IR 05000255/1991019

Ack Receipt of Informing NRC of Steps Taken to Correct Violations Noted in Insp Rept 50-255/91-19

ML18058A341
Person / Time
Site:	Palisades
Issue date:	04/10/1992
From:	Miller H NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:	Slade G CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
References
NUDOCS 9204160027
Download: ML18058A341 (28)

Text

• -*

Docket No. 50-255 consumers. Power company ATTN:

Gerald B.* Slade

. General Manager y

APR 1 0 19:!2 Palisades Nuclear Generating Plant 27780 Blue Star Memorial Highway Covert, MI 4904i

Dear Mr. Slade:

SUBJECT:

OPEN ITEMS (NRC INSPECTION REPORT

~O.

50-~~5/91019{DRS))

.

~his will acknowledge receipt of your.letter dated Mar~h 31,*

1992, in response to our Jetter dated January Ji, 1992, identifying open items associated with Inspection Report No.

50-255/91019 (DRS).. This rep'ort summarized the results of an electrical distribution system.functional inspection (EDSFI) at

.

.

. your Palisades Plant*.

We have reviewed your actions* to resolve.

these is~ues and have no further qu~~tions at this ti~e.

~he resolution of these op.en items will be examined during future inspections.

Enclosure:

tetter dated March 31, i.~92. **

See Attached Distribution R~~,,

• 111./171.

NJ'sler/jk 04/07/92 yr 9204160027 920410 PDR ADOCK 0500025*5.

Q

. PDR RIII

~*

Gardner 04/ Oi/92 s*incerely, * *

.\\...

.................... '::__;

H. * J ~

Miller~* *o+/-rector Division of Reactor.Safety

• lf

"Ring RIII rv'-.

• Martin 04/<) /92 04/\\/92 I

Consum*ers Power Company Distribution.

cc w/enclosure:

Dayid P. Hoffman, Vice Pre~ident

_Nuclear Operations P.. M. Oonnelly; Safety and Licensing Director DCD/DCB. (RIQS}

OC/LFDCB

.

Resident Inspector, RIII

• James R. Padgett, Michigan Public Service Commiss*ion Michigan De~art~ent o~

• Public Health

• arian Holian,. LPM, NRR SRI, Big Rock Point E. Imbro,. NRR W.. Hodges, Region I J. Durr, Region I A~ Gibson, Regioti lI C.. Ju.lian, Region II..

s. Collins, Region IV_

. T. Stetka, Region IV R.

Zim~ermari, Region V D. Kirsch, Regiori v..

)

'*

• '.*-~

'

• -

ATTACHMENT Consum~rs Power Company _

Palisades Plant

Do.cket 50-255 REPLY TO*OPEN ITEM~ -

INSPECTION REPORT No. 91019 March 31, 1992 *

24 Pages

• • -*

consumers Power POW ERi Nii

• NllCHlliAN"S PROGRESS Palisades Nuclear Plant:

27780 Blue Star Memorial Highway, Covert. Ml 49043 March 31, 1992

.

.

• Nuclear Regulatory tommis~ion. *

,Document Control Desk Washington, DC.20555 GB Slade Gmual Managtr

• DOCKET 50-255 - LICENSE DPR-20 - PALISADES PLANT - REPLY TO OPEN ITEMS; NRC INSPECTION REPORT No. 91019.*

NRC inspection Repdrt No.. 91019 *~rovided th~ resul~s of 'the sp~~ial el~ctric~l

.disttibution system functional inspection (EDSFI) Of the Palisad~s Plant. :The.

inspection report identified apparent violations and deviations from NRC requirements.. Consumers Po~er Company* responded to the violations and. :

deviations in ~ letter dated March 2, J992.

The ihspecti6n report also * *

identified ~ number of open items and requested a reply to those.items as well.

As was discussed between Mr. Bruce Jorgensen of Region III staff and Mr. Pat Donnelly of our staff, on February 12, 1992, a reply t~ the open items was to be pro~ided by March 31, 1992.

Attached to this lettei are the responses to_ the open item~ identified in the inspection report..

• I

• )~~\\~

Gerald B Sl~de.6 *

General Manager tC Administrator, Regiori III, USNRC NRC Resident Inspector -

P~lisades Attachment APR. 6 lS92.

A OJttS /V7?Gr" CO/l*1PANY

• .

. I REPLY TO OPEN ITEMS

• Open Item 91019-01 3.1.J Electrical Load Study

.

.

The team evaluated the licensee's. study conducted in 1988 and i989 of the adequacy of station power systems to sqpply adequate voltage under worst-case loading conditions and identified the following concerns:

The study assumed a temperature of 1s*c for a71. cables No. 8 and

. sma71~r and 65°C for a17 caP1e.s No~ 6 and larger. *However, the cab.les in the plant are ra.ted for a maximum conductor temperature of 90°C. *

The cable resistance at this temperature will be greater than the resistance at the assumed temperature. *

The resistance and*reactance values used were based on Westinghouse* T&D*

Handbook, Table 6.

The reactance values.shown in this table are for a grounded.neutral system and paper Insulated cables.

These values are not applicable to _the type of cable used in this plant.

• . The impedance of circuit breaker cdntacts and.fus.es wer~ not*..

considered.

.

.

• :The loads considered were not the "worst-case loads," (i.e., an* niotors running).

• The licensee identified in Audit Report QA-91-06 that the impedance of the buried cables from the switchyard to safeguards transformer was

. incorrectly specified in the calculations. Audit results indicated that the actuaT impedance was approximately 30 times greater than the impedance values used in the calculations.

In response to the team's concern, the licensee submitted new data which demonstrate that the. effect of higher resistance on system voltage wcfs negligible, agreed to update the study in 1992.

This item remains open pending NRC review of the updated study {255191019-0l(DRS)).

Open Item

"The study.assumed a temperature of 75°C for a77 cables No. 8 *and sma77er and 65°C for a77 cables No. 6 and larger~ However, the cables in the plant are.

rated for ~ maximum conductor temperature of 9o*c.

The cable resistance at this temperature will be greater than the resistance at the assumed -

tempera.ture."

..

CPCo*Response

.

..

.

.

Power System Simulation/Electrical (PSS/E) loadflow simulations with cables assumed at 65°C and 75°C (original inodel), versus simuhtions with cables *.* *

a~sum~d at 90°C (added conservatism) were run fbr comparison p~rposes~ -The.

simulation~ indicated that corr~cting th~ cable resistanc~s for go*c resulted in volta:ge changes of only 0.01% difference on 2400V buses, 0.02% difference*

on 480V bus~s, O.OI% worst case on Class IE 2300 volt motors, and 0.2% worst case on 460V Class IE motor terminal ~oltages: *Thus, the impact of changing the cable resistances to 90°C has minimal impact on the loadflow simulations.

.

.

'.

,

'

.

.

The use of 6S°C and 75°C for cable resiStances for the loadflow model was*

based on initial data available for model development.

It is agreed-that, for conservatism, *the cable resistances should be c~rrected for 90°C for l~adflow modeling in.all future studies. The corrections, as was shown, wi)l have

minimal impact on the overall result.s of the *_lbadflow studies currently available.

Pr~sently, there are I82 power ~ables modeled in the PSS/E loadflow model.**

Final correction of the resistances.of these cables due to temperature will require an engineering analysiS and model update.

Corrective Action *

.

.

.

By the end of I993, correct the cable ~esistances to 90'C in the PSS/E l oadfl ow model* for the station auxiliary system.

Open Item

"The resistance and reactance values _used are based on Westingho.use T&D Handbook, Table 6.

The reactance values shown in this table are for grounded neutral system and paper insulat_ed cables.

These values are not applicable to the type of cables used in this Plant."

. *.

~

CPCO Response Table 6 of the Westinghouse T&D Handbook was chosen fo~ typical cable

. constants.

The use of this table or similar tables from other references only*

provides typical cable constants.

The use ~f typical cable reactance is standard industry practice when developing loadflow and. short circuit models*.

• of station power ~ystems.

The Palisades loadflow model for the~ worst case start-up 'transformer (SUT) 1-2 supply during bloc~ l~ading of the Emergency Co~~ Cooling *system (ECCS) motor loads was field verified using steady state and transient voltage measurements.

The model assumed cable resistances at 65°C and used the cable resistance and reactance constants from Table 6 of the Westinghouse T&D Handbook.

Field verification of the ~odel indicated that voltages-measured on the 2400V ahd 48bV buses ~ere within I.07% of the simulated values and meet

• .

..

the NRC recommended*accuracY of 3% as outlined in of Br~nch Te~hnicil Positjon PSB-1, "Adequacy of Station ElectriC Distribution Syste111 Voltages."

In fa.ct, the model predicted lower voltages on both 2400V and.480V buses thari the actual field measurem~nts. This fi.eld veri.fication documents the validity of using the cable resistance (at 6S°C) and reactance *v.alues in the model from* *

Table 6 ~f the 'Westinghouse T&D Handbook.

.

.

Although the*reactance values in Table 6 of the Westinghouse T&D Bciokare for a grounded neut~al system, balanced three phase conditions are assumed in the loadflow and short circuit models.

A balanced three phase system will have zero neutral/ground current. Therefore, the effects of neutral currents on*

cable reactance are not required and Table 6 is appJicable for determining.

typical cable reactarice for balanced three phase loadflow and short circuit models.

We agree that the determination of positive~ negative, ~nd zero sequence cable reactance for unbalanced conditions would re~uire additional calculations, and the use of Table 6 would apply only for syst~ms using a grounded neutral system.

With re~pect to Table 6 applytng to paper i~sulated cables, it wis used for determining typical resistances and reactance for copper cable. It was not used for determining ampacity requirements for power cables..

Corrective Action No*further corrective action is required~

Open Item

"The* impedance of circuit breaker contacts and fuses were not considered. II CPCO Respons*e It is not standard. industry practice *to include the impedance of circuit breaker contacts and fuses in plant auxiliary loadflow or short circuit

• models.

They are considered to be negligible when compared to station power

. transformer or power cable impedances. *However, we agree that the effects of the impedances of circuit breakers and fuses should be documented in the loadflow and short circuit analyses.

• Corrective Action By the end of. the second ~uarter of 1994, docu~ent the effetts of circuit

• breakers and fuse impedances on the overall circuit im~edance*in the loadflow and short circuit models of the station auxiliary system.

.

...

• Open Item

"The loads were not the "worst-c.ase loads/* (i.e~, a77 motor running).

CPCO Response ihe response to the loadflow analy~es not considerihg "all motors running" as stated in the FSAR has*been evaluated in a corrective action document,

O-PAL~91-104. The evaluation concluded that the loidflo~ analyses assumed

• realistic loads with required system load configurations.for the given operating condition.

The FSAR statement can be int~rpreted to imply that all motors connected to a given bus were assumed to b~ running, regardless of the plant operating condition. The corrective action from D-PAL~91-104 is to ali~n the FSAR with the present ~nalys~~ and ~ubmit_a request to change the

. wording ih the FSAR, Section 8.3.2.

In addition, a re~i~w of Chapter 8 of the FSAR will be performed to ensure that current loadflow analyses align with FSAR statements and commitments.

• corrective Action By the end of 1992, (1) submit a r.equest _to change the *wording in the FSAR,.

Section 8.3.2 arid (2) review Chapter 8 of the FSAR to ens!Jre that current

• loadflow analyse~ alig~ with FSAR statements and com~itments.

Open Item

"The licensee identifie_d in audit report QA.-*91-06 that the impedance of the buried cables from the switchyard to safeguards transformer was incorrectly

• specified in the calculations. Audit results indicated that the actual impedance was approximately, 30 times greater than the. impedance values used in*

• the calculations."

..

CPCO Response C~rrective action documents O-PAL~90~122 and D-QG-91-12 were issued concerning the incorrect* safeguards cable impedance.

Field testing and analyses to determine. the correct impedance have been completed as a part of the corrective actions. Additional corrective actions included: (1) performing and documenting the results of t~chnical specifications surveillance test (TSSP) QOI, "Safety Injection System," documented in EA-O-PAL-90~122-01, (2)

determining the safeguards cable impedance using TSSP QOl test results documented in EA-O-PAL-QG-91-12, (3) validating the Power.System*

Simulation/Electrical {PSS/E) Dynamic Model of Safeguards supply during TSSP *

QOl testing documented in EA-D-PAL-90-1228, and (4) developing recommended

. second level undervoltage relay transient time delay during TSSP QOl testing to avoid diesel generato~ starts, dbcumented in EA-D-PAL-90-122A.

.5 Corrective* Action By the end of 1993, update the loadflow and short circuit inodels with the updated safeguards cable *;mpedance.

This will be coordinated with.the overall update of cable resistanc~s to 90°C (loadfl~w modeling) or 25*c (short circui~

. mode 1 i ng)..

Open Item 91019-02 3.1.3 Overvoltaqe on Class 1E Equipment The team was concerned that plant operating procedures did not adequately direct corrective actions to b~ taken in the event of higher than 240ov*on the*

Cl ass 1£. buses caused by a stuck safeguards or start-up transformer tap* * *

changer.

This condition could result in Class 1£ motors being exposed to voltages higher than their rating.. The team also noted that no formal calculations were in place identifying the expected voltages on Class 1£ motor.**

terminals during conditions o.f a stuck tap changer concurrent with high system voltages..

In response to the team's concern, the licensee determined that the voltage on

• the 2400V.. Class 1£ buses should be maintained at less than 2530V to prevent exceeding the voltage limitations of the 2300V and 460V motors.*. Th.e licensee-

. a 1 so agreed to revise the appropriate procedures to identify operator actions to be t~ken to maint~in voltages below2530V during stuck tap changer conditions.

This item remains open pending NRC review of the revised procedures (255/91019-02(DRS. CPCo Response Maximum station power voltages have b~en doctimented in ~ngi~eering analysis EA-E-ELEC-VOLT-10/91-01, "Palisades Maximum and Minimum Statibn Pbwer

• voltag*es" during a maximum historical 345 kV system voltag of 369 kV with the unit offline. *The loadflow simulations assumed plant cold shutdown loads fed by Safeguards Transformer (SGT) 1-1 (normal automatic tap changes) or Start-up.*

Transformer (SUT) 1-2.

ND overvoltages were id~ntified on Class IE equipment i~ EA-E-ELEC-VOLT-10/91-01 for the worst case 345 kV system voltage of 369 kV~ CPCo maintains historical records of its 345 kV switchyard voltages and periodically.r~views worst case voltage profiles using loadflow models of the. plant itation power system.

Based dn current analyses, the addition* of overvoltage ~rotection or alarms is not required.

We agre~ that expected voltages on Cl~ss IE motor terminals during conditions of a stuck tap changer should be identified through additiorial calculations.

Furthermore, we agree that voltages on the 2400V Class IE buses should be * maintained les~ than 2530V to prevent exceeding voltage limitations on the 2300V and 460V motors.

To assure these voltages are maintained, appropriate oparating procedures will be revised.

• • corrective Action

. . . . . . . Perform computer an~lyses to document expected voltages o~ Cl~ss. IE equipmeni d~e tp a stuck tap changer.. *Procedures will. be revised to include: (a) a stuck tap chariger. on SGT 1-1; or b) operation via SUT 1-t. The procedtir~s wiJl identify ne~essary operator a,ctions to maintain voltages below 2'530V during these conditions. Completion of these items is scheduled for the erid

• of the second qua~ter 6f 1994.

Open Item 91019-03 *

3. J.5 * Overl oadinq of Buses JC and ID Feeder Cab 1 es from Start-Up Transformer. During review of t~e "Steady State and Transient Cable Ampacities for Buses JC, JD & JE, Palisades Plant;" dat~d December J988, the tea~ noted that the 500 MCH cables from start-up transformer J-2 to buses JC and ID were heavily overloaded during sma77 LOCA transients.

An operator action was required to.

reduce the load within 11 hours to avoid damage to the cables*. Based on this study, the licensee instituted administrative loading limits pending.

replacement of the ca67es.

A subsequent spec1:a1 ampacity study,* "SUT J-2 500 MCH, Buses JC and ID via.SUT J-2~" Revision 0, dated Sept~mber 2J, J99J, was performed ut i 1 izing a 105 *c emergency overload temperatur~. Thi.s study - resulted in the cance11ation of the cable replac.ement. and removal of the administrative. loading.limits.

• . The team was concerned.that the analysis had not quantified the cable rating in terms of total. a77owable time at cable temperatures beyond 90°C.

The* team pointed out that the cable could-be operated at J05 *c for up to JOO hours. only and not indefinitely-as assumed by the analysis.

The licensee agreed that additional studies were required to quantify the time for operating beyond the. * 9o*c rating of the cable and agreed to incorporate these Jim.its into

• • appropriate operating proce~ures.

This item remains open pending NRC".review of the additional *studies and procedure revisions (255/~1019-03(DRS)). CPCo Response The plant's nQrmal operating cohfiguratio~ is via Safeguards 1ransformer (SGT) 1-1 stipplying 2400V Buses IC, IO and IE simultaneously.

The original 500 MCM cables feeding Btises IC and IO from this supply were.replaced with IOOO MCM

• cables as a part of_ facility change *FC-800, "Addition of 2400V Offsite Power."

The 2400V source from Start-up Transfo*rmer (SUT) * 1-2 is provided as a backup.

• to the normal power source fr6m SGT I-I.

Engine~ring analysis EA~sc~sa-OI9-b03 was performed to:~etermine the *netessity of replacing the 500 MCM cables*

• . from SUT 1-2 to Buses IC and 10 and conC"l uded that this was unnecessary. This conclusion was based on wor.st case accident loadings and the IOS°C emergency
• overload temp~rature rating of the cable. *The analysts showed that the cable would exceed its 90°C continuo~s rating, but would remain below its IOS°C emergency rating for the assumed accident loadings.

Subsequent review of loading the cable to its e~ergency rating of 105°C indicated the overload must be limited to 100 hours per year not to exceed a total of 5 occurrences (reference IEEE S-135, IPCEA P-46-426).* Even though the accident load resulted in temperatures above 90°C, but below 105°C, furthet investigation of the allowable ti~e above ~ontinuou~ ratirig is needed.

We agtee that r re-evaluation of p~st-LOCA loading and the effects on

management of the post-LOCA ~esponse is required.

Furthermore, we agree that both the maxi~um time and maximum ampere cable limits should be determined and incorporated as ap~ropriate in operating procedures.

• Corrective Acti~n By* the end of the second quarfer of 1994, update the post.;.LOCA loading analyses on SUT 1-2 and determine maximum ti~e and ampere cable limit~ for the 500 MCM cable feeding buses lC and ID from SUT 1-2.

Incorporate these limits* .into appropriate operiting procedures.as requir~d.

• Open Ite~ 91019-04

. 3.1.6 Switchyard Station Power Transformer Cable The team questioned the abi li"ty of the feeder cable to* switchyard station*. power transformer No.* 2 to withs.tand postulated fault. currents. Switchyard * station power transformer No. 2 is.fed from 2400V bus JC through three single conductor cables.

The maximum fault current at the 7-oad terminals of the

• .circuit breaker is 30,900 Amps- (5 cycle value).

The team performed an informal calculation which questioned the cable's ability to withstand the fault current caused by a fault located at the breaker's terminal.

The calculations submitted by the licensee in response to the team's concern . confirmed that for the postulated fault, the cable would.exceed its damage* temperature threshold *in approximately 2.8 hertz.

The licensee immediately issued Deviation Report D-PAL-91-196 to.further analyze the concern.

This item remains open pending resolution of D-PAL-91-196 (255/91019-04(DRS))'. CPCo Response There are two No. 1;o*cables ~eviewed as a result of correcti~e action document O~PAL-91-I96. Cable No. I runs from 2400 volt Bus IC i~ its -0wn conduit to the underg~ound duct and out to the switchyard.

Cable No. 2 r~ns.

from 2400 volt Bus IE through a conduit to a tray, through the turbine building to IC switchgear where it enters a conduit to the ~nderground duct. a11d out to the switchyard. A detailed review ofthe routing of the two cables. . in question indicated they are not ~outed with any ~lass IE cable~. Therefore, no operability concerns were identified.

.

.. 8 The root cause of the N6. 1/0 cable being undersized for short ci~cuit con.ditions is attributed to a design error. This is evidenced by the fact that the current FSAR, as well as the or'igi nal FSAR, states "The design prevents the conductor temper.ature from exceeding 200.*c for rubber insu'lated , cables for the f~ult current available from the source."

An industry recognized cable expert was contacted to provide possible failure * modes of the cable in question.

The ~xpert evaluation concluded:

. . . . . -

The cable would have to replaced ~ftei the (highly unlikely) short circuit took place.

The insulation will self-destruct and cease to function.

• No explos*ion is expected,. and only sporadic snioking is possible.*

- * Mi nor, sporadic flame,'.if any, in the conduit and duct bank will die out for lack of oxygeri.

• .While the temperature involved is high, the time~ involved are extremely
• short. It is this short duration of time that helps to alleviate the situation.
• The other cabl~s in the tray will contiriue to ~e operable.

The jackets.** of the cables in the tray that are, in.contac~ with this cable may * experience some jacket damage.. *.... *

• The cables should be examiried after such an occurrence.

Cable No~ 1 from*2400V Bus IC is in its own.coriduit to the underground duct and out to. the switchyard.

No 'othe~ cables *are exposed to potential damage in the e~ent of a short circuit. This cable; however, needs to be addressed as* it does not meet th~ requirements of the FSAR.. Cable No. 2 needs to.be addressed s i nee it does not meet the requirements of the FSAR and is in the turbine building cable tray-with other non-IE cables for approximately 300.

feet.

• The recommended corrective action is to revise the FSAR to reflect the exception to the 200°C requir~ment for both cables and accept the small

~isk

• that if a short circuit occurs the No~ 1/0 cable will need to b~ replaced and, if the short.circuit occurs on cable No. 2, the other cables in the*tray will need to be inspected for possible jac~et damage.
• Corrective Actions:

1. Initiate a change to the.FSAR to make a~ exception to the 2oo*c requirement* for the two cables feeding switchyard power. This action is scheduled for completion by June 1992.

.,.

.. . ' 2~ The CPCo System Protection Departmen~ will review th~ 480V ~y~tem fo~

• similar situations. Thi.s review iS scheduled for completion by September 1992.

- Open Ite~ 91019-05 3.1.1 Short Circuit Study The teain was concerned that non-conservative values for system voltage and . cable temper~ture were used ii1 calculation EA-E-ELECT-FLT-10191-1, dated . October 28~ 1991, which was.prepared to determine fault duties on the 4160V, 2400V, and 480V AC systems.* Jn response to,the team.'s concern, the 1i~ensee provided an analysis that demonstrated that the voltage assumed.in the calcu1ati-on w*as conservative in rel.ation to the maximum historical s~itchyard

• voltage.* Jn addition, the licensee*provided an analysis that demonstrated that the non-conservative high cable temperatures.would not have a significant effect on the overall results of the calculation.

The licensee committed to revise the c~lculation to provide a clarification of voltage criteria and to reflect.conservative. cable.temperatures. Pending further NRC review, these issues are consiqered an open Item (255/91019-05(DRS)).

.CPCo Response We agree that.the voltage criteria issumed in the short circuit calcul~tion should be clarified.

In addition, we agree that for conservatism, the cables should be ~orrected for~2s*c for short tircuit modeling in all future studies.

The correct1ons, however, wi.ll have minimal impact on the overall result.s of the short circuit studies currently available.

No station power system problems due *to these changes* are.expected.. Presently, there are 182 power cables modeled in the Auxiliary.Sy~tem Design Optimization Program (ASDOP) * short circuit computer model and Power System Simulation/Electrical (PSS/E) loadflow model.. *

Corr*ect i ve Action By the end of 1993, update the cable resistance in the short circuit model to 2s*c and clarify the voltage criteria used in the short circuit analyses.

Open Item 91019-06 3.1.8. Operator Response to Ground Fault Indication The team determined that existing plant *procedures do not provide adequate guidance on how to i.dent ify the 1 ocat ion of a f au 1 ted bus and do not a 7 ert * * operators.of the dangers of operating two safety buses from the same source when a ground fault exists on the ungrounded 2400V system.

In case of a solid single line to ground fault on an ungro~nded system, the location pf the fault cannot be immediately determined by observing system instruments. Also,

although continued operation is possible, an overvoltage of 73% wi71 occur on the unf'aLilted phases and a second fault on the system could cause the

• .. simultaneous loss of redundant loads.

In response to. this concern, the.1ice~see agreed to re\\;ise their procedures to direct operators to transfer the 2400V buses one at a time to the alternate offsite source while observing the status of the ground fault relays and to

• supply the faulted bus.from a separate supply until the fault is located.

The licensee has committed to complete thes.e actions in the first quarter of 1992.

Pending further NRC review, this is an open item (255!91019-06{DRS)). CPCo Resoonse Palisades 2400V bus design is such thai electr1cal se~aration of 2400V buses is not possible without transfer of an affected bu.s to a separate off-sfte power supply.

The procedure of conce~n, Alarm Response Procedure (ARP) 3~ "Electrical Auxiliaries And Diesel Generator Scheme EK-05," was inadequate in that it failed to provide direction to operators to electrically se~arate 2400V bus~s durinQ efforts to locate the source of the ground fault.

• ARP 3 has been revised to direct operators to transfer 2400V buses one at a time from the normal to the standby off-site power supply in conjunction with
• using ground fault indication to dete~mine the affected bus.* -This -revisiori also provi~es direction that the groLlnded bus, once identified, is to remain on the standby power supply until the fault is cleared to eliminate the *

. possibility of sim~ltaneous:loss of redundant l-0ads. *

Corrective Action . - The corrective action ~or this open item has been completed.

Open Item 91019-07 3.1.9. Overvoltage on Ungrounded 2400V System The team was concerned that the 2400V electrical system, which was designed to be ungrounded, was susceptible to high voltage transients caused by intermittent-ground faults.

The 2400V electrical distribution.system, including safety buses lC and JD, is designed as an.ungrounded system in order.

to permit continued operation with a single ground fault on the system.

• However, ungrounded systems are susceptible to severe overvo7tages caused by repetiti.ve intermittent ground faults such as can be produced in a piece or vibrating equipment.

This phenomenon can _rapidly produ~e voltages five or six.

times norma 1 and may cause failures in motors connected to the system before * operators can intervene. Also, since the safety buses are interconnected through cables or buses when being supplied from. t~e primary or alternate

off site source, an overvoltage originating anywhere* in the 2400V system wi 11 appear on both safety buses~ Although the condition described here is considered to be of low probability, it is of concern because it reptesents a

- potential common mode failure mecham:sm.

Pending further NRC revie.w, this matter is considered an open item (255191019-0l(DRS)). CPCo Response

• The ungrounded 2400V system design used at Palisades is a standard industry design and i~ u~ed at other nuclear power plants.

The main advantage of an* ungrounded 2400V system is that it allows continued operation of critical Class I~ Emergency Power Supply (EPS) loads during a LOCA, coincident with a sihgle-line-to-gr~und fault.* The present overvoltage relays will sense the ground~fault and alert operators who will isolate th~ ground fault through manual breaker operations.

We.disagree with the concern tha~ the 2400V ungrounded system is susceptible.

to.severe overvoltages due to repetitive intermittent ground faults.

The phenomenon of intermittent arcing faults is generally confined to systems of 15 kV or higher employing o~erhead lines or cable runs with high capacitive.

reactance. Theoretically, voltages can rise to 5 to 6 per unit peak-to-peak.

However, this will occur only if sufficient capacitive reactance is *present during the ground fault to provide energy for restrike.

"Protective Relaying - Principles and Applications;" by J. Lewis Blackburn, provides some guidance . rel~ted to the susc~ptibility of a power system to.an intermittent ground fault due to the system capacitive reactance. Although it represents a wye. grounded pr:im'ary source, it does provide insight into the amount of*capacitfve * * reactance ~equired in~ low voltage power system to sustain high voltages to* ground due to an -0pen conductor;

where Xe K. * -> Xe Xe = the equivalent capacitive r~actance per phase.

Xe = the equivalent exciting reactance per phase K *> 40 for 1 i mi ting voltage on a_n open phase to 1. 25 The above ratio was checked *for con~itions at the 2400V Bus IC when fed from the Safeguards Transformer 1-1 which includes the worst case cable run of 2500 feet.

The ratio of Xe/Xe was estimated to be 26,247 which is w~11 above 40 for limiting voltage on an operi phase to 1.25 per unit. This indicates that the present 2400V system has very low capacitance to ground to sustain transient voltages and an intermittent arcing ground fault as proposed in the stated concern.

The application of the groun~ fault detect6r schemes on the*2400V buses provides a high impedance ground path. The *resistors are _used to reduce the shift of the neutral fbr either unbalariced excitation paths of the voltage transforme~s or from ferro-resonanc~ between the inductive r~actarice bf the voltage transformer~ and relajs in the capacitive syste~. Circu~t resistance will.introduce dampening of the transierit, reducing the peak value of the* voltage.

The resistance of the relays and associated resistors.~lso helps to limit the transient overvoltages.

• Corrective Action

. ( No corrective action is. required.

Open Item 91019-08 *

• 3.1.11 Retransfer of Bus to Preferred Source.

The team determined that operating procedures provide insufficient gu*idance regarding the potential adverse effects, during a LOCA concurrent with a loss of offsite power, of retransferring from onsite power (EDGs) to restored off site power.

When }oads are being supplied from the EDGs during a*

concurrent LOCA and loss of offsite power, *operating procedures require retransfer to the preferred source,. should it becom.e available again. However, when LOCA loads are applied to the offsite source, a voltage drop slightly larger than 2% can occur on the safety bus.

If the bus voltage_ is too-lo'.¥, this *additional drop could cause the second level undervoltage relays to drop* out causing the loads to be transferred back to the onsite source.

(A rough calculation performed by.the team indicated that an initial voltage greater than.94 pu would*be required to inaintain loads, using startup transformer 1-2 as the offsite source.)

In this cas~, it would be preferable to leave the bus connected to the onsite source.

In response to this concern, the licensee agreed to perform analyses.to establish minimum required voltages to enable su.ccessful re.transfer, and to revise procedures accordingly.

Pending further NRC review of the revised procedures, this matter is considered an open item (255191019-0B(DRS)). CPCo Response Current Palisades operating procedures provide guidanc~ to operators as to the minimum voltage required on the offsite ~ource to*maintain loads. This guid~nce has beeri incorporated into a revision to System Operating Procedure

• (SOP) 30, "Station Power.

The graph summarizing Startup Transformer l-2 load limits assures a minimum operating voltage of 0.94 per unit (2256 volt~) is maintained on the 2400V buses.

The synchronization of the diesel generator to - the 2400V bus is accomplished by adjusting the automati~ voltage regulator to the bperating bus voltage prior to;paralleling.

The transfer of the diesel generator loads back to the offsite source is expected to be a smooth transition. A pre-transfer voltage of 0.94 per unit is expect~d to be * adequate to avoid actuation of the unJervoltage relays at 0.91883 per unit.

However, analyses will be co~pleted to-determtne recommended 2400V voltages. . prior to transferring diesel generator loads back to an available offsite power supply.

Corrective Action By the end of the third quarter of 1994, perform *analyse.s which identifies

• minimum reqtiired 2400V voltages prior to transferring diesel g~nerate loads back to an available offsite supply and incorporate volt~ge limits into appropriate operating procedures.

-

Open Item 91019-09

• 3.1.12 Diesel Generator Steady State Loading Calculation -

The team was concerned that the magnitude, start time, and duration of manually started loads identified in EDG steady state loading calculation DRS-010990-1 may not reflect worst case conditions. This could result in the - applicatiOn of heavier loadings on the EDGs than were reflected in the talculation~ The loads ~valuated w~re based on manual operator acti~ns listed in the EOPs, and the timing criteria used was based ~n expected plant -

conditions and the author's experience. * The l~censee agreed to perform * additional studies to assess worst case contingencies and to revise operating procedures as appropriate.

In addition, the licensee could not provide evidence that calculation

• EA-P-:SA-8602 which was used_as an input to the_ EDG loading calculation, was i

prepared in accordance with accepted design control procedures.

Pending further NRC review-of the rev.ised procedures and studies, this issue remain open (255 91019-09(DRS)).. CPCo Response ** * . . Engineering analysis EA-E-PAL-89~011-01,-"Evaluate EOG Load Profile fo~ Seven Days After an Accident {Large Break LOCA)," was used as an input to.

Calcul~tion No. DRS-010990-1, "Diesel Generator Load Analysis."

In o~der to

• provide a realistic {but also conservative) diesel generator 16ad profile, DRS-010990-1.used the magnitude arid timing of manu~lly-actuated Emergency*

Power System (EPS) loads from EA-E-PAL-89-011-01, Which in turn was based upon Emergency Operating Protedures {EOPs) and Off-Normal Operating Procedures (ONPs).

A wo_rst case 1 oad profile was performed for each dies_el generator.

This occurs when the opposite diesel generator was not av~ilable.

Since the EOPs and ONPs specify manual equipment actuations based upon plant or equipment co_nditions rather than t_ime criteria, it was necessary to assume the timing cif those manual equipment actuations ~s defined in EA-E~PAL-89-011-01.

These assumptions were based on the judgmerH of an experienced individua.l.

There was no attempt to quantify variations.in diesel ge.neratO-r electrical loads which may result from an operator failing to follow the load profile.

defined in DRS-010990-1, since it was intended to represent a conservative.

load profile~ However,. the load profile was not intended to constrain or limit the ope~ator in emetgency or off-normal c~nditio~s. * . .. . . ..

W~ agree that additional st~dies of potential post-LOCA operating contingencies should be performed with the Operations Department to assure ~11 possible manual addition or removal of post-LOCA loads are considered in the diesel generator steady state 'loading 1 imits. Results of these studies will.

provide maximum flexibilfty within diesel generator li~its. Changes to appropriate. operating proc~dures will be incorporated as required.

Correctiv~ Action By the end of 1993, complete.additional.analyses of post-LOCA operating contingencies and iricorporate diesel generator ltiading limits into appropriate operating procedures as required.

Open *Item 91019-10 3.J.ij *Diesel Generai~r Trip Logi~ The team noted that the EOG control. sc.heme employs the fo77owfng automatic trip mechanismsthat do not require two or more independent meas.urements ~f.

• the trip par.ameter:
• Generator trip on underspeed (<600*RPM) through the Field Shutdown
• Timer.
• Engine. and generator trip on engine underspeed (<120 RPM).
• Engin~ and generator ~rip on j~cket water low pressure, start cir~uit 8 *
• . only.
• Engine and generator trip on generator overcurrent.

Since these trips do not employ coincident logic, th_ey increase the potential* for spurious*EDGtrips, which could cause loss of a division during an.

emergency.

' ' In addition, 'the non-coincident jacket water low pressure signal provides an unintended DG trip mechanism and can result in the engagement of the air start motors while the engine is* running... This could cause equipment damage and deplete the starting air supply.

The licensee had previously identified these conditions and has committed to correct them by the fourth quarter of 1992.. Pending further NRC preview of the licensee's corrective action, these issues remain open (255191019-JO(DRS)).

-..

CPCo Response During* ihe development of Design Ba~is Document 080-5.06, "Diesel Generator "Controls," a Discrepancy Report No. (ORN) F-CG-91-106, entitled "Emergency. Non:..Coincident/Non-Bypassed Trips," was,issued to identify and evaluate *the diesel generator automatic shutdown logic As many as four process or system conditions can cause automatic diesel generator shutdown under circumstances.

which are unacceptable according to the NRC!~ technical criteria (i.e., Br~nch Technical Position No. ICSB-17; superseded by Regulatory Guide (RG) 1.9). * Because this aspect of the diesel gene~ator control circuit design represented a potential conflict in the plant's licensing bases, a corrective action doctiment, D-PAL~91-160, was initiated to.evaluate the issue.

The ¢orrective action which resulted from D-PAL-9'1-160 was to modify the diesel generator* control circuits to meet RG 1.9. *

. . Also,* during the development of 080-5.06, O~N F-CG-917107 was issued to . identify and provide a basis for *the process or *equipment conditions which can** cause automatic shutdown (i.e., low jacket* water pressure and-engine underspeed [<120 rpm]). These items were never intend~d as shutdown 1ignals.

Rather; an apparent (original design) oversight, wherein the.diesel generator "engine started" signal was ~ot designed to ~eal-in, makes the unit susceptible to automatic shutdown from these ptocess signals; This occurs if the "engine started" permissive conditions fall belo~ the sensing instrument settings-at any time duririg operation, following a legitimate start.

Additionally, a related cb~dition was identified where the lack of the.s~al-in feature can cau~e the re-engagemerit of the air start motcir, with the engine. running, -for the ~uration of the overcrank trip timer setting.

As a result, the potential for air start ~otor damage or needless starting air depletion,

• or both, is introduced.

The occurrence of a diesel generator unit trip on "low jacket water pressur~" o~ "engine underspeed" can occur due to rion-coincident sigrials during

emergenty operation. These trip are not in confor~ance with the NRC's technical criteria. Modifications to the control circuit described in Deviation Report No. O-PAL-91~160 will resolve th~1 issue. * Corrective Action A modification to correct theie*diesel generator control circuit deficiencies .will be completed duting the 1993 refueling outage. * Open Item 91019-11 3.1.14 Engineered Safeguards Testing The team identified that Survei71ance Procedure RT-BC&D.requires that less than 50% of the equipment required to be load shed during a Safety Injection System (SIS) actuation, coincident wit_h loss of off site power (LOOP}, be verified to have properly shed.

Additiona77y, during the load sequencing

• .portion of the test, only a few selected loads were verified to autd start.

In the event of a SIS actuation coincident with a LOOP, failure of one or more

. '

. ..

major loads to shed could effeet the EDGs' abi7 ity to maintain voltage levels during load sequencing.

Also, there is a potential of a generator overload trip, as the DG overcurrent relay trip signal is not bypassed*during an,ESF actuation.

• '

',* The liCensee's* response_ to the team's concern was that survei71ance procedure RT-BC&D meets' the TS objective to demonstrate overall automatic operation of the emergency power system based on initial.construction testing which verified that each relay contact operated properly.

The TS wording a77ows. for the automatic starting of only "Selected.Motors and.

Equipment" (apparently original TS wording) is not consistent with emergency diesel testing as stated in the CombustJon Engin~ering Standard TS.* . . .. Palisades has committed t~ implement a restructured TS that contains a more conservative EOG testing requirement.

This item is unreso}ved pending NRC review of Palisades implementation of the restructured TS (255/91019-ll(DR5)). CPCo Response The first part of the.operi item states that technical specifications surveillance pr6c~dures (T&SPs) RT-SC and ~T-SO r~qu~re less than 50% of the

• equtpment ~equired t6 be load shed be verified to h~ve ~roperly shed.
• However, Technical Specification 4.7.1.b states, "A test shall be condu~ted

.during each refueling 6utage to demonstrate the overall automatic op~ration of.

the emergency power sy~tem."

RT-SC and RT-SO verify' that at least one load controlled by.eac~ load shed* rel~y is properly shed~ Using t~is method, the logic of the scheme is verified and the overall automatic operati6n of the emergency power system is demonstrated.*

. . . The s~cond p~rt of the open item states.that du~ing the load seque~cing portion of TSSPs RT-SC and RT-SO only a few $elected loads were verified to automatically start. This is an incorrect statement.

Both procedures verify -*automatic start for every Safety lnjection*.system (SIS) actuated load* and also verify timing for every Design Basis Accident (OBA) sequ*encer actuated load. *. Corrective Action Prior to the 1993 refueling outage, TSSPs RT-SC and RT-SD will b.e reviewed and revised t6 ~emonstrate; to the maximum extent possible, load shedding of th~ e~ergency loads.

'*

Open Item 91019-17 . 3.2.1 Engineering O~awinqs The team noted various minor disc~epancies between single line diagram £-8, Sheets 1 and 2," and other relevant engineering documents.* Examples of observed di~crepancies included:.

Circuit breakers72-18 and 72-28 were shown on diagram E-8, Sheet.1 as

• having thermal and magnetic trips whereas FSAR* and coordination curves correctly indicated that there were only thermal trips in these.

breakers.

• The feeders to panels Dll-1, Dll-2, 021-1 and 021-2 were shown on

.. diagram E-8, Sheet 1 as 2#410 (le/pole) whereas ca-lCulations 011/SC and D21/SC c?rrectly indicated 2x~#4/0 (2c/pole).

• Main single line diagram f-1, Sheet 1 did not show the 1200.A fuses in series with the 500 A breakers on the feeders to de distributio.n panels DlO and 020, whereas diagram £-81. Sheet 1 correct 1 i showed the 1200 A fuses.*

. - . .

. . . The licensee agreed to revise and update* al 1 relevant drawings.* This item * . remains open pending NRC review of the Ucensee's corrective action .(255191019-ll(DRS)). -CPCo Response '* In re~ponse to the above ope~ item, a document chan~~ reque~t (OCR) 950-91-1194 was initiated to torrect plant electrical drawing E-8, Sheets 1 and. 2.

Electric~l Draw~ng E-1, Sheet 1, will also be reviewed ~nd cortected as* necessar,y.

Corrective Action In addition to the above action, a review of plant electrical single line drawings and single line meter and relay drawings will be performed to ensure* they contain the proper level of detail and consistency~ This corrective action is scheduled for completion by the end of 1993.

The review of electrical drawing E-1 will be completed by April 30, 1992.. Open Item 91019~18 3.2... 2 Cable Selection and Sizing Criterfa.

The team was concerned that the licensee's voltage drop and short circuit calculations developed fdr sizfng J25Vdc and 125Vac cables did not consider worst case temperature conditions.

Lr..isting design documentation did not identify important cable data such as cable resistance and temperature

• 18 -

ratings.

The licensee stated that for cables larger than 8 AWG, the. . temperature used. in the.voltage drop e,alculations was 65°C and for cables 8 AWG and.sma77er the temper(lture used.was 75°C.

The team.determined that for the XLPE and EPR cabl,es, a non-conservative value of 90°C was used. . In . addition, the team rioted that the Jicensee used a value of 30-.C in short circuit calculations instead of a more conservative value of 2s*c.

Fina77y, * the team noted thBt AC reiistance values were used in the short cirtuit calculations for be circuits resulting in lower than a~tual calculated short circuit currents.. The licensee is currently updating.calculations and agreed .to use cable resistances at 25°C.for short circuit calculations and cable

resistances at rated temperature for voltage drop calculations.

This item remains open pending NRC review of.the.-updated calculations (255/91019-18(DRS)). CPCo R*esponse We agree that the voltage drop and short circ~it calculations for sizing 125Vd~ cables should consider worst case temperatLlre conditions.

Calculations for the de system will* be updated using cable resistances at 25°C for short circuit and cable.resistances at rated te~perature for voltage drop calculations.

• Corrective Action By the end of the second quarter of J 994, update the de voltage drop and short cirtuit calculations with cable resistances at rated cable temperature

{voltage drop) and at 25~C (sho~t circuit).

Open Item 91019-19 3.2j e*attery Charger -Input and Output Cab 1 es The team was concerned that the battery charger's input and output cables were inadequately sized.

The input and output rating of the chargers wer~ 90 A and- . 200 A respectively, and the cabl.es used were 2 AWG and 410 rated 120 A and 253 A respectively.

Applying the derating factor of 0.7 used by the licensee in

• their calculations, the cables should not have been used for currents more than 84 A and 177 A respectively.. The licensee re-evaluated the sizing of the battery charger on September 11,. 1989, (Deviation Reports D-PAL-89-148 and-149) and concluded that the existing cables w.ere acceptable.

However, the team's informal calculations indicated that the cables were undersized for the battery charger's rating.* However,* since the battery chargers are not * operating at their fu77 rated capacity, the team had no imme.diate operability concerns for the cable.

The licensee's response to the team's concern*was that these cables would be included in their cabl*e tray ampacity study plan, a program in progress to be completed during the fourth quarter of 1993.

This issue remains open pending NRC of the results of the ampacity study for the battery chargers (255/91019-19(DRS)).

• • --**

CPCo Response* Work continues on the cable* tray amp.acity study.

Corrective Action* The cable tray ampacity ~tudy is scheduled for completion du~ing the fourth quarter of 1993.

Open* Item 91019~22 . .. 3.3.1 Diesel Engine Support-Systems 3.3.1.l Fuei Oi7 Supply System . . The team identified the following discrepancies in the design documentation associated with the EDG fuel oil storage tanks: . . . . . . ~

• Fuel consumptiont~sts were not documented;

. . 19...

The calculations regardf.ng the capacities* of the EDG day*tanks and be"71y tanks were inconsistent.

• e. *The low level day tank alarm setpoint did not provide an accurate tank.

inventory..

The UFSAR, TS and various engi-neering _analyses stat.ed different EOG * running time capabilities.

The.team noted that the day tank emergency supply lines and their external valves DE-115 and 116, were not included in a maintenance and testing program to assure their availability at a77 times.

The supply lines provide . :... compensatior:i for the fact that storage tank T-10 and its appurtenances are n*ot.

seisrilica77y or tornado qualified. *The team was also *concerned that the TS

• required minimum 16,000 ga77ons of fuel in storage tank T-10 would not assure 7 days of dedicated EDG fuel supply.

The licensee currently maintains tank levels above the TS minimum to satisfy calculated 7 day fuel supply requirements.

The licensee's responses to the team's concerns committed to evaluate and.provide necessary corrective action by the fourth quarter of 1992. This item remains open pending NRC review of the-licensee's corrective action (255/9)019-22(DRS)).

CPCo Response During the inspection, the NRC noted a number of discrepancies in the . calculations and documentation (i.e., FSAR and the technical specifications) associated with the diesel fuel oil storage analyses.

.,

20. The issue regarding the current Technical Specifitatibn 3.7.li li~it qn*diesel generator fuel oil supply was addressed by CPCo in 1989 as a result of the Cbnfiguration Control Project.

We determined that the technical specification limit of 16,000 gallons of diesel fuel oil in.T-10 would not support 7.day~ of diesel g~nerator operation. This is documented in LER 89-005, dated March 3,* 1989._ Cor~ective action for that event i~tluded performing ~ calculation to determine the actual diesel fuel oil storage capacity.

As a result, a technical specifications change request was submitted on November JI, 1989 to . increase the diesel fuel oil storage capacity requirement.

We are currently awaiting NRC approval of this technical specification change request.

In the i n~erfm, the more restrictive dies.el fue 1 oi 1 storage requirements are being administratively controlled via Standing.Order 54, Section 3.7.li and will serve to insure an adequate supply of di~sel fuel oil is maintained.

The * remaining.i~sues conterning diesel fuel oil storage arialysis are iurrently b~ing evaluated.

The NRCalso.identified a weakness in the calculations for the diesel* fuel oil sto~age capacity in that fuel oil consumption ~ates utilized in the

calculations had not been verified by testing.

CPCo had utilized the ori~inal diesel generator fuel consumption rate (obtained by te.st* at the factory) in our calculations; Data obtained during ~he performance of technical specification surveillance.~est (TSSP) M0-7A, "Emergency Diesel Generato~s," indicated the existing fuel consumption rates were nearly equal the factory test data.. A more formal test needs to be performed.

The NRC team* also identified a weakness concerning the emergency-fill lines to the diesel generator day tanks in that they are not included in the. * maintenance. and testing programs.

CPCo will develop a.maintenance and te~ting program commensurate with the significanc~ and probability of the failufe of the emerge,ncy fill lines.

Correttive Action The remaining iss~es ~oncerning diesel fuel oil. storage ariilysis are ~urrently being evaluated and wil 1 be comp 1 eted by the end of 1992.

Recommended corrective actions resulting from this evaluation will be documented and appropriate completio~ dates will be assigned.

The formal test to determine the diesel. generator *fuel oil consumption rate as well as the maintenance and testing program for the emergency fill lines to the diesel generator day tanks will be developed by the end of 1992.

• Open Item 91019-23 3.3.J.2 Diesel Gener*ator Room Heating, Ventilation*. and Air Conditioning The team questioned the ability of each EDG room heating. ventilation and air conditioning (HVAC} system to maintain the ambient air temperature below 104°F with only one of two fans fed by Class 1£ power, considering all heat sources

. in the room, and the design.maximum intake air temperature.of 95°F.

The* . information provided by the licensee did not provide confidence that fans

...

• V-24A (K-6A) or V-24C (K-68) would be able to provide adequate ventilation.

Most of the team's concerns had been previously identified by the licensee who retained the services of Bechtel Corporation for the preparation *of* an analysis* demonstrattng the capacity of the existing system.. After the completion of this analysis,.appropriate corrective. actions to resolve the concern wi71 be performed by the licensee. This issue* remains.open pending NRC review of the analysis and corrective action (255/91019-23(DRS)). . cPCo Response As indicated in the NRC discussion of this open item, CPCo had identified this issue ~rior to the EDSFI inspecti6n and had initiated crirrective action to address diesel generatof room HVAC capacity.

Corrective Action. CPCo has* received the results of Bechtel's analysis and is currently in the protess of reviewing.those r~sult$. Our initial review indicates that the . existing configur~tion i~ not adequate to maintain room temperature below 104"F ~hen the outdoor te~perature is greater than or equal tri 95'F.

As a result,.weare developing corrective*actions to ensure that the diesel generator room HVAC capacity is a.dequate prior to periods of high summer temperature. This wor:k will be completed by the summer of 1992. * Open Item 91019-24 3.3.1.3 Emergency Diesel Generator Air Intake and Exhaust During system walkdowns,.the team found the EDG exhaust mufflers µnbolted from .their pedestals.. The nuts had been removed and the ~nds of the bolts flame cut to prevent the reinsta11ation* of the nuts.

This raised a concern . regarding.the ability of the exhaust system to function after an earthquake.

The licensee expiained that the mufflers were left unbolted during preparations testing in order to accommodate thermal expansion of the exhaust piping.

No formal modifications documentation was available for review.

The licensee committed to include the EDG exhaust system in its Seismic* Verification.Project under the auspices of the Seismic Qualification Utility Group (SQUG), and to formally document the ~nchoring design of the mufflers * for both seismic loadings and thermal expansion during the first quarter of 1992.

This item remains open pending NRC review of the design documentation (255191019-24 (DRS)). . . CPCo Response We agree that a concern exists regardiilg the ability of the exhaust system to function after an earthquake.

Engineering analyses EA-SP-07003-01, . EA-SP-07003-02, and EA~IO-SP-07003-01 were performed on the diesel generator

.. * . i,

. ~.

exhaust mutflers to addiess seismic arid anchorage concerns and take into account thermal expansion.

The eng~neering analy~es toncluded that the piping meets the.Interim Operability Crjteria of CPCo Specification_M-195(Q), R~v 2, * but does not meet CPCo FSAR Design tlass II requirements.

This issue is docu~ented in a corrective action document, D-PAL-92~096.. Corrective Action We are currently developing modifications to the diesel generator.exhaust piping to meet the requirements of CPCO Design Class II. These modifications are scheduled for completion prior to the end of the 1992 refueling outage.

Open Item 91019-25 3!3.L4 Emergency Diesel Generator Starting The team noted that the EDGs have never been tested to demonstrate their ability to start at minimum *hot standby conditions as specified by the. manufacturer (i.e.; 90°F lube oil and jacket water temperature and 65°F rooin temperature).. EDG monthly testing does not verify these parameters prior to startup. *The licensee committed to test start the EDGs under these temperature conditions by the end of.the next refueling outage. The possibility that jacket water and room temperatures could fa77 below the minimum hot standbj ionditions w~s considered signifitant since these parameters are not under automatic alarm survei77ance.

The licensee's response to the team's concern committed to evaluate methods to assure that

• these temperatures do not fa77 below design temperatures, including modification or procedure revisions, as required.* This item remains open pending NRC review of the licensee s corrective action (255/91019-25(DRS)).

CPCo Response The diesel generator vendor has ~uppli~d information which indi~ates that the minimum ambient.temperature which a diesel generator can be expected to start without starting aids (eg warmed lube oil and jacket water) is S0°F.

However, it is not clear that this criteria applies to nuclear plant standby service (critically timed 10 second start).

The system operating procedure for the diesel gehe.rators, SOP-22, "Emergency Diesel Generators," states that.the lube oil and jackei water temperatures must be maintained above 90°F to insure diesel generator operability. The original purchase specification for the die~el generator states that the *

• . minimum expected ambient room temperature would be 60°F.

As indicated in the inspection rep6rt, and in our responses to cbncerns of the EDSFI inspection team, CPCo has not performed a documented test to qualify or validate diesel generator performance at these minimum temperature limits. Also, as indicated in the inspection report, low jacket water temperature and low room temperature are not alarmed parameters.

...

• -

CPCo believes that frequent testing at t~e mini~um temperatures is not necessary nor prudent in that frequent cold-fast starts have been.demonstrated t6 result in pre~ature diesel generatbr wear and failure.

Thi~ issue has been

• well documented in the industry and specifically in NRC Generic Letter 84,-15

. "Proposed Staff Actions to Improve and Maintain Diesel Generator Reliability.~ However, we will perform a test at th.e minimum-hot standby conditions on a one time basis as a "qualification" test for the diesel generators.

Corrective Action. . The following actions wil.l b~ ctimpleted.to resolve this issue.

• I. Perform a o~e ti~e "qualif~cation~ test to demonstrate that:an EOG is
• capable of adequately-st~rting at or below the following conditicin~~

Cold lube oil temperature s 90°F Cold jacket w~ter temperature ~. 90°F Room a~bient temperature s 60°P CPCo originally committed (in our responses to concern$ during the EOSF.I inspection) to perform this test by the end of' the 1~92 refueling o~tage; howeve~, d~e to work priorities ~e were not able to develop ahd schedule an adequate test to support performance durtng the 1992 refueling outage.

This testing wi 11 n.ow be comp*l eted by the end of the 1993 refueling outage. ** ~- The operator rounds sheet will be revised to include recording of the.. diesel generator jacket water temperature and the diesel generator room

• temperature. Action limits and recommended corrective actions will be specified.. These revisions will be completed by May 1, 1992.

3. An evaluation_will be performed t6 determine the need to pro~ide an automatic alarm on low jacket water and low room temperature conditions.

This evaluation will be completed by December 31 ~ 1992 and any

recommended modifications will be scheduled through the plant plann*ing * - process.

Open Item 91019-26 3.3.2 Limiting Conditions of Operation and Maintenance for Emergency Diesel Generators The team. was concerned that plant *procedures or policies neither prohibit nor control work in the switchyard or* on redundant systems when one EDG is

.inoperable because of maintenance or testing. Unnecessary risk of loss of offsite power should be clearly eliminated by procedures*when only one EOG is operable.* An incident of this nature is described in NRC Information Notice 91-34.

The licensee agreed to incorporate into plant administrative procedures the guidance to assure that testing or maintenance is avoided which

.- _.. . * . * ... has reasonable potential to affect redundant eqµipment.

This issue remains open pending NRC review of the* procedure and po 7i cy revfs ions (255191019-26{DRS) ). * . CPCo Response We agree that additional guidance regarding limiting conditions of operation and maintenance for emergency.diesel generators is desirable.

We will incorporate this additional guidance into.plant admiriistrative procedures~

The revised p~ocedures will serve* to ass~re that testing or maintenance, ~hich has reasonable potentfal to affect red~ndant equipment, is avoided. * Corrective Action Administrativ~ procedures will be revised in accordance with bur procedure biennial review process. This action will be completed by the third quarter. . of 1993.

}}