi Elbow 6.4.2 Submerged SRV Line The controlling stress for the submerged portion of the SRV line occurs in the ~ inclined lines and is a result of load case 25 in Table 1.

This case includes:

Design Basis Accident + Seismic + SRV STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Vertical Bending 25,085 psi 36,000 psi Section Above Reducer

Technical Report TE WE TR-5321-1 NBtNG S8WICES 6.4.3 Tee-Quencher Support The controlling stress that was calculated for the tee-quencher support is the result of load case 25 of Table 1. This case includes:

Design Basis Accident + Seismic + SRV The controlling stress for the beam is:

STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS At the Brace Bending 10,729 psi 36,000 psi Connection

WTE.EME NM Technical Report '

TR-5321-1 s VENT PIPE PEN ETR ATION TEE QUENCHER

, RING GIRDER SUPPORT BEAM '

RING GIRDER SUPPORT BE A M RING GIRDER FIG.6-1 FITZPATRICK AN ALYTlC MODEL_ _. _ _ _ _ _ _ . _ _

TM Tcchnical Report TR-5321-1 MM 7.0 OTHER STRUCTURES 7.1 Catwalk Analysis of the catwalk structure at Fitzpatrick showed that major modifications would be necessary to meet allowable stresses. The necessary modific'ations were judged to be prohibitive and the entire catwalk was removed except for three bays. This partial catwalk was analyzed and that analysis is reported here.

Like the original catwalk, the partial catwalk consists of a hori-zontal frame structure which supports sections of open grating. The partial

' catwalk is supported from the ring girders and is fitted with handrails as shown in Figure 2-16. The handrails are -inch cable to reduce froth drag

~

loads.

7.1.1 Computer Model The computer models of the catwalk are shown in Figures 7-1 l and 7-2 for the original and modified catwalks. They represent the structure '

fo'r one full bay, beginning at mid-bay. They include all of the load carrying structural members, but do not include the grating or handrails. Loads from these elements are calculated and applied to the frame as forces at the points of attachment.

All catwalk analysis was performed on these linear models.

l All analysis used static application of loads, increased to account for dyna-mic amplification, where appropriate.

7.1.2 Loads Analysis Loads analysis for the catwalk was performed for the direct effects of the following loads. Indirect effects due to motion of the ring girder at the attachments points were considered, but judged to be negli-gible.

Technical Report TN TR-5321-1 MM 7.1.2.1 Pool Swell Drag (4.3.4)

Pool swell drag loads are produced as the rising pool envelops the main frame, grating and handrails. Loads on the frame were calculated based on velocities taken from plant unique QSTF movies and the methods in Reference 1. These were multiplied by two to account for the dynamic effect. Loads on the grating were taken from Section 4.3.4 of Refer-ence 1; these loads already include a dynamic factor, since they are based on test data.

7.1.2.2 Pool Swell Fallback (4.3.6)

Pool f allback loads were calculated and applied in accordance with Reference 1, except in unusual cases where f allback loads exceeded upward loads. In these cases, the maximum values of upward load were used for fallback also. Fallback affects the main frame and grating as well as the handrails.

7.1.2.3 Froth Load (4.3.5)  !

Froth loads have their major effect on the catwalk handrails; and, when applied horizontally, can produce high bending stresses in the vertical handrail members. Froth loads were calculated in accordance with Reference 1, except that the froth 1 influence region was redefined using plant-unique QSTF movies. These movies show clearly that froth 1 loads do not reach the catwalk railing; the analysis was therefore performed with froth 2

. loads only.

Except for the handrails, the entire catwalk is submerged before froth loads reach this part of the torus; because of this, froth was only considered on the handrails.

Technical Report TE WE TR-5321-1 NBRNG sot \/ ICES 7.1.2.4 Drag Loads (Support Columns)

The submerged portion of the catwalk support col-umns are subject to loading from drag forces from the following sources:

(a) Pool Swell (b) SRV Discharge (c) Condensation Oscillation (d) Chugging Loads from these sources were calculated and applie.d to the support columns as static loads. The natural frequency of the support was calculated using hand calculations and compared to the fre-quency(s) of each source. The statically determined stress was then multi-plied by a dynamic amplification factor, developed by considering the worst case frequency ratio and the fact that this is a harmonic loading.

7.1.2.5 Weight and Seismic Loads

- Stress due to weight loads were analyzed using static analysis and the one-sixteenth computer model shown in Figure 7-2.

Seismic loads are small and were considered using hand analysis and scaling static stresses.

7.1.3 Results and Evaluation Table 1 allows stresses in the catwalk structure (excluding attachments) to exceed yield and, in certain cases, to exceed ultimate. Our analysis was based on a linear model and all stresses were maintained below the stress at which a plastic hinge would form. Controlling stress and load combination for various catwalk elements are listed here.

Technical Report TN TR-5321-1 MM 7.1.3.1 Main Frame The controlling stress in the catwalk frame occurs in the inboard supporting channel, Point A in Figure 7-2. It is a result of the combined condition that includes:

Pool Swell + SRV + Seismic + Weight (case 25, Table 1)

The maximum stress value is:

TYPE OF ACTUAL ALLOWABLE STRESS STRESS STRESS Bending + Axial 31,500 psi 56,700 psi 7.1.3.2 Support Columns and End Joints The controlling load case for the support column  ;

and end joints includes: I Pool Swell + SRV + Seismic + Weight (case 25)

Resulting stresses are:

TYPE OF STRESS ACTUAL ALLOWABLE STRESS LOCATION STRESS STRESS l

Bending Column 56,600 psi 56,700 psi Under Extension

Tcchnical Report TN TR-5321-1 MM 7.1.3.3 Welds to Ring Girder The controlling load combination for this stress is also case 25:

Pool Swell + SRV + Seismic + Weight For this condition, stresses are:

TYPE OF ACTUAL ALLOWABLE STRESS STRESS STRESS Shear 27,903 psi 42,000 psi 7.2 Internal Spray Header The internal spray header is attached to the ring girders and to a penetration on the shell. It is located at the top of the torus, above the vent header (Figure 2-3). i 7.2.1 Computer Model The computer model used to analyze the spray header is shown in Figure'7-3. It was constructed to allow determination of stresses in a typical multi-span area as well as at branch connections. This is a part of a piping system and piping elements were used in the model. All results were obtained through the use of static analysis, with factors applied to account for dynamic' response.

7.2.2 Loads Analysis The spray header is high enough in the torus so it does not experience direct water impact; froth is the only pool swell related load that is applied.

Technical R: port TN TR-5321-1 MM The motion of the ring girder that results from pool swell loads on the shell was considered but judged to be a negligible input to the spray header. Shell displacement at the nozzle connections was input to the computer analysis.

7.2.2.1 Froth Load (4.3.5)

Froth loads on the spray header were calculated as outlined in Reference 1. The worst stress condition existed for a vertically applied load. The loads were applied statically to the system.

7.2.2'.2 Weight, Seismic & Ring Girder Displacement The effects of weight, seismic, and shell displace-ment were all considered by using the model shown in Figure 7-3 and applying loads and displacments statically.

7.2.3 Results and Evaluation i

The controlling stress for the spray header piping i is a result of load case 19 in Table 1.

This case includes:

Froth (DBA), Weight, Seismic & Shell Motion The controlling stress is:

SPRAY HEADER PIPING STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Tee at Bending 2,420 psi 24,660 psi Branch Line

Technical Report WP WNE

.TR-5321-1 ENG#EERING SERVICES ATTACHMENTS TO RING GIRDER STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS Support Hold- Shear + 14,427 psi 18,0'00 psi down Plate Bending WELDS TO RING GIRDER STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS At Ring Girder Shear + 1,160 psi 18,000 psi Bending 7.3. Vent Pipe Bellows The vent pipe bellows forms the pressure seal between the vent pipe and. torus; and allows for relative motion between these parts. It is illus-trated in Figure 7-4. ,

7.3.1 Analysis Method Analysis of the bellows at Fitzpatrick considered the capa-bility of the bellows to respond to the dynamic motion applied to it; and also the possibility of direct water impact during pool swell (the bellows is inside the torus).

i. The effect of dynamic motion on the bellows was evaluated by considering the manufacturer's rating for differential motion both axially and radially. These ratings are intended to define static differences which occur over a long enough time so that dynamic response of the bellows itself can be ignored.

l l

L

Tcchnical R: port TF WE TR-5321-1 -100- ENGNEERNG SERVICES In the present analysis, both ends of the bellows are exper-iencing dynamic motion; one end is controlled by the vent pipe - the other by j the torus shell. We expect that the dynamic characteristics of the convoluted bellows should increase stresses over their static equivalents. We also expect that the convolutions will produce complex modes anc' stress patterns that will not couple efficiently with specific input frequencies, i.e., high dynamic response is not expected. Further, the "pogo" and " rolling" modes of the convolutions are non-linear, highly cross-coupled modes that would not be predicted by ordinary structural codes.

Our approach to the bellows evaluation for motion input is to compare the maximum calculated difference in dynamic response across the bellows to the manufacturers allowable. We accept the bellows as adequate for all cases where a large margin occurs between the predicted input motions and the static capacity, as stated by the manufacturer.

i Our evaluation of bellows stress, due to direct impact af ter !

I pool swell, began with calculations for pool impact velocity. This analysis '

showed that the rising pool would not reach the bellows for the full P

, case. There was slight impact for the 0 P case and the stress was insig-i nificant.

7.3.2 l_oads Analysis Calculation of vent pipe motion and torus shell motion was done as a part of the analysis work discussed in Sections 3.0 and 4.0 of this report. The analysis of the torus shell in Section 3.0 was based on a computer model of the non-vent bay and, therefore, did not account for the presence of the vent pipe hole, or the heavy shell reinforcement in that area.

7.3.3 Results and Evaluation The maximum differential motion across the bellows occurs as a result of case 25 in Table 1; this case includes:

l

TN 1-1 -101-Pool Swell Pressure on Shell + Water Impact on the Vent System + Vent System Thrust + Pressure + Weig'ht + SRV +

Seismic For this case, the following deflections occurred:

MAXIMUM MANUFACTURERS' DIFFERENTIAL STATIC MOTION ALLOWABLE Axial Compression (in.) .036 .375 Axial Extension (in.) .036 1.125 Lateral Motion (in.) .123 .625 All calculated values are less than 20% of the manu-facturer's allowables. We consider that this large difference demonstrates the acceptability of the bellows, especially if we consider that much of the load is either static or a single-pulse transient (maximum amplification of 2).

7.4 Monorail  !

The monorail is attached to the torus ring girders at about 45 above the water level. It is a non-containment related structure and there-fore in the same category as the catwalk. It is illustrated in Figures 2-18 and 7-5.

l 7.4.1 Computer Model i

i l The computer model used to analyze the monorail is shown in U

Figure 7-5. It is a beam model that represents the monorail through 180 of the torus structure. Two sections of the monorail, 180 apart, were removed to relocate the RHR return lines for circulating the torus pool. The monorail beam model ends are represented by the three foot cantilevered section with tne ends unrestrained.

TGehnical Report TN TR-5321-1 -102- MM All loads were applied to the monorail statically and dyna-mically.

7.4.2 Loads Analysis The monorail is high enough over the pool so that it does not experience direct water impact. The only pool swell related load is froth 1.

As with the catwalk, ring girder motion was considered, but judged to be negligible.

7.4.2.1 Froth Loads The monorail is located in the Froth 1 region of the torus and was analyzed for these loads, oriented to produce maximum stress. This orientation was 45U to the horizontal. Froth loads were calcu-lated in accordance with the methods of Reference 1, and applied statically to the computer model.

7.4.2.2 Weight & Seismic j Weight and seismic analysis was performed using the  !

model shown in Figure 7-5 and static analysis.

7.4.3 Results and Evaluation The combination of froth, weight and seismic SSE (load case 19, Table 1)producethefollowingcontrollingstresses:

MONORAIL BEAM STRESS STRESS ACTUAL ALLOWABLE LOCATION TYPE STRESS STRESS At Support Bending 37,310 psi 42,280 psi

Technical Report TR-5321-1 WM

-103- MM MONORAIL BUILT-UP COLUMN STRESS ACTUAL ALLOWABLE TYPE STRESS STRESS Axial + Bending 54,160 psi 57,990 psi WELD ON RING GIRDER STRESS ACTUAL ALLOWABLE TYPE STRESS STRESS Bending + Tension 53,067 psi 57,290 psi 1

WTEEME NSBMCES Technical Report TR-5321-1 -104-VENT BAY , RING GIRDER AT.TACHMENT POINTS QNON-VENT BAY

. RING GIRDER ATTACHMENT POINT i

l l

l SR2LL

ATTACHMENT l

POINT i

l FIG.7-1 UNMODIFIED CATWALK COMPUTER MODEL AT 315 AZ FITZ PATR ICK i

Technical Report Tn-o321-1 105-VENT BAY RING GIRDER-ATTACHMENT POINTS 4

S QNON-VENT BAY C

. RING GIRDER ATTACHMENT POINT f

SHELL ATTACHMENT POINT FIG . 7- 2 MODIFIED CATWALK COMPUTER MODEL AT 315 AZ DDD2M45Utu222

SPTE. EDGE ENCNEEMGSEMCES Technical Report TR-5321-1 -106-

[ PENETRATION X211 A ANCHOR e$s$ 26 q27 28 H2 ANCHOR WELDED TO WELDED TO aING Gi DER Y

~

T w

t ,

ANCHOR b E- -N 1

H 25 H 26,27,28 f

FIG. 7-3 INTERNAL SPRAY HE ADER DOUBLE TORUS BAY COMPUTER MODE L-FITZPATRIC K

Rep rt TR- 321-1

.123 CALCULATED BELLOW S * ,,,.f.036 MOTION y 9

$' .i2#3

[

.N 'N

\\

1 i

gv) v i

, EL. 2 4 4'- 9"

~ _ . . . . . ._

e t

4 FIG.7-4 M AXIMUM VENT PIPE -TORUS SH E LL REL ATIVE MOTI ON ZERO - AP POOL SWELL- FITZ PATRICK '

W TELE WPE Technical Report EN NEEfWG N TR-5321-1 -108-3 @. REACTOR

^v 3'

7

/

/

it

/v

=

BE AM isiaxs4.71 M

3 c/ _ MONORAIL g 2! ~ R ADIUS:63'-O"

>- CC . TORUS

_b)L l

l

! FI G. 7- 5 TORUS MONORAIL 180 COMPUTER MODEL-FITZ PATRICK

Technical Report YE WE TR-5321-1 -109- ENGNEERNG SERVCES 8.0 SUPPRESSION P00L TEMPERATURE EVALUATION The Mark 1 mod'ification which added tee-quenchers at the discharge end of

-the SRV lines required that we consider the high temperature performance characteristics of these devices. An important consideration in high tem-

.perature performance, is the mixing characteristics of the device and the attendent local-to-bulk temperature difference (At).

In response to these concerns and to assure reliable operation of these devices, the NRC has set limits on maximum pool temperatures for tee-quencher operation,.as well as guidelines for a temperature monitoring system for the suppression pool. TheserequirementsarestatedinNUREG0661(Reference.2).

8.1 Maximum Suppression Pool Temperature Reference 15 presents an evaluation showing the local temperature does not exceed the maximum pool temperature limits for tee-quencher opera-tion at different flow rates, and for several different plant conditions. The

~

evaluation of the Fitzpatrick Plant for these conditions was done by General f Electric Company under contract to the Power Authority of the State of New  !

York. The results of that work are reported here.

The local pool temperature limits for Fitzpatrick, and in accord-ance with NUREG-0661, is 200 F. The General Electric evaluation of the Plant for the following seven conditions shows the limit is not exceeded:

> lA Stuck-open SRV during power operation with one RHR loop available.

1B Stuck-open SRV during power operation assuming reactor isolation due to MSIV closure.

2A' Isolation / scram and manual depressurization with one RHR loop available.

Technical Report TE NE I TR-5321-1 -110- ENGNEERNG SERVICES 2B Isolation / scram and manual depressurization with the failure of an SRV to reclose (SORV).

2C Isolation / scram and manual depressu-ization with two RHR loops available. This case demonstrates the pool temperature responses when an isolation / scram event occurs under normal power operation, i.e., when all systems are operating in normal mode.

3A Small-break accident (SBA) with manual depressurization; accident mode with one RHR loop available.

3B Small-break accident (SBA) with manual depressurization and f ailure of the shutdown cooling system.

Table 8-1 summarizes the results of the evaluation of the condi-tions listed.

8.2 Pool Temperature Monitoring System 7 i

i The NRC criteria also presents guidelines for a monitoring system to constantly monitor pool temperature. A monitoring system has been installed at.Fitzpatrick which uses a network of 16 RTDs set in thermowells in the torus wall. The RTDs are hardwired to display consoles and an alarm in the control room. The system is described more fully in Section 2.2.1 of this report and is illustrated in Figures 2-9, 2-14 and 2-21.

4 1

(

am, l Number Maximum Maximum Maximum Local ",

4 $

Bulk Pool Pool g 3,.

of SRVs Cooldown Manually Rate 'i

Case No. Event Opened ( F/hr)

Tempe5ature-( F)

Tempe58ture

( F) d" E

1A S0RV at Power, 1 RHR Loop 2 1160 157 193 1B SORV at Power, Spurious 1 570 172. 198 Isolation, 2 RHR Loops l .

2A . Rapid Depressurization at 5 1500 166 203 w/o end cap holes l

Isolated Hot Shutdown, 195 w/end cap holes-i -

! 1 RHR Loop 2B SORV at Isolated Hot 1 570 174 194 ,

Shutdown, 2 RHR Loops C

\

T j 2C Normal Depressurization 5 100 179 193

at Isolated Hot Shutdown, 2 RHR Loops 3A SBA-Accident Mode, 6 (ADS) 5000 163 196 f i j 1 RHR Loop  ;

3B SBA-Failure of Shutdown 5 100 178 188 Cooling Mode, 2 RHR. Loops j l

OWhen the main condenser becomes available.

h  !

I i

l TABLE 8-1  ;

)

SUMMARY

OF.RESULTS FITZPATRICK P00L TEMPERATURE RESPONSE l

i

W F W NE 5 -

-112_ ENGNEstlNG SERVICES REFERENCES

1. G.E. Report NED0-21888, Rev. 2, " Mark 1 Containment Program Load Defi-nition Report", dated November 1981.

2. NRC " Safety Evaluation Report, Mark 1 Containment Long-Term Program",

NUREG 0661, dated July 1980.

3. G.E. Report NED0-24583-1 " Mark 1 Containment Program Structural Accept-ance Criteria Plant Unique Analysis Application Guide" dated October 1979.

4. G.E Report NE00-21944 "

...k Scale 2-D Plant Unique Pool Swell Test Report" dated August 1979.

5. G.E. Report NE00-24615 "....k Scale Suppression Pool Swell Test Program:

Supplemental Plant Unique Test", dated June 1980.

6. G.E. Report NEDE-24840 " Mark 1 Containmenti Program - Evaluation of Har-monic Phasing for Mark 1 Torus Shell Condensation Oscillation Loads" October 1980.

7. G.E. Report NEDE-24519-P " Mark 1 Torus Program Seismic Slosh Evaluation" dated March 1978.

8. G.E. Report NEDE-21968 " Analysis of Vent Pipe - Ring Header Inter-section" dated April 1979.

9. Deleted.

l

10. G.E. Report NED0-24578, Rev. 1, " Mark 1 Containment Program - Plant Unique Load Definition - James A. Fitzpatrick Nuclear Power Plant, dated April, 1981.

Technical Report TF WE TR-5321-1 -113- ENGNEERNG SERVICES REFERENCES (CONTINUED)

11. ASME B&PV Code,Section III, Division 1, through Summer 1977.

12. Structural Mechanics Report SMA-12101.05-R001, " Design Approach for FSTF Data for Combining Harmonic Amplitudes for Mark 1 Post Chug Response Calculations", dated May, 1982.

13. Mark 1 Containment Program Report WE8109.31 " Buckling Evaluation of a Mark 1 Torus", dated January, 1982.

14. Structural Mechanics Assoc. Report SMA-12101.04-R003D, " Response Factors Appropriate for Use with C0 Harmonic Response Combination Design Rules", dated March, 1982, pg. 3.

15. G.E. Report Nr.DC-24351-P " James A. Fitzpatrick Nuclear Power Plant Suppression Pool Temperature Response", dated August 1981.

16. Welding Research Council Bulletin No. 107, " Local Stresses in i

Spherical & Cylindrical Shells due to External Loadings", dated August 1965.

17. Welding Research Supplement, " Local Stresses in Spherical Shells from Radial and' Moment Loadings", P.P. Bijlaard, dated May 1957.

18. "On the Effects of Tangential Loads on Cylindrical & Spherical Shells", P.P. Bijlaard, Unpublished, Available from PVRC, Welding Research Council.

I 19. James A. Fitzpatrick, Final Safety Analysis Report Update,1978, Page 16.7-7.

. ___---_ _ , _ ,_w_.

TABLE I '

• r ,

say say ssA ssA + to Saa + saw Est + say + F4 .

Invi cNf ?*1rmi + lsA isA + (Q 154 + Sn 184 + SkV + tQ DBA DBA + E4 CSA + Say DBA + E4 + Sky aq 84 l co, Cu Co.Os Co.

CI Co.Of 75 (4)

Co.

Of r5 '

l CD.CN Ps Co.

Cu ts Co.Os ui =r OJ :3 TVrt of F.At N EA*& O 3 0 S 0 5 0 3 0 3 0 5 0 5 0 3 0 5 fu -**

Cu13tNATIb4 %Lhal&

3 3 J _-

4 ) _

[ _ _

'3 _h _

31 3) 1) [ l$ 36 If 18 19 20 21 22 23 24 25 26 27 Yb

~

M Normal (J) 3 I I I I I I I I I I I 2 E 1 I I I E I I I I I E I I I fD Elfthquabe gg g } 3 g g 3 g g g g g g g g g 3 3 g o Siv Disclisrae 5tv I I I I I 3 I g 3 3 d I 1 I I c+

W A h ermal T A

g I I T T 3 g g g g g 3 g 3 g I g I 3 I g 3 I E LIEA Seactione 8, 1 I I I I I I I E I I I I I I I I E I & I I I I

~'

IDCA Ques t-Static F fressers #

I I I I I I I I I E I E E I I I I I I I I I I I

~ ~ ~ ~ ~

rg

~ ~

l 1CCA Foot Swe11 I I I I E I 1 v a e w e.sett rCO Oscillation , .I 3 3 I 3 3 3 I I I I I

!.OCA Q.wggleg P gg 0

g 3 3 g g I I I I I 1 I .

57tvfUtst Ft rNfvT g ,

f: tercel Clgig E Torus. Enternal

[

O Veit Pipe. Se!- ,

lows. Drp. ell (at l Vent). Attachaeot 1 A B. C A & B C a C A *& 8 C 5 C A A B C 8 C C C C* C C C lolJo. Terma ( 3, (3 5.pporte. Salemic 6) 6)

Rastrelate Esternet want pape Cexeral soJ 2 & S C A A B C 8 C A A B C B C A A B C 3 C C C C C C C Attachseat Welde (3 ( 3.-

$1 S)

At Feastrettone 3 A B C A A B C 3 C A A 8 'C 8 C A A B C 5 C C C C C C C 1

l (2.g.. SeeJer) -

~

(3) (1)

Veer nesfer Ceneral anJ 6 A B C 'A A B C B C A A B C 5 C~ A A 8 C B C C C C C C C Attscheent WeIJe (3, (3 S) S)

At Fenettettens 5 A 3 C A A B C 8 C A A B C 8 C & A B C 5 C C C C C C C (3 3.. Downcosers) (4) (4) (4) (4) (3 (4) (3 (4) 4.5) 4.5)

Do m osers

• Cessret saJ 6 4 8 C A A B C 5 C A A B C 8 C A A 8 C S C C C C C C C l Attact.senat WelJe (). 43

$)  %)

lattrnst h porte 7 A Y C A A 7 [ B C A A B C 8 C A A B C 5 C C C C C C C hi!!La LJ LIII!vltt , i Cencral 8 A 3 C A A r 0 C D C C D E D E C E E E E E E E E E E E Vent Deflector  ? A 3 C A A C D C D C C D D D D D D D D 3 D D 0 0 0 D

_C .

0 l

l

~

l

~ - - - - - - -

Technical Report YE WE TR-5321-1 -115- ENGNEERNG SERVICES TABLE 2 PLANT PHYSICAL DIMENSIONS FITZPATRICK TORUS Inner Diameter 29'6" Number of Sections 16 Shell Plate Thickness Vent Pipe Penetration 1.125" Top Half .568" Bottom Half .632" SUPPORT COLUMNS Quant,ip Size Outer 16 12WF161 structural shape with fabricated Inner 16 I section near top Base Assembly Sliding, not anchored RING GIRDER  !

Quantity 16 Size T-Beam (8" x 1.5" Flange, 1.5" x 27.5" (Average) Web)

EARTHQUAKE RESTRAINT SYSTEM Quantity 4 Type Shell Mounted - Pinned to Floor DRYWELL VENT' SYSTEM Quantity Size Vent Pipe 8 6'9" I.D.

Vacuum Breakers 5 30" I.D.

Vent Header Support Columns 16 pairs 6" Sch. 80 Downcomers 96 24" 0.D.

Minimum Submergence 4.29'

, TGehnical Report TE @E l

TR-5321-1 -116- N W/lCES TABLE 3 PLANT ANALYSIS INFORMATION FITZPATRICK Seismic Acceleration Values (G's)

OBE SSE Vertical .06 .10 Horizontal .08 .15 Effective Water Mass for Horizontal Seismic Load (Reference 7) 22.1%

Effective Water Mass during Pool Swell Uplift (Reference 4)

Full /SP - 71%

Zero 29) - 63%

Plant Unique C.O. Multiplier (Reference 1)

.892

Technical Report TN TR-5321-1 -117- ENN SERVCES TABLE 4 SRV LOAD CASE / INITIAL CONDITIONS Any One ADS

• Multiple Design Initial Condition Valve Valves Valves 1 2 3 1 N0C*., First Act. A1.1 A3.1 A 2 SBA/IBA,* First Act. A1.2 A2.2 A3.2 3 DBA,* First Act. A1.3 1 NOC, Subsequent Act. C3.1 SBA/IBA, Sub. Act.

C 2 Air in SRV/DL C3.2

, SBA/IBA, Sub. Act.

• Steam in SRV/DL C3.3 (1) This actuation is assumed to occur coincidently with the pool swell  ;

event. Although SRV actuations can occur later in the DBA accident, the j resulting air loading on the torus shell is negligible since the air and water initially in the line will be cleared as the drywell to wetwell A P increases during the DBA transient.

• ADS = Automatic Depressurization System NOC = Normal Operating Condition SBA = Small Break Accident IBA = Intermediate Break Accident DBA = Design Basis Accident

.Tr.chnical Report TF WE TR-5321-1 Al-1 ENGNEERNG SERVICES APPENDIX 1 Use of SRV In-Plant Test Data for Analysis Test Data The in-plant SRV tests used to support structural analysis were run at Fitzpatrick in April,1982. The data was collected in a series of four tests, each consisting of one actuation with a cold line and a second about one minute later (hot line). Each of the four test pairs was run at approximately one week intervals.

The torus shell was instrumented with a combination of strain and pres-sure transducers as shown in Figure Al-1. Strain gages were mounted in pairs on both sides of the shell to allow separation of bending and membrane stresses. Additional gages were located on the tee-quencher and support (Figure Al-2), vent header support column (Figure Al-3), and downcomers (Fig-ureal-4). Pressure transducers were mounted on the shell and ring girder as illustrated in Figure Al-1.

Two independent data collection systems were used to provide a check on system accuracy. The major system was a multiplexed FM tape system on which all data was collected. The second system was a hard wired oscillograph to produce direct, quick-look readout on several channels.

In all, 100 transducers were used during the testing; as follows:

Strain Gages Shell 42 Internal Structures 28 Pressure Transducers Shell 19 Ring Girder 8

Technical Report TN TR-5321-1 Al-2 N set \/ ICES DCDT's Shell 1 Attached Piping 2 Use of Data - Applications The SRV test' data was used to calibrate computer analysis of the shell and support systems and also to establish actual numbers for SRV drag loads on submerged structures.

Use of Data - Shell & Support System Analysis Evaluation of shell stress and support system loads due to SRV actuation was done with a large detailed computer model as discussed in para. 3.2.4 of the report. Data collected from the in-plant tests was used to define the actual shell pressures and decay time for a benchmark (test) condition and to develop correction factors between these measured results and values pre-dicted by generic analytical methods. The steps involved are these:

i

1. Determine maximum average shell pressure, average frequency and j waveform for the four cold tests.

2. Calculate these same quantities for the test conditions using the generic computer programs (QBUBS 02).

3. Calculate calibration factors relating predicted-to-actual pres-sure and predicted-to-actual frequency.

4. Calculate predicted pressures and frequencies using the generic computer program, for other SRV conditions.

5. Apply the calibration factors calculated in step (3) to all other predictions for pressure and frequency. The duration of the pres-

Technical Report TN TR-5321-1 Al-3 mg sure transient, as measured in the test, is affected proportionally by the frequency correction and used as the basis for all computer model loading.

Verification of Computer Model The test data was also used to verify the accuracy of the computer model.

This was done by the following method:

1. The computer model was loaded with the measured shell pressures.

2. The model was run and stresses at all strain gage locations were calculated.

3. Comparisons were made between computer predicted shell stress and measured shell stress at the same points.

Correlations for shell stress were excellent - generally within 5%. [

Correlations to column loads were not so good - generally off by about 50%. l This cifference in computer results for test conditions was handled by devel-oping a second calibration factor for supports only, and combining it with the previous pressure calibration factor. The results were two different cali-bration factors to be applied to final analysis - one for the shell and one for the columns. The factors developed and used are:

Shell pressure = .34 x predicted l Support load = .4 x predicted Multiple Valve Contributions For cases where more than one valve actuates, the contributions from other valves were added directly (same signs). The maximum value used was 1.65 x the pressure from a single valve (Reference 2).

TGchnical Report TN TR-5321-1 Al-4 6M SRV Test Data for Drag 1.oads The data collected during the Fitzpatrick in-plant test included strains and pressures measured on submerged structures. Strain gages, positioned to

.show bending stress due to drag loads, were installed on the downcomer and vent header support column. Figure Al-4 shows the locations of these gages, relative to the quencher. Pressure transducers were mounted on the ring girder. The test data showed these results:

1. Structural response occurred at the natural frequency of the struc-ture only.

l 2. Responses were much less than would be predicted by Program analy-sis methods - as much as an order of magnitude lower.

The data collected from Fitzpatrick was evaluated along with the data collected by TES in three other in-plant tests. The matrix of data collected is as follows:

Ring Catwalk Vent Girder Supports Column Downcomer (Pressure)

Millstone X X X Nine Mile Point X X Vermont Yankee X X Fitzpatrick X X- X An important consideration in the application of this data was the pos-sibility that resonant structural response might occur at some other SRV condition. This was considered and dismissed based on two separate arguments; they are:

1. If a major frequency component existed in the drag force, it would be detectable on each of the structural responses for a given test.

This did not occur.

l 1

Technical Rep;rt TE NE TR-5321-1 Al-5 ENGREERNG SERVICES

2. The response frequencies of the structures tested (structural natu-ral freduencies) i ranged from 8.1 to 38 Hz.* If any single strong frequency existed.in the drag load, one of the structural responses should have demonstrated some degree of resonant response - none did.

We conclude from this that the structures involved are responding to a fairly uniform random field and that the test data represents usable data for all SRV conditions.

The next step in the process was to calculate an equivalent static load for each structure. This is the static load that produces the same bending stresses measured in the test, when applied uniformly to the submerged area.

These static pressure values were plotted against distance from the quencher and Figure Al-5 was developed. This curve represents the equivalent static drag pressures, including quencher jet loads. It is scaled upward from test conditions to more severe SRV cases by the ratio of the calculated shell pressures for the two casas, for application to structures under different ,

loading conditions.

• Actual values were U.1, 8.2,14.5,15, 21, 23, 24, 25, 29, 30, 34 and 38 Hz.

M Technical Report TR-5321-1 Al-6 i n e pll - B AY D

\ RG4\

., n

=-- BAY C 1

DC DT UNDER P5 TO p MEASURE SHELL nFlo DIS PLACEMEN T Pig,fV. SRV LINE"L"

/ / ,

/ / i

'/RG6 f ERG 8 PI, f RG5 e RG7 l

NP2 p[ f R4 B AY "B"

/ P7'k R7 P2Aok P5 /

R2 P5 k i P7A g P3 / P8' 2 LTO GIRDER-TYP.

6 g P8A p

R6 e

P9 P9A e PRESSURE 14 ROSETTE - INSIDES OUTSIDE FIG. Al-1 PRESSU RE 8 SHEL L GAGE LOCATlONS -FITZPATRICK

~

TR-532$.1

~

4 G AGES AT 90 SRV

, , , GAGES

' i i g i.

- -3" - -i--

m; i

QUENCHER SUPPORT B ^

B I- 'C FIG. Al-2 SRV TEST -TEE QUENCHER a SU PPORT GAGES F1TZPATRICK

6 WTBEME N SERVK115 kS32$15 Al-8 l

l f% +

nvn RPV 4 GAGES ,

90' APART QUENCHER {_

e .

FIG. Al-3 SRV TEST-VENT HE ADER SUPPORT GAGES

Technical Report M TR-5321-1 Al-9 MID LENGTH e RPV 4 GAGES 90' APART OVER P7 N #

l xp 1 C l FIG. Al- 4 DOWNCOMER GAGES

Technical Report TR-5321-1 Al-10 4

~ SRV CASE Al .1

$" NM w

Z D

y VY w

9

$ O l-- 1 O m MS FP o  !

y MS VY FP NM o 4 W

I MS O 2 4 6 8 HORIZONTAL DISTANCE FROM BUBBLE (f t)

EQUIVALENT SRV DRAG FROM IN PLANT TESTING FIG Al-5

WM l- A2-1 APPENDIX 2 Discussion of 32 Hz Frequency Cut-off for Condensation Oscillation and Post Chug Analysis TES made the decision to limit C0 and Post Chug response analysis to frequencies below 32 Hz early in the program. The decision was the result of several considerations that led to the conclusicn that the 32 Hz cut-off would produce realistic results.

The basis for use of a 32 Hz cut-off involved strong fundamental argu-ments, both in the loads used for the analysis, and in the stress analysis itself. The primary arguments are different for CO, and for Post Chug, and are given here:

For condensation oscillation analysis.

1. Load Definition - A PSD study of the C0 pressure data showed that '

frequencies above 25 Hz accounted for only 10% of total power (Ref-l I

erence 1, page 4.4.1-10). This means that a system with flat frequency response to 50 Hz would suffer a 10% unconservative stress error if a 25 Hz cut-off was used. Since we are using a 32 Hz cut-offandoursystemishighlyresponsiveatlowfrequencies(not

! flat), we should expect a much smaller error.

I

2. Structural Response Analysis - The relative importance of loads i

below and above 32 Hz can .be judged based on examination of the modal frequencies and generalized coordinates of the structure in both frequency ranges. If we consider the characteristics of a typical torus model in these ranges, we find:

l

_ J

TM l A2-2 Number 0f* Number 0f Max.

2 2 Number of GX GXd Value Frequencies >p 1000 >2000 GXf2 Below 32 Hz 44 25 14 167,858 32-50 Hz 34 5 1 4,594

• Product of generalized weight and.the square of the participation factor 'used as an indicator of modal response strength.

These figures show that for condensation oscillation, frequencies below 32 Hz clearly dominate the response and frequencies above 32 Hz are relatively insignificant. They provide a strong indication that the 10% worst-case unconservatism discussed above will be greatly reduced by the selective nat-ure of the structural response. We should logically expect the structural response characteristics, and the f act that we are using a 32 Hz cut-off, instead of 25, to reduce the 10% maximum error to less than 5%. An error of this magnitude is consistent with other assumptions which must be made in the  !

analysis and is considered acceptable. ,

A further statement regarding the validity of this approach may be found in References 11 and 14.

For the Post Chug load, the second consideration of structural response is also valid, but the load definition is not as heavily skewed toward the low frequency end as is C.0. The decision for handling post. chug was heavily i influenced by the fact that it produced very low stress. This is discus' sed further in Section 3.2.3.2 of this report.

Technical Report TN TR-5321-1 A3-1 MM APPENDIX 3 C0/CH Drag Loads for Ring Girder Analysis ,

TES did not follow the calculational methods of NUREG 0661 (Reference 2) for calculation of C0/CH drag loads on the ring . girder. This appendix describes the method that was used, the differences from the NUREG method and the basis for the change.

The.NUREG analysis method specifies that acceleration drag forces (and effective hydrodynamic mass) for flat plates be based on an equivalent cylin-der with radius equal to 2 times the radius of the circumscribed circle. It also specifies that the drag forces be increased by an additional factor of 2 for structures attached to the torus shell, to account for wall interference.

Application of the NUREG criteria produces a factor of 4 multiplier for drag force for flat plate structures in the fluid; and a factor of 8 multi-plier for flat plate structures in the fluid and attached to the shell. These values are referenced to a' drag force equal to 1.0 for flat plate calculations based on potential flow theory and neglecting interference effects.

These increases in loads are supported by data available in Reference A3-1 and A3-2. Keulegan and Carpenter show in Reference A3-1 that the drag forces on a plate in an oscillating flow may be as much as a factor of 4 higher than the theoretical force based on potential flow. Sarpkaya shows in Refer-ence A3-2 that forces on a cylinder near a boundary, may be twice as high as forces away from the boundary.

Technical Report TN TR-5321-1 A3-2 N S8WICES Both References A3-1 and A3-2 present results as a function of the VT/D ratio where:

V = maximum velocity T = period of flow oscillation D = diameter Keulegan and Carpenter (Reference A3-1) show the effective hydrodynamic mass coef.ficient for a plate varies from a maximum of 4 at = 125 to 1 at VT/D = 0. (pure potential flow). Sarpkaya (Reference A3-2) shows an increase in the hydrodynamic mass coefficient for a cylinder near a boundary that varies from a maximum. factor of 2 at h = 15 to a minimum of 1.65 at VT/D =

NUREG 0661 appears to use the bounding values from both of these refer-l ences to formulate its' analysis method. It implies by this that large values l of h will exist in the torus. In fact, this is not true for C0 and CH drag loads.on the ring girder. For this structure, under this load, V_T D

ratios are near zero and the use of maximum multipliers should not be neces-sary. It is on this basis that we have used an alternate method to calculate C0 and CH drag loads on the ring girdar.

The TES method to calculate these drag loads on the ring girder used the same references as above (A3-1 and A3-2), but accounted for calculated values of gVTrather than the values corresponding to the maximum increases. Consid-sideration of the actual f ratio for wall interference led to an interference factor of 1.65 (instead of 2).

Technical Report TN TR-5321-1 A3-3 M SB{\/ ICES Low values of fsuggest that the theoretical hydrodynamic mass coefficient for the ring girder is appropriate. The theoretical coefficient for this structure is estimated by an equivalent cylinder with a radius equal to the circumscribing radius. Use of this cylinder results in a hydrodynamic mass coefficient equal to two. The total factor used was related to the NUREG multiplier by:

1 2.0 1.65 gx 2.0 = .41 The factor used by TES was .41 x the NUREG 0661 factor.

?

I i

Technical Report TN TR-5321-1 A3-4 66 REFERENCES A3-1 Keulegan and Carpenter, " Forces on Cylinders and Plates in a Oscillating Fluid," National Bureau of Standards, Vol. 60, No. 5, May 1959.

A3-2 Sarpkaya, " Forces on Cylinders near a Plane Boundary in a Sinusoidally Oscillating Fluid", Journal of Fluids Engineering, September 1976.

I

_ I