ML20059H499

Safety Evaluation Accepting Revised Responses to IEB-80-04 Re MSLB Reanalysis

ML20059H499
Person / Time
Site:	Millstone
Issue date:	01/24/1994
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20059H498	List: ML20059H493 ML20059H499
References
IEB-80-04, IEB-80-4, NUDOCS 9401280080
Download: ML20059H499 (6)

Text

-. . _ _ .

f

• coq ,

f. t UNITED STATES f M .. j NUCLEAR REGULATORY COMMISSION gk-gv [f g WASHINGTON. D.C. 20555-0001 SUPPLEMENTAL SAFETY EVALVATION 1

l H  !

THE OFFICE OF NUCLEAR REACTOR REGULATION REVISED RESPONSE TO IE BULLETIN 80-04 l

l MAIN STEAM LINE BREAK WITH CONTINVED FEEDWATER ADDITION MILLSTONE NUCLEAR POWER STATION. UNIT NO. 2 l

DOCKET NO. 50-336 j l

1.0 INTRODUCTION

l On October 7, 1982, the staff issued its safety evaluation (SE) of the Northeast Nuclear Energy Company's (the licensee) response to IE Bulletin No.

80-04, " Analysis of a PWR Main Steam Line Break with Continued Feedwater Addition," for the Millstone Nuclear Power Station, Unit No. 2. The SE j l concluded that; (1) there is no potential for containment overpressurization '

! resulting from a main steam line break (MSLB) with continued feedwater I

addition, (2) no damage would be incurred by the auxiliary feedwater (AFW) pumps due to runout, (3) all potential water sources were identified and there '

I would be no violation of specified acceptable fuel design limits, and (4) no further action regarding IE Bulletin 80-04 is required.

The staff evaluation was based on review of a MSLB analysis prepared by the licensee in 1979 and submitted to the staff on January 25, 1980. In 1991, the licensee discovered errors in the 1979 MSLB analysis which indicated that the peak containment pressure and temperature resulting from a MSLB would exceed containment design conditions and the qualification temperature of safety related equipment. The MSLB event containment response has been reanalyzed and corrective plant modifications have been implemented. On January 13, 1993, the licensee submitted a description of the revised MSLB analysis.

l The staff's findings and conclusions regarding the revised MSLB analysis are presented below.

2.0 EVALUATION OF CONTAINMENT RESPONSE Previous MSLB Analysis: In 1979, a MSLB reanalysis was prepared as a result of implementation of TMI Short-Term Recommendation 2.1.7.a to provide l automatic initiation of auxiliary feedwater upon loss of main feedwater flow.

l 9401280080 940124 PDR ADOCK 05000336 O PDR

l l

1 l

l It was also used for the purpose of responding to IE Bulletin 80-04. The 1979 analysis assumed the following scenario: l (a) Initial conditions of hot zero power, end-of-cycle, and two-loop operation, (b) A 3 minute delay from the start of the event until the start of AFW flow, l

(c) No credit for isolation of the main feedwater (MFW) system, thereby resulting in a rampdown to continuous MFW flow of 772 gpm (5% flow) to the affected steam generator (S/G),

(d) Two single-failure cases were considered - (1) AFW control valve failure with loss of offsite power and (2) loss of one train of l safeguards equipment due to a diesel generator failure, (e) The highest worth control rod sticks in the fully withdrawn position, (f) AFW flow is 2800 gpm which is 35% higher than the maximum runout flow. (With the 772 gpm MFW flow, the total FW flow is 3572 gpm), and (g) The MSLB occurs at the vessel nozzle (i.e., upstream of the flow venturi for the affected main steam line (MSL).

The licensee did not reanalyze for a spectrum of break sizes, but only the original limiting Final Safety Analysis Report (FSAR) case for the nozzle break location. The results of the analysis indicated a peak containment pressure of 47 psig and a peak containment temperature of 274*F. These conditions are bounded by the containment design conditions of 54 psig and 289'F.

The staff reviewed the 1979 MSLB reanalysis, and in a Safety Evaluation dated October 7,1982, concluded that the analysis was acceptable.

In 1991, while performing S/G replacement studies using new S/G and containment models with revised moisture carryover assumptions, and ,

considering that different single-failures are limiting for different assumed  ;

power levels, the licensee determined that the 1979 assumptions regarding ,

power level, break location, break size and single-failure were non-  !

conservative. In 1992, two new postulated single-failures were identified i that could result in MSLB conditions exceeding the containment design  !

parameters. These events, and the licensee's corrective modifications, were  !

discussed in the staff's Safety Evaluation for Amendment 167 issued on December 23, 1992.

January 13. 1993. Changes to Millstone-2 MSLB Containment Analysis:

The licensee has: (a) performed extensive reanalysis, (b) implemented plant modifications, (c) implemented revised technical specifications, and (d) revised its response to IE Bulletin 80-04. Attachment 1 of the licensee's January 13, 1993, submittal provided a description of the licensee's new IE Bulletin 80-04 response for both the core and containment.

The new MSLB analysis was performed by the NSSS vendor using the SGN-III code.

SGN-III, which is referenced for use by the Standard Review Plan, is comprised of two interdependent programs, SGN and CONTRANS. SGN calculates the transient response of the reactor coolant system and S/Gs to develop mass-energy releases for use in the containment analysis. The model consists of fluid flow and heat transfer representations of the core, hot leg plena and piping, pressurizer, S/Gs, cold leg piping and plena, and S/G secondary side.

Reverse heat transfer from the intact S/Gs to the primary coolant is included.

A two-region core fuel model is used together with a >oint-kinetics representation and non-linear moderator, Doppler, boron and control rod reactivity effects. The secondary side of each S/G is represented by a finite two-node quasi-static balance of energy flux ud secondary fluid thermodynamics. The steam lines to the turbne are included in the secondary side representation.

CONTRANS is the containment response portion of the SGN-III code. CONTRANS receives mass and energy inputs from SGN and performs mass, energy, and momentum balances on the containment volumes and flow paths. CONTRANS includes active heat sinks such as containment sprays and heat exchangers, as well as passive heat sinks such as walls.

Major assumptions utilized in the new analysis were as follows:

1. Offsite power was assumed to be available for most cases. This is conservative in that it results in increased heat transfer due to  ;

forced flow at maximum RCS flow rate. j l

Confirmatory analyses with loss-of-offsite power were run for verification that the most conservative assumption was selected.

2. For purposes of determining peak containment pressure, the initial containment conditions were assumed to be at the technical specifications limit of 16.8 psig and 120*F. This results in later (than in the standard atmosphere case) spray initiation and higher containment temperature. Relative humidity was assumed at 30% except for the EQ analyses cases where it was set to 100% to provide conservatism in the containment temperature response calculation. 1

3. Moisture carryover was as determined by the SGN-III code.

I 1

i i

i I

i 4_

l

4. Feedwater spiking due to pressure imbalance between the ruptured  !

and intact S/G was accounted for by doubling the initial feedwater flow rate for each analysis.

5. Credit was taken for the main steam nonreturn valves to prevent blowdown of the intact S/G.

6. As with the earlier analyses, AFW initiation was delayed 3 l minutes. )

7. RCP heat was included, l l

8. Actuation signals are redundant and safety grade. Consistent with  !

the original licensing basis, credit is taken in some cases for ,

nonsafety grade components initiated by safety grade signals. l

9. The maximum RCS flow rate was conservatively assumed. This  ;

maximizes the heat transfer from the primary to secondary side.

To determine the limiting cases for containment pressure and temperature responses, the licensee analyzed a spectrum of cases for sensitivity to power level and break size and location, feedwater, electrical, and containmant heat 1

removal equipment single failures, and safety-related equipment temperature  ;

response.

The licensee has used previously accepted analytical methodology for calculation of mass-energy releases and associated containment pressure-temperature responses.

Results of new MSLB analyses:

\

The results show that for all cases, the peak containment pressure is less than the containment design pressure. The highest containment pressure was for the 50% power case in which the feedwater regulating bypass valve fails-open; the calculated peak pressure was 53.7 psig. For all the cases, the peak temperature exceeded the containment design temperature for a short time period, the maximum being for the case of failure of vital bus transfer, where the peak temperature was found to be 426*F. This peak temperature is essentially identical to that of the previous MSLB analysis, discussed in the Amendment 167 safety evaluation and accepted by the staff based on supporting information provided in a licensee letter dated December 4,1992.

In an attachment to the licensee's December 4, 1992 submittal, entitled,

" Evaluation of EQ Equipment Thermal Response to the Postulated MSLB and LOCA in the Millstone Unit 2 Containment," the licensee summarized its environmental qualification analysis. The analysis resulted in conditions I exceeding containment design temperature for a short period of time. The l licensee determined that a condensate layer would form on the surface of equipment that would act as a thermal barrier. The analysis showed that the temperature peak represents a short duration at superheated conditions and the l

I

surface temperature of safety related equipment will not exceed the saturated temperature of the steam pressure in containment during the event. The licensee concluded that equipment surface temperature would not exceed qualification temperature during a MSLB although the containment atmosphere temperature rises to a maximum of 426*F. The surface temperature of qualified equipment would not exceed its qualification temperature of 289'F. This method of analysis is described in NUREG-0588, Section 1.2(5)(b) as being acceptable when the containment atmosphere exceeds the qualification temperature. Therefore, the staff agrees that qualified equipment can be expected to perform its safety function during and after a MSLB.

3.0 EVALUATION OF CORE RESPONSE Item 2 of IE Bulletin 80-04 requested that the licensee review the core reactivity analysis for a main steamline break (MSLB) regarding runout from the auxiliary feedwater system or continuation of feedwater or condensate flow to the affected steam generator following isolation of the main steam system.

The original response to IE Bulletin 80-04 for Millstone 2 did not consider all potential water sources. The revised responses are discussed below.

a. The licensee reviewed the bounding conditions for the analysis and found that the existing analysis bounds the consequences of continued feedwater flow to the break.

b. The single failure assumed in the analysis of record is the loss of one of two HPSI pumps. The effect of this decreased delivery of high concentration boric acid solution to the reactor coolant system is that return-to-power is exacerbated. During the review the licensee identified an additional single failure that could impact core response. The failure of electrical bus 24G during the fast transfer from the normal to the reserve station service transformer. The licensee determined that this event is bounded by the analysis of record.

c. The licensee identified the primary effect of extended water supply to the affected steam generator as an increase in the duration of the event with a higher potential for return-to-power.

During the 1992 refueling outage, changes were made to the feedline isolation system to ensure redundant valve alignment, and also to the power supply to the feedwater regulating valves. The analysis of record includes the effect of runout flow for a duration of 180 seconds. The implemented changes would result in a less severe reactivity transient compared to the bounding case for the analysis of record.

d. The analysis of record indicates that the hot channel factors corresponding to the most reactive rod in the fully withdrawn position and Minimum Departure from Nucleate Boiling Ratio (MDNBR) values for the analyzed events bound the MSLB analysis for core response. The analysis predicts a slight return to power with no fuel failure.

em-v

The licensee evaluated the MSLB analysis and found that it bounds the core '

reactivity response in the event of an extended feedwater supply. The plant modifications during the 1991 refueling outage, resulted in the additional single active failure being less limiting than the analysis of record.

The staff has reviewed the licensee's evaluation and finds that the revised response to IE Bulletin 80-04 adequately demonstrates that the January 13, >

1993 MSLB analysis of record bounds the case of continued feedwater flow.

4.0 CONCLUSION

The licensee has provided a revised response to IE Bulletin 80-04. The '

updated response indicates that; (1) acceptable methodology has been used for  :

the revised MSLB analysis and that the results of the analysis indicate that  :'

plant systems, structures and components in containment will not be exposed to unacceptable pressure or temperature conditions resulting from a design basis MSLB, and (2) the core response to the reanalyzed MSLB is no more severe than shown by previous analysis and is also acceptable.  ;

Principal contributors: W. Long S. Brewer A. Dummer Dated: January 24, 1994 I

I e

t k

h

- - , - . .