TRACE: TRAC/RELAP Advanced Computational EngineNRC Reactor Transient and LOCA Analysis
U. S. Nuclear Regulatory Commission, Washington D.C.
W 412 Standard Plant
nSmall Breaks 2”, 3”, 4”, 6” breaksnTransition breaks SI, SI + 20%, SI – 20% , PSLnEmergency Diesel Generator (EDG) startup delay times of10 seconds and 60 seconds
Conclusions
nIncreasing EDG start up time has small impact on PCT results vResults sensitivity to loop seal clearing is a known phenomena and is independent of delay time.vDEG mitigation not considered.
nIncreasing containment spray setpoint or relying on operator action is feasible for LOCA with break sizes up to SI (cold leg) and PSL (hot leg)vContainment design specific
Vision and Long Term Direction
www.nrc.gov/reactors ...
Modeling Features
Modern ArchitecturenDecouple Computational Engine from Input Processor ‹
nParallelizable Flow Logic and Solution Scheme ‹
nObject-Based Architecture ‹
nXML-Based Automated Generation of Source Code
nEfficient List-Driven Internal Data Transfer Mechanism ‹
Component 1 Component 2 Component 3
Component 1 Component 2 Component 3
Data FlowCalculational Flow
Previous TRACE Calculationaland Data Flow
Component 1 Component 2 Component 3
Component 1 Component 2 Component 3
Data FlowCalculational Flow
Component 1 Component 2 Component 3
Flow Modi�ed forParallel Calculations
alpn
tln
tvn
Heat Structure Data
alpn
tln
tvn
Fluid Data Transfer Table
table(2)%from table(2)%totable(1)%from table(1)%to
table(3)%from table(3)%to
table(1)%to = table(1)%from
Interfacial DragCoe�cient
{component type,void fraction,phase velocities,�uid properties}
TRACE
InterfacialDrag
Module
Slug
Bubble Drag
Bubble Diameter
Pro�le FactorSlug
Fraction
IntFr(provides interface)
FricIF(sets �uid conditions)
HorizontalPipe
VerticalPipe Accumulator Pressurizer
Bubbly Bubbly/Slug Transition
Annular/Mist
(etc.)
Low-Level Object-Based Representation ofPhysical Property Evaluation Scheme
Decouple Computational Enginefrom Input Processor
Solution SchemeParallelizable Flow Logic and
Object-Based Architecture
XML-Based Automated Generation ofSource Code
E�cient List-Driven InternalData Transfer Mechanism
Modern Architecture
SNAP
TRACE InputProcessing
ComputationalEngine
Other SupportApplications
3D NeutronKinetics
SNAP SystemModel Database
RELAP5ASCIIInput
TRAC-PASCIIInput
TRAC-BASCIIInput
Interprocess MessagePassing Service
Platform IndependentBinary File
“To have the capability to perform T/H safety analysis in thefuture that allows for solutions to the full spectrum ofimportant nuclear safety problems in an efficient and effectivemanner, taking complete advantage of state-of-the-artmodeling, hardware, and software capabilities.”
We must be able to resources:
We must be able to reduce and consolidatepersonnel resources needed for solving any givenproblem and for maintaining code capability bydeveloping and/or improving:
Ease-of-useSpeedRobustnessFlexibilityMaintainability/upgradability
We must be able to accommodate the newchallenges and demands for best-estimate T/Hanalysis coupled to other related capabilities:
AccuracyFlexibilityMaintainability/upgradabilitySimplicityExpanded scope of capabilitiesQuality assurance
do more with less
LESS:
MORE:
NRC relied on 4 T/H codes
Over time the di�erences eroded but coding and inputvaried substantially
The suite was developed in the 70’s and 80’s and does nottake advantage of modern technology
Identi�ed modeling de�ciencies for the same phenomena
ameliorate the limitations
PWR
BWR
Old coding language and procedural styleLarge container arrayArchaic memory saving schemes (bit-packing)
NRC would have to expend 4 times the resources tocontinue making improvements to 4 separate tools
Evolve!Always have a running productTakes advantage of current knowledge centers
RELAP5 SBLOCA and transientsTRAC-P LBLOCA
RAMONA 3D Kinetics and stabilityTRAC-B LOCA’s and transients
Architectural and modeling improvements required to
Continue to support old technology or invest in newtechnology?
Evolve from existing code base or “develop fromscratch”?
ModernizationTRAC-P selected as basis for consolidationArchitectural modi�cations to take advantage of F90features
Input deck processingBWR-speci�c topology and modeling features
Recovered with PARCS (coupled to TRACE)
Input deck processingR5-speci�c topology and modeling features
Selection of physical models that provide the simulation�delity of TRAC-B and RELAP5 without degrading that ofTRAC-P for all targeted applications
TRAC-B functionality
RAMONA functionality
RELAP5 functionality
Developmental Assessment
Historical Perspective
Visionand LongTerm Direction
Consolidation Stages
Modeling FeaturesAdditional Component Types
Jet Pump (JETP)
Component (EXTERIOR)
Steam Separator (SEPD)Turbine (TURB)Feedwater Heater (HEATR)Containment (CONTAN)Fuel Channel (CHAN)Co-located Heat Structure (HTSTR)Radiation Enclosure (RADENC)Power (POWER)Fluid Power (FLPOWER)Single Junction (SJC)Exterior Code Coupling
Partial length fuel rodsSquare, cross, and round water rod geometriesRay-traced radiation view factors
Additional valve and pump typesActive pressure boundary conditionAdditional heat structure boundary conditionsSpherical heat structure geometryNew signal variable, control block, and triptypes
Command line argument supportExtended TRAC-B-style outputImproved code robustnessPlatform-independent graphics and dump �les
Automatic sorting of control blocks, signalvariables and tripsEnhanced input checking
Advanced BWR Channel Features
Extended Component Features
Usability Enhancements
RELAP5, TRAC-P, and TRAC-B Input DeckConversion
Additional Working Fluids (H O, D O, Air, N ,He, Na, PbBi)
Generalized Support for Coarse-GrainedParallel and Coupled-Code Computations
SETS & Semi-Implicit Numerical Schemes
User-De�ned Matrix Solvers
1D & 3D Kinetics (through PARCS coupling)
Advanced 1D & 3D Level Tracking
Trace Species Tracking
ASME Steam Tables
New Re�ood Model
Improved Choked Flow Model
Enhanced User-De�ned Material Tables
2 2 2
TRACE Tall.pdf 1/19/06 4:21:49 PM
TRACE Support for 50.46 Break Size Redefinition
nModern BWR Channel FeaturesvPartial length fuel rodsvSquare, cross, and round water rod geometriesvRay-traced radiation view factorsnExtended Component FeaturesvAdditional valve and pump typesvActive pressure boundary conditionvSpherical heat structure geometryvNew signal variable, control block, and trip types
nUsability EnhancementsvCommand line argument supportvExtended TRAC-B-style outputvImproved code robustnessvPlatform-independent graphics and dump files
vAutomatic sorting of control blocks, signal variables and tripsvEnhanced input checkingnRELAP5, TRAC-P, and TRAC-BInput Deck ConversionnAdditional Working Fluids (H2O, D2O, Air, N2, He, Na, PbBi)nGeneralized Support for Coarse-Grained Parallel and Coupled-Code ComputationsnSETS & Semi-Implicit Numerical SchemesnUser-Defined Matrix Solversn1D&3D Kinetics (through PARCS coupling)nAdvanced 1D & 3D Level TrackingnASME Steam TablesnNew Reflood ModelnImproved Choked Flow ModelnEnhanced User-Defined Material Tables
ESBWR Passive Safety Systems
nThe ESBWR passive safety systems have strong coupling between the reactor and the containment. nECCS SystemvRelies on depressurization like operating BWRsvGravity driven cooling system (GDCS) to refill reactor system after blowdown.nDecay Heat Removal SystemvLarge passive tube condensers (PCCS) to remove decay heat in long term cooling.
ESBWR Accident Phases
New TRACE Physical Models
Film condensation models appropriate for modeling tubes and containment walls havebeen added to TRACE.
TRACE Assessment for ESBWR
nSeparate Effects Test AssessmentvVoid fraction and level swellvTube condensationvFlat plate condensation
nIntegral Test AssessmentvFIST BWR full pressure blowdownvPUMA late GDCS to long term coolingvPANDA long term cooling
Current Status
nNew film condensation models have been added and are being assessed.nIntegral test assessments are in progress.nPlant calculations are in progress.
“To have the capability to perform T/H safety analysis in the future that allows for solutions to the full spectrum of important nuclear safety problems in an efficient and effective manner, taking complete advantage of state-of-the-art modeling, hardware, and software capabilities.”
TRACE Project Goals
We must be able to reduce and consolidate personnel resources needed for solving any given problem and for maintaining code capability by developing and/or improving:vEase-of-usevSpeedvRobustnessvFlexibilityvMaintainability/upgradeability
We must be able to accommodate the new challenges and demands for best-estimate T/H analysis coupled to other related capabilities:vAccuracyvFlexibilityvMaintainability/upgradeabilityvSimplicityvExpanded scope of capabilitiesvQuality assurance
nNRC relied on 4 T/HcodesvPWRwRELAP5 ‹ SBLOCA and transientswTRAC-P ‹LBLOCAvBWRwRAMONA ‹3D Kinetics and stabilitywTRAC-B ‹LOCA’s and transients
nOver time the differences eroded but coding and input varied substantially
nThe suite was developed in the 70’s and 80’s and does not take advantage of modern technologyvOld coding language and procedural stylevLarge container arrayvArchaic memory saving schemes (bit-packing)nIdentified modeling deficiencies for the same phenomena
nArchitectural and modeling improvements required to ameliorate the limitationsvNRC would have to expend 4 times the resources to continue making improvements to 4 separate tools
nContinue to support old technology or invest in new technology?
nEvolve from existing code base or “develop from scratch”?vEvolve!wAlways have a running productwTakes advantage of current knowledge centers
Historical Perspective
TRACE AssessmentnThe TRACE code is under development, but significant assessment has been performed with recent code versions.nApplicable integral assessment cases include:vROSA SBLOCA IETs (6 tests)vBETHSY ISP-27
ROSA SB-CL-18 (ISP 26) Assessment
Bethsy Test 9.1B (ISP-27) TRACE Simulation
TRACE Support for ESBWR Design Certification
Able to Model All Reactor Designs
Modern Architecture
TRACE_1A.indd 1 2/15/06 3:54:13 PM