slides for 1/29/2013 u.s. epr public meeting with areva to … · 2013. 1. 29. · public meeting...
TRANSCRIPT
Public Meeting to discuss U.S. EPR Postulated Piping Rupture Blast Effects Evaluation Methodology and Design Approach and the Closure Plan for the U.S. EPR FSAR Open Items on Blast Effects
Rockville, Maryland January 29, 2013
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 2
AgendaTopic Presenter
Introductions AREVA/NRC
Purpose R. Wells
Background R. Wells
Discussion of the AREVA NP blast effects methodology and approach for RAI 354
C. McGaughy
• Discussion of the application of loads and frequencies)
• Description of changes to ANP-10318P
Closure Plan for RAI 354, Question 3.6.2-35 R. Wells• Path Forward Activities and Schedule
Summary/Next Steps R. Wells
Comments and Questions
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 3
Purpose
Present and obtain NRC feedback on the AREVA NP blast effects methodology and approach for RAI 354, Question 3.6.2-35
Outline path forwardReview closure plan for U.S. EPR RAI 354, Question
3.6.2-35Confirm agreement with the NRC on the path to closure
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 4
Background RAI 354, Question 3.6.2-35 (Blast Effect Methodology)
3/16/2010 - Initially issued by NRC 5/20/10 – Initial Draft Response of RAI 354, Question 3.6.2-35 sent to NRC 11/1/2010 – Public Meeting with NRC 1/5/2011 - Revised Draft Response of RAI 354, Question 3.6.2-35 sent to NRC 1/26/2011 – Comments received from NRC on Revised Draft Response of RAI 354, Question
3.6.2-35 2/28/2011 – Second revised Draft Response of RAI 354, Question 3.6.2-35 and Revision 0 of
AREVA NP Technical Report ANP-10318P, “Pipe Rupture External Loading Effects on U.S. EPR Essential Structures, Systems, and Components,” submitted to NRC
4/20/2011 – NRC/AREVA telecon to discuss NRC comments on second revised Draft Response of RAI 354, Question 3.6.2-35 and Revision 0 of ANP-10318P
8/9/2011 - NRC/AREVA telecon to discuss path forward on RAI 354, Question 3.6.2-35 8/30/2011 – Interim Response to RAI 354, Question 3.6.2-35 sent to NRC along with proposed
changes to ANP-10318P 10/17/2011 – Follow-up NRC/AREVA telecon to regarding RAI 354, Question 3.6.2-35 (AREVA
informed NRC we plan to use a simplified program to evaluate the blast effects for the Main Steam Valve isolation room and validate these results using CFD).
6/5/2012 – RAI 354 Suppl. 35 defers final response to Question 3.6.2-35 to 7/30/13 11/9/2012 - NRC/AREVA telecon on the status of the of the final Response to RAI No. 354,
Question 3.6.2-35 (AREVA informed NRC that we no longer planned to do the simplified program, rather we were just going to the CFD)
1/29/13 - Public Meeting/Telecon on Blast Effect Methodology
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 5
Background SER Open Items related to Blast Effect Methodology
RAI Question No.
Date of Final RAI Response
NRC Classification
Audit Topic
354 3.6.2-35 7/30/13 Open Item
NRC requested that AREVA clarify how the plan to be used that will account for the different fluid properties in an U.S. EPR blast (steam and/or water) compared to those of air considered in the Army/Navy/Air Force manual.
Discussion of the AREVA NP Blast Effects Methodology and Approach for RAI 354
Chris McGaughyAdvisory Engineer
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 7
Blast Effect MethodologyIntroduction
Goals:Develop a methodology to determine blast loadings
caused by ruptures in steam linesValidate the Computational Fluid Dynamics (CFD)
methods based on current literature
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 8
Blast Effect MethodologyVariability Sources
Selection of discretization in space Grid quality (grid angles, aspect ratios, etc.) Grid design (structured, unstructured, hybrid mesh) Grid size (error versus computational time)
Selection of discretization in time Time step size (error versus computational time)
Selection of convergence criteria Courant number (CFL criterion) Deviations in numerical approach Reasonable values for target variables (especially in
validations)
Input parameter uncertainty Sensitivity on variations of boundary conditions or fluid
properties
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 9
Blast Effect MethodologyModeling the Physics
Compressibility Density model: ideal gas law or real gas law
High-speed flow Viscous model: inviscid or turbulent flow
Flow discontinuities (shock wave) Solver formulation (coupled pressure-based or density-based) Numerical scheme: implicit or explicit Time discretization: implicit or explicit Dynamic mesh adaption (mesh refinement e.g. by pressure
gradient)
For validation, combustion and/or detonation Reaction model: one-step, Finite-Rate
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 10
Blast Effect MethodologyStrategy to Approach the Physics
As a first approach, simple models are chosen: Ideal gas law to model density of the gas mixture. Inviscid flow (inertial forces dominant in high speed flow)
Comparison to experimental data will show the need to change them
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 11
Blast Effect MethodologyV & V Cases
Literature review to identify cases for verification and validation (V&V)
Adequate cases for V&V are few in the open literature Examples for V&V to show the CFD code ability to
determine blast effects: Flow in a shock tube
• 1-D behavior (exact analytical solution available) 28-Inch shock tube
• 3-D behavior (more difficulty in achieving comparable results) Hydrogen combustion in confined vessel
• Similar to shock tube with multiple obstructions Blast wave on a solid box (detonation of propane)
• External wave propagation
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 12
Blast Effect Methodology V&V Case 1: Flow in a Shock Tube
The shock tube is a device in which a normal shock wave is produced by the sudden bursting of a diaphragm separating a gas at high pressure from one at lower pressure.
The problem tests the CFD code ability to handle rarefaction waves, contact discontinuities and shocks.
The basis for this study is the one-dimensional analytical solution of a shock tube.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 13
Blast Effect MethodologyV&V Case 1: Flow in a Shock Tube
Pressure Ratio = 200
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5
Dimensionless-X
Mac
h N
umbe
r
Fluent
Analytical
Expansion wave front
Expansion wave tail
Contact surface
Normal shock
CFD
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 14
Blast Effect MethodologyV&V Case 2: 28-Inch Shock Tube
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 15
Blast Effect MethodologyV&V Case 2: 28-Inch Shock Tube
The sudden bursting of a diaphragm separating the driver gas at high pressure (328 psi) from one at lower pressure (14.5 psi) produces a blast wave.
The problem tests the CFD code ability to calculate the blast wave in a large three dimensional geometry.
Measurements of the pressure as function of time in the shock tube are available.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 16
Blast Effect MethodologyV&V Case 3: H2 Combustion in Confined Vessel
Homogeneous concentration of hydrogen
Flame direction: upwards 9 obstacles to generate
turbulence in the acceleration tube section
Pressure monitoring at the end of the tube
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 17
Blast Effect MethodologyV&V Case 3: H2 Combustion in Confined Vessel
The hydrogen combustion in the confined ENACCEF vessel generates hot gases behind the flame which accelerates upwards. The pressure and temperature rise rapidly in the vessel.
The overpressure history as function of time is monitored and compared to the experimental data.
H2-combustion in the ENACCEF vessel is calculated with CFD to show the ability of the code to handle reactive flow and to predict the time-dependent pressure loads and pressure oscillations in the vessel.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 18
Blast Effect MethodologyV&V Case 3: H2 Combustion in Confined Vessel
Maximum pressure comparable to measured one or higher (conservative)
First and second rate of pressure rise (dp/dt) well-predicted Pressure oscillations well-predicted CFD Model considers adiabatic walls (heat losses unknown in the
experiments)
Decrease due to heat losses in experiments
ENACCEF RUN153 H2 Conc. 13%
0
1
2
3
4
5
6
0 0.1 0.2 0.3 0.4 0.5time (s)
pres
sure
(ba
r)
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 19
Blast Effect MethodologyV & V Case 4: Blast wave on a solid box
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 20
Blast Effect MethodologyV & V Case 4: Blast wave on a solid box
A three dimensional blast wave is generated by the explosion of propane
Pressure load on the solid wall is measuredGeometry not confined (free atmosphere)Calculation requires the modeling of combustion
reaction of propane
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 21
Blast Effect MethodologyConclusions for the shown V&V Cases
Flow in the tube shock CFD results agree well with the exact analytical solution
28-Inch shock tube Positive pressure section well-predicted Negative pressure section more difficult to match accurately
Hydrogen combustion in ENACCEF facility Predicted pressure loads in the vessel are reasonably captured
by the CFD model AREVA NP has long experience in modeling hydrogen
combustionBlast wave on a solid box
Detonation of propane is required (reaction model)
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 22
Blast Effect Methodology Rupture of steam lines
Pipe contains only pressurized steam Pipe rupture occurs at once, cross-section of pipe is open Only steam discharges (no phase change) Room is filled with air and there is a wall close to the break
Wall
Break
Pipe
P, x=1
air
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 23
Blast Effect Methodology Rupture of steam lines (Comparison to
method used in the report)CFD calculation of blast pressure loading for
following example: Description in Technical report ANP-10318P, page 34 Pipe geometry, target structure at 3 ft from break, P=1250psi
and x=1 Calculation focuses on the blast wave that follows a pipe
failure.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 24
Blast Effect Methodology Rupture of steam lines (Comparison to
method used in the report)
Pipe conditions in CFD: Total pressure: 86.164 bar (1250 psi) Total Temperature: 264.91°C
P peak ~ 36 bar (522 psi), comparable to 540 psi as calc. in report
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 25
Discussion of the Application of Loads and Frequencies
Application of the Load Plate and Shell Models (Pressure Vessels and Walls) Overpressure applied to surfaces as a pulse with 0 rise time
and calculated duration (td) Beam Models (Structures and Components)
Overpressure applied as a distributed load along the length of the beam as a pulse with 0 rise time and calculated duration (td). Peak load calculated using peak overpressure multiplied by surface area.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 26
Discussion of the Application of Loads and Frequencies
Structural Analysis - Cutoff Frequency For a triangular pulse load with 0 rise time, the dynamic load
factor (DLF) is greatest when the pulse duration (td) is long, relative to the structure’s period.
With (td) typically being in the 1 ms range, the structure must have natural frequencies greater than 400 Hz for the DLF to be greater than 1 (Biggs - td/T=0.4). However, the DLF is applied to individual modes which, in high frequencies, usually have low effective masses and shapes that are not conducive to a distributed pressure loading.
Therefore, the cutoff frequency will be chosen such that 95% of the effective mass is accounted for. Remaining mass will be considered through the missing mass method.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 27
Discussion of the Application of Loads and Frequencies
Structural Analysis Damping
Although blast forces often cause nonlinear material behavior, damping is conservatively set to 1%, as for jet impingement.
LoadingLoading will be applied as a force time history using direct integration or modal integration, depending on geometric nonlinearities.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 28
Discussion of Acceptance Criteria
Acceptance Criteria Using NRC approved code acceptance criteria for structural
evaluation of safety related target SSCPressure Vessels – ASME CodeSteel Structures – AISC N690 CodeConcrete Structures – ACI 349 Code
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 29
Description of Changes to ANP-10318P
Jet Impingement partial Intersection (Section 3.2.3) -Incorporated the calculation of jet area for use in determining partial intersection areas. Also included discussion of pressure distribution in the jet.
Compressible Jet Properties (Section 3.2.4) - Added new Section 3.2.4.3 describing pulse loading for jet impingement
Retitled Section 3.2.5 as Jet Dynamic Loading and Resonance
Clarified the maximum nozzle pressure ratio for jet resonance in 3.2.5 and 3.2.5.1
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 30
Description of Changes to ANP-10318P
Blast Effects (Section 3.3) – Discussion updated to include CFD approach.
Structural Evaluation (Section 3.4) - Structural Evaluation of Typical Target Essential SSC (Section 3.4) – Added basic information concerning loading methodology.
Dynamic Finite Element Analysis (Section 3.4.1) – Added details
Loading Method (Section 3.4.1.3) – Separated into two subsections to define the loading for jet impingement and blast separately
Cutoff Frequency Criteria (Section 3.4.1.4) - Separated into two subsections to define the criteria for jet impingement and blast separately
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 31
Description of Changes to ANP-10318P
Applicable Codes and Standards (Section 3.5) – Additional information is added to cover the possibility of nonlinear elastic-plastic analysis for blast loading of steel and concrete structures.
Steel Structure Code Requirements (Section 3.5.2) –Additional information is added concerning the use of ductility ratios based on code requirements.
Concrete Structure Code Requirements (Section 3.5.3) –Additional information is added concerning the use of ductility ratios based on code requirements.
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 32
Discussion of Results of Blast Effects Methodology
Based on review of past precedent (e.g., GE ESBWR SER) it does not appear that results of the blast effect methodology are needed for NRC to reach a reasonable assurance finding
ITAAC exist in the U.S. EPR FSAR Tier 1 Section 3.8 to require that the pipe break hazards analysis report demonstrates that SSCs are protected or qualified to withstand the dynamic and environmental effects of postulated failures, including cubicle pressurization effects.
Similar to the GE ESBWR SER, ANP-10318P will require that converged, worst-case loading scenarios be applied to SSCs and neighboring structures when using the CFD analysis approach. In addition, AREVA will benchmark the CFD analysis approach prior to applying it to U.S EPR designs.
Russ WellsLead Licensing Engineer
Path Forward Activities and Schedule
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 34
Path Forward Activities and Schedule
Submit revised interim response to RAI 354, Question 3.6.2-35 to reflect blast effect methodology including markups to ANP-10318P (Note: revised interim response may not be required if results of the CFD analysis are not required for NRC to close the open item).
Follow-up public meeting/telecon to address NRC comments, if any, on the revised interim response
Submit advanced copy of final response of RAI 354, Question 3.6.2-35 to NRC (including markups to ANP-10318P)
Resolve NRC comments, if any Submit final response of RAI 354, Question 3.6.2-35 to NRC
(including formal revisions to ANP-10318P) See timeline on following slide
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 35
29 31 15 11 15 31
Jan-13 March-13 Apr-13 June-13 July-13 July-13
NRC public meeting
Submit advanced copy of final
response of RAI 354, Question 3.6.2-
35 to NRC
Timeline for Open Item Closure
Send revised interim response to RAI 354, Question
3.6.2-35 to NRC
Public meeting/Telecon
with NRC on revised interim response for
RAI 354 Q3.6.2-35
Receive NRC comments on
advanced copy of final response of
RAI 354, Question 3.6.2-35
Submit final response of RAI
354, Question 3.6.2-35 to NRC
Public Meeting to Discuss Blast Effect Methodology (RAI 354, Question 3.6.2-35) – 1/29/2013 36
Summary/Next Steps
Complete Revisions to RAI 354, Question 3.6.2-35 using the methodology described in this presentation.
Timely NRC feedback is essential to achieving closure.”
Advanced and final response to RAI 354, Question 3.6.2-35 will be submitted in a timeframe to support NRC closure of open item.