RBG-44230, Forwards Response to 970512 RAI Re Thermo-Lag Related Ampacity Derating Issues.Encl Also Contains Supporting Calculations,Rev 1 to G13.18.14.0-178 & Rev 1 to E-218
| ML20211J409 | |
| Person / Time | |
|---|---|
| Site: | River Bend  |
| Issue date: | 10/03/1997 |
| From: | King R ENTERGY OPERATIONS, INC. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20211J416 | List: |
| References | |
| RBF1-96-0362, RBF1-96-362, RBG-44230, TAC-M85596, NUDOCS 9710080112 | |
| Download: ML20211J409 (31) | |
Text
- --
O. Ent:rgy oper;tions, Inc.
R!ver Bord Staton
- og A
- -..Entergy : = = ze77e Fax 504 635 5068 E*,Y!; ""*
f4t#: lear $afety & Cegida!c4y Affars October 3,1997 U S. Nuclear Regulatory Commission Document Control Desk Mail Stop PI 37 Washington, DC 20555
Subject:
River Bend Station - Unit 1 Docket No. 50-458 License No NPF-47 Response to NRC Request for Additional Information Regarding Thermo-Lag Related Ampacity Derating issues (TAC No. M85596)
File Nos. 09.5, G9.33.4 RBG-44230 RBFi-96-0362 Ladies and Gentlemen:
Please find attached the response to the NRC Request for Additional Information (RAl) dated May 12,1997. Attachments A and B contain the specific responses to the RAl Attachments C and D contain supporting final calculations G13.18.14.0-178,"Ampacity Derating Factors for Thermo-Lag 330-1," Revision 1, and E-218, "Ampacity Verification of Cables within Raceways Wrapped with Appendix R Fire Protection Barrier," Revision 1, respectively.
Draft RBS calculations G13.18.14.0-178, "Ampacity Derating Factors for Thermo-l.ag 330-1," and E-218, "Ampacity Verification of Cables within Raceways Wrapped with
. Appendix R Fire Protection Barrier," were originally provided to your office as attaciunents to a [[letter::RBG-43571, Provides Responses to 961016 RAI Re Ampacity Derating Including Supporting Draft Calculations G13.18.14.0-178,Rev 0, Ampacity Derating Factors for Thermo-Lag 330-1 & E-128, Rev 1, Ampacity Verification....Fire Protection Barrier|letter dated December 19,1996]]. As stated in a subsequent letter dated June 16,1997, at the time of the December 1996 submittal, the draft calculations were ,j being finalized and reviewed for acceptance by the RBS engineering staff.
As was also discussed in the June letter, the RBS engineering staff identified a number of , yh concerns with the draft calculations some of which were similar to those discussed in the
,, RAI. The calculations are now approved. We expect that the revised calculations will
,' address all of the concerns identified in the RAI.
"(( hi jb
% Wp \ l pilF N1 \ i I!ElllHJR]RR]Ill 9710000ilb971003' 4 PDR ADOCK 05000458 P PDR
l Response to NRC Request for Additional Information Regarding Thermo-Lag Related i Ampacity Derating Issues (TAC No. M85596)
October 3,1997 RBG-44230 RBF1-96-0362
- Page 2 of 2 Should you have any questions or require additional information, please call Mr. T. W. Gates at 504-3bl-4866.
Sincerely, 4
I A 1
K/RMM/kym 44, r,g_ pa Attachments cc: U. S. Nuclear Regulatory Commission (w/o Attachments C & D)
Region IV 611 Ryan Plaza Drive, Suite 400 Arlington,TX 76011
- NRC Sr. Resident Inspector (w/o Attachments C & D)
P. O. Box 1050 St. Francisville, LA 70775 Mr. David Wigginton NRR Project Manager U S. Nuclear Regulatory Commission M/S OWFN 13-H-15 Washington DC, 20555
4 ATTACHMENT A Responses to Request for Additional Information
e Responses to Request for Additional Information Thermo-Lag Ampacity Derating issues (TAC No M82809)
The responses to the Request for Additional Information (RAI) comments are provided below, in cases where amplifying information on the RAI question was deemed appropriate, the information related to the question from Attachment 1 of the RAI has been included.
Req:testfor AdditionalInformation (1)
The licensee has not yet identified what course ofaction will be taken to resolve those cables that were identifled as overloadedfor application in Ihe plant. The staffagrees with the Sandia National Laboratories (SNL) assessment that the National Electri: Code overcurrent protection provisions [i.e., Articles 240-3(b) and 240-6] as described in the licensee 's submittal dated December 19,1996 cannot be used as a basisfor the resolution of the overloaded cable issue (see section 4.1.3 ofti.e SNL letter report (Attachment 1(a)]for details). In addition, the licensee is requested to estimate conservatively the remaining cable hfefor those cables that have operated under overloaded conditions.
Response (1)
EOI has incorporated the SNL comment related to the use of the NEC overcurrent protection provisions.
There are several cables that are thought to have operated at overload conditions in calculation E-218. On April 3,1997, RBS initiated a Condition Report (CR 97-0455) to document the cables that were identified as significantly overloaded in the draft revision 1 of E-218. These cables consisted oflow voltage pow cables (HVY).
The low voltage power cables identified as overloaded (HVY) are located in an area of the plant vehere they are not subject to the environmental influences of a design basis event. For cable that is not subject to the requirements of 10CFR50A9 (equipment qualification), diere is no real
" life" defined. The concept of" qualified life" as applied to electrical components defines that period after installation for which the comw nent can be shown to perform its function and still function as required during and followin 'esign basis event. The " qualified life" assures that sufficient capability exists within the equ., tent to survive the design basis event conditions after exposure to its complete " life." As a result, the specification of a " life" for these cables is not meaningful.
To address the potential for premature degradation of these cables, field testing of the cables was performed. In order to perform the testing, a small portion of the Thermo-Lag barrier was removed to allow access. A calculated temperature based upon the ampacity values in E-218 Anachment A Page1of13
~
l l
l showed that the cable conductor temperature would be approximately 105 C (275'F) to 130'C j (266'F ) under normal operating conditions. With the small thickness of the insulation and '
jacket material (approxi.nately 30 circular mils each), a temperature at this value s.ould be ;
immediately noticeable to the touch. No difference was noted between the cable and room I temperature (approximately 70-72 F). This indicated that the cable was not operating at the predicted temperature.
Te> ting of these cable w s also perfomied using the EPRI Cable Indenter mechanism. The EPRI cable indenter has been used in the past by SNL to measure degradation of cable, in NUREG/CR-5772, Volume 2, SNL concluded: "llardness and indenter modulus both increased with age for the materials tested. Indenter modulus measurements were clearly more sensitive than hardness measurements."
As a baseline for comparison, indenter modulus values were obtained for cable obtained at the same time as the installed cable and stored in the cable yard. The indenter modulus values for the overloaded cables were similar to the values obtained for the cable in storage. This clearly indicated that no significant aging has occurred.
The cable is manufactured with ethylene propylene rubber (EPR) insulation and chlorosulfonated polyethylene (CSPE) jackets. CSPE or liypalon* has been shown by SNL and others to be sensitive to thermal degradation. All physical testing (e.g., flexibility, indentation, indenter modulus) showed that the CSPE remains flexible and no significant thermal degradation has occurred. Assuming the temperwure calculated from the ampacity values for a ten year period (since construction), significant observable changes would be expected for the cable. Absent any observable change in the cable, and the results of the indenter modulus testing, no reduction in the life of the cable was taken.
As a final resolution for these cables, a corrective action was assigned to remove the Thermo-lag barriers from the affected cable trays. This action was completed.
Requestfor AdditionalInformation (2.a)
SNLfinds that the licensee has inappropriately applied convection coefficientsfor surface heat transfer in an unrestricted open (external) environment to the highly confined interior ofthefire barrier systems analy: ed This observation applies to both the conduit and cable tray analyses.
The licensee is requested to provide an explicitjustificationfor any such cases that includes a discussion ofitem-to-item, item-to-barrier surface, and item-to-wall / ceiling clearances. (See Sections 3.2.1 and 3.2.2 ofthe SNL letter report [ Attachment 1(a)]for details.)
Response (2.a)
EOI realizes that the heat transfer equation used in the calculation are for external surfaces where a free boundary layer is able to grow without interruption from nearby objects. We also agree Attachment A Page 2 of 13
s.
that this equation may be inappropriate in the confined interiors of a fire barrier where the scr.le of the confined space may interfere with free growth of a boundary layer and formation oflarge convective currents, llowever, we are reluctant to implement a cavity equation as suggested by the reviewer for the following reasons:
- 1. The method has been developed with emphasis on large barriers where the interior areas of the barrier are sumciently large (such as configurations Ul and U2 of the Calculation) to allow free development of a boundary layer and large convective currents. Therefore, the equation for exterior surfaces is more appropriate than the equations for cavities.
- 2. Because of their interior geometries, barriers are not ideal cavities for which empirical heat transfer equations could be applied without restrictions. In many cases, equations for cavities would be equally questionable (due to aspect satios or geometric irregularities) as the equation for exterior surfaces.
- 3. In many cases it is hard to distinguish between two distinct cavities at the top and the bottom (as suggested by the reviewer) due to communication through the openings between the cables and the gaps between the tray side rail and the barrier walls.
]
E--
- 4. In the calculation, where the width of the cavity is sumciently small so that heat transfer by conduction is dominant, the equation for exterior surfaces is replaced by the classic conduction equation. This applies to all individually wrapped conduits covered in the calculation.
The equation for the convective heat transfer used and the cavity equation suggested by the reviewer are described in Attachment B. The heat transfer coefficient calculated using the equation for exterior surfaces (Equations B.2 and B.3) was compared with the heat transfer equation for cavities (Equations B.9 and B.10). This comparison is shown in Table A.1 where the heat transfer coefficients for a typical cable tray are calculated. The cable tray is 24 inches wide and has a 4 inch cavity at the top and a one inch cavity at the bottom. Heat transfer coefficients are calculated at two different air temperatures (130 F and 160 F) for a range of temperature differences (10 to 30 F) that adequately cover the range of conditions encountered in raceway barrier installations. The input data and the equations used are identified in Table A.l.
p The results show that:
- 1. The heat transfer coefficient for the upper cavity calculated using the EOl method is substantially smaller than the heat transfer coefficient calculated using the cavity equation.
- 2. The heat transfer coefficient for the lower cavity calculated using the EOl method is less than or equal to the heat transfer coefficient calculated using the cavity equation for up to 20 F AT (surface to-air temperature difference). At AT=30 F the EOI equation yields about 10%
higher heat transfer coefficient than the cavity equation.
Attachment A Page 3 of13
.i l
1
- 3. The average interior heat transfer coemcient calculated using the EOI method produces a l cavity heat transfer coemeient that is about 10 to 20 percent smaller than the average heat transfer coemeient calculated using the cavity equation.
- 4. The convaum heat transfer coefficient is about 1/5 th of the radiative heat transfer coemeient.
From this comparison it is concluded that the heat transfer equation .used by EOl is more conservative than the cavity equation suggested by the reviewer. The EOl method produces a smaller heat transfer coemcient which results in a reduced heat dissipation rate through the barrier, and therefore, an increased ampacity derating factor for the ceN in the raceways. It is interesting, however, to notice that the effect of this conservatism on a a npacity derating factor is quite small dee to the fact that the convective heat transfer coemeient is in parallel with a much larger radiadve heat transfer coefficient and the thermal resistance of the cavity is only a small portion of the ocarall thermal resistance for the entire raceway-fire barrier system. This is demonstrated below:
. The overall heat transfer coemeient for the entire barrier system is (referring to the thermal modelin Figure A.1):
I I /'d'6
+
1
' AI + 1L +
1 U U,rA,, hc .,s + ha.,s < A,, ks h c-k + ha.s, (A.1) where Ab surface area of the barrier, ft2 2
Ar surface area of the raceway, ft U overall heat transfer coefficient for the entire raceway-barrier assembly (based 2
on Ab), Btu /hr-R ,op 2
Ur overall heat transfer coemeient for the raceway, Btu /h-ft ,op h,.m convective heat transfer coefficient from the raceway to the barrier wall (cavity heat transfer coefficient), Btu /hr-R2,op hrad-rb radiative heat transfer coefficient from the raceway to the barrier wall, Btu /hr-ft 2,op hc.ba convective heat transfer coefficient from the barrier wall to the ambient, Btu /hr-ft'- F hrad-ba radiative heat transfer coefficient from the barrier to the ambient, Btu /hr-R 2,op lb thickness of the barrier, ft
- kb thermal conductivity of the barrier, Btu /hr-ft 'F Attachment A Page 4 of 13
DifTerentiating the overall heat transfer coefficient U with respect to the cavity heat transfer coefficient hc.rb and ignoring the second order effects (change in other heat transfer coeflicients due to the change in temperature) gives:
AU U Ahc .,s ' As' U " h -rbc + h,as_,s h e-rb + h,as.,s &
A reasonable estimate of AU/U can be obtained by using representative values for the parameters in equations A.1 and A.2. From Table 4.6 of the Calculation G13.18.14 .78, Revision 1, for a 24 inch cable tray in a one-hour rated fire barrier:
2 Ur 4.32 Btu /h ft - F (Uf in Table 4.6) hc-rb 0.20 Btu /hr-ft2,op hrad rb 1.22 Btu /hr-ft:- F (hrad ni Table 4.6) hc.ba 0.38 Btu /hr-ft2,op hrad-ba 1.19 Btu /hr ft 2,op tb 0.041 ft (0.5 inch in Table 4.6) kb 0.122 Btu /hr-R- F 2
Ab 5.2 ft 2
Ar 4.0 R 2
From Equation A.1, U=0.46 Btu /h-ft - F, Corresponding te a 10 percent increase in the cavity heat transfer coefficient hc.rb, the change in U (therefore the beat dissipation rate) is only 0.6 percent. For a 20 percent change in hc-rb, the heat dissipation rate increases by 1.2 percent. The input data used to calculate these values and the results are shown in Table A 2.
The corresponding effect on the ampacity derating factor due to an increase in the heat dissipation rate can be found by noting that:
l' Rn c,,,nca - UA(T,, - T,,) (A.3) where I conductor current, Amperes R conductor resista6ce (ac) per unit length, ohm ncon number of conductors per cable ncable number of cables in the raceway Tn conductor temperature, F Ta ambient temperature, F Attachment A Page 5 of 13
-n
- 4. :
Taking the logarithmic derivative of Equation 13 (A.4)
Using the definition of the ampacity derating factor (ADF),
MDF = - Al - =
1AU (A.5) .
-ADF 1 2U Thus, a 1.2 percent increase in the overall heat transfer coefficient (corresponding to 20 percent increase in the convective heat transfer coeflicient) decreases the ampacity_ derating factor by only 0.6 percent. This is the main reason why a simple yet conservative convective heat transfer model based an the exterior heat transfer equation was preferred to a more sophisticated cavity equation. The incremental benefit gained from a sophisticated model is well within the L uncertainties introduced into these analysis due to approximations elsewhere.
i i Requestfor AdditionalInformation (2.b)
SNLfinds that the licensee 's treatment ofexternal convectionfor cable tray systems is unnecessarily crude and does not adequately treat the diferences associated with surface orientation. The licensee is requested to modify its thermal model so that more realistic external convective heat transfer coeficients are derived. (See sections 3.2.3 ofthe SNL letter report (Attachment 1(a)]for details.)
Response (2.b)
The equation used for the external heat transfer coefficient (Equation B.2 of Attachment B) -
should be regarded as representing a bounding average heat transfer coefficient for a raceway (or a barrier). This is implicitly consistent with the reviewer suggestion of a surface area weighted average coeflicient. The averaging is built into the equation itself rather than applied after the -
individual heat transfer coefficients accounting for different surface orientations are calculated.
In the EOl equation, the exponent n (=l/4) is conservatively chosen for the laminar heat transfer mode as opposed to the turbulent mode which yields a higher heat tiansfer coeflicient. The coefficient a (0.20)is chosen to bound the average value of the coefficient for rectangular or cylindrical objects.- For a horizontal cylindrical object a=0.27 and for vertical plates / cylinders a=29. For a heated horizontal surface a=27 (upward facing) and a=0.12 (downward facing).
Thus, the coefficient used in the model (0.20) is approximately 25 percent conservative for
. cylindrical items,30 percent conservative for vertical surfaces, and matches the average value for a heated top / bottom surface. The items considered in the model are either cylindrical (conduits),
or rectangular with two primary surfaces (cable bed) or four surfaces (cable tray barrier).
Therefore, the use of an average or a bounding heat transfer coefficient is reasonable from the
.. point of view of geometry. From the point of view of temperature, the approach is also Attachment A Page 6 of13
.w _.
reasonable since the items of concern (cable bed and the barrier) are exper d to be at their respective uniform temperatures. The wid:ly used ICEA standard [Ref. 3]: . sed on the same averaging approach. The heat transfer equation used in the ICEA standard is identical to the equation used in the EOI thermal model.
Table A.3 shows a comparison of heat transfer coefficients calculated using the EOl method with the heat transfer coefDeients that distinguish between the surface orientation. From the above discussion it should be obvious that the equation used by EOI produces a conservative, i.e.,
small, heat transfer coefficient when applied to cylindrical items and vertical surfaces.
Therefore, the comparison in Table A.3 is limited to horizontal top / bottom surfaces. The comparison is between the average heat transfer coefficient calculated by the EOl equation (Equation B.2) and the average value calculated using equations that account for surface orientation (Equations B.5 and B.6). The input data and the equations used are also indicated in this table. The results show that the EOI equation produces 10 to 25 percent smaller heat transfer coefficient than the equations that accoun. for surface orientation. As for the effect of this conservatism on the final ampacity derating factor, it should be noted that the external convective heat transfer coefficient is in parallel with a larger radiative heat transfer coefficient and it is only a small fraction of the overall barrier heat transfer coefTicient. Thus, the overall impact on the ampacity derating factor is of the same order of magnitude as the impact of the cavity heat transfer coefficient illustrated in Table A.3, i.e. less then one percent.
Requestfor Additionalinformation (2.c)
SNLfinds that the licensee treatment ofinternal heat transfer behavior within a conduit has not been adequatelyjustified. appears to be inappropriate, and is likely non-conservative. The licensee is requested to modify its analysis methodolog to conform to acceptedpracticesfor the analysis ofcable-to-conduit heat transfer. (See Section 3.2.4 ofSNL letter report [ Attachment 1(a)for details.)
Respons: (2.c)
It appears that the EOI heat transfer model for conduits was misinterpreted by the reviewer, probably due to insufficient discussion of the model in the calculation. The following paragraphs will try to compensate for this deficiency.
The EOI heat transfer model for conduits is illustrated in Figure A.1 and consists of the following distinct elements:
- 1. Cables inside the conduit
- 2. Air space between the cables and the conduit (called the enclosure in the calculation)
- 3. Conduititself
- 4. Air space between the conduit wall and the barrier (called gap in the Calculation)
- 5. Barrier wall
- 6. Surrounding air Attachment A Page 7 of 13
l 1
Contrary to the reviewer's statement, the model does account for the thermal resistance within I the cable bundle and the thermal resistance from the cables to the conduit wall. The model, however, does not distinguirh between these resistances individually, but combines them into a J
single equivalent resistance (defined by the heat transfer coefTicient Uconduiti n the Calculation G13.18.14 178, Revision 1). This value is back calculated from IEEE conduit ampacity Tables
[Ref.4]. A method based on the first principles as suggested by the reviewer would require several assumptions and approximations regarding the cable configuration within the conduit (how tightly they are bundled together, how symmetric they are, contact area, contact heat transfer coefficient, etc.), would be tedious, and would probably generate a larger list of questions. Instead a more direct approach was chosen as stated in the Calculation G13.18.14-178, Revision 1. This method,in essence, back calculates the cable conductor to-conduit thermal resistance from the ampacity data published by IEEE. This method is preferred to the direct method suggested by the reviewer since it is straight forward and easier to apply.
Furthermore, the method ensures consistency with the IEEE conduit ampacity tables, i.e., in the absence of a fire barrier the method produces an ampacity value consistent with the IEEE tables.
The method is also consistent with the Neher-McGrath methodology suggested by the Reviewer since the ampacity data used, i.e., IEEE tables, is generated using the Neher-McGrath methe i Requestfor AdditionalInformation (2.d)
SNLfinds that th msee has not calctdated radiation viewfactors correctly in particidar,for those analyses L 'mdtlple raceways in a common enclosure. The licensee is requested to correct its analysis on ums regard. (See sections 3.2.5 ofthe SNL letter report [ Attachment 1(a)]
for details.)
Response (2.d)
In the final issue of the calculation, the shape factors are calculated using parallel plate and parallel cylinder configurations (Equations B.11 and B.12 of Attachment B). These shape factors are based on reasonable estimates of the spacing between the raceways, EOI has obtained as-built dimensional data for the relative location of the raceways within the configurations Ul and U2 and has revised the shape factors used in the calculation based on the as-built data. The Thermo-Lag fire barriers have been removed from configurations Ul and U2.
With regard to the statement that the concrete walls should not be considered for rat 4tive heat transfer from the raceways because they are not credited for heat dissipation, we wcold like to point out that in the final issue of the calculation the assumption of no heat dissipation through the concrete walls has been removed. Regardless, the assumption of no heat dissipation through the concrete walls does not imply that they can not receive and teradiate/ convect the heat received from the raceways. This is also valid for the control / instrument raceways in the same enclosure. These raceways may block the view of the power raceway but still participate in radiative heat transfer. Accordingly, the calculation of the shape factors does not consider the concrete walls or the non-power raceways as radiation blocking elements.
Attachment A Page 8 of 13
1 e
l The radiation shape factors have been recalculated using the as built raceway configurations aac the equations in Attachment B. The calculations for the Unique configuration Ul are given in Table A.4. The summary provided below show that the values used in the calculation are in reasonable agreement with the values calculated in Table A.4.
Shape Factor (Fn)
Raceway ICK600NMI ICK600NA6 ITII200R ITK200R Calculation 0.85 0.70 0.69 0.38 Table A.4 0.71 0.62 0.68 0.35 (1) The raceway was mislabeled in the Calculation as 1CK600NAl The calculations have been revised to incorporate the shape factors calculated based on the as built raceway configurations and the equations given in Section B.3 of Attachment B.
Requestfor A dditional Information (2.e)
SNLfinds that the licensee comparison ofclad case ampacity limit estimates derivedfrom its own thermal model to tabulated base line ampacity limits is imippropriate. The licensee has failed to demonstrate that its thermal model is consistent with the thermal models used to develop the standard tables, and consistency between the clad case and base line case analyses is critical to the reliability and robustness ofthe calculations. The licensee is requested to explicitly determine base line ampacity limits using a thermal model consistent with that applied to the clad case analyses. (See Section 3.2. 7 ofthe SNL letter report [ Attachment 1(a)]for details.)
Response (2.e)
Baseline ampacity as defined in the Calculation, refers to the cable ampacity for an unprotected raceway. The baseline ampacity data used in the calculation are obtained from IEEE standard
[Ref. 4] for conduits, and from IPCEA standard [Ref. 3] for cable trays. Themial model used in the Calculation G13.18.14-178, Revision 1, are consistent with these data as discussed below:
The thermal model for conduits uses the IEEE standard to determine the combined thermal resistance of the cable bundle and the cable-to-conduit air space. This thermal resistance is then used (without alteration) to calculate the protected ampacity. The thermal resistance is intentionally calculated on the high side for conservatism by assuming that:
- 1. The data in the IEEE tables are based on conduit surface emissivity of 0.8 (as opposed to 0.2 for the aluminum conduits at River Bend)
Attachment A Page 9 of 13
-m e.
A
- 2. The coefficient a in the convective heat transfer coefficient in Equation 13.1 of Attachment B is 0.27 (protected ampacity is calculated using a=0.20).
- 3. The conductor ac resistance is the same as listed in NEC Tables (the conductor resistance used for River Bend is higher than the NEC values) liad the calculation used these same parameters to calculate a baseline ampacity, an identical value as listed in the IEEE tables would have been obtained. Due to the bias to calculate a higher thermal resistar.;e (therefore a lower protected ampacity), the baseline ampacity calculated by the thermal model is slightly less than the IEEE values. It should be noted, however, that the ampacity derating factors are calculated using the IEEE values, not the value calculated by the thermal model and are conservative due to the conservative estimate of the conduit thermal resistance and the use of the larger IEEE baseline ampacity.
Similar to the conduits, the baseline ampacity for cable trays is also obtained from industry standards; IPCEA [Ref. 3] for random filled trays, and IEEE [Ref. 4] for trays filled with maintained spacing. The thermal model itselfis fully consistent with the models used in these documents within the limits of the conservative approach followed in the methodology. The convective heat transfer equation used in the model is identical with the convective heat transfer equation used in ICEA. Similarly, the cable surface emissivity used in the model is identical l with the value used in ICEA. Therefore. the model is consistent with the model used to prepare the ICEA tables for baseline ampacities. This is demonstrated in Table 4.1 of Calculation G13.18.14-178, Revision 1, where the prediction of the model for an unprotected cable tray is compared with the ICEA ampacity value. The model's prediction (32.7 amps) agrees with the ICEA value (34 amps) within 4 percent. The reason why an exact match is not obtained is due to the thermal conductivity of the cable bed. This parameter is not published by ICEA but can be f back calculated from the heat flux data provided in (Ref. 3], EOI was reluctant to use the back l
calculated value since it would have implied building the desired answer into the model. Instead the data was taken from another reliable source [Ref. 5] which gave a slightly lower baseline ampacity than the ICEA value.
As in the case of the conduits, the ampacity derating factor for cable trays is based on the published data cited above, not on the value calculated by the thermal model, and since the ICEA baseline ampacity is greater than the baseline ampacity calculated by the thermal model, the calculated ampacity derating factor is also larger, it is, therefore, EOl' position that the baseline ampacities used in the calculation are consistent with the thermal model and slightly conservative.
- Requestfor AdditionalInformation (2.f)
I SNLfinds that the ampacity correctionfactors associated with the number ofcurrent carrying conductors in a conduit have either not been properly calculated or are still based on the older l! pre-1990 National Electric Code (NEC) correctionfactors. This item was also a concern identified in SNL 's earlier review, and the licensee response to the question in Section 2.3 ofthe l
J Attachment A Page 10 of 13 i
NRC staff's October 16.1996. RAI cited that the newer NEC correctionfactors would be used in all calctdations. (See Section 3.2.6 of the SNL letter report [ Attachment 1(a)]for details.)
Response (2.f) 4 When the number of current carrying conductors in a conduit is greater than three, E-218 uses the ampacity correction factors defined in the National Electrical Code, NFPA 70-1996, Article 310 15, Note 8(a) of Notes to Ampacity Tables of 0 to 2000 Volts, page 70-196. In E-218, these factors are found in Table 4.3.2. Plant Data Management System (PDMS) applies the correct factor based on its count of the number ofconductors in the conduit. These factors are used for all cables in conduit that are analyzed by PDMS in E-218.
The baseline ampacity for conduit ICK600NAl was inadvertently taken from IPCEA Table Vill for cables with maintained spacing which lists a correction factor of 0.7 for 10 to 24 conductors.
Thus, a baseline ampacity of 36.4 has been calculated instead of 26. Therefore, the revised calculation utilizes a baseline ampacity of 36.4 Amps.
Requestfor Additionalinformation (2.g)
SNLfinds thatfor cases involving multiple raceways (trays and/or conduits) in a single enclosure, the licensee 's independent treatment ofconvective heat transfer between each ofthe raceways and inner surface ofthefire barrier system is inappropriate. The licensee is retjuested to modify the thermal model to accountfor the simultaneous transfer ofthe total convective heat loadfrom allsources. (See Section 3.2.8 ofthe SNL letter report [ Attachment 1(a)]for details).
Response (2.g)
The statement by the reviewer that the heat transfer model used by EOl treats each raceway independently applies only to the calculation of the convective heat transfer coefficient from the raceway to the surrounding air . The barrier temperature (inside and outside) is based on the total heat generated simultaneously by all of the raceways enclosed within the barrier. Please refer to Figure A.1 illustrating the EOi thermal model. Since different raceways within the same barrier may have different thermal resistances, it is na' ural to expect that they may also have different surrounding air temperatures. This difference, however, is not very large. Examination of the inbles in the Calculation shows that the deviation from the mean air temperature is well within 10'F. This deviation from the mean air temperature does not introduce non-conservatism into the results. On the contrary, it introduces conservatism, though negligibly small, because the raceway-to-barrier convective heat transfer coefficient attains its highest value when based on the mean temperature. Any temperature above or below it produces a smaller heat transfer coef6cient. This is illustrated in Table A.5. Since the recommended approach (simultaneous heat transfer) would result in a single internal enclosure air temperature very close to the mean air temperature, the model used in the calculation is preferred to this recommended approach because ofits conservatism and practicality.
Attachment A Page 11 of 13
- e i
Requestfor A dditionalInformation (2.h)
SNLfinds that the licensee calculationsfor single aluminum conduits may be unnecessary. The licensee could instead apply test data availablefrom industryfor steel conduits as conservative estimates ofthe ADFfor an aluminum conduit (provided ofcourse, that thefire barrier configurations are roughly equivalent). Il'hlle this may actually result in less conservative estimates ofthe derating impact as compared to the current licensee estimates, this uvuld remove one significant source ofuncertainty in the licensee assessments, and wordd simphfy both this calculation and the overall licensee submittal. The licensee is requested to consider whether to abandon its calculationsfor single aluminum conduits and to instead rely on indt:stry datafor steel conduits in any case in which appropriately tested configuration can be identi., led for River Bendplant installations.
Response (2.h)
The ampacity derating factor used in the calculation for single conduits (21%) is based upon SNL's recommendation which includes an uncertainty margin in addition to the actual value measured by Texas Utilities Electric (TUE) to account for surface emissivity effects. The calculated ampacity derating factor based on 0.20 emissivity for aluminum conduits is 20%. EOl chose to use the value recommended by SNL since it was considered an industry accepted value.
Reqs.estfor AdditionalInformation (3)
It is noted that Calculation E-218 includes the use ofampacity deratingfactors derivedfrom Calculation G13.18.14.0-178, which SNL has determined to be sigmficantlyflawed in terms of model implementation. The licensee should revise Calculation E-218 in light ofary resolution for the concerns stated in item 2 above.
Response (3)
The responses to the comments regarding Calculation G13,18.14.0-178 are provided in paragraphs 2.a through 2.h above. Calculation E-218 has been revised as accordingly.
Requestfor AdditionalInfor. ation (4)
The licensee has documented conduit cablefills as high as 121 percent in I beir December 19, 1996, submittal It appearsphysically unrealistic to have a conduit load ofgreater than 100 percent. The licensee is requested to explain this apparent inconsistency.
Attachment A Page 12 o(13
r i
[ Response (4)
!: Conduit fill in the E 218 October 10,1996 draft indicate a number of conduits with fill greater
! than 100%. The reason for this is that the fill printed by PDMS for the E 218 attachments (and i observr.ble on-line in PDMS) is a corrected value equal to the actual percent fill (cross sectional 1 area of the contained cables divided by the cross sectional area of the conduit) divided by a 'ill i limit value which is based on the National Electrical Code, Chapter 9, Table 1,' The NEC and
[ PDMS filllimit values are:
i i e 53% for I conductor / cable in a conduit 1
- 31% for 2 conductors / cables in a conduit
- j. . 40% for more than 2 conductors / cables in a conduit i
- Therefore, if the actual percent fill (based on calculated cross sectional areas) for a single cable in I
conduit is 53%, the PDMS printed fill is 100% since the conduit is filled to 100% ofits allowable limit of 53%.
i In a number of cases at RBS. conduits are indeed filled beyond their allowable limit. Allowable fill limit is strictly a recommended limit for RBS associated with cable installation activities.
. The National Electrical Code conduit fill limits arc h: sed on " common coaditions of proper
,j cabling and alignment of conductors where the length of the pull and the number of bends are i within reasonable limits." Although cable installation is typically within the allowable percent fill limits, the absol.ae restriction is defined by the cable nanufacturer's allowable cable pulling j- tension. Utilities perform calculations and measure pulling tension during certain pulls to assure 4
that pulling tension does not exceed manufacturer's limits. ,
Ampacity of cable in conduit is not directly affected by conduit percent fill. Ampacity is based on the number of current-carrying conductors in the conduit. Refer to the response to item 2.(f).
- If percent fill is high, the number of current-carrying conductors is more likely to be high.
Baseline ampacity is not derated based on percent fill but rather the number of current carrying conductors in the conduit.
l Requestfor AdditionalInformation (5) l l In their December 19,1996, submittal, the licensee has documented numerous conduits that appear to be loaaed in excess ofthe loading limits established in the NEC (generally limited to 40-53 per cent loads depending on the conductor count). The licensee is requested to explain
- these apparent violations ofthe NEC i
4 Response (5)
' See response to question 4, above.
1 Attachment A Page 13 of 13
, _ .- _ . . . = . - - -- - . -. .
I I
(OsMTT rosDtfff4 tdest W wn(1 4
l 43 Oltt WLGTMo*IT
- tLL ,In06t nn t tes edl W 'll . A 10mrfin I $ ; Y f, ,,,b
, , , i
~
- vi i i
fo%Mffela t.
A o e,i
- m ,q
"*f -
g ej e.
r gal V
_ - ,A : p m idnas j utsotecavestm (b U" A -
- n. 2 % y T.
Y_
( L8LL IG (dLIlltf C l (L8LL S 4LaaEt @ W G WeHs j .ts(Lalito i l iu W 'd-l ,,3 _
_A
- i. _ "L SWBOLS SlllSCRIPTS 4 ThermalResisunce(=1JohorI AU) h heatinnsfercoct5cient a ambient a conductor q heat tiow rue b bamer r raceway
- TempnamSode i tempenrae e consectise rad radiatise C osera!Iheat transfer coefficient e enclosure Figure A.1 Thermal Model
Table A.1 Comparison of Convection llent Transfer Coefficients Interior Surfaces PHYSICAL CONSTANTS g, Gravitational Accel. (ft/hr') 4.17E +08 a, S. Boltzman C.(Btu /hr ft 'R*) 1.71 E-09 c, surface emissivity 0.8 w, Width of the tray (ft) 2 H, Height of the tray (ft) 1/2 -
xw width of upper air cavity (ft) 1/3 xn width oflower cavity (ft) 1/12 THERMAL PROPERTIES
. T. Film t..mperature ('F) 130 160 l
k, Thermal Cond. (Blu/hr-ft 'F) 0.017 0.017 v, viscosity (ft'/hr) 0.728 0.786 Pr, Prandtle number 0.70 0.70
- p. Compressibility (ifF) 1.69E 03 1.61E-03 (g/FvJ)Pr 9.32E+05 8.83E+05 '
ilEAT TRANSFER COEFFICIENTS T, Air T m. ('F) 130 160 T, Surfac. Temp. ("F) 140 150 160 170 180 190 CAVITY JT, (Tm-Tu). ('F) 10 20 30 10 20 30 EQN. Ra (=GrPr) Based on 2AT 6.90E+05 1.383+06 2.07E+06 6.54 E+05 1.31 E+06 1.96E+06 Nu,(Eq. B.9) 5.9 7.2 8.1 5.e 7.1 8.0 Nu,(Eq. B.10) 1.0 1.0 1.0 1.0 1.0 1.0 h,, 0.30 0.36 0.40 0.30 0.37 0.41 h, 0.20 0.20 0.20 0.21 0.21 0.21 I h., m,,o,(Note 1) 0.20 0.22 0.24 0.20 0.23 0.25 EOl h (Eq. B.2) 0.30 0.36 0.39 0.30 0.36 0.39 h,, m,,o,(Note 2) 0.17 0.20 0.22 0.17 0.20 0.22 hw(Eq. 8 of Ref. 2) 1.16 1.19 1.22 1.34 1.37 1.41 Deviation (Note 3) 19% 13% 10% 23 % 16 % 13%
Notes:
- 1. h,, m,,,,is the area weighted average of he, and h, and is based on the total tray area:
who + wha (A.1.1) j ' ,
2(w + H)
- 2. h,,p (EOl) is calculated from Equation B.3 assuming that h,_,, = h,j = h, This is equivalent to assuming that .dTis symmetrical which results in maximum h,3
- 3. Deviation is based on h,, ,,,,,,,,,(EOl)
-#'~ -' * ' ~
/Jevlat.'on(%) = (A.1.2) ha,.su <-(EOl)
Table A.2 Effect of Change in Convective Heat Transfer Coefficient on ADF Input Data (from Table 4.6 of Ref. 2)
Un Stu/hr ft' *F 4.32 he,,, Btu /hr ft' *F 0.2 he e., Btu /hr ftF 1.22 h,,,_e., Btu /hr ft' 'F 0.38 h,_,,, Blu/hr-ft'.'F 1,19 to in 0.5 k., Btu /hr ft 'F 0.122
, Aa, ft' 5.2 An ft' 4 Effect of Change in Convective Heat Transfer Coemclent l_ 4(h,_#1,_,, (%) 0% 10% 20%
h,,,,, Blu/ht ftN*F 0.20 0.21 0.24 U, Btu /hr-ft' *F 0.456 0.458 0.461 l
i 40.11 0.0% 0.6% 1.2%
A(ADF)/ADF 0.0% -0.3% -0.6%
.e .
Table A.3 Comparison of Convection Heat Transfer Equations -Esterior Surfaces PHYSICAL CONSTANTS g, Gravitational Accel. (ft/hr') 4.17E+08 a Stefan Boltzman C.(Btu /hr ft' *R') 1.71E-09 4 surface emissivity (Ref. 6) 0.8 w, Width of the tray (ft) 2 THERMAL PROPERTIES T. Film temperature (*F) 130 160 k, Thermal Cond. (Btulhr ft.*F) 0.017 0.017 v,viscositi(ft'/hr) 0.728 0.786 Pr, Prandtle number 0.70 0.70
/1, Compressibility (1/*F) 1.69E-03 1.61E-03 !,
, (gA'vJfr 9.32E+05 8.83E+05 ilEAT TRANSFER COEFFICIENTS Ta, Ambient Temp. ('F) 130 160 EQN. dT, Temp. diff. (*F) 10 20 30 10 20 30 D.5 & B.6 Ra (=GrPr) Based on W/2 9.32E+06 1.86E+07 2,79E+07 P. 83E+06 1.77E+07 2.65E+07 Nu,(Eqn. B.6) 29.8 39.8 45.5 29.4 39.1 44.7
! Nu,(Eq. B.9) 9.9 . 11.4 12.3 9.6 11.2 12.2 heu 0.50 0.66 0.76 0.51 0.67 0.77 ha 0.17 0.19 0.21 0.17 0.19 0.21 h,, ,,w,or .(h,,+he)/2 0.33 0.43 0.48 0.34 0.43 0.49 EOl h ,p(sh,in Eq. B.2) 0.30 0.36 0.39 0.30 0.36 0.39 hw(Eq.8 cf Ref.2) 1.16 1,19 1.22 1.34 1.37 1.41 Deviation 11 % l20% l22% l13% l22% l25%
~
Deviation (%) = ""~' -
x100 lla.
, nas,(EOl)
' Table A 4- Radiation shape Factors for Configuration UI Dimensions in Figure 3.1 (Inch) ra rb w 11 cc h cc v ct-h ct v tt v Side Raillleight 0.95 2.25 18.00 6.00 - 5.00 1.00 1.00 1.50 16.00 6.00 Parameters in Equations R.11, 3.12, B. I3 wi wj L el rj s si s2 R S C Shapel'.
a- ICK600NA 6
b ICK600NMI 0.95 2.25 1.90 2.37 2.00 5.37 Fab = 0.147 c ITil200R(top oO 1.50 0.95 19.0 1.0 Fac= 0.144 d iTK200R(side rails) 1.00 0.95 55.51.5 Fad 0.091 F/2= 0.62 b ICK600NMI l a ICK600NA6 2.25 0.95 1.90 0.42 0.84 2.27 Fba= 0.062 e 1Til200R(top 00 030 2.25 24.0 6.0 Fbe- 0.010 d iTK200R(side rails) 6.00 2.25 54.50.5 Fbg= 0.219 F 12= 0.71 l c ITil200R
! a ICK600NA6 1.50 0.95 19.0 1.0 fea= 0.048 l b 1CK600NM1 0.50 2.25 24.0 6.0 Fcb" 0.008 d 1TK200R 18.00 18.00 10.00 Fed = 0.59 F12= 0.68 d ITK200R a ICK600NA6 17.50 0.95 19.0 1.0 Fda= 0.04 b ICK600NMI 16.50 2.25 24.0 6.0 Fdb= 0.08 c ITil200R 18.00 18.00 10.00 Fde= 0.59 e ITC200R 18.00 18.00 10.00- Fec= 0.59 F/2= 0.35 Notes:
- 1. Refer to equations ]9, 20, 21, and Figure 3.3for nomenclature.
- 2. Conduit-to-tray shapefactors are bascd onfallowing conservative approximations:
- a. A solidplane extends along thefull width ofthe top ofIhe topmost trm>.
- b. A continuous solidplane covering the side rails extendsfor thefull height of
- the stack ofthe poner trms.
- 2. Trmsto-trm shapefactors are basedonfollowing:
- a. The spacing beturen the cable trm s is based on the actual measured dimension (not assumption 2.3.13)
- b. Tray isfilled to the top ofthe side rail andfor thefull width ofthe trm..
~
Table A.5_ Effect ofinterior Air Temperature on Heat Transfer Coefficient To"F 194 194 194 194 194 T., 'F - 146 . 156 1H 176 186 Tep*F ._ l140 _140 140 140 140 he,,, Stu/hr ft' *F 0.53 0.50 0.46 0.41 0.34 ht# Btu /hr ft' *F _0.31 0.40 0.45. 0,49 0.52 he, Blu/hr ftF 0.20 0.22 0.23 0.22 0.20
- References l
)
- 1. NRC Letter from David L- Wigginton to Mr. John R. hicGaha dated hiay 12,1997.
Subject:
Request for additional Information Regarding Thermo Lag Related Issues for River Bend
-(TAC NO. M82809) 4
- 2. Entergy Calculation Titled "Ampacity Derating Factors for Thermo-Lag 330-1 Enclosures", 1 i- Entergy Calculation No. G13.18.14.0 178, rev. 0 (March I 8,1997).
I
! 3. ICEA Standard Publication, Ampacities of Cables in Open top Cable Trays, ICEA P 54-440
- (Third Edition), NEMA WC 51-1986, t
3 .4. IEEE/IPCEA Standard S 135/P-46-426, Power Cable Ampacities: Volume I -Copper j Conductors and Volume 11 -Aluminum Conductors,1984.
! 5. Safety Evaluation Report by the Office of Nuclear Reactor Regulation,"Ampacity issues Related to Thermo-Lag Fire Barriers, Texas Utilities Electric Company, Comanche Peak Steam Electric Station, Unit 2, Docket No. 50-446," US Nuclear Regulatory Commission, June 14,1995.
a I
i-j--
i L
}
l i
4 I
i l
Attachment A - References
,,,---m-- , w w - - .-.r--, , --,. , - - - , , , ,-m
t ATTACIIMENT B Convective Heat Transfer Equations
. _ . . .. ... . . . _ . . . . . . . . . . . . u
B.I Convective Heat Transfer Equations Used by EOI B.I.1 Exterior Surfaces
- The equation used by EOI to calculate the convective heat transfer coeflicients for exterior surfaces is based on the free convection equation for air at atmospheric pressure. This equation,-
in its general form, is given below [Ref.1]:
h;= }(AT}" (B.1) where L is the characteristic length, ATis the temperature difference, and a and n are appropriate-constants as defined in (Ref.1, pp. 315]. These constants depend on the geometry of the saface (flat or cylindrical), dimetion of the heat flux (up or down), and flow regime (laminar or g turbulent). The exponent n is equal to 1/4 for laminar conditions and to 1/3 for turbulent L conditions. In the EOI method n=l/4 is used since it produces a smaller heat transfer L
coefficient. The constant a is set to 0.20. This value is the average of a heated upper surface (a=0.29), and a heated lower surface (a=0.12); Since the trays (and the barriers) always have an upper and a lower surface of equal area, using an average value is appropriate. The value used is
_ also smaller than the value for vertical plates (a=0.29) and cylinders (a=0.27).- With the coefTicient a=0.20 and the exponent n=l/4, the equation used in the RBS calculation is:
V -
g in h,-0.20 (B.2)-
.L, where he convective heat transfer coefficient (Btu /hr-ff 'F)
L- characteristic length (ft)
- AT temperature difTerence ('F)
The characteristic length L is set equal to the tray width (wt) for cable trays,- to the outside
' diameter (dcondult) for conduits, and to the largest dimension (width or height) for the enclosure -
walls. Equation B.2 is identical with the equation used in the ICEA/ NEMA standard [Ref. 2] to calculate the cable ampacities in the cable trays.
\
Attachment B Page1of5
)
B.I.2 Interior Surfaces Where the heat transfer is from one surface to another separated by an air space (as in the interior of a barrier), it is assumed that there are two film coefficients (in series) each defined by Equation B.2. Thus, for cavities (or gaps), between a raceway and the barrier the heat transfer coeflicient is calculated from:
h, ., = h""
h,., (B.3) h
]+(A,)(h,.,,,
where hc-rb cavity heat transfer coefficient (raceway to-barrier), Btu /hr ft2,op 2
hc.re heat transfer coefficient (raceway to-cavity air), Btu /hr 11,op hc-cb overall cavity heat transfer coefficient (cavity air-to-barrier), Bta/hr ft2,op 2
Ar surface area (raceway), ft 2
Ab surface area (barrier), fl The heat transfer coefficients hc.re and hc.cb are calculated using Equation B.2, When the width i of the cavity is small (as in the air gap between a conduit and a pre-formed round barrier), the heat transfer coefficient calculated from Equation B.3 is compared with the heat transfer f
i coefficient due to pure conduction across the air gap and the larger of the two is taken. The conduction heat transfer coefficient is calculated from:
k h, ,_ = ' '- (B.4) s where kg thermal conductivity of air, Btu /hr-A- F Ig width (thicl: ness) of the cavity, R Attachment B Page 2 of 5
_ _______j
.e s
s a
{ . B.2 : - Convective Heat Transfer Equations Suggested by SNL LB.2.1 Exterior Surfaces -
e
- The general equations'for exterior heat transfer coefficient that account for surface orientation and direction are given below. These equations are used in SNL's heat transfer model for ampacity derating factor [Refs. 3,4]. The SNL method distingu. hes between the upper and the ;
i . lower surfaces. The equation for an upward facing heated plate is:
Nu . " '2)
. o.u(or Pr)"' or (GrPr) s107 (B.S)
- -Nu= "(" )
= 0.15(GrPr)"" or (GrPr) >107 (B.6)
.k
! where Nu Nusselt number W width of the tray, ft h convection heat transfer coetricient, Btu /hr-ft *F k thermal conductivity, Btu /hr-fl 'F Pr Prandtle number Gr - Grashof numoer(defined below) l g, , xD(T. - Ta )(W/ 2)' zg,73
, v where g gravitational acceleration, ff/hr -
. p thermal expansion coefficient p = 1/(T, + 460) for air v viscosity of air, ff/hr The convection heat transfer coefficient for the a downward facing heated plate is:
y, , 0327(Gr Pr)0I __
(ggy (1+(1.9/ Pr)0'9]O2III i
Attachment B Page 3 of 5 d
g .
. B.2.2 Interior Surfaces There are two air gaps formed bet aeen the cable bed and the fire wrap. The heat transfet through the upper air gap is calculated by considering convection and radiation heat transfer
- according to [Ref. 3];
Nu = *[" =l 1 #f,+ + -1 ,
(B.9) where xlu is the width (thickness) of the air cavity (ft). The square brackets are set to zero if their contents are negative.
The heat transfer coefficient through the lower air gap is calculated using pure conduction. Thus, Nu= d*" = f (B.10) i k where x1/ is the width of the air gap (fl).
B.3 Radiation Shape Fa.: tors Radiation shape factors are applied to multiple raceways in a common fire barrier. Distinction is made between different configurations regarding tray-to-tray, conduit-to-conduit, and conduit-to-tray shape factors. Figure B.1 illustrates the configurations considered and the nomenclature used in the radiation shape factors. The equations for the shape factors are taken from Reference 5.
Radiation shape factor Fy for stacked trays in a commori enclosure is calculated using the equation for two parallel plates of widths "wi" and "wj" separated by distance L":
Fy = )(w + w,)# + 4 0 -l(w, - w )# + 4 0 g g 2 w, (B.Il) where Fy radiation shape factor from plate "i" to plate "j" w - width of the plate L perpendicular distance (gap) between the plates Attachment B Page 4 of 5
v i
i Radiation shape factor Fy for mult'ple conduits in a common enclosure is calculated using the equation for two parallel cylinders radius et and tj separated by distance (gap) s:
s,, (n + [C' -( R + 1)' J -(C' -(R - 1)' J
- (R-licosIf ') (11.\2)
-(R + 1)coi'[ + ))
I whem R=rfri S=s/rj C=l+k 6 and Fjf radinavi, shape factor from cylinder "i" to cylinder "j" r radius of the cylinder s gap between the cylinders Radiation shape factcr Fy from a tray (I) to a conduit (/) in a common enclor're is calculated using the relation for a parallel cy'inder and a plane:
F, = ,
tun - run (13.1 3 )
Equations 11.11,13.12, and 11.13 are applied, as :.ppropriate, to raceways in common enclosures in f
conjunction with the reciprocity theorem (A;/j=AjFjf) and the summation rule (IFy=1). An overall radiation shape factor for the raceway is, then, calculated as the area weighted average of the individual shape factors for each face of the raceway, i.e.,
A F,,=[(1)li, A (11.1 4 )
where F/2 overall shape factor between raceway (1) and its surrounding (2)
A/ total surface area of the raceway Af surface area of side "i" of the racewey F2f shape factor between the side "i" of the raceway and its surrounding Attachment B Page 5 of 5
d
't
- i
=-
rg L
= -__
- ) _. Parallel Planes
- -n -- rj 1 1 s_ ) Parallel Cylinders
- n-l 1 m si
(
wj s2 i Parallel Plane and Cylinder CC+h Cl*h r0a Nomenclature for Unique Configurations ct.,
e i H
I f
it v d
i
- Figure H,1 ConHgurations and Nomenclature Used to Describe the Radiation Shape Factors 4
Attachment B References
[,
11.3 REFERENCES
- 1. Ozisik, M. N., Basic # cat Transfer, McGraw liill,1977.
- 2. ICEA Standard Publication, Ampacities of Cables in Open top Cable Trays,ICEA P 54 440 (Third Edition), NEMA WC 51 1986.
- 3. Tanaka, T. J. et al.,' Fire Bonier System Cable Ampacity IMrating; A Review of Experimental and Analytical Studies", Sandia " !al Laboratories, August 25,1995.
- 4. Safety Evaluation Report by the Ollice of Nuclear Reactor Regulation,"Ampacity issues Related to Thermo Lag Fire Barriers, Texas Utilities Electric Company, Comanche Peak Steam Electric Station, Unit 2, Docket No. 50-446," US Nuclear Regulatory Commission, June 14,1995.
- 5. Incropera, F. Introduction to # cat Transfer, John Wiley & Sons,1985.
l l
Attachment D References I
J