ML20129B671

Rev 0 to Ltr Rept, Review of Plant Analysis of Fire Barrier Ampacity Derating Factors

ML20129B671
Person / Time
Site:	Duane Arnold
Issue date:	04/05/1996
From:	Nowlen S SANDIA NATIONAL LABORATORIES
To:
Shared Package
ML20129B636	List: ML20129B633 ML20129B671
References
NUDOCS 9610230083
Download: ML20129B671 (40)

Text

-- - .. .- . - . - - - . - . . . - . - _ - . _ _ . .. . _ _ _ -

c ...

4 1

1 i

A Review of the Duane Amold Energy Center Analysis of Fire Barrier Ampacity Derating Factors 2

i A Letter Report to the USNRC i ,

w j

Revision 0 t

April 5,1996 i

i Prepared by:

Steve Nowien l Sandia National Laboratories Albuquerque, New Mexico S7185-0737 (505)845-9 0 0 Prepared for.

Ronaldo Jenkins Electrical Engineering Branch  !

OfHce of Nuclear Reactor Regulation U. S. Nuclear Regulatory Conunission Washington, DC 20555 9610230083 PDR 961016 P ADOCK 05000331 PDR

, s' S a

t 4

TABLE OF CONTENTS:

Section Eggg i

FORWARD .................

................................ iii

}

l

1.0 INTRODUCTION

1.1

........'..... 1 Obj ecti ve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 i 1.2 Overview of the Utility Ampacity Derating Approach . . . . . . . . . . . . I

{ '3

.. Organization of Report ..............................

e .2 i

l 2.0 THE B ASIS FOR UTILITY ANALYSIS . . . . . . . . . . . . . . . . . . . . . 3 i 2.1 i

Overvi ew . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 3 2.2 Premiss of Total Heat Rejection Capacity . . . . . . . . . . . . . . . . . . . 3 l 2.3 Limitations to the DAEC Methodology . . . . . . . . . . . . . . . ..... 4 i 2.4 Extrapolation of Experimental Results . . . . . . . . . . . . . . . . . . . . . . 6 i

2.5 Validation of the " Watts /ft" Method . . . . . . . . . . . . . . . . . . . . . . . 7 l 2.6 Basis for Identification of Continuously Energized Power Cables . . . 7

2.7

, Summary of Methodology Limitations and Concerns . . . . . . . . . . . 8 i

j 3.0 A BRIEF REVIEW OF THE SPECIFIC CASES CITED BY DAEC . . . . . 9 j 3.1

?

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.2 Example Analysis for Cable Trsy 2H2D . . . . . . . . . . . . . . . . . . . . 9 3.3 Example Analysis for Conduit IC979 . . . . . . . . . . . . . . . . . . .. I1 i 4.0 a

SUMMARY

OF REVIEW FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . I '-

e k

5.0 REFERENCES

.......................................... 1.

)

1 i

1 1

l 4

4 i

i l

d a

4 1

i i

i i

11

'+ c FORWARD no United States Nuclear Regulatory Commission (USNRC) has of Sandia National Laboratories (SNL) in the review of utility submittals with fire protection and electrical engineering. This letter report docume of a SNL review of a set of submittals from the Duane Arnold Energy Cerit .

(DAEC). Dese submittals deal with the issues of normo Lag 330-1 fire bar endurance assessments and the assessment of ampacity loads for p and conduits. These documents were submitted by the utility in response t Generic Letter 92-08 and in response to a subsequent USNRC Request fo Information (RAI). His review has considered only those portions of th submittal directly related to the issue of ampacity derstmg. His work was p as Task Order 9, Subtask I of USNRC JCN J2017.

0 9

... I 111

^ ' ~

,+

i

• l k I l.0 INTRODUCTION

1.1 Objective s

In response to USNRC Generic Letter 92-08, the Duane Amold Energy Center

(DAEC) provided documentation of the utility position regarding both the fire j endurance rating and ampacity derating factors associated with its installed ' fire barrier j systems. The objective of this review was to assess the adequacy of those po.rtions of l the utility submittal related to the issue of fire barrier ampacity derating factors. In i particular, the submittals included documentation of an analytical methodology used to assess the adequacy ofin-plant cable ampacity factors for its various Appendix R l cable tray and conduit fire barrier systems. Also included were sample calculations  !

for one cable tray and for one conduit.  !

i ne submittals reviewed were documented m an untial response to Generic Letter  !

l '

92-08 provided by the utility and in a subsequent utility response to a USNRC i

Request for Additional Information (RAI). He relevant documents reviewed are:

-r Letter, February 14,1994 (item NG-94-0563), J. F. Franz, DAEC, to L.

J. Callan, USNRC, (with enclosures).

Letter, June 2,1995 (item NG 95-1223), J. F. Franz, DAEC'to W. T.

Russell, USNRC (with one ***achment).

SNL was requested to review these submittals under the terms of the general technical support contract JCN J-2017, Task Order 9, Subtask 1. nis letter report documents the initial results of this review. He intent of this review was to provide support to the USNRC in determining the adequacy of the utility submittals, and in the potential development of a supplemental RAL Based on the results of this review, it is recommended that such a request be pursued.

• 1.2 Overview of the Utility Ampacity Derstmg Approach The consideration of ampacity derating factors for fire barriers at DAEC is based on an analytical method with limited validation through available experimental data. De utility has noted that it plans to update its analyses once the results of the NEI normo-Las ampecity dernems test program become available. No testing has apparently been performed by the utility, and at the current time, no such testmg is planned.

For its assessment of ampacity derstms, the utility has swamined each cable tray or conduit which is both protected by a fire barrier system and houses at least one continuously energized power cable. He DAEC method is basically a variation of the methodology commonly referred to as the " Watts /ft" method. He analysis follows a simple two-step process:

Step 1: For each tray or conduit, the actual total heat load (due to cable resistance heating effects) is calculated based on the actual operating conditions m

s, ,

2.0 THE BASIS FOR UTILITY ANALYSIS 2.1 Overview he DAEC methodology is apparently based on a paper presented by Esteves Bechtel Power Corp., in 1982 (utility reference 6.1) and republished by IEEE in

[1]. However, while the basic methods used by Esteves are valid for the p'urp which he put them, the use of this methodology in the manner employed by DAEC not appropriate.

In particular, Esteves performed some simple analyses of the pioneering cable ampacity work of Stolpe [2]. Esteves' interest was in estimatmg the anticipated ampacity impact of fire barriers and fire penetration seals. His work was based on very simple models of the supplemental impact of a fire barrier (or fire stop) on heat e

transfer in a cable tray of the type tested by Stolpe. Using these simple calculations Esteves estimated the ampacity derating impact for certain specific fire barrier sys Inherent in Esteves' work is the base assumption that the cable tray and cable loa remained fixed as per the Stolpe tests. He utility's extension of Esteves' work to the assessment of cable ampacity adequacy for actual cable trays is incomplete and inappropriate because it violates this inherent assumption. He following sections outline the DAEC analysis methodology and identifty its limitations. He basis for the SNL finding that this methodology is inadequate and inappropriate to the problem of '

cable ampacity derating assessments is also provided. In particular, the inherent inability of this method to assess individual cable ampacity limits is discussed in

, detail.

2.2 Premiss of Total Heat Rejection Capacity he ampacity darstag assessment methodology employed by DAEC is based on an evaluation of the actual in-plant overall heat load for specific cable trays and conduits.

ne approach is largely analytical, although certain critical parameters in the analyses are based on limited test results as available to the utility through public test reports.

No supplemental testing by the utility has been performed, and at the current time, no such testas is apparently pisnoed. He utility does cite that once the NEI test results become available, a review of the ampacity calculations will be performed.

In order to understand the utility analyses, it is critical to recognias that the utility has

• not applied ampacity dorating factors direcdy to individual cables. Rasher, the utility has analyzed cable trays and conduits as composite systema. In effect, the ampacity derating factor is applied to the overall heat load of the system rather than to the ampacity of individual cables within the system. Dat is, the utility analyses assume that so long as the total thermal load for the cable tray or conduit as a system is

' within allowable limits, then one can conclude that the individual cables within that tray or conduit are all operating at acceptable limits. As will be noted below, this is not a valid assumption.

e 1

i, , e a '

l associated with each " continuously energized power cable." nat is, the heat load introduced by each individual cable is calculated and then the individual l loads are summed to obtain the overall heat load for a given tray or conduit.

Each individual cable heat load is calculated based on the actual in-plant current load and the cable's resistance per foot of length. The overall heat load

, is presented as the Watts of heat generated per linear foot of cable tray.

l (Hence, this approach is onen called the " Watts /ft" method.) '

Step 2: Once these overall in-plant heat loads have been calculated, they are compared to the estimated " permissible thermal output" for a tray or conduit.

I In the case of DAEC the " permissible" load is calculated based on a

} combination of one TSI/ITL test set and tabulated ampacity values. So long as j the actual heat load of the tray (or conduit) does not approach the heat rejection i

capacity of the generic tray (or conduit), then the ampacity of the protected cables is considered acceptable.

i ne basic premiss of the DAEC submittal appears to be that the acceptability of

! ampacity loads placed on individual in-plant cables can be demonstrated by showing i that the total in-plant heat load for the overall physical system housing the cable does j not exceed the estimated total heat rejection capacity of that physical system. SNL i takes issue with this assumption. Dat is, as discussed further below, the " Watts /A" l method is insufficient in and ofitself to demonstrate the adequacy of individual cable i

ampacity factors.

s -

j 1.3 Organization of Report his review has focussed on an assessment of the acceptability of the utility ampacity derating analyses and on a review of the two specific case examples provided by the utility. Section 2 presents a review of the basic utility paalysis methodology and identifies the technical shortcomings and concerns regarding the utility application of this methodology. Section 3 provides a review of the two sample calculations provided by the utility and identifies concerns associated with certain aspects of these analyses. It is expected that additional technical evaluations for each of the cases cited by the utility will be required to compless the assessments, and hence, thees reviews were limited to identification of potential problem areas, rather than to detailed case evn=iameions. Section 4 sn=marizes the SNL recommendations regarding the need for additional information to support the final assessment of the utility analyses.

l 2

)

y ,

subject tray or conduit. As a part of this e system on heat transfer behavior must be es .:ulated. For DA ampacity derating test set from 1982 is used as the. basis The for thi

" permissible thermal output" is then calculated sby mathe transfer effect onto the tabulated ampacity values yfor a given cable t or conduit eat rate, and this heat load is reduced mathem effects of the fire barrier. For the cable trays the generic er ved pow by Stolpe for a densely packed and uniformly powered or cable conduits, the tabulated ampacity limits for the actual installed ca whole. His calculation is based on a simpl applied to each of the individual " continuously energized power c that in the case of DAEC an additional .

. Note mult a e actual heat

. load is applied to account for potential intermittent loads.

The final assessment is then based on a simple and direct com total in-plant heat load for the tray or conduit to the estimated e thermal "permis the cable tray or conduit as a wholem allowable is lower tha heat load, then the ampacities of the~ individual cables are c .

2.3 Limitations to the'DAEC Methodology ne methodology used by DAEC is fundamentally limited and is inc total heat rejection capacity analysisent methodo of the behavior ofindividual cables within that system.overa inside the protected envelope. However, eve both the physical and electrical cable loading conditions in the g to those of the speci5c application (i.e., a dense mass es . of unifo vicinity of an individual powered cable, parti cable tray in which only a very few cables are actually energized ampacity it is the hot-spot behavior, not the average temperature beha drives the problem. Hence, this mesbod provides no assurance tha individual cable will not be overloaded to the point of damage ould Such easily amas without the overall " permissible thermal output" limit exceeded.

One way to illustrate this flaw in the methodology is to consider a hy tray with just one single power cable inside ofit. If the cable n a 12" were housed 4

wide cable tray, then a given ampacity limit could be calculated based on the

" permissible thermal output" of a 12" tray. If this same cable were then m 48" cable tray, according to the " Watts /ft" method a four-fold increase in the

" permissible thermal output" would result. The thermal i limit output of a cab the square of current, and hence, a two-fold increase in the allowable ampac ty would also result. If this " logic" is carried out further, any amperage could be

justified for any cable provided the tray holding th (

' behavior.

ne DAEC analyses do not address this fundamental methodological h limita provide no assessment of the anticipated behavior of individual cable utility analyses have not demonstrated that th.e ampacity values associat

• given cable are within acceptable limits, ne analyses as presented by incomplete h aeana they have only provided an assessment of the overall behavior of their cable trays and conduits. De utility must i provide suppleme analyses to ensure that individ2al cable ampacities remain within accept including consideration of me impact of the fire barrier system on the publi ampacity limits for thom cables. He " permissible thermal output" or " W method is simply not antuate to assassing this question.

In addition to the fundamental shortcomings of the methodology, there are als additional concerns which were identified during this review. In particular, th methodology provides no mechanism for including conaderation of the follow parameters, each of which can be expected to significantly impact bo cable ampacity limits and overall cable system heat rejection capacities:

Cable Lo Ann F#ects De DAEC analysis methodology does not consider the impact of cable loading on the allowable heat loads except extent that Stolpe's work does address different cable depths of fill. In g the " Watts /ft" method assumes that cable loading effectsi are largely irrele to the overall heat rejection capacity of the cable tray or conduit system. I expected that the total tray heat rejection capacity would be dependent various factors, especially including the power density within the cable mass nose factors should be accounted for in the methodology, or it should be demonstrated that these factors are not important to the analysis.

Cable Diversity F#ects: no DAEC analysis method provides for no significant treatment of cable diversity effects on the total allowable All of the available ampacity tests cited by DAEC are based on cable tr which all of the cables are powered uniformly (note that even though the TSI/ITL tests involved three cable mim, the power to each cable was set so as to maintain a uniform power density across the tray). In actual application cable trays will contain a mixture of loaded and unloaded cables. In g one typically assumes that diversity will introduce more margin into design. His is true so long as one is considering the behavior lof ind cables. However, one must recall that the " Watts /ft" method is only p an

==w" ment of the overall behavior of the cable tray as a system. De 5

_9

. i o 4

" Watts /ft" methodology would clearly lead to erroneous results for cases involving diverse loads.

Consider, for example, two cases involving a cable tray loaded with 50 power cables (an arbitrary number). In the first case, we assume that all 50 of the cables are powered uniformly. In the second case, we will assume that only a single cable is powered. He " Watts /ft" methodology would assume that the overall heat rejection capacity for these two cases would be ideritical. .

Hence, in effect, the heat IJad generated by the one cable which is powered in both cases could increase by a factor of 50 from case 1 to case 2, and still remain within the same overall heat load. This would imply a 7-fold increase in the ampacity of that cable from case I to case 2. (Heating load is proportional to the square of current so the " allowable" current increase would be given by the square root of 50, or 7.07 times the case 1 ampacity.) nis is clearly unrealistic. As the power of the one cable were increased a limit would quickly be reached beyond which the insulating effects of the surrounding .

unpowered cables would increase the powered cable temperature beyond its operating limits (90*C). The " Watts /ft" methodology in and ofitself would t conclude that each of these two cases was equally acceptable. I ne total overall heat rejection capacity of the cable tray as a system would very likely be reduced in cases involving diverse cable loads, even though the ampacity of the individual cables which are powered could be increased due to diversity arguments. De " Watts /ft" method as employed by DAEC simply contains no mechanism for assessing this effect.

2.4 Extrapolation of Experimental Results

' Additional concern regarding the utility analysis methodology derives from the fact that the utility has extrapolated a rather limited data base to all of their applications. It would appear that the extrapolation of the experunental results has been performed without adequate technical justification or validation. Without further technical justification, the utility extrapolation basis is questioned. Two specific areas of concern are identified here.

First, in the DAEC calculations, a single ampacity test set (one clad and one base line test) is used to characterize the added insulation effect of the fire barrier system, and this value is assumed to remain fixed for all subsequent configurations. Dat is one experiment is used to characterize the additional insulating effect for all cable trays, and the same experiment is used to characterize the effect for all conduitsi as well.

His treatment fails to recognize that there are many important factors which l

contribute to the barrier's overall impact. In particular, as the cable loadmg changes, l and as diversity in cable power levels is introduced, changes in both the convective i and radiative behaviors within the cable tray are expected. nose factors are not considered in the DAEC calculations.

A second concern arises when the assumed base line allowable heat loads are considered. In the DAEC analysis, these values are taken directly from Esteves (1]

who in turn calculated them based on Stolpe's work [2]. He assumption that these 3

values will apply to any tray as a function only of the fill depth is inappropriate. In l 6

e 6 particular, Stolpe's original work powered all of the cables in the test tray that the heat generation was distributed evenly throughout the tray. In realit particular in the DAEC cases, the cable power loads are highly diverse.

has stats,d that unpowered cables will act as a heat sink acting to cool powered cable. His is incorrect in the steady state condition. In fact, the s unpowered cables will act as thermal insulation causing an increase in the tem of the energized cables a compared to the case of a cable in open air or alone cable tray. In addition, the cable loading and power distributions will directly both the surface area and surface temperature of the cable mass. Mose will in tur directly impact the rates of both radiative and convective heat transfer away from cables. De DAEC m.ethod fails to account for these factors.

It should be noted that the DAEC application of the " Watts /ft" method is somewha different from that which has been observed elsewhere (see for example the SNL review of ampacity derating submittals from Palo Verde,9/27/94, and the Bechtel engineering design guide upon which the Palo Verde analyses were based). In o applications the " permissible thermal load" has been based directly on that which actually measured in an ampacity test for the clad test article. In the DAEC su the " permissible thermal loads" for the clad case are based on adjustment of the

" permissible thermal loads" allowed in the base cable ampacity tables (as re Esteves for cable trays and based on the NEC tables for conduits). He ampac experiments are used only to assess the additional impact of the fire barrier on the thermal resistance between the cables and the environment. Given that there is generally considered to be some margin of conservatism in the base ampacity tab the DAEC approach would be the more conservative. (Note that in the specific c tray calculation cited by DAEC this did in fact prove true as discussed in Section 3 below.)

2.5 Validation of the " Watts /ft" Method The utility has provided no direct results for the validation of the overall metho used in its analyses. Even given the. methodology limitations as outlined above, som validation of the method should be demonstrated. A statement that the "results of this mie'=- compere g favorably with results in references 6.8 through 6.12." However, no details of these comparisons are provided.

2.6 Basis for Identi6 cation of Contmuously Energized Power Cables In performing its calculations, the utility has only considered " continuously energ power cables" as sources of heet for the protected envelopes. Hence, only trays containmg such cables have been cor.sidered, all other cables in the tray ne considere to add no heat to the envelope, and trays with no such cables have not been analyz It is not clear that this assumption as implemented by the utility is appropriate.

In general, SNL does not take exception to the assumption that control and instrumentation cables will contribute no significant heat to the envelope. He exclusion of cables for items such as MOV's which would generally only be activated 7

e .

clear what basis was used to identify " cont ,

he concern here is whether or not all modes of plant operati this process. If only normal operation at full power is considere n of been overlooked. For example, during shu cables might be in operation than those active during power opera .

barrier envelopes, this may represent the limiting condition an envelope contains no power cables energized during normal opera during certain emergency situations cables which are normall energized for extended periods (i.e. hours or days). One such ex off-site power. It appears that such cables wo utility analysis.

l It is recommended that the utility be asked to provide additiona by which " continuously energized power cab  :

I manner in which various modes of plant operation were factored into 2.7 Summary of Methodology Limitations and Concerns ne DAEC analysis methodology is considered inherently inadequate the acceptability of ampacity loading factors for the cables installed given individual cable. His is an inherent limita cannot easily be corrected. He utility gates that it will reassess the anal the results of data expected to be forthcoming from NEL hise would n fundamental shortcomings in the DAEC analysis methodology which h the ampacity performance of individual cables as It was also noted that the utility has not provided a sufficient discussio whether or not adequate consideration has been given to all cables represent contributors to a cable tray's heat load. SNL does not take excepti exclusion Aom the best load analysis of control and instrumentation ca power cables to intermittent devices such as MOV's. However, it is unclear or not all possible modes of plant operation have been conadered, includ involving plant shutdown, startup, and emergency response proced off-site power) which might be active for extended periods.

Certain other con: erns related to the extrapolation oflimited experimenta of the DAEC cables and conduits were also raised. However, given the fu limitations and shortcomings in the overall DAEC methodology, these con considered of secondary importance.

8

. j 3.0 .A.BKIEFiFEFIEW Of"DlE:SEE3FJr CWRES CfTED BY DAEC 3.1 Oversaw no ud!ny sdtmimal!taminciinkhwe spr.ida cmompe 2nalyses, one for a cable tray and one for a rr.auish ~ngrimmyhraaddtes sesw was placed on the overall methodology, in /.arp taso ihwam h were mdeddogy was considered inherently deficient. However *ftunsenmr disussw x=rsais ism associated with the individual  !

calculations which were identified & ming a review of those calculations. In general it was found that insufficient information has been provided by the utility to perform a complete review of the calculations nose areas in which additional information is ,

needed to support a thorough review are identified below. l 3.2 Example Analysis for Cable Tray 2H2D ne first example given by the utility is for cable tray 2H2D. His tray apparently contains just two " continuously energized power cables." Each cable is a 3-conductor  !

350 MCM cable carrying 247.22 amps per conductor.

The analysis bgins by considenna the heat load determmed experimentally for a 12" wide cable tray as aported in a normal Science Inc. report from 1982 (TSI/ITL report 82-5-355F). ne test data (the measure currents and temperatures) are used to estimate the added thermal resistance effect of the fire barrier. Den the " permissible thermal output" for a protected tray is estimated by mathematically " adding" this calculated barrier resistance to th.e base line tray tests performed by Stolpe in his pioneenns work on cable' tray ampacity. He " permissible" heat load calculated by the utility is 52.6 W/ft for the 18" wide DAEC cable tray.

It should be noted that is one respect this approach is more conservative than other potential approaches which might be taken in the Watts /ft method. In particular, the

" permissible" heat loads calculated by DAEC are based on Stolpe's base line values of the cable tray heating effect rather than on the measured heat loads from the actual clad ampacity test of the TSI test set. In this case, the heat loads reported by Stolpe were significantly lower than those reported in the TSI/ITL base line test; Stolpe load of 61.6 W/ft* as compared to the TSI/ITL test value of 83.81 W/ft". Hence, the u:ility analysis is conservative in this regard.

The utility then turns to the actual cable tray under analysis. ne heat load for each of the two " continuously energized power cables" is calculated. The total heat load is multiplied by a factor of 1.25 to " add conservatism to account for intermittent load..."

Using the actual in-plant cable currents, an " actual" heat load of 17.8 W/ft is calculated for this tray. Tne " permissible" and " actual" heat loads are then compared.

Since the actual heat load is the smaller,17.8 as compared to 52.6, it is concluded that "no funber action is required."

On the surface, this analysis would indicate that this tray has significant available margin. That is, the " actual" heat load is less than 34% of the " permissible" heat load.

However, this indication of significant apparent marge is very misleading. As noted 9

.s .

above, it is SNL's fmding that this analysis has not demonstrated that these cables ar operating within acceptable ampacity limits. To illustrate the basis for this find consider that the ICEA P-54-440 "Ampacity of Cables in Open-top Cable Tr the ampacity limit for a 3-conductor 350 MCM cable in a tray of 1.5" depth of fill a 315 amps.' he DAEC cables are canying 247.22 amps. Hence, even without a m detailed analysis, one could conclude that there is a nominal ampacity margin of a least 21.5% available for these cables in the absence of any barrier systems: Ho it is this same margin which must encompass the fire barrier ampacity derating effe Based on the available data ampacity derating factors for a three hour normo-Lag barrier system should be on the order of 35% or more (depending on the barrier configuration).2 Hence, the nominal available ampacity margin is not sufficient to cover the estimated generic ampacity derating impact and further analyses of the DAEC cable ampacities se needed.

Recognize that this very simple comparison has not considered the various factors which the utility might wish to consider in a more thorough analysis. It should not be concluded on the basis of this comparison that the DAEC ampacity factors are not sufficient. Rather, this comparison simply highlights the fact that the Watts /ft method l

is insufficient to assess the ampacity limits of individual cables. For example, the utility might wish to credit diversity arguments for providing some added margin give that only two of the cables in the tray are continuously energized. Other mitigating

, factors may also come into play. Rose factors might contribute to an assessment that a larger ampacity margin is avaiable for these cables. In any case, the utility analysis '

must consider the current carrying capacity of the individual cables, not just the cable tray as a whole. l i

i nere is also a second way to illustrate the basic concern being raised by SNL in this

{

review. Consider that the utility estimste of the " permissible thermal output" for this '

cable tray was 52.6 W/ft. If this number is divided by 1.25 (consistent with the utility approach of multiplying the actual loads by 1.25 for conservatism) a value of 42.1 W/ft is obtained. This means that each of the individual conductors in the two, three-conductor continuously energized power cables could carry 1/6 of this thermal I load, or approximately 7 W/ft, and still the tray could remain within the overall

" permissible thermal output." Using the electrical resistance value for copper (as per l the utility analysis) this would correspond to a current in each conductor of 424 amps.

His mesas that any value of current lower than 424 amps would mean that the actual thermal load would be lower than the estimated " permissible thermal load" and l

1

• His value is the same in Tables 3-6,3-9 and 3-12 of the standard so the voltage rating of the cable, normally a consideration in the ampacity analysis, is not a factor in thiscase.

2 This value is based on the minimum ampacity dorating estimate identified in the SNL/USNRC tests documented in SAND 94-0146 [3]. Note that the ampacity values derived from those tests are not considered appropriate for use in the evaluation of actual cable tray application due to problems with the test article configuration. He SAND 94-0146 results are cited as rough estimated of the anticipated impact only.

10

. t acemdas:tretke.udihy.mr.shnh"moffuritem wion is needed." Note that this value,424 ampsfis T34%ci ikupernrap.umpads Ihmt of 315 amps published in the ICEA tables <fordtiamahiramSdiiratanth&airik tiny fill. Under any circumstances these ampacity:InWhwastf thramsdeuf ummettable for this cable. This hypothetical exarnple strx4 ii&rmxtsthe ibndturenmi%ortcoming and inadequacy of the

" Watts /F :matteri!

Another pona cdestescrearmpodhgh example analysis is the fact that the utility is

~

basing its analymcesone ces perftenersy the manufacturer, TSI, which has since been discredited. The manufacazed ust cited by DAEC has been the focus of various concerns related to the configuration of the test articles, the acceptability of the test methods, and the processing of the test results. In particular, these early TSI ampacity tests were not performed consistent with current testmg practices. The use of this test, or indeed any of the TSI tests for which significant technical concems have been raised, u the basis for current cable ampacity assessments is considered inappropriate. He utility should be asked to cite an ampacity test which is based on currently accepted test practices (for example the IEEE P848 draft standard) and which encompasses the configurations of the 3-hr fire barrier systems in use at DAEC (for example single 1" versus double 1/2" layer systems of the appropriate material density and thickness).

3.3 Example Analysis for Conduit IC979 ne utility has provided one example ofits ampacity dorating analysis of conduits by presenting the calculation for Conduit IC979. In certam respects the DAEC conduit analysis method is more ramaaaahle than is the corresponding DAEC cable tray method. However, the conduit analysis method still fails to consider the appropriateness of ampacity values for the individual cables being examined. Hence, the method is not considered an appropriate basis of analysis, and is not considered adequate to justify the ampacity factors for the cables installed at DAEC.

Here is one fundamental difference in the approach to analysis which results in the finding that the conduit analyses are more reasonable that the tray analyses. Recall that in the cable tray analysis the " permissible thermal load" was estimated based on Stolpe's tests in which the cable tray was heavily loaded and a uniform current was applied to all of the cables present (this creates a uniform " power density" which is the foundation of the Stolpe ampacity load values). It was than assumed that this same

" permissible thermal load" could be applied directly to the case of a tray with just two j

" continuously energized power cables." With just two active cables in an 18" wide tray the condition of uniform power density is clearly violated, and the extrapolation must be considered inappropriate. In contrast, for conduits DAEC has estimated the

" permissible thermal load" using tabulated NEC ampacity limits for ti4e cables actually installed in the subject conduit. While this may seem a mmor point, it is actually a fundamental difference in analysis approach.

Returning to the utility example, the conduit IC979 is described as housing just two cables; one 3-conductor 350 MCM and one 2-conductor #4/0 cable. Based on the NEC ampacity tables, the allowable ampacity limit for each cable in the absence of a 11

s .

fire barrier is calculated (236.6 and 171.1 amps per conductor respectively). This leads to an estimated permissible thermal output for-the unprotected conduit of 10.257 W/ft (see steps 5.1.3 - 5.1.5).

The utility has not justified the underlying assumptions which drive the next analy steps (steps 5.1.6 and 5.1.7). In step 5.1.6 the utility makes the basic assumption that the thermal impact of the fire barrier system for the conduit can be based oh the test results obtained for the 12" cable tray cited previously (the value is taken directly from step 4.1.4 of the cable tray analysis). Dat is, the same TSI/ITL cable tray ampacity test cited in its cable tray example is used to assess the added thermal impact of the fire barrier for a conduit as well. His assumption hu not been justified by the utility, and is of questionable merit. He fundamental thermal configuration is significantly different for cables in cable trays and those in conduits. He direct application of a value obtained in a cable tray experiment to a conduit analysis must be justified.

It would also appear that the utility has made an error in its calculations. In particular, in steps 5.1.6 and 5.1.7 the utility ha apparently adjusted the nominal fire barrier thermal conductance value to account for the circular geometry of the conduit in comparison to the rectangular geometry of the cable tray. This correction does not appear to have been properly performed.' The utility method makes an inherent assumption that all of the fire barrier effect is related to added thermal resistance due to conduction heat transfer through the fire barrier material. This is reflected in the manner by which the utility estimanas the thermal conductivity of the fire barrier material (step 5.1.6), and then calculates an equivalent thermal conductance for an annular region of 1" thickness (step 5.1.7). His ignores the fact that much of the barrier's thennal effect is related to degradation in the radiative " access" of the cables to the ambient and to interruption of the convective heat transfer process. His conversion appears inappropriate.

In step 5.1.8 the modified " permissible thermal output" is calculated as 8.74 W/ft.

Note that give this value, and the unprotected conduit heat output of 10.275 W/ft (from step 5.1.5) it is quite straight-forward to determine the effective ampacity derating factor being estimated for the protected conduit. Recall that the heat load is proportional to the square of current. Hence, the effective ampacity correcnon fac*ar (ACF) can be calculated as:

e2,, 4 , i 10. 257 f u.7a = 0.923 ACF = ) 73 and the ampacity deranna factor (ADF) is then given as:

ADF = (1. 0 - ACF) = . 077 = 7.7 %

Hence, while a rather round-about approach to the problem has been taken, the net effect of the utility conduit analysis is to assume that the ampacity derating impact of 12 l

1

= 0 ,

l I

l its 3-hr conduit fire barrier systems is 7.7%. Given that the conduit is still analyzed as a system this means that so long as the cables as a group have an average of at least l-j 7.7% margin in comparison to the NEC conduit ampacity limits, the conduit will be

judged by DAEC to be acceptable. Note that this ADF value will change for each

! conduit analyzed because the base thermal loads will change.

l In its subsequent analysis, the utility considers the actual ampacity loads fo'r each of l the two cables under analysis, and compares the at.tual heat load to the estimated j " permissible thermal output." Because the actual is less than the " permissible," the i conduit is judged acceptable.

L It is interesting to note that the actual cable ampacities cited 3b the utility for these two cables are 223.2 and 48.88 amps per conductor for the 350 MCM and #4/0 cables j respectively. Hence, in compprison to the NEC base conduit ampacity limits cited in

step 5.1.3 (236.6 and 171.1 amps respectively), these two cables have an available ,

j margin of 5.7% and 71.4% respectively. In the case of the 4/0 cable, the available l l margin of 71.4% would clearly be well within even the most conservative Thermo-Lag l conduit ADF values noted to date. However, for the 350 MCM cable, the available i margin of 5.7% might not cover the potential ampacity derstmg impact of the fire i barrier system. Even tests of one-hour conduit barriers have resulted in ADF values of l this magnitude or greater (see for example the Texas Utilities test results).-

In summary, the utility conduit analysis has again failed.co consider the actual ,

performance of each of the individual cables in comparison to published ampacity limits. Rather, because the conduit is analyzed as a system, in effect, the 350 MCM t

cable is able to " borrow" some of the available margin from the #4/0 cable without justification. In some ======, the " Watts /ft" methodology as applied by DAEC to conduits is more ramaaankle than the same methodology as applied to cable trays.

This is because the conduit analysis makes more direct use of the ampacity limits of the actual cables under analysis in its estimation of the " permissible thermal output" than does the cable tray analysia. However, the methodology still fails to consider the behavior of each individual cable in comparison to the ampacity limits of that cable in the presence of the fire barrier system.  ;

l l

l l

l l

l 13

f.5 -

4.0

SUMMARY

OF REVIEW FINDINGS l

j The SNL review hu assessed the general methodology employed by DAEC

!- evaluation of cable ampacity factors, and has examined the two spe j provided by the utility. Based on this review, SNL finds that the method i employed by DAEC in its evaluation of cable ampacity factors is inher to demonstrate that individual cable ampacity factors are within acceptable ..

l 1

i ne DAEC method is, ein ' ffect, a variation of the methodology often referred t

" Watts /ft" method. He fundamental shortcoming of this method is that it o provides a general assessment of the overall heat load for a cable tray or condu whole. No assessment of the individual cable operating conditions is provided j fact, the methodology might indicate a significant apparent ampacity margin i

when the cables within the system are actually operstmg well above accepta ampacity limits. He DAEC analyses are considered incomplete, and are no to demonstrate the existence of appropriate cable ampacity factors.

It was also noted that, at least for the two example cased provided by D limited margin was apparently incorporated into the design and selection of th used at DAEC. In particular, the one cable tray case provided by the utility ind a nominal available ampacity margin of approximately 21.5% for the two cables tray under analysis (before conaderation is given to the fire barrier impact). In the case of the conduit analysis, one of the two cables cited by the utility had an availa margin ofjust 5.7% (again, before consideration of the fire barrier impact). De ampacity derstmg impact of the 3-hour fire barriers in use at DAEC could eas

- exceed these' nominal margins, and hence, further analysis of the DAEC cable ampacities is clearly needed.

It was also noted that the level of detail provided with regards to the identification of individual cable characteristics was insufficient to support a complete review of the utility analyses. In particular, the utility has considered ampacity only for those cabl identified as " continuously energized power cables." He basis for identifying s cables has not been provided by the utility (for example, have all modes of powe operation been considered, and has the potential operation of various backup, emergency, and shutdown systems for extended time periods been considered). It is recommended that the utility be asked to provide a listing of all affected power cables.

which includes identification of the physical characteristics of each cable (size, insulation type, voltage rating, etc.) in addition to the actual in-plant ampacity fac Without this information, the individual cable heat load calculations cannot be reviewed in detail.

In addition to these findings related to the general applicability and acceptability o overall utility method, SNL has also identified certain more specific concerns associated with the example analyses provided by the utility. With regards to the example cable tray analysis, SNL noted the following concerns:

he DAEC analysis of cable trays is based entirely on a single fire barrier ampacity test set (one base line and one clad test) performed by 14

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

. t TSUD1:m '1981 The < manufacture:*nms of this vintage have been the subject of.stgnificant cdticismrnlam:l Wthe' rent article configuration, the test procedures uilined anddhe analysis (nfahe test data. Use of these tests as the bash!for, current euduatirnsrbsmpuitg factors is considered inappropriate and

~in con'lirswith dhe timenhnf tira Gamme Letter 92-08.

Me mtuity %s avamis titat % heat rejection behavior of a ' densely .

padked ' cum' tray wikurirfarn omant loads (the Stolpe tests) is identical to that of a diversa cohle tray o*Hu just two " continuously energized power cables.' His. assumption has not been justified, and is considered inappropriate.

ne example cable tray analysis has provided no direct comparison of the actual eable ampacity values to accepted standard ampacity limits (tabulated m:npacities). De utility should provide a direct assessment of the ampacity impact of the fire barrier system on the tabulated ampacity values for specific cables in use at DAEC.

In the case of the conduit analyses, four fundamental areas of concern were noted:

ne thermal effects of the dre barrier on the conduits is assumed to be 1 identical to the impact of the fire barrier on cable trays, his assumption is not  !

justified and as applied by DAEC appears to yield potentially non-conservative  !

results. Conduits and cable trays involve fundamentally different heat transfer '

behaviors. He assumption.that the impact of a fire barrier will be the same for both items is conadored inappropriate.

De conversion by DAEC of the fire barrier thermal impact from a rectangular to a radial geometry is performed in an inappropriate manner. De  !

DAEC conversion inherendy assumes that thermal conduction through the fire l barrier is the predomiamat mechanism contributing to the fire barriers insulating I effect. However, both radiative and convective heat transfer are critical aspects of the barrier's thermal effects. De DAEC methodology does not consider these effects. He DAEC conversion needs to be justified, or it needs to be demonstrated that this conversion is conservative.

Based on the DAEC methodology, an effective ADF of 7.7% for the example conduit has been calculated. Based on the available test results for conduits, this value may not be sufficient to cover the ADF effect of a 3-hour

normo-Las fire barrier system (even some 1-bour normo-Lag conduit systems have resulted in ADF values higher than this). Hence, the combination of inappropriate extrapolation of test results, and inappropriate geometric conversions appears to have resulted in a non-conservative estunees of the fire barrier impact. He utility should reconsider its assessments based on actual deranns tests for three-hour conduit fire barriers of the type and configuration used at DAEC.

15

As in the case of the cable tray analysis, the DAEC methodology fai to consider the performance of individual cables in comparison to accep ampacity performance limits, including the impact of the fire barrier system Rather, the cables are analyzed only as a composite system, and non-conservative results can easily be obtained.

e l

l 1

I 16

. 4

5.0 REFERENCES

1. Esteves, Oscar M., "Derating Cables in Trays Traversing Firestops or Wrapped in Fireproofing," IEEE Transactions on Power Apparatus and Systems, Vol. PAS-102, No. 6, June 1983, Pgs 1478-1481.

2. Stolpe, J., "Apacities for Cables in Randomly Filled Trays," IEEE Transactions on Power Apparatus and Systems, Vol. PAS-90, May 1972, Pgs 962-973.

3. Nowlen, S. P., An Evduation of the Fin BerierSystem Themo-Log 330-1, Sandia National Laboratories, SAND 94-0146, September 1994.

I b.

17

sl.

lb 1 _ .

v IJ i

• L t

1 -

2,I

a

• 1 CABLE DERATING ti -

PRACTICE

!jl * .

.J f n ,

5) iI

a. J l1 Le 15

.1 gi . v.ririu j

ATTACIM NT 1(b)

^

< l{- ^ i i j ^ 10-29-84 Revised Deratino of Covered Trav NL M',!

^ san j 8-#-83 General Ravision MR El 4#l'

( AA 7=7 75 sea. I mars rseuse As a we oasion suios navistoess um aos I4

, av  ; ms t" An

omsesas Jos n. 57AND An3 8'8""' CAL ' ~

sm Cable Derating Practice OE5fGN GUIDr. le SPC E.2.s.4 i

%# py ~

sacar i na 27

, , , _ * **seems4 .. ,, 05000528 y . .. .:.r. . .

._~ ~ .:~>_~ = -- :: -~---

. e RZYISTON 2

$1TMMARY Oy RIVISIONS i l 3 .-

J

1. General editing of entire document.

2. Changed " ventilated" cable trays to " opes top" cable J  : rays :hrougneut to confers to ICIA.

f 3. Paragraph 3.6.4 Deleted secties en dorating $ for solid tray covers, added reference to paragraphs 3.7, 3.8, 3.9 where 3.9 is a new sostise titled "Addtional

,.l I Derating !st Caeles Rested in Open Top Tray with Solid Covers". ,

s 4. Paragraph 3.6.6 b): Deleted derattag for solid tray g[ esvers free sampla calculaties.

]E 5. Added notations that tray severs should be removed prior to applying firestop asterials or enclosing raceway with fire protecting satorial, 1

6. Paragraph 3.83 Revised to inslade derettag for a 3-hour fjl fire rating fdr Therme-Lag. General revision to g separate discussion of Thorse-Las and seramic fiber blankats. ,

l'(

g w

7. Beaumbered "J.9 General Precautions" to "3.10 General Presauciosa".

e 1^

r l [t I il Ig I

.l 83 11

'it NUMetR E2. 6. 4 *a 1A. ,

SMstT 1A C' i

g, r

DATg Cc:. 29.

l -

s I

j 1. SU1 JECT i

2 g Cable Derating Practice .

I 5 2. PURPCst i

58

-J To establish a design guide to determine cable aspecity

{

hI ratings for cable directly buried, in underground ducts, j

)j fJ embedded and exposed conduits, and is open top cable trays.

(Ter industrial projects and utility buildings not scluded in l

Iy the power block.

the Natiemal Electrical Code should be used g for ampacity values and calculatiemal methods).

!  % i l 3. DESIGN CUIDE  !

i .=  !

l J.1 General l j

j s l J( The following is to set forth a deftaite end unifers precedure to deterniae cable ampacity rattags. It oncespasses various .!

j g$ types of cable installacies, aseely:

i I i l a) Daderground

1) Directly buried I 2) In duc'te g b) In Conduit

• f5

1) Embedded la slabs or walls Jk

= 2) Exposed Conduit 8

= c) In Cable Trays i g.s 1) With Maintained Cable Spacing

2) Randes Fill of Cables in Tray Il Pablicatica No. F-46-426 ef the ICIA contains tabulated

[ g .asposities for a variety of cable weltage classes, thermal

$vettags,andinstallations,and'previdesthebasicampacities jl] 9er asses a), b), and e)1) above. Two volumes comprise this publicaties. Volume 1 desis with sepper conductors and g

sentains sa introducties and Appendices applicable to both volmse. Volume 2 centsias sepacity tables for alusiaan conductors.

I

} -Sections II.D.2 and 3. of the Intredestion gesties in Telane 1 of the above publicaties set forth a. net.ged for asiculattat aspecities for case c)2)-Cable tray installations with random ii _

NUM8ER E2j. L. A eL 1 g if ) SHEET 2 os 27 g- l _,

DaTE_

Oct. N Q$$ {

- * . i

. e

.fi'11., Jhis sostion has baca supers ded by a scwor ICEA '

g pahiitution, publicatics No. F-5A-u t "Ampasities ist Cables g

as Opes Top 'Trsys.* The 1secar:sbev3d be used exclusively einse thsc elde.r setshed,, ar xst farto .in Sections D.2 and 3. ,

55 is is stror.,

tI i 3 ' :1 shivlidad wrius nehx3r parte machs are installed so cast J.l .ineindual tsuaw.eurs me wt .nsamtr:Leally disposed,

.I streitatig <sur.uAus la che dwarlac14a shields of the cables say resca a worav.de wtwis r.nr ::$vera1 effect of the !"1 J

f acur vege.uss Art.z:rhy si %*in eene conductor. Wherever Iy such case.s wesar, a takird *0lEA geeYLicaties, publication No.

I. PP-33-426, "Ampecities tuciuding 12fect of shield Losses for I, -

Single Conductor Solid Dielectric Power Cable 15kV through

35kV." should be consulted. The title seeld be aisleading in I that this effect applies at other voltages as well. Whenever i shielded power esbles are installed with individual conductors is a sea-symmetrical arrangement, these effects should be investigated and either takes into acceast if the losses are g[ significant, or scene should be taken to etiaisate the I (See Design Guide E-2.6.5 " Power

]y circulating shield currest.

Cable Shielding and Shield Grounding")

This design guide contains sassle anoacity caiculations.

E Although these are primartly based os copper conductors, the same precedures and considerations are applicable to aluminum.

II 3.1.1 It is important t o remember that correst-carryisc capacity, voltage regulaties, and short-circuit capacity of cables must N be considered indapendently is order to assure proper selecties of cabic sizes with various types of iussistions,

][

w voltage classes, and modes of installaties.

6

. 3.1.2 Altheogh it is set specifically called for is the Introduction

• j Secti.cs of Volume 1 of F-46-424 (ser elsewhere -is this 33 standard), combinina, circuit " sets" of power cables is the g same conduit er underground dust requires that the tabulated ampacity be reduced (derated) accordingly. The dorating effects of mutual hosting is addressed La ethat sections of I} the Introducties == dorattag for adjacent coeduits in air or is senereto-escased e'ust banks, etc.

g Staae the anthods set forth in Section II.D.2 and II.D.3 of I the, F-44-426 Introduction (including Table VIII) have been g superseded (and should be crossed est la all septes of the standard presently in use), the table is reproduced in this Deatgs Guide in Section 3.5.1.

3.1.3 Ampecity asiculatiese for underground cables, whether run in 3'* dust haaks ar directir !;uried, can be rapidly and conveniently

. ande by means of a pt Wately developed computer program nov

, ] available at some Rechtal effices.

( i Fit ~

NUMtE R E2. 6. 4 P. c

~

) ,sMEET 3 0F DATg Oct. 29 1 F.0 u

=-

w n - - - - - . - _ _ _ - _ - - _ _ _ . _ - _ _ - . - _ _ _ _ _ _ - -

3 e ,

s . scrd programmablo calculater, cad a printer euxiliary matt, l

I It calculates anpasitios fer sob',os to a large number of ducts i E er cable groups, with complete 41ezibility regarding duct i sine, spacing, and bank seafiguraties. Where sees sables are '

i,

]{

g s3 leaded to less than their permissible aspacity, this reduction is autual beating is taken into account to persit highar leading et ecer cables in the particular raa. Casselt with JJj your local stfice . Chief Ilestrical Eastaaer's staff regarding

{ possible utsi of this program.

J

! II Wen this preigram is available, it is reseemonded for use

)

.astead of th.e motheos set forth in Sectiess 3.2 sad 3.2.

l 3.2 Direct Serial Cable 3.2.1 i Typee of directly buried cable seafigurattens with typical l diosastees

1. 2 and 3. asFinal per ICIA Pub. No. F-46-426 are shows la Figures

, detail with respect to treaching and hat'. fill are to be supplied by the project.

I i i s

i h g w -% - wwm -.

f 34" as.

- N Y d -

d k /s i f'_p gdd 31 y{ 7 ,

2..

y ,*I,

~~ -

3,y yy

. . 4 , -

' (TYP)+ '

8j eJ I Firure 1 - Seriod 3/C Cables Firure 2 - Buried g 1/C Cables 1

• h a* Note that the arrangessat la Figure 2 may cause significant j] shield lesses if shielded cables are used and shields are  !

grounded at more than ese point.

II veem a wmwe dg ,,.

b 1 I

l w 24" a de . .

Firure 3 - Suried Triplex Cable 3 -

NUM9tR E2. L.L Rev

y SHEET 4 0F

'

I Io OA?g Oct. 29.19j:

F.D 22 C '

. y. . _ _ _ . _.- . -

m.- .

i .

l 3.2.2 !aformaties required to enter ICEA Pub. No. F-44-426 for

.J Direct Burial Cable E

a) Cable Desc-tytten (e.g. 1/C, 3/C, er Triplex)

! 5 l FI . b) Cable Operating Voltage (e.g. IkV, SkV, 15kV, er 25kV) 43 - .

l jJ c) Cable Insulaties (e.g. Rubber er Thermoplastic, faraished yJ Cloth, Paper, LP Cas er Oil-Filled) j lJ d) Conductor Temperature (e.g. 60, 65, 70, ~5, 30, 45 or 90C:

I e) burial j

• f The cableampacity are for savalues tabulated ambient ira ICIA for of earth temperature direct,C.

20 Adjustaests sudt be made for ambiest' earth temperatures j

whicharesuestantiallydifferestfreethis. A frequently ,

l 3 A

used guideline is to assume thg 20 C ambient for

! "Nerthern Us" locations and 30 C for "Seethern Us" l g I 1esations. Whils this approach may suffice for feeders in j j wetsk aspecity sargis fasters offset the importasse of

IJ this ites, impercast er critically leaded underground

! l sable systema should stilise testing er other methods (per

! EET " Underground Systems Referesse Seek" - Chapter 10).

i As important precauties is this regard is to ensure that .

i cable treaches er dust banks are set affected by close l l presisity t.e other underground systems creatisd a

((I higher-thas-sermal earth ambient. Aa asample of this i

l .

f

[{

g might be the installatice of a sable run from the power bleek to the intake staties or seeling cover in the same l

essavaties with the circulattag water discharge line.

d If ambient earth temperatures above 20'C are g ensenatered, ese mothed e2 deratias the cables is u lower

.A the saaductor operating temperature by the same soeuct as

• j the increase is ambient temerature (e.g. to find the p g aspecity of cable with seedugter temperature of 90 C and

• g w ambient tesgperature of 30 C, find the aspesgty of a cable with 30 C condester temperature and a 20 C ambient temperature which can be read directly free the

{} tables).

.I f) toad Faster:

Amposities are tabulated for 30, 50, 75 and 100% lead fasters. These are' indicated as 30LF, SOLF, 75LF, ans.

f5 4

100LF. It ta reconmended that 100LF be used for all selssistione involved with generating statise jA applicatiems.

d) Earth Thermal Resistivity:

( I' ICIA Pub. No. F-44-426 aspecities are tabulated for in-earth thermal resistivity EBO, is degree

'bh NUMBE R U . e . la . R_e g g SMEET 5 05:

[ gayg Oct. 29. :

( .

• r "

s .

b .

J* contigrede-cestimeters per tract, for R50-60, R50-90, sad 150-120.

j Procedures are gives for interpolation and estrapelation, if other than indicated values of 150 are '

encountered.

thermal resistivity ICEAisreceemends not known. that 350-90 he used when earth

{I I Iowever, in the instances of major cable installations, where engineering judgement and j economass dictate, we saould determine 150 as closely se possible.

Some of the factors which must be considered are:

j type of seil, type of backfill, moisture content of soil ,

gy desta below surface, and presence of searby concrete slabs or structures.

In addition. the " baking" of tt.e soil by the current-carrying conductors saa cause 350 to abange (for the

, worse) with time. 'Two referssee articles en this esbject are

.- " Rapid Messorement of Thermal Reetetivity of Seil" by Y

7. Maeos and M. Isrts, AIII Tranaaations, Tel. 71, 1952, page .

370s "Seil Thessal Basistivity Maasured Simply and Assurately" ,

by John Stolpe IIII Transaettene Vol. PAS-49, pueber 2 February 1970, pese 297.

I 3.2.3 3 ample calculation:

5 l Cives Directly Suried Cables 3-1/C 4.16kV, Rubber Insulated Ceedseter 1 Earth Temperature 30, C. Temperature 90'C, Ambient I Finds Ampacities for 2/0, 4/0, and 300 kemil at 100LF.

I s.l tions re A pub. u.. r-44-424 asipsetties buried cable are tabulated for 20,ef dire tly C Ambiest g(

w Earth Temperature. To seistais same temperature <

differesse between esaductog and earth, use a l conductor temperature of 40 C.

5 w

• & ICIA Pub. No. P-44-426 indes page 111 refers to table en page 202.

.s - .

Wire Stae omsacity 2/0 303 gg 4/0 393 130-90, 100LF, 1 Circuit, Sk?

300 kemil ett h

11 -

S *

~

n E '~

J ~

hUMBER 22.6.4 Rev.

SNtf7 OAT $

6 08 Oct. 1 11?

3.3 cantos te 'Jndortround Due:s 3.3.1 Type of, duct ec3figurati:ns aid typicci di onsions as por ICIA

$. Pub. No. F-46-426 for 5" duct are shown in Figures 4. 5, 6 and

7. Duct bank overall dimensions are approximate, to give 8

j minimum 3" encasement coverage:

y. - .-

NW\ A4WMW@\

J A+ Awm s x up ur enn ,

g 3" either encased in concrete or set

'f escased, typical V N

Jg I

? ,.

4 N@@) '?

4-5" E ' i I

- )

< !I g] Figure 4' Figure 5

• g 11.5" by 11.5" overall 19" by 19" overall i Js* Duet Bank Duck Bank l I

21 5j

^*'* N ^'4" * ** "" 4 9 "^" *

!J y.s 30" g.

hk 3r ,

'I y I 4

ii + q

v ,, lehh@ $

l s'

(TYP)

$ NHh [

l

) 3

\

i QL/ 7%~

4 j k hGD4)- __ _

! i-(TYP)q1 _ _v-GG J

c k 8 Figure 6 19" by 26.5" overa11 Figure 7 33.&" by 33.4" overall (Not a feasible design)

Duct Bank i

NUM8sA E2.6.4, ne I -

s_wsrt 7 og g

.g oArt oct. 29g S?

~ '

___!_-_ _' ~ ~_

~

" ~ ~ " ~ " " " - ' ' =

. .vo c e s ,

s .

a)' ,

• • Cable in individual duct ces be 1/C. S/C er Triplex. To 3

g find aspacity, use the appropriate ICIA pub. No. F-46-426 )

' table. If 3-1/C cables are used per duct, the table for i Triplea Cable is recoamended for use. .

)

i b) I 2 For any normal duct bank configuration, phase and

[{ =casuctor issalances will result if suiciale paralleled j castes for eacs phase are installed esca is a separate duct. To preclude this, paralleled ruas of sable should J he designed vsth all three pr.us installed is each of  !

Iy sultiple ducts (3/C. triplea, or 3-1/C), with esses and ,

lengths of all cable matched. *er large leads, such as  !

k t the secadary connescions for staties service

• transformers, this may require several more (smaller) conductors per phase but semperes favorably from a cost i viewpotat and avoids a possibly sortees problem. If for some reases paralleled single 1/C sables per dust must be used, the individual ducts theuld be transposed at g{ intervels along the dust run to balance the impedances of I the three phases - a slow and espessive dust installation

]g method.

Another way is to synestrically arrange the ducts as shown in the Duderground Systems Referesse Book, Figure 10-39, arrangements 4.5,6,7. 8 and 16.

1 c)

If cable eises larger ches tabulated in ICEA are required i{l or more than mise ocespied ducts per bank are required, entrapolaties of ICEA Pub. No. p-46-426 tables may be g{ c Maidered. It is resemaneaded that the estrapolation, fl g[

eacher aapacity versus table size er anpasity versue number of ducts is bank, be done sa leg-leg paper since an w appremiaste. straight line will be obtained. As with any g entrapeltion, this method is limited - the further the

. entrapelation, the lower the accarsey. For duct bank

• j arrangements other than those shows la p-46-426,

,g entrapelation should be limited to smaller duct banks with y met more than two, er at most three, layers of power y senduits. Beyond this, the Neher-MaGrath analysis should I. be applied, asemally, if necessary, er preferably by means j} of the'EE-700 05-80) seapoter program. If duct banks are run in parallel, the normal ampacity tables must be gI

. further dorated. The doratias will never be more severe thans I Depth of Additiemal 8 Also Applies Derating for 5 Distasse Between Serial of to Amy Ratie of all Cables In Nearest Ducts Lowest Ducts Distance / Depth Both Duet Bank jJ 1 ft 3 fa 1/3 0.79 2 ft 1- 2 ft 3 ft 2/3 0.37

• 3 ft 1 0.91 4 ft

' 3 ft 1-1/3 - 0.94 1' 5 ft 3 ft 1-2/3' O.95 dk g) NUMBER E2. 6. 43v .

I SMggT 8 M 27 g DATg Oct. 29. 198/

-= ~ .

~ _

~

Wan a horizantal separatica of 6 f t or greater is j- asistainsd ths cutual hasting offect of adjacent duct g banks can be safely ignored.

I d) la particular situations where the available tables and

• I extrapelations are inadequate, the general equations for 2

asfacity, as stated La " CIA pue. No. F 83-426. sect:.on jJ l F. A. are ree'essended as the best available approach.

J j e) *nformation required to enter ICIA Pub. No. p-46-426 i

!aeles for Caole in Unaerground Duct is the same as : hat required for Direct 3erial cables see section 3.2.2.

• f) Wes 1/C cable installations are designed, care sust be esercised to avoid placement of steel or other magnetic material between er around sesductors.

3) Tabulated aspacities apply esly to a single cable or

• slagle set of cables is each duet bank conduit. Where J[

I aeditional circuits are installed in tae same coseute, che J aspecity facters tabulated at the end of Section 3.5.1 suet be an,11ed.

Ig g 3.3.2 Sample Calculations iI g

Cives:

Underground Dust Installaties: 13.8kV Eubbgr Isastated Cables: Ceeductor emperature 90 C AmbiestEarthTemperature20{C: 16004 (full lead current,re,sirements).

F[

g w

Find Sise, number and seafigurattee of cables required.

j Soluttes: a) First cessider 3-1/c or 1 Triplan per duct An w

• j order to obeats balanced currests in each y individual phase.

,s

~g b) CIA Pub. No. p-66-426 indes pare tv refers to

• the Table es page 242 for Triplus for the given senditiosa stated above.

il jg c) We see the 3 Triples will set carry the current for the nazimum size tabulated. Bewever. 6_

Triples will give the required aspacity (i.e.

1. 500 kasil 150-90, 100LF aspacity is 288: 6x 5

t 288 = 1728 which is greater th's the 1600A re,.ir.d).

qI ' d) If dust sins will peraf r triplazed cable of j' larger sised, interpelaties of the tabulated data indicates that 6-150 kemil vill be

. marginally satisfactory and 4-1000 kemil will provide a conservative applicaties.

H

m. . n . %

7 f SHEET 9 O!_

I DATE Cet. 29. 19 A $ _. f -

.._ - c _- - - . ,

-> -] .

ED-22 (.

=

' ' e) Date that the tabulated data for triplexed cablos applies to corresponding sizes of 3-1/C ccblos installed ta a duct.

I 3.4 Cables in Conduit Embedded 1

3.4.1 Types of Embedded Conduit Installation:

12 j Imbedded conduit refers to costuit in coscrete alabs and walls. Bernal configurations of conduit.is underground duct g

installatiosa are illustrated in section 3.3.1. All provtstons of Section 3.J.1 are equally applicable to embedded conduit.

f 3.4.2 Inforastion required to enter IpCEA Tables for Cable is Embedded Ceeduits a) The aspecity of cable la embedded conduit should be takes from the ICIA p-46-426 tables for Jimilar cables is underground duct. The same data set forth ta Section g{ 3.2.2 for entertag the tables for direct burial cable is required for cable la embedded conduit.

]N b) It is recommended that 250-60 he used for cables in embedded conduit (130-60 is typical for "hardrock" 1 structural grade concrete).

I c)' As ambient temperature of greater than 20'C is frequently the case for embedded seeduit is a power er l g industrial plaat (i.e. esadmit is a senerete slab with a hegted room above and below any have ambient temperature 40 C er greater). Thea., it will be necessary la most d[ esses to derate the anpasities gives la ICIA pub. No.

p-46-426 for cable is underground dust, since tgese g aspeetties are for as amblest temperatare of 20 C. The precedure outlined ta 3.2.2e any be

• J

{,3 ambient temperature greater than 20'used te derate for C. or the any be found for sa amblest temperature of 20,ampacities C and thea dorated by the equaties shown below.

j] I' = I ' T* - T*' where T, - T,

=

I' derated ampacity (amperes)

I

• aspecity tabulated for T, and T, (amperes)

. J ,e .

rated . ti.e e . d t.r t _p.r.t.re ( C, a

T, tabulated aseient temperature (20'C)

T,' a actual ambient temperature ('0) 4 n NUMsgm E2. 6. /. . h= v

] SMttT 10_ 0F DATg Oct. 29. 19 g.

s o.__.n,

3.5 Cable in canduit Exposed 1 ,,

2-g 3.5.1 Information required to enter ICEA Pub. No. F 46-426 Tables i

$ forb, a, Cable in Conduit e, and is tha buried d for directly same as set forth La section 3.2.2 cables.

i N l I3 ,

j jjJ ! i Consultahe.'adox of : CIA Pub. No. ? .4-426 for tabulated ampacities for :riplemed or three conductor cables ta isolated conduit.

j Note that the tabulattees are based es as ambient Iyl j air :esprature of AO'C. *f ambient air temperatures higner than a0 C are encountered, them one of the same derating procedures outlined as 3.4. c saould be followed.

t le {

,F l Note that tabulated aspecities are for a sinsle three

t i conductor or triplexed sable is sa isolated conduit. If more

conductors are is cae same sesduiz and sescurrently loaded, the following ampacity factors (100% ampesity MINUS the

{ percentage derattag) must ha applieds jw i 3 Total Numeer

{ I J Ampacity ef Conductors Factor

! 1 i 3 t

4-6 1.00 7-9 0.80 j {jI g

10-24*

25-42*

o.70 0.70 O.60 i

g{ 43 & sp*

0.50 '

• Iselndes the effects of lead diversity.

i 2

A

=

i j Where a fourth coeduecer is included as the neutral in 3 phase 4 Vire systems, the neutral is set seented as a current j g carrying esaductor and se derattag is required.

i T. .gI here nominal lead diversity saanes resseeably be assumed, an g

} appropriate Ampesity Festar can be sales 1sted using the i 1msthods set forth ta Appendia 1 of the Asher-McGrath paper,

} I}

{g "The Celestation of the Temperature Rise and Lead Capability

.of Cable Systems." The matter should be reviewed with the

{

=

office Chief Electrical gagineer.

_he hos desating appreeshes 301, em alternative cable l routing er raseway arrangement shestd be seasidered.

3.5.2 Derattas Fasters for Cables in Esposed Creeps of Ceeduits in Airs Ia a) If the vertical and horisestal spesing between surfaces of p esoduits grouped em racks er other supports equals er g exceeds the outside diameter of the conduits, the aspecities for cables is iselsted conduits in air should be used without dorating.

n s ,

NuM8ER E2. 6. 4.h ,

$NgfT 11 OF I DATg Oct. 29, 19 !

_,, m.u a

s . .

b) Table I shows ampacity factors by which ampacities

j. tabulated for cable's in isolated conduit is air should be 1

eultiplied where conduits are Stouped more closely than outlined in a) above. The table is based es separaties *

]j y ,

between adjacent conduit enterior surfaces set less than one fourth of the outside diameter of the lar4er of the s3 two adjacent :oaduits. (d/4). TEIS SEOM.D SE CASEFULI,Y JJj NOTED IN PROJECT SACEWAY INSTALLATION " NOTES AND DETAILS." If separations are less that these minima, a g comptes heat transfer calculation is required to  !

accurately deternise espacity.

J If I

TA8LE 1 I

']

'g CABLES IN CONDUIT AMPACITY FACT 048 Number +

Vertically Number Serisestally I g i 2 3 4 5 6 1 i

1 1.00 0.94 0.91 0.88 0.87 0.86 l 2 0.92 .

. 0.87 0.84 0.81 0.80 0.79 l 3 0.85 0.81 0.75 0.76 0.75 0.74 4 0.82 0.78 0.74 0.73 0.72 0.72 d 3 0.80 0.76 0.72 0.71 0.70 0.70 6 0.79 0.75 0.71 0.70 0.69 v 3.5.3 sample Calculation 0.64 LI cives: Ceeduit installation la sir of 3 vertical and 4 j] heriaastal conduits, each seeduit separated by 1/2 gg eenduit diameter 3/C,6f,rubberinsulated

.a senduceer tempgrature 45 Cs ambient air temperature 40 C.

• Finds Ampaatties for M AUG and M AUG sables.

I 8 elations ICEA Pub. No. F-46-426, indes page Y refers to table jj en page 312 for isolated coeduit.

11 it -

~ "

NUMSER E2.6.L. P. e .

a 2 SNEg? 12 0F DATE Oct. 29 15 g -

, m. . ,

Aspecity in Aspacity Aapacity in J' sise of tach Cable Isolated Conduit Factor Grouped conduit

1. 4 87 x 0.76 I = 66

{I 2 d6 66 x 0.76 = 50.2 3.6 Cables in oeen ee Cable Trav j J.6.1 Catte say be installed in : ray with "saistained seasing" er gy randealy pulled or laid in ene : ray. In :he saistataed g spesing method, cable spacers of plastic, impregnated weed, or f4 y

perselata are inserted to asistain a selected vertical and l horisestal spacing dimensies between adjassat cables in the tray. Rows of each spacers are installed in the tray at intervals, depensing es the stiffness of the cables involved, sufficient to ensure that the design spacing is effectively

" maintained". The labor required to de this type of j, installaties is assy tians that required to tastall the same g I cables ressoaly La :he same : ray. It can only be economically

}I jastified for large, important feeders involving semperatively heavy electrical leads. The offsetting benefit is sabotaatially higher anpasity. It is suggested that cable 1 dust be seasidered whenever senditions are such that maissained spesing appears to be a desirable opties.

flg 3.6.2 If sable duet is selosted, the espasity used should comply

[1 with the resesmeadations of the sable dast asanfacturer. If q field-fabricated asistained spesing is to be used and the gl g

spesing is maintained to esseed the fall sable disaster, the ampesity will be this same as for the same sable iselsted in g air. For usistainind spesing free 1 diameter (cable e.d.) to 5 1/' dia=atar. areir tha ==***itF f****ra **b=1stad in Tabla g

VII en page 7 of Telume 1- (Copper), of the ICIA yI ampesity tables (7-44-426) to the ampacities tabelated in the n..g book (s) for (selsted cable la air.

y g 3.6.3 For poner circuits in tray other than the major, heavily I leaded rams which jastify the expense of asistained spacing, the method used is "readas spassag" or "rmadas tray fill".

Ig 1

Seattens 11.D.2 and 3 em page V of ICIA P-46-426 describe a method for determining anpasities for this aesdition, ustag i

Table VI!! free these same sostions, but the results are assaltable for ear applications and 3500LD WOT 53 USED. The g . eerrest referesse is ICIA Pub. 30. P-54-444 (NEMA WC-51) entitled "Ampsetties - Cables in Opes-Top Cable Trays".

Ia 3.6.4 Ampeetties for power cables installed is trays withest

] asistained spessas should be based as the methods and data ,

contained La ICIA Pub. No. 7-34-440. The ampacity : ables in this publication are generally based en the calculated depth

, 4

.it

.uussa E2.6. ' . Rev

/ sMEET 13 0F y DATE Oct. 29. 19.

__  :- . _ _ _ .____ :n "

1 el cables in trays carefully packsd to cyprezimate maximum J eable deIsity-of-installaties, seasidering this es the " worst case basts" for cesservative design.

{ i based en loof, lead facter and se diversity.The tables As the title are further

I indicates, the tables are based en "Opea-Tep Trays". ..The i

effects of tray cosers, fire protecting asterial wraps, or

3 testing of tray through firesteps reentre dorating to the l lj empeeiries as eenermaaed from ICgA p34 =4o . "h J e additional doratsas required for each is covered in the fellowing' sectiease i

]4

[

Iyg - 3.7 Additosal Derat ng for Trey Cables Transiting Firestepe

{ - 3.8 .

Additonal Derating for Cable Trays er Conduits

- 3.9 Iselesed la ytre protecting Material i

-1 Additiemal with Derating for Cables Routed is Open Top Tray Solid Covers.

3 ..

} 3.4.5 j' m seeee of ICIA Ameacit? Tables for Cable Trav i I{

}y a)., Data is tabulated for single sondetter, triplemed, and three sesductor cables. For um1tiseuductor power cable l

l ether than three conductor, a sesverstem formula is

! *E provided in the Introducties section of the Tables.

b)

' Data is tabulated based es the everall cable sises h{

g (outside diameters) corresponding to the aere commes sable i

gj esastreettees.

{

generally varies directly as the cable eatside disset

(other fasters betas equal), a simple proporties j g[ sultiplier seables determination of aspecities for outside' i .s diameters other than these tabulated for given coeductor sises.

j g A special case escurs where cable e.d. equals er a

w esseeds the design basis depth of fill. In these esses,

• j cables saa be laid parallel in the tray, ese layer deep.

, .s Ampasity will be as tabulated for depth serresponding to "table stas".e.d. regardless of pereest fill er easet table i *.g c) 1] Aspecities are tabulated for four different voltage ela..e.,

1 i

Ia8 15,oo0v.o-so0v. , sot-2,ooov. , 2,ool-s.ooor. , and Ampacities for seminal Sky class may be

! determined by applying the siaiag (cable everall e.d.)

i j a serreetten described is b) (above) to the aspeetty  ;

tabelsted for corresponding esaductor sises for 3.ooov. i j class table. (While this is met precise er theoretically j trae, the resulting errer will be negligible.) i

)J d)  !

q 3 Tahviatione are based es cables rated for 94'C maataua contissees tenductor teaterature operatier, and to C maatsum ambient air temperature.

Correcties coefficients

~

for each of these facters are tab'ulated la the Introduction section of the tables.

I

$t, uvutta u. u . nei k sweer u. er  :

1: .2 DATIr 0:t. 29 19!

g . .

. =. ~. a -

. e ,

o)

y. Tho 42pscicios are tabulated to tha basis of " depth" of cables in the tray with "4epth" defined in the **

Introduction section of the Tables as fellows:

• j I3 3epsh a.d. '* 2

• a.d.2

+ .... a ** d2

?* vhere

~41ech of Tray l

,,J i

d,,

.. 4 a

o.orail e.d. of 41ffere.t

..h1. 1s... ..d a,n, o n a g Number of cables of each

'I

• corresponding diameter All maits are is lasses.  ;

Our usual mothed of calculating tray fill is to select a o percentage of the usable areas secties area of the tray I{ig i unser consideraties, then to determine how sany caoles of j various percentage crossofsecties areas saa be accommodated in that the tray. Fok this, we use the actual cross sectiesareaofthecabjes,andsincetheseare(usually) circular, we sultiply d a ff /4 for each cable e.d.

'1I. Using the notations of the 1stroducties secties of the

'{l Tables, our " depth",,weeld bes j Iy a,d=4.a,d,24....sd,=c 1

Depth a e Width of Tray

. 2I (a gg d +ad 22 + ******"a d a )

TeI -

Width of Tray a.J h Note that our method differs from that es which the tables a .f are based by our imelastes of the faster 17/4 J] 8ecause of this differesse, eer calculated depth must be gI divided by Tf /4 (or 0.7854) to be consistost with the definities of " depth" on which the Tables are based.

1 (Sees find it easier to saltiply ser calculated depth by the reetpresal of Tf /4, er 1.273).

s y ,.... . .f t,. ,ab1.s J 1 freguest praattee is to select trays for raades fill with 1 fyouer cable which have a usable tray depth of 3 lashes and te

' design for a 30 poement fill. (While :he same tray usaole depth is very widely ased in the power industry, both of these I

I' '

parameters are selected arbitrarily. The 301 fill figure

.it -- -

NUMSE8l E2. 6. 4 Re3 SWEET ~If CD b

DATg Oct. 29 11 F.o 2: !

Wate that where oursthis is 0.33". is for a cable cutsido diameter of 0.45" Cuide that, Recall from 3.6.5 b) of this Design ignoring other parameters, ampacity is J directly proportional to cable overall o.d.:

] I sableaIcablex I table

" table

= 0.65 0.53 x 78A (as calculated thus far) 1

= 1.18 x 78A

,s

[J

= n 4. peres b) Cives: 2erisontal tray, 3" seabte depth, 2/C $4 AUC cogpersesductor600V.sableafunjasgated" singles,"

90 C sonductor temperature rattag. 60 C ambient air temperatura, sable fillers added to aske cable romsd in section with random withlay. overall o.d. of 0.75". Cebles are tactalled 4y ga Find Ampacity for 407, fill 2

l Solutises- Depth of cable for 401 fill 1 =(3"a0.i0)+ ff /4 .

= 1.2"+ 0.7854

{ = 1.53" '

.I (For practical purposes, the last 0.03" depth can be

• ignored and tabulated data for 1.5" depth can be used 5

without interpolation).

a.

.it7 ior 1.5" de,th #4 Eater Table 3 and acts the Av. JeC eI 1

a = 49 Asperes' t) 3 Note that se sorrections need be made for either 1 ambiest air temperatures or seoductor temperature rating.

5 Bewever, we must correst for a differsat I{

3 amaber of eenductors and for a different cable o.d.

These eerrections can be made is one step utilisias.

g the equation shown la the upper left corner of page a

11 of the tables:

g I' m = g a r, ,1 E

I*

]

• 0.75" 0.43"* z 49A f3 *From Table 3

}2 ,

= 0.905 AfA a 1.225 1

u = 54.3A.

NUMsg n E 2. 6. I. . Rev i

g v swerT 17 05 I . .. ..

supposedly represents filling th2 tray, esics resdos cable j- pullies er layics of cables into the tray, to thq. point where I covers any be installed wherever desired vitbeut particular difficulty. Many othere take 401 as apprezinating " complete" I tray fill. Se far, neither figure has been claimed to gI3 represent a " cost-effestive" optimum.)

3AMpt.g CA1,c11.ATIONS g a) Civeas Borisestal tray. 3" usable depth. 1/C #2 AWC s

Iy jacketed. 600V.

copper conductog cable. 0.33" I. o.d.,aarambgent temperature-30 c. insulation i rated for 125 C sesductor temperature, randos (f l fill.

?ind Ampacity at 301 tray fill.

soluttes: " Depth" of sable 4 301 fill I[g =

=

=

(3" x 301)+ ff/4 (3" a .0.3) + 0.7854 1.15 1I Enter Tables.ICIA pub. Ee. 7-54-440 Table 4.

line interpolaties to interpolate between ampacities Use straight

{I 3

tabulated for #2 AMC e 1.0" depth (75A.) and 1.5" depth (5SA.)

I I I.15 =

II.5)

II I.0 - 3.15 - 1.0 * (21.0

.. 3 - 1.0

}[

w

=

=

754 - 0.3 m (75A. - 584.)

75A. - 5.1A.

= 70 Amperes wa gJ Notethatghtsespasityisfor40'Cambientair,where

. .g ears is 50 C. Refer to pse i right hand colgan y g r

"corressies for Ambient Temperature" - for 50 C ambient, I, *, multiply the above result by 0.90s j} = 0.90 m 704. = 63A.

! Note that this ampasity is fe 90'C rated coeductor l' temperature where ours is 125 C. Refer to same page.

sang solana "Correcties for Camductor Temperature" - for g

115 C sesductor temperature multiply ear above result by 1.24:

jJ = 1.24 a 63A. = 78A.

k 3, .

-E .

NUMsEM E 2 , 6 , f, , p p. ,

5 2 SMEET 16 os *

.7 g DaTE Oct. 29, 195t.

, ~_ y -- _

4n U i3 7s

, s .

c) Civos Isrisental tray 4" uschle depth, 1/C #1/0 AVC j.s aluminus condugtor 8kV nominal rating shielded cable.

1 0.97" o.d. , 90 C conductor rating, 40*C ambient, open top randomly filled tray.

5 j Find: Aspacity at 307 fill S

j Sollicios Depth of caele for 207 fill g = (4" x 307.) + # /4 N = 1.53" I Again for practical purposes we can omit interpolation for

' the 0.03" incremescal '

depth.

= 1.5" I{l Inter Table 29 for 1/0 AUC conductor and 1.5" cable depth:

]E I1,3 - 94A. for sky cable v/0.72" .d.

1 (Assume that the voltage class difference between SkV and SkV has negligible ampacity effect and make correction I only for differesse in e.d's.):

g I uble

=

emble zItable p

og ta,1e

= 0.97 a 94A E U 5

g = 126.6A.

1E 33 = 127 Amperes d) Gives Eerisestal tray, 3" saable depth, 1/C 1,000 kes

'I copper saaductor 13gY ahielded cables 2.15" e.d. 90 C esaductor rating 40 C ambisat opes top randaaly filled Ig trap.

Flade Ampacity at 307 fill 8 Selation Depth of cable for 301 fill

= (3" x 0.30) + TI /4

= 1.1,-

J NUMBER E 22 6. l. , R,

~

) sNGET 18 O' g

oart oct. 29 g

tu

- - . . .=

j . .- ,

Note that the cable diameter exceeds the preteribed depth of fill. The cables can, however, be laid in the tray in i 3 a single layer. The aspacity may be calculated as i ] described in the second paragraph under section 3 "Use of

) j Tables" on page "i" of F-54-440. It may aers readily be j sj looked in directly in this standard by entering Table 33 j j3 l !sr 1.300 kcall :sadue:sr siger I * $33A E

Where the cable o.d. exceeds :he prescribed desch. :he i gl aspecity is 407, of be aspecity of the same casle in free l

D l sir, as tabulated in p-46-426. It is therefore independent of the sable e.d., se_no correction need be j -

" made for the differesse between the actual o.d. (2.15")

vs. ene taoulated o.d. (1.90").

l 3.7 Additional Deratism for Trar Cables Transitting Firestops

! j ,=

I Many of :he firestops commonly used for sealing vall and floor 3

I openings for cray cable passage aske use of a flame-retardant thermal insulating material such as silicone foam. Any solid l } or ventilated tray covera should be removed prier to forming i 31 the firestop.

! 31 i

i g Several assufacturers or installers of this type of material i

claim that the use of their predest er mothed does set require  !

, dorating of enclosed power cables. They base these claima es l

! a data from tests, imeleding ese er two in which the cables were

{ [

j leadedtothefullP-54-440ampsegty,witheetthefirestop E]k

.l hot-spot temperature esseedias 90 C.  !

)

El AsserdingtoStegpe,whoseanalysesandtestingarethebasis  !

5] f., p-54-44o, ,o C het s,.cs .111 eely .. cur vb.re a . ber og of cables are packed together - a typical "werst case" to use 1,

• as a design basis. Caumenting es his own test results, Stolpe stated, " Mete that eves though the majority of sables -- ran hh coeter them satsulated, there was a group of cables -- that

[8 did reach the salestated temperature. This points set the g} fast that all sables is a rendenly filled tray sanset be expected to have the ases thermally adverse environment, but II same of thee will.*. Stolpe aise demonstrated that ampacity j)-

3

~ should met be increased beessee of diversity (i.e., some or aany of the sables in the tray are only lightly leade'd er are

.3 sempletely ualoaded).

8 5esasse of these facts, good engineering practice requires I gl'

• that when thermal isselating asterial is used as a firestop,

, additiemal derating amat be semaidered and applied when a necessary.

h H -

NUM43R E2. 6. 4. ,Ry 3 J SMffT 19 05 I

Y oaTu oe c_. _2_9

.. 2 .5

M }es Angplas power. Division perferted a serios of tests in

y. 0 980

• Edeteriime the# thermal *efleet of two different types c,f i

firestepe se tray sables. Caetyperepresented5h*8I8C*

I firestop souprised of a 9" thiskassa of 17 lb/ft density -

j silicone fees. The other was the minimum thickness SpC i

d s2 firestep comprised of two laye$s of 1/2" Mariaits with a 2-1/4" thick layer of 17 *b/ft density silicose fees J between, plus Flamensstic seating on the sables en each

! esposed face of the firestep.

Alcheugh ebe hot-spot temperature of the SpC firestop was II slightly lower, the semelasions were that either type required I a seminal incremental (additiemal).derettag of 15%. ..It should

' .. aloe be ested that to the spostas test groep is this' series.

. as in Stolpe's tests, not-spot temperatures for trays witbeut firestepo er seramis fibeir blanket wrapping were found wata 3 -virtually me~ thermal marain when all sables were sentismessiv 3 Iggdgd to their F-54-440 ampseistes..Summaristas, these tests show that sables transisting either of these typical staissa I{I firesteps should be given as additional dorating of 151.

I (i.e., a tray cable with p-54-440 aspesity ef 100 esperes for

]y epen top tray should be dorated to SS amperes if it passes through a firestop).

I Another approach that can be seed is to analyse the II I heat il I-gata (in watts).for a ese foot length of tray. This is done by the following methods' A. Take de resistasse for OWI. foot of each individual

[1 .trand.d .o.da.t.t fro. a .abl. .s.is.eria. h.adbeen su.h as Okeette Cable Engineertag Data Booklet. Table 1-3 dg (tiased seeductors where appropriate).

3. Ceavert "A" to as (where appropriate), by multiplying by

• J the fasters tabulated in Okesite Data booklet. Table 1-5.

.3 C. Magtiplyegskvalueby1.25tosouvertRtabulatedfor h 25 C to 90 C maximum conductor temperature (1.258 for

&* almia m coedueter).

il D.

gI Multiply each "C" value by the SQUARE of the current

. serresponding to the actual full lead of the devi.e being ,

I served. Shere time intermittest leads (such as MOV a eperator ester leads) or leads that only escur during abeermally lightly leaded conditions, can be ignored.~ l E. Add all of these " watts per feet" (of tray).

E1 F.

) The total wattage for eacg 6" increment of tray width should be 24.5 Watts G *O C er less te ensure hot-spot conductor temperatures less than 90,C within the firestep. ' -

a I.

e l f, NUMBE R F.2. 6. 4. R e .

)

i $NEET DATE 20 0:C- 29. 15 OF

2

• ^

1 .

! . Isample:

i j- 70 - 3/C #12 avg cables are routed in a 12" wide tray with l 3 each circuit loaded to 9.5 amps / phase. Can a 9" thick j e i

silicosefosafirestopbeinstalledinthegraywithout l

]}I 1-4s creating , hot-spot internal temperatures >90 C1 j JJj 1 teach cable) 1.71 x 10' x 1.25 = 0.00215 (No det.:c correction required)

II (each cable) = 1.52 = 90.25 i ut 12R (each conduecer) = 90.25' s 0.00215 = 0.1940375 j

I

,l 121 total = 0.1940375 x 70 x 3 = 40.75

g 3

Maximum permissible watts for 12" w. tray = 2 x 24.5 =

l 3 49.0 watts I

I{ 40.75 < 49.J watts Therefore, the firestep hot-spot is less than 90'c i

8 CAUTICW: Since ~" watts per feet" or "wetts per feet per k unit width" serrelates with AVERACE temperatures, each

. I suah case should be analysed to ensure against bot-spots.

{ If many of'the sables are lightly leaded, ese er a few l ,

small cables saa be everleaded to the point of damage I{

i witheet the "wette per (square) feet" 11mitaties being l fld[

aseeeded. The analysis should verify that eks sables are eyesly distributed in the fire step. The review should be

' based en resseeable values of wette per linear feet per 4 j unit of cross-sectional area of each of the cables of w3 interest. .

• 4 I Some firesteps use silisene met as a foam but as a solid

1. J elastomer. either unfilled or filled with grasular metallic lead for resistasse to radioactivity. Because in h its sermal state it has thermal asaductivity much greater

&I j] than that of fosa, it dissipates the internally-generated heat of the sable such that se aspecity derating is

[ta required in the 4" to 12" thicknesses of sermal firestepe. Even with these materials, dorating may be I required is substantially thither sectione such as these 8 sometimes used for radiaties barrier seals. These

}4 asterials pyrelise whos exposed to fire, and the resulting "shar" is a good thermal insulater, thereby enabling the j asterial to fulfill its fanaties as a firestep.

if -

- 11

.gg!

l NUMSER E2.6.4, L.

EET 21 0 DATg Oct. 29, 1; g~- - w.n

{ *

3.8 .

~

J Fire Protectina MaterialAdditionni Deratics e n for{

l j

esaduits are seastimes eselesed ta fire

{ g,j l

l 43 Among the first such materials to be seed was ceramic fibe

J blanket asterial suca as Kaeweel or Cara-blanket . Its I

l f effective for precasting centrol. instrumenta communications type caeles.

3 Bewever, the seceed i

$I characteristic at least as a design makes baats.it generally unsuitable for power cable l

f ,

Asether fire protestise severing for trays or senduits sai plaster-like asterial asaed Theras-Lag 330-1.

{ In addities te 4

beias fiber bleakes, easierthe to derating install and required much foraere powerdatable s cables isthan mer reasonable due to a much higher thermal sesductivity.

{' Ig CAUTION:

3g Fire protecting sacertals should set be used j inseejunction with solid er weatilated type cable

}

tray severs en power tray. If used together, the cable woeld have to be darsted for both the tray severs and the fire protecting materials.

3.4.1 Ig Deraties Raoutred m en Usina The m -Las 330-1 Ia Res fire protectica esterial 1.s Tegaired on trays sentainin F}]g- power cable. Theras-Lag 330-1 to preferred ever seramic fiber ya blanket for Therne-Lag. materials as the sable derettag is sabstantially less 330-1, 1/2" Based g

U thickness willas ASTN provide a 1E-119 fire hour fire tests rating of1"Theras-1 and thickness will provide a 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> fire rating. Ampacity tests I have shows that 1/2" thickness of Theries-Las requires that 1g power sable b'e derated 12.5% for sable tray and 7.6%for 3' j esadmit. .

of 171 is required for eable tray.Teste asing 1" thieksess 5

jj since tests of seeduit with 1" thiskness of Therme-Lag were met sendested. the dorating for power sable la seeduit is II - estimated to be 10.5% based en the 341 increase ta the deratias tested es sable shove tray. between the 1/2" and 1" thicknesses w m 3.4.2 i perattaa teestred when Usina Ceramis Fiber 31sakets 1, . em e1.. wer mwisie. d t.d te.g. e,

.eerante fibee bleaketn (Gera-Blanket). These teets taditsted a requried dorattag of 73T, (2.28 watto per feet allevable g dissipaties per 6" width of tray) if ve 1" chick layers 4re n .

f h NUMSER E2.6.4 switT 22 Pev os g DATg Oct. .' 9

~ ._ _ .. 19

-- T .I".

• F.D C.213

___ _ ?-__ - - = -

~

used, er 6t.% (4.4 wr. cts par foot por 4" widen of cray) af a single 1" chick loycr is used for wrapping cable trays. For wrapping conduits, these tests indicated a maaimum permissible I. watts per (linear) feet of conduit internal heat generation of E 5.98 for 4" conduit with a single 1" layer wrap.

5 3j For the three 250 kcall cables involved in the testing. :his 43 represensa a,derating af 23.3% from :he actual seasured ampacity for the similar " unwrapped" conduit condition.

Since only a single caos was tested, it was necessary to 3 :alculate the sazimum permissible watts per foot of internal I '*. heat generation for other sizes of coseuit and other thickness l f of therusi lasulaties wrapping. A fairly detailed mathematical model was developed and the calculations

.f 3 perfereed for various trade sises of conduit, each wrapped with ese 1" thsch layer of seramis liber elasket. Conclustons l were as follows:

I Maaimum Allowable 'l I{ Conduit Sise !aternal test (13. IMT or INT) __I,2

( R ) Watts /Ft.*

• 1" J.28 E 1-1/2" J.44 2* 4.30

${l 2-1/2" 4.70

l g{

g 3" 5.18 3-1/2" 5.5A f 4" 5"-

. 5.88 6.53 6" 7.11 l d][

8 8j Where ese er two power strasics are installed la a wrapped eenduit, these limits saa be directly applied without

.J Sasteding the cable rating (S). Where three er more power  ;

} circuits are installed in the same counos conduit, two separate criteria suet be applied, as follows:

l

,3 2

• il (1) rue .= .f a.1 m 1....s .f ail .f a. Lasulat.d ga coedseters shall met easeed the tabulated Watts per foot 8 tabulated above, and

• (2) No insulated sonductor may be loaded to more than its

• ampeetty tabelated in ICIA F-46-426 for iselsted conduit ta air derated for the total amber of current-sarrying seedusters is asserdasse with Table VIII of the j !stredesties to Teluse I (Cepper) of F-46-426 and reproduced is this Design Guide ta secties 3.5.1

• *Use the method set forth La the latter'part of Section 3.7 of this Design Guide, steps "A" through "1".

.it NuansEm E2.6.4, L '

7 susti 23 '

1 DATE OCC. 29. j F.L 2: l

.I k

3.9 Additional Derstina for Cables touted in Open Top Trav with Solt-. Covers b 3 j ] Solid metal tray severs are often used en cable tray to provide mechanical protecties and prevent the accumulation of l g

, debris. Weslest power plast projects may consider the use of j hg solid ' tray covers to address the separaties criteria is Jj *as*=.5=m**M-j Jas prisy M M admotpd3 s.1980. by,the See Angeles power lE Divisi~os indicated that a dorating of 275 is regstred for j g2 aolid metal severs asusted directly en the tray sills. The l

j

• l test sensisted of a tray with a 12 feet less solid metal sover asusted directly se the tray silla with a 1/4" eyesing at each l

• ] and. For the seafiguraties tested, a maximum allowable dissipartes of 17.25 watts per linear feet for each 6"

-l increment of tray width was determised. Se hegesas to the l

i i

jg 1.Ap3 tests. IIIs paper No. 43 su 305-0 pres- ted test results ladisating that a 231-305 derating should b  : sed for solid jg metal esvers. m test configuraties for tas IEEE paper was a 3 24" wide tray with a 24 feet less solid sever meented es the I tray sills.

g

! h use of solid severs es tray containing power cable should

[ be avoided when it is practical and feasible to provide as j g esseemical layest resting the trays is areas set requiring

{ sewers. Bas 11 sing this is set always possible, projects which ,

i utilise solid messi tray severs for debris protection ebeeld i i I consider means other. thas sousting the severs directly en the -

l y tray sille. Instead of severs mounted directly es the tray i # ]g, sills, severs ~ er shields espported above the trays can be used i .J B with se addittamal dorating for power cables if a minium of l g] 4" clear space is asistained between the tray stil and the 0 sever. When adequate protection can be provided by a shield 3]5 er sever suspended above the tray, (supported a less or bessath walkways, etc.) this aise has the advantage that cable 1

23 may be added at same future ties without insurring the seat of removias severs.

! When solid tray severs greater than six (4) feet is length are i letilised os power, grays and are seented directly on the tray l Il . sills, the ICEA aspeetties should be dorated 271. Currest l [ 5eshtel practise for sable sising and selecties is such that g the 271 derating for solid tray severs may set require larger j .: seedseters. Based on as analysis of surrest 3FyD projects,

sable sising and selecties has been suffisiestly senservative

{ T as to set require larger sesductors to esapessate for the j gj additieaal derating for solid tray severs.

i li i

I .

4 l 8I .

l.

NUMBE R T 2. 6. la . kes SHEET 24 0F e

4 OATF Cet. 20, D

_ _ _ _ P._ __ _ . . _ _ _

i Prior to applying the 27*. derating for solid tri.y covers, the jg following items should be considered:

! a) Cables feeding actors are sized based on 125% of motor

]J full load current which provides a 25% margin over rated

'1 i l'I full load. Most motors are selected as the next larger

" trade size

• over :he horsenover requirements of the i i ~

jj driven ee'utpoent voich provides additional margin.

Mechanical eeuipment selection is also based os "verst

=ase consitions rather than normal operating conditions.

i IJj. 3riefly, motor feeders usus11y have a margin greater :han 27% aoove the actual motor load currest.

N b) Some motors are not continuous duty motors such as noter jll operated valves, cycling susp pumps, etc. and do not contribute to heating of cables in a covered tray (ie.,

ICEA tables are based on all cables at 1001~1oad with no tj j2 diversity).

g c) Cables feeding load centers, mocer control centers and

! I J power panels in power generating stations are usually h sized large enough to be capable of handling the full aspacity rating of the bus. It is recommended that this 1jg criteria be used to size such feeders as this permits the addition of leads throughout the life of the plant as long II as sufficient transformer capacity is available. This l{ technique typically results in cable ampacity margins greater than 27% above the actual lead currents.

Q In sumanary, the use of solid tray sewers asented directly on j[ the sills of power tray should be avoided. Men specifically w required, solid tray covers may be installed directly on the j sills of power trays (without concern) as long as proper w design techniques were employed in sising the power cables.

,J CAUTION: Wen selecting the size of cable feeders to panels that are dedicated to loads which are all simultaneously y

. energized, special care must be taken to assure that the E* conductors are of adequate size, and that all derating factors I} have been considered. Fanels which are dedicated to loads such as unit besters, ITAC equipment and freese protection are

[g a typical of those which require special attention in selecting the proper cable sise, f

g 3.10 Ceseral Precautions The consequence of overloading power cables insulated with the I high quality thesisosetting elastomers generally used in our industry is a reductice is full service life espectancy rather

}T than sudden catastrop6tc failure during startup. gven a

} .

seriously overloaded cable may function for many years before failing.

I

}T

I .I NUMSEA E2.6.4 Ra E

I

@ j SMgr-DATE 25 Oct. Q. 1 oc

( I y

performing these calculations.gewever, good E The relatiesship betweenu ness

{gj rather than linear, se that small evert s exponential -

s3 extended period of ties can seriouslyesshortes ever as c bl jJ Preasture approach to the end-of-service-lif thermal aging of casles is conside a e life.

providias abe potential for commesred by the JSWRC ase ten 3 circuits mader assident senditions. mode failure of cla N Prosa61y the highest possibility of fut

{f g regard is improper evaluation (or estination)ure problems i

,l sabia envireement temperature. from the point ofm ofview of a bi the actual ent g

g high ambient ersaa free the Plant FacilitiThe project ssinal "s es groep on the 3 There ar,e asay tight er iselstedeaareas selved. isdesig through which sables are routed where th as operattag plant I{I 5 substantially escoed the ETAC design figures e temperatures ,

]g of the whole plaat, but cables m ts for traversing design t be dorated accordingly es a spesial saae se areas b should 1, premature failure should be sattaipated asis er their layout te minimise hot-ssheeld beg and reviewed equipment and chec e.,,.s 1. t 7 are. ve pot espesures er verify that power

.. d.r.ted to I* . s.r e .

Our asposities for cablee is tray are bas d dg been preven that the sentrolling fastera ftray, but it has g sables la tray is the dggg. Oer eseal or fully leaded praatt uj] is 3" deep tray seaverte to 1.13" depth ee of per 301IC3A fill I P-54-440 (See preceding sample Celsalatte pub. No.

1J design basis is senserva'tive esty if nthe 3.6.4 a). Our 3

' #g ha project " sable installaties setes and detail "a n orm manner.

s should point

&I out to the installers the possibt's hasard of embl ta the trays (especially en inside eers e piling up

{} a fraettee of the maabte width of the trayfi ers of tray head

, for azample, for II .

a I We createshoulda seriousalso be aware that our swa ele heat probles.

If power table trays arectrical equipm

' (j II tray (s) will areate a higher temperaturerouted see(s). e the lower s

)J insulated lead aester (AA or FA rated) ass I transfThe 1 ambiest eseeeding the total temperature espabilit ormer creates an

" standard" aspecity. sables - the sables would e y of our verheat at sage,s l

layeur (conduit er tray) er special s providad.

vest e ay 14stieSu H

] _ _

NUMBER E2.6.4. Rev g gNEET 26 05 DATE _ Cg. 29, 19

$slar radiant heating can clos soricuity offect the cepatity of cables if their rasigt is :nne te Ihs exceeded. In mest I

cas a ss, .naturst ' breeze nr .f arved . air .s:12-culation substantially E reduces'the affect, thut considecatiet. mhould be give:Its both j Jedoor and utdnt whis, rune :ta ; tray nr conduit where solar yj .*zpaturs sn .eubstarctic] rmd 'pe My cascentuated by restricted 43 'venn lation, Orte method rl sar. list vit.h the probles is to jJ prov.ide-suasitelde of gnets :metat er scher suitable paneling.

J Another appsac's is u destraima 'the 1saximum estimated g :essersture fee a secland 4rn .in direct sunlight (or 3 usinaulased 4tr>1s ressentwre) froue project plant facilities II engineers . sed dexuts : he dahle far operation in this signer l ambiest, d mothed of divestly satseleting the required Dadditiemal doratia's is"se't forth is the Neher-Macrath Paper 5{j

, "The Calculaties of the Temperature Rise and Imad Capability 3 of Cable systems" es page 759 under " Aerial cables". To semplete the calculaties outlined thereia, it is helpful to refer to the " Method of Calculaties" seeties en page IT of the j Introdesties to ICIA Pub. So..P-46-424 and stiliae the values I{g i

in Table IV of that secties.

i l 4.0 REFEREscEs ,

4.1 "The Cales1 sties of the Temperature Rita sad the Lead

}I.

a Capability of Cable Systems", J. E. Neher and M. E. McGraths  !

l AIEE Transactions Part III, Telease 76, Osteher 1957.

I 4.2 "Ampeetties for Cables in Easdenly Filled Trays", J. Stolpe IEEE Transactions Paper No. 70 TF 557-FWE, April 1970.

4.3 " Engineering Data = copper.and Aluminum Conductor Electrical dg Cables" - Okesite Campany's Su11stin IIB-81. t 4.4 "Aapacity of Cable in Ceeered Tray". C. Eagnanas IEEE Paper ag I

so, 33 sg 303 0, Presented at the IEEE/ PES 1983 Summer Meeting g in Los Angeles, May 1933.

11

[8 al II l'

.l t.

il SP-E264d2 u NWSER E2. 6. /. . R av l p

SMst? 27 op DATE Oct. 29, 19 g

F.3 22 t3

W DOCUMENT NAME: G:\DUANEARN\DUA85547.RAI 1

ORIGINATOR NAME: G. Kelly SECRETARY NAME:

SUBJECT:

DUANE ARNOLD ENERGY CENTER - REQUEST FOR ADDITIONAL INFORMATION (RAI) ON THE DUANE ARNOLD ENERGY CENTER THERMO-LAG RELATED AMPACITY DERATING ISSUES (TAC NO. M85547)

NAME DATE i

1. D. Foster-Curseen

2. G. Kelly

3. Secretary - Dispatch PLEASE DO NOT REMOVE THIS SHEET FROM PACKAGE CAtl THIS DOCUMENT BE DELETED FROM SYSTEM 7 YES NO SHOULD THIS DOCUMENT BE ARCHIVED? YES N0 l l

l