Westinghouse Presentation Template Class 3
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End-to-End Benchmarking of an Acoustic Model for Analysis of BWR Steam Dryers David A. Suddaby Westinghouse, Principal Engineer
Karen K. Fujikawa Westinghouse, BWR Stress Analysis Manager
John C. Rommel Exelon, Power Uprate Director
1
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Overview • • • • •
Background Analysis Methods Benchmarking Results Conclusions
2
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background • Initial Event: BWR steam dryer cracking after implementation of extended power uprate • Root Cause: Acoustic pressure loads caused by resonances at the safety relief valves in the main steam lines (MSL) • Solution: Acoustic/structural analyses for qualification of existing or replacement steam dryer (RSD)
• Presentation Focus: Benchmarking of the acoustic/structural model 3
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background – Peach Bottom Atomic Power Station • Exelon and Westinghouse worked together to design, install, and qualify a new Westinghouse RSD at Peach Bottom Units 2 and 3
• The Westinghouse RSDs were critical component replacements for the planned extended power uprates (EPU) from 3514 MWt to 3951 MWt, yielding an increase in power of 12.4%
• The latest Westinghouse acoustic/structural model was benchmarked at Unit 2 as part of the power ascension to EPU
4
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background – Acoustic Pressure Loading on RSDs
Reactor Pressure Vessel Steam Dryer Acoustic Pressure
Steam Flow Main Steam Line
safety relief valves
5
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Analysis/Qualification • Instrumentation and measurement • Acoustic model of the MSL and RPV (above water)
• Structural finite element model (FEM) of the RSD
Westinghouse RSD FEM 6
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Instrumentation • Strain gauges • Pressure transducers • Accelerometers Hood
• On the MSLs, upstream of
Skirt
the SRVs • On the dryer Hood and Skirt surfaces
7
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Acoustic Model • Input: MSL acoustic pressure measurements • Combination of 3D and 1D acoustic modeling • Includes – MSL from the SRVs to the RPV – RPV steam dome above water level – Steam dryer
• Output: Acoustic pressure loads on the surface of the RSD (the forcing function for the structural FEM) 8
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Structural Analysis • The objective is to show that the maximum alternating • • • •
stress intensity anywhere in the dryer is less than the material endurance strength at 1011 cycles Analysis uses an endurance limit of 13.6 ksi FEM consisting mostly of shell elements Approximately 190,000 nodes and elements with varying mesh densities Harmonic analysis (~0-250 Hz)
Skirt Above Water
Skirt Below Water
Slots Belts
9
Drain Channel Above Water
Drain Channel Below Water
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Traditional Benchmarking for RSD Qualification MSL Measured Pressure Acoustic Model Pressure Structural Analysis
Alternating Stress
10
RSD Measured Pressure
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End-to-End Benchmarking for RSD Qualification MSL Measured Pressure
RSD Measured Strain
Acoustic Model Pressure Structural Analysis
Alternating Stress
Directional Strain The end-to-end benchmark ensures an accurate and conservative acoustic model and FEM 11
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Measured Data • The measured MSL data is used as input to process • The measured RSD data is used as the final comparison of the modeling accuracy and conservatism • Strain gauge arrays on all 4 MSLs • RSD was instrumented with ~25 strain gauges (both upper and lower dryer) ~15 pressure transducers ~5 accelerometers
Extensive instrumentation was used to benchmark the acoustic/structural models 12
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Acoustic Modeling • 3D model of the RSD and surrounding steam • Solution of the 3D wave propagation equation in the frequency domain, i.e. Helmholtz equation • There were comparisons made of the measured and predicted pressure on the RSD surface • The predictions were generally very conservative, but only used for reference as the key comparison was the final strain from the FEM Intermediate comparisons of predicted and measured pressure showed conservatism in the acoustic model 13
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Bias and Uncertainty • The true end-to-end comparison was of the predicted and measured strain on the RSD surfaces
• The overall model Biases and Uncertainty (B/U) were calculated based on predicted and measured strain
• Bias: over or under prediction • Uncertainty: how well the model predicts strain (error) The calculated B/U are combined into a total uncertainty and added back into the acoustic model 14
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Strain Prediction • The final strain prediction includes the calculated B/U • Validation by comparing measured and predicted strain – Over the entire frequency range
and – At specific key frequencies
Predicted Ɛ > Measured Ɛ at key frequencies
Strain PSD (Representative Example) 15
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – The Iterative Loop • The acoustic model contains many adjustable (heuristic) parameters • The FEM went through an initial “tuning” process, but remains fixed after that point • If the B/Us are unreasonable, i.e. the predicted strain is inaccurate or non-conservative in certain frequency ranges – The acoustic modeling parameters can be adjusted to help match measured data – Acoustic damping is most often the adjusted parameter 16
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – The Iterative Loop cont. • The key frequencies (where predicted strain must be greater than measured strain) are determined based on the results of the FEM stress analysis • The FEM is a part of the process that predicts the strain • If the benchmark attempt fails the key frequency criteria, changes can be made to the acoustic model • This can alter the key frequencies for the strain comparison Hundreds of benchmarking iterations were required to balance all criteria 17
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Results • Spring 2015: PBAPS Unit 2 completed power ascension to EPU • The Unit 2 EPU data was successfully used to benchmark the latest Westinghouse acoustic/structural methodology
• Acoustic Circuit Enhanced (ACE) Revision 3.1 – Meets all benchmarking criteria set forth by the US NRC
• ACE Rev 3.1 was successfully used to qualify the Unit 3 RSD at predicted EPU conditions
• Winter 2015: PBAPS Unit 3 completed power ascension to EPU 18
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Conclusions • End-to-end benchmarking – Include measurement at beginning and end of acoustic path – Compare measured vs predicted stress on the component of interest – Intermediate comparison of pressure for reference only – Ensures the entire analytical process is conservative
• The benchmarking process, including the acceptance criteria, will heavily involve the regulator – Technically challenging – Can be a lengthy process 19
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End
20
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Acronyms • • • • • • • • • •
ACE B/U EPU FEM MASR MSL PBAPS RSD RPV SRV
Acoustic Circuit Enhanced bias and uncertainty extended power uprate finite element model minimum alternating stress ratio main steam line Peach Bottom Atomic Power Station replacement steam dryer reactor pressure vessel safety relief valve
21
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End-to-End Benchmarking of an Acoustic Model for Analysis of BWR Steam Dryers David A. Suddaby Westinghouse, Principal Engineer
Karen K. Fujikawa Westinghouse, BWR Stress Analysis Manager
John C. Rommel Exelon, Power Uprate Director
1
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Overview • • • • •
Background Analysis Methods Benchmarking Results Conclusions
2
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background • Initial Event: BWR steam dryer cracking after implementation of extended power uprate • Root Cause: Acoustic pressure loads caused by resonances at the safety relief valves in the main steam lines (MSL) • Solution: Acoustic/structural analyses for qualification of existing or replacement steam dryer (RSD)
• Presentation Focus: Benchmarking of the acoustic/structural model 3
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background – Peach Bottom Atomic Power Station • Exelon and Westinghouse worked together to design, install, and qualify a new Westinghouse RSD at Peach Bottom Units 2 and 3
• The Westinghouse RSDs were critical component replacements for the planned extended power uprates (EPU) from 3514 MWt to 3951 MWt, yielding an increase in power of 12.4%
• The latest Westinghouse acoustic/structural model was benchmarked at Unit 2 as part of the power ascension to EPU
4
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Background – Acoustic Pressure Loading on RSDs
Reactor Pressure Vessel Steam Dryer Acoustic Pressure
Steam Flow Main Steam Line
safety relief valves
5
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Analysis/Qualification • Instrumentation and measurement • Acoustic model of the MSL and RPV (above water)
• Structural finite element model (FEM) of the RSD
Westinghouse RSD FEM 6
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Instrumentation • Strain gauges • Pressure transducers • Accelerometers Hood
• On the MSLs, upstream of
Skirt
the SRVs • On the dryer Hood and Skirt surfaces
7
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Acoustic Model • Input: MSL acoustic pressure measurements • Combination of 3D and 1D acoustic modeling • Includes – MSL from the SRVs to the RPV – RPV steam dome above water level – Steam dryer
• Output: Acoustic pressure loads on the surface of the RSD (the forcing function for the structural FEM) 8
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Structural Analysis • The objective is to show that the maximum alternating • • • •
stress intensity anywhere in the dryer is less than the material endurance strength at 1011 cycles Analysis uses an endurance limit of 13.6 ksi FEM consisting mostly of shell elements Approximately 190,000 nodes and elements with varying mesh densities Harmonic analysis (~0-250 Hz)
Skirt Above Water
Skirt Below Water
Slots Belts
9
Drain Channel Above Water
Drain Channel Below Water
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Traditional Benchmarking for RSD Qualification MSL Measured Pressure Acoustic Model Pressure Structural Analysis
Alternating Stress
10
RSD Measured Pressure
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End-to-End Benchmarking for RSD Qualification MSL Measured Pressure
RSD Measured Strain
Acoustic Model Pressure Structural Analysis
Alternating Stress
Directional Strain The end-to-end benchmark ensures an accurate and conservative acoustic model and FEM 11
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Measured Data • The measured MSL data is used as input to process • The measured RSD data is used as the final comparison of the modeling accuracy and conservatism • Strain gauge arrays on all 4 MSLs • RSD was instrumented with ~25 strain gauges (both upper and lower dryer) ~15 pressure transducers ~5 accelerometers
Extensive instrumentation was used to benchmark the acoustic/structural models 12
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Acoustic Modeling • 3D model of the RSD and surrounding steam • Solution of the 3D wave propagation equation in the frequency domain, i.e. Helmholtz equation • There were comparisons made of the measured and predicted pressure on the RSD surface • The predictions were generally very conservative, but only used for reference as the key comparison was the final strain from the FEM Intermediate comparisons of predicted and measured pressure showed conservatism in the acoustic model 13
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Bias and Uncertainty • The true end-to-end comparison was of the predicted and measured strain on the RSD surfaces
• The overall model Biases and Uncertainty (B/U) were calculated based on predicted and measured strain
• Bias: over or under prediction • Uncertainty: how well the model predicts strain (error) The calculated B/U are combined into a total uncertainty and added back into the acoustic model 14
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – Strain Prediction • The final strain prediction includes the calculated B/U • Validation by comparing measured and predicted strain – Over the entire frequency range
and – At specific key frequencies
Predicted Ɛ > Measured Ɛ at key frequencies
Strain PSD (Representative Example) 15
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – The Iterative Loop • The acoustic model contains many adjustable (heuristic) parameters • The FEM went through an initial “tuning” process, but remains fixed after that point • If the B/Us are unreasonable, i.e. the predicted strain is inaccurate or non-conservative in certain frequency ranges – The acoustic modeling parameters can be adjusted to help match measured data – Acoustic damping is most often the adjusted parameter 16
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Benchmarking – The Iterative Loop cont. • The key frequencies (where predicted strain must be greater than measured strain) are determined based on the results of the FEM stress analysis • The FEM is a part of the process that predicts the strain • If the benchmark attempt fails the key frequency criteria, changes can be made to the acoustic model • This can alter the key frequencies for the strain comparison Hundreds of benchmarking iterations were required to balance all criteria 17
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Results • Spring 2015: PBAPS Unit 2 completed power ascension to EPU • The Unit 2 EPU data was successfully used to benchmark the latest Westinghouse acoustic/structural methodology
• Acoustic Circuit Enhanced (ACE) Revision 3.1 – Meets all benchmarking criteria set forth by the US NRC
• ACE Rev 3.1 was successfully used to qualify the Unit 3 RSD at predicted EPU conditions
• Winter 2015: PBAPS Unit 3 completed power ascension to EPU 18
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Conclusions • End-to-end benchmarking – Include measurement at beginning and end of acoustic path – Compare measured vs predicted stress on the component of interest – Intermediate comparison of pressure for reference only – Ensures the entire analytical process is conservative
• The benchmarking process, including the acceptance criteria, will heavily involve the regulator – Technically challenging – Can be a lengthy process 19
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
End
20
Westinghouse Non-Proprietary Class 3
© 2016 Westinghouse Electric Company LLC. All Rights Reserved.
Acronyms • • • • • • • • • •
ACE B/U EPU FEM MASR MSL PBAPS RSD RPV SRV
Acoustic Circuit Enhanced bias and uncertainty extended power uprate finite element model minimum alternating stress ratio main steam line Peach Bottom Atomic Power Station replacement steam dryer reactor pressure vessel safety relief valve
21
