ML20239A546

From kanterella
Jump to navigation Jump to search
Safety Evaluation Supporting A-85-11, Retran Computer Code Reactor Sys Transient Analysis Model Qualification for Use in Performing plant-specific best-estimate Transient Analyses at Plant
ML20239A546
Person / Time
Site: Calvert Cliffs  Constellation icon.png
Issue date: 09/09/1987
From:
Office of Nuclear Reactor Regulation
To:
Shared Package
ML20239A543 List:
References
NUDOCS 8709170428
Download: ML20239A546 (3)


Text

.. . . . _ -

. [pm Rf Cgh i [',m i UNITED STATES

, r,, . NUCLEAR REGULATORY COMMISSION l WASHINGTON, D. C. 20655

%...../

SAFETY EVALUATION BY TPE OFFICE OF NUCLEAR REACT BALTIMORE GAS AND ELECTRIC COMPANY CALVERT CLIFFS NUCLEAR POWER PLANT, UNITS 1 AND 2 DOCKET N05. 50-327 AND 50-318 RELATING TO THE USE OF RETRAN COMPUTER CODE FOR BEST-ESTIMATE TRANS]ENT ANALYSES

1.0 INTRODUCTION

The BG&E Topical Report A-85-11. "RETRAN Computer Code Reactor Syst Transient Analysis Model Qualification", was submitted to demonstrate BG&

capability the RETRAN 02/to perform MOD best 03 Computer Code.estimate thermal-hydraulic transient analysis NRC Generic Letter 83-11. This submittal was made in response to to evaluate Power Plant, Unitsproposed 1 and 2 changes to operating procedures c ear for Calv The 1986 subject topical report was submitted by BG&E in a letter dated Ja In response to staff requests for information, additional supporting ,

materials February 24,were 1987, submitted in BG&E letters dated March 7,1986, and June 9, 1987. December 31

, 1986, BG&E submittals Technical Services, Inc. with technical assistance from its contractors onal at 2.0 COMPUTER MODELING BG&E Topical Re configurations, for Calvert Cliffs as follows: port A-85-11 briefly describes primary and 1

analyze synnetrical transients.RETRAN One Loop Model of Calvert The one loop model explicitly uses a single hot leg, cold leg and reactor coolant pump (RCP) to represent the the Calvert Cliffsofdesign.

combination the two hot legs, four cold legs and four RCPs actually i 2.

analyze loop asymmetric transients.RETRAN Two Loop Model of cold legs are modeled, one for each steam gene representing two cold legs and two RCPs.

Two N) A P

. 3. RETPAN Four Loop Model of Calvert Cliffs - This model was developed to analyze individual cold leg /RCP transients. The four loop model explicitly simulates the two hot legs and four cold legs geometry of the Calvert Cliffs design.

One of the above three models will be selected for analyzing each transient depending upon the expected plant thermal hydraulic behavior during the specific transient.

In response to staff requests, BG&E has provided sufficient information to demonstrate its understanding of the details and limitations of those models.

BG&E utilized the RETRAN 02/ MOD 03 computer code which was a modification of RETRAN 02/ MOD 02. More than two hundred changes were made in this modification.

These changes were a result of 1) user reported errors, 2) changes requested by the Utility Group for Regulatory Application (UGRA) Design Review Team, and 3) changes requested by NRC after the staff review of the RETRAN 02/ MOD 02 for error corrections. In a letter to NRC dated February 4,1985, Mr. T. W.

Schnatz of UGRA stated that the total effect of all those changes made in RETRAN modification as measured by sampling problems used to test each RETRAN version was insignificant in that the differences were well within the uncertainties in the original calculations. This corrected version of the code has not been approved by the staff. Nevertheless, we believe that there is reasonable assurance that this version of the code was adequate for use in this topical report. Unless some future NRC review of RETRAN 02/M0D 03 contradicts this assumption, its use shall be acceptable in the BG&E's best-estimate model.

3.0 VERIFICATION OF RETRAN MODELS FOR CALVERT CLIFFS BG&E has compared the results from analyses using the RETRAN models developed for Calvert Cliffs with actual plant operating data, results from analyses using the TRAC code, and the results from the LOFT tests.

BG&E performed comparisons of the RETRAN model analysis results with actual pisnt operating data for the following scenarios: 1) multiple secondary side malfunction event, 2) four pump coastdown from 20% rower, 3) reactor coolant operation pump combination flow tests at hot zero power, 4) one pump coastdown from 80% pcwer, and 5) total loss of flow / natural circulation test from 40%

power. Various RETRAN models were used in the comparison with the results of the analysis being reasonably close. Sensitivity studies were conducted by BG&E to determine which parameters would affect the transient results.

In comparing the results of the RETRAN model analyses developed for Calvert  ;

Cliffs with the results obtained from analyses performed by Los Alamos l National Laboratory using TRAC-PF1 nodalization of the Calvert Cliffs plant, the following cases were reviewed: 1) cooldown to residual heat removal entry using atmospheric steam dump valves and auxiliary pressurizer spray system, and

2) runaway main feedwater to one steam generator. Overall, these analyses exhibited similar trends despite large differences in the computer code models

and the degree of detail and capabilities in plant nodalization. BG&E has provided sufficient explanation to demonstrate its understanding of the results from its RETRAN models and of the reasons for differences from the TRAC-PF1 analyses results.

EG&E also compared RETRAN calculations with the data from LOFT Tests L6-1 and L6-3. In the LOFT Test L6-1 comparison the BCEE RETRAN model calculated a primary and secondary pressure transient behavior that did not precisely follow the test results. However, calculated results by RETRAN, with respect to both I

the general trends and the relative magnitudes of the data, are within accep-table bounds. In the LOTT Test L6-3 comparison, there was good agreement in the primary side parameter 3 of the test data and the RETRAN results. Also, the over611 RETRAN model results, calculated with respect to both the general trends and the relative magnitudes of the data, are within acceptable bounds.

4.0 CONCLUSION

The staff has determined that BG&E has demonstrated the capability to properly develop RETRAN models of Calvert Cliffs and perform best-estimate simulation of reactor system non-LOCA transient using these models. Therefore, the staff concludes that BG&E is cualified to use the RETRAN models documented in the BG&E Topical Report A-85-11 for performing system transient calculations to support the evaluations of proposed changes and revisions to operating procedures.

It is noted that BG&E used RETRAN 02/ MOD 03 which was developed as a result of making corrections necessitated after the NRC staff review of the RETRAN 02/ MOD 02 computer code. This corrected version of the code has not been approved by the staff. Nevertheless, we believe that there is reasonable assurance that this version of the code was adequate for use in this topical report and we accept its use in the BG8E's best-estimate model, unless some future NRC review of RETRAN 02/ MOD 03 contradicts this assumption.

PRINCIPAL CONTRIBUTOR:

l C. Liang Dated:

__ - - - - . . . _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _