ML20066D360
| ML20066D360 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 01/08/1991 |
| From: | Helwig D PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | Murley T Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9101150003 | |
| Download: ML20066D360 (9) | |
Text
__
PHILADELPHI A ELECTRIC COMPANY NUCLEAR GROUP HEADQUARTERS 955 65 CHESTERDROOK DLVD.
WAYNE, PA 19087 5691 (316) 64b 6650 o ^;' D ",, Hj,',i, ' a January 8, 1991
. m.......
,o. a... m.
50-278 License Nos. DPR-44 DPR-56 Dr. Thomas I. Murley, Director Office of Nuclear Reactor Regulation U.S.NuclearRegulatoryCommission Washington, DC 20555
SUBJECT:
Peach Bottom Atomic Power Station, Units 2 and 3 10CFR50.63, " Lots of All Alternating Current Power" Response to NRC Safety Evaluation and Claim of Backfit
Dear Dr. Murley:
The purpose of this letter is to submit our claim that the NRC positions used to determine that Peach Bottom Atomic Power Station (PDAPS), Units 2 and 3 do not comply with the station blackout (SBO) rule, 10CFR50.63, constitute backfits as defined in 10CfR50.109(a)(1).
Since we consider these NRC positions to be backfits, we request that the NRC rejection of our response to the SB0 rule for PDAPS, Units 2 and 3, be reversed.
As detailed in the attachment to this letter (which also provides additional background informatu,,i), the NRC positions in question were provided in the NRC Safety Evaluation (SE), transmitted by letter dated August B, 1990, and during a meeting with the NRC reviewers held on September 10, 1990 (documented in the NRC meeting summary dated November 6, 1990).
Specifically, the NRC conclusion, stated in its August 8,1990 letter, is that the existing Emergency Diesel Generators (EDGs) at PBAPS, Units 2 and 3. "...do not meet the excess redundancy requirement that would permit them to qualify as alternate AC (AAC) power sources under the 500 rule...." As a result of that conclusion, the August 8,1990 NRC letter states that "...a separate AAC source must be added or a re analysis should be conducted to show that the plant can cope with an SB0 for the required duration without an AAC source...."
ob umsnisass 9
1 p
..',ts.in
a i
U.S. Nucle 6r Re!,ulctory Commission January 8, 1991 Dr. Thomas E. Hurley Page 2 We consider, for the reasons summarized below, thdt the NRC's conclusion is based on positions which are inconsistent with or go beyond the requirements of the 500 rule. The 500 rule provides that at multi-unit sites (such as PBAPS),
existing EDGs may be used as AAC power sources wh$re the combination of emergency AC (EAC) power sources exceeds the minimum redundancy requirement (i.e., single failure criterion) for non-design basis accident safe shutdown of all units. We have determined, in conformance with NRC accepted SB0 rule guidance, that the combination of the existing four EDGs at PBAPS (which are completely shared between Units 2 and 3), exceeds the SB0 rule minimum redundancy requirement and therefore, any one of the four EDGs meets the SB0 rule requirements for an AAC power source.
The NRC's conclusion to the contrary is based on positions which, with respect to the EDG configuration at PBAPS, are not required by the SB0 rule or included in NRC guidance. The issues involved in these positions include the following.
o Two completely shared EDGs comprise the minimum number of EAC power sources needed to shut down both units during a Loss of Offsite Power (LOOP) event, rather than the three EDGs required to shutdown both units during a concurrent LOOP event and design basis accident at one unit.
o The minimum redundancy requirement is satisfied by one EDG, rather than two, since both of the two EAC EDGs power safe shutdown equipment for both units during a LOOP event.
Therefore, the PBAPS EDGs are categorized as a two-out-of-three EAC configuratiot, resulting in an 8-hour coping duretion for an SB0 event, rather that a two-out-of-four EDG category which wculd result in a fo, nour SB0
- duration, The use of the 200-hour EDG rating for determining that necessary o
safe shutdown loads can be powered during a LOOP event, rather than the 2000-hour or continuous rating, does not mean that the EDG will fail during or at the end of the 200-hour period, and is therefore reasonable, o
A reasonable number and type of operator actions can be credited in evaluating the station staff's response to an anticipated operational occurrence, and the acceptability of these operator actions can not be based on a comparison to the more limited scope of operator actions allowed to be taken during a design basis accident.
As explained in the attachment, we consider that the NRC's rejection of the SB0 analysis for PBAPS, Units 2 and 3, is based on positions which go beyond the requirements of the SB0 rule and are technically unjustified.
Accordingly, we request that the NRC's rejection be reversed.
l
U.S. Nuclear Regulatory Commission January 8. 1991 Dr. Thomat, E. Murley Page 3 A
If you have any questions or require additional information, please contact us.
Very truly yours, s
(
Attachment cc:
J. M. Taylor, Executive Director for Operations USNRC w/ attachment E. L. Jordan, Director Office for Analysis and Evaluation of Operational Data, USNRC J. G. Partlow, Associate Director for Projects, USNRC W. T. Russeil, Associate Director for Inspection and Technical Assessment, USNRC
- 1. T. Hartin, Administrator, Region 1. USNRC J. J. Lyash, USNRC Senior Resident inspector, PDAPS 64 t
ATTAOMENT FEAO! IOPJIM A*ItMIC IUeR SDLTICW, tNTIS 2 and 3 BIATICH IEAOCATT Ih pcmea to NRC Safety Evaluation and Claim of Backfit By letter dated August 8,1990, the }mC tranmitted its Safety Evaluation (SE) and Technical Evaluation Report (TER) related to our IDsponse to 10dR50.63 (i.e.,
the station blackout (SIC) rule) for the Peach Bottm Atomic Fwer Station (Pl%PS),
Units 2 ard 3.
'1he Atrjust 8,1990 Imc letter stated that IEMS, Units 2 and 3, do not ocrply with the SB0 rule, and that a revised response should be suhttitted. At rur request, a mooting was held between representatives of Ihi1%delpttia Eloctric Qqnny (PDCb) and the }GC on Septaber 10, 1990 to discuss the issues ard additional informtion requesta specified in the imC SE ard TER.
By letter dated October 15, 1990, we submitted an interim response to the imC SE stating that as a result of the 10C clarifications and positions provided in the 10C SE and during the Septembc.r 10, 1990 meetirrJ, w would re-evaluate the SIC rule response for PIRPS, Units 2 and 3.
We further ocanitted to assess various options ard subnit the dorcription and schedulo of the selected option by hh 31, 1990.
After further evaluation, we have concluded that PIAPS, Units 2 ard 3, do oxply with the SIC rule, as described in our letters dated April 17, 1989 ard April 3, 1990. We further fird that the positions cited in the 100 SE and during the September 10, 1990 mooting to support the }mc conclusion that PBMS, Units 2 ard 3 do not comply with the SEO rule, constitute backfits as defined in 10CIR50.109 since these positions go beyond the requirumonts of the SID rule, ard therefore are technically unjustified since conformnoe with these positions is not nooessary in order for PIRPS to ocrply with the SID rule nor to ensure adequate protection of the public health and safety. Accordingly, in accordance with lac Marmal Chapter 0514 ard IRmEG-1409, "Backfitting Guidelines," section 2.5, we are subnitting this backfit claim to you, the Director, Office of Nuclear Reactor Regulation, with a copy to the IRC Executive Director for Operations and the Director, Offico for Analysis and Evaluation of Operational Ibta, ard request that the imC rejection of our PIAPS SIC analysis be reversod.
LidGLU V 2 D PIRIs is a two unit station with four (4) Emtgency Diesel Generators (EDGs) that are ca pletely shared betwoon both units. 'Ihat is, when called upon to perform its safety function, each EDG starts and loads a safoguard electrical bus frun each unit.
Consequently, certain equipnent required to safely shut d wn both units, and mintain both units in a safe shutdown cordition, are powered by each ECG.
Furthet1nore, different safe shutdown equipment for both units is loaded on a particular safeguard electrical bus. For exanple, ED3 2 is capable of powering various safe shutdcun equipnent such as a Unit 2 Residual Heat Rem 3 val (MR) pump or a Unit 3 MR purp, ard a common Emargency Service Water (ESW) purp. Clearly, then, amergency AC power pInvided by the FDJs during anticipated operational occurrences as well as design basis accidents can mly be considered on a " station" basis, not on a "per unit" basis. Accordingly, we found that the ED3 loading configuration at PIAls, Units 2 ard 3, provides additional flexibility in the station responso to a Ioss of Offsite Power (IDOP) or SD0 event than would be the case with two dedicated EDGs por unit. As a result, we attenpted to pursue the NRC policy reflected in the SID rule Statement of Consideration that "...the door is open to licensecs kho believe their plant.s have additional (SDO) capability that should be considered by the (!mC) staff in denonstrating ocx:pliance with the rule."
Ibge 2 In dotomining the D:crgorcy AC (EAC) p:wcr cupplios for Pl%PS, Units 2 and 3 (i.e., the minitum number of EDGs rum to safely shutdwn both units in the avont of a LOOP), wo followod the stops in parts 2.A, 2.D, and 2.C of the Nuclear Management and Rosaurpos Courcil's (IUVJC's) guidarco document IUVJtC-8700, "Guidelinos ard Technical Bacos for IMUJC Initiativos Addrtssing Station Blacinit at Light Water Roactors." Wo corcluded that any two of tho four ED3s woru capable of pwcring the r--y nafo shutdwn eq.tiptont for bath units bacod on the folicwirg considerations.
o only oqalpment rcqaired to safely shutdwn ard unintain both units chutdwn durirg a ILOP went of exterdad duration rust be capablo of being powcrtd by the EAC pcecr sourcos. The capability to sinultarcously pcuer additional oquiptent khich is specifically noodod to mitigata the consequorcos of a design basis accident (e.g., locs of Ccolant Accident (LOCA)) doos tot nood to be ciansidered.
o Duo to the asyrmetry of the IIG loadirg describod above, a rtanomble number ard type of operator actions were noodod to connect ocrtain safo shutdcwn oqalp:ent to the FAC pcuer so.troos.
o Dependirq upon the particular ID3 combimtlan servirn as the EAC pcwor so.troos, the loadirg of a sirglo ED3 may bo within its 200-hour ratirn, rather than its 2000-hour or continuous rating. The current licensing basis for the ED3s at PIAIG, Units 2 ard 3, ircitdos conformnoo to a propacod IEEE stardard, "PIrpocod IEEE Criteria for Class lE Electrical Systems for Nuclear Power Generatirg Stations," dated Juno 1969, ard partial conformance to Atcanic Energy O:xtrnission (ADC) Safety Guide 9, "Soloction of Dioscl Generator Set Capacity for Stardby Pwer Supplies,"
dated March 10, 1971. Neither of these licensirg basis documents nor the PIAPS Updatcd Piral Safety Amlysis Report (UTSAR) ocamits us to mintain the PIRIS IIG loadirg within the 2000-hour or continuous ratiry for anticipated operational occurrorces such as a IDOP cvent.
Given that a LOOP ovent would, as designcd, result in the powering of ocrtain safe shutdwn ogalpacnt for both units by a minimum of two shared ED3s, we concluded that a third ED3 satisfies the roquiremont to account for minirum rodundarry of the FAC power courocs. This conclusion is based on the fact that the third ED3 satisfies the sirgle failure criterion. That is, sinoo two shared IIGs aru nooded to safely shutdown bcrth units in the event of a IDOP, a third EDG, which by design would power acrtain safe shutdcun equipacnt for both units, satisfies the ninimum redurdancy requirument arsumirg a failure of one of the two EAC EDGs.
In accordance with the SIO rule, where the cambiration of FAC pomr sources excoods the minimum Itdundancy requirrmants for non-dosign basis accident safe shutdwn of both units, the rmaining EDG (in this caso, the fou-th ED3) ray be uscd as the Altermte AC (AAC) pcuer source providad it uncts the AAC power scuroc requiruments. On this basis, we concludcd that one of the four existirq EDGs could corvo as the AAC power sourvo durirg a SD0 cvent. Wo should point out here that in the event of a SDO, i.e., a staticevide LOOP ovent ard the loss of throo EDGs khich constituto the EAC power sources, the rumainirg AAC EDG will, by design, start ard load ocrtain safe shutdcun ogalpnent for both units. Since certain safe shutdcun equip:ent will be powered ard other safe shutdwn equipcnnt will rot autantically be powered for both units, the dotermination of which unit is blackcd-out ard which unit is not blacked-out is not pocsible. Accordingly, we conservatively assumed that the 500 cvent affects both units at the same time.
rago 3
%o imC SE ard su;Tortirg mt, alorg with inforntico prwidcd by tho imC durity positions for the ImC's noetirn, prwid(d the follwiry;th the requirteents of the Sto tho Septaber 10, 1990 conclusion that Imis, Units 2 and 3, do rot cmply w rule as it portains to the JAC p:wcr scuroo definition. As explainod bel w, we consider those positions to constitute Lockfits in that those positions go beyord the rtquirteents of the Sto rule and are to&nically unjustificd.
Iy g PIT DI!n m Igj Hac_Irciticn Wo Imc relied on the fact that current IUJG 11oonrirg basis (i.e., the UPSAR) rhcus that the FAC p:Mor sourocs are oxprisod of throo ED3s, that is, throo of the four Ims are rtquired to safely rhut dwn toth unito durirg a LOOP cvent.
1000 lbeition 2e UIMR prwidos a safety amlysis to dw that there is a sufficient number of rms, assumiry a sirgle failure, for the nost limitirg fault (i.e., design basis accident). Accordingly, the UrrAR discussed the ability to safely shutdwn both units assuming cparation of three of the fcur IIna durity a station-wido I.00P cvent ard a concurrunt ifG at ano unit. 20 UIEAR dcos not curruntly includo an amlysis to detonnino the minirun number of Ims rrquind to safely rhut down both units during an anticipated operational occurrurce out as a IDOP cycnt. 20 FAC cvaluation perforncd in conformrco with the SIC rulo txquitumonts (i.e., the FAC pcuer source must be capable of powerirg ncn-design basis nacident safe shutdcun loads) shcus that two 12no can safely shut dwn both units durirg a IDOP cvent, accountirg for reasonable operator actions. Accordirgly, sitco the Slo rulo explicitly excitdes consideration of a concurrent 100P ovent ard a design basis accident, tho UPSAR safety analysis can not be uscd to support the conclusion that operation of thIro of the four IIns is r==ny for safo shutdcun of both units durirg a IOOP cvent.
NRC.IWitigo
%o Imc concludcd that two EDGs in addition to tho two Ims which servo as the FAC pcwor sources, are rtquirtd to satisfy the mininum rtdurdarcy requirements.
%croforo, the EAC Im category is two-out-of-four.
PIXb Positicn As dit-M above, the two rms which serve as the FAC power sourros can not be considered as pcucrity safe shutdcun equipent on a por unit basis Waum safe shutdwn loads for toth units are poworcd by both FAC ED3s. mis is clearly different frun a two unit station with two dodicated EDGs por unit (that is, safe shutdcun equipment for each unit is clearly ansociated with two ED3s).
In such a caso, based on the design whereby one EDG dodicated to each unit pcuers all the safe rhutdcun equipmnt for that t. nit, the pocord im dodicated to each unit satisfies the mininum rodurdarcy requiruments. Sinco safe rhutdcun loads at PDAPG, Units 2 ard 3, are not dedicated to cnrtain ED3s, ard two EDCs serve as the FAC pcuer catrees for both units, then onc* additional ED3 roots the minitum rtdandancy requirc:mnts.
1 l
Pago 4 Thoroforu, in acmrtiarce with titops 2. A, 2.D, and 2.C of IMARC-8700, the EAC EDG catcgory for PBAPS, Units 2 ard 3, is two-cut-of-throo. Note that ono ED3, the AAC pcuer sourco, has bcxan subtracted frun the total number of EDGs in acconiance with IMATC-8700, so as not to double count this ED3 as both an EAC and AAC poWur soutco.
In this way, PDAPS is cattgorized as an eight (8) hour copiry duration station.
If the EAC ED3 configuration is charged to two-aut-of-four, then PMPS, Units 2 and 3, would be classified in the less conservative four (4) hoitr copirg duration category.
NC Pociticn The Imc has taken the position that in detonniniry the number of EDGs which conprise the EAC power sources, only the 2000-hour or continuous ED3 ratify can be used.
PDoo Poniticn 7ho imC position that only the 2000-hour or continuous ratiry can be used in evaluatir the EAC pcwor sources was not pmvlously specified in the SD3 rule and accompany Statement of Ctnsideration, Regulatory Guido 1.155, " Station Blackout,"
dated Atqust 1988, or I M ARC-8700. Rtrthermore, as stated above, the current licensirg basis for PB PS, Units 2 and 3, does not limit EDG loads to the 2000-hour or continuous ratirg for anticipated operational mmrrences such as a LOOP cvent (i.e., UNAR Section 8.5). Operatirq an EDG loaded to its 200-hour rating for 200 hours0.00231 days <br />0.0556 hours <br />3.306878e-4 weeks <br />7.61e-5 months <br />, approximtcly one wock's time, would rot, accortlirg to the EDG verdor, result in its failure durirg or at the end of this period, but would only require noro extensivo preventive mintemroe.
ntrthermore, even -
- g the highly improbable cocurrenco of a IDOP event which lasted for one wock, plant operators would mnually add ard trip appropriate loads, as described in UNAR section 8.5.3, so as to reduce ED3 loads below the 200-hour ratiry. Therefore, we consider that use of the 200-hour EDG ratify for 100P ovent loads is reascnable and within the current PBAPS licensirg basis. Accortlirgly, this 100 position appears to im a now requiremont on the accepted PDAPS licensirg basis ard goes beyond the requirements of the SID nilo.
IEC IYnition The imC stated that additiomi loads needed to safely shut dcun both units, such as the Control Rod Drivo (CRD) pumps and Standby Gas Treatment System (SGTS) exhaust heater, were not includod in the tabulation of loads required to be poworod by the EAC power sources during a IOOP cvont. These additioml loads would exceed the EDG 200-hour ratirg.
PD00 Pooltigrg}
The acklitional loads identified by the 100 in its SE ard supportiry TER are associatcd with equipment needed to either mitigate the consequences of design basis accidents or other equipnont that has rot, urder the current PBAPS licensirg basis, previously been requitud to mintain safe shutdcun, ard are not autcznatically connectcd to the EDCs. Sirco the SID rule ard guidarro specifically excitdes consideration of design basis accidents in determiniry the EAC power sourros, and these additiom1 loads would not need to be stripped frun the EDGs since they are not autcratically loaded durity a LOOP event, the !GC identified additional loads should not be added to the IOOP shutdown load tabulation. Thereforu, EAC loads do not exceed the 200-hour EDG ratity.
l
Page 5 BBqfonitim the }mc stated that urder SIC carditions, one unit nust be assumd to bo blacked-out and one unit nust be assumed to be experiarcim a IDOP event but is not blacked-out.
HXb Iboitim As dimW above, for any IDOP event at HAPS, Units 2 ard 3, khcre at least one EDG starts ard loads its ammiated safeguard electrical bus, ocrtain safe shutdown equipnent at both units will be automatically powertd while other safo shutdown equipnent for both units will only be powered after speciflod manual operator actions are taken. Therefore, a blacked-out unit ard a non-blackod out unit cannot be distirguished. For this reason, our SIC analysis for PIRPS, Units 2 ard 3, conservatively assumos that both units are blacked-out. This is different frcrn the caso of a two unit station with two dedicated EDGs per unit.
In this case, the unit at khich the Indurdant EAC power sourtes are assumed to be lost durirg a IDOP event is clearly the blacked-out unit.
100 Positim 2he lac stated that an inoniinate nunber of operator actions, such as switchirn ono MR punp betwoon units, is rut acooptable in determining the EAC power courous.
Moo Positim Takim credit for operator actions in determining the FAC power sourecs was recognized as acceptable in the Statanent of consideration arvv=wnyim the SB0 rulo (Imc response to comment no.13, 531P23211). As part of our detonnination that any two EDGs could serve as the FAC power sources, we identified the operator actions that would be perfonned to power other safe shutdown equipment frun any ocanbination of two EDGs.- We then walked through these actions with licensed operators and found them to be reasonable, both in number and scope, ard with respect to the tino within which ocrtain actions must be taken. kkiitionally, as stated in our April 3,1990 lotter, operators would Dgt need to switch an RHR punp betwoon units in order to achieve and maintain safe shutdown of both units during a IDOP event, even though switchirg of an MR punp between units is already identified as operator actions taken durity a certain abrormal operaticnal transient in our licensing basis (i.e.,
i UPSAR section 14.5.7).
Thernfore, we have concitdod that the number of operator actions nooded to safely shutdown both units durirg a IDOP ovent usire two ED3s as the EAC power sources is not inoniinate nor unreasonable in scopo ard timing.
Furthermore, we consider that a ocrparison of the operator actions that would bo taken durirg an anticipated operational occurrence such as a IDOP cvont with the operator actions that would be taken in the event of a design basis accident is not valid, given that the range of operator actions that have boon fourd acceptable to achicvo ard maintain plant safe shutdown in accordance with 10CFR50.48, " Fire Protection," is significantly broader than those operator actions that are accountod for in mitigating a IDCA. This acrparison of the SIC rule with the fire protection rule is appropriate based on the discussion in the SID rule Statenant of Consideration that the SB0 rule "...should be viewod as boirg in the same accident prevention context as...the fire protection rule ( 50.48)..." In any event, we interd to p.irsue the prtoodural charges r="ry to incorporate the identified operator actions for a IDOP ovent since preliminary findirns of the Probabilistic
i Ibgo 6 Risk Annossent (IT%) bairy performad for IYAIS, Units 2 ard 3, rJus that prTooduralizirq the identified operator actions results in a substantial rodaction in the contribution of IroP events, ircludisq an SEO went, to the overall core clamage firquerry for IYRIS, Units 2 ard 3.
For the roanons explairxd ab:ne, we contsider that the 100 rejoction of SIO analysis for 14AIS, Units 2 ard 3, is based on positiers that go boyard the S10 rulo rcq.tirements ard are tahnically unj'2stified. Io:ordirgly, we rtquest that the imC rejection of the TYAls SIO aralysis be revertad.
._