ML20070C980

Forwards Summary of NRC Review of Exxon Nuclear Co Thermal Hydraulic Methodology for Mixed Cores in Pwrs.Encl Toprod Sensitivity & Thermal Margin Studies Submitted on 820618 & 29,respectively,remain Valid.Comments Requested

ML20070C980
Person / Time
Site:	Prairie Island
Issue date:	12/09/1982
From:	Cooke G SIEMENS POWER CORP. (FORMERLY SIEMENS NUCLEAR POWER
To:	Joseph Holonich Office of Nuclear Reactor Regulation
References
GCC:048:82, GCC:48:82, NUDOCS 8212140411
Download: ML20070C980 (7)

Text

~

, a E(DN NUCLEAR COMPANY,Inc.

11$1#ws/lquidrAusf

r. a au f.at acum< wadaptemmt Phone:(509)375-8100 Teler: 15 2878 December 9, 1982 GCC:048:82 Mr. J. J. Holonich Core Performance Branch Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D. C. 20555

SUBJECT:

ENC Thermal-Hydraulic Analyses for PWR Mixed Cores

Dear Mr. Holonich:

With your review of ENC's thermal-hydraulic methodology for mixed cores, including consideration of fuel with different rod sizes nearing completion, we thought it might be helpful to summarize the review as well as related analyses and developments from our perspective. This summary is provided in Attachment A.

As a result of your review, we found that automating the interface between core and subchannel analyses provides significant implementation advantages over our previous way of doing such analyses and this approach will be used in future work by ENC. As a result of our recent work with the automated interface in XCOBRA-IIIC, we believe our prior analyses to be accurate though they were cumbersome to perform. Thus, we believe our bounding analysis for ENC TOPROD fuel in Prairie Island Units 1 and 2 that was provided in submittals of June 18, 1982 to Gene Hsii and June 29, 1982 to Joe Holonich continue to be valid and thus supports future cycles of ENC TOPROD fuel in Prairie Island Units 1 and 2. For reference, I have provided these submittals again here as Attachments B and C.

~

We will be sending a copy of this letter and attachments to Northern States Power with a recommendation that it be used by them in supporting the validity of the current analysis. We would appreciate comments you may have regarding this matter.

Sincerely, G. C. Cooke, Manager Plant Transient Analysis GCC:GF0:gf l Attachments As noted l

8212140411 821209 goaaoocxosooom . . . _ . . . . .

f

ATTACHMENT A_

SUWiARY OF THE NRC'S REVIEW 0F ENC'S THERMAL HYDRAULIC METHODOLOGY FOR MIXED CORES IN PWRs The present detailed NRC review of EhC's thermal margin methodology for PWRs started with the review for ENC TOPROD reload fuel in Prairie Island Unit 1. The ENC calculations were performed with ENC's XCOBRA-IIIC code using a manual interface between core wide flow distribution analyses and limiting assembly MDNBR subchannel analyses. In the case of Prairie Island Unit 1, the original overpower MDNBRs being calculated with the W-3 correlation were on the order of 1.8 to 2.0. Reactor conditions were such that strong interassembly crossflow did not occur, and the use of an axial average flow factor from the core-wide flow distribution analysis was found to be adequate to establish limiting assembly flow in the MDNBR subchannel cnalysis.

Upon review of the TOPROD work, the NRC was concerned as to the general validity of the approach and was further concerned that crossflow at the interassembly boundary may impact MDNBR especially when the neighboring assembly has a different rod size from the limiting assembly. In response, ENC performed additional sensitivity studies. These calculations involved single and double octant subchannel models and were performed at much higher power conditions in order to produce MDNBRs near the 95:95 limit for the W-3 correlation. The purpose of going to a higher power beyond that applicable to Prairie Island was to examine the general validity issue.

In addition to going to higher power in the sensitivity studies, different axial power profiles were considered. At the higher power level

interassembly crossficw was much larger than in the original Prairie Island core thermal hydraulic calculations. For these new conditions ENC found it necessary to tighten the convergence parameter in the XCOBRA-IIIC calcula-tions and also found it necessary to use more care in selecting the limiting assembly inlet mass velocity in the MDNBR subchannel calculations. With respect to the latter item, selecting the inlet mass velocity, it was necessary to go beyond simply an axial average flow factor and to choose an inlet n. ass velocity which balanced conservatisms in enthalpy rise against non-conservatisms in assembly average mass velocity at the MDNBR elevation.

In view of the close agreement between single and double octant MDNBR results, it was concluded that the XCOBRA-IIIC-based core thermal hydraulics methodology with a manual interface between core and subchannel calculations (and without explicit account of interassembly crossflow) does provide an accurate calculation (approximately 1.0%) of MDNBR. Further, it was concluded that the single octant approach was generally applicable since in the case of non-peripheral MDNBR locations, it agreed closely with the double octant model and since it will be bounding for peripheral MDNBR locations. A less conservative but more accurate MDNBR calculation for peripheral MDNBR locations would require a double octant calculation.

Aside from the complication of choosing between single or double octant MDNBR subchannel models, the ENC methodology did have one major drawback.

Specifically, this was the fact that the analyst must have specialized training and experience to correctly establish the inlet mass velocity for the MDNBR subchannel analysis. The balancing of conservatisms in enthalpy rise against non conservatisms in mass velocity was straightforward in principle, but did require additional XCOBRA-IIIC sensitivity analyses to establish the variation of MDNBR with enthalpy rise and with assembly average mass velocity.

~

Furthermore, in the case of double octant calculations where the limiting assembly mass velocity varies with axial location, a trial and error process was required to establish the inlet mass velocity for the final MDNBR calculation.

In view of the potential for errors in application and the general cumbersomeness of the method, a new approach was taken by ENC. Specifically, an automated crossflow boundary condition feature was added to XCOBRA-IIIC.

This new approach did not change MDNBR results, but it did make it much easier and straightfoward to calculate MDNBR for any given core situation. General agreement has been found in audit calculations by the NRC between XCOBRA-IIIC analyses with automated crossflow boundary conditions and its own methods for core thermal hydraulics analysis.

To conclude, it is nrted that in its applications and in additional sensitivity studies for 0.C. Cook Unit 2 fuel geometries, ENC has found that the limiting assembly location in the core is not an important factor with respect to MDNBR. Secondly, it has been found that radial averaging of crossflow boundary conditions around the boundaries of the limiting assembly and applying this averaged crossflow boundary condition to the MDNBR l subchannel channel yields an accurate calculation of MDNBR. These findings l

l have consistently been found to be the case since the review of ENC's PWR l

l thermal hydraulic methodology was begun for ENC TOPROD fuel at Prairie Island Unit 1. Finally, it is noted that none of the developments during this review of ENC's methodology has resulted in any significant change to the very considerable MDNBR margin to limits in ENC's analyses for Prairie Island Unit l 1.

l

4 To : Gene Hsii, June 18, 1982. (informal)

ATTACHMENT B PRAIRIE ISLAND TOPROD SENSITIVITY STUDY Thermal margin results for the DNB limiting two pump coastdown event at Prairie Island Units 1 and 2 are reported in Appendix A.13, Table A.4, of the '

Exxon Nuclear Company TOPR00 thermal margin analysis document, XN-NF #-

56(P), Supplement 1. The results reported for a full TOPROD core are an MDNBR of 1.85 before accounting for rod bow and an MDNBR of 1.60 af ter a conservative account of rod bow as discussed in Appendix A.13 of XN-NF-80-56(P) Supplement

1. These results for a full core of TOPROD fuel have been confirmed as bounding of mixed cores of ENC TOPROD and ENC standard fuel in Prairie Island Units 1 and 2.

When calculating thermal margins, ENC methodology involves a core flow

' distribution analysis and an MDNBR subchannel analysis both of which are performed with ENC's XCOBRA-IIIC computer code. Crossflow boundary condi-tions are indirectly accounted for in the MDNBR subchannel analysis by selecting an input mass velocity that best matches the limiting assembly flow results from the core flow analysis. This is done by balancing conservatisms in enthalpy rise against nonconservatisms in mass velocity at the calculated MDNBR elevation in the MDNBR subchannel analysis.

In view of a concern that ENC's indirect approach to handling crossflow may not be accurate, a sensitivity study has been performed to bound the maximum potential error in the TOPROD analysis. Specifically, the impact to MDNBR of using the minimum mass velocity from the core flow analysis versus an average mass velocity which balances conservatisms in enthalpy rise against non conservatisms in flow, was evaluated. This evaluation was made for the limiting conditions at the time of MDNBR in the Prairie Island TOPROD two-pump coastdown transient event.

-~

~

The minimum flow over the active region of the fuel was found to be 3%

less than the average flow that balances enthalpy rise error against flow error at the NDNBR elevation (0.95 flow factor versus 0.98 flow factor). The associated reduction in MDNBR was evaluated to be 1.5%. Conservatisms(item d in Appendix A.10 of XN-NF-80-56(P), Supplement 1) in the initial MDNBR of 2.24 osed in the Prairie Island TOPROD transient analysis bound this 1.5%

reduction in MDNBR.

i e

l-4 0

~ .. ..." 4. "

To : J:e Holonich,. June 29, 1982 (informal)

ATTACHMENT C PRAIRIE ISLANL - THERMAL MARGIN As supplemental information to ENC's most recent evaluation of Prairie Island TOPROD therma'l margin (.0une 18, 1982, transmittal to Gene Hsii),

please note the following:

1) The calculated results for MDN3R at the time of limiting conditions during the two-pump coastdown transient are as follows:

MDNBR = 1.929 in transient XCOBRA-IllC core flow and MONBR subchannel calculations MDNBR = 1.892 in a quasi-steady core flow and MDNBR subchannel calculations.

These MDNBR results both reflect use of an axial minimum flow factor 1 of 0.95 in the single octant M3NBR subchannel calculations.

2) Examination of the core flow results shows that a 0.98 flow factor would balance conservatisms in enthalpy rise against non conservatisms in mass velocity at the MDNBR elevation. Use of the 0.98 flow factor in the quasi-steady MDNBR evaluation gave an MDNBR result of 1.919 (versus 1.892 for the 0.95 flow factor). Thus use of'a 0.95 flow factor yield a 1.5% reduction in MDNBR.

3) The MDNBR results cited above all pertain to a full core of TOPROD fuel at' Prairie Island Units I or 2. This case has been reconfirmed as bounding of mixed cores of TOPROD and ENC standard fuel. The radial

~

. power distribution used in the core flow analysis closely corresponds to and yields equivalent results to the Cycle 9 (2/3 TOPR00, 1/3 standard) distribution given in Figure A.1 of XN-NF-81-56(P), Supplement 1.

Use, reproduction. transtmttel or dec60sure of the above enformata is subsect to the restraction on the fwSt or title page of this document.

)