ML17053A023

Slides for Meeting with Tennessee Valley Authority to Discuss HEC-RAS Flood Model Project: Seismic Dam Failure Methodology

ML17053A023
Person / Time
Site:	Sequoyah
Issue date:	02/22/2017
From:	Tennessee Valley Authority
To:	Andrew Hon Office of Nuclear Reactor Regulation
Hon A, NRR/DORL/LPL2-2, 415-8480
References
Download: ML17053A023 (30)

Text

HEC-RAS Flood Model Project:

Seismic Dam Failure Methodology February 22, 2017

Outline

• Background Information

• Seismic Hazard

• Dam Stability Analyses

• Hydrologic & Hydraulic Models

• Single Dam Failures

• Multiple Dam Failures

• Results

• Schedule HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 2

=

Background===

In the March 12, 2012, 50.54(f) letter submitted to licensees, re-evaluation of flood hazard is requested. Interim Staff Guidance, JLD-ISG-2013-01, provides guidance to perform flood hazard due to dam failure.

• TVA submits Updated Flood Hazard March 12, 2015.

• TVA re-evaluates stability of 27 upstream dams to current TVA dam safety standards for flood and seismic loading as outlined in JLD-ISG-2013-01.

• Assumptions of failure are made for additional upstream dams (conservative assumption for the flood hazard at nuclear power plants)

HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 3

Updating Licensing Basis Proposed Update Current Basis HEC-RAS model SOCH model Current dam stability Deterministic nuclear methodology using power plant seismic input probabilistic site-specific for dam stability seismic hazard analysis evaluation for dam stability evaluation Multiple dam failure considerations using Multiple dam failure attenuation equations considerations through deaggregation of seismic hazard Use of the completed updated flood hazard for seismic dam failures aligns with current standards for both dam stability and hydraulic analysis. Planned hydrology license amendment requests will require review criteria other than SRP.

HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 4

Overview of Methodology Detailed Analysis per JLD-ISG-2013-01 When dam failure is not assumed, detailed analysis of the dams are made.

Detailed analysis on 27 upstream dams/dikes Seismic Source Model, CEUS-SSC [EPRI/NRC/DOE 2012]

Ground Motion Attenuation, Ground Motion Prediction Equations [EPRI 2004/2006]

Site Amplification, Site Response Analysis included detailed field investigation and laboratory test reports Detailed Analysis of Seismic Capacity of the Dam, Post-seismic dam stability with detailed field investigation and laboratory reports [TVA Dam Safety Criteria]

HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 5

Probabilistic Seismic Hazard Analysis 6

Inputs 5

Seismic source model: 2012 EPRI/DOE/NRC CEUS-SSC model 4

Ground motion prediction models: EPRI (2004, 2006) Series 1 3

Series 2 Series 3 2

Shown to the right are 10,000 year return period, 5% damped mean UHRS 1 No RLMEs located within about 200 km of the dam sites; however New Madrid fault 0 system, Reelfoot Rift, and Charleston Category 1 Category 2 Category 3 Category 4 source added to model HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 6

Seismic Source Catalog Updates Seismic Source Recently for seismic PRA work, catalog updates for the region have been completed Plot to the right shows predicted vs. observed earthquake counts for catalog updates to 4/30/2016 HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 7

Ground Motion Model Update Comparison Ground Motion Models EPRI 2013 vs. EPRI 2004/2006 Recently for seismic PRA work, ground motion models comparisons have been completed HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 8

SSE comparison 1.6 1.4 1.2 1

Acceleration, g SSE 0.8 Douglas E-4 Guntersville E-4 0.6 South Holston SD1 E-4 0.4 Wheeler E-4 0.2 0

0.1 1 10 100 Frequency, Hz HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 9

OBE comparison 0.8 0.7 0.6 0.5 Acceleration, g OBE 0.4 Douglas 1/2 E-4 Guntersville 1/2 E-4 0.3 South Holtson SD1 1/2 E-4 0.2 Wheeler 1/2 E-4 0.1 0

0.1 1 10 100 Frequency, Hz HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 10

Controlling Earthquake RG 1.208, Appendix D with following exceptions:

• Deagreggation at 1 and 10 Hz instead of 1 to 2.5 Hz and 5 to 10 Hz

- 10 Hz spectral frequency of interest, concrete dam

- 1 Hz spectral frequency of interest, earthen dam

• Modal magnitude and distance instead of mean magnitude and distance

- 1 Hz controlling earthquake, distant large magnitude New Madrid

- 10 Hz controlling earthquake, local moderate event in Eastern Tennessee Seismic Zone HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 11

Time histories Time History Development 1 Hz Controlling 1 1999 Chi Chi, Taiwan Earthquake 6 sets of time histories developed 2 2002 Denali, AK 3 1992 Landers, CA 3 sets for 1 Hz 3 sets for 10 Hz 4 1987 Superstition Hills, CA Spectrally matched per SRP 3.7.1 criteria 10 Hz Controlling Earthquake 5 2011 Mineral, VA Time History Use

1. Dynamic stability analyses of dams if 6 1988 Saguenay, Quebec founded on hard rock, or

2. Site response analysis to develop ground motions at foundation level beneath dam if dam not founded on hard rock HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 12

Dam Site Response Analysis

• Site-specific exploration and testing completed

• Time history sets developed for each dam at frequency of interest

• Response analysis software used to perform the wave propagation analysis to define seismic response at dam foundation for condition other than hard rock

• V/H ratios from NUREG-6728, Table 4-5

• Foundations considered massless but with stiffness

• Hydrodynamic masses representing the upstream normal pool reservoir are computed and applied to upstream face HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 13

Dam Stability Analysis Concrete Updated Current Licensing Basis 2-D or 3-D finite element analyses of dams 2-D rigid body analyses of critical locations prior including modifications to current modifications Headwater and tailwater - normal pool Headwater and tailwater - 25 year flood and 1/2 PMF flood without tailwater reduction Dynamic analysis of seismic loading and reservoir effects performed to define damaged Pseudo-static analysis with hydrodynamic state pressures

• Cohesion at rock-concrete interface neglected,

• Silt pressure not considered conservatively

• Cohesion at rock-concrete interface applied

• Degradation of friction angle

• Friction angle, no degradation

• Silt pressure effect included

• Increased uplift due to degraded drains and

• Increased uplift due to degraded drains and base base cracking not considered cracking Acceptance criteria - sliding FS=1.0 and If damaged state, no rock-concrete tensile resultant within dam base strength Acceptance criteria - sliding FS=1.3 and Resultant within dam base HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 14

Dam Stability Analysis Earthen Updated Current Licensing Basis Pseudo-static method of seismic coefficient Pseudo-static, standard slip circle method generally used Headwater and tailwater - normal pool Headwater and tailwater - 25 year flood and 1/2 Dynamic analysis for applied ground motion PMF without tailwater reduction with in-situ soil properties Amplification assumed based on nuclear plant Evaluation of liquefaction potential and studies (dam dynamic soil properties not reservoir effects available)

Post-seismic stability

• Stability FS 1.1

• If FS < 1.1, seismic deformation analysis is conducted

• Seismic deformation limited to < 2 feet and 1/2 filter thickness

• Seepage evaluations included HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 15

Hydrologic & Hydraulic Models Updated Current Licensing Basis Hydrologic Hydrologic Precipitation - 25 year flood updated with Precipitation - 1/2 PMF (Vintage HMR) additional gauge data Precipitation - 500 year flood Hydraulic Hydraulic Simulated Open Channel Hydraulics (SOCH)

HEC-RAS model model Debris fields not utilized Debris fields utilized HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 16

Simplified Screening

• Screening approach JLD-ISG-2013-01, Section 3

• Inconsequential dam failure combinations assessment based on reservoir volumes

• Entire volume transferred downstream to next reservoir, carried throughout system

• Outlet structures of stable dams assumed closed HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 17

Screening - Watts Bar HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 18

Screening - Sequoyah HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 19

Screening Results

• Possible Consequential Results (Single Failure)

- South Holston

- Watauga

- Chatuge

- Cherokee

- Fontana

- Norris

- Nottely

- Watts Bar HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 20

Seismic Single Dam Failure Updated Current Licensing Basis One half 10-4 AFE ground motion + 500 year flood OBE + 1/2 PMF Watts Bar Norris Chatuge Cherokee Douglas 10-4 AFE ground motion + 25 year flood Fontana No additional; above cases govern SSE + 25 year flood Watts Bar HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 21

Considerations of Multiple Dam Failures From Single Seismic Event (Updated)

JLS-ISG-2013-01 Section 5.5 Seismic influence refinement completed through deaggregation of hazard.

Frequency of Interest Concrete dams, 10 Hz Earthen dams, 1 Hz HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 22

Seismic Influence Deaggregation 10 Hz at 10-4 AFE HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 23

Seismic Influence Deaggregation 10 Hz at one half ground motion 10-4 AFE HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 24

Seismic Influence Deaggregation 1 Hz at 10-4 AFE & 1 Hz at one half ground motion 10-4 AFE

• All dams evaluated have large portion of 1 Hz seismic hazard influenced by the New Madrid Fault Zone. Therefore, events contributing to the hazard of one dam are likely to be a major contributor to the hazard at all the other dams.

HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 25

Multiple Dam Failure Combinations From Single Seismic Event Updated Current Licensing Basis OBE + 1/2 PMF 10-4 AFE eq + 500 year flood Fontana and Tellico Apalachia, Blue Ridge, Chatuge, Fort Norris and Tellico Loudoun, Fontana, Melton Hill, Tellico and Watts Bar Dams Fontana, Tellico, Hiwassee, Apalachia, Blue Ridge 1/2 10-4 AFE eq + 25 year flood Cherokee, Douglas and Tellico Apalachia, Blue Ridge, Chatuge, Fort SSE + 25 year flood Loudoun, Melton Hill, Tellico and Watts Bar Dams Norris, Cherokee, Douglas, Tellico HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 26

Results - SQN FHRR+ Current Licensing Basis Failure SQN Elevation (ft) Failure SQN Elevation (ft)

Combination Grade elevation 705 Combination Grade elevation 705 1/2 10-4 AFE eq + 500 year flood OBE+1/2 Probable Maximum Flood Douglas Centered 705.3 Norris, Tellico 706.3 (Apalachia, Blue Ridge, Fontana, Tellico 702.2 Chatuge, Fort Loudoun, Fontana, Tellico, 706.3 Fort Patrick Henry, Hiwassee, Apalachia, Melton Hill, Tellico, Blue Ridge Watts Bar) Cherokee, Douglas, 708.6 10-4 AFE eq + 25 year flood Tellico Fort Loudoun Centered 703.0 SSE+25 year flood (Apalachia, Blue Ridge, Cherokee, Douglas, 706.0 Chatuge, Fort Loudoun, Tellico, Norris Fontana, Melton Hill, Tellico, Watts Bar)

+FHRR shown only as an indication of expected updated results; HEC-RAS ineffective flow area issue requires addressing in updated HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 27

Results - WBN FHRR+ Current Licensing Basis Failure WBN Elevation (ft) Failure WBN Elevation (ft)

Combination Grade elevation 728 Combination Grade elevation 728

-4 1/2 10 AFE eq + 500 year flood OBE+1/2 Probable Maximum Flood Douglas Centered 729.9 Norris, Tellico 728.7 (Apalachia, Blue Ridge, Fontana, Tellico 720.7 Chatuge, Fort Loudoun, Fontana, Tellico, 722.0 Fort Patrick Henry, Hiwassee, Apalachia, Melton Hill, Tellico, Blue Ridge Watts Bar) Cherokee, Douglas, 729.1 10-4 AFE eq + 25 year flood Tellico Fort Loudoun Centered 727.7 SSE+25 year flood (Apalachia, Blue Ridge, Cherokee, Douglas, 731.2 Chatuge, Fort Loudoun, Tellico, Norris Fontana, Melton Hill, Tellico, Watts Bar)

+FHRR shown only as an indication of expected updated results; HEC-RAS ineffective flow area issue requires addressing in updated HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 28

Schedule - Planned Path HEC-RAS FLOOD MODEL PROJECT: SEISMIC DAM FAILURE METHODOLOGY l 29