ML072480518
| ML072480518 | |
| Person / Time | |
|---|---|
| Site: | Hope Creek  |
| Issue date: | 08/17/2007 |
| From: | Public Service Enterprise Group |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| Download: ML072480518 (30) | |
Text
ATTACHMENT 2 Hope Creek Generating Station Facility Operating License No. NPF-57 NRC Docket No. 50-354 Extended -Power Uprate Non-Proprietary Version of August 17, 2007 Presentation
ifiI introductions Objectives PSEG View of Acceptance Criteria Discussion of Key Points.
Bias & Uncertainty Acoustic Circuit Model (ACM) Rev 4 Methodology Loading Definition from ACM Finite Element Model (FEM)
Ell Limit Curves Final Reports RAI Responses
,,.H2peCreek.; 2
Demonstrate closure with May 2007 audit concerns as summarized by recent RAIs RAI matrix is provided at the end of the presentation Indicates ifa report was generated to address RAI response Broken down by category (FEM, ACM, etc)
Summarize the revised reports
,*lIndicates key reasons for resubmitting Solicit NRC feedback/observations HopeCreek 3
-,-- - i
Acoustic Circuit Model Biases and Uncertainties
@Frequency-dependent based on QC2 data i Applied to loads from in-plant CLTP strain gage data Finite Element Model o Frequency shift +/-10% to establish peak stresses Acceptable Margin Sufficient to accommodate expected increase in loads proportional to flow 2 Qualitative Assessment of Acoustic Loads h eCreek 4
The B&U for the load consists of:
LoStrain Gage B&U Acoustic Circuit Model Rev 4 B&U Based on QC2 data for dryer hood sensors P1 -P21 Calculated for discrete frequency ranges as defined by RAI 14.67 Which States "Frequency ranges are:((
I]
Averaged 6 pressure sensors in each frequency range Expected Standpipe Resonance for HC is at 118 Hz
.:HoeCreek5
. t ,6.i:-**
A' O M,l'!*
J
- Biases are added algebraically.
Uncertainties are combined by SRSS.
Summarized in CDI Report 07-09P &CDI Technical Note 07-29P Tables 2, 3 &4 FEM Load shifted in 2.5% intervals between -/+10% to account for frequency uncertainty HopeCreek6
I I I
((
Negative.bias is ACM over prediction; Positive bias is ACM under prediction Total includes strain gage bias and uncertainty NopeCreek 7
11 Summarized in CDI Report 07-09P
..HopeCreek 8
Input: 2007 in-plant data from all 4 MSLs (RAI 14.109)
No reliance on previous algorithm to justify data from only 2 MSLs No reliance on Scale Model Test (SMT)
The calculated B&U is added to the measured load before it is applied to the FEM
[ 1))
Summarized in CDI Report 07-18P 4.HopeCreek,: 9
FEM - ANSYS 10 in the Harmonic Domain Constant 1%structural damping across frequency range (RAI 14.78)
Addresses concerns on transient time at low frequencies 64 seconds of in-plant data are used vs 2 seconds (RAI 14.96)
Harmonic approach provides much more efficient means of producing accurate data Based on 2007 in-plant CLTP Loads
- HopeCreek*
10
Validation is provided in Appendix B of CDI Report 07-09 (RAI 14.110)
Compares results of the harmonic domain vs the transient in the time domain Stress histories are virtually identical in both Amplitude &
Phase Comparison is done using the Browns Ferry Unit 1 Dryer FEM Two Specific Base Frequencies were Compared HopeCreek:
1]
HopeCreek 12
Locations (All are Welds) %Frequency Alternating Shift Stress Ratio Outer vane bank/perforated inlet plate -7.5% 1.86 Middle base plate / middle vane bank -7.5% 1.89 Inlet Perforated Plate / vane bank top vertical -7.5% 2.11 plate Middle base plate / middle vane bank -7.5% 2.40 Outer base plate / outer vane bank -7.5% 2.50 All alternating stress ratios at nomi al frequency arO > 4 HopeCreek 13
Based on 2007 in-plant data No reliance on Scale Model Testing Based on Limiting Frequency shifts in 2.5% intervals between -4+10%
One limit curve for each of the 8 MSL strain gage locations Summarized in Technical Note No.07-29P hopeCreek" 14.
NNNNNNNEWWENNý . . ---
Minimum SR-a at CLTP = 1.86 mNo further reduction required due to loading B&U Level 2 is 80% of allowable (10,880 psi)
Level 1 is 100% of allowable (13,600 psi)
Load increase for 15% EPU is factor of 1.32 Projected stress ratio at EPU is 1.86 / 1.32 = 1.40 Load increase for 11.5% TPU is factor of 1.24
- Projected stress ratio at EPU is 1.86 / 1.24 = 1.50
- Target power in next operating cycle HopeCreek# 15
Load Methodology for ACM R4 CDI Report 07-09P Revision 1 Adds "low frequency hydrodynamic contribution" to increase fidelity at lower frequencies Provides bias and uncertainty information over specified frequency ranges (RAI 14.67)
Load Definition CDI Report 07-18P Based solely on 2007 in-plant data and from all 4 MSLs Addresses RAIs on SMT (14.70, 80, 88, 89, 95,105 and 109)
Figure 4.7 shows the ((R] (RAI 14.73)
HopeCreekl, 16
Finite Element Analysis CDI Report 07-17P Revised FEM based on CLTP 2007 in-plant data Does not rely on scale model tests, Eliminates reliance on using only 2 MS line data (RAI 14.85)
Includes frequency variations from -10% to +10%
Resolves issues with Rayleigh damping anchors (RAI 14.78)
I]
Adds explanation for model simplification (RAI 14.76)
Discusses Mesh Studies (RAI 14.79)
Uses 64 seconds of in-plant data (not 2 seconds) (RAI 14.96)
SeCreek:?
17
Limit Curves CDI Technical Note 07-29P Based solely on in-plant data Based on most conservative frequency shift Provides a limit curve for each of 8 locations Addresses RAI's 14.86, 14.104, and 14.109 HopeCreek 18
ACM, RAI 14.75 - Address problems that are prevalent in two- sensors acoustic measurement RAI's on benchmarking of SMT, etc Responses provided to specific questions, but revised submittal only uses SMT for QC and HC comparisons:
Add a 118 Hz peak to the HC predicted PSD curves at EPU for HopeCreekcomparison with the QC OLTP PSDs (RAI 14.107) 19
1]
,bopeq *.2
FEM, RAI 14.108, Use of higher factor at the root of a fillet weld WRC Bulletin 432 provides the basis for using a weld factor of 1.8 at the root of a fillet weld. The membrane plus-bending stresses and the thermal stresses are normally higher on the surface then at the embedded location and the free surface has less constraint than an embedded location. Low constraint allows the shear plane to be worked, distorted, and separations to be formed.
HCGS FEM shows that the bending stresses are much larger then the membrane stresses at all the limiting steam dryer locations.
H*oeCreek2
.. . Si A*J. %.
- 21
FEM, RAI 14.77, different thickness plates
%Provided two simple models Provided component thicknesses at highest stressed locations HopeCreek, 22 I
RAI's 14.74 SRV monitoring PSEG is in the process of developing a finite element model of the SRV's which will allow calculating vibration limits for theses valves.
Hop Creekl, 23
Vane Passing Frequency Concern on SSES Steam Dryer, RAI 14.83 Responded with added information about the SSES failure and the subsequent lessons learned applied at HCGS.
Recirculation pump speed range is not changing with EPU, the Steam dryer would not be exposed to any new vane passing frequencies. The Steam Dryer Inspections performed to date and those that will be performed following EPU implementation are adequate to address the concern on loads imposed by Vane Passing Frequency.
HopeCreek 24
- 14. Category Abridged RAI Abridged response 67 ACM Bias and uncertainty calculation does not New ACM report and new limit curves provide this follow VY method. Calculate bias and information. Load definition includes the revised bias uncertainty at frequency bands; and uncertainty.
73 ACM 11 I]
Load Report Figure 4.7 provides the ((
.]
75 ACM Concern on ACM definition at only two Provided as response to the RAI measurement locations; 84 ý ACM 11 1]
85 ACM Demonstrate 2006 Ioads/FEM is conservative New load definition/FEM based on 2007 data.
with regard to 2007 data; 96 ACM Use of 2-second interval for load definition; New calculations are over an order of magnitude longer.
<HopeCreek, 25
70 SMT Questions on limit curves and stresses based Provided limit curves based on CLTP data and on SMT; 80
. .. . ....... . . .......i .... . . ................. .
88 SMT Explain inconsistencies in CDI Report 07- New FEM and limit curves rely only on in-plant 01 between SMT and plant data; data 89 SMT Explain inconsistencies in CDI Report 07- New FEM and limit curves rely only on in-plant 01 between SMT and plant data; data 95 SMT Explain what appears to be errors in the Revised reports not relying on SMT data.
report; Provided as response to the RAI 105 SMT Questions on old limit curves based on SMT New limit curves provided based solely on in-plant data; data.
109 SMT Reliance on SMT for FEM; New FEM report based solely on in-plant data HopeCreek 26
66 FEM Accuracy of FEM; Provided as response to the RAI 76 FEM Explain FEM modeling simplifications; Discussed in new FEM report 77 FEM Impact of different thickness plates on Provided as response to the RAI stresses 78 FEM Justify selected Rayleigh damping "hi" New FEM report uses 1% thru the entire range anchor point below 150 Hz; S ... .........................
79 FEM Discuss mesh studies; Discussed in new FEM report 86 FEMW Do not treat freq shift as uncertainty; New limit curves generated that do not SRSS the Limit frequency with the load uncertainty curves 97 FEM Adjustment on weld factor for complex New FEM does not rely on this adjustment middle hood to end plates configuration 104 Limit Base limit curves on limiting frequency New limit curves provided are based on limiting Curves and one per MSL s/g location; frequency and a curve is provided for each MSL location.
106 TPU Commit to provide NRC with full FEM for Commitment made as response to the RAI FEM TPU; 108 FEM Justify fillet weld stress factor Provided as response to the RAI 110 FEM Validate freq versus harmonic based New FEM report appendix "B" FEM; HopeCreek 27 I
74 SRV's Discuss the SRV monitoring Provided as response to the RAI monitoring guidelines 82 SRV Qualitatively compare HC Provided as response to the RAI Cantilevered SRV to QC failed relief valve components 83 Recirc Pump Concern on VPF impact to Provided as response to the RAI VPF SSES failure 103 PATP Provide proposed licensing Provided as response to the RAI conditions; 107 MSL strain Provide MSL PSD Provided as response to the RAI gage comparisons to QC OLTP readings
-HopeCreek 28
2007 CLTP In-Plant Data Acoustic Circuit Model With Biases and Uncertainties Finite Element Model Limit Curves Acceptable Margin Revised reports May 2007 RAIs were answered Hopecreek 29