ML20199G259

Safety Evaluation Rept Accepting Revised SG Tube Rupture Analysis

ML20199G259
Person / Time
Site:	Byron, Braidwood
Issue date:	01/28/1998
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20199G211	List: ML20199G206 ML20199G259
References
NUDOCS 9802040247
Download: ML20199G259 (13)

Text

.

.. (.

l l

l l

SAFETY EVALUATION BY TliE OFFICE OF NUCLEAR REACTOR REGULATION

REVIDED STEAM GENERATOR TUBE RUPTURE ANALYSIS ,

i l

) COMMONWEALTH EDISON COMPANY .

! BYRON STATION. UNITS 1 AND 2 AND BRAIDWOOD STATION. UNITS 1 AND 2 i

j DOCKET NOS. STN 50 454. STN 50-455. STN 50456. AND STN 50457 i

1.0 INTRODUCTION

l

} Commonwealth Edison Company (the licensee) will be replacing the Westinghouse Model D 4

steam generators with Babcod & Wilcox Intemational steam generators at Byron, Unit 1 (fall of i

1997) and Braidwood Unit 1 (fall of 1998). The differences between thudginal steam generators (OSGs) and the replacement steam generators (RSGs) affet.: the analysis of the j steam 9enerator tube rupture (SGTR). As a result, the SGTR accident was re-analyzed.

I The analysis of the SGTR event is used to verify that the dose consequences are acceptable.

Following a tube rupture event at Ginna in 1982, where the steam generator (SG) filled and caused the code safety relief valves to lift and relieve water rather than steam, the tube rupture is i

also analyzed to verify that the SGs will not overfill as a result of a tube rupture event.

By letter dated November 13,1996, the licensee submitted the revised SGTR analyses for the i RSGs. Additional information was provided by letters dated March 20, June 24 August 19, November 3, November 26, and December 19,1997.

2.0 EVALUATION a

2.1 Operator Response Times 2,1.1 Backgrouno in response to a staff letter of March 30,1987 (Reference 1), Commonwealth Edison Company's i letter dated April 25,1990 (Reference 2), transmitted the plant specife analysit required by the

NRC to resolve the SGTR licensing lasues for the Byron /Braidwood plants. The staffs safety i evaluation dated April 23,1992, concluded that the licensee's demonstrated times for Byron /Braidwood operators, and its commitment to condect additional demonstrations, were

satisfactory, By letter dated July 6,1993, the licensee documented that the additional

- demonstrations had been complete 1 d

- 2.1.2 Evaluation 1

[ The staffs evA;ustion of the Westinghouse Owners Group WCAP-10698 (Reference 1) stipulates 4

four plant specific criteria for assessing c*perator response times in the event of an SGTR.

i

! 9902040247 990128

! PDR ADOCK 05000454 1

P PDR i_ __.- _ __ __.____. .______

t 1.

2 Those criteria were &TW,d to evaluate the information sent by Commonwealth Edison Company on operator response times during an SGTR at the Byron /Braidwood plants. The staffs evaluation by orHerion follows.

CrHerion 1. Provie simulator and emergency opereung procedure training related to a potential BOTR.

By letters dated November 13,1996, and March 20,1997, the licensee documented that onsite simulator and emergency operating procedures (EOP) training relevant to an SGTR are provided.

The staff finds that the licensee has satisfied CrHerlon 1.

CrHerion 2. Utilizing typloal control room staff as participants in demonstrations, show that the operator response times assumed in the SGTR analysis are realistic and achievable.

By letters dated November 13,1996, March 20,1997, and August 19,1997, the licensee submitted the assumed and demonstrated operator response times for the overfill scenario that follows:

I OVERFILL SCENARIO OPERATOR ACTIONS Assumed Time Demonstrated Time (in minutes) (in minutos)

Isolate AFW to ruptured SG 11.00 7.17 _

initiate RCS cooldown i8.00 18.07 _

inRiate RCS depressurization 2.00 1.65 ___

Terminate ECCS flow 2.00 1.87 __

Establish charging 2.00 2.02 _

Establish letdown 3.00 3.02 _ ,_

Roopen pressurtzer PORV 4.00 3.95 TOTAL 42.00 37.75 The licensee's letter of November 13,1996, indicated that two licensed reactor operator crews from Byron and two licensed reactor operator crews from Braidwood were evaluated for the overfill scenario. This evaluation yield 6d the operator response times reported in the table above. In addition, the licensee committed by letter o' Mamh 20,1997, to complete demonstrations on the overfill scenario for a minimum ai 60 percent of Byron /Braidwood licensed operator simulator crews by Dooomber 31,1997.

The demonstrated times for threw operator actions -initiate RCS cooldown, establish charging, and establish letdown -were not bounded by the licensee's assumed times. The licensee's letter of March 20,1997, explained that (1) the average rate of increase in ruptured SG fill rato does not vary significantly after auxiliary feedwater (AFW) to a ,vptured SG is isolated, (2) the final transient overfill results are dependent only on the total mitigation time required, and (3) the composite effect of three operator actions that exceeded the licensee's assumed time was a -

reduction in margin to overfill of 1.6 ft'.

The licensee also transmitted by letter dated March 20,1997, the operator response times for four crews. All of the crews had times that were bounded by the time assumed in the licensee's

.. I. .

3-analysis for isolating the ruptured SG One crew had five operator response times that exceeded thi licensee's assumed times as follows: total response time (31.67 minutes versus 31.00 minutes), initiate depressurization (2.25 minutes versus 2.00 minutes), terminate emergency core cooling system (ECCS) flow (3.43 mir*.utes versus 2.00 minutes), establish letdown (3.42 minutes ,

s versus 3.00 minutes), and reopen pressurizer power-operated relief valve (PORV) (4.72 minutes 4 versus 4.00 minutes). However, in a letter dated Aug .*st 19,1997, the licensee's sensitivity analysis, using all tne subject crew's response times, showed an increase in margin to overfill of 603 ft' above that in the assumed base analysis. The staff considered the differences between the assumed and demonstrated times under discussion to be acceptable because (1) the differences were relatively small and (2) the licensee's overall evaluation indicated that the margin to overfill would be maintained for 11 four crews.

As sieted previously, the licensee committed to complete demonstrations on the overfill scenario for a minimum of 80 percent of the Byron and Braidwood licensed operator crews. By letter dated December 19,1997, the licensee str'.ed that all operating crews at both Byron and Braidwood have receied the appropriate tralning and have successfully met the acceptanct -

cntmion. On the basis of the results of the licensee's sensitivity study, information in the al-tr.ble, explanation of operator response times, anu successful testing of the ope ating cres

• slaff f'ods that the licensee has satisfactorily addressed Criterion 2.

Qitaini 3. Complete demonstrations to show thct the postulated SGTR accide: t can be mitigated within a period of time comi.,atible with o,erfill prevention, using des:Jn-basis

.un..nptions iegarding available equipment and its impact on operator response timos, t hv W.ua demomtrated res;)omo times for the ovsfill scenario are consistent with the carator rmponse times assumed in the licensee's analysis, except for the response times to

..Jinta rr +: lor coolant us tm (RCS) cooldown, establish charging, and establish letdown. As pmice, e discv. sed, the I;censco's sansitivity study indicates that with these exceptions rnargin

'n 80 NMi'l would be maintained. Therefoie, the staff concJudes that Criterion 3 is satisfied.

Gitmian 4 If the emergency i>pe uting procedures (EOPs) specary SG sampling as a means of identify',ng the SG with the ruptured tube, provide the expected time period for obtaining the une roeutts and di= cuss the affect on the durmion of the accident.

The lic.ensee's letter dated March 20,1997, indicated ttti .ompling would take 3 to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> for

, fvut SGs and less thaq 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> for a single SG. Further, the licensee conveyed that (1) sampling is needed only when SG nt row range leve' indications are not adequate, (2) narrow ango Mvel instruments are Class 1E, and (3) tube rupures where narrow range level ,ndication M not suf6cient would invola sma!! leaks and would r,ot present an overfill cor>cem. On the M : of th4 infomstion, the staff finds thet Critorion 4 is satisfied.

A u G,nimary The staff has reviewed Commonwealth Edison Companf. responses regarding operator action times during an SGTR and concluded that the liceme(s demmstrated times for Byron /Braidwood operators are satisfactory.

- - - - i-m __ m

, 2.2 Marnin to Overfill 2.2.1 Background The initial revised analysis for the new RSGs resulted in a reduef!on in the margin to SG overfill.

Consequently, in order to maintain soceptable margin, certalri operator actions were revised.

The step sequences were revised to require the operators to act sooner by isolating AFW and taking other mitigating actions earlier in the everd. 'Ihe miugating actions taken earlier increase the margin to SG overfill. The aanlysis for both Units 1 and 2 for both sites were redone because of the new operator actions.

The original analysis of the SGTR and the margin to overfill wa's performed by the licensee and approved by the NRC (Reference 3). The analysis was y orformed using the NRC-aporoved methodology, WCAP 10689-P-A, "SGTR Analysis Methodology to Determine Margin to Steam Generator Overfill." There are changes to the analysis techniques that have also been made; however, the analysis was based on the same methodology as the original analysis.

Tne differences between the OSGs and the RSGs are summarized in the November 13,1996, submittal and result in a reduction in the postulated break flow because the RSGs tube dnmeter is smaller. The heat transfer surface area is increased with the RSGs and results, for the same

- average temperature, in the main steam presbure being higher. These factors tend to incesse the margin to overfill by reducing the primary to secondary flow; however, the RSGs have less secondary side volume becaun tnere are more moisture separators ir, the generators. As a result the margin to overfill for the RSGs is reduced.

2.2.2 Evaluation The re analysis was performed for both OSGs and RSGs because the new operator actions affect both Unit 1'a with the RSGs and Unit 2's with the OSGs. The analysis was performed with separate calculations to both minimize the margin to SG overfill and to maximize the release of '

radiation. The re-analysis of the SGTR was performed using RETRAN-02 MOD 5.1. This is an updated form the original methodology which used RETRAN-C1 MOD 3; however, the licensee

- described the changes to the analysis and verified that the inputs are conservative. The .

methodology, RETRAN-02 MOD 5.1, used for this analysis was approved (Reference 4) by the NRC. The licensee has verified that the application of the methodology continues to be performed in accordance wV., *ach of the limitations and restrictions associated with the NRC safety evcluation. As a row, the methodology is appropriate for use in this tube rup*ure analysis application. The onginal analysis of the SGTR and the margin to overfill was performed by the licenwe and approved by the NRC (Reference 3). The ane!ysis was performed using the NRC- approved methodology, WCAP-10689-P-A. There are changes to the analysis techniques; however, the analysis used for this submittal was based on the same methodology as the original analysis. The use of this methodology is acceptable to analyze the tu'oe rupture event.

The licensee has used a conservative. 'et of initial conditions and assumptions tha; include the input values for the RCS average temperature, reactor power, S3 pressure, SG level, and no assumed delays in AFW flow. In addition, conservative instrument and setpoint uncertainties are used. Decay heat was assumed to bs 120 percent nf the ig71 ANS standard. The analysis credits only equipment and instrumentation that is safety related and the new procedures did not

include equipment or instrumentatiori that was not already used for the original analysis.

l l

, - - r - ~ L

a' , '.

. Addnionally, a number of single failures were evaluated to determine the most limiting failures.

The failures considered included ine failure of an atmospheric dump valve (ADV), the failure of an main steam isolation valve (MSIV), and the failure of AFW control valve opened or closed.

The most limiting single failure for the margin to overfill cases for both the OSG and RSG cases was the failure of an ADV to open in one of the intact SGs. This reduces the SG coohng avmilable to the operators and slows the cooldown process. The limiting single failure for the offsite dose cases is the failure of an ADV in the open position in the faulted SG, This maximizes the amount of the release to the endronment and the offsite doses, A: nber of cases were run to determine the most limiting set of conditions The analysis was run for both the OSGs and RSGs. The analys's was run assuming the reactor tnpped on low '

RCS pressure and ca overtemperature delta-temperature (OTDT). For the low pressurizer pressure case a runheck to 30 percent wes assumed and for the OTDT case a runback to 95 <

percent was assumed. These assurr.ptions are conservative and maximize secondary inventory '

because it delays the reactor trip and causes an increased SG level at lower turbine inads. The analysis was run with single failures to minimize the margin-to-overfill and to maximize the dose consequences. The resalts indicate that for the OSG cases the low pressure case resumed in a margin to overfill of 230 ft*, and 2g3 ft* for the OTDT case. For the RSGs the margin to 3G overfill is 226 ft8 for the low pressurizer case and 60 ft8 for the OTDT case.

The emailest margin to overfill for all the calculated results is 60 ft'. Relative to the size of the SGs, this is not a significant amount of margin and it is a reduction frem 'he original analysis, However, the conservatisms described in the analysis assumptions ar.d snitial conditions have justified that a snargin of 60 ft8 is acceptable.

2.2.3 Summary The staff has reviewed the licensee submittal relating to the SGTR and found that the analysis presented is performed using an approved methodology with appmpilately ceservative inputs and assumptions. The use of the updated version of RETRAN is ccnsis'ent with the original use of RETRAN and the approved SGTR methodalogy An appropriate number of cases were evaluated and the results presented are acceptable for licensing appi;cstions. 'thre is acceptable margin to SG overfill, and the use of this analysis for inpm into the dose consequence analysis is also acceptable. As a resv't, based on the above discussion, the staff finds the revised analysis acceptable.

2.3 Radioloalcal Analysis 2.3.1Backaround The change in SG design affects the manner in which the reactoi responds in the event of certain accidents. This change in response can impact the radiological releases, thus affectir.g offsite and oasite radiological doses.

2.3.2 Evaluation The staff evaluated the consequences of a postulated SGTR accident. Two cases were aria;yzed. One case involved a p.m-existing iodine snike and the other case, an accident-initiated spike. For the pre-existing spike case, the reactor coolant iodine activity level of dose equivalent '8'l (iodine-131) was assumed to be at the full power level of 60 pCilgm of dore

, equivalent *l in Technical Specification (TS) Figure 3.41. The seconoary coolant lodine specific activity was based on the TS normal operation limit of 0.1 pC#gm dose equivalent *l.

Th a accident-initiated spike case assuraed the SGTR event itself initiated a concurrent lodine spike. The reactor coolant was assumed to be at the TS 46 hour5.324074e-4 days <br />0.0128 hours <br />7.60582e-5 weeks <br />1.7503e-5 months <br /> rea . tor coolant activity level of dose equivalent *l of 1 pCi/gm. The secondary system activity was assumed to be at the TS limit of 0.1 pCilgm dose equivalent *l. The SGTR is assumed to initiate an iodine spike which results in a release of lodine from the fuel gap to the reactor coolant at a rate in C# unit time which is 500 times the normal iodine release rate necessary to maintain primary coolant at 1pci/g. The licensee indicated that the SGTR did not result in any failed fusi.

For both analyses it was essumed that a primary to secondary leak occurred in the intact SGs at l.

a rate of 150 gallons per day (gpd) for the duration of the accident it was also assumed for these two analyses that offsite power was lost anc the main condenser was unavailable as a l cource to remove decay heat from the reactor. After 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> following the event, core heat had l'

' been sufficiently removed that the unit could be placed on residual heat removal (RHR) and no further steam release or activity release was assumed to occur. For completeness, the staff incorporated the contribution from the intact SGs for this (seriod even though the dose contribution would be small from this source. The licensee had excluded this contribution from its assessment. Table 1 presents the assumptions utilized by the staff in its assessment.

The potential consequences of an SGTR accident are presented in Table 2. For this accident, the staff did not perform an assessment of the whole-body dose associated witn the release of noble gases because the thyroid is so limiting with respect to compliance with General Design Criterion (GDC) ig and 10 CFR Part *00. The doses were found to be within a small fraction of 10 CFR Part 100 guidelines for the accident-initiated spike case and within Part 100 for the pro-existing spike case for offsite exposures and less than GDC ig guidelines in both cases end, therefore, acceptable.

2.3.3 Gummaty The staff has concluded that the doses would not exceed the dose guidelin*s presently contained in the Standard Review Plan for the SGTR accident and that in no case would the specific fraction of 10 CFR Part 100 be exceeded offsite nor would GDC ig of 10 CFR Part 50, Appendix A, doses be exceeded for control room operators, respectively.

3.0 CONCLUSION

The staff has reviewed the licensee's reanalysis of the SGTR accident analysis for Byron, Unit 1 and Craidwood, Unit 1 and concluded (1) that the opemtor action Smes during an SGTR are satisfactory; (2) the analysis performed in calculating SG margin to uc&, fill used an approved methodology with appropriately conservative inpute and assemptions. There is an ace ,) table margin to SG overfill and the use of this analysin (or input into the r%se consequence analysis are also acceptable; and (3) that the doses would nct exceed the cose guidelines presently contained in the Standard Re few Plan for the SGTR accident and that in no case would the specific fraction of 10 CF3 Part 100 be exceeded offsite nor would GDC ig of 10 CFR Part 50, Appendix A, doses be exc#.J for cor trol room operators, respectively.

o ,w - -

. Therefore, with regard to the SGTR acc8 dent analysis, the staff Ands the replacement of the SGs acceptable.

Prampal Contributors: G. West C. Jackson J. Hayes Date: January 28, 1998 j

e

, REFERENCES

1. '

Letter from C.E. Rossi (NRC), to A.E. Lcdieu (Westinghouse), transmitting WCAP-10698,

'SGTR Analysis Methodolopy to Determine the Margin to Steam Generator Overfill,"

March 30,1987,

2. Letter from T.K Schuster (Comed), to T.E. Murley (NRC), " Byron Station Units 1 and 2, and Braidwood Station Units 1 and 2, Steam Generator Tube Rupture Analysis," April 25, 1990.

3. NRC Letter dated April 23,1992," Byron, Units 1 and 2 and B sidwood, Unhs 1 and 2 -

Steam Generator Tube Rupture Anatysis".

4. NRC letter dated April 12,1994, " Acceptance fcr Referenchg the RETRAN-02 MOD 5.1 Code'.

l g - - -

TM21 Staff Assumptions Used for Steam Generator Tube Ruptum Accideilt Evaluation lodme Partition Factor Faulted SG 1.0 intact SGs 0.1 Steam and H2O Reloases from Faulted SG 0120 minutes 1.18E5 Steam Release from Intact SGs (Ibs)

C 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 4.10E5 2-8 hours 9.49E5 616 hours0.00713 days <br />0.171 hours <br />0.00102 weeks <br />2.34388e-4 months <br /> 6.77E5 16-24 hours 5.81E5 24-32 hours 5.22E5 32-40 hours 4.82E5 Primary to Secondary Leak Rate 150 (gpd/SG)

Tirne to isolate Faulted SG (min) 120 Flashing Fraction Variable with respect to time. Provided in Comed letter dated 11/3/97.

Scrubbing Fraction 0 Primary Bypass Fraction for Inl .ct 0 SGs Duration of Plant Cooldown (hrs) 40

' Primary coolant concentration of 60 pCilg of dose equivalent *l Pre-existino Solke Value fuci/a)

• l 5

= 46.2 1 = 51.7

• 1 = 73.9

• l ' = 11.1 "I = 40.6 Volume and mass of primary coolant and secondary coolant Primary Coolant Volume (ft') 12,062 @586.2 'F Primary Coolant Temperature (*F) 586.2 T1-1

. l a

t. ,

. TABLE 1- oontinued Mass of Primary Coolant (Ibs) 638,361 Primary Coolant Pressure (psla) 2,293 Pressurizer Temperature (Y) 657 Pressurizer Pressure (psia) 2,293 Pressurizer Volume (ft') 1,150 Secondary Coolant Steam Volume (ft') 2,780 Secondary Coolant Liquid Volume (ft') 2,423 Secondary Coolant Steam Mass /SG (Ibs) 5,571 Secondary Coolant Liquid Mass /SG (Ibs) 105,224 ,

Secondary C4olant Steam Temperature (T) 523 Secondary Coolant Feedwater Temperature (T) 440 TS limits for DE *l in tne primary and secondary coolant Primary Coolant DE '8'l concentration (pCilg)

Maximum Instantaneous Value 60 48-Hour Value 1.0 Secondary Coolant DE *l concentration (pCilg) 0.1 ,

TS value for the primary to secondary leak rate Primary to secondary leak rate, any SG (gpd) 150 Primary to secondary leak rate, total (gpd) 600 Maximum primary to secondary leak rate to the faulted and intact SGs Faulted SG (gpm) 150 Intact SGs (gpm/SG) 150 Letdown Flow Rate (gpm) 75 Equilibrium Release Rate from Fuel for 1 pCilg of Dose Equivalent *l

.QUA.tX

• l = 2,040 5 1 = 5,300 ml = 5,330

• l = 7,370

• 1 = 5,300 Control Room Free Volume (ft') 4.05Eb Filtered Recirce!stion Flow (cfm) 4.45E4 Recirculation Efficiency for All 90 T1-2

^

TARIA.16 Forms oflodme (%)

Makeup FMer Ef5ciency for All 99 Forms of lodme (%)

Makeup Air FMration Rate (cfm) 5400 Control Room SEtQ Braidwood Unfiltered Air infiltration Rate (cfm) 89 25 Occupancy Factors 01 day 1.0 1.0 14 days 0.6 0.6 Atmospheric Dispersion Fadors (sec/m')

Control Room 0 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> 4.05E-3 6.2E-3 8-24 hours 1.9E-3 3.2E-3 14 days 5.7E-4 8.4E-4 4-30 days -

1.4E-4 EAB 6.8E-4 7.7E 4 LPZ 0-6 hours 2.3E-5 7.0 E-5 8-24 hours 1.5E-5 5.2 E 5 14 days 6.4E-6 2.1 E-5 l

4-30 days -

5.6 E Spiking Fador for Accident-Initiated Spike 500 500 ledine Partition Factor 0.01 0.01 Steam Release from Defective SG 0 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (Ibs) 9.55E4

> 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (Ibs) 0 Steam Release from intad SGs (Ibs) 0-2 hours 1.61E5 2-40 hours Assumed to be at the same rate as for MSLa.

Reador Coolant Released to Fautted 1.41E5 SG (lbs)

Ti-3

TABLL1-continued 1

Primary to Secondary 1.eak Rate 150 (9Pd/SG)

, Time to isolate Faulted SO(sec) 3300 4

i I

J J

. T14 m

, TABLE 2 Potential Cc=equences of a steam Generator Tube Rupture Accident

, flRQN BRAIDWOOD fall LEZ Eall LPl Thyroid Dose

= Coincident Spike 1) 0.49 0.23 5.5 0.79 Pre-existing Spike 2) 24.7 0.85 2.8 2.9 Whole-Body Dose Coincident Sp!ke 3) <1 <1 <1

= <1 Pre-existing spike 4) <1 <1 <1 <1 Control Room Dose 5) THYROID WHOLE BODY THYROID WHOLE BODY Coincident Spike 0.15 <1 .11 <1 Pre Existing Spike 0.55 <1 0.41 <1

1) Acceptance Criterion is 30 rein

2) Acceptance Criterion is 3C') rem

3) Acc.eptance Criterion is 2.5 rem

4) Acceptance Criterion is 2.5 rem

5) Acceptance Criterion are 5 rem whole body and 30 rem thyroid.

.. 9