ML20217G617

Operability Assessment Process Form

ML20217G617
Person / Time
Site:	Byron
Issue date:	07/22/1997
From:	Adams B COMMONWEALTH EDISON CO.
To:
Shared Package
ML20217G615	List: ML20217G614 ML20217G617
References
97-044, 97-44, NUDOCS 9708080003
Download: ML20217G617 (23)

Text

. . . - _._. - - - - - - _._ ..- . - ..- . -

4 C mEd 97 044

Att: chm 2nt C

_Oper9bility Assessm:nt Proc:ss G/28/94 Operability Assessment Process Form 6.13.1

a. PlF # : B1997 01600 Assesstnent Expected Due Date/ Time: 08/21/97 1700 Operability Determination (check one): DD Operable O Inoperable O Operable, but degraded Approved Compensatory Actions - List those approved (also see attached list):

YES O NO 00 Approved Corrective Actions - List those approved (also see attached list):

YES D e NO DD OPERABILITY ASSESSMENT RECOMMENDATION: _7/u/p 6.13.2 Preparer / Evaluator: ./Im /d M 4% MU!97 (}O 6.13.3.b Regulatory Assur. Supv: _, N os (( M M $dadv ,

System Engineering Supv bao -h/n/97 Site Eng. Support Supv: /dNN A,A.w 74 I 4 7 Operating Engineer: n /fo dnM OPERABILITY ASSESSMENT APPROVAL:

6.13.5.d Shift Engineer: /f'OI M//r/A/A x

9700000003 970730 PDR ADOCK 05000454 P PDR u I

I<

CdmEd 97 044 Atttchmsnt C

}- Op;rtbility Ass:ssm:nt Proccss 6/28/94 I

, Operability Assessment Process Form l

de b 6.13.1 GeneralInformation: - '

[ b. Affected Station (s): RytoD i

c. Unit (s)

.1  !

l d. Description of problem, failure, defect, degraded or nonconforming condition:  ;

! Per the Byron Technical Specification (TS) 4.4.5.4, operability for steam generators is l j determined by the performance of an augmented inservice inspection program and i

. application of the TS listed acceptance criteria. TS 4.4.5.4.a.11 specifically addresses j use of the 3.0 Volt Interim Plugging Criterla (IPC). As part of the requirements for

) voltage based repair criteria for steam generator tubing with Outer Diameter Stress F

Corrosion Cracking (ODSCC), the potentialleakage during a MSLB accident with  :

! containment bypass must be calculated. This calculated leak rate value is then i compared to the site allowable leakage limit which results in a dose corresponding to a j small fraction _ (10%) of 10 CFR Part 100. The 10CFR Part100 requirements and the

[ caLilational methodologies are defined in Comed's TS Amendment Request L (Reference 1) and the NRC's subsequent approval for 3.0 Volt IPC implementation f (Reference 2). For Byron Unit 1 and Braldwood Unit 1, the voltage based repair criteria  ;

used is greater than that allowed by NRC Generic Letter 95 05 criteria for 3/4 inch ,

' tubing, (3.0 Volts as compared to 1.0 Volt). The increased repair criteria was reviewed and approved by the NRC for use at Byron Unit 1 and Braldwood Unit 1.

N

[ 3.0 Volt IPC was first implemented at Byron and Braidwood Stations in fall 1995 outages at each unit, respectively. As a result of recent steam generator inspectiont., ,

Braldwood determined that the allowable as found IPC tube leakage exceeded the cycle ,

predicted value. The predicted room temperature value was 6.99 gpm and the actual . ,

room temperature leakage identified at Braldwood was found to be 11.5 gpm. Since the site allowable leakage limit was calculated for operating temperature conditions, the ,

11.5 gpm at room temperature is scaled up by a factor of 1.406 to account for the density differences at the respective' operating temperatures.' The adjusted as-found leakage is then 16.2 gpm and was still below the allowable IPC leakage limit of 26.8 gpm for Braidwood Unit 1.

This operability assessment is being generated to document the applicability of the root cause performed for the Braidwood Unit 1 IPC (Reference 3) under prediction as it applies to Byron Unit 1. Note that a previous operability assessment,97-034,

determined that there was reasonable assurance of operability for Byron Unit i due to

. this issue. Subsequent to that initial determination, the root cause report was issued and an additional PlF and operability screening was performed at Byron Station

. (Reference 4). This Operability Assessment is generated to document the final determination of operability.

2-

_ m. _ _ . -

<, CSmEd 97 044 Att chm:nt C

_Operrbility Ass:ssm:nt Proc:ss 6/28/94 Operability Assessment Process Form G.13.1 (Cont'd)

c. Component (s) Affected:

Byron Unit 1 Steam Generators l l

I

f. Identification Number (s)(EPN, part/ serial number, etc.):

1 1RC01BA,1RC01BB,1RC01BC,1RC01BD Evaluation:

g. Describe the safety function (s) or safety support function (s), answering the following as part of the description: '

1. Does the affected SSC receive / initiate a RPS or ESF actuation signal? No

2. Is the affected SSC in the main flow path of an ECCS or support system? No

3. Is the affected SSC used to:

(a) Maintain containment integrity? No (b) Shutdown the reactor? Yes (c) Maintain it in a shutdown ~ condition? Yes (d) Prevent or mitigate the consequences of an accident that could result !n off site exposures comparable to 10CFR100 guidelines. Yes l 4. Does the SSC provide required support (i.e. cooling, lubrication, etc.) to a TS required SSC7 Yes

5. Is the SSC used to provide isolation between safety trains, or between safety and non-safety ties? Yes

6. Is the SSC required to be operated manually to mitigate a design basis event? No j 7. Technical specifications? 3/4.4.5.4

8. UFSAR or pending revisions? Various including Section 15.1.5 (MSLB)

\ .

t 3

_ i'

CNEd '97-044 Att: chm:nt C

_Opercbility Ass:ssm:nt Proc:ss 6/28/94 Operability Assessment Process Form G.13.1 (Cont'd)

Ensure the safety function (s) of the SSCs are included in the description:

The purpose of the steam generators at Byron Station is to provide a means for transfer of heat from the primaiy reactor coolant to the secondary cooling system.

In addition, the steam guerator tubes also serve as an RCS barrier to separate l primary coolant from the rest of the plant. The steam generators serve as a heat sink for a number of postulated events in the B/B Chapter 15 a1cidents, including the Main Steamilne Break (MSLB) accident, Section 15.1.5.

In combination with the auxillary feedwater system and steam generator power-L operated relief valves, the steam generators can be used to perform the safety function of removing residual heat following an accident and cool the plant down below 350'F, until the Residual Heat Removal System (RH) is made available.

h. Describe the effect the concern has on the SSC safety function (s):

The methodology used to calculate the 3.0 Volt IPC, EOC distribution and leak rate for Byron Unit 1, is the same as Braidwood and therefore, there is a potential that the EOC-8 distribution and leak rates are under-predicted. The Byron Unit i limiting leak rate is for steam generator C with an EOC 8 predicted leak rate of 30.8 gpm (21.9 gpm at room temperature) compared to the site allowable leak limit of 36.5 gpm. Note that leak rates from other steam generator tube degradation mechanisms must also be considered in evaluating the Byron Unit i EOC-8 leak rate condition. After 448.5 days of operation, a I leak rate of 31.9 gpm (22.7 gpm at room teinperature) must also be considered for TTS circumferential indications for the operating period to 479 days when the Byron unit 1 steam generators are being replaced. The total EOC 8 leakage attributed to IPC leakage, TTS circumferential leakage, and operational leakage of the three intact steam

- generators, is 63.1 gpm (44.9 gpm at room temperature). Therefore, for Cycle 8 operation beyond 448.5 days, a dose equivalent lodine (DEI) TS Amendment request was submitted to NRC (Reference 5) to reduce the RCS DE lodine limit to 0.2 pCIIgm which effectively increases the site allowable leak limit to 64 gpm. This amendment request is currently pending NRC review and approval.

4

, e.g ' CdmEd -97 044 Opercbility Ass:ssm:nt Process Att: chm:nt C 6/28/94 Operability Assessment Process Form 6.13.1 (Cont'd)

1. Tech Spec SSC affected.

DDYes D No Tech Spec Section(s): 3/4.4.5.4 & Bases

-- J. Equipment Degraded:

0 Yes - Enter Degraded Equipment Log or Equivalent 00 No

k. UFSAR System affected:

00Yes D No '

UFSAR Sections: Various including Section 15.1.5.3 (MSLB) 5

l t, rinEd 97 044

• • Attachmsnt C Operchility Ass ssm:nt Proc:ss 6/28/94 ,

Operability Assessment Process Form 6.13.1 (Cont'd)

1. Does the SSC meet its required design function?

00Yes D No  ;

1 m. Justification:

3 l

There is reasonable assurance that Byron Unit 1 can operate safely until the steam l generator replacement outage currently scheduled to begin in November,1997. This i j reasonable assurance is based on calculations that were reperformed for Byron Unit 1 l EOC 8 IPC leakage based on the results of the Braidwood Unit i root cause  !

determination. Note that the calculational methodology was revised for Byron Unit 1 as l a result of the root cause report. The original calculation for Byron Unit 1 IPC leakage i did not consider a voltage dependent growth rate for ODSCC cracking to determine end of cycle conditions. The revised calculations cantain design assumptions for a voltage dependent growth rate that is based on the results of the Braldwood Unit i root cause report. Note that the revised calculational methodology was benchmarked versus the Braldwood data to determine its acceptability. ,

The attached report (Reference 6) contains a documented review by Comed's Corporate 4

Engineering of the Byron Unit 1 IPC leakage for EOC 8. This report documents i Corporate Engineering's assessment of operability for Byron Unit 1 Cycle 8. - The previously calculated room temperature value for IPC leakage was 19.0 gpm. Using the .

revised calculational methodology, the new room temperature value is 21.9 gpm. Since the site allowable leakage limit was calculated at operating conditions, the 21.9 gpm at  ;

room temperature is adjusted to 30.8 gpm at operating conditions. This value remains  ;

below the site allowable leak limit of 36.5 gpm.

As documented in the attached report, for operation beyond 448.5 days, NRC approval I of the DEI TS amendment change (Reference 5) to 0.2 pcilgm is required to account for the additional leakage due to TTS circumferentialIndications. The total room temperature leak rate will be 22.7 + 21.9 + 0.3 gpm "frorn the intact SGs" = 44.9 gpm after 479.5 days. Adjusting the room temperature leakage of 44.9 to operating conditions results in a total leakage of 63.1 gpm, compared to the site allowable leak limit of 64 gpm at 0.2 pcilgm (assuming NRC approval of the TS amendment).

. Therefore, the identified discrepancy at Braidwood between predicted and actual TSP cracking does not create an operability concern for Byron Unit 1 with respect to Cycle 8 operation until expected shutdown in November 8,1997. Note that a NTS item (NTS

1. 454-20197-CAQS01254 03) has been assigned to Byron SEC to track implementation of the TS amendment at Byron Unit 1 prior to Cycle 8 operation of 448.5 days. No additional actions are required, i

6

4

Comed 97 044 Att chmtnt C ;

)' _Opercbility Ass:ssmrnt Proc:ss G/28/94  ;

Operability Assessment Process Form  :

i G.13.1 (Cont'd) '

, Disposition:

l  ;

3

1. I E the failed, DEGRADED or NONCONFORMING SSC prevents the accomplishment ~

, of the safety function (s) of a Tech Spec SSC specifically described by that TS,

.TJEN the SSC does NQI meet its required design furiction(s). Check No and 1

recommend to the shift engineer to declare the equipment inoperable; 3

a 1

[ 2. E the failed, DEGRADED or NONCONFORMING SSC alters a SSC from its description in the UFSAR ANQ the evaluation concludes that it does NOT affect the ability to meet the required design function (s)in the TS AND the intention is to 4

continue operating the plant in that condition, THEN check YES and a 50.59 evaluation is required and should be attached to this attachment. A review shall be j

performed to determine if a UFSAR update is required; j 3. E the failed, DEGRADED or NONCONFORMING SSC does NOT affect the ability to l

meet the required design function (s) and does NOT alter a SSC from its description

. In the UFSAR, THEN check YES and documant the recommended corrective actions to restore full qualification on this attachtrant.

4 i

V t

i

. 7.

-c...-.. . , , . _ _ . , , , , . . . . - , . _ _ _ , - - - _ , _ _ . . r -..,w,w-,.e.., .w._,,. . , ,., .-..r.-v,nw e-,,,-,-er,_,,,m.n- __ -,-,.m. .-

4  ;, C6mEd 97 044

_Op:rcbility Ass:ssm:nt Proc ss Att chmtnt C

, 6/28/94  ;

i Operability Assessment Process Form '

4 l n.-

REFERENCE DOCUMENT LIST '

f  !

1. Letter from D. Saccomando to USNRC, Supplement to Application for Amendment to Facility Operating Licenses, datea September 1,1995 l f

4

- 2. USNRC Letter to D. Farrar, Issuance of Amendments for 3.0 Volt IPC, dated  ;

i November 9,1995  !

j

3. Braidwood Unit 13.0 Volt IPC Leak Rate Under Prediction Root Cause Report, dated June 16,1997 i

q - 4. BSE 101, Attachment B, Operability Assessment Screening (97 044)in response to PIFs # B1997 01600, dated June 9,1997. '

5. Dose Equivalent lodine (DEI) Technical Specification Amendment Request for  ;

j Byron Unit 1, J.- Hosmer to U.S. NRC, dated January 31,1997.

l 6. " Byron Unit i Operability Ane4sment of Braldwood Unit 1 IPC Leak Rate Discrepancy," SGRVP 97 0007, J. Blomgren (Comed NES) to D. Wozniak (Byron l SEM), dated July 21,1997, e

4 j  ;

r 4

4 e

4 1

i i

8

,'bSmEd ' '

97 044

Att:chrn:nt C

.Op:rchility Ass:ssm nt Proc ss 6/28/94 Operability Assessment Process Forrn

o. COMPENSATORY ACTION ITEM LIST Compensatory action (s) or mitigating condition (s) required to ensure operability:

Dyes 00 No (refer to Corrective Action item List)

Listed below are compensation or mitigating conditions that are required to support operability of; Compensatory Action #1:

Rasponsible Dept./Supv.:

Action Due:

NTS:

Compensatory Action #2:

R:sponsible Dept./Supv.:

Action Due:

N_TS:

Compensatory Action #3:

Responsible Dept./Supv.:

Action Due:

NTS:

Compensatory Action #4:

R:sponsible Dept./Supv.:

Action Due:

NTS:

9

. Comed 97 044

Opercbility Ass:ssment Promss Att: chm:nt C 6/28/94 0; 'rability Assessment Process Form q, COkRECTIVE ACTION ITEM LIST Corrective Action (s) required?

Oyes 0D No (refer to the Compensatory Action Item List)

If any of these items cannot be accomplished in the specified time notify the Operations Manager.

Corrective Action #1: Track implementation of the Dose Equivalent lodine TS Amendment Request for Byron Unit i to ensure approval prior to Cycle 8 operation of greater than 448.5 days.

R2sponsible Dept./Supv.: SEC/Contrady/ Smith Action Due: September 15,1997 NTS: 454 20197.CAQS01254 03 Corrective Action #2:

R2sponsible Dept./Supv.:

Action Due:

NTS:

Corrective Action #3:

R:sponsible Dept./Supv.:

Action Due:

NTS:

Corrective Action #4:

R::sponsible Dept./Supv.:

Action Due:

NTS:

10

s

, July 22,1997 SGRVP.97-0007 To: D. Wozniak Byron SEC hianager From: J. Blomgren SG&RVP hianager

Subject:

Byron Unit i Operability Assessment of Braidwood Unit 1 IPC Leak Rate Discrepancy The SG&RVP Group have completed an operability assessment of the Braidwood Unit 1 3.0. Volt Interim Plugging Criteria (IPC) leak rate discrepancy. A discrepancy exists at Braidwood Unit I between the predicted and as found End of Cycle six (EOC 6) Steam Generator conditions. Because Byron Unit 1 implements the same 3.0 Volt IPC methodology an operability assessment has been performed. Previous inspection IPC predictions have resulted in conservative values when compared to the as found conditions.

This operability assessment applies a change to the 3.0 Volt IPC methodology developed to address the root cause of the Braidwood under. prediction. The operability assessment concludes that Byron Unit I will remain within site allowable leak limits for 479.5 days of operation > $00"F until the currently scheduled SG replacement outage.

cc: Jay Smith Lany Alexander Keith hioser hiike Sears Denise Saccomando Roman Gesior SG&RVP File Brad Adams mm

e Evron Unit 1 Operab.ility Asmsment of Braidwood UnitJ

~

l 3.0 Volt IPC LenidattJJadthhtditliDD '

l

1.0 Generallnformation

This Byron Unit 1 operability assessment is performed to address an issue at Braidwood Unit I where discrepancies exist between the predicted and actual End of Cycle six (EOC-6) Steam Generator conditions for the Tube Support Plate (TSP) Outside Diameter Stress Corrosion Cracking (ODSCC) 3.0 Volt Interim Plugging Criteria (IPC). As required by NRC Generic Letter (GL) 95 05, application of steam generator alternate repair criteria requires calculation of the potential primary to secondary leakage and the conditional probability of burst assuming a Main Steam Line Break (MSLB) accident with containment bypass. These calculations must be performed for both the "as found" condition of the steam generator and the condition as predicted at the end of the next operating cycle. These values are then compared to site specific acceptance criteria and are used to determine if accelerated corrective actions such as changing the repair limit or shortening the next operating cycle must be taken.

In the case of Braidwood Unit 1 Cycle 6, the "as found" conditional probability of tube burst was well within the predicted values and below the NRC Generic Letter Limits. The "as found" potentialleakage during a Main Steam Line Break accident with containment bypass was less than half the site allowable leakage limit, but was up to 65% higher for steam generator (SG) C than had been predicted based on information available during the previous inspection. Three of the four remaining SGs leak rates were also under-predicted but to a lesser cuent.

The methodology used to calculate the 3.0-Volt IPC, EOC distribution and leak rate for Byron Unit 1, is the same as Braidwood and therefore, there is a potential that the EOC 8 distribution and leak rates are under predicted. The Byron Unit I limiting leak rate is for SG C with an EOC-8 predicted a room temperature leak rate of 19.0 gpm. Since the site allowable leak limit is calculated at reactor coolant operating temperature and degradation mechanism leak rates are typically calculated at room temperature, the room temperature leak rates must be scaled up by a factor of 1.406 to account for the density differences between room temperature and opereting conditions. This adjustment is necessary in order to compare the predicted leak rates to the site allowable leak limit.

Therefore, the predicted EOC-8 leak rate is then 26.7 gpm, which is significantly less than the site allowable leak limit of 36.5 spm.

Leak rates from other steam generator tube degradation mechanisms must also be considered in assessment of the Byron Unit 1 EOC-8 leak rate condition. A leak rate of 31.9 gpm (22.7 gpm at room temperature) must be considered for TfS circumferential indications for the operating period of 479.5 days to SG replacement (Reference 8.9). In order to account for the additional leakage from TTS circumferential indications, associated with operating beyond the 448.5 days presently approved by NRC, a dose equivalent iodine (DEI) Tech Spec change was submitted to NRC (Reference 8.10) to

. I ., ,

increase the site allowable leak limit to 64 spm. The following discussion provides more details on the basis of the TTS circumferential indication leak rate value.

In a meeting on June 20,1996, regarding top of the-tube sheet (TfS) circumferential indications NRC concuned with Byron Unit 1 operation for a period equivalent to the previous cycle of operation or 448.5 days > $00*F (Reference 8.4), in subsequent conespondence (References 8.5, 8.6, 8.7 and 8.8) Comed requested and received NRC concurrence for operation of Byron Unit 1 beyond 448.5 days to $40 days of operation >

500*F. In the additional analysis to support 540 days of operation Comed performed two conservative TTS circumferential indications leak rate predictions.

In Reference 8.6 two conservative TTS circumferential indication analyses were performed to predict a leak rate after 600 days e f operation (original cycle length). The first analysis utilizes growth rates from full and nearly full cycles of operation (Detween UlR06 and B1P02 and between BIR06 and BIR07). A room temperature leak rate of 28.4 gpm was calculated from this analysis. The second case utilizes growth rates from the first analysis as well as shon cycle growth rates (100 days between BlP02 and D1R07). A discussion is presented in Reference 8.5 (Attachment C RAI Question 3) which explains why applying short cycle growth rates to a full cycle of operation is not appropriate, This second analysis results in a room temperature leak rate of 35.7 gpm.

The expected cycle length with the planned replacement outage starting on November 7, 1997 is 479.5 days > 500'F. Therefore, a TTS circumferentialindication leak rate of 22.7 gpm (28.4 gpm x (479.5 days /600 days)) at room temperature must be considered for EOC leak rate assessment. This conesponds to a leak rate of 31.9 gpm at operating 4

conditions.

2.0 Background As pan of the requirements for voltage based repair criteria for steam generator tubing with ODSCC, the potentialleakage during a MSLB accident with containment bypass must be calculated. This calculated leak rate value is then compared to the site allowable leakage limit which results in a dose corresponding to a small fraction (10%) of 10 CFR Part 100. These requirements and calculation methodologies are defined in our request (Reference 8.12) and NRC approval for 3.0 Volt IPC implementation (Reference 8.13) following the guidance of NRC Generic Letter 95 05 (Reference 8.11) which governs the approval and use of a voltage based alternate repair criteria. In the case of Byron 1 and Braidwood 1, the voltage based repair criteria used is greater than that allowed by the Generic Letter for 3/4 inch tubing, (3,0 Volts as compared to 1.0 Volt). This increased

repair criteria was reviewed and approved by th

NRC for Byron and Braidwood Unit 1.

The basis of these increased voltage repair criteria relics on ensuring the tube support plates do not displace more than 0.10 inch during accident conditions, and thus the ODSCC region of the tubing remains confined within the support plate. Expanding,

" locking" tubes at selected support plate intersections, and then removing the tube from service accomplished this. The tube then functions as an additional stay rod to limit support plate movement during accident conditions.

. .y -

l1 The required tube expansicus and subsequent 3.0 Volt IpC implementation was begun in the fall of 1995 at both Braidwood 1 (Cycle 5 Refuel Outage) and Dyron 1 Cycle 7A l

(Mid cycle inspection). As part of the development and licensing of the 3.0 Volt IPC, a revised leak rate calculation methodology was developed. This leak rate methodology i calculates the probability that an indication could open at MSLB conditions while contained within a tube support plate. This condition is ieferred to as an indication restricted from burst (IRB). Ifit is predicted that an IRB condition were to occur at an indication, a bounding leak rate value of 6.0 gpm is assigned to the indication. The 6.0 gpm boundim leak rate value was determined experimentally through a Comed / EPRI sponsored tut program. The test prograrn used laboratory produced throughwall cracking l

which was much larger than that seen in service throughout the industry to date.

Therefore, the 6.0 gpm bounding leak rate value is considered to be conservative.

Laboratory testing of Braidwood Unit 1 tube Row 28 Column 24 in an IRB configuration resulted in a leak rate of 0.9 gpm which is significantly below the assumed 6.0 gpm.

The 3.0 Volt IPC leak rate calculation methodology combines the NRC GL 95 05 "Freespan" leak rate vake taken at a 95% confidence level with the leakage contribution from IRB's to form a total Icak rate. This is performed on a steam generator specific basis.

The Braidwood 1 Cycle 6 SG C predicted leak rate value is 6.99 gpm as compared to the "as-found" leak rate value for Cycle 6 of 11.5 gpm (Reference 8.2).

A root cause investigation of the discrepancy between the predicted and as found leak rates was performed; the results are discussed in the next section.

3.0 Root Cause in order to assess the root cause for this under prediction, Comed has evaluated the calculation methodology as well as all inputs to the methodology to identify those, which could affect the EOC voltage distribution and leakage. Steam Generator C was selected to evaluate the root cause because it resulted in the largest deviation from the prediction.

The root cause of the Braidwood EOC distribution and leak rate under prediction, as l described in Reference 8.1, is presented below:

l Evaluation of voltage growth rates, as described in GL 95-05, was performed to ensure that the conservatism, ofusing absolute voltage difference (normalized to efTective full i

pvver years, EFPY), continues to be supported by operating experience. Evaluation of growth rates has identified a voltage dependent growth rate for ODSCC at the tube support plate (TSP) location. Braidwood EOC 6 predictions use a distribution of growth rates applied to the entire beginning of cycle voltage distribution. Previous Braidwood Unit 1 inspection cycles implemented a 1.0 Volt IPC and were shortened due to mid-cycle inspections. These two factors, which served to minimize the size ofindications left in service, along with growth rates which are now shown to be voltage dependent, led to an under-prediction of the Braidwood Cycle 6 EOC voltage distribution and leak rates l

N (i.e. if smaller indications are len in service the larger growth rates associated with the larger indications are not experienced).

In order to assess the impact of the root cause, a change to the 3.0 Volt IPC methodology was made to account for the voltage dependent growth rates. The methodology is discussed in the next section.

4.0 Voltage Dependent Methodology The present 3.0-Volt IPC methodology applies the entire growth rate distribution across .

the entire beginning of cycle (BOC) voltage distribution in calculating an EOC distribution. The change in the methodology, to account for voltage dependent growth iates, segregates the growth rates based upon the DOC voltage of the indication into separate distributions. The guidelines used to determine how the growth rate distributions are developed are listed below:

A hybrid growth distribution is created by ensuring that the limiting SG growth includes the largest three values found in all SGs; BOC voltage bins must have a minimum of 200 growth points. Therefore, select the highest BOC voltage bin such that it has over 200 growth points and meets the remainder of the criteria; Aner selecting the highest voltage bin select the next two DOC voltage bins of a width such that growth rates are voltage dependent for the three bins, this is typically from 0.4 to 0.6 volts; The highest voltage bin should include at least one of the largest three growth values; Cumulative probability distributions of binned growth rates should show increased voltage growth at least above 80% probability as the BOC voltage increases.

The above guidelines are incorporated into the standard GL 95-05 calculation methodology to determine the end-of cycle distributions and room temperature leak rates. The following steps for Byron Unit 1 EOC 8 predictions provide an example of how the methodology is applied, which are also shown in Figure 1;

1. Sort the Cycle 7B growth rates by BOC-7A indication voltage for each steam generator;

2. If the three largest growth points from all SGs do not occur in the SG with the worst average growth rates, add the largest growth points to the distribution of the worst SG;

3. Follow the guidelines discussed above to determine if there is a BOC l voltage dependency on the growth rates and to select BOC voltage bins if a voltage dependency exists (see Figure 2 for evaluation results);

4. For each of the three BOC-7B voltage bins determined in step 3 (0 0.5, 0.5-0.9, >0.9), calculate the EOC-8 distribution using the corresponding voltage dependent growth rates and NDE uncertainty; Combine the EOC i

i

,l distributions from the three distributions in step 3;

5. Calculate the burst probability and leak rate for the total EOC distribution.

The methodology discussed at>ove was benchmarked against Braidwood data, which is l discussed in the next section.

5.0 Methodology llenchmark in order to validate the vohage dependent methodology, for Byron Unit 1 the methodology was applied to Draidwood Cycle 6 data using cycle 6 voltage dependent growth rates. The data was then compared to the as found distribution and leak rate.

Inputs into the benchmark analysis include:

The beginning of cycle distribution selected was SG C DOC 6;

• A hybrid voltage dependent growth rate distribution was developed using the methodology in Section 4.0 by segregating Cycle 6 growth rates into three DOC-6 voltage bins: 0 - 0.7, 0.7 - 1.1, and > 1.1 (see Figure 3);

• Growth rates are normalized to number of days > 500*F; The analysis was performed for 413 days of operation > 500'F;

• The remainder of the inputs are consistent with the standard 3.0 Volt IPC methodology.

The EOC 6 distribution and leak rate predictions were performed using the methodology described in Section 4.0, The analysis resulted in a SG C, EOC 6 leak rate of 13.1 gpm with the largest indication at 10.3 volts (Reference 8.3). This compares to the as found leak rate of 11.5 gpm and the largest indication (field call) at 11.9 volts which includes an allowance for NDE uncertainty. The benchmark shows that there is significant EOC leak rate margin in the voltage dependent methodology and the largest EOC indication is nominally the same as the as found. See Figure 4 for a comparison of the predicted EOC-6 distribution for SG-C, using the voltage dependent growth rate methodology, and the SG C as-found distribution.

6,0 Application of Voltage Dependent Methodology to Hyron Unit 1 Conditions at Byron at BOC-8, are similar to Braidwood EOC-6, which led to the under-prediction of the EOC-6 leak rate and distribution. Both units were coming off short cycles due to mid cycle inspections. Draidwood Unit 1 implemented a 3.0-Volt IPC for the first time whereas Byron implements the 3.0 volt IPC for the second time. Having implemented the 3.0 Volt IPC during a previous outage the impact of the voltage dependent growth rates is less due to the fact that larger indications had been left in service leading to larger growth rates during the cycle. These larger growth rates were used in the Byron Cycle 8 predictions.

The following inputs were used in applying the voltage dependent growth rate

methodology to Byron Unit 1 Cycle 8:

• The most limiting beginning of cycle distribution selected was SG C, A hybrid voltage dependent growth rate distribution was developed by segregating Cycle 7B growth rates into three BOC-7B voltage 0 - bins:

0.5, 0.5 - 0.9, and > 0.9 (see Figure 2);

Growth rates are nonnalized to number of days > 500'F; The analysis was performed for 500 days of operation > 500"F; The remainder of the inputs are consistent with the standard 3.0 Volt IPC methodology.

The analysis resulted in an SG C, EOC 8 leak rate of 22.8 gpm and largest indication 11.0 volts shown in Figure 5 (Reference 8.3), which remains below the site allowable leak limit of 36.5 gpm. 'Ihe analysis used an operating cycle of 500 days > 500"F. To determine the leak rate for a period of 479.5 days (Reference 8.9) the leak rate m conected as follows 22.8 gpm x 479.5/500 = 21.9 gpm. Since this leak rate is calculated at room temperature, it is adjusted to 30.8 gpm to represent leakage at operating conditions.

7,0 Conclusions Applying a conservative voltage dependent growth rate methodology to Byron Cycle 8 results in a leak rate, when combined with leak rates from other degradation mechanism significantly below the site allowable leak limit of 36.5 gpm. For operation beyond 448.5 days, an approval of the DEI Tech Spec change to 0.2 pei/gm is required to account for the additionalleakage due to TTS circumferentialindications. The total room temperature EOC 8 leak rate will be 22.7 +21.9 + 0.3 gpm "from the intact SGs" = 44.9 i i

gpm after 479.5 days. Adjusting this leak rate to operating conditions, results in a total EOC 8 leak rate of 63.1 gpm, compared to the site allowable leak limit of 64 gpm at 0.2 pei/gm). Therefore, the identified discrepancy at Braidwood between predicted and actual TSP cracking does not create an operability concern for Byron Unit ! with respect l to Cycle 8 operation until expected shutdown in November 8,1997.

8.0 Referenees

8.) Braidwood Unit 13.0 Volt IPC Leak Rate Under Prediction Root Cause, June 16,1997 8.2 Westinghouse S. Srinivas, Braidwood Unit 1 EOC-6 Projections for Additional Cases, Fax Dated 5/30/97 8.3 Westinghouse S. Srinivas, Updated SLB Leak Rate Results for Braidwood-l and Byron-1, Fax Dated 7/13/97 8.4 Byron Letter, BYRON-96-0188, Byron Unit 1 Cycle Length, Dated June 27,1996, 8.5 Byron Letter, BYRON 96-0266, Operating Interval Between Eddy Current Inspections for Circumferential Indications in the Byron Unit 1 Steam Generators, Dated October 18,1996 8.6 J. Hosmer Letter to USNRC, Response to Request for Additional

.. . _ _ -- - - - - - - - - - - - - - - = = - - ---

,! l Information Pertaining to the Byron Cycle Length Byron Station Unit 1,  ;

_ Dated December 2,1996 8.7 J. Hosmer Letter to USNRC, Byron Unit 1 Cycle Length, December 20, 1996 8.8 USNRC Letter to Ms. Irene Johnson, Extension of Byron Unit 1, Operating Cycle Between Steam Generator Tube Eddy Current Inspections, Dated March 28,1997 8.9 Byron Nuclear Design Information Transmittal, NDIT BYR97-263, Dated 6/24/97 8.10 J. Hosmer to USNRC, DEI Tech Spec Change, Dated 1/31/97 8.11 NRC Generic Letter 95-05, Voltage Based Repair Criteria for 4

Westinghouse Steam Generator Tubes Affected by Outside Diameter Stress Corrosion Cracking 8.12 Letter from D. Saccomando to USNRC, Supplement to Application for Amendment to Facility Operating Licenses, Dated September 1,1995 8.13 USNRC Letter to D. Farrar, Issuance of Amendments for 3.0 Volt IPC, 1

hovember 9,1995

w

~

3.0 Vcit IPC Vcitage Dependent Growth Mcthodol:gy --

Byron Unit 1 Cycle 8 Figure 1 v.

E u

Ind Voltage Ind Voltage Ind Voltage

{

• BOC7B BOC7B BOC6 Distribution Distribution Distribution .

v9 < 0.5 Volts 0.5 - 0.9 Volts > 0.9 Volt

$ Y Y Y

=

0

~

=- Ind Voltage Ind Voltage Ind Voltage Growth Growth Growth n

v.

Distribution Distribution Distribution l 1 Leak Rate S + = Taken at

[ k k k /

Ind Voltage 95% Confidence k- Ind Voltage Ind Voltage Ind Voltage

[-

3 EOC 8 Distribution EOC 8 Distribution EOC 8 Distribution EOC 8 Distribution (Predicted) (Predicted) (Predicted) (Cumulative)

(Predicted) l

Figure 2 ,

Byron Unit -1 Cycle 7B Growth for Three BOC Distributions - SG-B (plus Largest growth in SG-C)

Cumulative Probability Distributions 3% s_ c. e, f_g f_[ e itC--m--*"' .

1. ~

g o D~ ,..w n-0.9 - y- .=- .

,m- a W 9_g. H .d 0.8 - c .c

/

/ .

W~W 0.7 - ,1 g /

5 0.6 -  ! .

/  : Up to 0.5 volt (826) -

g a

o .i a . g i:m 02 - V 3 .

/ - C- 0.5 to 0.9 volt (657) <

.t a W 0.4 - .

/ - -

E f d / . - *-- Over 0.9 volts (264) 0.3 - d #

1 1

E

/

-1 -

0.2 - d-/

/g i .

p!

0.1 -  ;- .d d

/

0'0 !

y Q 7 2 2 2 0 0

~

O 2 E O  % 2 O O  % 2 O I O Voltsge Growth / Days > 500F

2# - . m , ,. =

. .~ . ... ... . . .- -.

$2-ll*

Figure 3 Braidwood Unit -1 Cycle 6 Growth for Three BOC Distributions (SG-C + Largest Growth in SG-A) CPDF Distributions ,

hh[ bbb , , . -

~

0.9 - fr Ir *~x-x-x-x-x"*

• ,g. ,-

.4 cr

/ 5 x- x x-0.8 - -

d x-x'

/ **

Y y'

-G 0.7 - -i ,

• 5 1 y E d .-

• 0~6 - #

• 8 l E -*

a 5 5 0.5 ,'  : Up to 0.7 volt (1230) -

.n I Q l i

? I .-

-- 0'4 - ' --

.5 -

- c- 0.7 to 1.1 volts (665)

= F G

o .

U 0.3 .-

x

-- * --Over 1.1 volts (208) 0.2 - -

0.1 -

/ *

. * *h.

').0 d ., . . : '

- - - - n . --

=

=

n

=

=

- , ~ n , , .'

4 4 4 6 6 6 6 6 - a a n a a , a s a-Voltage Growth / Days > 500F a u e_=v_ . . .m.u u -

st-

~~

Braidwood Predicted EOC-6 Voltage Distribution Using Cycle 6 Voltage Dependent Growth and Cycle 6 As-Found Voltage Distribution Figure 4 20 O EOC-6 Predicted

, EEOC-6 Actual -

15 -- -

_E o

g 10 - -

.o E

5 5- -

l 0'.. ........ . . .. . ... . .

c. I' "- -

'?- 'I Ub '

ososococosassoooooooooooooo Voltage, volts 7/14/9711:26 AM Eoedistr.xts ch br6 br6 grwth

s:. -- .

)

. i Byron EOC-8 Voltage Distribution Using Byron Cycle 7B Voltago Dependent Growth Rates for  !

500 Days and 90 Day Report Results (600 Days)

Figure 5 l 8

D EOC-8 Byron Cycle 78 Voltage Dependent Growth E 90 Day Report Predictions

~

I  !

6 E  :

o

~~

t 3

E  !

-[-4 o

D n ,

E  !

m i z

2 -

i t

0 , , ,,, , , , ,

I I I ,

bNSA' - '**- '

'.; ' ' l I SEES I I I, ' ', , ,

a n e m N e m - ,e s e n o cn m e m - e s <n n o e m o m - v 5 5 5 4 4 4 5 5 5 5 5 e & 6 & 5 m & 6 6 & o g g g g i Bobbin Voltage, volts 7/18/9710:17 AM -ME135F.XLS ch by8 by7bvondepdgrwth

_ _ _ _ _ - _--