ML17313A795
| ML17313A795 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 02/04/1999 |
| From: | NRC (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML17313A793 | List: |
| References | |
| GL-92-08, GL-92-8, NUDOCS 9902120238 | |
| Download: ML17313A795 (9) | |
Text
gp,R RE0II
~4
~o Cy I
C1 C
O IA0
<r~
/J
~O 4~*~4 UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 SAFETY EVALUATIONBYTHE OFFICE OF NUCLEAR REACTOR REGULATION THERMO-LAGFIRE BARRIER CABLE AMPACITYDERATING ISSUES ARIZONAPUBLIC SERVICE COMPANY ET AL.
PALO VERDE NUCLEAR GENERATING'TATION UNIT NOS.
1 2 AND 3 DOCKET NOS. STN 50-528 STN 50-529 AND STN 50-530
1.0 BACKGROUND
1 The NRC notified Arizona Public Service Company (APS or the licensee) in a letter dated June 11, 1998, that the licensing actions associated with Generic Letter (GL) 92-08, "Thermo-Lag 330-1 Fire Barriers," were complete for Palo Verde Nuclear Generating Station (PVNGS),
Units 1, 2 and 3, with the exception of the cable ampacity derating issues.
The licensee's initial cable ampacity assessment, as documented in its submittal dated February 7, 1994, utilized a "Watts per foot" methodology.
This approach possessed fundamental weaknesses in providing an adequate assessment of the ampacity performance limits for individual cables.
In its submittal dated January 24, 1997, the licensee provided documentation regarding the implementation of a new ampacity derating methodology based upon an Institute of Electrical and Electronic Engineers (IEEE) Transactions on Power Delivery paper entitled, "Sizing of Cables in Randomly-Filled Trays With Consideration for Load Diversity" by H. C. Leake. The new approach, hereafter referred to as the Leake methodology is based upon a single modification of the Stolpe'ray modeling assumptions to allow some credit for actual load diversity in cable tray power loads in the ampacity assessment process.
By letters dated October 9 and December 24, 1997, APS responded to an NRC letter dated September 11, 1997, which requested further information associated with the implementation of the Leake methodology by the licensee.
The staff, in conjunction with its contractor, Sandia National Laboratories (SNL), has completed its review of APS's analytical approach for ampacity derating determinations.
The results of this review are documented below.
Attached to this evaluation are SNL reports, A Second Review of the Palo Verde Analysis of Fire Bamer Ampacity Derating Factors," dated August 14, 1997, and A Final Technical Evaluation of the Palo Verde Analysis of Fire Barrier Ampacity Derating Factors," dated January 9, 1998.
J.
- Stolpe, Ampacities for Cables in Randomly Filled Trays," IEEE Transactions on Power Apparatus and Systems.
Vol. PAS-90. Pt. 1. pp.962-974. 1971 9902i20238 990204 PDR ADQCK 05000528 P
2>>
2.0 EVALUATION The first part of this section provides a summary of the staff's questions and licensee's responses on the ampacity derating analysis, and the second part provides the staffs evaluation of this methodology as applied by the licensee.
Am acit Deratin Anal sis Review
~
Question t SNL found after a review of the licensee's cable ampacity assessment methodology (i.e., the Leake Methodology), that the licensee had failed to establish a basis for deciding when the subject method is appropriate or to have sufficient checks to ensure that unrealistic results are not credited via the analysis.
SNL recommended that the following set of limitations be established to resolve any potential misapplications or concerns regarding the use of the subject methodology:
It is recommended that in the application of the Leake method to diverse random fillcable trays, the maximum baseline capacity limit, or maximum baseline heat intensity, should under no circumstance be assumed to exceed 80% of the corresponding open air limits. That is, any calculation that estimates a baseline ampacity limit(or equivalently the corresponding heat intensity level) that exceeds 80% of the cable's open air ampacity should be discounted and disregarded.
s
~
The Leake method for crediting diversity should not be applied to the analysis. of any cable whose diameter is greater than or equal to N the tray filldepth as
'calculated using the ICEA definitions of cable cross-section and filldepth.
The Leake method should not be applied to any cable tray with a diversity of 50% or more where, in this case, diversity is defined as the ratio of the cross-sectional area of cables which are assumed to carry continuous loads to the total cable mass cross-section.
The licensee is requested to consider the incorporation of the proposed recommendations by SNL into the application of the subject methodology or alternativel, to address the app)icable concerns and criticisms as identified in the attached SNL report and to provide any direct test results which validates the Leake methodology.
Licensee Res onse In its submittal dated October 9, 1997, the licensee acknowledged the prudence of each of the recommended application limitations, and committed to implement each limitation.
The licensee did express some reservations regarding the third limitation as listed above, but has agreed to implement the limitation pending further industry research into
\\
the identified concerns.
The licensee's-supplementary submittal dated December 24, 1997, stated that a revision of the subject calculations has ensured compliance with the recommended limitations.
Staff Res onse The information provided by the licensee is fullyresponsive to the subject question.
Question 2 SNL noted several discrepancies in both the maintained spacing and diversity based examples provided in the licensee's submittal dated January 24, 1997.
Specifically, the following discrepancy was noted for the maintained spacing analysis:
~
The licensee should ensure that for the analysis of maintained spacing installed cables the ampacity correction factors from Table Vllof the IPCEA P<6-426 standard are being applied as appropriate to each case example.
It would appear from the cited example that these factors have not been included in the assessment.
The following discrepancies were noted for the diversity based analysis:
~
SNL was unable to reproduce the licensee's cited numerical results.
In particular, SNL's calculations estimated a baseline ampacity limitlower than that cited by the licensee even though the same methodology was applied.
It is recommended that the licensee resolve the apparent discrepancy regarding the cited diversity case example analysis, and ensure that no similar discrepancies have been introduced into the other in-plant applications of the methodology.
~
SNL's re-analysis of the example case indicated that the specific cable under study did not have sufficient margin to allow for estimated fire barrier derating impact.
It is recommended that in resolving the numerical results discrepancy described immediately above, the licensee provide a resolution of the apparently overloaded cable in the cited example tray.
The staff requested that the licensee resolve the subject discrepancies and ensure that the concerns cited are revised in any applicable licensee calculation.
Licensee Res onse In its submittal dated October 9, 1997, the licensee stated that where maintained spacing is between 25% and 100% of the cable diameter, the licensee utilizes the derating factors that are specified in Table Vllof the subject IPCEA standard.
The sample calculation inadvertently failed to state that cable tray 1EZA1DATKBBhas a maximum ambient temperature of 40'C. A comparison of the calculation performed by the licensee to the calculation performed by SNL indicates that SNL performed their
calculation using a 50 C ambient temperature.
The difference in the ambient temperature values used in the two calculations accounts for the difference observed by SNL.
The licensee recalculated the example case using 40'C as the maximum ambient temperature and confirmed that the results of the sample calculation were correct.
fI The information provided by the licensee is fullyresponsive to the subject question.
A lication of Am acit Deratin Methodolo The industry standards are derived from the Stolpe's model where all the cables in a tray are assumed to be loaded to an equal level based on the rate of heat generation per unit of cable cross-section or "heat intensity." This method assumes no toad diversity. For the heat transfer behavior internal to the tray, which is heat conduction within the cable mass, a simplified expression for heat transfer in a onedimensional mass with uniform heat generation is used to estimate the temperature rise from the surface of the mass to the hot spot. The same overall heat load is then used to estimate the temperature rise between the ambient and the cable surface, the tray-to-ambient environment heat transfer, based on simple convection and radiation correlations.
The resultant calculation is an estimate of the overall ambient to cable hot spot temperature rise. The heat load, or heat intensity, is adjusted until the predicted cable hot-spot temperature matches the maximum allowable temperature rating for the specific cables.
The general approach taken by the Leake methodology is to make one modification of the Stolpe tray modeling assumptions.
Specifically, the Leake methodology modifies the treatment of cable mass-to-ambient heat transfer by replacing the full load assumption with a reduced heat load based on actual cable loadings.
lt is this departure from the Stople model which represents the load diversity being credited by the licensee.
Since the actual heat load on a tray may be just a small fraction of the heat load assumed in the Stolpe calculations, this methodology reduces the role of tray-to-ambient heat transfer in the overall process.
Overall, the Leake model makes the same assumptions for the heat transfer behavior internal to the tray as the Stople model with no credit given to diversity. However, when the heat transfer between the cable mass and the ambient is considered the Leake methodology credits load diversity by using the actual heat load of the cables in the tray rather than a conservative estimate based on worst-case uniform heat generation.
Therefore, the Leake methodology represents a compromise solution to the more conservative Stople model. However, H. G. Leake acknowledged in the referenced IEEE paper that previously proposed diversity crediting methods can lead to non~nservative results, especially in the case of a highly diverse cable tray (a tray with only a few energized conductors).
lt is Leake's contention that by retaining Stolpe's cable mass thermal model, the licensee's methodology adequately assesses the ampacity load for individual cables.
~
~
Although the staff found that in general the Leake methodology represents a reasonable approach that can quantify some modest relaxation of the conservative assumptions of the Stolpe/ICEA methods by allowing credit for cable load diversity, there were some concerns
. regarding its implementation.
Given the potential for misapplication of the subject methodology there should be a basis for deciding when the method is appropriate, or that sufficient checks exist to ensure that unrealistic results are not credited in the final calculation. Therefore, pending further study, the staff agrees that the following set of limitations as recommended by SNL be established on the implementation of the licensee's methodology to resolve these potential concern:
It is recommended that in the application of the Leake method to diverse random fill cable trays, the maximum baseline ampacity limit, or the maximum baseline heat intensity, should under no circumstances be assumed to exceed 80% of the corresponding open air limits. That is, any calculation that estimates a baseline ampacity limit (or equivatently the corresponding heat intensity level) that exceeds 80%
of the cable's open air ampacity should be discounted and disregarded.
The Leake method for crediting diversity should not be applied to the analysis of any cable whose diameter is greater than or equal to N the tray filldepth as calculated using the ICEA definitions of cable cross-section and filldepth.
The Leake method should not be applied to any cable tray with a diversity of 50% or more where, in this case, diversity is defined as the ratio of the cross-sectional area of cables which are assumed to carry continuous loads to the total cable mass cross-section.
The licensee has committed to implement the recommended limitations described above to the Leake methodology. Therefore, the staff finds that the licensee has provided adequate information to resolve the ampacity-related concerns raised in GL 92-08.
3.0 CONCLUSION
S The licensee has been fullyresponsive to the concerns identTiied by the staff and its consultant SNL, and the licensee has committed to implement the recommended application limitations.
Therefore the staff concludes that no significant safety hazards are introduced through use of the licensee's ampacity derating methodology and that there are no outstanding ampacity related safety concerns at Palo Verde Nuclear Generating Station.
Attachments:
- 1. SNL Report, A Second Review of the Palo Verde Analysis of Fire Barrier Ampacity Derating Factors"
- 2. SNL Report, A Final Technical Evaluation of the
'alo Verde Analysis of Fire Barrier Ampacity Derating Factors'rincipal Contributor:
R. Jenkins Date:
February 4, 1999