ML19161A239
| ML19161A239 | |
| Person / Time | |
|---|---|
| Site: | Watts Bar, Sequoyah  |
| Issue date: | 06/10/2019 |
| From: | Tennessee Valley Authority |
| To: | Andrew Hon Plant Licensing Branch II |
| Hon A, | |
| References | |
| Download: ML19161A239 (53) | |
Text
Watts Bar and Sequoyah Nuclear Plants Pre-Submittal Meeting for Proposed License Amendments Hydrological Analysis Updates June 10, 2019
Introduction - Kim Hulvey Background and History of Sequoyah Nuclear Plant (SQN) and Watts Bar Nuclear Plant (WBN) Hydrology Issues - Dennis Lundy Probable Maximum Precipitation (PMP) Update - Dennis Lundy Translation of Gridded Rainfall Data to Inflows in the Simulation Model - Stu Henry WBN and SQN License Amendment Requests (LAR) - Key Updates - Roger Scott Open Issue to WBN and SQN LARs - Roger Scott Schedule for Submittal - Roger Scott Summary - Kim Hulvey Closed Portion - Dennis Lundy Agenda l 2
TVA Watershed l 3 Watershed is complex, draining over 20,000 square miles above Chickamauga Dam Major Tributaries French Broad, Little Tennessee and Hiwassee Rivers originating in the Blue Ridge mountains northeast and southeast of the plant site Holston and Clinch Rivers originating in Virginia northeast of the plant site TVA River Management provides operational control of the reservoirs in an integrated fashion Flood Control - Water Quality Navigation - Water Supply Hydropower - Recreation
=
Background===
Three Notice of Violations (NOVs) issued March 2008 on TVA hydrologic analysis following Bellefonte Nuclear Plant Combined License Application audit.
WBN/SQN 2012 LAR submittals to address NOVs.
NRC audit concerns with dam stability evaluations.
WBN 2014 Supplement to address dam stability.
TVA identifies an issue with the HEC-RAS model.
SQN 2015 Supplement to address dam stability placed on hold.
l 4
Background - HEC-RAS Volume Error Fall 2015 - TVA identified an error in HEC-RAS unsteady flow model used for the WBN 2014 LAR supplement and the proposed SQN 2015 LAR supplement flood simulation models l 5
- TVA Augmented Ineffective Flow Areas & HEC-RAS Post Processor calculates volume for left and right overbanks separately.
- HEC-RAS internal computation calculates volume using a simple arithmetic average of the right and left reach lengths.
Background - HEC-RAS Volume Error (cont)
TVAs placement of augmented ineffective flow areas on one side of channel and reservoir sinuosity resulted in significant overestimation of available flood storage in the Cherokee (-19.5%) and Douglas (-7.2%) reservoirs.
l 6
HEC-RAS Error Resolution Approach Resolution Approach for WBN and SQN Submit LAR for WBN
> Update HEC-RAS model overbank storage to be consistent with HEC-RAS internal algorithm for storage volume calculation
> Update PMP to use TVA-Topical Report - AWA16-A
> Replace SOCH model with HEC-RAS model for Seismic - Flood, Warning Time, and Loss of Downstream Dam analyses
> Update Seismic-Flood analysis to JLD-ISG-2013-01 guidance Submit Updated LAR (replacing August 2012 LAR) for SQN
> Replace SOCH PMF simulation model with HEC-RAS
> Update dam stability determinations to TVA Dam Safety standards
> Implement stream course model changes defined in SQN-TS-12-02
> Implement same changes as described above for WBN l 7
PMP: Topical Report TVA-NPG-AWA16-A Approval Chronology Topical Report TVA-NPG-AWA16, Revision 0 submitted to NRC on September 20, 2016 TVA revised TVA-NPG-AWA16 to Revision 1 to address NRC RAIs on June 22, 2018 NRC issued a Safety Evaluation Report on March 18, 2019 TVA submitted TVA-NPG-AWA16-A May 22, 2019 Topical PMP Defines gridded rainfall data for three storm types (local, general, and tropical) over TVA drainage basin PMP storm characterization l 8
Application of Gridded Rainfall Stuart Henry - Barge Design Solutions June 10, 2019
TVA Projects and River Network Model l 10
TVA Projects and River Network Model l 11
TVA Model Sub-Basins l 12
Gridded Rainfall Network l 13
Areal Distribution of Gridded Rainfall GIS tool provided in Topical Report (final SE ML19010A274)
Tool retrieves transpositionable event data for selected storm type at the point under review.
Interpolates depth from each event specific Depth-Area-Duration curve for selected area and durations.
Applies point specific adjustments to selected point value.
Reviews all transpositionable events and selects the maximum event depth found at point under review for that duration.
l 14
Areal Distribution of Gridded Rainfall l 15
Areal Distribution of Gridded Rainfall l 16
Areal Distribution of Gridded Rainfall l 17
Areal Distribution of Gridded Rainfall Methodology Similarities Both apply depth using GIS surface.
Volumes are calculated within irregular sub-basin boundary areas and applied uniformly over each sub-basin.
Methodology Differences Historical method prescribes storm event rainfall pattern regardless of location characteristics.
Historical patterns do not provide means for analysis of smaller area PMP embedded within larger area PMP.
Gridded rainfall application based on allowable event transposition adjusted for specific location and elevation.
l 18
Areal Distribution of Gridded Rainfall Gridded Rainfall Advantages Evaluated historical storm event data through 2014.
Avoids subjectivity in over/under application of rainfall on different sized watersheds based on assumed pattern.
Provides more detailed analysis by application of PMP rainfall based on transpositionable events to specific points.
Allows different project PMP comparison beyond project watershed for more effective review of embedded events.
Facilitates application of rainfall depth for embedded PMP events.
l 19
Areal Distribution of Gridded Rainfall Selection of a given Area of Interest (AOI) above a project defines the location of PMP depth from combined grid point DAD curves.
DAD curve is area specific and does not prescribe rainfall outside AOI.
Due to possibility of postulated failures during project PMPs, must assess impact of embedded PMP events.
Analysis of embedded PMP events requires development of a PMP event nesting methodology.
Objective:
Application of PMP rainfall volume in an AOI that meets the defined grid point DAD curve data for PMP volume.
l 20
Areal Distribution of Gridded Rainfall Due to number of possible combinations for 40 model sub-basins above Chickamauga simplifying assumptions were necessary for nesting criteria.
AOIs may be defined only by selecting adjacent sub-basins.
AOIs may be defined only by the set of sub-basins between TVA river projects.
l 21
Areal Distribution of Gridded Rainfall Nesting Criteria AOIs split into successive primary and secondary areas.
Primary AOI scenario may be a single projects set of one or more sub-basins or any combination of adjacent projects set of sub-basins meeting simplifying assumptions.
Secondary (larger) AOIs must include all primary AOI sub-basins (i.e., embedded PMP).
Simulations end with the boundary condition AOI at Wheeler with upstream sub-basins included.
Adjustments made based on individual sub-basin at each duration as part of total AOI PMP volume.
l 22
Areal Distribution of Gridded Rainfall l 23 FLTE_05
Areal Distribution of Gridded Rainfall l 24 CHH_02
Areal Distribution of Gridded Rainfall l 25 Example:
Areal Distribution of Gridded Rainfall l 26 FP = Fort Patrick Henry Sub-basin #
CR - Cherokee 9
10 11 12 13 14&15 for AOI Sub-basin Area (sq.mi.) 703.25 468.25 667.67 62.77 668.89 854.63 Total Area (sq.mi.)
PMP Depth (inches) 72hr Duration FP PMP Depths (inches) 14.30 18.74 14.34 15.07 1,901.94 15.43 72hr Duration CR PMP Depths (inches) 13.75 17.92 13.78 14.54 13.25 14.76 3,425.46 14.49 Raw 72hr Duration PMP at CR (inches) 14.30 18.74 14.34 15.07 13.25 14.76 3,425.46 14.84 Adjustment Factor = 0.95 72hr Duration Adjusted PMP at CR (inches) 14.30 18.74 14.34 15.07 12.52 13.95 3,425.46 14.49 Sub-basin #
9 10 11 12 13 14&15 Sub-basin Area (sq.mi.)
703.25 468.25 667.67 62.77 668.89 854.63 Volume (in sq.mi.-in.)
72hr Duration FP PMP Depths (inches) 14.30 18.74 14.34 15.07 29,350.73 72hr Duration CR PMP Depths (inches) 13.75 17.92 13.78 14.54 13.25 14.76 49,645.21 Unadjusted 72hr Duration CR Depths (inches) 13.25 14.76 21,470.85 Adjustment Factor (49645.2055 - 29350.7335) / 21470.8532 = 0.95
Areal Distribution of Gridded Rainfall Comparison with Previous HMR Based PMPs l 27 System 72hr PMP Above Projects (in inches)
Project Area (sq.mi.) 21400DSM 7980BGM FLTE_05 FNH_04 CHH_02 Douglas 4,543 17.33 17.23 16.32 12.53 16.32 Cherokee 352 13.31 18.75 11.91 11.48 11.56 FortLoudoun 9,549 16.84 19.04 13.76 11.99 13.76 Fontana 1,571 19.77 13.16 15.80 23.13 15.80 Tellico 2,626 20.01 14.20 14.60 21.03 14.60 Ft.LoudounTellico 12,175 17.52 17.99 13.94 13.94 13.94 Norris 2,913 13.64 16.49 9.41 9.41 9.06 WattsBar 17,294 16.78 17.19 12.52 12.52 12.43 AboveHiwassee 968 18.45 11.42 10.77 10.77 12.33 AboveApalachia 1,018 18.43 11.47 10.75 10.75 12.23 AboveBlueRidge 232 21.32 9.45 11.54 11.54 14.75 Chickamauga 20,781 16.73 16.26 12.06 12.06 12.06
Application of Gridded Rainfall Nested total rainfall depths developed for 1-, 6-, 12-, 24-, 48-and 72-hour durations for application to each TVA sub-basin.
Previously used HMR guidance 3-day antecedent, 3-day dry period and 3-day main sequence was adopted.
Total rainfall converted to 1-hour increments for each TVA sub-basin.
Antecedent scaled to 40% of PMP applied in same areal distribution.
l 28
Temporal Distribution of Gridded Rainfall Topical report historical event data used to develop temporal distributions.
Event data were statistically analyzed to develop storm type specific temporal distributions.
Four general event and one tropical event temporal distributions were developed.
One antecedent event distribution was applied.
l 29
Temporal Distribution of Gridded Rainfall l 30
Seasonality of Simulated Events Topical Report: three storm types; local, general and tropical Storm types are independent (i.e., do not mix)
Local limited to watersheds 500 sq.-mi. or less Tropical and general event types reviewed for seasonality Reviewed appropriate storm type run over weeks 11-23 Results used as guide for PMF simulation start dates Losses applied seasonally using Antecedent Precipitation Index (API) methodology.
l 31
Areal Distribution of Gridded Rainfall l 32 Duration(inhours) 1 6
12 18 24 48 72 Cum.Depth(ininches) 0.79 3.02 5.57 6.68 8.01 11.51 12.06 Increment(inhours) 1 26 712 1318 1924 2548 4972 Inc.Depth(ininches) 0.79 0.44 0.43 0.18 0.22 0.15 0.02 PMPRainfallfor20,780.8sq.mi.WatershedaboveChickamauga
Simulations Using Gridded Rainfall Summary Gridded rainfall allows more detailed analysis of PMP events.
89 documented simulations were performed with PMP depths at and above Chickamauga.
Simulations are directly comparable PMF candidates at locations where applied rainfall meets DAD volume.
Additional follow-up simulations were performed to analyze seasonality and warning time.
l 33
WBN and SQN LAR - Key Updates Roger Scott - TVA Licensing June 10, 2019
WBN and SQN LAR - Key Updates WBN and SQN UFSARs Section 2.4 being changed to reflect updates to the hydrologic analysis, including:
Change to PMP bases document and process of determining the controlling PMP storms.
Change to methodology for determining PMF.
Changes to correct HEC-RAS model geometry overbank volume error.
Update of overland wind speed data.
Changes to the seismically-induced dam failure flooding analysis.
Change to warning time hydrologic basis for rainfall on rivers and streams and seismically-induced dam failure floods.
l 35
WBN and SQN LAR - Key Updates Changes to PMP bases document and process of determining the controlling PMP storms Topical Report TVA-NPG-AWA16-A was approved for use in determining the PMP for the WBN and SQN sites NRCs Safety Evaluation Report dated March 18, 2019.
The design basis PMP used in the SQN and WBN site local intense precipitation (LIP) drainage analyses and the rivers and streams rainfall flooding hydrological analyses are updated to use Topical Report TVA-NPG-AWA16-A.
Due to the Topical Report gridded rainfall data format, the method used to determine the rainfall in PMP storm development has been updated.
l 36
WBN and SQN LAR - Key Updates Change to methodology for determining PMF The changes to the SQN UFSAR update the discussion of PMF on streams and rivers to reflect updated methodologies, such as the HEC-HMS and HEC-RAS software for elements of the hydrologic analysis for determining the PMF for streams and rivers for SQN.
The NRC approved the changes to the WBN UFSAR to update the hydraulic analysis methodology, including use of the HEC-HMS and HEC-RAS, on January 28, 2015.
Updates other analyses to HEC-RAS (warning time, seismically-induced dam failures, and loss of downstream dams).
l 37
WBN and SQN LAR - Key Updates Changes to correct HEC-RAS model geometry overbank volume error The HEC-RAS model geometry above Douglas Dam on the French Broad and Nolichucky Rivers and above Cherokee Dams on the Holston River and the downstream portion of the South Fork Holston River has been updated to more accurately characterize overbank storage through the additional cross-sections and storage areas, connected by lateral structures.
Overbank storage volume in the other modeled reservoirs have been adjusted by using the average reach length in the augmented ineffective flow area overbank storage volume calculation.
This update aligns the reservoir storage determined by the internal HEC-RAS software algorithm with the published reservoir storage volumes throughout the HEC-RAS unsteady flow model.
l 38
WBN and SQN LAR - Key Updates Update of overland wind speed data Meteorological wind speed data is being updated to include more current wind data and data sources from airports surrounding each site. As a result of this change, critical fetch and wind wave effects are updated for critical structures.
l 39
WBN and SQN LAR - Key Updates Changes to the seismically-induced dam failure flooding analysis Design basis at each site is being changed to reflect the guidance provided in JLD-ISG-2013-01, Guidance for Assessment of Flooding Hazards Due to Dam Failure, for the seismic dam failures combined with flooding analysis.
The methodology applied in the development of 25-year flood inflows is revised to align with more recent work performed for the TVA Clinch River Early Site Permit Application. Inflows for a 500-year storm are created to replace the one-half PMF inflows used in combination with the less severe seismic hazard, as defined in JLG-ISG-2013-01.
l 40
WBN and SQN LAR - Key Updates Changes to the seismically-induced dam failure flooding analysis (continued)
The stability analysis methods and the factors of safety required for stability of concrete and embankment dam structures are being updated to align with current TVA Dam Safety design.
The hydrological unsteady flow model used in the simulation of seismically induced dam failures combined with flooding is being updated to the industry standard HEC-RAS modeling software.
The model is also being updated and expanded to be consistent with the stream course model used in the evaluation of flooding due to rivers and streams rainfall.
l 41
WBN and SQN LAR - Key Updates Change to warning time hydrologic basis for rainfall on rivers and streams and seismically-induced dam failure floods The PMP used in the warning time analysis for each site is updated to Topical Report TVA-NPG-AWA16-A.
The target site elevation used in the development of flood warning notifications is updated to plant grade at each site.
The warning scheme is updated to utilize average rain on the ground on the basins above Chickamauga Dam as the trigger for the TVA River Management group to initiate flood modeling for potential impacts to the site.
l 42
WBN and SQN LAR - Key Updates Change to warning time hydrologic basis for rainfall on rivers and streams and seismically-induced dam failure floods (continued)
Average rain on the ground above Chickamauga is used, along with weather forecasts and flood steam course modeling, to provide Stage I and Stage II protective warnings.
The use of the revised Stage I and Stage ll action levels does not reduce the effectiveness of the warning plan, as the minimum time to prepare for operation in the flood mode is not changed at either site.
For WBN, the warning time analysis and response plan for seismically induced dam failure flooding is removed, because the resultant flooding from these events does not reach plant grade.
l 43
Open Issue Watauga Dam - Top of Dam Elevation Road atop Watauga Dam has settled below the top of dam elevation in hydrologic re-analysis calculation Does not impact current hydrologic analysis Resolution of this issue is required prior to use of the updated design basis for SQN and WBN Repair status will be provided in the 6-month status update (scheduled for January 1, 2020) l 44
Schedule for Submittal The SQN and WBN LARs are currently scheduled to be submitted to the NRC by mid-July 2019.
l 45
Summary l 46 The PMP for WBN and SQN is being updated to reflect use of NRC-approved Topical Report TVA-NPG-AWA16-A.
Due to the Topical Report gridded rainfall data format, the method used to determine the rainfall in PMP storm development is being updated.
TVA plans to submit WBN and SQN LARs to request changes to hydrologic analyses in mid-July 2019.
TVA will update the NRC on status of the Open Issue by January 1, 2020.
Closed Session Slides l 48
Low Margin Dam Failures - Controlling Simulation l 49 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390
Watts Bar Flood Hydrograph - Controlling PMF Simulation l 50 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390
Sequoyah Flood Hydrograph - Controlling PMF Simulation l 51 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390
Watts Flood Hydrograph - Controlling Seismically Induced Flooding Simulation l 52 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390
Sequoyah Flood Hydrograph - Controlling Seismically Induced Flooding Simulation l 53 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390 Security Sensitive - Withheld from Public Disclosure Under 10 CFR 2.390