2013-03-25 – NRC – Jocassee Dam – Duke Energy – Upstream Dam Failure Analysis – ML16237A007

2013-03-25-nrc-jocassee-dam-duke-energy-upstream-dam-failure-analysis-ml16237a007

Leave a Reply

Your email address will not be published. Required fields are marked *

> Date: Mon, 25 Mar 2013 09:57:52 ‐0400
> From: [email protected]
> To: [email protected]; [email protected]; [email protected]
> Subject: Duke’s slides for today @ 2
>
> Gents,
>
> Here are Duke’s slides for today
> Courtesy of NRC staff
>
> JPR
> ‐‐
> Jim Riccio Nuclear Policy Analyst Greenpeace (o) 202‐319‐2487 (email)
> [email protected]
Fukushima –
Flooding Hazard Reevaluation
Upstream Dam Failure Analysis
NCR
Technical
Presentation
NRC Headquarters
One White Flint North
Rockville, MD
March 25, 2013
Oconee Nuclear Station
For Information Only
For Information Only
Dave Baxter, VP, Regulatory Project Completion
Dean Hubbard, Oconee External Flood Licensing Manager
Ray McCoy, Principal Engineer, ONS Civil Design
Chris Ey, Civil Engineering Manager, HDR
Dana Jones, Oconee Fukushima Engineering Supervisor
Joe Ehasz, VP, URS Program Manager – Water Resources
2
For Information Only
Agenda
 Current Dam Failure Analysis – January 28, 2011
 Breach Analysis Summary
 Model Development
 Updated Dam Failure Evaluation – submitted March 12, 2013
 Models Considered
 Selection of Xu & Zhang
 Update Breach Parameters
 Sensitivity Analysis
 Independent Review
 Comparative Analysis – Large Modern Dam Failures
 Modifications Scope
3
For Information Only
2011 Breach Analysis Summary
 Breach parameters developed using regression methodology and
technical papers:
 Froehlich 2008
 Walder & O’Connor
 MacDonald & Langridge-Monopolis
 Breach analysis focused on maximizing flooding levels to provide a
very conservative and bounding analysis:
 Breach dimensions maximized to assume loss of most of the dam
embankment.
 Froehlich breach time of 5 hours was reduced to 2.8
 Maximum peak outflow was selected from all methods
 Breach times of Keowee dams/dikes adjusted to maximize water
directed at the site
 Tailwater effect below Jocassee dam was not considered
4
For Information Only
Jocassee Dam (postulated dam failure)
 Initial breach derived primarily from Froehlich regression
equations.
 Breach dimensions were adjusted based on physical
constraints of natural valley
 Jocassee breach parameters:
 Top Width – 1156 (64% of overall crest)
 Bottom Width – 431 feet
 Bottom Elevation – 800 msl
 Breach Formation Time – 2.8 hrs,
 Peak outflow 5,400,000 cfs
5
2011 Breach Analysis Summary
For Information Only
2011 SE Jocassee Dam Breach
Progression and Stage-Discharge Hydrographs
6
For Information Only
Keowee Dam/Dikes (postulated cascading dam failures)
 Overtopping failure trigger of two feet over the crest
 Cascading dam/dike failure on Keowee
 Keowee main dam- 2.8 hrs
 West Saddle Dam – 0.5 hrs
 Intake Canal Dike- 0.9 hrs
 Little River Dam – 1.9 hrs
 Conservative assumptions were made to maximize the water
directed toward the power block
7
2011 Breach Analysis Summary
For Information Only
8
Model Development
HEC-RAS 1D Model
For Information Only
Model Development
SRH 2D Model
(57 thousand elements)
9
For Information Only
2011 Breach Analysis Summary
2D Model
10
11
Updated Dam
Failure Evaluation
For Information Only
Updated Dam Failure Evaluation
Fukushima 2.1
Attributes of updated and refined dam failure analysis
 Updated methodology and present day regulatory guidance
 Performed to meet NUREG CR/7046, 2011 & ANS 2.8, 1992
 Realistic but still conservative assumptions
 Physical characteristics of the dams/dikes recognized
including materials and method/quality of construction
 Overtopping and Seismic are confirmed from the 2011 SE as
not being credible failure modes
12
For Information Only
Updated Dam Failure Evaluation
Fukushima 2.1
Overtopping of the Jocassee dam was confirmed not to be a credible failure mode
 The Jocassee dam and dikes include 15 feet of freeboard
 The Jocassee watershed is small relative to storage capacity – 148 square miles
 The top of the spillways are located at 1110 (full normal level)
 Four diverse methods of assuring spillway gate operation
 Rigorous spillway gate maintenance and surveillance testing as required and
monitored by FERC
 Lake management procedures require consideration of lower level to anticipate
additional storage needs for significant storms
 Weekly rain forecast are prepared by Duke Energy to project rainfall for the basin
 Precipitation monitoring has assured that no overtopping of the spillway gates has
occurred in 40 + years of operation
 PMF using current HRR-51,52 results in 3 feet of freeboard margin
 2011 SE also concluded that overtopping was not credible
13
For Information Only
Updated Dam Failure Evaluation
Fukushima 2.1
Seismic Failure of the Dam was confirmed not to be a credible failure mode
 Seismic evaluation based on current FERC criteria using the 1989 EPRI Hazard Curves
 The Jocassee dam is designed to a 0.12 g horizontal ground acceleration (Oconee site is designed to a
0.1g horizontal ground acceleration).
 2007 Updated Fragility Analysis
 High Confidence of a Low Probability of Failure (HCLPF) of the dam by sliding 0.305 g
 Evaluation was performed by Applied Research & Engineering Sciences (ARES) Corp., formerly EQE, a
respected consulting firm in the area of seismic fragility
 The ARES report concluded the median centered fragility value for failure of the dam is 1.64 g.
 Maximum Probabilistic Peak Ground Acceleration for a 2% probability of being exceeded within a 50 year
period is 0.197 g (using the United States Geologic Service hazard maps applicable to Jocassee).
 Jocassee dam is included in the seismic model of the Oconee Probable Risk Assessment.
 The combination of the updated seismic fragility with the seismic hazard curve results in a negligible risk
contribution from seismic events.
 In a letter dated 11/20/07 and in the 1/28/11 SE report, the NRC concluded that there is a negligible risk
14
For Information Only
Models Considered
Regression Analysis
 Froehlich 2008
 Walder & O’Connor
 MacDonald & Langridge-Monopolis 1984
 Xu & Zhang 2009
15
For Information Only
Selection of Xu & Zhang 2009
Basis
 Most current regression method developed and validated with
the largest data base of dam failures:
 182 earth and rockfill dam failures compiled
 75 failures w/ sufficient info to develop breach regression models
 Empirical formulas that account for physical characteristics of
dam/reservoir: dam type, failure mode, height, dam erodibility,
reservoir shape/storage)
 33 of the 75 failures were on large dams ( > 15 meters )
 Applies to multi-zoned dams
 Method yields realistic but conservative breach parameters
 Recognized by industry experts
16
For Information Only
Breach Parameters
Fukushima Update
 Jocassee Dam – Xu & Zhang
 Starting reservoir elevation 1110 (normal full pond)
 Rockfill dam with low erodibility classification
 Piping failure initiating at 1020 feet msl (Sunny Day Failure)
 Breach parameters:
Top Width – 701’ (39% of overall crest)
Bottom Width – 431’
Bottom Elevation – 870’
Breach Formation Time:
 Xu & Zhang – 29.2 hrs.(14.2 hours piping +16.0 open weir)
 Froehlich – 16.0 hours (open weir)
Peak outflow: 1,760,000 cfs
17
For Information Only
Jocassee Dam
Low Erodibility Classification
18
Densly compacted core
Densly compacted rockfill
Densly compacted random fill
Large graded rock used for rockfill
Downstream Slope 2H:1V
Quality control throughout construction
For Information Only
Fukushima Model
Structure
Crest
Elevation
(ft msl)
Reservoir
Starting
Elevation (ft
msl)
Failure Mode
Bottom Breach
Elevation (ft
msl)
Bottom
Breach Width
(ft)
Average
Breach
Width (ft)
Right Side
Slope (Zr)
Left Side
Slope (Zl)
Time to
Failure
(Hr)
Top of
Breach
Width (ft)
Breach
Progression
Breach
Initiation
Elevation (ft
msl)
Jocassee
Dam
1125 1,110 Piping 870 431 566 0.53 0.53 29.2 701 Sine Wave 1,020
JOCASSEE DAM BREACH PARAMETERS
19
Breach Formation Time
Xu & Zhang definition: 29.2 (13.2 hours piping + 16.0 hours open weir)
Froehlich definition: 16.0 hours open weir
For Information Only
Fukushima Model Jocassee Dam Breach
Progression and Stage-Discharge Hydrographs
20
Breach Formation Time ; Xu & Zhang definition: – 29.2 (13.2 hours piping + 16.0 hours open weir) Froehlich definition: -16.0 hours open weir
For Information Only
Breach Parameters
Fukushima Update
 Keowee Dam
 Starting reservoir elevation 800 (normal full pond)
 Homogeneous earth fill dam
 Overtopping failure trigger of two feet over the crest at 817 msl by
rapid rise of Keowee reservoir over the crest
 Multiple simultaneous breach initiation formation points across the
Keowee dam and West Saddle dam
 Cascading dam/dike failure on Keowee
 Keowee main dam- 0.75 hrs
 West Saddle Dam – 0.5 hrs (shorter than main dam, ratio of height)
21
For Information Only
Fukushima Model Keowee Dam
Breach Progression HEC-RAS
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Relative Breach Progression
Relative Time Progression
22
For Information Only
Fukushima 1D Modeling
650
660
670
680
690
700
710
720
730
740
750
760
770
780
790
800
810
820
830
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Elevation – feet msl
Model Time – hours
Keowee Dam – Headwater and Tailwater Stage Hydrographs
Final BEP LE 1-D Model Performance
BEP LE HW BEP LE TW
23
For Information Only
Fukushima 2.1 2D Modeling
Keowee Dam Breach Progression
24
For Information Only
Fukushima 2D
Modeling Velocity
and Flow Pattern
at 17 hrs.
25
For Information Only
Fukushima 2D
Modeling Velocity
and Flow Pattern
at 20 hrs.
26
For Information Only
Fukushima 1D-2D
Modeling Results
Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time
817 16.28 817 16.24 n/a n/a n/a n/a
Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time
818.4 16.53 820.1 16.58 810 17.17 807.2 17.67
Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time Elevation Decimal Time
817.5 16.55 815.5 16.53 787.4 17.52 790.4 18.41
Elevation Decimal Time Elevation Decimal Time Depth Decimal Time Depth Decimal Time
n/a n/a n/a n/a 0 n/a 0 n/a
HEC-RAS 2-D HEC-RAS 2-D
HEC-RAS 2-D
Breaching
HEC-RAS 2-D
Maximum Water Surfaces
Keowee Dam Intake Dike
Keowee Dam Intake Dike
HEC-RAS 2-D
HEC-RAS 2-D HEC-RAS 2-D
Swale Tailwater
Maximum Water Surfaces
2-D
Maximum Water Surfaces
SSF SSF
HEC-RAS
27
For Information Only
Sensitivity Analysis
Data in this table based on Wahl 2004, January 28, 2011 SE and updated Xu & Zhang data
100+ HEC-RAS studies performed with varied breach parameters and control variables
Erodiblity was the most significant factor influencing the breach parameters for Xu & Zhang 2009
Bias of conservatism with realism
28
Model Peak Outflow (cfs)
McDonald & Langridge-Monopolis 1984 1,566,381
Costa, 1985 1,634,480
Xu & Zhang, 2009 1,760,000
Evans, 1986 1,803,331
SCS, 1981 2,647,711
Bureau of Reclamation, 1982 3,046,462
McDonald & Langridge-Monopolis 1984 5,093,603 (upper envelope)
Froehlich (with additional conservatism), 2008 5,440,000
For Information Only
29
Independent Review
Breach Parameters
• Independent Peer Review
Joe Ehasz, P.E.
David Bowles, Ph. D P.E. P.H.
• FERC Board of Consultant Review
Gonzalo Castro, Ph.D., P.E.
James Michael Duncan, Ph.D., P.E.
James F Ruff, Ph.D., P.E.
Gabriel Fernandez, Ph.D., P.E.
For Information Only
Comparative Analysis
Large Modern Dam Failures
 Taum Sauk
 Overtopping failure initiated by human error (previous overtopping events had occurred)
 Random rockfill embankment supporting the inner concrete liner loosely placed by end dumping the material
without compaction except for the top 16’ of 84’ height
 The embankment was constructed on a very steep downstream slope of 1.3H to 1V with a 10 high concrete
parapet wall along the crest of the dam
 Embankment was highly erodible and contained over 45% sand sized material (also evident in unusual
level of surface erosion from rain events)
 .Teton
 earthen dam with majority of dam constructed of highly erodible windblown silt (infant mortality event)
 No transition zones (sand and/or fine filters) were included between the silt core and the sand & gravel
 Thin layer of small rock fill on both up and downstream faces with a majority of protection relied upon mix of
sand, gravel and cobble
 Piping failure at 130’ below the crest due to inadequate protection of impervious core trench material
 Breach top width 781’ (~25% of overall crest)
 Hell Hole
 True rockfill dam,with upstream sloping impervious core with massive rock fill sections up and down stream
to support and protect the core.
 Failure caused by overtopping during construction due to an intense rain event that could not be passed
through the construction diversion tunnel
 After overtopping of the core started, the dam took 26 hours to complete the breach and empty the upstream
reservoir 30
For Information Only
 Modifications for protection from dam failure (under review):
1. Relocation of external backup power transmission line
2. Intake Dike embankment protection
3. East embankment protection
4. Discharge Diversion wall
 Modifications for Local Intense Precipitation (under review):
 Transformer relocation
 Diversion walls and drainage canals
 Aux building and Turbine building protection
31
Modification Scope
Updated
For Information Only
Modification Options
32
Jocassee Dam
1
Questions and Feedback
33