ML20199G860

Rev 0 to Technical Evaluation of Oyster Creek Nuclear Generating Station Analysis of Cable Ampacity Limits

ML20199G860
Person / Time
Site:	Oyster Creek
Issue date:	04/10/1997
From:	Nowlen S SANDIA NATIONAL LABORATORIES
To:	NRC (Affiliation Not Assigned)
Shared Package
ML20199E440	List: ML20199E437 ML20199E456 ML20199G860
References
CON-FIN-J-2503 NUDOCS 9711250321
Download: ML20199G860 (12)

Text

._. -- .

, x.

A Technl:al Evaluation of the Oyster Creek Nuclear Generating Station Analysis of Cable Ampacity Limits l-Aimer Report to the USNRC l

i Revision 0 April 10,1997 Prepared by:

Steve Nowlen Sandia NationalLaboratories Albuquerque, New Mexico 87185-0737 (505fS45'9850 Prepared for: .

Ronaldo Jenkins Electrical Engineering Branch Office of Nuclear Reactor Regulation U. S. Nuchr Regulatory Commission Washington,DC 20555 JCN J2503

-9711250321 971113 PDR ADOCK 05000219 3 PDR  ;

' Enclosure 2

, w - . .,,,-w- .y,, - - - -

r.  % . .

l l

TABLE OF CONTENTS: i

. SE11on Pagt i

F O RWARD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...l u. . ,

1.0  !

OVERVIEW AND OBJFCTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1

2.0 UTILITY AMPACITY DERATING APPROACH . . . . . . . . . . . . . . . . . . . . . 2 i 2.1 Range of Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2  ;

2.2 Overview of Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 -

2.3 Determination of Fire Barrier ACF . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.4 Apparent Error in Base Line Estimate for 5 kV Cables . . . . . . . . . . . . . 3 2.5 Application of NEC Diversity. Based Conduit ACF Factors . . . . . . . . . 4 3.0 LICENSEE SPECIFIC RAI RESPONSES . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.1 RAI Item 1: Physical Comparison of Tested and In Plant Barriers . . . . . 6 3.2 RAI Item 2: Example Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.3 RAI Item 3: Ambient Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.4 RA' ae.= 4: Air Drop ADF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 3.5 RAI Item 5: Battery Charger Lod . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 4.0

SUMMARY

OF FINDINGS AND RECOMMENDATIONS . . . . . . . . . . . . . 9

_s ~ ~. - _ . . - _ - . .

l

, FORWARD

' ^

l The United States Nuclear Regulatory Commission (USNRC) has solicited the support of  ;

Sandia National Laboratories (SNL) in the review of utility submittals associated with Sre protection and electrical engineering. This letter report is the second in a series of '

submittal review reports related to the Oyster Creek Nuclear Generating Station j

(OCNGS). The submittals reviewed deal with the issues of Thermo Lag 3301 Era barriers, and in particular, the assessment of ampacity loads for protected cables. An -l initial review report was completed by SNL on June 13,1996 based on a licensee

. submittal ofDecember 8,1995. A subsequent USNRC Request for Additional Information (RAI) was forwarded to the licensee on August 26,1996. The current report documents an SNL review of a liunsee response to this RAI as provided in a licensee  !

document dated November 25,1996. This work was perfonned as Task Order 3, Subtask 5 of USNRC JCN J 2503, i

l I

I e

e .

[f e

r *

,a b

9 k

i- .

iii I E:

- . . _ - - - ~, ._

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

t_ . _ .,. m. .,-, m . _ .__ __ . . . _ __a. _ -.-_..m. .__ _ ....__. . .

+

1.0 OVERVEW AND OBECTIVE In ruponse to USNRC Generic Letter 92-08, the Oyster Creek Nuclear Generating Station (OCNGS) provided initial documentation of the utility position regarding ampacity derating facto., associated with its instt!!ed Thermo Lag fire barrier systems in a submitta!

dated 12/A/95. SNL reviewed this submittal and provided a review findings repon to the USNRC on 6/13/96 In large part as a result of this SNL review, a request for additional information (RAI) was forwarded to the licensee on 8/26/96.

'ine objective of this report la to document de findings and recommendations resulting from an SNL review of the licensee RAI response. The submittal reviewed was documented in a utility letter as follows:

i Letter, November 25,1996, (item 6730 96 2336), M. B. Roche, GPU Nuclear Corp, OCNGS, to USNRC Document Control Desk (with one Attachment).

SNL was requested to review this submittal under the terms of the general technical support coretact JCN J 2503, Task Order 3, Subtask 5.

e 4

4

_.- . . . - ... ._-.- - - - . - - - - - - -- ---l

• 1

.(

,, 2.0 UTILITY AMPACIT'lDERATING APPROACH i i

.2.1 Range ofAgh.es  !

i i

One somewhat unique aspect of the OCNOS nre baniers is that all of the installed Src l

banier systems are associated with either conduits or cable air drops. There are t apparently no cables trays clad in Thermo Lag Sre baniers. Thl is one feature that  ;

simpli6es the licensee's analysis problems.

]

It was also noted that relatively few cables are actually impacted by the 8te banier systems. In particular, because the cases generally involve conduits or air drops, the >

number of cables inside each banier system is relatively small. In contrast, when a typical  !

cable tray is clad, even though only one or two of the cables in the tray may be the actual '

targets of the cladding, many power cables may need to be assessed simply because they are co-located in the same tray. This, too, has led to a signlBeant reduction in the required t scope of the licensee's analyses.

2.2 Overview ofMethodology

, The consideration ofcable ampacity loads at OCNGS is based on a very simple method.

While some of the details of the assessment have been adjusted in respone ta the USNRC 1 RAI questions, the overall methodology remains largely unchanged. i h licemee's general process of analysis begins bv citing the base line ampacity limits for '

individur.1 cables &om the National Electrical Code (NEC) Handbook. While the actual ,

tables applied are not cited by the licensee, SNL has been able to determine that all of t'se -

ampacity values apparently derive kom either Tai >1e 310-16 (for 0-2000V cables) or 310-73 (for 2001 5000V cables) as identified in the 1996 version of the NEC. These .

values are appropriate to the licemee analysis in that all usume that the cables may be i.,

installed in a conduit. For each base line ampacity, the value is further adjusted for the -

assumed ambient in the plant,40'C. (See related P.A! item 3 as discussed in Chapter 3 - -

below, and note that SNL has identified an apparent conarvative error in the licensee ,

treatment as discussed in Section 2.4 below.) -

.,g h licemee analysis of base line ampacity limits has also included application of an ACF

'l based on the number ofconductors in the conduh. The basis for the values chosen by the e licensee la unclear, and this topic is discussed funher in Section 2.5 below. Finally, the :p licenses applies an ACF factor to reflect the impact of the Sre banier system. The Anal

'.6

- result is an estimate of the fully derated ampacity limit for individual cables. Given these '

s estimates of the derated cable ampacity limits, the utility simply compares the actual in- '

plant service loads to these ampacity limits. In all cases the utility has concluded that the  :-

estimated ampacity limits bound the octual in-plant service loads.

+

In general, thl approach la an a:ceptable means of demonstrating that cables are operating '

within acceptable amps:ity limits. - In particular, the utility analysis method is based on the ^h; '

comparison ofindividual cable loads to conesponding cable ampacity limits. However, ij.

g the level of documentation provided in the utility submittal remains quite sparse. 'three  :

'{:

i'.

2 S g

_ 3

. e.

.p 3

. )

.., complete example calculations have been provided, and SNL found no obvious errors in i

, these three exampla (beyond the apparent conservative error discussed in Section 2.4 i

. below). The licenses has also provided a summary table of results for other applications.

l While the dcemkm is somewhat sparse, SNL nonetheless Ends that the licenses has provided adequate dommentation to provide a reasonable assurance that cable ampacity limits have been given adequate considwation. In particular, the licensee has generally I demonstrated a sign 16 cant margin for all of the cable considered. SNL does not i recommend that any follow-up is regidred to fbrther clari.*y the general aspects of the  !

licenses analyses. Note that Section 2.4 belor discussed one apparent error in the licensee 1

analysis that is of a conservative nature, and Section 2.5 discusses one speciSc aspect of ,

the analysis for which a very limited follow-up is recommended.  !

2.3 Determination ofFire Barder ACF

. In the assessment of fire barrier ACF values, the licensee has directly applied the results of  !

the Texas Utilitin Electric (TUE) and Tennessee Valley Authority (TVA) ampacity  ;

dorating tuts. The lice,ises also provides an assessment of the derating impact based on the original manufheturer (Td!) tests, although these tests are no longer considered a valid ~

basis for analysis, and hence, SNL Ands these particular analyses should not be credited.

1 It is important to note that in its original submittal the utility did provide a compadson of the primary characteristics of the tested con 6gurations versus those installed at OCNOS.  ;

While certain questions were raised on this issue in the RAI, SNL has concluded that the licensee extrapolations are appropriate. (See fbrther discussion related to RAI item I in  !

Chapter 3 below.) Is. summary, SNL Snds that application of the TUE and TVA ampacity .

derating factors is an appropriate basis for analysis at OCNOS.  !

2.4 Apparent Error in Base Line Estimate for 5 kV Cables In determining base line cable ampacity limits, the licensee has adjusted all of the NEC i tabulated values using an ACF of 0.91 to reflect an assumed plant ambient of 40'C versus

, 30*C Table 31016 does assume a 30*C ambient, and hence, the adjustment of ampacitia derived tom this table is necessary and appropriate. However, arnpacity limits derived &om NEC Table 310 73 are already based on a 40*C ambient, and hence, ibrther adjustment is unnecessary. Nonetheless all of thus values have also been adjusted using the 0.91 temperature correction ACF. '

As a result of this apparent error all of the cited ampacity limits for the licensee's 4160 V cabin have an added, and presumably unintended, level of conservatism. This appears to impact the licensee cable identl6ed in the supporting tables as items 15,16,17, and 31.

Because this apparent error has had a conservative impact on the licensee's results, SNL '

recommends that no formal resolution is necessary. It is, however, recommended that the apparent discrepancy be broucht to the attention of the licensee. It would clearly be in the interest of the licensee to correct this apparent error in its internal dommentation.

l 3

k a # . ..m - -

+.___m____ _. '__--

. . 1

-a . .

.. 2.5 Application of NEC Diversity Based Conduit ACF Factors One correction applied by the Econses to the base line ampacity limits reDects a correction for the total count of current canying conductors in a single conduit. These values are nonnally taken directly tom the NEC, but there are two sets of values that can be used under different circumma==.

Prior to 1990, the NEC had published one set of values, but these values inherently assumed a 50% load diversity (no more than have of the conductors loaded at a given time). These values still appear in the NEC, but now appear in Appendix B as Table B-31011, and the diversity assumption is more explicitly stated. Since 1990, the NEC has .

published an alternate set ofmore restrictive values which do not include the diversity  :.

assumption for use in general applications (see NEC 1995 Section 31015, table under '

note Sa on page 70196). The changes only impact conduits with a conductor count of 10 or higher. - -

It is not clear which set ofvalues are being applied by the licensee because the licenses has .

not speci5ed the total conductor count for any of the conduits considered. However. -

based on the limited information available, it would appear that the licenses has applied the

. older pre-1990 diversity-based values: " .

Consider for example items 29 and 30 Rom the licensee's results summary table (see page 2 of 4 of Table 1 in the submittal). These two cases involve "ll 1/C #8" y,.

cables apparutly in a common condult. By " backing out" the " multiple t conductors in raceway" correction, it is apparent that an ACF of 0.7 has been '

I i applied in this step. This value does correspond to the ACF cited in the diversity- )fl; I based ACF table for a conductor count of 7 through 24. However, the updated v l post 1990 values would require application of a 0.5 ACF for a conductor count of 'I@.

10 through 20. Hence, if 11 is the total conductor count (it may actually be higher if there are other co located cables but it is certainly not lower) thea nominally an ACF of 0.5 should have been applied. .

@M

}l' ,' ,

Any application of the older, pre-1990, diversity-based values should be accompanied by speci5c consideration of the available load diversity. That is, the diversity based values h

.[.::

should only be applied when the diversity can be either explicitly demonstrated or JE-} l reasonably assumed. For many nuclear power plant applications, a reasonable assumption  :!/d .

of diversity is possible. 'Ihis would be especially true for control applications, and for Fi l certain power applications for which the equipment is never operated simultaneously; yj.;  !

However, the licensee has not provided any discussion of diversity in its conduit loads.

py;  ;

M. .

SNL Snds that the licenses has not provided adequate discussion of the conduit conductor t *)ij count ACF factors applied to its analyses. It is recommended that the USNRC ask the l licenses to explicitly state which set ofNEC values has been used in the analyses. Further, f/ i

• p{.

it is neommended that the USNRC ask the licensee that for each case in which the NEC llA

- diversity-based values have been applied to either (1) provide an explicit justi5 cation for

'if).

y use of the diversity-based NEC ACF values or (2) update the analysis using the newer $,

post 1990 NEC correction factors in which the diversity assumption has been negated.

M

$.h 4 yh

(

w.

s.m L _ _ _ _ .__ _ _ _ --- - - - - --- - -- E-

\ .. :.. _ _ .__*

'

• In general, it is not snticipated that this recommendation wi11 ultimately have a sign 16 cant Impact on the liceruee's sasessments. In particular, this observation only has the potential

, to impact conduits with 10 or more conductors (the factvis for 4 through 9 conductors are the same in both versions of the tables). Fuisher, for most cases it is anticipated that the available margin should be sufEclent to cover a change in this particular ACF value. For emnple, for the two cases cited above, items 29 and 30, if 11 is the total conductor count then appi; cation of a 0.5 conductor count ACF would still result in a dorated ampacity limit of 22 A, and this still exceeds the estimated load of 18.5 A. No cases can be veri 6ed with certainty, hnwever, based of the information provided in the submittal.

4 9

4 4

I e

9 5

3.0 LICENSEE SPECIFICRAIRESPONSES i

, , e 3.1 RAIItem 1: Physical Comparison of Tested and In Plant Baniers Synopsis ofQuestion: The licensee was asked to address aspects of the con 3 6 uration similarity issue related to air gaps in the conduit banier system as tested by TVA.

Synopsis of Response: The licensee has correctly pointed out that the specise tests being cited in the analysis did involve " dry St" and " post buttered" barrier installations, and -

hence, the air gap that normally forms between the conduit and the inner surface of the Are barrier system did, in fact, existing in the TVA tests cited.

• Assessment ofResponse: This response is fully adequate to resolve the identi6ed concern. ,

in fact, the licensee has identi6ed a very important aspect of the TVA tests that had not  ;

previously been appreciated by SNL. Most of the TVA tests involved a pre-buttered , :~.:,,

installation and elimination of the air gap. However, this particular set of 3.hr barrier tests did not, and hence, will have a much broader applicability to general industry installations. -

Findings and Recommendations: SNL finds that the licensee application of the TVA test *

, . results will conservatively bound the OCNGS installations as described in the submittals.

No further actions on this RAI item are recommended.

3.2 P '.! Item 2: Example Calculations  : ,

Synopsis ofQuestion: The licensee was asked to proside detalled examples ofits calculation process.

Synopsis ofResponse: The licensee has provided 3 case examples, one for a conduit with .:i; a three hour barrier, one for a conduit with a one. hour barrier, and one for an air drop. *

' hi* 1 Assessment of Response: The licensee response is fully adequrte to resolve the identined .

concern. SNL has reviewed the example calculations as discussed in Chapter 2 above.

Is i Findings and Recommendations: SNL Snds that the licensee has complied with the  !

is:.

USNRC request. No further actions on this RAI item are recommended.-

@f; 3.3 RAIItem 3: Ambient Temperature .9!-

-]g Synopsis ofQuestion: The licensee was asked tojustify its assumptions of a 30*C ambient h.,g.

used in certain of the cable analyses.

.l: ;,

Synopsis ofResponse: The licensee's updated analyses have uniformly assumed a 40*C Q ambient, p}

Assessment ofResponse: This response is fully adequate to resolve the identiSed concern.

A

.)i 2:}.

'N$ '

bl

• 6.:;,

6

g n.<

j

• A 5

. <* Findings and Recommendations: SNL Snds that the licensee has complied with the USNRC request, and no Barther actions on this RAI item are recommended.

3.4 RAIItem 4: Airdrop ADF Synopsis ofQuestion: The licenses assumption of an 11% ADF for air drops was not considered to represent a conservan ve bounding vabe. The licenses was asked to provide a more realistic assessment for these cases.

Synopsis of Response: De licenses has clarined that only two air drop Bre barrier systems am installed in the plant. Each is only 2 to 3 feet in length. The balance ofeach cable is apparently contained in clad conduits. The licenses has also cited a section of the NEC that allows for limited sections of a cable to be neglected in an ampacity assessment (no l more than 10% of the total length of a circuit up to a maximum of 10 feet). The licensee  !

has also cited the available margins for these cases as well. '

l Assessment of Response: The licensee assessment in this regard is consider <A acceptable ,

when taken as a whole. SNL has not previously encountered an invocation of this particular NEC passage, and is skeptical ofit applicability to this situation. However,

, SNL la swayed by three points in particular:

Re licensee ampacity assessment for these applications is based on derating of I scadult base line ampacity limits using sonduit ACF values. This should conservatively bound the equivalent air drop treatment based on derating apan. alt '

ampacity limits using aiuttop ACF values. That is, a penalty" is already paid for the conduit base line ampacity as compared to the open air ampacity limit of a cable. Thus, the combined Islatirs impact of both a conduit and a Sre barrier shoJd be more severe than the impact of a Are barrier alone. The primary concern raised by SNL was that air drops should not be assessed by derating casuit ampacity limits using conduit ACF values. The licenses has clearly not done this.

- The licensee has demonstrated a signi5 cant available margin for both cables impacted by this question. .

- The lengths of cable involved are very short, and hence, any lWi-t heating effects should be mitigated to a large extent.

L Findings and Recommendations: SNL Snds that the licensee response is adequate to conclude that air drop applications at OCNGS have been adequately treated. No further ,

actions on this RAI item are recommended.

3.5 RAIItem 5: Battery Charger Loads Synopsis of Question: ne licensee was asked to address the potential for higher initial i

battery charging current loads under conditions ofinitial battery discharge in its ampacity load assessments.

i Synopsis ofResponse: ne licensee has cited that bonditions of battery discharge are only rarely encountered during 2 year service outages. During such periods current loads are t

7

._. _ _ _ ___. . _ _ _ _ __ _ _ . _ _ -._._., _ ._._ _ _ _ . _ __ , n ._ .- _. _ _ . .. U

. . 1 ,-

. . l

/ monitored by plant procedure Further the licensee has considered a potential 10%

l overload condition on the maximum cunent load.

Assessment of Response: The licensee response is adequate to resolve the identi6ed concern. In particular the licenses has allowed for a 10% overload on a circuit that is by design current limited to the maximum rated cunent. Further, the licensee has cited existing plant procedures that would identify and resolve any overload conditions.

Findings and Recommendations: SNL finds that the licensee response is adequate to resolve the identi6ed concerns. No further actions on this RAI item are recommended.

$)

. z.:.

o..G..

l..* ..

. 'i(-

..s.

.i.

t.

'v iN.i

.C

.l.

vs.

8

>::.t .

8 Mi:

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

n'5@

u _K.f'C . . .

u ...;7 ',

<* 4.0

SUMMARY

OF FINDINGS AND RECOMMENDATIONS SNL 6nds that the !! censes has adequatdy addrased all of the speci6c RAI items forwarded under USNRC cover dated August 26,1996. Further, SNL has reviewed the licensee's example calculations, and Ands that the general methodology applied is an acceptable means of assessing individual plant cable ampacity loads. Finally, SNL Snds that the licenses application of both the TUE and TVA Sre barrier ampacity test roults has been adequatelyjusti6ed, and is appropriate, SNL did identifyt one point of potential concem for which a follow-up RAI is recommended:

SNL Snds that the licensee has not provided adequate discussion of the conduit conductor count ACF factors applied to its analyses. It is recommended that the USNRC ask the licenses to explicitly state which set of NEC values has been used in the analyses; either the pre 1990 50% diversity based values or the post 1990 no-diversity values. Further, it is recommended that the USNRC ask that, for each case in which the NEC 50% diversity-based values have been applied, the licensee either (1) provide an explicit justification for use of the diversity-based NEC ACF values or (2)"pdate the analysis using th: newer post 1990 NEC correction factors in which the diversity assumption has been negated.

This observation will only impact clad conduits with 10 or more conductors, and it is unknown how many such conduits exist at the plant. Given that a signi6 cant level of margin has been demonstrated for most cables, it is not anticipated that this concern will ultimately have a significant impact on the licensee's assessments. It is likely that a very limited scope RAI on this one item could easily resolve the concern.

Also as a part of the review SNL noted one apparent error in the licensee applications:

In determining base line cable ampacity limits, the licensee has adjusted d of the NEC tabulated va'ues using an ACF of 0.91 to reflect an assumed plant ambient of 40'C versus 30'C. Howeves, ampacity limits for the licensee's 5 kV cables appear to have been taken from NEC Table 310-73 and these values are already based on a 40'C ambient. Hence, further temperature adjustment is unnecessary.

This apparent error will impact the ampacity assessments for four speci6c cables; namely, licensee items 15,16,17, and 31. However, the apparent error has resulted in overly conservative estimates of the ampacity limits for these cables, and hence, a change would not impact the ultimate conclusions of the licensee assessments. Given the conservative nature of the error, SNL recommends that no formal resolution is necessary. It is, however, recommended that the apparent discrepancy be brought to the attention of the licensee. It would clearly be in the interest of the licensee to correa this apparent error in its intemal documentation.

9

_ ________ ______ ___