ML16102A330

From kanterella
Jump to: navigation, search

Presentation Slides for 04042016 Public Meeting on Resolution of TVA Flood Model Issue
ML16102A330
Person / Time
Site: Sequoyah, Watts Bar Tennessee Valley Authority icon.png
Issue date: 04/04/2016
From: Wetzel B A
Tennessee Valley Authority
To:
Office of Nuclear Reactor Regulation
References
Download: ML16102A330 (28)


Text

Resolution of Flood Model IssueApril 4, 2016 Agenda*Purpose*Background*Flood Model Issue*Path Forward/Resolution

  • Scope of Work
  • ScheduleFLOOD MODEL ISSUE RESOLUTIONl 2 PurposeThe purpose of this meeting is to discuss: *Technical details of the recently discovered external flood model issue using the HEC-

RAS code*Impacts to TVA nuclear sites current licensing basis*Impacts to NTTF Rec 2.1 Flood Hazard Reevaluation Reports*Resolution of IssueFLOOD MODEL ISSUE RESOLUTIONl 3 Watershed ModelFLOOD MODEL ISSUE RESOLUTIONl 4 Background*Watts Bar/Sequoyah LAR submittal -

July/August 2012*Watts Bar LAR supplement submittal -September 30, 2014-HEC-RAS and updated dam stability criteria adopted for PMF-NRC issues Safety Evaluation Report -January 28, 2015FLOOD MODEL ISSUE RESOLUTIONl 5 Background*NTTF Recommendation 2.1 Flood Hazard Reevaluation Reports (FHRRs) submitted for Browns Ferry, Sequoyah and Watts Bar

-March 12, 2015-NRC issues Interim Staff Response to Reevaluated Flood Hazards for Browns Ferry, Sequoyah and Watts Bar -September 3, 2015-NRC issues Staff Assessment of Response to Reevaluated Flood Hazards for Watts Bar -

December 1, 2015FLOOD MODEL ISSUE RESOLUTIONl 6 Background*Model Issue Identified -November 6, 2015-Condition Report 1101784*Model Issue Confirmed -December 16, 2015-Condition Reports 1116461 and 1116506FLOOD MODEL ISSUE RESOLUTIONl 7 TVA Current Hydraulic ModelFLOOD MODEL ISSUE RESOLUTIONl 8*Traceability of inputs from the SOCH model

influenced the

decision to

maintain historical cross section locations in TVA's current

hydraulic

model.*TVA manually augmented ineffective flow

areas to account for

reach storage

between cross sections to

match published

reservoir storage curves.

Volume Computation DifferencesFLOOD MODEL ISSUE RESOLUTIONl 9*TVA Augmented Ineffective Flow Areas

& HEC-RAS Post

Processor calculates volume for left and right overbanks separately.*HEC-RAS internal computations

calculates volume using a simple arithmetic average of the right and left reach

lengths.*Volume differences are compounded as left and right overbank length variance is

increased.

Flood Model Issue*Difference in volume computations between -HEC-RAS internal processor>computes a simple arithmetic average of left and right overbank lengths-HEC-RAS post processor >treats each overbank (and overbank length) separately providing a prismatic representation of the volume of each overbank*Differences exacerbated by TVA approach in accounting for actual reservoir storage by augmenting ineffective flow areas at Cherokee and Douglas where there are significant differences in right and left bank lengthsResult: TVA's HEC-RAS model overestimates the amount of reservoir storage volume available at Cherokee and Douglas FLOOD MODEL ISSUE RESOLUTIONl 10 HEC-RAS Bug Report*After discovering differences in volume checks, TVA submitted a HEC-RAS code Bug Report to USACE Hydrologic Engineering Center*USACE confirmed HEC-RAS internal and post processors calculate volume differently. However, in typical applications of the HEC-RAS software, this is not expected to significantly change the results*USACE concludes the issue is "not a bug, but a misunderstanding of the HEC-RAS internal computations"*USACE expert agreed to revise the user's manual to provide clarification on the volume equation and is considering a HEC-RAS code revision sometime in the future that would add an alternative solver algorithm which aligns the internal to the post processor*USACE expert recognized adjustment of augmented ineffective flow areas as a viable approachFLOOD MODEL ISSUE RESOLUTIONl 11 Cherokee Example -Worst CaseFLOOD MODEL ISSUE RESOLUTIONl 12Due to TVA's placement of augmented ineffective flow areas on one side of the channel, the HEC-RAS internal averaging of storage volumes in the overbank areas and the biased sinuosity in these reservoirs, TVA's hydraulic model issue

has the most pronounced impacts at Cherokee and Douglas.

Reservoir Extent of Condition ReviewFLOOD MODEL ISSUE RESOLUTIONl 13Extent of condition review performed for all reservoirs by comparing internal computed to published

reservoir storage curves. All produced less than 2% difference, except Cherokee and

Douglas.

Impacts*Watts Bar Current Licensing Basis*No Impact to Sequoyah and Browns Ferry Current Licensing Basis (SOCH code based)*Sequoyah HEC-RAS code based LAR Supplement*Watts Bar, Sequoyah and Browns Ferry FHRRsFLOOD MODEL ISSUE RESOLUTIONl 14 Technical Approach for Resolution*Updated Precipitation Data-Similar methods as HMR with an update to data set and orographic effects along with use of current

technology*Revise Hydraulic Model-For the most impacted reservoirs (Cherokee and Douglas) use a more standardized modeling approach, i.e. additional cross sections with storage nodes for large embayments-For the remaining reservoirs, adjust the augmented ineffective flow areas using the volume averaging equations consistent with HEC-RAS internal

processorFLOOD MODEL ISSUE RESOLUTIONl 15 Hydraulic Model Revision

-Cherokee/DouglasFLOOD MODEL ISSUE RESOLUTIONl 16*Add additional cross sections

between existing

cross-sections and

use storage nodes

for large

embayments*Provides a more standardized

modeling approach

utilizing current

industry practices*More realistically represents

geography of

storage volume in

the reservoirStoragenodesusedforlargeembayment Hydraulic Model Revision

-Other AreasFLOOD MODEL ISSUE RESOLUTIONl 17*Adjust the augmented ineffective flow areas

using the volume averaging equations

from HEC-RAS internal processor*Apply to entire model with the exception of

Cherokee and Douglas reservoirs Issue ResolutionTVA has confidence in the model issue resolution*Alternate method produces comparable results*Calibration shows stage and discharge match well to historic storms*USACE expert recognized adjustment of augmented ineffective flow areas as a viable

approach*Additional external reviews will be utilizedFLOOD MODEL ISSUE RESOLUTIONl 18 Issue Resolution*Precipitation update-Preliminary studies show margin between current HMR 41 and updated data*TVA has high confidence that: -Model issue is corrected with planned resolution-Final PMF simulations with use of the updated precipitation and update to the hydraulic model will produce elevations less than those shown in current FHRRs FLOOD MODEL ISSUE RESOLUTIONl 19 Scope of Work*Mitigating Strategies Assessment-TVA plans to use March 12, 2015 FHRR flood model and results to evaluate mitigation strategies>Elevation and flow results are bounding with use of the conservative precipitation data in HMR 41>NRC interim staff review has been completed for Watts Bar, Sequoyah and Browns Ferry with no open issues-MSAs will be completed by December 2016FLOOD MODEL ISSUE RESOLUTIONl 20 Scope of Work*New report/calculations -Precipitation data to replace HMR-41-PMP development to determine most critical inflows*Calculation revisions-Main Stem and Tributary Geometry >Cherokee and Douglas hydraulic model geometry revised for additional cross-sections with storage nodes>Adjust augmented ineffective flow areas for the balance of the model using HEC-RAS internal processor equations-Main Stem and Tributary CalibrationsFLOOD MODEL ISSUE RESOLUTIONl 21 Scope of Work, continued*Calculation revisions, continued-Model Setup-Probable Maximum Flood Simulations>Significant increase in number of events to review to determine controlling simulation-Warning Time-Controlling FHRR Seismic Combination Simulation

-UncertaintiesFLOOD MODEL ISSUE RESOLUTIONl 22 Scope of Work, continuedFLOOD MODEL ISSUE RESOLUTIONl 23Totalest.>25calculationsInputDataDevelopment(est.4calculations)SeasonalityAnalysisAntecedentSoilMoisture(seasonallyadjusted)TemporalDistribution(early/median/late)PMPRainfall(est.>60events)Inflow&PMFDevelopment(est.>20calculations)ApplyProjectCandidateCriticalBasin(s)RainfallApplyCandidateTemporalDistributionsApplyAntecedentEventCalculateSurfaceRunoffRoutetoModelLimitsDistributeInflowsRunHECRAS(est.>100eventsruns)SummarizeResultsforSite(s)WorstCaseWSEWarningTime(est.2calculations)ReviewallPreviousRunsforWarningTimeDevelopAnyAdditionalLocalInflowsRunHECRASSummarizeResultsforSite(s)WorstCaseWTUncertainties(est.1calculations)DevelopandApplyNonlinearUHAdjustmenttoControllingEvent(s)DevelopGateOperabilityRuns(est.6HECRASruns)

Scope of Work, continued*Deliverables:-License Amendment Request >Sequoyah>Watts Bar-FHRR report revisions-Flooding Integrated Assessment FLOOD MODEL ISSUE RESOLUTIONl 24 Additional Licensing Basis Scope*Update FSAR Section 2.4.4 Potential Dam Failures (Seismically Induced)-Simulations in HEC-RAS vs. SOCH-Use JLD-ISG-2013-01 guidance>Update seismic dam stability evaluations to current methods >Use dam specific probabilistic earthquake vs. nuclear site deterministic earthquake>>10-4annual exceedance seismic hazard in place of SSE>>Half the 10

-4ground motion in place of OBE>Use deaggregation to evaluate the potential for multiple dam failuresFLOOD MODEL ISSUE RESOLUTIONl 25 Additional Licensing Basis Scope*Update FSAR Section 2.4.11 Low Water Considerations-Simulations in HEC-RAS vs. SOCH*With this additional change to the licensing basis, legacy SOCH code use will be eliminated for Sequoyah and Watts Bar.FLOOD MODEL ISSUE RESOLUTIONl 26 Draft ScheduleFLOOD MODEL ISSUE RESOLUTIONl 27 QuestionsFLOOD MODEL ISSUE RESOLUTIONl 28 Resolution of Flood Model IssueApril 4, 2016 Agenda*Purpose*Background*Flood Model Issue*Path Forward/Resolution

  • Scope of Work
  • ScheduleFLOOD MODEL ISSUE RESOLUTIONl 2 PurposeThe purpose of this meeting is to discuss: *Technical details of the recently discovered external flood model issue using the HEC-

RAS code*Impacts to TVA nuclear sites current licensing basis*Impacts to NTTF Rec 2.1 Flood Hazard Reevaluation Reports*Resolution of IssueFLOOD MODEL ISSUE RESOLUTIONl 3 Watershed ModelFLOOD MODEL ISSUE RESOLUTIONl 4 Background*Watts Bar/Sequoyah LAR submittal -

July/August 2012*Watts Bar LAR supplement submittal -September 30, 2014-HEC-RAS and updated dam stability criteria adopted for PMF-NRC issues Safety Evaluation Report -January 28, 2015FLOOD MODEL ISSUE RESOLUTIONl 5 Background*NTTF Recommendation 2.1 Flood Hazard Reevaluation Reports (FHRRs) submitted for Browns Ferry, Sequoyah and Watts Bar

-March 12, 2015-NRC issues Interim Staff Response to Reevaluated Flood Hazards for Browns Ferry, Sequoyah and Watts Bar -September 3, 2015-NRC issues Staff Assessment of Response to Reevaluated Flood Hazards for Watts Bar -

December 1, 2015FLOOD MODEL ISSUE RESOLUTIONl 6 Background*Model Issue Identified -November 6, 2015-Condition Report 1101784*Model Issue Confirmed -December 16, 2015-Condition Reports 1116461 and 1116506FLOOD MODEL ISSUE RESOLUTIONl 7 TVA Current Hydraulic ModelFLOOD MODEL ISSUE RESOLUTIONl 8*Traceability of inputs from the SOCH model

influenced the

decision to

maintain historical cross section locations in TVA's current

hydraulic

model.*TVA manually augmented ineffective flow

areas to account for

reach storage

between cross sections to

match published

reservoir storage curves.

Volume Computation DifferencesFLOOD MODEL ISSUE RESOLUTIONl 9*TVA Augmented Ineffective Flow Areas

& HEC-RAS Post

Processor calculates volume for left and right overbanks separately.*HEC-RAS internal computations

calculates volume using a simple arithmetic average of the right and left reach

lengths.*Volume differences are compounded as left and right overbank length variance is

increased.

Flood Model Issue*Difference in volume computations between -HEC-RAS internal processor>computes a simple arithmetic average of left and right overbank lengths-HEC-RAS post processor >treats each overbank (and overbank length) separately providing a prismatic representation of the volume of each overbank*Differences exacerbated by TVA approach in accounting for actual reservoir storage by augmenting ineffective flow areas at Cherokee and Douglas where there are significant differences in right and left bank lengthsResult: TVA's HEC-RAS model overestimates the amount of reservoir storage volume available at Cherokee and Douglas FLOOD MODEL ISSUE RESOLUTIONl 10 HEC-RAS Bug Report*After discovering differences in volume checks, TVA submitted a HEC-RAS code Bug Report to USACE Hydrologic Engineering Center*USACE confirmed HEC-RAS internal and post processors calculate volume differently. However, in typical applications of the HEC-RAS software, this is not expected to significantly change the results*USACE concludes the issue is "not a bug, but a misunderstanding of the HEC-RAS internal computations"*USACE expert agreed to revise the user's manual to provide clarification on the volume equation and is considering a HEC-RAS code revision sometime in the future that would add an alternative solver algorithm which aligns the internal to the post processor*USACE expert recognized adjustment of augmented ineffective flow areas as a viable approachFLOOD MODEL ISSUE RESOLUTIONl 11 Cherokee Example -Worst CaseFLOOD MODEL ISSUE RESOLUTIONl 12Due to TVA's placement of augmented ineffective flow areas on one side of the channel, the HEC-RAS internal averaging of storage volumes in the overbank areas and the biased sinuosity in these reservoirs, TVA's hydraulic model issue

has the most pronounced impacts at Cherokee and Douglas.

Reservoir Extent of Condition ReviewFLOOD MODEL ISSUE RESOLUTIONl 13Extent of condition review performed for all reservoirs by comparing internal computed to published

reservoir storage curves. All produced less than 2% difference, except Cherokee and

Douglas.

Impacts*Watts Bar Current Licensing Basis*No Impact to Sequoyah and Browns Ferry Current Licensing Basis (SOCH code based)*Sequoyah HEC-RAS code based LAR Supplement*Watts Bar, Sequoyah and Browns Ferry FHRRsFLOOD MODEL ISSUE RESOLUTIONl 14 Technical Approach for Resolution*Updated Precipitation Data-Similar methods as HMR with an update to data set and orographic effects along with use of current

technology*Revise Hydraulic Model-For the most impacted reservoirs (Cherokee and Douglas) use a more standardized modeling approach, i.e. additional cross sections with storage nodes for large embayments-For the remaining reservoirs, adjust the augmented ineffective flow areas using the volume averaging equations consistent with HEC-RAS internal

processorFLOOD MODEL ISSUE RESOLUTIONl 15 Hydraulic Model Revision

-Cherokee/DouglasFLOOD MODEL ISSUE RESOLUTIONl 16*Add additional cross sections

between existing

cross-sections and

use storage nodes

for large

embayments*Provides a more standardized

modeling approach

utilizing current

industry practices*More realistically represents

geography of

storage volume in

the reservoirStoragenodesusedforlargeembayment Hydraulic Model Revision

-Other AreasFLOOD MODEL ISSUE RESOLUTIONl 17*Adjust the augmented ineffective flow areas

using the volume averaging equations

from HEC-RAS internal processor*Apply to entire model with the exception of

Cherokee and Douglas reservoirs Issue ResolutionTVA has confidence in the model issue resolution*Alternate method produces comparable results*Calibration shows stage and discharge match well to historic storms*USACE expert recognized adjustment of augmented ineffective flow areas as a viable

approach*Additional external reviews will be utilizedFLOOD MODEL ISSUE RESOLUTIONl 18 Issue Resolution*Precipitation update-Preliminary studies show margin between current HMR 41 and updated data*TVA has high confidence that: -Model issue is corrected with planned resolution-Final PMF simulations with use of the updated precipitation and update to the hydraulic model will produce elevations less than those shown in current FHRRs FLOOD MODEL ISSUE RESOLUTIONl 19 Scope of Work*Mitigating Strategies Assessment-TVA plans to use March 12, 2015 FHRR flood model and results to evaluate mitigation strategies>Elevation and flow results are bounding with use of the conservative precipitation data in HMR 41>NRC interim staff review has been completed for Watts Bar, Sequoyah and Browns Ferry with no open issues-MSAs will be completed by December 2016FLOOD MODEL ISSUE RESOLUTIONl 20 Scope of Work*New report/calculations -Precipitation data to replace HMR-41-PMP development to determine most critical inflows*Calculation revisions-Main Stem and Tributary Geometry >Cherokee and Douglas hydraulic model geometry revised for additional cross-sections with storage nodes>Adjust augmented ineffective flow areas for the balance of the model using HEC-RAS internal processor equations-Main Stem and Tributary CalibrationsFLOOD MODEL ISSUE RESOLUTIONl 21 Scope of Work, continued*Calculation revisions, continued-Model Setup-Probable Maximum Flood Simulations>Significant increase in number of events to review to determine controlling simulation-Warning Time-Controlling FHRR Seismic Combination Simulation

-UncertaintiesFLOOD MODEL ISSUE RESOLUTIONl 22 Scope of Work, continuedFLOOD MODEL ISSUE RESOLUTIONl 23Totalest.>25calculationsInputDataDevelopment(est.4calculations)SeasonalityAnalysisAntecedentSoilMoisture(seasonallyadjusted)TemporalDistribution(early/median/late)PMPRainfall(est.>60events)Inflow&PMFDevelopment(est.>20calculations)ApplyProjectCandidateCriticalBasin(s)RainfallApplyCandidateTemporalDistributionsApplyAntecedentEventCalculateSurfaceRunoffRoutetoModelLimitsDistributeInflowsRunHECRAS(est.>100eventsruns)SummarizeResultsforSite(s)WorstCaseWSEWarningTime(est.2calculations)ReviewallPreviousRunsforWarningTimeDevelopAnyAdditionalLocalInflowsRunHECRASSummarizeResultsforSite(s)WorstCaseWTUncertainties(est.1calculations)DevelopandApplyNonlinearUHAdjustmenttoControllingEvent(s)DevelopGateOperabilityRuns(est.6HECRASruns)

Scope of Work, continued*Deliverables:-License Amendment Request >Sequoyah>Watts Bar-FHRR report revisions-Flooding Integrated Assessment FLOOD MODEL ISSUE RESOLUTIONl 24 Additional Licensing Basis Scope*Update FSAR Section 2.4.4 Potential Dam Failures (Seismically Induced)-Simulations in HEC-RAS vs. SOCH-Use JLD-ISG-2013-01 guidance>Update seismic dam stability evaluations to current methods >Use dam specific probabilistic earthquake vs. nuclear site deterministic earthquake>>10-4annual exceedance seismic hazard in place of SSE>>Half the 10

-4ground motion in place of OBE>Use deaggregation to evaluate the potential for multiple dam failuresFLOOD MODEL ISSUE RESOLUTIONl 25 Additional Licensing Basis Scope*Update FSAR Section 2.4.11 Low Water Considerations-Simulations in HEC-RAS vs. SOCH*With this additional change to the licensing basis, legacy SOCH code use will be eliminated for Sequoyah and Watts Bar.FLOOD MODEL ISSUE RESOLUTIONl 26 Draft ScheduleFLOOD MODEL ISSUE RESOLUTIONl 27 QuestionsFLOOD MODEL ISSUE RESOLUTIONl 28