ML20238D415

Evaluation of Engineering Rept ERPT-E0220 Re Reactor Regulation of Util Approach to Compliance W/Reg Guide 1.75 for New Diesel Generator Installation at Plant.Licensee Approach to Demonstrating Compliance Acceptable

ML20238D415
Person / Time
Site:	Rancho Seco
Issue date:	09/03/1987
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20238D411	List: ML20238D408 ML20238D415
References
RTR-REGGD-01.075, RTR-REGGD-1.075 NUDOCS 8709110166
Download: ML20238D415 (16)

Text

.

\

STAFF EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION REACTOR REGULATION OF SMUD APPROACH TO COMPLIANCE WITH RG 1.75 FOR NEW NEW DIJSEL GENERATOR INSTALLATION AT RANCHO SECO SMUD (the licensee) has installed two new diesel generators (EDG) and associated electrical distribution equipment at Ranch Seco as part of a plant upgrade to meet NRC requirements. Details of the EDG installation were included as part of Proposed Amendment No. 147 to the Rancho Seco Technical Specifications.

During its review of Proposed Amendment No.147, the staff noted that the li- i censee had not adequately described how the electrical sep> ration positions of Regulatory Guide (PG) 1.75 had been met. Through discussions with the licensee, it was determined that the specific requirements of RG 1.75 and IEEE Std. 384-1974 (which is endorsed by RG 1.75) had not been met in all cases. On June 2 and 3, 1987, the staff met with the licensee to discuss how the licensee would demonstrate acceptability of the EDG installation even though the specif?-

criteria of RG 1.75 and IEEE 384 were not met in all cases. Subsequent to the above meeting, oa June 26, 1987, the licensee submitted Engineering Report ERPT-E0220 entitled " Report on Conformance of Nuclear Service Electric Building (NSEB) and OG Building Electrical Installation to Regulatory Guide 1.75." This report documents the discussions of June 2 and 3, 1987 regarding RG 1.75 i

concliance, and is the subject of this evaluation.

R.G.I.75 endorses, with specific exceptions, IEEE Std. 384. Engineering Report ERPT-E0220 addresses the sections of IEEE-384 which are applicable to Ranch Seco 8709110166 870903 PDR ADOCK 05000312 P PDR j

4 e

and-includes a discussion on how the specific requirements.are met.- The' staff's l

evaluation of the licensee's report is included below and is keyed to the IEEE-384 sections identified in the report.

[

At Rancho Seco, there are two separate and completely redundant safety trains, A and B. These safety trains are totally independent and meet all RG 1.75 physical separation criteria with respect to' redundant safety trains. Within each train, there are two redundant, safety related (Cla,ss 1) channels. The physical separation of these redundant channels, within a safety train, from each other or from non-safety (Class 2) circuits is the subject of this report.

While the equipment separation at Rancho Seco is in compliance with the posi-tion stated in RG 1.75, not all cabling is separated by the distances specified in the guide. However, the licensee has taken the approach to demonstrate by analysis, modifications and/or tests that the non-conforming installed cable /

wire configurations are acceptable. To demonstrate this approach, the licensee i has taken the position that tests conducted at other nuclear plants (River Bend /Vogtle) have demonstrated that less than optimum cable separataion is acceptable when certain modifications are made, are applicable at Rancho Seco.- '

The licensee also takes the position that any. differences between insulating i

and jacketing materials of the cables tested and the cables installed at Rancho Seco is acceptable without retest because all cables are IEEE 383-1974 qualified and because of the additional conservatism the licensee takes for each. specific, ,

non-conforming configuration. The licensee's submittel: Engineering pe,nnet

_ _ _ _ _ _ _ . __ __ . _ _ __a

3 ERRT-E0220, contains analyses of non-conforming configurations in the areas of raceway separation and separation of cables internal to control boards and cabinets, which are discussed below.

IV. EQUIPMENT SEPARATION The new diesel generators are completely independent of each other. They i

are located in separate rooms in the diesel generator building which are separated by a 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> rated fire wall. Each diesel. generator has its own auxiliary systems, including ventilation. There are no shared or common systems.

Electrical distribution equipment for each diesel generator is located in a separate room in the Nuclear Service Electric Building (NSEB). These rooms are separated by a 3-hour fire wall, and each room has its own ventilation system.

The "A" diesel powers safety train A, and the "B" diesel powers safety train l B. Battery chargers, batteries, inverters, and distribution panels asso-ciated with a safety train are located with other equipment of that train 1

l in their respective rooms in the NSEB, and are separated from redundant 4 l

equipment of that train by at least 10 feet horizontally. The above separ-I ation of safety trains by 3-hour fire walls and separation of redundant I l

l equipment within a safety train by specified horizontal distance is in I l conformance with RG 1.75 and is, therefore, acceptable.

I l

l l

1 1

1 I 7

4.'

V. RACEWAY SEPARATION-A. Separation Between Class l'and. Class 2 Raceways l 1. A minimum of 1" separation is maintained between Class'l-:

raceways (trays and conduits) and Class 2 enclosed raceways q used for power and control circuits. The staff will find this acceptable or confirmation thatLthe Class I trays are covered .

in accordance with RG-1.75.

In a limited number of cases, Class 1 raceways and Class 2 1

conduits used for instrumentation circuits are allowed'to touch 'l each other. The instrumentation circuits u'se shielded cables qualified per'IEEE 383, carry low energy signals and are routed in instrumentation raceways, 'In th'ese. cases, the classi2

' 1 instrumentation circuit raceways'which are in. contact with Class 1 )

i raceways carry only milliampere current even under faulted .

s conditions. At these low energy levels,-there is no credible I

fault that could 'cause overheating of: the Class. 2 cables' and-associated raceway with attendant damage to the Class l' cables, i Therefore, the staff finds-that. Class 2 conduits.in contact with:

~

Class 1 raceways under the above conditions is acceptable. .,

1 i

2. Class 2 trays are separated from Class l'. trays by 1)'a minimum- l distance of 3 feet horizontal and and 5 feet vertical, 2) sepa-i ration barriers installed in either the horizonte1 or vertical J

i t.--__---_____---.____-..- - _ - _ _ . . . - _ - _ _ _ - - _ _ _ - . _ _ _ . _

j l' .

^

]1 y

, planes,.or 3);by installing solid covers on the Class 2 trays. .i The staff will find this acceptable.on confirmation that the Class.1 and Class E. cable trays that;are .not separated by the minimum distances or by barriers have' covers installed in-l accordance with RG 1.75.

~

1 l.

In some cases, separation by one of the above methods is'not possible. .These cases are limited to situations where Class 1 cables in conduit are.in close proximity to an open' Class 2 cable tray. In these cases, the Class 1 conduit.will be wrapped with a 200% overlap (3 layers) of.siltemp 188 CH. ' Tests conducted at other nuclear plants have demonstrated that faulted cables carrying 2600A and located within .1 inch of target cables wrapped with a 100% overlap (2. layers) Lof. Siltemp 188CH did not ,f 1

l cause damage to the target cables. ' At Rancho Seco, .there is ]

additional protection as follows; 1). the target cables areL in.

conduit, 2) separation is 1" or more, and 3) the target cables have an extra layer of Siltemp 188CH. It.is the staff'sLview that this added protection adequately compensates for.any  !

differences between the insulation and jacketing material of the cables' tested and the cables installed at Rancho Seco. There-fore, the staff concludes that separation between Class 2 cable.

l trays and Class 1 conduits as described above and in ERPT- -

l E0220 will be acceptable.

I i

i

'b

m.

.6- 1:s

8. ' Separation Between Redundant' Class:1' Raceways

1. Trays of redundant Class l systems (trains A&B) are installed' in. separate rooms (separated by 3-hours fire walls). 'This' meets the positions of RG 1.75 an,d is: acceptable.

2. Class 1 channels in cable trays are' separated from redundant- .

Class 1 channels in enclosed raceways by: 1) a minimum distance d of 3 feet horizontally and 5-feet vertically, 2) separation q barriers installed in either the horizontal or. vertical planes, k i

or 3) by installing solid covers on the cable' trays, 'These- j a

methods of obtaining separation are in accordance with RG 1.75;- 'l l

and are acceptable.

In some cases, separation by the above methods is not possible.  !

These cases are limited and are described below,

a. In some cases, Class 1 channels in cable trays are. separated from redundant Class 1 channels in' rigid conduits by at-least3 inches,but'lessthantheminimumldistanceproposed l

-)

(1 ft.) by RG 1.75. In the'se cases, the voltage is limited to 480 VAC, fault cables in the tray are. limited to #2/0l AWG, and the fault cable. in the. rigid conduit is limited to" a single 500 mcm. Tests conducted at other nuclear plants. 1 for similar installations have demonstrated that'500 MCML

-l

!- j

f

(

I

' armored cables,without additional. enclosure e d #2/0'AWG~

faulted cables in free air located 3/4 inch'below target cables.in. trays and flexible conduits, respectively,-did not cause' damage to the target cable when subjected;to 2600 Amps until the faulted cables'open circuited. At' Rancho Seco, the faulted and. target cables are always in;a rigid conduit or a tray, and the minimum separation is 3 inches.

This additional protection' adequately compensates:for any; j differences between the. insulation'and jacketing material' ,3

.i of the cat,les tested and the cables installed at Rancho

.]

Seco. Therefore, the staff concludes that the~ separation 1

.1 between redundant Class 1 channels-in cable trays' and-rigid conduits as described in the above paragraph in ERPT-E0220. l.

.\

is acceptable.

b. In a limited number of cases, separation of redundant' Class 1 channels cables is less than that described in Paragraphs j B.2 or B.2.a, above. These cases fall.into 2 categories, as follows: 1) cable trays .with faulted cables at least 1 inch but less that 3 . inches from a redundant. Class 1 l channels in enclosed raceway, and 2) conduits with . faulted a

cables at least.1 inch but less than 3 inches'from a 'l redundant Class 1 channels in'a cable tray. To_ compensate for this lack of physical separation, the target cables or OL___-__

-__-_a_---____________-_____ --

,e 4

~

thefaultedcableswillfbewrappedwitha200% overlap (3 ,

layers) (vs_10J% overlap'-2 layers) of Siltemp 188CH.

In' addition, the maximum voltage is. limited to'.480 VAC, wrapped conduits are limited to-a single 500 MCM , faulted cable, and wrapped cables in the trays are limited to.

~

. cables of No. 2/0 AWG or smaller. Tests conducted at other

~

nuclear plants have demonstrated that faulted cables in L

L free air, when located 1 inch below target. cables in-flexible conduits and wrapped with.a '100% overlap (2 i

layers).of Siltemp 188CH,'did not cause damage to the.

target cables when subjected to a 2600A. fault currentuntil-the fault cables open circuited. At Rancho Seco,. the-faulted cables.are either in conduit or trays and . limited to No. 2/0 AWG or single conductor ~500 MCM-as' described; above, the separation is at leastL1 inch, and'a'200%-

overlap (3 layers) of Siltemp 188CH is used in all cases.

This additional protection' adequately compensates for any differences between the insulation .and jacketing material; of the cables tested and the. cables installed at Rancho Seco. Therefore, the staff concludes that the' separation between redundant Class 1 channels in cable trays land; enclosed raceways / conduit as described in the above para-graph and in ERPT-E0220 will be. acceptable.

c. In a few-isolated cases, separation between redundantLClass-1 channels circuits meets the configuration stated in y

-.__.mm_m__.____m.__

I' Paragraph B.2.a. above, but cables larger than No. 2/0'AWG are routed through the cable trays. These cases involve 120 VAC un-interruptable power supplies and are limited to the cables connecting the inverter and regulating transformer to the bypass switch, the transfer switch, and the distri-bution panel. The connections are made using a single 350 MCM conductor, or 2 single 250 MCM conductors. These circuits, however, are not subject to the same fault -

current used in the test discussed in Paragraph B.2.a; i.e., 2600A. The available fault current was used to a

calculate the cable temperature rise under worst case ]

loading. The results of the' calculation show that the  ;

L cables, in any case, would not get hot enough to ignite the ]

cable insulation and thus pose a threat to nearby circuits- . j Based on this, the staff concludes that routing of cables I larger than No. 2/0 AWG in cable trays when the available fault current to these cables is limited by inverters / l i

transformers and/or circuit breakers and fuses does not constitute a concern greater than that evaluated in Para-graph B.2.a, above, and is therefore acceptable.

d. At Rancho Seco there are four special cases which do not fit into any of the categories discussed in Paragraph B.2.a, b, and c above. These special cases involve four, sixteen inch diameter pipe ducts which run between the NSEB

10 and the Auxiliary Building. These four ducts are grouped in two pairs, with each pair representing one safety train.

The redundant safety trains are separated by a 3-hour fire barrier. Within each pair of ducts, one duct is used for instrumentation circuits only, and the other duct is used for power and control circuits. Within each duct, one group of redundant class I channel circuits is routed through a 4 inch flexible conduit. The other channel circuits are in the 16 inch duct outside of and touching the 4 inch flexible duct. The instrument circuits carry low current even under faulted conditions and, as such, there is no credible fault'which could cause a cable or cables in one redundant instrumentation channel to overheat and damage a cable or cables in the other redundant instru-mentation channel. The staff therefore concludes that the installation of instrumentation cables in their respective 16 inch pipe ducts is acceptable. The 4 inch flexible conduit in each 16-inch pipe ducts for power and control circuits each contain one power and two control circuits I for an auxiliary feedwater isolation valve. The power circuit (120 vac) in each of these flexible conduits have adequate current protection to preclude a faulted condition which could overheat these power cables and cause damage to the cables inside the pipe duct in contact with the flexible conduit. In reverse, a faulted condition on the power cables l

. l

. i

. l.

(t80 vac) in the pipe duct could conceivably damage the cables in the 4 inch flexible conduit. _In this case, loss of an auxiliary feedwater supply path to a steam generator . )

I would result. However, there are two parallel auxiliary- J feedwater supply paths to each steam generator. Therefore, loss of one supply path due to failure of an auxiliary feedwater isolation valve will not create a loss of feedwater event to a steam generator. A total loss of auxiliary _

l feedwater to a steam generator would require a coincident failure in the other, parallel auxiliary feedwater supply path whose power and control circuits are outside the 16 Based on tho fact that a fault in the j inch pipe duct',.

instrumentation pipe ducts will not propagate and that a )

i l

fault in the power and control pipe ducts will not result j

\

in unacceptable consequences, the staff concludes that the installation of the 16 inch pipe ducts between the NSEB and the Auxiliary Building is acceptable.

1

e. With one exception, enclosed raceway to enclosed raceway separation of at least one inch is maintained. In the single exception, a rigid conduit containing instrumentation circuits in one channel is allowed to touch a flexible con-duit containing redundant instrumentation circuits in the 1

other channel. The above.1' inch. separation;is inlconfor-

mance with RG 1.75 and is .therefore acceptable. The single <

exception is acceptable due .to the low energy . circuits invol-ved as, discussed in' Paragraph A.1, above. 1 VI. Internal Separati,on A. Wiring Separation Within Enclosures

1. Wiring separation is met within enclosures by maintaining a mini-i mun of 6 inches between redundant wiring and between Class 1 and' Class 2 wiring, or-by using a barrier'to separate redundant Class 1 wiring and Class 1.from Class 2 wiring. Both of these methods are in compliance with RG 1.75 and are, therefore, acceptable.

2. Some internal wiring does not meet'the' separation criteria i

discussed in Paragraph VI. A.1, above. The specific cases of devi-ation from the separation criteria.and the compensating measures proposed by the licensee are discussed below:. _,

l J

a. In some cases, Class I instrumentation and control. circuits.

(No. 8 AWG or smaller) in one channel are not separated from o

redundant channel Class .I circuits or Class.2 circuits tor.

the required distance or by barriers. In these cases, one- ,

1 l

3

l l *

~

i }

,i q

of the circuits ~is wrapped with a.10.0% overlap'.(2 layers) of-

]

1 Siltemp 188 CH and .then wrapped with:3M Scotch l69 glass cloth tape. ' Tests conducted at other nuclear plant's have j q

L demonstrated that a control cable No. 8 AWG or smaller which l

is in' contact with another control' cable can be subjectedito 1

the worst case electrical fault without' the fault propagat-ing to the touching cable provided one of the cables is: <

wrapped with a 100% overlap (2 layers)'of'Siltemp 188 CH.

At Rancho Seco, a layerfof. glass tape over the Siltemp provides additinnal protective margin. This additional protection adequately compensates for any differences between the insulation and jacketing material'of the cables tested and the cable installed at Rancho Seco. ~In' addition, the circuits to which this treatment will be-applied are low energy circuits with inherently low fault

~

current potential. Therefore, the staff. concludes that'the method for demonstrating adequate separation of instrumen-tation and control cables (No. 8 AWG or smaller) described above and in ERPT-E0220 will be acceptable.

b. In some cases, Class I multi-conductor instrumentation and control cables (No. 10 AWG & smaller)1are not separated from I. the redundant channel Class I or Class 2 cables by'the' l

required distance or by barrier. For these cases, effective separation is obtained by installing a copper braid over each

/

=

j 1

1 individual conductor from its termination back to the cable

'l breakout, an'd an additional copper braid over the cable jacket from cable breakout up to the point where at least 6 inches i

of separation is maintained. All copper braids touch each l other and are grounded at one end. The individual braids are covered with a flame retardant insulating tubing, and 1 1

Raychem WCSF-N shrink tubing is installed over the cable.

Jacket braid. In these cases, both the faulted and target i i

cables are protected. No tests have been conducted for this specific configuration. However, results of tests conducted on No. 8 AWG instrumentation and control cables demonstrate adequate protection when only one cable. (faulted or target) is protected and the cables touch each other.

For the cases covered in this section, both cables have {

l additional flame retardant insulation, and both'have a ground path for fault currents installed. In addition, the l circuits to which this treatment will be applied are low energy circuits with inherently low fault current potential.

Based on the above, the staff concludes that the cable .

treatment described in this paragraph provides protection which equals or exceeds the protection described in Paragraph VI.A.2.a. above and , therefore, will be acceptable, I

c. In four specific cases, Class 2 instrumentation cables are allowed to touch Class 1 cables without any compensating

i

r.

.y o

.e . d' j

provisions. In 4.16 KV. switchgear 54A2 and S4B2, Class 21 l1 output wiring from-4 isolating transducers _in each switch =  :.

gear is in contact.with class. l. wiring. :In TDI diesel- 'I generator control panels H2DGA2 and H2DGB2, Class 2: output  !

wiring from 5' isolating transducers in each panel is-

\.' The transducers will allowed to touch Class 1 cables.

i limit the short circuit' current to'such a low'value that.. y the Class 2 wiring-can not get overheated and cause damageo ,

I to the Class 1 cables. The staff. agrees'with'this' position-and concludes, therefore,'that Class 2 cables in contact >

with Class 1 cables in the above cases is acceptable. U

)

d. In multiplexer cabinets H4CDAR7. and H4CDAR9, cables carrying signals for redundant SPDS channels are not separated from' each other or from class 2 circuits. A' calculation (No.

Z-SEP-E0693) has been performed to show that'all circuits are low energy circuits which cannot damage each other, even under faulted conditions. Based on this' calculation, the fact that the.SPDS is properly isolated from safety protection systems, and that there is' redundant indication for the SPDS in the other safety train, the staff concludes that the cable arrangement in multiplexer cabinets H4CDAR7',

and H4CDR9 is acceptable.

1 f

. )

16-VII Raceway / Circuit IdentificaHon The staff has reviewed the licensee's description of raceway / circuit iden-tification included as part of ERPT-E0220. The staff has concluded that the licensee's identification scheme is in conformance with RG 1.75 and is, therefore, acceptable.

CCNCLUSION i

l l As noted above, not all cabling at Rancho Seco conforms to the positions of RG 1.75 with regard to physical separation. However, the licensee has taken the approach to demonstrate by analysis, modifications and/or tests that the non-conforming installed cable / wire configurations are acceptable. To demon-strate this approach, the licensee has taken the position that tests conducted l at other nuclear plants have demonstrated that less than optimum cable separa-tf on is acceptable when certain modifications are made, are applicable at

)

l Rancho Seco. The licensee also takes the position that any differences between l 1 insu'lating and jacketing materials of the cables tested and the cables installed j at Rancho Seco is acceptable without retest because all cables are IEEE 383-1974 -

qualified and because of the additional conservatism the licensee takes for i each specific, non-conforming configuration. The staff agrees with the licen-see's positions, and conclude that the licensee's approach to demonstrating  !

compliance with RG 1.75 is acceptable.

1 1

l i