Development of a Data Standard for V&V of Software to Calculate Nuclear System Thermal-Hydraulic Behavior www.inl.gov Richard R. Schultz & Edwin Harvego (INL) Ryan Crane (ASME)
Topics addressed Development of an experimental program designed to provide data to validate thermal-hydraulic software designed to analyze the behavior of nuclear plants. V&V30 Standard Committee
Nuclear Plant System & Calculational Envelopes System operational envelope: where the plant is designed to operate and produce power plus where the plant may experience various accident scenarios during its lifetime Anticipated operational occurrences Design basis accidents Beyond design basis accidents The system envelope is defined on the basis of the plant design characteristics and past history Calculational envelope: defined by numeric model physics Objective: Demonstrate calculational envelope encompasses the system envelope Calculational envelope = Domain of qualification Calculational envelope System envelopes System envelope = Domain where the plant will operate and also may experience various accident scenarios during plant lifetime.
Typical Operational and Accident Envelope State Conditions for Light Water Reactors Anywhere under steam dome at pressures less than ~17 MPa, limited superheated conditions, and above atmospheric pressures; subcooled conditions All two-phase flow regimes in vertical and horizontal direction
To Determine the Adequacy of Numerical Tools NRC s regulatory Guideline 1.203 breaks process down into 5 pieces: 1. Define requirements 2. Develop assessment base (define data requirements) 3. Develop numerical models 4. Assess adequacy of model 5. Make decision whether good enough. 6. Following completion of process UQ is performed.
This presentation, and V&V30 standard committee, focuses on Element 2 Specifying the scope and magnitude of the validation data base. Defining the scaling approach to relate experiments to prototype. Defining the integral effects and separate effects experimental requirements. Defining the experimental uncertainties. Ultimate objective: an experimental data matrix that provides the required validation data for numerical models to enable the numeric model validation to be performed and an adequacy decision to be made.
Sample Validation Matrix LWR Large Break LOCA Scenario
Experiments Used to Provide Data to the Experiment Matrix (1) Must be scaled, using an acceptable methodology, such that the data are in an appropriate range for the plant scenario of interest Must have acceptable measurement uncertainties to provide a reasonable range of acceptance when the data are used to judge whether numerical models are capable of calculating the measured phenomena
Experiments Used to Provide Data to the Experiment Matrix (2) Should be designed as a set to create a validation pyramid that is comprised of supporting levels: Fundamental experiments give data that describes the behavior of the key phenomena in an environment free of extraneous influences, e.g., influences from other phenomena Separate effects experiments provide data that describes the behavior of key phenomena in typical system components Integral effects experiments give data that demonstrates the interactions that occur between the key phenomena for the scenarios of interest. The different scales used in the experiments of the validation pyramid provide a check on the measured experimental phenomena scaling
Validation is an In-Depth Activity
TH Phenomena: Experiment Execution/Planning Normal operation at full or partial loads Coolant flow and temperature distributions through reactor core channels ( hot channel ) Mixing of hot jets in the reactor core lower plenum ( hot streaking ) Integral Facility/ RCCS Lower Plenum Exp Core Exp MIR Exp Loss of Flow Accident (LOFA or pressurized cooldown ) Mixing of hot plumes in the reactor core upper plenum Coolant flow and temperature distributions through reactor core channels (natural circulation) Rejection of heat by natural convection and thermal radiation at the vessel outer surface Loss of Coolant Accident (LOCA or depressurized cooldown ) Prediction of reactor core depressurized cooldown - conduction and thermal radiation Rejection of heat by natural convection and thermal radiation at the vessel outer surface Plenum-to Plenum Exp Air Ingress Exp
Experiment Design The H2TS (Hierarchical Two-Tiered Scaling ) methodology was developed in 1980s by Zuber and provides the following advantages: It is a method that is systematic, auditable, and traceable Provides a unified scaling rationale and similarity criteria Assures that important processes have been identified and addressed properly and provide a means for prioritizing and selecting processes to be addressed experimentally Creates specifications for test facilities design and operation (test matrix, test initial and boundary conditions) and provide a procedure for conducting comprehensive reviews of facility design, test conditions, and results Assures the prototypicality of experimental data for important processes and quantify biases due to scale distortions or to nonprototypical test conditions VHTR experiments designed using H2TS methodology. Ref: Zuber, N., 1991, An Integrated Structure and Scaling Methodology for Severe Accident Technical Issue Resolution, NUREG/CR-5809, November, Appendix D, Hierarchical, Two- Tiered Scaling Analysis.
Zuber s H2TS Scaling Methodology Decomposes and organizes the system Starting with the whole system Working downward through subsystems, components, until reaching the transfer processes Scale measures are assigned at each level Each phase characterized by one or more geometrical configurations. Address the effects caused by the interaction of its constituents which have been identified as important in the PIRT. Similarity criteria developed at appropriate scaling level. Each geometrical configuration described by 3 field equations (mass, momentum, and energy) Includes a top-down (system) and bottom-up scaling analysis development of similarity criteria for interaction of constituents and specific processes
Application of H2TS Methodology to Separate- Effects and Integral-Effects Experiments
Scaling Choices Preserving kinematic similarity Preserving friction and form loss similarity Elevation and length scale ratio Diameter scaling ratio Pressure scaling ratio (full pressure) a i = 1 ac R ( Π ) = 1 F R ( L ) = 1:4 R ( D ) = 1:4 R ( ) P 0 = 1:8 R
Separate Effects Experiments Designed to Match Key Phenomena Ranges of Prototype Designed to capture key phenomena Scaled to provide direct link between subscaled experimental facility and prototypical plant Low, quantified uncertainties Experiment design should consider decomposition of behavior in system component to lowest level that can be modeled by software to ensure each component is properly being calculated by software physics Courtesy C. Oh
Summary of Process Objective of NGNP methods is to validate numeric models that are capable of calculating VHTR behavior throughout its system operational/accident envelope For the NGNP, a rigorous process is used to design experiments that will provide validation data for both systems analysis and CFD numerical models The process is based on the H2TS methodology Experimental and validation matrices are defined Both separate-effects and integral-effects experiments are underway Validation of numeric models is also proceeding The NGNP approach is compatible with NRC guidelines and practices as used to qualify numeric models for Generation III+ systems
V&V30 Committee Overview Committee scope and objectives Committee structure Committee function Anticipated content of Verification and Validation (V&V) 30 Standard Relationship to NQA-1 and other Nuclear Standards and Regulations Summary
V&V30 Scope & Objectives (presently under discussion) Develop a standard that: Defines validation data based requirements for V&V of CFD and system analysis codes used in nuclear applications Defines requirements for experimental data used for software validation Is consistent with all Nuclear Regulatory Commission (NRC) regulatory requirements Is consistent with or complements related consensus standards (existing or under development) Specifically addresses requirements unique to High-Temperature Gas- Cooled Reactors (HTGRs) as a starting point; subsequently addresses requirements for LWRs. Considers the potential for coupling of CFD and system analysis codes as part of the analysis process Meets ANSI requirements
Committee Structure Consist of 8-20 members selected from Industry, national laboratories, academia, and government (NRC and/or Department of Energy), and serving a maximum 5 year term Selection of members will be based on technical qualifications and to ensure balanced representation Participation on the committee is as an individual technical expert, not as a representative of a particular organization or interest group A chair and vice chair will be elected for three-year terms by the voting committee membership Chair is also the committee representative to the Verification and Validation Standards Committee The secretary is a non-voting appointed position and a member of the American Society for Mechanical Engineers (ASME) staff with no defined term of office
Committee Function Committee members will typically meet two to three times a year to discuss and review progress on development of the standard Other group or individual meetings may be required to coordinate and/or present results to regulators and/or other consensus standards development organizations Based on committee deliberations, individual members will research and develop different topics to be included in the standard Most of the work in writing the standard will be done by individual committee members between meetings At the discretion of the committee, contributing (non-voting) members may be solicited (within or outside the U.S.) to serve in review and consulting roles All activities of the committee will of course be subject to availability of funds
Anticipated Content of V&V 30 Standard Definition of operational and accident domains that must be considered for licensing the nuclear reactor Description of calculation domain that is target for developing validation matrix Requirements for the experimental data sets to be used for validation of CFD and/or system analysis software Requirements for the ensemble of experimental data sets used to populate the validation matrix for the software in question Application of the software validation standard Direct reference to appropriate regulatory requirements for each topic addressed in the standard
Definition of Operational and Accident Domains USNRC Standard Review Plan (NUREG-0800) developed for pressurized water reacotrs and boiling water reactors, but provides the framework for specifying operational and accident domains in HTGRs Ensure a sufficiently broad spectrum of transients and accidents, or initiating events. Initiating events categorized according to expected frequency of occurrence and by type. Anticipated Operational Occurrence (AOO) one or more times during the life of the nuclear plant Anticipated transients without scram (ATWSs) AOOs with failure to scram (beyond design basis) Postulated accidents unanticipated accidents that are not expected to occur during the life of the nuclear plant
Definition of Operational and Accident Domains cont. Grouping of AOOs and postulated accidents by types: Increase in heat removal by the secondary system Decrease in heat removal by the secondary system Decrease in RCS flow rate Reactivity and power distribution anomalies Increase in reactor coolant inventory Decrease in reactor coolant inventory Radioactive release from a subsystem or component
Calculation Domain Calculation envelope of the thermal-hydraulic software must match or encompass the system operational and accident envelope Phenomena Identification and Ranking Table (PIRT) process provides a basis for ranking important phenomena associated with each scenario in the operational domain Software physics models must properly calculate the key phenomena over the entire range of conditions encompassed by the calculation envelope Basis for assessing the adequacy of the CFD and system analysis software models is experimental data
Requirements for Experimental Data and Experimental Data Sets Proposed standard should provide the processes and procedures for determining the data needed to populate software validation matrices for both system analysis and CFD software Processes and procedures should address evaluation of both existing experimental data and procedures for defining new data needs Software validation matrices should include both separate effects experiments for evaluating localized phenomena and integral effects experiments for evaluating global system response Experiment validation matrices will include data from experimental facilities at different scales, so that scaling effects can be evaluated to identify any scaling distortions and provide confidence in scaling assessments performed as part of the software validation process
Considerations for New Facilities Assure that the proposed experiment facility captures key phenomena being investigated Experiment is scaled to provide a direct link between the scaled facility and prototype plant Adequate high-quality measurements are available to ensure that experimental data uncertainties are quantifiable and acceptably low Experiment results can be decomposed to the lowest level modeled by the software to ensure that system behavior at the component level is properly being calculated by the governing software physics Quality assurance meets applicable requirements of ASME Standard NQA-1
Application of Software Validation Standard It is anticipated that the standard will be used by vendors as the basis for verification and validation of software for licensing advanced reactor designs Standard should conform to current NRC and other regulatory requirements and guidelines, but will provide additional detail on acceptable processes and procedures used to meet regulatory requirements Standard is being developed in conjunction with ASME V&V 10 and V&V 20 Standards, and should be consistent with and complement these standards V&V 30 Committee presently deciding upon the detailed scope and intent of this Standard and path forward
Relationship of V&V 30 Categories to Equivalent Regulatory Requirements and Guidelines for LWRs V&V 30 Categories /Federal Regs. System Envelope Calculation Envelope Experiment Matrix Data Validation 10CFR50 50.34 - Mitigation of accidents 50.46 - Acceptance criteria/ EM concept App. A General Design Criteria Appendix B QA criteria, testing Standard Review Plan Chpt. 15 - AOO and accident categories Chpt. 15 Defines model requirements Chpt. 15 EM variable ranges Reg. Guide 1.203 1.1 Defines application envelope Pg. 4(5) - Software peer review 1.2 (2) Need for assessment base 1.4 (4) - Scaling methodology Reg. Guide 1.157 Characteristics of BE Model Experimental data requirements
Relationship of V&V 30 Categories to Other Consensus Standards V&V 30 Categories/ Other Standards ASME NQA-1-2008 ASME V&V 10 System Envelope Describe physical phenomena Calculation Envelope PIRT to develop CSM models Experiment Matrix Experiment design Data Validation Expands on QA criteria in 10CFR 50, App. B Sources of experiment error ASME V&V 20 ASME PTC 19.1 Estimate simulation model error Experiment error quantification Standard for uncertainty evaluation
Relationship of V&V 30 Categories to Other Consensus Standards cont. V&V 30 Categories/ Other Standards System Envelope Calculation Envelope Experiment Matrix Data Validation ANSI/ANS-10.4-2008 V&V of non-safety related software for nuclear appl. ANSI/IEEE Std 1012-1998 V&V software processes ANS-10.7 (under development) Non-real time, high integrity software for Nuclear Industry
Summary V&V 30 Committee established and operational Committee presently discussing scope and content of standard. Have initiated interactions with other stakeholder committees and potential cooperation with NQA-1 Standard committee