ML20198K445

From kanterella
Revision as of 03:56, 21 November 2020 by StriderTol (talk | contribs) (StriderTol Bot insert)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
Reactor Bldg Cooling Fan Logic Mod Failure Modes & Effects Analysis, Rev 0
ML20198K445
Person / Time
Site: Crystal River Duke Energy icon.png
Issue date: 11/30/1997
From:
MPR ASSOCIATES, INC.
To:
Shared Package
ML20198K429 List:
References
MPR-1887, MPR-1887-R, MPR-1887-R00, NUDOCS 9801140419
Download: ML20198K445 (22)


Text

, . . . . . - . . . . . . , _. - - _ . . . . . . _ . _ - - _ . ~ . . . - - . _ . . . - .. ...

i I

WMPR ASSOCI Af f 8 thC fYOiNff48 '

l Crystal River 3 Reactor Building Cooling Fan j Logic Modification Failure klodes and Effects Analysis  :

l t

i MPR 1887 Revision 0 November 1997 i

Principal contributors

, Douglas Hill James Moroney QUALITY ASSU3ANCE DOCUMENT This document has been prepared, Teviewed, and approved in accordance with the Quality Assurance requirements of 10CFP50 Appen x B, as specified in the MPR Ouality Assurance Manual, Prepared by . /d b/d[/44 36 A/#P k Reviewed by Y Y lalarN OurW o?CNovyP- '

Approved by

>7- +4 As,2 w<u e w " w n2 Prepared for t Florida Power Corporation i

Crystal River Nuclear Power Plant 15760 W. Power Line Street Crystal River, FL 344284708 -

'O g

  • Glenn Ward g11gog g9 g E - Engineer p PDR i

- 3 O KING $1Ritt AL(XANO#iA. VA ~ 22314-3230 ' 703 619 4200 fax; 703 619-0274 g

j MMPR AllOCI Af t $ IldC I &NG1N&LR%

CONTENTS Pace Section EXECUTIVE SUhih1ARY 11 1

2 INTRODUCrlON AND CONCLUSIONS 21 Ilackground 21 Technical Approach 21 Scope lloundary 22 2-2 Con:lusions SYSTEhi/hiODIFICATION DESCRIPTION 31 3

4 COhiPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS 41 REFERENCES 51 5

O li

t WMPR

~

l ysoceAtas suc snoenstns Section 1 ,

EXECUTIVE

SUMMARY

l His report documents a Failure Modes and Effects Analysis (FMEA) of C.ystal River .

i Unit 3 (CR3) Reactor Building Cooling Fan Logic Modifications. This FMEA provides a qualitative assessment of the effects of failures of the various components that are part of the fan control logic modificatica detailed in MAR 97 09-05-01. The objective of the FMEA is to determine whether the design satisfies single failure criteria in accordance with the Crystal River Unit 3 Topical Design Basis Document for the Single Failure Criteria. i The rope of this FMEA is limited to those components affected by the modification and 1 i those that interact with the modified components.

The modified logic design is evaluated for its ability to accomplish the required functhn of l ;

preventing two Reactor Building (RB) cooling fans from running simultaneously following a Large Break Loss Of Coolant Accident (LBLOCA). Component level evaluations are i i

then performed for the various circuit components. Failures are postulated and their effects on the capability of the c:rcuits to perfoim their post accident design functions are determined and evaluated. .

The FMEA shows that there are no credible failures for the components in the RB cooling fan control circuits, as modified by MAR 97 09-05 01, that would result in two fans that  ;

could continue to run at the same time following the initiation of an Engineered Safeguards  ;

Actuation System (ESAS) signal. Single failures which result in no fans running are j '

discussed in Section 4, Note 1.

i

)

,f O  ;

11 y

. . , _ - . . _ - , , , . - . - , . . . _ _..---...,,m.-._..,-.-..=,. .. . . , - . _ .

I MMPR unocim s c.

V5 Te~rIIE R s Section 2 [

INTRODUCTION AND CONCLUSIONS BACKGROUND Recent analyses associated with Restart Issue D 28 (Error in Design Calculations for SW System IIcat Loads) have shown that the Scavice Water (SW) system can become .

overloaded due to excessive heat removal from the Reactor Building (RB) following an l LBLOCA. The RB heat removalis via the RB fan coolers Previous unalyses (including the original SW system design and sizing studies) assumed that the RB fan coolers operated in the worst casc degraded (fouled) condition and that fan coil cooling water flow rate was at the minimum design point (i.e., one of the two emergency SW pumps in service at minimum acceptable performance). However, the RB fan coolers could actually be in a clean, non degraded condition, and the two emergency SW pumps could be in service post-LOCA (the normal condition since both receive a start signal from ESAS). These assumptions result in a considerably greater heat transfer rate from the RB atmosphere via ,

O the RB fan coolers than pieviously evalutded. The SW heat exchangers may not be able to malttain design SW temperatures while transferring heat to the ultimate heat sink (UHS) through the Raw Water (RW) system at high UHS temperatures with two RB fans in operation post LOCA. The resultin6 higher temperature of the SW cooling water would exceed the analyzed cooling water temperature for SW system cooled equipmeat.

MAR 97 09 05-01 modifies the existing electrical controllogic for the RB fans to automatically start one RB fan following an ESAS actuation with the capability to automatically start another fan should the primary fan fail.

TECHNICAL APPROACil The FMEA was performed to confirm that the modified logic for the RB fan coolers meets single failure criteria in accordance with the CR3 Topical Design Basis Document for The Single Failure Criteria. This was accomplished in two phases. First, a system level evaluation was performed to verify that the changes accomplish the goal of ensuring that only one RB fan cooler will start following an ESAS s!gnal and that single failure considerations (train separation, train isolation, circuit protection, redundancy, ete ) were addressed in the modified circuit design. Then a component level evaluation of the ,

potential failure modes of the various circuit components and their effects was performed.

- These combine to confirm the integrity of the modified circuit design.

V 2-1

-9,., ,<.-t . , -----; y w r- t -' + - '~r e '

  • t' --
  • c Q Specifically, the purpose of ti.e Fh'EA is to provide assurance that the RB fan coole ts' circuit design, revised in accordance with MAR 97 09 05-01, will ensure that only one RB fan cooler will start following receipt of an ESAS signal and that no system or component level single failure will result in two fana continuing to run simultaneously.

SCOPE IlOUNDARY The scope of the FMEA is limited to ti ose components directly affected by the modifications in MAR 97 09 05-01. Items that interface with these components, such as, the fan rotor, fan motor, and motor control compartment Ocvices are also considered. The items included in the FMEA acope are listed in Section 4 of this rcj art. The general application of separation crheria for individual wiring, components, and raceway is not evaluated in detail. The FMEA assumes that the CR3 separation criteria have been properly implemented as defined in the MAR drawings and installation instructions and in accordance with other sta' ion drawings, proccuures, and practices.

The FMEA only addresses failure modes and effects following an ESAS signal. During normal power operation, the requirements for the RB Cooling Fans and responses to fan failures are covered by T chnical Specification Since the new relay logic is only utilized on receipt of individual ESAS signals for each fan, and interlock wiring is isolated by fan trains, a single failure of a control cr power component for one fan cannot cause spurious q

V operation of or Jisable another fan.

The impact of this modification on controlioom operator response to the various potential failures is considered to be encompassed by the existing operating procedures. The change from automatic start of two fans to automatic start of one fan following ESAS actuation is encompassed by the cxisting cast of one fan failing to start. The change in design that has the standby fan automatically staning if the preferred fan fails to start or stops after starting is also encompassed by the existing case of one fan failing to stan. In either case, the current design allows post LOCA Reactcr Building heat rem al using two RB fans, one RB fan and one Building Spray (BS) pump, or two BS pumps @eference 4). It is assunied that appropriate procedures are being changed as part of the modification process.

The failure mode! and effects detailed in Section 4 are applicable to all three fans, unless otherwise noted.

CONCl,USIONS The results of the FMEA (Section 4) have identified no credible single failures for the components in the modified RB fan coolers logic circuit that would result in two fans continuing to run at the same time following an ESAS actuation.

g t j The FMEA has identified one possible failure where a fan motor starter contactor could stick in the " hot" (fan running in slow speed) position following a trip signal and where the 2-2

i l

(l associated auxiliary relay (limit switch)" reads" the movement of the mechanicallinkage as i V having tripped. This could result in two fans running simultaneously. The results of the FMEA do not cons l der this credible. This failure is addressed in detail in Section 4.

The.*e are also several credible failures that could prevent the starting of either RB cooling fan upon an ESAS actuation,i.e., result in no RB cooling fan running. The effect of these failures does not result in operation outside of the design basis because other means to cool atid thereby reduce pressure in the RI3 (e.g., via the Building Spray pumps) would be available. Tliese failures an discussed in Section 4.

(3 V

O (O

2-3

[

EEMPR A S S OCI All S IN C.

tNGiN4iR$

Section 3 SYSTEM / MODIFICATION DESCRIPTION There are three Reactor Building (RB) cooling fans (AHF 1A, AHF 1B, and AHF 1C). In normal operation, two fans are running in fast speed. In the event of an initiation of ESAS, the fans will trip from fast speed. In the current configuration, the two fans would then be automa!!cally started in slow speed, based on the ESAS Block leading sequence. Under the proposed modification, only one fan will be started in slow speed, with a second f an as "back up." The back up fan will start automatically if the first fan does not start, or if it fails during operation. Due to Emergency Diesel Generator (EDO) loading considerations, the circuit modifications have been designed to make the A fan the preferred fan and the B fan the back up (the A diesel has more margin than the B).

Design modification MAR 97 09-05 01 mstalls interlocks to prevent more than one Reactor Building (RD) Cooling fan from operating at the same time on initiation of an ESAS signal, p This modification is intended to prevent excessive heat transfer to the SW system in the

'Q case of a Large Break Loss Of Coolant Accident (LBLOCA).

As currently configured, AHF 1A and AHF 1B are powered from ESAS Trains A (MCC 3A2) and B (MCC 3B3), respectively. The C fan is powered from the AB Motor Control Center (3AB), and can be manually selected to substitute for either the A or B fan.

MAR 97 09-05-01 modifies the fan control circuit so that ESAS signals to the fans " enable" a relay logic circuit which is designed to ensure that on!y one of the selected fans will b; running, on low speed. As long as therc is a standing ESAS sional, only one fan will run. If the running fan stops, the alternate fan will start. Reset of t.n dSAS signalleaves one fan running and " disables" the relay logic governing one fan operation, restoring normal manualcontrolof the fans.

The modification adds three relays to the A and B fan circuits and five relays to the C fan circuit. There are two timing relays (one to energize and one to de-energize) and one interlocking auxiliary relay being added to the A and B fans. Since the C fan mimics the circuitry for both the A and B fans, four timing relays are added te its circuitry. The auxiliary interlock relays receive their actuatior. directly from existing auxiliary contacts on both the FAST and SLOW speed contactors. The A fan time delay energize relays are set at a minimum setting of 1.5 seconds. The B fan time delay energize relays are set at a nominal 6 seconds to allow the A fan sufficient time to start first. The time delay m de-energize relays are set at 2 seconds to ensure against a potential relay race when ESAS Q is reset. The new logic is biased to select and run the A fan at slow speed on ES actuation.

31

i His is based on the assumption that both ESAS signals come in at the same time, if the B  ;

O actuation precedes the A actuation by more than six seconds, the B fan will start on slow speed first. Preliminary evaluation as part of the MAR process has determined that both .

E!X)'s have sufficient margin; MAR open item 1.b requires formal verification. Note: The  !

C fan has identical logic to both the A and the B fans, so that if the C fan was A selected, it would become the preferred fan and would be logically biased to start first as if it were the -[

A fan. If the C fan was B selected, it would become the back up fan and would be logically ,

biased to start only if the A fan failed to start. 1 A number of steps have been taken to assure that separation is maintained. _ Equipment {

boxes are being installed on the outside of the respective MCCs to house the new relays,  :

terminal strips, and associated wiring for each of the fans. Dese boxes are sized to allow sufficient separation between wiring and components from different electrical trains. The [

conductors which provide the power supply for the new interlock auxiliary relays will be  !

isolated from other wiring by routing through separate conduits and by using "Siltemp"  !

- wrap, which has been qualified and specified for use at CR3 as a separation barrier. . .

t No walkdowns were performed as part of this report. The physical equipment and wiring changes contained in the MAR and drawings contained in the Square D and Allen Bradley vendor manuals were reviewed. j i

i f

O  :

3 2-1

-.. - . [ .h e , , _,,.v., , - . . ~ . . , .., .. m... 4 , . . . , . . -.__,_..,_,.__~.,me

.__',,,4,,_,.'.

WMPR

  • *JREte!15JNc O iNG4 NEE 34S Section 4 COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS ,

O

{

O 4-1 i

i I. -__ ., -- ---

m m '

O Os COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS I C_-, s: Fa h M N w SW M Idestaficatese

1. Control Circuit a. Fail Open (Fast Fast speed schematic wiring fail open- Fan will not start /run l Wiring Speed, Slow Speed, in fast speed (logic is energize to start). His is aceptable Interlock) because ESAS actuation trips fast speed.

l Slow speed schematic, siring fail opere Fan with failed wiring will not start /run in slow speed; interlock relay not esu@d l enables second fan to run in slow speed on standing E.S.

Signal.

i I

New interlock wiring between fans fails open: Fan with failed Interlod wiring and devices are interlock will not run, enabling second fan to start. isolated from schematic wiring.

Shorts / failures in schematic wumg cannot affect interlock

$ wiring.

b. Short to Ground Control wiring short to ground: Control power fuse for Interlockwiring is rur. in affected fan opens; second fan is enabled to start. separate conduit and has coil to coilisolation.

I

c. Short to Another A single hot short to another conductor in the start schematic Interlockwiringisisolated so j Conductor wiring could cause spurious operation of the fan in fast or that failures in the start slow speeds, shortmg of the power conductors by actuation of schematicwiring cannot affect the high speed " shorting" contactor while in slow speed, or interlock wiring,and vice versa.

spurious actuation of the interlocking relay. Spurious operation of a fan in fast or slow speed will cause the al?rnate fan to stop through the interlocking relays. Shortmg the power conductors will cause catastrophic failure of the MCC wmy.ihnent or trip the Motor Circuit Protectsr MCP (either of which will enable -lart of the second fan.) Spurious actuation of the interlocking relay will prevent any fans from running. His is acceptable (See note 1).

,m ,m ,/7 N. Nv/

COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS (cont.)

I ENect on Systems Rennerks t Failere Mode j l Identdication

2. Start Dences- a. F.il to Operate (open This is buuuded by the case of control circuit wiring faHmg Control or close) open (la). Control circuit wiring failing open will open the (e.g., Push- current path to/from the device. The .csult is the same as if Button, Relay, the device did not operate. Fa.n either does not run (and other fan is ensbled), or runs spuriously (disabling other fan).

ES. Matrix)

If the interlock relay for one fan fails, the other fan may be enabled and start (two fans runmng). However, the second .

I fan startmg will disable the first (with a different relay).

leavinpc fan runnmg.

b. Fail to Release (stick) 'Ihis is bounded by the case of a short to another conductor (Ic). A short across two conductors will have the same effect as a control device stickmgin the actuated position. There

$ will be continuous current flow through the device. Spurious operation of a fan in fast or slow speed will cause the alternate fan to stop through the interlocking relays.

Spurious actuation of the interlocking relay will prevent any fans from running. This is acceptable (See note 1).

c. She:t to Ground, Shorting the contacts of a control dewee either to ground or Short across Contacts across two contacts of the device wdi have the same effect as a short between two conductors (Ic) or a conductor shotted to ground (Ib), both discussed above. Therefore, this case is bounded by the short to ground of a conductor and an open conductor and the consequences are acceptable.

n 7 7'^g (v) 7

(

ye

)

(v)

COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS (cont.)

Failere Mode E5ect ce System Rec. As identi5caties

3. Control Power a. Power Fail to either Fan stops, enabimg second fan to start. Note-The three fans Supply Fan are gu-ud from different Cass 1E busses.
b. Voltage High or Low Tins could cause intermittent fan operation or failure of the on either Fan fan control circuit. In both cases, the fan stoppmg enables the second fan to start. Once the second fan starts, the first far is disabled.
c. Spikes or Noise on This is not normally considered for relay logie (Reference 2) either Fan and would not be expected to result in any new failure modes not already addressed.
4. P:rwer a. Contactor Fail to Fast Speed: No effect on system. On ES. actuation, fast

[ Gose,or spurious speed fans are tnpped.

Components e.g. contactors, opening motor cucuit Slow Speed- Interlocking relay should be de<r m .g,ud, protector allowing second fan to start. If auxniary contact changes state when contactor does not, second fan will not start. This is (MCP's) acceptable (See note 1).

Shortmg- No effect on system. On ES. actuation, fast speed fans are tnpped.

APWm ~

,C

(

COMPONENT LEVEL FAILURE MODES AND EFFECTS ANAI VSIS (cont.)

h- C*"P'"*"' Fallarr Made Effect on System Remmarks Identification

4. Power b. L.,ntacter Stick Fast Speed: Two separate contactors a e required for f.st Componere Closed (Fast, Slow, speed operation. Sticking of fast contactor (single failure)

Shorting) will not allow fan operation because shorting contactor will c.g. contactors, motor circuit be open.

protector Slow Speed: Contactor sticks and keeps fan tunning in slow.

(MCP's) Secor.] fan will not run because auxiliary contact will drop out interlocking relay.

Shor&g- Contactor sticks closed. With fast contactor, fan See black box failure notes 2&3.

starts m fast. Wi:h slow contac:ar, shorting contactor la bolted fault which trips MCP, stopping fan and enabling second fan to start.

c. Contsetor Spurious This condition is bounded by the cases descnkd above (4b) Tbc effects of this failure will be

}

Closure (Fast, Slow, for contactor stuck closed. consideredin the black box J

failure section.

Shorting)

)

d. Motor Circuit Catastrophic failure would ca se powr " s and enable Protector second fan to start.
c. Catast;nphic Short in Damage te control eqj pment would trip MCP (4d). A fire MCC 7 tucket would cause hat shorts (Ic). Isolation of interlocx wiring / devices by the use of barriers (Siltemp, conduit, boxes, etc.) and space will prevent damage to schematic wiring from y

being propagated to interlock wiring.

f. Overload Relay Spurious Trip: Second fan start would be enabled.

Failure to Trip: Motor overheats and burns up. Would need to trip MCPS to start second fan If not, no fans would be J

runtdng. Tois is acce- * (See note 1).

.q ' ' "g l

i l

e -

COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS (cont.)

l l.

i "Ponent Effect on System Remarks Failure Mode 6

l

5. Fan / Fan Motor a. Short If MCP or overload relays trip, second fan start is enabled. If f rot, no fans are running. His is acceptable (See note 1).
b. Open Circuit Open field circuit (contactor c! ned): Motor not running, i second fan will not start because current is not monitored.

No fans running. This is acceptable (See note 1).

c. Locked Rotor Overload relays trip and second fan starts. See above for case in which overloads fail to trip (4f).
d. Overheat (IIigh The result of this fault will be a mechanical failure of the fan Ambient, Dense Air) and/or fan motor. His possibility is covered in the locked rotor (Sc), fan failure (mechanical) (Se), and overload relay

$ sections (4f).

c. Fan Failure Motor runs, but provides no air flow. Second fan would not (Mechanical) start. Therefore, no fans running. His is acceptable (See note 1).

i

%) ^ L ./

V.

COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALYSIS (cont.) .

l l- Comeponent FEcet on System Researks

). FaEare Mode identification

6. " Black Box" a. ControlCircuit Spurious shorting of control circuit wiring, e.g., in the case of Wiring: Spurious a loose wire on a terminal board, may cause intecittent Shorting actuation and/or de-actuation of control circuit componer.ts (i.e. relays, contactors). The effects for a range of these failures are covered in the above sections. No combination of postulated intermittent shorts has been identified wlich will resuit in the simultaneous energiration of two fans. This is the only unacceptable fan combination (See Note 1).
b. Start Devices- 'Ihe effects of spurious, intermittent failure of control relays

.J Spurious Failures of are the same as for i purious shorting of control circuit wiring.

Coatrol Relays which could result in ielermittent control relay operation.

Eerefore, the effects of this type of failure are the same as

.$' described above for control circuit wiring: spurious shorting.

c. ControlPowerSupply All postulated failures are addressed in the Control Power Suppiy section.
d. Fower Components: See Note 3 concerning this postulated failure.

Failure cf Main Contactors to Open, with auxiliasy contact energized.

_ _ _ _ - - _x, 7

/*% ' {""*

f* ,

~

i COMPONENT LEVEL FAILURE MODES AND EFFECTS ANALVSIS (cont.)

Notes:

1. A single failure of one RB tan would mean that, based on the single failure assumption, both EDGs and BS puraps will be operating.

He plant's 'icensing basis requires either both RB fans, one RB fan and one BS pump, or two BS pumps operating for containment -

cooling (Reference 4). Herefore, a failure preventing both fans from r.mning is acceptable, since it can be assumed that both BS pumps are operating, based on the single failure criteria.

2. Credit can be taken for separation features provided and consideration that coincident failure of auxiliary contacts is less than credible -

based on the considerations of the MAR Design Input Record Item #19.

3. Based on the results of analysis deliving from Restart Issue D-28, the only unacceptable failure is one that results in two fans continuing to run simultaneously. One " black box" type failure has been identified that could result in two fans running simultaneously. The fan starters are actuated when relay 42/S contactor) is energized by the fan control logic circuit. He yoke and armature assembly and the moveable conts et assembly move upward into the stationary contact block, where the three rc.oveable contacts connect the three upstream stationary contacts with their respective downstream contacts, starting the motor (Reference the f Allen-Bradley and Square D vendor manuals). He moveable contact support assembly will also actuate the auxiliary contactors, closing the auxiliary indication contact. This energizes an auxiliary relay, whicii blocks the starting of the other fans. A case in which the main contactor actuates, eneigizing the fan, and catastrophically fails, causing the main contactors to fail in the closed position, with the auxiliary contactor failed in the dc-energized posidon, has been postulated. This would resu!: in the fan running, but due to the failed auxiliary contactor, the auxiliary relay would not be energized. De second fan, receivirig the indication that the first is not running, would start. The second fan running would provide a signal to trip off the first fan, but due to the main contactors being failed closed, the first fan would continue to run. This occurrence would require the failure of both the main contactors and the auxiliary contactor simultaneously, a double failure not normally considered. However, the remote possibility that a catastrophic failure of the starter could cause this warrants consideration. Note that the starter and auxiliary contact actuation descnkd above pertains spdfically to the fans A and C starters, manufactured by Allen-Bradley. The fan B starter is manufactured by Square-D, and has a differeat actuation arrangement.' In the fan D starter, closing the main contactors actuates the auxiliary contacts with a rotary .

cam mechanism. Review of IEEE STD. 500 failure data states that the failure rate for circuit interrupter / relay is less than 1 per 10' year, and less than 0.5 per device per 10' operating cycles, covering all types of failures of this equipment. Based on review of contoctor designs and discussions with the vendors, no mechanistic failure mode has been identified which would result in the " black box" failure discussed above. Therefore, this case is not considered to be credible.

. ~ k

e ESMPR

~ A 8 80CI Af t $ INC.

g I h* G I N E E. R S V f Section 5 REFERENCES -

T 1.- ANSI /IEEE Std. 352-19Eri,IEEE Guide for General Principles of Reliability Analysis of Nuclear Power Generating Station Safety Systems.

2. IEEE Std. 500-1984, IEEE Guide to the Collection and Presentation of Electrical,-

Electronic, Sensing Component, and Mechanical Equipment Reliability Data for Nuclear Power Generating Stations.

L 3; L ANSI /IEEE Std. 242-1986, IEEE Recommended Practice for Protection and Coo.dination of Industrial and Commercial Power Systems.

4. Crystal River 3 Restart Action Plan, Revision 2, Issue Number D 28.

a

5. . Florida Power Corporation, Nuclear Engineering Department, Crystal River Unit 3,

- Instruction Manual 150, " Motor Control Center," Revision 1.

6. F:orida Power Corporation, MAR 97-09-05-01,"RB Fan E.S. Run Logic." FCN 1, "High Speed Interlock."
7. Florida Power Corporation, Crystal River Unit 3, MAR Sketch Numbei . 90501-05.

- 8.' Florida Power Corporation, Crystal River Unit 3, MAR Sketch Number 97090501-06.

9.- Florida Power Corporatica, Crystal River Unit 3, MAR Sketch Number 97090501-07.

10. Florida Power Corporation, Crystal River Unit 3, Interim Drawing D12-70201-A7,

" Wiring Diagram E.S. MCC 3B34AN RB Fan AHF-1B," Interim Revision B.

.11. NFPA-70, National Electric Code.1996 Edition.

5-1

-_____ __ _- . _ - _ _ _ _ .