3F0596-22, Informs NRC That FPC Completed New SBLOCA Analyses in Support of Efforts Re Improved HPI Instrumentation & Change as Stated in PCT in Accordance w/10CFR50.46 Requirements

Informs NRC That FPC Completed New SBLOCA Analyses in Support of Efforts Re Improved HPI Instrumentation & Change as Stated in PCT in Accordance w/10CFR50.46 Requirements

ML20117H503
Person / Time
Site:	Crystal River
Issue date:	05/22/1996
From:	Beard P FLORIDA POWER CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
3F0596-22, 3F596-22, NUDOCS 9605280232
Download: ML20117H503 (5)

Text

i Fl rida E933Y

• = =.:

May 22,1996 3F0596-22 U.S. Nuclear Regulatory Commission Attn: Document Control Desk Washington DC 20555-0001

Subject:

New Small Break Loss-of-Coolant Accident (SBLOCA) Analyses

References:

BAW-1976A & FTl 51-1245866-00

Dear Sir:

The purpose of this correspondence is to inform the Nuclear Regulatory Commission (NRC) that Florida Power Corporation (FPC) has completed new small break loss of coolant accident (SBLOCA) analyses in support of efforts related to improved high j

pressure injection (HPI) instrumentation and is reporting a change of 2: 50 F in peak clad temperature (PCT) in accordance with 10 CFR 50.46 requirements.

FPC contracted Framatome Technologies incorporated (FTI) to evaluate the new instrumentation configuration in terms of the HPl system capabilities. A preliminary evaluation concluded, given instrument uncertainty, assumed HPl pump degradation and overall assumed system configuration differences from previous analyses, that the j

HPl system performance was not consistent with the existing Emergency Core Cooling System (ECCS) analyses. The projected HPl flow deficiencies would require new ECCS analyses to demonstrate compliance with 10 CFR 50.46 requirements.

BACKGROUND On February 16,1996 FPC shutdown Crystal River Unit 3 (CR-3) because it was determined that operators might not be able to mitigate an HPI line break SBLOCA given coincident failure of a safety related DC train and a loss of offsite power (LOOP). Although the probability for such a condition is extremely small (Core Damage Frequency of 9.63E-15/ year) FPC concluded that a plant shutdown was warranted. FPC determined that an additional narrow range flow instrument would 9605280232 960522

/'

l PDR ADOCK 05000302 p

PDR

(

2 8 0 0 MTa' aivea raeaav cour'ex: ' > a-

- u

.a -

- >a a na. nr c.w.w l

U. S. Nuclsar Rsgulatory Commission 3F0596-22 Page 2 of 5 be added to each HPlline with power supplied from the redundant DC train. This new 3

instrument configuration would tolerate the proposed single failure of a battery or i

related DC train.

FTl was contracted to assure the new instrumentation would in fact be adequate for all postulated SBLOCA scenarios. As FTl began its evaluation, preliminary results indicated the existing HPl system configuration would not meet the minimum ECCS flow delivery to the core used in the previous SBLOCA analyses.

SBLOCA ECCS ANALYSES Previous SBLOCA analyses did not bound the actual CR-3 HPl flow split that occurs when the break is assumed in the cold leg pump discharge (CLPD) where HPI and normal makeup penetrate the RCS piping (see Figure-1). A 50/50 flow split was modeled which assumed that 50% of the HPl flow would enter intact CLPD piping with the other 50% entering the broken CLPD pipe before exiting the break during the first 10 minutes. The actual HPl flow split is 64/36, with 64% of the HPl flow exiting the break, in addition, reactor coolant pump (RCP) seal injection was incorrectly assumed to automatically isolate when HPl actuates. RCP s al injection is manually isolated within 20 minutes into the event. In addition, emergency feedwater (EFW) flow rate used in the analyses was higher than what would actually be supplied to the steam generators.

New SBLOCA analyses were performed using two evaluation models (EM), with the HPI system evaluated using a CR-3 specific hydraulic model. CRAFT 2 is the EM code of license and was used to evaluate the worst-case SBLOCAs. RELAP5/ MOD 2 was used as a screening / sensitivity tool, because of its faster speed, to determine what cases would be run using CRAFT 2.

The CR-3 HPl delivery to the core can be characterized in three time periods:

1) 0 - 10 minutes with 1 HPI pump supplying 2 injection lines,

2) 10 - 20 minutes with 1 HPl pump supplying 4 injection lines, 3)

> 20 minutes normal makeup and RCP seal injection isolated.

The above configurations are accomplished by operator action as follows:

Ensure all available HPl lines are open within 10 minutes, Isolate normal makeup and RCP sealinjection within 20 minutes, If a failed HPl line is indicated (highest line 75 gpm > lowest line using the narrow range HPl flow instruments), then isolate the high (broken) line.

After 20 minutes the HPI system delivers adequate flow to the core for any CLPD break size or location. Prior to 20 minutes the CR-3 system meets the minimum flow requirements for breaks in any cold leg except the A-1 leg. The flow deficit for the A-1 leg is largest for the first 10 minutes. Five most limiting CLPD breaks (0.07,0.1, 2

0.125,0.15 & 0.2ft ) were analyzed with RELAP5.The HPl line break with the most 2

limiting pinch area was also analyzed. The 0.125ft CLPD break defined the most limiting case. Using CRAFT 2, FOAM and THETA 1-B to evaluate this case, the results

U. S. Nuclear Regulatory Commission 3F0596-22 i ige 3 of 5 I

l dicated core uncovering (4 feet of the upper core uncovered at 730 seconds into the l

e<ent) with a peak clad temperature (PCT) of 1859 F.

All other cases indicated higher minimum core levels and lower PCTs. Previous analyses indicated that the core remained covered with PCTs < 1000*F. Nevertheless, this new limiting case PCT l

of 1859 F is less than the 10 CFR 50.46 criteria of 2200 F.

l Lower core mixture levels and elevated PCTs are the result of several differences l

between previous analyses and the analyses recently performed.

BAW-1976A i

documented SBLOCA analyses that were considered bounding for CR-3. Some of the assumptions used in those analyses were not reflective of actual CR-3 HPl system and emergency feedwater (EFW) system performance. FPC and FTl are continuing to j

evaluate why the previous values were used. FPC will forward the results to the NRC.

The specific nature of the differences are described below:

HPl system performance:

i As described earlier, an assumed 50/50 flow split was not representative of CR-3 l

configuration.

Isolation of normal makeup and RCP seal injection was not l

consistent with CR-3 operation. In addition, HPl pump degradation was not previously modeled. The overall effect results in a lower core liquid level.

l EFW system performance:

l Previous analyses assumed EFW flow of 570 gpm to each steam generator.

l Actual EFW flow is 200 gpm to each steam generator or 400 gpm to one. This difference accounts for the most significant inventory reduction in the core region.

The effect high EFW has on fluid volume in the core area is over-predicted in CRAFT 2. The prediction is primarily a function of the coarse modeling detail in the reactor vessel upper plenum region using the CRAFT 2 EM.

This was concluded by comparing RELAP5 cases varying EFW flow rate with an insignificant change in PCT.

I RELAP5 and CRAFT 2 predicted similar core mixture levels, however, when CRAFT 2 indicates core uncovering then the FOAM 2 and THETA 1-B codes are used to determine mixture level swell and PCT. This methoti is very conservative for PCT prediction. High EFW flow rates cause an over-prediction of core level using the CRAFT 2 code. This was determined through comparative analysis of the many cases run using RELAP5/ MOD 2 and CRAFT 2. It was also determined that the change in core mixture level due to EFW flow variations, using CRAFT 2, is over-exaggerated.

l This sensitivity is primarily attributed to a lack of modeling detailin the upper plenum l

area. Therefore, FPC through the B&W Owners Group will be pursuing the use of RELAPS/ MOD 2 as the licensed EM in the future.

i

U. S. Nuclear Regulatory Commission 3F0596-22 l

Page 4 of 5 l

SUMMARY

FPC contracted FTl to evaluate new HPl low flow instrumentation in terms of overall ECCS performance. This evaluation uncovered discrepancies in previous analyses indicating the need for new analyses to demonstrate compliance with 10 CFR 50.46.

The results of new SBLOCA analyses indicate elevated PCTs with some degree of core uncovery for some CLPD breaks. No core uncovery occurs for HPl line breaks.

in all cases the ECCS acceptance criteria of 10 CFR 50.46 were met.

i i

Sincerely, l

. Beard, Jr.

Senior Vice President Nuclear Operations PMB:PVF xc: Regional Administrator, Region II Senior Resident Inspector i

NRR Project Manager

\\

l

i t

1 l

U. S. Nucla::r R:gulatory Commission i*

3F0596-22 Page 5 of 5 l

Q Vessel Steam Generator Romer V

)

L_7 casas C__,, 7 i

< '3 Romer (3

v Cadas v

)

%sent B-2 A-2 Wong nest A1

% toggig a~

W f

B-1 LLL

\\

Q/

O MUV-23 MUV-24 MUV.25 MUV-26 MUV-31 MUV-16 l

MUV-27 MUV-18 MUV-9 MUV-3 MUP-IA MUP 18

)

Y k_.

k_.

MUV-58 MUV-73 e BWST l

Figure-1 l

i