A02724, Forwards Proprietary & Nonproprietary Responses to NRC 820819 Request for Addl Info Re Measurement Uncertainties & Shape Annealing Factor Component of Axial Shape Index.... Proprietary Response & Rept Withheld (Ref 10CFR2.790)

Forwards Proprietary & Nonproprietary Responses to NRC 820819 Request for Addl Info Re Measurement Uncertainties & Shape Annealing Factor Component of Axial Shape Index.... Proprietary Response & Rept Withheld (Ref 10CFR2.790)

ML20074A103
Person / Time
Site:	Millstone
Issue date:	05/03/1983
From:	Counsil W, Sears C NORTHEAST NUCLEAR ENERGY CO., NORTHEAST UTILITIES
To:	Clark R Office of Nuclear Reactor Regulation
Shared Package
ML19301C241	List: A02724, Forwards Proprietary & Nonproprietary Responses to NRC 820819 Request for Addl Info Re Measurement Uncertainties & Shape Annealing Factor Component of Axial Shape Index.... Proprietary Response & Rept Withheld (Ref 10CFR2.790) ML20074A106
References
A02724, A2724, TAC-48063, NUDOCS 8305120241
Download: ML20074A103 (16)

Text

r NORTHEAST tlTILITIES Gml OHiceo Selen Sum Bemn, Connemt

+.ca w m oo e o =c "'"

P O. BOX 270

(('7""((cQ',7'""" H ARTFO A D. CONNECTICUT 06141-0270

.o.w . uu u *a c:. . (203) 666-6911 L ' J ow . w.a i.m. cc-.e May 3,1983 <* g Docket No. 50-336 U A02724 ,

'll. N -

'~

% 75 Director of Nuclear Reactor Regulation Co f.

Attn: Mr. Robert A. Clark, Chief 6 Operating Reactors Branch #3 U.S. Nuclear Regulatory Commission Washington, D.C. 20555

References:

(1) W. G. Counsil letter to R. A. Clark, dated March 4,1982 (2) R. A. Clark letter to W. G. Counsil, dated August 19,1982 (3) W. G. Counsil letter to R. A. Clark, dated November 4,1982 (4) W. G. Counsil letter to R. A. Clark, dated February 22,1983 Gentlemen:

Millstone Nuclear Power Station, Unit No. 2 Request for Additional Information, Measurement Uncertainties Response to Question 6 In Reference (3), Northeast Nuclear Energy Company (NNECO) provided a partial response to the Staff's Reference (2) request for additional information concerning measurement uncertainties utilized in the Millstone Unit No. 2 safety analyses. Additional time was necessary to complete our responses to Questions 4 and 6 of Reference (2). Our response to Question 4 was docketed in Reference (4). Our response to Question 6, as per our mutually agreed upon schedule, as documented in Reference (2), is presented herein.

The results of the review of the total Axial Shape Index (ASI) uncertainty t

)g allowance have been completed and are attached. Details have been provided concerning uncertainties associated with plant calibration procedures, process A

equipment, the shape annealing factor, incore monitoring methodology, and ASI s separability. Results of this review indicate that these uncertainties result in a total uncertainty less than that reported in Reference (1). Thus, this information continues to support the measurement uncertainties utilized in the Millstone

,o" Unit No. 2 safety analyses.

Ob

$ Attachments I and 3 present the non-proprietary and proprietary versions, respectively, of the Total Axial Shape Index Review. Portions of the material in g

Attachment 3 are proprietary to both the Westinghouse Electric Corporation and Combustion Engineering, Incorporated. As discussed in telephone conversations a with your Staff, NNECO has included seven copies of this proprietary informa-

%] tion and requests that it be withheld from public disclosure in accordance with the provisions of 10 CFR 2.790 and that this material be safeguarded. The w reasons for the classification of this material as proprietary is delineated in the affidavits included in Attachment 3. 8305120241 830503 PDR ADOCK 05000336 P PDR t

In Reference (3), NNECO indicated that an evaluation of the feedwater measurement system was underway. Since that time NNECO has completed the evaluation and hereby presents a summary of this in Attachment 2. Results of the evaluation demonstrated that a reduction in temperature span in the main feedwater temperature measurement channel will provide a more accurate indication of feedwater temperature measurement. This was accomplished by recalibrating the temperature transmitter. The resulting increase in accuracy then provides smaller core power, RCS flowrate, neutron and delta-T power uncertainties. These results supercede those provided in Reference (4).

Attachment 2 also provides details concerning our Reference (4) commitment to inform the Staff of the results of NNECO's QA-verification of our response to Question 4 of Reference (1). This QA-verification was recently completed and the results support our conclusions of Reference (4). Quality assurance has, however, determined that slight differences exist with regard to the tabulation of historical drif t data used for justification of the instrument span drif ts as discussed in Reference (4). The QA-verified results indicate that two of the instrument span drif ts are, in fact, smaller than those values reported in Reference (4).

We trust you will find this information responsive to your Reference (1) request.

Very truly yours, NORTHEAST NUCLEAR ENERGY COMPANY LO . b . by W. G. Counsil Senior Vice President

, L By: C. F. Sears Vice President Nuclear and Environmental Engineering l

Docket No. 50-336 Attachment 1 Request for Additional Information Measurement Uncertainties Response to Question 6 (non-proprietary version)

May,1983

The following provides the response to question 6 of the NRC letter from R. A.

Clark to W. G. Counsil, Northeast Nuclear Energy Company, dated August 19, 1982.

QUESTION 6: Provide results of the review of the total axial shape index uncertainty allowance.

RESPONSE

The Millstone Unit No. 2 monitoring and protection systems utilize both incore and excore detectors to monitor core axial power shape. The parameter monitored is called the Axial Shape Index (ASI) and is mathematically defined by:

le = L-U L+U Y i= a

• Ie + b where le = excore axial shape index YI = monitored axial shape index L = lower excore nuclear instrument detector signal U = upper excore nuclear instrument detector signal a = shape annealing factor b = bias term Incore detector derived ASIS are used in Technical Specification 3.2.6 to provide initial conditions for the accident and safety analyses and to ensure that the assumed margins to DNB are maintained. Excore ASI is used as input to the Local Power Density trip which ensures that the peak local power density which corresponds to fuel centerline melting will not occur as a consequence of an axial power maldistribution. Excore ASI is also an input parameter to the Thermal Margin / Low Pressure trip which prevents operation when the DNBR is less than 1.30. The excore detector system can be utilized as described in Technical Specification 3.2.1 to monitor linear heat rate. The excore detector ASI is calibrated to the measured incore value via Millstone Unit 2 plant procedures.

Limiting conditions for operation and limiting safety system setpoint analyses assume core average axial power shapes (characterized by ASI) in their derived setpoints. The methodology for these setpoints is described in WCAP-9660(l).

There are five contributors to the ASI uncertainty, each of which is described in the following sections. The combination of the five uncertainties is provided in Section VI of this response.

1. CALIBRATION ALLOWANCE The Calibration Allowance is an Instrument uncertainty directly related to on-site, plant calibration procedures. This allowance includes calibration accuracy, drift, and voltage setting tolerances for a voltage sibnal from the excore detector through the output of the linear amplifier.

Using the data provided in Reference 2, Question 1, Section 11 (for upper and lower detector voltage calibration errors) and Reference 3, Tables 10 and 12 (for ASI drift allowance) it has been determined that a normal distribution can be used to describe the Calibration Allowance. The data provided in Table 12 of Reference 3 allows for the determination of a two sided, 95% probability /95% confidence limit value for the standard devia-tion of the ASI drif t allowance. The value of Ks determined is 0.00917 ASIU. Combining this allowance by Root Sum of the Squares (RSS) with the upper and lower detector voltage calibration errors results in a total uncertainty (1 Ks) of 0.0093 ASIU. However, a total uncertainty of 0.01 ASIU, with a degrees of freedom of 139, is used in this evaluation for conservatism. Table I provides a summary of the above.

11. EQUIPMENT PROCESSING ALLOWANCE The Equipment Processing Allowance quantifies the effect of the process-ing equipment upon ASI as well as the uncertainty in setting of trip or alarm values in the Reactor Protection System and the Power Ratio calculators. This allowance accounts for instrumentation or rack toler-ances from the output of the linear amplifiers through the trip bistables or LCO annunciators. Also included are calibration and drif t allowances for the signal comparators and setpoint settings. The allowance used in this evaluation is + 0.02 ASIU. The uncertainty distribution can be represented by a normal dI'stribution and the 0.02 ASIU is the equivalent i Ks value at a 95/95 level.

Ill. SHAPE ANNEALING FACTOR ALLOWANCE (SAF)

The Shape Annealing Factor (a) is an experimentally measured value which relates the peripheral shape index (Ip) to the measured value of the external shape index (le).

Ip = a*le The SAF uncertainty allowance accounts for the ability to measure the SAF experimentally on-site from xenon oscillation experiments. The Combustion Engineering report CEN-247 (N) is attached as Appendix A to provide a detailed discussion of this uncertainty allowance. CE has determined that the uncertainty distribution of this parameter can be represented by a normal distribution. CE also notes that at a 95/95 level the uncertainty (1 Ks) is 0.068 *Ip. Assuming a range in Ip = 1 0.2 ASIU results in a Ks value of 0.0136 ASIU. A value of 2 0.015 ASIU has been used in this evaluation for conservatism.

_4 IV. INCORE MONITORING METHODOLOGY UNCERTAINTY In addition to statistical fluctuations and instrument uncertainties which are evaluated in the previous sections, the use of four section fixed incore detector strings to reconstruct ASI can introduce a component of error.

To determine this error term, ( ) +a,b,c axial shapes were reconstructed utilizing the INCA algorithms. The mean error, standard deviation, and distribution type were determined. The data base is the full power subset of those shapes previously used in WCAP-9660-Addendum 1 W) to deter-mine the Fq uncertainty. The full power subset is used because the error values are more conservative than for all cases together. Figure 1 provides the result. The mean error is ( )+a,b,c and a standard deviation of

( )+a,b,c. The histogram of errors shows a range of +1 to -1.5 and can be characterized as normal with a mean of ( ) +a, b, c and a standard deviation of ( ) +a, b, c. The r, umber of degrees of freedom is ( )+a,b,c with N equal ( )+a,b,c. The two sided, 95/95 tolerance factor K equals ( )+a,b,c and Ks1 equals ( )+a,b,c.

V. SEPARABILITY This component of the ASI uncertainty accounts for the relationship between the core average ASI used in the safety analyses and the peripheral fuel assembly shape indices which directly affect the excore detector response. (See Figure 2.) The major subcomponents of this uncertainty are control rod shadowing effects, variations in the relation-ship between core average ASI (1) and peripheral ASI (Ip ) caused by off nominal conditions such as xenon oscillations, and the ability of the design methods themselves to correctly account for the nominal relationship between I and Ip.

The differences in peripheral assembly ASI's relative to core average ASI has been evaluated by performing a large number of 3-D neutronics calculations. Cycles 4 through 6 of Millstone Unit 2 were evaluated.

Power levels between 90% and 100% were modeled using various control rod depletion histories consistent with Technical Specification 3.1.3.6, CEA power dependent insertion limits (PDIL). Xenon oscillations were induced at the most limiting times in the cycle with respect to DNB. Control rod shadowing and xenon oscillation effects are thus implicitly included in these evaluations.

The results of these calculations provide the differences between core average and peripheral assembly ASIS. The results are summarized in Table 2 and Figure 2 provides the core map and excore detector locations.

The excore detector response will be related to a weighted sum of

peripheral assembly powers. It is conservatively assumed that the ASI variation in the " weighted" detector response assembly equals that of the population of peripheral ASI variations. Figure 3 compares the histogram of % ASI differences to a normal distribution with the same mean and standard deviation. The peripheral assembly power distributions are generally more bottom skewed than the core average power shapes. This y --- - . - -, ,--

bias is accounted for in the limiting conditions for operation and the limiting safety system setpoint analyses and is also measured as part of the determination of the shape annealing factor. (This is the bias term b as discussed in the Introduction.)

For this uncertainty component, it is conservative to assign a normal distribution with a standard deviation of ( ) +a,b,c and ( )+a,b,c degrees of freedom with N equal ( ) +a,b,c. The two sided, 95/95 tolerance factor K equals ( ) +a,b,c and KsSAF equals ( ) +a,b,c.

VI. ALLOWANCE AND UNCERTAINTY COMBINATION Table 3 summarizes the information noted in Sections I through V concerning means, standard deviations, degrees of freedom, and determined values for tolerance factors for those uncertainties where a Ks value must be calculated. In all cases, the population distributions are known to be normal or near normal and the Ks values are determined to be the standard deviation at a 95% probability and 95% confidence level.

Based on the data summarized in Table 3, the sum of the means can be calculated by:

xT = I xi = ( )+a,b,c It is possible to conservatively determine the value for s for the Shape Annealing Factor Allowance and the Equipment Processing Allowance.

However, the data required to determine the proper tolerance factors to arrive at a 95/95 value for ST are not presently available.

It is concluded that since all five uncertainties are independent, normal, two-sided distributions, with Ks values known at the 95/95 level, it is conservative to determine a combined Ks value using:

KsT = [I (Ksi)2] h The resulting value for KsT is therefore at least a 95% probability, 95%

confidence level value. Using the above noted equation, + KsT =

( *

) +a,b,c. Therefore, the total uncertainty for ASI is

( ) +a,b,c which is less than the + 0.06 ASIU noted in Reference 2.

I REFERENCES

1. Jacobs, G. V., et al. " Basic Safety Report - Millstone Nuclear Power j

Station Unit 2", WCAP-9660 (Proprietary), WCAP-9661 (Non-Proprietary),

February,1980.

l 2. W. G. Counsil letter to R. A. Clark, NRC, " Millstone Unit 2, Measurement Uncertainties," dated March 4,1982.

3. W. G. Counsil letter to R. A. Clark, NRC, " Millstone Unit 2, Additional Information on Measurement Uncertainties," dated February 22,1983.

4. Alsop, B. H., " Basic Safety Report - Millstone Nuclear Power Station Unit 2 - Power Peaking Factor Uncertainty Analysis," WCAP-9660 Addendum I i

(Proprietary), WCAP-9661 Addendum 1 (Non-Proprietary), May,1080.

i I

(

l l

l l

l 1

~

8 e TABLE 1 Calibration Allowance A) ASI Drift Allowance sD = 0.0042 ASIU n = 140 for n = 140, two sided, 95/95 tolerance factor, K = 2.184 Ks0 = 0.00917 ASIU B) Upper detector voltage calibration error: Ksu = 0.00096 ASIU Lower detector voltage calibration error: Kst = 0.0008 ASIU Total Uncertainty, VsCA

• E '

3 KsCA = 0.0093 ASIU For conservatism KsCA is rounded to 0.01 ASIU.

.-g

TABLE 2 -

n

! BREAKDOWN OF INDIVIDUAL ASSEMBLIES ASI RELATIVE TO THE CORE AVERAGE ASI FOR ASI UNCERTAINTY Condition Assembiv n Mean* Variance s

+a,b,c Xe Osc. HFP Cycles 5 and 6 Xe. Osc 90". Pc-er Cycles 5 and 6 i

1 HFP, All Rods Out (AR0)

Cycles 4, 5, and 6 1

3 HFP, Rod Insertion Limit (RIL)

Cycles 4, 5, and 6 I

d 0

-w -re -- -- -

-r

~

l v m

1 .- .

TABLE 2 (Cont)

BREAKCOWN OF INDIVIDUAL AESEMBLIES ASI RELATIVE TO THE CORE AVERAGE ASI FOR ASI UNCERTAINTY

+a,b,c Condition Assembly n Mean* Variance s 90% Pcwer RIL Cycies 4, 5, and 6 Cycle 5 Depleted with CEA7 in 12 Percent, then Cycle 5 ARO I

1 All Data

"% ASI core .% ASI assembly f

l t

l

i

TABLE 3

, 'ss ALLOWANCE AND UNCERTAINTY

SUMMARY

UhCERTAINTY E* s* n v k** + Ks*

4

1. Calibration Allowance 0.0 0.0046 140 139 2.184 0.01 .

2. Equipment Processing Allowance 0.0 0.02 ...

( . .

l 3. Shape Annealing Factor Allowance O' 0

• 4 0.015

4. Incore Monitoring Methodology .- +a;ba c Uncertainty .

j -

{ 5. Separability Allowance ._

5 .

l

• in ASIU .

• • determined from Tolerance Factors for Normal Distributions, CRC Handbook of Tables for Probability and Statistics, 2nd Edition,1976 .

p = 95% . .

A = 95%

l 1 .

i '.-

l l

1 .

l

- i y . ,

FIGURE 1 COMPARISON OF INCA RECONSTRUCTED ASI WITH REAL ASI 60

+a,b,c 50 40 l

S 30 E

W p

5 Y 20 10 0

-1.5 -1.0 .5 0 .5 1.0 1.5

% ASI RECONSTRUCTED - % ASI REAL

o "O" WICE SAFETY RANGE V W X Y A B C D E F GHJ KLMNPR S T I

!l -

1- - - Q Q -

gg 2 & ~

O O O

, ~

43 29

• 42 'N" 0 - 0 0e I

, CONTROL 4 - l . l . 3 - cHA;.NEL

? .

e  !-

O 01 6 l 41 ll 25 i .

C 0 O l 0 l 30 6 - 08 1, 09 25 1

- I i

~ - u :!"- -

'~

l 24 { l } } 07 l j i l l l l l 31]

8-l 40 l O I l j I  :  !

Cl .

l

I 11 l l l r# - l  ! CE l l l l l c, l l 10 - O ; O 3 0 ; Ol O O g

l 22 . Q l0"i

1. WICE ANOE II' *; 3 l , 23 03 10 ; , 12 i

. l } }

l O l j l  ! O 12 - t m

i i ,

i h

i Ol l l j O l l 33 C i  ! ,.

o~,. , i3 l, . ,

,I

.. x,, . = - ,3  ; i ' l . '

CONTROL CHANNEL *n ' O O O O I 15 A

l 16 04 03 , i 14 l l20 O C 17 16 34 U O O 18 3. 3p gn 38 18 17 ,

O Q

19 35 19 "S"WICE RANGE -

O

! O f D- C. 13 36 O l37 .

e n 43 n "A" WIDE RANGE i

l C 41 RHOOtU?.t DETECCRS

'/AN AOtu?.t CETE; 03 "A" SA F ETY 1 ~HER?.tCCOUPLE A Assembly 9, 2

- i O 4 RHCClu?.t OETECO35 BUILDING m

i 1sAcxGRcuNo NORTH 1 HEM.tOCOUPL5 Assembly 8, 2 c XCORE CETECTCR R AO! AL LOCATICNs NOT TO SCALE

^ '

Figure 1 . Inc:re and Excore Detector Core LcC2:icns and Designations D Assembly 8, 4 e

, , _j FIGURE 3 COMPARIS0N OF CORE AVERAGE AND PERIPHERAL ASSEMBLY ASI'S 50

+a b c 45 40

'~ ~~ '

35 -

3 30 E

S ,

E 25 W

p 4, -

w

= 20 15 10 i 5 j

3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 I. 5 - 3.0

% ASI CORE - % ASI ASSEiBLY er * . e .- e, e .

w . we - . . . - = = - = e. ew. m =m-e w .

- umw-we---gyv.. - - - 'e 3-my ._-,-mis -p.-w- - - -

• m p- -W.5 w - ,e p--,m m - -. , - -----em -

.0 i .

i  !

Docket 50-336 l ,.

> i 1

i r i

Appendix A '

The Shape Annealing Factor Component 3

of the Axial Shape Index Uncertainty i at Millstone Project Unit 2 1

1 l

1 i

I l

1 a

i .

i 1

t t

, I l

i i

May,1983 I

. _ _ ~ . _ . . . - . . _ _ _ _ _ _ _ _ _ . _ _ . , . . _ _ , _ . _ _ . - . . ~ _ , . . . . . _ , , _ . ,. _ _ , _ _ _ _ _ , _ . , _ . . _ _ _ _ _ _ _ . . , . . _ _ . _ , _ _