ML15224A370

Oconee, Supplemental Information on TIA 2014-05, Potential Unanalyzed Condition Associated with Emergency Power System

ML15224A370
Person / Time
Site:	Oconee
Issue date:	08/07/2015
From:	Batson S Duke Energy Corp
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
ONS-201 5-094
Download: ML15224A370 (34)

Text

~' EI\ERGYOconee Nuclear StationDuke EnergyONOIVP I 7800 Rochester HwySeneca, Sc 29672ONS-201 5-094 0: 864.873.3274f864.873.4208Scott.Batson@duke-energy.comAugust 7, 2015U.S. Nuclear Regulatory CommissionDocument Control DeskWashington, DC 20555

Subject:

Duke Energy Carolinas, LLCOconee Nuclear Station,Docket Nos. 50-269, 50-270 and 50-287Supplemental Information on TIA 2014-05, Potential Unanalyzed ConditionAssociated with Emergency Power System

References:

1. Oconee Nuclear Station -NRC Component Design Bases Inspection Report05000269/2014007, 05000270/2014007, and 05000287/2014007, dated June 27, 2014(ML14178A535).2. Letter from Duke Energy (Scott Batson) to USNRC (Document Control Desk), "TIA2014-05, Potential Unanalyzed Condition Associated with Emergency Power System,"dated May 11,2015 (ML15139A049).The purpose of this letter is to supplement information provided by Duke Energy Carolinas (DukeEnergy) in its May 11, 2015, letter (Reference 2) to the Nuclear Regulatory Commission (NRC).The focus of the May 11, 2015, letter was to provide additional Oconee licensing basis informationassociated with the 2014 Component Design Basis Inspection (CDBI) Unresolved Item that mightnot be readily available to NRC reviewers. Attachment 1 to the May 11, 2015, letter, "Review ofthe Design and Installation of Medium and Low Voltage Cables in Trench 3 at Oconee NuclearStation," described the Oconee licensing basis. Duke Energy also provided Attachment 1 tosummarize for the NRC, Oconee's licensing basis, which spans over 40 years of operation andincludes numerous plant modifications, including the emergency power system cableconfigurations that are the subject of the TIA. Included in Attachment 1 was a comparison of theOconee Trench 3 cable configuration to the Oconee licensing basis and to various NRCdocuments pertinent to cable design and cable faults. A key point made in the attachment is thedifference between the Oconee single failure licensing basis and the single failure licensingrequirements that currently exist but are not applicable to Oconee.Attachment 2 (to Reference 2), "UL 1569 Impact and Crush Tests on Keowee Underground TrenchPower Cables," provided the results of the cable armor testing conducted at the OkoniteCompany's High Voltage Laboratory in Paterson, New Jersey. This testing, performed in Spring2015, confirmed the mechanical properties of the bronze tape shielded 13.8 kVac emergencypower cables installed in Trench 3. More importantly, the testing demonstrated that the cableconfiguration with bronze tape shield provides adequate mechanical protection to perform aswww.duke-energy.com _.,

ONS-2015-094TIA 2014-05August 7, 2015Page 2armored cable per Underwriters Laboratory Test 1569, Sections 24, 25, and 26. It is noteworthythat, during cable testing, there were no instances where electrical continuity between the cableconductor(s) and the metallic shield occurred.Based on Oconee relevant licensing basis documents, cable testing results provided in Attachment2 to Reference 2, the properties of the Keowee generator impedance grounding system, and theOperability Determination conducted for the Trench 3 configuration, Attachment 1 to Reference 2concluded that the Trench 3 cables are capable of performing their intended functions and complywith the Oconee licensing basis.During the week of July 7, 2015, an Office of Nuclear Reactor Regulation (NRR) Peer ReviewTeam visited Oconee to conduct plant walkdowns and to review the Oconee design and licensingbasis with respect to the TIA. As a part of this visit, Duke Energy held detailed discussions with theNRC team on plant design and licensing basis issues. NRC feedback to Duke during thediscussions revealed that, during their visit, the NRC was provided new information of which theywere previously not aware. Therefore, Duke Energy is supplementing the May 11, 2015, letter withthe information provided herein to ensure the NRC has the pertinent facts and relevant informationto support NRC development of the TIA.Attachment 1 to this letter provides the following additional information to the May 11, 2015, DukeEnergy submittal:* A discussion of the Duke Power Company fleet design process with respect to the genericapproach taken regarding cable separation and the impact of cable armor on faultpropagation. This approach included the use of cable armoring to provide excellentprevention of cable to cable faults or failures.* A discussion of how Oconee design is consistent with plant licensing basis. Thissupplements the information provided in Reference 2, Attachment 1 and provides thehistory of the licensing basis relevant to the TIA beginning with the construction permitpreliminary safety analysis report (PSAR). It mirrors the information discussed with theNRR Peer Review Team on July 7, 2015.* An outline of the failure sequence necessary to damage DC circuits. A cable fault occursas a result of the failure of the insulation between the conductor and the bronze groundingtape on a single cable. If the fault is going to produce an interaction that will affect thecontrol cables that are on the bottom of the trench, it must propagate to adjacent powercables and/or grounded cable supports. The control cables are routed below the powercables. For the power cable to control cable interaction to occur, the fault would have tobypass the various grounding systems associated with the power cables and control cablesand the sequence of events would have to occur before the protective relaying clearedthe fault(s).* A discussion of certain design standards with respect to the Oconee licensing basis.Oconee was issued construction permits for all 3 units before January 1, 1971, andconsequently 50.55a(h)(2) only requires that the design of the protection system be inaccordance with the licensing basis.* A discussion of industry operating experience with respect to power cable failures. Thisdiscussion outlines the data bases researched, the selection of the incidents that weresimilar/applicable, and a comparison of the cable design related to the failure with theOconee design.

ONS-201 5-094TIA 2014-05August 7, 2015Page 3There are no new or revised regulatory commitments being made in this submittal.Should the NRC require any additional information, please contact Chris Wasik, OconeeRegulatory Affairs Manager, at 864-873-5789.Sincerely,Scott BatsonVice PresidentOconee Nuclear StationAttachment 1- Supplemental Information Related to TIA 2014-05, Potential Unanalyzed ConditionAssociated with Emergency Power System ONS-201 5-094TIA 2014-05August 7, 2015Page 4xc (with attachments):Mr. Victor M. McCreeAdministrator, Region IIU.S. Nuclear Regulatory CommissionMarquis One Tower245 Peachtree Center Ave., NE, Suite 1200Atlanta, GA 30303-1257Mr. Eddy L. CroweNRC Senior Resident InspectorOconee Nuclear StationMr. James R. HallNRC Senior Project Manager(File via E-mail)U.S. Nuclear Regulatory CommissionOne White Flint North, M/S O-8G9A11555 Rockville PikeRockville, MD 20852-2746Mr. Jeffrey A. WhitedNRC Project Manager(File via E-mail)U.S. Nuclear Regulatory CommissionOne White Flint North, M/S O-8B1A11555 Rockville PikeRockville, MD 20852-2746Ms. Holly D. CruzNRC Project Manager(File via E-mail)U.S. Nuclear Regulatory CommissionOne White Flint North, 12E111555 Rockville PikeRockville, MD 20852-2746 Ito ONS-2015-094Page 1 of 30Attachment 1Supplemental Information Related to TIA 2014-05, Potential Unanalyzed ConditionAssociated with Emergency Power System Attachment I to ONS-2015-094Page 2 of 30The 2014 Oconee Component Design Basis Inspection (COBI) is documented in the Reference 1NRC Inspection Report. This report initiated Unresolved Item (URI) 05000269,270,287/2014007-05, Potential Unanalyzed Condition Associated with Emergency Power System, which describesNRC concerns related to the design and installation of 13.8 kVac emergency power cables and125 Vdc control cables within the Keowee underground concrete raceway system (Trench 3). Asnoted in the report, Region II has requested assistance from NRR via TIA 2014-05, to review theemergency power system licensing basis to determine the acceptability of the design.Duke Power Company Design -Cable SeparationSimilarity to Other Duke Power Company Plant-related Approvals by the NRC:Duke Power Company was the Engineer/Constructor for the three Oconee Nuclear Station(ONS) units, as well as for the later units at McGuire Nuclear Station (MNS) and CatawbaNuclear Station (CNS). All three facilities used a similar design philosophy with respect to theuse of armored cable to provide adequate cable separation.ONS Updated Final Safety Analysis Report (UFSAR) Section 8.3.1.4.6.2, "Cable Separation,"states, in part:"It should be pointed out that the cable armors used provide excellent mechanical andfire protection which would not be provided with conventional, unarmored cable systems.It is our intent wherever physically possible to utilize metallically armored and protectedcables systems."MNS U FSAR Section 8.3.1.4.1.5, "Cable Application and Installation," states, in part:"Armored cable which has been demonstrated to be an excellent barrier to externallyand internally generated fires is used throughout the plant. Short circuit tests have beenconducted on the interlocked armor cable by Duke Energy. These tests havedemonstrated its acceptability as an adequate barrier by preventing damage to adjacentcables."ONS UFSAR Section 8.3.1.4.5.2, "Cable Separation," states, in part:"Interlocked armor cable has been demonstrated through short circuit testing conductedby Duke Power Company to provide an adequate barrier for preventing damage toadjacent cables."The basis of the statements in the MNS and CNS UFSARs regarding short circuit testing is aseries of tests documented in a November 8, 1977, Duke Power Company report, Report ofPower and Control Cable Overload and Short Circuit Tests Performed for McGuire NuclearStation. This testing was performed by the High Power Laboratory of the WestinghouseCorporation.A March 22, 1978, Duke Power Company letter to the NRC on the MNS docket stated thefollowing (emphasis added):"Testing performed by Duke Power Company at the Westinghouse High PowerLaboratory in East Pittsburgh, Pennsylvania demonstrated that adjacent armored cableswithin the same tray will not be damaged due to short circuiting of the power cables.These test reports have been submitted to the NRC and more than adequatelydemonstrate that these types of power cables pose no threat to the redundant safetytrains." to ONS-2015-094Page 3 of 30An August 1, 1978, Duke Power Company letter to the NRC on the MNS docket responded, inpart, to NRC questions on the High Power Laboratory testing. The letter states:"The test configuration was chosen to be ultraconservative with respect to actual powercable installations. Although power cable trays in the plant are installed above controltrays, the power tray was purposely placed between two control trays for this test toincrease combustible loading around the faulted cable."Supplement No. 2, dated March 1, 1979, and Supplement No. 5, dated April, 1981, to the SafetyEvaluation Report for the MNS Operating License (NUREG-0422) both state:"The applicant has conducted tests which demonstrate that no fire propagation fromcable to cable or tray to tray occurs as a result of an electrically initiated fire."Although the above discussion focuses on MNS, the same design philosophy was used on theONS design. Oconee was built to Duke Power Company fleet design specifications that werealso employed for the design and construction of the newer McGuire and Catawba nuclearpower stations; therefore, consideration of the NRC's review of the design of these other Dukeplants can inform a review of the Oconee design.In summary, the testing, performed on interlocked steel armored cable, demonstrates that cablearmoring provides excellent prevention of cable to cable faults or failures. While the ONSbronze tape shield/armor cable design configuration was outside of the scope of the 1977testing, the results of crush testing completed on the ONS 13.8kV cable design provided in theDuke Energy letter dated May 11, 2015, demonstrated that the bronze tape design of thesubject 13.8kV power cables provides equivalent mechanical protection as that provided by thearmored cable design that is called for in the Duke Power Company design standards usedduring construction.Previous NRC Inspection of ONS Specifically Reviewed the Cable Configuration Now AtIssue:ONS NRC Integrated Inspection Report 50-269/01 -05, 50-270/01-05 and 50-287/01-05 datedApril 29, 2002, documents an inspection of the then new Trench 3 cable installation. TheInspection Report notes the following:"1R17 Permanent Plant Modificationsa. Inspection ScopeThe inspectors evaluated NSM 0N-53065 (Replace Underground Power, AuxiliaryPower, & Control Cables from Keowee Hydro to Oconee Nuclear Station) to verifythat the emergency power system design basis, licensing basis, and performancecapability was not degraded due to the modification; and that the modification didnot leave the plant in an unsafe condition.The inspectors walked down the new trench and cables on several occasionsduring installation to verify that: (1) there was no effect on existing undergroundpower path during installation; (2) the new cables were protected from the effectsof external events, including tornados and water intrusion into the trench; (3) theampere rating of the new cables met design requirements of the modification; and(4) there were no unintended interactions.

Attachment i to ONS-2015-094Page 4 of 30The inspectors observed post-modification testing to verify that no cable damagewas done during installation or termination and that proper voltage and phaserotation was available after installation (i.e., the cables were not crossed).""b. FindinqsNo findings of significance were identified."Based on the above-referenced inspection, Duke Energy reasonably relied on the acceptabilityof the current design and its consistency with the ONS licensing and design basis.Compliance with the Oconee Licensing BasisArmor as SeparationWhile physical separation (distance) is considered a reliable method of providing circuit separationand isolation, cable armoring has historically been an integral component of Oconee's designstrategy. The Oconee licensing documentation has consistently credited the combination of arobust cabling system and defense-in-depth approach to power source availability [redundantpower sources] to meet the requirements of the Oconee license. In a December 22, 1970 letter tothe Atomic Energy Commission (AEC), Duke reiterated that "it is our intent wherever physicallypossible to utilize metallically armored and protected cable systems." This statement has existed inthe ONS FSAR since Revision 4 and was present at initial licensing. NRC awareness of the useand benefit of cable armoring is evidenced by statements within Duke/NRC correspondence (e.g.,"the cable armors used provide excellent mechanical and fire protection which would not beprovided with conventional, unarmored cable systems," "[t]he applicant is taking credit for armor oncables as barrier," and the "Safety significance of running two cables in the same tray wasmitigated by a unique design feature at Oconee of installing cable in armored jackets"). Context forthese examples is provided in the licensing excerpts cited below.Cable installation and routing requirements are documented in the ONS Updated Final SafetyAnalysis Report (UFSAR), OSS-0218.00-00-0019, "Cable and Wiring Separation Criteria," andDesign Criteria (DC) 3.13 "Oconee Nuclear Station Cable and Wiring Separation Criteria." Theserequirements are based on the use of armored cable. Most recently, DC 3.13 was provided as thebasis for cable separation in Duke Energy submittals associated with NFPA 805, Protected ServiceWater (P3W), and Tornado/HELB submittals. The installation of the cabling within the trenchadheres to these specifications that reflect the UFSAR requirements. Duke Energy's position hasbeen and continues to be that the current design configuration meets the design and licensingrequirements for cable routing and cable separation at Oconee and will not create a condition thatimpacts the ability of the emergency power system to perform its intended function.Routing of Control and Power CablesAn abbreviated chronology of correspondence related to the role of cable armor at ONS, whichhighlights NRC awareness of the subject configuration since initial licensing as it relates toseparation, and supplemental information on Oconee's single failure criteria are provided below.Additionally, key interactions that shape the Oconee license, including single failure discussions,are included in the timeline below. The information is generally presented in chronological orderfollowed by a summary of industry guidance. PSAPIFSAR Chapters 7 and 8 are highlighted tocompare and contrast how the single failure requirements and IEEE-279 discussions in these twosections are presented in the licensing history. to ONS-2015-094Page 5 of 30December 31, 1967 Preliminary Safety Analysis ReportPSAR Section 7.1.1.2.3:"Redundant protective channels and their associated elements shall be electrically independentand packaged to provide physical separation."PSAR Section 8.2.2.9, "Evaluation of the Physical Layout, Electrical Distribution SystemEquipment," part (e) states:"The application and routing of control, instrumentation and power cables will be such as tominimize their vulnerability to damage from any source. All cables will be applied usingconservative margins with respect to their current carrying capacities, insulation properties andmechanical construction. Cable insulations in the Reactor Building will be selected so as tominimize the harmful effects of radiation, heat and humidity. Appropriate instrumentationcables will be shielded to minimize induced voltage and magnetic interference. Wire andcables related to engineered safeguards and reactor protective systems will be routed andinstalled in such manner as to maintain the integrity of their respective redundant channels andprotect them from physical damage."On February 13, 1970, the Atomic Energy Commission's (AEC's) Division of Reactor Licensing(DRL) sent a letter to Duke requesting additional information on numerous sections of the ESAR aspart of their ongoing operating license review. Part 7.3 of this request questioned the level of detailin the ESAR on the installation of the reactor protection systems and is copied in part below."Submit your cable installation design criteria for independence of redundant RPS and ESEcircuits (instrumentation, control and power). (The protection system circuits should beinterpreted to include all sensors, instrument cables, control cables, power cables, and theactuated devices, e.g., breakers, valves, pumps.) Include the following:(a) Separation of power cables from control and instrument cables. (Describe anyintermixing within a tray --conduit, ladder, etc.-- of control and instrument cables, ofdifferent protection channel cables, or of nonprotection cables with protection cables.)(b) State how your design accomplishes separation of electrical penetration assemblieswithin the penetration rooms into areas, grouping of these assemblies in each area, andthe separation of assemblies with mutually redundant circuits.(c) Describe cable tray loading, insulation, derating, and overload protection for the variouscategories of cables.(d) Describe your design with respect to fire stops, protection of cables in hostileenvironments, temperature monitoring of cables, fire detection, and cable and wirewaymarkings."On April 20, 1970, Duke responded to the AEC's request. The Duke letter included ESARSupplement I (part of FSAR Revision No. 4) which detailed a listing of ESAR sections and thecorresponding questions they answered. Question 7.3, answered by FSAR Rev 4, changesincluded the following:FSAR Section 7.1.2.3.5 stated"Located under the control rooms between the outside of the reactor buildings and the cableand equipment rooms, four separate trays are provided per unit which carry nothing butnuclear, RPS and ES instrumentation cables. Three separate routes are followed by thesetrays... Ito ONS-2015-094Page 6 of 30..Equipment locations in the auxiliary building provide the basis for vertical arrangement oftrays following the same route from the reactor buildings. Switchgear for power equipment islocated at lower elevations and instrumentation cabinets are located at higher elevations.Therefore, vertical separation of classes of cables in trays is as follows from top trays down:a. Instrumentation cable traysb. Control cable traysc. Power and control cable traysd. Power cable traysInside the cable rooms cables from each protective channel are routed in trays separate fromthose carrying cables from any other protective channel. Included in these trays areinstrumentation cables from the reactor building, control and interconnecting cables associatedwith that protective channel, and non-protective instrumentation and control cables. Bothprotective and non-protective cables are individually armored and are flame retardant."ESAR Section 8.2.2.12 (h) "Evaluation of the Physical Layout, Electrical Distribution SystemEquipment" changes included the addition of the statement:"It is our intent wherever physically possible to utilize metallically armored and protected cablesystems. By this we mean the use of rigid and thin wall metal conduit, aluminum sheath cables,bronze armored control cables, steel interlocked armor power and control cables and eitherinterlocked armor or served wire armored instrumentation cables.Power cable trays are loaded with a single layer of cable. Each cable is clamped in place with aspacing being maintained between cables equal to 1/4 the diameter of the cable. Power cablesare derated based on IPCEA recommendations for Interlocked Armor Power Cables wheninstalled with one quarter cable diameter spacing in cable trays. The maximum fill in controland instrumentation, cable trays is such that trays will be filled to the top of the tray rails."The December 22, 1970, Duke response to item #1 regarding cable protection of the AEC letterdated November 25, 1970, is copied in part below:"1) Physical Protection Provided Safety-related Cables.a) We agree that Sections 7.1, 7.3, and 8.2 require that safety-related cables be routedand protected from physical damage.In our opinion, the installations for safety-related cables are adequately protected anddo conform to the designs described in the "Final Safety Analysis Report" and withAppendix B to 10 CFR 50, as well as those described in meetings with DRL.b) The cable system designs are described generally in Section 8.2 of the ESAR andstates that: 'It is our intent wherever physically possible to utilize metallically armoredand protected cable systems. By this we mean the use of rigid and thin wall metalconduit, aluminum sheath cables, bronze armored control cables, steel interlockedarmor power and control cables and either interlocked armored or served wire armoredinstrumentation cables. With this type of construction fire stops as such are notrequired.' Other references in Section 7 and Section 8 also state the requirements forrouting in cable trays to achieve separation and isolation of redundant circuits. Ito ONS-2015-094Page 7 of 30The primary method which has been used to achieve physical protection from damageis the metallically armored and protected cable systems. With these systems, either thecables are armored or they are installed in metal conduits. Armoring added to cables invarious forms provides additional mechanical physical protection in much the samemanner as does flexible conduit. Essentially, each cable with its armor has its own'built-in' conduit...Without the armors, the cables would be suitable for cable tray installations as hasbeen installed in existing nuclear power plants. However, with the armors, the cablesuse cable trays and other adequate means for support; i.e., cables strapped to beams,etc. The cables are suitable for running outside of the cable trays since each cable hasits own physical protection built in."The response is lengthy and discusses the various types of armor employed and cable installationrequirements.Subsequently, on December 29, 1970, the AEC issued the initial Safety Evaluation for OconeeNuclear Station Unit 1, licensing Oconee as described in the above correspondences. Included inthe Safety Evaluation was the following:"Section 8.5 Cable and Equipment Separation and Fire PreventionWe have reviewed the applicant's design provisions and installation arrangement plans relating(1) to the preservation of the independence of redundant safety equipment by means ofidentification and separation, and (2) to the prevention of fires through derating of power cablesand proper tray loading. We have found these design provisions and installation arrangementsto be acceptable."Following Oconee Unit I initial licensing, construction inspections continued on Units 2 and 3. AnAEC letter dated January 26, 1972, documents "a problem regarding adequate separation ofredundant instrumentation and control cables" that was observed during a site visit onNovember 30, 1971. The letter states that, during a meeting between AEC and Duke personnel,the following actions were discussed with the intent to resolve the issue:"1. Install 'Glastic' fire resistant barriers to the bottom of Oconee Unit 1 cable trays in all areaswhere the minimum spacing between the cables in the bottom of one tray is less than threeinches from the cables in the top of the tray immediately below it.2. Institute a cable temperature checking program in the critical areas of cable tray overfill inOconee Unit 1. This program will be carried out for a reasonable but limited period of timeand will include temperature checks during initial startup, normal and adverse operatingconditions.3. Revise the FSAR to incorporate the above Unit I cable tray modifications and the cabletemperature checking program and to show that for Oconee Units 2 and 3 the original cableseparation criteria will be met including no cable tray overloading and a minimum of fiveinches rail-to-rail space between all vertical trays."[Note: The identified issue is with inadequate spacing from the bottom of one tray to thetop of the tray below (see item 1 ); therefore, the five inch reference is the minimum verticalspacing between horizontal trays.]The January 26, 1972, letter concluded:"Based on our review of this matter, we conclude that your proposal as noted above isacceptable." Ito ONS-2015-094Page 8 of 30Duke staff responded by incorporating the requirements into FSAR revision 18, Section 3.2.2.13,"Evaluation of the Physical Layout, Electrical Distribution System Equipment," datedMarch 10, 1972."The maximum fill in control and instrumentation cable trays is such that trays will be filled tothe top of the tray rails except in some Unit 1 locations. Units 2 and 3 cable trays will not befilled above the side rails. A minimum of five inches rail to rail separation will be maintainedbetween all vertical trays on Units 2 and 3.....Armored cable was used at Oconee to achieve better mechanical protection and fireretardance. This caused the trays to fill faster than anticipated and in several locations the fillbecame excessive. Steps have also been taken to insure that no additional cables are routedthrough trays which are already overfilled.Where overfill situations exist in Unit 1 between vertically adjacent cable trays to the extent thatthe top cable in the lower tray is within three (3) inches of the bottom cable in the trayimmediately above, a 1/8" fire retardant fiberglass reinforced polyester barrier will be placedbetween the trays. These barriers will be attached to the bottom of the upper tray and fittedaround cables which may pass through the barrier."[Note: Again, minimum tray spacing is with respect to the vertical spacing between horizontaltrays as supported by references to tray above overfilled tray. Also see note with item 3above.]On July 6, 1973, the AEC issued the initial Safety Evaluation for Oconee Nuclear Station Units 2and 3. Section 8.0, Instrumentation, Control and Power System, of the evaluation opened with thestatement:"The staff review of the Oconee instrumentation, control, and power system on pages49-54 [Chapter 8] of the Oconee Unit 1 SER is applicable to Units 2 and 3. Additionalmatters related to the Units 2 and 3 review are discussed below."Section 8.5.1, Cable Separation, contained the following additional information:"The applicant has supplemented his cable installation criteria. ,Cable trays in Units 2 and 3will not be filled above the tray side rails; additional cable trays were installed to assurecompliance with this commitment. The staff concluded that the provisions for separation ofcables are acceptable."To summarize, by July 1973, all three Oconee Units had received Operating Licenses. Cable trayswere identified as being utilized for separation and isolation of redundant circuits. Additionally,armored cable was introduced as providing physical protection, similar to 'built-in' conduit.Oconee's cable routing and separation practices have been documented, reviewed, and foundacceptable during issuance of the Oconee licenses. Also, oversight in the area of cable routingand installation was evidenced by inspection issues with adequate cable separation which wereaddressed by the installation of "glastic" where required and future cable tray fill requirements.Single Failure CriteriaSubsequent to Oconee licensing, in January 1974, Appendix K to Part 50-ECCS EvaluationModels, was issued to define the Required and Acceptable Features of Evaluation Models. Alllicensees who were not under this rule were required to perform the evaluation. OnAugust 5, 1974, Duke submitted an evaluation of ECCS cooling performance calculated inaccordance with an evaluation model developed by the Babcock and Wilcox Company. Ito ONS-2015-094Page 9 of 30However, the regulatory staff concluded that B&W's evaluation model was not in completeconformity with the requirements of Appendix K. The December 27, 1974, Safety EvaluationReport required that Duke submit a re-evaluation of ECCS cooling performance calculated inaccordance with an acceptable evaluation model. It is within the on-going discussion of the re-evaluation of ECCS cooling performance that the discussion of single failure arises. In theMarch 10, 1976, letter to the NRC, Duke stated:"In Section 4.6 of BAW 10103 and Section 3.5 of the specific Oconee Unit 1 analysis, theworst single failure postulated is the loss of a diesel, following loss of off-site power, whichresults in the operation of only one LPI and one HPI pump. The Oconee emergency powersystem uses two hydro-electric generating units instead of diesels and a single failure ofone of these sources will have no effect upon ECCS performance. However, failure of a4160 volt switchboard could cause the loss of one HPI and one LPI pump, but there is nopossibility of a common mode failure which will result in the loss of more than one 4160 voltswitchboard. Therefore, although the failure mechanism for the Oconee units is differentfrom that described in BAW 10103 and the specific Oconee I analysis, the worst casesingle failure still results in the operation of only one HPI and one LPI pump, and theconclusions of the analyses remain valid."During this same time period, Duke and B&W were communicating as to how IEEE-279 should beapplied for single failure. In response to an April 6, 1976 RAI, B&W provided the followingexplanation of IEEE-279 to Duke:"IEEE-279 provides general, rather than specific, criteria for single failure analysis, so thatthe real acceptance standard is NRC approval of the analysis."The B&W response provides insight into industry perspective of how such criteria were historicallyinterpreted using licensing correspondence. Information in support of Duke's application of thesingle failure criteria can be found in a May 13, 1976, response to the April 6, 1976, RAI regardingthe ECCS analysis for Oconee. Duke was asked to describe the design of the ECCS actuationsystem and identify any non-conformance of this design with the single failure requirements ofIEEE Std 279-1971. Duke replied:"The design of the Oconee Nuclear Station Engineered Safeguards Protective System(ESPS) is described in FSAR Section 7.1.3. The ESPS includes the Emergency CoreCooling System in addition to the Reactor Building Isolation, Spray and Cooling Systems.The system logic for the ESPS is described in FSAR Section 7.1.3.2.1, and the specificdiscussion for the ECCS components is provided in ESAR Section 7.1.3.2.2. The safetyevaluation for the ESPS is provided in ESAR Section 7.1.3.3.The Oconee ECCS actuation system conforms to the single failure requirements of IEEE279-1971 ."Similarly, Duke was requested to describe the design of the onsite emergency power system, a-cand d-c and to identify any non-conformance of this design with the single failure requirements ofIEEE Std 279-1971. Duke replied:"The Oconee Nuclear Station onsite emergency AC power sources and distribution systemare described in FSAR Section 8.2.3. The emergency power distribution through theswitchboards is described in FSAR Sections 8.2.2.4, 8.2.2.5, and 8.2.2.6. The onsiteemergency DC power system is described in FSAR Section 8.2.2.7. A single failureanalysis of these systems is provided in Table 8.7. Ito ONS-2015-094Page 10 of 30The design of the Oconee onsite emergency AC and DC power systems conforms to thesingle failure requirements of IEEE 279-1971 ."Duke provided effectively the same response with respect to IEEE-279 Single Failure requirementsto both questions, with a high level statement regarding conformance to the generic criteria andspecific references to the ESAR defined system descriptions and bases in the responses. Theseresponses are consistent with the perspective that the IEEE-279 criteria as a general descriptionwith the details as listed in the ESAR references.In 1976, the Oconee licenses were amended based upon an acceptable Emergency Core CoolingSystem evaluation model conforming to the requirements of 10 CFR Section 50.46 and theoperating restrictions imposed by the Commission's December 27, 1974, Order for Modification ofLicense was terminated. The ECCS Reanalysis was accepted in three separate SafetyEvaluations. Within the Safety Evaluation for the amendments (with Unit 2 quoted), the followingexcerpts are highlighted:* "We reviewed the design of this system on the following basis: The design of the entireemergency electric power system, including generating sources, distribution system andcontrols, is such that a single failure of any single electric component will not preclude theEmergency Core Cooling System of either Units 2 or 3 from performing its function."* "We requested that the licensee determine if any single failure could compromiseredundant trains. The licensee provided a control circuit schematic typical of that whichwould be used for all safeguards equipment actuated by redundant trains. Since two relayfailures in redundant safeguards cabinets would be required to compromise redundanttrains, this design provides adequate isolation between trains. This configuration is similarto that used in other nuclear power plants whose designs have been found acceptable.Therefore, this portion of the actuation system is in conformance with the fundamentalsingle failure criterion at the electric component level."* "To preclude the likelihood of an undetected failure, Technical Specifications will berequired to include a monthly surveillance of this interlock [Keowee Underground Breaker].By including periodic testing of this interlock, we are satisfied that the same level of safetyhas been achieved for this interlock as exists for all other safeguards equipment that aretested monthly."* "Single Failure Conclusion -On the basis of our review, including the above indicatedchanges to Technical Specifications and commitments by the licensee, we find that there issufficient assurance that the ECCS will remain functional after the worst damaging singlefailure of ECOS equipment at the component level has occurred."The wording within this Safety Evaluation supplements the IEEE-279 single failure criteria with anunderstanding that, at the component level, the requirements are met by means of redundant andisolated trains. As designed and analyzed, Oconee met the single failure requirements and wouldsurvive the plant designed single failure.McGuire -Cable Separation Licensing BasisAn August 30, 1977, NRC-authored McGuire Site Visit Summary associated with a constructionrelated inspection clarified Duke's position with regard to cable armoring:"During our site visit, the staff has verified that routing of redundant safety related cablesconformed with the criteria described in the ESAR. The applicant is taking credit for armor oncables as barrier. The staff has requested the applicant to document the test results andconclusions reached with regard to the barrier integrity of the armored cables used in the plant.The evaluation of cable separation and identification will remain open till the fire protectionreview for this plant is completed." to ONS-2015-094Page 11 of 30Duke Power performed testing on armored cable with the intent to demonstrate the acceptability ofinterlocked armor cable as an adequate barrier to internally generated faults for both power andcontrol cables. The initial test report was issued November 8, 1977, and a follow up report wasissued March 3, 1982, all as MOM 1354.00-0029.001. The results have been referenced invarious correspondences between Duke and the NRC. The test report was provided to the NRCfor review as part of McGuire's Fire Protection licensing efforts. The March 28, 1978, RAIresponse stated:"Testing performed by Duke Power Company at the Westinghouse High Power Laboratory inEast Pittsburg, Pennsylvania demonstrated that adjacent armored cables within the same traywill not be damaged due to the short circuiting of the power cable. These test reports havebeen submitted to the NRC and more than adequately demonstrate that these types of powercables pose no threat to the redundant safety trains."Subsequent RAls and responses demonstrate that NRC personnel scrutinized the test method andresults. In a January 31, 1979, Fire Protection Review, McGuire stated:"The use of armor on cables ensures they are more resistant to mechanical damage andelectrostatic and electromagnetic interferences. The armor also provides protection from shortcircuits and overloads."On March 1, 1979, the NRC issued Supplement 2 to the McGuire ESAR, indicating acceptance ofthe Duke position on armored cable as supported by testing, stating"At the time the Safety Evaluation Report was issued we had not completed our review of thefire protection program. This review has now been completed. We find the applicant's fireprotection program to be acceptable and this item is resolved."Although the above references are in regard to McGuire, the design philosophy advocating the useof armored cable, licensed at McGuire, was applied to the Duke fleet. The cable test results havebeen referenced in correspondence between Oconee and the NRC.Note: While the bronze tape shield/armor cable design at Oconee was outside of the scope of the1977 testing, the robust design of the medium voltage power cables provides a commensuratelevel of protection from cable to cable faults or failures based on the current analysis and testing.ONS SSF LicensingDuring the fire protection review for McGuire, the Oconee fire protection review was also ongoing.On August 8, 1978, the NRC issued Oconee Amendment Nos. 64, 64, and 61 to add licenseconditions relating to the completion of facility modifications for fire protection. Within the SafetyEvaluation, the NRC recognized that the majority of the cables used in construction at Oconee areof the metallic armored type, reiterated the cable separation requirements as stated in the ESAR,and repeated the Duke position that the cable armors used provide excellent mechanical and fireprotection which would not be provided with conventional, unarmored cable systems. Within theSE, the construction cable seParation practices were specifically called out and the commitment toinstall a separate and independent facility to shut down the units (i.e., Standby Shutdown Facility)was documented."Throughout most of the plant there is good separation between redundant divisions such thata fire would not cause loss of redundant safe shutdown equipment. During the site visit, it wasdetermined that in a number of locations, redundant safe shutdown cables could bejeopardized due to a lack of sufficient separation. The separation criteria used in the to ONS-2015-094Page 12 of 30installation did not preclude the routing of redundant cables vertically over one another inadjacent trays."In 1981, the design and licensing requirements for the SSF were being established. During thisreview, a question was raised as to whether it was considered credible for a 4 or 7 kV voltagesource to be applied to one of the associated circuits as a result of a fire and subsequentlypropagate to and damage components of the shutdown train which was otherwise unaffected bythe postulated fire. Internally, Duke considered the cable routing criteria, robust cable design, andgrounded cable armor and determined that this type of event could not credibly result from a fire.In a letter from Duke Power to the NRC dated March 18, 1981, concerning licensing of Oconee'sStandby Shutdown Facility, Duke continued to credit the armor when discussion requiredseparation of associated circuits. The letter stated:"The cable used by Duke Power is of the armored type. We have performed tests thatdemonstrate the armor provides adequate protection to prevent a fault within a cable frompropagating into an adjacent cable, even if the breaker feeding the faulted cable fails to trip."Inspection Report 50-269, 270, 287/89-05, dated March 14, 1989, noted a potential probleminvolving safety related cabling. Two sets of redundant Main Feeder Bus control cables for lockoutrelaying were inappropriately routed through the same cable trays, in violation of FSAR Section8.3.1.4.6.2. The inspection report further documents the results of the operability analysis, with thefirst two items being "All cables concerned are grounded armored cables. ..Internal faultpropagation from one cable to another in a common tray will not occur. This is supported bytesting performed by DPC [Duke Power Company] and documented in Test Report MCM 1334.00-0029(sic).. ." The report concluded "No violations or deviations were identified." Oconeesubmitted LER 269/89-04 dated March 29, 1989, on the issue, characterizing it as a DesignDeficiency, Electrical Equipment Configuration Deficiency. The subsequent corrective actions wereto reroute the control cables through different cable trays and to perform a random inspection ofselected safety related cables throughout the plant.Electrical Distribution System Functional InspectionDuring the first quarter of 1993, the NRC conducted an approximately six week inspection toassess the capability of the Oconee Electrical Distribution System (EDS), including Keowee, toperform its intended functions during all plant operating and accident conditions. The teamreviewed the Oconee EDS design with respect to regulatory requirements, licensing commitmentsand pertinent industry standards. The review included the examination of EDS equipment size andrating, EDS as-built configuration, EDS material condition, maintenance, testing and calibrationprogram for EDS components, root cause analysis of EDS deviation reports, and the adequacy ofthe EDS design documentation.Three items of relevance were noted in NRC Inspection Report No. 50-269/93-02, 50-270/93-02,and 50-287/93-02, "Notice of Violation and Notice of Deviation," dated May 7, 1993. During theinspection a deviation from the requirements of FSAR, Section 8.3.1.4.6.2, "Cable Separation,"was identified. The power cables for the mutually redundant emergency core cooling recirculationsump isolation valves 2LP-19 and 2LP-20 were run in the same cable trays. However, the benefitof armored cable with respect to cable separation requirements was documented in the report withthe authors stating:"The safety significance of running the two cables in the same tray was mitigated by a uniquedesign feature at Oconee of installing cables in armored jackets."Another indication that the inspection team reviewed the Oconee design philosophy of utilizingarmored cable is provided by the observations quoted below from the same inspection report:

Attachment i to ONS-2015-094Page 13 of 30"Rigorous separation between the control and electrical cables of the units was not employed.In many cases, cables for the two units occupy the same trays. However, armored cables areused with voltage and current carrying ratings in excess of that required.""The team was concerned about cable size #2 that was used for most of the safety motors. If aground fault occurred near the feeding end of the #2 cable, the system short circuit currentcould exceed 45 kA. The temperature rise in the #2 size cable would well exceed 2500 limit.The team requested the licensee to verify that such a fault at the #2 size cable would not causea generated fire or source of propagation to its adjacent cables, which might belong to anothersafety load group. At the end of the inspection, the licensee provided a copy of the fault testreport, which was done for the McGuire Nuclear Station. The team reviewed this report anddetermined that the cables size was adequate."Keowee Underground Breaker ModificationIn this same time period, Oconee submitted a license amendment request associated with amodification to the Keowee underground power path breakers. RAIs associated with the changeclarified the Oconee single failure licensing basis. The RAl noted that Oconee's position was that"'smart failures' within the control system are not considered in failure analyses." In response, tothis and subsequent RAls, the failure-on-demand position was communicated.May 25, 1994 RAl: "In general, it has been the Oconee position to not consider smartfailures within control systems. The system is assumed to control as designed or to fail toits as-designed state."NRC in January 26, 1995 letter: "The design basis for the Keowee emergency powersystem was then presented as the ability to provide emergency power to the OconeeNuclear Station within the committed time under all applicable conditions assuming a singlefailure. After some discussion, the licensee stated the position that the switchyard yellowbus is part of the onsite power system at all times. Thus, a component failure which couldcause a loss of this bus would be considered the single failure of the onsite electricalsystem. In addition, the licensee stated that any single failure was assumed to occursimultaneously with the initiating event."March 8, 1995 response: "The second statement is that any single failure was assumed tooccur simultaneously with the initiating event. This should be changed to indicate that anysingle failure was assumed to occur immediately upon demand. Therefore, a failure wasassumed to occur when the equipment was called upon to perform its safety function. Thisfailure could be simultaneous to the initiating event or at some time during the mitigationevent."NRC acceptance that the modified system addressed the NRC's single failure concerns wasevidenced in the August 15, 1995 SER which stated:"It finds that the new circuitry, intended as part of the long-term corrective action foroverfrequency and overspeed concerns, does indeed eliminate those concerns, introducesno new single-failure vulnerabilities, and is acceptable conditioned on the proposal oftechnical specifications as discussed above."ONS Emergency Electrical Power System ReviewIn August 1995, the Office of Nuclear Reactor Regulation (NRR) undertook a formal review of theOconee Nuclear Station, Units 1, 2, and 3 (Oconee), Emergency Electrical Distribution System(EDS). The purpose of the review was to assess the overall reliability of the Oconee emergency Ito ONS-2015-094Page 14 of 30power system (EPS) and determine whether any additional staff actions might be required toaddress vulnerabilities or risks that may exist in the design or operation of the system and thestandby shutdown facility. In the Final Report issued in 1999, the staff did not identify anyvulnerabilities in the design or operation of the Oconee Emergency Power System or standbyshutdown facility that would require immediate corrective actions to be taken. Also, within thereport, the NRC recognized that the Oconee single failure criteria is plant-specific, with the originalfocus being on mechanical systems:"With regard to single failure, Oconee uses a plant-specific definition. The original staffreview of the ECCS for compliance with 10 CFR 50.46 and 10 CFR 50, Appendix Kchecked for single failure vulnerabilities in the piping system. It ultimately concluded that theplant should be analyzed assuming the limiting single failure was the same as the genericB&W analysis that assumed the loss of an emergency bus (actually a DG). The emphasisof the ECCS rulemaking, and the individual reviews at the time, was on the thermalhydraulic and physical treatment of LOCA analysis models in addition to the acceptancecriteria, not on the electrical design or the basis of the single failure criterion. Plantslicensed after 10 CFR 50, Appendix A, were required to meet requirements, of GDC 17 forelectrical systems (onsite and offsite) that encompass any single failure requirements onelectrical systems from 10 CFR 50, Appendix K or GDC 35. Because Oconee was notlicensed to 10 CER 50, Appendix A, the plant-specific single failure definition for Oconeeremains valid and in effect with no additional requirements on the electrical power systemsas a result of 10 CER 50.46 or 10 CFR 50, Appendix K."ONS LicensingAs part of ongoing licensing reviews (i.e., License Renewal, NFPA 805, Protected Service Water,and Tornado/High Energy Line Break), Oconee's cable system design continued to be thoroughlyreviewed.License RenewalThe Oconee units were issued Renewed Operating Licenses on May 23, 2000, prior to the Trench3 and PSW modifications. As a part of the license renewal process, aging management programswere established. Plant changes implemented subsequent to license renewal are required toconsider impacts on aging management programs. While not directly related to cable separation,the review and associated requirements indicate that the cables of concern are not subject tosignificant aging effects. UFSAR Section 18.3.14 states:"The Insulated Cables and Connections Aging Management Program includes accessible andinaccessible insulated cables within the scope of license renewal that are installed in adverse,localized environments.. .which could be subject to applicable aging effects from heat, radiationor moisture... Inaccessible or direct-buried, medium-voltage cables exposed to significantmoisture and significant voltage will be tested.. .Significant moisture exposure is defined asperiodic exposures to moisture that last more than a few days (e.g., cable in standing water).Periodic exposures to moisture that last less than a few days (i.e., normal rain and drain) arenot significant. Significant voltage exposure is defined as being subjected to system voltage formore than twenty-five percent of the time."These definitions are echoed in NUREG 1723, "Safety Evaluation Report Related to the LicenseRenewal of Oconee Nuclear Station, Units 1, 2, and 3," Section 3.9.3.2.1, "Aging ManagementProgram for Insulated Cables and Connections." Based upon these criteria, the 13.8 kV feedercables from Keowee to CT-4 and the 4.16 kV cables from Oconee to Keowee transformer CX wereevaluated and screened out from periodic testing due to lack of aging stressors from heat,radiation, or significant moisture exposure.

Attachment i to ONS-2015-094Page 15 of 30TornadolHigh Energy Line Break Licensing EffortsThe next major licensing effort under NRC review, with relevance to this cable issue, wasassociated with Tornado and High Energy Line Break (HELB). Numerous references wereidentified where the design and installation of cables was questioned. Duke repeatedly referencedDC 3.13 in response. Bronze tape was first introduced in licensing space as a part of thisexchange.In a September 2, 2009, response to a Request for Additional Information regarding LAR 2006-09,Tornado Licensing Basis, Duke provided the following responses:"RAI -3: In Enclosure 2, Section 3.3.4 of the LAR, the licensee states that the KeoweeHydroelectric Units (KHUs) will provide power supply to the PSW switchgear throughunderground cables. Provide analyses to show the kilo volt ampere (kVA) loading, new circuitbreaker rating, short circuit values, and voltage drop. In addition, provide information on theelectrical protection and coordination, and the periodic inspection and testing requirements.Further, explain how the redundancy and independence of the Class 1 E power system ismaintained as a result of the proposed modification. Provide applicable schematic and singleline diagrams.Duke Response:..4. The PSW electrical system has a normal (13.8kV Fant Line) and alternate (13.8kVKeowee Hydro Unit) 13.8kV power source breaker to each PSW Unit Substation. Theredundancy of the PSW power system is provided through these two power sources. TheKeowee Hydro Unit 13.8kV power source cables to the PSW electrical system will be routed ina combination of precast concrete trench boxes, duct banks, and manholes. The power feedsfrom Keowee Hydro Unit to both Oconee Nuclear Station and the PSW electrical equipmentare isolated by separate breakers and disconnect switches. Independence is maintained asrequired by Duke Design Criteria DC 3.13.RAI -5: Provide information on how the licensing basis for physical independence andseparation criteria are met for the PSW electrical system.Duke Response:The licensing basis for physical independence and separation criteria for the PSW electricalsystem meet the requirements of Duke Design Criteria DC 3.13, "Oconee Nuclear StationCable and Wiring Separation Criteria," that provides guidance for cable routing and installation.Refer to DC 3.13 [Enclosure]."[NOTE: Section 6.4 of DC 3.13 (provided in RAI response) states "Trenches maysimultaneously contain power, control and instrumentation cables. If cable tray, electray, orconduit is not provided for cable support, power cables within a trench shall be racked on theside of the trench or on unistrut cross members per Section 6.2 of this criteria. Control andinstrumentation cables shall be laid in the bottom of the trench. Mutually redundant safetycables shall be located on opposite sides of the trench."]In a June 24, 2010, response to additional Tornado RAIs, the following was provided to the NRC:"RAl 2-31: Provide information on how the licensing basis for physical independence andseparation criteria are met for the PSW electrical system.

Attachment i to ONS-2015-094Page 16 of 30Duke Energy Response [again referencing DC 3.13]"..Implementation of the physical independence and separation criteria for the PSW electricalsystem will be controlled by Duke Energy Design Criteria DC 3.13, "Oconee Station Cable andWiring Separation Criteria." DC 3.13 provides guidance for cable routing and installation whichhas been revised to include PSW-related cables."In an August 31, 2010, response to additional Tornado RAIs Duke addressed trench and cabledesign and construction, including the use of bronze tape, as well as long term monitoringpractices:"RAI 2-32: The licensee states in the June 26, 2008, LAR that the new PSW systemswitchgear will receive power from the KHUs via a tornado-protected underground feeder path.Provide the following Information:1) the type of underground cable installation, i.e., direct burial or in duct banks, manholes etc.,2) how the licensee will ensure that the proposed new underground cables remain in anenvironment that they are qualified for, 3) periodic inspections and testing planned for cables tomonitor their performance, and 4) details regarding cable size, type, maximum loadingrequirements, and cable protection devices.Duke Energy Response1. The underground cable route from Keowee Hydro to the PSW building will be acombination of precast concrete trench boxes, duct banks and manholes. This new routewill be an extension of the existing underground path from Keowee to the CT-4 block houseat the plant. Spare cables in the existing underground path will be spliced to new cables inthe underground path extension to the PSW building. None of the cables will be directburied.2. The Keowee underground path to the PSW switchgear will be designed to preclude waterentry that could wet the cables. The concrete trenches will have drains. The new duct bankconduits will be sloped towards manholes where drains are provided. Periodic inspectionswill be performed on the Keowee to PSW underground path to evaluate the condition of thetrenches, duct banks, manholes and drainage system.3. The cables will be evaluated for inclusion in the ONS Insulated Cables and ConnectionsAging Management Program. Since the underground path from Keowee to the PSWbuilding is designed to prevent significant exposure to moisture and most of the path isinaccessible, it is expected that the cables won't meet the criteria for periodic diagnostictesting. If subsequent periodic inspection of the Keowee to PSW building underground pathdetermines that these inaccessible cables are exposed to significant moisture, testing willbe in accordance with the ONS Cable Aging Management Program.4. The two (2) 13.8 kV circuits from Keowee to the PSW building consist of six singleconductor cables. Each conductor is 750 kcmil copper with Class B compact roundstranding. The conductor shield is a thermoset semi-conducting compound extruded overthe conductor. The insulation is ethylene propylene rubber (EPR) that provides aninsulation level of 173% above the 15 kV nominal insulation rating. The insulation is rated at90°C continuous and 130°C emergency overload. The insulation shield is a semi-conducting thermoset compound applied over the insulation. Two layers of non-magneticbronze tape shield are applied over the insulation shield. A thermoset chlorosulfonatedpolyethylene (Hypalon) jacket is applied over the cable core. Ito ONS-2015-094Page 17 of 30Maximum allowable cable loading will not exceed the continuous conductor insulationtemperature rating. Depending on the cable loading scenario, anticipated cable loading isexpected to range from 193 A to 559 A.Cable protection will be provided by Keowee and PSW switchgear breakers and protectivedevices, which includes time-overcurrent, instantaneous overcurrent and ground faultrelays."In a December 7, 2010, response to additional RAIs on the Tornado/HELB licensing effort, RAI2-31 and 2-32 responses were again provided as the responses to new RAIs 47 and 48 [T/H].RAI 47 [T/H] requested that Duke Energy provide information on how the licensing basis forphysical independence and separation criteria are met for the PSW electrical system. The DukeEnergy response referred back to PAl 2-31. RAl 48 [T/H] requested information on the 1 ) Type ofunderground cable installation, i.e., direct burial or in duct banks, manholes etc.; 2) How thelicensee will ensure that the proposed new underground cables remain in an environment that theyare qualified for; 3) Periodic inspections and testing planned for cables to monitor theirperformance; and 4) Details regarding cable size, type; maximum loading requirements, and cableprotection devices. The Duke Energy response was to refer to the RAl 2-32 response ofAugust 31, 2010.NFPA 805During this same time period, Oconee was pursuing transition from Appendix R to NFPA 805license. During this effort, Duke once again highlighted that armored cable is the prevalent cableutilized at ONS, and its use is addressed in the Duke design criteria."RAl 3-27:Provide a justification for your assumption that the use of armored cables, without furtherconsideration of their current installed configuration, is adequate to prevent inter-cable faultsdue to fire or, alternatively, provide information that reasonably demonstrates that theas-installed configuration of the armor cable grounding scheme is consistent with the originalplant design.The LAIR credits armored cables for precluding the occurrence of inter-cable shorts. As a result,only the effects of conductor-to-conductor shorts (intra-cable) within multi-conductor cableswere considered. Recent (CAROLFIIRE) test results demonstrate that this assumption may notbe valid if the armored cables are not appropriately grounded. From the CAIROLFIRE Report,Volume 1, Section 7.2.5, Grounded versus Ungrounded CPTs, "Grounded versus ungroundedcircuits may be a significant factor influencing the likelihood of spurious actuation for armoredcables," and Section 9.2.3, Grounded Versus Un-grounded Power Supply. It appears likely thatthe presence of the armor itself, which is grounded in typical applications, makes it more likelythat a short to ground and fuse blow failure will occur for the grounded power supply cases. Inthe absence of the armor, the ground plane is available only through either a groundedconductor or the raceway itself. For an un-grounded circuit, a single short to ground will not tripthe circuit protection (fuse) and therefore the likelihood of spurious actuation is somewhathigher.RAl 3-27 RESPONSE:Armored cable is the prevalent cable utilized at ONS. The interlocked armor on the cables atONS are terminated and grounded as per drawings OEE-014-04 and OEE-015 series. Thesedrawing series were in effect during the plants original construction. In addition, Section 6.4.1 in to ONS-2015-094Page 18 of 30Engineering Design Criteria DC-4.1 1, Generating Station Grounding, states that "The armor ofinterlocked armor cable shall be electrically continuous and grounded to equipment enclosureat each end of the cable." A similar design document exists for the Standby Shutdown Facility(OSS-0218.00-00-0010) and Radwaste (OSS-0218.00-00-009)."The NFPA 805 Safety Evaluation dated December 29, 2010, repeated the RAI 3-27 response andconcluded:"...the NRC staff finds that the licensee has adequately addressed the issue of grounding ofarmored cable to preclude inter-cable shorts."The Protected Service Water licensing review resulted in RAI response dated December 16, 2011.This question and response are copied below:"RAI 78: Provide a detailed discussion on how the electrical power systems of the PSWsystem will be installed such that they are physically separate and independent.Duke Energy ResponseThe PSW electrical system is a single train system; however, the PSW Main pump circuits,Booster pump circuits and associated valve circuits are mutually redundant to the SSF ASWpump and valve circuits. Red PSW cables are not to be routed in the West Penetration Roomsor the Cask Decontamination Rooms to ensure the mutually redundant cables are keptphysically separate. The alternate feeder from the PSW to the SSF may be routed without anyseparation requirements from SSF cables. This will be controlled by Duke Energy DesignCriteria DC 3.13, "Oconee Nuclear Station Cable and Wiring Separation Criteria." DC 3.13provides guidance for cable routing and installation which has been revised to include PSW-related cables. DC 3.13 references IEEE Standard 603-1980, IEEE Standard Criteria for SafetySystems for Nuclear Power Generating Stations."Finally, the August 13, 2014 PSW Safety Evaluation stated:"This SE provides the technical bases for the staff's approval of the changes to the ONSlicensing basis within the scope of these amendments. These amendments and the related SEdo not approve nor endorse the as-installed PSW electrical system cable configurations. DukeEnergy analyzed the PSW electrical system configurations and installed the PSW electricalcables and power supplies under the provisions of 10 CFR 50.59, and thus those parts of thesystem were not included in the scope of the staff's review for these amendments. Theinstalled configurations of the PSW cabling and associated onsite power supply systems arethe subject of a pending NRC inspection activity, as documented as an unresolved item in theNRC Component Design Basis Inspection Report, dated June 27, 2014, Section 1.2.b.v.(ADAMS Accession No. ML14178A535)."During the NRC's review of the PSW design, Duke Energy provided details on cable design andinstallation, again citing Duke Energy Design Criteria DC 3.13 as had been done in previouslicensing reviews. The RAl response provided in 2011 was not questioned by NRC reviewers untilthe final PSW Safety Evaluation was being drafted and a question was raised in the 2014 COBI,ultimately leading to the TIA.Applicable Industry and Regulatory GuidanceIn the numerous engagements between NRC and Duke Energy there have been discussionsinvolving various design standards associated with cable separation and single failure criteria. to ONS-201 5-094Page 19 of 30Oconee Nuclear Station was designed, constructed, and licensed prior to the development of manyof the standardized requirements. Because of this, understanding of the Oconee license oftenrequires knowledge of context (i.e., other ongoing discussions between Duke and the NRC viameetings, inspections, etc.). The listing below provides dates when NRC and industry guidanceand/or requirements were issued. Key dates associated with Oconee design and licensing aresuperimposed on the list to demonstrate which requirements are applicable to Oconee and whichmay only be applicable to newer plants. For these reasons, review of Oconee's design againstcurrent NRC design and licensing requirements is not an accurate reflection of Oconee'sadherence to licensing basis requirements.Timeline* 11/22/1965 -AEC press release H-252 27 documenting proposed General Design Criteria(GDC)* 12/1 /1 966 -Oconee Nuclear Station Preliminary Safety Analysis Report (PSAR)submitted* 7/11/1967 -Proposed rule-making, AEC updated the GDC's from the original 27 to 70criterion* 111611967 -Construction Permits issued for Units 1, 2,and 3 [but the GDC are still in aproposed state]* 12/29/1970 -Initial Safety Evaluation for ONS Unit I -issued, evaluated againstproposed GDC, dated 7/11/67, and Proposed IEEE -279, dated 8/28/68.* 1971 -IEEE Std 279-1971, Criteria for Protection Systems for Nuclear Power GeneratingStations [includes the commonly applied single failure definition]* June 1973, Reg Guide 1.53, Application of the Single-Failure Criterion to Nuclear PowerPlant Protection Systems, "It is recognized that IEEE Std 379-1972 has been publishedonly for trial use and as a draft American National Standard. As experience is obtained inits use, the standard may be modified to improve its usefulness by deleting provisionswhich prove to be unacceptable or by appropriately supplementing those provisions inwhich inadequacies are found."* 7/6/1 973 -Initial Safety Evaluation for ONS Units 2 and 3* 1974 -IEEE Std 308-1974, Criteria for Class 1 E Power Systems for Nuclear PowerGenerating Stations* 1/1 975 -Regulatory Guide (RG) 1.75, Revision 1 -Physical Independence of ElectricSystems. This guide describes a method acceptable to the Regulatory staff of complyingwith IEEE Std 279-1971 and Criteria 3, 17, and 21 of Appendix Ato 10 CFR Part 50 andendorses, with certain exceptions, IEEE Std 384-1974. RG 1.75 is "to be used by theRegulatory staff in evaluating all construction permit applications for which the issue date ofthe Safety Evaluation Report is February 1, 1974, or after." [Note: Timing makes this RGnot applicable to Oconee.]* 6/30/1977 -IEEE Std 384 -1977, IEEE Standard Criteria for Independence of Class I1EEquipment and Circuits* 11/8/1977 -MCM 1354.00-0029.001, "Report of Power and Control Cable Overload andShort Circuit Tests Performed for McGuire Nuclear Station," performed by the High PowerLaboratory of the Westinghouse Corporation.* 5/23100 -Issuance of Renewed Facility Operating Licenses for ONS Units 1, 2 and 3* July 2001 -NUREG 1801, Rev 0 -Generic Aging Lessons Learned (GALL) Report(NUREG-1 801, Initial Report)* 9/2005 -NUREG/CR-6850 (EPRI/NRC-RES Fire PRA Methodology for Nuclear PowerFacilities) endorsed by Oconee's NFPA 805 Safety Evaluation, endorsed by RG 1.205* 5/2006 -RG 1.205 -Risk-Informed, Performance-Based Fire Protection for Existing Light-Water Nuclear Power Plants Attachment i to ONS-2015-094Page 20 of 3010 CFR 50.55a(h)(2), Protection Systems, requires that, for plants with construction permits issuedbefore January 1, 1971 (e.g., Oconee), protection systems must be consistent with their licensingbasis. Thus, the requirements of the subsequently issued regulatory documents (e.g., IEEE279-1971, IEEE 308-1974, RG 1.75) are not generically applicable to Oconee. For example, theOconee Emergency Power System, as presented in FSAR prior to the initial Unit 1 SafetyEvaluation (SE), described a system with diverse power sources. The design was accepted by theNRC. Instrumentation and Control cable installation requirements, as re-stated in ESAR Chapter8, were accepted by the NRC in a 1972 letter and the 1973 SE's for Units 2 and 3. Oconee's cablerouting and separation practices have been documented, reviewed, and found acceptable sincethe original licensing of Units 2 and 3. At the component level, when applied to cabling designcriteria, this has translated to UFSAR-specified separation for cables supplying redundantfunctions. Subsequent Oconee design and licensing actions have, for the most part, remainedconsistent with the philosophy/approach licensed in the 1 960s and 1970s. Components of later,applicable guidance documents have been referenced or adopted as described in the UFSAR.Within the August 11, 1978, Safety Evaluation Report, armored cable was discussed with regard toelectrical cable combustibility. However, it was also covered within the section on separation ofequipment and consequences."It should be pointed out that the cable armors used provide excellent mechanical andfire protection which would not be provided with conventional, unarmored cablesystems.""..An unmitigated fire in the containment penetration areas could cause loss ofredundant safe shutdown instrumentation though this is unlikely due to the armor on thecable."With the migration of the Oconee license to NFPA 805, the governing documents were reviewedfor applicability or insights into this issue. NUREGlCR-6850 (EPRl/NRC-RES Fire PRAMethodology for Nuclear Power Facilities), as endorsed by NRC RG 1.205. NUREG/CR-6850 wascited in Oconee's NFPA 805 Safety Evaluation.RG 1.205 states:"The NRC and the Electric Power Research Institute (EPRI) have documented a methodologyfor conducting a fire PRA in NUREG/CR-6850/EPRI 1011989, "EPRI/NRC-RES Fire PRAMethodology for Nuclear Power Facilities," issued September 2005 (Ref. 17). However,recognizing that merely using the methods explicitly documented in NUREG/CR-6850/EPRI1011989 may result in a conservative assessment of fire risk, licensees may choose to performmore detailed plant-specific analyses to provide greater realism in the fire PRA model."As an NFPA 805 licensed facility, this guidance was also incorporated. Per NFPA 805, 2001Edition:"Plant-specific design features can preclude certain circuit failures from occurring. For example,the use of grounded, metallic, armored cable or dedicated conduit, shorting switches, or rugged(e.g., braided metal) shielding are considered in most cases to preclude external hot shortsfrom further consideration. However, multiple ground faults might still energize conductorswithin a grounded conduit, shield or armor if those conductors are associated with ungroundedcircuits."NUREG-6850 contains the following statements on rugged grounded shields precluding certaincable failure mechanisms used to govern circuit analysis methodology:

Attachment i to ONS-2015-094Page 21 of 301) "Three-phase proper polarity hot shorts on AC power systems: Case 3: Armored cable orcable in dedicated conduit. Three-phase proper polarity faults are not considered crediblefor armored power cable or a single triplex cable in a dedicated conduit. The basis forexclusion is that multiconductor-to-multiconductor hot shorts are not plausible given theintervening grounded barrier (i.e. the armor or conduit)."2) "Plant specific design features can preclude certain circuit failures from occurring. Forexample, the use of grounded, metallic, armored cable or dedicated conduit, shortingswitches or rugged (e.g. braided metal) shielding are considered in most cases to precludeexternal hot shorts from further consideration."3) "If the cable design can be verified as one that employs a rugged grounded metallic shield(e.g. armor, braid, etc.), then the analysis need only consider the effects of shortingbetween the conductors within the shield and shorting the conductors to ground, i.e., theeffects of shorts from external sources need not be considered."NUREG-6850 also contains the following statements which support the consideration of a threephase bolted fault only at the terminations:1) "Cable ducts: A power conductor configuration that provides a function like a bus duct butuses a length of insulated electrical cable in lieu of metal bus bars ... Cable ducts may beused in application conditions similarly to either a segmented or non-segmented bus duct."The medium voltage power cables in Trench 3 meet this definition of a cable duct.2) "Because nonsegmented bus ducts (category 1) and cable ducts (category 3) have notransition points other than the terminations at the end device, no treatment of bus ductfaults/fires independent from the treatment of fires for the end devices is required. That is,arc faults for these two categories of bus ducts, 1 and 3, are inherently included in thetreatment of the end device, and no further treatment is needed."3) "segmented bus ducts (category 2), a number of the identified fire events were manifestedat bus transition points (a point where two segments of the bus duct are bolted together)rather than at the bus termination points. These events were generally attributed to loosebolted connections, to failed insulators, or to the accumulation of dirt/debris/contaminants inthe bus duct. The key, however, is that the effects of the fault are manifested at transitionpoints along the bus duct length. Fire scenarios for segmented bus ducts should, therefore,be postulated to occur at duct transition points (i.e., bolted connections)."Frequently Asked Question (FAQ) 07-0035, Rev 2, determined to be acceptable for use bylicensees in transition (ML091620572), provides additional information:"For segmented bus ducts (category 2), a number of the identified fire events were manifestedat bus transition points (a point where two segments of the bus duct are bolted together) ratherthan at the bus termination points. These events were generally attributed to loose boltedconnections, to failed insulators, or to the accumulation of dirt/debris/contaminants in the busduct. The key, however, is that the effects of the fault are manifested at transition points alongthe bus duct length. Fire scenarios for segmented bus ducts should, therefore, be postulated tooccur at duct transition points (i.e., bolted connections)."Failure Sequence Necessary to Damaqe DC CircuitsIn order to qualify the potential sequence of consequential damages resulting from a cablefailure in Trench 3, one must consider the failure mechanism itself, the design of the cablesystem, and the design of the protection system. to ONS-2015-094Page 22 of 30As noted in NUREG 1723, "Safety Evaluation Report Related to the License Renewal ofOconee Nuclear Station, Units 1, 2, and 3" Section 3.9.3.1.3 "medium-voltage cables (2-ky to15-ky) are subject to changes in electrical properties from moisture, excessive heat, andradiation." No significant cable stressors related to radiation or temperature exist in Trench 3;therefore, this discussion will focus on moisture. The widely postulated cause of a cable failurein Trench 3 is a breakdown in cable insulation due to moisture induced water treeing resulting ina fault. NUREG 1723 states:"The effects of moisture on medium-voltage cables can result in water trees when theinsulating materials are exposed to long-term, continuous voltage stress and moisture,eventually resulting in breakdown of the dielectric and failure. The growth andpropagation of water trees is somewhat unpredictable and few occurrences have beendiscovered in cables operated below 15 kV."NEI 06-05, "Medium Voltage Underground Cable White Paper," contains further discussion onwater treeing, stating:"This water-enhanced degradation does not cause direct breakdown of the [cable]insulation, but rather reduces the dielectric strength of the insulation, eventuallyweakening the material to the point where it is susceptible to voltage surges that caninitiate partial discharging. Partial discharging causes relatively rapid electricaldegradation, leading to an electric tree and a faulted condition in weeks to months."The medium voltage cables in Trench 3 were constructed with 173% of rated voltage Pinkethylene propylene rubber (EPR) insulation which, as stated in NEI 06-05,".has treated clay fillers to preclude water absorption that makes the insulation lessprone to water-enhanced degradation."In addition, the conditions for significant moisture exposure, as defined in NUREG 1723, do notexist in Trench 3 due to the passive drainage system, as verified by periodic inspection.Referring again to NEI 06-05:"In systems provided with adequate and well-maintained drainage, short-termsubmergence consistent with post-storm runoff does little more than wet the surface ofthe cable, given the slow diffusion of the moisture through common jacketing systems."Therefore, based upon the factors of the cable insulation design (i.e. 173% Pink EPR) and thedesign of the trench drainage system, it can be concluded that the occurrence of a moistureinduced water tree resulting in a cable fault is of extremely low likelihood.However, for the purposes of this discussion, one can assume that a medium voltage powercable line to ground fault due to insulation degradation does occur so that the possibleconsequential effects can be examined. After the initial single failure (e.g. the proposed line toground fault) occurs, it falls to the protection systems in place to limit the effects of damage.When evaluating the protection systems in place for this scenario, one must consider theprotective relaying, the system grounding schemes, and the physical protection provided byboth barriers and distance. In the event of a cable fault, in order for there to be any significantenergy transfer, there must exist a path back to the energy source to generate current flow. Forthe scenario of a line to ground fault in Trench 3 this path back to source would be created bythe conductor of one of the medium voltage power cables faulting to its grounded shield, whichis tied to the same station ground as the neutral grounding of the source. For the 13.8 kV powercables in Trench 3, the neutral grounding scheme is a high resistance ground designed with theexpress purpose of limiting both the fault current and any system overvoltage occurring as a Ito ONS-2015-094Page 23 of 30result of the fault until the protective relaying can rapidly clear the fault, thereby severely limitingany consequential damages as a result. Meanwhile, the 4.16 kV power cables in Trench 3 havea solidly grounded neutral which does provide overvoltage protection but without the same faultcurrent limiting capabilities. Therefore, the 4.16 kV cable line to ground fault will experiencecomparatively higher fault currents; however, this in turn inherently means that the protectiverelaying will actuate even faster (on the order of two-tenths of a second), thus limiting damage.As the energy released during a fault is a function of both the square of the current and the timethe fault exists, it can be seen that as long as the protective relaying functions properlyconsequential damages can be limited as a result. Damage resulting from a fault in one of the4.16 kV cables will also be directly related to the distance from the energy source to the faultlocation. As the overall shield resistance increases with distance from the source, the availablefault current will decrease, thus causing distance to be the determining factor in whether or notthe fault current will exceed the shield's short circuit withstand capability.If one were to further progress down the sequence of consequential effects and assume that thefault energy is not contained within the initially faulted cable, either as part of the initial failure orby failure of the protective relaying, one must take a wider look at the other protective systemsin place. As stated above, each medium voltage cable has a grounded shield (furnished asbronze tape layered 20 mils thick instead of the typical 5 mils and connected to the same stationground) that any fault would have to bypass before propagating to an additional phase, whilealso providing another path to ground to quench the fault energy. In addition, each mediumvoltage cable is specified with 173% of voltage rating conductor insulation which would providea greater dielectric for any adjacent faulted cable to overcome. The 173% rating is the greatestof the three conductor insulation ratings discussed in IEEE-141 and is specified for situationswhere a fault could remain on a system indefinitely. Beyond the cables themselves, each trefoilbundle of medium voltage cables then rests on a set of unistrut supports, spaced horizontally infour foot intervals, that are furnished with their own #2/0 bare copper grounding conductors alsoconnected to station ground, thus providing another low resistance path to divert the faultenergy. If one were to postulate further that some force or event happened to cause the faultedpower cable(s) to come into direct contact with the DC control cables at the bottom of Trench 3,the fault energy would first have to bypass the outer jacket and the layer of galvanized steelinterlocked armor for each affected cable. Bypassing the galvanized steel interlocked armor ofthe DC control cables would be complicated by both the armor being grounded on both ends tostation ground, thus providing another low resistance fault path, and by the general materialproperty of steel having a higher melting temperature than the copper in the conductor(s) of themedium or low voltage cable(s) purportedly transmitting the fault energy. Thus, if an event wereto occur with the energy necessary to affect both the AC and DC circuits within Trench 3, it ismuch more probable for it to cause an open circuit than to create any electrical continuitybetween the two systems.However, if one were to suppose that the steel interlocked armor were to be bypassed andelectrical continuity were to exist between the medium voltage cables and the low voltagecables, then one would have to consider that the DC system itself is a floating, ungroundedsystem (except for the high resistance, ground detection circuit) and does not provide the samelow resistance path back to the source for fault energy to traverse, as compared to the otherexposed metallic surfaces within the trench. In addition, the spare conductors in the low voltagecables are grounded at the ends of the trench and would provide yet another low resistancepath to quench the fault energy. The lack of a low resistance path back to source within the DCsystem itself for current to flow would instead result in an overall rise in the electrical potential ofthe system as it equalized with the voltage of the faulted conductor. As the low voltage cablesenter termination cabinets on either end of Trench 3, the most likely outcome of this voltage riseon the system would be shorting or arcing from the terminal strips in the cabinet to the grounded to ONS-2015-094Page 24 of 30walls of the cabinet itself due to their voltage rating being exceeded. The same suppositioncould then be made for each subsequent cabinet as you progress down the DC system.Therefore, for a medium voltage power cable fault in Trench 3 to impart a transient on both setsof low voltage DC cables, the following sequence of events would be necessary: A line toground fault would occur, bypass the grounded metallic shield of the faulted cable, propagate tothe adjacent medium voltage power cables and bypass their respective insulation and groundedmetallic shields, bypass the grounded unistrut and bare copper grounding wires running thelength of the trench, all to arc across the distance remaining between the faulted cable(s) andthe DC cables following the event. This arc would then require the energy necessary to bypassthe outer jacket and grounded galvanized steel interlocked armor of the DC cables while stillkeeping the copper conductors intact for the electrical continuity needed to impart energy on theDC system without creating an open circuit, all while bypassing any grounded spare conductorswithin the cables. For the fault to propagate further into the DC system, the voltage presentwould have to stay within the rating of the terminal blocks in each DC terminal cabinet, elsearcing to the grounded surface of the cabinet could occur. In addition, for both sets of DCcables to be affected, the arc energy would require the size necessary to spread the width of thetrench. This entire sequence of events would have to transpire either within the time it wouldtake the protective relaying to clear the fault, or subsequent to an additional failure of therelaying. Due to both the multitude of grounded metallic surfaces and the protective relayingpresent, this sequence of events is considered to be improbable to the point of lackingcredibility.Industry Operating ExperienceOperating Experience was reviewed to determine applicability to the current concern. Theapplicability of two specific items highlighted by NRC personnel to Duke Energy (i.e., theSwedish paper and the German paper) is discussed first. Following these two items, the DukeOperating Experience search criteria, evaluation, and results are presented.Swedish paper -"Distribution System Component Failure Rates and Repair Times" 1The applicability of this paper to the medium voltage power cables in question at Oconee islimited. The paper presents a literature search of reliability information for the period1993 -2003 for electrical transmission and distribution system components. Its purpose isfocused on retail transmission and distribution system outages and repair times with noapparent consideration of data from power generation plants. Specific to undergroundcables, the paper cites a failure rate of 0.95 per 100 km /year from a Norwegian paperpublished in 2002 for 33-110 kV cables. This is equivalent to a failure rate of 3E-03 peryear per 1000 feet of cable, and is significantly higher than suggested by other data. The33-110 kV voltage levels almost certainly indicate cables of a different design (likely to beof XLPE or PILC insulation) and subject to external influences such as lightning, switchingsurges, dig-ins, underwater routing, etc., and thus would not seem to be directlycomparable to the Keowee Trench 3 installation. The more applicable failure rateinformation is that provided in the NRC Generic Letter (GL) 2007-01 industry response, therelated Nuclear Energy Institute (NEI) documents and the EPRI technical reports. Thisinformation is based on US commercial nuclear power plant data for cables using similardesigns, installation and operational modes.' F. Roos, S.Lindah, "Distribution System Component Failure Rates and Repair Times -An Overview" NordicDistribution and Asset Management Conference 2004, Finland August 2004 to ONS-2015-094Page 25 of 30German Paper -"Investigation of Higqh Energqy Arcing Fault Events in Nuclear Power Plants'2This paper is focused on the phenomenon of High Energy Arch Faults (HEAF) events andthe potential effects on plant equipment around them. It is specific to nuclear power plantsand provides a very broad review of the published literature (US and International) anddescribes a number of significant HEAF events that have occurred. The point of the paperis that HEAF events can cause much more severe damage than typical fire events anddeserve special attention and more research. Interestingly, a point is made in the paper isthat "fault clearing time plays the largest role in the arc-fault hazard category." Of theHEAF events described in the paper, the cases of severe plant damage all involved afailure or delay of the circuit protection system to clear the fault in a timely manner. In oneparticular example, an unisolated transformer fault resulted in a short circuit that lasted forabout 7.5 minutes. The fault overloaded downstream power cables which caught fire anddamaged adjacent control cables in the turbine building cable duct. In this case, the controlcables were damaged as the result of the circuit breaker failure that lead to the cable firebut were not failed by the HEAF directly. This paper generally supports the Duke positionthat ground faults do not cause significant damage to surrounding cables as long as therelay protection scheme operates as designed.Operating Experience Search ReviewOperating Experience related to medium voltage cable faults was reviewed to determineapplicability to ONS Keowee related cabling systems. In particular, incidents where faultspropagated to phase-to-phase faults and/or resulted in damage to nearby cables was evaluated.Operating Experience was reviewed as described below. OE deemed as most applicable toONS Keowee cabling systems are described in greater detail within Table 1 of this document.OE which resulted in phase-to-phase fault or was viewed as particularly noteworthy is furtherdescribed in results section of this document following Table 1.I. INPO ICES database with the following parameters:Database: INFO ICES Documents: Search All Document TypesTime Frame: Anytime Keywords: medium voltage cable failThe search specified above yielded OE from within the Construction Experience andOperating Experience areas within ICES. The ICES search engine uses an "And"operator for keywords specified above, but the words do not need to be in the exactorder as above. This search yielded 299 ICES incidents.The following types of incidents were omitted from further applicability consideration:-Faults caused by equipment failures that damaged medium voltage cables-Faults which occurred outside of the cable run (e.g. connection points, inequipment, etc.),-Faults due to non-electrical issues that are not applicable to the Keowee trenchor PSW Busduct (e.g. cable cut with a back hoe), and-Faults where there was insufficient information available to make anydetermination of the cause of the cable fault.I1. While reviewing the River Bend event (May 2012) additional OE references that werenot identified by the ICES search were identified. These additional OE items wereincluded as part of this OE review.2 Heinz Peter Berg and Marina R6wekamp (2011). Investigation of High Energy Arcing Fault Events in NuclearPower Plants, Nuclear Power -Operation, Safety and Environment, Dr. Pavel Tsvetkov (Ed.), ISBN: 978-953-307-507-5, InTech, DOI: 10.5772/20497. Available from: http://www.intechopen.com/books/nuclear-power- Ito ONS-2015-094Page 26 of 30Ill. EPRI data, INPO Topical Report TR10-69 "Cable Aging and Monitoring" (May 2010),and INPO SEN 272 "Underground Cable Ground Fault Causes Forced Shutdown" wereresearched. The only unique QE which was determined as applicable was an incidentinvolving rodent damage at Browns Ferry which was seen as an anomaly. This OE wasincluded in results, however, as rodent damage cannot be categorically disqualified.IV. Attachment to GL 2007-0 1 was reviewed to determine if there were additional failurecauses that were not already identified as part of the ICES search. No additional causeswere identified.V. Oconee response to GL 2007-01 -- This Generic Letter requested licensee informationon failures of inaccessible or undergound power cables. The response for Oconeeincluded a single failure of the power cable for HPSW Pump B. Details below:Cable Type: Ethylene Propylene Rubber (EPR) cable insulation, semi-conductiveand copper tape shielded.Manufacturer: Okonite.Date of failure: January 10, 1980.Type of service: Normally de-energized, High Pressure Service Water Pump B,I /c-#2AWGVolta~qe class: Nominal system voltage 4160 VAC, cable rating voltage 5000 VAC.Years of service: Approximately 7 years.Causes: No formal root cause was performed. Failure was discovered during periodicDC voltage testing of the pump motor. The feeder cable was included in the test circuitand its insulation "broke down" at a voltage above the operating level. Since the resultswere deemed unacceptable, the cable was replaced before the pump was placed backin service. The investigation report indicated that the most likely contributors wereexcessive exposure to moisture combined with a problem on the cable jacket.In addition, Corrective Action Program searches identified another Oconee mediumvoltage cable failure. This cable provides switchyard 4.16 kV auxiliary power. Duringsecurity upgrades, a guardrail post penetrated this cable which resulted in opening of thefeeder breaker. Due to the nature of the cable failure (i.e. external mechanically-induceddamage with no evidence of prior degradation), this was excluded from the response toGL 2007-01 and from further consideration of this review.VI. Oconee Corrective Action Program was searched for period after NRC GL 2007-01response to May 22, 2015 (where PIP system was retired). This research identified theCT-5 cable failure. This failure was in the cable run, was caused by a cut in the outerjacket which allowed water entry into the cable with formation of insulation water treesand a subsequent a single phase to ground fault. The cable insulation is black EPR.Background -Oconee Cabling System Overview:Description of Power Cable Installation in Trench 3:The medium voltage cables in Trench 3 are installed in a precast reinforced concretetrench/duct bank system and are protected from external influences such as seismic events,tornadoes, excessive heat and lightning strike. None of the cables are direct-buried. The powercables are racked on elevated unistrut supports that comply with the cable vendor'srecommendations. to ONS-201 5-094Page 27 of 30The trench/duct bank system is sealed to preclude water entry and is provided with a passivedrainage system at the low points. The drains are periodically inspected for properoperation. The cables have been evaluated by the Oconee Cable Aging Management Programto ensure that there are no environmental stressors that would cause premature aging ordegradation.The power cables are rated for wet or dry applications and have a modern pink EPR formulationthat is less susceptible to water treeing than other insulation systems such as black EPR orXLPE. The power cables were installed and inspected using QA-1 procedures. Rigorousfactory and post-installation cable testing was performed to provide assurance that the cablesare defect free and were not damaged during shipment or installation. There are no cablesplices.Description of Power Cable Installation in Protected Service Water Manholes 1-6:PSW Manholes 1-6 are sealed to preclude water entry and are provided with a passive drainagesystem at the low points. The drains are periodically inspected for proper operation. Per thePSW license amendment dated August 13, 2014, the accessible 13.8 kV PSW cables will beinspected every 10 years and have electrical testing every 6 years. The power cables aresupported by cable tray and/or unistrut.The power cables are rated for wet or dry applications and have a modern pink EPR formulationthat is less susceptible to water treeing than other insulation system such as black EPR orXLPE. The power cables were installed and inspected using QA-1 procedures. Rigorousfactory and post-installation cable testing was performed to provide assurance that the cablesare defect free and were not damaged during shipment or installation. Due to the length of thecable pulls, there are cable splices in Manholes 1 and 5; however, post installation testingconfirmed that the cable splices are discharge-free per IEEE standards.ResultsThe following CE are the medium-voltage cable faults (21) where the fault occurred in a cablerun and are determined to be most applicable to ONS underground cable system attributes.Armor type was not consistently documented in the CE items, so all cables were reviewedregardless of the fact the cable was armor or unarmored. Table 1 provides a summaryapplicability review of these CE. Additional detail is included following Table 1 for CE whichresulted in phase-to-phase faults or where the CE description indicated damage to adjacentcables.

Attachment i to ONS-2015-094Page 28 of 30TABLE 1: Medium-Voltage Cable Faults Identified in ICES SearchICES F lt/ Propagated from Portion of EventStation Report # Month Year Fault Froault ion LGt -vlae gisLocation Prpgtin LGooLLnvlaede gisCause Configuration* Cause #7 Yes ° Cause #6Cable run into

• Cause #14Robinson 242331 Mar-2010 4 kV Cabinet

• Cause #6Trillo 297938 Oct-1993 Connection

• Cause #13 Yes

• Cause #6* Cause #6Brunswick 216026 May-2005 Cable run

• Cause #1 Yes

• Cause #1* Cause #2

• Cause #2Limerick 309779 Feb-2014 Cable Run

• Cause #1 Yes

• Cause #1* Cause #8Sequoyah 198929 June-2002 Cable Run

• Cause #1 No

• Cause #1Browns Ferry 225047 Feb-2007 Cable Run

• Cause #1 No

• Cause #1North Anna 237341 Apr-2009 Cable Run

• Cause #7 No N/AChaulk River 238157 Jun-2009 Cable Run

• Cause #5 No

• Cause #5Harris 251793 Nov-2011 Cable Run

• Cause #1 No

• Cause #1River Bend 254057 May-2012 Cable Run

• Cause #3 No

• Cause #3 (P5W)* Cause #1

• Cause #1* Cause #2

• Cause #2Beaver 308418 Nov-2013 Cable Run

• Cause #12 No N/AQuad Cities 311048 Apr-2014 Cable Run

• Cause #10 No N/AHarris 314044 Nov-2014 Cable Run

• Cause #3 No

• Cause #3 (P5W)* Cause #9Oconee CT5 PIP-O Dec-2011 Cable Run

• Cause # 1 No

• Cause #115323

• Cause #5

• Cause #5Quad Cities 300595 Jul-2012 Cable Run

• Cause #1 No

• Cause #1* Cause #9Callaway 219822 Feb-2006 Cable Run

• Cause #3 (In No

• Cause #3 (PSW)manhole)North Anna 225471 Mar-2007 Cable Run

• Cause #4 No

• Cause #4* Cause #7Prairie Island 237726 May-2009 Cable Run

• Cause #1 No

• Cause #1* Cause #8* Cause #9 ___________Point Beach 230369 Jan-2008 Cable Run

• Cause #1 No

• Cause #1North Anna 242196 Mar-2010 Cable Run

• Cause #15 L-G N/A(Arcing DamagedCable Tray)___________EPRI -EPRI Report Cable Run

• Cause #11 No N/ABrowns FerryCauses:1)2)3)4)5)6)7)8)9)10)11)12)13)14)15)Wetting / Water TreeContaminants in water accelerated Water TreeInadequate Cable SpliceDamaged during installationAdditional Damage (Cut by knife, Walking on, etc)Breaker Out of Service / Failure (Mechanically or electrically) -This cause did not result in the fault, but contributed to thepropagation of the fault.Inadequate Support (Vertical or Horizontal)Cable Manufacturer DefectsDirect Bury ApplicationCable Bend Radius ExceededRodent DamageHigh Ambient temperatureHot Spot in shunt connectorInstalled Cable did not comply with cable specificationShield Grounded on one end of the cable to ONS-2015-094Page 29 of 30Additional Details of Selected QE ItemsThe following OE resulted in phase-to-phase faults or indicated damage to adjacent cables.The line to ground fault at Robinson propagated to a line to line due to the breaker being out ofservice and the next upstream breaker was required to clear the fault. This additional time thatthe fault was present increased the probability that the fault would propagate. This OE identifiedthat the cables adjacent to the faulted cable were damaged. The damaged cables jackets weredamaged due to the heat generated during the event but the cables would have been capableof providing their safety function (i.e. power or control) to mitigate a design basis event.The line to ground fault at Trillo propagated to line to line due to the a breaker mechanicalfailure and the next upstream breaker was required to clear the fault. This additional time thatthe fault was present increased the probability that the fault would propagate. This OE did notidentify that the cables adjacent to the faulted cable were damaged.The line to ground fault at Brunswick propagated to a line to line in the three conductor cablefeeding the Fire Protection System pump motor. The breakers are designed to trip once theground faults on two phases were connected by a low impedance path. The breakers did trip asdesigned. This OE did not identify that the cables adjacent to the faulted cable were damage.The line to ground fault at Limerick propagated to a line to line. Troubleshooting identified twoof the three phases of the power feed cables failed due to water treeing caused by amanufacturer defect. The ground fault overcurrent relay did trip to isolate the fault. No cablesadjacent to the faulted cable were damage.The line to ground fault at Beaver Valley occurred in the Husky Bus Enclosure. The cable hadbeen installed for 35 years and experienced diminished service life due to chronic exposure toohmic heating. The Arc Flash event due to the fault damaged the cable within the cubical inaddition to cubicle itself.The line to ground fault on the 120VAC system at Quad Cities in 2014 did not propagate to aline to line fault. This cable was damaged due to a bend exceeding the bend radiusrequirement and a subsequent steam leak injected moisture into the damaged section of cable.This OE did identify that the cables adjacent to the fault cable were damaged when the rung ofthe cable tray began heating due to the Steam Leak and the line to ground fault. This heatingled to the additional cables' insulation to melt and the cables eventually shorted as well.OE SummaryFor the 13.8kV ONS cables, the 18 Amps fault current due to a line to ground fault is well withinthe capability of the cable shield current capacity. Additionally, fault currents of this magnitudeare not sufficient to result in damage to nearby cables should the fault penetrate the cablejacket. For the fault to propagate to a phase-to-phase fault (and thus the fault current besufficient to expect damage to nearby cables), the initiating cable's shield and jacket would haveto fail coincident with simultaneous failure of additional (phase) cable or ACB breakers.The shielding on the 4kV cable feeding transformer CX is not capable of carrying the worst caseline to ground fault. The fault current may cause a breach of the outer cable jacket. For thisfault to further propagate, breaker 1TC-04 breaker would have to fail, allowing the fault tocontinue until the upstream breaker 1TC-14 isolated switchgear 1TC to clear the fault. Theadjacent cables may be damaged if the fault is allowed to propagate. It was not immediatelyclear in all cases from the four QE events that did propagate to a line to line fault the extent ofthe damage or the cause of the damage. Since adjacent cables were damaged, it can be Ito ONS-2015-094Page 30 of 30assumed that if the adjacent cables were control cables then they would be damaged by thefault. Damage due to the transient voltage induced on the low voltage cable due to the fault onthe medium voltage cable was not discussed in any case.Based upon the reviewed OE the line to ground faults that propagated to line to line faultsoccurred due to breakers being out of service, breaker failure, or detection was configured todetect line to line faults.ConclusionsAs can be seen in the licensing history of Oconee, there were several instances where the NRCreviewed the Oconee electrical design and even recognized the plant-specific single failurecriteria. The plant-specific nature of the design is not surprising, given that Oconee wasdesigned and built before most of the NRC-accepted guidance was developed. Documentationassociated with the NRC's review and approval of the Oconee design is relatively lacking indetail due to the plant's vintage; however, because Oconee was built to Duke Power Companyfleet design specifications that were also employed for the design and construction of the newerMcGuire and Catawba nuclear power stations, consideration of the NRC's review of the designof these other Duke plants can inform a review of the Oconee design. As documented above,the NRC reviewed the cable design for McGuire and accepted the use of armored cable as ameans of providing protection against fault propagation and damage to adjacent cabling.The concern for a MV power cable fault propagating to control cabling, resulting inconsequential failure(s) of DC systems is beyond the plant's design and licensing basis and isnot supported by operating experience (CE). Duke Energy conducted an extensive review ofCE, which concluded that cable faults resulting in damage to adjacent equipment requiredfailures in addition to the initiating cable failure (e.g., failure or disabling of circuit protectiveequipment) in order to cause the collateral damage. The consideration of multiple failures isoutside of the ONS single failure design criteria. In addition to active circuit protectiveequipment, passive design features exist to protect against the propagation of a cable fault toDC systems. As discussed previously, the ONS design employs the use of grounded cablesupports and grounded shielded cables that effectively provide a series of intervening groundedbarriers to protect against fault propagation. These barriers would also have to fail in order for afault to propagate to the point of damaging DC circuits.In conclusion, Duke Energy believes the configuration of cables in Trench 3 and the PSWductbank is safe and in accordance with design and licensing requirements for cable routing andcable separation at Oconee that have been reviewed and accepted by the NRC.