ML19093B817: Difference between revisions

From kanterella
Jump to navigation Jump to search
(Created page by program invented by StriderTol)
(StriderTol Bot change)
 
(One intermediate revision by the same user not shown)
Line 2: Line 2:
| number = ML19093B817
| number = ML19093B817
| issue date = 02/12/2019
| issue date = 02/12/2019
| title = RIC 2019 - TH35 Presentation - Overview of Upgrades to Maccs Dispersion Modeling Capabilities.
| title = RIC 2019 - TH35 Presentation - Overview of Upgrades to Maccs Dispersion Modeling Capabilities
| author name = Barr J, Bixler N, Clayton D, Compton K
| author name = Barr J, Bixler N, Clayton D, Compton K
| author affiliation = NRC/RES, Sandia National Labs (SNL)
| author affiliation = NRC/RES, Sandia National Labs (SNL)
Line 16: Line 16:


=Text=
=Text=
{{#Wiki_filter:2/12/2019 Overview of Upgrades to MACCS Dispersion Modeling Capabilities 2019 NRC Regulatory Information Conference Atmospheric Transport and Dispersion Modeling for Severe Accident Consequence Analysis K. Compton*, J. Barr*, N. Bixler**, and D. Clayton**
{{#Wiki_filter:2/12/2019 1
                  *U.S. Nuclear Regulatory Commission
OverviewofUpgradestoMACCS DispersionModelingCapabilities 2019NRCRegulatoryInformationConference AtmosphericTransportandDispersionModelingforSevere AccidentConsequenceAnalysis K.Compton*,J.Barr*,N.Bixler**,andD.Clayton**
                      **Sandia National Laboratories 1
*U.S.NuclearRegulatoryCommission
**SandiaNationalLaboratories


2/12/2019 Outline
2/12/2019 2
- Severe accident consequence analysis with the MELCOR Accident Consequence Code System (MACCS)
Outline
- Overview of current MACCS atmospheric transport and dispersion (ATD) model capabilities
- SevereaccidentconsequenceanalysiswiththeMELCOR AccidentConsequenceCodeSystem(MACCS)
- Ongoing dispersion model enhancements
- OverviewofcurrentMACCSatmospherictransportand dispersion(ATD)modelcapabilities
- Potential nearfield dispersion model enhancements 2
- Ongoingdispersionmodelenhancements
2
- Potentialnearfielddispersionmodelenhancements 2


2/12/2019 Overview of Severe Accident Consequence Analysis
2/12/2019 3
* Typically includes modeling the radioactive release to the atmosphere
OverviewofSevereAccidentConsequenceAnalysis 3
* Analyses typically estimate the doses and health effects from inhalation, cloudshine, groundshine, skin deposition, and ingestion (e.g., water, milk, meat, crops), as well as costs and other impacts associated with protective actions to reduce exposure 3
Typicallyincludesmodelingtheradioactivereleasetotheatmosphere Analysestypicallyestimatethedosesandhealtheffectsfrominhalation, cloudshine,groundshine,skindeposition,andingestion(e.g.,water,milk, meat,crops),aswellascostsandotherimpactsassociatedwithprotective actionstoreduceexposure
3


2/12/2019 Considerations for Severe Accident Assessments
2/12/2019 4
* Severe accident probabilistic consequence assessments
ConsiderationsforSevereAccidentAssessments
  - Realistic assessment
* Severeaccidentprobabilisticconsequenceassessments
  - Prospective analysis
- Realisticassessment
  - Multiple figures of merit
- Prospectiveanalysis
  - Wide temporal and spatial scales
- Multiplefiguresofmerit
* Computational efficiency needed for modeling multiple possible prospective weather conditions
- Widetemporalandspatialscales
* Traditional ATD approach for probabilistic consequence analysis is to use a Gaussian plume segment model 4
* Computationalefficiencyneededformodelingmultiple possibleprospectiveweatherconditions
4
* TraditionalATDapproachforprobabilisticconsequence analysisistouseaGaussianplumesegmentmodel 4


2/12/2019 MACCS Modules
2/12/2019 5
* ATMOS
MACCSModules ATMOS
  - Source term definition
- Sourcetermdefinition
  - Weather sampling algorithms
- Weathersamplingalgorithms
  - Atmospheric transport, dispersion, and deposition
- Atmospherictransport,dispersion,anddeposition EARLY(1to40days)
* EARLY (1 to 40 days)
- Dosesasmodifiedbyemergencyphasecountermeasuressuchas sheltering,evacuation,relocation,andpotassiumiodideingestion
  - Doses as modified by emergency phase countermeasures such as sheltering, evacuation, relocation, and potassium iodide ingestion
- Allowsmodelingofmultiplepopulationcohorts
  - Allows modeling of multiple population cohorts
- Acuteandlatenthealtheffectsfromearlyacuteexposure CHRONC(1weekto>50years)
  - Acute and latent health effects from early acute exposure
- Dosesasmodifiedbyintermediateandrecoveryphaseprotectiveactions suchasrelocation,temporaryandpermanentinterdiction,and decontamination
* CHRONC (1 week to >50 years)
- Latenthealtheffectsfromchronicexposuretodepositedmaterial
  - Doses as modified by intermediate and recovery phase protective actions such as relocation, temporary and permanent interdiction, and decontamination
- Economicimpactfromearlyandlatephaseprotectiveactions 5
  - Latent health effects from chronic exposure to deposited material
  - Economic impact from early and late phase protective actions 5
5


2/12/2019 ATMOS: Meteorological Data
2/12/2019 6
* MACCS ATMOS Gaussian plume segment model typically reads data from a file containing 1 years worth of observed hourly meteorological data (8,760 observations)
ATMOS:MeteorologicalData
  - Wind speed (at 10 meters)
* MACCSATMOSGaussianplumesegmentmodeltypically readsdatafromafilecontaining1yearsworthofobserved hourlymeteorologicaldata(8,760observations)
  - Wind direction (sector into which wind is blowing)
- Windspeed(at10meters)
  - Stability class (typically based on lapse rate)
- Winddirection(sectorintowhichwindisblowing)
  - Accumulated precipitation
- Stabilityclass(typicallybasedonlapserate)
* Each plume segment uses up to 1,000 hours of meteorological data from meteorological file
- Accumulatedprecipitation
* Up to 500 plume segments can be defined
* Eachplumesegmentusesupto1,000hoursofmeteorological datafrommeteorologicalfile
* Typical practice is to sample approximately 1,000 different accident initiation times from the 8,760 potential hourly start times 6
* Upto500plumesegmentscanbedefined
6
* Typicalpracticeistosampleapproximately1,000different accidentinitiationtimesfromthe8,760potentialhourlystart times 6


2/12/2019 ATMOS ATD Model Plume Dispersion
2/12/2019 7
* Straightline Gaussian plume segment model
ATMOS ATDModelPlumeDispersion StraightlineGaussianplumesegmentmodel Userspecifiedplumedispersionparameterization
* Userspecified plume dispersion parameterization
- Distancebasedpowerlawformulation
  - Distancebased power law formulation b                          d y = a x                    z = c  x
- Distancebasedlookuptables:allowformulationsthatdonotfitasimple powerlaw(e.g.,BriggsorEimutisandKonicekformulations)
  - Distancebased lookup tables: allow formulations that do not fit a simple power law (e.g., Briggs or Eimutis and Konicek formulations)
Horizontalandverticalscalingfactors(YSCALEandZSCALE)
* Horizontal and vertical scaling factors (YSCALE and ZSCALE)
- ZSCALEtypicallyusedtoaccountforsurfaceroughnesseffectsonvertical dispersion Longrangedispersion:timebasedplumegrowthmodel 7
  - ZSCALE typically used to account for surface roughness effects on vertical dispersion
y a
* Longrange dispersion: timebased plume growth model
x b
:      ,
* 7 7


2/12/2019 ATMOS Gaussian Plume Model Modifications Virtual source model
=
- Accounts for building wake effects and timevarying meteorological conditions
z c
- User can specify initial plume dimensions to simulate initial dispersion from wake effects Plume Meander
x d
- Userselectable model applies corrections to transverse dispersion coefficients
- Original model (accounts for sampling time)
- Alternative model based on NUREG/CR2260 (accounts for wind speed and stability)
MACCS also includes modifications to account for plume rise, wet deposition, and dry deposition 8
8


2/12/2019 ATMOS (Gaussian Plume Segment): Air Concentration
=
* Illustration showing how plume segments move with wind shifting from northwest to northeast
* Each segment has its own width depending on the amount of dispersion that has occurred as it experiences varying weather conditions
* Each segment has a unique length depending on wind speed
* Wet and dry deposition (not shown) results in plume depletion and buildup of ground concentration 9
9


2/12/2019 Accuracy of MACCS
2/12/2019 8
* Probabilistic consequence analysis applications generally require the generation of the statistical distribution of consequence values over annual weather data (i.e., weatheraveraged calculations)
ATMOS GaussianPlumeModelModifications 8
* ATD model fidelity for probabilistic consequence analysis is traditionally considered less significant than for emergency response tools because of the statistical nature of probabilistic consequence analyses
Virtualsourcemodel
* However, questions about the adequacy of the Gaussian plume model for probabilistic consequence analyses have been raised since at least the 1980s
- Accountsforbuildingwakeeffectsand timevaryingmeteorologicalconditions
* A previous study (NUREG/CR6853) showed that MACCS is generally within a factor of two of the National Atmospheric Release Advisory Centers Lagrangian Operational Dispersion Integrator (LODI) code for weatheraveraged calculations 10 10
- Usercanspecifyinitialplumedimensionsto simulateinitialdispersionfromwakeeffects PlumeMeander
- Userselectablemodelappliescorrectionsto transversedispersioncoefficients
- Originalmodel(accountsforsamplingtime)
- AlternativemodelbasedonNUREG/CR2260 (accountsforwindspeedandstability)
MACCSalsoincludesmodificationstoaccountforplumerise,wetdeposition, anddrydeposition


2/12/2019 Evaluation Attributes for Potential Alternative ATD Codes
2/12/2019 9
* Essential features
ATMOS(GaussianPlumeSegment):AirConcentration Illustrationshowinghow plumesegmentsmovewith windshiftingfromnorthwest tonortheast Eachsegmenthasitsown widthdependingonthe amountofdispersionthathas occurredasitexperiences varyingweatherconditions Eachsegmenthasaunique lengthdependingonwind speed Wetanddrydeposition(not shown)resultsinplume depletionandbuildupof groundconcentration 9
    - Treats Lagrangian particles
    - Code executable files can be distributed to users (either by Sandia National Laboratories or directly from the developer)
    - Source code available in case modifications are required for integration
    - Code has adequate quality assurance (QA)
    - Treats threedimensional (3D) wind field (gridded data)
* Desirable features
    - Can model both puffs and particles
    - Supports a variety of gridded weather data formats
    - Facilitates graphical postprocessing 11 11


2/12/2019 Features of Potential Alternative ATD Codes HYSPLIT          FLEXPART            LODI      CALPUFF        RASCAL        SCIPUFF Dispersion    Lagrangian/      Lagrangian        Lagrangian/  Gaussian      Gaussian      Gaussian Gaussian                            Gaussian Developer/    NOAA & Aus.       Norwegian Inst. LLNL/        TRC/          PNNL/        Titan/
2/12/2019 10 AccuracyofMACCS Probabilisticconsequenceanalysisapplicationsgenerallyrequire thegenerationofthestatisticaldistributionofconsequencevalues overannualweatherdata(i.e.,weatheraveragedcalculations)
Primary      Bureau of Met./  for Air Research/  DOE          EPA          NRC          DTRA, NASA,            Europe                                                      U.S. Army Customers Universities, Weather      Any gridded data  WRF, MM5          WRF, ADAPT    CALMET,      Surface data  Surface or Data                                                            MM5, etc.                  gridded data Surface      User input at any User input at any  User input at User input at User input at Unknown Roughness    resolution, or    resolution        any          any          any 1 degree                            resolution    resolution    resolution Availability/ NOAA/Yes          Norwegian Inst. Only via      TRC/No        NRC/Yes      Titan/No Source Code                    for Air Research/  Web/No Europe/Yes QA/Verifica  Yes, open        Open literature    Yes, open    Yes          Yes          Yes tion and      literature                          literature Validation Platforms    Unix/Linux,      Unix/Linux, PC    Unix/Linux    PC            PC            PC PC, Mac Graphics      Yes              User plots output  Yes          Yes          Yes          Yes Computer      Can be high/      Can be high/      Can be high/  Can be high  Moderate      Can be high Time          needs multiple    needs multiple    needs processors        processors        multiple processors 12 12
ATDmodelfidelityforprobabilisticconsequenceanalysisis traditionallyconsideredlesssignificantthanforemergency responsetoolsbecauseofthestatisticalnatureofprobabilistic consequenceanalyses However,questionsabouttheadequacyoftheGaussianplume modelforprobabilisticconsequenceanalyseshavebeenraised sinceatleastthe1980s Apreviousstudy(NUREG/CR6853)showedthatMACCSis generallywithinafactoroftwooftheNationalAtmospheric ReleaseAdvisoryCentersLagrangianOperationalDispersion Integrator(LODI)codeforweatheraveragedcalculations 10


2/12/2019 Selected Alternative ATD Code: HYSPLIT
2/12/2019 11 EvaluationAttributesforPotentialAlternativeATDCodes
* Selected for integration into MACCS as a modern alternative to the Gaussian plume segment model
* Essentialfeatures
* Developed and maintained by the National Oceanic and Atmospheric Administration (NOAA) Air Resources Laboratory
- TreatsLagrangianparticles
* Can compute trajectories as well as complex transport, dispersion, chemical transformation, and deposition
- Codeexecutablefilescanbedistributedtousers(eitherbySandia NationalLaboratoriesordirectlyfromthedeveloper)
* Accepts a wide variety of publicly available gridded meteorological data, including both North American and global datasets
- Sourcecodeavailableincasemodificationsarerequiredfor integration
* Models dispersion using either puffs or particles:
- Codehasadequatequalityassurance(QA)
    - [P]uffs expand until they exceed the size of the meteorological grid cell (either horizontally or vertically) and then split into several new puffs, each with its share of the pollutant mass. *
- Treatsthreedimensional(3D)windfield(griddeddata)
    - [A] fixed number of particles are advected about the model domain by the mean wind field and spread by a turbulent component. *
* Desirablefeatures
*https://www.arl.noaa.gov/HYSPLIT_info.php 13 13
- Canmodelbothpuffsandparticles
- Supportsavarietyofgriddedweatherdataformats
- Facilitatesgraphicalpostprocessing 11


2/12/2019 Potential MACCS Enhancements for NonLightWater Reactors (NonLWRs)
2/12/2019 12 FeaturesofPotentialAlternativeATDCodes 12 HYSPLIT FLEXPART LODI CALPUFF RASCAL SCIPUFF Dispersion Lagrangian/
* Evaluating potential code development needs for designspecific issues
Gaussian Lagrangian Lagrangian/
  -  Radionuclide screening
Gaussian Gaussian Gaussian Gaussian Developer/
  -  Radionuclide chemical form
Primary Customers NOAA&Aus.
  -  Aerosol size
BureauofMet./
  -  Aerosol shape factor
: NASA, Universities, NorwegianInst.
  -  Radionuclides with complex transport characteristics (e.g., tritium)
forAirResearch/
* Evaluating potential code development needs for siterelated issues
Europe LLNL/
  - Nearfield atmospheric transport
DOE TRC/
  - Decontamination modeling 14 14
EPA PNNL/
NRC Titan/
: DTRA, U.S.Army Weather Data Anygriddeddata WRF, MM5 WRF,ADAPT
: CALMET, MM5,etc.
Surfacedata Surfaceor griddeddata Surface Roughness Userinputatany resolution,or 1degree Userinputatany resolution Userinputat any resolution Userinputat any resolution Userinputat any resolution Unknown Availability/
SourceCode NOAA/Yes NorwegianInst.
forAirResearch/
Europe/Yes Onlyvia Web/No TRC/No NRC/Yes Titan/No QA/Verifica tionand Validation Yes,open literature Openliterature Yes,open literature Yes Yes Yes Platforms Unix/Linux, PC,Mac Unix/Linux,PC Unix/Linux PC PC PC Graphics Yes Userplotsoutput Yes Yes Yes Yes Computer Time Canbehigh/
needsmultiple processors Canbehigh/
needsmultiple processors Canbehigh/
needs multiple processors Canbe high Moderate Canbehigh


2/12/2019 NearField Atmospheric Transport
2/12/2019 13 SelectedAlternativeATDCode:HYSPLIT SelectedforintegrationintoMACCSasamodernalternativetothe Gaussianplumesegmentmodel DevelopedandmaintainedbytheNationalOceanicandAtmospheric Administration(NOAA)AirResourcesLaboratory Cancomputetrajectoriesaswellascomplextransport,dispersion, chemicaltransformation,anddeposition Acceptsawidevarietyofpubliclyavailablegriddedmeteorologicaldata, includingbothNorthAmericanandglobaldatasets Modelsdispersionusingeitherpuffsorparticles:
* MACCS currently has a simple virtual source model for building wake effects; its user guide (NUREG/CR6613) cautions against use closer than 500 meters
[P]uffsexpanduntiltheyexceedthesizeofthemeteorologicalgridcell(either horizontallyorvertically)andthensplitintoseveralnewpuffs,eachwithitsshareofthe pollutantmass.*
* Licensees for nonLWRs (and small modular reactors) desire a smaller emergency planning zone and site boundary than for large LWRs; therefore, better modeling of nearfield phenomena may be beneficial Wind tunnel simulation of streamlines near a cubic building Schulman, L.L., D.G. Strimaitis, and J.S. Scire (2000). Development and Evaluation of the PRIME Plume Rise and Building Downwash Model, Journal of the Air & Waste Management Association, 50:3, 378-390 15 15
[A]fixednumberofparticlesareadvectedaboutthemodeldomainbythemeanwind fieldandspreadbyaturbulentcomponent.*
13
*https://www.arl.noaa.gov/HYSPLIT_info.php


2/12/2019 NearField Atmospheric Transport Example QUICURB simulation of wind
2/12/2019 14 PotentialMACCSEnhancementsforNonLightWater Reactors(NonLWRs)
* Various options for addressing                vectors nearfield ATD
* Evaluatingpotentialcodedevelopmentneedsfor designspecificissues
  - Modifications to Gaussian plume segment ATD model
- Radionuclidescreening
  - Lagrangian particle tracking ATD using simplified 3D wind field models
- Radionuclidechemicalform
  - Lagrangian particle tracking ATD using CFD modeling of 3D wind field
- Aerosolsize
* Considerations for evaluating options        Example QUICPLUME simulation of urban
- Aerosolshapefactor
  - Extent of practical acceptance in the user transport and dispersion community
- Radionuclideswithcomplextransportcharacteristics(e.g.,tritium)
  - Simplicity of use
* Evaluatingpotentialcodedevelopmentneedsforsiterelated issues
  - Computational efficiency
- Nearfieldatmospherictransport
  - Cost and time efficiency
- Decontaminationmodeling 14
  - Accuracy
  - Feasibility for probabilistic application QUIC Factsheet, Los Alamos National Laboratory 16 16


2/12/2019 Summary
2/12/2019 15 NearFieldAtmosphericTransport MACCScurrentlyhasasimplevirtualsourcemodelforbuildingwake effects;itsuserguide(NUREG/CR6613)cautionsagainstusecloserthan 500meters LicenseesfornonLWRs(andsmallmodularreactors)desireasmaller emergencyplanningzoneandsiteboundarythanforlargeLWRs; therefore,bettermodelingofnearfieldphenomenamaybebeneficial 15 Schulman,L.L.,D.G.Strimaitis,andJ.S.Scire(2000).DevelopmentandEvaluationofthePRIMEPlumeRiseand BuildingDownwashModel,JournaloftheAir&WasteManagementAssociation,50:3,378-390 Windtunnelsimulationofstreamlinesnearacubicbuilding
* The current version of MACCS is an efficient and highly flexible probabilistic consequence analysis tool with many userconfigurable options
* Code enhancements are nearing completion to introduce results computed by HYSPLIT into MACCS
* Examining the applicability of current and potential approaches for modeling nearfield dispersion phenomena where building wakes may be important 17 17


2/12/2019 ACRONYMS AND REFERENCES Acronyms
2/12/2019 16 NearFieldAtmosphericTransport Variousoptionsforaddressing nearfieldATD
-   ATD: Atmospheric Transport and Dispersion
- ModificationstoGaussianplumesegment ATDmodel
FLEXPART: FLEXible PARTicle Dispersion Model
- LagrangianparticletrackingATDusing simplified3Dwindfieldmodels
HYSPLIT: Hybrid Single Particle Lagrangian Integrated Trajectory Model
- LagrangianparticletrackingATDusingCFD modelingof3Dwindfield Considerationsforevaluatingoptions
LODI: Lagrangian Operational Dispersion Integrator
- Extentofpracticalacceptanceintheuser community
MACCS: MELCOR Accident Consequence Code System
- Simplicityofuse
NOAA: National Oceanic and Atmospheric Administration
- Computationalefficiency
QUIC: Quick Urban & Industrial Complex Dispersion Modeling System References
- Costandtimeefficiency
Chanin, D., and M.L. Young (1998). Code Manual for MACCS2: Volume 1, Users Guide (NUREG/CR6613), Sandia National Laboratories, Albuquerque, NM, May 1998
- Accuracy
Molenkamp, C.R., N.E. Bixler, C.W. Morrow, J.V. Ramsdell, Jr., and J.A. Mitchell (2004). Comparison of Average Transport and Dispersion Among a Gaussian, a TwoDimensional, and a ThreeDimensional Model (NUREG/CR6853), Lawrence Livermore National Laboratory, Livermore, CA, October 2004
- Feasibilityforprobabilisticapplication 16 ExampleQUICURBsimulationofwind vectors ExampleQUICPLUMEsimulationofurban transportanddispersion QUICFactsheet,LosAlamosNationalLaboratory
Schulman, L.L., D.G. Strimaitis, and J.S. Scire (2000). Development and Evaluation of the PRIME Plume Rise and Building Downwash Model, Journal of the Air & Waste Management Association, 50:3, 378-390
 
Snell, W.G., and R.W. Jubach (1981). Technical Basis for Regulatory Guide 1.145, Atmospheric Dispersion Models for Potential Accident Consequence Assessments at Nuclear Power Plants (NUREG/CR2260),
2/12/2019 17 Summary
NUS Corporation, Rockville, MD, October 1981 18 18}}
* ThecurrentversionofMACCSisanefficientandhighly flexibleprobabilisticconsequenceanalysistoolwithmany userconfigurableoptions
* Codeenhancementsarenearingcompletiontointroduce resultscomputedbyHYSPLITintoMACCS
* Examiningtheapplicabilityofcurrentandpotential approachesformodelingnearfielddispersionphenomena wherebuildingwakesmaybeimportant 17
 
2/12/2019 18 ACRONYMSANDREFERENCES Acronyms ATD:AtmosphericTransportandDispersion FLEXPART:FLEXiblePARTicleDispersionModel HYSPLIT:HybridSingleParticleLagrangianIntegratedTrajectoryModel LODI:LagrangianOperationalDispersionIntegrator MACCS:MELCORAccidentConsequenceCodeSystem NOAA:NationalOceanicandAtmosphericAdministration QUIC:QuickUrban&IndustrialComplexDispersionModelingSystem References Chanin,D.,andM.L.Young(1998).CodeManualforMACCS2:Volume1,UsersGuide (NUREG/CR6613),SandiaNationalLaboratories,Albuquerque,NM,May1998 Molenkamp,C.R.,N.E.Bixler,C.W.Morrow,J.V.Ramsdell,Jr.,andJ.A.Mitchell(2004).Comparisonof AverageTransportandDispersionAmongaGaussian,aTwoDimensional,andaThreeDimensional Model(NUREG/CR6853),LawrenceLivermoreNationalLaboratory,Livermore,CA,October2004 Schulman,L.L.,D.G.Strimaitis,andJ.S.Scire(2000).DevelopmentandEvaluationofthePRIMEPlume RiseandBuildingDownwashModel,JournaloftheAir&WasteManagementAssociation,50:3,378-390 Snell,W.G.,andR.W.Jubach(1981).TechnicalBasisforRegulatoryGuide1.145,AtmosphericDispersion ModelsforPotentialAccidentConsequenceAssessmentsatNuclearPowerPlants(NUREG/CR2260),
NUSCorporation,Rockville,MD,October1981 18}}

Latest revision as of 03:50, 5 January 2025

RIC 2019 - TH35 Presentation - Overview of Upgrades to Maccs Dispersion Modeling Capabilities
ML19093B817
Person / Time
Issue date: 02/12/2019
From: Jonathan Barr, Bixler N, Clayton D, Keith Compton
Office of Nuclear Regulatory Research, Sandia
To:
Shared Package
ML19093B814 List:
References
Download: ML19093B817 (18)


Text

2/12/2019 1

OverviewofUpgradestoMACCS DispersionModelingCapabilities 2019NRCRegulatoryInformationConference AtmosphericTransportandDispersionModelingforSevere AccidentConsequenceAnalysis K.Compton*,J.Barr*,N.Bixler**,andD.Clayton**

  • U.S.NuclearRegulatoryCommission
    • SandiaNationalLaboratories

2/12/2019 2

Outline

- SevereaccidentconsequenceanalysiswiththeMELCOR AccidentConsequenceCodeSystem(MACCS)

- OverviewofcurrentMACCSatmospherictransportand dispersion(ATD)modelcapabilities

- Ongoingdispersionmodelenhancements

- Potentialnearfielddispersionmodelenhancements 2

2/12/2019 3

OverviewofSevereAccidentConsequenceAnalysis 3

Typicallyincludesmodelingtheradioactivereleasetotheatmosphere Analysestypicallyestimatethedosesandhealtheffectsfrominhalation, cloudshine,groundshine,skindeposition,andingestion(e.g.,water,milk, meat,crops),aswellascostsandotherimpactsassociatedwithprotective actionstoreduceexposure

2/12/2019 4

ConsiderationsforSevereAccidentAssessments

  • Severeaccidentprobabilisticconsequenceassessments

- Realisticassessment

- Prospectiveanalysis

- Multiplefiguresofmerit

- Widetemporalandspatialscales

  • Computationalefficiencyneededformodelingmultiple possibleprospectiveweatherconditions
  • TraditionalATDapproachforprobabilisticconsequence analysisistouseaGaussianplumesegmentmodel 4

2/12/2019 5

MACCSModules ATMOS

- Sourcetermdefinition

- Weathersamplingalgorithms

- Atmospherictransport,dispersion,anddeposition EARLY(1to40days)

- Dosesasmodifiedbyemergencyphasecountermeasuressuchas sheltering,evacuation,relocation,andpotassiumiodideingestion

- Allowsmodelingofmultiplepopulationcohorts

- Acuteandlatenthealtheffectsfromearlyacuteexposure CHRONC(1weekto>50years)

- Dosesasmodifiedbyintermediateandrecoveryphaseprotectiveactions suchasrelocation,temporaryandpermanentinterdiction,and decontamination

- Latenthealtheffectsfromchronicexposuretodepositedmaterial

- Economicimpactfromearlyandlatephaseprotectiveactions 5

2/12/2019 6

ATMOS:MeteorologicalData

  • MACCSATMOSGaussianplumesegmentmodeltypically readsdatafromafilecontaining1yearsworthofobserved hourlymeteorologicaldata(8,760observations)

- Windspeed(at10meters)

- Winddirection(sectorintowhichwindisblowing)

- Stabilityclass(typicallybasedonlapserate)

- Accumulatedprecipitation

  • Eachplumesegmentusesupto1,000hoursofmeteorological datafrommeteorologicalfile
  • Upto500plumesegmentscanbedefined
  • Typicalpracticeistosampleapproximately1,000different accidentinitiationtimesfromthe8,760potentialhourlystart times 6

2/12/2019 7

ATMOS ATDModelPlumeDispersion StraightlineGaussianplumesegmentmodel Userspecifiedplumedispersionparameterization

- Distancebasedpowerlawformulation

- Distancebasedlookuptables:allowformulationsthatdonotfitasimple powerlaw(e.g.,BriggsorEimutisandKonicekformulations)

Horizontalandverticalscalingfactors(YSCALEandZSCALE)

- ZSCALEtypicallyusedtoaccountforsurfaceroughnesseffectsonvertical dispersion Longrangedispersion:timebasedplumegrowthmodel 7

y a

x b

=

z c

x d

=

2/12/2019 8

ATMOS GaussianPlumeModelModifications 8

Virtualsourcemodel

- Accountsforbuildingwakeeffectsand timevaryingmeteorologicalconditions

- Usercanspecifyinitialplumedimensionsto simulateinitialdispersionfromwakeeffects PlumeMeander

- Userselectablemodelappliescorrectionsto transversedispersioncoefficients

- Originalmodel(accountsforsamplingtime)

- AlternativemodelbasedonNUREG/CR2260 (accountsforwindspeedandstability)

MACCSalsoincludesmodificationstoaccountforplumerise,wetdeposition, anddrydeposition

2/12/2019 9

ATMOS(GaussianPlumeSegment):AirConcentration Illustrationshowinghow plumesegmentsmovewith windshiftingfromnorthwest tonortheast Eachsegmenthasitsown widthdependingonthe amountofdispersionthathas occurredasitexperiences varyingweatherconditions Eachsegmenthasaunique lengthdependingonwind speed Wetanddrydeposition(not shown)resultsinplume depletionandbuildupof groundconcentration 9

2/12/2019 10 AccuracyofMACCS Probabilisticconsequenceanalysisapplicationsgenerallyrequire thegenerationofthestatisticaldistributionofconsequencevalues overannualweatherdata(i.e.,weatheraveragedcalculations)

ATDmodelfidelityforprobabilisticconsequenceanalysisis traditionallyconsideredlesssignificantthanforemergency responsetoolsbecauseofthestatisticalnatureofprobabilistic consequenceanalyses However,questionsabouttheadequacyoftheGaussianplume modelforprobabilisticconsequenceanalyseshavebeenraised sinceatleastthe1980s Apreviousstudy(NUREG/CR6853)showedthatMACCSis generallywithinafactoroftwooftheNationalAtmospheric ReleaseAdvisoryCentersLagrangianOperationalDispersion Integrator(LODI)codeforweatheraveragedcalculations 10

2/12/2019 11 EvaluationAttributesforPotentialAlternativeATDCodes

  • Essentialfeatures

- TreatsLagrangianparticles

- Codeexecutablefilescanbedistributedtousers(eitherbySandia NationalLaboratoriesordirectlyfromthedeveloper)

- Sourcecodeavailableincasemodificationsarerequiredfor integration

- Codehasadequatequalityassurance(QA)

- Treatsthreedimensional(3D)windfield(griddeddata)

  • Desirablefeatures

- Canmodelbothpuffsandparticles

- Supportsavarietyofgriddedweatherdataformats

- Facilitatesgraphicalpostprocessing 11

2/12/2019 12 FeaturesofPotentialAlternativeATDCodes 12 HYSPLIT FLEXPART LODI CALPUFF RASCAL SCIPUFF Dispersion Lagrangian/

Gaussian Lagrangian Lagrangian/

Gaussian Gaussian Gaussian Gaussian Developer/

Primary Customers NOAA&Aus.

BureauofMet./

NASA, Universities, NorwegianInst.

forAirResearch/

Europe LLNL/

DOE TRC/

EPA PNNL/

NRC Titan/

DTRA, U.S.Army Weather Data Anygriddeddata WRF, MM5 WRF,ADAPT
CALMET, MM5,etc.

Surfacedata Surfaceor griddeddata Surface Roughness Userinputatany resolution,or 1degree Userinputatany resolution Userinputat any resolution Userinputat any resolution Userinputat any resolution Unknown Availability/

SourceCode NOAA/Yes NorwegianInst.

forAirResearch/

Europe/Yes Onlyvia Web/No TRC/No NRC/Yes Titan/No QA/Verifica tionand Validation Yes,open literature Openliterature Yes,open literature Yes Yes Yes Platforms Unix/Linux, PC,Mac Unix/Linux,PC Unix/Linux PC PC PC Graphics Yes Userplotsoutput Yes Yes Yes Yes Computer Time Canbehigh/

needsmultiple processors Canbehigh/

needsmultiple processors Canbehigh/

needs multiple processors Canbe high Moderate Canbehigh

2/12/2019 13 SelectedAlternativeATDCode:HYSPLIT SelectedforintegrationintoMACCSasamodernalternativetothe Gaussianplumesegmentmodel DevelopedandmaintainedbytheNationalOceanicandAtmospheric Administration(NOAA)AirResourcesLaboratory Cancomputetrajectoriesaswellascomplextransport,dispersion, chemicaltransformation,anddeposition Acceptsawidevarietyofpubliclyavailablegriddedmeteorologicaldata, includingbothNorthAmericanandglobaldatasets Modelsdispersionusingeitherpuffsorparticles:

[P]uffsexpanduntiltheyexceedthesizeofthemeteorologicalgridcell(either horizontallyorvertically)andthensplitintoseveralnewpuffs,eachwithitsshareofthe pollutantmass.*

[A]fixednumberofparticlesareadvectedaboutthemodeldomainbythemeanwind fieldandspreadbyaturbulentcomponent.*

13

2/12/2019 14 PotentialMACCSEnhancementsforNonLightWater Reactors(NonLWRs)

  • Evaluatingpotentialcodedevelopmentneedsfor designspecificissues

- Radionuclidescreening

- Radionuclidechemicalform

- Aerosolsize

- Aerosolshapefactor

- Radionuclideswithcomplextransportcharacteristics(e.g.,tritium)

  • Evaluatingpotentialcodedevelopmentneedsforsiterelated issues

- Nearfieldatmospherictransport

- Decontaminationmodeling 14

2/12/2019 15 NearFieldAtmosphericTransport MACCScurrentlyhasasimplevirtualsourcemodelforbuildingwake effects;itsuserguide(NUREG/CR6613)cautionsagainstusecloserthan 500meters LicenseesfornonLWRs(andsmallmodularreactors)desireasmaller emergencyplanningzoneandsiteboundarythanforlargeLWRs; therefore,bettermodelingofnearfieldphenomenamaybebeneficial 15 Schulman,L.L.,D.G.Strimaitis,andJ.S.Scire(2000).DevelopmentandEvaluationofthePRIMEPlumeRiseand BuildingDownwashModel,JournaloftheAir&WasteManagementAssociation,50:3,378-390 Windtunnelsimulationofstreamlinesnearacubicbuilding

2/12/2019 16 NearFieldAtmosphericTransport Variousoptionsforaddressing nearfieldATD

- ModificationstoGaussianplumesegment ATDmodel

- LagrangianparticletrackingATDusing simplified3Dwindfieldmodels

- LagrangianparticletrackingATDusingCFD modelingof3Dwindfield Considerationsforevaluatingoptions

- Extentofpracticalacceptanceintheuser community

- Simplicityofuse

- Computationalefficiency

- Costandtimeefficiency

- Accuracy

- Feasibilityforprobabilisticapplication 16 ExampleQUICURBsimulationofwind vectors ExampleQUICPLUMEsimulationofurban transportanddispersion QUICFactsheet,LosAlamosNationalLaboratory

2/12/2019 17 Summary

  • ThecurrentversionofMACCSisanefficientandhighly flexibleprobabilisticconsequenceanalysistoolwithmany userconfigurableoptions
  • Codeenhancementsarenearingcompletiontointroduce resultscomputedbyHYSPLITintoMACCS
  • Examiningtheapplicabilityofcurrentandpotential approachesformodelingnearfielddispersionphenomena wherebuildingwakesmaybeimportant 17

2/12/2019 18 ACRONYMSANDREFERENCES Acronyms ATD:AtmosphericTransportandDispersion FLEXPART:FLEXiblePARTicleDispersionModel HYSPLIT:HybridSingleParticleLagrangianIntegratedTrajectoryModel LODI:LagrangianOperationalDispersionIntegrator MACCS:MELCORAccidentConsequenceCodeSystem NOAA:NationalOceanicandAtmosphericAdministration QUIC:QuickUrban&IndustrialComplexDispersionModelingSystem References Chanin,D.,andM.L.Young(1998).CodeManualforMACCS2:Volume1,UsersGuide (NUREG/CR6613),SandiaNationalLaboratories,Albuquerque,NM,May1998 Molenkamp,C.R.,N.E.Bixler,C.W.Morrow,J.V.Ramsdell,Jr.,andJ.A.Mitchell(2004).Comparisonof AverageTransportandDispersionAmongaGaussian,aTwoDimensional,andaThreeDimensional Model(NUREG/CR6853),LawrenceLivermoreNationalLaboratory,Livermore,CA,October2004 Schulman,L.L.,D.G.Strimaitis,andJ.S.Scire(2000).DevelopmentandEvaluationofthePRIMEPlume RiseandBuildingDownwashModel,JournaloftheAir&WasteManagementAssociation,50:3,378-390 Snell,W.G.,andR.W.Jubach(1981).TechnicalBasisforRegulatoryGuide1.145,AtmosphericDispersion ModelsforPotentialAccidentConsequenceAssessmentsatNuclearPowerPlants(NUREG/CR2260),

NUSCorporation,Rockville,MD,October1981 18