ML20058G258

Submits Changes to & Errors Discovered in Emergency Core Cooling Sys Evaluation Models for Units 1 & 2.Brief Summary of Changes Provided W/Addl Details & Peak Cladding Temp Margin Summary Sheets in Attachment

ML20058G258
Person / Time
Site:	Prairie Island
Issue date:	12/03/1993
From:	Richard Anderson NORTHERN STATES POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9312090152
Download: ML20058G258 (6)

Text

~

Northem States Power Company 414 Nicollet Mall Minnea,*s Minnesota 55401 1927 Telephone (612) 330 5500 December 3, 1993 10 CFR Part 56 Section 50.46 U S Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555 PRAIRIE ISLAND NUCLEAR GENERATING PLANT Docket Nos. 50-282 License Nos. DPR-42 50-306 DPR-60 Small Break LOCA Analysis Errors This letter is submitted in accordance with 10CFR50.46(a)(3)(ii) to report changes to, and errors discovered in, emergency core cooling systems evaluation models for Prairie Island Nuclear Generating Plant Units 1 and 2.

The model changes and errors discovered affect the small break loss of coolant accident (SBLOCA) model and are considered reportable under the 30-day reporting requirement. A brief summary of the changes is provided below with additional details and peak cladding temperature (PCT) margin summary sheets in the attachment.

The effect of safety injection (SI) in the broken loop of the reactor coolant system (RCS) during a SBLOCA modeled by NOTRUMP has been assessed a PCT penalty of 150 degrees F. Broken loop SI spilling to containment is an assumption historically judged to be conservative due to the loss of SI inventory. Delivery cf SI to the cold leg has been found to be_ a penalty in the NOTRUMP evaluation model because competition between steam venting out the break and SI entering the break increases the RCS pressure. Higher RCS pressure reduces the amount of SI delivered by the centrifugal SI pumps which increases the PCT.

The effect of an improved condensation model in NOTRUMP has been assessed a PCT benefit of 150 degrees F. Condensation reduces the RCS pressure and increases the amount of SI delivered by the centrifugal SI pumps. A comparison with test data shows that the PCT benefit from the improved condensation model more than offsets the PCT penalty due to SI in the broken loop.

Errors in the coding of the drift flux flow regime map caused discontinuities in the map under certain circumstances. The errors have been corrected and a PCT benefit of 13 degrees F has been assigned, though the effect of the changes may exceed the reportability threshold of 50 degrees F.

I 9312090152 931203 I t

PDR ADOCK 05000282 '

P PDR I

\

i Nodhem States Power Company i

'US NRC [

December 3, 1993 l Page 2 ,

We have made no new NRC commitments in this letter.

These issues are discussed in greater detail in the attached excerpts from l Westinghouse correspondence to us. The table (page 4 of the attachment)  ;

refers to other documents which we have not attached and other potential  :

errors which we have not discussed because their significance has not been f determined yet. If you need any of that additional information or have any. i questions related to this letter, please contact Jack Leveille (612-388-1121,  !

Ext. 4662).

I i

jh hf b Ro er O Anderson 4 Director (

Licensing and Management Issues  ;

l i

i c: Regional Administrator - Region III, NRC Senior Resident Inspector, NRC NRR Project Manager, NRC i J E Silberg j

Attachment:

Westinghouse summaries of the SBLOCA analysis errors (4 pages) i I

i l

J l

1 l

l f I h b b a

l

.. - - . - . ~ , . - .. <- , - , - , , , -,

d TECIINICAL DESCRIPTION

[

ISSUE DESCRIPTION Westinghouse emergency core cooling system (ECCS) evaluation models are considered to be composed of several features which include underlying assumptions. Westinghouse recently completed an evaluation of a potential issue concerning the modeling of Safety Injection (SI) flow into the broken RCS loop for small break loss of coolant accident (SBLOCA). Westinghouse previously assumed that SI to the broken RCS loop would result in a lower calculated PCT and, therefore, modeled the ECCS broken loop branch line to spill the SI to the containment sump. The basis for this assumption included consideration for the effect of back pressure on the spilling ECCS line for cold leg breaks, which would see a higher back pressure for SI connected to the broken RCS loop when compared to spilling against containment back pressure. Spilling to the higher RCS pressure would increase SI to the intact loops, which is a benefit for PCT. The effect on intact loop SI flow rates as well as the assumption that some of the SI to the broken loop would aid in RCS/ Core recovery resulted in the Westinghouse ECCS model assumption that Si to the broken loop was a benefit. However, when SI is modeled to enter into the broken loop, a significant PCT penalty is calculated by the NOTRUMP small break evaluation model (approximately 150 degrees F for a typical Westinghouse 3-loop design).

TECIINICAL EVALUATION An analysis by Westinghouse indicates that the penalty (as described above) occurs as a result of '

competition between the steam venting out the break and the SI to the broken loop, which also exits through the break. The competition between the steam and the SI results in higher RCS pressures for the identical core steammg rates. Sirce the ECCS uses centrifugal pumps, higher RCS pressure results in lower delivered SI flow rates to the intact RCS loops, leading to the calculated PCT penalty. His -

penalty is somewhat aggravated by the use of the Moody two-phase break flow model, which is a thermal equilibrium model being used to model a clearly nonequilibrium process. However, the penalty is large i enough such that a change to a nonequilibrium break flow model would not be expected to offset the break flow RCS pressure interaction seen when SI is assumed to enter into the broken loop. ,

However, when a newer conservative model based on prototypic test is used which modeled the configuration of the SI piping to the RCS cold leg in a Westinghouse designed PWR, a net PCT benefit is calculated. Improved condensation of the loop steam in the intact loops results in lower RCS pressure and larger SI flow rates. The increase in SI flow rates, due to lower RCS pressure, leads to'the lower '

calculated PCT. Thus, the negative effects of SI into the broken loop can be offset by an improved SI condensation model in the intact RCS loops.  :

~

The improved condensation model is based on data obtained from the COSI test facility, ne COSI test facility is a 1/100 scale representation of the cold leg and SI injection ports in a Westinghouse designed PWR. The COSI tests demonstrated that the current NOTRUMP condensation model under-predicted '

condensation in the intact loops during SI and thus is a conservative model. Use of the improved condensation model has demonstrated that the current NOTRUMP small break LOCA analyses without the improved condensation model and no SI into the broken loop is more conservative (higher calculated PCT) than a case which includes S1 into the broken loop and the improved condensation model.

ATTACHMENT PAGE 1 of 4

l Additionally, the effects of Si in the broken loop have been dearmined to not change RCP trip symptom developed in response to US-NRC Generic Letters83-10C and 85-12 or SI termination criteria found in I the Westinghouse Owners Group Emergency Response Guidelines.

ASSESSMENT OF SAFETY SIGNIFICANCE

'Ibe COSI tests demonstrated that the current NOTRUMP condensation model under-predicted condensation in the intact loops during SI and thus is a conservative model. Furthermore, recent evaluations have shown that the current NOTRUMP small break LOCA analyses without the improved condensation model and no SI into the broken loop is more conservadve (higher calculated PCT) than a case which includes SI into the broken loop and the improved condensation model. Based on these evaluations, Westinghouse determined that this issue does not involve a Substantial Safety Hazard as defined i110 CFR Part 21. Reanalyses are not necessary since current NOTRUMP based small break LOCA analyses have a conservatively calculated PCT and, therefore, remain valid.

t l

l i

l l

l 2

l ATTACILMENT PAGE 2 of 4

1 NOTRUMP DRIFT FLUX FLOW REGIME MAP ERRORS  :

Backcround Errors were discovered in both WCAP-10079-P-A and related coding in NOTRUMP SUBROUTINE DFCORRS where the improved TRAC-P1 vertical flow regime map is evaluated. In Evaluation Model  :

applications, this model is only used during counter-current flow conditions in vertical flow links. He  !

affected equation in WCAP-10079-P-A is Equation G'65 which previously allowed for unbounded values of the parameter C. contrary to the intent of the original source of this equation. His allowed a discontinuity to exist in the flow regime map under some circumstances. This was corrected by placing  !

an upper limit of 1.3926 on the parameter C. as reasoned from the discussion in the original source.  ;

As stated, this correction returned NOTRUMP to consistency with the original source for the affected equation.

Further investigation of the DFCGG5 uncovered an additional closely related logic error which led to discontinuities under certain other circumstances. This error was also corrected and returned the coding to consistency with WCAP-10079-P-A.

f This was determined to be a Nondiscretionary Change as described in Section 4.1.2 of WCAP-13451 l and was corrected in accordance with Section 4.1.3 of WCAP-13451. ,

i Affected Evaluation Models i

1985 Small Break LOCA Evaluation Model  !

\

Estimated Effect i

Representative plant calculations indicated PCT effects ranging from -13 degrees to -55 degrees. For the purpses of tracking PCT, the minimum benefit of -13 degrees has been assigned to these changes.

W hen considering reportability under 10 CFR 50.46(a)(3)(i), however, it has been demonstrated that the effect of these changes may exceed 50 degrees F. Westinghouse, therefore, recommends that these changes be considered significant with respect to 10 CFR 50.46(a)(3)(i) requirements.

~

ATTACHMENT PAGE 3 of 4 i

l 1

l

r

.i; S ma1LRtea khaM'la dle mpetAturg.hia rdal!!iliwion _

I Revision Date: 893  !

i Plant Name: Prairie Island Units 1 and 2 Eval.Model: NOTRUMP Fuel: 14X14 OFA I Utility Name: Northern States Power FO= 2.50 FAH= 1.70 SGTP= 10% l t

Reference

• Clad Temperature Notes l A. ANALYSIS OF RECORD (12/91) 1 PCT = 1077 *F i t

B. PRIOR PERMANENT ECCS MODEL ASSESSMENTS 2 APCT= 0 *F l

, C.10 CFR 50.59 SAIETY EVALUATIONS Table A APCT= 0 *F  !

D.199310 CFR 50.46 MODEL ASSESSMENTS  ;

(Permanent Assessment of PCTMargin)  ;

1. Effect of S1 in Broken Loop APCT= 150 *F  !

2. Effect of Improved Condensation Model APCT= -150 *F t

3. Drift Flux Flow Regime Errors APCT= -13 'F i E. TEMPORARY ECCS MODELISSUES"

l. Effect of Leaking Double Disk Gate Valves APCT=

(

3 0 *F 1 j l

F. OTi!ER MARGIN ALLOCATIONS

1. None APCr= 0 *F  !

i LICENSING BASIS PCT + MARGIN ALLOCATIONS PCT = 1064 *F i

• References for the Peak Clad Temperature Margin Utilization summary can be found in Table B.

" It is recommended that these temporary PCFallocations which address cunent LOCA modelissues not be f considered with respect to 10 CFR 50.46 reporting requirements. j i

Notes:

1. A temporary penalty of 0 *F will be assessed for this issue until the exact configuration of the Prairie Island Units I and 2 double disk gate valves has been verified.

l i

l' l

t 2

i i

d y i

v 1 ',

3 ATTACHMENT l

PAGE 4 of 4

< l

. )

,, -- __l