ML20216E495

Provides Suppl Info to RAI Re TS Change Request 203 for TS 24-month Surveillance Extensions

ML20216E495
Person / Time
Site:	Oyster Creek
Issue date:	08/29/1997
From:	Roche M GENERAL PUBLIC UTILITIES CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
6730-97-2219, TAC-M96906, NUDOCS 9709100208
Download: ML20216E495 (10)

Text

{{#Wiki_filter:e4

• I

,0 .,8 l GPU Nuclear. Inc. ( U S. Route #9 South NUCLEAR Post othee Box 388 ) Forked River. NJ 06731-0388 Tel 609-971-4000 August 29,1997 6730-97-2219 U. S. Nuclear Regulatory Commission Attn.: Document Control Desk Washington, DC 20555

Subject:

Oyster Creek Nuclear Generating Station Operating License No. DPR-16 Docket No. 50-210 Technical Specification Change Request No. 203 Tech. Spec. 24-Month Su. eeillance Extensions Supplemental Information Requested by NRC (TAC No M96906) 30k On May 9,1997 and August 14,1997 telephone conferences were hem between GPU Nuclear (GPUN) Inc. and the Nuclear Regulatory Commission (NRC). During the call, the NRC's Staff requested additional information to supplement the GPUN response letters dated March 25,1997 g and June 6,1997 (GPUN Letter Nos. 6370 97-2028, and 2165 respectively) to NRC's Request I \\ for AdditionalInformation(RAI). The requested information to supplement the RAI response is hereby provided by Attachment to this I,tter, as mutually agreed upon during the telephone conference call. This information is for clarification only and does not impact upon the safety evaluations or the guidance of GL 91-04, as applicable. ~ 9709100208 970829 PDR ADOCK 05000219 g tj kk k FDs No. 96069

6730 97-2165 TSCR No. 203 RAI Errata Page 2 - Pursuant to 10 CFR 50.91 (b)(1), a copy of this RAI response supplement has been sent to the State of New Jersey Department of Environmental Protection. If you have any questions - concerning this matter please contact Sr. Licensing Engineer Mr. G. M. Gurican at (201) 316 7972. r-Sincerely, Michael B. Roche Vice President and Director Oyster Creek t MBR\\gmg Attachment cc: NRC Project Manager Administrator, Region i NRC Sr. Resident inspector

1- -

4 4 Fue No. 96069.

,a- -{ GPU Nuclear. Inc. g-U.S. Route #9 South - NUCL AR '**'0"***3** Forked River. NJ 087310388 Tel 609-9714000 ~ August 29,1997 6730-97-2219 Mr. Kent Tosch, Chief Bureau of Nuclear Engineering . Department of Environmental Protection CN 411 Trenton, NJ 08625

Subject:

Oyster Creek Nuclear Generating Station Operating License No. DPR-16 Technical Specification Change Request No. 203 SupplementalInformation Enclosed herewith is one copy of the information which supplements GPUN's response to NRC's Request for Additional Information (RAI) on TSCR No. 203, for the Oyster Creek Nuclear Generating Station Operating License. The TSCR No. 203 RAI response was filed with the United States Nuclear Regulatory Commission on March 25,1997; and an RAI Errata was filed on June 6,1997. Sincerely, % LAMA Michael B. Roche Vice President and Director Oyster Creek MBR/gmg - Attaciunent - Fue No. 96069

__m.. _. - _. _ _ - - _. - _ _. - - ~ _ _ _.. _.. _.. _ _.. _. _._..... _. _ _. i i 4 l l GPU Nucioer,Inc. (. U.S. Route #9 South NUCLEAR P**' 0"* ' * ** 3** Forked River, NJ 067310388 Tel 609-9714000 e Augest 29,1997 6730-97-2219 i The Honorable L. Nick. Mayor of Lacey Township. 1: 818 West Lacey Road [ ' Forked River, New Jersey 08731

Subject:

Oyhter Creek Nuclear Generating Station l Operating License No. DPR 16 Technical Specification Change Request No. 203 SupplementalInformation i l Enclosed herewith is one copy of the information which supplements GPUN's response to the NRC's Request for Additional Information (RAI) on TSCR No. 203, for the Oyster Creek Nuclear Generating Station Operating License. The TSCR No. 20a RAI response was filed with the United States Nuclear Regulatory L Commission'on March 25,1997; and an RAI Errata was filed on June 6,1997. it i . Sincerely, M -h - Michael B. Roche i Vice President and Director y Oyster Creek MBR/gmg 5 Attachment 'Fue No. 96069

t 1 t ATTAClIMENT Oyster Creek Nuclear Generating Station (OCNGS) Operating License No. DPR-16 Docket No. 50-219 Technical Specificatior Change Request No. 203 SUPPLEMENTAL INFORMATION

1. SLOPE OF THE REGRESSION LINE BASED ON SURVEILIANCE DATA: Calculated parameters such as correlation coemeient, slope r.ssociated with instmment channels and sample size are listed below. These values are from a calculation C-1302-640-5350 009, Rev.1 instrument - Calculated Slope Tag No. Function /Vac Correlation-(X coemeient) Sample Size Trip System No. Coemeient (r) (b) (# of Data Points) LVA-1 Alarm /l 34.24 % -0.0004136 39 RSCSll Trip /l *(VTl) 35.98 % -0.0003585 39 RSCS 12 Trip /"(VTl) 37.92 % -0.003795 39 LVA-2 Alarm /2 5.22 % 0.0001026 33 RSCS-21 Trip /2*(VT2) 2.16 % 0.0000365 31 RSCS-22 Trip /2"(VT2) 3.46 % 0.0000584 31 VT-1 Turbine Trip 31.30 % 0.0004589 14 VT-2 Turbine Trip 6.53 % -0.0001055 10 Part of RPS Trip System 1 Part of RPS Trip System 2 VT1 Vacuum Trip System 1 VT2 Vacuum Trip System 2 Of the above, the Technical Specification instruments include: RSCS-11, RSCS-12, RSCS-21, RSCS-22. The correlation coemcient is not significant for any of the instruments including RPS instrumentation. The slope is indicative of dria. As cc.n be seen from the data on the Table above for instruments in the RPS trip system the slope is practically ZERO indicating small or nearly no drin at all. 2, DISCUSSION OF TiiE GENERAL METIlODOLOGY OF STATISTICAL ANALYSIS IN Tile ORIGINAL TSCR SUBMITTAL The GPUN description of the methodology of the statistical analysis in TSCR203 submittal and calculation C-1302-640-5350-009 Rev.1 interchangeably urca use 95% confidence interval as a " Tolerance Range." A Staff concern was that the " tolerance" represents a 95% confidence with a 95% proportion. The intent of GPUN's description of tolerance in the TSCR methodology and calculation is to represent an engineering tolerance since the calculated 95% confidence is comparable to the procedural tolerance which is as-found acceptance criteria. Use of the term tolerance range, meaning engineering tolerance range, is identical to the calculated 95% confidence interval. The TSCR 203 RAI Errata submitted on June 6,1997 by GPUN Letter No. 6730-97-2165 reflects the use of a confidence interval, where applicable, in place of tolerance range. However, the calculation which was submitted will not be revised at this time to reflect this wording change. l l

3. DERIVATION OF Tile SAMPLE SIZE 8 The low vacuum trip input to the reactor trip system comes from four vacuum switch assemblies (VSAs): RSCS ll, RSCS 12, RSCS-21, and RSCS-22. Two sequential numbers in the suflix represent vacuum trip system identification number and RPS trip system identification number, respectively. Calculation C-1302-640-5350 009 Rev. I Sheet 13 of 13 shows a pictorial configuration of the switch-bellows combination. Accordingly, the RSCS-ll and RSCS 12 both are a part of vacuum trip system # 1 and RPS trip systems 1 and 2 respectively, and RSCS 21, RSCS 22 are part of vacuum trip system # 2 and RPS trip systems 1 and 2 respectively. Vacuum from each condenser is fed to two bellows: l A & 2A bellows which are connected to A-condenser,1B & 2B bellows to B-condenser and IC & 2C to C-andenser. All 1 and 2 bellows are associated with vacuum trip s3 h # 1 & 2 respectively. Any of the l A, IB or 1C bellows, on low vacuum, win _auate VSAs RSCS-11 and RSCS-12 in RPS Trip System 1 & 2 respectively. Similarly, on low vacuum, any one of the 2A,2B or 2C bellows will actuate VSAs RSCS-21, RSCS-22 in RPS Trip System 1 & 2 respectively. During surveillance, each bellows in a vacuum trip system will actuate two switches in different RPS Trip System 1 & 2. During surveillance, all three bellows in a given vacuum trip system are maintained at 28" Hg vacuum and one out of the three is bled slowly to the trip setpoint. Thus at each surveillance, each switch gets checked three times, once for each bellows. If a switch actuation setpoint associated with a given bellows is outside the as-left tolerance, it is adjusted to as-left criteria. Every time a switch is adjusted with respect to a bellows, the nm time for that switch-bellows combination is set to zero (0). If a switch was adjusted at the previous surveillance its run cycle was set to zero then and at the next surveillance it was again adjusted, in such case there will be two sample or data points. Surveillance (0) As-len of run cycle (1) at time zero O ltput, y (1) As-found of runcycle (1) (1-R) JL (1 R) Reset value corresponding to (1) X(1)- X(0) = Run Cycle I in days (0) Ely. Ik As-lef) y II Target or Criteria Desired Setpoint y __ _bi g(1) x, days X(0) x(1) Process Run Cycle No.1, Days

1. of data points = 2, i.e.,(0) & (1)

'If a switch is not adjusted at the next two surveillances but was reset at the third surveillance, the process cycle has a longer continuous run and in that event there will be f6ur (4) sample points for that run cycle. See examples below. (1 R) As-left of nm cycle (2) at time zero ' Surveillance (2) As-found of runcycle (2) (2 R) Resetvaluecorrespondingto(2) (1-R) jfpu,y M2). M0) = Run Qcle 2 in days (S l (2) - 'I' (SI) J L As-left y Target or Criteria Desired II q. 4I(S2) U Setpoint (2-R) m ) ( ) ist Surveil. 2nd Surveil. 3rd Sun. x days No Adjustment No Adjustrnent Reset to (2 R) Process Run Cycle No. 2, Days

1. of data points = 4, i.e., (1-R), (SI), (S2) & (S3) i J

1,

1 4. CONSERVATISM IN TIIE CALCULATED RESULTS:

1.,

Data for a five year period from February 1989 to November 1994 was used for the analysis. I 2. The continuous run cycle duration for the last of the analyzed data sets of all 12 switch bellows in combination was 664 days, and this did not violate the as-found acceptance criteria. A duration of 645 days and 595 days of continuous run cycle 3 was experienced by 10 switch bellows sets and 2 switch-bellows sets, respectively, without violating the as-found acceptance criteria. 3. The predicted confidence interval at 30 months and elso at 36 months is within the as-found acceptance criteria. Sloye of the regression line based on the actual surveillance data is in the order of 4. 10 which justifies use of a mean modelindicating no drift. Use of a regression model with a 95% confidence interval is more conservative. In the confidence inten al calculation, there is an additional consenatism as discussed below: Predicted Confidence Intenal X days i = { Regression line at Xi days} 1 {to.a.on

• Std Deviation of predicted value @ Xi days}

= (a + b*Xi } {t<o.93,on * [s'(P,)f }, where s'(P,) = s'(p){l + 1 + Z ( A ' A), } = s#(y) + s'(j) + s#(b) * ( N where: s'(P,) = Variance of the predicted value s#(}) = Variance of the estimate s'(}) = Variance of } s'(b) = Variance of the slope, i.e., b Thus the variance of the predicted value is enlarged by the sum of: Variance of the estimate, Variance of } and Variance of the slope multiplied by the square of the deviation of the predicted time from the average value of X. ]

.l 4 ) +4 Surveillance

[

Output, y j l3 ~ +1 j / ~ Regression Line 7 3 2 -3 i l u j 4 4 4 m r 4 X(0) X bar-xi_ x, days 1-Avg X i* i Days l-In the figure above, 11 = Variability around regression line due to standard deviation of the estimate j i2 = Variability around regression line due to square root of (variance of the estimate + variance of y } : 1_ 13 = Variability around regression line due to square root of { variance of the estimate + variance of y + (variance of slope) x (square of the deviation from the X of time for which the prediction is made}} 7-l- 14 = 95% Confidence Interval around regression line l- = tg3,on x (13 from directly above) i e [ .-.}}