ML110280153

Staff Assessment of Duke'S Response to Confirmatory Action Letter Regarding Duke'S Commitments to Address External Flooding Concerns at the Oconee Nuclear Station, Units 1, 2, and 3 (ONS) (TAC Nos. ME3065, ME3066, and ME3067)

ML110280153
Person / Time
Site:	Oconee
Issue date:	01/28/2011
From:	Leeds E Office of Nuclear Reactor Regulation
To:	Gillespie P Duke Energy Carolinas
Stang J, NRR/DORL, 415-1345
Shared Package
ML110280163	List: ML110280153
References
TAC ME3065, TAC ME3066, TAC ME3067
Download: ML110280153 (17)

Text

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION January 28, 2011 Mr. Preston Gillespie Site Vice President Oconee Nuclear Station Duke Energy Carolinas, LLC 7800 Rochester Highway Seneca, SC 29672

SUBJECT:

STAFF ASSESSMENT OF DUKES RESPONSE TO CONFIRMATORY ACTION LETTER REGARDING DUKES COMMITMENTS TO ADDRESS EXTERNAL FLOODING CONCERNS AT THE OCONEE NUCLEAR STATION, UNITS 1, 2, AND 3 (ONS) (TAC NOS. ME3065, ME3066, AND ME3067)

Dear Mr. Gillespie:

By letter dated June 22, 2010, the U.S. Nuclear Regulatory Commission (NRC) issued a confirmatory action letter (CAL) to Duke Energy Carolinas, LLC (Duke, the licensee), associated with the mitigation of external flooding hazards at the Oconee Nuclear Station, Units 1, 2, and 3 (ONS) site, resulting from a postulated failure of the Jocassee Dam. The CAL confirmed your commitment to submit to the NRC by August 2, 2010, all documentation necessary to demonstrate to the NRC that the parameters and analysis used to evaluate the inundation of the ONS site resulting from the postulated failure of the Jocassee Dam was bounded.

Use of the term bounded, in this case, refers to conditions that bound the random sunny-day failure of the Jocassee Dam. The random sunny-day failure scenario was selected after evaluation of the failure modes determined that the potential failure of the Jocassee Dam from either an overtopping event or seismic event was not credible. Bounding reservoir levels were taken at the Federal Energy Regulatory Commission maximum allowable operating levels, not the absolute worst case.

By letter dated August 2, 2010, you provided the required information to the NRC. The NRC staff reviewed the information provided, and found that the documentation provided sufficient justification that the parameters chosen by the licensee and the analysis performed bound the inundation of the ONS site resulting from a potential failure of the Jocassee Dam and therefore providing reasonable assurance for the overall flooding scenario at the site. Enclosed is the staffs evaluation of the licensees documentation.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION T. Gillespie The Office of Nuclear Reactor Regulation (NRR) staff conducted the evaluation of this matter at the request of Region II since the NRCs technical expertise in this area is in NRR. In its evaluation, the NRC staff determined that the licensee provided the documentation necessary to demonstrate to the staff that the inundation of the ONS site resulting from the postulated failure of the Jocassee Dam was bounded. Therefore, the staff considers the CAL action associated with this issue to be closed. The NRC staffs assessment is based on the information that Duke provided to the staff by letter dated August 2, 2010.

If you have any questions, please call John Stang at 301-415-1345.

Sincerely,

/RA by JGrobe for/

Eric J. Leeds, Director Office of Nuclear Reactor Regulation Docket Nos. 50-269, 50-270, and 50-287

Enclosure:

Safety Evaluation OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION T. Gillespie The Office of Nuclear Reactor Regulation (NRR) staff conducted the evaluation of this matter at the request of Region II since the NRCs technical expertise in this area is in NRR. In its evaluation, the NRC staff determined that the licensee provided the documentation necessary to demonstrate to the staff that the inundation of the ONS site resulting from the postulated failure of the Jocassee Dam was bounded. Therefore, the staff considers the CAL action associated with this issue to be closed. The NRC staffs assessment is based on the information that Duke provided to the staff by letter dated August 2, 2010.

If you have any questions, please call John Stang at 301-415-1345.

Sincerely,

/RA by JGrobe for/

Eric J. Leeds, Director Office of Nuclear Reactor Regulation Docket Nos. 50-269, 50-270, and 50-287

Enclosure:

Safety Evaluation DISTRIBUTION:

DE R/F JStang ADAMS ACCESSION NO.: ML110280153 OFFICE NRR/DE/EMCB NRR/DE NRR/DRA NRR/DE NRR/DORL/LPL2-1 NAME MKhanna GWilson JMitman/JM PHiland GKulesa Non-Concur DATE 12/7/10 12/10/10 1/10/11 01/27/10 01/28/11 OFFICE NRR/DORL/LPL2-1 NRR/DORL NRR/DORL R-II NRR/DRA NAME MOBrien APersinko JGiitter VMcCree MCunningham DATE 01/27/11 01/27/11 01/27/11 01/27/11 1/27/11 OFFICE NRR NRR NAME JGrobe ELeeds (JGrobe for)

DATE 01/28/11 01/28/11 OFFICIAL RECORD COPY OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO DUKE ENERGY CAROLINAS, LLC CONFIRMATORY ACTION LETTER - COMMITMENTS TO ADDRESS EXTERNAL FLOODING CONCERNS CLOSURE OF INUNDATION SITE RESULTS OCONEE NUCLEAR STATION, UNITS 1, 2, AND 3 (ONS)

DOCKET NOS. 50-269, 50-270, AND 50-287

1.0 BACKGROUND

Duke Energy Carolinas, LLC (Duke or the licensee), performed an inundation study in 1992 to meet a Federal Energy Regulatory Commission (FERC) requirement for formulating an emergency action plan in the event that the Jocassee Dam failed. This study showed that approximately 16.5 feet of water would inundate the yard area surrounding the standby shutdown facility (SSF). This inundation of the ONS site would render all systems necessary to shut down and maintain all three units in a safe and stable condition inoperable.

In April 2006, the U.S. Nuclear Regulatory Commission (NRC) staff questioned the flood protection barrier for the SSF. The NRC identified that the licensee had incorrectly calculated the Jocassee Dam failure frequency and had not adequately addressed the potential consequences of flood heights predicted at the ONS site, based on the information provided by the 1992 inundation study.

Based on concerns raised by the NRC, by letter dated August 15, 2008 (Agencywide Documents Access and Managements System (ADAMS) Accession No. ML081640244), the NRC requested information from the licensee pursuant to Title 10 of the Code of Federal Regulations (10 CFR),

Section 50.54(f). By letter dated September 26, 2008, ADAMS Accession No. ML082750106),

the licensee responded to the request. The NRC staff reviewed the information and based on the review, the NRC staff found that the information provided by the licensee did not demonstrate that the ONS site would be adequately protected from external flooding events. Specifically, the licensee did not (1) provide an adequate inundation study, (2) provide a deterministic resolution of this matter, or (3) provide a schedule to resolve the external flooding issue in a timely manner.

Enclosure OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION By letter dated April 30, 2009 (ADAMS Accession No. ML090570779), the NRC requested the following additional information:

(1) a deterministic resolution of external flooding at the ONS site, and (2) a schedule to resolve the external flooding issue in a timely manner.

The NRC staff met with and had several telephone conversations with the licensee concerning the external flooding issue at the ONS site. By letter dated November 30, 2009 (ADAMS Accession No. ML093380701), the licensee provided its technical response to the NRCs April 30, 2009, letter. The NRC staff reviewed the licensees response and determined that although the licensee provided a more accurate estimate of the flooding caused by a failure of the Jocassee Dam, the NRC staff found that additional information was needed. By letter dated January 29, 2010 (ADAMS Accession No. ML100271591), the NRC requested additional information requiring that the licensee provide analyses to demonstrate, for the entire Jocassee earthen works, that the ONS site will be adequately protected from external flooding events. By letter dated March 5, 2010 (ADAMS Accession No. ML103430047), the licensee provided a partial response to the NRCs January 29, 2010, request for additional information (RAI).

On June 22, 2010, the staff issued a confirmatory action letter (CAL) to the licensee, requesting the following: the licensee to submit to the NRC all documentation necessary to demonstrate that the inundation of the ONS site from the postulated failure of the Jocassee Dam has been bounded; the licensee to submit a list of all necessary modifications to mitigate the inundation by November 30, 2010; and the licensee to make all necessary modifications by November 30, 2011.

The staff also requested that the compensatory measures (CMs) listed in the CAL remain in place until final resolution has been agreed upon between the licensee and the NRC staff.

By letter dated August 2, 2010, Duke provided its response to the remaining questions and its response to the CAL action requiring submittal of all documentation for the inundation of the ONS site from the postulated failure of the Jocassee Dam. The NRC staffs technical assessment of the information is provided below.

2.0 PURPOSE The purpose of this assessment is to verify that the licensee has provided adequate justification that the parameters chosen and the analysis performed bound the inundation of the ONS site resulting from a postulated random sunny-day failure of the Jocassee Dam. More specifically, the NRC staffs assessment includes the confirmation that the licensees parameters, used in the unmitigated Case 2 analysis, as discussed below, are conservative and provide reasonable assurance that the inundation of the ONS site from a random sunny-day failure of the Jocassee Dam will not exceed the levels predicted by the licensee. This Case 2 scenario will be the new flooding basis for the site. Results of the hypothetical dam failures provide inputs to surface water flow models used to simulate floodwater levels at the ONS site, specifically the water levels at locations that could have an effect on emergency shutdown capability, particularly, the SSF.

The SSF is a partially flood-protected structure which houses control systems to shut down the plant. Ground elevation at the base of the SSF is 796.0 ft. mean sea level.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION The dam breach parameters used for the Jocassee Dam and the earthen structures at the ONS site, as discussed below, were evaluated to ensure that they provide reasonable assurance for the flooding levels that the ONS site would see with a random sunny-day failure of the Jocassee Dam. The probable failure mode analysis (PFMA) stated that the most likely failure of the Jocassee Dam would be a piping failure through the left (east) or right (west) abutment. Since the most likely analyzed failure of the dam would be the piping failure, the staff determined that reasonable assurance is provided by using this as a failure mechanism. The last flooding inundation study performed had the starting reservoir level of the Jocassee Dam at 1110 feet, which is greater than the maximum power pool level during the hurricane season (1108 ft.). At 1110 ft., there is an additional 2 ft. above the normal pool level, and this is also the point where water also starts overtopping the flood control gates, therefore, the staff determined that there is sufficient conservatism at this reservoir level. In addition, the staff agrees that these parameters were appropriately used to start the hypothetical failure scenario associated with the Jocassee Dam.

The other parameters evaluated were breach dimensions, breach position, breach time, peak discharge flow rates, and Manning's n-values. The evaluation for these values is discussed later in this assessment. The main structures that were evaluated in the flooding scenario were the Jocassee Dam, the Hartwell Reservoir, and the structures around the ONS, which include the Keowee Dam, Little River Dam, ONS Intake Canal Dike, and the West Saddle Dam.

3.0 NRC STAFFS EVALUATION 3.1 HEC-RAS Modeling To accurately determine water levels over the ONS site resulting from a random sunny-day failure of the Jocassee Dam, unsteady (time varying) flow over approximately 44 miles of the river system had to be simulated by the licensee. The licensees simulation model included the Jocassee, Keowee, and Hartwell reservoir systems and incorporated the failure of the Jocassee, Keowee, and Little River Dams. The Hartwell Dam, which could also fail, was conservatively used as a downstream control and limited the size of the model. The model also incorporated flow bifurcation around the ONS site to the north and reunification of flows below the Little River Dam. To perform this river system modeling, the licensee chose the HEC-RAS program for this purpose. The HEC-RAS program was developed by the U.S. Army Hydrologic Engineering Center and it is one of the standards for flooding inundation studies.

The HEC-RAS simulations allowed for the efficient calculation of flow hydrographs and water elevations at various points of interest around the plant under various conditions of failure of the Jocassee Dam, as well as the failure of downstream structures such as the Keowee Dam, West Saddle Dam, Little River Dam, and the intake canal dike. Also included in the sensitivity studies was the effect of Mannings n-value variation for both the main channel and overbank for various reaches. Once the failure parameters for the Jocassee Dam were established, additional sensitivity studies were performed for additional failure modes for Keowee Dam. These cases included evaluation of breach geometries for the various earthen works (widths, bottom elevation, side slopes) and failure progression characteristics of the breaches (time to failure, linear or sine wave progression).

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION 3.2 Jocassee Dam, Oconee Site Dams, and Dike Breach Parameters and HEC-RAS Modeling The HEC-RAS computation was used to assess flooding at the ONS site for all three case scenarios. The three different case scenarios, provided by the licensee, were assessed with differing breaching parameters, such as time-to-failure and breach size. After analyzing the licensees case scenarios, the staff determined that Case 2 was acceptable based on the conservatisms included in the parameters used in the case study. The licensees parameter values, provided in Table 1, represent Case 2 for a "random sunny-day failure of the Jocassee Dam:

Table 1 Reservoir Elevation 1110 ft. msl Bottom Breach Elevation 800 ft. msl Bottom Breach Width 425 ft.

Side Slopes West slope (1.55:1)

East slope (0.7:1)

Time-to-- Failure 2.8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> Piping Elevation 1020 ft. msl.

Failure Progression Sine Wave Figure 1 below visually shows the breach dimensions of the Jocassee Dam.

Figure 1: Jocassee Dam Breach Dimensions The staff determined that the Jocassee Dam starting reservoir level of 1110 ft., as described above, is conservative.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION Based on the breaching parameters of the Jocassee Dam, the peak outflow was computed. The Jocassee Dam breach peak outflow was computed using the HEC-RAS model. The peak outflow was determined to be 5,440,000 cubic feet per second (cfs), which was greater than the empirically determined peak outflows, using several available models, as listed in Table 2, below.

Table 2: Empirical Equations for Predicting Peak Flows for a Jocassee Dam Breach Lake volume at stage = 1110 ft = 1,418,869,244 m3 - volume of water (vw)

Depth of water above invert at failure = 94.5 m - height of water (hw)

Peak Flow Equations:

Model Peak Outflow (m3/s) Peak Outflow (ft3/s)

MacDonald & 44328.60 1566381 Langridge-Monopolis, 1984 MacDonald & 144148.96 5093603 (upper envelope)

Langridge-Monopolis, 1984 Costa, 1985 46255.80 1634480 Bureau of Reclamation, 1982 86214.88 3046462 Evans, 1986 51034.28 1803331 SCS, 1981 74930.21 2647711 Equations embedded in this table are from Table 1 of Wahl, 2004 Based on a comparison with the values determined from empirical models (Table 2), the staff determined that the HEC-RAS model results for peak outflow are conservative.

The Jocassee Dam overall breach dimension assumes the entire loss of the dam embankment and massive erosion of bedrock at the dam base. The biotite gneiss which comprises the bedrock type at the base of the dam would be extremely resistant to erosion, so a large degree of conservatism was added to the breach size. The average width of the assumed dam breach

(~1137 ft) is one of the overall breaching parameters. This is larger than the average width estimated using Froehlichs 2008 methods (i.e., ~900 ft).

The Jocassee Dam breach hypothetical failure time of 2.8 hrs. is very short for a dam with the quality of construction, basal rock type, and degree of monitoring of the Jocassee Dam, so the staff determined that adequate conservatism was added to the breach size. The licensee used Froehlichs 2008 methods in their estimation. It is important to note that the breach dimensions and breach times are related.

As part of the model verification process, the licensee compared the volume of the outflow hydrograph from the Jocassee Dam failure with the total volumes of the flow hydrographs, through the connecting canal and over the Keowee Dam. This demonstrated that volume was properly being conserved in the flow routing by the model. The ability of the HEC-RAS geometric input to model the volumes of Lake Keowee and Lake Hartwell was also verified by comparing the volumes at normal pool level, as calculated by the model with the known volumes. Both lake volumes agreed within five percent, which is an indication of model alignment.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION Failure parameters for the downstream dams are provided in Table 3, below.

Table 3 Parameter Keowee Dam West Saddle ONS Intake Little River Dam Dam Canal Dike Breach Bottom 670 ft. msl 795 ft. 715.5 ft. msl 670 ft. msl Elevation Breach Bottom 500 ft. 1680 ft. 200 ft. 290 ft.

Width Side Slopes 1:1 1:0 1:1 1:1 Overtopping 817 ft. msl 817 ft. msl 817 ft. msl 817 ft. msl Trigger Main Dam 2.8 hrs. 0.5 hrs. 0.9 hrs. 1.9 hrs.

Failure Time Failure parameters for the dams and structures in Table 3 were developed based on Froehlichs 2008 methods. The bottom breach width for Keowee using Froehlich's 2008 methods was determined to be 1028 ft. with a failure time of 5 hrs. The physical size of the dam, however, limits the bottom breach width to 500 ft. Because the breach size and failure time are related, the failure time was reduced to 2.8 hrs., based on proportions. The breach widths and failure times of the Oconee Intake Dike, West Saddle Dam, and Little River Dam were determined in a similar manner. The staff concludes that the dam failure parameters for the other structures in Table 3 are conservative based on the physical constraints of the structures. Further information is provided in Attachment 1 of Dukes letter to the NRC dated January 15, 2010 (ADAMS Accession No. ML100210199).

The staff agreed that the overtopping trigger of two feet over the crest of the dam was found to be conservative based on the assumption that the slower (and/or later) the breach of the Keowee and West Saddle Dams, the greater the flow of water through the intake canal dike breach, which is the major contributor to the water level at the SSF.

At a water elevation of two feet over the crest, the water velocities are about 2 to 4 times the velocity required to initiate erosion on a grassed slope. This erosion will result in a breaching of the dam. The Mannings n-values assumed for the Case 2 study (run number 100), are provided in Table 4, below.

Table 4 Structure Mannings n-values Jocassee Tailrace 0.07 Keowee Reservoir and Little River Channels 0.025 Keowee Reservoir and Little River Overbank 0.08 Keowee Reservoir Tributaries 0.035 Keowee, Intake and Little River Tailraces 0.07 Hartwell Reservoir Channel 0.025 Hartwell Reservoir Overbank 0.08 A 60-foot threshold was chosen by the licensee to identify the change from stream to deep reservoir flow conditions. A deep water flow condition was modeled with a Mannings n-value of OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION 0.025. The modeled reservoir tributaries were considered for the streams, and their n-values remained at 0.035.

A Mannings n-value of 0.07 was used in the respective tailrace reaches below the Jocassee Dam, Keowee Dam, ONS Intake Canal Dike, and the Little River Dam to account for roughness associated with displaced dam breach material (suspended material and bed load). The affected reach lengths below each dam, where the higher roughness values were assigned, were assumed as the base length (upstream-downstream) dimension of each dam, followed by a second base length dimension to allow transition from 0.07 to the reservoir roughness coefficients of 0.025 in a linear fashion.

Based on its review, the NRC staff determined that the n-values chosen by the licensee are appropriate, as discussed below. Table 5-6 of Open Channel Hydraulics (Chow, 1959) tabulates the n-values for various conditions. A range of 0.025 to 0.060, which corresponds to a main channel that has no boulders or brush bracketing, bounds the values of 0.025 and 0.035 used by the licensee. Figure 5-4 in Chow (1959) shows a definite decrease in n-value with an increasing stage for 3 different rivers. This supports the reduction of the n-value with depth assumed by the licensee. Also, according to Chow (1959), a value of 0.08 corresponds to flood plains of cleared land, as might be expected after being swept with high velocity water from the tailrace of a breached dam. A comparison of the reaches with pictures and n-values in the U.S. Geological Survey (USGS) Water Supply Paper (1849) also assisted in narrowing the acceptable range of n-values by the staff. In addition, the sensitivity studies performed by the licensee showed a decrease in sensitivity to the downstream main channel n-values at higher flows.

Preliminary results from the licensee presented in October 2009 and again in August 2, 2010, showed a double peaked elevation hydrograph occurring at the SSF. From the timing of the peaks, it was determined that the first peak was primarily due to overtopping and breaching of the ONS intake canal dike. The second peak appeared to be primarily from the Keowee tailrace with flow from the Keowee Dam failure combined with flow over the site from the West Saddle Dam and the ONS intake dike failure.

At the request of the staff, the licensee further investigated the effects of a more rapid failure of the Keowee Dam to produce a greater Keowee tailrace contribution to the site flooding. The licensee added 6 more HEC-RAS runs and then selected two of the runs (100B and 100F) for more detailed 2-D modeling. HEC-RAS Case 2 study (run number 100) was used to set the 2-D boundary conditions. The additional HEC-RAS runs evaluated are provided in Table 5, below.

Table 5 100A Rapid failure (0.5 hrs) for Keowee Main dam 100B Median failure (1.65 hrs) for Keowee Main Dam 100C Rapid failure (0.5 hrs) for ONS Intake Canal east dike 100D Rapid failure (0.5 hrs) of additional breach (bottom width 400ft.) at ONS Intake Canal Dike 100E Rapid failure (0.5 hrs) of both east and north portions at ONS Intake Canal Dike 100F Rapid failure (0.5 hrs) for all Keowee structures, ONS Intake Canal Dike, and Little River Dam OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION 3.3 Additional Conservatisms in the HEC-RAS for Jocassee Breach Modeling The tailwater effect of the convoluted flood pathway below the Jocassee Dam was not considered.

A limit on the depth of hypothetical breach cutting at Jocassee (and, therefore, on the peak discharge) is provided by the tailwater elevation of 800 ft, which represents the Keowee normal pool level. This is the base elevation that was assumed for the hypothetical breach at Jocassee.

The staff agreed that it is conservative to not include the tailwater effect of the convoluted flow pathway below the Jocassee Dam, because excluding it would shorten the time required to empty the reservoir.

Figure 2: Satellite image of Jocassee Dam showing submerged quarry site and embayment area below the dam which has a restricted outlet connection to the rest of Keowee Lake The scouring effects of the flood waters were not considered. The effect of this scour would be to enlarge the channel system (compared to its present width and depth) and accelerate the transport of floodwaters southward away from the site. Therefore, conservatisms were added by the additional water levels at the site, since the licensee did not utilize this effect. Mapping of flood scouring effects has been documented by various authors, including the recent work of Krizanich (2010) downstream from the Taum Sauk reservoir. See the scoured flood path in Figure 3, below.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION Figure 3: Taum Sauk Breach 3.4 Two-Dimensional Modeling Early in the review, there was a concern about the ability of a one-dimensional (1-D) model to effectively simulate the flow regime immediately upstream of the canal to the north of the plant (connecting the Keowee River Basin to the Little River Basin), where the downstream velocity vector makes a 90-degree change in direction. Also, the potential for inundation of the site comes from many potential sources and is likely to flow in different directions without channelization.

Such overland flow may involve eddy patterns, flow recirculation, and spill over barriers. Alternate wetting and drying of area elements may also be required depending on the overland flow patterns. For these reasons, a two-dimensional (2-D) model was coupled with the HEC-RAS simulations at boundaries, sufficiently remote, where the hydraulic parameters of flow and depth would be relatively unaffected by flow over the site.

The 2-D model chosen was developed by the U.S. Department of Interior, Bureau of Reclamation, entitled Sedimentation and River Hydraulics-Two Dimensional River Flow Modeling (SRH-2D).

For the modeling effort, a 2-D mesh of triangular and quadrilateral elements was constructed of the area surrounding the station. The mesh size was selected to model the desired area, while keeping the computational array to a manageable size. The final computational mesh has approximately 57,500 unstructured elements. The mesh is coarser in areas that are farther away from ONS and finer in areas where more detail is required. The upstream boundary is about 6,200 feet wide and the upstream boundary condition consists of an inflow hydrograph. The downstream boundary conditions consist of stage hydrographs.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION Figure 4: FLOW MODEL SCHEMATIC (Duke, 2009)

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION Figure 5: TWO-DIMENSIONAL MODEL BOUNDARY DEFINITIONS (Duke, 2010) 3.5 2-D Model Computations The one-dimensional HEC-RAS and two-dimensional SRH-2D models are not dynamically coupled; and mass and momentum between the two models cannot be conserved. Hence, potential backflow between the inflow and outflow boundaries cannot be incorporated into the model. Also, the simulations represent an extreme and unobserved scenario, and parameters such as roughness coefficients over the site cannot be calibrated, requiring the selection of conservative values. The lack of coupling resulted in higher water levels from the 2-D simulations at the upstream boundaries than from the HEC-RAS runs (Wilson, 2010 and Young, 2009). The OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION peak of the higher values at the upstream boundary after a short dip, as plotted by Wilson, appeared about 20 minutes later than the peak at the Keowee Dam. This indicated a possible reflection of the flood wave at the Keowee Dam intensified by the rigid (no backflow) upstream boundary condition. This intensified reflection may have resulted in greater flow through the Keowee tailrace. The ONS peak appeared to occur earlier and was probably unaffected, although it was still about a foot higher than the HEC-RAS simulation.

Model runs 100B and 100F were selected for the more detailed 2-D assessment of the downstream Keowee tailrace area. Evaluation of these runs showed that faster failure times did not increase water depths at the SSF and that the original set of parameters with an updated computational mesh results in the greatest depth at the SSF of 19.5 ft.

The licensee confirmed that case 100M (original case) resulted in the highest 2-D water levels, and case 100W was formulated to incorporate improvements utilizing new boundary conditions and became the record model run. Based on its assessment, the staff agreed with the licensees approach, utilizing case 100W.

The final maximum water surface levels, determined from this updated computational mesh and the parameters listed earlier are:

2-D water levels 1-D water level Keowee Dam (upstream) 839.6 ft msl 834.8 ft. msl Keowee Dam (tailwater) 805.3 ft msl 791.5 ft. msl Intake Dike 823.0 ft msl 821.8 ft. msl Swale 828.5 ft msl 827.8 ft.msl SSF 815.0 ft msl N/A The model showed that there are key points that control the water level at the ONS site, which are discussed below.

The Keowee Dam (upstream) is the primary control on the water level upstream of the canal, which allows flow from the Keowee River Basin into the Little River Basin and eventually into the intake canal. The crest elevation of the Keowee Dam is 815 ft msl. The water level above Keowee Dam also controls the amount of water, which could flow through the swale near the World of Energy. The staff agreed that upstream of the Keowee Dam is a convenient location to compare water levels computed with HEC-RAS with those computed by the 2-D model showing the relative adequacy of boundary conditions.

The Keowee Dam tailwater is a possible source of flooding from the east side of the plant site across the switchyard. Flooding from the Keowee Dam tailrace resulted in a second (lower) flood peak at the SSF. It was noted that the 2-D simulation resulted in a higher second peak than the 1-D simulation at the tailwater. This was caused by retention of flow from the Oconee intake canal in the plant yard and delayed release into the tailrace, a result of simultaneous forward and lateral flow which could not be modeled in the 1-D simulation. The bounding scenario will be that which results in a higher water level.

The intake dike, which has a top elevation of 815 ft. msl, will allow flooding of the plant upon overtopping, independent of the breach location(s). Breaching to the east, however, will result in flow to the south east of the power block and eventually flow to the Keowee Dam tailrace.

However, breaching to the north does not appear to result in higher water levels at the SSF.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION The swale is a low lying pathway from the north of the plant to the World of Energy. Flooding from the swale would have an impact on the inundation levels at the ONS site. The swale has an invert or bottom elevation of 827 ft. msl. The licensees Case 2 scenario provided flow through the swale.

4.0 CONCLUSION

The NRC staff evaluated the information provided by Duke in their August 2, 2010, letter. The unmitigated Case 2 dam breach parameters that were used in the flooding models, provided by Duke for the ONS site, demonstrated that the licensee has included conservatisms of the parameters utilized in the dam breach scenario. These conservatisms provide the staff with additional assurance that the above Case 2 scenario will bound the inundation at ONS, therefore providing reasonable assurance for the overall flooding scenario at the site. This new flooding scenario is based on a random sunny-day failure of the Jocassee Dam. This Case 2 scenario will be the new flooding basis for the site.

The licensee has submitted to the NRC all documentation necessary to demonstrate that the inundation of the ONS site from the failure of the Jocassee Dam has been bounded. In addition, the licensee has committed to keep the compensatory measures in place until final resolution has been agreed upon between the licensee and the NRC staff. Therefore, this technical assessment officially closes the CAL action which stated, The licensee to submit to the NRC all documentation necessary to demonstrate that the inundation of the Oconee site from the failure of the Jocassee Dam has been bounded.

5.0 REFERENCES

1. Badr, A. W., A. Wachob, and J. A. Gellici, 2004. South Carolina Water Plan, Second Edition. Available at: http://www.dnr.sc.gov/water/admin/pubs/pdfs/SCWaterPlan2.pdf

2. Baecher et al., 1980. "Risk of Dam Failure in Benefit-Cost Analysis," Water Resources Research, 16, No. 3, pp. 449-456, 1980.

3. Bureau of Reclamation, 1982. Guidelines for defining inundated areas downstream from Bureau of Reclamation dams, Reclamation Planning Instruction No. 82-11, U.S. Dept. of the Interior, Bureau of Reclamation, Denver, p. 25.

4. Costa, J. E., 1985. Floods from dam failures, U. S. Geological Survey, Open-File Report No.85-560, Denver, p. 54.

5. Duke Energy Carolinas Letter to US NRC from D. Baxter, dated August 2, 2010, Regarding Oconee Response to Confirmatory Action Letter, 2-10-003.

6. Evans, S. G., 1986. The maximum discharge of outburst floods caused by the breaching of man-made and natural dams, Can. Geotech. J., 23(4), 385-387.

7. FERC, 2006. Taum Sauk Pumped Storage Project (No. P-2277), Dam Breach Incident FERC Staff Report, April 28, 2006, available online at:

http://www.ferc.gov/industries/hydropower/safety/projects/taum-sauk/staff-rpt.asp

8. Froehlich, D. C., 2008. Embankment Dam Breach Parameters and Their Uncertainties, Journal of Hydraulic Engineering, Vol. 134, No. 12, Dec. 1, 2008

9. Krizanich, G. W., 2010. Rapid Response Field mapping of the Taum Sauk Reservoir Failure, Paper No. 38-6, Geological Society of America, North-Central and South-Central Section Meetings (44th Annual), April 11-13, 2010.

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

OFFICIAL USE ONLY - SECURITY RELATED INFORMATION

10. MacDonald, T. C., and J. Langridge-Monopolis, 1984. Breaching characteristics of dam failures, J. Hydraul. Eng., 110(5), 567-586.

11. NRC Division of Risk Assessment, 2009. Generic failure rate evaluation for Jocassee Dam (Sensitive Information document - Official Use Only).

12. (Barnes, 1967) Harry H. Barnes Jr., Roughness Characteristics of Natural Channels, Geological Survey Water-Supply Paper 1849, United States Government Printing Office, Washington, D.C., 1967.

13. (Baxter, 2010a) Duke Energy Carolinas Letter to US NRC, Oconee External Flood, Response to Request for Additional Information, March 5, 2010.

14. (Baxter, 2010b) Duke Energy Carolinas Letter to US NRC, Oconee Responses to Confirmatory Action Letter (CAL) 2-10-003 August 2, 2010.

15. (Chow, 1959) Ven Te Chow, Open Channel Hydraulics, McGraw-Hill Book Company, New York, 1959.

16. (Lai, 2008) Yong G. Lai, SRH-2D version 2: Theory and Users Manual, Sedimentation and River Hydraulics - Two Dimensional River Flow Modeling, U.S. Bureau of Reclamation, November 2008.

17. (Young, 2009) Letter from Nathan C. Young to Dr. Andrew McCoy, Review of the HDR Engineering, Inc, Two Dimensional Simulation of Flow in the Vicinity of Oconee Nuclear Station Resulting from a Hypothetical Dam Failure Scenario at Lake Jocassee. October 16, 2009.

18. (Baxter, 2010c) Duke Energy Carolinas Letter to US NRC, Oconee External Flood Interim Actions, January 15, 2010.

19. (Wilson, 2010) Letter from Lloyd Wilson, Wilson Engineering, to Raymond l. McCoy, Independent Technical Review, HEC-RAS and SRH-2D Hydraulic Modeling, Wilson Engineering Project WE10009, June 14, 2010.

20. Wahl, T.L., 2004, Uncertainty of Predictions of Embankment Dam Breach Parameters.

21. MacDonald, T.C., and J. Langridge-Monopolis, 1984, Breaching Characteristics of Dam Failures, Journal of Hydraulic Engineering, Vol. 110, No. 5, pp. 567-586.

22. Costa, J.E., 1985, Floods from Dam Failures, U.S. Geological Survey Open-File Report 85-560, Denver, Colorado, p. 54.

23. Soil Conservation Service (SCS), 1981, Simplified Dam-Breach Routing Procedure, Tech.

Release No. 66 (Rev. 1), p. 39.

24. Evans, S.G., 1986, The Maximum Discharge of Outburst Floods Cause by the Breaching of Man-Made and Natural Dams, Canadian Geotech. J., 23(4), pp. 385-387.

Principal Contributors: Rex Wescott, NMSS Neil Coleman, ACRS Date: January 28, 2011 OFFICIAL USE ONLY - SECURITY RELATED INFORMATION