ML17354A430
| ML17354A430 | |
| Person / Time | |
|---|---|
| Site: | Turkey Point  |
| Issue date: | 03/07/1997 |
| From: | FLORIDA POWER & LIGHT CO. |
| To: | |
| Shared Package | |
| ML17354A428 | List: |
| References | |
| PTN-BFJM-96-028, PTN-BFJM-96-028-R00, PTN-BFJM-96-28, PTN-BFJM-96-28-R, NUDOCS 9703130426 | |
| Download: ML17354A430 (66) | |
Text
Page i CALCULATION COVER SHEET Ca 1 cul ati on No:
PTN-BFJH-96-028
Title:
Fire Barrier ACF for T-La 330-1/770-1 Assemblies Initial Issue Descri tion REVISIONS i
'F703130426
'770307 PDR ADOCK 05000250 P
PDR Form 82A.
Rev 6/94
0
Page ii LIST OF EFFECTIVE PAGES Ca 1 cul ati on No.
PTN-8FJH-96-028 Rev.
Title Fire Bar rier ACF for T-La 330-1/770-1 Assemblies Pa e
.1llill 1
2 3
4 5
6 7
1,2,3 4'4,5 5
5 5
6 Section Rev.
0
~
0 0
0 0
0 0
0 0
0 Pa e
Section Rev.
Al (23 Pages) 0 Form 828.
Rev 6/94
Page iii TABLE OF CONTENTS CALCULATION NUMBER PTN-BFJM-96-028 REV.
SECTION 1.0 2.0 3.0 4.0 5.0
'6.0 TITLE Cover Sheet List of Effective Pages Table of Contents Purpose/Scope References Methodology Assumptions/Bases Calculation Results PAGES ATlACH NO.
T-ITLE PAGES Fire Endurance
.and Ampacity Testing of One and Three-Hour Rated Thermo-Lag Electrical Raceway Fire Barrier Systems 23 Form 82C, Rev 6/94
0 0
0
CALCULATION SHEET SHEET NO.
1 CALCULATION NO.
PTN-BFJN-96-028 REV 0
1.0 Purpose/Scope Ampacity correction factors (ACFs) were developed for Turkey Point use based on TSI Thermo-Lag (T-Lag) 330-1 material for 1-hour and 3-hour rated fire barriers (References
- 2. 1 and 2.2).
The purpose of this calculation is to develo ACF values t'or T-Lag 330-1 with T-Lag 770-1 overlay (330-17770-1 composite) assemtlies for 3-hour rated fire barriers for power cables.
2.0 References 2.1 2.2 2.3 3.0 Calculation PTN-BFJM-96-005. Revision 0. "Fire Barrier Ampacity Correction Factors-Extrapolation of Test Results for 3 Hour Barrier" Evaluation JPN-PTN-SEEP-96-011.
Revision 0,
"Review of Ampacity Rating for Power Cables in Conduits and Trays with Thermo-Lag 330-1 Covering" Report, "Fire Endurance and Ampaci.ty Testing of One and Three-Hour 'Rated Thermo-Lag Electrical Raceway Fire Barrier Systems" (included as, Attachment 1)
Hethodology The approach of Reference
- 2. 1 was to use heat transfer analysis to extrapolate from Texas Utilities Electric test data and develop ACFs for single conduit. cable tray and banked conduit.
Heat transfer was calculated per foot of raceway length as a
convenient relation to,the test results.
The analysis was taken a step further by Reference 2.2 which evaluated power cable ampacity ratings.
This calculation begins with a similar approach to that of Reference
- 2. 1 and evaluates the effects of augmenting existing T-Lag 330-1 material with T-Lag 770-1 material.
However, this evaluation continues further by also considering the findings of Reference 2.3 to establish appropriate derating factors based on testing and to include boxes among the assembly configurations.
The analysis starts with the values calculated in Reference
- 2. 1 for trays and
- conduits, then considers the ampacity effect of overlaying with T-Lag 770-1 wrap.
As such. the heat transfer equation used in Reference
- 2. 1 is revised as follows:
q ~ (T,-T,) /
(R5 + R
+ R, + R5) where:
g Rate of heat transfer from raceway Temperature of conductor (fixed 9 90'C/194'F)
T,
=
Ambient temperature (fixed 9 40'C/104'F)
R,
=
Thermal resistance of all items within the raceway Thermal resistance of the air gap between the raceway and the T-Lag R,
=
Thermal resistance of either T-Lag 330-1 alone or with the T-Lag 770-1 overlay R,
=
Thermal resistance at the surface of the raceway The heat transferred from the raceway under steady-state conditions is essentially equal to the I'R losses within the conductors.
These heat transfer values were determined from the test data based on the measured current and size of conductor used; as documented in Reference
- 2. 1.
Form 83, Rev 6/94
0 0
CALCULATION SHEET CALCULATION NO.
PTN-BFJH-'96-028 REV 0
SHEET NO.
2 The thermal resistance values of R,.
R..
R, and R, are also based on those determined in Reference
- 2. 1 for 3-hour>ated t-Lag 330-7 assemblies.
The'value for R, with the T-Lag 770-1 overlay will be calculated assuming that the thermal conductivity for T-Lag 770-1 is the same as for T-Lag 330-1.
After thermal resistance values have been established.
the heat transferred can be calculated for the raceway with the 3-hour rated 330-.1/770-1 composite fire barrier.
4.0 4.1 4.2 4.3
~..
4.5 4.6 4.7 4.8 Since the heat is a function of the current squared.
the ampacity correction factor (ACF) will be determined by the following relationship.
ACF = I,/I = (q,/q)" where the subscript "p" refers to the protected raceway Assumptions/Bases The effects of inductive losses in the raceway and cabl'e sheath are negligible with respect to applying the test data to the Turkey Point configurations.
Surface emittance for cable,
- raceway, and T-Lag is assumed to be equal to 0.9.
Note that a high emittance value will reduce the heat transfer at the surface having an overall effect of maximizing the ampacity de-rating.
Heat transfer is assumed to flow perpendicular to the surface.
This allows one-dimensional
- analysis, is reasonable for single conduits.
and is conservative for cable tray and banked conduits because no credit is taken for heat flow through box corners'eat transfer through the sides of cable tray is assumed to be zero.
This reduces the heat transfer equation for tray to one dimension.
As the tested cable tray is relatively wide (24"), this is expected to be a good approximation for all cable tray.
Also, the horizontal tray configuration is more conservative than vertical in terms of reduced heat transfer, based on typical emperically developed formulae for natural convection coefficients over horizontal and vertical plane surfaces.
Convective heat transfer flow is assumed to be laminar.
This is reasonable for "stagnant" rooms and conservative in outdoor areas and rooms with forced ventilation, in terms of reducing the heat transfer rate.
Pre-buttering with trowel grade T-Lag material is assumed to virtually eliminate air gaps between T-Lag layers such that interface resistance effects are negligible.
The 3-hour...T-Lag 330-1 assembly is assumed to be at the 1/" nominal thickness in accordance 'with the manufacturer's tolerance
(+ 4").
It was judged to be excessively conservative.
considering application techniques and material cost, to assume a maximum thickness'f l~g".
Each layer of T-Lag 770-1 material is assumed to have a finished nominal thickness of 3/8" (including buttering).
Two or more layers of wrap are assumed to be.
consistent with the tested configurations (Reference 2.3).
~4. 10 Conduit which is banked in a single plane can be assumed to be equivalent to cable tray.
Both configurations involve a cable mass arranged in a shallow rectangular section and it is conservative to assume an air gap between cables in the tray and the T-Lag material.
4.9 Raceway is made of rigid steel.
which is typical for power. plant installations.
Form 83, Rev 6/94
0 0
0
CALCULATION SHEET CALCULATION NO.
PTN-BFJM-96-028 REV 0
SHEET NO.
3
~
~
4.]1 The thermal resistance values for all items within the raceway and for the gap between the conduit and the T-Lag material are assumed to remain constant as additional thickness of T-Lag is installed.
Considering that the geometry of, these areas is not
- changed, this approximation is
,reasonable for the purpose of extrapolating the thermal resistance from raceway with T-Lag 330-1 only to raceway with composite 330-1/770-1 wrap.
- 4. 12 The thermal conductivity of T-Lag 7?0-1 material is assumed to be the same as for T-Lag 330-1.
This is a reasonable approximation since the two materials seem to be chemically similar.
- Also, based on discussion with Kent Brown (co-author of Reference 2.3), material property data provided by TSI has been inconsistent, and is considered unreliable.
- 4. 13 The radiant heat from the sun does not adversely impact ACF values for wrapped components because the insulating-effect of T-Lag reduces sun load with respect to an exposed component.
5.0 Calculation Formulae-,
constants and parameter values are as presented in Reference
- 2. 1 unless noted otherwise.
,Detai'led calculations are performed by spreadsheet and the results presented in the table at the end of this section.
5.1 Determination of test heat loads Test heat losses for unwrapped raceway were calculated (q=I'RN) with the following results:
Raceway Size 2
tl 5
Il Tray (Conductor )
(1-3C/¹'10)
(1-3C/¹6)
(4-750 kCNil)
(126 -3C/¹6)
Heat/Ft BTU/Hr
- 22. 6 23.7 99.7 382.1 5.2 Determination of Thermo-Lag R values (R,)
For heat transfer through T-Lag cylinder:
R, - ln(r,/r,)/2vkL r0r kayo kv L
Outside Radius Inside Radius Thermal Conductivity = 0. 1 BTU/Hr-FT-'F Thermal Conductivity - 0. 1 BTU/Hr-FT-'F Length = 1 Ft.
(Per, Foot)
For heat transfer through T-Lag sheet:
R, = l/kA 1
4~
kno A
Thickness Thermal Conductivity = 0.1 BTU/Hr-FT-'F Thermal Conductivity. - 0. 1 BTU/Hr-FT-'F Surface Area Form 83, Rev 6/94
0 0
CALCULATION SHEET CALCULATION NP.
PTN-BFJN-96-028, REV 0
.SHEET NP.
4 5.3 Determination of surface R values (R,)
The surface resistance considers free convection andiradiation heat transfer:
qs = qc + qr q,
=
.Heat transferred by convection q
Heat transferred 'by radiation For free (laminar flow) convection:
q, - AT 0.27(hT/l )"MT For radiant energy:
A L
Convection heat transfer coefficient:
Horizontal cylinders in air h 0.27(BT/L)~
Vertical planes in air h =
0.29(hT/L)'orizontal planes facing up in air h
0.27(hT/L)"
Surface Area Characteristic length in feet -(diameter or width) 0
,q, = crAe(T,'-T,")
cr
=
1.714x10 'TU/Kr-Ft'-R4, Stefan-Bol tzmann Constant
,A
=
Surface Area Surface Emittance
= 0.9 T
=
Absolute Temperature.
R (460 + 'F)
For total heat transferred from the surface:
q, = 0. 27(hT/L)"MT +
- 1. 714x10'(0. 9)A(T,'-Tz')
q, = f0.27(DT/L)" + 1.714x10'(0.9) (T,'-T,')MT]MT hT/qs Rs 1/
LO 27(GT/L)
+
1 714x10 (0 9)(Ti Tz )/hT 5.4 Calculation of ACF The ACF is 'calculated in accordance with the methodol'ogy described above.
A description of the headings follows:
OD/M This is an input value of the conduit outside diameter or cable tray width.
Conduit diameters were developed in Reference
- 2. 1.
TH This value is the thermolag thickness.
For each raceway size a thickness representing 330-1 wrap (TH330) only and the 770-1 wrap (TH770) is entered.
ODT This is the outside diameter of the raceway with calculated from the OD and TH.
For cable tray QD'is not calculated since,it wi;11 always be equal to W.
Form 83, Rev 6I94
0
CALCULATION SHEET 0
CALCULATION NO.
PTN-BFJM-96-0 8 REV 0
SHEET NO.
5 A
The outer surface heat transfer area.
Note that for raceway both the top and bottom areas are included.
Area is calculated on the basis of a one foot length of raceway.
Ri Insi~thermal resistance as defined above.
The value was calculated from the test data in Reference
- 2. 1 and is not recalculated'ere.
Rg Gap thermal resistance as defined above.
The value was calculated from the test data in Reference
- 2. 1 and is not recalculated here.
Rt Thermo-Lag thermal resistance.
The value.is calculated in accordance with the following equations which were developed above.
Conduit Rt = 1n(ODT/OD)/2rrk k = 0. 1 Tray Rt = TH/kA k
'0.1 Rs Surface thermal resistance is calculated in accordance with the following equations which were developed above.
Note that the BT in this equation, is between the surface and ambient and the 7'alues must be in 'R.
The ambient temperature used is 104'F/564'R.
,Rs = 1/[0. 27((Ts-104) /ODT)'+ l.714xl0'(0. 9) ((Ts +460) 4-5644) /(Ts-104) 3A Ts Surface temperature of T-Lag, determined by iteration until q - q'.
q Heat transferred, cal'culated as follows:.
q = hT/(Ri + Rg + Rt + Rs),
Where hT = 90 F
q'eat transferred from the surface.
calculated as follows:
q' hT/Rs, Where BT = Ts - 104'F From continuity, the heat transferred from the surface is the same as the total heat transferred.
In order to solve the various cases.
Ts is adjusted by iteration until q - q'.
ACF Ampaeity correction factor calculated by the following equation which was developed above.
ACF (q,/q)~
The ampacity correction. factors for 3-hour T-Lag 330-1/7?0-1 composite assemblies extrapolated'y calculation are as follows:
Item 3," Conduit (2 wraps)
F" Conduit (3 wraps) 24" Tray Banked Conduit ACF 0,84 0'3 0.54 Form 83, Rev 6/94
Cl 0
I i
1 I
~
~
~
)
~
~RIREBIKHI~~~~~~~~~~KKK
~KKRKRZRMRMRRlaRISEEIEKESIQEBSKHES~HERIKBZa~
~WWW~~~~~~~
EiiKZIQESRK$9l&3WKSK'E5%RXER&KHME%&KRRE~RHHSRHH5&K%
KEG'HQK'lKKQ'RHEKE5%KKKRKK5iR&KQRKEESR~ZHRRERE%%5EESRKMK'Egg QQiigsmgi@iigi~i'gQ~~~~~ggQ~QQ~~gQQQQJ~~ggmmiigg EK~KHRRRHE9RRRRKR%KKCKRM2!%~~ME'%%3KRKEEKQKKERNRK%
MHRRRESEIQKEBRXGRREKRKRHSKKSRKMRKMRHK1KXRSK'ERR KRSRQRBHIHKKSRxMRKGRRB&RKMREkRRR)28%KEQKK~KIKEHQK~EB~Q
~WWW~~~
KRRZQHH)U~~m<<RESa3Umammaxmam5mazmazmaIEEI25aK~2RREEE~xm5%
%52tSM9KRH9RKt29RXEERISKkRKE2%RXGRREWRZERIMElEREKSRHEIRR2%
KRR~RKRKEHREHIRERRRKESRXEQRXKSK'EERREHSIKKQ'RRXRRZRIRZE5
~~~MR~~
ERRZRIKZlHRUCKBRXQHRZKSKKER~EWKZMREHSNHE15RKRRRKGSMIM aEZmiK23iEHRH99aKmaERaaKKamZmaKHaaEHXaEIEEIatE9%SEKRWRXZa EHQ>55iK23i5EHRE3M'ZZQRKKRKKESRXBRRRkMRXERKKER1SH'E5528KEKEEg
~RiRRIR~~~MHM~~WEWMEW~
~gggggg~~~s~ggggQgggmgggJgmgggg~~ggrimgglggg~
amiaEIRlKel~~aKeamKHam5mmxamHarHKKSRma mam REfRBK8IKW~RK9%
RRSRRRSRRRRERNHKH NEER RES
~+igigggiig~igiJ~~~i~ii i
i
~i~i~igg+~i.g+gp~i: ~i~i
~ gg~i~~i.g~
0 0
~O
CALCULATION SHEET CALCULATION NO.
PTN-BFJM-96-028 REV
.0 'HEET NO.
TYA performed ampacity tests as described in Reference 2.3.
The testing was performed using an even number of conductors.
pre-formed T-Lag 330.-1 dry-fit to the component.
and other such criteria per IEEE P-848 Draft 14.
TVA identified key parameters.
presented a reasonable rationale for:what constitutes conservative test configurations, and maintained key parameter limits while establishin ACF value for these configurations based directly from testing.
The ampacity correction factors tor 3-hour T-Lag 330'-1/770-1 composite assemblies based on tested configurations are as follows:
Item ACF Conduit 0.87 Tr ay 0.52 Banked Conduit 0.74 Although actual testing of conduits produced an ACF of 0.87, TVA established a value of 0.82 as their standard to allow for variations in emissivity for differe t conduit materials used.
or i
eren Banked conduit testing was performed in box-type configurations.
The ACF value of 0.74 is appropriate for banked conduit assemblies and conservative for larger boxes.
Therefore, since increasing box size increases convective cooling rate and the void fraction is larger than typically found for cable trays, an ACF value closer to that for banked conduits is reasonable for wrapped electrical boxes.
Accordingly, an ACF v'alue of 0'0 is selected for boxes.
Also, the tested configuration for 1" conduit appears to be with 3 wraps of T-Lag 770-1 material.
whereas Kent Brown (co-author), indicated that only the 2-wrap configuration was used for ampacity testing.
,Even so. the di.fference in calculated values, is not significant, and is conservative with respect to test values.
6 '
Results Calculated results for conduit and cable tray are reasonably close to those established through testing.
However, reliable testing automatically accounts for the myriad variables and the effects of their interactions, whereas calculated results depend on the validity of the model analyzed as well as the assumptions employed.
This is particularly the case here with the assumptions for material thermal conductivity and using a conservative value for emissivity.
As such, test data is considered to be a more rel.iable basis for determining ACF values.
Since conduit testing included only 1" and 4" conduits, the calculated value is used for the 5" conduit.
Therefore, the following ampacity correction factors shall be used for 3-hour rated T-Lag 330-1/770-1 composite assemblies:
ACF Conduit (<4")
Conduit (5")
Tray Banked Conduit Boxes (selected) 0.82 0.?4 0.52 0.74 0.70 Form 83, Rev 6/94
41 0'
~v~e I SSyy Oa 4V\\
813052488231 2
Mark H. Salley Senior Engineering Specialist Tennessee Valley Authority Calculation PTN-BFJM-96-028 Revision 0
Attachment 1
FIRE ENDURANCE ANDAMPACITYTESTINCPQF1 o< 23 ONE hND THREE HOVR RATEDTHERMO-LAG ELECTRICALRACEWAYFIRE BhRRIER SYSTEMS Kent O'. Brown Senior Elccuical Engineer Tennessee Valley Authority ABSTRACT The Nuclear Regulatory Commission (NRC) has determined there werc dcffciencics in the design, testing and installation of Ihermo-Lag 330-1 Electrical Raceway Fire Barrier Systems (ERFBS). Their two primacy concerns are:
~
The currently installed,Thermo-Lag 330-1 ERFBS may not ensure that tho eiecuical cables aro Gee from fire damage for the required rating and, The ampacity derating factors currently used forThermo-Lag 330-1 ERFBS may not bc correct.
Thermo-Lag 330-1 has been utilized in thc design ofthe Tennesseo Valley Authority's (TVA) Watts Bar (WBN),
Sequbyah (SQN) and Bmwns Ferry (BFN) Nuclear Power Plants to protect redundant cables required for safe shutdown.
ofth) reactor in the event ofa fire. Such protection is a licensing requirement per the Code ofFederal Rcguiationa,'0CIIR50, Appendix R.
After reviewing other avaiiablo ERFBS options, TVA determined that redesign and qualipcation of the Thermo-Lag 330-1 systems would provide the best available',ERFBS and minimize cablo replacement due to ampacity effica'-. Thc'to+Vmj.fire'and arnpacity. test prague'mYdeveloped one and three-hour rated Thermo-Lag ERFBS forcable trays condditiI,'iutdropj'ifaCtion'b'oxes an@ichir'uniqtio raceway configurations. The imptcIved designs and installation.iec&illucs.also Siuilinad ilrorrettuired.u+kip'hiaimg. These new designs will be used in inidal instalhtipns at WBN and for upgrading other systcins'g other TVANuclear Power Plants (NPP), The TVAresearch. and testing'also pr'ovided useful input to thc NRCs criterion for ERFBS'fire testmg and to an IEEE stanchrd being'eveloped which'gav'ertts ERFBS ampacity tosting.
Tho TVA'lnpacitj'estsrals'o <<stablishcd methcIdologics for use withiricewayeitifigurations which fal14utsidc the boundi ofthose'ccivered by tho draft IEEE standard.
Results ofampacity tests performed by Texas Utilities Electric (TU) were also compared to an existing mathcImatical model for cable trays which havo an ERFBS installed. Modification to kcy parameters ofthat model ate suggested which result in;good correlation with the TU tests.
~
q
~
,f480. ',!
dbms 'Rr; tB'd'h separinng"rediid8ait's'i cl" 1ated ERF5S'."'ha'dCutfonfeme'cting thyfiroendiirancol cablhr,a54't'quipitienL'mu'st. hiiaa'three Sik'" "- rc'qiiirements', the'EIPBS'6iuldnoc id'vcisilly'impiict-'aieas'without
&Au'tomatic'fire'.dejTe&i;""!-'-ttie'impacitj'dctat'iilgoAhec'abf5or'pr'esenr an impacti fon-'or'a Wi"Sour'.iiimjfoi'are'ai'%noh '-'r'-'-"hiza8 "4 '"sif~fe~"-'cqttiptneat duritig an "sutomatic"'fire'-'detbdtorr"and si" toaP "1"'-"- -earth '."-".'-"'-'1 -~".'-"8 App&
FSSO rating ib and"s pprcss insiallcd quak ppfl$5 whciiTAppendbt R.was prontulgaled,'tho titajority~F'-":
I L~ODUCTION
- '=
." '-- NPP werc'pcriting or in'fmal'stages'of construction:
i Whileredundant shutdown cucuits were a part oftheir Thoro Lag 330-1 Electrical'aceway FliiSarrleri
asfc'dciign, ieduhdant FSSD circuiti'were often not SysteiIss (ERFBS) hive 5ecii used in the'frtajoriiy'of separated by rated fire baniers. Thc most cost. ffectivo United States NGclhr'RhkYPlanls'ks iyaislvo ftro'
'bickfit'to meet the now rettuiremcnti was to protect bamer to protect. redundant electrical cables and
. ono ofthe installed redundant FSSD circuits. Areas in equipment rHgUgd fhfe ShutdoiIjn'(FMpuf" -;
the NPP that have," iiis6died automatic f&detcxtio'n:an'd the rijc6x. (d'"' '
.~ tho reqIliremedH'df th'"" '"'suppiession rsyste'tns refy'.ori 'odikloitr'ra'tcd'RFBS'
'\\>
~ ~ ~ ~
~
w
~ ~,
~
l a4
~
~~,
w
~
0 45 0
~J
vI
~ ~
~
IJ
~
'4 1V OO' 1
1 5 I&M outa. Engr.
8L. svcs.
813052488231iu 3
~ ~
Thermo-Lag 330-$
is a" proprietaiy'aterial manufactured by Thermal Science Incorporated (TSI)
St.
Louis MIssopi rizcd as a "sub Icvatcd temperatures ih lid into a vapor vithi tive shielding/cooling space industry d
Id for vehicle r~
Thc material is best ing ablator.". When exposed c materiaLtransposes from'h results in heat blockage
). The material originated in
,was primarilyiusrd as a'heat into'ihc aunosphcre'.
of ch fo a 5 (ab the shi CO CERNS Fi Protcctioa -,
8, deQciencies were identified in their Thermo-Lag I installation. ln the GSU testing,.30-inch wide inurn cable trays detected with 3-hour Thermo-Lag 30-I ERFBS exceeded pass/fail tcnipcratures and ost ircuit integrity witfiinapproximately 60 minuter.
phic: failure 'atId collapse of.the cable. tray within90.minutes.'he purpose ofthe GSU was to evaluate tlteiras-itistaHed.configurations.:
lmEdetermincd~g maintenanceactiVitiet dao bcontractot'who inktalled the Thertti'~330-I:
rcmov'ed:the-"sfsldn"'from thi::firer barrier:
.'Str'ess skin is a:steel wiie'mesh iYistalledhy.
dor.@apart:ofihe'ication of Se%hermo'-Lig'nc'-hour Thc'rIrio-Lhg'.330-'P l1as'stliss skin l9 33 alu I
C'S the mat a}.
33+i'.
$33~l'hree-hour Nermo-k.ag
.on both the-inside and repeated: the, fire.test by biles.in~rdancewith Iatiott;:manuaL:':Siinilar "tinftcd to7oifow"up-on'Diforrnaiie7Ndtkcd
'su'ppoit ot'the"Iicen'sing ontlucted-'ad4itio'n'al fHe.
"-FScrinWlt-:330-'P
'OfstnaIVcd'nduit and testing, tfiiPNRCBsued t'l"ilrJi&loand August QiitTherltio-Rig-330-"
'Ihsmaller tKm4-'inch wide;-wistma5Ie to:
R a rated'fbi'd on the inside~;
has itress skin Inpalied de surfaces GSU en ctingMHtionai m
endori published'al occurred'ltcrNKC Icon eject issttittg
'ntl Utllidc'iEsca@
'in anch'e 2";c
-tcst5rg of
'hour
- Based upoii lcd'Mures
'cable tray's dtukgf
'VLF' 92-Ol'~dsup cmen This Bulletiid~dminctf instttlled o~Ot condu ie trays:large'r d~k'l~
ItsdestgitMsis:t~tijjrequ outs the fall" this T
of cn ERF I
Bull l
I;.as and Begprupg m Augustaf I99I thc NRC alerted:.utilities to tcntial problemsassociated with Thermo-Lag 330-I E S'. Based upon fire testing performed by Gulf S
s Utilities'GSU) kiver Bend Station in October:
\\
Caicu1ation PTN-BFJH-96-028.
Revision p
Attachment 1
Page 2 of 23 barrier This Bullctm was then followed by Generic Lager 92<8'n December 1992. With all tho conccrtt Involving.Thermo-Lag 330-1 ERFBS and indcpcndcnt testing being performed by utilitics and the NRC the following question evolved: "What is the correct fire testing criteria 2:".
" Atthe time. this issue came to light the TVAwas in the final,stages. of.completing construction and. licensing WBN Unit.l located near Spring CityTennessee..The
...original.. design.of,,WBN. utilized... an. ERFBS
.. manufactured,,by the3M Company. Aftercomparing
.;. the ampacity deratfng factors. ofboth 3Mand Theimo-
.Lag;330-'I;, the decision.was made ta switch from the
.: 3M.ERFBS to Thermo-Lag.330-I with its lower ampacity derating factors.
Since WBN relics on Theimo-Lag 330-'1 ERFBS, TVAthen volunteered to assist the Nuclear UtilitiesManagement And Resource Council (NUMARC).with engineering and craft personnel in performing vendor sponsored'Ihenno-Lag 330-I fire testing that could be used at WBN as well as other NPPs. (Note; The Nuclear Energy Institute (NEI) is the successor organization to NUh4QLC). Afler
'umerous'" meetings:.
and discussions; with the
- NUMARCstaff and other utilitics 'represented oa the.
NUMARCadvtsoty.comnuttee, itwas determined that aconscnsttson:thc'appropriate firetestingmethodology an4 ttcceptance':criteria:could not: be reached'ia the sho'rt'term.iequiied.to support WBN licensing. TVA
~would have to 'proceed on its, own in qualifying
-: Thenao-Gag:330'. is'a rabbet ERFBR H
Thd'ecLsi65 toi~llrotect.tfLc:4kles,at.%9M'with aK
- ERFBS'occurred:afler Ihe'Cab%i w'eie installed'ut their raceways.'v'tiiuatioii oftlioe~of thc ERFBS on'he initalled caSfcsrw'as" initially'p'eifortnetin themid 19804 utiliziiigaiiipacity cbttectfott actors based on thiiesults of~ a+ducted by oi'forHs mattufhcturer; TSf I'n thi earIJJ 1990s Notices, Bulletins and Generic Leneri fro'm the HRC atuf position'papers'rom various industty~upficKntified:potential probletns with the originaV'ampacity'ert'rith-the'mbst 'significant coaccrni fn the tuea ofpiotected'trays' comparison oftho'cotrocdotr facrora o'rigirutllyutilizedis TVAanti thermo:which were snbscquendy published is yven in ta.>>
i m "ont irz:tisane
~ "
lg
~ ue p \\
~
( "$( tl
~. ". ~.!i"""c.':~'~~ia< '"!'>"ws se:W' ~">:"
0 0
0
'V I I ~
~
~
0
~ M ov
~ I
~
'JAg~
Ll I
~ t
~ I
~
uuxo engr.
o ~vcs. >>
81305246623.1:o 4
.Table t Comparisob orAapacisy Corrccsios Fiorors'(hCF)
Calfiounaks OI1 TVAACP~
OAer SowesACt TSsyosaOVC,
-0.475
~
~:
~
" OTIC"'
V S
5 Calculation PTN-BFJM-96-02B
'Revisiori 0-Attachment 1
~
Page 3 of 23 derariag, a layer was included in rhc TU ampacity tests.
The TU"correctioa" factor;(Table 2) was therefore conservative: for uscaat.ZVA'asWBN plant"where no siliqa.biankoawoufd bp isrilizecLl oi
. TU (fata Hoar,.Tbcrspo-Lag 330-t Tray.
~
5/8-iitdr la singlcI layer top oftha wa's aged yer'otdS%9ctmM+330-'1 jiirieli'withiF
'ofiMiMesPMbhtiikct placed'over th'
"~:A~gfo'-m; th~slmet to priv'id'idditiana1protectioa'Yor'cables in high/ filledtrays from the heat transmitted through the ba'meta'ff tlie firis-Sin'ce4e.tNerttfhl.asistiyity of tho sill@a oiotht"is about Advice ttuttwf:t5'e cabi~
itic!IARd'coulci'.tli'ctii7oRMuit"-iti'%YmSMdi%M".
l4oie'l.
~
ACF Wrappafhmpi'PSasclino imps"
.Notaz.
Baiiut on TSi sOoososcd sassing.'.. " n A review ofthe early tests reveals that there was iiule'onsistency between the various protocols us'cd for they amp/rcity assessment.,
This is not surprising:given thc absentee at that time'of an industry consensus method.
forthe perfonnance'of such tests.
Diffcrenccs existed in tlIe length ofthe test specimens;.tlie control'of end heatmg effects"the method'of ambient temperature control, the size and nmqber ofcables used; the size of
- raceway, the location! of thcrmocoupies and the defujition ofequilibrium conditions.
Variances in'such kcy Iparameters inevitably resulted in significant.'iffcreaccs in the derived correction factors.o:
i ln jrcparing;;for-theo,suhj act, test.,prograar..:TVA-pcrfcIrtned a parameter by paraineter review of-'th' prougufs" which'"h&e~- been issued'yhce=th origipattcstiiqp"T1re &a;:Uh 17M'.was-an issietl4 1984l ychile.the second, IEEE P-848'as still:in draft?
form[ Based.oa a 'camparislt of the'key"paraectcrs,'Eicti is"dexsscrribed'in an coariier'paper~'TVA'chSh
'to utilize the draft IEEE dopuncat,";:~-;
....'.:...-',< <~;=.'-, -
~
Atthe same time that TVAwas evaluating the varioui methodologies, TUw'af'prepatmg to pe'rfoim-'aaIpac~
tests".TU had also deterniined'that""title-'F448:draft""
provided the best'available'guidahce fdr=the:testssm5'ad Based its progr'am'ott Draft'1'I ---""i""'
XZ~;- Draft 11 reqttk5fSti'use'hf a 4=incH'x'24'-in'eh" tray filledw'8K1205A%9tWG'600'ihNco~ cabh5 arranged in four3~~TU baiiier'designs utilized a
..'.C~asa!SO4)-'
Trsy aosO4 as)-:
~ca i
'2.45 4
i
~
o 445 Quubljl.~:~The, TVA and TU condui< configuratioas differed,, Both"used the same-.&erma.-.Lag 330-1 preformed sections. however. TVAprocedures dictated that the interior ofthe barrier have a complete layer of trowclable grade Thermo-Lag 330-.l applied prior to installation., TU used a mare conventional method, employing only prebuttered joints.
Because of the resulting differenc in Qt, TVAdetermined to pcrfortn its own.ampacity, tests on canduits (using Draft 12) with the same configurations as those which were evaluated during.,thri.fire endurance. testing as follows:
,. a,sioaioktksch bye.,
~ '., 'i 38Rnc5 base covered Qiils i 3/S.hctf c.~~.i.o<<
'~iA (zyNach',bag;covered wtrh a 3/Bach C'
I As perttutrecfbyhft12, separiita coaduits were used for each. of, thise assembltes.. Tbo. utilization of differen-conduit:: segments.for cadi con'figs'uration aHowed.thetesbprograin~ymcccd'abc much;faster pact, thanj+they.bye,beyn. condgctpt.sertucntially on ihc same segment, due to the 30 day cure time (for the tmwelablc grade matemd), between eachitesr.
Given:.
that-.the thermal'esistrvity. of/water is fairly,low, its elimination..&m. theiseuns and= interior.-.prior to.
ampacity'testing'is required'in 'order. to:obtain results.
which accaratelydepict the effects oftheERPBS...
- ln the early stages ofthe program development it had beea assama'hat the configuratians:autlinedln.PA4$ meuM;envelope those
. which wouidibeiencarmtcredaduria.construction at WBN):Asactual,hisailbahaef the ERFMl.neared, lt became cleat that. many unique, enclosures would be rcquirecL.'These.unique eaciosttres Jlt hto,two broad categories. Fust~:thc qiidug4&odcfuitspacing; was.. not, suQicicat;".to. gcgp'k,instalhtsait, of, the 3
hg
0
~O
t
~ ~t
~ ~et ~
4 I 4Vg jjLOOLd 1 t
Calculation PTN-BFJM-96-028 Revision.0
.Attachment 1.
Page 4.of. 2$
'reformed sections'he only viable alternate was the enclosure of mangle conduits within-a canimon ERFBS. Second, where cabtes.in trays extended. above sideraits such that enclosure ofindividual trays was not 1.
pospible it.was determined,that multiple trays within the lame vemcal'stack would have to.be wrapped in a conImon ERFBS,
.Therefore, given'he:myriad"of combinations which can exist, onemajor component of TVA'.s test program became: tlic'.identifications of boupding test configucations for.:such<<non-standard arrangements ancl the development df some general rules, for,extrapolation ofthe fmdings TVkAPPROACH XO.RESOLUVfOY.:tn -".
t As IIrcviously stated Thermo-Lag 330-l is used at all three TVA NPP..Since thc problem was generic to more than onc TVA'NPP,, the "ownership" of the pro/em resided with Corporate Engineering (CE)
Chief Engineer.
The 6rst step in addressing the issue was to assembk a self-managing taun ofengineers to ownl this issue.
In order. to be successful Ae team:
woufd have to work at a'fast pace,'quickly overcoming any-mistakes or failures-'s'. At:TVA;thc respottsibility-forFire Protection Engineering/Appendix R is assign'cd:
to tbsp Mechanical/HucIcar. Enginecririg.
Department'M/NE):
Because of.the poteatuti:impact'onccabling.
and pn raceway seismic'integrity, theiEIcctr'ibad-'and CiviltEtiginmingDepartments;iri a'dditlon to'Nucleai LicetIslng, would 'hy: -key roke in isucces'sfully:
resolfing4m Issue.'h@
is "appa'rent'sinco Thcim&-
La'g )30 1 w'oald have tp.meet tmuitiple perfdimancc.
criteria to bc successfuL 'TTiaefote; lntorder to address and 've all the ues.assoc'atal vnth Th o- '"
rcso tss i
erm Lag 330-'f, 'thR spcc1aliiccL'self trfatiigiiig'fBmiappt'oach was scca as thc most ajpropriatc choicii; %ith thi NIfCtasuaiicc of"RiqiMforAdditionilInfoimatiot'i.
Rcgafding Gcncria: LcCjr.92<5, Issuai Pursuant to 10CPp0.'54(f) &5ekfa6er 22; 1994". aa-eifginecr.
from thc CE IvtaQrh5c&.5ipcctionMpaitmcnt wtN C
to the teatts! " '~"
jnn'tjnttnnjninjjtjtenp'ptttptjete:ftiet:tint'lttitjtit'tjojnjp-and acceptance criteri fbr ERFBS'thI. et'arting potntp'asp tci perform a lit'eraturts review-aF tHc=<<vailibS.
stan t7VA EngTti'cering'rc'yiewecf the:NRC I
~'
ouidancc itt Generic Lett'cr 86-10<<, ASTM E 119>>
alid the American Nuctcar tnsurcrs guidance provided in "ANI/MAERPRA Standard Fire Endurance Test MetIiod'to'Qualify a'Protective Envelope for Class lE Electrical Circuits""., Undctwriteis Laboratories (UL) also'had a methodology. avaIlable foriteIting ERFBS.
'IYieu cfocumcnt, UL Subject f724'r, vias reviewed and
,detcnnined to be the best starting point for developing tti'a methodology.'."TVA cfeqidcd that testing would bc
'erfo'ttned to the Standard Lime/Temperature curve as iisted'in,most fire testing standards (i. e. ASTM E119).
For'he TV4 Pfiase I testIng the decision.was made to test'the.racev/ass without cables...in their place woukl be a single bare /t8 AWQ stranded copper conductor.
This provides two.benefits: fiat, thc fire test would be conservative since'there is no thcntial mass ofcables to act as a heat sink during thc thamal exposure, and second, the testing would bc mde pendent ofcable types and fills. The bare conductor is instrtunented at 6 or l2-inch intervals and acts as a surrogate cable,. It' thennocouples provide a profile of the temperature environment, inside the raceway during the. fire test.
The exterior of. the raceway. is also insuumented.with thcnnocouples at expected high:temperature locations'nd 6 'oi. I2-Inch s'pactngntherciftcr,=
Because the ERFBS'encloses the entire riccway, there'is no easily dcfmable cold side ofthc-btftricr as thercr is with fire wall tests. Thcrefote by definition, thc exterior. oftho raceway is the cold side ofthe ERFBS:"Thc test data is.
thea i'ntapreted ae follows - -- '. '...::".
~
~
~ t l)"Sinco: the'.'ERFBS'tis ittstallcd-'for. physicaL
~ "scparauott-'and'fi?actions~-a,fire w~ if the average tempaature rise is ass than 25'12VC)
=-aiid the.highest single'thartnocouph risc is less';
tfian 325; V:(l63.'C) me'iisu'rcd:oii the 'exterio ofhc'raceway,'(ri.
ASTMEH9:Crttertts); ifshaH bc
'consider'ef rited"foriIItypcjs'aad fills'of cables.')
- Iftile'ERFBS'excccch these tanpjetature liraits; ts
. - time/temperature proSe cats be developed and thc.
specific
'cabl6'types then:qualified in'accordance with ttie requiiemcnt5 of UL'124:Appendix B
"(with'sctmc.enhancement/
Fire:bamer tcsttng~icallp Iticludes:anpcxposutts to a hose"stream 'to modcI itnpact; aoeiots'-and. cooling.
effects.
TVA;detaiminccLthat a rcaHstic hose stteatnt test eh'ould bc'pcrfonned atthecnd'of the'ftreexposute; n
I t
" ln
~et e
~
~ r
0
~~
ae:hi aY 6
10-66 t.i.i
." BA4M Gute Eager.
4.6~cs....,.
8,a306246623i 'c 6
~
w
~
-'=" "'CaTcuTat'tcIh&TN-"BFM<6-t}28-Revision 0 Attachment 1.
Page 5 of 23
~etio fait.
f-1992.
t co leto (
~ '
U n o'fhc first three Pha& I:test decks in:fifer.
c"Ttr'erforming fcstfttg't mttit.,
tS'bit%KenM'Gtg'330-1'hifsra:hhg lead'fimo with each'test hdtv tbo'nstalbttlaa-waa appioxitnatep ono w'eek"pei'test'decTtt for
'ased on the experience of the March 2~, 1975, BFN toe, the electrical raceways did not receive any severe impact force fitxtr'Mlingobjects induced by the fire.
This can be attiibuteif ro thc solid ctcsigned'suppon systems used for eitithquake requirements.
The one-hap"ERFBS will bc installed. in sprinkler protected areas ofthe plant to meet Appendix R requirements.
fn an )tctual NPP fire the ERFBS willbc'exposed'to the mechanical impact, erosion and cooling effects'of water spray fram sprinkler'system opcrauon and fire'brigade hose streams.
Therefore, it was determined that thc most appropriate hose stream test that would represent theimechanical impact, erosion and coaling effects in.
a T~VA NPP environment would bc the one and.one-half inch fog nazile testas describe'd in NRCs NUREG 0800'
~ ~
Aft/rthe completion af the TVAphase I testing and prior to the start of Phaso li, the NRC issued Supplement I to Generic Letter 86-10'~.
In this Su+lemon!
thc NRC defined the appropriate methodology and acceptance criteria for ERFBS.
tcstilig. Comparison of Supplement 1 to TVA'stesting pas4iaa demonstratef:a great deaf of similanty,-the majIIr diffcrc'nce being the nomimtf'spacing, of.
theqno'couples.
TVA's Phase 4 testing'tttilizetf UL'.
1724 guidance ofkey'locationi on the raceway'.ind'I2='nchI spacing thereafter.
Supplcmeiif%requires 8-inc}i*
spacfng.
Based Qpon the uiuform temperatuie prbfiies.
dev)hped in the Phascl I',-'tc'sts'lh'h differenc'~as:
detcimined to be insignificant'".- IIIevcrdiekis,'5%ii'o fF ancf )R testing would'utilizti'th'o 6-'inch thenaaca'uplo'V FERE.TEST RESUKTS. -:"-" -:
PhsuIe I Fire Testa"'
The fire testing waa dfv+xlinto three phases,.Pttaee I
~ basic elcctsfcak raceways such as; conduits, jungian boxes anil~ profcctedwith OOH~
ERFBS;.
'rcabhv.trays'nd'ultiple:
candItits inising& eniIQuro proWted witfta~M ERFQS;" -Pftaeai.IN tcstttd cotufuits and caMo. fnya:
prouictccf with "tfae&NQS ERPHK TVA bcgais two craAsmen to install thc ERFBS), thc assemblics must cure for a minimum 30 days. This curing time is required'or the.trowel grade Thermo-Lag 330 1
material to,dry. The fiat decks consisted oftwo each:
1-inch conduits, 5-irich conduits. and 2-inch au drops.
Varia'tians in the ERFBS Cosign would be ippfied to each ser.
Thc'goal 'was to determine the effects of conduit siae.on identical ERFBS. The first test was run on December 21, 1992: Thii test had a baseline, nominal 5/8'-inch Tliermo-Lag 330-1 ERFBS installed on each'set ofcomponents. Additia'nally, the second sct ofcomponents'(I-inch and 5-inch concfuits and 2-inch air drop) had an external'layer of stress skin applied over the completed. Thermo-Lag 330-1 ERFBS.
The purposo af t&is.variatioa'was to ancmpt to achieve
'etter. performance, as suggested by, thc smaII scale test rc'suits as reported in Information Notice 92-55'. This additional stress skin would. also act as an cxtcrnal skeleton should any structural problems develop durmg the test.
The test demottstratcd that a single layer of nominal 5/8-inch Thermo-Lag 330-1 was adequate to protect the 5-inch conduit but insufficient to ptatecr the l-inch.conduit or tho 2-inch air drop for the required period oftinlo'(i.oone hour)
An'intc'rcsifng prebfetn was dBcoverek'during some earltcr- 'tcstlllg "Lnvol~"..ianotnalies'weal:
thc thcimocou'plo. r'eadingi. -TU'-*iacxifcricficingerratic temperatuio dibs in sornb oAMir'tests. Aftcrsotne researcher by'c-liboratory'On@a" Point) it"was:
dctetmutc'tf-'r fhtf..'twitch. thc i ThctTttyiLag undergo'tisubliming ablaifvo shielding process the tniiertal saturates the fiberglass insiihited~acoupio lea& causing theth to:short. 3oftifngthc testing, TYA ha'd expeiienced. tfti.'same an6ehUes: with fiberglass
-lrisulafcdthmn~h iastalJctf Me+er eoastrtlctcd:
~:Aaatcsuif:of these'probletni+yiiigg4+
a Tea Chap@
A1iteik TVA:continuetf'constructing: Phaso I test decks and performin'g~tesfs4iniugh:tho spring-'o$ 1993 For the first "upgrade ~'a nomfnaf 3/8-inch preformed section-wae ahtef aves the nominal 5/8-inch first laycrx'.i An interesting discovery.wis made during the.
first:5/8-incft+3lMxJitese~ccrning tho two layer.
'Ofcttno-Kag330-'I ERFBS perfortnance (see Figure 1).
Ther two-layer ERFBS CHtnbcd stcadtly up to approximately.2~ (40PC} (asnbfcnr tcmpcntturc,~
tempera&to sfre). Acthat pointthe curve Qattctiof whh
41
~
lc.
a
~ N~Qc40023i
~ 4 7
~
~P
' i""='"""(ad julftao~Tt7-'fggM-Obr028'Revision 0
~
- Attachment'1 page 6 of 23" only a small rise for the last 1/3 of the test.
Upon disassembly it wa! also noted the outside layer of material was cocnpiacaly consusncd and l/2-inch to 3/8>>
inch of the insidle layer was un-reacted.
Based upon this tost, subsequent tests wcte performed and qualified using a 3/8-inch ~ 3/8-inch ERFBS. design.
lt is theorized that the interface resistance bctwecn the two layers coupled: with the outside hyer" charring and stress skin maxiinizes the efficiency o'f the subliming ablative cooling on the protected electrical raceway.
The graphs shown in Figure 1 were dupficatid for all two;layer Thermo-Lag 330-'1 ERFBS installed on varying sizes ofconduitL Thc only difference in the curves is at what time into the test the sample reaches apprOximately 212'F (10'}, Due'to their smaller therntal mass (inertia+ smaller cond'uits. (and air drops}
reaclI this point faster th3n larger,'conduits.,Table 3
shows the TVA Thermo-Lag 330-1 ERFBS. conduit designs qualified in the test prognun.
r Ffipua I 5 Nar Cease~ rNtCeHas g5 i
5,,0 5
I
~
5
~~
~,
5 '
~
1
~ i 0 ~
'-i~
c555555I '
X 1 ls5555" r.flIloI I
~ I'il lI5e5'hasci I also performed tests to de'tertaine ifthere was any difFerence in thcrtaal erfartaance.'bctwcctk;:
aluminum aad steet. cotiduiti protected.by.ifctitical ~.
ERFHS. Steet is apprtaiaatidy three times heavier~
equal )ized contQlita 55gehttninutnhas approxiniatsly;u,:
ce
>the; hest:carr t:~g';firtt5;1css on 3 utch conduits with the saaSa 'Ihamo-Lag 330-1. ERFBS exposed to the same~ heat fhx demonstrated thc-aluminum conduita wiR:experience slightly higher temperatures (Reference ~ATest:6."1 A).
~ \\"
Tabic 3
'rhctlao Lac 330.1 CoaltllitS~ acaiaaa a Wo.l~~
5' Vl +l4 55'
)/p X
i x
i x
'5, x
i X'
X 5
4
- )c i'-;1-'
5 X
Nba i. Retina caMe qualNcadoa.
luactfoti box testing demonstrated a single layer of nominal 5/8-'inch Ihertao-Lag 330-1 with an exteriia}
hyerofstainlc'ss steel stress shnind trowel grade skim cow was adequiie to protect sizci rariy'tig from 6-ia*
x 6-inchx 6-inch through 24-inch-x-.t8 Inchz; }2<<inctL..
Phase'I also. determined junction 'boxes" with a diiri'ension gteatcr=than 24-.inch.up.through.48-hch x 36 inch x 12-inch can be success5dly ptotccted with a nominal:.5/8-inch-"+'3/8-inch:-- Thettao-Rag "330 1
Afterthe successful compietioa ofPhase Itesting, TVA.
catered into an agreemeat with the, manufacturer, TSI,.
forthe Phase I15and IIItesting, In this atrangemcnt TSl would provide ail accessary.'Thermo-Lag materials and laboratory fees-whili TYA.. was.-responsible for p
'~..th ~gm ~g,a m~ndiaion~
thc necessary electrical raceways.
'the results ofthii testing would birmade avaihible to any interested NPP'"
through NEL,. This.'5aizangetnyg would provide a "Win/%iii"philosophy for thi prolect and all patties involvecP'.-. In addldoo to the-fhwtests, corresponding ampacity dcratiag would'e pcr&rmcd on the configutationL5 The shy'f the.'testhtg was sec a.:
sevctr"orie-hottr5firi tests in Phaia 8 aalu-two/threeti'-
three hour firetests in Phase K':: -: =
5
~
a'0 i V
~
~
~ 'ZAVt"'
%1
'I
~
~
~
~ ~
Calculat>on,PTN, BFJH,96 028 Revision 0
~- 'ttachment 1
Page 7 of 23 Phase IIFire Testa The design for the seven one-hour tests was based on exiping installations itSFN and SQN in addition: io.the expected configurations'at WBN'--Based upon: ihe success of the TVA'h'ase: t: testingr: the design
-,philosophy was ta keep the "upgrades"'minimal aad geqeric as possible.
The tests would also aaempt to establish installation rprincipals through bounding configurations. The "Generic TVAupgrade"- consisted of adding a layer ofexternal stainless. steel stress skin with a skim coat ofThermo-Lag 330- l trowel grade.
Thq external stainless steel stress skin and trowel grade skim coat would:
not including the weight of the Therma-Lag 330-1 ERFBS which was constant for each tray) versus its temperature at 60 minutes an'expression for tho effect ofcables can be developed; r
Figiire?
1 Hert~~
CANISLyylIIIy C+
r
.r lryyy I r W%
thlOah I~ ILKSeO e
- ~
Provide a thermal boundary layer during thc fire exposure.
This will maxuniae the ablative shielding provided to thc raceway by the nominal 5/8-inch Thermo-Lag 330-1, i~
Ensures the 'RFBS rcntauts sttucturaHy intact during.thc dynamic process of the material subliming during the fire exposure and subsequent hose stream test, and
~ l -
Provide the necessary strength aad sttucturai integrity needed to meet seismic requireinents.'VA verified.. this'design featureh thesummei of'19944y canthcting
-'".. "'- fall sade. shake ta8'at Wyie Liabo'ratorics,'
"" Huntsville, Alabaini.
were.dedicated ta cable triyconfiguratiotts." TVA'-'Test 6:l:0 consisted'of thr'ee N-inch wide;hdder back, s&l'ab'rays with ldcatfcihupgraded ERFBS and Varying.
cabl) fiiL The lefttray uz the.test deck'represented'rf mm
- u. f-if~~(i.c.'2894CW6.AWOyn6.Lbe-cabti/'linear Ft.)). The center tray in the test deck represented a single layer filled tray (i.e. 26 4/C 416 AWG (624 Lbe cibkfhaatr Fto; TSo rijh'I'tray in the test deck repreaseihk'i4 empcy tray'(Le. ni'cables):
The results are'shavikltm Figuri-2 " "
The only theitnot~IIlea to 'exceetf -tSe acceptance'riteria weri the Me h iniide tfte:empty tray. This:
occurred 56"mm~ into"the-test.
-The ambfcnt'ern
'at th'e ~;of'thb tist was 83'6 TMs di a mIiiXitniim'avetago'fempiratiire'of33$'F at 60 mIautes (ambient tcmyeriiture plus 250'F allowabler tcmpcntture~J Bypkltthtgthe weight of+M1Jcabliet tray system-'(f.e the weig&ofthe'tray and cables attiL' r
I r
~
I
., ~
~I
~
O yr S
O S
e I
The followiag equation is based on a "best fit curve approach with a logarithmi relationship (i.er result of linerregrcssioa, method ofleast squitres).
Final Temp.'87&'- 86.752 Log'(Weight)
~ Final. Tenqa
': Degrees Fahrenheit" Weight -'ounds/Foot of cable tray aaC cables r
- This equation is'valid far eighteen-inch cable trays protected with the TVAdciigned Thermo-Lag 330-1 ERFBS having-cable.fiHi ranging from 624 Lbs/ftup thru 69.36 Lbs/k.
~ I Further review waa. performed'on the. results, of the single layer fiHed cable tray (624 Lbs/ftofcable) and the empty cabR'triy (0.0 Lbs/ft ofcables).
This was determined necessary since the cf5xts ofadding cables over thc first layer becotnes less importaat due to the cable insulation skrwhg the hear transfer to the capper conductors. Conservatively, a plot was constructed of the temperanues fortbc empty cable tray (0,0 Lbs/ftof cable) and the single'laya cable tray (624 Lbs/ft of caMo).
The:resultiag linear'equatioa givctt below coaservanveiy predicts the system's thermal response at law cable'Qih (he. les! than 624'f4'/ft ofcable).
I
~
Find Z'emp ~ 385;10'- 97756 ~ {Weight)
~
~
~ I
~ ~
I
~,
~'
a '
I
~ ~
0 0
I'I s Cr Cs 3 v C S.
J OhOOoo S
~ 'e Ca 1 cul ation PTN-BFJM-96;028 Revision 0
Attachment 1
Page 8 of 23 where:
'Final Temy...
Degrees Fahrenheit'"
Weight ~" -
Pounds/foot, of cable tray and cable Solving this linear equation in the range ofacceptable temperatures indicates that a cable tray system.with a weight of 5.33 Lbs/Ft would maintain acceptable rem at 60 minutes. Subtracting the weight of the Ie tray (4.00 Lbs) from ihksystem yields a cable l
ing of 133 Lbs/Ft.. Based on the cables used in the test(4/C 816 AWG>>024 Lbs/Ft) a minimum 6 cables are Iteeded to produce acceptable temperatures (i.e. b,T
<2 0;F).
The gext fourtire~ ofPhase tt'were dedicated ta-'ultgle condug'~guraaoas,"
Each
'assembly'ons~
of aomhsiL544fch'Tlienitd-'Lragj330-1'idi'he Peneric TVA 0nBe'.'-VVX,.
Vest 6:l.l0'ern the'erfankiiii'""of'twi encl area".'hhr type-;ciidosuie.is typrceally"usxf.
w
'conduiti are kcatcrlM a c'orner,'ic'alorig"w'hats the c)iling and wall meet. Thc coaduits ark <<nclescd on twp sides by coricrato irtdtwo sideiby'7heimo-Laj:
330-panels attachert to"a lrnistrutrtrame.'iMe 4"'he average aad alaximum'tern'peraatrc'ase of'
, ~,n'
~ vr',
~
'V Test 6.1.7 also contained a 3-inch conduit with onethalf protcctecLby Thermo-Lag 330-1 and the remaning halfprotected by 3M M20-A. The different firefarrier materials were butted together and a layer of3IVt M20-Awas wrapped around the joint. The test demonstrated compatibility between the materials with no failureLThe thermal performance during tho t t was representative ofeach'unique materiaL TV Tests 6. LS and'6;19 demonstrated'that bilge cable tray Ifittings (doublmross), vertical stack oftHW 1g-'nch) trays, and two side by siCk 18.inch trays can succCssfuUy be ptotkcteckin a single Therm~330-l encgure upgraded with thc TVAsystem.-These'tests:.
wereI all performed with nirinstaHed cables; TVATe'st-6.1.$ 'also demonstrated that'a cabli tray with a raised.
cover can bc protected with Thermo'-Lag MO-1:."Thar test Simulated a "randota Glled'cabfe'Tray (ieb, the'ablg are all located in the'center ofthe:triy with'the'eist ofthc bundle above the side raigr Onk-inch a'nd'fvk-tach'ir drop
.Cabla.Sandie's owere also dern asrrated acceptable in this test,
~I the asscnl biles.
Tlris test was perfortned horizontal position.
TaMe 4 Two Sieicat Eadocarca Sriadarat Coadoit Coafiaaradoaa Coehcsreeess' hvmye 'fasyaaare I&
Seaeai ass eeeea Tss I
$eeea eeeceeee seeaa aces ushers Fraae Oaesseeea Tmeerasen aies FFI
- Lares seas ssCa Seas s'L Coeaaoeo Laras aos I/siesso F~
Table 5 Tiraa SiatcaS EacSocarca Cos/iocsase
~
hsseaos Tmyease alas rFS stoa
" "'.IJ T
.I Sax~
- -" I)a-.
"- t --"o-i1a "
"Tine 1'veksees Coooic
\\ ~
rr I
So,r a
~
P Iis' i
,3?a
- ~,
TVA Test Kl".Ik-dhaoastraRd'the plarformance of:
"four sideYenclosurcs>>
- This-type-.enclosure ls.
typically used wltese coaduits are free:standmg..Thk conduits 'ar6 enciosedea all four,sides by Thermo-Lag.
330-1 aaached'o the. cottduita; The-. CoaBguratioas-TVA Test 6.1.11 demonstrated thc performance of "three sided enclosures".
This type enclosure is typically used where conduits are located agrunst a wall or ceiling. The conduits are enclosed oa one side by concrete and three sides by Thermo-Lag 330-1 aaached to thc conduits, Tiblo 5 shows thc average aad maxhnum temperatures ofthc assemblies.
This test was performed in the vertical position. This test also
. contained a large junctioabosr(60-inch x 36-inch x 24-inch) protected vtith the Gihcric TVAUpgrade. Me success ofthis junction box itesiga eliminated thc need for-a'secondlayer of nominal 3/S-iadr.:thermo-Lag 330-1 as previously deterauned la TVATest 6.1.5;
~
~
~
e v
~
0 0
~
~
~
~
~
~
~
~
P I,
~
~
~
'I
~
. ~
~
~
~
~ I
~
~
I tI
~
~
~
~
~
I a
~
~
~
~
I
~
~
~
~
I ~
~
~
~ L
~
~
~ I
~
~
~
~ I
~
~
~
~
'4
~
~ I
~
~
I ~
~
~
~
~
~
~ I
~ I
~ al ~
~
~
~
~
~
~
I
~
~
~ t
~
I
~ '
I ~
~
~
~
I I
~
~
~
~
~
.'~
~
I
~
~
~
I. ~ ~
~
P
~
PL
~
I
~
~
~
~
I
~
~
I
~
~
~
~
~
~
~
~ P
~
I
~
~ I
~ I
~
~
~
~ I
~
~
I I
~
~
I
~
L
~
~
4 I
~
~
~
~
~
~
~
'II
~ II~
~
I
~
~
~ ~
~
~
~
I
~
~
~
~
~
~
I
~
~
~
~
~
~
~
~
~
~
~ 'I
~
' I'
~ I I ~
~
~ I
~
~
~'
I
~
~
~
~ I I
I II
~ I ~I I
'.L
~
~
I~
~
~
~
~
~
~
I
~
~
~
I ~
P ~
~
~
~
~
~
~ ~
~ ~
~
I.
~
'I~
~
~ I.
II'
~
~
~
I
~
~
~ I
~ I ~ ~
I ~
I ~
~
P
~ I ~
~
~
~
~ I
~
~
I
~
~
~
~ ~
I '
P
~I
~
~
~
~
~ ~
~ ~
~
~II
~
I
~
I
~
~
I
~
~
I
~ I~
~ I II '
~
~
~
~
~
p
~
~
~
I
~
~
~ ~
~
~
~
~
~
~
~
~
~
~
~ I P
~
~ ~
~ I
~
~
~
~
~
I ~
I
~ I
~
~
I
~
~
II~
I II
~I
~
~ ~ ~
~
~
~
~
~
~
II '
~
~
I I I ~
I
~
'I
~ I~ I
~
~
~
> ~
~ ~
~ ~
~
~
~
~
~
~
~ ~
~ I
~
~
~
~
~
I ~
~ '
~
~
' I
~ ~
~
~
~
~
I I
~
~
~ I
~
~
I I
~
I
~
I
~
~
~
~ ~
~
~
~ I
~
~
~
~
I
~
~
~
~ ~
~. ~
~
~
~
~ I
0 0
0
a0
~
~
J
<<1e av 4v 'S.
44044 I ~,
~
~
- Calculation PTH-8FJM-96-028 Revision 0,-
Attachment 1
Page l0 of 23 t
using nominal l-i/4-Inch Ther)no-Lag 330-I. "Worst case" was detinod m utstalling configurations with die lease desirable auributes.
This included post-buttering and stainless steel banding 12-inches onwenter.
Thc second step was defined as "R'einforoemcnt'.- ln %is step, areas of known weakness in thc worst case'design were reinforced with external stainless steel stress skin and Thermo-Lag 770-1 trowel grade materiaL The third step in the installation process was the "Upgrade".
fn this step, a ncw material Thettno-Lag 770--1 Mats",
manufactured by TSi, would be used.
The The'rmo-Lag 770-1 mats arc a fiexible, nominal 3/g-inch sheet material that is installed over the exisung worst case Thermo-Lag 330-1 installations.
The number oflayers ofInat applied to each raceway is given in Table 8. All layers werc, pre-buttered with Thermo-Lag 770-1 trowel grade and assembled using nominal 1-inch staples and 16 gage annealed stainless steel tie wire.
DiKerent assembly techniques were used on the cable trays i,e. multiple pieces vs. continuous wrapping. The final step of the installation was a skim coating of Thcjmo-Lag 770-1 trowel grade.
Tabti S. -:
~ ~
4gcra oC,TTLLovce bcocltee l:"c,,,a a.,
~ ~
tYAfa<< '
4Cc)clat)O Iear~
h>> 4Ma<<h>> 4 Mea.
ta<<aa Ahern')
'acy. Nae rS)
<<)SSS.
aa)9)
~
I) cay I45 A>>
, IS)aeaev 2I'aZ,
)5 h>>
Ia) MvaL l)ah>>
Ia) laaa l4Ih>>
InM IS) h>>
Ial Mea, IS I h>>
I SS Maa, Tcatc 10 Rca!tta orTVATc¹+2 Ceal!cava<<aa,, '>> 4 Mae A>>a Maea T~ Neers) t~ ciaerf)
,. <<110 I..
~. aa)DX!~
l1 ~..
v,,
lee A>>'
IS)A>>
l$4 Meea 4
- )OI Mass Prior to Ihe start ofthe second test it was decided to run d!e test until the first assembly was close to its allowabl'e maximi)rn average temperature rise of250'F (l21'C). 7YA.Test 62.2 was run for 4-hours and 10-minutes. The assemblies easily passed the hose stream test. Test results are listed in Table lO TAQa 0 Reseal!4 orTVATcs~ "
I2 aae<< Iaea<<4.). I iiaC can rrt ', 4.21
', 4.).l
~
421
~.y
~ 1
.'. ~ ~
11'a ll a40' 4J.!"
psactaa loa
~.
a) '
. Il~eeehsk';
4111,,
~
'v.q I
~
a 1'aaael'caaafaia '" '22 1"
~,
~
, I)S A>>
(lel MSL
'It)A>>
ISS Mea.
)'aaeal eea4<<S..;..i!,I)4A>>
. --."...2)4 Aw.
',<<OMa<<
'.))SM>>
, 1'wel eaec<>,!
)44h>>
14l Maaa I.
)14 Maac: '
vvvvtv
)'<<<<SvaeaJaea,,.....
I44A>>
. I.;I%eh>>
144~
a
')4) Mia
)
aaeel eeaaMI
...421 '2 1'I'w~, ':.
411 1 ah'iaaf I)4A>>
')l Mec,'
v leS A>>
lee Mac'he composite
'?70't'ce& 330-'1'deiIg'ii out jerfoase'4'rab~'cq)cctlioe.
'TVA Test 52.l was iud 'S~ and'~iiiutts.'t tha1 paine the test was stopp'cd sinco no one ltad prepared to run loiigcr'han thc'equired three-hours.
The assemb Hes had a large thermal margin remaining,and.
easily passed the hose streitam ppt.. Test'results are l~ in Table 9.
l I
P
~
Athird test assembly had been prepared using only the new Theimo-Lag'75 l materials"tri conastruct the ERFBS.
This test assctnbly served as a Backup" in the event thc previous Phase iEtcst asscmblics did not perform satisfactorily.
Based oa the successes ofthe previous hvo.tests,~assctnb~vgs not used.
10
~ Q Y.
v
~
~
~ v
45 Ci
v
~ ~
7Q 4
I
'H c<oocs
~
Calculation PTH-BFJM-96-028 Revision 0
Attachmegt 1
~
Page ll of. 23 TVAELECTRICALTEST RESULTS Hav of TN.
ng witnessed a po were peffoflned ion ofthe TU work, a variety TVA's labs in Chattanooga, Standard Condula C'onfIguratlons I
DorIng the early TU am pacity tests. which were based on graft I I ofP-848, it was noted that rhe'use'of three conrIuctors connected in series resulted in excessive inductive coupling bctwcen the cables and the conduit.
The resultant conduit hiLiting and surpresscd currents in their early tests led Ttr lo add'a fourth conductor to minpnize the conduit'rIffecrs.
These findings, later supIIorted by TVA testing led thc P-848 working group to issue drift l2, IArhich stipulated the usc ofan everI number ofconductors.
,Table Il
~
.TSl IVreppcd Coodeit - TVATutf Ja fit'-.
sn
}s
~ ACP I ~ iVo 4'CFR Ale 1Ori:I.Oi iO)
~
'.,I j
i04 " r 'ol'tT While the correlation between tests conducted'using the
'ifferent conductor and power supply artangcments was good, the differenc between thc two baseline conduits was greater than is desirable for test repeatability and seemed to indicate that some conduit specific effects were still being seen.
Whig the bulk ofthe d
'Is and findings ofthat early TVA effort werc re in our earlier paper, there wcrcI several key ations in that program which may[now bc updatecL gn the Qrst progrant, because TVA had assumed thd the clror'ce of test conduits wouitrI make litt!e diffcreIice in thc detertnination ofthe finalj correction "factor;( no effort'made to "match" condliit eats ie. sc((ect all from the same vendor).
A coasistent 3A%'iffc'pp factoI was scen for the[
resu!Is are sho'wtf ih'Tg connected sihgle p6ii: I baseline conduitt-ee respectively.
TbaI.
(whep only"a'singfi also iven in Table I'I;"
co in the tesuhant correction two baseline conduits.
The O':Il frtr'ihe'8-'9(c'cables'ulations'or thc diffcrttt abclled ACFI and
- ACF2, of the I-inch conduit test llhe condlllt'w'as used) are c span of ACF8'mcasrrrek's performed'using chtssical lUgkfjthe corlectiotl'factors' baicrfon-TSI an'd'-K TVA close y wer~w noted 'thac though,th mirrored'an McGrath mcrhodo ell above that ex segm
(
in orIferto confirm that a'peatable results werc indeed:
being obtained, a second <'-inch baseline"conduit was prepared and tested with['a variety ofcable'and power=
supp y arrangemeati, as rfollows:-' t/c aria corinecied'powered iriigfc'phese:" -.
4-t/c saia coo~4 ypeeed iiiisre~.;
. ~...'
8-3/o series coeqted pmraed siaaie ph+
8 3/c powered direc piief4r I'i Given thc confinnauon provided by the variety of configurnions evaluated, TVAconcluded that ACFs at or near unity were indeed possible. TVAbclicves that a significant contributor to this improvement was the elimination of thc annular air space between th~
conduit outer surface and the inner surface of the I Thermo-Lag. This was accomplished by thc complete i
prebuttering ofthi intaiorofthe'preformed'sccnons of Thetmo-.~~or: ta.placing. them over the conduit.
Having i:i[mutated the air gap ai describ'cd above,'the'ddftiorial thermal resistince from installation of the" Thermo-Lag.330-1 appears to have been.ofQct by thc.
siggifrcant it~ace ia die..overall.surhcc area and.by.
the inaeased surface emissivity of thc ERFBS compared to bare conduit.
Duriitg-'a literatu're 'mirchv tVwas-observers that the:
recommended value forcceduit emhsivity haschangcd
- signiflcintlyova the yeat3. Whai the Neher-McGrarh work~was performccVin l957 a-value of 0.9$ was utilized:
In':-I962; when':ICEA P~26 was generated, a value of0.82,was appiioL kore recently, IREE 835'as utilized a value of-OW. Discussions with those fimiiiar-with the conduit"m'anufacturing.
proces's reveals'hat:the chinge waa driven by improvements'in'he-hot~
ppxcsa and by a trend-.
towards the use'of electro-galvanized Intermediate Metal Conduit (IMC)and Electrical. Metallic Tubing.
(EMT) in lieu ofhot~ rfghfsteel.conduits.. Thus itappears that the.difference observed between ACFI
~ ~ '
v
~
I I
I
~
0 41 0
acssa I aY:
o-s u-oo s
s 3: a 'sr 4 GtaKQ Ktlgc.
3 aves.
~
3 a3u 240aC3 I
~
Calculation PTN-BFJH-96-028 Revision 0
Attachment 1
Page 12 of 23 and ACF2 was a %action ofvariations in the surface emissivities of the, segments used in making the baseline measurementi.
emijsivitfes) can be c for ]teel were frequ 835)
For aluminum, WhQe the low emissi corri:ction factors whee do sums to warrant basclampacities.
Since specify an acceptable 848 has been revised to be taken to facilitate in f'ronI differen tests: "'
~
scca.
Second, the readings below that assumed in lEEE the difference is even greater.
itttas will result in favorable testing an ERFBS, the results somp consideration wlicnchoosing Ihanufacturing stindards'do not range of conduit emissiyity; P-require that emissivity readings
)formed compa'riso'n ofresults obi0 l1
~~:
~ " ~ '
Kmisavtucs:,,
7 Coadtljt During the course ofseveral test programs. TVA has measured the surface I',mlssivities of thirty-two rigid steel and five aluminum conduits. Allmeasurements werp made with a Mikron model M80AL-2FH hand hei) gun. This device has a target zone of0.2$-inchres at g distance of 2.0-ijlches.
The results of those measurements arc shown in Table l2.
Several observations are in order.
First, thlt cmissivity.'of the rigiIIsteel conduits varills over a wide rangr. l!may be.
thatl this disparity wasI the result of. the age of the various specimens or( perhaps reflective of their respective storage conditions, ln either
- case, the impprtance of "matcHing" conduit specimens (ie draping from the e vendor stock or verjfying The results are given in Table 13 Tottic D Three Hoor Coodilit Upgrade Syslcla' IalaaraeeaO>>
ne I louse her oo Tabie l4 Scloctlee or Haai ACF-Coa4at t SIP I.tease '
O.O0 I'.n llrtel e IIter>>r
~
on al tli0.~ Iboar..r v...',,ay...a.i..
an I
I sheet....
aof I
an
~l Hoa-Stan datdrCo adult.CoaOgaratlons.
I
- Consistent with the P-848, the worst'case ACF was.identified for one aad four inch conduits for each thickness of ERFBS.
Those
'actors were furih'cr reduced'by f/o to account for p'ossible vanations in the 'surface 'emissivities of installed conduiti.'iTTie"fhal 'values, which were included in TletA's interttal design standard, are shown in Table 14.
ert I lees~
riaarei>>-
. isro, I
ae eeeeI.
e O~
rs IL
.I enact i.
44t r
10-'i " 'gael f. -
400 l4 LS eiial '84
~ 0 4 v ji.Z'e0aoaw, O.N
" - %Re subicqucnt test pmgfaii-a!
dcd'2-iieet 44bch'caiId5 ~~4'with
'0edt" "sy~ '&> phrviouatf "llo48 leering fQtc8'hfcotnplhaw
-14 of
':wltich'includcdwlhof the Ifhini'that'U.and earlier TVAefforts 0
inclu TSP '-:thieve w~~uubb~
lessIIns learned raeaItbntt'-specimen.
configuration. aad'.connection TVAI
" -'"A: Sas) '"!"OS'a4inaaaa t.'m'.
The. TVANPL'rogtaia also inoluded'assemblies to evaluate muMple conduit! encased'withh a common Thermo-Lag ERFBS..Two types of such.enclosures may. be utiUzcd.
rn the, fortner, the Qat Thermo-Lag 330-l.panels. are mounted directly on the surface ofthe conduits, with aaintaatlonal pciiphetal air gap. This construction willbe used oe all three-aad four-sided boxes (thc additional side ofthc three-sided box being provided by a concrete wall or ceiling). The second enclosure type consists ofThermo-Lag 330-1 panels aaachcd to a Unistrut fhune installed itouad (but aot touching),gabe conduita,, This augaucaoa win bc used where the conduits to be prptcctqi ere routed such that only two sides ofthe box wiltbe Thermo-Lag 330-l, the other, sider being concrete walla or a wall and the ceiling. The. substantial heat,rapacity nfthe concrete and rebar, coupled with its low thermal resistivity s
I I
a s
'sa sl
~
M s
0 0
Ca 1 cul ation PTN-BFJM-96-028
.....Revision 0
- Attachment 1
Page 13 of 23 compared to Thermo.Lag 330-I, w>li ensure that such
'two-,and three-sided enclosures are not thc limiting cases and may beseprasented during testing through the)use of four.-sided assemblies.
I I
There is no specific guidance for the performance of such tests in P-848. Thus, TVAfirst identified the kcy parameters regarding such arrangcmcnts and'selected theivariables in a manner as to ensure conservative results.
Those parameters are thc selection ofconduit
.size; num&sr of coadiihs (and their anangement into rows and/or columns), conduit~lag (whero multiple conduits are used) and. box size (in the case of two-sid@ boxes on Unistrut &ames).
Those'ssues were addressed by, TVA's selection in the followingmanner, I
&cfosurcr, Using drc. Conduir Jor Supporr - When encfosurcs are constructed ofpanels mounted directly to the conduit, the air gaps (between adjacent conduits) ten< to vary in accordance with the <<nciosed conduit size,'Wth small conduits, tho conespondingly smak air gyps result la tho heat uaasfcr across the gap being a fbnctioa oftadiatioa aad coadactioa only, rather thea a
I'ncaa ofradiatiao,ccevcctiaa iad~xhctsNL Th tho ACFderivedusiag "sma11"- c'oaduits would be '
- e. ln kcepia'g with the P-848'philosophy; TV/chose to utilize.1 inch Geduits since they aro the smaQest size ia which power circuits are typically routed.
For such enclosures,'surface area ii low'est for low numbers of small conduits: A single con'duit ~ould.
provide.for the lowest suiYace arka bht'oufd'nos-inclu'do any intetnal air gip-sinor TVA's m'eth'o'd of appiicatioa is to utilize preformed sections'pre-butfered' withItrowel grado'ater'iaL" WMle a::twcr conduk-encaaemcntwouid result in aa caclosaf air gap; each '-
condliit'as an adjs~end wa0" Gem'which to" radiate.
Thus, a eel ofee coatkkl"ia'a row wai' qg judgjk~jxinrlM~RaAMKci'%84.whilestill-includmg a coaduit4hfc59&it idjaccnl'to'one oftNe:.
'acldsure ead'aBhj'Sich'atr arrangcm@t'using.h-inch conduits,'wiihC5i&ithe OKtestatg;. TVA'-
recogntzed that ~5QQSt'ae baakad m mulaplo."
mabel atr addltioaal
'alt"gap Ia taloctod bettveea tho'.
2,a3 raw' 'babas; ht onhrta asaaaa thatstlsa'ri'asaembly
'aa htchdat 4hkh'hittlgdoubQ~5f throiHI'-ihcM coathhtL While'this arraagctaent leaves each r'owwith '-
adjaccat fb a radladng:surhice, it wis. fcl!'that'ho likelihood'of'encounterin" thror rows of ERFBS 1
'I yl~
protected power conduits (which could noc bo individually wrapped) within a typical generating station was-low; Thc presence'of multiple Ealduits in ciosi proximiyto ono.another (even without a barrier) results in mutual heating ance tea inaoductioa" of'hat Nch'cr an't McGrath described as an "interfereace tempcraatro rise"., This eO'cct is typica0y accounted for through the uso ofa grouping factor, such as is given in PA&426.
The tendency ofthe effect to-'dominate the correction factor when maay conduita~ wrapped'ogether, further contributed to TVA'i decision to uso small conduit banks (lx3:aaC2x3).
A spacing between coaduita ofon&talftheir aotaiaal diameter was utilized: Lesser spacings would result in a
greater interferenco tetnperature riso effect, as dcscabcd above (and'hus appear to minimize the effect of the barrier) and are not feasible below oao-fourth ofthe nominal diameter bccausc ofthe physical interfcreaco ofcoupHngs aad supports and as a reek of the need for tool clearance Larger spacings wouM generally support individual wrapping. ofthe conduits.
&closur'es on a Unltrrut Fraimt -'Ia contrast to the boxes formed by Thermo-Lag 330-l panels in direct contact with multiple coaduits (which were necessary to assess tho resultant dead'air spaces'bctweea conduits) enclosures constructed ofpanels mounted oa a UnJstrtN ihaao poteadally result in: a-" lirge gap between ihe'conduit swhcc aad: the inner wallofthe.
paneh. Fice air exchange willexist between that larger'.
space aad thtr'smaH conduit~duit gaps, rendering the latter insignificant Tests'te asiesa" the effect ofthe large'ap aad'tho. Thermo-Lag 330-1 ERFBS were conducted using a single conduit;-.thus avoiding tho concern for"interfereaco temperature risc" mentioned above.'-
Those tesat were coaducted 'itsing-I-inch'oaduits, givctrthat its thermal resistance ta the surrounding air is higher ihart that of a 4-inch (duo to the smaller sur&cc area ofthe foitacr); A'itlted cariicr, this is the smallest size: conduit--ia which@neet: circuit. are typ re~
For a given size caaduit; the tnhthnum box size. (and therefore thc minimum air gap thickness) is established by the. diameter of the. conduit; thoth~ of the 1
~
~
~
~
.ap t
's
.1 'q. '
< ~
0 0
0
Unistruc and the gap between the conduit and the Unistcut. For a single 1-inch conduit with Thermo-Lag mounted over 1.CLS" P1000 Unistcut.
the smallesc Possible box wouldbe approximate.ly 4.75-inch x 4.75-inch (as measured a
. thc Unistrut). tn contrast. thc larjest box (regacdl of conduit size or numbers of conduits) which TVA ticipated was be approximately 30~inch x 30-inch, TIIough ic was believed that the
.lee box would factl (and therefore the to ravidc thc needed Th) final ACF for box wapld thea be mast co It tc better convective cooling
~ ACF), both-boxes were built onficmation.
canduits on a Unistcut frame nsccvative ofthe two determined adjusnnent for the appropriate t with 1EEB P.-848, bodr in length.
above tests,a cl ing factor C
tsten mblies.were 12 ee)
Th results of the non-standard conduit testing are I
given in Table 1S.
I I
kaMe ts Cotrcctloe Foots Noa4taadard Coodotco
~ t, SOU Aisec lac. aa.l 4aaiia
',4rasuaenc bee.tet t Ofiecc~e1Ll e I OIIIccexwal o I8 I~ coeLw's 0.'N 4~
QCS 0.01 The ccsttita are gi IO ~
tandarO Toy C in Table.15.
Qgttradona Con tructabtTity aB.
t witLt.tho.~
walkdo that not s
acaci005'
~ ~i incluIsioa.of~~-
o-Lag'33&1 or, Tho same. coYMwf at 1VBN had determined
'bo-'individually wrapped gg~aipy outlined In,P-had showi that elccirical 1-top cover. undemeathc the wrapping of solid bottom shown that the maunding
~
'ta dard Three-Hour ray The test pcognun at OPL,also included trays. wrapped" I
withja composite Thermj;Lag 330-1/770-1. three-hour upgtIsdc system as was IprcsIiously. described.
Those tests'/erc configured ictqmpUanco widtDry 14 ofP-Ca 1 culati on PTN-8FJM-96-028'evision 0
Attachment 1
Page 14 of 23 ofcables above the sufecatls could sometunes preclude the encasement,of individual trays (necessitating the enclosure of several trays in vertical stack).
ln some cases, the mounding could be addressed by raising the siderails, though the effect ofincreasing the thickness ofthe air space above the cable mass would have to be evaluacccL Finally, inadequate space between adjacenr trays could necessiate their common enclosure.
As with thc non-standard conduit configuradons,'here is na specific guidance for thc performance of such tests in. P-848.'hus, TVA identified thc major varianans and evaluated each ta detenninc ifadditional tests werc required.,
For thbse configurations which, requited additional testmg, itwas noted that only minor excension of the P-848 logic was necessary.
Those issues were addressed.
by TVA's selccdan in the followingnuumcr, with the results given in Table 16; S/dd-by Side-Upon review, it was determined that the results ofthe TU tray casts cotdd bc usccL to represent thc common cnclosuce oftrays which are horixontally adjacent (ie cun side-'by-side).
This acrangemeat is consistent. with the.Stolpo modeL (an. which tray ampacities given.in ICEA P-$4-440 are derived).which considers that heat is dissipated out the top.and bottomc surfaces.only. "
~ -..:
Solid:Bottom frccyr I-Zsough;the TU: testa. were-perfocmcd on ladder type~ i@was determined that thotfe. roatlts'ould'also 4e~ to represent solid boaotn. trays.
This extension is conservative in that tttto solid bottom uaysdo not have an airgap bcovecn.
thc bottoim ofthc cable mass and the ERFBS barrier as.
a result ofthe presence ofthe tray rungs.:.
Coviemt1'rayr <<The test pcogcacn at OPL addressed the enclosure ofladder type trays over which a sheet steel cover was applied pria'rto tho'applicadoa ofany barrier materiaL During application ofthe ERFBS, ithad bcea TVA's objecdve to" ensure that no alrgap existed between the upper surface of the cover and the underside of the barrie.
This vras accomplished by applying trowel grado Thermo-Lag 330-1 to tho top of thc cover prior to ittstalling the panoLL Unfoctuttatcly, when thc masan was applied, thc dpi sheet stccl cover sagged and moctcmasdo was ccquhed ta.obtain.a level, surface sa that no~ wouldaxlst. Thls yraccss was.
rcpeated
= undl..tho'umtdadve.
thickness;. greatly exceeded tho nominaL 5/8-inch; Thus, cho <<elm depth which accumulated'during its installation resulted in a iw
~ s 14
0
a final multiplierof0.60, well below that which had been predicted.
Caariyvasaee Soleil Ooaaaa ~ Iboar Sba)a waI ooo ave I boar 049(sew I)
Vcr)icalStack-Thctests also included a vertical stack consisting ofooe control and two power trays within a common Thett)N-Lag 330-1 ERF BS; with'th'e control tray on the bottom. Though a measured baseline was detgrmined, for this multiple tray arrangement, the actual ACF utilized by TVA'was based on thc morc co(tservative equilibrium cunent for a single tray.
Thg ACF developed for'thc stacked triple tray arrangement will also be used for cases where the common enclosure includes two power trays without the. control tray:
The 'applicatroit of this ACF is conservative since: the two'ray arrangement affords more direct heat dissipation'or the l'ower power tray (hau ttte'tested three tray enciosut'e'.
~.. '
Talc l6'orrccaoa fatrosi-TroyI Calcula6on PTH-8FJH=96-028 Revision' Attachment 1:
Page, 15. of.23 region's
'and modelled using conventional heat trutsfcr prir)ciples. Their or)alysis suggested corrcctio(t factors ranging from 0;7$ to 0.78 for onc-hour. Thcrtno-Lag 330-1 systems applied to tray.
These values were
'conservative compared
'with the.vendor's.
original published'ntimbeta (Tabio 1). However, data published
'ihce release of'their paper, appeared to indicate that th'e mo'del was non~onsetvative.
Thus, in the absence of confirmatory tests, TVA's eQorts tumed back to identification ofacceptable test methodologies.
" After completion of'the'TU,and TVA tray tests, the
'nowledge that'dditional. tray configurations were liki:lytt) be enc'ouantcred as thc fire barrier installation procesos pro'gressed at WBN,:lcd TVA to;revisit the Save-Engmana.analysis...
With a cable depth ef 2.95-
'inch, the 'ACF established by the TU tests can readily be compared to the 3.0-inch computation in the Save-Engmann paper.
The inital review, the results of which ate given in Table 17, indicated that there was a significant discrepancy. bctweca thc inodel and the TU results.
,. PrcHafaary Coaoparfooa ofhCFa YorOao) gaa a I)NMr 5a)ebs'de I boar SwaaO Iroo -1 boar 4)9
~a 0.4! (RW I)'
4$f=
" Cob)o Oa)ab (4)
~ Trav Oa9OI(4)
~., ACf.
S~oww.. '.',.'+,
)
O.ra a'ore.).
Those. values are by aukyati. no )pectN test.
pcrfonnocL RaLfafSickrai/s - During the course of'preparaiion for'he non-standard fray tcsthg, good conctation between the KU'results and a mathematicat model~as.note(L
-'"'ince thc model had'the abilityto address'the effects'of""
raiscct sideraiis, no special testing was'erfotme(L Results ofthe analysis performed are ghcn bclow.'"-
a
'l Savh-'En'gmina Modil-'
As TVAidentified tha pcef to.obtaia pew correction factors forwtmppefftlya)isutvcy h)d been conducted whirrhco'uQboueekc4c(s ofmolesting.
a aa I
A litcratute scatch revealed that such a model had bcctt pro+ised'y PhB-Save" and" Gary "Engmantt
...'ollowing'a'review ofthe other appDcable. litetaa)re for..:
f)re;it)rotected trays, Save and Engmann suggested that.
'he ttay/Qrc barrier systcta couldbe divided into seven 4
i a4 J
+0 '
';RC vAV'"ra Q T(r '
}.99 Note)
. 'r
'Aitokco 6eiIITobfe'1 oftt(o rebrer)ceo The value ofThermo-Lag 330-1 thermal conductivity used by, Sivc and Engmann may hive been the cause of much ofthfsvariance; When the paper was prepared; they. liad obtained a value".of thcimaf conductivity of 0.430'Waits/{meter-'C) from TSh'ubsequently,'TSl has refuied: their estimate oE this parameter 'a'nd have id(a)tificd 'that it"h)fs )trfairly st'rarig temperature" dependency:. 'ased od the Thcmfo-Lag'30-1 temperatuie.measured during the tcstitig at(OH-a correspo'ndihg conductiv)ty of0232 watts/(meter-'C) has been used'ia4e foHowingrc-analysis:.:-
The only other modiflcatioa to the Save-Engmana model necessary for,direct coinjolson to the TU data's the additioa ofi sunple"c(tuatlotr to 'evrahiate the cffectrofthe silicablanket=.Th~ee dlNiact refyona.
ofthe model were reducecf by Save aiid Engmann tr)a system ofhvo cquitfoiiswitttv'l(r)o%rS. Thi two a
I a
~
~
rI I p
ran a
a V
o
0
~
~
~
~ v~r ~
~
uur v ciliif~, a 4'v ca. ~
013052466231:a) 7 Calculation PTN-BFJM-96-028..
~-"
'= "-. Revision.Q.......
Attachment 1
Page 16 of 23
, th e mass and heat intensity, uation determines an cFcctive combined mass (cables and tern is solved by guessing d iterating until the system anger; and using a baseline ds to the actual baseline test inal value from P-54-440),
well with the.TU tests (see bk ta Save-Knarnaan <<uh TltTesta.
g Png asymmetry of the l
res ppctfvcfy. Thc ncw CIl conducdvity, km3, of c
ct}. As before, thk sys for the unknow/ an rgcs; With these bh ofs which cortespt)n nt (rather than a num odel correlates very 18).
bi+
valup conge valub' CufrC the'ab Ta 0m paT(eon or MortUI Cnrle Daonr iirr)
Trav Drorrrall O.0T (reae 1) equations, Noc. 5 and'12 from their paper, have been rcptoduccd below (defmiuons are given in Appendix t). I Thc system cao. be solved'teratively by first uc i
a value fee th two unknowns'a and s e
thc TV tests, the ACFsshown in Table l ofthe Save-Eiigmann paper ca'n be recalculated using the modified parameters.
For these standard configurations (ie no silica blanket) 'equation 2 is not required. The results arc shown in Table 19.
One diAiculty in establishing a w'orkable'model for protected raceways'has been the P-848 requirement that the bascTine'used'in dctctminatioa 'of the ACF be cmpIrically cstabHshed.
Such a requirement works a particular'hardship on the dcvc:lopment ofa consistent model foruse with cable tnys, given the conscrv:uistns inhcrint in the Stolpe modeP.
Stolpe assumes no air flow through the cable'mass.
and his original tests l
included thc use ofa plastic sheet across the bottom of the tray to block such flow"."Th'e usc"'of'a'measured baseline current as imposed by'-'848 results in a loss of rcpcatibility since the efficiency'fcable packing (and thus air flowduough the mass) willvary from tert to test Sav a00 I
4 0.4li (rro<<()
reer vr0llNialre l29$
4'
~
~ 0
~
o(c t.
Vati>>a us)nod>> actnatZU.base(tnc heat.
Ir g.shown thitthcmpfel favorably. co(itpares with ngmann.equation((5)'.;
As an example of that conservatistn (and thus the potential variability induced by the use ofa mecLsurod baseline current), the normalized current in the TVh single open top tray test was approximately 11%0 above.
the. adjusted. 1CEA cunent:fl)r: the same.diameter and depth 4
~
I
~
1 iianf
'v((1)rr-
)-1'c)4, +sbviob((5)ir-;
) -1Xt })"'
(.1)
Ncw I
uation use@to evaluate the effect oftho silica blanket.
~I
~
II 16
~
W I
~ '
~ ~ !
VDI
0 0
vaI ~ I V ~
s~-~v o i d:
~ vnrM s
UVKQ angl.
5 SVCS.
813052488231:F18 Save/Enynann cctuation (12)
Ca 1 cul ation PTN-SFJM-96-028 Revision 0
Attachment 1
Page 11 of 23
'-(t-
)-
a(r -rc-1.911 s' (t o) td(t-O) i
)
J~ r~h~rr ~
W 4
d (
~ ~2 TKIF ltd q kg d t -cern-OAQ ae IM 3
(3)'...;...
~'
Table l9 Seve/Eaauseaa Theres~ ACFO
., (+ ). 'ear Sa~aySIdaet a JLS'd
~5weCseaaee" 12$'d 34odr~
Ma&A(IAs&ale MddCnde Aaaheis S~+
0.1$
1 0.1$
0.41 a01 0.10 ass I
O.s1 ase 0s1 5
0.11 1.5 a10 a14 041 0.10 0.11 0.11 0.11 2
1 0.14 4
ate
~
~ >
5 a10-"
0.10
$ 0 4
on
~ an
..... a00..... ".'..9u.........j 0.11 asl on OAF oc
. a$ 4 as4 05$
0.55 0.$ $
Note 1.
These foctots ere based on the nombril 0.30'lthica'ofhw'baricr and orisiaal Ttxnn~a conducdvtty.
Nose +,. Theec Ssclors ~ conscrvsclvcgbescdonlbamaahaen
%chiesa(0625 LofQw 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> berrlccen4 tba revised Thcrrno4aa conductklflty.
Note 3.
Thew factors enrbescdunttw&ainsH4ourdcsiasu-.
~
'1 Note 4.
These tbctore are based'onew'nominal thickncse oMe 34owupgredesynenc, Notch:,-:AlbcncskebescdonnomhaHCRh;vstucs otbcet bucnslsyds:...r.:.".;= --.:~-.:-:-:--."--".--'-~-=-.-. ~,
Tltc ~eQt ofsnc)~odd is that cable sizing fora- -
- only simple modification ofthe input parameters.
d&ld I, Idd-,N
. d time and expensi'of-testing per P-84$
once.i Inthecascofanupgradesystem,thethicitncssofthe con aroty testa liive beeii 'p'eBoimccC
"'Such -'
EK:BS;ib'and sr;inust b'e adjusted," L'Qcewise, where conifg'uratient$ $2ight hclude various upgrade- -- ---raised sidcraiis exist, the paiametcr dmust be adjusted.
traZL~gggsei.'Qyqno-~330-I side rsHs Table 19 showa that such extensions typicallyresult ln (to c car undcriying cables which axtcnd above then an additional dcrating of Bo or less.
raUs) and trays with"removable'sficct metal top or.-; "-- -'
"i botto covers under the ERFBS."Each castFrcquircs --*
- Application.-of-a..tap cover 'or '.battotn..respires h 0 4t V
0
~f 0
I I
~
~ '
~ ~
pe Ana)ysis (TG'A) and I resp pnsc to the N
information". Thc full frared (IR) Spectroscopy'n s
request for., additional, scale fire test v/illattempt to minor, difference. =. in the rent vintagcs of Thermo-Lag effect the perfonnance ofthe e
demI)nstrate tliat the composition of the diQ'30-tf willnot adverse]
S.
ERI'8 adjusunent ofthe parameter err. though in the case of.
the TVAtesting; the sagging ofthe top cover fumhcr rcq 'rcd cho odjustmens forbmricr thickhcss.
e FUTURE WOE At t is writing TV% is plannmg onc more fire.'test in late )995. This testis necessitated by diffcrcncesdound<
'n cika 1985 Thermo-f.ag 330-1 installed at SQN and the @erma-Lag 330-1 materials used in thc testing and
-installations at %BN BFN,. The anomaly was'iscovered while reforming Thermogravimetric Cat'cul idion PTH-SFJH-96-028 Revision 0
Attachment 1
Page 18 of 23 design ofThermo-Lag ERFBS.
I
SUMMARY
The 'HA tes! program has demonstrated Thermo-Lag materials, when properly designed, tested and installed are'c'apablc of providing rated one and three-hour ERFBS.
TVA ampacity test program and results'<<have been
eviewed"itang with a rationale for tests af multiple raceways. within.a.comtlton enclosure.
Recent ampaci(y tests conducted an,cables in tray by TU have
" been'a'nafyicd and compared withan existing thermal'model.
Modifications.to that model have'een proposed which. provide good carrchtion with the TU
"'results.""
ACKNOWLEDGEMENTS cuhtion methodoloIDp based lectrical raccwaysl abtte) protected wiSP apcci An appro by the'Amcricru) stec] protect]a'j w
and electrictII en
'I
~
~
Aca of pcfffed sauclura]
pro tion an the mass and size including'he cable fil]
ercno-Lag materials should ach'imilar to the work Iron and Steel Institute for'uldbc, v'e'ty uscfu] to fire gineers involved in the The authors would like to acknowledge tho contributions made ta the Thermo-Lag project by the other TVAteam members, Thermal Science Inc., Texas Utilitiesand Omega Paine Lalsarataries.
Special thanks are extended to Gary Engmann ofB]ack and Veatch for his time and patience in the discussion ofhcs work...":
'PP NDIXI - SYMB
~
e ca LS'avet a
Engmann distance from cbc b'ocrori othe packed'cable mass'to i'pl~Nufacc rINs ls'a'ihe maximum'npctaturc'n tbe pack'cd cable mass s
cm cob cu I
ke sb c
Tc Tw km ksb kss d
TKAI TKw sb W
Ra Nu (meters)
~
~. hcatgcclctaticus le' cmissiricy ofcbc~
cmissiricy ofthe ~ide cmisciricyofibc outside
- cmis&kyofcbe iaswdc o
-'. thermal cooducciviiy ofalt.
Scc~
siuckiiesaof rbc '
cable temperature oftbe lelnperanuc ofcbc lbcimN
.ofde.
chasmal
'ot'chc
'bcrnud bfwrap toadiag'de'psbHC Cab]e
.cernperacurc ofcb~
4eemm lcmpcracum ofcbe chicknesi ofcbc file
~ ch)ckncs'a ofcbc tire
. loading.wkiib of cable Rayibigb atunbct(d mens Nussels numbct (duSIcasloai
~
~
oiume ofpacked qsbtc macs-()sects/merce')
~.:,,,...,,..-
~
~
cabtc mass surtbccs (dbncaskultcss)
OfibCWrari'maceriatcri'iheseiiOmOfrbetray(dintCnciaaieSS)" '-.'-..a ' ". S'i" ofrbc wrap nascsiai orf cb'e lap of lbc tray (dlmcas(onlcss) "..: "-:,":" 'v'.",'-
fcbemasaial on cop of sbe slay aide rslts - lray curst; wrap naccriak ccc. (dbrlcaslonlcss),
~.
'~tmrrct~C)3...,
p
-'. j,y4je-ti)'[w'aej~~ K')
i(miner}'t t%
a
~
~
(degroes C) cable mesa (degrees C) pecked cable mms Iwau/(~grec C)3 macci9fon sbe boccorn ofcbe tray Iwaa/(meccr4k'gicc C)]'namrfironcbe cop ofibc tray (wacs/(mc~cgrcc C)]:
"." ~
~ ~
. ~
'ruy'(mcccra)
(dclcca *...
~.
~"
~
~ <<
~
'ass (degrees,K) mscendal as sbc boaom ot'rbc riiy (merers) craacsctfat ac rfrcsopor ibc tray(mcccrs)
~ a
' '<<r"3:. e'.le r
'r.'ray (meters) I
~. "
~
~ ra,
~
t'oatcss)
~
ess) us,e ~
1 e
e s \\,
~ <<r e
~
w
~
der rent.
(
0 0
0
Calculation P7N-HAH-96-028 Revision 0
Attachment 1
Page 19,of 23 hcrwcc33 ihc iop fshe packed cd i ofihe iray side raiis Addltioaal Parasactcm kc cffccuvts 3er333aicotd33csiviq ofihesiiica cloih fwaal(33cicr-degree C)l zc u3ldc33eee ofshc nlica doth (333cicisS i,ml errccshttadheisiial oot3dttciivisy ofihe33ppcr potsiotsofihe packed cah(omahas consider(nit ihc <<(fees of the siiica doth (waa/(mcicrwsrcc c))
n ioiai th ckntsss ofihe cabie mass and silica cloih (o3cicis) r APPENDIX2-REFERENCED TESTS",
6.1.1 6.LQ I
6.1.3 ONE HOUR FIRE VESTS "Fire Endurance Test.of a Thermo-Lag.330-1 Fiie Protective Envelope (1-in. and 5-in. conduit configurations3 and 2-in. air drop coaflguratIoa)"; Omega Point Laboratories Project No, 11210-94S54c.
"Fire Endurance Test of'a Thermo-Lag 330-1 Fire Protective Envelope (1-in. and S>>ia, conduit configurations and 2-in. air drop configuration)", Omega Point Laboratories Project No. 11210-'94554m "Fire Endurance Test ofa Thermo-Lag 330-1 Fire Protective Envelope (3<<inee 2-in., 1-into and 4-in. conduit 19 e
r r
t.l
~,
~
~
W ~
~
configurations).", Omega Point Laboratories Project No. 11210-94943m 6.l.
"Fire Endurance Test ofa Thermo-Lag 330-1 Fire Protective Envelope (3-in. steel, 3-in. aluminum, and 1 1/2-in. steel configurations and generic 2-in. and 4-in. tube steel support members)", Omega Point Laboratories Project No, l l210-94943b.
6,1.
"Fire Endurance Test ofa Thermo-Lag 330-1, Fire ProtcctiveSnvciope (l-.in 2-In 3-in., and 5-.ia,.conduit and fivejunction boxes ofvarying sizes)",.Omega Joint Laboratones Project No, 11210-94943cL 6.1.
"Fire Endurance Test ofa Ihcrmo-Lag 330-1 Fire Protective Envelope (large junction box and three 4-in.
- -- conduit'sections), Omega Point Laboratories Project No. 1 1210-94943e.
6.1'.,
"Fire Endurance. Test of'a Thermo-Lag 330'-F Fire Pietcctive Envelope(Tltree 18-in. Cable Trays and a 3-in.
conduit)", Omega Point Ldaborato'ries Project No. 1 1960;9718$ ; ',.
6.1.
"Fire Endurance Test ofa Thermo-Lag 330-1 Fire Protective Envelope(Special Tray Fitting Pith Two 18-in.
Cabte Tray Sections)"; Omega. Point Laboratories Project No.l 1960-97186.
6.1... "Fire Endurance Test ofa.Thermo.Lag 330-,1.Eire.protccttve Envelope (Common.Enclosure withThree 18-in..
Cable Trays and Covered 18-ia. Tray with 1sin. and S-in. AirDrops)", Omega Point Laboratories Project No.
11960'-92187'.-'.t.lf
'Fire Endurenee Tert or n Thermo'tutg 33ntt Fire Protective Enveiope (Two Sided Mutdpte Conduit Enclosures and Cable Tray Support System)"; Omega Pomt Laboratories Project No. 11960-97257.
6.1.1 "Fire Endurance. Test ofa Thermo-Lag 330-1 Rue Ptotcctive Envelope(Three Sided Box Enclosure Encasing'Grou ps-opf Horimntal'Conduits and'a Large Junction Box)"";Omega Point Laboratories Project No. 1'1960 97258!
6.1.15 "Fire Endurance Tm'ofa Tlfc'rmo-Lag 330-1 Fire Protective"Envelope(ThfetiFour Sided Box Enclosures.
Eaca'sing Qsstups ofVertical Coaduits and an Enclosure Eacasing a4 ia Conduit and a function Box)",Omega Point Eaboratarics Projec! No; I 1960-972593"'.1.1 "FireEashtraaal Tcstofhlltcrmo-Lig330-I: Fire Protective Envclo'pe'(Box'Enclosure Encasing a Gang of Sevea 4 iss'.Ssisal Coaduits and Iadividuat Enclosures on 3/4-in Alufninuinand Steel Coaduits)"; Omega Point
%11960-'972M;. '
w
'". '"We~wc -"'" a':-': VHREF HOUR FIRETKSTS- -
'>>-= '-:, --,t..
62.1
'FireEisduiatQSTc'sio'fi'ihunisktg 330-1.Fire Protective Envelope(12ciis aad 24-in..Cablss Toys,'and a 12-ia.'x:12M:x 60'-in Jusictiott'3oss)<<",=Om'egst'.Poh)t Laboratories:ProjcctBo. 11960-97555.
622 "Fire Ehduriuicsfrest afallseHA~330-1'Fusr Protective Eavclope(12 ia'.ands-ia Cable Trays,aui5-ia.-2-'in:; and'1-:hr Steel Coadtiia)",'.Omega PoiitLaboratorIcs Project No. 11960-97553; r
't',
~
~
en'~in'~e
0
~
II,
Calcul'ation PTN-8FJH-96-'028
'evision 0
Attachment '.1:
Page 20 of 23.
usrent Fue Endurance Test Results for Thermo-Lag Fue Barner Materia, July 27, 1992. NRC Information Notice 8I2, "Results ofThermo-Lag 330-'1'Combustibility Testing". NRC Information Notice 93~;"Results ofThcnno-gin e
rays 30-1 Combustibility Testing", December f$;.:1992;"
Bulletin No. 92@i,."Failure ofThermo-'Lag330 F'uc Barrier Systems to Maintain CabHng in Wide Cable Trays mall,Conduits Free from Fire Damage", Sune 24, 1992.
C Bulletin No. 9241, Supplement I,"Failure ofThermo-Lag 330 Fue Bamcr Systems to Perform its Speci6ed
~
~
~
durance Function", August 28, l992.
~
t C Generic Letter 9248, "Thermo-Lag 330-1 Fire Bamers", December.17, 1992.
~
~
Subject 1712, "OutlineofInvcsngation forTests forAmpacity ofIitsuLucd Electrical Conductors Installed in Fire tive Systems", luly,1984...
~
P-848, Draft lEES 'Standard.entitle d, "Procedure for the Determination ofthe Ampacity. Derating of Fire Cables"; under preparation hy Task Force 12<5 ofthc Tests and Mcasuriments Subcomnitte'e afdt'e hiuiatcd
's Commiaee ofthe IEEE, -,,;
crmo-Lag Elcctncal Raceway Fire Bamer Systems. Testing and QualiQcation":, Mark,H. Salley. and Kent %;
delivered at the February 1993 WATTECconference in Knoxville,TN; "
I
~
pacity af Wrapped Cables", Keith A.gcttysiIEEEyapcr 86 SM 398-2, IEEE TrattsaTctlons on Power Delivcryi 3, number 1, January 1988 pp 3$-38:;......
iT.~ 'l"l": tt ll' Request forAdditio'nal fnfonnitionRegarding Gen'eric Letter 9248; hsucd Pursuint i'.NCHVM4(0dated aber 22,1994.. -"~- -.
C Gcncric Letter 86-10, "irriplemeniation ofPire Piotection.Requirements",'datedAprt
-241986..,
TME-119-~'.Standard Test Method forFiie Tests ofBuilding Construction and Materials".
IIMAERPRAQifdalincs forFuestop and Crap Systems ofNuclear Ficilitici;Article B, Revisioii 0, 1987' derwritm;Labeaary'Subject 1724, "Outline of Investigation for,Fire Tests for Electrical CIreuit Protective
", Issue nutnber 2, August 1991.
C.Standard. Review Plan, NUREG<860, Ication'9.$.1,'Fi're Prate'ction Proyim".'" "
CGeneric Letter 86-10 Supplcmcnt 1 "Fire Endurance Test Acceptance Cnteria forFire Bamer Systems used
'edundan't'S'afe Sliutdown'-Tiaini'withinthe'Samo Fire Area" dated" Mirchl25<!1994. ': ":
~
20 7.N 8.
- 9. IREE
~.+
- 10. M
- BTOTITI, voiuke
- 12. P 'ters 13.N C
- 14. N
- 15. A 16.
17.
Systc 18.
19.
to Se REFERENCES
- l. Cade ofFederal Regulatians (CFR), Title 10 Part 50, Appendix R. Paragraph lll.G.2: '
~
~ ~
P II
- 2. A detailed discussion on thc mechanics ofThermo-Lag 330-1'an be found in the NRCJNRR Og
'al roceedings Meeting with Thermal Science, Inc.. io Discuss.issues tnvolving Thermo-Lag 330", dated October 17 in c
Ncial Transcript of l991. NRC Public Document Room Accession Number 920205305.
- 3. NRC Information Notice 91<7; "Failure ofThermo-Lag Fire Barrier Material to Pass Fire Endu T
ass
.,lre utance Test"q Augiist I
I
~
~
I
~
~
I
- 4. NRC Information Notice 91-79 "Deficiencies in the Procedure for installing Thcnno-Lag Materials", December 6, C
199. NRC lnfarmation Notice 92M, "Thermo-Lag Fire Barrier Materiah Special Review Team Rc F
d ent Fue Endurance Testmg, and Ampacity Chlculatioii'Eirors~ Sune 23, 1992. NRC Information Notice 92-5$
92
'I I ~
~ t
~ l
0 II
Cal cul ati on PTN-BFJH-96-028 Revision 0
'ttachment"1 Page 21 of 23 2p, TVArefers ta tho niatcrial as nominal 5/8-inch or 3/8-.inch rather than minimum 1/2-inch or l/4-inch TMs '
-ol o
is duo ta TVA QA acceptance requirements.
The taleranccs used are 5/&-inch + 1/8-inch <<nd 3/8-inch + 1/8 Inch.
ments taken durin the construcuan ofihc test assemblies indicated the material was typically 5/8-'nch.
y
-inc or 3/8-
The ambient starting temperature of83'F must bc used for thc equations to be valid. Figure 2 has been simplified (i.e. the ambient.temperature subtracted} ta graphically show tile:allowable tcmperatu're rise (i.e; BT 250'F),
23..The Calcuhtian ofthc Tem pcrarure.Rise and E.oad Capability ofCable Systems",, L H; Neher and M. H. McGrath, AIEE Transactians, volume 76, October 1957, pp 752-773.
~
i
- 24. ICEA P46-426, 1962 edition, entitled, "Pawer. Cable. Ampacities".
Published by thc Insulated Cable Engineers Association'.
1
', e
~
~
- 25. IEEE Standard 835-1994, entitled "Power Ampacity Cable. Tables", dated, December 1994, section 33.5.
- 26. For a single open top tray, an equilibrium current of29.48 am ps was determined.
Withtho triple tray. vertical stack, an equilibrium'cur'rent of26.19 was'ineasurcd.
This'I I~to difference is the result ofa mutual heating effec. This effect is'not presently addressed by any standard.
It is iho author's experienco that some companies apply a SN detate to account for this effect whereas others ignore the influence ofother trays in close proximity because ofthc signi Geant diversity that exists in nuclear station power trays.
- 27. "Fire Pratecdan Wrapped Cable Tray Ampacity", Phil Save and Gary Engmann, IEEE Traiisactians on Energy Conversion, volume 4, number 4, Decanber 1989, pp 575<<584.
- 28. The seven regions o'Fthe thermal mixfeI proposed by Save and Ehgmann were:. I)'who kiwerjunction ofthe cable.
mass; 2) Thd Qiewrapbelaw the bottom afthe tray." 3) Theolitside mr&ccoHiewrsp bchnv th'e 'bottatn'af thc'tray.
- 4) TItc upper portian ofthe cable masL 5) The ah'space between the top ofthe cable mass and tho firewrap, 6) TItc firowrap over the top ofthe tray. 7).Thc outsidyaurhcc ofthe fire.wrap over the top ofthe tray,
- 29. "$mpacitics for Cables in R@uhmly Filled,Trays", S,Stolpe, IEEE Trattsacttons on. Power Apptiratus and Systems,.
Volutne 90, Piit I, 197), pp. 962-9)4...
r
- 30. SUPE frst cditian, 1)88 National FireProtection Association, Quincy, MassaIchuseas 02269, Chapter 34...
~ ~ ~i
~
li
\\
V I
~
~'i
~
A
0 0
1 jS c
l
~~,
~
~
50'. ATMhXNUMALLOWED
~ 4
~ i'>; l 'i,
~ f,r, T '
~
~1" dia condut I -1 taye<, 516" (TVATost 6.1.1) 1 dta cooduit - 2 tayefs, 516)
+ 3Itt"(TVATest 6.1.3)
~3 die collduit-1 taye<,5/6" (TVATest 6.1 4)
~3'dia conduit - 2 laye4. 31tt
+ 3IS" gVATesl6.1.3)
~ 5"dia coeduc 1 taye<,
5fQ'TVA Test 6.1.1)
~ 4 I
~.4 jw
~
~
NKNR RATED THERMOUS 33M CONDUITERF SS AVERAGE NTERNALTEMPERATURE RISE (5T}
NO CABLEFILL
~~'AA gy ~ID of
~at%~
gg gg~O O MP g~~
o~ -~
~~OO FO I O
I CPTl
+IlO Ol co
0 Cl
~O
250'T MAXNUMAllOWED C
I
~'t i' I
r
~" i i
i
~
~ ~
~18 Cabhlray-Nocabhhtt gVATeal 6.1.7)
18'able tray - 6.24 hs/ft cabh 68 (TVATost 6.1.7) 1
~18" Cabh tray - 69.36 Ibs/ft cable Q (TVATeal 6.1.7)
~
~
r
< ~ '. a',I.'-HOURRATED THERMO4AG 330-1 CABLETRAYER1:BS AVERAGENTERHALTEMPERATURE RlSE (hT)
~+NA as <<+ iD A i't< ~
(ii gy ~A g1 Vl C
~BOA
$ 5 8 oa
~~WO
~l D
I n
C I