ML20138L176

Requests Approval for Use of Voltage Dependent Probability of Detection for Application in SG Alternate Repair Criteria Leak & Burst Projections

ML20138L176
Person / Time
Site:	Beaver Valley
Issue date:	02/13/1997
From:	Jain S DUQUESNE LIGHT CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9702210128
Download: ML20138L176 (4)

Text

e 4

vet VaRoy Power Station Shippingport, PA 16077-0004

$USHIL C.. LAIN (412) 393-5512 OMe60n V6w Prooksent i fax (412) 643 8069 UiN.N*o*ivwon February 13, 1997 i U. S. Nuclear Regulatory Commission i I

Attention: Document Control Desk Washington, DC 20555-0001

Subject:

Beaver Valley Power Station, Unit No. I and No. 2 BV-1 Docket No. 50-334, License No. DPR-66 BV-2 Docket No. 50-412, License No. NPF-73 l Voltage Dependent Probability of Detection The purpose of this submittal is to request NRC approval for use of a voltage dependent probability of detection (POD) for application in steam generator alternate repair criteria (ARC) leak and burst projections. Technical Specification Amendment No.198 implemented the Generic Letter 95-05 ARC requirements for Beaver Valley l Power Station Unit I which provide for a fixed 0.6 POD. Section 3.3 of the Unit 1 l "1996 Steam Generator Tube Inspection 90 Day Report," provided by letter dated August 2,1996, provides an evaluation of the voltage dependent probability of prior cycle detection (POPCD) and concludes that the fixed POD should be upgraded to a voltage dependent POD.

Section 3.3 of the 90 day report has been updated and restated in Attachment 1.

l 'Diis request is consistent with Generic Letter 95-05 Section 2.b.1 which provides for an attemative POD function as long as the function is approved by the NRC.

]

Approval for use of this voltage dependent POPCD is requested for September 5,1997, for Unit I and by March 7,1998, for Unit 2 for use in the respective refueling outages.

ARC is not yet approved for Unit 2; however, our request was provided by letter dated June 18,1996, which requested ARC approval by the above date.

Sincerely, 1

I Sushil C. Jain [ <

c: Mr. D. M. Kern, Sr Resident Inspector DE W AlNS Mr. II. J. Miller, NRC Region I Administrator 0UAlIII Mr. D. S. Brinkman, Sr. Project Manager ENERGV 9702210128 970213 PDR ADOCK 05000334 P PDR

l ATTACHMENT I Beaver Valley Power Station, Unit No. I and No. 2 Voltage Dependent Probability of Prior Cycle Detection (POPCD)

The inspection results at a given end-of-cycle (EOCn) permit an evaluation of the probability of detection at the prior EOC (EOCn.1) inspection. For ARC applications, the important indications are those that could significantly contribute to EOC leakage or burst probability. These significant indications can be expected to be detected by bobbin and confirmed by RPC inspection. Thus, the population of interest for ARC POD assessments is the EOCnRPC confirmed indications that were detected or not detected at the prior EOCn.i inspection. The POPCD can then be defined as:

EOCnRPC Confirmed and + EOCn-1 RPC Confirmed Detected at EOCn.1 and Plugged at EOCn-1

=

POPCD(EOCn-1)

{ Numerator} + New EOCnRPC Confirmed Indications (i.e., not detected

! at EOCn.i)

POPCD is evaluated at the EOCn.1 voltage values since it is an EOCn-1 POPCD assessment. The indications at EOCn.1 that were RPC confirmed and plugged are included as it can be expected that these indications would also have been detected and confirmed at EOCn. It is also appropriate to include the plugged tubes for ARC applications since POD adjustments to define the beginning of cycle (BOC) distribution are applied prior to reduction of the EOC mdication distribution for plugged tubes.

Indication in tubes deplugged and returned to service at EOCn are not included in the evaluation since it is not appropriate to include them.

It should be noted that the above POPCD definition includes all new EOCn

, indications not reported in the EOCn-1 inspection. The new indications include EOC n-1 indications present at detectable levels but not reported, indications present at EOCn.1 below detectable levels and indications that initiated during Cycle 'n'. Thus, this definition, by including newly initiated indications, differs from the traditional POD definition. Since the newly initiated indications are appropriate for ARC applications, POPCD is an acceptable definition and eliminates the need to adjust the traditional POD for new indications.

The above definition for POPCD would be entirely appropriate if all EOCn indications were RPC inspected. Since only a small fraction of bobbin indications (e.g.,

> 2.0 volts) are RPC inspected, POPCD could be distorted by using only the RPC inspected indications. Thus, a more appropriate POPCD estimate can be made by assuming that all bobbin indications not RPC inspected would have been RPC confirmed. This definition is applied only for the EOCn indications not RPC inspected since inclusion for the EOCn.1 inspection could increase POPCD by including indications on a tube plugged for non-outside diameter stress corrosion cracking

i i .

. Attachment i Voltage Dependent Probability of Detection Page 2 i

(ODSCC) causes. This POPCD can be obtained by replacing the EOC, RPC confirmed l 4

by RPC confirmed plus not RPC inspected in the above definition of POPCD. Inclusion l' of the indications not RPC inspected in POPCD primarily influences detectability below the threshold for RPC inspections (currently 2.0 volt for 7/8" tubing). The only bobbin

' indications excluded from the POPCD evaluation are those RPC inspected but not confirmed as a flaw indication.

At this time, POPCD evaluations are available for fifleen inspections of eight plants. The available data include ten inspections of plants with 7/8" diameter tubing i

(including BVPS #1) and five inspections of plants with 3/4" diameter tubing. The

POPCD evaluations performed since 1992 show significant improvement over the earlier ,

l assessments which represent the first ARC inspections. Thus, it is appropriate to assess POPCD for inspections performed since 1992. Eleven of the fifteen inspections for which POPCD has been evaluated were performed since 1992.  ;

j The Electric Power Research Institute (EPRI) initiated the development of a l database report EPRI-7480-L, Addendum 1 S$50-17 Final Report November 1996 titled, l

" Steam Generator Tubing Outside Diameter Stress Corrosion Cracking at Tube Support l Plates Database for Alternate Repair Limits," to compile industry data on this subject.

i This report has been submitted to the NRC for review; therefore, the figures and tables l l

identified in the discussie i that follows are from that report.

l 9

Figure 7-2 (based on data of Table 7-2) shows the combined POPCD evaluation for plants with 7/8" diameter tubing and includes results for six inspections performed since

1992. The POPCD evaluation is in good agreement with the EPRI POD except in the

one to two volt range where POPCD is about 0.83 and the EPRI POD increases from about 0.82 to about 0.97. For the data of Figure 7-1, a POD approaching unity is supponed above 3 volts. The average POPCD independent of voltage is about 0.67. The definition of POPCD includes indications which were not present at the prior inspection and thus could be expected to be somewhat lower than the EPRI POD which is based on

" expert" evaluations ofinspection results and does not include indications clearly below detectable levels.

The combined data for the eleven inspections since 1992 are given in Table 7-3 and the POPCD evaluation is shown in Figure 7-3 for RPC confirmed plus not inspected indications. It is seen that the inspections since 1992 yield a POPCD in good agreement with the EPRI POD which was a 1994 evaluation. POPCD supports a POD approaching unity at about 3.5 volts while the EPRI POD is about 0.98 at 2 volts and unity at 3 volts.

Figure 7-3 also includes POPCD evaluated at the lower 95% confidence limit on the data forindividual voltage bins.

Attachment I  !

Voltage Dependst Probability of Detection [

Page 3 ,

i The POPCD evaluations shown in Figures 7-1 to 7-3 are based on the definition of c " truth" as RPC confirmed plus not RPC inspected indications. Since many of the ,

indications not RPC inspected would be expected to be found with no detectable i

! degradation (NDD) if inspected, this represents a lower bound POPCD evaluation. j Figure 7-4 shows the POPCD evaluation for all eleven inspections since 1992 based only ,

on RPC confinned indications. This results in a significant increase in POPCD below l 1.0 volt and a modest increase above 1.0 volt. The data of Table 7-3 show 400 to 5700  ;

indications in all, voltage bins below 2 volts, more than 200 between 2.0 and 3.2 volts l and about 30 indications above about 3.2 volts. Thus, the collective data provide a substantial database for defining a POD. l The results of Figure 7-3 clearly support an increase in the POD for ARC applications above the POD =0.6, independent of voltage, required by the NRC Generic l

Letter 95-05. For indications above 1.0 volt, the POD exceeds 0.9 and is 0.97 to near unity at 2.0 volts. A POD of 0.6 is only applicable to indications below about 0.6 volts.

! A recommended voltage dependent POD is developed as POPCD evaluated at the i lower 95% confidence level and the mid-voltage of each voltage bin. The result is then l

smoothed to obtain the recommended POD as shown in Figure 7-5. The recommended l 1 POD is tabulated in Table 7-4 and compared with the EPRI POD in Figure 7-6.

l The POPCD evaluations for eleven inspections since 1992, together with the EPRI

' POD evaluation provide a database for updating the NRC generic letter requirements on POD. The POD of Figure 7-5 and Table 7-4 is recommended for ARC applications as a .

, replacement for the constant POD of 0.6. i h

l l

4 t

i 5

. , _ . - ,_