ML12226A562

Sequoyah, Units 1 and 2, Enclosure 1, Evaluation of Proposed Changes Attachment 1 Proposed SQN Units 1 and 2 UFSAR Text Changes (Markups), Page 2.4-1 Through 2.4-41

ML12226A562
Person / Time
Site:	Sequoyah
Issue date:	08/10/2012
From:	Tennessee Valley Authority
To:	Office of Nuclear Reactor Regulation
References
TAC ME8200
Download: ML12226A562 (42)

Text

ENCLOSURE I EVALUATION OF PROPOSED CHANGES ATTACHMENT I Proposed SQN Units I and 2 UFSAR Text Changes (Markups)I SQN-2.4 HYDROLOGIC ENGINEERING SON is located on the riaht hank of Chickamauaoa L ake at Te~nnessee Rivr mile (TRM'/484 F5 with plant grade at elevation 705.0 ft MSL. The plant has been designed to have the capability for safe shutdown in floods UD to the computed maximum water level, in accordance with reaulatorv position 2 of RG 1.59.Determination of the maximum flood level included consideration of postulated dam failures from seismic and hydrologic causes. The maximum flood elevation of 722.0 ft would result from an occurrence of the probable maximum storm. Coincident wind wave activity results in wind waves of up to 4.2 ft (crest to trouah). Wind wave run uo on the Diesel Generator Buildina reaches elevation 723.2 ft. Wind wave run up on the critical wall of the ERCW Intake Pumping Station and the walls of the Auxiliary, Control and Shield Buildinqs reaches elevation 726.2 ft.The nearest surface water user located downstream from SQN is East Side Utility at TRM 473.0, 11.7 miles downstream.

All surface water suoBlies withdrawn from the 98.6 mile reach of the mainstream of the Tennessee River between Dayton, Tennessee (TRM 503.8) and the Mead Corporation in Stevenson, Alabama (TRM 405.2) are listed in Table 2.4.1-1.2.4.1 Hydrologic Description 2.4.1.1 Site and Facilities The location of key plant structures and their relationship to the original site topography are shown on Figure 2.1.2-1. The structures which have safety-related equipment and systems are indicated on this figure and are tabulated below, along with the elevation of major exterior accesses.Number of Accesses Structure Access Intake.p.

• Pumping (1) Stairwell entrance st-UGtUleStructure (2) Access hatches (3) Cable tunnel Auxiliary and Control Buildings (1)(2)(3)(4)(5)(6)(7)Railroad access opening Doors to turbine building Doors to turbine building Doors to turbine building Personnel lock to SB General vent or intake Doors to AEB and MSW 1 6 1 1 2 2 2 4 2 4 1 1 1 4 1 4 1 Elevation (ft)705.0 705.0 690.0 Shield buildingBuilding Diesel gene-ate4 buldi4! RGenerator Building ERCW r Rakelntake Pumping Station (1) Personnel lock (watertight)

(2) Equipment hatch (3) Personnel lock (1) Equipment access door (2) Personnel access door (3) Emergency exit (4) Emergency exit (1) Access door (2) Trash sluice (3) Deck drainage (sealed for flood)706.0 706.0 732.0 685.0 690.0 714.0 714.0 691.0 730.0 732.0 722.0 722.0 722.0 740.5 725.0 723.5 720.0 1 1 Exterior accesses are also provided to each of the class IE electrical systems manholes and handholes at elevations varying from 700 ft MSL to 724 feet MSL, depending upon the location of 2.4-1 SQN-each structure.

The relationship of the plant site to the surrounding area can be seen in Figures 2.1.2-1 and 2.4.1-1. It can be seen from these figures that significant natural drainage features of the site have not been altered. Local surface runoff drains into the Tennessee River.2.4.1.2 Hydrosphere The Sequoyah Nuclear Plant (SQN) site comprises approximately 525 acres on a peninsula on the western shore of Chickamauga Lake at Tennessee River Mile (TRM) 484.5. As shown by Figure 2.4.1-1, the site is on high ground with the Tennessee River being the only potential source of flooding.

SQN is located in the Middle Tennessee Chickamauga watershed, U.S. Geological Survey (USGS) hydrologic unit code 06020001, one of 32 watersheds in the Region 06 -Tennessee River Watershed (Figure 2.4.1-2).The Tennessee River above SQN site drains 20,650 square miles. The drainage area at Chickamauga Dam, 13.5 miles downstream, is 20,790 square miles. Three major tributaries-, Hiwassee, Little Tennessee, and French Broad Rivers,_-rise to the east in the rugged Southern Appalachian Highlands.

They flow northwestward through the Appalachian Divide which is essentially defined by the North Carolina-Tennessee border to join the Tennessee River which flows southwestward.

The Tennessee River and its Clinch and Holston River tributaries flow southwest through the Valley and ridge physiographic province which, while not as rugged as the Southern Highlands, features a number of mountains including the Clinch and Powell Mountain chains. The drainage pattern is shown on Figure 2.1.1-1. About 20 percent of the watershed rises above elevation 3 000 ft with a maximum elevation of 6,684 ft at Mt. Mitchell, North Carolina.

The watershed is about 70 percent forested with much of the mountainous area being 100 percent forested.The climate of the watershed is humid temperate.

Mean annual prccipitation for the Tonncsscc Valley is shown by Figuro 2.4.1 2. Above Chickamauga Dam, annual rainfall averages 51 inches and varies from a low of 40 inches at sheltered locations in the mountains to high spots of 85 inches on the southern and eastern divide. Rainfall occurs relatively evenly throughout the year. See Section 2.3 for a discussion of rainfall.

The lowest monthly average is 2.9 inches in October. The highest monthly average is 6.8 inches in March, with January a close second with an average of 6.0 inches.Major flood-producing storms are of two general types; the cool-season, winter type, and the warm-season, hurricane type. Most floods at SQN, however, have been produced by winter-type storms in the main flood-season months of January through early April.Watershed snowfall is relatively light, averaging only about 14 inches annually above the plant.Snowfall above the 3,000-ft elevation averages 22 inches annually.

The maximumhlighest average annual snowfall efin the basin is 63 inches GGeim at Mt. Mitchell, the highest point east of the Mississippi River. The overall snowfall average above the 3,000-foot elevation, however, is only 22 inches annually.

Individual snowfalls are normally light, with an average of 13 snowfalls per year.Snowmelt is not a factor in maximum flood determinations.

The Tennessee River, particularly above Chattanooga, Tennessee, is one of the most highly-regulated rivers in the United States. The TVA reservoir system is operated for flood control, navigation, and power generation with flood control a prime purpose with particular emphasis on protection for Chattanooga, 20 miles downstream from SQN.Chickamauga Dam, 13.5 miles downstream, affects water surface elevations at SQN. NermalSummer full pool elevation is 683.0682.5 feet. At this elevation the reservoir is 58.9 miles long on the Tennessee River and 32 miles long on the Hiwassee River, covering an area of 3&40036.050 acres, with a volume of 628,0622.500 acre-feet.

The reservoir has an average width of nearly 1 mile, ranging from 700 feet to 1.7 miles. At SQN-4the SQN site the reservoir is about 3,000 feet wide with depths ranging between 12 feet and 50 feet at normal pool elevation.

The Tennessee Rioer abo've Chattanooga, Tenne , .f the best regulated rives in the United 2.4-2 SQN-States. A prime purpose of the TVA water cntrol system 69 is flood, control#0, With pa.ticular emphasis On protection for Chattanooga, 20 milos, downstroam fromn SQN.There are 2-Q17 major in the TVA system from the plant, 13 Of Whi-h ha e&s.bstas ial reevdflood detention capacity during the m~ain flood season. Table 2.4.1 1 list pertinent data for TVA's mnajor dams prior to moedifications made by the Dam Safety Programn (see Table-2.4.-5 dams (South Holston, Boone, Fort Patrick Henry, Watauga, Fontana, Norris, Cherokee, Douglas, Tellico, Fort Loudoun, Melton Hill, Blue Ridge, Apalachia, Hiwassee, Chatuge, Nottely, and Watts Bar) in the TVA system upstream from SQN, 14 of which (those previously identified excluding Fort Patrick Henry, Melton Hill, and Apalachia) provide about 4.8 million acre-ft of reserved flood-detention (March 15) capacity during the main flood season. Table 2.4.1-2 lists pertinent data for TVA's dams and reservoirs.

Figure 2.4.1-3 presents a simplified flow diagram for the Tennessee River system. Table 2.4.1-3 provides the relative distances in river miles of upstream dams to the SQN site. Details for TVA dam outlet works are provided in Table 2.4.1-4. In addition, there are six major non-TVA dams, previously owned by the Aluminum Company of America (ALCOA). The ALCOA reservoirs often contribute to flood reduction, but were ignored in this analysis do not have dependable reserved flood detention capacity.

The lccations of these dams and the minor dams, Nolichucky and Walters (Waterville Lake), shawn on Fiure 2.1.1-1. Table 24.-122.4.1-5 lists pertinent data for the major and minor ALCOA dams and Walters Dam (Waterville Lake). The locations of these dams are shown on Fiaure 2.1.1-1.The flo-od d+etention capacity rese..ed in the TVA system arie. .....ally, with the greatest amounts during the flood.se.ason. 2.4.1 3, containing 14 sheets, shows tributa.y and ma.n.rve seasFon coprl atig guides fr thoese reservis having major ifuence on SON flood flows. Table 2.4.1 3 shows the fleouds cotrol reseratieon at the multiple PRipvce projecto above SON at the beginning and end of the wirtervie d season and in the summer. Assured -ystem detention capacity abeve the plart varies fuom 5.6 inches OR Jaruary 4to 1.5 incrhes on Muarch 15, decreasing to 1.0 in during the summer and fall. ActuaI detention capacity may exceed these amoneats, depending upon infloGws and powerdea~nds&.

Flood control above SON is provided largely by 4412 tributary reservoirs.

Tellico Dam is counted as a tributary reservoir because it is located on the Little Tennessee River, although, because of canal connection with Fort Loudoun Dam, it also functions as a main river dam. On March 15, near the end of the flood season, these provide a minimum of 44 004818500 acre-feet of detention capacity, equivalent to ximatey.5 inches on on the 4476approximately 19,500 square-mile area they control. This is 90 percent of the total available above Chickamauga Reservoir.

The two main river reservoirs, Fort Loudoun and Watts Bar, provide 490,000 acre-feet, equivalent to 4-.5proiael

.inches of detention capacity on the remaining area above the thaesChickamaua Dam.The flood detention capacity reserved in the TVA system varies seasonally, with the greatest amounts during the January through March flood season. Figure 2.4.1-4 (16 sheets) shows the reservoir seasonal operating guides for reservoirs above the plant site. Table 2.4.1-6 shows the flood control reservations at the multi ple-purpose proalcts above SON at the beginning and end of the winter flood season and in the summer. Total assured system detention capacity above Chickamauga Dam varies from approximately

5.5 inches

on January ito approximately 5 inches on March 15 and decreasing to approximately

1.5 inches

during the summer and fall. Actual detention capacity may exceed these amounts, depending upon inflows and power demands.Chickamauga Dam, the elevation of which affects flood elevations at the plant, has a drainage area of 20,790 square miles, 3,480 square miles more than Watts Bar Dam. There are seven maior tributary dams (Chatuge, Nottely. Hiwassee, Apalachia.

Blue Ridge, Ocoee No. 1 and Ocoee No. 3) in the 3,480-square-mile intervening watershed.

of which four have substantial reserved capacity.

On March 15, near the end of the flood season, these provide a minimum of 379,300 acre-ft eguivalent to 5.9 inches on the 1 ,200-sguare-mile controlled area. Chickamauga Dam contains 345,300 acre-ft of detention capacity on March 15 equivalent to 2.8 inches on the remaining 2,280 square miles. Figure 2.4.1-4 (Sheet 1) shows the seasonal operatina auide for Chickamauaa.

Elevation-storaqe relationships for the reservoirs above the site and Chickamauqa, downstream, are 2.4-3 SQN-shown in Figure 2.4.1-5 (17 sheets).Daily flow volumes at the plant, for all practical purposes, are represented by discharges from Chickamauga Dam with drainage area of 20,790 square miles, only 140 square miles more than at the plant. Momentary flows at the nuclear plant may vary considerably from daily averages, depending upon turbine operations at Watts Bar Dam upstream and Chickamauga Dam downstream.

There may be periods of several hours when there are no releases from either or both Watts Bar and Chickamauga Dams. Rapid turbine shutdown at Chickamauga may sometimes cause periods of Up-sr-eamreverse flow in Chickamauga Reservoir.

Based upon discharge records since closure of Chickamauga Dam in 1940, the average daily streamflow at the plant is 32,600 cfs. The maximum daily discharge was 223,200 cfs on May 8, 1984.Except for two special operations on March 30 and 31, 1968, when discharge was zero to control milfoil, the minimum daily discharge was 700 cfs on November 1, 1953. Flow data for water years 1951-1972 indicate an average rate of about 27,600 cfs during the summer months (May-October) and about 38,500 cfs during the winter months (November-April).

Flow durations based upon Chickamauga Dam discharge records for the period 1951-1972 are tabulated below.Average Daily Percent of Time Discharge, cfs Equaled or Exceeded 5,000 99.6 10,000 97.7 15,000 93.3 20,000 84.0 25,000 69.3 30,000 46.8 35,000 31.7 Channel velocities at SQN average about 0.6 fps under normal winter conditions.

Because of lower flows and higher reservoir elevations in the summer months, channel velocities average about 0.3 fps.As listed on Table 2.4.1-41, there are 23 surface water users within the 98.6-mile reach of the Tennessee River between Dayton, TN and Stevenson, AL. These include fifteen industrial water supplies and eight public water supplies.The industrial users exclusive of SQN withdraw about 497500 million gallons per day from the Tennessee River. Most of this water is returned to the river after use with varying degrees of contamination.

The public surface water supply intake (Savannah Valley Utility District), originally located across Chickamauga Reservoir from the plant site at TRM 483.6, has been removed. Savannah Valley Utility District has been converted to a ground water supply. The nearest public downstream intake is the East Side Utility (formerly referred to as U.S. Army, Volunteer Army Ammunition Plant). This intake is located at TRM 473.0.Groundwater resources in the immediate SQN site are described in Section 2.4.13.2.2.4.1.3 T-/A Dam Safety Proqram Most of the dams upstrcam from SON Were designed and built before the hydr metorgica.

approach to spillway design had gained- its. currenpt level of acceptance.

Spillway design capacity was generally less than weuld be provided today. The original FSAR analyses were based on the existing dam system before dam safety modifications wcrc made and included failure of som~e upstream dams from oVertopping.

In 1982, P/A officially began a safety review of its dams. The P/.IA DmS-afety Program was designed to be consistent with Fcderal Guidelines for Dam Safety and simil!ar efforts by ether Federal 2.4-4 SQN-agnis. T-echnical 6tudios and enginerin 1nay6,86 Were conRducted and physical mod~ifications.mplomon9td to onsure the hydFr4oloi anrd seismic inRtgrity of thA TVA dams and dm-nGStratc that P./..'s dans can boe .perat.d in accor.Pd.ance.

With Managem"ent Agoncy (FEMA)guidelines.

Table 2.4.1 5 provides the status of P/A. Damrn Safety AS Of 1 A98 These modifications enable these to safely pass the probable maximum flood. The raMining hyd~ologic-mo-dific-ations pla~nnd for Be-ar Crpeeak Da~m andd C-hic-kamauga Dam will not affect SQN i an. manner whiGh Might invalidate the reanalysis deScribed below.In 1997 98, TPA reanalyzed the nu.lcar plant design basis flood events. The purpose of the reanalysis was to evaluate the effects of the hydrologic dam safety modifications on the floo The following methods anRd assumptions wee applied to the FeaRalysis:

1. The computer programs and modeling methods were the sam speiusly used and do-cumenRted in the FS9A.R.2. Probable maimum timae distrib`ution of precipitaoniO, PreGipitation losses and rscR'oi operating procedures nwere unchanged fonm the original

3. The orFiginal stability analyses and postulated seismic dam failure assumptions woee GOnsepoatively assumed to occur in the same manner and in combination with the saepevosy RA-If- U;II rt---f-A -KI- M__ C-f-4,.u u. A UPPI: IAF W A Q RM i f; f I-.,- LI n i C ý4 14 T U;Ridne D~rmr Which dmntt thpir sCtrnwturl intporitv for ;; ~C;4Mire Pvnt Aih rpt.-rn n~riind of approXimately 10,000 years.* m 2 ii i i i i Ii A rle nlaRRed Fnoa1;iravGR 0" "hiGI-amauna r1arn aFFAGFIR the embankment te nQFFRi+ OVORG IRG)was cOnservatively assumed to have been implemented for the purpose Of calculating flEood effects. IUnder present existing conditions, the Chikamauga embankment would be eoroded- in the overtopping PMF event and the mxumfloo-d ele-vation at SON woul be lowerF than that With the miannedl modlification.

2.4.2 Floods

2.4.2.1 Flood Historyvt,-.ia4)

The nearest location with extensive formal flood records is 20 miles downstream at Chattanooga, Tennessee, where continuous records are available since 1874. Knowledge about significant floods extends back to 1826, based upon newspaper and historical reports. Flood flows and stages at Chattanooga have been altered by TVA's reservoir system beginning with the closure of Norris Dam in 1936 and reaching essentially the present level of control in 1952 with closure of Boone Dam, the last major dam with reserved flood detention capacity constructed above Chattanooga.

Tellico Dam provides additional reserved flood detention capacity; however, the percentage increase in total detention capacity above the Watts BarSequoyah site is small. Thus, for practical purposes, flood records for the period 1952 to date can be considered representative of prevailing conditions.

Table 2.4.2-1 provides annual peak flow data at Chattanoogqa.

Figure 2.4.2-1 shows the known flood experience at Chattanooga in diagram form. The maximum known flood under natural conditions occurred in 1867. This flood eaeGhedwas estimated to reach elevation 690.5 ft at SQN site with a discharge of about 450,000 cfs. The maximum flood elevation at the site under present-day regulation reached elevation 687.9 ft at the site on May 9, 1984.2.4-5 SQN-The following table lists the highest floods at SQN site under present-day regulation:

Estimated Elevation 7 at SQN (Feet)Discharge, at Chickamauga Dam (cfs)Date Before RcgulatiO.

March 11, 1867 6 0. 5 450,000 MaFerh , 1975 686.2 105,000 April 3, 1886 684.5 385,000 March 7, 1917 680.0 335,000 April 5, 1920 676.5 270,000 Since PreSent Regulation February 3, 1957 683.7 180,000 March 13, 1963 684.8 205,000 March 18, 1973 687.0 219,000 April 5, 1977 685.0 150,000 May 9,1984 687.9 250,000 April120.

1998 685.9 180,000 May 7, 2003 687.8 225,000 There are no records of floodina from seiches. dam failures, or ice Jams. Historic information about icing is provided in Section 2.4.7.2.4.2.2 Flood Design Considerations TVA has planned the SQN project to conform with regulatory position 2 of Regulatory Guide 1.59 including position 2.The types of events evaluated to determine the worst potential flood included (1) Probable Maximum Precipitation (PMP) on the total watershed and critical subwatershedssub-watersheds, including seasonal variations and potential consequent dam failures and (2) dam failures in a postulated Safe Shutdown Eathquake (SSE) or eae-half-SSEOBE with guide specified concurrent flood conditions.

Specific analysis of Tennessee River flood levels resulting from ocean front surges and tsunamis is not required because of the inland location of the plant. Snow melt and ice iam considerations are also unnecessary because of the temperate zone location of the plant. Flood waves from landslides into upstream reservoirs required no specific analysis, in part because of the absence of maior elevation relief in nearby uostream reservoirs and because the orevailino thin soils offer small slide volume potential compared to the available detention space in reservoirs.

Seiches pose no flood threats because of the size and configuration of the lake and the elevation difference between normal lake level and plant grade.The computed maximum stillwatermaximum PMF plant site flood level in the reserVoir at the plant site fFGM aRY-ae is elevation 741.6722.0 ft. This elevation would result from the PMP critically centered on the watershed as described in Section 2.4.3.Maxim'um le-Vel including wave height is 722.4. This elevation would result from the probable maximum nrecinitntion critica!lv centered on the waters'hed and a Wind waves based on an overland wind speed of 45 Mile per hour ..e...... indF from the most critical direction miles per hour were assumed to occur coincident with the flood peak of the resulting flood. This would create maximum wind waves up to 4.2 ft high (trough to crest).All safety related facilities syste~ms and eoiiinme.nt are hnuisped in stnr'ctures whinh nrnvide nrntection from flooding for all flood conditions up to plant grade at elevation 705.0 ft. See Section 2.4.10 for more specific information.

Other rainfall floods will also exceed plant grade 7 elevation 705.0 ft- and will nocessitaterequire plant shutdown.

Flood warning cr!teria and forecasting techniques have been developed to assure that 2.4-6 SQN-grade and are deScrib.,d in Subsections 2.4.10 and 2.4.14, and Appendix 2.4A.Section 2.4.14 describes emeraencv protective measures to be taken in seismic events exceedina plant arade.Seismic and GenRu-rent-flood events could create flood levels Which weould excedcause dam failure surges exceeding plant grade elevation 705.0 ft. The maximum elevation reached in such ar event is elevation 707.9, 2.9 feet above plant grade and 11.7 feet below the controlling event probal maximum flood (PMF), eXcluding Wind wave conRsiderationAs.

In all s.uch e-Vents. therFe is adequate time-fer safe plant shutdowA~n aafte-r the seismicG event and before plant grade would be crossed. The emnergencY protective mneasures and warning criteria are desc'ribed i'n Sub-sections 2.4.10 and 2.4.14, and Appendix 2.4A.Section 2.4.14 describes emergency protective measures to be taken in seismic events exceeding plant grade.Most safety related bWuiding accesses are flocatedi at elevatioan 706 orabeove.

The acnessecs below elevation 706 afe withir the powerhouse and Will not be exposed to flooedwater util plant grade is exceeded.

Therefore, the structres are protected fromm floedingprior touthce hnd of the cheutdoW pe~ied.Drainage to the Tennessee River has been provided te accommoAdate-runoff from the probal Specific aralysis of Tennessee River flood levels resulting from oceanfrnet surges and tunamis is Rte reqed beGause of the inland location Of the plant oWmeltand ice jamu cursiderations ate alDs uneGecessary because of the temperate zone location Of the plant. Floo~d waves from landslides"E inoupstreamn reser.'eirs required no specific analysis, in pa~t bewicauhs Of th absve nhe of major elevatior relief int earby upstream run uwils and because tfet pfevailing thin soils ofer small slide volume potential cempared to the available detentio ae inte All safety related facilities, systems, and equipment are housed in structures which provyide protectionR fromR flooding forF all flood conditions up to plant grade at elevation 705.For the condition where flooding exceeds plant grade, as described In Subsections 2.4.3 and 2.4.4, all equipment required to maintain the plant safely duFrig the flood, and for 100 days after the beginning of the flood, is either designed to operate submerged, located above the maximum flood level, or other'.iso protected.

Safety Felated For the condition where flooding exceeds plant grade, as described in Sections 2.4.3 and 2.4.4, those safety-related facilities, systems, and equipment located in the containment structure are protected from flooding by the shield buildfing.

AlShield Buildingq structure with those accesses and penetrations below the maximum flood level in the s-hield- building are designed and constructed as wateF tkjtwatertight elements.Wind wave run up during the PMF at the Diesel Generator Building would reach elevation 723.2 ft which is 1 .2 ft above the operating floor. Conseguently, wind wave run up will impair the safety functions of the Diesel Generator Buildino.

The accesses and penetrations below this elevation in the Diesel Generator Building are designed and constructed to minimize leakage into the building.Redundant sump pumps are provided within the buildinq to remove minor leakage. Protective measures are taken to ensure that all safety-related systems and equipment in the Emergency Raw Cooling Water (ERCW) Intake Pumping Station will remain functional when subjected to the maximum flood level.Those Class 1 E electrical system conduit banks located below the PMF plus wind wave run up flood level are designed to function submerqed with either continuous cable runs or qualified, type tested splices.The turbine, cont.ro, and auxilia.y build...Turbine, Control, and Auxiliary Buildings will be allowed to flood. All equipment required to maintain the plant safely during the flood, and for 100 days after the 2.4-7 SQN-beginning of the flood, is either designed to operate submerged, is located above the maximum flood level, or is otherwise protected.

Wind WaVe FRu up during the PMF= at the diesel generator building reaches olevation 7-24.8 which is 0.2 feet below the operating flori .C-Rsoquently, Wind wave rUn up will not impair the safety funRction of systems in the diesel generator building.The accesses and penetratiodns exbelw this elevation iR the diesel generator buMildig are designed and conStruoted to minimize leakage into the buildings.

Redundart sUMp pumcps ae pfrovided within te building to remove minor leakage. roterative mneasures acr takenon to enue that all safety related systems and equipmnt in the Emrergeny Raw Coolinqg Water (ERvW) pumt p station Will remstain functional when subjected to the maximum flnod level.Class E electrigal cables, lo-cated belo the Proebsable Maximumen Flood (PMF) plus wind wave acivgi and requiFed in a flood, are designed for submrerged opeiratio.

2.4.2.3 Effects of Local Intense Precipitation Maximum water levels at buildings expected to result from the local plant PMP were determined using two methods: (1) when flow conditions controlled, standard-step backwater from the control section using peak discharges estimated from rainfall intensities corresponding to the time of concentration of the area above the control section or (2) when ponding or reservoir-type conditions controlled, storage routing the inflow hsdrograph equivalent to the PMP hydrograph with 2-minute time intervals.

Structures housing safety-related facilities, systems, and equipment are protected from flooding during a local PMF by the slope of the plant yard. The yard is graded so that the surface runoff will be carried to Chickamauga Reservoir without exceeding the elevation of the external accesses given in Paragraph 2.4.1.1 except those at the intake pumping station whose pumps can operate submerged.

PMP for the plant drainage system and roofs of safety-related structures was determined from Hydrometeorological Report No. 45 [21. The probable maximum storm used to test the adeguacy of the local drainage system would produce 27.5 inches of rainfall in six hours with a maximum one-hour depth of 14 inches. Depths for each of the six hours in sequence were 1.5, 2.3, 5.0. 14.0, 3.0, and 1.7 inches.The separate watershed subareas and flowpaths are shown on Figure 2.4.3-22.Runoff from the 24.5 acre western plant site will flow either northwest to a 27-foot channel along the main plant tracks and then across the main access highway or to the south over the swale in Perimeter Road near the 161-ks switchyard and across Patrol Road to the river. Because the 500-kv switchyard and TEACP building areas are essentially level, peak outflows from this subarea were determined using method (2). These peak outflows were then combined with discharge estimates from the remaining areas, using method (1), to establish peak water surface profiles from both the north channel and south swale. The maximum water surface elevation is below critical floor elevation 706 and occurs near the east-west centerline of the Turbine Building.The 28.9 acre eastern plant site was evaluated as two areas. Area 1 (19.7 acres) including the diesel generator, unit two reactor building, field services/storage buildings and adiacent areas. Runoff from area 1 will flow to the south along the perimeter road and across the pavement with low point elevation 705.0 ft to the discharge channel. Maximum water surface elevations computed using method (1)were less than elevation 706.0 ft. Area 2 (9.2 acres) includes the office/service, unit one reactor building, office/power stores buildings, intake pumping station, and adiacent areas. Runoff from area 2 will flow to the north and west along the ERCW pumping station access road to the intake channel and river. Maximum water surface elevation computed using method (2) is less than elevation 706.Underground drains were assumed clogged throughout the storm. For fence sections, the Manning's n value was doubled to account for increased resistance to flow and the potential for debris blockage.2.4-8 SQN-2.4.3 Probable Maximum Flood (PMF) on Streams and Rivers The guidance of Appendix A of Regulatory Guide 1.59 was followed in determining the PMF.-Plan surface drainage was evaluated and found capable of passing the locGal probable maximu Ftr without FrGchiRg Or eXceeding the floor elevation 706, as fu-ther described, in 2-.235, Evaluation of areal variations Of prohbable maximum sterms showed that the probable-ocring in March centered in the mountains, east of the plant. The flood crest at the plant would be ,hximum, disc.harge is 1,236,000

.fS. The probable maximum el1va a e n 719.6, Dam downstream.Two basic storm situations were found to have the potential to produce maximum flood levels at SON. These are (1) a sequence of storms producing PMP depths on the 21,400-square-mile watershed above Chattanooga and (2) a sequence of storms producing PMP depths in the basin above Chattanooga and below the five major tributary dams (Norris, Cherokee, Douglas, Fontana, and Hiwassee), hereafter called the 7,980-square-mile storm. The maximum flood level at the plant would be caused bv the March PMP 21,400-square-mile storm. The flood level for the 7,980-square-mile storm would be slightly less.In both storms, the West Saddle Dike at Watts Bar Dam would be overtopped and breached.

No other failure would occur. Maximum discharge at the plant is 1,331,623 cfs for the 21,400-square-mile storm. The resultinq PMF elevation at the plant would be 722.0 ft excludinq wind wave effects.2.4.3.1 Probable Maximum Precipitation Probable maximum precipitation (PMP) for the Tennessee River watershed above SQN has been defined for TVA by the Hydrometeorological Branch of the National Weather Service in Hydrometeorological Report No. 41 Reference

[1]. Two basic storm positions were evaluated.

One.would produce maximum rainfall over the total watershed.

The other would produce maXi the part of the basin downstreamn fromn major TVA tributary reservoirs, hereafter referred to asth 7,980 square milo storm. Snowme#tThis report defines depth-area-duration characteristics, seasonal variations, and antecedent storm potentials and incorporates orographic effects of the Tennessee River Valley. Due to the temperate climate of the watershed and relatively light snowfall, snowmelt is not a factor in generating maximum floods for the Tennessee River at the plant site.Centrsolling PwouP depths for 21,00 square mile and 7,980 square mile areas are tabulated below.72 HouJr Main Storm Si. Moiles Antecedent Storm. 6 H-ou 21 Hour 72 Ho-u, 21,400 6.7 5-0-3 11.18- 16.70 7,980) 8-1 7.02 14.01 20.36-Two basic storms with three possible isohyetal patterns and seasonal variations described in Hydrometeorological Report No. 41 were examined to determine which would produce maximum flood levels at the SQN site. One would Droduce PMP deoths on the 21.400-sauare-mile watershed above Chattanooga.

Two isohyetal patterns are presented in Hydrometeoroloqical Report No. 41 for this storm. The isohvetal pattern with downstream center would produce maximum rainfall on the middle portion of the watershed and is shown in Figure 2.4.3-1.Two possible isohyotal patterns producing the total area depths are presented in Report No. 11 .The second storm described in Hvdrometeoroloaical Reoort No. 41 would oroduce PMP deoths on the 7,980-square-mile watershed above Chattanooga and below the five major tributary dams. The-eae to this study is the "downstream pattern" shown in Figure 2..3- 1 The isohyetal pattern for the 7,980-square-mile storm is shown in Figure 2.4.3 2. The pattern is not Orographicallygeographically 2.4-9 SQN-fixed and can be moved parallel to the long axis, northeast and southwest, along the Tennessee Valley. The isohyetal pattern centered at Bulls Gap, Tennessee, would produce maximum rainfall on the uoDer oart of the watershed and is shown in Fioure 2.4.3-2.~"~'~ir ~'vvm 'nrn" 'r1"~' '~iflTr~r.nnnflT T'~ PM' ~'gIP ~r"~m '~'g~ '~'~" '"~ r'~' ""'~' "" '~*with steor All PMP storms are nine-day events. A three-day antecedent storm was postulated to occur three days prior to the three-day PMP storm in all PMF determinations.

Rainfall depths equivalent to 40 percent of the main storm were used for the antecedent storms with uniform areal distribution as recommended in Reoort No. 41.Potential sto~rm amounts differing by seasons were analyzed in SUfficient number to Make ccrtain that the March storms would be controlling.

Eno~ugh centeFRngs wcrc investigated to assure that a most position was used.Seasonal variations were also considered.

Table 2.4.3-1 provides the seasonal variations of PMP. The two seasons evaluated were March and June. The March storm was evaluated because the PMP was maximum and surface runoff was also maximum. The June storm was evaluated because the June PMP was maximum for the summer season and reservoir elevations were at their highest levels. Although September PMP is somewhat higher than that in June, less runoff and lower reservoir levels more than compensate for the higher rainfall.Storms producinig PMP abov. upstream tributaFry dams, ,hose failure has the to create maximm flood levels, were evaluated in the original FSAR analysis.

Dam safety moedifications at upstream tributaFry dams have eliminatod these potential failures aRnd subsequent plant site flood A standard time distribution pattern was adopted for alithe storms based upon major observed storms transposable to the Tennessee Valley and in conformance with the usual practice of Federal agencies.The adopted distribution is shewn on Figure 2.4.3 3 within the limits stipulated in Chapter VII of Hydrometeorological Report No. 41. This places the heaviest precipitation in the middle of the storm.The adopted sequence closely conforms to that used by the U.S. Army Corps of Engineers.

A typical distribution mass curve resulting from this approach is shown in Fiqure 2.4.3-3.The PMF discharae at SON was determined to result from the 21.400-souare-mile storm Droducina PMP on the watershed with the downstream storm pattern, as defined in Hydrometeorological Report No. 41. The PMP storm would occur in the month of March and would produce an average of 16.25 inches of rainfall in three days on the watershed above Chickamauga Dam. The storm producing the PMP would be preceded by a three-day antecedent storm producing an average of 6.18 inches of rainfall, which would end three days Drier to the start of the PMP storm. Precioitation temporal distribution is determined bv aDDIvina the mass curve (Fiaure 2.4.3-3) to the basin rainfall depths in Table 2.4.3-2.The critical probable Maximum stoFrm was deAtermined to be A total basin6 s~to*rm with downstreamn eregraphically fixed pattern (Figure 2.4.3 1) which would fellowA an antecedent stoFrm comm~encing On March 15. Translatien ef the PMP from Report Ne. 4 1 to the basin rosults in an antecedent storm producGing an aege preiitation of 6.4 inches in three days, followed by a three day dry period,an t h en by t hPe man Pstorm~ producing an average precipitation of 16.45 inc-hes inA three days. Figure 2.4.3 4 is an iehyetal map of the three day PhMP. Bar sin rainfall depths are given in Table 2.4.3 1.PMP for the plant drainage system and roofs of safety related structures was deteFrmined froM Hydrmeteor'elgc3al Report No. 45 [2]. The p"r bable maximum storm used to test the adequaGy of the IGal ,drai.age system would -27.5 inches f rainfall in six hours with a maximu.m one hou depth of 14 inchcs. Depths for each of the six hour ien sequence e 1.5, 2.3, 3.0, and 1.7 2.4.3.2 Precipitation Losses F-FeGlpitatIGR lesses IR tHe pFebable maximum stoFm aFe esumatedWKH multivariable Felati()n6HIPS us. Aedh in the day to day operation of the TVA system. There relationships, develeped from a study of ft a" ýWl- M, f-M C3 -, "- " W " W C1 W" "-Atýwan am -wttm to C2 V" -"--/ -2.4-10 SQN-the; week of the year, an antoccdonlt procipitation index (API), and geographic location.

The FelationGhips ar0 suc.h thit the Io6. subtraction from rainfall to compute precipitatioex is greatest at thp nf the sto"rm and decreases to Ro subtFactin When the rainfall totals from 7 to 16 inches.

losses become zer ORi the late part of ex(trem% e For this probable maximum flood analysis, m.edian m.oisture onditions as determined frm past records were used to determine the API at the start of the storm sequence.

The artecedet hs is so large, however, that the precipitation excess computedr fer the later main stom irot sensitive4t variations in adopted initial moisture conditions.

The precipitation loss in the critical probable m Aimm stor~m totals 1.13 inches, 2.30 inches in the antccedcnt storm amonOUting to 36 percent ot the 3 day 6.44 inch rainfall, and 1.83 inches in the mnain stoFrm amRounting toll1 percent of the 3 day, 16.46 1nch rainfall.

Table 2.4.3 1 displays the API, rain, and precipitation excGess for: each of the 1 5 subwatersheds of the hydrologic model forF the SQN probable maximum floed.No precipitation loss; was applied in the pro-bable ma;ximumA stormB o the local area used to test th adequacy of the site drainage system and roofs of safety rolated struc~tures.

R-unoff w..as made equa tO FainfalU, A multi-variable relationship, used in the day-to-day operation of the TVA reservoir system, has been aoolied to determine orecioitation excess directly.

The relationshios were develoned from observed storm and flood data. They relate orecioitation excess to the rainfall, week of the year. aeoaraohic location, and antecedent precipitation index (API). In their application, precipitation excess becomes an increasinq fraction of rainfall as the storm prowresses in time and becomes equal to rainfall in the later part of extreme storms. An API determined from an 11-year period of historical rainfall records (1997-2007) was used at the start of the antecedent storm. The precipitation excess computed for the main storm is not sensitive to variations in adooted initial moisture conditions because of the larae antecedent storm.Basin rainfall, precipitation excess, and API are provided in Table 2.4.3-2. The average precipitation loss for the watershed above Chickamauga Dam is 2.33 inches for the three-day antecedent storm and 1.86 inches for the three-day main storm. The losses are approximately 38% of antecedent rainfall and 11% of the PMP. resoectivelv.

The orecioitation loss of 2.33 inches in the antecedent storm compares favorably with that of historical flood events shown in Table 2.4.3-3.2.4.3.3 Runoff Model The runoff model used to determine Tennessee River flood hydrographs at SQN is divided into 4540 unit areas and includes the total watershed above Chickamauga Dam. Unit hydrographs are used to compute flows from these areas. The watershed unit areas are shown in Figure 2.4.3-5. The unit area flows are combined with appropriate time sequencing or channel routing procedures to compute inflows into the most upstream tributary reservoirs, which in turn are routed through the reservoirs, using standard routing techniques.

Resulting outflows are combined with additional local inflows and carried downstream using appropriate time sequencing or routing procedures, including unsteady flow routina. Figu-re 2.4.3 5 shows u-nit areas of the watershed uastream from SON.'rk 4FF A 1 A ; fk; .4 + .4 CC A Q A;ff- f k A i.-.-.-.-..-.--..............

~......made in some elements of the m.del during PMF. studios fo.r other n.ear plants and those made from inforsmatio gained from the 1973 flood, the largest that has- occured during present reservoir GG~ditiORG&

Changes are identified when appropriate in the text. They include both additional and revised unit hydroGQaphs and additional and revised unsteady flew stream; course models.Unit hydrographs were developed for each unit area for which discharge records were available from maximum flood hydrographs either recorded at stream n stations or estimated from reservoir headwater elevation, inflow, and discharge data using the procedures described by Newton and Vineyard Reference

[231. For non-gaged unit areas synthetic unit graphs were developed from relationships of unit hvdrographs from similar watersheds relatinQ the unit hvdrograph peak flow to the 2.4-11 SQN-drainaae area size. time to oeak in terms of watershed slooe and lenoth. and the shade to the unit hydrograph peak discharge in cfs per square mile. Unit hydrograph plots are provided in Figure 2.4.3-6 (11 Sheets). Table 2.4.3-4 contains essential dimension data for each unit hydrograph.T-he number of u-nit areA hasb in crea..d frorm 3 used previou.ly to 15. The d,4e...r.on.e include: 1. Use of the model developed for the Phipps Bend study Which cOmbined the wo uRnit areas for W.1atauga River (Sugar Grovye and Watauga local) into one Unit area and divided the Chroeet Gate Gity uR~tarea into two unit areas (SurgoinSVille locnal and- Cheroekee local below SUrgoinsville);

2 UsIeA of the model,4 developed for the Clinch River Reactor which increased the unit areas On the Clinch River from 3 to 11 and the Watts Bar locGal from 1 to 2;3. Changes to add an unsteady flow model for the PFot Loudoun T-ellico DamA complex which included dividi4ng the lower L-ittle TeAnnesseet Rivemr unit Area inotoui ra (Fontaqna to Chilhowee and Chilhowoe toRlioad the Fort Loudoun locGal unit area into three Unit areas (FrFenchBra River Iocal, HolstOn RiveFr Ioal aRnd Fot Ioal); 4. Combining the te unit areas above ..c.e No. 1 (Oce No. 1 Coh.e No. 3) int on unit area (Ocoee No. 1 to Blue Ridge).In addition, eight of the unit graphs have bocn revised. Fmigure 2.1.3 6, which contains 11 sheet-s, shows the unit HY9Froywph.

Table 2.4.3 Lo GEuflati esu and identification of those hydregraphs which are new or-+; I i S H- un d r ia" H y ~ F"S ri Tributary reservoir routings, except for Tellico and Melton Hill, were made using the-GGGd~iGh

...th.dstandard reservoir routing procedures and flat pool storage conditions.

M44The main river reservoirs.

a.id-Tellico, and Melton Hill routings were made using unsteady flow techniques.

This dffe-rs from the previous in that: 1. An unsteady flow model has been added for the Fort Loudoun Tellico com..lex, and 2. The Chickar from tho HEC 2 I I I I I IL am] inn I If.--I- -hRrkvP;;t 4 4oW mRoo8t Ras P88R revksee using Me 4913 Nood data ana resuims u-.tpr Amroornm.In the original study, the failure wave hydrograph of the moeuth of the Hiwassee River was approximate, for the postulated failures of Hiwasee, Apalachia and Blue Ridge dams a, r described in2.4.4.2.1.

In the eassessment,an unRsteady flow model developed dUring the dam.afety studios was used as an adjunct to rou..tel the HiW..assee, Apalachia and Blue Ridge failures in the one half SSE. The model was verified by comparing model elevations in a state of steady flow with eleatinn-.omnutd b h~e staprnda;rd stpm method. ýThis; %WaS dAne for s~tpadv flow-si r~noino fromn 25,000 cfs to 1,000.000 cfs.ýj ý7ý ý ý7;ýý ýUnsteady flow routings were computer-solved with athe Simulated Open Channel Hydraulics (SOCH)mathematical model based on the equations of unsteady flow, [3]. Boundary conditions proscribed tiOW HYUFOgFaphs at the 61P6tFeaFA GOURGaFy, GGal IR 9 , aA HeadwateF tjtSGHaFge.j-.---...---...--...

.---...--..

J -~----I-.~.-.-I--.-...,~.-.----------

cuPres when geometry cGntrolled.The SOCH model inputs include the reservoir geometry, upstream boundary inflow hydrograph, local inflows, and the downstream boundary headwater discharge relationshios based uoon ooeratino auides or ratina curves when the structure aeometrv controls.Seasonal operating curves are provided in Figure 2.4.1-4 (16 Sheets).Discharge rating curves are provided in Figure 2.4.3-7 (17 Sheets) for the reservoirs in the watershed at and above Chickamauga.

The discharge rating curve for Chickamauga Dam is for the current lock configuration with all 18 spillway bays available.

Above SQN, temporary flood barriers have been installed at four reservoirs (Watts Bar. Fort Loudoun. Tellico and Cherokee Reservoirs) to increase the height of embankments and are included in the discharge rating curves for these four dams.Increasing the height of embankments at these four dams prevents embankment overflow and failure of the embankment.

The vendor supplied temporary flood barriers were shown to be stable for the most severe PMF headwater/tailwater conditions using vendor recommended base friction values. A 2.4-12 SQN-sin-gle postulated Fort Loudoun Reservoir rim leak north of the Marina Saddle Dam which discharges into the Tennessee River at Tennessee River Mile (TRM) 602.3 was added as an additional discharge component to the Fort Loudoun Dam discharge rating curve. Seven Watts Bar Reservoir rim leaks were added as additional discharge components to the Watts Bar Dam discharge rating curve. Three of the rim leak locations discharge to Yellow Creek, entering the Tennessee River three miles downstream of Watts Bar Dam. The remaining four rim leak locations discharge to Watts Creek, which enters Chickamauga Reservoir iust below Watts Bar Dam. A single postulated Nickajack Reservoir rim leak lust northeast of Nickalack Dam and back into the Tennessee River below Nickaiack Dam was added as an additional discharge component for the Nickaiack Dam.The unsteady flow mathematical model for the 49.9 mile long Fort Loudoun Reservoir was divided nto twenty four 2.08 m~ile reaches. The moedel was verified at thrce gauged points within Fort Loudoun Reser'oir using 1963 and 1973 flood data configuration for the Fort Loudoun-Tellico complex is shown by the schematic in Figure 2.4.3-8. The Fort Loudoun Reservoir portion of the model from TRM 602.3 to TRM 652.22 is described by 29 cross-sections with additional sections being interpolated between the original sections for a total of 59 cross-sections in the SOCH model, with a variable cross-section spacing of about 1 mile. The unsteady flow model was extended upstream on the French Broad and Holston Rivers to Douglas and Cherokee Dams, respectively.

The French Broad and Holston Rivet unsteady flow moedels werc verified at one gaged point each at m~ile 7.4 and 5.5, respectively, using 1963 and 1973 flood data River from the mouth to Douqlas Dam at French Broad River mile (FBRM)32.3 was described by 25 cross-sections with additional sections being interpolated between the original sections for a total of 49 cross-sections in the SOCH model, with a variable cross-section spacing of about 1 mile. The Holston River from the mouth to Cherokee Dam at Holston River mile (HRM) 52.3 was described by 29 cross-sections with one additional cross-section being interpolated between each of the original sections for a total of 57 cross-sections in the SOCH model, with a variable cross-section spacing of about 1 mile.The Little Tennessee River was modeled from Tellico Dam, mile 0.3, through Teo!co Reservoir to COhilho...oe Dam at mAile 33.6. a d upstream Fn-to FonA"tanAR DamR at mAi 6.0. The mode1l AIh,,.'for Tellico Resep.vir to C-1hilho..ee Dam was tested for, adequacy by .cm"paring its result. with steady state P.! #innnfAr n n .4 k 4 ~ARA; A .conveyance in the unsteady flow moedel yielded good agreement.

The average conveyance correctic R.U.E. neceesa,'y In t. .a. he.. ..o. -h.....ee Quamn to make the unsteady flow model agree wth then standaFrd step methd was also used in the river reach fro.m Ghilhowee to Fontana Dam Little Tennessee River mile (LTRM) 0.3 to Chilhowee Dam at LTRM 33.6. The Little Tennessee River from Tellico Dam to Chilhowee Dam at LTRM 33.6 was described by 23 cross-sections with additional sections being interpolated between the original sections for a total of 49 cross-sections in the SOCH model, with a variable cross-section spacing of up to about 1.8 miles.T-he-Fort Loudoun and Tellico unsteady flow models wefeare joined by a canal unsteady flow model.The canal was modeled with five equally spaced cress Sectin6 at 525 foot itrasfoMr the 2,100 foot long canal an interconnecting canal. The canal was modeled using nine cross-sections with an average cross-section spacing of about 0.18 miles.The Fort Loudoun-Tellico complex was calibrated by two different methods as follows: (1) Using the available data for the March 1973 flood on Fort Loudoun Reservoir and for the French Broad and Holston rivers. The calibration of the 1973 flood is shown in Figure 2.4.3-9 (2 Sheets). Because there were limited data to verify against on the French Broad and Holston rivers, the steady-state HEC-RAS model was used to replicate the Federal Emergency Management Agency (FEMA) published 100- and 500-year profiles.

Tellico Dam was not closed until 1979, thus was not in place during the 1973 flood for calibration.

(2) Using available data for the May 2003 flood for the Fort Loudoun-Tellico complex. The calibration of the May 2003 flood is shown in Figure 2.4.3-10 (3 Sheets). The Tellico Reservoir steady-state HEC-RAS model was also used to replicate the FEMA published 100-and 500-year profiles.A schematic of the steadv-state SOCH model for Watts Bar Reservoir is shown in Fiqure 2.4.3-11.2.4-13 SQN-The unsteady flow routing model for the 72.4-mile-long Watts Bar Reservoir was divided into thilry four 2.13 mnilo roaches. The moedel was verified At b o gaugod points within the reservoir using 1963 flood datadescribed by 39 cross-sections with two additional sections being added in the upper reach for a total of 41 sections in the SOCH steady state model with a variable cross-section spacing of up to about 2.8 miles. The model also includes a iunction with the Clinch River at Tennessee River mile (TRM) 567.7. The Clinch River arm of the model goes from Clinch River mile (CRM) 0.0 to CRM 23.1 at Melton Hill Dam with one additional section being interpolated between each of the original 13 sections and cross-section spaces of up to about 1 mile. Another iunction at TRM 601.1 connects the Little Tennessee River arm of the model from the mouth to Tellico Dam at LTRM 0.3 with cross-section spaces of about 0.08 miles. The time step was tested between 5 and 60 seconds which produced stable and comparable results over the full range. A time step of 5 seconds was used for the analysis to allow multiple reservoirs and/or river segments to be coupled together with different cross-section spacing. The verification of Watts Bar Reservoir for the March 1973 and the May 2003 floods are shown in Figure 2.4.3-12 and Figure 2.4.3-13, respectively..

TheA schematic of the unsteady flow mathematical model for Chickamauga Reservoir is shown in Figure 2.4.3-14.

The model for the total 58.9-mile-long Chickamauga Reservoir was di~ided-OitO

..enty eight 2.1 mile reaches prviding twenty nine equally spaced grid points. The grid point at mile 483.62 is nearest to the plant, mnile 481.5. The unsteady fleow mo-delviwas verified at four gauged poit W'ithin Chickamauga Reservier using 1973 flood data. This differs from the previous submission in that the 19732lo a de fner vierific-ation, replacing the 1963 flood. The 1973 flood occurred drn preparation of the FSAR and therefore, was not available fo-r ve~rification.

The 1973 floo)d is the !argest;.ohic-h has occ-rurre-d si.nce, closure of SoAuth HAolrstPn -Da~m in 19A50. Comparisons between observed and computed stages in Chickamauga Rese.voir are shown in Figure 2.4.3 7described by 29 cross-sections with one additional section being interpolated between each of the oriainal 29 sections for a total of 53 sections in the SOCH model with a variable cross-section spacing of up to about 1 mile. The model also includes a *unction with the Dallas Bay embayment at TRM 480.5. The Dallas Bay arm of the model goes from Dallas Bay mile (DB) 5.23 to DB 2.86, the control point for flow out of Chickamauga Reservoir.

Another 'unction at TRM 499.4 connects the Hiwassee River arm of the model from the mouth to the Charleston gage at HRM 18.9. The time step was tested between 5 and 50 seconds producing stable and comparable results over the full range. A time step of 5 seconds was used for the analysis to allow multiple reservoirs and/or river segments to be coupled together with different cross-section spacing. The verification of Chickamauga Reservoir for the March 1973 and the May 2003 floods are shown in Figure 2.4.3-15 and Figure 2.4.3-16, respectively.

it is impossible to verif' the modelsVerifying the reservoir models with actual data approaching the magnitude of the probable maximum..

flooed. The- remaining alternative was to compare the 6 4i a state of steady flow with elevationS omputed by the step mnethod. This was-done for steady fliw rang-Ingu to 1,500,000 cfs. An example shown by the rating curve of Figure 2.4.3 A shows the good ag.reement PMF is not possible, because no such events have been observed.

Therefore, using flows in the magnitude of the PMF (1,200,000

-1,300,000 cfs), steady-state profiles were computed using the HEC-RAS [241 steady state model and compared to computed elevations from the SOCH model. An example of the comparison between HEC-RAS and SOCH profiles is shown for Chickamauga Reservoir in Figure 2.4.3-17.

This approach was applied for each of the SOCH reservoir models. Similarly, the tailwater rating curve was compared at each project as shown for Watts Bar Dam in Figure 2.4.3-18.

In this figure, the initial tailwater curve is compared to results from the HEC-RAS or SOCH models.The reservoir operating guides applied during the SOCH model simulations mimic, to the extent possible, operating policies and are within the current reservoir operating flexibility.

In addition to spillway discharge, turbine and sluice discharges were used to release water from the tributary reservoirs.

Turbine discharges were also used at the main river reservoirs up to the point where the head differentials are too small and/or the powerhouse would flood. All discharge outlets (spillway gates, sluice gates, and valves) for proiects in the reservoir system will remain operable without failure up to the point the operating deck is flooded for the passage of water when and as needed during the flood. A high confidence that all gates/outlets will be operable is provided by periodic inspections by TVA plant personnel, the intermediate and five-year dam safety engineering inspections consistent with Federal Guidelines for Dam Safety, and the significant capability of the emergency response teams to direct and manage resources to address issues potentially impacting gate/outlet functionality.

2.4-14 SQN-Median initial reservoir elevations for the appropriate season were used at the start of the PMF storm sequence.

Use of median elevations is consistent with statistical experience and avoids unreasonable combinations of extreme events.The flood from the antecedent storm occupies about 70% of the reserved system detention capacity above Watts Bar Dam at the beginning of the main storm (day 7 of the event). Reservoir levels are at or above guide levels at the beginning of the main storm in all but Apalachia and Fort Patrick Henry Reservoirs, which have no reserved flood detention capacity.The watershed runoff Model was verified by using it to reproduce the March 1963 and March 1973 floods; the largeSt closure of South Helston Dam. This from the submission in that the 1973 floodM F ia s ddeded fo verificatioR, reping the 1957 flood. Obsen1ed volumes Of precipitation texess were used in Vreifiation.

Comparchisons beaweeowbsterVed and foxputed outflows from Wals Bar and Chickamauga Dams for the 1973 and 1963 floods arc shown in FigUrcs 2.4.3 9 and 2.4.3 10, respectively.

From~ a study of the basic unis of the predicting system and its response to alteratiosivaou basic elements, iiscnlddth-at the Model servýes adequately and conservatively to determinRe mnaximum flood levels.2.4.3.4 Probable Maximum Flood Flow The probable maximum floodPMF discharge at SQN was determined to be I 0001W1 331623 cfs.This flood would result from the 21 ,400-sguare-mile storm in March with a downstream orographically-fixed storm Dattern (Fioure 2.4.3-1).The PMF discharge hydrograph of this flood is shown in Figure 2.4.3-4419.

This flood would result I tir" 0 cy CI OCY M ww"o rvczm URY9 p va y Av 5 OFF" Pa .Or '-. ga ..,r"" "Wr" I t, dosc,"bcd in Soetion 2.1...! Tho dam safet'-. modification to Fort Lodon. Teiiice. and Watts Bar Dams enable them to safely pass the PMF=. The west saddle dikeThe West Saddle Dike at Watts Bar Dam (Figure 2.4.3-20) would be overtopped and the earth embankment breached.

The discharge from the failed West Saddle Dike flows into Yellow Creek which ioins the Tennessee River at mile 526.82, 41.82 miles above SQN.Chickamauga Dam downstream would be overtopped but was assumed net to fail as a failure would reduce the flood level at the site. The dam was oostulated to remain in olace, and any notential lowering of the flood levels at SQN due to dam failure at Chickamauga Dam was not considered in the resulting water surface elevation.

In the original FSAR analysis, the flood would overtop and breach the earth emnbarkmnents of Fort Loudon, T-ellicO, and Watts Bar Dams upstream..A. seond candidate is tho 7,980 square mile sto ca...n -tered at Bul1l Gap, Tennessee, 50 miles nrFtheast of Knoxville, shownn in Figure 2.4.3 2. The flood from this storm would overtop and breach the west saddle dike at Watts Bar Dam. The flood from the 7,980 square m~ile- storm.is the loss critical storm and would produce a prbable maximum discharge less th-ian fro-mn the total basin storm.The previous PMF. evaluations considered

' ,nvolving upstream tributary dams Douglas and Watauga. These two situations were shown at that time to be non governing.

Dam safet mod-ifica3-tions have Since eliminated the potential failures-of those d-am~s.. Therefore, those two candidate situa-tions have been eliminated.

Reservoir routings started at median observedevi frthe large area P"MP sterms.Median levels were revauaeduinG Ongeratina exacrience for: '1 Tha tnt.,I nrir aid ,"-J--- I 2.4-15 SQN--2. The five year period, 1972 1976, for those projects whose operating guiddes were changed i Becau.se of the wet years of 1972 1975 and the operating guide changeG, Median eleVat,,ns were higher fo'r A- of the 13 tributa' y reseroir.

w.her rohutiRg iS involved.Normal reeviroeating prOcedurcs were used In the antecedent storm. These used turbine and sluice discharge in the tributary re.e..oirs.

TurbiRe discharges ae. ued in.the main riv.r reservois after large, ficoPd- flAýAows develop becausc-;P Ihead- diffre.F-nfials A-rc too t- Small. lNormal operating proceduFre were used in the stefrm, exept that tufrbine disharge was not Used in either the tributa,"/or main river dams.Concrete Section Analysis For concrete dam sections, factors of safety in sliding were determined by comparison of the existinq design headwater/tailwater levels to the headwater/tailwater levels that would occur in the PMF as described in Section 2.4.3. The structures were considered safe against failure if a factor of safety greater than 1.0 for sliding was demonstrated.

The dams upstream of SQN passed this test.A IJ S All oates were aetermineG to be oeerable witnout fa~iuere duFRin the TiOOO. ~Ate m~ain river cams*j ....I I [ tl ' I it:OUic oe TuiiV raisec. tnus rccuiFRGn no accmoiinal ocerations Dv. the i-ast c-a4 9f tnc StoFrm. WnIcn 5 i I I I nerore me structures ano access roads would be !nUncatec.

Spillway Gates During peak PMF conditions, the radial spillway gates of Fort Loudoun and Watts Bar Dams are wide open with flow over the qates and under the gates. For this condition, both the static and dynamic load stresses in the main structural members of the Watts Bar Dam spillway gate are determined to be less than the yield stress and the stress in the trunnion pin is less than the allowable design stress.The Open radial soillwav oates at other dams upstream of Watts Bar Dam were determined to not fail bv comparison to the Watts Bar Dam sDillwav cate analysis.iAeoiaR Rintal reservo~ir eiovailORS were uscc at Me sia~i Of Me sIEorm sequence u68co to Gaene Me PMF to be consistent With statistical experience and to avoid unreasonable combinations ofexrm events. As a result, 53 perccnt of the total reserved system flood detention capacity was occupied at thc start of the mnain flood. This is considered to be amply conservative.

The statement made in the PSAR and subsequent versions of the FSAR that 67 percent of the reseVed system detention capacit wacuid at the start of the main storm was in erro. The correct percentage as; 233 The remaining reserved system detention c~apacity was 67 percent. This e~r-roeous statement was first made in the PSAR and was copied in subsequent statements where the routings were the same.In the revised analysis submitted in Amendment 51, all resroirS are higher or about the same leainat the beginning of the maRin soto-rm as a res~ult of the revised- starting levels explained in Sertfion 24.3.4 of the FSAR. ThiS cOnservative change results in 53 perent of the reservoir system detention capacity being upied at the Fstart -f the Main filE)d rFathe thaR 33 in.eis estudies.Neither the initial reservoir levels nor the operatRingules would have signigficaRnteffecn max, imum flood discharges and elevations at the plant site because hpillway apa"ities, anRd h8nce, conRditions, wore reached earl" in the flood.The procedures used to determine if and when an overtopped earth embankment would fail and the proaedures for omputinRg the effect Of such are desc-ribed in 2 4 2 a ntd 2. 4.,.3 In testing the adequacy of the yard drainage system, to safely pass the site PMP, all underground drains were assumed Glaced anRd the Gfqae drFain"ace to be full Waterborne Obiects 2.4-16 SQN-Consideration has been given to the effect of waterborne obiects striking the spillway gates and bents supportinq the bridge across Watts Bar Dam at peak water level at the dam. The most severe potential for damage is postulated to be by a barge which has been torn loose from its moorings and floats into the dam.Should the barae aDoroach the sPillwaY Portion of the dam end on. one bridae bent could be failed bv the barae and two spillwav gates could be damaaed and possibly swept away. The loss of one bridae bent will likely not collapse the bridce because the bridge girders are continuous members and the stress in the girders is postulated to be less than the ultimate stress for this condition of one support being lost. Should two -gates be swept away, the nape of the water surface over the spillway weir would be such that the barge would likely be grounded on the tops of the concrete spillway piers and provide a partial obstruction to flow comparable to un-failed spillway gates. Hence the loss of two gates from this cause will have little effect on the peak flow and elevation.

Should the barqe approach the spillway portion broadside, two and possibly three bridge bents may fail. For this condition the bridae would likelv collaDse on the barae and the barae would be arounded on the tops of the soillwaY Diers. For this condition the barae would likelv around before strikina the spillway gqates because the gqates ale about 20 ft downstream from the leg of the upstream bridgqe bents.Lock Gates The lock gates at Fort Loudoun, Watts Bar, and Chickamauga were examined for possible failure with the conclusion that no potential for failure exists. The lock gate structural elements may experience localized vieldina and may not function normally followina the most severe headwater/tailwater conditions.

2.4.3.5 Water Level Determinations The controlling PMF elevation at the SQN was determined to be 722.0 ft, produced by the 21,400-sauare-mile storm in March and coincident with overtopping failure of the West Saddle Dike at Watts Bar Dam. The PMF elevation hydrograph of the controlling PIIF, ..esting at elevation 719.66, is shown on Figure 2.4.3-4-221.

Computation of both the probable maximuRmF discharge hydrOgraph (FiguIe 2.4.3 11) and the elevation was accoFmplished using the unsteady flow techniques Elevations were computed concurrently with discharges using the SOCH unsteady flow reservoir model described in Section 2.4.3.3.The leG critical arca producing P,- MP deptho n. the 7,980 square Mile watershed would produce crest elcation 718.9 at the plant site.Maximum water levels at buildings expected t , result frm the local plant PI.-P were determined siR-n l~mo methods: (1) when flow conditions controlled , standard step backwater ferom the con-trol ecio usn eak discharges estimated from rainfall1 intensities corresponding to the time9 Of concenA-tration of the araabove the cOnMro section or (2) when pending Or r8eseVoir type conditions controlled, storage routing the inflow hydrograph equivalent to the PMVP hydregraph with 2 min'ute time intervals-.

The separate watershed,-

subareas fTowpaths are ..shown.. o. F 2.4.3 1a Runoff fromF the 21.5 acre western plant site will flew e-ither norFthwest to a 27- foot channFel along the mna in p la nt tracA-k s anAd- thePn across-r, .the- mai acces s h ig hwa y orF to the sou th P'over the sw abe in PeFrimeter Road n ear the 16-1 kA/ &9.Aitchyard and acror-ss Pat-rol Roaid to the river. Because the 500 01 switchyard and T-EACP building areas are essentially level, peak eutffowis fro~m this. su,-ba-rea were determined using method (2). These peak euA94ows; wAere then combn-ined-w~ith discharge estimates from the rmiigresusnmehd()tostablish peak water surface profiles from both the north channel and south swae. The maximum water surace elevation is belo-w critical floor elevation 706 and occurs near the east west centerline of the T-urbine Building.The 28.9 acre eastern plant site was evaluated as two areas. Area 1 (19.7 acrFes) inRluding the diesel 2.4-17 SQN-generator, unit tWO reactor building, fiel.d soe.PxrvicosIstorage buildings and adjacen.t areasb. Runoff from area-i-will flow to the south along the perimeter road and across the pavoment With l0W point elevation 705.0 to the diScharge channol. Maximum water su~face elevationS cOMputed using m~ethod (1) werc less than elevation 7-06. Area 2 (9.2 acres) includes the of:AGce/seVice, unit one reactor building, officeipower stores buildings, intake pumnping station, and adjacent areas. Runoff frmarea 2 will flow to the north and west along the E=ROW pumping station access road to the intake channe! and riVcr.Un~derground drains were assum~ed clogged throughout the sto-rm. Fo-r fenc~e ssect-ions, the MeaRn4gs R value was doubled to account for increased resistance to flew and the potential for debris bloc~kage.

The only stream adjacent to SQN is the Tennessee River. There are no streams within the site. The 1 percent chance floodplain of the Te~nnejssee River at then site is delineated on Figure 2.4.3 14.DePt-ails nof the analyses used in the computation of the 1 percent chance flood flow and water ctcvatien are described in a study made by TVA for the Federal Insurance Administration (P!A) and publishedi Februa~' 1979[5].The only structures located in the 1 percent c-ha;nceP floodplain are transmission towers, the intake pming station skimmer wall, and the ERCWV pump station deck. The ERCW pumps are located On trUct-res are shown on Figure 2.4.3 14.The strUctures that Iocated in the ill net flond flows nr elevaltinns The 20,650 square mile drainage area is not altered and the reduction in flow area at the site i generated from it will be minimal and will present no problem to downstream facilitis 2.4.3.6 Coincident Wind-Wave Activity Some wind waves are likely when the probable maximum flood crests at SQN. The flood would be near its crest for a day beginning about 2-1/2 days after cessation of the probable maximum storm.The day of occurrence would most likely be in the month of March or possibly the first week in April.A conservatively high velocity of 45 miles per hour over water was adopted to associate with the probable maximum flood crest. A 45-mile- per-hour overwater velocity exceeds maximum March one-hour velocities observed in severe March windstorms of record in a homogeneous region as reported by the Corps of Engineers

[6].That a 45-mile-per-hour overwater wind is conservatively high, is supported also by an analysis of March day maximum winds of record collected at Knoxville and Chattanooga, Tennessee.

The records analyzed varied from 30 years at Chattanooga to 26 years at Knoxville, providing samples ranging from 930 to 806 March days. The recorded fastest mile wind on each March day was used rather than hourly data because this information is readily available in National Weather Service publications.

Relationships to convert fastest mile winds to winds of other durations were developed from Knoxville and Chattanooga wind data contained in USWB Form 1001 and the maximum storm information contained in Technical Bulletin No. 2 [6]. From the wind frequency analysis it was determined that the 45-mile-per-hour overwater wind for the critical minimum duration of 20 minutes had an 0.1 percent chance of occurrence on any given March day.The probability that this wind Might occur on the8 specific day that the probable maXimRum flood would c-rest is, extremely remeto. E.ven assuming that the floo9d w.Aas- to crest once during the 40 ye-ar plant life, the probability Of the wind occurring on that particular day is in the order of 1 X 1 0~-TA es6t*iate that the probability, of the food and wind occurring n givena O the same day to be in the ordJer of 1 x e 10~4--4~Computation of wind waves was made using the procedures of the Corps of Engineers

[7]. The critical directions were from the north-northwest and northeast with effective fetches of 1.7 and 1.5 miles, 2.4-18 SQN-respectively.

For the 45-mile-per-hour wind, 99.6 percent of the waves approaching the plant would be less than 4.2- and 4.0-foot-high crest to trough for the 1.7- and 1.5-mile fetches as shown on Figures 2.4.3-4-524 and 2.4.3-4625.

Maximum water su, a...es. in the .eser..i.

approaching the plant would be 2.8 and 2.7 feet abo'e the maximum computed leVel' or ele-Vations 722.4 and 722,3, F8GPeGtively.OnlV the most critical fetch length of 1.7 miles is used to determine the design basis flood elevations The maximum water level attained due to the PMF plus wind-wave activity is elevation 72-,.726.2 ft at the ERCW pump station and the nuclear island structures (shield, auxiliary, and centrol buildingqShield Auxiliary, and Control Buildings).

The wind waves approaching the Diesel Generator Building and cooling towers break before reaching the structures due to the shallow depth of water. The topography surrounding these structures is such that the wind waves will break on a steeper slope (4H:1V) than the slope immediately adjacent to the structures.

This is shown by Figure 2.4.3-4-726.

The runup estimates are on the basis that th incminnd waves break hbfnre reaching the StrutuFe aRnd then reform for a shallwer wateFr deptfh. This reformed wave then approaches the Structure.

The runups are lower than the maximum reseo-oir level due to the small wave height for the refo.r.m.e.,d wa.ve, the shallow wat.., andh, very shallow slope b'for, reaching the structure..

Wind-wave runup coincident with the maximum flood level for the diesel generator buildirgDiesel Generator Building and cooling towers (Figure 2.4.3-26) is elevation 72-1.723.2.

The level inside structures that are allowed to flood is elevation 720.1. The flood elevations used as design bases are given in Section 2.4A.4 2.4.14.1.1.

Dynamic Effect of Waves 1. Nonbreaking Waves The dynamic effect of nonbreaking waves on the walls of safety- related structures was investigated using the Rainflow Method [8]. As a result of this investigation, concrete and reinforcing stresses were found to be within allowables.

2. Breaking Waves The dynamic effect of breaking waves on the walls of safety-related structures was investigated using a method developed by D. D. Gaillard and D. A. Molitar. The concrete and reinforcing stresses were found to be less than the allowable stresses using this method.3. Broken Waves The dynamic effect of broken waves on the walls of safety-related structures was investigated using a method proposed by the U.S. Army Coastal Engineering Research Center [7]. This method of design yielded concrete and reinforcing stresses within allowable limits.All safety-related structures are designed to withstand the static and dynamic effects of the water and waves as stated in Section 2.4.2.2.2.4.4 Potential Dam Failures, (Seismically anAd Otherise-Induced)

The procedures described in Appendix A of Regulatory Guide 1.59 were followed when evaluating potential flood levels from seismically induced dam failures.The plant site and upstream reservoirs are located in the Southern Appalachian Tectonic Province and, therefore, subject to moderate earthquake forces with possible attendant failure. Upstream dams whose failure has the potential to cause flood problems at the plant were investigated to determine if failure from seismic events would endanger plant safety.2.4-19 SQN-It should be clearly understood that these studies have been made solely to ensure the safety of SQN against failure by floods caused by the assumed failure of dams due to seismic forces. To assure that safe shutdown of the SQN is not impaired by flood waters, TVA has in these studies added conservative assumptions to be able to show that the plant can be safely controlled even in the event that all these unlikely events occur in must the proper sequence.By furnishing this information TVA does not infer or concede that its dams are inadequate to withstand earthquakes that may be reasonably expected to occur in the TVA region under consideration.

TVA believes that multiple dam failures are an extremely unlikely event. The TVA Dam Safety Program (DSP), which is consistent with the Federal Guidelines for Dam Safety [251, conducts technical studies and engineering analyses to assess the hydrologic and seismic integrity of agency dams and verifies that they can be operated in accordance with Federal Emergency Management Agency (FEMA)guidelines.

These guidelines were developed to enhance national dam safety such that the potential for loss of life and property damage is minimized.

As part of the TVA DSP, inspection and maintenance activities are carried out on a regular schedule to confirm the dams are maintained in a safe condition.

Instrumentation to monitor the dams' behavior was installed in many of the dams during original construction and other instrumentation has been added since. Based on the implementation of the DSP, TVA has confidence that its dams are safe against catastrophic destruction by any natural forces that could be expected to occur.2.4.4.1 Dam Failure Permutations There are 20 major dams above SQN. Dam locations with respect to the SQN site are shown in Figure 2.4.1-3. These are Watts Bar and Fort Loudoun Dams on the Tennessee River: Watauga, South Holston, Boone, Fort Patrick Henry, Cherokee, and Douglas Dams above Fort Loudoun; Norris, Melton Hill, Fontana, and Tellico Dams between Fort Loudoun and Watts Bar: and Chatuqe, Nottely, Hiwassee, Apalachia, Blue Ridge, Ocoee No. 1, Ocoee No. 2, and Ocoee No. 3 emptying into Chickamauqa Reservoir.

These were examined individually, and in getufpscombinations, to determine if failure might result from a seismic event and. if suhso would failure or failures occurring Gcnurent4Iyconcurrent with storm runoff weu4-create nritalmaximum flood levels at the plant.Two sotuation wer Reaid: (1) a one half Safe Shutdown Earthquake (SSE=) as definod in Subsection 2.5.2, imposed concurrently With one half th p ,robable m :axim.um.

flood and (2) a Safe Shutdwn Earfthquake (SSE) as defined ir Subsectien 2.5.2, imposed contUrhrentl with a 25 year flood. Ncither of these conditionRS would croate levcls greater than the h;Ydrologic probable maximAum flood at SQN, described previously in 2.4.3. Details of the da~m f-ailur-e analysis aro discussed in n oOGBOR lt.4.4.4, Dam FamuFe r-eFFRUt=tIC'R II he~ proiUdueUes r~ei eU LO in r-\eru~dLUIX V~~d r-\%U IP.AJ II .Appendix A, were followed for evaluating potential flood levels from seismically induced dam failures.In accordance with this guidance, seismic dam failure is examined using the two specified alternatives:

(1) the Safe Shutdown Earthquake (SSE) coincident with the peak of the 25-year flood and a two-year wind speed applied in the critical direction, (2) the Operating Basis Earthquake (OBE) coincident with the peak of the one-half PMF and a two-year wind speed applied in the critical direction.

The OBE and SSE are defined in Section 2.5.2.4 as having maximum horizontal rock acceleration levels of 0.09 q and 0.18 q respectively.

As described in Section 2.5.2.4, TVA agreed to use 0.18 q as the maximum bedrock acceleration level for the SSE.Failure Of Chickamauga Da,,, downstream, can affoct cooling wat.r supplies at the plant.Consequently for conser.atism, an arbitray fail'ue was imposed. This resulting would not be critical to plant operation, as discussed in Section 2.4.11.6.

From the seismic dam failure analyses made for TVA's operating nuclear plants, it was determined that five separate, combined events have the potential to create flood levels above plant grade at Watts Bar Nuclear Plant. These events are as follows: 2.4-20 SQN-(1) The simultaneous failure of Fontana and Tellico Dams in the OBE coincident with one-half PMF.(2) The simultaneous failure of Fontana, Tellico, Hiwassee, Apalachia, and Blue Ridge Dams in the OBE coincident with one-half PMF.(3) The simultaneous failure of Norris and Tellico Dams in the OBE coincident with one-half PMF.(4) The simultaneous failure of Cherokee, Douglas, and Tellico Dams in the OBE coincident with one-half PMF.(5) The simultaneous failure of Norris, Cherokee, Douglas, and Tellico Dams in the SSE coincident with a 25-year flood.Tellico has been added to all five combinations which was not included in the original analyses for TVA's operating nuclear plants. It was included because the seismic stability analysis of Tellico is not conclusive.

Therefore, Tellico was postulated to fail.2.4.4.1 Reservoir

_Description Charactoristics of dams that influence ivoer cenditions at SQN are contained in Tables 2.4.1 1 and 2.44 2. location with respect to the plant as shewn on Figure 2.1.1 1. safety criteFia.:ere not inore~porated in the design of dams6 upstream froM SQN, eXcopt TellicO and_ Norris. Thosehaving a potential to influehnce plant flodinRg levels examined, as de,. ribcd in Scctio-in Elevation storage relationship, and seasenally varying steoage allocatiens in the major projects are shown On the 14 she-ets of Figure 2.4.1 3.2.4.4.2 Dam F~ailure Permutations The plant site and upstreamA reservoirs arc located in the Southern AppalachiaR T-ectonic ProVince and, therefore, su1bject to moderate ea~thquake for~es with possible attendant failure. All upstrea dams, whose fa.iure has the potential to cause flood problems at the plant, were indestigated to determine if failure fromn seimSFic Or hydrologic eventS would endanger plant safety. Potential failure frem both seiscn and hydraonlgic eeAnts and the esuflinsg conseque nce are discused in this section.it should be cleary understood that these studies have been made solely to ensure the safety of SQN against failWe by floods. caused frFo exessive rainfall Or by the assumphed failure Of dams due to seismFic forces. To assure that safe shutdown of SQN is, not impaired by flood wators, TVA has in these stuidios; a-d-deAd censervative assumptio-ns to coenservative assumptions to be able to show that the plant can be safely controlled even in the event that all these unlikely events occur in just the pro)per sequence.

TVA is of the stronRg opinionR that the chances~_

of the assumed events occurring approach zero9 probability.

B3y furnishing this- info~rm~ation, TVA does RAM inferF Or concede that its dams6 are inadequate to w~ithstand great floods, and~er ea~thquakes that may be reasonably expected to occur in the TV reio unjder consideration.

TVA has a programn Of inspection and main~tenance carried out on a regular schedlule to keep its damsG safe. Isrmnainof the dams6 to help keep check en their behavioar Wars, installedin many of the dams6 duFrig original construction.

O-the-r instruSmenaBItionA has been added since and isstill being added a6 the need may appear or as new techniques becom~e avafilab~e In sheot, TVA has confidence that its dams, are safe against catastrophic destruction by any natural forces6 th-at coeuld- beP oXoocte~d to_ occur.2 4 4 2 1 RAORmir_ F=iiiir An;;IvrRi 2.4-21 SQN-Seismic failure aRna!yss consisted of the following:

r i i i

• il l I I 4 I IP mP r iI 1. _eterminaven of tne water level at te plant during one half the M~t With f u, reser o.rs i its, creSts were aumimented by noee waves tro4e tnh Peostuilateet tailure Oft u,6toamn edars drinG a enie halfSSE.J 2. 'Ueterm:nAtien of the w.atper level -at the olant drinp Ea 2!iea;r flood full rpeervirs if iU restsf.... r .. ....-- ý --- j were augmented by flood waves ftrom the postulated faiUreto upstreamn damns during a Saf ShutdoWn Earthquake (SSE+.The one half SSE Identified in conditionR 1 is defincd in FSAR Section 2.5.2.4 as having a peak horizontal acceleration value of 0.09 g at the rock foundation.

The discussion in Stin2-5,2.4 v shows the eXtFreRe conservatism contained in the analy6is.In the 1998 reanalysis all potentially critical seismic events involvYing dam failure upstream of the plant site were reevaluated.

The six events included the postulated one half SSE, failure of (1) Norris, (2)Fontana, (3) Cherokee Douglas, and (1) Fontana Hiwassee Apalachia Blue Ridge during one half the Loudoun Tellico during a 25 year flood.Seismic- failure of upstream dams during nonflood periods pose no threat to the plant-.,summtreof the result of the ......an.lysis isgiven ,in T-able2.41.

1. SU0Nandupstream dams ae ,located asshownR, .. Figure 2..1 1. The highest flood level.at SON from different seismic dam failure and flood combinations; would be elevationF 70)7.9 from simultaneous failure Of Fon~tana Darn On the Little Tennessee River and Hiwassee, Blue Ridge, and Apalachia Dams OR the Hiwassee River during a ene half 6afe shutdown earthquake coincGident with one half the PMF=. This includes improv~emets resulting fromn modifications pe~fermed-fo-r the Dam Safety Programn.

Wind waves could raie te eevaionto7096 i th reervir Runup could reach elevation 710.4 On a 4:4 slope to elevation 712.8 On a VeFtical wall in shallow (4.9 feet) water, and to elevation 710.4 en Oa vFtiGal wall in deep wateF.Only stn ather seismic dam failure combination With conident foimmdbs could cause elevatiors above pat g~ade.Plant safety would be assured by shutdOgn prier to these floods crossing plart grade, elevation 705, using the warning system described in Appendix 2.1A.T-he effect of postulated seismic bridge failure and resulting failure Of spillway gate anchors at Watts Bar and Fort Loui-doaun Dams would not cr-eate a safety hazard at SQN.Prer-e dwes Concrete Structures The standard method of computing stability is used. The maximum base compressive stress, average base shear stress, the factor of safety against overturning, and the shear strength required for a shear-friction factor of safety of 1 are determined.

To find the shear strength required to provide a safety factor of 1, a coefficient of friction of 0.65 is assigned at the elevation of the base under consideration.

AG stated in Sec~tion 2.4.1.2, all of the original stability analyses and postulated dam failue assum...ptions in the 1998 FeaRalyses were coRseprvtively assumed to occu in the same manner and in combination with the same postulated rainfall events.The analyses for earthquake are based on the pseudo-static analysis method as given by Hinds [10]2.4-22 SQN-with increased hydrodynamic pressures determined by the method developed by Bustamante and Flores [11]. These analyses include applying masonry inertia forces and increased water pressure to the structure resulting from the acceleration of the structure horizontally in the upstream direction and simultaneously in a downward direction.

The masonry inertia forces are determined by a dynamic analysis of the structure which takes into account amplification of the accelerations above the foundation rock.No reduction of hydrostatic or hydrodynamic forces due to the decrease of the unit weight of water from the downward acceleration of the reservoir bottom is included in this analysis.Waves created at the free surface of the reservoir by an earthquake are considered of no importance.

Based upon studies by Chopra [12] and Zienkiewicz

[13], it is eu*TVA's judgment that before waves of any significant height have time to develop, the earthquake will be over. The duration of earthquake used in this analysis is in the range of 20 to 30 seconds.Although accumulated silt on the reservoir bottom would dampen vertically traveling waves, the effect of silt on structures is not considered.

There is only a small aFmo-nt of silt noW present, and th The accumulation rate is slow, as measured by TVA for many years [14].Embankment Embankment analysis was made using the standard slip circle method, except for Chatuge and Nottely Dams where the Nemark method for the dynamic analysis of embankment slopes was used.The effect of the earthquake is taken into account by applying the appropriate static inertia force to the dam mass within the assumed slip circle (pseudo-static method).In the analysis, the embankment design constants used, including the sheer strength of the materials in the dam and the foundation, are the same as those used in the original stability analysis.Although detailed dynamic soil properties are not available, a value for seismic amplification through the soil has been assumed based on previous studies pertaining to TVA nuclear plants. These studies have indicated maximum amplification values slightly in excess of two for a rather wide range of shear wave velocity to soil height ratios. For these analyses, a straight-line variation is used with an acceleration at the top of the embankment being two times the top of rock acceleration.

As discussed in Section 2.4.3, temporary flood barriers are installed on embankments at Cherokee, Watts Bar, Fort Loudoun and Tellico Reservoirs.

However, the temporary flood barriers are not required to be stable following an OBE or SSE and are not assumed to increase the height of the embankments for these loading conditions.

Flood Routing The runoff model described in Section 2.4.3.3, which includes unsteady flew medels for critical rospoer/irs and riVer roaches6, Was used to reevaluate plant Site flood- leVels re-sulting from the pestulated SSE and one;, half SRI= dam failu- r cob The remaining eventS prodIuced plant 6site flooed levels sufficiently lower than the controlling9 events and wore not evaluated, was used to reevaluate five potentially critical seismic events involving dam failures above the plant. Other events addressed in five earlier studies (the postulated OBE single failures of Watts Bar and Fort Loudoun;the postulated SSE combination failure of Fontana and Douglas, the postulated SSE combination failure of Fontana, Fort Loudoun, and Tellico; the SSE combination failure of Norris, Douglas, Fort Loudoun and Tellico, and the single SSE failure of Norris) produced plant site flood levels sufficiently lower than the controlling events and therefore were not re-evaluated.

The procedures prescribed by Regulatory Guide 1.59 require seismic dam failure to be examined using the SSE coincident with the peak of the 25-year flood, and the OBE coincident with the peak of one-half the PMF.Reservoir operating procedures used were those applicable to the season and flood inflows.2.4-23 SQN-ThiG s-ction was rev..,, with a major rearrangoment to l'cato the ..nrol +.n, e...t.. .Valu.,atod in the v vI

• J Ill 1 uuQ anayiS.. .irst AnA te non. contro,- in. eVentS. Wnin wo not re cacuate later. I Te no no conroiiing events aro ieR in tne zi-'oN- Tar nlistry.One-half SSE Concurrent With One-Half the Probable Maximum Flood Previous evaluations have been made which determined flood levels at SQN for potentially critical events. Re-evaluations made later using the updated runoff model described in Section 2.4.3.3 and including the Dam Safety Program modifications did not determine flood levels for those events which were previously shown to clearly not be controlling.

The 1998 analysis for determining the effects of the Dam Safety Program modifications determined that non-flood related seismic dam failure events clearly pose no threat to the plant. Flood levels were determined for six combined seismic/flood events. Only two of these controlling seismic/flood events would exceed plant grade. These two events consist of multiple dam failures on (1) Little Tennessee/Hiwassee, and (2) Clinch/Upper Tennessee rivers with flood levels at SQN of El. 707.9 and 706, respectively.

The following is detailed descriptions of the potentially critical controlling events including reevaluated flood levels, followed by brief descriptions of the non-controlling failure events previously evaluated.

• I I D" I lJ f li *ml ......Altflugfl conirde Fe, as aiscussed iR the TolloWing paragrapfls, T VA iDelieVos thlat mnultiple damn failuros aro an extremely unlikely event. TVA's search of the literature reeal no recOrd- of failure-o concrete dams erm earthquake.

The postulation of an SSE Of 0.18 g acceleration is a very oRnserFative upperF im it in itself (as stated in Scrtion., 2.5.2). In addition, the SSE must beo ol atod in a.Fcn' precise seg!pa to have the potential for multiple dam failures.SSE= In onrder to fail three dams Nforri, heenkte, and Douglas the epiceste r of a SSE m'ust bo confined to a relatively small area, the shape of a football, about 10 miles wide and 20 miles long. In order to fail four dams Norris, Douglas, Fort Loudoun, and Teolco the epicenter of an SSE= m:ust b cnfinedJ to a tiangular arean With sides -f approximately 1 mile in length. However, as ar extreme upper limit the above tWo com~binatfionRs of dams are postulated to fail as well as the Gembinatiens of (1) Fort Loudoun, Tollico, and Fontana; (2) Fontana and Douglas; and (3) Fontana ;And the six Hiwa;rsse River dams& The 1098 Fe analysis, determined that only the fir-st ton com.rbinations are controlling and need to be consi;dereFPd_.

Only the Norris Cherokee Douglas event would oxceoed plan g -deelevatin One half SSE= Attenu atioin studies of the one half SSE= show that there arc three combinatiOns at somultaneous failures Of morGe than one dam which need to be conRsidered With respect to SQN safety.:hich are discussed below. These are (1) Cherokee Douglas, (2) Fontana Hi4wassec Apalachia Bu Ridge, and (3) Hiwassee Apalachia Bllue Ridge Gsoee No 1. Nottely. The 19998 re analysis determined that only the first two comFbinationRs are controlling and need to be considered.

Only the Foentana Hiwasse Apalachia Blue Ridge evenAt woul)exeed plant grae The following descriptionRs arc first forthe contro9lling events for which flooAd- 'levels o~realuaefo the_ 1998 rFeanalysis, followed by the non contro9lling evenIts ýw~hic~h wer-e not Fe analyzed in 1998.One half S-SEFOBE Concurrent With One-Half the Probable Maximum Flood (Controlling Events)Watts Bar Dam Stability analyses of Watts Bar Dam powerhouse and soillwav sections result in the iudament that these structures will not fail. The analyses show low stresses in the spillway base, and the powerhouse base. Original results are given in Figure 2.4.4-1 and were not updated in the current analysis.

Dvnamic analysis of the concrete structures resulted in the determination that the base acceleration is amplified at levels above the base. The original slip circle analysis of the earth embankment section results in a factor of safety greater than 1, and the embankment is iudged not to fail as shown in Figure 2.4.4-22.2.4-24 SQN-For the condition of peak discharge at the dam for one-half the probable maximum flood the spillway gates are in the wide-open position with the bottom of the gates above the water. This condition was not analyzed because the condition with bridge failure described in the following paragraphs produces the controlling condition.

Analysis of the bridge structure for forces resulting from the OBE, including amplification of acceleration results in the determination that the bridge could fail as a result of shearing the anchor bolts. The downstream bridge girders are assumed to strike the spillway gates. The impact of the girders striking the gates is assumed to fail the bolts that anchor the gate trunnions to the pier anchorages allowing the gates to fall on the spillway crest and be washed into the channel below the dam. The flow over the spillway crest would be the same as that prior to bridge and gate failure, i.e., peak discharge for one-half the probable maximum flood with gates in the wide-open position.

Hence, bridge failure will cause no adverse effect on the flood.Previous evaluations determined that if the dam was postulated to fail from embankment overtopping in the most severe case (gate opening prevented by bridge failure) that the resulting elevations at SQN would be several ft below plant grade elevation 705.0 ft. Therefore, this event was not reevaluated.

Fort Loudoun Dam Stability analyses of Fort Loudoun Dam powerhouse and spillway sections result in the iudgment that these structures will not fail. The analyses show low base stresses, with near two-thirds of the base in compression.

The original results given in Figure 2.4.4-2 were not updated for the current analysis.Slip circle analysis of the earth embankment results in a factor of safety of 1.26, and the embankment is iudged not to fail. The original results given in Figure 2.4.4-3 were not updated in the current analysis.The spillway gates and bridge are of the same design as those at Watts Bar Dam. Conditions of failure during the OBE are the same, and no problems are likely. Coincident failure at Fort Loudoun and Watts Bar does not occur.For the potentially critical case of Fort Loudoun bridge failure at the onset of the main portion of one-half the probable maximum flood flow into Fort Loudoun Reservoir, in an earlier analysis it was found that the Watts Bar inflows are much less than the condition resulting from simultaneous failure of Cherokee, Douglas, and Tellico Dams as described later.Tellico Dam Although, not included in the original analyses for TVA's operating nuclear plants, The concrete portion of Tellico is iudged to fail completely because the seismic stability analysis of Tellico is not conclusive.

No hydrologic results are given for the single failure of Tellico because the simultaneous failure of Tellico Dam with other dams discussed under multiple failures is more critical.

The embankments at Tellico are stable (Figure 2.4.4-23).

.4-Norris Dam Reuls f the Norris Dam stability analyses for a typical 6pillWay block and a typical non oeverfow section of height ;;r r.hwn n Figure 2.4.4 8. Because only a small percentage Of the spillway bare is in compre..ion, this structure is judged to fail. The high non ov..ow section .ith -small percentage o the base in compression ande with high comprE) si.. and shearing stre.ses is also judged to ail Although an evaluation made in 1975 by Agbabian Associates concluded that Norris Dam would not fail in an OBE (with one-half PMF) or SSE (with 25-year flood), the original study postulated failure in both seismic events. To be consistent with prior studies, Norris was conservatively postulated to fail.2.4-25 SQN-Figure 2.4.4-94 shows the likelypostuIated condition of the dam after OBE failure. Based on stability analys 6, the non oVeoflOW blocks rem~aining in placo aro judged to withstand thc- o-no half SSE.v i,^ t^:l k ...... ,,4. .... * .v PAtf.,"L -A,2 ";4 t f.i ; -rV rlI,.The location of the debris is not based on any calculated procedure of failure because it is believed that this is not possible.

It is TVA's judgment, however, that the failure mode shown is one logical assumption; and, although there may be many other logical assumptions, the amount of channel obstruction would probably be about the same.The discharge rating for this controlling, debris section was developed from a 1:150 scale hydraulic model at the TVA Engineering Laboratory and was verified closely by mathematical analysis.in the hydrologic routing for this failure, MeltOn Hill DamR was postulated to fail when the flood wave reached headwater elevation 804, based On Structural analysis.

The headwator at Watts Bar Damn w:ould reach elevation 758.1, 8.9 fcct below top of dam. The west saddle dike at Watts Bar Dam w:ould be overtopped and brcachcd.

A cOmnplete washout of the dike was assumed. The resultinlg w~ater level at the nuclear plant site is 698.1. 6.9 fect bclow plant grade 705.No hydrologic results are given for the single failure of Norris Dam because the simultaneous failure of Norris and Tellico Dams, discussed under multiple failures, is more critical.2. Fontana Dam Fon-tana; Da~m w~as a4s~su-med-to fail; in the one half SSE, although no stability analysis was made.is a high darn constructed with three longitudinal contraction joints in the highcr blGorc.A structJural defect in Fontana Dam ws;as found in October of 972 and conssist of a crngitudinal crack~~~~~~~~~~~f pn Af t .-. , 1,- ,+ ~ ..h,- r.., 1.,-. r..,.-71 A A 19\ C4r~ fI,-,.,i.-.....

.,4S'£rr J.*l nlIX 1 fl Figure 2.4.4 17). Only these three blocks ahe cracked, and there is no evidence that any other po-tion of the dam is weakened.Studies and tests, undertaken with the concurrence of a bo~ard Of pF!Yate consulting engioeerS, that this cracking was caused by a longitudinal thru st crpeatd by a cOmbinati*n of long concrete growth and expansion due to temperature rise in the summer months. This thrust tends to push the cuR'ed blocks upsteam. The studies and tests Will continue unti there is established a basis for design of permnanent m:easures to conAtrol:)

the fu-ture behayior of the damn.The strengthening Work has reestablished the structural integrity of the cracke-d blc l.~though the joints are keyed and grouted, it is possible that the grOUting was not fully effective.

Consequently, there Is some question as to how this structure Will respond to the motion of a seVere earthquake.

To be ..nse..atiy., therefore, i is.assumed that Fonta-a Dam Will nOt the one half SSE without Figure 2.1 ,t16 s;how the part of Fontana Dam judged to remain oiginal position after failure and the assumed location of the debris of the failed portion. The location of the after falure is one a6sumption based on a failure Of the dam at the InOgitudinal joints. There may be other assumptions, but the amonl int of channel obstruction would probably be aboh ut the sam~e.The, higher blocks 9 27, containing either two Aor three longitudinal joints, are assu..med to fail. Right abutment blocks 1 8 and left abutment blocks 28 and beyond were judged to be stable for the 1. Their heights are less than one half the maximum height of the dam.22. Nne of theA-se b.ks haven more than- one lAgitudinAal cRtFacGtio joit, anAd ISo have no longitudinal joints.2.4-26 SQN-3. The back slope o~f Fontana Dam is one On 0.76, Which the original stability analysis shows is flatter than that required for stability for the normal static loadings.Although not investigated, it was assumed that Nantahala Dam, from Fontana ,nd Santeotlah On a downstream tributary, and the three ALCOQA dams, downstream On the Little Tennessee River, Cheoah, Ca~derweod, an~d Chilhowoe, would fail along with Fontana in the one half SSE. Instant Vanishment was assumed. Tel"icI ad Watts Bar Dam spillway gates would be operable during and afte~r the onRe half RSS1E. Failure of the bridge at Fort Loudoun Dam would rendcr the spillway gates inopcrablo in the wide open position.The flood wave would overtop Tellico Dam and its saddle dikes. Tran.fe of water into Fo- t Leudoun w~ould occur but would not be s~uffic~ient to ovYertop the damn or to prevent failure of Tellico Dam. Tellico was postulated to completely fail. Watts Bar headwater would reach elevation 761 .3, 5.7 feet belo top of d m. The Watts Bar west saddle dike would be overtopped and breached.

A complet w~ashout of the dike was assumed. The elevation at the plant site would be 702.8, 2.2 feet below plant g~ade.3.-Cherokee-ueul.as Dam The simultaneous failurc of Cherokec and Douglas Dams could occur WhcR the half SSE is located midway be,-een the dams which are just 15 miles apart.Results of the original Cherokee Dam stability analysis for a typical spillway block are shown in Figure 2.4.4-405.

B edo this analysis, theThe spillway is judged stable at the foundation base elevation 900.0 ft. Analyses made for other elevations above elevation 900.0 ft, but not shown in Figure 2.4.4-4-05, indicate the resultant of forces falls outside the base at elevation 1010.0 ft. The spillway is assumed to fail at that elevation.

The non-overflow dam is embedded in fill to elevation 981.5 ft and is considered stable below that elevation.

However, original stability analysis indicates failure will occur above the fill line.The powerhouse intake is massive and backed up by the powerhouse.

Therefore, it is judged able to withstand the GRe4half-SSEOBE without failure.Results of the originlanalysis for the highest portion of the south embankment are shown on Figure 2.4.4-146.

The analysis was made using the same shear strengths of material as were used in the original analysis and shows a factor of safety of 0.85. Therefore, the south embankment is assumed to fail during the one-half SSE. Because the north embankment and saddle damsSaddle Dams 1, 2, and 3 are generally about one-half or less as high as the south embankment, they are judged to be stable for the e ehalf- SSOBE.Figure 2.4.4-1-27 shows the assumed condition of the dam after failure. All debris from the failure of the concrete portion is assumed to be located downstream in the channel at elevations lower than the remaining portions of the dam, and therefore, will not obstruct flow.No hydrologic results are given for the single failure of Cherokee Dam because the simultaneous failure of Cherokee, Douglas, and Tellico Dams discussed under multiple failures, is more critical.Douglas Dam Results of the original Douglas Dam original-stability analysis for a typical spillway block are shown in Figure 2.4.4-4138.

The upper part of the Douglas spillway is approximately 12 feet higher than Cherokee, but the amplification of the rock surface acceleration is the same. Therefore, based on the Cherokee analysis, it is judged that the Douglas spillway will fail at elevation 937.0 ft, which corresponds to the assumed failure elevation of the Cherokee spillway.The Douglas non-overflow dam is similar to that at Cherokee and is embedded in fill to elevation 927.5 ft. It is considered stable below that elevation.

However, based on the Cherokee analysis, it is 2.4-27 SQN-assumed to fail above the fill line. The abutment non-overflow blocks 1-5 and 29-35, being short blocks, are considered able to resist the e half-SSE-OBE without failure.The powerhouse intake is massive and backed up downstream by the powerhouse.

Therefore, it is considered able to withstand the ene-half-SSEOBE without failure.Results of the original analysis of the saddle damSaddle Dam shown on Figure 2.4.4-4-49 indicate a factor of safety of one. Therefore, the saddle damSaddle Dam is considered to be stable for the ene-half SSEOBE.Figure 2.4.4-4-510 shows the portions of the dam judged to fail and the portions judged to remain. All debris from the failed portions is assumed to be located downstream in the channel at elevations lower than the remaining portions of the dam and, therefore, will not obstruct flow.These-failures, in with one half the probable maximum. flood, would oyertop Fort Loudon for only 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, but would not fail the dam. At Watts Bar the werst saddie dike would be ovcrtoppe and broached.

A complete washout of thedie as assurncd.

Crest level at SON would be elovation 701.1, 3.9 feet below plant. No hydrologic results are given for the single failure of Douglas Dam because the simultaneous failure of Cherokee, Douglas, and Tellico Dams as discussed later under multiple failures.

is more critical.Fontana Dam The original hydrological analysis used a conservative seismic failure condition for Fontana Dam. A subsequent review which takes advantage of later earthquake stability analysis and dam safety modifications performed for the TVA DSP has defined a conservative but less restrictive seismic failure condition at Fontana. This subsequent review used a finite element model for the analysis and considered the maximum credible earthquake expected at the Fontana Dam site. Figure 2.4.4-11 shows the part of Fontana Dam iudged to remain in its original position after postulated failure.No hydrologic results are given for the single failure of Fontana Dam because the simultaneous failure of Fontana and Tellico Dams, as discussed later under multiple failures, is more critical.Multiple Failures Previous attenuation studies of the OBE above Watts Bar Dam result in the iudgment that the following simultaneous failure combinations require reevaluation:

(1) The Simultaneous Failure of Fontana and Tellico Dams in the OBE Coincident with One-Half PMF Figure 2.4.4-11 shows the postulated condition of Fontana for the OBE event. Tellico was conservatively oostulated to comoletelv fail.The seismic failure scenario for Fontana and Tellico include postulated simultaneous and complete failure of non-TVA dams on the Little Tennessee River (Cheoah, Calderwood, and Chilhowee) and on its tributaries (Nantahala and Santeetlah).

Failure of the bridge at Fort Loudoun Dam would render the spillway gates inoperable in the wide-open position.

Watts Bar Dam spillway gates would be operable during and after the OBE.Watts Bar Dam headwater would reach 756.13 ft, 13.87 ft below the top of the embankment.

The West Saddle Dike at Watts Bar Dam with top elevation of 757.0 ft would not be overtopped.

The peak discharge at SQN would be 775,899 cfs. The elevation at SQN would be 702.2 ft, 2.8 ft below plant grade elevation 705.0 ft.(2) The Simultaneous Failure of Fontana, Tellico, Hiwassee, Apalachia, and Blue Ridge Dams in the OBE Coincident with One-Half PMF 2.4-28 SQN-Fontana, Tellico, Hiwassee, Apalachia and Blue Ridge Dams could fail when the OBE is located within a flattened oval-shaped area located between Fontana and Hiwassee Dams (Figure 2.4.4-12).

Failure scenarios for Fontana, Tellico, Hiwassee, Apalachia, and Blue Ridge Dams include postulated simultaneous failure of non-TVA dams on the Little Tennessee River (Cheoah, Calderwood and Chilhowee) and on its tributaries (Nantahala and Santeetlah).

Based on previous attenuation studies, the OBE event produces maximum ground accelerations of 0.09 q at Fontana, 0.09 q at Hiwassee, 0.07 q at Apalachia, 0.08 q at Chatuge, 0.05 q at Nottely, 0.03 q at Ocoee No. 1, 0.04 q at Blue Ridge, 0.04 q at Fort Loudoun and Tellico, and 0.03 q at Watts Bar.Figure 2.4.4-11 shows the postulated condition of Fontana Dam after failure. Hiwassee, Apalachia, Blue Ridge, and Tellico Dams are postulated to completely fail. Chatuge Dam is iudged not to fail in this defined OBE event.Nottely Dam is a rock-fill dam with large central impervious rolled fill core. The maximum attenuated ground acceleration at Nottely in this event is only 0.05 g. A field exploration boring program and laboratory testing program of samples obtained in a field exploration was conducted.

During the field exploration program, standard penetration test blow counts were obtained on both the embankment and its foundation materials.

Both static and dynamic (cyclic) triaxial shear tests were made. The Newmark Method of Analysis utilizing the information obtained from the testing program was used to determine the structural stability of Nottely Dam. It is concluded that Nottely Dam can resist the attenuated ground acceleration of 0.054 q with no detrimental damage.Ocoee No.1 Dam is a concrete gravity structure.

The maximum attenuated ground acceleration is 0.03 g. Based on past experience of concrete dam structures under significantly higher seismic ground accelerations, the Ocoee No. 1 Dam is 'udged to remain stable following exposure to a 0.03 q base acceleration with amplification.

Ocoee No. 1 and Ocoee No. 3 Dams, downstream of Blue Ridge Dam, would be overtopped and were postulated to completely fail at their respective maximum headwater elevations.

Ocoee No. 2 Dam has no reservoir storage and was not considered.

Fort Loudoun and Watts Bar spillways would remain operable.

The Fontana failure wave would transfer water through the canal from Tellico into Fort Loudoun, but it would not be sufficient to overtop Fort Loudoun Dam. The maximum headwater at Fort Loudoun would reach elevation 817.13 ft, 19.87 ft below the top of the dam. Watts Bar headwater would reach elevation 756.13 ft, 13.87 ft below the top of dam. The West Saddle Dike at Watts Bar with a top elevation of 757.00 ft would not be overtopped.

The peak discharae at the SQN site oroduced bv the OBE failure of Fontana. Tellico. Hiwassee.Apalachia, and Blue Ridge coincident with the one-half PMF is 918,880 cfs. The peak elevation is 706.3 ft, 1.3 ft above 705.0 ft plant grade.(3) The Simultaneous Failure of Norris and Tellico Dams in the OBE Coincident with One-Half PMF Figure 2.4.4-4 shows the postulated condition of Norris Dam for the OBE event. Tellico was conservatively postulated to completely fail in this event.In the hydrologic routing for this failure, Melton Hill Dam would be overtopped and was postulated to fail when the flood wave reached headwater elevation 817.0 ft, based on the structural analysis and subsequent structural modifications performed at the dam as a result of the Dam Safety Program.The headwater at Watts Bar Dam would reach elevation 762.96 ft, 7.04 ft below top of dam. The West Saddle Dike at Watts Bar with top at elevation 757.0 ft would be overtopped and breached.

A complete washout of the dike was assumed. Chickamauga headwater would reach 701.05 ft, 4.95 ft below top of dam. The embankments at Nickalack Dam would be overtopped but was postulated not to breach which is conservative.

2.4-29 SQN-The peak dischar-ge at the SQN site produced by the OBE failure of Norris and Tellico Dams coincident with the one-half PMF is 912.939 cfs. The oeak elevation is 706.3 ft. 1.3 ft above 705.0 ft plant grade.(4) The Simultaneous Failure of Cherokee, Douclas, and Tellico Dams in the OBE Coincident with One-Half PMF Fioures 2.4.4-7 and 2.4.4-10 show the postulated condition after failure of Cherokee and Doualas Dams, respectively.

Tellico was conservatively postulated to completely fail.In the hydrologic routing for these postulated failures, the headwater at Watts Bar Dam would reach elevation 763.1 ft. 6.9 ft below the too of the dam. The West Saddle Dike at Watts Bar with a too elevation of 757.0 ft would be overtopped and breached.

A complete washout of the dike is assumed.Chickamauga Dam headwater would reach 702.95 ft, 3.05 ft below the top of the dam. The embankments at Nickaiack Dam would be overtopped but were conservatively postulated not to breach.The peak discharge at the SQN site produced by the OBE failure of Cherokee, Douglas, and Tellico with the one-half PMF is 930.585 cfs. The oeak elevation is 708.6 ft. 3.6 ft above 705.0 ft olant arade.This is the highest flood elevation resulting from any combination of seismic events. The flood elevation hydrograph at the plant site is shown on Figure 2.4.4-18.-. Fontana, Hiwa.see, Apalachia, and Blue Ridge Damsi, Fontana, Hiwassee, Apalachia, and Blue Ridge Dams coul-d- fail wA.hen the ono half SSE is located ,,thi. the fotball shaped area shown .i Figure 2.4 1.4 T-his cvcnt prodUccs maximum ground accelerations of 0.09 g at Fontana, 0.09 g at Hiwassce, 0.07 g at Apalachia, 0.08 g at Chatuge, 0.05 g at Not+ely, 0.03 g at Ocoee NO. 1, 0.04 g at Blue Ridge, 0.04 -g at Fort Loudoun and Tellico), and 0.03 g at Watts Bar. Failuro is postulated for Fontana and Hiwassec for an oarthguakc epicenter located anywherc within the- foogtball 1shaped arca shown on F~igure -2.4.4 18. Ground accelerations shown for the various dams6 arc maximum that Gould occur for cpiccnter locGated at Yarious pointS in the deScribed area and would not occur simultaneouslY.

Fort LoudoUn, Tellico, and Watts Bar Dams and spillway gates would rem~ain intact. The degrec of Fontana failure and likely position of are judged to be comparable to that shown for single failure in Figure 2.4.4 16. Hiwassee, Apalachia, and Blue Ridge Dams were assumed to com~pletely disappear.

Chatuge was judged not to fail as the acceleration is less than for the one half SSE=: centered at the dam., Nottely Dam is a rockfill dam with large ceRntal iFmpeViou shOlled fill Gore. The FmaXimum attenuate ground aceleatioen at Nottely is only 0.054 g. A field exploration bering program and laborato5g' testing programn of samples obtaine-d in -A field exploration was conducted#_ý_.

Dur4_ing the field exploration program, standard penetration tests blow counts were o-btained-oM both the_ embankment andit foudatonmaterials.

Both static and dynamic (cyclic) triaxgial s-hear tests were made. This informnationR was u se'd ithe Newmark Me..thod of Analysis.

The "Newma. Method of Analysis" (Newma.., N. M.,"Effects of Earthquake On Dam.. of Embankments," Getechnique 15A140 141, 156, 1065) utilizingthe informnation obtained-from the testing program was used_ to deAter4m.ine_

th~e structuMI-ral stability of Nettely_Dam. W.~e concF~lude Nottely DamA can easily resist the attenuated ground accGeleration of 0.054 g wt no detrimental damage.Qrcooot No. 1 Dam is aconcete gravity strucwtu re. The maximuim attenuated groun~d acceleration is 0.03 g. The 0.03 g with the proper am~plification wvas u-sed- to- analyze the structural stability-Ot structures at Ocooc No. 1. The metho of ana'ysis used was the same as deScribed previously under m i v IAI 1 A L, .4 R! ;Ii g gl MICI "Ot 0 RQ el" E)Vet-tUFR RG. 8 GORG U e t e aFA W RGI a...In the original analysis, the failure wave hydregraph was approximated for the H;iw..as..ee Rivr a..t its M.outh for thet; fal ures o. ef H iwasse.e, Apalachia ad Blue Ridge DaM. In the 1.9-9 e analyis 2.4-30 SQN-unte-ady flow model in S-ctifn 2.4.3.3 developed during the da-m s"afty studi.o Was Ue as an adjunc~t to routo the Hiwassoe, Apalachia and Blue Ridge faIlures in the simultAneous-1 f-ailurc of Fontana, HiWassec, Apalachia, and Blue Ridge Dams, the Fontan-a fail.ur wavo would oe.top aRd fil the- Ttellico .mbankmonts.

TraRnfer of water i'to Fo="rt Loudoun would occur but would not b-e suiffic-ient to overtop the dam Or to preVent failurze of Tollico. Tellico was postulated to com~pletely fail. Watts Bar headwater would- re;ach elevatio~n 761.3, 5.7 foot below top-of damn. The west sa-dd-le d-iko at Watts BRar would be overtepped.

A comAplete washout of the dike down to ground elevation was assumed. This flooed woave comFbined with th-at of HiwassqCee, Blue Ridge, and ApaJa~hi Dams would pe-,rP a mnaxi~m-um flood- ee at the pl~ st of 70., .feet above plant grade. This is the highest flood resultiRg from any combination of and flood Peunts Th satae .h-hvdrnor;nh

t the leant site is sho,.w.n on Fieure 2 4 4 21 SSE Concurrent With 25-Year Flood (Controlling Events)The SSE will produce the same postulated failure of the Fort Loudoun and Watts Bar bridqes as described for the OBE described earlier. The resulting flood level at the SQN was not determined because the larger flood during the OBE makes that situation controlling.

Watts Bar Dam A reevaluation using the revised amplification factors was not made for Watts Bar Dam for SSE conditions.

However, even if the dam is arbitrarily removed instantaneously, the level at the SQN based on previous analyses would be below elevation 705.0 ft plant grade.Fort Loudoun Dam Results of the original stability analysis for Fort Loudoun Dam are shown on Fiaure 2.4.4-13.

Because the resultant of forces falls outside the base, a portion of the spillway is 'udged to fail. Based on previous modes of failure for Cherokee and Douglas, the spillway is judged to fail above elevation 750.0 ft as well as the bridae suDported bv the soillwav niers.The results of the original slip circle analysis for the highest portion of the embankment are shown on Fiqure 2.4.4-14.

Because the factor of safety is less than one, the embankment is assumed to fail.No analysis was made for the powerhouse under SSE. However, an analysis was made for the OBE with no water in the units. a condition believed to be an extremely remote occurrence durino the OBE.Because the stresses were low and a large percentage of the base was in compression, it is considered that the addition of water in the units would be a stabilizing factor, and the powerhouse is 4udged not to fail.Figure 2.4.4-15 shows the condition of the dam after assumed failure. All debris from the failure of the concrete portions is assumed to be located in the channel below the failure elevations.

No hydroloqic routing for the sinqle failure of Fort Loudoun, includinq the bridge structure, is made because its simultaneous failure with other dams is considered as discussed later in this subparagraph.

Tellico Dam No hvdrologic routing for the sinqle failure of Tellico is made because its simultaneous failure with other dams is more critical as discussed later in this sub-paragraph.

Norris Dam Although an evaluation made in 1975 by Agbabian Associates concluded that Norris Dam would not fail in the SSE (with 25 year flood), Norris Dam was postulated to fail. The resulting debris downstream would occupy a greater span of the valley cross section than would the debris from the 2.4-31 SQN-OBE but with the same top level, elevation 970.0 ft. Figure 2.4.4-16 shows the part of the dam judged to fail and the location and heiaht of the resultina debris.The discharae ratinq for this controllinq, debris section was developed from a 1:150 scale hydraulic model at the TVA Engineering Laboratory and was verified closely by mathematical analysis.

The somewhat more extensive debris in SSE failure restricts discharge slightly compared to OBE failure conditions.

No hydrologic routing for the single failure of Norris was made because the simultaneous failure with Cherokee, Douqlas and Tellico Dams, discussed under multiple failures, is more critical.Cherokee The SSE is iudaed to oroduce the same Dostulated failure of Cherokee as was described for the OBE.The sinale failure does not need to be carried downstream because elevations would be lower than the same OBE failure in one-half the probable maximum flood.Douglas The SSE is iudged to produce the same postulated failure of Douglas as was described for the OBE.The single failure does not need to be carried downstream because elevations would be lower than the same OBE failure in one-half the probable maximum flood.Multiple Failures 5. NorriG, Cherokee, and Douglas Norris, Cherokee, and Douglas Dams werc also pestulatcd-to fail soimultancously.

Figure 2.4.4 2 shows the !Eoca;tion.

Of _An SSE_, and it6 attenuation, which produ~ec 0.15 g at Norris, 0.09 g at Cherokee and Douglas, 0.08 g at Fort Loudeun and Tollico, 0.05 g at Fontana, and 0.03 g at Watts Bar. Fort Loudoun, TelliGG, and Watts. Bar have boen judged not to fail for the one half SSE (acceleration value of 0.0-9 g) (see following discussion of non controling events). The bridge at Fr Loudoun DamR, however, m~ight fail under 0.08 g forces, falling OR any open gates, and on gate hofisting m~achinery.

TFrunion anchor belts of open gates would fail and the gates would be washed downstream, leaVing an open Spillway.

Cloced gates cou-ld not be opened. The mest coeorasetive assumption war, used that at the time of the seismAic event on the upstreamA tributary dame, the Gcrest Of the 25 year glood would likely have passed Fort Leudoun and flows would have been reduced to turbine capacity.

Hcnce spillway gates would be closed. As stated before, it is believed that multiple d-am. fiailure is extremely remote, and it seems reasonable to exclude Fontan~a on the basis of bein'g thc most distant in the cluster of dams under consideration.

For the postulated failur~es of Norris, Cherokee, and Douglas, the pertions judged to rem~ain and debris arran~gemenRt are as given i Figures 2.4.4 9, 2.4.4 12, and 2.4.4 15, respectively.

The SSE will produc~e the same postulated failures of Chero~kee and Douglas Dams6 as were described fo-r the onRe half SSE.For N~rris under SSE condibonsG, blgcks, 31 45 (883 feet of length) are judged to fail. The resulting debris downstream would occupy a greater span of the valley cross section than would the debris from the one half SSE but with the same top level, elevation 97-0. Figure 2.4.4-280 siheows the part of the controlling debris section was developed from a 1:150 scale hydraulic moedel at the TVAEnierg Labe~atory and was verfified closely by mathematical analysis.

The somewhatimoare exesiedebri in SSE= failre restricts discharge slightly comnparedtonehlSEfaurcndins The flooed- fo-r the postulated f-aiure cmitonwould overtep and breac;;-h Fort Loudoun Damn.Although transfer of water into Tellfico would occur, it would not be sufficpient to oveto th dam At Wlatts Bar Dam the; headwator wouldd reach 761.9, 2.1 feet below the top of the ea;rth erhmbnmt ofA tha ~ ~ ~ +k --im -4 m IUniarta4a~e r 4 IaiL rII~a'~ AI.,tt D- k-!, --n ,,I n ,a4---4-4n krc-hAM--rI 2.4-32 SQN-Resulting Wate-r SurlfaceP at SQN would reach elevation 706. This is 1.0 foot higher than plant grade This is the highest flood resulting from any combination of SSE seismic and flood eve-nts. The flood elevatiOn Flow and stage hydrOgraphs at the plant site is shown On Figure 2.4.4 30.6. Norris, Douglas, Fort Loudoun, and Tellico NorriS, Douglas, Fort Loudoun, and Tellico Dams were pestulatcd to failsmlanosy FigUre 2.4.4 31 shOWs the location of an SSE, and Its attenuation, which produces 0.12 g at Norris, 0.08 g at Dougla.0.12g ateFortLoudou and Telio,0.07gatCherokee, 0.06 g at.Fontana,and.0.

g Waftt Bar. Cherokee i6 judged net to fail at 0.07 g; Watts Bar has previously been judged not to fall at 0.09 g; and, for the same resn as given above, It seems Fea6onable to exclude FmOntana in this-failure combination.

For the postulattedd flaillurmes Of Norris, Douglas, ForFt Loudoun, and TellicO, the portiens judged to remain and the debFris arrangements are gi vn in F 9, 2.4.4 15, 2.4.4 26, and 2.4.4 27, respectively.

For analysis Fort Loudoun and Telli.o were postulated to fail com+pletely as the potions judged to remain are relatively small.The SSE will produce the same postulated failure of Douglas Dam as was described for the one hal SSE, Results of the stability analysis for Fort Loudun Dam are shownR o F.igue 24. Because the mresultantof foce falls outside the base, a paFtion of the spillway is judgs tod Rfil. Ba previos, modes of failure for Cherokee and Douglas, the spillway is judged to fail above elevation 7-50 as wel as the bridge supported by the spillway piers.The results of the Slip Gircle analysis for the highest poretio of the embankment aFe sho-wn o In Figure 2.4.4 25. Because the factOr of safety is less than one, the embankment is assumed to fai.No analysis was made for the powerhouse under SSE. However, an analysis was; Made for the one hal;f SS2-E with no water in the units, a condition believed to beP eXtrPemcly remote to occur durig h one half SSE. Because the stresses were lOW and a large percentage of the base was in compression, it is considered that the addition of water in the units woulId be a stabilizing factor, and the pGeweho)use is judged net to fail.Fmigure 2.4.4 26 shows the codtio f the- rdam after asmdfailu,_re.

All debris from the failure of the conrGete poFtion is assume-- -d to be locGated in the channel below the failure elevatios No structural analysis was made for Tellico Dam failure in the SSE=. Ber.aus~e of the similarity to Fr Loudoun, the spillway and entire embankment are judged to fail In a mnanner similar to Fort LOUdGun.F~igure 2.4.4 27 sheows afterF failure conpditions with all debris assumed locatpd inA the channel below the This postulated failure cmiaonresul-ts in Wattsý Bar htqadwater elevation 758.9, 8.1 foet below above the top o-f the_ em-bank~menpt of the mnain damn. The west saddle dike at Watts Bar Dam would be overtopped and broac~hed.

A com:plete washout of the dike woas asssumed.

The resulting water- level at SQN would be elevation 699.3, 5.7 feet below plant grade 705.O~ halIf SSE QGRGUFF8,f VAJth G~ HlJIf th P~bal Maiu .Ev M .~I., 1. Watts Bar Dam Stability analyses of Watts Bar Dam powerhouse and spillway s~ections result in the judgment that these structures will net fail!. The- a;nalyses ShOW loW stresses with about 38 percent of the sp"iway base in co.mpreSion

and- about 42 percent of the powerhouse bion. Results are given in Figure 2.4.4 1. Dynamic analysis of the concrete structures resulted in the determination that the ba;se acreoerationR is amplified at levels above the base.The slip cirle analysis of the eaFth embankment section results in Ra fator Of safety of 1.52, and the 2.4-33 SQN-!-- ^ A A ýAM-14uAAK-Mj04T2Ims iut;

uuuu uf h4A +;!I kil "UU~.I PrIVRAI !F4 rtuuru- 4 -J Z7 Normally for the condition of peak discharge at the damR f4or one half the PMF=, the spillway gates wul be in the wide open position (Figure 2.4.4 3). But, a~ly4 of the bridgo Structure for for~es resulting from a on-e h~alf SSE, including amnplificat'on of acceleration results in the determination that the bridge could fail as a result of shearing the anchor bolts. The dOWnstrcam bridge girder-s could 6trike the spillway gates. The impact of the girders striking the gates, could fail' the bolts Which anchor the gat tru-nninps, to the pier anchorages allowing the gates to fall. The flow ever the spiliWay crest would be the same as that prier to bridgc and gate f-ailure.

Hencc, bridge failure Will cause no adverse effect On the flood.A potentially con~eGedition is the bridge falling when most spillway gates would be closed. The gate heisting machine~'

would be inoperable after being struck by the bridge. As a result, the flood w:ould crest with the gates closed and the bridge deck and girders lying OR top of the spillway piers Analysis of the concrete podtions of the dam for the headwM.atfor for this condition shows that theyw l iRot f4a Floo-d lvl at SON for all the conditions described above is safely below plant grade elevation 705 2. FEA Loudoun Dam Stability analyses of Foed Loudoun Darn po~werhouse and spillway sections result in the judgment that these structures will Rat fail. The analyses show low base stresses, With near tWO thirds of the as i compesson.Results are giiven in FiOgure 2.4.4 4.Slip circle an~alysis of the cadth embankment result in a factor of safety of 1.26, and the emba~knkent is judged not to fail. Results are given in Fiegure 2.4.4 5.The spiliway gates and bridge arc of the same design as those at W~atts Bar DamA. ConAditions Of failure during a one half SSE are the same, and no problems-aRep li6kely. Coincident f-aUilure-at FGed Loudoun and Watts Bar does nOt occur.ForF the potentially critisal caste Of FAed Loudoun bridge failure at the onRset of the main peodion of one half the probable m-axi~m.um flooed- flow, into Faed Leudeun ReceF'eiF, It was found that the Watts Bar inflowVs areFP mHuch less than the condition resulting fromR simYultaneous failur-e of Chwerkee and Douglas-.3. Toellico Dam No padt of Tellico Dam is judged to fail. Results of the stability analyses for a typical non GVe~fOW block and a typical spillway blocGk are shown in Figure 2.4.1 6. The result of the stability analysiso the earth emnbankment is shown in Figur~e 2.4.4 7 and ind.icates6 A- factor of safety of 1.28.4. Cherokee Dam No hydrologic r~esults are give-n fo-r the GIngle failure of Cheroakeee Damri bec-a-usee the simul---ta-neous--r failure of Chero~kee and Douglas is mne-re c~ritic-,al.

5. etiqa6 Dam4 No) hydrolo~gic results are given for the single failure o~f Douglas Dam because the simultaneous failure o-f Che'rokee and- Douglas. is more critical.6.Hiwassee~t River Dams-Hiwaccee Dam was assumed to fail in the one half SSE. No hydrologic results are given for the single fad, rA efHiwAgslso Da;m because its4r simul.taneous failre wth ther dams is more critical.7. Apalahi~a 2.4-34 SQN-A.ala.chi DWam, was ;- isA'Jmd to tfail an thA on, hiltf SSE Nio hdrloic rul- t;, ar rivn ftr thp Sirle-Z A --t a GA. la -am A- re 9-;-; W-6 61 1 + A -I I N; A A A A a Ii 9 M.B. Blue R~ge Blue Ridgc Dam was assumed to fail in the oRne half SSE. No are given for the singlo failur eof Blue Ridge Dam bccausc its simultanoouIS failu-re-with otho-r dsis mon-rR- criticalR-.

9. Ocoee No.-1 Ocnoecl ro-sh. 1 Dam was assnuodI to fail in the Pon half SSE. No hydrologic result6 arc giren for the Single failure Of Ococe No. 1 Dam bccause its simultanous failure with other dams is more critical.Nottely Dam wau assumed to fail in the one half SSE. No hydrologic results afe giVcn for the single failure of Nottely Dam because issimultaneous failure with other damns is more critical.11. Chatuge Ghatuge Dam is a ho gee o, nmpepwieur.

rolled fill dam. With the epicenter of the one half SSE located at the dam, the maximum ground accelcration atI Cha;tuge is 0.09g g. Gron, d accelerationsg o this mtagnitude should have no detrimcn~ta!

effects On a well cOnstructed compacted oadthfill embankment.

We knew of no failueresf compacted carth embankment slopes fromn earthquake motions. F~ailures to date have been associated with o~ther liquefaction of hydraulic fill embankmcn~tS of liquefaction of loose granular foundatien materials.

The rolled embankment materials in Chatuge are no~t sensitive to liquefaction.

To Yerif' these conclusion analysis using the "Ncwmark Methodfo the Dynamic Analysis of EmbakmeRnt Slopes" (New'mark, N. M.,

of Ea-thquake

o. [Dams of FEmbankments," Geetechnique 15-440 141, 156, 1965) was made to determine the structural stability of Chatuge. We conducted a field exploration boring program and laboraton.'

testing program of samples obtained in the field exploration.

DuFrig the field exploration program, standard penetratio tests blow cn,,ts were obtained On both the embankment and its foundatioRn mFateials.

Both static and dynamic (cycIli) triaxial shear tests were made. This information was used in the Newmar Method of Analysis.

We concluded from the Analysis that the Chatuge Dam can easily resist the qround acceleration Of 0.09 g With no detrimnental damageK .v v 12. Hiw:assee.

Aoalachia.

BluJe Ridee. Ocoee Ne. 1. and .Noelev Hiwassee, Apalachia, Blue Ridge, Ococe No.1, and Ngttely Damsr could fail when the one half SSE is critcaly !oted Allfiv das wore assumned to cOmnpletely disappear in this event. Resulting crs level at SQN would be below plant grade 705.1. Watts Bar Dam A reeAv'aluation waS nnt made ..r Dam condi SS t Apevieus evaluation had determined that even if the dam is arbitrarily remeved instantaneously, the level at the nucl'ear plant site w~ould- be below plant grade.2. Fo-rt Loudoun Dam No hydrologic routing fo the single failure- of Fot L=-udeun, including the bridge ifs Made because Its simulitaneous failure with Tellico and Fontana, as well as with Tolfico), Norris, and Douglas, 3. Toqllico Da;m 2.4-35 SQN-No ro)uting for the Gingle failure of Tellice is made fo-r the re-arons given above for Fort Loudo-un.4. Norris Dam This postulated Single failure would result in peak headwator at Watts Bar bolow the top of the earth PG~t4GR&-Gf the damn. Routing waS not carried furAther bec~ause it was evident that flood levels6 at the plant site would be considerably lower than for the Norris failure in the one half SSE combined with the o~e half MF.5. Hiwassee River Dams Considered Separately No structural analyses were made for Chatuge, Nottely, Blue Ridge, Ocoee No. 1, Hiwassee, and A palachia in the SSE= Instead, all six dams were postulated to fail Opely No roguting for the failure of the six Hiwassee dams alone i6 m~ade because their simultaneous failure 4:th FonRtana is considered as discussed earlier in this subparagraph.-

6. Cherokec, Douglas, and Fontana Considered Separately The SSE= will produce the same postulated failures of Chero~kee, Douglas, and Fon)tana Dams as were described for the one half SSE. None of these failures need to be carried downstream, however, because elevations would be lower than the same failures in one hal;f the probable mxumflood.4

7. Fort Loudoun. Tellico. and Fontana A~ c~~p "nnt'~rnd hctnnn Fnntin~ inn th~ Fnrt I nirlni'n Thilirn comnI~" '*'i~' nn5~t'¶I1tn~d In fiji thr~'r three dams. The four ALCOA dams downstreamR fromn Fn~tana and Nantahala, an ALCOA dam, ups~treamn were also postulated to fail completely in this event. W~atts Bar Dam and spillway gatesw:ould remain intact, but tailure of the roadway bridge was postulated whic-hwoould-render the spillway gates inoperable.

At the time of seismnic failure, discharges would be small in; the 25 year flood. F-or GOnseR'atism, Watts Bar gates w~ere _assumed inoperable in the closed position after the S-S-Emet This event would result in a flooed- level at the nuclear plant site below 705 plant grade.8. Douglas and Fontana Douglas and Fontana wore postulated to fail si~m.ultanReously.

The locration ofRan SSE required tofi both darns would produce 0. 14 g at Douglas, 0.009 gat Fontana, 0.07 g at Chero~kee, 0.05 g at Norris, 0.06 g at ranF I=auaUn) aRia i ciiica, aria 0.03 g at Watts Bar. i-or Mne postluated m.aIiures of Deugias and Font.ana, the potions ju"dged to roman and the dbris arrangements aed as given in Figures 2.4.4 15 and 2.1.4 16. Fort Loudoun, T-ellico, and WaI;tts, Bar haepviulbenjddnotoal fonr the_ OBE (0).09 o ). The bridge at Loud.oun Damn, however, might fail u. n d 06 g f -oes, falling OR gates and OR gate hoisting mnachinery.

Fort Loudoun gates wore assumned inoperable in the closed po)sition folloEwing the SSE Gevet. ReSUltn~g water sU~face at SQN would be below plant grade.91 Fontana and Hiwoassee River Dams FonRtana and six Hiwassec River dams Hiwassee, Apalachia, Chatuge, Nottly, Blue Ridge, and Ocoe-e- No. 1 were postulated to fail simultaneously.

For) the postulated failure of Fontana, the portion judged to remain and the_ debris_ a;rraneet ar as given in Fmigure 2.4.4 16. The six Hiwassee daR-msR wereM as~sume-d to fail completely.

Fort Loudeun, Toellico, and Watts Bar are judged not to fail'NA1,th all gates operable.

The=Fontana surge comhined;, with that of ther six ia,,ee Ri;vr dar-mswounIld reach an elevation at the plant site below the plant grade.2.4.4.2.2 Failue Analys , All upstream auu GW~tiuam dams6 which could have Sinfraton flood level at SQ wore examined for po-te-ntial; fa-ilu-reA during all fiood conditions, which would have the potential to v 2.4-36 SQN-Produ~e mnaximum plant flood levels includinG the dam PMF at the ind-iVidui-al unstroam; dams 1 6EncRGcte secUonS W8er eoxam~noo Tar ovenuMrng -Ana- norizentai snear and sldang. zspillway gates.:ere examined for stability at potentially critical wa-ter lelsand against failure fromn being Struck by water borne objects. Leocks and- lock] gts wer xamnined forF stability, and earth em~bankmonets wcrc exa nd for erosion due to overtopping.

During the SQN PMFl the only failure would be the west saddle dike a;t Waifs Bar. C-hicka;mauga Dam would be overtoonoed but wsxas ~oprse~vatively assumed not to fa4l.I I Ii POF conrGete dam sections, comparisonG were made oetween thle original design fleadwater and taihA4ater le'els and those that would prevail in the PMF. if the oveFturning moments and horizontal forces were not inrGeased by more than 20 percent, the structure eecnidered safe against failure. All upstFeam dams passed this test except Douglas, Feor Loudoun, and Watts Bar Original designs showed the spillway section.s÷ of these dhams to be mos..t vulnerable.

These splway sections:erae examined in futher detail and judged to be stable.During peak PMF= conditions the radial spillway gates of Fmort Loudoun and Watts Bar DamsG will be wide epenn with flow over the gates and under the gates. For this cnGdition both the static and dynamic load stresses in the main structural mnembers of the gate will be less than the yield stress by a f-ctor of three. The stFres in the trunniop pin is less than the allowable design stress by a factor grFeater than 10. The trunnion pin is prevented from disledgment by a key inft the gate anchorage assembly and fitting into a slot in th. in.The gates were also investigated for the cndition n when rising headwatFr firsFt begins te eXceed the bottom of the gates in the wide open position.

This condition produces the largest forces tending to retate the Fradial gate upward. InR the wide Poition the gates are dogged against steel gate steps anchored to the conrGete piers. The stressqes in the gate stop members are less than the yield stress Of the- material by afactOr Of 2L.It is concluded that the above listed margins are sufficient to provide assurance also that the gates will net f-;il as a esult of additional stresses whiGh may result fFom possible vibrations of the gates Consideration has been given to the effect of water borne objectS striking the spillway gates and bentS supporting the bridge- ac-ross W.atts Bar Dam at pealk ;.watte~r llevel at the dam. The most severe potential for damage would be by a barge which has been torn loose from its moo..rgs and floats the-a Should the barge h the spillway portion of the dam end 9o, one bridge bent could be failed by the barge and two spillway gates could be damaged and possibly swept away. The loss of onRe bridge bon-;t will1 not collapse the bridge because the bridge girders are continuou 10SMembers; and the str~ess in the girders Will be less than the ul-tima;tetrs forF this condition Of one support befing lost. Should two gates be swept away, the nappe of the water su~face over the sp~iway Weir would be such that the barge would be grounded on the tops of the conrGete spillway weir-s and provide a partial obstuction to- flowN comparable to unfailed spillway gates. Hence the loss of two gates frolm this nau se will have little effect on the peak flow and loVa;tion.

Should the barge approach the spillway portion broadside, two and possibly thFree bidge bohnts couild be failed. For this condition, the bridge would collapse On the barge and the barge would be grounded onA the tops of the spillway weirs. This would be probable because the approach velocity Of the barge w:ould be fromF 4 to) 7 Mi les per hour and the bottomA of the barge would be about six inches above the tops of the weirs. For this- coend-ition the barge would be grounded before striking the spillway gatos 2.4-37 SQN-henause theo gates aro abut20- feePt dontcmfom the leg of the upstroamA bridcio bents.w Legk-Gates The lock gates at Fort Loudoun, Watts Bar, aRd Chickamauga were examined for possible fai'ure with the conclusion that no potential for failure exists because the gates are designcd for a differential hydrostatic head greater than that which exists d-ring the probable m.aximum flood, Eimbnkhnel mnn D a~nh4n in the 1998 reanalysis, the only embankment would be the wesot saddle dike at Watts BaF Dam.Chickamauga Dam, downstreamR of the plant, would be overtopped but was assumed not to fail. Thiss is conse awtive as failure of Chickamauga Dam would slightly lower flood elevations at the plant.The adopted relationship to c.mpute the rate of erosion in an earth dam, failure is that developed and used by the Bureau of Reclamation in With 4is safety of dams prOgram The expression relates the volume of eroded fill material to the volume of water flowing through the breach.The equabo i~&L Qsoil-Ke-Qw'ater-whe~e A S..- Vol.um.e of soil eroded n.each tme period Pý -nnQiRS4if (4r MFnnnrtionamV~r 4 Tflot solnl in .. ..diScharge relationships in this study e -Base Of natural logarithm systemr b Ha~Wher-e b -Base length of evefl(eW channel at any given time H- H"daui head at any given time-~j Developed angle Of friction Of 6ol0atril A GOnsewative value of 13 degrees was adopted for materials in the dams investiciated.

Solving the equation, which was computerized, involves a tria! and error procedure over short depth anRd time in.rements.

In the program., depth changes of 0.1 foo0t Or less are used to keep time incr emets to less than one second during rapid failure and up to about 350 seconds prior to breaGhýýThe so..lution of an earth embankment breach begins by so.v.ng the er.ion equation ursn a headwatorel..

ation hydr.graph a..um.ing no failure. Erosion is pestulated to occur across the entire earth secAtioA ;;Ad to start at the down..tream.

edge; whAe heA;d4 wat-r elevations reac.hed a se-leted depth above the dam top elevation.

Subsequently, when erosio;An reachesr the upstream edge of the embankment, breaching and rapid lowerin~g of the emnbankmnent begins. T-hereafter, comnputations, incl'ude headwater adiustments for inrGeased reservoir Gut-flGw resulting from the breach.Watts Bar West Saddle Dike Embankment Failure 2.4-38 SQN-Figure 2.4 A 37 is a general plan of Waftt Bar .hoWing elevations and sectiens.

Figure 2.4.4 38 is a topographic Map of the general Vicinity of Watts Bar Dam, Figure 2.4.4 39 is a general plan and section of the west saddle dike.The west saddle dike .a 3xmnd and found subject to aiur from.r GYcropping.

This failure was assumed to be a complete washout and add to the discharge fromn Watts Bar Dam.Some v...fi.ation for the breaching com.putational p.rocedures ilustrated above was obtained b comparison with actual failures reported in the literature and in inforFm~al d-iscGussion With hydrologic-engi neer. These repo ..ts show that ove.topped earth cmbankmentsdo n ot necessarily fail. EaFth emb-ankm.e.-nts have sustained overtopping Of several feet f.o several hours before failurFe ocurred.An extreme example is Ores earth dam in Brazil [17] which was, overtopped to a depth otf aprxmately 2.6 feet along a 2,000 foot length fo-r 12 ho-urs beforep bre-ac-hing began. Once an earth embhank~men-t s breached, failure tends to progress rapidly, however. How rapidly depends upon the material and headwate~rdenths dijrnn faiure Gemniete fiailt res comniited in this and other stujd Fn hav. va...ri.d from about one h-af t. six hou.rs a.ft.r initi ......... I'" .................vv This, is ncnsistent with Artual ChiGkamauqa EmbaRkmeRt F=ai!uFe-in the dewRn---- ý I I --..rlnl; I analysis:I thela iIluru Af [A;t-RR [:rmeajikmRtl-fI:

11 "MPr M--...ruKJl:u;

.... .-a rurrM- ONP rP"HIPAC rAHUuu 44000 uI.'uiI. AT TPA ulAri[ F- u.9 M-1ui.ruiri iriKridiiAi im.pro.em.ents are planned for Chickamauga Dam, which if implemented, would prevent failrc.Therefore, although overtopped in the PMVF, the damA was assumed net to fail in deteFrmining flood elevations at the plant. This assumRption is conservative.

TVA considered the following multiple SSE dam failure combinations.

(5) The Simultaneous Failure of Norris, Cherokee, Douglas and Tellico Dams in the SSE Coincident with 25-year Flood The SSE must be located in a very precise region to have the potential for multiple dam failures.

In order to fail Norris. Cherokee.

Doualas. and Tellico Dams. the eoicenter of SSE must be confined to a relatively small area the shaoe of a football.

about 10 miles wide and 20 miles Ione.Figure 2.4.4-17 shows the location of an SSE, and its attenuation, which produces 0.15 g at Norris, 0.09 q at Cherokee and Douqlas, 0.08 q at Fort Loudoun and Tellico, 0.05 q at Fontana, and 0.03 q at Watts Bar. Fort Loudoun and Watts Bar have previously been judged not to fail for the OBE (0.09 g).The bridge at Fort Loudoun Dam, however, might fail under 0.08 q forces, falling on any open gates and on aate hoistina machinery.

Trunnion anchor bolts of ooen aates would fail and the aates would be washed downstream, leaving an open spillway.

Closed gates could not be opened. By the time of the seismic event at upstream tributary dams the crest of the 25 year flood would likely have passed Fort Loudoun and flows would have been reduced to turbine capacity.

Hence, spillway qates would be closed. As stated before, it is believed that multiple dam failure is extremely remote, and it seems reasonable to exclude Fontana on the basis of beina the most distant in the cluster of dams under consideration.

For the postulated failures of Norris, Cherokee, and Douglas the portions iudged to remain and debris arrangements are as given in Figures 2.4.4-16, 2.4.4-7, and 2.4.4-10, respectivelv.

Tellico is conservatively postulated to completely fail.As discussed in Section 2.4.3, temporary flood barriers are installed on embankments at Fort Loudoun and Tellico Reservoirs.

The temoorarv flood barriers are assumed to fail in the SSE and are thus not credited for increasing the height of the Fort Loudoun or Tellico Reservoirs embankments.

The flood for this postulated failure combination would overtop and breach the south embankment and Marina Saddle Dam at Fort Loudoun. At Watts Bar Dam, the headwater would reach elevation 765.54 ft, 4.46 ft below the top of the earth embankment of the main dam. However, the West Saddle Dike with top at elevation 757.0 ft would be overtopped and breached.

The headwater at Chickamauga Dam would reach elevation 701.14 ft. 4.86 ft below too of dam. The embankments at Nickaiack Dam would be 2.4-39 SQN-overtopped but was conservatively postulated not to breach.The maximum discharge at SQN would be 974,937 cfs. The elevation at the plant site would be 706.0 ft, 1.0 ft above 705.0 ft plant grade. This is the highest flood elevation resulting from any combination of seismic events.In addition to the SSE failure combination of Norris, Cherokee, Douglas, and Tellico identified as the critical case, three other combinations were evaluated in earlier studies. These three originally analyzed combinations produced significantly lower elevations and were therefore not reevaluated.

These include the following:

1. Norris, Douglas, Fort Loudoun, and Tellico 2. Fontana, Fort Loudoun, and Tellico 3. Fontana and Douglas In order to fail Norris, Douglas, Fort Loudoun, and Tellico Dams, the epicenter of an SSE must be confined to a triangular area with sides of approximately one mile in length. However, as an extreme upper limit the above combination of dams is postulated to fail as well as the combination of (1)Fontana, Fort Loudoun, and Tellico; and (2) Fontana and Douglas.Norris, Douglas, Fort Loudoun, and Tellico Dams were postulated to fail simultaneously.

Figure 2.4.4-19 shows the location of an SSE, and its attenuation, which produces 0.12 q at Norris, 0.08 q at Douglas, 0.12 q at Fort Loudoun and Tellico, 0.07 q at Cherokee, 0.06 q at Fontana, and 0.04 q at Watts Bar. Cherokee is iudged not to fail at 0.07 q: Watts Bar has previously been iudged not to fail at 0.09 q; and, for the same reasons as given above, it seems reasonable to exclude Fontana in this failure combination.

For the postulated failures of Norris, Douglas, Fort Loudoun, and Tellico, the portions iudged to remain and the debris arrangements are as given in Figures 2.4.4-16, 2.4.4-10, 2.4.4-15 and 2.4.4-20 for single dam failure. Fort Loudoun and Tellico were postulated to fail completely as the portions iudged to remain are relatively small. This combination was not reevaluated because previous analysis showed it was not controlling.

An SSE centered between Fontana and the Fort Loudoun-Tellico complex was postulated to fail these three dams. The four ALCOA dams downstream from Fontana and Nantahala, a Duke Energy dam (formerly ALCOA) upstream were also postulated to fail completely in this event. Watts Bar Dam would remain intact. This flood level was not reevaluated because previous analysis showed it was not controlling.

Douglas and Fontana Dams were postulated to fail simultaneously.

Figure 2.4.4-21 shows the location of an SSE and its attenuation, which produces 0.14 q at Douglas, 0.09 q at Fontana, 0.07 g at Cherokee, 0.05 g at Norris, 0.06 g at Fort Loudoun and Tellico, and 0.03 q at Watts Bar. For the postulated failures of Douglas and Fontana Dams, the portions judged to remain and the debris arrangements for Douglas Dam are as given in Figures 2.4.4-10 and 2.4.4-11 for Fontana dam failure.Fort Loudoun and Watts Bar Dams have previously been iudged not to fail for the OBE (0.09 q).Postulation of Tellico failure in this combination has not been evaluated but is bounded by the SSE failure of Norris, Cherokee, Douglas and Tellico.2.4.4.32 Unsteady Flow Analysis of Potential Dam Failures Unsteady flow routing techniques

[261 were used to evaluate plant site flood levels from postulated seismically induced dam failures wherever their inherent accuracy was needed. FoF PMF determinations unsteady In addition to the flow models described in Section 2.4.3.3. wore used- f routing floodsG from.. postulated 6eiSm°ically ind"uc-.ed.

damn failures, of tributary dams, additional unsteady flow models were used as adjuncts to those described in Section 2.4.3.3models described below were used to develop the outflow hydrographs from the postulated dam failures.

The HEC-HMS storage routing was used to compute the outflow hydrograph from the postulated failure of each dam except main river dams. In the case of dams which were postulated to fail completely (Hiwassee, Apalachia and Blue Ridge), HEC-RAS or SOCH was used to develop the outflow hydrograph.

For Tellico Dam, the complete failure was analyzed with the SOCH model.2.4-40 SON-Un'stoeay ,ow tecnigques w..e applie, in Neoris Resew(*i.

The Norr-s Resepe.ir model wM%developed in sufficiot detail to do-fine the m~annor in Which the roserv~oir mouln upyan uti ,4,,,,,,,,,,,.,,4~~~~~~

~ h..,. ., .-,.. ,_,-1 suppl a.nd,, ,, sustain.u.flW fo...Wi, g postulated dam fa..u.e. The mRO .. utdowas vefrified by fa Mpaire g its outed heoadwater level in the one half PMF= with thoso using storagc routing techniqueS.

Headwater level agreed within a foot, and the model was co~nsidered adequate for the purpose-.Unstoady flow techniques wcrc also applied in Cherokee, Douglas, and Fon~tana Reser~'oirs.

The reseR'oier mo-dels; werP-e developed in sufficient detail to define the mtanner in w~hich the re~sepvoir-s would supply and sustain outflowA.

followinAf'g postulated dam failure. The failure time and initial reservoir elevations for each dam were determined from a pre-failure TRBROUTE analysis.

HEC-HMS was used to develop the post failure outflow hydrographs based on the previously determined dam failure rating curves. The outflow hydrographs were validated by comparing the HEC-HMS results with those generated by simulations using TRBROUTE.2.4.4.43 Water Level at Plant Site Maxim water level at the plant from- differet postulated co,-mhination Of e.-.-, seismicdafilur

,.es coincident With floods would be elevation 707.8, excGluding wind wave effects. it would result from the one half SSE fafilure of Fontana, Hiwassee, Apalachia, and Blue Ridge Damns coincident with one-half the probable maximum flood. March win~d with one percent eXceodance probability over the 1.4 mile effective fetch from the critical nrOth northwest direction is 26 miles per hour over land. This would cause reev )rwvs to reach elevation 7-09.6. Runup could reach elevation

710.1 on a smoo~th 4:1 slope, elevation 712.8 9R a Yectical wall in shallow (4.9 feet) water, and elevation 710.4 OR a Ye~tical w-all in deep water.The unsteady flow analyses of the five postulated combinations of seismic dam failures coincident with floods analyzed yields a maximum elevation of 708.6 ft at SON excludino wind wave effects. The maximum elevation would result from the OBE failure of Cherokee.

Doualas and Tellico Dams coincident with the one-half PMF flood postulated to occur in March. Table 2.4.4-1 provides a summary of flood elevations determined for the five failure combinations analyzed.Coincident wind wave activity for the PMF is described in Section 2.4.3.6. Wind waves were not computed for the seismic events, but superimposed wind wave activity from guide specified two-year wind speed would result in water surface elevations several ft below the PMF elevation 722.0 ft described in section 2.4.3.2.4.5 Probable Maximum Surqe and Seiche Floodinq (HISTORICAL INFORMATION)

Chickamauga Lake level during non-flood conditions could be no higher than elevatin 685.441, top gates, and i; not likely to would not exceed elevation 682.5 ft, normal summe maximum pool level, for any significant time. No conceivable hurricane Or cyclonic typo windsmeteorological conditions could produce the over 20 feet ef wave height required tea seiche nor reservoir operations a surge which would reach plant grade elevation 705.0 ft, some 22 ft above normal maximum pool level.2.4.6 Probable Maximum Tsunami Floodin.q (HISTORICAL INFORMATION)

Because of its inland location, SON is not endangered by tsunami flooding.2.4.7 Ilee -it 9 o ~a Rd LanRds idesQ (141 S- TO RI GAL IN F RMIA T. 1 -0 14 ect Because of the location in a temperate climate; significant amounts of ice do not form on4the TeRnessee Valley rivers and lakes. SON is In no danger from ice floodig,, .lakes and rivers in the plant vicinity and ice iams are not a source of maior flooding.Flood waves from landslides into upstream reservoirs pose no danger because of the absence of major elevation relief in nearby upstream reservoirs and because the prevailing thin soils offer small slide volume potential compared to the available detention space in reservoirs.

2.4-41