LD-88-049, Flow Distribution & Tube Vibration:Evaluation of Sys 80 Steam Generator Tube Lane/Economizer Corner Region
| ML20150E833 | |
| Person / Time | |
|---|---|
| Site: | 05000470 |
| Issue date: | 07/01/1988 |
| From: | ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY |
| To: | |
| Shared Package | |
| ML20150E730 | List: |
| References | |
| LD-88-049, LD-88-49, NUDOCS 8807180001 | |
| Download: ML20150E833 (86) | |
Text
{{#Wiki_filter:. . . . . . . . -- P Enclosure (1) to LD-88-049 .
, , , 78 Pages
-i
< ;o 1
ENCLOSURE 1 FLOW DISTRIBUTION AND TUBE VIBRATION
' EVALUATION OF SYSTEM 80 STEAM
. GENERATOR TUSE LANE / ECONOMIZER CORNER REGION COMPUSTION ENGINEERING, INC.
WINDSOR, CT. 4 t l -- 88o7180001 880701 0 PDR ADOCK 0500
-A
,==. - -, . . - . _ _ . . . , , - . . , , , . . . ~ . _ _ _ ,,________. . . - _ _ _ . _ . _ _ . _ .
.i
.e l
; ,6 TABLE OF CONTENTS l
i
- 1. EXECUTIVE
SUMMARY
. 1-l'
- 2. INTRODUCTION 2-1~
- 3. FLOW OISTRl8UT!0N ANALYSIS J-1 3.1 'Generat~ 3-1 '
3.1.1 Steam Generator Description 3-3 3.1.2 ATH0S !! Code 3-6
. 3.t.3 FLOW 3 Code 3-7 3.2 Palo Verde System 80 Steam Generator Analysis 3-8 3.2.1 ATH0S Analysis of Palo Verde Steam Generator 3-8 3.2.2 FLOW 3 Analysis of Pato Verde Steam Generator 3-9 3.3 Flow ofstribution Analysis Results 3-10 4 TUBE VIBRATION EVALUATION b-1 4.1 Design Basis Evaluation b-1 4.1.1 ~ Generic Vibration Testing b-l' 4 .1.' 2 Palo Verde Analysis b-3 4.2 Downcomer Modet Vibration Test 4-h 4.3 ECT Inspections at Palo Verde h-O 4.4 Eigenvalue Analysis b-10 4.5 Critical Velocity Evaluation b-13
- 5. CONCLUSIONS AND RECOMMENDAfl0NS 5 6. REFERENCES 6-1 APPENDIX A A-1 i
a
, 34 6'
List - of'Fiaures'- m jb Floure No. Title ELLL.
~
21 - System-80 Economizer Region 2-2 22 Economi.or S.G. Tubelane 2b Geometry 5
.., 23 Tube sundle - and tube s'upport 2-6' Geometry
- Vertical Legs 24 Tube Support construction 2-7 Details 3.1 1 Schematic of a Combustion 3-13
- Engineering System 80 Steam Generator.
3.1+2 ' System 80 Steam Generator Cold . 3-lh side Recirculating Fluid
' Entrance Region 3.2*1 ATHOS 11'Model of System 80 3-15 Steam Generator 3.2 2 Flow Distribution Analysis 3-16 M o'd e t Econoriter Local Corner Region of System 80 Steam Generator 11 1
4 s # = , u,- --en---- ,y, y ,,w,, m,w,,m....,,w,,.~~.-r,, .,r, , -~,--~-~.% - , ,.,, ~ . . ., _ . .y.. -- _ _ - ,,-e.--,-~*
t . List of Floures (Cont 8d) FIsure No. TItie Ef.jLg, 3.2 3 Economizer Steam' Generator 3 17 Local Flow Distribution Analysis Results
' 3.2 4 Economizer' Steam Generator 3 18 Local Flow Distribution Analysis Results i
' 3.2 5- -Economizer. Steam Generator 3 19 Local Flow Distributton Analysis Results 3.2 6 Economizer Steam Generator 3 20 Local Flow Distribution Analysis Results 4.1 1 CE Tube Bundle Configurations 4 16 Evaporator and Economizer Sections Water Environment Testing 4.1 2 Response Acceleration 4 17 4.1 3 Critical Damping Ratios vs. 4 18 Frequency Test No.
I l l l-111 l l-( l
. 'i :
e' '
.g
. -f .
List of Floures (Cont'di Fieure No. ' Tit le P,.,313, x 4.1 4 System 80 Steam' Generator b-19 Potential Regions for F.I.V Addressed by Testing and Analys{s 4.1 5 Air Water Flow Testing h-20 ConfIguratton 4.1 6 Air Water Flow Testing b-21 Configuration 4.1 7 off Line Computer Processed b-22 Data 4.1 8 -fube Deflection vs. Fluid h-23 Pressure 4.1 9 Upperbound Effective Force b-23 Pressure 4.1?10 StabitIty Diagram b-2b 4.2 1 System 80 Downcomer Hydraulic b-25 Test Model
- 4.2 2 Downcomer Flow Test Model b-26 Plar View 4.2 3 Velocity Distribution at h-27 Cold Leg DJwncomer Intet 100% F1ow iv
.p. . .= .. .- , ., .~ _
. l
+
l L L'st of Flaures'(Cont'd) i I s: hk ' Finure-No. M
-l 4
M 4.2 4. Tube vibration Amplitude h-28
- i Profile 4.2 5 System ~80 Economizer Eggerate 4-29 Map 4.3 1 Pato Verde 112 Steam h-30 Generator Cold Recirculation Fluid Entrance Corner Region -
Inspection Results 4.3 2 Pato verde 182 Steam h-31 Generator Cold Recirculation Fluid Entrance Corner Region - Inspection Results 4.3 3 Tube support stay. Rod h-32 Assembly 4.4 1 System 80 Tube support L-33 Locations 4.4 2 A.v s Y S Finite Element Model h-34 of Typical Tube Vertical Span with Support Locations l' i l l _ _ . . - _ ~
,{ r f'
+ .
(
\
- Lfst of Floures (Cont'd$ ;
s l.- a Ffoure~ No. Tftie W.
' 4 '. 4
- 3 '
system 80 Modal Analysis k-35 Resulta
. 51 Palo Verde 3 Steam Generator 5-3 Plugging Pattern list of Tables Table No. Title M
, 'T !
2*1 Palo Verde Steam Generators a 2-2 Summary of the Number of Worn Tubes a February 1987 Inspection 311 System 80 Steam Generator 3-4 Operating Conditions - 100% Power 3.3*1 CE System 80 Steam Generator 3-12 Flow Disbribution Analysis Summary vi (, ,i .s
*, pee er ~py ewy --.,,y e s,- -ivy,rpwwe.----m**T---*--*'--wm"'* ~ - * " wY-N"'--'-'-*MF T1 1T*
k f
< k .!! .,)
i-ci SECTION 1 EXECUTIVE
SUMMARY
} .! 1he purpose of this report is to present the basis for closing out the steam generator tube vibration issue for plants which reference CESSAR F. This report, which provides this basis, overviews (1) relevant portions of the System 80 steam generator design / analysis / test program, (2) the tube vitration which was experienced at Pa'lo Verde Units 1 & 2 durirg the early stages of operation, (3) Combustion Engineering's flow and vibrational aastysis program ' which ensued, and (4) the' plugging, staking, inspection, and monitoring program which was implemented as the short and long term corrective actions. Following the discavery of primary to secondary leakage at Palo Verde Unit 1, steam generator tube eddy current inspections were perf3rmed at Units 1 and 2. Wear indications were found at the periphery of the cold side tube bundle adjacent to the economizer divider plate. Subsequently, flow distribution and structural analyses were performed to understand the root cause of the observed tube wear and to determine a potential for further propagation into the tube bundte. Analyses were performed for the system 80 tube and tube support seemetry to determine vibrational characteristics, modal frequencies, and moce shapes for each design. This information, in conjunction with the wear scar data obtained from eofy current examinations, was utilized to evaluate the critical or threshold velocity for onset of tube instability due to flow induced vibrations. The flow distribution analyses using the ATH0S II and FLOW 3 computer codes indicated fluid velocities were higher than the critical velocity of 25.5 ft/sec in the affected region. Analysis also demonstrated that the high velocities dropped off rapidly away from the "corner" region. This indleates the cause of vibration and wear to be very localized in nature and limited to only a few tubes in the corner resfrn. It is expected therefore, that this vibration induced tube wear will not propagate further into '. h e tube bundle. 1-1
This review again demonstrates that the basis for CE-vibration analysis and test. data used for the CE System 80 tube bundle geometry and the vibration
.aqalysis nethodology'used for Palo Verde are valid. The tube wear pinblem at Palo Verde is not due to any breakdown.in data or technique, but rather to a lack of detall in.the calculation of the tube-Lane bypass velocities in the
-Local corner region. The high velocities were invisibte to ATHOS !! since the local region is much smaller than ATHOS 11 node size uc.ed in the original analysis.
a
-In conclusion, combustion Engineering and Arizona Public Service determined the root cause of the tube wear. Based on the results of analysis and' inspections, Combustion Engineering has not identified any design change which would be necessary.to assure their continued safe operation of the steam gener a t ors,~ nor do we have any reason to belleve that any changes will be necessary in the future. The plugging, staking, monitoring, and inspection program assures continued safe operation for all plants which reference CESSAR F. This report documents the final results and conclusions for the System 80. steam generator tube vibration issue. It is requested that this issue be elosed out on the on the CESSAR F Docket.
l i. i 1-2
e e SECTION 2 INTRODUCTION Palo Verde Units ~1 and 2 steam sensrators experienced localized wear at the corners of the cold side recirculating fluid intet region. The cold side recirculating fluid entrance region is shown in Figure 2-1. The corner region is described.in Figure 22, with the arrow depicting the much higher than average local velocities preferentially attracted to the low flow resistance in the tube lane above the divider plate. Table 2 1 lists the affected tubes in the corners of the cold leg sides of four Pato Verde Unit 1 and 2 steam generators. Note that not all tubes in the Figure 2*2 pattern have worn in att the cold side corners in all four steam generators. Based on the summary contained in Table 11, an average of approximately nine (9) tubes have worn in each of the eight cold side corners. The randomness of wear within the affected region is attributed to two factors: (1) variation of tube support clearances near the chord boundary of the economizer half eggerates and (2) the proximity to a tube support stay rod assembly, represented by solid circles in Figure 2 2. The variation in tube support clearance occurs when the thin l eggerate strips are edge welded to the bars which frame the boundary of the i eggerate support at top and bottom causing local distortion of the thin strips. Proximity to a stay rod assembly reduces the flow gap adjacent to the tube from 0.25 In. to 0.0625 In. which tends to further accelerate the fluid through the tight gap caused by the oversized stay rods. i l Two types of analysis have been performed in order to quantify the structural l l and fluid parameters associated with the vibration and wear in Palo Verde Units 1 and 2 steam generators, tube vibration Eigenvalue analysis and secondary fluid flow distribution analysis. The Eigenvalue analysis results, in combina-tion with the characteristics of wear Indications discussed in section 4 explain how the highest tube wear can occur at support Elevation 2 while the recirculating fluid entrance is located between Levels 3 and 4 The flow distribution analysis discussed in Section 3 consisted of a full steam l l 2-1
)
a
' 1 II. ~
t. s..
- l. 6 TABLE 2-1 PALC VERDE STEAM GENERATORS SUlHARY OF THE NUMBER OF WORN TUBES FEBRUARY 1987 INSPECTION
- Range of Wear Depth Unit 1 Unit 2
(% Through Wall) SG #1 SG #2 SG #1 SG #2_
<20 1 4 7 4 20-39 5 1 4 2 40-59 4 3 9 13
, 60-79 2 1 9 2 80-100 1 0 1 0 , t TOTALS 13 9 30 21 GRAND. TOTAL 73 I S 2-2 +
i l l
- 9enerator ATH0$=II analysis and a more detailed FLOW 3 analysis of the economiz- j er local. corner' region.
The flow distribution analysis results.ldentify the
' bypass flow' paths.which bound the affected corner tubes and quantify the much Greater than anticipated local flow velocities in the Palo Verde steam genera-tors.
[, l i e 1. 2-3 l L
4l 1 c - TUBE BUNDLE SHROUD- 'a I- '. k a h 5 Q COLD SIDE CENTER ; SUPPORT l RECIRCULATING CYUNDER 'g FLUID ENTRANCE r 'ECCCRATE" TUBE f % SUPPORT p o- . i
"' i s'
[ , I/ A L k h 4 i I l l I
/
1 i _ .
/ h FLOW DISTRIBUTION
! / PLATE l f ; as -w+ . .. pb #7 l d ' O J L l , ZL t .AiCE)
?///&/// J//f1/ / 6 SYSTEM 80 ECON 0MIZER REGION FIGURE 2-1 1
2-h
BYPASS now mggER I R E /S A / SIDE EC0 ZER 000 OOO I[ essOOO G6000000000000 l se ,
- egeggoseasO
@H OOO i eseaMoH@Xo
@O00000000000 000 GO00000000000 000 @O0000000000 000 9900000000000 000 @O0000000000 000 I 000000000000 000 000000000000 000 000000000000 000 i O00000000000
' 00000000 000 I OOOOOOnnnn OOO 000000 OOO OOOOOd TU!!$NI$0!ETRY 000 000 I OOOOOE @ s^"ca'"E"u?#S OOOOOC l ggg OOO OOOOOg gggg $ Il fid isis' FIGURE 2-2 2-5
/?.O*
p: O ! O . r l D s [f "V \ (j ~ e i Nj
/
l
'J i
V lA'
// i /%
,v i I
Oh[M. LD TUBE 0.0. = 0.75 INCHES 37 l 1.0 INCH PITCH A N TUBE BUNDLE AND TUBE SUPPORT GEOMETRY VERTICAL LEGS FIGURE 2-3 2-G
k k WEDGE
BLOCK
" '~
"EGGCRATE" GRIO l ~
y J
, ...... }
g STRIP m - - _ . ~ -' ' f(2"X0.090") ~- _ . _. f(I.X0.090-) m m O i
.~ J fWnraer SHROUD STRIP WELD l FRAMING BAR (0.875" X 0.875")
~
~.
FULL GRIO LOWER HALF GRID TUBE SUPPORT CONSTRUCTION DETAILS ff
SECTION 3-FLOW DISTRIBUTION ANALYSIS J3.1 General-Combustion Engineering has been involved with steam generator flow distribution code development and applications for'over 15 years. During this time, CE has co sponsored the' development of two such codes by CHAM of London, England, worked with CHAM on the development of a third code sponsored by. Electric Power Research Institute (EPRI), and analyzed
-numerous steam generators of both CE and non CE designs using these codes. > Additionally,-CE has obtained a numbe' of codes developed by various. Laboratories for the U.S. Government.
In 1974, CE and Kraftwerk Union (KWU) of West Germany sponsored the development of-the.. steady state, homogeneous, three dimensional, two phase flow distribution code HELIOS. The code became operational in the Fall of 1975. HELIOS and associated software were able to compute
.the secondary fluid velocity components in three coordinate directions, pressure, temperature / quality, heat flux, tube wall and primary temperature at approximately 1000 grid locations. The code also calculated the macrescopic steam generator parameters, such as the circulation' ration, downconer flow rates, etc. HELIOS was used to analyze and design the CE Systen 80 generator.
-In 1978, the CALIPSOS code became operational, the code was developed by CHAM under CE/KWU sponsorship. CALIPSOS had all the same capability as HELIOS and also offered improved geometrical modeling. In addition, an option provided for slip flow by means of Interphase friction factors and separate momentum equations for the liquid and vapor phases. However, convergence problems when using the slip option restricted the applications to the homogeneous flow model.
3-1 m
In 1978, CHAM in London and Huntsville, Alabama began development of an advanced flow distribution code under sponsorship of EPRI. CE served as a technical advisor to EPRI throughout development of this code, which eventually became known as.ATHOS. Unlike its predecessors, HELIOS and , CALIPSOS, ATHOS had transient and algebraic slip capability. A different numerical procedure provides consistently converged solution, under a wide range of conditions with both homogeneous and slip options. The code has been checked and verffled by CE and others (References 2 through 7). The tube and tube support vibration analysis performed as part of the ASME code (Reference 9) design report utilized CRIBE results because the HELIOS and CALIPSOS codes were not yet verifled. CAIBE (Eross connected Recirculation instabilities in Lofter) is a computer program to calculate the steady internal conditions and has an option to predict the presence of oscillations in the downcomer flow, steam flow, or water level in the secondary side of a steam generator. The CRI8E code was developed by the Bettis Atomic Power Laboratory (Reference 15). Combustion Eng!neering modified the code for the System 80 steam generator analysis. The modified code, CRIBE, is a two flowpath version that provides separate modeling of the hot and cold legs. The CRIBF output includes hot and cold side downcomer flow rates which were used to evaluate susceptibility to tube vibration at the cold side recirculating fluid entrance region. The CRIBE calculated average cold side downcomer entrance velocity of 2.45 ft/see yletded a tube gap velocity of 9.8 ft/sec for the maximum flow rate. The CRIBE and ATHoS calculated overall steam generator characteristics, f.e., circulation ratio, the hot and cold side downcomer flow rates, mass inventories, etc., are in good agreement. However, both codes lack a capability to calculate localized flow conditions, i.e., fluid velocities and other thermal hydraulic parameters within the gaps between the ind ;! dual tubes or the tube lane. 3-2
c.
.s s 1
The eddy current inspections of the Palo Verde steam generators ind{cated tube wear at the corner regions formed by the tube lane and the cold side downconer opening. In order to understand the causes of tube wear a flow distribution analysis.of the System 80 steam generator was performed to determine the fluid velocities in the affected region. The analysis consists of a full steam generator ATHOS 11 analysis and a more detailed FLOW 3 analysis of the steam generator corner region. The FLOW 3 model utflized the boundary conditions from ATHOS analysis and modeled the rectangular tube lane and the annuter gap between the tube bundle and the shroud in detalt. The model more accurately assessed the fluid velocities in the "open" tube lane above the economizer divider plate and the annulus at the cold side recirculating fluid entrance elevation where the frictional resistance to flow is negligible. 3.1.1 Steam Generator Description A schematic of the System 80 steam generator is illustrated in Figure 3.'1 1, and the operating conditions at 100% power are shown in Table 3.1 1. The System 80 generator is a U tube, recirculating steam generator with an oxlat flow economizer. The steam generator consists of 11,012 inverted U* tubes supported by nine full and partlet horizontal tube supports. The secondary fluid is partially evaporated by the primary fluid flowing inside the tubes. The steam is separated from the liquid vapor mixture by the centrifugal separators. Most of the remaining moisture is removed by steam dryers. The dry saturated steam exits the generator through the two outlet nozzles and flows to the high pressure turb{ne. The separated liquid and a small portion of the feedwater flows downwards through the annulus between the tube bundle shroud and the steam
- generator shell, f.e., downcomer. The downcomer fluid enters the tube bundle section at the tubesheet levet on the hot side and above the economizer region on the cold side.
3-3 s~ ,.
t 4 9 '
.e n
TABLE 3.1 1 SYSTEM 80 STEAM GENERATOR OPERATING CONDITIONS 100% POWER 1 Steam Pressure 1070 pela
- 2. Feedwater Flow Rates Economizer 2147.5 lb/see Cold Side Downconer 238.5 lb/see Total 2386 lb/sec
- 3. Feedwater Temperature 450*F
- 4. Primary Flow Rate 22,778 lb/sec
- 5. Primary Pressure 2250 psia
- 6. Thermal output 1906 MW / generator i
J T 3-4 _y+ ,-m: w,.-mm.-,w
" ^
TU. r . - 0< The economizer region is comprl' sed of the cold side tube bundle from the top of the tubesheet to-the cold side downcomer fluid entrance region 96" above the tubesheet. The cold and hot sides are separated by the economizer divider plate up to 103" above the tubesheet. The cold feedwater enters the economizer at the tubesheet level and is heated by the primary fluid inside the tubes as it flows upwards. As discussed in Section 2, most of the affected tubes at Pato Verde Units 1 & 2.were located adjacent to the economizer dividtr plate and'st the periphery of the cold side tube bundle. These tubes seem to have experienced excessive vibration and wear due to high fluid velocities caused by an open tube lane above the divider plate and proximity to the cold side recirculating fluid entrance window. The geometrical details of this region are Illustrated in Figure 3.1 2. In order to evaluate the fluid velocities in the affected region, thermat hydraulic analysis of the System 80 steam generator was performed for the es* designed configuration of Pato Verde System 80 steam generator (Figures 3.1 1 and 3.1 2). 4 The ATHOS 11 code was utilized to determine the overall steam generator performance and to obtain the boundary conditions for the detailed model of the "corner" region which was analyzed using the FLOW 3 code. i 3-5 l.
,.c -
,1b +
V4 3.1.2 ATNot'll Code ATN08 (inelysis of the Lhermal Lydraulles gf iteam Generators) is a three dimensional, two phase, steady state and transient ecde for thermat hydraulic analysis of recirculating U tube steam generators. The 1. code was' developed for the Electric Power Research Institute (EPRI) by CHAM of North America. ATMOS was further modified by CE to incor pora te modeling of plugged, sleeved, removed and putted tubes. The CE version of the code also includem capability to model sludge deposits on the tubesheet. A more detailed description of the mathematical and physical models, finite difference equations, the code structure and solution procedure is presented in Reference 1. ATH0S !! is a state of the art 3D code avaltable for steady state and transient analysis of a steam generator. The code is widely used by vendors.and utilities to evaluate overalt steam generator performance and to understand flow related problet expertenced by opetating steam generators. The ATH0S 11 codo utilites a "porous media" approach and lacks "zoom In" capability, hence, aferoscopic thermat hydraulle charac-teristics, i.e., flow velocities and other parameters within the gaps between the individual tubes cannot be calculated by the code. The ATH0S code has been checked and verffled by CE and others (References 2,3,4,5,6,7). The check out studies included comparing the ATH0S geometry pre processor computed values of the steam generator geometric parameters against hand calculations, checks on the mass, momentum, and energy balances, and consistency and plausibility of steady state and ! transient solutions for a number of different cases. For code verification, measured data from several small scate experiments, modet steam generators, and full scale steam generators were compared with ATN05 results. In one case, an analytical solution was avaltable for comparison with ATH0S calculations. In general, the agreement between the ATH0S results and available global experimental data were good, in addition, consistent trends were found in all parametric studies. More detalls of ATH0S verification may be found in References 2 through 7 and Appendix A. l 3-6
i, . c Most of tne tubos with wear indications are located near the "tube lane" region (Figure 3.1 2), where, due to lack of flow resistance, the fluid velocities are higher than in the tube bundle. The tube lane has a rectangular chape, while the ATH0S 11 code models the steam generator using the polar coordinate system, which is consistent with the overall steam generator-geometry. The ATH0S 11 code cannot model the rectangular tube lane region without tubes accurately. In order to assess fluid velocities in the "tube lane" region more escurately than avaltable from the'AtHOS 11 analysis, the FLOW 3 code was utilized to modet a 22 sector around the tube lane. 3.1.3 FLOW 3 code l FLOW 3 is a computer program for steady state thernet and hydraulle anaty- { sls of two phase flow in three dimensions. The program employs a network representation of the flow field and predicts local fluid conditions based on a bourdary value solution method. The code output includes nodal pressures, enthalples, qualities, and mass velocitics. The individual link pressure drops and flows are also a part of the output. In the current version, the flow fletd may be simulated with up to 350 nodes and 700 links. The conservation equations of mass, energy and momentum form the basis for the FLOW 3 network calculations. Enerpy and mass flow are conserved at each node and momentum is conserved y 'in each link. The conservation equations are supplemented by correlations for vapor fraction, friction, two phase multiplier, and heat transfer. The code was developed and vertfled by thi Bettis Atomic Power Laboratory. Verification was accomplished by checking results with hand calculations, evaluating the convergence criteria, comparing the predictions with other computer programs, and running a variety of problems (Reference 8). The FLOW 3 code was used by Combustion Engineering to simulate the FRIGG heated rod bundle experiment. The FLOW 3 calculated results were in good agreenent with the measurements and ATNoS 11 predictions (Reference 2). 3-7
b t 3.2 Palo Verde System 80 Steam Genet s tor Analysis
- ATMOS and FLOW 3 analysis of the steam generator was performed for the operating conditions at 100% power shown in Tabte 3.1 1. The analysis utilftes both the-full steam generator ATH0S model and the detailed "corner" region FLOW 3 model.
3.2.1 atmos Analysis of system 80 Steam Generator The ATMOS modet of the Systes 80 steam generator includes speelfications' for the geometric details, transport correlations, and operating conditions. Details of the geometric speelfications are input to the ATM05 geometry tre Processor (ATHOSGPP) program. The necessary geometric data for use in the ATMOS code are calculated by ATMOSGPP and written on a tape. The finite difference grid selected for the model was 12x10x2'6 in the circumferentist (C), radial (R) and axial (Z) direations, respec-tively. The arrangement of the finite difference grid in the R.Z and R*0 planes is shown in Figure 3.2 1. The steam generator operating conditions utillted for the analysis are' included in Table 3.1 1. 4 The ATHOSGPP output includes the finite difference grid, steam generator shelt and shroud details, tube bundle details, and all the geometri: parameters utilized by the ATMOS !! code. The ATMOS 11 output includes a summary of geometry (sta, physical properties of the primary and secondary fluids and tube metal, friction correlations, and numerical parameters used in the computational procedures to obtain a converged solution. s Results from the ATMOS Il code are then used to obtain the boundary conditions for the FLOW 3 mcdel of the lccalized region. The boundary conditions include the hot and cold side downcomer flow splits, circumferentist flow rates for the hot and cold side of tuce bundle at 22 from the tube lane, and axial flow at the upper boundary of the local model. 3-8
s Mr 3.2.2 FLOW 3 Analysis of Palo Verde steam Generator i The FLOW 3 model of a 22 sector around the tube lane is depleted in Figure 3.2 2. The model included the economizer local corner region from the tubesheet through 159" above the tubesheet at the 4th aggcrate or 5th tube support including the flow distribution oeffte. The model consisted of 169 nodes and 338 links. Twenty *four nodes were , used to"model each Level;"six nodes were used to represent the "tube lane
~
region" and six nodes to represent the 22 sector of the "tube bundte" region for the hot and cold sides eac5. The hot and cold side downcomer nodes are appropriately included in the model. The central cavity, the tube lane and the annutus region between the tube bundle and the shroud are modeled as separate nodes to study the flow characteristics in the regions without the tube resistance to the secondary side flow. All the
- p. geometric details Ittustrated in Figure 3.1 2, such as the economizer divider plate, the downcomer partition, the cold side downcomer opening, etc., are represented in the FLOW 3 model. These local geometric verlations have a significant effect on local flow; however, they do not have a major influence on the overall steam generator performance, hence, this capability to include the "detalls" is not available in the ATH0S type codes.
The secondary side boundary flow rates, enthalples, pressures and heat fluxes used by the FLOW 3 model were obtained from the ATHOS analysis as discussed in the previous section. The pertinent results for this case are presented in Figures 3.2 3 through 3.2 6. Figures 3.2 3 and 3.2 4 Ittustrate the circumferentlet and radiat components of the fluid velocities at the cold side downcomer 3 fluid entrance elevation. At Level D, 96" through 103" above the tubesheet, which includes the economizer divider plate, fluid velocities are relatively Low, as shown in Figure 3.2 3. At Level E, 103" through 112" above the tubesheet, the local fluid velocities in the "corner" 3-9
h a L o, a region are much greater than anticipated, Figure 3.2 4. Level E incorporates the cold side downcomer opening without the economiter divider plate, thus providing an open tube lane for flow to cross from the cold side to hot side tube bundle, and an open channel for flow to penetrate radlally towards the central cavity. The highest calculated velocity.of 46.2 ft/sec occurs in the corner region formed by the annular gap between the cold side tube bundle and the shroud and the open tube Lane. There is practically no frictional resistance to flow'in this ragion. 'The corner region may also cause high turbulence and eddles which are not calculated by either the FLOV3 or ATHOS !! codes. The.
=-
fluid velocities decrease from 46.2 to 30.3 to 27.1 ft/sec as the fluid in the annutus algrates from the cold side tube bundle to tube lane to the hot side tube bundle. The circumferentist and radial e velocity components are significantly tower in the tube bundle because of
^
the high cross flow frictional resisthnce and boiling which makes the axial (parallet to vertical tubes) direction to be the dominant flow direction in the tube bundle. The velocities in the tube lane also decrease rapidly from 26.1 ft/sec at the outer edge of the cold side tube bundle (Radius = 81.4") to 13.3 ft/sec at 7.8" away, i.e., at 76.8" radius. The rapid decline in the velocity is due to some of the fluid entering the tube bundte and traveling upwards /downwards in the tube lane. In summary, the high fluid velocities experienced by the tubes in the "corner" region diminish rapidly. The phenomena was also predicted by the ATHOS !! code, although not in such detall. The fluid velocities in the downcomer and in the annulus region are presented in Figures 3.2 5 and 3.2 6, respectively. These velocities provide comprehensive three dimensionst flow characteristics in the System 80 steam generator. 3.3 _F L O W DISTRIBUTION ANALYSIS RESULTS The pertinent results from the flow distribution analyses performed for this study are summarized in Table 3.3 1. For the as designed system 80 3-10 _n-
- 9..
l-steam generator, the fluid velocities-In the corner region and'In the tube tone-are high. This is due to's combination of local? geometry, lttustrated.In Figure'3.1'2, and the downcomer flow split, f.e., 66% of l the-flow enters the tube bundle on the cold side. For the reference case analysis model in Tebte 3.3.1, the steen generator design is essentially l 1-Identical to System 80 with the exception of not having a hot side flow l distribution plate. Thus, the results indicate that the presence of the l , hot side flow distribution plate is the major reason for the higher cold l- i side downcomer flow rate and not the economizer. i I i o 3-11
4 y
*; e -~
TABLE 3.3-1 CE SYSTEM.80 SI'EAM GENERATOR FLOW DISTRIBUTION ANALYSIS
SUMMARY
STEAM GENERATOR FEATURES Sealed Raised Hot Side Downconer Economizer flow Distribution Steam Generator Partition Divider Plate Baffle
- 1. System 80 Steam Generator No Yes u
[, 3. Reference case No no r0 no PERTINENT SECONDARY FLOW CHARACTERISTICS Not Side Cold Side Maximum Velocity Circulation Downconer Flow Downconer Flow Ft/Sec Steam Generator Ratio Lb/Sec % Lb/Sec X Corner Tube Lane Steam Generator 3.1 1782 34 3521 66 46.2 26.1
- 3. Reference case 3.3 2766 47 3187 53 *N/A *N/A 3
l
*Results not calculated whea Reference case was run.
I
- - - - . - - - - - - . - - - - - - _ . . . - . - - - _ - . _ - _ - - - - - . - - - - - - - - _ - m - ' _ * - - w - , -n-
STEAM GENERATOR COMPONENTS - W r STE AM O UTLET N0ZZLES
-STE AM ORYERS u c cs cs c3 c c cs c;;;rJ, ad)
CENTRIFUG AL SEPAR ATORS - j .
-5 TEAM ORYER OR AINS 2
Cu. c I MOISTURE SEPAR ATOR # SUPPORT PLATE g g .% -00WNCOMER FEE 0 WATER
.d
. _. I ..
.I
..-%. I EGGCRATES .___
I L l 1 l j i ! i TURE BUNDLE SHROUD
' i ..
i !
- -=
I .N NE INDOW FLOW DISTRIBUTION BAFFLE v : 9/- _ TURESHEET g , g "9g
, [ ,
.~
\ -
1
- STAY CYLINDER
[ \ FIGL'RE 3.1-1. SCHEMATIC 0F A COMBUSTION ENGINEERING SYSTEM 80 STEAM GENERATOR 3-13 - . . ~ . - . . _ . _ - . - . _ . - . _ . - . . - _ _ , . . . , - . . . . _ - _ , . -
1 j - _ _- -- L_-] _ ii 1 - _ -_ / I l l l l
)
1 I DQWNCOMMER, I PARTIS 10N I h-L V- - - k) l l I I C [ l
/ SUPPORT LEVEL 4 POSITION l h
I MAIN g n FLOW
, SHROUD HOT SIDE 9 IN.
COLD SIDE RECIRCULATING 16 IN* WATER ENTRANCE
/
; 7 IN.
,j e
.. , y v II !
ECONOMIZER DIVIDER l 1 PLATE i SUP. LEV. 3 POS. )' 1 ECONOMIZER i,
- FLOW SHROUD i ~,
f ~ ~__ __- __ _Y s CORNER
, s TUBE l
i i FIGURE 3.1-2 SYSTEM 80 STEAM GENERATOR COLD SIDE RECIRCULATING FLUID ENTRANCE REGION 1 3-lh
, NOT BIDE 567 COLD SIDE n-I 2345,70 100el20NTAL N00ALIZATION SEPARATOR DECE- r 26
\. / -
24
\ / 23 TOP OF U-BENO RE010N 22 7 3
\ l s r\ / it PARTIAL E80 CRATES 1 is r 17 15 i
- 15 14 2 13 l
FULL EBOCRATES gg 2 Il it r 3 g COLO O!DE EOSCRAT* I i s ! WOT BIDE EOSCRAT* r 5 I COLO BIDE ESCCEAT* r e
^
WCT SIDE EOSCRATE r 3 j FLOW O!BTRIBUTION SAFFLC I i II f VERTICAL N00All!ATION FIGURE 3.2-1. ATHOS II MODEL OF SYSTEM 80 STEAM GENERATOR 3-15
REGIONS OF MODEL
- m m * " ,
l
$ $! E! EE v6 b
a g rg re rg g 5 < a W u i M
~
wh COLD SIDE
. TUBE SECTOR BUNDLE ,, ,
I /- HOT S!DE s TUBE P BUNDLE SEC /f % SECTOR III s SECTOR N
!Y s
i
, E F TUBELANE TUBE DIVIDER / {
BUNDLE PLATE f SHROUD D l COLD SIDE TUBE SUPPORTS @g i TUBESHEET - B0T OF A , S FLOW DISTRIBUTION PLATE
- BOT OF B B f .
EGGCRATE 1 - B0T OF C EGGCRATE 2 - 4 INCHES A I BELOW BOT Or D DEL EGGCRATE 3 - 4 INCHES LEVELS ABOVE TOP OF E EGGCRATE 4 - TOP OF G l FIGURE 3.2-2 FLOW DISTRIBUTION ANALYSIS MDDEL ECONOMIZER LOCAL CORNER REGION l OF SYSTEM 80 STEAM GENERATOR 3-16
' ^
VESSEL SHELL NCOWERl l COLD S106 i / i 00WNCOW2R
- R0W 12 2.2 FT/SEC 3 '
giig.3FT/SECg ijjgr e: a+ ECONOWlZER D VIDER f
# W- l
/
I FIGURE 3.2-3 ECONOWIZER STEAW GENERATOR LOCAL FLOW DISTRIBUTION ANALYSIS COLD RECIRCULATION INLET REGION j CE SYSTEM 80
- LEVEL 0, 7 INCH DEPTH WEAN ELEY 99.5 INCHES ABOVE TS 3-17
~
VESSEL SHELL
^
t i i l COLD SIDE WNCOWERl l DOWNCOWER i i , i i i
---h,, R0W 12 2.2 FT/SEC 26.1
27.1 +-l 6.2 FT/SEC r-- LINE 1,189 l l . Og l 13.3 U
-- -- LINE 10,180 i .
TUBE (T) T LANE Q I I I I I I I I I I I I i l l I I i FIGURE 3.z-4 ECONOWIZER STEAW GENERATOR LOCAL FLOW DISTRIBUTION ANALYSIS COLD RECIRCULATION INLET REGION CE SYSTEM 80 LEVEL E,9 INCH OEPTH WEAN ELEY 107.S INCHES ABOVE TS l 3-18
SECTORS IV III 11 I DOWNCOWER PARTITION ASSEWBLY 8.1 10.4 10.6 l'.6 i ki i k i 1 l l 1 l I I HOT SIDE i COLD S10E 7-l , -. -- -- -- - l
' I s , l LEVEL F 1 1 l
iw3.5 FT/SEC c l l 1 1 l
._____l__I_h_____
LEVEL E lN10.6 ' 10.1 I i 4.6 l I l U l
,__U__.
1 l l LEVEL 0 I I ______i- i__.._ ECONOWIZER DEAD SPACE DIVIDER LEVEL C PLATE i i FIGURE 3.2-s ECONOWIZER STEAW GENERATOR LDCAL FLDW DISTRIBUTION ANALYSIS COLD RECIRCULATION INLET RE010N CE SYSTEW 80
- RADIAL VIEW FROW CENTER 00WNCOWER RE010N l
3-19
SECTORS l IV !!! !! ! l HOT SIDE 0.40.2 COLO SIDE l
- _ _L' y uj _ _ _a _ _ ,
n , , , 3.2 I I i 1.4 FT/SEC l i I l i I bll M i LEVEL F l l l 1 1 1 l l l 1 1 I l l I I l l 1 _L _I _ _ _ _ _ . < l l l l '30'.3 1 6.4 i I H l l LEVEL E 9 I I__ _. i i 7.6FT/SEC LEVEL D l l ______i_ _i_ _ _ _ _ _ ECONOMIZER l l LEVEL C DIVIDER PLATE j i
+
l t FIGURE 3.2-6 ECONOWlZER STEAW GENERATOR LOCAL FLOW DISTRIBUTION ANALYSIS COLD RECIRCULATION INLET RE010N CE SYSTEW 80 RADIAL VIEW FROW CENTER ANNULUS RE010N 3-20
4 SECTION 4 TUBE V!BRATION EVALUATION 4.1 Desfon Basis Evalua' ion The ASME Code, Section 111 Design Analysis Report for plants with the System 80 steam generators included a comprehensive evaluation of tube vibration. In addition to vibration analysis of tubes and tube supports using the latest findings and techniques for flow induced tube bundle vibration analysis available in the technical literature, every effort was made to bring CE vibration testing results to bear on uncertainties in previously used analytical correlations. Where possible, cross checking between analytical and test results was performed and reported upon in the dealen report. The following sections overview the vibration testing which was performed during the System 80 steam generator design program. 4.1.1 Generic Vibration Testino During the design phase, CE performed a steam generator tube vibration test program which included multispan dynamic response testing using mechanical shakers to obtain tube response characteristics, eggerate tube support characteristics and appropriate values for damping as a function of frequency of tube vibration. A typical test setup is shown in Figurg 4.1 1. The testina, which was performed in both air and water, produced result such as those illustrated in Figures 4.1 2 and 4.1 J. The dynamic response testing complimented several flow induced vibration design baals test programs, each of which addressed specific flow conditions in various regions of the tube bundle (Figure 4.1 4). A brief description of the flow tests which were performed are as follows: ! l 1 h-1 l l J
v e
'A. Air Water Flow Testing (Figure 4.1 5):
Model No. 1, weter flow, fluid entrance regions (Figure 4.1 6) - This modet addressed cross ftow entering the tube bundle in the hot and cold side recirculating fluid entrance regions. Model No. 2, Alt Water Flow, Upper sundle Cross Flow Region - this model addressed flow of verlable vold fraction across straight horltontal tube spans. Model No. 4, Alr Water Flow, Rows 1 11 U Bend Tubes - This model addressed flow of variable vold fraction across central bundle U bend tubes. The results from this test program and related work were published in i the transactions of the ASME (Reference 13) as well as being applied to the Palo Verde steam generator tube support design. Typical test results are shown in Figures 4.1 7 thrcugh 4.1 10. These figures ' demonstrate the verlety and level of complexity of tests which were performed during the System 80 steam generator design program.
- 8. 30 segment Economiter Model Flow Test -
This model addresses the feedwater entrance region just above the tubesheet on the coid side. C. 1 1/2 Mwt Model Test - This modet addressed two phase (steam and water) axial flow in the evaporator region of the tube bundle as well as providing performance data at System 80 operating conditions. As a result of performing a rigorous vibration testing progrom, it was felt that all bases had been evaluated with respect to the System 80 steam generator design and its susceptibility to vibration induced tube wear. The test data developed from the described vibration tests was apptled to the design and analysis of the System 80 steam generators as described in section 4.1.2. k-2
e 4.1.2 system 80 The tube and tube support vibration analysis performed as part of the-ASME Code (Reference 9) design report addressed the cold side recirculating fluid entrance region with regard to susceptibility to excessive tube vibration. The entrance flow velocity was taken ftom the CRIBE performance and stability code. However, the values used are very close to the current ATH0S !! predictions. An average entrance velocity of 2.45 f t / :t e c . became a tube gap velocity of 9.8 ft/sec. for the maximum flow rate. The vibration analysis methodology developed by connors (Reference 12) was used to determine critical velocity with regard to fluid elastic instability. The insteollity constant was taken from the flow testing discussed in 4.1.1 (Reference 13). Tube support characteristics, used to determine the effective gap velocity (see 4.5), and damping vetues were taken from the CE dynamic response testing (Reference 14). The virtus! m6cs coefficient for external fluid was taken from the work of Moretti 8 Lcwery (Reference 11). With this information a critical i velocity of 17.9 ft/sec. was calculated. White average, rather than local, velocity was used te compare with the critical velocity, the entering fluid was conservatively.;=sumed to be totally horizontal in i direction. A critical velocity ratio of 0.55 was Judged to have sufficient margin. Using the calculated critical velocity ratio and Figure 4.1 9 (Reference 13), an amplitude of tube vibration was predicted to be 1.6 mits for the tube spans adjacent to the cold side recirculating fluid entrance window. Such a small amplitude is normal and would not be predicted to l cause tube wear. The tube stress calculated for this level of vibration is less than one kal. The accuracy of the described tube vibration analysis was subsequently vertfled by ac^ual measurements in the downconer model test program ! described in section 4.2. k-3 !
4.2 Downeomer Modet Vibration Test For the axial flow economiter reglen, flow vibration tests had been previously performed on a 30 sector, fullascats model. Although no tube vibration of consequence was measured in these tests, occurrences of fretting and wear of tubes in the preheaters of some (non C E) operating units prompted C E to expand the scope of its investigations to more thoroughly assess the susceptibility of the System 80 design to similar damage mechanisms. Accordingly, thit investigation was under-taken to experimentally determine the vibrational response of steam generator tubes when subjected to water flow issuing from inlet open- ' ings, and to acquire velocity distributions for comparison with previ-ously (experimentally) established critical velocities for fluid elastic coupling. There are two locations in this region where water enters the tube bundle as indicated in Figure 2 1. At the tubesheet, feedwater enters from the feedwnter distributor below the flow distribution plate and flows upward through the bundle. At the top of the economizer, auxtllary feedwater mixed with the cold leg recirculated water enters from the downcomer through an opening in the shroud. The region of the steam generator which was modeled includes both the feedwater and cold leg downcomer inlets to the tube bundle. Figure 4.2 1 shows the general arrangement of the full scale geometric modet. Tubes, tube supports, tube support spacing and shelt side intet openings are the same as for the System 80 steam generator. The model is rectan-gular in shape and constructed from structural steel with plexiglas sides to permit visual studies. It consists of 144 tubes, each 175 inches long which are arranged I r. a 7 tine pattern as shown in Figure 4.2 2. The tube array is representative of a bundle with a depth of 20 rows from the periphery. Selected tubes near the flow inlets are instrumented as shown in Figures 4.2 2 with semi conductor strain gages and bi directional acetterometers. Penetrations through the plexiglas side are provided at 8 elevations downstream of the two inlet openings for Insertion of a 4-L
d' i pitot proba which can be moved horizontally for measuring velocities at positions across a section. Intet flow may be admitted to both economiser and downcomer inlet
-regions. System control valves are manipulated to achieve predetermined axial and radial mass fluxes. The capability to = partition" the flow makes it possible to simulate near prototypic hydraulic conditions in the model.
Hydraulic testing was performed at room temperature with nominal flow rates equivalent to 100% power and for downcomer flows up to 200% nominal. Modeling stallitude was based on equality of dynanic pressure, i.e., (pv ) modet = (pv )-S,G. For the 100% case, the specified System 80 feedwater flow rate was used. The cold leg downcomer flow was determined from the ATHOS (Reference 1) analysis of the System 80 steam generator. The partition retto for cross flow at the downcomer inlet elevation (Span 4) was also determined from ATHOS results. For the increased downcomer flow cases (!50% and 200%), the nominal flow was increased linearly while holding the feedwater flow ar.d partition ratio constant. Velocity distributions of the shell side fluid downstream of the two inlet openings were established from measurements made at the eight vertical and four horizontal Intersecting locations shown in Figure 4.2 2. A two dimensional "wedge" pitot probe was used for measuring the direction and magnitude of flow veloc;ty at each grid point. In Figure 4.2 3, totat velocity distribution is presented in vector form with the horizontal component shown plotted vs opening height. I The maximum tube vibration amplitudes occurred in tube no. 4 (Figure 4.2 4). In addition, the following observations can be mede
- 11) The tube motion was orbital with the major axis in the transverse direction. It is believed that this preferential orientation is j due in part to the eggcrate grid orientation which allows h-5
,s o=
1 approximately twice the transverse displacement within the support that.ls attowed in the direction of flow. Eggcrate grid orienta-tion chosen for the model is shown in Figure 4.2 5. Grid orienta-tion in the steam generator varies, with respect to Intet flow,
- around the periphery of the tube b>;ndle.
(2) The largest observed vibration amplitudes occurred in the span above the cold side downconer fLuld entrance region (Span 5, Figure 4.2 4). This span is longer and more flexitte than Span 4, which is subjected to entrance ftow. The amplitude of vibration in Span 5 increased from 1 1/4 mits at 100% flow to 2 1/4 mits at 200% flow. (3) The level of vibration in the tube span subjected to cold side downcomer fluid (Span 4) was relatively constant at 0.4 mit up to ' approximately 150% flow. Between 150% and 200% flow there was an increase in vibration amplitude to 1.1 mils. White this amplitude is less than that for Span 5, it is more highly stressed due to the shorter length. (4) In this test, the vibration mechanism which was cbserved is similar to that reported in Reference 13. At low flow rates, there is random orbital tube movement due to turbulence caused by the flowing fLuld which must negotiate a sinuous path between the rows of tubes. Vibration emptitude increases gradually as flow rate is increased. As anticipated, no vortex shedding induced vibration was observed for two reasons: (1) the fLuld approaching the bundle was too turbulent, and (2) the triangular pitch tube array is so tightly packed that vortices cannot build up. (5) The velocity profite in Figure 4.2 4 was examined using the method-ology described in Reference 12. Due to the quantity of non radial flow at the entrance region, it was found that the effective uniform tube gap velocities were 60% or less (depending on flow rate) than one would calculate by assuming a uniform radiat h-6
4~ . p. velocity. The original design calculation assumed.an effective uniform velocity of 67%. Thus, the original calculation was conservative with regard to velocity distribution on the tube span. This conservatit.m somewhat offsets the 10% increase in fluid velocity from that used in the original calculation. Calculations of tube deflections using the measured velocity distributions to predict tube vibration, and observations of measured tube movea ments, Indicate that there is at least a 50% margin to Instability at 100% power, assuming a uniform circumferential flow distribu-tion. Vibration emptitudes of tubes are less'thaa 2 mits. For average entrance velocities, the downcomer flow modet served as a proof test for the CE vibration test data and flow leduced i vibration analysic methodology described in section 4.1. It also quantiflad the degree of conservatism due to the non horizontal velocity component of the flow entering the cold side recirculating fluid entrance window. The downconer flow model test program did not identify the vibra + tion Levet increase in the local corner region due to the high velocity of fluld bypassing the tube bundle and streaming down the tube lane. The model was not constructed with a geometry that coutd have discovered the bypass flow probtem. Figure 4.2 5 ittustrates the location of the test model gecmetry (upper, right) with the region of concern (lower, left). A paratLel flow model geometry was chosen since test data indicated it was the most i susceptible geometry to flou induced vibration (see Figure 4.1 8). l l l 1 4-7 i
+ -
J 4.3 ECt Inspections at Pato Verde 4.3.1 Detailed ECT fvaluatfon An eddy current inspection of Palo verde's steam generator tubes for Units 1 and 2 was conducted earty in 1987. Units 1 and 2'had been-{n operation for 359 and 175 days, respectively. Figures 4.3 1 and 4.3 2 summarize the findings of this inspection by < Identifying the location and wear depth Indications in mits for atL four steam g e r.o r a t o r s . The primary locations of wear were the tubes in the corner region and tube Lane adjacent to the economiter divider plate on the cold side of the steam generator. Based on detailed review of a sample from the Con Am ECT tapes by CF (4 tubes each s.G.), the following observations can be made: Most tubes, that have wear, have Indications at two or three support elevations. Deepest wear occurs at second support elevation (first eggerate), with typical scar 1.0 inch in length. Essentlatly no wear at flow distribution plate. Less deep wear at support elevations 3, 4 and 5, with typical scar 0.25 to 0.75 inches in length. Three of eight tubes examined had multiple wea.a scars at either elevation three or fsur. Wear at elevations three and four similar in scar geometry to abetwing" wear scars observed at sen onofre 2 and 3, Waterford and st. Lucie 2. 4-8 ,
s 14 :
. e:
Wear at elevation two nearly flat a suggesting a mode of vibration with zero or necr zero slope.
-An eddy current inspection of Unit i steam generator tubes was performed curing the Fall 1987 outage after en additional 135 days of operation.
In generet, there were no significant changes'in the condition of the-tubes since tho' February 1987 ECT inspection. 4.3.2-Tie tod Eftects 1 Isolated cases of tube wear occurred at locations adjacent to tie rod supports (See Figures 2*2 and 4.3*3). The clearance between a tube and t i e.. r od is only 0.0625 inches whereas the nominal clearance between tubes is 0.25 inches. It is believed that the smaller clearances result in higher local velocities leading to excessive vibration and account for the isolated tube wear. i k-9
U~ ,
}
l i 1 4.4 Elaenvalue Analvels a A vibration analysis was perf ormed ces the Palo Verde tubes and tube support configuration to nhtain information relative to evaluntion of
~
tube wear caused by fluid elastic instability. The general purpose computer program ANSYS (Reference 10) was used to perform a finite element analysis. A description of the procedures, model, and results are presented.
-4.4.1 Geometry A single tube with 0.75 Inch 0.D. x 0.042 inch wall is supported at ten locations (see Figure 4.4 1), by a flow distribution plate, eight eggerates and the tubesheet. Att eggerates and the flow distribution plate are assumed to provide simple support to the tube Laterally white the tubesheet restrains movement in all directions rotational as well as translational.
4.4.2 Effeettve Tube Mass for vibration purposes the tube effective mass density is a combination of tube materlat mass, water mass inside the tube, and fluid virtual mass I outside the tube. The tube elements effective mass is determined from the following expression.
= T + $
( T $ + M f f) A Tg i where: 3 W = Density of tube material -
.305 lbs/in T
i W = Density of fluid In tube -
.026 lbs/in '
W, = Density of displaced fluid - variable l l 1 l l 4-10
A = Area of tube material -
.0934 in.
A = inside area of tube .
.348 In.
A = Qutside area of tube - .442 in. f C, a Virtual mass coefficient A virtual - xm- coefficient of 1.52 is applied to the tubes based on the work of Moretti and Lowery, Reference 11. This accounts for fluid mass surrounding the tube which tends to lower its frequency of vibration. The fluid density outside of the tubes varies from 45.8 lbs/ft at the tubesheet to 12.0 lbs/ft at the upper most support location. 4.4.3 Modet The finite element model is composed of 59 structural pipe elements with mass characteristics as described above. The modulus of elasticity for each element is assumed to be constant and is equal to 29.4X 10 lbs/in . Staty nodes define the elecent locations aJ shown in Figure 4.4 2. The horizontal span of the tube was not included in the finite element model since its motion is, for the most part, decoupled from the region of Interest near the cold side recirculating fluid entrance window. 4.4.4 Results The Eigenvalue analysis determines mode shapes and frequencies for the l tube model described above. The first eight modes of vibration vary in l I frequency from 33.7 to 123 Mr. Figure 4.4 3 shows a compilation of the first 8 mode shapes for ease of comparison. The span between the 3rd and 4th support locations is of particular interest since the cold side 1 h-ll
-s I l e recirculating flow lapinges directly on the tubes In'a radial or cross
- flow manner. Based on Connor's theory (Reference 12) the mode most likely to cause damaging vibrations is one'in which the displacements are maximum in the region of cross flow. This analysis indicates that the shth mode with a frequency of 123 Ha is the most likely candidate. It Leo significant._to note that the mode shape at support number two,-
'frSt eggerate above the flow distribution plate, has a zero slope.
seidering .the nominal clearance between the tube and eggerate, an orbital vibratory motion of the tube at this frequency could lead to a flat wear scar as observed at this location from eddy current inspection data. The model'results presented are based on the assumption tnat all supports are effective (i.e., provide simple support), however, consideration was given to the possibility that the first eggerate was ineffective as a support initially at low vibration levels. An Eigenvalue analysis for this condition indicates that the most probable mode of vibration, which occurs at 131 Hz, would produce flat wear at support levels two and three and is therefore inconsistent with eddy current findings, h-12
4.5 criticat Velocity l When tubes in a heat exchanger are subjected to fluid cross flow, there f exists a threshold velocity where the onset of fluid elastic unstable vibrations occur. .This is defined as the critical velocity and is given by the equation
= .
~
V =f K d o o (Reference 12) Cr n 2 o where f 9 Natural Frequency of Nth Mode of vibration (u z) k = Threshold instability Constant (Dimensionless) d = Tube 0.D. (in.) 2 2 l M = Reference Kass of Tube Per Unit Length tb-sec /In o
= Logarithmic Decrement =
[ 2 4"p
= Damping Ratio of Tube in Fluid (Dimensiontess)
Q = Reference Fluid Density lb sec /In If the cross flow occurs over a partial span or only one span of a multi span tube, the effective velocity must be determined since the critical velocity is greater for partial than full span flow. Reference l 12 presents a method for determining this value which is denoted by the l term, V p. The general equation for this evaluation is given bg g n
~
3 = i 9. FF YYl . 2. Nod () b 4 3E;
- ftl For the problem under consideration, the fluid density variation ( and tube mass per unit length a can be assumed constant in determining V ,p.
Performing the above evaluation for the System 80 eighth mode of vibra-tion (Reference Figure 4.4 3) results in an effective velocity of 0.656 h-13 _ _ _ _ - - - -----------------------._----a
V where V is the critical velocity of cross flow for span flow between supports 3 and 4 i Using the equation for critical velocity discussed earlier, the values for. full span flow can be determined. l 1 f = 123.4 Hz. (Palo Verde) l n ' K = 5 (Reference 13) d = 0.75 Inches M = 0.0001484 lb sec /In o f
=
2W{=0.04398 (Reference 14) 2 4
= 0.00006134 lb sec /In S0 The Instability constant used in this analysis (K = 5) was developed from a CE test program using prototypical geometry and materials.
The critical velocity for full span flow is catculated to be 16.7 ft/sec for Systen 80. Using the effective velocity factor gives a partial span critical velocity of 25.5 ft/sec. (Velocities in the cold side corner region were greater than 25.5 ft/sec as shown in Figure 3.2 4. The early ATHOS modeling did not predict flow velocities in this localized region.) For radial flow into the bundle away from.the localized high flow tube lane region one can determine the gap velocities using the downcomer flow velocities calculated in Section 3 of this report. This velocity, denoted by V Is converted to gap velocity using the fotlowing reta-tionship. GAP D.C. PD ~ where P = tube pitch D = tube outside diameter i
- l, g h-14
O From section 3 of this report, the cold side recirculating fluid velocity entering the window has an average rate of 2.2 ft/sec. for the System 80 design. To obtain gap velocities, V must be multiplied by (P/P D) D.C. for the cases considered. The table below summarizes the results of the vfbration analyses. All vetoeftles are in units of fps. The parameters in Table 4.5 1 are defined below: V DC - Downcomer Velocity (partial span entrance velocity)
= 2.2 V
GAP - Gap Velocity Setween Tubes = 8.8 (partial span velocity averaged around circumference) EFF/ GAP - Effective Velocity Constant = .656 V cr/F.S. - Critical Velocity for Full Span Flow = 16.7 Ver/P.S. - Critical Velocity for Partial Span Flow = 25.5 V MAX /f.L. - Maximum Velocity in and around the Tube Lane = 46.2 Determined by Flow Distribution Analysis An observation can be stated from the results presented in Table 4.5 1 A coeparison of gap velocity with partial span critical velocity shows a substantial margin for both configurations. Note that this is away from the region of high localized flow down the tube lane. h-15
4 l l
]
vi t VA gggo O posmoN i [W^ GRao e O position s g c'^ '^ sRao i g O rosmoen s 3e- SPAN 7 3.- SPANS O rossTeoN 4 m i 1 d QGRes 'n 'n GRao 5 O - rossim 3
! O - rossisoN 2 ; 40- SPAN s 4e- SPAN s j
Ig O- roSITIoM l ' r/s/rn GRIOs V 9 In GRIO 4 V/D/A GR40 l 7 28- SPAN s 28- y SPAN 4 V8EW A-A O rosarsoN 3 Os REFERENCE l A O 4
- 'Is / rn GRIO 4 TI>//M 3
GHBO3
' 2 emm i
O - rosmoN i 32- SPAN 4 48 S- m3 PLATE j r// / r/J ggg 3 } ' r/> / II' GRIO 2 READ:NGSWERE TAKEN AT ALL SUPPORTS 33- SPAN 3 AND AT i AND 3 POSIT 80NS GETWEEN TWO "E SPAN 2 ADJACENT EGGCRATES AND PLATES. l " GR10 2 g - l! -PLATE 1
' ^ '
HESPECTIVELY l' 6 Is- l l Q SPAN I his" p SPAN 1 l { l _S8NusotDAL A CONT'M. ACCELEROMETER ' VIBRAT80N INPUT *
$4880 LATED l CONSTANT I e O REN MLENTER TEST No.s TUDE SHEET TEST No. 8 ACCELERATION 9 DISTR 88UTIoM PLATE: 0.030-l C E SYSTEM 80 C E SYSTEM 80 DIAAAETRAL GAP W8TH TusES
{ ttOT LEG SUPPOH T SPACING ECoNOGNZER SurFORT ARRANGEMENT l C-E TUBE SUNDLE CONFIGURATIONS EVAPORATOR ANO ECONOtel2ER SECTIONS WATER ENVIRONRAENT TESTING
4 6 1 Z ) 5 s/ T R E 7 H o( 4 N2 1 T N E1 S U1 0 T1
~ 3 E
M N O
- 2 R
I V N E R
> / 1 E
T , A n I l i 0 W 7 D N < l 1
]
A 's s 2 N l N h 8 2 Os E T B U T 8 E
/ N 3 E A
R L 2 1 L E 4 o D C N N C E T U 4 A R S B TH E U G E S E BIG
-N I
T N F S UE 5 O T O TH S P L E P l R E 6 P A H 3 S E D O ) Z 2 M 2
.T K oR A
E P NEH E ( / 1 O T K E A^ S 4 U1 T 1
')ll '
[, " 0 P g 1 f( \ 2 3 3 2 T 6 S 4 1 E D I D D D I D I E E l I R I R R T i R R A S G G G G G L E P B U T
?4
v i . . , , . iy . - g I I 2
=
-n I
a
- a. n . . _,
I . 4 - I 5 ...'
.I St , '.
-n
% ,e.
3 4
= _
-~ ~
E
.A 4. 1 . ~;
a !!! . i
. J.o ;-
7
-n v -
J
.- q. I, J m
_s- g e * .. .:.
- a g
v i .. .. . p 3 I
,2-g _ . . . _
I 4
.. 94 e 2
4 9 e f f f f
= . . =
. e m ~ ~ - - .
5 ouww em.mc w:wvs 1-18 4
\ /
Ah A mca ra Ir p -- E
;4 G "M e e
3 su is ir l \O O O O O O[; -
, i i 4 . , m ja.
TWO PHASE j , r*~ ' -~ i ~7 ' CROSS FLOW ~ '
, , . . =
N g 1 [."* -,
, i=:
I l
-p= =a j ===;
']
/ <
/
AXIAL FLOW N[$:% [
,$)i > '~
f'
- 11 1 -
sd s l"I ; p ' COLD SIDE 4 RECIRCULATillG j ' s
.,j i
EE
) EffTRAllCE FLOW w- ..,;{ s s ,
i __-: I' HOT SIDE I Rh 4 : FEEDWATER ti iE
^
, a
< l }
./ 5
'N,'
l
. ,) (f., . , .,
/ i i
SYSTEM 80 ST'EAM GEtlERATOR P0TEtiTIAL REGIO!IS FOR F.I.V. ADDRESSED BY TESTIfiG AllD AffALYSIS FIGURE 4.1-4 h-19
)
i
) EXPANSION 1 JOINT RETURN LINE FLOW ~
CONTROL VALVE WITH REMOVABLE = TUBE MANUAL BY-PASS SEC ON . 4 BUNDLE g w g ', AIR SUPPLY
@ y PRESSURE FILTER 2,000 GALLON AIR-WAT8ER AIR HEADER REGULATOR TANK SEPARATOR I I a
g LJ 3 I I l RTO TEMPERAlURE I SENSOR LOW FLOW (__; BY-PASS LINE (4" SCH. 10) r a 1 4 d) VENTURI FLOW GLOBE EXPANSION CENTRIFUGAL METERS JOINT PUMP [ VALVES MITERED ELBOW WITH VANES / - FLOW lvj l } 12" SCH. 10 PIPING - - SAND LINED CONCRETE PEDESTAL FIGURE 4.1~5
O O O -
. I 2.: '
- 5- ,
. t.
- .=
g~zy y-Q}'*M V. ~
,.c ,
i.f 'il 2 , -;
-. ,3 t 9 ,
.g ._
N: -
~~
i .e - - ,,
% I a.
.,_ l 4 '-~
L,^ '
,r.--__. ,
_. J ': k.'?_.:r;-]-* "g *
. 'I
, .- y .
- e 1 .
y h t ; - , l
. * . =
G* **
. t
'.i~ . y.. .
'\
~ Q . '. .
FRONT OF MODEL TWO-PHASE FLOW FIGURE 4.1-6
1 x 10 3 8 8 i 8 . . , 1 x 10 0 , , , . , , , , , , , , , , t.i rSo rLOT (ca m rLOr 1 x 10 2 _ RUN 46 - 1 x 10'l - Oo"= _ ,12 FT/SEC .84 FT/SEC - G = 0.722 G = E724 6v= 0.018 IN. RMS = 002 N. RMS 1 x 10 1 - 1 x 10-2 t V y N LOW DAMPING _ N HIGH DAMPING e3 E 1 x 10 0 -
$3 1 x 100 -
1 x 10^l 1 x 10 4 1 x 10-2 ' ' ' ' 0 20 40 60 80 1 x 10-5 e i i e i e i i di i , 100 120 140 0 20 40 60 80 100 120 140 i FREQUENCY, HZ FREQUENCY,HZ T i Pd 10 . . . . .. ... ... 3 , , , , , , , , , , , , , , (b) COHERENCE PLOT RUN 46 g 270 - g,,pHASEetOr - 0.8 - n INDICATES STRONG , W RUN 42 COUPLING BETWEEN tas DISCONTINUITIES TUSES E INDICATE COHERENCE w l U [ m 180 -
< 10%. NO COUD LlNG ,
S 0.G - h - O BENEEN TUBES
' k -
= li _,
I $ 0 g% o 0.4 - [ - u I 0.2 -
- 0 0.0 .go i e i e i i 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 FREQUENCY,HZ FREQ'.lENCY, HZ Figure 4.1-7 O F F - L I N E COM? UTER PROCESSED DATA
I i 1 I I a i i a t tt U 80 - Q- 2.0 - IMPACTING A ARRAY Nor.1,2,3,4 O _ 7, p (12s mils) _ i.8 -
" O ARRAY Nos.5,6,7 3
x FULL BUNDLE ou. 1.6 - OQ ARRAY No.8 ARRAY Nos. 9,10 - E 60 - O 60* - p 3 ARR AY No. 5 e O 50 =
- o_ i4 _ (TWO PHASE) -
' Q go 2 6%0 I >
- - m F
[o 50 O 1.2 - O E' - 6 w ROW m VORTEX ( g uj u SHEDDING RESONI.NCE O"$ b 40 - - o 1.0 u- ARRAY No.9 - o Q-,. s O a Y Bg e b 30 - - U~ ,'s O $ 9 $ I W
- s u. A O Go y
- 03 -
Y g O o On 5 - e 20 - -
# u.
0.4 d 6 - 10 - - 0 0.2 - O -
' I ' ' 0 ' ' '
0 O 10 20 30 40 50 60 0 .2 .4 .6 .'8 1.0 1.2 NORMALIZED GAP VELOCITY, VG/Vc FLUID PRESSURE (GAP) p VG /2g (LB/FT2) (a) MODEL I - WATER FLOW Figure 4.1-9 UPPERBOUND EFFECTIVE FORCE Figure 4.1-8 TUBE DEFLECTION vs FLUID PRESSURE COEFFICIENT
l I 100 g i g g i g g O Vc MoSo 50 =K 2
,a ,
O g UNSTABLE / f o
> /
s'
/
$ 20 -
p
/ -
; /
2 K = 9.9 #
< /p t-E x
k
- n. /
/
i m 10 i K = 5.0 # / - i N P b s
~ O / O CONNORS' O b / O ARRAY No.1
# & ARRAY Nos. 2,3,4
$ @ #p d 5 - O / A ARRAY No.11 (NEAR O) _
! o b b p'p O ARRAY Nos. 5,6, a = 0 ARRAY No. 5, a = 0.724
, $ g p' @ ARRAY No. 5, a = 0.840 i
O*p 'p# @ ARRAY No. 5 a = 0.870
@ ARRAY No.7 (NEAR O) 2 - /, STABLE 8 ARRAY No.13 -
/ O ARRAY No. 8 f
# @ ARRAY No.9 (NEAR O )
O ARRAY No.10 0 ARRAY No.12 I I I I I I I 1 l .1 .2 .5 1.0 2 5 10 20 50 DAMPING PARAMETER, MoSo/a d2 ST ABI LITY DI AG R AM
1
...e - -r . ,- . , , , . - . . , ,
~ -
<.v.=. % =.
1
,y f
, y==1 ,
7
=
, &,) L, m .
[ g I. l 4 _ _
, 'r, o , + )
!. t g F l T'1 .
t t l!f I [ _
't1 ,'-) '
/ IC
;, }f
- lu t g .
x s , l l
,a ;.- J Ai.
!. = g% s ,. "' ' D I T h
Iit f-J
- 1 w; A- -
L 5;I ) c#
~~-- , # t!TJ i; ; ) g
-/
I \I N g o
==' = '
\
g b< l i
"* N l;s. ~
... ,m8
[ni p i li , J[I g," 8 B
- f,pI..i l *. A
- p . . . .
t> ' ' * <t,
,, ,p .
;_ ._.r_-= .! y.
jff f l I t--- 7 j f i.i. 4 1.i- 1I gi J ,l _ l E ( ' l h- l a. \ - lll Ill' l- ! q-
~~ t l ::
c :
! 5 L
E' n; l l
-. sj YC g ,
5_ "~gD' F~- b '-
~
f@Le l-r- ,. i (,M di] [n" ."% r : 6, f[py14:yf ,2 l i- ~ h I [!i I$i k ,', I/k
~
ih.l f I i j +i l h"'[lj 33; 9'>' ' " AJ IlI
% i -
t ai lasji I i-i b>* l .l
,l .
A i at . II l ( sg _ . - . ! .. lj.
$i-- j Ig -l_ l o 5. Il . t ,
l ll_ FIGURE 4.2-1 l 1 i i , i i , h-25
l 00WNCOMER FLOW _ _ TEST H0 DEL l PLAN VIEW
.a .an
~
6.27' .
' de i I O.I O I l a> l 4 s\ !
........- % j 750' Kg O I
.. L
_ s-Q
.25 ! 8 O x.43 ,
O
= 3.Is- l O .,_4.89" '
O - O' ae e o O , O 3 n O O 2 k O ' ' O 4 -- g O O O O _.....f..._....... N 6 INLET N FLOW' VELOCITY GRID
- l
%E] Q Q$ > $ - ACCELEROMETER
@ . STRAIN GAGE FIGURE 4.2-2 b-26
VELOCITY DISTRIBUTION AT COLD LEG DOWNCOMER INLET 100% FLOW
/ /
E66 CRATE AVERAGE VELOCITY = 8.1 FNEC.
- v E -
\
,- N s
\
T
/ I is~ l / !
t y l
/ \
Y
/ !
?
i , 5 3 -2.-l 00 1 2 34 5 6 HOR'lZONTAL VELOCITY - FgEC. EG GCRATE l FIGURE 4.2-3 do l i34 5 TOTAL VELOCITY-FT/SEC. h-27
9 U-TUBE COLD LEG v
) G / N0h N
g s c SPAN 5 100% k0 FLOW g /
/ O a 4
COLD SIDE s b 100% AND 150% R C
) ) m
@ SPAN 4 FLOW ENTRANCE /
g % t s
? @ t 200%
FLOW ACCELEROMETER / POSITIONS "EGGCRATE" S RTS TYPICAL l @ FLOW DISTRIBlTTION 3 PLATE 4,
@ i N
\
@ k FEEDWATER \
ENTRANCE / / @
@ i ,
i i TUBESHEET 2 o 1 2 MAXIMUMDISPLACEMENT(MILS) FIGURE 4.2-4 TUBE VIBRATION AMPLITUDE PROFILE h-28 i
STEAM GENERATOR TUBE BUNDLE ENTERING FLUID CROSSFLOW ORIENTATION l PITCH l TRIANGLE 7 PERPENDICULAR l FLOW (0UT-OF-LINE) d PARALLEL i FLOW (IN-LINE) l DOWNCOMER MODEL REGION l' R0W _- T ' 150 _-
^
i
-9 13
- t 120 g 110
~-
I i 100 ._.s 90 l 1 80
. . ~
g 70 60 50 40 30 20
#E!!!bT!
--x- -
0-l
. . . . . _. . . . . . . . . -- -;-mm-;-;-;-:- . . . . .-
'LINE 10 20 30 40 50 SO 70 80 90 100
' I TUBE LANE < TUBE LANE :
BYPASS REGION SYSTEM 80 ECONOMIZER EGGCRATE MAP FIGURE 4.2-5 1-29 4
6 l A3SS31 SH311 1 l l l dV10 A3803 I y 2 i 513 H
"" L 10M3B 3010 5103 5HB000 N r R 80M t eet s9 z S*D'I'0.
l " l l s 6 10 li 13 8 e e e,I l I 11N3
;^el- <
8'8888 - e
@88888 e 888 8 ,
*' 888 88 s i 888 8 e i 98888 z
! 88888 e
- i. 8888 e ni 8888 te 888 <<
888 ie 88 te 88 1, 8 <, 8 1, 310083 #'C-1 9-EO
l l 1 A3SS31 SH311 nT!!!! l l dY 0 A 803 I ? 2 DNB 0 u 10M3B 3010 S!03 SHd000 3 8
~ !..[
~
83D10N 80M ( ac,s9 z s'0'I'0-l l l l l e l6 lol M MI lIN3
,86
/N
@@@@DD -
~
@88 8 tee
*' @@@ 88 -
i @88 @ -
! 088@8 -
! 88888 -
!- @888 let ni 8888 -
@88 me i 888 me tu
- 88 me l
8 tzS VlSO S'D*IV' 80M I' lIN3 tzO 410003 t*C-2 1 tt ~ ET l
TUBE SUPPORT STAY ROD ASSEMBLY STAY ROD i SPACER SLEEVE l I I l l STAY RCD "EGGCRATEn l l NUT l TUBE H I I SUPPORT I i i I
? < DIA, = 1.125 IN.
' ~
SPACER SLEEVE STAY ROD l x FIGURE FIGURE 4.3-3 h-32
4 t C 3 I 0018i l PARTIAL
); EGGCRATE r1 .
h SUPPORTS TIE RODS We A;' // t
,)
7
, . FULL CGGCRATE
# f SUPPORTS
. ,,, (
. a
" ~ " --
-. / I' CONOMIZER
# DIVIDER PLATE l-
-- k PARTIAL 7'.' g F -
j - EGGCRATE SUPPORT "j' 4 _ . , L _
; j
( / ' l4 ( ;
; j . '. . .
l ~~ 1 's
,/: i ; ;
'; i 3,
s .1- i- . s . g 4
.,,.I 7+9 sia M 2,.
3 .' i. >
,, s n i
'i' i s T..
(VERTICAL GRIDS AND BATWING SUPPORTS NOT SHOWN) FIGURE 4.4-1 SYSTEM 80
+- TUBE SUPPORT LOCATIONS h-33
9 360 ,. ELEVATI0tl 0F SUPPORT N00E SUPPORT SUPPORTS AB0VE TYPE NUfEER R NUMBER TUBESHEET (IN.) E.C. D 660 9th 332.00 320 " E.C. 9454 8th 287.00 280 E.C. 94 48 7th 2A2.00 240 E.C. ->4 42 6th 201.00 200 p 160 - E.C. P4 36 5th 159.00 120h E.C. pd 30 oth 116.09 E.C. De 21 3rd 92.00 80 - E.C. >e 15 2nd A8.75 40 F.D.P. 9 49 1st 16.00 Tubasheet Secondary Side 0- , , , i , i i < / s i s i s ii isi<i Scale (in.) I t AflSYS FINITE ELEf'E!'T liODEL OF TYPICAL TUBE VERTICAL SPAN WITH SUPPORT LOCAT10!45 FIGURE 4.4-2 41 1-34
. l l
l 1
=_. __. SUPPORT l 1-
~-
3__ 5. 55 Z :r----- sc-J:. =.w .- ==I-l== -- r A: ~ ~ - ~ " =-n ' fi~~~ ~ .~ - ~ . LOCATIONS
=== 9
. _-L=;
V,.5 7 5--- - CE C_Ef fEf- _ EEp.i'pE E h T/ ~=4 _33
% =.=== a= - --
=pl ...
Y__ y lg _ : _-- \~_
- l l .. -
--:Ja$. ..
_: g= - _ _%a_- _ _.-_. g .
=;
__ _ _ . , _ . _ _ jf-E=4 =.- _-*--- ==-1 ;- g._ :
+-(E EISQ == f:=
. - - - , _ . - - - - _ - . - - . - - ' - . _x .A. ,
g
._-t..
a_ / ._ _2 _;_-- - _ - _ L
-..m .__4 . ._
y _ g e . _ . _ - . . 1 _ _ _ __ k . _ _ _ _ l . d= =' ==-
.=E Z-tr _ _ _ .. _ _ _ . _L. _ _
+= = ---
; .;. . i ___ _ ;._=._ _ .
- - _ - - 1_p-
__.p\=_- 1. _. ,
. +. ..
_...___k _- ._ _ n ._.
~ - -
,..---~_ . . _ ,
f =- - ._ - =
- b -
..__ _ (_ f$ _
5_T' .___,._...j .2 ._. .__ f
'. .=~ Y.... - - - 3
__ h=_= :=j. _ . - _._ 7 -_~=-:_=_-, EE ._ __ q ag. - _ ~. g[_. 3 M_
- T ;
d_ _ _ _ _M. * ' " -
- . .. _ _ . _ . _ _ k b
- - ~
=1 '
M .J . - f. 7 -===,b -- -._ l
.. - =
4-' v- x - - - __ j.____ ___. f. _ ___;___._.y n _= .
=r- /__ = . . ._. .7 - -- g __
=-
6 y --- - __; .___ 4 __..y 54 .__y __ _=- - - - - . __ g
~*~ '
E@= ^---TT . l-L a q.
'== -
%-- .% V -
_._ : _ -- ; <(.-= =. ; )g__ u>_____ L- _% _ 6 =_=__=a_ s . ... _ _s _ . p. _ . = - .z.
= .3_ __x ,-- _.- = = ,= - J=
.=__
. = h. =%_: .- == -
t
~t . s
- ,I.
i-- M =- =- - - -} --=4_=: ==.,=.-H = . , =/. ' _s .
- = \ === ==> = .a={--
2: ---=EEEiER.:t
- Q: = - I _ ._ ;2 2 -
. . f_=:t:-~_
-- is= ,=
=
EEEi ==.~. _ _ x= = -
- . :_=_ . ..4___
=-y:3 . __., _.._.a
_ w r-J _. g _: ..
,/ ._ _ _ _ _ _ ._ . _ _ . .
- -i > - -
y - .- u~ _
'- -' 4 x :--/
~
'x ._-_ .---
2 f- -" RECIRC
. _ .==
==mi _: - __ '}
- .._ . _ _ . = r - -J -- - _ := )=. .- r-- - } =g intg (
.=*
- ' FLOW g !
- =*===3 = _ _ --==- /.-E- z .: 5.c= p.~ f.. J : __ _ f ===.4 sH =
N 3 WINDOU 53 ?=2M.\'E;5E ~s W rE W - i "i_" ff"i"4 ""-=.'==.F-=. h_ F=E /: y.=.- = . = = = _-=3=
= :
EEqu.. i- -- -
. = __z _E. ___ .
~ -- -- - - - . - .
3-_ _
=gg.
?== C=- == =_ . =; f='_:: =__-==m
= --__ -
. ._ . s - . . -: ==:-iE~f- 5 c= " = -
i-GnE E =-
-- ~ ~
/ -F~-K_ ___ ==i==E 5ETEiEE=f =s- + 9 Z-i [ IEA _
l T . : . :: := .
. _ __ V+ =< _ .a =- --
+-- ;-r _ = - _ .== ' ? - e_= , +.-. 2
=j _7 -
~
-=;=i_] ~-=[ _.
E=.f _ ===-E= &! + -mf K# 25 : ^-~; l__ ._ . A,
- = = =
== . m A =.=
i
.. _ = .-~~ _.. ,
_ %= i . . _ . 7
=- - - - --
? .__ : --. . _ - _ . _ _ . . _ _ . ___ _ ._. . . . _ + _ _ _ _ .-
5 ~ . . _ _.-= :- m
~~~~
~'~. 5- ~T '= T- '( 5~ h ~
5 _gjp __. _ . w y g e - . .___ g ;;.__ y 3 y =.= q y u. q _ p. . gg _ 1 i I l 33.7 37.7 43.9 49.6 63.9 83.4 110.2 123.4 l FREQUENCY (HZ) SYSTEM 80 MODAL i ANALYSIS RESULTS (NORMALIZED MODE SHAPES) l FIGURE 4.4-3 h-35 1 1
e-SECTION 5 CONCLUSIONS AND RECOMMENDATIONS CONCLU$f0NS The ATH0S !! and FLOW 3 analysis of the System 80 steam generators have identified the potential for local high velocity flow streaming around the cold side corner of the tube bundle and along the tube lane at the cold side recirculating fluid entrance window. ECT tube testing and evaluation of Palo Verde 1 and 2 steam generators have Identified wear indications at some of the tubes in the cold side corner region, predominantly at tube support levels 2, 3 and 4 The tube indication location coincides with the region of high velocity flow. e Recent tube vibration analysis predicts the potential for fluid elastic instability at votocity levels predicted by flow distributtoa analysis in the cold side corner regions, assuming idealized tube support conditions. Since tube to tube support clearances are typteelly variable near the edge and corner of the "eggerate" tube support grids, tighter than-nominal clearances could explain why not all tubes have worn due to excessive vibrations. The high fluid velocities experienced at the "corner" region decline rapidly and are significantly lower, only inches away. This Indicates the tube vibration and wear to be very localized and limited to only a few tubes in the cold leg corner region and these phenomena are not expected to propagate further into the tube bundle. I l l Results from a second steam generator ECT tube inspection at Pato Verde 1 l-
- indicate little change in previously unplugged corner tubes after eight t
months of additional operation. No prevlously plugged tubes were inspected, l 5-1
p: e-e- but all plugged. tubes were previously staked to prevent severance, regard-
'Less of wear.
RECOMMENDATIONS Palo Verde Units 1 and 2 - continue to operate with no further modifications. Palo Verde UnitL3 continue to operate with the preventative tube plugging pattern'shown in Figure 5-1. All Palo Verde units should continue inspecting unplugged tubes.in and bounding the cold side corner region as part of the periodically required steam generator tube Inservice inspection. Based on satisfactory results from subsequent inservice inspections, APS may choose to modify their inspection Intervals for this corner region as appropriate. Any corner' region tubes which are plugged, for either required or preventa-tive reasons, should be staked and plugged to preclude the possibility of gross failure in a non monitored tube. i 5-2
y / / / / / / / /(/ / / ///////// ' L(sO ' vee
\
COI l PA 0 YERDE 3 p/'// ....g; h -R o sis se,.
~~,jjjiik-n J u J -~
' i ROW 1 2 3 4 e f' 7 8 l l l 91011 l 12 LlHE
@@@@ + 1. te9 00@@@@8 "e
==" @@@ G 4 1" 98:ana _i es e . 1ee
@@@ @ e.1*
@ ! @@8@@ ' "'
?!!?c!%. @@88@ e ""
-e g eeeg .. le1 l
g _i egee 1 . le, m R. d=1,125 IN. 088 " 1"
- 12. 178 g ge m. uz auviv - @8 " " '
eU,gge g 1,. 17, 1 _9 ,e. 1,. FIGURE 5-1 5-3
ii SECTION 6 REFERENCES
- 1. Singhal, A.K., et.al., "ATHOS -
A Computer Program for Thermal Hydrau-tic Analysis of Steam Generators. Volume 1: Mathemat # cal and Physi- [- cal Models and Methods of Solution, volume 2: Programmer's Manual, Volume 3: User 8s Manual", EPRI Report, EPRI NP 2698 CCM, October, 1982.
- 2. Miestand, J.W. and Thakkar, J.G., "ATH0S and FLOW 3 Simulation of the FRIGG Heated Rod Bundle Experiment, EPRI Report, EPR!aNP 3514, May, 1984.
- 3. Hopkins, G.W., et.al., "Thermal and Hydraulle Verification: ATH0$2 and Model Boller No. 2 Data, volumes 1, 2, 3", EPRI Report, EPRI-NP*2887, February, 1983.
- 4. Katra, S.P., et.al., "Thermal Hydraulic Performance of a U Tube Steam Generator Under Contro'. led Feed Flow Oscittations", A Paper Presented at the 1983 National Heat Trans4er Conference at Seattle, Washington, July 1983.
- 5. Mastello, P.J., "Thermat Hydaaulle Code Qualification ATHOS2 and Data from Bugey 4 and Tricastin 1", EPR! Report, EPRI NP 2872, Febru-ary, 1983.
- 6. Singhet, A.K., et.al., "ATHOS -
A Computer Program for Thermal *Hydrau-lic Analysis of Steam Generators, Volume 4: Applications, EPRI Report EPRI NP*2698, August, 1984.
- 7. Thakkar, J.G., et.al., "Thermal Hydraulic Analysis of the New York Power Authority Indian Point No. 3 Steam Generator", CE Technt:st Report, CENC 1633, April, 1984.
6-1
l 1
",.t.
8.' Beus, S.G. and Anderson, J.H., "FLCW3 - A Computer Program for Network Analysis of Two Phase Flow In Three Dimensions", Bettis Atomic Poyer , Laboratory Report WAPD TM 1177(L), October _1974.
- 9. ASME Bolle. and Pressure'Vesset Code, Section I!! for Nucterf Vessels, 1966 Edition.
- 10. ANSYS Finite Element Computer Code, Revislan 4.1C, 1983 by Swanson A t.a t y s i s Systems Inc.
- 11. "Hydrodynamic Inertia Coefficients for a Tube Surrounded by Rigid Tubes", Moretti and Lowery, June 23, 1975, ASME Publication 75 47,
- 12. "Fluidelastic Vibration of Heat Exchanger Tube Arrays", C o r.n o.- s , N.J.,
Jr., 1977, ASME Publication 77 DET 90.
- 13. -"Vibration in Nuclear Heat Exchangers Due to Liquid and Two Phase Flow", Nellker, W. J., and Vincent, R. Q., April 1981, Engineering for Power, Vol. 103, No. 2
- 14. System 80 Steam Generator tube Basic vibration Tests, TR ESE 073, TR ESE 092, by K. H. Hastinger, CE Nuclear Laborator, December 1975.
- 15. Pulick, d. A., and Margotts, 5.G., "CRIB 1 A Steam Generator Stability Analysis Program for the PHILCO 2000 computer", Bettis Atomic Power Laboratory Report, WPAD TM 530, December 1965.
ibJ n
'g APPENDIX A COMPUTER CODE VERIFICATION OF ATH0S II AND FLOW 3
.This Appendix.' includes information on projects to verify the ATHOS 11 and FLOW 3 codes.
The ATH0S code has been checked and vertfled by Combustion Engineering and others (References A1 through A*7). The check out studies included comparing
'the ATH0S geometry pre processor computed values of the steam generator geomet-ric parameters against hand catculations, checks on the mass, momentum, and energy balances, consistency and plausibility of steady state and transient i
4F solutions for a number of different cases. For code verification measured data from several small scale experiments, modet steam generators, and *ull scale steam generators were compared with ATH0S results. In one case, an analyticat solution-was available for camparison with ATH0S calculations. In general, the agreement between the ATH0S results and available experimental data was good. 4 In addition, consistent trends were found in att parametric studies. More details of ATHOS verification may be found in References A1 through A-7. (Excerpts from References A 1, A 2, and A 3 are included to acquaint the reader with ATHOS II veriffcation efforts. The FLOW 3 code was verified by the Bettis Atomic Power Laboratory. Verifica-tion was accomplished by checking results with hand calculations, evaluating the. convergence criteria, :omparing the predictions with other computer pro-grams, and running a variety of problems (Reference A 8). The FLOW 3 code was used by Combustion Engineering to simulate the FRIGG heated rod bundle experi-ment. The FLOW 3 calculated results were in good agreement with the measure-ments and ATHOS predictions (Reference A 2). I l.-1
r,. N-REFERENCES A1' Hiestand, J.W. and Thakkar, J.G., "ATHOS and FLOW 3 $1mulation of the FRIGG Heated Rod Bundle Experiment, EPRI Report, EPRI NP 3514, May 1984. A 2. 81nghel, A.K., et.al., "ATHOS - A Computer Program for Thermal Hydraulle Analysis of Steam Generators, Volume 4: Applications, EPRI Report EPRl*NP 2698, August, 1984. A 3. Mastello, P.J. and Chan, P.C., "Determination of Secondary Flow Fletd in Steam Generator 2 of Millstone Unit 2". Jaycor Final Report J 5391 87 001, Jaycor Engineering Services Group, San Diego, California, February 1987. A 4. Hopkins, G.W., et.al., "Thermat and Hydraulle Verification: ATHOS2 and Model Boller No. 2 Data, Volumes 1, 2, 3", EPRI Report, EPRL NP 2887, February, 1983. A $. Katra, S.P., et.al., "Thermal Hydraulic Performance of a U Tute Steam Generator Under controlled Feed Flow Oscillations", A P?per Presented at the 1983 National Heat Transfer Conference at Seattle, Washington, July 1983. A 6. Mastello, P.J., "Thermal hydraulic Code Qualification: ATHOS2 and Data from Bugey 4 and Tricostin 1", EPRI Report, EPRI NP 2872, Februsty, 1983. A 7. Thakkar, J.G., et.al., "Thermal Hydraulic Analysis of the New York Power [ Authority Indian Point ko. 3 Stesa Generator", CE Technical Report, CENC 1633, Aprit, 1984 Beus, and Anderson, A Computer Program for Network I A 8. S.G. J.M., "FLOW 3 - Analysis of Two Phase Flow in Three Dimensions", Bettis Atomic Power Laboratory Report WAPD TM 1177(L), October 1974. A-2
- p. ,,:, - - - = - -
;q
--s. .
- Enclosure (2)
- to LD-88-049 -
- 5 Pages ENCLOSURE' 2 CESSAR CHANGES TO ADDRESS STEAM. GENERATOR TUBE VIBRATION IN THE COLD SIDE CORNER REGIONS j:
L' l-i
- p. 1 of 5
a I Insert A After design and fabrication, it was determined during the initial operation of System 80 steam generators, that a potential exists for vibration induced tube wear to-occur in the economizer at the periphery of the cold side tube bundle adjacent to the divider plate. Owners are recommended to 1,stitute a program to plug and stake all susceptible tubes with indleated tube wear or preventively plug and stake tubes in these corner regions which have the potential to experience this wear. It is also recommended that this program consist of monitoring and periodic inspection during plant outages as part of the required in service inspection program and, where further unacceptable wear is found, ensure that the tube (s) is plugged and stake.d appropriately. i I
- p. 2 of 5 L
< +,
[ h. Four thousand ressure trarfsients of 85 psi across the prin.ary divider plate in either. directiof'dition). coolant pumps (normal con caused by starting and stopping reactor The steam generator was esigned to ensure that critical vibration frequencies are well out of the va e expected during normal operation and during abnormal conditions, fI he tubing and tubing supports are designed and fabricated with considerations given to both secondary side flow induced vibration and reactor coolant pump induced vibrations. In addition, the steam generator assemblies are designed to withstand the blowdown forces resulting from the soverance of a steam nozzle. T blies are also designed to withstand the severance,he of steam any one generator of theassem-feedwater nozzles. The two accidents are not considered simultaneously. The steam generator tubes are Ni-Cr-Fe alloy, 3/4-inch 00, with 0.042-inch nominal wall thickness. A steam generator tube rupture incident is a penetration of the barrier between the reactor coolant system and the main steam system. The integrity of this barrier is significant from the standpoint of radiological safety in that a leaking steam generator tube allows the transfer of reactor coolant into the main steam system. Radioactivity contained in the reactor coolant would mix with water in the shell side of the affected stean generator. This radioactivity would be transported by steam to the turbine and then to the condenser, or directly to the condenser via the Turbine Bypass System. Noncondensible radioactive gases in the condenser are removed by the main condenser's evacuation system and discharged to the plant ventilation system. Experience with nuclear steam generators indicates that the probability of complete severenco of a tube is remote. A double-ended rupture has never occurred in a steam generator of this design. The more probable modes of failure, which result in str. aller penetrations, are those involving the occurrence c.' pinholes or small cracks in the tubes, and of cracks in the seal welds between the tubes and tube sheet. Detection and control of ! steam generator tube leakage is described in Section 5.2. The concentration of radioactivity in the secondary side of the steam generators is dependent upon the concentration of radionuclides in the reactor coolant, the primary-to-secondary leak rate, and the rate of steam generator blowdown. The expected specific activities in the secondary side of the steam generators during periods of normal operation are given in Section 11.2. The recirculation water within the steam generators will contain volatile additives necessary for proper chemistry control. These and other chemistry considerations at tr.e main steam system are discussea in section 10.3.5. ( 5. 4. 2. 2 Descriotion The steam generator is illustrated in Figure 5.4.2-1. Moisture-separit)7 equipment in the shell side of the steam generators limits moisture coM r of the exit steam. Manways and handholes are provided for access to t"e 5.4-9 p. 3 od 5 t
~
i:. , e Insert B
- 5. As part of the in service inspection program, the owner should monitor and periodically inspect steam generator tubes for vibration induced wear at the periphery of the cold side tube bundle adjacent to the economizer divider plate. Criteria should be established such that tubes found to exhibit wear are preventively plugged and staked to preclude the potential for leakage to occur. The owner,- as a result of continuous satisfactory inspection results, may revise the inspection intervals for the steam generator corner regions.
- p. 4 of 5
~15. If the isolation valves upstream of the ADV's are electrically controlled and operated, the valve operator and control systems 10 shall be designed to the same IEEE standards as applied to the ADV's.
J. Inspection and Testing
- 1. All ASME B&PV Code, Section III, Class 1 and 2 valves shall be designed, fabricated and installed such that they are capable of being periodically tested in accordance with ASME Code, Section XI.
- 2. Adequate clearances shall be provided for in;ervice inspection of the Reactor Coolant Pressure Boundary and the ASME B&PV Code Section III, Class 2 portions of the Main Steam, Main Feed, Emergency Feed, and Blowdown systems' piping, in accordance with the provisions of Section XI of the ASME Boiler and Pressure Vessel Code.
- 3. Biological shielding and all other insulation, if installed around the Reactor Coolant Pressure Boundary, shall be designed to afford access for inservice inspection as defined by Section XI of the ASME Boiler and Pressure Vessel Code.
4 The pressurizer manway shall be accessible for internal examination of the pressurizer. K. Chemistry / Sampling
- 1. A sampling system which provide a means of obtaining remote liquid samples from the RCS for chemical and radiochemical labor-NM atory analysis shall be provided. The sampling system shall be designed to allow for the following tests: corrosion product 6 activity levels, dissolved gas, fission product activity, chloride concentration, coolant pH, conductivity levels and boron concen-tration. The pressurizer steam space sample lines shall contain 7/32" x l orifice as close to the pressurizer as possible. The sample system shall be as shown on Figure 5.1.2-1.
- 2. A system or systems shall be provided to maintain the steam generator secondary water chemistry within Section 10.3.4 specifi-cations during plant operation. The system or systems shall incorporate steam generator blowdown, chemical addition, and monitoring.
- 3. Provisions shall be made to allow sampling of the RCS during Shutdown Cooling 9 stem operation.
4 Provisions shall be made to allow sampling of the RCS during startup. L. "aterials
- 1. The materials used for the containrent and its internal structures shall be compatible with both the normal operating environment and the most severe thermal, chemical, and radiation environment expected during post-accident conditions (refer to Section 3.11
- p. 5 of 5 June 28, 1985 5.1-13 Amendment No. 10
- 1. *
, Enclosure (3) to LD 049 ENCLOSURE 3 STEAM GENERATOR TUBE BURST TEST DATA (DATA IS FROM BATWING TUBE WEAR EVALUATION PROGRAM)
? l
- p. 1 Of 5
-9 10-The following data represents results from actual and simulated tube burst testing performed as part of the steam generator Bat-Wing Evaluation Program. The testing shows the effects of tube leakage from flat wear scars of varying depths, both for normal and accident primary to secondary pressure differentials. The data demonctrates that with 100%
through wall wear, leakage rates remain minimal. Also, burst pressures for the 50 and 75 percent through wall wear are considerably higher than the expected normal and accident differential pressures. l l l l l
- p. 2 of 5 1
t
4 ENCIDSURE 3 '-
- c, ,
SUMMARY
OF TUBE LEAKAGE AND BURST TESTING 3/4 INCH 0.D. 0.048 INCH WALL TUBES
. TUBE WEAR AT BATWING i
LEAKAGE RATE (GPM) . SAMPLE- % ' BURST DEFECT THRU- NORMAL OPERATION MSLB - SPIKE POST MSLB S.S. PRESSURE I.D. WALL- AP = 1350 PSI AP = 1750 PSI AP = 1300 PSI (PSI) A-1 100 0.23 - - - A-2 100 0.63 - - - A-3 100 0.15 - - - A-4 50 - - - No Burst A-5 50 - - - 9825 A-6 75 - - 7200 A-7 100 0.11 0.11 0.08 - A-8 100 0.37 0.42 0.38 - A-9 100 0.12 0.12 0.13 -
~
B-1 100 0.13 0.44 0.38 - B-2 100 0.49 0.62 0.55 - B-3 100 0.61 0.94 0.77 - B-4 50 - - - 8725 B-5 50 - - - 8450 B-6 75 - - - 6700 C' 100 0.15 - - - C-2 100 0.28 -- - - C-3 100 0.39 - - - C-4 50 - - - No Burst C-5 50 - - - 8800 C-6 75 - - - 7000 i
- p. 3 of 5
- t;
$o SUPNARY _.0F TUBE BURST, COLLAPSE AND LEAKAGE TESTING 3/4 INCH 0.0., 0.042 INCH WALL TUBES (SYSTEM 80) { TUBF WEAR AT BATWING Leakaae Rate (GPM) Burst Collapse Sample - % -- Defect Thru Normal Operation MSLB Steady State Press. Press I.D. Wall AP = 1180 psi AP = 1900 psi -(psi) (psi) D-1 100 0.35 0.92 - - D-2 100 0.49 3.86 - - D-3 100 0.09 0.94 - - D-4 50 - - 7800 - D-5 75- - - 5300 - E-1 100 0.25 0.77 - - E-2 100 0.10 2.65 - - E-3 100 0.14 2,23 - - E 50 - - 8200 - E-5 75 - - 48S0 - F-1 100 0.50- 2.95 - - F-2 100 0.18 0.61 - - F-3 100 0.11 1.98 - - F-4 50 - - 8225 - F-5 75 - - 4950 - i G-1 90 - - - 2500 G-2 90 - - - 2525 G-3 90 - - - 2500 G-4 75 - - - 2800 G-5 75 - - - 3050 G-6 75 - - - 3100 t
- p. 4 of 5
.o ,
h SUPNARY OF TUBE BURST AND LEAKAGE TESJgG 3/4 INCH 0.0., 0.048 INCH WALL TUBE SIMULATED FLAT TUBE WEAR Normal operation MSLB Spike Post MSLB S.S. Sample AP = 1350 psi (3) 3p 1750 psi (3) AP : 1300 psi Burst Defect Actual Actual Pressure
% Bar Leak Leak Actual Leak I.D. Thru Width AP Rate AP Rate AP Rate (psi)
Wall (in.) (psi) (GPM) (psi) (GPM) (psi) (GPM)
?
- VXI 100 0.5 1160 0.37(I) - - - - -
a vX2 100 0.5 1360 0.12 - - - - - w VX3 100 0.5 1370 0.27 - - - - - VXC 100 0.5 1365 0.13 1760 0.82 1330 0.66 - VXS 100 0.5 1410 0.30 1760 0.47 1300 0.41 - VX6 70 0.5 - - - - - - 6700 VY1 100 1.0 1300 0.64 - - - - - VY2 100 1.0 1350 0.31 - - - - - VY3 100 1.0 1370 0.56 - - - - - VY4 100 1.0 1370 0.16 1000 >20(2) _ _ _ , VY5 100 1.0 1355 0.17 1760 0.82 1320 0.66 - Y VY6 70 1.0 - - - - - - 5700 Notes: (1) Unintentional pressure spike caused an increase in leak rate to 5.14 GPM. (2) Tube defect "opened up" before 1750 psi was achieved and pressure dropped off to 1000 psi. (3) 3410 Mwt parameters shown change to 0.042 inch wall, AP = 1180 psi and AP = 1900 psi norm MSLB for System 80 steam generators. N . -_ _ _ _ _ ---_--____- _}}