ML050960052
| ML050960052 | |
| Person / Time | |
|---|---|
| Site: | Vermont Yankee  |
| Issue date: | 03/31/2005 |
| From: | Hwang H, Mehta H, Schrag M, Shekhtman I Entergy Nuclear Operations |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| BVY 05-034, DRF Section 0000-0038-0936 GE-NE-0000-0038-0936-NP | |
| Download: ML050960052 (101) | |
Text
Attachment 6 to BVY 05-034 Docket No. 50-271 Attachment 6 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 26 Extended Power Uprate Steam Dryer Analyses and Monitoring GE-NE-0000-0038-0936-NP Vermont Yankee Nuclear Power Station Modified Steam Dryer Stress Analysis Non-Proprietary Version l Total number of pages in Attachment 6 l (excludina this cover sheetl is 54.
I . i a, .7 1 ,. o -I z GE-NE-0000-0038-0936-NP Class I DRF Section 0000-0038-0936 March 2005 Non-proprietaryVersion Entergy Nuclear Operations Incorporated Vermont Yankee Nuclear Power Station Modified Steam Dryer Stress Analysis W&W, Principal Contributors:
I I H. S. Mehta ,,K I I I. Shekhtman A/
Principal Verifier:
H.L. Hwang I iII Manager: I / I A0 fJ.A:
/ I r Mike Schrag I I
U ',tw-
, 'It.
I 9 .,is*,.9
. 0, I* - -
7K
GE-NE-0000-0038-0936-NP Non-proprietary Version NON PROPRIETARY NOTICE IMPORTANT NOTICE This is a non-proprietary version of the document GE-NE-0000-0038-0936P, which has the proprietary information removed. Portions of the document that have been removed are indicated by an open and closed bracket as shown here (( )).
IMPORTANT NOTICE REGARDING THE CONTENTS OF THIS REPORT Please Read Carefully The only undertakings of the General Electric Company (GE) respecting information in this document are contained in the contract between Entergy Nuclear Operations Incorporated and GE, Order 4500528282, Schedule A-2, as amended to the date of transmittal of this document, and nothing contained in this document shall be construed as changing the contract. The use of this information by anyone other than Entergy Nuclear Operations Incorporated, for any purpose other than that for which it is furnished by GE, is not authorized; and with respect to any unauthorized use, GE makes no representation or warranty, express or implied, and assumes no liability as to the completeness, accuracy, or usefulness of the information contained in this document, or that its use may not infringe upon privately owned rights.
i
GE-NE-0000-0038-0936-NP Non-proprietary Version Table of Contents 1.0 Executive Summary .................... .I 2.0 Background and Introduction .................. 2 3.0 Dryer Analysis Approach ................... 6 3.1 Operational Pressure Loading .................................................................... 6............6 3.1.1 Acoustic Pressures ................................. 6 3.1.2 Vortex Shedding Pressures ................................. 7 3.2 Time History Analysis with Modal Superposition ................................. 9 3.2.1 Time Step Selection and Sensitivity .......... ....................... 9 3.2.2 Damping ....................... 10 3.3 Model Material Properties ....................... 10 3.4 Design Criteria ........................ 11 3.5 DryerLoads ........................ 13 3.5.1 Transient Pressure Loads ....................... 13 3.5.2 Seismic Loads ...................... 13 3.6 Load Combinations .. 13 3.7 FEA Analysis Model .. 14 3.8 FIV Stress Determination ......... .. 15 3.9 Weld Stress Concentration Factors ...................... 16 4.0 Modified Dryer Analysis Results. ...................................................................................... 22 4.1 Acoustic Pressure Stresses ................................................ 22 4.2.1 Sensitivity Results..................................................................................... 22 4.2 Vortex Shedding Stresses ............................................... 22 4.3 Evaluation of Primary Stresses under ASME Code Section I Loads .. 22 4.4 Evaluation of FIV Stresses vs. Endurance Limit . ...............................23 5.0 Conclusion ............................... 30 5.1 Fatigue Alternating Stresses ............ ................. 30 5.2 ASME Primary Stresses Evaluation ........... .. ................ 30 6.0 References......................................................................................................................... 31 Appendix A Vermont Yankee Steam Dryer Differential Pressure . .32 Appendix B Acoustic Pressure Stress Intensity Contour Plots . .35 iii
GE-NE-0000-0038-0936-NP Non-proprietary Version List of Figures Figure 2-1 Modified Dryer.....................................................................................................4 Figure 2-2 Modification Details ............................. 5 Figure 3.1-1 Examples of Pressure Contour Plots ............................ 17 Figure 3.1-2 Pressure Differential Across Vertical Plate ............................ 18 Figure 3.1-3 Pressure Differential Across Left Gusset ............................ 18 Figure 3.1-4 Pressure Differential Across Right Gusset ............................ 19 Figure 3.1-5 PSD for Point 6 of LES Model ............................ 20 Figure 3.7-1 Modified Full Dryer Analysis Model ............................ 21 Figure A-1. Typical TSV Load Time Histories ............................ 34 Figure B-I Inner hood base plate ............................ 36 Figure B-2a Modified outer cover plate, includes tips ............................ 36 Figure B-2b Modified outer cover plate, exclude tips ............................ 37 Figure B-3a Original top hood (all hood) ............................ 37 Figure B-3b Modified top hood (outer hood) ............................ 38 Figure B-3c Hood top plates (Inner hood) ............................ 38 Figure B4a Original outer hood, strips ........................... 39 Figure B4b Modified outer hood, top weld ........................... 39 Figure B4c Modified outer hood, bottom weld ............................ 40 Figure B-5 Hood vertical plates (inner hood) ........................... 40 Figure B-6 Hood end plates,(inner hood) ........................... 41 Figure B-7 Hood end plates (outer hood) ........................... 41 Figure B-8 Inner hood brackets (gussets) ........................... 42 Figure B-9 Steam'darn ........................... 42 Figure B-10 Steam 'dam' gussets............................................................................................43 Figure B- I1 Hood partition plates ........................... 43 Figure B-12 Baffle plates ........................... 44 Figure B-13 Outlet plenum ends ........................... 44 Figure B-14 Dryer support ring ........................... 45 Figure B-15 Bottom cross beams ........................... 45 Figure B-16 Cross beam gussets ........................... 46 Figure B-17a New gusset on cover plate and front hood ........................... 46 iv
GE-NE-0000-0038-0936-NP
- Non-proprietary Version Figure B-17b Gusset foot weld to cover plate ............................. 47 Figure B-I8 Outer hood, bottom weld ............................. 48 Figure B-19 Cover plate ............................. 49 List of Tables Table 3.3-1 Properties of SS304L and SS316L ............................................ 1I Table 3.4-1 Primary Stress Limits ............................................ 12 Table 3.6-1 ASME Code Section III Load Combinations ............................................ 14 Table 4.2-1 Time Step Sensitivity Results ............................................ 24 Table 4.3-1 Modified Dryer Stresses at the Outer Cover Plate ............................................ 24 Table 4.3-2 Modified Dryer Stress at the Original Front Hood Both Side Strips ................. 25 Table 4.3-3 Modified Dryer Stress at the 1.0 Inch Front Hood Lower Weld ........................ 25 Table 4.3-4 Modified Dryer Stress at the Unmodified Top hood .......................................... 26 Table 4.3-5 Modified Dryer Stress at the 1" Thick Top hood ............................................ 26 Table 4.3-6 Modified Dryer Stress at the Long Gussets ............................................ 27 Table 4.3-7 Modified Dryer Stress at the Vertical Side Hood ............................................ 27 Table 4.4-1 FIV Stress Summary, Nominal for CLTP ............................................ 28 Table 4.4-2 FIV Stress Summary, +10%/o Time Step ............................................ 29 Table 5.1-1 FIV Stresses for Key Dryer Components ............................................ 30 Table 5.1-2 FIV Stresses (+10%/o Time Step) for Key Dryer Components ............................ 30 Table A-i. Steam Dryer Pressure Differentials for Normal Conditions at EPU .................. 32 Table A-2. Dryer Delta-P for EPU Forward Flow Upset Conditions .................................. 32 Table A-3. Maximum TSV Acoustic Load on the Dryer Face at EPU ................................ 33 V
GE-NE-0000-0038-0936-NP Non-proprietary Version 1.0 Executive Summary As a result of BWR/3 steam dryer failures in the industry, Entergy has performed a comprehensive structural evaluation to demonstrate the adequacy of the Vermont Yankee (VYNPS) dryer integrity for extended power uprate (EPU) steam flow changes. The key issue has been the modeling of the flow-induced vibration (FMy) loading during normal operation.
The purpose of the analysis reported here is to update the FIV stresses using pressure fluctuating loads generated with the combination of an acoustic circuit model and a computational fluid dynamics (CFD) large eddy simulation (LES) model for vortex shedding. The acoustic circuit model, LES model and resulting alternating stresses will provide Entergy with a more realistic load definition and a basis to systematically monitor changes to predicted dryer stresses during power ascension to EPU conditions.
The VYNPS acoustic circuit model was developed based on plant design/operation configuration and used VY-specific measured pressure fluctuation data as input. The acoustic circuit methodology was benchmarked with scale model test results developed by GE for Entergy. The LES model characterizes the nature and magnitude of unsteady flow effects across the face of the dryer at the entrance to the main steam line nozzles.
Each model generated time history pressure profiles that were input to the ANSYS finite element program to determine associated FIV stresses.
Maximum acoustic pressure stresses and vortex shedding stresses were extracted from separate finite element analyses. The stresses were conservatively combined and, where appropriate, multiplied by stress concentration factors that account for weld shape and size. The time history analysis was done with a +/-10%/o time step change to conservatively account for uncertainty in the frequency content of the FIV loads.
The resulting maximum FIV stresses calculated for current power were about 25% of the 13,600 psi ASME endurance limit. In addition, normal, upset and faulted stresses were calculated and compared to ASME Code allowable, with all conditions showing acceptable stresses.
I
.GE-NE-0000-0038-0936-NP Non-proprietary Version 2.0 Background and Introduction As a result of significant steam dryer cover plate fatigue cracking at Quad Cities Unit 2, GE issued SRL 644 in August of 2002 to provide information to all BWR utilities on cover plate related failures. In September of 2003, GE added Supplement I to SIL No.
644 in order to describe additional steam dryer fatigue cracking at Quad Cities 2, and to explain that the root cause of the second event was different than the first. SIL 644 applied to BWRJ3-style steam dryer design plants. Supplement I to SR,No. 644 provided recommendations applicable to plants with BWR/4 and later design steam dryers. The objective of this report is to detail the latest analyses of the Vermont Yankee Nuclear Power Station (VYNPS) steam dryer that were performed for the modified dryer configuration. The purpose of these analyses is to confirm that the modified dryer meets ASME criteria for fatigue initiation and other ASME-based design criteria.
The VYNPS steam dryer modifications included replacing the 1/2 inch outer vertical plates and portions of the top hood plates with 1-inch plates, removing internal brackets that attached the internal braces to the outer hood plates, replacing the 1/4/inch thick cover plate with 5/1 inch thick material, and adding three long gussets at the outer vertical hood plate and cover plate junction. Each gusset is triangular in shape, 53 inches high, with no more than 1.0-inch width at the top. The top of each gusset is welded around and has a smooth transition to the modified front hood. Each gusset extends to within 5.5 inches of the top of the modified front hood. Transition between each gusset and the modified lower cover plate is accomplished via the use of a U-shaped "gusset extension" that is welded to both the lower cover plate and to each gusset. The replacement lower cover plate is attached with 1/2-inch welds all around except for the cbmer intersection with the dryer support ring where 5/8-inch welds are used for a distance of four inches. The existing tie-bars are replaced with a modified tie bar design. Figures 2-1 and 2-2 show the modified dryer configuration.
A brief summary of the previous analyses done is provided as background for the analysis results documented in this report. Reference [1] documents a stress analysis of the modified dryer based on equivalent static load methodology. The equivalent static pressure was based on the estimated pressure fluctuations on the dryer faces and an estimated dynamic amplification factor. In order to address the dynamic amplifications explicitly and to account for the sensitivity in modal frequency variations, a response spectrum approach based analysis was documented in [2]. However, the enveloping load definition used in this approach led to very high fatigue stresses such that fatigue failures were predicted at locations where the field history had shown no such failures.
Therefore, a scale factor developed consistent with known field failures had to be used to calculate realistic values of fatigue stresses. A review by the NRC [3] found this technical approach to be inadequate in that the scaling factor may not be the same at 2
GE-NE-0000-0038-0936-NP
- Non-proprietary Version various locations in the dryer. Several other deficiencies in that analysis were also identified in [3] (e.g., need for adding FIV primary stress in the ASME load combinations and consideration of loads due to a postulated break in the steam line outside the containment). The NRC review also identified the need for a plant specific identification of acoustic and other fluctuating pressure loading sources.
The analysis presented in this report addresses the deficiencies identified by the NRC audit. Specifically, an acoustic circuit model was used with VY-specific plant signals to develop acoustic loading on the dryer. This analysis and the ENVY steam dryer ascension test monitoring accounts for acoustic sources that may impact the dryer.
Additionally, vortex-induced loading on the front face of the dryer was also estimated through a transient CFD analysis. The stresses from the two sources were then combined to obtain total FIV fatigue stresses.
The acoustic model includes the affect of unsteady flow forces that originate at the exit to the vessel and the CFD model includes compressible flow and therefore accounts for vortex pressure forces. The addition of both sets of loads is conservative. The advantage of using both sets of loads is that the CFD model provides a more detailed assessment of unsteady dryer loads along the face of the dryer. Therefore lateral loads on the gussets and local vortex pressure gradients are considered. The acoustic model accounts for external piping acoustic pressures from plant data. The CFD model provides more conservative low frequency loads and the acoustic model provides more conservative high frequency loads. Therefore the addition of these load cases applies conservatism in both magnitude and frequency content for the FIV load.
The purpose of the analysis reported here is to update the current licensed thermal power (CLTP) analysis using pressure fluctuating loads generated with the combination of an acoustic circuit model for acoustic pressures and a large eddy simulation (LES) model for vortex shedding pressures. The acoustic circuit model, LES model and resulting alternating stresses will provide Entergy a basis to systematically monitor changes to dryer stresses while they increase power to extended power uprate (EPU) conditions.
3
GE-NE-0000-0038-0936-NP Non-propri etary Version
[1 11 Figure 2-1 Modified Dryer 4
GE-NE-0000-0038-0936-NP Non-proprietary Version 1]
Figure 2-2 Modification Details 5
GE-NE-0000-0038-0936-NP Non-proprietary Version 3.0 Dryer Analysis Approach The analysis presented follows the same approach as described previously for the modified dryer, except in the area of fatigue stresses, where the loads and the method of applying those loads to the finite element analysis (FEA) model have been updated.
3.1 Operational Pressure Loading The flow-induced vibration (FMy) loading as a result of the passage of steam through the steam dryer vane banks is a significant cyclic loading that has the potential to initiate and grow fatigue cracking. This FIV loading in the form of distributed fluctuating pressures is highly complex in that it varies as a function of the location and phasing, and has complex frequency content. The first analyses of the modified dryer [11] used an equivalent static pressure approach to represent the FIV loading. The second analysis [2]
used a response spectrum based on four instrumented dryer plants' measured pressure time histories, applied to VYNPS with a scaling factor based on steam line velocity differences. However, both technical approaches were judged inadequate from the point of view of actual plant loadings.
In the updated analysis, the FIV pressure loading was calculated from two sources: (I)
Acoustic Circuit Modeling, and (2) Large Eddy Simulation (LES). The total FIV pressure loading is represented by a combination of acoustic pressures and vortex shedding pressures. These pressure loads were provided by Entergy. Brief overview descriptions of the acoustic circuit model and the LES model are provided herein. The methods used to apply the pressures to the ANSYS FEA model are described as well.
3.1.1 Acoustic Pressures The acoustic circuit based pressure fluctuations were developed by Continuum Dynamics Inc. (CDI) [41 and supplied by Entergy as input to GE. The acoustic circuit model for the VYNPS steam path (i.e., steam dome, dryer and the steam lines) was developed by CDI and used VY-specific measured pressure fluctuation data as input. The ability of the acoustic circuit model to predict dryer pressure loads from main steam system fluctuating pressure measurements was benchmarked with scale model test results by Entergy [5].
The acoustic circuit model loads were applied to the GE finite element model of the dryer as time varying pressure vectors.
6
GE-NE-0000-0038-0936-NP Non-proprietary Version The CDI acoustic model solved for pressures in the RPV dome with a 3"x3" detailed grid. Differential pressures on components in the area of the dryer front face were then calculated at a 3x3 inch mesh. Based on discussion with GE, CDI provided coarser 6"x6" and 12"x12" summaries in areas away from the face where fluctuating loads were much lower and the pressure gradients permitted less refined grid information. This variable load grid help optimize the data processing efficiency of ANSYS load vectors.
Based on the CDI model and Grid information GE then developed an ANSYS Macro to translate the CDI supplied pressures into input pressures for the GE ANSYS finite element model. This macro was checked by comparing the resulting ANSYS load vectors at 28 key locations in the dryer against the CDI loads. The comparisons showed that the ANSYS load vectors and CDI pressure fluctuations matched, or that the ANSYS values were conservative.
3.1.2 Vortex Shedding Pressures Entergy has performed a CFD LES compressible flow simulation of the VY dryer. The purpose of this model is to assess vortex shedding and other unsteady flow effects along one face of the dryer and into the main steam -
nozzles. The CFD model was developed by Fluent and includes the dryer RPV dome and main steam piping to the steam equalizer approximating the total pressure loss in the system. This model includes a detailed refined mesh along one side of the dryer. This is depicted as the foreground in the pressure contour plots shown below.
The balance of the model includes sufficient detail to simulate flow throughout the vessel dome and piping.
This lower resolution portion of the model was required to assure flow in and out of the front face region was depicted accurately.
The purpose of this model is to characterize the nature and magnitude of unsteady flow effects across one face of the dryer at 100% CLTP conditions. This CFD model is very large. It has nearly 4.5 million cells and on a system with up to 12 parallel processors requires a full day to produce about 0.7 second of simulation.
7
GE-NE-0000-0038-0936-NP Non-proprietary Version J
Included in Figure 3.1-1 is an example of figures that depict differential pressure across major plate components of the VY dryer at 17 time points (at roughly 0.25 second intervals). These contour plots provide a good sense of how the predicted differential pressure variation changes with time. In general it can be noted the more severe pressure gradients are adjacent to the steam nozzles. These are caused by a vortex as the steam exits the vessel. This vortex forms then disappears then reappears less pronounced at the adjacent nozzle. In earlier CFD analysis performed with a clipped domain incompressible LES solution it was noted that after 8 seconds the vortex pressure oscillations subsided. It is likely that with more time, some of the oscillations currently observed with this much larger model may also subside.
In general the other contours away from the nozzle do not undergo significant change.
The dryer face is subject to a -0.8 psi pressure differential as the higher pressure, slower moving steam within the dryer comes over the top and accelerates into the steam nozzles.
The 17 snapshots of data presented here are just a data sample and not continuous record.
After eight months of model development and months of computer time, this CFD results permit quantification of the impact of CFD loads variations on dryer fatigue.
In the fatigue assessment for CFD loads, the alternating stress caused by the unsteady CFD differential pressures was evaluated. The CFD model provides -18,000 pressure values across the vertical plate and -2500 element areas of the two outer gussets.
Therefore to help interpret the delta-P data we have provided three graphs, Figures 3.1-2 through 3.1-4. These graphs depict both the average pressure and maximum pressure differential across the vertical plate, the left gusset, and right gusset. These are the components most affected by the CFD forces. The coming and going of the vortex 8
GE-NE-0000-0038-0936-NP Non-proprietary Version directly impacts the maximum pressure differential across each component. The change in average pressure impacts the ballooning of the vertical face and gussets.
The large changes in the vortex movement occur very slowly. The low pressure regions and high differential pressures take over a second to change. In addition to the very slow vortex changes, there are much lower magnitude, higher frequency changes occurring also. Entergy has generated detailed pressure data on the dryer face at a 0.0005 sec interval (2000 Hz). The power spectral density (PSD) plots, for this data are included in Figure 3.1-5.
Point 6 represents frequency content close to the vortex. The vortex forces have negligible impact to the I" dryer face and 5/8" cover plate. The vortex pressures do have a localized effect resulting in lateral bending loads on the gussets. This would impact the 2nd mode natural frequency of 99 Hz of the gussets'. The majority of the Point 6 energy is below 45 Hz.
3.2 Time History Analysis with Modal Superposition The dynamic structural response of the dryer to the applied pressure fluctuations from acoustic loading was calculated using a time history method with modal superposition.
Normal modes up to 200 Hz were included in calculating the response. Entergy has collected transient data during multiple full power operation periods. CDI used a representative 16 second set of data and performed a coarse transient solution. GE requested that Entergy provide approximately 2 seconds of transient fine mesh data for their analysis. Therefore CDI provided a 2.4 second subset of fine mesh data that contained the highest Prms and peak data of their 16-second analysis. The plant data acquisition for the strain gages and pressure transmitters was at 1024 Hz and 2000 Hz respectively. From this data, CDI developed the dryer transient data at 1024 Hz time rate. The selection of modes up to 200 Hz in the modal superposition analysis is consistent with the preceding time interval between the acoustic pressure values and captures the important mode shapes that contribute to fatigue stress in the model.
3.2.1 Time Step Selection and Sensitivity The time step value used in the time history analysis was 0.000977 second. It is an accepted engineering practice to assess the sensitivity of the calculated stresses to such factors as differences in modal frequency due to geometric and material variations, l GE-NE 000-0026-7944-1 pgs 44. Vermont Yankee Nuclear Povcr Station Steam Dryer Modification, Revision 1,March 2004.
9
GE-NE-0000-0038-0936-NP Non-proprietary Version random variations in pressure time history, etc. A sensitivity assessment was conducted by varying the time interval between the pressure time steps by (( )). This is equivalent to peak broadening in the response spectrum analysis method. The (( ))
variation was judged to be a reasonable value to capture instances where a structural mode that contributes significant response may have its frequency very close to any one of the frequencies present in the fluctuating pressure time history. The ENVY Acoustic Model Benchmark Report [5] demonstrates that the (( )) variation provides a conservative envelope of transient data. For dryer components with the higher FIV stresses and positive changes in stress intensity with time step change, the FIV stress increase was assessed relative to the endurance limit.
3.2.2 Damping A conservative damping value of (( )) of the critical damping was used in the modal superposition analysis. Technical justification for the selection of this value is provided in [6]. This value was applied uniformly for all the modes.
3.3 Model Material Properties The original VYNPS dryer assembly was manufactured from solution heat-treated SS304 conforming to applicable ASTM standards at the time of manufacture. The modification plate is made from SS316L. Minimum of SS304L and 316L properties from [7] were used to conservatively envelop the properties of the unmodified components and the modification plate. In actuality the stress intensity limit, S.,, of SS304 is slightly greater than SS304L2 . Therefore the use of SS304L material properties for unmodified dryer components is conservative. The applicable properties are shown in Table 3.3-1.
2 The ASME code 1971 and 1989 editions have the following material properties for SS304:
ASME 1971 - S,, = 17.4 ksi at 5000F,S,, = 16.4 ksi at 6000F ASME 1989 - S. = 17.5 ksi at 5000F,S,, = 16.4 ksi at 6000F 10
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 3.3-1 Properties of SS304L and SS316L Room temperature Design temperature Material I Property 700F5751F SS304L S,, Stress intensity limit, psi 16700 14,200 Sy, Yield strength, psi 25000 15,700 Su, Ultimate strength, psi 70000 57,200 Sem, Endurance limit, psi 13,600 E, Elastic modulus, psi 28300000 25575000 SS316L Sm, Stress intensity limit, psi 16700 13,725 Sy, Yield strength, psi 25000 15.225 S", Ultimate strength, psi 70000 61600 S,,, Endurance limit, psi 13,600 E, Elastic modulus, psi 28300000 25575000 3.4 Design Criteria The analysis uses the ASME Code [8] as a design guide although the dryer is not an ASME Code component. Specifically, structural adequacy for Service Level A and B loads was investigated using the corresponding stress limits of [8] with the exception of application of the weld quality factors. Weld quality factors are described in the ASME code Table NG-3352-1 for safety components, such as the reactor pressure vessel, that contain radioactive fluid. Because the steam dryer is not a safety-related pressure retaining component ((
)) The requirement of 'no loose parts' during Service Level D events was investigated using stress limits of Subsection NG and Appendix F of the ASME Code [9]. The stress limits are summarized in Table 3.4-1. (Note that for completeness, application of the seismic loading in the (( )) direction was considered).
11
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 3.4-1 Primary Stress Limits Service level Stress category Stress limit Service levels A & B Pm Sm Pm + Pb 1.5 Sm Shear stress 0.6 Sm Bearing stress 1.5 Sy
£ fatigue damage 1.0 Service level D Pm min (2.4Sm, 0.7S%)
Pm + Pb min (3.6%m, 1.05Su)
Shear stress 1.2 Sm Bearing stress 3.0 Sy Pm: Primary membrane stress intensity Pb: Primary bending stress intensity Sm: Stress intensity limit Sy: Yield strength S.,: Ultimate strength 12
GE-NE-0000-0038-0936-NP
- Non-proprietary Version 3.5 Dryer Loads 3.5.1 Transient Pressure Loads The FIV pressure loading on the dryer components is described in Section 3.1. The other normal operating and transient pressure loadings on the dryer components for other service conditions are described for EPU conditions in Appendix A. They are used here as conservative CLTP loads.
3.5.2 Seismic Loads The seismic loads on the VYNPS dryer are documented in [10]. These seismic loads are unchanged with EPU, were used in the [1] and [2] evaluations and are used here. The accelerations are listed below.
1]
Because the modified dryer first mode frequency is in the ZPA (Zero Period Acceleration) range of the seismic load ((( ))), the time history maximum acceleration (g) load is the dryer acceleration (g) load considered in a static.
analysis.
3.6 Load Combinations The VYNPS steam dryer was originally procured and supplied as a non-safety related, non-seismic category I, non-ASME component There was no design specification for the dryer and, as such, the service conditions for the steam dryer were not specifically defined. However, as late as 1969, an internal GE design report, 257HA760, was prepared for BWR-3 style steam dryers. In the 257HA760 report the following service condition and acceptance criterion were stated:
- The principal design loads considered in the analysis of the steam dryer assembly are the weight loads and the pressure loads, which are present during accident conditions.
- In the event of a guillotine steam line break outside the drywell, dryer design must preclude the possibility of dryer debris entering the steam line and interfering with isolation valve closure.
- The structural elements, which hold the dryer in place, are designed to accommodate the pressure loading due to a break outside the isolation valves within the ASME Code,Section III stress criteria. The flat panels, which form partitions in the dryer, 13
GE-NE-000-0038-0936-NP Non-proprietary Version are designed so that the elastic collapse loading on these panels is not exceeded under these same pressure loadings.
The above criteria continue as design bases for the modified VYNPS steam dryer, both at CLTP and EPU conditions. The VYNPS steam dryer design basis continues to be structural integrity after a steam line break outside of containment. However, due to the operating experience related to steam dryer structural integrity associated with normal operation, this analysis includes Normal and Upset case loading combinations. Analysis of Normal and Upset cases addresses the concern that frequent and moderately frequent events which do not require an immediate inspection of the dryer should not degrade the dryer condition to the point that it might not meet its Level D design criteria.
Table 3.6-1 provides the load combinations and load cases used in the VYNPS dryer analysis.
Table 3.6-1 ASME Code Section III Load Combinations Case Service Level Load Combination 3.7 FEA Analysis Model Transient and static stress analyses were performed using the ANSYS finite element code (Versions 6 and 8) running under Windows 2000 Operating System [I1]. The model is a full dryer model, as shown in Figure 3.7-1, which incorporates the modifications made to the VYNPS dryer in 2004. To allow more efficient time history execution, the model in
[121 was simplified by effectively removing the detailed tie bar and steam dam repair modeling on the top hood. The half of the model without these details was replicated by symmetry to replace the half with the details.
14
GE-NE-0000-0038-0936-NP Non-proprietary Version In final verification of the model and completed analysis cases, one node was found in the modified hood that was inappropriately constrained by the beam below. The node was corrected and a static case run to assess the stiffness changes. The modified hood stress decreased as a result and the attached end plate stress increased. FIV stresses were scaled accordingly in these two components.
3.8 FIV Stress Determination The FIV stresses are taken as a combination of the acoustic and vortex shedding contributions. The acoustic pressure FEA results for each dryer component of interest are screened for the maximum stress intensity throughout the -2500 time steps. The stress intensity is conservatively used as the acoustic contribution to the FIV stress amplitude.
Because the majority of the vortex shedding load energy is well below the natural frequency of the affected components, static analysis alone provides a reasonable assessment of alternating stress and fatigue impact under the changing pressure loads described in Section 3.1.2. Therefore, the ANSYS FE model of the dryer was first used to statically assess the pressure cases to determine the magnitude of the CFD load related alternating stresses. Based on review of the pressure contours and preliminary static stress analyses, Entergy selected 7 of the 17 cases, which represent the key load conditions for the determination of representative alternating stress ranges. The conditions selected for ANSYS analysis included 100%b CLTP time steps at 2.20sec, 3.20 sec, 4.45 sec, 4.70 sec, 4.95 sec, 5.20 sec. and 5.23 sec.
Based on the ANSYS results the following 10 load case combinations for the alternating stress calculation were selected:
- 1. 2.20 sec to 3.20 sec
- 2. 2.20 sec to 5.23 sec
- 3. 3.20 sec to 5.20 sec
- 4. 3.20 sec to 5.23 sec
- 5. 4.45 sec to 4.70 sec
- 6. 4.45 sec to 5.20 sec
- 7. 4.70 sec to 4.95 sec
- 8. 4.70 sec to 5.20 sec
- 9. 4.95 sec to 5.20 sec
- 10. 4.95 sec to 5.23 sec These cases are selected to maximize alternating stress but also biased to later time points under the assumptions that as time progressed the CFD model would be converging on a steady state solution.
15
GE-NE-0000-0038-0936-NP Non-proprietary Version The stress intensity from the acoustic pressure analysis is added by absolute sum to the maximum stress range from the vortex shedding pressure analysis. This is conservative because the acoustic stress intensity is larger than the maximum acoustic FIV stress of interest and because the acoustic and vortex shedding maximum stresses likely do not occur at the same location in the component.
3.9 Weld Stress Concentration Factors The following stress concentration factors (SCFs) were used in this evaluation: ((
)) These SCFs were applied to the calculated peak stresses ftom the finite element analyses. The use of peak stress to multiply with SCFs provides alternating stresses at welds consistent with recommended SCFs in the ASME Code. The technical basis for these values is provided in [13]. These values are also consistent with those used in the previous stress analyses for VY dryer (e.g., [I]).
16
GE-NE-0000-0038-0936-NP Non-proprietary Version
, 1!_ .: .!-
. . i : . !i ,
Z _ ' -' .:'
_ -- . . . l
- ' - -, . ]
a3 : : -. l 4 _ -
{_ : -
33: .
E_ 'l W_
E_
8 a3ns3_ t :
a_
=_ . n .
a_ w
_s 4 R
S_
l
- . s s; : t ! 11 l
i';i i e5, i "
_eCE_ ..
1_: .' ' ' _
=5D ' i i
' - i
- - . Z.
1 : l e [ R s
lulk E
3_
f
- x5 k_ '
. SH l ' 'S' a _ 't; ZIIiC^
Y 1 . . Si i3 A 11 1 l] =, !, ,
ill ia l i l i; 1 i. i Bi;l11X"i Ili;i; Figure 3.1-1 Examples of Pressure Contour Plots 17
GE-NE-0000-0038-0936-NP Non-proprietary Version CFD Cifferenial Pressure 100% CLUP Front Face of the Dryer 30_ .
I' "-S'--; --- 0'"}
2- J J. I Ig 04 a0.3 10.31 *43 "Z46 P256 p2.7 P2.21 *42 " 462 1.380 t-420 1P4.4 1.4.7*-US 0*420 P623 0APM".3.
TV" 36w I -MaWmumO~fe(rei1 '-----AverageI Figure 3.1-2 Pressure Differential Across Vertical Plate CFD Differential Pressure 100% CLUP Gusset X-23" Left I
II I 2I 25
-:i I
IU 15 m er !34T r4m l I l 0 i l l l l 02 al t l ll ll }l l l l l l lit l l l l l ip.1.6 0.220 0.2.46 1.214 1.2)0 1.195 V12 .3-45 W.12 1.3.70 PS"0 1.430 14-45 t47 1.4.95 "G.: 1$.6 lTMOSP
- Maximum giffevental . Average I Figure 3.1-3 Pressure Differential Across Left Gusset 18
GE-NE-0000-0038-0936-NP Non-proprietary Version CFD Offerential Pressure 100n/. CLTP Gusset Xz+23' Right 461 0.8I
.1 I t l. 1.120 0245 1.2. 1 2.70 I."5 t42 6345 4.52 t4.70 t4.5 t-4.20 6-4.45 14.7 W4J5 1.20 .523 Th. slp j_-Mixunun 0erOtaenth AverageI Figure 3.1-4 Pressure Differential Across Right Gusset 19
GE-NE-0000-0038-0936-NP Non-proprietary Version PM data forCFD Results Front Fae of the Dryer. resarnled to 2000Hz
&OOE.04 N
I?
5 25 45 65 a5 105 125 145 165 185 Frequency Hz I -Ptt 2 opofFoce - Ptt Botom ofFae I Figure 3.1-5 PSD for Point 6 of LES Model 20
GE-NE-0000-0038-0936-NP Non-proprietary Version
[C 1]
Figure 3.7-1 Modified Full Dryer Analysis Model 21
GE-NE-0000-0038-0936-NP Non-proprietary Version 4.0 Modified DryerAnalysis Results 4.1 Acoustic Pressure Stresses The ANSYS model produces stress intensities from the acoustic pressure zero-to-peak amplitudes. Stress intensity contour plots for each component are included in Appendix B. The maximum stress intensity for each component was extracted from the analysis for use in Table 4.4-1. The applicable primary stresses from the extractions were used as part of the primary stress evaluation for the components shown in Section 4.3.
4.2.1 Sensitivity Results Table 4.2-1 shows the relative change in maximum stress intensity for the acoustic pressure stresses as a result of the ((
ff time step changes. In a couple of instances, the stresses increased in the f[ )) time step case. The potential impact to FIV results for these increases is discussed in Section 4.4.
4.2 Vortex Shedding Stresses The 10 load case combinations described in Section 3.8 were screened for the maximum surface stress range for each component. Then from the 10 maximum results, the largest was selected to combine, as an amplitude, with the acoustic stress intensity value in Table 4.4-1. The vortex shedding maximum stress may not occur at the same location as the acoustic pressure maximum stress, so this is an additional conservatism in the FIV stresses.
4.3 Evaluation of Primary Stresses under ASME Code Section III Loads Each of the load combination tabulated in Table 3.6-1 has been analyzed for the modified dryer in [1]. With the exception of the FIV stresses, the other stresses are calculated for EPU conditions, which is conservative across the board for CLTP evaluation. In the ASME Code Section III load analysis, the dynamic loads, such as OBE, SSE and TSV loads should be combined by square root of the sum of the square (SRSS). This analysis combined the dynamic loads by algebraic sum. Because the OBE and SSE have been input in both the positive and negative directions and both results are compared with the allowable limits, the results are equivalent to absolute sum results. Therefore, the load combinations are conservative. The TSV pressure impulse load has been multiplied by a dynamic load factor of 1.5 in the input for ANSYS analyses.
22
GE-NE-0000-0038-0936-NP Non-proprietary Version Tables 4.3-1 through 4.3-7 tabulate the stresses for each weld with the undersize weld factor included. The maximum stress ratio is 0.76, at the gusset, due to the Service Level B-3 load combination. There is more than 20% margin for any ASME Level A or Level B load combination.
4.4 Evaluation of FIV Stresses vs. Endurance Limit Table 4.4-1 shows the CLTP acoustic and vortex shedding maximum stress amplitudes for each key dryer component. These stresses are conservatively combined by absolute sum and the sum is multiplied by the weld undersize factor and the weld stress concentration factor to determine the FIV stress on each component. The maximum value for CLTP conditions is well below the endurance limit.
If the stress increases to acoustic stress of the increased time step are factored into the FIV stresses, the results are the values in Table 4.4-2.
The FEA for FIV stresses is a linear analysis, so the Entergy approach of monitoring main steam system fluctuating pressure measurements during power ascension to EPU conditions, in accordance with the VYNPS Steam Dryer Power Ascension Test Plan (SDPATP) is appropriate using these results. If predicted dryer FIV pressures approach the endurance limit, FEA can be employed to determine FIV stresses.
23
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 4.2-1 Time Step Sensitivity Results 1]
Table 4.3-1 Modified Dryer Stresses at the Outer Cover Plate
((
1D 24
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 4.3-2 Modified Dryer Stress at the Original Front Hood Both Side Strips
((
I]
Table 4.3-3 Modified Dryer Stress at the 1.0 Inch Front Hood Lower Weld I]
25
GE-NE-0000-0038-0936-NP
- Non-proprietary Version Table 4.3-4 Modified Dryer Stress at the Unmodified Top hood
[E 1]
Table 4.3-5 Modified Dryer Stress at the 1" Thick Top hood 11 26
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 4.3-6 Modified Dryer Stress at the Long Gussets
[I 1]
Table 4.3-7 Modified Dryer Stress at the Vertical Side Hood
[I 27
GE-NE-0000-0038-0936-NP Non-proprietary Version -
Table 4.4-1 FIV Stress Summary, Nominal for CLTP 28
GE-NE-0000-0038-0936-NP Non-proprietary Version Table 4.4-2 FIV Stress Summary, +10% Time Step 11 29
GE-NE-0000-0038-0936-NP
- INon-proprietary Version 5.0 Conclusion 5.1 Fatigue Altemating Stresses The FIV alternating stresses shown in Table 5.1-1 are well below the endurance limit.
Similarly, the FIV alternating stresses for a (( )) time step in Table 5.1-2 are below the endurance limit of 13,600 psi.
Table 5.1-1 FIV Stresses for Key Dryer Components Item Current FIV Endurance
. Results (psi) Limit (psi)
Front Vertical Hood Bottom Weld 2,222 13,600 Front Hood Gusset Weld 3,238 13,600 Front Vertical Hood Top Weld 2,417 13,600 Table 5.1-2 FIV Stresses ((( D] Time Step) for Key Dryer Components Item Current FIV Endurance
_ Results (psi) Limit (psi)
Front Vertical Hood Bottom Weld 2,810 13,600 Front Hood Gusset Weld 3,535 13,600 Modified Hood Top Hood 2,900 13,600 5.2 ASME Primary Stresses Evaluation The modified dryer meets the ASME Code Section 111, Service Level A, B, C and D, primary stress criteria for all the load combinations tabulated in Table 3.6-1 using conservative methods of analysis.
30
GE-NE-0000-0038-0936-NP Non-proprietary Version 6.0 References
[1] GE-NE 0000-0024-7944-1, Vermont Yankee Nuclear Power Station Steam Dryer Modification, Revision 1,March 2004.
[2] GE-NE 0000-0024-7944-3, Vermont Yankee Nuclear Power Station Steam Dryer Dynamic Response Spectrum Analysis, Revision 0, June 2004.
[3] Docket No. 50-271, Technical Audit of Vermont Yankee Steam Dryer Analysis in Support of Extended Power Uprate Request, USNCR, August 24-26, 2004.
[4] C.D.I. Technical Memorandum No. 05-06, Analysis of Steam Dryer Differential Pressure Loads at Vermont Yankee, Revision 0
[5] Entergy Engineering Report No. VY-RPT-05-00006 Rev 0, Acoustic Model Benchmark, Dryer Acoustic Load Methodology, 3-29-2005
[6] Wu, Richard, Damping Value for Steam Dryer Dynamic Analysis, GE eDRF 0000-0034-1855, November 2004.
[7] American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code,Section III, Appendix 1, 1989 Edition with no Addenda.
[8] American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code,Section III, Subsection NG, 1989 Edition with no Addenda.
[9] American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code,Section III, Appendix F, 1989 Edition with no Addenda.
[10] MPR Calculation Number 319-002-01, transmitted by MPR letter H. William McCurdy to JR Hoffman, "Vermont Yankee Seismic Data for Increased Core Flow Evaluation," May 29, 1997.
[11] ANSYS, Releases 6.1 and 8.1, ANSYS, Incorporated, 2002.
[12] GE-NE 0000-0024-7944-3, Vermont Yankee Nuclear Power Station Steam Dryer Modification with Response Spectrum Analysis, June 2004.
[13] "Recommended Weld Quality and Stress Concentration Factors for Use in the Structural Analysis of Exelon Replacement Steam Dryer," GENE Report, e-DRF#0000-0034-6079, February 2005.
31
GE-NE-O000-0038-0936-NP Non-proprietary Version Appendix A Vermont Yankee Steam Dryer Differential Pressure The pressure differentials across the steam dryer are calculated for three categories of events; normal, upset, and faulted conditions. Normal conditions are the steady-state operating conditions. Upset conditions are the anticipated transient events. ((
)) Faulted conditions are the design basis accident events (e.g. main steam line break). The pressure differentials across the steam dryer for the normal conditions at EPU power level are summarized in Table A-1.
Table A-I. Steam Dryer Pressure Differentials for Normal Conditions at EPU I_
11 The pressure differentials across the steam dryer due to forward flow for upset conditions at EPU power level are summarized in Table A-2.
Table A-2. Dryer Delta-P for EPU Forward Flow Upset Conditions 11
.1 The maximum acoustic loads on the dryer face at EPU power level are summarized In Table A-3. Typical pressure time history Isshown In Figure A-1.
32
GE-NE-0000-0038-0936-NP Non-proprietary Version Table A-3. Maximum TSV Acoustic Load on the Dryer Face at EPU I!
I I I I I I I I 11 33
GE-NE-0000-0038-0936-NP Non-proprietary.;Version
[I 11 Figure A-1. Typical TSV Load Time Histories 34
GE-NE-0000-0038-0936-NP
- -Non-proprietary Version Appendix B Acoustic Pressure Stress Intensity Contour Plots
((
35
GE-NE:-0000-0038-0936-NP Non-proprietary Version 11 11 Figure B-1 Inner hood base plate H
1]
Figure B-2a Modified outer cover plate, includes tips 36
GE-NE-0000-0038-0936-NP Non-proprietary Version
[I 11 Figure B-2b Modified outer cover plate, exclude tips 11 1]
Figure B-3a Original top hood (all hood) 37
. GE-NE-0000-0038-0936-NP
- -Non-proprietary Version
((
11 Figure B-3b Modified top hood (outer hood) 11 Figure B-3c Hood top plates (Inner hood) 38
GE-NE-0000-0038-0936-NP Non-proprietary Version 11 1]
Figure B-4a Original outer hood, strips
((
I]
Figure B-4b Modified outer hood, top weld 39
GE-NE-0000-0038-0936-NP Non-proprietary Version 1[
I]
Figure B-4c Modified outer hood, bottom weld I]
Figure B-5 Hood vertical plates (inner hood) 40
GE-NE-0000-0038-0936-NP Non-proprietary Version R
11 Figure B-6 Hood end plates (inner hood) 11 Figure B-7 Hood end plates (outer hood) 41
GE-NE-0000-0038-0936-NP
-^ Non-proprietary Version 1[
11 Figure B-8 Inner hood brackets (gussets)
U 11 Figure B-9 Steam 'dam' 42
GE-NE-0000-0038-0936-NP Non-proprietary Version
[1 11 Figure B-10 Steam 'dam' gussets
[1 11 Figure B-11 Hood partition plates 43
GE-NE-0000-0038-0936-NP Non-proprietary Version
[E I]
Figure B-12 Baffle plates U
1]
Figure B-13 Outlet plenum ends 44
GE-NE-0000-0038-0936-NP Non-proprietary Version 1D I]
Figure B-14 Dryer support ring 11 Figure B-16 Bottom cross beams 45
GE-NE-0000-0038-0936-NP Non-proprietary Version
((
I]
Figure B-16 Cross beam gussets
((
11 Figure B-17a New gusset on cover plate and front hood 46
GE-NE-0000-0038-0936-NP Non-proprietary Version If 11 Figure B-17b Gusset foot weld to cover plate 47
GE-NE-0000-0038-0936-NP Non-proprietary Version 1[
11 Figure B-18 Outer hood, bottom weld 48
GE-NE-0000-0038-0936-NP Non-proprietary Version 1[
]1 Figure B-19 Cover plate 49
Attachment 7 to BVY 05-034 Docket No. 50-271 Attachment 7 Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 26 Extended Power Uprate Steam Dryer Analyses and Monitoring Technical Memorandum No. 05-06, Analysis of Steam Dryer Differential Pressure Loads at Vermont Yankee Total number of pages in Attachment 7 L (excludinc this cover sheet) Is 45. l
C.D.I. Technical Memorandum No. 05-06 Analysis of Steam Dryer Differential Pressure Loads at Vermont Yankee Revision 1 Prepared by Continuum Dynamics, Inc.
34 Lexington Avenue Ewing, NJ 08618 Prepared under PO No. 4500531980 / 07/02/2004 for Entergy Nuclear Vermont Yankee LLC 320 Governor Hunt Road Vernon, VT 05354 Approved by b.J%
ya^ Alan J. Bilanin March 2005
SUMMARY
A previous report [1] detailed the development of the acoustic circuit analysis on the Vermont Yankee (VY) instrument lines. That analysis corrected the venturi instrument line data by a validated acoustic circuit technique [2] that includes all of the branch lines and Rosemount transmitters on the instrument lines, from the main steam lines to the pressure measurement instruments. These corrected data are then coupled with the strain gage data collected on the main steam lines to predict the differential pressure loads anticipated on the steam dryer. The technical approach for the acoustic circuit analysis of the instrument lines, main steam lines, and steam dome is summarized in [3]. Here the application of these techniques to the Vermont Yankee plant is presented.
MODELING CONSIDERA TIONS The VY steam supply system is broken into two distinct analyses: a Helmholtz solution within the steam dome and an acoustic circuit analysis in the main steam lines.
In the Helmholtz analysis the complex three-dimensional geometry of the steam dryer and steam dome are rendered onto a uniformly-spaced rectangular grid (with mesh spacing of 3 inches),
and a solution is obtained for the Helmholtz equation 4 (1)
V2P+ 2P=0 a2 where P is the pressure at a grid point, o is the circular frequency, and a is acoustic speed in steam. This equation is solved for incremental frequencies from 0 to 200 Hz, subject to the boundary conditions dP = 0 (2a) dn normal to all solid surfaces (the steam dome wall and interior and exterior surfaces of the dryer),
dP in (2b)
- cc-OP dn a normal to the nominal water level surface (steam-water interface), and unit pressure applied to one inlet to a main steam line and zero applied to the other three.
The Helmholtz solution is coupled to an acoustic circuit solution in the main steam lines. The acoustic circuit analysis divides the main steam lines into elements, which are each characterized by length L, cross-sectional area A, fluid mean density p, fluid mean flow velocity U, and fluid 2
acoustic speed a. Application of acoustic circuit methodology generates solutions for the fluctuating pressure Pn and velocity un in the nth element of the form Pn = AneikInXn + B.eik2nXn Jict (3a)
In1,[(0+Unki. )AneiklnXn ("+Unk2n)Bneik2nXne (3b) u= t [kin k2 n e (b where harmonic time dependence of the form eI"t has been assumed. The wave numbers kin and k2 n are the two complex roots of the equation kn2 +i 2 (CO+Unkn)--1/2 ('l)+UnknY =0 (4)
Dna a where fn is the pipe friction factor for element n, Dn is the hydrodynamic diameter for element n, and i = I I. An and Bn are complex constants which are a function of frequency, and are determined by satisfying continuity of pressure and mass conservation at element junctions.
The resulting system is driven by shear layer motions at geometric discontinuities. These geometric discontinuities exist in the steam delivery system where convective velocities are high.
Solution is obtained with eight pieces of data collected at eight separate locations, two on each main steam line. In the case of VY, each main steam line has been instrumented with one strain gage, in addition to the pressure reading at the end of the venturi instrument line. These eight pieces of data uniquely determine the four inlet sources at the entrances to the main steam lines.
As may be seen from the above equations, there are eight unknown coefficients (An and Bn on each main steam line) that must be determined before the pressure and velocity are known in the main steam lines. These unknown coefficients are uniquely determined (and unique inlet sources are obtained) when eight independent pieces of data are supplied at each power level. Previously
[4], only six pieces of data (pressure at the four venturi instrument lines and two water reference legs) and an assumption regarding inlet source phasing had to be made to close the problem.
That restriction no longer applies.
INPUTDA TA VY mounted four strain gages on the main steam lines upstream of the SSV/SRVs, four pressure transducers on the main steam lines at the venturi (venturi instrument lines), four upstream of the turbine (turbine instrument lines), one in the steam chest, and one on each of the two water reference legs. These data sets are tabulated in Table 1. Only the four strain gage data and four venturi instrument line pressure data are examined here. The pressure data were taken at the ends of instrument lines whose detailed lengths were verified by VY personnel. Lines were assumed filled with water to the condensing chambers, and the data were corrected for the 3
instrument line effects of line length, line connectivity and layout geometry, acoustic speed, transmitter compliance, and frictional losses along the line, by correcting the data in frequency space and then reconstructing the time signal at the instrument line location on the main steam line, as explained below.
Venturi pressure data for five power settings (80%, 85%, 90%, 95%, and 100% power) were supplied previously by VY [4]. Strain gage data were acquired independently, but contained a common signal (the venturi pressure in the A main steam line). The pressure data were corrected to accommodate the analysis by resampling the data from 2000 samples per second to the strain gage sample rate of 1024 samples per second, and shifting the pressure data in time by a constant delay that matched the common signal. The pressure data were filtered before resampling to minimize aliasing of the high frequency content in the original data. The time delays were determined by analyzing the common signal in the frequency domain to estimate the phase delay based on spectral analysis methods. The synchronized strain gage offset times are also shown in Table 1.
Table 1: Summary of VY Data Sets Data Feed Flow Date/Time Data Rate Power Level Strain Gage Set (106 lbs/hr) (samples/sec) (%) Offset (sec)
-- 6.605 07/09 08:36 2000 100.0 0.889 1 6.605 07/09 08:50 2000 100.0 0.404
-- 6.605 07/09 08:51 2000 100.0 0.170
-- 6.262 07/09 09:59 2000 95.0 0.026 2 6.262 07/09 10:05 2000 95.0 0.750
-- 6.262 07/09 10:06 2000 95.0 0.689
-- 5.857 07/09 10:18 2000 90.0 1.161 3 5.857 07/09 10:24 2000 90.0 0.196
-- 5.857 07/09 10:24 2000 90.0 0.766
-- 5.473 07/09 10:37 2000 85.0 0.960 5 5.473 07/09 10:37 2000 85.0 0.444
-- 5.473 07/09 10:38 2000 85.0 0.045
-- 5.088 07/09 10:39 2000 80.0 0.639 4 5.088 07/09 10:50 2000 80.0 0.462
-- 5.088 07/09 10:51 2000 80.0 0.512 4
These data were smoothly notch filtered to remove signal content (noise) at the recirculation pump vane passage frequency (and its harmonics) and at frequencies corresponding to electrical interference (60, 120, 180 Hz). Second order notch filters were constructed and applied using a forward-backward digital filtering analysis, by processing the input data in both the forward and reverse directions, and resulting in fourth-order attenuation and no phase delay. The resulting time sequence has precisely zero-phase distortion and double the filter order. The filter stopband was I Hz.
Sixteen seconds of data were analyzed at each power setting, centered on the second set of 20 seconds of data supplied, and resulting in 16384 samples with a time increment of 0.0009765625 sec. A Fast Fourier transform [5] was used for both the transform and its inverse. To present a clear picture of the frequency effects, the PSD sampling size was reduced to 1024 samples with a 50% overlap window [5]. The effects of these operations may be seen in the figures presented later in this report.
COMPLIANCE EFFECTS Transfer of the venturi pressure data from the instrument line reading to the main steam lines requires the application of an acoustic circuit model to the instrument line geometry shown schematically in Figure 1. This circuit required an analysis of the response of the Rosemount transmitters, located at the ends of every branch on every instrument line. The transmitters are modeled by a spring-mass-damper system, thereby permitting the development of a pressure-velocity boundary condition that represents the compliance (and damping) effect of the transmitter on the instrument line.
The compliance modeling begins with the equation AP = Mx + cx + kx (5) where A is the cross-sectional area of the transmitter, P is the pressure against the transmitter, M is the apparent mass of the transmitter, c is the damping of the transmitter, k is the effective spring constant of the transmitter, x is the displacement of the transmitter diaphragm, x is the velocity of the transmitter diaphragm, and x is the acceleration of the transmitter diaphragm.
The apparent mass of the transmitter may be defined as M = pfAL + pdAt (6) where pf is the density of the fluid, Pd is the density of the diaphragm, L is the depth of the fluid, and t is the thickness of the diaphragm. It can also be seen that conservation of mass requires rA = uAi (7) where Ai is the cross-sectional area of the inlet pipe to the transmitter. The governing equation is transformed to frequency space with the two substitutions 5
LL p4 L5 L8 L6 4 c d e L7 a b Figure 1. Schematic of typical VY venturi instrument line. The notation is MSL = Main Steam Line, CC = Condensing Chamber, P = Penetration, Gray Box = Whitey Valve Manifold (on lines a, b, d, and e) and Dragon Manifold (on line c), Empty Box =
Rosemount Transmitter (on line c: positioned with empty box - not shown - off L12 ,
with empty box off L13 representing the Sensotec FP2000 pressure measuring instrument), Open Circle = Valve, S = Location where detailed instrument line sketch begins.
6
(8a)
X X = (8b) ici)
The effective spring constant can be represented by the transmitter compliance K KAV A~x which gives A2 (9)
K AP A iP Substituting these expressions into Equation 5 gives the equation P =((pfL+pdt)to~i- +cAi +___) (10)
Dividing both sides of the equation by pa2 gives an equation relating PP pa2 U
u =-
a which is p ((pfL +pdt)io) Ai cA1 A; -
(1 1) pa A paA 2 io)Kpa)
By defining
- Kpa A;
- = pfL+Pdt pa 7
cAj paA' and multiplying through by iwoK, the equation becomes ioJKP = (1+i+oKc o2KM-Ai a (12)
This equation represents the effect of each Rosemount transmitter in the acoustic circuit model of the instrument lines at VY. The physical inputs for the transmitter were supplied by Rosemount from either their catalogue, email, or telephone conversation. The damping parameter was developed from the solution for oscillating flow through a pipe [6].
TRANSFER FUNCTIOAS To test the effects of compliance on the four instrument lines, a sensitivity study was undertaken by varying the amount of compliance in the Rosemount transmitters. In this study the Sensotec compliance was always set at 1 x 107 in3 /psid (its default value), with little change seen between the no compliance case and the no compliance case with the Sensotec compliance turned on.
The results are presented in Figure 2. Here the transfer function magnitude H(co) is defined such that JT(w)j = H(Co)IS(o)I where T(o) is the fast Fourier transform coefficient of pressure recorded at the pressure measurement device and S(o,) is the fast Fourier transform coefficient of pressure at the main steam line. It may be seen that the shape of the transfer functions is a strong function of compliance, as the amplitude changes an order of magnitude from no compliance to 100%
compliance. The transfer function is complex and includes phase as well as magnitude. Based on the sensitivity runs conducted in this report, the dryer loads are primarily influenced by the transfer function amplitude. Therefore the magnitude plots are presented here.
STRAIN GA GE AND CORRECTED PRESSURE DA TA PSDs of the strain gage data (corrected to pressure) are shown in Figure 3 for 100% power and in Figure 4 comparing 100% power with 80% and 90% power. The notch frequencies are accentuated by the 1 Hz bin size Log PSD plots. The filtering used was optimized to eliminate only +1-0.5 Hz at the electrical noise frequencies. This filtering in fact, slightly increases the 0.01 Prms2 /Hz noise floor. Therefore this filtering resulted in negligible reduction of calculated dryer load. The comparison in Figure 3 suggests little difference in driving pressure across the power settings supplied for analysis by VY.
8
For the most part, above 50 Hz, the strain gage data are at or below their noise floor. It appears that the primary loading of the VY dryer will result from loads which have frequency content below 50 Hz.
9
100 I I 'Ii,'
. -- - 5()"' ,.(ozlrli111 .t1AC Full Compliance
.O 10 0 -----
t--------------- --------------------- -- -- --
0v°. 1 -- - - -
0 50 100 150 200 Frequency (Hz)
-- Compliance
No Venturi B i (i CiinIrI1,1, 100 - ' ' ' ' ! : ' 1 j'i,,
50%' CompNlianlce
!.Full Compliance 10 -- ----
¢ ~~~0 .1 -------------------- ----.------------------. ...--/....
001 0 50 100 150 200 Frequency (Hz)
Figure 2a. Transfer function magnitude H(Q)) for venturi A (top) and venturi B (bottom) for several values of Rosemount transmitter compliance.
10
100 rull Compiiance 00 4ij 0 50 100 150 200 Frequency (Hz)
- No Compliance Venturi D 100 I ' ' )( Vi .)I :
IXA(" 0Compliance Full Compliance
- °10 10 -- ----- ,----------j: - -------- V---- --------------------
1 --- -- --------.-.. .- - X-1 10.01 0 50 100 150 200 Frequency (Hz)
Figure 2b. Transfer function magnitude H(co) for venturi C (top) and venturi D (bottom) for several values of Rosemount transmitter compliance.
11
Vermont Yankee MSL A Strain Gage 100% Power 0.1 CA 0.01 0.001 . . . . . . . . . . . . . . . .
. 100 150 200 0 50 Frequency (Hz)
Vermont Yankee MSL B Strain Gage 100% Power 1
N 0.1 0.01 C/)
0.001 0 50 100 150 200 Frequency (Hz)
Figure 3a. PSD of filtered strain gage data on the A main steam line (top) and the B main steam line (bottom) for 100% power.
12
Vermont Yankee MSL C Strain Gage 100% Power 1
/1s N
Z 0.1
.7-V) 0.01 0.001 0 50 100 150 200 Frequency (Hz)
Vermont Yankee MSL D Strain Gage 100% Power 0.1 N
CA C._ 0.01 C)
C'/
P-0.001 L 0 50 100 150 200 Frequency (Hz)
Figure 3b. PSD of filtered strain gage data on the C main steam line (top) and the D main steam line (bottom) for 100% power.
13
Vermont Yankee MSL A Strain Gage 0.1 --
N 0.01
.._4 Mn ti) 0.001 -
0.0001 ) 200 50 100 150 Frequency (Hz)
Vermont Yankee MSL B Strain Gage 1
N 0.1 0.01 Ci) 0.001 0.0001 0 50 100 150 200 Frequency (Hz)
Figure 4a. Comparison of PSD of filtered strain gage data on the A main steam line (top) and the B main steam line (bottom) for 100%, 90%, and 80% power.
14
Vermont Yankee MSL C Strain Gage 1
N 0.1
.W-Cn 0.01 0.001 0.0001 0 50 100 150 200 Frequency (Hz)
Vermont Yankee MSL D Strain Gage 0.1 N
.1
"/ 0.01 0.001 0 50 100 150 200 Frequency (Hz)
Figure 4b. Comparison of PSD of filtered strain gage data on the C main Steam line (top) and the D main steam line (bottom) for 100%, 90%, and 80% power.
15
PSDs of the corrected venturi pressure data through the instrument lines are shown in Figure 5 for 100% power and in Figure 6 comparing 100% power with 80% and 90% power. The strong similarities between signals across the supplied power settings are clear.
The difference between the venturi pressures and the strain gage equivalent pressures is in the fact that the venturi pressures dip around 50 Hz (for the B and C lines) and 100 Hz (for the A and D lines), and increase with frequency thereafter, whereas the strain gage pressures do not. This behavior suggests that the instrument line correction is adding acoustic energy at higher frequencies, resulting in a conservative signal transferred to the main steam line.
It may be postulated that the pressure recorded at the end of the instrument line is always interpreted as originating from an acoustic wave in the main steam line. In fact, 3D effects such as main steam line turbulence would propagate an acoustic signal down the instrument line that, in this analysis, would be interpreted as acoustic, when in fact it is noise. Therefore, the pressure time histories recorded at the ends of the instrument lines contain both acoustic and turbulent noise. At 100% power, main steam line flow velocities are U = 144 ft/sec, and the largest eddies scale with the main steam line diameter D = 1.34 ft. Therefore, flow noise is anticipated at frequencies above U/D = 100 Hz.
LOW RESOLUTION GRID MESH RESULTS These data were used to drive an acoustic circuit model of the main steam lines and steam dome, to construct low resolution grid mesh results on the steam dryer geometry illustrated in Figures 7 and 8. The summary results of this effort are shown in Figure 9. Very little difference may be seen in maximum steam dryer loads between power levels, consistent with similar behavior observed in Dresden 2 (DR2). Figure 10 compares the PSD at the peak load point on the low resolution grid (node numbers 98 on the A-B side and 6 on the C-D side) for 100%, 90%, and 80% power. The consistency across power level is maintained in the load predictions as well.
HIGH RESOLUTION GRID MESH RESULTS A high resolution load result at 100% power was supplied to GE to support the finite element analysis of the VY dryer. This result included time histories at 11,118 nodes located on the dryer, separated by the grid mesh increments illustrated in Figure 8. The loads were delivered separately to GE.
The selected time interval (10.0 to 12.5 seconds) within the 16 seconds examined represents the overall signal content of the entire 16 seconds predicted. This conclusion can be reached by a detailed examination of the peak load prediction on the low resolution grid (node numbers 98 on the A-B side and 6 on the C-D side). The RMS and peak pressure values for the selected 2.5 seconds are 0.0767 and 0.3220 psid, respectively, for the A-B side, and 0.0845 and 0.3010 psid, respectively, for the C-D side, while the RMS and peak pressure values for the entire 16 second time interval are 0.0734 and 0.3365 psid, respectively, for the A-B side, and 0.0794 and 0.3086 psid, respectively, for the C-D side. The time history slice is plotted in Figure 11.
16
Vermont Yankee MSL A Venturi 100% Power 10 1
N 0.1 "i 0.01 CA 0.001 - - - - - - - -
vo 0.0001 10-5 10-6 0 50 100 150 200 N Frequency (Hz)
P- Vermont Yankee MSL B Venturi 100% Power 10 1 --- ------------------------------------ -------------------------------------------
7- - - - - - - - - - - - - - - - - - - - - -
0.1 - - - - - - - - - - - - --- - - - - - - - - - - - - - - - - - -
0.01 - - - - - - - - - - - - -
0.001 0.0001 10-5 0 50 100 150 200 Frequency (Hz)
Figure 5a. PSD of instrument line corrected venturi pressure data on the A main steam line (top) and the B main steam line (bottom) for 100% power.
17
Vermnont Yankee MSL C Venturi 100% Power 10 1
N 0.1 ti 0.01 0.001
.....V . . .- ---------
0.0001 - -- -- -- , - -- -- - -- -- --- - - -- - - - - - - -- - - - - --
10-5 10-6 0 50 100 150 200 Frequency (Hz) rd) Vermont Yankee MSL D Venturi 100% Power 10 1
0.1 0.01 0.001 0.0001 - - - - - - -- - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
10-5 104 10-6 1 7 0 50 100 150 200 Frequency (Hz)
Figure 5b. PSD of instrument line corrected venturi pressure data on the C main steam line (top) and the D main steam line (bottom) for 100% power.
18
Vermont Yankee MSL A Venturi 10 - r I I I I r r I I r , I I I r -
1 - ------------- 100% Power -----------
N I tt ~90% Powser .-
0.1 p.4 0.01 Ut 0.001 ---------------------------
0.0001 - -- --- ---- ------ -- - - - - - - - - - - -- - --
10 10-6 0 50 100 150 200 Frequency (Hz)
Vermont Yankee MSL B Venturi 10 _ I I 1 T r T r j ! r , r N
0.1 - ----------- - - - - - - --- - --,--
_ I,
--- -- -- - -- T . .. . .. . .. . ... .
._ 0.01 0.001 Q ...... .. 7.. -- - -
u/) 0.0001 _...... ..... ....
-- _ 100% Power 10 -- ----------- 90% Power .. _
... 801P\ ,1, we A-6 0 50 100 150 200 Frequency (Hz)
Figure 6a. Comparison of PSD of instrument line corrected venturi pressure data on the A main steam line (top) and the B main steam line (bottom) for 100%, 90%, and 80% power.
19
Vermont Yankee MSL C Venturi 10 ,100% Power
--h---------- 900/0, pomme- -------------
0.01
- [l- . , -- - - - - -
6 ;.;
10-5 .~;~;~~
.~~~;~~~~~~~;~ .~;~~
0 50 100 150 200 Frequency (Hz)
Vermont Yankee MSL D Venturi 1 - -- - - - -- - - - ---------- -----------
_ 0.0001 Cot -- -- -- -- - -- - -- - - - -- -- -- -- -- - - - ------- owe 100% Power.
106 ...........-......... ..... 90% Pow e
...... -o--- 800 Ioxvel o\e P0) _
10-8 0 50 100 150 200 Frequency (Hz)
Figure 6b. Comparison of PSD of instrument line corrected venturi pressure data on the C main steam line (top) and the D main steam line (bottom) for 100%, 90%, and 80% power.
20
TOP VIEW:
2,3 Pressure data MSL B 900 MSL A SIDE VIEW: Pressure data locations E D C B A 3 105 27009 JEJ 900 J MSLC&D MSLA&B Nominal Water Level
( 3G-(E----------& E~----------(------3 ---- (gD SL 2 104 Figure 7. Top and side view schematic of pressure node locations on the VY steam dryer.
21
LZ i393 6 98 30 82 Figure 8a. Bottom plates pressure node locations on the VY dryer, with pressures acting downward in the notation defined here. The high resolution grid mesh is 3 inches on the cover plates, 6 inches on the first bottom plates, and 12 inches on the rest of the bottom plates.
22
47r-7 ---- "F---7 I 74 10 1628 3446 5 5 773 8891 102 Figure 8b. Upper plates pressure node locations on the VY dryer, with pressures acting downward in the notation defined here. The high resolution grid mesh is 3 inches on the outer top plates, 6 inches on the first inner top plates, and 12 inches on the rest of the inner top plates.
23
Figure 8c. Vertical plates on the VY dryer: Pressures acting left to right on panels 6-11, 22-29, and 40-47; acting right to left on panels 64-71, 82-89, and 98-103. The high resolution grid mesh is 3 inches on the outer bank hoods, 6 inches on the first inside hoods, and 12 inches on the inside hoods.
24
86 103
.') 2(
'~_ 21 81 41 10
- 937 S.-.-7
~
I,L~,
-97 A
E 36 6 1 L2ŽiILntI LI2 IL SJ Figure 8d. Skirt plates on the VY dryer: Pressure acting on the outer dryer 0/180 surfaces and the skirt. The high-resolution grid mesh is 3 inches on the outer portion of the skirt closest to the main steam lines, 6 inches on the sections nearer the center of the dryer, and 12 inches on the center of the dryer.
25
0.35 I T I I I - - 1 I I
- 1. 0.3 - -80% - - ------------
Power 0 -,T . 1.- -l t-0)
. _, 0.25
. , 9- (3 ) flow'cr
._ 0.2 I 00% Power C/1 Ut _ _. -------- r - ----
- -- @ _-------------*- ----- ---- ------------ ~~~~
0.15 a)t 1-Iq a) 0.1 rE 0.05 -I - _
0 0 20 40 60 80 100 120 Node Numbers 0.08 ,T - , ,, ,, , , , , I ,-, ,
7:
0.07 80% Power I - _
\ ---- - - _--
5-4 0.06 --- --- , -
0.05 95 .'0Power- .
0)
En I 00% Power 0.04 co uO 0.03 - - - - - - - - - - - - - - - - - - - - -.- -. . . . . . . . . ...............
Aq 0.02 - I - -- -- -- -- -- -- -- -- -- -.--------------- - - - ---
0.01 - - --- ; - - - - - -- - - - - - ------ ----
n v
0 20 40 60 80 100 120 Node Numbers Figure 9. Low resolution grid mesh results for VY for 80% power (black curve), 85% (red curve), 90% power (gray curve), 95% power (green curve), and 100% power (blue curve).
26
Vermont Yankee Peak Load A-B Side of Dryer 0.001 o.000 1 --------- -------------------------------------- I--------\-- /A -Sri I10-6 . ..........
10 -- I -- -l-l
.------ -- - -- - -- -- - - - -- - --- I---------
100% Power 10- 90% Power 10-8 I l I l l l l l I l I _
0 50 100 150 200 Frequency (Hz)
Vermont Yankee Peak Load C-D Side of Dryer 0.001 - ' ' ' ' ! '
N 10- 5
~ ~ f- SX V~oll 11 I 10 -6 . .. . . . .. . . - -- - - . .. . . .
100% Power M-- 90% Power 80O1/~ Power 10.8 0 50 100 150 200 Frequency (Hz)
Figure 10. Comparison of PSD at the low resolution dryer locations where peak load is predicted: X on the A-B side (top) and Y on the C-D side (bottom).
27
Vermont Yankee Peak Load A-B Side of Dryer AA 032 "C,
a.)
0.1~
0 ah uf I~LIIL 1-4
-0.1
-0.2
-0.3 111.11-TIA I It. 011Sli' 6S8N1 2 I- - - - - -
4tl
I - - - - .-.-..
-0.4 I'.
.I . . . I
. l 0 2 4 6 .8 lo 12 14 16 Time (sec)
Vermont Yankee Peak Load C-D Side of Dryer A
I. It Il I g r, j r ,
0.3.
11 ,, I. I, 1 1 1-.7 0.2 LA,- :Ij 'i li-: In-r-r I I;--1rI-l i [I- 4,111- iI-i "Ci 1-_4 0.1 0
CA, -0.1I CD,
$.)
-0.3
-0 1--. -V f 2 ][I ~Hi Y-~-p
~ip ri'>' I -y 1 :I1 IYj2 .1 . pL +II.' l' fl A I I I I
-U.14 0 2 4 6 8 10 12 14 16 Time (sec)
Figure 11. Time slice (in blue) selected for high resolution grid mesh results, compared with the entire time history prediction at 100% power (in black): peak load on the A-B side of the dryer (top); peak load on the C-D side of the dryer (bottom), both based on the low resolution grid mesh results.
28
The high resolution grid mesh results add additional load definition only approximated by the low resolution grid mesh results, and finds a more accurate prediction of the maximum load amplitude across the VY dryer. For example, the maximum differential pressure load predicted from the low resolution loads is 0.336 psid, while for the high resolution loads it is 0.648 psid.
COMPARISON WVITHDRESDENA2AND QUAD CITIES 2 Recently, strain gage data were collected at DR2, and low and high resolution load definitions were supplied to Exelon [7]. These strain gage data at OLTP are compared with VY strain gage data at 100% power in Figure 12. The feed flow rate for VY is 6.605 Mlbs/hr (from Table 1),
while the feed flow rate for DR2 at OLTP is 9.71 Mlbs/hr. It may be seen that the DR2 strain gage data decrease in magnitude with frequency increase, while the VY data do not. Clearly, the VY noise floor on the strain gages is higher than the DR2 noise floor on their strain gages, effectively increasing the higher frequency energy content of the pressure signal supplied to the steam dryer. Structural Integrity Associates estimates the noise floor at 0.008 psid2 /Hz [8].
The low-resolution grid mesh results for VY may be compared with similar nodal locations on DR2 as shown in Figure 13. Here the nodes associated with the center bank and the top skirt have been removed from the low resolution results previously obtained for DR2 [7], reducing the total number of nodes on these dryers to 104, the same as on VY (and at the same relative locations on the dryer - the VY dryer is smaller than the DR2 dryer, however). It may be seen from Figure 13 that VY at 100% power has maximum pressure loads that are 0.730 of the predicted DR2 load at OLTP, on the C-D side of the dryer (0.3086 psid on VY and 0.4228 psid on DR2), and 0.650 of the predicted DR2 load at OLTP, on the A-B side of the dryer (0.3365 psid on VY and 0.5179 psid on DR2). The average load reduction factor is 0.636 across the 104 nodes in the low resolution result, suggesting that VY dryer loads are reduced considerably from those predicted for the DR2 dryer.
The low-resolution grid mesh results for VY may also be compared with similar nodal locations on Quad Cities 2 (QC2) as shown in Figure 14. Again, the feed flow rate for VY is 6.605 Mlbs/hr, while the feed flow rate for QC2 at OLTP is 9.92 Mlbs/hr. Itinay be seen that VY at 100% power has a maximum pressure load of 0.3086 psid on the C-D-side of the dryer, as opposed to 2.1298 psid in QC2, and 0.3365 psid on the A-B side of the dryer, as compared to 2.6294 psid in QC2. A comparison of peak loads on the dryer is shown in Figure 15.
Peak load behavior is quite different, reflecting the added energy content present in QC2. This content is the result of distinct deterministic mechanisms that exist at the feed flow rates at which the QC2 plant operates. These mechanisms increase the dryer loads, and are simply not excited at the much lower feed flow rates at which the VY plant operates; hence, the difference in frequency content. The QC2 load at 92 and 138 Hz is not present in VY.
29
Main Steam Line A Strain Gage Comparison 0.1 ' I ' ! I 'I 0.01 a0.001ol lVY __ YA vi \, i\
- -DR2A upper 0.0001 , I I I ,
0 50 100 150 200 Frequency (Hz)
Main Steam Line B Strain Gage Comparison 1 , I , r I , I I I, , ,
VY B
DR2 B upper _ -
4 0.0 1 ---------------
v 0.001 o~~oo 1 .... ............ ........ ,.. ... .... .... .n.. . , . . -----
C- '
0.0001 ,1 1 I 0 50 100 150 200 Frequency (Hz)
Figure 12a. PSD comparisons of strain gages on VY (black curve), DR2 at their upper locations (blue curve, closer to the steam dome), and DR2 at their lower locations (red curve, farther from the steam dome): main steam line A (top); main steam line B (bottom).
30
Main Steam Line C Strain Gage Comparison 1
1-1-11 N 0.1 tO "O
. '..4 rn 0.01 15 C)
Cn 0.001 P-4 0.0001 0 50 100 150 200 Frequency (Hz)
Main Steam Line D Strain Gage Comparison 1
N 0.1 0.01 p".4 0.001 0.0001 0 50 100 150 200 Frequency (Hz)
Figure 12b. PSD comparisons of strain gages on VY (black curve), DR2 at their upper locations (blue curve, closer to the steam dome), and DR2 at their lower locations (red curve, farther from the steam dome): main steam line C (top); main steam line D (bottom).
31
0.8 0.7 TI------r--
T T T rI r 1 00% Pow er we r
-- r-~~~~
--------r---------------r---
. _ 0.6 .. DR2: EPU -
.-7: 4 0.5
- - i 0
Ue $-4 s3 0.4 W) 0.3 0.2 -------- -------- -------- -------
0.1 ;-
0 0 20 40 60 80 100 120 Node Number 0.14 0.12 _ 1l 100% Power 1
ll--------.------ ... R :EU-----
2-------------
0.1 *DR2: EPU --
v :.
- _4 0.08 CA -I- ---r--------------r-- ----------- r--------------r--------- --- ----'-
0.06 0.04 0.02 - !1S ,,AI n _ . .. . . . . . .. . .- -L - - ,
0 20 40 60 80 100 120 Node Number Figure 13. Low resolution grid mesh results for VY at 100% power (black curve), compared with DR2 at OLTP (red curve), and DR2 at EPU (blue curve), from [7].
32
4
'-4
- -4 3.5 ------ - 100% Power 3 --- I .-------------- *QC2:QC--EPU-.--- EPU
.'4 2.5 a) 2 1l / IJ 1;--------
1.5 a)
'- -~~ ~~~~. . . . .. ... .------ 1,-!----
x 1 0.5 0
0 j 0I 20
-1.i I /
40 60 80 I
100 120 a) Node Number 1.6 1.4 -----r ----------- 1 00% Power
_ .- , - I I ,
1.2 - -- - -- -- - -C
- -- l----- -P- -
.'4 1
- Q:: ::
0.8 0.6 -- - - - - - - - - - - - - - .... . . ...........
t: jX ;(wi --
0.4 0.2 0
0 20 40 60 80 100 1:20 Node Number Figure 14. Low resolution grid mesh results for VY at 100% power (black curve), compared with QC2 at OLTP (red curve) and QC2 at EPU (blue curve), from [7].
33
Peak Load Comparison A-B Side of Dryer 1
0.1 VY
.-N 0.01 0.001 - --------- \ - -
0.0001 10-5 10-6 10-7 N 10-8 0 50 100 150 21DO Frequency (Hz) r,)
Peak Load Comparison C-D Side of Dryer 1
0.1
-QC2) 0.01 0.001 0.0001
~ Y'\4-'-'F ------
10o 10-6 10-7 0-8 0 50 100 150 21)O Frequency (Hz)
Figure 15. Comparison of peak loads for VY at 100% power (black curves) and peak loads for QC2 OLTP (blue curves): A-B side of dryer (top); C-D side of dryer (bottom).
34
SUMAIAR Y BEHA VIOR WITH POoVER SETTING An indication of the effect of power level on the steam dryer may be seen in Figure 16, where the maximum differential pressure and RMS pressure are plotted as a function of power level. The effect of power level appears as an increase in pressure (although a small one), with 100% power giving the largest loads (these curves are consistent with Figure 9). Previously, in [4], it was thought that 90% power gave the highest steam dryer loads, but with the corrected venturi pressure time histories developed here, this result no longer holds. The reason for this may be traced to the conservatism built into the venturi instrument line analysis. The instrument line analysis is more accurate than what was developed previously [4] but includes conservatism that tends to increase dryer loads. This effect appears to be increasing as power level increases (Figure 16), although the actual increase in maximum differential pressure is small (0.075 psid on the A-B side of the dryer and 0.035 psid on the C-D side of the dryer from 90% to 100%
power level).
Finally, VY results are compared against DR2 and QC2 results in Figure 17. The lower loads anticipated on VY are clearly seen (DR2 and VY analyses include strain gage data; QC2 results do not). This figure summarizes the behavior of the three steam dryers examined, and illustrates that there is little load difference expected for VY and DR2 as a function of feed flow rate (also percentage power in VY), but significant load difference expected for QC2.
35
Vermont Yankee 0.34 1--4 ai) 0.32 .. .. l A-B Side-- -- ------------------ ---- -
.-a 0 -* - C-D Side U
-/-
0 0.3 "0 \- - - - ---- -- - --------------
0.28 ci) -- - --- - - - , - -- - - - -- - - - - -- - - - - ...--.--- ----..- ;--.- --
ci) 0.26 0.24 80 85 90 95 100 Percent Power Vermont Yankee 0.062
/I--,
. .. I 0.06 ---..... ° A -B Side ----- ---- ---- ----- ---- --------- ---- ----
(*I 0 -C -D Sid e./
15 +C4 Q> gi.,4 0.058 _ *-,-- --- -*-- - - *-
$--4 0
0.056 .. . . . . . . . .-- . - - - -- - -- -- --- -- -- --
0 U
0.054 C'n 00 0.052 A AC 80 85 90 95 100 Percent Power Figure 16. Summary comparison of maximum dryer pressure loads at VY as a function of power level.
36
4 , I, I I, I,, I I I I I a.)
ct 3.5 --------- ° V Y: A-B Side -----------------
VY: C-D Side a.) "C~ 3 _ -- -- DR2:-A- DR2: A-B Sd. Side ---------- --- -- L~ --------
1 2.5 _------ -- -A DR2: C-D Side - ----------- -------
a)-4 M ->-- QC2: A-B Side C-2 _ .- Q C 2: C-D -'QC2: D Side ----
Side-. ---- ----------- ------------
1-4 U2 1.5 e
1 - ---- --- ---- --- ----- - - -- - . ... .. . . . . . . . . --- --- ------
cU 0.5 _ - - - - - ,-. - - - - -
0 5 6 7 8 9 10 11 12 Feed Flow (Mlbs/hr) 1.6 *I I I F I 1.4 ---..... -° VY: A-B Side . ---
VY: C-D Side 1.2 -- DR2: A-B Side 1W-4 a) 1 _ -------. A, DR2: C-D Side . - ..
-V QC2: A-B Side--
cn 0.8 _ -- QC2: C-D Side IEa) 0.6 0.4 _ ---.-.--.-.
. ----- . -- ------- - -. - - - - - - -. - - - -. - - - - - - - - - I-.............
- - - - -I.-. - - - - - - -
0.2 0
5 6 7 8 9 10 11 12 Feed Flow (Mlbs/hr)
Figure 17. Comparison of maximum differential pressure loads and RMS for VY, DR2, and QC2 to 200 Hz.
37
QUALITY ASSURANCE The work documented in this report meets requirements of Continuum Dynamics, Inc.'s Nuclear Quality Assurance Program.
38
REFERENCES
- 1. Continuum Dynamics, Inc. 2005. Acoustic Circuit Analysis of Vermont Yankee Instrument Lines (Rev. 0). C.D.I. Technical Memorandum No. 05-05.
- 2. Continuum Dynamics, Inc. 2004. Test Report for Validating an Instrument Line Acoustic Transmission Model (Rev. 0). C.D.I. Report No. 04-12.
- 3. Continuum Dynamics, Inc. 2005. Methodology to Determine Unsteady Pressure Loading on Components in Reactor Steam Domes (Rev. 6). C.D.I. Report No. 04-09 (Proprietary).
- 4. Continuum Dynamics, Inc. 2005. Hydrodynamic Loads on Vermont Yankee Steam Dryer (Rev. 0). C.D.I. Report No. 04-13.
- 5. Press, W. H., S. A. Teukolsky, W. T. Vettering, and B. P. Flannery. 1992. Numerical Recipes in FORTRAN: The Art of Scientific Computing. Second Edition. Cambridge University Press. 498-502, 542-55 1.
- 6. Schlichting, H. 1968. Boundary-Layer Theory. Sixth Edition. McGraw-Hill. 419-421.
- 7. Continuum Dynamics, Inc. 2005. Revised Hydrodynamic Loads on Quad Cities Unit 2 Steam Dryer to 200 Hz, with Comparison to Dresden Unit 2 and Dresden Unit 3 Loads (Rev.
0). C.D.I. Report No. 05-01.
- 8. Structural Integrity Associates, Inc. 2005. Report No. SIR-05-045 (Rev. A).
39
APPENDIX: VY HIGHRESOLUTION GRID MESH Model predictions are collected into ASCII data files containing the time history data for the predictions of differential pressure (psid) across various locations on the plates in the dryer. A cross-sectional schematic of the dryer geometry is shown in Figure 7. Perspective views of the steam dryer panels are provided in Figure 8.
For horizontal plates, pressure differences are computed by subtracting the pressure below the plate from the pressure above the plate. For vertical plates, pressure differences are computed by subtracting the pressure to the right of the plate from the pressure to the left of the plate, for the vertical plates on the MSL C and D side, and in the opposite direction on the MSL A and B side (the direction designated "left" faces MSL C and D; "right" faces MSL A and B). For skirt plates, pressure differences are computed by subtracting the pressure inside the skirt from the pressure outside the skirt. Pressure differences are provided at all plate intersections, and along the steam dryer centerline (the 900 and 270° directions).
Each output file contains 2.5 seconds of data with a time difference of 0.0009765625 seconds to extract results to 200 Hz. There are 105 columns of data:
- 1. Time (seconds)
- 2. Pressure difference across the skirt at the steam dryer centerline below MSL C and D at the nominal water level (Figure 8d)
- 3. Pressure difference at the skirt and cover plate intersection at the steam dryer centerline below MSL C and D (Figures 8a and 8d)
- 4. Pressure difference across the skirt on the MSL B-C side below the E outer bank hood at the nominal water level (Figure 8d)
- 5. Pressure difference across the skirt on the MSL A-D side below the E outer bank hood at the nominal water level (Figure 8d)
- 6. Pressure difference across the skirt and the E outer bank hood on the MSL B-C side at the cover plate intersection (Figures 8a, 8c, and 8d)
- 7. Pressure difference across the E outer bank hood at the steam dryer centerline at the cover plate intersection (Figures 8a and 8c)
- 8. Pressure difference across the skirt and the E outer bank hood on the MSL A-D side at the cover plate intersection (Figures 8a, 8c, and 8d)
- 9. Pressure difference across the skirt and the E outer bank hood on the MSL B-C side at the top plate intersection (Figures 8b, 8c, and 8d)
- 10. Pressure difference across the E outer bank hood at the steam dryer centerline at the top plate intersection (Figures 8b and 8c)
- 11. Pressure difference across the skirt and the E outer bank hood on the MSL A-D side at the top plate intersection (Figures 8b, 8c, and 8d)
- 12. Pressure difference across the E bottom plate open edge at the skirt on the MSL B-C side (Figure 8a)
- 13. Pressure difference across the E bottom plate open edge at the steam dryer centerline (Figure 8a)
- 14. Pressure difference across the E bottom plate open edge at the skirt on the MSL A-D side (Figure 8a) 40
- 15. Pressure difference across the E top plate corner edge at the skirt on the MSL B-C side (Figure 8b)
- 16. Pressure difference across the E top plate corner edge at the steam dryer centerline (Figure 8b)
- 17. Pressure difference across the E top plate corner edge at the skirt on the MSL A-D side (Figure 8b)
- 18. Pressure difference across the skirt on the MSL B-C side below the D inner bank hood at the nominal water level (Figure 8d)
- 19. Pressure difference across the skirt on the MSL A-D side below the D inner bank hood at the nominal water level (Figure 8d)
- 20. Pressure difference across the skirt on the MSL B-C side at the D inner bank hood intersection with the E bottom plate (Figure 8d)
- 21. Pressure difference across the skirt on the MSL A-D side at the D inner bank hood intersection with the E bottom plate (Figure 8d)
- 22. Pressure difference across the D inner bank hood on the MSL B-C side at the E bottom plate intersection (Figures 8a and 8c)
- 23. Pressure difference across the D inner bank hood at the steam dryer centerline at the E bottom plate intersection (Figures 8a and 8c)
- 24. Pressure difference across the D inner bank hood on the MSL A-D side at the E bottom plate intersection (Figures 8a and 8c)
- 25. Pressure difference across the skirt on the MSL B-C side at the D inner bank hood intersection with the D top plate (Figure 8d)
- 26. Pressure difference across the skirt on the MSL A-D side at the D inner bank hood intersection with the D top plate (Figure 8d)
- 27. Pressure difference across the D inner bank hood on the MSL B-C side at the D top plate intersection (Figures 8b and 8c)
- 28. Pressure difference across the D inner bank hood at the steam dryer centerline at the D top plate intersection (Figures 8b and 8c)
- 29. Pressure difference across the D inner bank hood on the MSL A-D side at the D top plate intersection (Figures 8b and 8c)
- 30. Pressure difference across the D bottom plate open edge at the skirt on the MSL B-C side (Figure 8a)
- 31. Pressure difference across the D bottom plate open edge at the steam dryer centerline (Figure 8a)
- 32. Pressure difference across the D bottom plate open edge at the skirt on the MSL A-D side (Figure 8a)
- 33. Pressure difference across the D top plate open edge at the skirt on the MSL B-C side (Figure 8b)
- 34. Pressure difference across the D top plate open edge at the steam dryer centerline (Figure 8b)
- 35. Pressure difference across the D top plate open edge at the skirt on the MSL A-D side (Figure 8b)
- 36. Pressure difference across the skirt on the MSL B-C side below the C inner bank hood on the MSL C/D side at the nominal water level (Figure 8d)
- 37. Pressure difference across the skirt on the MSL A-D side below the C inner bank hood on the MSL C/D side at the nominal water level (Figure 8d) 41
- 38. Pressure difference across the skirt on the MSL B-C side at the C inner bank hood intersection with the D bottom plate (Figure 8d)
- 39. Pressure difference across the skirt on the MSL A-D side at the C inner bank hood intersection with the D bottom plate (Figure 8d)
- 40. Pressure difference across the C inner bank hood on the MSL B-C side at the D bottom plate intersection (Figures 8a and 8c)
- 41. Pressure difference across the C inner bank hood at the steam dryer centerline at the D bottom plate intersection (Figures 8a and 8c)
- 42. Pressure difference across the C inner bank hood on the MSL A-D side at the D bottom plate intersection (Figures 8a and 8c)
- 43. Pressure difference across the skirt on the MSL B-C side at the C inner bank hood intersection with the C top plate on the MSL C-D side (Figure 8d)
- 44. Pressure difference across the skirt on the MSL A-D side at the C inner bank hood intersection with the C top plate on the MSL C-D side (Figure 8d)
- 45. Pressure difference across the C inner bank hood on the MSL B-C side at the C top plate intersection on the MSL C-D side (Figures 8b and 8c)
- 46. Pressure difference across the C inner bank hood at the steam dryer centerline at the C top plate intersection on the MSL C-D side (Figures 8b and 8c)
- 47. Pressure difference across the C inner bank hood on the MSL A-D side at the C top plate intersection on the MSL C-D side (Figures 8b and 8c)
- 48. Pressure difference across the C bottom plate open edge at the skirt on the MSL B-C side on the MSL C-D side (Figure 8a)
- 49. Pressure difference across the C bottom plate open edge at the steam dryer centerline on the MSL C-D side (Figure 8a)
- 50. Pressure difference across the C bottom plate open edge at the skirt on the MSL A-D side on the MSL C-D side (Figure 8a)
- 51. Pressure difference across the C top plate open edge at the skirt on the MSL B-C side on the MSL C-D side (Figure 8b)
- 52. Pressure difference across the C top plate open edge at the steam dryer centerline on the MSL C-D side (Figure 8b)
- 53. Pressure difference across the C top plate open edge at the skirt on the MSL A-D side on the MSL C-D side (Figure 8b)
- 54. Pressure difference across the C top plate open edge at the skirt on the MSL B-C side on the MSL A-B side (Figure 8b)
- 55. Pressure difference across the C top plate open edge at the steam dryer centerline on the MSL A-B side (Figure 8b)
- 56. Pressure difference across the C top plate open edge at the skirt on the MSL A-D side on the MSL A-B side (Figure 8b)
- 57. Pressure difference across the C bottom plate open edge at the skirt on the MSL B-C side on the MSL A-B side (Figure 8a)
- 58. Pressure difference across the C bottom plate open edge at the steam dryer centerline on the MSL A-B side (Figure 8a)
- 59. Pressure difference across the C bottom plate open edge at the skirt on the MSL A-D side on the MSL A-B side (Figure 8a)
- 60. Pressure difference across the skirt on the MSL B-C side below the C inner bank hood on the MSL A-B side at the nominal water level (Figure 8d) 42
- 61. Pressure difference across the skirt on the MSL A-D side below the C inner bank hood on the MSL A-B side at the nominal water level (Figure 8d)
- 62. Pressure difference across the skirt on the MSL B-C side at the C inner bank hood intersection with the B bottom plate (Figure 8d)
- 63. Pressure difference across the skirt on the MSL A-D side at the C inner bank hood intersection with the B bottom plate (Figure 8d)
- 64. Pressure difference across the C inner bank hood on the MSL B-C side at the B bottom plate intersection (Figures 8a and 8c)
- 65. Pressure difference across the C inner bank hood at the steam dryer centerline at the B bottom plate intersection (Figures 8a and 8c)
- 66. Pressure difference across the C inner bank hood on the MSL A-D side at the B bottom plate intersection (Figures 8a and 8c)
- 67. Pressure difference across the skirt on the MSL B-C side at the C inner bank hood intersection with the C top plate on the MSL A-B side (Figure 8d)
- 68. Pressure difference across the skirt on the MSL A-D side at the C inner bank hood intersection with the C top plate on the MSL A-B side (Figure 8d)
- 69. Pressure difference across the C inner bank hood on the MSL B-C side at the C top plate intersection on the MSL A-B side (Figures 8b and 8c)
- 70. Pressure difference across the C inner bank hood at the steam dryer centerline at the C top plate intersection on the MSL A-B side (Figures 8b and 8c)
- 71. Pressure difference across the C inner bank hood on the MSL A-D side at the C top plate intersection on the MSL A-B side (Figures 8b and 8c)
- 72. Pressure difference across the B top plate open edge at the skirt on the MSL B-C side (Figure 8b)
- 73. Pressure difference across the B top plate open edge at the steam dryer centerline (Figure 8b)
- 74. Pressure difference across the B top plate open edge at the skirt on the MSL A-D side (Figure 8b)
- 75. Pressure difference across the B bottom plate open edge at the skirt on the MSL B-C side (Figure 8a)
- 76. Pressure difference across the B bottom plate open edge at the steam dryer centerline (Figure 8a)
- 77. Pressure difference across the B bottom plate open edge at the skirt on the MSL A-D side (Figure 8a)
- 78. Pressure difference across the skirt on the MSL B-C side below the B inner bank hood at the nominal water level (Figure 8d)
- 79. Pressure difference across the skirt on the MSL A-D side below the B inner bank hood at the nominal water level (Figure 8d)
- 80. Pressure difference across the skirt on the MSL B-C side at the B inner bank hood intersection with the A bottom plate (Figure 8d)
- 81. Pressure difference across the skirt on the MSL A-D side at the B inner bank hood intersection with the A bottom plate (Figure 8d)
- 82. Pressure difference across the B inner bank hood on the MSL B-C side at the A bottom plate intersection (Figures 8a and 8c)
- 83. Pressure difference across the B inner bank hood at the steam dryer centerline at the A bottom plate intersection (Figures 8a and 8c) 43
- 84. Pressure difference across the B inner bank hood on the MSL A-D side at the A bottom plate intersection (Figures 8a and 8c)
- 85. Pressure difference across the skirt on the MSL B-C side at the B inner bank hood intersection with the B top plate (Figure 8d)
- 86. Pressure difference across the skirt on the MSL A-D side at the B inner bank hood intersection with the B top plate (Figure 8d)
- 87. Pressure difference across the B inner bank hood on the MSL B-C side at the B top plate intersection (Figures 8b and 8c)
- 88. Pressure difference across the B inner bank hood at the steam dryer centerline at the B top plate intersection (Figures 8b and 8c)
- 89. Pressure difference across the B inner bank hood on the MSL A-D side at the B top plate intersection (Figures 8b and 8c)
- 90. Pressure difference across the A top plate corner edge at the skirt on the MSL B-C side (Figure 8b)
- 91. Pressure difference across the A top plate corner edge at the steam dryer centerline (Figure 8b)
- 92. Pressure difference across the A top plate corner edge at the skirt on the MSL A-D side (Figure 8b)
- 93. Pressure difference across the A bottom plate open edge at the skirt on the MSL B-C side (Figure 8a)
- 94. Pressure difference across the A bottom plate open edge at the steam dryer centerline (Figure 8a)
- 95. Pressure difference across the A bottom plate open edge at the skirt on the MSL A-D side (Figure 8a)
- 96. Pressure difference across the skirt on the MSL B-C side below the A outer bank hood at the nominal water level (Figure 8d)
- 97. Pressure difference across the skirt on the MSL A-D side below the A outer bank hood at the nominal water level (Figure 8d)
- 98. Pressure difference across the skirt and the A outer bank hood on the MSL B-C side at the cover plate intersection (Figures 8a, 8c, and 8d)
- 99. Pressure difference across the A outer bank hood at the steam dryer centerline at the cover plate intersection (Figures 8a and 8c) 100. Pressure difference across the skirt and the A outer bank hood on the MSL A-D side at the cover plate intersection (Figures 8a, 8c, and 8d) 101. Pressure difference across the skirt and the A outer bank hood on the MSL B-C side at the top plate intersection (Figures 8b, 8c, and 8d) 102. Pressure difference across the A outer bank hood at the steam dryer centerline at the top plate intersection (Figures 8b and 8c) 103. Pressure difference across the skirt and the A outer bank hood on the MSL A-D side at the top plate intersection (Figures 8b, 8c, and 8d) 104. Pressure difference across the skirt at the steam dryer centerline below MSL A and B at the nominal water level (Figure 8d) 105. Pressure difference at the skirt and cover plate intersection at the steam dryer centerline below MSL A and B (Figures 8a and 8d) 44
The grid used for the Helmholtz solver for pressure is a rectangular grid with spacing of 3 inches in all three directions. The grid coordinate system is x and y horizontal (x points toward 900 and y points toward 00) and zero at the center of the dryer, with z out of the paper in Figure 7 and zero at the steam-water interface. The 104 low-resolution nodes are shown on these figures for reference. The bottom center of the steam dome (at the steam-water interface) has a coordinate of n, = 37, ny = 37, and nz = 1. The cover plate is at the level height of nz = 26, while the outer hood top plate is at the level height of nz = 47.
The transmitted data files contain a title to each column of data. The first column in each file is time (in seconds), designated by the title "0", and numerous additional columns titled with a number. For every pressure difference provided, its location is designated by multiplying its x grid location (nt) by 10000, adding its y grid location (ny) multiplied by 100, and adding its z grid location (ni). Thus, for example, the heading 713726 corresponds to nx = 71, ny = 37, and n,
= 26, and is the position of node 105 in the low-resolution grid. Node 2 is 33702.
The transmitted data files consist of four sets: AB (for the A-B main steam line side of the steam dome), CD (for the C-D main steam line side of the steam dome), and ML and EX (for the middle of the dryer).
45